Durvalumab Consolidation Treatment Following Concurrent Chemoradiotherapy in Limited-Stage SCLC: ADRIATIC China Subgroup Analysis

Abstract

Introduction

At the first interim analysis of the phase 3 ADRIATIC trial, durvalumab consolidation significantly improved overall survival (OS) and progression-free survival (PFS) compared with placebo in patients with limited-stage SCLC (LS-SCLC) without progression after concurrent chemoradiotherapy (cCRT). We report a prespecified subgroup analysis of patients enrolled in China.

Methods

Patients with stage I to III LS-SCLC were randomized to durvalumab, durvalumab plus tremelimumab (arm remained blinded), or placebo for up to 24 months. The dual primary end points were OS and PFS by blinded independent central review for durvalumab versus placebo.

Results

Of the global population, 95 of 530 patients (17.9%) were randomized in China; 49 of 264 versus 46 of 266 received durvalumab versus placebo. For prior cCRT, 69.4% versus 69.6% of the China subgroup received cisplatin-etoposide, and 53.1% versus 69.6% had once-daily radiotherapy; post-cCRT, 55.1% versus 67.4% received prophylactic cranial irradiation. The hazard ratio for OS (durvalumab versus placebo; median follow-up 35.5 mo) was 0.71 (95% confidence interval: 0.37–1.37). Median OS was not reached in either arm (36-mo OS: 63.7% versus 55.4%). The hazard ratio for PFS (median follow-up 27.7 mo) was 0.67 (95% confidence interval: 0.39–1.14); median PFS was 22.9 versus 14.3 months (24-mo PFS: 45.8% versus 37.6%). With durvalumab versus placebo, 16.3% versus 17.4% of patients had maximum grade 3 or 4 adverse events (AEs), 42.9% versus 21.7% had serious AEs, and 10.2% versus 13.0% discontinued treatment due to AEs; 55.1% versus 63.0% had pneumonitis/radiation pneumonitis (grade 3 or 4: 4.1% versus 4.3%).

Conclusion

Consolidation durvalumab demonstrated a favorable benefit to risk profile in patients with LS-SCLC without progression after cCRT who were enrolled in China.

Introduction

Lung cancer is the most frequently diagnosed malignancy and the leading cause of cancer-related death in China in both men and women. An estimated 1.06 million new cases were diagnosed and 0.73 million deaths were attributed to lung cancer in 2022.^1,2 SCLC accounts for 5.7% to 18.3% of all lung cancers in China.3, 4, 5, 6, 7, 8, 9 Approximately 26.8% to 49.8% of patients diagnosed with SCLC in China present with limited-stage disease (LS-SCLC).^5,6,10,11

For most patients with LS-SCLC, surgery is not an option. The recommended treatment in both Chinese and international guidelines has for many years been chemoradiotherapy (CRT), followed by prophylactic cranial irradiation (PCI) where indicated.12, 13, 14 Concurrent CRT (cCRT), with thoracic radiotherapy starting no later than the end of the second cycle of platinum-etoposide chemotherapy, is preferred based on evidence revealing improved overall survival (OS) when radiotherapy is begun earlier rather than later.^15,16 Although LS-SCLC is typically responsive to CRT, most patients relapse within 2 years of starting treatment.17, 18, 19 In China, reported 5-year OS rates for LS-SCLC following CRT (either concurrent or sequential) are approximately 17% to 24%,^10,20 highlighting the need for new treatments to improve outcomes in this patient population.

Durvalumab is a selective, high-affinity human IgG1 monoclonal antibody that inhibits programmed cell death-ligand 1 (PD-L1) binding to programmed cell death 1 (PD-1) and CD80.²¹ Consolidation treatment with durvalumab in patients with unresectable, locally advanced NSCLC whose disease had not progressed following platinum-based cCRT significantly prolonged both progression-free survival (PFS) and OS when compared with placebo in the phase 3 PACIFIC trial, with the benefits sustained through approximately 5 years of follow-up.22, 23, 24 These findings informed the rationale for investigating durvalumab in this setting in LS-SCLC. Furthermore, in the phase 3 CASPIAN study in patients with extensive-stage SCLC (ES-SCLC), addition of durvalumab to first-line platinum-etoposide significantly improved OS compared with chemotherapy alone.²⁵ Results from PACIFIC and CASPIAN led to approvals for durvalumab in the respective treatment settings in China^26,27 and elsewhere globally.

The ongoing phase 3 ADRIATIC trial (NCT03703297) investigated durvalumab, with or without tremelimumab (another monoclonal antibody that targets cytotoxic T-lymphocyte antigen 4), compared with placebo as consolidation therapy for patients with LS-SCLC who had not progressed after cCRT.²⁸ The results from the first planned interim analysis (data cutoff: January 15, 2024) showed that consolidation durvalumab monotherapy significantly improved OS (hazard ratio [HR]: 0.73; 95% confidence interval [CI]: 0.57–0.93; p = 0.0104; median 55.9 versus 33.4 mo) and PFS (HR: 0.76; 95% CI: 0.61–0.95; p = 0.0161; median 16.6 versus 9.2 mo) compared with placebo.²⁸ The safety profile for durvalumab was consistent with that established in the post-cCRT setting.^22,23 In ADRIATIC, durvalumab resulted in numerically higher rates of adverse events (AEs) leading to treatment discontinuation than placebo (16.4% versus 10.6%) and higher rates of immune-mediated AEs (any grade, 32.1% versus 10.2%).²⁸ Pneumonitis/radiation pneumonitis (grouped term) was among the most commonly reported AEs in both treatment groups, occurring in 38.2% versus 30.2% of patients in the durvalumab versus placebo arms, respectively; grade 3 or 4 events occurred in 3.1% versus 2.6% of patients. Results from the durvalumab plus tremelimumab arm remained blinded at the first interim analysis.²⁸ Based on these data, the ADRIATIC durvalumab regimen has received multiple regulatory approvals, and is recommended in treatment guidelines,29, 30, 31 including the 2025 Chinese Society of Clinical Oncology Guidelines for the Diagnosis and Treatment of Small-Cell Lung Cancer. Here, we report the results of an exploratory, prespecified subgroup analysis of durvalumab compared with placebo in patients enrolled in ADRIATIC in China.

Patients and Methods

Study Design and Patients

ADRIATIC is an ongoing phase 3, randomized, double-blind, placebo-controlled study, with 939 patients enrolled and 730 patients randomized between September 28, 2018, and August 18, 2021. Efficacy and safety results from the first interim analysis of ADRIATIC, including detailed methodology with full eligibility criteria, have already been published.²⁸ In brief, eligible patients were aged 18 years or older with a WHO performance status (WHO PS) of 0 or 1, had histologically or cytologically confirmed stage I to III LS-SCLC (inoperable if stage I or II) per the American Joint Committee on Cancer Staging Manual, eighth edition,³² or the International Association for the Study of Lung Cancer Staging Manual in Thoracic Oncology,³³ and had no disease progression after receiving definitive cCRT. The trial was conducted in accordance with the Declaration of Helsinki, the International Council for Harmonisation guideline for good clinical practice, and applicable local laws and regulations. The protocol and all amendments were approved by the relevant ethics committee or institutional review board at each trial center. All patients provided written informed consent.

Patients had to have completed prior definitive cCRT according to local or regional standards of care for patients with LS-SCLC prior to randomization. Chemotherapy comprised 3 or 4 cycles of cisplatin-etoposide or carboplatin-etoposide. Thoracic radiotherapy (60–66 Gy once daily over 6 wks or 45 Gy twice daily over 3 wks) had to start no later than at the end of cycle 2 of chemotherapy. Additionally, patients could receive PCI after completing cCRT and before randomization, per local or regional standards of care and at the discretion of the investigator.

Patients were required to be randomized in ADRIATIC within 42 days of completing prior cCRT (the last day of radiotherapy or the last day of the final cycle of chemotherapy, whichever was later). Randomization was stratified by disease stage (I or II versus III) and receipt of PCI (yes versus no). Patients were initially randomized in a 1:1:1 ratio to the durvalumab, durvalumab plus tremelimumab, or placebo arms. After the randomization of the first 600 patients, an additional 130 patients were randomized in a 1:1 ratio to the durvalumab and placebo arms only.

Patients in the durvalumab and durvalumab plus tremelimumab arms received durvalumab 1500 mg every 4 weeks. Patients in the durvalumab plus tremelimumab arm also received tremelimumab 75 mg every 4 weeks for four doses, whereas patients in the durvalumab arm received a tremelimumab-matched placebo. Patients in the placebo arm received a durvalumab-matched placebo plus a tremelimumab-matched placebo. Treatment continued for a maximum of 24 months, or until investigator-determined disease progression, intolerable toxicity, or withdrawal of consent.

End Points and Assessments

The dual primary end points were OS and PFS per Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST v.1.1), as assessed by blinded independent central review (BICR), with durvalumab versus placebo. Key secondary end points were OS and PFS by BICR with durvalumab plus tremelimumab versus placebo. However, the durvalumab plus tremelimumab arm remained blinded at this first interim analysis because the statistical boundary for OS with durvalumab plus tremelimumab versus placebo had not been crossed, as determined by the Independent Data and Safety Monitoring Committee.²⁸ Other secondary end points included OS at 24 and 36 months, PFS at 18 and 24 months, and safety and tolerability. This prespecified exploratory subgroup analysis compared efficacy and safety with durvalumab versus placebo in patients enrolled in China. The data cutoff date was the same as that for the first interim analysis of the global population (January 15, 2024).

Tumor assessments, using computed tomography (preferred) or magnetic resonance imaging, were done at baseline (1–42 d before randomization), every 8 weeks for the first 72 weeks after randomization, at 84 and 96 weeks, and then every 24 weeks until RECIST v.1.1–defined disease progression, with one follow-up scan at least 4 weeks later. Following the last dose of the study treatment, patients were followed every 8 weeks for subsequent anticancer therapy and survival status. Patients were monitored for safety throughout their treatment and until 90 days after the last dose of study treatment. AEs were graded using the National Cancer Institute Common Terminology Criteria for Adverse Events version 4.03.

Statistical Methods

The China subgroup included all patients randomized at sites in mainland China. Efficacy was analyzed in the intent-to-treat population, and safety was analyzed in all patients who received at least one dose of study treatment. All statistical analyses of the China subgroup were considered exploratory, with no α allocated for end point comparisons; consequently, no p values are reported. OS and PFS were analyzed using the Kaplan-Meier method; 95% CIs for median values were derived based on the Brookmeyer-Crowley method with log-log transformation. HRs with corresponding 95% CI were calculated for OS and PFS using stratified Cox proportional hazards models adjusted for receipt of PCI.

Results

Patients and Treatment

Of the global ADRIATIC population randomized to durvalumab or placebo (N = 530), 95 patients (17.9%) were randomized at 24 sites in China between April 20, 2020, and July 9, 2021, to receive durvalumab (n = 49) or placebo (n = 46). Demographics and baseline disease characteristics in the China subgroup were generally balanced between the treatment arms (Table 1). Most patients were male (74.7%), former or current smokers (67.4%), and had stage III disease at diagnosis (92.6%). Overall, the China subgroup was somewhat younger than the global population, with a median age of 58.0 years (range, 28–73) and 23.2% of patients aged 65 to <75 years; the global population²⁸ had a median age of 62.0 years [range, 28–84], with 39.2% aged 65 years or over. It also included more patients with WHO PS 1 (63.2% compared with 51.3%) and those who were never smokers (32.6% compared with 9.2%).²⁸

Table 1.

Baseline Patient Demographics and Disease Characteristics, and cCRT Received Before Randomization, in the China Subgroup of the ADRIATIC Trial

Patient Characteristics	Durvalumab (n = 49)	Placebo (n = 46)
Age
Median (range), y	59.0 (28–73)	55.0 (28–73)
Mean (SD), y	58.4 (8.79)	56.6 (8.98)
<50 y	5 (10.2)	8 (17.4)
50–<65 y	32 (65.3)	28 (60.9)
≥65–<75 y	12 (24.5)	10 (21.7)
Sex
Male	38 (77.6)	33 (71.7)
Female	11 (22.4)	13 (28.3)
WHO performance status
0	20 (40.8)	15 (32.6)
1	29 (59.2)	31 (67.4)
Smoking history
Smoker	35 (71.4)	29 (63.0)
Current smoker	3 (6.1)	3 (6.5)
Previous smoker	32 (65.3)	26 (56.5)
Never smoked	14 (28.6)	17 (37.0)
TNM stage at diagnosisa
I or II	5 (10.2)	2 (4.3)
III	44 (89.8)	44 (95.7)
Previous cCRT, chemotherapy regimen
Cisplatin-etoposide	34 (69.4)	32 (69.6)
Carboplatin-etoposide	15 (30.6)	14 (30.4)
Previous cCRT, number of chemotherapy cycles
3	3 (6.1)	2 (4.3)
4	46 (93.9)	44 (95.7)
Previous cCRT, radiotherapy fractionation schedule
Once daily	26 (53.1)	32 (69.6)
Twice daily	23 (46.9)	14 (30.4)
Previous cCRT, radiotherapy total dose, Gy
≥60–≤66	21 (42.9)	30 (65.2)
45	21 (42.9)	13 (28.3)
Other	7 (14.3)	3 (6.5)
Best response to cCRT
Complete response	1 (2.0)	5 (10.9)
Partial response	42 (85.7)	34 (73.9)
Stable disease	6 (12.2)	7 (15.2)
Time from end of cCRT to randomization
<14 d	10 (20.4)	6 (13.0)
≥14 d to <28 d	14 (28.6)	8 (17.4)
≥28 d	25 (51.0)	32 (69.6)
Received PCI before randomizationa	27 (55.1)	31 (67.4)

Note: Data are n (%), unless indicated otherwise.

cCRT, concurrent chemoradiotherapy; PCI, prophylactic cranial irradiation.

^aStratification factor, as recorded on the electronic case-report form.

In the China subgroup, prior cCRT comprised cisplatin chemotherapy in 69.5% of patients and once-daily radiotherapy in 61.1%; following cCRT, 61.1% of patients received PCI prior to randomization (Table 1). Respective rates in the global population were 66.2%, 72.1%, and 53.8%.²⁸ There were some minor imbalances between arms in the China subgroup (Table 1). In the durvalumab versus placebo arm, the percentages of patients receiving once-daily radiotherapy (53.1% versus 69.6%; global population, 73.9% versus 70.3%) and PCI (55.1% versus 67.4%; global population, 53.8% versus 53.8%) were lower. Additionally, a lower percentage of patients in the durvalumab arm achieved a complete response to prior cCRT (2.0% versus 10.9%; global population, 11.7% versus 12.8%) and had an interval from end of cCRT to randomization of 28 days or more (51.0% versus 69.6%; global population, 58.0% versus 57.9%).

All patients in the China subgroup underwent study treatment; patients in the durvalumab and placebo arms received a median of 12.0 (range, 1–26) and 9.5 (range, 1–26) infusions, respectively, and means of 13.3 and 11.7 infusions (Table 2). In both arms of the global population, patients received a median of 9.0 (range, 1–26) infusions, and the mean number of infusions was 12.9 and 11.8 in the durvalumab and placebo arms, respectively.²⁸ At data cutoff (January 15, 2024), all patients were off study treatment. In the durvalumab and placebo arms of the China subgroup, respectively, 18 (36.7%) and 14 (30.4%) patients completed the maximum 24 months of treatment allowed per protocol (33.5% and 26.4%, respectively, in the global population²⁸). Of the 31 (63.3%) and 32 (69.6%) patients who discontinued treatment with durvalumab and placebo, respectively, the reasons for treatment discontinuation were disease progression in 24 (49.0%) and 26 (56.5%), AEs in five (10.2%) and six (13.0%), and patient decision in two (4.1%) and zero. In the durvalumab and placebo arms, 21 (42.9%) and 27 (58.7%) patients had received any subsequent anticancer therapy, which included cytotoxic chemotherapy in 18 (36.7%) and 26 (56.5%) patients and immunotherapy in six (12.2%) and 15 (32.6%) patients, respectively. In the global population, 36.0% and 47.7% of patients in the durvalumab and placebo arms, respectively, had received subsequent anticancer therapy,²⁸ which included cytotoxic chemotherapy in 88 of 264 (33.3%) and 121 of 266 (45.5%) patients and immunotherapy in 23 (8.7%) and 39 (14.7%) patients, respectively.

Table 2.

Summary of Treatment Exposure, Safety, and Common AEs^a in the China Subgroup of the ADRIATIC Trial

	Durvalumab (n = 49)	Placebo (n = 46)
Treatment exposure – number of infusions
Median (range)	12.0 (1–26)	9.5 (1–26)
Mean (SD)	13.3 (9.0)	11.7 (8.9)
AEs
Any AE	48 (98.0)	44 (95.7)
Any AE of maximum grade 3 or 4	8 (16.3)	8 (17.4)
Any serious AE	21 (42.9)	10 (21.7)
Any AE with outcome of death	0	0
Any AE leading to treatment discontinuationb	5 (10.2)	6 (13.0)
Any AE leading to dose interruption	23 (46.9)	15 (32.6)
Any imAEc	21 (42.9)	5 (10.9)
Any imAEc of maximum grade 3 or 4	4 (8.2)	0
Common any-grade AEsd
WBC count decreased	16 (32.7)	11 (23.9)
Anemia	15 (30.6)	10 (21.7)
Hyperthyroidism	13 (26.5)	1 (2.2)
Radiation pneumonitis	12 (24.5)	19 (41.3)
Hypothyroidism	10 (20.4)	1 (2.2)
Decreased appetite	9 (18.4)	8 (17.4)
Hypoalbuminemia	9 (18.4)	2 (4.3)
Hyponatremia	9 (18.4)	7 (15.2)
Lymphocyte count decreased	9 (18.4)	6 (13.0)
Weight decreased	9 (18.4)	5 (10.9)
ALT increased	8 (16.3)	11 (23.9)
Blood bilirubin increased	6 (12.2)	10 (21.7)
Pneumonitis	6 (12.2)	9 (19.6)
Hyperglycemia	5 (10.2)	7 (15.2)
AST increased	4 (8.2)	7 (15.2)
AEs of maximum grade 3 or 4 (≥2 patients)e
Immune-mediated lung disease	2 (4.1)	0
Pulmonary embolism	2 (4.1)	0
Lipase increased	1 (2.0)	2 (4.3)
Radiation pneumonitis	0	2 (4.3)
Pneumonitis or radiation pneumonitis (grouped terms)f
Any AE	27 (55.1)	29 (63.0)
Pneumonitis grouped term	18 (36.7)	10 (21.7)
Radiation pneumonitis grouped term	12 (24.5)	19 (41.3)
Any AE of maximum grade 3 or 4	2 (4.1)	2 (4.3)
Any AE leading to treatment discontinuation	5 (10.2)	4 (8.7)

Note: Data are n (%), unless indicated otherwise.

AE, adverse event; ALT, alanine aminotransferase; AST, aspartate aminotransferase; imAE, immune-mediated adverse event; WBC, white blood cell.

^aAEs of any cause with an onset date after the first dose of durvalumab or placebo, as well as any AEs before the receipt of the first dose that increased in severity after the first dose, through to 90 days after the last dose or until the start of the first subsequent systemic anticancer therapy (whichever occurred first).

^bDurvalumab: immune-mediated lung disease (n = 2), pneumonia plus radiation pneumonitis (n = 1), and radiation pneumonitis and pneumonitis (each n = 1). Placebo: radiation pneumonitis (n = 3), and metabolic encephalopathy, platelet count decreased, and pneumonitis (each n = 1).

^cAEs of special interest (excluding infusion-related, hypersensitivity, or anaphylactic reactions) that were consistent with an immune-mediated mechanism and that led to treatment with systemic glucocorticoids, other immunosuppressants, or endocrine therapy.

^dOccurring at any grade in 15% or more of patients in either treatment group.

^eAdditionally, AEs of maximum grade 3 or 4 of ALT increased, amylase increased, appendicitis, diabetic ketoacidosis, diarrhea, foot fracture, hepatitis, hyperglycemia, hypertension, liver injury, lymphocyte count decreased, lymphopenia, esophagitis ulcerative, type 1 diabetes mellitus, and upper respiratory tract infection were reported in 1 (2.0%) patient each in the durvalumab group, and of cerebral infarction, depression, electrolyte imbalance, hyponatremia, metabolic encephalopathy, and platelet count decreased were reported in 1 (2.2%) patient each in the placebo group.

^fPneumonitis grouped term includes events with the preferred terms of immune-mediated lung disease, interstitial lung disease, and pneumonitis. Radiation pneumonitis grouped term includes the preferred terms of radiation pneumonitis and radiation fibrosis; no cases of radiation fibrosis were reported in the China subgroup.

OS and PFS

At data cutoff, in the durvalumab and placebo arms of the China subgroup, 17 (34.7%) and 20 (43.5%) patients had had OS events, respectively. Median follow-up for OS in censored patients was 35.7 months (range, 11.3–45.0) and 35.2 months (range, 30.3–43.0), respectively. Median OS was not reached in either arm; the stratified HR for death with durvalumab versus placebo was 0.71 (95% CI: 0.37–1.37) (Fig. 1). In the durvalumab and placebo arms, the respective 24-month OS rates (95% CI) were 77.2% (62.6–86.7) and 69.6% (54.1–80.7), and the 36-month rates were 63.7% (48.0–75.8) and 55.4% (38.5–69.4), while in the global population, the respective 24-month OS rates (95% CI) were 68.0% (61.9–73.3) and 58.5% (52.3–64.3), and the corresponding 36-month rates were 56.5% (50.0–62.5) and 47.6% (41.3–53.7), and the OS HR was 0.73 (95% CI: 0.57–0.93).²⁸

Figure 1.

Kaplan-Meier curves for OS with durvalumab and placebo in the China subgroup of the ADRIATIC trial. Dashed lines show the times of the landmark analyses at 24 and 36 months. CI, confidence interval; HR, hazard ratio; NE, not evaluable; NR, not reached; OS, overall survival.

By BICR assessment, PFS events were observed in 26 (53.1%) and 30 (65.2%) patients in the durvalumab and placebo arms of the China subgroup, respectively. Median follow-up for PFS in censored patients was 33.1 months (range, 1.8–38.9) and 24.8 months (range, 0–38.7), respectively. Median PFS by BICR was 22.9 months (95% CI: 7.4–not evaluable) in the durvalumab arm and 14.3 months (95% CI: 5.7–26.9) in the placebo arm, and the stratified HR for progression or death was 0.67 (95% CI: 0.39–1.14) (Fig. 2). In the durvalumab and placebo arms, 18-month PFS rates (95% CI) were respectively 50.9% (35.7–64.2) and 40.3% (25.1–55.1) and 24-month rates were 45.8% (30.8–59.5) and 37.6% (22.8–52.4), while in the global population, respective 18-month PFS rates (95% CI) were 48.8% (42.2–55.0) and 36.1% (29.9–42.2) and 24-month rates were 46.2% (39.6–52.5) and 34.2% (28.2–40.3), and the PFS HR was 0.76 (95% CI: 0.61–0.95).²⁸

Figure 2.

Kaplan-Meier curves for PFS with durvalumab and placebo, as assessed by blinded independent central review, in the China subgroup of the ADRIATIC trial. Dashed lines show the times of the landmark analyses at 18 and 24 months. CI, confidence interval; HR, hazard ratio; NE, not evaluable; PFS, progression-free survival.

Safety

AEs of any cause were reported in 98.0% and 95.7% of patients in the durvalumab and placebo arms of the China subgroup, respectively, with maximum grade 3 or 4 AEs occurring in 16.3% and 17.4% of patients (Table 2); the respective rates of any-grade AEs and maximum grade 3 or 4 AEs in the global population were 94.3% and 88.3%, and 24.4% and 24.2%.²⁸ AEs that occurred in more than 25% of China subgroup patients in either arm were radiation pneumonitis (24.5% and 41.3% in the durvalumab and placebo arms, respectively), white blood cell count decreased (32.7% and 23.9%), anemia (30.6% and 21.7%), and hyperthyroidism (26.5% and 2.2%) (Table 2). AEs were mostly grade 1 or 2 in severity; events observed at maximum grade 3 or 4 in more than one patient were immune-mediated lung disease, pulmonary embolism (n = 2 each in the durvalumab arm), lipase increased (n = 1 in the durvalumab arm, n = 2 in the placebo arm), and radiation pneumonitis (n = 2 in the placebo arm) (Table 2).

Serious AEs (SAEs) were reported in 42.9% and 21.7% of patients in the durvalumab and placebo arms, respectively (Table 2); the most common were radiation pneumonitis (10.2% and 2.2% in the durvalumab and placebo arms, respectively), interstitial lung disease (10.2% and 0%), pneumonitis (4.1% and 6.5%), and immune-mediated lung disease (6.1% and 0%) (Table 3). Respective rates of SAEs in the global population were 29.8% and 24.2%.²⁸ AEs led to treatment discontinuation in 10.7% and 13.0% of patients in the durvalumab and placebo arms of the China subgroup, respectively (Table 2) (16.4% and 10.6% in the global population²⁸). No patients in either arm of the China subgroup died due to AEs.

Table 3.

All Serious AEs of Any Cause in the China Subgroup of the ADRIATIC Trial

Event, n (%)	Durvalumab (n = 49)	Placebo (n = 46)
Any serious adverse event	21 (42.9)	10 (21.7)
Interstitial lung disease	5 (10.2)	0
Radiation pneumonitis	5 (10.2)	1 (2.2)
Immune-mediated lung disease	3 (6.1)	0
Pneumonia	2 (4.1)	0
Pneumonitis	2 (4.1)	3 (6.5)
Appendicitis	1 (2.0)	0
Diabetic ketoacidosis	1 (2.0)	0
Esophagitis ulcerative	1 (2.0)	0
Foot fracture	1 (2.0)	0
Hepatitis	1 (2.0)	0
Liver injury	1 (2.0)	0
Lymphocyte count decreased	1 (2.0)	0
Pleural effusion	1 (2.0)	0
Pulmonary embolism	1 (2.0)	0
Sinus node dysfunction	1 (2.0)	0
Type 2 diabetes mellitus	1 (2.0)	0
Weight decreased	1 (2.0)	0
Cerebral infarction	0	1 (2.2)
Depression	0	1 (2.2)
Electrolyte imbalance	0	1 (2.2)
Hemoptysis	0	1 (2.2)
Hyponatremia	0	1 (2.2)
Metabolic encephalopathy	0	1 (2.2)
Sternal fracture	0	1 (2.2)
Venous thrombus limb	0	1 (2.2)

Note: Serious AEs of any cause with an onset date after the first dose of durvalumab or placebo, as well as any AEs before the receipt of the first dose that increased in severity after the first dose, through to 90 days after the last dose or until the start of the first subsequent systemic anticancer therapy (whichever occurred first). Events are listed alphabetically in descending order of total number of events in the durvalumab group.

AE, adverse event.

The incidence of AEs of any grade within the pooled grouped terms of pneumonitis and radiation pneumonitis was 55.1% in the durvalumab arm and 63.0% in the placebo arm of the China subgroup (Table 2). Most of these AEs were grade 1 or 2 in severity, with 4.1% and 4.3% of patients having AEs of maximum grade 3 or 4. In the global population, overall rates were 38.2% and 30.2%, respectively, with 3.1% and 2.6% having an event of maximum grade 3 or 4.²⁸ AEs within the pooled grouped terms of pneumonitis and radiation pneumonitis led to treatment discontinuation in 10.2% and 8.7% of patients in the durvalumab and placebo arms of the China subgroup. Immune-mediated AEs (imAEs) were more common in the durvalumab arm of the China subgroup (42.9%) than in the placebo arm (10.9%); rates of imAEs by group term are shown in Table 4.

Table 4.

Immune-Mediated AEs in the China Subgroup of the ADRIATIC Trial

imAEa Group Term, n (%)	Durvalumab (n = 49)		Placebo (n = 46)
	Any Grade	Maximum Grade 3 or 4	Any Grade	Maximum Grade 3 or 4
Any imAEb	21 (42.9)	4 (8.2)	5 (10.9)	0
Pneumonitis	10 (20.4)	2 (4.1)	4 (8.7)	0
Hypothyroid events	9 (18.4)	0	1 (2.2)	0
Hyperthyroid events	3 (6.1)	0	0	0
Hepatic events	1 (2.0)	1 (2.0)	0	0
Type 1 diabetes mellitus	1 (2.0)	1 (2.0)	0	0

AE, adverse event; imAE, immune-mediated adverse event.

^aimAEs are defined as AEs of special interest (excluding infusion-related or hypersensitivity or anaphylactic reaction) that are consistent with an immune-mediated mechanism and that require treatment with systemic corticosteroids, other immunosuppressants, or endocrine therapy.

^bimAEs with an onset date following first dose of study treatment, or pre-treatment AEs that increased in severity after the first dose of study treatment, through to 90 days after the last dose of study treatment or until the start of the first subsequent systemic anticancer therapy (whichever occurred first). Events are listed in descending order of total number of events by group term in the durvalumab group.

Discussion

In this prespecified exploratory subgroup analysis of the ADRIATIC trial, consolidation durvalumab following cCRT appeared to have a favorable benefit to risk profile in patients from China with LS-SCLC. The magnitude of OS and PFS benefit with durvalumab versus placebo, in terms of HRs, was consistent with the findings reported for the global population,²⁸ where durvalumab demonstrated a statistically significant and clinically meaningful benefit versus placebo; however, the 95% CIs of both HRs in the China subgroup were notably wider due to the substantially smaller sample size compared with the global population,²⁸ with the upper boundaries crossing one. Although there were some differences in overall rates of AEs (notably, the higher rate of SAEs observed in the durvalumab arm of the China subgroup) and rates of individual AEs, safety findings were broadly in line with those in the global population and with the established safety profile for durvalumab in the setting of post-cCRT consolidation.^22,23

While the patient and disease characteristics of the China subgroup were generally similar to those of the global study population, there were some notable differences. The China subgroup was younger compared with the global population; only 23.2% compared with 39.2% of patients were aged 65 years or over, none of whom in the China subgroup were aged 75 years or over, and the median age was 58.0 (range, 28–73) compared with 62.0 (range, 28–84) years.²⁸ Furthermore, the China subgroup included a notable number of patients who were less than 50 years of age (n = 13; 13.7%). These findings are in line with previous provincial registry data from 2015,⁴ a SCLC patient information questionnaire,⁵ and real-world findings¹⁰ in China showing 17.1% to 27.9% of patients with LS-SCLC or SCLC were aged 65 years or over. The China subgroup included a higher proportion of never-smokers (32.6%, compared with 9.2% in the global population²⁸), consistent with previous data on patients with SCLC in China and patients in Chinese phase 3 studies of immunotherapy for ES-SCLC with no smoking history (20.4%–42.7%^5,6,8,34, 35, 36, 37, 38, 39).

In terms of prior cCRT components and PCI use, data were largely consistent with the global ADRIATIC population,²⁸ with a once-daily radiotherapy schedule more commonly used than twice-daily, cisplatin-etoposide used more frequently than carboplatin-etoposide, and most patients receiving PCI, in line with local practice and treatment guidelines.^5,40, 41, 42 However, there were some between-arm differences that may potentially have differentially affected survival; some may have benefited patients in the placebo arm and others may have benefited those in the durvalumab arm. Firstly, a higher proportion of patients in the placebo versus durvalumab arm received PCI, which has been shown in a meta-analysis to significantly improve OS in patients with LS-SCLC.⁴³ Secondly, in the durvalumab versus placebo arm, a greater proportion of patients received twice-daily (versus once-daily) radiotherapy, and, as a result, more patients received a total radiotherapy dose of 45 Gy versus 60 to 66 Gy. Although this regimen has historically been the standard of care in the CRT setting, after twice-daily radiotherapy was shown to be superior to once-daily radiotherapy in the Intergroup 0096 study,⁴⁴ the more recent CONVERT and CALGB 30610/RTOG 0538 phase 3 studies failed to show a significant difference in OS between once- and twice-daily regimens.^45,46 It is worth noting that the higher rate of use of a once-daily radiotherapy schedule overall is consistent with previously reported practice patterns for LS-SCLC in China.⁴⁰ Lastly, a lower proportion of patients in the durvalumab versus placebo arm was randomized (and thus started consolidation treatment) 28 days or more after the end of cCRT. However, the time interval from completion of cCRT to randomization would have been influenced by whether PCI was administered, and it is therefore difficult to understand its individual contribution. Unfortunately, it is impossible to assess the overall cumulative effect of these between-arm differences on survival outcomes in the China subgroup.

Consolidation with durvalumab resulted in a 29% reduction in the risk of death and a 33% reduction in the risk of progression or death compared with placebo in the China subgroup (although the small sample size led to wide CIs around the HRs), consistent with the respective 27% and 24% reductions in risk in the global population.²⁸ Rates of OS in both arms of the China subgroup appeared numerically higher at 24 months and 36 months than in the global population.²⁸ Possible reasons for the numerically higher OS in the China subgroup include their younger age and the higher proportion of never-smokers compared with the global population. Both age47, 48, 49 and smoking history⁵⁰ have been previously reported as independent prognostic factors in patients with LS-SCLC. Furthermore, treatment duration in the durvalumab arm was somewhat longer in the China subgroup than in the global population, and a higher proportion of patients received any subsequent anticancer treatment, including a higher rate of subsequent immunotherapy use, in the China subgroup than in the global population (the rate of subsequent treatment was slightly higher in the placebo versus the durvalumab arm of the China subgroup; however, this likely reflects the greater proportion of patients who had disease progression in the placebo arm). Notably, the OS data for the China subgroup were less mature (34.7% and 43.5% of patients had had events in the durvalumab and placebo arms, respectively) than OS data for the global population (43.6% and 54.9%, respectively²⁸). PFS landmark data were broadly similar to those for the global population.²⁸ As patients in this study were randomized after completion of cCRT, data on the number of PFS events within versus outside the CRT radiation field were not collected.

The overall safety profile for durvalumab in the China subgroup was broadly consistent with that of the global ADRIATIC population,²⁸ with some differences in AE rates, and with that established in the post-cCRT setting in patients with unresectable locally advanced NSCLC.^22,23 In the China subgroup, while the rates of any-grade AEs were slightly numerically higher, notably in the placebo arm, than in the global population, the rates of maximum grade 3 or 4 AEs were numerically lower. The rates of AEs leading to treatment discontinuation were slightly numerically lower in the durvalumab arm and similar in the placebo arm in the China subgroup compared with the global population. However, the rate of SAEs in the durvalumab arm of the China subgroup appeared somewhat higher than that in the global population.²⁸ Additionally, the overall rates of pneumonitis/radiation pneumonitis (grouped terms) were numerically higher in the China subgroup than in the global population, but the rates of maximum grade 3 or 4 AEs and AEs leading to treatment discontinuation were similar in both treatment arms.²⁸ This finding is consistent with data from PACIFIC and real-world studies using the PACIFIC regimen,^51,52 and from a meta-analysis comparing pneumonitis in Asian and non-Asian patients.⁵³ Overall, these data may reflect regional differences in diagnosis and reporting or monitoring of pneumonitis/radiation pneumonitis. However, the small size of the China subgroup and the exploratory nature of the analysis limit the interpretation of the safety findings and the differences in the observed frequencies of AEs presented here.

In conclusion, the findings of this China subgroup analysis of the phase 3 ADRIATIC trial suggest that consolidation durvalumab for up to 24 months provided a consistent magnitude of benefit to that observed in the global population, where there was a statistically significant and clinically meaningful benefit in terms of OS and PFS compared with placebo in patients with LS-SCLC and no progression after cCRT. These results support consolidation durvalumab as the new standard of care in this setting, including in patients in China.

A plain language summary and a graphical abstract for this manuscript are included in the Supplement.

CRediT Authorship Contribution Statement

Lixia Ma: Investigation; Data curation; Writing – review & editing.

Qiming Wang: Investigation; Data curation; Writing – review & editing.

Jian Fang: Investigation; Data curation; Writing – review & editing.

Anhui Shi: Investigation; Data curation; Writing – review & editing.

Lin Wu: Investigation; Data curation; Writing – review & editing.

Shun Lu: Investigation; Data curation; Writing – review & editing.

Zhengfei Zhu: Investigation; Data curation; Writing – review & editing.

Yunpeng Liu: Investigation; Data curation; Writing – review & editing.

Qifeng Wang: Investigation; Data curation; Writing – review & editing.

Hao Jiang: Investigation; Data curation; Writing – review & editing.

Xiaoling Li: Investigation; Data curation; Writing – review & editing.

Mengzhao Wang: Investigation; Data curation; Writing – review & editing.

Yuan Chen: Investigation; Data curation; Writing – review & editing.

Hongda Lu: Investigation; Data curation; Writing – review & editing.

Yuzhong Duan: Investigation; Data curation; Writing – review & editing.

Rui Meng: Investigation; Data curation; Writing – review & editing.

Liyan Jiang: Investigation; Data curation; Writing – review & editing.

Lujun Zhao: Investigation; Data curation; Writing – review & editing.

Wenbin Zhu: Formal analysis; Data curation; Writing – review & editing.

Yunru Chen: Formal analysis; Data curation; Writing – review & editing.

Leah Szadkowski: Formal analysis; Data curation; Writing – review & editing.

Yuka Olivo: Formal analysis; Data curation; Writing – review & editing.

Suresh Senan: Conceptualization; Investigation; Data curation; Writing – review & editing.

Disclosure

Dr. S. Lu has received consulting fees for advisory boards/consultancy from AstraZeneca, Boehringer Ingelheim, GenomiCare, Hutchison MediPharma, InventisBio, Menarini, Pfizer, Roche, Yuhan Corporation, and Zai Lab; has received speaker fees from AstraZeneca, Hansoh Pharmaceuticals, Jiangsu Hengrui Pharmaceuticals, and Roche; has participated on advisory boards for AstraZeneca, Boehringer Ingelheim, GenomiCare, Hutchison MediPharma, InventisBio, Menarini, Pfizer, Phanes Therapeutics, Roche, Shanghai Fosun Pharmaceutical, Simcere Zaiming Pharmaceutical, Yuhan Corporation, and Zai Lab; has received research support from AstraZeneca, BeiGene, Hansoh Pharmaceuticals, Jiangsu Hengrui Pharmaceuticals, Hutchison MediPharma, and Roche; and is a director/board member of Innovent Bio. Dr. W. Zhu, Dr. Y. Chen, Ms. L. Szadkowski, and Ms. Y. Olivo are employees of AstraZeneca, and Dr. Y. Chen and Ms. Y. Olivo own stock in AstraZeneca. Prof. S. Senan has participated in advisory councils or committees for AstraZeneca, Lilly, and MSD; has received honoraria from AstraZeneca and MSD; has received grants or funds from AstraZeneca; has participated as a member of a steering committee for Lilly; and has been a principal investigator for AstraZeneca. The remaining authors declare no conflict of interest.