| |
Mitochondrial DNA Changes in PBMCs and Lipoatrophy & Nukes
|
| |
| |
"Changes in Mitochondrial DNA in Peripheral Blood Mononuclear Cells from HIV-Infected Patients with Lipoatrophy Randomized to Receive Abacavir"
The Journal of Infectious Diseases August 15, 2004;190:688-692
Jennifer F. Hoy,1,2 Michelle E. Gahan,1,2 Andrew Carr,5 Don Smith,6 Sharon R. Lewin,3,a Steve Wesselingh,4 and David A. Cooper5,6
1Infectious Diseases Unit, Alfred Hospital, 2Department of Medicine, Monash University, 3Department of Microbiology and Immunology, University of Melbourne, and 4Burnet Institute for Medical Research, Melbourne, and 5St. Vincent's Hospital and 6National Centre in HIV Epidemiology and Clinical Research, Sydney, Australia
ABSTRACT
It has been suggested that lipoatrophy associated with exposure to nucleoside analogues is caused by depletion of mitochondrial DNA (mtDNA). The aim of the present study was to determine whether switching treatment from a thymidine analogue to abacavir was associated with an increase in the mtDNA copy number in peripheral blood mononuclear cells (PBMCs). Of 111 patients with lipoatrophy who were randomized to have treatment switched to abacavir or to continue treatment with thymidine analogues, 94 patients had PBMCs obtained at baseline and at weeks 4, 12, and 24, for quantification of the mtDNA copy number. During the 24-week study, there was no significant change in mtDNA copy numbers in PBMCs in either treatment group, despite improvement in peripheral lipoatrophy among patients whose treatment was switched to abacavir.
"... Changes in the mtDNA copy numbers in PBMCs did not correlate with improvement in peripheral subcutaneous fat...There is an urgent need to find a suitable assay with which to measure the presence of, as well as the change in, nucleoside analogue toxicity. Although PBMCs are easy to obtain on successive occasions, they do not appear to be a useful source for the quantitative mtDNA assay, either to predict or monitor changes in lipoatrophy associated with mitochondrial dysfunction resulting from exposure to nucleoside analogues."
BACKGROUND
Combination antiretroviral therapy, which includes nucleoside reverse-transcriptase inhibitors (NRTIs), has had a dramatic effect on HIV-associated morbidity and mortality. NRTIs have been shown to variably inhibit mtDNA polymerase-g, which is the enzyme responsible for replication of DNA in mitochondria. This inhibition of mtDNA polymerase-g, is manifest in vitro as a reduction in the mtDNA copy number per cell. Zalcitabine, didanosine, and stavudine, followed by zidovudine and lamivudine, are the most potent inhibitors of mtDNA polymerase-g. In in vitro assays, abacavir and tenofovir appear to be the least potent inhibitors of mtDNA polymerase. Mitochondrial toxicity has been proposed as the underlying mechanism of a range of NRTI-associated adverse side effects, including fat wasting, lactic acidosis, peripheral neuropathy, pancreatitis, hepatitis, cardiomyopathy, and encephalopathy in neonates.
There is an urgent need for minimally invasive tests to assess mitochondrial toxicity. Venous lactate measurements are not sensitive and specific for identification of early mitochondrial dysfunction. Muscle and liver biopsies are considered to be the reference standard for the evaluation and diagnosis of mitochondrial toxicity in muscle and liver, but they are impractical for routine and repeated evaluations. We developed a real-time polymerase chain reaction (PCR) assay to accurately quantify the mtDNA copy numbers per cell in peripheral blood mononuclear cells (PBMCs). This assay has been shown to give consistent and reproducible results, through internal studies conducted in our laboratory (Neurovirology Research Unit, Infectious Diseases Unit, Alfred Hospital, Melbourne, Australia) and participation in an international quality-control program.
The MITOX study (which involved substitution of abacavir for thymidine analogues in the treatment of HIV-infected patients with lipoatrophy) was performed by the Clinical Trials Network of the National Centre in HIV Epidemiology and Clinical Research (Sydney, Australia), and the results of the study were recently published elsewhere. The aim of the present randomized, controlled study was to determine whether substitution of abacavir for the thymidine analogue (stavudine or zidovudine) resulted in an increase in peripheral subcutaneous fat without a loss of HIV virological control. Switching the thymidine analogue to abacavir was associated with a significant (about 10%) increase in limb fat mass, as measured by a dual-energy x-ray absorptiometry (DEXA) scan, but the increase was not detected clinically during the 24 weeks of the study [8]. Evaluation of the effect of substitution of abacavir for the thymidine analogue component of the antiretroviral regimen, on the mtDNA content in PBMCs, was performed as a 24-week substudy of the main study. The aim of the substudy was to determine whether switching treatment to abacavir would be associated with a significant increase in the mtDNA content of PBMCs that would parallel improvements in lipoatrophy.
RESULTS
The assay showed excellent interassay consistency among the 12 PCR runs performed, as determined by the coefficient of variation calculated from the threshold cycle values of the mtDNA (1.9%) and nuclear DNA (2.4%) standards and 2-control samples as well as the Y intercept (2.5% for mtDNA PCR runs and 1.7% for nuclear DNA PCR runs) and the slope of the standard curves (3.0% for mtDNA PCR runs and 4.6% for nuclear DNA PCR runs). The mtDNA real-time PCR assay had a lower degree of variability, compared with the nuclear DNA assay.
Specimens from 94 of the 111 randomized patients were available for analysis of mtDNA; 39 of the patients had their treatment switched from a thymidine analogue to abacavir (at baseline, 34 of the patients were receiving stavudine, and 5 were receiving zidovudine), and 55 patients continued receiving a thymidine analogue (45 continued receiving stavudine, and 10 continued receiving zidovudine). The characteristics of the 94 subjects, at baseline, according to treatment group, are shown in table 1. There was no significant difference, at baseline, in the mtDNA copy numbers, venous levels of lactate, peripheral limb fat mass, or duration of known HIV infection, either between the treatment intervention groups or between individuals receiving zidovudine or stavudine at baseline (data not shown). Of note, 25 of the 94 patients had an elevated venous level of lactate (>2.0 mmol/L), and hyperlactatemia was significantly associated with administration of stavudine at baseline (P < .001).
There was no correlation between patient characteristics (e.g., age and exposure to stavudine or zidovudine at baseline), lactate levels, and results of mtDNA quantification, for the group as a whole. In particular, there was no correlation between peripheral limb fat mass and quantification of mtDNA in PBMCs at study entry (r = 0.045; P = .668), suggesting that depletion of mtDNA in PBMCs is not a marker of the severity of loss of subcutaneous limb fat mass.
There was no significant difference in the mean mtDNA log copy number between individuals who continued receiving the thymidine analogue that they were receiving at baseline and individuals whose treatment was switched from a thymidine analogue to abacavir at weeks 4, 12, and 24 of the study (table 2). During the 24-week study, there was no significant change from baseline, in the mtDNA copy number in either of the groups. Although there appeared to be a trend for an increasing mean difference in mtDNA between the 2 treatment groups, the difference was extremely small (about 0.1 log). The lack of a significant change in the mtDNA copy number persisted when analysis was done according to whether stavudine or zidovudine was received at baseline and treatment group allocation (i.e., continued receipt of a thymidine analogue vs. a switch in treatment from a thymidine analogue to abacavir).
In the MITOX study, the finding of significant improvement in limb fat mass among individuals randomized to switch treatment from stavudine or zidovudine to abacavir prompted an analysis of the mtDNA copy number in individuals with or without improvement in limb fat mass [8]. A change in limb fat mass did not correlate with a change in the mtDNA copy number in PBMCs (r = -0.035; P = .754; Spearman's correlation). All analyses were repeated only for individuals who had an elevated venous level of lactate at baseline (11 individuals whose treatment was switched to abacavir and 14 individuals who continued receiving stavudine or zidovudine), with similar nonsignificant changes in mtDNA occurring during the 24-week study (data not shown).
AUTHOR DISCUSSION
We were unable to show significant changes in the mtDNA content in PBMCs, during a 24-week study, in patients with lipoatrophy who were randomized to have the nucleoside analogue treatment backbone switched from stavudine or zidovudine to abacavir, despite improvement in peripheral limb fat mass (as measured by DEXA scan). There was no significant correlation between mtDNA copy number and lactate levels, nor was an association found between exposure to stavudine and depleted mtDNA at baseline.
Several other groups have attempted to correlate changes in the mtDNA content of PBMCs with other changes in mitochondrial toxicity. Only 1 group has found that changes in mtDNA can be used as a marker of nucleoside toxicity in HIV-infected patients. This group also found that ratios of mtDNA to nuclear DNA were significantly lower in patients who were receiving stavudine, compared with patients who were not receiving stavudine. We were unable to replicate this finding in the population with lipoatrophy in the present study. We observed no difference in the mean or median mtDNA levels, at baseline, between the groups of patients receiving zidovudine or stavudine.
The findings of the present study are similar to those of McComsey et al, who failed to detect evidence of mtDNA depletion in leukocytes from HIV-infected patients with lipoatrophy or hyperlactatemia who were treated with nucleoside analogues. Other investigators have demonstrated significant depletion of mtDNA in the subcutaneous fat of HIV-infected individuals receiving stavudine or didanosine, without noting concomitant changes in the mtDNA content of PBMCs from the same individuals.
These findings corroborate the findings of other reports that indicate that measurement of mtDNA in PBMCs does not differentiate individuals on the basis of proposed mitochondrial dysfunction resulting in lipoatrophy. There may be a compensatory mechanism available to mitochondria that enables proliferation via a mechanism that is not dependent on the polymerase-g enzyme; this may explain the observation of elevated mtDNA levels in individuals exposed to thymidine analogues.
The present randomized, controlled, open-label study compared the effects of substitution of abacavir for the thymidine analogue component of the antiretroviral regimen in patients with moderate to severe lipoatrophy. The primary end point of the study was changes in subcutaneous limb fat mass, and the outcome at 24 weeks was a statistically significant, albeit clinically small, increase in peripheral fat in the group of individuals whose treatment was switched to abacavir. Measurement of mtDNA in the PBMCs from these patients was found to be precise, with a low coefficient of variation; however, significant changes in mtDNA were not apparent when mtDNA was measured 4 times during the 24-week study, and there was no difference between study groups. It has been suggested that platelet contamination of PBMCs results in falsely elevated mtDNA levels. We did not control for platelets in the present assay; however, all assays from each individual were performed in parallel, and analysis included the change, from baseline, in the mtDNA copy number for each individual. Changes in the mtDNA copy numbers in PBMCs did not correlate with improvement in peripheral subcutaneous fat or with changes in venous levels of lactate. The reasons why quantification of mtDNA in PBMCs has been shown to be an insensitive assay for mitochondrial toxicity expressed as lipoatrophy include the following: (1) mtDNA assays measure mtDNA copy number and not function, (2) lipoatrophy is an organ-specific mitochondrial toxicity, and (3) measurement of mtDNA in fat may have yielded different results. At this point, assay for the detection of mtDNA in PBMCs is clearly less sensitive than a DEXA scan, for assessment of changes in HIV-infected patients with lipoatrophy.
here is an urgent need to find a suitable assay with which to measure the presence of, as well as the change in, nucleoside analogue toxicity. Although PBMCs are easy to obtain on successive occasions, they do not appear to be a useful source for the quantitative mtDNA assay, either to predict or monitor changes in lipoatrophy associated with mitochondrial dysfunction resulting from exposure to nucleoside analogues.
Patients, materials, and methods
A total of 111 patients were recruited from hospital outpatient clinics and community-based primary care. Patients had moderate or severe clinical lipoatrophy noted on physical examination, and randomization to abacavir substitution was stratified according to current use of stavudine, current use of protease inhibitors, and presence of elevated serum levels of lactate. All patients provided written, informed consent, following approval of the study by the research and ethics committee of each study site. The research was conducted according to the principles of the National Statement on Ethical Conduct in Research Involving Humans (from the Australian National Health and Medical Research Council; Canberra, Australia).
Analysis of the effects of the change in treatment from a nucleoside analogue to abacavir was performed on an as-treated basis. Data were included until the time that patients changed the treatment regimen that they were randomized to receive. The mtDNA copy numbers were log transformed to obtain normally distributed data. Changes from baseline were compared between treatment groups, by use of 2-sample t tests. Analysis of variance and regression analysis were used to assess associations between variables at baseline, and Spearman's correlation was used to assess the correlation between the change in limb fat mass, the change in the venous level of lactate, and the change in the mtDNA copy number. All hypotheses were 2 sided, and statistical significance was denoted by P<05. Analyses were performed using SPSS statistical software (version 11.5; SPSS).
|
|
| |
| |
|
|
|
