Ileal neuroendocrine tumors (INETs) are the most common type of neuroendocrine neoplasms in the gastrointestinal tract, with a male prevalence and a median age at the time of diagnosis of 66 years. They are mainly composed of enterochromaffin cells (EC) producing serotonin and substance P [1,2]. A distinctive feature of INETs, especially when involving the liver, is their capability of causing distinct clinical syndromes [1,3,4], which can be faithfully monitored measuring the relevant hormones in the bloodstream. Surgical resection can be curative in early stage patients [5,6], but most of them present with liver involvement at the time of diagnosis, so tumor grading and staging according to WHO/AJCC/ENET criteria are likely to play the most important role in the prognostic and therapeutic assessment of INETs [6]. Accordingly, neuroendocrine carcinomas, which show high proliferative activity as reflected by Ki67 labeling index (LI) over 20%, are treated with cisplatin-etoposide combination chemotherapy similarly to small cell lung cancer (SCLC) [7,8], whereas most ileal neuroendocrine tumors are slowly growing neoplasms, which mainly depend on angiogenesis for their maintenance and growth [9,10,11,12,13,14]. Hence the search for new markers capable of getting new insights into the biological properties of INETs may be clinically warranted.

The succinate dehydrogenase (SDH) enzyme (also known as succinate ubiquinone oxydoreductase) is a highly conserved heterotetrameric protein, with SDHA and SDHB functioning as catalytic subunits, which protrudes into the mitochondrial matrix and is anchored to the inner membrane by means of SDHC and SDHD subunits, the latter also providing the binding site for ubiquinone [15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40]. SDHB is normally ubiquitously expressed with granular cytoplasmic immunostaining reflecting its mitochondrial location [41,42]. It is also been shown that silencing SDHB expression induces tumor-like phenotypic traits in cell cultures [43], and that the loss of any subunit protein, especially B, leads to the loss of SDH expression due to destabilization of its complex [20,44]. However, data on the prevalence of SDHB in INETs and its implications on tumor differentiation and prognosis are still lacking, to the best of our knowledge.

This study was aimed at evaluating the distribution of SDHB by immunohistochemistry (IHC) in 31 INETs and corresponding lymph node or liver metastases in order to explore its diagnostic and prognostic implications.

There were no differences between functioning (FT) and nonfunctioning (NFT) groups in age, gender, clinical outcome and medical treatment. The only significant association was the greater liver tumor load (p = 0.0000026) and the increased basal chromogranin A (CgA) serum level (p = 0.0113) in the NFT group (Table 1).

SDHB immunoreactivity was found in tumor cells of all cases under assessment, but there was a significant relationship between SDHB intensity and percentage of immunoreactive cells (test for trend) (Table 2). Representative pictures of SDHB and Ki-67 antigen immunoreactivity are depicted in Figure 1. In particular, the more the intensity of immunoreactivity, the more the percentage of positive cells. The percentage of tumor cells was associated significantly with the site of tumors and Ki67 LI, since primary lesions bearing a proliferative activity ≤1.3% showed over 50% SDHB immunoreactive tumor cells (Table 3). Likewise, significant associations were found between the site of tumors (p < 0.0001) or Ki-67 LI (p < 0.0001) and SDHB immunostaining intensity (Table 3). No significant associations were found with age, gender, type of therapy, presence of clinical syndrome, and CgA level.

Survival analysis showed that the SDHB intensity but not the SDHB percentage of tumors cells impacted inversely on the patients’ prognosis, with marginal poorer prognosis being observed in individuals loosing SDHB immunoreactivity when considered as a whole (p = 0.1106) and significant shorter survival in the subset of metastatic diseases (p = 0.0387) (Figure 2). Other variables correlating with reduced survival included the lack of transplant treatment (p = 0.0094) and CgA levels >200 (p = 0.001). Although proliferative activity by means of Ki-67 labeling index was not a prognostic factor in this subset of patients, it was however marginally related to survival in metastases rather than primary tumors (p = 0.086). Multivariate analysis according to Cox’s model let emerge CgA level but not SDHB immunostaining intensity as an independent factor of survival (Table 4).

Interesting findings of our study were that SDHB expression correlated with the tumor cell differentiation and malignant potential of G1 INETs, and that the percentage of immunoreactive cells was associated with the staining intensity (Table 2). As a matter of fact the higher was the loss of SDHB immunoreactivity, the higher the proliferative activity (Table 3), the higher the likelihood of facing with metastatic sites (Table 3), and the shorter the survival (Figure 2a,b). Accordingly, SDHB was likely to behave as tumor suppressor gene in this category of neuroendocrine tumors, in that the lack of this protein was associated with parameters of clinical aggressiveness in metastatic tumors and reduced life expectation.

Although the switch from respiration to glycolysis in tumor cells has often been considered a consequence rather than a cause of cancer [45,46], the discovery that germline, inherited mutations in the genes encoding SDH enzyme subunits may cause paragangliomas and phaeochromocytomas [17,19,45,46,47], whether hereditary or sporadic, has however revolutionized this assumption [45,46]. Interestingly, all SDH subunits have no cytosolic counterpart unlike most Krebs cycle enzymes, but are imported into the mitochondria where they are modified, folded and assembled. Hence, they are able to deeply affect the capability of producing energy, whenever these subunits are downregulated in their expression. In our investigation, we observed that metastatic sites showed loss of SDHB expression along with a higher Ki-67 LI thereby supporting the contention that the development of anaerobic metabolism mechanisms could favor an increase of clinical aggressiveness.

In our paper, we provide evidence for the first time that the assessment of SDHB immunoreactivity in well-differentiated INETs may identify a subset of tumors characterized by reduced life expectation, which are worth treating more aggressively with multimodality therapy. Moreover, this paper challenges the common credence that G1 neuroendocrine tumors of the ileum are uniformly poorly aggressive tumors, while they are likely to be a heterogeneous tumor group harboring different lesions with different degrees of malignancy. The loss of function of SDHB as indicated by its relevant down-expression might be responsible for a pseudo-hypoxic drive via succinate-induced glycolysis and HIF stabilization in normoxic conditions [48], so favoring angiogenesis, tumor growth and progression of malignancy [48]. Recent evidences have corroborated the notion of a possible role for succinate accumulation due to SDHB activity loss also in epigenetic changes of chromatin via the histone H3 methylation in succinate-accumulating tumor cells [48]. Further investigation dealing with somatic mutation analysis of SDHB gene is currently in progress in our laboratory to better clarify additional molecular mechanisms underlying the loss of SDHB expression in this subset of INETs patients.

Our study provides the first evidence of a down-regulation of SDHB in well-differentiated INETs as likely mechanism contributing to the development of a subset of biologically more aggressive tumors as heralded by increased proliferative activity and reduced survival. Further investigation is currently in progress in our laboratory on a larger cohort of INETs patients, as well as in other types of neuroendocrine tumors, in order to confirm and expand these preliminary data.