Cross-sectional Study of Patient- and Physician- Collected Cervical Cytology and Human Papillomavirus

Transcription

1 Cross-sectional Study of Patient- and Physician- Collected Cervical Cytology and Human Papillomavirus Francisco Garcia, MD, MPH, Bel Barker, MD, Carlos Santos, MD, Elena Mendez Brown, MD, Thomas Nuño, MA, Anna Giuliano, PhD, and John Davis, MD OBJECTIVE: To compare the performance of patient- and physician-obtained cytology and human papillomavirus (HPV) testing for the detection of high-grade cervical intraepithelial neoplasia. METHODS: A cross-sectional study was performed involving 334 women seen at three colposcopy clinics (Tucson, Arizona; Hermosillo, Mexico; and Lima, Peru). All women used a cytobrush to self-collect specimens for cervical cytology and HPV testing. Subjects subsequently underwent physician collection for cytology and HPV, followed by a complete colposcopic evaluation with directed biopsy. Cytology was processed using thin-layer technology, and HPV was determined using the polymerase chain reaction technique. Test performance characteristics were determined using the histopathologic diagnosis as the reference standard and designating high-grade cervical intraepithelial neoplasia as clinically significant disease for the purpose of the analysis. RESULTS: The sensitivity of patient-collected cytology was significantly lower (55.0%) and specificity was significantly better (84.1%) than those of physician-directed sampling (85.2% and 73.4%, respectively). Patient-collected HPV had significantly lower sensitivity (49.0%) than physician sampling (82.2%), although specificity did not significantly differ. From the Departments of Obstetrics & Gynecology and Pathology, Arizona Cancer Center, and Hispanic Center of Excellence, University of Arizona Health Sciences Center, Tucson, Arizona; Departmento de Ginecologia, Instituto Nacional de Enfermedades Neoplasicas, Lima, Peru; and Departamento de Citologia, Instituto de Seguridad Social al Servicio de los Trabajadores del Estado ISSSTE, Hermosillo, Sonora, Mexico. The authors express their sincere appreciation to the women in Tucson, Hermosillo, and Lima who participated in this study. They also acknowledge the contributions of the technical team (S. Vanzzini, B. Monaco, T. Fernandez, J. Wein, M. Abramsen, L. Vaught, K. Harrigill, C. Kavanagh, and K. Hatch) and the statistical review provided by J. Ranger-Moore and S. Carvajal. The Population Council, Latin America Office, provided seed funding for this study. Thin-layer liquid cytology collection materials were provided at no cost by Cytyc Corp. (Boxborough, Massachusetts). This study was partially supported by grants from the National Institutes of Health (no. CA82715) and the Health Resource Services Administration (no ). BB s participation was partially underwritten by a career development grant from the University of Arizona. CONCLUSION: Patient collection is a feasible although inferior alternative to physician-collected cervical cytology and HPV testing. (Obstet Gynecol 2003;102: by The American College of Obstetricians and Gynecologists.) Cervical carcinoma is the leading cause of gynecologic cancer-related morbidity and mortality worldwide. 1 The prevalence of cervical cancer in the developing world is 59.4 per 100,000, and in Latin America it represents the second most common malignancy among women. 1 For 2003 in the United States it is estimated that there will be 12,200 new cases and 4100 cervical cancer deaths. 2 The relatively prolonged and well-characterized premalignant phase of this disease allows for the detection and treatment of precancerous lesions, with a potentially dramatic effect on both morbidity and mortality. The five-fold reduction in invasive cervical cancer incidence in North America over the last 4 decades is largely attributed to the effectiveness of cytology-based screening programs. 3 Current recommendations include regular cervical cytologic screening in reproductive-aged women to detect cervical intraepithelial neoplasia (CIN) before the development of invasive disease. 4 6 Although cytology-based screening programs have been successful in the developed world, their performance in low resource settings, including Latin America, has been disappointing. 7 Most frequently this has been primarily due to a failure to screen large segments of the population, especially among women at greatest risk. 7,8 In the United States, cervical cancer continues to be a significant burden for poor women and those who have difficulty accessing screening services. 9 Although many factors contribute to this problem, a major barrier is the need for a speculum examination provided by an experienced practitioner. An ideal screening method would be relatively inexpensive, would be easily performed by the individual or a rudimentarily trained practitioner, and would correlate highly with cervical disease on histological evaluation. 266 VOL. 102, NO. 2, AUGUST /03/$ by The American College of Obstetricians and Gynecologists. Published by Elsevier. doi: /s (03)

2 Our purpose was to assess the feasibility of patientperformed sampling for cervical cytology and oncogenic human papillomavirus (HPV) as a screening modality for premalignant cervical disease in a population of women at risk for cervical cancer and its precursors. We compared the performance of patient sampling with physician-obtained cytology and HPV testing, using colposcopy with directed biopsy as the reference standard. MATERIALS AND METHODS Between January 1999 and June 2000 we performed a cross-sectional study of patient- and physician-performed sampling for cervical cytology and oncogenic HPV. A total of 334 women seeking care at three colposcopy clinics (University of Arizona Health Sciences Center, Tucson, Arizona; Instituto Nacional de Enfermedades Neoplasicas, Lima, Peru; and ISSSTE, Hermosillo, Mexico) were enrolled in the study. The University of Arizona Human Subjects Committee approved this study, as did the institutional authorities at the participating clinics in Peru and Mexico. Eligible subjects were all 18 years or older, scheduled for colposcopic examination, and able to provide written informed consent. Women with a history of hysterectomy, current pregnancy, or a history of vaginal trauma or laceration were excluded from the study. Subjects were enrolled sequentially at each site and were instructed individually and verbally on the collection of the vaginal specimen. Specifically, women were asked to separate their labia and gently insert a soft endocervical collection brush (Cytobrush Plus; Medscand, Malmo, Sweden) into the vagina approximately 5 to 6 cm and rotate the brush five times with their dominant hand. The patient then placed the brush in a container filled with methanol buffer solution (PreservCyt Solution; Cytyc, Boxborough, MA), the end of the brush was broken off, and the container was closed tightly. All self-collections took place in a private examination room and without the assistance of medical personnel. All subjects underwent their scheduled examination after the collection was completed. The gynecologic examination included the placement of a speculum and visualization of the cervix, assessment for vaginal trauma, and collection of cytology using a plastic Ayre spatula and endocervical brush. These collection devices were subsequently rinsed in the methanol buffer solution per the manufacturer s instructions (Cytyc). Once this was completed, all subjects underwent a thorough, standard colposcopic examination with directed cervical biopsies as described by Hatch. 10 Colposcopic examination results were classified as negative, low-grade CIN (grade 1); high-grade CIN (grade 2 or 3); or carcinoma. All specimens were labeled with a unique subject identification number and the date of the collection. It should be noted that any colposcopically visible lesion was biopsied in the US and Peruvian study sites, whereas only lesions colposcopically consistent with suspected CIN grade 2 or 3 and carcinoma were biopsied in the Mexican study site in concordance with the local standard of care. Colposcopists with a minimum of 6 years of experience conducted all examinations. All thin-layer cervical cytology preparations were performed maintaining strict adherence to the manufacturer s recommendations (ThinPrep Pap Test; Cytyc) at the University of Arizona, Department of Pathology. All slides were screened by one of two experienced cytotechnologists, and all abnormal slides and a random sampling of 10% of normal ones were reviewed by the experienced supervising cytopathologist (JD). Cytologic diagnoses were reported using the terminology of the Bethesda System. 11,12 Each specimen was fractionated for HPV testing before the preparation of the cytology slides to minimize the theoretic possibility of cross contamination. Human papillomavirus deoxyribonucleic acid (DNA) analyses of exfoliated cervical cell samples were conducted using polymerase chain reaction (PCR) testing. Genomic DNA was extracted following standard techniques. Specimens were tested for the presence of HPV by amplifying 5 L of the DNA extracts with the PGMY09/11 L1 consensus primer system 13 and Ampli- Taq Gold polymerase (Perkin-Elmer, Foster City, CA). The samples were amplified using the GeneAmp PCR System 9700 (Perkin-Elmer). Human papillomavirus genotyping was conducted using the reverse line blot method 14 on all samples that were positive by PCR. This detection method used the HPV L1 consensus PCR products labeled with biotin to detect 27 HPV types. The HPV genotype strip contained 29 probe lines, detecting 27 individual HPV genotypes and two concentrations of the -globin control probe. All reagents were kindly provided by Roche Molecular Systems Inc. (Alameda, CA). The following types were detected: 6, 11, 16, 18, 26, 31, 33, 35, 39, 40, 42, 45, 51 59, 66, 68, 73, 82, 83, and 84. Histopathologic analyses with standard hematoxylin eosin preparations were performed at each institution. All slides were forwarded for evaluation to a central gynecologic pathologist (JD) who was blinded to the cytologic diagnosis, colposcopic findings, or prior pathologic interpretation. Histopathologic diagnoses were reported using the terms normal, low-grade squamous intraepithelial neoplasia (CIN grade 1), high-grade squamous intraepithelial neoplasia (CIN grade 2 or 3), and cancer. 15 VOL. 102, NO. 2, AUGUST 2003 Garcia et al Self-collected Cytology and HPV 267

3 For the purpose of this analysis, the reference standard diagnosis was determined from the histological interpretation of the biopsy specimen. Colposcopic impression was used as a surrogate in nine cases from Mexico that were colposcopically suspicious for low-grade CIN but not biopsied. High-grade disease (CIN grade 2 or 3) or carcinoma constituted clinically significant disease for the purpose of this analysis. For the cytologic analysis a diagnosis of atypical squamous cells of undetermined significance (ASCUS) or higher was considered a positive screening test. Oncogenic HPV status was characterized as present or absent. Test performance characteristics were calculated for the patient- and physiciancollected cytology and HPV samples and compared with the reference standard diagnosis. Statistical analyses were performed using SAS System for Windows 8.1 (SAS Institute Inc., Cary, NC). Oneand two-way frequency counts, sensitivity, specificity, positive predictive values, and negative predictive values with respective exact 95% confidence intervals (CIs) were calculated as needed. In addition, Pearson and likelihood ratio 2 tests of association, analysis of variance, the Fisher exact test, and the McNemar test were done and statistics calculated. The predetermined level of significance was set at P.05 for all statistical comparisons. RESULTS Three hundred thirty-four women were enrolled into the study: 108 from Tucson, Arizona; 100 from Lima, Peru; and 126 from Hermosillo, Mexico. The mean age of women enrolled in the study was 36.9 (range 18 67, standard deviation [SD] 11.3), and the mean parity was 2.3 (range 0 10, SD 1.9). An analysis of variance revealed that mean age did not statistically differ between the three study sites (P.05), although parity was significantly higher (2.7) among women from Peru (P.01) than among women at the other sites. Colposcopic diagnoses were available for all subjects. Biopsy specimens were collected from 75.5% of study participants, including all women with a colposcopic impression consistent with CIN grade 2 or 3 or carcinoma. In accordance with the local standard of care, ten of 126 subjects in the Mexican study site with a colposcopic diagnosis of CIN grade 1 were not biopsied. Five had normal results and five had low-grade squamous intraepithelial lesions on cytology done at colposcopy. Table 1 summarizes the reference standard diagnoses by study site. The proportion of subjects in each diagnostic reference category significantly differed (P.001) by study site. The Mexican study site contributed a lower proportion of subjects with CIN grade 2 or 3 and cancer Table 1. Reference Standard Diagnoses (and Percentages) by Country United States Mexico Peru Negative 40 (37) 79 (63) 24 (24) Low-grade disease 42 (39) 28 (22) 20 (20) (CIN grade 1) High-grade disease 19 (18) 19 (15) 45 (45) (CIN grade 2/3) Cancer* 7 (6) 0 (0) 11 (11) Total CIN cervical intraepithelial neoplasia. 2 P.001. * Includes microinvasive and overtly invasive carcinoma. (15.1% and 0%, respectively) than the US (17.6% and 6.5%) and Peruvian (45.0% and 11.0%) sites. Additionally, all participants were evaluated for vaginal lacerations after self-sampling, during the colposcopic examination. Only one woman in the Arizona group was noted to have a small superficial laceration that did not require repair. A comparison of the quality of cytology is summarized in Table 2. The proportion of satisfactory but limited cytology was significantly higher for the self-collected group (P.01). Consequently, the proportion of entirely satisfactory cytology was significantly higher for the physician-collected group. Unevaluable, unsatisfactory cytology did not significantly differ between the two groups. A secondary analysis was performed using the new definitions of cytologic adequacy recently published from the National Institutes of Health Consensus Conference. 16 Under this reporting scheme adequacy is reported as a dichotomous outcome, either satisfactory or unsatisfactory. Using these definitions, there was no difference in the proportion of satisfactory cytology when comparing physician (329 of 334 [98.5%]) and Table 2. Quality of Cytology Using the Bethesda 1992 Definitions Definitions Satisfactory (without diagnostic limitations) Satisfactory but limited by absent endocervical cells Patient collected Physician collected % 84% % 14% Satisfactory but limited by other 8 3 2% 1% Unsatisfactory 3 5 1% 2% Total Likelihood ratio 2 test, P.012. Percentages do not add up to 100% because of rounding. 268 Garcia et al Self-collected Cytology and HPV OBSTETRICS & GYNECOLOGY

4 Table 3. Test Performance Characteristics (and Exact 95% Confidence Intervals) for Self- and Physician-Collected Cytology and Human Papillomavirus Testing Self-collected cytology Physiciancollected cytology Self-collected HPV Physiciancollected HPV Self-collected combination* Sensitivity 55% (45.3, 64.8) 85% (78.2, 92.1) 49% (39.2, 58.8) 82% (74.7, 89.6) 75% (66.5, 83.5) Specificity 84% (79.4, 88.8) 73% (67.7, 79.1) 73% (51.7, 79.1) 67% (60.9, 73.1) 66% (60.2, 72.3) PPV 60% (49.8, 69.8) 58% (50.2, 66.1) 44% (34.9, 53.4) 52% (44.1, 59.6) 49% (40.8, 56.5) NPV 81% (76.4, 86.3) 92% (88.0, 95.9) 77% (71.5, 82.6) 90% (85.1, 94.2) 86% (81.1, 91.2) HPV human papillomavirus; PPV positive predictive value; NPV negative predictive value. Abnormal cytology for the purpose of this analysis is any diagnosis of atypical squamous cells of undetermined significance or greater. Abnormal HPV testing represents the detection of any oncogenic HPV type. Assumes a reference standard diagnosis of abnormal for any high-grade intraepithelial abnormality (cervical intraepithelial grade 2 or 3) or cancer. * Presence of an abnormal cytology and/or positive oncogenic HPV on the same collection. patient (330 of 333 [99.1%]) collection techniques (P.83). Sensitivity and specificity were calculated for patientand physician-collected cytology. All calculations were performed assuming that clinically significant disease was limited to CIN grade 2 or 3 or cancer on histopathology. Given that cytology was used as a screening test, the threshold for abnormality was set as ASCUS or higher. With that in mind, the sensitivity of patientcollected cytology was 55.0% with a specificity of 84.1% for the diagnosis of CIN grade 2 or 3 or cancer. This compares to physician-collected cytology with a sensitivity of 85.2% and specificity of 73.4%. The positive predictive value of abnormal cytology (ASCUS or higher) did not significantly differ between the two groups. The negative predictive value was significantly better for the physician-collected cytology than the patient-collected (91.9% and 81.3%, respectively). Test performance characteristics are summarized in Table 3. A comparison of oncogenic HPV positivity, by reference diagnosis and patient or physician collection, is summarized in Table 4. For each diagnostic category clinician sampling yielded a significantly higher proportion of positive results. The sensitivity of patient-collected oncogenic HPV (49.0%) was significantly lower than that of the physician-collected specimen (82.2%) for diagnosing CIN grade 2 or 3 or cancer. By contrast, specificity and positive predictive values did not significantly differ. Negative predictive value was significantly lower for the patient-collected oncogenic HPV test than the physician-collected one (77.2% and 89.7%, respectively). When either cytology (abnormality threshold of ASCUS or greater) or HPV was positive, the sensitivity of the self-collected specimens was estimated at 75.0%, with a specificity of 66.2%. For the clinician collection the sensitivity was significantly higher (95.1%); however, specificity was 56.8% (overlapping 95% CIs with selfcollection). Positive predictive values were similar for patient and clinician sampling (48.7% and 48.7%). Negative predictive values were significantly lower for patient-collected specimens (86.1%) than for physician-collected ones (96.4%). There was good agreement between patient- and physician-collected specimens. Comparing self- and clinician-collected cytology specimens, a of 0.50 (95% CI 0.39, 0.58) was calculated, with an overall concordance of 76.0%. Similarly for oncogenic HPV, the was estimated at 0.51 (95% CI 0.43, 0.61), with an overall concordance of 77%. Although HPV testing was performed on all specimens, there was a trend toward higher rates of no recoverable DNA for analysis in 2.7% of patient-collected specimens (nine of 334), compared with 0.9% of the physician-collected (three of 333 specimens) (P.10) (data not shown). Table 4. Oncogenic Human Papillomavirus Positivity by Collection Technique and Reference Diagnosis Patient collected Physician collected McNemar 2 Normal 25/144 (17%) 27/144 (19%) 0.22 (NS) Low-Grade Disease (CIN grade 1) 37/90 (41%) 50/90 (56%) 8.05 (P.01) High-Grade Disease (CIN grade 2/3) 40/82 (49%) 62/83 (81%) (P.001) Cancer* 9/18 (50%) 16/18 (89%) 5.44 (P.05) Total 111/ /335 NS nonsignificant; CIN cervical intraepithelial neoplasia. * Includes microinvasive and overtly invasive squamous carcinoma. VOL. 102, NO. 2, AUGUST 2003 Garcia et al Self-collected Cytology and HPV 269

5 DISCUSSION Eighty percent of cervical cancer occurs in developing world settings, where it is responsible for 200, ,000 deaths each year. 1 By comparison, last year in the United States there were approximately 4900 deaths from cervical cancer. 2 This dramatic disparity has been attributed to the ubiquity of cytology-based screening programs in developed nations. 7 Efforts to implement cytologic screening programs in the developing world have met a variety of barriers, including inadequate material resources and supplies, deficient infrastructure, numerically insufficient and/or poorly trained providers, and low patient acceptability due to cultural and educational barriers. The lack of a pelvic examination performed by an adequately trained examiner may be among the most important obstacles to many women accessing cervical cancer screening. Self-collection addresses some of these considerations and may make cervical cancer screening more accessible to hard-to-reach women and women who have poor access to health care services. Self-collection overcomes barriers to screening associated with physical access to clinical settings, transportation, and, perhaps most importantly, the requirement for a pelvic examination by a skilled provider. Before the implementation of self-collection for a cervical cancer screening program we must establish the efficacy of self-screening for the detection of clinically significant premalignant cervical disease. Although this has been largely addressed for HPV testing, the utility of self-collected cervical cytology has not been adequately evaluated. Our study demonstrates the feasibility and limitations of patient-collected cervical cytology for cervical cancer screening. Patient collection of cervicovaginal specimens yields material adequate for cytologic analysis and the performance of HPV testing. The major limitation is the sampling of the endocervical canal, which is deficient in as many as 66% of self-collections, compared with 14% of clinician-collected specimens. Indeed, what is remarkable is that approximately 34% of self-collections had adequate endocervical material for cytologic evaluation. This would suggest that both ectocervical and endocervical cells are exfoliated and may circulate within the vagina, where they are available for sampling. The presence of endocervical cells is no longer a criterion for adequacy under the new Bethesda guidelines 16 ; however, their absence may at least partially explain the observation of lower rates of recovery of oncogenic HPV across diagnostic categories. In this study, selfcollected cytology has a high degree of concordance relative to clinician-collected cytology but lacks the sensitivity and negative predictive value necessary in a primary screening tool. Screening with self-collected HPV has been studied in a variety of settings with mixed results. Gravitt et al 17 found an excellent correlation between patient-collected Dacron swabs and physician-collected cervical swabs in a population of patients with known cervical cancer. Sellors et al 18 also compared self-collected HPV to physician-collected HPV for the detection of CIN grade 2 or 3. They investigated a variety of patient collection techniques and found vaginal collection of HPV by Dacron swab to be superior to vulvar or urine sampling. As in the present study, physician-collected sampling yielded significantly superior results for the detection of CIN grade 2 or 3. Harper et al 19 published the results of a randomized controlled trial of patient and clinician sampling for HPV in a group of 103 women undergoing colposcopy. They specifically examined a variety of sampling devices including one and two Dacron swabs as well as a tampon. Among six cases of CIN grade 2 or 3, there were no significant differences between patient- and clinician-collected specimens, regardless of technique used. Most importantly, they document consistently high levels of patient acceptability with self-sampling regardless of specific technique. Although the number of patients with clinically significant disease in this study was small, Harper et al s work confirms a favorable response to the concept of self-sampling by its potential users. Other investigators have recently documented equally high levels of user acceptability among women in central Mexico. 20 Recent publications have confirmed acceptable test performance characteristics for self-collected oncogenic HPV testing using the commercially available DNA hybrid capture technique in large population screening trials in South Africa and China. 21,22 For both of these studies sensitivity was reported to be higher (66% and 83%, respectively) and positive predictive value lower (23% and 21%, respectively) than those of the current study. 21,22 This may be explained in part by differences in the types of populations being tested (screening versus referral population) as well as the specifics of the collection techniques. More recently, authors in Brazil working with a screening population have reported poor correlation of oncogenic HPV testing with CIN grade 2 or 3 and cancer. 23 These authors used cotton-tip applicators for sampling and PCR for HPV testing. It is unclear whether such important differences in test performance characteristics can be ascribed to differences in HPV detection method, collection device, or sampling method. The most important limitation of our study was its conduct in colposcopy clinic populations with a high 270 Garcia et al Self-collected Cytology and HPV OBSTETRICS & GYNECOLOGY

6 prevalence of clinically significant cervical disease. Test performance will likely be different in the setting of population-based screening. Specifically, the predictive value of testing is likely to be lower in a screening population with a lower prevalence of disease. Human papillomavirus testing is an important and established adjunct to the management of ASCUS, especially when collected in a medium amenable to reflex viral testing. 16,24 Recent work has also suggested that HPV positivity, among cytology-negative women, has a high predictive value for the detection of women at risk for future CIN grade 2 or 3. 25,26 Age group targeted screening strategies employing self-sampling for HPV, followed by visual inspection with acetic acid or cytology, are currently under evaluation. Such secondary screening may be simplified if the same specimen used for HPV testing is also used for cytology. Moreover, the recovery of cellular material for cytology serves as objective confirmation that self-sampling was performed. In the future, recovery of cellular material from self-collected cervical specimens may lend itself to a new generation of molecular oncoprotein testing that would improve the performance of cervical cytology as it is currently practiced. The tremendous burden of cervical cancer and its devastating impact on the lives and productivity of women in developing world settings and in large pockets of predominantly poor communities within our own country demand a dramatic rethinking of our approach to this disease entity. Self-sampling for HPV testing and cytology is an attempt to overcome some of the very significant barriers that keep women from entering the screening process. These technologies are not by themselves proposed as a solution, but they are part of the spectrum of options that should be evaluated in an effort to serve hard-to-reach populations of women. Our study suggests that self-collection with an endocervical brush is an alternative to Dacron or cotton swab, cervical lavage, and tampon for the purpose of HPV testing and that self-collection techniques deserve further evaluation. REFERENCES 1. Pisani P, Bray F, Parkin DM. Estimates of the world-wide prevalence of cancer for 25 sites in the adult population. Int J Cancer 2002;97: American Cancer Society. Cancer Facts and Figures Atlanta: ACS, Saslow D, Runowicz CD, Solomon D, Moscicki A, Smith RA. American Cancer Society guideline for the early detection of cervical neoplasia and cancer. CA Cancer J Clin 2002;52: U.S. Preventive Services Task Force, Office of Disease Prevention and Health Promotion. Guide to clinical preventive services: Report of the U.S. Preventive Services Task Force. Washington: U.S. Department of Health and Human Services, Office of Public Health and Science, Office of Disease Prevention and Health Promotion, American College of Obstetricians and Gynecologists. Primary and preventive care: Periodic assessments. ACOG committee opinion no Washington: American College of Obstetricians and Gynecologists, The American Cancer Society guideline for the cancerrelated checkup: An update. Atlanta: American Cancer Society, Sankaranarayanan R, Budukh AM, Rajkumar R. Effective screening programmes for cervical cancer in low- and middle-income developing countries. Bull World Health Organ 2001;79: Lazcano-Ponce EC, Moss S, Alonso de Ruiz P, Salmeron Castro J, Hernandez Avila M. Cervical cancer screening in developing countries: Why is it ineffective? The case of Mexico. Arch Med Res 1999;30: Sawaya GF, Grimes DA. New technologies in cervical cytology screening: A word of caution. Obstet Gynecol 1999;94: Hatch KD. Handbook of colposcopy: Diagnosis and treatment of lower genital tract neoplasia and HPV infections. Boston: Little, Brown, The 1988 Bethesda System for reporting cervical/vaginal cytological diagnoses. National Cancer Institute Workshop. JAMA 1989;262: Kurman RJ, Henson DE, Herbst AL, Noller KL, Schiffman MH. Interim guidelines for management of abnormal cervical cytology. The 1992 National Cancer Institute Workshop. JAMA 1994;271: Gravitt PE, Peyton CL, Alessi TQ, Wheeler CM, Coutlee F, Hildesheim A, et al. Improved amplification of genital human papillomaviruses. J Clin Microbiol 2000;38: Gravitt PE, Peyton CL, Apple RJ, Wheeler CM. Genotyping of 27 human papillomavirus types by using L1 consensus PCR products by a single-hybridization, reverse line blot detection method. J Clin Microbiol 1998;36: Wright TC Jr. Precancerous lesions of the cervix. In: Kurman R, ed. Blaustein s pathology of the female genital tract. New York: Springer-Verlag, 1994: Solomon D, Davey D, Kurman R, Moriarty A, O Connor D, Prey M, et al. The 2001 Bethesda System: Terminology for reporting results of cervical cytology. JAMA 2002;287: Gravitt PE, Lacey JV Jr, Brinton LA, Barnes WA, Kornegay JR, Greenberg MD, et al. Evaluation of selfcollected cervicovaginal cell samples for human papillomavirus testing by polymerase chain reaction. Cancer Epidemiol Biomarkers Prev 2001;10: VOL. 102, NO. 2, AUGUST 2003 Garcia et al Self-collected Cytology and HPV 271

7 18. Sellors JW, Lorincz AT, Mahony JB, Mielzynska I, Lytwyn A, Roth P, et al. Comparison of self-collected vaginal, vulvar and urine samples with physician-collected cervical samples for human papillomavirus testing to detect highgrade squamous intraepithelial lesions. CMAJ 2000;163: Harper DM, Noll WW, Belloni DR, Cole BF. Randomized clinical trial of PCR-determined human papillomavirus detection methods: Self-sampling versus cliniciandirected biologic concordance and women s preferences. Am J Obstet Gynecol 2002;186: Dzuba IG, Diaz EY, Allen B, Leonard YF, Lazcano Ponce EC, Shah KV, et al. The acceptability of self-collected samples for HPV testing vs. the pap test as alternatives in cervical cancer screening. J Womens Health Gend Based Med 2002;11: Wright TC Jr, Denny L, Kuhn L, Pollack A, Lorincz A. HPV DNA testing of self-collected vaginal samples compared with cytologic screening to detect cervical cancer. JAMA 2000;283: Belinson J, Qiao YL, Pretorius R, Zhang WH, Elson P, Li L, et al. Shanxi Province Cervical Cancer Screening Study: A cross-sectional comparative trial of multiple techniques to detect cervical neoplasia. Gynecol Oncol 2001;83: Lorenzato FR, Singer A, Ho L, Santos LC, Batista Rde L, Lubambo TM, et al. Human papillomavirus detection for cervical cancer prevention with polymerase chain reaction in self-collected samples. Am J Obstet Gynecol 2002;186: Cuzick J, Sasieni P, Davies P, Adams J, Normand C, Frater A, et al. A systematic review of the role of human papilloma virus (HPV) testing within a cervical screening programme: Summary and conclusions. Br J Cancer 2000;83: Castle PE, Wacholder S, Sherman ME, Lorincz AT, Glass AG, Scott DR, et al. Absolute risk of a subsequent abnormal pap among oncogenic human papillomavirus DNApositive, cytologically negative women. Cancer 2002;95: Sherman ME, Lorincz AT, Scott DR, Wacholder S, Castle PE, Glass AG, et al. Baseline cytology, human papillomavirus testing, and risk for cervical neoplasia: A 10-year cohort analysis. J Natl Cancer Inst 2003;95: Address reprint requests to: Francisco Garcia, MD, MPH, University of Arizona Health Sciences Center, 1501 North Campbell Avenue, PO Box , Tucson, AZ ; fcisco@u.arizona.edu. Received January 17, Received in revised form March 21, Accepted March 27, Garcia et al Self-collected Cytology and HPV OBSTETRICS & GYNECOLOGY