P53 GENE MUTATION IN FEMALE GENITAL-TRACT CARCINOSARCOMAS (MALIGNANT MIXED MULLERIAN TUMORS) - A CLINICOPATHOLOGICAL STUDY OF 74 CASES

Mutations of the p53 suppressor gene are involved in carcinogenesis by inactivating p53 protein (p53P), which is involved in normal cell gro wth control. Mutant p53P, detectable by immuno-histochemistry due to l onger half-life as compared to wild-type, is a marker for p53 gene mut ation. Seventy-four female genital tract carcinosarcomas (FGTCSs) (41 of the heterologous type exhibiting 33 rhabdomyosarcomatous, 13 chondr osarcomatous, one osteosarcomatous, and one liposarcomatous component) were stained using two commercially available monoclonal antibodies: p53P, clone DO7 (p53P-DO7) and p53P, Ab-6 (p53P-P6). p53P-DO7 and p53P -P6 stained 33 and 36, respectively, of 56 endometrial, 8 of 11 ovaria n, 2 of 5 cervical, and 2 of 2 fallopian tube carcinosarcomas. Conside ring all 74 FGTCSs, p53P-DO7 and p53P-P6 stained both the carcinomatou s component (CC) and the sarcomatous component (SC) in 46% and 54%, th e CC only in 9.5% and 8.1%, and the SC only in 5.4 and 2.7%, respectiv ely. The two antibodies for p53P showed the following concordance for staining of FGTCSs (either CC or SC or both) (p53P-DO7/p53P-P6): +/+, 58%; +/-, 1.3%; -/+, 6.7%; -/-, 34%. p53P immunoreactivity was not ass ociated with histological features or grade of the CC or SC. Clinical follow-up was available in 72 cases, which showed 48.5% and 70.8% of p atients recurred or died of disease by 12 and 80 months, respectively. 20.8% of patients were disease-free after 19 to 307 months of follow- up (median, 62; mean, 92). The remaining 8.4% of patients were disease -free but had insufficient follow-up (<1 year). Recurrent disease was associated with high stage, III or IV (P < 0.001), vascular space inva sion in resection specimens (P < 0.01), large tumors (P < 0.01), and d eep myometrial invasion in hysterectomies (P < 0.01). p53P staining wa s not a predictor of disease-free survival, even when other prognostic factors were held constant. p53 gene mutations are detected and presu med to be involved in the carcinogenesis of 66% of 74 FGTCS. Both the CC and the SC show mutations in 81.6% of these 49 p53P-positive FGTCS, which suggests a similar carcinogenic mechanism for both components. p53 gene mutation neither correlates with histopathology nor predicts disease recurrence in this patient group.