Cervical Cancer Screening with DNA-HPV Testing and Precancerous Lesions Detection: A Brazilian Population-based Demonstration Study

Objective  To evaluate the rates of precancerous lesions, colposcopy referral, and positive predictive value (PPV) by age groups of a population-based screening with DNA-HPV testing. Methods  The present demonstration study compared 16,384 HPV tests performed in the first 30 months of the program with 19,992 women tested in the cytology screening. The colposcopy referral rate and PPV for CIN2+ and CIN3+ by age group and screening program were compared. The statistical analysis used the chi-squared test and odds ratio (OR) with 95% confidence interval (95%CI). Results  The HPV tests were 3.26% positive for HPV16-HPV18 and 9.92% positive for 12 other HPVs with a 3.7 times higher colposcopy referral rate than the cytology program, which had 1.68% abnormalities. Human Papillomavirus testing detected 103 CIN2, 89 CIN3, and one AIS, compared with 24 CIN2 and 54 CIN3 detected by cytology ( p  < 0.0001). The age group between 25 and 29 years old screened by HPV testing had 2.4 to 3.0 times more positivity, 13.0% colposcopy referral, twice more than women aged 30 to 39 years old (7.7%; p  < 0.0001), and detected 20 CIN3 and 3 early-stage cancer versus 9 CIN3 and no cancer by cytology screening (CIN3 OR= 2.10; 95%CI: 0.91–5.25; p  = 0.043). The PPV of colposcopy for CIN2+ ranged from 29.5 to 41.0% in the HPV testing program. Conclusion  There was a significant increase in detections of cervix precancerous lesions in a short period of screening with HPV testing. In women < 30 years old, the HPV testing exhibited more positivity, high colposcopy referral rate, similar colposcopy PPV to older women, and more detection of HSIL and early-stage cervical cancer.


Introduction
In Brazil, 16,590 new cases of cervical cancer are expected annually and a 6.17 per 100,000 women mortality rate, which means that 1 woman dies of cervical cancer every 90 minutes. 1 A national screening program started in the 1980s and expanded significantly after that. The current Brazilian Guidelines for Cervical Cancer Screening (2016) recommends conventional cytology (Pap test) every 3 years in women aged 25 to 64 years old. 2 The Brazilian Unified Public Health System (SUS, in the Portuguese acronym) registers a yearly number of cytology tests performed sufficient to cover 80% of all targeted women. However, the estimated coverage does not exceed 30%, with a significant rate of excess tests as they are performed in women outside the age range and at an inadequate periodicity. 3 The mortality rates have not reduced, and > 60% of cervical cancer cases are diagnosed in advanced stages. 4,5 The World Health Organization (WHO), supported by the consolidated evidence since 2013, recommends that the new programs replace the cytology-based strategy with the Human Papillomavirus (HPV) test-based screening, starting at 30 years old. 6 Several studies and scientific associations indicate the use of HPV tests with caution in the age group up to 29 years old, due to possible limitations in achieving the expected efficacy. However, the age group between 25 and 29 years old is a frequent target of cervical cancer screening programs. [7][8][9] Our research group, seeking to produce evidence to support improving the current scenario, started a population-based demonstration study in 2017. The objective was to raise epidemiological, cost-benefit, and life gain indicators to support the transition of the current Brazilian cytology-based opportunistic screening to a DNA-HPV test-based organized screening program starting at 25 years old.
The "PREVENTIVO" program (PREvention of HPV Viruses in ENTire Indaiatuba by Vaccination and Organization of the screening) is based on primary DNA-HPV testing screening. It was implemented in Indaiatuba, a medium-sized city (250,000 inhabitants) in São Paulo State, Brazil. 10 Indaiatuba is included in the macro-region of Campinas, attended by the University of Campinas. It has health care facilities networked with information from computerized systems and individual digital records. A cost-effectiveness analysis based on real-life costs was published in 2021, showing that HPV testing every 5 years had lower costs than cytology every 3 years. 11 In sequence, just after 50% completion of the 1 st round in high coverage (> 80%), there was an impressive detection of 21 prevalent cancers, two-thirds as microinvasive carcinomas, a washing effect anticipating the diagnosis in 10 years. 12 The current paper presents the detection of precancerous lesions, particularly in women < 30 years old, during the first round of the ongoing population-based screening program with DNA-HPV testing, just before the pandemic started.

Methods
The present study considered information from the first 30 months of the population-based demonstration study, an HPV test-based screening program, from October 2017 to March 2020, just before the impact of the pandemic. The aim was to verify the detection rate of cervical intraepithelial neoplasia grade 2 or 3 and worse (CIN2 þ , CIN3 þ ), the colposcopy referral rate, and the positive predictive value (PPV) of a colposcopy by age groups, comparing with the previous information from the cytology screening.
The HPV test-based screening protocol from the PREVEN-TIVO program was previously published. 10 Summarily, the program retrieves women between 25 and 64 years old, relying on the municipal public health system (50% of the total population). 13 Therefore, considering the population estimated of 70,573 women for 2020, the target population was 35,000 women. 14 There is no information about HPV vaccination in the screened women. However, the Brazilian National Vaccination Program started in 2014 for girls aged up to 13 years old, so women will reach the minimum age of 25 for screening in 2025. Therefore, we consider that close to 100% of the women studied were unvaccinated.

HPV Test and Flowchart of the Screening Program
Based on the previously published validation study, the Cobas HPV Test (Roche Molecular Systems, Pleasanton, CA, USA) was adopted as the screening test. 15 This test simultaneously provides individual results on the highest risk genotypes -HPV 16 and HPV 18-and aggregate result (positive or negative) on the twelve other (12OT) high-risk (hr) HPV genotypes (types 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68). All genotypes were tested simultaneously. Screening samples were collected at primary healthcare facilities, and the tests were performed at a regional laboratory. Clinical procedures according to the first result of the DNA-HPV test were: 1. Negative DNA-HPV test: advised to return after 5 years to repeat the test; 2. Positive HPV16 and/or HPV18 test: referred for colposcopy and biopsy if needed; 3. Positive DNA-HPV test for the 12OT hr-HPV: liquid-based cytology (LBC) was performed on the same sample. If a cytological abnormality (Atypical Squamous Cell of Undetermined Significance [ASC-US] or worse) was detected, the woman was referred for colposcopy. When the cytology was negative, the woman was advised to return after 12 months to repeat the DNA-HPV test.
The flagged women were referred to a colposcopy clinic. Excision of the transformation zone (ETZ) under local anesthesia was performed when needed. Histopathologic results were negative, low-grade (LSIL or CIN1) and high-grade intraepithelial lesions (HSIL or CIN2, CIN3, and adenocarcinoma in situ [AIS]), and cancer. Cases suspicious or diagnosed with cervical cancer or in a more complex clinical situation were referred to the regional gynecological cancer center at the Women's Hospital of the Universidade de Campinas, where oncological surgeries, radiotherapy, and chemotherapy are available.
The municipal health secretary managed the screening program at the central level. The PREVENTIVO program used algorithms according to study protocols, which did not allow a sample collection of a woman outside the established flowchart, avoiding testing twice the same woman or out of the recommendation. The research team from the Universidade de Campinas had access to all data and worked as a quality control surveillance system of the progress of the program. The present study obtained information about age, colposcopy, and histology results from all women with abnormal screening tests. The researchers do not interfere with the management or care of the screening program. Therefore, the level of organization of the PREVENTIVO program helps to achieve complete information about the procedures and results of women with a positive screening test. Fourteen women moved out of the city and were discharged by the program, following the regulatory recommendations and with adequate referral to the new city.

Comparison Group: Cytology-based Screening
As a historical reference, information was obtained from the previous screening program with conventional cytology in Indaiatuba, performed in the period from October 2014 to March 2017, 30 months just before the new program started. The cytology results and information of the women were obtained from the regional Cytology Laboratory at the Universidade de Campinas, where 100% of tests from the SUS were read. The cytologic diagnoses followed the Bethesda Reporting System, 16 and some follow-up information are missing due to the lack of central regulatory coordination. Only one cytology test was considered per woman, the one with the worst diagnosis. The cytology program lacks information about the number of negative colposcopies and the total of colposcopies performed; the only available information was about positive colposcopies with histologic results from biopsy or ETZ performed (335 colposcopy referrals with 112 known diagnoses).
After the end of all regular procedures, the definitive diagnosis was considered the worst grade of the histological evaluation of tissue obtained from colposcopy-directed cervical biopsy or ETZ (CIN2þ and CIN3 þ ). All CIN2 diagnoses in the PREVENTIVO program had an immunohistochemical confirmation with automatized biomarker p16INK4a detection (CINtec Histology; Ventana Medical Systems Inc, Tucson, AZ, USA).
The colposcopy referral rate and the CIN2þ and CIN3þ detection rates by age group were calculated in both screening programs with an odds ratio (OR) with a 95% confidence interval (CI). For the DNA-HPV test screening program, the number of colposcopies performed as well as the positive predictive value (PPV) of the colposcopy to detect CIN2þ or CIN3þ were calculated. The CIN2-3 detection rate by the positive screening test in the age group between 25 and 29 years old was calculated and compared with the age group between 30 and 64 years. The statistical analysis was performed with the chi-squared test using StatsDirect statistical software 3.0 (England, www.statsdirect.com). P-values < 0.05 were considered significant.
The Ethics Committee of the University of Campinas approved the present study (number 1045580, May 1, 2015; CAAE: 43815315.9.0000.5404). The mayor of Indaiatuba sanctioned a law instituting the HPV test as the standard for screening in 2017, which replaced conventional cytology in public healthcare. 17

Results
A total of 16,384 DNA-HPV tests in the PREVENTIVO program and 19,992 conventional cytology tests in the previous cytology screening were performed, considering 30 months between each program. The test outcomes by screening program are shown in ►Table 1. There were 86.84% negative DNA-HPV tests, 3.26% positive by HPV16 and/or HPV18, and 9.92% positive for 12OT hr-HPV that performed LBC at the same sample. The cytology program exhibited 1.68% abnor-mal tests and 0.26% suggesting HSIL (CIN2-3). The colposcopy referral rate was 3.7 times higher in the DNA-HPV test program: 6.26 (1,026/16,384) versus 1.68 (335/19,992) in the cytology program (p < 0.0001). The PREVENTIVO program had a compliance of 78% (788/1,012) of colposcopies performed, considering data until May 2021, while this information is unknown regarding the previous cytology program. The detection of CIN2þ was 1.31% in the PREVEN-TIVO program and 0.45% in the cytology program (rate ¼ 2.9). The number of women with any lesion was 405 CIN1þ detected by DNA-HPV testing screening and 106 CIN1þ cases in the cytology program, which means 14 times more LSIL (p < 0.0001) and 2 to 5 times more HSIL (p < 0.0001) (►Table 1).
Excluding 21 cervical cancers detected in the PREVEN-TIVO program and 12 cancer cases detected in the cytology program, the DNA-HPV screening detected more precursor lesions, 103 CIN2, 89 CIN3, and one AIS, compared with 24 CIN2 and 54 CIN3 detected in the cytology screening program (p < 0.0001). The HPV16 colposcopy referral rate was 2.48, 39.5% (405/1,026) of all colposcopies indicated, and was 48% higher than the rate of colposcopy referrals for the entire cytology program. The 12OT hr-HPV positive test plus ►Table 2 compares the programs regarding the colposcopy referral rate by age group. The highest colposcopy referral rates were associated with younger women in both programs. The age group between 25 and 29 years old had a 13.0% colposcopy referral rate, twice more than that of women aged between 30 and 39 years old (7.7%; p < 0.0001). In a similar analysis for the cytology program, the age group between 25 and 29 years old had a 3.0% colposcopy referral rate, almost twice more than that of women between 30 and 39 years old (1.7%; p < 0.001). Comparing the programs, the DNA-HPV test screening referred more colposcopies in all age groups, with similar ORs for the age groups between 25 and 29 years old (OR 4.80; 95% CI: 3.59-6.51) and between 30 and 39 years old (OR 4.81; 95% CI: 3.70-6.32). The cytology screening program tested more women outside the target age group: 16.2% (n ¼ 3,249) up to 24 years old with a 3.0 to 3.7% of colposcopy referral rate and eight CIN3 detected (no cancer); and 4.8% (n ¼ 995) of women ! 65 years old, with no abnormalities detected (►Tables 2 and 3).
The cervical lesions detected by age group in each program are shown in ►Table 3. The PREVENTIVO program screened 2,117 women aged between 25 and 29 years old with HPV test, detected 32 CIN2, 20 CIN3, and 3 cervical cancer cases: (1) microinvasive squamous-cell carcinoma, (1) microinvasive adenocarcinoma, (1) invasive adenocarcinoma Stage IB1. Comparatively, the results of 1,992 women     CIN 3, among 7 colposcopies performed in women 24 years old. No HSIL was detected in 7 colposcopies performed in women ! 65 years old. The age group between 25 and 29 years old had 2.4 to 3.0 times more HPV testing positive than women between 30 and 64 years old, with an OR of 2.1 to 3.0 for CIN2-3 lesions diagnosed according to the HPV type detected (►Table 6). There was no difference in the proportion of CIN2-3 detected by HPV type and age group among positive tests, ranging from 23.3 to 27.1%.

Discussion
The present article describes the first 30 months of a screening program with primary HPV testing, organized and implemented into routine practice, outside the research setting, compared with a previous program based on conventional cytology. The HPV-based program performed fewer screening tests than the cytology screening with high age compliance, detecting 2.9 times more CIN2 þ , which means 193 CIN2-3 lesions compared with 78 cases detected in the previous cytology screening. These achievements in unvaccinated women were associated with a 3.7 times higher colposcopy referral rate for the new program. In summary, there was 13.2% (2,156/16,384) of positive HPV testing, with 6.26% (1,026/16,384) of colposcopy referral, and 1.31% of CIN2þ detected.
The detection of CIN2-3 is the primary goal of screening programs since 20 to 50% of them have the potential to progress to invasive lesions. [18][19][20] This more significant number of HSIL detected is associated with the higher sensitivity of the HPV-DNA test, 9,15 and the expansion of the population coverage, with a 90% projection of the SUS target population, as previously published. 12 The compliance with the flow-chart of the program was high, with 78% of all colposcopies indicated already performed, which can be considered a milestone of the program. In a similar implementation screening study in Argentina (2011), with 49,000 HPV tests performed between 2012 and 2014, the researchers described a 74.6% of colposcopy compliance rate. 21 It is important to note that most women screened with the HPV test had been participating in the previous screening with cytology, which is standard in the country. Although the numbers of lesions detected were high, a significant reduction may be expected in the next round of the program, including the colposcopy referral rate, as already demonstrated in other countries with ongoing programs. [22][23][24] From the higher colposcopy referral rate observed in the demonstration program, 51.9% (532/1026) were indicated by HPV16 and/or 18 positive tests. According to the program guideline, in these cases, the colposcopy is performed without any cytology information, which can be considered an additional challenge for the colposcopists. Nevertheless, the colposcopies performed in the PREVENTIVO program showed higher CIN3 PPV for HPV16 and/or 18 positive tests (19.3%) compared with the cases from the 12OT hr-HPV positive with abnormal cytology (12.5%). There were five times more CIN2, and twice more CIN3 detected in the HPV testing compared with the cytology-based screening.
Although the cytological result can help the colposcopic evaluation and even partially reduce the colposcopy referral, 25,26 colposcopy for HPV16/18 positive test was performed without cytology. This decision is based on the higher risk for HSIL-or cancer-related, the ability to enable risk stratification for women aged 25 to 29 years old, 15 and the program flowchart simplification, avoiding issues related to coexisting two different tests, the conventional cytology and HPV test with LBC. The CIN2-3 detection rate in women Another concern of the study was the screening performance of the HPV test in the age group between 25 and 29 years old, considered a novelty at the beginning of the PREVENTIVO program. This age group exhibited 22.3% positive screening tests (473/2,117) and a colposcopy referral rate of 13.0%. It was almost double the 7.7% for the age group between 30 and 39 years old, and a PPV for CIN2þ of 34.2%. Although the cytology program exhibited a similar double of colposcopy referred to younger women (3.0% versus 1.7% for between 30 and 39 years old), there was 5 times less. When comparing the programs, the OR was equivalent between the age groups between 25 and 29 and between 30 and 39 years old. Similarly, a recently published population-based UK study evaluating the screening performance with HPV testing in unvaccinated women reported 26.9% of positive tests and 10.4% of directly referred to colposcopy for the age group between 24 and 29 years old. 26 A recent review of the French program with primary HPV testing stated that lesions in young women such as CIN2 have a high proportion of regression and that they could dispense an immediate colposcopy and wait for a repeat HPV test in 12 months. 24 Nevertheless, these results are influenced by the population prevalence of HSIL and cancer and, shortly, by the proportion of women adequately vaccinated against HPV. Our study proposed a single flowchart for all age groups. Although there was a high colposcopy referral rate, the HPV test in the age group between 25 and 29 years old demonstrated a higher performance, detecting 52 CIN2-3 and 3 cervical cancers at early stages, compared with 16 precursor lesions and no cancer detected by the cytology screening.
The PREVENTIVO program exhibited a colposcopy PPV of 37.1% for CIN2þ and 19.2% for CIN3þ for all age groups, slightly lower than that reported for the British study with 43 and 27%, respectively, the last using cytology triage to colposcopy referral for all HPV positive tests. 26 The present study is the first population-based demonstration program to replace conventional cytology with a DNA-HPV test for primary screening in a real-life scenario in the Brazilian public health system. The program structure and its results add evidence to the previously published information about coverage, program compliance, earlystage cervical cancer detection, 12 and cost-effectiveness. 11 It supports the adoption of the HPV test for a new national screening program.
The limitation of the present study is the use of a historical cohort as a comparison. There was a significant lack of detailed information about the previous cytology-based screening program, such as information about the management of positive tests. However, comparing the colposcopy referral rate and the cases with histologic results available from cervix procedures was possible. Another limitation was that the PREVENTIVO program considered only the SUS users, which comprise 50% of the women population.
The strength of the present study was the organization of the PREVENTIVO screening program in implementation providing access to crucial information for surveillance, not available in the opportunistic programs. The new program applies processes such as a simplified flowchart, and a computerized central management system. It is being performed in the same health units, by the same staff, and in the same population as the previous program, which highlights the results achieved, indicating that they are related to the process of the new program.
The results presented are preliminary but the best possible without the influence of the pandemic that started in Brazil in March 2020. The effect of the pandemic on the program will be the subject of an ongoing detailed analysis.
All results reported from the present demonstration study are highly impactful, even with the short program running time. The question remains on how to expand this action to other regions of Brazil, where the current situation could be worse. As the PREVENTIVO program used all the existing structures for the cytological program and simplified the work process, we believe the program can be replicated in all urban areas of the country. The main challenge to its implementation is the initial budgetary impact of the cost of HPV testing, developing a system to integrate the health units, and creating a digital record with information from the population to manage and monitor the program. The first round of the PREVENTIVO program will finish in late 2022 and additional rounds will be necessary to confirm the findings. Keeping the program under monitoring can serve as a sentinel study for further regions starting their new program, assessing any additional actions needed, and establishing an effective and long-lasting program.

Conclusion
There was a significant increase in detections of cervical cancer precursor lesions in a short period after the shift from opportunistic screening with conventional cytology to an organized program with HPV DNA test. In addition, women < 30 years old screened by HPV testing exhibited more positive tests and a higher colposcopy referral rate. However, the PPV of colposcopy was similar to that of older women, with more HSIL and early-stage cervical cancer cases detected.

Contributions
Each author has participated actively and sufficiently in the present work, and all had final approval of the manuscript version being submitted. Teixeira J. C. contributed to the conception and design, data collection and analysis, interpretation of data, and writing of the article; Vale D. B. contributed to the conception and design, interpretation of data, and writing of the article; Disgaciatti M. G contributed to data collection, interpretation of data, and writing of the article; Campos C. S. contributed to data collection and writing of the article; Bragança J. B. contributed to the interpretation of data and writing of the article; Zeferino L. C. contributed to the conception and design, interpretation of data, and critical review of the article.