Real-world treatment patterns and clinical outcomes in patients with locally advanced or metastatic urothelial carcinoma in Germany: retrospective CONVINCE study

Abstract

Purpose

CONVINCE is a retrospective medical chart review study that examined demographics, treatment patterns, and outcomes in patients who received first-line (1L) treatment for locally advanced or metastatic urothelial carcinoma (la/mUC) in Germany.

Methods

Eligible patients were adults with confirmed la/mUC who received any systemic 1L anticancer treatment between January 1, 2019, and September 30, 2021, outside of a clinical trial. Patients were grouped by type of 1L treatment: platinum-based chemotherapy (PBC), immune checkpoint inhibitor (ICI), or other treatments. Follow-up was ≥ 6 months after end of PBC or start of ICI or other treatments. The primary objective was measurement of real-world progression-free survival (rwPFS).

Results

Data were collected from 188 patients treated at 27 sites (hospitals or office-based practices). First-line treatment was PBC in 76.1% of patients, ICI in 19.1%, and other treatments in 4.8%. The most common PBC regimen was cisplatin + gemcitabine (72.7%), and the most common ICI was atezolizumab (44.4%); 4.2% of PBC-treated patients received avelumab 1L maintenance. In patients who received 1L PBC, ICI treatment, or other treatments, median (95% CI) rwPFS was 10.5 months (9.2–11.6), 12.6 months (8.9–22.9), and not evaluable; median (95% CI) real-world overall survival was 18.1 months (16.5–19.0), 15.9 months (11.1–24.5), and not evaluable; and objective response rates were 56.6%, 60.0%, and 83.3%, including complete response in 14.0%, 20.0%, and 0%, respectively.

Conclusion

PBC was the most common 1L treatment in patients with la/mUC in Germany, consistent with treatment guidelines. Future studies are needed to assess outcomes with newer treatments.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00432-025-06131-y.

Introduction

Bladder cancer is the fifth most common cancer in Germany, with > 30,000 new cases diagnosed in Germany each year (approximately 18,000 cases excluding in situ carcinoma) (Robert Koch Institut 2024). Bladder cancer was responsible for 5.6% of all cancers and 3.4% of all cancer-related deaths in Germany in 2020 (World Health Organization 2023). Urothelial carcinoma (UC) accounts for approximately 90% of bladder cancer cases (Saginala 2020).

Until recently, platinum-based chemotherapy (PBC) was the preferred first-line (1L) treatment option for patients with unresectable locally advanced or metastatic UC (la/mUC) (Cathomas 2022; German Guideline Program in Oncology 2019; Powles 2022). Standard platinum-containing regimens include cisplatin + gemcitabine or dose-dense methotrexate + vinblastine + doxorubicin + cisplatin (ddMVAC) for cisplatin-eligible patients, and carboplatin + gemcitabine for cisplatin-ineligible patients. In 2024, treatment guidelines were updated to recommended enfortumab vedotin plus pembrolizumab as the preferred 1L treatment for la/mUC, irrespective of cisplatin eligibility. Alternative recommended options are 1L PBC followed by avelumab 1L maintenance in patients without disease progression, or 1L nivolumab in combination with cisplatin and gemcitabine for cisplatin-eligible patients only (NCCN 2024; Powles 2024a). Previously, immune checkpoint inhibitor (ICI) monotherapy (pembrolizumab or atezolizumab) was recommended for cisplatin-ineligible patients with PD-L1 + tumors or platinum-ineligible patients and single-agent chemotherapy for platinum-ineligible patients; however, this approach is no longer recommended (NCCN 2024; Powles 2024a). Options for second-line (2L) treatment depend on the classes of agents received as 1L treatment, time to progression, and patient health/performance status (NCCN 2024; Niegisch 2024c; Powles 2024a).

Several studies have shown that a high proportion of patients with la/mUC do not receive any systemic anticancer treatment (Bilen 2023; Geynisman 2022; Kearney 2023; Maráz 2023; Morgans 2023; Tricotel 2024). In a systematic literature review, proportions of patients not receiving systemic treatment in several European studies ranged from 40 to 74% (Kearney 2023). Since the approval of ICIs for the treatment of patients with la/mUC, but before more recent changes in the treatment landscape, several real-world studies have assessed treatment patterns and outcomes in the US (Bilen 2023; Geynisman 2022; Morgans 2023) and Europe (Maráz 2023; Milloy 2024; Niegisch 2024a; Puente 2023; Tricotel 2024). In general, these studies found that PBC was the most common 1L systemic treatment, patients receiving 1L PBC had longer overall survival (OS) than those receiving 1L ICI treatment, and outcomes were worse in cisplatin-ineligible vs cisplatin-eligible patients. In a real-world study conducted in Germany, which examined patients with la/mUC who received any 1L treatment between 2009 and 2016 (before the availability of ICIs), 77% received 1L platinum-based combination chemotherapy, including cisplatin + gemcitabine in 83%, and 29% received 2L treatment (Niegisch 2018). Median OS with 1L treatment was 16.1 months (95% CI 13.7–19.2) overall and 17.7 months (95% CI 14.4–24.2) with cisplatin + gemcitabine. In Germany, patients with la/mUC are usually treated by medical oncologists or urologists in hospitals or office-based community practices; however, data on treatment patterns according to medical discipline or setting are limited.

Here we report findings from CONVINCE (retrospeCtive ObservatioNal study in patients with locally adVanced or metastatIc urothelial caNCEr receiving 1L systemic anticancer treatment under real world conditions in Germany), a medical chart review study of patients with la/mUC who received systemic 1L treatment in hospitals or office-based community practices in Germany between 2019 and 2021, after the approval of 1L ICIs but before avelumab 1L maintenance treatment was widely available and more recent developments in the treatment landscape (Merck 2021; Powles 2024a). We describe the demographics, treatment patterns, and clinical outcomes in patients with la/mUC treated in routine clinical practice.

Methods

Study design

CONVINCE is a retrospective, noninterventional, multicenter, medical chart review study conducted in Germany. Eligible patients were aged ≥ 18 years, had histologically or cytologically confirmed la/mUC (International Classification of Diseases, Tenth Revision, code C67.1), and had received ≥ 1 systemic 1L treatment as part of routine care between January 1, 2019, and September 30, 2021. Patients who received treatment as part of a clinical trial within 30 days before the start of 1L treatment were excluded. Invitations to participate in the study were sent to a wide range of hospitals and office-based practices across Germany, and those who accepted the invitation were contracted.

Patients were grouped based on the type of 1L treatment received: (1) PBC (combination regimen or monotherapy), with end of treatment between January 1, 2019, and September 30, 2021; (2) ICIs, with start of treatment between January 1, 2019, and September 30, 2021; and (3) other treatments, with start of treatment between January 1, 2019, and September 30, 2021. A follow-up period of  ≥ 6 months after completion of PBC (group 1) or start of treatment (groups 2 and 3) was required. Anonymized data collection and analysis were performed between February 3, 2022, and February 28, 2023.

Study objectives

The primary objective was measurement of real-world progression-free survival (rwPFS) with 1L treatment. Secondary objectives included measurement of real-world OS (rwOS), objective response, and duration of response (DOR) with 1L treatment; description of prior treatments and subsequent treatments (after completion of or progression during or following 1L treatment); and description of reasons for receiving nonplatinum-based 1L treatment.

Outcomes

The index date for outcomes was the date of 1L treatment initiation. Duration of rwPFS was measured from the index date until disease progression or death due to any cause. Duration of rwOS was measured from the index date until the date of death. Best overall response was defined as the best response recorded from the start of treatment until disease progression or recurrence, based on the investigator’s assessment. The objective response rate was calculated as the proportion of patients who had a best overall response or complete or partial response. DOR was defined as the time from documentation of tumor response (complete or partial) to disease progression.

Safety analyses

Safety analyses were not part of this study; however, adverse events (AEs) of grade ≥ 3 (per National Cancer Institute Common Terminology Criteria for Adverse Events version 5.0 [US Department of Health and Human Services 2017]) that resulted in dose modification or treatment discontinuation were collected.

Statistical analyses

Patient characteristics, treatment patterns, and best response were analyzed descriptively. Because of the 1-point retrospective documentation and full anonymization, with no option for longitudinal follow-up, median rwPFS and rwOS were calculated as arithmetic medians (not via Kaplan–Meier analyses) from the date of first anticancer drug administration to the date of progression or death; if none of these events occurred, the last contact before documentation was used. Thus, median values for rwPFS and rwOS represent minimum median values. Confidence intervals (CIs) for median values were obtained using the Brookmeyer-Crowley method, with 95% CIs generated using the log(-log) method (Brookmeyer 1982). Data shown in Kaplan–Meier curves are descriptive because the number of events at the timepoint of documentation was too small, and longer follow-up was not possible due to anonymization.

Results

Patients

Between January 1, 2019, and September 30, 2021, a total of 188 patients who received 1L systemic treatment for la/mUC at 27 sites (8 hospitals and 19 practices) were identified. At the time of analysis, 59 patients (31.4%) were still alive, 79 (42.0%) had died, and 50 (26.6%) had unknown survival status. First-line treatment was PBC in 143 patients (76.1%), ICI in 36 (19.1%), and other treatments in 9 (4.8%). Follow-up ranged from 6.0 months to 42.8 months (95% CI 17.5–19.8) and was < 12 months in 47 patients (25.0%) and ≥ 12 months in 141 (75.0%).

Patient characteristics by treatment group and in the total population are summarized in Table 1. Compared with the PBC group, patients in the ICI group had a higher median age at the start of treatment (77 vs 69 years) and a higher proportion of patients with Eastern Cooperative Oncology Group performance status of ≥ 2 (13.9% vs 4.9%), PD-L1 + tumors (92.6% vs 61.8%), and prior neoadjuvant or adjuvant treatment (30.6% vs 7.7%); patient numbers in the other treatments group were too small for meaningful comparisons. In evaluable patients (n = 187), 126 (67.4%) had stage IV disease at the start of 1L treatment (Supplementary Table 1).

Table 1.

Patient characteristics by 1L treatment and in the total population

	PBC group (n = 143)	ICI group (n = 36)	Other treatments group (n = 9)	Total (N = 188)
Sex, n (%)
Female	39 (27.3)	12 (33.3)	1 (11.1)	52 (27.7)
Male	104 (72.7)	24 (66.7)	8 (88.9)	136 (72.3)
Age, median (range), years
At first diagnosis	68 (48–84)	74.5 (39–87)	70 (68–79)	69 (39–87)
At start of 1L treatment	69 (48–84)	77 (41–87)	70 (68–80)	70 (41–87)
ECOG PS, n (%)
0	74 (51.7)	19 (52.8)	2 (22.2)	95 (50.5)
1	62 (43.4)	12 (33.3)	5 (55.6)	79 (42.0)
≥ 2	7 (4.9)	5 (13.9)	1 (11.1)	13 (6.9)
Not available	0	0	1 (11.1)	1 (0.5)
Prior neoadjuvant or adjuvant treatment, n (%)	11 (7.7)	11 (30.6)	3 (33.3)	25 (13.3)

Smoking history, n (%)	n = 72	n = 16	n = 3	n = 91
Active smoker	26 (36.1)	5 (31.3)	0	31 (34.1)
Former smoker	30 (41.7)	6 (37.5)	1 (33.3)	37 (40.7)
Never smoker	16 (22.2)	5 (31.3)	2 (66.7)	23 (25.3)

PD-L1 status, n (%)	n = 68	n = 27	n = 3	n = 98
Negative	26 (38.2)	2 (7.4)	2 (66.7)	30 (30.6)
Positive	42 (61.8)	25 (92.6)	1 (33.3)	68 (69.4)

Tumor grade, n (%)	n = 110	n = 25	n = 2	n = 137
1	3 (2.7)	0	0	3 (2.2)
2	25 (22.7)	6 (24.0)	1 (50.0)	32 (23.4)
3	82 (74.5)	19 (76.0)	1 (50.0)	102 (74.5)

Tumor status at first diagnosis, n (%)	n = 119	n = 26	n = 5	n = 150
Nonmuscle invasive	39 (32.8)	8 (30.8)	2 (40.0)	49 (32.7)
Muscle invasive	80 (67.2)	18 (69.2)	3 (60.0)	101 (67.3)

1L first-line, ECOG PS Eastern Cooperative Oncology Group performance status, ICI, immune checkpoint inhibitor, PBC platinum-based chemotherapy

Treatment setting

The total population (N = 188) included patients treated in 13 of 17 regions in Germany (Fig. 1A). The treatment setting was office-based community hematology-oncology or urology practices for 110 patients (58.5%) and hospitals for 78 (41.4%) (Fig. 1B); 113 patients (60.1%) were treated by hematologist-oncologists and 75 (39.9%) were treated by urologists (Fig. 1C). The proportions of patients treated in each setting were similar between the PBC and ICI groups.

Fig. 1.

Distribution of patients receiving 1L treatment according to A region, B treatment setting, and C medical specialty of the treating physician. ICI immune checkpoint inhibitor; PBC platinum-based chemotherapy

Treatment patterns

Overall, 1L treatment comprised a platinum agent + gemcitabine in 126 of 188 patients (67.0%; Table 2). In the PBC group (n = 143), 104 patients (72.7%) received cisplatin + gemcitabine, and 22 (15.4%) received carboplatin + gemcitabine; 6 patients (4.2%) received avelumab 1L maintenance, and all had received cisplatin + gemcitabine. In the ICI group (n = 36), 16 patients (44.4%) received atezolizumab, 15 (41.7%) received pembrolizumab, and 5 (13.9%) received nivolumab. In the other treatments group (n = 9), 6 patients (66.7%) received gemcitabine monotherapy, 2 (22.2%) received vinflunine, and 1 (11.1%) received mitomycin. The most common reported reasons for not administering 1L PBC in the ICI and other treatments groups combined (n = 45) were physician’s decision (28.9%), impaired renal function (20.0%), preexisting disease (15.6%), and advanced age (11.1%) (Supplementary Table 2).

Table 2.

Summary of 1L treatments received by treatment group and in the total population

Patients, n (%)	PBC group (n = 143)	ICI group (n = 36)	Other treatments group (n = 9)	Total (N = 188)
Cisplatin + gemcitabine	104 (72.7)	0	0	104 (55.3)
Followed by avelumab 1L maintenance	6 (4.2)	0	0	6 (3.2)
Carboplatin + gemcitabine	22 (15.4)	0	0	22 (11.7)
Atezolizumab	0	16 (44.4)	0	16 (8.5)
Pembrolizumab	0	15 (41.7)	0	15 (8.0)
Gemcitabine	0	0	6 (66.7)	6 (3.2)
Cisplatin	5 (3.5)	0	0	5 (2.7)
Nivolumab	0	5 (13.9)	0	5 (2.7)
MVAC	4 (2.8)	0	0	4 (2.1)
Cisplatin + etoposide	4 (2.8)	0	0	4 (2.1)
Carboplatin + etoposide	2 (1.4)	0	0	2 (1.1)
Carboplatin + paclitaxel	2 (1.4)	0	0	2 (1.1)
Vinflunine	0	0	2 (22.2)	2 (1.1)
Mitomycin	0	0	1 (11.1)	1 (0.5)

1L first-line, ICI immune checkpoint inhibitor, MVAC methotrexate + vinblastine + doxorubicin + cisplatin, PBC platinum-based chemotherapy

The median number of treatment cycles received was 4.0 (range 1–7) in the PBC group (3-week cycles) and 13.0 (range 2–42) in the ICI group (21-day cycles with pembrolizumab and atezolizumab and 14-day cycles with nivolumab). A dose reduction or delay was required by 36 patients (25.2%) in the PBC group, and a dose interruption was required by 1 patient (2.8%) in the ICI group (Supplementary Table 3). Reasons for end of treatment are shown in Supplementary Table 4.

Subsequent treatments are shown in Table 3. In the total population (N = 188), 96 patients (51.1%) received 2L treatment and 64 (34.0%) received third- or later-line treatment. The most common 2L treatments following 1L PBC were pembrolizumab (35 of 76 patients [46.1%]) and nivolumab (18 of 76 [23.7%]). The most common 2L treatments following 1L ICI treatment were cisplatin + gemcitabine (6 of 14 patients [42.9%]) and carboplatin + gemcitabine (3 of 14 [21.4%]).

Table 3.

Distribution of subsequent treatments in the PBC, ICI, and other treatments groups

	PBC group (n = 143)		ICI group (n = 36)		Other treatments group (n = 9)
Line of treatment	2L	≥ 3L	2L	≥ 3L	2L	≥ 3L
Patients who received subsequent treatment, n (%)	76 (53.1)	56 (39.2)	14 (38.9)	4 (11.1)	6 (66.7)	3 (33.3)
Cisplatin + gemcitabine	7 (9.2)	3 (5.4)	6 (42.9)	1 (25.0)	0	0
Carboplatin + gemcitabine	1 (1.3)	1 (1.8)	3 (21.4)	1 (25.0)	2 (33.3)	0
MVAC	0	3 (5.4)	0	1 (25.0)	0	0
Carboplatin + paclitaxel	0	5 (8.9)	0	0	0	0
Atezolizumab	9 (11.8)	3 (5.4)	0	0	1 (16.7)	0
Pembrolizumab	35 (46.1)	7 (12.5)	0	0	3 (50.0)	1 (33.3)
Nivolumab	18 (23.7)	6 (10.7)	0	0	0	0
Cisplatin	1 (1.3)	3 (5.4)	0	0	0	0
Gemcitabine	0	2 (3.6)	0	0	0	0
Vinflunine	1 (1.3)	11 (19.6)	1 (7.1)	0	0	0
Mitomycin	0	2 (3.6)	0	0	0	0
Docetaxel	0	3 (5.4)	2 (14.3)	0	0	0
Enfortumab vedotin	0	2 (3.6)	1 (7.1)	1 (25.0)	0	0
Paclitaxel	0	3 (5.4)	0	0	0	2 (66.7)
Carboplatin	0	1 (1.8)	0	0	0	0
Agent not reported	4 (5.3)	1 (1.8)	1 (7.1)	0	0	0

2L second line, 3L third line, ICI immune checkpoint inhibitor, MVAC, methotrexate + vinblastine + doxorubicin + cisplatin, PBC platinum-based chemotherapy

Clinical outcomes

Median rwPFS (95% CI) was 10.1 months (9.2–11.6) in the total population, 10.5 months (9.2–11.6) in the PBC group, 12.6 months (8.9–22.9) in the ICI group, and not evaluable in the other treatments group, and 12-month rwPFS rates (95% CI) were 63.8% (54.9%−70.6%), 60.1% (50.7%−68.5%), 77.8% (59.2%−90.1%), and 88.9% (33.4%−97.9%), respectively (Fig. 2A). Median rwOS (95% CI) was 18.0 months (16.3–18.7) in the total population, 18.1 months (16.5–19.0) in the PBC group, 15.9 months (11.1–24.5) in the ICI group, and not evaluable in the other treatments group, and 12-month rwOS rates (95% CI) were 87.8% (82.0%−92.0%), 89.5% (83.4%−93.8%), 80.6% (59.2%−90.1%), and 88.9% (27.3%−97.4%), respectively (Fig. 2B). The objective response rate in evaluable patients in the total population (n = 172) was 58.1%, and the disease control rate (complete response, partial response, or stable disease) was 85.5% (Table 4). Objective response rates in the PBC, ICI, and other treatments groups were 56.6%, 60.0%, and 83.3%, including complete response in 14.0%, 20.0%, and 0%, respectively. DOR was evaluable in 32 patients, of whom 30 patients were in the 1L PBC group; median DOR in the evaluable 1L PBC population was 9.7 months (range 3.5–22.6).

Fig. 2.

Real-world A PFS and B OS in in the PBC, ICI, and other treatments groups. 1L first line, ICI immune checkpoint inhibitor, NE not evaluable, PBC platinum-based chemotherapy, rwOS real-world overall survival, rwPFS real-world progression-free survival. ^aCalculated from date of first drug administration to date of tumor progression or last contact. ^bCalculated from date of first drug administration to date of death or last contact

Table 4.

Summary of responses in evaluable patients by 1L treatment and in the total population

	PBC group (n = 136)	ICI group (n = 30)	Other treatments group (n = 6)	Total (n = 172)
Best overall response, n (%)
CR	19 (14.0)	6 (20.0)	0	25 (14.5)
PR	58 (42.6)	12 (40.0)	5 (83.3)	75 (43.6)
SD	41 (30.1)	5 (16.7)	1 (16.7)	47 (27.3)
PD	18 (13.2)	7 (23.3)	0	25 (14.5)
ORR, n (%) [95% CI]	77 (56.6) [47.0–66.2]a	18 (60.0) [44.3–77.7]	5 (83.3) [70.3–96.4]	100 (58.1)
Disease control rate, n (%)b	118 (86.8)	23 (76.7)	6 (100)	147 (85.5)

1L first-line, CR complete response, ICI immune checkpoint inhibitor, ORR objective response rate, PBC platinum-based chemotherapy, PD progressive disease, PR partial response, SD stable disease

^aBased on ORRs for the different PBC regimens administered

^bCR, PR, or SD

Patients in the PBC group had similar outcomes irrespective of treatment setting or whether they were treated by a hematologist-oncologist or urologist (Supplementary Tables 5 and 6).

Safety

Grade ≥ 3 AEs are summarized in Supplementary Table 7. In the PBC group (n = 143), 49 patients (34.3%) had a grade ≥ 3 AE; the most common grade ≥ 3 AEs were leukopenia (4.9%), neutropenia (4.2%), and thrombocytopenia (4.2%). In the ICI group, 1 of 36 patients (2.8%) had a grade ≥ 3 AE (deterioration of performance status); no patients in the other treatments group had a grade ≥ 3 AE.

Discussion

The CONVINCE study provides valuable insights about 1L systemic treatment in patients with la/mUC in Germany following the introduction of ICIs. Consistent with treatment guidelines from the time period (Cathomas 2022; German Guideline Program in Oncology 2019; Powles 2022), PBC was administered as 1L systemic treatment in most patients (76.1%), with approximately half of all patients receiving cisplatin + gemcitabine. Baseline characteristics in the study population were consistent with the known epidemiology of la/mUC and previous findings from real-world studies, including studies in Germany using statutory health insurance databases (Niegisch 2024a; b). Overall, 67.4% of evaluable patients had stage IV disease at the start of 1L treatment, similar to other real-world studies of la/mUC populations (Geynisman 2024; Mahmoudpour 2024) but lower than in phase 3 populations (86–95%; Galsky et al. 2020; Powles et al. 2021; Powles et al. 2024b; van der Heijden et al. 2023). A notable proportion of patients received 1L ICI treatment (19.1%); however, recently updated guidelines do not recommend 1L ICI monotherapy for routine clinical care (Powles 2024a). The group treated with 1L ICIs was older, had a worse performance status than the group treated with 1L PBC, and included a notably higher proportion of patients who had received prior neoadjuvant or adjuvant treatment. The most common stated reasons for not administering PBC, other than physician’s decision, were impaired renal function and preexisting disease.

The median number of PBC treatment cycles received was 4.0 (range 1–7), consistent with treatment guidelines that recommend 4 to 6 cycles (Cathomas 2022; Powles 2022). The median number of ICI treatment cycles received was 13.0 cycles (range 2–42), consistent with recommendations to administer ICI treatment for 2 years for pembrolizumab or until disease progression for atezolizumab (Powles 2022). High attrition rates were observed across lines of treatment. Approximately half of all patients received 2L treatment and one-third received third-line treatment, with a greater proportion receiving subsequent treatment after 1L PBC or other treatments compared with after 1L ICI treatment. The most common 2L treatment following 1L PBC was ICIs, most often with pembrolizumab, also consistent with treatment guidelines (German Guideline Program in Oncology 2019; Cathomas 2022; Powles 2022). The most common 2L treatment following 1L ICI treatment was PBC.

Clinical outcomes were generally similar between patients treated with 1L PBC or ICIs, although direct comparisons between groups are not appropriate because of their different characteristics and bias associated with selecting patients for different treatment options. Previous real-world studies that reported multivariable analyses of outcomes that were adjusted based on population characteristics concluded that PFS and OS were longer in patients treated with 1L PBC vs ICI treatment (Geynisman 2022; Morgans 2023). Across all treatment groups in this study, the highest frequency of grade ≥ 3 AEs was observed in the PBC group.

To our knowledge, this is one of the first studies to assess treatment patterns and outcomes in Germany based on treatment setting, including office-based community treatment vs hospital-based treatment and treatment by hematologist/oncologist vs urologist specialty. The participating study sites spanned most of Germany. Overall, no major differences were observed in treatment patterns or outcomes in patients treated in different settings.

The treatment landscape in la/mUC is evolving rapidly and several novel options have become available in recent years. In this study population, only 6 patients received avelumab 1L maintenance, which was expected because regulatory approval in Germany only occurred in January 2021 (Merck 2021), shortly before the end of the study period in September 2021; thus, no meaningful data about avelumab 1L maintenance could be collected. Since this study was conducted, enfortumab vedotin monotherapy has been approved in Europe and the US for the treatment of patients with la/mUC who have received prior PBC and ICI treatment (Astellas 2023, 2024a). Additionally, enfortumab vedotin in combination with pembrolizumab has been approved recently for 1L treatment of patients with la/mUC, and it is now recommended in international guidelines as the preferred treatment, irrespective of cisplatin eligibility (Astellas 2024b; US Food and Drug Administration 2023; NCCN 2024; Powles 2024a). In addition, 1L nivolumab in combination with cisplatin + gemcitabine has been approved in the US and Europe and is a further treatment option for cisplatin-eligible patients (Bristol Myers Squibb 2024a, 2024b, NCCN 2024; Powles 2024a). The effects of these novel treatment options on real-world patient outcomes and economic healthcare burden in la/mUC warrant evaluation in future studies.

This study has several limitations. As a retrospective study, available data are limited to what was recorded in patient medical records as part of routine clinical care. Although measures were taken to ensure that information obtained was as complete and accurate as possible, the potential for bias from missing data cannot be excluded. Data collection relied on completion of electronic case report forms, based on information recorded by treating physicians and medical staff. Bias may have been introduced during the selection of patient records, and errors may have been introduced during the completion of electronic case report forms. Practice patterns described in this study may not be generalizable outside of Germany. Analyses of outcomes were not adjusted for variability in patient characteristics, including demographics or disease characteristics, which may have influenced outcomes in treatment groups. Analyses by treatment setting or specialty did not consider factors that may have influenced where patients were treated (selection bias), which may have influenced outcomes. Because tumor assessments in clinical practice may be less stringent than in clinical trials, analyses of ORR and rwPFS may represent overestimates. Additionally, data for rwOS should be interpreted with caution because of immortal time bias being introduced by the 6-month follow-up period specified as an inclusion criterion (Suissa 2007). Detailed safety data were not collected. Lastly, analyses conducted in several groups and subgroups were limited by small patient numbers.

Conclusion

Overall, findings from the CONVINCE study show that PBC was the most common 1L treatment in patients with la/mUC in Germany irrespective of treatment setting or type of specialist, consistent with treatment guidelines and available treatment options at the time of the study. Approximately 1 in 5 patients received 1L ICI monotherapy. Further studies are needed to assess treatment outcomes, including safety, following the introduction of newer treatment options.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization: U.O., S.G., M.K., R.L. Data curation: A.E., R.L. Formal analysis: A.E., M.K., R.L. Investigation: K.S., S.M., T.K., M.R., C.S., S.S. Methodology: U.O., S.G., M.K., R.L. Supervision: M.K., R.L. Writing—review & editing: All authors.

Funding

This study was sponsored by Merck (CrossRef Funder ID: 10.13039/100009945).

Data availability

Any requests for data by qualified scientific and medical researchers for legitimate research purposes will be subject to Merck’s Data Sharing Policy. All requests should be submitted in writing to Merck’s data sharing portal (https://www.merckgroup.com/en/research/our-approach-to-research-and-development/healthcare/clinical-trials/commitment-responsible-data-sharing.html). When Merck has a co-research, co-development, co-marketing or co-promotion agreement, or when the product has been out-licensed, the responsibility for disclosure might be dependent on the agreement between parties. Under these circumstances, Merck will endeavor to gain agreement to share data in response to requests.

Declarations

Conflict of interest

K.S. has served in speakers bureaus and consulting or advisory roles for AAA, Amgen, Apogepha, Astellas, AstraZeneca, Bayer, Bristol Myers Squibb (personal and institutional), Eisai, EUSA-Pharma, Fosanis, Ipsen, Janssen Cilag, Merck, MSD, Novartis (personal and institutional), Pfizer (personal and institutional), and Roche, and has received travel, accommodations, and expenses from Astellas, AstraZeneca, Bayer, Janssen Cilag, Ipsen, Merck, and Sanofi-Aventis. S.M. has served in consulting or advisory roles for Amgen, AstraZeneca, Bayer, BD, Bebig, Janssen-Cilag, MSD, Pfizer, and Sanofi-Aventis, and has participated in speakers bureaus for Amgen, Apogepha, Astellas, AstraZeneca, Bayer, BD, Bebig, Ipsen, Janssen-Cilag, Merck, MSD, Novartis, Pfizer, Sanofi-Aventis, and Takeda. T.K. has served in speakers bureaus and consulting or advisory roles for AstraZeneca, BMS, Boehringer-Ingelheim, GSK, Janssen, MSD, Roche, and Takeda. M.R., C.S., and A.E. have nothing to disclose. U.O. reports employment with Merck Healthcare Germany GmbH, Weiterstadt, Germany, an affiliate of Merck KGaA. S.G. reports employment with Merck Healthcare KGaA, Darmstadt, Germany when the study was conducted. M.K. reports employment with Merck Healthcare KGaA, Darmstadt, Germany and reports stock and other ownership interests with Merck, Novartis, and UCB. R.L. has served in consulting or advisory roles for AstraZeneca, Gilead, Janssen-Cilag, Merck, Omnicare, Sanofi, and Takeda; has served in a speakers bureau for Merck; and has provided project support for AbbVie, AstraZeneca, Bristol Myers Squibb, Gilead, Janssen-Cilag, Merck, and Pfizer. S.S. has nothing to disclose.

Ethical approval

The study protocol was submitted to the independent ethical committee of the Chamber of Physicians in Northern Rhine, which confirmed that because data were fully anonymized, patient informed consent was not required. The protocol was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice Guidelines.