Prognostic markers for coexistent carcinoma in high-risk endometrial hyperplasia with negative D-score: significance of morphometry, hormone receptors and apoptosis for outcome prediction.

OBJECTIVES: Hysterectomy represents the current routine therapy for high-risk endometrial precancers. More sophisticated methods are needed for treatment decision among women who want to preserve fertility and seriously ill patients. Among women diagnosed with high-risk hyperplasia, approximately 40% show signs of endometrial cancer in the hysterectomy specimen. Thus, more sophisticated methods are needed to select the women at risk. SETTING: University Hospital of Tromsø, Regional Center for Gynecological Oncology in northern Norway. POPULATION: From 1999 to 2004, 258 consecutive patients had endometrial hyperplasia diagnosed by D-score; 57 among these were high-risk cases (D-score < 0) and 10 had coexisting endometrial carcinoma. No further cancers were detected after long-term follow-up (4-10 years). DESIGN: From the initial histological specimens, material from the 10 patients with cancer and from the 13 cases without cancer (high-risk D-score < 0) was analyzed with selected histomorphometric (architectural and nuclear) and immunohistochemical (hormone receptors and apoptotic) features blinded to the investigator. METHOD: Original slides were used for computerized histomorphometry (4-class rule and related procedures). Serial sections from the paraffin embedded material were used for immunohistochemical investigations. Immunohistochemical expression in glands and stroma was evaluated by the semi-quantitative H-score (ER-alpha, ER-beta, PR-A, PR-B, RCAS-1, Bcl-2, BAX, and Caspase-3). RESULTS: The histomorphometric 4-class rule differentiates between presence and absence of cancers with a sensitivity of 80% and specificity of 77%. Several morphometric and immunohistochemical features were significantly different in cases with cancer and hyperplasia. CONCLUSIONS: Histomorphometry seems superior in predicting coexistent carcinoma in high-risk endometrial hyperplasia and should be considered for clinical use.

Effect of levonorgestrel IUD and oral medroxyprogesterone acetate on glandular and stromal progesterone receptors (PRA and PRB), and estrogen receptors (ER-alpha and ER-beta) in human endometrial hyperplasia.

OBJECTIVES: The effect on progesterone and estrogen receptor expression in glands and stroma after two different treatment regimens of endometrial hyperplasia was determined. METHODS: Pre- and post-treatment paraffin-embedded endometrial hyperplasia specimens from women treated with levonorgestrel (LNG) intrauterine device (n = 21) and women treated with 10 mg medroxyprogesterone acetate (MPA) for 10 days per cycle (n = 29) were examined immunohistochemically and evaluated by H-score (a semi-quantitative microscopical method evaluating staining intensity and number of stained cells, scale 0-3) for changes in expression of PRA (progesterone receptor A), PRB (progesterone receptor B), ER-alpha (estrogen receptor-alpha), ER-beta (estrogen receptor-beta) and AR (androgen receptors) after 3 months of treatment. RESULTS: All the patients in the LNG IUD group responded to treatment with no sign of hyperplasia after 3 months, while only about half of the patients given MPA orally responded. Expression of PRA, PRB, ER-alpha and ER-beta were markedly reduced after progestin treatment in both treatment groups but the reduction was much more pronounced in the LNG group (H-score for PRA was reduced from 2.61 to 0.11 in glands and from 2.26 to 0.09 in stroma in LNG group. Corresponding reduction for PRB in the LNG group was from 1.96 to 0.11 and from 0.83 to 0.01. PRA was reduced from 2.53 to 1.78 in glands and from 1.93 to 1.30 in stroma in the MPA group. Corresponding reduction for PRB in the MPA group was from 2.02 to 1.25 in glands and from 0.80 to 0.34 in stroma). Weak and focal stromal expression of AR was demonstrated in 22% of the specimens before but not after therapy. There was a statistically significant reduction in both PR and ER among the responders whereas non-responders showed no statistical change after treatment. CONCLUSION: The present study shows that LNG IUD causes an almost complete down-regulation (lack of immunohistochemical expression) of PR expression and a considerable down-regulation of ER expression in both glands and stroma. The changes in receptor expression were markedly less pronounced after treatment with intermittent oral MPA. The differences in receptor expression among responders and non-responders might serve as possible markers for therapy response.

Bcl-2, BAX, and apoptosis in endometrial hyperplasia after high dose gestagen therapy: a comparison of responses in patients treated with intrauterine levonorgestrel and systemic medroxyprogesterone.

OBJECTIVES: The aim of the study was to investigate apoptosis as a growth regulatory mechanism of gestagen in endometrial precancers and to compare differences in the apoptotic cascade after high and low dose gestagen regimens. METHOD: Pre- and post-treatment paraffin-embedded endometrial hyperplasia specimens from women treated with levonorgestrel intrauterine device (n = 26) and women treated with 10 mg medroxyprogesterone for 10 days per cycle (n = 31) were examined for changes in the expression of Bcl-2 and BAX and the extent of apoptosis after 3 months of treatment. Immunohistochemical expression in tissue specimens for Bcl-2 and BAX was evaluated by H-score. Average number of apoptotic cells per hundred cells within ten different high power field (40 x) was evaluated for each section after in situ apoptosis detection (TUNEL method). A second group of patients with endometrial hyperplasia was examined after 1 week treatment with levonorgestrel IUD (n = 6) or medroxyprogesterone (n = 5) to determine early effects on expression of Bcl-2 and BAX and the extent of apoptosis. RESULTS: All the patients in the IUD group (n = 31) but only about half of the patients in per oral group (16 of 26) responded to treatment. The glandular reduction in Bcl-2 expression was markedly greater for the IUD patients than for the patients who received oral gestagen. The decrease in BAX expression after IUD treatment was less than the reduction of Bcl-2. Decrease in glandular Bcl-2 after 3 months of treatment was coincident with a significant increase in the measurable amount of apoptosis. In stromal cells, the increase in expression of Bcl-2 and BAX was found after gestagen treatment, the response being much more marked for the IUD group. The non- responders to per oral gestagen had no Bcl-2 expression in stroma after 3 months of therapy whereas an increase was observed for the responders. After 1 week, glandular Bcl-2 expression was significantly reduced after treatment in the IUD group. As for the rest, no changes were detected in either of the groups. CONCLUSION: Our results indicate that proteins in the apoptotic cascade are regulated by gestagen therapy in human endometrial precancers. Expression of these proteins is shown to be dependent on administration form and/or type of gestagen. Stromal Bcl-2 expression appears to be a potential biomarker which can separate responders of gestagen treatment from non-responders after oral administration.