Clinicopathologic Characteristics and Prognostic Features of Well-Differentiated Pancreatic Neuroendocrine Tumors in MEN1 Syndrome: A Cohort Study of 39 MEN1 Patients and 778 Non-MEN1 Patients with 7752 Person-Year Follow-Up

1. Introduction

Multiple endocrine neoplasia type 1 (MEN1) syndrome (MIM#131100), also known as Wermer syndrome, is an autosomal dominant hereditary tumor syndrome caused by inactivating germline variants in the MEN1 tumor suppressor gene [1,2]. The syndrome is characterized by the development of multiple endocrine tumors of the parathyroid glands, pituitary gland, and foregut neuroendocrine system [3,4]. The diagnosis of MEN1 can be made based on clinical, family history, and/or genetic criteria as defined by clinical practice guidelines [5].

Well-differentiated pancreatic neuroendocrine tumors (PanNETs) have been reported to occur in 30–80% of patients with MEN1 over their lifetime and are the leading cause of MEN1-related disease-specific mortality [6,7]. The clinical behavior of MEN1-associated PanNETs presents unique management challenges. MEN1-associated PanNETs, in contrast to sporadic PanNETs, typically arise at a younger age, are more frequently multifocal, and may secrete functional hormones causing distinct clinical syndromes [8,9]. Due to the multifocality and heterogeneity of MEN1-associated PanNETs, there remains no clear consensus regarding the optimal timing, extent, and type of surgical intervention [10,11,12]. Current clinical guidelines recommend surveillance protocols for patients with MEN1 and a PanNET, with surgical intervention generally reserved for tumors exceeding 2 cm for non-functional tumors, or for functional tumors causing clinically significant hormone-related symptoms [11,13]. These recommendations are based on the premise that smaller, non-functional tumors are likely to follow an indolent clinical course, whereas larger tumors carry increased risk of metastasizing. However, the evidence supporting these guidelines derives predominantly from series with relatively smaller cohort size, variable follow-up durations, and heterogeneous patient populations [14,15].

A fundamental question that remains incompletely resolved is whether MEN1-associated PanNETs are a biologically distinct entity compared with sporadic PanNETs. Prior studies comparing MEN1-associated and sporadic PanNETs have yielded conflicting results regarding tumor size, grade, metastatic potential, and survival outcomes. Two recent multi-institutional large cohort studies suggested that survival for patients with MEN1-associated PanNETs is comparable to those non-MEN1 associated tumors [16,17]. However, both studies had relatively short-term follow-up (both 5 years), and were multi-institutional, introducing significant inter-institutional heterogeneity in pathologic assessment and lack of uniformly applied standard surgical criteria. Furthermore, direct pathologic comparison of tumor characteristics including histologic grade, patterns of invasion, and staging parameters between MEN1 and non-MEN1 tumors in large surgical series remains sparse.

Therefore, this study aims to comprehensively characterize the clinicopathologic features and long-term prognostic outcomes of surgically resected well-differentiated PanNETs in patients with MEN1compared with a large cohort of sporadic tumors, with extended follow-up from a single institution. We aimed to define whether MEN1 syndrome status independently influences tumor biology, pathologic characteristics, or long-term survival outcomes in patients undergoing surgical resection for PanNETs. Our goal is to improve the understanding of MEN1-associated PanNETs with the hope of guiding patient management.

2. Materials and Methods

This study was approved by the Johns Hopkins Institutional Review Board (IRB 00017139). Patients who underwent surgical resection for well-differentiated pancreatic neuroendocrine tumors between 1990 and 2025 at the Johns Hopkins Hospital were retrospectively identified. A total of 817 patients were included, comprising 39 patients with MEN1 syndrome and 778 patients with sporadic (non-MEN1) PanNETs. MEN1 diagnosis was established based on clinical criteria, genetic testing, or both, in accordance with established guidelines [5,18]. Patients with poorly differentiated neuroendocrine carcinomas (NECs) or mixed neuroendocrine–non-neuroendocrine neoplasms were excluded to maintain a homogeneous well-defined cohort. The electronic medical record was retrospectively reviewed for each patient to collect clinical data including demographics, MEN1 status, tumor functionality, type of surgery, and duration of follow-up. Functional tumors were defined by clinical syndromes and biochemical evidence of hormone hypersecretion [16].

The well-differentiated PanNETs were reviewed and assessed for the following histopathologic features: tumor size (T stage), histologic variant, tumor grade (Ki-67 proliferation index), lymphovascular invasion, perineural invasion, lymph node status, and pathologic stage [19]. Tumors were graded according to World Health Organization (WHO) classification criteria using mitotic activity and Ki-67 index [20]. Tumor staging followed the American Joint Committee on Cancer (AJCC) staging system and TNM staging for neuroendocrine tumor was applied [21].

Demographic and clinical variables were compared between patients with and without MEN1 diagnosis using Fisher exact tests for categorical measures and Student t test or Wilcoxon rank sum tests for continuous measures. Overall and disease-specific survival were defined as the time from surgical resection to death (from any cause or due to PanNET, respectively) or last follow-up as of 07/01/2025. Survival was estimated using the Kaplan–Meier method and compared between patients with and without MEN1 diagnosis with Cox proportional hazards models. Time to relapse was defined as the interval from surgical resection to disease progression or last follow-up, with death before progression treated as a competing event. The cumulative incidence of relapse was estimated using Fine and Gray’s method [22] and compared between patients with and without MEN1 diagnosis using a subdistribution hazards model. All time-to-event models adjust for age and sex. Two-sided p values < 0.05 were considered statistically significant. Analyses were completed using R software version 4.5.1 [23].

3. Results

3.1. Demographic and Clinical Characteristics

MEN1-associated PanNETs comprised 4.8% of the cohort (39/817). The female-to-male ratio was similar between MEN1 and non-MEN1 groups. MEN1 patients were significantly younger at the time of surgery, with a mean age of 41.8 years compared to 57.4 years of non-MEN1 patients, and 89.7% of the patients with MEN1 were younger than 60 years (p < 0.001). Most MEN1-associated PanNETs (64.1%) were located in the distal pancreas compared to 50.9% in non-MEN1 associated tumors (p <0.001). Functional PanNETs were more common in MEN1 patients (46.2% in MEN1 vs 13.5% in non-MEN1, p<0.001). Insulinomas and gastrinomas predominated, while pancreatic polypeptide–, glucagon-, ACTH-, and VIP-secreting tumors were rare. The MEN1 associated gastrinoma cohort could be divided into three categories: 1) Eight of the 11 (72.7%) had lymph node–metastatic gastrinomas with a known duodenal primary; 2) two of the 11 (18.2%) had lymph node–metastatic gastrinomas without an identifiable primary; and 3) one (9.1%) had a gastrinoma originating from the pancreas. Excluding gastrinomas from duodenal origin and metastatic to a peri-pancreatic lymph node, functional tumors were still more frequent in MEN1 patients (25.6% vs. 13.5%, p=0.04). Importantly from a surgical standpoint, R0 resection rates were similar between two groups (94.9% in MEN1 versus 91.4% in non-MEN1, p=0.42) (Table 1).

3.2. Pathological Features and Outcomes

Histological variants, including cystic and sclerosing patterns, were observed in both groups without statistically significant differences. T-stage distribution except T4, as reflected by tumor size, was virtually identical between two groups with a mean size of 2.9 cm in MEN1 and 3.1 cm in non-MEN1 (p=0.42); however, tumors larger than 2 cm were more prevalent in patients with MEN1. Grade distribution based on Ki-67 labeling indices was essentially superimposable between MEN1 and non-MEN1 tumors (p>0.99). The majority of MEN1-associated PanNETs were low-grade (grade 1 or 2). MEN1-associated PanNETs demonstrated significantly lower rates of lymphovascular and perineural invasion compared with sporadic tumors (2.6% vs 22.1%, p=0.002, and 10.3% vs 29.7%, p=0.006, respectively. After excluding the peripancreatic lymph node metastasis from duodenal primary gastrinomas, the rate of lymph node metastasis and stage III disease are similar between two groups. These findings align with previous observations that MEN1-associated duodenal gastrinomas may preferentially spread to lymph nodes despite limited local invasiveness [24] (Table 1).

With a total follow-up of 7,752 person-years, overall survival (OS) was similar between patients with and without MEN1 (Adjusted Hazard Ratio (aHR) = 1.49, 95% Confidence Interval (CI): 0.77, 2.89, p = 0.24, Table 2 and Figure 1A). Five-year overall survival from the time of surgery was 91.4% in MEN1 and 92.3% in non-MEN1, consistent with prior 5-years follow up outcome studies [16,17]. Ten-year overall survival rates from the time of surgery were 83.9% for patients with MEN1 and 79.9% for patients without MEN1. Disease-specific survival from time of surgery to death was 77.7% for patients with MEN1 and 80.1% for patients without MEN1 at 10-years (p=0.60) (Table 3 and Figure 1B). The cumulative incidence of relapse (CSS) was 0.20 [95% CI: 0.03, 0.36] for patients with MEN1 and 0.22 [95% CI: 0.19, 0.25] for patients without MEN1 at 10-years (p=0.88) (Table 4 and Figure 1C). The OS and CSS with PSW are similar with no significant statistical difference with OS of 1.59 (regression adjustment) vs 1.52 (PSW), and CSS of 1.62 (regression adjustment) vs 1.24 (PSW) (data not shown).

4. Discussion

This study is one of the largest single institution comparative analyses of surgically resected MEN1-associated and sporadic PanNETs with extended long-term follow-up (average 10 years), providing robust data to characterize the clinicopathologic features and prognostic outcomes in these two patient populations. Our findings demonstrate that while MEN1-associated PanNETs exhibit certain distinctive clinical and pathologic characteristics, including younger age at presentation, higher frequency of functional tumors, larger tumor size, higher nodal disease burden from duodenal primaries, long-term overall survival remains comparable to sporadic tumors following surgical resection.

The observation that patients with MEN1 underwent surgery at a significantly younger age than non-MEN1 patients is consistent with the natural history of hereditary tumor syndromes, in which germline deleterious variants predispose to earlier tumor development and surveillance programs facilitate earlier detection [25,26]. The predominance of functional tumors, particularly insulinomas and gastrinomas, among MEN1 patients in our cohort is consistent with the well-documented MEN1 tumor biology [3,4,27]. These functional tumors often prompt surgical intervention independent of size criteria due to refractory symptoms or incorrect medical management, potentially contributing to the surgical case mix observed in our series.

Although the average size was similar between patients with and without MEN1, a notable finding in our study is the higher proportion of tumors exceeding 2 cm among patients with MEN1 compared with sporadic cases. The higher prevalence of larger tumors in our surgical cohort may reflect selection bias inherent to guideline-concordant practice, whereby smaller MEN1-associated tumors are managed conservatively under surveillance while larger tumors are preferentially referred for resection. Alternatively, this finding could indicate that a subset of MEN1-associated PanNETs demonstrate more accelerated growth kinetics requiring intervention. A third possibility could be that MEN1 associated PanNETs begin to grow earlier in life than non-MEN1 PanNETs and thus have a growth advantage of the non-MEN1 tumors that start to grow later. Distinguishing between these possibilities requires prospective studies incorporating comprehensive surveillance data and serial imaging assessments or an extensive morphological analysis of PanNETs in large resection specimens from MEN1 patients of various ages compared to matched non-MEN1 patients.

While we observed only one primary pancreatic gastrinoma, peripancreatic lymph node metastasis of gastrinomas were much more frequent in patients with MEN1. The majority of gastrinomas in MEN1 patients arise not within the pancreatic parenchyma but rather in the duodenum, particularly within the so-called “gastrinoma triangle” [28,29,30]. These duodenal gastrinomas, which are characteristically small and often multifocal, demonstrate a well-documented propensity for early lymph node metastasis to peripancreatic and periduodenal lymph node basins, frequently disproportionate to their diminutive primary tumor size [28,29,31]. When MEN1 patients undergo pancreatic resection for concurrent PanNETs, the regional lymphadenectomy specimen may harbor metastasis originating from synchronous duodenal neuroendocrine tumors rather than from the resected pancreatic primary. The higher rate of nodal involvement may therefore reflect, at least in part, the inclusion of cases with concurrent duodenal gastrinoma-derived nodal metastasis rather than indicating a higher metastatic potential of the PanNETs themselves. Second, it is also possible that different tumor populations within MEN1 enteropancreatic axis may be contributing to the observed nodal disease burden. MEN1 often involves multiple synchronous micro- and macro-PanNETs; the “dominant” resected lesion may not always be the biologic driver of nodal spread, and nodal metastasis could reflect dissemination from one of the patient’s multiple separate primaries. The former mechanism appears to be the primary driver of the higher lymph node metastatic rate observed in our study, as the metastatic rates became comparable between two groups once we excluded positive nodes derived from known duodenal primaries as well as involved peripancreatic lymph nodes for which no primary tumor could be identified. Regardless of mechanism, the finding is clinically relevant because nodal positivity did not translate into inferior long-term survival outcomes in this cohort. From a practical standpoint, this finding underscores the importance of comprehensive preoperative evaluation in patients with MEN1 to identify synchronous duodenal tumors. Additionally, careful gross examination of specimens in MEN1 patients may help clarify the origin of nodal metastases.

In this large cohort study, we demonstrate that surgically resected MEN1-associated PanNETs share key histopathologic features with sporadic PanNETs, including tumor grade and T-stage distribution. Despite being diagnosed at a younger age, and frequently presenting with larger tumors, patients with MEN1-associated PanNETs do not exhibit inferior long-term overall survival, disease-specific survival, and cumulative incidence of relapses. These data argue against an assumption that MEN1 status alone confers worse post-resection prognosis. Instead, they reinforce that tumor grade and stage remain central anchors of risk stratification across both MEN1 and sporadic settings, at least among surgically treated, well-differentiated tumors. The results also support current strategies emphasizing individualized surgical decision-making and careful surveillance, particularly for small, low-grade lesions in patients with MEN1.

A major strength of this study is the large cohort and extended time horizon within a single institution, which reduces heterogeneity in pathology review and surgical approach compared with multi-institution studies. Several limitations of our study should be acknowledged. First, the retrospective design introduces potential selection biases, as our cohort comprises only surgically resected tumors and excludes non-surgically managed cases that may follow different clinical trajectories. Second, while our MEN1 cohort of 39 patients represents a substantial series for this rare syndrome, the sample size limits statistical power for comparing outcomes. Third, treatment heterogeneity including adjuvant therapies, management of recurrent disease, and systemic therapy for metastatic progression could not be fully standardized across the study period. Fourth, molecular alterations beyond MEN1 status were not assessed and may provide additional prognostic insight.

5. Conclusions

In conclusion, our large comparative study with extended follow-up demonstrates that surgically resected MEN1-associated PanNETs exhibit distinctive clinicopathologic characteristics including younger patient age, higher prevalence of being functional, larger tumor size, lower lymphovascular and perineural invasion, and similar rates of lymph node metastasis compared with sporadic tumors. Despite these differences, long-term overall survival is comparable between patients with and without MEN1. These findings underscore the importance of tumor grade and stage over hereditary status in prognostic assessment and support current surgical management approaches for MEN1-associated PanNETs. Future studies should incorporate molecular profiling to determine whether specific genomic features beyond MEN1 mutation status stratify prognosis within the MEN1-associated PanNET population. Additionally, prospective registries capturing comprehensive surveillance data, growth kinetics, and outcomes for conservatively managed tumors would complement surgical series in defining the natural history of MEN1-associated pancreatic neuroendocrine disease.