Original Article Japanese Journal of Clinical Oncology Advance Access published December 17, 2010 Jpn J Clin Oncol 2010 doi:10.1093/jjco/hyq228 The Clinical Impact of the Classification of Carcinoma In Situ on Tumor Recurrence and their Clinical Course in Patients with Bladder Tumor Nozomi Hayakawa, Eiji Kikuchi *, Shuji Mikami, Kazuhiro Matsumoto, Akira Miyajima and Mototsugu Oya Department of Urology, Keio University School of Medicine, Tokyo, Japan *For reprints and all correspondence: Eiji Kikuchi, Department of Urology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan. E-mail: eiji-k@kb3.so-net.ne.jp Received August 16, 2010; accepted November 16, 2010 Objective: To evaluate recurrence and clinical course of bladder carcinoma in situ according to the current carcinoma in situ classification by analyzing the patients with diagnosed carcinoma in situ in our hospital. Methods: Between January 1993 and September 2008, 93 patients were initially diagnosed with bladder carcinoma in situ in our hospital. All specimens underwent an additional review by one uro-pathologist. Primary, secondary and concurrent carcinoma in situ were found in 26, 21 and 46 patients, respectively. Sixty-nine patients (74.2%) underwent bacillus Calmette Guérin instillation therapy. Results: The multivariate analysis determined that the secondary carcinoma in situ and the absence of bacillus Calmette Guérin therapy were the independent unfavorable risk factors for tumor recurrence. The 5-year recurrence-free survival rates on primary, concurrent and secondary carcinoma in situ were 60.9, 63.2 and 25.4%, respectively, and the differences between secondary and primary carcinoma in situ, and secondary and concurrent were significant (P ¼ 0.023 and P ¼ 0.006, respectively). During the median follow-up period of 47 months, 19 patients had tumor recurrence in the bladder after the first bacillus Calmette Guérin therapy, and 13 of them were treated with a second bacillus Calmette Guérin therapy. After the second bacillus Calmette Guérin therapy six patients eventually had distant metastasis and three had upper tract recurrence, whereas totally four had a tumorfree status after the second bacillus Calmette Guérin therapy. Conclusions: The first induction course of bacillus Calmette Guérin therapy proved to be effective for the prevention for bladder cancer recurrence, however, the efficacy of a second bacillus Calmette Guérin therapy on a recurrent tumor was somewhat limited. Key words: carcinoma in situ recurrence progression classification bacillus Calmette Guérin INTRODUCTION Five to 10% of patients with non-muscle invasive bladder cancer have carcinoma in situ (CIS). Until the appearance of bacillus Calmette Guérin (BCG), their average development rate to invasive disease was about 50% and cystectomy had been recommended as the initial treatment of choice for CIS (1). At present, intravesical BCG instillation has been established as the first choice of treatment for the bladder CIS and has contributed to their conservative therapy. Only 10% of CIS patients with initial complete response after BCG therapy have muscle invasive disease later, whereas muscle invasive disease develops in approximately half of patients who do not respond to initial intravesical BCG instillation (2). However, the # The Author (2010). Published by Oxford University Press. All rights reserved.
Page 2 of 6 Carcinoma in situ of bladder long-term natural course of CIS has not been fully evaluated yet. Although several studies indicated some risk factors that can predict recurrence or progression of CIS, they were not abundant to analyze the clinical outcome of CIS with much circumstance because the proportion of CIS in the whole urinary bladder tumor is small, and moreover, CIS is heterogeneous pathema. So, it is important that the full and impartial classification to analyze CIS be applied. Recently, CIS was clearly classified into three differential clinical types; primary, secondary and concurrent CIS according to the current guidelines (3). This classification helps us to evaluate the clinical course of heterogeneous categories of CIS more comprehensibly. The objective of this report was to evaluate our CIS cases to investigate their clinical course and determine whether the clinicopathological feature of different types of CIS or the extent of CIS influences the course of the disease. PATIENTS AND METHODS PATIENTS Between January 1993 and September 2008, 125 patients were diagnosed with bladder CIS by transurethral biopsy at Keio University hospital. Of them, 93 patients (71 males and 22 females) had an initial diagnosis of CIS and we take them as an object of this study. The mean age of the patients was 68.7 years (range 40 95 years). The median follow-up was 47.0 months (range 4 178 months). None of them had previous or coexisting muscle-invasive urothelial carcinoma at the time of diagnosis. PATHOLOGY We classified 93 CIS in three differential clinical types; primary, secondary and concurrent CIS according to the current guidelines (3). Primary CIS is defined as an isolated CIS with no previous or concurrent papillary bladder tumor. Secondary CIS is a lesion which is found during the follow-up of previous papillary bladder tumor. Concurrent CIS accompanies papillary tumors. Primary, secondary and concurrent CIS were found in 26, 21 and 46 patients, respectively. We estimated the extent of CIS by calculating the percentage of positive samples from more than six different sites and defined that more than half of the sites were extensive, referring to the previous report of Takenaka et al. (4). If patients had less than six random biopsy sites, they were excluded from this classification of extent. Eighty patients received random biopsy of more than six sites, and extensive and limited CIS were found in 26 and 54 patients, respectively. For pathological diagnosis, The 1999 World Health Organization (WHO) classification was used (5). All the specimens were re-reviewed for this study by an expert uro-pathologist in a blind fashion on the clinical data in our hospital. TREATMENT BCG was scheduled for weekly administration at a dose of 80 mg Tokyo strain or 81 mg Connaught strain for 6 8 weeks into the bladder by way of urethral catheter with retention for 1 2 h. In our hospital, we used only Tokyo strain before 2004 and after that, we used both the Tokyo and Connaught strains, which mainly depended on the attending doctors preference. Complete response for treatment was defined as both negative urine cytology and no evidence of malignancy by random biopsies of the urinary bladder. Tumor recurrence was defined as any bladder cancer appearance on cystoscopy or biopsy, and tumor progression was defined as progression to muscle invasive or metastatic disease. Basically our treatment strategy for CIS was BCG intravesical instillation, however, some patients were treated with intravesical chemotherapy because of their advanced age and their poor general conditions. Of the 93 patients, 69 patients received a BCG instillation therapy on schedule, and 10 patients received intravesical chemotherapy. One patient was treated with total cystectomy and four patients with G3 pt1 coexistent tumors received radiation therapy. The remaining eight patients did not receive any additional therapy because of their poor general condition. While the remaining one patient received some therapy at another other hospital; he was excluded from analysis for recurrence and progression because the modality of treatment was unclear. Finally, during the follow-up periods, 11 of 92 patients underwent radical cystectomy because of the progression to muscle invasion (6 patients), perpetual multiple high-grade tumor recurrence (3 patients) and atrophic bladder as a severe side effect (2 patients). Of two patients who underwent radical cystectomy because of severe side effects of BCG therapy, one patient had no residual tumor cells and one patient had negligible residual tumor. Finally, two patients were diagnosed as having progression later to metastatic disease after the operations. STATISTICAL ANALYSIS Kaplan Meier method was used to assess recurrence-free survival and progression-free survival. Independent predictors of disease recurrence and progression were selected by stepwise multivariate Cox proportional hazard regression models. P value,0.05 was considered statistically significant. RESULTS BACKGROUND OF THREE TYPES OF CIS Primary, secondary and concurrent CIS were found in 26, 21 and 46 patients, respectively. Accompanying papillary tumor of concurrent CIS was as follows; pta/g2 tumors were seen in 3 patients, pta/g3 tumors in 14 patients, pt1/g2 tumors
Jpn J Clin Oncol 2010 Page 3 of 6 in 6 patients and pt1/g3 tumors in 23 patients. Forty-three of them (93.5%) had G3 element and/or T1 tumors. With regard to secondary CIS, 5 cases (23.8%) did not have accompanying papillary tumor, and in 16 cases (76.2%) accompanying papillary tumors were observed. Of the 16 cases, five patients had pta/g2 tumors, four patients pta/g3 tumors, two patients pt1/g2 tumors and five patients pt1/ G3 tumors. Eleven of them (68.8%) had G3 element and/or T1 tumors. Table 1 shows the background according to the three types of CIS. Concurrent CIS occupied more extent than secondary CIS, and the difference was significant (P ¼ 0.016). Seventy-three, 57 and 50% of patients with primary CIS, secondary CIS and concurrent CIS, respectively, had positive urine cytology. Patients with primary CIS tended to show positive preoperative urine cytology compared with those with concurrent CIS (P ¼ 0.056). On the other factors of background and therapy, there were not any significant differences among the three types of CIS. RISK FACTORS FOR TUMOR RECURRENCE IN PATIENTS WITH CIS In patients receiving any adjuvant treatment for CIS, 34 patients experienced recurrence of bladder cancer during the follow-up periods. The 5-year recurrence-free survival rate was 53.3% (Fig. 1). The multivariate analysis determined that the secondary type of CIS and the absence of BCG therapy were the independent unfavorable risk factors for tumor recurrence (Table 2). The 5-year recurrence-free survival rates with primary, concurrent and secondary CIS were 60.9, 63.2 and 25.4%, respectively (Fig. 2), and the differences between secondary and primary CIS, and secondary and concurrent were significant (P ¼ 0.023 and P ¼ 0.006, respectively). The 5-year recurrence-free survival rate in the BCG therapy group was 60.3% and that in the group without BCG therapy but with any other treatment was 24.8% (P ¼ 0.006). Totally 17 patients who were conservatively managed for CIS had stage progression (to muscle invasive cancer in 9 patients and metastatic disease in 8) during the follow-up periods. Of Figure 1. The recurrence-free survival in overall patients treated for carcinoma in situ (CIS). The 5-year recurrence-free survival rate was 53.3%. Table 2. Univariate and multivariate analyses of tumor recurrence in patients treated for CIS Univariate P value Sex 0.816 Age 0.289 Extent of CIS (50 vs.,50%) Preoperative urine cytology 0.524 0.563 Type of CIS 0.011 0.043 Multivariate P value Hazard ratio (95% CI) Primary 0.471 (0.196 1.131) Secondary 1.000 Concurrent 0.375 (0.169 0.830) BCG therapy (no vs. yes) Intravesical chemotherapy (no vs. yes) 0.006 0.027 2.322 (1.102 4.892) 0.020 Table 1. Characteristics of the study patients according to type of CIS Total Primary CIS Secondary CIS Concurrent CIS Number of patients 93 26 21 46 Male/female 71/22 (75.5/24.5%) 19/7 (73.0/27.0%) 14/7 (66.7/33.3%) 38/8 (82.6/17.4%) Average age (range) 68.7 (40 95) 70.2 (51 95) 69.5 (40 88) 67.4 (41 89) Extent of CIS (50 vs.,50%) (n ¼ 80) 26/54 (32.5/67.5%) 7/16 (30.4/69.6%) 2/16 (11.1/88.9%) 17/22 (43.6/56.4%) Preoperative urine cytology 54/39 (58.0/42.0%) 19/7 (73.1/26.9%) 12/9 (57.1/42.9%) 23/23 (50.0/50.0%) BCG therapy (yes vs. no) 69/24 (74.2/25.8%) 22/4 (84.6/15.4%) 13/8 (61.9/38.1%) 34/12 (73.9/26.1%) Intravesical chemotherapy (yes vs. no) 10/83 (10.8/89.2%) 4/22 (15.4/84.6%) 3/18 (14.3/85.7%) 3/43 (6.5/93.5%) CIS,carcinoma in situ; BCG, bacillus Calmette Guérin.
Page 4 of 6 Carcinoma in situ of bladder Figure 2. The difference of recurrence-free survival rates among three types of CIS in treated patients. The 5-year recurrence-free survival rate of primary, concurrent, secondary CIS were 60.9, 63.2, 25.4%, respectively. Secondary CIS recurred significantly more than concurrent CIS (P ¼ 0.006) and primary CIS (P ¼ 0.023). them, six patients had concurrent CIS, four patients primary CIS and seven patients secondary CIS. BCG FAILURE CASES In a subgroup of 69 patients treated with a first induction course of BCG therapy (first BCG) for CIS, two patients had recurrence of tumor within 6 months (BCG-refractory cases). These two patients were treated with total cystectomy or Table 3. Clinical outcome of BCG failure cases No. of patients Type of CIS at first presentation Presence of coincident papillary tumor at first presentation chemoradiotherapy for the residual bladder tumors. During the mean follow-up period of 50 months after the first BCG therapy, 21 patients had tumor recurrence or stage progression thereafter. Of them, upper tract recurrence (UTR) was observed in one patient and distant metastasis in one without tumor recurrence in the bladder. Of the remaining 19 patients with tumor recurrence in the bladder, 13 patients were treated with a second course of BCG therapy (second BCG) (Table 3). Pathological features of tumor recurrence after the first BCG therapy includes pure CIS in four, papillary tumor with CIS in one and papillary tumor alone in eight. The mean period from the first BCG therapy to the next tumor recurrence was 22.0 months (4 91 months). All three patients who had UTR during the follow-up period had primary CIS in the first tumor and two of their recurrent tumors were cured with a second course of BCG therapy. After the second course of BCG therapy, six patients eventually had distant metastasis, though two of them had a tumor-free status of the bladder. Totally four patients had a tumor-free status of the bladder without UTR as well as without distant metastasis after the second course of BCG therapy. DISCUSSION We classified our CIS patients into three types of CIS including primary, concurrent and secondary according to the current guideline. In patients with concurrent CIS, more than 90% of them had higher tumor grade and T1 papillary tumor. In patients with secondary CIS who had accompanying Duration in months from first BCG to recurrence Pathology of recurrent tumor after first BCG Local status (in the bladder) after 2nd BCG Tx Progression 1 Primary 33 CIS Tumor-free UTR 2 Primary 32 CIS Tumor-free Mþ 3 Primary 91 CIS Tumor-free 4 Primary 22 G3 pta G3 pt1 UTR, Mþ 5 Primary 23 G2 pt1 Tumor-free UTR 6 Primary 4 G3 pt1 G3 pt1 Mþ 7 Concurrent G3 pt1 12 CIS G3 pt2 Mþ 8 Concurrent G3 pt1 18 G2 pta G2 pt1 9 Concurrent G3 pta 26 G3 pta Tumor-free Mþ 10 Concurrent G2 pt1 4 G2 pta Tumor-free 11 Secondary (pure) 10 G3 pta and Tumor-free CIS 12 Secondary (concomitant) G3 pta 11 G3 pt1 G3 pta Mþ 13 Secondary (concomitant) G2 pta 48 G1 pta Tumor-free UTR, upper tract recurrence; Mþ, presence of metastasis.
Jpn J Clin Oncol 2010 Page 5 of 6 papillary tumors, 70% of them had G3 element and/or T1 papillary tumors. These results demonstrated that patients having CIS with accompanying papillary tumor tend to have higher grade and stage non-muscle invasive bladder tumors. In our multivariate analysis, we found that the secondary type of CIS and absence of BCG instillation were the independent unfavorable risk factors for recurrence. Takenaka et al. (4) classified their 185 patients into three types: primary, concomitant and secondary CIS such as our study and demonstrated that the extent of CIS was the only independent prognostic factor in patients treated with BCG therapy by multivariate analysis. However, it is unclear whether all patients in their report were initially diagnosed with CIS and how many pathologists agreed with the diagnosis. Meanwhile, Gils-Gielen et al. (5) reported that the extent of CIS had no prognostic significance in their 52 patients with CIS. In their study, the presence of CIS was confirmed by a referee pathologist clearly like this study, but they only performed univariate analysis. In our series, the extent of CIS was not significant in predicting recurrence in a multivariate analysis. We analyzed the prognostic value of the extent of CIS in a subgroup of patients treated with BCG therapy again. The results showed that the extent of CIS was not selected as an independent prognostic factor. In our series, the patients with secondary CIS were more likely to experience tumor recurrence. Previously there were a few reports evaluating whether the recurrence pattern of CIS was different among the types of CIS classification. With regard to the contribution of the type of CIS to patients prognosis, there were almost three differential conclusions, including primary (or pure) CIS had good prognosis (6 8), there was no difference in prognosis with regard to classifications (9,10) and primary CIS had bad prognosis (1). In patients having a history of bladder cancer with secondary CIS, the secondary CIS lesions are thought to be a type of recurrence in the whole course of bladder tumor, indicating that their recurrent malignant potential is higher than that of initially diagnosed bladder cancers, and may result in frequent recurrence. BCG induction therapy for CIS is clearly known to be effective in prevention against bladder tumor recurrence especially in the short term (11), and our results also support this. Previous reports have shown that approximately 22 35% of the patients treated with BCG induction therapy eventually had progression in 3 5 years (10,12,13). Griffiths et al. (6) reported in 135 patients with CIS without maintenance BCG a 3-month response rate of 73.3%, but a 5-year progression rate of 36.4%. The meta-analysis of the published randomized clinical trial (14) and the current clinical guidelines (3) pointed out the high response rate of BCG induction therapy, compared with intravesical chemotherapy in patients with CIS, however, no difference in prevention of stage progression between them. The current issue is whether maintenance BCG therapy could affect the prevention of stage progression. Lamm et al. (15) advocated maintenance BCG therapy as a precaution against stage progresssion, and suggested that maintenance BCG therapy may be necessary for high-risk patients, such as those with secondary CIS. Although there are cases for which maintenance BCG treatment is required for, the optimal maintenance schedule needed is unknown yet. In our study, 13 patients received a second course of BCG therapy against the remaining bladder tumor. Only four of them had a tumor-free status, six had eventually distant metastasis and three had UTR. These results indicated that in patients with CIS additional BCG therapy for recurrent tumor was somewhat limited. Our study has several limitations. First, it was performed in a retrospective fashion, and in a small number of patients. Disparities in intravesical therapies among the type of CIS may have introduced into the results, and no patients received second-look TUR for papillary tumor, which may have improved the results. To investigate CIS pathologically, the most considerable problem is that the pathological consensus of diagnosis for CIS has not been completely established yet. There is both intra-observer and inter-observer variability, even between severe dysplasia and CIS. That is a cause for a variety of results of research on CIS. For example, Sharkey and Sarosdy (16) reported a significant discrepancy between the original pathological report of a multi-institutional clinical study and the central review of them in 60 of 159 biopsies (38%). In our study, to overcome the inconsistence of pathological reports by several pathologists, the specimens were re-reviewed by an expert uropathologist in the blind fashion of clinical data and we are certain that our report has gone up in value. We concluded that CIS with accompanying papillary tumors tends to have higher grade and stage non-muscle invasive bladder tumor. Secondary type of CIS had a statistically significant risk for tumor recurrence. In the concurrent CIS, the CIS lesion was more extended than in the other types of CIS, but the extension of CIS was not associated with tumor recurrence. First induction course of BCG therapy proved to be effective for prevention for bladder cancer recurrence, however, an additional second course of BCG therapy in the recurrent tumor was somewhat limited. Conflict of interest statement None declared. References 1. Lamm DL. Carcinoma in situ. Urol Clin North Am 1992;19:499 508. 2. Witjes JA. Bladder carcinoma in situ in 2003: state of the art. Eur Urol 2004;45:142 6. 3. van der Meijden AP, Sylvester R, Oosterlinck W, Solsona E, Boehle A, et al. EAU guidelines on the diagnosis and treatment of urothelial carcinoma in situ. Eur Urol 2005;48:363 71. 4. Takenaka A, Yamada Y, Miyake H, Hara I, Fujisawa M. Clinical outcomes of bacillus Calmette Guérin instillation therapy for carcinoma in situ of urinary bladder. Int J Urol 2008;15:309 13. 5. van Gils-Gielen RJ, Witjes WP, Caris CT, Debruyne FM, Witjes JA, Oosterhof GO. Risk factors in carcinoma in situ of the urinary bladder. Dutch South East Cooperative Urological Group. Urology 1995;45:581 6.
Page 6 of 6 Carcinoma in situ of bladder 6. Griffiths TR, Charlton M, Neal DE, Powell PH. Treatment of carcinoma in situ with intravesical bacillus Calmette Guerin without maintenance. JUrol2002;167:2408 12. 7. Orozco RE, Martin AA, Murphy WM. Carcinoma in situ of the urinary bladder. Clues to host involvement in human carcinogenesis. Cancer 1994;74:115 22. 8. Ovesen H, Horn T, Steven K. Long-term efficacy of intravesical bacillus Calmette Guerin for carcinoma in situ: relationship of progression to histological response and p53 nuclear accumulation. J Urol 1997;157:1655 9. 9. Cheng L, Cheville JC, Neumann RM, Leibovich BC, Egan KS, Spotts BE, et al. Survival of patients with carcinoma in situ of the urinary bladder. Cancer 1999;85:2469 74. 10. Gofrit ON, Pode D, Pizov G, Zorn KC, Katz R, Duvdevani M, et al. The natural history of bladder carcinoma in situ after initial response to bacillus Calmette Guérin immunotherapy. Urol Oncol 2009; 27:258 62. 11. Jakse G, Hall R, Bono A, Höltl W, Carpentier P, Spaander JP, et al. Intravesical BCG in patients with carcinoma in situ of the urinary bladder: long-term results of EORTC GU Group phase II protocol 30861. Eur Urol 2001;40:144 50. 12. Herr HW, Wartinger DD, Fair WR, Oettgen HF. Bacillus Calmette Guerin therapy for superficial bladder cancer: a 10-year followup. JUrol1992;147:1020 3. 13. Kaasinen E, Wijkström H,Malmström PU, Hellsten S, Duchek M, Mestad O, et al. Alternating mitomycin C and BCG instillations versus BCG alone in treatment of carcinoma in situ of the urinary bladder: a nordic study. Eur Urol 2003;43:637 45. 14. Sylvester RJ, van der Meijden AP, Witjes JA, Kurth K. Bacillus calmette guerin versus chemotherapy for the intravesical treatment of patients with carcinoma in situ of the bladder: a meta-analysis of the published results of randomized clinical trials. J Urol 2005;174:86 91. 15. Lamm DL, Blumenstein BA, Crissman JD, Montie JE, Gottesman JE, Lowe BA, et al. Maintenance bacillus Calmette Guerin immunotherapy for recurrent TA, T1 and carcinoma in situ transitional cell carcinoma of the bladder: a randomized Southwest Oncology Group Study. JUrol2000;163:1124 9. 16. Sharkey FE, Sarosdy MF. The significance of central pathology review in clinical studies of transitional cell carcinoma in situ. J Urol 1997;157:68 70.