Switching from Kaletra to raltegravir improves lipid profiles at the risk of viral rebound

People whose viral load is well controlled on a lopinavir/ritonavir (Kaletra) regimen may see significantly improved cholesterol and triglyceride levels if they switch to raltegravir (Isentress), but at the risk of losing viral suppression, delegates to the Sixteenth Conference on Retroviruses and Opportunistic Infections heard on Monday.

Raltegravir, the first-in-class integrase inhibitor, has been showing strong virologic results and a very good safety and tolerability profile in clinical trials. The ongoing BENCHMRK trials have investigated raltegravir in highly treatment-experienced individuals with viral loads higher than 1000 copies/ml and resistance to NRTIs, NNRTIs and protease inhibitors (PIs). The STARTMRK trials are also studying raltegravir as first-line treatment.

In this CROI session, Joseph Eron presented reports from two trials in a different patient population – HIV-positive individuals who are already virologically well controlled on a gold-standard PI-based regimen. In these studies (SWITCHMRK 1 and 2, also known as P032 and P033, respectively), all participants had – at baseline – undetectable viral loads (below 50 copies/ml by PCR or below 75 copies/ml by b-DNA) maintained for at least three months on a stable regimen containing lopinavir/ritonavir (Kaletra, 400/100 mg, twice daily). Any number of prior combinations or treatment failures were allowed, but lipid-lowering therapies in the past twelve weeks were excluded.

Participants were randomised either to remain on their successful Kaletra-based regimen, or to switch to raltegravir while leaving their other antiretrovirals (which had to include at least two NRTIs and no other PIs) unchanged. Primary endpoints were: mean percentage change in lipids at week 12; the proportion of patients whose HIV RNA remained below 50 copies/mL at week 24; and safety and tolerability up to 24 weeks. SWITCHMRK 1 and 2 were multicentre, double-blind, randomised clinical trials; the two studies involved different geographic regions but were otherwise identical in design.

Data from the two trials were presented separately. Patient characteristics and outcomes were largely consistent apart from the geographic differences. Most of the participants (approximately 82%) had been on Kaletrafor at least a year. About 80% were taking 3TC or FTC and very few had taken T-20 (enfuvirtide); otherwise treatment experience was quite varied. Some participants had as many as 19 years of prior therapy; a very few in the raltegravir arm (contrary to study protocol) were only on one other drug.

In SWITCHMRK 1 (protocol 032), 348 patients in North America, Europe and Australia were randomised and treated; 174 switched to raltegravir (of whom 25 discontinued) and 174 remained on Kaletra (of whom 17 discontinued).

In SWITCHMRK 2 (protocol 033), which also included Latin American, African and southeast Asian patients, 355 were randomised and 354 treated; 176 switched to raltegravir (with ten discontinuations), 178 remained on Kaletra (with six discontinuations).

Virologic results

The virologic outcomes were analysed as intent-to-treat, with everyone who did not complete treatment considered as a treatment failure.

At week 24 in SWITCHMRK 2, viral RNA was undetectable (Kaletra, for a difference of 5.8% (95% CI, 12.2 to 0.22). At week 24 in SWITCHMRK 1, the outcomes were 81% for raltegravir and 87% for Kaletra. The losses of viral suppression generally happened fairly early – most often by week four after the switch to raltegravir. A combined analysis of observed failure in both trials found a -4.8% difference overall (95% CI, -9.1 to -0.7).

Due to the poorer viral control seen with raltegravir, the SWITCHMRK studies have been halted, and the data are being analysed to better understand the reasons. The results were fairly consistent across many subgroups, but there were some variations, which are now being analysed further. (For example, women actually did slightly better on raltegravir, and the outcomes were much closer in Africa than in other regions.) Drug resistance to raltegravir was not seen frequently, but NRTI-, NNRTI-, and PI-associated resistance mutations were frequent. Eron did point out that many study participants had quite extensive histories of treatment failure; however, it remains to be seen whether this explains the differences in outcome.

Safety and lipid profiles

As in other studies, raltegravir use resulted in better triglyceride and cholesterol profiles than Kaletra. In SWITCHMRK 2, the mean change from baseline to week 12 was (for raltegravir versus Kaletra) -12% vs +1% for total cholesterol, -43% vs +8% for triglycerides, and -15% vs +3% for non-HDL cholesterol (p

Overall safety was good in both groups, with less than 5% grade-3 or 4 laboratory abnormalities, and no serious adverse events or deaths in either group. Clinical adverse experiences of all severities were relatively similar (70% for raltegravir vs 63% for Kaletra) as were drug-related adverse experiences (13% vs 20%).

Conclusions

While the trials have been halted, investigators will continue to analyse the existing data to explain the poorer virologic performance of raltegravir. The difference may well be explained by the antiretroviral background therapy. In the BENCHMRK studies, which also involved highly treatment-experienced individuals, raltegravir was accompanied by an optimised background therapy (OBT) selected to maximise the chances of treatment success, with no restrictions on what it might contain. In SWITCHMRK, other protease inhibitors were excluded and raltegravir was supported only by a nucleoside backbone. This may have been enough to produce a fully suppressive regimen with boosted Kaletra but not with raltegravir.

In the meantime, when asked about recommendations for clinical practice, Eron said that these results did not impact the first-line use of raltegravir, but that "we need to be extremely cautious about making one-drug switches to raltegravir. If you have an intact background, especially one containing T-20, that will likely continue to be safe and effective [with such a switch] – but otherwise it's unclear."