The genotype of the hepatitis C virus (HCV) affects treatment time and sustained virologic response (SVR). It is known that the rate of response to treatment is higher for genotypes-2 and -3 compared to genotype-1, and infection with genotype-1b is related to low response to interferon (IFN) or unresponsiveness. In patients infected with genotype-1, a relation was found between treatment response based on IFN(α) and interleukin (IL) 28B encoding IFN-lambda(λ)-3.¹ Genetic variations have been shown in the IL28B gene region encoding IFN(λ), which is related to spontaneous clearance of HCV.² With these advances, genetic polymorphism in the IL28B region is currently being mentioned in addition to viral load and genotype among factors affecting treatment.³ IL28B gene, localized on the 19th chromosome, encodes IFNλ3. ^4–7 Some polymorphisms in the IL28B gene (rs12979860 and rs8099917) were found to be related to high SVR rates in patients with HCV genotype-1.⁸ It has been shown specifically that the CC genotype appropriate for rs12979860 single nucleotide polymorphism (SNP) increased the possibility of SVR by 2.5-fold in patients with HCV genotype-1.⁹ In patients of Asian origin with genotype-1, who show poor response to treatment, the rates of SVR are higher compared to patients of European origin. This difference is attributed to specific polymorphic changes in the IL28B gene region.^9,10

The rate of response to treatment can be predicted by IL28B genotyping. Thus, decisions related to treatment can be tailored, and the cost can be reduced.¹¹ In this study, the frequency of IL28B polymorphism and its effect on response to treatment were investigated in patients with chronic HCV. Clarification of relations between these polymorphisms and response to treatment may be one of the decisive points in chronic HCV infection treatment.

MATERIAL AND METHOD

The study was conducted by GATA Education Hospital, Infectious Diseases and Medical Genetic Clinics after obtaining the local ethics committee approval (Local ethics committee: September 16, 2010, 1491-1030-10/1539). Data of the patients with chronic HCV who had received or were receiving treatment were reached via the hospital's computer database system and outpatient follow-up archive. The patients were divided into two groups: the group with a response to pegilated (peg)-IFN + Ribavirin (RBV) treatment and those with no response to peg-IFN + RBV treatment.

Study inclusion criteria were as follows:<list list-type="order"> </list>

• Anti-HCV positive, HCV-RNA negative chronic HCV patients who were cured by treatment; and

• Anti-HCV positive, HCV-RNA positive chronic HCV patients with no response to treatment.

Patients were included in the study if they met the condition that treatment had been administered completely and genotype analyses had been done. Following receipt of informed consent from the patients, tests were done in the Genetic Laboratory on the blood samples with EDTA, obtained during routine follow-up visits. After obtaining informed consent from chronic HCV hepatitis patients, DNA isolation was performed by “spin column” method. The DNAs obtained were kept at −20°C until polymerase chain reaction (PCR) was performed.

Analysis of rs12979860 Polymorphism

The “Tetra primer amplification refractory mutation system” method developed by Galmozzi et al.¹² was used. PCR process was performed by complying with mixed components, rates, and PCR programs (primers used in the analysis of rs12979860 polymorphism: rs12-OUTER-FW: 5′-AACTCAACGCCTCTTCCTCCT-3′, rs12-OUTER-RV: 5′ TTCCCATACACCCGTTCCTGT-3′, INNER-FW T: 5′-AGGAGCTCCCCGAAGGAG T-3′, INNER-RV-G: 5′-GTGCCATTCAACCCTGGTACG-3′). A single reaction was performed for analysis of rs12979860 polymorphism. PCR products obtained after PCR were subjected to electrophoresis in 2% agarose gel using a constant flow of 100 V. Ethidium bromide (EtBr) staining method was used. An ultraviolet (UV) transilluminator was used for the imaging process, and pictures of the images were taken.

Analysis of rs8099917 Polymorphism

A rapid and cheap PCR process was developed to determine the polymorphism in IL28B locus. The primers were designed for the reaction based on the “amplification refractory mutation system” (ARMS) method. The primers were designed using the “Web-based allele specific primer designing tool” program.¹³ Two separate reactions were formed for C/T polymorphisms in the PCR process (primers used for analysis of rs8099917 polymorphism: rs80-OUTER-F: 5′-CCCACTTCTGGAACAAATCGTCCC-3′, rs80-G-ALELR: 5′-GCATGGTTCCAATTTGGGTCAC-3′, rs80-T-ALELR: 5′-GCATGGTTCCAATTTGGGTCAA-3′). Two separate reactions were performed for analysis of rs8099917 polymorphism. PCR products obtained after PCR were subjected to electrophoresis in 2% agarose gel using a constant flow of 100 V. EtBr staining method was used. An UV transilluminator was used for the imaging process, and pictures were taken. If amplification was present in both tubes, the patient was considered to be a heterozygote. If amplification was present in C/T reaction tubes, the patient was considered to be a homozygote for that allele.

Statistical Analyses

Data were evaluated using the SPSS 15.0 statistical package software. Statistical significance between response states of the patients and rs8099917 and rs12979860 was evaluated using a χ² test. The relation between genotypes in rs8099917 polymorphism was evaluated using logistic regression analysis. The prediction value of rs8099917 for treatment response was evaluated using ROC curve. A p-value of <.05 was considered to be statistically significant.

RESULTS

Fifty-three chronic HCV patients (25 female, 28 male) who received peg-IFN + RBV treatment in different years were examined. Thirty-seven of the patients (69.8%) responded to treatment and 16 (30.2%) had no response to treatment. Twenty of the patients who responded to treatment (54.1%) were male and 17 (45.9%) were female. Eight of the patients (50%) who did not respond to treatment were male and 8 (50%) were female (p=.786).

The results of liver biopsies, which show the stage of the disease, were not analyzed. Patients were included in the study if they met the condition that treatment had been administered completely. Genotype analyses were performed.

All patients included in the study were found to have genotype-1b. rs12979860 and rs8099917 polymorphisms localized in the IL28B gene region were investigated in the patients. When we evaluated genotype frequency in rs8099917 polymorphism in 53 patients, a major homozygote (TT) genotype was found in 25 of the patients (47.2%), a heterozygote (TG) genotype was found in 20 (37.7%), and a minor homozygote (GG) genotype was found in 8 (15.1%) (Fig. 1). In examining rs12979860 polymorphism in these patients, a major homozygote (TT) genotype was found in 12 patients (22.6%), a heterozygote (CT) genotype was found in 38 (71.7%), and a minor homozygote (CC) genotype was found in 3 (5.7%) (Fig. 2).

When we evaluated allele frequency, in patients with a response in rs8099917, 52 of 74 alleles (70%) were found to be T alleles and 22 (30%) were found to be G alleles. In patients with no response, 18 of 32 alleles (56.25%) were found to be T alleles and 14 (43.75%) were found to be G alleles. In examining rs12979860 in patients with a response, 44 of 74 alleles (59.5%) were found to be T alleles and 30 (40.5%) were found to be C alleles. In patients with no response, 18 (56.25%) were found to be T alleles and 14 (43.75%) were found to be C alleles (Table 1).

In patients with and with no response to treatment, the relation of rs8099917 polymorphism with genotypes was analyzed using a χ² test. The TT genotype was found in 84% of the patients with a response to treatment and in 16% of the patients with no response to treatment. The rate was found to be 50% in patients with and with no response for the TG genotype and 75% in patients with a response and 25% in patients with no response for the GG genotype (p=.045) (Fig. 3).

The analysis of the relation of rs12979860 polymorphism with genotypes in patients with and with no response using a χ² test revealed that the TT genotype was found in 66.7% of the patients with a response to treatment and in 33.3% of the patients with no response to treatment. The CT genotype was found in 71.1% of the patients with a response alnd in 28.9% of the patients with no response. The CC genotype was found in 66.7% of the patients with a response and in 33.3% of the patients with no response (p=.952) (Fig. 4).

When the difference between rs8099917 polymorphism and other genotypes was evaluated using logistic regression analysis, the GG genotype increased response to treatment threefold compared to TG, but the difference was not statistically significant (p=.238). It was found that the TT genotype increased response to treatment 5.25-fold compared to TG. The difference was considered to be statistically significant (p=.019) (Table 2). On ROC curve, the area under the curve value was found to be 0.693.

DISCUSSION

The immune response of the host to HCV infection shows variation.¹⁴. Recently, studies about polymorphism in the IL28B gene have drawn great attention.^15,16 Studies have shown that more than 3.3 billion nucleotides and more than 10 million SNP are present in the human genome.¹⁷

In contrast to some studies that considered gender as one of the factors affecting response to treatment,^2,5 we found that gender difference did not affect response to treatment statistically significantly when we evaluated demographic properties of chronic HCV hepatitis patients with and with no response to treatment who received peg-IFN + RBV treatment (p=.786). This may be explained by the fact that the number of our patients was small and the study was performed in a patient group with a single genotype, which is known to have a low response to treatment.

Independent studies have shown a close relation between IL28B SNP and response to peg-IFN + RBV treatment clearly. This could lead a way to individualized treatment based on genotype in patients with chronic HCV. However, it should be kept in mind that response to treatment is dependent to many factors, including genotype and hepatic fibrosis.²

Kawaoka et al.¹⁸ proved the effect of HCV RNA viral load and the rs8099917 genotype on treatment in the study they performed. They found that rs8099917 and viral load are determinant factors for SVR in genotype-2b patients treated with peg-IFN + RBV and that this was not true for genotype-2a. Moreover, they found that viral load and rs8099917 were determinant factors in genotype-2a patients treated with IFN monotherapy. While the rate of SVR was 42%–46% in genotype-1 patients with the current combination treatment, it was found to be 76%–82% in genotype-2 and -3 patients.19 In the Turkish population, HCV genotype-1b is observed predominantly. Approximately, 69.8% of our patients responded to treatment and 30.2% did not respond to treatment. This result showed that treatment success was above the average. This suggests that genotype-1b does not have such a poor response to treatment as was believed.

Honda et al.²⁰ defined IL28B rs8099917 genetic variation in 91 of 168 chronic hepatitis C patients who received peg-IFN + RBV treatment. They determined IL28B polymorphisms as 66% major homozygote (TT), 30% heterozygote (TG), and 4% minor homozygote (GG). Frequencies of alleles show variation among ethnic groups. The C allele has been reported to be observed frequently in Caucasians and to be related with SVR.⁷ IL28B polymorphism is also strongly related with HCV clearance.¹⁷ When we evaluated the frequency of alleles in our study, we found 52 of 74 alleles (70%) were T and 22 (30%) were G in patients who responded to treatment in rs8099917. In patients who did not respond to treatment, 18 of 32 alleles (56.25%) were found to be T and 14 (43.75%) were found to be G. In rs12979860, 44 of 74 alleles (59.5%) were found to be T, and 30 (40.5%) were found to be C in patients who responded to treatment. In patients who did not respond, 18 (56.25%) were found to be T and 14 (43.75%) were found to be C. This result showed that a high frequency of the T allele, which constitutes the TT genotype and can be named as the “good genotype” especially in rs8099917 polymorphism, is important in terms of response.

In our study, analysis of the relation of rs8099917 polymorphism with genotypes in patients who responded and did not respond to treatment using a χ² test revealed TT at a rate of 84% in patients who responded to treatment and at a rate of 16% in patients who did not respond. When compared with other studies, it was found that the patients who responded to treatment had a higher rate of the TT genotype, especially in rs8099917 polymorphism, and this result was statistically significant (p=.045).

In chronic HCV hepatitis under control by treatment, carrying rs8099917 SNP is a strong genetic determinant. In our study, we found that the TT genotype increased response to treatment 5.25-fold compared to the TG genotype when we evaluated the difference between genotypes in rs8099917 polymorphism using logistic regression, and we found the difference to be statistically significant (p=.019).

In a cohort study performed by Thomas et al.,⁸ in individuals with spontaneous viral clearance or treatment-resistant infection that compared rs12979860 variation in HCV, response to treatment was found to be better in patients with the CC genotype compared to CT and TT genotypes. In a study performed by Mangia et al.,²¹ which they investigated the relation of response to treatment in 268 patients with rs12979860 SNP, the frequency of IL28B genotypes was found to be 48% CT, 37% CC, and 15% TT. While SVR could be achieved in 82% of the patients with the CC genotype, this rate was found to be 75% in patients with the CT genotype and 58% in patients with the TT genotype (p=.0046). A rate of the CC genotype, HCV-infected individuals, was seen to be less than the non-infected.²² In addition, the CC genotype can be said to have a sustained virological response to increase.²¹

Some studies reported that response to treatment had a negative relation with rs12979860 polymorphism and a positive relation with rs8099917 polymorphism.^23,24 In our study, we could not find the CC genotype to be the strongest determinant of SVR before treatment. Our result was found to be statistically insignificant (p=.952). This result may be explained either by the fact that the number of our patients was small or by the differing genetic structure of the Turkish population. In our study, the relation of rs12979860 polymorphism with genotypes in patients who responded and did not respond to treatment was analyzed using a χ² test. In both patient groups, a statistically significant difference was not found in terms of rs12979860 gene polymorphism (p=.952).

In evaluation of the difference between genotypes in rs8099917 polymorphism using logistic regression, the GG genotype increased response to treatment threefold compared to the TG genotype, but the difference was not statistically significant (p=.238). The TT genotype was found to increase response to treatment 5.25-fold compared to the TG genotype and the difference was found to be statistically significant (p=.019). On the ROC curve, the area under the curve value was found to be 0.693. If this rate were above 0.7, a better prediction could be obtained.

An important point to consider is the addition of IL28B evaluations into the algorithm included in the treatment guides. Presence of the TT genotype in rs8099917 polymorphism in patients infected with HCV genotype 1 may lead us to start treatment earlier instead of waiting for the discovery of new drugs. With the possibility of a high success rate, the “good TT genotype” could encourage us to continue antiviral treatment of the patients independent of HCV genotype. Determining the IL28B genotype may also provide shortening of treatment and reduction of resistance with toxic action.

The healthcare systems should support the rational use of new markers instead of additional expenses. We cannot recommend routine IL28B polymorphism tests for the present time, but we think that the test would be appropriate if early viral response is not achieved in the fourth week of treatment.

The genotype basis of patients should be investigated to develop new treatment approaches. Genomic inventions related to hepatitis C will render development of safer, more efficient treatment strategies for clinicians and investigators.

Disclosure: The work we submitted “INVESTIGATION OF INTERLEUKIN (IL) 28B GENE POLYMORPHYSIM IN PATIENTS WITH CHRONIC HEPATITIS C” is original and is not submitted elsewhere for publication. There are no conflict of interest statements

REFERENCES

1. Chung RT. Reaping the early harvest of the genomics revolution. Gastroenterology. 2010;138(5):1653–1654.
2. Rauch A, Kutalik Z, Descombes P, et al. Genetic variation in IL28B is associated with chronic hepatitis C and treatment failure: a genome-wide association study. Gastroenterology. 2010;138(4):1338–1345, 45 e1–45 e7.
3. Chary A, Holodniy M. Recent advances in hepatitis C virus treatment: review of HCV protease inhibitor clinical trials. Rev Recent Clin Trials. 2010;5(3):158–173.
4. Imazeki F, Yokosuka O, Omata M. Impact of IL-28B SNPs on control of hepatitis C virus infection: a genome-wide association study. Expert Rev Anti Infect Ther. 2010;8(5):497–499.
5. Rauch A, Rohrbach J, Bochud PY. The recent breakthroughs in the understanding of host genomics in hepatitis C. Eur J Clin Invest. 2010;40(10):950–959.
6. Grebely J, Petoumenos K, Hellard M, et al. Potential role for interleukin- 28B genotype in treatment decision-making in recent hepatitis C virus infection. Hepatology. 2010;52(4):1216–1224.
7. Zhang L, Jilg N, Shao RX, et al. IL28B inhibits hepatitis C virus replication through the JAK-STAT pathway. J Hepatol. 2011;55(2):289–298.
8. Scherzer TM, Hofer H, Staettermayer AF, et al. Early virologic response and IL28B polymorphisms in patients with chronic hepatitis C genotype 3 treated with peginterferon alfa-2a and ribavirin. J Hepatol. 2011;54(5):866–871.
9. Mangia A. IL28B: a new wager in the skyline of hepatitis C virus infection. Dig Liver Dis. 2011;43(3):177–179.
10. Hsu CS, Hsu SJ, Chen HC, et al. Association of IL28B gene variations with mathematical modeling of viral kinetics in chronic hepatitis C patients with IFN plus ribavirin therapy. Proc Natl Acad Sci U S A. 2011;108(9): 3719–3724.
11. Ahlenstiel G, Booth DR, George J. IL28B in hepatitis C virus infection: translating pharmacogenomics into clinical practice. J Gastroenterol. 2010;45(9):903–910.
12. Galmozzi E, Menico BD, Rametta R, et al. A tetra-primer amplification refractory mutation system polymerase chain reaction for the evaluation of rs12979860 IL28B genotype. J Viral Hepat. 2011;18(9):628–630.
13. http://bioinfo.biotec.or.th/WASP. Accessed [2011].
14. Dill MT, Duong FH, Vogt JE, et al. Interferon-induced gene expression is a stronger predictor of treatment response than IL28B genotype in patients with hepatitis C. Gastroenterology. 2011;140(3):1021–1031.
15. Grimm D, Blum HE, Thimme R. [Genome-wide association studies]. Dtsch Med Wochenschr. 2011;136(3):95–98.
16. Clark PJ, Thompson AJ, McHutchison JG. Genetic variation in IL28B: impact on drug development for chronic hepatitis C infection. Clin Pharmacol Ther. 2010;88(5):708–711.
17. Clark PJ, Thompson AJ, McHutchison JG. IL28B genomic-based treatment paradigms for patients with chronic hepatitis C infection: the future of personalized HCV therapies. Am J Gastroenterol. 2011;106(1):38–45.
18. Kawaoka T, Hayes CN, Ohishi W, et al. Predictive value of the IL28B polymorphism on the effect of interferon therapy in chronic hepatitis C patients with genotypes 2a and 2b. J Hepatol. 2011;54(3):408–414.
19. Akuta N, Suzuki F, Hirakawa M, et al. Amino acid substitutions in the hepatitis C virus core region of genotype 1b affect very early viral dynamics during treatment with telaprevir, peginterferon, and ribavirin. J Med Virol. 2010;82(4):572–582.
20. Honda M, Sakai A, Yamashita T, et al. Hepatic ISG expression is associated with genetic variation in interleukin 28B and the outcome of IFN therapy for chronic hepatitis C. Gastroenterology. 2010;139(2):499–509.
21. Mangia A, Thompson AJ, Santoro R, et al. An IL28B polymorphism determines treatment response of hepatitis C virus genotype 2 or 3 patients who do not achieve a rapid virologic response. Gastroenterology. 2010;139(3):821–827, 827 e1.
22. Sarrazin C, Susser S, Doehring A, et al. Importance of IL28B gene polymorphisms in hepatitis C virus genotype 2 and 3 infected patients. J Hepatol. 2011;54(3):415–421.
23. Suppiah V, Moldovan M, Ahlenstiel G, et al. IL28B is associated with response to chronic hepatitis C interferon-alpha and ribavirin therapy. Nat Genet. 2009;41(10):1100–1104.
24. Tanaka Y, Nishida N, Sugiyama M, et al. Genome-wide association of IL28B with response to pegylated interferon-alpha and ribavirin therapy for chronic hepatitis C. Nat Genet. 2009;41(10):1105–1109.

Autoimmune hepatitis (AIH) is a chronic inflammatory liver disease of unknown aetiology. Its clinical presentation is often non-specific. Biochemical characteristics are elevated transaminases, increased levels of IgG as well as liver-specific and -non-specific autoantibodies (antinuclear antibodies—ANA, smooth muscle antibodies—SMA, liver-kidney antibodies—LKM). ^1–3 There is a female predominance in AIH.⁴ The pathophysiological mechanisms leading to AIH are complex and only partially understood. AIH's association with specific human leucocyte antigen (HLA) alleles (eg, HLA B8, DR3, DRB1 alleles⁵) implicates certain genetic predisposition to autoimmunity.

Diagnosis of AIH¹ is made by exclusion of other liver diseases, elevated levels of ANAs and IgGs and specific histological features, for example, plasma-cell hepatitis. The excellent therapeutic response to immunosuppressive therapy has long been considered as a helpful diagnostic tool.

Liver transplantation (LTx) is the standard therapy for end-stage liver diseases. Long-term survival is excellent with >50% of patients surviving for more than 10 years.⁶ However, long-term survival can be compromised by side-effects of immunosuppression (IS) and recurrence of the disease, notably by hepatitis C virus (HCV) and to some extent by hepatitis B virus (HBV) infections. In autoimmune liver diseases, such as, AIH, primary sclerosing cholangitis (PSC) and primary biliary cirrhosis, recurrence has also been reported with almost no impact on patient's survival.

Primary hyperoxaluria (PHO) is a rare metabolic disorder with autosomal recessive inheritance. This enzymatic defect results in enhanced conversion of glyoxalate to poorly soluble oxalate, which is then excreted in the urine.^{7, 8} The treatment of choice in the case of end-stage renal disease is either kidney transplantation or combined liver-kidney transplantation as the transplanted liver would provide the missing enzyme.

In adults and children with pre-existing autoimmune liver disease, recurrence of AIH has been well described.^{9, 10} However, the occurrence of de novo AIH in adults without pre-existing autoimmunity has been controversially discussed and has been described in only a few cases.^{9, 10} Here, we report a young female patient who received a liver-kidney graft in 1995 for PHO type I and who developed the full blown picture of AIH without any evidence of a liver disease 12 years after transplantation.

CASE REPORT

The 30-year-old female patient was transplanted for PHO type 1¹¹ in June 1995. The combined liver-kidney transplantation (LKTx), to treat end-stage renal disease and to perform gene therapy for the missing enzyme, from a blood-group compatible, local donor was performed. In the pre-transplant workup, she was negative for all virologic tests except cytomegalovirus (CMV). The histology of the explanted liver was reported normal. The post-operative course was uncomplicated and IS was initiated with cyclosporin A (CyA), azathioprin and prednisolone. The dose of Prednisolone was initially tapered to a maintenance dose of 7.5 mg. In December 1995, an Epstein–Barr virus (EBV) hepatitis was diagnosed and the dose of CyA was further reduced to about 100 µg/L. After normalisation of transaminase levels with normal kidney function in early 1996, steroid therapy was discontinued and azathioprin was switched to mycophenolate mofetil (MMF) 2 g/d.

Consequently, the patient underwent follow-up check-ups at least every 6 months at the outpatient clinic. In the following years, the IS regimen remained unchanged. After showing normal liver- and kidney-function tests for 4 years it was decided to taper the MMF dose by 500 mg for every 6 months. Six months after the reduction of IS (CyA + 500 mg MMF) the level of alanine aminotransferase (ALT) increased to 223 U/L.

A swift increase of MMF dose to 1g twice a day, lead to the rapid normalisation of ALT levels. After excluding hepatitis A, D, E, EBV and CMV infections the diagnosis of a mild rejection was carried out on clinical grounds. In the following year, the MMF dose was slowly reduced to 500 mg twice a day. However, from early 2005 to mid 2007 undulating elevation of ALT levels up to 100 U/L (Figure 1) were documented despite increasing the MMF doses and CyA through levels up to 120-160 µg/L; no specific cause could be detected.

In November 2007, the patient presented with an asymptomatic sharp rise in the ALT level of 558 U/L (Figure 1) and with a CyA dosage of 242 mg/L and an MMF dosage of 500 mg twice a day. A complete screen for viral infections based on serological and microbiological studies was negative. The liver biopsy showed a lymphoplasma cellular infiltration and a grade-two fibrosis without any evidence of rejection (Figure 2) compatible with AIH.

A retrospective analysis for the presence of ANA in sera stored at −70°C obtained at every outpatient visit revealed that all these sera collected until August 2004 are negative, but became positive thereafter with a homogeneous nuclear pattern at titers greater than 1:160 (Figure 1).

After having established the diagnosis of de novo AIH, a treatment with prednisolone at a dosage of 50 mg/d and a subsequent tapering of 10 mg/wk was initiated. After 3 weeks of therapy, ALT levels normalised, but increased again when the patient stopped corticosteroid therapy and normalised with a maintenance dose of 5 mg/d.

DISCUSSION

In this article, we present a case of a female patient with acute, histologically proven AIH 12 years after combined LKTx for PHO. This event occurred while she was on stable IS therapy with CyA and MMF. She made a full recovery after administration of corticosteroid therapy. According to the diagnostic criteria of the international AIH group,¹ she fulfils all diagnostic criteria with the exception of IgG being unspecifically elevated in the post-transplant situation. Graft rejection and other causes of “graft-hepatitis” were excluded histologically. The clinical course is remarkable as the diagnosis was established on histology when she presented with an acute hepatitic ALT profile and positivity for serum ANA (titre ≥ 1:160). Interestingly, on retrospective analysis, serum samples tested positive for ANA as early as 3 years prior to the diagnosis correlating with an undulating elevation of transaminases, tested on stored samples.

Disease recurrence is a common phenomenon after LTx and is established for almost all causes except for genetic disorders. The description of recurrence of AIH after LTx has always been met with scepticism because of the persistence of autoimmune features (eg, autoantibodies), the IS therapy, grafting across HLA barriers and the non-specificity of histological features. The occurrence of possible de novo AIH after LTx was first described in 1998 in children.⁹ Kerkra et al⁹ reported on seven children transplanted for inborn liver diseases (biliary atresia and metabolic liver disease), who developed a de novo AIH between 6 and 45 months after transplantation. Interestingly, in four children either elevated perinuclear anti-neutrophil cytoplasmic antibodies (PANCAs),, ANAs, SMAs or atypical LKMs were reported prior to LTx. In 2001, another publication presented seven patients with a so-called de novo autoimmune liver disease again with evidence of a pre-existing autoimmunity.¹⁰ In 2002, in a series reporting on the response to steroid treatment for de novo AIH, one out of 11 patients were transplanted for PSC, and three were tested positive for ANAs prior to LTx.¹² A large matched case-controlled study regarding de novo AIH in children, published in 2007, further emphasised this controversial issue.¹³ It showed that out of 619 paediatric patients, transplanted for reasons other than a primary autoimmune disease, astonishingly 41 patients developed a de novo AIH according to AIH diagnostic criteria. None of these children, however, showed evidence for autoimmunity prior to LTx.¹³

On the other hand, the diagnosis of de novo AIH after LTx is straight forward; if a patient has been transplanted for a non-autoimmune liver disease, but histology fulfils all the required criteria, rejection has been ruled out. Nevertheless, there has been discussion on the possibility of pre-existing “occult” autoimmunity with other causes of chronic hepatitis presenting as de novo AIH after LTx.¹² However, the clinical course of our patient is extraordinary because she had no evidence of autoimmunity or of a liver disease prior to LTx as she was transplanted for a genetic defect, and she presented with the typical picture of AIH 12 years after LTx responding to corticosteroid therapy.

In adults, however, only a very few cases of “true” de novo AIH have been reported.^{12, 14} A male patient grafted for malignant alcoholic cirrhosis¹⁴ developed serologic and histologic de novo AIH 18 months after LTx while on CyA and MMF treatment with a good response to corticosteroid therapy. The diagnosis of de novo AIH in our patient is even more clear cut; female patient, transplanted for oxalosis without liver diseases and occurrence of all diagnostic features 12 years after Ltx with a good response to corticoid treatment.

The reason for the development of AIH after LTx remains unclear because so far no risk factors have been identified. In theory, the same scenario of viral triggers could be contemplated as in the non-LTx situation¹⁵ leading to an autoimmune reaction through polyclonal stimulation, enhancement and induction of membrane expression of MHC class I and II antigens, interference with immunoregulatory cells and/or with idiotype–anti-idiotype network.⁴ It is noteworthy to mention that our patient had an episode of EBV infection many years prior to the diagnosis of AIH. Certain HLA haplotypes, such as HLA-DR3, are considered to predispose to AIH and may play an important role in the initiation of an allo-/auto-immune response.¹⁶ However, our patient has no typical HLA alleles,⁵ providing susceptibility to autoimmune liver disease. Sex mismatch and ABO blood type mismatch are not associated with de novo AIH.¹⁷

In children who developed de novo AIH, antibodies directed to glutathione S-transferase T1 have been found, suggesting an immune reaction against this protein synthesised by the donor liver.¹⁸ The author of this study also proposed that an antibody-mediated immune response represents a foreign antigen-driven anti-graft reaction–an “alloimmune” response–rather than an autoimmune response.

In summary, this case report demonstrates that the development of a true de novo AIH can occur many years after LTx even with a stable maintenance of IS. In contrast to the paediatric patients, de novo AIH represents a rare event in adults and the diagnosis can be missed. Corticosteroid therapy is necessary to induce control of inflammation, while CNI and MMF treatment might be useful to maintain remission. De novo AIH has to be considered as a cause for graft dysfunction even years after transplantation in patients treated for a non-autoimmune liver disease.

Disclosure: The authors declare no conflict of interest.

REFERENCES

1. Manns MP, Czaja AJ, Gorham JD, et al. Diagnosis and management of auto-immune hepatitis. Hepatology. 2010;51(6):2193–2213.
2. Czaja AJ. Diverse manifestations and evolving treatments of autoimmune hepatitis. Minerva Gastroenterol Dietol. 2005;51(4):313–333.
3. Zachou K, Eirini R, Dalekos GN. Autoantibodies and autoantigens in autoimmune hepatitis: important tools in clinical practice and to study pathogenesis of the disease. J Autoimmune Dis. 2004;1(1):2.
4. Mieli-Vergani G, Vergani D. De novo autoimmune hepatitis after liver transplantation. J Hepatol. 2004;40(1):3–7.
5. Czaja AJ, Manns MP. Advances in the diagnosis, pathogenesis, and management of autoimmune hepatitis. Gastroenterology. 2010;139(1): 58–72.e4.
6. European Liver Transplant Registry. http://www.eltr.org 7. Danpure CJ. Molecular and clinical heterogenity of primary hyperoxaluria type1. Am J Kidney Dis. 1991;17(4):366–369.
8. Cochat P. Primary hyperoxaluria type1. Kidney Int. 1999;55(6):2533–2547.
9. Kerkar N, Hadzić N, Davies ET, et al. De-novo autoimmune hepatitis after liver transplantation. Lancet. 1998;351(9100):409–413.
10. Heneghan MA, Portmann BC, Norris SM, et al. Graft dysfunction mimicking autoimmune hepatitis following liver transplantation in adults. Hepatology. 2001;34:464–470.
11. Lhotta K, Rumsby G, Vogel W, et al. Primary hyperoxaluria type 1 caused by peroxisome-to-mitochondrion mistargeting of alanine: glyoxylate aminotransferase. Nephrol Dial Transplant. 1996;11(11):2296–2298.
12. Salcedo M, Vaquero J, Ban˜ares R, et al. Response to steroids in de novo autoimmune hepatitis after liver transplantation. Hepatology. 2002;35(2): 349–356.
13. Venick RS, McDiarmid SV, Farmer DG, et al. Rejection and steroid dependence: unique risk factors in the development of pediatric posttransplant de novo autoimmune hepatitis. Am J Transplant. 2007;7(4): 955–963.
14. Di Cocco P, Barletta A, Clemente K, et al. De novo autoimmune hepatitis following liver transplantation: a case report. Transplant Proc. 2008;40(6): 2073–2074.
15. Czaja AJ. Autoimmune hepatitis after liver transplantation and other lessons of self-intolerance. Liver Transpl. 2002;8:505–513.
16. Gonzales-Koch A, Czaja AJ, Carpenter HA, et al. Recurrent autoimmune hepatitis after orthotopic liver transplantation. Liver Transpl. 2001;7: 302–310.
17. Miyagawa-Hayashino A, Haga H, Egawa H, et al. Outcome and risk factors of de novo autoimmune hepatitis in living-donor liver transplantation. Transplantation. 2004;78:128–135.
18. Aguilera I, Sousa JM, Gavilan F, Bernardos A, Wichmann I, Nunez- Roland A. Glutathione S-tranferase T1 mismatch constitutes a risk factor for de novo immune hepatitis after liver transplantation. Liver Transpl. 2004;10:1166–1172.

Vasoactive intestinal polypeptide-oma (VIPoma) is an extremely rare neuroendocrine tumour (NET). In a recent nationwide epidemiological study from Sweden, the incidence of NET was estimated to be approaching three per 100 000 individuals per year, with a slight predominance in women. These tumours have a strong predilection for the gastrointestinal tract. VIPomas affect the pancreas in 90% of the cases. VIPoma, also called as WDHA (watery diarrhoea, hypokalemia and achlorhydria) syndrome, which was described by Verner and Morrison in 1958,¹ presents with profuse watery diarrhoea and serious electrolyte disturbances, especially, hypokalemia, and dehydration. Localised tumour can be cured completely by surgical resection. It needs an expert and patient assessment of the case along with a high index of suspicion for an early diagnosis, which, in turn, improves the prognosis, as 40–70% of the cases may have metastasis at the time of diagnosis or surgery.

CASE REPORT

A 39-year-old Indian woman admitted to the emergency ward with an acute episode of diarrhoea, four to five times a day for the past one year; not associated with vomiting, fever or abdominal pain. Stools were typically of large volume (>2 L/day), watery, not associated with blood or mucus, or steatorrhoea, not related to meals, with no nocturnal episodes of diarrhoea. The diarrhoea persisted during periods of religious fasting. It was not related to ingestion of milk/milk products. She gave no history of major weight loss, and her appetite was normal. She was being treated at an outside clinic but was referred to us in view of worsening diarrhoea and development of acute, purely motor, flaccid quadriparesis with areflexia, due to severe hypokalemia. On deeper questioning, she also reported of facial flushing intermittently. She was reported to have diabetes and was put on oral hypoglycemic agents a month ago. Her serum potassium (S.K⁺) on admission was 1.9 mEq/L, which was corrected with intravenous and oral potassium supplementation. She had metabolic alkalosis with the serum bicarbonate being 30.5 mmol/L. She was managed with careful correction of hyperglycemia with insulin as it could worsen hypokalemia further. The patient's hydration status, muscle power and ECG changes of hypokalemia gradually improved. She was up and ambulant by the next 24 hours, and her concurrent S.K⁺ was 3–3.7 mEq/L by day 2. Clinical suspicion of WDHA syndrome was considered in view of chronic watery diarrhoea, with severe syptomatic hypokalemia, facial flushing and hyperglycemia. As we expected, an abdominal ultrasound showed a 3×2.9 cm mass (Figure 1) in the uncinate process of the pancreas, and a contrast-enhanced computer tomography of the abdomen confirmed the ill-defined lobulated isodense mass lesion, 3.5×3.2 cm in the uncinate process of the pancreas (Figures 2 and 3), close to the third part of the duodenum, compressing the inferior vena cava. With the head, body and tail of the pancreas being normal, the main pancreatic duct was mildly prominent at head. There was no evidence of liver metastasis. Fasting serum vasoactive intestinal polypeptide (S.VIP), a diagnostic test, was performed. The S.VIP done by immunoassay method is reported as 173 pg/ml (range: 20–42 pg/ml). The collection of blood sample for the test was done when the patient did not have a severe bout of diarrhoea, which could possibly explain the relatively lower value in this case. Other tests for evaluation of chronic diarrhoea, namely, a thyriod function test, serum calcium, HIV ELISA, were normal. A proctoscopy was negative for rectal villous adenoma. Her chest x-ray was normal. As a part of Multiple endocrine neoplasia type 1 (MEN-1) work up, serum parathormone and serum prolactin were sent and the results awaited. The patient still continues to have diarrhoea almost daily. Now, the patient is awaiting surgery. As it is a localised lesion without metastasis, the prognosis is good. However, approach may not be easy as the mass is located in the uncinate process, posteriorly.

DISCUSSION

VIPoma is a rare NET.¹ It has an incidence of 0.05–0.2 per million adults in the United States,² with a female preponderence.¹ In a study conducted in Japan published in 1998, 241 cases of VIPoma reported in international literature were analysed.⁴ It occurs in less than 10% of the pancreatic islet cell tumours.¹ A more recent Chinese study reported 31 cases of VIPoma in Chinese literature from 1977 to 2002. The acronym WDHA (Watery Diarrhoea, Hypokalemia and Achlorhydria) syndrome was proposed in 1967.⁴ The mean age is 42–51 years.^{2, 4–7} It is usually located in the pancreas (90%),^{4, 6, 8} with 42%–75% located in the tail,^{4, 7} although, in our patient it was located in the uncinate process. The symptomatic tumours are >3 cm in size and usually solitory tumours,^{2, 5 7–10} and sporadic. This may be inherited as a part of MEN-1 syndrome. In a report of 580 cases of MEN-1, 2% were associated with a VIPoma, which highlights the importance of basic screening for parathyroid and pituitary adenomas in the patient and, if strongly suspected, in their family members as well. In patients, the cardinal features of large-volume (>1 L/day) secretory diarrhoea (89%–100%)^{5, 11}; hypokalemia (67%–100%); dehydration (44%–100%); weight loss (36%–100%); and flushing (14%–33%) are thought to be due to the vasodilatory action of VIP.⁸ Our patient had all these features except weight loss. The VIP secreted by the tumours stimulate 3′,5′-cyclic adenosine monophosphate (cAMP) production by the intestines, resulting in profuse watery diarrhoea leading to dehydration and symptomatic hypokalemia.¹² Typical laboratory findings include hypokalemia, hypochlorhydria, hypercalcemia and hyperglycemia.^{2, 4, 8, 13} Our patient had hyperglycemia, which could be attributed to the glycogenolytic effect of VIP in the liver.¹⁴

VIPoma is diagnosed by an elevated fasting S.VIP levels^{3, 6, 8, 15} and a history of large-volume (>3 L/day) secretory diarrhoea. Stool volume <700 g/day rules out the diagnosis of VIPoma. A normal S.VIP level is between 0 and 190 pg/ml.^{2, 11} The mean S.VIP among patients with VIPoma in two large studies were 956 and 675 pg/ml, respectively.^{2, 11} Ultrasound, computer tomography and somatostatin receptor scintigraphy are the imaging modalities used to localise the tumour. In case of very small tumours, exploratory laparotomy with intra-operative ultrasound is very sensitive. It is important to know that 29%–78% of cases have metastatised at diagnosis to the lymph nodes, liver, lung or bone.^{2, 3, 5–8, 13, 16, 17} Symptoms can be controlled medically, by using long-acting somatostatin analogues, like octreotide and lanreotide.^{2, 6, 9 18–21} The newer somatostatin analogue, pasireotide or SOM-230, has a more prolonged half-life and higher affinity for somatostatin receptors, which may be considered in the future. In unresponsive cases, glucocorticoids added with octreotide may be effective in some cases.¹⁸ Surgical resection of localised tumour is the definitive treatment with complete cure in 30%–33% of cases,^{2, 16} as is possible in our patient. In advanced metastatic disease de-bulking surgery with Radiofrequency ablation (RFA) and cryotherapy is used for the managed distant metastasis.^{5, 8} Long-term octreotide and chemotherapy (5-FU + streptozotocin, interferon-α) decrease medical cost by 50%.²² Finally, early diagnosis and management is very essential in suspected patients who present with chronic diarrhoea, as early resection of the tumour may be curative for the patient.

Disclosure: The authors declare no conflict of interest.

REFERENCES

1. Levi F, Te VC. Epidemiology of carcinoid neoplasms in Vaud, Switzerland, 1974–97. Br J Cancer. 2000;83:952–955.
2. Long RG, Byrant MG, Mitchell SJ, et al. Clinicopathological study of pancreatic and ganglioneuroblastoma tumors secreting vasoactive intestinal polypeptide. Br Med J (Clin Res Ed). 1981;282:1767–1771.
3. Jensen RT. Carcinoid and pancreatic endocrine tumors: recent advances in molecular pathogenesis, localization, and treatment. Curr Opin Oncol. 2000;12:368–377.
4. Soga J, Yakuwa Y. Vipoma/diarrheogenic syndrome: a statistical evaluation of 241 reported cases. J Exp Clin Cancer Res. 1998;17:389–400.
5. Peng SY, Li JT, Liu YB, et al. Diagnosis and treatment of VIPOMA in China: (case report and 31 cases review) diagnosis and treatment of VIPOMA. Pancreas. 2004;28:93–97.
6. Verner JV, Morrison AB. Non-B islet tumors and syndrome of watery diarrhea, hypokalemia and hypochlorhydria. Clin Gastroenterol. 1974;3:595.
7. Verner JV, Morrison AB. Endocrine pancreatic islet disease with diarrhea. Report of a case due to diffuse hyperplasia of the non-beta islet tissue with a review of 54 additional cases. Arch Intern Med. 1974;133:492–499.
8. Matuchansky C, Rambaud JC Vipomas and endocrine cholera: clinical presentation, diagnosis and advances in management. In: Mignon M, Jensen RT, eds. Endocrine Tumors of the Pancreas: Recent Advances in Research and Management. Series: Frontiers in Gastrointestinal Research. Basel, Switzerland: S Karger; 1995:166.
9. Jensen RT. Overview of chronic diarrhea caused by functional neuroendocrine neoplasms. Semin Gastrointest Dis. 1999;10:156–172.
10. Smith SL, Branton SA, Avino AJ, et al. Vasoactive intestinal polypeptide secreting islet cell tumors: a 15-yr experience and review of the literature. Surgery. 1998;124:1050–1055.
11. Jensen RT, Norton JA Pancreatic endocrine tumors. In: Feldman M, Friedman LS, Sleisenger MH. Sleisenger and Fordtran’s Gastrointestinal and Liver Disease: Pathophysiology/Diagnosis/Management. 8th ed. 643–666.
12. Adam N, Lim SS, Ananda V, Chan SP. VIPoma syndrome: challenges in management. Singapore Med J. 2010;51(7):e129.
13. Verner JV, Morrison AB. Islet cell tumor and a syndrome of refractory watery diarrhea and hypokalemia. Am J Med. 1958;25:374–380.
14. Jenson RT, Norton JT. Endocrine tumors of the pancreas. In: Feldman M, Scharschmidt BF, Sleisenger MH, eds. Gastrointestinal and Liver Disease. 6th ed. Philadelphia, PA: WB Saunders; 1998:871.
15. Kroneberg HM, Melmed S, Polonsky KS, Larsen PR. Williams Textbook of Endocrinology. 11th ed:1691.
16. Capella C, Polak JM, Buffa R, et al. Morphological patterns and diagnostic criteria of VIP–producing endocrine tumors. A histologic, histochemical, ultrastructural, and biochemical study of 32 cases. Cancer. 1983;52:1860.
17. Bloom SR, Chritofides ND, Demarter J, et al. Diarrhea in vipoma patients associated with cosecretion of a second active peptide (peptide histidine isoleucine) explained by single coding gene. Lancet. 1983;2:1163–1165.
18. Manton PN, Garner JD, Jensen RT. Use of the long-acting somatostatin analog, SMS 201-995 in patients with pancreatic islet-cell tumors. Dig Dis Sci. 1989;34(suppl 28):28S–39S.
19. Manton PN, O’Dorisio TM, Howe BA, et al. Effect of a long-acting somatostatin analogue in a patient with pancreatic cholera. N Engl J Med. 1985;312:17–21.
20. Schonfeld WH, Elkin EP, Woltering EA, et al. The cost-effectiveness of octreotide acetate in the treatment of carcinoid syndrome and vipoma. Int J Tech Assess Health Care. 1996;14:514–525.
21. Jensen RT. Peptide therapy. Recent advances in the use of somatostatin analogues and other peptide receptor agonists and antagonists. In: Lewis JH, Dubois A, eds. . Current Clinical Topics in Gastrointestinal Pharmocology. Malden, MA: Blackwell Science;1997:144.
22. Manton PN, O’Dorisio T, Malarkey WB, et al. Successful therapy of pancreatic cholera with long acting somatostatin analogue SMS 201-995: relation between plasma concentrations of drug and clinical and biochemical responses. Scand J Gastroenteol. 1986;21(suppl 119):181.

Diagnosis of chronic pancreatitis is classically based on the demonstration of either morphological and/or functional changes that typically develop over time in the course of the disease. Exocrine pancreatic secretion impairs progressively as chronic pancreatitis develops, and a reduced pancreatic function may be seen as a sign of the disease. However, exocrine pancreatic function is frequently disturbed not only in chronic pancreatitis but also in other diseases of the exocrine and endocrine pancreas (cystic fibrosis, pancreatic tumors, after acute necrotizing pancreatitis, and diabetes mellitus), as well as extrapancreatic diseases like celiac disease, Crohn's disease, and Zollinger-Ellison Syndrome. In addition, a secondary exocrine pancreatic insufficiency develops frequently after gastrointestinal and pancreatic surgery (partial or total gastrectomy, duodenopancreatectomy).

The role of the functional evaluation of the exocrine pancreas for the diagnosis of chronic pancreatitis in clinical practice is currently limited. This is mainly due to the development and optimization of imagine techniques, as magnetic resonance imaging, secretin-stimulated magnetic resonance cholangiopancreatography (s-MRCP), and endoscopic ultrasonography (EUS), which are highly sensitive for detecting morphological changes of the gland at early stages of the disease. In this new scenario, functional evaluation of the exocrine pancreas could still be important to support the diagnosis of chronic pancreatitis in cases of inconclusive morphological findings on imaging methods. However, it must be accepted that the most relevant role of the functional evaluation of the pancreas today is the detection of primary or secondary pancreatic insufficiency in patients with known pancreatic disease or after gastrointestinal and pancreatic surgery to aid in the indication of enzyme substitution therapy and to control the efficacy of this therapy.

Exocrine pancreatic secretion may be evaluated by means of direct and indirect methods. Direct methods are those based on the quantification of stimulated pancreatic secretion of volume, bicarbonate, and/or enzymes. Direct methods are invasive, since duodenal intubation and duodenal juice sampling are required. Indirect methods are those evaluating either the digestive ability of the exocrine pancreas on a meal, or pancreatic secretion by quantifying pancreatic enzymes in stool samples. Most indirect methods are noninvasive, but some of them require blood sampling and should be then considered as invasive tests. The clinical usefulness of each one of the available methods is related to diagnostic accuracy, applicability to clinical routine and cost. Direct pancreatic function tests, mainly the secretin-pancreozymin test, are the gold standard for the evaluation of exocrine pancreatic secretion.¹ However, these tests are invasive, cumbersome, time-consuming, expensive, and thus limited to some specialized centers. Indirect pancreatic function tests are more easily applicable to clinical routine and therefore more widely used. Among these are oral and breath tests that, together with the fecal fat quantification, evaluate the digestive ability of the exocrine pancreas, and fecal tests that measure the activity or concentration of pancreatic enzymes in feces. The sensitivity and specificity of these indirect tests for the diagnosis of chronic pancreatitis are variable and lower than those of direct tests. Since the information provided by each test is different, it is important to select the optimal test to be performed in each clinical situation.

In patients with clinical suspicion of chronic pancreatitis but normal or inconclusive imaging, only the secretin-pancreozymin test and the endoscopic pancreatic function test can be sensitive enough to support the diagnosis of the disease. In the opposite site, the diagnosis of primary or secondary exocrine pancreatic insufficiency and, in this context, the indication or control of the efficacy of enzyme substitution therapy, require a test able to detect maldigestion. Between both situations, tests with an intermediate sensitivity (eg, fecal elastase test) may be useful for the screening of chronic pancreatitis in patients with compatible clinical picture and the long-term follow-up of patients with known chronic pancreatitis.

SECRETIN-PANCREOZYMIN TEST

Secretin-pancreozymin test is based on the direct measurement of pancreatic enzymes and bicarbonate output in samples of duodenal juice obtained after stimulation of the gland with intravenous administration of secretin and CCK or caerulein.¹ The endogenous stimulation by a test meal (Lundh test) is no longer used outside clinical research because of a lower diagnostic accuracy, although it is the most physiological direct test.

The secretin-pancreozymin test protocol differs among centers. A double lumen naso-duodenal tube should be placed for constant aspiration of gastric juice and complete and fractionated collection of duodenal juice on ice during intravenous continuous infusion of secretin and CCK or caerulein.^{1, 2} Despite duodenal juice being continuously aspirated, collection may be incomplete. The amount of juice that is lost toward the jejunum may be calculated by constant duodenal perfusion of a dilution nonabsorbable marker, usually polyethilenglycol.³ This requires a triple lumen tube and further complicates the performance of the test.

The secretin-pancreozymin test allows the classification of exocrine pancreatic dysfunction in different degrees (Table 1). Sensitivity and specificity of this test for the diagnosis of chronic pancreatitis both exceed 90% (Table 2).⁴

Since direct pancreatic function tests are invasive, cumbersome, time-consuming, and expensive and since the development of novel sensitive imaging methods (ie, endoscopic ultrasonography) has markedly improved the diagnosis of chronic pancreatitis, the usefulness of the secretin-pancreozymin test is currently mainly limited to the validation of new pancreatic function tests.

ENDOSCOPIC PANCREATIC FUNCTION TEST

The endoscopic pancreatic function test has been developed in order to avoid the problems of the secretin-pancreozymin test related to intubation, duration, and clinical applicability. It is based on the measurement of bicarbonate concentration and/or pancreatic enzymes activity in samples of duodenal juice obtained during upper gastrointestinal endoscopy after intravenous secretin or cholecystokinin stimulation.^{5, 6}

The protocol of the endoscopic pancreatic function test is based on the following four steps^{5, 6}: (1) standard endoscopy to the descending duodenum with the patient under conscious sedation; (2) intravenous administration of secretin (1 U/kg or 0.2 µg/kg/h) or CCK (40 ng/kg/h); (3) endoscopic duodenal fluid collection at 0, 15, 30, 45, and 60 minutes after secretin injection; and (4) fluid analysis for bicarbonate concentration and/or pancreatic enzymes activity.

Peak of bicarbonate concentration over 60 minutes is lower in patients with advanced chronic pancreatitis than in those with abdominal pain of extrapancreatic origin.⁷ Measurement of lipolytic activity in duodenal juice collected for 10 minutes after intravenous CCK is also significantly lower in patients with chronic pancreatitis compared with patients with normal pancreas, but it is not accurate enough for routine clinical use.⁸

Although the endoscopic pancreatic function test is a promising procedure, it is far from being the current standard. Whether the peak bicarbonate concentration, instead of the output over time, is a reliable marker of the exocrine pancreatic function is questionable. In fact, most experts in this field support bicarbonate and enzyme output and not concentration as the most reliable marker of exocrine pancreatic function. This is due to the inverse relationship between bicarbonate concentration and rate of juice secretion in response to secretin. The endoscopic pancreatic function test, in addition, requires the endoscope to be maintained in the duodenum for 1 hour, which is at least as uncomfortable for patients as naso-duodenal tubing. Nevertheless, it must be underlined that the endoscopic pancreatic function test produces complementary information to endoscopic ultrasound for the evaluation of chronic pancreatitis.⁹ In fact, combining the endoscopic pancreatic function test with endoscopic ultrasound allows improving the diagnostic sensitivity for detection of early fibrosis in the context of chronic pancreatitis.¹⁰

S-MRCP

Together with the tests mentioned above, s-MRCP allows for a quantitative measurement of pancreatic secretion volume and should be thus considered as a tubeless direct test. The major advantage of s-MRCP is that of providing information about both ductal changes and function. Monitoring of the pancreatic flow dynamics after intravenous secretin administration and measurement of the subsequent filling of the duodenum during s-MRCP can therefore be used to evaluate the exocrine pancreatic function. The duodenal filling is significantly reduced in patients with exocrine pancreatic insufficiency as compared with healthy subjects. It has been shown that the reduced duodenal filling correlates with the reduced exocrine pancreatic function measured by direct intubation tests.^{11, 12} In addition, s-MRCP appears to be sufficiently sensitive to identify patients with a mild exocrine pancreatic insufficiency. The advantage of s-MRCP over other direct function tests is that it is a noninvasive technique, except for the need of infusing intravenous secretin, which provides both functional and morphological dynamic information.

INDIRECT TESTS

Indirect tests evaluate exocrine pancreatic function by quantifying either the digestive ability of the gland or the levels of pancreatic enzymes in feces.¹³ From a methodological point of view, these tests can be thus classified as oral tests and fecal tests. In oral tests, a substrate is orally given together with a test meal. Pancreatic enzymes hydrolyze the substrate within the duodenum, and the released metabolites are absorbed from the gut and can be measured either in serum, urine, or breath. Among oral tests, the ¹³C-mixed triglyceride breath test is to be underlined. Other tests like the pancreolauryl test, the NBT-PABA test, and the amino acid consumption test¹⁴ are no longer commercially available and/or have a too low diagnostic accuracy to be recommended for clinical use.

Several extrapancreatic factors are known to limit the accuracy of oral pancreatic function tests, mainly those interfering with a normal digestion (slow gastric emptying rate, decreased bile acid secretion) and intestinal absorption (intestinal diseases) as well as those affecting the elimination of the digestion products (renal insufficiency). Variability in gastric emptying rate can be avoided to some extent by administration of metoclopramide or any other prokineticum in the context of the test. The potential negative role of renal disturbances is avoided by the quantification of the digestion products in serum or breath instead of urine.

Fecal tests are based on the quantification of pancreatic enzyme concentration (elastase) or activity (chymotrypsin) in feces. Enzymes are deactivated and diluted or concentrated in a variable degree during the intestinal passage, which must be taken into account for an adequate interpretation of the test results. Exocrine pancreatic function can also be measured indirectly in feces by means of the fecal fat quantification. The amount of fat eliminated within the feces reflects indirectly the degree of fat digestion and, therefore, the pancreatic lipase secretion.

FECAL ELASTASE CONCENTRATION

Pancreatic elastase is highly stable along the GI transit and the fecal concentration of this enzyme significantly correlates with the amount of enzyme secreted by the exocrine pancreas. ^15–17 Furthermore, since the methodology used to quantify this enzyme is based on human specific monoclonal antibodies, the oral enzyme substitution therapy does not interfere with the test. Therefore, interruption of this therapy previous to stool collection is not needed, which is an important advantage of the test.

Quantification of fecal elastase is performed in a single small stool sample by a specific enzymimmunoassay. A fecal elastase concentration higher than 200 µg/g is considered as normal. Concentrations lower than 50 µg/g are related to exocrine pancreatic insufficiency. Although fecal elastase quantification is not sensitive enough to detect patients with mild chronic pancreatitis, its sensitivity in cases of moderate to severe disease is very high, reaching values close to 100%. ^15–17 The specificity of fecal elastase is also high, only limited by dilution in cases of watery diarrhea.

Fecal elastase based on the use of human specific monoclonal antibodies is therefore a very adequate test for the diagnosis of exocrine pancreatic dysfunction in the context of chronic pancreatitis. Since this test is easy to apply to the clinical routine, it may be used as a first step test for the study of patients with clinically suspected chronic pancreatic disease and for the follow-up of patients with known chronic pancreatitis. In situations of secondary exocrine pancreatic dysfunction (ie, after gastrointestinal surgery), fecal elastase allows evaluating pancreatic secretion but not detecting maldigestion.

FECAL CHYMOTRYPSIN ACTIVITY

Quantification of fecal chymotrypsin is a simple test, easy to apply to the clinical routine. This test is based on the enzymatic quantification of chymotrypsin activity in an isolated small stool sample. Because of that, fecal chymotrypsin has been widely introduced in clinical routine as an exocrine pancreatic function test. Chymotrypsin is however variably inactivated during the intestinal passage in such a way that fecal chymotrypsin activity does not accurately reflect pancreatic secretion of the enzyme.¹⁵ In addition, dilution in patients with diarrhea of any etiology will also decrease the fecal activity of the enzyme.¹⁵

Because of that and in order to keep an adequate specificity of the test, a low cut-off (3 U/g of stool) is generally accepted to define a test as abnormal. Patients with fecal chymotrypsin activity less than 3 U/g of stool are thus considered as suffering from exocrine pancreatic dysfunction, but the sensitivity obtained with the test is too low to recommend it for clinical practice. In fact, the test is not able to detect a single case of mild chronic pancreatitis and only slightly more than half of the patients with moderate or severe disease (Table 2).¹⁵

Last but not least, orally administered exogenous pancreatic enzymes as a treatment of exocrine pancreatic insufficiency interact with the determination of chymotrypsin in stool and thus this therapy should be interrupted for at least the 48 hours preceding the stool sample collection. This is not always easy to accomplish for patients with exocrine pancreatic insufficiency.

In conclusion and after taking into consideration all aspects mentioned above, fecal chymotrypsin quantification should no longer be considered as adequate for evaluating exocrine pancreatic function in clinical routine. Nevertheless, quantification of chymotrypsin in feces is a simple way to evaluate patient's compliance to enzyme substitution therapy.¹⁸ In this context, fecal chymotrypsin activity should increase significantly if oral enzymes are correctly taken.

Fecal fat quantification (coefficient of fat absorption)

Quantification of the coefficient of fat absorption (CFA) is still considered as the gold standard for the diagnosis of exocrine pancreatic insufficiency with fat maldigestion. Despite that, this test has several and important disadvantages limiting its clinical applicability. Patients must keep on a standard diet containing 80–120 g of fat daily during 5 consecutive days. This is an important handicap since the majority of patients with chronic pancreatitis are alcoholics and thus with a limited compliance. Furthermore, patients should collect the whole amount of feces produced over the last 3 days of diet. Again this is not easy to comply for many patients. A 3-day collection is needed to reduce errors and variability that may occur if a shorter collection period is used.

Not only patient compliance is a limitation for the fecal fat quantification but also the handling of stool samples in the lab. Stool samples collected over 3 days must be first homogenized and then processed following a manual method that makes this test unpleasant and cumbersome. A methodology based on near infrared reflectance analysis (NIRA) has simplified the quantification of fat in stool and thus could make feasible the application of this test in clinical routine.¹⁹ Nevertheless, difficulties associated to patients’ compliance remain the same.

Interpretation of the test requires keeping a record of all dietary intakes over the 5-day period to determine fat intake and thus to calculate the fractional fat absorption.²⁰ It should be noticed that fecal fat quantification is a nonspecific pancreatic function test since any other cause of maldigestion (ie, obstructive jaundice) or malabsorption (ie, sprue, Crohn's disease) may also lead to abnormal fecal fat excretion.

¹³C-MIXED TRIGLYCERIDE BREATH TEST

Several substrates, mainly ¹³C-labeled, have been used to evaluate exocrine pancreatic function by means of breath tests. ^21–24 In these tests, the labeled substrate is given orally together with a test meal. After intraduodenal hydrolysis of the substrate by specific pancreatic enzymes, ¹³C-marked metabolites are released, absorbed from the gut, and metabolized within the liver. As a consequence of the hepatic metabolism ¹³CO₂ is released and eliminated with the expired air. The amount of ¹³CO₂ expired, which indirectly reflects the exocrine pancreatic function, can be measured by means of mass spectrometry or infrared analysis.

Most substrates used in breath tests, among them the mixed ¹³C-triglyceride, the cholesteryl ¹³C-octanoate, the ¹³C-hiolein, and the ¹³C-triolein, are hydrolyzed by pancreatic lipases. ^21–24 In this way pancreatic function breath tests should be seen as fat digestion tests and thus be considered as an alternative to fecal fat quantification.

The mixed ¹³C-triglyceride (¹³C-MTG) breath test is the only breath test that has been optimized so far. In our experience, this is the optimal substrate for the diagnosis of fat maldigestion and thus the ¹³C-MTG-breath test has been developed as a simple alternative to the fecal fat quantification.²⁴ An abnormally low 6h-recovery of ¹³CO₂ indicates the presence of fat maldigestion with a sensitivity and specificity higher than 90%.²⁴ The test is also highly accurate for the diagnosis of maldigestion in clinical situations of secondary exocrine pancreatic insufficiency such as partial or total gastrectomy or duodenopancreatectomy.

The ¹³C-MTG-breath test is a simple, noninvasive, and accurate method for the diagnosis of exocrine pancreatic insufficiency. This test is easily applicable to the clinical routine and can be repeated as frequently as needed. In this way, utility of the test is not only limited to the diagnosis of exocrine pancreatic insufficiency but can also be extended to the control of the efficacy of oral enzyme substitution therapy in these patients. Therefore, the ¹³C-MTG breath test may play a relevant role in the management of patients with maldigestion secondary to chronic pancreatitis, cystic fibrosis, pancreas cancer, after acute necrotizing pancreatitis, or after gastric or duodenal surgery.

CONCLUSIONS

A wide variety of tests are available for the evaluation of exocrine pancreatic function that have been classically used for the diagnosis of chronic pancreatitis. However, functional testing for the diagnosis of chronic pancreatitis plays a minor role today. Only the secretin-pancreozymin test and the endoscopic pancreatic function test, together with the s-MRCP, are sensitive enough to be of help still in cases of inconclusive morphological findings of chronic pancreatitis. The role of other function tests like the quantification of the coefficient of fat absorption (CFA) and the ¹³C-mixed triglyceride breath test is limited to the diagnosis of exocrine pancreatic insufficiency with maldigestion. Fecal elastase is still useful for detecting reduction of pancreatic secretion in patients with different pancreatic diseases, but its sensitivity is too low to be used as a diagnostic test in early chronic pancreatitis. The use of fecal chymotrypsin is limited currently to the evaluation of patients’ compliance to enzyme replacement therapy. Other indirect tests used in the past (NBT-PABA test, pancreolauryl test) are no longer commercially available.

Disclosure: The author declares no conflicts of interest.

REFERENCES

1. Gullo L. Value and clinical role of intubation tests in chronic pancreatitis. In: Beger HG, Buchler M, Ditschuneit H, Malfertheiner P, eds. Chronic Pancreatitis. Springer Verlag: Berlin-Heidelberg; 1990:287–290.
2. Domý´nguez-Mun˜oz JE, Martý´nez S, Leodolter A, Malfertheiner P. Quantification of pancreatic zinc output as pancreatic function test: making the secretin-caerulein test applicable to clinical practice. Pancreatology. 2004;4:57–62.
3. Jin HO, Song CW, Chang TM, Chey WY. Roles of gut hormones in negative-feedback regulation of pancreatic exocrine secretion in humans. Gastroenterology. 1994;107:1828–1834.
4. Domý´nguez-Mun˜oz JE. Pancreatic function tests for diagnosis and staging of chronic pancreatitis, cystic fibrosis, and exocrine pancreatic insufficiency of other etiologies: which tests are necessary and how should they be performed in clinical routine? In: Dominguez-Munoz JE, ed. Clinical Pancreatology for Practising Gastroenterologists and Surgeons. Oxford: Blackwell-Publishing; 2005:259–266.
5. Stevens T, Conwell DL, Zuccaro G Jr, et al. A prospective crossover study comparing secretin-stimulated endoscopic and dreiling tube pancreatic function testing in patients evaluated for chronic pancreatitis. Gastrointest Endosc. 2008;67:458–466.
6. Conwell DL, Zuccaro G Jr, Vargo JJ, et al. An endoscopic pancreatic function test with cholecystokinin-octapeptide for the diagnosis of chronic pancreatitis. Clin Gastroenterol Hepatol. 2003;1:189–194.
7. Conwell DL, Zuccaro G Jr, Vargo JJ, et al. An endoscopic pancreatic function test with synthetic porcine secretin for the evaluation of chronic abdominal pain and suspected chronic pancreatitis. Gastrointest Endosc. 2003;57:37–40.
8. Raimondo M, Imoto M, DiMagno EP. Rapid endoscopic secretin stimulation test and discrimination of chronic pancreatitis and pancreatic cancer from disease controls. Clin Gastroenterol Hepatol. 2003;1: 397–403.
9. Stevens T, Dumot JA, Parsi MA, Zuccaro G, Vargo JJ. Combined endoscopic ultrasound and secretin endoscopic pancreatic function test in patients evaluated for chronic pancreatitis. Dig Dis Sci. 2010;55: 2681–2687.
10. Albashir S, Bronner MP, Parsi MA, Walsh RM, Stevens T. Endoscopic ultrasound, secretin endoscopic pancreatic function test, and histology: correlation in chronic pancreatitis. Am J Gastroenterol. 2010;105: 2498–2503.
11. Csakó L. Diagnosis of early-stage chronic pancreatitis by secretinenhanced magnetic resonance cholangiopancreatography. J Gastroenterol. 2007;42(suppl 17):113–117.
12. Balci NC, Smith A, Momtahen AJ, et al. MRI and s-MRCP findings in patients with suspected chronic pancreatitis: correlation with endoscopic pancreatic function testing (ePFT). J Magn Reson Imaging. 2010;31: 601–606.
13. Domý´nguez-Mun˜oz JE. Noninvasive pancreatic function tests. In: Büchler MW, Friess H, Uhl W, Malfertheiner P, eds. Chronic Pancreatitis: Novel Concepts in Biology and Therapy. Oxford: Blackwell Science; 2002:225–232.
14. Gullo L, Pezzilli R, Ventrucci M. Diagnostic value of the amino acid consumption test in pancreatic diseases. Pancreas. 1996;12:64–67.
15. Domý´nguez-Mun˜oz JE, Hyeronimus C, Sauerbruch T, Malfertheiner P. Fecal elastase test: evaluation of a new noninvasive pancreatic function test. Am J Gastroenterol. 1995;90:1834–1837.
16. Lankisch PG, Schmidt I. Fecal elastase 1 is not the indirect pancreatic function test we have been waiting for. Dig Dis Sci. 2000;45:166–167.
17. Löser C, Möllgaard A, Fölsch UR. Faecal elastase 1: a novel, highly sensitive and specific tubeless pancreatic function test. Gut. 1996;39: 580–586.
18. Keller J, Layer P. Pancreatic enzyme supplementation therapy. Curr Treat Options Gastroenterol. 2003;6:369–374.
19. Stein J, Purschian B, Bieniek U, Caspary WF, Lemcke B. Near-infrared reflectance analysis (NIRA): a new dimension in the investigation of malabsorption syndromes. Eur J Gastroenterol Hepatol. 1994;6:889–894.
20. Safdi M, Bekal PK, Martin S, Saeed ZA, Burton F, Toskes PP. The effects of oral pancreatic enzymes (Creon 10 capsule) on steatorrhea: a multicenter, placebo-controlled, parallel group trial in subjects with chronic pancreatitis. Pancreas. 2006;33:156-162. Erratum in: Pancreas. 2007;34:174.
21. Lembcke B, Braden B, Caspary WF. Exocrine pancreatic insufficiency: accuracy and clinical value of the uniformly labeled 13C-Hiolein breath test. Gut. 1996;39:668–674.
22. Vantrappen GR, Rutgeerts PJ, Ghoos YF, Hiele MI. Mixed triglyceride breath test: a noninvasive test of pancreatic lipase activity in the duodenum. Gastroenterology. 1989;96:1126–1134.
23. Ventrucci M, Cipolla A, Ubalducci GM, Roda A, Roda E. 13C-labelled cholesteryl octanoate breath test for assessing pancreatic exocrine insufficiency. Gut. 1998;42:81–87.
24. Domý´nguez-Mun˜oz JE, Iglesias-Garcý´a J, Vilarin˜o-Insua M, Iglesias-Rey M. 13C-mixed triglyceride breath test to assess oral enzyme substitution therapy in patients with chronic pancreatitis. Clin Gastroenterol Hepatol. 2007;5:484–488.

The incidence of esophageal adenocarcinoma (EA) has escalated over the last three decades in the Western world. Even more concerning is the high mortality of EA that suggests that the majority of people diagnosed with this disease will die from it. The incidence of adenocarcinoma arising in the setting of Barrett's esophagus (BE) is approximately 0.5% per patient-year.¹ The key to prevention and early treatment of EA is the detection and eradication of neoplasia found in patients with BE.

The management of neoplasia in BE has rapidly evolved to embrace endoscopic therapy and requires a multidisciplinary approach to provide appropriate care. Highly trained pathologists, endoscopists, and surgeons can offer effective, minimally invasive, and individualized care to patients withBE. This review outlines the critical functions of pathology, endoscopy, and surgery in this field and how they interact with each other to provide care to patients with BE with neoplasia.

Pathologists With Expertise in Gastrointestinal Pathology

Pathologists play a crucial role in the initial diagnosis of patients with Barrett's neoplasia, the staging of disease, and monitoring of therapy. In order to appreciate the need for a pathologist with expertise in gastrointestinal pathology, it is important to understand the pathologic definitions utilized for Barrett's associated neoplasia. Adenocarcinoma related to BE develops via a process of transformation from metaplasia to dysplasia to cancer. The proper identification of BE in stages of specialized intestinal metaplasia (IM), low grade dysplasia (LGD), high grade dysplasia (HGD), intramucosal carcinoma (IMC), and submucosal carcinoma (SMC) is fundamental in discussions of therapeutic options for Barrett's neoplasia. Dysplasia is defined as neoplastic cytologic and architectural atypia without evidence of invasion past the basement membrane. The diagnosis of low-grade dysplasia (LGD) or HGD is based on the severity of cytologic and architectural changes that suggest neoplastic transformation of the columnar epithelium as previously described.² High grade dysplasia and carcinoma in situ are regarded as equivalent in terms of pathologic significance as both are superficial to the basement membrane. Since the clinical implication of these two diagnoses is equivalent and the diagnoses are difficult to reproducibility distinguish, the diagnosis of carcinoma in situ is no longer used by gastrointestinal (GI) pathologists in the context ofBE. Intramucosal carcinoma is defined as a tumor that has penetrated the basement membrane but is still superficial to the lamina propria.

The exact definitions of the histological stage have a profound impact for treatment. For example, continued surveillance is recommended for nondysplastic BE and LGD, whereas HGD is an indication for intervention. Intramucosal cancer carries a minimal risk of nodal metastasis ^3–5 and, therefore, is considered amenable to endoscopic management in appropriately selected patients.⁶ In contrast, submucosal carcinoma (SMC) defined as a tumor invading past the muscularis mucosa into the submucosa but not into the muscularis propria, significantly increases the risk of nodal metastasis and therefore is usually an indication for surgery and/or systemic therapy.^{3 7–10}

Although the gold standard is histology, a high interinterpreter variability in diagnosing dysplasia is reported in the literature.^2,11–14 Therefore, it is important to include the opinion of a second pathologist with experience in dysplasia in gastrointestinal pathology. There are several challenges that present themselves in the diagnosis of dysplasia inBE. Neoplastic transformation represents a biologic and histological continuum, but the pathologist must utilize discrete categories of BE, LGD, HGD, IMC, and invasive cancer. The cytologic features of dysplasia can be confused with reactive inflammatory changes. Often endoscopic biopsy specimens are small and poorly oriented, making proper pathologic interpretation problematic. The identification of invasion of single cells through the basement membrane into the lamina propria, the feature that separates HGD and IMC, is especially difficult to recognize reproducibly. Invasion into the submucosa also may or may not be captured in superficial endoscopic biopsy specimens. Diagnoses based on biopsies obtained during initial pretreatment endoscopy reportedly differ from 39% to 48% of the time when compared to the histopathologic evaluation based on examination of the subsequent therapeutic EMR specimen, and there is a greater interobserver agreement between pathologists in examining EMR specimens than biopsy specimens. ^15–18 Therefore, endoscopic resection often provides a more accurate histologic diagnosis by providing large, well-oriented, and intact tissue specimens.

One unique feature specifically associated with BE is the high incidence of a duplication of muscularis mucosa, originally described as “musculo-fibrous anomaly.”^{19, 20} This feature is often not recognized by non-GI pathologists and has the potential to lead to a misdiagnosis of IMC as SMC—particularly on endoscopic biopsy specimens. A mistaken impression of submucosal invasion will be rendered if the duplicated muscularis mucosae is confused with the native muscularis mucosae and the native muscularis mucosa is regarded as the muscularis propria.^{21, 22} Furthermore, there may be a higher risk of lymphatic invasion in the presence of metastices in muscularis mucosa.^{21, 22} This underscores the need for review by expert gastrointestinal pathologists in the diagnosis of Barrett's associated neoplasia, the utility of endoscopic resection specimens to provide accurate staging, and further studies to examine how this anomaly impacts clinical staging. The grade of the IMC or SMC, presence of lymphatic or perineural invasion, and the depth of submucosal invasion all portend poorer outcomes with endoscopic treatment. Special attention to these details is helpful in fine-tuning and individualizing management plans for patients.

The pathologist plays a critical role in gauging adequacy of surgical and endoscopic resection of Barrett's related neoplasia. Details specific to endoscopic resection include the type of margins that are described. Positive lateral margins may be an indication for retreatment with endoscopic therapy. The remaining disease suggested by involvement of the deep margins of the endoscopic resection specimen, however, would not be amenable to endoscopic therapy and would be an indication for surgery.

Given these challenges in diagnosing dysplasia in BE, a second opinion by a pathologist with experience in gastrointestinal pathology is critical.

Appropriate Endoscopic Staging

Specialized centers offer improved detection of neoplasia with tissue acquiring endoscopic resection readily available to appropriately stage visible lesions and enhanced endoscopic imaging modalities to survey the remainder of the Barrett's epithelium for subtle or occult lesions.

Pathologists face challenges in distinguishing between HGD, IMC, and SMC on a single endoscopic biopsy specimen due to limited size and poor orientation. As part of an appropriate staging of a patient with a visible lesion, endoscopic resection of a lesion is an essential diagnostic evaluation. Specimens from endoscopic resection may provide a larger, intact, and well-oriented tissue from which pathologists may more reliably provide the histopathological stage of a lesion.

Endoscopic ultrasound (EUS) has a limited role in the assessment of superficial depth of a lesion in BE, but does play an important role in the identification of lymph node involvement. The endosonographic appearance of BE is a thickened mucosal layer. The EUS is not able to reliably distinguish between an IMC tumor and a submucosal tumor, and EMR is better suited for depth staging at this range.²³ The IMC carries a minimal risk of lymph node metastasis, ^24–26 EUS with fine needle aspiration (FNA) may identify patients not eligible for endoscopic therapy.²⁷ An EUS with or without FNA is a part of the diagnostic evaluation in patients with IMC and patients with visible lesions who have a higher risk of occult cancer. Evidence for lymph node metastasis is a contraindication for endoscopic therapy. The utility of EUS in HGD without any visible lesions is limited.²⁸

In addition to proper staging of visible lesions, the remainder of the Barrett's segment should be carefully investigated to evaluate for subtle or occult neoplasia using a rigorous protocol. Although the current standards are based on the random four quadrant biopsies every one to two centimeters to detect occult neoplasia, this method is still subject to sampling error. Detailed white light examination (WLE) is the foundation in detection of neoplasia in the setting of BE. Other modalities include chromoendoscopy, narrow band imaging (NBI) or other digital chromoendoscopy, autofluorescence imaging (AFI), and confocal laser endomicroscopy (CLE). These modalities demonstrate having the potential to improve the yield when detecting occult neoplasia in BE. Narrow band imaging (NBI) is one of the more accessible and commonly used enhanced imaging modalities. Narrow band imaging is based on the principle of filtered light at 415 nm and 540 nm that is absorbed by hemoglobin. The NBI enhances the mucosal surface pattern and the superficial capillaries. An irregular or distorted mucosal pattern is associated with HGD or cancer.²⁹ The NBI may be used for differentiation of HGD/cancer from LGD or non-neoplastic BE with a sensitivity of 97% based on a pooled analysis of five studies.³⁰ However, some studies question the improvement with NBI over detailed high resolution WLE and the yield of detection of neoplasia by NBI.^{31, 32} Advantages to this modality are that there are no additional risks to the patient, no additional preparation or sedation, wide availability on standard endoscope systems, ease of combination with other modalities, short learning curve, and relative ease of use.

Autofluorescence imaging (AFI) is a technology where the endogenous substances in the tissue, fluorophores, emit a light of longer wavelength when excited by short wavelength.³³ The AFI produces images in which normal squamous and nondysplastic BE appear green, areas that are suspicious for HGD/EA are a bluish purple color, and squamous islands amidst IM appear pinkish. A combined approach with HRE, AFI, and NBI has been studied that may assist with targeting of lesions, however, is not sufficient to replaced targeted biopsies of detected lesions or random biopsies. ^34–36

Another modality that is gaining increased attention in BE is confocal laser endomicroscopy (CLE), which allows for imaging of the GI lining with microscopic detail. The CLE requires the use of a contrast agent, most commonly intravenous fluorescein sodium. This agent carries a rare risk of anaphylaxis, but safety in GI imaging has been demonstrated.³⁷ Endoscopists who perform CLE must also have some familiarity with tissue architecture of the GI tract in order to interpret images. Endomicroscopy classification systems for BE with and without neoplasia have been described for endomicroscopy.^{38, 39} Prospective randomized trials have been performed demonstrating CLE as a tool that can potentially improve the diagnostic yield of endoscopically occult Barrett's neoplasia and decrease the number of negative biopsies.^{40, 41} Disadvantages of CLE technology include additional procedural time and cost. Further studies are needed for the clinical utility of these imaging modalities. A major limitation to many of the studies is the lack of a true gold standard. The common histological standard from targeted biopsies and random four quadrant biopsies is still fraught with sampling error as much of the mucosa remains not sampled.

Of the currently available technologies, detailed high resolution WLE examination with a trained observer is the principal foundation for recognizing suspicious areas. The success of these additional modalities is based on both the technology as well as the ability of the endoscopist to recognize where to look and how to interpret the findings. These modalities may be used prior to endoscopic therapy as part of the initial assessment or may have a role during endoscopic to guide therapy.^{42, 43}

APPROACH TO SURGICAL VERSUS ENDOSCOPIC Therapy

Until the current decade, esophagectomy had been the traditional standard treatment for BE with HGD and IMC. The rationale for the surgical treatment of HGD was based on the suspected risk of harboring occult invasive cancer with estimations in excess of 40% in those patients undergoing a prophylactic esophagectomy for the treatment of HGD.⁴⁴ However, in a systematic review of 23 studies applying both standardized criteria and strict definitions, the majority of these patients with occult invasive cancer had, in fact, IMC potentially amenable to endoscopic therapy. The rate of submucosal invasive cancer was 12.7%.⁴⁴ It is critical to distinguish the presence of BE with either HGD, IMC, or submucosal invasion as each of these has different implications for therapy.

Patients should be thoroughly counseled on the risks and benefits of therapy as well as the appropriate alternate modalities or approaches. Patients’ age, comorbidities, and willingness to undergo surgery also should be taken into account. The available surgical and endoscopic expertise at specific institutions also plays a role into which therapy may benefit a patient more. There is a strong trend today for limiting esophagectomy to more selected patients with submucosal invasion, evidence for lymph node metastasis, or unsuccessful endoscopic therapy.⁴⁵ High risk characteristics of Barrett's neoplasia should also be considered. For example, long segment Barrett's and multifocal neoplasia have been associated with higher risks of cancer and recurrence of neoplasia after endoscopic therapy. ^46–48 Also, poorly differentiated tumors or invasion of the lymphatic channels, blood vessels, or neural structures are associated with lymph node metastasis.^{26, 49} Ideally, the surgeon and endoscopist can coordinate to provide the most appropriate care for each patient after a careful endoscopic and histological assessment. Currently the approach to HGD and IMC is more complex and provides much more individualized care of patients than previously was available.

Armamentarium of Endoscopic Therapeutic Modalities

The approach to the management in BE has rapidly evolved based on the paradigm shift toward endoscopic therapy. The goal is to preserve the esophagus while ablating or removing the known neoplasia as well as the entire Barrett's segment.^{6, 50} Endoscopists also should be ready to offer a range of therapeutic modalities that include both tissue acquiring and nontissue acquiring therapies. Endoscopic resection, either endoscopic mucosal resection (EMR) or endoscopic submucosal dissection (ESD), is a tissue acquiring modality in which there is not only eradication of the lesion but a histologic specimen also is obtained. Nontissue acquiring modalities such as photodynamic therapy (PDT), radiofrequency ablation (RFA), and cryotherapy treat the lesion by tissue destruction. Multiple sessions and multiple modalities may be required to achieve total Barrett's eradication. Patients undergoing endoscopic therapy for Barrett's neoplasia need endoscopic surveillance indefinitely while in remission.

Focal EMR can be performed for endoscopically visible lesions that are suspicious for malignancy. Due to limited efficacy of focal EMR alone, treatment should be directed toward the known lesions and also the remainder of the Barrett's epithelium to address any metachronous or synchronous lesions. Additionally, esophagectomy should be readily available at these centers to expeditiously treat complications such as full thickness perforation resulting from endoscopic resection.

Complete Barrett's eradication EMR (CBE-EMR) has been performed to treat all of the known and at risk Barrett's epithelium by EMR. Complete responses have ranged from 76% to 100%.^15,51–54 The complication profile of EMR includes stricture formation with an incidence rate that approaches 50%, bleeding and, rarely, perforation.^{15 51–54} Of note, most esophageal stenoses and bleeding are amenable to endoscopic treatment.

Of the currently available ablative methods, RFA of BE tissue has demonstrated initial successful efficacy. The RFA can be achieved by the application of thermal therapy either circumferentially with a balloon-based catheter or locally with a focal device. Complications include noncardiac chest pain, nontransmural lacerations, and stricture formation. Radiofrequency ablation was investigated in patients with BE with dysplasia.^{55, 56} At a 1-year follow-up in 142 patients with BE HGD, there was remission of HGD in 90.2%, remission of dysplasia in 80.4%, and remission of IM in 54.3%.⁵⁵ In a recent landmark multicenter, sham-controlled trial, 127 patients with dysplastic BE were randomly assigned to receive either RFA or a sham procedure.⁵⁶There was complete eradication of dysplasia in 90.5% of the patients undergoing RFA as opposed to only 22.7% of the patients in the control group. There was also less disease progression in the RFA group and fewer cancers as compared to the control group (3.6% vs 16.3% and 1.2% vs 9.3%, respectively).⁵⁶

Performance of EMR of any visible lesions in the setting of dysplastic Barrett's with subsequent RFA of the remaining flat segment has been studied. In an examination of RFA with or without focal EMR, complete histological eradication of all dysplasia and IM was achieved in 43 patients (98%).⁵⁷ Given this efficacy and complication profile, RFA may provide a prime alternative in the treatment of BE with flat HGD as compared with EMR. Visible lesions in the setting of dysplasia are still best addressed by EMR for accurate histological diagnosis and treatment. The issue of subsquamous Barrett's glands remains a low, but possible, risk in patients who have undergone RFA treatment.^{58, 59} Random biopsies are subject to sampling error and may miss subsquamous glands.

In photodynamic therapy (PDT), a light sensitizing reaction sequence destroys the epithelium. This ablative modality did demonstrate success with BE with neoplasia,^{60, 61} but with the advent of other available endoscopic ablative options, PDT usage has gone out of favor in recent years due to its photosensitivity side effects, stricture rates, and presence of buried glands after therapy. A newer ablative therapy is cryotherapy, which is the spray application of liquid nitrogen that induces injury from the freeze–thaw cycle.⁶² A recent multicenter trial demonstrated a 97% complete eradication rate for HGD, 87% complete eradication of all dysplasia with persistent nondysplastic IM, and 57% had complete eradication of all IM.⁶³ Cryotherapy has demonstrated initial success with HGD but longer term studies are needed.

There is a low risk of cancer for patients with nondysplastic BE or with LGD, and ablative therapy has been studied for treatment in nondyspastic Barrett's.⁶⁴ At the same time, endoscopic therapy still carries possible complications. The benefit and risk profiles suggest that ablative therapies do not provide the same level of benefit for patients with nondysplastic BE or LGD compared with patients with HGD or cancer.⁶⁵ Future investigations on efficacy, recurrence of disease, cost-effectiveness, and quality of life will need to be performed to better address interventions in these low risk populations.

Surgical Expertise With Esophagectomy

Selected patients with HGD and IMC who are appropriate surgical candidates can benefit from esophagectomy. The surgical specimen enables accurate staging of disease to diagnose areas of occult cancer and confirms treatment adequacy with negative margins and lymph nodes. Esophagectomy for HGD and IMC also prevents death fromEA. The issue of the morbidity and mortality from esophagectomy is the major concern. Adverse outcomes include pulmonary complications, hemorrhage, anastomotic leak, infection, and recurrent nerve palsy. Although the morbidity rate associated with esophagectomy may be as high as 50% and mortality may be 10%,⁶⁶ the expertise and volume of the center, the experience of the surgeon, the patient risk factors, and the indications for the esophagectomy all should be taken into account. ^67–69 The mortality rate of esophagectomy for cancer is lower in institutions with high esophagectomy volumes and surgical expertise.^{69, 70} It is also important to note that esophagectomy specifically for HGD has a lower risk profile with a pooled mortality rate of 1%,⁶⁷ compared to esophagectomy performed for cancer. Comorbid diseases, debilitation from cancer and/or neoadjuvant therapy, and issues with locally advanced disease are issues of greater relevance for patients who present with cancer. Furthermore, postoperative quality of life indicators for patients who have undergone esophagectomy for HGD and IMC are equivalent to those of the general population.⁷¹

The best surgical option for HGD or IMC is the one that produces the least morbidity balanced against the best long-term survival. Conventional approaches are transhiatal esophagectomy and transthoracic esophagectomy. Minimally invasive esophagectomy (MIE) techniques include video-assisted thoracoscopy surgery with laparotomy or with laparoscopy, laparoscopy with a right thoracotomy, or laparoscopic transhiatal resections. The MIE procedures have potential advantages because of their perceived benefits of reduced pain, lower incidence of postoperative complications, and faster recovery, but any superiority of MIE over open esophagectomy has yet to be proven due to lack of randomized, prospective trials.^{72, 73} The use of vagal-sparing techniques, especially for HGD, has potentially interesting advantages with regards to quality of life, but has not been adequately evaluated in terms of staging accuracy and long-term outcomes. The more extensive resections likely improve nodal staging accuracy. Long-term functional status is similar regardless of the surgical approach. Ultimately, the needs of the patient and the surgeon's experience determine the most appropriate approach to esophagectomy for HGD or IMC. Most importantly, esophagectomy when needed should be performed by individuals with experience and in centers with high esophagectomy volume.

CONCLUSION

The management of Barrett's neoplasia must start with a sound endoscopic and histological evaluation. The management strategy increasingly utilizes endoscopic therapy as a means to preserve the esophagus while ablating or removing the entire Barrett's segment. Proper patient selection and assessment of the Barrett's epithelium is required for successful therapy. Endoscopic therapy may require multiple sessions and may include both tissue acquiring endoscopic resection and ablative methods. Surgery plays an important role in the treatment of Barrett's neoplasia for those patients who have submucosal invasion, lymph node metastasis, or have failed endoscopic therapy. Ideally, centers should incorporate expertise from gastrointestinal pathology, endoscopy, and surgery into a multidisciplinary approach in order to provide the most appropriate and effective care for patients with Barrett's associated neoplasia.

Acknowledgements

Dr. Konda, Hart, and Waxman have received honoraria from Mauna Kea Technologies. Dr. Waxman has also received support from Cook Medical and Olympus

REFERENCES

1. Shaheen NJ, Crosby MA, Bozymski EM, Sandler RS. Is there publication bias in the reporting of cancer risk in Barrett’s esophagus? Gastroenterology. 2000;119(2):333–338.
2. Montgomery E, Bronner MP, Goldblum JR, et al. Reproducibility of the diagnosis of dysplasia in Barrett esophagus: a reaffirmation. Hum Pathol. 2001;32(4):368–378.
3. Paraf F, Flejou JF, Pignon JP, Fekete F, Potet F. Surgical pathology of adenocarcinoma arising in Barrett’s esophagus. Analysis of 67 cases. Am J Surg Pathol. 1995;19(2):183–191.
4. Feith M, Stein HJ, Siewert JR. Pattern of lymphatic spread of Barrett’s cancer. World J Surg. 2003;27(9):1052–1057.
5. Stein HJ, Feith M, Bruecher BL, Naehrig J, Sarbia M, Siewert JR. Early esophageal cancer: pattern of lymphatic spread and prognostic factors for long-term survival after surgical resection. Ann Surg. 2005;242(4): 566–573; discussion 573–575. PMCID:1402356.
6. Bergman JJ, Sharma P. Total Barrett eradication: is it ready for prime time? Endoscopy. 2008;40(12):991–993.
7. Holscher AH, Bollschweiler E, Schneider PM, Siewert JR. Early adenocarcinoma in Barrett’s oesophagus. Br J Surg. 1997;84(10):1470–1473.
8. Siewert JR, Stein HJ, Feith M, Bruecher BL, Bartels H, Fink U. Histologic tumor type is an independent prognostic parameter in esophageal cancer: lessons from more than 1,000 consecutive resections at a single center in the Western world. Ann Surg. 2001;234(3):360–367; discussion 368–369. PMCID:1422027.
9. Rice TW, Zuccaro G Jr, Adelstein DJ, Rybicki LA, Blackstone EH, Goldblum JR. Esophageal carcinoma: depth of tumor invasion is predictive of regional lymph node status. Ann Thorac Surg. 1998;65(3): 787–792.
10. van Sandick JW, van Lanschot JJ, ten Kate FJ, et al. Pathology of early invasive adenocarcinoma of the esophagus or esophagogastric junction: implications for therapeutic decision making. Cancer. 2000;88(11):2429– 2437.
11. Alikhan M, Rex D, Khan A, Rahmani E, Cummings O, Ulbright TM. Variable pathologic interpretation of columnar lined esophagus by general pathologists in community practice. Gastrointest Endosc. 1999; 50(1):23–26.
12. Ormsby AH, Petras RE, Henricks WH, et al. Observer variation in the diagnosis of superficial oesophageal adenocarcinoma. Gut. 2002;51(5): 671–676. PMCID:1773435.
13. Reid BJ, Haggitt RC, Rubin CE, et al. Observer variation in the diagnosis of dysplasia in Barrett’s esophagus. Hum Pathol. 1988;19(2):166–178.
14. Kerkhof M, van Dekken H, Steyerberg EW, et al. Grading of dysplasia in Barrett’s oesophagus: substantial interobserver variation between general and gastrointestinal pathologists. Histopathology. 2007;50(7): 920–927.
15. Chennat J, Konda VJ, Ross AS, et al. Complete Barrett’s eradication endoscopic mucosal resection: an effective treatment modality for highgrade dysplasia and intramucosal carcinoma—an American single-center experience. Am J Gastroenterol. 2009;104(11):2684–2692.
16. Hull MJ, Mino-Kenudson M, Nishioka NS, et al. Endoscopic mucosal resection: an improved diagnostic procedure for early gastroesophageal epithelial neoplasms. Am J Surg Pathol. 2006;30(1):114–118.
17. Moss A, Bourke MJ, Hourigan LF, et al. Endoscopic resection for Barrett’s high-grade dysplasia and early esophageal adenocarcinoma: an essential staging procedure with long-term therapeutic benefit. Am J Gastroenterol. 2010;105(6):1276–1283.
18. Wani S, Mathur SC, Curvers WL, et al. Greater interobserver agreement by endoscopic mucosal resection than biopsy samples in Barrett’s dysplasia. Clin Gastroenterol Hepatol. 2010;8(9):783–788.
19. Rubio CA, Riddell R. Musculo-fibrous anomaly in Barrett’s mucosa with dysplasia. Am J Surg Pathol. 1988;12(11):885–889.
20. Takubo K, Sasajima K, Yamashita K, Tanaka Y, Fujita K. Double muscularis mucosae in Barrett’s esophagus. Hum Pathol. 1991;22(11): 1158–1161.
21. Lewis JT, Wang KK, Abraham SC. Muscularis mucosae duplication and the musculo-fibrous anomaly in endoscopic mucosal resections for Barrett esophagus: implications for staging of adenocarcinoma. Am J Surg Pathol. 2008;32(4):566–571.
22. Abraham SC, Krasinskas AM, Correa AM, et al. Duplication of the muscularis mucosae in Barrett esophagus: an underrecognized feature and its implication for staging of adenocarcinoma. Am J Surg Pathol. 2007;31(11):1719–1725.
23. Pech O, Gunter E, Ell C. Endosonography of high-grade intra-epithelial neoplasia/early cancer. Best Pract Res Clin Gastroenterol. 2009;23(5):639–647.
24. Westerterp M, Koppert LB, Buskens CJ, et al. Outcome of surgical treatment for early adenocarcinoma of the esophagus or gastroesophageal junction. Virchows Arch. 2005;446(5):497–504.
25. Leers JM, Demeester SR, Oezcelik A, et al. The prevalence of lymph node metastases in patients with T1 esophageal adenocarcinoma: a retrospective review of esophagectomy specimens. Ann Surg. 2011;253(2):271– 278.
26. Ancona E, Rampado S, Cassaro M, et al. Prediction of lymph node status in superficial esophageal carcinoma. Ann Surg Oncol. 2008;15(11):3278– 3288.
27. Shami VM, Villaverde A, Stearns L, et al. Clinical impact of conventional endosonography and endoscopic ultrasound-guided fine-needle aspiration in the assessment of patients with Barrett’s esophagus and high-grade dysplasia or intramucosal carcinoma who have been referred for endoscopic ablation therapy. Endoscopy. 2006;38(2):157–161.
28. Thomas T, Gilbert D, Kaye PV, Penman I, Aithal GP, Ragunath K. Highresolution endoscopy and endoscopic ultrasound for evaluation of early neoplasia in Barrett’s esophagus. Surg Endosc. 2010;24(5):1110–1116.
29. Sharma P, Bansal A, Mathur S, et al. The utility of a novel narrow band imaging endoscopy system in patients with Barrett’s esophagus. Gastrointest Endosc. 2006;64(2):167–175.
30. Curvers WL, van den Broek FJ, Reitsma JB, Dekker E, Bergman JJ. Systematic review of narrow-band imaging for the detection and differentiation of abnormalities in the esophagus and stomach (with video). Gastrointest Endosc. 2009;69(2):307–317.
31. Curvers WL, Bohmer CJ, Mallant-Hent RC, et al. Mucosal morphology in Barrett’s esophagus: interobserver agreement and role of narrow band imaging. Endoscopy. 2008;40(10):799–805.
32. Herrero LA, Curvers WL, Bansal A, et al. Zooming in on Barrett oesophagus using narrow-band imaging: an international observer agreement study. Eur J Gastroenterol Hepatol. 2009;21(9):1068–1075.
33. Falk GW. Autofluorescence endoscopy. Gastrointest Endosc Clin N Am. 2009;19(2):209–220.
34. Curvers WL, Singh R, Song LM, et al. Endoscopic tri-modal imaging for detection of early neoplasia in Barrett’s oesophagus: a multi-centre feasibility study using high-resolution endoscopy, autofluorescence imaging and narrow band imaging incorporated in one endoscopy system. Gut. 2008;57(2):167–172.
35. Curvers WL, Herrero LA, Wallace MB, et al. Endoscopic tri-modal imaging is more effective than standard endoscopy in identifying earlystage neoplasia in Barrett’s esophagus. Gastroenterology. 2010;139(4):1106– 1114.
36. Curvers WL, van Vilsteren FG, Baak LC, et al. Endoscopic trimodal imaging versus standard video endoscopy for detection of early Barrett’s neoplasia: a multicenter, randomized, crossover study in general practice. Gastrointest Endosc. 2011;73(2):195–203.
37. Wallace MB, Meining A, Canto MI, et al. The safety of intravenous fluorescein for confocal laser endomicroscopy in the gastrointestinal tract. Aliment Pharmacol Ther. 2010;31(5):548–552.
38. Kiesslich R, Gossner L, Goetz M, et al. In vivo histology of Barrett’s esophagus and associated neoplasia by confocal laser endomicroscopy. Clin Gastroenterol Hepatol. 2006;4(8):979–987.
39. Pohl H, Rosch T, Vieth M, et al. Miniprobe confocal laser microscopy for the detection of invisible neoplasia in patients with Barrett’s oesophagus. Gut. 2008;57(12):1648–1653.
40. Dunbar KB, Okolo P III, Montgomery E, Canto MI. Confocal laser endomicroscopy in Barrett’s esophagus and endoscopically inapparent Barrett’s neoplasia: a prospective, randomized, double-blind, controlled, crossover trial. Gastrointest Endosc. 2009;70(4):645–654. PMCID:2755622.
41. Sharma P, Meining AR, Coron E, et al. Real-time increased detection of neoplastic tissue in Barrett’s esophagus with probe-based confocal laser endomicroscopy: final results of an international multicenter, prospective, randomized, controlled trial. Gastrointest Endosc. 2011;74(3):465–72.
42. Ross AS, Noffsinger A, Waxman I. Narrow band imaging directed EMR for Barrett’s esophagus with high-grade dysplasia. Gastrointest Endosc. 2007;65(1):166–169.
43. Konda VJ, Chennat JS, Hart J, Waxman I. Confocal laser endomicroscopy: potential in the management of Barrett’s esophagus. Dis Esophagus. 2010; 23(5):E21–E31.
44. Konda VJ, Ross AS, Ferguson MK, et al. Is the risk of concomitant invasive esophageal cancer in high-grade dysplasia in Barrett’s esophagus overestimated? Clin Gastroenterol Hepatol. 2008;6(2):159–164.
45. Konda VJ, Ferguson MK. Esophageal resection for high-grade dysplasia and intramucosal carcinoma: when and how? World J Gastroenterol. 2010; 16(30):3786–3792. PMCID:2921090.
46. Pech O, Behrens A, May A, et al. Long-term results and risk factor analysis for recurrence after curative endoscopic therapy in 349 patients with high-grade intraepithelial neoplasia and mucosal adenocarcinoma in Barrett’s oesophagus. Gut. 2008;57(9):1200–1206.
47. Weston AP, Krmpotich PT, Cherian R, Dixon A, Topalosvki M. Prospective long-term endoscopic and histological follow-up of short segment Barrett’s esophagus: comparison with traditional long segment Barrett’s esophagus. Am J Gastroenterol. 1997;92(3):407–413.
48. Tharavej C, Hagen JA, Peters JH, et al. Predictive factors of coexisting cancer in Barrett’s high-grade dysplasia. Surg Endosc. 2006;20(3):439–443.
49. Gockel I, Domeyer M, Sgourakis GG, et al. Prediction model of lymph node metastasis in superficial esophageal adenocarcinoma and squamous cell cancer including D2–40 immunostaining. J Surg Oncol. 2009;100(3):191–198.
50. Waxman I, Konda VJ. Mucosal ablation of Barrett esophagus. Nat Rev Gastroenterol Hepatol. 2009;6(7):393–401.
51. Lopes CV, Hela M, Pesenti C, et al. Circumferential endoscopic resection of Barrett’s esophagus with high-grade dysplasia or early adenocarcinoma. Surg Endosc. 2007;21(5):820–824.
52. Peters FP, Kara MA, Rosmolen WD, et al. Stepwise radical endoscopic resection is effective for complete removal of Barrett’s esophagus with early neoplasia: a prospective study. Am J Gastroenterol. 2006;101(7):1449– 1457.
53. Seewald S, Akaraviputh T, Seitz U, et al. Circumferential EMR and complete removal of Barrett’s epithelium: a new approach to management of Barrett’s esophagus containing high-grade intraepithelial neoplasia and intramucosal carcinoma. Gastrointest Endosc. 2003;57(7): 854–859.
54. Pouw RE, Seewald S, Gondrie JJ, et al. Stepwise radical endoscopic resection for eradication of Barrett’s oesophagus with early neoplasia in a cohort of 169 patients. Gut. 2010;59(9):1169–1177.
55. Ganz RA, Overholt BF, Sharma VK, et al. Circumferential ablation of Barrett’s esophagus that contains high-grade dysplasia: a U.S. multicenter registry. Gastrointest Endosc. 2008;68(1):35–40.
56. Shaheen NJ, Sharma P, Overholt BF, et al. Radiofrequency ablation in Barrett’s esophagus with dysplasia. N Engl J Med. 2009;360(22):2277– 2288.
57. Pouw RE, Gondrie JJ, Sondermeijer CM, et al. Eradication of Barrett esophagus with early neoplasia by radiofrequency ablation, with or without endoscopic resection. J Gastrointest Surg. 2008;12(10):1627–1636; discussion 1636–1637.
58. Semlitsch T, Jeitler K, Schoefl R, et al. A systematic review of the evidence for radiofrequency ablation for Barrett’s esophagus. Surg Endosc. 2010; 24(12):2935–2943.
59. Adler DC, Zhou C, Tsai TH, et al. Three-dimensional optical coherence tomography of Barrett’s esophagus and buried glands beneath neo-squamous epithelium following radiofrequency ablation. Endoscopy. 2009;41(9):773–776. PMCID:2890226.
60. Overholt BF, Panjehpour M, Halberg DL. Photodynamic therapy for Barrett’s esophagus with dysplasia and/or early stage carcinoma: longterm results. Gastrointest Endosc. 2003;58(2):183–188.
61. Overholt BF, Wang KK, Burdick JS, et al. Five-year efficacy and safety of photodynamic therapy with Photofrin in Barrett’s high-grade dysplasia. Gastrointest Endosc. 2007;66(3):460–468.
62. Gross SA, Wolfsen HC. The role of photodynamic therapy in the esophagus. Gastrointest Endosc Clin N Am. 2010;20(1):35–53, vi.
63. Shaheen NJ, Greenwald BD, Peery AF, et al. Safety and efficacy of endoscopic spray cryotherapy for Barrett’s esophagus with high-grade dysplasia. Gastrointest Endosc. 2010;71(4):680–685.
64. Fleischer DE, Overholt BF, Sharma VK, et al. Endoscopic radiofrequency ablation for Barrett’s esophagus: 5-year outcomes from a prospective multicenter trial. Endoscopy. 2010;42(10):781–789.
65. Wani S, Puli SR, Shaheen NJ, et al. Esophageal adenocarcinoma in Barrett’s esophagus after endoscopic ablative therapy: a meta-analysis and systematic review. Am J Gastroenterol. 2009;104(2):502–513.
66. Bailey SH, Bull DA, Harpole DH, et al. Outcomes after esophagectomy: a ten-year prospective cohort. Ann Thorac Surg. 2003;75(1):217–222; discussion 222.
67. Fernando HC, Murthy SC, Hofstetter W, et al. The Society of Thoracic Surgeons practice guideline series: guidelines for the management of Barrett’s esophagus with high-grade dysplasia. Ann Thorac Surg. 2009;87(6):1993–2002.
68. Reavis KM, Smith BR, Hinojosa MW, Nguyen NT. Outcomes of esophagectomy at academic centers: an association between volume and outcome. Am Surg. 2008;74(10):939–943.
69. Birkmeyer JD, Stukel TA, Siewers AE, Goodney PP, Wennberg DE, Lucas FL. Surgeon volume and operative mortality in the United States. N Engl J Med. 2003;349(22):2117–2127.
70. Law S. Esophagectomy without mortality: what can surgeons do? J Gastrointest Surg. 2010;14(suppl 1):S101–S107.
71. Moraca RJ, Low DE. Outcomes and health-related quality of life after esophagectomy for high-grade dysplasia and intramucosal cancer. Arch Surg. 2006;141(6):545–549; discussion 549–551.
72. Decker G, Coosemans W, De Leyn P, et al. Minimally invasive esophagectomy for cancer. Eur J Cardiothorac Surg. 2009;35(1):13–20; discussion 20–21.
73. Biere SS, Cuesta MA, van der Peet DL. Minimally invasive versus open esophagectomy for cancer: a systematic review and meta-analysis. Minerva Chir. 2009;64(2):121–133.

Protein-losing enteropathy is characterized by hypoproteinemia and malabsorption from gastrointestinal tract without any failure of protein synthesis. This description, however, is a general term for a group of diseases characterized by protein lose, resulting from local or systemic conditions.¹ Pathologic findings of protein-losing enteropathy are lymphatic obstruction, mucosal ulceration, or idiopathic mucosal damage. Followed by protein lose, the oncotic pressure decreases and hypoproteinemia causes symptomatic edema. The levels of serum immunoglobulins, transferrin, seruloplasmin, and fibrinogen are decreased and immune functions deteriorate. ^2–4

Intestinal lymphangiectasia is a commonly seen cause of protein-losing enteropathy in children considering the adults.⁵ Primarily it is seen in childhood due to congenital obstruction of lymphatics and the secondary form can be done from various etiological factors.

In this study, we report the case of a 52-year-old man consulted to us with dyspnea and bilateral leg swelling from emergency care unit and searched for etiological factors that cause hypoalbuminemia. In this case, it is emphasized that intestinal lymphangiectasia can be seen as a rare cause of protein-losing enteropathy with pleural effusion without any abdominal symptoms.

CASE

A 52-year-old man consulted to our clinic with complaints of dyspnea and bilateral leg swelling. He had this complaint for two years. In his medical history, there was no more remarkable finding. His vital signs were normal. On physical examination, bilateral pulmonary sounds were not heard beginning from middle zones. He had pitting edema of scrotal, sacral, and bilateral lower extremities. The laboratory findings showed remarkable changes.

The white blood cell count was within the normal range; however, lymphocytopenia was evident (800/mm³). Erythrocyte sedimentation rate was normal (8 mm/h). Laboratory examination showed the decreased levels of total serum protein (4.6 g/dl) and albumin (1.9 g/dl). Liver, renal function tests, and electrolyte levels were normal, but some of mineral levels were decreased [Zn^+ +: 6 ?g/dl (N: 10.4–16.4) Cu^+ +: 9 ?mol/L (N: 11–22)]. Serum total cholesterol and triglyceride levels were 82 and 67 mg/dl, and below the normal range. Urinalysis was normal. Autoimmune serological markers, including antigliadin, antiendomysial, and tissue transglutaminase antibodies, were negative. Serum levels of vitamin A, E, D were decreased [vitamin A: 0.77 µmol/L (N: 1.05–3.32), vitamin E: 5 µmol/L (N: 11.6–46.4), 25-hydroxyvitamin D: 10 nmol/L (N: 25–137)]. He had anergic puri?ed protein derivative skin test. He had transudative pleural effusion without any atipic cells. Examination of eye ground was normal. On chest X-ray examination, bilateral cardiophrenic was closed and fluid sequestration was evident (Fig. 1). No lesion was reported on thorax–abdominal computerized tomography (CT). Ecocardiography results were normal. Gallium-67 whole body scan for the suspicion of sarcoidosis was also not remarkable.

There was not any kind of protein lose related to renal system, so etiologic factors are oriented to gastrointestinal system. Upper gastrointestinal system revealed pangastritis and mucosal biopsies were taken from morphologically normal duodenal tissue. Biopsy results revealed marked dilatation of lymphatic vessels composing lacteals located at the end of villi (Fig. 2). Abdominal ultrasonography and abdomen CT results were normal. In colonoscopy, there was an ulcer found in terminal ileum so biopsy was taken. No source explaining severe hypoalbuminemia, anemia, lymphopenia, and hypolipidemia was found through imaging studies so Tc99m T-lymphocyte marked HIC scintigraphy was taken. Extravasation of human immunoglobulins was found in images taken on fourth, sixth, eighth, and twenty fourth hours. Enteroscopy showed advanced jejunum of 70 cm and revealed white color areas of lymphangiectasia located on mucosa (Fig. 3).

Treatment regimen of the patient was revised, following diagnosis of intestinal lymphangiectasia. A low-fat diet enriched with protein and medium-chain triglycerides was ordered. After two months from discharge, the patient had no abnormal pulmonary sounds and any other complaints (Fig. 4). Control laboratory and imaging studies were normal. The patient oral dietary therapy is advised to keeping on and controls by three month intervals were arranged.

DISCUSSION

In intestinal lymphangiectasia, lymphatic fluid and lymphocytes escape from dilated lymphatic vessels to intestinal lumen or peritoneal space.⁶ As a result of lymphatic fluid leakage from gastrointestinal tract plasma albumin, immunoglobulins, and other proteins decrease. Whole body is affected as a consequence of third space fluid collections. Lymphocytopenia and hypogammaglobulinemia occurred. Opportunistic infections due to impaired chemotaxis and cellular immunity can also be seen.^{7, 8}

In some cases, intermittent diarrhea and steathore can occur.⁹ In cases with lymphangiectasia primary symptoms are edema and diarrhea. In our case, edema is the primary symptom and severe dyspnea caused by bilateral pleural effusion worsens the disease process. Gold standard diagnostic tool for intestinal lymphangiectasia is endoscopic biopsy.¹⁰ Taking mucosal or submucosal dilated lymphatic vessels by biopsy has diagnostic value.

Other diagnostic methods are useful especially for excluding secondary factors. In this case, whole gastrointestinal tract is searched by endoscope and multiple biopsies are taken. Duodenal mucosal biopsies are taken twice and dilated lymphatics are reported. So it is pathologically diagnosed as lymphangiectasia. Intestinal lymphangiectasia should be considered by patient symptoms, laboratory, and radiological methods.

Essential secondary causes of lymphangiectasia are congestive heart failure, constrictive pericarditis, fontan procedure, lymphoma, sarcoidosis, tuberculosis, radiotherapy, and chemotherapeutics.⁴ In this case, the patient is examined primarily for these secondary causes but any secondary cause cannot be found for dilatation of lymphatics.

In literature duodenal lymhangiectasia is also described.¹¹,¹² In this case, endoscopic and histopathologic findings are similar but hypoalbuminemia and lymphocytopenia are not seen.

In our case, clinic and histopathologic findings confirm primary lymphangiectasia although age at diagnosis and few lesions are unusual for it. In literature, symptomatic functional lymphangiectasia is not reported yet. Nutritional arrangement is a main component in the treatment of protein malabsorption caused by intestinal lymphangiectasia. A diet involving low fat, high protein, and enriched with middle chain fatty acids is suggested.¹² Patients do not succesfully respond nutritional therapy octreotid administration is found effective. This drug was first used by Bac et al. in a case with lymphangiectasia.^{13, 14} With octreotid administration, intestinal blood and lymphatic stream decrease and hence protein lose is prevented. However, considering its side effects and cost-effectiveness, its role in the treatment is still controversial.¹⁵ In the cases with segmental involvement, surgery is also a choice of treatment.

In conclusion, intestinal lymphangiectasia is a rare cause of protein-losing enteropathy which is developed primary or secondary. It should always be considered in cases with generalized edema, ascites, pleuritis, and hypoalbuminemia. Considering the patients age, symptoms, and unknown underlying factor, intestinal lymphangiectasia is always remembered as a cause of protein-losing enteropathy.

Acknowledgements

Disclosure: The authors declare that they have no conflict of interest.

REFERENCES

1. Zeng Y, Wang F, Williams ED, et al. Lymphatics in the alimentary tract of children in health and disease: study on mucosal biopsies using the monoclonal antibody d2-40. Pediatr Dev Pathol. 2005;8(5):541–549.
2. Goldberg RI, Calleja GA. Protein losing gastroenteropathy. In: Haubrich WS, Scahffner F, Berk JE, eds. Bockus gastroenterology. Philadelphia: WB Saunders; 1995:1072–1086.
3. Patel AS, DeRidder PH. Endoscopic appearance and significance of functional lymphangiectasia of the duodenal mucosa. Gastrointest Endosc. 1990;36(4):376–378.
4. Milovic V,Caspary WF,Stein J,et al. Protein-losing gastroenteropathy. In:Up To Date Gastroenterology(Online) Available at: http:// www.uptodate.com/ 5. Hardikar W, Smith AL, Murch S Disease of small intestine in childhood. 4th ed. Oxford: ISIS; 1999.
6. Greenwald DA. Protein losing gastroenteropathy. In: Feldman M, ed. Sleisenger and Fordtran’s gastrointestinal and liver disease. 8th ed. Philadelphia: Saunders; 2006:557–562.
7. Rossi P, Covarelli P, Cirocchi R, et al. Intestinal lymphangiectasis in adults. G Chir. 1996;17(4):171–174.
8. Nazer HM, Abutalib H, Hugosson C, et al. Intestinal lymphangiectasia masquerading as coeliac disease. Ann Trop Paediatr. 1991;11(4):349–355.
9. Brasitus TA, Bissonnette BM. Protein losing enteropathy. In: Feldman M, Scharschmidt BF, Sleisenger MH, eds. Gastrointestinal and liver disease, 6th ed. Philadelphia: W.B Saunders 1619–1620.
10. Aoyagi K, Iida M, Yao T, et al. Characteristic endoscopic features of intestinal lymphangiectasia: correlation with histological findings. Hepatogastroenterology. 1997;44(13):133–138.
11. Barnes RE, DeRidder PH. Fat absorption in patients with functional intestinal lymphangiectasia and lymphangiectic cysts. Am J Gastroenterol. 1993;88(6):887–890.
12. Ohno S, Nakahara S, Kasahara K, et al. A case report of primary intestinal lymphangiectesia successfully treated with low fat diet. Nihon Shokakibyo Gakkai Zasshi. 1997;94(11):767–771.
13. Bac DJ, Van Hagen PM, Postema PT, et al. Octreotide for protein-losing enteropathy with intestinal lymphangiectasia. Lancet. 1995;345(8965):1639.
14. Kuroiwa G, Takayama T, Sato Y, et al. Primary intestinal lymphangiectasia successfully treated with octreotide. J Gastroenterol. 2001;36(2):129–132.
15. Ballinger AB, Farthing MJ. Octreotide in the treatment of intestinal lymphangiectasia. Eur J Gastroenterol Hepatol. 1998;10(8):699–702.

Syphilis is an infection caused by the spirochete bacteria Treponema pallidum through sexual contact, microtrauma of mucous membranes, or ulcerated genital lesions. Recent data from the Centers for Disease Control and Prevention (CDC) show that the number of cases of primary and secondary syphilis in the United States increased during 2001–2009, mainly as a result of high-risk sexual behaviors, particularly among men who have sex with men (MSM). In that population, syphilis cases reported to CDC increased from 13 500 in 2008 to 13 997 in 2009, an increase of 3.7% (accounting for 4.6 cases per 100,000 population).¹

Secondary syphilis represents a progressive phase of early acquired syphilis and is characterized by multisystem involvement, with manifestations appearing about eight weeks after transmission (or even within two years of infection). Clinical manifestations often include maculopapular rash of the palms and soles, condylomas, and generalized lymphadenopathy, being hepatitis an uncommon finding.² Cases of syphilitic hepatitis are more frequently observed among patients infected with human immunodeficiency virus (HIV), especially those with acquired immunodeficiency syndrome (AIDS).^{3, 4}

We report a case of syphilitic hepatitis diagnosed in an immunocompetent homosexual patient during the investigation of acute cholestatic syndrome.

CASE REPORT

A 22-year-old white male presented to the emergency department of a tertiary reference Hospital in Porto Alegre, Rio Grande do Sul, in the South of Brazil, with a seven day history of jaundice and pruritus. The patient reported initial right-sided hypochondral pain that began 30 days earlier and evolved with fatigue, anorexia, night sweats without fever, and weight loss (4 kilograms) during that period. Over the last seven days he developed jaundice, choluria, acholia, and pruritus. His medical history included hepatitis A in childhood. The patient denied other comorbidities, regular use of medications, and current and/or previous intravenous drug use. The patient informed being a sexually active homosexual, with a steady partner and regular condom use (including a few incidents of condom breakage).

Physical examination revealed good general condition, presence of tattoos, and jaundice. Small, mobile, painless lymph nodes were palpated in the left posterior cervical chain, occipital (left/right), and right inguinal glands. Findings included small round brown maculae on the left sole and abdomen without visceromegaly or pain on palpation. Genital examination revealed superficial ulcers on the penis glans. Furthermore, a recent total abdominal ultrasound showed hepatosplenomegaly.Laboratory tests included: AST (114 mg/dL [<37]) and ALT (177 mg/dL [<41]), total bilirubin (5.64 mg/dL [<1.2]), immediate direct bilirubin (4.63 mg/dL [<0.4]), erythrocyte sedimentation rate (66 mm/h [<20]), C-reactive protein (20.7 mg/L [<5]), Gama GT (729 mg/dL [8–61]), and alkaline phosphatase (808 mg/dL [40–129]). Serological tests: anti-HIV negative, anti-HCV negative, HBsAg negative, anti-HBc IgM negative, anti-HAV IgM negative/ IgG positive; negative results for atypical lymphocytes and mononucleosis, toxoplasma-IgM negative/ IgG positive; cytomegalovirus IgM negative/ IgG positive, VDRL positive 1:64. After clinical and laboratory findings suggestive of secondary syphilis, a positive treponemal test (hemagglutination assay) confirmed the diagnosis. The evaluation of the patient's clinical features, with increased transaminase levels, cholestasis, and positive treponemal tests, led to the diagnosis of syphilitic hepatitis.

Treatment was initiated with deep intramuscular injections of benzathine penicillin G, 2.4 million units, at weekly doses for three weeks, with outpatient follow-up screening for monitoring and control tests, which showed negative VDRL and normalization of liver function tests two weeks after the third dose of benzathine penicillin G treatment.

After confirmation of the diagnosis, the patient and his partner were interviewed in order to investigate high-risk sexual behaviors and similar lesions. We found that his partner was an HIV-infected (non-AIDS) patient, on treatment, and had skin lesions suggestive of syphilis. The partner was referred for evaluation and investigation with an infectious disease specialist.

CONCLUSION

Liver involvement in syphilis has been recognized for 400 years, but is uncommon and tends to be overlooked.⁵

Patients diagnosed with syphilis at any stage of the disease may present varying degrees of abnormalities in liver tests; however, the development of clinically evident hepatitis and cholestasis is quite uncommon.^5,6

In the literature, the incidence of syphilitic hepatitis reportedly ranged from 0.24% to 17% . ^7–10

Syphilitic hepatitis is characterized by a significant increase in alkaline phosphatase (AP), in contrast to a slight increase in aminotransferase levels, rarely leading to increased bilirubin levels. A disproportionate increase in levels of AP, sometimes ten times the normal level, in the absence of accountable cholestasis has been suggested as a feature of syphilitic hepatitis,¹¹ although many of the cases reported in the literature may have had elevated AP due to sources other than the liver; thus, an elevated AP level in cases of secondary syphilis cannot be a reliable index of liver damage.¹¹ Some authors^{12, 13} emphasize that the AP activity should be interpreted with care because periostitis, or more severe bony lesion, may be seen in secondary syphilis, and many patients have no microscopic cholestasis on liver biopsy.¹²

After clinical suspicion, and usually after positive non-treponemal tests (VDRL), the diagnosis is confirmed by serological treponemal tests (FTA-Abs, TPHA).¹⁴

The pathogenesis of liver involvement in syphilis has not been fully elucidated, and several mechanisms have been considered, including direct injury to the portal venous system by spirochetes and immune complex-mediated autoimmune reaction. Studies describing liver biopsies in these patients have demonstrated portal lymphocytic infiltrates, hepatocellular necrosis, noncaseating granulomas with multinucleated giant cells, and inflammation of intrahepatic bile ducts, changes to which laboratory abnormalities have been attributed.^{14, 15}

Secondary syphilis is typically a systemic disease, with the patient presenting with a variety of symptoms, such as malaise, sore throat, headache, weight loss, low grade fever, pruritus, and muscle aches, in addition to the dermatologic manifestations. Lymph node enlargement is present in the great majority of patients.¹⁰ Jarisch-Herxheimer reaction (manifested by fever, chills, headache, myalgia and/or exacerbation of skin lesions), despite very common in the setting of syphilis, was not found in our patient after antibiotic treatment.

Once the diagnosis is confirmed, treatment with deep intramuscular injections of benzathine penicillin, 2.4 million units, at weekly doses for three weeks, should be promptly initiated. Sexual partners should be interviewed about symptoms and clinical changes, which should be investigated.

The relevance of the present report is based on the rare occurrence of hepatitis secondary to syphilis, especially in immunocompetent patients. Although rare, this diagnosis should always be considered when investigating hepatitis and/or cholestasis, because of its severity and unfavorable outcome if left undetected, particularly in young people with high-risk sexual behaviors. In Brazil, there is no mandatory reporting of syphilis, except congenital and gestational syphilis, reason why there are no accurate data on disease incidence. A literature search was conducted for medical articles involving Latin American countries, especially Brazil, using major international medical databases (LILACS, SciELO, BIREME, and MEDLINE). The search returned no reports on hepatitis due to syphilis in immunocompetent patients in our country and this is, therefore, the first report on the topic.

REFERENCES

1. Centers for Disease Control and Prevention (CDC). 2009. Sexually transmitted diseases surveillance-Syphilis. Available from: http:// www.cdc.gov/std/stats09/syphilis.htm. Accessed Dec 24, 2010.
2. Lee V, Kinghorn G. Syphilis: an update. Clin Med. 2008;8:330–333.
3. Crum-Cianflone N, Weekes J, Bavaro M. Syphilitic hepatitis among HIVinfected patients. Int J STD AIDS. 2009;20:278–284.
4. Mullick CJ, Liappis AP, Benator DA, et al. Syphilitic hepatitis in HIVinfected patients: a report of 7 cases and review of the literature. Clin Infect Dis. 2004;39:e100–e105.
5. Noto P, Nonno FD, Licci S, Chinello P, Petrosillo N. Early syphilitic hepatitis in an immunocompetent patient: really so uncommon? Int J STD AIDS. 2008;19:65–66.
6. Agrawal NM, Sassaris M, Brooks B, Akdamar K, Hunter F. The liver in secondary syphilis. South Med J. 1982;75:1136–1138.
7. Irgang S. The problem of involvement of the liver in syphilis. Arch Dermatol Syphiol. 1937;36:685–707.
8. Hahn RD. Syphilis of the liver. Am J Syphilol. 1943;27:529–562.
9. Feher J, Somogyi T, Timmer M, Jozsa L. Early syphilitic hepatitis. Lancet. 1975;2:896–899.
10. Baughn RE, Musher DM. Secondary syphilitic lesions. Clin Microbiol Rev. 2005;18:205–216.
11. Veeravahu M. Diagnosis of liver involvement in early syphilis. A critical review. Arch Intern Med. 1985;145:132–134.
12. Terry SI, Hanchard B, Brooks SE, McDonald H, Siva S. Prevalence of liver abnormality in early syphilis. Br J Vener Dis. 1984;60:83–86.
13. Baker AL, Kaplan MM, Wolfe HJ, McGowan JA. Liver disease associated with early syphilis. N Engl J Med. 1971;284:1422–1423.
14. Ridruejo E, Mordoh A, Herrera F, Avagnina A, Mando OO. Severe cholestatic hepatitis as the first symptom of secondary syphilis. Dig Dis Sci. 2004;49:1401–1404.
15. Haburchak DR, Davidson H. Anorectal lesions and syphilitic hepatitis. West J Med. 1978;128:64–67.

Familial adenomatous polyposis (FAP) is a well-known entity, featuring widespread neoplastic growth in the gastrointestinal tract. In the colon, 1000 of polyps develop, of which some invariably develop into carcinomas. Prophylactic colectomy at a young age is generally advised.

Upper gastrointestinal polyps have increasingly been recognized in these patients. In the stomach, multiple small, mostly cystic gland polyps apparently cause little, if any harm, although adenomas are found in approximately 5% of cases. However, the majority of patients also demonstrate duodenal polyps. These are often dysplastic and warrant follow-up.¹ Duodenal cancer is the main cause of death in colectomized FAP patients, but duodenal polyps are far less amenable to curative surgery, and Whipple's procedure or other variations of duodenectomy are reserved for patients with severe duodenal disease.² Otherwise, endoscopic resection of prominent or rapidly expanding lesions is usually advised, usually by piecemeal snare resection techniques.

Although the literature on chemoprevention on duodenal polyposis is scarce, some studies have indicated an inhibitory effect of NSAIDs and a selective Cox-2 inhibitor, celecoxib, on duodenal polyp growth.^{3, 4}

The primary aim of this study was to evaluate the effects of treatment with another selective Cox-2 inhibitor, rofecoxib, in colectomized patients with FAP and duodenal polyposis on the polyp load in the duodenum by endoscopy as well as assessment of immunohistochemical changes and genetic Cox-2 expression in duodenal biopsies. We also wanted to relate the results in biopsies to the expression of Cox-2 and prostaglandins in blood cells to further elucidate the cellular mechanisms of selective Cox-2 inhibition.

MATERIAL AND METHODS

Most FAP patients in Norway have been under the care of the National Polyposis Center at Oslo University Hospital-Rikshospitalet where surgical and endoscopic data are collected and available. All FAP patients between the ages of 18 and 70 with documented duodenal polyposis were eligible for inclusion. Exclusion criteria were duodenal findings amenable to endoscopic or surgical resection, known hypersensitivity to the study drug, or unwillingness to comply with the protocol.

Study Design

After inclusion, the patients were randomized to treatment with rofecoxib (Vioxx) 25 mg once daily or placebo for 1 year, in a double-blind double dummy design. Before treatment was initiated, baseline demographics, diet history, and blood samples were obtained and upper endoscopy performed.

Patient inclusion was started January 2003. The study drug was withdrawn in September 2004, due to data indicating an increased risk of cardiovascular events. The study was terminated immediately resulting in an incomplete treatment period for some of the patients. We decided to include data from study objects completing at least 9 months of treatment, based on previously published studies and treatment duration therein.

Endoscopic, clinical, and biochemical assessment were repeated at completion of the study. Moreover, phone follow-up was performed at 1 and 3 months, primarily to identify drug side effects or other adverse events.

Endoscopy

Upper endoscopy was done after an overnight fast with a standard forward-viewing endoscope under conscious sedation and with 20 mg Buscopan for enteral paralysis. Accurate imaging of the duodenal bulb and descending part of the duodenum were performed on insertion and retraction with digital video recording of the entire procedure. Then, two to four biopsies were taken from the most prominent lesions, as well as two samples from endoscopically normal duodenal mucosa. Spigelman classification was performed after histological assessment (Table 1).

At the end of the study, all videos were reviewed blindly by two highly experienced (>25 years of academic endoscopic practice) endoscopists by comparing before/after recordings in all patients, without being aware of the study order. Based on this, a conclusion of “no change,” “little less,” “much less,” “little more,” and “much more” was made. This assessment model was chosen to improve on the inherent crudeness of the Spigelman classification, in particular as regards endoscopic findings.

Histology and Grading of Dysplasia

Blind assessment of duodenal biopsies was done on 5 µm thick sections from paraffin embedded tissue blocks from adenomatous tissue. The samples were assessed as to degree of dysplasia, at baseline, and at completion of trial therapy. Grading was “mild,” “moderate,” or “severe” to concur with the Spigelman classification. A total of 175 biopsies were assessed. If adenomas from the same patient and time point showed different grades of dysplasia, the categorization was based on the most severely dysplastic area.

Immunohistochemistry

The sections were stained with mouse anti Cox-2 (Zymed Laboratoris, Inc., San Francisco, CA) 1/50 on a Ventana Nexes machine with Ventana “I” view DAB detection kit (Ventana Medical Systems Inc., Tucson AZ) and counterstained for 10 sec with Harri's Hematoxiline. All sections were scanned on a Hamatsu Nanozoomer HT Model: C9600-02, and stored on a Western digital power book. The analysis was preformed with NDP Viewer Dual (Nano Zoomer Digital Pathology ver. 1.0.3 Hamamtsu Photonics K.K.).

Blind assessment of anti-Cox-2 staining was performed, comparing normal and adenomatous tissue at baseline and after active treatment or placebo. The assessment was performed with regard to intensity, distribution, and granularity. The following comparisons were performed:<list list-type="order"> </list>

• Staining of adenomatous tissue within patients before or after treatment.

• Staining of normal mucosa within patients before and after treatment.

• Staining of normal compared to adenomatous tissue within patients before and after treatment.

RNA Isolation, cDNA Synthesis, and Real-Time PCR (or Gene Expression Analysis)

Leucocytes were isolated from fasting blood samples by lysing red blood cells and centrifugation. Cell pellets containing leucocytes were immediately placed on dry ice and stored at −80°C. Duodenal biopsies were obtained from the patients during endoscopy and put directly in tubes containing RNA later (Ambion, Applied Biosystems, Austin, TX) for storage at −80°C until RNA isolation.

The RNA was isolated by the Trizol/RNeasy hybrid protocol. Briefly, leucocytes or biopsies were mixed with chilled 1 ml Trizol reagent (Invitrogen, TM Life Technologies) in an MM 400 Retsch Mixer Mill (Haan, Germany). Chloroform (0.2 ml) was added to the solution (1 ml), mixed, and followed by incubation for 5 min at room temperature ending with phase separation by centrifugation (12,000g) at 4°C for 10 min. The upper aqueous phase was mixed with an equal volume of 70% ethanol (4°C). Each sample was transferred to an RNeasy mini column (Qiagen), and further RNA isolation was according to the manufacturer's instructions. All samples were treated with Rnase-free Dnase (Qiagen) to remove traces of genomic DNA. Capillary electrophoresis (RNA 600 Nano LabChip, Agilent) on an Agilent Bioanalyser 2100 system showed RNA Integrity Number (RIN) values between 6.2 and 9.4 (mean: 8.1) for leucocytes, and 6.7 and 9.4 (mean: 8.7) for the biopsies. The RNA concentration was measured on a Nanodrop ND-1000 Spectrophotometer (NanoDrop Technologies Wilmington, Delaware). For cDNA synthesis, reversed transcription of RNA was carried out by Super Script (Invitrogen) according to the manufacturer's protocol.

Real-Time Polymerase Chain Reaction

Gene expression analyses of leucocytes and duodenal biopsies were carried out on a 7900HT real-time PCR machine from Applied Biosystems. Glucuronidase (GUS) was chosen as the housekeeping gene. The primers and probes were initially designed as three assays per gene, and validated for efficiency and specificity. The best performing sets of the three were then chosen for further analysis (details available upon request).

TaqMan Universal PCR Master Mix (Applied Biosystems) was added as reaction mix. The reaction conditions were initiated by a step of 2 min at 50°C and 10 min at 95°C, followed by 40 cycles of denaturation at 95°C for 15 sec and annealing at 60°C for 1 min.

For Cox-2 expression analysis, standards and samples were analyzed in triplicate for all assays. A combination of cDNA from several samples was prepared and diluted in order to make a dilution curve that was included on each plate. The average of the three values for each gene was divided by the average of the corresponding GUS values, generating a normalized value of the gene expression, which is a unit-less value used to compare the relative amount of mRNA for each gene in the different samples.

Prostaglandin Analysis

The EDTA plasma was isolated from the patients and controls by centrifugation and stored at −80°C. The samples were thawed and analyzed for PGE₂ concentration (pg/ml) by ELISA, according to the manufacturer's instructions (R&D, London, UK).

Ethics

The study was performed in accordance with the Helsinki Declaration. All patients were informed personally by one of the study physicians, in addition to thorough written information given prior inclusion.

At withdrawal of the active study drug, the principal investigator was notified immediately. The following day, all patients were addressed in writing and by phone to discontinue the trial drug if still taking it, and to have their final assessment done as rapidly as practically possible. This included the trial specific assessment, in addition to any symptoms or signs indicative of coronary heart disease. No such pathology was detected.

RESULTS

Demographics

Thirty-eight patients were initially included in the study. Due to the drug withdrawal, only 32 completed at least 9 months of treatment and attended a final assessment. Gender distribution was 17 males and 15 females. Gender, ages, and other clinical parameters were comparable between the Vioxx treated and placebo groups and are shown in Table 2.

Endoscopic Findings

The macroscopic lesions appeared unchanged after intervention, both in the treatment and placebo group (Table 3). Moreover, no differences were found regarding Spigelman classification, neither between groups nor within the individual groups before and after treatment.

Histology

All of the biopsies were scored as mild to moderate dysplasia (Fig. 1), with no significant change after treatment, between or within groups.

Cox Expression

Immumohistochemistry

Both adenomatous and normal tissue showed unchanged COX-2 expression after treatment with rofecoxib, whereas an increase was observed both for adenomatous and normal tissue in the placebo group (p=.007 and p=.01). A higher concentration of COX-2 was shown in the normal compared to adenomatous mucosa, both at baseline (p<.01) and after treatment (p<.01) (Fig. 2). No differences were found for staining distribution in adenomatous tissue, whereas a tendency of more scattered distribution was shown in normal mucosa after placebo treatment (p=.054) (Fig. 3A). A tendency of increased granularity of adenomatous tissue (Fig. 3B) was shown in the placebo group (0.053).

RNA Expression and Real-Time PCR

Biopsies and Leucocytes: The COX-2 expression was unchanged after treatment with rofecoxib and slightly reduced after placebo (ns) in both tissues (Fig. 4 and Fig. 5).

Plasma PGE ₂ : A nonsignificant increase of plasma PGE2 was found in the placebo group, and a slight decrease in the rofecoxib group (Fig. 6).

DISCUSSION

In the present study, no significant effect of 9–12 months treatment with Cox-2 inhibitors and no evidence of severe dysplasia were found among 32 patients with a follow-up of a median of 20 years after colectomy and approximately 12 years after the diagnosis of duodenal polyposis.

Previous trials have demonstrated reduced adenoma formation and growth, in vitro and in vivo, after treatment with selective Cox-2 inhibitor in duodenal adenomas,⁴ and in Min mice models,⁵ as well as in colorectal adenomas.⁶

Several murine models that resemble FAP have been used to determine whether various NSAIDs and selective COX-2 inhibitors suppress the development of spontaneous intestinal adenomas.⁷ In Apc-Min mice, Cox-2 inhibition clearly attenuate tumor development.^{8, 9}

Numerous epidemiologic studies presented in a recent meta-analysis reported consistently that individuals on regular use of aspirin and other NSAIDs have a lower incidence of adenomatous polyps and death from colorectal cancer compared with nonusers.¹⁰ Sustained use of NSAIDs was associated with a 30–50% reduction in incidence of adenomatous polyps and death from colorectal cancer, in all but 1 of 13 epidemiologic studies.

After a proctocolectomy, the majority of FAP patients are left with duodenal adenomas. In a previous large multinational study, a cumulative incidence of duodenal adenomatosis in FAP patients was shown to increase from 65% at the age of median 38 years, to 90% at the age of 70.¹¹ Moreover, during a 7.5 years observation period, a significant increase in Spigelman score along with increasing size, multiplicity of duodenal adenomas, as well as a 4.5% cumulative incidence rate of duodenal cancer at age 57 years formed the basis for the authors’ recommendation of surveillance of these patients. Currently, the general recommendation is to perform regular upper endoscopy in all FAP patients from the age of 25–30 years¹ and, in addition, some centers are advocating prophylactic COX-2 inhibitors in the Spigelman I–III group, as well as prophylactic surgery in the Spigelman IV group.¹¹

On the basis of comparisons between endoscopy and later surgery in duodenal FAP patients, it was also postulated that understaging of malignancy is a serious danger in these patients, due to the high number of advanced stage cancers in patients subjected to surgery on the indication of “severe duodenal adenomatosis.”¹¹ One criticism of that study was the relatively high median age of the cohort, with the risk of self-selection of the cases with bad prognosis at an earlier age before selection of the screening cohort.

Previous studies have demonstrated diverging patterns, regarding the frequency of advanced lesions in the duodenum of FAP patients. A study from France showed advanced lesions in 50% of cases after a mean follow-up of 48 months among 58 FAP patients,¹² whereas only 6.2% of 32 patients showed advanced lesions after 73 months in an Italian study.¹³

In agreement with Saurin et al.¹² the Danish experience¹¹ reported on a lifetime risk of about 50%, emphasising the significance of age as the most important risk factor for progression.

In comparison to previous studies, the median age of the present material at time of tissue sampling was relatively low at approximately 41 years. Correspondingly, the median age at diagnosis of duodenal polyposis was 29 years, less than 10 years after a colectomy. The relatively low dysplastic score and few significant lesions, may be explained by the low age and limited duration of follow-up in comparison to the reported materials showing a high prevalence of advanced lesions and cancer, both related to age.¹¹

In another randomized placebo-controlled study over 6 months, celecoxib, another selective Cox-2 inhibitor, was administered at a dose of 400 mg twice daily and shown to reduce the area covered by adenomatous tissue by 14.5% compared to placebo. Moreover, a 31% reduction in affected area was shown for those with significant disease compared to 8% on placebo.³ This is in contrast to the present study where neither rofecoxib nor placebo caused any change in macroscopic polyp load compared to baseline. This lack of effect might have been caused by a suboptimal dose or lower efficacy of rofecoxib compared to celecoxib.³ Lack of drug compliance also appears unlikely given the drug intake monitoring applied in our study.

The unchanged macroscopic status was supported by an unchanged grade of dysplasia demonstrated at histology, and the unchanged immunohistochemical Cox-2 expression in adenomatous and normal tissue after treatment with rofecoxib. On the other hand, a small but significantly increased immunohistochemical staining was shown in the placebo group, possibly reflecting a lack of growth inhibition in the untreated group.

A somewhat unexpected higher concentration of Cox-2 was shown in histologically normal compared to adenomatous mucosa, both at baseline and after treatment. The lower immunohistochemical Cox-2 concentration in adenomas, may in part explain the lack of effect of the selective Cox-2 inhibitor in the present study, provided that a minimum of neoplastic activity is needed as a basis for an effect of treatment. A recent study¹⁴ showed an effect of aspirin on reducing the development of colorectal cancer (CRC), mainly in tumors demonstrating abnormal Cox-2 expression. If extrapolated to adenomas, the results may suggest that the individual Cox-2 expression may be a prerequisite for CRC prevention by selective Cox-2 inhibitors, which has not been generally addressed in previous pharmacologic studies. A marked individual variability in the response to selective inhibitors of COX-2 has previously been reported, as regards plasma concentrations of CYP2C9, with an even greater interindividual variability.¹⁵

There is also evidence for variation in genetic polymorphism between populations that could partly explain differences in response to treatment.¹⁶ Genetic polymorphism of the Cox-2 gene is largely unknown in the Norwegian population, and the rational for cancer prevention with Cox-2 inhibitors may be related to differences between populations or to the individual Cox-2 expression.

The surprisingly reduced Cox-2 level in adenomatous compared to normal mucosa, might also be explained by genetic differences between populations, however, larger studies must be performed to evaluate this question. Based on the very low levels of duodenal Cox-2 in the present study, the results might be in conflict with the previously demonstrated increased Cox-2 expression in duodenal compared with colonic tissues within FAP patients.¹⁷ On the other hand, their demonstration of a significantly higher expression in normal duodenal compared to colonic mucosa related to the −765GG genotype, more common among Caucasians, might be in concordance with the present results.

Nevertheless, one cannot rule out the possibility of endogenous mechanisms behind the low Cox-2 level in the present duodenal adenomas. One explanation might be changes in the bile acid composition in colectomized patients. A relationship to the presence of bile was demonstrated in a previous study, in which a correlation was found between the site of duodenal adenomas and the site of exposure to bile acid.¹⁸ Moreover, ex vivo experiments have shown that Cox-2 expression increases in response to bile acids and stomach acid.^{19, 20} Other studies have indicated a chemopreventive effect of ursodeoxycholic acid in mouse and rat intestine, alone²¹ or in combination with sulindac.²²

In the present material, a change in serum fatty acid composition was found, with a significantly increased level of arachidonic acid and docosahexanoic acid, and a significant reduction in palmitic acid, stearic acid, alpha-linoleic, and linoleic acid compared to healthy controls.²³ It is an interesting question if this fatty acid composition could have had an effect on the cyclooxygenase expression in the present material.

It might be speculated that changes in the fatty acid metabolism as shown in a recent study,²³ might have had an influence on the the arachidonic acid/Cox-2 pathway, and the balance between Cox-1 and Cox-2 within the lipid metabolism, which in turn could have affected the mucosal Cox-2 expression in FAP patients.

The present study has several limitations. The sample size is moderate. It comprises, however, all patients available in Norway with the diagnosis in question.

The lack of clinical effect of the treatment was somewhat disappointing. A reason for this might be that the dose applied was too low, that the duration of treatment was too short (partially due to the withdrawal of the drug), or that more advanced dysplasia combined with a longer duration of disease and higher age of the patients might be necessary to obtain an effect of Cox-2 inhibitors. In accordance with this, a surprisingly low immunohistochemical and genetic Cox-2 expressions was found in adenomatous tissue of the present material. Future studies should focus on genetic polymorphisms or endogenous factors, as possible explanations for discrepancies in Cox-2 expression between studies, in order to improve our understanding of the basis for future treatment with Cox-2 inhibitors.

In conclusion, the present study did not support an effect of the selective Cox-2 inhibitor, rofecoxib, on the size and multiplicity of duodenal adenomas in FAP patients over a 9-month period of treatment.

Acknowledgements

Disclosures: The study was supported by an unconditional grant from Merck Sharp & Dome, Norway.

Conflicts of Interest: None to report.

REFERENCES

1. Vasen HFA, Moslein G, Alonso A, et al. Guidelines for the clinical management of familial adenomatous polyposis (FAP). Gut. 2008;57: 704–713.
2. Spigelman AD, Talbot IC, Penna C, et al. Evidence for adeno-carcinoma sequence in the duodenum of patients with familial adenomatous polyposis. The Leeds Castle Polyposis Group (Upper Gastrointestinal Committee). J Clin Pathol. 1994;47:709–710.
3. Phillips RKS, Wallace MH, Lynch PM, et al. A randomised, double blind, placebo controlled study of celecoxib, a selective cyclooxygenase-2inhibitor, on duodenal polyposis in familial adenomatous polyposis. Gut. 2002;50:857–860.
4. Brosens LAA, Keller JJ, Offerhaus GJA, Goggins M, Giardielleo FM. Prevention and management of duodenal polyps in familial adenomatous polyposis. Gut. 2005;54:1034–1043.
5. Swamy M, Patlolla JM, Steele VE, Kopelowich L, Reddy BS, Rao CV. Chemoprevention of familial adenomatous polyposis by low doses of atorvastatin and celecoxib given individually and in combination to APCMin mice. Cancer Res. 2006;66:7370–7377.
6. Steinbach G, Lynch PM, Phillips RK, Wallace MH, Hawk E, Gordon GB, Wakabayashi N, Saunders B, Shen Y, Fujimura T, Su LK, Levin B. The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. NEJM. 2000;342:1946–1952.
7. Kensler TW, Tsuda H, Wogan GN. United States-Japan workshop on new rodent models for the analysis and prevention of carcinogenesis. Cancer Epidemiol Biomarkers Prev. 1999;8:1033–1037.
8. Jacoby RF, Seibert K, Cole CE, Kelloff G, Lubet RA. The cyclooxygenase-2 inhibitor celoxib is a potent preventive and therapeutic agent in the Min mouse model of adenomatous polyposis. Cancer Res. 2000;60:5040–5044.
9. Oshima M, Murai N, Kargman S, et al. Chemoprevention of intestinal polyposis in the Apc(deka)716 mouse by rofecoxib, a specific cyclooxygenase-2 inhibitor. Cancer Res. 2001;61:1733–1740.
10. Thun MJ, Henley J, Patrono C. Nonsteroidal anti-inflammatory drugs and anticancer agents: mechanistic, pharmacologic, and clinical issues. J Natl Cancer Inst. 2002;94:252–266.
11. Bulow S, Bjork J, Christensen IJ, et al. Duodenal adenomatosis in familial adenomatous polyposis. Gut. 2004;53:381–386.
12. Saurin J-C, Gutknecht C, Napoleon B, et al. Surveillance of duodenal adenomas in familial adenomatous polyposis reveals high cumulative risk of advanced disease. J Clin Oncol. 2004;22:493–498.
13. Santucci R, Volpe L, Zannoni U, Paganelli GM, Poggi B, Calabrese C. Cell proliferation of the duodenal mucosain patients affected by familial adenomatous polyposis. Gastroenterology. 1997;113:1159–1162.
14. Chan AT, Shuji Ogino MPH, Fuchs CS. Aspirin and the risk of colorectal cancer in relation to the expression of cox-2. NEJM. 2007;356:2131–2142.
15. Lee CR, Goldstein JA, Pieper JA. Cytochrome P450 2CP (CYP2C9 polymorphisms: a comprehensive review of the in vitro and human data. Pharmacogenetics. 2002;12:251–263.
16. Ulrich CM, Bigler J, Sibert J, et al. Cyclooxygenase-1 polymorphisms in African-American and Caucasian populations. Hum Mutat. 2002;20: 409–410.
17. Brosens LAA, Iacobuzio-Donahue CA, Keller JJ, et al. Increased cyclooxygenase-2 expression in duodenal compared with colonic tissues in familial adenomatous polyposis and relationship to the 756G-C cox-2 polymorphism. Clin Cancer Res. 2005;11:4090–4096.
18. Spigelman AD, Williams CB, Talbot IC, Domizio P, Phillips RK. Upper gastrointestinal cancer in patients with familial adenomatous polyposis. Lancet. 1989;2:783–785.
19. Hao X, Bishop AE, Wallace M, et al. Early expression of cyclo-oxygenase-2 during sporadic colorectal carcinogenesis. J Pathol. 1999;187:295–301.
20. Zhang F, Subbaramaiah K, Altorki N, Dannenberg AJ. Dihydroxy bile acids activate the transcription of cyclooxygenase-2. J Biol Chem. 1998; 273:2424–2428.
21. Earnest DL, Holubec H, Wali RK. Chemoprevention of azoxymethaneinduced colonic carcinogenesis by supplementary dietary ursodeoxycolic acid. Cancer Res. 1994;54:5071–5074.
22. Jacoby RF, Cole CE, Hawk ET, Lubet RA. Ursodeoxycholate/sulindac combination treatment effectively prevents intestinal adenomas in a mouse model of polyposis. Gastroenterology. 2004;127:838–844.
23. Almendingen K, Høstmark AT, Fausa O, Mosdøl A, Aabakken L, Vatn MH. Familial adenomatous polyposis patients have high levels of arachidonic acid and docosahexanoic acid and low levels of linoleic acid and a-linoleic acid in serum phospholipids. Int J Cancer. 2006;120: 632–637.

Uveal melanoma (UM) is a rare tumor but is the most frequent primary intraocular malignancy in adults, accounting for about 6% of all melanoma diagnoses.¹ Aggressive treatments of localized disease with ocular enucleation and radiotherapy often allow the cure but unfortunately almost 50% of the patients will develop distant metastases, with 5-year overall survival for newly diagnosed patients ranging from 50% to 70%.^2,3 Among all the possible sites of distant recurrence, the liver is by far the most predominant. Up to 40% of patients have been reported to have the presence of hepatic metastases present since the initial diagnosis, and the liver becomes involved in up to 95% of individuals who develop spreading disease, with liver failure as the most common cause of death.^4,5 Prognosis for patients with liver metastases is extremely poor with median survival reported to be 2–7 months and with less than 10% of patients alive at 1 year since diagnosis. ^6-8

Metastatic UM has shown to be refractory to immunotherapy and chemotherapy regimens. Many systemic treatment strategies using immunotherapy such as interferons and interleukin-2; chemotherapy, including dacarbazine (DTIC), cisplatin, temozolomide, or lomustine; or the antiangiogenic agent thalidomide, alone or in combination have been tested with disappointing results.^9,10

Because of the limited efficacy of systemic treatments and considering that liver failure is the main cause of death in these patients, locoregional treatments directed against liver metastases seems to be a rational approach. Surgical resection has been investigated but with no apparent benefit due to the high frequency of liver relapses and the morbidity and mortality associated.11–13"> ^11-13

Transarterial chemoembolization (TACE) is a relatively new locoregional approach that has shown efficacy in hepatocellular carcinoma. Some reports investigated this strategy also in UM patients with liver metastases with interesting results. ^14-16 TACE combines hepatic artery embolization with simultaneous infusion of concentrated doses of chemotherapeutic drugs.^17,18 The theoretical advantages of this technique include tumor ischemia, high drug concentration in tumor microenvironment, and reduced systemic toxicity. Irinotecan (IRI) has been reported to be active in a variety of advanced chemoresistant solid tumor including melanoma.¹⁹ When IRI-loaded microembolization is performed to treat liver lesions, tumor drug concentrations are reported to be 10–25 times higher than those achieved by systemic simple infusion alone and the dwell time of the agents is markedly prolonged with very few side effects. ^20-24

We have reported that TACE with IRI-loaded microspheres (DEBIRI) is active and well tolerated in the treatment of colorectal cancer liver metastases.^25,26 According to these interesting data and to our previous experiences adopting TACE with Irinotecan preloaded microspheres in LM from colorectal cancer and primary liver tumors, we planned a phase II study to assess the safety and efficacy of this new kind of TACE in this disease.

MATERIAL AND METHODS

Patient Characteristics

The study was a multicenter, single-arm, open-label phase II clinical study, conducted in compliance with the protocol and the principles laid down in the Declaration of Helsinki, in accordance with the ICH Harmonized Tripartite Guideline for Good Clinical Practice (GCP). Fifty two patients were recruited from four centers. All the patients has been treated by one center (Ferrara C.A. and G.B. MD). Empoli and Padova Hospitals were reference centers for UM. Patients came from the north and central Italy. Informed consent was obtained from the subject prior to evaluation, screening, and treatment.

Inclusions criteria were normal leukocyte and platelets count, hemoglobin levels >12 g/dl, preserved renal and liver function with creatinine <2 mg/dl, total bilirubin <3mg/dl, transaminases <1.5×the upper limit of normal, albumin levels >3 g/dl, normal PT and PTT, a Karnofsky score ≥60%, ejection fraction >55% measured at a cardiac EUS, absence of significant signs or symptoms of cardiac failure (≤NHYA class 2), absence of pulmonary hypertension, absence of hypersensitivity to iodine contrast enhancer and irinotecan, no known prior or concomitant malignancies, no pregnancy or lactation at the time of enrollment, and at least one prior systemic treatment with immunotherapy and/or chemotherapy.

Clinical evaluation was performed before the procedure and 1 month after TACE by means of contrast-enhanced, multidetector computed tomography (MDTC). Follow-up assessments included verification of the clinical-laboratory status (complete blood cell count, creatinine, total bilirubin, albuminemia, PT, PTT, alkalin phosphatase, and transaminases) and a MDTC (Brilliance 64 slice, Philips Medical Systems, Needherland). The CT scans included triphasic study of the liver every 3 months, with evaluation of the enhanced pattern of target lesions and tumor response rate according to modified RECIST criteria. Clinical complications were classified according to the WHO scale Version 2009. Each patient was asked to fill in an ESAS (Edmonton Symptom Assessment System) questionnaire a commonly used symptom assessment tool for advanced cancer and palliative patients to assess quality of life.²⁷

Treatment Administration and Drug Preparation

We adopted the same program of supportive treatment and intra-arterial lidocaine used in our previous experience.²⁵ The prophylactic treatment to prevent renal failure was hydration given intravenously, which started on the day before TACE and continued on days 0–2 with 2000 ml (1000 ml of saline solution, 1000 ml of glucose 5%) with the addition of ranitidine 900 mg, infused for 24 hours. The prophylactic treatment against nausea and vomiting was based on 5 mg tropisetron, one vial before TACE and one vial after 6 hours on day 0, and 8 mg dexamethasone at 8.00 a.m. and 8.00 p.m. on days 0–5. The prophylactic treatment against pain was based on 10 mg morphine, one vial 30 minutes before and 6 hours after TACE. Intra-arterial lidocaine (5 ml) was infused selectively to the vascular bed to be treated immediately before TACE. Prophylactic treatment against infections was based on 2000 mg cefazolin at 8.00 a.m. and 8.00 p.m. on days 0–2. The supportive treatment was maintained whenever required on days 3–5.

Treatment consisted of intra-arterial infusion of slow-release irinotecan loaded microspheres (Biocompatibles UK, Surrey, UK). Treatments could be repeated at intervals of 1 month depending on the total amount of liver involved to allow treatment of all lesions. No concomitant chemotherapy or immunotherapy was allowed. The saline suspension of slow-release microspheres was removed and the microspheres were mixed with an IRI solution at a dose of 100 mg per 2 ml at least 4 hours before the procedure to permit drug adsorption.

Digital subtraction angiography (DSA) was performed transfemoral. The study performed was extended to the abdominal aorta to assess hepatic circulation and blood supply to the tumor, with subsequent therapeutic planning. Following insertion of a 5-F Cobra or Simmons catheter, a 3-F micro-catheter (Renegade Hi-Flo, Boston Scientific, MEDI-TECH, USA) was then placed and selective hepatic catheterization of lobar or segmental branches afferent to the metastases was performed in all patients.

Irinotecan preloaded slow-release microspheres infusion was preceded by selective intra-arterial infusion of 5 ml of 1% lidocaine to reduce the local pain, and verapamil 5 mg was delivered to induce arterial vasodilatation and to prevent vasospam related to the contact of the drug with the endothelium of the vessel.

Subsequently, for every TACE, 100 mg IRI preloaded in 2–4 ml beads of 100–300 µm was administered to the first 10 patients. Then 200 mg IRI preloaded beads were administered to all other patients. We observed in the first group of patients a high percentage of response. Due to low toxicity we decided to increase the dose of IRI, hoping to obtain longer duration of responses and survival.

Study Endpoints and Statistical Design

The primary objective of this study was the response rate defined as the percentage of partial responses (PR) plus complete response (CR), calculated on an intention-to-treat basis. Secondary endpoints were progression-free survival, overall survival, and safety. The best overall response from each patient was reported. All results were reviewed by an independent radiologist. Progression-free survival (PFS) was calculated from the first day of treatment to first evidence of progression or death. Overall survival (OS) was calculated from first day of treatment until death of any cause. Both OS and PFS were estimated using the Kaplan–Meyer method.

The mini-max two-stage sequential design described by Simon (1989) was used to determine the number of patients to be included. We assumed a 35% chance of response, obtained by the most active treatments evaluated in metastatic melanoma. According to this, a response rate of 40% would be considered promising. The design parameters p0 (response rate in null hypothesis) and p1 (response rate in alternative hypothesis) selected were 0.20 and 0.40. Also considering an alpha error of 0.05 and 0.20, respectively, the first stage of the study required 28 patients and, if at least seven responses were observed, an additional 22 patients had to be enrolled in the second step of the study. The treatment was considered interesting for further investigation if more than 22 responses were observed.

RESULTS

Between January 2007 and February 2010, 52 patients presenting LM from UM with a median age of 65 (range 37–86) were enrolled into this study. Twenty-five males and 27 females, median age 65 (37–86) were evaluated and treated. The percentage of liver substitution was up to 25% in 30/52 (57.6%), up to 50% in 20/52 (38.4%) cases, and more than 50% in 4/52 (7.6%) cases (Table 1).

A total of 85 TACE procedures were performed and we obtained 100% technical success of the TACEs performed. All the 52 patients received at least one procedure and were available for responses. The median number of treatments per patient was 1.6 (range 1–3). Treatment was well tolerated and toxicities were mostly of grade 1 or 2 and easily manageable. No complications due to the procedure occurred. Among the most frequent adverse events, we recorded abdominal pain (often in the right-upper quadrant) and nausea and vomiting. Serious adverse events were extremely rare. Only two cases of acute chemical cholecystitis have been reported and spontaneously resolved. No neutropenic fever and no grade 5 adverse events were observed. We observed one evidence of transient paralytic ileus lasting 4 days, and two cases of increase of transaminases (Nx 4), regressed in 7 and 12 days, respectively.

No alopecia was related to treatment. A complete report of toxicities recorded is shown in Table 2. We used IRI 100 mg in the first group of 10 patients then, after observing only limited toxicity, we adopted IRI 200 mg.

Clinical evaluation was performed before the procedure and 1 month after TACE by means of contrast-enhanced multidetector computed tomography (MDTC). The responses were measured following RECIST criteria. All the 52 patients were available for responses and progression-free survival. We observed 45 objective responses. Five patients obtained a stabilization. Following RECIST criteria, 17 patients had a complete or near-complete response, with evidences of necrosis and reduction of contrast enhancement >90%, 30 patients had a reduction of >80% and three patients presented a minor reduction of 60%. Progression outside the liver was a very rare event.

All the patients were requested to fill in an ESAS questionnaire to assess their quality of life. The QoL assessment has been done after 1, 3, and 6 months from TACE. A clear benefit of QoL of more than 50% in 40 patients. The remaining patients experienced no changes. After a median follow-up of 23.5 months, median PFS was 7.5 months (95%, CI. 4.6–9.7). Observed median OS was one of the highest ever reported for this setting with 13.9 (95%, CI 9.2–16.8). Kaplan–Meyer curve for OS is shown in Fig. 1.

DISCUSSION

Liver is the most prevalent site of metastases in UM with liver failure being the most frequent cause of death.5 The prognosis of untreated patients with LM is poor and none of the systemic treatments used for cutaneous melanoma have shown activity in this peculiar type of melanoma, except some conflicting reports in small series. Given the predominant liver involvement of this disease, locoregional approaches have been investigated in the clinical setting.

Agarwala et al conducted a randomized phase I/II trial evaluating escalating doses of intrahepatic chemotherapy with cisplatin with or without gelatine sponge (PVS) in 19 patients with liver metastases from ocular melanoma and liver metastases.¹⁶ The cisplatin dose was initiated at 100 mg/m2 and was increased of 25. Patients were randomized to receive cisplatin alone or cisplatin plus PVS. Seven patients were treated with intrahepatic cisplatin at 100 mg/m2, four with embolic agent, and three without. The dose was escalated to 125 mg/m2 with or without embolic agent in the remaining 12 patients. The maximum tolerated dose for intrahepatic cisplatin was determined to be 125 mg/m2 with or without embolic agent. The overall response rate was 16%. Dose-limiting toxicities included renal, hepatic, and hematological effects. They concluded that this therapy produces an interesting response rate in patients with LM from UM. The toxicities was higher than in our study and suggests the limitations in the use of PVS respect new generation embolic agents.

Patel et al conducted a phase II clinical trial using chemoembolization with BCNU dissolved in ethiodized oil.¹⁷ Gelatin sponge particles were used as a transiently occlusive agent. Twenty-four patients out of 30 completed at least one treatment to all targeted liver metastases and were evaluated for hepatic response. Eighteen of these 24 patients experienced regression or stabilization of hepatic metastases for at least 6 weeks (1 CR in hepatic metastases, 4 PRs, 13 SDs). One of the 13 patients with SD was rendered free of disease by surgical removal of metastases after chemoembolization (surgical CR). The overall response rates (complete and partial responses) for intention-to-treat patients and for patients who were evaluated for response were 16.7% and 20.4%, respectively. The median overall survival of the entire intention-to-treat group of patients was 5.2 months (range 0.1–27.6 months), for the patients with complete or partial response in hepatic metastases 21.9 months (range 7.4–27.6 months), for the patients with stable disease 8.7 months (range 2.9–14.4 months), and for the patients with progressive disease 3.3 months (range 1.6–5.6 months). This paper reports 20% of responses and shows the possibility of benefits from a further surgical approach. On the contrary, there was no surgical removal of metastases because of the small number of our patients. We will continue to pursue this neo-adjuvant approach.

Thirteen out of the 18 patients who achieved CR, PR, or SD, subsequently developed progression of extrahepatic metastases with control of hepatic metastases. The authors conclude that chemoembolization with BCNU is a useful palliative treatment for the control of hepatic metastases in patients with UM. However, progression in extrahepatic sites after stabilization of hepatic metastases requires further improvement in the therapeutic approach to this disease.

Vogl et al treated 12 patients with liver metastases of UM with TACE.¹⁸ Six patients presented with solitary liver metastases (6–12 cm in size) and six patients with oligonodular metastases (n ≤ 6). The embolization suspension consisted of a maximum of 10 mg/m2 mitomycin C, 10 ml Lipiodol, and an injection of 200–450 mg resorbable microspheres for vascular occlusion. In the follow-up, magnetic resonance imaging was performed at 3-month intervals. They reported that TACE procedure was well tolerated in all patients without any relevant side effects. Three patients responded to TACE with a size reduction of more than 50% (PR), five patients with stable disease, and four patients with progressive disease with an increase in volume of more than 25%. Mean survival following primary tumor treatment was 32.9 months and after first embolization 19.5 months. Lower survival rates were recorded for the progressive group (16.5 months). They conclude that repeated TACE offers a palliative treatment option in patients with oligonodular liver metastases of uveal malignant melanoma. The TACE proposed has been performed with mitomycin that is not efficacious in UM. Furthermore, the mixture of lipiodol and resorbable microspheres seems to induce a short-lasting vascular occlusion. In this study, the rationale, the pharmacokinetics properties, and the actual drug uptake advantage of resorbable microspheres are not clearly reported.

Comparing our data with all the previous reports, our study introduces a new type of embolic agent preloaded with IRI, presenting more detailed and widely studied pharmacological aspects. ^25-29

In the present study, we report the first clinical evaluation of TACE-DEBIRI in patient with liver metastases from uveal melanoma in one of the larger series of patients. Updating our first report,³⁰ the primary endpoint was successfully reached, with a high percentage of responses that surely deserve clinical evaluation in further prospective randomized studies. We also reported more than 7 months of progression-free survival and a median survival exceeding 1 year. These are among the best results ever reported for this poor-prognosis disease, even in untreated patients. There are no comparable results of such magnitude in pretreated patients, which makes TACE-DEBIRI probably the most active treatment available in this setting. Even though there is the possibility that selection bias could be at least partially responsible for these very good results, we found no apparent characteristics that can explain these observations. Even if patients were pretreated and 21% of them had a poor Karnofsky score of less than 60%, toxicities were mild and well tolerated. No toxic deaths were observed and grade 3 or 4 events were extremely rare, with the exception of nausea, vomiting, and abdominal pain that were, however, well controlled by commonly used analgesic or antivomiting agents like tramadol and metoclopramide. Finally, quality of life was investigated in every patient and a significant improvement was seen in the vast majority of patients. We conclude that TACE-DEBIRI is active and safe for the treatment of liver metastases from uveal melanoma. Its use should be strongly considered in the case of symptomatic liver metastases, even in patients with a poor Karnofsky score or heavily pretreated, but must be confined to experienced centers. Further studies are needed to address the survival benefit that this treatment seems to provide.

Acknowledgements

Disclosure: The entire work was funded by the Azienda Unità Sanitaria Locale 11, Empoli, Italy; the Azienda Sanitaria Locale of Ferrara, Italy and the IRCCS Foundation, Padova, Italy

Conflict of interest statement: All the authors declare no conflict of interest

REFERENCES

1. Singh AD, Topham A. Incidence of uveal melanoma in the United States: 1973–1997. Ophthalmology. 2003;110:956–961.
2. Shields JA, Shields CL, De Potter P, Singh AD. Diagnosis and treatment of uveal melanoma. Semin Oncol. 1996;23:763–767.
3. Seregard S, Kock E. Prognostic indicators following enucleation for posterior uveal melanoma. Acta Ophthalmol Scand. 1995;73: 340–344.
4. McCartney A. Pathology of ocular melanoma. Br Med Bull. 1995;51: 678–693.
5. Becker JC, Terheyden P, Kampgen E, et al. Treatment of disseminated ocular melanoma with sequential fotemustine, interferon alpha, and interleukin 2. Br J Cancer. 2002;87:840–845.
6. Soni S, Lee DS, DiVito JJ, et al. Treatment of pediatric ocular melanoma with high-dose interleukin-2 and thalidomide. J Pediatr Hematol Oncol. 2002;24:488–491.
7. Bedikian AY, Legha SS, Mavligit G, et al. Treatment of uveal melanoma metastatic to the liver. Cancer. 1995;76:1665–1670.
8. Gragoudas ES, Egan KM, Seddon JM, et al. Survival of patients with metastases from uveal melanoma. Ophthalmology. 1991;98:383–390.
9. Pyrhonen S. The treatment of metastatic uveal melanoma. Eur J Cancer. 1998;34(suppl 3):S27–S30.
10. Nathan FE, Berd D, Sato T, et al. Interferon in the treatment of metastatic uveal melanoma: first report of active systemic therapy. J Exp Clin Cancer Res. 1997;16:201–208.
11. Fournier GA, Albert DM, Arrigg CA, Cohen AM, Lamping KA, Seddon JM. Resection of solitary metastasis. Approach to palliative treatment of hepatic involvement with choroidal melanoma. Arch Ophthalmol. 1984;102:80–82.
12. Gunduz K, Shields JA, Shields CL, Sato T, Mastrangelo MJ. Surgical removal of solitary hepatic metastasis from choroidal melanoma. Am J Ophthalmol. 1998;125:407–409.
13. Aoyama T, Mastrangelo MJ, Berd D, et al. Protracted survival after resection of metastatic uveal melanoma. Cancer. 2000;89:1561–1568.
14. Carrasco CH, Wallace S, Charnsangavej C, Papadopoulos NE, Patt YZ, Mavligit GM. Treatment of hepatic metastases in ocular melanoma. Embolization of the hepatic artery with polyvinyl sponge and cisplatin. JAMA. 1986;255:3152–3154.
15. Mavligit GM, Charnsangavej C, Carrasco CH, et al. Regression of ocular melanoma metastatic to the liver after hepatic arterial chemoembolization with cisplatin and polyvinyl sponge. JAMA. 1988;260:974–976.
16. Agarwala SS, Kirkwood JM. Phase I/II randomized trial of intrahepatic arterial infusion chemotherapy with cisplatin and chemoembolization with cisplatin and polyvinyl sponge in patients with ocular melanoma metastatic to the liver. Melanoma Res. 2004;14:217–222.
17. Patel K, Sullivan K, Berd D, et al. Chemoembolization of the hepatic artery with BCNU for metastatic uveal melanoma: results of a phase II study. Melanoma Res. 2005 Aug;15(4):297–304.
18. Vogl T, Eicheler K, Zangos S, et al. Preliminary experience with transarterial chemoembolization (TACE) in liver metastases of uveal malignant melanoma: local tumor control and survival. J Cancer Res Clin Oncol. 2007;133:177–184.
19. Dumez H, Awada A, Piccart M, et al. A phase I dose-finding clinical pharmacokinetic study of an oral formulation of irinotecan (CPT-11) administered for 5 days every 3 weeks in patients with advanced solid tumours. Ann Oncol. 2006;17(7):1158–1165.
20. Konno T. Targeting cancer chemotherapeutic agents by use of lipiodol contrast medium. Cancer. 1990;66:1897–1903.
21. Egawa H, Maki A, Mori K, et al. Effects of intra-arterial chemotherapy with a new lipophilic anticancer agent, estradiol-chlorambucil (KM2210), dissolved in lipiodol on experimental liver tumor in rats. J Surg Oncol. 1990;44:109–114.
22. Nakamura H, Hashimoto T, Oi H, Sawada S. Transcatheter oily chemoembolization of hepatocellular carcinoma. Radiology. 1989;170:783–786.
23. Sasaki Y, Imaoka S, Kasugai H, et al. A new approach to chemoembolization therapy for hepatoma using ethiodized oil, cisplatin, and gelatin sponge. Cancer. 1987;60:1194–1203.
24. Daniels JR, Sternlicht M, Daniels AM. Collagen chemoembolization: pharmacokinetics and tissue tolerance of cis-diamminedichloroplatinum( II) in porcine liver and rabbit kidney. Cancer Res. 1988;48:2446–2450.
25. Fiorentini G, Aliberti C, Turrisi G, et al. Intraarterial hepatic chemoembolization of liver metastases from colorectal cancer adopting irinotecan- eluting beads: results of a phase II clinical study. In Vivo. 2007;21(6):1085–1091.
26. Aliberti C, Benea G, Tilli M, Fiorentini G. Chemoembolization (TACE) of unresectable intrahepatic cholangiocarcinoma with slow-release doxorubicin-eluting beads: preliminary results. Cardiovasc Intervent Radiol. 2008;31(5):883–888.
27. Nekolaichuk C, Watanabe S, Beaumont C. The Edmonton symptom assessment system: a 15-year retrospective review of validation studies (1991–2006). Palliat Med. 2008;22(2):111–122.
28. Varela M, Real MI, Burrel M, et al. Chemoembolization of hepatocellular carcinoma with drug eluting beads: efficacy and doxorubicin pharmacokinetics. J Hepatol. 2007;46(3):474–481.
29. Taylor RR, Tang Y, Gonzales MV, Stratford PW, Lewis AL. Irinotecan drug-eluting beads for use in chemoembolization: in vitro and in vivo evaluation of drug release properties. Eur J Pharm Sci. 2007;30(1):7–14.
30. Fiorentini G, Aliberti C, Del Conte A, et al. Intra-arterial hepatic chemoembolization (TACE) of liver metastases from ocular melanoma with slow-release irinotecan-eluting beads. Early results of a phase II clinical study. In Vivo. 2009;23(1):131–137.
31. Simon R. Designs for efficient clinical trials Oncology (Williston Park). 1989 Jul;3(7):43–9; discussion 51–3. Review.

In the healthy intestine, estimates suggest that the natural microflora comprises approximately 15,000 to 36,000 bacterial species,¹ numbers more than 100 trillion and weighs an average of 1–2 kg.² A survey of bacterial genes suggests that the majority (60–90%) of the bacteria are from only two divisions (Bacteroidetes and Firmicutes).² However, many of the species existing in small numbers remain to be characterised.¹

Intestinal bacteria are not merely commensal. A key role for the intestinal microbiota is to suppress numbers and activity of species that would otherwise harm the human host.^3-5 This may explain why imbalances in the microbiota (dysbiosis) may play a role in the pathophysiology of some gastrointestinal disorders and why suppression of microbial numbers following systemic antibiotic treatment allows pathogenic bacteria to emerge in the intestine and/or increase production of toxins, resulting in antibiotic-associated diarrhoea (AAD).

Diarrhoea occurs in 5% to 39% of patients treated with antibiotics and is associated with the use of broad-spectrum antibiotics.⁶ Of these cases of diarrhoea, approximately 20–30% are caused by Clostridium difficile, ⁷ which is now a leading cause of nosocomial infection. AAD, and C. difficile-associated diarrhoea (CDAD) in particular, more frequently occur among the old, ^8-11 patients with Comorbidities,⁹ those who have had abdominal surgery¹⁰ and those with recent exposure to broad-spectrum antibiotics^10,11. Some antibiotics, like clindamycin and ampicillin predispose to CDAD, which may be explained by specific induction of colonising factors of C. difficile, such as three adhesins, Cwp66, the S-layer protein P47 and Fbp68, and by induction of a cysteine protease, Cwp84, which may facilitate translocation.¹²

Current prevention strategies include stopping or changing the precipitating antibiotic treatment.^13,14 Approximately three-quarters of patients respond to such measures, but the remainder may develop recurrent CDAD.¹⁵

C. difficile is an increasing problem for healthcare providers, impacting on patient safety and treatment costs.¹⁴ Rates of CDAD are increasing, as is the severity of these infections, with more patients at risk of sepsis.¹³ With this in mind, new treatments and strategies are being sought to avoid or treat AAD. Probiotics, that according to the World Health Organization definition are ‘live microorganisms which, when administered in adequate amounts, confer a health benefit on the host’,¹⁶ are now being considered as treatment and preventative measures for at-risk patients. Treatment with probiotics is rationalised on the basis that probiotics help maintain the healthy gut flora and/or by competitively inhibiting the growth of pathogens.¹⁴

There is clear evidence from a large number of (randomized, double-blind, controlled) clinical studies and several meta-analyses on beneficial effects of probiotics on AAD, including CDAD. ^17-26 However, published evidence that discriminates the different categories of evidence in relation to medical claims (according to the EU commission ²⁷ and the European Food Safety Authority (EFSA) guidance on gut and immune function 2011²⁸) is scarce.

This distinction makes understand apparent conflicts between negative opinions by EFSA on the evidence for beneficial health effects of probiotics and positive outcome in well conducted, published clinical trials. A recent example for such a conflict was a negative opinion by EFSA on the substantiation of a health claim related to fermented milk containing L. casei DN-11400 in patients receiving antibiotics pursuant to article 14 of Regulation (EC) No 1924/2006.⁸ This negative opinion was expressed in spite of statistically significant risk reduction of CDAD and was mainly based on lacking assessment of C. difficile (as risk factor) in those patients not suffering from diarrhoea. Accordingly the study design of clinical trials, supposed to demonstrate a beneficial effect, may depend on the goals differing between legislations.

In relation to AAD or CDAD, drug legislation covers prevention and treatment of diseases, whereby prevention is defined by reduction of the incidence of disease endpoints or reduction of recurrence of disease (secondary prevention) by prior intervention and treatment is defined by reduction of disease endpoints in overt disease (AAD or CDAD). By contrast, the health claims according to the health claim regulations for food, now categorised by the EFSA, include evidence for efficacy in “defence against pathogens of a specific site of the body” and reduction of disease risk (defined by “reduction of a risk factor for infection”).²⁸ This review aims to discuss the available evidence for probiotics in the light of this discrimination and the options addressed by recent EFSA guidance.²⁸ The focus in this review on the efficacy of probiotics in AAD is on level A evidence according to the generally applied recommendations of evidence-based guidelines;²⁹ this is usually provided by meta-analyses. Level B evidence, from some individual randomised, controlled trials is also reviewed.

Treatment of AAD (Including C. Difficile-Associated Diarrhoea) with Probiotics

Evidence for efficacy in the treatment of AAD or CDAD comprises the studies in which administration of probiotics has been started when diarrhoea occurred during or after administration of antibiotics and the endpoints were a reduction in severity or duration of AAD or CDAD.

Despite the above mentioned randomised, controlled trials and meta-analyses on beneficial effects of probiotics on AAD, including CDAD,^17-26 few meta-analyses have been published that, arguably, showed in a clear and methodologically acceptable manner that probiotics are effective in the treatment of AAD or CDAD, and this was also the consensus at a recent meeting of a group of experts.³⁰

A meta-analysis by McFarland¹⁷ discovered six randomized trials, which together showed that probiotic treatment (administration of probiotic for at least 3 wk during and after CDAD) had significant efficacy for reduction of recurrence of CDAD (RR=.59, 95% CI .41–.85, P=.005; Table 1). In this analysis, the primary outcome was diarrhoea, defined as ≥ 3 loose stools/d for at least 2 d or ≥ 5 loose stools/48 h within 2 mo of antibiotic exposure. This is consistent with the definition of diarrhoea given by the World Health Organization.³¹ The author categorized these trials as dedicated to treatment of CDAD. Of the six trials included in this meta-analysis, however, five reported on CDAD recurrences after specific anti-C. difficile treatment and thus tested the effect on secondary prophylaxis. One trial³² investigated the effect of administering probiotics or placebo within 72 h of prescribing antibiotics because of various infections. In this trial, which was stated to be underpowered by the authors (with n = 150 instead of n = 400), on the basis of development of diarrhoea, the incidence of samples positive for C. difficile-associated toxins was 2.9% in the probiotic group compared with 7.25% in the placebo-control group.

When samples from all patients were tested (rather than just those developing diarrhoea) 46% of probiotic patients were toxin-positive compared with 78% of the placebo group. Regarding the EFSA guidance, the design of this trial would fit to a risk reduction claim, but does not fulfil the definition of treatment in the strict regulatory sense.

Thus, presently no evidence from meta-analyses for an effect of the treatment by probiotics on CDAD exists.

Probiotics in the Prevention of AAD

Evidence for prevention of AAD comprise the studies in which probiotics have been administered before diarrhoea occurred (before or at the beginning of the administration of antibiotics). In this category of drug legislation, endpoints are incidence of AAD or CDAD. The meta-analyses providing evidence for prevention of AAD and CDAD are summarised in Table 2.

Whereas a number of meta-analyses of effects by probiotics on the prevention of AAD exist, there are few published randomized, double-blind, controlled trials and only three meta-analyses which looked at the prevention of CDAD.^13,20 The above mentioned meta-analysis¹⁷ is a third one which is serving to this target, although it was supposed to be dedicated to “treatment” of CDAD.

In the most recent analysis, the specific aims were to: 1. Assess whether probiotics, when administered with antibiotics, will reduce the incidence and severity of AAD and CDAD in an adult, hospitalized population. 2. Assess the incidence of adverse events when probiotics are co-administered with antibiotics in adults. Eight double-blind randomised, controlled trials were included in the analysis. The administration of probiotics led to a statistically significant relative risk reduction of 44% for the incidence of AAD and 71% for CDAD.¹³ This reduction in risk of CDAD due to antibiotic exposure was associated with an effect size (Z score) of 5.18, P ≤ .001. Four studies were included in this meta-analysis, which used probiotics comprising S. boulardii, several Lactobacillus spp. and Bifidobacterium bifidum. ^8,32-34 The definition of diarrhoea in this analysis was > 2 liquid stools/d for at least 3 d in excess of normal for each patient.

The other, earlier meta-analysis of prevention of AAD and CDAD with the probiotic Saccharomyces boulardii in adults and children comprised 1076 participants²⁰ and included data from one study³⁵ on the incidence of CDAD in children following antibiotic treatment. The definition of diarrhoea in this study was ≥ 3 loose or watery stools per day for a minimum of 48 h. In this study, the addition of S. boulardii as an adjunct to antibiotic therapy reduced the risk of AAD compared with those given placebo (3.4% of children compared with 17.3% developed ADD). Most probably due to limited power, this study found that the difference was of borderline significance, although the relative risk of C. difficile was lower in the probiotic treatment group compared with placebo (.3; 95% confidence intervals, .10–1.04).

Another eight meta-analyses (Table 2) also found that probiotics appeared to prevent AAD.^17,19,21-26 Some of these meta-analyses reviewed data concerning a single probiotic species (most usually S. boulardii, ^20,24 Lactobacillus rhamnosus GG²¹ or other Lactobacillus strains^22,26), others took a broader approach, analyzing relevant data from a wide range of probiotic species.^17,19,23,25

The definition of diarrhoea varied in the trials included in these eight meta-analyses; the most commonly used definition was the presence of three or more loose stools (either for 24 or 48 h).^17,19,21-26

One meta-analysis found that probiotics were not associated with a statistically significant preventive effect on antibiotic-associated diarrhoea in children, when using extreme-case assumptions for imputation of missing data for intention-to-treat analysis (P=.99).³⁶ This meta-analysis analysed two additional trials not incorporated in previous meta-analyses. The per-protocol analysis showed a significant difference in favour of probiotic over placebo (P=.003) but the data showed substantial heterogeneity (Table 2).

One systematic review of a specific strain (L. rhamnosus GG) ruled out a meta-analysis because of heterogeneity.²¹ The heterogeneity was due to the data from one trial.³⁷ This trial differed from the other trials in that all the patients in the study were inpatient (and on intravenous antibiotics rather than oral antibiotics) and the study was negative. The authors suggested that the participants in this study may have had a greater severity of infection at enrolment, and that such serious infections do not benefit from Lactobacillus rhamnosus GG therapy to the same degree as milder infections.

Thus overall there is evidence for a preventive effect of probiotics on AAD and CDAD.

Probiotics in the Reduction of Risk of AAD as Defined by the EFSA Based on C. Difficile as Risk Factor

According to the EFSA-guidelines, health claims for nutritional purposes, which address a risk reduction of disease, may only be claimed thief there is evidence for a reduction of a risk factor (“risk reduction claims”).²⁷ Therefore, evidence for the reduction of risk of AAD comprises the studies in which probiotics have been administered with the aim of reducing C. difficile, which is defined as a risk factor for diarrhoea.²⁸ In this EFSA category (EFSA guidance 2011, article 3.4.),²⁸ endpoints are supposed to be C. difficile or its toxin in all subjects studied independent of whether the subjects experienced diarrhoea.

In the most recent meta-analysis of four randomised controlled studies (471 participants) on AAD, in which the presence of C. difficile toxin A and B in stool samples was measured, probiotics reduced the risk of diarrhoea associated with C. difficile toxins significantly by 71% (P ≤.001) compared with placebo (Table 3).¹³

In an earlier meta-analysis published in 2002, the combined efficacy demonstrated a significant protective effect of probiotics against C. difficile associated diarrhoea (Table 3).¹⁷ In this meta-analysis, the endpoint was defined as a new episode of diarrhoea associated with a positive C. difficile culture or a positive C. difficile toxin A or B assay within 1 mo of exposure to antibiotics. Consequently, the outcome for risk reduction was a new episode of C. difficile positive diarrhoea within 1 mo of a previous episode of C. difficile associated diarrhoea.

In another meta-analysis²⁰ only one study (a paediatric trial)³⁵ investigated the effect of probiotics in the prevention of C. difficile diarrhoea (Table 3). Although the risk of C. difficile diarrhoea was lower in the probiotic group compared with the placebo group, the difference was of borderline significance (not significant).

None of these meta-analyses differentiated between studies in which C. difficile or its toxins were assessed in all the subjects taking antibiotics and studies in which C. difficile or its toxins were assessed only in those experiencing diarrhoea. Thus, there is insufficient evidence from meta-analyses to determine the effect of probiotics on the risk of AAD as defined by EFSA, although there is evidence on reduction of the risk of AAD and CDAD as defined by the more valid disease endpoints.

The few single trials in which C. difficile or its toxins were assessed in all the subjects taking antibiotics showed inconsistent results (Table 3). Plummer et al.³² found a non-significant reduction of C. difficile toxin (46% of the participants in the group receiving L. acidophilus and B. bifidum versus 78% in the control group were tested positive for the toxin). De Vrese et al.³⁸ found no increase of C. difficile colony forming units in faeces in the group receiving B. lactis BB-12 and L. acidophilus LA-5 containing yoghurt whereas there was an increase in the group receiving the acidified milk. A statistical comparison was not done due to the low number of positive C. difficile samples. Another group of investigators³⁴ also found no difference of toxin detection between the group administered S. boulardii and the placebo group.

Probiotics Improving Defence Against Gastrointestinal Pathogens

In this EFSA category (EFSA guidance 2011, article 3.3.),²⁸ defence against pathogens is thought to be a physiological function of the body, of which endpoints are gastro-intestinal infections. These endpoints can be measured as the number and/or severity or duration of episodes of antibiotics—or CDAD, not necessarily by demonstration of a decrease in pathogen concentration.

In case of the endpoint is the number of episodes this, evidently, is identical with the endpoints addressed in “probiotics in prevention of AAD” (see above and Table 2).

Concerning the effect of probiotics on severity and duration of infection, the EFSA guidance does not specify whether this applies to probiotic administration before onset of infection or to administration at the beginning or during infection (AAD). The latter case is evidently identical with the endpoints addressed in “treatment of AAD with probiotics” (see Table 1).

CONCLUSION

This review summarises data from clinical studies (with a focus on meta-analyses) on effects of probiotics on the gastrointestinal tract when antibiotics are given. These effects were reviewed following the categories for medical claims (according to drug legislation) and health claims for foods according to the EU commission²⁷ and the European Food Safety Authority (EFSA) guidance on gut and immune function 2011.²⁸

Regarding the evidence for efficacy of probiotics in the treatment of AAD, these comprise the studies in which probiotics were administered when diarrhoea occurred during or after administration of antibiotics. A meta-analysis of six randomised, controlled trials¹⁷ found that probiotics had a significant effect in the treatment of AAD. Five of the six trials included in this meta-analysis, however, were not dedicated to treatment in the above defined sense, but reported on a reduction of recurrence of CDAD in the probiotic group. Thus, presently, evidence from meta-analyses for an effect of ADD or CDAD treatment by probiotics is lacking.

For the prevention of AAD, studies in which probiotics were administered before diarrhoea occurred (before or at the beginning of the administration of antibiotics) were analysed. The majority of meta-analyses, and particularly the most recent, found a significantly beneficial effect of probiotics.^17,19,21-26 There were few studies specifically investigating the effect on prevention of CDAD but, nevertheless, one meta-analysis of these studies¹³ found a highly significant reduction in the incidence of CDAD.

For the reduction in risk of diarrhoea (the studies in which probiotics have been administered with the aim of reducing C. difficile, which is defined as a risk factor for diarrhoea), there are no meta-analyses, because most of the studies were not designed to consider this discrimination, in which the pathogen (defined as risk factor) was investigated in all subjects receiving antibiotics, and those, in which C. difficile was only assessed in subjects experiencing AAD. Thus although there is clear evidence in risk reduction (prevention) as defined by relative risk (RR) or odds ratio of the incidence of AAD or CDAD, there is not sufficient evidence on C. difficile appearance defined as risk factor for infection.

There is clear evidence of the efficacy of probiotics on gastrointestinal infections in general, and AAD and CDAD in particular, and thus on “improving defence against pathogens of the gastrointestinal tract”.

The current drawbacks of the data include the relative lack of studies (and meta-analyses), the heterogeneity of the data, including the definitions of diarrhoea used, the wide variety of probiotic agents in single meta-analyses (which may increase the heterogeneity of the data), the different dosages and formulations of probiotics, timing and length of treatment, and heterogeneity of populations studied. Due to the strain specificity in mediating effects one may expect higher effect size for some than found in meta-analyses of studies including all the strains ever assessed in studies with appropriate design. In this respect meta-analyses of strain-specific effects would give a more precise notion on the attainable effect size. Despite the fact that the probiotics used in the studies reviewed include a wide array of species and strains, each of which may have different immunological and physiological effects in disease states, there is statistically significant evidence for those effects for which more than four randomised, controlled trials have been available for meta-analysis: prevention of AAD and CDAD, and enhancing defence against pathogens of the gastrointestinal tract. In case of treatment of AAD and CDAD and the risk reduction claims, based on C. difficile as risk factor, studies are still needed for enabling assessment of the evidence on a sound basis.

With the pressure on reducing treatment and hospitalisation costs unlikely to end in the near future, together with the spectre of increasing levels of antibiotic resistance, the use of probiotics to reduce the incidence and attenuate the effects of diarrhoea remains attractive. Clostridium difficile infection is increasingly prevalent, can cause morbidity and in some instances, mortality. It is also problematic for healthcare providers, because CDAD increases the length of hospital stays with subsequent impact on costs. Probiotics may represent a low-cost, low-risk preventive measure that appears to have the potential to provide considerable benefits by reducing AAD by approximately 43% and CDAD by approximately 71%. Nevertheless, further studies may be needed to demonstrate more clearly which strains, when and in which patients, prevention is most effective.

Acknowledgements

JS received fees for consulting Danone, Yakult, Dr. Fischer Health Care, Campina, Infectopharm, Merck, funds and grants for research from Campina, Chr. Hansen. Danisco, Danone, Merck, MONA, Morinaga, Mueller, Nestle, NÖM, Wakunaga and Yakult for studies on probiotics, and fees as a speaker at symposia sponsored by Yakult, Danone, Nestle, Orthomol, and Merck. MV had received funds for research from Campina, Chr. Hansen, Danisco, Danone, Merck, MONA, Morinaga, Mueller, and NÖM.

REFERENCES

1. Frank DN, St Amand AL, Feldman RA, Boedeker EC, Harpaz N, Pace NR. Molecular-phylogenetic characterization of microbial community imbalances in human inflammatory bowel diseases. Proc Natl Acad Sci USA. 2007;104:13780–13785.
2. Neish AS. Microbes in gastrointestinal health and disease. Gastroenterology. 2009;136:65–80.
3. Ismail AS, Hooper LV. Epithelial cells and their neighbors. Iv. Bacterial contributions to intestinal epithelial barrier integrity. Am J Physiol Gastrointest Liver Physiol. 2005;289:779–784.
4. Kaper JB, Sperandio V. Bacterial cell-to-cell signaling in the gastrointestinal tract. Infect Immun. 2005;73:3197–3209.
5. Lawley TD, Clare S, Walker AW, et al. Antibiotic treatment of Clostridium difficile carrier mice triggers a supershedder state, spore-mediated transmission, and severe disease in immunocompromised hosts. Infect Immun. 2009;77:3661–3669.
6. McFarland LV. Epidemiology, risk factors and treatments for antibioticassociated diarrhea. Dig Dis. 1998;16:292–307.
7. McFarland LV. Diarrhoea associated with antibiotic use. BMJ. 2007;335:54–55.
8. Hickson M, D’Souza AL, Muthu N, et al. Use of probiotic lactobacillus preparation to prevent diarrhoea associated with antibiotics: randomised double blind placebo controlled trial. BMJ. 2007;335:80.
9. Bauer MP, Notermans DW, van Benthem BH, et al. Clostridium difficile infection in europe: a hospital-based survey. Lancet. 2011;377:63–73.
10. Kent KC, Rubin MS, Wroblewski L, Hanff PA, Silen W. The impact of Clostridium difficile on a surgical service: a prospective study of 374 patients. Ann Surg. 1998;227:296–301.
11. Sansone S, Aschbacher R, Staffler M, et al. Nosocomial diarrhoea in adult medical patients: the role of Clostridium difficile in a north italian acute care teaching hospital. J Prev Med Hyg. 2009;50:117–120.
12. Deneve C, Delomenie C, Barc MC, Collignon A, Janoir C. Antibiotics involved in Clostridium difficile-associated disease increase colonization factor gene expression. J Med Microbiol. 2008;57:732–738.
13. Avadhani A, Miley H. Probiotics for prevention of antibiotic-associated diarrhea and Clostridium difficile-associated disease in hospitalized adults—a meta-analysis. J Am Acad Nurse Pract. 2011;23:269–274.
14. Hickson M. Probiotics in the prevention of antibiotic-associated diarrhoea and Clostridium difficile infection. Therap Adv Gastroenterol. 2011;4:185–197.
15. Bartlett JG. Clinical practice. Antibiotic-associated diarrhea. N Engl J Med. 2002;346:334–339.
16. Guidelines for the evaluation of probiotics in food. In Edition London, Ontario: Joint FAO/WHO Working Group on Drafting Guidelines for the Evaluation of Probiotics in Food, World Health Organization, 2002.
17. McFarland LV. Meta-analysis of probiotics for the prevention of antibiotic associated diarrhea and the treatment of Clostridium difficile disease. Am J Gastroenterol. 2006;101:812–822.
18. Szajewska H, Skorka A, Dylag M. Meta-analysis: Saccharomyces boulardii for treating acute diarrhoea in children. Aliment Pharmacol Ther. 2007;25:257–264.
19. Szajewska H, Ruszczynski M, Radzikowski A. Probiotics in the prevention of antibiotic-associated diarrhea in children: a meta-analysis of randomized controlled trials. J Pediatr. 2006;149:367–372.
20. Szajewska H, Mrukowicz J. Meta-analysis: non-pathogenic yeast saccharomyces boulardii in the prevention of antibiotic-associated diarrhoea. Aliment Pharmacol Ther. 2005;22:365–372.
21. Hawrelak JA, Whitten DL, Myers SP. Is lactobacillus rhamnosus gg effective in preventing the onset of antibiotic-associated diarrhoea: a systematic review. Digestion. 2005;72:51–56.
22. Cremonini F, Di Caro S, Nista EC, et al. Meta-analysis: the effect of probiotic administration on antibiotic-associated diarrhoea. Aliment Pharmacol Ther. 2002;16:1461–1467.
23. D’Souza AL, Rajkumar C, Cooke J, Bulpitt CJ. Probiotics in prevention of antibiotic associated diarrhoea: meta-analysis. BMJ. 2002;324:1361.
24. McFarland LV. Systematic review and meta-analysis of saccharomyces boulardii in adult patients. World J Gastroenterol. 2010;16:2202–2222.
25. Sazawal S, Hiremath G, Dhingra U, Malik P, Deb S, Black RE. Efficacy of probiotics in prevention of acute diarrhoea: a meta-analysis of masked, randomised, placebo-controlled trials. Lancet Infect Dis. 2006; 6:374–382.
26. Kale-Pradhan PB, Jassal HK, Wilhelm SM. Role of lactobacillus in the prevention of antibiotic-associated diarrhea: a meta-analysis. Pharmacotherapy. 2010;30:119–126.
27. Anon. Regulation (ec) no 1924/2006 of the European Parliament and of the Council. Official Journal of the European Union. 2006;404:9–25.
28. Anon. Guidance on the scientific requirements for health claims related to gut and immune function. EFSA Journal. 2011;9:1984.
29. Harbour R, Miller J. A new system for grading recommendations in evidence based guidelines. BMJ. 2001;323:334–336.
30. Rowland I, Capurso L, Collins K, et al. Current level of consensus on probiotic science—report of an expert meeting—London, 23 November 2009. Gut Microbes. 2010;1:436–439.
31. Health topics: Diarrhoea. In Organization, WH (ed) Edition 2008.
32. Plummer S, Weaver MA, Harris JC, Dee P, Hunter J. Clostridium difficile pilot study: effects of probiotic supplementation on the incidence of C. difficile diarrhoea. Int Microbiol. 2004;7:59–62.
33. Can M, Besirbellioglu BA, Avci IY, Beker CM, Pahsa A. Prophylactic saccharomyces boulardii in the prevention of antibiotic-associated diarrhea: a prospective study. Med Sci Monit. 2006;12:I19–22.
34. Lewis SJ, Potts LF, Barry RE. The lack of therapeutic effect of saccharomyces boulardii in the prevention of antibiotic-related diarrhoea in elderly patients. J Infect. 1998;36:171–174.
35. Kotowska M, Albrecht P, Szajewska H. Saccharomyces boulardii in the prevention of antibiotic-associated diarrhoea in children: a randomized double-blind placebo-controlled trial. Aliment Pharmacol Ther. 2005; 21:583–590.
36. Johnston BC, Supina AL, Vohra S. Probiotics for pediatric antibioticassociated diarrhea: a meta-analysis of randomized placebo-controlled trials. CMAJ. 2006;175:377–383.
37. Thomas M, Litin S, Osmon D, Corr A, Weaver A, Lohse C. Lack of effect of lactobacillus gg on antibiotic-associated diarrhea: a randomized, placebo-controlled trial. Mayo Clin Proc. 2001;76:883–889.
38. de Vrese M, Kristen H, Rautenberg P, Laue C, Schrezenmeir J. Probiotic lactobacilli and bifidobacteria in a fermented milk product with added fruit preparation reduced antibiotic associated diarrhoea and helicobacter pylori activity. J Dairy Res. 2011;78:396–403.