Vercellino, Giuseppe F. MD; Erdemoglu, Evrim MD; Chiantera, Vito MD; Vasiljeva, Katharina MD; Drechsler, Inka MD; Cichon, Guenter MD; Schneider, Achim MD; Böhmer, Gerd MD, MPH
Colposcopy is the basis for the correct identification of the atypical transformation zone, for the definition of the grade of the underlying lesion, for targeted biopsy in case of high-grade cervical intraepithelial neoplasia (CIN), and for excisional therapy. The literature on the accuracy of colposcopy is contradictory.1–3 As a result of the transient character and duration of the changes of the abnormal cervical tissue, induced by acetic acid application and the subjective evaluation of colposcopic findings such as margins, vessel, and color, colposcopy is a subjective “art” and results in interobserver and intraobserver variability.4–6
Over the years, a number of grading indices, which are based on a variety of morphologic phenomena, have been created to help differentiate among normal, minor change, major change, and cancer.7–12 One of these that has attempted to quantify grade by assigning points to each morphologic finding7–10 has not been confirmed to be highly predictive of the ultimate histology.3,11 Thus, International Federation for Cervical Pathology and Colposcopy13 has revised the grading nomenclature by adding two recently described pathognomonic criteria, inner border sign and ridge sign, highly associated with the presence of high-grade CIN.14,15 Pathognomonic signs differ from the other existing criteria in that they are simply present or absent: they should be dichotomously reported, and not graduated, their presence being significantly associated with high-grade CIN.
We identified another pathognomonic nongradeable criterion associated with high-grade CIN, the “rag sign,” an iatrogenic small erosion of the epithelium generally caused by its partial ablation during mechanical trauma of the cervix at the time of the collection of the smear for cytology or human papillomavirus (HPV) testing, application of acetic acid or Lugol's solution, or both.16
We undertook the present study to evaluate the detection rate for high-grade CIN of three objective, pathognomonic, nongradeable features of the abnormal transformation zone: inner border sign, ridge sign, and rag sign. Additionally we evaluated a possible association with the age of the patient.
PATIENTS AND METHODS
Between November 2010 and July 2011, 389 patients were referred to the colposcopy and lower genital tract disease center of the Charité University, Berlin and the Colposcopy Clinic Wagner Stibbe, Bad Muender, Germany, with a history of an abnormal Pap smear, detection of high-risk HPV in a cervical smear, or both. Each patient underwent video exoscopy using the VITOM System and histologic evaluation by VITOM-guided biopsy or loop excision.
The VITOM System is a video exoscope based system consisting of the VITOM 25 scope, Xenon 300 light source, FULL HD camera system, AIDA HD documentation system, FULL HD monitor, a mechanical support arm, and a three-chip HD camera head providing a resolution of 1,920×1,080 pixels (FULL HD, progressive scan) with a frame rate of 50/60 frames per second. The camera has a 2× parfocal zoom and weighs 246 g. The VITOM System supporting a 25-cm to 60-cm working distance provides comfortable workspace. The VITOM System has shown to be accurate for the diagnosis of high-grade CIN17 and hence has replaced traditional colposcopy in our center since July 2010. This system combines two decisive and equally important advantages over conventional colposcopy: it provides a FULL HD video documentation of colposcopic examination, and it allows its storage and reanalysis both for teaching and scientific purposes.
Patients with a history of hysterectomy or previous radiation were excluded. In addition, we decided to exclude from our analysis patients with transformation zone type 3 (n=44), as defined by International Federation for Cervical Pathology and Colposcopy 2003,18 because colposcopic diagnosis is insufficient and unreliable in these instances.19 Patients with invasive cancers (n=10) were also excluded, because the focus of this study is the evaluation of colposcopic criteria for the detection of high-grade CIN and not of invasive disease. Thus, 335 of 389 patients (86.1%) of the total cohort were eligible for evaluation.
All patients undergoing video recording procedures in our institution give written consent to the saving of both images and videos of their examination or operation. This study was approved by our internal review board.
All video exoscopies stored in the AIDA HD documentation system were evaluated independently, on a 26-inch HD monitor, with a luminance of 400 cd/m², by two senior colposcopists (S.A., B.G.) in a darkened, artificially lit room, with back-screen illumination, to provide optimal viewing conditions and to standardize the examination, thus enhancing reproducibility of the results. Both parties were blinded to all clinical data of the patients such as referring diagnosis, results of Pap smear, HPV test, medical history, and histology apart from age.20
The overall colposcopic impression was categorized as normal or benign abnormality (inflammation, atrophy), minor change, or major change. All video exoscopies were specifically checked for the presence of inner border, ridge sign, and rag sign at the level of the transformation zone. Video colposcopic findings and transformation zone (TZ) type were reported according to the criteria of the Committee on Nomenclature of the International Federation of Cervical Pathology and Colposcopy: “Type 1 TZ is completely ectocervical and fully visible; type 2 TZ has an endocervical component, is fully visible; type 3 TZ has an endocervical component that is not fully visible.”18
The inner border is a dull, oyster white area, inside a less opaque acetic white area, in its turn demarcated from the normal squamous epithelium.14 The peripheral area represents an earlier, minor grade change, the central area being the subsequent evolution of a high-grade CIN at the advancing edge of the new squamocolumnar junction with aging (Fig. 1).
Atypical transformat...Image Tools
The ridge sign is an opaque lesion, direct adjacent to the squamocolumnar junction, which resembles a mountain ridge15 (Fig. 2).
Atypical transformat...Image Tools
The rag sign is an opaque acetowhite area at the squamocolumnar junction, mechanically abraded during either collection of the smear for cytology or HPV testing, applying acetic acid or Lugol's solution, or both. In the rag sign, part of the epithelium is sloughed off, and either the underlying erosion or the detached epithelium, which resembles a rag, are visible (Figs. 3–5).
All punch biopsies or loop excisions were made under colposcopic or VITOM guidance21 and the material was fixed in formalin, paraffin-embedded, sectioned in series, and stained with hematoxylin and eosin. CIN was classified according to the criteria described by Crum.22 Statistical analysis was performed using PASW 20 and Medcal. Descriptive statistics including the mean age of patients and the frequency of each sign in different age groups was performed. Chi square analysis was used to compare categorical and ordinal data. We evaluated colposcopic performance for each colposcopic sign separately. For high-grade CIN, we estimated the sensitivity and specificity, positive predictive value (PPV), and negative predictive value (NPV), negative likelihood ratio (LR−), and positive likelihood ratio (LR+), and the 95% confidence intervals for those statistics in different age groups.
Likelihood ratios determine whether presence of a colposcopic sign usefully changes the probability that CIN exists by using sensitivity and specificity. The LRs have advantages over predictive values because they are less likely to change with the prevalence of the disorder, and they can be calculated for several levels of the symptom or sign or test.23 An LR between 1 and 5 indicates a small increase in the likelihood of disease; however, an LR between 5 and 10 and more than 10 indicates a moderate and strong conclusive likelihood of disease, respectively. A negative LR between 0.5 and 1 indicates a minimal decrease in the likelihood of disease, and an LR between 0.2 and 0.5 indicates a small decrease in the likelihood of disease. However, a negative LR between 0.1 and 0.2 and less than 0.1 indicates a moderate and large decrease in the likelihood of disease, respectively. A P value <.05 was considered statistically significant.
A total of 335 patients were included in the study. The mean age of patients was 33.02±7.7 years. 14.3% of patients were younger than 25 years, 53.1% of patients were between 25 and 35 years, 25.1% of patients were between 35 years and 45 years, and 7.5% of patients were older than 45 years of age. Median age for patients with inner border was 31.5 years, 30 years for patients with the ridge sign and 31 years in patients with the rag sign. The Pap smear result was normal, atypical squamous cells of uncertain significance, low-grade squamous intraepithelial lesion, and high-grade squamous intraepithelial lesion in 12.4%, 5.9%, 32.3%, and 49.4% of patients, respectively. Of 335 patients, 287 patients were tested for the presence of HPV DNA: 64.5% were high-risk HPV-positive and 1.8% were low-risk HPV-positive. In 19.4% of patients, no HPV was detected. In 285 patients (85%), a single VITOM-directed biopsy was taken; 50 patients (15%) underwent two biopsies. One hundred sixty-two patients (48%) underwent subsequent magnification-guided loop excision and had thus two histologic results (Table 1). In these patients, the most severe histologic diagnosis was recorded. Cervical intraepithelial neoplasia 1 or high-grade CIN were diagnosed in 10.1% and 69.8% of patients, respectively; in the rest of the patients, the histologic diagnosis was normal. In women younger than 25 years old, 23%, 10%, and 67% had normal histology, CIN 1, or high-grade CIN, respectively. In women between 25 and 35 years old, 16%, 10%, and 74% had normal histology, CIN 1, or high-grade CIN, respectively. In women older than 35 years old, 25%, 11%, and 64% had normal histology, CIN 1, or high-grade CIN, respectively.
One hundred thirty-eight patients (41.2%) had transformation zone type 1 and a median age of 30 years; 197 patients (58.8%) had transformation zone type 2 zones. Inner border, the ridge sign, and the rag sign were seen in 14.3% (n=48), 37.9% (n=127), and 28.1% (n=94) of women, respectively.
The frequency of inner border sign and rag sign was not different between younger than 25 years old, 25–35 years old, and older than 35 years old. Ridge sign was the only pathognomonic criterion, which showed a correlation with age (P<.05). Ridge sign was significantly less common in women older than 35 years old. Inner border was present in 12.5% 14.6%, and 14.7% of women younger than 25 years old, 25–35 years old, or older than 35 years old (P=.92). Ridge sign was found in 39.6%, 43.8%, and 27.5% of women younger than 25 years old, 25–35 years old, or older than 35 years old (P<.05). The rag sign was found in 27.1%, 26.4%, and 31.2% of women younger than 25 years old, 25–35 years old, or older than 35 years old (P=.67).
A total of 2.1%, 0%, and 97.9% of inner border positivity was found in women with normal, CIN 1, and high-grade CIN histology, respectively (P<.001). A total of 1.6%, 1.6%, and 96.8% of ridge sign positivity was found in women with normal, CIN 1, and high-grade CIN histology, respectively (P<.001). A total of 4.3%, 0%, and 95.7% of rag sign positivity was found in women with normal, CIN 1, 2, and 3 histology, respectively (P<.001) (Table 2).
Of 234 patients with high-grade CIN, 47 showed inner border in colposcopic examination. Sensitivity, specificity, PPV, and NPV of inner border to detect high-grade CIN were 20%, 99%, 97.9%, and 34.8%, respectively. The LR+ was 20.3 and the LR− was 0.81 (Table 3).
Table 3-a Sensitivit...Image Tools
Of 234 patients with high-grade CIN, 123 showed the ridge sign. Sensitivity, specificity, PPV, and NPV of the ridge sign to detect high-grade CIN were 52.5%, 96.4%, 96.8%, and 46.6%, respectively. The LR+ was 13.2 and the LR− was 0.49 (Table 3).
Table 3-b Sensitivit...Image Tools
Of 234 patients with high-grade CIN, 90 showed the rag sign. Sensitivity, specificity, PPV, and NPV of the rag sign to detect high-grade CIN was 38.4%, 96%, 95.7%, and 40.2%, respectively. The LR+ was 9.7 and the LR− was 0.6 (Table 3).
Inner border, the ridge sign, and the rag sign are good diagnostic tests to determine high-grade CIN (Tables 3 and 4). Diagnostic test performance of all three colposcopic signs was significantly improved in women 25–35 years old and older than 35 years old compared with women younger than 25 years old (Table 3) (P<.05). Of 234 patients with high-grade CIN, 182 (77.8%) patients had a combination of at least one of three pathognomic signs; 114 (49.5%) patients had one single sign and 58 (24.7%) patients had two of three signs and 10 (4.2%) patients had all three signs.
Inner border sign alone was seen in 20 patients, ridge sign alone was seen in 61 patients, and rag sign alone was present in 38 patients. Inner border and ridge sign were combined in 24 patients, inner border and rag sign were combined in 14 patients, and ridge sign and rag sign were combined in 52 patients. A combination of all three criteria was seen in 10 patients.
All 24 patients with a combination of the ridge sign and inner border were diagnosed with high-grade CIN. Of 52 patients with a combination of the ridge sign and rag sign, 51 patients were diagnosed with high-grade CIN. In 14 patients, a combination of the rag sign and inner border was seen and 13 of these patients were diagnosed with high-grade CIN. All 10 patients with a combination of ridge sign, inner border, and rag sign were diagnosed with high-grade CIN.
A combination of inner border sign and ridge sign has a specificity of 100%, a sensitivity of 10.3%, PPV of 100%, and NPV of 32.5% for detection of high-grade CIN. A combination of the ridge sign and rag sign has a specificity of 99% a sensitivity of 21.8%, PPV of 98.1%, and NPV of 35.3% for detection of high-grade CIN. A combination of the rag sign and inner border sign has a specificity of 99.0%, a sensitivity of 5.6%, PPV of 92.9%, and NPV of 31.2% for detection of high-grade CIN.
When at least one of the pathognomonic criteria was present, specificity was 93%, sensitivity was 77.8%, PPV was 96.2%, and NPV was 63%. Positive likelihood ratio was 11.2 and LR− was 0.2. Presence of at least two of three criteria has a specificity of 98.0%, a sensitivity of 29.1%, PPV of 97.1%, and NPV of 37.4%. Positive likelihood ratio was 14.6 and LR− was 0.7. A patient with a combination of two of the three pathonogmonic signs had a significant likelihood for the presence of high-grade CIN (Table 4). Therefore, the diagnosis of high-grade CIN should be considered in the presence of any two signs such as inner border, ridge sign, or rag sign.
Twenty-seven years ago Reid introduced the sharpness of the peripheral margins in his improved colposcopic graduating index and, although not separately analyzing the diagnostic power of each colposcopic sign, Reid described the importance of rolled edges, peeling edges, and internal border between lesions of different severity to differentiate minor and major change lesions.7
We highlight the rag sign, a sloughing off of atypical squamous epithelium in a rag mechanically abraded during either collection of the smear for cytology or HPV testing, applying acetic acid or Lugol's solution, or both. Its biological basis is unclear. The rag sign is a new pathognomonic, objective, colposcopic criterion, whose presence, alone or combined with one or both of the two other objective signs, “inner border” sign and “ridge” sign,14,15 is strongly associated with high-grade CIN. Because the most prominent areas of colposcopic change does not always coincide with the areas of greatest histologic abnormality, if present, pathognomonic signs can help colposcopists to focus their attention on the most prominent areas of morphologic change and to correctly diagnose high-grade CIN in up to 70% of the patients.
A multitude of different grading systems have been proposed and implemented in the daily praxis over the last 40 years24,25 but have failed to provide reproducible results. Therefore, some colposcopists recommend random biopsies.23,26,27
Diagnostic comparability between colposcopy and the VITOM System has been previously described.17 The efficacy of pathognomonic signs to suggest the presence of high-grade CIN can be better objectified through real time videos than by means of static photograms.6
To prove the simplicity and reproducibility of detection of pathognomonic signs, we opted for the evaluation of unselected, consecutive, previously recorded video exoscopies of patients referred for diagnostic colposcopy to our outpatient center for cervical and lower genital tract pathology. Videos were recorded by eight different colposcopists, in a diagnostic setting, and are thus fully representative of everyday colposcopic practice, which allows generalization of our results. In addition, because the video exoscopies are original and full length, not edited videos, it minimizes the retrospective nature of our study.
Colposcopy is mainly useful in a diagnostic setting with high prevalence of high-grade CIN as opposed to screening setting.23 The prevalence of high-grade CIN in our cohort is 69%, which will influence positively the predictive values of colposcopic criteria for high-grade CIN. Therefore, we calculated the LRs for each colposcopic sign, which, unlike PPV and NPV, are independent of prevalence of disease. It is remarkable that 77.8% of patients (182/234) with high-grade CIN showed at least one of the three criteria: inner border, ridge sign, or rag sign. This means that despite being very specific for high-grade CIN, a combination of pathognomonic signs is present in more than 75% of patients with high-grade CIN.
Simultaneous presence of pathognomonic criteria was highly indicative of high-grade CIN.
Because of the known high regression rate of CIN 2 in women younger than 25 years28 and of the significant difference of the acetic white reaction in women younger than 35 years,20 we set different age cutoffs and analyzed the association of pathognomonic signs with age. The ridge sign was the only pathognomonic criterion that showed a correlation with age (P<.05).
This study has a number of limitations. We exclusively focused our attention on three pathognomonic signs, disregarding standard grading signs, which are still the most widely used colposcopic diagnostic tools for diagnosis of high-grade CIN.13 Specifically we did not analyze the grade of the acetic white reaction20,29,30 the location,11,31 the size of the lesion,32,33 and the presence of cuffed glandular openings.13,34 Such comparisons are the focus of other ongoing prospective studies. In addition, the reviewer had no influence on the actual punch guided biopsy site, which in some occasions differed from the best possible target. The statistical limitation of the study is a high prevalence of high-grade CIN. However, the prevalence of high-grade CIN is always higher in a reference colposcopy clinic (20–60%) compared with a screening population. The prevalence of diseases effects PPV and NPV, but not the LR. At the same time, our study has some decisive strengths. First, the dynamic process of colposcopy is recorded by video exoscopy in real time, which allows subsequent evaluation of the cervix and of underlying pathology as opposed to static photograms.6 Second, to fully evaluate the three pathognomonic signs objectively, reviewers were blinded to smoking habits, cytologic index,35 and HPV status.36 Previous studies have shown that the knowledge of these results may increase the sensitivity for detection of high-grade CIN.2,36,37 Third, the search for pathognomonic signs resulting from their intrinsic dichotomy, present compared with absent, can increase the specificity of colposcopic grading.
Inner border sign, the ridge sign, and the newly defined rag sign yield a sensitivity of 77.8%, specificity of 93%, PPV of 96.2%, and NPV of 63% for detection of high-grade CIN if at least one of these signs is present. Positive likelihood ratio was 11 and LR− was 0.2. Pathognomonic signs, introduced in the latest International Federation for Cervical Pathology and Colposcopy terminology (2011), are objective, present or absent, easy to see and learn, with smaller risk for subjectivity.13 Their presence is highly suggestive for high-grade CIN, and their validity is independent of the age of the patient.
1. Mitchell MF, Schottenfeld D, Tortolero-Luna G, Cantor SB, Richards-Kortum R. Colposcopy for the diagnosis of squamous intraepithelial lesions: a meta-analysis. Obstet Gynecol 1998;91:626–31.
2. ASCUS-LSIL Triage Study (ATLS) Group. Results of a randomized trial on the management of cytology interpretations of atypical squamous cells of undetermined significance. Am J Obstet Gynecol 2003;188:1383–92.
3. Ferris DG, Litaker MS. Prediction of cervical histologic results using an abbreviated Reid Colposcopic Index during ALTS. Am J Obstet Gynecol 2006;194:704–10.
4. Ferris DG, Litaker M. Interobserver agreement for colposcopy quality control using digitized colposcopic images during the ALTS trial. J Lower Genit Tract Dis 2005;9:29–35.
5. Sellors JW, Nieminen P, Vesterinen E, Paavonen J. Observer variability in the scoring of colpophotographs. Obstet Gynecol 1990;76:1006–8.
6. Etherington IJ, Dunn J, Shafi MI, Smith T, Luesley DM. Video colpography: a new technique for secondary cervical screening. Br J Obstet Gynaecol 1997;104:150–3.
7. Reid R, Scalzi P. Genital warts and cervical cancer. VII. An improved colposcopic index for differentiating benign papillomaviral infections from high-grade cervical intraepithelial neoplasia. Am J Obstet Gynecol 1985;153:611–8.
8. Coppleson M. Colposcopic features of papillomaviral infection and premalignancy in the female lower genital tract. Obstet Gynecol Clin North Am 1987;14:471–94.
9. Stafl A, Mattingly RF. Colposcopic diagnosis of cervical neoplasia. Obstet Gynecol 1973;41:168–76.
10. Reid R, Fu YS, Herschman BR, Crum CP, Braun L, Shah KV, et al.. Genital warts and cervical cancer. VI. The relationship between aneuploid and polyploid cervical lesions. Am J Obstet Gynecol 1984;150:189–99.
11. Hong DG, Seong WJ, Kim SY, Lee YS, Cho YL. Prediction of high-grade squamous intraepithelial lesions using the modified Reid index. Int J Clin Oncol 2010;15:65–9.
12. Massad LS, Jeronimo J, Katki HA, Schiffman M. The accuracy of colposcopic grading for detection of high-grade cervical intraepithelial neoplasia. J Lower Genit Tract Dis 2009;13:137–44.
13. Bornstein J, Bentley J, Bosze P, Girardi F, Haefner H, Menton M, et al.. 2011 colposcopic terminology of the International Federation for Cervical Pathology and Colposcopy. Obstet Gynecol 2012;120:166–72.
14. Scheungraber C, Glutig K, Fechtel B, Kuehne-Heid R, Duerst M, Schneider A. Inner border—a specific and significant colposcopic sign for moderate or severe dysplasia (cervical intraepithelial neoplasia 2 or 3). J Lower Genit Tract Dis 2009;13:1–4.
15. Scheungraber C, Koenig U, Fechtel B, Kuehne-Heid R, Duerst M, Schneider A. The colposcopic feature ridge sign is associated with the presence of cervical intraepithelial neoplasia 2/3 and human papillomavirus 16 in young women. J Lower Genit Tract Dis 2009;13:13–6.
16. Wagner M. Wertigkeit kolposkopisch definiert Muster und Zeichen in Frühformen invasiver Karzinome der Zervix uteri. Jena (Germany): Doktor Arbeit, Friedrich-Schiller-Universität; 2008.
17. Vercellino GF, Erdemoglu E, Kyeyamwa S, Drechsler I, Vasiljeva J, Cichon G, et al.. Evaluation of the VITOM in digital high-definition video exocolposcopy. J Lower Genit Tract Dis 2011;15:292–5.
18. Walker P, Dexeus S, De Palo G, Barrasso R, Campion M, Girardi F, et al.. International terminology of colposcopy: an updated report from the International Federation for Cervical Pathology and Colposcopy. Obstet Gynecol 2003;101:175–7.
19. Hopman EH, Kenemans P, Helmerhorst TJ. Positive predictive rate of colposcopic examination of the cervix uteri: an overview of literature. Obstet Gynecol Surv 1998;53:97–106.
20. Zahm DM, Nindl I, Greinke C, Hoyer H, Schneider A. Colposcopic appearance of cervical intraepithelial neoplasia is age dependent. Am J Obstet Gynecol 1998;179:1298–304.
21. Vercellino GF, Chiantera V, Gaßmann J, Erdemoglu E, Drechsler I, Frangini S, et al.. Prospective comparison of loop excision under colposcopic guidance vs VITOM guidance. Geburtsh Frauenheilk 2012;72:945–8.
22. Crum C. Papillomavirus-related changes and premalignant and malignant squamous lesions of the uterine cervix. In: Clement PB, Young RH, editors. Tumors and tumorlike lesions of the uterine corpus and cervix. New York (NY): Churchill Livingstone; 1993:51–83.
23. Cantor SB, Cardenas-Turanzas M, Cox DD, Atkinson EN, Nogueras-Gonzalez GM, Beck JR, et al.. Accuracy of colposcopy in the diagnostic setting compared with the screening setting. Obstet Gynecol 2008;111:7–14.
24. Coppleson M. Colposcopic features of papillomaviral infection and premalignancy in the female lower genital tract. Dermatol Clin 1991;9:251–66.
25. Sianturi R. Colposcopic index of HPV and CIN patients. Asia Oceania J Obstet Gynaecol 1993;19:127–31.
26. Nam K, Chung S, Kwak J, Cha S, Kim J, Jeon S, et al.. Random biopsy after colposcopy-directed biopsy improves the diagnosis of cervical intraepithelial neoplasia grade 2 or worse. J Lower Genit Tract Dis 2010;14:346–51.
27. Pretorius RG, Belinson JL. Colposcopy. Minerva Ginecol 2012;64:173–80.
28. McAllum B, Sykes PH, Sadler L, Macnab H, Simcock BJ, Mekhail AK. Is the treatment of CIN 2 always necessary in women under 25 years old? Am J Obstet Gynecol 2011;205:478.e1–7.
29. Wu T, Cheung TH, Yim SF, Qu JY. Clinical study of quantitative diagnosis of early cervical cancer based on the classification of acetowhitening kinetics. J Biomed Opt 2010;15:026001.
30. Shaw E, Sellors J, Kaczorowski J. Prospective evaluation of colposcopic features in predicting cervical intraepithelial neoplasia: degree of acetowhite change most important. J Lower Genit Tract Dis 2003;7:6–10.
31. Autier P, Coibion M, Huet F, Grivegnee AR. Transformation zone location and intraepithelial neoplasia of the cervix uteri. Br J Cancer 1996;74:488–90.
32. Kierkegaard O, Byralsen C, Hansen KC, Frandsen KH, Frydenberg M. Association between colposcopic findings and histology in cervical lesions: the significance of the size of the lesion. Gynecol Oncol 1995;57:66–71.
33. Hammes LS, Naud P, Passos EP, Matos J, Brouwers K, Rivoire W, et al.. Value of the International Federation for Cervical Pathology and Colposcopy (IFCPC) terminology in predicting cervical disease. J Lower Genit Tract Dis 2007;11:158–65.
34. Dexeus S, Cararach M, Dexeus D. The role of colposcopy in modern gynecology. Eur J Gynaecol Oncol 2002;23:269–77.
35. Nazeer S, Shafi MI. Objective perspective in colposcopy. Best Pract Res Clin Obstet Gynaecol 2011;25:631–40.
36. Gage JC, Hanson VW, Abbey K, Dippery S, Gardner S, Kubota J, et al.. Number of cervical biopsies and sensitivity of colposcopy. Obstet Gynecol 2006;108:264–72.
37. Monsonego J, Pintos J, Semaille C, Beumont M, Dachez R, Zerat L, et al.. Human papillomavirus testing improves the accuracy of colposcopy in detection of cervical intraepithelial neoplasia. Int J Gynecol Cancer 2006;16:591–8.