Journal Logo

Article

Familial Mediterranean Fever and Cryptogenic Cirrhosis

Tweezer-Zaks, Nurit MD; Doron-Libner, Anat MD; Weiss, Perez MD; Ben-Horin, Shomron MD; Barshack, Iris MD; Lidar, Merav MD; Livneh, Avi MD

Author Information
doi: 10.1097/MD.0b013e31815be056
  • Free

Abstract

INTRODUCTION

Familial Mediterranean fever (FMF) is an autosomal recessive, ethnically related disease affecting Sephardic Jews, Armenians, Turks, Arabs, Druzes, and other populations of a Mediterranean origin44. The disease is characterized by acute episodes of fever and severe pain usually located in the abdomen, chest, joints, skin, and muscles44. The gene causing FMF, MEditerranean FeVer (MEFV), was mapped to the short arm of chromosome 16. It encodes a 781 amino acid protein, entitled pyrin/marenostrin, which is mainly expressed in the cytoplasm of mature neutrophils and monocytes and is thought to regulate neutrophil-mediated inflammation40. To date, at least 50 disease-associated mutations in the MEFV gene have been described (http://fmf.igh.cnrs.fr)45. Yet, in about 30% of FMF patients only 1 or none of the mutations can be found, suggesting a different genetic setting in these patients41,48. Thus, the diagnosis of FMF is still determined clinically, and genetic analysis remains an ancillary tool27,28.

Amyloidosis of the amyloid A (AA) type is the gravest manifestation of FMF, affecting many organs, including the kidneys, adrenal glands, intestines, spleen, liver, thyroid, heart, and lungs16. Kidney involvement is usually the first to be noted and the most significant. It presents as proteinuria, progressing successively through phases of nephrotic syndrome, azotemia, and uremia to end-stage renal failure within about 10 years. Treatment with colchicine, when initiated in time, prevents both disease attacks and amyloidosis13,26,49. In general, amyloidosis of the liver and the gastrointestinal (GI) tract becomes clinically overt only many years after the more advanced stages of kidney disease have been reached. Clinical manifestations include hepatosplenomegaly, mild disturbances in the liver function tests, diarrhea, malabsorption, and weight loss32. Clinically overt amyloid liver disease without amyloid nephropathy has not been reported in patients with FMF.

Nonamyloid chronic liver disease, either as part of the spectrum of FMF manifestations or as an unrelated illness, has not been systematically studied in FMF, as far as we know. The current report aims to evaluate the possible association, rare as it may be, between FMF and cryptogenic cirrhosis.

PATIENTS AND METHODS

Setting

The National Center for FMF at the Sheba Medical Center, Tel-Hashomer, Israel, holds a computerized registry of most FMF patients in Israel. At the time of the study, approximately 6000 FMF patients were enrolled in the registry. One-third of this population is being followed annually, while the rest are examined once every 4 years on average. Data are updated routinely at each clinic visit, using a standardized questionnaire.

Study Population

We conducted a wide computerized search in our registry looking for terms related to liver abnormalities, including elevated transaminases, hepatomegaly, splenomegaly, ascites, hepatitis, altered liver function, and cirrhosis. The clinical, histologic, and laboratory parameters of all patients detected were retrieved from the outpatient clinic and inhospital patient records, as well as from the referral letters of the primary care physicians and notes of the gastroenterology/hepatology consultants. In selected cases a revision of the histopathology specimens was performed. We used the METAVIR3 scoring system to grade the stage of fibrosis and the degree of inflammation, and the Child-Pugh36 classification to grade the severity of liver disease. FMF genetic diagnosis was based on direct determination of the 5 most common mutations of MEFV in our population, including M694V, V726A, E148Q, M694I, and M680I2. FMF disease severity score was calculated according to the Mor criteria33. Cumulative colchicine levels were calculated by multiplying daily dose by the treatment duration. Average daily colchicine dose was calculated by dividing the cumulative dose by the total number of treatment days. The ethical committee of the Sheba Medical Center approved the study.

Inclusion and Exclusion Criteria

Patients were included if 1) they met the widely accepted established clinical criteria for diagnosis of FMF27, whether or not MEFV mutations were looked for or found; 2) they were diagnosed by their gastroenterologist as suffering from cryptogenic cirrhosis after an extensive workup, as detailed below.

Cirrhosis was determined histologically. In patients declining liver biopsy, or when liver biopsy was deemed unnecessary by the gastroenterologist, the presence of cirrhosis was established clinically, based on findings suggestive of portal hypertension on gastroscopy and/or liver scan, coinciding with cirrhotic appearance of liver in ultrasonography (US) or computerized tomography (CT).

All patients receiving a diagnosis of cryptogenic cirrhosis were evaluated at the hepatology clinic; cirrhosis was designated as cryptogenic following exclusion of known etiologies. Specifically, this workup included the following:

  1. Serology for hepatitis B and hepatitis C viruses (HBV and HCV). The tests used for screening were hepatitis B surface antigen (HBsAg) and anti-HCV in all patients. Anti-hepatitis B core (HBc) was additionally determined in 3 patients.
  2. Laboratory studies for metabolic causes, investigated by testing for serum levels of ceruloplasmin, 24-hour urinary copper secretion, iron, ferritin, levels of transferrin saturation, and of alpha-1 antitrypsin.
  3. Serology for autoimmune hepatitis, determined by the presence of abnormal levels of antinuclear antibody, antismooth muscle antibody, and/or anti-liver-kidney microsomal antibody9.
  4. Studies directed at detection of nonalcoholic steatohepatitis, including liver imaging, glucose, triglyceride and cholesterol serum levels, and presence of obesity.
  5. History supporting hepatotoxic damage,including alcohol abuse and certain drugs or substance exposure.

This extensive evaluation will be subsequently referred to as the workup for known causes of cirrhosis.

Patients with known secondary systemic AA amyloidosis, whether determined by Congo red staining of tissue specimen or by evidence of proteinuria plus kidney dysfunction, were excluded.

Statistical Analysis

All the comparisons performed in the study were analyzed using the 2-tailed chi square test or the Fisher exact test according to the size of the population examined. Findings were considered significant when p values were 0.05 or lower.

RESULTS

Seventy patients with liver and splenic disease or abnormalities were identified by the computerized search. The medical records of all these patients were retrieved and pertinent data were re-evaluated, yielding a variety of liver disorders (Table 1). Cirrhosis (per biopsy) was present in 16 of the 70 patients. Of these, in 1 patient the cirrhosis was related to autoimmune hepatitis, and in 6 it was combined with amyloidosis. Cryptogenic cirrhosis was diagnosed in the remaining 9 patients after exclusion of possible viral, metabolic, toxic, amyloid, or autoimmune causes for liver disease, as described in the methods section. All of the other 54 patients were excluded because their findings did not agree with the inclusion criteria (see Table 1).

TABLE 1
TABLE 1:
FMF Patients With Liver Disease or Splenomegaly

Table 2 summarizes the FMF characteristics of the 9 patients with cryptogenic cirrhosis comprising the study cohort. The onset of FMF occurred relatively late in childhood in most patients (mean age, 17.2 ± 10.4 yr). Five of the 7 patients with genotype available were homozygous for the M694V mutation, the mutation associated with higher disease severity. Most patients had multisite involvement of the FMF attacks, a further marker of a more severe disease. Nevertheless, only 4 patients could be formally defined as having severe FMF by the Mor severity score. The mean colchicine daily dose was 1.4 ± 0.4 mg, and the mean cumulative colchicine dose before the first appearance of liver symptoms was 8.2 ± 6.3 g.

TABLE 2
TABLE 2:
FMF Characteristics of Patients With Cryptogenic Cirrhosis

The main features of cirrhosis in these patients are summarized in Table 3. Child-Pugh classification (based on serum albumin and bilirubin levels, prothrombin time, and the presence of ascites and hepatic encephalopathy) was documented in 6 patients at the time of cirrhosis diagnosis, and was classified as A in 4 of the 6. In the remaining 3 patients the score was A as well, but was based on data from a later stage of the disease (less than 5 years from diagnosis in all). Most patients had episodes of upper GI bleeding from gastroesophageal varices. Liver biopsies were performed in 5 cases, and the pathology findings are described in Table 4. The medical history of each of the 9 patients is summarized briefly below.

TABLE 3
TABLE 3:
Liver Abnormalities in Patients With Cryptogenic Cirrhosis
TABLE 4
TABLE 4:
Histopathologic Characteristics of Liver Biopsies*

CASE REPORTS

Case 1

This 48-year-old man of Jewish-Libyan origin suffered from FMF since early childhood. Over the years he experienced episodes of peritonitis and erysipelas-like erythema, but colchicine therapy was initiated, upon clinical diagnosis, only at the age of 34 years. Despite this delay, he had no proteinuria and the serum creatinine level was normal. Rectal biopsy was negative for amyloid. MEFV mutation analysis showed homozygosity to M694V. At the age of 44 years, pancytopenia, splenomegaly, and mildly elevated liver enzymes were noted. Abdominal US revealed a nodular liver of normal size. The liver size by the isotope scan was 12 cm. CT scan showed an enlarged spleen and widened portal vein. Gastroscopy revealed grade 3 esophageal varices. The patient refused liver biopsy. The workup for known causes of cirrhosis was negative, and cryptogenic cirrhosis was diagnosed. During the 4 years of follow-up, the patient's condition remained stable, under treatment with propranolol and colchicine.

Case 2

A 67-year-old woman of Jewish-Moroccan origin had symptoms compatible with FMF since she was 20 years old, but FMF was formally diagnosed when she was aged 39 years. Over the years she suffered episodes of peritonitis, pleuritis, arthritis, erysipelas-like erythema, and myositis. Her disease-associated genotype was found to be M694V/M694V. Colchicine treatment, 1 mg/d, was started at the age of 43 years and successfully prevented her attacks. She had no proteinuria, and kidney function remained normal. Her medical history included chronic ischemic heart disease with stable angina pectoris and mild chronic obstructive lung disease.

At the age of 63 years, pancytopenia and splenomegaly were noted. Abdominal CT scan showed a moderate amount of peritoneal fluid, normal liver texture, normal portal vein, and an enlarged spleen. Liver scintigraphy showed a nonhomogeneous liver pattern. Liver biopsy demonstrated chronic hepatitis with mild to moderate activity and moderate to severe bridging fibrosis. The workup for known causes of cirrhosis was negative, and cryptogenic cirrhosis was thus diagnosed. Diuretics, angiotensin-converting enzyme inhibitor, statin, H2 blocker, and calcium channel blocker were initiated and colchicine therapy continued. At last follow-up, the patient remained in stable condition under this treatment.

Case 3

In this man of Jewish-Tunisian origin, FMF first manifested during childhood and was diagnosed at the age of 16 years. He suffered episodes of peritonitis and protracted arthritis until colchicine treatment was initiated at the age of 19 years. FMF genotype was found to be M694V/M694V. At the age of 45 years he had a rectal biopsy performed due to proteinuria with normal renal function tests; it was negative for amyloid.

At the age of 46 years, gastroscopy, performed after the first episode of upper GI bleeding, revealed grade 2 esophageal varices and stage 3 congestive gastropathy. A huge spleen (21 cm) and an enlarged hyperechoic liver were found on US examination. He had pancytopenia, mildly elevated bilirubin, and abnormal coagulation tests. Liver enzymes were within the normal limits. Liver biopsy was consistent with cirrhosis, and Congo red stain for amyloid was negative. The workup for known causes of cirrhosis was negative, and the patient received a diagnosis of cryptogenic cirrhosis. At last follow-up his treatment consisted of propranolol, spironolactone, nitrates, and omeprazole, as well as colchicine, 2 mg/d. No other episodes of bleeding or clinical signs of progression were noted.

Case 4

This man of Jewish-Egyptian origin suffered from FMF attacks with peritonitis, pleuritis, and arthritis from the age of 8 years, but FMF was diagnosed only at the age of 27 years, and he commenced colchicine treatment at the age of 39 years. He had mild persistent proteinuria, but small and large bowel biopsies revealed no evidence of amyloidosis. A bone marrow biopsy (performed when he was aged 70 years for chronic anemia) was also negative for amyloidosis. His FMF genotype was M694V/M694V.

Elevated transaminases and bilirubin levels were noted when he was aged 70 years, and treatment with amiodarone, given for paroxysmal atrial fibrillation during the preceding 4 years, was discontinued. Abdominal US revealed an enlarged spleen, a small amount of peritoneal fluid, and a normal flow in portal vein. Gastroscopy demonstrated esophageal varices and gastric telangiectases. Liver biopsy confirmed the presence of active cirrhosis (chronic hepatitis with stage 4 fibrosis). No granulomas or steatohepatitis, typical for amiodarone-induced liver damage, were observed, and Congo red stain for amyloid was negative. The workup for known causes of cirrhosis was negative, and a diagnosis of cryptogenic cirrhosis was entertained. At last follow-up the patient's condition remained stable, and he was being treated for cirrhosis and FMF.

Case 5

This woman of Jewish-Iraqi origin presented at the age of 30 years with episodes of peritonitis, typical rash, and arthritis. She was diagnosed with seronegative polyarthritis as well as FMF at the age of 36 years, and started taking colchicine. Her MEFV genotype was found to be M694V/V726A. Rectal biopsy was negative for amyloid. Liver abnormalities were first noted at the age of 50 years, with mildly elevated liver enzymes.

During the following years, short episodes of transaminase elevations with spontaneous resolution were recorded, but were unrelated to the administration or discontinuation of her arthritis medications, administered intermittently and for a short duration (none exceeding 6 mo). These medications included sulphasalazine, minocycline, hydroxychloroquine, methotrexate, (cumulative dose of only 187.5 mg, 15 injections 12.5 mg each), cyclosporine, etanercept, and azathioprine.

When she was aged 59 years (under no arthritis therapy), pancytopenia appeared and prompted a full investigation. Abdominal CT showed an enlarged spleen and enlarged left liver lobe, with normal texture and a small amount of peritoneal fluid. A liver biopsy revealed chronic hepatitis grade 2-3 and stage 4 fibrosis. Fatty degeneration was observed in less than 10% of liver cells, and no specific histologic features suggestive of drug toxicity were found. Congo red and Prussian blue staining (for iron) were negative. The workup for known causes of cirrhosis was negative, and a diagnosis of cryptogenic cirrhosis was considered. At last follow-up the patient was being treated with colchicine, calcium channel blockers, omeprazole, and spironolactone, and her condition remained stable.

Case 6

An 80-year-old woman of Jewish-Syrian origin suffered from episodic fever and peritonitis from the age of 30 years, and FMF was diagnosed when she was aged 48 years. Her FMF genotype was V726A/M694V. She had no proteinuria, and kidney function was normal. Her medical history included type 2 diabetes and breast carcinoma, at the age of 70 years, treated with lumpectomy and radiotherapy.

When she was aged 75 years, chronic liver disease was diagnosed that manifested with weakness, edema, and ascites. Laboratory tests showed anemia, thrombocytopenia, and abnormal coagulation tests. On abdominal US, the liver appeared nodular but with normal size, the spleen was large (14.9 cm), and there was normal flow in the portal and hepatic veins. Liver and spleen radioisotope scan revealed hepatosplenomegaly, with filling defects in the liver. A shift of the radioactive material from the liver to bone marrow, compatible with cirrhosis, was detected. Rectal biopsy at that time was negative for amyloid. Gastroscopy revealed grade 1-5 gastroesophageal varices. Biopsy of the liver was not performed. Nevertheless, based on these findings and a negative workup for known causes of cirrhosis, cryptogenic cirrhosis was diagnosed.

Case 7

A 52-year-old man of Jewish-Turkish origin was diagnosed with FMF at the age of 27 years, 10 years after he experienced the first symptoms of the disease that included short, self-limited febrile episodes accompanied by arthritis. Over time, the arthritis became protracted. His FMF genotype was M694V/M694V. Colchicine treatment was started at the age of 27 years. He had no proteinuria, and kidney function was normal. Oral aphthae, prolonged episodes of headaches and fever, and epilepsy, for which he was given antiepileptic medications, supported the diagnosis of Behçet disease in addition to FMF.

At the age of 50 years, following an episode of upper GI bleeding, he underwent gastroscopy, in which grade 2 gastroesophageal varices were demonstrated. At that time, an enlarged spleen was detected, and laboratory tests showed anemia and thrombocytopenia. Liver biopsy was compatible with cirrhosis, with mild to moderate inflammatory activity. No amyloid or iron deposits were identified. The workup for known causes of cirrhosis was negative, supporting a diagnosis of cryptogenic cirrhosis. At last follow-up the patient was being treated with propranolol and low-dose colchicine and remained stable.

Case 8

This 83-year old Jewish-Syrian woman was diagnosed with FMF and started treatment with colchicine at the age of 58 years, some 30 years after the beginning of symptoms that included fever, peritonitis, pleuritis, arthritis, and erysipelas-like erythema. Genotype was not determined. She had no proteinuria and her kidney function was normal.

At the age of 60 years, she was hospitalized for evaluation of hepatosplenomegaly, elevation of transaminases and alkaline phosphatase, and pancytopenia. Abdominal US revealed enlarged liver and spleen and the presence of peritoneal fluid. However, cirrhosis was diagnosed only when she was aged 74 years, following an episode of upper GI bleeding and a gastroscopy revealing grade 3 varices, although liver biopsy was not performed. Because the workup for known causes of cirrhosis was negative the cirrhosis was labeled cryptogenic. The patient died at the age of 83 years of acute myocardial infarction after an episode of severe GI bleeding.

Case 9

This 70-year-old woman of Jewish-Moroccan origin suffered from episodes of fever, peritonitis, and pleuritis from the age of 12 years, and FMF was diagnosed when she was aged 20 years. Her FMF genotype is not known. Although the exact stage of colchicine therapy initiation was unclear, the treatment lasted many years. Her medical history was marked by hypertension and an episode of amaurosis fugax for which she was treated with propranolol and aspirin.

At the age of 58 years she was noted to have splenomegaly. Transaminases and alkaline phosphatase were mildly elevated. On US there was evidence of portal hypertension and large spleen. Abdominal CT demonstrated a small nodular liver, an enlarged spleen, a widened portal vein, and small (<8 cm) kidneys. Rectal biopsy was negative for amyloid. Liver biopsy was not done, and cirrhosis was diagnosed clinically. The workup for known causes of cirrhosis was negative, and the cirrhosis was therefore thought to be cryptogenic. When she was aged 67 years, esophageal varices were found after an episode of upper GI bleeding. One year later the patient died of cerebral infarct.

DISCUSSION

Among 6000 patients with FMF, we identified 16 patients with liver cirrhosis. Nine patients complied with the definition of cryptogenic cirrhosis (see Table 1). Since to our knowledge only 1 population-based study of the prevalence of cirrhosis was done in Israel, more than a decade ago31, it is not clear whether a prevalence of 0.27% (16/6000) for cirrhosis, or 0.15% (9/6000) for cryptogenic cirrhosis, in our cohort is higher than would be expected in the general Israeli population at present. However, in the previous study, performed to evaluate the need for liver transplantation in Israel, cirrhosis was found in 650 of 4.4 million individuals (0.015%) comprising the Israeli population at that time, which is significantly less than the prevalence of cirrhosis in the current cohort. Even if the study underestimates the need for liver transplantation by a factor of 2-315,30, the differences remain overwhelming (p < 0.0000, for the 2 comparisons). Furthermore, in several series7,39 cryptogenic cirrhosis accounts for 5%-30% of all causes of cirrhosis, a rate much lower than the rate in our cohort of 56% (9/16). Taken together, these data may suggest a role for FMF, or its treatment, in instigating this chronic liver disease.

Notably, some of the 10 patients with nephropathic amyloidosis plus chronic liver illness of unknown etiology who were excluded from this study (assuming that their chronic liver disease was amyloid related or due to lack of data), could in fact have been diagnosed as having cryptogenic cirrhosis. This might further increase the rate of cryptogenic cirrhosis in FMF.

To our knowledge, no association has been described yet between FMF and cirrhosis of the liver or any other chronic liver disease, unrelated to amyloid. Liver disease concomitant with FMF has been reported only rarely, including recurrent acute cryptogenic hepatitis in 2 children21,34, lone elevation of liver enzymes and bilirubin in a few patients29, dilatation of liver sinusoids12, fatty degeneration21, and hepatitis on liver biopsy, performed during laparotomy in FMF attacks in several other patients38. The questionable association of these reports to FMF has not been further studied or described. As far as we know, the current study is the first to address systematically the relation between FMF and chronic liver disease.

Could any feature in FMF explain this association? Phenotypically, there were no specific traits in the patients of our study cohort that could be considered risk factors for developing chronic liver disease. As in other unselected Israeli FMF cohorts, most of our patients originated from North Africa and Iraq (see Table 2). The mean age of FMF disease onset was before the age of 20 years, as is the case in most patients with FMF. Most patients had more than 1 site involved in FMF attacks, and the distribution of severity scores was not statistically different from the proportion of patients with severe disease in the randomly selected cohort of 100 patients that served to establish the severity scoring system by Mor et al33. All in all, it is unlikely that phenotypic factors of FMF, such as profuse systemic inflammation that usually parallels disease severity, are associated with the development of cirrhosis in our patients.

However, on a genetic level, most (7 of 9) patients underwent MEFV genotyping, with 2 mutations in all of them and M694V homozygosity in 5 (see Table 2). The rate of double mutations and of homozygosity to M694V is therefore higher than expected in the general FMF population, since carriage of 2 mutations is encountered in only 50% of the Israeli FMF population, and carriage of 2 M694V mutations is found in only 25% of double carriers40,48 (p < 0.05 for both comparisons). These findings suggest that carrying 2 MEFV mutations in general, and 2 M694V in particular, may confer an increased risk for liver cirrhosis in FMF patients. How this genetic flaw may lead to cirrhosis is currently unknown, but as noted above, its effect is not mediated by causing a more severe disease or an unusual FMF phenotype.

Could colchicine have a role in the development of cryptogenic cirrhosis in our patients? FMF is the only disease model where colchicine is given on a daily basis, often from early childhood and throughout life. This alkaloid is partially demethylated in the liver by the cytochrome P450 system (particularly the cytochrome P3A4 isoform of the enzyme), and both the unchanged drug and the metabolite enter the GI tract via biliary and intestinal secretions and undergo enterohepatic circulation. The drug is eliminated via both hepatic and renal systems4. In animal models, colchicine may affect Kupffer cell function, thereby leading to reduced toxin clearance by the liver6. Colchicine is widely distributed in the body including the liver tissue. Elevations in liver enzymes and hepatomegaly have been described in patients treated with colchicine34, and liver damage in colchicine intoxication was also reported18. These data make it hard to exclude a detrimental effect of colchicine on liver cells.

Nevertheless, although colchicine is metabolized and utilized by hepatic cells, and is harmful to hepatic cells in high concentrations, there is no evidence that prolonged colchicine use in the acceptable manner leads to liver injury. On the contrary, colchicine has been suggested to be a beneficial treatment for cirrhosis and other fibrosing disorders (including pulmonary fibrosis, Dupuytren contracture, scleroderma, and retroperitoneal fibrosis)25. Severe liver involvement in colchicine intoxication is extremely rare, and occurs only in the second stage of colchicine toxicity37. None of the other systems, more commonly affected by colchicine overdose, were symptomatic in any of the patients presented. Moreover, our experience suggests that liver enzyme fluctuations do not correlate with colchicine dose variations. The mean cumulative colchicine dose before the first appearance of liver symptoms was 8.2 ± 6.3 g, which is a dose comparable to 10 years of treatment with colchicine at a daily dose of 2 mg/d, a situation applicable to more than 2000 patients in our registry not known to suffer liver disease. The mean daily dose used by our patients was 1.4 ± 0.4 mg (see Table 2), which is an even lower dose than that taken by most of our FMF patients20, and much lower than 2-3 mg/d, which is the highest dose permitted. On the whole, the data do not support a role for cumulative dose-dependent colchicine-induced hepatic toxicity as mediating liver damage in the reported patients.

As for other hepatotoxic medications occasionally used by FMF patients, they may also be excluded as a possible cause for liver damage due to the lack of detectable transaminase abnormalities before cirrhosis was diagnosed, or the absence of a temporal association between medication exposure and the elevation in liver enzymes. Nonsteroidal antiinflammatory drugs, frequently used by FMF patients, could be associated with transient liver enzyme elevations, but are not known to cause cirrhosis.

Theoretically, amiodarone in Patient 4 and methotrexate in Patient 5 could be implicated in the development of cirrhosis. However, the nonsupportive histology and the low cumulative dose of methotrexate both argue against this possibility43,46,47. Similarly, liver amyloidosis was not definitely excluded in Patient 9, who had proteinuria with negative rectal biopsy for amyloid A. However, clinically overt AA amyloidosis of the GI tract typically occurs only after many years of overt kidney disease, and is usually accompanied by hepatosplenomegaly and signs of bowel involvement, such as diarrhea, malabsorption, and weight loss32, none of which were present in this patient.

Recent progress in the understanding of cryptogenic cirrhosis must be accounted for in our cohort. Occult HBV infection (defined as the presence of serum or liver HBV DNA in HBsAg-negative individuals) is found in up to 30% of patients with cryptogenic cirrhosis, particularly in endemic areas5,11,42. However, the role of this infection in causing cirrhosis or chronic liver disease has not been clarified. Conflicting reports on this matter have been published: some authors found the rate of cirrhosis or degree of fibrosis in affected individuals to be comparable to rates in the control population23,42, while others found the opposite8. Moreover, some authors reported that serologic testing for anti-HBsAg or anti-HBc detected all patients harboring the occult infection23, while others found that a complete battery of serologic tests for HBV failed to identify 20% of patients with occult HBV DNA1,8. All our patients underwent extensive serologic studies, including anti-HBc in some. Given these circumstances, and based on the expected frequency of occult viral infection and the expected rate of serologic testing failure, we can safely exclude occult HBV infection in the large majority of our patients, leaving only 1 patient with a possible occult infection of a questionable effect.

Two other relatively recent entities that must be accounted for in cryptogenic cirrhosis are nonalcoholic fatty liver disease and its more advanced form, nonalcoholic steatohepatitis, found in about 15% of patients with hypertransaminasemia of unknown etiology5. This, however, may be relevant to only 1 of our patients, Patient 5, whose liver biopsy revealed mild fatty changes. Obesity, diabetes mellitus, and hyperlipidemia, considered risk factors for fatty liver, were not present in any of the patients, and the frequency of fatty liver changes in our cohort (1/5 in biopsy material and none in US imaging) was not different from that found in the general population (17%-33%)17, casting doubt as to their role in causing cirrhosis in the study patients, even in the 1 patient with the fatty liver changes. Overall, the extensive workup and prolonged clinical follow-up support our decision to refer to the cirrhosis in the patients comprising this cohort as cryptogenic.

Curiously, FMF has been associated with a number of autoinflammatory diseases, including Behçet disease, Henoch Schönlein purpura, and Crohn disease32. In all these conditions, mutated MEFV was found to be associated with either precipitating or aggravating the allied condition, suggesting a broader role for the FMF gene beyond inducing the FMF attacks. The exact role of MEFV mutations still remains to be elucidated, yet these mutations were shown to be associated with the induction of various cytokines10,14,19,22,24,35, all playing a central role in the inflammatory cascade probably shared in part by these diseases. This study may add cryptogenic cirrhosis to the list of diseases modified by MEFV.

In conclusion, the current study suggests a possible association of FMF and homozygous M694V MEFV mutation with cryptogenic cirrhosis. This association, albeit rare, may suggest a role for mutated MEFV as a modifier gene in cryptogenic cirrhosis. In contrast, neither colchicine usage nor the various phenotypic features of FMF seem to be related to the development of cirrhosis in FMF patients. Future studies are needed to corroborate and extend these findings, by performing MEFV gene analysis in a large cohort of patients with cryptogenic cirrhosis in Israel and other Mediterranean communities.

REFERENCES

1. Agarwal N, Naik S, Aggarwal R, Singh H, Somani SK, Kini D, Pandey R, Choudhuri G, Saraswat VA, Naik SR. Occult hepatitis B virus infection as a cause of cirrhosis of liver in a region with intermediate endemicity. Indian J Gastroenterol. 2003;22:127-131.
2. Aksentijevich I, Torosyan Y, Samuels J, Centola M, Pras E, Chae JJ, Oddoux C, Wood G, Azzaro MP, Palumbo G, Giustolisi R, Pras M, Ostrer H, Kastner DL. Mutation and haplotype studies of familial Mediterranean fever reveal new ancestral relationships and evidence for a high carrier frequency with reduced penetrance in the Ashkenazi Jewish population. Am J Hum Genet. 1999;64:949-962.
3. Bedossa P, Poynard T. An algorithm for the grading of activity in chronic hepatitis C. The METAVIR Cooperative Study Group. Hepatology. 1996;24:289-293.
4. Ben Chetrit E, Levi M. Colchicine update. Sem Arthritis Rheum. 1998;28:48-59.
5. Berasain C, Betes M, Panizo A, Ruiz J, Herrero JI, Civeira MP, Prieto J. Pathological and virological findings in patients with persistent hypertransaminasaemia of unknown aetiology. Gut. 2000;47:429-435.
6. Bikhazi AB, Jurjus AR, Kamal MT, Al-Housseini AM, Saab RN, Jaroudi WA, Bitar KM. Kinetics of lipopolysaccharide clearance by Kupffer and parenchyma cells in perfused rat liver. Comp Biochem Physiol Toxicol Pharmacol. 2001;129:339-348.
7. Browning JD, Kumar KS, Saboorian MH. Ethnic differences in prevalence of cryptogenic cirrhosis. Am J Gastroenterol. 2004;99:292-298.
8. Cacciola I, Pollicino T, Squadrito G, Cerenzia G, Orlando ME, Raimondo G. Occult hepatitis B virus infection in patients with chronic hepatitis C liver disease. N Engl J Med. 1999;341:22-26.
9. Caldwell SH, Oelsner DH, Iezzoni JC, Hespenheide EE, Battle EH, Driscoll CJ. Cryptogenic cirrhosis: clinical characterization and risk factors for underlying disease. Hepatology. 1999;29:664-669.
10. Chae JJ, Wood G, Masters SL, Richard K, Park G, Smith BJ, Kastner DL. The B30.2 domain of pyrin, the familial Mediterranean fever protein, interacts directly with caspase-1 to modulate IL-1beta production. Proc Natl Acad Sci USA. 2006;103:9982-9987.
11. Chan HL, Tsang SW, Leung NW, Tse CH, Hui Y, Tam JS, Chan FK, Sung JJ. Occult HBV infection in cryptogenic liver cirrhosis in an area with high prevalence of HBV infection. Am J Gastroenterol. 2002;97:1211-1215.
12. Clotet B, Navas J, Grifol M, Rubies-Prat J, Foz M. Liver sinusoidal dilatation in familial Mediterranean fever. Arch Intern Med. 1986;146:1243.
13. Drenth JPH, Van der Meer JWM. Hereditary periodic fever. N Engl J Med. 2001;345:1748-1757.
14. Erken E, Ozer HT, Gunesacar R. Plasma interleukin-10 and interleukin-12 levels in patients with familial Mediterranean fever. Rheumatol Int. 2006;26:862-864.
15. Evans RW. Need for liver transplantation. Lancet. 1995;346:1169.
16. Falk RH, Comenzo RL, Skinner M. The systemic amyloidoses. N Engl J Med. 1997;337:898-909.
17. Farrell GC, Larter CZ. Nonalcoholic fatty liver disease: from steatosis to cirrhosis. Hepatology. 2006;43(Suppl 1):S99-S112.
18. Ferrannini E, Pentimone F. Marrow aplasia following colchicine treatment for gouty arthritis. Clin Exp Rheumatol. 1984;2:173-175.
19. Gang N, Drenth JP, Langevitz P, Zemer D, Brezniak N, Pras M, van der Meer JW, Livneh A. Activation of the cytokine network in familial Mediterranean fever. J Rheumatol. 1999;26:890-897.
20. Gershoni-Baruch R, Brik R, Shinawi M, Livneh A. The differential contribution of MEFV mutant alleles to the clinical profile of familial Mediterranean fever. Eur J Hum Genet. 2002;10:145-149.
21. Guardia J, Vilaseca J, Moragas A, Martinez-Vazquez JM, Bombi JA, Calders J, Bacardi R. Hepatitis in exanthematous Mediterranean fever. Hepatogastroenterology. 1981;28:81-83.
22. Haznedaroglu S, Ozturk MA, Sancak B, Goker B, Onat AM, Bukan N, Ertenli I, Kiraz S, Calguneri M. Serum interleukin 17 and interleukin 18 levels in familial Mediterranean fever. Clin Exp Rheumatol. 2005;23(Suppl 38):S77-80.
23. Kao JH, Chen PJ, Lai MY, Chen DS. Occult hepatitis B virus infection and clinical outcomes of patients with chronic hepatitis C. J Clin Microbiol. 2002;40:4068-4071.
24. Koklu S, Ozturk MA, Balci M, Yuksel O, Ertenli I, Kiraz S. Interferon-gamma levels in familial Mediterranean fever. Joint Bone Spine. 2005;72:38-40.
25. Lange U, Schumann C, Schmidt K. Current aspects of colchicine therapy classical indications and new therapeutic uses. Eur J Med Res. 2001;6:150-160.
26. Livneh A, Zemer D, Langevitz P, Laor A, Sohar E, Pras M. Colchicine in the treatment of AA amyloidosis of FMF: An analysis of factors affecting outcome. Arthritis Rheum. 1994;37:1804-1811.
27. Livneh A. Langevitz P, Zemer D, Zaks N, Kees S, Lidar T, Migdal A, Padeh S, Pras M. Criteria for the diagnosis of familial Mediterranean fever. Arthitis Rheum. 1997;40:1879-1885.
28. Livneh A, Langevitz P. Diagnostic and treatment concerns in familial Mediterranean fever. Baillieres Best Pract Res Clin Rheumatol. 2000;14:477-498.
29. Majeed HA, Halabi I, al-Taleb O. Recurrent hyperbilirubinaemia, a feature of familial Mediterranean fever: report of a child and review of the literature. Ann Trop Paediatr. 1998;18:13-15.
30. Matesanz R, Miranda B. Need for liver transplantation. Lancet. 1995;346:1168-1169.
31. Modan B, Shpilberg O, Baruch Y, Sikuler E, Anis E, Ashur Y, Chetrit A, Luxenburg O, Rosenberg E, Rosenthol N. The need for liver transplantation: a nationwide estimate based on consensus. Lancet. 1995;346:660-662.
32. Mor A, Gal R, Livneh A. Abdominal and digestive system associations of FMF. Am J Gastroenterol. 2003;98:2594-2604.
33. Mor A, Shinar Y, Zaks N, Langevitz P, Chetrit A, Shtrasburg S, Rabinovitz E, Livneh A. Evaluation of disease severity in familial Mediterranean fever. Semin Arthritis Rheum. 2005;35:57-64.
34. Neequaye J, Jelly AE. Acute hepatitis in recurrent hereditary polyserositis (familial Mediterranean fever). J Trop Pediatr. 1994;40:243-245.
35. Notarnicola C, Didelot MN, Seguret F, Demaille J, Touitou I. Enhanced cytokine mRNA levels in attack-free patients with familial Mediterranean fever. Genes Immun. 2002;3:43-45.
36. Pugh R, Murray-lyon I, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg. 1973;60:646-649.
37. Putterman C, Ben Chetrit E, Caraco Y, Levy M. Colchicine intoxication: clinical pharmacology, risk factors, features, and management. Semin Arthritis Rheum. 1991;21:143-155.
38. Reimann HA. Hepatitis-a feature of periodic peritonitis. JAMA. 1961;178:334-335.
39. Sakugawa H, Nakasone H, Nakayoshi T, Kawakami Y, Yamashiro T, Maeshiro T, Kobashigawa K, Kinjo F, Saito A. Clinical characteristics of patients with cryptogenic liver cirrhosis in Okinawa, Japan. Hepatogastroenterology. 2003;50:2005-2008.
40. Shinar Y, Livneh A, Langevitz P, Zaks N, Aksentijevich I, Koziol DE, Kastner DL, Pras M, Pras E. Genotype-phenotype assessment of common genotypes among patients with familial Mediterranean fever. J Rheumatol. 2000;27:1703-1707.
41. Shohat M, Magal N, Shohat T, Chen X, Dagan T, Mimouni A, Danon Y, Lotan R, Ogur G, Sirin A, Schlezinger M, Halpern GJ, Schwabe A, Kastner D, Rotter JI, Fischel-Ghodsian N. Phenotype-genotype correlation in familial Mediterranean fever: evidence for an association between Met694Val and amyloidosis. Eur J Hum Genet. 1999;7:287-292.
42. Silva C, Goncales NS, Pereira JS, Escanhoela CA, Pavan MH, Goncales FL Jr. The influence of occult infection with hepatitis B virus on liver histology and response to interferon treatment in chronic hepatitis C patients. Braz J Infect Dis. 2004;8:431-439.
43. Singhal A, Ghosh P, Khan SA. Low dose amiodarone causing pseudo-alcoholic cirrhosis. Age Aging. 2003;32:224-225.
44. Sohar E, Gafni J, Pras M, Heller H. Familial Mediterranean fever. A survey of 470 cases and review of the literature. Am J Med. 1967;43:227-253.
45. Touitou I, Lesage S, McDermott M, Cuisset L, Hoffman H, Dode C, Shoham N, Aganna E, Hugot JP, Wise C, Waterham H, Pugnere D, Demaille J, Sarrauste de Menthiere C. Infevers: An evolving mutation database for auto-inflammatory syndromes. Hum Mutat. 2004;24:194-198.
46. Zachariae H, Sogaard H, Heickendorff L. Methotrexate-induced liver cirrhosis. Clinical, histological and serological studies-a further 10-year follow-up. Dermatology. 1996;192:343-346.
47. Zachariae H, Heickendorff L, Sogaard H. The value of amino-terminal propeptide of type III procollagen in routine screening for methotrexate-induced liver fibrosis: a 10-year follow-up. Br J Dermatol. 2001;144:100-103.
48. Zaks N, Shinar Y, Padeh S, Lidar M, Mor A, Tokov I, Pras M, Langevitz P, Pras E, Livneh A. Analysis of the three most common MEFV mutations in 412 patients with familial Mediterranean fever. Isr Med Assoc J. 2003;5:585-588.
49. Zemer D, Livneh A, Langevitz P. Reversal of the nephrotic syndrome by colchicine in amyloidosis of familial Mediterranean fever. Ann Intern Med. 1992;116:426.
© 2007 Lippincott Williams & Wilkins, Inc.