Final results confirm twice-daily boosted fosamprenavir is comparable to twice-daily Kaletra in treatment-naive

A large randomised open-label non-inferiority head-to-head comparison of two protease inhibitors (PIs) - ritonavir-boosted fosamprenavir (Telzir) with ritonavir-boosted lopinavir (Kaletra) - as first-line therapy in treatment-naïve individuals, has found that the two are comparable in terms of efficacy, safety, and tolerability over 48 weeks. The final results of the KLEAN study were published in the August 5^th edition of The Lancet, and were also presented last week to the Sixteenth International AIDS Conference in Toronto. The investigators believe that fosamprenavir should be elevated to the same status as lopinavir in HIV treatment guidelines: from an alternative choice to a preferred choice.

In May, GlaxoSmithKline (GSK) and Vertex Pharmaceuticals, who market fosamprenavir (known as Telzir in Europe and Lexiva in North America) announced the results of the KLEAN (Kaletra versus Lexiva with Epivir and Abacavir in ART-Naive patients) study in a press release, which focused on the study's primary endpoint - the proportion of individuals with a viral load below 400 copies/ml after 48 weeks. The press release said that in intent-to-treat analysis, 73% of participants receiving fosamprenavir/ritonavir achieved a viral load of less than 400 copies/ml compared with 71% of those on lopinavir/ritonavir.

The press release provided little information on toxicity and tolerability other than reporting that "both regimens were generally well-tolerated," with 6% of all participants withdrawing from the study due to adverse events.

The full results confirm that twice-daily ritonavir-boosted fosamprenavir is similar to the twice-daily dose of soft-gel ritonavir-boosted lopinavir used in the study when taken in combination with GSK's once-daily fixed dose combination nucleoside backbone pill (known as Kivexa in Europe, Epzicom in North America) containing 600mg abacavir and 300mg 3TC (lamivudine).

Baseline characteristics

Baseline patient characteristics were similar in both arms, with an identical proportion of women (22%), and both arms had a median HIV viral load of 5.1 log₁₀ copies/ml, although a slightly higher proportion of patients in the fosamprenavir arm had a viral load above 100,000 copies/ml (55% vs. 53%).

The median age of those randomised to fosamprenavir was slightly older (38 vs. 37) and there were slightly more people of white ethnicity in the fosamprenavir arm (61% vs. 56%) and of black and Hispanic ethnicity in the lopinavir arm (33% and 9% vs. 29% and 7%, respectively).

Participants in the fosamprenavir arm had a slightly lower median CD4 cell count (188 vs. 194 cells/mm³), although the lopinavir arm had slightly more people with CD4 counts below 50 cells/mm³ (18% vs. 15%).

Broadly similar proportions of participants had hepatitis B (4% on fosamprenavir, 3% on lopinavir), hepatitis C (12% on fosamprenavir, 9% on lopinavir) or both (

None of the baseline differences between the arms were statistically significant, however.

Intent-to-treat exposed analysis

In total, 887 patients were enrolled in the study, 443 were randomised to receive fosamprenavir and 444 received lopinavir. However, nine patients in the fosamprenavir arm did not receive treatment (five due to patient decision, one due to job relocation, one due to non-compliance, and one due to genotype results). In addition, another patient who had difficulty swallowing the ritonavir capsule (it was regurgitated whole within minutes of the first dose) was not included in the intent-to-treat analysis. This led the investigators to provide a slightly modified intent-to-treat analysis which they termed, 'intent-to-treat exposed (ITT-E)'.

Resistance data

A total of 26/434 (6%) on fosamprenavir and 30/444 (7%) on lopinavir had unconfirmed virological failure, although the proportion of participants with study protocol-defined confirmed virological failure were fewer: 16/434 (4%) and 24/444 (5%), respectively. Samples taken at baseline and at 48 weeks were analysed for genotypic and phenotypic failure for 14 of the 16 confirmed fosamprenavir failures and 21 of the 24 confirmed lopinavir failures.

Ten of the 14 fosamprenavir failures and 14 of the 21 lopinavir failures produced no measurable mutations. Of the remaining four fosamprenavir failures analysed, three lamivudine-related mutations (M184I, V, or M/V) and two PI-associated mutations were found.

Of the remaining seven lopinavir failures analysed, one TAM-associated mutation (M41M/L), four lamivudine-associated mutations, two NNRTI-associated mutations (V106V/A) and two PI-associated mutations were found.

However, none of the PI mutations detected (I54I/L, K20K/R, and I62I/V) were associated with reduced susceptibility to either fosamprenavir or lopinavir.

The median time to virological failure could not be estimated, say the investigators, due to "the small number of virological failures."

Efficacy data

Of the 434 patients who started fosamprenavir, 315 (73%) were classified as responders (defined as remaining on treatment throughout the study and having a viral load below 400 copies/ml at 48 weeks). Of the 444 patients who started lopinavir, 317 (71%) were responders at 48 weeks. These data established non-inferiority of fosamprenavir to lopinavir.

Using the gold standard measurement of virological suppression of the proportion of participants with less than 50 copies/ml at 48 weeks, 285 (66%) achieved this in the fosamprenavir arm and 288 (65%) achieved this on lopinavir.

Efficacy was also similar when participants were stratified according to baseline viral loads or according to their CD4 count at baseline. Equivalent efficacy rates were found using other analyses, including missing/discontinuation equals failure, and per-protocol observed. This indicates, say the investigators, that “the primary efficacy results were robust regardless of the analysis method used.”

CD4 cell count increases from baseline to week 48 were also broadly similar between the arms, with a median increase of 176 cells/mm³ in the fosamprenavir arm and a median increase of 191 cells/mm³ in the lopinavir arm.

Similarly, progression to AIDS or death was comparable in both arms: 11/434 (3%) in the fosamprenavir arm and 11/444 (2%) in the lopinavir arm.

Five deaths occurred during the study, four in the fosamprenavir arm and one in the lopinavir arm, none of which were considered to be related to the study medication. However, one of the deaths in the fosamprenavir arm appeared to be related to HIV disease progression as a result of virological failure. This individual died of MAI, CMV, oesophageal herpes, Kaposi's sarcoma, and Stevens-Johnson syndrome.

Comment

The equivalence of ritonavir-boosted fosamprenavir with soft-gel Kaletra suggests to the study's authors that fosamprenavir should be elevated in US and UK treatment guidelines to an equally preferred regimen alongside Kaletra.

This may well be the case, but the results of a recent study comparing efavirenz (Sustiva) with soft-gel Kaletra alongside two nucleosides as a first-line regimen found that efavirenz may be more durable over two years.

Nevertheless, when patients and clinicians choose between fosamprenavir and lopinavir, it may come down to the drugs' relative costs, and the fact that the new tablet formulation of Kaletra does not need to be refrigerated, unlike the ritonavir capsule that must be taken to boost fosamprenavir.

Will the end result be a marketing and pricing war between Abbott (who market both Kaletra and ritonavir) and GSK?

Adverse events

The other primary outcome of the study was the proportion of participants who permanently discontinued treatment due to adverse events.

In total, 53/436 (12%) on fosamprenavir discontinued the study early due to adverse events, compared with 43/443 (10%) on lopinavir. Diarrhoea was the most commonly reported side-effect, with 55 (13%) of those on fosamprenavir and 50 (11%) of those on lopinavir reporting this as a grade 2 - 4 (moderate-to-severe) event. However, only five (1%) participants in the fosamprenavir arm and seven (2%) in the lopinavir arm discontinued the study due to diarrhoea.

Another one percent in each arm discontinued due to suspected abacavir hypersensitivity reactions. Although the investigators fail to disclose the other adverse events that led to study discontinuation, they do report the following moderate-to-severe clinical adverse events:

• nausea (6% fosamprenavir/5% lopinavir)
• headache (3% fosamprenavir/1% lopinavir)
• fatigue (2% fosamprenavir/1% lopinavir)
• vomiting (2% fosamprenavir/2% lopinavir)
• rash (3% fosamprenavir/

A higher proportion of grade 3 - 4 laboratory abnormalities were reported in both arms, however, most notably fasting blood fat levels:

• cholesterol (11% fosamprenavir/9% lopinavir)
• triglycerides (8% fosamprenavir/8% lopinavir)

Both arms showed similar increases from baseline in fasting blood fat levels at week 48, and 11% of patients in both arms were prescribed lipid-lowering medications.

Individuals coinfected with hepatitis B, C or both, were more likely to experience liver enzyme increases. Approximately 11% of coinfected patients on fosamprenavir and 14% of coinfected patients on lopinavir experienced grade 3 - 4 liver enzyme increases.