Jean‐Luc Prétet

Reproducibility of HPV Testing With Two Versions of the Papilloplex High‐Risk HPV Assay

ABSTRACT Validation of HPV tests usable in cervical cancer screening require demonstration of noninferior clinical accuracy and sufficient reproducibility. The Papilloplex HR‐HPV assay [version1] (Genefirst, UK) has been validated only for clinical accuracy. Here, we assessed its reproducibility, aiming to complete the validation process. A panel of 550 cryopreserved cervical cell samples collected from women attending the cervical cancer screening program in Belgium was used to assess the intra‐ and inter‐laboratory reproducibility of Papilloplex HR‐HPV [version1], a full genotyping assay that identifies separately 14 high‐risk HPV (hrHPV) types using multiple probe amplification technology. We assessed whether the reproducibility fulfils validation criteria (lower 95% confidence interval [CI] bound ≥ 87% and κ ≥ 0.50). Subsequently, we compared the concordance between version1 and a new version6 of the assay and between three analysis methods for PCR curve interpretation. Papilloplex HR‐HPV version1 assay showed an excellent reproducibility for hrHPV (97.5% [CI: 95.8%–98.7%], κ = 0.94 for intra‐ and 93.5% [CI: 91.0%–95.4%], κ = 0.85 for inter‐laboratory reproducibility). Concordance analyses exhibited an excellent agreement between two assay versions and between three PCR curve analysis methods. Papilloplex HR‐HPV version1 assay exhibited excellent reproducibility, completing the international validation criteria. Papilloplex HR‐HPV version6 showed excellent concordance with version1 but still lacks clinical validation.

The Sansure® Human Papillomavirus DNA Diagnostic Kit offers excellent reproducibility performance for the detection of high‐risk HPV

AbstractCervical cancer screening is a cornerstone of cervical cancer elimination. Detection of high‐risk human papillomavirus (hrHPV) is recommended as the first step in screening provided that the assay used has been adequately validated. The Sansure® Human Papillomavirus DNA Diagnostic Kit is a new assay designed to detect HPV16, HPV18 and 13 other HPV in aggregate. The study aimed to evaluate the intra‐ and interlaboratory reproducibility of the assay according to international guidelines. Five hundred and fifty cervical residual cell samples from women attending cervical cancer screening were selected from the biobank of the HPV National Reference Centre in Belgium and used in this study. After DNA extraction, HPV was tested using the Sansure® Human Papillomavirus DNA Diagnostic Kit. The lower 95% confidence limit around the general reproducibility of this assay should be greater than or equal to 87%, with κ ≥ 0.50. Five hundred and thirty‐three samples had valid results. The Sansure® Human Papillomavirus DNA Diagnostic Kit demonstrated an excellent intra‐laboratory reproducibility of 93.8% (95% confidence interval [CI]: 91.4–95.7, κ = 0.85). The interlaboratory reproducibility was 93.4 (95% CI: 91.0–95.4, κ = 0.84). Intra and interlaboratory reproducibility were also excellent at the genotype level. Excluding HPV53 single infection samples from the analyses also resulted in excellent agreement. These data show that the Sansure® Human Papillomavirus DNA Diagnostic Kit is highly reproducible.

Human papillomavirus genotype distribution among high-grade cervical lesions in French women born between 1972 and 1993

Implementation of HPV vaccination programmes may induce significant changes in the distribution of HPV genotypes detected in cervical (pre-)cancers. In a multicentre study in France, we investigated the distribution of HPV genotypes using a sensitive molecular technique in 611 histological specimens diagnosed with high-grade lesions (CIN2/3) in a large cohort of women born between 1972 and 1993. Women born after 1983 were potentially vaccinated because of their young age. High-risk HPV genotypes were detected in 93.4 % of women and 24.4 % had a multiple infections. HPV16/18 genotypes were more frequently associated with CIN3 than with CIN2 lesions (66 % vs. 57 %, p = 0.045). The prevalence of HPV16 and HV18 genotypes did not change according to the year of birth. Contrary to previous descriptions in countries with a high HPV vaccination coverage, we did not observe difference in the prevalence of HPV16 and HPV18 genotypes in CIN2/3 lesions after the introduction of HPV vaccination in France, in a context of low national vaccination coverage. NCT04167501.

The level of expression of HPV16 early transcripts is not associated with the natural history of cervical lesions

Abstract The natural history of cervical cancer is closely linked to that of high‐risk human papillomaviruses (HPV) infection. It is recognized that upon HPV DNA integration, partial or complete loss of the E2 open reading frame precludes expression of the corresponding protein, resulting in upregulation of the E6 and E7 viral oncoproteins. To better characterize HPV16 infection at the cervical level, viral load, viral DNA integration, and viral early transcript expression (E2, E5, and E6) were analyzed in a series of 158 cervical specimens representative of the full spectrum of cervical disease. Overall, the frequency of early transcript detection varied from 45% to 90% and tended to increase with lesion severity. In addition, the levels of E2, E5, and E6 transcript expression were slightly higher in high‐grade lesions than in cervical specimens without abnormalities. Notably, early transcript expression was clearly associated with viral load, and no inverse correlation was found between the expression of E2 and E6 transcripts. No clear association was found between early transcript expression and HPV16 DNA integration, with the exception that samples with a fully integrated HPV16 genome did not harbor E2 or E5 transcripts. In conclusion, early HPV16 transcript expression appears to be associated with viral load rather than lesion grade. From a practical point of view, quantification of HPV16 early transcripts is difficult to translate into a relevant biomarker for cervical cancer screening.

Criteria for second generation comparator tests in validation of novel HPV DNA tests for use in cervical cancer screening

Abstract While HC2 and GP5+/6+ PCR‐EIA were pivotal in test validation of new HPV assays, they represent the first generation of comparator tests based upon technologies that are not in widespread use anymore. In the current guideline, criteria for second‐generation comparator tests are presented that include more detailed resolution of HPV genotypes. Second‐generation comparator tests should preferentially target only the 12 genotypes classified as carcinogenic (IARC‐group I), and show consistent non‐inferior sensitivity for CIN2+ and CIN3+ and specificity for ≤CIN1 compared to one of the first‐generations comparators, in at least three validation studies using benchmarks of 0.95 for relative sensitivity and 0.98 for relative specificity. Validation should take into account used storage media and other sample handling procedures. Meta‐analyses were conducted to identify the assays that fulfill these stringent criteria. Four tests fulfilled the new criteria: (1) RealTime High‐Risk HPV Test (Abbott), (2) Cobas‐4800 HPV test (Roche Molecular System), (3) Onclarity HPV Assay (BD Diagnostics), and (4) Anyplex II HPV HR Detection (Seegene), each evaluated in three to six studies. Whereas the four assays target 14 carcinogenic genotypes, the first two identify separately HPV16 and 18, the third assay identifies five types separately and the fourth identifies all the types separately.

Intra‐ and interlaboratory reproducibility evaluation toward international validation status of the AmpFire assay

Abstract To meet the screening goal of WHO's 90‐70‐90 strategy aimed at eliminating cervical cancer (CC) by 2030, clinical validation of human papillomavirus (HPV) assays is essential to provide accurate and valid results through fulfilling three criteria of the international validation guidelines (IVGs). Previously, the clinical accuracy of the AmpFire® HPV Screening 16/18/HR assay (AmpFire assay) was reported but reproducibility data are lacking. Here, we aim to evaluate the intra‐ and inter‐laboratory reproducibility of the AmpFire assay. The reproducibility of the isothermal AmpFire assay was assessed using 556 cervical cell samples collected from women attending CC screening and biobanked in a Belgian HPV national reference center. This assay detects HPV16, HPV18, and 12 other high‐risk HPV (hrHPV) types (31/33/35/39/45/51/52/56/58/59/66/68) in aggregate. Lower 95% confidence interval bound around the assay's reproducibility should exceed 87%, with κ  ≥ 0.50. Additionally, a literature review of the assay's clinical performance was performed. The AmpFire assay showed an excellent intralaboratory (96.4%, 95% CI:94.5–97.8%, κ  = 0.920) and interlaboratory (95.3%, 95% CI:93.2–96.9%, κ  = 0.897) reproducibility. One study demonstrated noninferior sensitivity of a prototype AmpFire assay targeting 15 hrHPV types (including HPV53) to detect CIN2+. However, clinical specificity became similar to the comparator after removing HPV53 from analyses. The low‐cost and easy‐to‐use AmpFire assay presents excellent reproducibility and—after removing HPV53 from the targeted types—fulfills also clinical accuracy requirements. Inclusion of HPV53, which is not recognized as carcinogenic, comprises clinical specificity of screening assays.

Prevalence and significance of HPV DNA detection below the clinical threshold of the commercial kit Alinity m HR‐HPV assay (Abbott)

Abstract The positive clinical threshold of human papillomavirus (HPV) tests validated for primary cervical cancer screening (CCS) is designed to offer an optimal balance between clinical sensitivity and specificity. However, there may be a gap between the analytical sensitivity of the test and the positive clinical threshold, referred to here as the “gray‐zone.” This study aims to determine the prevalence and significance of HPV results obtained in the gray‐zone in routine practice. Cervical samples obtained in our institution for CCS over a 22‐month‐period were tested with the Alinity m HR‐HPV Assay (Abbott). Clinical and biological data, including cytological results and patients' HPV history were collected. Of the 6101 samples collected, 1.7% had an HPV result in the gray‐zone (102 patients). The proportion of gray‐zone results varied according to HPV genotype, reaching 11.8% of samples with detectable HPV DNA in the case of HPV31/33/52/58 genotypes. Reflex cytologies showed no abnormalities or Atypical Squamous Cells of Undetermined Significance results in 74.6% and 17.9% of cases, respectively. A previous or subsequent HPV‐positive result with a (possibly) identical genotype was observed in 58% and 38% of cases, respectively. Two women with a history of persistent HPV detection had a CIN2+ lesion 1 year after the gray‐zone result. In conclusion, the proportion of HPV results in the gray‐zone varies according to genotype. No cytological abnormality is observed in the majority of cases, but a few rare patients with a history of persistent HPV infection should be closely monitored even if the HPV result is transiently located in the gray‐zone.