abstract
- BACKGROUND: Persistent cervical infection with high-risk [HR] HPV is a causative factor for cancer. Liquid-based [L-Pap] Pap samples are convenient for HPV testing and SurePath samples have been least studied. Most HPV tests have multiple step protocols and testing laboratories experience large volumes of samples. OBJECTIVES: Using SurePath L-Pap residual samples the objectives were as follows: [1] to test the performance of AMP-HPV and LA-HPV. [2] To perform an agreement study between two laboratories for the AMP-HPV test and [3] to compare agreement of results between AMP-HPV and LA-HPV and HC2. STUDY DESIGN: Samples from 657 women were tested for Pap cytology then assayed for HR-HPV using AMP-HPV and LA-HPV tests. AMP-HPV performance was compared between 2 laboratories and agreement studies were conducted between AMP-HPV, LA-HPV and HC2. RESULTS: HR-HPV genotypes were associated with L-Pap readings as follows: HSIL 92% [23/25], LSIL 73.6% [162/220], ASCUS 70.4% [131/186], normal 31.9% [72/226]. More women less than 30 were infected with HR-HPV and multiple genotypes regardless of the L-Pap reading. AMP-HPV and LA-HPV testing had an overall raw agreement with each other of 84.2% [Kappa 0.66] and each had agreement of 94% with HC2 testing of 133 samples [Kappa 0.86/0.87]. AMP-HPV agreement between two laboratories was better at 93% [Kappa 0.84] compared to 76.1% [Kappa 0.40] when extraction was standardized. CONCLUSION: It is feasible to perform AMP-HPV and LA-HPV on SurePath samples to detect HR-HPV genotypes. HC2, AMP-HPV and LA-HPV showed strong agreement. The extraction component of the AMP-HPV assay needs careful attention to yield consistent results.