icon-    folder.gif   Conference Reports for NATAP  
 
  17th CROI
Conference on Retroviruses
and Opportunistic Infections
San Francisco CA
February 16-19, 2010
Back grey_arrow_rt.gif
 
 
 
Large-Scale Application of "Deep" Sequencing Using 454 Technology to HIV Tropism Screening
 
 
  Reported by Jules Levin
17th CROI 2010 Feb 16-19 SF
 
Luke C Swenson1, Winnie Dong1, Theresa Mo1, Alexander Thielen2, Mark Jensen3, Doug Chapman4, Ian James5, Jayvant Heera4, Hernan Valdez4, P Richard Harrigan1 1BC Centre for Excellence in HIV/AIDS, Vancouver, Canada; 2Max-Planck Institute for Informatics, Saarbrucken, Germany; 3Fortinbras Research, Buford, GA; 4Pfizer, Inc. New York, NY; 5Pfizer Research and Development, Sandwich, UK
 
AUTHOR CONCLUSION
 
"Deep" sequence analysis of the HIV V3-loop gives improved prediction of virological response to MVC relative to the original Trofile assay, and similar to the currently-available Trofile assay (ESTA).
 
BACKGROUND
 
HIV tropism testing is required prior to treatment with CCR5-antagonists such as maraviroc (MVC). Currently, the recombinant phenotypic Trofile assay (Monogram Biosciences) is the most commonly used method to test for HIV tropism/coreceptor-usage.
 
Genotypic testing using either population-based sequencing or novel "deep" 454 sequencing of the V3 loop of HIV env may be viable alternatives, with some practical advantages over Trofile.
 
"Deep" sequencing can quantify minority HIV variants within an individual, including non-R5 variants.
 
We retrospectively tested response to a MVC-containing regimen across 4 clinical trials, with HIV tropism determined by deep sequencing. Results were compared to those obtained using the original Trofile assay, population-based V3 sequencing and, where available, the Enhanced Sensitivity Trofile Assay (ESTA).
 
The 4 trials of MVC included 1 large study in treatment-naïve (TN) patients, MERIT, and 3 large studies in treatment-experienced (TE) patients: MOTIVATE-1, -2 and A4001029.
 
MERIT compared MVC (QD or BID) & AZT/3TC vs. efavirenz (EFV) & AZT/3TC.
 
The TE studies compared MVC plus an optimized background therapy (OBT; based on resistance and treatment history) vs. placebo plus OBT.
 
All patients entering the trials were originally screened with the original Trofile assay as having R5-only virus, except those entering A4001029, who were screened as having non-R5 HIV.
 
Screening samples from MERIT were subsequently re-assayed using ESTA.
 
METHODS
 
Screening plasma samples from the MVC arms of trials in treatment-naïve "TN" (MERIT) and treatment-experienced patients "TE" (MOTIVATE-1 & -2 and A4001029) were examined. The EFV arm was also examined for comparison.
 
The V3 loop of HIV env was amplified in triplicate with nested RT-PCR and combined in equal quantities before sequencing using a 454/Roche GS-FLX (N=1560), blinded to any clinical data.
 
Tropism was inferred from V3-loop sequences, with samples classified as non-R5 if ≥2% of the virus population scored ≤3.5 using the geno2pheno algorithm. The population-based sequencing cut-off was 5.75 using geno2pheno.
 
The 3 TE studies were combined and analyzed up to week 48, and MERIT up to week 96.
 
The MVC QD & BID arms were analysed together for TE, and separately for the TN studies.
 
The ability of both V3 genotyping methods to infer tropism and outcome on MVC was assessed using a variety of parameters including:
 
- change in plasma viral load (pVL) from baseline (LOCF) (Figs. 2B & 3A)
 
- proportion of patients with pVL<50 HIV-RNA-copies/mL (Figs. 2C & 3B) time to tropism change from R5 to DM/X4 (as defined by Trofile) (Figs. 2D & 3C)
 
RESULTS
 
"Deep" sequence analysis discriminated between responders and non-responders to MVC regardless of treatment experience across all 4 clinical trials.
 
Where assay results were discordant, virological outcomes were better predicted by deep V3 sequencing and ESTA compared to the original Trofile (Table 1). Discordance of deep sequencing with ESTA or population-based sequencing slightly favoured the 454 method (Fig. 1).
 

At week 24, more patients screened R5 by "deep" sequencing had pVL <50 c/ml vs. non-R5 patients (Figs. 2C & 3B).
 
- For TN patients: 72% (221/308) vs. 54% (19/35); for TE patients: 50% (335/675) vs. 27% (58/214).
 
- Using population-based sequencing, these were: 71% (229/322) vs. 55% (16/29) (TN), and 43% (241/557) vs. 32% (37/117) (TE).
 
- For TN patients, ESTA results were similar to 454 results: 72% (215/297) vs. 54% (25/46), for ESTA-R5 and -DM, respectively.
 
Similar results were obtained at week 48. The MVC BID and EFV arms had similar proportions of TN patients with a pVL <50 copies/mL at week 48 when re-screened as R5 by deep sequencing: 67% (205/308) and 69% (205/299).
 
Of MVC patients screened as R5 by the original Trofile, 10% and 25% changed tropism to DM during the TN (Fig 2D) and TE (Fig. 3C) studies. On-study tropism changes were more common if any method identified these samples as non-R5 (40% & 66% for "deep" sequencing; 31% & 45% for population V3; and 38% & N/A for ESTA in TN and TE studies respectively).
 
The MVC QD arm in MERIT was originally discontinued because it did not meet protocol-defined criteria for non-inferiority to EFV. Re-analysis by "deep" sequencing appears to give similar results for these two arms: 80% (115/143) of R5 MVC QD and 78% (234/399) of EFV patients had a pVL <400 c/ml at week 16. This held for the analysis censored to only include patients remaining on MVC QD (Fig. 2E), and was similar for the uncensored analysis that included patients who switched to MVC BID.
 

Figure 3: Virological Response of Treatment-Experienced Patients in MOTIVATE-1, -2, and A4001029