Lung cancer is the leading cause of cancer-related deaths among men and women in the United States, despite being the second most commonly diagnosed cancer.

 Based on Surveillance, Epidemiology, and End Results (SEER) Program data from 2009 to 2013, cancers of the lung and bronchus account for 57.3 new cases of cancer per 100,000 individuals and 46.0 deaths per 100,000 individuals in the United States each year, and the risk of developing some form of lung or bronchus cancer over the course of a lifetime is about 6.5%. Further, in 2016, it was estimated that more than a quarter of all cancer-related deaths (26.5%; 158,000) will be attributed to lung and bronchus cancer.

 Overall, according SEER data, an estimated 224,390 new cases of lung and bronchus cancer will occur in 2016 in the United States, which represents 13.3% of all new cancer diagnoses.

Lung cancer is divided into 2 major classes: non—small cell lung cancer (NSCLC) and small cell lung cancer. NSCLC is the most common form of lung cancer (83%).1 The estimated 5-year survival rate is 21%.

Molecular targeted therapy is a strategy for cancer treatment that relies on exploiting the pathways and mutations that enable tumors to grow and progress. Prior to its introduction, treatment for locally or advanced-stage NSCLC was limited to chemotherapy or radiotherapy.

 Therefore, the introduction of agents that target mechanisms of tumor cell (TC) immune evasion has led to additional options for patients with NSCLC.

 NSCLC TCs exploit the programmed death-1 (PD-1) pathway to evade immune surveillance and destruction through upregulation of suppressive cell surface immune checkpoint regulators, making inhibitors of this pathway an attractive target in NSCLC cancer therapy.

 Many investigators are turning attention toward the identification of additional molecular biomarkers to predict targeted therapeutic efficacy in patients. Currently, programmed death ligand-1 (PD-L1) expression from NSCLC TCs and tumor-infiltrating lymphocytes has been most explored as a predictor of efficacy of anti—PD-1/PD-L1 immunotherapy treatment in clinical trials.

In the development of cancer therapies, NSCLC was not traditionally considered to be an immunotherapy-responsive tumor type; however, major responses are well documented in reaction to blockade of the PD-1/PD-L1 pathway with anti-PD-L1 monoclonal antibodies.

Immune checkpoint proteins are located on the surfaces of lymphocytes and other immune cells (ICs). These proteins bind specific membrane-bound ligand partners on the surface of neighboring immune effector cells. These ligand-receptor pairs potentially interact in conjunction with antigen presenting cells (APCs) through specific antigen peptide complexes with the T-cell receptor. Checkpoint pathways regulate coregulatory signals that determine if the T cell will be activated for proliferation, cytokine production, and functional activity or will be inactivated. Coregulatory signals are important to the immune response, as the intensity and duration of the response are controlled through this signaling complex. If the immune system is not activated at the appropriate time, normal tissue may be damaged or it may elicit a chronic autoimmune inflammatory response. Checkpoint pathways are attractive targets for therapeutic intervention, as they are important in regulating the physiological immune response to tumor antigens. An inhibitory immune checkpoint that has been targeted in tumor escape is PD-1 and its ligand, PD-L1.

The PD-1 receptor is expressed on the surface of cells involved in the anti-tumor immune response, including activated T cells (such as regulatory T cells), natural killer cells, and B cells. Its high-affinity ligand, PD-L1, is expressed on the surface of APCs, such as dendritic cells and macrophages, as well as TCs. When PD-L1 binds to the PD-1 receptor, the receptor complex activation induces intracellular inhibitory signaling cascades that lead to dysfunctional tumor-infiltrating T cells. These T cells have exhibited poor proliferation and a reduced capacity to produce cytokines.

 As PD-L1 is reported to be overexpressed on TCs in patients with NSCLC, the enhanced interaction with highly expressed PD-1 on attacking tumor-infiltrating T cells, allows cancer cells to inactivate T cells and thus avoid destruction.

Atezolizumab (Tecentriq, Roche/Genentech)

At the European Society for Medical Oncology (ESMO) 2016 Congress in Copenhagen, Denmark, investigators presented important updates on several clinical trials of anti-PD-L1/PD-1 immunotherapies. These included improved survival of patients with advanced NSCLC given atezolizumab in the OAK trial, a 2-year update on nivolumab in the CheckMate 057 and 017 trials, and promising evidence for pembrolizumab in the update to the KEYNOTE-10 trial. Some investigators suggest that the clinical benefit of anti-PD-1/PD-L1 monotherapy is not expected in the population of patients without elevated PD-L1 expression.6 Updated findings from these clinical trials indicate large variations in the expression of PD-L1 from diseased tissue. However, multiple trials support high PD-L1 expression as an effective tumor biomarker in predicting clinical response.

Atezolizumab is a human monoclonal immunoglobulin (Ig) G1 antibody specifically targeted to PD-L1 on tumor and tumor-infiltrating ICs.

 By binding to PD-L1, atezolizumab blocks the PD-L1/PD-1 and B7-1 receptor-activation complexes and downstream signaling responsible for suppression of T-cell proliferation, antitumorigenic T-cell activity, and cytokine production. In this way, atezolizumab restores the tumor-specific T-cell immune response.

 Additionally, atezolizumab is designed to eliminate fragment, crystallizable region (Fc)-receptor binding that facilitates antibody-dependent cell-mediated toxicity, thereby enhancing anticancer immunity through preventing the loss of T-effector cells that express PD-L1.

 Based on the results of numerous clinical trials (FIR, POPLAR, BIRCH, OAK), atezolizumab is indicated for use in patients with metastatic NSCLC who experience disease progression during or following treatment with platinum-containing chemotherapy.

The BIRCH trial was a single-arm, phase 2 study evaluating the efficacy and safety of atezolizumab as first-line or subsequent therapy. Atezolizumab was evaluated in 659 patients with recurrent, locally advanced, or metastatic (stage IIIB/IV) NSCLC with medium to high PD-L1 expression in diseased tissue.

An investigational immunohistochemistry (IHC) assay with the SP142 antibody, developed by Roche Diagnostics, allowed investigators to assess PD-L1 expression of both tumor and tumor-infiltrating inflammatory cells from patients with NSCLC.10 The reproducible method allows investigators to score TCs as a percentage of PD-L1—expressing TCs and tumor-infiltrating ICs expressing PD-L1 as a percentage of tumor area. In this way, the subgroup of patients with NSCLC TCs expressing the highest PD-L1 level (at least 50%) are scored as TC3, and patients with PD-L1 expression from 5% to 50% are scored as TC2. Patients scored as TC1 have less than 5% PD-L1 expression, and tumors scored as TC0 have less than 1%. Tumor-infiltrating ICs with a score of IC3 have PD-L1 expression and represent at least 10% of tumor area; IC2, at least 5%; IC1, at least 1% but less than 5%; and IC0, less than 1%.

As the goal of this study was to provide atezolizumab to patients with NSCLC who would have the greatest response and treatment benefits, those with PD-L1 expression scoring of TC2/3 or IC2/3 were treated with a 1200-mg intravenous (IV) dose of atezolizumab monotherapy every 3 weeks. Patients receiving atezolizumab as a first-line treatment continued therapy until they experienced progression of the disease or toxicity, whereas the drug was continued until loss of clinical benefit in patients with prior platinum-based chemotherapy. The primary endpoint of the BIRCH study was the objective response rate (ORR), assessed by an independent review facility per RECIST v1.1 criteria. Secondary endpoints included duration of response (DOR), progression-free survival (PFS), overall survival (OS), and safety.

Atezolizumab was found to be an effective monotherapy in patients with advanced NSCLC and PD-L1 overexpression. At the data cutoff on May 28, 2015, patients with TC3/IC3 PD-L1 expression had greater reduction of tumor burden compared with those with lower PD-L1 expression levels when receiving atezolizumab as first-line or subsequent therapy. The ORR in patients treated with atezolizumab was greater with increased expression of PD-L1, with similar rates in those with or without previous chemotherapy. Of patients who received first-line atezolizumab treatment, those with the highest PD-L1 expression had a stronger overall response, as 26% of TC3-/IC3-scored patients had reduction of tumor burden compared with 19% in the lower PD-L1—expressing subgroup. This pattern was also observed in the patients treated with atezolizumab as a second-line treatment: the ORR was 24% in patients with higher PD-L1 TC or IC expression levels and 17% in those with lower PD-L1 expression. In patients who had undergone 2 or more prior chemotherapy regimens, ORRs increased when stratified by PD-L1 expression: 27% in the TC3 or IC3 subgroup compared with 17% in those with lower PD-L1 expression scored as TC2/3 or IC2/3.

After 6 months of atezolizumab treatment, greater rates of PFS were also achieved by patients with higher PD-L1 expression in tumors scored as TC3 or IC3 compared with TC2/3 and IC2/3. PFS was achieved by 48% of patients receiving first-line atezolizumab with higher PD-L1 expression levels compared with 46% of this cohort with lower PD-L1 expression. In patients receiving atezolizumab as a second-line therapy, PFS was achieved by 34% of patients with TC3 or IC3 PDL1 expression versus 29% in patients with lower PD-L1 expression. Among those receiving atezolizumab as a third- or higher-line therapy, PFS was achieved by 39% of patients with PD-L1 expression of TC3 or IC3 versus 31% of those with TC2/3 and IC2/3 scoring.

 In patients with TC2/3 or IC2/3 expression levels, 6-month OS was achieved by 82% of treatment-naïve patients, 76% with 1 prior line of chemotherapy, and 71% receiving atezolizumab as a third- or higher line of therapy.

Results of the BIRCH trial showed that atezolizumab had no unexpected toxicities in this population. Adverse events (AEs) of severity grade 3 and 4 were reported in 38% of patients after treatment with atezolizumab. The most commonly reported AEs were fatigue (18%) and nausea (10%). One patient reported treatment-related pneumonia of grade 5 severity.

The findings of the BIRCH trial indicate that PD-L1 may serve as a biomarker to identify patients most likely to benefit from atezolizumab monotherapy and that this biomarker can be used to determine which patients may benefit most from a combination of atezolizumab and another medicine. Additionally, the benefit of atezolizumab was pronounced, with a durable response in tumors with pre-existing immunity. The POPLAR trial further assessed the value of PD-L1 expression levels on tumor and tumor-infiltrating ICs in predicting atezolizumab efficacy.

The phase 2, open-label POPLAR trial evaluated the efficacy and safety of atezolizumab as a second- or thirdline treatment in patients with NSCLC. The trial included 287 patients in 13 countries with advanced or metastatic NSCLC who had disease progression after platinum-based chemotherapy treatment. A key strength of POPLAR was that it enrolled patients irrespective of PD-L1 status, which was assessed on both TCs and tumor-infiltrating ICs. Patients were stratified by PD-L1 expression on ICs and TCs and scored using a highly reproducible Roche Diagnostic IHC assay.

Patients with NSCLC were randomized 1:1 to receive either 1200 mg of atezolizumab (n = 144) or 75 mg/ m² of docetaxel (n = 143) intravenously every 3 weeks. Atezolizumab was continued in patients until loss of clinical benefit, while patients received docetaxel until disease progression or unacceptable toxicity.8 The primary endpoint in the POPLAR trial was OS in both arms, and the results were stratified by varying PD-L1 expression. Investigators assessed tumors through imaging at baseline, every 6 weeks for 36 weeks after randomization, and then every 8 to 10 weeks. Secondary endpoints of ORR, PFS, and DOR were assessed using RECIST v1.1 criteria.

After a minimum follow-up of 13 months, investigators found that atezolizumab was associated with significant (P = .04) improvements in OS, with a rate of 12.6 months compared with 9.7 months in patients treated with docetaxel (HR, 0.73; 95% CI, 0.53-0.99).

 Much like the results of the BIRCH trial, POPLAR showed that OS with atezolizumab increased among patients with higher TC and IC PD-L1 expression. In 195 patients (68%) with at least a 1% level of PD-L1 expression, survival was significantly (

 = .005) improved, with a median OS of 15.5 months in the atezolizumab group and 9.2 months in docetaxel group. Within the study population, 37% of patients (n = 105) were found to have PD-L1 expression in at least 5% of cells, scored as TC2/3 or IC2/3. Improvements in OS were significant (

 = .014) within this subgroup of patients, with a median OS of 15.1 months with atezolizumab compared with 7.4 months with docetaxel.

 Approximately 16% of patients were scored as TC3 or IC3. This subgroup had significant improvement (

 = 0.068), with a median OS of 15.5 months for those receiving atezolizumab and 11.1 months for those receiving docetaxel. Importantly, investigators found that PD-L1 expression was not associated with docetaxel efficacy, as patients without PD-L1 expression (scored as TC0 or IC0) who received atezolizumab or docetaxel had the same OS of 9.7 months.

In the entire treatment population, the ORR was 15% in both the atezolizumab (n = 21) and docetaxel groups (n = 21). In the 21 patients achieving an objective response with atezolizumab, the median DOR was 14.3 months and 12 of these responders had an ongoing response at the cutoff date. This response was greater than in the 21 patients who achieved an objective response with docetaxel treatment, as the median DOR was 7.2 months and only 5 of the responders had an ongoing response at the time of cutoff. Overall median PFS was similar between treatment groups, with 2.7 months for atezolizumab and 3.0 months for docetaxel.

 As with OS, atezolizumab was associated with improvements in PFS and ORR, with increasing PD-L1 expression. The greatest benefits were observed in the subgroup of patients with the highest PD-L1 expression of TC3 or IC3 (n = 47). Patients with the highest PDL1 expression had a median PFS of 7.8 months with atezolizumab compared with 3.9 months with docetaxel. In this subgroup, 38% of patients given atezolizumab had a confirmed partial or complete response per RECIST v1.1 criteria compared with 13% of those given docetaxel.

The POPLAR trial demonstrated that atezolizumab was well tolerated and associated with fewer AEs and related discontinuations than docetaxel. Atezolizumab was discontinued in 1% of patients due to treatment-related AEs, whereas 18% of patients were withdrawn from docetaxel. Treatment-related AEs of grade 3 or 4 severity were reported in 11% of patients receiving atezolizumab compared with 39% receiving docetaxel. AEs reported with atezolizumab, and at a incidence of 5% or more than docetaxel, included decreased appetite, shortness of breath (dyspnea), nausea, fever (pyrexia), arthralgia, musculoskeletal pain, diarrhea, insomnia, pneumonia, asthenia, myalgia, and hypothyroidism.

Investigators discovered distinct patient populations with PD-L1 expression on TCs only or on tumorinfiltrating ICs only across all PD-L1 expression scores. For example, in patients with scoring of TC1/2/3 or IC1/2/3, 30% expressed PD-L1 only on ICs and not TCs. In addition, less than 1% of patients with NSCLC had the highest expression levels of PD-L1 co-expressed on both TCs (TC3) and tumor-infiltrating ICs (IC3). This suggests that TC3 and IC3 are distinct subpopulations in NSCLC and that both cell types can be used as a predictor of atezolizumab response.

Clinical Trial Data Presented at ESMO 2016

The POPLAR trial demonstrated that atezolizumab provided a survival benefit in previously treated patients with NSCLC and that differences in PD-L1 expression in diseased tissue can predict response to atezolizumab as a biomarker. The subsequent larger phase 3 OAK trial was conducted to better assess the efficacy of atezolizumab within different subgroups.

The phase 3 OAK study enrolled 1225 patients with locally advanced or metastatic NSCLC who experienced disease progression during or following 1 or 2 lines of previous platinum-based chemotherapy. Twenty-five percent of patients had received 2 prior chemotherapy regimens. Patients were randomized 1:1 for IV treatment every 3 weeks with 1200-mg atezolizumab or 75-mg/m2 docetaxel.

Using the same method as the BIRCH and POPLAR trials, patients were subgrouped according to PD-L1 expression scoring assessed by a reproducible IHC assay using the Roche SP142 antibody. Enrolled patients were stratified by tumor PD-L1 expression status, whether they had 1 or 2 previous chemotherapy regimens, and by histological differences in either squamous (SQ) or nonsquamous (non-SQ) disease. Co-primary endpoints included OS in all patients and in patients subgrouped by expression level of PD-L1 on TCs and tumor-infiltrating ICs. Secondary endpoints were PFS, ORR, DOR, and safety.11 The analysis of primary efficacy in the first 850 patients enrolled was presented at the ESMO 2016 Congress. Atezolizumab improved OS in patients with SQ and non-SQ NSCLC as a second- or third-line therapy, regardless of PD-L1 expression.

In the overall population, investigators reported a significant (

 = .0003) improvement in OS in patients given atezolizumab (n = 425) compared with patients given docetaxel (n = 425). Median OS was 13.8 months with atezolizumab and 9.6 months with docetaxel (HR, 0.73). Atezolizumab had similar effects on OS in patients with SQ and non-SQ histology: in patients with SQ NSCLC (n = 112) treated with atezolizumab, median OS was 8.9 months compared with 7.7 months in those treated with docetaxel (n = 110) (

 = .0383). In patients with non-SQ NSCLC (n = 313), median OS was 15.6 months, a significant improvement (

 = .0015), compared with 11.2 months in those given docetaxel (n = 315).

When patients were stratified by PD-L1 expression on at least 1% of TCs or ICs, atezolizumab significantly improved OS in patients scored TC1/2/3 or IC1/2/3 compared with docetaxel (HR, 0.74; P = .0102). This benefit was shown to be significant (

 = .0102), with patients scored for PD-L1 expression on at least 5% of TC or ICs (TC2 or IC2), with 15.7 months in the atezolizumab group compared with 10.3 months in the docetaxel group. A pronounced benefit in OS was found to be correlated with the highest PD-L1 expression of at least 50% of TCs or at least 10% of ICs (TC3 or IC3). Median OS was significantly (

 <.0001) greater among atezolizumab-treated patients (n = 72) at 20.5 months compared with 8.9 months in docetaxel-treated patients (n = 65). Investigators also found that patients treated with atezolizumab (n = 180) with less than 1% PDL1 expression (TC0 and IC0) had significant (

 = .0215) improvements in median OS (12.6 months) compared with those treated with docetaxel (n = 199; 8.9 months).

Atezolizumab improved PFS in treated patients, with a duration of 2.8 months compared with 4 months in patients given docetaxel (HR, 0.95). The reduction in tumor burden in the treatment population was observed at similar rates of partial or complete objective responses in 13.6% of patients given atezolizumab and 13.4% of patients given docetaxel. However, atezolizumab was found to substantially improve the DOR: time to disease progression was 16.3 months with atezolizumab compared with 6.2 months in those given docetaxel.11 Atezolizumab was well tolerated. It was associated with improved outcomes, as treatment-related AEs of grade 3 or 4 severity occurred in 15% of patients in the atezolizumab group and 43% in the docetaxel group. Furthermore, there were no reports of patient deaths related to atezolizumab and 1 death associated with docetaxel treatment.

From previous phase 2 studies in patients with NSCLC, response rates indicated that the efficacy of atezolizumab treatment varies depending on PD-L1 protein expression on TCs and tumor-infiltrating ICs. The POPLAR, FIR, and BIRCH studies demonstrated the tolerability and efficacy of atezolizumab monotherapy and showed that increased OS is correlated with TC or IC PD-L1 expression.

 In the OAK trial, atezolizumab significantly improved patient survival in those with less than 1% of cells expressing PD-L1.

 There is a need for additional biomarkers that can identify patients likely to benefit from atezolizumab therapy.

Data presented at the ESMO 2016 Congress indicated that tumor mutational burden (TMB) can be a predictor of improved response to atezolizumab independent of PD-L1 expression in patients with NSCLC. Mutation load (the number of genetic mutations found within a tumor) may be a predictive biomarker of response to immune checkpoint inhibition. TMB in NSCLC can help determine response to treatment, as mutations can increase tumor immunogenicity through neoantigen formation and increased response to immunotherapy.

Archival tumor samples were evaluated for TMB in 465 patients with NSCLC who were enrolled in the POPLAR (n = 92), BIRCH (n = 340), or FIR (n = 33) trials for a second-line or higher treatment with atezolizumab monotherapy (as they received prior platinum-based chemotherapy and experienced disease progression). The population of patients treated with docetaxel from the POPLAR study was also included.12 Using the panel of targeted short variants in the FoundationOne genomic profiling assay, TMB was quantified from all 3 trials and defined by percentiles of the study-specific mutational burden as high (75% quantile), median (50% quantile), and low (25% quantile) to assess efficacy in different patient groups.

ORR was confirmed from the samples per RECIST v1.1 criteria. Investigators found that 20% of patients from the POPLAR study with a high mutation load achieved a response to atezolizumab compared with 8% of patients given docetaxel. In patients who had a median TMB, 20% had a response to atezolizumab compared with 4% receiving docetaxel. Low mutation load was associated with a response rate of 14% in patients given atezolizumab compared with 10% given docetaxel. Patients in the docetaxel group did not experience an improvement in ORR with increasing TMB, and those with low tumor burden had similar rates of response, suggesting that TMB is not a prognostic marker.

Using tumor-infiltrating IC and TC samples from all trials, TMB and PD-L1 expression levels were associated with increasing ORRs to atezolizumab. In patients with the highest level of TMB and greater than 10% PD-L1— expressing cells (scored as TC3 or IC3), 40% achieved an objective response compared with those with the highest TMB and TC2 or IC2 (at least 5% PD-L1 expression), who had an ORR of approximately 17.5%. However, even in patients grouped by the lowest TMB, those with TC3 or IC3 scores demonstrated an ORR greater than 30%, double the ORR of those with TC2 or IC2 scores (approximately 15%). In the patients identified with median TMB, those with TC3 or IC3 PD-L1 expression had an ORR of approximately 37.5%, and 15% of those with TC2 or IC2 scores achieved a response. Therefore, the results indicate that ORRs are tightly associated with PD-L1 expression across TMB levels, with increasing responses in the highest TMB and highest PD-L1 expression groups. High TMB may be used as a candidate biomarker in addition to tumor PD-L1 to identify patients for treatment with atezolizumab.

Nivolumab is a humanized IgG4k monoclonal antibody that binds to the PD-1 receptor expressed on T cells. It inhibits receptor complex activation with its ligands, PDL1 and programmed death ligand-2 (PD-L2), upregulated in TCs. Nivolumab-mediated blockade of PD-1/PD-L1 signaling facilitates anti-tumor immune responses, including T-cell immune surveillance for tumor recognition.13 Based on the results of the CheckMate clinical trials, nivolumab is indicated for use in patients with metastatic NSCLC who experience progression during or following platinum-based chemotherapy.

The phase 3 CheckMate 017 trial compared the safety and efficacy of nivolumab with docetaxel in patients with advanced SQ NSCLC who experienced progressive disease during or after treatment with platinumbased chemotherapy. Patients enrolled had an Eastern Cooperative Oncology Group (ECOG) performance status score of 0 or 1, and most were current or former smokers. Patients with stage IIIB/IV SQ NSCLC were randomized 1:1 to receive nivolumab (n = 135) at 3 mg/kg intravenously every 2 weeks or IV docetaxel (n = 137) at 75 mg/m

 every 3 weeks. Treatment was continued in all patients until progressive disease or unacceptable toxicity.

PD-L1 protein expression was evaluated retrospectively from archival tumor biopsy specimens, and all randomized patients were grouped by baseline PD-L1 expression level. Samples were categorized as positive when IHC staining at the tumor-cell membrane at any intensity was observed, and PD-L1 expression was defined by the percentage of diseased TCs with PDL1 staining of any intensity at prespecified expression levels for the PD-L1 biomarker analysis, which were 1%, 5%, and 10% of cells in a section with at least 100 TCs that could be evaluated.

The primary outcome was OS (time in months from randomization until death from any cause) in patients with SQ NSCLC treated with nivolumab and docetaxel. Results from this trial demonstrated that nivolumab significantly improved patient survival over treatment with docetaxel. The median OS was 9.2 months in patients given nivolumab compared with 6.0 months for patients receiving docetaxel. The risk of death was significantly lower with nivolumab than docetaxel (HR, 0.59; 95% CI, 0.44-.079; 

 <.001). Nivolumab also was found to nearly double the rate of 1-year survival, as 42% of patients were likely to be alive following randomization after 12 months compared with 24% of patients treated with docetaxel.

 Among the secondary endpoints were investigatorassessed ORR and PFS, the correlation between PD-L1 expression and these parameters, and safety. Nivolumab significantly (

 = .0004) extended median PFS, with patients accruing 3.5 months until disease progression, whereas docetaxel was associated with a median PFS of 2.8 months (HR, 0.62; 95% CI; 0.47-0.81; 

 <.001). Within this study population, the 1-year PFS rate was 21% for patients treated with nivolumab; this was a significant improvement compared with docetaxel, which was associated with only a 6% probability that participants would experience no disease progression or death following randomization. In all participants, nivolumab treatment significantly (

 = .008) improved patient response compared with docetaxel: 1% of patients treated with nivolumab had a complete response, where targeted lesions were eradicated, and 19% had a partial overall response, where there was at least a 30% decrease in the sum of target tumor diameters versus 9% of patients given docetaxel.

PD-L1 expression was quantifiable in 83% of the study patients with SQ NSCLC. Although the rates of OS and PFS in the PD-L1 subgroup favored nivolumab, they were similar to the overall population and PD-L1 expression was not prognostic or predictive of efficacy in this trial. A limitation of this trial was that PD-L1 expression was assessed in archival tumor tissue, not in tumors at the time of nivolumab treatment. Unlike the BIRCH, POPLAR, and OAK trials, tumor-infiltrating ICs were not assayed. Investigators determined that the benefit of nivolumab may be independent of PD-L1 expression.

Nivolumab was found to have a better safety profile in this study population than docetaxel. Patients treated with nivolumab had no reported grade 5 severity events compared with 2% treated with docetaxel. Treatmentrelated AEs of grade 3 to 4 severity occurred in 7% of patients given nivolumab compared with 55% given docetaxel. Fatigue and decreased appetite (both 1%) were the most frequent treatment-related AEs of grade 3 or 4 severity reported in patients treated with nivolumab. Patients given docetaxel most commonly reported grade 3 or 4 AEs of neutropenia (30%), febrile neutropenia (10%), and fatigue (8%). Although there were no treatmentrelated deaths in patients treated with nivolumab, 2% of patients treated with docetaxel had fatal AEs.

Results from the CheckMate 017 study showed that nivolumab was the first PD-1 inhibitor with survival benefit compared with standard-of-care docetaxel as a continual line of therapy in patients with advanced SQ NSCLC who previously failed chemotherapy. However, tumor PD-L1 expression was not found to be a biomarker for selecting patients for nivolumab treatment in this study.

The phase 3 CheckMate 057 trial investigated the efficacy and safety of nivolumab in patients with non-SQ NSCLC who had progression of disease after treatment with platinum-based chemotherapy and a tyrosine kinase inhibitor. Patients were randomized to receive 3 mg/kg nivolumab intravenously every 2 weeks (n = 292) or docetaxel 75 mg/m

 every 3 weeks (n = 290) until progression or discontinuation due to toxicities. The primary endpoint was OS. Secondary endpoints of the trial, as evaluated by investigators per RECIST v1.1 criteria, were ORR, PFS, efficacy by PD-L1 expression, and safety.

Nivolumab significantly improved patient survival in all randomized patients. The median OS in patients treated with docetaxel was 9.4 months compared with 12.2 months in patients treated with nivolumab, a significant increase (HR, 0.73; 95% CI, 0.59-0.89; 

 = .00155). Additionally, the 1-year survival rate was 50.5% in patients treated with nivolumab (95% CI, 44.6%- 56.1%) compared with 39.0% in patients treated with docetaxel (95% CI, 33.3%-44.6%). In all randomized patients, significant improvements (

 = .0235) in ORR were shown in patients who received nivolumab (19.2%) compared with patients given docetaxel (12.4%). One-year PFS was observed in 8.1% of patients receiving docetaxel (95% CI, 5.1%-12.0%) versus 18.5% of patients receiving nivolumab (95% CI, 14.1%-23.4%).

An endpoint of the CheckMate 057 trial was the correlation of nivolumab efficacy with tumor PD-L1 expression: 78% of the randomized study population had quantifiable PD-L1 expression. Using the same methods as the CheckMate 017 trial, PD-L1 biomarker evaluation by IHC was assessed in archival or recent samples, and expression was defined as the percentage of TCs with PDL1 staining of any intensity to predefined levels. When patients were grouped by baseline PD-L1 expression for OS and ORR, investigators found that PD-L1 expression was predictive of nivolumab treatment benefit.

Approximately 29 months following randomization, the OS of patients with greater than 1% TC PD-L1 expression was 9 months with docetaxel and 17.2 months with nivolumab (HR, 0.59; 95% CI, 0.43-0.82). This benefit in OS was not observed for patients with less than 1% PD-L1 expression or those with no quantifiable PD-L1 expression, as they had similar rates of OS. ORR was defined as the percentage of all randomized participants whose best overall response was a confirmed complete response or partial response.

In patients treated with nivolumab, 30.9% with greater than 1% PD-L1 expression had a partial or complete response to treatment compared with 12.2% of patients treated with docetaxel. This benefit of nivolumab was not seen in patients with less than 1% PD-L1 expression or those with no quantifiable PD-L1 expression compared with those subgroups in patient receiving docetaxel. As seen in the CheckMate 017 trial, it is important to point out that these analyses are mostly from archival tumor samples. Because PD-L1 expression can change after exposure to chemotherapy, it is difficult to draw firm conclusions on the predictive value of this biomarker in this pretreated population.

2-Year Follow-up of CheckMate 017 and Checkmate 057

The CheckMate 057 trial demonstrated a similar safety profile in patients given nivolumab as in previous trials. Treatment-related AEs of grade 3 to 5 severity were reported in 10.5% of patients given nivolumab and 53.7% given docetaxel. Although no fatalities occurred in patients treated with nivolumab, 1 death was reported with docetaxel treatment.

Based on the results from CheckMate 057 and CheckMate 017, the FDA approved nivolumab for the treatment of patients with non-SQ and SQ NSCLC who failed treatment with platinum-based chemotherapy.

 In these trials, nivolumab was superior to docetaxel in both safety and efficacy. Updated long-term outcomes based on a 2-year follow-up of patients enrolled in CheckMate 017 and Checkmate 057 were presented at ESMO 2016.

In a 2-year follow-up of patients from the CheckMate 017 trial with SQ NSCLC, 8% of patients remained on nivolumab therapy; no patients continued a docetaxel regimen. Updated survival results indicated that nivolumab treatment had continued benefit: patients who received nivolumab were reported to have a 2-year OS rate of 23% compared with 8% in the docetaxel group.

The median OS of patients treated with nivolumab was 9.2 months (95% CI, 7.3-12.6 months) compared with 6 months (95% CI, 5.1-7.3 months) for patients on docetaxel treatment (HR, 0.62; 95% CI, 0.47-0.80). In the 2-year follow-up of the CheckMate-017 trial, PD-L1 status was still not found to be linked to survival.

The rate of 2-year PFS indicated that 16% of patients treated with nivolumab had no disease progression; however, as no patients remained on docetaxel treatment, the 2-year PFS could not be calculated for that patient cohort or compared between groups. The median PFS was 3.5 months in the nivolumab group and 2.8 months in the docetaxel group.

After 2 years of treatment, 20% of patients continued to have a partial or complete response to nivolumab treatment compared with 9% of patients given docetaxel. Additionally, an extended DOR of 25.2 months (time between the first documented tumor reduction in response to nivolumab until disease progression or death) was observed in this follow-up study; in contrast, docetaxel treatment had a DOR of 8.4 months. Although no patients had an ongoing response to docetaxel, 37% of responders (10/27 patients) to nivolumab had not progressed or initiated subsequent therapy at the time of analysis. Progressive disease was observed in 41% of patients given nivolumab and 35% given docetaxel (HR, 0.96; 95% CI, 0.65-1.43).

Longer mean therapeutic durations were found in the nivolumab patient cohort: 7.5 months compared with 2.5 months with docetaxel. Treatment-related AEs of grade 3 and 4 severity at 2 years’ minimum follow-up were 8% with nivolumab versus 56% with docetaxel. The most common treatment-related AEs were endocrine- (5% with nivolumab vs 0% with docetaxel), hepatic- (5% vs 2%, respectively), pulmonary- (7% vs 1%), and skin-related (10% vs 9%).

In a 2-year follow-up of the CheckMate 057 trial in patients with non-SQ NSCLC, 9% of patients remained on nivolumab and no patients continued treatment with docetaxel. The 2-year rate of OS was 29% with nivolumab compared with 16% in patients treated with docetaxel (HR, 0.75; 95% CI, 0.63-0.91). The median OS was extended at 12.2 months (95% CI, 9.7-15.1 months) for patients treated with nivolumab compared with 9.5 months (95% CI, 8.1-10.7 months) in patients treated with docetaxel.

The 2-year PFS results indicated that 12% of patients in the nivolumab group had no disease progression compared with 1% of patients in the docetaxel group. However, the median PFS was 2.3 months in patients given nivolumab (HR, 0.89; 95% CI, 0.75-1.07) compared with 4.3 months in patients given docetaxel. After 2 years of treatment, 19% of study patients continued to have a partial or complete response to nivolumab compared with 12% given docetaxel. In the patient cohort treated with nivolumab, progressive disease was observed in 44% and stable disease in 25% compared with 29% and 42%, respectively, in patients receiving docetaxel. Although no patients had an ongoing response to docetaxel, responders to nivolumab had ongoing responses without progression or initiation of subsequent therapy at the time of this 2-year analysis. Of the 56 total responders to nivolumab treatment, 19 patients had an ongoing response in this follow-up study.

Patients with PD-L1—positive staining tumors had stronger OS outcomes in the 2-year follow-up. Among patients with less than 1% of tumors expressing PD-L1, the OS rate was 25% in patients given nivolumab and 18% in patients given docetaxel). Among patients with PD-L1 expression on at least 1% of TCs, the 2-year OS rate was 37% in patients treated with nivolumab compared with 17% in those treated with docetaxel (HR, 0.62; 95% CI, 0.47-0.83). The 2-year OS in patients with at least 5% of tumors expressing PD-L1 was 44% with nivolumab compared with 14% with docetaxel (HR, 0.48; 95% CI, 0.34-0.68). In patients with PD-L1 scoring of at least 10%, OS was 45% in patients treated with nivolumab versus 13% treated with docetaxel (HR, 0.43; 95% CI, 0.30- 0.62).

 AEs of grade 3 to 4 severity were less frequent with nivolumab compared with docetaxel (11% vs 54%).

The phase 1 CheckMate 012 study measured the efficacy and safety of nivolumab as a monotherapy and in combination with ipilimumab as a first-line treatment in patients with advanced NSCLC. The CheckMate 026 trial was an open-label, randomized, phase 3 study that evaluated the efficacy of nivolumab as a first-line therapy in patients with stage IV/recurrent PD-L1—positive NSCLC compared with investigator’s choice of platinumbased doublet chemotherapy based on histology characterization. Patients with SQ NSCLC received gemcitabine with cisplatin or carboplatin and paclitaxel with carboplatin, and patients with non-SQ NSCLC were treated with pemetrexed with cisplatin or carboplatin.

 Patients enrolled in this trial were treatment-naïve for stage IV or recurrent NSCLC, had an ECOG performance status score of 0 to 1, and were positive for PD-L1 expression. Like prior CheckMate clinical studies, PD-L1 expression in patients was assessed using an IHC assay.

A total of 541 patients were randomized 1:1 to receive 3-mg/kg nivolumab treatment intravenously every 2 weeks or up to 6 cycles of platinum-based doublet chemotherapy every 3 weeks until disease progression, unacceptable toxicity, or completion of 6 cycles of treatment.

Nivolumab used as a first-line therapy in patients with NSCLC was not superior to treatment with platinumbased chemotherapy in the 423 patients selected as they had PD-L1 expressed on at least 5% of tumors.

The primary endpoint was to compare PFS per RECIST v1.1 criteria between treatment groups. The results showed no significant improvements. Patients treated with nivolumab had a reported 4.2 months before first documented tumor progression or death compared with 5.9 months reported in patients treated with chemotherapy (HR, 1.15; 95% CI, 0.91-1.45).

 The secondary outcome of OS was also similar among the 2 treatment groups, as patients given nivolumab had an OS of 14.4 months compared with 13.2 months for patients given chemotherapy (HR, 1.02; 95% CI, 0.80-1.30).

 The safety prolife from the CheckMate 026 trial was consistent with observations from previous studies.

Among all patients given nivolumab, AEs of any severity occurred in 71% of patients, and grade 3 or 4 treatmentrelated AEs were reported in 18% of patients. Nivolumab treatment was favorable to chemotherapy in this study population, as 92% of patients given chemotherapy reported AEs of any severity and 51% had grade 3 or 4 severity events.

CheckMate 026 did not meet its primary endpoint, and nivolumab was not superior to chemotherapy in improving PFS and OS for patients with greater than 5% of tumors expressing PD-L1. As with other CheckMate trials, this study used fresh or archival tissue samples obtained 6 months prior to enrollment for PD-L1 biomarker evaluation by IHC. More effort is needed to explore the potential predictive tumor and peripheral biomarkers of clinical response to nivolumab and to evaluate the correlation between PD-L1 tumor expression level and PFS in patients with PD-L1—positive stage IV or recurrent NSCLC given nivolumab monotherapy.

Due to the success of the KEYNOTE trials, pembrolizumab is approved as a first-line treatment for patients with metastatic NSCLC whose tumors have high PD-L1 expression (as assessed by the IHC 22C3 pharmDx immunohistochemistry assay and found to have greater than 50% of TCs expressing PD-L1). Pembrolizumab is also approved for metastatic NSCLC in patients who have experienced disease progression on or after platinumcontaining chemotherapy, with at least 1% of tumors expressing PD-L1. Pembrolizumab is a monoclonal antibody that binds to the PD-1 receptor found on T cells and blocks its interaction with PD-L1 and PD-L2. Through inhibition of the PD-L1/PD-1 receptor activation complex, pembrolizumab facilitates T-cell proliferation and cytokine production, and releases the anti-tumor immune response.

At ESMO 2016, investigators presented data on the prevalence of PD-L1 expression in patients with NSCLC enrolled in the KEYNOTE trials. In 3 KEYNOTE trials, PDL1 expression was used to identify patients who would benefit from pembrolizumab treatment. Archival or newly obtained tumor samples were obtained as entry criterion for PD-L1 expression testing, assessed for percentage of TCs with any positive membranous staining, and given tumor proportion scores (TPSs).

A total of 4784 patients with advanced NSCLC whose tumors were evaluable for PD-L1 expression from the KEYNOTE-001 (n = 909), KEYNOTE-010 (n = 2222), and KEYNOTE-024 (n = 1653) trials were evaluated. Of these patients, 33% were found to have the lowest amount of TCs expressing PD-L1, for a TPS less than 1%, and 38% had a TPS of 1% to 49%. The highest expression of PD-L1 was found in 28% of patients who had a TPS of at least 50%. The prevalence of PD-L1 expression was similar in patients treated with pembrolizumab as a first-line therapy and those who had disease progression with previous lines of platinum-based chemotherapy: 30% had the highest level of PD-L1 expression, 40% had PDL1 expression from 1% to 49% of diseased tumors, and 31% had less than 1% of tumors expressing PD-L1. In patients who had previous lines of chemotherapy and experienced disease progression, 27% had the highest level of PD-L1 expression, 38% had PD-L1 expression from 1% to 49% of diseased tumors, and 35% had less than 1% of tumors expressing PD-L1.

 The findings showed that 67% of patients with advanced NSCLC enrolled for pembrolizumab treatment in the KEYNOTE trials had detectable PD-L1 expression from at least 1% of TCs and 28% had high levels from at least 50% of tumors.

The phase 3 KEYNOTE-024 trial assessed the efficacy of pembrolizumab monotherapy compared with standardof- care platinum-based chemotherapies in patients previously untreated for advanced NSCLC. Enrolled patients had an ECOG performance score of 0 or 1, life expectancy of at least 3 months, no prior treatment regimen, and high PD-L1 expression.

Of the 1653 patients screened for PD-L1 expression, 500 (32%) had a TPS of 50% or more tumors expressing PD-L1, as assessed by a specific 22C3 pharmDx IHC assay. A total of 305 patients from 16 countries met all inclusion criteria for enrollment and were randomized 1:1 to receive an IV dose of 200-mg pembrolizumab every 3 weeks for up to 35 cycles (n = 154) or investigator’s choice of platinum-based chemotherapy (n = 151) for 4 to 6 cycles of either carboplatin plus pemetrexed, gemcitabine, or paclitaxel, or cisplatin plus pemetrexed or gemcitabine. Treatment continued in both arms until disease progression, treatment-related AEs of unacceptable severity, or a patient withdrew consent.24 The primary endpoint of this trial was PFS; the median follow up was 11.2 months. Investigators observed that pembrolizumab significantly (

 <.001) improved PFS by approximately 4 months compared with chemotherapy, with a 10.3-month median PFS in patients treated with pembrolizumab compared with 6.0 months in patients treated with chemotherapy (HR, 0.50; 95% CI, 0.37-0.68).

At 6 months following randomization, an estimated 62.1% of patients treated with pembrolizumab were alive and did not have progression of disease compared with 50.3% of patients treated with chemotherapy.

 Secondary endpoints were OS, ORR, and safety.

Pembrolizumab was associated with significant (

 = .005) reduction in the risk of death compared with chemotherapy (HR, 0.60; 95% CI, 0.41-0.89). An estimated 80.2% of patients treated with pembrolizumab were alive at 6 months compared with 72.4% of patients treated with chemotherapy. A higher ORR was observed in patients given pembrolizumab compared with chemotherapy: 44.8% of those in the pembrolizumab group had a confirmed complete or partial response based on RECIST v1.1 criteria compared with 27.8% of patients in the chemotherapy group. Median time to response was 2.2 months in both treatment groups, with 69 and 42 patients in the pembrolizumab and chemotherapy groups, respectively, having an objective response. At the time of data cutoff, responses were ongoing in the pembrolizumab group and the median DOR was 6.3 months in the chemotherapy group.

Treatment-related AEs occurred in 73.4% of patients given pembrolizumab and 90.0% of patients given chemotherapy. The most common treatment-related AEs in patients given pembrolizumab were diarrhea (14.3%), fatigue (10.3%), and pyrexia (10.4%). Certain AEs also occurred at a higher incidence with chemotherapy, and included anemia (44.0%), nausea (43.3%), and fatigue (28.7%). AEs of grade 3 to 5 severity occurred in twice as many patients in the chemotherapy group as in the pembrolizumab group, with 53.3% and 26.6% of patients affected, respectively.

Based on results from the KEYNOTE-024 trial, pembrolizumab proved to be a promising first-line treatment in patients with advanced NSCLC, with superior improvements in survival and response to treatment compared with platinum-based chemotherapy.

At the ESMO 2016 Congress, investigators presented results of the phase 2 KEYNOTE-021 study, which evaluated the efficacy and safety of pembrolizumab in combination with platinum-based doublet chemotherapy in a cohort of previously untreated patients with metastatic non-SQ NSCLC.

The KEYNOTE-021 trial included 123 patients randomized to receive 4 cycles of 500 mg/m

 pemetrexed plus carboplatin (AUC of 5 mg/mL/min) every 3 weeks alone (n = 60) or with maintenance treatment of 200 mg of pembrolizumab every 3 weeks for 24 months (n = 63). Randomization was stratified by PD-L1 expression, with positive expression defined as a TPS of 1% or more and negative expression defined as a TPS of less than 1%. Using the same methodology as in other KEYNOTE trials, samples were collected at the time of diagnosis of metastatic disease and assessed for PD-L1 expression with the IHC 22C3 pharmDx assay. Patient response to treatment was assessed every 6 weeks for the first 18 weeks, then every 9 weeks through the first year and every 12 weeks through the second year using RECIST v1.1 criteria. The primary endpoint was ORR and secondary endpoints included PFS, DOR, and OS.

After a median follow-up of 10.6 months, researchers observed a significantly greater objective response (

 = .0016) in patients who received pembrolizumab plus chemotherapy (55%) compared with chemotherapy alone (29%). As assessed per RECIST v1.1 criteria, a partial response was observed in 55% of patients given pembrolizumab plus chemotherapy compared with 29% of patients given chemotherapy alone (95% CI, 9%- 42%). By the median follow-up time point, responses in both groups were durable, with 29 of 33 responders in the pembrolizumab plus chemotherapy group and 14 of 18 responders in the chemotherapy alone group experiencing an ongoing response (defined as alive and progression-free) at the time of data cutoff.

No apparent relationship between PD-L1 expression and pembrolizumab response was established in this cohort analysis. Among the patients receiving pembrolizumab plus chemotherapy (n = 60), 21 patients had a TPS of less than 1%, and of these, 12 (57%) achieved response. The response rate was similar among the 39 patients who had a PD-L1 TPS of 1% or greater, of whom 21 (54%) achieved a response. However, this group had 20 patients with a TPS of 50% or greater, and 16 (80%) achieved an objective response. Of the 19 patients who had a TPS between 1% and 49%, 5 (26%) achieved a response. In the chemotherapy alone group (n = 63), 3 (13%) of 23 patients with a TPS of less than 1% achieved a response, whereas 9 of 23 (39%) patients who had a TPS of 1% to 49% and 6 of 17 (35%) patients who had a score of 50% or greater responded.

Patients treated with pembrolizumab experienced an improved median PFS of 13.0 months compared with 8.9 months in patients receiving chemotherapy alone. PFS at 6 months was estimated at 77% (95% CI, 64%-86%) of patients treated with pembrolizumab plus chemotherapy and 63% (95% CI, 49%-74%) of patients receiving chemotherapy alone. However, OS rates were the same between the 2 treatments: 22% of patients died in both treatment groups. At 6 months, estimated survival was 92% with pembrolizumab plus chemotherapy or chemotherapy alone.

Combination therapy with pembrolizumab and was associated with a higher incidence of AEs of grade 3 to 5 severity (40%) compared with chemotherapy alone (25%). The most common treatment-related AEs were of grade 1 to 3 severity and generally occurred at a higher incidence in patients treated with pembrolizumab. Fatigue occurred in 64% and nausea in 58% of patients treated with combination therapy compared with 40% and 44%, respectively, in patients treated with chemotherapy alone. Anemia was more common in the chemotherapy group: 54% compared with 32% among patients receiving pembrolizumab therapy. There was 1 treatment-related death from sepsis in a patient receiving pembrolizumab plus chemotherapy and 2 fatalities in patients receiving chemotherapy alone, with 1 due to pancytopenia and 1 from sepsis.

The phase 2/3 KEYNOTE-010 trial compared the efficacy and safety of pembrolizumab with that of docetaxel in patients with PD-L1—expressing advanced NSCLC who experienced disease progression after 2 or more previous lines of platinum-based doublet chemotherapy. Enrolled patients had an ECOG performance status score of 0 or 1 and PD-L1 expression on at least 1% of TCs. PD-L1 expression was assessed by an IHC 22C3 pharmDx assay and scored using the same methodology as previous KEYNOTE trials. A total of 1034 patients were randomized 1:1:1 to pembrolizumab treatment intravenously over 30 minutes every 3 weeks at either 2 mg/kg (n = 345) or 10 mg/kg (n = 346) or docetaxel intravenously at 75 mg/m

 (n = 343) over 1 hour every 3 weeks. The primary endpoints were OS and PFS in the total population of patients with a TPS of 1% or greater and in patients with a TPS of 50% or greater. Secondary endpoints were safety, response rate, and DOR.

Investigators noted that median OS was similar between the patient groups treated with either dose of pembrolizumab; however, pembrolizumab demonstrated superior OS compared with docetaxel regardless of tumor PD-L1 expression. At a median follow-up of 13.1 months, the OS in patients with a TPS of 50% or greater was 14.9 months in patients given pembrolizumab at the 2-mg/kg dose and 17.3 months for the 10-mg/kg dose compared with 8.2 months with docetaxel. These survival rates were higher than those in the overall study population receiving pembrolizumab, where median OS was 10.4, 12.7, and 8.5 months, respectively. Pembrolizumab was observed to improve the rate of 1-year OS, with 52.3% survival in the 10-mg/kg group and 43.2% survival in the 2-mg/kg group compared with 34.6% survival with docetaxel.

 Pembrolizumab also provided superior PFS over docetaxel treatment in patients with a TPS score of 50% or greater. Patients with higher PD-L1 expression experienced a median PFS of 5.0 months with 2-mg/kg pembrolizumab, 5.2 months with 10-mg/ kg pembrolizumab, and 4.1 months with docetaxel.

Significantly longer survival occurred at the 2-mg/kg (

 <.0001) pembrolizumab groups compared with docetaxel. The effect on PFS between these groups was unchanged with tumor histology. Within the population of those with a TPS score of 1% or greater, median PFS was not significantly different between treatment groups: 3.9 months for the 2-mg/kg dose and 4.0 months for the 10-mg/kg dose versus 4.0 months in the docetaxel group.

In the KEYNOTE 010 trial, patients who responded to treatment (as assessed by RECIST v1.1 criteria) had only partial responses. Significant objective responses were noted in the pembrolizumab groups at either dose compared with docetaxel for both the total patient population and the subgroup of patients with a TPS of 50% or greater. Although no significant differences in responses were observed between the pembrolizumab doses in either PD-L1—expressing population, the subgroup of patients with higher PD-L1–expressing tumors had a greater response to pembrolizumab compared with the total population. Among patients with a TPS of 50% or greater, responses to pembrolizumab occurred in 30% patients treated with the 2-mg/kg dose and 29% of patients treated with the 10 mg/kg-dose compared with 8% of patients in the docetaxel group. In the total population, 18% of patients receiving either dose of pembrolizumab had partial responses, compared with 9% in the docetaxel group (

 = .0002 for 10 mg/kg). In patients with a TPS of 50% or greater, responses were longer in the pembrolizumab groups than in the docetaxel group, with a median DOR not reached for either pembrolizumab group compared with 8 months in the docetaxel group.

Treatment-related AEs of grade 3 to 5 severity occurred at a rate of 13% in the pembrolizumab 2-mg/ kg group, 16% in the pembrolizumab 10-mg/kg group, and 35% in the docetaxel group. AEs of special interest, of grade 3 to 5 severity, that occurred in 1% or more of patients, were pneumonitis and severe skin reactions.

In the pembrolizumab arm, 3 deaths occurred (2 cases of pneumonitis, and 1 case of pneumonia) at the 2-mg/kg dose and 3 at the 10-mg/kg dose (1 case each of myocardial infarction, pneumonia, and pneumonitis). In the docetaxel group, 5 patients died, with 1 case each of acute cardiac failure, dehydration, febrile neutropenia, interstitial lung disease, and respiratory tract infection.

To better characterize the long-term outcomes of patients from the KEYNOTE-010 trial, investigators performed an analysis based on an additional 6 months of follow-up and presented these updated results at ESMO 2016. In a median follow-up of 19.2 months, 60% of all patients treated with pembrolizumab at either dose were alive, progression-free, and had no additional anticancer therapy compared with 15% of responders with docetaxel.

Regardless of PD-L1 expression level, treatment with pembrolizumab at either dose continued to show superiority in survival outcomes compared with docetaxel after follow-up of an additional 6 months. Moreover, no difference was observed in OS between the pembrolizumab doses in either population expressing PD-L1. In the group of patients with a TPS greater than 50%, the 18-month median OS rate was significantly improved with pembrolizumab: 52% with the 10-mg/ kg dose (

 <.00001) and 46% with the 2-mg/kg dose (

 = .00004) compared with 24% with docetaxel.

Pembrolizumab treatment also was associated with a longer median OS of 18.8 months (HR, 0.48; 95% CI, 0.35- 0.66) at the 10-mg/kg dose and 15.8 months (HR, 0.54; 95% CI, 0.39-0.73) at the 2-mg/kg dose compared with 8.2 months with docetaxel.

In the population with the highest PD-L1 expression, pembrolizumab 2 mg/kg and 10 mg/kg were superior (

 = .00001, respectively) in the 18-month rate of PFS compared with docetaxel. PFS also was similar between the 2 doses of pembrolizumab, with 27% and 29% in the 10-mg/kg and 2-mg/kg groups, respectively, compared with 5% in the docetaxel group. The median PFS was 5.2 months for both pembrolizumab arms and 4.1 months in the docetaxel arm.

In this follow-up, ORR remained significantly higher (

<.00001) with pembrolizumab in the population with a TPS greater than 50% for PD-L1 expression: 29% and 32% of patients had tumor reductions with the 2-mg/kg and the 10-kg/mg dose, respectively, compared with 9% in the docetaxel group. Median DOR was not reached in either pembrolizumab group in this population of patients; however, response to docetaxel treatment lasted 8.1 months. Additionally, in the population with a TPS of at least 1%, ORR was higher with pembrolizumab treatment at either dose (2 mg/kg, 19%; 

 = .00004) compared with docetaxel (10%). In this population, median DOR was not reached in the pembrolizumab 10-mg/kg group at the time of follow-up; however, the response lasted 21.7 months with pembrolizumab 2 mg/kg and 7.1 months with docetaxel.26 Despite the longer follow-up and longer treatment exposure with pembrolizumab, rates of treatmentrelated AEs and AEs leading to treatment discontinuation remained lower with pembrolizumab than with docetaxel.

Among the 343 patients treated with docetaxel, 81% of patients reported AEs, with 36% experiencing AEs of grade 3 to 5 severity and 12% discontinuing treatment due to AEs. Of the 344 patients treated with 2-mg/kg pembrolizumab, 64% experienced AEs of any severity, with 13% reporting AEs of grade 3 to 5 severity and 4% discontinuing treatment due to AEs. Of the 346 patients treated with 10-mg/kg pembrolizumab, 67% reported AEs, with 17% being grade 3 to 5 in severity and 6% leading to discontinuation.26 Extended results of the KEYNOTE-010 trial confirmed that pembrolizumab is superior to docetaxel in patients with previously treated advanced NSCLC who have tumors expressing PD-L1.

Durvalumab is a selective, high-affinity, engineered human IgG1 monoclonal antibody that blocks PD-L1 binding to PD-1 and cluster of differentiation 80 (B7-1). Durvalumab does not block the interaction between PD-1 and PD-L2 so as to avoid the immune-related toxicity associated with the PD-1/PD-L2 blockade. In addition, durvalumab has an engineered triple mutation in the Fc domain to reduce antibody-dependent cell-mediated cytotoxicity and complement-dependent cytotoxicity.

In the phase 1/2 trial presented at ESMO 2016, durvalumab was evaluated as a first-line or subsequent line of therapy in patients with advanced (stage IIIB/IV) NSCLC. Fresh or archival tumor tissue was assessed for PD-L1 expression using the SP263 assay, where high PDL1 expression was defined as at least 25% of TCs staining for PD-L1 at any intensity and low PD-L1 expression was defined as less than 25% of TCs staining. Patients were intravenously treated with 10-mg/kg durvalumab monotherapy every 2 weeks for up to 12 months or until unacceptable toxicity or disease progression.

Of the 304 patients enrolled to treatment with durvalumab, 47% had non-SQ NSCLC, 53% had SQ histology, 76% had an ECOG performance status score of 1, and 85% were current or prior smokers. The primary endpoints were the safety and tolerability of durvalumab treatment, which were assessed by incidence of AEs and physical examinations through 90 days after the last dose or until initiation of alternative anticancer therapy. Among the secondary endpoints were ORR (defined as a confirmed response based on investigator assessment using RECIST v1.1 criteria) and OS.

 The safety profile of durvalumab was reported by investigators to be consistent with previous reports.

However, treatment with durvalumab induced AEs of any grade severity in 57% of patients and led to discontinuation of therapy in 5% of patients; serious AEs occurred in 4.6% of patients. The most frequent treatment-related AEs were fatigue (17%) and decreased appetite and diarrhea (9% each). Durvalumab was associated with AEs of grade 3 severity and higher in 10% of patients; these were most commonly fatigue, hyponatremia, and colitis (1% each). One fatal pneumonia event occurred after 3 doses of durvalumab treatment.

Out of 305 patients treated, 50 patients had a confirmed response to durvalumab per RECIST v1.1 criteria. The ORR was higher in patients with high PD-L1—expressing tumors compared with low PD-L1–expressing tumors. Among the 154 patients with high PD-L1 expression, 39 (25%) achieved a complete or partial response to durvalumab, including 14 patients receiving durvalumab as first-line therapy, 12 as second-line, and 13 as thirdline or higher therapy. However, of the 115 patients with low PD-L1 expression, only 7 (6%) achieved an objective response. The responses ranged from 1.4 months or longer to 22.0 months or longer, and of the responders, 48% had ongoing responses at time of data cutoff.

In the patients receiving durvalumab as second line of treatment for NSCLC, the 12-month rates of OS were 56% (95% CI, 39%-69%) for patients with high PD-L1expression and 39% (95% CI, 19%-58%) in patients with low PD-L1 expression. Additionally, the median OS was 17.8 months (95% CI, 7.9-22.4 months) in patients with high PD-L1 expression compared with 8.2 months (95% CI, 4.9-15.5 months) in patients with low PD-L1 expression. Similar rates of 12-month OS were observed in patients receiving third-line or higher durvalumab treatment: 51% (95% CI, 36%-64%) among those with high PD-L1 expression and 37% (95% CI, 26%-48%) among those with low PD-L1 expression. In patients with high PD-L1 expression, median OS was extended to 13.0 months compared with 7.6 months in patients with low PD-L1 expression.

 Although these preliminary data show durvalumab may be promising in the treatment of patients with advanced NSCLC, survival results for its use as first-line therapy were not mature and larger trials are needed to confirm its efficacy and safety.

Avelumab is an investigational fully human anti—PD-L1 IgG1 monoclonal antibody. It is believed to elicit activation of T cells and the adaptive immune system by blocking PD-L1 interactions. By retaining a native Fc region, avelumab may engage the innate immune system and induce antibody-dependent cell-mediated cytotoxicity.28 Avelumab is currently being investigated in the multicenter, phase 3 JAVELIN Lung 200 clinical trial.

This trial aims to demonstrate superiority with regard to the OS of avelumab versus docetaxel in patients with advanced (stage IIIB/IV) NSCLC who have experienced disease progression after receiving platinum-containing doublet chemotherapy. Secondary endpoints include ORR according to RECIST v1.1 criteria, PFS, quality of life, and safety. Enrolled patients must have an ECOG performance status score of 0 or 1 and PD-L1—positive tumors. Patients with recurrent NSCLC are randomized 1:1 for IV treatment with 10 mg/kg of avelumab for 1 hour every 2 weeks or docetaxel at 75 mg/m

 once every 3 weeks. The trial has an estimated primary completion date of January 2018.

In updates presented at the ESMO 2016 Congress, evidence from the first 850 patients enrolled in the OAK trial indicated that atezolizumab is superior to chemotherapy as a second- or higher line of therapy in patients with stage IIIB/IV NSCLC. Atezolizumab influenced survival benefit and extended treatment response time in patients with both non-SQ and SQ disease. Significant improvements were observed regardless of PD-L1 expression levels in tumor-infiltrating ICs and TCs. Pronounced survival benefit and response were seen in patients with greater than 50% of PD-L1−expressing tumors and at least 10% of ICs expressing PD-L1. Much like the results in the BIRCH and POPLAR trials, the results of the OAK trial revealed that PD-L1 is an effective predictive biomarker for identifying patients with NSCLC most likely to benefit from atezolizumab therapy, which could aid clinicians in selecting personalized treatment plans.

Suggestions that PD-L1 expression is variable and cannot act as a predictive biomarker for treatment efficacy came from updates of the Checkmate trials in patients with NSCLC given nivolumab. In the 2-year follow-up of the Checkmate-017 and CheckMate-057 trials, nivolumab was observed to be superior to chemotherapy in terms of survival outcomes and response rates in patients with non-SQ and SQ NSCLC who had disease recurrence with platinum-based chemotherapy. However, the 2-year survival outcomes were only improved (compared with chemotherapy) in those with non-SQ NSCLC and 5% to 10% of tumors expressing PD-L1. PD-L1 expression had no influence on nivolumab efficacy in SQ NSCLC. Additionally, updated findings from the CheckMate-026 trial revealed that the primary endpoint was not met and nivolumab was not effective as a first-line therapy in patients with advanced NSCLC and PD-L1 expression in at least 5% of tumors.

A supplementary study of nivolumab presented at ESMO 2016 showed evidence that patients with NSCLC enrolled in the KEYNOTE clinical trials had variable PD-L1 expression levels: only 28% of patients had detectible PDL1 expression from at least 50% of tumors. Updated results from the KEYNOTE-024 trial demonstrated superiority in response to pembrolizumab as a first-line therapy in patients with NSCLC and at least 50% of tumors expressing PD-L1 compared with platinum-based chemotherapy. Previously untreated patients with metastatic non-SQ NSCLC enrolled in the KEYNOTE-021 trial had a superior partial response to treatment with platinum-based doublet chemotherapy and continued pembrolizumab as maintenance treatment compared with chemotherapy alone. However, survival outcomes and durability of responses to treatment were the same with and without pembrolizumab maintenance treatment. Patients with 50% or more tumors expressing PD-L1 had an enhanced response to pembrolizumab maintenance treatment; however, patients with less PD-L1 expression had similar rates of response as chemotherapy alone.

In the additional 6-month follow-up of patients with advanced NSCLC from the KEYNOTE-010 trial, pembrolizumab was superior to chemotherapy as a secondor higher line of treatment. Results similar to those from the KEYNOTE-021 trial were revealed, indicating PD-L1 expression was a predictor of clinical response. Patients with 50% or more tumors expressing PD-L1 had enhanced survival outcomes and tumor responses compared with the overall population with 1% or greater of TCs expressing PD-L1; however, both populations had superior outcomes compared with chemotherapy.

Anti-PD-1/PD-L1 Immunotherapy in Lung Cancer: Updates From the ESMO 2016 Congress | Special Issues

Anti-PD-1/PD-L1 Immunotherapy in Lung Cancer: Update From the ESMO 2016 Congress