Abstract
The treatment of advanced or refractory non–small cell lung cancer has been historically difficult owing to the lack of studies on effective systemic cure. The progress in lung cancer treatment has plateaued, necessitating new options for additional benefits. Immune checkpoint proteins are co-inhibitory factors that can diminish the antigen-specific immune responses by attenuating the regulatory role of cytotoxic T-lymphocyte-associated protein 4, programmed cell death-1, lymphocyte-activation gene 3, and T-cell immunoglobulin mucin-3. The therapeutic strategies targeting immune checkpoints mainly focus on the monoclonal antibody of these regulatory factors, which may facilitate clinical decision making. An enhanced understanding of the drug-resistance mechanisms and the therapeutic efficacy regulation will provide opportunities to improve the clinical outcomes of non–small cell lung cancer patients. Preclinical and clinical trials on these key immune-regulatory agents, which has heralded a new era in immuno-oncology in non–small cell lung cancer treatment, are currently in development.
Keywords
Introduction
Lung cancer is a major global health problem, with nearly 1.82 million new cases and 1.6 million deaths annually. 1 Almost 85% of the cases are of non–small cell lung cancer (NSCLC), consisting of nonsquamous (70%) and squamous (30%) histological subtypes.2,3 Refractory NSCLC patients suffer very disappointing prognoses, with a 5-year survival rate in 11%–15% of cases. 4 Although systemic surgeries and chemotherapy have provided modest therapeutic gains, patients with advanced or refractory NSCLC often progress after conventional strategies, with less treatment options afterward. Thus, discovering novel immunotherapies is urgent to provide the host immune system with better clinical outcomes.
The immune checkpoint inhibitors targeting cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), programmed cell death-1(PD-1), and PD-1 ligand (PD-L1) are widely applied in clinical cancer treatments, including melanoma, breast cancer, ovarian cancer, and NSCLC. CTLA-4 and PD-1 are usually expressed on the surface of effector T cells. The ligands for PD-1 have known to be PD-L1 and PD-L2, with more attention for PD-L1. Immunotherapies that block these factors (CTLA-4, PD-1, and PD-L1) can restore and augment cytotoxic T-cell responses, leading to potentially resilient responses and prolonged overall survival (OS) with tolerable toxicity. Some other immune checkpoint molecules, including lymphocyte-activation gene 3 (LAG-3), T-cell immunoglobulin mucin-3 (TIM-3), and V-domain immunoglobulin suppressor of T-cell activation (VISTA), have been characterized as well. 5 Numerous clinical and preclinical trials have obtained significant success in blockading these factors, thus serving as a potential component of clinical efficacy for patients with advanced or refractory NSCLC.
This study summarizes the immune checkpoint signaling in NSCLC and the current efforts to predicting the clinical prognosis of these therapies. Our objective is to highlight the similarities and to distinguish among key immune-regulatory agents, such as ipilimumab, tremelimumab, nivolumab, pembrolizumab, and other anti-PD-L1 antibodies (Abs), that might facilitate therapeutic decision making. Accumulating evidence imply the clinical benefits in the treatment of some malignancies, including advanced NSCLC. In addition, the resistance mechanisms and factors that affect the clinical responses can possibly improve the efficacy of these therapies. Thus, an enhanced understanding of their functions may shed light on the therapeutic strategies against NSCLC.
CTLA-4-blocking inhibitors and clinical relevance
CTLA-4 is mainly expressed on activated CD8+ effector T cells, and negates the early-stage T-cell activation and cell cycle progression. CTLA-4 is homologous to CD28, and it possesses the same ligands, including B7-1 (CD80) and B7-2 (CD86), thus blocking the activation and proliferation of antigen-activated T cells secondary to CD28/B7 interaction. Monoclonal Abs-blocking CTLA-4 is regarded as a comprehensive therapy for NSCLC. Numerous studies have evaluated the associations between CTLA-4 inhibitors and cancer development. The inhibition of CTLA-4 signaling is extremely promising to recover the suppression of T-cell activation, proliferation, and infiltration into tumors. This process can result in enhanced anti-tumor immunity and tumor cell eradication.6,7 Herein, we emphasized the clinical improvements in therapies of ipilimumab and, another CTLA-4 inhibitor, tremelimumab, which are under extensive investigations for NSCLC treatment.
Ipilimumab
Ipilimumab (BMS-734016), a fully humanized immunoglobulin (Ig)G1 monoclonal Ab directly against CTLA-4, is the first immune checkpoint inhibitor tested in clinical trials. Clinical trials have shown that ipilimumab can positively function in treating metastatic melanoma (MM), 8 renal cell carcinoma, 9 and prostate cancer. 10 However, numerous studies have focused on the efficacy of ipilimumab on MM patients, whereas NSCLC has not been historically considered to be responsive to it. However, the current combination therapies with chemotherapy have drawn great attention. Ipilimumab plus chemotherapeutic agents can induce anti-tumor activity and therapeutic synergy based on distinct and complementary mechanisms of action for drugs and unique cellular targets. 11 Hence, in a phase II study, chemotherapy-naive stage IIIB/IV NSCLC patients were administered with 10 mg/kg ipilimumab every 3 weeks (q3w), concurrently with or sequentially to carboplatin and paclitaxel, or given chemotherapy alone as the control group (NCT00527735). 12 The study showed improved immune-related progression-free survival (PFS) for phased ipilimumab versus the control group (hazard ratio (HR) = 0.72; p = 0.05), but not for concurrent ipilimumab (HR = 0.81; p = 0.13). Phased ipilimumab also improved PFS according to the modified World Health Organization (WHO) criteria (HR = 0.69; p = 0.02). Similarly, the recommended dose of ipilimumab in a combination regimen with paclitaxel/carboplatin among Japanese patients (NCT01165216) is 10 mg/kg with similar adverse events (AEs) compared with the data from the former trials. A phase III trial comparing ipilimumab plus paclitaxel and carboplatin with placebo plus paclitaxel and carboplatin (NCT02279732) 13 is ongoing to further investigate the clinical benefits of the combination therapy. On the basis of the therapy described earlier, a phase III trial termed NCT01285609 is ongoing, which administrates phased ipilimumab with chemotherapy in first line for patients with squamous NSCLC. Clinical trials are also evaluating ipilimumab in combination with ionizing radiation. 14 Efficacy improvements and clinical AEs of the combined therapy should be further investigated and elucidated. The ongoing trials of ipilimumab monotherapy or combination therapies are summarized in Table 1.
Ongoing clinical trials on main ICIs involving monotherapy or in combination with chemotherapy, radiotherapy, and targeted therapy.
ICI: immune checkpoint inhibitor; CTLA-4: cytotoxic T-lymphocyte-associated protein 4; PD-1: programmed cell death-1; EGFR: epidermal growth factor receptor; PD-L1: programmed cell death-ligand 1; BSC: best supportive care.
Tremelimumab
Tremelimumab (CP-675,206), a fully humanized IgG2 anti-CTLA-4 Ab, is similar to ipilimumab in clinical activity. 15 Tremelimumab 16 tends to enhance the proliferative response of effector T cells upon stimulation and to abrogate the immunosuppressive functions of regulatory T cells (Tregs). In addition, the frequencies of interleukin (IL)-2-secreting CD4+ and interferon (IFN)γ-secreting CD4+ and CD8+ T cells increased with tremelimumab treatment. Thus, tremelimumab may be vital in the tumor environment and beneficial in the anti-tumor response. However, limited information on the long-term outcomes of NSCLC patients in tremelimumab treatment is available, such that it requires more evidence to support its clinical efficacy. A phase II study 17 has been conducted to compare the efficacy of tremelimumab with best supportive care (BSC) among stage IIIB/IV NSCLC patients with platinum-based first-line therapy (NCT00312975). A total of 87 patients are randomized to receive either tremelimumab or BSC. However, PFS at 3 months was similar in each arm, with 20% and 14.3% in tremelimumab and BSC arms, respectively. As for the safety parts, the overall incidence of grade 3/4 was 20.5% and 0% in the tremelimumab and BSC groups, respectively. 18 No significant benefits were observed collectively in the tremelimumab group, although the incidence of AEs increased. With regard to the limited number of patients and few related studies, the combined tremelimumab treatment and the AEs necessitate further study. Clinical trials are also evaluating tremelimumab in combination with durvalumab (a type of Abs-targeting PD-L1) in NSCLC patients (NCT02000947). It implied that durvalumab 20 mg/kg every 4 weeks (q4w) plus tremelimumab 1 mg/kg exerted anti-tumor activity with a manageable tolerability profile, as two of six patients (33%) indicate dose-limiting toxicities. More ongoing trials of tremelimumab are summarized in Tables 1 and 2.
Ongoing clinical trials involving ICIs with other ICIs.
ICIs: immune checkpoint inhibitors; CTLA-4: cytotoxic T-lymphocyte-associated protein 4; PD-1: programmed cell death-1; Abs: antibodies; NSCLC: non–small cell lung cancer; NR: not reported; PT-DC: platinum-based doublet chemotherapy; PD-L1: programmed cell death-ligand 1; q3w: every 3 weeks; SoC: standard of care (erlotinib, gemcitabine, or vinorelbine); EGFR: epidermal growth factor receptor; TK: tyrosine kinase; ALK: anaplastic lymphoma kinase.
In general, the clinical efficacy of anti-CTLA-4 inhibitors in NSCLC treatment, either delivered as monotherapy or combination therapies, has not been thoroughly elucidated. Considering the limited number of patients involved, more clinical trials are needed to further explore a promising prospect of clinical application of CTLA-4-blocking inhibitors.
PD-1/PD-L1-blocking inhibitors and clinical relevance
PD-1 is more broadly expressed than CTLA-4 on activated T cells, which inhibits cell immune response upon interaction with its ligands: PD-L1 (B7-H1, CD274) or PD-L2 (B7-DC, CD273). PD-1 primarily inhibits T-cell activity in the effector phase within tumor tissues, whereas CTLA-4 regulates immune responses early in T-cell activation. 19 As for its ligands, PD-L1 tends to be correlated with the poor prognosis of patients with some malignancies, including esophageal and gastric cancer, 20 renal cell carcinoma, 21 hepatocellular carcinoma, 22 melanoma, 23 and, most importantly, NSCLC. 24 Human tumors have also been documented to express PD-L2, whereas its roles in tumorigenesis may share some differences with PD-L1 that are not very coherent.25,26
Clinical studies on NSCLC show that the promising response rate (RR) is negatively correlated with PD-L1 expression on tumor-infiltrating immune cells (TILs) or tumor cells. Furthermore, anti-PD-1/PD-L1 therapy may not be confined to NSCLC cell types, given that no difference in the pooled objective response rates (ORRs) was observed between nonsquamous and squamous cell lung cancers in a meta-analysis. 27 In this article, we evaluated the clinical applications of the molecules targeting the PD-1 and PD-L1 pathways in advanced or metastatic NSCLC. A better understanding of these factors might provide potential options for better clinical outcomes.
Anti-PD-1 Abs
Nivolumab and pembrolizumab are the first two members of the anti-PD-1 pathway family to gain approval from the US Food and Drug Administration (FDA) for the treatment of ipilimumab-refractory melanoma. For NSCLC treatment, the first approved inhibitor is nivolumab, based on a survival advantage over docetaxel in recurrent squamous NSCLC. 28
Nivolumab
Nivolumab (BMS936558) is a fully humanized IgG4-blocking Ab against PD-1 checkpoint protein. Nivolumab binds PD-1 on to activated immunocytes to disrupt their interactions with PD-L1 and PD-L2, thereby activating host immune response. 6 The FDA granted nivolumab traditional approval on 4 March 2015, for treating metastatic squamous NSCLC with progression during or after platinum-based chemotherapy. The approval may provide an important treatment option for patients, which affects routine care and clinical trials. 29
The first study to investigate the efficacy of nivolumab in humans was conducted by Brahmer et al. 30 In the phase I study, 6 out of the 39 participants with treatment-refractory NSCLC were infused with nivolumab at doses of 0.3, 1, 3, and 10 mg/kg. This is the first in-human trial, which indicated the high affinity of nivolumab for PD-1 and its ability to produce a robust response. Although the clinical efficacy was not a primary end point of this trial, they found evidence of anti-tumor activity, with significant lesional or mixed regressions in one additional patient with NSCLC (1 mg/kg). Nivolumab was well-tolerated, with no dose-limiting toxicity. Several clinical trials have been designed to demonstrate the clinical efficacy and good tolerance of nivolumab therapy in patients with advanced or refractory NSCLC. To observe the clinical benefits of nivolumab monotherapy, 31 129 patients with heavily pretreated advanced NSCLC randomly received three doses of nivolumab (1, 3, or 10 mg/kg, every 2 weeks (q2w) in an 8-week cycles for upto 96 weeks). Except for 6 patients who suffered from unconventional immune-pattern responses, the ORRs among 110 patients were 3% (1 mg/kg), 24% (3 mg/kg), and 20% (10 mg/kg), and the OS were 9.2, 14.9, and 8.6 months, respectively. The PD-1 therapy seems to be a clinically relevant target in NSCLC, and demonstrated objective response (OR) to nivolumab monotherapy of 17% (22/129), lasting for a median of 17.0 months. However, the current second-line standard therapies for advanced NSCLC generated ORRs of 7% to 9%, with a median OS of approximately 8 months. 32 These factors demonstrate robust responses and encourage survival rates in nivolumab treatment. A phase II trial, termed NCT01721759, 24 was recently conducted on 117 patients with advanced, refractory stage IIIB/IV squamous NSCLC. Patients who received two or more previous treatments received an intravenous nivolumab (3 mg/kg, q2w) until tumor progression or intolerable toxicity. Currently, the OR was 14.5% (95% CI (confidence interval) = 8.7–22.2) and 1-year OS was 40.8% (95% CI = 31.6–49.7). All results indicated that nivolumab has meaningful activities and manageable safety profile in previously treated NSCLC patients, resulting in new areas of anti-tumor strategies.
Clinical trials are also underway or being planned to evaluate the combinations of nivolumab with chemotherapy drugs, such as docetaxel, and platinum-based therapies. Docetaxel is the standard-of-care agent for nonsquamous NSCLC patients who are refractory to platinum-based first-line chemotherapy regimens. The NCT01673867 (CheckMate 057) 33 trial results indicated that nivolumab produced more significant clinical benefits compared with docetaxel. A total of 582 patients were randomized (1:1) to receive nivolumab (3 mg/kg, q2w) as monotherapy or docetaxel. As a result, the median OS of the nivolumab group was longer than that of the docetaxel group, for 12.2 months versus 9.4 months (p = 0.002). Although the median PFS was unfavorable to nivolumab over docetaxel (2.3 and 4.2 months, respectively), the rate of PFS during the first year was higher with the nivolumab group (19% and 8%, respectively). A statistically significant improvement in ORR was also observed, with an ORR of 19% (95% CI = 15%–24%) in the nivolumab arm and 12% (95% CI = 9%–17%) in the docetaxel arm. Based on these results, nivolumab represents a new treatment option for patients requiring second-line treatment for metastatic NSCLC. 34 In another phase III clinical trial (NCT01642004), 33 272 patients with metastatic squamous NSCLC were randomized to receive nivolumab (n = 135) or docetaxel (n = 137). The recommended dose of nivolumab was 3 mg/kg q2w, which was similar to that of previous research. The primary end point was an OS of 9.2 months in the nivolumab group versus 6.0 months in the docetaxel group (p < 0.001). Moreover, nivolumab also reported promising RR and PFS benefits. Based on such encouraging results, the potential therapeutic function of nivolumab holds promise to fully realize the potential efficacy of immunotherapy in NSCLC. The ongoing trials are summarized in Table 1.
A number of ongoing trials evaluating the efficacy of the combined ipilimumab and nivolumab are in progress. Both can exert potent anti-tumor activity and good tolerability in NSCLC treatment as monotherapies. Furthermore, their combination therapy is promising in enhancing efficacy. A phase III trial (NCT02785952) compares nivolumab with or without ipilimumab in treating patients with recurrent squamous NSCLC. Another clinical trial (NCT02477826) is recruiting participants with stage IV NSCLC for evaluating nivolumab, nivolumab plus ipilimumab, or nivolumab plus platinum-based doublet chemotherapy (PT-DC) versus PT-DC. Details of the ongoing trials of combined immune checkpoint inhibitors are summarized in Table 2.
As a result, nivolumab has been proven to be effective in metastatic or refractory NSCLC patients and considered as a more superior therapy compared to conventional chemotherapies. Besides, the application of combined therapies with other types of checkpoint inhibitors may further give rise to improved efficacy in NSCLC patients.
Pembrolizumab
Pembrolizumab (MK-3475) is a highly selective humanized mAb of IgG4-kappa isotype, which directly blocks the interactions between PD-1 and the two ligands PD-L1 and PD-L2. 35 Pembrolizumab 2 mg/kg given q3w was granted accelerated approval in the United States for the treatment of patients with metastatic NSCLC whose tumors express PD-L1 with disease progression during or after platinum-containing chemotherapy. Pembrolizumab is a potentially powerful and innovative strategy in cancer therapy. In a phase I study (NCT01295827), 36 495 previously treated and treatment-naive patients with advanced or metastatic NSCLC were randomized to receive pembrolizumab (at a dose of either 2 or 10 mg/kg, q3w, or 10 mg/kg, q2w) to either a training (n = 182) or a validation group (n = 313). In this trial, the overall ORR was 19.4% (95% CI = 16.0–23.2), whereas the ORR was 24.8% (95% CI = 16.7–34.3) in treatment-naive patients (n = 101) and 18.0% (95% CI = 14.4–22.2) in previously treated patients (n = 394). As for the pembrolizumab-involved group, this trial demonstrated that ORRs at 2 mg/kg (q3w; n = 6) or 10 mg/kg (q3w; n = 287) were 33.3% or 19.2%, respectively, whereas ORR at 10 mg/kg (q2w; n = 202) was 19.3%. Although limited patients were involved in the trial, there was promising clinical outcome with an acceptable side-effect profile for the pembrolizumab treatment in NSCLC patients. The recent NCT01905657 (KEYNOTE-010) trial 37 indicates improved OS with pembrolizumab treatment when compared to docetaxel group. A total of 1034 patients who have progressed after PT-DC were randomly assigned to three groups, received, and treated with a pembrolizumab dose of either 2 (n = 345) or 10 mg/kg (n = 346) or docetaxel of 75 mg/m2 (n = 343). The two pembrolizumab groups (2 and 10 mg/kg) exhibited significantly superior OS versus the docetaxel group (HR = 0.71; p = 0.0008 and HR = 0.61; p < 0.0001). Moreover, pembrolizumab was associated with less high-grade toxicity than docetaxel. Unfortunately, no significant difference in PFS was observed between the pembrolizumab and docetaxel groups. The results described above demonstrate the promising clinical efficacy of pembrolizumab treatment, which may be superior to traditional chemotherapy in NSCLC patients. The lack of a benefit for PFS despite a potent OS benefit was also reported in the NCT01673867 study.
Randomized clinical trials, evaluating the efficacy of pembrolizumab as a single agent or in combination with chemotherapy and radiation therapy, are currently underway with much-awaited results (NCTC02220894, NCT02142738, NCT02504372, and NCT02444741). These trials aim to learn whether combination therapy can deliver clinical improvements and promising clinical prognoses in metastatic NSCLC. Other ongoing trials are summarized in Table 1.
Anti-PD-L1 Abs
Atezolizumab, durvalumab, and avelumab are fully humanized IgG1 Abs, whereas BMS936559 is an IgG4 Ab. 38 These blocking Abs have presented promising benefits in NSCLC patients. Plenty of Abs-targeting PD-L1, including atezolizumab, durvalumab, avelumab, and BMS936559, are currently widely studied and gaining attention as cancer treatments. In a phase I trial (NCT00729664), Brahmer et al. administered anti-PD-L1 Ab intravenously (at escalating doses ranging from 0.3 to 10 mg/kg) to patients with advanced cancers, including NSCLC. 39 It was administered q2w for upto 16 cycles or until the patients had a complete response (CR) or confirmed disease progression. As a result, an OR was observed in 5 out of 49 patients. Ab-mediated blockade of PD-L1 also induced durable tumor regression (ORR of 6% to 17%) and prolonged stabilization of disease (rates of 12% to 41% at 24 weeks) in patients with advanced cancers, including NSCLC. Herein, we demonstrate the concrete clinical trials and studies focusing on the key blocking Abs, heralding a new and smart approach to NSCLC treatment.
Atezolizumab (MPDL3280A), a humanized monoclonal Ab that targets PD-L1, has been evaluated as a new anti-tumor strategy. FDA has approved atezolizumab for the treatment of NSCLC patients with PD-L1+ expression that has progressed during or after standard first-line treatments. This finding was based on results from early studies on the efficacy and safety of atezolizumab. In trial NCT01375842, 53 efficacy-evaluable patients received atezolizumab 0.01, 0.03, and 0.1 mg/kg doses; 40 confirmed responses (CR plus partial response (PR)) were observed in 11 out of 53 (21%) patients. The responses were rapid and robust. In a phase II randomized controlled trial (NCT01903993), Fehrenbacher et al. 41 assessed the efficacy of atezolizumab versus docetaxel on patients who progressed on post-platinum chemotherapy. A total of 142 patients received atezolizumab (1200 mg fixed dose, q3w) and 135 received docetaxel. Compared with docetaxel, atezolizumab significantly improved patient survival, with an OS of 12.6 months (95% CI = 9.7–16.4) for atezolizumab versus 9.7 months (8.6–12.0) for docetaxel (HR = 0.73; 95% CI = 0.53–0.99; p = 0.04); a safety profile distinct from chemotherapy was also shown. However, there was no difference between the two groups in PFS (2.7 months with atezolizumab vs 3.0 months with docetaxel; HR = 0.94; 95% CI = 0.72–1.23). A comparative study (NCT02409342) of atezolizumab with cisplatin or carboplatin plus pemetrexed is currently recruiting patients. Other studies (NCT01846416 and NCT01375842) focusing on the efficacy, safety, and pharmacokinetics of atezolizumab are currently underway. Detailed descriptions are provided in Table 1. Herein, we believe that more trials will support the improved clinical benefits of atezolizumab treatment for NSCLC.
Durvalumab (MEDI4736) is an engineered humanized IgG1 Ab that blocks PD-L1 from binding to PD-1, thereby allowing T cells to recognize and kill malignant cells; 42 it is currently being evaluated in clinical trials as a monotherapy and in combination with other agents for first or subsequent lines of NSCLC treatment. In an ongoing phase 43 I/II study (NCT02352948), the section of substudy A aimed to assess the safety and clinical activity of durvalumab (10 mg/kg, q2w) versus SoC in stage IIIB/IV patients with PD-L1+ tumors (≥25% of tumor cells with membrane staining). This study is expected to confirm the anti-tumor activity of durvalumab with better clinical efficacy compared with SoC treatment, with a manageable tolerability profile. Furthermore, in trial NCT02000947, 44 the safety and tolerability of durvalumab in combination with tremelimumab are evaluated in patients with advanced solid tumors including NSCLC. AEs occurred in 34% of all patients, with a remarkably limited toxicity of grades 1 to 2. It is noteworthy that durvalumab proved to be clinically effective by inducing four partial remissions and five additional minor responses. These occurred not only in melanoma but even more in NSCLC patients. Besides, the evidence of clinical activity was noted both in patients with PD-L1+ tumors and PD-L1− tumors. Their data provided a promising therapeutic option for patients with PD-L1− tumors, which are not expected to obtain a benefit from anti-PD-1/PD-L1 therapies. Another phase II trial 45 is currently designed to deliver durvalumab in addition to neoadjuvant chemotherapy in patients with stage IIIA NSCLC. Thus, it further adds great interest in the utilization of durvalumab in treating NSCLC. Other ongoing trials on durvalumab participation are shown in Tables 1 and 2.
Avelumab (MSB0010718C) is another IgG1 anti-PD-L1 monoclonal Ab, with the ability to lyse human tumor cells via the antibody-dependent cell-mediated cytotoxicity (ADCC) mechanism. 46 PD-L1 expressed on tumor cell surfaces determines ADCC sensitivity mediated by avelumab, which increases its potential functions in anti-tumor immunity. Several studies are expected to demonstrate the superiority of avelumab in terms of the clinical efficacy and safety versus chemo-based doublet on PD-L1+ tumors (NCT02576574 and NCT02395172). Other clinical trials focusing on the efficacy of its monotherapy or in combination with targeted therapies are demonstrated in Table 1.
BMS936559 (MDX-1105) is a high-affinity IgG4 Ab that inhibits the binding of PD-L1 to both PD-1 and CD80. In a multicenter phase I trial, named NCT00729664, 39 75 patients with advanced NSCLC were administered with BMS936559 (at doses ranging from 1, 3, to 10 mg/kg). Among those with evaluable responses, an OR (CR or PR) was observed in 5 out of 49 (10.2%) patients. The blockade of PD-L1 induced strong tumor regression (ORR of 6% to 17%) and prolonged the stabilization of disease (rate of 12% to 41% at 24 weeks). These findings validated the pathway between PD-1 and PD-L1 as an important target for therapeutic intervention in some NSCLC patients. Additional studies are needed to identify the patients who are likely to have a better response and to determine an appropriate clinical dose. However, no clinical trials on NSCLC patients are in progress.
The current main clinical trials on anti-PD-1/PD-L1 Abs in NSCLC treatment are demonstrated in Table 3.
Currently main clinical trials on anti-PD-1/PD-L1 antibodies and the description of PD-L1 expression as a potential biomarker.
PD-1: programmed cell death-1; PD-L1: programmed cell death-ligand 1; ORR: objective response rate; OS: overall survival; PFS: progression-free survival; NSCLC: non–small cell lung cancer; q2w: every 2 weeks; OR: objective response; PT-DC: platinum-based doublet chemotherapy; NR: not reported; q3w: every 3 weeks; TIL: Tumor-infiltrating immune cells.
As described above, the studies of anti-PD-L1 Abs focused on the clinical efficacy, safety, and tolerability and the optimized dosage for clinical application. As these are confined research trials, we cannot currently evaluate the clinical benefits of PD-L1 blocking therapies.
Predictive markers of checkpoint inhibitor response
Despite the promising application prospects of checkpoint inhibitors in NSCLC patients, only a subset of patients derives clinical benefits (ORRs = 20%), underscoring the need for more specific predictive biomarkers. Therefore, biomarkers that identify patients with favorable clinical responses to therapies are essential to guide treatment decisions and to improve patient outcomes. This study lists several well-known factors that may predict the clinical activities of checkpoint inhibitors in NSCLC.
Predictive markers of anti-CTLA-4 therapy
Inducible costimulator
Inducible costimulator (ICOS) molecule is mainly expressed on activated T cells, working as a biomarker for therapeutic benefits in anti-CTLA-4 therapy for melanoma and bladder cancer.47,48 In particular, the frequency of ICOS+ CD4+ T cells in the ipilimumab treatment shows a statistically significant increase. 48 The increased frequency is positively correlated with improved clinical outcomes and the absence of ICOS or its ligand leads to impaired tumor rejection. An increased frequency of ICOS+ CD4+ T cells in blood samples can work as a pharmacodynamic biomarker, which promisingly can be used for making decisions regarding different doses, schedules, and combinations tested in clinical trials with anti-CTLA-4. In CT26-HER-2 lung metastasis mouse model, 49 the results revealed that treatment with CTLA-4 blockade augmented the number of ICOS+ CD4+ T cells but not ICOS+ CD8+ T cells in the tumor, blood, and spleen. It further indicated that higher expression of ICOS on effector cell populations may be correlated with improved survival of combining poxvirus immunotherapy with CTLA-4 inhibitors.
However, few studies have focused on the relationship between NSCLC treatment and ICOS expression. The polymorphisms in ICOS genes (ICOSc.1554+4GT{8_15}{>10}) were previously associated with susceptibility to disease and OS rates of NSCLC in a Polish population 50 but not as an independent risk factor. Therefore, this merits more studies to define the therapeutic ability for ICOS in NSCLC patients with anti-CTLA-4 therapy.
Myeloid-derived suppressor cells
CTLA-4 is not exclusively expressed on T cells; anti-CTLA-4 monoclonal Ab may directly target additional cell populations. Several studies have focused on myeloid-derived suppressor cell (MDSC) frequency changes after the CTLA-4 inhibition. In patients with MM, ipilimumab treatment resulted in lower MDSC frequency after the treatment. 51 Furthermore, the Lin−CD14+HLA-DR− monocytic MDSC frequency was reduced in ipilimumab-treated patients, working as a predictive marker of response, as patients with low frequency tend to easily benefit from ipilimumab treatments. 52 Another study reported that circulating CD33+CD11b+HLA-DR− MDSCs from peripheral blood can be used as biomarkers to predict response and survival of stage IV melanoma patients in ipilimumab treatments. 53 Low levels of MDSCs prior to CTLA-4 therapy correlated with an objective clinical response, long-term survival, and improved clinical status. The presence of more than 55.5% of circulating CD33+CD11b+ out of the HLA-DR− cells in 56 patients was associated with significantly short OS (p < 0.003), a median of 6.5 months, in comparison to the group showing lower MDSC frequencies, with a median survival of 15.6 months. These studies represent a prospect that circulating MDSCs can serve as a promising predictive metric for CTLA-4 inhibition.
MDSCs are associated with a poor prognosis; therefore, the effects of CTLA-4 blockade on these cellular populations are likely to provide insights into possible predictive biomarkers among NSCLC patients. We are looking for more detailed studies focused on NSCLC treatment and more specific biomarkers to identify the efficacy of anti-CTLA4 treatments.
Although there are some proofs of ICOS and MDSCs serving as predictive markers in anti-CTLA-4 inhibitor treatments, the specific knowledge and inner mechanisms of the two molecules have not been studied well in NSCLC patients. More researches are needed to deliver a more detailed guideline for CTLA-4 blocking therapies.
Predictive markers of anti-PD-1/PD-L1 therapy
PD-L1
PD-L1 expression on tumor cells appears to be a predictor of the response to PD-1/PD-L1 inhibitors. In NCT01905657 (KEYNOTE-010) trial, Herbst et al. 37 first reported that PD-L1 can serve as a prospective biomarker for pembrolizumab treatment and can identify therapeutic benefits. Consistently, in the NCT01295827 (KEYNOTE-001) trial, 36 it showed that pembrolizumab represented superiority to docetaxel on PFS in patients with more than 50% PD-L1 expression, but not in the total population. Besides, the responses are durable, regardless of PD-L1 expression level. However, in lung adenocarcinoma, patients with stage I to III lung carcinomas 54 harboring PD-L1+ tumor cells show poor recurrence-free survival (p < 0.001) and OS (p < 0.001) on univariate analysis, which reported an inconsistent conclusion.
Several studies have focused on the differences of PD-L1 predictive functions on tumor cells and TILs. In a study 41 assessing clinical efficacy of atezolizumab versus docetaxel on patients who progressed on post-platinum chemotherapy, baseline PD-L1 expression was scored by immunohistochemistry in tumor cells (as percentage of PD-L1-expressing tumor cells: TC3 ≥ 50%, TC2 ≥ 5% and < 50%, TC1 ≥ 1% and < 5%, and TC0 < 1%) and in TILs (as percentage of tumor area: IC3 ≥ 10%, IC2 ≥ 5% and < 10%, IC1 ≥ 1% and < 5%, and IC0 < 1%). OS improvement was significant in subgroups with high PD-L1 expressions (TC3 or IC3: HR = 0.49 (95% CI = 0.22–1.07; p = 0.068); TC2/3 or IC2/3: HR = 0.54 (95% CI = 0.33–0.89; p = 0.014); TC1/2/3 or IC1/2/3: HR = 0.59 (95% CI = 0.40–0.85; p = 0.005), TC0 and IC0 HR = 1.04 (95% CI = 0.62–1.75; p = 0.871)). OS was also improved in atezolizumab-treated patients with high PD-L1 gene expression as well as in patients with high expression of other PD-L1/PD-1 pathway genes in tumor tissues. It indicates that the two sources of PD-L1 expressions may work as a predictive biomarker for anti-PD-L1 Ab treatments. The predictive function of PD-L1 is also evaluated in another type of anti-PD-L1 Ab termed MPDL3280A. 40 In this case, the association of the clinical responses to MPDL3280A treatment and PD-L1 expressions on TILs reached statistical significance (p = 0.015), whereas no such association was achieved in tumor cells (p = 0.920). Furthermore, this trial described that responses are associated with T-helper type 1 gene, and CTLA-4 expressions, as well as reported the absence of fractalkine in baseline tumor specimens. In addition, the differential efficacy of checkpoint inhibitors targeting PD-1 (nivolumab and pembrolizumab) and PD-L1 (MPDL3280A) were evaluated in a sensitivity analysis 55 according to the tumor expressions of PD-L1. ORR is significantly higher in PD-L1+ in comparison to PD-L1− patients (23.2% vs 14.5%, p = 0.0216). However, when the relationship between PD-L1 expression and OS among chemo-treated patients was evaluated, no obvious association was observed in median or tertile PD-L1 expression levels. 56
Several investigations to ascertain whether polymorphisms of genes involved in immune checkpoints may affect anti-tumor immune activity, thereby influencing clinical outcomes of chemotherapy in NSCLC patients, have been conducted. Lee et al. 44 investigated the effects of the genetic polymorphisms of immune checkpoints on patients after first-line chemotherapy. In total, 12 single nucleotide polymorphisms (SNPs) of PD-1, PD-L1, and CTLA-4 genes were selected. As a consequence, two polymorphisms in the PD-L1 gene (rs2297136T > C and rs4143815C > G) were found to be significantly associated with the response and survival of NSCLC patients. This study revealed that the SNPs of PD-L1 may be used as independent predictive biomarkers for NSCLC patients receiving first-line chemotherapy, and thereby helped to identify the subgroup of patients who may benefit from chemotherapy from those who may suffer from unnecessary toxicity. PD-L1 gene copy number can also work as a new reliable biomarker. The increased PD-L1 copy number 57 is associated with PD-L1 expression in NSCLC, which leads to poor prognosis. In addition, there was a high consistency of PD-L1 copy numbers in tumor cells between paired primary and metastatic lesions in contrast to a low agreement of PD-L1 expression, which attracts great attention. In univariate analyses, with regard to post-operative survival, both amplification of the PD-L1 gene and PD-L1 overexpression led to poor prognosis. Therefore, this study concluded that the increased PD-L1 gene copy number can be a feasible alternative biomarker for predicting response to PD-1/PD-L1 inhibitors. Given the studies described above, PD-L1 expression tends to predict the response of the anti-checkpoint therapies, especially concerning the efficacy of immunotherapy combined with chemotherapy, which needs further study.
PD-L1 seems to be an encouraging predictive biomarker for anti-PD-L1/PD-1 therapies in NSCLC, but there are some exceptions which need to be paid attention to. Recently, a phase Ib study 44 aiming to evaluate the anti-tumor activity of durvalumab plus tremelimumab in NSCLC patients noted that the clinical outcomes were irrespective of PD-L1 expression status. In addition, pulmonary pleomorphic carcinoma (PC) patients 58 with a high ratio of PD-1+/CD8+ TILs showed a shorter PFS (p = 0.036), whereas PD-L1 and PD-L2 expression had no prognostic implications. Another study also indicated that intratumoral expression of PD-L2 did not affect the response to anti-PD-L1 treatments, and the expressions of other T-cell negative regulators also failed to be correlated with poor responses. 40 In conclusion, it is not yet clear whether the expression of PD-L1 (or PD-L2) on the membrane of tumor cells and lymphocytes in the tumor microenvironment may be useful in the correct stratification of patients to address specific immunotherapies. The clinical relevance of PD-L1 expression in response to anti-PD-1/PD-L1 has not been well established and the expression on tumor cells or TILs may also affect the predictive function. More researches are needed to explore it further. The main clinical trials on the description of PD-L1 expression as a potential biomarker are shown in Table 3.
Other potential biomarkers
As described above, PD-L1 expression may emerge as the most promising candidate for predicting the response to anti-PD-1/PD-L1 inhibitors. However, the lack of PD-L1 expression cannot exclude patients who were highly responsive to the immunotherapeutic approach from the group. Other components of tumor environment are also being investigated.
Transforming growth factor-β
The transforming growth factor (TGF)-β signaling pathway is vital in regulating cell growth, differentiation, and apoptosis, as well as in development and tumorigenesis of cancers, including NSCLC. 40 The TGF-β plasma level is increased in NSCLC patients compared to that in healthy volunteers. 59 TGF-β is involved in epithelial–mesenchymal transition (EMT) induction, 60 and EMT is proven to be associated with higher expression of PD-L1 on tumor cells, suggesting that immunotherapies targeting immune checkpoints may benefit these patient subgroups with mesenchymal characteristics. 61 This report may emphasize the potential functions of TGF-β as a biomarker in anti-PD-1/PD-L1 therapies.
IFN-γ
Currently, high tumoral IFN-γ messenger RNA (mRNA), PD-L1 protein, and combined IFN-γ mRNA with PD-L1 protein expression are reportedly associated with the response to durvalumab monotherapy in NSCLC patients. 62 In an exploratory analysis, 41 atezolizumab improved the OS of patients with tumors characterized by high expression of T-effector and IFN-γ-associated genes (HR = 0.43; 95% CI = 0.24–0.77). The two gene signatures were also associated with PD-L1 expression on TILs. However, the potentially predictive functions of IFN-γ and its specific connections with PD-L1 expression require more research.
Smoking status
History of smoking appears to affect the probability of response to anti-PD-1/PD-L1 therapies. Several studies have detected that the former and current smokers respond better to nivolumab, 63 pembrolizumab, 64 and MPDL3280A.25,40 Consistent with these findings, an analysis 65 identified that the ORR among nonsmokers or light smokers was 4.2% versus 20.6% among heavy smokers (p = 0.123) within PD-1/PD-L1 inhibitors treatment. Although there was no significant difference, improved survival was generally observed among heavy smokers. Among subgroups of epidermal growth factor receptor (EGFR)-mutant or anaplastic lymphoma kinase (ALK)+ patients, one unconfirmed PR was seen among nonsmokers/light smokers (ORR 4.5%), and no responses were seen among six heavy smokers. Within the EGFR and ALK− group, seven PRs were seen among heavy smokers (25%). Only two nonsmokers/light smokers were in the negative cohort, neither of whom responded to PD-1/PD-L1 inhibitors. These findings showed that whether heavy smokers benefit from anti-PD-1/PD-L1 inhibitors or not may be in connection with EGFR/ALK status. Given the limited sample number, we concur that more assessments are needed. In addition, smoking status may relate with well-known predictive markers such as PD-L1 expression. Calles et al. 25 found that higher PD-L1 expression intensity was more frequent in smokers, which is associated with increased pack years in MPDL3280A treatment. Cha et al. 54 also confirmed that in lung adenocarcinoma, higher PD-L1 expression was more prevalent among former or current smokers (p = 0.026), which was associated with more pack years (p = 0.016). These studies described the potential interactions between PD-L1 expression and smoking status. Under these circumstances, the predictive function of smoking status needs further study to be clearly identified.
EGFR mutations or ALK rearrangement
A retrospective analysis 65 has shown that EGFR mutations and ALK rearrangements result in low RRs to PD-1 pathway blockade in NSCLC. The ORs were observed in 1/28 (3.6%) EGFR-mutant or ALK+ patients versus 7/30 (23.3%) EGFR wild type and ALK−/unknown patients (p = 0.053). Other researchers have also focused on the relationship between PD-L1 expression and EGFR status. Ji et al. 66 have detected that PD-L1 expression is negatively correlated with EGFR status (p = 0.012) in 100 surgically resected lung adenocarcinoma patients, with a higher mutation rate in patients with lower PD-L1 expressions. However, D’Incecco et al. delivered that PD-L1+ was strongly associated with the presence of EGFR mutations (p = 0.001), potentially modulating sensitivity to anti-EGFR agents. 67 Given that PD-L1 expression appears to be a promising biomarker for anti-PD-1/PD-L1 treatments, there may be potential associations between EGFR status and PD-1/PD-L1 inhibition response. The upregulation of PD-L1 may be due to the activated PI3K-AKT and MEK-ERK signaling pathways in NSCLC by EGFR mutations, which revealed a direct link between oncogenic drivers and PD-L1 expression. 68 The surface expression of PD-L1 in cell lines with EGFR mutations is higher than that in most lines that are wild type for EGFR and ALK, with average mean fluorescence intensity values of 265 versus 48. Given the current results, the efficacy of EGFR mutations or ALK rearrangement as biomarkers for these therapies is uncertain and needs further study.
Current clinical trials mandate these potential biomarkers of the response to checkpoint inhibitors. However, further studies are needed to identify their complex predictive values and thereafter guide the selection of NSCLC patients.
Anti-checkpoint therapy: the underlying drug-resistance mechanisms and the regulation of the therapeutic efficacy in NSCLC
Anti-CTLA-4 therapy
As described above, CTLA-4-blocking Abs achieved durable responses in advanced or refractory NSCLC patients, although therapeutic benefits have been limited to a fraction of patients. This calls for elucidation of specific resistance mechanisms of these therapeutic strategies and the necessity to select specific factors that affect clinical response. Improving the clinical response to anti-checkpoint therapy is a central theme in the field of cancer immunology and immunotherapy. This study selected several widely studied factors and presented their potential resistance mechanisms for either the anti-CTLA-4 or the anti-PD-1/PD-L1 therapies of NSCLC.
Indoleamine 2,3-dioxygenase
Indoleamine 2,3-dioxygenase (IDO) is a critical resistance mechanism in anti-tumor T-cell immunotherapy targeting CTLA-4. In IDO-knockout mice 69 treated with anti-CTLA-4 Abs, there was a striking delay in the growth of B16 melanoma tumors and increased OS compared with wild-type mice. This phenomenon is also observed with Abs-targeting PD-1/PD-L1. The study emphasized that IDO strongly contributes to the resistance of anti-CTLA-4 therapies. The combination of anti-CTLA-4 Abs and IDO inhibitor (1MT) resulted in the rejection of tumors and resistance to secondary challenge in B16 melanoma mice. Moreover, another study 70 utilizing murine melanoma and colon cancer cell lines showed that targeting the receptor tyrosine kinase colony stimulating factor-1 receptor (CSF-1R) can sensitize the tumors to immunotherapy with CTLA-4, PD-1, or IDO blockades. Maturation and tumor recruitment of MDSCs are dependent on signaling through CSF-1R. MDSCs are the critical cell population in IDO-expressing B16 tumors in mediating accelerated tumor outgrowth and resistance to immunotherapy. The use of inhibitors of CSF-1R sensitizes IDO-expressing tumors to immunotherapy with T-cell checkpoint blockade, and the combination of CSF-1R blockade with IDO inhibitors both potently elicited tumor regression. These findings validated the potential roles of IDO in immunoresistance. Recently, IDO enzymatic activity 71 was reported to be associated with IFN-γ-induced apoptosis and death receptor 5 (DR5) expressions in NSCLC cells, which suggested that IDO may affect the anti-NSCLC cancer effects of IFN-γ. Thus, we see IDO expression as a possible source of resistance to anti-CTLA-4 therapy in NSCLC, with pending additional research.
Microbiota
A number of studies have suggested that manipulating the microbiota may modulate cancer immunotherapy. Within the tumor microenvironment, microbiota composition was shown to enhance the efficacy of anti-CTLA-4 and PD-L1 therapies by upregulating T-cell related responses. The oral administration of
Anti-PD-1/PD-L1 therapy
Immunosenescence
Immunosenescence is a phenomenon that emphasizes the decreased immune functions resulting in age-associated alterations to the immune system. The process involved is mainly due to the diminished T cell-mediated immunity. A meta-analysis 74 evaluated the efficacy of checkpoint-blocking therapies both on younger and older patients with an age range of 65–70 years, including two trials in NSCLC. Although the OS benefit is significant in both groups, an exception is the group of older patients aged 75 years (HR = 0.86; 95% CI = 0.41–1.83), in which an OS benefit was not observed when treated with anti-PD-1 monoclonal Abs. However, a consistent survival benefit was observed in patients aged 65 years and 65 to 75 years. Thus, immunosenescence may negatively impact the efficacy of anti-PD1 monoclonal Abs. Given that the immune checkpoint inhibitors exert their anti-tumor effects through effector T cells, there might be potential interactions of immunosenescence and drug effect. As a result, more elderly patients should be recruited for the ongoing and future trials of these agents.
Tregs
The infiltration of Foxp3+ Tregs into local tumor sites has been widely demonstrated in human lung cancer, which predicts worse clinical outcomes. A meta-analysis 75 demonstrated that high Foxp3+ Treg infiltration was significantly associated with poor OS in NSCLC (HR = 3.88; 95% CI = 2.45–5.40; p = 0.000). Therefore, we speculated that the level of Tregs in peripheral blood or tumor sites may result in the resistance to anti-PD-1 therapies. A recent study considering this issue was conducted in other mice tumor models instead of NSCLC. The induction of intratumor Tregs 76 was shown to be partly responsible for the development of anti-PD1-resistant tumors. The comparison of anti-PD1-sensitive and anti-PD1-resistant tumors has suggested that intratumoral Tregs might be responsible for limiting anti-PD-1 Ab efficacy in cases where intratumoral T-cell numbers appear sufficient. In NSCLC treatments, the TNFR2+ Tregs 77 present in the peripheral blood of lung cancer patients were more proliferative and expressed higher levels of the immunosuppressive molecule CTLA-4 compared to healthy donors; this can directly suppress the function of immune cells. All these results showed the potential connections between Treg accumulations and poor anti-PD-1 therapy outcomes in NSCLC, which still merits further investigations.
Phosphatase and tensin homolog
Phosphatase and tensin homolog (PTEN) can exert important functions in the regulation of cell proliferation, adhesion and invasion, apoptosis, and DNA damage repair. The loss of PTEN is seen in many types of tumors including NSCLC. A meta-analysis indicated that lung cancer patients with reduced PTEN expression exhibited shorter OS and PFS. 78 However, the PTEN function in promoting checkpoint inhibitor resistance has been mainly observed in melanoma. In melanoma patients, 79 PTEN loss was correlated with decreased T-cell infiltration at tumor sites and inferior PD-1 inhibitor therapy outcomes. The treatment with a selective PI3Kβ inhibitor improved the efficacy of both anti-PD-1 and anti-CTLA-4 Abs in murine models. These findings demonstrated that PTEN loss can promote immune resistance and supported the rationale to explore the combinations of immunotherapies and PI3K-AKT pathway inhibitors. As PTEN loss and PI3K-AKT pathway activation occur in multiple tumor types including NSCLC, the results confirmed that the rationale for PTEN participating in the resistance of anti-checkpoint therapies need further observations.
Conclusion
Immune checkpoint inhibitor agents have currently achieved scientific and promising prospect for monotherapy or combination therapy treatments in several solid tumors, including NSCLC, to date. However, whether they will be feasible to achieve more efficacious clinical benefits and whether immune-related AEs will become more problematic remain unknown. Moreover, the predictive markers of the anti-CTLA-4 and anti-PD-1/PD-L1 responses are still unclear among NSCLC patients. Owing to the complexity of the anti-tumor immune response, the predictive value of factors related to cancer cells or tumor microenvironment and the resistance mechanisms of the therapies require further studies. Further investigations are necessary to determine whether these co-inhibitors can be used in combination with each other or with conventional treatments and whether the outcomes would be synergistic. Given the numerous questions that need to be addressed, an improved understanding of the pathogenic mechanisms and clinical benefits of targeting immune checkpoints will help develop novel and efficient approaches to tumor activity monitoring and therapy.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
This work was supported by grants from the National Key Technology R&D Program (No. 2015BAI12B12) and the Natural Science Foundation of China (No. 81472471 and No. 81272221).