Agreement Between High-Risk Human Papillomavirus Testing in Paired Self-Collected and Clinician-Collected Samples from Cervical Cancer Screening in Spain

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript presented for review discusses the up-to-date topic of gynecology and reproductive medicine. Although cervical cancer screening programs and  HPV vaccination is well developed, cervical cancer incidence is increasing in many countries worldwide. Since 2006 when the approval of the first HPV vaccine occurred by June 2020, 55% of the WHO Member States reported implementation of partial or national HPV vaccination..

 

Thus, any efforts to improve the cervical cancer screening, public knowledge of cervical cancer and HPV vaccination remain valuable. Therefore, the manuscript has significant value for modern medical practice and could be published after some requested revisions are done.

Please find below my comments:

1.      The title is clear and concise and represents the study.

2.      The abstract in general clearly presents the study, however, needs improvement in its structure and writing style.

3.      The introduction does not explain a full rationale for the study. Please restructure it. Include data on cervical cancer  epidemiology in Spain and updated data on cervical cancer incidence, cervical cancer screening practices and cervical cancer screening coverage.

4.      Provide details of the screening practices in Spain- age group, frequency, and method used for cervical cancer screening (Pap test, HPV genotyping, or co-testing with both).

5.      As a factor contributing to the increased cervical cancer incidence, HPV vaccination coverage in Spain also could  be mentioned here as Cervical cancer could be prevented not via screening but through vaccination. 

6. In the methods section, please justify your choice of self-sampling device. There are many available. This source could help you - doi: 10.1016/j.pmedr.2024.102590

7. Results are interesting and supported by clear tables, however, a figure or two could improve the data comprehension.

8. In the discussion part is narrow and difficult to comprehend in the current version. It requires restructuring. The following outline is suggested:

Outline for the Discussion part

1.1 Rationale of the study (why it was done)

1.1.1 Main findings of the study

1.1.2 What makes your study unique

1.1.3 What it adds to what we already know

1.2 Study subjects

1.3 Subject of the discussion

Comparison of your results with previous studies in the field. Agreement and disagreement with the studies compared 

 

1.4 Study strengths and limitations

1.5 Clinical implication

 

 

Author Response

Reviewer 1:

The manuscript presented for review discusses the up-to-date topic of gynecology and reproductive medicine. Although cervical cancer screening programs and HPV vaccination is well developed, cervical cancer incidence is increasing in many countries worldwide. Since 2006 when the approval of the first HPV vaccine occurred by June 2020, 55% of the WHO Member States reported implementation of partial or national HPV vaccination.

Thus, any efforts to improve the cervical cancer screening, public knowledge of cervical cancer and HPV vaccination remain valuable. Therefore, the manuscript has significant value for modern medical practice and could be published after some requested revisions are done.

Answer: Thank you for your detailed feedback and constructive suggestions regarding our manuscript. We appreciate your interest in our study and your recommendations to enhance its quality. Below, we address your comments point by point:

1. The title is clear and concise and represents the study.

Answer: Thank you for noting that the title is clear and accurately represents the study.

 

2. The abstract in general clearly presents the study, however, needs improvement in its structure and writing style.

Answer:  We have revised the abstract to improve its structure and writing style.

The abstract has been modified as follows:

“Background: Implementing self-sampling (SS) in cervical cancer screening requires comparable results to clinician-collected samples (CCS). Agreement measures are essential for evaluating HPV test performance. Previous studies on non-paired samples have reported higher viral cycle threshold (Ct) values in SS compared to CCS affecting sensitivity for detecting cervical intraepithelial neoplasia grade 2 or worse (CIN2+).

Objectives: We aimed to evaluate the agreement of high-risk (hr) HPV testing results between SS and CCS using paired samples and to explore differences in Ct values.

Methods: Women aged 30 to 65 years attending cervical cancer screening in two regions of Spain were invited to participated in the study. For each woman there was: CCS collected during the screening visit using liquid-based cytology and cytobrush, and a SS using a brush at home one month later. A PCR-based assay was used for hrHPV detection. Agreement in hrHPV results among both samples, Ct value differences and their association with screening outcomes were analyzed.

Results: The study included 981 women with paired samples. SS had a higher hrHPV prevalence than CCS (overall ratio of 1.3). Positive agreement for all hrHPV genotypes, HPV16, HPV18 and other hrHPV types were 85%, 91.3%, 66.7% and 83.3%, respectively. Negative agreement was >95% for all results. Median Ct values was slightly higher in SS than in CSS (32.9 vs 30.6, p=0.02). Seven CIN2+ cases HPV positive were detected by both methods. One CIN3 case was missed by SS..

Conclusions: This study showed a good agreement between SS and CCS for hrHPV testing in a routine screening in Spain. Despite the slightly higher Ct values for SS, no significant impact on sensitivity could be determined due to the low incidence of CIN2+ cases. Further research on larger paired samples is needed to assess the implications of Ct values on test sensitivity.”

 

3. The introduction does not explain a full rationale for the study. Please restructure it. Include data on cervical cancer epidemiology in Spain and updated data on cervical cancer incidence, cervical cancer screening practices and cervical cancer screening coverage.

Provide details of the screening practices in Spain- age group, frequency, and method used for cervical cancer screening (Pap test, HPV genotyping, or cotesting with both).

As a factor contributing to the increased cervical cancer incidence, HPV vaccination coverage in Spain also could be mentioned here as Cervical cancer could be prevented not via screening but through vaccination. 

 Answer: Thank you very much for your comment. We have revised and completed the introduction following the reviewer's suggestions.

“Introduction

Cervical cancer remains a significant public health concern globally, despite advances in screening and vaccination. In Spain, cervical cancer continues to have a substantial impact, with the age-standardized incidence rate in 2022 estimated at 5.4 cases per 100,000 women (crude rate 8.5) and an age-standardized mortality rate of 1.6 ^[1], being the eleventh leading cause of cancer in women of all ages.

Cervical screening coverage in Spain varies by age, with the highest rates observed between 25 and 65 years, the target age group for the cervical cancer screening program. According to the European Survey of Health in Spain in 2020, 72% of women aged 25 to 65 had a cervical cytology in the past three years, and 80% in the past five years ^[2] .It is estimated that 9,380,239 sexually active women aged 25 to 65 will undergo cervical  cytology every three years in Spain, representing 3,583,145 women annually ^[2]. However, it is noteworthy that in 2020, when the survey was conducted, the predominant cervical cancer screening recommendation in Spain was to perform a cervical cytology every three years on an opportunistic basis. Currently, following the modification of the cervical cancer screening program of the National Health System (Order SCB/480/2019), regions are transitioning to a population-based screening program, using liquid-based cytology for primary screening in women aged 25-29/34 years and detection of carcinogenic genotypes of human papillomavirus (hrHPV) every five years starting at age 30/35, with cytological triage for screen positives ^[3]. Screening ends at 65 years. Most regions uses HPV testing use risk-based clinical management algorithms, with limited genotyping ^[4].

In Spain, the systematic HPV vaccination for adolescent girls was implemented in 2008, with the vaccination age set at 12 years. In 2018, the recommendation was extended to certain conditions with a higher risk of developing anogenital cancer from HPV, such as women treated for cervical intraepithelial neoplasia grade 2 or 3 (CIN2-3). In 2022, systematic vaccination was extended to adolescent boys aged 11 to 12 years ^[5,6].In July 2024, a single-dose schedule was approved up to age 25, with a two-dose schedule for those aged 26 and older, except for individuals who are immunocompromised or undergoing treatment for HSIL/CIN2-3, who will continue with the three-dose schedule ^[5]. Since the start of the program, the overall HPV vaccination coverage in Spain has ranged between 70-80%. In 2023, the national coverage for the full vaccination schedule at age 15 was 86%, and the coverage for one dose was 91% ^[7].

Although the incidence of cervical cancer has declined in many countries due to the implementation of regular screening programs and the introduction of HPV vaccination, the disease remains a leading cause of cancer-related deaths, especially among underscreened or never screened women ^[8,9]. Nowadays, many scientific societies and organizations worldwide, including the World Health Organization (WHO), recommend using the detection of hrHPV as primary screening test rather than cervical cytology in women aged 30 years and above ^[4,10–13]. This shift has facilitated the introduction of new sample collection methods, such as cervico-vaginal self-sampling (SS). Numerous studies have evaluated the acceptability of SS among women in different regions and countries worldwide, demonstrating its feasibility and high acceptance among participants ^[14,15]. In most of the studies, women reported ease of use, reduced embarrassment, increased privacy, enhanced comfort, and overall convenience ^[15–17]. SS can be used to improve screening coverage ^[18] particularly for the underscreened women in whom the majority of cervical cancers are diagnosed ^[9,19].

A critical aspect is that to confidently implement SS, results need to not inferior to those seen when using CCS. A systematic review including data from 56 studies and 25 randomized controlled trials (RCT) conducted in under or never screened women showed that the detection of cervical intraepithelial neoplasia grade 2 or worse (CIN2+) through HPV testing on SS samples versus CCS was comparable, provided HPV DNA assays were based on a highly sensitive approach using a polymerase chain reaction (PCR)^[20]. Further RCT and population studies conducted in regular screening populations confirmed these findings ^[21,22].

However,  a study from The Netherlands where hrHPV-based screening programs used PCR as the HPV assay using either SS or CCS ^[23]. The study observed a 6% lower sensitivity and a 2% higher specificity for cervical intraepithelial neoplasia grade 3 or worse (CIN3+) in unpaired SS compared to clinician-based hrHPV testing. The authors suggested the detection threshold values (the Ct values), for both sampling approaches may need to be adapted to the sampling approach as SS samples had on average, higher Ct values ^[23].

The main objective of the study here presented was to evaluate the agreement of HPV test results between SS and CCS using paired samples and to explore the differences in Ct values between both samples. This study was part of a wider investigation that evaluated the implementation of SS in cervical cancer screening in Spain ^[17]”.

4. In the methods section, please justify your choice of self-sampling device. There are many available. This source could help you - doi: 10.1016/j.pmedr.2024.102590

Answer: Thank you very much for your question and suggestion. The following paragraph has been included in the Methods section to justify the choice of self-sampling device used in the study:

“The self-sample was collected using the Rovers Medical Devices Evalyn Brush (Rov-ers Medical Devices B.V, Oss, Netherlands). It is a dry self-sampling device. In 2018, when the study was approved and initiated, the Evalyn Brush was the most established and evidence-supported self-sampling device. It was validated for high-risk HPV detection across multiple PCR assays and is widely used in research [24,25]. Additionally, it was included in the national cervical screening program of the Netherlands [21] and served as the comparator device in the 2018 Valhudes protocol [26]. Its sample stability for up to 32 weeks at temperatures ranging from 4°C to 30°C ensured reliable results [27], making it particularly suitable for multicenter studies where sample processing may be delayed compared to a clinical diagnostic setting.”

5. Results are interesting and supported by clear tables, however, a figure or two could improve data comprehension.

Answer: Although figures could improve the visual representation of the data, tables are generally more informative for this type of study. They can later be included in systematic reviews and meta-analyses, where their detailed presentation is more relevant for the broader analysis of the findings.

6. In the discussion part is narrow and difficult to comprehend in the current version. It requires restructuring. The following outline is suggested:

Outline for the Discussion part

1.1 Rationale of the study (why it was done)

1.1.1 Main findings of the study

1.1.2 What makes your study unique

1.1.3 What it adds to what we already know

1.2 Study subjects

1.3 Subject of the discussion

Comparison of your results with previous studies in the field. Agreement and disagreement with the studies compared 

1.4 Study strengths and limitations

1.5 Clinical implication

Answer: We appreciate the reviewer’s valuable feedback regarding the structure and clarity of the discussion section. In response, we have reorganized the discussion to align with the suggested outline:

“Discussion

The agreement and concordance between SS and CCS in paired samples were evaluated in a cohort of women aged 30-65 years attending a routine cervical cancer screening in Spain. Samples were collected as part of the regular screening process, with SS per-formed at home by the women without clinician assistance. Results showed a high level of agreement and concordance between the two sample collection methods, both overall and across different HPV genotypes. The only exception was observed in the few HPV18 samples, where the positive agreement was lower compared to other genotypes. These findings support the use of SS as a reliable alternative to CCS in cervical cancer screening, particularly in routine screening.

Our study contributes to the growing body of literature supporting the use of SS for hrHPV testing. Our findings were consistent with previous research, including a study from Kaiser Permanente Northern California (N=144 paired collections) [32], which re-ported an overall positive agreement of 84.2% between SS and CCS with channel agreement rates of 90.3% for all HPV genotypes. The study also found high agreement for various HPV genotypes, with 97.8% agreement for HPV16, 99.3% for HPV18 and 92.4% for other hrHPV types. A similar study in Australia (303 paired samples) [33] also found comparable agreement rates between SS and CCS. In line with the meta-analysis previously mentioned in the introduction, which includes 26 studies with over 10,000 participants [31], a positive concordance of 59.3% (95% CI: 44.4%-73.5%) and a negative concordance of 85.6% (95% CI: 77.9%-91.8%) were reported. These findings, along with our study, highlight the use of SS as a tool for HPV testing in screening populations.

According to the authors of this meta-analysis, evaluating agreement/concordance parameters could also be used to validate new self-sample collection devices intended for use with a previously validated HPV test for self-sampling. This approach would eliminate the need for larger validation studies requiring histological confirmation, thus reducing both time and costs, and accelerating the validation process [31].

While agreement/concordance measures are essential for evaluating HPV test performance, assessing Ct values can provide valuable additional insights. Some studies have found that there is a linear correlation between the Ct values and the logarithm of HPV viral copies in the sample. Thus, a higher initial number of viral copies is associated with lower Ct values, and a higher viral load is linked to a greater presence of lesions [34–36]. Nevertheless, our study confirms the findings of Inturrisi et al. [23] in the Netherlands, showing that Ct values for SS were significantly higher than for CCS. The Dutch study re-ported a mean Ct difference of 2.73 (95% CI: 1.75-3.72) for CIN2 and 3.59 (95% CI: 3.03-4.15) for CIN3+, with a relative sensitivity of 0.9 (95% CI: 0.90-0.97) [23]. In our study, one CIN3 case missed by SS had a Ct value of 35.2 in the CCS sample.

Eight women were diagnosed with CIN2+ (2 cases of CIN2, 5 cases of CIN3, and 1 case of infiltrating squamous cell carcinoma). All but one CIN3 case tested HPV-positive using both sampling methods, with HPV16 detected in the carcinoma case and one CIN2 case, while all eight CIN2+ cases were also positive for other hrHPV types excluding HPV16/18. To draw more accurate conclusions regarding the relationship between higher Ct values in SS and potential sensitivity loss, larger studies involving paired samples in a screening context would be necessary.

Although the median viral Ct values in SS were slightly higher than those found in CCS, the study by Inturrisi et al. found no differences among age groups [23]. We observed variations in median Ct values among women under 50 years of age; however, linear regression analysis did not reveal statistically significant age-related differences underscoring the contribution of age associated with sampling methods.

In The Netherlands, 46.3% of the participants cited doubts about obtaining the SS correctly [16], while in our study 29,3% of the participants showed doubts about whether they had correctly taken the sample [17]. The proportion of SS with an invalid result was very low (4/981, 0.4%) in our study and lower than the 1.1% (95% CI 0.4%-2.1%) observed in a meta-analysis conducted in 20 RCTs [37]. This low rate of invalid results should re-assure women and providers about the quality of SS, although adequate health education remains crucial for obtaining reliable results [17].

A strength of our study was that it was conducted among regular screening users providing paired samples to comply with an agreement analysis. HPV SS was used in the context of primary screening, and women performed it at home without clinician assistance. Furthermore, we included a detailed analysis of Ct values, a measure that is not always assessed but can provide valuable insights into viral load and test sensitivity. All Ct values from HPV-positive samples, both from SS and CCS, were included. Only one CCS sample did not have a Ct value available, as Ct values are not automatically provided with the HPV result and must be extracted separately. Additionally, the use of two liquid-based cytology media available on the market (SurePath and ThinPrep) and the lack of differences between them contribute to the consistency of the study’s results.

However, a limitation of our study was limited number of cases of CIN2+ due to its focus on women undergoing routine screening, correlated by the low incidence rate in our country [1]. This limitation hindered our ability to conduct a comparative analysis of sensitivity and specificity between SS and CCS. Nonetheless, although our study was not primarily designed to assess the sensitivity and specificity of SS compared to CCS, but rather to evaluate the agreement of results, our data are consistent with the realities observed in clinical practice within the screening context in Spain [38] which strengthens the applicability of our results to real-world settings.

Despite the limitations, this study provides evidence supporting the use of SS for hrHPV testing in routine cervical cancer screening. The high agreement between SS and CCS, and the low rate of invalid results, suggest that SS can be a reliable alternative to CCS. This could increase coverage and accessibility of screening, especially in populations with barriers to attending a clinic.

Conclusions

In conclusion, our study confirmed that hrHPV testing on SS demonstrated a good concordance, in terms of agreement parameters, with CCS in population-based screening of women aged 30 to 65 years. However, SS tends to have a weaker amplification signal (higher Ct values in the PCR assays) compared to CCS. This highlights the need for further investigation into the implications of Ct values in SS through larger paired samples studies to determine whether these differences may indicate a potential loss of sensitivity.”

Reviewer 2 Report

Comments and Suggestions for Authors

It’s an interesting article about a current topic,the concordance of hrHPV testing results between self-collected(SS)and clinician-collected samples(CCS) for cervical cancer screening.This study is correctly structured and fluent,easy to follow.The summary is significant for the study that follows.Introduction is comprehensive and highlights two essential aspects;if is a similar performance between SS andCCS for hrHPVtesting and if there is lower sensitivity for SS method.Materials and methods chapter is correctly structured,with accuracy for the intended purpose.Results are clear presented, with very detailed and concise tables.Discussion is the key of the study with a detailed and critical analysis of the literature.The authors emphasize the weak and strong points of this study,showing the need for larger future randomized trials to clarify the debatable questions.Conclusions are clear and concise.The English language is fluent.References are  quite onerous and correctly cited.I do not find weak points of this article.

 

concordance of hrHPVtestig results between se

Author Response

Reviewer 2

It’s an interesting article about a current topic, the concordance of hrHPV testing results between self-collected (SS)and clinician-collected samples (CCS) for cervical cancer screening. This study is correctly structured and fluent, easy to follow. The summary is significant for the study that follows. Introduction is comprehensive and highlights two essential aspects; if is a similar performance between SS and CCS for hrHPV testing and if there is lower sensitivity for SS method. Materials and methods chapter is correctly structured, with accuracy for the intended purpose. Results are clear presented, with very detailed and concise tables. Discussion is the key of the study with a detailed and critical analysis of the literature. The authors emphasize the weak and strong points of this study, showing the need for larger future randomized trials to clarify the debatable questions. Conclusions are clear and concise. The English language is fluent. References are quite onerous and correctly cited. I do not find weak points of this article.

Answer: We thank the reviewer for their positive feedback. We are pleased to hear that the article’s structure, clarity, and fluency were appreciated, as well as the comprehensive presentation of the results and discussion.

Thank you again for your review.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Dear Authors, I read with great interest you manuscript entitled AGREEMENT BETWEEN HIGH-RISK HUMAN PAPILLOMAVIRUS TESTING IN PAIRED SELF-COLLECTED AND CLINICIAN-COLLECTED SAMPLES FROM CERVICAL CANCER SCREENING IN SPAIN.

Your work appears interesting and relevant, particularly considering the fact that self sampling will likely increase the uptake of this important screening in the future.

Few minor revisions are needed prior to the manuscript publication:

- the abstract claims that 981 women aged 30-65 years were included in the study, while based on Table 1 it appears that 18 out of these 981 participants were aged <30 years and 67 were >60 years (rather than 61-65 years); please clarify this point

- please specify if and possibly how participants were instructed on specimen self collection

- row 303-304: "...median viral Ct values was slightly higher for SS than for CCS 302 (32.9 vs 30.6 respectively, p=0.02) and among the samples of women under 50 years pared to women below 50 years (32.8 vs 30.0 respectively; p=0.02)..." I guess below should be substituted with "above"

- text font and dimension need to be uniformed across the manuscript

Author Response

Reviewer 3

Dear Authors, I read with great interest you manuscript entitled AGREEMENT BETWEEN HIGH-RISK HUMAN PAPILLOMAVIRUS TESTING IN PAIRED SELF-COLLECTED AND CLINICIAN-COLLECTED SAMPLES FROM CERVICAL CANCER SCREENING IN SPAIN.

Your work appears interesting and relevant, particularly considering the fact that self sampling will likely increase the uptake of this important screening in the future.

Few minor revisions are needed prior to the manuscript publication:

- the abstract claims that 981 women aged 30-65 years were included in the study, while based on Table 1 it appears that 18 out of these 981 participants were aged <30 years and 67 were >60 years (rather than 61-65 years); please clarify this point.

Answer: Thank you for your comment. We apologize for the confusion. Upon reviewing Table 1, we noticed the discrepancy regarding the age ranges. The 18 participants aged <30 years were indeed included in the study due to clinical discretion, as most of them were 29 years old or turned 30 during the year of the study. Regarding the participants over 60 years old, all of them were between 60-65 years, except for two women aged 67 and 68 years.

- please specify if and possibly how participants were instructed on specimen self collection

Answer: Thank you for your interest. At the beginning of the Materials and Methods section, we mention that this study is part of a randomized controlled trial conducted to evaluate the acceptability of self-sampling in women attending routine cervical cancer screening in Spain, and we reference the article where this trial is explained in more detail (https://doi.org/10.1016/j.ypmed.2023.107571). However, we have included a brief paragraph in this article to briefly explain how self-sampling was offered to the women.

“Briefly, all participants were not pregnant, with no history of cervical disease, and were not hysterectomized. Sociodemographic information was collected through a self-completed questionnaire, including the date and country of birth, nationality, educational level, data on previous history of cervical cancer screening among other information, and an acceptability questionnaire was completed when they self-collected a sample at home [17]. All participating women had a liquid-based cervical sample collected by a clinician during their screening visit. In addition, the clinician explained how to use the SS device, provided a leaflet with pictorial instructions, and gave them a SS kit to take home, along with instructions to return the sample to the health center one month after the initial visit.”.

 

- row 303-304: "...median viral Ct values was slightly higher for SS than for CCS 302 (32.9 vs 30.6 respectively, p=0.02) and among the samples of women under 50 years pared to women below 50 years (32.8 vs 30.0 respectively; p=0.02)..." I guess below should be substituted with "above"

Answer: Thank you for your comment. We appreciate your suggestion. The paragraph has been revised to improve clarity, and we have made the correction as follows:

"Among the samples of women under 50 years, SS showed higher median Ct values compared to CCS (32.8 vs 30.0 respectively; p=0.02). However, for women above 50 years, no significant difference was observed between SS and CCS (34.8 vs 33.6 respectively; p=0.57)."

 

- text font and dimension need to be uniformed across the manuscript

Answer: Thank you for your comment. We used the journal's provided template, which specifies the font and dimensions for different sections, including titles, tables, and text. We have ensured that all elements follow the template's guidelines.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for addressing the comments.