In the phase III CASPIAN study, durvalumab added to etoposide and platinum-based chemotherapy delayed the development of new lesions and improved patient-reported outcomes compared to etoposide and platinum-based therapy alone in untreated patients with extensive-stage small cell lung cancer.
Marina Chiara Garassino, MD
In the phase III CASPIAN study, durvalumab (Imfinzi) added to etoposide and platinum-based chemotherapy delayed the development of new lesions and improved patient-reported outcomes (PROs) compared to etoposide and platinum-based therapy alone in untreated patients with extensive-stage small cell lung cancer (ES-SCLC).1
“Numerically fewer patients developed new lesions at first progression with the combination of durvalumab plus etoposide and platinum over the chemotherapy-alone arm, including in lung, liver, and bone, and the incidence of new brain/central nervous system [lesions] was similar between the 2 arms,” said Marina Chiara Garassino, MD, of Fondazzione IRCCs Instituto Nazionale die Tumori, Milan, Italy, in a presentation during the 2019 ESMO Congress.
In addition, PD-L1 expression was found to be low in this patient setting and did not appear to be a predictive biomarker for the benefit of the durvalumab combination on clinical outcomes.
In the phase III, global, open-label multicenter CASPIAN study, 805 treatment-naïve patients with ES-SCLC were randomized 1:1 to 4 cycles of etoposide and platinum-based chemotherapy plus durvalumab at 1500 mg every 3 weeks followed by maintenance durvalumab every 4 weeks (n = 268), or up to 6 cycles of etoposide/platinum-based chemotherapy every 3 weeks plus prophylactic cranial irradiation at the investigator’s discretion (n = 269). More than half (56.8%) of patients in the control arm received 6 cycles of etoposide-platinum.
Following a preplanned interim analysis by an Independent Data Monitoring Committee, a third arm of the trial, in which patients received durvalumab, tremelimumab, and etoposide/platinum-based chemotherapy remains blinded and continues to final analysis.
The median age was between 62 and 63 years; about 70% of patients were male, about 45% were current smokers, about 47% were former smokers, and 10% of patients had brain/central nervous system metastases. PCI was administered to 8% of patients in the chemotherapy-alone arm. Patients were stratified by whether they received carboplatin or cisplatin as their planned platinum-based agent.
Prior CASPAIN results showed that durvalumab in combination with etoposide plus either cisplatin or carboplatin in patients with treatment-naïve ES-SCLC was found to significantly improve overall survival (OS), the trial’s primary endpoint, versus chemotherapy alone without additional toxicity.
The median OS increased from 10.3 months with etoposide and platinum-based chemotherapy alone to 13.0 months with the addition of durvalumab, translating to a 27% reduction in the risk of death (HR, 0.73; 95% CI, 0.591-0.909;P= .0047) at a median follow-up of 14.2 months.2
The earlier data also showed that the median progression-free survival (PFS) was 5.1 months with durvalumab versus 5.4 months without (HR, 0.78; 95% CI, 0.645-0.936), which was not formally tested for statistical significance. Analysis of the 12-month PFS rate showed an advantage favoring durvalumab (17.5% vs 4.7%).
Secondary endpoints included PFS, objector response rate, safety and tolerability, and PROs; PD-L1 biomarker analysis was an exploratory endpoint.
In the analysis presented during the 2019 ESMO Congress, findings showed that progression patterns were similar between the two arms, but numerically fewer patients in the durvalumab arm developed new lesions at first progression (41.4% vs 47.2%). The rates of total progression events were 84.3% in the durvalumab/chemotherapy arm versus 86.6% of the chemotherapy-alone arm, and progressionas defined by RECIST criteria—occurred in 71.6% of the durvalumab-plus-chemotherapy arm compared with 72.1% of the chemotherapy-alone arm. The rates of target lesion progression were 42.9% and 39.4%, respectively, and the rates of nontarget lesion progression were 24.6% and 2.7%, respectively. Death in the absence of progression occurred in 12.7% and 14.5%, respectively.
New lesions appeared in 41.4% of the durvalumab-chemotherapy arm and 47.2% of the chemotherapy-alone arm. In the durvalumab/chemotherapy arm, there were numerically fewer new lung lesions (8.6% vs 15.2%), liver lesions (5.6% vs 8.9%), and bone metastases (4.5% vs 7.1%) than in the etoposide-platinum arm.
There was a similar number of brain metastases between the durvalumab/chemotherapy and chemotherapy-alone arms (11.6% vs 11.5%) but prophylactic cranial irradiation was permitted only in the control arm, said Garassino.
The rates of new lesions in regional lymph nodes were 5.6% in the durvalumab/chemotherapy arm versus 4.5% in the chemotherapy-alone arm.
“The relevance of PD-L1 expression in SCLC and optimal cutoff still has to be defined,” said Garassino.
In CASPIAN, tumor tissue was mandated at screening if available. PD-L1 expression on immune cells and tumor cells was assessed using the Ventana PD-L1 (SP263) assay. Tumor tissue was available and tested for PD-L1 expression in 277 (51.6%) of randomized patients across the durvalumab/chemotherapy and chemotherapy-alone arms. Baseline characteristics of these 277 patients resembled those of the intent-to-treat population, and PD-L1 was analyzed as a continuous variable.
“Eighty percent of patients had negative PD-L1 on tumor cells and 58.1% on the immune cells,” said Garassino. Due to the low PD-L1 expression, a 1% cutoff was used in a post-hoc analysis of OS based on PD-L1 expression. Some 5.1% of patients had PD-L1 expression ≥1% on tumor cells and 22.4% had ≥1% PD-L1 expression on immune cells.
“PD-L1 is not a predictive biomarker to select patients to receive chemotherapy plus durvalumab or chemotherapy,” she said, as the durvalumab regimen was associated with improved OS versus etoposide-platinum without durvalumab, regardless of PD-L1 expression with a ≥1% cutoff. “No significant interaction was observed for OS and PFS and also response rate.”
Symptoms, health-related quality of life, and functioning were assessed using EORTC QLQ-C30/LC13 with changes from baseline analyzed by time to deterioration per Cox proportional hazards. Time to deterioration (TTD) was defined as the time from randomization until the first clinically meaningful deterioration confirmed at the following visit or death.
Baseline PROs scores were comparable between the experimental and control arms across all symptoms and functional domains.
The combination of durvalumab plus chemotherapy was favored across all functioning scales. In particular, the Global health status/QoL was superior with the addition of durvalumab over chemotherapy alone (HR, 0.81; 95% CI, 0.626-1.054), and the TTD of physical functioning (HR, 0.75; 95% CI, 0.581-0.970), cognitive functioning (HR, 0.61; 95% CI, 0.472-0.776), emotional functioning (HR, 0.61; 95% CI, 0.464-0.800), role functioning (HR, 0.71; 95% CI, 0.550-0.904), and social functioning (HR, 0.70; 95% CI, 0.549-0.897) favored combination treatment.
Baseline symptom scores were also comparable between arms. Durvalumab plus etoposide and platinum-based chemotherapy was favored across all symptoms collected on the QLQ-C30 and QLQ-LC13.