Quad fixed-dose integrase combination: phase 3 studies at week 48

Simon Collins, HIV i-Base

The fixed dose, single-pill, four-drug formulation of elvitegravir/cobicistat/tenofovir/FTC, developed under the name Quad, is likely to be closest to regulatory approval (and has an FDA hearing in May).

At CROI, results were presented from two randomised, double blind, placebo controlled Phase 3 studies. One study comparing Quad to efavirenz/tenofovir/FTC (Atripla) was an oral session and another comparing Quad to atazanavir/ritonavir plus tenofovir/FTC was a poster. [1, 2]

The primary endpoints in both studies were the proportion of patients with undetectable viral load (< 50 copies/mL) at week 48 by intention- to-treat analysis, with non-inferiority defined by a lower margin of -12% and that included patient stratification by baseline viral load above and below 100,000 copies/mL. Virological efficacy was around 90%, tolerability was good and discontinuations were notably low in all arms and Quad was found to be non-inferior compared to the comparator combinations in both studies.

Study 236-0102 compared Quad to Atripla enrolled 700 treatment-naive patients in the US and Puerto Rico and Paul Sax from Brigham and Women’s Hospital, Boston, presented the results. [1]

Baseline characteristics included: mean age 38 years and low median viral load (31,000 copies/mL) although one third of participants started at >100,000 c/mL. Mean CD4 count was just under 400 cells/mm3 with 12% of participants starting below 200, 32% starting at both 200 to 350 and 350-500 and 23% starting at >500 (percentages for Quad arm but similar to Atripla). The study was largely male (88%) with ethnicity 61% white, 31% African Americans and 8% other. Less than 5% of participants in each arm had either HBV or HCV coinfection.

Discontinuations before week 48 were similar in 11% vs 13% in the Quad vs Atripla arms for broadly similar reasons.

Viral load was suppressed to undetectable in 88% vs 84% patients (difference +3.6%, 95%CI -1.6 to +8.8) meeting criteria for non-inferiority, with 7% of patients in each arm having virological failure and 5% vs 9% having missing data (all Quad vs Atripla, respectively). Responses by subgroup (viral load, CD4, race, sex, age and adherence level were not significantly different but trended to favour Quad. CD4 increases favoured the Quad arm with +239 vs + 206 cells/mm3 respectively (p=0.009).

Approximately half of the patients in each arm failed with mutations associated with resistance to either integrase inhibitors (mainly E92Q) or NNRTIs (mainly K103N) in 8/14 vs 8/17 respectively.

Most side effects were reported as mild (grade 1) with statistically significant differences including more nausea in the Quad arm (21% vs 14%) and more abnormal dreams (15% vs 27%), insomnia (9% vs 14%), dizziness (7 vs 14%) and rash (6% vs 12 %) in the Atripla arm.

Discontinuations related to side effects occurred due to rash (0 vs 1.4%), renal abnormalities (1.4% vs 0), depression (0.3% vs 0.9%), abnormal dream (0 vs 0.6%) in the Quad vs Atripla arms respectively with 3% in each arm stopping due to both fatigue and paranoia.

The most frequent grade 3 or 4 laboratory abnormalities occurring in greater than five patients in each arm were broadly similar and generally low including creatinine kinase (5% vs 11%), AST (2% vs 3%), ALT (1%vs 3%), GGT (2% vs 5%), neutrophils (2% vs 3%), amylase (2% each arm) and haematuria (2% vs 1%), all in Quad vs Atripla respectively.

Serum creatinine increased by approximately 0.1-0.2 mg/dL by week 2 in the Quad arm which was maintained through to week 48 compared to no change with Atripla (p<0.001).

Increases in fasting total cholesterol, LDL and HDL cholesterol were significantly greater in the Atripla compared to the Quad arms but there was no difference between groups in the more clinically significant TC:HDL ratio or in triglycerides (+ 7 mg/dL in each arm).

The second Quad study, called 236-0103, compared Quad to atazanavir/ritonavir plus tenofovir/FTC (Truvada). It enrolled 708 treatment-naive patients and results were presented by Edwin DeJesus in a poster. [2] Baseline characteristics were broadly similar to the 236-0102 study: mean age 38 years, 90% male, and 74% white. CD4, viral load and hepatitis coinfection were also similar, with 40% of participants having viral load ≥100,000 copies/mL. Exclusion criteria for this study included eGFR < 70 mL/min.

Virological efficacy (<50 copies/mL) at week 48 was 92% vs 88% (difference +3.5%, 95%CI –1.0% to +8.0%) in favour of Quad, which met criteria for non-inferiority. In patients with baseline viral load ≥100,000 copies/mL, response rates were 85% vs 82% (NS). Virologic failure (FDA snapshot algorithm) was 5%, in both arms. Median CD4 increases were similar at + 207 vs 211 cells/mm3 and discontinuation rates for side effects were 4% vs 5% (both in Quad vs atazanavir/r arms, respectively).

Side effects occurring in ≥5% of patients, were similar in each arm, apart from elevated bilirubin levels which were significantly higher in the atazanavir/ritonavir arm. Discontinuations occurred due to diarrhoea (4% vs 5%), pyrexia (1% vs <1%), nausea (1% vs 0), nausea, vomiting and fatigue (each <1% vs 1%) and jaundice, dizziness, ocular icterus and drug eruption (each 0 vs <1%). The most frequent grade 3 or 4 laboratory abnormalities occurring in at least 2% in either arm were broadly similar including creatinine kinase (6% vs 7%), haematuria (4% vs 2%), AST (2% vs 3%), ALT (2% vs 2%), amylase (2% each arm) and increased bilirubin (1% vs 58%), all in Quad vs atazanavir/ritonavir arms respectively. Serum creatinine increased by approximately 0.08 mg/dL by week 2 in the Quad arm which was 0.12 mg/dL at week 48 compared to 0.05 with atazanavir/ritonavir (p<0.001). Median change in CLCr from baseline was –12.7 mL/min in Quad and –9.5 mL/min (p <0.001). Lipid increases were similar for TC, LDL and HDL cholesterol (all p=NS) but triglycerides increased by less in the Quad arm (+5 vs +23 mg/dL, p=0.006).

Median changes in bone mineral density were similar in each group. Spine changes reduced by about 3% at week 24 and remained stable, with reductions at week 48 of -2.45% vs -3.48% (p=0.25 for between arm comparison). Reductions at the hip were continuous slopes for both combinations of about -1.5 vs -2.0% at week 24 and -2.87 vs 3.59% at week 48 (p = 0.12).

In an answer to a question about clinical management of increases in serum creatine, in reference to increases seen in both Quad studies, Paul Sax stated that a statistical analysis of the data suggested that an increase of 0.4 mg/dL or greater was a cut-off for concerns about potential tenofovir renal tubular toxicity.

When asked whether paired cystatin C in serum would distinguish between tenofovir and cobicistat associated changes Sax said that this was a possibility, but that it might be affected by other HIV-related factors. The question of cost was also raised in the context of results that had not demonstrated superiority, perhaps related to study size.

comment

Quad has already been submitted for regulatory approval to Western regulatory agencies with an FDA decision expected by August 2012.

References

Sax P et al. Elvitegravir/cobicistat/emtricitabine/tenofovir (Quad) has non-inferior efficacy and favorable safety compared to efavirenz/emtricitabine/tenofovir in treatment-naïve HIV-1+ subjects. 19th CROI, 5–8 March 2012, Seattle. Oral abstract 101.
http://www.retroconference.org/2012b/Abstracts/43143.htm
Webcast: State of the “ART” and Drug Resistance (Tuesday 10 am).
http://retroconference.org/static/webcasts/2012/
DeJesus et al. Week 48 results of an ongoing global phase 3 study comparing elvitegravir/cobicistat/emtricitabine/tenofovir (Quad) with atazanavir/ritonavir plus emtricitabine/tenofovir in treatment-naïve HIV-1+ subjects showing efficacy, safety, and pharmacokinetics. 19th CROI, 5–8 March 2012, Seattle. Poster abstract 627.
http://www.retroconference.org/2012b/Abstracts/43391.htm