Ureteral Frozen Section Analysis During Cystectomy: A Reassessment

0022-5347/96/1554-1218$03.00/0 THE JOURNAL OF UROLOGY Copyright 0 1996 by AMErUrAN UROLOGICAL ASSOCIATION,

Vol. 155, 1218-1220, April 1996 Printed in U.SA.

INC

URETERAL FROZEN SECTION ANALYSIS DURING CYSTECTOMY: A REASSESSMENT MARK P. SCHOENBERG,* H. BALLENTINE CARTER

AND

JONATHAN I. EPSTEIN

From the James Buchanan Brady Urological Institute and Department of Pathology, Johns Hopkins Medical Institutions, Baltimore, Maryland

ABSTRACT

Purpose: Frozen section analysis of the distal ureteral margins is commonly performed at cystectomy to exclude involvement of tumor in the retained ureter. We reviewed our experience with 101 consecutive cystectomies to determine the contemporary incidence and clinical significance of the urothelial abnormalities detected by frozen section analysis performed at operation. Materials and Methods: The pathology reports for 101 consecutive patients treated with nerve sparing cystectomy between 1982 and 1989 were reviewed. Frozen section and final ureteral analyses were compared. Results: Of the patients 8%had evidence of a urothelial abnormality ranging from mild atypia to frank carcinoma in situ involving the distal ureters on frozen section. Only 4 patients had documented carcinoma in situ at the final margin and all 4 ultimately died of disease. The frozen section false-positive and false-negative rates were 2 and 6%, respectively. In 6 patients with ureteral urothelial abnormalities documented on frozen section ureterointestinal anastomosis was performed despite persistent abnormalities a t the ureteral margins, frequently after multiple frozen analyses failed to clear the margins definitively. None of the 6 patients in this group experienced upper tract recurrence during a mean followup of 41 months. Conclusions: These data suggest that routine frozen section analysis of the ureteral margins at cystectomy may not be necessary for most patients undergoing cystectomy. KEYWORDS:bladder neoplasms, ureteral neoplasms, cystectomy, frozen sections Radical cystectomy remains the gold standard for management of high grade recurrent superficial and muscle invasive bladder cancer. Durable post-cystectomy survival has been reported by numerous groups for patients with organ-confined disease.'-6 During cystectomy, frozen section analysis is traditionally performed to exclude possible invasion of the distal ureters by tumor. Frozen section analysis was popularized in the late 1960s by investigators who reasoned that the presence of urothelial abnormalities at the ureteral margins would increase the risk of local and distant recurrence if these contaminated margins were not excised at operation.?-" Although this supposition is logical, the urological literature contains scant data to support this routine practice and no substantial proof that contaminated ureteral margins themselves pose a significant long-term threat to patients undergoing cystectomy for cure of bladder cancer. To address this issue and to ascertain the ability of frozen section analysis of the ureters at cystectomy to provide useful clinical information, we reviewed our experience with 101 patients who underwent nerve sparing cystectomy for clinically organ-confined bladder cancer. MATERIALS AND METHODS

Between March 1982 and November 1989,101 nerve sparing radical cystoprostatectomies were performed at our hospital for treatment of carcinoma of the bladder. Median followup for the entire group was 67 months (range 3 to 148). Mean patient age was 59 years (range 25 to 78). Of the patients 50 received additional preoperative radiation therapy, or neoadjuvant or adjuvant chemotherapy. The technique of nerve sparing cystectomy has been described previAccepted for publication October 27, 1995. * Re uests for reprints: Department of Urology, Marburg Bldg. 145,d n s Hopkins Hospital, 600 North Wolfe St., Baltimore, Mary. land 21287-6101.

ously in detai1.12-16 Routine practice at our department has been to send a portion of each pelvic ureter for frozen section analysis before performing ureteroenteric anastomosis as part of urinary diversion or reconstruction of the lower urinary tract. Frozen section analyses were performed in 97 of the 101 patients studied. The pathology report from each cystectomy was reviewed, and the results of the permanent and frozen section analyses were compared. Our current method of reporting urothelial atypia at the time of frozen sections of ureters has been modified. We currently do not record mild urothelial atypia, since this condition is subjective and clinically insignificant. Rather, this degree of atypia is reported as normal. We also generally combine the classification of marked atypia with carcinoma in situ. The difference between marked atypia and carcinoma in situ is subjective and clinically both are believed to have the same implications. Any degree of significant atypia short of carcinoma in situ is still reported as moderate at our institution. We currently described urothelial histology as benign, or reflecting the presence of moderate atypia or carcinoma in situ. Actuarial 10-year survival data for these patients have recently been obtained and were used to evaluate the pathological findings.16 RESULTS

Accuracy of frozen section. Of the 97 patients evaluated 8

had frozen sections that revealed evidence of urothelial abnormalities ranging from mild atypia to carcinoma in situ and invasive cancer (see table). In 2 of these 8 cases final histological analysis failed to demonstrate the urothelial abnormality identified at frozen section (2 of 97, for a 2% false-positive rate). Six patients had negative frozen sections but positive final histological analyses of the ureteral segments examined, for a 6% false-negative rate. Of these 6 patients 3 had mild and 2 had marked urothelial atypia on

1218

URETERAL FROZEN SECTION ANALYSIS DURING CYSTECTOMY Patient data

1219

short. Sharma et a1 reviewed the Memorial Hospital experience and discovered carcinoma in situ in 17 of 205 patients Pt. No. Stage Followup (mos.) Frozen Section (8.5%)examined.* They reported that ureteral margin carcinoma in situ was most frequently observed in patients with 1 P3B/N+ Dead of disease Marked atypia Ca in situ multifocal tumors, and those with high stage and high made (3) disease. Of the 17 patients with carcinoma Fn situ reported on PO No evidence of No tumor seen Mild atypia 2 by Sharma et a1 1 had uretero-ileal recurrence. Cooper et a1 disease (65) described a group of 45 patients who underwent cystectomy No evidence of Moderate P3B 3 No tumor seen disease (74) atypia and found 8 (18%)with epithelial abnormalities at the ureP3B 4 No evidence of No tumor seen Mild atypia teral margins, again associated with high grade and stage of disease (72) disease.9 Three of 13 patients with carcinoma in situ a t the Dead of disease No tumor seen Ca in situ PIS 51 (381 ureteral margin had upper tract disease after cystectomy and 6 Dead of disease Mild atypia P3A Mild atypia required nephroureterectomy. Cooper et a1 suggested that (8) frozen section analysis at cystectomy would facilitate more PIS 7 No evidence of No tumor seen Marked atypia thorough removal of tumor and decrease the likelihood of disease (67) PIS Dead of disease Moderate 8 upper tract recurrence. In a followup review of the Memorial Moderate (29) atypia atypia experience, Linker and Whitmore suggested that ureteral P3Bm + No evidence of Mild atypia 9 No tumor seen carcinoma in situ encountered a t cystectomy probably had disease (81) little overall effect on the clinical outcome of the patients 10* PIS Dead of disease Moderate Moderate (54) atypia atypia studied, and that a conservative approach with scrupulous P3BN + No evidence of No tumor seen Mild atypia I1 clinical followup was probably most appropriate for these disease (89) patients." 12' P3BlN + Dead of disease Ca in situ Ca in situ Johnson et a1 at M. D. Anderson Hospital reviewed 403 (6) 13 P3A No evidence of Mild atypia Moderate cystectomies performed between 1969 and 1985, and discovdisease (76) atypia ered unsuspected malignant disease on ureteral frozen sec14 P3B/N+ Dead of disease No tumor seen Marked atypia tions in 8 patients (2%).'7Among the 24 patients undergoing (7111 ureterenteric anastomosis despite the presence of persistent * Local recurrence. urothelial dysplasia or evidence of malignant disease at the ureteral margin, only 1had clinically recognized tumor at the final analysis. One patient had unappreciated carcinoma in anastomotic site within the 6-year followup. The authors situ at the ureteral margin that was not diagnosed by frozen concluded that frozen section analysis of the ureters was only section analysis. He subsequently had pelvic recurrence, dis- indicated in high risk situations (diffuse carcinoma in situ or tant metastases and radiographic evidence of upper urinary prostatic ductal involvement by tumor). Our experience suggests that the benefit of ureteral frozen tract recurrence before death from disease 38 months postsection analysis may have been overestimated by earlier operatively. Diagnosis of carcinoma in situ and marked atypia. One investigators. Although patients with high grade and high patient had carcinoma in situ and 1 had marked urothelial stage disease appear more likely to have ureteral involveatypia a t the ureteral margin on frozen section. Both patients ment, they are also patients in whom surgery alone is unhad high stage disease and died of metastatic bladder cancer likely to affect cure no matter how extensive the operation. In at 3 and 6 months postoperatively (see table). Neither patient patients with diffuse carcinoma in situ the risk of ureteral involvement may be greater than average, and we cannot had evidence of upper tract recurrence at death. Diagnosis of mild or moderate atypia. Among the 8 pa- argue strongly against frozen sections in this group. Simitients with positive frozen sections 6 had mild to moderate larly, frozen section analysis may be useful in the unusual atypia at the ureteral margins, of whom 3 are alive and 3 died patient in whom induration or frank tumor infiltration of the of disease (see table). The degree of atypia does not appear to distal ureter is discovered unexpectedly at operation. Pabe related to extent of disease or clinical outcome in this tients with transitional cell carcinoma of the prostate have a small group. generally poor prognosis, largely because of an increased risk Ureteral status at anastomosis. Six of the 97 patients stud- of metastatic disease. Although obtaining tumor-free ureied had persistent urothelial abnormalities after frozen sec- teral margins in this high risk group may seem sensible, the tion analysis and underwent ureteroenteric anastomosis de- development of effective systemic chemotherapy may provide spite these findings. In addition, 6 patients had apparently a substantially greater survival benefit to these patients negative frozen sections but urothelial abnormalities ranging than the absolute eradication of distal ureteral urothelial from atypia to carcinoma in situ were diagnosed at final atypia. Our patient group was highly selected, and the surhistological analysis. Of the 14 patients in this group none vival and negative ureteral margin rates may reflect a n inexperienced isolated recurrence at the ureteral anastomosis, herent selection bias. Nonetheless, the data imply that frozen and only 1 with carcinoma in situ at the final ureteral margin section analysis of the ureters does not often provide addihad radiographic evidence of upper tract carcinoma during tional useful clinical information in our patients. followup. Finally, contemporary urological surgeons must face the sobering fact that their every clinical decision will now be DISCUSSION scrutinized by third party payers who can be expected to Investigators 25 years ago theorized that frozen section require data rather than intuition to justify expenditures analysis of the ureteral margins a t cystectomy would facili- related to medical practice. Frozen section analysis of the tate complete removal of tumor, which would in turn trans- ureters at our institution costs approximately $800 for the late into improved survival of bladder cancer patients. c u b evaluation of 2 ureteral margins with added charges for each et a1 first addressed this issue after discovering that 38 of231 additional section. In many of our patients multiple frozen patients (17%)had unanticipated ureteral epithelial abnor- sections were performed during cystectomy. We are unable to malities at cystectomy.7 They suggested that unappreciated document any case of ureteral anastomotic site recurrence carcinoma in situ could result in recurrence in the remaining that resulted from a contaminated ureteral margin during ureter and renal pelvis, although their series was small, followup. Only 1 patient had upper tract disease in our semany of the patients had advanced disease and followup was ries. Final Histological Analysis

TNM

_-"I

1220

URETERAL FROZEN SECTION ANALYSIS DURING CYSTECTOMY CONCLL'SIONS

Our data. coupled with observations from M. D. Anderson Hospital, s u g g e s t that ureteral frozen section a n a l y s i s i s probably unnecessary i n the majority of p a t i e n t s undergoing cystectomy. In p a t i e n t s at high risk for recurrence, s u c h as those with diffuse carcinoma i n s i t u or prostatic ductal invasion b y t u m o r , frozen section analysis m a y be helpful in tailoring the operation. REFERENCES

1. Montie, J. E.. Straffon, R. A. and Stewart, B. H.: Radical cystectomy without radiation therapy for carcinoma of the bladder. J. Urol., 131: 477, 1984. 2. Skinner, D. G. and Lieskovsky, G . : Contemporary cystectomy with pelvic node dissection compared to preoperative radiation therapy plus cystectomy in management of invasive bladder cancer. J. Urol.. 131: 1069, 1984. 3. Pagano, F., Bassi, P., Galetti, T. P., Meneghini, A,, Milani, C., Artibani, W. and Garbeglio, A,: Results of contemporary radical cystectomy for invasive bladder cancer: a clinicopathological study with an emphasis on the inadequacy of the tumor, nodes and metastases classification. J. Urol., 145: 45, 1991. 4. Brendler, C. B., Steinberg, G. D., Marshall, F. F., Mostwin, J. L. and Walsh, P. C.: Local recurrence and survival following nerve-sparing cystoprostatectomy. J. Urol., 144: 1137, 1990. 5. Waehre, H., Ous, S., Klevmark, B., Kvarstein, B., Urnes, T., Ogreid. P.. Johansen, T. E. and Fossa, S. D.: A bladder cancer multi-institutional experience with total cystectomy for muscle-invasive bladder cancer. Cancer, 7 2 3044, 1993. 6. Soloway, M. S., Lopez, A. E., Patel, J. and Lu, Y.: Results of radical cystectomy for transitional cell carcinoma of the bladder and the effect of chemotherapy. Cancer, 73: 1926, 1994. 7. Culp, 0. S.. Utz. D. C. and Harrison, E. G., Jr.: Experiences with ureteral carcinoma in situ detected during operations for ves-

ical neoplasm. J. Urol.. 97: 679, 1967. 8. Sharma, T. C., Melamed, M. R. and Whitmore, W. F.. Jr.: Car. cinoma in-situ of the ureter in patients with bladder carci. noma treated by cystectomy. Cancer, 26: 583, 1970. 9. Cooper, P. H., Waisman, J., Johnston, W. H. and Skinner, D. G,: Severe atypia of transitional epithelium and carcinoma of the urinary bladder. Cancer. 31: 1055. 1973. 10. Skinner, D. G., Richie, J. P.. Cooper, p. H.. Waisman, J. and Kaufman, J. J.: The clinical significance of carcinoma in situ of the bladder and its association with overt carcinoma. J . Urol., 112: 68, 1974. 11. Linker, D. G. and Whitmore. W. F.: Ureteral carcinoma in situ, J. Urol., 113 777, 1975. 12. Walsh, P. C. and Mostwin, J. L.: Radical prostatectomy and

cystoprostatectomy with preservation of potency. Results US. ing a new nerve-sparing technique. Brit. J. Urol., 56: 694, 1984. 13. Walsh, P. C.: Radical retropubic prostatectomy. In: Campbell's

Urology, 5th ed. Edited by P. C. Walsh, R. F. Gittes, A. D, Perlmutter and T. A. Stamey. Philadelphia: W. B. Saunders Co., vol. 3, chapt. 76, pp. 2769-2791, 1986. 14. Schlegel, P. N. and Walsh, P. C.: Neuroanatomical approach to radical cystoprostatectomy with preservation of sexual function. J . Urol., 138: 1402, 1987. 15. Walsh, P. C. and Schlegel, P. N.: Radical cystectomy. In: Operative Urology. Edited by F. F. Marshall. Philadelphia: W. B. Saunders Co., chapt. 20, pp. 129-142, 1991. 16. Schoenberg, M. P., Walsh, P. C., Breazeale, D. R., Marshall, F. F., Mostwin, J. L. and Brendler, C. B.: Local recurrence and survival following nerve sparing radical cystoprostatectomy for bladder cancer: 10-year followup. J. Urol.. 155: 490, 1996. 17. Johnson, D. E., Wishnow, K. I. and Tenney, D.: Are frozensection examinations of ureteral margins required for all patients undergoing radical cystectectomy for bladder cancer? Urology, 33: 451, 1989.