 |
IAPAC Monthly - Vol. 8, No. 11, November 2002
Mark Mascolini
Introduction
From bench to bedside
From bedside to bench
Hard Science
Hard Arteries
More good lipid and insulin scores with atazanavir
Lipid subclass quiz (correct answer: efavirenz)
Lipid liabilities with d4T
HAART, the heart, and other muscles
Hard Choices
Switching from d4T for atrophy and high lactates
Diet and exercise: those who can, win
Options for facial atrophy
Alendronate for thin bones
Statins popular, but how potent?
APPENDIX A. Appendix
APPENDIX B. Bone
References and Notes
Every morning from 7 till 9 San Diego's Hotel Del Coronado ("the Del" to locals) hosts an allegory on risk taking in scientific research. Of course the Del's crackerjack convention staff is much too savvy to call it that. Instead they present the allegory in the guise of a "lap swim" in the big shoreside pool.
This works out well because most of the Del's vacationing guests are still abed at 7 AM, and most of them don't like to swim in the fog, the prevailing atmospheric at that hour on the Pacific coast. But plenty of hydrophilic meeting attendees, stripped to their hydrophobic hides, show up to put in their laps—including a few from the 4th Lipodystrophy Workshop. After four or five jump in, the staff's clever planning becomes clear. No, the towel boy isn't smiling because he's paid to; he's smiling because the allegory of scientific research has begun to unfold.
Instead of roping off the pool's entire length in lap lanes, the staff deploys only two lines, yielding two lanes. Further, they have painted no lap lines on the bottom. Two swimmers, unnerved by the foggy, lineless pool, head straight for the lanes and stay there, crawling contentedly. A slightly more daring duo follows, one sighting off the outside lap rope, the other off the far wall, to stay on track. The most adventurous plunge into the pool's chartless center, hoping to steer a straight course despite fog, lack of guidelines, and the pother of less bold paddlers. Some of them swim straight as dolphins. Others meander aimlessly. Late arrivals jump cheerfully into the froth, contributing nothing but congestion.
Though tender in years, the towel boy can probably guess which swimmers will play it safe, which will push the envelope just a little, which will take the boldest plunge, and maybe even which will succeed when they do. He's seen enough fog swimmers to spot the telltale traits. It's not as easy guessing which scientists will set a true and steady course, though one might discriminate the daring from the diffident.
Research into the metabolic disturbances of HIV and antiretrovirals has reached the point where players abound. Some favor the laid-back backstroke, others the gaudy butterfly, and others the relentless Australian crawl. Some stick to established lanes; some swim uncharted waters. A few pace setters have emerged, even snagging some early trophies. But it's much too soon to tell which of the boldest bathers have the grit, guile, and good luck to prevail.
The quickening pace of their exploits set the tone at this year's Lipodystrophy Workshop, where basic research commanded the center lanes for the first time in this meeting's four years (Table 1). Whereas cohort studies, case series, and clinical trials accounted for a big majority of slide talks in 1999, 2000, and 2001, they filled fewer than one third of the slide slots in 2002. Meanwhile, basic research— defined here as studies in engineered cells—surged from a bare 5.5 percent of talks in 1999 (one presentation) to 41 percent this year (12 talks). On top of that, three other PowerPoint excursions featured statistical analyses or assays, and two probed animal models of drug
toxicity.
People enduring the metabolic malfeasance of antiretrovirals, and the clinicians who care for them, may see this fountainhead of research in two ways. On the one hand, it means the study of antiretrovirallinked metabolic disorders has matured. Now the unglamorous grunt work of cell sorting and real-time fluorescence can eke out data that may define actual mechanisms, instead of cohort-calculated "associations." And finding mechanisms not only points the way to remedies, it also helps define risk, progression, and prognosis.
On the other hand, studies of 3T3- F44A2 cells sure lack the sizzle of a big randomized trial pitting drug suspects A and B against drug suspects X and Y, or even a tidy case-control exercise that ferrets out risky dietary indiscretions or genetic jeopardies. Of course clinical trials and other human studies must rest on hypotheses, and well-reasoned hypotheses often rely on primal benchwork. Yet understanding clinical trials can seem (the operative word) so much easier than understanding what a drug may or may not do to a dishful of hepatocytes. When signifi- cantly more people taking certain drugs in an adequately powered randomized trial end up with a side effect of interest, prescribers make decisions. But in cell studies questions can outnumber conclusions:
Worrying too much about these pitfalls would freeze research on the starting block, but the pitfalls can't be ignored. Take, for example, the ambitious study by Jacqueline Capeau (Saint-Antoine Hospital, Paris), an acknowledged leader in this field who tied lipoatrophy in people taking PIs to lower levels of sterol regulatory element binding protein-1 (SREBP-1).1 Her new study sought to discern the effect of didanosine (ddI) and stavudine (d4T) on adipocytes when given with indinavir [abstract 9, L6*]. Among the findings:
Ultimately these findings may say more about the interactions of nucleoside reverse transcriptase inhibitors (NRTIs) and protease inhibitors (PIs) than the bestdesigned multinational placebo-controlled trial, because the findings may explain not only what happens, but how. Independent confirmation of the results would make them even more compelling.
But clinicians can't look at these data and decide, for example, that ddI/d4T plus indinavir makes sense because that lipoatrophic risk may be offset by appeased insulin resistance. A representative of these NRTIs' maker was prompt to observe that the peak concentrations achieved with 10 µM of each drug may reflect peaks in humans. But no human goes through the day with peak drug levels. Session cochair Steven Grinspoon (Massachusetts General Hospital) noted that technicians administered the rosiglitazone along with ddI and d4T in these experiments, not as a clinician may someday use rosiglitazone, months and months after an NRTI's lipoatrophic effects become evident. True, Capeau replied, but her bench-to-bedside cogitation had already clicked to the next implication: Should people start taking glitazones when they start taking potentially lipoatrophic drugs?
Leagues away on the other side of the research spectrum sat a single case report proffered by Stefan Mauss (Center for HIV and Hepatogastroenterology, Düsseldorf). In its humble way, this simple, stark case study probably speaks more directly to HIV clinicians than the most elegant and provocative lab work. Mauss chronicled the story of a trim young man (there is a picture) who took two courses of postexposure prophylaxis after risky sex with his HIV-infected partner [abstract 77, L51].
He did not become infected. But six weeks after his second three-week course of d4T, lamivudine (3TC), and efavirenz, his abdomen ballooned into the classic picture of central lipohypertrophy. CT scans disclosed a large depot of visceral adipose tissue. Although lipids and blood glucose remained normal, the central fat buildup persisted to the time of Mauss's report, five months after the man took his last antiretroviral. Mauss advanced the case as "a limited proof of principle that the changes in adipose tissue [characteristic of lipohypertrophy] are caused by antiretroviral treatment rather than HIV infection itself."
Perhaps just as disturbing as this sudden- onset lipohypertrophy in an antiretroviral- treated man without HIV infection are the drugs he took—two nucleosides and the nonnucleoside (NNRTI) efavirenz, not a PI. Is the case just a wretched quirk? Or will inspired benchwork have to sort out how six weeks of d4T, 3TC, and efavirenz may rile adipocytes? The case seems unlikely to touch off a wholesale abandonment of efavirenz. But it shows how an utterly straightforward, n-of-one study can simultaneously make a shocking clinical point and raise questions that can be fathomed only by deep-diving basic research.
Results of large clinical trials can be less ambiguous than Stefan Mauss's disturbing case report. Take, for example, AIDS Clinical Trials Group (ACTG) protocol A5005s, reported by Michael Dubé (Indiana University, Indianapolis) [abstract 27, L18]. Although a mere substudy of ACTG 384, A5005s had 330 participants randomized to take zidovudine (ZDV)/ 3TC or ddI/d4T plus nelfinavir, efavirenz, or both. Intent-to-treat and on-treatment analyses showed a significant drop in DEXA-measured limb fat in the ddI/d4T group after 48, 64, and 80 weeks of followup. Both the ddI/d4T group and the ZDV/3TC group, who started the study with no antiretroviral experience, gained limb fat in the first 16 weeks, then started losing fat. At week 48 the limb fat in the ddI/d4T group was down 0.28 kg (7.5 percent), while fat in the ZDV/3TC group still stood 0.26 kg above baseline (4.7 percent) (P = 0.027). The ZDV/3TC group continued to lose peripheral fat through week 80, but the difference from baseline had not reached statistical significance at that point.
People assigned to nelfinavir lost a median 0.92 kg of limb fat (18 percent) by week 80, a significant change from baseline (P = 0.019). People randomized to efavirenz lost a median 0.50 kg (10.7 percent) at that point, but that change from baseline lacked statistical significance. Whites did not differ from nonwhites in limb fat changes. Trunk fat tended to rise in all treatment groups. Total and lumbar bone mineral density did not change significantly in any group.
The results of this randomized trial add to the pile of evidence implicating d4T in peripheral lipoatrophy. And more such evidence came in the next report, a longitudinal study of 27 people starting their first triple-drug regimen with d4T and 26 starting with ZDV [abstract 28, L18]. Measuring limb fat as a percentage of body mass index, David Nolan (Royal Perth Hospital) used DEXA scans to chart a drop from 22 percent to 13 percent after two years in the d4T group (P < 0.005) and from 22 percent to 19 percent in the ZDV group, a nonsignificant change. Equivalent proportions of both groups took PIs or non-PI regimens, and limb fat changes did not differ between PI and non-PI groups.
Studies like these, and others showing compound toxicities with ddI/d4T,2 led clinical investigator Judith Currier (University of California, Los Angeles) to propose removing ddI/d4T from guideline lists of preferred nucleoside pairs.
At the same time, David Nolan reminded Workshop attendees, one can't ignore nondrug factors that promote lipodystrophy. In another study he con- firmed earlier findings3 that a polymorphism in tumor necrosis factor alpha (TNF-α) spurs a quicker onset of lipodystrophy [abstract 26, L17]. He found the TNF-α promoter polymorphism 238G/A in 13.1 percent of 220 Western Australian Cohort members, all of them white. A Cox proportional hazards analysis figured that heterozygosity for the polymorphism raised the risk of clinically identified lipodystrophy 1.73 times (P = 0.041) and did so independently of age, duration of NRTI therapy before highly active antiretroviral therapy (HAART), or duration of d4T therapy.
How this polymorphism promotes lipodystrophy remains an open question. But TNF (or cachectin) has a long-established role in wasting (or cachexia). And work by Pere Domingo (Hospital de la Santa Creu i Sant Pau, Barcelona) traced a link between adipocyte death and circulating levels of two soluble TNF receptors in 45 HIV-infected men and women with lipoatrophy [abstract H-1920]. Receptor levels were significantly higher in those with diffuse apoptosis than in those with moderate, focal, or no apoptosis. In another study, though, levels of soluble TNF receptors did not correlate with lipoatrophy in people with HIV (see the next section).
The four related studies just summarized span the gamut from straightforward clinical trial, through simple longitudinal analysis, back to a bed-lab interface exploring genes or cytokines. The murky origins of metabolic disorders, and their miasmatic evolution—especially in people taking toxic drugs—mean HIV clinicians must grapple with the hard science, get knee-deep in adipocytology, and recall lost lessons in dyslipidemia, never forgetting two decades of revelation in immunology. And what better launching pad than leptin?
On day one of the Lipodystrophy Workshop, no word leapt from more speakers' lips than leptin. Discovered only eight years ago, touted and dismissed as a panacea for obesity, this potent peptide has become the darling of non-HIV lipodystrophy research. Inevitably, leptin landed on the HIV research agenda, and workshop organizers allotted it a keynote lecture, a plenary lecture, and a slide talk—all in the meeting's first session.
O'Rahilly summarizes things this way6:
"Thus," O'Rahilly concludes, "we now know of two distinct human diseases where the correction of near aleptinemia is of major therapeutic benefit."
A study by Grace McComsey (Case Western Reserve University, Cleveland) may offer the clearest hint so far about the potential for leptin replacement in HIV-infected people with lipoatrophy [abstract 78, L51]. Her retrospective analysis of 137 people starting their first antiretrovirals suggested leptin replacement will not help most lipoatrophic people with HIV because low leptin probably did not cause their atrophy.
After a median follow-up of 1,045 days, 40 people in this group had clinically apparent lipoatrophy, and after a median 743 days, 97 did not. In a univariate analysis, several factors, including lower leptin, favored emergence of lipoatrophy:
But when McComsey parsed those parameters in a multivariate analysis, only older age, treatment duration, and lower pretreatment body mass index independently predicted lipoatrophy. The results, she concluded, do not support an independent role for leptin in lipoatrophy, perhaps because the low leptin levels merely reflect the lower pretreatment body mass index in the lipoatrophy group. Some people with lipoatrophy clearly had severely low leptin levels, as low as 1 ng/mL. But the average leptin level in the whole lipoatrophy group, 5 ng/mL, falls above the severe deficits measured in seronegative people who have benefited from leptin replacement so far. Levels of soluble TNF receptors, which correlated with fat cell death in Pere Domingo's study (above), did not correlate with atrophy in McComsey's analysis.
Leptin wasn't the only item on the workshop's basic research agenda. Given McComsey's findings, it may not even be the most important for people with HIV lipodystrophy. Several other studies stood out:
i. Genesis
Genesis of fat cells slowed to a crawl when Simon Jones (University of Liverpool) daubed 3T3-F44A adipocytes with nelfinavir, saquinavir, or ritonavir, with or without ZDV or d4T [abstract 3, L2]. With the method Jones used, measuring glycerol-3-phosphate dehydrogenase, indinavir curtailed adipogenesis less. But indinavir synergized with ZDV or d4T in slowing adipogenesis. The PIs also boosted levels of TNF-α and IL-6, which vex adipocyte metabolism. Jones's colleague Omar Janneh showed that ZDV and d4T phosphorylated to their active metabolites in cultured fat cells, but they didn't inhibit adipogenesis or promote lipolysis—the spilling of fatty acids from adipocytes into the circulation [abstract 2, L1]. PIs slowed glucose uptake in fat cells, in this order: indinavir > saquinavir > ritonavir > nelfinavir.
ii. Lamentation
Jacqueline Capeau's work shows that altered expression and nuclear localization of a transcription factor, sterol regulatory element-binding protein-1 (SREBP-1), may explain why indinavir stymies adipocyte differentiation and response to insulin.1 Pursuing these leads in 3T3-F44A2 cells, she found that indinavir stifles maturation of lamin A/C, which interacts with SREBP-1 and makes up part of the fibrous web lining the shell of fat cell nuclei [abstract 1, L1]. The resulting disruption of nuclear architecture could alter adipocyte function. People with a much-studied non-HIV insulin-resistant lipodystrophy—familial partial lipodystrophy—have mutated lamin A/C.
iii. Over PPAR
Comparing subcutaneous fat from the thighs of antiretroviral-treated people with lipoatrophy, from antiretroviral-naïve people without lipoatrophy, and from healthy controls without HIV infection, Ellen Li (Washington University, St. Louis) found significantly higher levels of peroxisomal proliferator-activated receptor gamma (PPAR-γ) in the lipoatrophic individuals than in the other groups [abstract 4, L3]. (Insulin sparks formation of new fat cells via PPAR-γ.) Levels of lipoprotein lipase (a fat-splitting or lipolytic enzyme) also proved significantly higher in the lipoatrophy group than in the other two groups.
iv. Idle islets
Beta cells live to secrete insulin in the pancreas. They lie in the storied islets of Langerhans, and their failure foments insulin resistance. Measuring glucosestimulated insulin secretion in mouse and rat islets and a murine cell line, Joseph Koster (Washington University, St. Louis) found a dose-dependent muffling of insulin release with indinavir [abstract 6, L4]. Stopping indinavir unleashed insulin.
v. Zucker rats undone
Zucker rats become hyperglycemic at five to 10 weeks of age. By the time they're 14 weeks old, they have diabetes. Kevin Yarasheski (Washington University, St. Louis) shortened their disease-free days even more by giving them NRTIs (ZDV/3TC, ddI/3TC, or d4T/3TC), indinavir alone, or indinavir plus two NRTIs [abstract 7, L4]. Compared with rats taking placebo, those taking two NRTIs had significantly higher glucose levels by week six to seven. Adding indinavir to the NRTIs significantly boosted glucose by week three.
vi. Lesion lessons
PIs may promote atherosclerotic lesions without boosting lipid levels in plasma. That surprising proposal evolved from a study of mice who drank water spiked with low-dose amprenavir, indinavir, or ritonavir for eight weeks and ended up with significant cholesterol jumps inside peritoneal macrophages [abstract 11, L7]. The mechanism may be up-regulation of the CD36 receptor, which Richard Greenberg (University of Kentucky, Lexington) documented in both a human cell line and in the mouse macrophages. After eight weeks, atherosclerotic lesion area proved significantly higher in PIexposed mice than in those who swilled only the PI-transporting vehicle. But similar studies by Sharon Walmsley (Toronto General Hospital), reported last year in AIDS,8 found that PIs lowered expression of CD36 in human monocytes and in cell lines.
vii. Art and artifice
Walmsley's discrepant finding (Greenberg attributed it to differing methods) was not the only challenge to these and other reports of basic research at the Lipodystrophy Workshop. Jacqueline Capeau, for example, doesn't know why SREBP-1 lands in the right nuclear niche in half of the adipocytes she studies but not in the other half. Omar Janneh acknowledged that his cell studies may have been too brief to spot the slowlyevolving mitochondrial toxicities of NRTIs. Kevin Yarasheski noted that the high antiretroviral doses he needed to outpace rapid rat metabolism (10 times the human dose) make it more difficult to extrapolate his results to humans.
But artificial experimental conditions cannot be divorced from the art of research—basic or clinical. Even big placebo-controlled trials set exclusion and inclusion criteria that can't be tacked to an HIV clinic's door. ("People with a CD4+ count under 100 cells/mm3 or HCV infection, please apply elsewhere.") And trial participants typically benefit from more rigorous follow-up and high-tech monitoring than harried clinicians can provide. Stepping down a rung to the cohort study, whether longitudinal or cross-sectional, grays outcomes even more.
No finer example of cross-sectional ambiguity emerged in the past year than divergent results of the EMEA lipodystrophy case-definition study9 and the Fat Redistribution and Metabolic Change in HIV (FRAM) Study.10 Both were crosssectional. Both compared hundreds of lipodystrophy-afflicted "cases" with lipodystrophy-free "controls." Both pinpointed peripheral lipoatrophy as a distinguishing hallmark of HIV lipodystrophy. But the case definition study confirmed clinical impressions that central fat gains also separate people with HIV lipodystrophy from those without it, and FRAM did not.
Why the difference? A first step toward resolving this conundrum came at a roundtable scheduled in conjunction with the Lipodystrophy Workshop by the Forum for HIV Collaborative Research. A summary of that meeting will appear on the Forum's Web site, http://www.hivforum. org. But no one needs deep thinking and tension-packed panels (a fair description of the Forum session) to discern the glaring difference between these cross-sectional exercises. The case definition study compared HIV-infected people without signs of lipodystrophy and HIV-infected people with one or more signs. FRAM compared people with HIV infection, randomly selected without regard to fat changes, and a non-HIV cohort in a big heart disease study.
Experimental choices, mind-numbing analyses, and consequently foggy findings can make human studies every bit as challenging as hard-core science. The lingo may be more familiar, but the doubts linger. Anyone who thinks otherwise will have an easy time with the following clinical questions:
Please answer in 50 words or less before reading the rest of this article.
Few questions have spawned bigger HIV cohorts than whether the virus or antiretroviral therapy heightens the risk of cardiovascular disease. The five largest cohort probes so far came up with different answers to this question. Two linked PI therapy to higher rates of myocardial infarction,11,12 and three did not.13-15 But two of the latter three studies did see more heart disease in people with HIV infection than in those without it.13,14 At the Lipodystrophy Workshop, and a few days later at ICAAC, these findings emerged:
The record review involved 28,513 people with HIV infection who had no episode of coronary heart disease for at least one year after enrolling in California's Medi-Cal program [abstract 54, L37]. Splitting the cohort into those whose records showed antiretroviral prescriptions and those whose did not, Judith Currier (University of California, Los Angeles) figured the risk of coronary heart disease in a multivariate analysis controlling for diabetes, hyperlipidemia, kidney disease, and hypertension. She also split the cohort into four age groups: 18 to 33 years, 34 to 49 years, 50 to 65 years, and 66 years or older. Only in the youngest group did antiretroviral therapy raise the risk of heart disease, 2.06 times (P < 0.001). Still, the incidence of heart disease among 18- to 33-year-olds was low: 1.08 cases per 100 patient years. In that age group and all others, hyperlipidemia, kidney disease, and hypertension independently raised the heart disease risk. Diabetes upped the risk in the three youngest age groups.
Currier did not try to define a link between PI therapy and heart disease, although two thirds of the 18- to 33-year-olds taking antiretrovirals had taken a protease drug. Three quarters of the three older age groups with treatment experience had taken PIs. A limitation of this analysis, Currier observed, is its inability to pin down other important variables, including smoking status, family history of heart disease, and precise lipid levels.
Yet other cohort comparisons suggest that men taking antiretrovirals don't differ much from the American men at large in cholesterol readings. In a plenary talk at the Workshop, statistician James Neaton (University of Minnesota, Minneapolis) compared four cohorts—two big heart risk studies in the general population, MRFIT and NHANES, and two CPCRA trials, SMART and FIRST. SMART enrolls antiretroviral-experienced people and FIRST recruits people naïve to antiretrovirals. Graphing cohort cholesterol levels of men by age, Neaton found that the lines for MRFIT, NHANES, and SMART essentially overlap. Despite their antiretroviral experience, SMART participants had cholesterol scores no worse or better than age-matched men from the general population. But the treatment-naïve FIRST men had cholesterol tallies substantially below those of men in the other cohorts. For example, the average reading for 40- to 44-year-old FIRST enrollees measured 160 mg/dL, compared with 200 to 210 mg/dL in the other cohorts.
These findings parallel pre-HAART data showing that HIV-infected people with and without AIDS had significantly lower cholesterol levels than uninfected controls.16 Whatever the effects of HIV and antiretrovirals on lipids, Neaton reminded attendees, the ties between cardiovascular risk, HIV infection, HIV therapy, and individual patient traits remain hard to untangle. Although "cardiovascular events" now comprise only a small percentage of "adverse events" due to antiretrovirals, he counseled, the long-term cardiovascular consequences of therapy still lie somewhere over the horizon [see note 17]. And because many people taking antiretrovirals have histories thick with heart risk factors, the other side of the horizon may not be pretty.
More good lipid and insulin scores with atazanavir
Two 48-week studies of atazanavir in treatment-naïve people showed that this once-daily PI doesn't share the lipid-inflating potential of other protease drugs.18 At the Lipodystrophy Workshop, follow-up in one of those studies showed that atazanavir can flatten lipid spikes inspired by nelfinavir [abstract 15, L10]. Robert Murphy (Northwestern University, Chicago) reported that people randomized to 1,250 mg of nelfinavir twice daily in an open-label comparison with once-daily atazanavir had a 25 percent leap in total cholesterol, a 23 percent gain in low-density lipoprotein cholesterol (LDL-C), and a 50 percent jump in triglycerides after 48 weeks. People randomized to 400 or 600 mg of atazanavir once daily, on the other hand, had only 5 to 8 percent elevations in these lipids through week 48.
After week 48, people taking nelfinavir could switch to 400 mg of atazanavir. Twelve weeks later, fasting lipids in switchers had dropped significantly (Table 2). Before the switch, 35 percent of those taking nelfinavir were candidates for antilipid therapy according to the National Cholesterol Education Program (NCEP);19 12 weeks after swapping nelfinavir for atazanavir, only 10 percent still ranked as candidates for lipid lowerers.
Despite the good lipid numbers with atazanavir, clinical signs of lipodystrophy do emerge in people taking this PI with d4T and 3TC, as they were in this trial. Clinicians noted 14 cases among people originally randomized to atazanavir and six among those who traded nelfinavir for atazanavir among the 346 people still in follow-up. Those low numbers may represent underreporting because the study protocol did not ask clinicians to track body shape changes. Murphy did not distinguish between peripheral atrophy and fat accumulation in this summary. Among people randomized to 400 or 600 mg of atazanavir, 23 percent and 35 percent respectively had grade 3 or 4 bilirubin elevations. In the nelfinavir-atazanavir switch group, 10 percent had high bilirubins. Murphy said hyperbilirubinemia rarely caused people to drop out of the study.
Analysis of 24-week trends in a placebo-controlled comparison of atazanavir and efavirenz plus ZDV and 3TC confirmed good lipid trends with both regimens [abstract 36, L26]. Michael Sension (North Broward Hospital District, Fort Lauderdale) reported modest gains in total cholesterol with atazanavir (1 percent) and efavirenz (20 percent), with means for both groups remaining below 200 mg/dL. Mean fasting LDL-C also stayed below the NCEP danger threshold of 130 mg/dL in both treatment arms.
Six-month data from this 810-person international trial showed that neither regimen induced insulin resistance as measured by fasting insulin, C peptide, and glucose concentrations. These markers hardly changed over 24 weeks with either treatment. Unlike some other protease inhibitors, atazanavir does not stymie insulin-mediated glucose traffic by inhibiting a transporter labeled GLUT-4. That may explain the lack of insulin resistance with atazanavir in this trial, Sension proposed.
Lipid subclass quiz (correct answer: efavirenz)
Clinicians who have done their lipid homework know that LDL-C, very lowdensity lipoprotein cholesterol (VLDL-C), and high-density lipoprotein cholesterol (HDL-C) particles come in large, small, or intermediate sizes. And which size predominates determines whether one's risk of heart disease is low, intermediate, or high. Gary Simon and colleagues (George Washington University, Washington, DC) have been sizing up lipid particles in people taking efavirenz and other antiretrovirals. The good news for efavirenz partisans is that particle size and numbers look good in people taking this nonnucleoside. The good news for people with HIV and their clinicians is that Simon cooked up a nice particle size cheat sheet (Table 3).
Simon studied 22 treatment-naïve people, 15 starting nelfinavir and 17 starting efavirenz, in two different trials [abstract 14, L9]. So this is not a randomized comparison, and the number of study participants is small. Measuring lipid subclasses before treatment and periodically for the first 16 weeks, he determined that HIV infection itself conferred a low lipid-related risk in these people. Overall, LDL-C rose 45 percent in the nelfinavir group and 17 percent in the efavirenz group, while HDL-C rose 15.5 percent with nelfinavir and 35.4 percent with efavirenz (P < 0.05 for both comparisons). Looking at subclasses, Simon found that ominous small particle HDL-C rose 44.8 percent with nelfinavir and 25.2 percent with efavirenz, while salutary large-particle HDL-C rose 8.2 percent with nelfinavir versus 40.9 percent with efavirenz (P < 0.05). The number of LDL-C particles rose 45 percent with nelfinavir and 11.7 percent with efavirenz (P < 0.05), again denoting a higher risk with the PI. Most people in both treatment groups were taking ZDV/3TC, although a few in the nelfinavir group were taking ddI/d4T.
Simon's colleague Angelike Liappis confirmed these efavirenz-induced subclass shifts in a 16-week study of 17 treatment-naïve people starting the NNRTI with ZDV/3TC, or ZDV, 3TC, and abacavir [abstract H-1917]. LDL-C rose modestly though significantly from 98 to 114 mg/dL (P = 0.0014), but most of the gain came in the lower-risk large LDL-C particles (38 to 60 mg/dL, P = 0.0234). The total number of LDL-C particles barely budged. Meanwhile, the more dangerous small HDL-C particles multiplied by only 26 percent, compared with a 40 percent jump in risk-lowering large HDL-C particles.
A 48-week open-label study of 10 girls and seven boys who traded a PI for efavirenz showed a sustained virologic response and healthier lipid profiles after the switch [abstract H-1081]. Sixteen of the 17 maintained a viral load below 50 copies/mL after taking efavirenz for 48 weeks, and the viral load stood at 62 copies/mL in the remaining child. Before the switch 12 children (71 percent) had triglyceride tallies above the 95th percentile for their age group, and five (29 percent) had cholesterol levels above the 95th percentile. After 48 weeks of efavirenz, average cholesterol fell from 203 to 175 mg/dL and only one child had a level above 200 mg/dL. LDL-C and triglycerides fell significantly (P < 0.05), and the group's cholesterol-to-HDL-C ratio fell 21 percent. Grace McComsey reported that no one stopped efavirenz because of side effects, though one child with a family history of epilepsy had a seizure.
PI therapy affects not only lipids in children, but also, as in adults, insulin sensitivity. Comparing 33 PI-treated children with 15 PI-naïve children, Ari Bitnun (Hospital for Sick Children, Toronto) linked PIs with decreased insulin sensitivity among pubertal children (P = 0.0163) but not prepubertal children [abstract 8, L5]. In the whole group, three variables independently predicted lowered insulin sensitivity: older age (P = 0.0002), worse HIV immunologic category (P = 0.0186), and PI therapy (P = 0.0333).
Efavirenz came home with a decent metabolic report card in a long, singlecenter Spanish study. These 84 people beginning antiretroviral therapy with efavirenz had a modest but significant rise in total cholesterol (186 to 205 mg/dL, P = 0.011), a drop in uric acid (5.4 to 4.9 mg/dL, P = 0.014), and no substantial change in triglycerides or glucose after two years [abstract 46, L32]. Bernardino Roca (General Hospital of Castellon, Spain) did not report changes in LDL-C, HDL-C, or subclasses thereof.
Two studies showed that d4T drives triglycerides and cholesterol in the wrong direction in treatment-naïve people starting a regimen including this nucleoside. Sharon Walmsley and colleagues randomized untreated people to start efavirenz or saquinavir/ritonavir (1600/100 mg once daily) [abstract 52, L35]. Fifty of them took d4T/3TC and the rest took a non-d4T NRTI combo, usually ZDV/3TC. Forty eight weeks later, both triglycerides (P = 0.014) and total cholesterol (P = 0.021) were significantly higher in the d4T group than in the non-d4T group. Lipid differences between people taking efavirenz and saquinavir/ritonavir did not reach statistical significance.
Outlining interim 48-week results of a placebo-controlled trial comparing tenofovir with d4T (plus 3TC and efavirenz), Joel Gallant (Johns Hopkins University, Baltimore) reported good immunologic and virologic results in both groups [abstract LB-2, ICAAC]. But triglycerides climbed an average 74 mg/dL in the d4T arm while staying flat in the tenofovir group (P < 0.001). Total cholesterol rose 53 mg/dL with d4T and 25 mg/dL with tenofovir (P < 0.001). The incidence of peripheral neuropathy also proved signifi- cantly higher in the d4T arm (7 versus 2 percent, P < 0.001).
HAART, the heart, and other muscles
Other Workshop studies looking at how HIV and its treatments affect the heart and other muscles offered greater challenges to clinicians a few years removed from physiology lectures. One study found a (possible) deficit with treatment, one a (possible) benefit, and one (possibly) neither.
Let's start with arteriovenous oxygen difference (a-vO2) which, all will recall, measures the gap between how much oxygen the heart pumps out in refreshed blood and how much oxygen downstream muscles can suck up. To measure a-vO2 rates and cardiovascular function, Todd Cade (University of Maryland, Baltimore) recruited 45 people, yoked them to the requisite sensors, and worked them on treadmills "to exhaustion" [abstract 16, L10]. Then he compared results in these 15 HAART-treated people, 15 untreated HIV-infected people, and 15 healthy controls matched for age, gender, and activity level. Although cardiac output and stroke rate proved equivalent in the three groups, the HAART group scored worst in a-vO2 and peak heart
rate (Table 4).
Among people on HAART, three more were taking an NNRTI than a PI, but that did not affect results in this small sample. Neither did race. Cade believes the results implicate HAART rather than HIV itself as the prime culprit in oxygen extraction by peripheral muscles, a deficit that may be mediated by NRTI-induced failings in mitochondrial function. He noted, though, that he did not match the groups for lean mass, which can affect oxygen extraction. But no one in the study exercised enough to avoid the "sedentary lifestyle" label. Michael Dubé suggested that Cade's findings may reflect the relative health of the three groups rather than antiretroviral therapy per se. The HAART group, he conjectured, may have been sicker than the untreated and healthy controls, and antiretrovirals may have improved their oxygen-extracting prowess. Although Cade could not rule out that possibility, he noted that no one in the study had a CD4+ nadir below 200 cells/mm3.
A noncomparative study by Kevin Yarasheski (Washington University, St. Louis) found that treatment-induced drops in viral load do help muscles in one way—by improving their ability to metabolize amino acids [abstract 18, L11]. His analysis of 10 men and one woman starting antiretroviral therapy suggested that every 10,000-copy drop in viral load allowed muscles to synthesize an extra 3 g of protein daily. The group's baseline CD4+ count of 94 cells/mm3 climbed to 137 cells/mm3 with treatment (P = 0.021), while the median viral load nosedived from 171,758 copies/mL to 87 copies/mL (P = 0.002). During an average 4.3 months of treatment, mixed muscle protein synthesis rose from 16 to 20 percent daily (P = 0.012), still below the healthy norm of 25 percent. Plasma glutamine efflux, another measure of muscle moxie, improved significantly with treatment. Nine of the study participants were starting a salvage regimen, and two were starting their first antiretrovirals. Yarasheski recorded modest, nonsignificant gains in body muscle mass and thigh strength.
Even in people with marked peripheral atrophy, Giorgos Sakkas (University of California, San Francisco) found, muscle strength and intramuscular energy metabolism were not compromised [abstract 19, L12]. This small study compared six men with clinical and MRI evidence of severe atrophy and six HIV-infected men without atrophy. The atrophy group was signifi- cantly older (48 versus 37 years, P = 0.02), but the groups didn't differ in body mass index, CD4+ count, duration of HIV infection, or current PI or NNRTI use. Nor did they differ in strength (maximum voluntary contraction) of the tibialis anterior muscle (which moves the foot up and down). Several measures of mitochondrial function did not differ between groups.
The results suggest that mitochondrial depletion or damage recorded in people with lipoatrophy doesn't hurt muscles. But the study has limits, Sakkas observed. First, the sample is small. Statisticians told Sakkas he would need 120 volunteers to show statistically significant differences in a study like this. Second, measuring muscle function at rest, as he did, doesn't exclude the possibility that muscles of people with lipoatrophy come up short during exercise. Finally, the study did not compare the atrophic men with healthy volunteers.
Management of antiretroviral-induced toxicities looks less bleak than it did only a few years ago. But even with more options on the table, the choices are still hard.
Swapping PIs for a nonnucleoside or abacavir can work as a fix for high lipids or insulin resistance, and more evidence shows that retiring d4T improves lactates and even fat atrophy. Yet switching is hardly risk free, as Graeme Moyle (Chelsea and Westminster Hospital, London) observes in a recent review.20 The risk of losing viral control may not be great in people responding to their first PI regimen—a patient species on the wane. Risks are higher, maybe too high, in people with single- or dual-nucleoside experience before their PI. And the drugs replacing a PI (or d4T) have their own toxicities, which may be worse, in the short or long term, than the toxicity one hopes to reverse by switching. Finally, Moyle notes, open-label switch studies till a rich breeding ground for bias. Consent forms of randomized open-label trials pique concern about the toxicity of interest, and that may lead to faster dropouts or prompter reporting of side effects in nonswitch arms.
Diet and exercise almost invariably yield morphologic and metabolic dividends, and two Workshop studies (reviewed below) show that both can be practical and cost effective. But they take guidance, time, and dedication, three items often in short supply among anxious people with an incurable disease.
Antilipid agents may ease antiretroviral-induced hyperlipidemia. But results of early statin studies have been inconsistent, and the statin that interacts least with antiretrovirals may be the weakest drug in the class.
Ironically, the side effect that seemed least treatable when the Lipodystrophy Workshop started four years ago, facial atrophy, has emerged as the one with potentially the most effective, least complicated, and most tolerable remedy. At least three surgical procedures are reversing facial atrophy, though they remain largely unavailable in some places, like the United States.
Switching from d4T for atrophy and high lactates
Before the Lipodystrophy Workshop and ICAAC, five studies—two of them randomized —showed that replacing d4T with ZDV, ZDV/3TC, or abacavir slowly but significantly reversed fat atrophy.21-25 One of those studies updated 24-week data to 48 weeks at the two meetings, and a new case-control study logged increased fat trends after switching from d4T.
The noncomparative TARHEEL trial assigned 118 people with physically identified lipoatrophy, symptomatic lactates of 2.2 mmol/L or more, or lactates above 3.2 mmol/L to trade d4T for ZDV/3TC (as Combivir) or abacavir (if they had ZDV experience) [abstracts 21, L13; H-1929]. Grace McComsey reported that everyone had taken d4T for at least six months, and all had a viral load below 400 copies/mL on consecutive readings. Only 16 people in the study had high lactates, and they were more likely to be women (odds ratio 2.64) and African-American (odds ratio 4.50).
Eighty-six people switched to abacavir and 32 to Combivir. The tradeoff came with some risk. Five people had to stop because of hypersensitivity to abacavir, five stopped because of other side effects or clinical setbacks (pancreatic cancer, B-cell lymphoma, and bacteremia), and 14 others stopped returning for visits or left for "other" reasons. Only one person had a virologic breakthrough.
DEXA-measured subcutaneous fat in the arms, legs, and trunk grew in both switch groups, but more with abacavir (Table 5). Total CT-measured subcutaneous fat for both groups jumped by 23.5 cm2 (P < 0.01), while visceral adipose tissue dipped by 2.8 cm2. Most study participants (87 percent) gained subcutaneous fat by week 48, although about 60 percent claimed they saw no improvements in face, legs, arms, or buttocks on a body image questionnaire. Another 20 percent believed they gained "some" or "a lot" of fat, and the rest thought they lost fat. The liver enzyme aspartate
aminotransferase (AST) fell from a median 40 U/L at baseline to 35 U/L at week 48, while alanine aminotransferase (ALT) dropped from 51 to 46 U/L. But not all lab values improved. Triglycerides rose from a median 244.5 mg/dL to 268.5 mg/dL at week 48.
A 12-month case-control study presented by Teresa García-Benayas (Carlos III Hospital, Madrid) compared 49 people who traded d4T for abacavir with 63 who continued d4T [abstract 57, L39]. Everyone had clinician-confirmed lipoatrophy, and no one changed their other antiretrovirals during follow-up. Skinfold measures showed a significant drop in triceps fat among the nonswitching controls, and nonsignificant drops in biceps, thigh, and calf fat. The people who switched to abacavir had nonsignificant fat gains in these four areas. Total body fat measured by bioelectrical impedance fell significantly over 12 months in the nonswitchers and rose slightly in the switchers.
Tracking lactate changes in the TARHEEL cohort, Tyler Lonergan (University of California, San Diego) found that levels stayed flat through 48 weeks in the 102 people who started the study with normal lactates and switched to abacavir or Combivir [abstract 21, L13]. The median lactate measured 2.9 mmol/L in the 16 people who started with high lactates. Ten of them stopped treatment until their lactates fell, and the median at the switch to abacavir or Combivir measured 2.1 mmol/L. After 48 months of abacavir or Combivir, the group's median lactate level stood at 1.3 mmol/L.
In another study Lonergan showed that replacing withdrawn NRTIs in people with symptomatic high lactates usually does not kindle recurrent symptoms [abstract H-1080]. From 1998 through 2002, he tracked 12 people with one or more symptoms of hyperlactatemia (nausea/ vomiting, abdominal pain, anorexia, weight loss, fatigue) plus at least two consecutive venous lactates twice the upper limit of normal (4.2 mmol/L in his lab). All 12 were taking d4T, six with 3TC, four with ddI, and two with abacavir. People stopped antiretrovirals until the lactates subsided, then five restarted regimens containing 3TC/abacavir, five ZDV/3TC/abacavir, and two ZDV/3TC. Two people stopped the rechallenge regimen for reasons unrelated to high lactates. One stopped because of recurrent symptoms. The other nine continued their new regimens for 127 to 1,161 days without recurrent symptoms. One of these nine had lactic acidosis before the NRTI switch.
Diet and exercise: those who can, win
One study explored diet and another exercise for lipid and weight abnormalities in people taking antiretrovirals. The diet study [abstract H-1929], involving 230 people with antiretroviral-induced dyslipidemia, appeared in print shortly after ICAAC.26 Ana Barrios and colleagues (Carlos III Hospital, Madrid) made two important findings:
The Carlos III team enrolled people with cholesterol levels at or above 200 mg/dL (mean 257 mg/dL) or triglycerides at or above 200 mg/dL (mean 320 mg/dL). Two thirds had body fat abnormalities, and two thirds had a viral load below 50 copies/mL. About half were taking a PI regimen and half an NNRTI. No one switched antiretrovirals or started antilipid drugs during the study.
Only 161 people made it to the threemonth mark, and only 36 percent of those reported good adherence with the low-fat diet. By month six only 70 people remained in the study, and 45 percent of them reported good adherence. Among people who stayed in the study and stuck with the diet, total cholesterol, triglycerides, and body mass index fell significantly (Table 6). People taking a PI started the study with much higher lipid levels, and enjoyed greater lipid improvements, than people taking an NNRTI. (The dietary delinquency in this study would not surprise Graeme Moyle, who noted during the Workshop that people he randomized to add pravastatin to a prescribed diet27 stopped the diet when they
started the drug.)
An exercise program thoughtfully planned by clinicians at New York's Harlem Hospital Center met with the same success as the Carlos III diet: Most people could not keep to the program, but those who did reaped rewards. Led by Sai Subhasree Raghavan, the Harlem group got money from the city department of health for an exercise program that includes free membership at a fitness center one block from the hospital, exercise and nutrition evaluation and counseling, supervised twice-weekly training by a professional familiar with the local HIV community, and followup phone calls to encourage adherence [abstract 61, L41]. Despite these enticements, only 20 of 45 people with self-reported fat changes or weight gain enrolled in the program when invited, and only 10 exercised at least once weekly for eight weeks. Raghavan figured the cost per client at a thrifty $33 per encounter hour.
The 16 men and four women who enrolled averaged 1.4 visits per week [abstract 60, L40]. They started with an average body mass index of 26.4 kg/m2, 32.8 percent body fat, and a waist-to-hip ratio of 96. HDL-C, LDL-C, triglycerides, and glucose did not improve significantly over eight to 12 weeks, possibly because of the short duration. But study participants shed weight, mostly in subcutaneous fat. Skinfold measures showed significant improvements in triceps, biceps, thigh, abdomen, and subscapular and suprailiac areas. Hip circumference waned by 0.69 cm (P = 0.004), waist size by 1 cm (P = 0.04), and weight by three pounds (P = 0.039).
French researchers have chronicled successful treatment of facial atrophy with polylactic acid (PLA, New-Fill) since the second Lipodystrophy Workshop. Now the first UK study—and the first randomized study anywhere—confirms these good results. Graeme Moyle randomized 30 people with facial atrophy to get PLA injections in the cheek and nasolabial regions immediately (at day one, week two, and week four) or later (at weeks 12, 14, and 16).
At week 12, ultrasound scans gauged a 4- to 5-mm increase in dermal thickness near the injection sites in the immediatetreatment group, but not in the delayed group. At week 24 the two groups had equivalent gains in dermal thickness. Before-and-after photographs evaluated by three clinicians not involved in the study confirmed these trends, with the best improvement seen in the lower buccal fat pad area. Moyle called the photo evaluation a "semiobjective" measure, and with good reason. The readers also noticed improvement in the temples, even though that area had not been treated. Moyle speculated that the improved cheek appearance "gives the assessor the impression of a generally ‘fuller' face."
The clinician panel did not rate the improved appearance as normal. Earlier studies suggest the need for four or five PLA treatments to remedy the severe atrophy seen in the 28 men and two women who had three treatments so far in Moyle's study. Only two people had treatmentrelated problems—a case of transient cellulitis that did not require antibiotics and delayed the next injection by one week, and superficial bruising that did not delay the next injection.
Although PLA injections remain the best-studied remedy for facial atrophy, not everyone in the field believes it's the best. Derek Jones, a Los Angeles dermatologist who treats people with lipoatrophy, told the Workshop audience he has seen reversal of PLA's effects within one year among people treated elsewhere. Parisian surgeon Patrick Amard, on the other hand, reports that he sees sustained improvement well beyond a year.28 Jones favors a silicone microdroplet product called Silskin, which, unlike PLA, is available in the US. He claimed that migration of injected silicone has not been a problem with experienced practitioners. Migration, he said, results from injecting too much silicone in one session or using adulterated silicone. Jones and others plan a US trial of Silskin for people with HIV lipoatrophy.
A third technique involves collecting subcutaneous fat from a person's abdomen, separating it from blood, and injecting it into the face. Drawbacks of this procedure compared with PLA injections include the need for hospitalization and a sufficient supply of abdominal fat. Parisian clinicians who tried fat implants in 12 men and three women excluded anyone with wasting or a CD4+ count below 100 cells/mm3.29
An independent five-person panel, who reviewed before-and-after photos, and study participants themselves generally agreed on results six months after surgery. Thirteen of 15 patients felt they had good (9) or very good (4) atrophy correction, while the panel gave good scores to 11 and a very good score to one. Eleven study participants thought they had good (7) or very good (4) global results, while panelists judged global results good in 11 and very good in two. MRI scans recorded an average six-month fat gain of 10.5 mm on the left side of the face and 10 mm on the right.
A small randomized trial sized up the bisphosphonate alendronate in 22 men with osteoporosis or severe osteopenia [abstract H-1936]. Eugenia Negredo (Germans Trias i Pujol University Hospital, Barcelona) excluded people who had an antiretroviral treatment break within the preceding four weeks, a record of poor adherence, or secondary causes of osteopenia. She randomized 12 people to 10 mg of alendronate daily plus dietary counseling and 10 to counseling alone. The treatment group was older than the control group (46 versus 38 years), had a higher calcium intake (1,262 versus 1,040 mg/day), and smoked less (35 percent versus 55 percent).
After 24 weeks neither group had a significant change in bone mineral density. Three of 12 people taking alendronate improved from an osteoporosis ranking to osteopenia, while no one in the control group improved. At best the study suggests that, in a group like this, longer follow- up will be needed to see significant improvement in bone mineral density with alendronate.
Statins popular, but how potent?
HIV clinicians wondering how many of their colleagues use statins can stop wondering. Bristol-Myers Squibb did a timetrend analysis to find out, and the answer is "more and more." Uchenna Iloeje and colleagues used electronic pharmacy, physician, and hospital claims from a US managed care organization to track antiretroviral and antilipid trends from 1998 through 2001 [abstract 42, L29]. The number of people with HIV in the organization remained fairly constant—between 3,000 and 4,000—over those years, but antiretroviral prescriptions dropped while antilipid prescriptions climbed (Table 7). Part of the jump in lipid-lowering therapy may reflect the aging of the cohort, because
antilipid use also grew among HAART-naïve people and among those taking non- PI regimens. In an analysis adjusted for year, people taking a PI were 2.36 times more likely to be on lipid therapy than those not taking a PI.
About 90 percent of the people in this survey were men. In another study of 24,862 men and 8,735 women with HIV enrolled in California's Medi-Cal program between January 1996 and June 2000, Iloeje found that men taking PIs were three times more likely to get an antilipid prescription than women taking PIs [abstract 43, L30].
Are the lipid drugs working? Results so far are mixed. A few years ago a randomized trial by Graeme Moyle found that pravastatin (40 mg daily) plus dietary advice for 24 weeks lowered cholesterol (mostly LDL-C) more than dietary advice alone in 31 men who started with levels above 250 mg/dL.27 But a study that randomized 170 people with high lipids to pravastatin (40 mg daily) or fenofibrate (200 mg daily) found that the drugs helped few people achieve NCEP lipid goals in 12 weeks.30
At the Lipodystrophy Workshop a placebo-controlled trial of pravastatin in HIV-infected people with high lipids found an incomplete response to the drug [abstract 50, L34]. The results led James Sosman and coworkers (University of Wisconsin, Madison) to conclude that "combinations of lipid-lowering therapies may be needed to normalize the lipoprotein abnormalities observed in patients taking PIs."
The study included 15 men and one woman with triglycerides above 150 mg/dL, LDL-C above 130 mg/dL, and HDL-C below 40 mg/dL while taking a suppressive PI regimen for at least three months. After four weeks of good adherence with a low-fat, low-cholesterol diet, they continued the diet and took either pravastatin (40 mg daily) or placebo for eight weeks. Total cholesterol, LDL-C, non-HDL-C, and triglycerides fell significantly with pravastatin but not with placebo. But pravastatin did not improve an atherogenic profile characterized by mildly elevated LDL-C particle concentration, decreased large HDL-C, and increased large VLDLC (see Table 3). Sosman added that the "magnitude of improvement [in the main lipid measures] appears less than in clinical trials of HIV-negative patients."
In a Workshop review lecture, Peter Reiss (Academic Medical Center, Amsterdam) noted the irony that pravastatin interacts with PIs less than other statins, but it may be the weakest drug in the group.
So how about fluvastatin? Again, the drug fell short by some lipid measures in a placebo-controlled crossover trial published just after the Workshop.31 Fourteen men and two women with hyperlipidemia while taking a PI took fluvastatin (40 mg daily) or placebo for four weeks, then crossed over to placebo or fluvastatin after a twoweek washout. Although triglycerides did not drop with treatment, total cholesterol fell from 8.0 mmol/L (309 mg/dL) to 6.6 mmol/L (255 mg/dL) (P < 0.001) and the total cholesterol-to-HDL-C ratio dropped from 7.3 to 6.4 (P < 0.01). But the authors write that "the amplitude of this lipidlowering effect was limited" because eight of the 16 study participants still had cholesterol readings above 7.0 mmol/L (270 mg/dL) and triglycerides above 3.0 mmol/L (266 mg/dL) after finishing their course of fluvastatin, compared with 10 of 16 before treatment.
Other morsels of good news for people suffering PI side effects turned up in the literature around the time of the Lipodystrophy Workshop. Rosiglitazone improved glucose disposal and added subcutaneous fat in eight people with HIV infection and insulin resistance.32 Study participants had a glucose disposal rate of 3.8 mg per kg of lean body mass per minute compared with 11.08 mg in healthy age- and weight-matched controls (P < 0.001). After six to 12 weeks of rosiglitazone at a dose of 8 mg daily, glucose disposal improved by 59 percent to 4.99 mg (P = 0.02). CT-measured subcutaneous adipose tissue rose by 23 percent (P = 0.05), and visceral adipose tissue dropped by 21 percent (P = 0.04). Rosiglitazone did not affect viral load or CD4+ count in study participants.
An earlier placebo-controlled trial of rosiglitazone at the same dose in 30 people with HIV lipodystrophy found improved insulin levels.33 In this study subcutaneous peripheral fat, visceral fat, and waist-to-hip ratio did not improve significantly with rosiglitazone, but the trial was powered to detect only substantial fat changes. A large study of rosiglitazone in people with HIV infection is under way in Australia.
This is not really an appendix; it's really about the appendix. Daniel Klein, the clinician whose digital delvings into the Kaiser Permanente health care system database furthered the understanding of antiretrovirals and heart disease,13 has turned his attention to the intestinal cul-desac that does nothing but cause trouble [abstract H-1154]. He found that it causes trouble more often in men with HIV than in men without HIV.
Comparing 6,436 HIV-infected men with 18,438 uninfected men, Klein calculated a crude appendicitis rate of 3.3 hospital admissions per 1,000 person-years in the HIV group versus 0.86 in the non-HIV group. Adjusted for age, the relative rates measured 2.9 with HIV and 0.92 without HIV (P < 0.001). Because rates in men with HIV did not increase from 1991 to 2001, it appears that the more potent HIV therapies used since 1996 did not worsen the incidence of appendicitis seen with earlier regimens. Klein charted use of one or more PIs in 66 percent of HIV-infected men without appendicitis and in 72 percent of those with appendicitis, a nonsignificant difference. Nor did CD4+ count, duration of HIV infection, or immune reconstitution appear to influence rates of appendicitis. But because a higher proportion of men with than without appendicitis were taking some antiretroviral (81 percent versus 65 percent, P < 0.05), Klein proposed that "the role of [antiretroviral] use in general needs further study."
Since 1996, hospital stays for appendicitis lasted no longer in HIV-infected men than in uninfected men. The HIV group had a higher percentage of perforations than the non-HIV group (27 percent versus 16 percent), but that difference lacked statistical significance. A lower CD4+ count did not predict perforation. In the HAART era men with HIV and appendicitis were much less likely than uninfected men to have a white blood cell (WBC) count above 12.5 x 103/µL upon admission to the hospital (26 percent versus 70 percent, P < 0.001). Klein concluded that "clinicians should be quick to consider appendicitis when presented with right lower quadrant pain in HIV-infected patients but should not rely on elevated WBCs in making a diagnosis."
Daniel Klein's data-packed poster (see Appendix A) reminds us again that sheer clinical number crunching can offer ready insights into clinical practice. But, as he observes, this revealing inquest does not explain why appendicitis appears to favor people with HIV. This need to know the why's is one thing that keeps bench scientists on the payroll in academia and industry. Thanks to some of them, the cellular nuts and bolts of HIV lipodystrophy and other toxicities are finding their way to the right bins in the pathogenic toolbox.
That's why it's important to tough through those slide talks on 3T3-F44A2 cells, as a largely clinical audience did at the 4th Lipodystrophy Workshop. You never know how close that bench will get to your bedside. Take, for example, a study of 3T3-F44A2 cells by Rik Derynck and colleagues in the School of Dentistry at the University of California, San Francisco.34 Derynck found that he could take these undifferentiated fat cells, or pre-adipocytes, juice them with a little retinoic acid and something called bone morphogenetic protein, and radically change their destiny. Instead of becoming fat, the 3T3-F44A2s became bone.
By tinkering with these cells' signaling systems, Derynck said, "one can do things that one has not thought of before. In effect, the life of these fat cells has been changed from fat to bone. Now we can begin to think about therapeutic applications."35 HIV docs can think of one.
*Abstracts from the Lipodystrophy Workshop are published in the 2002 volume of Antiviral Therapy. Citations here give the abstract number and the journal page number, which is preceded by an L. Workshop abstracts are online at http://www.intmedpress.com/journals_contents_avt7(3)cont.cfm#lipo. Abstract numbers beginning with an H are from the 42nd Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC), held September 27-30 in San Diego.
Mark Mascolini writes about HIV infection (mailmark@ptd.net).
1. Bastard JP, Caron M, Vidal H, et al. Association between altered expression of adipogenic factor SREBP1 in lipoatrophic adipose tissue from HIV-1-infected patients and abnormal adipocyte differentiation and insulin resistance. Lancet 2002 Mar 23;359(9311):1026-31.
2. Moore RD, Wong WM, Keruly JC, McArthur JC. Incidence of neuropathy in HIV-infected patients on monotherapy versus those on combination therapy with didanosine, stavudine and hydroxyurea. AIDS 2000 Feb 18;14(3):273-8.
3. Maher B, Alfirevic A, Javier Vilar F, et al. TNF-α promoter region gene polymorphisms in HIV-positive patients with lipodystrophy. AIDS 2002 Oct 18;16(15):2013-8.
4. Farooqi IS, Jebb SA, Langmack G, et al. Effects of recombinant leptin therapy in a child with congenital leptin deficiency. N Engl J Med 1999 Sep 16;341(12):879-84.
5. Oral EA, Simha V, Ruiz E, et al. Leptin-replacement therapy for lipodystrophy. N Engl J Med 2002 Feb 21;346(8):570-8.
6. Savage DB, O'Rahilly S. Leptin: a novel therapeutic role in lipodystrophy. J Clin Invest 2002 May;109(10):1285-6. http://www.jci.org/cgi/content/full/ 109/10/1285.
7. Shimomura I, Hammer RE, Ikemoto S, et al. Leptin reverses insulin resistance and diabetes mellitus in mice with congenital lipodystrophy. Nature 1999 Sep 2;401(6748):73-6.
8. Serghides L, Nathoo S, Walmsley S, Kain KC. CD36 deficiency induced by antiretroviral therapy. AIDS 2002 Feb 15;16(3):353-8.
9. Carr A. An objective case definition of HIV lipodystrophy. 9th Conference on Retroviruses and Opportunistic Infections. February 24-28, 2002. Seattle. Abstract 31.
10. Grunfeld C. Basic science and metabolic disturbances. XIV International AIDS Conference. July 7-12, 2002. Barcelona. Abstract TuOr158 - This was a non-abstract driven session by an invited speaker; no abstract was submitted.
11. Mary-Krause M, Cotte L, Partisani M, et al. Impact of treatment with protease inhibitor (PI) on myocardial infarction (MI) occurrence in HIV-infected men. 8th Conference on Retroviruses and Opportunistic Infections. February 4-8, 2001. Chicago. Abstract 657.
12. Holmberg SD, Moorman AC, Tong TC, et al. Protease inhibitor use and adverse cardiovascular outcomes in ambulatory HIV patients. 9th Conference on Retroviruses and Opportunistic Infections. February 24-28, 2002. Seattle. Abstract 698.
13. Klein D, Hurley LB, Quesenberry CP Jr, Sidney S. Do protease inhibitors increase the risk for coronary heart disease in patients with HIV-1 infection? J Acquir Immune Defic Syndr 2002 Aug 15;30(5):471-7.
14. Woolley IJ, Johnsen SP, Sorensen HT, et al. Cardiovascular events in the MACS cohort and prior prophylaxis with macrolides for Mycobacterium avium complex. XIV International AIDS Conference. July 7-12, 2002. Barcelona. Abstract ThPeB7282.
15. Bozzette SA, Ake C, Carpenter A, et al. Cardio- and cerebrovascular outcomes with changing process of anti-HIV therapy in 36,766 US veterans. 9th Conference on Retroviruses and Opportunistic Infections. February 24-28, 2002. Seattle. Abstract LB9.
16. Grunfeld C, Pang M, Doerrler W, et al. Lipids, lipoproteins, triglyceride clearance, and cytokines in human immunodeficiency virus infection and the acquired immunodeficiency syndrome. J Clin Endocrinol Metab 1992 May;74(5):1045-52.
17. In a study of 642 HAART-treated people enrolled in five US trials and tracked for 30 months, cumulative rates of grade IV "cardiac-related events" and "cardiovascular events" both measured 1.6 percent, trailing grade IV liver-related toxicity (6.1 percent), neutropenia (3.9 percent), pancreatitis (2.2 percent), anemia (2.1 percent), and psychiatric problems (2.1 percent). Reisler R, Han C, Burman W, et al. Incidence of grade IV events, AIDS, and mortality in a large multicenter cohort receiving HAART. 9th Conference on Retroviruses and Opportunistic Infections. February 24-28, 2002. Seattle. Abstract 36.
18. Piliero PJ, Cahn P, Pantaleo G, et al. Atazanavir: a once-daily protease inhibitor with a superior lipid profile—results of clinical trials beyond week 48. 9th Conference on Retroviruses and Opportunistic Infections. February 24-28, 2002. Seattle, 2002. Abstract 706.
19. National Cholesterol Education Program. Executive summary of the third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). JAMA 2001 May 16;285(19):2486-97. http://jama.ama-assn.org/issues/v285n19/ffull/jsc10094.html.
20. Moyle G. The switching spiral: a triumph of hope over benefit? AIDS Read 2002 Apr;12(4):147-50.
21. Saint-Marc T, Touraine JL. The effects of discontinuing stavudine and the development of lipodystrophy. AIDS 1999 Oct 22;13(15):2188-9.
22. Carr A, Workman C, Smith DE, et al. Abacavir substitution for nucleoside analogs in patients with HIV lipodystrophy: a randomized trial. JAMA 2002 Jul 10;288(2):207-15.
23. McComsey G, Lonergan T, Fisher R, et al. Improvements in lipoatrophy are observed after 24 weeks when stavudine (d4T) is replaced by either abacavir or zidovudine. 9th Conference on Retroviruses and Opportunistic Infections. February 24-28, 2002. Seattle. Abstract 701.
24. Moyle GJ, Baldwin C, Gazzard BG, et al. A randomized open label study of 3 substitution strategies in hypercholesterolemic persons virologically controlled on first line antiretroviral therapy. XIV International AIDS Conference. July 7-12, 2002. Barcelona. Abstract ThPeB7322.
25. John M, James I, McKinnon E, et al. A randomized, controlled, openlabel study of revision of antiretroviral regimens containing stavudine (d4T) and/or a protease inhibitor to zidovudine/lamivudine/abacavir to prevent or reverse lipoatrophy: 48-week data. 9th Conference on Retroviruses and Opportunistic Infections. February 24-28, 2002. Seattle. Abstract 700.
26. Barrios A, Blanco F, García-Benayas T, et al. Effect of dietary intervention on highly active antiretroviral therapy-related dyslipidemia. AIDS 2002 Oct 18;16(15):2079-81.
27. Moyle GJ, Lloyd M, Reynolds B, et al. Dietary advice with or without pravastatin for the management of hypercholesterolaemia associated with protease inhibitor therapy. AIDS 2001 Aug 17;15(12):1503-8.
28. Amard P. Role of plastic surgery in treatment of fat wasting and fat accumulation. 2nd European Workshop on Lipodystrophy. April 19-21, 2001. Cannes. Abstract O-24.
29. Levan P, Nguyen TH, Lallemand F, et al. Corrections of facial lipoatrophy in HIV-infected patients on highly active antiretroviral therapy by injection of autologous fatty tissue. AIDS 2002 Sep 27;16(14):1985-7.
30. Aberg J, Zackin R, Evans S, et al. A prospective, multicenter, randomized trial comparing the efficacy and safety of fenofibrate versus pravastatin in HIV-infected subjects with lipid abnormalities: ACTG 5087. XIV International AIDS Conference. July 7-12, 2002. Barcelona. Abstract LbPeB9018.
31. Doser N, Küble S, Telenti A, et al. Efficacy and safety of fluvastatin in hyperlipidemic protease inhibitor-treated HIV-infected patients. AIDS 2002 Sep 27;16(14):1982-3.
32. Gelato MC, Mynarcik DC, Quick JL, et al. Improved insulin sensitivity and body fat distribution in HIV-infected patients treated with rosiglitazone: a pilot study. J Acquir Immune Defic Syndr 2002 Oct 1;31(2):163-70.
33. Sutinen J, Hakkinen AM, Westerbacka J, et al. Rosiglitazone in the treatment of HAART-associated lipodystrophy: a randomized, doubleblind, placebo-controlled study. 9th Conference on Retroviruses and Opportunistic Infections. February 24-28, 2002. Seattle. Abstract LB13.
34. Skillington J, Choy L, Derynk R. Bone morphogenetic protein and retinoic acid signaling cooperate to induce osteoblast differentiation of preadipocytes. J Cell Biol 2002 Oct 14;159(1):135-46.
35. Stern T. UCSF study converts fat cells to bone cells, could lead to help in bone repair. UCSF News Service. October 14, 2002.
021110
IA021102
Copyright © 2002 - International Association of Physician's in AIDS Care (IAPAC). Reproduction of this article (other than one copy for personal reference) must be cleared through the IAPAC.
AEGiS is made possible through unrestricted grants from Boehringer Ingelheim, iMetrikus, Inc., John M. Lloyd Foundation, the National Library of Medicine, and donations from users like you. Always watch for outdated information. This article first appeared in 2002. This material is designed to support, not replace, the relationship that exists between you and your doctor.
AEGiS presents published material, reprinted with permission and neither endorses nor opposes any material. All information contained on this website, including information relating to health conditions, products, and treatments, is for informational purposes only. It is often presented in summary or aggregate form. It is not meant to be a substitute for the advice provided by your own physician or other medical professionals. Always discuss treatment options with a doctor who specializes in treating HIV.
Copyright ©1980, 2002. AEGiS. All materials appearing on AEGiS are protected by copyright as a collective work or compilation under U.S. copyright and other laws and are the property of AEGiS, or the party credited as the provider of the content. .