Loss of PAX-2 Expression Is a Sensitive Marker for Neoplastic Endometrium
KH Allison, SD Reed, K Upson, LF Voight, CD Jordan, KM Newton, RL Garica. University of Washington Medical Center, Seattle; Group Health Center for Health Studies, Seattle
Background: Immunohistochemical markers of to better define areas of neoplastic endometrium would be of diagnostic use. The PAX-2 gene has been previously reported to be involved in estrogen and tamoxifen driven endometrial carcinogenesis, but there are limited studies on PAX-2 expression by IHC in endometrial hyperplasia.
Design: PTEN, PAX-2, BCL-2 and MLH-1 expression were evaluated in a subsample of endometrial biopsies from women enrolled in the Endometrial Hyperplasia Outcomes Cohort Study. Cases with an index biopsy consensus diagnosis of either complex or atypical hyperplasia with both neoplastic and adjacent normal endometrial tissue present were included. Women receiving high dose progestin were excluded. Antibody staining was scored by one of two study pathologists in both the lesional tissue and the adjacent normal tissue as 0%, 1-25%, 26-50%, 51-75%, or 76-100%. Separate analyses were conducted for individual markers using an exact McNemar's test of the paired binary data.
Results: We found a statistically significant difference in loss of PAX-2 expression between lesional and normal tissue in 76 cases of complex hyperplasia and 47 cases of atypical hyperplasia (P<0.005). Using >25% loss of staining as a threshold, 95% of complex and 96% of atypical hyperplasias had loss of expression of Pax-2 in neoplastic tissue, compared with 21% and 9% of the respective matched adjacent normal endometrium. 53% of cases of complex and 73% of atypical hyperplasias had >25% loss of PTEN expression, while 21% and 29% of adjacent normal endometrium had loss. Loss in the normal endometrium occurred more frequently in secretory endometrium for PTEN (7 of 11 normal secretory controls had loss), while PAX2 had less frequent loss in secretory endometrium (1 of 10 normal secretory had loss). There was not a statistically significant difference in the loss of MLH-1 expression between normal and complex or atypical hyperplasia. We found a statistically significant difference in BLC-2 loss of expression only among women with atypical hyperplasia. Of the four IHC markers, only PTEN had a significant difference in loss between those with complex and atypical hyperplasia (P=0.045).
Conclusions: PAX-2 may be more sensitive and specific than PTEN as a marker of neoplastic endometrium. However, PTEN may be diagnostically useful in the distinction between complex and atypical hyperplasia.
Wednesday, March 11, 2009 1:00 PM
Poster Session VI # 140, Wednesday Afternoon