Bcl-2 and Noxa are potential prognostic indicators for patients with gastroenteropancreatic neuroendocrine neoplasms

Bcl-2 family proteins are of great significance in the pathogenesis and development of tumors. In this study, the correlations between the expression of Bcl-2 family proteins and clinicopathological features and prognosis of neuroendocrine neoplasms (NENs) were further investigated. 105 Patients diagnosed with gastroenteropancreatic NENs (GEP-NENs) with the paraffin specimen of the tumor available were retrospectively included. Immunohistochemistry (IHC) was performed to detect the expression of Bcl-2 family proteins in paraffin-embedded samples. Student’s t-test and Chi-square test were applied to compare the difference of quantitative and categorical variables, respectively. Survival analysis was conducted according to Kaplan–Meier method. Univariate and multivariate cox regression analysis were used to identify the independent prognostic factors. The IHC score of Bcl-2 was significantly higher in neuroendocrine carcinoma (NEC) patients (65.6%), while a higher IHC score of Noxa was more common in neuroendocrine tumor (NET) patients (49.3%). Survival analysis indicated that patients with higher Bcl-2 expression and lower Noxa expression had worse 5-year survival (39.3% vs. 75.6%, p < 0.001; 40.6% vs. 84.9%, p < 0.001). Multivariate cox analysis indicated that high Bcl-2 expression was an independent factor associated with inferior DFS (hazard ratio [HR]: 2.092; 95% confidence interval [CI]: 1.106–3.955; p = 0.023) and OS (HR: 2.784; 95% CI: 1.326–5.846; p = 0.007), while higher Noxa expression was associated with superior DFS (HR:0.398; 95% CI: 0.175–0.907; p = 0.028) and OS (HR: 0.274; 95% CI: 0.110–0.686; p = 0.006). Higher expression of Bcl-2 and lower expression of Noxa were associated with unfavorable prognosis of GEP-NENs patients.


Introduction
Neuroendocrine neoplasms (NENs) are a rare group of tumors with high heterogeneity, originating from neuroendocrine cells or peptide-energetic neurons. In a series of 64,971 cases with NENs reported by the Surveillance, Epidemiology, and End Results (SEER) program of the National Cancer Institute, the annual incidence rate increased from 1.09 per 100,000 in 1973 to 6.98 per 100,000 in 2012 [1]. The therapeutic effects of chemotherapy and approved target drugs are still limited. For instance, the median progression-free survival (PFS) of patients received targeted therapies such as sunitinib and everolimus in well-differentiated pancreatic neuroendocrine tumors (NETs) were 11.4 months and 11.0 months, respectively [2,3]. In non-pancreatic neuroendocrine tumors, the median PFS of patients received everolimus and surufatinib was 11.0 months and 9.2 months, respectively [4,5]. Thus, development of novel targeted small molecule drugs is still an arduous subject that needs to be resolved urgently.
Apoptosis, one of the important forms of cell death, is tightly regulated by the balance of proteins in the Bcl-2 family which includes pro-apoptotic proteins (e.g., Bax, Bak), anti-apoptotic proteins (such as Bcl-2, Mcl-1, Bcl-xL) and BH3-only proteins (such as Noxa, PUMA, etc.) [6,7]. The imbalance between cell proliferation and apoptosis, or apoptosis evasion, is of great importance for tumorigenesis [8]. In some solid and hematologic tumors, Bcl-2 overexpression is associated with more malignant phenotypes and worse prognosis [9][10][11]. Based on this theory, several small molecules targeting Bcl-2 were designed to re-activate apoptosis which have also been developed and tested in clinical trials. Venetoclax, for instance, has been approved by the Food and Drug Administration (FDA) for the treatment of adult chronic lymphocytic leukemia (CLL), small lymphocytic lymphoma (SLL), as well as acute myeloid leukemia (AML) by combining with azacitidine or decitabine [9,12]. Although previous studies found that pancreatic small cell neuroendocrine carcinomas (NECs) had a higher Bcl-2 expression level than pancreatic NETs [13], the expression of Bcl-2 family proteins in GEP-NENs is still less understood. Therefore, we aimed to further study the expression of Bcl-2 family proteins, including Bcl-2, Bcl-xL, Mcl-1, Noxa, and PUMA, in patients with GEP-NENs, and investigate their relationships with the prognosis of patients.

Patient cohort and data collection
Parraffin-embeded tumor tissue samples were collected from 105 patients who were diagnosed with GEP-NENs at The First Affiliated Hospital, Sun Yat-sen University from January 2008 to November 2016. All patients included were pathologically reassessed according to the 2019 world health organization(WHO) criteria [14] by a pathologist with extensive experience, and TNM staging was reperformed according to the eighth edition AJCC guidelines [15]. All the clinicopathological data, including gender, age, tumor grade, tumor size, lymphatic and distant metastasis status, and follow-up data, were retrospectively collected from the medical record system of The First Affiliated Hospital, Sun Yat-sen University. This study was approved by the clinical ethics committee of The First Affiliated Hospital, Sun Yat-sen University (2020489) and complied with the ethical standards of the World Medical Association Declaration of Helsinki.

Immunohistochemical staining
Immunohistochemistry (IHC) was carried out as followed. In brief, first, the paraffin-embeded samples were sectioned into 4μm-thick sections and dewaxed in xylene, rehydrated in rinsed graded ethanol solutions. Endogenous peroxidase activity was then blocked using 3% hydrogen peroxide solution for 10 min, and rinsed in phosphate-buffered saline (PBS) 3 times for 5 min each. Then the tissue sections were heated at 100°C for 5 min in citrate (10 mmol/L, pH 6.0) solution to retrieve the antigens. After cooling to room temperature, serum blocker was added to block nonspecific antigen, and the sections were incubated with the primary antibody Bcl

Immunohistochemical analysis
The slides were evaluated independently by two observers blinded to clinicopathological information. Any disagreement was resolved by a joint reevaluation. The expression of markers was evaluated by combining the percentage of positive cells and staining intensity. The percentage of positive cells was evaluated quantitatively and scored as 0 for staining of <2% of total cells counted, 1 for staining of 2-25%, 2 for staining of 26-50%, 3 for staining of 51-75%, and 4 for staining of > 75% of the cells examined. The intensity was graded as follows: 0, negative staining; 1, weak; 2, moderate; and 3, strong staining. A total "staining score" of 0-12 was calculated by multiplying staining intensity score and staining percentage score [16,17].

Follow-up
All patients were followed up every three to six months in out-patient clinic or by telephone, with the last telephone follow-up in June 2020. The time of death was recorded for patients who died, and the last follow-up date and status of patients who could not be reached were obtained from the hospital system. The primary outcome was overall survival (OS), defined as the time from the date of diagnosis to death or the last follow-up. The secondary outcome of the study was disease-free survival (DFS), defined as the time from the date of diagnosis to disease recurrence or death or the last follow-up.

Statistical analyses
Statistical analyses were performed using IBM SPSS software, version 25.0 (IBM, Chicago, IL, USA). The cut-off value of markers expression was determined by the median IHC score as previous study [18,19]. The IHC score which was greater than cut-off value was defined as high expression, while lower than cut-off value was defined as low expression. Quantitative variables were presented as mean ± average and categorical variables were presented as percentages. Chi-square test (or Fisher's exact test) was applied to compare categorical variables. Survival analyses were performed via Kaplan-Meier method with long-rank. Univariate and multivariate analyses were carried out based on cox proportional hazard regression. Results were presented as hazard ratio (HR) and 95% confidence intervals (CI). Two sides p-value < 0.05 was considered statistically significant.

Correlation analysis between Bcl-2 family markers and clinicopathological characteristics
As shown in Fig. 1, immunohistochemical staining for Bcl-2 and Noxa were found in cytoplasm of tumor cells. The result of Chi-square test analysis showed that there was no correlation between the expression of Bcl-2 and Noxa (p = 0.161). In order to determine the relationship between the expression of Bcl-2 family markers and clinicopathological parameters, we then took the Bcl-2 family markers expressions and clinical characteristics into Chi-square test analysis (  Fig. 2A). Next, we divided the whole cohort into NETs and NEC subgroups for further analysis, and the results showed that the 5-year DFS rates of high and low Bcl-2 expression were significantly different only in the NET group (57.1% vs. 78.8%, p = 0.032, Fig. 2B), same trend was observed in the NEC group, but not statistically different (9.5% vs. 24.2%, p = 0.090, Fig. 2C). Similarly, the 5-year DFS rate of Noxa high expression was significantly higher than that of Noxa low expression in the NET group (88.5% vs. 51.9%, p = 0.001, Fig. 2E), and NEC group (37.5% vs. 8.3%, p = 0.071, Fig. 2F), although not statistically significant. We also assessed the prognostic value of Bcl-2 and Noxa for OS in all cases. Kaplan-Meier survival analysis showed that both in the NET group and in the NEC group, patients in the low Bcl-2 expression group and high Noxa group had superior 5-year OS (Fig. 3). The expressions of Bcl-xl, Mcl-1, and PUMA were not correlated with DFS (S- Fig. 1) or OS (S- Fig. 2).

Discussion
In order to further explore prognostic biomarkers for longterm survival in GEP-NENs, our study evaluated the Bcl-2 family protein expressions in tumor tissue of GEP-NENs patients. Our results indicated that Bcl-2 and Noxa expressions correlate with tumor grade and survival prognosis. Overexpression of Bcl-2 protein and low expression of Noxa protein indicated poor tumor differentiation and poor prognosis. Bcl-2 is the fundamental member of Bcl-2 family of apoptosis and is classified as an oncogene [20]. Dysregulation of Bcl-2 family proteins has been found in a variety of tumors such as lung cancer, melanoma, and AML [21][22][23]. In this study, we found that the expression of Bcl-2 in NEC was significantly higher than that in NET, and the same trend was also observed in previous studies which focused on pancreatic NENs, suggesting that overexpression of Bcl-2 may be responsible for higher proliferation rate and more malignant phenotype of NENs [13]. This phenomenon was also observed in NENs derived from lung. The expression of Bcl-2 in small cell lung cancer (SCLC) was higher than that in typical carcinoid (TC) and atypical carcinoid (AC). What's more, several studies found a positive correlation between Bcl-2 and chromogranin A (CgA), which indicated the expression of Bcl-2 may be involved in neuroendocrine differentiation [13,21,24]. In addition, overexpression of Bcl-2 also induces resistance to chemotherapy and targeted therapies [25,26]. Currently, several Bcl-2 inhibitors, including Venetoclax, have  [27][28][29]. We found overexpression of Bcl-2 in poorly differentiated GEP-NEC, suggesting that inhibitors targeting Bcl-2 may also be a potential treatment option for these patients.
On the contrary, Noxa, which belongs to a subclass of BH3-only proteins, selectively binds to Mcl-1 and and plays a pro-apoptotic effect through the neutralization of Mcl-1 [30,31]. Previous studies have revealed that Noxa gene expression and protein function have been linked to cell death in kinds of hematopoietic and solid cancers, such as melanoma, multiple myeloma (MM), and CLL [31][32][33]. In melanoma and breast cancer studies, induced upregulation of Noxa enhanced the pharmacological effects of BH3 analogue ABT-737 [34,35]. In our study, we also observed that the expression of Noxa was significantly higher in NET, compared with NEC, and was positively associated with prognosis. Restoring or enhancing Noxa expression may significantly increase treatment efficacy and may serve as a worthwhile strategy to be explored in GEP-NENs.
This study still has some limitations. First of all, this is a retrospective study, with inherent limitations of retrospective research, such as missing some data. Secondly, given the limitations of the understanding of NENs in the early years, some patients did not receive standard treatment. Therefore, further large sample and multi-center studies are of great importance to validate these conclusions. Fig. 3 Kaplan-Meier analysis showing high Bcl-2 expression is associated with poor OS (A). When patients were grouped by NET and NEC, this relationship existed in the NET group (B), and NEC group (C). On the contrary, high Noxa expression is associated with better OS (D), and this correlation is still existed not only in NET group (E) but also in NEC group (F)

Conclusion
Taken together, our study, for the first time, systematically detected the expression of Bcl-2 family proteins in GEP-NENs and further evaluated the relationship between the expression of Bcl-2 family proteins and the prognosis of GEP-NENs patients. Our results demonstrated that Bcl-2 and Noxa were valuable and independent prognostic markers of DFS and OS in GEP-NENs. It gave us a new understanding of NENs and laid the ground for the application of drugs targeting Bcl-2 family proteins in treating NENs.