Evaluation of NAV-2, Cyclin D2, FABP3 and TGLN antibodies for differantial diagnosis between endometrial stromal sarcoma and leiomyosarcoma and for prognosis prediction.

Aim: Leiomyosarcomas (LMS) constitute 1-3% and endometrial stromal sarcomas (ESS) constitute less than 1% of all uterine malignancies. Currently, stage is the most powerful prognostic indicator for both tumors. Five year survival rate for LMS is about 40%, for low grade endometrial stomal sarcomas, it is approximately 80%. For the two above-mentioned tumors, which have different survival rates, differential diagnosis is mainly made by morphological features. In borderline cases immunohistochemical tests are helpful. Currently, CD10 and smooth muscle markers are main immunohistochemical markers for differential diagnosis. Some of the smooth muscle tumors show CD10 positivity whereas some endometrial stromal tumors have immunohistochemical and morphological smooth muscle differantiation. So additional tests are needed. Recently, molecular genetic features about uterine mesenchymal tumors have been described. Studies about uterine sarcomas showed that CDK2NA, FABP3, TGLN, JPH2, GEM, NAV-2, RAB23 are overexpressed in LMS's and SLC7A10, EFNB3, CCND2, ECEL1, ITM2A, NPW, PLAG-1 ve GCNR are overexpressed in ESS's. In our study, we aimed to investigate the NAV-2, TGLN, FABP3 and CCND2 genes products in uterine sarcomas, and its association with disease progression and differential diagnosis characteristics. Materials and methods: The digital archive was searched for patients with a diagnosis of either LMS or ESS, who were examined in Istanbul University Faculty of Medicine Pathology Department between 1992-2018. Then, the slides of these cases were retrieved from the archive and re-evaluated morphologically. A total of 118 uterine sarcomas were obtained. Of these, 71 were LMS, 39 were low grade endometrial stromal sarcoma (LG-ESS) and 6 were high grade endometrial stromal sarcoma (HG-ESS). The remaining two are mixed cases of LMS-ESS and LMS-Angiosarcoma. Paraffin blocks that include smooth muscle and endometrial tissues as internal controls were selected for immunohistochemical examination, using antibodies against to Nav-2, transgelin, FABP3 and cyclin D2. Morphological features were also obtained from the original reports and review of the slides. Survival, follow-up and treatment data of the patients in our study were obtained from the file archives of Istanbul University Istanbul Medical Faculty Obstetrics Department of Gynecologic Oncology and Istanbul University Institute of Oncology. The follow-up information of 49 LMS, 22 LG-ESS and 4 HG-ESS was obtained. Immunohistochemical and morphological findings were compared with clinical data. Results: The majority of uterine sarcomas were LMS cases (60%). LG-ESS cases constituted 33% of the total group. Mean age of cases was 51,17 for LMS's and 45,03 for LG-ESS's. The average follow-up period was 49.6 months for LMS, 128.1 months for LG-ESS and 77 months for HG-ESS. 37 out of 49 patient in LMS group, 0 out of 22 patients in LG-ESS and 1 out of the 4 patients in HG-ESS cases died during follow-up. Mortality rates were significantly different between LMS's and LG-ESS's (p<0,001). The prognosis largely depended on the stage at the initial diagnosis. Because of the small number of cases, Stage 1-2 and Stage 3-4 groups were formed in our study. Advanced stage cases constituted 33% of LMS and 21% of LG-ESS cases, respectively. Of the 6 HG-ESS cases, 3 were stage 1-2 and the other 3 were stage 3. There was no stage 4 patient in all ESS's. For LMS cases, the mean survival time was 89.2 months in Stage 1-2 cases and 23.7 months in Stage 3-4 cases. A significant difference was found between Stage 1-2 and Stage 3-4 case groups on 48 LMS cases (p <0.001). Because of no deaths were detected in LG-ESS cases and the low number of HG-ESS cases, prognostic evaluations were not performed for these tumors. In our study, overall survival (OS) was found to be shorter in patients older than 55 years (p = 0.044) and disease-free survival (DFS) was shorter in patients over 60 years of age (p = 0.013) for LMS. One of the important parameters in the staging is tumor size. We demonstrated that LMS cases larger than 10 cm had recurrence in a shorter period (p = 0.024). From a morphologic view of point, we demonstrated that LMS cases with necrosis had worse prognosis than the others (p = 0.028). In LMS cases with lymphovascular invasion (LVI), mean OS duration was 33.2 months and DFS duration was 19.7 months. These values were calculated as 75.37 and 62.8 months for cases without LVI (for OS: p=0,391, for DFS p=0,331). There was a moderate degree of inverse correlation between mitosis and OS (r =-0.353, p = 0.013). In our study, Nav-2 antibody showed mild reactions in all types of uterine sarcomas. However, moderate and strong reactions were seen in only 18 cases of LMS and one case of HG-ESS. Nav-2 expression was found to be useful for differential diagnsosis between LMS and LG-ESS when only moderate or strong expression was taken to account. There was no relationship between Nav-2 reactions with OS or DFS. Transgelin has been studied in uterine and soft tissue tumors and has been described as a marker of smooth muscle differentiation. In our study, 18 out of 64 LMS cases were to be widespread positive and 18 out of 64 cases were focally positive. All ESS cases were negative for transgelin. The results showed that it has 56,2% sensitivity and 100% spesifity for differentiating LMS's from ESS's. The presence of transgelin expression was not significant with regard to OS or DFS of LMS cases. FABP3 protein reacted strongly in 21 out of 33 LMS cases, 7 out of 11 LG-ESS cases and 1 out of 3 HG-ESS cases. In the remaining cases, weak reactions were observed. We found out that there was no significant difference in FABP3 expression between uterine sarcoma types. Furthermore, there was no association between FABP3 reactions and disease progress in LMS cases. Cyclin D2 is a nuclear-intense protein that plays an important role in the cell cycle. In our study, 68 LMS, 38 LG-ESS and 6 HG-ESS cases were evaluated and nuclear positivity was observed in 47 out of LMS cases, 26 out of LG-ESS cases and 4 out of HG-ESS cases. There was no difference in expressions between tumor groups. But nuclear expression loss was associated with short OS in LMS cases (p = 0.034). Conclusion: LMS cases are the most common in uterine sarcomas. They have significantly worse prognosis than LG-ESS cases. The stage significantly affects the course of the disease in LMS's. Cellular atypia was not associated with survival times. On the other hand, LMS cases with necrosis were also found to have a poor prognosis. LVI does not affect the survival time, statistically. Increased mitosis values were inversely corralated with OS. Of the four markers tested in our study, Nav-2 was found to be insufficient to predict the progress of the disease. On the other hand, cyclin D2, which was previously shown to be associated with poor prognosis in epithelial tumors, was associated with short OS in LMS cases. To our knowledge, this finding is the first in the literature. Therefore, Cyclin D2 molecule is promising for prediction of the disease course for LMS. Transgelin molecule was found to be very useful in detecting smooth muscle differentiation. The FABP3 molecule did not show any significant difference in terms of differential diagnosis and clinical course of uterine sarcomas. In conclusion, the results of immunohistochemical evaluation of cyclin D2 and transgelin molecules are promising. In particular, these two markers need to be studied more in uterine sarcomas and tumors of other regions. They may be useful markers for prediction of prognosis and differential diagnosis and may become useful tests in daily practice.