At Retrovirus there were 4 studies reported on switching individuals to efavirenz and 3 studies on switching to abacavir from PI regimens. These studies monitored lipodystrophy, fat redistribution, and insulin & lipid values. As well, there were 2 studies on switches to nevirapine. The Ruiz nevirapine switch study was reported in extensive detail elsewhere the NATAP web site in the Conference Reports section. It showed increasing improvements in cholesterol and triglycerides. There were no significant (statistical) body-shape changes through anthropometric or DEXA measurements at 12-months follow-up. But if you look at both the DEXA and anthropometric changes from baseline through follow-up, you will note in a number of cases that the NVP arm showed a stabilization while for the individuals in the PI control arm the trend was a continuing worsening. Ruiz and others have suggested that perhaps body changes will improve in more time. A second nevirapine switch study with only 6 months follow-up so far showed similar results.

P Tebas reported on this second study at Retrovirus on switching the PI to nevirapine 200 mg bid. The study included 40 patients with undetectable viral load for at least 6 months. Nucleosides were maintained. All patients received prednisone 40 mg once daily for 7 days. Virologic failure was confirmed VL>200 copies/ml. The median follow-up is 6 months. Six patients (15%) developed severe rash and switched to EFV. 1 patient left the study because of drug related hepatitis. There was 1 virologic failure whose VL became undetectable after switching back to a PI regimen. Fasting triglycerides decreased by 31% (p=0.0005). Cholesterol decreased 11% (p=not significant). There was a decrease in insulin, proinsulin and C-peptide levels. 3/10 patients had evidence of insulin resistance at baseline using an insulin tolerance test. All 3 normalized their insulin response. Fat redistribution measured with DEXA (central to appendicular fat ratio) did not improve in the 6 patients with measurements at baseline and 6 months. But as stated above, in the Ruiz study they started to see a trend towards improvements in body changes.

In the nevirapine switch studies improvements in cholesterol, triglycerides and insulin appear to begin by 12 weeks. But rash and hepatitis can occur. So far in the efavirenz studies discussed below, which are out to 6-10 months in follow-up, they don't observe improvements in lipids. But viral load suppression is well maintained. NVP is approved for twice daily dosing but is used once daily by some doctors. EFV is approved for once daily dosing. There two switch studies to abacavir discussed at Retrovirus. As Graeme Moyle reported in his report on lipodystrophy for the NATAP web site, the switch to abacavir shows some preliminary promise. The data shows improved cholesterol, triglycerides and insulin sensitivity. In the Opravil study described below, preliminary follow-up was 12 months and there were more virologic failures in the abacavir arm than the PI arm (9 vs 5). They did not monitor parameters to measure changes in body habitus. But, in the other study also discussed below follow-up was only 6 months and they did monitor changes in body habitus. They saw no significant changes in fat redistribution, but again maybe it takes longer to see actual improvements in fat redistribution.

None of the studies described here address the situation of a person with experience with several anti-retroviral regimens. For example, the person who started with AZT monotherapy, added 3TC later, and then switched to HAART. Switching regimens is more complicated for them because they have less treatment options. When considering a treatment change for fat redistribution and elevated lipids, you must weigh the risk of losing virologic control. Using lipid lowering agents is an option for elevated lipids but there is no evidence that they nor Human Growth Hormone reverse fat redistribution. In preliminary study, HGH did appear to reduce fat accumulation but did not nor is it expected to reverse fat depletion in the face, limbs, etc.

There are two additional reports on lipodystrophy and fat redistribution at the Retrovirus Conferences on the NATAP web site. One is authored by Dr Moyle and the other written by me. They're both posted in the Conference Reports section in the section on reports from the 7^th Conference on Retroviruses and Opportunistic Infections.

Identification of Fat Redistribution / Metabolic Anomalies in a Cohort Treated by 2 NRTIs + 1 PI, and Absence of Significant Modification Following PI-Substitution (to EFV)

S Gharakhanian and others from the Rothschild Hospital in Paris set out to identify clinical/laboratory features and risk factors of lipodystrophy in a group treated by 2 NRTIs & 1 PI. Itνs a cross-sectional study of patients treated by PI for > 3 months with comprehensive monitoring of relevant lab values including 14 clinical & anthropometric endpoints, lipid, glucose (75gr OGTT), insulin dosage. In a subset of these patients with HIV RNA < 500cp/ml and at least one clinical and laboratory abnormality, PI was substituted for efavirenz, with NRTI unchanged, and patients were followed-up prospectively for a year.

There were 624 patients studied in a referral center -- (84% M, 16% F), aged 40 yrs. (±9), AIDS in 31%. The frequency of clinical abnormalities were high (85%) and there were differences between the sexes. The authors reported 3 types of clinical abnormalities-- 20% had atrophy (fat depletion), 22% of patients had fat accumulation (22%), and 58% were what the authors called mixed/heterogeneous which I assume means they had fat redistribution meaning fat depletion and fat accumulation. 34% had glucose intolerance/diabetes. 42% had hyperinsulinemia; Cholesterol was NCEP high/very high in 36% and triglycerides in 14% of patients. Duration of therapy is a notable risk factor.

Thirty-three patients from this cohort treated by PI for a median of 24 months [5-35] mainly by indinavir (84%), associated with d4T/3TC in 81% of cases underwent substitution of PI by efavirenz. No significant difference was observed at month ten for weight, fat redistribution patterns, lipid & glucose/insulin abnormalities. 4 patients withdrew for adverse events of laboratory anomalies.

(n=34, 6 months after switch--VL suppression maintained, no change in lipids or fat redistribution)

P Viciana and others from Hospital University Virgen del Roclo in Seville, Spain, reported on this studying of switching people with undetectable viral load (200 copies/ml) and fat redistribution from their PI regimen to an efavrienz regimen. Nucleoside analogue therapy was unchanged.

The 39 patients in this study were on PI therapy and had <200 copies/ml for a median of 11 months (at least 6 months). Patients were assessed every 3 months by BMI (body mass index), waist/hip ratio (WHR), face photography, and a questionaire of subjective body changes was completed by patient. As well, fasting cholesterol, HDL-cholesterol (good cholesterol), triglycerides, glycemia (sugar), CD4s, and HIV viral load were performed at baseline and months 1, 2, 3, 4, 6, 9, and 12. For this analysis viral load below the level of detection (BLD) was equal to 20 copies/ml.

Fifteen women and 24 men were actually enrolled, and completed 6 months of follow-up. The median duration of previous PI-containing regimen was 17 months (range: 13-21 months). Nucleoside therapy with d4T and 3TC was used by 34 (87.1%) patients. Viral load was below 200 copies/ml in 92.3% and below 20 copies/ml in 81% of patients at day 0. At month 6 VL was BLD in 82.3%, and <200 in 91.1%. The median CD4 count rose from 475 cells to 601cells. Most of patients (76%) subjectively reported a partial improvement in their body shape (although never back to their prior appearance before PI therapy) and WHR decrease from 0.94 to 0.93 (p=ns). However, BMI did not change (21.96 versus 22.06 kg/m2) and metabolic parameters showed a slightly impairment: mean triglycerides arise from 169 mg/dl at baseline to 201 mg/dl at month 6, cholesterol 202 mg/dl to 209 mg/dl, HDL cholesterol 36 mg/dl to 50 mg/dl, LDL cholesterol 113 mg/dl to 117 mg/dl and glucose 85 mg/dl to 92 mg/dl.

Bonnet and others from Hospital Pupan in Toulouse, France reported these preliminary results from a prospective non-comparative study of 43 (male=35, female=8) HIV-infected patients. They had at least 1 year on PI regimen, viral load <50 copies/ml (by 2 consecutive blood samples- 3 months before study and on day of starting study), clinical lipodystrophy including loss of peripheral fat and increase in central fat, and triglycerides > 2 mmol/l. The nucleosides in their regimens were unchanged and efavirenz was substituted for the PI.

On day 0 (D0), month 3 (M3) and month 6 (M6), all patients were given a routine physical examination (including body weight and body mass index), measurements of legs, thighs, abdomen (ombilic, hips) and chest perimeters, routine biological parameters (including CD4-cell count and viral load). Additionally they received complete evaluation of lipid profile (including triglyceridaemia, total cholesterolaemia, HDL, LDL, VLDL, Apo A1, Apo B, Apo CIII, Apo E, LpB:e, LpB:C3, Lp:a measurements, and v) evaluation of fat/thin masses distribution by dual X-ray absorptiometry (DEXA). Bonnet reported most patients felt "better" after the switch to EFV because medications were easier to take and compliance was easier. He reported no serious adverse side effects although some patients reported insomnia several months after starting EFV. There were no discontinuations reported at month 6. Of 37 patients from whom viral load was attainable 84% (31/35) were still <50 copies/ml. Bonnet said that although he didn't show the data a comparison between those taking AZT or d4T showed no differences.

CD4 count and undetectable viral load suppression was maintained in all patients, but preliminary results show no improvement in lipids, or lipodystrophy syndrome by physical exam and DEXA (at 3 to 6 months after study started).

Martinez and others from Barcelona reported on 20 patients (11 males, 9 females) treated with 2 NRTI and at least 1 PI with sustained viral suppression who developed metabolic abnormalities and body changes consistent with lipodystrophy and accepted to switch from PI to EFV. Body mass index (BMI), waist:hip ratio (WHR), regional fat thickness (echography), fasting total and HDL cholesterol, triglycerides, glucose, insulin, CD4 cells and viral load were measured at baseline and every 3 months.

In this study, after 6 months following the switch to EFV the authors reported that triglycerides and fasting insulin resistance improved, and viral load suppression was generally maintained, but glucose and cholesterol was unchanged. Waist/hip ratio decreased from 0.92 to 0.87 (p=0.06). Eleven patients (55%) self-reported a partial improvement in body changes that was related to a significant decrease in WHR and intra-abdominal fat thickness. Perhaps most importantly, peripheral fat depletion did not change.

At baseline, the median age was 40 years. Body changes observed included one or more of the following: buffalo hump (n=2), breast enlargement (n=8), abdominal obesity (n=19), and fat loss in arms (n=19), legs (n=20), buttocks (n=18), and face (n=20). Most frequent therapy at baseline was d4T+3TC+indinavir (n=11, 55%). Hypertriglyceridemia (≥200 mg/dL) was present in 17 (85%), hypercholesterolemia (≥200 mg/dL) in 14 (70%), and impaired fasting glucose (≥110 mg/dL) in 8 (40%) patients; CD4 T cells were 280/mL (range: 64-942); and HIV-1 RNA had been <200 copies/mL a median of 14 months (range: 3-24).

Six months after switching, triglycerides (decreased 31%) (p=0.03) and fasting insulin resistance index (decreased 28%) (p=0.03) improved, but total and HDL cholesterol and glucose did not change. WHR decreased from 0.92 to 0.87 (p=0.06). Eleven patients (55%) reported a partial improvement in body changes that was related to a significant decrease in WHR and intra-abdominal fat thickness. Peripheral fat did not change. CD4 cells remained stable and viral load became detectable only in one patient (555 copies/mL). Severe adverse effects occurred in 3 patients in whom EFV was substituted by nevirapine.

Rozenbaum and others from Rothschild Hospital in Paris and Glaxo Wellcome

reported preliminary results from a substudy analysis of 31 patients who were randomized to continue their PI regimen or switch the PI to abacavir (16 in the abacavir arm, 15 in continuing PI arm). None experienced viral load above 400 copies/ml.

CNA30017 is an ongoing, international, randomized study evaluating the efficacy and tolerance of 2NRTI/abacavir (ABC) vs continued 2NRTI/PI for>6 months with no virological failure (HIV-RNA<50cop/ml) since initiation of ART therapy. Patients were randomized to continue current treatment (PI) or switch PI to ABC. This sub-study was conducted in France to evaluate lipid, glucose and insulin levels and morphologic (skin folds, body circumference) anomalies at Day 1 of study and week 4, 12, 24, 36, and 48. Prior NRTI for ABC vs PI groups were -- Combivir 11 vs 7, 3TC 4 vs 8, ZDV 3 vs 2, d4T 2 vs 7 and ddI 1 v s1. The median CD4+ cell count was 604 vs 564 at baseline.

Week 24 results-

• there was a median reduction in cholesterol in the abacavir group of -0.75 mmol/l vs no change in the PI group
• median reduction in triglycerides of -0.09 mmol/l in abacavir arm vs -0.04 in PI arm that was not statistically significant (p=0.41)
• 4/6 subjects with „ 1 glucose/insulin abnormality in the abacavir (ABC) arm experienced a normalization at week 24 vs 1/2 in the PI group. No subjects in the ABC group reported development of „ 1 glucose/insulin abnormality vs 3 in the PI group. Again, this is a trend
• median change in fasting glycemia between baseline and week 24 was +0.27 mmol/l in ABC vs +0.12 mmol/l in PI (p=0.37, not significant)
• median change in glucose following the OGTT between baseline and week 24 was -0.05 mmol/l in ABC vs 0.61 mmol/l in PI (p=0.72, not sig.)
• the median change in insulin tolerance test was -1.0 gU/ml for ABC vs +8.3 gU/ml for PI (p=0.20, not significant)
• median increase of the HOMA test of 0.06 in the ABC group vs 0.46 in the PI group (p=0.49, not significant)

There were no statistically significant differences observed in all skin folds measurements, chest/waist/hip circumferences and weight. Median waist/hip ratios (0.95cm) were similar at week 24 in both groups. The authors reported central obesity resolved in 50% of subjects switching to ABC while in the PI group it remained stable. Nine subjects in each arm had clinical signs of lipodystrophy at baseline. At week 24, 3 in the ABC arm reported improved lipodystrophy and 1 reported a worsening. None of the subjects in the PI arm showed an improvement, and 3 had a worsening.

Opravil representing the Swiss HIV Cohort Study and others from Glaxo Wellcome reported orally on this study of switching individuals to an abacavir triple nucleoside regimen. Patients (n=164; 84 in PI group and 80 in abacavir group) with viral load <50 copies/ml for 6 months or more following initiation of PI regimen, and no codon 215 AZT resistance mutation at baseline were randomized to either switching to abacavir (300mg bid) + Combivir (300mg AZT+150mg 3TC) or to continue with their PI regimen. Study goals were to-- explore induction/maintenance; to monitor and compare immunological and virologic parameters after the randomization; to look at lipid abnormalities; and, the possible prevention of the development of or reversal of preexisting lipodystrophy; the effect on adherence; and, to look at the ability to salvage virologic failures. Viral load was measured every 4 weeks and virologic failure was defined as 2 consecutive values >400 copies/ml. The mean duration of prior HAART therapy was 22 months, and the mean duration of viral load below 400 copies/ml was 15.5 months. HIV-RNA was <50 copies/ml at screening for study. CD4 were about 515 in both groups at baseline.

Opravil reported interim results. After a median of 48 weeks follow-up, there were 5 virologic failures in the group continuing on their PI regimen (n=44„ 48 weeks; 87„ 36 weeks) and 9 in the group switching to the abacavir regimen (n=49„ 48 weeks; 71„ 36 weeks)--ITT analysis, despite 95% reported adherence. Discontinuation/death-- 5 in PI arm and 4 in abacavir arm.

There was a drop in cholesterol (non-fasting) which appeared at about 4 weeks after switch to the ABC group (p£ 0.001). It was a significant change compared to baseline and between groups. There was a small improvement in HDL cholesterol at week 48 in ABC arm but it was not significant. Non-fasting triglycerides improved at week 48. Significant for change from baseline and between groups (p£ 0.05). Decreased triglycerides again occurred at about 4 weeks. Adherence was reported at a mean >97% for both groups by self-assessment and nurse assessment.

Opravil reported on 8 individuals in the ABC arm who were virologic failures and how they responded to salvage therapy. Failures occurred at weeks 0, 5, 8, 9, 9, 16, 17, and what appeared to be 86 weeks. As required by study design each individual had to have two viral loads above 400. For 7/8 the first VL failure ranged from 400 to 2200 copies/ml. For 1 person it was 15,000 copies/ml. Their 2^nd VL values were usually a little higher and ranged from 800 to 2400. The person with 15,000 had 7,000 on the second test. 6/8 switched to a PI regimen (NFV, IDV, RTV, RTV/SQV). 5 of those 6 had <50 copies/ml. One person remained on ABC but switched their nucleoside regimen from d4T/3TC to Combivir. Their initial failing viral tests were 480 and 890 and they also had <50 copies/ml upon follow-up. One person had viral loads of 480 and 1200, poor adherence reported, made no change to their regimen, and had a VL of 117 upon follow-up.