Noncirrhotic Portal Hypertension Associated with Didanosine: a Case Report and Literature Review

Transcription

1 Jpn. J. Infect. Dis., 65, 61-65, 2012 Short Communication Noncirrhotic Portal Hypertension Associated with Didanosine: a Case Report and Literature Review Hui-Min Chang 1, Hung-Chin Tsai 2,3 *, Susan Shin-Jung Lee 2,3, Shue-Ren Wann 2,3, and Yao-Shen Chen 2,3 1 Department of Pharmacy and 2 Section of Infectious Diseases, Department of Medicine, Kaohsiung Veterans General Hospital, Kaohsiung; and 3 National Yang-Ming University, Taipei, Taiwan (Received February 23, Accepted October 14, 2011) SUMMARY: Noncirrhotic portal hypertension (NCPH) has recently been reported as a liver complication in human immunodeficiency virus (HIV)-infected patients and has been found to be associated with exposure to didanosine. Here, we describe the case of an HIV-infected patient with portal hypertension who initially presented with massive ascites and portal vein thrombosis. The patient's HIV-1 infection was well-controlled with highly active antiretroviral therapy (lamivudine/didanosine plus nevirapine) for 3 years since its diagnosis in He had no history of alcoholism, drug abuse, or liver diseases. An extensive work-up for other possible causes of liver disease was performed, but the results were inconclusive. In addition to reporting this case, we have reviewed the literature on didanosine-related NCPH and analyzed the findings of 61 similar previously reported cases. Didanosine is a nucleoside/nucleotide reverse transcriptase inhibitor that is effective against HIV-1 and was often used in combination with other antiretroviral drugs as a part of highly active antiretroviral therapy (HAART). However, on January 29, 2010, the Food and Drug Administration updated the labeling regulations for didanosine because the drug was suspected to cause noncirrhotic portal hypertension (NCPH) in HIVinfected patients. We recently encountered this adverse effect of didanosine in a Chinese HIV patient. An 80-year-old Chinese man who had been experiencing abdominal fullness and nausea for 1 month was admitted to the Kaohsiung Veterans General Hospital, Taiwan. He was diagnosed with HIV-1 infection in 2007, and since then, he had been undergoing HAART with nevirapine, lamivudine, and didanosine. Didanosine was administered for 34 months. However, the patient did not report any adverse effects after administration of these antiretroviral drugs, and his viral load was undetectable. His CD4 lymphocyte count 1 month before admission was 236 cells/ml, and his HIV-1 RNA viral load was undetectable. The patient had no history of opportunistic infections, liver disease, or alcohol or drug abuse. The exact route of HIV transmission in this patient was unknown. Upon arrival at our clinic, the patient's aspartate aminotransferase (AST) level was 50 U/L, alanine aminotransferase (ALT) level was 36 U/L, alkaline phosphatase (ALP) level was 188 U/L, and g-glutamyltransferase (ggt) level was 332 U/L. The patient's International Normalized Ratio was 1.1 and his platelet count was /mm 3. Blood urea *Corresponding author: Mailing address: Section of Infectious Diseases, Department of Medicine, Kaohsiung Veterans General Hospital, 386 Ta-Chung First Rd., Kaohsiung 813, Taiwan. Tel: ext 2029, Fax: , hctsai1011@yahoo.com.tw nitrogen, creatinine, albumin, globulin, electrolyte, and total bilirubin levels were normal. The results of assays for the hepatitis B surface antigen (HBsAg) and hepatitis B surface antibody (HBsAb) were negative, while those for the hepatitis B core antibody (HBcAb) were positive. Under the impression of occult hepatitis B infection, the hepatitis B virus DNA (HBV-DNA) level was measured and was negligible (less than 10 copies/ml). The anti-hepatitis C virus antibody (anti- HCV) test result was negative. The serum antinuclear antibody test result was negative and the stool examination did not show any ova of schistosomiasis. He had no AIDS defining illness before. Gastrointestinal endoscopic examination revealed 4 esophageal varices (1F3 and 3F2) with positive red-color sign and red wale markings. However, the upper abdomen sonogram and upper GI series performed 6 months before this presentation did not show any evidence of cirrhosis and portal hypertension. Abdominal sonography conducted after admission revealed slightly increased liver echogenicity, a patent portal vein, and moderate ascites, but no splenomegaly. However, abdominal computerized tomography confirmed the presence of thrombosis in the portal vein as well as massive ascites (Fig. 1). Subsequently, combined treatment with spironolactone and furosemide was started. The ascitic fluid was analyzed and was yellow and clear; the erythrocyte was 61/mm 3 ; and the leukocyte was 203/mm 3,ofwhich69z were lymphocytes and 31z were mononuclear cells. Biochemical testing of the ascitic fluid revealed a glucose level of 123 mg/dl, total protein content of 1 g/dl, lactate dehydrogenase (LDH) level of 164 U/L, and amylase level of 36 U/L. The serum-ascites albumin gradient was 2.1 g/dl. All bacterial, tuberculous, and fungal cultures yielded negative results. On the sixth day after admission, the patient felt considerably better, and he was discharged without a liver biopsy and prescribed a regimen of abacavir, nevirapine, and lamivudine. 61

2 Fig. 1. The abdominal computerized tomography revealed massive ascites, portal vein thrombosis, and normal spleen size in our patient with noncirrhotic portal hypertension. However, 3 months later, the patient died of upper gastrointestinal bleeding and respiratory failure. We used the PubMed database to find published studies (relevant articles published up to November 2010) with the following key words: NCPH, nodular regenerative hyperplasia, portal hypertension, and didanosine. Our search was limited to articles in English. Additional studies were selected from the reference lists in the reports. Only studies with data on didanosine-related adverse reactions in HIV-infected patients were included in this review. We included all published studies that described didanosine treatment in detail. The following data were extracted and tabulated from each study: author name, year of publication, patient sex and age, study design, inclusion and exclusion status, clinical presentation, liver function, didanosine exposure time, and antiretroviral drugs. In all the studies that were conducted, NCPH was confirmed by liver biopsy. The process of retrieval and selection of the included articles is depicted in detail in Fig. 2. Of the 42 identified articles, 13 met the inclusion criteria for our review. Twenty-nine articles were excluded from the selection for various reasons (Fig. 2). A total of 8 case reports or case series and 5 case control studies (1 13) shortlisted in the literature search were reviewed. Table 1 and 2 show the demographic, clinical, and laboratory characteristics of the patients. Among the 61 cases with didanosine-associated NCPH (from 8 case reports or case series) included in our review, most (92z) were either from the United States or Europe. The patient's ages were between 26 and72years,with76z being male. All median baseline CD4 counts at the time of diagnosis of NCPH or nodular regenerative hyperplasia (NRH) were above 200 cells/ml. Hepatitis B and C co-infection was absent in 88z of the patients. The interval between diagnosis of HIV infection and diagnosis of NCPH or NRH ranged between 5.7 and 23 years. Most of the patients showed stable virological suppression and mildly elevated liver Fig. 2. Flow chart showing how the articles were selected and excluded. enzyme levels. The duration of didanosine exposure was between 13 to 111 months in these patients. Among the 5 case control studies included in our review, one was conducted by Cesari et al. (9), who studied 11 HIV-infected patients with idiopathic NCPH and 10 age- and gender-matched HIV-infected controls with no liver disease on exposure to didanosine. The median duration of exposure to didanosine was 72 months in the patients with NCPH and 52 months in the controls (P º 0.01). The median baseline CD4 count at the time of diagnosis of NCPH was 271 cells/ml. All patients had achieved stable virological suppression. Cotte et al. (10) studied 13 HIV-infected patients with NRH and 78 matched controls. The median duration of exposure to didanosine was 48 months in the patients with NRH and 15 months in the controls (odds ratio [OR], 1.3; 95z CI, ; P º 0.013). The median baseline CD4 count at the time of diagnosis of NCPH was 327 cells/ml. Most of the patients (84z) had achieved stable virological suppression. Dinh et al. (11) studied 9 HIV-infected patients with cryptogenic liver disease. The mean duration of exposure to didanosine was 72 ± 32.1 months. Kovari et al. (12) studied 15 patients with NCPH and 75 matched controls. In their study, didanosine exposure was prominent in NCPH patients (OR, 3.44; 95z CI, ). (13) studied 17 patients with cryptogenic liver disease and 17 matched 62

3 Table 1. Clinical characteristics of didanosine associated noncirrhotic portal hypertensionsion in the 13 published reports Authors (ref. no.) (year) Country Design No. of patients No. of didanosine exposure Co-treatment Age Gender (male/ female) AIDS diagnosis Risk group (HS/MSM/ IVDU/Alc) Time on HAART ddi exposure (mo) Viral load 1) (copies/ml) CD4 cell Count (cells/ml) Arey et al. (1) (2007) Bihl et al. (2) (2010) Cesari et al. (9) (2010) Cotte et al. (10) (2010) Dinh et al. (11) (2009) Sandrine et al. (3) (2007) Garvey et al. (4) (2007) Kovari et al. (12) (2009) (13) (2006) (5) (2008) Saifee et al. (6) (2008) Tateo et al. (7) (2008) Vispo et al. (8) (2010) USA CR 1 1 Emtricitabine Tenofovir Swiss CR 1 1 Lamivudine Abacavir Italy CC NCPH: 11 Control: 10 France CC NRH: 13 Control: 78 USA CC CLD: 9 Control: male NR MHSM/Alc NR NR Undetectable female 7 No Alc intake , Any HAART 44 2) 7/4 NR 6/3/2/ ) 72 2) Undetectable: ) 13 Any HAART ) 10/ ) 6/6/0/ ) 48 2) Undetectable: ) 8 Any HAART (4 patients: hydroxyurea concomitantly) France CR 1 1 Zidovudine Nevirapine UK (4/6 from Uganda) 50 1) 9/ ) 8/0/0/0 10 1) 72 1) Undetectable: 8 1 patient: female 11 NR , CR NCPH: 6 5 NR NR 10 NR NR 58 2) NR 445 2) Swiss CC NCPH: 15 Control: 75 Spain CC CLD: 17 Control: Any HAART 52 2) 13/ ) 4/11/0/NR 8 2) 60 2) Undetectable: ) 17 Nevirapine Stavudine Ritonavir (3 patients: hydroxyurea concomitantly) Spain CS CLD: Nevirapine Stavudine Ritonavir USA CR NCPH: Lamivudine and Zidovudine 42 2) 14/3 15 2) 3/13/1/0 4 2) 51 2) Undetectable: 14 Other: 116, 893, 694, 14, ) 27/ ) 5/22/4/ ) 44 2) Undetectable: ) 51 1) 10/1 14 1) NR NR 61 1) Undetectable: ) Other: 317 1) France CR NRH: 3 3 Any HAART 38 2) 1/2 14 2) NR 10 2) 60 2) Undetectable: ) Spain CS NCPH: NR 47 2) 10/2 11 2) 1/9/1/0 NR 53 2) Undetectable: 11 1 patient: 3282 NR 368 2) 232 2) 1) : Values are the mean. 2) : Values are the median. 3) : Upper normal range. HS, heterosexual; MSM, men having sex with men; IVDU, intravenous drug users; Alc, alcohol; CC, case control CS, case series; CR, case report; CLD, cryptogenic liver disease; NRH, nodular regenerative hyperplasia; NCPH, noncirrhotic portal hypertension; HAART, highly active antiretrovial therapy; NR, not recorded. 63

4 Table 2. Laboratory characteristics and findings of liver biopsy in patients with didanosine-associated noncirrhotic portal hypertension in the 13 published reports Authors (ref. no.) (year) HBV/HCV coinfection ALT IU/L AST IU/L Total bilirubin (mg/dl) Platelets (10 3 /ml) Liver biopsy Clinical presentation Arey et al. (1) (2007) Bihl et al. (2) (2010) Cesari et al. (9) (2010) Cotte et al. (10) (2010) Dinh et al. (11) (2009) Sandrine et al. (3) (2007) Garvey et al. (4) (2007) Kovari et al. (12) (2009) (13) (2006) (5) (2008) Saifee et al. (6) (2008) Tateo et al. (7) (2008) Vispo et al. (8) (2010) Negative NR Sinusoidal dilation Patchy hepatocellular necrosis and congestion. Negative 1.5 times 3) 1.5 times 3) 16 NR NRH Portal fibrosis NR 49 2) NR Liver biopsy in 5 of 11 Microvacuolar steatosis: 1 NRH: 5 1/3 39 2) 33 2) NR 169 Fibrosis: 5 NRH: 13 Sinusoidal dilatation: 7 Steatosis: 7 Negative NR NR Sinusoidal dilatation: 1 Cirrhosis: 1/9 NRH: 3/9 Fibrosis: 6/9 Perisplenic varices Portal gastropathy Ascites : 9 gastropathy: 2 Portal thrombosis: 2 : 9 Variceal bleeding: 2 Portal thrombosis: 5 Ascites: 4 : 8 : 9 Ascites: 6 Encephalopathy: 2 : 8 Portal vein thrombus: 1 : 8 Spontaneous bacterial peritonitis: 2 Variceal bleeding: 4 NR 1.5 times 3) 2.5 times 3) NR 70 NRH Ascites 5/1 NR NR NR NR Normal liver: 1 NRH: 2 Sinusoidal dilatation: 3 Venous outflow obstruction: 3 NR 51 2) NR NR 166 2) Ductopenia and ductular proliferation: 4 Hepatocellular necrosis: 4 Macrovacuolar and/or microvesicular steatosis: 6 Normal liver: 1 NRH: 1/15 Periportal fibrosis: 10/15 Perisinusoidal fibrosis: 1 Portal and lobular inflammation: 7 Sinusoidal dilatation: 1 Negative ) 58 2) ) 304 2) Liver biopsy in 5 of 17 Fibrosis: 5 Liver cirrhosis: 2 Microvesicular steatosis: 5 Negative 71 2) 56 2) NR NR Liver biopsy in 12 of 32 Cholangiopathy: 1 Cirrhosis: 4 Liver fibrosis: 3 Microvesicular steatosis: 2 Normal liver: 1 NRH: 2 Periportal fibrosis: 3 1/0 53 1) 56 1) 1.2 1) 177 1) NRH: 11 Portal venulopathy: 5 Steatosis: 2 Negative 62 2) 49 2) 2.1 2) 165 2) NRH: 3 Perisinusoidal fibrosis: 2 Sinusoidal dilatation: 1 Negative ) ) NR NR Liver biopsy in 10 of 12 NRH: 4 Perisinusoidal fibrosis: 1 Portal venopathy: 4 : 5 gastropathy: 1 : 6 Ascites: 8/15 : 15/15 Hepatic encephalopathy: 2/15 Portal vein thrombosis: 3 : 13/15 Variceal bleeding: 7/15 Ascites: 8/17 : 5/17 Hepatic encephalopathy: 2/17 Portal thrombosis: 6/17 Variceal bleeding: 5/17 Ascites: 8 : 13 Portal thrombosis: 8 : 32 Variceal bleeding: 5 Ascites: 6 : 11 : 3 gastropathy: 1 Portal vein thrombosis: 2 : 1 Variceal bleeding: 6 Portal vein thrombus: 4 Footnotes are in Table 1. HBV, hepatitis B virus; HCV, hepatitis C virus; ALT, alanine aminotransferase; AST, aspartate aminotransferase. 64

5 controls. The mean duration of exposure to didanosine was 47 months in the patients with cryptogenic liver diseaseand25monthsinthecontrols(p = 0.009). The median baseline CD4 count at the time of diagnosis of cryptogenic liver disease was 368 cells/ml. The symptoms associated with HIV-related NCPH were esophageal varices, splenomegaly, ascites, portal vein thrombosis, and variceal bleeding. In the 61 patients who underwent a liver biopsy, the most commonly seen pathological findings were NRH (75z, 46/61), portal fibrosis (59z, 36/61), and microvacuolar steatosis (37z, 23/61). In conclusion, these results show a strong correlation between exposure to didanosine, the presence of NRH in liver pathology and the development of NCPH. Similar to didanosine, several other drugs, including the adenosine analogue azathioprine, have been associated with this form of portal hepatopathy. The exact mechanism underlying the liver toxicity of azathioprine is unclear, but it may involve the depletion of intracellular gluthatione, leading to death of sinusoidal endothelial cells (14). The OR for development of NCPH within a year of exposure to didanosine is 3.4 (95z CI, ) (12). Exposure to didanosine has recently been shown to enhance proinflammatory status and increase cardiovascular events. One hypothesis to explain these effects is that purine analogs could interfere with the proinflammatory signaling molecules adenosine triphosphate and adenosine diphosphate present in vascular endothelial cells, leading to vascular damage (15). Recurring endothelial portal vein infections, such as pylephlebitis, which may be caused by microbial translocation from the gut, could play a role in NCPH in the HIV-infected individuals (5). Finally, a genetic predisposition cannot be ruled out as an important contributor to NCPH in HIV-infected persons. The inosine triphosphatase deficiency may lead to an increased risk of purine analog toxicity (16). Genetic polymorphisms affecting the expression and/or function of other critical enzymes in the purine metabolic pathway might also contribute to potentiate the toxicity of didanosine over the portal vessels. Thromboembolic events have been reported in some HIV-infected individuals with NCPH (6), as seen in our patient. Patients receiving didanosime may also show enhancement of the levels of proinflammatory mediators, which might create a prothrombotic state (17). In another report, the authors proposed that the prothrombotic state could be the result of an autoimmune functional inactivation of circulating protein S levels due to abnormal B-cell activation (18). The results of our review revealed that didanosine exposure was a risk factor for development of NCPH. Our patient had been exposed to didanosine for 34 months; prolonged exposure to didanosine is one of the main risk factors for developing NCPH. Since the therapeutic options are not yet clear, symptomatic treatment of portal hypertension and discontinuation of didanosine should be considered as the primary options for treatment. In summary, didanosine-associated NCPH is rarely reported in Asian people. Further investigation is needed to clarify the roles of genetic polymorphisms in enzymes involved in the purine metabolic pathway. HIVinfected patients with NCPH may experience potentially life-threatening gastrointestinal bleeding. Early diagnosis is important, as preventive measures can be undertaken to detect esophageal varices and prevent bleeding, since portal vein thrombosis is a frequent complication in patients with NCPH. The benefit of anticoagulation and the risk of bleeding warrant further investigation. Didanosine-associated portal damage should be taken into consideration for diagnosis in patients with NRH, portal fibrosis, or steatosis in liver pathological examinations and in patients with a history of didanosine use. Conflict of interest None to declare. REFERENCES 1. Arey, B., Markov, M., Ravi, J., et al. (2007): Nodular regenerative hyperplasia of liver as a consequence of ART. AIDS, 21, Bihl, F., Janssens, F., Boehlen, F., et al. (2010): Anticoagulant therapy for nodular regenerative hyperplasia in a HIV-infected patient. BMC Gastroenterol., 10, Sandrine, P.F., Sylvie, A., Andre, E., et al. (2007): Nodular regenerative hyperplasia: a new serious antiretroviral drugs side effect? AIDS, 21, Garvey, L.J., Thomson, E.C., Lloyd, J., et al. (2007): Nodular regenerative hyperplasia is a new cause of chronic liver disease in HIV-infected patients. AIDS, 21, Maida, I., Garcia-Gasco, P., Sotgiu, G., et al. (2008): Antiretroviral-associated portal hypertension: a new clinical condition? Prevalence, predictors and outcome. Antivir. Ther., 13, Saifee, S., Joelson, D., Braude, J., et al. (2008): Noncirrhotic portal hypertension in patients with human immunodeficiency virus 1 infection. Clin. Gastroenterol. Hepatol., 6, Tateo, M., Sebagh, M., Bralet, M.P., et al. (2008): A new indication for liver transplantation: nodular regenerative hyperplasia in human immunodeficiency virus-infected patients. Liver Transpl., 14, Vispo, E., Moreno, A., Maida, I., et al. (2010): Noncirrhotic portal hypertension in HIV-infected patients: unique clinical and pathological findings. AIDS, 24, Cesari, M., Schiavini, M., Marchetti, G,. et al. (2010): Noncirrhotic portal hypertension in HIV-infected patients: a case control evaluation and review of the literature. AIDS Patient Care STD, 24, Cotte, L., Benet, T., Billioud, C., et al. (2010): The role of nucleoside and nucleotide analogues in nodular regenerative hyperplasia in HIV-infected patients: a case control study. J. Hepatol., Sep 19 [Epub ahead of print]. 11. Dinh, M.H., Stosor, V., Rao, S.M., et al. (2009): Cryptogenic liver disease in HIV-seropositive men. HIV Med., 10, Kovari, H., Ledergerber, B., Peter, U., et al. (2009): Association of noncirrhotic portal hypertension in HIV-infected persons and antiretroviral therapy with didanosine: a nested case-control study. Clin. Infect. Dis., 49, Maida, I., Nunez, M., Rios, M.J., et al. (2006): Severe liver disease associated with prolonged exposure to antiretroviral drugs. J. Acquir. Immune Defic. Syndr., 42, DeLeve, L. (2007): Hepatic microvasculature in liver injury. Semin, Liver Dis., 27, Goicoechea, M. and McCutchan, A. (2008): Abacavir and increased risk of myocardial infarction. Lancet, 372, Fellay, J., Thompson, A., Ge, D., et al. (2010): ITPA gene variants protect against anaemia in patients treated for chronic hepatitis C. Nature, 464, Strategies for Management of Antiretroviral Therapy/INSIGHT, DAD Study Groups (2008): Use of nucleoside reverse transcriptase inhibitors and risk of myocardial infarction in HIV-infected patients. AIDS, 22, F17 F Mallet, V., Varthaman, A., Lasne, D., et al. (2009): Acquired protein S deficiency leads to obliterative portal venopathy and to compensatory nodular regenerative hyperplasia in HIV-infected patients. AIDS, 23,