Moreover, participants in the Women’s Health Initiative had a median age of 63 years, which was much older than the mean age of 38 years for the BRCA carriers studied. The former "are women who had gone through their whole natural life with estrogen and then [had taken] more, so potentially, it’s not relevant to this population of patients."
Discussant Dr. Lynn Hartmann, an oncologist at the Mayo Clinic in Rochester, Minn., cautioned about the pitfall of unknown biases in observational studies. "I can tell you from participating in cohort studies myself that there are biases that one cannot even imagine that can seep into your study sets," she said.
In the PROSE study, the types of cancers resulting from a BRCA mutation in a family might have influenced which women underwent oophorectomy. And a woman’s breast history (for example, atypia) might have influenced whether her physician offered HRT after oophorectomy.
Dr. Hartmann commended the investigators for developing a large, multi-institutional registry; conducting a high-quality study; and addressing an important, relevant clinical question.
"But I think we do have to have some skepticism when treatment questions are tried to be answered from these types of [study] designs," she said. "I would at least ask the PROSE team ... to consider whether or not they could move into prospective clinical trials with their cohorts."
Dr. Domchek and Dr. Hartmann reported that they had no relevant conflicts of interest.