- Email: [email protected]
Usefulness of the Spanish Urological Club for Oncological Treatment Scoring Model to Predict Nonmuscle Invasive Bladder Cancer Recurrence in Patients Treated With Intravesical Bacillus Calmette-Guérin Plus Interferon-α
Usefulness of the Spanish Urological Club for Oncological Treatment Scoring Model to Predict Nonmuscle Invasive Bladder Cancer Recurrence in Patients Treated With Intravesical Bacillus Calmette-Guérin Plus Interferon-␣ Henry M. Rosevear, Andrew J. Lightfoot, Kenneth G. Nepple and Michael A. O’Donnell*,† From the Department of Urology, University of Iowa, Iowa City, Iowa
Purpose: The Spanish Urological Club for Oncological Treatment recently developed a scoring model to stratify the recurrence risk in patients treated with intravesical bacillus Calmette-Guérin using gender, age, grade, tumor status, T category, multiplicity and associated carcinoma in situ. We investigated the ability of this model to stratify the recurrence risk in patients with nonmuscle invasive bladder cancer undergoing combination bacillus Calmette-Guérin plus interferon ␣-2B therapy. Materials and Methods: We retrospectively reviewed data from a national multicenter phase II trial of bacillus Calmette-Guérin plus interferon ␣-2B in patients with nonmuscle invasive bladder cancer to identify 718 with the data required to use the model. Recurrence was defined as visible tumor on cystoscopy unless histologically confirmed as benign, definitive positive cytology or biopsy proven disease even with negative cystoscopy. Time to recurrence was indexed to the first intravesical treatment date. Patients were assigned points based on the model and then divided into 4 groups based on total score, including 0 to 4, 5 or 6, 7 to 9 and 10 or greater. Results: The model successfully stratified the recurrence risk into 4 statistically different groups based on score with a 3-year recurrence-free rate of 58%, 52%, 42% and 26% for scores of 0 to 4, 5 or 6, 7 to 9 and 10 or greater, respectively (p ⬍0.001). Conclusions: The Spanish Urological Club for Oncological Treatment scoring model is a useful prognostic tool to stratify recurrence risk in patients with nonmuscle invasive bladder cancer who are treated with combined intravesical bacillus Calmette-Guérin plus interferon ␣-2B. Larger, prospective trials are required for full model validation. Key Words: urinary bladder, urinary bladder neoplasms, BCG vaccine, interferon alfa-2b, prognosis INTRAVESICAL Mycobacterium bovis BCG immunotherapy, the most commonly used intravesical treatment for high risk NMIBC, is endorsed in European Association of Urology1 and American Urological Association practice guidelines.2 While intravesical BCG dramatically decreases the risk of tumor re-
currence, 25% to 45% of patients experience recurrence despite BCG treatment3 with a recurrence rate as low as 17% when 3-week maintenance therapy is given.4 In 2006 to stratify patients undergoing intravesical immunotherapy into risk groups Sylvester et al developed
0022-5347/11/1851-0067/0 THE JOURNAL OF UROLOGY® © 2011 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
Vol. 185, 67-71, January 2011 Printed in U.S.A. DOI:10.1016/j.juro.2010.08.083
AND
RESEARCH, INC.
Abbreviations and Acronyms BCG ⫽ bacillus Calmette-Guérin CIS ⫽ carcinoma in situ CUETO ⫽ Spanish Urological Club for Oncological Treatment EORTC ⫽ European Organization for Research and Treatment of Cancer IFN ⫽ interferon NMIBC ⫽ nonmuscle invasive bladder cancer TURBT ⫽ transurethral resection of bladder tumor Submitted for publication April 12, 2010. * Correspondence: Department of Urology, University of Iowa, 200 Hawkins Dr., 3 RCP, Iowa City, Iowa 52242-1089 (telephone: 319-384-6981; FAX: 319-356-3900; e-mail: michael-odonnell@ uiowa.edu). † Financial interest and/or other relationship with Abbott Laboratories, Alynylam Pharmaceuticals, Viventia, Anadys Pharmaceuticals, Spectrum, Loras, Endo Pharmaceuticals and Medical Enterprises.
For other article on a related topic see page 298.
www.jurology.com
67
68
SCORING MODEL TO PREDICT BLADDER CANCER RECURRENCE
a scoring system based on analysis of more than 2,500 patients with NMIBC enrolled in European Organisation for Research and Treatment of Cancer approved trials.5 European Organisation for Research and Treatment of Cancer defines risk strata based on tumor grade, stage, size, previous recurrence, number of tumors and concomitant CIS. Concerns about the validity of this scoring system were raised based on contamination of this cohort by simultaneous chemotherapy treatment and the lack of maintenance BCG therapy in a large number.5 With that in mind Fernandez-Gomez et al recently developed a scoring model to predict the NMIBC recurrence rate in patients treated with BCG immunotherapy, including a similar maintenance schedule.6 They reviewed data between 1990 and 1999 from 4 CUETO trials on a total of 1,062 patients, and developed a scoring model based on readily available clinical and pathological data, including age, gender, previous bladder tumor, number of current tumors, CIS and tumor grade. Since the maintenance BCG schedule used in these trials differed from that normally used in the United States and 27% of the patients received 9 or fewer BCG installations, concern exists about the usefulness of this model in populations undergoing the BCG regimens that are standard in the United States.7 We retrospectively reviewed data from a national multicenter phase II trial of BCG plus combination IFN-␣2B therapy to investigate the usefulness of the CUETO scoring model in this geographically distinct population.
PATIENTS AND METHODS A total of 1,106 patients with NMIBC were enrolled in an American multicenter phase II study of combination BCG plus IFN-␣2B from May 1999 to February 2001. Study results and methods were previously reported.8 Briefly, all forms of NMIBC, including Ta, T1 and CIS, were included in analysis. Immunotherapy commenced within 3 to 8 weeks of confirmatory cystoscopy, biopsy or positive cytology and best efforts were made to render the patient grossly disease free with TURBT. The starting dose in the BCG induction cycle was tailored according to prior BCG treatment, in which BCG naïve patients were given full dose standard BCG mixed directly with 50 MU IFN-␣2B. Those with recurrent NMIBC in whom therapy had failed despite having received at least 1 prior BCG treatment at any time in the past but who were not deemed BCG intolerant received a third the standard dose of BCG plus 50 MU IFN-␣2B. Patients believed to be BCG intolerant received a tenth of the standard BCG dose plus 100 MU IFN-␣2B. If patients had BCG intolerance or toxicity during induction, 1 further BCG dose reduction by approximately a third the prior dose was permitted after a 2-week rest period, eg a full dose to a third of a dose, a third to a tenth and a tenth to a thirtieth. Additional 2-week treatment delays were allowed for repeat intolerance episodes
as long as the entire induction cycle was completed within 10 weeks of initiation. All patients without recurrence received reduced dose maintenance therapy, consisting of 3 sets of 3-week mini cycles of maintenance therapy begun 3, 9 and 15 months after the end of induction cycle. Of eligible patients 90% received all 3 maintenance cycles, as fully described in our previous report.8 Patients underwent initial assessment 4 to 6 weeks after induction and then every 3 months for the first 2 years. For Kaplan-Meier tumor-free survival analysis tumor recurrence was scored positive when certain events occurred, such as visible tumor unless histologically confirmed to be benign, definitive positive urine cytology, positive biopsy even with negative cystoscopy and any urothelial cell carcinoma regardless of presenting site, including the upper tract, prostate, urethra or metastasis. Time to recurrence was indexed to the first intravesical treatment date. Each patient was assigned a CUETO score using the published model, which stratifies recurrence based on 6 characteristics, including gender, age, tumor status (recurrent or primary), number of tumors, CIS and grade (see Appendix).6 The CUETO system uses the T category of tumor to stratify progression risk but, since our cohort did not include data on disease progression, we could not investigate the validity of this portion of the CUETO model. Each risk factor in the CUETO model is given a weighted value and the sum of the risk factors is then calculated with a total score of 0 to 16. Four risk groups were then created based on a total CUETO score of 0 to 4, 5 or 6, 7 to 9 and 10 or greater with higher scores correlating with a greater risk of progression recurrence. We evaluated recurrence-free survival using Kaplan-Meier analysis.
RESULTS Of the 1,106 patients in the original study8 718 had complete data available on all 6 factors in the CUETO scoring model (see table). Reasons for study exclusion included unknown age at first diagnosis in 96 patients, pathological findings in 61, tumor stage in 66, tumor grade in 155 or primary vs recurrent tumor status in 10. The overall recurrence rate was 48.9% at the median 24-month followup. The CUETO recurrence score was 0 to 4 in 289 patients, 5 or 6 in 175, 7 to 9 in 167 and 10 or greater in 87. The figure shows the Kaplan-Meier curve for time to first recurrence. The curve revealed a significant difference among the 4 categories based on CUETO score (log rank test p ⬍0.001). Stratification based on the CUETO scoring system showed a 58%, 52%, 42% and 26% 3-year recurrence-free rate in our patients with a CUETO score of 0 to 4, 5 or 6, 7 to 9 and 10 or greater, respectively. In our population the recurrence rate was 47% (89 of 190) in females and 50% (262 of 528) in males, which was not significantly different (p ⫽ 0.51).
SCORING MODEL TO PREDICT BLADDER CANCER RECURRENCE
Patient demographics Factor Gender: M F Age: 60 or Less 61–70 71–80 Greater than 80 Prior treatment: No Yes No. tumors: 1 2–3 4–7 Greater than 7 Tumor size (cm): Less than 1 1–3 Greater than 3 T category: Ta T1 Isolated CIS Associated CIS: No Yes* Grade: G1 G2 G3 Isolated CIS
No Pts (%) 528 (74) 190 (26) 183 (25) 240 (33) 230 (32) 65 (10) 412 (57) 306 (43) 399 (56) 156 (22) 118 (16) 45 (6) 79 (11) 420 (58) 219 (31) 482 (67) 211 (29) 25 (4) 622 (87) 96 (13) 166 (23) 268 (37) 259 (36) 25 (4)
* Including isolated disease.
We would have liked to validate the CUETO model using entire our data set. Unfortunately the retrospective nature of our review introduced a significant number of potential sources of bias into that validation attempt. Some sources of bias include the large number of patients (388 of 1,106) that we excluded from analysis due to missing data as well as differences in patient populations. For example, in our patients BCG doses were tailored according to previous BCG exposure while CUETO patients were randomized by treatment arm.6 Our use of combination immunotherapy (BCG plus IFN-␣) also introduced potential bias. With those caveats in mind the observation that the CUETO model successfully stratified our different patient population into groups with unique recurrence risks (58% vs 26% 3-year recurrence-free rate in the lowest vs the high risk group) provides evidence that the CUETO model may be a powerful clinical tool. We stress that our patient population differed significantly from the population analyzed by the CUETO group6 in geographic location, ethnic background and treatment algorithm. These differences may explain the variation in recurrence-free rates between our population and that in the CUETO study (58%, 52%, 42% and 26% in our study vs 79%, 64%, 53% and 33% in the CUETO study in the 4 groups, respectively). Important information on progression to muscle invasive disease was not available in our patient cohort and so the usefulness of that portion of the CUETO scoring system remains unknown. The recurrence rate in our study was higher than the overall recurrence rate in the original study by
DISCUSSION An easy to use prognostic tool based on readily available clinical data that can predict which patients with NMIBC are likely to respond to intravesical therapy would be useful.9 Fernandez-Gomez et al significantly advanced that goal when they developed the CUETO model using 1 of the largest published multivariate analyses to stratify the risk of recurrence based on readily available clinical and demographic data.6 Most risk factors used in the CUETO model are not novel since patient age, tumor status, tumor burden, CIS and grade were each independently shown to influence the response to BCG treatment.10 –14 Gender, which is included in the CUETO model, was previously reported not to be a prognostic factor for determining the outcome of the response in patients with NMIBC treated with BCG, which makes its inclusion in this scoring system novel. Reports of variance in the urinary immunological response to intravesical BCG between the genders may help explain its inclusion.10,15,16 Fernandez-Gomez et al6 deserve credit for consolidating these factors into a clinical tool with prognostic ability.
69
Kaplan-Meier Curve shows recurrence-free survival
70
SCORING MODEL TO PREDICT BLADDER CANCER RECURRENCE
Fernandez-Gomez et al (48.9% vs 32.6%).6 Possible explanations for this discrepancy are the higher percent of patients with CIS and/or high grade disease in our population compared to theirs, including 23.2% vs 10.3% for CIS and 43% vs 23.5% for high grade disease. The published CUETO model6 and our patient cohort included patients treated from 1999 to 2001, before the currently accepted practices of immediate post-resection intravesical chemotherapy17 and maintenance intravesical therapy.4 Each practice was shown to decrease the recurrence rate so that the contemporary recurrence rate should theoretically be less than predicted by the CUETO model. Also, the standard of care has progressed to obligatory repeat resection for T1 high grade bladder cancer since 30% to 40% of cases are up-staged to muscle invasive disease.18,19 Given that the CUETO data set and our data included cases that were likely not re-resected for high grade disease, the impact of this practice pattern change on the CUETO model is unknown. The practice of re-resection TURBT for T1 high grade bladder cancer avoids unnecessary delay for intravesical therapy in patients for whom it is destined to fail and may even improve the BCG response rate.1 Furthermore, the CUETO study group used the older version of the WHO/International Society of Urologic Pathology reporting system for tumor grade (grades 1, 2 and 3)6 vs the newer high vs low grade system. To our knowledge the impact of this change on the CUETO model is unknown. Additional factors not included in the CUETO model or in our data could be added to a prognostic model to enhance its usefulness. Lymphovascular invasion in the TURBT specimen is associated with more aggressive disease and poorer prognosis.20 Other possible prognostic factors are Ki67 or tumor necrosis factor-␣ polymorphisms.21,22 Our study has several limitations, as partially described. Our data were from a retrospective review of an American multicenter trial of BCG plus IFN-␣2B. Due to incomplete data collection a number of patients were excluded from analysis due to a lack of the data required for the CUETO model. Our study lacked progression data, which is clinically
important. Followup was moderate compared to the excellent followup in the study by Fernandez-Gomez et al.6 Also, all of our patients received combination BCG plus IFN-␣2B. In addition to the differences in induction and maintenance cycle timing, this makes direct comparison between the CUETO study6 and our study challenging, and may explain some differences in predictive values between our cohort and the CUETO cohort. However, differences in cohorts provide some evidence on model generalizability.
CONCLUSIONS Overall the CUETO scoring model successfully stratified the patient risk of recurrent NMIBC after BCG plus interferon ␣ therapy. It can be used to help counsel patients on treatment decisions. A large, prospective trial should be considered to determine the true validity of this scoring model, taking into account the advances made in NMIBC treatment in the last decade.
APPENDIX CUETO Scoring Model6 Factor Gender: M F Age: Less than 60 60–70 Greater than 70 Recurrent tumor: No Yes No. tumors: 3 or Less Greater than 3 Associated CIS: No Yes Grade: G1 G2 G3 Total range
Score 0 3 0 1 2 0 4 0 2 0 2 0 1 3 0–16
REFERENCES 1. Babjuk M, Oosterlinck W, Sylvester R et al: EAU guidelines on non-muscle-invasive urothelial carcinoma of the bladder. Eur Urol 2008; 54: 303.
3. Grossman HB, O’Donnell MA, Cookson MS et al: Bacillus Calmette-Guerin failures and beyond: contemporary management of non-muscle-invasive bladder cancer. Rev Urol 2008; 10: 281.
2. Hall MC, Chang SS, Dalbagni G et al: Guideline for the management of nonmuscle invasive bladder cancer (stages Ta, T1, and Tis): 2007 update. J Urol 2007; 178: 2314.
4. Lamm DL, Blumenstein BA, Crissman JD et al: Maintenance bacillus Calmette-Guerin immunotherapy for recurrent TA, T1 and carcinoma in situ transitional cell carcinoma of the bladder: a ran-
domized Southwest Oncology Group Study. J Urol 2000; 163: 1124. 5. Sylvester RJ, van der Meijden AP, Oosterlinck W et al: Predicting recurrence and progression in individual patients with stage Ta T1 bladder cancer using EORTC risk tables: a combined analysis of 2596 patients from seven EORTC trials. Eur Urol 2006; 49: 466.
SCORING MODEL TO PREDICT BLADDER CANCER RECURRENCE
6. Fernandez-Gomez J, Madero R, Solsona E et al: Predicting nonmuscle invasive bladder cancer recurrence and progression in patients treated with bacillus Calmette-Guerin: the CUETO scoring model. J Urol 2009; 182: 2195. 7. Kamat AM: Editorial comment: predicting nonmuscle invasive bladder cancer recurrence and progression in patients treated With bacillus Calmette-Guerin: the CUETO scoring model. J Urol 2009; 182: 2203. 8. Joudi FN, Smith BJ and O’Donnell MA: Final results from a national multicenter phase II trial of combination bacillus Calmette-Guerin plus interferon alpha-2B for reducing recurrence of superficial bladder cancer. Urol Oncol 2006; 24: 344. 9. O’Donnell MA: Optimizing BCG therapy. Urol Oncol 2009; 27: 325. 10. Saint F, Patard JJ, Maille P et al: Prognostic value of a T helper 1 urinary cytokine response after intravesical bacillus Calmette-Guerin treatment for superficial bladder cancer. J Urol 2002; 167: 364. 11. Joudi FN, Smith BJ, O’Donnell MA et al: The impact of age on the response of patients with superficial bladder cancer to intravesical immunotherapy. J Urol 2006; 175: 1634.
71
12. Martinez-Pineiro JA, Flores N, Isorna S et al: Long-term follow-up of a randomized prospective trial comparing a standard 81 mg dose of intravesical bacille Calmette-Guerin with a reduced dose of 27 mg in superficial bladder cancer. BJU Int 2002; 89: 671.
17. Sylvester RJ, Oosterlinck W and van der Meijden AP: A single immediate postoperative instillation of chemotherapy decreases the risk of recurrence in patients with stage Ta T1 bladder cancer: a meta-analysis of published results of randomized clinical trials. J Urol 2004; 171: 2186.
13. Lamm DL: BCG immunotherapy for transitionalcell carcinoma in situ of the bladder. Oncology (Williston Park) 1995; 9: 947.
18. Nepple KG and O’Donnell MA: The optimal management of T1 high-grade bladder cancer. Can Urol Assoc J 2009; 3: S188.
14. Brake M, Loertzer H, Horsch R et al: Recurrence and progression of stage T1, grade 3 transitional cell carcinoma of the bladder following intravesical immunotherapy with bacillus CalmetteGuerin. J Urol 2000; 163: 1697. 15. Boorjian SA, Zhu F and Herr HW: The effect of gender on response to bacillus Calmette-Guerin therapy for patients with non-muscle-invasive urothelial carcinoma of the bladder. BJU Int 2010; 106: 357. 16. Rosevear HM, Lightfoot AJ and O’Donnell MA: Cancer recurrence disparities among racial and gender groups with non-muscle invasive bladder cancer (NMIBC) after intravesical bacillus Calmette-Guerin (BCG) and interferon alpha-2B therapy. Presented at annual meeting of Society for Urologic Oncology, Washington, D.C., December 8 –10, 2009.
19. Herr HW: Tumor progression and survival of patients with high grade, noninvasive papillary (TaG3) bladder tumors: 15-year outcome. J Urol 2000; 163: 60. 20. Streeper NM, Simons CM, Konety BR et al: The significance of lymphovascular invasion in transurethral resection of bladder tumour and cystectomy specimens on the survival of patients with urothelial bladder cancer. BJU Int 2009; 103: 475. 21. Margulis V, Lotan Y, Karakiewicz PI et al: Multiinstitutional validation of the predictive value of Ki-67 labeling index in patients with urinary bladder cancer. J Natl Cancer Inst 2009; 101: 114. 22. Ahirwar DK, Mandhani A, Dharaskar A et al: Association of tumour necrosis factor-␣ gene (T-1031C, C-863A, and C-857T) polymorphisms with bladder cancer susceptibility and outcome after bacille Calmette-Guerin immunotherapy BJU Int 2009; 104: 867.
Purpose: The Spanish Urological Club for Oncological Treatment recently developed a scoring model to stratify the recurrence risk in patients treated with intravesical bacillus Calmette-Guérin using gender, age, grade, tumor status, T category, multiplicity and associated carcinoma in situ. We investigated the ability of this model to stratify the recurrence risk in patients with nonmuscle invasive bladder cancer undergoing combination bacillus Calmette-Guérin plus interferon ␣-2B therapy. Materials and Methods: We retrospectively reviewed data from a national multicenter phase II trial of bacillus Calmette-Guérin plus interferon ␣-2B in patients with nonmuscle invasive bladder cancer to identify 718 with the data required to use the model. Recurrence was defined as visible tumor on cystoscopy unless histologically confirmed as benign, definitive positive cytology or biopsy proven disease even with negative cystoscopy. Time to recurrence was indexed to the first intravesical treatment date. Patients were assigned points based on the model and then divided into 4 groups based on total score, including 0 to 4, 5 or 6, 7 to 9 and 10 or greater. Results: The model successfully stratified the recurrence risk into 4 statistically different groups based on score with a 3-year recurrence-free rate of 58%, 52%, 42% and 26% for scores of 0 to 4, 5 or 6, 7 to 9 and 10 or greater, respectively (p ⬍0.001). Conclusions: The Spanish Urological Club for Oncological Treatment scoring model is a useful prognostic tool to stratify recurrence risk in patients with nonmuscle invasive bladder cancer who are treated with combined intravesical bacillus Calmette-Guérin plus interferon ␣-2B. Larger, prospective trials are required for full model validation. Key Words: urinary bladder, urinary bladder neoplasms, BCG vaccine, interferon alfa-2b, prognosis INTRAVESICAL Mycobacterium bovis BCG immunotherapy, the most commonly used intravesical treatment for high risk NMIBC, is endorsed in European Association of Urology1 and American Urological Association practice guidelines.2 While intravesical BCG dramatically decreases the risk of tumor re-
currence, 25% to 45% of patients experience recurrence despite BCG treatment3 with a recurrence rate as low as 17% when 3-week maintenance therapy is given.4 In 2006 to stratify patients undergoing intravesical immunotherapy into risk groups Sylvester et al developed
0022-5347/11/1851-0067/0 THE JOURNAL OF UROLOGY® © 2011 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
Vol. 185, 67-71, January 2011 Printed in U.S.A. DOI:10.1016/j.juro.2010.08.083
AND
RESEARCH, INC.
Abbreviations and Acronyms BCG ⫽ bacillus Calmette-Guérin CIS ⫽ carcinoma in situ CUETO ⫽ Spanish Urological Club for Oncological Treatment EORTC ⫽ European Organization for Research and Treatment of Cancer IFN ⫽ interferon NMIBC ⫽ nonmuscle invasive bladder cancer TURBT ⫽ transurethral resection of bladder tumor Submitted for publication April 12, 2010. * Correspondence: Department of Urology, University of Iowa, 200 Hawkins Dr., 3 RCP, Iowa City, Iowa 52242-1089 (telephone: 319-384-6981; FAX: 319-356-3900; e-mail: michael-odonnell@ uiowa.edu). † Financial interest and/or other relationship with Abbott Laboratories, Alynylam Pharmaceuticals, Viventia, Anadys Pharmaceuticals, Spectrum, Loras, Endo Pharmaceuticals and Medical Enterprises.
For other article on a related topic see page 298.
www.jurology.com
67
68
SCORING MODEL TO PREDICT BLADDER CANCER RECURRENCE
a scoring system based on analysis of more than 2,500 patients with NMIBC enrolled in European Organisation for Research and Treatment of Cancer approved trials.5 European Organisation for Research and Treatment of Cancer defines risk strata based on tumor grade, stage, size, previous recurrence, number of tumors and concomitant CIS. Concerns about the validity of this scoring system were raised based on contamination of this cohort by simultaneous chemotherapy treatment and the lack of maintenance BCG therapy in a large number.5 With that in mind Fernandez-Gomez et al recently developed a scoring model to predict the NMIBC recurrence rate in patients treated with BCG immunotherapy, including a similar maintenance schedule.6 They reviewed data between 1990 and 1999 from 4 CUETO trials on a total of 1,062 patients, and developed a scoring model based on readily available clinical and pathological data, including age, gender, previous bladder tumor, number of current tumors, CIS and tumor grade. Since the maintenance BCG schedule used in these trials differed from that normally used in the United States and 27% of the patients received 9 or fewer BCG installations, concern exists about the usefulness of this model in populations undergoing the BCG regimens that are standard in the United States.7 We retrospectively reviewed data from a national multicenter phase II trial of BCG plus combination IFN-␣2B therapy to investigate the usefulness of the CUETO scoring model in this geographically distinct population.
PATIENTS AND METHODS A total of 1,106 patients with NMIBC were enrolled in an American multicenter phase II study of combination BCG plus IFN-␣2B from May 1999 to February 2001. Study results and methods were previously reported.8 Briefly, all forms of NMIBC, including Ta, T1 and CIS, were included in analysis. Immunotherapy commenced within 3 to 8 weeks of confirmatory cystoscopy, biopsy or positive cytology and best efforts were made to render the patient grossly disease free with TURBT. The starting dose in the BCG induction cycle was tailored according to prior BCG treatment, in which BCG naïve patients were given full dose standard BCG mixed directly with 50 MU IFN-␣2B. Those with recurrent NMIBC in whom therapy had failed despite having received at least 1 prior BCG treatment at any time in the past but who were not deemed BCG intolerant received a third the standard dose of BCG plus 50 MU IFN-␣2B. Patients believed to be BCG intolerant received a tenth of the standard BCG dose plus 100 MU IFN-␣2B. If patients had BCG intolerance or toxicity during induction, 1 further BCG dose reduction by approximately a third the prior dose was permitted after a 2-week rest period, eg a full dose to a third of a dose, a third to a tenth and a tenth to a thirtieth. Additional 2-week treatment delays were allowed for repeat intolerance episodes
as long as the entire induction cycle was completed within 10 weeks of initiation. All patients without recurrence received reduced dose maintenance therapy, consisting of 3 sets of 3-week mini cycles of maintenance therapy begun 3, 9 and 15 months after the end of induction cycle. Of eligible patients 90% received all 3 maintenance cycles, as fully described in our previous report.8 Patients underwent initial assessment 4 to 6 weeks after induction and then every 3 months for the first 2 years. For Kaplan-Meier tumor-free survival analysis tumor recurrence was scored positive when certain events occurred, such as visible tumor unless histologically confirmed to be benign, definitive positive urine cytology, positive biopsy even with negative cystoscopy and any urothelial cell carcinoma regardless of presenting site, including the upper tract, prostate, urethra or metastasis. Time to recurrence was indexed to the first intravesical treatment date. Each patient was assigned a CUETO score using the published model, which stratifies recurrence based on 6 characteristics, including gender, age, tumor status (recurrent or primary), number of tumors, CIS and grade (see Appendix).6 The CUETO system uses the T category of tumor to stratify progression risk but, since our cohort did not include data on disease progression, we could not investigate the validity of this portion of the CUETO model. Each risk factor in the CUETO model is given a weighted value and the sum of the risk factors is then calculated with a total score of 0 to 16. Four risk groups were then created based on a total CUETO score of 0 to 4, 5 or 6, 7 to 9 and 10 or greater with higher scores correlating with a greater risk of progression recurrence. We evaluated recurrence-free survival using Kaplan-Meier analysis.
RESULTS Of the 1,106 patients in the original study8 718 had complete data available on all 6 factors in the CUETO scoring model (see table). Reasons for study exclusion included unknown age at first diagnosis in 96 patients, pathological findings in 61, tumor stage in 66, tumor grade in 155 or primary vs recurrent tumor status in 10. The overall recurrence rate was 48.9% at the median 24-month followup. The CUETO recurrence score was 0 to 4 in 289 patients, 5 or 6 in 175, 7 to 9 in 167 and 10 or greater in 87. The figure shows the Kaplan-Meier curve for time to first recurrence. The curve revealed a significant difference among the 4 categories based on CUETO score (log rank test p ⬍0.001). Stratification based on the CUETO scoring system showed a 58%, 52%, 42% and 26% 3-year recurrence-free rate in our patients with a CUETO score of 0 to 4, 5 or 6, 7 to 9 and 10 or greater, respectively. In our population the recurrence rate was 47% (89 of 190) in females and 50% (262 of 528) in males, which was not significantly different (p ⫽ 0.51).
SCORING MODEL TO PREDICT BLADDER CANCER RECURRENCE
Patient demographics Factor Gender: M F Age: 60 or Less 61–70 71–80 Greater than 80 Prior treatment: No Yes No. tumors: 1 2–3 4–7 Greater than 7 Tumor size (cm): Less than 1 1–3 Greater than 3 T category: Ta T1 Isolated CIS Associated CIS: No Yes* Grade: G1 G2 G3 Isolated CIS
No Pts (%) 528 (74) 190 (26) 183 (25) 240 (33) 230 (32) 65 (10) 412 (57) 306 (43) 399 (56) 156 (22) 118 (16) 45 (6) 79 (11) 420 (58) 219 (31) 482 (67) 211 (29) 25 (4) 622 (87) 96 (13) 166 (23) 268 (37) 259 (36) 25 (4)
* Including isolated disease.
We would have liked to validate the CUETO model using entire our data set. Unfortunately the retrospective nature of our review introduced a significant number of potential sources of bias into that validation attempt. Some sources of bias include the large number of patients (388 of 1,106) that we excluded from analysis due to missing data as well as differences in patient populations. For example, in our patients BCG doses were tailored according to previous BCG exposure while CUETO patients were randomized by treatment arm.6 Our use of combination immunotherapy (BCG plus IFN-␣) also introduced potential bias. With those caveats in mind the observation that the CUETO model successfully stratified our different patient population into groups with unique recurrence risks (58% vs 26% 3-year recurrence-free rate in the lowest vs the high risk group) provides evidence that the CUETO model may be a powerful clinical tool. We stress that our patient population differed significantly from the population analyzed by the CUETO group6 in geographic location, ethnic background and treatment algorithm. These differences may explain the variation in recurrence-free rates between our population and that in the CUETO study (58%, 52%, 42% and 26% in our study vs 79%, 64%, 53% and 33% in the CUETO study in the 4 groups, respectively). Important information on progression to muscle invasive disease was not available in our patient cohort and so the usefulness of that portion of the CUETO scoring system remains unknown. The recurrence rate in our study was higher than the overall recurrence rate in the original study by
DISCUSSION An easy to use prognostic tool based on readily available clinical data that can predict which patients with NMIBC are likely to respond to intravesical therapy would be useful.9 Fernandez-Gomez et al significantly advanced that goal when they developed the CUETO model using 1 of the largest published multivariate analyses to stratify the risk of recurrence based on readily available clinical and demographic data.6 Most risk factors used in the CUETO model are not novel since patient age, tumor status, tumor burden, CIS and grade were each independently shown to influence the response to BCG treatment.10 –14 Gender, which is included in the CUETO model, was previously reported not to be a prognostic factor for determining the outcome of the response in patients with NMIBC treated with BCG, which makes its inclusion in this scoring system novel. Reports of variance in the urinary immunological response to intravesical BCG between the genders may help explain its inclusion.10,15,16 Fernandez-Gomez et al6 deserve credit for consolidating these factors into a clinical tool with prognostic ability.
69
Kaplan-Meier Curve shows recurrence-free survival
70
SCORING MODEL TO PREDICT BLADDER CANCER RECURRENCE
Fernandez-Gomez et al (48.9% vs 32.6%).6 Possible explanations for this discrepancy are the higher percent of patients with CIS and/or high grade disease in our population compared to theirs, including 23.2% vs 10.3% for CIS and 43% vs 23.5% for high grade disease. The published CUETO model6 and our patient cohort included patients treated from 1999 to 2001, before the currently accepted practices of immediate post-resection intravesical chemotherapy17 and maintenance intravesical therapy.4 Each practice was shown to decrease the recurrence rate so that the contemporary recurrence rate should theoretically be less than predicted by the CUETO model. Also, the standard of care has progressed to obligatory repeat resection for T1 high grade bladder cancer since 30% to 40% of cases are up-staged to muscle invasive disease.18,19 Given that the CUETO data set and our data included cases that were likely not re-resected for high grade disease, the impact of this practice pattern change on the CUETO model is unknown. The practice of re-resection TURBT for T1 high grade bladder cancer avoids unnecessary delay for intravesical therapy in patients for whom it is destined to fail and may even improve the BCG response rate.1 Furthermore, the CUETO study group used the older version of the WHO/International Society of Urologic Pathology reporting system for tumor grade (grades 1, 2 and 3)6 vs the newer high vs low grade system. To our knowledge the impact of this change on the CUETO model is unknown. Additional factors not included in the CUETO model or in our data could be added to a prognostic model to enhance its usefulness. Lymphovascular invasion in the TURBT specimen is associated with more aggressive disease and poorer prognosis.20 Other possible prognostic factors are Ki67 or tumor necrosis factor-␣ polymorphisms.21,22 Our study has several limitations, as partially described. Our data were from a retrospective review of an American multicenter trial of BCG plus IFN-␣2B. Due to incomplete data collection a number of patients were excluded from analysis due to a lack of the data required for the CUETO model. Our study lacked progression data, which is clinically
important. Followup was moderate compared to the excellent followup in the study by Fernandez-Gomez et al.6 Also, all of our patients received combination BCG plus IFN-␣2B. In addition to the differences in induction and maintenance cycle timing, this makes direct comparison between the CUETO study6 and our study challenging, and may explain some differences in predictive values between our cohort and the CUETO cohort. However, differences in cohorts provide some evidence on model generalizability.
CONCLUSIONS Overall the CUETO scoring model successfully stratified the patient risk of recurrent NMIBC after BCG plus interferon ␣ therapy. It can be used to help counsel patients on treatment decisions. A large, prospective trial should be considered to determine the true validity of this scoring model, taking into account the advances made in NMIBC treatment in the last decade.
APPENDIX CUETO Scoring Model6 Factor Gender: M F Age: Less than 60 60–70 Greater than 70 Recurrent tumor: No Yes No. tumors: 3 or Less Greater than 3 Associated CIS: No Yes Grade: G1 G2 G3 Total range
Score 0 3 0 1 2 0 4 0 2 0 2 0 1 3 0–16
REFERENCES 1. Babjuk M, Oosterlinck W, Sylvester R et al: EAU guidelines on non-muscle-invasive urothelial carcinoma of the bladder. Eur Urol 2008; 54: 303.
3. Grossman HB, O’Donnell MA, Cookson MS et al: Bacillus Calmette-Guerin failures and beyond: contemporary management of non-muscle-invasive bladder cancer. Rev Urol 2008; 10: 281.
2. Hall MC, Chang SS, Dalbagni G et al: Guideline for the management of nonmuscle invasive bladder cancer (stages Ta, T1, and Tis): 2007 update. J Urol 2007; 178: 2314.
4. Lamm DL, Blumenstein BA, Crissman JD et al: Maintenance bacillus Calmette-Guerin immunotherapy for recurrent TA, T1 and carcinoma in situ transitional cell carcinoma of the bladder: a ran-
domized Southwest Oncology Group Study. J Urol 2000; 163: 1124. 5. Sylvester RJ, van der Meijden AP, Oosterlinck W et al: Predicting recurrence and progression in individual patients with stage Ta T1 bladder cancer using EORTC risk tables: a combined analysis of 2596 patients from seven EORTC trials. Eur Urol 2006; 49: 466.
SCORING MODEL TO PREDICT BLADDER CANCER RECURRENCE
6. Fernandez-Gomez J, Madero R, Solsona E et al: Predicting nonmuscle invasive bladder cancer recurrence and progression in patients treated with bacillus Calmette-Guerin: the CUETO scoring model. J Urol 2009; 182: 2195. 7. Kamat AM: Editorial comment: predicting nonmuscle invasive bladder cancer recurrence and progression in patients treated With bacillus Calmette-Guerin: the CUETO scoring model. J Urol 2009; 182: 2203. 8. Joudi FN, Smith BJ and O’Donnell MA: Final results from a national multicenter phase II trial of combination bacillus Calmette-Guerin plus interferon alpha-2B for reducing recurrence of superficial bladder cancer. Urol Oncol 2006; 24: 344. 9. O’Donnell MA: Optimizing BCG therapy. Urol Oncol 2009; 27: 325. 10. Saint F, Patard JJ, Maille P et al: Prognostic value of a T helper 1 urinary cytokine response after intravesical bacillus Calmette-Guerin treatment for superficial bladder cancer. J Urol 2002; 167: 364. 11. Joudi FN, Smith BJ, O’Donnell MA et al: The impact of age on the response of patients with superficial bladder cancer to intravesical immunotherapy. J Urol 2006; 175: 1634.
71
12. Martinez-Pineiro JA, Flores N, Isorna S et al: Long-term follow-up of a randomized prospective trial comparing a standard 81 mg dose of intravesical bacille Calmette-Guerin with a reduced dose of 27 mg in superficial bladder cancer. BJU Int 2002; 89: 671.
17. Sylvester RJ, Oosterlinck W and van der Meijden AP: A single immediate postoperative instillation of chemotherapy decreases the risk of recurrence in patients with stage Ta T1 bladder cancer: a meta-analysis of published results of randomized clinical trials. J Urol 2004; 171: 2186.
13. Lamm DL: BCG immunotherapy for transitionalcell carcinoma in situ of the bladder. Oncology (Williston Park) 1995; 9: 947.
18. Nepple KG and O’Donnell MA: The optimal management of T1 high-grade bladder cancer. Can Urol Assoc J 2009; 3: S188.
14. Brake M, Loertzer H, Horsch R et al: Recurrence and progression of stage T1, grade 3 transitional cell carcinoma of the bladder following intravesical immunotherapy with bacillus CalmetteGuerin. J Urol 2000; 163: 1697. 15. Boorjian SA, Zhu F and Herr HW: The effect of gender on response to bacillus Calmette-Guerin therapy for patients with non-muscle-invasive urothelial carcinoma of the bladder. BJU Int 2010; 106: 357. 16. Rosevear HM, Lightfoot AJ and O’Donnell MA: Cancer recurrence disparities among racial and gender groups with non-muscle invasive bladder cancer (NMIBC) after intravesical bacillus Calmette-Guerin (BCG) and interferon alpha-2B therapy. Presented at annual meeting of Society for Urologic Oncology, Washington, D.C., December 8 –10, 2009.
19. Herr HW: Tumor progression and survival of patients with high grade, noninvasive papillary (TaG3) bladder tumors: 15-year outcome. J Urol 2000; 163: 60. 20. Streeper NM, Simons CM, Konety BR et al: The significance of lymphovascular invasion in transurethral resection of bladder tumour and cystectomy specimens on the survival of patients with urothelial bladder cancer. BJU Int 2009; 103: 475. 21. Margulis V, Lotan Y, Karakiewicz PI et al: Multiinstitutional validation of the predictive value of Ki-67 labeling index in patients with urinary bladder cancer. J Natl Cancer Inst 2009; 101: 114. 22. Ahirwar DK, Mandhani A, Dharaskar A et al: Association of tumour necrosis factor-␣ gene (T-1031C, C-863A, and C-857T) polymorphisms with bladder cancer susceptibility and outcome after bacille Calmette-Guerin immunotherapy BJU Int 2009; 104: 867.