Share this article on:

Risk Factors for Symptomatic Mitochondrial Toxicity in HIV/Hepatitis C Virus-Coinfected Patients During Interferon Plus Ribavirin-Based Therapy

Bani-Sadr, Firouzé MD*; Carrat, Fabrice MD, PhD*; Pol, Stanislas MD, PhD; Hor, Ravy MD*; Rosenthal, Eric MD; Goujard, Cécile MD§; Morand, Patrice MD, PhD; Lunel-Fabiani, Françoise MD, PhD; Salmon-Ceron, Dominique MD, PhD#; Piroth, Lionel MD, PhD**; Pialoux, Gilles MD, PhD††; Bentata, Michèle MD‡‡; Cacoub, Patrice MD††; Perronne, Christian MD, PhD§§The ANRS Hc02-Ribavic Study Team

JAIDS Journal of Acquired Immune Deficiency Syndromes: September 1st, 2005 - Volume 40 - Issue 1 - p 47-52
doi: 10.1097/01.qai.0000174649.51084.46
Clinical Science

Objective: To evaluate the incidence, clinical features, and risk factors for symptomatic mitochondrial toxicity in HIV/hepatitis C virus (HCV)-coinfected patients receiving anti-HCV therapy.

Methods: All cases of symptomatic mitochondrial toxicity reported in 416 patients participating in an open, randomized trial of peg-interferon α-2b plus ribavirin vs. interferon α-2b plus ribavirin for 48 weeks were reviewed. Associations with antiretroviral treatments and with clinical and laboratory findings were sought by univariate and multivariate analysis.

Results: Eleven of the 383 patients who received at least 1 dose of anti-HCV treatment developed symptomatic mitochondrial toxicity (symptomatic hyperlactatemia and pancreatitis in 6 and 5 patients, respectively). All cases occurred in patients being treated for HIV infection, and the incidence of symptomatic mitochondrial toxicity was 47.5 per 1000 patient-years. In multivariate analysis, symptomatic mitochondrial toxicity was significantly associated with didanosine-containing antiretroviral regimens (odds ratio 46; 95% CI, 7.4 to infinity; P < 0.001), but not with stavudine or with nucleoside reverse transcriptase inhibitor regimens not containing didanosine. The incidence of symptomatic mitochondrial toxicity was 200.2 per 1000 patient-years in patients receiving didanosine. Demographic characteristics were not associated with symptomatic mitochondrial toxicity.

Conclusions: Coadministration of ribavirin with didanosine should be avoided. If unavoidable, patients should be monitored closely for mitochondrial toxicity. Didanosine should be suspended if clinical signs or symptoms of mitochondrial toxicity occur.

From the *Groupe Hospitalier Universitaire Est, Université Paris 6, INSERM U444, Paris, France; †Groupe Hospitalier Universitaire Ouest, Université Paris 5, INSERM U370, Paris, France; ‡Hôpital de l'Archet, Faculté de Médecine, Nice, France; §Groupe Hospitalier Universitaire Sud, Université Paris 11, Paris, France; ∥Centre Hospitalier Universitaire, Grenoble, France; ¶Groupe Hospitalier, Faculté de Médecine, Angers, France; #Centre Hospitalier Universitaire Cochin, Paris, France; **Centre Hospitalier Universitaire, Dijon, France; ††Groupe Hospitalier Universitaire Est, Université Paris 6, Paris, France; ‡‡Centre Hospitalier Universitaire, Avicenne, France; and §§Centre Hospitalier Universitaire Raymond Poincaré, Université de Versailles, Garches, France.

Received for publication February 14, 2005; accepted June 8, 2005.

Supported by Agence Nationale de Recherche sur le SIDA and SIDACTION.

Reprints: Firouzé Bani-Sadr, Inserm U444, Faculté de Médecine, Hôpital Saint Antoine, 27 rue de Chaligny, 75475, Paris Cedex 12, France (e-mail:

Combination therapy with peg-interferon (pegIFN) and ribavirin is currently the standard of care for patients with hepatitis C virus (HCV) infection.1-3 Although IFN-α impairs mitochondrial DNA transcription and ribavirin is a guanosine nucleoside analogue, symptomatic mitochondrial toxicity has not previously been described in HCV-monoinfected patients receiving this combination.1,2,4 In contrast, several recent reports show that the use of the ribavirin-interferon combination for chronic hepatitis C in HIV-seropositive patients receiving highly active antiretroviral therapy can cause mitochondrial toxicity.5,6 The aims of this study were to study the incidence of symptomatic mitochondrial toxicity in HIV/HCV-coinfected patients participating in a large trial comparing pegIFN α-2b plus daily ribavirin with thrice-weekly of 3 million units of IFN α-2b plus daily ribavirin; to describe the laboratory and clinical outcomes associated with symptomatic mitochondrial toxicity; and to identify associated risk factors.

Back to Top | Article Outline


The results of the Agence Nationale de Recherche sur le SIDA HC02 Ribavic trial have been reported elsewhere.3 In brief, 416 HIV/HCV-coinfected patients who had never received IFN or ribavirin were randomly assigned to receive either weekly subcutaneous injections of 1.5 μg/kg pegIFN α-2b (ViraferonPeg, Schering-Plough, Kenilworth, NJ) plus daily ribavirin (Rebetol, Schering-Plough) (800 mg) (n = 206), or thrice-weekly subcutaneous injections of 3 MU of IFN α-2b plus daily ribavirin (Rebetol, Schering-Plough) (800 mg) (n = 210) for 48 weeks. Patients were eligible if they had detectable serum HCV RNA, liver biopsy results showing at least mild activity or fibrosis, a CD4 cell count above 200/mm3, and stable HIV RNA load (<1-log variation in the previous 3 months) at randomization, and stable or no antiretroviral treatment during the previous 3 months. The main ineligibility criteria were active narcotic consumption or self-reported daily alcohol intake exceeding 40 g (women) or 50 g (men) within 3 months before entry to the study, decompensated cirrhosis, and hepatitis B surface antigen antigenemia.

For this study, we reviewed all recorded cases of symptomatic mitochondrial toxicity, defined as symptomatic hyperlactatemia or pancreatitis. Symptomatic hyperlactatemia was defined as a lactate elevation (venous lactate >2 mM) with clinical signs. Pancreatitis was diagnosed on the basis of typical clinical findings (eg, abdominal pain in a patient with no other causes of abdominal pain) with supporting biochemical signs or amylase or lipase activity more than 3 times the upper limit of normal.

The following data were collected at inclusion and every 4 weeks during follow-up: age, sex, body mass index, antiretroviral drugs received, anti-HCV treatment received, duration of HIV and HCV treatment, Metavir score at inclusion, symptoms and signs, biochemical test results, liver enzyme activities, CD4 lymphocyte counts, plasma HIV RNA load, and plasma HCV RNA load. Lipase, amylase, and lactate were assayed at the request of the treating physician. Serum bicarbonate and lactate were assayed in venous blood.

The χ2 test or the Fisher exact test was used to analyze qualitative variables and the Mann-Whitney U test was used for quantitative variables. A logistic regression model was used to test possible associations between symptomatic mitochondrial toxicity (independent variable) and antiretroviral regimens (dependent variables). An exact method was used to calculate odds ratios and 95% CIs.7 Incidence rates of mitochondrial toxicity events were calculated by using the number of person-days of ribavirin therapy as denominator, and the number of events as numerator, in patients who took at least 1 dose of anti-HCV treatment. All statistical tests were 2-sided, with a type I error of 5%.

Back to Top | Article Outline


Symptomatic mitochondrial toxicity occurred in 11 of the 383 patients who received at least 1 dose of study medication (symptomatic hyperlactatemia and pancreatitis in 6 and 5 patients, respectively). The clinical and biologic characteristics of these 11 patients are summarized in Tables 1 and 2, respectively. The mean baseline CD4 cell count was 426 cells/mm3 (±176), and plasma HIV RNA load ranged from less than 200 copies/mL (5 patients) to 33,200 copies/mL. Severe liver fibrosis (Metavir score F3 or F4) was present in 55% of the 11 patients. All the patients were receiving antiretroviral treatment. The regimens included didanosine (ddI) in every case, stavudine (d4T) in 8 cases, lamivudine (3TC) in 3 cases, abacavir in 1 case, protease inhibitors (PI) in 3 cases, and nonnucleoside reverse transcriptase inhibitors (NNRTIs) in 5 cases. The most common dual-NRTI regimen was ddI plus d4T (8 patients). The mean duration of ddI therapy at the diagnosis of symptomatic mitochondrial toxicity was 23.9 months (±14.4). ddI was used at the recommended doses, namely 400 mg (tablets) or 500 mg (powder) daily for patients weighing more than 60 kg and 250 mg daily (tablets) for patients under this weight.





Nine of the 11 patients were in the pegIFN-ribavirin arm. The mean interval between the outset of anti-HCV treatment and the diagnosis of symptomatic mitochondrial toxicity was 26.5 weeks (±13.4).

The clinical manifestations of symptomatic hyperlactatemia were general weakness (5 patients), weight loss (5 patients; ≥6 kg in 4 cases), nausea (2 patients), severe rapid-onset lipoatrophy (2 patients), dyspnea (1 patient), and peripheral neuropathy (1 patient). No intercurrent acute health disorders were diagnosed. Five patients had hyperlactatemia (>3 mM; ≥5 mM in 2 cases). Two patients (lactate level unknown in 1) had evidence of metabolic acidosis (bicarbonate levels 15 and 16 mM). In patients with pancreatitis, the main symptoms of pancreatitis were abdominal pain (4 patients) with or without nausea/vomiting (2 patients). The median serum lipase elevation was 10 times the upper limit of normal (range, 3.53-10 IU/L). A large increase in gamma-glutamyl transferase (GGT) levels was observed in 6 of the 11 patients (median elevation from 4.95 [2-7.4] to 14.35 [6-30] the upper limit of normal). By contrast, alanine aminotransferase (ALT) and alkaline phosphatase (alk Ph) levels were not significantly different from baseline values.

All but 1 of the patients (the patient with neuropathy) made a full recovery. However, one patient (patient 6) died 15 months later from liver failure, despite a sustained HCV virologic response.

Back to Top | Article Outline

Factors Associated With Symptomatic Mitochondrial Toxicity

Univariate analysis showed no association with age, gender, the body mass index, HCV transmission group, AIDS status, CD4 cell count, HIV viral load, mean duration of antiretroviral treatment, mean duration of HIV or HCV infection, ALT, GGT, or alk Ph levels, HCV viral load and HCV genotype, mean Metavir inflammation and fibrosis scores, or the HCV treatment arm (pegIFN plus ribavirin or standard IFN plus ribavirin) (Table 3).



The possible role of antiretroviral agents was analyzed by both single NRTI, NNRTI, and PI used at the initiation of anti-HCV treatment. Univariate analysis showed no association with PI- or NNRTI-containing regimens. Among the NRTIs, only ddI was significantly associated with an increased risk of symptomatic mitochondrial toxicity. To study the impact of specific NRTI combinations, 4 groups of NRTI regimens were determined a priori. Two contained ddI (with or without d4T), 1 contained d4T but not ddI, and 1 contained neither ddI nor d4T. Eighty-three patients received regimens that included ddI, either with d4T (n = 52) or without d4T (n = 31). Symptomatic mitochondrial toxicity occurred in 8 (15.4%) of the 52 patients treated with ddI and d4T and in 3 (9.6%) of the 31 patients treated with ddI without d4T. No cases occurred among the 122 patients receiving d4T without ddI or in the 112 patients receiving an NRTI regimen containing neither d4T nor ddI. Multivariate analysis showed that ddI, with or without d4T, was associated with symptomatic mitochondrial toxicity (OR 46, 95% CI, 7.4 to infinity; P < 0.001) compared with regimens with neither ddI nor d4T. The risk of symptomatic mitochondrial toxicity was not increased in patients treated with d4T without ddI (OR 1.8; 95% CI, 0.4-11.3). Likewise, triple-NRTI regimens without ddI were not associated with symptomatic mitochondrial toxicity. The risk of symptomatic mitochondrial toxicity associated with ddI was similar whether patients were receiving 1 or 2 other NRTIs.

Back to Top | Article Outline


The incidence of symptomatic hyperlactatemia in adults receiving at least 1 NRTI has been estimated in 2 prospective studies to be 8 to 20.9 per 1000 patient-years.8,9 In our study, as in HCV-monoinfected patients treated with the IFN-ribavirin combination, we observed no cases of symptomatic mitochondrial toxicity in patients receiving anti-HCV treatment without concurrent antiretroviral therapy. Conversely, the incidence of symptomatic mitochondrial toxicity in patients receiving both anti-HCV and anti-HIV therapy was higher (47.5 per 1000 patient-years, based on 278 patient-years of ribavirin therapy) than that reported in patients receiving anti-HIV treatment alone, suggesting that an interaction between the 2 treatments increases the risk of symptomatic mitochondrial toxicity.

We found that concomitant ddI therapy in patients receiving both IFN (standard or pegylated) and ribavirin was associated with an adjusted 46-fold increase in the risk of symptomatic mitochondrial toxicity compared with patients receiving antiretroviral regimens that did not contain ddI. The incidence of symptomatic mitochondrial toxicity was 200.2 per 1000 patient-years (based on 55 patient-years of ddI and ribavirin therapy), pointing to synergism between ddI with ribavirin. Indeed, although the lack of a no-ribavirin arm theoretically undermines the conclusion that these events are related to ribavirin rather than interferon, in vitro studies had shown that ribavirin enhances both the efficacy and the toxicity of ddI.10-13 Furthermore, all previously reported HIV/HCV-coinfected patients who developed symptomatic mitochondrial toxicity while receiving IFN (standard or pegylated) combined with ribavirin, in addition to anti-HIV therapy, were also receiving ddI.5,6,14-16

In our study the incidence of symptomatic mitochondrial toxicity was not significantly different between patients who were taking both ddI and d4T and those taking ddI alone or combined with NRTIs other than d4T. No cases were observed among patients treated with d4T without ddI, or with a triple-NRTI combination without ddI. This is consistent with the results of an in vitro study showing that mitochondrial DNA depletion in hepatocytic cells due to the ddI-d4T combination was not more severe than with ddI alone.17 Our results are also in keeping with those of a large cross-sectional study, in which ddI-containing regimens doubled the relative risk of hyperlactatemia, whereas the risk was similar with d4T and zidovudine; the incidence in the 2 groups depended on whether the latter drugs were combined with ddI or 3TC.18

Mitochondrial toxicity can produce a broad spectrum of clinical disorders, but initial symptoms (nausea, vomiting, and abdominal pain or, in cases with more insidious onset, fatigue and weight loss) are generally nonspecific.8,19 In our patients, the clinical manifestations usually consisted of weight loss, weakness, nausea, and abdominal pain and were therefore indistinguishable from common adverse effects of IFN-ribavirin combination therapy.1,2 Serum lactate and lipase should therefore be assayed in patients presenting with these manifestations or with more acute symptoms. Attention must also be paid to gradually increasing GGT levels, as this is not a common effect of HCV infection. As in previously reported cases of symptomatic mitochondrial dysfunction, we noted a gradual significant increase in GGT levels (mean 14.35, range 6-30 upper limit of normal (ULN) in 6 of our patients.5,15 The pancreatic toxicity of ddI is dose related: its frequency is 1% to 7% with currently recommended doses, and most cases occur within the first 6 months of therapy.20-22 In our study the observed incidence of pancreatitis in ddI-treated patients was 6%; it might have been higher, however, as amylase and lipase were not routinely assayed during the trial. The median duration of ddI therapy at the diagnosis of pancreatitis was 14 months (range, 3-24 months), suggesting that the addition of ribavirin increased the risk (median duration of HCV treatment: 26 weeks; range, 15-48 weeks).

Back to Top | Article Outline


In this study, no cases of symptomatic mitochondrial toxicity occurred in anti-HCV-treated patients who were not also taking antiretroviral therapy or whose antiretroviral regimen did not contain ddI. Conversely, concomitant treatment with ddI and the IFN-ribavirin combination resulted in a 46-fold increase in the relative risk of symptomatic mitochondrial toxicity. These results, together with other published data, suggest that coadministration of ribavirin and ddI should be avoided; if unavoidable, patients should be monitored closely for signs of mitochondrial toxicity.

Back to Top | Article Outline


The authors thank David Young for editing the manuscript.

Back to Top | Article Outline


1. Manns MP, McHutchison JG, Gordon SC, et al. Peginterferon alfa-2b plus ribavirin compared with interferon alfa-2b plus ribavirin for initial treatment of chronic hepatitis C: a randomised trial. Lancet. 2001;358:958-965.
2. Fried MW, Shiffman ML, Reddy KR, et al. Peginterferon alfa-2a plus ribavirin for chronic hepatitis C virus infection. N Engl J Med. 2002;347:975-982.
3. Carrat F, Bani-Sadr F, Pol S, et al. Pegylated interferon alfa-2b versus standard interferon alfa-2b, plus ribavirin for chronic hepatitis C in HIV-infected patients: randomized controlled trial. JAMA. 2004;292:2839-2848.
4. Fromenty B, Pessayre D. Impaired mitochondrial function in microvesicular steatosis: effects of drugs, ethanol, hormones and cytokines. J Hepatol. 1997;26(Suppl 2):43-53.
5. Lafeuillade A, Hittinger G, Chadapaud S. Increased mitochondrial toxicity with ribavirin in HIV/HCV coinfection. Lancet. 2001;357:280-281.
6. Salmon-Ceron D, Chauvelot-Moachon L, Abad S, et al. Mitochondrial toxic effects and ribavirin. Lancet. 2001;357:1803-1804.
7. Mehta CR, Patel NR. Exact logistic regression: theory and examples. Stat Med. 1995;14:2143-2160.
8. Gerard Y, Maulin L, Yazdanpanah Y, et al. Symptomatic hyperlactataemia: an emerging complication of antiretroviral therapy. AIDS. 2000;14:2723-2730.
9. Lonergan JT, Behling C, Pfander H, et al. Hyperlactatemia and hepatic abnormalities in 10 human immunodeficiency virus-infected patients receiving nucleoside analogue combination regimens. Clin Infect Dis. 2000;31:162-166.
10. Hartman NR, Ahluwalia GS, Cooney DA, et al. Inhibitors of IMP dehydrogenase stimulate the phosphorylation of the anti-human immunodeficiency virus nucleosides 2′,3′-dideoxyadenosine and 2′,3′-dideoxyinosine. Mol Pharmacol. 1991;40:118-124.
11. Harvie P, Omar RF, Dusserre N, et al. Ribavirin potentiates the efficacy and toxicity of 2′,3′-dideoxyinosine in the murine acquired immunodeficiency syndrome model. J Pharmacol Exp Ther. 1996;279:1009-1017.
12. Klein MB, Campeol N, Lalonde RG, et al. Didanosine, interferon-alfa and ribavirin: a highly synergistic combination with potential activity against HIV-1 and hepatitis C virus. AIDS. 2003;17:1001-1008.
13. Kakuda TN, Brinkman K. Mitochondrial toxic effects and ribavirin. Lancet. 2001;357:1802-1803.
14. Moreno A, Quereda C, Moreno L, et al. High rate of didanosine-related mitochondrial toxicity in HIV/HCV-coinfected patients receiving ribavirin. Antivir Ther. 2004;9:133-138.
15. Guyader D, Poinsignon Y, Cano Y, et al. Fatal lactic acidosis in a HIV-positive patient treated with interferon and ribavirin for chronic hepatitis C. J Hepatol. 2002;37:289-291.
16. Summary of safety-related drug labelling changes approved by FDA Center for Drug Evaluation and Research (CDER) Available at: videx. Accessed October 12, 2004.
17. Walker UA, Setzer B, Venhoff N. Increased long-term mitochondrial toxicity in combinations of nucleoside analogue reverse-transcriptase inhibitors. AIDS. 2002;16:2165-2173.
18. Moyle GJ, Datta D, Mandalia S, et al. Hyperlactataemia and lactic acidosis during antiretroviral therapy: relevance, reproducibility and possible risk factors. AIDS. 2002;16:1341-1349.
19. Coghlan ME, Sommadossi JP, Jhala NC, et al. Symptomatic lactic acidosis in hospitalized antiretroviral-treated patients with human immunodeficiency virus infection: a report of 12 cases. Clin Infect Dis. 2001;33:1914-1921.
20. Kahn JO, Lagakos SW, Richman DD, et al. A controlled trial comparing continued zidovudine with didanosine in human immunodeficiency virus infection. The NIAID AIDS Clinical Trials Group. N Engl J Med. 1992;327:581-587.
21. Schindzielorz A, Pike I, Daniels M, et al. Rates and risk factors for adverse events associated with didanosine in the expanded access program. Clin Infect Dis. 1994;19:1076-1083.
22. Dassopoulos T, Ehrenpreis ED. Acute pancreatitis in human immunodeficiency virus-infected patients: a review. Am J Med. 1999;107:78-84.

HIV/hepatitis C virus coinfection; hyperlactatemia; lactic acidosis; pancreatitis; didanosine and ribavirin

© 2005 Lippincott Williams & Wilkins, Inc.