JAVELIN Solid Tumor (NCT01772004) is an international, multicohort, open-label phase I trial. In this phase Ib NSCLC expansion cohort, patients, who were unselected for PD-L1 expression, had histologically confirmed stage IV (per International Association for the Study of Lung Cancer classification, seventh edition)18 or recurrent NSCLC, no prior treatment for metastatic or recurrent disease, and no activating EGFR mutation or ALK translocation/rearrangement (tumors with non-squamous cell histology were tested if mutational status was unknown). General eligibility criteria for the JAVELIN Solid Tumor trial have been reported previously.12

Patients received avelumab 10 mg/kg by intravenous infusion every 2 weeks until disease progression, unacceptable toxicity, or other criteria for withdrawal were met. Patients were permitted to continue treatment despite progression according to the investigator’s decision and in agreement with the patient if no new symptoms appeared, existing symptoms did not worsen, Eastern Cooperative Oncology Group (ECOG) performance status did not decrease, and the investigator did not consider it necessary to administer a salvage therapy. Dose reductions were not permitted. Premedication with an antihistamine (diphenhydramine or equivalent) and acetaminophen was given 30 to 60 min before each infusion. Treatment was permanently discontinued for any grade ≥3 adverse event (AE) except for specified transient AEs (reported previously).12 13 Grade 2 AEs were managed by treatment delays of ≤2 subsequent omitted doses; events that did not resolve to grade ≤1 or recurred resulted in permanent discontinuation of avelumab. Clinical activity and safety were analyzed in all patients who received ≥1 dose of avelumab.

Tumor assessments were performed every 6 weeks for the first year and every 12 weeks thereafter by investigators according to Response Evaluation Criteria in Solid Tumors 1.1 (RECIST 1.1) and modified immune-related response criteria.19 Safety was assessed every 2 weeks at each visit, and AEs were graded according to National Cancer Institute Common Terminology Criteria for Adverse Events V.4.0. Immune-related AEs (irAEs) were identified using a prespecified list of Medical Dictionary for Regulatory Activities (MedDRA)-preferred terms followed by comprehensive medical review. Infusion-related reactions (IRRs) were identified using an expanded definition that included both a prespecified list of MedDRA-preferred terms (IRR, drug hypersensitivity, or hypersensitivity reaction) that occurred post infusion within 48 hours, and additional signs or symptoms that occurred on the day of infusion and resolved within 2 days.

PD-L1 expression was assessed using a proprietary immunohistochemistry assay (PD-L1 IHC 73-10 pharmDx; Dako, Carpinteria, California). In previous studies comparing the 73-10 PD-L1 assay with the 22C3 assay used in pembrolizumab trials, the 73-10 assay showed greater sensitivity, and the ≥80% PD-L1 cut-off for the 73-10 assay was found to be comparable to the ≥50% PD-L1 cut-off for the 22C3 assay (manuscript in press).20 21 PD-L1-positive status was predefined as PD-L1 expression of any intensity on ≥1% of tumor cells; PD-L1 expression status was also assessed using cut-offs of ≥50% and ≥80% in post hoc analyzes.

Prespecified endpoints assessed in this expansion cohort included confirmed best overall response, duration of response (DOR), and PFS based on investigator assessment according to RECIST 1.1, best overall response based on investigator assessment according to modified immune-related response criteria, OS, PD-L1 expression, and safety (all secondary endpoints in the overall JAVELIN Solid Tumor trial protocol).

Planned enrollment was 150 patients, which was based on the anticipated sample size required to estimate and provide 95% Clopper-Pearson CIs for potential objective response rates (ORRs). Safety data were summarized using descriptive statistics. Time-to-event endpoints were estimated using the Kaplan-Meier method, while 95% CIs for medians were calculated using the Brookmeyer-Crowley method. P values for the association between PD-L1 status and ORRs were determined using the Fisher’s exact test. Comparisons between other subgroups were not prespecified and are reported descriptively.

Between March 18, 2015, and November 19, 2015, 156 patients were enrolled from seven countries in North America, Europe, and Asia (table 1). The median age was 69.5 years (range, 41 to 90 years), 83 patients (53.2%) were male, and most patients (108 (69.2%)) had an ECOG performance status of 1. Tumor histology was squamous in 46 patients (29.5%) and non-squamous in 110 (70.5%). A total of 139 patients (89.1%) were ever smokers (current or previous) and 17 (10.9%) had never smoked. PD-L1 expression was evaluable in 111 patients (71.2%), of whom 88 (79.3%) had PD-L1-positive tumors based on a ≥1% cut-off. EGFR and ALK mutation status were unknown in 18 (11.5%) and 16 (10.3%), respectively. At data cut-off (February 15, 2017), the median follow-up was 18.6 months (range, 15 to 23 months). Patients received a median of 12 avelumab infusions (range, 1 to 49) over a median duration of 5.5 months (range, 0.5 to 22.5 months). At data cut-off, 26 patients (16.7%) remained on study treatment. Reasons for permanent treatment discontinuation were disease progression (82 (52.6%)), AEs (24 (15.4%)), withdrawal of consent (9 (5.8%)), death (9 (5.8%)), and other reasons (6 (3.8%)).

Of 156 patients, 3 (1.9%) had a confirmed complete response (CR) and 28 (17.9%) had a confirmed partial response (PR), resulting in an ORR of 19.9% (95% CI, 13.9% to 27.0%); 17 patients (10.9%) were not evaluable for response per RECIST (missing evaluations or not assessable; table 2). ORRs were observed in 17.4% (95% CI, 7.8% to 31.4%) of patients with squamous and 20.9% (95% CI, 13.7% to 29.7%) of patients with non-squamous histology. ORRs in ever smokers and never smokers were 20.9% (95% CI, 14.4% to 28.6%) and 11.8% (95% CI, 1.5% to 36.4%), respectively. In patients who had unknown EGFR or ALK mutation status, ORRs were 16.7% (95% CI, 3.6% to 41.4%) and 18.8% (95% CI, 4.0% to 45.6%), respectively. Of the three patients who had a CR, two had a preceding PR. Response was ongoing in 15 of 31 patients at data cut-off (figure 1A). The median time to response was 11.4 weeks (range, 5.1 to 29.6 weeks) and the median DOR in patients with confirmed responses was 12.0 months (95% CI, 6.93 months to not estimable). Of 142 patients who were evaluable for changes in target lesions (ie, those with a baseline and on-study tumor assessment available), 93 (65.5%) had a reduction in tumor size of any level, while 43 (30.3%; including 12 patients with unconfirmed responses) had ≥30% reduction (figure 1B and online additional file 1), with no notable trends based on tumor histology or smoking status (online additional file 2). The median PFS in all patients was 4.0 months (95% CI, 2.7 to 5.4 months) and the 6-month PFS rate was 38.5% (95% CI, 30.7% to 46.3%) (online additional file 3). The median OS was 14.1 months (95% CI, 11.3 to 16.9 months) and the 12-month OS rate was 56.6% (95% CI, 48.2% to 64.1%) (online additional file 3).

In analyzes of efficacy in subgroups with PD-L1 expression levels of ≥1%, ≥50%, and ≥80%, ORRs by RECIST 1.1 were 19.3% (95% CI, 11.7% to 29.1%), 22.6% (95% CI, 12.3% to 36.2%), and 26.3% (95% CI, 13.4% to 43.1%), respectively (table 3). Patients who were not evaluable for PD-L1 status had an ORR of 26.7% (95% CI, 14.6% to 41.9%). Best change in target lesions by PD-L1 status is shown in online additional file 2. In PD-L1+ and PD-L1− subgroups (≥1% and <1% cut-offs, respectively), median PFS was 4.0 months (95% CI, 2.7 to 6.0 months) and 1.5 months (95% CI, 1.35 to 5.4 months), and median OS was 14.1 months (95% CI, 11.2 to 18.2 months) and 11.3 months (95% CI, 1.6 to not estimable), respectively.

In analyzes of response by immune-related criteria, four additional patients, who did not achieve a response according to RECIST 1.1, had an objective response, resulting in an immune-related ORR of 22.4%, including immune-related CRs in four patients (2.6%) and immune-related PRs in 31 patients (19.9%). In PD-L1 subgroups with expression levels of ≥1%, ≥50%, and ≥80%, immune-related ORRs were 23.9% (95% CI, 15.4% to 34.1%), 28.3% (95% CI, 16.8% to 42.3%), and 34.2% (95% CI, 19.6% to 51.4%), respectively. In the overall population, median PFS based on immune-related criteria was 6.9 months (95% CI, 5.4 to 9.7 months).