Clinicopathological analysis of c-kit expression in carcinosarcomas and leiomyosarcomas of the uterine corpus.

OBJECTIVE

The purpose of this study was to evaluate the expression of the protooncogene, c-kit, in carcinosarcomas and leiomyosarcomas of the uterine corpus and determine the associations between c-kit expression and clinicopathologic factors, including clinical outcome.

METHODS AND MATERIALS

Using a polyclonal anti-KIT-antibody, immunohistochemical staining was performed on formalin-fixed paraffin-embedded tissue blocks from 21 carcinosarcomas, 17 leiomyosarcomas, and 1 endometrial stromal sarcoma. KIT-positive tumors were defined as those tumors demonstrating immunopositivity in > or =30% of tumor cells examined. KIT-negative lesions demonstrated immunopositivity in <30% of tumor cells. Two authors independently scored the slides as positive or negative. Staining was repeated on all specimens and independently scored, and in the occasion of a mismatch, a third staining was performed. The carcinosarcomas were catalogued as to whether the sarcomatous and/or carcinomatous elements expressed c-kit. Clinical data were abstracted for those patients with uterine carcinosarcomas. The associations between clinicopathologic characteristics and c-kit expression were compared using univariate and multivariate analyses. Kaplan-Meier curves based on c-kit expression were plotted for progression-free and overall survival and compared using the log-rank test.

RESULTS

Nine of 21 (43%) carcinosarcomas demonstrated immunopositivity for the KIT receptor, although staining was relatively weak. In contrast, only 1/17 (6%) leiomyosarcomas demonstrated KIT immunopositivity (P = 0.029). The solitary endometrial stromal sarcoma evaluated did not demonstrate significant KIT positivity. The majority of KIT-positive carcinosarcomas (6/9 (67%)) demonstrated KIT presence in the sarcomatous portion as compared to the carcinomatous portion (4/9 (44%)). No clinical factor had a statistically significant association with c-kit expression. The lack of c-kit expression was the only factor that was significantly associated with disease recurrence in univariate and multivariate analyses (P < 0.05), although there appeared to be a trend toward a low stage associated with kit positivity. The median progression-free interval for the KIT-negative cohort was 8 months, while it had not been reached for the KIT-positive cohort after median follow-up of 15 months (P = 0.0462).

CONCLUSIONS

A significant proportion of carcinosarcomas of the uterine corpus display immunoreactivity for c-kit. Patients with KIT-positive carcinosarcomas may have an improved progression-free survival compared to KIT-negative tumors; however, further data are needed to determine whether this finding is confounded by stage.