Summary
Background
Immunotherapy with atezolizumab and durvalumab has significantly advanced the treatment of extended-stage small cell lung cancer (esSCLC). We aimed at studying the real-world use and comparative effectiveness and safety of these treatments.
Methods
We utilised the French National Health Data System (SNDS) to include all patients treated with atezolizumab or durvalumab for esSCLC from May 1, 2019 to December 31, 2023, with follow-up until June 30, 2024. Baseline characteristics were described. Effectiveness (time to treatment discontinuation and overall survival) and safety (all-cause and cause-specific overnight hospitalisations) were assessed using Cox proportional hazards models.
Findings
A total of 8612 patients were included, with 5188 initiating atezolizumab and 3424 initiating durvalumab. The mean age at inclusion was 66·2 years (standard deviation 8·7), and most patients were men (63·8%, n = 5493). The median time to treatment discontinuation was 5·4 months (95% CI, 5·3–5·5) for atezolizumab and 5·5 months (95% CI, 5·4–5·6) for durvalumab, with an adjusted hazard ratio (aHR) of 0·97 (95% CI, 0·92–1·02) for durvalumab vs. atezolizumab. The median overall survival was 11·1 months (95% CI, 10·6–11·4) for atezolizumab and 11·4 months (95% CI, 10·9–11·9) for durvalumab, with an aHR of 0·93 (95% CI, 0·88–0·98) for durvalumab vs. atezolizumab.
While 61·8% (n = 5323) of patients received systemic steroids, testifying of immune-mediated side effects, no significant safety differences were observed between the groups.
Interpretation
Our study found no clinically relevant differences in effectiveness or safety between atezolizumab and durvalumab for first-line esSCLC. Both treatments appearing interchangeable, treatment decision could be based on drug availability, local institutional protocols, administrative organisation, or cost considerations.
Funding
None.
Keywords: Extended-stage small cell lung cancer, Immunotherapy, Atezolizumab, Durvalumab, Epidemiology, French National Health Data System (SNDS)
Research in context.
Evidence before this study
Extended-stage small cell lung cancer (esSCLC) is a highly aggressive form of lung cancer. While conventional chemotherapy had remained the standard treatment for over 30 years, the advent of immunotherapy marked a significant advancement in the management of esSCLC. In France, two PD-L1 inhibitors have been approved since 2019–2020: atezolizumab, in combination with carboplatin-etoposide, following the IMpower133 trial, and durvalumab, combined with platinum-etoposide, following the CASPIAN trial. Both treatment schemes showed overall survival benefits over platinum-etoposide treatment alone. Early access programs were granted based on these pivotal clinical results, subsequently leading to market authorisation. However, these treatments have not been studied in large-scale real-world settings, leaving clinicians with limited guidance on which treatment to initiate, as no comparisons have been conducted to date.
We conducted a PubMed search for peer reviewed papers published from inception to April 25, 2025, using the following search terms: ((PD-L1[Title/Abstract]) OR (atezolizumab[Title/Abstract]) OR (durvalumab[Title/Abstract])) AND ((extensive-stage small cell lung cancer[Title/Abstract]) OR (esSCLC[Title/Abstract])). The search yielded 280 records, including many case reports, meta-analyses, and clinical trials, but few observational studies. Most real-world studies were descriptive, focusing on the use of either atezolizumab or durvalumab. Only three studies actually compared the two treatment schemes, and those had limited sample sizes (n = 40–194) and included patients in not more than three hospital facilities.
Given the paucity of evidence and data on the use, comparative effectiveness, and safety of PD-L1 inhibitors in esSCLC, we felt it was important in terms of public health to carry out a large-scale real-life study in this population.
Added value of this study
This study provides the first large-scale description of the use of PD-L1 inhibitors for first-line treatment of extensive-stage small cell lung cancer. Small hospital facilities play a major role in the treatment initiation, to ensure a territorial network of healthcare provision, especially in the case of this highly deprived target population. This study is also the first comparison of atezolizumab versus durvalumab in this indication. Our findings demonstrate that durvalumab and atezolizumab are comparable in terms of both effectiveness and safety. Our results show that PD-L1 inhibitors appear less effective than in clinical trials, reflecting differences in patient selection. We observed median durations of treatment of five months, median overall survivals of eleven months, and a rising use of systemic steroids, along with PD-L1 infusions, testifying immune-mediated side effects.
Implications of all the available evidence
Immunotherapy represents a significant advancement in the treatment of patients with esSCLC, leading to improved overall survival. Real-world studies such as the one we performed are essential for public health purposes, as conventional therapeutic trials are never conducted to compare new medicines and indirect comparisons are not suitable. Although not identical, atezolizumab and durvalumab, combined with platinum-etoposide, show similar effectiveness and safety profiles, and thus appear to be interchangeable. The decision to initiate one product over the other could be based on drug availability, local institutional protocols, administrative organisation, or cost considerations. Future qualitative studies should focus on understanding the factors influencing the choice of PD-L1 inhibitors in clinical practice.
Introduction
Small cell lung cancer (SCLC) is a highly aggressive form of lung cancer. The global estimate for SCLC in 2015 is 260,000 new cases, with over 11,000 in Europe, and its aggressive nature and association with smoking make it a significant public health concern.1, 2, 3 In France, lung cancer was estimated to be responsible for 33,000 deaths in 2018, and SCLC accounts for approximately 15% of these cases.4 Patients experience a 5-year survival rate of only 5%. Extensive-stage SCLC (esSCLC), constitute two-thirds of all cases.5,6
Historically the first line treatment for esSCLC has been chemotherapy with etoposide and platinum-based drugs in the 1980s, with a limited survival outcome.7 With the recognition of SCLC as having a high mutation burden, significant advancements in immunotherapy have provided new treatment alternatives. In the late 2010s checkpoint inhibitors targeting PD-L1, such as atezolizumab and durvalumab, have been introduced in therapeutic arsenal, marking an important shift in the treatment and prognosis of esSCLC. These therapies, when combined with chemotherapy, have showed survival improvement in clinical trials. The IMpower133 trial showed that atezolizumab combined with carboplatin-etoposide led to a median overall survival (OS) improvement of 2 months (HR: 0·70, 95% CI: 0·54–0·91) and a progression-free survival (PFS) improvement of 0·9 months (HRs: 0·77, 95% CI: 0·62–0·96).8 Similarly, the CASPIAN trial showed that combining durvalumab with platinum-etoposide improved OS by 2·7 months compared to chemotherapy alone (hazard ratio [HR]: 0·73, 95% CI: 0·56–0·95).9
Immunotherapy thus represented an advancement in the treatment of esSCLC. In France, atezolizumab received temporary authorisation for use in May 2019 and marketing approval in September 2019, followed by durvalumab's approval in August 2020 after temporary use authorisation in March 2020.
Although clinical trials have established the efficacy of PD-L1 inhibitors in esSCLC, and integration of atezolizumab and durvalumab into clinical practice, comparisons of their effectiveness in the real-world clinical practice remain limited to small-scale observational studies.10, 11, 12
The objective of this study was to conduct a comprehensive real-world comparison of durvalumab and atezolizumab in the first-line treatments for esSCLC. Specifically, we aimed to (i) evaluate and compare treatment effectiveness between the two treatment groups, using overall survival and treatment duration as surrogate outcome, (ii) identify prognostic factors associated with effectiveness outcomes, and (iii) evaluate potential toxicities related to the use of these treatments.
Methods
Study design
This is a retrospective, observational cohort study, based on a national health data system. The study is reported following the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidance.
Data source
We used data from the French National Health Data System (SNDS). The SNDS covers almost the totality (>99%) of the French population—68 million residents. Each person is identified by a unique and anonymous number. The SNDS records comprehensive outpatient and inpatient reimbursements and also contains sociodemographic data (age, sex, place of residence, deprivation) and vital status, among others. Since 2019, the SNDS provides medical indications for hospital-level deliveries of expensive or early-access program drugs.13,14, 15, 16, 17, 18 We were thus able to identify patients treated with a PD-L1 inhibitor for esSCLC.
Study periods and population
Patients were included if they initiated treatment with either atezolizumab (ATC L01FF05) or durvalumab (ATC L01FF03) for esSCLC between 1 May 2019 and 31 December 2023. Patients with any prior exposure to either of these drugs were excluded (Supplementary Table S1). We restricted the analysis to patients with a history of lung cancer within 10 years before inclusion. Patients were followed until their death or 30 June 2024 (last date for which data were available at the time of analysis) (Supplementary Table S2). The first delivery of treatment defined the index date for the study participants.
Patients characteristics
Inpatient treatment induction was used as a proxy for ECOG performance status.4 Hospital facilities were ranked by the volume of patients included in the study, in three categories: high volume (top 10% centres); intermediate volume (10–40% centres); low volume (bottom 60% centres)4 (Supplementary Table S3).
Presence of metastases at baseline was defined as any hospital record with a discharge code related to metastasis in the past five years (Supplementary Table S4). Baseline medical history was assessed over the three years preceding inclusion in the study, including radiotherapy, lung surgery, number of intravenous chemotherapy sessions and immunosuppressive treatment. Baseline exposure to systemic steroids, antibiotics and proton-pump inhibitors19 were assessed within three months before index date. Baseline comorbid conditions were identified with the diseases and expenditures mapping tool developed by the French National Health Insurance and described elsewhere20,21 (Supplementary Tables S5–S7).
Outcomes
Treatment discontinuation was defined as a period of 81 days with no delivery of treatment (21-day theoretical coverage plus 60-day washout). The date of treatment discontinuation was defined as the last date of administration plus 21 days (or, if the patient died, the date of death). The time-to-treatment discontinuation (TTD), used as an imperfect proxy for progression-free survival (PFS) or unbearable toxicity, was defined as the time from the index date to the date of treatment discontinuation. Overall survival (OS) was defined as the time from the index date to death or the end date of the study. Severe adverse events were defined as all-cause (except cancer) overnight hospitalisations and cause-specific overnight hospitalisations (mainly for immune-related adverse event (ir-AE) and infections) during treatment (Supplementary Table S8). Immunosuppressive treatment during the follow-up was used as a proxy for ir-AE of treatment. We used hospitalisation as a proxy of severe potential related events while acknowledging that not all hospitalisations were directly attributable to checkpoint inhibitors therapy.
Statistical analysis
We used descriptive statistics to characterise patients at baseline, namely numbers and percentages for qualitative variables and mean and standard deviation and median and quartiles for quantitative variables. Chi2 and Student p-values were reported to highlight the differences between exposure groups.
Crude Kaplan Meier survival curves were used to estimate median TTD and OS, and percentages of survival at different time points. Univariate and multivariable Cox proportional hazard models were applied to estimate the hazard ratios of treatment discontinuation and death associated with the main exposure and other covariates. Inverse probability of treatment weighted (IPTW) Cox proportional hazard models were used to estimate the hazard ratios for the safety events.
Subgroup and sensitivity analyses
We carried out several subgroup analyses on time to treatment discontinuation and overall survival, in line with those performed in clinical trials: by age (<60 years, 60–69 years, ≥70 years); by sex; by presence or absence of brain metastasis. In addition, we completed the analyses of time to treatment discontinuation and overall survival with IPTW Cox proportional hazard models as statistical sensitivity analyses.
Ethics approval
The French National Health Data System (SNDS) is a strictly anonymous database set, including all mandatory health insurance reimbursement data of the French population, in particular data from the processing of healthcare reimbursements (electronic or paper treatment sheet) and data from health establishments (PMSI).
EPI-PHARE has permanent regulatory access to the data from the French National Health Data System (SNDS) via its constitutive bodies ANSM and CNAM, in application of the provisions of the French Decree No. 2016–1871 of December 26, 2016 relating to the processing of personal data called the “National Health Data System”15, the French law articles Art. R. 1461–1316 and R. 1461–1417 from the French Public Health Code and the French Data Protection Authority (CNIL) decision CNIL-2016-316·18.
In accordance with the permanent regulatory access granted to EPI-PHARE via ANSM and CNAM, this work did not require any specific opinion from the French Ethical and Scientific Committee for Research, Studies and Evaluations health (CESREES) nor approval from the CNIL.
All requests in the database were made by duly authorised people.
The study was registered on the study register of EPI-PHARE concerning studies from SNDS data under the reference T-2025-01-542.
Role of the funding source
The authors are employees of the French National Health Insurance Centre (CNAM), the French National Agency for Medicines and Health Product Safety (ANSM), and the Avec Groupe Hospitalier Mutualiste and received no specific funding for this study.
Results
The study flow chart is presented in Supplementary Figure S1.
Patients characteristics at baseline
A total of 8612 patients were included in the study, 5188 patients initiating atezolizumab and 3424 initiating durvalumab. Baseline characteristics were similar across treatment groups. The mean age at inclusion was 66·2 years (standard deviation, SD 8·7). Most patients were men (n = 5493, 63·8%), had been diagnosed with lung cancer within the past year (n = 7663, 89·0%), and had no prior chemotherapy (0–1 session, n = 6264, 72·7%). We observed a deprivation gradient among patients receiving a PD-L1 inhibitor, with 13·9% (n = 1194) in the least deprived quintile of French residents and 24·3% (n = 2089) in the most deprived quintile. At baseline, 18·5% (a total of 1591 patients) had received at least the equivalent of 1 g of predisone within 3 months before inclusion; 22·0% (n = 1895) had brain metastases; 39·4% (n = 3391) had a cardiac condition; 19·5% (n = 1676) had diabetes; 6·0% (n = 518) had an autoimmune disease; and 11·0% (n = 950) were diagnosed with an alcoholic abuse disorder. The only differences between the groups concerned the treatment delivery setting. Most patients (n = 5977, 69·4%) were treated in small facilities (i.e., local, for-profit and non-profit hospitals), compared with university, regional or specialised centres. Atezolizumab was more often administered in low-volume facilities (n = 1255, 24·2%), especially in local hospitals and private institutions, whereas durvalumab was mainly initiated in high-volume facilities (n = 1420, 41·5%), particularly in university or regional hospitals. Initiation as part of a full inpatient hospitalisation occurred more frequently with durvalumab than with atezolizumab (n = 2186, 63·8% vs. n = 2978, 57·4%) (Table 1).
Table 1.
Patients characteristics at baseline.
| Durvalumab | Atezolizumab | All | p-value | |
|---|---|---|---|---|
| Number of patients | 3424 | 5188 | 8612 | |
| Year of treatment initiation | <0·01 | |||
| 2019 | 2 (0·1) | 118 (2·3) | 120 (1·4) | |
| 2020 | 122 (3·6) | 1424 (27·4) | 1546 (18·0) | |
| 2021 | 726 (21·2) | 1347 (26·0) | 2073 (24·1) | |
| 2022 | 1158 (33·8) | 1152 (22·2) | 2310 (26·8) | |
| 2023 | 1416 (41·4) | 1147 (22·1) | 2563 (29·8) | |
| Mean age (SD) | 66·1 (8·8) | 66·2 (8·6) | 66·2 (8·7) | 0·39 |
| Median age (IQR) | 66·0 (60·0–72·0) | 67·0 (61·0–72·0) | 67·0 (60·0–72·0) | |
| Age categories | 0·35 | |||
| <50 | 116 (3·4) | 165 (3·2) | 281 (3·3) | |
| 50–59 | 656 (19·2) | 924 (17·8) | 1580 (18·3) | |
| 60–69 | 1425 (41·6) | 2173 (41·9) | 3598 (41·8) | |
| ≥70 | 1227 (35·8) | 1926 (37·1) | 3153 (36·6) | |
| Sex: male | 2172 (63·4) | 3321 (64·0) | 5493 (63·8) | 0·60 |
| Deprivation index | 0·12 | |||
| 1 (least deprived) | 505 (14·7) | 689 (13·3) | 1194 (13·9) | |
| 2 | 599 (17·5) | 899 (17·3) | 1498 (17·4) | |
| 3 | 699 (20·4) | 1033 (19·9) | 1732 (20·1) | |
| 4 | 737 (21·5) | 1151 (22·2) | 1888 (21·9) | |
| 5 (most deprived) | 815 (23·8) | 1274 (24·6) | 2089 (24·3) | |
| Missing/not estimated | 69 (2·0) | 142 (2·7) | 211 (2·5) | |
| Type of hospital facility | <0·01 | |||
| University/regional hospital | 978 (28·6) | 894 (17·2) | 1872 (21·7) | |
| Specialised cancer centre | 303 (8·8) | 460 (8·9) | 763 (8·9) | |
| Local hospital | 1369 (40·0) | 2469 (47·6) | 3838 (44·6) | |
| Private facility | 552 (16·1) | 1051 (20·3) | 1603 (18·6) | |
| Non-profit private facility | 222 (6·5) | 314 (6·1) | 536 (6·2) | |
| Size of hospital facility | <0·01 | |||
| High volume (top 10% facilities) | 1420 (41·5) | 1478 (28·5) | 2898 (33·7) | |
| Intermediate volume (top 10–40%) | 1410 (41·2) | 2455 (47·3) | 3865 (44·9) | |
| Low volume (bottom 60%) | 594 (17·3) | 1255 (24·2) | 1849 (21·5) | |
| Inpatient initiation | 2186 (63·8) | 2978 (57·4) | 5164 (60·0) | <0·01 |
| Duration of lung cancer | 0·56 | |||
| ≤1 years | 3051 (89·1) | 4612 (88·9) | 7663 (89·0) | |
| 4 years | 301 (8·8) | 480 (9·3) | 781 (9·1) | |
| >4 years | 72 (2·1) | 96 (1·9) | 168 (2·0) | |
| Cancer treatments in the past 3 years | ||||
| Lung surgery | 115 (3·4) | 200 (3·9) | 315 (3·7) | 0·25 |
| Radiotherapy | 748 (21·8) | 1031 (19·9) | 1779 (20·7) | 0·03 |
| Sessions of chemotherapy | 0·03 | |||
| 0–1 | 2529 (73·9) | 3735 (72·0) | 6264 (72·7) | |
| 2–5 | 410 (12·0) | 637 (12·3) | 1047 (12·2) | |
| 6–9 | 169 (4·9) | 266 (5·1) | 435 (5·1) | |
| 10–13 | 166 (4·8) | 243 (4·7) | 409 (4·7) | |
| ≥14 | 150 (4·4) | 307 (5·9) | 457 (5·3) | |
| Immunosuppressive treatment in the past 3 years | 45 (1·3) | 57 (1·1) | 102 (1·2) | 0·42 |
| Cumulative dose of systemic steroids in the past 3 months, prednisone equivalent | 0·02 | |||
| 0 g | 1647 (48·1) | 2603 (50·2) | 4250 (49·3) | |
| [0–1] g | 1115 (32·6) | 1656 (31·9) | 2771 (32·2) | |
| [1–2] g | 342 (10·0) | 419 (8·1) | 761 (8·8) | |
| [2–3] g | 184 (5·4) | 279 (5·4) | 463 (5·4) | |
| >3 g | 136 (4·0) | 231 (4·5) | 367 (4·3) | |
| Antibiotics–at least 2 deliveries in the past 3 months | 748 (21·8) | 1023 (19·7) | 1771 (20·6) | 0·02 |
| Proton-pump inhibitors–at least 2 deliveries in the past 3 months | 1072 (31·3) | 1602 (30·9) | 2674 (31·0) | 0·69 |
| Metastases in the past 5 years | 2482 (72·5) | 3767 (72·6) | 6249 (72·6) | 0·92 |
| Lymph node | 1050 (30·7) | 1520 (29·3) | 2570 (29·8) | 0·18 |
| Bone | 846 (24·7) | 1275 (24·6) | 2121 (24·6) | 0·91 |
| Digestive | 1104 (32·2) | 1634 (31·5) | 2738 (31·8) | 0·48 |
| Brain | 775 (22·6) | 1120 (21·6) | 1895 (22·0) | 0·26 |
| Surrenal | 454 (13·3) | 648 (12·5) | 1102 (12·8) | 0·31 |
| Other | 222 (6·5) | 351 (6·8) | 573 (6·7) | 0·64 |
| Duration of metastatic disease | 0·28 | |||
| ≤1 year | 2305 (67·3) | 3460 (66·7) | 5765 (66·9) | |
| 1–4 years | 118 (3·4) | 222 (4·3) | 340 (3·9) | |
| >4 years | 59 (1·7) | 85 (1·6) | 144 (1·7) | |
| Missing | 942 (27·5) | 1421 (27·4) | 2363 (27·4) | |
| Comorbid condition | 3119 (91·1) | 4681 (90·2) | 7800 (90·6) | 0·19 |
| Cardiac | 1351 (39·5) | 2040 (39·3) | 3391 (39·4) | 0·92 |
| Diabetes | 678 (19·8) | 998 (19·2) | 1676 (19·5) | 0·54 |
| Respiratory | 1571 (45·9) | 2246 (43·3) | 3817 (44·3) | 0·02 |
| Demence/Parkinson | 34 (1·0) | 43 (0·8) | 77 (0·9) | 0·50 |
| Other neurologic | 92 (2·7) | 103 (2·0) | 195 (2·3) | 0·04 |
| Psychiatric | 304 (8·9) | 449 (8·7) | 753 (8·7) | 0·75 |
| Autoimmune diseases | 231 (6·7) | 287 (5·5) | 518 (6·0) | 0·02 |
| Renal | 5 (0·1) | 10 (0·2) | 15 (0·2) | 0·81 |
| Liver/pancreas | 189 (5·5) | 273 (5·3) | 462 (5·4) | 0·64 |
| Anti-hypertensive treatment | 1792 (52·3) | 2688 (51·8) | 4480 (52·0) | 0·65 |
| Lipid-lowering treatment | 1387 (40·5) | 2058 (39·7) | 3445 (40·0) | 0·45 |
| Psychiatric treatment | 1041 (30·4) | 1630 (31·4) | 2671 (31·0) | 0·33 |
| Opioids | 27 (0·8) | 53 (1·0) | 80 (0·9) | 0·32 |
| Obesity | 272 (7·9) | 449 (8·7) | 721 (8·4) | 0·26 |
| Alcohol abuse | 388 (11·3) | 562 (10·8) | 950 (11·0) | 0·49 |
| Tobacco use | 1897 (55·4) | 2766 (53·3) | 4663 (54·1) | 0·06 |
Figures are presented in numbers (percentages), unless stated otherwise.
SD: standard deviation; IQR: interquartile range.
Time to treatment discontinuation and overall survival
Median follow-up was similar between the groups, at 10·0 months (IQR 5·7–17·8) in the atezolizumab group and 9·6 months (IQR 5·8–16·2) in the durvalumab group. A total of 5819 patients (67·6%) had received the induction protocol, namely 4 injections of anti-PD-L1 with chemotherapy within the first 3 months of follow-up (Table 2).
Table 2.
Patients characteristics during the follow-up.
| Durvalumab | Atezolizumab | All | |
|---|---|---|---|
| Number of patients | 3424 | 5188 | 8612 |
| Mean follow-up (SD) | 12·0 (9·1) | 13·8 (11·9) | 13·1 (10·9) |
| Median follow-up (IQR) | 9·6 (5·8–16·2) | 10·0 (5·7–17·8) | 9·8 (5·8–17·2) |
| Treatment discontinuation | 2986 (87·2) | 4769 (91·9) | 7755 (90·0) |
| Median time to treatment discontinuation (95% CI) | 5·5 (5·4–5·6) | 5·4 (5·3–5·5) | 5·4 (5·4–5·5) |
| Death during follow-up | 2281 (66·6) | 3930 (75·8) | 6211 (72·1) |
| Median overall survival (95% CI) | 11·4 (10·9–11·9) | 11·1 (10·6–11·4) | 11·2 (10·9–11·4) |
| Survival rates (%, 95% CI) | |||
| At 6 months | 73·4 (72·0–74·9) | 73·6 (72·4–74·8) | 73·5 (72·6–74·5) |
| At 12 months | 47·9 (46·2–49·6) | 46·3 (44·9–47·7) | 46·9 (45·8–48·0) |
| At 18 months | 34·2 (32·5–36·1) | 31·4 (30·1–32·7) | 32·4 (31·4–33·5) |
| At 24 months | 25·3 (23·6–27·2) | 23·4 (22·2–24·7) | 24·2 (23·1–25·2) |
| At 36 months | 17·7 (15·9–19·7) | 16·3 (15·2–17·5) | 16·8 (15·8–17·8) |
| Treatments during the follow-up | |||
| Induction (4 PD-L1 injections within 3 months after initiation) | 2302 (67·2) | 3517 (67·8) | 5819 (67·6) |
| Mean number of injections (SD) | 8·0 (7·0) | 9·7 (10·7) | 9·0 (9·4) |
| Median number of injections (IQR) | 6·0 (4·0–9·0) | 7·0 (4·0–10·0) | 7·0 (4·0–10·0) |
| Immunosuppressive treatment during the follow-up | 34 (1·0) | 87 (1·7) | 121 (1·4) |
| Cumulative dose of systemic steroids within 6 months after treatment initiation, in prednisone equivalent | |||
| 0 g | 1296 (37·9) | 1993 (38·4) | 3289 (38·2) |
| [0–1] g | 886 (25·9) | 1367 (26·3) | 2253 (26·2) |
| [1–2] g | 482 (14·1) | 737 (14·2) | 1219 (14·2) |
| [2–3] g | 331 (9·7) | 448 (8·6) | 779 (9·0) |
| >3 g | 429 (12·5) | 643 (12·4) | 1072 (12·4) |
| Topotecan/irinotecan during follow-up | 54 (1·6) | 146 (2·8) | 200 (2·3) |
| PD-L1 switch during the follow-up | 25 (0·7) | 30 (0·6) | 55 (0·6) |
Figures are presented in numbers (percentages), unless stated otherwise.
SD: standard deviation; IQR: interquartile range; 95% CI: 95% confidence interval.
The median time to treatment discontinuation (mTTD) was 5·4 months (95% confidence interval CI, 5·3–5·5) in the atezolizumab group and 5·5 months (95% CI, 5·4–5·6) in the durvalumab group (Fig. 1a). The adjusted hazard ratio (aHR) of treatment discontinuation associated with durvalumab exposure relative to atezolizumab exposure was 0·97 (95% CI, 0·92–1·02) (Fig. 2, Supplementary Table S9). Treatment discontinuation at the end of follow-up concerned 91·9% (n = 4769) of atezolizumab users and 87·2% (n = 2986) of durvalumab users (Table 2).
Fig. 1.
Crude Kaplan–Meier survival curves comparing atezolizumab and durvalumab in terms of overall survival (a) and treatment discontinuation (b).
Fig. 2.
Hazard ratios for overall survival (a) and treatment discontinuation (b) associated with the exposure (durvalumab initiation as compared to atezolizumab initiation) and the main variables included in the Cox model. ∗In the past 3 years; ∗∗In the past 3 months, converted in predisone equivalent; §In the past 5 years. TTD: time to treatment discontinuation; OS: overall survival; aHR: adjusted hazard ratio; 95% CI: 95% confidence interval. The model was additionally adjusted on year of treatment initiation, radiotherapy, antibiotics (≥2 deliveries in the past 3 months), proton-pump inhibitors (≥2 deliveries in the past 3 months) and duration of metastatic disease (See Supplementary Table S9). Reference categories are indicated by a hazard ratio of 1. For metastasis sites and comorbid conditions, the reference category was the absence of the specific metastasis or comorbidity.
The median overall survival (mOS) was 11·1 months (95% CI, 10·6–11·4) in the atezolizumab group versus 11·4 months (95% CI, 10·9–11·9) in the durvalumab group, with 3930 deaths (75·8%) in the atezolizumab group and 2281 deaths (66·6%) in the durvalumab group (Fig. 1b). The adjusted hazard ratio (aHR) of overall survival associated with durvalumab exposure relative to atezolizumab exposure was 0·93 (95% CI, 0·88–0·98). Advanced age, male sex, deprivation, being hospitalised for treatment administration, having received previous chemotherapy sessions, having received high dose steroids, having metastases at baseline from various localisations and presenting comorbidities at treatment initiation were associated with a higher risk of death (Fig. 2, Supplementary Table S9).
Overall survival rates at 12, 18, 24, and 36 months were 46·3% (95% CI, 44·9–47·7), 31·4% (95% CI, 30·1–32·7), 23·4% (95% CI, 22·2–24·7), and 16·3% (95% CI, 15·2–17·5), respectively in the atezolizumab group and 47·9% (95% CI, 46·2–49·6), 34·2% (95% CI, 32·5–36·1), 25·3% (95% CI, 23·6–27·2), and 17·7% (95% CI, 15·9–19·7), respectively, in the durvalumab group (Table 2).
Subgroup and sensitivity analyses were consistent with the main results (Supplementary Figure S2, Supplementary Table S10).
Safety
Within the first 6 months of follow-up, 5323 patients (61·8%) received at least one delivery of systemic steroids, and 1072 (12·4%) received a cumulative dose of at least 3 g of prednisone equivalent, which might testify of immune-related adverse events among anti-PD-L1 users.
While treated, 4408 patients, representing 51·2% of the patients having initiated an anti-PD-L1 (51·5%, n = 2674 in the atezolizumab group; 50·6%, n = 1734 in the durvalumab group), have had at least one treatment-emergent serious adverse event (SAE) requiring overnight hospitalisation (all causes except cancer related hospitalisation). Among them, 2312 patients (26·9%) were hospitalised for a selected potential ir-AE: 1408 patients (27·1%) in the atezolizumab group and 904 (26·4%) in the durvalumab group. All-cause and potential immune-related adverse events rates did not differ between atezolizumab users and durvalumab users (respective weighted HR, wHR, with atezolizumab as a reference, 0·97 [95% CI, 0·91–1·04] and 0·95 [95% CI, 0·87–1·05]). Severe infection was encountered in 1125 patients (13·1%), including 645 (12·4%) in the atezolizumab group and 480 (14·0%) in the durvalumab group, again with no difference between exposures (wHR 1·06 [95% CI, 0·92–1·21]) (Table 3, Supplementary Table S11).
Table 3.
Hazard ratios for the main safety events associated with durvalumab exposure, as compared to atezolizumab exposure.
| Event | Number of events (%), atezolizumab | Number of events (%), durvalumab | Number of events (%), all | Incidence per 100 person-years, atezolizumab | Incidence per 100 person-years, durvalumab | Incidence per 100 person-years, all | Weighted hazard ratio [95% confidence interval] |
|---|---|---|---|---|---|---|---|
| Immunosuppressive treatment | 45 (0·87) | 23 (0·67) | 68 (0·79) | 1·24 | 1·03 | 1·16 | 0·91 [0·50–1·63] |
| All-cause hospitalisation | 2674 (51·54) | 1734 (50·64) | 4408 (51·18) | 105·56 | 110·82 | 107·57 | 0·97 [0·91–1·04] |
| Serious infection | 645 (12·43) | 480 (14·02) | 1125 (13·06) | 19·03 | 23·26 | 20·63 | 1·06 [0·92–1·21] |
| Immune-specific hospitalisation | 1408 (27·14) | 904 (26·4) | 2312 (26·85) | 45·78 | 47·78 | 46·55 | 0·95 [0·87–1·05] |
| Ischaemic heart diseases | 31 (0·6) | 39 (1·14) | 70 (0·81) | 0·85 | 1·76 | 1·20 | 1·91 [1·12–3·24] |
| Pulmonary circulation diseases | 75 (1·45) | 36 (1·05) | 111 (1·29) | 2·08 | 1·62 | 1·90 | 0·80 [0·50–1·27] |
| Pericardial diseases | 28 (0·54) | 17 (0·5) | 45 (0·52) | 0·77 | 0·76 | 0·77 | 1·00 [0·50–2·02] |
| Non-alcoholic liver diseases | 13 (0·25) | 22 (0·64) | 35 (0·41) | 0·36 | 0·99 | 0·60 | 3·27 [1·51–7·04] |
| Acute pancreatitis | 24 (0·46) | 5 (0·15) | 29 (0·34) | 0·66 | 0·22 | 0·49 | 0·24 [0·09–0·66] |
| Aplasia | 514 (9·91) | 320 (9·35) | 834 (9·68) | 14·97 | 15·17 | 15·05 | 0·91 [0·77–1·06] |
| Coagulation anomalies | 51 (0·98) | 40 (1·17) | 91 (1·06) | 1·41 | 1·80 | 1·56 | 1·03 [0·66–1·62] |
| Other blood diseases | 40 (0·77) | 24 (0·7) | 64 (0·74) | 1·10 | 1·08 | 1·09 | 0·95 [0·54–1·65] |
| Chronic lower respiratory diseases | 104 (2) | 79 (2·31) | 183 (2·12) | 2·88 | 3·59 | 3·15 | 1·05 [0·75–1·47] |
| Respiratory conditions due to other external agents | 17 (0·33) | 9 (0·26) | 26 (0·3) | 0·47 | 0·40 | 0·44 | 0·91 [0·39–2·14] |
| Interstitial tissue diseases | 27 (0·52) | 12 (0·35) | 39 (0·45) | 0·74 | 0·54 | 0·66 | 0·58 [0·27–1·24] |
| Pleural conditions | 122 (2·35) | 66 (1·93) | 188 (2·18) | 3·38 | 2·99 | 3·23 | 0·78 [0·55–1·11] |
| Other respiratory diseases | 145 (2·79) | 76 (2·22) | 221 (2·57) | 4·02 | 3·43 | 3·80 | 0·76 [0·55–1·05] |
| Diabetes mellitus | 35 (0·67) | 32 (0·93) | 67 (0·78) | 0·96 | 1·44 | 1·15 | 1·64 [0·95–2·83] |
| Diseases of other endocrine glands | 47 (0·91) | 31 (0·91) | 78 (0·91) | 1·30 | 1·39 | 1·33 | 0·86 [0·52–1·41] |
| Tubulo-interstitial diseases | 36 (0·69) | 14 (0·41) | 50 (0·58) | 0·99 | 0·63 | 0·85 | 0·50 [0·26–0·96] |
| Renal failure | 21 (0·4) | 15 (0·44) | 36 (0·42) | 0·58 | 0·67 | 0·61 | 0·99 [0·48–2·05] |
| Inflammatory diseases of the central nervous system | 25 (0·48) | 14 (0·41) | 39 (0·45) | 0·69 | 0·63 | 0·66 | 0·77 [0·37–1·62] |
| Other nervous system disorders | 47 (0·91) | 40 (1·17) | 87 (1·01) | 1·29 | 1·80 | 1·49 | 1·38 [0·85–2·22] |
Not all hospitalisations were directly attributable to checkpoint inhibitor therapy. Immunosuppressive treatments include methotrexate, thalidomide, etanercept, infliximab, adalimumab, certolizumab, golimumab, cyclophosphamide, and azathioprine.
Boldface entries indicate statistical significance.
Discussion
Our study represents the largest real-world analysis to date of PD-L1 inhibitors use in first line esSCLC with a comparison between atezolizumab and durvalumab including 8612 patients.
Our findings demonstrate a widespread and increasing utilisation of PD-L1 inhibitors for the treatment of first line esSCLC. Characteristics of patients at treatment initiation reflect real-world differences with clinical trials. Mean age at inclusion was 66·2 years, slightly older than the median ages reported in clinical trials such as CASPIAN (64 years)9 and IMpower133 (63 years)8; Additionally, we observed a higher frequency of comorbid conditions at baseline, with 39·4% of patients having cardiac diseases, 19·5% diabetes and 6% an autoimmune diseases. This latest condition was an exclusion criterion in clinical trials.
The observed socioeconomic gradient among patients initiating PD-L1 inhibitors for esSCLC in this study highlights significant but known disparities. Patients from the most deprived quintile accounted for 24·3% of the cohort compared to 13·9% from the least deprived quintile. This result suggests that socioeconomic deprivation may not only influence access to treatment but confirmed that it is associated with a higher burden of esSCLC itself, mediated by tobacco use.2,10,22
The mTTD and mOS were similar between the atezolizumab and durvalumab groups, but lower than the ones shown in clinical trials: median OS was 11·1 months for atezolizumab versus 11·4 months for durvalumab, with aHR indicating no clinically significant difference. Although the mOS were lower in our study, our findings show similar figures to those observed in the CASPIAN and IMpower133, where mOS ranged from 12·3 to 13·0 months.8,9
Previous observational studies evaluating PD-L1 inhibitors in esSCLC have been limited by small sample sizes with cohorts ranging from 40 to 367 patients. Only a few have compared atezolizumab and durvalumab.10, 11, 12 One study on 100 patients treated with durvalumab and 43 with atezolizumab reported a significantly longer survival with durvalumab (22·0 vs. 10·0 months; p = 0·03).12 However, the absence of adjustment for baseline characteristics, such as comorbid conditions and deprivation factors, could limit the interpretability due to the selection bias.
In our study, the aHR of TTD associated with durvalumab exposure, compared to atezolizumab exposure, was 0·97 (95% CI, 0·92–1·02), and the aHR of OS was 0·93 (95% CI, 0·88–0·98). Although the difference in OS was statistically significant, the effect size remained very modest. After clinical review and discussion with oncology experts, we concluded that this difference was not substantial and did not suggest a clinically meaningful superiority of one treatment over the other.
In clinical trials, the benefits of theses PD-L1 inhibitors over chemotherapy alone were characterised by relatively modest absolute gains in overall survival, but showed an increase in long-term survival rates: CASPIAN trial showed a 36-months OS of 17·6% with durvalumab compared to 5·8% for chemotherapy alone.23 Similarly, the IMpower133 trial reported an 18-month OS of 34% for atezolizumab versus 21% for chemotherapy.24 Our real-world analysis were in line with these clinical trial outcomes (18-month OS rate of 31·4% for atezolizumab and a 36-month OS rate of 17·7% for durvalumab).
In terms of safety, high-dose steroid use during anti-PD-L1 treatment (>3 g prednisone equivalent) was observed in 12·4% of patients, reflecting the burden of immune related adverse events (ir-AEs) in routine practice and aligning with trial-reported ir-AE rates. In exploratory safety analyses, a few statistically significant associations between the primary exposure and safety events were observed. However, these associations concerned small numbers of events and showed opposing directions of risk, considering them to be no clinically relevant.
Although university and specialised cancer centres manage the highest number of patients on a per-institution basis, a non-negligible proportion of patients are treated in smaller public or private local hospitals. These smaller facilities play a major role in ensuring treatment accessibility, particularly for patients living in socio-economically deprived areas with limited infrastructure. While public authorities tend to concentrate health services in larger institutions, greater resources should be allocated to smaller local hospitals to better meet treatment needs.
The primary strength of our study was the large population size, made possible by the comprehensive nature of the SNDS database, which covers the entire French population. It provides complete information on hospitalisations, dates of death, and treatment deliveries, making it possible to estimate time to treatment discontinuation and overall survival accurately. Furthermore, we had access to the medical indications documented for each treatment delivery, enabling us to identify patients treated specifically for first line esSCLC.
The main limitations of our study lie in the inability to account for unmeasured variables such as the type of chemotherapy infusions (which are included in the hospital stay package and thus untraceable in the SNDS), performance status or esSCLC subtype. However, the baseline characteristics of the two groups were similar, which may help to reduce the risk of residual confounding. In this real-world study, based on the SNDS, we did not have access to the necessary clinical or imaging data for assessing progression and evaluation of the clinical endpoints, such as progression-free survival. Consequently, we used the time-to-treatment discontinuation, a commonly used surrogate endpoint in real-world studies, which has been shown to correlate well with PFS in several large analyses.25
The SNDS database does not provide detailed data to distinguish between conventional platinum chemotherapies (cisplatin vs. carboplatin, used in combination with the PD-L1 inhibitor). However this is unlikely to affect our results significantly, as a meta-analysis showed no difference in efficacy between cisplatin and carboplatin in SCLC, although differences in toxicity profiles were observed.26
In addition, the performance status, a crucial prognostic factor in esSCLC commonly used in clinical trials and practice to assess functional capacity and inform treatment choices, was not directly measured in our dataset. Instead, initiation of a first-line PD-L1 inhibitor during inpatient hospitalisation likely acted as a proxy for poorer baseline health, given that hospitalised patients often have more advanced disease or significant comorbidities. Our proxy indeed correlated with prognosis, aligning with prior research.4
Another potential limitation of our study is the lack of information on esSCLC subtypes. Atezolizumab was shown to be more effective in the inflamed subtype of SCLC.27 However, it is important to note that currently, in routine clinical practice, subtypes are not typically determined before treatment by immunotherapy. Therefore, while this limitation may impact the granularity of our findings, it likely does not affect the general applicability of our results to standard clinical practice. Finally, we did not adjust for multiple comparisons in safety analyses; therefore, these findings must be interpreted with caution.
This study represents the first large-scale comparison of atezolizumab versus durvalumab, combined with chemotherapy, in the treatment of first-line extensive-stage small cell lung cancer (esSCLC), addressing a significant gap in the existing evidence. Our findings demonstrate that durvalumab and atezolizumab are comparable in terms of both effectiveness and safety. Although not identical, the two molecules appear to be interchangeable, and the decision to initiate one product over the other could be based on drug availability, local institutional protocols, administrative organisation, or cost considerations.
Contributors
Literature search: HJ and NH. Study design: HJ and NH. Data extraction and collection: HJ and DD. Statistical analysis: HJ. Interpretation of the results: HJ, NA, AM, and NH. Drafting of the manuscript: HJ, NA, AM, and NH. Revision of the manuscript: all. Supervision: NH and MZ.
HJ, NH, and MZ confirm that they had full access to all the data and verified the data. HJ, NH, and MZ take responsibility for the integrity of the data and the accuracy of the data analysis. HJ, NH, and MZ take final responsibility for the decision to submit the manuscript for publication.
Data sharing statement
In accordance with data protection legislation and the French regulation, the authors are not allowed to release or make public the data from the SNDS. However, any person or structure, public or private, for-profit or non-profit, is able to access SNDS data in order to carry out a study, research or an evaluation in the public interest, upon authorisation from the French Data Protection Office (CNIL), via the French Health Data Hub (https://www.health-data-hub.fr/).
All information for data subject is available on the EPI-PHARE website at https://www.epi-phare.fr/en/regulation-snds/.
Declaration of interests
The authors have no conflict of interest to report.
Acknowledgements
The authors acknowledge their respective employers: the French National Health Insurance Centre (CNAM), the French National Agency for Medicines and Health Product Safety (ANSM), and the Avec Groupe Hospitalier Mutualiste. The authors received no specific funding for this study.
Footnotes
Translation: For the French translation of the abstract see the Supplementary Materials section.
Supplementary data related to this article can be found at https://doi.org/10.1016/j.lanepe.2025.101484.
Appendix A. Supplementary data
References
- 1.Wang Q., Gümüş Z.H., Colarossi C., et al. SCLC: epidemiology, risk factors, genetic susceptibility, molecular pathology, screening, and early detection. J Thorac Oncol. 2023;18(1):31–46. doi: 10.1016/j.jtho.2022.10.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Howlader N., Forjaz G., Mooradian M.J., et al. The effect of advances in lung-cancer treatment on population mortality. N Engl J Med. 2020;383(7):640–649. doi: 10.1056/NEJMoa1916623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Alvarado-Luna G., Morales-Espinosa D. Treatment for small cell lung cancer, where are we now?-a review. Transl Lung Cancer Res. 2016;5(1):26–38. doi: 10.3978/j.issn.2218-6751.2016.01.13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Rousseau A., Michiels S., Simon-Tillaux N., et al. Impact of pembrolizumab treatment duration on overall survival and prognostic factors in advanced non-small cell lung cancer: a nationwide retrospective cohort study. Lancet Reg Health Eur. 2024;43 doi: 10.1016/j.lanepe.2024.100970. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Saida Y., Watanabe S., Kikuchi T. Extensive-stage small-cell lung cancer: current landscape and future prospects. Onco Targets Ther. 2023;16:657–671. doi: 10.2147/OTT.S272552. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Chen D., Zou B., Li B., et al. Adebrelimab plus chemotherapy and sequential thoracic radiotherapy as first-line therapy for extensive-stage small–cell lung cancer (ES-SCLC): a phase II trial. EClinicalMedicine. 2024;75 doi: 10.1016/j.eclinm.2024.102795. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Evans W.K., Osoba D., Feld R., Shepherd F.A., Bazos M.J., DeBoer G. Etoposide (VP-16) and cisplatin: an effective treatment for relapse in small-cell lung cancer. J Clin Oncol. 1985;3(1):65–71. doi: 10.1200/JCO.1985.3.1.65. [DOI] [PubMed] [Google Scholar]
- 8.Horn L., Mansfield A.S., Szczęsna A., et al. First-line atezolizumab plus chemotherapy in extensive-stage small-cell lung cancer. N Engl J Med. 2018;379(23):2220–2229. doi: 10.1056/NEJMoa1809064. [DOI] [PubMed] [Google Scholar]
- 9.Paz-Ares L., Dvorkin M., Chen Y., et al. Durvalumab plus platinum–etoposide versus platinum–etoposide in first-line treatment of extensive-stage small-cell lung cancer (CASPIAN): a randomised, controlled, open-label, phase 3 trial. Lancet. 2019;394(10212):1929–1939. doi: 10.1016/S0140-6736(19)32222-6. [DOI] [PubMed] [Google Scholar]
- 10.Wang Y., Li L., Hu J., et al. Comparison of efficacy and safety between PD-1 inhibitors and PD-L1 inhibitors plus platinum-etoposide as first-line treatment for extensive-stage small-cell lung cancer: a multicenter, real-world analysis. BMC Cancer. 2023;23(1):1196. doi: 10.1186/s12885-023-11709-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Igata F., Inoue H., Ikeda T., Nakao A., Ebi N., Fujita M. Comparison of real-world efficacy and safety of atezolizumab and durvalumab in combination with chemotherapy for first-line treatment of extensive-stage small-cell lung cancer. Anticancer Res. 2024;44(7):3175–3183. doi: 10.21873/anticanres.17132. [DOI] [PubMed] [Google Scholar]
- 12.Zou Y., Ren X., Zhang H., et al. Efficacy and safety of durvalumab + chemotherapy vs. atezolizumab + chemotherapy in the treatment of small-cell lung cancer: a retrospective comparative cohort study. J Thorac Dis. 2023;15(6):3339–3349. doi: 10.21037/jtd-23-588. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tuppin P., Rudant J., Constantinou P., et al. Value of a national administrative database to guide public decisions: From the système national d’information interrégimes de l'Assurance Maladie (SNIIRAM) to the système national des données de santé (SNDS) in France. Rev Epidemiol Sante Publique. 2017;65:S149–S167. doi: 10.1016/j.respe.2017.05.004. [DOI] [PubMed] [Google Scholar]
- 14.Swital M., Drouin J., Miranda S., Bakchine S., Botton J., Dray-Spira R. Use of multiple sclerosis disease-modifying therapies during pregnancy in France: nationwide study between 2010 and 2021. Mult Scler. 2024;30(2):227–237. doi: 10.1177/13524585231223395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Jourdain H., Hoisnard L., Sbidian E., Zureik M. Persistence and safety of anti-TNF biosimilars versus originators in immune-mediated inflammatory diseases: an observational study on the French National Health Data System. RMD Open. 2024;10(1) doi: 10.1136/rmdopen-2023-003531. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Roland N., Neumann A., Hoisnard L., et al. Use of progestogens and the risk of intracranial meningioma: national case-control study. BMJ. 2024;384 doi: 10.1136/bmj-2023-078078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Rios P., Herlemont P., Fauque P., et al. Medically assisted reproduction and risk of cancer among offspring. JAMA Netw Open. 2024;7(5) doi: 10.1001/jamanetworkopen.2024.9429. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Jourdain H., Mansouri I., Di Meglio A., Zureik M., Haddy N. Utilization and safety of trastuzumab emtansine (T-DM1): a nationwide population-based study using the French National Health Data System. ESMO Real World Data Digit Oncol. 2024;4 doi: 10.1016/j.esmorw.2024.100045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Takada K., Takamori S., Shimokawa M., Pinato D.J., Cortellini A. Prior antibiotics, proton pump inhibitors, and probiotics in patients with extensive stage small cell lung cancer treated with immune checkpoint blockade: a post-hoc analysis of the phase I/III IMpower 133 trial. Int J Cancer. 2025;156(5):914–919. doi: 10.1002/ijc.35249. [DOI] [PubMed] [Google Scholar]
- 20.Semenzato L., Botton J., Drouin J., et al. Chronic diseases, health conditions and risk of COVID-19-related hospitalization and in-hospital mortality during the first wave of the epidemic in France: a cohort study of 66 million people. Lancet Reg Health Eur. 2021;8 doi: 10.1016/j.lanepe.2021.100158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Rachas A., Gastaldi-Ménager C., Denis P., et al. The economic burden of disease in France from the National Health Insurance Perspective: the healthcare expenditures and conditions mapping used to prepare the French social security funding act and the public health act. Med Care. 2022;60(9):655–664. doi: 10.1097/MLR.0000000000001745. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Bryere J., Dejardin O., Launay L., Colonna M., Grosclaude P., Launoy G. Socioeconomic status and site-specific cancer incidence, a Bayesian approach in a French Cancer Registries Network study. Eur J Cancer Prev. 2018;27(4):391–398. doi: 10.1097/CEJ.0000000000000326. [DOI] [PubMed] [Google Scholar]
- 23.Paz-Ares L., Chen Y., Reinmuth N., et al. Durvalumab, with or without tremelimumab, plus platinum-etoposide in first-line treatment of extensive-stage small-cell lung cancer: 3-year overall survival update from CASPIAN. ESMO Open. 2022;7(2) doi: 10.1016/j.esmoop.2022.100408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Liu S.V., Reck M., Mansfield A.S., et al. Updated overall survival and PD-L1 subgroup analysis of patients with extensive-stage small-cell lung cancer treated with atezolizumab, carboplatin, and etoposide (IMpower133) J Clin Oncol. 2021;39(6):619–630. doi: 10.1200/JCO.20.01055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Blumenthal G.M., Gong Y., Kehl K., et al. Analysis of time-to-treatment discontinuation of targeted therapy, immunotherapy, and chemotherapy in clinical trials of patients with non-small-cell lung cancer. Ann Oncol. 2019;30(5):830–838. doi: 10.1093/annonc/mdz060. [DOI] [PubMed] [Google Scholar]
- 26.Rossi A., Di Maio M., Chiodini P., et al. Carboplatin- or cisplatin-based chemotherapy in first-line treatment of small-cell lung cancer: the COCIS meta-analysis of individual patient data. J Clin Oncol. 2012;30(14):1692–1698. doi: 10.1200/JCO.2011.40.4905. [DOI] [PubMed] [Google Scholar]
- 27.Gay C.M., Stewart C.A., Park E.M., et al. Patterns of transcription factor programs and immune pathway activation define four major subtypes of SCLC with distinct therapeutic vulnerabilities. Cancer Cell. 2021;39(3):346–360.e7. doi: 10.1016/j.ccell.2020.12.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.