| |
Mitochondria and the Toxicity in HIV Disease and Antiretroviral Agents
|
| |
| |
Modena, Italy May 2005
Written for NATAP by Mike Youle, MD
Royal Free Hospital, Director Centre for HIV Medicine
London, UK
Contributions by Andrea Cossarizza, MD PhD
Univ. of Modena and Reggio Emilia
Modena is known for fast cars. Ferrari and Maserati have their headquarters there and there was no shortage of automatic excitement the weekend of the meetingwith the "Mille Miglia", a classic Italian car rally, passing through. On the outskirts of this town of 200,000 inhabitants lies the Policlinico, which hosted this meeting, which drew the glitterati of the mitochondrial world to discuss where we are with this field which has spurned so much interest. Andrea Cossarizza, the Chair of Immunology of the University of Modena and Reggio Emilia brought together, from May 19-21st 2005, a wide-ranging group of basic scientists and clinical researchers interested in the area of mitochondria and the toxicity engendered by HIV disease and antiretroviral agents.
The opening lecture given by Vladimir Skulachev from Moscow focused on programmed death of the organism and concluded though an elegant series of examples from lower organisms demonstrated that the reduced efficacy and number of mitochondria lead inexorably to deterioration in the function of organisms resulting ultimately in death.
The Friday morning session was opened by Kees Brinkman from Amsterdam, he illustrated his talk by a case of a female Ethiopian who in 1999 developed anaemia, then polyneuropathy and finally after one month on D4T/DDI and saquinavir a fatal lactic acidosis with hepatic steatosis at autopsy. At that time only 10 cases had been reported but as of this month Pub-med showed 140 publications on the subject of presumed mitochondrial toxicity, showing the rise of interest in and importance of the area, although 25% were reviews! Whilst there seems to be a general consensus of the importance of nucleoside analogue reverse transcriptase inhibitors many controversies remain. The link between polymerase-_ inhibition and mitochondrial DNA (mtDNA) depletion is not absolute and in addition the incidence and severity of clinical toxicities appears to vary also, for instance the well known association with DDI and pancreatitis does not link closely with a reduction in mtDNA in those patients, whilst other drugs clearly show a tissue preference such as AZT for muscle and bone marrow. "Not all patients develop toxicity."
The next presentation from Giuseppe Passarino from Cosenza in Calabria gave an overview of mitochondrial DNA and human history where he pointed out that mtDNA variability "has been the best tool to reconstruct the history of human populations". The important facts underlying this statement are that mtDNA is uniparentally inherited and is such a history of human females; that general evolution rate is greater than that seen in mtDNA and that mtDNA escapes recombination and is therefore "more sensitive to genetic drift". Since mtDNA lineages evolve separately from each other and in small isolated human populations mtDNA lineages are easily lost it is possible to track lineages backwards to view the movement of human populations. Essentially there are three groups worldwide L-Africa; M Asia/Africa; N- Europe/Asia and this division occurred around 20,000 years ago. This pattern at least in Europe and parts of Asia was formed by the last Glacial Maximum (25-13,000 years ago) or Ice age, which forced mass movements of populations and often isolated small groups for hundreds of years. Some populations remain as links from one group to another and in fact the Tsunami of 2004 possibly almost eradicated one of the last links between two sub-groups when the Andaman Islands were destroyed. It appears that evolutionary rates are affected by cold climatic conditions when child production rates increase and diet changes.
Why are these mtDNA evolutionary patterns important? One reason is that there is a relationship between mtDNA variability and the efficiency of the OXPHOS energy production pathway. There may also be some association with Parkinson's disease, longevity (for instance there are pockets of extreme long life in Sardinia), diabetes and sperm motility. Finally there may be a different phenotypic (physical expression) of disease according to on which mtDNA molecule (there are two in each cell) the changes occur. With regard to the effect of mtDNA on energy production, and therefore the likelihood of cellular toxicity, there may be feedback mechanisms to abrogate the genetic defects and at least in part increase the efficiency of the system. So in conclusion it would seem that a proportion of the outcomes of mtDNA damage mediated by HIV and antiretrovirals will be also predetermined by our genes.
The underlying message, of the next few presentations, was the complexity of evaluation of changes that occur in general and in specific tissues. This story has been one throughout the history of HIV. Without an antibody test who was infected? Without a viral load test, who had the more serious risk of disease progression; without resistance tests which drugs were no longer useful? So, one challenge is to produce a clear battery of tests to assess mitochondrial function and to assess the consequences of this.
William Lewis, from Emory University School of Medicine, has been a pioneer in this area and presented elegant data on NRTI mt changes as a function of toxicity in HIV disease. He showed the wise range of aetiologies of AIDS related cardiomyopathy and the work he has conducted in a Sprague-Dawley rat model to asses the addition effect of AZT in studies stretching back to 1991. Not only does DNA _-polymerase inhibition have a role but also various issues regarding import phosphorylation with TP-mutants, TK-mutants, TK2 (cytoplasmic) restriction and DNC mutations which result in a predilection for the import of DDNT (AXT into the cell). DNA pools are limited in a number of ways through import restrictions and phosphorylation making the whole process extremely complex. One potential clinical consequence from these rat experiments was that although D4t and to a lesser degree AZT and 3TC caused changes in heart muscle, nelfinavir and efavirenz also showed some changes and so did not further support NRTI sparing as a potential approach to avoid mitochondrial toxicity. He was asked about the importance of the cellular thymidine pool and agreed that this may be one of the key factors, which predicts toxicity.
Paddy Mallon, from Sydney, via Belfast, next gave a review of the methodological issues around some of the assays used in this area, focusing on mtRNA and its' relationship to HIV-associated lipodystrophy. He presented first the data from the work his group had performed in HIV uninfected individuals given 6 week courses of NRTI's. Already by week 2 a reduction in cytochromes had occurred suggesting an immediate functional impairment of energy metabolism although by the time no changes could be seen in mtDNA levels. In fact even after the withdrawal of the NRTI's some marrow suppression remained for at least 12 weeks, suggesting the system takes time to re-equilibrate; by week 66 normal levels of function had been resumed. The rest of his talk focussed on the difficulties of working with RNA, requiring higher tissue amounts, rapid processing and agreeing with the next speaker that it was vital that quality assurance systems are set up to help validate some of these newer assays across different laboratories across the World.
The chair of the conference, Andrea Cossarizza, gave a colourful lecture on Mt RNA as a new marker for drug toxicity and echoed the previous speakers concerning the increasing publication rate in this area. Many methodologies are under evaluation but most focus on the measurement of RNA by real time PCR, which allows for both DNA and RNA to be measured simultaneously.
He presented some data comparing fat samples form normal controls and both HIV infected men and women who during surgical procedures had had sample of fat removed from the back, breast and abdomen. In these fat cells the expression of mitochondrial function as measured by various parameters was reduced in the infected subjects. When asked which of the many tests currently gives the best measure of mitochondrial function he answered that no single test could yet be given that role.
Jordi Casademont from Barcelona continued the theme by discussing some of the other potential methods for assessing the functioning of mitochondrial and or oxidative phosphorylation. He noted the poor link between geno- and phenotype in this area and that due to tissue heteroplasmy and partial deficiencies that occur it may be necessary to use multiple methods to delineate the functional consequences of mitochondrial damage.
He then ran through the series of alternative methods that exist to assess mitochondrial toxicity. These commenced with a wide selection of biochemical markers (of which the most tested are lactate/pyruvate ratios and the quantification of ketone bodies). These are difficult to perform routinely but deserve further studies comparing them with more complex physiological assessments such as modified cardiopulmonary stress testing. He described a relatively simple ischaemic forearm exercise test where lactate produced appears to clear more slowly in HIV infected individuals with mitochondrial dysfunction. Some other techniques, such as the methionine C13 breath test have limitations since they cannot be used in those with underlying liver disease. Finally there is a whole panoply of imaging techniques, for instance PET and SPECT scanning, but these remain prohibitively expensive, even within studies.
Massimo Galli showed the LIPO-ICONA data, which although it consists of a huge (845) representative cohort of 71% men, 38% IDU, 40% HCV suffers from the fact that only 9% received NNRTI-HAART and <1% boosted PI-HAART. However, useful data was gleaned over a 4 year follow-up with 40% of this cohort showing an abnormal glucose over time. He tried to predict if it was possible to predict lipodystrophy. Some factors were predictive; for lipoatrophy a 3.08 RR was seen for IDU (p=0.03) and an 11-fold risk existed for women to develop lipoaccumulation (p<0.001). Other correlates were age and PI use, although there was little boosted-PI data to assess. He showed fascinating early data on a polymorphism (FAS-670), which appeared to predict the development of lipoatrophy, thereby strengthening the argument that TNF-a is involved in the process.
Cristina Mussini from Modena showed data on mtDNA content in platelet-free peripheral blood lymphocytes from patients who underwent a CD4-guided therapy interruption. Thirty patients had been treated for a mean of 107 months; they had a median CD4+ T cell count at discontinuation of 702 cells/mL, and the median observational time from HAART discontinuation was 11.3 months. The analysis of mtDNA content revealed that discontinuation of the treatment provoked a significant increase in mtDNA among CD8+ but not CD4+ T cells, which started only 6 months after therapy discontinuation. This study is the first showing that mtDNA content can increase in peripheral blood lymphocytes, but only after, at least, 6 months of treatment interruption. She concluded that interruptions of shorter periods, due to clinicians or patients decision, are unlikely to allow restoration of mtDNA and thus decrease HAART-related toxicity.
Simon Mallal travelled al the way from Perth in Australia to provide an elegant lecture which attempted to use the know pathogenesis of lipodystrophy to evaluate therapeutic approaches. He likened the evolution of the syndrome to a sieve where genetic and environmental factors are shaken up and then compounded by HIV infection and finally by HAART. As HIV therapy is given for prolonged periods, and the nucleosides at least after cellular metabolism, the initial weight gain seen is gradually negated as DNA depletion and dysfunction takes out the beneficial effects of the drugs (at least in some tissues). Some newer agents, such as abacavir and tenofovir appear, at least in the short-term, to have no effect on the ultrastructure of fat cells, which bodes well for their risk to contribute to lipodystrophy. If damage exists then clinical reversal appears slow and may depend on the degree of end-organ damage. He showed data suggesting the pro-inflammatory state seen in the syndrome is slow to reverse due possibly to macrophage recruitment, with loss of adipocytes, reduced to adipokines (particularly adiponectin). TNF-a is co-localized in macrophages and even with cellular regeneration this pro-inflammatory activity remains.
Mike Youle from the Royal Free Hospital in London described some of the pathophysiological options for the clinical treatment of peripheral neuropathy. The first of these, human nerve growth factor (HNGF), although sadly now not being developed, provided valuable information on how to conduct studies in this disease, with a multifaceted clinical picture with ultrastructural nerve damage which is slow to recover. Therapy is therefore based on a balance of reducing abnormal sensations (mainly pain and paraesthesiae), whilst attempting to reverse nerve damage. The use of fluorescence staining of structural and functional components of nerve fibres within the skin was validated for use in studies of HIV neuropathy within the ACTG study of HNGF. He showed published open label data of marked and persistent improvement in antiretroviral toxic neuropathy (ATN) in a small cohort using acetyl-L-carnitine (ALCAR). This has now been augmented by a placebo controlled study of intramuscular ALCAR showing significant reduction in pain parameters in a two week study. In addition a 50 patient, 48-week placebo controlled trial to assess the potential of ALCAR to both prevent nerve damage and also to alter lipid metabolism in subjects commencing HAART, is almost fully enrolled. Clearly it is vital to assess other agents, such as uridine or alpha-lipoic acid in carefully controlled clinical studies with validated endpoints.
Giovanni Guaraldi presented the metabolic clinic running in Modena, which uses a model of a multidisciplinary approach to the problem of lipodystrophy. In a single day, patients are seen for diagnosis, education and treatment by a health educator, psychologist, dietician, metabolic physician, physical trainer and an HIV doctor. He showed data on 402 individuals (173 women, 289 men) who have follow up from the clinic (107 out to 48 weeks). There were equal numbers of Buffalo humps but lipoaccumulation was seen in 71% of females compared to 47% of men. Lipoatrophy was significantly more common in men (45% vs. 22%, p<0.001) but a mixed picture was more common in women (57% vs. 34%, p<0.001). Smokers had higher rates of metabolic syndrome (98% vs. 77%, p<0.001). However, over follow up little normalisation was seen in metabolic parameters although significant improvements were seen in body image assessment following surgical interventions. Fat and blood has been stored for detailed mitochondrial studies.
A switch study from d4T 40 twice daily to d4T 30 twice daily or to tenofovir was discussed by Ana Milinkovic from Barcelona. All subjects had lipoatrophy and whilst 22 remained on full dose d4T, 19 switched to 30, b.d and 17 to tenofovir 300, once daily. There were no major differences in age (around 44) pre-treatment viral load (250,000) and T4 at baseline (590) or metabolic parameters. By six months a significant reduction of 20% was seen in triglycerides for the tenofovir arm (p=0.039) with no change in the group allocated to 30 twice daily and a rise of 15% in those who continued on d4T 40 twice daily. Similar changes were seen in total cholesterol but only trends in HDL-LDL ratios. Serum lactate levels followed the same pattern with a 25% reduction in the T group (p<0.01), stable levels in the 40-30 switch arm and a 15% rise in the d4T 40 arm; 4 lactic acidosis were also seen in the latter group. As measured be DEXA scan a significant improvement occurred in the tenofovir arm (p 0.003). A rather perplexing median 1000g loss of lean muscle tissue occurred in the tenofovir group with all subjects showing some loss and some discussion took place as to the reasons for this, one suggestion being that the ability of DEXA scans to adequately assess lean mass could be questioned. No changes were seen, however, in any marker of mitochondrial function. Her final comment was that although d4T is less used in the developed world it continues to be a major first line regimen in many developing countries and date on lower doses such as 20, b.d could be important.
Vincent Soriano gave an excellent overview of the potential for viral hepatitis to exacerbate oxidative stress in those already infected with HIV. Not only does HCV appear to cause oxidative stress per se but also activates cell death pathways. In Apricot DDI use was associated with hepatic decompensation and d4T appeared to result in weight loss. In his cohort mitochondrial DNA depletion was significantly greater in co-infected individuals and worse in those taking treatment for HCV.
He closed his presentation with a description of 17 cases of cryptogenic liver disease, a presentation that had also been seen by Paddy Mallon in Sydney and Mike Youle in London. These cases mainly seen in gay males with a median of 15 years of HIV infection all occurred in subjects who had had extensive experience with antiretrovirals (especially ddi/d4T/nevirapine). Fifty-three percent had progressed to severe liver disease (8 ascites, 2 encephalitis, 5 variceal bleeds and 6 partial thromboses). In a case control study DDI use was the only independent risk factor to emerge.
Finally Uri Walker walked through the potential of uridine to avert or treat mitochondrial toxicity. A controlled study is ongoing.
This meeting provided an excellent forum for the discussion of all aspects of mitochondrial dysfunction in HIV disease and produced many more questions than answers. Hopefully improvements in assays, markers and therapeutic options will help to make the field clearer in the coming years. What is clear is that all companies with nucleosides analogues both for HIV and hepatitis should take note of the current science and attempt to address concerns now rather than wait for these toxicities to develop in the future.
Mike Youle
|
|
| |
| |
|
|
|
