Abstract
Neoadjuvant cisplatin-based chemotherapy (NAC) confers a survival benefit in muscle-invasive bladder cancer (MIBC), however a relevant proportion of patients experience early disease recurrence following radical cystectomy. We conducted a multicenter observational study to evaluate clinical, pathological, and laboratory factors associated with early recurrence defined as disease-free survival [DFS] ≤ 12 months in patients with MIBC treated with neoadjuvant cisplatin plus gemcitabine followed by radical cystectomy. Early recurrence was significantly associated with aggressive pathological features, including higher ypT stage, nodal involvement, lymphovascular invasion, and lower rates of pathological complete response. Patients experiencing early recurrence predominantly presented with systemic dissemination and had significantly shorter DFS and overall survival (OS). Variant histology and the presence of post-treatment lymphovascular invasion were independently associated with early recurrence. Early recurrence after NAC identifies a subgroup of patients with aggressive clinicopathological features and poorer oncological outcomes, highlighting the need for improved postoperative risk stratification and novel perioperative treatment strategies.
Subject terms: Cancer, Oncology, Urology
Introduction
Muscle-invasive bladder cancer (MIBC) is an aggressive malignancy associated with a high risk of disease recurrence and cancer-related mortality1. Radical cystectomy with pelvic lymph node dissection remains the cornerstone of curative-intent therapy; however, a substantial proportion of patients experience disease relapse, frequently within the first year after surgery, reflecting the presence of occult micrometastatic disease at the time of diagnosis2,3.
Neoadjuvant chemotherapy (NAC) with cisplatin-based combinations has demonstrated a significant survival benefit in patients with MIBC and is currently recommended for eligible patients4. Among available regimens, cisplatin plus gemcitabine is widely adopted due to its favorable efficacy-toxicity profile4. Nevertheless, responses to NAC are heterogeneous, and a considerable subset of patients derives limited benefit, ultimately developing early disease recurrence after radical cystectomy3. Identifying patients at high risk of early relapse remains a clinically relevant unmet need, as this population may benefit from intensified surveillance or alternative perioperative approaches.
Although pathological complete response after NAC is a well-established surrogate marker of favorable prognosis, increasing evidence suggests that additional clinical, pathological, and biological factors may influence long-term outcomes beyond pCR alone5–8. Adverse pathological features at cystectomy, including residual tumor burden, nodal involvement, and lymphovascular invasion, have been associated with early disease progression7–9. Moreover, the timing and pattern of recurrence may reflect distinct biological phenotypes, with early relapse often characterized by rapid systemic dissemination and poor survival outcomes.
In this context, the present multicenter observational study aimed to investigate factors associated with early disease recurrence (DFS ≤ 12 months) in patients with MIBC treated with neoadjuvant cisplatin and gemcitabine followed by radical cystectomy. Additionally, we sought to characterize recurrence patterns and survival outcomes according to time to relapse, with the goal of refining risk stratification and informing post-surgical management strategies in this patient population.
Materials and methods
Study design and population
This was a multicenter, observational study conducted across seven academic and non-academic institutions in Italy. The study aimed to identify clinical, pathological, and laboratory factors associated with early disease recurrence in patients with MIBC treated with NAC followed by radical cystectomy. Consecutive patients who received cisplatin- and gemcitabine-based NAC and subsequently underwent radical cystectomy with curative intent between June 2018 and June 2023 were included. Neoadjuvant cisplatin plus gemcitabine was administered according to local institutional practice. Treatment schedule, number of administered cycles, and supportive measures were managed at the discretion of treating physicians. Data were collected through structured review of electronic medical records at each participating center. This was a multicenter observational cohort study. Although clinical data were collected from both retrospectively reviewed medical records and prospectively maintained institutional databases, the present analysis was conducted retrospectively. Given the observational design, all treatment decisions were made according to routine clinical practice and in compliance with national and international guidelines. No protocol-mandated interventions were applied. The study was approved by the Institutional Review Board of the Comitato Etico Regionale per la Sperimentazione Clinica della Regione Toscana (Sezione Area Vasta Centro; Careggi University Hospital, Florence; approval number ESR-20-20559). All methods were performed in accordance with the relevant guidelines and regulations.
Eligibility criteria
Eligible patients were aged 18 years or older and had histologically confirmed muscle-invasive urothelial carcinoma of the bladder. Patients with predominant urothelial carcinoma, including tumors harboring variant histology, were eligible for inclusion. Pure non-urothelial histologies were not included. For the purpose of analysis, histology was categorized as pure urothelial carcinoma versus urothelial carcinoma with variant histology. All patients received NAC with cisplatin plus gemcitabine followed by radical cystectomy with curative intent. Tumor diagnosis and staging were established using standard clinical, radiological, and pathological assessments.
Patients with evidence of distant metastatic disease at diagnosis or with locally advanced tumors invading the pelvic or abdominal wall (clinical stage T4b) were excluded. The indication for NAC and the planned number of treatment cycles, were determined by the treating physician based on performance status, overall clinical fitness and renal function, in accordance with contemporary clinical guidelines.
Data collection
Demographic information, including age and sex, was recorded for all patients. Clinical and tumor-related variables included clinical T and N stage, histological subtype, and overall stage at diagnosis. Baseline functional status was assessed using the Eastern Cooperative Oncology Group (ECOG) performance status. Information on comorbidities and previous oncologic treatments was also collected.
Treatment-related variables included the number of NAC cycles administered, radiologic response to chemotherapy, and treatment-related toxicity10,11. Radiologic response after neoadjuvant chemotherapy was collected from routine imaging assessments performed at participating centers according to local clinical practice. Response categories were extracted from radiology reports and clinical documentation available in medical records. Although response evaluation was broadly consistent with RECIST 1.1 principles, formal centralized RECIST 1.1 reassessment was not systematically performed across all centers. Adverse events were retrieved from medical records and graded when possible, with particular attention to grade 3–4 toxicities and hematologic adverse events.
Laboratory parameters routinely assessed in clinical practice were collected both before and after completion of NAC. These included hematologic indices and biochemical markers, allowing exploratory analyses of systemic inflammatory and metabolic parameters in relation to oncologic outcomes.
Surgical and pathological assessment
All patients underwent radical cystectomy with pelvic lymph node dissection following completion of NAC The surgical approach (open, laparoscopic, or robotic) and perioperative characteristics were documented. Pathological evaluation of cystectomy specimens was performed locally at each institution in accordance with standard pathological procedures. Histopathological evaluation, including variant histology classification, was performed according to the World Health Organization (WHO) Classification of Urinary and Male Genital Tumours (5th edition).
Pathological staging was reported using the TNM classification, including assessment of residual primary tumor (ypT), nodal involvement (ypN), and the presence of lymphovascular invasion. Pathological complete response was defined as the absence of residual tumor in both the bladder and lymph nodes (ypT0N0). Additional pathological response categories, including non–muscle-invasive downstaging, were also recorded.
Follow-up and outcome measures
Patients were followed according to institutional standards after radical cystectomy. Disease-free survival (DFS) was defined as the time from surgery to the first documented disease recurrence or death from any cause, whichever occurred first. For the purpose of the present analysis, patients were categorized according to the occurrence of early recurrence, defined as DFS ≤ 12 months after radical cystectomy. Patients with DFS > 12 months, irrespective of whether recurrence subsequently occurred, were included in the comparator group (no early recurrence). OS was defined as the time from surgery to death from any cause. Patterns of disease recurrence were recorded and categorized according to site and extent of metastatic involvement.
Statistical analysis
Descriptive statistics were used to summarize patient characteristics. Continuous variables were reported as medians with ranges, while categorical variables were expressed as absolute numbers and percentages. Comparisons between patients with early recurrence (DFS ≤ 12 months) and those without early recurrence (DFS > 12 months) were performed using appropriate statistical tests based on data distribution. The 12-month cutoff was selected a priori to define a clinically relevant early recurrence phenotype, as relapse within the first postoperative year is generally considered to reflect aggressive disease biology and poor prognosis. Logistic regression was therefore used to explore factors associated with early disease recurrence (DFS ≤ 12 months). Candidate variables were first assessed in univariate logistic regression analyses. Variables with a p value < 0.10 in univariate analysis were subsequently entered into a multivariate logistic regression model to identify independent predictors of early recurrence. Odds ratios (ORs) with corresponding 95% confidence intervals (CIs) were reported.
All statistical tests were two-sided, and a p value ≤ 0.05 was considered statistically significant. Statistical analyses were performed using Stata version 18.0 (StataCorp LLC, College Station, TX, USA).
Results
Patient population and baseline characteristics
A total of 115 patients with MIBC treated with neoadjuvant cisplatin plus gemcitabine followed by radical surgery were included in the study. Early recurrence (DFS ≤ 12 months) occurred in 26 patients (22.6%), whereas 89 patients (77.4%) did not experience early recurrence within the first postoperative year. Overall, 39 patients (33.9%) developed disease recurrence, including 13 patients who relapsed beyond 12 months.
Baseline demographic and clinical characteristics according to early recurrence status are summarized in Table 1. No significant differences were observed between patients with early recurrence and those without early recurrence regarding age, sex, ECOG performance status, clinical T or N stage, or overall clinical stage at diagnosis. Patients with early recurrence more frequently presented with urothelial carcinoma with variant histology compared with patients without early recurrence (23.1% vs. 9.0%, respectively; p = 0.083).
Table 1.
Characteristics of patients.
| Characteristics | Total N = 115 |
Early recurrence DFS ≤ 12 months (n = 26) |
No early recurrence DFS > 12 months (n = 89) |
P | |
|---|---|---|---|---|---|
| Age | Median (range) | 67 (32–81) | 63.5 (53–79) | 67 (32–81) | 0.624 |
| Sex | Male | 98 (85.2) | 19 (73.1%) | 79 (88.7%) | 0.061 |
| ECOG PS | 1 | 5 (4.3%) | 3 (11.5%) | 2 (2.2%) | 0.075 |
| cT | 2 | 87 (75.7%) | 20 (76.9%) | 67 (75.3%) | 0.575 |
| 3 | 25 (21.7%) | 5 (19.2%) | 20 (22.5%) | ||
| 4 | 2 (1.7%) | 1 (3.8%) | 1 (1.1%) | ||
| NA | 1 (0.9%) | 1 (2.9%) | 1 (1.1%) | ||
| cN | 0 | 86 (74.8%) | 18 (69.2%) | 68 (76.4%) | 0.202 |
| 1 | 11 (9.6%) | 5 (19.2%) | 6 (6.7%) | ||
| 2 | 11 (9.6%) | 3 (11.6%) | 8 (9.0%) | ||
| NA | 7 (6.1%) | 0 | 7 (7.9%) | ||
| Stage | III | 45 (39.1%) | 11 (42.3%) | 34 (38.2%) | 0.820 |
| Hystology | Non-urothelial | 14 (12.1%) | 6 (23.1%) | 8 (9.0%) | 0.083 |
N: number; ECOG PS: Eastern Cooperative Oncology Group Performance Status; NA: not available.
Neoadjuvant chemotherapy and surgical outcomes
Chemotherapy-related characteristics are reported in Table 2. Both groups received a median of four NAC cycles. Radiologic response and grade 3–4 toxicity rates were comparablebetween groups.
Table 2.
Chemotherapy characteristics.
| Characteristics | Total N = 115 |
Early recurrence DFS ≤ 12 months (n = 26) |
No early recurrence DFS > 12 months (n = 89) |
P | |
|---|---|---|---|---|---|
| Cycles | Median (range) | 4 (2–5) | 4 (2–5) | 4 (2–5) | 1.000 |
| Radiologic response after NAC | CR | 14 (12.2%) | 2 (7.7%) | 12 (13.5%) | 0.162 |
| SD | 49 (42.6%) | 10 (38.5%) | 39 (43.8%) | ||
| PD | 7 (6.1%) | 4 (15.4%) | 3 (3.4%) | ||
| PR | 33 (28.7%) | 6 (23.1%) | 27 (30.3%) | ||
| NA | 12 (10.4%) | 4 (15.4%) | 8 (9.0%) | ||
| Toxicity (any grade) | 66 (57.4%) | 11 (42.3%) | 55 (61.8%) | 0.114 | |
| Toxicity (grade 3–4) | 11 (9.6%) | 3 (11.5%) | 8 (9.0%) | 0.710 | |
| Hematologic toxicity (grade 3–4) | 23 (20.0%) | 5 (19.2%) | 18 (20.2%) | 1.000 |
N: number; NAC: neoadjuvant chemotherapy; CR: complete response; PR: partial response; SD: stable disease; PD: progressive disease; NA: not available; RC: radical cystectomy.
Surgical and pathological findings are detailed in Table 3; Fig. 1. Patients with early recurrence were significantly less likely to achieve a pathological complete response (pCR) compared with patients without early recurrence (3.8% vs. 31.5%, p = 0.004). Moreover, early recurrence was strongly associated with adverse pathological features, including higher ypT stage (p < 0.001), nodal involvement (p < 0.001), and the presence of post-treatment lymphovascular invasion (p = 0.008). These findings suggest a more aggressive residual disease phenotype in patients who subsequently developed early relapse.
Table 3.
Surgery and pathological characteristics.
| Characteristics | Total N = 115 |
Early recurrence DFS ≤ 12 months (n = 26) |
No early recurrence DFS > 12 months (n = 89) |
P | |
|---|---|---|---|---|---|
| Surgery technique | Laparoscopy | 12 (10.4%) | 3 (11.5%) | 9 (10.1%) | 0.770 |
| Open | 48 (41.7%) | 10 (38.5%) | 38 (42.7%) | ||
| Robotic | 36 (31.3%) | 6 (23.1%) | 30 (33.7%) | ||
| NA | 19 (16.6%) | 7 (26.9%) | 12 (13.5%) | ||
| Comorbidity | Yes | 21 (18.6%) | 7 (26.9%) | 14 (16.1%) | 0.252 |
| Pathological complete response | Yes | 29 (25.2) | 1 (3.8%) | 28 (31.5%) | 0.004 |
| ypT | ypTis. ypTa | 15 (9.7%) | 0 | 15 (16.9%) | < 0.001 |
| ypT0 | 31 (20.0%) | 2 (7.7%) | 29 (32.6%) | ||
| ypT1 | 8 (5.2%) | 1 (3.8%) | 7 (7.9%) | ||
| ypT2 | 28 (18.1%) | 6 (23.1%) | 22 (24.7%) | ||
| ypT3 | 24 (15.5%) | 13 (50.0%) | 11 (12.4%) | ||
| ypT4 | 9 (5.8%) | 4 (15.4%) | 5 (5.6%) | ||
| ypN | ypN0 | 91 (58.7%) | 11(42.3%) | 80 (89.9%) | < 0.001 |
| ypN1 | 8 (5.2%) | 5 (19.2%) | 3 (3.4%) | ||
| ypN2 | 13 (8.4%) | 8 (30.8%) | 5 (5.6%) | ||
| ypN3 | 1 (0.6%) | 1 (3.8%) | 0 | ||
| NA | 2 (1.3%) | 1 (3.8%) | 1 (1.1%) | ||
| Residual disease | 0 | 108 (93.9%) | 20 (76.9%) | 88 (98.9%) | 0.008 |
| 1 | 2 (1.3%) | 2 (7.7%) | 0 | ||
| 2 | 1 (0.6%) | 1 (3.8%) | 0 | ||
| NA | 4 (2.6%) | 3 (11.5%) | 1 (1.1%) |
NA: not available.
Fig. 1.
Pathological features according to early recurrence status. Rates of pathological complete response (pCR), ypT ≥ 3 disease, and nodal involvement (ypN+) in patients with early recurrence (DFS ≤ 12 months) versus patients without early recurrence (DFS > 12 months).
Recurrence patterns
The distribution of recurrence sites according to time to recurrence is reported in Table 4; Fig. 2. Patients with early recurrence exhibited a highly heterogeneous and predominantly systemic pattern of disease progression. Lung metastases represented the most frequent site, occurring in 8 out of 26 patients (30.8%), either as isolated pulmonary relapse or in combination with additional metastatic sites such as lymph nodes or bone. Lymph node involvement was observed in 11 patients (42.3%) and included pelvic, iliac, para-aortic, retroperitoneal, and distant nodal stations. Pelvic and/or peritoneal recurrences occurred in 5 patients (19.2%), frequently as part of multisite disease. Bone metastases were documented in 5 patients (19.2%), rarely as an isolated site.
Table 4.
Patterns of recurrence among patients experiencing disease relapse.
| Site of recurrence | Early recurrence DFS ≤ 12 months (n = 26) |
Late recurrence DFS > 12 months (n = 13) |
|---|---|---|
| Lung (isolated or combined) | 8 (30.8%) | 3 (23.1%) |
| Lymph nodes (any site) | 11 (42.3%) | 3 (23.1%) |
| Pelvic / Peritoneal | 5 (19.2%) | 2 (15.4%) |
| Bone | 5 (19.2%) | 6 (46.2%) |
| Multisite recurrence | 15 (57.7%) | 4 (31.0%) |
Fig. 2.
Sites of recurrence in bladder cancer patients treated with neoadjuvant chemotherapy (NAC), stratified by time to recurrence: early (DFS ≤ 12 months) versus late (DFS > 12 months). Data are shown as percentages of patients.
Overall, more than half of patients with early recurrence (57.7%) developed multisite progression, indicating an aggressive disease course with rapid systemic dissemination.
In contrast, patients with recurrence beyond 12 months showed a more homogeneous and localized pattern of relapse. Bone was the most frequently involved site, affecting 6 out of 13 patients (46.2%), either as isolated disease or associated with a single additional metastatic site. Pelvic and/or lymph node recurrences were observed in 3 patients (23.1%), predominantly as locoregional progression. Pulmonary recurrence occurred in 3 patients (23.1%), generally as an isolated metastatic site. Multisite disease was less common in this group (31.0%), suggesting a more indolent biological behavior compared with early recurrent patients.
Survival outcomes
Survival analyses demonstrated marked differences according to time to recurrence. Patients with early recurrence had a median DFS of 8.1 months (95% CI, 6.8–9.4), whereas median DFS was not reached in patients without early recurrence (95% CI, not reached–not reached; Fig. 3). Similarly, median OS was 40 months (95% CI, 13.8–not reached) in the early recurrence group and was not reached in patients without early recurrence (95% CI, not reached–not reached; Fig. 4).
Fig. 3.
Disease-free survival (DFS) according to early recurrence status. Patients with early recurrence (DFS ≤ 12 months) showed significantly shorter DFS compared with patients without early recurrence (DFS > 12 months).
Fig. 4.
Overall survival (OS) according to early recurrence status. Early recurrence was associated with significantly worse OS compared with patients without early recurrence.
Predictors of early recurrence
Univariate and multivariate logistic regression analyses evaluating factors associated with early recurrence (DFS ≤ 12 months) are reported in Table 5. In univariate analysis, variant histology, higher ypT and ypN stages, post-treatment lymphovascular invasion, absence of pCR, and higher pre-treatment platelet count were associated with an increased risk of early recurrence.
Table 5.
Univariate and multivariate analysis of the relationship of various variables with Early recurrence DFS ≤ 12 months in patients treated with neoadjuvant cisplatin and gemcitabine.
| Variable | Odds ratio | 95% CI | P value |
|---|---|---|---|
| Age | 1 | 0.95–1.06 | 0.89 |
| Sex (male vs. female) | 0.34 | 0.12–1.02 | 0.05 |
| BMI | 0.88 | 0.76–1.02 | 0.08 |
| cT | 1.09 | 0.44–2.70 | 0.86 |
| cN | 1.41 | 0.75–2.64 | 0.29 |
| Stage at diagnosis (III vs. II) | 1.19 | 0.49–2.88 | 0.70 |
| Hystological variant (yes vs. no) | 3.04 | 0.95–9.75 | 0.06 |
| Lymphovascular invasion (yes vs. no) | 1,00 | 0.22–4.54 | 1.00 |
| Number of comorbidity | 0.77 | 0.45–1.31 | 0.33 |
| ECOG (1 vs. 0) | 5.67 | 0.89–35.99 | 0.07 |
| Previous treatment in MBIC (yes vs. no) | 0.17 | 0.02–1.34 | 0.09 |
| Pre-treatment leukocytes | 1.11 | 0.90–1.36 | 0.34 |
| Pre-treatment neutrophils | 1.09 | 0.86–1.39 | 0.47 |
| pre-treatment lymphocytes | 0.99 | 0.45–2.19 | 0.98 |
| Pre-treatment N/L | 0.95 | 0.60–1.48 | 0.81 |
| Pre-treatment monocytes | 1.15 | 0.79–1.67 | 0.46 |
| Pre treatment eosynophils | 0.83 | 0.12–5.91 | 0.86 |
| Pre-treatment N/eos | 0.99 | 0.99-1 | 0.56 |
| Pre-treatment platelet | 1.00 | 1.00-1.01 | 0.02 |
| Pre-treatment P/L | 1.00 | 0.99-0.1.02 | 0.30 |
| Pre-treatment HB | 0.70 | 0.53–0.92 | 0.01 |
| Pre-treatment glycemia | 0.97 | 0.93-1.00 | 0.08 |
| Pre-treatment clearance of creatinine | 0.99 | 0.95–1.03 | 0.59 |
| Pre-treatment sodium | 1.01 | 0.95–1.26 | 0.21 |
| Pre-treatment potassium | 0.89 | 0.34–2.33 | 0.81 |
| Pre-treatment calcium | 1.17 | 0.61–2.24 | 0.64 |
| Pre-treatment ALT | 0.95 | 0.89–1.02 | 0.19 |
| Pre-treatment bilirubin | 1.32 | 0.80–2.17 | 0.27 |
| Pre-treatment LDH | 1.00 | 0.99–1.02 | 0.54 |
| Symptomatic (yes vs. no) | 0.97 | 0.40–2.35 | 0.95 |
| 4 Cycles (yes vs. no) | 0.83 | 0.32–2.16 | 0.70 |
| Post-treatment leukocytes | 1.03 | 0.96–1.11 | 0.43 |
| Post-treatment neutrophils | 1.05 | 0.94–1.18 | 0.34 |
| post-treatment lymphocytes | 1.02 | 0.77–1.35 | 0.90 |
| Post-treatment N/L | 1.07 | 0.83–1.40 | 0.59 |
| Post-treatment monocytes | 2.31 | 0.69–7.74 | 0.17 |
| Post-treatment eosynophils | 14.83 | 0.14-1620.04 | 0.26 |
| Post-treatment N/eos | 0.99 | 0.97-1.00 | 0.26 |
| Post-treatment platelet | 1.00 | 0.99–1.01 | 0.07 |
| Post-treatment P/L | 0.99 | 0.99–1.01 | 0.97 |
| Post-treatment HB | 0.99 | 0.77–1.29 | 0.96 |
| Post-treatment glycemia | 0.99 | 0.97–1.01 | 0.49 |
| Post-treatment clearance of creatinine | 0.99 | 0.95–1.03 | 0.68 |
| Post-treatment sodium | 1.08 | 0.91–1.28 | 0.41 |
| Post-treatment potassium | 0.90 | 0.37–2.19 | 0.82 |
| Post-treatment calcium | 0.60 | 0.19–1.91 | 0.38 |
| Post-treatment ALT | 0.99 | 0.94–1.03 | 0.58 |
| Post-treatment bilirubin | 1.28 | 0.94–17.26 | 0.85 |
| Post-treatment LDH | 1.00 | 0.99–1.01 | 0.88 |
| Surgical procedure (robotic vs. open/laparoscopy) | 0.59 | 0.21–1.62 | 0.31 |
| Comorbidity surgery (yes vs. no) | 1.92 | 0.68–5.43 | 0.23 |
| Post Lymphovascular invasion (yes vs. no) | 4.98 | 1.60–15.50 | 0.01 |
| Toxicity any grade (yes vs. no) | 0.45 | 0.19–1.01 | 0.08 |
| Emathological toxicity any grade (yes vs. no) | 0.94 | 0.31–2.83 | 0.91 |
| Grade 3–4 toxicity (yes vs. no) | 1.32 | 0.32–5.38 | 0.70 |
|
Response rate after chemotherapy (yes vs. no) |
0.62 | 0.23–1.63 | 0.33 |
| ypT | 2.43 | 1.58–3.75 | < 0.001 |
| ypN | 3.96 | 2.11–7.43 | < 0.001 |
| PCR (yes vs. no) | 0.08 | 0.01–0.68 | 0.02 |
| Multivariate analysis | |||
| Sex (male vs. female) | 0.46 | 0.10–1.78 | 0.72 |
| BMI | 1.02 | 0.61–1.78 | 0.93 |
| Hystological variant (yes vs. no) | 2.57 | 1.25–4.83 | 0.02 |
| ECOG (1 vs. 0) | 4.44 | 0.78–5.50 | 0.78 |
| Previous treatment in MBIC (yes vs. no) | 0.45 | 0.33–1.45 | 0.32 |
| Pre-treatment platelet | 1.01 | 0.99–1.03 | 0.14 |
| Pre-treatment HB | 0.55 | 0.16–1.85 | 0.34 |
| Pre-treatment glycemia | 0.85 | 0.72–1.01 | 0.10 |
| Post-treatment platelet | 0.99 | 0.98–1.01 | 0.30 |
| Post lymphovascular invasion (yes vs. no) | 3.57 | 1.33–8.89 | 0.05 |
| Toxicity any grade (yes vs. no) | 0.12 | 0.01–6.53 | 0.30 |
| ypT | 1.20 | 0.58–12.4 | 0.72 |
| ypN | 2.70 | 0.58–12.4 | 0.20 |
PCR has been excluded for collinearity.
On multivariate analysis, variant histology (OR 2.57, 95% CI 1.25–4.83; p = 0.02) and post-treatment lymphovascular invasion (OR 3.57, 95% CI 1.33–8.89; p = 0.05) remained independently associated with early recurrence. pCR was excluded from the multivariate model due to collinearity with pathological stage.
Discussion
In this multicenter observational study, we identified distinct clinical and pathological features associated with early disease recurrence following neoadjuvant cisplatin and gemcitabine chemotherapy and radical cystectomy in patients with muscle-invasive bladder cancer. Early recurrence, defined as disease relapse within 12 months from surgery, was associated with aggressive pathological characteristics, unfavorable patterns of relapse, and significantly worse survival outcomes, underscoring the clinical relevance of early post-surgical disease progression as a surrogate of intrinsic tumor resistance.
Consistent with prior real-world and retrospective series demonstrating that the benefit of NAC is largely driven by pathological response and residual disease burden at cystectomy12–14. Multiple studies have shown that patients achieving pathological complete response or downstaging to non–muscle-invasive disease experience durable disease control and improved survival, whereas those with residual ≥ypT3 or nodal disease remain at high risk of early relapse13.14. Our data reinforce the concept that pathological response integrates both chemosensitivity and underlying tumor biology.
Notably, post treatment LVI was independently associated with early recurrence in multivariate analysis. LVI is increasingly recognized as a robust marker of biological aggressiveness and micrometastatic potential in urothelial carcinoma15. Recent analyses from large institutional and national cohorts have consistently linked LVI to worse oncologic outcomes after radical cystectomy, even in the context of perioperative chemotherapy16. Notably, post-treatment LVI was independently associated with early recurrence in multivariate analysis. LVI is increasingly recognized as a marker of biological aggressiveness and micrometastatic potential in urothelial carcinoma. Our findings suggest that post-treatment LVI may help identify a subgroup of patients at particularly high risk of rapid postoperative relapse despite NAC. However, because LVI is assessed on cystectomy specimens, its clinical role should primarily be viewed as postoperative prognostic stratification rather than a pre-treatment predictive tool for perioperative treatment selection.
The presence of variant histology was independently associated with early relapse. Variant histologies, including squamous differentiation and other recognized urothelial differentiation patterns, have been associated with higher pathological stage and inferior survival outcomes following cystectomy in several retrospective series. Variant histologies, including squamous differentiation, have been associated with higher pathological stage and inferior survival outcomes following cystectomy in several retrospective series17. Although some molecular and clinical data suggest that basal/squamous tumors may retain sensitivity to cisplatin-based chemotherapy, real-world evidence indicates marked heterogeneity in treatment response18. Our findings suggest that, in routine clinical practice, variant differentiation may identify tumors with increased risk of rapid systemic progression despite standard NAC, supporting the need for intensified perioperative strategies in this population. In addition, in contrast to postoperative pathological variables, variant histology is available before treatment initiation and may therefore contribute to upfront risk stratification and individualized perioperative treatment planning.
The observed pattern of failure may suggest biological differences between early and later relapse; however, given the limited sample size, these findings should be considered exploratory. Early recurrence was predominantly systemic, frequently multisite, and commonly involved lung and distant lymph nodes. In contrast, late recurrence was more often limited and localized. This divergence suggests that early relapse largely reflects occult metastatic disease present at diagnosis rather than inadequate local control, underscoring the limitations of surgery alone or chemotherapy-only perioperative approaches in biologically aggressive tumors19.
The therapeutic landscape of MIBC is rapidly evolving. The phase III NIAGARA trial demonstrated improved event-free and OS with the perioperative durvalumab combined with NAC establishing chemo-immunotherapy as a new standard in eligible patients20. Within this evolving framework, our findings may help refine postoperative prognostic stratification, particularly in patients with adverse pathological features such as LVI or nodal involvement who remain at high risk of rapid relapse despite NAC. Whether these factors should guide treatment intensification strategies warrants prospective validation. Whether patients with adverse postoperative pathological features may derive greater benefit from treatment intensification strategies remains speculative and warrants prospective validation.
Moreover, emerging evidence from metastatic disease further challenges the traditional chemotherapy-centric paradigm. The EV-302/KEYNOTE-A3921 trial demonstrated unprecedented survival outcomes with enfortumab vedotin plus pembrolizumab compared with platinum-based chemotherapy in the first-line metastatic setting, suggesting that antibody–drug conjugate–based combinations can overcome resistance mechanisms intrinsic to conventional chemotherapy. While EV-303 is currently exploring this strategy in earlier disease settings, our findings raise the hypothesis that patients experiencing early relapse after NAC may represent ideal candidates for intensified or alternative perioperative approaches incorporating antibody–drug conjugates and immunotherapy.
This study has limitations inherent to its retrospective design and relatively limited sample size, including potential selection bias and reduced statistical power for subgroup analyses. In addition, the limited number of early recurrence events may have increased the risk of model overfitting and unstable estimates in multivariable analyses; therefore, these findings should be interpreted cautiously and considered hypothesis-generating pending external validation. Detailed information regarding chemotherapy delivery, including cisplatin schedule, dose reductions, treatment delays, reasons for treatment discontinuation, and the interval between NAC completion and radical cystectomy, was not consistently available across participating centers. As these variables may influence pathological response and recurrence outcomes, their role should be prospectively evaluated in future studies. Similarly, recurrence pattern analyses among patients who experienced relapse should be considered exploratory due to the limited sample size. The inclusion of multiple laboratory variables in univariate analyses may also have increased the risk of multiple-testing bias and false-positive findings. Furthermore, the use of a dichotomized DFS endpoint at 12 months may have resulted in loss of time-dependent information and reduced statistical efficiency compared with time-to-event approaches.
Finally, radiologic response assessment was retrospectively extracted from routine imaging reports and clinical documentation across participating centers. Although response categorization was generally consistent with RECIST 1.1 principles, formal centralized RECIST 1.1 reassessment was not systematically performed, potentially introducing variability in response evaluation. Nonetheless, the multicenter real-world setting enhances the generalizability of the findings and reflects routine clinical practice.
Identifying patients at high risk of early relapse remains a clinically relevant unmet need, as this population may benefit from intensified surveillance, postoperative risk stratification, or alternative perioperative approaches. Taken together, our findings highlight the heterogeneity of outcomes following neoadjuvant cisplatin and gemcitabine and underscore the importance of distinguishing pre-treatment predictive features from postoperative prognostic markers. Early recurrence identifies a subgroup of patients with poor oncological outcomes for whom improved surveillance, refined postoperative risk stratification, and hypothesis-generating perioperative strategies warrant further investigation.
Acknowledgements
We thank the participating investigators and clinical staff from all contributing centers for their valuable collaboration.
Author contributions
Giandomenico Roviello: Writing – original draft.Elisabetta Gambale: Data curation, Writing – original draft, Visualization, Writing – review & editing.Roberta Giorgione: Data curation.Umberto Basso: Data curation.Maria Oliveri: Data curation.Malvina Cremante: Data curation.Francesco Atzori: Data curation.Sarah Scagliarini: Data curation.Cristina Masini: Data curation.Valentina Baldazzi: Data curation.Federico Scolari: Data curation.Marinella Micol Mela: Data curation.Eleonora Lai: Data curation.Ismaela Anna Vascotto: Data curation.Virginia Rossi: Data curation.Chiara Calandrelli: Data curation.Daniele Rossini: Writing – review & editing.Daniele Lavacchi: Conceptualization, Methodology, Formal analysis, Validation, Writing – review & editing.Sergio Serni: Supervision.Andrea Minervini: Supervision.Serena Pillozzi: Conceptualization, Methodology, Formal analysis, Validation, Writing – review & editing.Lorenzo Antonuzzo: Conceptualization, Methodology, Supervision.
Funding
This study was conducted with the unconditional support of Astrazeneca.
Data availability
The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.
Declarations
Competing interests
GR: Scientific advisory boards, consulting, travel expenses: MSD, Merck Serono, Bayer, BMS, AAA, Astra Zeneca, Pfizer, Astellas. LA has received honoraria or consultation fees for speaker, consultancy, or advisory roles from Amgen, Bayer, Eisai, Merck Serono, Pierre Fabre, Roche, Servier, Incyte, Ipsen, and MSD; has also received travel grants from AstraZeneca, MSD, Merck Serono, and Ipsen; research funding was provided to LA’s institution by Novartis and AstraZeneca. FB has received consulting fees or speaker bureau honoraria from Eli Lilly, MSD, Eisai, Bristol Myers Squibb, AstraZeneca, Pierre Fabre, Servier, and AAA Novartis. DM has participated in advisory boards for Roche, MSD, and Merck. SS has received honoraria or consultation fees for speaker, consultancy, or advisory roles from MSD, BMS, IPSEN, EISAI, MERCK, ASTELLAS, J&j. DR : Scientific advisory boards, consulting, travel expenses: MSD, Merck Serono, Bayer, BMS, AAA, Astra Zeneca, Pfizer. The other Authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Informed consent statement
Written informed consent was obtained from all patients prior to inclusion in the study.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Cigliola, A. et al. The management of muscle invasive bladder cancer: State of the art and future perspectives. Cancers (Basel). 17 (24), 4017. 10.3390/cancers17244017 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kaczmarek, K., Małkiewicz, B. & Lemiński, A. Adequate pelvic lymph node dissection in radical cystectomy in the era of neoadjuvant chemotherapy: A meta-analysis and systematic review. Cancers (Basel). 15 (16), 4040. 10.3390/cancers15164040 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Roviello, G. et al. Neoadjuvant treatment in muscle-invasive bladder cancer: From the beginning to the latest developments. Front. Oncol. 12, 912699. 10.3389/fonc.2022.912699 (2022). (eCollection 2022). [DOI] [PMC free article] [PubMed]
- 4.de Angelis, M. et al. Real-world impact of cisplatin-based neoadjuvant chemotherapy on bladder cancer survival: A 20-year study. World J. Urol.43 (1), 636. 10.1007/s00345-025-05992-9 (2025) (European Association of Urology–Young Academic Urologists Urothelial Carcinoma Working Group (EAU-YAU)). [DOI] [PubMed] [Google Scholar]
- 5.Voskuilen, C. S. et al. Multicenter validation of histopathologic tumor regression grade after neoadjuvant chemotherapy in muscle-invasive bladder carcinoma. Am. J. Surg. Pathol.43 (12), 1600–1610 (2019). [DOI] [PubMed] [Google Scholar]
- 6.Waingankar, N. et al. The impact of pathologic response to neoadjuvant chemotherapy on conditional survival among patients with muscle-invasive bladder cancer. Urol. Oncol. 37(9), 572.e21-572.e28 (2019). [DOI] [PubMed]
- 7.Shan, L., Xu, H., Piao, C., Liu, Z. & Xie, S. Nomogram for the pathological complete response after neoadjuvant chemotherapy in muscle-invasive bladder cancer patients. Ann. Surg. Oncol.32 (1), 589–597. 10.1245/s10434-024-16429-9 (2025) (epub 2024 Oct 28). [DOI] [PubMed] [Google Scholar]
- 8.Piao, C., Liu, D., Liu, Z. & Shan, L. Markers in identifying pathological complete response status in muscle invasive bladder cancer patients who achieved clinical complete response after neoadjuvant chemotherapy. Clin. Genitourin. Cancer. 22 (6), 102211. 10.1016/j.clgc.2024.102211 (2024). Epub 2024 Aug 23. [DOI] [PubMed] [Google Scholar]
- 9.He, W. et al. Neoadjuvant treatment patterns and biomarker selection in muscle-invasive bladder cancer. Discov. Oncol.16 (1), 1197. 10.1007/s12672-025-02796-6 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Schwartz, L. H. et al. RECIST 1.1 – Update and clarification: From the RECIST Committee. Eur. J. Cancer. 62, 132–137. 10.1016/j.ejca.2016.03.081 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Common Terminology Criteria for Adverse Events (CTCAE). Version 5. Published November 27. US Department of Health and Human Services, National Institutes of Health, National Cancer Institute.
- 12.Joffe, B. I. et al. Management of patients with muscle-invasive bladder cancer achieving a clinical complete response after neoadjuvant therapy: Evidence and consideration for active surveillance. Curr. Urol. Rep.26 (1), 36. 10.1007/s11934-025-01264-6 (2025). [DOI] [PubMed] [Google Scholar]
- 13.Saito, R. et al. Japan Urological Oncology Group. Efficacy of cisplatin-based neoadjuvant chemotherapy and risk factors for residual extravesical disease in muscle-invasive bladder cancer: Insights from a nationwide cohort. Int. J. Clin. Oncol.30 (10), 2106–2117. 10.1007/s10147-025-02833-y (2025). Epub 2025 Jul 21. [DOI] [PubMed] [Google Scholar]
- 14.Haugnes, H. S. et al. Real-world data on utilization of neoadjuvant chemotherapy for muscle invasive bladder cancer: Impact on surgical complications and oncological efficacy. Acta Oncol.64, 13–21. 10.2340/1651-226X.2025.42052 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mathieu, R. et al. The prognostic role of lymphovascular invasion in urothelial carcinoma of the bladder. Nat. Rev. Urol.13, 471–479. 10.1038/nrurol.2016.126 (2016). [DOI] [PubMed] [Google Scholar]
- 16.Furubayashi, N. et al. Perioperative chemotherapy in bladder and upper tract urothelial carcinoma: Outcomes by nodal status and lymphovascular invasion. Cancers (Basel). 17 (24), 3986. 10.3390/cancers17243986 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Agrawal, P. et al. Clinicopathologic and survival after cystectomy outcomes in squamous cell carcinoma of the bladder. Clin. Genitourin. Cancer. 21 (6), 631–638e1. 10.1016/j.clgc.2023.05.019 (2023). Epub 2023 May 30. [DOI] [PubMed] [Google Scholar]
- 18.Holmsten, K. et al. Molecular subtypes are associated with clinical benefit in cisplatin-treated metastatic urothelial cancer patients. JCO Precis Oncol.8, e2400209. 10.1200/PO.24.00209 (2024). Epub 2024 Oct 2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Stein, J. P. et al. Radical cystectomy in the treatment of invasive bladder cancer: long-term results in 1,054 patients. J. Clin. Oncol.19 (3), 666–675. 10.1200/JCO.2001.19.3.666 (2001). [DOI] [PubMed] [Google Scholar]
- 20.Powles, T. et al. NIAGARA Investigators. Perioperative Durvalumab with neoadjuvant chemotherapy in operable bladder cancer. N Engl. J. Med.391 (19), 1773–1786. 10.1056/NEJMoa2408154 (2024). Epub 2024 Sep 15. [DOI] [PubMed] [Google Scholar]
- 21.Powles, T. et al. Enfortumab Vedotin and Pembrolizumab in untreated advanced urothelial carcinoma. N Engl. J. Med.390 (10), 875–888 (2024). [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.