HPV Genotype Data Improves Cervical Cancer Prediction Tool

The addition of viral genotype data boosts the ability of a screening tool to predict cervical cancer among women with high-risk HPV infections, a new study in JAMA Network Open reports. Researchers from Southern Medical University in Guangzhou, China, and elsewhere developed a stacking machine learning model to predict cervical cancer diagnoses that includes a combination of epidemiological factors, pelvic exam findings, and HPV genotypes and trained it on a set of 14,533 women, 2.4 percent of whom were diagnosed with cervical intraepithelial neoplasia grade 3 or worse (CIN3+) and 4.6 percent of whom were diagnosed with CIN2+. In a validation set of 7,167 women, they found the inclusion of viral genotypes increased the tool's accuracy and reported it had a sensitivity of 80.1 percent and a specificity of 83.4 percent for predicting CIN3+ and a sensitivity of 80.4 percent and a specificity of 81.0 percent for CIN2+. The researchers note that the World Health Organization recommends high-risk HPV DNA testing for screening for cervical cancer and say their tool could be a "triage tool" for after testing. "Including HPV genotypes in the model markedly improved the prediction ability, suggesting that this prediction model may be an important auxiliary tool in screening for and early diagnosis of cervical cancer in low-resource settings when cytological and colposcopic examination results are unavailable," they write.