As more agents are evaluated either in addition to or following PD-1/PD-L1 inhibitors, more information may be learned about how best to leverage alternate biomarkers, according to Marina C. Garassino, MD.
Although PD-L1 has evolved to become a biomarker of great interest and clinical utility in the treatment of patients with non–small cell lung cancer (NSCLC), several mechanisms contribute to immunotherapy resistance. As more agents are evaluated either in addition to or following PD-1/PD-L1 inhibitors, more information may be learned about how best to leverage alternate biomarkers, according to Marina C. Garassino, MD.1
“We have to integrate all our sciences to create an individualized medicine [approach],” Garassino said in a presentation delivered during the 20th Annual Winter Lung Cancer Conference®. “I hope that we will be able to integrate all our knowledge to have a true personalized medicine [approach], because it’s clear now that we have an absence of consistent biomarkers. We unfortunately only have PD-1/PD-L1 as biomarkers able to address proper strategies. Maybe in the future, there will be not a one-size-fits-all approach, but we will go with the right treatment for each patient because we understand the mechanism of resistance.”
Over the course of her presentation, Garassino, director of the thoracic program and a professor of medicine in the department of Hematology and Oncology at the University of Chicago Medical Center in Illinois, gave an update on findings observed with anti–PD-1 combination regimens and outlined the efficacy of other treatment approaches for patients with NSCLC, including bispecific T-cell engagers (BiTEs), cancer vaccines, and tumor-infiltrating lymphocyte (TIL) therapy.
Despite PD-1/PD-L1 inhibitors providing life-altering survival benefits for many patients with NSCLC, a significant portion of patients do not elicit the same efficacy. Thus, clinicians have evaluated the agents in combination with other treatment modalities in several clinical trials.
In the phase 2 TACTI-002 trial (NCT03625323), the anti–PD-1 agent pembrolizumab (Keytruda) was combined with the soluble LAG-3 protein eftilagimod alpha and examined for patients with untreated metastatic NSCLC unselected for PD-L1 expression. Investigators hypothesized that the combination could display a stronger antitumor response than pembrolizumab monotherapy due to the ability of eftilagimod alpha to bind to a subset of MHC class II molecules to mediate antigen presenting cell and CD8 T-cell activation, subsequently leading to T cell recruitment. In preclinical models, study authors noted the antitumor activity of the PD-1 agonist was synergistically enhanced in combination with eftilagimod alpha.2
In part A of the study, adult patients with advanced/metastatic (stage IIIb/IV) NSCLC (n = 114), both squamous and nonsquamous, were enrolled from March 2019 to November 2021. Patients received subcutaneous eftilagimod alpha 30 mg every 2 weeks plus intravenous (IV) pembrolizumab 200 mg every 3 weeks for 8 cycles, followed by the same dosage strength for both agents administered every 3 weeks for 9 cycles. After 1 year, patients received pembrolizumab monotherapy every 3 weeks for 16 cycles, then progression-free survival (PFS) and overall survival (OS) follow-up were performed after another year.
The primary end point was overall response rate (ORR) per immune-modified RECIST (iRECIST) criteria. Secondary end points included ORR by RECIST 1.1, duration of response (DOR), safety, PFS, and OS.
Findings from the trial showed that per iRECIST, the ORR was 38.6% (95% CI, 29.6%-48.2%), with 2 patients achieving a complete response. Additionally, the disease control rate (DCR) was 73.7% (95% CI, 64.6%-81.5%). Nineteen patients experienced disease progression.2
In terms of safety, most patients (99.1%) experienced an any grade treatment-emergent adverse effect (TEAE). Serious TEAEs occurred in 39.5%, grade 3 and above TEAEs occurred at a rate of 51.8%, and 10.5% of patients experiences a grade 5 TEAE. TEAEs leading to discontinuation of study treatment (20.2%) were also reported.
Study authors wrote that the combination displayed a promising ORR compared with the historical control of single-agent pembrolizumab, with deep, durable responses and a well-tolerated safety profile. Late-stage clinical investigation is warranted, they concluded.2
For patients with NSCLC with a PD-L1 tumor performance score (TPS) of at least 50%, PD-1 inhibitors have been showing promise across multiple clinical trials when administered in combination with immunotherapy agents, Garassino acknowledged.
In the phase 2 CITYSCAPE trial (NCT03563716), the novel anti-TIGIT antibody tiragolumab was evaluated in combination with the anti-PD-L1 antibody atezolizumab (Tecentriq) in patients with locally advanced or metastatic NSCLC and PD-L1 TPS of at least 1% (n = 135). Patients were randomly assigned 1:1 to receive either IV tiragolumab 600 mg every 3 weeks plus IV atezolizumab 1200 mg every 3 weeks or the same regimen with placebo in place of tiragolumab. Treatment continued until disease progression or loss of clinical benefit.3
The coprimary end points were ORR and PFS. Key secondary end points included safety, DOR, and OS. Efficacy analysis by PD-L1 status and patient-reported outcomes served as exploratory end points.
At the June 30, 2019, data cutoff, at a median follow-up of 5.9 months, the investigator-assessed median PFS for patients with a TPS of at least 50% was 16.6 months (95% CI, 5.5-22.3) vs 4.1 months (95% CI, 2.1-6.8) for patients in the tiragolumab arm (n = 29) and the placebo arm (n = 29), respectively (HR, 0.29; 95% CI, 0.15-0.53). Moreover, the median DORs were 15.7 months (95% CI, 9.1-not evaluable [NE]) compared with 8.2 months (95% CI, 5.6-10.4), respectively. ORRs were 69.0% vs 24.1%, respectively.
However, patients with a PD-L1 TPS of less than 50% did not experience a similar level of benefit with the tiragolumab-containing regimen; patients who received the combination (n = 38) achieved a median PFS of 4.0 months (95% CI, 1.6-5.6) compared with 3.6 months (95% CI, 1.4-5.5) in the 39-patient placebo group (HR, 1.07; 95% CI, 0.67-1.71). The median DORs were 17.8 months (95% CI, 8.3-24.2) vs 18.8 months (95% CI, 15.9-22.8), respectively, and the ORRs were 15.8% vs 17.9%, respectively.
In another phase 2 study, ARC-7 (NCT04262856), investigators enrolled 150 patients with stage IV NSCLC with treatment-naïve metastatic disease and a PD-L1 TPS of at least 50%. Patients were randomly assigned 1:1:1 to receive either the anti–PD-1 monoclonal antibody zimberelimab (arm 1; n = 44); zimberelimab in combination with the anti-TIGIT monoclonal antibody domvanalimab (arm 2; n = 44), or etrumadenant, an antagonist of the A2a and A2b receptors, plus zimberelimab and domvanalimab (arm 3; n = 45). Zimberelimab was given at a dose of 360 mg IV every 3 weeks, domvanalimab at a dose of 15 mg/kg IV every 3 weeks, and etrumadenant 150 mg was given orally once daily.4
Patients in arm 1 experienced an ORR of 27% (95% CI, 15%-42.8%) compared with 41% (95% CI, 26.3%-56.8%) and 40% (95% CI, 25.7%-55.7%) in arms 2 and 3, respectively. Additionally, patients in both arms 2 and 3 experience a benefit in terms of PFS compared with the median figure of 5.4 months (95% CI, 1.8-8.6) in arm 1. Patients in arm 2 achieved a median PFS of 12.0 months (95% CI, 5.5-NE; HR vs arm 1, 0.55; 95% CI, 0.31-1.0) and those in arm 3 experienced a median PFS of 10.9 months (95% CI, 4.8-NE; HR vs arm 1, 0.65; 95% CI, 0.37-1.1).
The multitargeted tyrosine kinase inhibitor cabozantinib (Cabometyx), both with and without atezolizumab, was evaluated in patients with advanced NSCLC who were previously treatment with an immune checkpoint inhibitor as part of the phase 1/2 COSMIC-021 study (NCT03170960). Patients in cohort 7 (n = 81) received oral cabozantinib 40 mg daily plus IV atezolizumab 1200 mg every 3 weeks. In cohort 20 (n = 31), patients were treated with oral cabozantinib 60 mg daily.5
The primary end point was ORR. The secondary end point was safety, and DOR, PFS, and OS served as exploratory end points.
Overall, patients in cohort 7 experienced an ORR of 19% with a DCR of 80%. Comparatively in cohort 20, the ORR and DCR were 6% and 65%. When cohort 7 was stratified by PD-L1 status, patients with a PD-L1 TPS of at least 1% (n = 41) experienced an ORR of 20% compared with 11% for patients with a PD-L1 TPS of less than 1% (n = 19). The DCRs were 80% and 74%, respectively.
Finally, in the phase 2 S1800A trial (NCT03971474) pembrolizumab was combined with the VEGFR-targeted monoclonal antibody ramucirumab (Cyramza) for the treatment of patients with stage IV, previously treated NSCLC. Patients were randomly assigned 1:1 to receive either investigators choice of standard-of-care chemotherapy or pembrolizumab 200 mg every 3 weeks for up to 35 cycles plus ramucirumab 10 mg/kg every 3 weeks. OS was the primary end point; secondary end points included DCR, DOR, and PFS.6
Efficacy-evaluable patients in the combination arm (n = 69) achieved a median OS of 14.5 months (95% CI, 13.9-16.1) vs 11.6 months (95% CI, 9.9-13.0) in the 67-patient standard-of-care arm (HR, 0.69; 80% CI, 0.51-0.92; log-rank P = .05).6
After talking through the combination regimens, Garassino briefly spoke on the rationale of using BiTEs for the treatment of patients with lung cancer. BiTEs recognize and bind to the tumor cell with one arm, recognizing targets including CEACAM5 or HER2, and T cells (CD3, PD-1) or natural killer cells (CD16) with the other, thus overcoming the major histocompatibility complex, Garassino explained in the presentation. Additionally, these agents can activate the effector T-cells within the tumor microenvironment, minimizing systemic immune activation.1
Tarlatamab (AMG 757) is a half-life extended BiTE that binds to DLL3 on tumor cells and CD3 on T cells, eventually leading to T-cell mediated tumor death. The agent was examined in patients with small cell lung cancer after progression with at least 1 platinum-based regimen as part of a phase 1 study (NCT03319940). Patients received tarlatamab intravenously every 2 weeks with or without step-up dosing.7
At the January 11, 2021, data cutoff, the DCR was 43%. Partial responses were seen across 13% of patients treated at all dose levels (n = 60), with patients 5 of 8 who received a 100-mg dose of tarlatamab experiencing a response. The estimated DOR was at least 6 months in 71% of patients (95% CI, 26%-92%) with any response. Study authors concluded that the agent had an acceptable safety profile up to 100 mg with rapid and durable responses. The trial is still ongoing.
Garassino then touched on cancer vaccines, which have also shown some promise in NSCLC. These agents have multiple designs, all with the goal of eliciting specific immune responses to tumor antigens. The 2 primary target types of cancer vaccines are tumor-associated antigens and tumor-specific antigens, she noted.8
Vaccines targeted toward tumor-associated antigens typically target either overexpressed proteins/differentiation antigens or cancer testis antigens. Tumor specificity for vaccines aimed at overexpressed proteins is variable, but these agents usually have a high central tolerance. Conversely, those directed at cancer testis antigens have good specificity with a low tolerance. Both classes have a high prevalence in multiple patients.
Tumor-specific antigen vaccines can target oncoviral antigens, shard neoantigens, or private neoantigens. All of these have ideal tumor specificity and no central tolerance. The prevalence in multiple patients is also high, except for private neoantigen-targeted vaccines, where it is low.
Multiple clinical trials of cancer vaccines are ongoing for patients with NSCLC, both as monotherapies and as components of combination regimens.
TIL therapy has previously shown clinical activity in patients with melanoma but has not been extensively studied in NSCLC, Garassino said. Investigators have hypothesized that this class of agent may be effective in NSCLC because of its ability to release T cells from a suppressive microenvironment and reactivate them against the tumor. TIL treatment for patients metastatic lung cancer not previously treated with an anti–PD-1/PD-L1 agent is being evaluated in a phase 1 trial (NCT03215810.9
Patients with stage IV NSCLC who were not candidates for surgical resection will receive cyclophosphamide and fludarabine lymphodepletion, TIL infusion, and interleukin-2, followed by maintenance nivolumab (Opdivo). Twenty patients received nivolumab, but only 16 patients proceeded to TIL therapy; all 16 were eligible for safety and survival analysis and 13 were evaluable for response post TIL.
The primary end point is safety. Secondary end points included ORR and DOR.
Safety findings showed that the severe toxicity rate was 12.5%, with most TEAEs resolving within 1 month following TIL. Most adverse effects were attributed to lymphodepletion and interleukin-2. Common any grade TEAEs included decreased lymphocyte count (100%), decreased white blood cell count (100%), anemia (100%), decreased platelet count (94%), and decreased neutrophil count (88%). No grade 5 TEAEs were reported.
Radiographic response was observed occurred for 6 of 13 evaluable patients, including 2 complete responses that were still ongoing 1.5 years later. Tumor burden was reduced in 11 patients, with a median best change of 35%. Study authors concluded that TIL infusion displayed a manageable safety profile and that further study of the approach in NSCLC is warranted.9
Garassino concluded her presentation noting that there are many ongoing stratagies to watch for patients with NSCLC as novel biomarkers and therapies begin to read out longer term data. The future depends on a combination of coinhibitory strategies leveraging bispecific antibodies, cytokines, and cellular therapies.