Benefit of Adjuvant Chemotherapy After Radical Cystectomy for Treatment of Urothelial Carcinoma of the Bladder in the Elderly –An International Multicenter Study

INTRODUCTION

Radical cystectomy (RC) is the standard treatment for muscle-invasive bladder cancer (MIBC); but due to its heterogeneous nature and high rate of occult metastases, approximately half of all patients will ultimately succumb to disease progression despite apparent cure with extirpative surgery [1–4]. There is an unmet need to provide reliable risk-stratification tools for patient selection towards perioperative systemic therapy [5], as biomarkers that add sufficient value on outcome prediction are still missing [6–13]. Furthermore, clinical stage is discrepant with final pathologic stage and only postoperative pathologic features offer the highest prognostic value [4, 13–15]. Still, due to the aggressiveness of MIBC, in all eligible patients, RC and cisplatin-based neoadjuvant chemotherapy (NAC) is considered as the standard of care, due to level one evidence demonstrating a net-benefit in overall survival (OS) and recurrence-free survival (RFS), relative to no NAC [3, 17].

Adjuvant chemotherapy (AC) has frequently been favored over NAC as treatment decisions can be based on pathological staging, however, there are only weak data comparing the efficacy of both treatment modalities [18]. This could be especially true for older patients, as treating physicians may find the potential detrimental effect of NAC and the risk of overtreatment particularly disadvantageous in this specific group of patients [19]. As MIBC is considered more aggressive in the elderly population, older patients may therefore especially benefit from AC [20, 21]. However, only observational studies and meta-analyses have demonstrated a clear benefit to RFS and OS for the use of AC vs. surgery alone [22, 23]. Multiple prospective studies failed to confirm its efficacy over deferred chemotherapy at time of recurrence due to poor accrual [24–26]. Patient selection for the use of AC is of highest importance, as especially patients with lymph node metastases and/or ≥pT3 disease seem to benefit from AC [5, 28]. Older patients are known to be less likely to receive appropriate treatment for MIBC, including a less frequent administration of AC, even though it has been demonstrated that they can tolerate platinum-based chemotherapy sufficiently well [29, 30]. However, the real-world benefit of AC among these patients remains poorly defined.

We hypothesized that administration of AC can improve survival outcomes in elderly patients treated with RC for urothelial carcinoma of the bladder (UCB). To test this hypothesis, we compared survival outcomes of patients treated with or without AC after RC in a large, international-multicenter study. We also conducted multiple subgroup analyses in order to evaluate which patients may particularly benefit from AC.

METHODS

RESULTS

Survival analyses

Median follow up of patients alive was 42.4 months (IQR 18.3–85.1) for the entire cohort. The 5-year estimates for RFS, CSS and OS were 60.8%(95%CI 59.1–62.5%), 66.9%(95%CI 65.3–68.6%) and 55.9%(95%CI 54.2–57.6%), respectively. On survival analyses, patients aged ≥75 years had significantly worse survival outcomes with respect to RFS, CSS and OS compared to patients <75 years (p < 0.001, Fig. 1).

OPEN IN VIEWER

Fig. 1

Kaplan-Meier Curves for 5-Year Recurrence-Free Survival; Cancer-Specific Survival and Overall Survival by Age Groups (< 75 vs. ≥75 years old).

Effect of adjuvant chemotherapy

On univariable survival analyses of the entire cohort, patients who had received AC had a significantly worse 5-year CSS (47.3%vs. 73.6%, p < 0.001) and 5-year OS (42.2%vs. 60.3%, p < 0.001) in comparison to patient who had not received AC. 5-year RFS was found to be similar for both groups (60.5%vs. 61.8%, p = 0.37). In a subgroup analyses of patients aged≥75 years, patients who had received AC also suffered significantly worse 5-year RFS (30.9%vs. 58.1%, p < 0.001), 5-year CSS (34.2%vs. 64.2%, p < 0.001) and 5-year OS (28.7%vs 43.6%, p = 0.007).

In a subgroup analyses of patients ≥75 years with lymph node metastases, patients who had received AC showed a significantly better 5-year OS (41%vs. 30.9%, p = 0.02). While 5-year RFS and CSS were also favorable for patients who had received AC, this did not reach statistical significance (Fig. 2). For patients ≥75 years with either ≥pT3 disease or any NOCD (non-organ confined disease), there were no significant group differences with respect to the administration of AC and survival outcomes (Figs. 3,4). In comparison, patients <75 years with lymph node metastases who had received AC also showed a significantly better 5-year RFS (32.1%vs. 26.6%, p < 0.001), 5-year CSS (38.2%vs. 31.5%, p = 0.011) and 5-year OS (34.3 vs. 24.4%, p < 0.001). However, again in patients <75 years with either ≥pT3 disease or any NOCD, there were no significant group differences with respect to the administration of AC and survival outcomes (p > 0.05 for all endpoints).

OPEN IN VIEWER

Fig. 2

Kaplan-Meier Curves Demonstrating The Effect of Adjuvant Chemotherapy on Recurrence-Free Survival; Cancer-Specific Survival and Overall Survival in Patients ≥75 Years Old with Lymph Node Metastases.

OPEN IN VIEWER

Fig. 3

Kaplan-Meier Curves Demonstrating The Effect of Adjuvant Chemotherapy on Recurrence-Free Survival; Cancer-Specific Survival and Overall Survival in Patients ≥75 Years Old with ≥pT3 Disease.

OPEN IN VIEWER

Fig. 4

Kaplan-Meier Curves Demonstrating the Effect of Adjuvant Chemotherapy on Recurrence-Free Survival; Cancer-Specific Survival and Overall Survival in Patients Over 75 Years with any Non-Organ Confined Disease.

In a multivariable Cox model that was adjusted for several established outcome predictors, patients ≥75 years who had received AC showed favorable survival outcomes with respect to OS compared to patients ≥75 years who had not received AC (HR 0.75 [95%CI 0.56–0.99] p = 0.045, Table 2). This effect was even more pronounced in the subgroups of patients with either lymph node metastases (HR 0.64 [95%CI 0.45–0.9] p = 0.011) or ≥pT3 disease (HR 0.63 [95%CI 0.46–0.86] p = 0.003, Table 2). However, there was no significant effect on either RFS or CSS and the association remained insignificant in all further subgroup analyses. The same model in patients aged <75 years also showed a favorable effect for the administration of AC with respect to OS (HR 0.89 [95%CI 0.79–1.0] p = 0.047) and on subgroup analyses (lymph node metastases: HR 0.66 [95%CI 0.56–0.79] p < 0.001; ≥pT3disease: HR 0.84 [95%CI 0.73–0.96] p = 0.013). However, there was again no significant association between the administration of AC and RFS or CSS and the association remained insignificant in all further subgroup analyses.

OPEN IN VIEWER

Table 2

Multivariable Cox Regression Analyses of the Association of Adjuvant Chemotherapy with Clinicopathologic Characteristics in 820 Patients ≥75 Years Old Treated with Radical Cystectomy for Urothelial Carcinoma of the Bladder

Subgroups	Variable	n	Recurrence-free Survival			Cancer-specific Survival			Overall Survival
			HR	95%CI	p	HR	95%CI	p	HR	95%CI	p
All patients ≥75 years old (n = 820)	Use of adjuvant chemotherapy (Ref.: no)	102	0.95	0.7–1.3	0.77	0.85	0.61–1.2	0.35	0.75	0.56–0.99	0.045
	Gender (Ref.: female)	621	0.97	0.74–1.3	0.8	1.01	0.76–1.3	0.96	1.01	0.83–1.25	0.89
	≥pT3 disease (Ref.: < pT3)	447	1.81	1.37–2.4	< 0.001	2.15	1.58–2.9	< 0.001	1.7	1.38–2.11	< 0.001
	LNM (Ref.: negative)	221	2.28	1.75–3.0	< 0.001	2.62	1.98–3.5	< 0.001	1.97	1.58–2.46	< 0.001
	LVI (Ref.: negative)	328	1.81	1.41–2.3	< 0.001	1.8	1.38–2.4	< 0.001	1.52	1.25–1.85	< 0.001
	Surgical margins (Ref.: negative)	73	1.41	1.02–1.9	0.039	1.6	1.14–2.2	0.006	1.2	0.89–1.61	0.23
	CIS (Ref.: negative)	413	0.96	0.76–1.2	0.74	1.02	0.79–1.3	0.88	1.25	1.04–1.5	0.02
All patients ≥75 years old with lymph node metastases (n = 221)	Use of adjuvant chemotherapy (Ref.: no)	69	0.85	0.59–1.2	0.39	0.75	0.51–1.1	0.14	0.64	0.45–0.9	0.011
	Gender (Ref.: female)	164	0.83	0.57–1.2	0.35	0.89	0.59–1.3	0.55	1.16	0.8–1.7	0.44
	≥pT3 disease (Ref.: < pT3)	178	1.93	1.17–3.2	0.01	2.33	1.33–4.1	0.003	2.27	1.42–3.6	< 0.001
	LVI (Ref.: negative)	143	1.61	1.09–2.4	0.016	1.86	1.23–2.8	0.003	1.78	1.24–2.6	0.002
	Surgical margins (Ref.: negative)	42	1.37	0.91–2.1	0.126	1.72	1.13–2.6	0.012	1.45	0.99–2.1	0.055
	CIS (Ref.: negative)	109	0.93	0.65–1.3	0.68	0.9	0.62–1.3	0.6	1.02	0.73–1.4	0.91
All patients ≥75 years old with ≥pT3 disease (n = 447)	Use of adjuvant chemotherapy (Ref.: no)	81	0.78	0.56–1.1	0.15	0.72	0.5–1.0	0.071	0.63	0.46–0.86	0.003
	Gender (Ref.: female)	341	1.06	0.77–1.4	0.83	1.13	0.81–1.6	0.48	1.24	0.94–1.63	0.13
	LNM (Ref.: negative)	178	2.41	1.79–3.3	< 0.001	2.77	2.01–3.8	< 0.001	2.15	1.66–2.79	< 0.001
	LVI (Ref.: negative)	252	2.08	1.54–2.8	< 0.001	2.17	1.58–3.0	< 0.001	1.82	1.42–2.34	< 0.001
	Surgical margins (Ref.: negative)	62	1.4	1.0–2.0	0.052	1.58	1.12–2.2	0.01	1.18	0.86–1.61	0.3
	CIS (Ref.: negative)	193	0.91	0.69–1.2	0.52	0.97	0.72–1.3	0.83	1.17	0.92–1.48	0.2

Ref. = Reference, LNM = Lymph Node Metastases, CIS = Carcinoma in situ, LVI = Lymphovascular invasion, HR = Hazard ratio 95%CI=95%Confidence interval, p = P-Value.

DISCUSSION

Despite cumulative evidence to higher incidence and mortality in the ever-growing elderly population, there is very little data concerning the effectiveness of multimodal treatment strategies for UCB for these patients [19, 36]. With this retrospective analysis, we aimed to inform the debate concerning the optimal management of UCB in the elderly with real world evidence on the benefit of AC after RC. Management of UCB in the elderly should take into consideration that, while aging is in fact a heterogeneous process, many elderly patients are frail and their tumors show more aggressive behavior [19–21]. In our large multicenter database, we verified that elderly patients are more likely to suffer from advanced disease. Still, we found that elderly patients were less likely to receive AC. These findings are in line with previous findings by Leveridge et al., who analyzed the outcomes of 1,331 patients aged ≥75 years undergoing RC for UCB from 1994 to 2008 [37].

On univariable survival analyses of patient aged ≥75 years, we were not able to demonstrate a significant benefit for the administration of AC with respect to RFS, CSS or OS. Several previous studies that analyzed a non-age specific cohort had similar findings, and it was only on meta-analyses that a small, yet significant survival benefit could be demonstrated [22, 26]. The likely explanation is that only patients with poor prognostic factors were selected for administration of AC and subsequently suffered worse survival outcomes due to advanced and aggressive disease. In a subgroup analysis of patients with lymph node metastases, we found favorable outcomes after the administration of AC with respect to OS. Contrary to data by Leveridge et al., in our study, this association did not reach statistical significance for RFS or ≥pT3 disease. These conflicting results may be due to different cut off groups that were used for classification of age. Overall, our findings emphasize that the effect of AC in the elderly is significantly modified by the individual’s risk of disease recurrence and optimal patient selection is of the utmost importance in clinical decision making.

After adjusting for established prognostic variables, we were able to demonstrate a significant favorable association between the administration of AC and OS on multivariable Cox regression analyses. This effect was even more pronounced in the subgroups of patients with either lymph node metastases or ≥pT3 disease and similar to the effect in patients aged < 75 years. While an OS benefit is arguably the most unambiguous and ultimately most important endpoint for many patients, it may not be the optimal endpoint in the elderly population and could potentially be affected by a selection bias. In contrary, even on subgroup analyses of patients with advanced disease there was no benefit with respect to RFS or CSS. These conflicting results may be due to the fact that the optimal chemotherapeutic regimen for elderly patients in this specific setting remains unknown. A recent non-age specific meta-analysis of the benefit of AC reported that only the regimen of cisplatin, gemcitabine and paclitaxel is associated with significant improvement in both RFS and OS [38]. However, patients aged ≥70 years are known to be less likely to receive cisplatin-based chemotherapy than younger patients [37]. However, just like the curative treatment of RC should not be withheld from elderly patients, they should also not be withheld a potentially life-saving AC only based on chronological age alone. The treatment of UCB in the elderly should be individualized, focusing on biological age and performance status [19, 20]. A multidisciplinary approach and a geriatric evaluation are needed to identify patients eligible for AC [20]. Our study shows further trials concerning the optimal chemotherapeutic regimen and the value of presumably more tolerable agents, such as non-cisplatin-based chemotherapy, are required to investigate adjuvant treatment strategies in the management of UCB in the elderly. The demonstrated inefficacy of AC to improve RFS or CSS in the elderly population also warrants the investigation of the benefit of novel immunotherapeutic drugs in the adjuvant and neoadjuvant setting, as such agents might demonstrate a more compelling clinical net-benefit for all endpoints and thus change clinical practice.

While the strength of this cohort is its homogeneity in treatment allocation and its international, multicenter nature, the study is limited by its retrospective design and the short median follow-up of 42.4 months. However, previous data suggest that over two-third of patients experience disease recurrence after RC within 12 months and ≥90%within 24 months [39]. We do not have any data, why patients who had an indication for administration of AC did not receive it. This could reflect a selection bias. Nevertheless, we feel that this does not necessarily contradict our main finding, which is that optimal patient selection is necessary to identify patients who are most likely to tolerate and benefit from AC. Another limitation is the missing information concerning the specific chemotherapeutic regimen administered and about the administration of deferred chemotherapy at time of recurrence. Furthermore, more appropriate endpoints such as quality of life or as functional independence would be preferable in the elderly population [40]. Well-designed, randomized trials would be superior to fully establish the administration of AC in the elderly population and further improve patient selection, however, the advent of novel agents suggests that such a randomized clinical trial is unlikely to ever be successfully concluded.

CONCLUSION

In this large observational study, AC was associated with improved OS, but not RFS or CSS, in elderly patients treated with RC for UCB. This is of clinical importance, as elderly patients are more likely to have adverse pathologic features and experience worse survival outcomes. Treatment of UCB should include both a multidisciplinary approach and a geriatric evaluation to identify patients who are most likely to tolerate and benefit from AC. Elderly patients should not be precluded from AC due to their chronological age alone.