failure

ADU-S100 was the first STING agonist in the clinic. Injected directly into tumours, it produced almost no responses, alone or with checkpoint blockade.

Adding a MEK inhibitor to boost iodine uptake before ablation did not improve complete remission rates, cooling the 'redifferentiation for everyone' idea.

Bempegaldesleukin was a re-engineered interleukin-2 meant to be a safer version of a famous old immunotherapy. It added nothing to nivolumab in three phase 3 trials.

Biotech that ran one of the early GPNMB ADCs in triple-negative breast cancer, which failed, and has since pivoted to mast-cell (KIT) antibodies for allergic disease.

An enzyme blocker meant to stop tumours starving T cells of tryptophan. Its 2018 phase 3 failure ended an entire class overnight.

Eprenetapopt (APR-246) was a drug meant to refold mutant p53, the most common broken protein in cancer. Its phase 3 in blood cancer failed in 2020.

A second LAG-3 blocker failed to beat pembrolizumab alone, a warning that the Opdualag result is not automatically a class effect.

Neoantigen-vaccine pioneer (GRANITE, SLATE) that ran out of money and filed for bankruptcy in October 2024; a cautionary tale for the personalised-vaccine field.

The first adjuvant immunotherapy trial in liver cancer looked positive early, then the benefit vanished with longer follow-up.

IMpassion131 was the sister trial to the first immunotherapy success in breast cancer. It failed, and the approval it was meant to confirm was withdrawn.

Billed as the first PARP inhibitor for triple-negative breast cancer, it failed its phase 3 in 2011. It turned out not to inhibit PARP at all.

Adding a PD-L1 blocker to curative chemoradiation did not help, the first of several such failures in head and neck cancer.

KRISTINE tested whether an antibody-drug conjugate could replace chemotherapy before surgery in HER2-positive breast cancer and found it could not: fewer tumours disappeared and more grew before the operation.

Keyhole surgery, assumed equivalent, turned out to be worse: more recurrences and more deaths than open surgery, reversing practice overnight.

LEAP-002 randomised 794 patients with untreated advanced liver cancer to lenvatinib with or without pembrolizumab. Adding the immunotherapy did not lengthen life beyond lenvatinib alone, partly because the control arm did unusually well, a reminder that not every immunotherapy plus anti-angiogenic pairing works.

The first 'don't eat me' signal blocker. Gilead paid $4.9B for it; it was stopped in 2024 after trials showed more deaths, not fewer.

A myeloma drug given accelerated approval in 2021 and withdrawn in the US months later when its confirmatory trial suggested it shortened survival.

Adding a liposomal chemotherapy did not help in second-line bile duct cancer, contradicting an earlier Korean trial.

Nintedanib is a triple angiokinase inhibitor pill (VEGFR, FGFR, PDGFR) approved for pulmonary fibrosis and, in the EU, for second-line lung adenocarcinoma with docetaxel. In mesothelioma the phase 2 part of LUME-Meso suggested slower progression in epithelioid disease, but the phase 3 part showed none, a much-cited warning about small randomised phase 2 signals.

CAR-NK company that shelved its oncology programmes after falling response rates and pivoted to autoimmune disease.

Giving extra chemotherapy after chemoradiation added side effects and no survival benefit, so timing matters: chemotherapy helps before radiation (INTERLACE), not after.

Two large trials found that adding palbociclib after surgery does not prevent recurrence, even though the same drug helps in metastatic disease.

An oral SERD did not beat the aromatase inhibitor in first-line treatment when both were combined with palbociclib. Oral SERDs earn their place after resistance, not before it.

Adding relatlimab to adjuvant nivolumab did not reduce recurrence, so the LAG-3 combination stays a treatment for measurable disease.

Rovalpituzumab tesirine (Rova-T) was the first drug against DLL3 in small-cell lung cancer. AbbVie bought it for $5.8B and abandoned it after two failed phase 3 trials. The target later worked with a different weapon.

Two large trials found radioactive beads were gentler than sorafenib but did not help patients live longer.

An immune-brake blocker that looked excellent in a phase 2 lung cancer trial and then failed every phase 3.

A HER2 ADC with a DNA-alkylating payload that beat chemotherapy in a phase 3 trial yet never reached the market, because eye and lung toxicity and a stronger rival arrived first.

TrilynX tested xevinapant, a drug meant to make tumours more sensitive to radiation, with cisplatin chemoradiation in 730 patients with locally advanced head and neck cancer. Despite a striking phase 2 result, the phase 3 was stopped for futility in 2024 with worse outcomes on xevinapant, a reminder that early wins in this disease often do not replicate.

Tusamitamab ravtansine was Sanofi's ADC against the classic CEA tumour marker, stopped for futility in lung cancer in 2023.

XPORT-EC-042 tested maintenance selinexor against placebo in TP53-normal advanced endometrial cancer after response to platinum chemotherapy, chasing a subgroup finding from the earlier SIENDO trial. It missed its primary endpoint in July 2026: a trend favoured selinexor but did not reach statistical significance, a lesson in the limits of post-hoc biomarker subgroups.