Background

As single agents, durvalumab (MEDI4736), a human IgG1 anti-PD-L1 antibody, and tremelimumab, a human IgG2 anti-CTLA-4 antibody, have shown acceptable safety profiles and antitumor activity. Similar to other anti-PD-L1/anti-PD-1 monotherapies, durvalumab has shown greater objective tumor response rates in PD-L1-positive patients compared with PD-L1-negative patients. Anti-CTLA4 therapies activate T-cells and may increase immune infiltrate and PD-L1 expression in tumor cells and tumor infiltrating immune cells. Thus, combination therapy with durvalumab and tremelimumab could be active in NSCLC regardless of baseline PD-L1 expression.

Methods

This is a phase 1, open-label, dose-escalation/expansion study (NCT02000947) of D+T in patients with Stage III/IV NSCLC (any number of prior lines of therapy; immunotherapy-naïve). The primary endpoint is safety and tolerability; secondary endpoints include investigator-reported RECIST v1.1 response. PD-L1 expression was tested retrospectively using an immunohistochemical assay (Ventana).

Results

As of 1 June 2015, 102 patients received treatment in the dose escalation phase; combinations of durvalumab [3 mg/kg (D3) to 20 mg/kg (D20) every 2 (q2w) or 4 weeks (q4w)] and tremelimumab [1 mg/kg (T1) to 3 mg/kg (T3)] q4w, plus a D15 + T10 combination, were explored. Across all cohorts, 80% and 42% of patients had ≥1 treatment-related AE (any Grade and Grade 3/4, respectively); 28% discontinued treatment due to a related AE. A greater frequency of AEs, without a corresponding increase in tumor response, was seen with increasing T dose. In the combined T1 cohort (D10–D20), 73% and 30% of patients had ≥1 related AE (any Grade and Grade 3/4, respectively); 16% discontinued treatment due to a related AE. There were 3 treatment-related deaths (myasthenia gravis, T1; pericardial effusion, T1; neuromuscular disorder, T3).

84 patients (73 EGFR/ALK wild-type; 77 non-squamous; 48 with ≥2 prior lines of therapy) were evaluable for response (Table 1). The overall response rate (confirmed+unconfirmed) was 25%. Higher response rates were observed in those with 1 vs ≥2 prior therapies. Response rates do not appear dependent on PD-L1 status: 35% (PD-L1-positive), 22% (PD-L1-negative, <25% tumor cell staining) and 33% (PD-L1-negative, 0% tumor cell staining). Similar findings were observed for the combined T1 cohort. D+T also showed good durability of response similar to that seen for monotherapy.

Table 1 Response rates (Confirmed/unconfirmed with ≥16 weeks follow-up)
Full size table

Conclusions

D+T at selected phase 3 dose (D20, T1) has a manageable tolerability profile and anti-tumor activity in NSCLC. Unlike anti- PD-1/PD-L1 monotherapies, the combination of D+T appears to be active regardless of PD-L1 status, including even in patients with no tumor cell membrane PD-L1 staining, a setting where patients would not be expected to derive significant benefit from anti-PD-1/PD-L1 monotherapy over current standard of care [1, 2].