Human papillomavirus load in Hybrid Capture II assay: does increasing the cutoff improve the test?

Abstract

Objective: The aim of the study was to evaluate whether the results of the Hybrid Capture II (HCII) assay for detecting high-grade squamous intraepithelial lesions (H-SIL) or cervical carcinoma can be improved by increasing the relative light units (RLU) level.

Study design: We included 2271 women (mean age 38.7 +/- 12.3, range 15-92) referred to a colposcopic clinic due to cytology of atypical cells of unknown significance, SIL or carcinoma. All women underwent a new Pap test, HR-HPV detection using HCII and colposcopy with biopsy of suspicious areas when present.

Results: HR-HPV was detected in 91.7% of carcinomas, 96.6% of H-SIL, 85.1% of low-grade SIL and 21.6% of cases with no lesion. The probability of harboring an H-SIL or a carcinoma significantly increased as RLU increased (P = 0.01). The sensitivity and specificity for H-SIL or carcinoma at different cutoffs were 95.7 and 54.6 at 1 RLU, 93.9 and 59.6 at 2 RLU, 90.1 and 65.1 at 5 RLU and 85.7 and 68.7 at 10 RLU. The percentage of cases not detected with HCII increased from 2.4% for cases with <1 RLU to 9.5% for cases between 1 and 2 RLU, 14.8% between 2 and 3 RLU, 21.7% between 3 and 5 RLU and 28.4% between 5 and 10 RLU.

Conclusion: The use of a higher cutoff (higher viral load) in the HCII should not be recommended because it significantly increases the number of cases with H-SIL or carcinoma not detected, reducing the sensitivity and negative predictive value of the test.