Peptide PD-1/PD-L1 Inhibitors vs Antibodies

Cancer Immunotherapy Peptides

5.9 nM EC50

The macrocyclic peptide JMPDP-027 restores T-cell function with an EC50 of 5.9 nM, comparable to the antibody pembrolizumab.

Miao et al., RSC Advances, 2021

Miao et al., RSC Advances, 2021

View as image

Antibodies targeting the PD-1/PD-L1 checkpoint pathway (pembrolizumab, nivolumab, atezolizumab) have transformed cancer treatment. But antibodies are large (~150 kDa), expensive to manufacture, have long half-lives that make dose adjustment difficult, and penetrate solid tumors poorly. Peptide-based PD-1/PD-L1 inhibitors, typically 1-3 kDa, could address all of these limitations. The question is whether they can match the clinical efficacy of antibodies while delivering on these theoretical advantages. The research is advancing from patent filings to preclinical validation, with one macrocyclic peptide (BMS-986189) having reached clinical trials. For context on the broader landscape of peptide-based immune checkpoint modulation, see our pillar article on peptide-based immune checkpoint inhibitors.

Why Antibodies Have Limitations as Checkpoint Inhibitors

PD-1/PD-L1 antibodies work. That is not in dispute. Pembrolizumab and nivolumab have produced durable responses in melanoma, lung cancer, and dozens of other indications. But they have structural limitations that peptides could address.

Size and tumor penetration. Antibodies at ~150 kDa are too large to efficiently penetrate solid tumors. They accumulate at tumor peripheries and in perivascular regions, leaving the hypoxic tumor core largely untreated. This matters because the tumor core often contains the most immunosuppressive microenvironment.

Half-life inflexibility. Antibody half-lives of 2-4 weeks are an advantage for patient convenience (dosing every 2-6 weeks) but a disadvantage for managing immune-related adverse events. When a patient develops autoimmune colitis or hepatitis, the antibody cannot be quickly cleared from circulation. Peptides with half-lives of hours to days would allow faster dose adjustment and recovery from adverse events.

Manufacturing cost. Antibodies require mammalian cell culture, extensive purification, and cold-chain storage. Peptides can be chemically synthesized, are more stable, and cost less to produce at scale.

Immunogenicity. Antibodies can themselves trigger anti-drug antibody responses that reduce efficacy over time. Smaller peptides are generally less immunogenic.

Shaabani et al. (2018) reviewed the patent landscape for PD-1/PD-L1 antagonists beyond antibodies, documenting over 30 patents filed between 2015 and 2018 for small molecules, peptides, and macrocycles.^[1] The field grew rapidly after the 2015 publication of the crystal structure of the human PD-1/PD-L1 complex, which enabled structure-based design of smaller blocking molecules.

Macrocyclic Peptides: The Leading Peptide Format

Linear peptides are rapidly degraded by proteases in the blood, limiting their therapeutic utility. Macrocyclic peptides (peptides constrained into a ring structure) resist enzymatic degradation and adopt well-defined conformations that can bind protein targets with high affinity. This makes them the dominant format for PD-1/PD-L1 peptide inhibitors.

Miao et al. (2021) reported the rational design of JMPDP-027, a macrocyclic peptide that blocks the PD-1/PD-L1 interaction.^[2] The peptide restores T-cell function with an EC50 of 5.9 nM, a potency comparable to the anti-PD-1 monoclonal antibody pembrolizumab. JMPDP-027 also demonstrated high resistance to enzymatic hydrolysis in human serum, addressing the stability problem that has plagued linear peptide approaches.

The significance of achieving antibody-comparable potency in a peptide format cannot be overstated. It demonstrates that the PD-1/PD-L1 interaction can be blocked by molecules ~100x smaller than antibodies without sacrificing binding affinity or functional activity.

Computational Design: Building Peptide Inhibitors from Scratch

Traditional peptide drug discovery relies on screening natural peptide libraries or modifying known sequences. Guardiola et al. (2021) developed a fundamentally different approach: using computational methods to design macrocyclic peptide inhibitors de novo.^[3]

Their framework generates macrocyclic peptides containing both D- and L-amino acids (the mirror-image forms), creating topologies not found in natural peptides. D-amino acids are resistant to proteases that specifically recognize L-amino acids, providing an inherent stability advantage. The approach discovered drug-like inhibitors of PD-1 that bind with high affinity and have favorable pharmacokinetic properties.

This computational approach is important because it bypasses a fundamental limitation of conventional peptide design. Natural peptides evolved for biological functions unrelated to checkpoint blockade, so adapting them is inherently inefficient. De novo design starts from the target structure and works backward to the optimal peptide, incorporating non-natural building blocks that evolution never explored. The intersection of this approach with how peptide-MHC complexes present cancer targets opens possibilities for dual-function molecules.

BMS-986189: The First Macrocyclic Peptide in Clinical Trials

Bristol-Myers Squibb developed BMS-986189, a macrocyclic peptide inhibitor of PD-L1, which became the first peptide-based checkpoint inhibitor to enter clinical trials. The peptide was identified through a combination of phage display and medicinal chemistry optimization.

BMS-986189 binds PD-L1 and blocks its interaction with PD-1 on T cells. In preclinical studies, it demonstrated anti-tumor activity in syngeneic mouse models and showed favorable pharmacokinetic properties for subcutaneous administration. The clinical trial (Phase 1) evaluated safety and pharmacokinetics in patients with advanced solid tumors.

While detailed clinical results have not been fully published, the progression of a macrocyclic peptide to human testing validates the approach. It demonstrates that regulatory agencies accept the peptide format as a viable therapeutic modality for checkpoint blockade, not just an academic curiosity.

Peptide Combinations: Vaccines and Checkpoint Blockade

Peptides play dual roles in cancer immunotherapy: they can directly block checkpoint interactions (as inhibitors) and prime immune responses (as vaccines). The combination of peptide vaccines with checkpoint blockade is a major area of investigation.

Hailemichael et al. (2018) made a critical discovery: the formulation of peptide cancer vaccines determines whether they synergize with or undermine checkpoint blockade.^[4] In the landmark ipilimumab registration trial, adding gp100 peptide vaccine in incomplete Freund's adjuvant (IFA) actually reduced the benefit of checkpoint blockade. The authors showed this was because IFA creates a persistent antigen depot that traps and kills vaccine-specific T cells at the injection site rather than allowing them to migrate to tumors. Different adjuvant formulations that promote T-cell migration rather than trapping produced strong synergy with checkpoint blockade.

This finding reshaped how peptide vaccines are combined with anti-PD-1/PD-L1 therapy. Zhou et al. (2024) developed a tumor-targeting peptide (TMTP1)-modified nano-vaccine combined with PD-1 blockade for ovarian cancer.^[5] The nano-vaccine format avoids the IFA depot problem while targeting antigen delivery directly to tumors, and the combination with checkpoint blockade enhanced anti-tumor immune responses.

Next-Generation Approaches

Two recent studies illustrate where the field is heading:

Ji et al. (2026) combined the oncolytic peptide LTX-315 with PD-L1 targeting to enhance anti-tumor immune responses alongside nanosecond pulsed electric field ablation.^[6] LTX-315 is a cationic amphipathic peptide that directly lyses tumor cells while simultaneously modulating the tumor immune microenvironment. By targeting PD-L1, LTX-315 combines direct tumor killing with checkpoint modulation in a single peptide, a dual-function approach that antibodies cannot replicate.

Otani et al. (2026) developed a peptide-based photoimmunotherapy drug targeting PD-L1.^[7] Near-infrared photoimmunotherapy (NIR-PIT) uses photosensitizer-conjugated targeting molecules to kill specific cell populations when activated by light. Current NIR-PIT uses antibodies, but their large size limits tissue penetration. Peptide-based NIR-PIT drugs could reach deeper into tumor tissue, improving treatment efficacy for solid tumors.

Esfahani et al. (2023) explored combining peptide receptor radionuclide therapy with checkpoint inhibitors in neuroendocrine tumors.^[8] This approach uses radiolabeled peptides (like DOTATATE) that bind somatostatin receptors on tumor cells to deliver targeted radiation, combined with anti-PD-1 therapy to enhance the immune response against radiation-damaged tumor cells. For more on how peptides are combined with different immunotherapy modalities, see our article on bispecific peptides in cancer.

The Structure-Activity Challenge

Designing peptides that block protein-protein interactions is inherently difficult. The PD-1/PD-L1 binding interface spans approximately 1,970 square angstroms, a large, flat surface without the deep pockets that small molecules typically target. Antibodies can cover this entire surface because they are large enough to make contacts across the full binding interface.

Peptides must achieve the same blocking effect with a much smaller molecular footprint. This requires identifying the critical "hot spots" on the PD-1/PD-L1 interface where binding energy is concentrated and designing peptides that engage these hot spots with maximum efficiency. The crystal structure of the PD-1/PD-L1 complex, published in 2015, revealed that a relatively small number of residues contribute disproportionately to binding energy, making peptide-based disruption feasible.

Macrocyclization addresses another structural challenge: the entropic penalty of binding. Linear peptides lose conformational entropy when they fold into a binding conformation, reducing their effective affinity. Macrocyclic peptides are pre-organized into binding-competent conformations, paying the entropic cost during synthesis rather than at the time of binding. This is why cyclic peptides consistently outperform their linear counterparts in PD-1/PD-L1 binding assays.

The incorporation of D-amino acids adds a further dimension. D-amino acids are mirror images of the natural L-amino acids and are not recognized by most human proteases. Guardiola et al. (2021) showed that incorporating D-amino acids into macrocyclic peptide PD-1 inhibitors improved both metabolic stability and binding affinity, because the non-natural backbone geometry can sometimes achieve better complementarity with the target surface than natural L-peptides.^[3]

Can Peptides Actually Replace Antibodies?

The honest answer: not yet, and possibly not entirely. Antibodies have decades of clinical validation, established manufacturing infrastructure, and proven efficacy across dozens of cancer types. Peptide inhibitors would need to match this in clinical trials.

What peptides can do is fill gaps that antibodies leave:

• Combination therapy where a short-acting, quickly adjustable checkpoint inhibitor is paired with other treatments
• Intratumoral delivery where size matters and antibodies cannot penetrate effectively
• Resource-limited settings where antibody manufacturing costs make checkpoint blockade inaccessible
• Novel modalities like photoimmunotherapy and peptide-drug conjugates where the small size of peptides is a functional requirement, not just an advantage

The most likely near-term outcome is not replacement but complementarity: peptide checkpoint inhibitors used alongside antibodies, or in specific clinical contexts where their unique properties provide an advantage.

The economic argument alone may drive adoption in specific markets. The global immune checkpoint inhibitor market exceeded $35 billion in 2023, with antibody manufacturing costs contributing to treatment prices of $100,000-$250,000 per year. Chemically synthesized peptides could reduce manufacturing costs by an order of magnitude, potentially making checkpoint blockade accessible in healthcare systems that currently cannot afford antibody-based immunotherapy.

The pace of development is accelerating. The number of peptide-based checkpoint modulators in preclinical pipelines grew from fewer than 10 in 2015 to over 50 by 2024. As macrocyclic peptide chemistry matures and computational design becomes more powerful, the quality of peptide candidates entering clinical development will continue to improve. The question is no longer whether peptides can block PD-1/PD-L1 at the molecular level, but whether they can do so safely and effectively in patients with cancer.

The Bottom Line

Peptide PD-1/PD-L1 inhibitors have achieved antibody-comparable potency (EC50 5.9 nM for JMPDP-027), demonstrated superior tumor penetration, and reached clinical trials (BMS-986189). Computational design now generates de novo macrocyclic peptides with non-natural building blocks. The field is expanding into dual-function molecules that combine checkpoint blockade with direct tumor killing or photoimmunotherapy. Peptides are unlikely to fully replace antibodies in the near term but are carving out specific roles where their small size, tunability, and lower manufacturing costs provide distinct advantages.

Frequently Asked Questions

Can peptides block PD-1/PD-L1 as effectively as antibodies?

Yes, in preclinical settings. The macrocyclic peptide JMPDP-027 blocks PD-1/PD-L1 interaction with an EC50 of 5.9 nM, comparable to the antibody pembrolizumab (Miao et al., RSC Advances, 2021). It also resists enzymatic degradation in human serum. However, comparable potency in a test tube does not guarantee equivalent clinical efficacy, and no peptide PD-1/PD-L1 inhibitor has yet completed pivotal clinical trials.

What is BMS-986189?

BMS-986189 is a macrocyclic peptide developed by Bristol-Myers Squibb that binds PD-L1 and blocks its interaction with PD-1 on T cells. It is the first peptide-based checkpoint inhibitor to enter clinical trials in patients with advanced solid tumors. The peptide format offers potential advantages over antibodies including better tumor penetration and more flexible dosing.

Why are peptide checkpoint inhibitors smaller than antibodies?

Monoclonal antibodies like pembrolizumab weigh approximately 150 kDa, while macrocyclic peptide inhibitors typically range from 1-3 kDa. This 50-150x size difference translates to better penetration into solid tumor tissue, faster clearance that allows dose adjustment, lower manufacturing costs from chemical synthesis rather than cell culture, and reduced immunogenicity.

What are macrocyclic peptides?

Macrocyclic peptides are peptides constrained into a ring (cyclic) structure, typically through chemical bonds between side chains or backbone atoms. This cyclic shape makes them more resistant to enzymatic degradation than linear peptides and allows them to adopt defined conformations for high-affinity target binding. Most advanced PD-1/PD-L1 peptide inhibitors use macrocyclic formats.

Do peptide PD-L1 inhibitors penetrate tumors better than antibodies?

Preclinical data indicates they do. Studies using 3D tumor spheroid models showed that peptide PD-L1 inhibitors reached tumor cores that antibodies could not access. Solid tumors have dense stroma and poor vasculature that restricts large molecule penetration, giving smaller peptides a physical access advantage. This could be relevant for treating tumors that respond poorly to antibody-based immunotherapy.

Are peptide checkpoint inhibitors available for cancer treatment?

No peptide PD-1/PD-L1 inhibitor is currently approved for clinical use. BMS-986189 entered Phase 1 clinical trials, and several other macrocyclic peptide candidates are in preclinical development. All currently available checkpoint inhibitors are monoclonal antibodies (pembrolizumab, nivolumab, atezolizumab, durvalumab, avelumab, cemiplimab).