NEW YORK (GenomeWeb) – Researchers from the University of California, San Francisco involved in ongoing longitudinal research on the human papillomavirus have published a study describing their development of a Luminex-based HPV genotyping assay to replace their previous testing using Roche's Linear Array.
The study, published online as an accepted manuscript earlier this month in the Journal of Clinical Microbiology, describes the team's creation of a Luminex assay they call PGMY-LX using the same PGMY09/11 primer/probe system and offering comparable sensitivity and specificity to the Roche Linear Array test that they had been using previously.
According to the authors, the switch to Luminex has allowed them to test for HPV using the same primers, maintaining critical continuity with older samples in their long-term research, but with added benefits in terms of cost and throughput associated with Luminex technology, which allows multiplexed testing of up to 100 unique assays within a single, small-volume sample.
Sepideh Farhat-Nozzari, the study's first author, told GenomeWeb this week that the group hopes sharing their approach may inform other groups involved in lengthy longitudinal studies using Linear Array who also might want to switch to a higher-throughput and cost-effective system.
While other groups have also adopted Luminex technology for HPV testing, the UCSF group wanted to develop a test specifically using the same primer/probe system as the Roche Linear Array HPV assay, in order to maintain the continuity of their longitudinal results.
The team has been involved in a study using PGMY09/11-based LA to prospectively characterize the natural history of cervical HPV infection in adolescents and young women in a cohort setting since 2000.
The authors wrote that their study is the first they know to directly compare a Luminex-based HPV assay with the Roche PGMY09/11-based LA, and to demonstrate their close concordance.
In the study, Nozzari and her colleagues reported on their development of PGMY-LX and how the method performed in comparison with Roche's Linear Array.
According to the authors, PGMY-LX, like LA, allows type-specific detection of 33 oncogenic and non-oncogenic HPV types. The group took as an example previous groups' successful development of Luminex HPV assays, but adapted the technique to the specific primers and probes of the Roche LA test.
Detection of HPV type-specific plasmid DNA was highly specific using PGMY-LX, the group reported, with high signal-to-noise ratios for all types except non-oncogenic type 40 and no cross-reactivity between types.
Overall, the team compared LA and PGMY-LX results for 378 cervicovaginal lavage samples from 251 women in their longitudinal study.
They did not examine agreement between the two tests for HPV types 40 and 52 due to inability to establish a satisfactory PGMY-LX cutoff value for type 40 and to the absence of a specific probe for type 52 in the Roche LA assay.
According to the authors, the samples tested by both the Linear Array and PGMY-LX showed "almost perfect type-specific agreement" for seven HR HPV types (16, 31, 33, 39, 45, 58, 59), six possible HR (53, 66, 67,70, 73, 82) and 11 LR HPV types (6, 11, 42, 54, 55, 62, 72, 81, 83, 84, and CP6108).
The two tests also showed substantial agreement for LR type 61 and the remaining HR 18, 35, 51, 56, and probable HR, 68b. The sole type for which the two methods did not substantially agree was HR type 26, despite the fact that PGMY-LX appeared to perform well for this type with plasmid samples, displaying low background and a high signal-to-noise ratio on par with other types.
To examine the sensitivity of PGMY-LX, the team also performed a dilution experiment using a high copy-number HPV16 plasmid.
A five-fold dilution series of the HPV16 plasmid was prepared from 10,000 copies/µl to 1 copy/µl, and the group found that dilutions with 80 copies/µl or higher tested positive using PGMY-LX above the established MFI cutoff.
Nozzari explained that the Luminex approach has a slightly higher sensitivity than LA, but because the group wanted to replicate LA exactly, they adjusted their MFI cutoff to bring the threshold higher so that they would see positivity at the same rate as the Roche test.
The team then examined agreement between the two methods in samples with zero, one, two, or multiple HPV types.
Agreement between LA and PGMY-LX was highest when only a single HPV type was present by LA, decreasing somewhat in cases with two types present, and even further in cases with multiple types present. This was not surprising, the authors wrote, as they have observed that reproducibility diminishes as the number of HPV types goes up, even when a sample is retested by the same method.
The team also investigated the inter-run reproducibility of PGMY-LX, showing an inter-run reproducibility rate of 92%, 98%, and 85% for samples with zero, one, or two or more types, respectively.
According to Nozzari, the group hoped, in publishing their results, to potentially provide a roadmap for other studies that rely on Linear Array but might need or want to adopt an equivalent alternative.
"For cohort studies that have been using Roche Linear Array, this is a very good alternative if they have to switch over [and] it's half the price per person," she said "[PGMY-LX] allows for testing 34 HPV types. Different groups, depending on their area of interest, could use any probe they want to target more or different types," she added.
The study that Nozarri and her team initially developed PGMY-LX for has closed, and she is now joining a different UCSF lab focused on HPV testing in the HIV-positive gay community.
She said that she is in discussion with this project's PI about whether their project might also benefit from a Luminex-based testing method.