mtDNA Depletion & Coenzyme Q10
"Mitochondrial DNA Levels in Fat and Blood Cells from Patients with Lipodystrophy or Peripheral Neuropathy and the Effect of 90 Days of High-Dose Coenzyme Q Treatment: A Randomized, Double-Blind, Placebo-Controlled Pilot Study"
Clinical Infectious Diseases Nov 1, 2004;39:1371-1379
Eva Rabing Christensen, Marc Stegger, Søren Jensen-Fangel, Alex Lund Laursen, and Lars Ostergaard
Research Unit Q, Department of Infectious Diseases, Skejby Sygehus, Aarhus University Hospital, Aarhus, Denmark
"... The objectives of this study were to measure the mtDNA level in PBMCs and fat tissue with real-time PCR before and after 3 months of treatment with coenzyme Q (at a dosage of 200 mg daily) and to assess clinical changes in 3 groups of HIV-infected patients receiving HAART: those with lipodystrophy, those with peripheral neuropathy, and those without adverse effects. For comparison, mtDNA quantification was also done in a group of HIV-seronegative control subjects...
... We were not able to find any changes in mtDNA levels after 90 days of treatment with coenzyme Q. In the peripheral neuropathy group, however, we saw a marked aggravation of symptoms shortly after introduction of coenzyme Q therapy in 3 of 3 patients receiving the drug and in none of the patients receiving placebo.
Background. Mitochondrial toxicity can be induced by reverse-transcriptase inhibitors, and an association between levels of mitochondrial DNA (mtDNA) per cell and lipodystrophy, peripheral neuropathy, and HIV infection per se has been suggested. Studies aimed at increasing the oxidative capacity in HIV-infected patients have been sparse.
Methods. Levels of mtDNA in fat and peripheral blood mononuclear cells (PBMCs) from 25 HIV infected patients and 10 healthy control subjects were studied with real-time PCR analysis. A placebo-controlled and double-blind design was used to assign individuals to receive either 100 mg of coenzyme Q twice daily for 3 months or a matching placebo regimen. Levels of mtDNA and other parameters were assessed before and after the intervention period.
Results. The mean number of mtDNA copies per cell was lower in fat tissue obtained from patients with peripheral neuropathy (1547 mtDNA copies/cell; P = .045), patients with lipodystrophy (1732 mtDNA copies/cell; P = .003) and in HIV patients with no complications associated with highly active antiretroviral therapy (2935 mtDNA copies/cell; P = .078), compared with healthy control subjects (6198 mtDNA copies/cell). No clear difference was seen in mtDNA content in PBMCs. Coenzyme Q therapy improved the general condition of patients (P = .005) and caused a reversible increase in peripheral neuropathy pain (P = .048). Compared with placebo, treatment with coenzyme Q did not result in changes in mtDNA levels in fat cells or in PBMCs after the treatment period.
Conclusions. Levels of mtDNA in fat tissue, but not in PBMCs, were associated with peripheral neuropathy and lipodystrophy. High-dose coenzyme Q therapy increased well-being in asymptomatic HIV-infected patients and those with lipodystrophy, as well as in control subjects, but aggravated pain in patients with peripheral neuropathy.
In this study we showed that the mtDNA content per cell in abdominal fat, but not in PBMCs, was lower in HIV-infected patients who had lipodystrophy or peripheral neuropathy, compared with non--HIV-infected healthy control subjects. We also found that treatment with coenzyme Q, a drug theoretically able to increase the oxidative capacity at the cellular level, improved the general condition of patients, generally resulted in increased weight in patients with lipodystrophy, but initially and reversibly aggravated neuropathy symptoms.
The advantage of this study is that data and samples for assessment of mtDNA were collected systematically and prospectively. For the evaluation of the effect of treatment with coenzyme Q, the randomized, blinded, and placebo-controlled design diminished bias.
The major drawback of this study is the relatively limited number of subjects. It was a pilot study, and even with use of a randomized design, it may not have been possible to detect a beneficial effect of coenzyme Q treatment on mtDNA levels of participants. Furthermore, the lack of standardized measuring methods for lipodystrophy and neuropathy limits the study. On the other hand, the blinded and placebo-controlled design contribute to the validity of the results based on subjective data.
Our finding of an association between low mtDNA levels and both lipodystrophy and peripheral neuropathy has also been observed in other studies. There is good evidence that lipodystrophy is caused by the use of NRTIs—in particular, stavudine—and several studies have also shown an association between use of NRTIs and low mtDNA levels. Thus, the observed differences in mtDNA levels in patients with lipodystrophy could be explained by the current choice of NRTI, because stavudine was used by more patients in this group.
Although the finding was not statistically significant, we also found a trend towards a lower mtDNA levels in adipose tissue in patients with stable HAART regimens, compared with healthy control subjects. This could indicate that measurement of mtDNA levels could be a sensitive marker for preclinical mtDNA depletion and predict an increased risk of lipodystrophy or peripheral neuropathy. If such assessments are to be done in subsequent studies of time series, it is important to use fat tissue rather than PBMCs, because we were not able to find any difference when only assessing the PBMCs in this study.
We were not able to find any changes in mtDNA levels after 90 days of treatment with coenzyme Q. In the peripheral neuropathy group, however, we saw a marked aggravation of symptoms shortly after introduction of coenzyme Q therapy in 3 of 3 patients receiving the drug and in none of the patients receiving placebo. The aggravation was reversible. It could be due to regeneration pain, but no mechanism can be resolved on the basis of data from this study. Certainly, subsequent dose-finding studies that include objective monitoring of pain and performance of nerve biopsies are needed.
In the group of patients with lipodystrophy, 1 patient had a marked redistribution of fat with a diminishing dorsocervical fat pad ("buffalo hump"). Among those who did not observe any redistribution of fat, a greater weight change was seen in the coenzyme Q arm, compared with the placebo arm of the lipodystrophy group. Because no dual-energy absorptiometry scans or MRIs were done in this study, we were not able to tell whether this weight gain was distributed as decreases in lipoatrophy or increases in abdominal fat. Another explanation, however, could be that, because patients had a significant ability to place themselves in the correct treatment group, even though there was no significant change between day 1 and day 90 in any of the variables measured at baseline variables, the general well-being improved in the group receiving coenzyme Q, which in turn might have increased their appetite.
Measurement of mtDNA levels in fat biopsy specimens seems superior to measurement of mtDNA levels in PBMCs for predictive purposes. Coenzyme Q treatment has an effect on patients with lipodystrophy or peripheral neuropathy. The exact benefit and the mechanisms behind this benefit, however, will need to be resolved in subsequent studies.
Nucleoside reverse-transcriptase inhibitors (NRTIs) are important drugs in the treatment of HIV infection. To various extents, NRTIs inhibit human DNA polymerase γ, resulting in impaired replication of mtDNA. This results in mitochondrial depletion and dysfunction. Many studies have pointed to this mtDNA depletion as a potential cause of long-term complications seen in HIV-infected patients, such as peripheral neuropathy, lipodystrophy, hyperlactatemia, lactic acidosis, hepatic steatosis, and pancreatitis, and combinations of complications. However, the optimal type of cell or tissue to study has not yet been clarified.
The mitochondria supply the energy (which is stored in the molecule adenosine triphosphate) that is necessary for the bioenergetic processes taking place in the cell. When the level of normal mitochondria falls below a certain threshold, the cell exhibits dysfunction and a decreased capacity to produce adenosine triphosphate. This dysfunction could potentially be reversed by coenzyme Q. This compound is an electron carrier involved in oxidative phosphorylation, which takes place in the inner membrane of the mitochondrion.
Design and inclusion criteria.
The study was a pilot study designed as a randomized, double-blind, placebo-controlled trial. The inclusion criterion was receipt of at least 12 months of stable NRTI treatment with no treatment interruptions during the previous 6 months. Patients were selected if they currently had no adverse effects (i.e., "asymptomatic") or had either lipodystrophy or peripheral neuropathy. Excluded were patients not exposed to any NRTI treatment currently or during the previous 6 months; pregnant women and women who did not use safe contraception; patients with hemophilia; and patients with bowel disease or a history of bowel-reducing operations.
The patients were selected in the outpatient clinic at the Department of Infectious Diseases, Skejby Hospital (Aarhus, Denmark). Of the 25 included HIV-infected patients, 13 had lipodystrophy, 5 had peripheral neuropathy, and 7 had neither lipodystrophy nor peripheral neuropathy. In addition, 10 HIV-seronegative, healthy control subjects were included, who were recruited among the people affiliated with the research laboratory at the Department of Infectious Diseases of Skejby Hospital.
The tablets used in the study, both coenzyme Q and placebo, were provided by Pharma Nord in 100-mg tablets. The tablets were identical in appearance and odor: 1.5 × 0.5 cm in size, dark-brown, oval capsules with a smooth surface facilitating ingestion. The patients and control subjects were instructed to take 2 tablets daily, 1 at breakfast and 1 at dinner, for a duration of 90 days. The participants within the 4 groups were randomized to receive either placebo or coenzyme Q treatment.
In this trial, lipodystrophy syndrome was defined by patient report through a standardized questionnaire definition of lipodystrophy from Carter et al. and matched with the findings of a physical examination. Waist-to-hip ratio and body mass index were calculated before and after the 90-day intervention period. Patients with peripheral neuropathy were recruited from among those who had clinically diagnosed peripheral neuropathy that could not be ascribed to other causes, such as alcohol abuse, diabetes, or peripheral neuropathy that was present before the institution of HAART.
The patients were asked to score their symptoms on a visual analogue scale from 0 to 10, with 0 indicating no symptoms and 10 indicating the worst symptoms. The extent of neuropathy was determined with a standardized questionnaire to evaluate sensory neuropathy in HIV/AIDS (C. Cherry, personal communication) along with the score. The score was done individually for each limb before, during, and after the intervention.
Sample collection and handling.
A blood sample and a biopsy specimen of subcutaneous fat tissue from the abdomen were collected from each patient on day 1 and day 90 of the study. PBMCs were isolated by centrifugation over a layer of Ficoll-Paque Plus (Amersham Biosciences) at 800 g for 20 min at 4°C. Aliquots of 1--2 × 106 PBMCs/mL were stored in PBS at -70 °C until use. A specimen of subcutaneous fat tissue was collected from the abdominal skin over McBurney's point using a sterile biopsy needle. The tissue was frozen immediately in liquid nitrogen and stored at -70°C until use.
Evaluation of the method.
In fat tissue, the mean coefficient of variation for assessment of mtDNA was 15.0%, and for assessment of nDNA it was 14.3%. In blood, the mean coefficient of variation for mtDNA was 14.8%, and for nDNA it was 19.6%. Between different PCR runs assessing the same sample in duplicate, the mean coefficients of variation overall were 52.4% for mtDNA and 34.5% for nDNA. The interassay coefficient of variation in fat tissue was 54.3% for mtDNA and 14.0% for nDNA; the interassay coefficient of variation in blood was 50.4% for mtDNA and 61.8% for nDNA.
Of the 25 HIV-infected patients included in the study, 13 had lipodystrophy, 5 had peripheral neuropathy, and 7 had no symptoms. In addition, 10 healthy control subjects were included. Baseline characteristics were similar between the coenzyme Q group and the placebo group.
Comparison between disease groups.
The ratio of mtDNA to nDNA in adipose tissue differed between the 4 subject groups (P = .005, by analysis of variation): the mean level of mtDNA in subcutaneous abdominal fat was highest in the control group (6198 mtDNA copies/cell; 95% CI, 3089--12,436 copies/cell), followed by the group with no adverse effects (2935 mtDNA copies/cell; 95% CI, 1976--4358 copies/cell; P = .078), the lipodystrophy group (1732 mtDNA copies/cell; 95% CI, 1051--2853 copies/cell; P = .003), and the peripheral neuropathy group (1547 mtDNA copies/cell; 95% CI, 262--9136 copies/cell; P = .045). Furthermore, the ratio of mtDNA to nDNA in adipose tissue was significantly lower in the patients treated with stavudine (mean mtDNA level, 1081 copies/cell; 95% CI, 515--2269 copies/cell), compared with the patients treated with zidovudine (mean mtDNA level, 2583 copies/cell; 95% CI, 1916--3481 copies/cell; P = .007). No such difference in mtDNA levels could be demonstrated in blood samples from these groups. The limited number of patients did not justify a multivariate analysis. The difference found for the disease groups may therefore be attributable to the higher rates of current stavudine exposure in these patient groups.
Mean triglyceride levels were significantly higher in both the peripheral neuropathy group (2.86 mmol/L; 95% CI, 1.64--5.0 mmol/L; P = .002) and the lipodystrophy group (2.15 mmol/L; 95% CI, 1.57--2.92 mmol/L; P = .008), compared with in the control group (1.22 mmol/L; 95% CI, 0.90--1.63 mmol/L). The mean cholesterol level was found to be significantly lower in the control group (4.78 mmol/L, 95% CI, 4.27--5.35 mmol/L) than in the group with no adverse effects (5.69 mmol/L; 95% CI, 4.84--6.69 mmol/L; P = .048) and the peripheral neuropathy group (5.99 mmol/L; 95% CI, 4.85--7.39 mmol/L; P = .028); the lipodystrophy group had a mean cholesterol level that was not significantly higher (5.53 mmol/L; 95% CI, 4.77--6.41 mmol/L; P = .12).
The mean low-density lipoprotein level was significant higher in the lipodystrophy group (3.36 mmol/L; 95% CI, 2.56--4.40 mmol/L; P = .039) and among the asymptomatic subjects (3.57 mmol/L; 95% CI, 2.75--4.64 mmol/L; P = .038), compared with the control group (2.35 mmol/L; 95% CI, 1.84--2.99 mmol/L). The lack of additional low-density lipoprotein values did not permit us to do statistical analysis in the peripheral neuropathy group, but the single value was in the same range as the mean for other 2 patient groups, suggesting there was a similar level in the peripheral neuropathy.
The mean high-density lipoprotein level was lowest in the peripheral neuropathy group (1.08 mmol/L; 95% CI, 0.73--1.58 mmol/L; P = .065), with trend toward ascending values in the lipodystrophy group (1.2 mmol/L; 95% CI, 1.0--1.45 mmol/L; P = .039), the group with no adverse effects (1.36 mmol/L; 95% CI, 0.97--1.90 mmol/L; P = .29), and the control group (1.67 mmol/L; 95% CI, 1.25--2.22).
There was a trend toward a lower lactate level in the control group (1.12 mmol/L; 95% CI, 0.89--1.37 mmol/L) and the asymptomatic patients (1.13 mmol/L; 95% CI, 0.98--1.29 mmol/L; P = .57), compared with the lipodystrophy group (1.48 mmol/L; 95% CI, 1.14--1.91 mmol/L; P = .08) and the peripheral neuropathy group (1.60 mmol/L; 95% CI, 0.86--3.0 mmol/L; P = .08). There was no difference in glucose levels between the 4 groups.
In the coenzyme Q arm, none of the blood parameters showed a significant difference between day 1 and day 90 in the lipodystrophy, asymptomatic, and control groups. In one patient with lipodystrophy who was randomized to receive coenzyme Q, a remarkable redistribution of fat occurred, whereas no redistribution of fat was observed in the 5 other patients with lipodystrophy randomized to receive coenzyme Q. The latter 5 patients, however, experienced an average weight gain of 〜3 kg (P = .011), whereas the 7 patients receiving placebo had an average weight gain of 〜0.2 kg.
There was no significant difference in weight from day 1 to day 90 in the coenzyme Q group or the placebo group among HIV-infected patients with no adverse events and among the healthy control subjects.
After the trial was closed but before unblinding the treatment given to the lipodystrophy, the asymptomatic, and the healthy control groups, the participants in these 3 groups were asked to state if they noticed any relevant changes after 3 months. Twenty-five answered this question, and, of the 13 who received coenzyme Q, 9 (90%) reported improvement, in contrast to only 1 of 12 who received placebo (P = .0054; 95% CI, 0.342--0.924). The HIV-infected patients with peripheral neuropathy were not included in the test.
Patients with peripheral neuropathy.
The first patient's case shows the patient's degree of peripheral neuropathy according to the visual analogue scale score before, during, and after the intervention with coenzyme Q. Because symptoms were aggravated in 3 of 5 patients, we were requested to break the code for this group and stop recruitment. All 3 patients with aggravation received coenzyme Q, and the remaining 2 patients with no changes received placebo.
Patient 1 experienced an aggravation in peripheral neuropathy symptoms beginning 24--48 h after intake of the tablets given. The symptoms escalated both in intensity (from 2 to 5 on the visual analogue scale) and in extent over the next 7--8 days. Forty-eight hours after cessation of coenzyme Q therapy, the symptoms returned to the same level as they were before the patient entered the study. Similar worsening in symptoms (visual analogue scale score, 4--5) occurred when the patient was administered half the daily dose (i.e., 100 mg); symptoms returned to baseline level 24--48 h after cessation of treatment. Patient 2 experienced an aggravation in peripheral neuropathy symptoms beginning 12--24 h after intake of the tablets given (visual analogue scale score increase from 4 to 8--9); the exacerbation in symptoms disappeared within <24 h after the patient stopped taking the tablets.
Patient 3 experienced a light aggravation in peripheral neuropathy symptoms during the first week of treatment (visual analogue scale score increase from 7 to 8--9). The patient continued to take the tablets for 90 days, and peripheral neuropathy symptoms returned to baseline level (visual analogue scale score, 7) after 2--3 weeks. Patients 4 and 5 did not note any changes in their peripheral neuropathy symptoms.