Identifying Squamous Cancer at Colposcopy

During colposcopy the practitioner must determine the surface extent of the lesion; identify the most abnormal colposcopic area(s) for biopsy(ies); and identify any areas suspicious for invasive cancer. After histological reporting, correlation determines whether the lesion can be safely ablated or if it must be excised to ascertain if a microinvasive or frankly invasive cancer is present. Patients usually present with abnormal cytology suggesting squamous disease of some grade.

Cancer Clues

(1) Patient's Age

Cervical cancer is rare in patients younger than 25. The peak incidence of CIN 3 (cervical intraepithelial neoplasia grade 3) occurs in patients between ages 28-32. Microinvasive cancers and occult cancers increase in proportion as patients age. In the author's experience, only 4 percent of CIN 3 cases occurred in patients 19 or younger. Studies indicate that for every 100 CIN 3 patients, 3 cancers (3%) will be found. Half will be microinvasive and half more advanced disease.

(2) Cytology

The Bethesda System for Reporting Cervical/Vaginal Cytological Diagnosis describes abnormal squamous cells of undetermined significance (ASCUS), low grade squamous intraepithelial neoplasia/CIN 1 (LSIL), high grade squamous intraepithelial neoplasia/CIN 2-3 (HSIL), and cells reflecting a squamous cell cancer. Although cancer can be found in patients presenting with any grade of cytology, the more advanced the cytology, the more reliable the predictor of cancer. Furthermore, a glandular lesion may be present despite abnormal squamous cytology; 2-3% of CIN 3 cases confirmed histologically will have a glandular component.

(3) Linear Length and Surface Area

The linear length of CIN 3 lesions (the distance over the tissue surface between caudal and cephalad edges) varies from 2-22 mm (mean 6-10 mm). Long linear lengths, >10 mm and particularly with endocervical canal involvement, are always suspect for cancer. As the surface area increases, so does the proportion of cancers, particularly with a surface area >40 sq. mm (Figure 1).

(4) High Grade Lesions with Canal Involvement

The transformation zone recedes into and up the canal as women age. The worst disease is located centrally. In cervical sampling, the innermost/uppermost margin of the transformation zone potentially provides the worst histology regardless of its location (Figure 2).

(5) Surface Contours

Microinvasive and occult squamous cancers can produce irregular surfaces, erosions, granularity and, in more advanced disease, necrotic areas (Figure 3).

(6) Color

The color can indicate the disease process. A yellowish or grey hue can indicate necrosis, a white surface keratin, and a red one excessive vascularity (Figures 3-4).

(7) Atypical Blood Vessels

The degree of change in vasculature reflects severity of disease. Regular/irregular, and fine/coarse mosaicism and punctation indicate an intra-epithelial lesion. In microinvasive cancer the punctate and mosaic patterns become disrupted and disorganized. They start to break out of their geometric patterns, transmutating into more bizarre vessels, referred to as corkscrew, spaghetti-like, irregular coarse, irregular parallel, or comma-like vessels (Figures 4-6). Other atypical forms have nonbranching vessels that bulge and contract. They sometimes appear within an area of uneven surface contour because these vessels are supporting an active, proliferating tumor (Figure 3).

(8) Colposcopic Grading Systems

Different colposcopic grading systems deal with squamous lesions in terms of angioarchitecture, surface patterns, acetowhiteness, and demarcation of the lesion (using acetic acid and iodine solutions). All indicate that the more severe the lesion, the more pronounced are the findings, and that any atypical vessels should lead to the suspicion of malignancy. None address glandular lesions.

(9) Persistence/Recurrence After Treatment

When HSIL disease persists or recurs after treatment, colposcopy is frequently unsatisfactory. Any suspicion that a malignancy may be present warrants some method of conization even if the transformation zone is colposcopically identifiable. Excision is always required in cases of unsatisfactory colposcopy.

(10) Colposcopic Mimics

The colposcopic mimics of malignancy usually relate to surface and vasculature. Common look-alikes are condyloma, post-radiation changes (Figure 7), cervical or endometrial polyps, decidual tissue, and fibroids prolapsed into or through the endocervical canal.

Summary

Proper colposcopic assessment means: 1) studying vascular patterns using the colored filter first; 2) determining satisfactory versus unsatisfactory patterns; 3) grading the lesion; 4) selecting biopsy site(s); and 5) correlating cytology, colposcopy, and histology. This process should, in most cases, identify or cause one to suspect even the earliest cancer.

Figure 1. The linear length of this HSIL measures 20 mm on the ectocervix and it extends into the endocervical canal. The canal extension potentially represents the worst histology. Lesions with long linear lengths are more likely to be cancer than smaller ones.

Figure 2. A large, elevated, dense acetowhite and well demarcated lesion exists on the ectocervix and extends into the endocervical canal.

Figure 3. This invasive squamous cell cancer demonstrates thickened acetowhite epithelium with surface ulcerations.

Figure 4. A high magnification view of an invasive squamous cell cancer that has a yellowish hue. Numerous waste thread-like blood vessels are apparent.

Figure 5. This geometric pattern is breaking out of its mosaic form into the irregular, tortuous vessels (waste thread and corkscrew) of an invasive squamous cell cancer.

Figure 6. These vessels demonstrate numerous corkscrew and waste thread formations in an invasive squamous cancer recurrent 20 years following radiation.

Figure 7. A cervix previously irradiated for cancer. The corkscrew and waste thread vessels mimic a squamous cell cancer.

References:

Recommended Reading

1. Abdul-Karium FW, Fu YS, Reagan JW, et al. Morphometric study of intraepithelial neoplasia of the uterine cervix. Obstet Gynecol 1982; 60:210.

2. Benedet JL, Anderson GH, Boyes DA. Colposcopic accuracy in the diagnosis of microinvasive and occult invasive carcinoma of the cervix. Obstet Gynecol 1985; 65:557.

3. Coppleson M, Dalrymple JC, Atkinson KH. Colposcopic differentiation of abnormalities arising in the transformation zone. In Wright VC (Ed): Contemporary Colposcopy. Philadelphia: WB Saunders, 1993; p. 99.

4. Gusberg SB, Moore DB. The clinical pattern of intraepithelial carcinoma of the cervix and its pathological background. Obstet Gynecol 1953; 2:1.

5. Kolstad P, Stafl A. Atlas of Colposcopy, Third Edition. Oslo: Universitetsforlaget, 1982, p. 33.

6. Przybora LA, Plutowa A. Histological topography of carcinoma in situ of the cervix uteri. Cancer 1959; 12:268.

7. Wright VC, Riopelle MA. The geometry of cervical intraepithelial neoplasia as a guide to its eradication. Cervix 1986; 4:21.

Dr. Wright is a gynecologic oncologist and professor, Department of Obstetrics and Gynaecology, The University of Western Ontario, London, Ontario, Canada. He is principal colposcopist at the St. Joseph's Health Centre Colposcopy Clinic. Dr. Wright is a former president and founding member of the Society of Canadian Colposcopists. He was recently awarded the American Society of Colposcopy and Cervical Pathology's Distinguished Scientific Achievement Award for his life-long academic contributions to the understanding and treatment of lower genital tract disease.