The Body: The Complete HIV/AIDS Resource
Follow Us Follow Us on Facebook Follow Us on Twitter Download Our App 
Professionals >> Visit The Body PROThe Body en Espanol
  • Email Email
  • Printable Single-Page Print-Friendly
  • Glossary Glossary

Debate Widens Over
Protease Inhibitor Side Effects

July/August 1998

A note from The field of medicine is constantly evolving. As a result, parts of this article may be outdated. Please keep this in mind, and be sure to visit other parts of our site for more recent information!

Reports of multiple metabolic abnormalities in HIV-positive individuals on highly active antiretroviral therapy (HAART) continue to mount. At the 12th World AIDS Conference in Geneva, presentations from all parts of the world depicted a group of unusual and distressing symptoms. The syndrome of abnormal redistribution of body fat, termed lipodystrophy, is one of the most dramatic occurrences. Other metabolic abnormalities also present in some HAART users may or may not be associated with lipodystrophy. No clear-cut cause has been established, and there is still no standardized and widely accepted case definition or diagnostic criteria, which may explain why the reported prevalence ranges from below 5% to over 60%.

Lipodystrophy manifests clinically as subcutaneous fat depletion in the face, limbs and upper trunk (peripheral lipodystrophy), prominent subcutaneous veins in the legs (normal veins that protrude due to the fat loss), increased abdominal girth, enlarged dorsocervical fat pad ("buffalo hump" between the shoulders), breast enlargement in women and fat accumulations in the lining of the abdominal cavity and surrounding the organs (visceral fat). The changes in body shape are generally observed several months after initiation of HAART and can be quite disfiguring.

Metabolic abnormalities have also been widely reported and include hyperlipidemia (an excess of fats in the blood, specifically cholesterol and triglycerides); hyperglycemia (elevated blood glucose levels); diabetes; and hyperuricemia (increased levels of uric acid, which can lead to gout). The elevated lipid levels may contribute to premature cardiovascular disease and heart attacks (see Treatment Issues, May 1998).

The Australian Hypothesis

Two theories have been posited as possible causes of the unusual symptoms. David Cooper, M.D., and colleagues from St. Vincent's Hospital, Sydney, Australia were among the first to report cases and to use the term peripheral lipodystrophy. His group has consistently found very high prevalence rates among its cohort of protease inhibitor-treated patients. At the 5th Retrovirus Conference last winter, Dr. Cooper reported that 64% of his patients on protease inhibitors showed signs of lipodystrophy (abstract 410). At a late breaker in Geneva, his colleague Andrew Carr, M.D., reported over 80% prevalence by patient self-report (LB 8/abstract 12462). Dr. Cooper has stated that he expects nearly all patients on protease inhibitors will eventually experience these types of metabolic abnormalities.

Dr. Cooper, Dr. Carr and colleagues published their hypothesis this summer (A. Carr et al. The Lancet. June 20, 1998; 351(9119):1881-3) and Dr. Cooper presented the theory at an oral session in Geneva (abstract 176). According to their research, the region of the HIV protease enzyme to which protease inhibitors bind is structurally similar to regions on two human proteins that regulate the breakdown of lipids: low density lipoprotein-receptor-related protein (LPR) and cytoplasmic retinoic-acid binding protein type 1 (CRABP-1). They hypothesize that protease inhibitors inhibit LPR and CRABP-1 as well as the HIV protease enzyme.

CRABP-1 presents retinoic acid to cytochrome P450 3A (the P450 enzymes in the liver metabolize drugs and other fat-soluble substances). The retinoic acid is then synthesized into cis-9-retinoic acid, which helps regulate cell differentiation and programmed cell death in peripheral fat cells. Inhibition of retinoic acid binding to CRABP-1 could lead to decreased cis-9-retinoic production and to reduced differentiation and increased death of peripheral fat cells. The authors speculate that the subsequent altered fat metabolism would be proportional to the degree of cytochrome P450 3A (CYP3A) inhibition of each protease inhibitor (as the protease inhibitors themselves also inhibit CYP3A). And, in fact, hyperlipidemia appears be worse in patients receiving the most potent of the CYP3A inhibitors, ritonavir. The theory also proposes that protease inhibitor binding to LPR impairs triglyceride clearance from the blood.

The resulting hyperlipidemia contributes to central fat deposits (and breast enlargement in the presence of estrogen), insulin resistance, and, in susceptible individuals, diabetes. The authors call for validation of their hypothesis and the development of protease inhibitors that do not cause this syndrome.

Another Theory

Donald Kotler, M.D., of St. Luke's-Roosevelt Hospital in New York, points out the major flaw in the Australian hypothesis, namely that some people with this group of symptoms, including one of Dr. Carr's patients, are not taking protease inhibitors. There also are some reports of patients with similar manifestations from the pre-HAART era and from patients with other conditions. Dr. Kotler, who has noted this phenomenon in survivors of such life-threatening illnesses as childhood leukemia and breast cancer, considers it a type of post-traumatic stress syndrome. (He suggests cortisol [a naturally occurring steroid that affects the metabolism of glucose, proteins and fat] may be involved and is looking further into this theory using special lab tests to detect abnormal cortisol levels.)

Dr. Kotler believes it is still unclear if the symptoms are specifically linked with the presence of the protease inhibitors in the body or are the result of effective control of HIV brought about by the use of HAART. While not discounting the Cooper/Carr hypothesis, he suggests that simply concluding that the abnormalities are protease inhibitor-related may be missing something of great importance.

In Geneva, Dr. Kotler presented the results of a retrospective analysis of 96 participants (abstract 32173). The study compared body composition and fat distribution from current patients to records dating back to 1984 of matched non-protease inhibitor-treated HIV-positive patients. There was also a healthy control group. The results demonstrated that there were alterations in fat distribution in people who were treated many years ago similar to those seen in patients today. A multiple regression analysis did not find a significant predictive effect for CD4 count or drug therapy on fat redistribution. Only viral load had a significant independent predictive value. Participants with more abnormal fat redistribution had lower viral loads.

Dr. Kotler theorizes that the association between protease inhibitors and the metabolic abnormalities is indirect, and might be inversely related to HIV replication. Although viral load has been effectively suppressed, the systems of the body may not completely revert to "normal" and metabolism remains altered. He states that these symptoms were not commonly observed in the pre-HAART era because patients usually died quickly once they developed AIDS. The cases that were seen were in long-term survivors.

Visceral Fat Deposits

This spring, Kirk Miller, M.D., and colleagues from the NIH reported on a study of visceral fat accumulation in 30 HIV-positive men (K. Miller et al., The Lancet. March 21, 1998; 351(9106):871-5). Dr. Miller used abdominal computed tomography (CT) scans to measure the ratio of visceral to subcutaneous adipose (fatty) tissue in ten indinavir-treated patients who developed increased abdominal girth and gastrointestinal symptoms three months after starting indinavir in the absence of substantial overall weight gain. As a control, scans were also conducted on ten asymptomatic indinavir-treated patients and ten patients not using indinavir.

Comparison of these scans confirmed an increase of fat within the abdomen of symptomatic indinavir users, an intermediate amount in asymptomatic indinavir users and a normal amount in the non-indinavir group. The higher amount of visceral fat correlated with a longer duration of indinavir use and elevated serum triglyceride levels. An overabundance of visceral fat is associated with metabolic and endocrine abnormalities such as hyperlipidemia and glucose intolerance. Given the fact that the fat is in the lining of the abdomen and surrounding internal organs rather than in the subcutaneous area, liposuction is not a viable solution.


Various lipid abnormalities were reported at the Geneva Conference. Jane Pollner and colleagues of Walter Reed Army Center in Washington conducted a chart review of 340 HIV-positive patients (abstract 12269). Mean increase in cholesterol level was statistically greater in the protease inhibitor users (average increase was 28.9 mg/dl). Sixteen percent of the protease inhibitor group had elevations of greater than 65 mg/dl and 5.7% had increases of greater than 100 mg/dl. The changes were most dramatic for patients on either a ritonavir/saquinavir combination or ritonavir or saquinavir alone, with less pronounced changes for those on nelfinavir and indinavir.

Antonio Carota from Switzerland presented his group's findings on the prevalence of metabolic abnormalities in 131 protease inhibitor users and 25 non-protease inhibitor treated patients (abstract 12375). Among the protease inhibitor users, 40% experienced marked triglyceride elevation and 56% experienced high cholesterol levels. The prevalence of hyperlipidemia was over 80% in the ritonavir or ritonavir/saquinavir subgroups. Diet modification was ineffective in reducing elevated lipid levels.

A report from Keith Henry, M.D., of Regions Hospital, Minnesota, underscored the potential for serious clinical consequences in patients experiencing lipid abnormalities. Dr. Henry presented cases of three men ranging in age from 26 to 40 years old who developed coronary artery disease while on protease inhibitors (abstract 12319). Two of the patients developed elevated triglycerides and cholesterol after starting HAART. Dr. Henry noted that the patients had a history of smoking or a family history of heart disease. Traditional risk factors for heart disease also include high fat diets (necessary for maximum absorption of ritonavir and saquinavir), testosterone use, hypertension and diabetes.

Of 135 patients in the Regions Hospital Clinic on protease inhibitor-containing regimens, 48% had elevated lipid levels (compared to 17% of non-protease inhibitor treated patients). Lipid abnormalities were more frequent and severe in patients on the combination of ritonavir and saquinavir versus single protease inhibitor regimens (nelfinavir and indinavir).

Forty-four patients were enrolled in a lipid management study: half were placed in a diet/exercise group, half received lipid-lowering agents (gemfibrozil and atorvastatin). Dr. Henry stated, "Our experience suggests that some of the standard lipid-lowering agents might be 'safe,' as we haven't seen any negative drug reactions with the protease inhibitors. However, while there is improvement in the lipid profiles, it's not spectacular. The treatment does not seem to be quite as effective as you would see in the general population." (For more on lipid-lowering agents, see Treatment Issues, June 1998.) In the diet/exercise group, even after several months' concerted effort, Dr. Henry noted that, "there was not any effect whatsoever in most patients." Dr. Henry is planning clinical trials later in the year for patients who would like to switch off protease inhibitors to other regimens due to poor lipid profiles.

Finally, four patients in Dr. Henry's clinic developed pancreatitis. Two had a history of alcohol abuse, which can lead to pancreas inflammation. The average triglyceride level was 506 mg/dl. While this is above the normal range, it is not considered dangerously high. Dr. Henry stated he could find no evidence to support the widely held belief that high triglyceride levels cause pancreatitis. All four patients were on aggressive antiretroviral therapy with either a two-protease regimen or a four-drug combination. Dr. Henry revised the treatment regimens by shaving off some of the nucleoside analogs and is watching the patients closely. He recommends that all patients obtain baseline fasting lipid profiles prior to initiating protease inhibitor therapy in order to be able to compare to subsequent lab values.

Is There a Benefit in Switching to Nelfinavir?

Two posters addressing the question of whether switching protease inhibitors could improve the symptoms seemed to indicate some advantage in using nelfinavir. Chris Duncombe and colleagues from Sydney, Australia, conducted an observational study on 21 participants with hyperlipidemia and visual evidence of lipodystrophy to evaluate changes in serum lipids after switching protease inhibitor therapy (abstract 12287). All participants were using indinavir or saquinavir plus ritonavir in addition to two other antiretroviral agents. Participants were switched to nelfinavir and at least one new antiretroviral. The median serum triglyceride level went down three months after the switch, but serum cholesterol levels remained unchanged. After three months, the visual appearance of lipodystrophy was stabilized or unchanged in 57% of participants, partially reversed in 33% of participants and progressive in 10% of participants.

Michael Dube, M.D., of the University of Southern California, Los Angeles, and colleagues reported on eight men without prior histories of glucose intolerance who were treated with nelfinavir after developing new-onset hyperglycemia on indinavir (abstract 32172). Glucose levels greater than 180 mg/dl developed one to seven months after commencing on indinavir. Seven of the eight had experienced potent viral suppression with indinavir (six had viral loads below 200 copies/ml and one had a greater than 2 log decrease).

Two participants stopped indinavir and the hyperglycemia resolved within six weeks and did not recur eight months after starting nelfinavir. Six patients were switched directly from indinavir to nelfinavir without interruption. Three of these had complete resolution of the hyperglycemia within two to three weeks and it has not recurred at six, seven and eleven months. The other three did not have resolution of the hyperglycemia and are being treated with either oral drugs or insulin.

Two persons experienced viral rebound from less then 200 copies/ml after the switch. In both cases only the protease inhibitor was changed and the background nucleoside analogs remained the same. The two participants drove their viral loads back below 200 copies/ml by intensifying their regimens. Dr. Dube recommends monitoring closely for viral rebound and switching to new antiretroviral agents at the time of the protease inhibitor change. He cautions that when drug options are limited, it may be preferable to continue indinavir and treat the diabetes rather than risk viral rebound.

Treatment for Fat Redistribution

Aside from surgery or liposuction, which have numerous drawbacks (including the possibility of recurring fat deposits) and are often impractical, there are very few options for the treatment of fat redistribution associated with HAART. Recombinant human growth hormone (HGH) has been used to treat AIDS-related wasting because it promotes increases in lean, protein-rich tissue while fat is broken down. Importantly, HGH causes the metabolism of visceral as well as subcutaneous fat.

Gabriel Torres, M.D., of the Bentley-Salick Medical Practice in New York, noticed while treating a wasting patient with HGH that there was a reduction in the size of her buffalo hump. Dr. Torres subsequently used HGH at 5-6 mg/day in four other patients with buffalo humps and reported his results in Geneva (abstract 32164). All five patients experienced a reduction in the size of the buffalo hump and a regression of truncal obesity ranging from 25% to 100%. There were some adverse effects of the treatment including carpal tunnel syndrome, hyperglycemia and elevated pancreatic enzymes. No significant changes in cholesterol or triglyceride levels were observed and there were no effects on HIV viral load. It is unclear what the optimal dose or duration of treatment should be for fat redistribution.

If HGH does not address the underlying cause of the problem, it is possible that this will be a short-term and very expensive solution. Serono, the makers of HGH, quoted a price of $18,000 for three months of treatment. The drug has an indication for AIDS-related wasting, but not lipodystrophy. A third-party payer would not be required to cover the cost of HGH treatment, unless there was a concurrent diagnosis of wasting. In addition, side effects of HGH use include hyperglycemia. This could be problematic for patients whose blood glucose levels are already elevated.

Dr. Kotler and the Community Research Initiative on AIDS (CRIA) in New York City have launched an open-label safety and efficacy pilot study of HGH for the treatment of truncal obesity. Twenty-five participants will receive human growth hormone therapy daily (by self-administered injection) for 24 weeks. For more information on this study, known as FAB (Fat Accumulation in the Belly), call 212/924-3934. If HGH seems effective, Serono will design a double-blind placebo-controlled trial and seek FDA approval.

Despite the growing mass of information, it is still unclear if all the symptoms are linked with one particular syndrome or are the result of several distinct syndromes. New and possibly related symptoms continue to emerge as well, such as ingrown toenails, dry skin and cracked lips. The relative health hazard of each of the symptoms is unknown. Patients should be monitored closely and symptoms treated when appropriate. Most physicians do not now advise discontinuing a successful protease inhibitor-containing regimen when metabolic abnormalities emerge.

Back to GMHC Treatment Issues July/August 1998 contents page

A note from The field of medicine is constantly evolving. As a result, parts of this article may be outdated. Please keep this in mind, and be sure to visit other parts of our site for more recent information!

  • Email Email
  • Printable Single-Page Print-Friendly
  • Glossary Glossary

This article was provided by Gay Men's Health Crisis. It is a part of the publication GMHC Treatment Issues. Visit GMHC's website to find out more about their activities, publications and services.
See Also
An HIVer's Guide to Metabolic Complications
More Research on Lipodystrophy and Other Metabolic Complications