Oncology Trials Terminate at 4x the System Rate, and 'Business Decision' Beats 'Safety' as the Dominant Reason

USE CASE. Three groups need to know per-condition termination rates: (1) patient-advocacy organizations and disease foundations advising patients on the realistic probability that a trial they enroll in will run to completion; (2) IRBs and trial design researchers calibrating expectations for new trials in specific disease areas; (3) journalists and policy researchers analyzing the structural causes of trial attrition. ClinicalTrials.gov publishes the underlying records but does not publish per-condition termination rates with reason analysis. RESULT 1 (the rate). System-wide termination rate across the 225,264 trials starting 2020-2025: approximately 3.0% (6,709 terminated). The top 10 conditions by termination rate among conditions with ≥100 total trials: advanced solid tumor 12.1% (63/519), myelodysplastic syndromes 11.4% (12/105), coronavirus infection 11.4% (12/105), acute myeloid leukemia 11.1% (46/414), solid tumor 10.9% (46/422), solid tumors 10.3% (14/136), melanoma 10.0% (25/250), non small cell lung cancer 10.0% (31/311), COVID 9.8% (23/234), carcinoma non-small-cell lung 9.7% (14/144). 8 of the top 10 are oncology conditions and the other 2 are COVID variants. The advanced solid tumor rate of 12.1% is approximately 4 times the system mean of 3.0%. RESULT 2 (the reason). The WhyStopped free-text reasons for the top-10-rate conditions tell a structural story. For the 8 oncology conditions in the top 10, the most common reason words are 'decision', 'business', 'strategic', 'sponsor' — i.e. the sponsor decided to stop the trial for business or strategic reasons unrelated to safety or efficacy of the drug. Solid tumors example: 'decision: 6, business: 6, sponsor: 5, study: 4, terminated: 3, reasons: 2, unrelated: 2, patient: 2'. Melanoma example: 'decision: 7, business: 5, terminated: 5, study: 4, patient: 3, recruitment: 3, sponsor: 3, strategic: 3'. NSCLC example: 'decision: 6, study: 6, strategic: 5, safety: 5, accrual: 5, terminated: 4, trial: 3, concerns: 3'. By contrast the COVID terminations are dominated by recruitment failure: 'study: 11, patients: 7, eligible: 3, standard: 3, subjects: 3, plasma: 2, greatly: 2, recruitment: 2' — i.e. the trial couldn't enroll patients because community transmission patterns shifted, the standard of care evolved, or eligibility criteria became impossible to meet. STRUCTURAL READING. The conventional narrative around oncology trial failure is medical: drugs are toxic, efficacy doesn't materialize, patients can't tolerate the regimen. The 2020-2025 ClinicalTrials.gov ledger says something different. The dominant reason oncology trials in the highest-termination-rate conditions stop is 'business decision' — sponsor strategic re-prioritization, portfolio rationalization after a corporate merger, drug development pipeline cuts. This means the typical patient who consented to an oncology trial that later stopped was failed by sponsor strategy rather than by the science. The 4-times-the-system-rate termination signal in oncology is structurally driven by industry portfolio decisions, not by drug failure rates per se. CAVEATS. (1) The first-listed-condition key is coarse — many oncology trials list a generic condition like 'solid tumor' even when the actual indication is more specific. The top 10 has multiple variants of 'solid tumor' / 'solid tumors' / 'advanced solid tumor' / 'solid tumor, adult' that represent overlapping populations. (2) WhyStopped is free-text and the word-count analysis is heuristic; manual coding of each terminated trial would give a cleaner reason classification. (3) The 'business decision' framing depends on how sponsors choose to disclose stop reasons in the public registry; some terminations reported as 'business decision' may have an underlying safety or efficacy component the sponsor chose not to disclose. (4) The 225,264 denominator includes trials in all phases, all sponsor types, and all geographies; phase-stratified analysis would refine the per-condition rates.

When a clinical trial enrolls patients, the patients are told the trial will run to completion if at all possible, and that any decision to stop early will be for a serious reason — usually patient safety, futility (the drug isn't working), or unforeseen toxicity. Sometimes a trial stops because the sponsor company decides for unrelated business reasons that the program isn't worth continuing — that gets coded as 'sponsor terminated for business decision' on ClinicalTrials.gov. I downloaded every trial that started between 2020 and 2025 from the federal ClinicalTrials.gov database (225,264 trials in total) and asked: which medical conditions have the highest fraction of terminated trials? The system-wide termination rate is about 3%. But for advanced solid tumor (a generic cancer category), the termination rate is 12.1% — four times the system mean. 8 of the top 10 conditions by termination rate are oncology categories: advanced solid tumor, myelodysplastic syndromes, acute myeloid leukemia, melanoma, non-small-cell lung cancer, renal cell carcinoma, and so on. The other 2 in the top 10 are COVID variants. So oncology trials are 4x more likely to terminate than the typical trial. Why? I read the WhyStopped reasons sponsors filed for each terminated trial in the top 10. For the 8 oncology conditions, the most common words are 'decision', 'business', 'sponsor', 'strategic'. The trials weren't stopped because the drug was unsafe or didn't work — they were stopped because the sponsor company made a business decision to terminate the program. By contrast, the COVID terminations are dominated by 'patients couldn't be recruited' — the standard of care for COVID changed faster than trials could enroll. So the structural finding: the typical patient who enrolled in an oncology trial that later stopped between 2020 and 2025 was failed by sponsor strategy, not by the science. The drug might have been fine. Pharma companies cut programs for portfolio reasons (a merger, a pipeline rationalization, a strategic pivot to a hot new modality), and the patients enrolled in those abandoned trials are a downstream cost that doesn't show up in the company's earnings call. Why this matters: patient-advocacy groups counseling cancer patients on whether to enroll in a specific trial should know the risk that the trial will be terminated for non-medical reasons in the oncology space is roughly 1 in 10. Trial design researchers should treat 'business decision' as a first-class termination cause in oncology, not just an edge case.

advanced solid tumor 12.1% (63/519) · myelodysplastic syndromes 11.4% (12/105) · coronavirus infection 11.4% (12/105) · acute myeloid leukemia 11.1% (46/414) · solid tumor 10.9% (46/422) · solid tumors 10.3% (14/136) · melanoma 10.0% (25/250) · non small cell lung cancer 10.0% (31/311) · covid 9.8% (23/234) · carcinoma, non-small-cell lung 9.7% (14/144)

Oncology (top 8 oncology conditions): dominant words 'decision', 'business', 'strategic', 'sponsor' — sponsor portfolio decisions, not safety or efficacy · COVID variants (covid, sars-cov-2): dominant words 'patients', 'eligible', 'recruitment', 'standard' — recruitment failure as standard of care evolved

225,264 trials started 2020-01-01 through 2025-12-31 · 6,709 terminated · system rate 2.98% · 43,343 distinct first-listed conditions · 266 conditions with ≥100 trials in the window · top oncology rate (12.1%) is 4.0x the system mean