To the Editor: We commend Akira Kugaya, M.D., Ph.D., and colleagues (1) on their brief report on serotonin 2A (5-HT_2A) receptors in postmenopausal women. Characterization of the effects of gonadal steroids on neuroreceptor pharmacology and cognitive function holds great importance to clinical care and cognitive neuroscience research. Consistent with the article by Dr. Kugaya et al., we previously reported that 5-HT_2A receptor binding potential significantly increased in postmenopausal women during administration of transdermal estradiol, 0.1 mg/day for 8 to 14 weeks, followed by combined transdermal estradiol and progesterone, 100 mg b.i.d. (2 to 6 weeks), with positron emission tomography and the selective 5-HT_2A receptor radioligand [¹⁸F]altanserin. Dr. Kugaya et al. correctly cited this publication as reporting a "subthreshold" effect of estradiol alone on increasing 5-HT_2A receptor binding potential in specific brain regions of interest. However, we alert readers to our more recent publication (2), in which our original image data were analyzed voxel by voxel using statistical parametric mapping, a technique similar to that applied by Dr. Kugaya et al. Consistent with the work of Dr. Kugaya et al., this approach showed that administration of estradiol alone increased 5-HT_2A receptor binding potential in multiple brain regions that had not been examined in our initial region-of-interest analysis (3), which included the right superior frontal gyrus, the right ventrolateral prefrontal cortex, the left inferior parietal cortex, and the left temporal polar cortex. Furthermore, in a post hoc analysis that employed a lower significance threshold for identifying voxels with increased 5-HT_2A receptor binding potential after estradiol treatment (the same threshold used by Dr. Kugaya et al.), we observed widespread increases in estradiol-related cortical 5-HT_2A receptor binding potential.

Given the clinical risks and benefits associated with the hormone replacement therapy regimens investigated in recent large-scale clinical trials (4, 5), mechanistic studies such as these may help shift the focus to alternate hormone replacement therapy regimens that may exert potentially beneficial effects on brain function. Both our study and that of Dr. Kugaya et al. lack the sensitivity needed to establish relationships between changes in 5-HT_2A receptor binding and cognitive function or emotional behavior, which minimizes the potential for making clinical inferences. The sensitivity of both studies was limited by small group sizes and ceiling/floor effects on neuropsychological test performance in cognitively intact euthymic women. The findings of these studies nevertheless support the initiation and inform the design of future studies aimed at investigating neurobiological bases for the potential neuropsychiatric benefits of such treatments in specific patient populations.