Oncology failure signals

Oncology Phase II clinical trial failure signals in stopped trials

Oncology is the largest disease area in the stopped-trial dataset. Phase II is where many programs start to show whether the biology is promising enough to keep moving.

2026-07-127 min readKeyword: oncology clinical trial failures
Five facts from the dataset
  • Oncology accounts for 7,871 stopped records in the current dataset.
  • Oncology has 581 likely biological failure signals: 304 efficacy/futility and 277 safety records.
  • Phase II appears in 4,524 oncology stopped records.
  • Within oncology Phase II records, 204 are efficacy/futility and 159 are safety stops.
  • The largest oncology sponsors by stopped-record count include M.D. Anderson Cancer Center, National Cancer Institute (NCI), and Novartis Pharmaceuticals.

Why oncology Phase II is worth separating

Oncology is not just another disease area in this dataset. It is the largest one, with 7,871 stopped records. That makes it useful for search demand, but also easy to misread if everything is grouped together.

Phase II is especially important because it often sits between early safety/tolerability work and larger confirmatory trials. A stop at this point can be a stronger signal about efficacy, futility, dose, endpoint, or patient-selection problems.

What the oncology slice shows

In oncology, operational and unclear stop reasons are still the biggest groups: 4,175 operational records and 3,099 other/unknown records. But the biological signal is large enough to study directly, with 304 efficacy/futility stops and 277 safety stops.

That is why the oncology view should not be just a list of terminated cancer trials. It should separate scientific signals from administrative noise.

The Phase II signal

Phase II appears in 4,524 oncology stopped records. Inside that slice, the dataset includes 204 efficacy/futility stops and 159 safety stops.

Those are the records I would inspect first when looking for failed endpoints, weak activity, tolerability problems, or early signs that a program was not strong enough to continue.

Sponsor context matters

The largest oncology stopped-trial sponsor counts include M.D. Anderson Cancer Center with 326 records, National Cancer Institute (NCI) with 311, Novartis Pharmaceuticals with 138, Hoffmann-La Roche with 97, and Washington University School of Medicine with 91.

Those counts should not be read as a simple ranking of bad performance. Large research centers and active sponsors naturally run more studies. The better use is comparison inside a reason bucket, phase, and disease context.

Oncology stop-reason buckets

Reason bucketOncology records
Operational4,175
Other/unknown3,099
Efficacy/futility304
Safety277
Regulatory16

Oncology Phase II stop signals

Reason bucketPhase II oncology records
Operational2,456
Other/unknown1,697
Efficacy/futility204
Safety159
Regulatory8

Largest oncology stopped-trial sponsor counts

SponsorOncology stopped records
M.D. Anderson Cancer Center326
National Cancer Institute (NCI)311
Novartis Pharmaceuticals138
Hoffmann-La Roche97
Washington University School of Medicine91

FAQ

How many oncology stopped trial records are in the dataset?

The current dataset contains 7,871 oncology stopped trial records.

How many oncology records are likely biological failure signals?

There are 581 oncology records classified as likely biological failure signals: 304 efficacy/futility records and 277 safety records.

Why focus on Phase II oncology trials?

Phase II often tests whether the treatment signal is strong enough to justify larger trials, so efficacy, futility, and safety stops in this phase can be especially useful for failure analysis.

Source note: counts are generated from the current ClinicalTrials.gov-derived stopped-trial dataset used by ClinicalTrialFailures.com. These labels are analytical screening signals, not medical advice.