Frequency and Clinical Implication of SOX11 Expression in Burkitt Lymphoma

Introduction: The transcription factor SOX11 is normally expressed during embryogenesis in humans. Several studies have shown that SOX11 is aberrantly expressed in various types of haematopoetic and solid malignancies, and appears to affect clinicopathological characteristics. For example, in mantle cell lymphoma, SOX11 serves as a diagnostic antigen and has been shown to be associated with superior outcome (Nordström et al. British Journal of Haematology 2014), although its role as a prognostic indicator remains controversial. In chronic lymphocytic leukemia, SOX11 was associated with inferior overall survival (Roisman et al. Tumor Biol 2015), whereas SOX11 expression correlated with superior outcome in epithelial ovarian cancer (Sernbo et al. BMC Cancer 2011). Various possible target genes and transcriptional programs have been identified, by which SOX11 may exert both repressive, and exaggerative, functions on proliferation, and potentially function as therapeutic targets in the future (Kuo et al. Oncogene 2015, Vegliante et al. Blood 2013). In Burkitt lymphoma (BL), two small series have reported expression of SOX11 in 30-50 % of cases (Mozos et al. Haematologica 2009, Dictor et al. Haematologica 2009). As of yet, no study has been performed specifically to evaluate the implication of SOX11 expression in BL. The aim of this study was to investigate the frequency of SOX11 positive BL cases in a joint cohort from Denmark and Sweden, and correlate its expression to clinical and pathological parameters.

Methods: 45 BL cases were collected from Sweden and Denmark. Samples were analysed for expression of SOX11 and other immunohistochemical markers. Cases were classified classic BL or B-cell lymphoma, unclassifiable, with features intermediate between DLBCL and BL (BCLU) after pathology review. Clinical data were obtained from the Danish Lymphoma Database (LYFO) and the Swedish Lymphoma Registry. Additionally, 9 pediatric BL cases were analyzed for SOX11 expression, but with no clinical data available.

Results: In the adult BL study population 14/45 (31 %) expressed nuclear staining of SOX11. The SOX11 positive subgroup had a higher median age (53 compared to 41), as well as a larger proportion with elevated LDH. Other prognostic factors were equally distributed between the groups. Although a similar proportion of patients received high-intensive chemotherapy in both subgroups, 3 (21 %) received no treatment in the SOX11 positive cohort. In this BL material, 28 (62 %) were classified classic BL and 17 (38 %) BCLU. There was a significant difference in SOX11 expression between classic BL and BCLU (p=0.03). Among classic BL SOX11 was seen in 12/28 (43 %) whereas in BCLU only 2/17 (12 %) expressed SOX11.

Five-year overall survival (OS) in the SOX11 positive group was 57 %, compared to 77 % in the SOX11 negative cohort. When adjusting for treatment, BCLU and prognostic factors in multivariable analysis, no significant difference in outcome was seen between groups (hazard ratio: 2.5 95 % confidence interval: 0.4-13.9, p=0.3). When restricting analysis to classic BL cases only, there was no remaining difference in median age, and no significant difference in OS between SOX11 positive and negative patients in neither univariable nor multivariable analysis. In the pediatric cohort, 5/9 (56 %) expressed SOX11.

Conclusion: SOX11 expression was found in less than half of classic BL, and in a small minority of BCLU. In contrast to SOX11 expression in other malignancies, SOX11 expression had no significant impact on outcome, in our cohort, when adjusting for prognostic factors, treatment and BCLU.