Parallel assessment of the impact of different hormone replacement therapies on breast density by radiologist- and computer-based analyses of mammograms.

ABSTRACT

OBJECTIVES: First, to compare the impact of nasally and orally dosed estradiol on breast density; second, to investigate the utility of computer-based automated approaches to the assessment of breast density with reference to traditional methods. METHODS: Digitized images from two 2-year, randomized, placebo-controlled trials formed the basis of the present post hoc analysis. Active treatments were 1 mg estradiol continuously combined with 0.125 mg trimegestone (oral hormone replacement therapy, HRT) or low-dose (150 or 300 microg estradiol) nasal estradiol cyclically combined with 200 mg micronized progesterone (nasal HRT). The effects on breast density were assessed by a radiologist, providing the BI-RADS score and the interactive threshold, and by a computer-based approach, providing the measure of stripiness and the HRT-effect specific measure of breast density. RESULTS: In the oral HRT trial, active treatment induced a significant increase in breast density, which was consistent in all methods used (all p < 0.05). In contrast, none of the methods detected significant changes in women receiving nasal HRT. The sensitivity of automated methods to discriminate HRT- from placebo-treated women was equal or better than the sensitivity of methods performed by the radiologist. CONCLUSIONS: The markedly different pharmacokinetic profile of nasal estrogen seems to be associated with better breast safety. Automated computer-based analysis of digitized mammograms provides a sensitive measure of changes in breast density induced by hormones and could serve as a useful tool in future clinical trials.