Immune Checkpoint Blockade: The Hope for Immunotherapy as a Treatment of Lung Cancer?

Abstract

Over the past 20 years, immunotherapy has not played a role in the treatment of lung cancer outside of clinical trials. Early trials with vaccines yielded promising results, but phase III trials have yet to show an improvement in survival. Recently, immune checkpoint pathway inhibitors have yielded exciting and consistent activity across this class of antibodies. However, phase III trials are now ongoing. Currently, the hope of bringing immunotherapy to lung cancer patients lies in this class of drugs. Only time will show us if these antibodies will yield an improvement in long-term survival. This review will focus on checkpoint pathway inhibitors that have completed early-phase trials.

Immunotherapy has yet to find its place in the treatment of lung cancer. Vaccines once thought to be the mainstay of immunotherapy have yet to yield an improvement in long-term survival in lung cancer. Even as recently as 2013, a vaccine, lip-osomal-BLP25, used in the ideal setting in low-volume disease after definitive therapy did not show a survival advantage compared to placebo.¹ Other vaccines remain in phase III trials and their results are awaited. The largest trial conducted in the adjuvant setting in non-small cell lung cancer (NSCLC) using the MAGE-A3 vaccine is one such trial. Thus NSCLC continues to be considered a non-immunogenic tumor by many.

NSCLC is able to thwart the immune system through many mechanisms. One such mechanism is through aberrant major histocompatibility complex (MHC) class I expression. MHC class I molecules are required for antigen presentation to cytotoxic T cells. Without MHC class I antigens, tumors are able to escape cell lysis by these T cells.² Aberrant MHC class I expression can occur via deficiency or lack of expression of MHC molecules.^3,4 Another way that NSCLC can thwart the immune system is by adapting immune inhibitory pathways called immune checkpoints. Some checkpoints are costimulatory. These costimulatory pathways are required for T-cell activation such as CD 28 and its ligands B7.1 (CD80) and B7.2 (CD86).⁵ Other checkpoints inhibit T-cell activation such as cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) and programmed death 1 (PD-1) immune checkpoints.

CTLA-4 is a checkpoint pathway that is important early on in T-cell activation.⁵ Through upregulation of CTLA-4, it is able to out compete for its ligands (B7.1 and B7.2) with the costimulatory receptor CD28 after which effector T-cell response is decreased. Regulatory T cells are also known to upregulate CTLA-4 that suppresses activation and expansion of cytotoxic T cells.^6,7 CTLA-4 is only known to be upregulated on T cells and its ligands are expressed on antigen-presenting cells (APC). Preclinically, CTLA-4–deficient mice are known to die early in life from widespread autoimmune syndromes.⁸

Another key checkpoint receptor is PD-1. PD-1 is known to be expressed on activated T cells and mediates immune suppression. In the periphery, the PD-1 receptor binds to its ligands PD-L1 (B7-H1) and PD-L2 (B7-DC), which can be expressed on APCs, as well as tumor cells.⁹ Binding of PD-1 with its ligands results in downregulation of activated T cells. Pre-clinically, PD-1–deficient mice are known to develop modest strain and organ-specific autoimmunity later in life.¹⁰ Tumors are able to coopt the PD-L1 ligand to use it to bind to PD-1 and thus able to down-regulate the immune response.¹¹

ANTI–CTLA-4 INHIBITORS

Antibodies have been developed to block the CTLA-4 pathway by binding to the CTLA-4 receptor. By blocking CTLA-4, this allows binding of B7.1 to its costimulatory receptor CD28 that causes an overriding stimulatory signal and T-cell activation.¹²CTLA-4 blockade is analogous to releasing the breaks on the immune system. Two different antibodies have been developed to block CTLA-4. Currently, ipilimumab is being studied in phase III trials in combination with chemotherapy in both NSCLC and small cell lung cancer (SCLC).

Ipilimumab

Ipilimumab is an anti–CTLA-4 antibody that is approved for use in melanoma and also has been tested in combination with chemotherapy in NSCLC. In a randomized phase II trial of patients with never-treated stage 4 NSCLC, patients were randomized to either combination chemotherapy (paclitaxel 175 mg/m² and carboplatin [AUC 6]), or the same chemotherapy combined with ipilimumab (10 mg/kg) given once every 3 weeks either in combination with cycle 1 through cycle 4 (concurrent regimen) or starting later with cycle 3 and continuing on through cycle 6 (phased regimen)¹³ (Table 1) The trial enrolled 204 patients. A total of 73 patients were treated with all six cycles of combination therapy and continued on ipilimumab or placebo once every 12 weeks until cancer progression during the maintenance phase of the trial. The primary endpoint of immune related progression-free survival (irPFS) took into account the ability of immune-based therapy to initially cause a tumor flare or growth followed by response. Improvement in irPFS was noted in favor of the phased arm compared to chemotherapy, hazard ratio (HR) of 0.72 (P = .05). An improvement in survival was also noted in the patients treated on the phased arm, but this was not statistically significant (median overall survival [OS] of 12.2 months compared to 8.3 months). The concurrent arm did not result in an improvement in irPFS or OS. Differences in irPFS and OS were noted by histology in a preplanned subset analysis. Squamous cell carcinoma patients who were treated with the phased treatment had a significantly improved irPFS (HR 0.55; 95% confidence interval [CI], 0.27–1.12) and OS (HR 0.4; 95% CI, 0.22–1.03). The patients with nonsquamous cell carcinoma did not benefit with the addition of ipilimumab. The modified World Health Organization (WHO) best overall response rate (ORR) was 32% in the phased arm compared to 14% in the chemotherapy along arm. The immune-related ORR was 32% in the phased schedule compared to 18% for chemotherapy alone. The concurrent schedule yielded a response rate of 21% and immune-related ORR of 21%.

Table 1.

Checkpoint Inhibitors Currently Being Evaluated in NSCLC

Agent	Phase	NSCLC Population	Activity
Anti–CTLA-4
Ipilimumab13	II	Stage 3B-4	Median irPFS
		No prior treatment	Phased regimen = 5.7 mo
			Chemotherapy    alone = 4.6 mo
		N = 204	HR 0.72, P = .05
Anti–PD-1
Nivolumab21	I	Stage 4	ORR = 17.1%
	Expansion cohorts    at 1, 3, 10 mg/kg	2–5 prior therapies allowed	Median duration of    response = 74 wk
			Median OS = 9.6 mo
			1-yr survival = 42%, 2-yr    survival = 24%
		N = 129	Across all dose levels
MK-347526	I	Stage 4	ORR = 21%, irORR = 24%
	Expansion 10 mg/kg    every 2 and every    3 weeks	At least 2 prior therapies N = 38	Median PFS = 9.7 wk Median OS = 51 wk
Anti-PD-L1
MPDL-3280A31	I	Stage 4	ORR = 23%
	Expansion cohorts at    10,15,20 mg/kg    every 3 weeks	N = 53	SD rate at 24 wk = 12% 24-wk PFS = 46%
MEDI-473634	I	Stage 4	Activity seenin 2 NSCLC    patients early in dose    escalation
			Expansion cohorts ongoing

Abbreviations: irPFS, immune-related progression-free survival; HR, hazard ratio; ORR, overall response rate; irORR, immune-related response rate; OS, overall survival; SD, stable disease rate.

Ipilimumab did not add significant toxicity to chemotherapy. In general, the grade 3 or 4 side effect rate was similar across all arms with the rate in the control arm of 27%, concurrent 41%, and phased arm 39%. Only 6% of patients in the phased arm had to discontinue the drugs due to related side effects. Two treatment-related deaths were noted, one in the control arm due to neutropenic sepsis and one in the concurrent treatment arm due to septic shock. Based on these results, further testing using the phased schedule of giving ipilimumab in combination with paclitaxel and carboplatin is planned in patients with metastatic squamous carcinoma of the lung. This international trial is enrolling 920 patients with a primary endpoint of survival (Table 2).

Table 2.

Ongoing Phase III Trials of Checkpoint Inhibitors

	Description	Population
Ipilimumab	Randomized to paclitaxel/carboplatin or    the same combination plus    ipilimumab starting    in the third cycle (phased schedule)	Stage 4 squamous carcinoma patients    —no prior therapy
Ipilimumab	Randomized to platinum/etoposide or    the same combination plus    ipilimumab starting    in the third cycle (phased regimen)	Extensive stage SCLC—no prior therapy
Nivolumab	*Randomized to nivolumab v docetaxel	Stage 4 squamous—1 prior line of    platinum-based combination    therapy
		Stage 4 non-squamous—prior    platinum-based combination    therapy
MK-3475	Randomized to MK-3475 v docetaxel	Stage 4 disease—1 prior platinum-    based combination therapy
MPDL-3280A†	Randomized to MPDL-3280A v    docetaxel	Stage 4 disease—1 prior platinum-    based combination therapy

^*Two separate trials: one for squamous cell carcinoma patients and another for non-squamous cell carcinoma patients.

^†Phase II trial.

The phase II trial described above also included patients with SCLC.¹⁴ Again, the phased regimen improved irPFS but not OS. The median irPFS of the phased regimen treated group was 6.4 months compared to 5.3 months for the chemotherapy-treated patients. The resulting HR was 0.64 (P = .03). The OS was 12.9 months for the phased ipilimumab-treated group compared to 9.9 months for the chemotherapy-treated group. The patients treated with the concurrent ipilimumab regimen did not improve their irPFS or OS compared to control. Tumor response favored the phased regimen again with a modified WHO best ORR of 57% versus 49% in the control arm. Now a phase III trial for patients with metastatic SCLC is combining the standard small cell regimen of platinum and etoposide with ipilimumab using the phased regimen compared to the standard chemotherapy alone. This trial is ongoing¹⁵ (Table 2)

Tremelimumab

Another anti–CTLA-4 antibody is tremelimumab. Tremelimumab is a fully human immunoglobulin G2 (IgG2) antibody. In one phase II study performed in lung cancer patients with stable or responding disease after first-line chemotherapy, it was compared to observation as maintenance therapy.¹⁶ The PFS was not significantly improved. At the time, future development in NSCLC was placed on hold, but now trials combining it with the MEDI-4736 anti–PD-L1 antibody are planned¹⁷ (see details below).

ANTI–PD-1 AND PD-L1 INHIBITORS

Antibodies have been developed to block the interaction between the PD-1 receptor and its ligand, PD-L1¹⁸ (Table 1) At this time, there are two different ways of blocking the PD-1 pathway (Figure 1). Blockade of PD-L1 inhibits the binding of PD-L1 to its known receptors, PD-1 and B7.1. However, an anti– PD-L1 antibody does not block the binding of PD-1 to its other ligand PD-L2. Alternatively, by blocking PD-1, an anti–PD-1 antibody blocks the binding of PD-1 to both of its ligands, PD-L1 and PD-L2. However, the anti–PD-1 antibodies do not block the binding of PD-L1 to B7.1. These differences in blockade may or may not be clinically relevant in terms of efficacy and toxicity. Further studies are needed to compare the efficacy and possible toxicity differences in these two ways of blocking the PD-1 pathway. The first in human, first in class antibody was directed to block PD-1. This antibody demonstrated activity in a multiple tumor types including NSCLC with a short lived mixed response in one NSCLC patient.¹⁹ This trial led the way for further trials using different antibodies developed to block the PD-1 pathway.

Figure 1.

There are two different ways of blocking the PD-1 pathway. By blocking PD-L1, the antibody blocks the binding of PD-L1 to its known receptors, PD-1 and B7.1. However, the antibody does not block the binding of PD-1 to its other ligand PD-L2. Alternatively by blocking PD-1, an anti–PD-1antibody blocks the binding of PD-1 to both of its ligands, PD-L1 and PD-L2. However, the anti–PD-1antibodies do not block the binding of PD-L1 to B7.1. These differences in blockade may or may not be clinically relevant. Further studies are needed to compare the efficacy and possible toxicity differences in these two ways of blocking the same pathway.

Anti–PD-1 Inhibitors

Nivolumab (BMS-936558), a fully human IgG4 molecule, was the first anti–PD-1 antibody to report activity in NSCLC²⁰ (Table 1). In this trial, nivolumab was given intravenously once every 2 weeks in 8-week treatment cycles. Because of early activity seen during the dose-escalation phase in this phase I trial, expansion cohorts of NSCLC patients at 1, 3, and 10 mg/kg accrued. A response rate of 17.1% was noted in the NSCLC population and similar response rates were seen in squamous (16.7%) and non-squamous (17.6%) patients. The median duration of response of 74 weeks is notable. A stable disease rate at 6 months of 10.1% was also seen. Interestingly, a difference in response was seen between the 1-mg/kg (response rate [RR] of 3%), 3-mg/kg (RR 24.3%), and 10-mg/kg (20.3%) lung cancer cohorts. Survival was also reported in this large phase I NSCLC cohort. The 1-year survival rate was 42% and the 2-year survival rate was 24%.²¹

In general, this antibody was easy to tolerate when given once every 2 weeks (Table 3). Of the 129 patients with NSCLC, 71% experienced some type of treatment-related side effect and only 14% experienced grade 3 or 4 toxicities. The most common side effects were fatigue (24%), decreased appetite (10%), and diarrhea (10%). Only 5% experienced significant (grade 3 or 4) side effects that were felt to be immune-mediated. Unfortunately, two patients died due to treatment-related pneumonitis early on in the trial. In general, pneumonitis was not common and was only seen in 6% of patients, with 2% experiencing grade 3 or 4 pneumonitis.

Table 3.

PD-1 Checkpoint Inhibition Toxicities in NSCLC population

Agent and Population, N	Treatment-Related AE, All and Grade 3/4	Most Common Treatment-Related AE	Select AE, All Grade and Grade 3/4	Pneumonitis Rate
Nivolumab21	71%	Fatigue: 24%	53%	All grade: 6%
NSCLC- 129	14%	Decreased    appetite: 10%	5%	Grade 3/4:2% 2 deaths
		Diarrhea: 10%
MK347526     NSCLC-38	53% NR: only 1 patient with grade 3 pulmonary edema	Rash: 21% Pruritis: 18% Fatigue: 16%	NR	1 patient with    grade 2    pneumonitis No deaths
MPDL-3280A31     NSCLC-85	66% 11%	Fatigue: 20% Nausea: 14% Decreased     appetite: 10%	NR 1%	All grade: NR No grade 3–5

Abbreviations: AE, adverse events; NR, not reported; Select AE, adverse events associated with immune mechanism of action.

Further studies in NSCLC are ongoing. A third-line and beyond study in squamous cell carcinoma patients is testing the activity of single-agent nivolumab.²² Two randomized phase III trials are testing nivolumab against docetaxel in the second-line treatment setting (Table 2). One trial is specifically in squamous carcinoma patients²³ and the other trial is only in non-squamous patients.²⁴ Another phase I trial is ongoing and testing the safety of combining nivolumab with different standard therapies such as chemotherapy, bevacizumab, and erloti-nib.²⁵ Also in this trial, ipilimumab is being combined with nivolumab for the first time in NSCLC.

MK-3475, an anti–PD-1humanized IgG4 antibody, also has activity in NSCLC²⁶ (Table 1). A total of 38 patients with NSCLC were treated with MK-3475 at 10 mg/kg given once every 3 weeks. Using standard Response Evaluation Criteria in Solid Tumor (RECIST) criteria, a response rate of 21% was noted in the 33 patients with evaluable disease. A higher response was seen in squamous patients (33%), but the total patient numbers were small (n = 6). In non-squamous patients, a response rate of 16% was also seen. In general, MK-3475 was easily tolerated with rash (21%), pruritis (18%), and fatigue (16%) being the most common side effects (Table 3). Drug-related side effects of any grade were experienced in 53% of the NSCLC patients. No patient experienced a treatment-related death. Further studies are testing the activity of two different doses of MK-3475 versus docetaxel in the second-line treatment setting²⁷ (Table 2) as well as a phase I trial combining MK-3475 with standard platinum chemotherapy combinations.²⁸

Anti–PD-L1 Inhibitors

BMS-936559, a fully human IgG4 molecule, was the first anti–PD-L1 antibody to have activity in NSCLC.²⁹ A response rate of 10% was reported in 49 NSCLC patients in the phase I trial that included patients treated during the dose-escalation phase and expanded cohorts specifically of NSCLC patients treated at the 1-, 3-, and 10-mg/kg dose levels. At 6 months, 12% of NSCLC patients had stable disease. Activity was also similar in squamous (RR 8%) and non-squamous (RR 11%) histologies. Of the total 207 patients on this trial, treatment-related side effects occurred in 61% of patients and 9% of patients experienced grade 3 or 4 toxicities. No treatment-related deaths were noted. Fatigue (16%), infusion reactions (10%), and diarrhea (9%) were the three most common side effects noted on this trial. This antibody is no longer being developed in lung cancer.

Since the activity of BMS-936559 was reported in 2012, other anti–PD-L1 antibodies have demonstrated activity in NSCLC. MPDL-3280A, an engineered human IgG1 antibody, has activity in NSCLC.³⁰ In this phase I trial, MPDL-3280A was given intravenously once every 3 weeks in 6-week treatment cycles for a total of 1 year of treatment (Table 1). Of 53 evaluable patients at the time of the presentation, 23% had a RECIST response and a stable disease rate at 6 months of 17%.³¹ Similar activity was demonstrated in non-squamous and squamous histologies (21% and 27% RR, respectively). Interestingly RRs differed by smoking status. Former or current smokers had a higher RR than never smokers (26% v 10%, respectively). In general, this antibody had similar side effects as the rest of the antibodies in this class (Table 3). Fatigue, nausea, and decreased appetite occurred in 16%, 14%, and 10% of the patients, respectively. Of the 85 NSCLC patients evaluable for side effects, 66% experienced some type of side effect and only 11% experienced grade 3 or 4 toxicities. No deaths occurred due to pneumonitis. Further trials are evaluating this antibody in combination with bevacizumab or compared to docetaxel in the phase II trial in second-line treatment in NSCLC³² (Table 2). Another study is looking at MPDL-3280A activity in only a PD-L1– positive NSCLC population.³³

Another antibody, MEDI-4736, is early on in its phase I trial. MEDI-4736 is an engineered human IgG1 antibody given once every 2 weeks in 6-week treatment cycles. At the European Cancer Conference (ECC) in 2013, reports of two lung cancer patients’ tumor response were presented early on in the dose-escalation phase of the trial.³⁴ Further results of this phase I trial are anticipated as expansion cohorts of patients continue to be enrolled.³⁵ A phase I trial combining this antibody with the anti-CTLA4 antibody, tremelimumab, is planned.¹⁷

THE FUTURE

The future of immunotherapy beyond single-agent checkpoint blockade is in combinations with other therapies. Certainly, these antibodies are being combined with standard chemotherapy. However, the ideal combination, sequence, and schedule have yet to be found. Combinations with radiation therapy also are being explored based on the reports of an abscopal effect of radiation on systemic disease.³⁶Combinations with other immune-based therapies either vaccines, T-cell–based therapies, or other checkpoint inhibitors are being developed. Currently, several phase I trials are ongoing combining two different checkpoint inhibitors, for example, ipilimumab and nivolumab. This combination demonstrated exciting activity in melanoma patients.³⁷Ideally, the use of these combinations would be based on the evaluation of a tumor specimen. With the discovery of ways to detect which checkpoint pathway is being used by an individual’s tumor to shield itself from the immune system and the presence or absence of tumor-infiltrating lymphocytes, personalized immunotherapy might be a reality in the future.³⁸

CONCLUSION

Immunotherapy for the treatment of lung cancer is not yet standard. Vaccines still hold promise even though recently reported results have not shown an improvement in survival. Further refinement of the right population, either with gene signature or using the optimal standard clinical treatment in trials with vaccines, may help yield the survival improvement. Currently, the excitement and hope of bringing immunotherapy to the clinic for the treatment of lung cancer lies in the early results of the anti–PD-1and anti–PD-L1 inhibitors. Long-lived responses in NSCLC seem to be consistent across this class of inhibitors. However, phase III trials comparing these inhibitors as single agents against chemotherapy are still ongoing. The results of these trials are heavily anticipated. Until that time, oncologists and patients have to remain hopeful that checkpoint inhibitors may someday become a routine part of lung cancer care.