Estimation of maspin's subcellular localization in invasive ductal breast cancer via light microscopy and computerized image analysis: a comparative study.

PURPOSE Contradictory results have been reported concerning the role of maspin and its cellular distribution in breast cancer. The purpose of this study was to examine the subcellular localization (nuclear-cytoplasmic) of maspin in breast cancer and to compare the evaluation of maspin immunostaining via light microscopy (LM) to the estimation via computerized image analysis (CIA) system. We also examined correlations between maspin expression and several clinicopathological parameters. METHODS The sample consisted of 48 primary invasive ductal carcinomas (IDC) of the breast. Maspin immunostaining was quantified and graded via LM by two pathologists, separately in the nuclear and cytoplasmic compartments. Total maspin expression was also estimated via CIA system. Univariate non-parametric statistics and stepwise multivariate ordinal logistic regression were performed. RESULTS Both maspin components (nuclear and cytoplasmic) were closely associated with each other (p<0.001). Total maspin score was positively and closely associated with nuclear maspin (p<0.001) and cytoplasmic maspin (p<0.001). Total maspin , nuclear maspin and cytoplasmic maspin did not correlate significantly with either age, grade, T, N and M status, stage, micro vessel density (MVD) (CD34), ki-67, p53, estrogen receptor (ER) and HER-2 status, or with any of the 4 groups of the molecular classification. The only factor that showed a borderline inverse correlation with nuclear maspin (p=0.059) was progesterone receptors (PR) positivity. CONCLUSION The cytoplasmic and nuclear fractions of maspin seem to be closely interwoven. Evidently, both mutually intertwined counterparts were independently reflected upon the total maspin levels measured by CIA. Future studies should ideally encompass all three approaches (nuclear, cytoplasmic, total) adopted herein.