This retrospective study was conducted from January 2014 to January 2019 (5 years interval). Patients who were included to our study diagnosed with primary DLBCL that were aroused from the head and neck region. All cases included to our study had complete medical records and with confirmed diagnosis of DLBCL by sample immunophenotyping. After exclusion of cases had insufficient biopsy speciemen and those with missing medical records, Forty-five paraffin blocks for DLBCL cases with thorough medical and clinicopathological data were retrieved from archives of the Pathology laboratory and Oncology unit of Oncology Center, Faculty of Medicine, Mansoura University. Patients’ follow up data were obtained from the medical reports belong to the internal medicine department. All patients who were included in our study were treated with R-CHOP-like therapies (rituximab, cyclophosphamide, prednisone, doxorubicin hydrochloride and vincristine). After completion of treatment, patients were followed every 3 months intervals. The follow up comprised clinical examination and periodic ultrasonography for the head and neck region. In addition, chest X-ray, bone scan, and abdominal ultrasonography were performed when relapse was suspected. Overall survival (OS) and disease-free survival ( DFS ) data were retrieved from patients’ medical reports. Five years OS was calculated from the dates of diagnosis to death or to the dates of the last follow up. PFS was calculated from the dates of first progression, death, relapse or the last follow up. Previous clinicopathological and follow-up data were obtained after getting approval from the institutional review board. This study was approved by the local ethical committee.

All selected blocks were cut at 4 µm thickness. The slices were placed on coated slides. After xylene deparaffinization, the sections were rehydrated in descending grades of alcohol followed by water. Antigen retrieval was performed by using 0.01 M citric acid buffer (pH = 6.0) and heated for 10 minutes in the microwave. The sections were then incubated in a blocking medium (3% H₂O₂) for five minutes followed by washing with distilled water. Rabbit monoclonal anti-human antibody (clone; EPR 5111, 1:50, dilution, Abcam, 1 Kendall Square, Suite B2304, Cambridge, MA 02139-1517, USA (was used against BCL2 antibody. Assessment of BCL2 positivity was performed by staining sections of hyperplastic human tonsillar tissue that used as a positive control at the same time and under the same conditions. Negative control slides obtained by replacement of the primary antibodies by plain phosphate buffer saline. Immuno-detection was executed using Power-stain TM1.0 poly HRP DAB kit for mouse + rabbit (Cat No 52-0017, Genemed Biotechnologies, Inc., 458 Carlton Ct., South San Francisco, CA 94080, USA). Immune staining was performed based on the manufacturer’s instructions. Immunoreaction was visualized by adding DAB Counterstaining of slides that were performed with the Mayer hematoxylin.

Immunohistochemical results were evaluated semiquantitatively. The BCL2 positivity was determined by cytoplasmic staining (brown) of neoplastic cells. The internal control was T cells. The percentage of positive cells in the whole section after the exclusion of the areas of reactive T cells was determined. It was scored negative if 5% or less of neoplastic cells were stained. In this study, the value of BCL2 was considered weak positive if 6 to less than 50% were brown stained. For statistical analysis, we considered negative and weak positive (low expression) and high expression if ≥ 50% of tumor cells were brown stained [14]. The histopathological and immunohistochemical evaluation was done by two pathologists independently and blindly. The images were acquired utilizing a Nikon Eclipse microscope equipped with a 5-megapixel cooled CCD camera and the Image-Pro + AMS7 software.

Statistical analysis of the data was done by using the Excel program and the Statistical Package for Social Science (SPSS) version 22 program. The description of the data was done in the form of mean (+/−) Standard Deviation (SD) for quantitative data, while in the form of frequency and proportion for qualitative data. The analysis of the data was done by one way ANOVA, to test statistical significant difference between groups and Wilcoxon multiple comparisons test for comparison of each two groups. Pearson correlation co-efficiency test was used to test the association between the different variables. A chi-square test was used to compare between two groups for qualitative data. The construction of survival curves was conducted using the Kaplan-Meier method, and the significance was assessed with the log-rank test. Univariate and multivariate survival analyses were performed with the Cox proportional hazards model to detect an independent prognostic factor. Using the univariate and multivariate analysis, the outliers were detected and cleaned from the data set. A P-value less than or equal to 0.05 was considered statistically significant.

A total of 45 patients of DLBCL were identified and included in the study. Regarding gender, our study included 24 males (53.3%) and 21 females (46.7%) with male to female ratio 1.14:1. The research included 35 cases (77.8%) were younger than 60 years old, while the rest were older than 60 years old (10 cases, 22.2%). DLBCLs that aroused from lymph nodes of head and neck region comprised 28 cases (62.2%), while the rest aroused from extranodal location (Table 1). The majority of the studied cases presented as localized disease in stage I and

II according to Ann Arbor clinical staging system (33 cases, 73.3%). DLBCLs of advancing clinical stages (III and IV) demonstrated in 12 cases (26.7%).

All participant of the study received chemotherapy. Patients’ response to chemotherapeutic treatment was assessed during the follow up visits in accordance with international working group recommendations for response criterion for NHL. Patients were divided into 3 groups according to their response to the administered chemotherapy as follow: respond to less than 3 chemotherapeutic courses (14 cases, 31.1%), respond to three or more chemotherapeutic courses (16 cases, 35.6%) and chemoresistant group (15 cases, 33.3%). Recurrence of the tumor during follow up demonstrated in 18 cases (40%). Nine of the followed cases (20%) were died during follow up. Patients’ clinicopathological data is shown in Table 1.

We selected a cutoff of ≥50% protein expression for BCL2 positivity. In the total cases BCL2 positive rate was 80% (36 out of 45 cases; Figures 1(D)-(F)). The relationship between different clinicopathological variables and BCL2 expression was assessed using Chi-square test. BCL2 expression had no statistically significant associations with tumor location either nodal or extranodal (p = 0.758), Ann Arbor clinical stage (P = 0.736) and patient’s age (p = 0.37). Table 2 illustrates the data.

BCL2 expression was positively expressed in 17 out of the 18 cases (94.4%) that had recurrence during follow up. Nineteen out of 27 non-recurring cases (70.4%) had positive BCL2 expression. Chi-square test revealed a statistically

Chi-square test; p-value significant if ≤ 0.05.

significant difference in BCL2 expression according to incidence of recurrence (P = 0.048).

All cases that died during follow up demonstrated positive BCL2 expression (9 cases, 100%). On the other hand, BCL2 was positively expressed in 27 (75%) out of 36 alive cases. Although Chi-square test revealed no statistically significant difference in BCL2 expression between the two groups (died and alive), but the P-value (p = 0.09) approximates the value required to reach significance.

BCL2 expression was varied among the three different patients’ responses to chemotherapy. Fourteen out of 15 chemoresistant cases (93.3%) had positive BCL2 expression. Fourteen out of 16 Patients (87.5%) who received and responded to ≥3 chemotherapeutic courses had positive BCL2 expression. Eight out of 14 patients (57.1%) who responded to less than three chemotherapeutic courses had positive BCL2 expression. Chi-square test revealed a statistically significant difference in BCL2 expression among the mentioned three groups of patients according to their response to the taken drugs (P = 0.033; Figure 2). BCL2 expression in the studied DLBCLs could predict the patients’ chemotherapeutic response (ANOVA test P-value = 0.015).

Patients’ OS were analyzed in relation to status of BCL2 expression using Kaplan Meier method and log rank test was used for comparing the variables. The mean OS in patients with positive BCL2 expression was 37.5 months, while those who had negative BCL2 expression was 56 months. In other words, the mean OS was greater in BCL2 negative patients (n = 9) as compared to the positive ones (n = 36, 56 Vs 37.5 months; p = 0.000) Table 3, Figure 3. BCL2 expression in tumor could predict the 5 years OS (R² = 40.5%, p = 0.000, Table 4).

a. Predictors: (Constant), BCL2 expression; Dependent Variable: 5 years overall survival.

Patients’ OS was analyzed in relation to patient age, gender, tumor location nodal or extranodal, tumor clinical stage and different patient’s responses to the taken chemotherapeutic drugs. None of these variables had effect on patient’s OS except for patient’s age and response to chemotherapy. In more words, mean survival was greater in patients under age of 60 years as compared to those above 60 years old (n = 42, 8 Vs 35.7 months; p = 0.044). Also, chemoresistant cases had shorter OS as compared to cases that respond to 3 or more chemotherapeutic courses and cases that respond to less than three courses (32.2 Vs 42.18 Vs 49.71 months respectively; p = 0.000).

Patient PFS was analyzed in relation to BCL2 expression using Kaplan Meier survival curves. The mean PFS in patients with positive BCL2 expression was 27.4 months as compared to those who had negative BCL2 expression 52.3 months, a high statistically significant difference in PFS according to the status of BCL2 expression (p-value = 0.000; Table 5, Figure 3).

Patients’ PFS was analyzed in relation to patient age, gender, tumor location either nodal or extranodal, clinical stage and different patient responses to chemotherapy. All these independent variables had significant difference in relation to PFS except for gender and tumor location. In more words, patients who diagnosed with tumor in advanced clinical stage (III and IV), the age ≥ 60 years old, chemoresistant cases had shortened PFS when compared to patients who had tumors in localized clinical stage (I and II), patient age younger than 60 years, patients who respond to chemotherapy (P < 0.05, log Rank test; Figures 4-7). BCL2 expression in tumor could predict patients PFS (R² = 28.7%, p-value = 0.000, Table 6).

Finally, we analyzed patients’ OS in addition to PFS in relation to BCL2 expression, Ann Arbor clinical stage, incidence of recurrence and different patients’ responses to chemotherapy as covariables using COX regression model. BCL2 expression and different patients’ chemotherapeutic response were the only two independent variables that affect patients’ OS and PFS (p < 0.05, Table 7 and Table 8).

ANOVA a. Dependent Variable: 5 years progression free survival; b. Predictors: (Constant), BCL2 expression.