Clinical Characteristics and Survival Associations of Pancreatic Neuroendocrine Tumor: Does Age Matter?

Background: Pancreatic neuroendocrine tumor (pNET) is the second most common epithelial neoplasm of pancreas. As in pancreatic adenocarcinoma, patients with different onset ages display different clinical features and prognosis, we divided pNET patients into early-onset pancreatic neuroendocrine tumor (EOpNET) and typical age-at-onset pancreatic neuroendocrine tumor (TOpNET) to investigate the effect of onset age on its clinical characteristics and prognosis. Methods: Data were collected from the Surveillance, Epidemiology, and End Results (SEER) database (2004–2015; cohort 1) and Fudan University Shanghai Cancer Center (FUSCC) (2005–2018; cohort 2) retrospectively. The clinical characteristics were compared using chi-squared tests. Cox proportional hazards regression was used to evaluate hazard ratios (HRs) and 95% condence intervals (CIs), and overall survival were formulated by Kaplan–Meier curves. Results: In total 5,368 and 330 patients’ data were included from the SEER database and FUSCC. Gender had no effect on survival in EOpNET group. Tumors locating in the tail (HR: 0.721, 95%CI: 0.63-0.83, P < 0.001) and body (HR: 0.712, 95%CI: 0.60-0.85, P = 0.001) had lower risks of death compared to tumors in the head in TOpNET group. The overall survival of EOpNET (136 (3-143) months) was better than TOpNET (85 (3-143) months) ( P < 0.001) in SEER database. The results in FUSCC was similar with SEER cohort. Conclusions: EOpNET group had signicantly better overall survival than TOpNET though their clinical characteristics were not signicantly different. In the future personalized treatment of pNET, the onset age of patients will the and prognosis.


Abstract
Background Pancreatic neuroendocrine tumor (pNET) is the second most common epithelial neoplasm of pancreas. As in pancreatic adenocarcinoma, patients with different onset ages display different clinical features and prognosis, we divided pNET patients into early-onset pancreatic neuroendocrine tumor (EOpNET) and typical age-at-onset pancreatic neuroendocrine tumor (TOpNET) to investigate the effect of onset age on its clinical characteristics and prognosis.
Methods Data were collected from the Surveillance, Epidemiology, and End Results (SEER) database retrospectively. The clinical characteristics were compared using chi-squared tests. Cox proportional hazards regression was used to evaluate hazard ratios (HRs) and 95% con dence intervals (CIs), and overall survival were formulated by Kaplan-Meier curves.
Results In total 5,368 and 330 patients' data were included from the SEER database and FUSCC. Gender had no effect on survival in EOpNET group. Tumors locating in the tail (HR: 0.721, 95%CI: 0.63-0.83, P < 0.001) and body (HR: 0.712, 95%CI: 0.60-0.85, P = 0.001) had lower risks of death compared to tumors in the head in TOpNET group. The overall survival of EOpNET (136 (3-143) months) was better than TOpNET (85 (3-143) months) (P < 0.001) in SEER database. The results in FUSCC was similar with SEER cohort.
Conclusions EOpNET group had signi cantly better overall survival than TOpNET though their clinical characteristics were not signi cantly different. In the future personalized treatment of pNET, the onset age of patients will become another important factor for guiding the treatment and prognosis.

Background
Pancreatic neuroendocrine tumor (pNET) is the second most common epithelial neoplasm of the pancreas with relatively inert biological behavior compared to pancreatic adenocarcinoma. [1] However the incidence of it has increased steadily in recent decades [2], probably due to more sensitive detection methods and more frequent routine examinations among people. [3,4] Recent reports suggest that earlyonset pancreatic cancer (EOPC) which was de ned as pancreatic adenocarcinoma (PDAC) with onset age before 50 and accounts for 5-10% of all pancreatic cancer cases, displays a distinct clinical and/or biological subset of the disease [5][6][7]. With a mean age at diagnosis of 56.8 for pNET, also a small proportion of cases are diagnosed in younger age [8]. However, research on the differences between earlyonset pancreatic neuroendocrine tumor (EOpNET) and typical age-at-onset pancreatic neuroendocrine tumor (TOpNET) are very limited, probably due to di culties in studying rare disease subset with lower incidence. Most studies are descriptive and only include relatively small populations lacking a typical age at diagnosis comparison group. In the recent repot of Wang et al., age plays a critical in uence on the overall survival and should be considered as a factor in future staging systems of pNETs [9]. To identify whether onset age of pNET has similar effects in patients as in PDAC, we conducted this study to evaluate the associations between clinical characteristics and prognosis by age at diagnosis of pNET, and how early-onset pNET (EOpNET) differs from typical age-onset pNET (TOpNET).

Patients and Data Collection
Pathologically con rmed pNET cases from 2004 to 2015 were retrieved from the SEER database. Patients And then a single center series of Fudan University Shanghai Cancer Center (FUSCC) was analyzed and patients who were pathologically diagnosed of pNET were included. The data of demographics, including age, sex, grade, location of the primary site and surgery condition were retrieved. And the data of tumor T stage, nodal status and metastases were also retrieved and classi ed on the basis of the AJCC staging classi cation (8th). The follow-up data were con rmed by review of medical records monthly and contacting the patients or their relatives to ascertain disease progression and vital status or date of death if applicable.
The study was approved by the ethics committee of Fudan University Shanghai Cancer Center. In both cohorts, patients with overall survival time less than 3 months were excluded to rule out perioperative mortality.

Statistical Analysis
For statistical analysis we de ned patients with age < 50 years as early-onset pNET (EOpNET) and age ≥ 50 years as typical age-at-onset pNET (TOpNET) at the time of diagnosis. Clinical characteristics were compared using chi-squared tests. Survival time was calculated from the date of initial diagnosis until the date of last follow-up time or time of death. Kaplan-Meier method and log-rank tests were used to analyze the overall survival and compare survival proportions. Univariate survival analysis of age, sex, grade, surgery, tumor location and so on were performed by Cox proportional hazards regression. Hazard ratios (HRs) and corresponding 95% con dence intervals (95% CIs) were calculated. All the statistical analyses were performed using IBM SPSS Statistics version 21.0 software (IBM Corp, Armonk, NY, USA) and GraphPad Prism 7.0. All tests were two-sided and tests with P values < 0.05 were considered statistically signi cant.

Results
In total, 5,368 patients from the SEER database were pathologically diagnosed of pNET and included in the study (Table 1), including 1203 (22.4%) EOpNET patients and 4165 (77.6%) TOpNET patients, respectively. While the male/female ratio was around 1:1 in EOpNET patients (1:1.05) which was similar to previous reports [7,8], the proportion of male in TOpNET patients was signi cantly higher than female (1.30:1, P < 0.001). Approximately half of the patients had a tumor located in the body and/or tail of the pancreas in both groups (45.2% and 48.1%). 92.5% and 89.4% patients had low or intermediate differentiated tumors in EOpNET group and TOpNET group respectively. Furthermore, signi cantly higher proportion of patients in EOpNET group received surgery treatment (P < 0.001) than TOpNET group. The median survival period for EOpNET group was 136 (3-143) months and for TOpNET group was 85 (3-143) months (P < 0.001). Univariate Cox proportional hazards models were used to evaluate factors associated with survival in both groups. We found that even though gender had no effect on the survival of EOpNET group, it seems that females had a lower risk of death than males in TOpNET group (HR: 0.871, 95%CI:  We further analyzed the overall survival associations by age at diagnosis of pNET in the two cohorts and Kaplan Meier method and survival curves were formulated (Fig. 1). We found in SEER cohort, the overall survival in EOpNET patients was signi cantly better than among TOpNET patients (136 (3-143)   Discussion pNETs are relatively rare tumors with an incidence of 5.25/100,000, and majority have satisfying prognosis [4]. However, they are highly heterogeneous with small percentage of patients displaying aggressive properties. For example, the overall 5-year survival in metastatic non-functional pNET was only 30%, which bring forth challenges to clinical practice [10]. Therefore, clarifying the clinical characteristics and survival associations could help greatly for the management of pNETs.
In our study, the proportion of EOpNET in FUSCC cohort (38.2%) was signi cantly higher than SEER cohort (28.9%). The average age at diagnosis of pNET in our cohort was 52.6 ± 12.6 years, which is earlier compared to SEER database (62 ± 15) years [4]. The mean age of diagnosis of PDAC in US is reported as 71years, meanwhile in china is 62-65 [11,12]. Hence, it is reasonable to consider differences among different subgroups of races and the similar condition in PDAC. The overall survival of EOpNET (136 (3-143) months) was signi cantly better than TOpNET (85 (3-143) months) (P < 0.001) in SEER database, and also in the FUSCC cohort with median follow-up time of 39.585 (3.6-152.8) months (N = 330), though not statistically signi cant (P = 0.245), which was consistent with previous reports [9]. In this study, we analyzed both cohorts at 50 years age to keep their consistency, however this could be the reason that certain characteristics in FUSCC cohort are not consistent with or obvious as the SEER database. Therefore, exploring the reasons of different age of onset of disease in different states and races could provide new strategy for therapy in the future.
Surgery is the most effective therapy for localized tumors and even for metastatic NETs that more than 90% of liver tumors can be resected, cytoreductive surgery is recommended [2]. In our study, we found surgery can lower the risks of death signi cantly in both EOpNET or TOpNET group. Furthermore, the tumors located in the body and tail had lower risks of death compared to head in TOpNET patients while not in EOpNET patients. This may be associated with worsened physical status in TOpNET group as the with pancreatic head tumor patients underwent relatively more invasive surgery of pancreaticoduodenectomy.
Additionally, patients diagnosed between 2010 to 2015 had signi cantly better prognosis than those diagnosed between 2004 to 2009, which re ects that with earlier tumor detections due to more prevalent routine medical examinations and the systemic treatment options for pNET have also been adopted considerably in recent years. Somatostatin analogs, peptide receptor radionuclide and everolimus have been demonstrated to effectively improve the prognosis of pNET [13]. Combined chemotherapy with temozolomide and capecitabine is recently reported to prolong the progression-free survival of pNETs [14,15]. Moreover, our study also shows unmarried patients had signi cantly higher death risk than the married, which coincides with previous reports suggesting that social support could potentially impact cancer survival signi cantly [16].
There is no consensus yet on optimal follow-up strategy for pNET. Most guidelines suggest performing enhanced CT or magnetic resonance imaging (MRI) of the abdomen yearly for the rst 3 years, then every 1 to 2 years for a total of 10 years [17]. For tumors smaller than 2 cm, previous reports are controversial about observation or operation [18]. While from the perspective of our research, operation is recommended, as EOpNET group had signi cantly better overall survival than TOpNET.
pNETs are usually sporadic, but increasing evidences suggest the importance of genetic mutations in the initiation and development of pNETs. Jiao et al. performed whole exome sequencing in 68 sporadic pNETs and found multiple endocrine neoplasia type 1 (MEN1) and death domain-associated protein/αthalassemia mental retardation syndrome (DAXX/ATRX) mutations in more than 40% of pNETs. Also 14% of samples had mutations in the genes implicated in mTOR pathway [19]. Scarpa et al. performed whole genome sequencing of 102 primary pNETs and identi ed larger than anticipated germline contribution to clinically sporadic pNETs [3]. According to the "two-hit" hypothesis [20], gene mutation was closely associated with tumor age of onset. Therefore, future studies on the variations of genetic mutation among different age groups, especially the early-onset tumor, could help identify pNET driver genes.
The patients were all histology diagnosed in this retrospective study, hence all the patients were either from a surgical database or with positive biopsy results; thus, there may be some bias in the proportion of different stages. However, we queried the SEER database and our institution to minimize the in uence of inadequate sample size and geography. Thus, the clinical characteristics of EOpNET and TOpNET and their effects on prognosis were supported and complemented by both cohorts. However, when exploring the survival associations with age at diagnosis, our database is not signi cant due to limitations on sample size and number of deaths. The tumor function was not distinguished, though only a small percent of pNETs were functional; this may cause the bias of prognosis, because functioning neoplasms usually have better prognosis compared to nonfunctioning tumors as a result of early diagnosis.
EOpNET constitute only 28-38% of the total pNET patients and have better prognosis, however due to young age of onset, the disease is responsible for much more of the total number of years-of-life-lost.
Identifying the clinical characteristics and survival associations by age at diagnosis of pNET could help us better understand pNET, especially those early onset ones, and potentially yield strategies to delay the onset of it, or help put forward new therapeutic methods.

Conclusions
EOpNET group had signi cantly better overall survival than TOpNET though their clinical characteristics seem to be not signi cantly different, except that the proportion of female and lymph node positive status in EOpNETs was higher than in TOpNET. In the future personalized treatment of pNET, the onset age of patients will become another important factor for guiding the treatment and prognosis.

Declarations
Ethics approval and consent to participate The study was approved by the ethics committee of Fudan University Shanghai Cancer Center. All patients provided written informed consent.

Consent for publication
There will be no con ict of commercial interest for any of the patient with the publication of the manuscript.