ELMO3 – a Negative Prognostic Marker in Minor Salivary Gland Carcinoma

Fifty patients with newly diagnosed carcinoma of the minor salivary glands, treated between 1974 and 2013 at the Medical University of Vienna, were included in this retrospective study. Clinicopathologic information was derived from medical charts and patients were classified according to the TNM classification of the Union for International Cancer Control (UICC). Patients that were lost to follow up (1 patient) or patients with missing tissue samples (3 patients) were excluded from statistical analysis. Subsequently, 46 patients were eligible for further investigations. The detailed patients data and histological classification are summarized in Table 1. This study was approved by the institutional ethics committee (EK 1926/2015). Informed consent was obtained from all patients and the study was performed in accordance with the Declaration of Helsinki.

Immunohistochemical staining was performed as described elsewhere using the Lab Vision Ultra V Block kit (Thermo Fisher Scientific, Waltham, MA, USA) and the Lab Vision Ultravision LP detection system (Thermo Fisher Scientific, Waltham, MA, USA) according to manufacturers protocol [16]. Briefly, citrate buffer (pH 6.0) was used for antigen retrieval. Anti-ELMO3 antibody (Sigma-Aldrich, St.Louis, MO, USA) was diluted to 1:500. Slides were incubated for one hour at room temperature. Paraffin-embedded samples of non-small-cell lung carcinoma and squamous cell carcinoma served as positive control. For negative control primary antibody was replaced by rabbit immunoglobulin G isotype control (Abcam, Cambridge, United Kingdom). Samples were analyzed using an Olympus BH-2 microscope (Olympus, Tokyo, Japan). Based on the intensity of the cytoplasmic staining of neoplastic cells, all samples were assigned to one of four categories: 0: negative; 1: weak; 2: moderate; 3: strong. The assignment was performed by two independent investigators (FO and LK). Subsequently, the percentage of stained neoplastic cells was taken into account (<10% negative; >10% positive) and specimens were characterized as ELMO3 positive or ELMO3 negative. For statistical analysis we combined patients with negative and weak expression of ELMO3 to the group of “ELMO3 low” patients and patients with moderate and strong expression to the group of “ELMO3 high” patients.

For clinical and patients’ data, we used descriptive statistics. Additionally, Fisher’s exact test was used to compare categorical data between two groups and chi square test was used in case of 3 or more groups. Rates of OS and DFS were calculated by means of the Kaplan-Meier method. We used the log-rank test (Mantel-Cox) to assess statistical differences between the established patient groups. Furthermore, hazard ratios (HR) were calculated and according 95% confidence intervals (CI) were established to reflect a significance level of 0.05. Moreover, 2 × 2 tables with incidence data of death and recurrence were constructed and analyzed by using Fisher’s exact test. Multivariate analyses were performed using the Cox regression model in order to identify independent variables. Thereby all variables were included into multivariate analysis. Statistical significance was determined as p < .05. SPSS software version 21.0 (SPSS Inc., Chicago, IL) and Prism Graphpad (Graphpad Software Inc., La Jolla, CA) were used to analyze and visualize the data.

Forty-six patients with complete medical data were stained for ELMO3 expression. Out of them, 25 (54.3%) patients were diagnosed with an adenoid cystic carcinoma, 9 (19.6%) patients had a mucoepidermoid carcinoma, 6 (13%) patients had an adenocarcinoma, 4 (8.7%) patients had a carcinoma ex pleomorphic adenoma, and in one (2.2%) patient each basal cell adenocarcinoma and clear cell carcinoma was found. Of these patients, 22 were male (48%) and 24 were female (52%). The mean age was 57 years (range 25–88 years, median 61 years). At time of diagnosis 18 (39.1%) patients presented with T1 disease, 14 (30.4%) patients had a T2 classification and 7 (15.2%) patients each were diagnosed with T3 and T4. Forty-two patients (91.3%) did not show any lymph node metastases and were therefore classified as N0. Four patients (8.6%) presented themselves with N2b classification. Two patients (4.3%) had distant metastases. All patients received either surgical resection with or without irradiation, depending on the staging, or radiochemotherapy. Of the patients with surgical therapy, four had a R1 resection. The mean follow-up period was 5.3 years (range 0.2–31.6 years). Next, rates of 10-year survival were calculated for our cohort. Median OS measured 6.3 years and median DFS was calculated as 5.0 years. A total of 27 (58.7%) patients developed recurrent disease and 23 patients (50%) died during the observation period.

All samples were analyzed independently by two examiners. The calculated kappa value was 0.652. ELMO3 expression was detected in 39 (85%) of 46 samples. The expression of ELMO3 could always be found in the cytoplasm of the cancer cells (Fig. 1). Weak expression was found in 17 (37%), moderate expression in 15 (32.6%) and strong expression was found in 7 (15.2%) of the analyzed samples. Seven (15.2%) samples showed no staining. Adjacent connective tissue, nerves, and vessels showed no staining. Inflammatory cells showed a weak to focal moderate staining. For further statistical analysis the groups “no expression” and “weak expression” were summarized to an “ELMO3 low” staining group. “Moderate expression” and “strong expression” were summarized to an “ELMO3 high” staining group.

OS and DFS were calculated. The OS rate after 10 years was 39.4% for patients with low ELMO3 expression and 18.4% for patients with high ELMO3 expression (Table 2). These results were not statistically significant (p = .2502). However, in terms of DFS, a significantly worse outcome for patients with high ELMO3 expression was observed. DFS after 10 years was 39.4% for patients with low ELMO3 expression and 10.7% for patients with high ELMO3 expression (p = .0495) (Fig. 2). Moreover, patients with high ELMO3 expression showed a significantly increased recurrence rate (p = .0071). After total follow up, 80% of the patients with high ELMO3 staining had a recurrence as compared to 40% of patients with low ELMO3 staining (Fig. 3). Regarding the OS, there was no statistically significant difference between the high and low staining groups. However, more patients with high ELMO3 staining died compared to the low staining ELMO3 group (p = .2362).

Moreover, univariate and multivariate analyses were performed to evaluate if ELMO3 expression serves as an independent marker for DFS. ELMO3 expression, T classification, N classification, histological classification, tumor localization and resection margins were included into calculation. Since malignancies of the minor salivary glands located in the nasal cavity and/or paranasal sinuses have a significantly worse survival [17], tumors in this localization were compared to tumors in the remaining localizations. As aforementioned univariate analysis showed that ELMO3 is a negative prognostic factor for DFS. Beside ELMO3 expression, adenoid cystic carcinoma (p = .033) and adenocarcinoma (p = .018) were associated with reduced DFS rates in univariate analysis. However, multivariate analysis revealed that DFS is significantly depending on the histologic subtype (Table 3).