Boosted saquinavir has similar potency, better lipid profile than lopinavir/r in naïve patients: interim results from GEMINI study

Early results from a head-to-head comparison of boosted saquinavir with boosted lopinavir (Kaletra) suggest that the drugs achieve similar virological suppression, but that saquinavir produces significantly fewer cases of raised lipids.

The GEMINI trial, sponsored by Roche, who market saquinavir as Invirase, a 48-week open-label international study of the two drugs in 337 antiretroviral naïve patients. The other drugs in the regimen were tenofovir and FTC (Truvada).

Twenty-four week interim data were presented for the first 150 patients recruited at the Eighth Glasgow International Congress on Drug Therapy in HIV Infection.

The bottom line virological results are that on an intent-to-treat analysis, 75.3% of lopinavir patients achieved viral loads under 50 copies/ml compared with 69.4% of patients on saquinavir. The difference was not statistically significant (p=0.427). The results for viral loads under 400 copies/ml were 83.6% for lopinavir (LPV) and 80.6% for saquinavir (SQV).

The 150 patients were 47% (LPV) and 42% (SQV) female, with a median age of 36, and with quite advanced disease. Median CD4 cell counts were 121 cells/mm³ (LPV) and 134 cells/mm³(SQV) and viral load was 5.1-5.2 logs (around 150,000). Thirty per cent of the lopinavir patients had CD4 cell count under 50 cells/mm³ while 24% of the SQV patients did. Over half (57%) of the LPV patients had viral loads over 100,000 copies/ml while only 43% of the SQV patients did.

On an on-treatment analysis, counting only those patients who stayed on the assigned treatment, 85.9% of the LPV patients and 79.4% of patients on SQV achieved viral loads under 50.

There were 16 (22%) discontinuations of saquinavir, of which two were for adverse events, including one death (a boating accident), and 17 (22%) on lopinavir, of which four were for adverse events. These also included one death, which was drug-related: a Thai woman stopped her treatment at week 17 due to raised liver enzymes, ascites and jaundice and died of liver failure at week 24. She had mildly abnormal liver function at baseline but was not co-infected with hepatitis B or C.

There were seven patients who were counted as virological failures, two on lopinavir and five on saquinavir. Of note, however, is that the definition of virological failure was a viral load above 400 copies/ml after week 16 of the trial, so patients only had eight weeks in which to meet this definition.

Presenter Jihad Slim said that one of the failures on lopinavir and three on saquinavir had poor adherence, while the other failure on lopinavir had severe diarrhoea. Amongst the two adherent failures on saquinavir was one patient with a baseline viral load of 24 million, so 16 weeks in this case may not have been sufficient time in which to achieve undetectability. Adherence was similar between the two arms with 83-85% of patients achieving full adherence.

The one significant difference between the two treatment arms so far is in the proportion of patients with raised lipids. The proportion of patients with a higher than grade 1 lipid elevation at baseline was 17% (SQV) and 13% (LPV). By 24 weeks this had increased to 21% (SQV) and 38% (LPV) and this difference was statistically significant (p=0.036).

In detail, patients had cholesterol rises of 21 mg/dl (SQV) and 29 mg/dl (LPV); triglyceride rises were 29mg/dl (SQV) and 83 mg/dl (LPV). The proportion of patients with triglyceride elevations higher than grade 1 (400mg/dl, which suggests lipid-lowering therapy) was zero and 4% at baseline for SQV and PLV respectively; by 24 weeks it was 1% on SQV but 13% on LPV.

In terms of significant adverse events in general (grades 2-4), 48% of patients on SQV and 58% of patients on LPV experienced them; this difference was largely driven by gastrointestinal intolerance (13% on SQV and 23% on LPV: there were two discontinuations due to GI events on the LPV arm).

Roche were already giving the GEMINI trial significant publicity at the conference, though one questioner at the live presentation queried the necessity for publishing the interim trial results at all when there were no significant safety or potency signals sent by the trial.