A total of 76 MEN-1 and 320 Sp-GEP-NEN patients were evaluated. Forty-one MEN-1 patients were diagnosed with GEP-NEN, of which 20 were non-functioning pNENs and 21 were functioning NENs. Among functioning NENs, three patients had insulinomas and 18 had gastrinomas with clinical manifestation of ZES. All the patients with MEN-1 syndrome were genetically diagnosed, except for one patient who had clinical and familiar diagnostic criteria. Primary hyperparathyroidism was present in all the MEN-1/ZES patients, while a pituitary tumor was found in five of them. Among the Sp-GEP-NEN patients, 19 were diagnosed with Sp-ZES. A total of 37 ZES patients were studied and followed up for a median of 152 months [interquartile range (IQR) 46.5–200 months], with a median follow-up of 174 months (IQR 92.2–228) in MEN-1/ZES patients and 108 months (IQR 45–168) in Sp-ZES patients.

Among the Sp-ZES, 17 (89.4%) patients had peptic ulcer disease at the time of diagnosis, of whom one had concomitant mild diarrhea and one had severe esophagitis (Los Angeles C) with a major acute bleeding event, while two (10.6%) were paucisymptomatic, having already been treated with PPIs. In the MEN-1/ZES group, 15 patients (83.3%) had peptic ulcer disease at the time of diagnosis, of which two patients had major bleeding, two patients had esophagitis (Los Angeles C), and three patients had diarrhea, while three patients (16.7%) were paucisymptomatic due to ongoing PPI treatment. Clinical manifestation at the time of diagnosis was not related to patient outcome.

At the time of diagnosis, a total of 20 patients underwent CT imaging [11 Sp-ZES (57.9%) and 9 MEN-1/ZES (50%)], while 19 patients underwent MRI [9 patients with Sp-ZES (47.4%) and 10 patients with MEN-1/ZES (55.5%)]; two patients underwent both TC and MRI. Gallium-68 PET was performed in six Sp-ZES (31.6%) and three MEN-1/ZES patients (16.7%), while Octreoscan® was performed in 12 Sp-ZES (63.2%) and 13 MEN -1/ZES patients (72.2%), somatostatin receptors functional imaging was not available in three patients diagnosed before 1995. Gallium-68 PET missed one patient with a positivity rate of 87.5%, whereas Octreoscan® missed six patients with a positivity rate of 76%.

After diagnosis, all patients received appropriate PPI treatment with a good clinical response. The characteristics of the patients are shown in Table 1.

MEN-1/ZES patients were younger than Sp-ZES (mean age 43.7 ± SD 12.3 vs 53.5 ± SD 14.7 years, p = 0.035), and primary gastrinoma was smaller in MEN-1/ZES compared to Sp-ZES (median13 vs 18 mm, p = 0.03). Primary dNEN was found in ten of 18 MEN-1/ZES patients and seven of 19 patients with Sp-ZES (55.5% and 36.8%, respectively, p = ns). The other patients with MEN-1/ZES had pancreatic (5/18), hepatic (1/18) and unknown primary localization (2/18). By contrast, the other patients with Sp-ZES had pancreatic (5/19), hepatic (1/19), gastric (1/19) and unknown primary localization (5/19). Gastrin plasma levels at diagnosis were higher in Sp-ZES than in MEN-1/ZES (median 874 vs 374 pg/ml, p = 0.012).

Liver metastases (either synchronous or metachronous) were found in ten (synchronous in eight and metachronous in two) Sp-ZES patients and in ten (synchronous in five and metachronous in five) MEN-1/ZES patients, respectively. Only the Sp-ZES patients developed extrahepatic metastases (6 versus 0, p = 0.019).

Two of 19 patients (10.5%) with Sp-ZES were found to have additional NENs at diagnosis or during follow-up, both gastric carcinoids. Eight of 18 patients (44.4%) with MEN-1/ZES had additional NENs, three non-functioning pNENs, two gastric carcinoids, two lung carcinoids, and one thymic carcinoid. Four of the 18 MEN-1/ZES also had a prolactinoma and one had a non-functioning pituitary adenoma.

Non-neuroendocrine tumors were observed during follow-up in two patients with MEN-1/ZES [one with gastrointestinal stromal tumor (GIST) and one with breast cancer] and in two Sp-ZES patients (one with prostate and one with breast cancer).

Surgery was performed in 13 Sp-ZES and eight MEN-1/ZES. Of the 13 Sp-ZES, 11 underwent duodenal-pancreatic (DP) surgery (one Whipple, two distal pancreatectomies, four pancreatic enucleations, four duodenotomies), one liver resection, and one gastrectomy. Six of 13 patients (46%) showed metastatic lymph nodes and seven patients had liver metastases (53%). Seven of the eight MEN-1/ZES patients (87%)underwent DP surgery (two distal pancreatectomies and five duodenotomies) and one liver resection; six (75%) had positive lymph nodes at the time of surgery and four (50%) showed liver metastases. At the initial postoperative follow-up (2–4 weeks), surgery resulted curative in all the patients with MEN-1/ZES and in 10/13 with Sp-ZES. However, at subsequent long-term follow-up, recurrences occurred in four Sp-ZES and four MEN-1/ZES, at 12–288 and 12–144 months, respectively. Among surgically treated patients, the PFS of the MEN-1/ZES patients was not statistically different from that of the Sp-ZES group, although a trend toward better PFS was observed for the MEN-1/ZES patients in the entire cohort (median PFS of 96 months for Sp-ZES and 126 months for MEN -1/ZES patients).

A total of 15 patients received SSAs (either octreotide 28 mg or lanreotide 120 mg every 28 days), eight of whom were in the Sp-ZES group and seven in the MEN-1/ZES group. Of the eight Sp-ZES patients, seven had liver metastases and one had liver and extrahepatic metastases. Of the seven MEN-1/ZES patients, six had liver metastases and two had multifocal pNEN, which was treated surgically in only one case. Two patients with Sp-ZES had a partial objective response while four Sp/ZES and six MEN-1/ZES showed stable disease. Disease progression occurred in two Sp-ZES and one MEN-1/ZES, and recurrence occurred in five Sp-ZES after 24–120 months (median 39).

During SSAs therapy, one MEN-1/ZES patient developed a lung carcinoid and another developed an esophageal GIST.

With pharmacological treatment, MEN-1/ZES patients showed a better PFS than Sp-ZES (median PFS not reached vs 34.5 months, p = 0.0227 by log-rank-test) (Fig. 1).

Progression-free survival of patients treated with somatostatin analogs in both MEN-1/ZES patients (not reached) and Sp-ZES (median 34.5 months, p = 0.0227 at log-rank-test)

The median OS (mOS) of the entire cohort was 204 months, and it was higher in MEN-1/ZES than in Sp-ZES (310 vs 168 months, p = 0.0347) (Fig. 2).

Median overall survival (OS) in MEN-1/ZES patients compared with Sp-ZES (310 vs 168 months, p = 0.0347 at log-rank-test)

At univariate-logistic regression, age at ZES diagnosis [p = 0.01, odds ratio (OR) 1.05, CI 1.0–1.1], G3 grading (p = 0.003, OR 21.3, CI 2.8–162.0), MEN-1/ZES (p = 0.02, OR 0.3, CI 0.1–0.9) and extrahepatic metastases (p = 0.001, OR 7.2, CI 2.2–23.4) were significantly related to OS (Table 2), while TNM, gender, gastrin level, location and size of primary tumor did not matter.

The grading affected the mOS, which was 310 months in G1 patients, 168 months in G2 patients, and 39 months in G3 patients (p = 0.003 by log-rank-test) (Fig. 3). Moreover, even though TNM did not affect OS, the presence of extrahepatic metastases, which occurred only in Sp-ZES, resulted in a significant shortening of OS. However, among patients with liver metastases, OS was better in the MEN-1/ZES group (mOS Sp-ZES 96 months vs MEN-1/ZES not reached, p = 0.0069 by log-rank-test).

Overall survival in the entire study population, according to grading

Gastrin levels at diagnosis did not correlate with OS, although patients who died due to ZES progression had significantly elevated circulating gastrin levels before death (median 16,815 pg/ml, IQR 695.5–59,375) compared to median at diagnosis (1090 pg/ml, IQR 625–1835, p = 0.0098).

In multivariate regression analysis, none of the variables was independently associated with OS (Table 2).

Regarding PFS, at univariate-logistic regression, age (p = 0.04, OR 1.0, CI 1.0–1.1), size (p = 0.039, OR 1.0, CI 1.0–1.1), G3 grade (p = 0.009, OR 14.6, CI 2.0–107.6) and the presence of extrahepatic metastases (p = 0.005, OR 4.6, CI 1.6–13.8) were independently associated with PFS (Table 3). At multivariate regression analysis, only the presence of extrahepatic metastases (p = 0.05, OR 3.4, CI 0.99–11.6) resulted independently associated with PFS (Table 3).

At the end of the study, eight Sp-ZES (42%) and 14 MEN-1/ZES (78%) were alive (p = 0.0448 Fisher-test), 9/19 (47%) and 3/18 (17%) patients died of ZES in Sp-ZES and MEN-1/ZES, respectively. In the MEN-1/ZES group, another patient died from malignant thymic carcinoma. In the Sp-ZES group, besides the nine ZES-related deaths, two other patients died because of other non-endocrine malignancies. Age at death was not significantly different in the four MEN-1/ZES patients (median 61.5 years, IQR 52.7–67.2) and the 11 Sp-ZES (median 71 years, IQR 51–80).