Login to Access Video or Poster Abstract: MP04-13
Sources of Funding: None

Introduction

Current guidelines for surveillance after radical cystectomy (RC) are based on pathologic stage and do not take into account other patient specific factors. We previously reported Weibull models determining duration of cancer surveillance. We now present a novel surveillance protocol outlining the frequency and duration of follow-up based on stage specific recurrence risk and competing risks of non-cancer mortality.

Methods

We identified 2205 patients who underwent RC for urothelial carcinoma (UC) at the Mayo Clinic from 1980 to 2010. The risks of abdomen/pelvis and chest recurrence were estimated using accelerated failure-time (AFT) models, stratified by pathologic stage (pT0, pTa/Cis/1, pT2, pT3/4, pTanyN+). Similarly, the risks of non-cancer death according to age (<60, 60-69, 70-79, >80) and Charlson comorbidity index (CCI <1 versus ≥1) were also calculated. Surveillance intervals were calculated for each conditional 1%, 3%, and 5% recurrence risk increase up to 10 years follow-up. Specific surveillance recommendations balance estimated risk of non-cancer death with recurrence risk where allowable recurrence risk is up to the risk of non-cancer death.

Results

At a median follow-up of 4.7 years (IQR 1.1, 10.3), disease recurrence was diagnosed in 852 (38.6%) patients. Using AFT models for recurrence, surveillance intervals for all stage, age, and comorbidity risk groups were generated. The surveillance strategy (e.g. 1%, 3%, or 5%) for an individual patient is selected based on probability of non-cancer specific mortality as determined by age and CCI (Tables 1 and 2). For example, a patient less than 60 with CCI ≤ 1 would follow a 1% recurrence risk schedule for the full 10 years. Conversely, a patient age 60-69 with a CCI > 1 would follow a 5% schedule for the full 10 years.

Conclusions

Using AFT modeling we have developed a risk-adapted protocol for surveillance frequency and duration after RC taking into account both the risk of recurrence and the risk of non-cancer mortality.

Funding

None