Cancer Immunotherapy: Evolving Treatment Options in Squamous NSCLC

This article is part II of a series. View parts I and III:Evolving Paradigms in Squamous NSCLC> >

Activation of T-cell function has become a main focus of immunotherapeutic approaches aimed at enhancing an anticancer immune response and suppressing mechanisms used by tumor cells for immune evasion.^96-97With the recognition of immune evasion as a key pathogenic mechanism in NSCLC, cancer immunotherapy strategy has evolved to a main therapeutic focus, leading to the integration of nivolumab into clinical practice for the treatment of platinum-resistant SCC, and ongoing evaluation of multiple other immunotherapeutic agents. Similar to other cancers susceptible to immunotherapy, such as melanoma, the upregulation of immune inhibitory molecules in the tumor microenvironment appears to mediate immune resistance of NSCLC.^98-100Key components of inhibitory immune checkpoints that have been implied in the escape from an humoral antitumor response include PD-1, its ligands PDL1 and PD-L2, and cytotoxic T lymphocyte-associated antigen-4 (CTLA- 4) (FIGURE 4).^98-100The CTLA-4 and PD-1–negative regulatory checkpoints control the T-cell activity induced by inflammatory signals. CTLA-4 acts mainly within the lymph nodes to regulate the magnitude of early activation of naïve and memory T cells, whereas PD-1 mainly affects T-cell activity in peripheral tissues through interaction with PD-L1 and PD-L2 (FIGURE 4). Both PD-L1 and PDL2 expression are often upregulated in NSCLC.¹⁰¹Furthermore, NSCLC tumors are infiltrated by immunosuppressive leukocytes expressing factors such as PD-L1, contributing to the inhibition of the host antitumor immune response.^102,103Immune resistance in NSCLC may also be mediated by downregulation of components of the antigen-presenting machinery, particularly among smokingassociated lung cancers and SCC.^20,104-105Different immunologic properties of squamous and nonsquamous lung cancers suggest that SCC may be particularly amenable to immunotherapeutic approaches. Often, SCC contains inactivating mutations in HLAA class I genes, which may lead to loss of function and reduced expression of tumor antigens, a possible immune-evading strategy.^20,105Compared with nonsquamous tumors, SCC also exhibits frequent upregulation of tumor antigens, including melanomaassociated antigens A3 and A4 and NY-ESO-1.^38,106More extensive infiltration of SCC with both immunosuppressive leukocytes and CD8 effector cells has been linked to poor survival, suggesting the presence of an immune-suppressive environment.^{38,102,103,106}

Figure 4. CTLA-4 and PD-1 immunologic checkpoints. (A) Activation of T cells by antigen-presenting cells requires the interaction of antigen presentation via MHC and costimulatory interaction of B7 and CD28. This T-cell response is negatively regulated by surface expression of CTLA-4, which interacts with B7 and and downregulates the T-cell response, preventing autoimmunity. (B) The PD-1 receptor is expressed by T and B lymphocytes, natural killer cells, and T regulatory cells. Among various functions of the PD-1 checkpoint, downregulation of effector T-cell function by interaction of PD-L1 with PD-1 and B7 is a presumptive key mechanism leading to immunosuppression in the tumor microenvironment.

Four antibodies targeting PD-1 or its main ligand, PD-L1, are in advanced clinical development in NSCLC; these include the anti–PD-1 antibody pembrolizumab and the anti–PD-L1 mAbs MPDL3280A and MEDI4736. The anti-PD-1 antibody nivolumab was recently approved for subsequent-line treatment in patients with advanced platinum-resistant SCC and remains in investigation for other lines of treatment. Anti–CTLA-4 agents in development include the CTLA- 4–blocking mAbs ipilimumab and tremelimumab.

Clinical investigation involves both single-agent and combination strategies: the CTLA-4 and PD-1 checkpoints are nonredundant pathways for the regulation of T-cell responses, and preclinical data suggest enhanced activity of combined pathway inhibition. Clinical evidence for additive and synergistic potential of combination regimens is emerging.^{98-99,107-108}Immunotherapy also has synergistic potential when combined with taxanes and platinum-based compounds. The latter induce the release of tumor-specific antigens from dying tumor cells, which promotes T-cell activation and increased immune response.¹⁰⁹

PD-1 checkpoint agents in development for SCC

Exploration of PD-1 checkpoint blockade in SCC has led to the addition of the first immunotherapeutic agent for platinum-resistant SCC, nivolumab. Beginning with the PD-L1 antibody BMS-936559, which was the first to show durable tumor regression in patients with a variety of solid tumors, including NSCLC and melanoma,¹¹⁰multiple trials have shown clinical efficacy of single-agent PD-1 and PD-L1 blockade in unselected NSCLC, with response rates ranging around 16% to 23%.³⁹Positivity for tumors for PD-L1 and PD-L1 expression levels are being evaluated as potential biomarkers for the selection of tumors sensitive to PD-1 blockade; outcomes remain controversial with differences between studies and agents. Subgroup analyses of trials with pembrolizumab¹¹¹and MPDL3280A^112-113suggest higher response rates among patients with increased tumor PD-L1 expression. In contrast, PD-L1 expression was not prognostic or predictive of any efficacy endpoint in trials with nivolumab.³⁷These discrepancies may in part derive from inconsistencies with detection methods and stratification thresholds used across trials. Primary and secondary endpoints of ongoing clinical trials therefore include stratification or subgroup analysis based on PD-L1 expression.^3,114History of smoking, correlated with SCC, has been associated with response to anti–PD-1/PD-L1 mAbs, with generally higher response rates in current or former smokers compared with never or light smokers; this has been observed for nivolumab (30% vs 0%),⁸⁹pembrolizumab (22.5% vs 10.3%),¹¹¹and MPDL3280A (42% vs 10%).¹¹²This observation has been attributed to a larger number of somatic point mutations in the lung cancers of smokers—including largely patients with SCC—leading to the formation of more immunogenic neoepitopes.^40,115

Outcomes of the phase II CheckMate 017 and randomized phase III CheckMate 063 trials have demonstrated that in the setting of advanced platinum-resistant SCC, the anti–PD-1 mAb nivolumab is effective and superior over docetaxel.^37,42Furthermore, evidence for synergistic activity of nivolumab with CTLA-4–targeted agents is emerging. Results from the phase I/II CheckMate 032 trial in patients with recurrent NSCLC after platinum chemotherapy suggest higher activity of double checkpoint inhibition. According to interim results, nivolumab monotherapy and nivolumab/ipilimumab combination therapy produced ORR of 18% and 17%, respectively, and disease control rates of 38% and 54%, respectively.^107-108The median OS was 4.4 months with monotherapy (95% CI, 2.9- 9.4) and 8.2 months with combination therapy (95% CI, 3.7, not reached).^107-108The efficacy of this combination regimen versus standard chemotherapy in the first-line treatment of advanced or recurrent NSCLC is being evaluated in an ongoing, large phase III trial (CheckMate 227, NCT02477826; n = 1980) A second randomized, phase III trial will evaluate single-agent nivolumab versus chemotherapy (investigator’s choice, based on histology) in firstline treatment for PD-L1–positive advanced or recurrent NSCLC (CheckMate 026, NCT02041533; n = 535); primary and secondary objectives include PFS, ORR, and OS among patients with strong PD-L1 expression. Combination regimens of nivolumab with various chemotherapeutic regimens in the first-line setting in advanced NSCLC are being explored in a phase I, multiarm trial (CheckMate 012, NCT01454102; n =412).

A second anti–PD-1 mAb in advanced development in NSCLC is pembrolizumab (MK-3475), a humanized IgG4 kappa isotype. The phase I KEYNOTE-001 trial evaluated outcomes of histologically unselected patients (n = 495) with pretreated (n = 304) or previously untreated (n = 101) advanced NSCLC in correlation to PD-L1 expression.¹¹¹The overall ORR was 19.4%, with a median duration of response of 12.0 months; median PFS and OS durations were 3.7 and 12.0 months, respectively. PD-L1 expression in at least 50% of tumor cells was associated with a higher ORR of 45.2% in a validation cohort for this cutoff, and a median PFS of 6.3 months.¹¹¹Pembrolizumab was associated with fatigue, pruritus, and decreased appetite; grade 3 or greater severity occurred in less than 10% of patients, and the incidence of pneumonitis was 3.6%.¹¹¹The phase II/III KEYNOTE-010 study (NCT01905657) will compare the efficacy of two dosing schemes of pembrolizumab with single-agent docetaxel in patients who have progressed after platinum-containing therapy. The phase III KEYNOTE-024 study (NCT02142738; n = 300) has been designed to validate pembrolizumab versus five different platinum-based chemotherapies in metastatic disease of all NSCLC subtypes. Both trials are limited to patients with PD-L1–positive tumors, with KEYNOTE-024 further stratifying for strong expression.

Agents that target PD-L1 may mediate different effects than PD-1 mAb, because it also blocks interaction of PD-L1 with other receptors, which is not prevented by PD-1–blocking agents (FIGURE 4). Additionally, PD-L1 antibodies do not affect the interaction of PD-1 and PD-L2; disruption of this interaction has been implicated in autoimmune lung toxicity. Anti–PD-L1 agents in development for NSCLC include MPDL3280A (atezolizumab), an engineered human anti–PD-L1 mAb, and MEDI4736, a human IgG1 anti-B7H1 mAb that blocks PD-L1 binding to PD-1 and CD80 with high affinity and selectivity. Single-agent MPDL3280A was assessed versus docetaxel in a randomized phase II study (POPLAR, NCT01903993) among patients with NSCLC (n = 287) who had previously received chemotherapy with docetaxel. Tumors included squamous and nonsquamous tumors and were scored in respect to PD-L1 expression on tumor cells (TC) and tumor infiltrating immune cells (IC). A higher grade of TC and IC PD-L1 expression correlated with higher efficacy of MPDL3280A compared with docetaxel (in patients with highest TC/IC; OS HR, 0.47; PFS HR, 0.56; and ORR, 38% vs 13% on docetaxel), whereas response to treatment did not differ among patients with low PD-L1 levels (OS HR, 1.22; ORR 8% vs 10% on docetaxel). Fewer patients receiving MPDL3280A experienced grade ≥3 AEs (43% vs 56% on docetaxel). A phase Ib study assessing MPDL3280A in combination with platinum-doublet chemotherapy in patients (n = 37) with previously untreated locally advanced or metastatic NSCLC (81% nonsquamous; 19% SCC) reported good tolerability, with most AEs resembling those associated with chemotherapy. MPDL3280A-related grade 3-4 AEs included anemia, neutropenia, and thrombocytopenia. Across all arms, the ORR (95% CI) was 67%; responses were seen in each arm independent of PD-L1 expression. Evaluation of MPDL3280A is ongoing in a single-arm phase II study among patients with advanced or metastatic NSCLC with PD-L1 expression (BIRCH, NCT02031458; n = 635) and in comparison to docetaxel in a randomized phase III trial among unselected patients with platinum-resistant disease (OAK, NCT02008227; n = 1225). A maintenance trial will assess MPDL3280A to BSC after adjuvant cisplatin chemotherapy in PD-L1–selected patients with resected 1B-IIIA NSCLC (NCT02486718; n = 845). Two phase III trials are focused on the evaluation of MPDL3280A as first-line treatment for patients with SCC. IMpower 111 (NCT02409355; n = 400) will compare single-agent MPDL3280A to gemcitabine plus cisplatin or carboplatin in patients who are chemotherapy-naïve with stage IV SCC. IMpower 131 (NCT02367794; n = 1200) is a three-arm study, comparing MPDL3280A in combination with carboplatin/sb-paclitaxel or in carboplatin/nab-paclitaxel to chemotherapy with carboplatin/ nab-paclitaxel in the first-line setting in advanced SCC.

An ongoing phase I/II trial (NCT01693562) evaluating singleagent MEDI4736 for the treatment of multiple solid malignancies reported promising outcomes in patients with NSCLC (n = 198; including 82 patients with SCC). Among 149 patients evaluable for response (≥24 weeks of follow-up), the ORR was 14%, with higher rates among patients with PD-L1–positive tumors (23%).41 Patients with SCC exhibited a higher response rate than patients with nonsquamous tumors (21% vs 10%). Responses were durable with 76% ongoing (range 0.1 to >35 weeks).⁴²The AE profile was manageable, with drug-related AEs reported in 48% of patients; most frequently fatigue (14%), decreased appetite (9%), and nausea (8%). Grade 1-2 pneumonitis occurred in 2 (1%) of patients; 6% of patients experienced grade ≥3 drug-related AEs.⁴²Promising clinical activity, particularly among patients with PD-L1–negative tumors, was observed in an ongoing phase Ib study (NCT02000947) evaluating MEDI4736 in combination with the anti–CTLA-4 mAb tremelimumab.^107-108Across all doses, a substantial proportion of patients achieved a partial response or stable disease (disease control rate, 41%). Patients with PD-L1–negative tumors had a comparable ORR to the whole cohort (9/33, 27% and 17/63, 27%, respectively) and a disease control rate of 48%. Grade 3/4 toxicities occurred in 41 of 102 evaluable patients but were less frequent among patients receiving the combination selected for phase III evaluation (4/22 patients).^107-108This combination regimen is under further clinical investigation in a randomized, phase III, multiarm trial in patients with advanced platinum-resistant NSCLC. Allocation of patients will be performed based on PDL1 expression in tumors, comparing MEDI4736/tremelimumab versus standard of care treatment in patients with PD-L1–negative tumors, and single-agent MEDI4736 versus standard of care treatment in patients with PD-L1–positive tumors (NCT02352948; n = 900). The phase III MYSTIC trial will evaluate single-agent MEDI4736 and MEDI4736/tremelimumab in the first-line setting, with standard of care platinum-doublet chemotherapy as active comparator (NCT02453282; n = 675).

CTLA-4 blockade in SCC

Ipilimumab is an anti–CTLA-4 mAb that binds to the CTLA-4 receptor, thereby restoring downstream signaling and allowing immune activation to persist.^37,89In NSCLC, ipilimumab has been evaluated in combination with paclitaxel and carboplatin as first-line intervention in stage IIIB/ IV NSCLC, with outcomes suggesting activity, particularly in SCC. The randomized trial was designed to compare efficacy of three treatment regimens, including a phased ipilimumab regimen of 2 chemotherapy cycles followed by 4 chemotherapy cycles, an ipilimumab concurrent arm with 4 cycles of chemotherapy in combination with ipilimumab followed by 2 chemotherapy cycles with a placebo, and chemotherapy plus placebo. Patients in the ipilimumab arms received maintenance ipilimumab every 3 months until progression.¹¹⁶The study demonstrated PFS benefits for the phased ipilimumab combination regimen versus platinum-doublet chemotherapy (per immune-related [ir] response criteria; irPFS, HR, 0.72;P<.05); the concurrent regimen did not yield benefits over chemotherapy. PFS benefits in the phased regimen were larger among patients with SCC than with nonsquamous NSCLC (irPFS: HR, 0.55 [95% CI, 0.27-1.12] versus HR, 0.82 [95% CI, 0.52-1.28]; mWHO-PFS, HR, 0.40 [95% CI, 0.18- 0.87] vs HR, 0.81 [95% CI, 0.53-1.26]). Overall, phased ipilimumab, concurrent ipilimumab, and control treatments were associated with a median OS of 12.2, 9.7, and 8.3 months and with grade 3 and 4 immune-related AEs of 15%, 20%, and 6%, respectively. Two patients (concurrent, one patient; control, one patient) died from treatment-related toxicity.¹¹⁶The clinical activity of ipilimumab plus paclitaxel and carboplatin is being investigated in two randomized phase III trials limited to patients with advanced or recurrent SCC (NCT01285609; n = 920; NCT02279732; n = 867).

Immune vaccines

TG4010 is a cancer vaccine based on a modified attenuated vaccinia of Ankara (MVA) poxvirus that encodes the MUC1 tumor-associated antigen and interleukin-2. The randomized phase II/III TIME study is evaluating the addition of TG4010 to chemotherapy in first-line therapy of advanced NSCLC (NCT01383148; n = 1000). Recently reported interim analyses indicate that a PFS benefit of TG4010 was limited to patients with low levels of triple-positive activated lymphocytes and to patients with nonsquamous tumors; phase III of the trial will therefore only continue in patients with nonsquamous tumors.¹¹⁷

Targeted therapies in SCC of the lung

Targeted therapy for SCC has become a focus of research, based on the identification of recurrent genetic alterations that may confer susceptibility to investigational agents. Clinical trials of multiple targeted agents, including inhibitors of PIK3CA pathway components, FGFR1, and EGFR inhibitors, among others, are ongoing (TABLE 1). The phase II/III SWOG biomarker-targeted Lung-MAP trial has been specifically designed to evaluate targeted agents in second-line therapy of recurrent, platinum-resistant, advanced SCC (NCT02154490; n = 1000), with PFS as main outcome measure. Based on tumor-specific gene expression profiles, patients with SCC (validated by IHC) will be randomized to a corresponding targeted agent or docetaxel. Targeted agents in experimental arms include the PI3 kinase inhibitor GDC-0032 (tumors positive for PI3KCA), palbociclib isethionate (tumors positive for CDK4/6, CCND1, CCND2, and CCND3), FGFR inhibitor AZD4547 (tumors positive for FGFR1, FGFR2, and FGFR3), and rilotumumab/erlotinib (tumors positive for HGF/c-MET). Patients with tumors without one of these biomarkers will be randomly assigned to PDL1–targeted immunotherapy with the anti-B7H1 mAb MEDI4736 versus docetaxel. Similarly, the phase II SAFIR02 Lung intergroup trial (NCT02117167; n = 650) is comparing potential PFS benefits with targeted therapy versus standard treatment (pemetrexed in nonsquamous disease and erlotinib in squamous disease) in patients with advanced or recurrent NSCLC. Patients who are chemotherapy naïve are eligible for first-line chemotherapy. Targeted treatments include the mTOR inhibitor AZD2014, the FGFR inhibitor AZD4547, the AKT inhibitor AZD5363, the EGFR inhibitor AZD8931, the MEK inhibitor selumetinib, and the VEGF/EGFR inhibitor vandetanib. The choice of targeted agent will be based on molecular anomalies identified by high-throughput genome analysis (CGH array, next-generation sequencing).

Table 1. Genetic Alterations in Lung Squamous Cell Carcinoma

Adapted and expanded from Gandara⁵and Chen⁴

Targeted treatment in SCC: emerging clinical trial data

EGFR inhibitors, specifically the small molecule inhibitors erlotinib, afatinib, and gefitinib, are standard-of-care treatment for NSCLC withEGFR-activating mutations, mostly adenocarcinomas.³Such sensitizing mutations are not common in SCC; however, wild-type EGFR is expressed and EGFR amplifications have been identified in a subset, and based on moderate activity, singleagent erlotinib and gefitinib are considered acceptable later-line treatment options.^3,33,81Several other members of the ErbB family of receptors, specifically HER2, HER3, HER4, and the ligand NRG1, may contribute to SCC pathogenesis; therefore, broader inhibition of the ErB family has been pursued therapeutically.¹¹⁴Recent outcomes from the randomized phase III LUX-Lung 8 trial have demonstrated superior outcomes with afatinib, an irreversible ErbB family blocker that affects signaling through all homodimers and heterodimers, compared with erlotinib as second-line treatment for patients with advanced SCC.³³Patients were selected based on squamous cell histology but notEGFRmutation status. At a median follow-up of 18.4 months, patients randomized to afatinib (n = 398) had significantly higher OS (median 7.9 months [95% CI, 7.2-8.7] vs 6.8 months [5.9-7.8]; HR, 0.81 [95% CI, 0.69- 0.95],P= .0077), PFS (median 2.6 months [95% CI, 2.0-2.9] vs 1.9 months [1.9-2.1]; HR, 0.81 [95% CI, 0.69-0.96];P=.0103), and disease control rate (51% vs 40%; P = .0020) compared with those receiving erlotinib. ORR did not differ between groups (6% vs 3.0%, P = .0551). Adverse event profiles were similar between groups; grade 3 diarrhea and stomatitis were more common with afatinib (10% vs 2% and 4% vs none, respectively), and grade 3 rash or acne occurred more frequently in the erlotinib arm (10% vs 6% with afatinib).¹¹⁴

The EGFR mAb cetuximab has previously shown small improvements in OS benefit (HR, 0.871; 95% CI, 0.762-0.996;P= .044) when added to first-line cisplatin and vinorelbine inEGFR-expressing NSCLC; greatest benefit was observed among patients with SCC (HR, 0.80 [95% CI, 0.64-1.00]).¹¹⁸Because of the high frequency of febrile neutropenia, cetuximab-containing regimens are not recommended in NSCLC 3 Recent positive outcomes from the SQUIRE trial evaluating the second-generation EGFR mAb necitumumab indicate that its addition to first-line gemcitabine/cisplatin improves survival. Patients randomized to the experimental arm (n = 545) had improved OS compared with patients receiving chemotherapy alone (median, 11.5 vs 9.9 months; HR, 0.84; 95% CI, 0.74-0.96;P= .01).³⁶The addition of necitumumab was associated with a higher rate of AEs and serious AEs (72% vs 62% and 48% vs 38%, respectively), including grade 3 to 4 hypomagnesaemia (9 % vs 1%) and grade 3 rash (4% vs one <1%).

The phase II BASALT-1 study (NCT01820325) evaluated the clinical activity of the pan-PI3K inhibitor buparlisib (BKM120) in patients with PI3K pathway-activated, relapsed NSCLC. Pathway activation was defined asPIK3CAmutation,PTENmutation, or PTEN negative (<10% protein expression by IHC). The study did not meet its primary objective; treatment of selected patients (63 of 1242 prescreened; 13.5%) resulted in 12-week PFS rates of 23.3% (95% CI, 9.9-42.3) and 20.0% (95% CI, 7.7-38.6) in patients with squamous and nonsquamous tumors, respectively.³⁴These clinical outcomes were considered comparable with those involving chemotherapy in unselected patients; stage II was therefore not initiated in either group. Selection of patients was primarily based on archival primary tumors; and the authors detected discordance of mutation patterns with those detected in circulating tumor DNA samples collected during the trial, which were more closely related to metastatic tumors. These findings suggest that circulating tumor DNA may be a more accurate sample for the molecular stratification of recurrent and metastatic SCC.³⁴

Preliminary results from early-phase studies have reported on the clinical activity of pan-FGFR TKIs AZD4547, BGJ398, and JNJ- 42756493 in lung cancer.^119,120A multicenter phase I study evaluating AZD4547, selective FGFR1-3 inhibitor in patients with previously treated stage IVFGFR1-amplified SCC (NCT00979134), reported a manageable safety profile, but only 1 of 15 patients had a partial response such that the prespecified ORR efficacy endpoint was not met. Among patients withFGFR1-amplified advanced or metastatic platinum-resistant SCC, who received BGJ398 in a phase I study, partial response was obtained by 4 of 21 patients. Responses of NSCLC patients to JNJ-42756493 were limited to stable disease. A phase II study of dasatinib, an inhibitor of Src family kinases, found an overall disease control rate of 43% among patients with advanced NSCLC (n = 34), including one patient with a partial response.120 The trial included six patients with SCC; however, the presence of potentially sensitizing mutations in DDR was not evaluated.

Click here to view references:Evolving Paradigms in Squamous NSCLC> >