Clinical Trial Intelligence

How is Event-Free Survival Used in Oncology Trials Authorized in 2024-2026?

16 June 2026

Event-free survival (EFS) appears in 221 of 2,529 oncology trials, or 8.7%. It is most often a secondary endpoint, but when used as a primary endpoint it is associated with larger target sample sizes and late-stage confirmatory settings. The EFS signal is concentrated in haematologic malignancies, paediatric/orphan oncology, and multi-modality drug programs.

221 / 2,529
EFS penetration
8.7% of oncology trials include EFS
142 secondary-only
Endpoint role
64.3% of EFS trials use EFS only as secondary
133 haematology trials
Disease center
60.2% of EFS trials are haematologic malignancies
118 include death
Definition anchor
53.4% explicitly include death as an EFS event

EFS is mainly a Phase II/III and Phase III endpoint

EFS is rare in pure Phase I oncology trials at 4 of 300 trials (1.3%), but becomes common once trials move into efficacy-driven development: 95 of 862 Phase II trials (11.0%) and 83 of 787 Phase III trials (10.5%) include EFS.

Share of oncology trials with EFS by phase
Phase I
1.3%
4 / 300 trials
Phase I/II
5.9%
26 / 440 trials
Phase II
11.0%
95 / 862 trials
Phase II/III
11.9%
10 / 84 trials
Phase III
10.5%
83 / 787 trials
Percentage of trials in each phase category that include at least one EFS endpoint.
Interpretation

EFS behaves like an efficacy-maturity endpoint: it is uncommon in early dose-finding trials but reaches about one in ten trials in Phase II and Phase III settings where durable disease control, relapse prevention, or treatment failure can support clinical decision-making.

EFS use declined across authorization years

The EFS share falls from 158 of 1,689 trials in 2024 (9.4%) to 47 of 603 in 2025 (7.8%) and 16 of 237 in 2026 (6.8%).

Share of oncology trials with EFS by year
2024
9.4%
158 / 1,689 trials
2025
7.8%
47 / 603 trials
2026
6.8%
16 / 237 trials
Authorization year based on first CTIS authorization date.
Interpretation

The downward pattern suggests the 2026 oncology cohort is more weighted toward trials where EFS is not the dominant efficacy frame, while 2024 carried a broader mix of late-stage and haematology-heavy studies using EFS.

EFS is usually secondary, not the main endpoint

Among 221 EFS trials, 142 use EFS as a secondary-only endpoint (64.3%), 51 use it as primary-only (23.1%), and 28 include EFS in both primary and secondary endpoint sets (12.7%).

Endpoint hierarchy among EFS trials
Secondary only
64.3%
142 / 221 trials
Primary only
23.1%
51 / 221 trials
Primary + secondary
12.7%
28 / 221 trials
Role based on whether EFS appears in primary, secondary, or both endpoint lists.
Interpretation

EFS is more often used to complement survival, response, and safety endpoints than to carry the full primary efficacy claim. Still, 79 of 221 EFS trials (35.7%) place EFS in the primary endpoint set.

Haematology is the core EFS use case

Haematologic malignancies account for 133 of 221 EFS trials (60.2%) and have an EFS rate of 12.4% across the oncology dataset. Breast and genitourinary cancers follow at 22 EFS trials each.

EFS rate within major oncology disease clusters
Haematologic malignancies
12.4%
133 / 1,073 trials
Breast cancer
9.4%
22 / 234 trials
Genitourinary cancer
9.1%
22 / 243 trials
Sarcoma / paediatric solid tumors
26.3%
15 / 57 trials
Gastrointestinal cancer
2.9%
10 / 348 trials
Melanoma / skin cancer
9.0%
6 / 67 trials
CNS tumors
4.9%
4 / 81 trials
Other oncology
1.4%
4 / 290 trials
Bars show the percentage of trials in each disease cluster that include EFS.
Most frequent disease labels among EFS trials
Disease labelEFS trialsRate
Acute myeloid leukemia32 / 7940.5%
Other oncology31 / 6334.9%
Acute lymphoblastic leukemia28 / 5352.8%
Breast cancer22 / 2947.5%
DLBCL / large B-cell lymphoma17 / 6327.0%
Bladder / urothelial cancer16 / 10115.8%
Non-small cell lung cancer14 / 3384.1%
Follicular lymphoma10 / 3627.8%
Hodgkin lymphoma10 / 6216.1%
CLL / SLL8 / 5913.6%
Disease labels are non-exclusive when trials list multiple diseases.
Interpretation

EFS is especially aligned with blood cancers where relapse, response failure, molecular persistence, and treatment failure are clinically meaningful events before overall survival matures.

EFS is enriched in paediatric and orphan oncology

EFS appears in 54 of 191 paediatric oncology trials (28.3%) versus 167 of 2,338 non-paediatric trials (7.1%). It also appears in 61 of 290 orphan-drug oncology trials (21.0%) versus 160 of 2,239 non-orphan trials (7.1%).

EFS rate by special-population flag
Paediatric oncology
28.3%
54 / 191 trials
Orphan-drug oncology
21.0%
61 / 290 trials
Non-paediatric oncology
7.1%
167 / 2,338 trials
Non-orphan oncology
7.1%
160 / 2,239 trials
Population flags are trial-level context variables.
Interpretation

The enrichment is consistent with settings where relapse-free or failure-free disease control can be more feasible and clinically interpretable than waiting for mature overall survival.

EFS appears across modalities, led by small molecules and antibodies

EFS trials most often include small molecules (176 of 221, 79.6%) and monoclonal antibodies (112 of 221, 50.7%). Bispecific antibody and ADC programs show the highest EFS penetration among major modalities: 14.6% and 13.9%, respectively.

EFS rate within major drug modalities
Small molecule
10.9%
176 / 1,614 trials
Monoclonal antibody
10.6%
112 / 1,054 trials
ADC
13.9%
40 / 288 trials
Bispecific antibody
14.6%
23 / 157 trials
Peptide/protein/enzyme
9.5%
18 / 190 trials
Cell therapy
13.4%
16 / 119 trials
Modalities are non-exclusive because trials may include combination regimens.
Interpretation

EFS is not tied to one therapeutic technology. Its higher rate in bispecifics, ADCs, and cell therapy reflects use in immuno-oncology and haematology settings where event definitions can capture durable control before survival readout.

EFS definitions most often anchor on death, relapse, and progression

Across 221 EFS trials, death appears in 118 EFS definitions (53.4%), relapse or recurrence in 100 (45.2%), and progression in 85 (38.5%).

Components named in EFS endpoint wording
Death
53.4%
118 / 221 trials
Progression
38.5%
85 / 221 trials
Relapse / recurrence
45.2%
100 / 221 trials
Treatment failure / non-response
13.6%
30 / 221 trials
Next therapy
1.4%
3 / 221 trials
Secondary malignancy
5.4%
12 / 221 trials
MRD event
9.5%
21 / 221 trials
Surgery / operability
14.9%
33 / 221 trials
Toxicity / discontinuation
13.6%
30 / 221 trials
Components are not mutually exclusive; one EFS definition may include several event types.
Interpretation

The dataset shows EFS as a composite endpoint rather than a single standardized event construct. Death, relapse, and progression dominate, but surgery-related failure, treatment failure, toxicity, MRD, and secondary malignancy appear in narrower clinical contexts.

Only one quarter of EFS trials specify a fixed time window

A fixed EFS time window appears in 50 of 221 trials (22.6%). The most common windows are two-year EFS (20 trials, 9.0%), three-year EFS (15 trials, 6.8%), and one-year or 12-month EFS (14 trials, 6.3%).

Explicit time windows in EFS endpoint wording
1-year / 12-month
6.3%
14 / 221 trials
2-year / 24-month
9.0%
20 / 221 trials
3-year / 36-month
6.8%
15 / 221 trials
5-year
5.0%
11 / 221 trials
10-year
0.5%
1 / 221 trials
Percentages use 221 EFS trials as denominator.
Interpretation

Most EFS endpoints are framed as time-to-event endpoints without a named landmark year. Landmark EFS is present, but it is a minority pattern and appears mainly where relapse prevention or early treatment failure is operationalized at defined follow-up points.

EFS assessment relies more on clinical event framing than named criteria

Imaging-related wording appears in 91 of 221 EFS trials (41.2%). Explicit investigator assessment appears in 27 (12.2%), independent review in 15 (6.8%), and RECIST in 12 (5.4%).

Assessment or criteria references in EFS endpoints
PET / imaging-linked
41.2%
91 / 221 trials
Investigator-assessed
12.2%
27 / 221 trials
Independent review
6.8%
15 / 221 trials
RECIST
5.4%
12 / 221 trials
MRD-linked
8.1%
18 / 221 trials
Lugano
2.3%
5 / 221 trials
ELN
2.3%
5 / 221 trials
STEEP
0.9%
2 / 221 trials
Assessment categories are not mutually exclusive.
Interpretation

Unlike progression-free survival, EFS wording often depends on protocol-defined clinical events rather than a single cross-tumor measurement standard. Imaging is common, but named criteria such as RECIST, Lugano, ELN, or STEEP appear in smaller subsets.

EFS travels with OS, safety, response, and biomarker endpoints

Among EFS trials, overall survival (OS) appears in 201 of 221 (91.0%), safety or adverse-event endpoints in 186 (84.2%), and response endpoints in 173 (78.3%).

Other endpoint families appearing in EFS trials
Overall survival (OS)
91.0%
201 / 221 trials
Safety / adverse events
84.2%
186 / 221 trials
Response endpoint
78.3%
173 / 221 trials
Progression-free survival (PFS)
34.4%
76 / 221 trials
Biomarker / genomic analysis
29.4%
65 / 221 trials
MRD / molecular response
31.2%
69 / 221 trials
Duration of response (DOR)
22.6%
50 / 221 trials
PK / immunogenicity
25.3%
56 / 221 trials
Endpoint families are non-exclusive and counted at trial level.
Interpretation

EFS rarely stands alone. It is usually embedded in a broader endpoint architecture that preserves OS as the survival anchor while adding response, safety, molecular, and biomarker evidence around the EFS result.

EFS trials have higher endpoint complexity

EFS trials average 11.3 total endpoints with a median of 10, compared with 7.0 total endpoints and a median of 6 in non-EFS oncology trials.

Average endpoint count by trial group
EFS trials: total endpoints
11.3%
221 / 221 trials counted as group average
Non-EFS trials: total endpoints
7.0%
2,308 / 2,308 trials counted as group average
EFS trials: secondary endpoints
9.3%
221 / 221 trials counted as group average
Non-EFS trials: secondary endpoints
5.5%
2,308 / 2,308 trials counted as group average
Bars show average endpoint count, not percentage.
Interpretation

EFS is associated with broader endpoint packages. The difference is driven mainly by secondary endpoints, consistent with EFS trials needing supporting survival, response, safety, biomarker, and patient-centered measures.

EFS trials are more often randomized and larger when EFS is primary

Among EFS trials, 128 of 221 are randomized (57.9%), compared with 1,126 of 2,529 in the overall oncology set (44.5%). Median target sample size is 136 in EFS trials versus 105 in non-EFS trials; when EFS is in the primary endpoint set, the median rises to 278.

Design features in EFS trials
Randomised
57.9%
128 / 221 trials
Open label
51.1%
113 / 221 trials
Biomarker-stratified
24.0%
53 / 221 trials
Adaptive
19.0%
42 / 221 trials
Dose escalation / combined-dose
15.4%
34 / 221 trials
Real-world control
1.4%
3 / 221 trials
Design features counted at trial level across 221 EFS trials.
Median target sample size by endpoint role
EFS primary endpoint set
278.0%
79 / 79 trials median target sample size
Any EFS trial
136.0%
215 / 215 trials median target sample size
EFS secondary-only
85.0%
142 / 142 trials median target sample size
Non-EFS oncology trials
105.0%
2,277 / 2,277 trials median target sample size
Bars show median target sample size, not percentage.
Interpretation

EFS is operationally heavier when it becomes decision-bearing. Primary-EFS trials have a median target sample size more than three times higher than secondary-only EFS trials, indicating a shift from supportive efficacy characterization to confirmatory endpoint strategy.

Definitions appendix

EFS means event-free survival. OS means overall survival. PFS means progression-free survival. MRD means minimal or measurable residual disease. ADC means antibody-drug conjugate. RECIST, Lugano, ELN, and STEEP are endpoint or response-assessment frameworks used in selected oncology settings.