SWOG S1400A (NCT02154490): A phase II study of durvalumab for patients with previously treated stage IV or recurrent squamous cell lung cancer (Lung-MAP sub-study)

Abstract

Purpose:

The objective of Lung-MAP sub-study S1400A was to evaluate the response rate to durvalumab, an anti PD-L1 antibody, in squamous non-small cell lung cancer (SqNSCLC) patients.

Methods:

Patients, who progressed on at least one prior platinum-based chemotherapy were eligible. The study was designed as a phase II/III trial comparing durvalumab to docetaxel but was modified to a single arm phase II trial with primary endpoint of objective response when immunotherapy became an approved treatment.

Results:

116 patients were registered to this sub-study; 78 to durvalumab and 38 to docetaxel. Of the 78, 9 were ineligible and 1 was not evaluable for endpoints. Responses were achieved in 11 among the 68 eligible and evaluable patients on durvalumab (16% ORR, 95% Confidence Interval [CI] 7%–25%). Disease control rate was 54% (95% CI 43%–66%), median OS was 11.6 months (95% CI 10.2–14.3 mos), and median PFS was 2.9 mos (95% CI 2.0–4.0 mos). PD-L1 data was available for 43 patients on durvalumab, with 14 patients (33%) PD-L1 positive (≥25%) and 2 responses (14% ORR, 95% CI 0%–33%), DCR was 57% (95% CI 31%–83%), median OS and PFS were 10.7 mos (95% CI 9.2–14.3 mos) and 2.3 mos (95% CI 1.4–4.2 mos), respectively. Grade ≥3 treatment-related AEs occurred in 22 (32%) patients on durvalumab with 6 discontinuing due to Drug-related AEs (9%, 95% CI 2%–16%).

Conclusions:

Durvalumab shows single agent activity and toxicities in this sub-group of patients that is comparable to other anti PD-1/PD-L1 antibodies.

Introduction

Squamous cell disease accounts for 20% of NSCLC with approximately 40,000 new cases diagnosed each year in the United States. Identifying targeted agents with activity in advanced squamous non-small cell lung cancer (SqNSCLC) has been challenging. Patients with this histology have not benefited from the advances in molecular- and genomics-based treatments that are available for patients with adenocarcinoma of the lung.

The Lung Master Protocol (Lung-MAP, SWOG S1400), led by SWOG, was developed as a cooperative effort by the National Cancer Institute (NCI) and National Clinical Trials Network (NCTN). The Lung-MAP study is an umbrella protocol which contains a screening component and multiple independently conducted and analyzed treatment sub-studies. The overarching hypothesis is that the umbrella master protocol will establish genomic screening for a large population of previously treated squamous lung cancer patients to evaluate targeted therapies (or combinations) in biomarker-matched subpopulations (biomarker-driven sub-studies) and therapies thought to have broad activity for patients not eligible for the biomarker-driven sub-studies (non-match sub-studies) within one infrastructure (“umbrella”). The ultimate goal of Lung-MAP is lead to efficient approval of efficacious regimens. Here we report on results of SWOG S1400A, a phase II study evaluating durvalumab in patients with chemo-refractory SqNSCLC not eligible for the biomarker-driven sub-studies conducted as part of Lung-MAP.

S1400A was the first non-match sub-study included in Lung-MAP and was activated in June of 2014 along with 4 biomarker-driven sub-studies. Patients without one of the biomarker-driven sub-study target alterations or who did not satisfy the eligibility criteria for a biomarker-driven sub-study were eligible to participate in S1400A. At the time that the trial was inaugurated, no checkpoint inhibitor had been approved for treatment in lung cancer.

Patients and Methods

Patients who had received at least one line of platinum-based chemotherapy for SqNSCLC were eligible for the Lung-MAP screening study. Patients were aged 18 or older, had histologically documented SqNSCLC (stage IIIB or IV) or recurrent or progressive disease following multimodal therapy. Initially, patients were allowed to have only a single line of prior therapy and Zubrod PS 0–2, however, the study was amended to allow any number of lines of prior therapy and restrict to PS 0–1 in May 2015. Patients could be registered at the time of progression or could be pre-registered while receiving therapy. Availability of adequate tumor tissue with at least 20% tumor cells and ≥ 0.2 mm³ tumor volume was required. Adequate organ function as demonstrated by an absolute neutrophil count (ANC) ≥ 1500 mcL, platelet counts ≥ 100,000 mcL and hemoglobin ≥ 9 g/dL was required. Other criteria included: adequate hepatic function as defined by serum bilirubin ≤ Institutional Upper Limit of Normal (IULN) and either alanine aminotransferase (ALT) or aspartate aminotransferase (AST) ≤ 2 × IULN within 28 days prior to sub-study registration if no liver involvement; and for patients with liver metastases, bilirubin and either ALT or AST ≤ 5 × IULN and a calculated creatinine clearance ≥ 50 mL/min. Measurable disease (by Response Evaluation Criteria in Solid Tumors RECIST 1.1) was required within 28 days of sub-study registration. Key exclusion criteria were: presence of an EGFR mutation or ALK gene fusion, mixed squamous/non-squamous histology, active or previously documented autoimmune disease, and presence of untreated central nervous system (CNS) metastases. No patient could be enrolled at an institution prior to review and approval by either the local Institutional Review Board or by the National Cancer Institute Central IRB. Written informed consent was required from all patients prior to enrollment in the master protocol and a separate consent for the specific sub-study. Full eligibility criteria are discussed in the appendix.

All patients had tissue that underwent NGS testing by Foundation Medicine, Inc. Mutational analysis was performed on archival formalin-fixed paraffin-embedded (FFPE) tumor specimens in a Clinical Laboratory Improvement Amendments (CLIA)-certified and College of American Pathologists (CAP)-accredited laboratory (Foundation Medicine, Cambridge, MA). Genomic DNA (≥50 ng) was extracted from FFPE specimens and sonicated to ~200bp fragments. Material underwent whole-genome shotgun library construction and hybridization-capture of at least 236 genes and selected introns of 19 genes involved in rearrangements. Using the Illumina® HiSeq 2000, 2500, and 4000 platforms, hybrid-capture-selected libraries were sequenced using 49 × 49bp paired-end reads to high uniform depth. Sequence data were processed using a customized analysis pipeline designed to accurately detect base substitutions, small insertions and deletions, focal copy number amplifications, homozygous gene deletions, and genomic rearrangements. Tumor mutational burden (TMB) was calculated as the number of somatic, coding, short variants, excluding known driver mutations, per megabase of the genome interrogated. PD-L1 testing was retrospectively added to the protocol in May 2015 with Revision # 2. PD-L1 testing was performed using the immunohistochemistry (IHC)-based assay for PD-L1 determination developed by Medimmune in partnership with Ventana Medical Systems Inc., a CAP-accredited/CLIA-certified laboratory (Tucson, AZ). Tumors with at least 25% of tumor cells with membranous staining, at any intensity, for PD-L1 expression were considered PD-L1 positive.

Durvalumab was administered as 10 mg/kg infusions every 2 weeks until progression or for up to 1 year. Patients able to complete the first year of treatment could be re-treated upon progression with durvalumab for up to an additional year of treatment. Patients who discontinued treatment with durvalumab prior to completing 12 months of therapy, regardless of reason, were not eligible for re-treatment. Disease assessment occurred every 3 cycles, and treatment could continue beyond progression as long as in the opinion of the treating investigator the patient is continuing to clinically benefit from treatment. Patients with progression of disease that is confirmed by a second determination of progression at least 4 weeks from the first documentation of progression had to be removed from protocol treatment.

Statistical Considerations

S1400A was activated in June 2014 as a phase II/III trial with randomization between durvalumab and docetaxel and co-primary objectives to compare OS and PFS between the arms with an accrual goal of 400 patients. In response to the approval of nivolumab in the second-line setting, the study was amended to be a single arm phase II trial in May 2015 with response per RECIST 1.1 (confirmed and unconfirmed, complete and partial) as the primary endpoint. The revised accrual goal to the study was 100 patients including at least 30 PD-L1 positive patients. The new design specified that the lower bound for the confidence intervals excluding a 10% response rate in both the overall and within the PD-L1 positive will be considered promising evidence of an improved response rate with durvalumab over docetaxel.

Secondary objectives include an evaluation of investigator-assessed progression-free survival (IA-PFS), overall survival (OS), and toxicities by CTCAE 4.03 in all patients and an evaluation of IA-PFS and OS in the subset determined to be PD-L1 positive. Immune-related adverse events were defined based on the set of toxicities commonly attributed as immune related in the literature. ^1–3 IA-PFS was defined as the duration from randomization/registration to first occurrence of progression per RECIST 1.1, symptomatic deterioration, or death due to any cause. IA-PFS for patients last known to be without an event were censored at the date of last disease assessment. OS was defined as the duration from randomization/registration to death due to any cause with censoring at the date of last contact. Duration of response was defined as the duration from first occurrence of a response (confirmed or unconfirmed) to first occurrence of progression per RECIST 1.1, symptomatic deterioration, or death due to any cause with censoring following the same rules as for IA-PFS. Additional objectives include a comparison of IA-PFS, OS, response, and toxicity between durvalumab versus docetaxel in the subset of patients enrolled during the randomized portion of the study.

Binary proportions are summarized along with 95% confidence intervals. The method of Kaplan-Meier was used to estimate survival distributions and the Brookmeyer-Crowley method was used to estimate confidence intervals around medians. Logistic regression was used to evaluate factors associated with response. A Cox proportional hazards model was used to evaluate the potential prognostic factors and their association with IA-PFS and OS and to compare treatment arms among the randomized subset. A threshold of 0.05 was used to interpret statistical significance.

Results

The study was opened to accrual in June 2014. The docetaxel arm was closed to accrual on May 25, 2015 when the study was converted to a single arm design. The entire trial was administratively closed in December 2015 by the National Cancer Institute despite not having reached the accrual goal. While the study was actively accruing, 511 patients were screened for Lung-MAP and 116 patients registered to S1400A; 78 patients were randomized/assigned to durvalumab and 38 to docetaxel.

Of the 78 patients randomized or assigned to receive durvalumab, nine patients were ineligible due to: inadequate baseline data (3), prior treatment (2), Hepatitis C positivity (1), brain scan not done within protocol-specified timeframe (1), inadequate baseline laboratory values (1), and had not progressed on most recent line of therapy (registered before pre-screening was an option) (1). One additional patient did not receive any protocol treatment due to symptomatic deterioration and is not included in any analyses.

Of the 38 patients randomized to docetaxel, 30 were eligible and analyzable for response and toxicity. Two patients were ineligible due to no evidence of measurable disease, two patients due to brain metastases and 6 patients did not receive any protocol treatment.

Baseline characteristics are shown in Table 1. Average age of patients was 66 in the durvalumab arm and 71 in the docetaxel arm of the trial. About 12% of the patients had Zubrod PS 2, and 70% were classified as former smokers. Fifty-seven percent of patients had received exactly one line of prior therapy and 19% had received two or more lines of therapy for stage IV disease.

Table 1.

Patient Demographics

	Durvalumab (n=68)	Docetaxel (n=30)
Age
Median (Range)	66 (35–92)	71 (50–83)
>= 65	36(53%)	21(70%)
Female Gender	27(40%)	10(33%)
Race/Ethnicity
White	55(81%)	27(90%)
Black	6(9%)	1(3%)
Asian	5(7%)	1(3%)
Multi-Racial	0(0%)	1(3%)
Unknown race	2(3%)	0(0%)
Hispanic	2(3%)	0(0%)
Number of lines of prior therapy for stage IV disease
0	16(24%)	N/A*
1	39(57%)
2	11(16%)
3+	2(3%)
Zubrod Performance Status
0	18(26%)	10(33%)
1	42(62%)	16(53%)
2	8(12%)	4(13%)
Weight Loss in Past 6 months
<5%	48(71%)	29(97%)
5–10%	12(18%)	1(3%)
10–<20%	6(9%)	0(0%)
>=20%	2(3%)	0(0%)
Smoking Status
Current Smoker	17(25%)	6(20%)
Former Smoker	47(69%)	21(70%)
Never Smoker	4(6%)	3(10%)
Brain Metatases Reported at Baseline	4(6%)	1(3%)
PD-L1 **
≥25 (positive)	14(33%)
< 25 (negative)	29(67%)
Tissue not evaluable	10(15%)
Not tested	15(22%)	30 (100%)

^*While the docetaxel arm was enrolling, the eligibility stated that patients must have received exactly one line of prior therapy for stage IV disease or platinum-based chemotherapy for stage I-III with progression within a year of completing therapy. Therefore, the number of lines of prior therapy for stage IV disease was not captured prior to the revision in May 2015 that closed the docetaxel arm to accrual.

^**PD-L1 testing was only performed on patients enrolled on the durvalumab arm.

PD-L1 results were available for 43 (63%) patients in the durvalumab arm. PD-L1 was not evaluated for patients on the docetaxel arm. Of the patients with known PD-L1 expression, 14 (33%) were classified as high (≥25%) and 29 (67%) were considered as having low expression (<25%).

Gene alterations targeted in sub-studies open in the original version of the umbrella Lung-MAP included PI3K mutations (S1400B), cell cycle gene alterations (S1400C), and FGFR alterations (S1400D) and these were seen in a minority of patients included in the currently reported study. Patients who were eligible based on their biomarker profile for one of the sub studies but did not qualify based on other eligibility criteria, were able to be enroll in S1400A. Additional gene alterations and their frequencies were as previously reported⁴ and are included in Table 2.

Table 2.

Gene Alterations Detected on FMI NGS Screening

	N (%) (n=98)
Other Sub-study Biomarker Eligible
PI3K + (S1400B)	3(3%)
CELL CYCLE GENE ALTERATION + (S1400C)	0(0%)
FGFR + (S1400D)	5(5%)
Tumor Mutation Burden Score (N=68) **
Median	10.88
Range	1.21–59.25
Interquartile Range	4.84–15.72
<10	28(48%)
≥10	30(52%)
Not Evaluable	10(15%)
Other Concomitant Gene Alterations
Short Variants
TP53	87(89%)
MLL2	20(20%)
CDKN2A	18(18%)
NFE2L2	15(15%)
ARID1A	11(11%)
LRP1B, PTEN	10(10%)
RB1	9(9%)
BRCA2, FBXW7	6(6%)
CREBBP, NF1, SETD2	5(5%)
APC, ASXL1, PBRM1, PIK3CA, SMARCA4, STK11	4(4%)
CDK12, KRAS, NOTCH1, PALB2, PTCH1, SPEN, TSC2, WT1	3(3%)
ARID2, ATM, ATRX, BAP1, BRAF, BRIP1, CASP8, CIC, EGFR, ERBB4, KEAP1, MLH1, MYST3, NCOR1, PIK3R1, RAD51	2(2%)
ABL1, ATR, AXL, BCORL1, CTNNB1, DNMT3A, EP300, EPHA3, EPHA5, EPHB1, ERBB2, FAT3, FGFR2, FGFR3, GNAQ, HGF, IKZF1, IL7R, KDM5A, KDM5C, KDM6A, KIT, MAP3K1, MED12, MUTYH, MYCN, NOTCH2, NOTCH3, PARP4, PDGFRA, PPP2R1A, PRKDC, RAD50, RAD51C, RNF43, RUNX1, RUNX1T1, TBX3, TET2, TNFAIP3, TRRAP, TSC1	1(1%)
Copy Number Alterations
SOX2	25(26%)
PIK3CA	17(17%)
CDKN2A	16(16%)
CDKN2B, FGF12	14(14%)
CRKL, LRP1B, PTEN	5(5%)
EGFR, FGF10, MYC, MYST3	4(4%)
FGFR1, NFKBIA, REL, RICTOR, ZNF703	3(3%)
AKT2, BCL2L2, HGF, JAK2, KDR, KIT, MEN1, NKX2–1, PDGFRA, SRC	2(2%)
AKT1, AKT3, ARFRP1, AURKA, AURKB, CCNE1, CDK6, CTNNA1, EPHB1, ERBB2, GNAS, IRS2, KDM5A, KDM6A, KRAS, MAP2K4, MCL1, MDM2, MET, NF1, PBRM1, RB1, RET, RPTOR, STK11	1(1%)
Rearrangements
LRP1B	5(5%)
MLL2	3(3%)
CDKN2A, CTNNA1, EGFR, NF1, PTEN, STK11	1(1%)

^**TMB was only evaluated on the durvalumab arm (n=68).

On the durvalumab arm, 10 patients completed the initial year of treatment as planned, 6 discontinued due to adverse events, 1 refused further treatment unrelated to adverse events, 49 came off due to progressive disease, 1 due to death, and 1 due to physician decision. Of the 10 patients who completed treatment, 4 registered to be re-treated with durvalumab after progression. Of these 4, two completed the additional year of treatment, and 2 are off treatment due to progression or death. The median (range) number of cycles for patients on the durvalumab arm was 7 (1–27) and 3 (1–6) for patients on the docetaxel arm.

Toxicities were in line with the known side effect profiles of both agents. On the durvalumab arm, 68 patients were evaluable for adverse events and 30 patients on the docetaxel arm were evaluable for adverse events. Treatment-related adverse events were reported in 60 (88%) of the durvalumab treated group and 30 (100%) of the docetaxel group. Six patients (9%) on the durvalumab arm and three (10%) in the docetaxel arm discontinued treatment due to toxicity. On the durvalumab arm, there was one treatment-related death due to bronchopulmonary hemorrhage (1%), 4 patients (6%) experienced Grade 4 treatment-related adverse events (one case each of dyspnea, decreased ejection fraction, hyponatremia and decreased lymphocyte count) and 17 patients (25%) experienced Grade 3 treatment-related adverse events; 32% of patients with a Grade 3 or higher treatment-related adverse event. On the docetaxel arm, there was one treatment-related death due to sepsis (3%), 9 patients (30%) experienced Grade 4 treatment-related adverse events and 12 patients (40%) experienced Grade 3 treatment-related adverse events; 22 patients (73%) experienced a Grade 3 or higher treatment-related adverse event (Table 3).

Table 3.

Adverse Events Attributed to Treatment

	Durvalumab (n = 68) Grade
	1–2	3	4	5
NON-IMMUNE-RELATED ADVERSE EVENTS Abdominal pain	1(1%)	1(1%)
Anemia	11(16%)	4(6%)
Anorexia	13(19%)
Bone pain		1(1%)
Bronchopulmonary hemorrhage				1(1%)
Ejection fraction decreased	1(1%)		1(1%)
Hypercalcemia	1(1%)	1(1%)
Hyponatremia	5(7%)	2(3%)	1(1%)
Leukocytosis		1(1%)
Lung infection		1(1%)
Nausea	16(24%)	1(1%)
Syncope		1(1%)
Urinary tract infection		1(1%)
Weight loss	10(15%)	1(1%)
IMMUNE-RELATED ADVERSE EVENTS	1–2	3	4	5
ALT increased	5(7%)	2(3%)
AST increased	5(7%)	2(3%)
Alkaline phosphatase increased	7(10%)
Diarrhea	12(18%)
Dyspnea	5(7%)	5(7%)	1(1%)
Fatigue	25(37%)	3(4%)
Hyperthyroidism	9(13%)
Hypoxia		2(3%)
Infusion related reaction	1(1%)	1(1%)
Lymphocyte count decreased	5(7%)	2(3%)	1(1%)
Rash maculo-papular	10(15%)
MAX. GRADE ANY IRAE	34(50%)	11(16%)	2(3%)	0
MAX. GRADE ANY ADVERSE EVENT	38(56%)	17(25%)	4(6%)	1(1%)

Grade 3 immune-related adverse events (irAEs) occurred in 11 (16%) and Grade 4 irAEs occurred in 2 (3%) patients of the durvalumab group. These included elevated ALT, AST, dyspnea, hypoxia, fatigue, decreased lymphocyte count and infusion related reaction (Table 3).

There were 11 responses among the 68 eligible and analyzable patients on the durvalumab arm (ORR= 16% [95% CI 7–25%], Figure 1), the median time to response was 3.6 months (95% CI 2.8–5.7 mos). The median duration of response with durvalumab among responders was 9.3 months (95% CI 4.4–18.6 mos). There were 2 responses among the 14 patients in the PD-L1 high group (ORR=14% [95% CI 0–33%]) and 2 responses (OR=7% [95% CI 0–16%]) among the 29 PD-L1 negative patients. Tumor Mutational Burden (TMB) data are available for 58 (85%) on the durvalumab arm with a median score of 10.88 mutations/mB.

Figure 1. Waterfall Plot of Response to Durvalumab.

All patients who received at least one dose of durvalumab are represented in this plot. For patients with at least one follow-up disease assessment, the unfilled vertical bars represents a patient’s best percent decrease in tumor burden when compared to baseline as defined by RECIST 1.1. Best percent decrease in tumor burden for patients who had new lesions appear at their first follow-up assessment are represented as cross-hatch filled bars to the left of patients with best response of progressive disease based a the change in measurements. Patients who expired prior to the first scheduled disease assessment and the death can reasonably be assumed to be due to disease progression are represented graphically to the left of patients with new lesions. Patients who did not have follow-up tumor disease assessment are presented at the very left of the plot marked with ‘IA’. Negative numbers represent decrease in tumor burden from baseline while positive numbers represent increase in tumor burden from baseline. Dashed reference lines at −30% and +20% indicate thresholds for partial response and disease progression.

With progression-events reported for 66/68 patients on the durvalumab arm, median IA-PFS was 2.9 months (95% CI 2.0–4.0 mos) overall and 2.3 months (95% CI 1.4–4.2 mos) among PD-L1 positive patients (Figures 2a and data not shown). Additionally, with events reported for 58/68 (85%) patients on durvalumab, median OS was 11.6 months (95% CI 10.2–14.3 mos) overall and 10.7 months (95% CI = 9.2–14.3) among PD-L1 positive patients (Figures 2c and data not shown).

Figure 2.

Progression-Free Survival and Overall Survival

As stated prior to an amendment that changed this trial to a single arm study, patients were randomized to durvalumab vs docetaxel (30 subjects). In this subset, median IA-PFS was 3.1 (95% CI 2.3–6.4) and 2.6 months (95% CI 1.4–4.6 mos) for the durvalumab and docetaxel arms, respectively with a hazard ratio of 0.66 (95% CI 0.40–1.11, Figure 2b). Also, in this subset, median OS was 12.1 (95% CI 9.2–17.7 mos) and 7.7 months (95% CI 6.6–10.7 mos) for durvalumab vs docetaxel with a hazard ration of 0.64 (95% CI 0.37–1.11, Figure 2d)

In the evaluation of association between demographic and tumor characteristics with overall survival, smoking status (ever versus never) was associated with better survival. (HR=0.28 [95% CI 0.1–0.8], p=0.02). PD-L1 expression (high versus low) was not associated with OS (p=0.81) (Figure 3b).

Figure 3. Forest Plot for Durvalumab-Treated Patients.

S1400A was designed to evaluate the activity of durvalumab in patients with squamous cell carcinoma of lung. All eligible patients had to have evidence of disease progression after first line treatment with a platinum doublet regimen. This trial was originally designed a s a randomized study, however, with the approval of other checkpoint inhibitors it was re-designed a single arm study. An overall response rate of 16% was observed in this study with a median OD of 11.6 months in unselected patients. PD-L1 data was available for 43 patients on durvalumab, with 14 patients (33%) PD-L1 positive (≥25%) and 2 responses (14% ORR, 95% CI 0%–33%), DCR was 57% (95% CI 31%–83%), median OS and PFS were 10.7 mos (95% CI 9.2–14.3 mos) and 2.3 mos (95% CI 1.4–4.2 mos), respectively. Toxicity was in line with what has been reported for other drugs in this class. Durvalumab, therefore, has single agent activity and toxicity comparable to other drugs in this class.

Discussion

The American Cancer Society estimated that about 222,500 new cases of lung cancer were diagnosed in the US in 2017.⁵ Although significant advances in our understanding of the molecular and genomic changes in this disease have led to the discovery of very effective treatments for certain populations (e.g. EGFR activating mutations, ALK translocations), these advances are mainly applicable to patients with non-squamous histology of NSCLC.

Durvalumab is a high affinity human IgG1 antibody that is capable of blocking PD-L1 binding to PD-1 and CD80.⁶ It does not block PD-L2 binding and because of this feature it is thought to cause less toxicity. In a phase I/II trial of patients with advanced NSCLC, durvalumab monotherapy was reported as having an objective response rate (ORR) of 23% in PD-L1 positive patients, and 5% in PD-L1 negative patient at a dose of 10 mg/kg every 2 weeks (q2w). Responses were durable and toxicity was manageable.⁷

In our study, the ORR was 16.2% in the durvalumab arm which is consistent with the recently reported single-arm study of durvalumab as third-line or later treatment for advanced NSCLC (ATLANTIC) where the response rate in the EGFR/ALK negative cohort was 7.5% for PDL1<25% and 16.4% for PDL1≥25%.⁸ This response rate is also similar to that reported in CheckMate-057 with nivolumab (19%), CheckMate-017 (20%) and Keynote-001 with pembrolizumab (18%).^1–3 In our study the intensity of PD-L1 staining did not correlate with clinical outcomes (response, IA-PFS, OS). This is in contrast with some of the other reported studies.^2,3 This could be due to the smaller number of PD-L1 positive patients on this trial (14). In the randomized portion of the study, with 30 patients in each arm, OS was 12.1 months in the durvalumab arm and 7.7 months in the docetaxel arm. The hazard ratios and therefore the relative benefit observed in this group and in the single arm part of the study are very much in line with prior studies.

Tumor Mutational Burden has been investigated in several retrospective studies as a potential biomarker of response to immunotherapy in lung cancer.^9,10 In the current trial, 58 (85%) of the patients on the durvalumab arm were evaluable for TMB and of them, 30 patients (44%) were classified as having high TMB, defined as ≥10 mutations/Mb (Table 2). Given the small number of patients in each subgroup in this trial, it is hard to truly assess the impact of TMB on clinical efficacy of durvalumab. Prospectively designed studies are needed to establish the role of TMB as a biomarker for predicting clinical efficacy of currently available checkpoint inhibitors.

Overall, a higher frequency of treatment-related adverse events (88%) were observed in this cohort of patients compared to the recently reported single agent experience with durvalumab as part of the ATLANTIC trial⁸ (55%) but a similar incidence of grade 3–4 immune-related adverse events. This rate is also slightly higher than that reported in ChekMate-057 (69% any grade), CheckMate-017 (58%) and Keynote-001 (63% any grade). This could be a reflection of the broader patient population enrolled on this study. The tolerability profile is acceptable and consistent with other anti-PD-1 and anti-PD-L1 therapies.

Durvalumab has proven to be highly effective in a number of studies including the PACIFIC trial.¹¹ As reported, consolidation treatment with durvalumab after concurrent chemotherapy and radiation in patients with locally advanced disease leads to a significant and clinically meaningful improvement in both PFS and OS thus establishing a new standard of care for this population.

Subsequent to the activation of this trial the treatment landscape in NSCLC changed with the introduction of immune checkpoint inhibitors, leading to the design changes described above. Initially validated in the second-line setting and now in the front-line setting, the results of phase III trials have led to FDA approval for checkpoint inhibitors in NSCLC. Current data suggest that these drugs have a similar profile both in terms of safety and efficacy in all squamous and non-squamous histologies.^1–3,12 Our results are consistent with these studies.

The Lung-MAP trial has demonstrated the feasibility and utility of an umbrella trial in advanced NSCLC. Enrollment has been rapid. Newer sub-studies for both genomically matched and unmatched patients are activated frequently. Rapid changes in the treatment landscape of NSCLC require frequent and thoughtful changes to this important trial to ensure that highest level of care is provided for patients. Data obtained from this trial will pave the way for the rapid development of new treatment options for this population.