Immunotherapy roadmap: Coley's toxins → checkpoint inhibitors → engineered immunity
The immunotherapy roadmap is a 130-year arc from injecting bacteria into tumours to releasing immune brakes, and now to designing the immune response itself with vaccines, engagers, and cells.
Checkpoint inhibitors proved the immune system can cure metastatic cancer. The current decade is about the 60-80% who do not respond: supplying antigen (vaccines), supplying T cells (engagers, CAR, TIL), and remodelling the microenvironment.
- 1891-2010historic
Prehistory
The prehistory runs from Coley's toxins (1891) through BCG for bladder cancer (1976) and interferon and IL-2 (1980s-90s) to sipuleucel-T (2010). The era brings sporadic cures, high toxicity and scepticism.
- 2011-2018historic
Checkpoint revolution
Ipilimumab (2011), nivolumab and pembrolizumab (2014) and atezolizumab (2016) are approved; the tumour-agnostic MSI-H approval follows (2017) and the Nobel Prize goes to Allison and Honjo (2018). Durable responses follow across >15 tumour types.
- 2018-2024current
Earlier lines and combinations
Immunotherapy moves earlier and into combinations: neoadjuvant/adjuvant IO in melanoma, NSCLC, TNBC, bladder; IO + chemo, IO + VEGF, IO + IO (LAG-3); T-cell engagers in myeloma and lymphoma; and the first solid-tumour engager (tarlatamab) and TIL (lifileucel).
- 2024-2027emerging
Engineered immunity arrives
Personalised mRNA vaccines pass phase 3 (intismeran, 2026); PD-1×VEGF bispecifics challenge pembrolizumab; ADC + IO becomes first line; oncolytic virus RP1 approved; TCR-T and solid-tumour CAR-T approvals; in vivo CAR first-in-human.
- 2028+speculative30%–55%likely
Speculative
Off-the-shelf shared-neoantigen vaccines for KRAS and TP53; interception vaccines in high-risk carriers; myeloid-reprogramming drugs that make cold tumours hot; immune-PET-guided IO selection; engineered cytokines with tumour-restricted activity.
Probability ranges are named estimates that the claim is borne out on roughly a five-year horizon. They are meant to be argued with: propose a revision with your name and reasoning via a pull request to src/data/confidence.ts.
Story
topPrehistory
The prehistory runs from Coley's toxins (1891) through BCG for bladder cancer (1976) and interferon and IL-2 (1980s-90s) to sipuleucel-T (2010). The era brings sporadic cures, high toxicity and scepticism.
Checkpoint revolution
Ipilimumab (2011), nivolumab and pembrolizumab (2014) and atezolizumab (2016) are approved; the tumour-agnostic MSI-H approval follows (2017) and the Nobel Prize goes to Allison and Honjo (2018). Durable responses follow across >15 tumour types.
Ipilimumab was the first checkpoint inhibitor (2011), and proved the immune system could be unleashed against cancer.
The most widely used cancer immunotherapy, approved in more than 40 settings, including before and after surgery for triple-negative breast cancer.
Nivolumab was the second PD-1 blocker and is often combined with ipilimumab. Long-term data show about half of advanced melanoma patients alive at 10 years on the combination.
The trial with the longest immunotherapy follow-up: about half of patients on the combination are alive at 10 years.
Earlier lines and combinations
Immunotherapy moves earlier and into combinations: neoadjuvant/adjuvant IO in melanoma, NSCLC, TNBC, bladder; IO + chemo, IO + VEGF, IO + IO (LAG-3); T-cell engagers in myeloma and lymphoma; and the first solid-tumour engager (tarlatamab) and TIL (lifileucel).
The trial that added immunotherapy to pre-surgery chemotherapy for triple-negative breast cancer and, uniquely, improved survival.
Opdualag combines relatlimab, the first drug targeting the LAG-3 immune brake, with nivolumab for melanoma.
The first T-cell engager to improve survival in a common solid tumour, small-cell lung cancer.
Lifileucel was the first approved TIL therapy: the patient's own tumour-fighting immune cells are expanded to billions and given back.
Engineered immunity arrives
Personalised mRNA vaccines pass phase 3 (intismeran, 2026); PD-1×VEGF bispecifics challenge pembrolizumab; ADC + IO becomes first line; oncolytic virus RP1 approved; TCR-T and solid-tumour CAR-T approvals; in vivo CAR first-in-human.
INTerpath-001 was the first positive phase 3 trial of a personalised cancer vaccine, announced 19 August 2026.
A Chinese bispecific that beat Keytruda head-to-head on progression-free survival in lung cancer, the first drug ever to do so.
Showed that pairing an ADC with immunotherapy beats chemotherapy plus immunotherapy in first-line PD-L1-positive TNBC.
Vusolimogene oderparepvec is an engineered herpes virus injected into melanoma tumours, approved in August 2026 with nivolumab after immunotherapy failure.
Instead of engineering T cells in a factory, an injection reprograms them inside the patient's body.
Speculative
Off-the-shelf shared-neoantigen vaccines for KRAS and TP53; interception vaccines in high-risk carriers; myeloid-reprogramming drugs that make cold tumours hot; immune-PET-guided IO selection; engineered cytokines with tumour-restricted activity.
Vaccines against mutations or proteins shared by many patients, so they can be made in advance.
PET scans built from radiolabelled antibodies or their fragments, to see any protein an antibody can reach, including immune cells inside tumours.
'Hot' tumours are full of immune cells and respond to immunotherapy; 'cold' tumours have kept the immune system out.