Pieretti A, Krasnow R, Drumm M et al: Complications and outcomes of salvage cystectomy after trimodality therapy. J Urol 2021; 206: 29.

Bladder-sparing trimodality therapy (TMT) is part of the National Comprehensive Cancer Network guidelines and an accepted alternative in appropriately selected patients with muscle invasive bladder cancer (MIBC). TMT oncologic outcomes are comparable to radical cystectomy series.^1–3 As TMT protocols and patient selection over the past 2 decades have evolved, the current complete response rate equals ∼88%, and 5-year disease-specific survival (DSS) has increased to ∼84%.¹ Due to the risk of incomplete response to TMT or subsequent intravesical recurrence, close surveillance is required. Salvage cystectomy (SC) is ultimately necessary for approximately 11%–16% of TMT patients in contemporary series.^1,2

Concern about the safety and oncologic efficacy of post-TMT SC has been a reason for the lack of more prevalent use of TMT in appropriate patients. Previous retrospective analysis on SC after TMT identified a 69% 90-day complication rate with major complications comprising 16%. This risk of acute complications was higher than other primary radical cystectomy (PC) series; however, this report lacked a control group for comparative analysis.⁴ Similarly, the survival outcomes of post-TMT SC are incompletely described. Our goal was to describe short (≤90 days) and long-term (>90 days) complications of SC post-TMT and compare outcomes to patients who received primary cystectomy and PC with prior history of non-TMT abdominal/pelvic radiotherapy (PC with Hx XRT).

Our retrospective study included 265 patients with clinical stage cT1-4N0M0 urothelial bladder cancer who underwent radical cystectomy at Massachusetts General Hospital between 2003 and 2013. A total of 216 patients had PC, 28 patients had PC with Hx XRT and 21 patients had SC after TMT. All TMT patients initially had MIBC. Among the 3 groups, there was no difference in the clinical stage prior to cystectomy, but there was a difference in age, lymphovascular invasion, use of neoadjuvant chemotherapy and type of urinary diversion. Patients who received primary cystectomy were younger and had a higher incidence of lymphovascular invasion. Neoadjuvant chemotherapy was only used in PC, and ileal conduits were the only type of urinary diversion in SC patients.

Comparing intraoperative events and estimated blood loss, there was no difference among the 3 groups. For early (≤90 days) complications, there was no difference in the predicted probability of an overall early complication (p=0.17; fig. 1). Early respiratory, infectious and neurological complications were relatively more common in the SC group (p=0.01, p=0.02, p=0.007, respectively). The risk of major early complications (Clavien-Dindo 3–5) was similar among the groups.

Figure 1. Predicted probability of any early complication (≤90 days).

Figure 2. Adjusted freedom from any late complications for PC, salvage cystectomy and PC with past medical history of non-TMT abdominal/pelvic XRT.

Figure 3. Adjusted freedom from of late major (Clavien-Dindo Grade 3–5) complications for PC, salvage cystectomy (SC) and primary cystectomy with past medical history of non-TMT abdominal/pelvic XRT (PC with Hx XRT)

Figure 4. Adjusted disease-specific survival (A) and overall survival (B) for PC, SC and PC with past medical history of non-TMT abdominal/pelvic XRT (PC with Hx XRT).

For late (>90 days) complications, adjusted freedom analysis demonstrated a difference among the 3 groups (p=0.03; fig. 2). On a secondary analysis comparing SC with PC with XRT, there was no statistical difference in the incidence of late complications (p=0.5). Comparing SC with PC, SC had a significantly higher incidence of late complications (p=0.002). Previous TMT was associated with higher incidence of late complications on multivariable Cox regression analysis (HR 2.3, 95% CI 1.2–4.4, p=0.02). SC had a relatively higher incidence of infectious, gastrointestinal and genitourinary complications (p=0.03, p <0.01, p=0.03, respectively). Adjusted freedom analysis of major late complications did not demonstrate a difference among the 3 groups (p=0.1). On secondary analysis of major late complications comparing SC with PC, SC was associated with a higher late complication rate (p=0.049; fig. 3). On multivariable Cox regression analysis, SC was associated with an increased risk of major late complications (HR 2.1, 95% CI 1.0–4.2, p <0.05).

The DSS and overall survival from the time of cystectomy were similar among the 3 groups (p=0.8, p=0.9, respectively) on adjusted analysis (fig. 4).⁵

In summary, SC after TMT is often curative with an acceptable complication rate. SC has a similar immediate and early safety profile when compared to SC and PC with XRT. There are more late complications, both all and major, in SC patients. However, this does not impact overall survival. This study supports the use of SC for post-TMT intravesical recurrences that require cystectomy.