Randomized controlled trial evaluating the utility of urine HPV DNA for cervical cancer screening in a Pacific Island...

Non-invasive, self-collection sampling methods for human papillomavirus (HPV) DNA detection have the potential to address logistical and cultural barriers to Pap screening, particularly in under resourced settings such as Yap state in the Federated States of Micronesia - a population with low levels of screening and high incidence of cervical cancer.

A randomized controlled trial was conducted among adult women in Yap to compare cervical HPV DNA in self-collected urine and clinician-collected liquid cytology. Adult women aged 21-65 (n=217) were randomized by the order of sample collection. Concordance of HPV DNA, evaluated by the Roche Linear Array, was compared in paired self-collected urine and clinician-collected liquid cytology samples. The sensitivity and specificity of urine HPV DNA for prediction of cervical HPV and abnormal cytology was also evaluated. p16 in urine cytology samples was additionally assessed.

Overall, HPV DNA detection was significantly lower in urine than cervical samples for any HPV (27.8% and 38.3%, respectively) and high-risk HPV (15.1% and 23.8%, respectively). For paired samples, there was moderate agreement for the overall study population (Kappa=0.54, 95% confidence interval CI=0.40-0.68) and substantial agreement for women >40 years (Kappa=0.65, 95% CI=0.46-0.85). The sensitivity and specificity of urine for the detection of cervical high-risk HPV was 51.0% and 96.2%, respectively. The sensitivities of HPV DNA in urine and liquid cytology for prediction of abnormal cytology (ASCUS/LSIL/HSIL) were 47.4% (95% CI=31.0-64.2) and 57.9% (95% CI=40.8-73.7), respectively; specificities were 92.0% (95% CI=86.9%-95.5%) and 83.5% (95% CI=77.2-88.7). Urine p16 was poorly correlated with urine HPV DNA positivity.

Urine is less sensitive but more specific than directed cervical sampling for detection of cytologic abnormalities and may have utility for screening in older populations within low-resource communities when clinically-collected samples cannot be obtained.