Clinical performance of the HPV DNA Array genotyping assay in detection of CIN2+ lesions with BS GP5+/6+ MPG Luminex tested cervical samples

Human papillomavirus (HPV) detection is used for screening of cervical cancer and genotype‐specific persistence has shown to be mandatory for dysplasia development. Aim of this study was to evaluate the clinical performance of HPV DNA Array for cervical intraepithelial neoplasia 2+ (CIN2+)  lesion detection. HPV DNA Array is a polymerase chain reaction‐based assay that targets E1 sequences of 29 HPV types (6, 11, 16, 18, 26, 31, 33, 35, 39, 40, 42, 44, 45, 51, 52, 53, 54, 56, 58, 59, 66, 67, 68, 69, 70, 73, 82, 85, and 97). The clinical evaluation was performed against the reference assay, BS‐GP5+/6+ multiplex genotyping (MPG)‐Luminex, with 600 cervical smear samples of a referral population. HPV DNA Array detected CIN2+  lesions with a sensitivity of 90.2%, identical to that of MPG‐Luminex. Detection of CIN3+  lesions was with a sensitivity of 90.3%, as compared with 88.7% of MPG‐Luminex. It demonstrated very good agreement for HPV detection, irrespective of type, of 91.5% (κ  =  0.832). HPV DNA Array is a simple and robust assay, with a short protocol of 4 hours hands‐on time and automated readout by ELISpot AiDot software. It permits testing of up to 96 samples in one run and may be considered for use in organized screening programs and low resource settings.


| INTRODUCTION
The main cause of cervical cancer is persistent human papillomavirus (HPV) infection. 1 In the majority of women, HPV infections will clear within 2 years. 2 Even if a low-grade intraepithelial neoplastic lesion develops, in most women, it will regress within 3 years. 3 However, if genotype-specific HPV infections and lesions persist, women are at higher risk of developing cervical cancer. 4 The most clinically significant types, labeled high-risk (HR) types are HPV- 16,18,31,33,35,39,45 and -18, for primary cervical cancer screening. 7 A shift in the paradigm from cytology to HPV detection in primary cervical cancer screening is evident. 8 The high number of HPV assays available in the market, 9 challenges the health care professionals to determine which assays are most efficient for the detection of high-grade lesions. To evaluate the performance of any HPV test, a comparison against a wellvalidated reference HPV assay is warranted. An assay validation guideline has been established by Meijer et al. 10 In this study, the performance of HPV DNA Array was validated against BS GP5+/ 6+polymerase chain reaction (PCR) followed by Luminex-based hybridization assay termed multiplex genotyping (MPG), an internationally recognized and clinically validated HPV test. 11

| HPV DNA Array
The assay is capable of genotyping 18 HR (16, 18, 11 An internal control, the cellular β-globin, is included to follow if sufficient DNA amount of each sample is present. Testing was performed as described. 13,14 However, in our laboratory the final PCR volume was adjusted to 25 µL vs 50 µL used in the publications.
In our laboratory, MPG assay performance was validated by participation in EQUALIS proficiency panel testing. 15

| Ethics statement
Patients consented to use residual diagnostic material for research (IRB Charite-Universitätsmedizin Berlin, no. EA1/168/13).  Finally, after retesting by both assays, in 69 samples a discordance was concluded. We present the analysis on the reevaluated results.
Agreement between assays was 91.5% with κ 0.832 (95% CI, 78.7%-87.6%) showing very good agreement ( In most cases, we observed that when results changed it was either a single infection that was lost, or a negative sample becoming a single infection, or losing an HPV type in a sample with multiple infections. We could theorize that the HPV types initially missed were present in low copy numbers, hence missed during pipetting for the first time, but not the second time or vice versa. Or the sequences of the missed HPV types could be more difficult to amplify within PCR due to competition with other HPV types. Or there was an initial operational mistake while pipetting.
It is noteworthy to mention that the MPG assay seems to be more sensitive and therefore, prone to detecting low copy numbers that may fluctuate around the detection limit.
This retesting was performed and included in the analysis with the aim to have the most accurate HPV results, especially for validation purposes. However, we are aware such retesting would not be feasible as part of a real-life screening program.
This might be of importance for epidemiological studies, but not relevant for clinical routine as histologically important lesions were detected.
It seems that HPV genotyping assays demonstrate a lower agreement for HPV type-specific detection, 24 they, however, show a very good agreement for detection of CIN. This could be explained by the higher number of viral copies in such lesions. 10

| CONCLUSION
HPV DNA Array demonstrated a very good clinical performance for CIN2+/CIN3+ lesion detection and a very good agreement to the MPG test. HPV DNA Array is a full genotyping assay and may be competitive to other full genotyping assays due to high throughput and ease of handling, what may allow its use in organized screening programs and low resource settings as well.