Abstract
Context.—African American women with breast cancer have worse prognoses than non–African Americans and might benefit with a race-specific prognostic marker such as PTEN (phosphatase and tensin homologue), a tumor suppressor protein. Reduced PTEN expression is associated with worse outcomes and resistance to trastuzumab in human epidermal growth factor receptor 2–positive breast cancers. Standardized PTEN evaluation is therefore needed.
Objective.—To assess PTEN as a race-specific prognostic marker in breast cancer by using a novel semiquantitative score and a percent staining assessment.
Design.—Between 1991 and 1996, 146 patients with invasive ductal adenocarcinoma were grouped by race and recurrence; there was a median follow-up of 7.2 years with 63 recurrences. Immunostaining of PTEN in tissue microarrays was correlated with race, recurrence, node positivity, stage, size, age, estrogen/progesterone receptor status, grade, and DNA ploidy.
Results.—No significant racial difference was detected in mean PTEN values using either the semiquantitative score (P = .46) or the percent staining (P = .54). Unrelated to race, the percentage of tumor cells with positive PTEN expression correlated with longer time to recurrence (P = .047), positive estrogen receptor status (P = .009), and lower tumor grade (P = .005). The semiquantitative score correlated with positive estrogen receptor status (P = .01) and lower tumor grade (P = .001).
Conclusions.—PTEN expression is not a race-specific biologic prognostic marker for invasive ductal adenocarcinoma. Increased PTEN expression correlates with longer time to recurrence, positive estrogen receptor status, and lower tumor grade. The novel semiquantitative score may be used to evaluate PTEN expression, but the approximate percentage of tumor cells with any PTEN staining may be the most useful measure of PTEN expression.
Versatile multicharacterization platform involving tailored superhydrophobic SU-8 micropillars for the investigation of breast cancer estrogen receptor isoforms
