Checkpoint Therapy’s Second Act: Hope, Hype, or Hard Truths?

Introduction

A hallmark of cancer is the ability to evade or subvert immune recognition and response through a range of countermeasures such as altered tumor antigen expression, the recruitment of regulatory immune cells into the tumor microenvironment (TME) and production of immunosuppressive factors.

In the mid to late 1990s, landmark translational research from the laboratories of James Allison and Tasuku Honju (joint recipients of the 2018 Nobel Prize in Physiology or Medicine) established that antibody-mediated blockade of the interaction between the T cell receptors CTLA-4 (cytotoxic T lymphocyte–associated protein 4) and PD-1 (programmed death 1) and their ligands could counter inhibition of T cell function and bring anti-tumour immune responses into play, thereby kickstarting a conceptual revolution in cancer drug development.

By 2010, clinical studies had established that “taking the brakes off” T cells inhibition by targeting so-called immune checkpoints could achieve impressive responses in advanced melanoma, leading to FDA approval of the first immune checkpoint inhibitor (ICI), the CTLA-4 antibody ipilimumab (Yervoy™: Bristol Myers Squibb- “BMS”) in 2011. FDA approvals for the PD-1 blocking antibodies pembrolizumab (Keytruda™: Merck) and nivolumab (Opdivo™: BMS) were granted in 2014, with the first PD-1 ligand-directed ICI, atezolizumab (Tecentriq™: Roche) receiving approval in 2016.

Clinical and commercial enthusiasm for ICI therapy catalysed an unprecedented increase in mono- and combination therapy studies, leading to the approval of ICIs in a still growing number of cancer indications, and to further ICI products entering the market. Keytruda™ is currently approved for the treatment of some 17 tumour types, a breadth of utility which drove global revenues to over $25 billion in 2023¹.

Beyond the classic - still waiting after all these years

“Classic” (CTLA-4, PD-1/L1 directed) ICI therapy has significant shortcomings. ICI therapy can achieve complete, near curative responses in some individuals, but overall response rates remain modest. Less than half of American cancer patients meet the criteria for ICI therapy and of those treated, only around 20% achieve objective responses, with around 13% experiencing durable multiyear responses.

Combination with chemotherapy, radiotherapy and targeted small molecule and biologic treatments has been evaluated in thousands of clinical studies, but to date only a small number of combinations have usefully improved durable response rates. As with other cancer treatments, ICI therapy can be thwarted through primary and acquired resistance. Unlike chemotherapy, ICIs are associated with immune-related adverse events of varying severity requiring intense medical management and which may necessitate treatment discontinuation².

Complementing classic ICI therapy through the blockade or agonism of alternative immune checkpoints has been widely explored as a route to more effective treatments. Promising “next- generation” targets which have been the subjects of substantial development efforts include LAG-3, TIGIT, TIM-3 and VISTA (“co-inhibitory” receptors which contribute to T cell exhaustion among other functions), and GITR, OX40, 4-1BB and ICOS (“co-stimulatory” receptors involved in T cell activation)³.

Success to date can reasonably be described as unspectacular: only the anti-LAG-3 antagonist, relatlimab, (co-formulated with the PD-1 antagonist, nivolumab) has so far secured regulatory approval, the single winner out of a LAG-3 antagonist clinical development pipeline once featuring upwards of 16 candidates.

Why has next-generation ICI success proved elusive?

An obvious barrier is the complexity of tumour biology. The TME is dynamic and highly heterogenous with respect to immune cell populations, and the interplay between inhibitory and stimulatory checkpoint pathways is far from completely understood. Animal models cannot reliably gauge the likelihood of ICI clinical response or adverse events. IC expression itself is of undependable prognostic value, and alternative biomarkers remain to be validated.

Early next-generation ICI clinical studies have rarely reflected promising translational results. A recent review of available data from first in human studies of ten different GITR antagonist candidates, a once attractive target due to its expression on both effector and regulatory T cells, found no compelling evidence of therapeutic activity⁴. Exploitation of VISTA, expressed by both immune and tumour cells, has been confounded by lack of understanding of its role in cancer, with higher VISTA expression being associated with poor outcomes in melanoma and renal cell cancer, but conversely with higher survival rates in other tumour types⁵.

Although modest responses were observed in early studies of the TIM-3 antagonists LY3321367 and MGB453 (sabatolimab), developed by Eli Lilly and Novartis, respectively, the former was abandoned after Phase I study and the latter after failing a Phase III study in haematologic cancers. No responses were observed in melanoma patients receiving the anti-TIM-3 antibody cobolimab, (GSK4069889, acquired along with the 2018 purchase of Tesaro) although GSK continues its evaluation in combination with the PD-1 antagonist, dostarlimab (Jemperli®) in other solid tumour indications. Beigene’s LBL-007 in combination with the anti-PD-1 antibody, tislelizumab is in early evaluation in a subset of colorectal cancer patients.

Initial forays into 4-1BB agonist development (urelumab: BMS and utomilumab: Pfizer) were curtailed due to lack of efficacy at safe doses. Despite a clinical development pipeline compromising more than 40 antibody candidates and bispecific constructs, advancement of 4-1BB agonism remains largely stalled by the need to balance efficacy against hepatotoxicity⁶.

OX40 is expressed on activated T cells and binding to its ligand, OX40L has multiple beneficial effects on T cell function. Exploitation of OX40 agonism in cancer has been hampered by its variable expression, with less than 20% of cancer patients being in the therapeutic “sweet spot” of high OX40 and low OX40L expression No or only very low responses have been observed in OX40 agonist monotherapy studies, with meaningful clinical improvement not observed in the majority of combination studies⁷.

ICOS agonists have been another disappointment. Even with enrichment for patients with ICOS-expressing tumours, objective response rates in a Phase I/II study of vopratelimab (Concentra Biosciences) in combination with anti-PD-1) were in low single figures. GSK and Merck abandoned enrolment of patients in an early study of feladilimab (GSK3359609) in head and neck cancer patients on the advice of its data monitoring committee.

Early clinical failure and discontinuation is not unusual for novel cancer treatments, and next-generation ICI development may have suffered unduly from the sheer scale and momentum of contemporary classic ICI development, with weak or ambiguous early study data compromising the ability to compete for funding and patient recruitment against the myriad investigator and industrially-sponsored studies focused on broadening and optimising the proven modalities of PD-1/L-1 and CTLA-4 blockade across tumour types.

The high-profile failures of expensively acquired non-ICI assets such as indoleamine-2,3-dioxygenase-1 (IDO1) inhibitors, and the IL-15 receptor agonist, bempegaldesleukin (Nektar Therapeutics), on which BMS lavished an estimated $3.8 billion prior to abandonment, may have contributed to steelier pragmatism over finding the next big thing in immuno-oncology. A McKinsey & Co analysis postulated that the 46% decrease in asset-focused venture investment between 2021 and 2022 was due to loss of faith in immuno-oncology⁸.

Narrowed scope, still hope

Given the depth and diversity of current oncology portfolios, showcasing antibody-drug conjugates, radiopharmaceuticals, cellular therapies, and small molecules aimed at once-undruggable targets such as KRAS and SHP2⁹, is next-generation ICI development still relevant?

Flawed it may be, but classic ICI therapy is and will remain an integral element of cancer treatment. A next-generation ICI which can be slotted into established best standard of care ICI regimens to improve response rates and outcomes without additional toxicity could see relatively rapid uptake as first-line through third-line treatment. Efficacy in those cancers poorly responsive to classic ICI therapy offers opportunities for faster approval and dominance of significant niches.

The anti-LAG-3 antibody, relatlimab in fixed dose combination with nivolumab (Opdualag™) received FDA approval in 2022 for advanced melanoma on the back of superior progression-free survival over nivolumab alone. Promising melanoma data continues to accrue, and Q3 2024 sales data points to blockbuster status by year end, although colorectal, liver and gastric cancers studies have been abandoned due to low efficacy.

Despite only modest efficacy in Phase II study, BMS has initiated a pivotal study in which Opdualag™ plus chemotherapy will be compared with pembrolizumab plus chemotherapy in late-stage and recurrent non-small cell lung cancer (NSCLC) patients exhibiting less than 50% PD-L1 expression. Merck’s favezelimab, in fixed-dose combination with pembrolizumab failed to improve overall survival in microsatellite stable metastatic colorectal cancer patients (KEYFORM-007), resulting in the scrapping of all further development.

Regeneron’s fianlimab in combination with cemiplimab (Libtayo™) showed early promise in melanoma, with clinical responses appearing to be independent of LAG-3 or PD-L1 expression and is under evaluation as first-line therapy. After some 20 years in development and less than stellar efficacy in solid tumour studies, Immutep intends to conduct a pivotal study of the soluble LAG-3 protein, eftilagimod alpha, in combination with pembrolizumab as first-line therapy in NSCLC patients.

TIGIT blockade remains of active interest, although a recent cascade of study failures is likely to curb remaining enthusiasm. The antagonist tiragolumab (Roche) failed to improve either overall or progression-free survival over pembrolizumab plus chemotherapy as first-line therapy in NSCLC in the SKYSCRAPER-06 study, with other SKYSCRAPER studies involving combination with atezolizumab in first-line NSCLC, first-line small cell lung cancer (SCLC), first-line non-squamous NSCLC, first-line maintenance in Stage III NSCLC, and second-line cervical cancer showing poor efficacy or being discontinued.

Merck discontinued the KEYVIBE-008 Phase III study of vibostolimab (MK-7684) co-formulated with pembrolizumab plus chemotherapy in patients with extensive-stage SCLC due to futility, along with a higher than anticipated rate of immune-mediated adverse events, an issue also observed in the KEYVIBE-010 melanoma study. The Phase III studies KEYVIBE-003 and KEYVIBE-007 in NSCLC patients were halted due to adverse event rates, with KEYVIBE-006, also in NSCLC, discontinued due to futility. Merck recently announced the abandonment of further vibostolimab development.

Arcus Bioscience and Gilead canned a Phase III study of domvanalimab in first line NSCLC cancers with high PD-L1 expression but remain encouraged by results from combination with an investigational anti-PD-1 antibody, zimberelimab where the doublet achieved superior progression-free survival, overall survival, and objective response rates in NSCLC patients compared with zimberelimab or chemotherapy alone. An AstraZeneca-partnered Phase III study evaluating combination of zimberelimab with durvalumab following radiotherapy in patients with NSCLC is ongoing.

Confidence in TIGIT appears to remain high within iTeos and its partner, GSK, with initiation of a global pivotal study with a target enrolment of one thousand patients in which belrestotug (EOS-448/ GSK4428859A) will be evaluated as first-line therapy in NSCLC patients in combination with PD-1 blockade. And despite rejection by Novartis, Beigene’s ociperlimab continues in early study, as does Merck KGaA’s dargistotug (M6223).

Faith remains in improvement through antibody engineering. iTeos attributes the success of belrestotug to its high affinity binding and Fc-mediated interaction with macrophages and other immune cells, although Fc engineering does not appear to have sufficiently enhanced efficacy of the anti- TIM-3 antibody, INCAGN2385 (Incyte and Agenus) to warrant continued development.

Table 1. Active label expansion or late-stage clinical studies evaluating next-generation immune checkpoint inhibitors (December 2024).

 

Bispecifics have long featured in next-generation ICI pipelines, but as with monospecific candidates, study success has been patchy. Constructs discontinued or offered for licensing after early studies include the PD-1 x LAG-3 bispecifics, tobemstomig (Roche); tebutelimab (Macrogenics and Zai Lab); INCA32459 (Incyte and Merus), FS118 (acquired along with F-Star Therapeutics by Invox), and Abl Bio’s ABL510.

Other bispecifics remain promising. AstraZeneca has initiated a pivotal study of rilvegostomig, a PD-1 x TIGIT bispecific, as first-line therapy in NSCLC in combination with chemotherapy. Genmab is continuing development of the PD-L1 x 4-1BB bispecific acasunlimab in combination with Keytruda™ despite losing BioNTech as its development partner. BioNTech continues to see merit in bispecifics, having acquired a PD-1 x VEGF bispecific from Chinese biopharma, Biothesus and now outright ownership of the company, bringing an earlier stage pipeline featuring PD-L1 x TIGIT, PVRIG x TIGIT, PD-L1 x 4-1BB and claudin 18.2 X 4-1BB bispecifics.

 

Table 2. Selected bispecific antibodies in clinical development which target co-stimulatory or co-inhibitory immune checkpoints (December 2024).

ICI 2.0 validation on the near horizon?

While product revenues are unlikely to rival those of the PD-1/PD-L1 bonanza, and in apparent defiance of historical failure (most recently highlighted by Merck’s abandonment of late-stage anti-TIGIT and anti-LAG-3 development), mono- and bispecific next-generation ICI candidates still feature in the pipelines of BMS, AstraZeneca, Roche/Genentech, Pfizer and other oncology majors and those of small-cap biopharma.

For those with a presence in immuno-oncology, the prize is competitive advantage through expansion and protection of their existing ICI franchises through novel combinations, the gamble being that incremental clinical improvement will not to be rapidly overtaken by other modalities. Near-horizon pivotal and label expansion data may validate current confidence in the potential of next-generation immune checkpoint targeting in anti-PD-1/L-1 combination therapy, and in the longer term, bispecific success may finally see ICI 2.0 make a long-anticipated and significant contribution to cancer treatment.

References

1. Best-selling pharmaceuticals of 2023 reveal a shift in pharma landscape. Buntz B. May 21, Drug Discovery and Development online May 2024 https://www.drugdiscoverytrends.com

2. Estimation of the percentage of US patients with cancer who are eligible for and respond to checkpoint inhibitor immunotherapy drugs. Haslam A & Prasad V. JAMA Network Open. 2019;2(5): e192535.

3. LAG-3= Lymphocyte activation gene 3; TIGIT= T cell immunoreceptor with immunoglobulin and ITIM domains; TIM-3= T cell immunoglobulin and mucin domain 3; VISTA= V-domain immunoglobulin suppressor of T-cell activation; GITR= Glucocorticoid-induced tumour necrosis factor related protein; OX40= Also known as tumour necrosis factor receptor superfamily, member 4 and CD134; 4-1BB= Also known as tumour necrosis factor receptor superfamily 9 and CD137; ICOS= inducible costimulator of T cells.

4. Targeting GITR in cancer immunotherapy - there is no perfect knowledge. Davar D & Zappasodi R. Oncotarget. 2023 Jun 19; 14:614-621.

5. Clinical and research updates on the VISTA immune checkpoint: immuno-oncology themes and highlights. Noelle RJ et al. Front Oncol. 2023 Sep 15; 13:1225081.

6. The emerging landscape of novel 4-1BB (CD137) agonistic drugs for cancer immunotherapy. Claus C et al. mAbs 15(1) 2023.

7. OX40/OX40 ligand and its role in precision immune oncology. Thapa B et al. Cancer Metastasis Rev 43, 1001–1013 (2024).

8. What early-stage investing reveals about biotech innovation. Capra E et al. McKinsey & Co Insights. Online December 12, 2023 https://tinyurl.com/m59c4dxb

9. SHP2= Src homology-2 domain-containing protein tyrosine phosphatase-2.