ABSTRACT
Aim
Describe real-world epidemiology, treatment patterns, health care resource utilization, and costs of locally advanced or metastatic urothelial carcinoma (la/mUC) in France.
Patients & methods
Retrospective study including all adults with la/mUC diagnosis during January 2017 to December 2020 in the PMSI database.
Results
Annual prevalence and incidence ranged from 36.4 to 38.9 and 16.4 to 18.5 cases per 100,000 people, respectively. Of the 25,314 patients with incident la/mUC, 37.6% did not receive first-line systemic treatment. Of the 14,656 patients who started first-line systemic treatment, 66.6%, 22.5%, and 10.9% received 1, 2, and 3 lines of therapy, respectively. Annual per-patient costs in second-/third-line setting ranged from €8803 to €16,012.
Conclusion
The substantial disease burden of la/mUC in France highlights the unmet need for new therapies.
KEYWORDS: Incidence, prevalence, costs, urothelial carcinoma, France, treatment
Plain Language Summary
What is this article about?
Urothelial carcinoma (UC) is a type of cancer affecting the urinary system. It can spread to other parts of the body, described as locally advanced or metastatic (la/m). We used information from a French database recording hospitalizations in France to find out how many people have la/mUC, how many new cases develop each year, what treatments they receive, how many die in the hospital, and how much their care costs.
What were the results?
Based on database information, 37 to 39 of every 100,000 people have la/mUC and 17 to 19 of every 100,000 people are identified with a new case yearly. Slightly more than one-third of patients with la/mUC did not receive recommended treatment (chemotherapy) when first diagnosed. Chemotherapy was the most common treatment type for the first, second, or third treatment; checkpoint inhibitors (a unique treatment) became more commonly used as a second treatment over time. Yearly in-hospital death rates were high, ranging from 47.8% of patients who died within 1 year from diagnosis to 62.9% dying within 3 years. Yearly cost of care was high (costing €8803 to €16,012) in patients starting a second or third treatment.
What do the results of the study mean?
The study shows many patients may not be fit enough or choose not to receive treatment. Even those receiving treatment are at high risk for poor outcomes. The burden of la/mUC in France is high, underscoring the need for more therapies and better supportive care early in disease management.
GRAPHICAL ABSTRACT
1. Introduction
In 2020, bladder cancer comprised 4% of cancer diagnoses in France and was the fifth most commonly diagnosed cancer, with approximately 16,490 new cases [1]. Urothelial carcinoma (UC) accounts for more than 90% of bladder cancer cases [2,3].
Cisplatin-based chemotherapy is the standard first-line treatment for locally advanced or metastatic UC (mUC) in France per the 2022 to 2024 Comité de cancérologie de l’Association Française d’Urologie guidelines [4]. However, approximately 50% of patients with locally advanced or mUC are ineligible for cisplatin-based chemotherapy [3,5,6]. In this case, carboplatin-based chemotherapy is the recommended first-line treatment in patients ineligible for cisplatin [4]. Maintenance immunotherapy with avelumab is recommended for patients without disease progression on first-line platinum-based chemotherapy [4]. Monotherapy with pembrolizumab immunotherapy is the recommended second-line treatment for patients who progress during or after first-line platinum-based chemotherapy [4]. Enfortumab vedotin is available (early access program) for patients who relapse after previous treatments, including platinum-based chemotherapy and immunotherapy [4].
Based on published clinical trial data, the median overall survival is 13.2 months for patients eligible for cisplatin and 9.7 to 12.0 months for patients not eligible for cisplatin treatment, although these data may not be reflective of the use of maintenance therapy in the first-line setting [7]. In a recent real-world analysis, median overall survival in patients treated in the first-line setting was also lower in those ineligible for cisplatin (8.6 months) compared with those eligible for cisplatin (14.4 months) [8]. Real-world evidence provides information about the use, benefits, and risks of therapies in more clinically diverse settings and patients, such as older patients and those with more comorbidities [9].
Little is known about the real-world epidemiology, treatment patterns, health care resource utilization (HCRU), and health care costs in patients with locally advanced or mUC in France, particularly after first-line systemic treatment. As treatment strategies evolve very quickly, there is a need to understand the burden and treatment pathways of locally advanced or mUC in France. The objectives of this study were to describe the annual incidence and prevalence, patient characteristics, treatment patterns, in-hospital deaths, and HCRU and costs of locally advanced or mUC in France from 2017 to 2020.
2. Methods
2.1. Study design
This was a national, retrospective, longitudinal, noninterventional study of patients with locally advanced or mUC using data extracted from the Programme de Médicalisation des Systèmes d’Information (PMSI), the French national hospital discharge database, from 1 January 2013, to 31 December 2020 (study period; Figure 1). The inclusion period was from January 2017 to December 2020. Depending on the study population, 2 different index dates were defined: either the first occurrence of locally advanced or mUC during the inclusion period (for populations 1 and 2) or the start date of second- or third-line treatment (for populations 3 and 4). A historical period (from 1 January 2013, to index date) allowed characterization of patient baseline characteristics at index date. Patients were followed from the index date and censored at in-hospital death or at study end date, whichever came first. For longitudinal analyses (treatment patterns, in-hospital deaths, HCRU, and costs), only patients with an index date before 31 December 2019, were included to allow for at least 1 year of follow-up (Supplemental Figure S1).
Figure 1.
Study design. aFor population 1, which includes patients with la/mUC diagnosed before and during the inclusion period, the index date was defined depending on patient status at the start of the inclusion period. For prevalent patients (ie, those diagnosed before 2017), the index date was 1 January 2017. For incident patients (ie, those newly diagnosed on or after 1 January 2017), the index date was the date of first la/mUC diagnosis.
la/mUC: locally advanced or metastatic urothelial carcinoma.
2.2. Data source/setting
The PMSI is a national medicoadministrative discharge database that covers the French population and is managed by the Agence Technique de l’Information sur l’Hospitalisation [10]. Approximately 20 million to 25 million anonymous hospital discharge summaries are collected each year. Patients are pseudonymously linked across all health care facilities, and all diagnostic and treatment pathways can be tracked.
Each record contains data about the patient’s demographics, management, and hospital stay [10]. Management data include information about pathologies, procedures, and treatment. Data regarding the patient’s hospital stay include information about diagnosis-related groups (DRGs), length of stay, type of admission/discharge, specific units attended, and date of in-hospital death. The reasons for hospitalization are coded as principal diagnoses (condition for which the patient was hospitalized) and, optionally, related diagnoses (any underlying condition that may be related to the primary reason for hospitalization) or significant associated diagnoses (comorbidities or complications that may affect the course or cost of hospitalization) using the International Classification of Diseases, 10th Revision (ICD-10) [10,11]. Procedures are coded using the Classification Commune des Actes Médicaux (CCAM) [10]. Hospital health care consumption corresponds to hospitalizations in acute-stay (médecine, chirurgie, obstétrique [MCO]) institutions, after-care, and rehabilitation (soins de suite et de réadaptation [SSR]) clinics, and at-home hospital (hospitalization à domicile [HAD]) units for public and private sectors [12]. Data regarding drugs received on top of DRGs (liste en sus), drugs in early access, medical devices, and biological tests are also available in the PMSI but were not used in this analysis.
2.3. Study population
The study population included all patients in the PMSI database with a diagnosis of locally advanced or mUC during the inclusion period. Patients who had an initial diagnosis of locally advanced or mUC from 1 January 2013, to 31 December 2016 could be included, provided they had at least 1 hospital stay during the inclusion period to ensure the patient was alive and not lost to follow-up. Patients younger than 18 years of age at initial diagnosis of locally advanced or mUC and those residing in French overseas regions or abroad were excluded.
There is no ICD-10 code for locally advanced or mUC, so patients with locally advanced or mUC were identified using an algorithm. We identified patients with a diagnosis of UC (ICD-10 codes C65–68, D090–091 as a principal, related, or significant associated diagnosis) during the study period (1 January 2013–31 December 2020). Patients with locally advanced or mUC were identified as those with (1) a diagnosis of metastasis (ICD-10 codes C77–79) concomitant with a diagnosis of UC, and/or (2) a chemotherapy session or administration of a specific advanced-stage treatment received in addition to the DRG (pembrolizumab, avelumab, atezolizumab, and nivolumab) concomitantly with a diagnosis of UC (Supplemental Table S1). Patients with UC who underwent radical cystectomy were considered to have locally advanced or mUC only if the administration of chemotherapy occurred 4 or more months after the surgery. The date of the first diagnosis of locally advanced or mUC during the study period was the date of the first diagnosis of metastasis or the first treatment administration, whichever occurred first.
Four cohorts were evaluated. Population 1 included all patients with locally advanced or mUC diagnosed before and during the inclusion period (i.e., prevalent and incident locally advanced or mUC). Population 2 consisted of all adults newly diagnosed with locally advanced or mUC during the inclusion period (i.e., patients with incident locally advanced or mUC only). Population 3 included all patients with prevalent or incident locally advanced or mUC who were starting second- or third-line treatment during the inclusion period. Population 4 was derived from population 3 and included all patients with locally advanced or mUC who were previously treated with first-line chemotherapy and were starting second- or third-line treatment with a checkpoint inhibitor during the inclusion period.
Identification of populations 3 and 4 required treatment sequences to be reconstructed for all patients identified in population 1. In the absence of relapse data and identifiable standard chemotherapy drugs in the PMSI database, an algorithm was developed from the PMSI MCO database based on dates of chemotherapy and checkpoint inhibitor therapy (Supplemental Figure S2). First-line systemic treatment was defined as the first treatment regimen, either chemotherapy or checkpoint inhibitor, beginning less than 3 months after the initial locally advanced or mUC diagnosis. If the first systemic treatment regimen began more than 3 months after the diagnosis, the treatment line was considered to be indeterminate. Patients for whom no systemic treatment (chemotherapy or checkpoint inhibitor therapy) was identified were assigned to the group of “untreated” patients (i.e., those who did not receive first-line systemic treatment). Identification of the switch from first- to second-line treatment depended on whether first-line treatment could be identified in the PMSI database. The switch from first-line checkpoint inhibitor therapy to second-line treatment was identified using the first occurrence of switch to another identifiable drug (namely checkpoint inhibitor therapy, excluding avelumab), switch to a nonidentifiable drug (i.e., standard chemotherapy), or an interval of more than 2 months between 2 treatment sessions with the same checkpoint inhibitor. The switch from unidentifiable first-line treatment (i.e., first-line standard chemotherapy) to second-line treatment was identified using the first of switch to an identifiable drug (namely checkpoint inhibitor therapy, excluding avelumab) or an interval of more than 2 months between 2 chemotherapy sessions (i.e., switch to second-line standard chemotherapy).
2.4. Study outcomes
The demographics and clinical characteristics of patients with locally advanced or mUC were examined. Patient medical data included history of UC management, comorbidities, and age-adjusted Charlson Comorbidity Index (CCI), all of which were identified during the historical period using procedure codes and/or diagnoses or drugs prescribed in addition to the DRG (Supplemental Table S2). Comorbidities included other cancers (ICD-10 codes C00–D48, except malignant neoplasms of the urinary tract), with a focus on lung (ICD-10 codes C33, C34, D021, D022) and prostate (ICD-10 codes C61, D075) cancers, as well as comorbidities used to compute age-adjusted CCI (Supplemental Table S3). Adjusted CCI was estimated per an algorithm adapted for data from the Système Nationale des Données de Santé [13]. Because available diagnostic codes could not distinguish patients with locally advanced UC from those with mUC, the age-adjusted CCI was systematically set to a minimum of 6 (equivalent to a metastatic solid tumor in the malignancy variable).
Epidemiology outcomes were incidence and prevalence of locally advanced or mUC. The incidence outcomes included the number of patients newly diagnosed with locally advanced or mUC each year in France and the incidence per 100,000 people. The prevalence outcomes included the number of patients reported with a diagnosis of locally advanced or mUC at any time during the year and the prevalence per 100,000 people. Patients with incident locally advanced or mUC were distinguished from patients with prevalent locally advanced or mUC using the historical period.
Treatment patterns included time to start of first-line systemic therapy, lines of therapy received, type of treatment, time of each line, reason for discontinuation (end of line), time between 2 lines of treatment, and numbers of patients treated in each line (first-, second-, and third-line settings and beyond). We used an interval of 2 months between 2 chemotherapy sessions to allow identification of the start of a new line of chemotherapy.
In-hospital deaths (all-cause mortality) identified during the patient follow-up period were also of interest, including number of deaths, annual mortality rates, and median survival.
The HCRU and costs related to locally advanced or mUC management included hospitalization characteristics (number of stays, sector, length of stay, establishment type, admission type) and treatments (diagnostic imaging, radiotherapy, chemotherapy, palliative care, drugs reimbursed in addition to DRG-based tariffs). These were identified within 1 year of follow-up through procedure codes, diagnoses, or both in an anonymous discharge summary in an acute hospital care setting or a corresponding structure in a rehabilitation center or home care setting.
2.5. Statistical analysis
Patient demographics, medical history of UC management, and comorbidities were described in all 4 study populations overall (Supplemental Figure S1) and stratified by age group at index date and sex. Continuous and count variables were described with standard descriptive statistics; counts and percentages were computed for categorical variables. All analyses were performed with SAS v9.4 (SAS Institute, Cary, NC, USA). Analyses of hospital HCRU and costs, in-hospital deaths, and treatment patterns were restricted to patients with index dates before or on 31 December 2019.
2.5.1. Incidence and prevalence
The annual hospital incidence and prevalence of locally advanced or mUC in the general French population were estimated using populations 1 and 2 (Supplemental Figure S1). The general adult French population was defined using census data from metropolitan France [14]. For any given year, the general population (i.e., denominator) was estimated from census data as the average adult population from 1 January of that year to 1 January of the following year. This estimate did not consider individual-level time at risk and did not exclude patients with prevalent locally advanced or mUC. Annual prevalence and incidence were defined as the proportion of the general population with a locally advanced or mUC diagnosis or a new locally advanced or mUC diagnosis, respectively, at any time during the year. Patients were considered prevalent until the year of their last hospitalization (MCO, SSR, or HAD) retrieved in the PMSI. Incidence and prevalence were described overall and stratified by age and sex.
2.5.2. Treatment patterns
Treatment patterns were analyzed using data from patients in population 2 with index dates before or on 31 December 2019 (Supplemental Figure S1). All patients were followed from the date of first diagnosis of locally advanced or mUC to in-hospital death or end of study. The numbers of treated and untreated patients were reported; treatment patterns were not reported for untreated patients or those with an indeterminate treatment line. Treatment patterns were described overall and according to each respective index date. Total numbers of treatment lines per patient were calculated. For each line, the type of treatment, time from previous to next line, treatment duration, and reason for discontinuing treatment were provided.
2.5.3. In-hospital deaths
The numbers of in-hospital deaths were analyzed in patients in population 2 with index dates before or on 31 December 2019 (Supplemental Figure S1). In-hospital mortality rates were calculated at 1, 2, and 3 years of follow-up; the denominator was the number of patients with sufficient theoretical follow-up for the considered timepoint, and the numerator was the number of deaths that occurred within that timepoint. The 1-year mortality rate was calculated as the number of deaths occurring within 1 year after index date divided by the total number of patients in the population with a theoretical follow-up of 1 year (i.e., those with an index date in 2017, 2018, or 2019). The 2-year mortality rate was calculated as the number of deaths occurring within 2 years after index date divided by the total number of patients in the population with a theoretical follow-up of 2 years (i.e., those with an index date in 2017 or 2018). The 3-year mortality rate was calculated as the number of deaths occurring within 3 years after index date divided by the total number of patients in the population with a theoretical follow-up of 3 years (i.e., those with an index date in 2017). Median survival was estimated using Kaplan–Meier methods. The numbers of in-hospital deaths and the annual in-hospital mortality rate, but not median survival, were also calculated for population 3.
2.5.4. Hospital health care resource utilization and costs
Hospital HCRU and costs were estimated using data from patients in population 3 with index dates before or on 31 December 2019 (Supplemental Figure S1). Only hospital stays related to locally advanced or mUC management (i.e., diagnosis of UC or metastatic cancer required as a reason for hospitalization) were included in the analysis. Costs were estimated from the health care payer perspective using data from the Agence Technique de l’Information sur l’Hospitalisation on official French hospital tariffs and tariffs for drugs in addition to the DRG, depending on the sector (public or private) and hospitalization year. Official costs in effect for each year from 2017 to 2020 were applied to each hospitalization depending on the year of discharge. Costs are presented as total cost, mean and median cost per stay, and mean and median costs per patient at the 1-year follow-up for each type of unit (MCO, SSR, HAD) overall and by year of index date. Costs were reported in Euros for each year of interest (e.g., for the year 2018, the 2018 French hospital tariffs were used, and costs are reported in 2018 Euros).
2.5.5. Missing data
Missing data were expected to be scarce given the exhaustive nature of the PMSI database [10]; therefore, they were not imputed. The numbers and percentages of missing values were reported for each variable and are only displayed when missing values were identified. For categorical variables, missing values were included in the denominator of the proportion calculations. No results are presented for variables with fewer than 11 patients or hospitalizations due to patient data protection rules.
2.5.6. Sensitivity analysis
A sensitivity analysis was conducted excluding patients who had any concurrent malignancy at the time of first diagnosis of UC to identify the potential for misclassification of patients with locally advanced or mUC using the algorithm.
3. Results
3.1. Study population
Study population 1 included 44,919 patients with a diagnosis of locally advanced or mUC (Figure 2). Of these patients, 78.6% were men, and the mean age at index date was 72 years, with most patients (61.1%) aged between 60 and 79 years (Table 1). Overall, 18.3% of patients in population 1 had a history of another malignancy. Population 2 included 34,795 patients newly diagnosed with locally advanced or mUC during the inclusion period, which could include patients with de novo locally advanced or mUC or those who progressed from earlier disease. On average, incident patients who did not receive first-line systemic treatment were older (mean age, 77.4 vs 70.1 years), had more comorbidities (mean age-adjusted CCI score, 11.1 vs 9.7), and had undergone slightly fewer surgeries and other treatment for UC compared with the treated group (Supplemental Table S4). Population 3 included 8094 patients who were treated in the second- or third-line setting during the inclusion period. At the start of the second- or third-line treatment, 80.1% of patients were men and the mean age was 70 years, with most (70.7%) patients aged 60 to 79 years. Population 4 included 2621 patients treated with a checkpoint inhibitor after receiving first-line chemotherapy during the inclusion period.
Figure 2.
Study population.
la/m: locally advanced or metastatic; UC: urothelial carcinoma.
aFor patients undergoing cystectomy, only those with a reported code of chemotherapy >4 mo following surgery have been included (n = 6413).
Table 1.
Baseline characteristics and medical history at index date.
| Characteristica | Population |
|||
|---|---|---|---|---|
| 1: Prevalent and incident la/mUC (N = 44,919) | 2: Incident la/mUC (N = 34,795) | 3: Starting second or third lines (N = 8094) | 4: Second- or third-line checkpoint inhibitor (N = 2621) | |
| Age, y | ||||
| Mean (SD) | 72.2 (10.7) | 72.8 (10.7) | 69.8 (9.5) | 70.0 (9.0) |
| Age group, y | ||||
| 18–39 | 233 (0.5) | 165 (0.5) | 37 (0.5) | NRb |
| 40–59 | 5054 (11.3) | 3603 (10.4) | 1064 (13.1) | 305 (11.6) |
| 60–79 | 27,441 (61.1) | 20,760 (59.7) | 5724 (70.7) | 1943 (74.1) |
| ≥80 | 12,191 (27.1) | 10,267 (29.5) | 1269 (15.7) | 366 (14.0) |
| Sex | ||||
| Male | 35,301 (78.6) | 27,185 (78.1) | 6483 (80.1) | 2134 (81.4) |
| Region of residence | ||||
| Auvergne-Rhône-Alpes | 5072 (11.3) | 4005 (11.5) | 901 (11.1) | 285 (10.9) |
| Bourgogne-Franche-Comté | 2371 (5.3) | 1862 (5.4) | 446 (5.5) | 139 (5.3) |
| Bretagne | 2007 (4.5) | 1596 (4.6) | 361 (4.5) | 130 (5.0) |
| Centre-Val de Loire | 2345 (5.2) | 1689 (4.9) | 495 (6.1) | 127 (4.8) |
| Corse | 326 (0.7) | 237 (0.7) | 67 (0.8) | 19 (0.7) |
| Grand Est | 4042 (9.0) | 3096 (8.9) | 620 (7.7) | 183 (7.0) |
| Hauts-de-France | 4010 (8.9) | 3123 (9.0) | 732 (9.0) | 230 (8.8) |
| Île-de-France | 6894 (15.3) | 5339 (15.3) | 1354 (16.7) | 501 (19.1) |
| Normandie | 2173 (4.8) | 1682 (4.8) | 316 (3.9) | 79 (3.0) |
| Nouvelle-Aquitaine | 4589 (10.2) | 3579 (10.3) | 871 (10.8) | 299 (11.4) |
| Occitanie | 4487 (10.0) | 3506 (10.1) | 812 (10.0) | 224 (8.5) |
| Pays de la Loire | 2149 (4.8) | 1745 (5.0) | 321 (4.0) | 113 (4.3) |
| Provence-Alpes-Côte d’Azur | 4449 (9.9) | 3332 (9.6) | 797 (9.8) | 291 (11.1) |
| Missing | 5 (<0.1) | 4 (<0.1) | 1 (<0.1) | 1 (<0.1) |
| Time since la/mUC diagnosis, dc | ||||
| Median (Q1–Q3) | NA | NA | 277 (177–445) | 234 (140–357) |
| Previous treatment for UC | ||||
| Radical cystectomy | 6716 (15.0) | 5583 (16.0) | 2166 (26.8) | 795 (30.3) |
| Partial cystectomy | 601 (1.3) | 474 (1.4) | 225 (2.8) | 62 (2.4) |
| Total nephroureterectomy | 2164 (4.8) | 1776 (5.1) | 747 (9.2) | 298 (11.4) |
| Tumor resection | 30,092 (67.0) | 23,541 (67.7) | 6007 (74.2) | 1952 (74.5) |
| Radiotherapy | 1751 (3.9) | 1378 (4.0) | 1385 (17.1) | 390 (14.9) |
| Intravesical instillation | 3474 (7.7) | 2752 (7.9) | 604 (7.5) | 205 (7.8) |
| Neoadjuvant/adjuvant chemotherapy | 2412 (5.4) | 1934 (5.6) | 1402 (17.3) | 499 (19.0) |
| Comorbidity | ||||
| Myocardial infarction | 4684 (10.4) | 3939 (11.3) | 880 (10.9) | 319 (12.2) |
| Congestive heart failure | 3471 (7.7) | 3082 (8.9) | 554 (6.8) | 203 (7.7) |
| Peripheral vascular disease | 6597 (14.7) | 5555 (16.0) | 1303 (16.1) | 488 (18.6) |
| Cerebrovascular disease | 3301 (7.3) | 2845 (8.2) | 531 (6.6) | 206 (7.9) |
| Dementia | 854 (1.9) | 784 (2.3) | 52 (0.6) | 17 (0.6) |
| Chronic pulmonary disease | 6795 (15.1) | 5613 (16.1) | 1374 (17.0) | 506 (19.3) |
| Hepatic disease | 1674 (3.7) | 1433 (4.1) | 380 (4.7) | 124 (4.7) |
| Diabetes | 8418 (18.7) | 6881 (19.8) | 1485 (18.3) | 483 (18.4) |
| Moderate/Severe renal disease | 5944 (13.2) | 5192 (14.9) | 1363 (16.8) | 510 (19.5) |
| Other cancersd | 8198 (18.3) | 6725 (19.3) | 3453 (42.7) | 1050 (40.1) |
| Lung | 1330 (3.0) | 1104 (3.2) | 807 (10.0) | 389 (14.8) |
| Prostate | 3440 (7.7) | 2809 (8.1) | 1064 (13.1) | 319 (12.2) |
| Adjusted CCI score | ||||
| Mean (SD) | 10.0 (2.2) | 10.2 (2.3) | 9.9 (2.1) | 10.1 (2.2) |
CCI: Charlson Comorbidity Index; la/m: locally advanced or metastatic; NA: not applicable; NR: not reported; UC: urothelial carcinoma.
aValues are n (%) unless otherwise indicated.
bResults for variables with <11 patients not included.
cTime from diagnosis to start of second- or third-line treatment.
dAny other type of cancer, including lung and prostate cancers.
3.2. Epidemiology
Annual prevalence of locally advanced or mUC ranged from 18,443 to 19,909 cases overall and 36.4 to 38.9 cases per 100,000 people between 2017 and 2020 (Table 2). Prevalence was stable over the study period and increased with age. Prevalence was higher in men versus women. Annual incidence of locally advanced or mUC ranged from 8319 to 8660 cases overall and 16.4 to 16.9 cases per 100,000 people from 2017 to 2019. A slightly higher incidence was observed in the year 2020, with 9481 cases overall and 18.5 cases per 100,000 people (Table 2). As with prevalence, incidence increased with age and was higher in men versus women.
Table 2.
Prevalence and incidence of locally advanced or metastatic urothelial carcinoma in metropolitan France stratified by age and sex (2017–2020).
| Year |
||||
|---|---|---|---|---|
| Characteristic | 2017 | 2018 | 2019 | 2020 |
| Prevalence | ||||
| Cases, n | ||||
| Overall | 18,443 | 19,441 | 19,909 | 19,741 |
| Age at index date, y | ||||
| 18–39 | 104 | 109 | 98 | 102 |
| 40–59 | 2375 | 2458 | 2402 | 2266 |
| 60–79 | 11,543 | 12,215 | 12,675 | 12,731 |
| ≥80 | 4421 | 4659 | 4734 | 4642 |
| Sexa | ||||
| Male | 14,952 | 15,419 | 15,765 | 15,682 |
| Female | 3847 | 4017 | 4138 | 4054 |
| Cases per 100,000 people, n | ||||
| Overall | 36.4 | 38.2 | 38.9 | 38.5 |
| Age at index date, y | ||||
| 18–39 | 0.6 | 0.6 | 0.6 | 0.6 |
| 40–59 | 14.0 | 14.5 | 14.2 | 13.5 |
| 60–79 | 90.5 | 94.0 | 95.8 | 94.4 |
| ≥80 | 111.8 | 116.6 | 117.0 | 114.2 |
| Sexa | ||||
| Male | 60.5 | 63.6 | 64.8 | 64.2 |
| Female | 14.5 | 15.1 | 15.5 | 15.1 |
| Incidence | ||||
| Cases, n | ||||
| Overall | 8319 | 8335 | 8660 | 9481 |
| Age at index date, y | ||||
| 18–39 | 36 | 44 | 39 | 46 |
| 40–59 | 924 | 899 | 848 | 932 |
| 60–79 | 4862 | 4870 | 5215 | 5813 |
| ≥80 | 2497 | 2522 | 2558 | 2690 |
| Sexb | ||||
| Male | 6476 | 6536 | 6719 | 7454 |
| Female | 1841 | 1797 | 1938 | 2026 |
| Cases per 100,000 people, n | ||||
| Overall | 16.4 | 16.4 | 16.9 | 18.5 |
| Age at index date, y | ||||
| 18–39 | 0.2 | 0.3 | 0.2 | 0.3 |
| 40–59 | 5.4 | 5.3 | 5.0 | 5.5 |
| 60–79 | 38.1 | 37.5 | 39.4 | 43.1 |
| ≥80 | 63.2 | 63.1 | 63.2 | 66.2 |
| Sexb | ||||
| Male | 26.8 | 27.0 | 27.6 | 30.5 |
| Female | 6.9 | 6.7 | 7.2 | 7.5 |
aData missing for 20 patients with prevalent disease.
bData missing for 8 patients with incident disease.
3.3. Treatment patterns
Treatment patterns were analyzed among 25,314 patients in population 2 with an index date on or before (i.e., included ≥1 year before the end of the study period). More than one-third (37.6%) of these patients did not receive any first-line systemic treatment and 4.5% had indeterminate first-line treatment. Therapeutic pathways are described for the 14,656 patients who started an identifiable first-line treatment between 2017 and 2019 (Figure 3). Of these patients, 66.6% received only 1 line of therapy, 22.5% received 2 lines of therapy, and 10.9% received 3 lines of therapy (Table 3). Of patients who received first-line treatment, 33.4% (n/N = 4896/14,656) received second-line treatment; of those receiving second-line treatment, 32.3% (n/N = 1598/4896) received third-line treatment. Mean (SD) time from locally advanced or mUC diagnosis to start of first-line therapy was 10.6 (20.9) days. Median times to start of next treatment were 131.0 days between first- and second-line treatment and 76.0 days between second- and third-line treatment. Median time between first- and second-line treatment in patients newly diagnosed with locally advanced or metastatic UC decreased over time, from 147 days in 2017 to 141 days in 2018 and 108 days in 2019.
Figure 3.
Therapeutic pathways for patients newly diagnosed with la/mUC between 2017 and 2019.
la/mUC: locally advanced or metastatic urothelial carcinoma.
Table 3.
Treatment patterns by line in patients with identifiable treatment lines (population 2).
| Overall |
Year of index dateb |
|||
|---|---|---|---|---|
| 2017–2019 | 2017 | 2018 | 2019 | |
| Variablea | (N = 14,656) | (n = 4848) | (n = 4818) | (n = 4990) |
| Treatment line per patient | ||||
| 1 | 9760 (66.6) | 3262 (67.3) | 3226 (67.0) | 3272 (65.6) |
| 2 | 3298 (22.5) | 1022 (21.1) | 1060 (22.0) | 1216 (24.4) |
| 3 | 1598 (10.9) | 564 (11.6) | 532 (11.0) | 502 (10.1) |
| First line | ||||
| Time between diagnosis of la/mUC and start of first line, d | ||||
| Mean (SD) | 10.6 (20.9) | 10.7 (21.0) | 10.6 (21.0) | 10.4 (20.8) |
| Median (Q1–Q3) | 0 (0–9) | 0 (0–10) | 0 (0–9) | 0 (0–9) |
| Min–max | 0–90 | 0–90 | 0–90 | 0–90 |
| Type of treatment | ||||
| Chemotherapy | 14,389 (98.2) | 4794 (98.9) | 4724 (98.0) | 4871 (97.6) |
| Checkpoint inhibitor | 267 (1.8) | 54 (1.1) | 94 (2.0) | 119 (2.4) |
| Pembrolizumab | 116 (43.4) | <11 | 41 (43.6) | 66 (55.5) |
| Atezolizumab | <11 | <11 | <11 | <11 |
| Nivolumab | 143 (53.6) | 45 (83.3) | 52 (55.3) | 46 (38.7) |
| Avelumab | <11 | <11 | <11 | <11 |
| Duration of treatment, d | ||||
| Mean (SD) | 94.3 (105.2) | 100.1 (117.2) | 94.6 (109.2) | 88.5 (87.0) |
| Median (Q1–Q3) | 67 (35–122) | 69 (36–128) | 67 (34–120) | 66 (33–120) |
| Min–max | 1–1359 | 1–1359 | 1–1091 | 1–721 |
| End of treatment | ||||
| Treatment discontinuation | 6192 (42.2) | 2238 (46.2) | 2175 (45.1) | 1779 (35.7) |
| Switch to the next line | 626 (4.3) | 101 (2.1) | 106 (2.2) | 419 (8.4) |
| End of follow-up | 7838 (53.5) | 2509 (51.8) | 2537 (52.7) | 2792 (56.0) |
| Second line | ||||
| Patients in 1 L | 4896 (33.4) | 1586 (32.7) | 1592 (33.0) | 1718 (34.4) |
| Time between first- and second-line treatment, d | ||||
| Mean (SD) | 187.8 (176.0) | 230.7 (221.3) | 201.6 (169.6) | 135.3 (108.8) |
| Median (Q1–Q3) | 131.0 (77.0–234.0) | 147.0 (88.0–289.0) | 140.5 (84.0–252.5) | 108.0 (62.0–190.0) |
| Min–max | 1–1411 | 4–1411 | 1–943 | 1–652 |
| Type of treatment | ||||
| Chemotherapy | 3558 (72.7) | 1431 (90.2) | 1330 (83.5) | 797 (46.4) |
| Checkpoint inhibitor | 1338 (27.3) | 155 (9.8) | 262 (16.5) | 921 (53.6) |
| Pembrolizumab | 1109 (82.9) | 87 (56.1) | 186 (71.0) | 836 (90.8) |
| Atezolizumab | 36 (2.7) | <11 | <11 | 23 (2.5) |
| Nivolumab | 193 (14.4) | 60 (38.7) | 71 (27.1) | 62 (6.7) |
| Avelumab | 19 (0.4) | <11 | <11 | 13 (0.8) |
| Duration, d | ||||
| Mean (SD) | 103.8 (126.6) | 118.4 (151.1) | 110.2 (132.7) | 84.5 (87.8) |
| Median (Q1–Q3) | 64.0 (27.5–135.0) | 71.0 (29.0–148.0) | 67.0 (29.0–141.0) | 56.5 (22.0–113.0) |
| Min–max | 1–1302 | 1–1302 | 1–879 | 1–537 |
| End of treatment | ||||
| Treatment discontinuation | 1495 (30.5) | 688 (43.4) | 534 (33.5) | 273 (15.9) |
| Switch to the next line | 585 (11.9) | 113 (7.1) | 154 (9.7) | 318 (18.5) |
| End of follow-up | 2816 (57.5) | 785 (49.5) | 904 (56.8) | 1127 (65.6) |
| Third line | ||||
| Patients in 1 L | 1598 (10.9) | 564 (11.6) | 532 (11.0) | 502 (10.1) |
| Time between second- and third-line treatment, d | ||||
| Mean (SD) | 105.2 (109.7) | 135.7 (131.3) | 114.3 (106.1) | 61.4 (63.0) |
| Median (Q1–Q3) | 76.0 (33.0–131.0) | 98.0 (63.5–168) | 84.0 (43.5–140.5) | 37.0 (21.0–77.0) |
| Min–max | 0–1029 | 0–1029 | 0–741 | 0–475 |
| Type of treatment | ||||
| Chemotherapy | 1068 (66.8) | 432 (76.6) | 316 (59.4) | 320 (63.7) |
| Checkpoint inhibitor | 530 (33.2) | 132 (23.4) | 216 (40.6) | 182 (36.3) |
| Pembrolizumab | 472 (89.1) | 103 (78.0) | 199 (92.1) | 170 (93.4) |
| Atezolizumab | 12 (2.3) | <11 | <11 | <11 |
| Nivolumab | 46 (8.7) | 27 (20.5) | 13 (6.0) | <11 |
| Avelumab | <11 | <11 | <11 | <11 |
| Duration of treatment, d | ||||
| Mean (SD) | 89.7 (106.8) | 102.6 (128.5) | 92.5 (107.3) | 72.2 (71.2) |
| Median (Q1–Q3) | 57.0 (21.0–123.0) | 64.0 (22.0–127.0) | 57.0 (22.0–126.5) | 44.0 (15.0–110.0) |
| Min–max | 1–960 | 1–960 | 1–820 | 1–383 |
| End of treatment | ||||
| Treatment discontinuation | 332 (20.8) | 171 (30.3) | 118 (22.2) | 43 (8.6) |
| Switch to immune checkpoint inhibitor | 196 (12.3) | 70 (12.4) | 85 (16.0) | 41 (8.2) |
| End of follow-up | 1070 (67.0) | 323 (57.3) | 329 (61.8) | 418 (83.3) |
1 L: first line; la/mUC: locally advanced or metastatic urothelial carcinoma.
aValues are n (%) unless otherwise indicated.
bResults for patients diagnosed in 2020 not reported due to insufficient follow-up duration (<1 y).
Most (98.2%) patients who received first-line treatment received chemotherapy. The proportion of patients receiving a checkpoint inhibitor in the second-line setting increased over the study period, from 9.8% in 2017 to 53.6% in 2019. Although the proportion of patients treated with checkpoint inhibitors in the third-line setting also increased over the study period, chemotherapy remained the most common third-line treatment.
3.4. In-hospital deaths
Among patients in population 2 who were included at least 1 year before the end of the study period, annual in-hospital death rates were 47.8% (n/N = 12,079/25,314) at 1 year, 58.1% (n/N = 9666/16,654) at 2 years, and 62.9% (n/N = 5239/8319) at 3 years of follow-up. Median survival was 18 months (Q1–Q3: 61 days–not reached) from locally advanced or mUC diagnosis.
Among patients in population 3 who were included at least 1 year before the end of the study period, annual in-hospital death rates were 43.6% (n/N = 2133/4896) at 1 year, 55.2% (n/N = 1754/3178) at 2 years, and 60.4% (n/N = 958/1586) at 3 years of follow-up. Median survival was not calculated for this population.
3.5. Health care resource utilization and costs
A total of 72,085 MCO, 857 SSR, and 1016 HAD hospitalizations reported with a primary or a related diagnosis code of UC were identified in the PMSI database among patients in population 3 who initiated second- or third-line therapy between 2017 and 2019 (Figure 4). Of these 59,286 MCO (n = 4949 patients; 12.0 stays per patient), 719 SSR (n = 523 patients; 1.4 stays per patient), and 885 HAD (n = 600 patients; 1.5 stays per patient) hospitalizations related to locally advanced or mUC management were included in the analysis. More hospitalizations were observed in the public versus private sector; hospitalizations in the public sector accounted for 69.9%, 60.1%, and 55.0% of hospitalizations in MCO, SSR, and HAD units, respectively (Table 4). In MCO units, 81.6% of hospitalizations were day cases (i.e., same-day admission and discharge).
Figure 4.
Flow diagram of select stays for health care resource utilization and health care cost analysis (population 3).
HAD: hospitalization à domicile; la/mUC: locally advanced or metastatic urothelial carcinoma; MCO: médecine, chirurgie, obstétrique; SSR: soins de suite et de réadaptation.
Table 4.
Health care resource utilization in the second- and third-line settings at 1 year of follow-up (population 3).
| Type of facility |
|||
|---|---|---|---|
| Variablea | Acute-stay institution (MCO) (N = 59,286 stays) |
After-care/Rehabilitation clinic (SSR) (N = 719 stays) |
At-home hospital unit (HAD) (N = 885 stays) |
| Sector | |||
| Public | 41,412 (69.9) | 432 (60.1) | 487 (55.0) |
| Private | 17,874 (30.1) | 287 (39.9) | 398 (45.0) |
| Type of establishment | |||
| CHU/CHR | 10,054 (17.0) | 31 (4.3) | 57 (6.4) |
| CH | 14,557 (24.6) | 199 (27.7) | 175 (19.8) |
| ESPIC | 15,588 (26.3) | 186 (25.9) | 276 (31.2) |
| Private | 17,567 (29.6) | 282 (39.2) | 332 (37.5) |
| Other | 1520 (2.6) | 21 (2.9) | 45 (5.1) |
| Type of admission | |||
| Day case | 48,355 (81.6) | NA | NA |
| Overnight | 10,931 (18.4) | NA | NA |
| Length of stay, d | |||
| Mean (SD) | 1.8 (6.4) | 29.1 (27.6) | 26.4 (32.9) |
| Median (Q1–Q3) | 0 (0) | 21 (12–37) | 14 (7–33) |
| Range (min–max) | 0–344 | 1–257 | 1–284 |
| Diagnostic imaging | 3678 (6.2) | 49 (6.8) | 12 (1.4) |
| Therapy | |||
| Radiotherapy | 7153 (12.1) | 11 (1.5) | 14 (1.6) |
| Chemotherapyb | 42,709 (72.0) | 38 (5.3) | 66 (7.5) |
| Palliative care | 3057 (5.2) | 300 (41.7) | 631 (71.3) |
CH: center hospitalier (hospital center); CHU/CHR: center hospitalier universitaire/center hospitalier régionaux (teaching hospital/regional hospital center); ESPIC: etablissement de santé privé d’intérêt collectif (private health establishment of collective interest); HAD: hospitalization à domicile; MCO: médecine, chirurgie, obstétrique; NA: not applicable; SSR: soins de suite et de réadaptation.
aValues are n (%) unless otherwise indicated.
bIncludes immuno-oncology treatments (immune checkpoint inhibitors).
Treatments, including radiotherapy and chemotherapy, were mainly administered in MCO units (84.1% of MCO hospitalizations), whereas palliative care was mainly provided in HAD units (71.3% of HAD hospitalizations) and, to a lesser extent, in SSR units (41.7% of SSR hospitalizations). Mean estimated annual costs of care per patient at 1 year from the start of second- and third-line treatments were €16,012 in MCO units, €8803 in SSR units, and €9029 in HAD units (Table 5).
Table 5.
Costs of management of locally advanced or metastatic urothelial carcinoma in second- and third-line settings at 1 year of follow-up (population 3).
| Variable | Cost of stay,a € |
||
|---|---|---|---|
| Acute-stay institution (MCO) (N = 59,286) |
After-care/Rehabilitation clinic (SSR) (N = 719) |
At-home hospital unit (HAD) (N = 885) |
|
| Total cost of hospital stays | 79,241,520 | 4,604,078 | 5,417,516 |
| Mean per stay | 1337 | 6403 | 6122 |
| Median per stay | 383 | 4250 | 3850 |
| Mean per patient | 16,012 | 8803 | 9029 |
| Median per patient | 11,114 | 5860 | 5260 |
| Intragroup | 59,233,306 | 4,592,238 | 5,396,340 |
| Mean per stay | 999 | 6387 | 6098 |
| Median per stay | 383 | 4250 | 3817 |
| Mean per patient | 11,969 | 8781 | 8994 |
| Median per patient | 9399 | 5859 | 5238 |
| Drugs funded on top of group-based tariffs | 20,008,214 | NRb | NRb |
| Mean per stay | 4435 | NRb | NRb |
| Median per stay | 5294 | NRb | NRb |
| Mean per patient | 22,182 | NRb | NRb |
| Median per patient | 15,715 | NRb | NRb |
| Palliative care | 17,914,607 | 2,276,875 | 4,292,666 |
| Mean per stay | 5860 | 7590 | 6803 |
| Median per stay | 5243 | 5612 | 4330 |
| Mean per patient | 8939 | 8791 | 9075 |
| Median per patient | 6730 | 6139 | 5079 |
HAD: hospitalization à domicile; MCO: médecine, chirurgie, obstétrique; NR: not reported; SSR: soins de suite et de réadaptation.
aDerived from publicly available diagnosis-related group costs available for each year of the inclusion period from a French public health care insurance perspective. Total costs were estimated at 1 y of follow-up (all calendar years pooled) in the subpopulation of patients who had started a second- or a third-line treatment before 2020 by applying the official French hospital tariffs in effect at the time of each hospital stay (2017–2020). No specific cost year was used to describe the distribution of costs across the study period.
bResults for variables with <11 stays not included.
3.6. Sensitivity analysis
In the analysis, excluding patients with a history of malignancy before their first UC diagnosis, similar proportions of patients with a concomitant cancer diagnosis were seen across the 4 study populations (population 1: 17.3% vs 18.3%; population 2: 18.3% vs 19.3%; population 3: 35.8% vs 42.7%; population 4: 35.2% vs 40.1%). In the years 2017 to 2020, there were 27,585 patients without any concurrent cancer who were newly diagnosed with locally advanced or mUC. Incidence ranged from 13.1 to 13.5 cases per 100,000 people in 2017 to 2019, with a slightly higher incidence observed in 2020 (14.6 new cases per 100,000 people).
4. Discussion
This study included the entire population with locally advanced or mUC in the exhaustive PMSI database. The results provide recent data on the epidemiology of locally advanced or mUC in France. From 2017 to 2020, annual locally advanced or mUC prevalence ranged from 18,443 to 19,909 cases (36.4–38.9 per 100,000 people) and incidence ranged from 8319 to 9481 cases (16.4–18.5 per 100,000 people). More than one-third (37.6%) of patients with locally advanced or mUC did not receive any systemic treatment; of those with identifiable treatment line data, only 10.9% reached third-line treatment. Patients may not be fit enough or willing to receive systemic therapy; for example, in our sample, patients who did not receive first-line systemic treatment were older and had more comorbidities than those who were treated. Furthermore, these findings appear to be consistent with treatment rates reported in a recent systematic literature review, reporting between 40% and 74% of patients not receiving first-line systemic treatment for locally advanced or mUC in European studies [15]. These results show the need to reinforce supportive care early in the management of advanced UC.
Chemotherapy was the most common treatment received in the first-, second-, and third-line settings. Use of checkpoint inhibitors in the first-line setting was very low (1.8%), likely due to lack of reimbursement in France during the study period. Although it is undetermined how those patients had access to first-line checkpoint inhibitor therapy, they may have been provided access via compassionate use programs, clinical trial participation, or through off-label use. The proportion of patients receiving a checkpoint inhibitor in the second-line setting increased during the study period (9.8% in 2017 vs 53.6% in 2019), coinciding with the Haute Autorité de santé (HAS) reimbursement appraisal of pembrolizumab for this indication in August 2017 and its funding in November 2019 [16]. Pembrolizumab is currently the only immunotherapy funded in France in the post–platinum-based chemotherapy setting.
Despite the fact that the proportion of patients treated with a checkpoint inhibitor in a third-line setting also increased over the study period, chemotherapy remained the most common third-line treatment. These results are consistent with an analysis of Adelphi mUC Disease Specific Programme cross-sectional survey data (November 2020–April 2021) for France, in which most patients received first-line chemotherapy and the proportions of patients receiving a checkpoint inhibitor were 6%, 66%, and 24% in the first-, second-, and third-line settings, respectively [17].
The treatment paradigm in France has changed since the end of our study period (December 2020). Avelumab maintenance therapy was recommended by the HAS in 2022 as monotherapy for first-line maintenance treatment of adult patients with locally advanced or mUC whose disease has not progressed after platinum-based chemotherapy and was funded for this indication starting in September 2022. Enfortumab vedotin received a reimbursement appraisal and early-access program renewal from the HAS in December 2022 as monotherapy in the treatment of adults with locally advanced or mUC who have previously received platinum-based chemotherapy and a programmed cell death 1/ligand 1 inhibitor [18,19]; the early access program for enfortumab vedotin began in July 2021. Based on the results of the EV-302 pivotal trial [20], enfortumab vedotin in combination with pembrolizumab was granted European marketing authorization in August 2024 for the first-line treatment of adult patients with unresectable or mUC who are eligible for platinum-containing chemotherapy [21]; this combination received a favorable early access authorization opinion from the Transparency Commission in September 2024 but has not yet been authorized by the HAS [22]. As the treatment paradigm continues to evolve, future research is needed to understand changes in the burden and treatment pathways of locally advanced or mUC in France.
In the second- and third-line settings, the estimated mean annual costs of care per patient at 1 year of follow-up were €16,012 in MCO units, €8803 in SSR units, and €9029 in HAD units. There are very limited HCRU and cost data in Europe in the locally advanced or mUC setting. In a study of the economic burden of bladder cancer in the European Union in 2012, the estimated total health care cost per prevalent bladder cancer case in France was €13,370 [23], which is within the range of costs observed in the EVOLVE study. Although estimates are comparable, results were reported using different data sources and over a different study period, thereby limiting the interpretation of differences between the costs reported. This PMSI study provides the first estimate of the mean annual costs of hospital care per patient with locally advanced or mUC in the second- and third-line locally advanced or mUC settings in France.
Annual in-hospital death rates were 47.8% at 1 year, 58.1% at 2 years, and 62.9% at 3 years of follow-up. The estimated median survival from locally advanced or mUC diagnosis in our analysis (18 months) is longer than the median overall survival observed with first-line locally advanced or mUC regimens in a 2023 network meta-analysis (13.2 months for cisplatin-eligible patients and 9.7–12.0 months for cisplatin-ineligible patients) [7]. The longer median overall survival seen in our analysis may be explained by the algorithm used to identify patients with locally advanced or mUC, which does not allow for differentiation between locally advanced and metastatic cancer, and/or the impact of new checkpoint inhibitor treatments. In addition, median survival in our analysis was only calculated among patients with at least 1 year of follow-up and was based on in-hospital deaths only, i.e., deaths that were captured within PMSI. There is limited information on the proportion of deaths not recorded in the PMSI system. A study reported that 23.4% of cancer-related deaths occurred outside hospital settings (where hospital settings include short-stay hospital, hospital at home, or rehab) in France in 2013: 7.0% of deaths occurred in nursing homes, and 16.4% occurred in other settings, including home [24]. Therefore, it is possible the median overall survival results may be an overestimation because deaths that occurred in settings that are not captured in the database were not included [25]. However, a recent systematic literature review reported a wide range of real-world survival estimates from 2.0 to 6.9 months without systemic therapy and from 9.2 to 34.5 months with systemic therapy [15].
Although we report data from a nationwide – and thus comprehensive – cohort of patients with locally advanced or mUC in France, limitations associated with the use of the PMSI dataset should be considered when interpreting our results. Patients with locally advanced or mUC were identified using an algorithm based on one developed for pembrolizumab (validated by the HAS) and reviewed by experts of the scientific committee [16]. Misclassification was possible in the absence of information regarding cancer stage. Patients were not required to have a code for metastatic cancer, so identification of patients with mUC was not likely to be exhaustive; when it was coded, determination of whether the associated location was the primary or secondary location was uncertain. It is possible that patients with muscle-invasive bladder cancer were included in the sample if they had codes for metastasis, chemotherapy, or both. Nevertheless, the estimated incidence of locally advanced or mUC (range, 8319–9481 patients) identified in the present study was consistent with the findings of the 2014 pembrolizumab algorithm (n = 8118 patients) [26].
At UC diagnosis, 18% to 19% of patients in populations 1 and 2 and 40% to 43% of patients in populations 3 and 4 had a code indicating concomitant cancer. In the sensitivity analysis excluding these patients, the proportions of patients with concomitant cancer at the time of UC diagnosis did not substantially change, but the mean annual locally advanced or mUC incidence was lower (12.9–14.6 vs 16.4–18.5 per 100,000 people), highlighting a possible misclassification of some patients and potential overestimation of the number of patients with locally advanced or mUC. It is possible that the chemotherapy sessions or administration of advanced-stage treatments used in the study population identification algorithm could have been administered for another concurrent cancer diagnosis but, as it cannot be ruled out that these other cancers were metastatic sites of UC, these patients were not excluded from analyses.
Use of chemotherapies in locally advanced or mUC could be identified only by chemotherapy session codes; the type of drug is not identifiable in the PMSI. In clinical practice, the change from one treatment line to another occurs when tumor progression is present, but this is not identifiable in the PMSI database. We used an interval of 2 months between 2 chemotherapy sessions to allow identification of the start of a new line of chemotherapy, but this approach does not allow identification of patients who progressed while on treatment or those who switched from first-line standard chemotherapy to second-line standard chemotherapy. Due to the difficulty of differentiating between lines of treatments for chemotherapy, there was potential for underestimation of the number of patients initiating second- or third-line treatment. We did not specifically analyze avelumab maintenance use because it was not generally available during the study period; similarly, data on erdafitinib use were not available.
The PMSI database did not allow inclusion of patients with locally advanced or mUC who were only treated in community settings (i.e., not hospitalized), so there was potential for selection bias. However, considering the severity of locally advanced or mUC, few patients were expected to be treated in the community setting alone, so the data are still expected to be fairly representative. There was potential for estimation bias in the calculation of locally advanced or mUC incidence because the denominator was the entire French population, including all prevalent patients and not only patients at risk. However, given the low prevalence and median survival of locally advanced or mUC, the use of the entire French population as a denominator was not expected to affect estimations of incidence. Although PMSI data are geared toward the billing process rather than more complex requirements for scientific research [10], they are a valuable source that combines data on patient profiles, treatment pathways, HCRU, and health care costs in real-world settings.
5. Conclusions
The analyses presented herein allow for a better understanding of the real-world epidemiology, characterization of patients, and therapeutic management of locally advanced or mUC in France from 2017 to 2020. The disease burden of locally advanced or mUC was substantial. More than one-third of patients with locally advanced or mUC did not receive any first-line systemic treatment, demonstrating that many patients may be ineligible for or are not choosing therapy. Patients who do receive therapy are still at high risk for poor outcomes, and most treated patients received only one treatment line. Annual in-hospital death rates and annual costs of care were high in patients starting second- or third-line treatment during the inclusion period. Given the treatment landscape has evolved since this study period, updated research in this population is required. These results highlight the unmet need for additional therapies for treatment of locally advanced or mUC in France and for optimal supportive care early in disease management.
Supplementary Material
Acknowledgments
Astellas study management was provided by Franck Bruon.
Funding Statement
The study and this manuscript were funded by Astellas Pharma Inc. and Seagen, which was acquired by Pfizer in December 2023. The funders were involved in the study design, analysis, decision to publish, and preparation of the manuscript.
Article highlights
Little is known about the epidemiology, treatment patterns, health care resource utilization (HCRU), and health care costs in patients with locally advanced or metastatic urothelial carcinoma (mUC) in France, particularly after first-line systemic treatment.
This retrospective study included the entire population with locally advanced or mUC in France from 1 January 2017, to 31 December 2020, in the exhaustive PMSI national hospital discharge database.
The study included 4 cohorts: (1) all patients with locally advanced or mUC diagnosis between 2017 and 2020, (2) patients with new locally advanced or mUC diagnosis only, (3) patients starting second-/third-line treatment, and (4) patients treated with first-line chemotherapy who were starting second-/third-line treatment with a checkpoint inhibitor.
From 2017 to 2020, annual locally advanced or mUC prevalence ranged from 36.4 to 38.9 cases per 100,000 people; annual incidence ranged from 16.4 to 18.5 cases per 100,000 people.
More than one-third (37.6%) of patients with locally advanced or mUC did not receive any systemic treatment, demonstrating that many patients may be ineligible for or are not choosing therapy, and those who do receive therapy are still at high risk for poor outcomes.
Annual in-hospital death rates ranged from 47.8% at 1 year to 62.9% at 3 years of follow-up, and annual costs of care were high (€8803–€16,012, depending on facility type) in patients starting second- or third-line treatment during the inclusion period.
The disease burden of locally advanced or mUC in France remains substantial, highlighting the need for additional therapies.
The treatment landscape has changed since this study period, so further research is still needed in this population.
Author contributions
Study design: Florence Joly, Stéphane Culine, Morgan Rouprêt, Aurore Tricotel, Emilie Casarotto, Sandrine Brice, Marthe Vuillet, Marie-Catherine Thomas, Kirsten Leyland, Anil Upadhyay, Vicki Munro, Torsten Strunz-McKendry
Data acquisition: Aurore Tricotel, Emilie Casarotto, Sandrine Brice
Analysis: Aurore Tricotel, Emilie Casarotto, Sandrine Brice, Rafaël Minacori, Marthe Vuillet, Kirsten Leyland, Vicki Munro
Interpretation and writing: All authors provided interpretation of the data, reviewed and revised the manuscript for important intellectual content, gave final approval for the version to be published, and agreed to be accountable for all aspects of the work.
Disclosure statement
Florence Joly reports consulting fees from GSK; payment or honoraria for lectures and presentations (symposium) from 3A, Amgen, Astellas, AstraZeneca, GSK, Ipsen, MSD/ESAI, and Pfizer; support for attending meetings and/or travel from GSK, Ipsen, MSD/ESAI, and Pfizer; participating on an Advisory Board for AstraZeneca, Bayer, GSK, Ipsen, MSD/ESAI, and Pfizer; as well as an unpaid leadership or fiduciary role for GCIG. Stephane Culine reports consulting fees from Bayer; payment or honoraria from AstraZeneca, Ipsen, Janssen, Merck, MSD, and Takeda; support for attending meetings and/or travel from Janssen; and participating on a Data Safety Monitoring Board or Advisory Board for Amgen and BMS. Morgan Roupret reports consulting fees from Astellas, Bayer, BMS, Ipsen, and Janssen. Aurore Tricotel, Emilie Casarotto, and Sandrine Brice are employees of IQVIA, contracted by Astellas Pharma Inc. to conduct the study. Rafael Minacori, Marthe Vuillet, and Marie-Catherine Thomas are employees of Astellas Pharma France. Kirsten Leyland, Anil Upadhyay, Vicki Munro, and Torsten Strunz-McKendry are employees of Astellas Pharma Europe Ltd. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
Medical writing support was provided by Catherine Mirvis, BA, and Beth Lesher, PharmD, BCPS, from OPEN Health and was funded by Astellas Pharma Inc. and Pfizer.
Ethical declaration
The authors state that they have followed the principles outlined in the Declaration of Helsinki for the research described. Ethics approval for the use of PMSI pseudonymized data was covered under the French regulations (MR-006), and informed consent was not required.
Data availability statement
Researchers may request access to anonymized participant-level data, trial-level data, and protocols from Astellas sponsored clinical trials at www.clinicalstudydatarequest.com. For the Astellas criteria on data sharing, see https://clinicalstudydatarequest.com/Study-Sponsors/Study-Sponsors-Astellas.aspx.
Supplementary material
Supplemental data for this article can be accessed online at https://doi.org/10.1080/14796694.2025.2459058
References
Papers of special note have been highlighted as either of interest (•) or of considerable interest (••) to readers.
- 1.Dyba T, Randi G, Bray F, et al. The European cancer burden in 2020: incidence and mortality estimates for 40 countries and 25 major cancers. Eur J Cancer. 2021;157:308–347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Berdik C. Unlocking bladder cancer. Nature. 2017;551(7679):S34–S35. [DOI] [PubMed] [Google Scholar]
- 3.Witjes JA, Bruins HM, Carrión A, et al. Muscle-invasive and metastatic bladder cancer [Internet]. Arnhem, The Netherlands: European Association of Urology; [cited 2022 Jul 8]. Available from: https://uroweb.org/guidelines/muscle-invasive-and-metastatic-bladder-cancer [Google Scholar]; • Updated 2022 guidelines for bladder cancer from the European Association of Urology (EAU).
- 4.Neuzillet Y, Audenet F, Loriot Y, et al. French AFU cancer committee guidelines – update 2022–2024: muscle-invasive bladder cancer (MIBC). Prog Urol. 2022;32(15):1141–1163. [DOI] [PubMed] [Google Scholar]; • 2022–2024 recommendations for muscle-invasive bladder cancer from the Comité de cancérologie de l’Association Française d’Urologie (CCAFU).
- 5.Powles T, Bellmunt J, Comperat E, et al. Bladder cancer: ESMO clinical practice guideline for diagnosis, treatment and follow-up. Ann Oncol. 2022;33(3):244–258. [DOI] [PubMed] [Google Scholar]; • Updated 2022 guidelines for bladder cancer from the European Society for Medical Oncology (ESMO).
- 6.Rouprêt M, Pignot G, Masson-Lecomte A, et al. French CCAFU guidelines - update 2020–2022: bladder cancer. Prog Urol. 2020;30(12S):S78–S135. [DOI] [PubMed] [Google Scholar]
- 7.Bloudek L, Wright P, McKay C, et al. Systematic literature review (SLR) and network meta-analysis (NMA) of first-line therapies (1L) for locally advanced/metastatic urothelial carcinoma (la/mUC). Curr Oncol. 2023;30(4):3637–3647. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Sonpavde GP, Galsky MD, Wright P, et al. Real-world treatment patterns and clinical outcomes with first-line therapy in cisplatin-eligible and ineligible patients with advanced urothelial carcinoma. J Clin Oncol. 2022;40(16_suppl):4565. [Google Scholar]
- 9.Dreyer NA. Advancing a framework for regulatory use of real-world evidence: when real is reliable. Ther Innov Regul Sci. 2018;52(3):362–368. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Boudemaghe T, Belhadj I.. Data resource profile: the French national uniform hospital discharge data set database (PMSI). Int J Epidemiol. 2017;46(2):392–392d. [DOI] [PubMed] [Google Scholar]; • Profile of the Programme de Médicalisation des Systèmes d’Information (PMSI) database used in our analysis.
- 11.Bergot E, De Leotoing L, Bendjenana H, et al. Hospital burden of pulmonary arterial hypertension in France. PLOS ONE. 2019;14(9):e0221211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Palmaro A, Moulis G, Despas F, et al. Overview of drug data within French health insurance databases and implications for pharmacoepidemiological studies. Fundam Clin Pharmacol. 2016;30(6):616–624. [DOI] [PubMed] [Google Scholar]
- 13.Bannay A, Chaignot C, Blotière P-O, et al. The best use of the Charlson comorbidity index with electronic health care database to predict mortality. Med Care. 2016;54(2):188–194. [DOI] [PubMed] [Google Scholar]
- 14.Bilan démographique 2020 [2020 demographic report] [Internet]. Montrouge, France: Institut national de la statistique et des études économiques; 2021. [cited 2022 Jul 8]. Available from: https://www.insee.fr/fr/statistiques/5007688?sommaire=5007726 [Google Scholar]
- 15.Kearney M, Zhang L, Hubscher E, et al. First-line treatment patterns among patients with locally advanced or metastatic urothelial cancer (la/mUC): a systematic literature review. In: Poster session presented at: ISPOR Europe 2022; Vienna, Austria. 2022. Nov 6–9. [Google Scholar]; • Systematic literature review on real-world first-line treatment patterns and survival among patients with locally advanced or metastatic UC.
- 16.KEYTRUDA - Avis de la commission de la Transparence adopté le 21 février 2018 [KEYTRUDA - Opinion of the Transparency Commission adopted on February 21, 2018] Internet]. Paris, France: Haute Autorité de santé; 2018. [cited 2020 Nov 20]. French. Available from: https://www.has-sante.fr/upload/docs/evamed/CT-16530_KEYTRUDA_carcinome_urothelial_PIC_EI_Avis3_CT16530.pdf [Google Scholar]; • The February 2018 Commission de la Transparence opinion for pembrolizumab included an algorithm used to identify patients eligible for pembrolizumab. This algorithm was the basis for the one used in our analysis.
- 17.Milloy N, Kirker M, Unsworth M, et al. Real-world analysis of treatment patterns and platinum-based treatment eligibility of patients with metastatic urothelial cancer in 5 European countries. Clin Genitourin Cancer. 2024;22(1):e136–e147. [DOI] [PubMed] [Google Scholar]
- 18.Avis sur les médicaments: enfortumab vedotin PADCEV 20 mg et 30 mg, poudre pour solution à diluer pour perfusion. Première évaluation [Drug reviews: enfortumab vedotin PADCEV 20 mg and 30 mg, powder for solution for dilution for infusion. First assessment]. French. Paris, France: Haute Autorité de santé; 2022. Dec 7. https://www.has-sante.fr/upload/docs/evamed/CT-19842_PADCEV_PIC_INS_AvisDef_CT19842.pdf [Google Scholar]
- 19.Décision n° 2022.0467/DC/SEM du 8 décembre 2022 du collège de la Haute Autorité de santé portant renouvellement de l’autorisation d’accès précoce de la spécialité PADCEV [Decision No. 2022.0467/DC/SEM of December 8, 2022, of the college of the Haute Autorité de santé renewing the authorization for early access to the PADCEV specialty] [Internet]. Paris, France: Haute Autorité de santé; 2022. [cited 2023 Mar 15]. Available from: https://www.has-sante.fr/upload/docs/application/pdf/2022-12/padcev_ap136_decision_et_avisct_renouvellement_aa.pdf [Google Scholar]
- 20.Powles TB, Perez Valderrama B, Gupta S, et al. LBA6 EV-302/KEYNOTE-A39: open-label, randomized phase III study of enfortumab vedotin in combination with pembrolizumab (EV+P) vs chemotherapy (chemo) in previously untreated locally advanced metastatic urothelial carcinoma (la/mUC). Ann Oncol. 2023;34:S1340. [Google Scholar]
- 21.European Commission approves Astellas’ . PADCEVTM (enfortumab vedotin) in combination with KEYTRUDA® (pembrolizumab) for first-line treatment of advanced urothelial cancer [Internet]. Tokyo, Japan: Astellas Pharma Inc; 2024. Aug 28 [cited 2024 Sep 11]. Available from: https://www.astellas.com/en/news/29371 [Google Scholar]
- 22.Avis sur les médicaments . Enfortumab vedotin PADCEV 20 mg et 30 mg, poudre pour solution à diluer pour perfusion. Accès précoce post-AMM [Drug reviews: Enfortumab vedotin PADCEV 20 mg and 30 mg, powder for solution for dilution for infusion. Early access post-marketing authorization]. French. Paris, France: Haute Autorité de santé; 2024. Sep 4. https://ansm.sante.fr/tableau-acces-derogatoire/impression/317 [Google Scholar]
- 23.Leal J, Luengo-Fernandez R, Sullivan R, et al. Economic burden of bladder cancer across the European Union. Eur Urol. 2016;69(3):438–447. [DOI] [PubMed] [Google Scholar]
- 24.Poulalhon C, Rotelli-Bihet L, Moine S, et al. Use of hospital palliative care according to the place of death and disease one year before death in 2013: a French national observational study. BMC Palliat Care. 2018;17:75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Carrigan G, Whipple S, Taylor MD, et al. An evaluation of the impact of missing deaths on overall survival analyses of advanced non–small cell lung cancer patients conducted in an electronic health records database. Pharmacoepidemiol Drug Saf. 2019;28:572–581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.KEYTRUDA avis d’efficience, carcinome urothélial (2ème ligne) [Internet]. Paris, France: Haute Autorité de Santé; 2018. [cited 2023 Sep 28]. Available from: https://www.has-sante.fr/upload/docs/application/pdf/2018-10/keytruda_10072018_avis_efficience.pdf [Google Scholar]; • In July 2018, the Commission Evaluation Economique et de Santé Publique (CEESP) validated the locally advanced or mUC population identification algorithm used in the February 2018 Haute Autorité de santé (HAS) opinion (which was the basis of our identification algorithm) and included an estimated incidence of locally advanced or mUC in France in 2014 based on PMSI data.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Researchers may request access to anonymized participant-level data, trial-level data, and protocols from Astellas sponsored clinical trials at www.clinicalstudydatarequest.com. For the Astellas criteria on data sharing, see https://clinicalstudydatarequest.com/Study-Sponsors/Study-Sponsors-Astellas.aspx.