HPV Primary Screening Pilot Study: molecular testing of potential triage strategies for HPV-positive women.
Citation:REYNOLDS, STEPHEN HUGH, HPV Primary Screening Pilot Study: molecular testing of potential triage strategies for HPV-positive women., Trinity College Dublin.School of Medicine, 2020
Stephen Reynolds PhD Thesis MASTER FILE.pdf (PDF) 5.169Mb
Cytological based screening programmes are being phased out in favour of HPV-based screening both in Ireland and internationally. Infection with HPV is the primary aetiological agent in the development of cervical cancer. Screening for the virus through HPV testing is more sensitive than our current cytological approach. However, due to high rate of transient HPV infection, HPV DNA testing for cervical cancer or precancer suffers from a lower specificity. Several possibilities exist to improve specificity including the use of more specific tests such as HPV E6/E7 mRNA detection or the use of triage tests such as methylation testing of specific markers for HPV positive women. Evidence has showed that methylation testing may be a superior triage test compared to cytology triage. The project aimed to assess the prevalence of HPV prevalence in the Irish population using the Cobas 4800 HPV DNA test and the Aptima HPV E6/E7 mRNA test on a population of over 12,000 women attending for their routine cervical smear test and compare both platforms for their overall performance in the context of screening this population. The population prevalence of HPV DNA was 15.98% compared to 13.22% for HPV mRNA. Both platforms had comparable sensitivities for CIN 2+ (96.26% and 96.55%), as well as a reasonable level of agreement between each other (kappa: 0.701). HPV mRNA however, had a higher specificity for detection of CIN 2+ (76.42%) in women <30 compared to HPV DNA testing (70.82%), due in part to the lower number of cytology normal women with a positive HPV mRNA test. Overall, both platforms were comparable and viable for a HPV primary screening population however, both would require adequate triage testing of HPV positive women. Cytology will likely be the initial triage test for HPV positive women. However, cytology will still suffer from a low specificity for detection of CIN 2+ as well as variability. An alternative approach to triaging women using methylation markers was assessed in this study. The aim was to validate a three-marker methylation panel (CAD M1-M18, MAL M1 and hsa-mir-124-2), using standard qMSP techniques and determine a clinically relevant threshold for positive methylation for detection of CIN 2+. These individual methylation markers were investigated in several different and novel combinations. Two such combinations were termed the Total Methylation Score and the Paired Methylation Score. Cut-off points were determined for each individual methylation marker, as well as for the Total and Paired Methylation Score by ROC curve analysis. From this, the Total Methylation Score had a sensitivity and specificity of 78.57% and 80.00% for detection of CIN 2+. The Paired Methylation Score performed well with a sensitivity and specificity of 71.43% and 76.00%. Of the three individual methylation markers, hsa-mir-124-2 had the highest individual performance with a sensitivity and specificity of 66.67% and 78.00% respectively. These three methylation approaches were brought forward to be assessed in a HPV positive triage population. When applied to the HPV positive population, the Paired Methylation Score had the best performance with a sensitivity of 59.26% and specificity of 66.80%. The combination of HPV 16/18 genotyping with methylation testing of the other hrHPV types increased the sensitivity to 80.56% for both the Total and Paired Methylation Scores. Referral rates for each of the three methylation approaches were comparable to the referral rates of cytology (37.76% to 41.59% vs 38.27% respectively). The use of HPV 16/18 genotyping with methylation or cytology showed comparable referral rates also. Only the Paired Methylation Score had sufficient risk following a positive test to refer women to colposcopy in this study (>20%). However, in women <30 years the PPV increased sufficiently for all three approaches to safely refer women on to colposcopy. None of the methylation approaches with or without HPV genotyping were able to safely refer women back to the routine recall population. However, the Paired Methylation Score, in women >45, had a reduced negative triage test risk of 2.9%. Though no triage approach tested here reached the <2% cut off for the safe return to routine screening, the majority of other triage tests including cytology triage fail to do so also. In conclusion, HPV DNA and mRNA have been shown to have both a comparable sensitivity, specificity and PPV with an NPV of >99%. However, in women <30 years, HPV mRNA testing is more specific than HPV DNA testing and overall refer fewer women for triage. Methylation triage testing has been shown to be both sensitive and specific for detection of CIN 2+ in a HPV positive population with comparable referral rates to cytology. The combination of HPV 16/18 genotyping with methylation testing greatly increases the sensitivity for detection of CIN 2+. In women <30 methylation testing has been shown to be more sensitive, while in women >45 a negative methylation test imparts a higher NPV which approaches the threshold for returning women safely to routine screening. The use of methylation testing shows great potential in the triage of HPV positive women.
Health Research Board (HRB)
Author: REYNOLDS, STEPHEN HUGH
Publisher:Trinity College Dublin. School of Medicine. Discipline of Histopathology
Type of material:Thesis
Availability:Full text available