Cervical cancer is the third most common cancer in women worldwide, accounting for 8% of new female cancer cases1 and is the fourth most common cause of cancer death among women.1,2
In Latin America and the Caribbean, more than 25,000 women die annually of cervical cancer; approximately 6100 of these deaths occur in Mexico. Even though Mexico has had a cervical cancer national early detection program based on primary cytology for more than 36 years, this program has had many logistic problems and barriers to effective implementation.3–5
High-risk types of the human papillomavirus (HR-HPV) are the necessary cause of cervical neoplasia.6 Multiple studies have shown that screening for HR-HPV is much more sensitive but less specific than cervical cytology for the detection of cervical intraepithelial neoplasia (CIN) 3 and cancer. In addition, primary screening for HPV can lead to an extension of the screening interval because of its high negative predictive value.7–12 When compared with cervical cytology, HR-HPV tests have a lower subsequent risk of CIN3 or cancer, thus allowing a longer screening interval.12,13
In our first study, “The Mexican Cervical Cancer Screening Study” (MECCS I), we combined a self-collected upper vaginal sample with a high-risk HPV assay for the primary screen.14 The management of women testing positive for HR-HPV then becomes the critical issue. In most situations, this will require a triage test also referred to as a secondary screen. In MECCS I, after the primary screen, the triage test was acetic acid–aided visual inspection (VIA). To further reduce the number of HR-HPV–positive women requiring treatment, women initially HR-HPV positive and VIA negative had repeat HR-HPV tests in 1 year. Although utilizing VIA and retesting for HR-HPV at 12 months decreased the proportion of women requiring treatment (from 14.3% to 0.73%), the algorithm (which required as many as 4 visits) had a high rate of lost to follow-up (31.1%). In addition, if appropriate screen–positive women had cryotherapy without biopsy, the algorithm would have overtreated many women with HR-HPV or CIN1 and, in a few women, delayed the diagnosis of CIN3 and cancer.14
The objectives of MECCS II are (1) to insert the APTIMA HR-HPV assay into our algorithm and establish its characteristics when used with self-collected vaginal specimens and (2) determine the efficacy of cryotherapy in the treatment of HR-HPV–positive and VIA-positive and -negative women after VIA triage. This article addresses the first of the objectives.
The institutional review boards for human research subjects from the Cleveland Clinic, the State of Michoacán, and Mexican Federal Government (COFEPRIS) approved the study. The study flow diagram is presented in Figure 1. Nonpregnant women between the ages of 30 and 50 years, residing in the state of Michoacán, Mexico, without history of Pap smear screening or knowledge of their Pap results within the last 3 years, and with no history of hysterectomy and no history of prior pelvic radiation, were invited to participate. All the women who qualified based on inclusion criteria and agreed to participate signed an informed consent.
Study recruitment was done with the assistance of Desarrollo Integral de la Familia (DIF is a family welfare agency), Colegio Nacional de Educación Profesional Técnica (Conalep is a national college of professional technical education), and the Secretary of Health of the State of Michoacán. The recruitment methods included posted public community announcements, radio and television advertisement, local meetings within communities/villages led by DIF staff, and promotion through DIF offices. The districts in Michoacán, Mexico, that participated in the study were Pátzcuaro, Santa Clara de Cobre, Tzintzuntzan, Erongarícuaro, Quiroga, Zitácuaro, Ocampo, Juarez, Angangeo, Jungapeo, and Tuxpan. Transportation for the screening visit and recall of participants for the second visit was provided by local DIF authorities.
Women who were interested in participating and met eligibility criteria were given an oral presentation of the protocol. Participants then saw an instructional video in their native language that showed how to collect the vaginal sample using the fourth-generation Preventive Oncology International (POI)/National Institutes of Health self-sampler.15 Trained staff from DIF and Conalep obtained informed consent from the women and then collected demographic data that included age, marital status, and obstetric and gynecologic history. After obtaining the self-sample (unobserved), the women placed the brush of the self-sampler in a transport media (PreservCyt) for transfer to the Prueba Para La Vida laboratory in Morelia, Michoacán, Mexico, established during MECCS I project. After collection of the self-sample, the participants completed a questionnaire assessing their impressions of the self-sampler. Next, a nurse or a physician obtained 2 direct cervical samples in alternating order based on study ID number, using either a “broom” sampler (Rovers Cervex-Brush; Rovers Medical Devices, Oss, the Netherlands) placed in PreservCyt, or direct to Gen-Probe STM using the Gen-Probe cervical brush. This article is based on the data from APTIMA assay using the PreservCyt media (although both samples performed in a similar fashion; unpublished). In the laboratory in Morelia or our collaborating Hospital de Mujer in Morelia, the following procedures were accomplished:
- A ThinPrep slide was made from the direct PreservCyt sample. The unstained slides were then transported to Cleveland Clinic in Cleveland, OH, where they were stained and then read by the gynecologic cytopathology staff according to the Bethesda System.16
- In the laboratory in Morelia, 4 mL of the remaining ThinPrep sample was then removed for Hybrid Capture 2 (HC2), and 2 mL removed for the Gen-Probe APTIMA assay. Hybrid Capture 2 was done in Morelia according to the manufacturers’ recommended procedures, and the Gen-Probe APTIMA HR-HPV assays were done at Gen-Probe in San Diego, CA. All HPV testing was performed by technicians who had no knowledge of, or access to, cytology results.
Women who were HR-HPV negative on all assays (self and direct collected) with normal cytology were not evaluated further. Those women who tested positive for HR-HPV (on any assay) or had ≥atypical squamous cells of uncertain significance (ASCUS) on cervical cytology were recalled for a second visit. At the second visit, acetic acid 4% was applied to the cervix; VIA was performed after a 1-minute delay to determine (1) satisfactory for cryotherapy or (2) unsatisfactory for cryotherapy (defined as a high-grade lesion >3 quadrants unable to be covered by a 1.8-cm cryoprobe, or suspicion for cancer). Then 4% acetic acid was reapplied, and colposcopy performed according to the POI microbiopsy protocol of directed and random biopsies.17 All pathology specimens were processed in Mexico and read by a local pathologist and 2 gynecologic pathologists who traveled to Mexico from Cleveland Clinic. Immunostaining for p16 was done on all available CIN1 blocks, as well as all CIN2 and CIN3 specimens after transporting the specimens back to Cleveland Clinic.
According to protocol, cryotherapy was done for all women who were HC2 positive by self-sample or by direct sample, before knowing the results of the biopsies; unless on VIA, they were diagnosed as unsatisfactory for cryotherapy. Cryotherapy was done using a freeze-thaw-freeze technique, with the freeze cycle lasting 3 minutes or until a 5-mm frozen ring extended beyond the cryotherapy probe. All patients were treated using N20 as the gas. With the exception of 11 patients, all were treated using a Wallach LL100 cryotherapy system. The 11 patients treated with a different system (Leisegang) had longer freeze times to accommodate an apparent slower-freeze cycle.
The sensitivity and specificity for CIN3 or cancer of APTIMA were compared with HC2 with McNemar test. In addition, partial areas under the receiver operating characteristic (ROC) curves for the 2 assays were compared using the Roccomp test in STATA. Roccomp compares the equality of 2 or more ROC curves using the χ2 statistic. P < 0.05 was considered significant.
Between February 2009 and April 2009, 2522 women were screened. Four hundred seventy-three women with missing data were excluded from final analysis. (Missing data include 15 HC missing samples, 259 APTIMA missing samples, 137 missing ThinPrep Pap samples, 4 patients missing demographic information, and 58 missing biopsies including in this group 2 patients with grossly advance cervical cancer who were referred for further medical care and excluded from POI biopsy protocol.) Of the 2049 women with no missing data or unsatisfactory samples, the median age was 39.2 years (range, 30–50 years); 84.7% were married or living with a partner, 91.4% were nonsmokers, and 75% reported only 1 lifetime sexual partner.
Seven percent of the 2049 women reported no history of screening for cervical cancer; 14.4% had been screened more than 3 years ago, and 78.6% had been screened within the last 3 years.
Of the 2049 women with complete data, 7.7% had cervical cytology with ASCUS or worse (≥ASCUS); 1.8% had cytology of low-grade squamous intraepithelial neoplasia (LGSIL) or worse (≥low-grade squamous intraepithelial neoplasia), and 0.5% had cytology of high-grade squamous intraepithelial neoplasia, atypical glandular cells of uncertain significance, adenocarcinoma in situ, or cancer (≥high-grade squamous intraepithelial neoplasia). Five percent of the cervical cytology slides were unsatisfactory.
One tenth of 1% (2/2049) of the women included in the analysis had cancer, 0.68% (14/2,049 had CIN3, 1.2% (25/2,049) had CIN2, 6.6% (136/2,049) had CIN1 or HPV (+), and 91.3% (1872/2049) were negative.
Histology results available after the recall visit identified that 2.0% of patients recalled had ≥CIN2, and 0.78% had ≥CIN3. Eight patients were referred for further treatment. (Three patients were referred after VIA, 2 with signs suggestive of invasive cervical cancer, and 1 with unsatisfactory VIA due to anatomy difficult for cryo (later found to have CIN2). Five patients who were treated with cryotherapy were found to have CIN3 on endocervical curettage when the biopsies were read and then were referred for further evaluation and treatment.
Figure 1 illustrates patient population that participated in each step of the screening process. The DIF attempted to locate the 503 women with positive HPV test or with negative HPV and abnormal Pap smear (≥ASCUS) for a secondary screening visit and possible treatment. Thirty-four of 503 either were unable to be located or simply chose not to return. A total of 470 VIA evaluations (469 women and 1 patient with double cervix and vagina) and 291 cryotherapies were performed during the second clinic visit.
The sensitivity, specificity, and positive predictive value for ≥CIN3 of the 5 screening tests are presented in Table 1. The sensitivity for CIN3 or cancer of physician-collected endocervical specimens tested for HR-HPV by APTIMA (100% [16/16]) does not differ from that of physician-collected endocervical specimens tested for HR-HPV by HC2 (100% [16/16]; P = 1). Likewise, the sensitivity for CIN3 or cancer of self-collected vaginal specimens tested for HR-HPV by APTIMA (62.5% [10/16]) does not differ from that of the same specimens tested for HR-HPV by HC2 (62.5% [10/16]; P = 1). In contrast, the sensitivity for CIN3 or cancer of physician collected-specimens tested for HR-HPV by APTIMA or HC2 (100% [16/16]) exceeds that of self-collected vaginal specimens tested for HR-HPV by APTIMA or HC2 (62.5% [10/16]; P = 0.03) but does not exceed the sensitivity of cervical cytology with a cut point of ≥ASCUS (87.5% [14/16]; P = 0.5). The specificity of APTIMA exceeds that of HC2 for both the endocervical specimen (93.5% [1900/2033] vs 92.2% [1875/2033]; P = 0.008) and the vaginal specimen (93.0% [1890/2033] vs 90.5% [1840/2033]; P = 0.0001) and is similar to that of cervical cytology with cut point of ≥ASCUS (93.5% ([1900/2033] vs 94.1% [1914/2033]; P = 0.09). In addition, the specificity of the endocervical specimens exceeds that of the specificity for vaginal self-collected specimens for HC2 (92.2% [1875/2033] vs 90.4 [1840/2033]; P = 0.002) but does not differ for APTIMA (93.5% [1900/2033] vs 93.0% [1890/2033]; P = 0.44).
To further explore the differences between the 2 HR-HPV assays, we constructed ROC curves comparing physician-collected endocervical specimens tested for HR-HPV (Fig. 2) and those comparing self-collected vaginal specimens (Fig. 3). As shown in Figure 3, the partial area under the ROC curve from 1 − specificity of zero to 0.85 for diagnosis of CIN3 or cancer for APTIMA (0.1303) was greater than that of HC2 (0.1146, P = 0.015). As shown in Figure 3, the partial area under the ROC from 1 − specificity of zero to 0.85 for diagnosis of CIN3 or cancer for APTIMA (0.0590) was greater than that of HC2 (0.0879, P = 0.003).
Even though we currently have an effective prophylactic HPV vaccine, the only way the mortality associated with cervical cancer among women already infected with HPV virus will be reduced is through effective and sustainable cancer screening programs. The overall goal of our research program has focused on the design of a sensitive and specific screening algorithm able to be adaptable to all socioeconomic levels in the state of Michoacán, Mexico. To that end, we have explored a variety of screening and treatment technologies as well as the acquisition of specimens using self-collection instead of physician (healthcare worker)–collected samples. The current trial algorithm (a primary HR-HPV screen followed by VIA as secondary triage for cryotherapy treatment) used in MECCS II was based on the results of our prior work in Mexico (MECCS I).14 MECCS II used 3 cervical samples (2 collected directly from the cervix and 1 self-collected), APTIMA and HC2 for HR-HPV screening and ThinPrep for liquid-based cytology in the initial clinic visit.
MECCS II evaluate the characteristics of the APTIMA HR-HPV assay into our algorithm when used with direct and self-collected vaginal specimens. MECCS II demonstrated that APTIMA assay has superior specificity compared with HC2 in both endocervical specimens and vaginal self-collected specimens. Our results from endocervical specimens are consistent with previous work performed by our group during the Shenzhen Cervical Cancer Screening Trial I18. The Aptima human papilloma virus test was more specific and significantly more accurate by ROC curve comparison when compared with liquid-based cytologic examination and HC2 (P = 0.005).18 These findings are consistent with the results of multiple studies that have used direct endocervical collection including the FASE trial,19,20and work published by Ratnam et al21 in Canada and Clad et al22 in Germany. The low false-positive rate makes APTIMA suitable for use as a primary assay for detecting cervical disease in a screening setting.21,22
Our group has been studying the use of self-collection HPV screening for over a decade before this trial. The Shanxi Province Cervical Cancer Screening Study I and II demonstrated the potential for acceptable sensitivity and specificity for the detection of ≥CIN2 using HPV self-test.23Both studies identified sensitivities that range between 83% and 87.5% and specificities that range between 77.2% and 86%.23Based on the analysis of the data obtained, we concluded that self-sampling for HPV was less sensitive for ≥CIN2 than the direct test, but similar to liquid-based cytology.23–25 In SPOCCS III, we identified the discordance in results between self- and directly collected samples to be primarily the presence of HR-HPV solely in the vagina, unassociated with ≥CIN2 in the cervix.24 We developed a self-collection device based on the data generated by our prior studies (POI/NIH self-sampler) and tested versions 4 to 6 in MECCS II and the SHENCCAST II study.26 In SHENCCAST II, we conclusively demonstrated that a self-collected sample was equivalent to a physician-obtained specimen when tested using a polymerase chain reaction–based assay such as matrix-assisted laser desorption/ionization–time of flight 26 or linear array24 not dependent on the sampling device used. These data were obtained after the MECCS II trial was completed as we continued to explore self-collection algorithms for future consideration in primary cervical cancer screening. MECCS II presents the first comparison of HC2 and APTIMA assays using self-collected vaginal samples. The POI fourth-generation self-sampler demonstrated excellent ability to collect a highly cellular sample as had been previously hoped (data not reported), and the device had high acceptance among the participants based on the acceptability questionnaire performed as part of the initial visit.15 At the time of this project, we still believed the sampling device was the key issue, but as noted above, we have since demonstrated it is not.26
When compared with our previous studies in China, MECCS II identified a relatively low sensitivity for both assays (62.5%).15,23,25,26The sensitivity of this part of the study was affected in large part by participants’ noncompliance with the use of the self-sample secondary to a low understanding of the instructions of how to obtain an adequate vaginal sample with the use of POI self-sampler. Although we had oral instructions along with a video, these results made us realize the need for absolute simplicity of design along with the education of women for successful screening using self-collection. Fortunately, our subsequent research further clarified a simple collection brush will work fine matched with a proper assay for a self-collected sample.
We chose to use ≥CIN3 as an end point because of the recognized 50% regression rate of CIN2, making a ≥CIN3 diagnosis more reproducible and a superior surrogate for invasive disease.27,28In addition, the sensitivity of all the assays using a cut point of ≥CIN2 can be compromised because of the marked inflammation as was present in this population. Our pathologists were experienced with the reading of small biopsies based on the POI protocol but needed on many occasions to utilize p16 staining in CIN2 samples to arrive at a definitive diagnosis (most frequently due to severe inflammation). In fact, the sensitivities and specificities of CIN2 samples increased significantly (data not shown in current paper) when staining was used. No change in sensitivity or specificity was identified among ≥CIN3 samples after p16 staining.
Without focused efforts to increase the education and awareness among the communities, prevention programs will consistently be unable to reach their full potential. Development of strong partnerships between community groups and the health departments should be a priority to ensure successful results using the technologies that are now available.
1. Jemal A, Bray F, Center MM, et al.. Global cancer statistics. CA Cancer J Clin
. 2011; 61: 69–90.
2. Laara E, Day NE, Hakama M, et al.. Trends in mortality from cervical cancer in the Nordic countries: association with organized screening
. 1987; 1: 1247–1249.
3. Lazcano-Ponce EC, Castro R, Allen B, et al.. Barriers to early detection of cervical –uterine cancer in Mexico. J Womens Health
. 1999; 8: 399–408.
4. Murillo R, Almonte M, Pereira A, et al.. Cervical cancer screening
programs in Latin America and the Caribbean. Vaccine
. 2008; 26: L37–L48.
5. Lazcano-Ponce E, Palacio-Mejia LS, Allen-Leigh B, et al.. Decreasing cervical cancer mortality in Mexico: effect of Papanicolaou coverage, birthrate, and the importance of diagnostic validity of cytology. Cancer Epidemiol Biomarkers Prev
. 2008; 17: 2808–2817.
6. Walboomers JM, Jacobs MV, Manos MM, et al.. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol
. 1999; 189: 12–19.
7. Mayrand MH, Duarte-Franco E, Rodrigues I, et al.. Human papillomavirus DNA versus Papanicolaou screening
tests for cervical cancer. N Engl J Med
. 2007; 357: 1579–1588.
8. Bulkmans NW, Berkhof J, Rozendaal L, et al.. Human papillomavirus DNA testing for the detection of cervical intraepithelial neoplasia grade 3 and cancer: 5-year follow-up of a randomized controlled implementation trial. Lancet
. 2007; 370: 1764–1772.
9. Sankaranarayanan R, Nene BM, Shastri SS, et al.. HPV screening
for cervical cancer in rural India. N Engl J Med
. 2009; 360: 1385–1394.
10. Schiffman M, Herrero R, Hildesheim A, et al.. HPV
DNA testing in cervical cancer screening
: results from women in a high-risk province of Costa Rica. JAMA
. 2000; 283: 87–93.
11. Katki HA, Kinney WK, Fetterman B, et al.. Cervical cancer risk for women undergoing concurrent testing for human papillomavirus and cervical cytology: a population-based study in routine clinical practice. Lancet Oncol
. 2011; 12: 663–672.
12. Dillner J, Rebolj M, Birembaut P, et al.. Long term predictive values of cytology and human papillomavirus testing in cervical cancer screening
: joint European cohort study. BMJ
. 2008; 337: a1754.
13. Arbyn M, Sasieni P, Meijer CJ, et al.. Chapter 9: clinical applications of HPV
testing: a summary of meta-analyses. Vaccine
. 2006; 24 (suppl 3): S3/78–S89.
14. Belinson JL, Pretorius RG, Enerson C, et al.. The Mexican Cervical Cancer Screening
Trial: self-sampling for human papillomavirus with unaided visual inspection as a secondary screen. Int J Gynecol Cancer
. 2009; 19: 27–32.
15. Arriba LN, Enerson CL, Belinson S, et al.. Mexican Cervical Cancer Screening
Study II: acceptability of human papillomavirus self-sampler. Int J Gynecol Cancer
. 2010; 20: 1415–1423.
16. Solomon D, Davey D, Kurman R, et al.. The 2001 Bethesda System: terminology for reporting results of cervical cytology. JAMA
. 2002; 287: 2114–2119.
17. Belinson J, Qiao YL, Pretroius RG, et al.. Shanxi Province Cervical Cancer Screening
Study: a cross-sectional comparative trial of multiple techniques to detect cervical neoplasia. Gynecol Oncol
. 2001; 83: 439–444.
18. Wu R, Belinson SE, Du H, et al.. Human papillomavirus messenger RNA assay for cervical cancer screening
: the Shenzhen Cervical Cancer Screening
Trial I. Int J Gynecol Cancer
. 2010; 20: 1411–1414.
19. Monsonego J, Hudgens MG, Zerat L, et al.. Evaluation of oncogenic human papillomavirus RNA and DNA tests with liquid-based cytology in primary cervical cancer screening
: the FASE study. Int J Cancer
. 2011; 129: 691–701.
20. Monsonego J, Hudgens MG, Zerat L, et al.. Risk assessment and clinical impact of liquid-based cytology, oncogenic human papillomavirus (HPV
) DNA and mRNA testing in primary cervical cancer screening
(the FASE study). Gynecol Oncol
. 2012; 125: 175–180.
21. Ratnam S, Coutlee F, Fontaine D, et al.. APTIMA HPV
E6/E7 mRNA test is as sensitive as Hybrid Capture 2 assay but more specific at detecting cervical precancer and cancer. J Clin Microbiol
. 2011; 49: 557–564.
22. Clad A, Reuschenbach M, Weinschenk J, et al.. Performance of the APTIMA high-risk human papillomavirus mRNA assay in a referral population in comparison with Hybrid Capture 2 and cytology. J Clin Microbiol
. 2011; 49: 1071–1076.
23. Belinson JL, Qiao YL, Pretorius RG, et al.. Shanxi Province cervical cancer screening
study II: self-sampling for high-risk human papillomavirus compared to direct sampling for human papillomavirus and liquid based cervical cytology. Int J Gynecol Cancer
. 2003; 13: 819–826.
24. Belinson JL, Hu S, Niyazi M, et al.. Prevalence of type-specific human papillomavirus in endocervical, upper and lower vaginal, perineal and vaginal self-collected specimens: implications for vaginal self-collection. Int J Cancer
. 2010; 127: 1151–1157.
25. Belinson JL, Wu R, Belinson SE, et al.. A population-based clinical trial comparing endocervical high-risk HPV
testing using Hybrid Capture 2 and Cervista from the SHENCCAST II Study. Am J Clin Pathol
. 2011; 135: 790–795.
26. Belinson JL, Du H, Yang B, et al.. Improved sensitivity of vaginal self-collection and high-risk human papillomavirus testing. Int J Cancer
. 2012; 130: 1855–1860.
27. Castle PE, Schiffman M, Wheeler CM, et al.. Evidence for frequent regression of cervical intraepithelial neoplasia-grade 2. Obstet Gynecol
. 2009; 113: 18–25.
28. Syriänen KJ. Spontaneous evolution of intraepithelial lesions according to the grade and type of the implicated human papillomavirus (HPV
). Eur J Obstet Gynecol Reprod Biol
. 1996; 65: 45–53.
29. Denny L, Kuhn L, Hu CC, et al.. Human papillomavirus-based cervical cancer prevention: long-term results of a randomized screening
trial. J Natl Cancer Inst
. 2010; 102: 1557–1567.
Keywords:Copyright © 2013 by IGCS and ESGO
HPV; Screening; Cryotherapy