Novel Combinations of Immunotherapies or DNA Damage Repair Inhibitors in Platinum-Refractory Extensive-Stage Small Cell Lung Cancer: The Phase II BALTIC Study

Abstract

Purpose: The phase II, multiarm, signal-searching BALTIC study (NCT02937818) assessed novel treatment combinations for platinum-refractory/resistant extensive-stage small cell lung cancer (ES-SCLC).

Patients and methods: Patients with ES-SCLC with progressive disease during or within 90 days of completing first-line platinum-based chemotherapy received one of three regimens: durvalumab plus tremelimumab followed by durvalumab monotherapy (arm A), adavosertib plus carboplatin (arm B), or ceralasertib plus olaparib (arm C). The primary endpoint was the objective response rate. Prespecified exploratory biomarker analyses were conducted in arms A and C.

Results: In arm A (n = 41), arm B (n = 10), and arm C (n = 21), the confirmed objective response rates were 7.3%, 0%, and 4.8%, respectively. Safety profiles in all arms were consistent with those of the individual drugs. In arm A, patients with PD-L1 expression (tumor cells or immune cells) ≥1% seemed to have a greater likelihood of achieving disease control with durvalumab plus tremelimumab than those with PD-L1 (tumor cells and immune cells) <1%, and lower baseline ctDNA and reduction in the on-treatment ctDNA level were both associated with longer overall survival. Among patients treated with ceralasertib plus olaparib in arm C, specific immune response-relevant circulating chemokines and cytokines were identified as early biomarkers of survival and pharmacodynamic biomarkers.

Conclusions: In BALTIC, all combination regimens demonstrated tolerable safety profiles, but antitumor activity was limited in refractory/resistant ES-SCLC. Among patients treated with durvalumab plus tremelimumab, an association of on-treatment reduction in ctDNA with longer overall survival suggests the potential use of ctDNA as a surrogate of treatment response, warranting further investigation.

Figure 1.

Best percentage change from baseline in selected target lesion size per investigator assessment—full analysis set. A, Arm A, durvalumab + tremelimumab. B, Arm B, adavosertib + carboplatin. C, Arm C, ceralasertib + olaparib. In arms A, B, and C, respectively, 10, 3, and 4 patients had no measurements of target lesion size after baseline and were excluded from the plots. Best change in target lesion size is the maximum reduction from baseline or minimum increase from baseline in the absence of a reduction. Response includes confirmed CR or PR. Dotted reference lines at −30% and 20% indicate thresholds for PR and PD, respectively.

Figure 2.

Response and OS by PD-L1 expression status, baseline ctDNA level, and by change in ctDNA levels (ΔMaxVAF) from baseline to on treatment—arm A PD-L1 BEP and ctDNA BEP. A, Best percentage change in tumor size from baseline in patients evaluable for response (n = 32)* by PD-L1 expression status. B, Kaplan–Meier curves for OS by PD-L1 expression status. C, Best percentage change in tumor size from baseline in patients evaluable for response (n = 32)* by baseline ctDNA level. D, Kaplan–Meier curves for OS, by tertiles of baseline ctDNA level [low (MaxVAF ≤ 21.62%; n = 9)], medium (MaxVAF > 21.62% to <54.86%; n = 8), or high (MaxVAF ≥ 54.86%; n = 9). E, Best percentage change in tumor size from baseline in patients evaluable for response (n = 32)* by change in the ctDNA level from baseline to on treatment [MaxVAF decrease (ΔMaxVAF < 0) or increase (ΔMaxVAF > 0)]. F, Kaplan–Meier curves for OS, by change in the ctDNA level from baseline to on treatment [decrease (negative ΔMaxVAF) or increase (positive ΔMaxVAF)]. *Nine patients did not have postbaseline tumor size measurements, so are not included in the figure; in A, C, and E, dashed reference lines at −30% and 20% indicate thresholds for PR and PD.

Figure 3.

Circulating biomarkers associated with ceralasertib + olaparib treatment response and pharmacodynamics—arm C circulating biomarker BEP. A, Baseline cytokine and chemokine distribution according to OS. B, Kaplan–Meier curves for OS by baseline PD-L1 expression status. C, Kaplan–Meier curves for OS by baseline CXCL5 expression status. D, Change in GDF15 expression over time in cycles 1 and 2. E, Change in TR expression over time in cycles 1 and 2. F, Change in MCP1 expression over time in cycles 1 and 2. BLM, bleomycin; CCL23, C–C motif chemokine ligand 23; CXCL, C–X–C motif chemokine ligand; FABP4, fatty acid–binding protein 4; HGF, hepatocyte growth factor; MUC16, mucin 16; PAI, plasminogen activator inhibitor; TNFRSF14, TNF receptor superfamily member 14; vWF, von Willebrand factor.