Somatostatin analogs in patients with Zollinger Ellison syndrome (ZES): an observational study

Zollinger Ellison syndrome (ZES) is a rare syndrome caused by gastrin hypersecretion from a gastrinoma. Gastrinoma treatment has two goals: the control of acid hypersecretion and the control of tumor growth. While therapy for the syndrome is univocally based on proton pump inhibitors, the one for disease control is still debated. We here aimed at evaluating the role of somatostatin analogs (SSAs) in the control of tumor progression in a series of ZES patients. A retrospective analysis of a prospectively collected database of ZES patients, followed and managed from 1990 to 2019, was performed. The patients’ clinical, pathological, treatment, and follow-up data were analyzed. Data regarding SSAs therapy start, dosage, duration, and side effects were collected. 33 patients with ZES were diagnosed. Fourteen patients (42%) had a grade 1 (G1) neuroendocrine neoplasm (NEN), five had G2 (15%), none had G3. Fifteen patients (45%) had metastatic disease. Overall, 12 (36%) underwent SSAs therapy. The median treatment duration was 36 months. Eight patients (67%) had a sustained response to SSAs, four (33%) showed an early progression, with a significant difference in terms of PFS between the patients with early and late progression (84 vs 2 months, p = 0.004). No differences in terms of OS and PFS were observed between the treated and non-treated patients, despite the proportion of metastatic patients was greater in the SSAs-treated group (75% vs 29% in the non-treated group, p = 0.01). Present data support the use of SSAs in ZES, considering that gastrinoma is mainly a well-differentiated low-grade tumor (G1 or G2), with a high expression of somatostatin receptors.


Introduction
Gastrin secreting tumors, or gastrinomas, are rare functioning neuroendocrine neoplasms (NEN), usually arising from the duodenum or the pancreas. Gastrin hypersecretion from gastrinoma is responsible for Zollinger Ellison syndrome (ZES) characterized by hypersecretion of gastric acid and severe recurrent peptic disease [1]. Gastrinoma occurs in one to three cases per million people in the USA annually, making them the second most common functional, pancreatic neuroendocrine neoplasm [2]. Approximately 75% of gastrinomas are sporadic, while 25% of patients present with multiple endocrine neoplasia type 1 (MEN-1) [3].
There are two therapeutic goals in the management of patients with gastrinoma: the control of gastric acid hypersecretion and the treatment of the tumor itself. While the therapy for syndrome control is univocal and based on proton pump inhibitors (PPI), which have significantly decreased the morbidity and mortality resulting from severe ulcer disease [4], the management of the neoplastic disease is still debated.
Surgical treatment is the only curative approach for patients with gastrinoma, even if surgery is of limited value in presence of metastatic disease or multiple NEN in the contest of MEN-1 [1]. In addition, in up to 30% of cases, the primary tumor can not be precisely located [5].
Since 1980, long-acting somatostatin analogs (SSAs) have been used as a symptomatic treatment of welldifferentiated functioning NENs. More recently, the PRO-MID and CLARINET phase III randomized trials demonstrated an increase of PFS in NEN patients treated with SSAs [6,7], even if among these studies only four cases of gastrinoma were included in the CLARINET study.
However, based on these studies, the use of long-acting SSAs is currently recommended in all advanced NEN patients irrespective of their functional status [8]. Despite different case reports and case series suggested the role of SSAs on control of gastrin secretion and symptoms in ZES patients, as well as on tumor progression [9][10][11][12], to date only a few studies with a very low number of patients investigated specifically the role of SSAs in ZES [13].
We here aimed at evaluating the role of SSAs in the control of tumor progression in a series of consecutive enrolled ZES patients.

Methods
A retrospective analysis of a monocentric prospectively collected database was performed on patients with ZES, followed and managed from 1999 to 2019, at the Division of Gastroenterology and Digestive Endoscopy at Fondazione IRCCS Ca' Granda, Ospedale Policlinico, Milan, Italy.
ZES syndrome was diagnosed based on clinical, biochemical, and imaging data. Each tumor has been retrospectively immunohistochemically classified according to the WHO (World Health Organization) 2010 classification based on Ki-67 index [14], when histological/cytological specimens were available, and the disease was staged based on the presence/absence of nodal or distant metastases, i.e., according to the current ENETS (European Neuroendocrine Tumor Society) TNM (Tumor/Node/Metastasis) clinical staging [15]. Other characteristics evaluated were the site and the size of the primary tumor and the number of lesions. Symptomatic, biochemical [circulating chromogranin A (CgA) and gastrin levels] and objective (tumor size) responses were evaluated according to the criteria of the Italian Trials in Medical Oncology (ITMO) group [16], based on evaluation of biochemical, clinical, and tumor responses. For 'biochemical response', complete remission (CR) was defined as the return of circulating specific peptides to the normal range for at least one month, and partial response (PR) as >50% decrease for at least one month. For 'symptomatic response', CR was defined as complete relief from all symptoms, and PR as reduction by at least 50% of both frequency and intensity of flushing and/or diarrhea. 'Tumor response' was defined according to the number and size of a preexisting lesion on computed tomography (CT) or magnetic resonance imaging (MRI) as CR when a complete disappearance of all known disease was observed, PR when facing with a decrease >50%, stable disease (SD) with a decrease of <50% or an increase of <25%; and progressive disease (PD), in presence of an increase of more than 25%. The tumor response was also assessed according to Response Evaluation Criteria in Solid Tumors (RECIST) criteria, evaluating retrospectively the iconographic documentation of the patients [17]. Results are presented according to the ITMO biochemical, clinical and tumor responses and RECIST criteria.
All the patients underwent a complete clinical, biochemical, and imaging evaluation at diagnosis and at regular intervals during follow-up. Plasma CgA levels and gastrin levels were prospectively collected at diagnosis, then after during the follow-up. The venous blood samples for gastrin and CgA assay were collected in EDTA-containing tubes between 8h00 and 10h00 a.m. after an overnight fast. The samples were centrifugated at a temperature of 4°C and the plasma was separated and stored at −30°C until assayed. CgA and gastrin were measured using commercially available kits. Imaging examinations, such as ultrasound (US), CT, MRI, and functional nuclear medicine imaging (either Gallium-68 positron emission tomography (PET) or Octreoscan®) were collected at the diagnosis and then, during the follow-up. The rate of surgical interventions, the medical regimen, the presence of associated diseases (including other malignancies), the presence of genetic syndrome, i.e., MEN-1, the cause of death, and the length of survival were also recorded.
The types of treatment were namely radical surgery, longacting SSAs, combined SSAs and non-radical surgery, liverdirected therapies, or others (Peptide receptor radionuclide therapy (PRRT) or targeted-therapies). Octreotide was given i. m. at a dose of 30 mg and lanreotide s.c. at a dose of 120 mg once every four weeks. SSAs were given until complete disease remission occurred, significant disease progression was observed, or severe side effects occurred.
Data regarding SSAs therapy start, dosage, duration, and side effects were collected. Finally, data on the median follow-up, overall survival (OS), and progression-free survival (PFS) were collected.
All data for each patient were anonymized after collection, recorded, evaluated, and analyzed.
The data collection was closed in November 2019.

Statistical analysis
Continuous variables were reported as median and range unless otherwise stated. Continuous data were analyzed using the non-parametric Mann-Whitney test. Kruskal-Wallis statistics followed by Dunn's multiple comparison test were used to compare variables repeated at different times.
Survival curves were estimated using the Kaplan-Meier method and the log-rank test was used for comparison of survival curves between groups of patients.
A p value < 0.05, two-sided, was considered statistically significant.
Among the 33 patients, 12 (36%) underwent SSAs therapy, due to metastatic disease in 10, and concomitant second neuroendocrine neoplasm in two. Six (18.2%) underwent SSAs therapy alone and six in combination with non-radical surgery. In the entire cohort of the patients, nineteen patients (58%) underwent surgery: in nine patients surgery was radical (47%), even if ten patients (53%) have relapsed after a median time of 54 months. Two patients underwent PRRT for disease progression to SSAs therapy.
SSAs (either octreotide LAR and lanreotide) were given at full/standard doses of 30 mg and 120 mg, respectively, every four weeks. The median treatment duration of SSAs treatment was 36 months (range 30-84). Lanreotide was given in two patients for a median period of 21 months, whereas octreotide in ten patients for a median of 60 months (See Table 2).
The biochemical response was documented in nine (75%) of the studied patients who achieved partial response  Other 4 (12.1) Fig. 1 Flow chart diagram of the studied population in serum gastrin levels, presenting a decrease of more than 50% in serum gastrin levels [median value from 532 pg/ml at baseline (range 709-1713 pg/ml) to 206 pg/ml (median 253-1280 pg/ml)] after three months of treatment; in three patients no change had occurred and presented a rapid progression of the disease. In responder patients, a longlasting (36 months, range 30-84) biochemical response was observed during the prolonged follow-up. Before SSAs commencement, two patients had persistence of ZES symptoms despite high-dose PPI treatment (omeprazole 60 mg/daily in 15 patients, esomeprazole 120 mg/daily in 8, rabeprazole 60 mg/daily in 10 patients). After SSAs introduction, complete clinical control (with no evidence of peptic disease, no upper gastrointestinal symptoms, or diarrhea) was obtained in all but one patient, by concomitant PPI.
As concerned objective response, after a median follow up of 144 months, three patients (25%) had a sustained objective response to SSAs, with stable disease documented by RECIST criteria, whereas six (50%) showed an objective late progression of the disease and three (25%) an early progression. No patient had tumor regression.
Considering both clinical, biochemical, and radiological response, eight patients (67%) had a sustained overall response to SSAs, with a mean duration of response of 48 months (range 24-120), whereas four (33%) showed an early clinical or radiological progression after 3 months of treatment, with a significant difference in terms of median PFS between the patients with early and late progression (mPFS 84 vs 2 months, p = 0.004).
Overall, between SSAs-treated and non-treated patients, no differences in terms of median OS and PFS were observed, despite the proportion of metastatic patients was greater in the SSAs-treated group (75% vs 29% in the nontreated group, p = 0.01).
When considering only metastatic patients, the median OS after the onset of metastases in treated patients resulted of 144 months compared to 76 months in untreated patients (p = 0.06).
In the entire series, two patients had a type 2 gastric NEN (gNEN) diagnosed. In one case, gNEN was treated with endoscopic resection, in the other case SSAs treatment was started with gNEN complete response at 12 months. No patients had gNEN diagnosed on SSAs therapy.
No significant side effect development was observed in the 12 treated patients and in none of the patients, it was necessary to withdraw the treatment. One patient developed asymptomatic cholelithiasis.
At the end of the long follow-up period, 14 patients (42%) died: eight in the not-treated group and six in the SSAs-treated group, without significant difference.

Discussion
The present study supports the role of SSAs treatment in ZES patients, showing an excellent safety profile with a prolonged biochemical and clinical response in the majority of the patients treated with SSAs with a sustained objective response in 3/12 cases, after a long period of observation. A possible role of SSAs in ZES treatment has been previously suggested due to the reported high expression of somatostatin receptors (SSTR) in gastrinoma [18][19][20][21]. In 2002, Kulaksiz et al. observed sstr2 expression in 100% of gastrinomas, while sstr3 and sstr5 were expressed in 79 and 76% of cases, respectively [18]. Accordingly, somatostatin receptor scintigraphy (SRS) has been known for years as the most sensitive technique for the detection of both primary tumors and metastatic disease in patients with gastrin secreting tumors [19][20][21]. More recently, the 68Gallium-PET demonstrated a good detection rate in a series of 25 patients with ZES and negative CT findings [20]. In our series we observed a high rate of patients with positive functional imaging (88.5%), confirming a high SSTR expression in ZES patients. Moreover, as previously reported in the literature [21], we observed most patients with gastrinoma having a grading G1 and G2 (42.5 and 15%, respectively), while no patients had G3 carcinoma. The inverse relationship between tumor grade and SSTRs expression is well known [22]. These observations represent a strong rationale for SSAs therapy in gastrinoma.
In the last years, the use of PPI has dramatically improved the control of gastric acid hypersecretion, significantly reducing the morbidity and mortality of ZES related peptic disease. However, a portion of patients has been reported not to tolerate high doses of PPI or present symptoms of gastrin hypersecretion on PPI [23]. Since 1985, several case reports and small case series have suggested SSAs be effective in symptoms control in ZES patients. [11,[24][25][26], thus it is reasonable to think that SSAs may further help obtain a complete clinical response. In this series, one patient with the persistence of ZES symptoms on high-dose PPI achieved a complete clinical control after SSAs commencement. SSAs treatment thus could be of value in patients experiencing relevant side effects from PPI therapy or with a syndrome not completely controlled by PPIs alone even with dose escalation.
In our series, 12 of 33 ZES patients (36%) underwent SSAs therapy for a median treatment duration of 36 months (range 30-84). Overall, SSAs treatment was well-tolerated without significant side effects during the prolonged follow-up.
In 75% of cases, a reduction in gastrin levels of more than 50% was observed after 3 months of treatment compared with pre-treatment levels. In responder patients, a long-lasting (36 months, range 30-84) biochemical response was observed in all cases. Similar results have been previously reported in ZES patients treated with octreotide (300-600 μg s.c. daily, in divided doses) with a reduction in gastrin levels ranging from 53 to 87% [11,24,27]. In most cases, the SSAs effect on gastrin reduction was sustained for at least 12 months [11,24,27]. More recently, two studies on GEP-NET patients including ZES patients, suggested also long-acting SSAs be effective in achieving a reduction in gastrin levels by ≥50% [28,29].
In addition, ZES patients are prone to develop gNEN, due to the trophic action of gastrin on ECL cells in the gastric body [30]. Interestingly, we observed the occurrence of gNEN in two of 21 cases in the group of patients nontaking SSAs, while no gNEN were observed in the SSAstreated group. Since, the demonstrated effect of SSAs on gastrin reduction in ZES patients, SSAs therapy may have had a protective/therapeutic role on gNEN occurrence.
Although the low incidence of gNEN does not allow to draw definitive conclusions, SSAs could be of help in preventing gNEN occurrence or in treatment of multiple gNEN or recurrent disease.
A possible role of SSAs therapy on tumor growth control and disease progression has been suggested by different studies [31]. However, studies focusing on ZES patients are scanty and heterogeneous, with a limited number of patients, moreover, their results are hardly comparable, because of the use of different SSAs formulations and heterogeneous criteria to define response or progression of the disease. In 2002 Shojamanesh et al. reported the largest prospective study on ZES patients treated with SSAs. A total of 15 consecutive patients with gastrinoma (three MEN-1) underwent octreotide treatment (initially 200 μg s.c. twice a day, since 1999 LAR formulation 20-30 mg in monthly) [32]. In all cases, progression over the 3-6 months before enrollment was documented. SSAs treatment was continued until severe side effect development, complete disease remission, or significant progression. Tumor response was observed in 53% of patients (stabilization in seven and partial response in one patient) lasting 25 ± 6 months (range 5.5-54.1 months) [32]. In the present series, a sustained objective response to SSAs was observed in 3/12 patients (25%) with an mPFS of 84 months, whereas six (50%) showed an objective late progression. As expected, none of the patients had tumor regression [32,33]. Moreover, we observed a lower rate of overall tumor response, probably due to the remarkably longer follow-up period, compared to other studies. In fact, we observed a longer mean duration of tumor response (48 months), when compared to other studies, such as the one by Shojamanesh et al. (25 months)…. Interestingly, non-responder in most cases presented an aggressive course with early disease progression (mPFS 2 months), probably reflecting a more aggressive tumor behavior even before SSAs treatment. On the other hand, the patients who showed an early response maintained it for a long period (mPFS of 84 months). A statistically significant difference between responder and non-responder ZES patients was observed p = 0.004.
Interestingly, we observed a higher rate of biochemical response (75%) as compared to objective tumor response (25%) this probably reflecting a mechanism of tumor growth independent from SSAs inhibition. Moreover, all the patients that experienced an early progression did not show a biochemical response. Similarly, Vinik et al. reported gastrin level reduction in 6/8 ZES patients treated with SSAs with tumor response observed only in patients with demonstrated biochemical response [33]. We did not observe any significant differences in terms of median OS and PFS between SSAs-treated and nontreated patients. This observation should not be interpreted as lack of efficacy of SSAs therapy, as in the SSAs-treated group a significantly higher rate of metastatic patients was observed (75% vs 29%, p = 0.01). In addition, to date, none of the currently existing regimens for progressive metastatic gastrinoma has demonstrated an unequivocal improvement in survival [32]. The occurrence of metastatic disease represents the most important prognostic factor in GEP-NEN and it is associated with a significantly reduced OS and PFS [34,35]. Taking into consideration the observed comparable course of disease between SSAs-treated and non-treated patients, even in presence of a significantly higher rate of metastatic disease in SSAs-treated patients, it could be postulated that SSAs did actually have an impact on survival leveling up the two curves.
In conclusion, data from the present series suggest SSAs be safe and effective in obtaining a sustained reduction of gastrin secretion and a sustained objective response in 25% of cases. Further prospective studies are needed to clarify the impact of SSAs treatment on survival in patients with ZES. Funding This research did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector.

Compliance with ethical standards
Conflict of interest The authors declare no competing interests.
Ethics approval This study was approved by the institutional review board and carried out following the approved guidelines.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.