PAX5 expression and its diagnostic utility in gastroenteropancreatic neuroendocrine neoplasm Authors: Kun Wang, Xiaoying Liu, Jie Zhao, Zhihong Chen Published: 2025 ABSTRACT BackgroundGastroenteropancreatic neuroendocrine neoplasms (GEP-NENs) are common tumors of the digestive system. Although their histological diagnosis is usually straightforward, the differential diagnosis between neuroendocrine tumors G3 (NET G3) and neuroendocrine carcinomas (NECs) may be challenging at times due to a similar immunophenotype and morphology in a small number of cases.Aims and methodsThe objective of this study was to assess the expression and diagnostic value of PAX5 in GEP-NENs. Immunohistochemical (IHC) analysis for PAX5 was carried out in a large cohort of GEP-NENs, including 39 GEP-NECs and 62 GEP-NETs.ResultsUsing a semi-quantitative scoring method, PAX5 was shown to be expressed in 30 of 39 GEP-NECs (77%), with 27 of 39 cases (69%) having positivity ≥ 10% in the cells. In particular, GEP-NECs with a high proliferative index had more PAX5-positive cells. None of the cases of GEP-NET G1 and G2 expressed PAX5, while only 2/12 cases of GEP-NET G3 (17%) were positive for PAX5, with PAX5 positivity observed in 1–10% and 10–50% of cells, respectively. The difference in PAX5 expression between GEP-NETs and GEP-NECs was statistically significant (3% vs. 77%, P < 0.001).Conclusions PAX5 can be used as a complementary diagnostic IHC marker for GEP-NECs. In addition, PAX5 can be considered in cases when it is difficult to distinguish between GEP-NECs and GEP-NET G3. Introduction Neuroendocrine neoplasms (NENs) are a highly heterogeneous group of tumors that originate from peptidergic neurons and neuroendocrine cells and exist in all parts of the body. Gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs) account for approximately 60% of all NENs [1]. The gastrointestinal tract and pancreas are common sites of occurrence for NEMs, with their reported incidence and prevalence increasing due to likely improvements in detection and diagnosis. Over the past 40 years, the incidence of GEP-NENs has increased 3.65–4.8 fold [2]. The diagnosis of NENs usually requires a series of immunohistochemical (IHC) stains, as many traditional markers used for these tumors often lack sufficient sensitivity or specificity to be used alone. Currently, the most commonly used neuroendocrine IHC markers are synaptophysin (Syn), chromogranin (CgA), and insulinoma-associated protein 1 (INSM1) [3, 4]. However, difficulties still exist in a small but significant number of NENs, especially differentiating between GEP-NECs and GEP-NET G3, in which cell morphology, structural arrangement, and IHC staining are very similar. Previous studies have reported that p53 protein mutations and loss of Rb protein are often present in neuroendocrine carcinomas (NECs), whereas these features are absent in neuroendocrine tumors (NETs) [5–7]. Other studies have also reported that the somatostatin receptor 2 (SSTR2) protein is usually expressed in NETs but is generally not expressed in NECs [8]. In fact, there are significant variations in the results reported by different researchers, and therefore, a consensus has not yet been reached. This variability prompted us to conclude that additional IHC markers are still required. Paired box 5 (PAX5) is a specific IHC marker for B-cell lymphomas. We also found previous reports that demonstrated expression of PAX5 in poorly differentiated NECs of the gastrointestinal and pancreatobiliary tract, small cell carcinomas of the lung and Merkel cell carcinomas [9–11]. However, most of these studies focused on poorly differentiated NECs or well-differentiated NETs. The aims of the present study were therefore to assess the IHC expression of PAX5 in a series of GEP-NENs diagnosed at our institution and to determine the utility of this expression for diagnosing GEP-NENs. Materials and methods Sample selection criteria. The 101 cases of GEP-NENs included in the study were collected retrospectively from the Department of Pathology, Hunan Provincial People’s Hospital (The First Affiliated Hospital of Hunan Normal University) between January 2017 and December 2023. The study was approved by the Ethics Committee of the Hunan Provincial People’s Hospital (The First Affiliated Hospital of Hunan Normal University). Inclusion criteria Cases graded according to the 2019 WHO classification of digestive system tumors, with clear differentiation between NETs G1, G2, G3 and NECs.Formalin-fixed paraffin-embedded (FFPE) tissue blocks of sufficient quality (tissue integrity > 70%, no severe autolysis or artifacts) that allowed IHC staining and morphological analysis.Cases confirmed as GEP-NENs by histopathological examination, with complete clinical and pathological data including tumor location, size, and histological type. Cases graded according to the 2019 WHO classification of digestive system tumors, with clear differentiation between NETs G1, G2, G3 and NECs. Cases graded according to the 2019 WHO classification of digestive system tumors, with clear differentiation between NETs G1, G2, G3 and NECs. Formalin-fixed paraffin-embedded (FFPE) tissue blocks of sufficient quality (tissue integrity > 70%, no severe autolysis or artifacts) that allowed IHC staining and morphological analysis. Formalin-fixed paraffin-embedded (FFPE) tissue blocks of sufficient quality (tissue integrity > 70%, no severe autolysis or artifacts) that allowed IHC staining and morphological analysis. Cases confirmed as GEP-NENs by histopathological examination, with complete clinical and pathological data including tumor location, size, and histological type. Cases confirmed as GEP-NENs by histopathological examination, with complete clinical and pathological data including tumor location, size, and histological type. Exclusion criteria Cases with incomplete clinical or pathological data.FFPE tissues with severe degradation, insufficient tumor cells (< 500 tumor cells in the section), or extensive artifacts (e.g., edge effect, uneven staining) that might affect IHC evaluation.Cases with a history of neoadjuvant therapy (chemotherapy, radiotherapy, or targeted therapy) before tissue sampling, as therapy may alter tumor morphology or marker expression. Cases with incomplete clinical or pathological data. Cases with incomplete clinical or pathological data. FFPE tissues with severe degradation, insufficient tumor cells (< 500 tumor cells in the section), or extensive artifacts (e.g., edge effect, uneven staining) that might affect IHC evaluation. FFPE tissues with severe degradation, insufficient tumor cells (< 500 tumor cells in the section), or extensive artifacts (e.g., edge effect, uneven staining) that might affect IHC evaluation. Cases with a history of neoadjuvant therapy (chemotherapy, radiotherapy, or targeted therapy) before tissue sampling, as therapy may alter tumor morphology or marker expression. Cases with a history of neoadjuvant therapy (chemotherapy, radiotherapy, or targeted therapy) before tissue sampling, as therapy may alter tumor morphology or marker expression. Immunohistochemistry The IHC of sections from formalin-fixed paraffin-embedded tissues was performed using the EnVision two-step system. All IHC labeling was performed by a Benchmark ULTRA automated IHC station (Ventana, Roche Diagnostics, Switzerland). The ready-to use labeled antibodies for IHC were all obtained from Maixin China and included PAX5 (clone MX017, mouse monoclonal antibody, synaptophysin (Syn clone MX038), mouse monoclonal antibody, Ki67 (clone MXR002), rabbit monoclonal antibody, chromogranin A (CgA clone MX018), mouse monoclonal antibody, SSTR2 (clone EP149), rabbit monoclonal antibody, and p53 (clone MX008, mouse monoclonal antibody. The PAX5 protein staining was defined by nuclear immunoreactivity, with the other markers showing localized staining in the cell membrane, cytoplasm or nucleus. The IHC staining was assigned a semi-quantitative score of 0, 1+, 2+, or 3 + based on the proportion of the tumor cells stained, with 0 representing < 1%, 1+: 1–10%, 2+: 10–50%, and 3+: >50% [12, 13](Fig. 1). Any obvious nuclear staining was considered positive. Two pathologists independently evaluated all the sections according to the semi-quantitative scoring system and were blinded to clinical information of the patients, such as age, sex, tumor location, initial pathological diagnosis, and grading of the case. The kappa coefficient for the final statistical results was 0.939.Fig. 1Representative images of each PAX-5 staining score in gastroenteropancreatic neuroendocrine carcinomas. A 0. B 1+. C 2+. D 3+. The images are 100x magnifications Representative images of each PAX-5 staining score in gastroenteropancreatic neuroendocrine carcinomas. A 0. B 1+. C 2+. D 3+. The images are 100x magnifications Statistical analysis The data were analyzed using SPSS (v.25.0; IBM Corp, Armonk, NY, USA). Between-group comparisons of the positivity rates for the markers were performed using the chi-squared test (Fisher’s exact probability test), with p < 0.05 considered statistically significant. Results Tissue samples’ composition Of the 101 cases, 58 were male and 43 were female. The age of onset of the NETs ranged from 28 to 76 years, with a median age of 42 years, while the age of onset of the NECs ranged from 36 to 81 years, with a median age of 55 years. The NETs were graded according to the 2019 WHO classification [14] and classified as follows: 23 with WHO grade NET G1, 27 with WHO grade NET G2, 12 with WHO grade NET G3, and 39 with WHO grade NEC. Of the 101 cases, 32 cases were biopsy tissue and 69 cases were surgical resections, including 62 GEP-NETs (6 stomach, 38 pancreas, 4 ampulla, 14 colorectum), 23 GEP-NET G1 (5 stomach, 7 pancreas, 2 ampulla, 9 colorectum), 27 GEP-NET G2 (23 pancreas, 4 colorectum), 12 GEP-NET G3 (1 stomach, 8 pancreas, 2 ampulla, 1 colorectum), and 39 GEP-NECs (4 stomach, 10 pancreas, 12 ampulla, 13 colorectum) (Table 1).Table 1Tissue sample compositionGEP-NENsGEP-NECsTotalSex (male/female)26/3632/758/43Age, median (range) (y)46(28–78)61(36–81)Tumor siteG1G2G3Stomach*501410Pancreas**72381048Ampulla***2021216Colorectum****9411327Total23271239101*2 surgical resection cases and 8 biopsy cases; **38 surgical resection cases and 10 biopsy cases; ***16 surgical resection cases; ****13 surgical resection cases and 14 biopsy cases Table 1Tissue sample compositionGEP-NENsGEP-NECsTotalSex (male/female)26/3632/758/43Age, median (range) (y)46(28–78)61(36–81)Tumor siteG1G2G3Stomach*501410Pancreas**72381048Ampulla***2021216Colorectum****9411327Total23271239101*2 surgical resection cases and 8 biopsy cases; **38 surgical resection cases and 10 biopsy cases; ***16 surgical resection cases; ****13 surgical resection cases and 14 biopsy cases Tissue sample composition 46 (28–78) 61 (36–81) *2 surgical resection cases and 8 biopsy cases; **38 surgical resection cases and 10 biopsy cases; ***16 surgical resection cases; ****13 surgical resection cases and 14 biopsy cases Immunohistochemistry Findings IHC expression grouped according to tumor grade and the location of the GEP-NENs and GEP-NECs is shown in Table 2. We showed that PAX5 was positive in only 3% (2/62) of GEP-NENs (Fig. 2). Both of the positive cases were observed in NET G3. According to the semi-quantitative scoring analysis, one case was scored as 1 + and the other as 2+. Compared to GEP-NENs, we showed that PAX5 was positive in 77% (30/39) cases of GEP-NECs. According to the semi-quantitative scoring analysis, 3 cases were scored as 1+, 19 cases as 2+, and 8 cases as 3+. The expression of PAX5 in GEP-NENs and GEP-NECs was significantly different (3% vs. 77%, P < 0.001) (Table 3; Fig. 3). The results of the correlation analysis between Ki67 and PAX5 are presented in Fig. 4.Fig. 2H&E and IHC staining in gastroenteropancreatic neuroendocrine tumors and carcinomas. H&E staining of NET-G1, NET-G2, and NET-G3, respectively A, E, and I. The Syn staining of NET-G1, NET-G2, and NET-G3 was diffusely positive B, F, and J. The Ki-67 proliferation index staining of NET-G1, NET-G2, and NET-G3 was 1% and 20%, respectively C, G, and K. The PAX5 staining of NET-G1, NET-G2, and NET-G3 was negative D, H, and L. The images are 200x magnificationsFig. 3H&E staining in gastroenteropancreatic neuroendocrine carcinoma A B-D represent Syn, Ki67, and PAX5 staining, respectively. PAX5 shows diffuse positivity in the GEP-NECs. The images are 200x magnificationsFig. 4A scatter plot and correlation analysis present the correlation results between Ki67 and PAX5. The Pearson r value is 0.8191, indicating a relatively strong positive linear correlation. The 95% confidence interval for this correlation ranges from 0.6586 to 0.9084. The two-tailed P value is less than 0.0001, so the correlation is statistically significantTable 2Immunohistochemical expression grouped according to tumor grade and locationTumor gradeTumor locationNumber of casesPAX5(n, %)Syn(n, %)Cga(n, %)p53 abnormal(n, %)SSTR2(n, %)Ki-67 index range (%)NET G1Stomach505503<3%Pancreas707705Ampulla202200Colorectum909503NET G2Pancreas23023220163–20%Colorectum404201NET G3Stomach11100020–40%Pancreas807526Ampulla212121Colorectum101000OverallAll locations NETs(G1 + G2 + G3)622(3%)61(98%)49(79%)4(10%)35(76%)NECStomach43223050–90%Pancreas108101056Ampulla12912973Colorectum131012492OverallAll locationsNECs3930(77%)36(92)25(64%)24(61%)11(35%)Table 3PAX5 expression in GEP-NENs and GEP-NECsPAX-5NENsNECsG3vsNECsNET G1NET G2NET G3θP < 0.001Grade++++++++++++Positive001103198Negative2327109θThe values represent significantly different positive staining, NET G3 vs. NECs (Fisher’s exact test) H&E and IHC staining in gastroenteropancreatic neuroendocrine tumors and carcinomas. H&E staining of NET-G1, NET-G2, and NET-G3, respectively A, E, and I. The Syn staining of NET-G1, NET-G2, and NET-G3 was diffusely positive B, F, and J. The Ki-67 proliferation index staining of NET-G1, NET-G2, and NET-G3 was 1% and 20%, respectively C, G, and K. The PAX5 staining of NET-G1, NET-G2, and NET-G3 was negative D, H, and L. The images are 200x magnifications H&E staining in gastroenteropancreatic neuroendocrine carcinoma A B-D represent Syn, Ki67, and PAX5 staining, respectively. PAX5 shows diffuse positivity in the GEP-NECs. The images are 200x magnifications A scatter plot and correlation analysis present the correlation results between Ki67 and PAX5. The Pearson r value is 0.8191, indicating a relatively strong positive linear correlation. The 95% confidence interval for this correlation ranges from 0.6586 to 0.9084. The two-tailed P value is less than 0.0001, so the correlation is statistically significant Table 2Immunohistochemical expression grouped according to tumor grade and locationTumor gradeTumor locationNumber of casesPAX5(n, %)Syn(n, %)Cga(n, %)p53 abnormal(n, %)SSTR2(n, %)Ki-67 index range (%)NET G1Stomach505503<3%Pancreas707705Ampulla202200Colorectum909503NET G2Pancreas23023220163–20%Colorectum404201NET G3Stomach11100020–40%Pancreas807526Ampulla212121Colorectum101000OverallAll locations NETs(G1 + G2 + G3)622(3%)61(98%)49(79%)4(10%)35(76%)NECStomach43223050–90%Pancreas108101056Ampulla12912973Colorectum131012492OverallAll locationsNECs3930(77%)36(92)25(64%)24(61%)11(35%) Immunohistochemical expression grouped according to tumor grade and location All locations NECs Table 3PAX5 expression in GEP-NENs and GEP-NECsPAX-5NENsNECsG3vsNECsNET G1NET G2NET G3θP < 0.001Grade++++++++++++Positive001103198Negative2327109θThe values represent significantly different positive staining, NET G3 vs. NECs (Fisher’s exact test) PAX5 expression in GEP-NENs and GEP-NECs θThe values represent significantly different positive staining, NET G3 vs. NECs (Fisher’s exact test) Discussion NENs are most commonly seen in the gastrointestinal system, but may also occur sporadically in other sites. The cell morphology and structural arrangement of NENs are diverse, which makes IHC testing necessary for their diagnosis, especially in atypical cases. It is well known that regardless of the site of origin, NETs are characterized by robust and consistent expression of neuroendocrine markers, including synaptophysin (Syn), chromogranin A (Cga), and the recently identified INSM1 [3]. However, these traditional markers can only confirm a neuroendocrine origin and cannot assist in tumor grading. Specifically, this limitation poses significant challenges in differentiating between NET G3 and NECs, as the two entities often have overlapping morphology and IHC. Given their distinct treatment strategies and clinical outcomes, accurate differentiation is therefore critical for their clinical management. Previous studies have reported that p53 and Rb are the most reliable IHC markers for making this distinction [15]. In addition, SSTR2 is expressed commonly in well-differentiated NETs but is reduced significantly in poorly differentiated carcinomas, making it another useful marker for distinguishing between NET G3 and NECs [6, 8]. The molecular differences between GEP-NET G3 and NECs are as follows [5, 16, 17]: NET G3 frequently harbors mutations in MEN1, DAXX/ATRX, and mTOR pathway genes (e.g., TSC2, PTEN), whereas NECs mainly have mutations in TP53 and RB1, with higher mutation burdens, and also often involve KRAS, PIK3CA/PTEN, and BRAF. NET G3 is associated with dysregulation of the MEN1, DAXX/ATRX, and PI3K/mTOR pathways, while NECs are linked to dysregulation of the Rb/E2F pathway. Similarly, the current study showed that the positivity rate of p53 in NET G3 was significantly lower compared to that in NECs (10% vs. 61%, P < 0.001). The positivity rate of SSTR2 was also significantly higher in NET G3 compared to that in NECs (76% vs. 35%, P < 0.05). These findings therefore reconfirm the diagnostic utility of p53 and SSTR2 IHC to distinguish between NET G3 and NECs. Although p53, Rb, and SSTR2 IHC usually help pathologists make the correct diagnosis, a small number of cases are difficult to classify, and therefore, additional practical markers are needed to facilitate accurate differential diagnosis. We observed that PAX5 was not only expressed in haematopoietic neoplasms, but also in alveolar rhabdomyosarcoma, Wilms tumor, small cell carcinoma of lung and Merkel cell carcinoma [9, 10, 13]. In addition, previous studies have shown that PAX5 is expressed in poorly differentiated NECs of the gastrointestinal tract and pancreaticobiliary ducts [11], as well as in the cytoplasmic expression of well-differentiated NETs of the rectum [18]. The paired box (PAX) family is a gene family that encodes nine transcription factors, which play important roles in multiple stages of embryonic development, human organogenesis, and development of the central nervous system [19, 20]. The PAX5 gene, one of nine members of the PAX family, is located on chromosome 9q13 and plays a key role in B-cell development. Recent studies using RT-PCR have demonstrated that PAX5 shows significantly different expression in colorectal NETs and colorectal NECs [21]. It is also likely that PAX5 is an important transcription factor that regulates the expression of neurogenic genes and exhibits specificity in induced neuroendocrine-like prostate cancer (t-NEPC) [22]. However, the aforementioned studies only focused on poorly differentiated NECs or well-differentiated NETs. To our knowledge, our study is the first to evaluate PAX5 expression in a large series of GEP-NENs. The results of immunohistochemistry showed a high percentage of PAX5 positivity in NECs, regardless of the tumor location, and a low percentage of PAX5 positivity in NET G3. None of the NETs G1 or NETs G2 cases tested positive for PAX5. Additionally, We also showed that in these cases, the percentage of PAX5-positive tumor cells correlated positively with the Ki67 proliferation index. Only 2 cases of NET G3were positive for PAX5, with staining scores of 1 + and 2+, respectively. The corresponding Ki67 proliferation index for these two cases was approximately 40%. Previous studies have suggested that the expression of PAX5 is associated with tumor aggressiveness in pulmonary NETs [23]. Specifically, the expression level of PAX5 reflects the tumor grade and proliferative activity, both of which increase as the tumor grade, Ki67 proliferation index, and the proportion of tumor cells with positive PAX5 rise [24]. This relationship suggests that PAX5 can be a useful biomarker for assessing the aggressiveness and proliferative potential of NETs, aiding in their classification and potentially guiding treatment strategies. On the other hand, an interesting finding in our study was the low sensitivity of chromogranin A(CgA) expression in colorectal NETs, especially in colorectal NECs. We showed that only 11 of 27 colorectal NEN cases, were positive for CgA. Jisup et al. [25] demonstrated that the sensitivity of CgA expression in rectal NETs was poor, a similar result to that of ours. Furthermore, they showed that CgA expression in rectal NETs was associated with more aggressive clinical behavior and a poorer prognosis. However, in our study, we found that 10 of 13 cases of colorectum NECs were positive for PAX5 and that compared with CgA, PAX5 expression was more robust in these tumors. Therefore, PAX5 could be used as a complementary diagnosis IHC marker in colorectum NECs when CgA expression is not ideal. In addition to compensating for the shortcomings of traditional markers for NECs at specific sites (such as the colorectum), the expression characteristics of PAX5 also have unique value for differentiating between tumors with similar morphologies. This advantage is due to PAX5 being expressed in a variety of small round cell tumors, with the pathological diagnosis of these tumors often facing challenges due to sample limitations, such as core needle biopsies (CNB). It is well known that PAX5 has been widely used as a diagnostic marker for evaluating B-cell lymphoma. PAX5 is also expressed in alveolar rhabdomyosarcoma, Wilms tumors, Merkel carcinoma, and GEP-NECs. These neoplasms frequently exhibit overlapping morphological features of small round cells, which complicates their distinction. Considering the high expression rate of PAX5 in GEP-NECs, it is important to make appropriate differential diagnoses between GEP-NECs and the other types of neoplasms, particularly in CNB and biopsy tissues. Compared with complete surgical excision, CNB and biopsy samples are often limited in quantity and quality, making the diagnosis less conclusive and thereby increasing the risk of a misdiagnosis [26]. However, as an alternative to surgical excision for pathological diagnosis, CNBs and biopsies have been widely accepted over the past decades. As a consequence, clearer guidance for pathologists is essential regarding the interpretation of PAX5 expression in the setting of morphologically similar tumors. Limitations and future directions. The study had several limitations that should be acknowledged. First, it was a single-center retrospective design with a relatively small sample size in each subgroup. Data collected from a single institution may have introduced selection bias and limited the generalizability of the findings. Multi-center studies are therefore needed to verify the stability of PAX5 expression. Second, the molecular mechanisms underlying the expression patterns of markers such as PAX5 in GEP-NENs were not investigated, which limited the in-depth understanding of their biological functions. Moreover, the potential association between marker expression and patient prognosis, including overall survival and progression-free survival, was not analyzed, meaning that the prognostic implications of PAX5 expression in GEP-NECs remained uninvestigated. While the results of this study are mainly relevant to the daily diagnostic work of pathologists, it is necessary for future studies to investigate the mechanism of expression and prognostic value of PAX5 in GEP-NECs. Conclusion In conclusion, this study demonstrates that PAX5 is expressed frequently in GEP-NECs but rarely in GEP-NETs, with no expression in NET G1/G2 and only occasional positivity in NET G3. These distinct expression patterns identify PAX5 as a valuable complementary IHC marker for GEP-NECs, particularly in challenging differential diagnoses between GEP-NECs and GEP-NET G3. Importantly, in order to optimize the accuracy of GEP-NENs diagnosis, PAX5 should be used in conjunction with established markers (e.g., Ki67, p53, and SSTR2) rather than being used as a standalone test.