In the Cancer Prevention Regional Laboratory of ISPO (Florence,Italy), after the transition from Hybrid Capture® 2 test (HC2®, Qiagen®), that detects 13 HPV types (12 HR-HPV+HPV68), to Cobas® 4800 HPV test (Roche®), that detects 14 HPV types (12 HR-HPV+HPV66 and 68), a decrease in HR-HPV positivity was observed in all the Local Sanitary Areas on which the HR-HPV test was already performed by our laboratory. The aim of this study is to compare the performance of the two methods using the same set of samples.
620 routine screening samples,HC2® positive, were retested on Cobas®. The samples that resulted negative to Cobas® (discordant) were typed by a Reverse Line Blot (RLB) method (Ampliquality HPV-Type express 3.0®-AB Analitica®), that detects the presence of 40 HR and non HR-HPV types. These results were linked with cytological and histological data.
419 samples (67.6%) was confirmed HR-HPV positive to re-test with Cobas®,while 201 samples (32.4%) resulted HR-HPV negative (discordant). This decrease of the positivity to HR-HPV is also confirmed in the screening samples analyzed after the introduction of Cobas®.
The discordant samples reported the following results of cytology triage: 165(82.1%) resulted with normal cytology and 36(17.9%) with abnormal cytology (31/36 LSIL (86.1%), 3/36 ASC-H (8.3%), 1/36 HSIL (2.8%), 1/36 AGC (2.8%)).
All discordant samples were typed using RLB. 14/201 samples(7%) resulted HR-HPV positive by typing and only one of these (HPV58+) had an abnormal cytology (LSIL), but resulted negative to colposcopic exam. The typing results of the other 187 discordant samples were: 88/187(47%) HPV negative, of which 5 with abnormal cytology triage (2 ASC-H, 1 AGC, 2 LSIL) and none with CIN2+ lesion; 99/187(53%) were non HR-HPV positive, of which 30(30.3%) with abnormal cytology triage (1 ASC-H, 1 HSIL and 28 LSIL) and 2 with CIN3 lesions (HPV26 and HPV54+73+90 respectively).
Furthermore, a considerable number of discordant samples (20/201, 10%) resulted HPV68a positive by RLB (5 with abnormal cytology and none CIN2+).
Cobas® is a reliable method and is more specific than HC2® (92.5% of discordant samples are HR-HPV negative, so we would have registered a lower rate of false HR-HPV positive samples). Between Cobas® HPV negative samples, but HC2® HR positive, we found two CIN3 lesions resulted associated with LR-HPV types by PCR RLB. HPV 68 is a target type for Cobas® but it is not clear if 68a and 68b are detected at the same level, because, in our set of samples, several HPV 68a resulted negative by Cobas®.