Quality of contemporary clinical trials in neuroendocrine tumors.

Abstract

e14617 Background: The heterogeneity of neuroendocrine tumors (NETs) makes the interpretation of clinical trials for this disease group challenging. Our aim was to review the quality of phase II and III trials in NETs to inform design of future studies in this area. Methods: We identified studies by conducting a computerized search of MEDLINE. We considered all phase II or III trials of systemic antineoplastic treatments for NETs that were written in English and published between 2000-2010. Information on trial design, description of the study population, primary and secondary endpoints, statistical considerations and results was abstracted from each article using a standardized form. Results: Six phase III and 34 Phase II trials were identified. The majority were single-arm studies (78%). The make-up of the study population was variable: only 22% of trials included patients with one type of tumor (islet cell carcinoma or carcinoid tumor), 43% of trials included patients with both tumor types and 35% of trials included others endocrine cancers. Disease progression at baseline was often not reported (50%) and was present for all patients in only 20% of trials; status of functioning tumor, tumour differentiation and Ki67 index were reported in 35%, 37% and 12% of trials, respectively. The primary end point (PEP) was clearly defined in 75% and was standard (OR, TTP, PFS by WHO or RECIST criteria) in 50% of trials. An identifiable statistical design and predefined sample size were reported in 62% and 65% of trials, respectively. Toxicities were clearly defined and reported in 65% of trials. Interpretation of study results in concordance with predefined criteria described in the Methods section was poor – in 45% of trials the authors’ interpretation was discordant. Among endpoints, OR, PFS or TTP, OS, Biochemical response (CgA), and clinical response were respectively reported in 97%, 77%, 75%, 37%, and 32% of trials. Conflict of interests and study sponsorship were reported in 42% and 55% of trials. Conclusions: The quality of the design and reporting of phase II/III NETs trials is poor. Future trials should focus on more homogenous patient populations and adhere to rigorous definition, reporting and interpretation of population, and trial parameters as well as results.