Switch to PI-Sparing Studies

There was a good deal of discussion and abstracts at Lisbon on switching people from PI to non-PI regimens (either abacavir or NNRTI based triple regimens) in the hopes of alleviating metabolic complications, lipodystrophy syndrome, improving adherence, reducing pill burden, and improving quality of life.

Switching from PI to NVP Regimen in HIV+ Subjects under long-term successful treatment. In this study 138 patients were randomized to switch to NVP with the same NRTIs (n=104) or to continue with their current PI-containing regimen (n=34). They were on their PI regimen for more than 1 year, had above 200 CD4s, and had below 50 copies/ml for more than 6 months. All were NNRTI-naïve.

Prior exposure to PI therapy was about 17 months for both groups. Those in NVP group had 9.4 months below 50 copies/ml and those in PI group had 8.7 months below 50 copies/ml. CD4 count was not significantly different at baseline (640 in NVP group; 500 in PI group; CD4 % was 25 in NVP group and 22% in PI-group). Cholesterol and triglycerides were about the same in both groups ( 245 chol. and 290 trig.; about 75% in both groups had above 200 mg/dL chol.) About 60% in both groups had trig.above 200 mg/dL. And about 70% in both groups had lipodystrophic body-shape changes. There was an inversion of the HDL/LDL ratio.

Results.  Barreiro reported (ITT analysis) 29% in the PI arm and 11% in the NVP arm experienced a virologic rebound to over 50 copies/ml at 6 months. The difference was statistically significant. At the time of failure median values of viral load tended to be higher for those continuing on PI (54% on PI and 31% on NVP above 500 copies/ml). A higher incidence of poorer compliance was reported for those failing PI regimen (90% vs 22%). In 3/7 patients on NVP regimen they detected codon mutations to one of the drugs in combination. Two patients had NVP mutations and 1 with 3TC mutation. No resistance was detected in those failing PI regimen. Those switching to NVP had a mean loss of 35 CD4s but those continuing on PI had mean increase in CD4s of 64 at 6 months.

A trend towards a reduction in both trig. and chol. (10-15%) was seen in both groups but there was no statistical significance. In the PI group, 73% had chol above 200 at baseline and 64% at month 6. In the NVP group. 77% had chol above 200 at baseline and 58% at month 6. Regarding trig., 61% in the PI group had above 200 at baseline and 47% at month 6. In the NVP group. 57% had above 200 tig. at baseline and 45% at 6 months. None received lipid lowering drugs. But all were advised to follow a low-fat diet which could explain this observation.

Based on patient survey and physician examination, 72% had lipodystrophy features at baseline and 6 months later 50% in NVP group reported at least partial improvement while no one in PI group reported improvement. Half of them reported some improvement in these features after 6 months switching to NVP. Barreiro reported the pill burden reduction had a positive impact on quality of life. Patients were asked to rate quality of life from 1 to 10. Mean scores were 4.4 for PI regimen and 9.1 for NVP regimen.

Switch to NVP due to Lipodystrophy. Lydia Ruiz presented this study at Boerhinger Ingelheim symposium. HIV+ adults suffering clinically evident lipodystropy were enrolled. They had been on the same protease inhibitor therapy at least 9 months, had baseline CD4 above 100, and HIV-RNA below 400 copies/ml for 6 months. Patients were excluded for prior NVP experience and an OI within prior 4 weeks. Patients were randomized to ddI+d4T+NVP +antihistimines (first 15 days to prevent rash) or to d4T+3TC+same PI. Patients were assessed at screening, baseline, and every 4 weeks thereafter to check whether there was a rebound in HIV-RNA. 3TC was substituted by ddI in order to spare 3TC for rescue therapy in case virologic failure occurred. CD4s, triglycerides (TG), total cholesterol (CHOL), DEXA and anthropometric changes were determined at baseline and every  3 months.

Results.  63 of 106 patients recruited (31 in group 1; 32 in group 2) have completed 24 weeks of the study. Evidence of NVP rash was seen in 3 patients in group 1. Plasma viral load was above 400 in 4 patients in group 1 and 3 in group 2.

Treatment Failures and Toxicity (See Table 6):  

Table 6.



PI Group

Viral Rebound






Acute Hepatitis






Anemia (AZT)



Polyneuropathy (d4T)






Diarrhea (>3)



Prolonged fever *



*only 1 discontinued treatment for rash
** not associated with NVP

In NVP Group, they observed a significant decline in CHOL from 224 to 203 (p=0.05) at week 24 which continued to week 36. The change in TG was not statistically significant but reduced from 267 at baseline to 210 at week 24; trend appeared to continue on graph at week 36. CD4s increased by small amounts in both groups. About 95% of patients in NVP group had below 20 copies/ml at week 24 and 36. In PI group, about 82% had <20 copies/ml at week 24 and 36. At baseline, about 50% of patients in both group had below 5 copies/ml, while at week 24 70% in NVP group and 60% in PI group had below 5 copies/ml. Anthropometric measurements and DEXA parameters improved in NVP group but differences did not reach statistical significance by week 24. Improvements did occur in measurements at week 24 in abdomen pleat, waist circumfrance, hip circumfrance, and trunk DEXA change. Quality of life improved significantly in the NVP group when patients were surveyed.  Ruiz and Joep Lange, who chaired the session, offered their anecdotal experience. Lange said he has seen improvements in fat redistribution with patients who switched from PI to NVP. He also said, and Ruiz agreed, it could take a long time to see improvement: one year or more. And Ruiz agreed that some patients show improvements. Some researchers would likely say they want to see longer-term more objective data before they conclude switching to PI-sparing regimen can reverse fat redistribution. Ruiz reported normal lactic acidemia and L-carnitine were found in both groups at baseline and week 24. Later in session a report was presented that the use of antihistimines reduced the incidence of NVP rash.

Switch from PI based regmen to Efavirenz regimen. Graeme Moyle reported no improvements in overall cholesterol and triglycerides due to switch to EFV from a PI but viral load suppression was well maintained (9/9 <50 copies/ml at week 48; 19/20 <50 copies/ml at week 24). Patients reported they had some improvement in their body appearance. There were improvements in glucose tolerance test. Previous reports have observed improvements in good cholesterol and decreases in bad cholesterol.

This is a study of individuals who were on a PI therapy and requested to either stop therapy or be switched to another due to experiencing lipodystrophy. It's a non-comparative study group because participants did not want to be on their current regimen. So, its an open-label single arm prospective observational cohort of persons with clinical lipodystrophy and/or metabolic abnormalities virologically controlled on PI based regimen.

Patents were either self-reported or their doctor referred them to study as having lipodystrophy from a specialist lipodystrophy assessment clinic.  For the study they were clinically examined to confirm clinical body shape changes. There were a few individuals who only had metabolic elevations (cholesterol >6.5 mmol/l; fasting triglycerides >2.2 mol/l; diabetes). Participants had viral load <500 copies/ml (bDNA).

EFV 600 mg was substituted for the evening dose of PI with no overlap and people stayed on nucleoside analogues. Abacavir was added in some patients on dual PI or with prior transient NNRTI exposure.

At baseline and every 3 months they performed—fasting lipids, CD4s, viral load, insulin, testosterone, 2 hour glucose tolerance test, assessment of body composition by DEXA, anthropometrocs, and single slice-CT scan.

Moyle reported on two groups of patients: 9 (1 female, 8 male) who completed 48 weeks and 20 individuals ( 2 female, 18 male) who were out to 24 weeks. 8/9 and 18/20 were taking d4T/3TC. 7/9 and 15/20 were taking IDV as their PI. Three in the 20 person group were taking dual PI.  CD4 counts were 231 in 9 person group and 408 in 20 person group. All participants had received mean 20 months prior PI therapy.

At baseline in cohort 1 (n=9, 48 weeks), clinical manifestations were: facial (5), legs (6), arms (5), abdomen (5), all above (1), metabolic only (1).  Cohort 2 (24 weeks, n=20) clinical manifestations: facial (12), legs (13), arms (15) abdomen fat accumulation (16), all above (8), metabolic only  (2).

Results. In cohort 1 (48 weeks) 9/9 had viral load <50 copies/ml at week 48 and their CD4s increased 114. Two of 9 changed therapy: one person switched from EFV to NVP after 4 weeks due to concentration problems due to EFV, and a second person who switched from d4T to abacavir after 1 month due to neuropathy. In cohort 2 (24 weeks, n=20) 5 people added abacavir as mentioned above. One person switched from d4T to abacavir at week 40 due to elevated triglycerides. 19/20 had <50 copies/ml at week 24. One person stopped therapy due to PCP.

In both cohorts, cholesterol and triglycerides were modestly but not significantly increased at week 48 compared to baseline. Moyle reported significant improvements in glucose tolerance. By patient survey partial improvements in appearance were reported. Also Moyle reported significant increased weight that was initially mostly in fat free mass although in cohort 1 there was non-significant fat increase. In cohort 1 Moyle reported significant improvement (12%) in VAT (visceral adipose tissue- stomach fat) by CT scan. Moyle reported a significant correlation between baseline insulin resistance and triglycerides, consistent with diabetes, and insulin resistance and VAT (CT-scan). Overall weight increase despite declining fat accumulation in the stomach, may be due to reported improved dietary patterns. Also, Moyle is suggesting the improved glucose is related to reduced fat in stomach, suggesting elevated glucose may be associated with elevated lipids and fat redistribution.

Moyle reported 6/7 individuals who had abnormal glucose tolerance at baseline (>7.8 mmol/l) had a normal glucose tolerance test at week 24. One person had family history of diabetes. Regarding triglycerides 2/14 individuals reduced triglycerides from above to below normal (2.2 mmol/l). As well 2 individuals had normal baseline values that became abnormal at week 24. The same response was seen with cholesterol.

Switching Protease Inhibitor to Abacavir: Preliminary 16-Week Data (LISBON).  This report contains results reported on a study switching individuals with below 50 copies/ml from a PI to abacavir.

N Clumeck from the Hospital St. Pierre in Brussels, Belgium reported on preliminary study data at 16 weeks from CNA30017 which explored switching individuals to abacavir in place of their PI. 211 individuals were randomized in an open-label comparison to continue their PI+2 NRTIs regimen or to continue their 2 NRTIs plus abacavir in place of the PI. The objectives of the study are to compare long-term durability of antiviral response (48 weeks), plasma HIV-1 RNA and CD4 profiles, safety and tolerance, and quality of life. Treatment failure is defined as above 400 copies/ml on two consecutive occasions or premature discontinuation of study treatment for any reason.

Individuals had to have a minimum of 6 months prior duration on their current PI+2 NRTI regimen, HIV-RNA undetectable since the beginning of ART, and below 50 copies/ml at screening. 80% in the abacavir (ABC) arm and 78% in the PI arm had received ART for over 12 months prior to study entry: I'm assuming the PI regimen people were taking at study entry was their first treatment, and that they were treatment-naïve previously. I did not see this question addressed in the presentation or abstract. Here is a list of drugs people were taking at prior to study (See Table 7):    

Table 7. Prior Drug Use

NRTIs (%)






• D4T/3TC



PI (%)  















At baseline median CD4s in both arms were about 505 (ranges varied from 70-1500); median triglycerides (mmol/l) were 1.68 in the ABC arm and 1.64 in the PI arm; median cholesterol (mmol/l) was 5.18 in ABC arm and 5.34 in PI arm. It appears that lipids testing were non-fasting.

Results.  105 were randomized to ABC and 106 to continue PI. 102 received ABC treatment and 103 PI treatment. There were 3 discontinuations due to adverse event in the ABC arm and 2 in PI arm. There were 2 hypersensitivity reactions in the ABC arm. There were 2 virologic failures in each arm. One withdrew from the ABC arm and 3 in the PI arm. And 1 was lost to follow-up in the PI arm. In total, 6 in the ABC arm and 8 in the PI arm were considered treatment failures. There was a trend for decreasing triglycerides and cholesterol at week 16 following the switch in the ABC arm. While it appeared on the slide graph that the lipids in the PI arm remained about the same.

75% in the ABC arm and 57% in the PI arm experienced at least 1 adverse event. The most frequent adverse events were nausea (75% in ABC arm v 57% in PI arm), diarrhea (16% v 6%), headache (5% v 10%), and malaise & fatigue (14% v 5%). Individuals in ABC arm initiated ABC just 16 weeks ago.