A case of extragastrointestinal stromal tumor complicated by severe hypoglycemia: a unique presentation of a rare tumor

The true incidence of NICTH is unclear, but has been estimated to be extremely rare at approximately one per one million person years [10]. Hypoglycemia is thought to arise from excessive tumor production of IGF-II or its precursor high molecular weight “big” IGF-II. Hence the term “IGFII-oma” is often used to describe these tumors [1]. Both IGF-I and IGF-II are structurally related to insulin and act at peripheral insulin receptors to mimic insulin action, leading to increased glucose uptake, suppression of lipolysis, and inhibition of glycogenolysis, gluconeogenesis, and ketogenesis in the liver [11].

Under normal conditions, the liver produces IGF-II, in a growth hormone-independent manner, and IGF-II forms a biologically inactive ternary complex with IGF binding proteins (IGFBP), the most common of which are IGFBP3 and acid labile subunit (ALS) [12, 13]. Tumor-produced IGF-II has an equal affinity for IGFBPs compared to IGFI and II. However, instead of forming an inactive ternary complex, IGF-II forms smaller binary complexes with IGFBP. In NICTH, there is an increase in the quantity of unbound and active IGF-II as well as increased formation of these active binary complexes, which exhibit increased vasculature permeability and significantly increased bioavailability. As a side mechanism, increased IGF-II in turn further suppresses hepatic IGFBP production, thus propagating further insulin like activity and creating a vicious cycle of severe and symptomatic hypoglycemia [12].

NICTH is typically a diagnosis of exclusion, but should be suspected when hypoglycemia without hyperinsulinemia is present. Biochemically, insulin, proinsulin, C-peptide, growth hormone and ?-hydroxybutyrate levels are low at the time of hypoglycemia, as was observed in this patient. In contrast, IGF-II or IGF-II precursor levels are often elevated. Typically an IGF-II:IGF-I ratio of 10:1 is considered pathognomonic, while a normal ratio is around 3:1 [1, 12]. In our patient, the levels were 9.7:1, which were quite high, though perhaps not diagnostic.

The range of tumors associated with NICTH is broad. In recent years, a handful of case reports and case series have described GIST-associated NICTH [4, 9, 11]. GIST tumors arise from the gut wall. In rare instances, they arise outside the GI tract in the omentum, mesentery, and retroperitoneum and are referred to as EGISTs. EGISTs share the same histological and immunotypic features and biological behavior as GISTs [4, 14]. The diagnosis of GIST/EGIST tumors relies on CD117 (c-KIT) positivity on immunohistochemical staining, which remains a highly sensitive and specific marker.

Discovered on GIST-1 (DOG1) protein is another recently discovered tumor marker that has particularly utility in identifying tumors harboring mutations in platelet-derived growth factor-? (PDGFR-?) [15]. Whereas staining for both CD117 and DOG1 were observed in this patient’s tumor, formal mutation analysis was not performed. Surgery remains the mainstay for localized GIST/EGIST. However, immunotherapy with tyrosine kinase inhibitors (TKI), especially imatinib, has emerged as a promising neoadjuvant or alternative therapy. TKIs are especially useful in surgically unresectable or malignant tumors, and the advent of TKIs has increased median survival in such cases by nearly 50% [16, 17].

Besides resection and therapy aimed at shrinking the underlying tumor, management of NICTH is typically independent of its location. A variety of approaches have been described in literature, including steroids, octreotide and human recombinant growth hormone (hGH) [18, 19]. Notably, here we describe the use of continuous glucagon infusion as an effective treatment in the acute setting, especially when intravenous dextrose is ineffective alone. Among long-term interventions, glucocorticoids have demonstrated the most efficacy in terms of reversing the biochemical abnormalities associated with tumor production of IGF-II [18, 20]. The proposed mechanism of glucocorticoids is thought to be suppression of IGF-II production by the tumor and/or its increased sequestration. Indeed published studies have demonstrated a significant fall in circulating big IGF-II, accompanied by an increase in serum ALS in response to prolonged glucocorticoid use. Interestingly, these changes were found to be reversible with glucocorticoid withdrawal. Furthermore, the required dose of steroid appears to be somewhat individualized and based on both tumor debulking strategies and the co-administration of other agents [18, 20]. Table 1 summarizes available published studies in which steroids were used either solely or in conjunction with other therapies in cases of inoperable NICTH (Table 1). In aggregate, the initial dose of glucocorticoid was at least 30 mg or its steroid equivalent dose, with attempts made to taper therapy based on the maintenance of euglycemia [19–21]. The minimum effective daily dose in cases employing only steroids was 20 mg once daily or 5 mg three times a day, with further dose reductions resulting in relapse of hypoglycemia [18, 20]. The ideal duration of therapy remains unclear from the literature, with most studies having variable follow-up.

Our patient continued to have severe hypoglycemia as an inpatient. This was acutely treated with continuous infusions of both glucagon and dextrose. Once the diagnosis of recurrent NICTH was established, she was successfully transitioned to prednisone 40 mg once daily, and euglycemia was achieved. Since surgical resection was not a viable option given the extent of her disease, imatinib was restarted after discussion with oncology with the aim of decreasing tumor burden. Over the next few months, she was successfully weaned off all steroids and repeat imaging demonstrated some shrinkage of the tumor with TKI therapy.