Long-Term Outcomes of Tuberous Sclerosis Complex-Associated Non-functional Pancreatic Neuroendocrine Tumors: Should We Be More Conservative?

This single-center, retrospective, comparative study of TSC-PNET and spPNET patients was approved by the Institutional Review Board. Patients undergoing pancreatic resection for PNETs at the Massachusetts General Hospital (MGH) between December 1998 and December 2020 were evaluated, and patients with TSC-PNETs were compared with patients with non-functional PNETs. Patients with functional tumors or other hereditary syndromes were excluded from this study.

Clinicopathologic factors including sex, age at operation, body mass index (BMI), American Society of Anesthesiologists (ASA) score, and type of operation (pancreatoduodenectomy, distal pancreatectomy, total pancreatectomy, central pancreatectomy, or pancreatic enucleation) were evaluated.

Histopathological parameters evaluated were tumor size, tumor location (pancreatic head, body or tail, and multifocal tumor), lesion type (solid, cystic, and mixed solid/cystic), TNM stage (American Joint Committee on Cancer [AJCC] 8th edition), Ki-67 proliferation index, tumor grading according to the 2010 World Health Organization (WHO) classification, number of harvested lymph nodes, and number of positive lymph nodes. Negative resection margin (R0) status was defined according to the Union for International Cancer Control (UICC)/AJCC if no tumor cells were detected within 1 mm of the resection margin, and R2 status was defined as macroscopic evidence of a tumor left in situ. Postoperative complications were determined according to the Clavien–Dindo Classification and dichotomized. Local or distant tumor recurrences were classified based on pathology or imaging. Disease-free survival (DFS) was defined as the time between the operation and the first tumor recurrence, while overall survival (OS) time was defined from the date of operation until death of any cause.

Continuous variables were expressed as medians and interquartile range (IQR), and categorical variables were expressed as absolute numbers and relative frequencies. A case-control matching was performed between patients with TSC-PNETs and spPNETs based on the pathological size of the largest tumor. Group analyses of categorical variables were performed using the Chi-square test or Fisher’s exact test with Bonferroni correction. Continuous variables were compared among groups, using the non-parametric Kruskal–Wallis test, and the median follow-up period was calculated using the reverse Kaplan–Meier method. Survival analyses were performed using the Kaplan–Meier method with log-rank test, and statistical analysis was performed using IBM SPSS Statistics (version 28.0; IBM Corporation, Armonk, NY, USA).

A total of 347 patients underwent pancreatectomy for PNETs. After excluding patients with functional and other hereditary syndrome-associated PNETs, 255 patients were considered eligible for this study; 14 were TSC-PNETs and 241 were non-functional spPNETs. Patient and operative characteristics are shown in Table 1. Males constituted 52.9% (n = 135) of the cohort, and the median age at operation was 56.0 years (IQR 21.0). TSC patients were operated on at a younger age (24 years vs. 57 years; p < 0.001) and were more likely to undergo minimally invasive operations compared with spPNETs (78.6% vs. 21.6%; p < 0.001).

Tumor characteristics and distribution of prognostic factors between spPNETs and TSC-PNETs are presented in Table 1. The median tumor size for the whole cohort was 2.7 cm (IQR 2.5). Final pathology revealed a well-differentiated (WHO grade I) tumor in most (62.9%) patients, and there were no significant differences between the TSC-PNET and spPNET cohorts (p = 0.7). Cases of poorly differentiated (grade III) tumors were rare and were only observed in patients with spPNETs (n = 7, 3.0%). TSC-PNETs were more frequently multifocal (28.5% vs. 0.0%; p < 0.001). Lymph node (N1) and distant metastases (M1) were only observed among patients with spPNETs (25.3% vs. 0.0%, p = 0.027; and 10.0% vs. 0.0%, p = 0.215) respectively). The median follow-up period of the entire cohort was 103.8 months (95% confidence interval [CI] 89.2–118.6), and the median follow-up period for TSC-PNET and spPNET patients was 78.0 (95% CI 52.2–103.9) and 106.7 (95% CI 89.3–124.0) months, respectively. Despite a higher rate of R1 resections (28.6% vs. 6.2%; p = 0.004), TSC-PNETs did not recur locally or distantly. Local (n = 8, 3.3%), distant (n = 30, 12.4%), and local + distant (n = 6, 2.5%) recurrence were only observed among spPNETs. Median OS and DFS were not reached in both cohorts. Kaplan–Meier survival curves are shown in Fig. 1. In 14 resected TSC-PNETs, three patients developed new-onset diabetes, and one patient who underwent subtotal pancreatectomy was taking pancreatic enzyme supplements (Table 2).

Case-control matching based on the largest tumor size was performed for 14 patients with TSC-PNETs and 194 patients with spPNETs. All PNETs were <5.1 cm in size, which was the maximum size observed in the TSC-PNET cohort. TSC-PNETs were operated on at a younger age (24.0 years vs. 57.5 years; p < 0.001) and were more likely to undergo minimally invasive operations compared with spPNETs (78.6% vs. 24.3%; p < 0.001). There were no significant differences in tumor type (solid vs. cystic; p = 0.13) and tumor WHO grade (p = 0.795). The number of harvested lymph nodes was higher in the spPNET group (5.0 [IQR 13.0] vs. 1.5 [IQR 3.0]; p = 0.016). Lymph node metastases were only observed in the spPNET group (21.1% vs. 0.0%; p = 0.023), while postoperative complications (p = 0.31) and length of hospital stay (p = 0.12) were comparable between the two groups.

Among spPNETs, tumor recurrence was observed in 30 (16.0%) patients, of whom 20 patients presented with isolated liver metastases, 4 had a local and liver recurrence, and 5 had local recurrence only. One patient developed recurrent multiorgan disease involving the liver, peritoneum, and bone.

Only one TSC-PNET patient presented with a new 8 mm T2-weighted hyperintense non-enhancing pancreatic lesion adjacent to the resection margin on magnetic resonance imaging (MRI) 33.8 months after the R0 resection of a well-differentiated PNET in the pancreatic tail. The 5- and 10-year mortality rates of spPNET and TSC-PNET patients were 6.2% vs. 0.0% and 11.3% vs. 0.0%, respectively. No PNET-related deaths were observed among the TSC-PNET patients. OS and DFS were not reached.

Given the fact that none of the TSC patients had nodal metastases or poorly differentiated grade 3 histology, we further compared outcomes of patients with spPNETs, matched for tumor size and absence of lymph node metastases and grade 3 histology, with the TSC group. After exclusion of patients (n = 47) with the latter risk factors, a lower overall tumor recurrence rate of 8.5% was observed. However, tumor recurrence from spPNETs was still significantly higher compared with TSC-PNETs (9.3% vs. 0.0%; p = 0.035). Similar findings applied for 5- (4.0%) and 10-year (9.3%) mortality rates from spPNETs, which slightly improved in the case of grade 1–2 tumors and lymph node-negative disease.