Phase II trial use of historical data often flawed
NEW YORK (Reuters Health) - While many phase II trials of cancer agents use historical data to establish null response rates, new research now suggests that these trials often do not cite a source for the historical data or provide a clear single historical estimate.
"The choice of the target response rate, (which is often based on historical data), is a key aspect of phase II design," lead author Dr. Andrew J. Vickers, from Memorial Sloan-Kettering Cancer Center in New York, and colleagues note. "Poorly chosen targets reduce the ability of phase IIs to determine which agents or approaches should be considered for definitive phase III trials."
As they report in the February 1 issue of Clinical Cancer Research, Vickers' team conducted a systematic review of phase II trials appearing in the Journal of Clinical Oncology or Cancer between June 1, 2002 and June 1, 2005.
A total of 134 eligible trials were identified, including 70 (52%) that required historical data for their design, the report indicates. Thirty-two of these trials (46%) did not cite a source for the historical data that was used. Only 9 (13%) trials clearly provided a single historical estimate as the basis for the null response rate.
Failure to cite historical data properly was strongly associated with finding the trial agent to be active: 82% vs. 33% (p = 0.005).
None of the studies employed statistical methods to account for sampling errors or for possible differences among cases or differences between the study cohort and the historical controls.
"More appropriate use of historical data in phase II design will improve both the sensitivity and specificity of phase II for eventual phase III success, avoiding both unnecessary definitive trials of infective agents and early termination of effective drugs for lack of apparent benefit," the authors conclude.
Clin Cancer Res 2007;13:972-976.