Frozen section examination of the endocervical margin of cervical conization specimens

Available online at www.sciencedirect.com R

Gynecologic Oncology 90 (2003) 305–309

www.elsevier.com/locate/ygyno

Frozen section examination of the endocervical margin of cervical conization specimens Roman Rouzier,a,* Estelle Feyereisen,a Elisabeth Constancis,b Bassam Haddad,a Philippe Dubois,a and Bernard-Jean Paniela a

Department Obstetrics and Gynecology, Centre Hospitalier Intercommunal de Cre´teil, 40 avenue de Verdun, 94010 Cre´teil Cedex, France b Department of Pathology, Centre Hospitalier Intercommunal de Cre´teil, 40 avenue de Verdun, 94010 Cre´teil Cedex, France Received 21 August 2002

Abstract Objective. We conducted this retrospective study to determine accuracy of frozen section examination of endocervical margin during cold knife conization. Methods. Between June 1993 and June 2001, 310 consecutive patients underwent cervical conization for squamous intraepithelial lesion or stage IA1 cervical cancer. Before 1997, the surgical specimens of 149 patients were processed following a standard pathological procedure (historical group). After 1997, a frozen section of the upper endocervical margin was processed during surgery for 161 patients. If the upper endocervical margin was involved with intraepithelial neoplasia, the surgeon performed a second resection if possible. Results of the frozen section examination were compared with the final diagnoses to determine sensitivity, specificity, and positive and negative predictive values. The usefulness of this procedure was evaluated by comparison of positive margin status rate with the one of the historical control group. Results. For the diagnosis of intraepithelial neoplasia involving the endocervical margin, the sensitivity, specificity, and positive and negative predictive values of frozen section were 91%, 100%, 100%, and 98%, respectively. Eleven patients had definitive positive endocervical margin in the frozen section group (three false negatives, six patients without additional resection, and two patients with intraepithelial neoplasia involving the upper margin of the additional resection) and 17 patients in the historical group (P ⫽ .16). Conclusion. Frozen section examination of the endocervical margin of cervical specimen obtained during cold knife conization is highly accurate. Its clinical relevance has to be demonstrated in a multicenter study. © 2003 Elsevier Science (USA). All rights reserved. Keywords: Cervical intraepithelial neoplasia; Endocervical margin; Frozen section examination

Introduction Cervical conization is one of the recognized treatments of high-grade cervical intraepithelial neoplasia. High success rates of conization of the transformation zone have been reported in many studies. A cone margin positive for cervical neoplasia usually indicates the presence of residual disease in the remaining cervix [1]. During postconization hysterectomy, the prevalence of residual disease is 8%– * Corresponding author. Department of Surgery, Institut Gustave Roussy, 39 rue Camille Desmoulins, 94805 Villejuif Cedex, France. Fax: ⫹33-1-42-11-52-13. E-mail address: [email protected] (R. Rouzier).

84.8% [1]. New methods to achieve negative margins on cone specimen are still necessary. Several studies have reported frozen section examination to be a reliable method to rule out invasive cancer before hysterectomy [2–7]. To our knowledge, no data are available concerning the accuracy of intraoperative assessment of endocervical margin status of conization specimen by frozen section and we report our experience.

Materials and methods The records of 310 consecutive patients who underwent cervical conization for squamous intraepithelial lesion or

0090-8258/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0090-8258(03)00324-X

306

R. Rouzier et al. / Gynecologic Oncology 90 (2003) 305–309

Table 1 Clinical and pathologic characteristics of the study group consisted of the frozen section group and the historical group (standard pathologic examination group)

Period Number Age (mean ⫾ SD) Menopausal status Premenopausal Menopausal Preoperative biopsy CIN I CIN II/CIN III/IA1 Transformation zone fully visualized Yes No Specimen height (mean ⫾ SD) Specimen width (mean ⫾ SD) Operative time (mean ⫾ SD)

Historical group

Frozen section group

1993–1997 149 42.0 years ⫾ 1.7

1997–2001 161 42.8 years ⫾ 1.6

ns

133 (89%) 16 (11%)

134 (83%) 27 (17%)

ns

24 (16%) 125 (84%)

27 (17%) 134 (83%)

ns

103 (70%) 46 (30%) 21 mm ⫾ 1 28 mm ⫾ 1 42 min ⫾ 5.5

98 (61%) 63 (39%) 22 mm ⫾ 1 29 mm ⫾ 1 48 min ⫾ 3.1

ns ns ns .049

stage IA1 cervical cancer in our department between June 1993 and June 2001 were retrospectively reviewed. Patients with adenocarcinoma in situ were excluded. Indications for conization were persistent cervical intraepithelial neoplasia (CIN) I or CIN I and transformation zone not fully visualized for 51 patients and presence of CIN II or III for 259 patients. Prior to conization, Lugol’s iodine was used to delineate abnormal epithelium. All the cone biopsies were performed with a scalpel, taking care to avoid trauma to the specimen. All specimens were marked for correct orientation. In the pathology department, the apical portion (approximately 5 mm thick) of the conization specimen was amputated from the major portion of the cone. The remaining specimen was cut longitudinally at 3 mm intervals. The samples were embedded in paraffin in toto and serially sectioned. After 1997, a frozen section of the upper endocervical margin was processed as follows for 161 patients: the caudal part of the apical portion separated from the cone specimen was cut into a thin slice approximately 2 mm thick. The piece was prepared in water-soluble embedding material and 10 ␮m sections were obtained with the cryostat and stained with the toluidine blue. The pathologist examined the entire circumference of the endocervical canal and indicated to the surgeon the endocervical margin status. If either low-grade or high-grade intraepithelial neoplasia was present on the frozen section, the surgeon performed a second resection if possible. Reparative procedure was begun during frozen section examination: hemostasis was achieved and vagina and cervix were reapproximated using anterior and posterior running sutures. The study group consisted of the 161 patients with frozen section examination of the upper margin while the 149 patients operated on before 1997 and who underwent a standard pathologic examination of the conization specimen comprised the historical group. A margin was considered positive when intraepithelial neoplasia, regardless of grade, was present at the margin of the specimen. The diagnoses

P

reported to the surgeon after examination of the frozen section material were compared with the final diagnoses obtained after examination of the paraffin-fixed specimen of the apex of the cone to determine sensitivity, specificity, and positive and negative predictive values. The usefulness of frozen section analysis of the endocervical margin was evaluated by comparison of definitive margin status between the two groups: standard pathologic procedure group (historical) and frozen section procedure group. All continuous data were expressed as means ⫾ standard deviation (SD) and compared by Student’s t test. A ␹2 test was used for comparative analysis of proportions. Results were considered statistically significant if P ⬍ 0.05.

Results Age, preoperative biopsies data, visualization of the transformation zone rates, mean specimen height and width, and mean operative time according to the pathologic procedure are given in Table 1. Frozen section significantly increased operative time (⫹6 min). The additional time was only 6 min because reparative procedures were begun before frozen section results. The final pathological diagnosis, defined as the most advanced lesion on the specimen, was Table 2 Comparison of frozen and final pathology of conization specimen Final pathology of endocervical margin

Frozen section examination Negative CIN I CIN II/CIN III/IA 1 Historical group

Negative

CIN I

CIN II/CIN III/ IA 1

128

0 2

132

2

3 9 19 15

R. Rouzier et al. / Gynecologic Oncology 90 (2003) 305–309 Table 3 Sensitivity, specificity, and positive and negative predictive value of frozen section examination of the endocervical margin of conization specimen N Sensitivity Specificity Positive predictive value Negative predictive value

%

30/33 128/128 30/30 128/131

91 100 100 98

CIN I for 36 patients, CIN II or III for 250 patients, and stage IA1 cervical cancer for 22 patients. No squamous intraepithelial lesions were found on two specimens. Table 2 lists the frozen and final pathologic diagnoses for the 161 patients who had frozen section evaluation of the upper margin of the specimen. Only three false negative results were recorded, but final pathologic results revealed highgrade intraepithelial neoplasia in nine of the 11 patients with low-grade intraepithelial neoplasia at frozen section evaluation. For the diagnosis of intraepithelial neoplasia involving the endocervical margin, the sensitivity, specificity, and positive and negative predictive values of frozen section evaluation were 91%, 100%, 100%, and 98%, respectively (Table 3). Twenty-four of the 30 patients with positive margin at frozen section had additional resection. Reasons for not performing additional resection were pregnancy (one case), concomitant anal carcinoma (one case), multiple genital intraepithelial neoplasia (one case) and additional resection requiring hysterectomy because all the cervix had been removed (three cases). On the additional resections, frozen section examination of the endocervical margin had been performed on 14 specimens; they were all negative. For the other patients, frozen section examination of the endocervical margin was not performed because the whole cervix had been removed. On the final pathologic analyses of second resection, intraepithelial neoplasia was diagnosed on six specimens, whereas 18 did not show any evidence of intraepithelial neoplasia. Finally, there were 11 patients with definitive positive margins in the frozen section group: three false negative results of frozen section examination, six patients without additional resection, and two patients with intraepithelial neoplasia involving the upper margin of the additional resection (no frozen section examination of the additional resection specimen). In the standard pathologic procedure group (historical), 17 patients had involved

307

upper margin (Table 4). The difference was not significant (P ⫽ .16) but the power of the test was under .80. Clinical outcomes are reported in Table 5; we did not perform any reexcisions, but eight patients had completion hysterectomy for involved margins (six in the historical group and two in the frozen section group). Residual disease was found on only two hysterectomy specimens (historical group). A recurrence/persistence of intraepithelial neoplasia at 6 months was observed for seven patients, four in the historical group and three in the frozen section group.

Discussion The use of frozen section techniques has greatly impacted on the care of gynecologic patients. The ability to distinguish malignant from benign processes intraoperatively on histologic ground allows the more appropriate selection of surgical procedures. Determination of margin status by frozen section has been reported in breast [8,9] and prostate [10] carcinomas but, to our knowledge, no data are available for cervical cone biopsy specimens. With sensitivity, specificity, and positive and negative predictive values of, respectively, 91%, 100%, 100%, and 98%, our study demonstrates that frozen section evaluation of the endocervical margin of cervical conization specimen is highly accurate. Several studies regarding cone biopsy and recurrence have demonstrated that follow-up colposcopy and cytologic examination may be more reliable than margin status in predicting recurrence because the majority of patients with positive squamous intraepithelial neoplasia on margin will resolve their disease [11–13]. Other authors have suggested that human papillomavirus testing is potentially an effective tool in predicting residual intraepithelial neoplasia after conization and could assist in the decision between hysterectomy and conservative follow-up in women with CIN III [14,15]. Jain et al. have reported a prospective analysis undertaken on 79 cone biopsies of women with high-grade lesions [14]. Forty-seven of these patients underwent a hysterectomy for positive margins. None of the patients with negative postconization human papillomavirus testing had residual disease on hysterectomy specimen. In most of the studies, residual disease [1] and risk of recurrence [16 –18] are correlated with margin status. In the report of Andersen et al., when excisional margins were involved, recurrence

Table 4 Final margin status of standard procedure pathologic group and frozen section procedure group Endocervical margin

Historical group Frozen section group

Ectocervical margin

Negative n (%)

Positive n (%)

132 (89%) 150 (93%)

17 (11%) 11 (7%)

P

.16

Negative n (%)

Positive n (%)

137 (92%) 153 (95%)

12 (8%) 8 (5%)

P

.26

308

R. Rouzier et al. / Gynecologic Oncology 90 (2003) 305–309

Table 5 Clinical outcomes

Completion hysterectomy for involved margins Residual disease n(%) Recurrence/persistence of intraepithelial neoplasia at 6 months Treatment of recurrent/persistent intraepithelial neoplasia

Historical group

Frozen section group

P

6 2 (33%) 4 One reconization and Three hysterectomies

2 0 (0%) 3 Three hysterectomies

ns

was observed in 8.7% of cases, compared to 2.3% when margins were uninvolved [16]. In our study, positive endocervical margin rates were 11% and 7%, respectively, without and with frozen section evaluation of the upper margin. The difference between the two groups is not significant because of a lack of statistical power partly due to the low rate of positive margin in the control group. Consequently, demonstrating the impact of frozen section examination of the endocervical margin in terms of recurrence is not conceivable because of the same lack of statistical power. Even if the clinical impact of the frozen section examination is difficult to demonstrate, converting patients to a diseasefree status remains a valid concept. One could argue that accomplishing frozen section may damage the specimen. We did not find that frozen section had damaged any of the specimen examined. The consistence of cervix and the fact that thickness of the part taken from the apical portion of the cone specimen may be adapted to its diameter may explain that no difficulties have been encountered. On the other hand, cervical dilatation and/or reparative procedure to reapproximate vagina and cervix may damage remaining endocervival mucosa after conization. Consequently, frozen section and definitive pathologic examination of reexcision specimen are far more difficult. Concerning the surgical aspect of reexcision, we did not encounter any difficulties, whereas complete hysterectomy is associated with an increased risk of bladder injury [19]. Ectocervical margin was positive in 8(5%) of the 161 patients who had frozen section of the upper margin. Performing frozen section examination of the endocervical margin, whereas ectocervical margin is possibly involved, may be a cause of concern. The residual or recurrent intraepithelial neoplasia actually is associated with the location of the involved margin; many studies have reported that endocervical margin involvement is more frequent and more strongly correlated with residual disease than ectocervical involvement [20 –23], suggesting that margin status is more important for endocervix than for ectocervix. Moreover, careful lugol’s iodine application and attention to ectocervix incision may provide clear ectocervical margins. Additional resection was not performed in six of the 30 patients with intraepithelial neoplasia involving the upper margin at frozen section examination. These cases increase the definitive rate of positive margin and demonstrate the difficulty to achieve a negative margin status in extended cervical intraepithelial neoplasia more than inadequacy or

ns

uselessness of frozen section examination of the upper margin. Some authors focused on factors which may be used to predict residual intraepithelial neoplasia after conization. Husseinzadeh et al. advocated that endocervical curettage should be performed routinely on every patient undergoing conization in order to base subsequent management on histologic findings of the cone margins and postcone endocervical curettage [24]. Thirty-nine of 63 patients (62%) with CIN involving the endocervical margin of the cone biopsy had residual disease. When both the inner margin and postcone endocervical curettage (ECC) were involved, residual disease was present in 80% of hysterectomy specimens compared to 36% of hysterectomy specimens when the inner margins of the cone were involved with CIN-III and the postcone ECC was negative. ECC was not performed routinely in our study because it may alter quality of additional specimen in case of involved upper margin at frozen section. It may be performed if endocervical margin is negative at frozen section. In conclusion, frozen section examination of the endocervical margin of cervical specimen obtained during cold knife conization is highly accurate. Its clinical relevance has to be demonstrated in a multicenter study.

References [1] Jakus S, Edmonds P, Dunton C, King SA. Margin status and excision of cervical intraepithelial neoplasia: a review. Obstet Gynecol Surv 2000;55:520 –7. [2] Gupta RK. Frozen sections in cervical conization. A cytohistopathologic approach in the early diagnosis of carcinoma cervix. Obstet Gynecol 1971;38:248 –50. [3] Hannigan EV, Simpson JS, Dillard EA Jr, Dinh TV. Frozen section evaluation of cervical conization specimens. J Reprod Med 1986;31: 11– 4. [4] Hoffman MS, Collins E, Roberts WS, Fiorica JV, Gunasekaran S, Cavanagh D. Cervical conization with frozen section before planned hysterectomy. Obstet Gynecol 1993;82:394 – 8. [5] Kaufman RH. Frozen section evaluation of the cervical conization specimen. Clin Obstet Gynecol 1967;10:838 –52. [6] Neiger R, Bailey SA, Wall AM 3rd, Jennings JB, Gallup DG. Evaluating cervical cone biopsy specimens with frozen sections at hysterectomy. J Reprod Med 1991;36:103–7. [7] Woodford HD, Poston W, Elkins TE. Reliability of the frozen section in sharp knife cone biopsy of the cervix. J Reprod Med 1986;31: 951–3.

R. Rouzier et al. / Gynecologic Oncology 90 (2003) 305–309 [8] Weber S, Storm FK, Stitt J, Mahvi DM. The role of frozen section analysis of margins during breast conservation surgery. Cancer J Sci Am 1997;3:273–7. [9] Noguchi M, Minami M, Earashi M, Taniya T, Miyazaki II, Mizukami Y, Nonomura A, Nishijima H, Takanaka T, Kawashima H, Saito Y, Takashima C, Nakamura S, Michigishi T, Yokoyama K. Pathologic assessment of surgical margins on frozen and permanent sections in breast conserving surgery. Breast Cancer 1995;2:27–33. [10] Shah O, Melamed J, Lepor H. Analysis of apical soft tissue margins during radical retropubic prostatectomy. J Urol 2001;165:1943– 8. [11] Buxton EJ, Luesley DM, Wade-Evans T, Jordan JA. Residual disease after cone biopsy: completeness of excision and follow-up cytology as predictive factors. Obstet Gynecol 1987;70:529 –32. [12] Naumann RW, Bell MC, Alvarez RD, Edwards RP, Partridge EE, Helm CW, Shingleton HM, McGee JA, Higgins RV, Hall JB. LLETZ is an acceptable alternative to diagnostic cold-knife conization. Gynecol Oncol 1994;55:224 – 8. [13] Flannelly G, Langhan H, Jandial L, Mana E, Campbell M, Kitchener H. A study of treatment failures following large loop excision of the transformation zone for the treatment of cervical intraepithelial neoplasia. Br J Obstet Gynaecol 1997;104:718 –22. [14] Jain S, Tseng CJ, Horng SG, Soong YK, Pao CC. Negative predictive value of human papillomavirus test following conization of the cervix uteri. Gynecol Oncol 2001;82:177– 80. [15] Bekassy Z, Ahlgren M, Eriksson M, Lindh E. Carbon dioxide laser miniconization for treatment of human papillomavirus infection associated with cervical intraepithelial neoplasia. Acta Obstet Gynecol Scand 1995;74:822– 6.

309

[16] Andersen ES, Pedersen B, Nielsen K. Laser conization: the results of treatment of cervical intraepithelial neoplasia. Gynecol Oncol 1994; 54:201– 4. [17] Livasy CA, Maygarden SJ, Rajaratnam CT, Novotny DB. Predictors of recurrent dysplasia after a cervical loop electrocautery excision procedure for CIN-3: a study of margin, endocervical gland, and quadrant involvement. Mod Pathol 1999;12:233– 8. [18] Andersen ES, Nielsen K, Larsen G. Laser conization: follow-up in patients with cervical intraepithelial neoplasia in the cone margin. Gynecol Oncol 1990;39:328 –31. [19] Tancer ML. Observations on prevention and management of vesicovaginal fistula after total hysterectomy. Surg Gynecol Obstet 1992; 175:501– 6. [20] Lapaquette TK, Dinh TV, Hannigan EV, Doherty MG, Yandell RB, Buchanan VS. Management of patients with positive margins after cervical conization. Obstet Gynecol 1993;82:440 –3. [21] Lopes A, Morgan P, Murdoch J, Piura B, Monaghan JM. The case for conservative management of “incomplete excision” of CIN after laser conization. Gynecol Oncol 1993;49:247–9. [22] Murdoch JB, Morgan PR, Lopes A, Monaghan JM. Histological incomplete excision of CIN after large loop excision of the transformation zone (LLETZ) merits careful follow up, not retreatment. Br J Obstet Gynaecol 1992;99:990 –3. [23] Paterson-Brown S, Chappatte OA, Clark SK, Wright A, Maxwell P, Taub NA, Raju KS. The significance of cone biopsy resection margins. Gynecol Oncol 1992;46:182–5. [24] Husseinzadeh N, Shbaro I, Wesseler T. Predictive value of cone margins and post-cone endocervical curettage with residual disease in subsequent hysterectomy. Gynecol Oncol 1989;33:198 –200.