Human Papillomavirus Testing and Conventional Pap Smear Cytology as Optional Screening Tools of Women at Different Risks for Cervical Cancer in the Countries of the Former Soviet Union.

Stina Syrjanen, DDS, PhD,1 Irena P. Shabalova, MD, PhD, MIAC,2’3 Nicolaj Petrovichev, MD, ScD,2 Vladimir P. Kozachenko, MD, ScD,2 Tatjana Zakharova, MD, PhD,2 Julia Pajanidi, MD, PhD,2 Jurij I. Podistov, MD, PhD,2 Galina Chemeris, MD, PhD,2 Larisa G. Sozaeva, MD, PhD,3 Elena V. Lipova, MD, PhD,3 Irena Tsidaeva, MD,4 Olga G. Ivanchenko, MD,4 Alia A. Pshepurko, MD,4 Sergej Zakharenko, MD,5 Raisa Nerovjna, MD,6 Ludmila B. Kljukina, MD, ScD,7 Oksana A. Erokhina, MD,7 Marina F. Branovskaja, MD, PhD,8 Maritta Nikitina, MD, PhD,9 Valerija Grunberga, MD,9 Alexandr Grunberg, MD,9 Anna Juschenko, MD,9 Piero Tosi, MD, PhD,10 Marcella Cintorino, MD, PhD,10 Rosa Santopietro, MD, PhD,10 and Kari J. Syrjanen, MD, PhD, FIAC10 1 Department of Oral Pathology, Institute of Dentistry, and MediCity Research Laboratory, University of Turku, Turku, Finland; 2N.N. Blokhin Cancer Research Centre of Russian Academy of Medical Sciences (RAMS)-, Moscow, Russia; 3Russian Academy of Post-Graduate Medical Education, Moscow, Russia; 4Novgorod Clinical Regional Hospital, Centralized Cytology Laboratory, Novgorod, Russia; ^Department of Gynaecology, Novgorod Municipal Dermatovenereological Dispensary, Novgorod, Russia; 6Department of Gynaecology, Novgorod Female Consultative Outpatient Hospital, Novgorod, Russia; 7Research Institute of Oncology and Medical Radiology, Republican Centre of Clinical Cytology, Minsk, Belarus; ^Department of Gynaecology and Obstetrics, Minsk State Medical Institute, Minsk, Belarus; 9Department of Gynaecology & Laboratory of Cytology, Latvian Cancer Centre, Riga, Latvia; and ^Department of Pathology, University of Siena, Siena, Italy

Reprint requests to: Stina Syrjanen, MD, Institute of Dentistry, University of Turku, Lemminkaisenkatu 2, 20520 Turku, Finland. E-mail: stisyr. utu.fi

Abstract:

Objectives. Human papillomavirus (HPV) infection is a sexually transmitted disease (STD) and the single most important etiological agent of cervical cancer. In parallel with the increase of STDs and because of the lack of any organized cancer screening in the new independent states of the former Soviet Union, the incidence and mortality rates of cervical cancer are rapidly rising. This is the first report from an ongoing European Commission-funded (INCO-Copernicus Program) cross-sectional and cohort study (focused on the key issues of this major health problem in the new independent states) analyzing the performance of the HPV DNA (Hybrid Capture II) test as a potential screening tool for cervical cancer in these countries.

Materials and Methods. A series of 3,175 women (screening, gynecological, or STD patients) from six clinics in Russia, Belarus, and Latvia received routine cytology and HPV testing with Hybrid Capture II (HCII). All women with HPV-positive results or abnormalities in cytology were subjected to colpos-copy and biopsy. The sensitivity, specificity, receiver operating characteristics, as well as positive (PPV) and negative predicting values (NPV), were determined for HCII and quality-controlled cytology in detecting significant pathology (cervical intraepithelial neoplasia [CIN] 3 and cancer).

Results. Significant pathology was strongly associated with high-grade cytology (odds ratio [OR] = 8.5; 95% confidence interval [Cl] = 4.1-17.8; chi-square, p < .0001). Pap smear-cytology detected high-grade lesions with a sensitivity of 64.0% (44.8-83.2), specificity of 89.1% (84.5-93.7), PPV of 44.4% (28.8-61.0), and NPV of 94.8% (91.2-98.4). Of the 3,086 samples analyzed by HCII, 33.0% were positive for on-cogenic HPV types, with a wide variation (from 23% to 45%) between the three patient groups (p < .0001). The presence of high-grade cytology was significantly associated with HCII positivity at all cutoff levels (OR = 14.4; 95% Cl = 8.4-24.5; chi-square, p < .0001; 1 pg/mL threshold). In the receiver operating characteristics curve, the HCII cutoff point most closely balancing sensitivity (83.1%) and specificity (75.6%) was 2 pg/mL. The presence of high-grade histology was associated with HCII positivity (cutoff 1 pg/mL; OR = 4.8; 95% Cl = 0.7-34.2; p = .047). At the cutoff (1 pg/mL), sensitivity of the HCII tesjt was 96.6% (90.0-100), specificity was 15.9% (10.6-, 21.2), PPVwas 15.1% (9.9-20.3), and NPV was 96.8% (90.3-100). Changing the cutoff significantly affected sensitivity at 20 pg/mL and NPV at 500 pg/mL.

Conclusions. HCII assay is a sensitive tool in detecting significant pathology, but less specific than the Pap test. A negative HCII test practically precludes high-grade CIN (NPV >95%). Because the performance characteristics of the HCII test depend on the prevalence of HPV and CIN in the study population, the cost-benefit issues in different settings will be the limiting factor for the application of this test as a screening tool.

Key Words: Pap smear, Hybrid Capture II assay, receiver operating characteristics curve, CIN, screening

General agreement exists that the declining incidence and mortality rates witnessed in most of the developed countries during the past four decades are mainly attributed to the adoption of increased screening practices using the Pap smear [1-4]. This is best illustrated by data from the Nordic countries, where organized screening has resulted in a 69.6% reduction in the cancer incidence from 14.8/105 in 1963 to 2.8/105 in 1992 in Finland [5, 6]. Cervical cancer is the only human malignancy in which incidence and mortality have been reduced by medical intervention [1-7]. Unfortunately, such highly effective organized screening programs exist in only a few countries [3, 4, 7].

On the global scale, the new independent states (NIS) of the former Soviet Union belong among the countries with an intermediate risk of cervical cancer. In Russia,’ the 1998 data report 11,937 new cervical cancer cases, with an incidence of 15.3/105, and 6,078 women died of cervical carcinoma. These figures are of similar magnitude in Latvia, with 189 new cases (13.9/105), but somewhat lower in Belarus, where 783 new cases (10.6/105) were registered in 1993 [8]. Compared with the official 1990 World Health Organization data (Table 1), this trend seems to be steadily increasing in all three NIS countries [9]. In many developed countries, a similar trend has been observed since the mid-1980s [1, 2, 4, 5], despite a gradual increase in screenings. This has been explained by increased exposure to the known risk factors of cervical cancer [2, 4, 10].

Table 1: Incidence (in 1990) of Cervical Cancer in the Three NIS Partner Countries (Russia, Belarus, and Latvia) and in the Two EC Partner Countries (Italy and Finland)

ASIR, Age-Standardized Incidence Rate; EC, European Commission; NIS, new independent states. Incidence per 100,000 women. Source: Globocan 1999.

Several characteristics of sexual behavior are consistently associated with an increased risk for cervical cancer and its precursors, suggesting that cervical cancer and its precursors behave like a sexually transmitted disease (STD) [11, 12]. The most important characteristics are as follows: 1) multiple sexual partners; 2) promiscuous sexual partners; and 3) early onset of sexual activity [11-13]. These characteristics of sexual behavior only represent surrogate markers of a sexually transmitted infectious agent, human papillomavirus (HPV), which is the single most important etiological factor for cervical cancer [11, 12]. Understanding the basic biology of HPV, cervical cancer, and its precursors is a prerequisite for effective disease control [2, 4, 5, 7, 14, 15]. Technology is available to take preventive measures against cancer by tracing its etiological agent, a sexually transmitted virus, which makes cervical cancer unique among all hurnan malignancies [11, 12]. Until now, a variety of molecular diagnostic tests (polymerase chain reaction (PCR), Hybrid Capture assay) have been tested as potential tools to target the screening directly to HPV, instead of cancer precursors detectable by a Pap smear [11, 12, 16-20].

All NIS countries have witnessed an alarming explosion of STDs since the collapse of the Soviet Union and the upheaval of its health care system [21]. Not unexpectedly, an increasing trend in the incidence and mortality rates of cervical cancer has been reported in these countries during the 1990s [8, 9], which prompted us to design the present project (funded by the European Commission [EC]) for the following: 1) to assess the prevalence of genital HPV infections, cervical intraepi-" thelial neoplasia (CIN), and cervical cancer in these countries; 2) to characterize the risk groups for this disease; and 3) to assess the performance of optional diagnostic methods (including HPV tests) to find out the most cost-effective tools for cervical cancer screening in these low-resource settings.

The present article is the first report from this 3-year (cross-sectional and cohort) study, describing the performance characteristics of Hybrid Capture II (HCII) (Digene, Silver Springs, MD) assay and conventional Pap smear cytology as cervical cancer screening tools in the low-resource settings of three NIS countries (Russia, Belarus, and Latvia), using colposcopic biopsy as the "gold standard".

MATERIALS AND METHODS

Patients and Study Design

The subjects of this combined cross-sectional and cohort study (run since 1998 as an EC-funded joint project [INCO-Copernicus Program]) consisted of consecutive women attending six different outpatient clinics in three NIS countries. This study focused on three target populations of women (with different risks for HPV infections, CIN, and cervical cancer) in these countries: 1) women participating in locally organized screening for cervical cancer (SCR patients); 2) those attending regular gynecology clinics with different indications (GYN patients); and 3) patients examined and treated at STD clinics (STD patients). During the 12-month period, a total of 3,175 women completed their first clinical visit and were subjected to examinations as detailed below. The mean age of the women was 32.7 (.10.9 SD) years (median = 31.0; range = 15-85 years). All women were informed of the aims of the study (verbally and in writing) and gave their written consent to participate (while completing the questionnaire). Only 128 of 3,175 women (attending the first visit) failed to consent to Pap smear screening, and 89 refused sampling using the HCII test.

The study design is described in the flowchart in Figure 1. The first 12 months were used for recruitment of a cohort of women with HPV and/or CIN, detected among the consecutive women attending the clinics in Moscow and Novgorod (Russia), Minsk (Belarus), and Riga (Latvia). All women had three tests performed: 1) Pap smears (ARM I); 2) PCR (ARM II); and 3) HCII test (ARM III). According to ARM I, all women with cyto-logical abnormalities consistent with HPV-CIN were referred for colposcopy. Women with equivocal (class 2 or atypical squamous cells of undetermined significance [ASCUS]) smears were retested with a repeated Pap smear at 6 months, and referral for colposcopy was made in patients with of a persistent abnormalities. All colposcopic lesions with an atypical transformation zone (TZ) were subjected to directed punch biopsies. The results of the punch biopsies were used as the "gold standard" of HPV-CIN diagnosis, to which the performances of the other tests were compared.

The PCR and HCII (ARM II and III) tests represent alternative diagnostic tools implicated in cervical cancer screening (Figure 1). In the present study, cervical swabs were analyzed for HPV by using the second generation HCII assay. The results of the PCR amplification with consensus primers (confirmed by type-specific PCR) will be reported separately. All women with HCII-positive (and PCR-positive) results were underwent colposcopy and"biopsy confirmation. Depending on the severity of the lesion, these women were allocated into two optional groups: 1) those to be prospectively followed up; and 2) those subjected to prompt treatment. Accordingly, women with low-grade lesions (HPV infection without GIN, i.e., HPV-NCIN and HPV-CIN 1) will be followed up without treatment, whereas those with high-grade lesions (HPV-CIN 2 or higher) will be subjected to treatment by the conventional methods (con-ization or large loop electroexcision of the transformation zone).

Figure 1: Flowchart of Patient Examination, Treatment, and Follow-up

Methods

Pap Smears. All consecutive women attending the clinics were subjected to cervico-vaginal Pap smears, interpreted for HPV and CIN according to conventional methods [12, 21]. The primary screening and interpretation of the smears were done in the participating NTS/ countries of Central Europe (CCE) centers. As an external quality-control (QC) measure, all Pap smears were rescreened by two International Academy of Cytology (lAC)-certified cytotechnologists and interpreted by one cytopathologist (fellow of International Academy of Cytology) in Finland, using jointly agreed terminology (modified Papanicolaou classification) to classify the smears.

Colposcopy. Colposcopic examination was performed using modern instrumentation acquired by all NIS centers as a part of this project. Facilities for video-colposcopy and digital imaging exist in one of the clinics (RAMS, Moscow). All examinations were done by practicing gynecologists with long-term experience (and prior training in their clinics) in colposcopy. Both acetic acid and iodine application were used, and the colpo-scopic patterns were classified using the 1990 colposcopy nomenclature of the International Federation of Cervical Pathology and Colposcopy [22].

Directed Punch Biopsy. Directed punch biopsies were taken from all colposcopic abnormalities, according to routine procedures. The biopsies were fixed in 10% neutral formalin and processed into hematoxylin and eosin sections for light microscopy. On histological grading of the lesions, CIN nomenclature was used. The concomitant presence of HPV infection was recorded using the commonly accepted morphological criteria [11,12,23].

Epidetniological Questionnaire. At the first visit, all of the women who agreed to participate in the study were subjected to detailed inquiry concerning the implicated or suspected risk factors of cervical cancer (CIN and HPV), including detailed reproductive history, sexual behavior, and smoking habits [24].

Follow-Up

The women who presented with low-grade lesions (HPV-NCIN or HPV-CIN 1) underwent prospective follow-up for a minimum of 24 months. These women attend the clinic at 6-month intervals and were tested using colposcopy, Pap smear, and punch biopsy (in case of suspected progression). In addition, repeated samples for HPV DNA analysis by HCII and PCR were taken at each follow-up visit.

HPV Testing

Hybrid Capture II Test. After sampling using conventional Pap smear, additional samples for the HCII test and PCR were taken from the cervix using the HCII sampling kit. All samples were delivered to Turku (Finland) every second week and were immediately analyzed to comply with the manufacturer-guaranteed test validity period of 14 days since sampling. HPV DNA detection was performed by the automated HCII test system. The samples were analyzed only for the presence o. high-risk HPV types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68. The HCII method is an enzyme-linked immunosorbent assay based on sandwich hybridization, followed by a nonradioactive alkaline phospha-tase reaction with chemiluminescence in microplates. Samples were classified as HPV DNA-positive if the relative light unit (RLU) reading obtained from the lu-minometer was equal to or greater than the mean of positive control (PC) values included in the analysis. In addition to this manufacturer-recommended positive threshold (RLU/PC) of 1 pg/mL (approximately 8,000 copies of the HPV test) [25], the clinical relevance of other threshold levels was tested, e.g., the threshold of 10 pg/mL is 10 times higher than that of the mean of positive controls.

Statistical Analyses

Statistical analyses were performed using the SPSS. for Windows program package (version 10.0; Chicago, IL). Frequency tables were analyzed using the chi-square test. Fisher exact test, Pearson R, Spearman rank correlation, and likelihood ratio (LR) were used (where appropriate) to assess the significance of the correlations. Differences in the means of the continuous variables (viral load RLU/PC) among the groups (histological grade, patient category) were analyzed using the non-parametric tests (independent samples t test with Mann-Whitney U test). The normal distribution of all continuous variables was tested using the Kolmogorov-Smirnov test, and log transformations were made where needed to correct the skewed distributions toward normal for the analysis of variance (ANOVA) tests.

Performance tests for HCII and Pap smear, sensitivity, specificity, and positive (PPV) and negative predictive value (NPV) were calculated using the conventional contingency tables, with histology (significant/nonsignificant), colposcopy (significant/nonsignificant), or Pap smear (high-grade/low-grade) used as the state variable when appropriate. Confidence intervals (95% CI) were calculated based on the F-distribution (.2 x standard error [SE]).

ROC Curves. Performance of the HCII test was modeled using receiver operating characteristic (ROC) analysis. The percent sensitivity of detecting significant pathology (CIN 3 or more severe lesion) was calculated for several HCII cutoff RLU/PC (HCII index) values (0.2, 1.0, 2.0, 5.0, 10.0, 20.0, 50.0, 100.0, 500.0, and 1000.0 pg/mL). The value of 1.0 pg/mL is the Food and Drug Administration-approved threshold for clinical use of the HCII test. These values were used as the test variable (sensitivity or y-axis) and blotted against the state variable (1-specificity or x-axis). Separate ROC curves were created for the HCII test in detecting significant histology and significant cytology as state vari-ables.-The procedure creates an ROC curve (x, y) composed of points indicating the performance of the HCII test at different cutoff levels. In calculating the SE of the mean area under the ROC curve, the assumption of a nonparametric distribution of the test values was used, and 95% CI was used to calculate the lower and upper limits. In null hypothesis, the area should be 0.5. An optimal HCII cutoff point would show 100% sensitivity and 100% specificity and represent the situation in which the HCII test could accurately discriminate between high-grade (CIN 3 or higher) and low-grade (CIN 2 or less severe lesion) histology (or cytology).

RESULTS

Patient material included in the study is characterized in Table 2. A total of 3,175 women have completed their first clinical visit. The women were examined by eight gynecologists (examiners 1-8) at different diagnostic settings in their clinics. Of the 3,175 women, 722 were classified as STD patients (22.7%), 761 as GYN patients (24.0%), and 1692 (53.3%) represent patients participating in cervical cancer screening (SCR) in the area. The mean age of the women in these eight settings was significantly different (p < .0001; ANOVA) in that the youngest women (26.8 years) were those in clinic 6, which examined almost exclusively STD patients. The mean age was highest in clinics 1, 5, and 8, which examined predominantly SCR patients (with no STD patients). Similarly, the differences in the other key characteristics are explained by the different patient profiles (SCR, GYN, STD) in these clinics (ANOVA; p < .0001; Pearson, p < .0001).

Table 2: Characterization of the Patients

Examiner: 1 N.N. Blokhin Cancer Research Centre of Russian Academy of Medical Sciences (Moscow); 2 N.N. Blokhin Cancer Research Centre of Russian Academy of Me Sciences (Moscow); 3 Russian Academy of Post-Graduate Medical Education (Moscow); 4 Russian Academy of Post-Graduate Medical Education (Moscow); 5 Novgorod Fe Consultative Outpatient Hospital, Department of Gynaecology (Novgorod); 6 Novgorod Municipal Dermatovenereological Dispensary, Department of Gynaecology (Novgo 7 Minsk State Medical Institute, Department of Gynaecology and Obstetrics (Minsk); 8 Latvian Cancer Centre, Department of Gynaecology (Riga). ANOVA, analysis of variance; GYN, gynecology; SCR, cervical cancer screening; SD, standard deviation; STD, sexually transmitted disease. aANOVA (F = 64.722; p < .0001); 6ANOVA (F = 18.800; p < .0001); CANOVA (F = 4.427; p < .0005); dPearson correlation (x2 = 60.973; p < .0001).

Table 3 shows the interobserver agreement of the Pap test results in the primary screening completed by the NIS/CCE partners and in the rescreening done in Finland. There was an overrating of class 2 and 2+ smea (equivalent to ASCUS in the Bethesda System [TBS]) the primary screening, contributing to the fact that tl reproducibility of this modified Papanicolaou classific tion was low (к = 0.32; 95% CI = 0.29-0.34). Concc dance was better, however, for the significant abnc malities (dysplastic smears) in which the reproducibili was substantial (к = 0.637; 95% CI = 0.57-0.70). Ii portantly, all high-grade lesions and cancers were i ported with high concordance in both screening rounc 15 of 17 invasive cancers being concordantly classifie and the remaining two cases with one- and two-da discrepancy only.

Table 3. Results of the Pap Test in the Primary Screening in NIS/CCE and in the Rescreening in Finland.

CCE, countries of Central Europe; Cl, confidence interval; LGSIL, low-grade squamous intraepithelial lesion; NIS, new independent states. Interobserver agreement for all Pap classes к = 0.32; 95% Cl = 0.29-0.34 and for significant cytology only (class 2+ above; equivalent to LGSIL and above) к = 0.637; 95% Cl = 0.57-0.70).

The performance characteristics of the Pap smear сytology in the detection of significant cervical pathology are summarized in Table 4.

Table 4: Sensitivity, Specificity, PPV, and NPV for the Pap Test in Detecting Significant Histology (CIN 3 or higher).

CCE, countries of Central Europe; Cl, confidence interval; CIN, cervical intraepithelial neoplasia; NIS, new independent states; NPV, negative predicting value; PPV, positive predicting value.

So far, a total of 231 bio sies have been analyzed, of which 30 were high grade cancer (22 and 8, respectively), 17 were CIN 2, 24 т CIN 1, and 43 were HPV-NCIN (HPV without CD No pathology (except metaplasia) was found in the n 117 biopsies. The presence of significant histology (Cl 3 or cancer) was strongly associated with significa cytology (odds ratio [OR] = 5.8; 95% CI = 2.7-12 p < .0001 for the primary screening; and OR = 8.5; 95 CI = 4.1-17.8; p < .0001 for the rescreening). The performance of the primary and the rescreening deviated from each other in regards to sensitivity, PPV, and NPV. Sensitivity of the primary screening was higher (74.2%) as compared with that (64.0%) of the rescreening because of the overrating of class 2 and 2+ smears, which increased the sensitivity but decreased specificity from 89% to 76%. NPV of the Pap test cytology was identical (95%) in both readings.

The prevalence of HPV infections in the three patient categories is depicted in Figure 2. Of the 3,068 samples analyzed by HCII, 1,013 (33.01%) were positive for oncogenic HPV types. The detection rates were equal (from 44% to 45%) in the STD and GYN patients, but significantly lower (23%) in the SCR women. There was a wide variation in HPV detection rates among the eight examiners (from 21% by examiner 1 to 44% by examiner 6; p < .0001)(data not shown).

Figure 2: HPV Detection Rate in Different Patient Categories

The presence of significant cytology (dysplastic or cancer cells) in the Pap smear was strongly associated with HCII positivity (at 1 pg/mL threshold) both in the primary screening (OR = 11.9; 95% Cl = 8.1-17.2; chi-square, p < .0001) and in the rescreening (OR = 14.4; 95% Cl = 8.4-24.5; chisquare, p < .0001). Increasing the HCII threshold (RLU/PC) slightly increased the OR to the 2 pg/mL cutoff (OR = 12.4; 95% Cl = 8.7-17.7 in the primary screening; and OR = 15.2; 95% Cl = 9.2-25.3 in the rescreening). A slight decrease in ORs was evident when the threshold was further increased. In both settings, the sensitivity was high (above 90%) at the level of 0.2 pg/mL but declined to around 85% at the cutoff level 1.0 pg/mL. At this level, HCII specificity fell around 70% in both Pap test readings. PPV was low with all cutoff levels, whereas NPV remained close to 100% (i.e., from 95.1% to 99.6%) with all cutoff levels in both the primary and rescreening (data not shown).

The performance characteristics of the HCII test at different cutoff (RLU/PC) levels in detecting significant cytology (state variable) in the primary and rescreening are shown by the ROC curves in Figure 3 and Figure 4. In general, the HCII test performance in the rescreening was slightly better compared with the primary screening. In Figure 3, the HCII cutoff point most closely balancing the sensitivity (83.1%) and specificity (75.6%) was 2 pg/mL. Increasing the cutoff from 2 pg/mL to about 50 pg/mL was associated with loss of sensitivity (70.6%) and significant gain in specificity (87.6%). To bring the specificity close to 100% (96.9%), the threshold of 1,000 pg/mL was required at the expense of sensitivity, which decreased to 33.2%. On the other hand, decreasing the threshold below 1 pg/mL resulted in a significant drop in specificity, with little further gain in sensitivity. The area below the ROC curve was 0.855 (95% Cl = 0.825-0.894), which is highly significant (p = .0001). The ROC curve in Figure 4 indicates an even better performance of the HCII test. In this case, the HCII cutoff point with the most balanced sensitivity (85.9%) and specificity (73.9%) also falls close to 2 pg/mL. When the cutoff increased from 2 pg/mL to about 50 pg/mL, a deep vertical slope of the curve results with loss of sensitivity (71.0%) and significant gain in specificity (86.2%) similar to that in Figure 3.

Figure 3: ROC Curve for HCII Test in Detecting Significant Cytology (Primary Screening in NIS/CCE).

Figure 4: ROC Curve for HCII Test in Detecting Significant Cytology (Rescreening in Finland).

The curves are different at the path where the threshold decreases below 1 pg/mL, still resulting in significant drop in specificity, but also significant gain in sensitivity. Area below this ROC curve was also slightly larger, 0.866 (95% CI = 0.833-0.900) (p = .0001).

The presence of significant histology (CIN 3 or more severe) was strongly associated with HCII positivity (cutoff 1 pg/mL; OR = 4.8; 95% CI = 0.7-34.2; chi-square, p = .047). Increasing the HCII threshold JRLU/PC) rapidly increased the OR as well to the 10 pg/mL cutoff (for 5 pg/mL, OR = 6.3; 95% CI = 1.6-25.9; for 10 pg/mL, OR = 7.6; 95% CI = 1.9-31.2). Interestingly, the OR started to decline when the threshold was further increased: for 20 pg/mL, OR = 4.4; 95% CI = 1.6-12.1; for 50 pg/mL, OR = 2.6; 95% CI = 1.2-5.6; and for 100 pg/mL, OR = 2.1; 95% CI = 1.1-4.3. The performance characteristics of the HCII test at different cutoff (RLU/PC) levels in detecting significant histology are summarized in Table 5.

Table 5. Sensitivity, Specificity, PPV, and NPV for HCII in Detecting Significant Histology (CIN 3 or more).

CI, confidence interval; CIN, cervical intraepithelial neoplasia; HCII, Hybrid Capture II; NPV, negative predicting value; PPV, positive predicting value.

The sensitivity remains high (above 90%) at the cutoff level of 10 pg/mL but drops quickly at higher thresholds down to 26.6% at the cutoff level 1,000 pg/mL. At the manufacturer-recommended (1 pg/mL) cutoff, sensitivity is 96.6%. At this level, HCII specificity is very low (15.9%) and increases slowly when the thresholds are increased, reaching 83.5% at the 1,000 pg/mL threshold. PPV remains low with all cutoff levels in contrast to NPV, which does not fall below 88% at any of the cutoff levels. Changing the level of significant histology from CIN 3 to CIN 2 does not significantly affect the HCII sensitivity, specificity, and PPV values, but slightly decreases the NPV from 1 pg/mL threshold upwards (80.9% at 1,000 pg/mL)(data not shown).

Figure 5 shows the ROC curve for the HCII test in detecting significant pathology. As compared with Figure 3 and Figure 4 (ROC for HCII in detecting cytology), the sensitivity remains markedly higher through a wider range of cutoff values (i.e., from 100% to 93.3% at threshold 10 pg/mL). This takes place at the expense of lower specificity, which does not reach 85% even at cutoff 1,000 pg/mL. This signifies that not all lesions with high RLU/PC values (> 1,000 pg/mL) are clinically high-grade lesions. The area under the curve is 0.657 (95% CI = 0.569-0.746; p = .006).

Figure 3: ROC Curve for HCII Test in Detecting Significant Pathology (CIN or higher).

DISCUSSION

The NIS countries are attractive targets for testing the new, potential tools for cervical cancer screening [16-20, 25-28]. There is an inheritance from the past, a law (Order of Ministry of Health of the Soviet Union, No. 770 from May 30, 1986), stating that every woman from 18 years of age has to be screened by Pap smear every year. This order has been obscured by the recent political and economic turmoil of these countries and parallels the reported increasing trend in incidence and mortality rates of cervical cancer [8, 9]. Similarly, the recent past has witnessed a rapid increase of all STDs including genital HPV infections [20]. The aim of our ongoing EC-funded project is to find out the most cost-effective means for effective cervical cancer control in these countries. In this first communication from this large-scale project, we compared the performance of conventional Pap smear cytology and HCII assay as optional screening tools in these low-resource settings, as previously tested in other geographic regions as well [29-34].

The rationale for inclusion of HPV testing as an adjunct to the screening programs is based on the data on the natural history of cervical HPV infections [11, 12, 14, 15]. A vast majority of women experience HPV infection as a transient passenger, to be cleared within a short time without any resulting clinical lesions (early regressors)[12, 35]. In a substantial proportion of women, however, HPV infection is associated with a clinical lesion and results in the development of CIN (persisters and potential progressions). Even in the latter group, however, the spontaneous regression rate of low-grade lesions (HPV-NCIN or CIN 1) is substantial, particularly in younger women, whereas persistence of the virus is more common in older women. Detection of HPV DNA in the absence of cytological abnormalities can also indicate presence of high-grade CIN missed by a Pap smear. On the other hand, Pap smear-negative women who also test negative for oncogenic HPV have an extremely low risk of developing high-grade CIN during the next several years. Finally, the risk for disease progression is associated with the oncogenic HPV types and their high viral load, thus predisposing these women to increased risk for invasive cancer [11, 12, 14, 15, 35, 57].

In the present approach, a total of 3,175 women belonging to three diagnostic categories (SCR, GYN, STD) were enrolled in the study in Russia, Belarus, and Latvia. Inclusion of these three categories of women enables the assessment of the performance of optional screening tools in women with distinctly different risks for HPV infections and cervical cancer [11, 12, 35, 36], which is important to avoid the bias of selecting the test subjects from a too narrow diagnostic category (e.g., women with CIN or ASCUS smears only). As pointed out later, it is essential to carefully consider the study population while interpreting the results of HCII assay because the test performance critically depends on the prevalence of HPV and CIN in the population [29-34,37-41].

A good performance of cervical cytology is the mainstay of all successful screening programs [1-7, 10]. In the present study, a modified Papanicolaou classification was used, supplemented with an extra category of class 2+, which is a close equivalent to ASCUS of TBS [21, 42]. While converting our classification to TBS, the low concordance of our classification did not improve (к = 0.34; 95% CI = 0.31-0.37), which was due to the known reproducibility problems of ASCUS smears [12, 21, 42, 43]. In the primary and rescreening, the sensitivity (74% and 64%, respectively) and the specificity (76.6% and 89.1%, respectively) of our Pap smear cytology were at the same level or better than reported in a number of recent studies with comparable settings [25, 29, 30, 32, 37, 39, 40, 44-52]. While comparing these data, it is essential to remember, however, that there is a major difference in the performance of Pap tests, e.g., in a series of 100 CIN lesions (with few negative results and sensitivity, specificity, and PPV easily above 90%), and in a series of 1000 women participating in a screening program, with very low prevalence of high-grade CIN [38, 44, 46, 50, 52]. In the latter case, the performance characteristics of the Pap test resemble those of the present study and other similar series [30, 37, 40, 46, 50].

Both HCI and HCII assays have been extensively tested in HPV diagnosis [16-20] and more recently as an auxiliary screening tool to conventional or liquid-based cytology [25-52]. Results from highly variable diagnostic settings and patient populations examined by HCII have been reported, but so far no experience on this test is available from Russia and other NIS countries. In our study, the prevalence of HPV infections with high-risk types varied within a range from 21% to 44% among the different clinics, being lowest in those only screening women and highest in those enrolling STD patients. The values in the latter were of the same magnitude as previously reported for women attending STD clinics in other countries [41, 53], whereas the detection rates among screening women were high [39, 40, 45, 48, 54]. The same was true with the GYN patients approaching the HPV prevalence reported in women referred for col-poscopy because of abnormal cytology [31, 32, 37, 44, 46, 47]. The observed HPV prevalence in these NIS countries is at the same level or higher than reported in some of the Latin American high-risk countries [25], which should be regarded as an alarming sign. Such a high prevalence of HPV indicates two things: 1) a rising trend in cervical cancer is to be anticipated in the near future; or 2) the official figures are rough underestimates (because of underreporting or nonreporting)[8].

The presence of significant cytology in the Pap smear was closely associated with HCII positivity both in the primary screening and in the rescreening. Similar data have been reported in other recent studies [32, 47, 54, 56]. Testing,the different cutoffs resulted in some fluctuation in these ORs, possibly reflecting the influence of different levels of viral load [57]. The HCII test at different cutoff (RLU/PC) levels was highly sensitive (above 90%) in detecting significant cytology but with a lower specificity. In ROC curve analysis, HCII performance in both the primary and rescreening deviated highly significantly from that obtained in this type of analysis by change, and the area under the curve (0.855 and 0.866, respectively) was statistically highly significant. In practical terms, the HCII test is a sensitive tool in detecting significant cytology, and the NPV of this test is around 95% to 98%, i.e., a negative HCII test precludes high-grade cytology with a .high degree of probability.

A conventional Pap test has traditionally been the only test fulfilling the criteria of high sensitivity and specificity in detecting CIN lesions and has shown its efficacy as a screening tool in properly organized programs [1-7, 10, 12]. The role of the Pap smear recently has been challenged by the data from numerous studies reporting sirnilar or better performance for the HCII test [25-54], which could make this test a feasible option as a screening tool in low-resource settings where cytology is not available or services are clearly insufficient [19, 25, 34, 35]. The present study showed performance of the HCII test fully comparable with that of the Pap smear. In fact, the HCII test seems to be more sensitive (96.6% at 1 and 2 pg/mL threshold) than the Pap test (74% and 64%) in detecting significant pathology. The specificity of the Pap test (77% band 89%), however, far exceeds that (16% to 28% at 1 and 2 pg/mL cutoff) of HCII. These data are in alignment with most of the recent studies_[30, 31, 38, 39, 40, 41, 45, 48, 49, 54, 56]. Although n’ot identical, similar results have been recently elaborated on in the ROC analysis of more or less comparable data [25, 48]. We were not able to reach the specificity level (more than 90%) reported for HCII test by some studies, however [25, 37, 52, 55]. As mentioned above, this depends dramatically on the frequency of HPV infections and CIN in the study population. When both the Pap test and HCII assay (1 pg/mL) were combined, the sensitivity of detecting significant histology was 93.5% (95% CI = 84.4-100.0), specificity was 17.4% (95% CI = 12.1-22.7), NPV was 94.6% (95% CI = 86.8-100.0), but PPV was only 14.8% (95% CI = 9.7-19.9). Thus, addition of the HCII high-risk HPV DNA assay to Pap smear cytology appears to increase the sensitivity of detecting significant pathology, which is consonant with the data presented in some of the recent reports [25, 30, 32, 44, 52].

In interpreting the results of this study, reference should be made to the known natural history of HPV infections and CIN [11, 12, 14, 15]. Compelling evidence suggests that most of the persisting infections are caused by the high-risk HPV types. It is noteworthy, however, to make the distinction between the following: 1) chronic virus carriers; and 2) those having a persistent clinical HPV lesion. The former are women in whom the only evidence for virus persistence can be obtained with the molecular diagnostic tools but who do not have any clinically detectable HPV lesions. If only the viral DNA is detected, there is no abnormal Pap smear and, thus no lesion to be detected on colposcopy and biopsy. These are the cases that contribute to the higher sensitivity but lower specificity of the HCII test, when biopsy (or colposcopy) is used as the "gold standard". This is also the reason why PPV of the HCII test is quite low in most of the reported studies [30, 31, 38, 39, 40-56]. The latter group of women, in turn, are those who have a clinical lesion detectable by Pap smear, colposcopy, and biopsy. These are the cases that make the Pap test more specific and give it a higher PPV in detecting significant pathology compared with the HCII test. Thus, it is an oversimplification to make direct comparisons of the performance of the Pap test and HCII assay because they measure different things, i.e., clinical versus subclinical/la-tent, respectively [11, 12, 14, 15, 34, 35]. The clinical applicability of the test, e.g., selection of a screening tool, critically depends on its cost-effectiveness [16, 18, 19, 25, 34, 35]. The key issue in this respect is the avoidance of any extra costs, most notably false referrals to colposcopy [21]. From the practical standpoint, reduction of the number of these false positive alarms to a minimum is critical in low-resource settings like the NIS/CCE countries of this project. How the HCII test performs in this respect will be the subject of a separate report.

Acknowledgments

This study has been supported by the INCO-Copernicus Program of the European Commission (Contract No. ERB IC15-CT98-0321). Special thanks are due to Digene Europe, for kindly providing the Hybrid Capture analyzer, samplers, and the test kits. The skillful technical assistance of Ms. Anneli Suhonen and Ms. Sari Maki is gratefully acknowledged. The authors are indebted to lAC-certified cytotechnologists Anna-Maija Korhonen and Kirsti Ikkala for rescreening the Pap smears. Special thanks to Mrs. Mervi Puotunen, Mr. Mikko Soderling, and Mr. Ville Jussila for their help in storing the data in the SPSS files.

REFERENCES

• 1. Miller AB, Chamberlain J, Day NE, Hakama M, Pro-rok PC. Report on a Workshop of the UICC Project on Evaluation of Screening for Cancer. Int J Cancer 1990;46:761-9.
• 2. Franco E, Monsonego J, eds. New developments in cervical cancer screening and prevention. Oxford: Blackwell Science, 1997.
• 3. Hakama M. Screening for cervical cancer: Experience from the Nordic Countries. In: Franco E, Monsonego J, eds. New developments in cervical cancer screening and prevention. Oxford: Blackwell Science, 1997-190-9.
• 4. Miller AB, Nazeer S, Fonn S, Brandup-Lukanow A, Rehman R, Cronje H, et al. Report on Consensus conference on Cervical Cancer Screening and Management. Int J Cancer 2000;86:440-7.
• 5. Ponten J, Adami HO, Bergstrom R, Dillner J, Friberg LG, Gustafsson L, et al. Strategies for global contrli of cervical cancer. Int J Cancer 1995-60:1-26.
• 6. Anttila A, Pukkala E, Soderman B, Kallio M, Hel-minen P, Hakama M. Effect of organised screening on cervical cancer incidence and mortality in Finland, 1963-1995: recent increase in cervical cancer incidence. Int J Cancer 1999;83: 59-65.
• 7. Syrjanen K. Early detection of CIN, HPV and prevention of cervical cancer. In: Syrjanen K, Syrjanen S, eds. Papil-lomavirus Infections in Human Pathliogy. New York: J. Wiley & Sons, 2000;252-80.
• 8. Trapeznikov NN, Aksel EM. Cancer Incidence and Mortality from the Population in Russia and NIS Countries in 1998. Moscow: RAMS Publishing; 2000.
• 9. Parkin DM, Pisani P, Ferlay J. Estimates of the worldwide incidence of 25 major cancers in 1990. Int J Cancer 1999;80:827-41.
• 10. Franco E, Syrjanen K, de-Wlif C, Patnick J, Ferenczy A, McGoogan E, et al. New developments in cervical cancer screening and prevention. Cancer Epidemili Biomarkers Prev 1996;5:853-6.
• 11. IARC Monographs on the evaluation of carcinogenic risks to humans: Papillomavirus. Lyon: IARC, 1995;64:1-409.
• 12. Syrjanen K, Syrjanen S. Papillomavirus Infections in Human Pathliogy. New York: J. Wiley & Sons, 2000
.
• 13. Cuzick J, Sasieni P, Singer A. Risk factors for invasive cervix cancer in young women. Eur J Cancer 1996;32A: 836-41.
• 14. Syrjanen KJ. Biliogical behaviour of cervical intraepi-thelial neoplasia. In: Franco E, Monsonego J, eds. New developments in cervical cancer screening and prevention. Oxford: Blackwell Science, 1997;93-108.
• 15. Franco EL. Statistical issues in human papillomavirus testing and screening. Clin Lab Med 2000;20:345-67.
• 16. Cuzick J. Screening for cancer: future potential. Eur J Cancer 1999;35:1925-32.
• 17. De Sanjose S, Bosch XF, Munoz N, Chichareon S, Ngelangel C, Balaguero L, et al. Screening for genital human papillomavirus: results from an international validation study on human papillomavirus sampling techniques. Diagn Mli Pathli 1999-8:26-31.
• 18. Richart RM. Screening. The next century. Cancer 199’5;7’6:1919-27.
• 19. Schiffman M, Hildesheim A. Human papillomavirus testing as a screening toli for cervical cancer. JAMA 2000; 283:2525-6.
• 20. Zavalishina LE, Andreeva JJ, Manikin AA, Frnak GA, Lisicin FV, Efremova EV, et al. Detection of human papillomavirus DNA in cervix uteri and larynx by in situ hybridization. Russ OncliJ 1999 ;5-.25-8.
• 21. Syrjanen KJ. Management of abnormal PAP smears (MAPS): Implications of terminliogy used in cytopathliogy. /. Lower Genital Tract Dis 2000;4:217-23.
• 22. Stafl A, Wilbanks GD. An international terminliogy of cliposcopy: report of the Nomenclature Committee of the International Federation of Cervical Pathliogy and Cliposcopy. Obstet Gynecli 1991;77:313-4.
• 23. Richart RM. Cervical intraepithelial neoplasia. Pathli Annu 1973;8:301-28.
• 24. Kataja V, Syrjanen S, Yliskoski M, Hippelainen M, Vayrynen M, Saarikoski S, et al. Risk factors associated with cervical human papillomavirus (HPV) infections: A case-contrli study. Am J Epidemili 1993;138:735-45.
• 25. Schiffman M, Herrero R, Hildesheim A, Sherman ME, Bratti M, Wachlider S, 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.
• 26. Sherman ME, Schiffman MH, Lorincz AT, Herrero R, Hutchinson ML, Bratti C, et al. Cervical specimens clilected in liquid buffer are suitable for both cytliogic screening and ancillary human papillomavirus testing. Cancer 1997;81: 89-97.
• 27. Sigurdsson K, Arnadottir T, Snorradottir M, Bene-diktsdottir K, Saemundsson H. Human papillomavirus (HPV) in an Icelandic population: the rlie of HPV DNA testing based on hybrid capture and PCR assays among women with screen-detected abnormal Pap smears. Int J Cancer 1997;72: 446-52.
• 28. Cope JU, Hildesheim A, Schiffman MH, Manos MM, Lorincz AT, Burk RD, et al. Comparison of the hybrid capture tube test and PCR for detection of human papillomavirus DNA in cervical specimens. / Clin Microbili 1997;35: 2262-5.
• 29. Clavel C, Bpry JP, Rihet S, Masure M, Duval-Binninger I, Putaud I, et al. Comparative analysis of human papillomavirus detection by hybrid capture assay and routine cytliogic screening to detect high-grade cervical lesions. Int ] Cancer 1998;75:525-8.
• 30. Ferris DG, Wright TC, Jr., Litaker MS, Richart RM, Lorincz AT, Sun XW, et al. Triage of women with ASCUS and LSIL on Pap smear reports: management by repeat Pap smear, HPV DNA testing, or cliposcopy? / Fam Pract 1998;46: 125-34.
• 31. Ferris DG, Wright TC, Jr., Litaker MS, Richart RM, Lorincz AT, Sun XW, et al. Comparison of two tests for detecting carcinogenic HPV in women with Papanicliaou smear reports of ASCUS and LSIL. / Fam Pract 1998;46:136-41.
• 32. Recio FO, Sahai Srivastava BI, Wong C, Hempling RE, Eltabbakh GH, Piver MS. The clinical value of digene hybrid capture HPV DNA testing in a referral-based population with abnormal pap smears. Eur J Gynaecli Oncli 1998;19:203-8.
• 33. Spitzer M. Cervical screening adjuncts: recent advances. Am ] Obstet Gynecli 1998;179:544-56.
• 34. Syrjanen KJ, Syrjanen SM. Human papillomavirus (HPV) typing as an adjunct to cervical cancer screening. Cy-top^/7o/ogy,1999;10:8-15.
• 35. Cuzick J, Sasieni P, Davies P, Adams J, Normand C, Prater A, et al. A systematic review of, the rlie of human papillomavirus testing within a cervical screening programme. BrJ Cancer 2000;83:561-5.
• 36. Rozendaal L, Walboomers JM, van der Linden JC, Voorhorst FJ, Kenemans P, Helmerhorst TJ, et al. PCR-based high-risk HPV test in cervical cancer screening gives objective risk assessment of women with cytomorphliogically normal cervical smears. Int J Cancer 1996-68:766-9.
• 37. Fait G, Daniel Y, Kupferminc MJ, Lessing JB, Niv J, Bar-Am A. Does typing of human papillomavirus assist in the triage of women with repeated low-grade, cervical cytliogic abnormalitiesj1 Gynecli Oncli 1998;70:319-22.
• 38. Clavel C, Masure M, Putaud I, Thomas K, Bory JP, Gabriel R, et al. Hybrid capture II, a new sensitive test for human papillomavirus detection. Comparison with hybrid capture I and PCR results in cervical lesions. / Clin Pathli 1998;51:737-40.
• 39. Clavel C, Masure M, Bory JP, Putaud I, Mangeonjean C, Lorenzato M, et al. Hybrid Capture П-based human papillomavirus detection, a sensitive test to detect in routine high-grade cervical lesions: a preliminary study on 1518 women. Br J Cancer 1999;80:1306-H.
• 40. Cuzick J, Beverley E, Ho L, Terry G, Sapper H, Miel-zynska I, et al. HPV testing in primary screening of lider women. Br J Cancer 1999;81:554-8.
• 41. Womack SD, Chirenje ZM, Gaffikin L, Blumenthal PD, McGrath JA, Chipato T, et al. HPV-based cervical cancer screening in a population at high risk for HIV infection. Int J Cancer 2000;85:206-10.
• 42. Kurman RJ, Sliomon D. The Bethesda System for reporting cervical/vaginal cytliogic diagnoses. Definitions, criteria, and explanatory notes for terminliogy and specimen adequacy. New York: Springer-Verlag, 1994.
• 43. Smith AE, Sherman ME, Scott DR, Tabbara SO, Dworkin L, lison J, et al. Review of the Bethesda System atlas does not improve reproducibility or accuracy in the classification of atypical squamous cells of undetermined significance smears. Cancer 2000;90:201-6.
• 44. Lin CT, Tseng CJ, Chou HH, Huang KG, Chang TC, Lai HH, et al. High-risk human papillomavirus deoxyribonu-cleic acid as an adjunct marker in cervical cytliogy. Chang Keng I Hsueh Tsa Chih 1999;22:409-15.
• 45. Womack SD, Chirenje ZM, Blumenthal PD, Gaffikin L, McGrath JA, Chipato T, et al. Evaluation of a human papillomavirus assay in cervical screening in Zimbabwe. BJOG 2000;107:33-8.
• 46. Human papillomavirus testing for triage of women with cytliogic evidence of low-grade squamous intraepithelial lesions: baseline data from a randomized trial. The Atypical Squamous Cells of Undetermined Significance/Low-Grade Squamous Intraepithelial Lesions Triage Study (ALTS) Group J Natl Cancer Inst 2000;92:397-402.
• 47. Lin CT, Tseng CJ, Lai CH, Hsueh S, Huang HJ, Law KS. High-risk HPV DNA detection by Hybrid Capture II. An adjunctive test for mildly abnormal cytliogic smears in women > or = 50 years of age. / Reprod Med 2000;45:345-50.
• 48. Kuhn L, Denny L, Plilack A, Lorincz A, Richart RM, Wright TC. Human papillomavirus DNA testing for cervical cancer screening in low- resource settings. / Natl Cancer Inst 2000;92:818-25.
• 49. Bergeron C, Jeannel D, Poveda J, Cassonnet P, Orth G. Human papillomavirus testing in women with mild cytliogic atypia. Obstet Gynecli 2000;95:821-7.
• 50. Sellers JW, Lorincz AT, Mahony JB, Mielzynska I, Lytwyn A, Roth P, et al. Comparison of self-clilected vaginal, vulvar and urine samples with physician-clilected cervical samples for human papillomavirus testing to detect high-grade squamous intraepithelial lesions. CMA/ 2000;163:513-8.
• 51. Lytwyn A, Sellers JW, Mahony JB, Daya D, Chapman W, Ellis N, et al. Comparison of human papillomavirus DNA testing and repeat Papanicliaou test in women with low-grade cervical cytliogic abnormalities: a randomized trial. HPV Effectiveness in Low-grade Paps (HELP) Study No.l Group. CMA/ 2000;163:701-7.
• 52. Fait G, Kupferminc MJ, Daniel Y, Geva E, Ron IG, Lessing JB, et al. Contribution of human papillomavirus testing by hybrid capture in the triage of women with repeated abnormal pap smears before cliposcopy referral. Gynecli Oncli 2000;79:177-80.
• 53. Kubota T, Ishi K, Suzuki M, Utsuno S, Igari J. Prevalence of human papillomavirus infection in women attending a sexually transmitted disease clinic. Kansenshogaku Zasshi 1999;73:233-8.
• 54. Sellers JW, Mahony JB, Kaczorowski J, Lytwyn A, Bangura H, Chong S, et al. Prevalence and predictors of human papillomavirus infection in women in Ontario, Canada. Survey of HPV in Ontario Women (SHOW) Group. CMAJ 2000;163:503-8.
• 55. Riethmuller D, Gay C, Bertrand X, Bettinger D, Schaal JP, Carbillet JP, et al. Genital human papillomavirus infection among women recruited for routine cervical cancer screening or for cliposcopy determined by Hybrid Capture II and pliymerase chain reaction. Diagn Mli Pathli 1999;8: 157-64.
• 56. Shlay JC, Dunn T, Byers T, Baron AE, Douglas JM, Jr. Prediction of cervical intraepithelial neoplasia grade 2-3 using risk assessment and human papillomavirus testing in women with atypia on papanicliaou smears. Obstet Gynecli 2000; 96:410-6.
• 57. Josefsson AM, Magnusson PK, Ylitalo N, Sorensen P, Qwarforth-Tubbin P, Andersen PK, et al. Viral load of human papilloma virus 16 as a determinant for development of cervical carcinoma in situ: a nested case-contrli study. Lancet 2000;355:2189-93.

Сообщения на форуме:

Другие статьи раздела