Pancreatic neuroendocrine tumors: A case-based evidence review

Abstract

Pancreatic neuroendocrine tumors (pNETs) are rare, presenting significant challenges in timely diagnosis and subsequent treatment. The clinical and pathobiological behavior of these tumors varies significantly, making follow-up and therapeutic approaches challenging for clinicians. Although the majority of these neoplasms are hormonally inactive, some can be associated with endocrine dysfunction. Very rarely, a nonfunctional tumor can later become hormonally active, further complicating prognostication and management. Depending on the character of the disease, clinical picture and prognosis, different treatment modalities are instituted with varying effectivities. We recently came across a unique case of nonfunctioning malignant pNET at an advanced stage, metastatic disease upon diagnosis, managed medically with somatostatin analog therapy (Octreotide) and targeted therapy (Everolimus) with stable disease for 40 months that subsequently turned out to become functional (insulinoma). With the aid of this unique case, we update the current clinical, diagnostic and therapeutic approach to pNETs in this evidence-based review.

Key Words: Pancreatic neuroendocrine tumour; Somatostatin analog therapy; Insulinoma; Nonfunctional pNET; Metastatic disease; Somatostatin receptor imaging

Core Tip: Pancreatic neuroendocrine tumors (pNETs) are rare neoplasms with variable clinical and pathobiological characteristics posing diagnostic and therapeutic challenges. Mostly nonfunctional with low growth rates, some cases can be functional while others can run an aggressive clinical course. Hormonal, imaging and histological assessments are essential for planning appropriate diagnosis, prognosis and management. Therapeutic strategies depend on the tumor grade, functionality and presence of metastatic disease. Nonfunctioning pNETs very rarely turn functional during the advanced disease course. Presenting a unique nonfunctioning metastatic pNET case, later evolving to an insulinoma, we update the diagnostic, prognostic and management strategies for pNETs in the evidence-based review.

INTRODUCTION

Epidemiology of pancreatic neuroendocrine tumors

Neuroendocrine neoplasms (NENs) refer to neoplasms originating from the enterochromaffin cells of the embryonic gut exhibiting both neural and endocrine differentiation. While the majority of NENs are well-differentiated, less aggressive, but potentially malignant NETs (neuroendocrine tumors), the remaining are poorly differentiated and aggressive NECs (neuroendocrine carcinomas). As per the latest classification, NENs can be divided into gastroenteropancreatic NENs (GEP-NENs) and lung NENs[1]. GEP-NENs constitute 55%-70% of all NENs, 12%-20% of which are formed by pancreatic NENs (pNENs)[2].

The pNENs account for only 1%-2% of all pancreatic neoplasms, and pancreatic NETs (pNETs) account for nearly 90% of all pNENs[3]. The pNENs are mostly diagnosed in the sixth decade of life, and they exhibit a minimal male preponderance. Nearly 70% of the pNENs are non-functioning, whereas 30% are functioning, secreting hormones or peptides associated with the clinical syndrome[4]. Those tumors with detectable peptides on biopsy specimens but that do not exhibit features of hormone hypersecretion syndrome are still considered non-functioning. While most pNENs are sporadic, nearly 10% are associated with hereditary syndromes, including Multiple Endocrine Neoplasia type 1 or type 4 (MEN1 or MEN4), von Hippel-Lindau syndrome, neurofibromatosis type 1 and tuberous sclerosis. The hereditary pNENs are often multifocal, well-differentiated, functioning, possibly secreting multiple peptide hormones simultaneously, with a young age of onset, and are associated with ectopic production of hormones and accompanied by other endocrine disorders or malignancies. Though pNENs are less aggressive with better 5-year survival rates compared to pancreatic adenocarcinomas, they are associated with nearly 23.1% recurrence rates within 8.1 years, even after curative resection[5].

A population-based study using surveillance epidemiology and end results (SEER) 18 registry data showed that in comparison to 2000-2008, the annual incidence of pNETs has increased from 0.27 per 100000 to 1.00 per 100000 in the study period 2009-2016[6]. This increase in annual incidence is accompanied by the early detection of particularly the asymptomatic non-functioning NETs in their early stages, thereby resulting in an improvement in median overall survival from 46 months to 85 months. Moreover, the improved median overall survival was consistent over different stages of the disease - localized disease at 83%, regional disease at 67%, and metastatic disease at 28%[6].

The overall incidence of GEP-NECs is rare, but several studies report increasing incidence particularly across North America, Asia and Europe[7,8]. The SEER 18 registry reported an incidence rate of 9.36 per 100000, nearly 90% of which are lung NECs (8.36 per 100000, 95% small cell NECs and 5% non-small cell NECs), whereas pNECs constitute only 0.07 per 100000 (small cell NECs nearly 50%)[9]. The pNECs have one of the shortest median survivals at 5.7 months [multivariate hazard ratio (HR): 1.10; 95% confidence interval (CI): 1.03-1.18, P < 0.001]. This could be attributed to the advanced stage at diagnosis (75.6% of the pNEC patients presenting with distant metastasis) in comparison to other GEP-NEC subtypes[9].

Due to this cumulative disease burden, studying the pathophysiology, clinical presentation, and response to treatment of pNETs and pNECs make it interesting and relevant.

Etiopathogenesis of pNETs

The pNETs are derived from primitive endoderm, which differentiate into several gastrointestinal (GI) epithelial lineages, inclusive of enterochromaffin cells. These enterochromaffin cells are categorised based on morphology, location and expression of peptide hormones. The proliferation of these cells is dependent on various transcription factors, for example, Neurogenin 3 in the pancreas, which delineates the Notch signalling pathway responsible for secretory and absorptive cell lineages[10].

Germline mutations in tumour suppressor genes MEN1, VHL, NF1, and TSC1/2 account for 10% of familial syndrome-derived pNETs. The majority of pNETs are sporadic, resulting from somatic cell mutations for proteins involved in chromatin modelling, for example, MEN1, DAXX and ATRX. Encoding protein mutations in the mTOR pathway have shown a correlation to the formation of pNETs[11]. Alongside the presence of serum biomarkers, these have vital implications for targeted therapy. NECs can present with specific genetic alterations, for example, TP53 and RB1 in small cell NECs and a variable genetic profile in large cell NECs[12].

Dependent on secretory products and hypersecretion of specified hormones, the pNET are called by different names. Insulinoma is the most common functional pNET (30%-40%), with an annual incidence of 1-32 cases per million population[13-17]. Functional expression means symptomatic management can be achieved effectively with surgical excision. In the case of insulinoma, tumor origin within β-cells of Islets of Langerhans means parenchymal sparing pancreatectomy can relieve symptoms of the classical Whipple’s triad of hypoglycaemia (plasma glucose < 4 mmol/L), neuroglycopenic clinical signs and resolution after administration of glucose[18].

Classification of pNETs

Functionality is an important predictive factor in the presentation of pNENs[19]. Patients with nonfunctioning pNENs have larger, poorly differentiated tumors with hepatic and lymph node involvement associated with reduced survival outcomes compared to those with functioning pNENs[20]. Tumor grade is another important prognostic factor that can indicate cellular morphology according to classes of differentiation[21]. According to the World Health Organization, the pNENs are broadly classified into well-differentiated pNETs (three subtypes: G1 pNETs, G2 pNETs and G3 pNETs) and poorly differentiated pNECs (two subtypes: Small cell and large cell types and are high-grade by definition)[22]. These classifications are summarized in Table 1.

Table 1 Classification of pancreatic neuroendocrine tumors, modified from World Health Organisation Classification of Neuroendocrine Neoplasms[20].

Terminology	WHO grade	Differentiation	Mitotic rate	Ki-67 index
NET	G1	Well-differentiated NET	< 2/2 mm²	< 3%
	G2		2-20/2 mm²	3%-20%
	G3		> 20%/2 mm²	> 20%
Small cell NEC	High grade	Poorly differentiated NEC	> 20/2 mm²	> 70%
Large cell NEC	High grade

NET: Neuroendocrine tumors; NEC: Neuroendocrine carcinoma.

DIAGNOSTIC APPROACH

Role of serum biomarkers

Generally, serum biomarkers can be used to identify the presence of NENs with the presentation of proteins in the secretory granules, synaptic-like vesicles, or cytoplasm of neuroendocrine cells. These have low specificity as these biomarkers can also be elevated in inflammatory conditions like inflammatory bowel disease, pancreatitis, gastritis, proton pump inhibitor and steroid treatment, liver failure, and other neoplasms[12]. Specific pancreatic neuroendocrine markers[23] can be used to distinguish functional vs non-functional pNETs, as summarised in Table 2.

Table 2 Biomarkers for the diagnosis of pNEN[23].

Circulating biomarkers
Non-specific biomarkers	Specific to functional pNEN	Specific to nonfunctional pNEN
Chromogranin A	Insulin	Pancreatic polypeptide
Pancreastatin	Glucagon	Human chorionic gonadotropin
Chromogranin B	Gastrin	Neurotensin
Neuron-specific enolase	Somatostatin	Ghrelin
Alpha fetoprotein	Vasoactive intestinal peptide	Calcitonin
	Growth hormone or GHRH
	Adrenocorticotropic hormone
Immunohistochemical biomarkers
Differentiation markers	Site of origin markers	Prognostic markers
Chromogranin A	NESP55	ATRX/DAXX
Synaptophysin	PGR	SSTR2a
INSM1	PDX1	PD-L1
	ARX
	ISL1
Molecular biomarkers
Circulating tumor cells (CTCs)
Circulating cell-free DNA (cfDNA)
Circulating microRNAs (miRNAs)
Circulating transcripts (NETest score)
G protein-coupled receptor-associated sorting protein 1 (GPRASP1)
Delta-like protein 3 (DLL3)
Tumor-associated macrophages (TAMs)
Glucose transporter 1 (GLUT1)

INSM1: Insulinoma-associated protein 1; NESP55: Neuroendocrine secretory protein 55; PGR: Progesterone receptor; PDX1: Pancreatic and duodenal homeobox 1; ISL-2: Insulin gene enhancer binding protein 2; ATRX/DAXX: Alpha-thalassemia/mental retardation X-linked and death domain-associated protein; SSTR2a: Somatostatin receptors 2a; PD-L1: Programmed cell death ligand 1.

The diagnosis of NEN is established by immunohistochemical staining showing positivity for markers of neuroendocrine differentiation, including INSM1 (highly sensitive and specific), synaptophysin (highly sensitive but not specific), and chromogranin A (less sensitive than synaptophysin but highly specific)[21]. This should be followed by a stain for transcription factors to know the site of origin: For example, pNENs may express PDX1 (a transcription factor typically expressed by the β cells), ARX (transcription factor typically expressed by the α cells), and ISL1 and stain for the peptide hormones[24].

The typical (indolent) insulinomas are epigenetically similar to β cells: PDX1 positive and ARX negative[25]. They are small (below 2 cm), have a favourable prognosis after resection, and are characterized by YY1 mutations in nearly 30% of cases. On the other hand, the rare aggressive insulinomas are ARX positive, larger (3.5-9.0 cm), have higher metastatic rates, and are characterized by genetic alternations seen characteristically in non-functioning pNENs, including ATRX/DAXX mutations, alternative lengthening of telomeres or CDKN2A deletions[25]. Hence, assessment of ATRX/DAXX mutations and alternative lengthening of telomeres are recommended in pNENs.

The origin, differentiation and tumorigenic mechanisms of the aggressive insulinomas are more closely related to non-functioning pNENs[25]. It is possible that they existed as non-functioning pNENs for a while before exhibiting the functional behaviour (recurrent hypoglycaemia). These tumours might have been producing insulin, though at an asymptomatic level. It is hypothesized that the tumorigenesis of aggressive AXR-positive insulinomas might have happened from a non-functioning β-cell tumor acquiring α-cell characteristics after ATRX/DAXX mutations and the presence of alternative lengthening of telomeres, or from a non-functioning α cell/intermediate tumor acquiring β cell characteristic of insulin secretion[25].

Role of imaging

Imaging plays an important role in the diagnosis, follow-up and assessment of response to treatment of pNETs. High-resolution computed tomography (CT) of the thorax, abdomen, and pelvis is the first line in providing radiological imaging of the location and size of the primary tumor, nodal disease, and sites of metastasis. Intravenous contrast, particularly in the late arterial phase, can prove useful for detecting NENs in the pancreas and liver[26]. Magnetic resonance imaging (MRI) has increased specificity and sensitivity compared to CT (100% specificity for pancreas and 98% for liver), and diffusion-weighted (T1/T2) imaging can aid in small hepatic metastasis identification, along with greater detail into tumor margins[19]. Positron emission tomography (PET) with F-fluorodeoxyglucose can show uptake in combination with CT/MRI.

Due to the expression of biomarkers, nuclear imaging is useful in providing functional response information of pNETs. Somatostatin Receptor (SSR) imaging is most commonly used in combination with single-photon emission computed tomography (SPECT)/CT or PET/CT and includes the use of octreotide, ¹¹¹In-pentereotide or Gallium-DOTA tracer[13]. Receptor positivity has important applications for the use of somatostatin analogues or targeted therapy and can measure treatment response/remission.

Staging

Following physical examination, biopsy and imaging for diagnosis, pNETs are staged, as per TNM and AJCC classifications[27], as recorded in Table 3.

Table 3 Tumor, node, metastasis and American Joint Committee on Cancer classifications for staging of pancreatic neuroendocrine tumors[27].

AJCC	TNM	Description
I	T1 N0 M0	Tumor < 2 cm, local to pancreas. No nodal disease/metastasis
II	T2 N0 M0	2 cm < Tumor < 4 cm, local to pancreas. No nodal disease/metastasis
	T3 N0 M0	Tumor > 4 cm, extended to duodenum/common bile duct. No nodal disease or metastasis
III	T1-3/4 N1 M0	Tumor extended to adjacent organs (stomach, spleen, colon/adrenals) or spread haematogenously (T4). It may present with lymph node spread (N1) but no distant metastasis
IV	Tx Nx M1	Any size (Tx), any nodal spread (Nx), but distant metastasis (M1)

Prognostic factors

SEER stages of localized, regional and distant disease can be used to categorize pNETs based on 5-year relative survival rates. Localized tumors to the pancreas report the best prognoses of up to 95%, followed by regional (adjacent organ/Lymph nodal spread; 72%) and distant metastasis to lungs, liver and bone at 23%. Overall, a 53% 5-year survival rate is reported across all stages, meaning that effective diagnosis and treatment can achieve stable disease[28].

Therapeutic approach

Treatment of pNETs falls under two main aims: Curative intent and reduced disease burden with symptomatic management. Underpinning all treatment modalities is multidisciplinary care, involving surgeons, radiologists, endocrinologists and clinical oncologists[16]. As patients with pNETs, including insulinoma, may present with varied symptoms, this is crucial to developing a personalized management plan for the patient in follow-up, maintenance and evaluating disease stability/remission status.

Surgery remains the only treatment modality with curative intent. Indications include small pNETs (< 2 cm), symptomatic non-functioning pNETs, and all functional pNETs, excluding those with positive MEN1 genes. Techniques include parenchymal-sparing pancreatic resection, either enucleation if the tumor is superficial to the main pancreatic duct or central pancreatectomy with lymph node plucking. Larger tumors with extensive margins can be excised via pancreaticoduodenectomy or distal/total pancreatectomy with regional lymphadenectomy[18].

Where surgery (open/Laparoscopic/robotic) is not indicated as in poor general health, high American Society of Anesthesiologists/Eastern Cooperative Oncology Group performance score, extrahepatic metastases or pNEC, local measures can be trialled such as ablation using radiofrequency, microwave, laser or percutaneous or endoscopic cryotherapy[19]. Indications include hepatic metastases < 5 cm and ablation margins > 1 cm. Transarterial embolization can afford debulking via chemo/selective internal radiotherapy with Yttrium-90 isotope. Pancreatitis is the most common reported risk post-local measures. The consensus is that surgery is performed with curative intent for G1/2 tumors even with nodal/hepatic metastasis, coupled with hepatectomy or staged liver resection, followed by liver transplantation if necessary[19]. The role of whether curative intent can be achieved via surgery for G3 tumors over debulking is still debated in the literature.

Medical measures are indicated for symptomatic, functional pNETs or unresectable or incompletely resected tumors to suppress post-operative functionality. Recurrence can be managed symptomatically. In the case of hypoglycaemia secondary to insulinoma, this includes dietary adjustments of frequent small meals, followed by diazoxide (sulphonylurea receptor 1 blocker), which raises glucose secretion by hyperpolarising K⁺-ATP (potassium-adenosine triphosphate) channels on pancreatic islet β-cells. For tumours that express somatostatin receptors, somatostatin analogs (SSAs) such as octreotide, lanreotide, or pasireotide have anti-proliferative effects. Long-acting SSA is used as standard therapy, and short-acting can be used as rescue drugs; the frequency of both can be increased if the biochemical/clinical picture does not show improvement. Pasireotide has activity against all somatostatin receptors, except SSTR4, and octreotide/Lanreotide has selective action towards SSTR2, 3 and 5, with maximal selectivity towards SSTR2[29]. According to the CLARINET study, greater progression-free survival of 32.1% is achieved by lanreotide Autogel compared to octreotide long-acting[30].

For locally advanced, unresectable or metastatic tumors that do not respond to SSA, molecular-targeted therapies can be trialled. Particularly for insulinomas, everolimus acts as an mTOR (mammalian target of rapamycin) pathway inhibitor, and sunitinib as a VEGF/PDGF (vascular endothelium growth factor/platelet-derived growth factor) inhibitor under the broader category of tyrosine kinase pathway inhibitors[13,19]. These can be used individually or in combination. Before these molecular-targeted therapies are established, neoadjuvant or adjuvant chemotherapy with alkylating agents (streptozotocin), in combination with 5-fluorouracil or doxorubicin (e.g., FOLFOX/FOLFIRI) showed up to 69% response rate in tumors with high disease burden, warranting systemic mechanism of action[13,19].

Recent research is showing the promise of peptide receptor radionuclide therapy (PRRT) for the management of functioning metastatic insulinomas with refractory hypoglycaemia. This treatment modality couples the use of SSA therapy, labelled with a radioactive isotope (Luteium-177-DOTA-Tyr3-octreotate) or Indium-111 octreotide, to act as targeted systemic radiation. The choice of this 177 Lu-DOTATE isotope is preferred due to the emission of concomitant beta and gamma-rays, reduced renal and haematological toxicity and modification of C-terminal threoninol on octreotide to threonine to increase affinity to SSRT2[31]. Positive results include improved quality of life secondary to better glycaemic control (70.6%) and stable disease (23.5%)[32]. This is further supported by the fact that 81% of patients having a reduced hypoglycaemic score, 58% requiring reduced antihypoglycaemic medication and median overall survival and progression-free survival of 19.7 months (95%CI: 6.5-32.9 months) and 11.7 months (95%CI: 4.9-18.5 months), respectively, post-PRRT[29].

As per current accepted European neuroendocrine tumor society (ENETS) guidelines[33], an algorithm for the management of pNETs is summarized in Figure 1, according to functional status.

Figure 1 Management of pancreatic neuroendocrine tumors, according to functional status, adapted from European neuroendocrine tumor society 2023[33,34]. SSR: Somatostatin receptor; SSA: Somatostatin analog; PRRT: Peptide receptor radionuclide therapy.

Treatment strategy

Based on the latest available guidance, the treatment strategy for non-functioning pNETs can be summarized for ease of understanding[34]. Patients with locally advanced NF-Pan-NET (stage T3 and T4) can be resected safely with low mortality and acceptable morbidity risk in expert centres. Radical local resection (R0) including portal-venous resection could be considered in selected cases. To help estimate risk of recurrence post-local measures, and guide follow-up schedules, nomograms after resection are recommended. In locally advanced or oligometastatic cases, preoperative treatment with PRRT can be beneficial in reducing tumor bulk.

For slow-growing, advanced G1-G2 non-functioning pNETs that are SSR positive, SSA is the recommended upfront treatment. In progressive G1-G2 non-functioning pNETs with SSR positivity, targeted therapy with everolimus and sunitinib are recommended. These agents should be considered in G3 progressive disease as well. PRRT may be considered second-line in non-functioning pNETs that are SSR positive. Finally, for metastatic disease, systemic chemotherapy (temozolomide in combination with capecitabine or streptozotocin + 5-FU) may be considered for patients with progressive/metastatic or symptomatic non-functioning G1-G2. For G3 metastatic non-functioning pNETs, temozolomide in combination with capecitabine can also be considered for upfront treatment[34].

CASE SUMMARY AND MANAGEMENT

A 52-year-old male presented with new recurrent, severe hypoglycaemic episodes, including nocturnal symptoms and loss of consciousness. Relevant background is a 5-year history of weight gain (body mass index 33 kg/m²), fatigue, general malaise, loss of appetite and intermittent undiagnosed abdominal pain. Investigations revealed new-onset diabetes and a suspicious lesion on the liver and pancreas on ultrasound. The patient was prescribed insulin (Abasaglar and Humalog Kwikpen) and referred for a CT scan of the thorax, abdomen, and pelvis (TAP), which identified a large (10 cm × 9.5 cm) primary mass replacing the pancreatic tail (Figure 2A), extending diffusely into the pancreatic body. A fasting gut hormone profile suggested that the tumor was nonfunctional at the initial diagnosis. Bulky pathological lymph nodal disease was noted in the periportal territory (4.8 cm max diameter) around the superior mesenteric artery and coeliac axis. A large central necrotic mass was noted in segment 8 of the liver, measuring 15 cm (Figure 2B), with small benign cysts seen in the left hepatic lobe.

Figure 2 Initial CT-TAP confirmed 10 cm × 9.5 cm 1° mass in the pancreatic tail, extending diffusely into the pancreatic body. A and B: Evidence of metastasis – 15 cm necrotic mass in segment 8 of the liver (B); C: Increased Octreotide uptake is shown on the OctreoScan, confirming the presence of somatostatin receptor positivity.

An ultrasound-guided liver biopsy of the right lobe of the liver, with histological analysis, confirmed Grade 2 NET, Ki67, 5%-10%. Planar imaging at 0004h, 0024h, and 0040h with SPECT CT Abdomen and Pelvis at 24 hours confirmed the pancreatic tail mass, periphery of large hepatic metastasis in segment 8, large portocaval node and additional 2 areas of nodal disease in the upper abdominal mesentery, with increased Octreotide uptake confirming SSTR positivity (Figure 2C). As per the guidelines, no detectable peptides on biopsy specimens and clinical picture at the time of diagnosis did not exhibit features of hormone hypersecretion syndrome, the diagnosis was non-functioning NET with metastasis[20].

In line with literature, surgical resection was not performed due to widespread liver metastasis and the advanced disease stage[35,36]. Systemic therapy was initiated to prolong the survival and for symptomatic control. As per RADIANT-2 trial[37], SSA therapy (Octreotide LAR 30 mg, subcutaneously monthly) and targeted therapy (Everolimus 10 mg, oral, for 28-day cycles). The patient developed thrombocytopenia and a generalized itchy rash with maculopapular lesions, so Everolimus was subsequently changed to Sunitinib. 47 cycles of SSA and Everolimus were completed, and he was monitored bimonthly via CT-TAP (Figure 3, showing the smallest size of the primary tumor at 20 months post-treatment).

Figure 3 The smallest 1° tumor size 20 months post-initial treatment with Everolimus and somatostatin analogue therapy.

Disease progression was observed at 48 and 49 months (Figure 4), in line with a worsening clinical picture of new hypoglycaemic episodes (> 4 times weekly) and the development of biliary sepsis requiring hospitalization. Insulin was stopped due to frequent hypoglycaemic attacks during the most recent admission, but the hypoglycaemic episodes continued. He was treated with diazoxide up to 200 mg daily. Even with this and continuous intravenous glucose infusions, he developed recurrent hypoglycaemic episodes while in the hospital.

Figure 4 48 months post-treatment showing disease progression (increase in 1° tumor size and secondary hepatic metastasis and nodal disease). A and B: Increased size in secondary hepatic mass (12.3 cm × 9.2 cm) at 49 months.

Further investigations were performed due to the evolving clinical picture from the initial presentation. A plasma C-peptide level of 6975 pmol/L with an insulin level 147 mU/L when plasma glucose was 2.4 mmol/L with a negative urine sulphonylurea screening test confirmed endogenous hyperinsulinism from a functional insulinoma. The patient was referred for peptide receptor radionuclide therapy (PRRT), in conjunction with SSA, due to disease progression despite systemic therapy, and died due to intractable hypoglycaemia and cardiac arrest.

DISCUSSION

The reported case highlights the initial presentation of non-functioning pNET subsequently turning out to become an insulinoma. The rate of transformation in case series is reported between 3.34% to 6.8%[38-40]. A review of the literature only highlights recent case reports with similar clinical pictures for comparison, highlighting the rarity and dilemma in clear management.

This case can be compared to that of a G3 Ki-67 index of 40% pNET with liver, lung and spinal metastasis, which was seen to turn into functional insulinoma 3 months post-diagnosis, post ineffective treatment with SSA (octreotide) to manage recurrent hypoglycaemia[40]. Glycaemic control was achieved to somewhat stable limits by Day 10 on diazoxide, octreotide and prednisolone. The case reflected reluctance from patients to try molecular-targeted therapies (everolimus/sunitinib) as a second line but also highlighted aggressive disease course in comparison to a median of 15 months in other reports[38]. The patient was given a one-off course of FOLFOX chemotherapy to manage the systemic oncological burden but had denied palliative chemotherapy and progressively declined.

A more recent case report documents a patient diagnosed with well-differentiated non-functional NET who underwent total pancreatectomy and started on an insulin regimen following surgery[41]. Two years post-surgery, the patient was identified to have recurrent disease in the mesentery and liver and received sunitinib along with PRRT. Symptoms of intermittent hypoglycaemia began a year post-treatment from recurrent disease, resulting in stopping the insulin regimen. Severe neuroglycopenic episodes and acute refractory hypoglycaemia in months following confirmed an insulinoma. Octreotide was trialled but not tolerated due to GI upset, thus, the patient’s hypoglycaemia was managed with diazoxide, dexamethasone and capecitabine (for palliation). No further chemotherapy, hormonal or immunologic therapy, was suitable and refractory hypoglycaemia was managed symptomatically as and when.

Overall, these cases, including our own, report the importance of recognizing new hypoglycaemia as a red-flag symptom for an investigation into malignant insulinoma developing from a previous nonfunctional pNET. Biochemical analysis is necessary and prompts a multidisciplinary care approach to tailor individualized treatment plans as the transformation from non-functioning to functioning pNET confers a poor prognosis and high symptomatic burden for patients. Ultimately, the prognosis is multifactorial and dependent on grade, stage and prognostic factors, but a coherent investigation can afford the most targeted treatment for tumor characteristics.

CONCLUSION

In conclusion, functioning neuroendocrine tumors are rare, and prompt management is difficult due to presenting signs and symptoms. Although extremely rare, a transformation of a nonfunctioning pNET to a functioning pNET (insulinoma) should be considered when unexplainable hypoglycaemia ensues in such a patient. Thus, this review and novel case highlight that further studies are warranted to understand the frequency of functional transformation of non-functioning pNETs. Moreover, managing refractory hypoglycaemia from malignant insulinomas is challenging, and there is an avenue for constant evidence-based recommendations of therapeutic efficacy.