Clinical evaluation of p16 INK4a immunocytology in cervical cancer screening: A population-based cross-sectional study from rural China

BACKGROUND Cervical cancer screening with cytology suffers from low sensitivity, whereas the efficiency of human papillomavirus (HPV)-based screening is limited by low specificity. The authors evaluated a novel p16INK4a immunocytology approach in cervical cancer screening compared with HPV-based and cytology-based screening. METHODS In total, 2112 women aged 49 to 69 years from Shanxi, China were screened from March to July 2019. HPV testing, liquid-based cytology (LBC), and p16INK4a immunocytology were performed on samples from all women. Any positive result triggered a referral to colposcopy with biopsy, if indicated. Screening performance for detecting cervical intraepithelial neoplasia grade 2 and 3 or worse (CIN2+/CIN3+) was evaluated using multiple algorithms. RESULTS p16INK4a had a lower positive rate (10.0%) than LBC abnormality (vs 12.1%; P = .004) and a high-risk HPV positivity (21.4%; P < .001). For the detection of CIN3+, the relative sensitivity of p16INK4a compared with HPV and LBC was 0.93 (95% CI, 0.82-1.07) and 1.12 (95% CI, 0.95-1.32), respectively. The specificity of p16INK4a was significantly higher than that for HPV and LBC, with a relative specificity of 1.13 (95% CI, 1.11-1.16) and 1.02 (95% CI, 1.01-1.04), respectively. In addition, p16INK4a alone yielded a clinical performance very similar to that of the current mainstream strategy of using HPV16/18 with reflex cytology (ASC-US+, atypical squamous cells of undetermined significance or worse). The immediate risk of CIN3+ was 14.6% if p16INK4a results were positive and 0.2% if p16INK4a results were negative. CONCLUSIONS With minimal colposcopy referrals, p16INK4a screening demonstrated promising utility for risk stratification and yielded a better balance between sensitivity and specificity compared with HPV and LBC primary screening. Moreover, with accuracy and efficiency similar to what is achieved using mainstream cotest algorithms, p16 may simplify the screening practice. More evidence will be required before clinical recommendation.