PREDICTING PERIOPERATIVE COMPLICATIONS IN PATIENTS RECEIVING RADICAL CYSTECTOMY USING PRE-OPERATIVE CT-MEASURED ADIPOSE TISSUE INDICES

View Poster

INTRODUCTION

Obesity is a global epidemic, however the link between obesity and bladder cancer outcomes remains controversial. Recent studies have suggested that body mass index (BMI) may not be the most accurate measure of obesity. In this study, visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) levels were measured using CT scans prior to radical cystectomy (RC). The hypothesis was that patients with higher adipose tissue levels would have poorer perioperative and survival outcomes.

METHODS

There were 202 patients undergoing a RC at a single institution that were included in this retrospective study (2000-2012). Multivariable logistic regression analysis was used to generate odds ratios (OR) for predictors of 30-day grade III-V Clavien-Dindo (CD) complications, and linear regression analysis was used to assess predictors of increasing length of stay (LOS). Multivariable competing risks and Cox proportional hazards models were used to assess disease-specific (DSS) and overall survival (OS), respectively.

RESULTS

The median age was 70 (IQR 78-60) years, VAT 165 (IQR 223-114) cm2, SAT 233 (IQR 316-182) cm2, LOS 9 (IQR 12-7) days, and age-adjusted Charlson Comorbidity Index (CCI) score was 6 (IQR 8-5). 76% of patients were male, 59% were ever smokers, there were 32 (16%) 30-day CD grade III-V complications, 71% had ≥pT2-4 disease, and 40% of patients received chemotherapy. Over a median follow-up of 37 (IQR 54-27) months for alive patients, there were 43 (21%) bladder cancer specific deaths and 65 (32%) all-cause deaths. Adjusting for CCI score and smoking status, VAT was not predictive of grade III-V 30-day CD complications (OR 1.004, 95%CI 0.999-1.008), whereas SAT (OR 1.004, 95%CI 1.001-1.008) was predictive, with similar adjustments. VAT was predictive of increasing LOS (β-coeff 0.0233, 95%CI 0.0002-0.0463) when adjusted for CCI score and gender, whereas SAT was not (β-coeff 0.0159, 95%CI -0.0024-0.0342). Neither VAT nor SAT were predictive of DSS or OS on multiple scenario survival analyses.

CONCLUSION

We demonstrated that higher VAT was predictive of longer post-operative LOS and SAT was predictive of worse complications (CD III-V) 30 days after RC. There was no difference in cancer-specific or overall survival between groups. VAT and SAT measurements may be useful in conjunction with existing modalities to improve pre-operative risk assessment for predicting immediate post-operative outcomes.

Funding: None