Across 38 CTIS-authorized Phase III lung cancer trials in 2025, the dominant design pattern is randomized, multi-country, combination-heavy development. Randomization appears in 33 of 38 trials (86.8%), combination treatment in 23 of 38 (60.5%), and CRO support in 26 of 38 (68.4%). Among trials with primary endpoint data, overall survival and progression-free survival remain the central efficacy anchors, appearing in 19 of 35 (54.3%) and 18 of 35 (51.4%) trials respectively.
Of 38 trials, 33 were randomized (86.8%), 18 were open-label (47.4%), 14 were biomarker-stratified (36.8%), and 4 used an adaptive element (10.5%). No trial in this cohort used a real-world control or crossover design.
The cohort is operationally mature: most trials are randomized and controlled, but the 47.4% open-label rate shows how often Phase III lung cancer protocols still accept treatment-visibility tradeoffs, likely reflecting complex combinations, route differences, or active standard-of-care comparators.
Among 35 trials with primary endpoint data, overall survival (OS) appeared in 19 trials (54.3%) and progression-free survival (PFS) in 18 trials (51.4%). Blinded independent central review or blinded independent review committee assessment appeared in 13 of 35 trials (37.1%), while disease-free survival (DFS), objective response rate (ORR), and safety/adverse-event endpoints each appeared in 4 of 35 trials (11.4%).
Phase III lung cancer trial design remains centered on hard survival and time-to-event efficacy endpoints. The 37.1% central-review signal shows that imaging adjudication is a major operational dependency, especially when PFS is used as a primary endpoint.
Among 30 trials with secondary endpoint data, safety/adverse events appeared in 25 trials (83.3%), PFS in 21 trials (70.0%), duration of response in 21 trials (70.0%), OS in 17 trials (56.7%), quality of life or patient-reported outcomes in 17 trials (56.7%), and ORR in 16 trials (53.3%). The median secondary endpoint count was 5 across all 38 trials and 7.4 on average among trials with secondary endpoint data.
Secondary endpoint design is more operationally complex than the primary endpoint layer. Trial teams should expect heavy safety, imaging, response-duration, survival follow-up, and patient-reported outcome workloads even when the headline primary endpoint is a single survival measure.
Biomarker-stratified design appeared in 14 of 38 trials (36.8%). Across trial records, PD-L1 appeared in 18 of 38 trials (47.4%), EGFR in 17 of 38 (44.7%), ROS1 in 6 of 38 (15.8%), and KRAS G12C in 4 of 38 (10.5%). Treatment modalities were multi-label: small molecules appeared in 28 of 38 trials (73.7%), monoclonal antibodies in 23 of 38 (60.5%), antibody-drug conjugates in 7 of 38 (18.4%), and bispecific antibodies in 6 of 38 (15.8%).
| Signal | Trials | Share |
|---|---|---|
| Small molecule | 28/38 | 73.7% |
| Monoclonal antibody | 23/38 | 60.5% |
| Combination treatment | 23/38 | 60.5% |
| PD-L1 signal in trial records | 18/38 | 47.4% |
| EGFR signal in trial records | 17/38 | 44.7% |
| ADC | 7/38 | 18.4% |
| Bispecific antibody | 6/38 | 15.8% |
The 2025 Phase III lung cancer landscape is not simply immunotherapy or chemotherapy. It is a layered targeted-treatment environment where small molecules, antibodies, ADCs, biomarker-defined subgroups, and combination protocols coexist.
The 38 trials planned 13,741 participants in total, with a median planned sample size of 381 participants per trial. The median recruitment window was 56.5 months, with a range from 17 to 121 months. Digital or remote recruitment appeared in 14 of 38 trials (36.8%), while registry or advocacy recruitment appeared in 5 of 38 trials (13.2%).
The median 56.5-month recruitment window shows that Phase III lung cancer studies remain long-cycle execution programs. Digital recruitment is meaningful but not universal, suggesting many sponsors still depend heavily on site referral networks and country-specific recruitment materials.
Across country-level site counts, Spain contributed 331 sites, France 277, Italy 259, Germany 250, and Poland 146. Together, these five countries accounted for 1,263 of 1,887 country-level sites (66.9%). Trial participation was also broad: Spain appeared in 28 of 38 trials (73.7%), France and Italy in 26 each (68.4%), Poland in 25 (65.8%), and Germany in 23 (60.5%).
For European Phase III lung cancer planning, Spain, France, Italy and Germany are the practical site-density core. Poland adds breadth and appears in nearly two-thirds of trials, but with fewer total sites than the four largest site countries.
The most recurrent site was Hospital Universitari Vall D Hebron, appearing in 16 of 38 trials (42.1%). Institut Catala D'oncologia appeared in 14 trials (36.8%), Hospital Universitario 12 De Octubre in 13 trials (34.2%), and both Institute Of Oncology Prof. Dr. Ion Chiricuta Cluj-Napoca and Athens Medical Center S.A. appeared in 12 trials each (31.6%).
| Site | Trials | Share |
|---|---|---|
| Hospital Universitari Vall D Hebron | 16 | 42.1% |
| Institut Catala D'oncologia | 14 | 36.8% |
| Hospital Universitario 12 De Octubre | 13 | 34.2% |
| Institute Of Oncology Prof. Dr. Ion Chiricuta Cluj-Napoca | 12 | 31.6% |
| Athens Medical Center S.A. | 12 | 31.6% |
A small group of oncology centers repeatedly appears across 2025 Phase III lung cancer trials. For sponsors, these sites likely represent high-demand partners where feasibility, contracting speed, and competing-trial burden matter as much as scientific fit.
CRO support appeared in 26 of 38 trials (68.4%). Sponsors used a median of 8 third parties per trial, with an average of 8.4 and a maximum of 27. The most frequent sponsor countries were the United States with 12 of 38 trials (31.6%), followed by Sweden, France and Ireland with 4 trials each (10.5% each).
| Most recurrent CRO / vendor names | Trials |
|---|---|
| Icon Clinical Research Limited | 7 |
| Fortrea Inc. | 5 |
| PPD Development LP | 5 |
| Medidata Solutions Inc. | 5 |
| WCG Clinical Inc. | 5 |
Phase III lung cancer execution is vendor-intensive. The typical trial is not just a sponsor-site relationship; it depends on CRO, imaging, lab, recruitment, ePRO/eCOA, pharmacovigilance, and operational technology layers.
OS means overall survival. PFS means progression-free survival. DFS means disease-free survival. ORR means objective response rate. DoR means duration of response. PRO means patient-reported outcome. BICR means blinded independent central review; BIRC means blinded independent review committee. CTIS means Clinical Trials Information System.