Kaletra monotherapy studies yield conflicting results

Two studies, both published in the January 30^th edition of AIDS, have reported disparate findings on the 48-week effectiveness of lopinavir/ritonavir (Kaletra) monotherapy, as compared to Kaletra used in conjunction with dual nucleosides.

A Spanish study found monotherapy statistically non-inferior to triple therapy at maintaining viral suppression in patients who were already virologically suppressed. In contrast, the French MONARK study found lower rates of suppression in treatment-naive patients who were started on monotherapy versus those started on Kaletra, AZT and 3TC.

Although combinations of three or more antiretroviral agents remain the standard of care for HIV infection, Kaletra monotherapy has appeared to be surprisingly effective, showing similar rates of virologic suppression when compared to standard triple therapy in many preliminary clinical trials. Ongoing results from two such studies – France's MONARK (MOnotherapy ANtiretroviral Kaletra) - and Spain's OK (Only Kaletra) – have been generally encouraging.

However, emerging differences between the findings of these two trials were reported at last year's EACS conference. Published 48-week results from both studies now confirm the EACS reports.

The MONARK study

In this prospective clinical trial, 136 treatment-naive patients were randomised to either Kaletra alone or with 3TC/AZT (Combivir) as an initial regimen. Participants aged 18 years or older were recruited from centres in France, Poland, Italy, Germany and Spain between October 2003 and February 2005. At baseline, all participants had HIV RNA levels below 100,000 copies/ml and CD4 cell counts above 100 cells/mm³. The median age was 36 years, 35% were female, the median viral load was 25,000 copies/ml and the median CD4 count was 232 cells/mm³.

Eight-three and 53 patients were assigned to the monotherapy and triple-drug arms, respectively (an initial 1:1 randomisation was changed to 2:1 to include more patients in the monotherapy arm). The primary study endpoint was viral suppression, defined as HIV RNA levels below 400 copies/ml at week 24 and below 50 copies/mL at week 48. Follow-up will continue until 96 weeks; current published results reflect 48-week data.

At 48 weeks, fewer participants were virologically suppressed on monotherapy than on triple therapy, although the difference was only statistically significant for on-treatment analysis. By intent-to-treat analysis, 53 of 83 patients (64%) in the monotherapy group and 40 of 53 patients (75%) in the triple-drug group achieved the primary endpoint (p = 0.19). The difference between the groups was 11.6% (95% confidence interval [CI], -0.04% to 27%).

There were 16 discontinuations before week 48 in the monotherapy arm and twelve in the triple therapy arm, resulting in 67 patients on monotherapy and 41 on triple therapy at week 48. There was a significant difference by on-treatment analysis, with 80% and 98% response in the two arms, respectively (p = 0.02) – a difference of 17.3% (95% CI, 7% to 28%).

Genotypic PI resistance mutations (L76V, M46I) were seen in three of 21 patients on monotherapy who underwent resistance testing; phenotypic changes in lopinavir sensitivity were either absent or ranged from 1.13- to 2.69-fold increases from reference.

Serious adverse events were reported in ten patients (12%) on monotherapy and four (8%) on triple therapy.

The MONARK investigators conclude that their results "suggest that lopinavir/ritonavir monotherapy demonstrates lower rates of virological suppression when compared with lopinavir/ritonavir triple therapy and therefore should not be considered as a preferred treatment option for widespread use in antiretroviral-naive patients."

The OK study

This open-label clinical trial studied 205 patients who were already successfully virally suppressed (HIV viral load below 50 copies/ml for at least six months) on antiretroviral therapy consisting of Kaletra plus two nucleoside analogues (or one nucleoside plus tenofovir). Patients were then randomised to either continue their treatment, or discontinue the two nucleosides and continue on Kaletra alone.

Participants were recruited from 28 sites in Spain between December 2004 and July 2005. The median age was 41.5 years, 20% were female, the median baseline viral load was 158,000 copies/m before starting antiretroviral therapy, and the median CD4 count at enrollment was 474 cells/mm³. None had a history of suspected or confirmed virologic failure while on protease inhibitors. Of the original 205 patients, 100 were randomised and began treatment on monotherapy; 98 began treatment on triple therapy.

The primary endpoint was lack of therapeutic failure, where failure was defined as HIV RNA higher than 500 copies/ml, discontinuation, or loss to follow-up. For the primary analysis, patients in the monotherapy arm who restarted their nucleosides due to lack of response were not counted as "treatment failures", as long as they were successfully resuppressed when the nucleosides were reintroduced. Lack of resuppression, or any other change in therapy besides this reintroduction, was counted as failure.

At 48 weeks, the percentage of patients without therapeutic failure by this definition was 94% in the monotherapy group versus 90% in the triple therapy group. The difference was not statistically significant: the 95% CI for the difference was -11.8% to 3.4% (p = 0.28), "fulfilling the criteria for non-inferiority". (In this analysis, those who switched back to triple therapy and reachieved suppression below 50 copies/ml were not counted as treatment failures.)

The lack of significant difference persisted in a more rigorous intent-to-treat analysis, with missing data or reinduction of triple therapy counted as failure. In this analysis, the percentage of patients with HIV RNA 50 copies/ml at 48 weeks was 85% in the monotherapy group versus 90% in the triple therapy group (95% CI for difference, -4.4% to 14%; p = 0.31).

At week 48, six out of the 100 patients (6%) in the monotherapy group had lost virological suppression, compared to three out of 98 (3%) in the triple therapy group (p = 0.31). In the monotherapy arm, five of the six resumed nucleoside treatment; four of these regained suppression, which was maintained for a median of 56 weeks thereafter. The remaining patient showed genotypic resistance to lopinavir and was removed from the trial.

Moderate or severe drug-related adverse events were reported in three patients (3%) in the triple therapy group and none (0%) in the monotherapy group (p =.08).

The investigators concluded that "for subjects who are currently suppressed while receiving lopinavir-ritonavir and two nucleosides, lopinavir-ritonavir monotherapy followed by reintroduction of nucleosides as needed is a therapeutic strategy as effective as continuing triple therapy." However, "episodes of low level viremia were more common in patients receiving monotherapy."

Why the differences?

The critical difference between the studies is the participant treatment history – those in MONARK were purely treatment-naive, while the OK participants were already successfully virally suppressed on Kaletra-based treatment. This may be the predominant factor in explaining the apparently poorer efficacy of monotherapy in MONARK.

However, other differences should be noted. MONARK was designed as a 96-week pilot study to demonstrate preliminary safety and efficacy of Kaletra monotherapy in treatment-naive patients; it was "not specifically powered to demonstrate either equivalence or non-inferiority but to identify whether lopinavir/ritonavir monotherapy provided adequate safety and efficacy relative to recommended lopinavir/ritonavir triple therapy." By contrast, the OK trial, while relatively small, was statistically designed as a non-inferiority trial.

The primary analysis in the OK trial allowed for virologically failing patients on monotherapy to switch back to triple therapy: these cases were still counted as successful as long as suppression was reachieved. Monotherapy, however, did not require this broad definition of "success" to remain comparable; results remained similar (although lower for monotherapy) when these cases were removed in the intent-to-treat analysis.

Baseline viral loads were about 1 log₁₀ higher in the OK study (before starting antiretroviral therapy; all were below 50 copies/ml at actual study commencement); however, baseline characteristics were comparable between study arms in each trial. The EACS report mentioned above comments extensively on other factors – including adherence and HIV viral subtype – which may have affected the outcomes seen in these trials.