A 56-yr-old man was admitted to our department because of recurrent and severe hypoglycemia. Ten years previously the patient had undergone surgery for an intraabdominal hemangiopericytoma, a mesenchymal tumor. An ileocolectomy with colostomy was performed about twenty months later because of local recurrence of the tumor. He was apparently well thereafter until 2 years before the admission to our department, when he was treated with chemotherapy and radiotherapy because of a small mass in the pelvis evidenced at a computed tomography scan. Two months before admission to our department, he was brought to the emergency room in a state of unconsciousness and with a plasma glucose level of 20 mg/dL (1.1 mm). He was given 20% dextrose iv, and he promptly recovered. The hypoglycemic episodes recurred and became more frequent. Upon admission to our department the patient was first treated with 50% dextrose infusion through a subclavian vein. Nevertheless episodes of hypoglycemia persisted, especially at night, when fasting was longer than 4–5 hr. Increased meal frequency with complex carbohydrates administration plus continuous dextrose infusion were no longer able to control hypoglycemia. The patient had normal-to-low plasma insulin levels and very high levels of insulin-like growth factor (IGF)-II, especially in the isoform 10–17 kDa, around 80% of total measured IGF-II (Fig. 1and Table 1 ). The diagnosis of nonislet cell tumor hypoglycemia (NICTH) was proposed. Evaluation of liver showed metastatic lesions, but because of the size, a palliative surgical attempt was not advised. Therapy with both prednisone (15 mg/day per os) and, after informed consent, biosynthetic GH (2 U/day sc) at bedtime was soon started. Dextrose infusion was gradually reduced and eventually suspended 4 days after beginning continued hormone treatment. No further episode of hypoglycemia occurred in the following 3 days, and he was discharged from the hospital in apparent good condition with instructions for blood glucose self monitoring and glucagon use in the case of severe hypoglycemia. In the following month only one episode of hypoglycemia was reported. The patient was admitted again to the hospital for control (Table 1). Blood glucose levels were low only sporadically, especially during afternoon and the night. The patient showed a state of relative well-being and no relevant side-effect of GH treatment. The prednisone and GH dose were reduced to 10 mg and 1.5 U/day, respectively. Three months later the patient decided to discontinue GH treatment because of concern of an effect on tumor growth. The patient’s condition rapidly worsened, with increasingly frequent episodes of hypoglycemia, and iv glucose treatment was again required. The patient died 5 months later. It has been reported that mesenchymal tumors may produce and release an excessive amount of IGF-II as well as the isoform called big IGF-II (1). This high molecular weight IGF-II is considered the causative agent of NICTH because of its insulin-like effect, exherted primarily by stimulating glucose uptake at peripheral tissue level (2). In addition, the increased serum IGF-II inhibits both GH and insulin secretion and, as a consequence, lowers the plasma concentration of insulin and GH-dependent IGF binding proteins (mainly IGFBP-3). This in turn increases the unbound IGF-II fraction and worsens its hypoglycemic effect (3). When surgical removal or radiotherapy is not advised, many other strategies may ameliorate a patient’s severe hypoglycemia, including corticosteroid administration (which, in addition to its antihypoglycemic action, may also suppress IGF-II production by the tumor) (4) and subcutaneous glucagon infusion (5). GH treatment has also been tried in such patients (6, 7), with the postulated mechanism of reducing IGF-II availability to tissues by increasing IGFBP-3 levels and the acid labile subunit (ALS) that will form the “ternary complex” containing bound (biologically nonactive) IGF-II. In vitro GH does not appear to have any significant mitogenic effect (8, 9). Increased cancer prevalence in acromegaly is probably a consequence of elevated IGF-1 levels (10). In our patient, however, plasma IGF-1 remained low during hGH treatment at the doses used (Table 1). In this very severe case of NICTH the combined treatment with prednisone and GH proved effective in ameliorating the hypoglycemic syndrome without side-effects. It can therefore be proposed as a cost-effective palliative treatment in these patients. Western blot analysis of IGF-II of the patient. After acid-ethanol extraction of 300 μL of serum, IGF-II was immunoprecipitated with an anti IGF-II antibody 2H-11 conjugated resin. The samples were solubilized, electrophorezed, and analyzed by an anti-IGF-II antibody (2H-11). Immunoblotting indicates that the serum of the patient shows an abundance and a size-heterogeneity of IGF-II compared with control serum. The molecular weight was between 10–17 kDa, rather than the 7.5 kDa found in two normal control subjects. Western blot analysis of IGF-II of the patient. After acid-ethanol extraction of 300 μL of serum, IGF-II was immunoprecipitated with an anti IGF-II antibody 2H-11 conjugated resin. The samples were solubilized, electrophorezed, and analyzed by an anti-IGF-II antibody (2H-11). Immunoblotting indicates that the serum of the patient shows an abundance and a size-heterogeneity of IGF-II compared with control serum. The molecular weight was between 10–17 kDa, rather than the 7.5 kDa found in two normal control subjects. Hormonal measurements before and after 1 month therapy with GH, 2 U/day sc, and prednisone, 15 mg/day po (m± sd of 4 different measurements) Normal values. Insulin: 2–10 μU/mL; IGF-I: 78–250 ng/mL (age 40–70); IGF-II: 290–740 ng/mL; IGFBP-3: 1.7–4.0 μg/mL. Assay performed with standard commercial immunoassays. Hormonal measurements before and after 1 month therapy with GH, 2 U/day sc, and prednisone, 15 mg/day po (m± sd of 4 different measurements) Normal values. Insulin: 2–10 μU/mL; IGF-I: 78–250 ng/mL (age 40–70); IGF-II: 290–740 ng/mL; IGFBP-3: 1.7–4.0 μg/mL. Assay performed with standard commercial immunoassays.
No takes yet. Share an insight, caveat, or question.
Gullo et al. (1999) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: