Histopathological features of gastrointestinal and pancreatic neuroendocrine neoplasms

ABSTRACT

Objectives:

To evaluate the clinicopathological features of gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs) over 11 years at a single tertiary institution in Saudi Arabia.

Methods:

In a retrospective review, data from 71 patients diagnosed with GEP-NEN, at a tertiary academic center in Jeddah, Saudi Arabia, between January 2014 to December 2024, were analyzed for demographics, clinical presentation, tumor location, size, pathology, grade, and immunohistochemistry. The statistical analyses included Chi-square, Mann-Whitney-U, and Spearman’s correlation tests.

Results:

The cohort had a mean age of 54.9±15.1 years and comprised 62% males. Most tumors originated in the foregut (62.0%) and were unifocal (83.1%). Abdominal pain (29.6%) and weight loss (9.9%) were the most common symptoms, and 18.2% of cases were incidentally detected. Well-differentiated neuroendocrine tumors represented 77.5% of the cases, with grade 1 being the most prevalent (60.0%). Synaptophysin was detected in 100% of patients and chromogranin A was detected in 90.1% of patients. The tumor size was significantly positively correlated with the tumor grade (r=0.57, p=0.001), whereas tumor grade was not significantly associated with gender or other clinicopathological features.

Conclusion:

Well-differentiated neuroendocrine tumors originating in the foregut were predominant in this cohort. These findings provide important regional insights into GEP-NENs and support the need for further research to enhance the diagnostic and treatment strategies.

Neuroendocrine neoplasms (NENs) originate from neuroendocrine cells distributed throughout the body, and approximately two-thirds of NENs are located in the gastrointestinal tract and pancreas.^1,2 Gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs) are rare. However, their incidence has recently increased, largely owing to advances in diagnostic modalities and heightened clinical awareness.³ The estimated annual incidence of GEP-NENs is 1-2 per 100,000 individuals and constitutes 2% of all gastrointestinal tract malignancies.^3,4

Neuroendocrine neoplasms may arise in various anatomical locations and can produce diverse hormones, leading to variable clinical presentations. Most are nonfunctional and present with nonspecific symptoms such as abdominal pain or gastrointestinal bleeding. According to the current World Health Organization (WHO) classification, NENs are categorized as well-differentiated neuroendocrine tumors (WDNETs), poorly differentiated neuroendocrine carcinomas (PDNECs), and mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs), the latter comprising both components, each accounting for at least 30% of the tumor.^5,6

The MiNENs are often aggressive and require a multidisciplinary diagnostic and therapeutic approach. However, data on their incidence and behavior remain limited.⁵

Grading of GEP-WDNETs, according to the WHO criteria, is based on 2 key proliferative indices: the mitotic rate, expressed as mitoses per 10 high-power microscopic fields (HPFs) or per 2 mm², and the Ki-67 index, which quantifies the proportion of neoplastic cells immunolabeled for the Ki-67 proliferation marker. Tumors are classified as grade 1 (G1: <3%), grade 2 (G2: 3-20%), or grade 3 (G3: >20%) based on the Ki-67 index. Similarly, mitotic rates <2 per 10 HPFs correspond to G1, 2-20/10 HPFs to G2, and >20/10 HPFs to G3.⁷

This study evaluated the clinicopathological and demographic features of patients diagnosed with GEP-NENs over an 11-year period at a tertiary academic institution in Saudi Arabia. Regional data on the pathological spectrum of GEP-NENs have remained limited to date, making the present analysis a valuable contribution to the understanding of these rare neoplasms in the local context.

Methods

This retrospective study was carried out in the Pathology Section of the Department of Clinical Laboratories at a large academic medical center in Jeddah, Saudi Arabia. This study included all patients diagnosed with GEP-NENs between January 2014 and December 2024.

The cases were identified using a pathology laboratory database and verified using archived pathology reports and electronic medical records. Eligible cases included endoscopic biopsy and surgical resection specimens with a histopathological diagnosis of gastrointestinal or pancreatic NENs. The following parameters were collected based on the WHO criteria: age, gender, nationality, clinical presentation, tumor location, tumor size, pathological diagnosis, immunohistochemical findings, and tumor grade.

This study was approved by the institutional review board of Biomedical Ethics (IRB approval number: 337-24). All data were anonymized and handled in accordance with the institutional and ethical guidelines.

Statistical analysis

Data analysis was carried out using the Statistical Package for the Social Sciences, version 26.0 (IBM Corp., Armonk, NY, USA). Categorical variables were summarized as frequencies and percentages. Associations between categorical variables (such as gender, clinical presentation, tumor focality, embryological origin, pathological classification, and tumor grade) were evaluated using the Chi-square test. Continuous variables were reported as means ± standard deviation (SD), and group comparisons (namely, age or tumor size across grades) were analyzed using the Mann-Whitney-U or Kruskal-Wallis tests, as appropriate. Correlations between continuous variables (namely, tumor size, age, and tumor grade) were assessed using Spearman’s rank correlation coefficient. A p-value of <0.05 was considered significant.

A tumor size of 4 cm was used as a categorical cut-off, as lesions ≥4 cm have been associated in previous literature with more aggressive behavior and higher risk of metastasis in neuroendocrine tumors.⁸

Results

A total of 71 patients with GEP-NEN were included in this study. The majority (62.0%) were over the age of 50, with a mean age of 54.9±15.1 years. Most patients were male (62.0%), and 59.2% were Saudi nationals (Table 1).

Table 1.

- Patients’ demographic characteristics (N=71).

Variables	n (%)
Age (years)
<30	4 (5.6)
30-50	23 (32.4)
>50	44 (62.0)
Gender
Female	27 (38.0)
Male	44 (62.0)
Nationality
Non-Saudi	29 (40.8)
Saudi	42 (59.2)

Values are presented as numbers and percentages (%)

The tumors most commonly originated in the foregut (62.0%), followed by the midgut (23.9%), and hindgut (14.1%). The tumor size exceeded 4 cm in 28.2% of cases, while 30.9% had an undocumented size owing to the limited nature of endoscopic biopsy specimens versus resection samples. The mean tumor size was 4.01±2.73 cm. Most tumors were unifocal (83.1%), and 45.1% had metastasized at the time of diagnosis. Pathologically, 77.5% were WDNETs, 18.3% were PDNECs, and 4.2% were MiNENs. Chromogranin A was positive in 90.1% of cases, while synaptophysin was positive in all tumors (Table 2).

Table 2.

- Tumor’s pathological characteristics (N=71).

Variables	n (%)
Embryological origin
Foregut	44 (62.0)
Hindgut	10 (14.1)
Midgut	17 (23.9)
Size (cm)
≤2	14 (19.7)
2.1-4	16 (22.5)
>4	20 (28.2)
Not mentioned	21 (30.9)
Focality
Unifocal	59 (83.1)
Multifocal	12 (16.9)
Pathological classification
WDNET	55 (77.5)
PDNEC	13 (18.3)
MiNENs	3 (4.2)
Immunostaining
Chromogranin A
Positive	64 (90.1)
Negative	7 (9.9)
Synaptophysin
Positive	71 (100)
Metastasis
Present	32 (45.1)
Absent	39 (54.9)

Values are presented as numbers and percentages (%).

WDNET: well-differentiated neuroendocrine tumor,

PDNEC: poorly differentiated neuroendocrine carcinoma,

MiNENs: mixed neuroendocrine-non-neuroendocrine neoplasms

Among the 55 WDNETs, grade 1 tumors were the most common (60.0%), followed by grade 2 (38.2%) and grade 3 (1.8%) tumors (Table 3).

Table 3.

- Distribution of well-differentiated neuroendocrine tumors’ grades (n=55).

Variables	n (%)
Grade 1	33 (60.0)
Grade 2	21 (38.2)
Grade 3	1 (1.8)

Values are presented as numbers and percentages (%). Based on WHO 2019 grading criteria.

No statistically significant associations were found between gender and clinical presentation, embryological origin, tumor size, focality, metastasis, pathological classification, or tumor grade (p<0.05). For example, while PDNECs (20.5%) and multifocal (20.5%) tumors were more frequent in male patients than in female patients (PDNECs: 14.8% and multifocal: 11.1%), the differences were not significant. Similarly, the chromogranin A and synaptophysin expression did not differ significantly between male and female patients (Appendix 1).

In the WDNETs subgroup, the tumor grade was significantly associated with multiple clinical and pathological variables. Grade 2 tumors were more frequently associated with abdominal pain (p=0.002), incidental discovery (p=0.008), and bowel obstruction (p=0.001). The tumor grade was also significantly related to embryological origin (p<0.001), with grade 1 tumors more commonly arising from the foregut or midgut, and only grade 3 tumors originating from the hindgut. The tumor size also significantly correlated with the grade: grade 2 tumors were more likely to measure >4 cm or fall within the 2.1-4 cm range, while grade 1 tumors were predominantly ≤2 cm (p=0.001). The mean tumor size for grade 2 tumors was 4.15±1.40 cm, significantly larger than the 2.45±1.43 cm observed for grade 1 tumors (p=0.001). No statistically significant associations were found between the tumor grade and focality or the presence of metastasis. Chromogranin A positivity did not vary according to the grade, and synaptophysin was universally expressed (Table 4).

Table 4.

- Relationship between well-differentiated neuroendocrine tumor grade and clinical presentation, tumor embryological origin, size, focality, and metastasis at diagnosis (n=55).

Variables	Tumor grades			χ2	P-values
	Grade 1	Grade 2	Grade 3
Clinical presentation *
Abdominal pain	9 (52.9)	8 (47.1)	0 (0.0)	12.88	0.002 ‡
Incidental finding	7 (58.3)	5 (41.7)	0 (0.0)	9.75	0.008 ‡
Gastric ulcer	2 (100)	0 (0.0)	0 (0.0)	6.00	0.05
Other/mixed symptoms†	6 (42.9)	7 (50.0)	1 (7.1)	6.64	0.036 ‡
Anemia/IDA	3 (75.0)	1 (25.0)	0 (0.0)	5.25	0.072
Bowel obstruction	5 (100)	0 (0.0)	0 (0.0)	15.00	0.001‡
Vomiting	1 (100)	0 (0.0)	0 (0.0)	3.00	0.223
Embryological origin *
Foregut	16 (53.3)	14 (46.7)	0 (0.0)	22.80	0.00‡
Hindgut	5 (55.6)	3 (33.3)	1 (11.1)	4.00	0.135
Midgut	11 (73.3)	4 (26.7)	0 (0.0)	18.60	0.00‡
Size (cm) *
≤2	13 (92.9)	1 (7.1)	0 (0.0)	13.70	0.001‡
2.1-4	8 (61.5)	5 (38.5)	0 (0.0)
>4	3 (23.1)	10 (76.9)	0 (0.0)
Not mentioned	10 (58.8)	6 (35.3)	1 (5.9)
Focality
Multifocal	7 (21.9)	3 (14.3)	0 (0.0)	0.64	0.726
Unifocal	26 (78.1)	18 (85.7)	1 (100)
Metastasis
Absent	20 (60.6)	11 (52.4)	1 (100)	1.09	0.580
Present	13 (39.4)	10 (47.6)	0 (0.0)
Immunostaining§
Chromogranin A
Positive	31 (93.9)	20 (95.2)	1 (100)	0.10	0.951
Negative	2 (6.1)	1 (4.8)	0 (0.0)

Values are presented as numbers and percentages (%).

^*For clinical presentation, embryological origin, and size (categorical), the percentages were calculated using the total number of patients in each row as the denominator (namely, row-wise percentages).

^† Some patients presented with multiple symptoms.

^§ Synaptophysin immunostaining was positive in all cases across tumor grades; hence, statistical testing was not applicable.

^‡ Indicate statistical significance at p<0.05. IDA: iron deficiency anemia

Spearman’s correlation analysis showed no significant association between age and tumor size (r= -0.02, p=0.881) or tumor grade (r= -0.26, p=0.051). In contrast, a statistically significant moderately positive correlation was identified between the tumor size and the tumor grade (r=0.57, p=0.001), supporting the finding that larger tumors tend to have a higher histologic grade. Although the association between age and tumor grade did not reach statistical significance (p=0.051), the result was marginal and suggested a possible trend that may reach significance in larger studies.

Discussion

The GEP-NENs are a rare category of tumors derived from cells in the gastrointestinal tract and the pancreatic endocrine system.^1,2 The annual incidence of GEP-NENs in Saudi Arabia has not been well-studied to date. However, these tumors are thought to be rare, as is also the case in other parts of the world. This retrospective study identified only 71 cases over an 11-year period, highlighting their scarcity. Studies carried out in other countries have reported a male predominance, a trend also observed in our institutional population (62.0%).^9-11 In this series, 62.0% of patients were over 50 years of age, a finding that aligns with much of the existing literature, including a prominent study by Akin et al.¹⁰ However, this observation contrasts with the findings of other studies, such as that of Rafique et al.¹²

The GEP-NENs can present as hormonally active or inactive tumors with clinical manifestations that vary depending on the tumor site of origin. However, these tumors frequently remain asymptomatic for prolonged periods. Raphael et al¹³ reported that 9-39% of gastrointestinal neuroendocrine tumors are asymptomatic. Furthermore, 87% of symptomatic patients exhibit nonspecific symptoms.¹³ In this study, 14 (19.7%) patients were asymptomatic, consistent with the published ranges. The symptoms reported in the literature include abdominal discomfort in 28-79% of patients, bowel obstruction in 18-24%, diarrhea in 10-32%, weight loss in 2-58%, and gastrointestinal bleeding in 4-10%.^13-15 Among the symptomatic patients in this study, 23 (33.8%) experienced abdominal pain and 8 (11.8%) reported weight loss. The less frequently observed symptoms included vomiting, bowel obstruction, rectal bleeding, melena, anemia, jaundice, constipation, and dysphagia.

The most common locations of gastrointestinal neuroendocrine neoplasms, excluding the pancreas, are the appendix and ileum. However, a recent case series from a single institution in the southern region of Saudi Arabia reported that the most common site was the pancreas (29.1%), and the least common was the appendix (6.9%).^16-19 In this study population, most tumors originated in the foregut (62.0%), including 15 in the stomach, 13 in the pancreas, and 11 in the liver. The hindgut accounted for 10 cases, including tumors of the rectosigmoid and anal canals, while the appendix was the least prevalent site, with only 3 cases. These findings suggest that the distribution of GEP-NENs varies significantly between countries, likely owing to differences in ethnicity, race, and geographical location. The foregut predominance observed in the studied cohort may be influenced by several factors. First, upper gastrointestinal symptoms tend to prompt earlier endoscopic evaluation, increasing the likelihood of detecting foregut lesions. Second, dietary habits and high rates of Helicobacter pylori infection in certain populations may contribute to increased gastric NET incidence. Third, genetic predispositions and regional differences in healthcare-seeking behaviour may also play a role in the anatomical distribution observed. Further multicenter research is needed to determine whether these patterns are unique to our region or represent broader trends. Albishi et al¹⁹ reported that 54.2% of patients presented with metastatic disease at diagnosis, which is higher than what has been reported in other regions. In this study, 45.1% of the patients had metastases at the time of diagnosis, and this slightly lower rate may reflect differences in tumor biology, earlier detection, or better healthcare access for our institution’s population.

Unlike in the lungs, where poorly differentiated neuroendocrine carcinomas are more common, WDNETs are the predominant form of NENs in the gastrointestinal and pancreatobiliary tracts. In a study of the Turkish population, Akin et al¹⁰ found that grade 1 NETs constituted the majority of cases (61.8%), followed by grade 2 NETs (18.8%), and grade 3 NETs (19.4%), which included both grade 3 NET and NEC cases. These findings align with the results, in which grade 1 tumors comprised 60.0% of the WDNETs, followed by grade 2 (38.2%), and only one grade 3 tumor (1.8%). Akin et al¹⁰ also examined the correlations between the tumor grade and variables, such as gender, age, pT stage, and tumor location, demonstrating a significant association between tumor location and grade. However, in the current study, no significant correlations were found between the tumor grade and gender, age, or tumor location.

Only 3 cases of MiNENs were identified in our population, each arising from a distinct site. The first case was in the esophagus, where the tumor coexisted with squamous cell carcinoma. The second was in the pancreas, which was mixed with pancreatic ductal adenocarcinoma, although pancreatic MiNENs are more commonly associated with acinar cell carcinoma and pancreatoblastoma.^20,21 The final case involved the liver, where the tumor was mixed with cholangiocarcinoma. The MiNENs continue to pose diagnostic and therapeutic challenges owing to their rarity and complex histologies.

Similar to other neuroendocrine tumors, GEP-NENs typically stain positively for neuroendocrine markers including synaptophysin, chromogranin A, CD56, and the newer marker INSM-1. Although our laboratory does not use INSM-1, our pathologists rely primarily on synaptophysin and chromogranin A to diagnose GEP-NENs. In this study, synaptophysin was detected in 100% of tumors, confirming its utility as a sensitive marker. Chromogranin A was detected in 90.1% of cases. These findings are consistent with those reported in the literature and support the reliability of these markers for routine diagnostics.

The study also demonstrated that the tumor size was significantly associated with the tumor grade, a relationship supported by both cross-tabulation and Spearman’s correlation analyses (r=0.57, p=0.001). Larger tumors were more likely to be of higher grade, with grade 2 tumors having a mean size of 4.15±1.40 cm, compared to 2.45±1.43 cm for grade 1 tumors. This reinforces the prognostic utility of tumor size and may have implications for clinical decision-making. For instance, patients presenting with tumors larger than 2 cm (especially those exceeding 4 cm) might benefit from more aggressive surveillance, earlier surgical intervention, or closer multidisciplinary assessment. These findings support the potential role of tumor size as a surrogate marker for biological behaviour, emphasizing its relevance in risk stratification and management planning.

No statistically significant differences were observed between the male and female patients in terms of clinical presentation, tumor size, embryological origin, focality, metastasis, or pathological classification. This highlights the gender-independent behavior of these tumors in our population.

Study’s limitations

Its retrospective design and execution at a single institution. The modest cohort size limits the statistical power and generalizability of the findings. Additionally, the low number of hindgut tumors and MiNEN cases limits our ability to draw robust subtype-specific conclusions. The high percentage of biopsy-based specimens also limits full histological evaluation, including accurate tumor sizing in some cases.

In conclusion, the stomach, pancreas, liver, and rectosigmoid represented the predominant sites of GEP-NENs in the studied cohort. Well-differentiated tumors were the most common subtype, and the tumor size showed a significant correlation with the histologic grade. No significant associations were observed between the tumor grade and patient age, gender, or focality. These findings provide valuable regional insights into the clinicopathological spectrum of GEP-NENs and underscore the importance of continued multicenter research to inform diagnostic pathways and optimise patient management.

Appendix 1. - Gender-based distribution of clinical and pathological features of gastroenteropancreatic neuroendocrine neoplasm patients (N=71)

Variables	Gender		χ2	P-values
	Female	Male
Clinical presentation
Abdominal pain	7 (25.9)	13 (29.5)	0.01	0.954
Vomiting	0 (0.0)	1 (2.3)	0.01	1.000
Weight loss	2 (7.4)	5 (11.4)	0.02	0.894
Incidental finding	4 (14.8)	8 (18.2)	0.01	0.967
Gastric ulcer	1 (3.7)	1 (2.3)	0.01	1.000
PR bleeding	0 (0.0)	2 (4.5)	0.15	0.700
Jaundice	0 (0.0)	2 (4.5)	0.15	0.700
Anemia/IDA	1 (3.7)	1 (2.3)	0.01	1.000
Bowel obstruction	1 (3.7)	1 (2.3)	0.01	1.000
Constipation	0 (0.0)	1 (2.3)	0.01	1.000
Other/mixed symptoms†	18 (66.7)	23 (52.3)	0.89	0.345
Embryological origin
Foregut	20 (74.1)	24 (54.5)
Hindgut	3 (11.1)	7 (15.9)	2.82	0.244
Midgut	4 (14.8)	13 (29.5)
Size (cm)
≤2	7 (25.9)	7 (15.9)	2.37	0.499
2.1-4	5 (18.5)	11 (25.0)
>4	9 (33.3)	11 (25.0)
Not mentioned	6 (22.2)	15 (34.1)
Focality
Multifocal	3 (11.1)	9 (20.5)	0.48	0.488
Unifocal	24 (88.9)	35 (79.5)
Metastasis
Absent	14 (51.9)	25 (56.8)	0.03	0.871
Present	13 (48.1)	19 (43.2)
Pathological classification
WDNETs	22 (81.5)	33 (75.0)	0.41	0.815
PDNECs	4 (14.8)	9 (20.5)
MINENs	1 (3.7)	2 (4.5)
Tumor grade (WDNET, n=55)
Grade 1	14 (63.6)	19 (57.6)	0.78	0.677
Grade 2	8 (36.4)	13 (39.4)
Grade 3	0 (0.0)	1 (3.0)
Immunostaining
Chromogranin A
Positive	25 (92.6)	39 (88.6)	0.02	0.894
Negative	2 (7.4)	5 (11.4)
Synaptophysin
Positive	27 (38.0)	44 (62.0)	0.00	1.000

Values are presented as numbers and percentages (%). No statistically significant differences were observed.

^† Some patients presented with multiple symptoms. IDA: iron deficiency anemia, WDNET: well-differentiated neuroendocrine tumor,

PDNEC: poorly-differentiated neuroendocrine carcinoma, MiNENs: mixed neuroendocrine–non-neuroendocrine neoplasms