A clinical trial simulation study of a new pediatric phase I trial methodology: The rolling six design

9564

Background: Phase I cancer trials have traditionally used the 3+3 study design in which cohorts of three patients at a time are enrolled. The study is suspended to accrual with every 3rd patient, and again suspended to replace inevaluable patients. As the pediatric phase I starting dose is based on adult phase I safety data, we hypothesized that pediatric study timelines could be shortened without increasing risk using a novel rolling six design, in which up to 6 patients are concurrently enrolled per dose level. Methods: A trial simulation approach comparing the standard 3+3 design to the rolling six design was performed. Discrete time events (accrual/enrollment, evaluation and/or time to dose limiting toxicity (DLT) or inevaluability) with outcome probabilities (DLT/inevaluability) assigned to each subject based on historical data from phase I pediatric oncology trials. The probability of DLT was a function of dose level and varied distributions were used to assign discrete time elements. Metrics for study efficiency (time to complete study, total number of patients required) were determined for each design. Simulated trials and decision rule logic were coded using PC/Windows SAS v9.1. Results: The rolling six design outperformed the 3+3 design for the most relevant metrics. Average (±SD) time to study completion was 373±90 days vs. 431±98 days and the total number of patients required was 20±5 vs. 17±4 for the rolling six vs. 3+3 designs, respectively. There was no difference in the number of DLTs per study (3±1) between designs. Conclusions: The rolling six design is predicted to shorten pediatric phase 1 study timelines and will be tested prospectively in upcoming COG phase I trials.

No significant financial relationships to disclose.

Increased proteasome degradation of cyclin-dependent kinase inhibitor p27 is associated with a decreased overall survival in mantle cell lymphoma

Mantle cell lymphoma (MCL) is an aggressive neoplasm characterized by the deregulated expression of cyclin D1 by t(11;14). The molecular mechanisms responsible for MCL's clinical behavior remain unclear. The authors have investigated the expression of p53, E2F-1, and the CDK inhibitors p27 and p21 in 110 MCLs, relating their expression to proliferative activity (Ki-67). For comparison, they have similarly analyzed low-grade (12 MALT, 16 CLL/SLL) and high-grade (19 DLCL) lymphomas. p53 was detected more frequently in large-cell MCL (l-MCL; 5 of 7) than in classical MCL (s-MCL; 13 of 103) and DLCL (8 of 19). In MCL and DLCL, the percentage of E2F-1+ nuclei was high, correlating with high Ki-67 expression. Most MCLs (91 of 112) and DLCLs (12 of 19) showed a loss of p27; MALT and CLL/SLL, however, were p27 positive. Reverse transcription-polymerase chain reaction and in vitro protein degradation assays demonstrated that MCLs have normal p27 mRNA expression but increased p27 protein degradation activity via the proteasome pathway. Correlation of MCL p53 and p27 expression with clinical data showed an association between reduced overall survival rates and the overexpression of p53 (P =.001), the loss of p27 (P =. 002), or both. Loss of p27 identified patients with a worse clinical outcome among p53 negative cases (P =.002). These findings demonstrated that MCL has a distinct cell cycle protein expression similar to that of high-grade lymphoma. The loss of p27 and the overexpression of p53 in MCL are prognostic markers that identify patients at high risk. The demonstration that low levels of p27 in MCL result from enhanced proteasome-mediated degradation should encourage additional clinical trials. (Blood. 2000;95:619-626) (Blood. 2000;95:619-626)