Reporting of Clinical Trials
This is an interesting editorial from Journal of Clinical Oncology about the right time to report the results of clinical trials for adjuvant therapy. I am quite interested in the clinical trial process and usually read everything that crosses my desk (or my computer screen), this was a perspective that I had not considered. You, too, may find it interesting.
Here is the introduction and a link:
Reporting of Adjuvant Breast Cancer Trials: When Is the Right Time?
Helena M. Earl
The article by Joensuu et al published in Journal of Clinical
Oncology (JCO) is an example of the timely reporting of an adjuvant breast cancer trial adding oral capecitabine to standard chemotherapy.
Although there is no effect in the whole trial population,
subgroup analysis demonstrates a statistically significant benefit for the addition of capecitabine in triple-negative patients, an interesting result first presented at the American Society of Clinical Oncology Annual Meeting in Chicago, IL, in 2010.However, this is not the first time the Finland Capecitabine Trial, also known as FinXX, has
been presented. It was reported in full in Lancet Oncology in 2009,and at that point, capecitabine showed a benefit for the whole population.
How can the result of a trial change? Has it really changed, or was the earlier report simply premature? The most likely conclusion is the latter, which raises questions about the right time to report trial results.
As a lead clinician on adjuvant and neoadjuvant breast cancer
trials over the past 15 years I have slowly and rather reluctantly (on my part) been schooled in patient perseverance, not only with regard to what is required to complete large trials, but also in terms of waiting for that all-important first analysis. Gregory et al published on the dangers of early reporting in a fascinating analysis of adjuvant breast cancer data and demonstrated the misleading results of early and repeated analyses. Only as time passed and follow-up matured did the survival curves become stable and provide a robust estimate of progression-free (PFS) and overall survival
(OS). Is that what has happened here? Were the initial positive data simply the results of a few random events overly influencing the curves?