Is DFS an Appropriate Surrogate for OS in Early HER2+ Breast Cancer?

Disease-free survival has a reasonably high correlation with overall survival in trials of patients with early HER2-positive breast cancer treated with adjuvant trastuzumab, according to a new meta-analysis. The findings suggest that disease-free survival (DFS) can continue to be used as a surrogate endpoint in this field.

“The development of novel adjuvant regimens is a lengthy process, and the analysis of overall survival requires a long follow-up,” wrote study authors led by Everardo D. Saad, MD, of the International Drug Development Institute in Louvain-La-Neuve, Belgium. “Although DFS has often been used as the primary endpoint in adjuvant trials of breast cancer, to our knowledge it has not been formally validated in this setting.”

The researchers conducted a review and meta-analysis evaluating the role of DFS as a surrogate for overall survival (OS) in the adjuvant treatment of HER2-positive breast cancer. They investigated both patient-level associations between those outcomes, meaning whether a patient with prolonged DFS would also expect to have prolonged OS, and trial-level associations, meaning whether treatment-induced changes in the surrogate endpoint are mirrored by such changes in the final endpoint. The results of the analysis were published in Lancet Oncology.

The study included a total of 8 trials encompassing 21,480 patients; the trials included 12 contrasts between varying therapies, including between chemotherapy or observation plus trastuzumab, between 12 months and a shorter course of trastuzumab, or in one case, between lapatinib and that agent plus trastuzumab.

The patient-level associations between DFS and OS were “strong,” according to the authors. The correlation coefficient for this measure was 0.90 (95% CI, 0.89–0.90). For trial-level associations, the coefficient of determination (R²) was 0.75 (95% CI, 0.50–1.00). A post-hoc reduced set analysis that excluded one trial due to methodological limitations yielded an R² of 0.84 (95% CI, 0.67–1.00).

The authors noted that their analysis is limited by the reliance on the individual trials included and some of their methodological differences. Still, they wrote that “our results suggest that DFS might have good overall statistical validity to be used as a surrogate for OS in the adjuvant treatment of HER2-positive early breast cancer.”

In an accompanying editorial, Eitan Amir, PhD, of the Princess Margaret Cancer Centre and the University of Toronto, wrote that the inclusion of the patient- and trial-level analyses is a “major strength of the analysis and provides robust support for the adequate surrogacy of DFS and OS in HER2-positive early-stage breast cancer.”

He noted that it remains to be seen whether the data from this study can be extrapolated to other settings. For example, triple-negative breast cancer is associated with DFS event rates similar to those seen in untreated HER2-positive breast cancer, but “the magnitude of relative treatment effect in this subgroup is smaller.” In other settings, lower event rates might result in a poorer performance of the surrogate endpoint.

“Attempts to validate DFS as a surrogate for OS should be pursued in a broader group of breast cancers while ensuring ... that assessment includes not only the strength and consistency of correlation between the surrogate and definitive endpoints at the trial level, but also the prediction of the net effect of treatment at a patient level,” Amir wrote.