ABSTRACT
The Alinity m HR HPV assay (Alinity) is a novel human papillomavirus (HPV) assay that individually identifies genotypes HPV16, HPV18, and HPV45 while reporting on 11 other high-risk HPV (hrHPV) genotypes in two aggregates: HPV31/33/52/58 and HPV35/39/51/56/59/66/68. The clinical performance of Alinity for screening for cervical cancer was evaluated in population-based settings. For women aged ≥30 years, the clinical sensitivity (n = 68) and specificity (n = 3,077) for the detection of cervical intraepithelial neoplasia grade 2+ (CIN2+) of Alinity were 100.0% and 92.4%, respectively, and were not inferior to those of the Qiagen Digene Hybrid Capture 2 high-risk HPV DNA assay (hc2) (P = 0.0006 and P < 0.0001, respectively). The intralaboratory reproducibility and interlaboratory agreement of Alinity were 96.7% (kappa, 0.92) and 98.7% (kappa, 0.97), respectively. In the group ≥30 years of age, women who were baseline hrHPV negative had a lower risk for CIN2+ at 3 years using Alinity (0.04%) than those with a normal baseline cytology (0.65%) and had a risk comparable to that determined by the Abbott RealTime High Risk HPV assay (0.04%), hc2 (0.08%), or the Roche Cobas 4800 HPV assay (0.04%). High-risk HPV16/18 infection was associated with a significantly higher baseline and 3-year CIN2+ and CIN3+ risk than the absence of HPV16/18 or the presence of hrHPVs at the baseline (all P values were <0.05). The baseline CIN2+ risk was 8.8% for those with HPV31/33/52/58 infection and 2.5% for those with HPV35/39/51/56/59/66/68 infection, while the 3-year CIN2+ risk was 17.0% and 4.9%, respectively (relative risk, 3.4 [P = 0.03] and 3.5 [P = 0.003], respectively), suggesting that extended genotyping by Alinity may be valuable in improving patient risk stratification. Alinity fulfills international consensus guideline criteria for primary cervical cancer screening and can be considered clinically validated, demonstrating safety comparable to that of other clinically validated HPV tests.
INTRODUCTION
Cancer of the cervix is a common gynecological cancer worldwide (1). Persistent infection with high-risk human papillomaviruses (hrHPV) is the cause of virtually all cervical cancers and their immediate precursors (2), with 2 of the 14 hrHPV genotypes, HPV16 and HPV18, being linked to approximately 70% of the cases of invasive cervical cancer (ICC) worldwide (3).
Evidence shows that screening for hrHPV DNA has a greater sensitivity than cytology for detecting cervical intraepithelial neoplasia (CIN), providing greater protection against cervical cancer (4, 5). hrHPV DNA testing can also be used effectively in triaging patients with equivocal cytology, for posttreatment follow-up, and for monitoring vaccine efficacy (6–8). Accordingly, several countries (e.g., Australia and Netherlands) recently adopted hrHPV-based testing and/or are in the transition from cytology- to hrHPV-based testing for primary cervical cancer screening. Although at least 250 commercial HPV tests and over 150 test variants are currently available (9), a great majority of these do not have the documented clinical performance characteristics required for HPV testing and are therefore considered unreliable for use in primary cervical cancer screening.
HPV test performance can be evaluated in several ways. For regulatory approval, the United States Food and Drug Administration (FDA) follows a rigorous evaluation process, while the conformity assessment applied in the European Conformity Mark for In Vitro Diagnostics (CE-IVD) marking process is less demanding than FDA’s procedure and largely technical (9). The most widely adopted and internationally accepted guidelines describing the minimum requirements of hrHPV DNA tests for primary cervical cancer screening were introduced in 2009 and provide guidance on the validation of new candidate HPV assays (10). More recently, there has been growing interest among laboratory experts and manufacturers of HPV assays in the validation of HPV genotyping tests (VALGENT) project, which offers a framework for comparing and validating HPV assays, including those that permit limited or extended genotyping (11).
To the best of our knowledge, 11 hrHPV assays have been fully clinically validated according to the international 2009 guidelines (10), based on their reproducibility and sensitivity/specificity relative to the results of two hrHPV tests, the Digene Hybrid Capture 2 high-risk HPV DNA assay (hc2; Qiagen, Gaithersburg, MD, USA) and/or the GP5+/6+ PCR-enzyme immunoassay (GP5+/6+ PCR-EIA; Diassay, Rijswijk, the Netherlands) (12, 13), which have been clinically validated in cohort studies and large, well-designed randomized clinical trials with a follow-up period of at least 8 years and which are considered the standard comparators for evaluating the clinical performance of novel HPV tests (12, 14, 15).
The Alinity m HR HPV assay (Alinity; Abbott Molecular, Des Plaines, IL, USA) is a novel HPV assay launched in 2019 as part of a family of next-generation molecular assays for use with the automated continuous random-access Abbott Alinity m system. The novel HPV assay offers an innovative approach to the detection and partial genotyping of the 14 hrHPV genotypes, providing information on 3 main individual hrHPV genotypes (HPV16, HPV18, and HPV45) and reporting information on the remaining 11 targeted hrHPV genotypes in two aggregates, HPV31/33/52/58 (other HR HPV A) and HPV35/39/51/56/59/66/68 (other HR HPV B) (16). The channel for HPV31/33/52/58 together with the three separate readings for HPV16, HPV18, and HPV45 covers all hrHPV genotypes included in the nonavalent HPV vaccine (Gardasil 9; Merck & Co., Inc., Whitehouse Station, NJ, USA), offering an additional risk assessment. Here, we describe the findings from the first evaluation of Alinity’s clinical performance in population-based cervical cancer screening settings.
MATERIALS AND METHODS
Study population and protocol.The study population has been described in detail previously (17, 18). Briefly, between December 2009 and August 2010, a total of 4,510 Slovenian women aged 20 to 64 years, attending routine screening for cervical cancer in 16 outpatient gynecology services covering the entire country, were prospectively enrolled in the baseline screening round of the study. After visually inspecting the cervix, two samples were collected, a sample for conventional cervical cytology and a sample for HPV testing, with the latter sample being stored in ThinPrep PreservCyt solution (Hologic, Marlborough, MA, USA). After 36 months, baseline screening participants were invited to the second round of screening, using a similar approach, between December 2012 and October 2014. Patient- and physician-based questionnaires gathered information on events of clinical relevance (e.g., colposcopy, cytology, histology, HPV testing, treatment) occurring between the two screening rounds. After 3 years of follow-up from the baseline enrollment, final disease status was determined for all study participants.
The study was approved by the National Medical Ethics Committee of the Republic of Slovenia (consent numbers 83/11/09 and 109/08/12) and conducted according to the principles of the Declaration of Helsinki. All study participants provided written informed consent prior to enrollment.
hrHPV testing.Specimens collected during the baseline screening round were tested immediately upon arrival in the laboratory with two clinically validated hrHPV tests, the Digene Hybrid Capture 2 high-risk HPV DNA assay (hc2; Qiagen, Gaithersburg, MD, USA) and the Abbott RealTime High Risk HPV assay (RealTime; Abbott, Wiesbaden, Germany), and samples that tested positive for hrHPV with either test were genotyped using the algorithm described previously (17). The remaining samples were split into aliquots and stored at −70°C. The approach used in the second screening round was similar, while testing was performed with RealTime only. In December 2016, all samples from the baseline screening round were additionally tested with another clinically validated and FDA-approved HPV test, the Cobas 4800 HPV test (Cobas; Roche Molecular Systems, Alameda, CA, USA) at Amedes MVZ Wagnerstibbe für Laboratoriumsmedizin, Hämostaseologie, Humangenetik und Mikrobiologie, Hannover, Germany. All results generated in the initial studies were used in the evaluation of Alinity performance.
Testing of samples from both screening rounds with Alinity was performed in August 2018 and followed the manufacturer’s instructions. The results obtained using Alinity were determined by the assay software and are based on comparison of the specimen’s cycle number (CN) values for each HPV signal against signal-specific, established cutoff values. For each sample, five HPV signals in separate fluorescence channels, corresponding to (i) HPV16, (ii) HPV18, (iii) HPV45, (iv) HPV31/33/52/58 (a pooled signal, consisting of single signal corresponding to any individual genotype or combination of genotypes within the group, reported as “other HR HPV A”), and (v) HPV35/39/51/56/59/66/68 (a pooled signal reported as “other HR HPV B”), are measured. Each signal is determined either as “HPV detected,” if the CN is less than or equal to a fixed assay cutoff cycle for that signal, or as “HPV not detected,” if either the CN is not generated or is greater than the assay cutoff cycle (16). The endogenous human beta-globin sequence serves as an internal control to evaluate cell adequacy, sample extraction, and amplification efficiency and is measured in a separate channel.
Intralaboratory reproducibility and interlaboratory agreement.The intralaboratory reproducibility of Alinity over time was assessed with a total of 550 samples that comprised 183 hrHPV-positive and 367 hrHPV-negative randomly selected samples that were retested after 369 to 379 days from the baseline testing, following the guideline recommendations of Meijer et al. (10). Additionally, a set of coded ThinPrep aliquots was prepared from the same samples and shipped on dry ice to a collaborating laboratory (Viollier, Allschwil, Switzerland) in August 2019, where additional HPV testing was performed to assess interlaboratory agreement (retesting range, 368 to 376 days). All samples used in the reproducibility testing were recoded, and all the technicians performing the assay in the two laboratories were blind to HPV status.
Cytological examination and colposcopic referral.Cervical smears were assessed under standard conditions by certified cytologists, who were blind to the results of HPV testing (17, 18). In accordance with Slovenian screening standards, referral for colposcopy occurred if cytological grading reached or exceeded atypical squamous cells, cannot exclude a high-grade epithelial lesion (ASC-H), threshold (19). The study protocol also strongly recommended immediate colposcopy for all women with HPV16 and/or HPV18, irrespective of the cytology findings. For women who were positive for other hrHPV genotypes, colposcopy took place at the discretion of the physician. Certified pathologists, blind to the participants’ HPV status, assessed the histopathology of colposcopy-directed punch biopsy specimens obtained from any areas of suspicion. We considered histologically confirmed CIN grade 2+ (CIN2+) to be the clinical disease outcome.
Data analysis.The clinical performance of Alinity and hc2 was assessed using noninferiority testing (10) involving women with histologically confirmed high-grade cervical disease (CIN2+, who were considered cases) and those without high-grade cervical disease (CIN1 or lower, who were considered controls). The number of samples included in our analysis was in line with international 2009 guidelines (10), which require testing of at least 60 samples and 800 samples to achieve 80% power for the assessment of clinical sensitivity and specificity, respectively. Based on guideline requirements, the noninferiority of Alinity was demonstrated if the clinical specificity for the detection of lesions of CIN2 or lower was ≥98% and the clinical sensitivity for the detection of CIN2+ lesions was ≥90% compared with the results of hc2 (10). For intralaboratory reproducibility and interlaboratory agreement, a lower confidence bound of ≥87% and a kappa value of >0.5 were used as thresholds, as recommended in international guidelines (10).
Three separate analyses were performed to assess whether the baseline presence of specific hrHPV genotypes predicts the risk of developing CIN2+ and CIN3+ at 3 years (18). Following stratification by the baseline HPV result, women were included in the analyses if they had either (i) a diagnosis of CIN2+ by consensus or (ii) ≥1 valid hrHPV DNA and/or cytology result obtained 36 to 48 months after enrollment in the baseline screening round and had had a colposcopy according to the study protocol. Vaccination against HPV and/or a high probability or actual occurrence of a therapeutic procedure on the cervix between screening rounds was an exclusion criterion.
The baseline and the 3-year risk of CIN2+ and CIN3+ with the corresponding 95% confidence intervals (CIs) were assessed according to the baseline cytology and HPV results as described in detail previously (18). Briefly, (i) the number of incident CIN2+ cases was divided by the number of study participants at risk to assess the safety of a negative result according to distinctive baseline characteristics, (ii) the number of baseline CIN2+ cases was divided by the number of study participants at risk (women who comprised the baseline screening round analysis population) to assess the baseline risk, and (iii) the number of all CIN2+ cases was divided by the number of study participants (women who comprised the second screening round analysis population) to assess the 3-year risk (18). Statistical analyses were performed using R software (version 2.12.0; Free Software Foundation, Boston, MA, USA), with significance being set at a level of 0.05.
RESULTS
Clinical performance of Alinity and hc2 for women aged ≥30 years.The primary study population used to evaluate the clinical performance of Alinity and hc2 was women aged ≥30 years (n = 3,145; mean and median age, 41.5 and 40.0 years, respectively). The overall prevalence of hrHPV infection, regardless of HPV genotype, determined using Alinity was 9.6% (301/3,145; 95% CI, 8.6 to 10.7%). The prevalence of hrHPV infection was the highest among women aged 30 and 34 years and then declined with age.
Alinity HPV results for women aged ≥30 years, stratified by cases and controls, compared to the results of hc2 are shown in Table 1. For calculation of clinical sensitivity, 68 CIN2+ cases were detected in the two screening rounds in our cohort: 28 had CIN2, 31 had CIN3, 1 had carcinoma in situ (CIS), 4 had invasive carcinomas (squamous cell carcinoma [SCC]), 3 had adenocarcinomas in situ (ACIS), and 1 had adenocarcinoma (ADC). For calculation of clinical specificity, 3,077 women who did not have histologically confirmed CIN2+ (controls) were identified from a population-based Slovenian cohort. The clinical sensitivity and specificity of Alinity for the detection of CIN2+ in women aged ≥30 years were 100.0% (68/68; 95% CI, 92.2 to 100.0%) and 92.4% (2,844/3,077; 95% CI, 91.4 to 93.3%), respectively, and those of hc2 were 95.6% (65/68; 95% CI, 87.6 to 99.1%) and 91.9% (2,829/3,077; 95% CI, 90.9 to 92.9%), respectively.
Comparison of Alinity and hc2 results stratified for cases and controls in women aged ≥30 yearsc
Alinity demonstrated noninferiority for both relative clinical sensitivity (P = 0.0006) and clinical specificity (P < 0.0001) compared to the hc2 reference test.
Intralaboratory reproducibility and interlaboratory agreement.The intralaboratory reproducibility of Alinity over time was 96.7% (532/550; 95% CI, 94.8 to 98.0%), with a kappa value of 0.92 (95% CI, 0.89 to 0.96). The interlaboratory agreement of Alinity was 98.7% (543/550; 95% CI, 97.3 to 99.4%), with a kappa value of 0.97 (95% CI, 0.95 to 0.99). Thus, the Alinity HPV assay displayed a high degree of reproducibility and met the requirements for the lower confidence bound (≥87%) and the kappa value (>0.5) (10).
Clinical comparison of Alinity with cytology and other clinically validated HPV tests. (i) Three-year CIN2+ risk stratified by various baseline characteristics.The first and second screening rounds identified 68 and 36 CIN2+ cases, respectively. Of the latter, 17 were detected passively through clinical records and 19 were detected actively as a result of findings from cytology and/or HPV tests in the second screening round (18).
Regardless of the age restriction (women aged ≥30 years and the total study population), women with a negative baseline hrHPV result, regardless of the HPV assay used (Alinity, RealTime, hc2, or Cobas), had a lower CIN2+ risk at 3 years than women who had a normal baseline cytology (Table 2). In addition, the 3-year CIN2+ risk was similar for women with a normal baseline cytology and a negative hrHPV result and women testing hrHPV negative only at the baseline using Alinity, RealTime, hc2, or Cobas alone (Table 2).
Three-year risk for CIN2+ according to various negative baseline characteristics calculated separately for women aged >30 years and for the total study populationa
(ii) CIN2+ and CIN3+ risks for women with baseline HPV infection according to distinct hrHPV genotypes.The baseline and 3-year CIN2+ and CIN3+ risks for women who tested positive for distinct hrHPV genotypes at the baseline are shown in Fig. 1 and 2. Baseline HPV16 and HPV16/18 infection was significantly associated with a greater 3-year risk for both CIN2+ and CIN3+ than hrHPV positivity alone or non-HPV16/18 hrHPV baseline infection (all P values were <0.05; see Fig. S1 and S2 in the supplemental material). An HPV16- and/or HPV18-positive result at the baseline strongly predicted either an underlying high-grade lesion or a high possibility of developing CIN2+ and CIN3+ in the short term. Regardless of the test used (Alinity, RealTime, Cobas) and time (baseline, 3 years), the risk for CIN2+ was significantly higher for women positive for HPV16 and HPV16/18 than for non-type 16/18 hrHPV-positive women (Fig. S1 and S2). Interestingly, women positive for HPV31/33/52/58 and HPV35/39/51/56/59/66/68 exhibited remarkably different risks for CIN2+ and CIN3+, as shown in Fig. 1 and 2. For CIN2+, the baseline risk was 8.8% for HPV31/33/52/58 and 2.5% for HPV35/39/51/56/59/66/68, while the 3-year risks were 17.0% and 4.9%, respectively (relative risk [RR], 3.4 [95% CI, 1.2 to 10.2; P = 0.03] and 3.5 [95% CI, 1.5 to 7.8; P = 0.003], respectively). For CIN3+, the baseline risk was 4.4% for HPV31/33/52/58 and 1.3% for HPV35/39/51/56/59/66/68, while the 3-year risks were 9.6% and 2.8%, respectively (RR, 3.4 [95% CI, 0.7 to 16.3; P = 0.12] and 3.4 [95% CI, 1.1 to 10.2; P = 0.03], respectively). Thus, for example, 1 out of 11 women positive for HPV31/33/52/58 and 1 out of 40 women positive for HPV35/39/51/56/59/66/68 will have underlying CIN2+, yielding a 3.4-fold increase in risk for underlying CIN2+ for women positive for HPV31/33/52/58 compared to that for women positive for HPV35/39/51/56/59/66/68. Similarly, 1 out of 6 women positive for HPV31/33/52/58 and 1 out of 20 women positive for HPV35/39/51/56/59/66/68 will be diagnosed with CIN2+ at the baseline and/or within 3 years, yielding a 3.5-fold increase in risk for CIN2+ at the baseline and/or within 3 years for women positive for HPV31/33/52/58 compared to that for women positive for HPV35/39/51/56/59/66/68.
Risk (95% CI) for CIN2+ at the baseline and at 3 years for women with various baseline HPV results according to the HPV test used (Alinity, RealTime, Cobas, hc2) in the total study population. The risk for CIN2+ at the baseline was assessed in women enrolled in the baseline screening round, based on cases of CIN2+ identified in that round only. The risk for CIN2+ at 3 years was assessed in women included in the final analysis of the second screening round, based on the CIN2+ cases identified in both screening rounds. Only samples with valid results by all HPV assays were included in the analysis. *, samples non-HPV16/18 hrHPV positive by Alinity were analyzed separately for channel HPV31/33/52/58 (other HR HPV A) and channel HPV35/39/51/56/59/66/68 (other HR HPV B). CI, confidence interval; CIN2+, cervical intraepithelial neoplasia grade 2+; Alinity, Alinity m HR HPV assay; RealTime, Abbott RealTime High Risk HPV assay; Cobas, Roche Cobas 4800 HPV test; hc2, Qiagen Digene Hybrid Capture 2 high-risk HPV DNA assay; HrHPV, high-risk human papillomavirus.
Risk (95% CI) for CIN3+ at the baseline and at 3 years for women with various baseline HPV results according to the HPV test used (Alinity, RealTime, Cobas, hc2) in the total study population. The risk for CIN3+ at the baseline was assessed in women enrolled in the baseline screening round, based on cases of CIN3+ identified in the baseline screening round. The 3-year risk for CIN3+ was assessed in women included in the final analysis of the second screening round, based on CIN3+ cases identified in both screening rounds. Only samples with valid results by all HPV assays were included in the analysis. *, samples non-HPV16/18 hrHPV positive by Alinity were analyzed separately for channel HPV31/33/52/58 (other HR HPV A) and channel HPV35/39/51/56/59/66/68 (other HR HPV B). CI, confidence interval; CIN3+, cervical intraepithelial neoplasia grade 3+; Alinity, Alinity m HR HPV assay; RealTime, Abbott RealTime High Risk HPV assay; Cobas, Roche Cobas 4800 HPV test; hc2, Qiagen Digene Hybrid Capture 2 high-risk HPV DNA assay; HrHPV, high-risk human papillomavirus.
DISCUSSION
International guidance recommends a clinical validation strategy for novel hrHPV DNA assays which involves a comparative noninferiority analysis of samples from a population-based screening cohort using a clinically validated reference hrHPV assay, such as hc2 (10). Although a large number of HPV assays are commercially available, only a few of them currently meet the criteria defined in the guidelines and can be viewed as clinically validated and reliable for use in primary screening for cervical cancer (12).
The recently launched Alinity provides individual information for three main HPV genotypes (HPV16, HPV18, and HPV45) and concurrent detection of the remaining 11 targeted HPV genotypes in two aggregates: HPV31/33/52/58 and HPV35/39/51/56/59/66/68. The rationale for this approach is 2-fold. First, the HPV31/33/52/58 channel, together with the three individual readings for HPV16, HPV18, and HPV45, covers all hrHPV genotypes included in the nonavalent HPV vaccine (Gardasil 9). In addition, based on the findings for more than 6,000 women diagnosed with the CIN2+ or invasive cervical cancer (ICC) from 17 European countries, the most common HPV types in women with ICC were HPV16, HPV18, and HPV45 (63.3, 15.2, and 5.3%, respectively), with the HPV16, HPV18, and HPV45 prevalence being 1.1, 3.5, and 2.5 times higher in women with ICC than in women with CIN2+, respectively. In Europe, HPV16 predominates in women with CIN2+ and in women with ICC, whereas HPV18 and HPV45 are associated with a low median age of ICC. Thus, findings from this study support the need for a special focus on the prevention of HPV16-, HPV18-, and HPV45-related cervical lesions (20). Second, evidence from several studies has demonstrated a wide variation in the risk associated with different hrHPV genotypes (21–25). Although HPV16 is the type associated with the highest risk for underlying as well as subsequent CIN3+, several prospective studies with different follow-up periods also reported a high absolute risk associated with HPV18, HPV31, and HPV33 (23, 26, 27). In Swedescreen, a population-based randomized clinical trial where more than 11,000 women aged 32 to 38 years from five regions of Sweden were enrolled and followed for up to 14 years, it has been shown that hrHPV types cluster in three different risk groups. Women with HPV16/18/31/33 had 14‐year cumulative incidences of CIN3+ above 28% (highest risk), those with HPV35/45/52/58 had 14‐year cumulative incidences of between 14% and 18% (medium risk), and HPV39/51/56/59/66/68 was associated with a risk below 10% (limited risk) (28). Thus, the extended genotyping capability of Alinity may be of value in improving patient risk stratification.
In our study, the clinical sensitivity and specificity of Alinity for the detection of CIN2+ in women aged ≥30 years were 100.0% and 92.4%, respectively, and both parameters demonstrated noninferiority to hc2 at the recommended thresholds (sensitivity, ≥98%; specificity, ≥90%). In addition, our manufacturer-independent evaluation of the intralaboratory reproducibility and interlaboratory agreement of Alinity performed on 550 ThinPrep samples (including more than 30% HPV-positive samples) showed a high overall percent agreement when retesting was performed approximately 1 year later in the same laboratory (96.7% [lower bound of the 95% CI, 94.8%]) as well as in a collaborative laboratory (98.7% [lower bound of the 95% CI, 97.3%]). Based on the international guideline criterion of a lower confidence bound of ≥87% (10), our data suggest that Alinity has a robust and reliable performance. Considering the results of our study, Alinity fulfills international consensus guideline requirements for HPV tests in primary cervical cancer screening (10) and therefore can be considered clinically validated.
Our study presents the first data on the longitudinal clinical performance of Alinity, determined using a cohort of 4,510 Slovenian women aged 20 to 64 years who attended routine screening for cervical cancer with >70% coverage. Significantly fewer cases of CIN2+ disease were detected in the second round of screening than at the baseline (36 versus 68, respectively), which is unsurprising, given that prevalent cases are primarily identified in the baseline screening round and only incident cases are found in the second round of screening. In both study populations (women aged ≥30 years and the total study population) and regardless of the HPV test used (all P values were <0.05), HPV screening offered better protection than cytology, and all hrHPV tests evaluated showed comparable results. The 3-year CIN2+ risk in participants who tested hrHPV negative and who had a normal cytology at the baseline did not significantly differ from that in women who tested Alinity negative only at the baseline for both study populations analyzed (P = 0.97 and P = 0.70, respectively). This indicates that a negative cytology result does not provide any additional value in protecting against CIN2+ over and above a negative Alinity HPV result.
The findings from this study are in agreement with earlier 3-year data on the performance of RealTime and hc2 in primary cervical cancer screening settings. A previous analysis of data from our cohort of Slovenian women showed that women aged ≥30 years who had a negative baseline hrHPV DNA result had a significantly lower CIN2+ risk at 3 years (0.04% [95% CI, 0.00 to 0.22%]) than women who had a normal baseline cytology (0.68% [95% CI, 0.40 to 1.08%]) (18). The ATHENA trial in women aged ≥25 years (n = 42,209) found cumulative 3-year CIN3+ risks of 0.3% (95% CI, 0.1 to 0.7%), 0.8% (95% CI, 0.5 to 1.1%), and 0.3% (95% CI, 0.1 to 0.6%) following a negative Cobas 4800 HPV, negative cytology, or double-negative Cobas/cytology result, respectively (29). In the CLEAR study with 10,860 women aged >30 years, women who had a normal baseline cytology plus a negative hc2 or Aptima HPV assay result were at a similarly low 3-year CIN2+ risk (<0.3%), whereas the risk was 5.1% and 6.3% in women with a positive HPV result using the hc2 and Aptima HPV assays, respectively (30). Other longitudinal data from 14 years of follow-up of a large screening cohort of women aged ≥29 years from the Dutch Population-Based Screening Study Amsterdam (POBASCAM) trial showed that the cumulative incidence of cervical cancer (0.09%) and CIN3+ (0.56%) in women who were HPV negative in the intervention group (HPV and cytology cotesting) after three rounds of screening at 5-year intervals was similar to that in cytology-negative women in the control group (tested by cytology only) after two rounds (0.09% and 0.69%, respectively). The RRs for cervical cancer and CIN3+ were 0.97 (95% CI, 0.41 to 2.31; P = 0.95) and 0.82 (95% CI, 0.62 to 1.09; P = 0.17), respectively (31).
In our previous analysis, HPV16 infection at the baseline was associated with a significantly higher baseline CIN2+ risk (21.9% [95% CI, 15.2 to 30.4%]) and a baseline plus future 3-year CIN2+ risk (33.3% [95% CI, 24.7 to 44.0%]) than the absence of high-risk HPV16/18 infection at the baseline (7.0% [95% CI, 4.6 to 10.2%]) or baseline hrHPV positivity (11.7% [95% CI, 9.1 to 14.9%]) (18).
For women who are hrHPV positive and cytology negative after HPV testing in cervical cancer screening settings, subsequent clinical management is critical (32). HPV16 and HPV18 genotyping is supported by evidence from long-term observational studies showing an elevated risk of CIN3+ compared to other hrHPV genotypes, particularly for HPV16 (32). Accordingly, interim U.S. guidance on the use of primary hrHPV testing for cervical cancer screening recommends that a clinical management plan for women who are hrHPV positive be guided by a combination of HPV genotyping for those who are HPV16/18 positive and reflex cytology for women who are positive for 12 other hrHPV genotypes (33). However, HPV tests allowing genotyping should not focus only on HPV16/18 but should also focus on HPV45 (34) and HPV31/33 that, in combination, rival even HPV16 as a risk for CIN3+ and cervical cancer in women ≥30 years of age (26, 28). In addition, HPV16, HPV18, and HPV45 are the three most common types in cervical cancer, together account for 75% of SCC cases and 94% of ADC cases (34), and are more frequently associated with a diagnosis of cervical cancer in younger women than with a diagnosis of cancers associated with other HPV types (35). Thus, further information on which other hrHPV genotypes could provide useful risk stratification in patient management is needed. Here, we demonstrated that the 3-year CIN2+ and CIN3+ risk was significantly higher for those HPV31/33/52/58 positive than for those HPV35/39/51/56/59/66/68 positive (P = 0.003 for CIN2+ and P = 0.03 for CIN3+). These findings might warrant different approaches for the triage of these women.
A potential limitation of this study is that Alinity testing was performed up to 8 years subsequent to the hc2 and RealTime testing and up to 2 years subsequent to Cobas testing. However, due to the observed clinical noninferiority of Alinity in comparison to all three comparator assays, it is highly unlikely that prolonged specimen storage at −70°C influenced its performance. In addition, results from the clinical assessment of 15 different HPV assays evaluated within the third iteration of the VALGENT project, where testing with different HPV tests was performed up to 7 years apart, imply that archival ThinPrep samples, when appropriately aliquoted, handled, and stored, can be safely used for evaluation of the clinical performance of HPV DNA tests several years after sample collection (36–39).
In conclusion, results from the first comparative assessment of Alinity’s clinical performance for cervical cancer screening in a population-based study suggest that Alinity can be considered a clinically validated HPV assay, while 3-year longitudinal data demonstrate that Alinity offers safety comparable to that of other clinically validated HPV tests. Furthermore, Alinity provides extended genotyping beyond HPV16/18 that may be valuable in improving the risk stratification of hrHPV screen-positive women.
ACKNOWLEDGMENTS
We thank our colleagues Arndt Gröning and Alexander Bertram for excellent laboratory assistance.
This study was funded by Abbott Molecular and the Institute of Microbiology and Immunology, Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia. Mario Poljak and Anja Oštrbenk Valenčak were supported by the Seventh Framework Program of DG Research of the European Commission through the COHEAHR Network (grant no. 603019).
Abbott Molecular was not involved in the study design, data collection and analysis, interpretation of the results, or writing of the manuscript.
Editorial assistance in the development of this manuscript was provided by Sarah Stowell of Ashley Medical Communications.
The authors’ institution received several research grants from Abbott Molecular and Qiagen. Anja Oštrbenk Valenčak has received reimbursement for travel expenses for attending conferences and honoraria for speaking from Abbott Molecular, Qiagen, and Seegene. The other authors declare no competing interest.
FOOTNOTES
- Received 11 July 2019.
- Returned for modification 14 August 2019.
- Accepted 20 October 2019.
- Accepted manuscript posted online 30 October 2019.
Supplemental material is available online only.
- Copyright © 2019 American Society for Microbiology.
