Why This Study Matters
Cancer clinical trials are foundational to improving outcomes for people with cancer and have contributed to decreases in cancer mortality over the past two decades. They provide the highest level of evidence for treatment efficacy, informing FDA approvals and clinical practice guidelines. But who funds those trials shapes which questions get asked. Industry sponsors have traditionally focused on new drug development and regulatory approval, while federally sponsored trials — largely through the National Cancer Institute (NCI) and its cooperative trial networks — have addressed broader questions with less commercial incentive: treatment de-escalation, multimodality strategies, and care for rare and pediatric cancers.
Federal sponsorship dominated the field four decades ago. Today, the enrollment ratio of industry- to federally sponsored trials stands at roughly 9 to 1, a ratio that doubled between 2010 and 2020. Despite this shift, the substantive differences between what federal and industry sponsors actually study had not been systematically quantified — until this analysis.
Study Design
This was a retrospective, registry-based comparative study — not a clinical trial itself, but an analysis of the entire landscape of US cancer clinical trials. The authors downloaded ClinicalTrials.gov data in March 2025 for all interventional cancer trials with at least one US study site that were initiated between January 1, 2008, and December 31, 2024. The analysis followed relevant portions of the STROBE observational-study reporting guideline. No institutional review board approval was required, since the study relied entirely on publicly available, deidentified registry data.
Each trial’s lead sponsor was classified as either federal (NIH or another US federal agency, including R01-funded and cooperative group trials) or industry (a pharmaceutical or biotechnology company). Trials with lead sponsors in ClinicalTrials.gov’s heterogeneous "Other" category (individuals, academic centers, foundations, hospitals, non-US governments) were excluded as not forming a coherent comparison group. Differences in trial characteristics — study purpose, phase, intervention modality, de-escalation design, rare-cancer focus, and pediatric population — were tested with χ² tests and reported as odds ratios (ORs) with 95% confidence intervals. Temporal trends across three periods (2008–2012, 2013–2018, 2019–2024) were assessed with logistic regression interaction terms.
Study population analyzed: The dataset comprised 11,681 trials meeting inclusion criteria: 2,112 (18.1%) federally sponsored and 9,569 (81.9%) industry-sponsored. Of these, 10,588 (90.6%) were conducted for the primary purpose of cancer treatment; the remainder addressed basic science, device feasibility, diagnostics, health services research, prevention, screening, or supportive care. Because this is a registry analysis rather than a single interventional trial, conventional trial statistics such as enrollment-per-arm, number of study sites, randomization ratio, and dosing do not apply — the figures reported here reflect the scope of the trial portfolio analyzed.
Primary Comparison: Federal vs. Industry Trial Portfolios
Federally sponsored trials were significantly less likely than industry trials to be conducted for cancer treatment purposes at all (85.1% vs 91.9%; OR, 0.50; 95% CI, 0.43–0.58; P < .001), instead skewing toward nontreatment interventions: diagnostic trials (3.6% vs 2.1%; OR, 1.74; P < .001), health services research trials (1.2% vs 0.2%; OR, 7.63; P < .001), prevention trials (4.7% vs 1.1%; OR, 4.43; P < .001), screening trials (1.1% vs 0.5%; OR, 2.28; P = .002), and supportive care trials (2.5% vs 0.9%; OR, 2.92; P < .001).
| Characteristic | Federal (n=2,112) | Industry (n=9,569) | Odds Ratio (95% CI)ᵃ |
|---|---|---|---|
| Cancer treatment purpose | 85.1% | 91.9% | 0.50 (0.43–0.58) |
| Phase 1 trialsᵇ | 26.8% | 37.6% | 0.61 (0.54–0.68) |
| Phase 2 trialsᵇ | 59.4% | 42.9% | 1.95 (1.76–2.16) |
| Phase 3 trialsᵇ | 13.1% | 17.7% | 0.70 (0.60–0.81) |
| Single-agent drug trials | 48.6% | 77.4% | 0.28 (0.25–0.31) |
| Drug/biologic + radiotherapy | 10.5% | 1.4% | 8.22 (6.51–10.37) |
| Drug/biologic + surgery | 2.5% | 0.2% | 14.09 (7.95–24.98) |
| Drug/biologic + transplant | 1.4% | 0.1% | 10.32 (5.18–20.58) |
ᵃ Odds ratios describe the federal share relative to industry. ᵇ Phase percentages are calculated among the 10,588 treatment-purpose trials only.
Temporal Trends, 2008–2024
This study analyzes trial-level registry data rather than patient-level data, so it does not contain patient subgroup analyses (for example, by biomarker or demographic status) in the way a clinical trial report would. The closest analog reported in the source is a stratified analysis of how sponsor-type differences evolved across three time periods: 2008–2012, 2013–2018, and 2019–2024.
| Characteristic | Trend | Interaction P Value |
|---|---|---|
| Nontreatment intervention studies (any) | Increasingly federally sponsored | P = .03 |
| Basic science purpose | Increasingly federally sponsored | P = .04 |
| Prevention purpose | Increasingly federally sponsored | P = .04 |
| Cancer treatment purpose | Increasingly industry-sponsored | P = .002 |
| Phase 3 trials (among treatment trials) | Increasingly federally sponsored | P = .03 |
| Multimodality trials | Increasingly federally sponsored | P = .01 |
| De-escalation trials | Increasingly federally sponsored | P = .002 |
| Single-agent drug trials | Increasingly industry-sponsored | P = .01 |
| Phase 1 trials | Increasingly industry-sponsored | P = .009 |
| Rare-cancer trials | Increasingly industry-sponsored | P = .02 |
The federal-vs-industry gap in phase 3 trial share had narrowed to the point of no longer being statistically significant by 2019–2024 (16.2% vs 17.9%; OR, 0.89; 95% CI, 0.69–1.14; P = .35). A sensitivity analysis using alternative sponsor-classification rules (crediting primary or secondary sponsor roles) reversed the direction of one finding: biologic-only trials were more common among industry in the base-case analysis (8.5% vs 7.1%; P = .04) but more common with federal involvement under the sensitivity analysis (7.9% vs 6.8%; P = .01) — a reminder that this single comparison was sensitive to classification method, even as the study’s other findings were described as largely consistent across approaches.
Safety Profile
Not applicable to this study. This is a registry-based comparative analysis of clinical trial design characteristics — it does not enroll or treat patients and therefore reports no adverse event data, toxicity rates, or safety endpoints. Readers looking for the safety profile of a specific cancer therapy should consult the primary trial reports for that treatment.
This study found that federally sponsored and industry-sponsored cancer clinical trials differed substantially in scope and purpose.
— Discussion, Federal and Industry Sponsorship in US Cancer Clinical Trials, JAMA Oncology (2026)
Interpretation and Broader Context
The authors frame federal and industry sponsorship as complementary: federally supported research contributes to basic science discovery, clinical trial infrastructure, and methodologic innovation, while industry plays the central role in therapeutic development and regulatory approval. Notable federally led, de-escalation-driven trials cited by the authors as establishing standards of care include TAILORx, RxPONDER, and PROSPECT — studies that showed selected patients could safely avoid chemotherapy or radiotherapy without compromising outcomes.
But the data also raise a policy concern. Between 2018 and 2022, industry sponsorship accounted for roughly 9 of 10 enrollments among adult patients with cancer. Because the pool of patients willing to enroll in trials is finite, the authors note this imbalance raises the risk that research questions without commercial incentive — de-escalation strategies, multimodality regimens, rare and pediatric cancers — may go increasingly unaddressed as the federal share of the portfolio continues to shrink. The authors also point to an interdependency: counties with NCI-funded trial sites were nine times more likely to also host industry-sponsored trials between 2015 and 2024, suggesting federally supported infrastructure underpins industry-run research as well.
Limitations
The authors note the study relied on registry coding, which may misclassify trial characteristics; it did not fully account for multisponsor trials, potentially obscuring collaborative partnerships; and it examined differences in trial portfolios at the point of initiation rather than downstream effects on clinical practice or patient outcomes. Because this is a retrospective, non-randomized comparison of registered trial characteristics rather than a study of trial conduct or results, it cannot say whether the observed sponsorship differences translate into differences in trial quality, completion, or patient benefit. The reversal of the biologic-only finding under an alternative sponsor-classification rule also shows that at least one comparison in this analysis is sensitive to methodological choices the authors themselves flag as open to interpretation.
Funding & Disclosures
Funding: Public Health Sciences Division, Fred Hutchinson Cancer Center; the Hope Foundation for Cancer Research. The funding organizations had no role in study design, conduct, analysis, or the decision to submit the manuscript for publication, per the article’s disclosure statement. This analysis is not itself a registered interventional trial and carries no ClinicalTrials.gov identifier.
Future Directions
The authors call for continued attention to the balance of sponsor support in cancer research, framing federally supported infrastructure as a platform that also enables industry-sponsored research at the same sites. They suggest their portfolio-level findings provide an empirical basis for evaluating the complementary roles of federally sponsored and industry-sponsored research as funding priorities evolve.
Among 11,681 US interventional cancer trials initiated from 2008 to 2024, industry now sponsors nearly 5 times as many trials as the federal government, and that gap has widened over time. Federally sponsored trials disproportionately pursue de-escalation designs and rare and pediatric cancer research — questions with less commercial incentive — while industry trials concentrate on single-agent drug development. The analysis is a registry-level, retrospective comparison of trial characteristics at initiation, not of trial conduct or outcomes, and at least one of its findings reversed under an alternative classification method, so its conclusions describe what gets studied and by whom rather than what that difference means for patient outcomes.