Even though overall survival is considered to be the "gold standard" for determining the effectiveness of a treatment, in the case of adjuvant cytotoxic agents for gastric cancer, disease-free survival could be an acceptable surrogate end point, according to an international group of researchers.
The team, led by Koji Oba, PhD, from Hokkaido University Hospital in Sapporo, Japan, conducted a meta-analysis of 14 randomized clinical trials comparing adjuvant chemotherapy with surgery alone in patients with curatively resected gastric cancer. There were more than 3000 patients in the cohort.
The meta-analysis was published in the November 6 issue of the Journal of the National Cancer Institute.
The team found that a large proportion of relapses occurred before 3 years, and that there was a strong correlation between overall and disease-free survival at the individual and trial levels.
There was a high individual-level association between disease-free and overall survival (Spearman's rank correlation coefficient, 0.974; 95% confidence interval [CI], 0.971 - 0.976). In individual patients, this indicates that disease-free survival is highly predictive of overall survival, the researchers report. In an external validation, the 6 hazard ratios for overall survival that were predicted on the basis of disease-free survival "were in very good agreement" with the actual ratios observed for overall survival.
This strong correlation between disease-free and overall survival can, at least in part, be attributed to the short time span from relapse to death (median, <12 16="" a="" addition="" all="" analyzed="" and="" died="" disease-free="" documented="" explain.="" for="" identical.="" in="" included="" months="" of="" overall="" p="" patients="" relapse="" survival="" they="" thus="" trials="" were="" without="">
"Similar results have led to the adoption of the 3-year disease-free survival as a surrogate for 5-year overall survival in evaluating new treatments for resectable colon cancer," write Dr. Oba and colleagues. "Our results are based on fewer trials and smaller sample sizes, but they include a broader range of treatment options."
On the basis of these results, Dr. Oba and his team believe that disease-free survival should be a primary end point for gastric cancer studies. "Since we only investigated cytotoxic agents, we need to continue the validation based on future trials with agents that have different mechanisms of actions, such as targeted therapy," he told Medscape Medical News.
These findings could primarily influence the design of future trials. "If studies are conducted and approved based on disease-free survival, the new treatment can be used sooner than if they are based on overall survival," he said. "We also evaluated the correlation between disease-free and overall survival and found that a high correlation exists. This will enable us to use disease-free survival as a prognostic factor in the management of patients, which has already been done in current treatment practice."
But it is unclear at this point whether the model of a disease-free survival surrogate end point can be extrapolated to other cancer studies. Dr. Oba explained that it depends on many factors. "There have been so many investigations about the surrogacy against overall survival," he said. "For example, disease-free survival is regarded as an acceptable surrogate for overall survival in the adjuvant setting for colon cancer."
Statistical Evidence Not Sufficient
This meta-analysis demonstrates the "impressively close concordance" of treatment on disease-free and overall survival in gastric cancer, writes Colin B. Begg, PhD, chair of the Department of Epidemiology and Biostatistics at the Memorial Sloan-Kettering Cancer Center in New York City, in an accompanying editorial.
"These results exemplify the fact that in the adjuvant setting disease recurrence is a pivotal event in the natural history of the disease for most solid tumors," writes Dr. Begg.
Because the rank correlation between disease-free and overall survival was very high, the evidence is "indeed strong that trials that use disease-free survival as the primary end point" will provide conclusions that can reliably predict the effect on overall survival, he notes.
These studies are heavily focused on the various statistical comparisons, but are "light" when it comes to certain aspects of the clinical setting, which "might help to influence a conclusion that is ultimately a judgment call," Dr. Begg writes. "In studies of this nature, there is no accepted statistical test or definitive statistical standard that a surrogate end point must meet to justify its use in future studies."
Statistical evidence alone isn't sufficient. For instance, "How valuable is delay of progression to the quality of life of patients generally?" he asks. Information is also needed on the amount of variation that exists in the techniques used to measure progression and on clinician confidence in the available methods.
"Decision making in this context cannot be relegated to purely statistical comparisons without substantive judgment of this nature," Dr. Begg concludes.
Can Reduce Trial Duration
Although 5-year overall survival is generally the most common metric for judging the success of a particular therapeutic intervention, Dr. Oba and colleagues point out that its main disadvantage is that it requires an extended follow-up period. Additionally, its measurement can be diluted by other causes of death and by therapies for recurrent/advanced disease.
They hypothesized that if disease-free survival replaced overall survival in the evaluation of new therapies for gastric cancer, both trial duration and cost would be reduced.
Their findings suggest that an early measurement of disease-free survival, such as at 2 years, might be too soon to make an accurate prediction of overall survival at 5 years. Few disease-free survival events are available at very early time points, which could result in imprecise predictions, the researchers note. However, very late time points are of less use because they are closer to the final end point.
An analysis at 3 or 4 years would probably reduce the overall duration of a trial by about 15% to 30%, if the accrual was short enough, they add.
Dr. Oba and colleagues point out that the use of disease-free survival as a surrogate can be used independent of the geography of the trial, even though there is a "large heterogeneity about the prognosis between Asian and non-Asian patients."
"The relationship between the hazard ratio for overall survival and the hazard ratio for disease-free survival were clearly consistent throughout all trials" included in the meta-analysis, they add.
This work was partially supported by the French Institut National du Cancer, the Clinical Research Support Unit, and the Epidemiological and Clinical Research Information Network. The study authors and Dr. Begg have disclosed no relevant financial relationships.
Δεν υπάρχουν σχόλια:
Δημοσίευση σχολίου