A Case Report of Insulinoma Relapse on Background Nesidioblastosis: A Rare Cause of Adult Hypoglycemia Free

Abstract

Gastroenteropancreatic neuroendocrine tumors (GEP-NETs) are rare overall (up to 5 cases per 100,000 individuals per year), although they are steadily increasing in incidence (1). Insulinomas account for a small proportion of GEP-NETs, 90% of them are benign, and tumor enucleation is generally sufficient to recover fully from the recurrent hypoglycemic events that characterize the clinical presentation (2). The detection of insulinomas on background nesidioblastosis is a clinical finding of even more rare occurrence (3), thus representing a considerable challenge still for the physicians nowadays, despite recent advancements in diagnosis and therapy (4). Standard medical therapy with diazoxide (DZX) and increased meal frequency can mitigate the recurrence of hypoglycemic attacks, although the DZX-induced side effects, such as fluid retention and hypertrichosis and the unpredictability of (pro)insulin bouts from the lesion(s), are severely limiting an optimal, long-term disease management. Somatostatin receptor (SSTR) ligands, such as octreotide and lanreotide, are accounted as a second-line treatment option for refractory insulinomas and nesidioblastosis. However, efficacy is often hampered by the weak binding affinity and the limited spectrum of SSTRs covered by these drugs. A multi-SSTR ligand, pasireotide, has been developed for the treatment of GEP-NETs (5–7) and successfully applied in at least one case for the long-term management of adult nesidioblastosis (8).

We herein provide a case study, summarizing our experience in the multidisciplinary management of a rare case of adult hypoglycemia at a third-level clinical referral center.

Case Report

A 63-year-old white woman presented at the Endocrinology Clinic of the Verona University Hospital on April 2017 with a history of sporadic insulinoma (6 mm in diameter), surgically removed by distal spleno-pancreatectomy at another institution in 2013. The patient’s history revealed that the intervention was followed by further hypoglycemic episodes just before the home discharge. Histopathological examination confirmed the presence of a well-differentiated, proinsulin-secreting endocrine neoplasm. However, residual β cell hyperplasia, in the absence of additional insulinomas, was detected near the resection margins, and it was compatible with nesidioblastosis (Fig. 1). Therefore, a combined therapy with scaled-up DZX, up to 225 mg/d, and hydrochlorothiazide (HCT) was promptly started. Nevertheless, the hypoglycemic episodes recurred, progressively increasing in severity and frequency, up to 10 events per month. At a follow-up visit on April 2016, subcutaneous injections of the somatostatin analog octreotide (100 μg once daily) were introduced. However, octreotide therapy failed to prevent further hypoglycemic attacks, and it was therefore discontinued in favor of the preceding DZX + HCT therapy, coupled with increased meal frequency.

Figure 1.

Tissue morphology and immunohistochemistry of the primitive pancreatic lesion. The transition between (A) insulinoma [hematoxylin and eosin (H&E) stain] and (B) adjacent pancreatic parenchyma (insulin antibody, polyclonal guinea pig, dilution 1:200; Dako/Agilent, Santa Clara, CA) is shown. (C and D) Nesidioblastosis directly expands from a ductal structure, forming a ductuloinsular complex, as indicated by the arrows (C, H&E; D, chromogranin A antibody, clone DAK-A3, dilution 1:2500; Dako/Agilent). (E and F) Insulinoma develops around a ductal structure with typical amyloid deposits (E, arrow, H&E; F, synaptophysin antibody, clone 27G12, dilution 1:100; Novocastra/Leica Biosystems, Newcastle Upon Tyne, UK). Formalin-fixed tissue sections were embedded in paraffin blocks and sectioned at a thickness of 3 µm. Staining was performed in an automated stainer (Leica Bond-III; Leica Biosystems, Newcastle Upon Tyne, UK) using the “Bond Polymer Refine Detection” system, according to the manufacturer’s protocol.

Open in new tabDownload slide

The patient came to our attention 1 year later. The anamnesis revealed fatigue, sweating, craving for sweets, and a weight gain of about 6 kg since the surgical intervention. The patient reported that the fear of hypoglycemia was progressively limiting her daily life. The patient’s comorbidities included idiopathic chronic pericardial effusion and mixed anxiety-depressive syndrome. The family history was positive for type 2 diabetes (brother). The physical examination only revealed mild DZX-induced hypertrichosis at both extremities and on the upper lip. Physical parameters included the following: ambulatory blood pressure 115/70 mm Hg, heart rate regular at 80 beats/min, weight 58 kg, body mass index 22.1 kg/m². The results of the laboratory workup for rare forms of hypoglycemia, other than functional GEP-NETs, were all within the physiological range.

On the basis of a recently published, successful experience in the long-term medical treatment of adult nesidioblastosis (8) and in light of recent literature (4, 7), an in-hospital rescue therapy with the multi-SSTR ligand pasireotide was proposed.

As shown in Fig. 2 and Table 1, DZX 225 mg/d was continued for 4 days (period A). Then, DZX was scaled down to 150 mg/d, and concomitant subcutaneous pasireotide 0.6 mg twice per day (BID) was started (period B; 4 days). DZX was then further scaled down to 75 mg/d, and pasireotide 0.6 mg BID was continued (period C; 4 days). The daily glucose profiles were monitored every 5 minutes by a subcutaneous continuous glucose-monitoring (CGM) system (Enlite™ Sensor, MedTronic, Minneapolis, MN), coupled with a transmitter (Guardian™ Link; MedTronic). Severe hypoglycemia (sensor glucose ≤ 50 mg/dL) occurred only during Period C. Thus, pasireotide was discontinued, and DZX was scaled up to the original dose.

Figure 2.

Continuous glucose monitoring (CGM) with (black) and without (gray) pasireotide therapy. The plot shows the in-hospital glucose sensor-level time courses recorded while the patient was assuming DZX 225 mg/d (period A; 4 days; gray line) or DZX 75 mg/d + pasireotide 0.6 mg twice per day (BID; period C; 4 days; black line). Period B is not shown for visual clarity, and it is detailed in Table 1. Note that period C was characterized by an episode of severe hypoglycemia and by recurrent “saw blade” glycemic phases, reflecting mild–moderate hypoglycemic episodes, promptly corrected by the patient with repeated oral glucose intake.

Open in new tabDownload slide

A new instrumental and diagnostic workup was subsequently performed. The morphologic and functional imaging of the pancreas included abdominal MRI, ⁶⁸Ga-DOTATOC CT/positron emission tomography, ^99mTc-nanocolloid scintigraphy, and echo-endoscopy. A 22-mm nodular lesion with unclear significance for insulinoma or splenule was identified close to the pancreas. The functional imaging was characterized by both moderate uptake of ⁶⁸Ga-DOTATOC (9) and endothelial-like uptake of ^99mTc-nanocolloids (9). The echo-endoscopic examination identified an additional hypo-echogenic nodulation (9.9 mm maximum diameter) adjacent to the major lesion and comprised within the pancreatic parenchyma (9). Fine-needle ago-biopsy (22 gauge) of the major lesion showed splenic parenchyma, whereas the smaller one displayed the immunophenotype of a well-differentiated insulinoma (9).

In December 2017, the patient underwent laparoscopic intervention: The lesion was identified by intraoperative ultrasound, and an 8.0-mm benign insulinoma relapse (Ki-67 = 1%; 2010 World Health Organization grading = G1) on background nesidioblastosis, with adjacent accessory spleen, was confirmed by intraoperative histopathological examination (9). Subtotal pancreatectomy with preservation of the pancreatic head was performed, and the patient was discharged with a pancreatic enzyme replacement therapy. No further hypoglycemic episode occurred, and altered glucose regulation within the prediabetes range (hemoglobin A1c = 41 mmol/mol) was detected at the 6- and 9-month follow-up visits in the absence of DZX therapy.

Discussion

The cornerstone of current medical therapy for benign insulinomas is DZX, which acts on the β cells of the endocrine pancreas by opening the ATP-sensitive K⁺ channels, thus inhibiting insulin release. However, DZX therapy is often insufficient to control the abrupt hyperinsulinemic bouts by the tumor, and it is characterized by fluid retention and hypertrichosis as major side effects.

Insulinomas are known to express predominantly SSTR subtypes 2 and 5 (10). Therefore, SSTR ligands, such as octreotide or lanreotide, both displaying high affinity for SSTR2 and to a lesser extent, for SSTR5, are considered as an additional treatment option, alone or in combination with DZX. The major drawback of these drugs, however, relies on the inhibitory effects (octreotide > lanreotide) on glucagon and growth hormone secretion, which may result in a defective counter-regulatory response to hypoglycemia. Moreover, approximately one-third of insulinomas does not express SSTR2. In these cases, octreotide/lanreotide therapy may have the detrimental effect of worsening frequency and amplitude of hypoglycemic episodes (11).

Pasireotide is a multi-SSTR ligand with a higher binding affinity for SSTR subtypes SSTR5 > SSTR2 > SSTR3 > SSTR1 (5), and it is known to display a major hyperglycemic effect compared with octreotide and lanreotide (7). Although hyperglycemia is generally accounted as a side effect, this action may be helpful in hyperinsulinemic hypoglycemic syndromes refractory to standard treatment (4).

The absence of the hyperglycemic effect by pasireotide observed in this case lacks a univocal explanation: it is indeed possible, given the co-occurrence of both nesidioblastosis and insulinoma, that the insulinoma lesion was unresponsive to pasireotide, whereas the surrounding hyperplastic β cells were actually sensitive to the treatment, or vice versa. As an alternate hypothesis, neither the insulinoma nor the nesidioblastosis foci might have been sensitive to pasireotide. A post hoc comprehensive determination of the SSTRs, although desirable, was not possible, although it would have provided useful insights on the immunophenotypic signature of both insulinoma and nesidioblastic aggregates. It should also be noted that the primary lesion was stemming from a ductal structure, very close to the hyperplastic foci of nesidioblastosis. Therefore, although this is an event of very rare occurrence (12, 13), which may recognize a genetic background (3), insulinoma origin from nesidioblastosis can be suspected. Whether this is the causative process, the pathological characteristics of both the primary and the secondary lesions were clearly diagnostic for a neuroendocrine tumor compatible with insulinoma.

However, it should be pointed out that nesidioblastosis is a histologic appearance that can be observed in asymptomatic subjects, as well as in patients with hypoglycemia. Nesidioblastosis is typically characterized by postprandial hypoglycemia as a distinctive clinical feature, whereas insulinomas are more frequently and typically characterized by fasting hypoglycemia (14). In this specific clinical case, the profile of the glucose time courses recorded by CGM would thus play in favor of insulinoma as a major determinant of the clinical symptoms. Nevertheless, clinical reports on nesidioblastosis are very limited—this condition probably rarer than insulinoma—and it was reported that fasting cannot distinguish definitely between these conditions. Indeed, rare cases of insulinoma may have negative fasts (15), whereas rare cases of nesidioblastosis may have positive fasts (16–18).

Furthermore, whereas prediabetes was detected 9 months after surgery, follow-up might have been too short to confirm cure. Indeed, although ∼90% of the pancreatic tissue was removed as a result of the second-look surgery, the histopathological examination revealed that nesidioblastosis nests were dispersed around the secondary insulinoma lesion. Thus, the long-run development of further insulinomas from a hyperplastic β cells’ nest cannot be completely excluded.

This experience taught us that insulinomas/nesidioblastosis treatment failure may occur because a differential diagnosis is neglected or masked by local anatomic variation. Hence, it further reinforced the relevance of a careful, intraoperative histopathological characterization of the focal lesion and surrounding tissue. Finally, the availability of CGM measures should be mentioned among the strengths of this case study and underscores the usefulness of such technological devices to assist clinical decisionmaking.

Conclusions

The co-occurrence of insulinoma and nesidioblastosis is a rare clinical finding, and relapses may occur after surgery. This case study highlighted the limited efficacy of a rescue therapy with the multi-SSTR ligand pasireotide, when a (masked) insulinoma relapse occurred on background nesidioblastosis. Although we did not determine the expression pattern of SSTRs, it is possible that the failure of pasireotide at preventing hypoglycemia was a result of the unfavorable immunophenotype of the insulinoma relapse, the nesidioblastosis, or both. The optimization of disease outcomes in the era of personalized medicine warrants further studies to unveil whether the systematic determination of the expression pattern of SSTRs on the lesion(s) may eventually improve the identification of the most appropriate medical treatment of insulinoma/nesidioblastosis when surgery is impracticable/unaccepted.

Abbreviations:

Abbreviations:
 
• BID

  twice per day

 
• CGM

  continuous glucose monitoring

 
• DZX

  diazoxide

 
• GEP-NET

  gastroenteropancreatic neuroendocrine tumor

 
• HCT

  hydrochlorothiazide

 
• SSTR

  somatostatin receptor

Acknowledgments

The support of the administrative and clinical personnel of the Division of Endocrinology, Diabetes and Metabolism and of the Verona Pancreas Institute (University and General Hospital of Verona, Verona, Italy) is gratefully acknowledged. Parts of this study were presented in poster form at the “Endocrinology & Metabolism Scientific Meeting 2018,” organized by the Italian Society of Endocrinology, Bari (Italy), 10–11 May 2018.

Financial Support: This study was supported, in part, by research grants of the University of Verona (to P.M. and E.B.). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Author Contributions: M.D. researched and analyzed data, provided care for the study patient, and drafted the manuscript. C.M. researched data, provided care for the study patient, and discussed the manuscript. C.C. and A.T. analyzed all pancreatic tissue sections, provided high-quality histopathological images, and edited the manuscript. A.F. performed ultrasound-guided transesophageal tissue biopsy and discussed the manuscript. L.L. and G.M. performed the laparoscopic surgical intervention, provided care for the study patient, and discussed the manuscript. M.F. was responsible for the radionuclide imaging. E.B. edited the manuscript and provided substantial contribution to the overall discussion. P.M. researched and analyzed data, provided care for the study patient, and edited the manuscript. P.M. and M.D. are the guarantors of this work and as such, had full access to all of the data in the study and take responsibility for the integrity and the accuracy of the data analysis.

Disclosure Summary: The authors have nothing to disclose.

References