Two major pathways of recurrent high-grade squamous intraepithelial lesions of the cervix.

Abstract

Dear Sir, We read with interest the article by Herfs et al.1 reporting the outcome of 131 consecutive patients with SIL after electrosurgical excision procedure (LEEP) during a mean follow-up period of 1.7 (0.4–3.5) years. About 94.3% of patients had high-grade squamous intraepithelial lesions (HSIL) at time of primary treatment. In four patients with positive resection margins in their initial cone specimen, HSIL was detected after 3 months and interpreted as persistent disease. Both primary and persistent lesions were positive for so called squamo-columnar junction (SCJ) makers. Twelve women with complete resection of their initial HSIL developed a low-grade squamous intraepithelial lesion (LSIL) negative for SCJ markers. Herfs et al.1 propose that HSIL/invasive cervical cancer arise from a small cell population at the SCJ whereas high-risk human papillomavirus (HR-HPV) infection of the original squamous epithelium and metaplastic squamous epithelium of the transformation zone (TZ) usually produces a primary or recurrent LSIL, and that a prophylactic destruction of SCJ may be an effective prevention of SCC. In our opinion, this study has two major problems: Firstly, the short follow-up and the low number of patients reported by Herfs et al.1 are serious limitations which may account for the absence of recurrent HSIL, adenocarcinoma in situ (AIS) and invasive cervical cancer after LEEP in their series. The short observation time may be responsible for some underreporting of recurrent HSIL and does not allow the postulation of a unique recurrence pattern of squamous intraepithelial lesion (SIL) following excision of the SCJ, in particular since the results of their study contradicts those of observations with long follow-up. In a long-term study of 4,417 patients with cervical intraepithelial neoplasia (CIN) 3 after cold knife conization (CKC) with clear margins, 15 (0.35%) women developed a new HSIL after a median follow-up of 107 months (range 40–201). AIS developed in 2 (0.05%) patients 14 and 17 years after conization.2 In another long-term follow-up study on 390 patients with cervical intraepithelial neoplasia (CIN) 3 and involved margins after CKC, 306 (78%) patients remained free of CIN 3, and 84 (22%) had persisting or recurrent CIN 3 (n = 78) or developed invasive carcinoma (n = 6). Fifty-three of the 84 patients had persisting CIN 3 (diagnosed within 1 year of conization), 25/84 developed recurrent CIN 3 after a median of 3 (range 2–28) years, 5/84 developed microinvasive carcinomas (at 3, 6, 7, 12 and 23 years), and 1/84 developed a FØGO stage IB carcinoma at 8 years.3 HSIL can be treated by different techniques, including ablative techniques (e.g., cryosurgery, electrocautery, cold coagulation or laser ablation) and excisional techniques (e.g., large-loop excision of the transformation zone (LLETZ) or CKC). When excisional techniques were compared, recurrent HSIL has been reported after LLETZ up to 30 times higher than after CKC, most likely as a result of less aggressive treatments in the last two decades (LEETZ vs. CKC).2, 4 The difference of recurrent disease after complete excision of abnormal epithelium between LLETZ and CKC is probably due to the extent of removal of the sub-columnar reserve cells. Sub-columnar reserve cells as target cells of HR-HPV infection are the second issue we argue with Herfs et al.1 Martens et al.5 studied the distribution of cervical reserve cells in the region from ectocervix to lower uterine cavity in sagittal sections of 10 normal uteri histologically and immunohistochemically with antibodies to p63, bcl-2 and CK 5, 7, 8, and 17. They observed that sub-columnar reserve cells are present along the entire cervical canal. Although the highest concentration of sub-columnar reserve cells was near the SCJ, reserve cells were also located in the upper third of the cervix. These observations along with our own description of CK17 and p63 positive reserve cells in the TZ6 suggest that HR-HPV infection of sub-columnar reserve cells are responsible for primary and recurrent HSIL after excision of the SCJ. Herfs et al.7, 8 described CK7 along with other biomarkers as specific for the SCJ (so called SCJ marker). In our experience, nearly all glands underneath the metaplastic and dysplastic squamous epithelium of the TZ and the adjacent columnar epithelium consistently stain positive for CK7, while the original squamous epithelium of the cervix is negative for CK7 (Dako CK7 Clone OV-TL 12/30). Similar results were described by van der Marel et al.9 We therefore challenge the statement that CK7 is a SCJ marker and that only HR-HPV infected CK7 positive SCJ cells give rise to primary HSIL/invasive cervical cancer or recurrent HSIL/invasive cervical cancer after LEEP treatment. It is more likely that the entire cervical glandular field is at risk for the development of cancer, rather than just a few cells at the SCJ. In summary, we propose two separate pathways for development of primary and recurrent SIL: The first type is the well described and accepted pathway of HSIL developing via LSIL after HR-HPV infection of basal cells within the original and metaplastic squamous epithelium. Not all LSIL will progress to HSIL, which is particularly true for those arising in the original squamous epithelium outside the TZ. The second pathway involves HR-HPV infected columnar epithelium, either at the SCJ, at the surface or deep within glandular crypts, producing immediately a HSIL without a low-grade precursor lesion. This second pathway is not well recognized as these lesions are often misdiagnosed as atypical immature metaplasia. Correct identification is possible with immunohistochemical demonstration of p16 INK4a over-expression.6, 10 HSIL occurring after complete or incomplete excision and destruction of the TZ can develop through this pathway. Prophylactic destruction of SCJ cells will not prevent all HSIL/invasive cervical cancers. A prophylactic vaccination against HR-HPV would seem much more effective in the prevention of cervical cancer than destruction of SCJ. Yours sincerely, Olaf Reich Sigrid Regauer