icon_folder.gif   Conference Reports for NATAP  
  3rd International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV
Athens, Greece - October 2001
Back grey_arrow_rt.gif
Report on Lactic Acidosis and other Mitochondrial Toxicity Issues
Written for NATAP by Cecilia Shikuma, MD, the University of Hawaii and the ACTG
  While the 3rd International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV saw very little that was new and earth-shaking in the area of lactic acidosis and other mitochondrial toxicity issues in HIV, this conference served to confirm many previous findings, better define and enhance our understanding of the disease process, as well as suggest a few new areas of concern that perhaps merit further investigation.
Several abstracts were available on the incidence and natural history of hyperlactatemia and lactic acidosis. Graeme Moyle [Abs 98] presented cross sectional data on lactic acid measurements obtained from patients in his clinic between Aug 1999 to June 2001. Of a total of 4361 lactate measurements performed on 2069 patients, 8.7% had lactates > 2.5 mmol/l while 0.8% (9 individuals) had severe hyperlactatemia defined as > 5 mmol/l. The median lactate in 1239 subjects receiving antiretroviral medications for at least 4 months was 1.4 mmol/l whereas in the population of 253 HIV-positive untreated individuals the median lactate as 1.1 mmol/l. The results of repeat measurements in individuals with high lactates indicated that a single elevated lactate had limited predictive value suggesting that many elevations of lactates were transient. In contrast, in those with normal lactate levels, there was a high predictive value for subsequent normal levels. Events were more common with ddI based regimens, while abacavir was associated with less risk for hyperlactatemia. The use of ABC + 3TC as well as d4T + 3TC was associated with a lower relative risk for hyperlactatemia as compared to ddI + d4T. Elevated glucose, increased ALT and a wider anion gap were associated with hyperlactatemia. Females were over-represented in those with hyperlactatemia. Of interest, of the 9 individuals identified through this process with severe hyperlactatemia, 4 individuals had evidence of acidosis, and interestingly all 4 had evidence of an infectious process (3 respiratory and 1 urinary). Comment: This study highlights the transient fluctuating nature of lactates. High values should be followed by a confirmatory lactate measurement. Processing of the lactate needs to be done according to strict guidelines both in the preparation of the patient (ideally, fasting and with no recent exercise) and in the blood drawing/processing procedure (ideally, blood drawn without a tourniquet, no hand clenching, blood drawn into a chilled gray top [NaF/Kox] tube and processed immediately) to minimize false elevations of lactates. The importance of mildly abnormal lactates in an ASYMPTOMATIC patient remains unclear. The jury is still out on whether such individuals are at risk of developing fulminant lactic acidosis syndrome. The majority of investigators continue to go on record that routine lactates are NOT indicated. Finally, the possible role of intercurrent illness in the pathogenesis of lactic acidosis clearly needs to be studied as such information may have direct clinical relevance. Should future research suggest that intercurrent illness superimposed on a background of mild lactate elevation is adequate to "tip over" an individual into fulminant lactic acidosis, a case could be made for routine lactate monitoring with the intent that extra vigilance for infectious processes would be desirable in those individuals identified to have mild to moderate lactate elevations.
(Editorial comments: lab range for lactate used by Moyle was 1-2.4 mmol/l. Values of >2.5 mmol/l were considered elevated for this analysis. Values >5 mmol/lhave been associated with acidosis in the HIV and critical care settings and were categorized as severe hyperlactatemia. regarding risk for having a raised lactate, Moyle reported AZT+ddI or 3TC, AZT+3TC+ABC and d4T+ABC did not significantly differ from ddI+d4T. He also reported event rates for AZT (11/138), 8%) and d4T (14/145, 9.7%) with either ddI or 3TC were similar. This is contrary to reports from other studies suggesting d4T is a culprit in raising lactate. The event rate for raised lactate (>2.5) was 6.7 for 3TC+AZT, 6.5 for 3TC/d4T, 15.8 for AZT/ddI, and 15.4 for d4T+ddI. This question needs further investigation.
Moyle also suggested that increased lactate without acidosis may result from increased lactate production, increased release or diminshed clearance. Hyperlactatemia may be observed in normal physiologic circumstances, such as during or immediately post-exercise or in hypermetabolic states, and as an accompaniment of pathologic conditions. Hypertriglyceridemia and insulin resistance may lead to steatohepatitis and diminished liver function, hence reduced lactate clearance. Hyperinsulinemia may also increase cellular lactate release).
GA McComsey et al. (Abs 88) submitted an abstract on baseline lactate values, performed by the strict ACTG criteria, obtained on HIV infected subjects randomly selected from a large university clinic who were entering a 2 year longitudinal observational study designed to characterize the natural history of lactate elevation. In their first 250 subjects, 5% of NRTI-treated subjects and 5% of ART-na´ve subjects had lactate levels > 2.4 mmol/l. Significant correlations were found between lactate levels and age, duration of NRTI therapy and duration of d4T therapy. Comments: Utilizing the strict ACTG criteria for drawing/processing of lactate specimens, this study is interesting in that the prevalence of hyperlactatemia is lower than other studies which have suggested a higher prevalence of 10-36%. Even more interesting is the occurrence of hyperlactatemia in individuals not on anti-retroviral therapy. While other drugs and disease states (for ex. liver disease and diabetes) cause lactate elevations, whether HIV infection per se may also contribute to hyperlactatemia require further study.
Tyler Lonergan et al. [Abs 81] presented a poster on the short term (12 week) improvement in lactate levels and ALTs in 16 individuals with symptomatic hyperlactatemia following replacement of d4T with abacavir (ABC) or ZDV. Comments: The efficacy of this "switch" modality of treatment remains to be proven. While short term success in controlling hyperlactatemia is suggested with a switch of d4T use to a "less toxic" NRTI, it should be pointed out that lactic acidosis has been reported with other NRTIs besides d4T and in particular, with ZDV. Thus, the long-term safety of these types of in-class anti-retroviral switches remains to be seen.
G Biesecker et al [Abs 37] of Gilead presented a poster on the evaluation for mitochondrial toxicity (mitochondrial enzyme cytochrome C oxidase and citrate synthase levels and mitochondrial DNA levels) of its company drug Tenofovir DF (up to 6X the normal human dose) in the liver, kidney, skeletal muscle and cardiac muscle of rats (for 28 days) and rhesus monkeys (for 56 days). The investigators concluded that treatment did not affect mitochondrial enzymes or DNA content and that there was no evidence of tissue pathology associated with mitochondrial injury at any dose level for tenofovir DF. Comments: These data are encouraging but, as is the case in all non-human studies of short duration, whether these findings will translate to human cases remain to be seen.
B van Gemen et al [Abs 19] reported on a new rapid duplex real-time NASBA assay, which they utilized to demonstrate decrease in mtDNA in blood PBMCs from a handful of individuals on NRTIs. Comments: Assessments of mitochondrial DNA in blood as well as in other tissues are being conducted in various different ways by different researchers. Simon Mallal of Perth, Australia currently has an on-going effort to try to standardize these techniques. It is unclear at the current time whether assessment of mtDNA levels in PBMCs will ever have any clinical utility. Whether decreases in mtDNA in PBMCs will consistently correlate with hyperlactatemia or whether evidence of such mtDNA decrease gives any more clinical information than a simple lactate measurement remains to be seen.
The data on the effect of storage, processing delays and inter-laboratory variation on plasma lactate concentration (ACTG 5099) was presented in poster form by M. Dube et al. (Abs 53). Three months of storage of NaF/Kox plasma had no significant effect on lactate measurements although at 6 months some meaningful decline was seen. Plasma collected on EDTA and subjected to delays in processing was unacceptable. Comments: The results of this study are important as it indicates that batch analyses, as commonly done in research studies, are possible for lactate measurements. This study disappointingly indicates, however, that banked EDTA (purple top) blood specimens from older studies cannot be used for retrospective analysis of lactates.
C. Shikuma et al (Abs 20) presented data on their re-analysis of mtDNA from a cross-sectional study of individuals with lipoatrophy utilizing a double quantitative-competitive PCR method. Utilizing this method, individuals with lipoatrophy had a 49.8% decrease in mtDNA copies/cell as compared to HIV sero-negative controls. Comments: While limited by the cross-sectional nature of the study and not indicative of a direct cause and effect link, a significant decrease in mtDNA content in subcutaneous adipose tissue in lipodystrophic subjects as compared to control groups suggests the possibility of NRTI-induced mitochondrial toxicity as one etiology for the development of lipoatrophy. Recent epidemiologic data strongly suggests however that there is a substantially increased risk for the development of lipoatrophy in individuals on combined PI and NRTI therapy (Mallal et al. AIDS 2000; 14:1309-1316) while therapy with PIs alone is associated with minimal tendency towards lipoatrophy (van der Valle et al. AIDS 2000; 15:847-855). Thus, one high research priority area in the near future should be the elucidation of the PI-NRTI interaction responsible for this increased risk in the development of lipoatrophy.
Acetyl carnitine deficiency and mitochondrial DNA-depletion in HIV patients with antiretroviral therapy (Abs 63) was studied by G Friese et al. In this study, mean free carnitine, acetyl-carnitine and total carnitine levels were significantly decreased in HIV infected patients irrespective of anti-retroviral therapy. However, in 5 patients with anti-retroviral therapy in whom mtDNA depletion was found in leukocytes as performed by Southern blotting, a even greater decrease in acetyl carnitine deficiency was found. Comments: Recent data from the 8th CROI suggests an association between elevated plasma lactate levels (a more common marker of mitochondrial toxicity) and NRTI-induced peripheral neuropathy. Acetyl carnitine has been suggested as a possible treatment modality for NRTI-induced peripheral neuropathy and the association in this study of mtDNA depletion in leukocytes and lower acetyl carnitine levels may be important if verified in future studies.
(Editorial comments: then we would have to conduct studies to see if taking carnitine supplements actually results in clinical improvements)
Measurement of hepatic b-oxidation by [13C]-octanoic acid breath test was reported to be a reliable tool to assess impaired hepatic b-oxidation in NRTI-treated patients. Abnormal test results, manifested by delayed occurrence of 13CO2 determined at t1/2 > 200 min, were seen in 5/20 studied patients who had either severe lipoatrophy, mild unspecific gastrointestinal symptoms, a history of severe lactic acidosis or manifest lactic acidosis shortly following the study.
Several studies on the effect of d4T and other NRTIs on mitochondrial toxicity from tissue culture to animal studies were presented by Bristol-Myer Squibb. In Abs 61, inhibition of adipocyte function and viability was seen in differentiated adipocytes exposed to d4T + ZDV only at levels 16-30x the plasma Cmax. Furthermore, NRTI inhibited TG accumulation after 7 days of exposure while reduction in mtDNA occurred only after day 20. In BMS's study on undifferentiated adipocytes, no significant change in mtDNA or in the differentiation of preadipocytes was seen upon exposure to up to the top conc of 30 uM of d4T or AZT for 8 weeks. In another BMS study, this time in a mice model (Abs 23), obese (ob/ob) mice were administered d4T (100 mg/kg/day) in drinking water for 6 weeks and underwent fat body mass assessment by DEXA as well as mtDNA content of fat, muscle and liver by real-time PCR. After 6 weeks, no difference in regional fat content by DEXA was seen although a 20% decrease in mtDNA content of adipose tissue was seen. The authors concluded that these findings do not support a NRTI-induced depletion of mtDNA as a mechanism for lipodystrophy in HIV-HAART. Comments: It is unclear whether short term studies (especially in the context of the ob/ob mice studied for only 6 weeks) can mimic what is, in humans, a very chronic condition that often becomes manifest in the 2nd year of HAART therapy.
Ulrich Walker et al [Abs 18] explored the long-term mitochondrial toxicity of pyrimidine nucleoside combinations in HepG2 human hepatoma cell line in 3 different concentrations equivalent to peak plasma levels, 1/3 or 10-fold peak plasma levels. MtDNA depletion was most rapid and pronounced with zalcitabine and declined in the order of ddI > d4T > 3TC = ZDV. ZDV and ZDV/3TC behaved peculiarly in increasing lactate and cell death independent of a decline of mtDNA and without dramatically augmenting intracellular lipids. The toxicity of all the other NRTIs and their combinations was largely related to mtDNA depletion, which preceded or coincided with a decline in COX II-expression, a decrease in cell growth, increased lactate production and increased intracellular lipids. The effects of ZDV/3TC, d4T/3TC, and ddC/d4T were more pronounced than those of either component alone. Cell growth and mtDNA depletion continued to worsen in some assays with no steady state of mitochondrial damage at day 30. No effects were observed with efavirenz. The authors concluded that this in vitro data indicated possible additive or synergistic long-term mitochondrial toxicity of the pyrimidine nucleoside combinations and that mitochondrial damage due to some NRTI-concentrations/combinations can worsen beyond 1 month of incubation. Comment: This elegant study has significance in several ways. Under conditions equivalent to those obtained under clinical conditions, this in vitro study demonstrates that all NRTIs do not behave in the same manner, and that there may be additive and sometimes synergistic toxicity that may occur with some NRTI combinations. Second the possibility of continued worsening mitochondrial toxicity beyond 30 days underscores a need perhaps to investigate the long term consequences of NRTI combination anti-retroviral regimens.
Finally, some observations were presented in abstract form on the possibility of increased symptomatic hyperlactatemia in HIV/HCV co-infected individuals upon treatment with interferon and ribavirin. C Quereda et al [Abs 103] reported that in a series of 43 individuals treated with Peg-IFN-a or IFN-a plus ribavirin, symptomatic hyperlactatemia developed in 3 patients (6.9%) a median of 5.6 months after initiation of therapy. Another abstract [Abs 35] reported new development or possible worsening of lipoatrophy with or without other symptoms of mitochondrial toxicity within the lst 3 months of treatment of IFN (+ ribavirin) suggesting that INF or peg INF could have acted synergistically with anti HIV-NRTIs to develop the observed lipoatrophy. Comments: These observations warrant further investigation to substantiate these concerns.