| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
CLINICAL CASE SEMINAR |
Department of Medicine (M.A.N., A.M.F., W.S.), Ruhr-University, Knappschafts-Krankenhaus, D-44892 Bochum, Germany, and Diabeteszentrum Bad Lauterberg (M.A.N.), D-37431 Bad Lauterberg im Harz, Germany; Division of Neuroendocrinology (M.R., G.B., J.Z.), Institute of Anatomy, University of Zurich, and Department of Internal Medicine (C.Z., J.Z.), University Hospital, Zurich, Switzerland; Institute of Pathology (A.P.), University Hospital, CH-8057 Zürich, Switzerland; Medical Research Laboratories (J.F., A.F.), Institute of Experimental Clinical Research, Aarhus University Hospital, DK-8000 Aarhus, Denmark; Department of Internal Medicine (P.G.L.), Städtische Kliniken D-21339 Lüneburg, Germany; Department of Pediatrics (W.F.B.), University of Giessen, Giessen, and Eli Lilly & Co., D-61350 Bad Homburg, Germany; and Department of Pathology (G.K.), Christian-Albrecht-University, D-24105 Kiel, Germany
Address all correspondence and requests for reprints to: Michael A. Nauck, M.D., Diabeteszentrum Bad Lauterberg, Kirchberg 21, D-37431 Bad Lauterberg, Germany. E-mail: M.Nauck{at}diabeteszentrum.de.
| Abstract |
|---|
|
|
|---|
Objective: The objective was to investigate whether the patients hypoglycemia was due to excessive IGF-I production by the tumor.
Methods: Serum IGF- I and IGF-II, insulin, and GH were measured by RIA; the distribution of IGFs between IGF binding protein complexes in serum was analyzed after neutral gel filtration. Tissue IGF-I was identified by immunohistochemistry and in situ hybridization, and by RT-PCR after RNA extraction.
Results: Total and free serum IGF-I, but not total, free, and big IGF-II, was increased, and the IGF-I content of the two IGF binding protein complexes was elevated. Immunohistochemistry demonstrated IGF-I peptide in situ hybridization IGF-I mRNA in the lymph node metastasis. Combined GH/glucocorticoid treatment prevented hypoglycemia, but did not lower IGF-I. After chemotherapy with carboplatinum/etoposide, the lung nodules largely regressed, and serum IGF-I and the IGF-I content of the two binding protein complexes became normal. Hypoglycemia did not recur despite discontinuation of GH/glucocorticoid treatment.
Conclusion: Our findings are compatible with a new form of tumor hypoglycemia caused by circulating tumor-derived IGF-I.
| Introduction |
|---|
|
|
|---|
NICTH often results from the overproduction and secretion by the tumor of incompletely processed or unprocessed pro-IGF-II peptide (6, 7, 8), termed "big" IGF-II (molecular mass, 1017 kDa, in contrast to mature IGF-II of 7.5 kDa). Big IGF-II impairs formation of a heterotrimeric 150-kDa IGF binding protein (IGFBP) complex in the circulation (9), which consists of IGF-I, IGF-II, IGFBP-3, and an acid-labile subunit (ALS). In normal individuals, most of the circulating IGFs are bound to this complex. Thereby, IGFs are prevented from displaying their insulin-like potential. However, when formation of the 150-kD complex is impaired, IGFs are sequestered in a binary complex with IGFBP-3 (7). In the latter complex, the bioavailability of IGFs is increased and their insulin-like potential is unmasked leading to enhanced peripheral glucose consumption (10, 11, 12) and suppressed hepatic glucose production (10, 11). Typically, serum insulin levels are low or unmeasurable in NICTH, and serum GH levels are suppressed.
Patients with clinical NICTH do not always present with elevated "big" IGF-II (13). Some cases are due to ectopic insulin production, e.g. from a bronchial carcinoid (14), and ovarian carcinoma (15), or a small-cell carcinoma of the cervix (16).
So far, hypoglycemia due to an IGF-I-producing tumor has not been reported. We recently identified a patient with severe recurrent hypoglycemia who had elevated IGF-I but normal big IGF-II serum levels. Extensive clinical testing excluded an insulinoma, as well as pituitary or adrenal insufficiency. We therefore set out to investigate whether the patients recurrent hypoglycemia was due to overproduction and secretion of IGF-I by a tumor and to characterize a hitherto undiscovered form of NICTH.
| Patient and Methods |
|---|
|
|
|---|
A 64-yr-old woman was admitted to the emergency room in February 1999. She was unconscious and had convulsions. Hypoglycemia was diagnosed and corrected by iv glucose administration. Hypoglycemia frequently recurred, especially during the night.
The patient had a history of Sjögren syndrome (lymphocytic sialadenitis with xerostoma) but was otherwise healthy. She was slightly overweight (76.6 kg, 168 cm, body mass index 27.1 kg/m2) and had no signs of acromegaly. No abdominal mass or enlarged lymph nodes were palpated. A chest x-ray, an abdominal ultrasound, a mammography, and a somatostatin receptor scintigraphy did not show any tumor. Liver and kidney function was unimpaired. Thyroid disease and pituitary or adrenal insufficiency were excluded.
Treatment
Because serum IGF-I was elevated, GH and big IGF-II were normal, and serum insulin and C-peptide were suppressed, we considered IGF-I-induced tumor hypoglycemia, although no tumor had initially been found. Therefore, in analogy to IGF-II-induced tumor hypoglycemia, treatment with GH (17.4 µg/kg·d im Norditropin; NovoNordisk, Mainz, Germany) and prednisone (40 mg/d) (9), was started. Hypoglycemic episodes ceased. GH and prednisone were gradually tapered to 5.2 µg/kg·d and 2.5 mg/d, respectively. This dose was given until October 2004.
Seven months after the first admission, the patient palpated a mass in the right axilla. A lymph node (diameter, 3.5 cm) was surgically removed. Histological examination revealed an undifferentiated large-cell carcinoma. A computed tomography scan showed several small nodules in the right lung. Sequential therapy with Gemcitabin/Vinorelbine, Docetaxel, and Iressa (a vascular endothelial growth factor receptor antagonist) failed to stop further dissemination in the lungs despite transient partial remission. Finally, treatment with four cycles of carboplatinum (500 mg)/etoposide (190 mg) from December 2003 to March 2004 led to the disappearance of the lung nodules as documented by computed tomography. GH and prednisone was stopped in October 2004, and hypoglycemia did not recur. Fasting was tolerated for 24 h, and 24-h profiles of glucose, insulin, and C-peptide concentrations showed normal values throughout the day (Fig. 1
). In 2005 a cerebellar metastasis was surgically removed, and the patient died in 2006.
|
Glucose was measured by the glucose oxidase method. Plasma insulin was determined by a microparticle enzyme immunoassay (IMx Insulin; Abbott Laboratories, Wiesbaden, Germany), C-peptide by ELISA (Dako Diagnostics Ltd., Cambridgeshire, UK). Total serum IGF-I was determined by RIA (17) after chromatography of serum samples over Sep-Pak C18 cartridges (Waters Corp., Milford, MA). Free IGF-I and IGF-II levels were determined as described previously (8). Total ALS was determined by ELISA (Diagnostic Systems Laboratories, Webster, TX). GH was measured using Immulite hGH (EURO/DPC; Llanberis, Gwynedd, UK). IGFBP-1, IGFBP-2, and IGFBP-3 were determined by RIA as described (18, 19).
Gel filtration. Acidic Biogel P-60 gel filtration (Bio-Rad Laboratories, Richmond, CA) and neutral (pH 7.4) Sephadex G-200 gel filtration (Pharmacia LKB Biotechnology Inc., Uppsala, Sweden) were performed as described (7). IGF-I and IGF-II in pooled fractions were measured by RIA (17) after chromatography over Sep-Pak C18 cartridges (Waters Corp.). SDS-PAGE and ligand blot analysis were performed as described (20).
Histology and immunohistochemistry. Four-micrometer sections from formalin-fixed paraffin-embedded resection specimens (axillary lymph node) were stained with hematoxylin-eosin and periodic acid Schiff reagent. IGF-I peptide was identified by immunohistochemistry (21).
In situ hybridization. Digoxigenin-labeled probes for in situ hybridization were prepared by in vitro transcription. Templates were generated by PCR amplification of total RNA using human IGF-I-specific reverse and forward primers with T7 and T3 RNA polymerase binding sites attached to the five prime ends (21): T7, reverse primer: GGATCCTAATACGACTCACTATAGG-GCATGTCGGTGTGGC; T3, forward primer: AAT TAA CCC TCA CTA AAG G-CA GTC TTC CAA CCC AAT.
The primers amplify a segment of 374 nucleotides spanning the B and C domains and the first half of the A domain. From the promoter sequences, antisense and sense probes were transcribed using the digoxigenin in vitro transcription kit (Roche, Basel, Switzerland) as described (21).
Identification of IGF-I by PCR
RT-PCR primers spanned intron 1 of the IGF-I gene or the GAPDH gene (IGF-I forward: 5'-CTGAGCTGGTGGATGCTCTT-3'; IGF-I reverse: 5'-CACTGCTGGAGCCATACCC-3'; GAPDH forward: 5'-CCATGGAGAAGGCTGGGG-3'; GAPDH reverse: 5'-TTCACACCCATGACGAACAT-3'), leading to PCR products of 84 bp and 97 bp, respectively. RNA was extracted from paraffin-embedded tissue (22). RNA from a papillary thyroid carcinoma and blood leukocytes was used as positive and negative controls. After RT, semiquantitative RT-PCR was performed, and 5 µl of the amplificate after 30, 35, 40, and 45 PCR cycles was loaded on a 2% agarose gel. The bands for IGF-I products were excised, purified, and cycle-sequenced (22).
| Results |
|---|
|
|
|---|
On admission of the patient, total and free serum IGF-I was increased (692 ng/ml and 27.2 ng/ml, respectively), whereas total and free IGF II was normal (548 and 0.73 ng/ml). Four weeks after initiating GH/prednisone treatment, total and free IGF-I was still elevated. However, after chemotherapy with carboplatinum/etoposide and discontinuation of GH/prednisone treatment (October 2004), total IGF-I had fallen to 283 ng/ml. Total IGF-II stayed within the normal range. IGFBP-1 and IGFBP-2 were decreased. IGFBP-3 was normal but rose during GH/prednisone treatment.
|
The patients ternary (150 kDa) and binary (50 kDa) IGFBP complexes obtained by neutral Sephadex G-200 gel filtration each contained more IGF-I than the corresponding complexes of normal serum (Fig. 2A
). IGF-II in the 150-kDa complexes of the two sera was similar; the 50-kDa complex of the patients serum contained somewhat less IGF-II than that of the normal serum (Fig. 2B
). After the last chemotherapy, the IGF-I content of either complex was similar to that of normal serum. As shown by ligand blotting of the Sephadex G-200 serum fractions (Fig. 2
, C and D), most of the IGFBP-3 was associated with the 150-kDa complex in both the patients serum and normal serum.
|
IGF-I peptide and mRNA in the tumor metastasis.
Immunohistochemistry demonstrated IGF-I peptide in some tumor cells as well as in extratumoral tissue (Fig. 3A
). IGF-I mRNA was detected by in situ hybridization in virtually all tumor cells and much less frequently in extratumoral tissue (Fig. 3B
). In the latter, the number of positive cells was in a similar range as that found by immunohistochemistry. mRNA extracted from the tumor-infiltrated lymph node, reverse-transcribed, and amplified by PCR (Fig. 3C
) gave the cDNA sequence of a fragment that was identical to intron 1 of the published IGF-I cDNA sequence.
|
| Discussion |
|---|
|
|
|---|
The lymph node metastasis detected several months after the first hypoglycemic symptoms most likely originated from a pulmonary large cell carcinoma. The metastasis expressed IGF-I mRNA in nearly all tumor cells but contained only small amounts of immunoreactive IGF-I peptide. These findings suggest a high rate of newly synthesized IGF-I, which was rapidly released from the tumor into the circulation, resulting in elevated total and free IGF-I serum concentrations and finally in hypoglycemia. The failure of GH to rise after arginine and GHRH administration is probably due to feedback inhibition through circulating tumor-derived IGF-I (23, 24, 25). The absence of acromegalic features in our patient may be explained by 1) the absence of a sustained combined elevation of GH and IGF-I (Table 1
); 2) the relatively short duration of paraneoplastic hormonal changes; 3) a relatively high background insulin sensitivity; and 4) hormonal changes induced by repeated hypoglycemic episodes added on to chronic elevations in IGF-I.
Normally, 8090% of the circulating IGF-I and IGF-II is bound to IGFBP-3 and ALS in a ternary 150-kDa complex. In big IGF-II-related NICTH, big IGF-II impairs the formation of the 150-kDa complex (9). IGF is thereby sequestered into a binary 50-kDa complex (3, 8). In this complex the bioavailability of the bound IGF is enhanced (26) because it can cross the capillary wall (27) and reach insulin target tissues to exert insulin-like effects. Indeed, the patients 50-kDa complex contained a 3-fold higher amount of IGF-I than the 50-kDa complex of normal serum. In combination with the elevated concentration of free IGF-I, this has probably contributed to the development of hypoglycemia.
The same symptomatic therapy (GH and glucocorticoids) as in IGF-II-mediated NICTH was effective in our patient. The mechanisms likely include GH-induced stimulation of IGFBP-3 (28), which may have prevented sequestration of free IGF-I, but above all GH-induced and glucocorticoid-induced insulin resistance (29, 30) and, thus, resistance to the insulin-like actions of IGF-I. IGF-I was not further elevated by GH treatment, indicating that IGF-I secretion from the tumor was autonomous and, therefore, relatively independent from GH as a stimulus. After successful chemotherapy, when GH treatment was discontinued, IGF-I concentrations fell, almost into the normal range.
After "successful" chemotherapy and after stopping GH/prednisone treatment, total serum IGF-I had fallen into the normal range, the IGF-I content of the 50-kDa complex was no longer elevated, and hypoglycemic episodes did not recur. All of these findings support the conclusion that secreted paraneoplastic IGF-I had caused hypoglycemia.
| Acknowledgments |
|---|
| Footnotes |
|---|
Disclosure Statement: The authors have no conflict of interest to declare in relation to the content of this manuscript.
First Published Online February 13, 2007
Abbreviations: ALS, Acid-labile subunit; IGFBP, IGF binding protein; NICTH, nonislet cell tumor hypoglycemia.
Received November 22, 2006.
Accepted February 1, 2007.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
A. Ashare, A. B. Nymon, K. C. Doerschug, J. M. Morrison, M. M. Monick, and G. W. Hunninghake Insulin-like Growth Factor-1 Improves Survival in Sepsis via Enhanced Hepatic Bacterial Clearance Am. J. Respir. Crit. Care Med., July 15, 2008; 178(2): 149 - 157. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. W. B de Groot, B. Rikhof, J. van Doorn, H. J G Bilo, M. A Alleman, A. H Honkoop, and W. T A van der Graaf Non-islet cell tumour-induced hypoglycaemia: a review of the literature including two new cases Endocr. Relat. Cancer, December 1, 2007; 14(4): 979 - 993. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. H. Daughaday Hypoglycemia due to Paraneoplastic Secretion of Insulin-Like Growth Factor-I J. Clin. Endocrinol. Metab., May 1, 2007; 92(5): 1616 - 1616. [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |