| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
U.S. Food and Drug Administration-CDER (E.A.K., L.G., S.N.M.), Rockville, Maryland 20857; New York Hospital, Cornell Medical Center, Pediatric Endocrinology (J.M.G.), New York, New York 10021; and Central University Hospital-Department of Pediatrics (M.B.), Grenoble cedex 09, France
Address all correspondence and requests for reprints to: Elizabeth Koller, U.S. Food and Drug Administration-CDER, Parklawn Building, HFD 510, Room 14B04, 5600 Fishers Lane, Rockville, Maryland 20857. E-mail: kollere{at}cder.fda.gov
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
| Materials and Methods |
|---|
|
|
|---|
| Results |
|---|
|
|
|---|
The patient was a 31-yr-old white man, who incurred a traumatic
head injury at age 9. Hypothalamic injury was presumed because the
previously healthy patient experienced a retardation of linear growth
and the absence of a pubertal growth spurt (free testosterone, 33
pg/dL; range, 50260 pg/dL) resulting in a final height of 162 cm;
soon after the injury, he developed hyperphagia, and his final weight
was 185 kg. GH deficiency was suspected because of a poor response to
clonidine (<1.5 ng/mL) and subnormal serum levels of insulin-like
growth factor (IGF)-I (64 ng/mL) and IGF-binding protein 3 (1.2 mg/L)
(Nichols Institute, San Juan Capistrano, CA). Before initiation of GH
therapy at age 30, a 75-g oral glucose tolerance test (OGTT) was
performed and found to be normal. The fasting glucose was 89 mg/dL, and
the glucose peaked at 30 min, with a value of 149 mg/dL. Insulin levels
were modestly elevated and were consistent with the degree of obesity.
The fasting level was 30.4 µU/L (3.517), and the peak at 60 min was
172.1 µU/L. Serum creatinine was normal. Human GH therapy was
initiated at 0.009 mg/kg·day. Responsiveness to the initial dose of
GH was indicated by the rise in serum IGF-I to 265 ng/mL and IGF-BP3 to
3.1 mg/L. A repeat OGTT done on this dose was again normal, but the
hyperinsulinemia was more prominent. The fasting insulin was 75.9
µU/L, and the peak insulin at 90 min was 424.7 µU/L. A HgbA1c
level, obtained 3 weeks after beginning therapy, was 6.2% (<6.5%).
After 8.5 months, the dose was increased to 0.012 mg/kg·day and,
after another 4.3 months, to 0.017 mg/kg·day. The patients course
was otherwise complicated by an episode of pericarditis and
intermittent hypertension, with values ranging from 128/72 to 170/99 mm
Hg. Concomitant medications for obesity included phentermine,
fenfluramine, and pemoline. After 14 months of GH therapy, the patient
presented with a 4-week history of blurred vision [OD (right eye)
20/401; OS (left eye) count fingers]. Visual field quadrants were
intact to confrontation. Color vision was impaired (OD 2/8 correct; OS
0/8 correct) and suggested macular involvement. Pupillary reflexes were
normal. Papilledema and arteriovenous nicking were absent. Bilateral
macular edema, cotton wool spots, 12+ dot hemorrhages, and
microaneurysms were present on funduscopic exam (Fig. 1
). There was slight retinal pigment
epithelial dropout in the right eye. There was no definitive evidence
for vascular occlusion. There was leakage of microaneurysms OU (each
eye) with fluorescein angiography. Emergent laser surgery was performed
on the left eye. Visual acuity was not restored to normal with laser
surgery (best corrected: OD 20/40, OS 20/100). An OGTT at the time of
visual complaints was consistent with normal glucose tolerance
[glucose = 101 mg/dL (t = 0 min), 162 mg/dL (t = 30),
138 mg/dL (t = 60), 108 mg/dL (t = 90), and 89 mg/dL (t
= 120)]. Triglyceride and cholesterol levels were normal at 152 mg/dL
and 84 mg/dL. HgbA1c, 2 months after discontinuation of GH, was normal
(5.9%). Thirteen months later, visual acuity without correction was
20/40 OU, and eye grounds as visualized by oral angiography were
markedly improved. Eighteen months after the initial presentation, the
patients visual acuity was 20/30 OD and 20/25 OS (Humphrey
autorefractor); his funduscopic exam was essentially normal.
|
The patient was a 9-yr-old girl with Turners Syndrome (45 X/46 X mar, SRY +). The patient had short stature (<-2 SD, by National Center for Health Statistics Percentiles) but was weight-height proportionate (19 kg; 116 cm). There was no history of trauma, and her medical history was otherwise unremarkable. The patient was not receiving any concomitant medications. Baseline blood pressure was 105/60 mm Hg. Baseline HgbA1c assay was normal at 4.3% (<5.3). Fasting glucose and insulin levels were unremarkable at 77 mg/dL and 5.1 µU/L, respectively. Human GH therapy had been initiated at 0.033 mg/kg·day. The dose remained unchanged for 17 months and then was increased to 0.043 mg/kg·day. Twenty-two months after starting GH therapy, the patient presented with decreased visual acuity (OD 20/10020/200; OS 20/20). Color vision in the right eye was impaired. Angiography in the right eye was notable for edema of the optic disk, cicatrization, and subretinal and peripapillary neovascularization; there was no definitive evidence for vascular occlusion. Angiography in the left eye was normal. At the onset of visual problems, blood pressure readings were 95/60 mm Hg, and the HgbA1c remained unchanged at 4.5%. Cryoglobulin and fibrinogen levels were normal.
Literature-MedWatch data
Although nonproliferative retinopathy has been reported with pheochromocytoma and severe hypertension, the blood pressure readings of these two patients seem to be be insufficient to account for the retinal pathology (6). Phentermine and pemoline, taken by patient 1, have adrenergic properties and may be associated with blood pressure increases, but we could find no published literature or MedWatch reports associating use of these medications with retinal changes mimicking diabetic retinopathy. Furthermore, the clinical presentations of visual impairment observed with these drugs are more suggestive of cerebrovascular accidents or transient ischemic attacks (a similar stroke-like picture, with documented retinal vascular occlusion in the setting of normo-tension, has been reported with fenfluramine) (7). Finally, proliferative changes, albeit unilateral, were seen in the patient with Turners syndrome, and neovascularization is not typically seen with the retinopathy of hypertension (6).
| Discussion |
|---|
|
|
|---|
In addition to the initial report by Poulsen (4), several other clinical observations suggest that GH (and/or its hormonal mediator, IGF-I) has a proliferative effect on retinal vasculature. In a controlled trial of hypophysectomy, retinal changes were inhibited for 5 yr in patients with equivalent levels of glucose control (8). Similarly, diabetics treated effectively with pituitary implantation of radioactive yttrium had fewer hemorrhages, microaneurysms, and new vessels (9). In 1973 and 1978, Merimee (10) reported that, although GH-deficient ateliotic dwarfs exhibited carbohydrate intolerance, they did not have retinopathy. He later demonstrated that a subset of insulin-dependent diabetes mellitus (IDDM) patients, those with accelerated retinopathy, had elevated serum IGF-I levels and that the mean vitreal IGF-I concentration was greater for patients with proliferative diabetic retinopathy than in age-matched nondiabetic controls (11, 12). In addition, Alzaid et al. (13) reported that diabetic retinopathy is approximately five times more prevalent in IDDM patients who are GH sufficient than those who are GH deficient during an insulin tolerance test. Similarly, diabetics who had an increase in GH, factor VIII-related antigen, and plasminogen activator activity in response to exercise went on to develop retinopathy (14). Conversely, SRIF analogue has been used to reverse new onset retinopathy that presented after initiation of intensive insulin therapy (15)
Several laboratory models are also consistent with the aforementioned clinical observations. GH, itself, in physiologic concentrations, can stimulate the in vitro proliferation of human retinal microvascular endothelial cells (16). Similar to Merimees (17) observations in humans, dwarf transgenic mice that express a GH antagonist have a lower rate of neovascularization. Retinal microvascular cells also have receptors for GHs mediator, IGF-I (18, 19). IGF-I can induce the release of plasminogen activator and increase DNA synthesis in retinal endothelium (3, 18, 20). IGF-I also induces dose-dependent chemotactic activity in human capillary endothelial cells, and this activity is potentiated by FCS (21). Most notably, rabbits injected by the intravitreal route with supraphysiologic doses of IGF-I show retinal vascular proliferation within 10 days, and mice treated with a SRIF analogue and either GH or IGF-I had higher levels of neovascularization than animals treated with SRIF analogue alone (17, 22). Also important to the establishment of causality of retinopathy is the demonstration that such growth peptides also can be produced locally. Cultured retinal microvascular endothelial cells, pericytes, and retinal pigment epithelial cells release IGF-I, and IGF-I production increases with hypoxia (23).
Although there are compelling data to suggest that GH and/or IGF-I have a role in the pathogenesis of diabetic retinopathy, the relationship between these peptides and retinopathy seems to be complex. Indeed, the profound elevations in serum IGF-I reported by Merimee were seen only in those patients with accelerated retinopathy, and some subsequent investigators have failed to confirm his observations (11, 24, 25, 26). Notably, Hyer et al. (24) observed that seven of eight diabetic patients with preproliferative disease exhibited a transient rise in serum IGF-I proximate to the appearance of neovascularization, but that these serum concentration levels could not be distinguished from the mean value of their cohort. Moreover, Wang et al. (25) found that a single, baseline IGF-I level was not predictive for the development or progression of diabetic retinopathy in 1260 IDDM and NIDDM patients over a 6-yr span. Finally, in a cross-sectional analysis, Arner et al. (26) found that, despite equivalent glucose control, total IGF-I levels were lower in IDDM subjects with retinopathy than in controls with no or minimal background retinopathy. The patients with retinopathy, however, exhibited an alteration in the pattern of IGF-I binding to the 40-kDa and 150-kDa carrier proteins.
The contradictory nature of this information may be caused by the complexity of the GH-IGF system, as well as the evolving nature of IGF-I determinations. Historically, IGF measurements have been limited by technical difficulties, including the presence of multiple binding proteins. Most importantly, at the present time, we are still unable to assess local changes in the level of IGF-I and its binding proteins.
In conclusion, the proliferative action of GH or its hormonal mediator, IGF-I, on retinal vasculature is supported by the development or acceleration of retinopathy in these two nondiabetic patients. Hyperglycemia is not mandatory for initiating retinopathy (although it may alter its course). Baseline and periodic funduscopic evaluation should be performed in subjects receiving GH.
| Acknowledgments |
|---|
| Footnotes |
|---|
Received December 15, 1997.
Revised March 26, 1998.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
Z. T. Bloomgarden Diabetic Retinopathy and Diabetic Neuropathy Diabetes Care, March 1, 2007; 30(3): 760 - 765. [Full Text] [PDF] |
||||
![]() |
M. E. Molitch, D. R. Clemmons, S. Malozowski, G. R. Merriam, S. M. Shalet, M. L. Vance, and for The Endocrine Society's Clinical Guidelines Su Evaluation and Treatment of Adult Growth Hormone Deficiency: An Endocrine Society Clinical Practice Guideline J. Clin. Endocrinol. Metab., May 1, 2006; 91(5): 1621 - 1634. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Hansen, E. A. Koller, and S. Malozowski Full Remission of Growth Hormone (GH)-Induced Retinopathy after GH Treatment Discontinuation: Long-Term Follow-Up J. Clin. Endocrinol. Metab., July 1, 2000; 85(7): 2627 - 2627. [Full Text] |
||||
![]() |
G. Radetti, L. Gentili, and M. Lepidi Comment on Growth Hormone Therapy and Retinal Changes Mimicking Diabetic Retinopathy J. Clin. Endocrinol. Metab., February 1, 2000; 85(2): 923 - 923. [Full Text] |
||||
![]() |
E. A. Koller, L. Green, and S. Malozowski Comment on Growth Hormone Therapy and Retinal Changes Mimicking Diabetic Retinopathy J. Clin. Endocrinol. Metab., February 1, 2000; 85(2): 923a - 923. [Full Text] |
||||
![]() |
D. LeRoith Insulin-like Growth Factors in Pediatric Health and Disease J. Clin. Endocrinol. Metab., December 1, 1999; 84(12): 4355 - 4361. [Full Text] |
||||
![]() |
A. Hellström, E. Svensson, B. Carlsson, A. Niklasson, and K. Albertsson-Wikland Reduced Retinal Vascularization in Children with Growth Hormone Deficiency J. Clin. Endocrinol. Metab., February 1, 1999; 84(2): 795 - 798. [Abstract] [Full Text] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |