Clinical Trial Intelligence

OS vs PFS as Primary Endpoints in Phase III Colorectal Cancer Trials

29 May 2026

Across 68 colorectal, colon, and rectal cancer Phase III studies authorized in CTIS from 2024–2026, PFS appeared as a primary or co-primary endpoint in 17 trials, while OS appeared as a primary or co-primary endpoint in 6 trials. OS/PFS co-primary designs were uncommon, appearing in only 2 trials, while most non-OS/PFS primary endpoints reflected curative-intent, organ-preservation, response, diagnostic, surgical, or safety-driven study designs.

68
Included Phase III trials
25.0%
Used PFS as primary or co-primary
8.8%
Used OS as primary or co-primary
2
OS + PFS co-primary designs

PFS was the dominant survival-primary endpoint, while OS-primary designs were selective.

PFS appeared as the only OS/PFS primary endpoint in 15 of 68 trials (22.1%). OS appeared without PFS as a primary endpoint in 4 of 68 trials (5.9%), and OS plus PFS co-primary designs appeared in 2 of 68 trials (2.9%).

Primary endpoint strategy
PFS-only primary15 / 68 · 22.1%
OS-only primary4 / 68 · 5.9%
OS + PFS co-primary2 / 68 · 2.9%
Other primary endpoint architecture47 / 68 · 69.1%
OS vs PFS primary endpoint use · colorectal Phase III trials
Interpretation

The dataset shows PFS as the practical dominant primary endpoint for interventional advanced-disease colorectal cancer trials, while OS-primary use is more selective and generally appears where survival separation is central enough to justify longer follow-up or where PFS alone may not capture the decisive clinical question.

OS/PFS co-primary designs appeared only in the later cohort years.

In 2024, OS/PFS primary use was mainly PFS-only or OS-only. In 2025 and 2026, the dataset included explicit OS + PFS co-primary designs, with one such trial in each year.

OS/PFS primary endpoint use by authorization year
Year
PFS-only
OS-only
OS + PFS
2024
12
3
0
2025
1
0
1
2026
2
1
1
Yearly endpoint strategy · OS and PFS as primary/co-primary endpoints
Interpretation

The later-year appearance of OS + PFS co-primary designs suggests that some contemporary colorectal Phase III programs are anchoring both disease-control and survival claims at the primary level, but this remains a minority pattern rather than the standard endpoint architecture.

PFS clustered around advanced, metastatic, unresectable, or biomarker-selected colorectal cancer trials.

Among the 17 trials using PFS as a primary or co-primary endpoint, the dominant context was advanced-disease drug development: metastatic colorectal cancer, locally advanced unresectable colorectal cancer, KRAS G12C-mutant disease, BRAF/KRAS/NRAS-defined disease, or trials using RECIST-based independent radiology review.

Observed context of PFS primary use
Advanced / metastatic / unresectable disease context Dominant
RECIST / BICR / IRRC-based radiologic assessment Frequent
Targeted or antibody-based combinations Common
Curative-intent localized disease Less typical
PFS-primary context · advanced colorectal trial design
Interpretation

PFS appears to function as the principal disease-control endpoint when the trial question depends on earlier radiographic separation, targeted-therapy effect, or randomized comparison in metastatic or unresectable disease. disease-control endpoint when the trial question depends on earlier radiographic separation, targeted-therapy effect, or randomized OS frequently remains present, but more often as a secondary endpoint than as the primary decision endpoint.

OS-primary designs were uncommon but concentrated in survival-defining advanced disease programs.

OS appeared as a primary or co-primary endpoint in 6 of 68 trials. Only 4 trials used OS without PFS as the primary OS/PFS endpoint, while 2 trials paired OS directly with PFS as co-primary endpoints.

OS primary endpoint architecture
4
OS-only primary endpoint pattern
2
OS + PFS co-primary pattern
8.8%
All OS-primary or co-primary trials
OS-primary architecture · colorectal Phase III trials
Interpretation

The low OS-primary share suggests that OS is reserved for a narrower set of colorectal Phase III trials where mortality is the central efficacy claim, while many studies use OS as confirmatory secondary evidence behind PFS, DFS, RFS, response, or organ-preservation endpoints.

Most colorectal Phase III studies did not use OS or PFS as the primary endpoint.

Non-OS/PFS primary endpoint architectures covered 47 of 68 trials (69.1%). These included DFS, RFS, TTR/TFS, organ preservation, pathological response, ORR, diagnostic sensitivity/specificity, surgical outcomes, toxicity, functional outcomes, and procedure-related endpoints.

Endpoint families outside OS/PFS primary use
DFS / RFS / recurrence
Curative-intent and post-resection designs
Organ preservation
Rectal cancer and watch-and-wait strategies
pCR / cCR / response
Neoadjuvant or tumor-response studies
Surgical / diagnostic / safety
Procedure, imaging, and tolerability-focused trials
Endpoint displacement pattern · non-OS/PFS primary designs
Interpretation

The large non-OS/PFS group reflects the diversity of colorectal Phase III development: localized colon cancer often moves toward DFS/RFS, rectal cancer toward organ preservation or pathological response, and surgical or diagnostic studies toward procedure-specific endpoints rather than survival endpoints.

Executive Interpretation

The colorectal Phase III endpoint landscape is not a simple OS-versus-PFS split. PFS is the most common OS/PFS primary endpoint, OS-primary use is selective, and most trials use endpoint architectures shaped by disease setting, resectability, treatment intent, and whether the question is systemic control, survival, organ preservation, surgery, imaging, or pathological response.