HTB

Accuracy of Monogram’s Trofile assay in determining tropism and new tropism assays in development

Mark Mascolini, for NATAP.org

Monogram’s Trofile test dominates the coreceptor tropism assay field after ongoing screening and monitoring use in numerous clinical trials. Eoin Coakley from Monogram outlined reasons for confidence in Trofile in a review of several studies [1].

Using a panel of three clonal R5, X4, or dual/mixed-tropic (D/M) viruses, Coakley documented concordant results in more than 1000 pairwise comparisons. Testing Trofile in 46 R5 or D/M patient samples, he found discordant results in only 3 samples (6.5%). Monogram determined that the assay successfully amplifies viral samples from people with low viral loads, including 45 of 48 samples (94%) from people with 500 to 1000 copies/mL and 104 of 109 samples (95%) from others with 1000 to 10,000 copies/mL.

Working with R5- and X4-tropic patient-derived gp160 clones, Monogram scientists created a 10%/90% X4/R5 mixture and a 5%/95% X4/R5 mixture. Trofile correctly read the 10% mixture in all of 80 assays, while calling the 5% mixture correctly in 83% of assays. The Monogram assay has successfully tested about 95% of more than 16,000 samples submitted from clinical trials. Coakley could not explain the 5% failure rate because Monogram does not have viral load readings on all patients sampled. Some people did have loads below 500 copies, which could explain Trofile’s failure to read those samples.

Screening viral samples for trials of the CCR5 antagonist maraviroc, Trofile confirmed a substantially higher prevalence of R5-only virus in 1428 antiretroviral naive people (85%) than in 2560 with antiretroviral experience (56%). Coreceptor expert Rolf Kaiser (University of Cologne) suggested a growing consensus that R5 virus dominates in early infection because the healthier immune system at that point copes well with X4-favoring HIV. As the immune system falters, it loses control of X4 virus, which may then expand to represent a growing proportion of a person’s viral population.

Trofile is hardly the only tropism assay clamoring for clinical trial attention. Virco in Belgium has developed a tool for quantitative (genotyping and phenotyping) and qualitative (phenotyping) tropism testing [2]. BioalliancePharma of Paris showed that its Phenoscript Env assay works well in populations with different HIV-1 subtypes and relatively low viral loads [3]. InPheno scientists in Basel offered findings on two tests, a hybridization-based method called XtrackC and a replicative phenotyping test dubbed PhenoTecT (Phen-XR) [4]. Both tests detected minority population of both X4 and R5 virus.

None of these assay makers divulged projected costs of their tests. But none will be cheap.

As reported separately from the European Resistance Workshop, a few online genotype-based algorithms have 80% accuracy in calling coreceptor tropism correctly (see “Several Genotyping Systems Offer Hope as Tropism Predictors”).

References:

  1. Petropoulos C, Limoli K, Whitcomb J, et al. Validation studies defining the performance of Monogram’s HIV coreceptor tropism assay, Trofile. 5th European HIV Drug Resistance Workshop. March 28-30, 2007. Cascais, Portugal. Abstract 57.
  2. Rondelez E, Van Eygen V, Van Baelen K, et al. New quantitative and qualitative platform to test co-receptor usage of HIV-1 strains. 5th European HIV Drug Resistance Workshop. March 28-30, 2007. Cascais, Portugal. Abstract 58.
  3. Roulet V, Rochas S, Labernardiere J, et al. HIV-1 Phenoscript Env: a sensitive assays for the detection of HIV minority species and determination of B and non-B subtypes viral tropism. 5th European HIV Drug Resistance Workshop. March 28-30, 2007. Cascais, Portugal. Abstract 59.
  4. Hamy F, Vidal V, Hubert S, et al. Hybridization-based assay and replicative phenotyping as diagnostic platform for determination of co-receptor tropism. 5th European HIV Drug Resistance Workshop. March 28-30, 2007. Cascais, Portugal. Abstract 60.

Links to other websites are current at date of posting but not maintained.