Pancreatic metastases in patients with neuroendocrine neoplasms: A multi-centre cohort study.

Pancreatic metastases (PMs) from neuroendocrine neoplasms (NENs) are rare but the increased sensitivity of new diagnostic tools such as 68 Ga-DOTATATE PET/CT has resulted in their increased recognition at initial diagnosis or follow-up. A retrospective analysis of the data of patients from six tertiary referral centres was performed in order to identify the characteristics and the prognostic significance of PMs in patients with NENs. We used a control group of 69 age-, sex- and primary tumour - matched NEN patients from the same cohort with stage IV disease but no PMs. Overall survival (OS) was assessed using the Kaplan-Meier method log-rank analysis was used to assess the impact of various clinical and histopathological variables in OS. We identified 25 patients (11 females) with PMs with a median age at diagnosis of 60 years. The small intestine was the most common primary (80%) with a prevalence of 4.2% PMs (21/506). Fourteen patients presented with synchronous PMs whereas 11 developed metachronous PMs after a median time of 28 months (range: 7-168 months). Grading was available in 24 patients; 16 patients had G1 tumours, four G2, two atypical lung carcinoid, one typical and one atypical thymic carcinoid. Most patients had other concomitant metastases (12 hepatic, 4 lung and 6 bone) while five patients exhibited peritoneal carcinomatosis. Median OS in the PMs group was not reached compared with 212 months in the control group (95% CI: 26-398). The univariate analysis identified no prognostic factors statistically significantly associated with the OS. In conclusion, PMs are encountered with a low prevalence among NEN patients mostly developing in patients with advanced metastatic disease. The presence of PMs does not seem to be associated with a negative prognostic impact in OS.

Association of lymph node metastases, grade and extent of mesenteric lymph node dissection in locoregional small intestinal neuroendocrine tumors with recurrence-free survival.

We aimed to assess the prognostic impact of tumor- and patient-related parameters in patients with stage I-III small intestinal neuroendocrine tumors (SI-NETs), who underwent locoregional resective surgery (LRS) with curative intent. We included 229 patients with stage I-III SI-NETs diagnosed from June 15, 1993, through March 8, 2021, identified using the SI-NET databases from five European referral centers. Mean ± SD age at baseline was 62.5 ± 13.6 years; 111/229 patients were women (49.3%). All tumors were well-differentiated; 160 were grade 1 (G1) tumors, 51 were G2, two were G3 and 18 tumors were of unspecified grade (median Ki-67: 2%, range 1%-50%). One-hundred and sixty-three patients (71.2%) had lymph node (LN) involvement. The median number of retrieved lymph nodes was 10 (0-63), whereas the median number of positive LNs was 2 (0-43). After a mean ± SD follow-up of 54.1 ± 64.1 months, 60 patients experienced disease recurrence at a median (range) of 36.2 (2.5-285.1) months. The 5- and 10-year recurrence-free survival (RFS) rates were 66.6% and 49.3% respectively. In univariable analysis, there was no difference in RFS and overall survival (OS) between LN-positive and LN-negative patients (log-rank, p = .380 and .198, respectively). However, in stage IIIb patients who underwent mesenteric lymph node dissection (MLND) with a minimum of five retrieved LN (n = 125), five or more LN metastases were associated with shorter RFS (median RFS [95% CI] = 107.4 [0-229.6] vs. 73.7 [35.3-112.1] months; log-rank, p = .048). In addition, patients with G2 tumors exhibited shorter RFS compared to patients with G1 tumors (median RFS [95% confidence interval (CI)] = 46.9 [36.4-57.3] vs. 120.7 [82.7-158.8] months; log-rank, p = .001). In multivariable Cox-regression RFS analysis in stage IIIb patients, the Ki-67 proliferation index (hazard ratio = 1.08, 95% CI = 1.035-1.131; p < .0001) and the number of LN metastases (hazard ratio = 1.06, 95% CI = 1.001-1.125; p = .047) were independent prognostic factors for RFS. In conclusion, LRS with a meticulous MLND and a minimum number of five harvested LNs appears to be critical in the surgical management of SI-NET patients with locoregional disease. In patients who underwent LRS and MLND, the Ki-67 proliferation index and the LN metastases count were independent predictors of RFS.