1410RADICAL CYSTECTOMY FOR CLINICALLY MUSCLE INVASIVEBLADDER CANCER: DOES PRIOR NON-INVASIVE DISEASEAFFECT CLINICAL OUTCOMES?

Ahmed Kotb, Evan Kovac*, Wassim Kassouf, Montreal, Canada;Joe Chin, London, Canada; Yves Fradet, Quebec City, Canada;Jonathan Izawa, London, Canada; Eric Estey, Adrian Fairey,Edmonton, Canada; Ricardo Rendon, Halifax, Canada;Ilias Cagiannos, Ottawa, Canada; Louis Lacombe, Quebec City,Canada; Jean-Baptiste Lattouf, Montreal, Canada; David Bell,Halifax, Canada; Darrel Drachenberg, Winnipeg, Canada;Armen Aprikian, Montreal, Canada

INTRODUCTION AND OBJECTIVES: To compare clinical andpathologic outcomes of radical cystectomy for muscle invasive bladdercancer in relation to prior history of non-invasive urothelial carcinoma.

METHODS: The Canadian Bladder Cancer Network was usedto retrospectively analyze data from 1150 patients managed by radicalcystectomy for urothelial carcinoma of the bladder. Patients with clinicalstage T2 or more were included and divided into two groups: (Group 1)patients with prior history of non-invasive urothelial carcinoma(N�365), and (Group 2) patients with clinical muscle invasive cancerde novo (N�785). Variables analyzed included patient age, gender,pathologic stage, adjuvant chemotherapy, recurrence and mortality.

RESULTS: Both groups were nearly equal in mean age andgender distribution, with mean ages of 67.2 and 66.7 years, and 79.7%and 79.5%, respectively (P 0.4 and 0.9, respectively). The presence ofpreoperative hydronephrosis was 20.8% and 32.6% (P 0.0007) forgroups 1 and 2, respectively. The rate of higher pathological stage (T3or T4) was 36.3% and 58% (P �0.0001), positive lymph nodes was20.1% and 28.8% (P 0.002) and lymphovascular invasion was 31.7%and 46.2% (P 0.0001) for groups 1 and 2, respectively. The rate ofadjuvant chemotherapy was 15.5% and 23.3% (P 0.002) for groups 1and 2, respectively. None of the sampled patients received neoadju-vant chemotherapy.

The overall survival (OS) and disease-specific survival (DSS)rates at 5 years was 62% and 70% for group 1 and 51% and 60% forgroup 2, respectively, while at 10 years, OS and DSS was 46% and66% for group 1 and 35% and 49% for group 2, respectively (P 0.0001and 0.0002 respectively).

Using multivariate analysis examining factors affecting recur-rence and survival, we found that previous non-invasive bladder tumourhistory was associated with a significantly reduced risk of mortality andrecurrence (Hazard ratio of 0.7 for all risks, P 0.0002).

CONCLUSIONS: Our retrospective study suggests that pa-tients with non-invasive urothelial carcinoma of the bladder that prog-ress to muscle-invasion and require radical cystectomy appear to havebetter pathologic and clinical outcome than patients presenting withclinical muscle invasive disease de novo.

Source of Funding: None

1411PREDICTION OF SURVIVAL OUTCOMES AFTER RADICALCYSTECTOMY USING THREE COMORBIDITY INDICES: APROPOSAL FOR A STANDARDIZED INSTRUMENT

Adrian Fairey*, Siamak Daneshmand, Kenneth Faber,Isuru Jayaratna, Dennis Lee, Anne Schuckman, Gary Lieskovsky,Jie Cai, Gus Miranda, Eila Skinner, Los Angeles, CA

INTRODUCTION AND OBJECTIVES: The optimal tool to as-sess comorbidity status prior to radical cystectomy for bladder cancerhas not been studied. Three comorbidity indices (Charlson ComorbidityIndex [CCI], Adult Comorbidity Evaluation-27 [ACE-27], and NationalInstitute on Aging and National Cancer Institute [NIA/NCI] ComorbidityIndex) were evaluated to identify the one that best explained survivaloutcomes.

METHODS: A retrospective analysis of prospectively collecteddata from the University of Southern California Bladder Cancer Data-

base was performed. Between 2003 and 2008, 557 patients underwentradical cystectomy and extended pelvic lymph node dissection forbladder cancer. Comorbidity information was obtained through a med-ical record review. Multivariable Cox regression analysis was used topredict survival outcomes. Separate models were developed for eachcomorbidity index. A goodness-of-fit (R2) statistic was used to deter-mine the Cox model (and thereby Comorbidity index) that best ex-plained overall survival and recurrence-free survival.

RESULTS: The median follow-up duration was 3.2 years(range, 0.1 to 7.3 years). For the CCI and ACE-27 instruments, thehighest comorbidity burden was independently associated with pooreroverall survival (CCI: HR 1.83, 95% CI 1.28 to 2.62, p�0.001; ACE-27:HR 1.52, 95% CI 1.10 to 2.11, p�0.012, NIA/NCI: HR 1.14, 95% CI0.83 to 1.55, p�0.425). The CCI best predicted overall survival (CCI:R2�0.18; ACE-27 R2�0.15; NIA/NCI R2� 0.15). For each index,comorbidity status was not independently associated with recurrence-free survival.

CONCLUSIONS: Comorbidity status was independently asso-ciated with overall survival but not recurrence-free survival. The CCIbest explained overall survival among 3 comorbidity indices after con-trolling for clinical variables. We propose that the CCI be adopted as thestandard for reporting outcomes of radical cystectomy.

Source of Funding: None

1412VALIDATION OF NEW STAGING SYSTEM FOR PATIENTS WITHINVASIVE UROTHELIAL CARCINOMA OF THE PROSTATE

Ahmed Abd El Latif, Ranko Miocinovic*, Cleveland, OH;Hosni Salem, Amr Massoud, Cairo, Egypt; Andrew J Stephenson,Donna Hansel, Cleveland, OH

INTRODUCTION AND OBJECTIVES: To investigate whetherthe outcome of patients undergoing radical cystectomy (RC) withcontiguous involvement of the prostatic urethra by urothelial cancer ofthe bladder (UCB) varies by the extent of ductal/stromal invasion, andto verify the changes in the new staging system.

METHODS: A retrospective review identified 103 consecutivepatients who underwent RC at two high-volume hospitals who werefound to have contiguous involvement of the prostatic urethral ducts�/� stroma with UCB. Patients were divided into two groups accordingto extent of prostatic invasion: 1) superficial N�48 (ductal involvement[N�6], glandular invasion [N�7] or focal stromal invasion [N�35]), and2) deep N�55 (deep stromal invasion (N�32), extra capsular invasionor seminal vesicles invasion (N�23)). Multivariable Cox proportionalhazards model was used to determine the association of extent ofprostatic involvement with mortality after controlling for age, institution,pathological stage, surgical margin status, and lymph node status.

RESULTS: The median follow-up was 18 months (IQR: 8-37).Lymph node metastasis was observed in 27% and 40% of patients ingroups 1 and 2, respectively. The 5-year overall survival for groups 1and 2 was 63% and 40%, respectively (p�0.02). In multivariableanalysis, patients with deep stromal invasion had a significantly worsemortality than those with superficial involvement of the prostatic ure-thra/stroma (HR: 2.6; 95% CI: 1.2-5.9).

CONCLUSIONS: Patients with superficial involvement of theprostate by contiguous UCB have a significantly improved survival incomparison to deep invasion. This supports the recent changes instaging system in which patients with ductal and focal stromal invasionare classified as pT2 sage.

Source of Funding: None

Vol. 187, No. 4S, Supplement, Tuesday, May 22, 2012 THE JOURNAL OF UROLOGY� e573