Introduction:

Most studies in urothelial cancer have examined smoking status (current, former, never), but often fail to quantify smoking burden/exposure, which has been shown to be prognostic in other malignancies. The impact of smoking burden on oncologic outcomes for upper tract urothelial carcinoma (UTUC) remains understudied with current literature providing mixed results. Limitations of the existing literature include the lack of patients who received neoadjuvant chemotherapy (NAC) and only a low proportion of patients receiving adjuvant therapy (AC). The relationship between smoking burden and pathologic downstaging after NAC in UTUC is largely unknown, and survival data has been mixed. We therefore performed a contemporary analysis of smoking burden on oncologic outcomes in patients with UTUC undergoing radical nephroureterectomy (RNUx) utilizing a larrge multicentered, multinational cohort (ROBUUST 2.0). 

Methods:

We performed a retrospective analysis from a large multicentered cohort of 1,730 patients with UTUC across 17 institutions from 2005-2022. We excluded patients with incomplete smoking history, pathologic data, non-urothelial histology, and prior or concurrent cystectomy. All patients underwent RNUx. Smoking history included current smoking status, cigarettes per day (CPD), total pack-years (TPY), and cumulative smoking exposure. Based on prior studies, cumulative smoking exposure was stratified as light (≤19 CPD & ≤19 years smoking), or heavy (>20 CPD & >20 years smoking), with all other cases being moderate exposure. For analysis, light and moderate smoking exposure groups were combined. Survival outcomes of cancer specific (CSS) and overall survival (OS) were assessed using the Kaplan-Meier method and multivariable competing risk regression to adjust for competing risk of non-cancer mortality from smoking exposure. A multivariable regression analysis was performed to examine odds of achieving pathologic down-staging (<ypT2) after cisplatin NAC. 

Results:

1,041 patients met criteria for analysis. Median follow up (IQR) was 24(10-48) months. 5 year CSS was 97% in non-smokers, 89% in light to moderate smokers, and 75% in heavy smokers (p<0.001, Figure 1). 5 year OS was 91% in non-smokers, 68% in light to moderate smokers and 60% in heavy smokers (p<0.001). On multivariable competing risk regression, both light to moderate smoking burden (HR: 2.98, p=0.02) and heavy smoking burden (HR: 3.24, p=0.02) were associated with greater cancer mortality compared to non-smokers (Table 1). Further, both light to moderate smoking burden (HR: 3.58, p<0.001) and heavy smoking burden (HR: 2.69, p<0.001) were associated with greater overall mortality compared to non-smokers. Models adjusted for active smoking status, NAC, pathologic T & N stage, grade, necrosis, lymphovascular invasion, and AC usage. Smoking burden was not associated with pathologic down staging after NAC on multivariable regression.

Conclusion:

In this contemporary multicentered UTUC cohort, increasing smoking burden was associated with worse cancer specific and overall survival, even among former smokers. This contemporary analysis supports existing literature that increasing smoking burden is associated with worse survival outcomes and underscoring continued need for more aggressive smoking cessation and prevention programs. Smoking burden was not associated with lower odds of achieving pathologic down staging after NAC, suggesting that NAC and AC should be continued to be emphasized in this patient population at high risk of cancer specific mortality. Future and on-going UTUC trials should assess the influence of smoking burden on response to NAC and oncologic outcomes. 

Funding: N/A