[41685508]
We are reporting the suboptimal reproducibility of roche cobas hpv assay and the risk of false-negative results. We have recently noticed the suboptimal reproducibility of the roche cobas hpv assay. We have informed roche, provided our data to them and requested their attention to this matter. The company has taken our concerns seriously, escalated it to their scientific team and provided assistance in analyzing our data. We feel that we should also bring this issue to your attention since this test is fda cleared. Below is a short summary of our finding related to the suboptimal reproducibility of this assay. The roche official response to our complaint (complaint# (b)(4) filed on 01/22/2016) is for your review. The fda cleared, cobas hpv assay (roche diagnostics, (b)(4)) has the ability to detect 14 high risk (hr) hpv and to distinguish hpv 16 and 18 from the other 12 high risk hpvs. It is a qualitative, multiplex, real time (rt) pcr. The cobas hpv results are reported as positive or negative for hpv 16, 18 or for the other 12 high risk hpvs. Samples that have a crossing point (cp) value less than the manufacturer's assigned cut off value, are reported positive and if they have no cp or cp greater than cut off value, will be reported negative. Upon retesting of clinical samples, we discovered suboptimal reproducibility of cobas hpv assay. Our original data revealed that up to 4% of results may change upon retesting. These discrepancies occur in samples with cp values near the cut off value. Although such small changes in cp value are usually inconsequential, we have determined that minor fluctuation in cp values upon retesting of the same specimen by cobas hpv assay may change test result from negative to positive and vice versa. With the new proposed screening algorithm that uses hpv with genotype as the first-line primary screening tool for cervical cancer, false negative hpv results can have significant impact on patient care. Following the discovery of suboptimal reproducibility of roche hpv assay, and to avoid false negative results, we started to test all hpv samples in duplicate. We have observed around (b)(4) overall discrepancy between results upon testing of (b)(4) clinical samples in duplicate during 01/27/2016 to 03/21/2016. All discrepant results belong to the samples with cp values close to the cut off of the assay. We discussed the issue with roche and requested access to cp value or the samples that are reported negative (cp value of samples reported as negative by cobas is not visible to the laboratory personnel). Our goal was to limit repeat testing to negative samples that have cp greater than cut off. Our request was temporarily approved by roche, for research and investigational purposes, but they feel that since it is not fda approved, it constitutes an off label use.
Patient Sequence No: 1, Text Type: D, B5