Pharmaceutical involvement in phase II breast cancer clinical trials.
6104 Background: Pharmaceutical industry (PHI) involvement in breast cancer clinical trials is increasing and may correlate with positive outcomes. We sought to further investigate the prevalence and potential impact of PHI involvement in breast cancer research through evaluation of recently published breast cancer phase II trials (BP2T). METHODS: Phase II clinical trials from June 2005 - June 2010 were identified in MEDLINE by search terms "breast cancer" and "phase II" and limits of English language, human subjects, and "clinical trials". Trials were abstracted by 2 independent reviewers for industry involvement, report of possible conflicts of interest (COI), trial design, setting, interventions, accrual, and outcomes. PHI trials were defined by any report of industry involvement. "Positive" outcomes was defined by assessment of authors conclusions. Descriptive statistics and Fisher's exact test are reported. RESULTS: 298 eligible trials were abstracted. 181 (61%) reported PHI including 44 (15%) with industry authorship. For 114 (38%) trials a COI was identified. COI was reported more frequently among PHI trials vs. non-PHI trials (58% vs. 7%, p < 0.001 ). Among PHI trials, 132 (73%) involved metastatic disease, 44 (24%) neoadjuvant, and 9 (5%) other settings. 125 (69%) PHI trials were single arm, 29 (16%) randomized, 15 (9%) Phase I/II, 13 (7%) nonrandomized with multiple cohorts. PHI trials were more likely to have US sites (78% vs. 52%, p < 0.001), to include > 50 patients (54% vs. 38%, p = 0.006), to evaluate biologic therapy (38% vs. 25%, p = 0.04) and to include correlative science (34% vs. 17%, p = 0.002) compared to non-PHI trials. Average time to accrual was 27 months for PHI vs. 30 months for non-PHI trials. 140 (77%) PHI trials and 93 (79%) non-PHI trials were positive (p = 0.8). CONCLUSIONS: The majority of BP2T involve pharmaceutical industry support, and many include an author with COI. The prevalence of PHI support and frequency of COI highlights the importance of managing COI while supporting academia/industry collaboration. PHI support correlates with some aspects of study design and study characteristics, but correlation with positive outcomes as noted in some prior studies is not seen among this more homogenous dataset.
Duke Scholars
Published In
EISSN
Publication Date
Volume
Issue
Start / End Page
Location
Related Subject Headings
- Oncology & Carcinogenesis
- 1112 Oncology and Carcinogenesis
- 1103 Clinical Sciences
Citation
Published In
EISSN
Publication Date
Volume
Issue
Start / End Page
Location
Related Subject Headings
- Oncology & Carcinogenesis
- 1112 Oncology and Carcinogenesis
- 1103 Clinical Sciences