What Have We Learned About Lipodystrophy, High Cholesterol, Diabetes, and Long-Term Complications of Protease Inhibitors? Or, What About Protease Paunch?

One of the important issues discussed at the 12th International AIDS conference was long term complications of protease inhibitors. Key questions are: What are the complications? How frequent are they? Are they dangerous or just annoying or disfiguring? Are they really specific to PI's or are they a consequence of any good "HAART"? If they are caused by PI's, are any of the drugs worse or better than the others? Not all of the answers are in, but the picture is a little clearer. One group of complications are changes in body shape, or "lipodystrophy" Several groups described the same constellation of findings: Loss of peripheral fat (fat in the face, legs, buttocks) and a tendency to accumulate central fat, including inside the abdomen ("visceral fat"), on the belly, the breasts, and sometimes the back of the neck (leading to the"buffalo hump appearance)

Poster Session 12462: Prevalence and severity of protease inhibitor (PI) induced lipodystrophy and insulin resistance

In my opinion, one of the best studies of lipodystropy came from the group in Sydney, reported by Andrew Carr. They studied 84 men who had been on PI therapy for a mean of 21 months using a questionnaire, DEXA scanning, and measurements of lipids, blood sugar and insulin. They found subjective changes in body shape (i.e., the patients thought they looked different) in a staggering 76% of patients, beginning on average 13 months into treatment. DEXA scanning found a clear loss in total body fat, but this was a loss in peripheral fat, often associated with an increase in central fat. Changes were graded as mild 20%, moderate 17%, and severe 13%. One patient of 84 had diabetes but 16% had abnormal glucose tolerance. Hyperlipidemia was common but did not correlate clearly with lipodystrophy. Carr and colleagues have found a possible explanation that remains to be tested and proven. There is a structural similarity between the protease inhibitors and a specific retinol binding receptor on fat cells. If confirmed, this would help prove that it is in fact the current generation of PI's which cause this syndrome in some patients. However, no other group has seen anything as high as the 76% rate of changes. Are Australians more susceptible? More body conscious? Could it just be all that Australian beer and barbecue?

Session 177/12373: Changes in body habitus in HIV(+) women after initiation of protease inhibitor therapy

Dong and colleagues from Brown University reviewed 118 women on PI's. Nineteen of 118 women on PI treatment complained of a change in body shape after starting therapy. 8/19 patients were currently taking indinavir, 10/19 had switched to another PI after having discontinued indinavir, and 1/19 was taking another PI. The nineteen patients with a change in body shape reported the following subjective changes: increase in breast size (71%), increase abdominal size (71%), weight gain (47%), peripheral wasting (47%), gluteal wasting (29%), and buffalo hump (23%). Mean total body fat ( by bioelectrical impedance analysis -- BIA) was 38% (range 15-57%) and mean superficial fat was 38% (range 6-55%), both indicative of an abnormally high body fat. Abnormalities of lipids included: low HDL in 76%, elevated LDL in 47%, elevated triglycerides in 35%. The overall prevalence of body changes or lipid changes at an average of one year on therapy was 16%.

Poster Session 12319: Lipid abnormalities associated with use of protease inhibitors: Prevalence, clinical sequalae and treatment

This Minnesota study was a careful review of the charts, designed to address the question of what happens to serum lipids in patients who start protease inhibitors. Seventeen patients were on no treatment. Their average cholesterol was 181 mg (240 is considered the level at which intervention is important, less than 200 is what is desirable). Twenty-five patients on Ritonavir/Saquinavir had an average cholesterol of 162 before treatment and 239 on the double PI's. For 31 people on single PI and 2 nukes, the average cholesterol increased from 179 to 211. In addition 23% of patients had abnormal triglycerides. It appears from this, and other studies (including from our institution) that many people on PI's have cholesterol increases, usually moderate, but usually reflecting an increase in LDL cholesterol (sometimes called "bad cholesterol"). But it is not clear what the clinical impact is. Henry reported one new heart attack and one new case of angina. It is likely that other risk factors beyond cholesterol will determine the outcome: smoking, sedentary life style, diet, family history. For now, I don't know how best to manage moderate changes in cholesterol. If the PI's are working, I suggest stopping smoking, improving your diet, and exercising. There is no proof, but many feel that exercise helps control "protease paunch" as well.

Session 336/32173: Alterations in body fat distribution in HIV-infected men and women

Kotler's group in New York used dual xray absorbtiometry and body measurements to calculate the body composition of HIV infected patients and HIV negative controls. Recent patients had higher body weights, higher lean body mass and lower fat than HIV negative controls and HIV positive men and women from an earlier era. However, they stated that there was no association with use of protease inhibitors. The findings of this group seemed to be in contrast to many other reports, and the role of androgens and exercise, which may be more common in HIV postive people, were not accounted for.

What to do about these changes. We really do not know, although anecdotally, they often get better if the person changes to an NNRTI containing regimen. Two limited reports were interesting.

Poster Session 32164: Treatment of dorsocervical fat pads and truncal adiposity with serostim [recombinant human growth hormone] in patients with aids maintained on HAART

Because recombinant human growth hormone (rhGH or Serostim) is known to promote fat catabolism and lean body mass growth, these doctors treated one man and one woman with growth hormone. Both had developed buffalo humps, central fat and peripheral muscle wasting, associated with fatigue and hypertriglyceridemia. One patient also had mild hypercholesterolemia (serum cholesterol 200-350). Treatment was injections of growth hormone 6 mg/day subcutaneously for 12-24 weeks. After 12 weeks of GH therapy woman had total resolution of the buffalo hump, decreased central obesity, increased energy and improved strength. After 6 weeks of GH therapy the male patient had a notable 50% regression of the buffalo hump, and increased muscle strength. Side effects include mild joint pain in one patient and mild hypertension and elevation of pancreatic enzymes in the other. This response is interesting, but the staggering cost of growth hormone makes this of limited use unless Serono changes its pricing.

Poster Session 12287: Reversal of hyperlipidaemia and lipodystrophy in patients switching therapy to nelfinavir

This study reviewed 21 patients with lipodystrophy (it is important to note that there is no agreed on definition). All were taking combination therapy which included indinavir or saquinavir plus ritonavir, in addition to two nucleosides or non-nucleosides. Patients were switched to nelfinavir as the only PI and at least one other antiretroviral was also changed at baseline. Median serum triglyceride level before switching to nelfinavir was 3.7 nmol/l and dropped to 2.8 nmol/l after 3 months. Serum cholesterol did not change. At three months, the treating physician judged the lipodystrophy as stabilised/no change in 12 (57%) of patients, partially reversed in 7 (33%) patients, and progressive in 2 (10%) patients. The authors concluded that switching from ritonavir or indinavir to nelfinavir resulted in improvement in triglycerides and body shape in some, but definitely not all of the patients. This matches my experience, where one person had a dramatic return to their more normal body shape after switching, but others did not. See also the review by Cal Cohen on the virologic effects of switching from indinavir to nelfinavir in patients with suppressed viral loads.