Checkpoint Inhibitors
Immune checkpoint inhibitors (ICIs) are monoclonal antibodies that block inhibitory receptors — PD-1, PD-L1, or CTLA-4 — on T cells or tumor cells. By removing these "immune brakes," ICIs reinvigorate tumor-specific CD8+ T cells that have been exhausted by chronic antigen exposure. Since ipilimumab's approval in 2011, the class has transformed oncology, producing durable complete responses in previously untreatable malignancies. Over 50 tumor types now have approved ICI indications.
Overview
The immune system normally keeps T cell activation under tight control through co-inhibitory receptors — molecular "checkpoints" that prevent autoimmunity and limit the intensity of immune responses. Tumors exploit these same brakes to escape immune destruction. The two principal axes exploited by cancer are the PD-1/PD-L1 axis (peripheral exhaustion, operating in the tumor microenvironment) and the CTLA-4/B7 axis (priming phase, operating in lymph nodes).
Checkpoint inhibitors are immunoglobulin antibodies — predominantly IgG4 or engineered IgG1 — that block these inhibitory interactions. They do not directly kill tumor cells; instead, they remove immunological constraints so the patient's own immune system can mount and sustain anti-tumor responses.
The landscape encompasses three validated target axes with multiple approved agents, plus emerging targets (LAG-3, TIM-3, TIGIT). Combination approaches — particularly PD-1 + CTLA-4 co-blockade — produce synergistic activity by targeting non-redundant phases of T cell activation.
Targets and Approved Agents
| Target | Agent (Brand) | Antibody Type | Key Approvals |
|---|---|---|---|
| PD-1 | Pembrolizumab (Keytruda) | Humanized IgG4κ | Melanoma, NSCLC (PD-L1 ≥1%), TNBC, MSI-H/TMB-H pan-tumor, cervical, HCC, biliary, TNBC, RCC, esophageal, CRC (MSI-H), HNSCC, cHL, MCC, EC, TMB-H |
| PD-1 | Nivolumab (Opdivo) | Fully human IgG4 | Melanoma, NSCLC, RCC, urothelial, HNSCC, cHL, HCC, gastric/GEJ, MSI-H CRC, esophageal, MPM, NSCLC adjuvant |
| PD-1 | Cemiplimab (Libtayo) | Fully human IgG4 | CSCC, BCC, NSCLC (PD-L1 ≥50%), cervical |
| PD-L1 | Atezolizumab (Tecentriq) | Humanized IgG1 (Fc-engineered) | NSCLC (IMpower150 combination), urothelial, TNBC (PD-L1+), HCC, SCLC (withdrawn US), AML (withdrawn) |
| PD-L1 | Durvalumab (Imfinzi) | Human IgG1κ (Fc-silent) | Unresectable Stage III NSCLC (PACIFIC), SCLC, BTC, urothelial |
| PD-L1 | Avelumab (Bavencio) | Fully human IgG1 | MCC (maintenance), urothelial maintenance (JAVELIN Bladder 100) |
| CTLA-4 | Ipilimumab (Yervoy) | Fully human IgG1 | Melanoma (1st and 2nd line), RCC (+nivo), CRC (MSI-H, +nivo), NSCLC (+nivo+chemo), MPM (+nivo), ESCC (+nivo) |
| CTLA-4 | Tremelimumab (Imjudo) | Fully human IgG2 | Unresectable HCC (+ durvalumab, STRIDE regimen) |
| LAG-3 | Relatlimab (in Opdualag + nivo) | Fully human IgG4 | Melanoma (unresectable/metastatic, 1st line; RELATIVITY-047 — PFS 10.12m vs 4.63m vs nivolumab alone) |
Mechanism of Action
PD-1 / PD-L1 Axis — Peripheral Exhaustion
In the tumor microenvironment (TME), chronic antigen stimulation of CD8+ T cells drives upregulation of PD-1 (programmed death-1), an inhibitory receptor on T cells. Tumor cells and tumor-infiltrating myeloid cells counter-express PD-L1 (CD274/B7-H1) via multiple mechanisms:
- IFN-γ-mediated induction: Tumor-infiltrating T cells release IFN-γ → JAK1/2-STAT1/3 signaling → IRF1 transcription factor → PD-L1 transcription (adaptive immune resistance)
- Oncogenic signaling: EGFR, ALK fusions, MYC amplification, PTEN loss → constitutive PI3K-AKT and MAPK signaling → PD-L1 upregulation independent of IFN-γ
- Hypoxia: HIF-1α directly transactivates the CD274 (PD-L1) promoter
PD-L1 binds PD-1 on tumor-specific CD8+ T cells → PD-1 recruits SHP-2 phosphatase to its cytoplasmic ITSM motif → SHP-2 dephosphorylates CD28 and proximal TCR signaling components (ZAP-70, LAT, PLC-γ) → blunted TCR signal → reduced IL-2 production → T cell exhaustion (co-expression of PD-1, TIM-3, LAG-3, TOX transcription factor).
Anti-PD-1/PD-L1 blockade effect
Antibody blockade of PD-1 or PD-L1 prevents SHP-2 recruitment → restores TCR/CD28 signaling → reactivation of exhausted CD8+ T cells → tumor cytolysis via granzyme B/perforin and FasL → regression. Also reinvigorates "progenitor exhausted" T cells (TCF1+PD-1+) capable of clonal expansion.
CTLA-4 / B7 Axis — Lymph Node Priming Phase
CTLA-4 (CD152) is upregulated on T cells shortly after activation and constitutively on regulatory T cells (Tregs). It competes with the co-stimulatory receptor CD28 for binding to B7-1 (CD80) and B7-2 (CD86) on antigen-presenting cells (APCs) in lymph nodes. CTLA-4 binds B7 ligands with ~20-fold higher affinity than CD28, effectively out-competing CD28 and terminating the priming signal. Anti-CTLA-4 antibodies block this competition → sustained CD28 co-stimulation during T cell priming → larger and more diverse anti-tumor T cell clones expand from lymph nodes.
CTLA-4 vs. PD-1: Non-redundant mechanisms
CTLA-4 acts during the priming phase in draining lymph nodes (affects naive/priming T cells); PD-1 acts in the effector phase within the tumor (affects pre-exhausted effector T cells). This mechanistic non-redundancy explains the synergy of combined anti-CTLA-4 + anti-PD-1 therapy (CheckMate-067).
LAG-3 — Third Checkpoint Axis
Relatlimab + Nivolumab (RELATIVITY-047)
LAG-3 (CD223) binds MHC-II on APCs and fibrinogen-like protein 1 (FGL1) on tumor cells → inhibits T cell activation synergistically with PD-1. Dual blockade doubled PFS vs nivolumab alone in treatment-naive melanoma (10.12m vs 4.63m, HR 0.75). Opdualag approved March 2022 — first approved LAG-3 inhibitor.
Key Trial Results
| Trial | Drug(s) | Population | Key Outcome |
|---|---|---|---|
| KEYNOTE-001 | Pembrolizumab | Advanced melanoma (ipilimumab-naive and pretreated) | 33% ORR overall; durable responses — 5-year OS 34% (treatment-naive cohort ~41%) |
| CheckMate-057 | Nivolumab | Pretreated advanced non-squamous NSCLC | OS 12.2m vs 9.4m docetaxel (HR 0.73); landmark 2014 survival benefit in NSCLC |
| KEYNOTE-024 | Pembrolizumab vs platinum chemo | Untreated NSCLC, PD-L1 TPS ≥50% | PFS 10.3m vs 6.0m (HR 0.50); OS 30.0m vs 14.2m — established pembro as 1st-line in PD-L1-high NSCLC |
| CheckMate-067 | Nivolumab + ipilimumab | Untreated advanced melanoma | 5-year OS 52% (combo) vs 44% (nivo) vs 26% (ipi); durable benefit — 6.5-yr data confirm 49% OS for combo; PFS 11.5m vs 6.9m vs 2.9m |
| KEYNOTE-522 | Pembrolizumab + chemo → pembro maintenance | Early-stage TNBC (neoadjuvant) | pCR 64.8% vs 51.2% (chemo alone); EFS HR 0.63 at 36-month follow-up — FDA-approved regardless of PD-L1 |
| IMpower150 | Atezolizumab + bevacizumab + carboplatin/paclitaxel | Untreated metastatic non-squamous NSCLC | PFS 8.3m vs 6.8m (HR 0.62 in biomarker-selected); benefit regardless of PD-L1 status; EGFR/ALK altered subgroup also benefited |
| KEYNOTE-158 | Pembrolizumab | Pan-tumor MSI-H/dMMR; TMB-H ≥10 mut/Mb | MSI-H ORR ~40% (dMMR Lynch tumors); TMB-H ORR 29% — first pan-tumor agnostic approvals (2017 dMMR, 2020 TMB-H) |
Predictive Biomarkers
PD-L1 Expression
PD-L1 immunohistochemistry (IHC) is the most widely used predictive biomarker, but scoring systems and thresholds differ by drug-tumor combination. TPS (Tumor Proportion Score) counts only tumor cell staining (used with pembrolizumab in NSCLC, HNSCC); CPS (Combined Positive Score) counts tumor cells + immune cells + stromal cells per 100 tumor cells (used with pembrolizumab in gastric, cervical, urothelial, TNBC, esophageal). Key thresholds: TPS ≥50% for pembro monotherapy 1st-line NSCLC; CPS ≥10 for pembro in TNBC (KEYNOTE-522 approved regardless); CPS ≥1 for gastric/GEJ.
MSI-H / dMMR
Microsatellite instability-high (MSI-H) or deficient mismatch repair (dMMR) status predicts robust ICI response across tumor types. Mechanism: MMR deficiency (MLH1, MSH2, MSH6, PMS2 loss) → accumulation of insertions/deletions at microsatellite loci → high frameshift neopeptide burden → high TMB → immunogenic tumors. Lynch syndrome is the hereditary cause; sporadic MSI-H occurs in ~15% of colorectal cancers (MLH1 hypermethylation) and variable rates in endometrial (25–30%), gastric (~9%), biliary (~5%). KEYNOTE-158 ORR ~40% across 11 MSI-H tumor types → 2017 pan-tumor approval of pembrolizumab.
Tumor Mutational Burden (TMB)
TMB-High (≥10 mutations/Mb) by Foundation Medicine CDx was approved as a pan-tumor biomarker for pembrolizumab in 2020 (KEYNOTE-158). Higher TMB correlates with greater neoantigen load → greater chance of immunogenic mutations recognized by T cells. Importantly, MSI-H and TMB-H are not synonymous: only ~50% of MSI-H tumors are TMB-H by the 10 mut/Mb cutoff, and TMB-H can occur in MSS tumors (e.g., bladder, NSCLC). TMB utility is tumor-type dependent — strong signal in NSCLC, TMB, skin; weak in colorectal.
EBV Association
EBV-positive gastric cancers and nasopharyngeal carcinoma (NPC) show high PD-L1 expression driven by viral LMP1 signaling → particularly high ICI response rates (NPC ORR ~30% with pembro monotherapy; EBV+ gastric ~50% in early series).
Negative predictors: STK11/KEAP1 mutations (NSCLC — limit IFN-γ response), β2-microglobulin loss (MHC-I loss → T cell invisibility), PTEN loss (prostate, endometrial), EGFR mutations in NSCLC (immunologically "cold" TME, ICI generally not beneficial as monotherapy in EGFR+ NSCLC).
Immune-Related Adverse Events (irAEs)
ICIs unleash autoreactive T cells that can attack normal tissues, causing a distinct pattern of immune-related toxicities (irAEs) that differ mechanistically and clinically from cytotoxic chemotherapy side effects. Anti-CTLA-4 agents (particularly ipilimumab) have higher irAE rates than PD-1/PD-L1 agents; combination ICI further increases frequency and severity.
Colitis / Diarrhea
Most common grade 3-4 irAE with CTLA-4 blockade. Grade 3-4 colitis: ~12% with ipilimumab 3 mg/kg, ~2-3% with anti-PD-1 monotherapy. Colonic biopsy reveals lymphocytic infiltration ± crypt abscesses. Management: grade 1-2 — loperamide + oral steroids; grade 3 — IV methylprednisolone 1-2 mg/kg; grade 4 — permanent discontinuation + consider infliximab if steroid-refractory.
Pneumonitis
2–5% any grade with anti-PD-1; ~1% grade 3+ with monotherapy; higher with combination ICI and with lung cancer (pre-existing lung disease). CT pattern: organizing pneumonia or CIP. Prednisone 1-2 mg/kg/day → taper over 4-6 weeks; grade ≥3: bronchoscopy + IV steroids + hold; grade 4: permanent discontinuation.
Endocrinopathies
Hypothyroidism most common (5–10% with anti-PD-1; usually subclinical initially). Hyperthyroidism/thyroiditis in 3-4%. Hypophysitis: predominantly anti-CTLA-4 (~8% with ipi); headache + fatigue + panhypopituitarism on MRI. Primary adrenal insufficiency: rare but life-threatening — check morning cortisol. Steroid replacement, not withholding ICI, is standard for most endocrine irAEs.
Myocarditis (Rare but Fatal)
Incidence ~0.04-1%; significantly higher with combination ICI (~0.27% vs ~0.06% monotherapy). Mortality ~50% if severe (complete heart block, cardiogenic shock). Mechanism: T cell infiltration of cardiac myocytes sharing antigen with tumor. Troponin is the most sensitive early marker (weekly in first 12 weeks recommended by ASCO). Management: high-dose methylprednisolone 500-1000 mg/day + cardiac monitoring in ICU; abatacept/rATG for steroid-refractory cases.
Hepatitis
Grade ≥3 hepatitis in ~1-2% (anti-PD-1) to 4-5% (combination ICI). Transaminase elevation pattern; biopsy if needed. Hold ICI; prednisone 1-2 mg/kg → taper. Grade 4: permanent discontinuation + consider mycophenolate mofetil if steroid-refractory.
Dermatologic
Most common any-grade irAE (30-40%): maculopapular rash, pruritus, vitiligo (especially in melanoma — paradoxically associated with better tumor response). Severe bullous pemphigoid, SJS/TEN rare. Topical steroids for mild; oral prednisone for extensive grade 2+.
irAE Management Principle: Grade 1 — continue ICI + symptomatic treatment. Grade 2 — hold ICI + oral prednisone 0.5-1 mg/kg/day. Grade 3 — hold ICI + IV methylprednisolone 1-2 mg/kg/day + GI/pulm/cardiology consult. Grade 4 or life-threatening — permanently discontinue ICI + IV methylprednisolone + ICU where needed. Steroid taper over ≥4-6 weeks to prevent rebound.
Connections
References
- Topalian SL, Hodi FS, Brahmer JR, et al. Safety, activity, and immune correlates of anti-PD-1 antibody in cancer (KEYNOTE-001 basis). N Engl J Med. 2012;366(26):2443-54. doi:10.1056/NEJMoa1200690 · PubMed 22658127
- Borghaei H, Paz-Ares L, Horn L, et al. Nivolumab versus docetaxel in advanced nonsquamous NSCLC (CheckMate-057). N Engl J Med. 2015;373(17):1627-39. doi:10.1056/NEJMoa1507643 · PubMed 26412456
- Reck M, Rodríguez-Abreu D, Robinson AG, et al. Pembrolizumab versus chemotherapy for PD-L1-positive NSCLC (KEYNOTE-024). N Engl J Med. 2016;375(19):1823-33. doi:10.1056/NEJMoa1606774 · PubMed 27718847
- Wolchok JD, Chiarion-Sileni V, Gonzalez R, et al. Long-term outcomes with nivolumab plus ipilimumab in advanced melanoma (CheckMate-067). N Engl J Med. 2022;387(25):2349-2359. doi:10.1056/NEJMoa2109441 · PubMed 36546651
- Schmid P, Cortes J, Pusztai L, et al. Pembrolizumab for early triple-negative breast cancer (KEYNOTE-522). N Engl J Med. 2020;382(9):810-21. doi:10.1056/NEJMoa1910549 · PubMed 32101663
- Socinski MA, Jotte RM, Cappuzzo F, et al. Atezolizumab for first-line treatment of metastatic nonsquamous NSCLC (IMpower150). N Engl J Med. 2018;378(24):2288-2301. doi:10.1056/NEJMoa1716948 · PubMed 29863955
- Marabelle A, Le DT, Ascierto PA, et al. Efficacy of pembrolizumab in patients with noncolorectal high microsatellite instability/mismatch repair-deficient cancer (KEYNOTE-158). J Clin Oncol. 2020;38(1):1-10. doi:10.1200/JCO.19.02105 · PubMed 31682550
- Tawbi HA, Schadendorf D, Lipson EJ, et al. Relatlimab and nivolumab versus nivolumab in untreated advanced melanoma (RELATIVITY-047). N Engl J Med. 2022;386(1):24-34. doi:10.1056/NEJMoa2109970 · PubMed 34986285