Plasma Polyamines as an Additional to Imaging Biomarker in MEN1 Patients With Duodenopancreatic Neuroendocrine Tumors.

Multiple endocrine neoplasia type 1 (MEN1) is an inherited tumor endocrine syndrome caused by pathogenic variants of the MEN1 gene, mainly involving the parathyroids, anterior pituitary, and pancreas (1). Duodenopancreatic neuroendocrine tumors (dpNETs) constitute an important clinical entity of MEN1 as 70% to 80% of patients will develop a dpNET that can be nonfunctioning or secrete gastrin, insulin, or other hormones leading to distinct syndromes (2). The majority of dpNETs are nonfunctioning pancreatic NETs (NFpNETs) and have a 20% probability of developing metastatic disease that is more evident when they reach a size of 2 cm or greater or exhibit a high growth rate (2-4). Early detection and continuous follow-up are required because surgery is the only curative treatment for localized dpNETs (1, 2). Current evidence suggests initiating radiological screening at approximately age 10 years, preferentially with magnetic resonance imaging, and when required with endoscopic ultrasound (EUS) or earlier according to the presence of clinical symptoms (1, 2). Screening should be performed every 1 to 3 years in patients without dpNETs and more frequently in patients with symptoms related to hormonal secretion or evident tumors (1, 2). This approach, besides the inconvenience and requirement for continuous patient adherence, entails exposure to the risks of gadolinium deposits in several tissues and EUS-related complications (3). Currently existing serum biomarkers do not have the necessary negative predictive value to replace imaging as a mean of detecting patients at risk for progression or metastatic disease, making imperative the search for new biomarker classes (2, 3).

Polyamines are polycationic alkylamines with tumor-promoting actions that are associated with disease progression in several cancers (5).

Fahrmann and colleagues (6) evaluated the prognostic value of blood-based polyamines as a means for monitoring patients with MEN1-related dpNETs for harboring or developing aggressive disease. For initial testing, plasma samples were obtained from 14 MEN1 patients with liver metastases and 2 types of controls: MEN1 patients with nonmetastatic (distant or regional) indolent dpNET (n = 28; control group 1), and MEN1 patients without dpNET (n = 14; control group 2) (6). The presence of metastases was confirmed either through histological examination or morphological and functional imaging. Nine patients with metastases had a functional tumor (7 gastrinomas), 11 a tumor size smaller than 2 cm, and 3 had previously received several forms of treatment, whereas a further one was on active treatment. Almost all controls had nonfunctioning dpNETs and tumor size was smaller than 2 cm in group 1 controls. The absence of distant or regional metastases for control group 1 was based on a minimum 3-year follow-up period after the diagnosis of dpNET and imaging performed one or more years after blood draw. Control group 2 had no imaging-verified dpNET at least 6 months or more after blood draw.

Measurement of plasma polyamines was conducted using mass spectrometry (6). Of the 5 quantifiable polyamines, 3 (NAcPut, AcSpmd, and DiAcSpmd) were significantly elevated in the plasma of metastatic cases compared to controls, especially in patients with concurrent gastrinoma. Using logistic regression models, this 3MP model was 28.6% sensitive and 95% specific in distinguishing cases from controls. The discrimination performance of individual polyamines as well as the combined 3MP was further assessed in an independent test set of plasmas from 6 metastatic cases, 5 of whom were on active treatment with various regimens, and 22 controls showing a 66.7% sensitivity and 95% specificity. The 3MP signature identified MEN1 cases without a dpNET with 66.7% sensitivity and 95% specificity, respectively.

To further validate their findings, the authors studied a cohort of Men1Pdx1-Cre mice that mirror MEN1 disease, by progressing from hyperplastic islets (5-6 months) to insulinomas (10-12 months) (7). They evaluated time-dependent changes in plasma polyamine levels throughout MEN1-pNET disease progression. When the 3MP model developed in the human cohort was applied to the Men1Pdx1-Cre mice, they showed an increase in levels in comparison to Men1 controls, starting at four-months, remaining elevated throughout disease progression (6). In addition, elevated polyamines were associated with a worse prognosis, as at 8 months the median survival of the Men1Pdx1-Cre mice was significantly reduced (15.7 months vs 21.8 months for the controls). This was obtained without any concomitant differences in blood glucose levels, indicating that the polyamines increase occurred before the onset of any hypoglycemia. These studies demonstrated a concordance between a mouse model and human MEN1 patients regarding an association between polyamines and pNET progression, implicating mechanisms such as upregulation of oncogenic MYC due to loss of MEN1 (6).

The findings of the present study are of particular interest as they may help identify MEN1 patients with dpNET who are at a higher risk for a more aggressive course irrespective of tumor size. Currently, the most widely used parameters to identify such patients are tumor size (≥ 2 cm), functionality, growth rate, and grade, but not all high-risk patients will develop metastases, and metastases may also occur in low-risk patients with lesions smaller than 2 cm (3, 8). Additional biomarkers for predicting disease progression are currently missing, whereas evolving biomarkers such as the NETest have not been studied extensively in this patient group (3). Although polyamine levels were particularly elevated in gastrinomas, their levels were still able to differentiate metastatic cases without gastrinomas from controls, suggesting that polyamines are not affected by the functional status of the tumors. However, it is still not clear whether other clinical manifestations of MEN1 may affect polyamine levels having an impact on their signature, or whether polyamine levels can predict early future metastasis and/or correlate with metastatic load. In addition, although the 3MP signature showed similar predictive performance in treatment-naive patients with metastases and those on active treatment, whether treatment affects its performance characteristics needs further clarification. Furthermore, additional studies may need to be performed in treatment-naive patients with sporadic NFpNET with tumor size smaller than 2 cm to define whether polyamines could be similarly used in this expanding group of patients.

The findings of this study are encouraging but merit further investigations to delineate whether plasma polyamines present a reliable noninvasive biomarker to predict aggressive behavior in MEN1-related dpNETs, particularly in patients with high levels due to their relatively low sensitivity. Until such data become available, imaging still remains an integral part of the follow-up for such patients.