Progression of Retinopathy in Insulin- Treated Type 2 ... - Diabetes Care

Pathophysiology/Complications O R I G I N A L

A R T I C L E

Progression of Retinopathy in InsulinTreated Type 2 Diabetic Patients MARIANNE HENRICSSON, MD, PHD1 KERSTIN BERNTORP, MD, PHD2

PER FERNLUND, MD, PHD3 GORAN ¨ SUNDKVIST, MD, PHD2

OBJECTIVE — To study the progression of retinopathy 3 years after initiation of insulin therapy. RESEARCH DESIGN AND METHODS — In a prospective, observational case-control study, 42 type 2 diabetic patients were examined at baseline and 1, 3, 6, 12, 24, and 36 months after change to insulin therapy. Retinopathy was graded based on fundus photographs using the Wisconsin scale; HbA1c and IGF-1 were measured. RESULTS — During the observation period of 3 years, 26 patients progressed in the retinopathy scale; 11 patients progressed at least three levels. After 3 years of insulin therapy, HbA1c and IGF-1 were significantly lower than at baseline. Progression of retinopathy greater than or equal to three levels was related to high IGF-1 levels. CONCLUSIONS — A relationship was found between high IGF-1 levels at 3 years and progression of retinopathy in type 2 diabetic patients.

In a randomized controlled study, Grant et al. (18) showed that the progression to high-risk proliferative diabetic retinopathy (PDR) was diminished in patients treated with ocreotide, a somatostatin analog decreasing the secretion of GH. We have now followed the type 2 diabetic patients of our previous 2-year study regarding the progression of retinopathy. In this prospective case-control study, we conducted eye examinations 3 years after insulin had been initiated in type 2 diabetic patients previously treated with oral agents. The aim was to assess potential risk factors for retinopathy progression in this patient group. RESEARCH DESIGN AND METHODS

Diabetes Care 25:381–385, 2002

I

n diabetic retinopathy (DR), retinal capillaries become occluded, resulting in large areas of retinal nonperfusion and prompting new vessel formation (1). Hyperglycemia is regarded as a major cause of retinopathy. Improved glycemic control retards the development and progression of retinopathy in both type 1 and type 2 diabetes (2– 4). However, worsening of retinopathy has been reported after rapidly improved glycemic control (2,5–9). A relationship between the degree of improvement in HbA1c and early worsening of retinopathy was found in the Diabetes Control and Complications Trial (DCCT) (8). In agreement, we found a relationship between the degree of improvement in HbA1c and progression of retinopathy 2 years after the start of insulin therapy in type 2 diabetic patients (9).

Endocrine and growth factors may also play a role in the pathogenesis of DR. The secretion of growth hormone (GH) is increased in poorly controlled diabetes (10,11), and DR correlates with the magnitude of GH hypersecretion (12). A role for pituitary-associated factor in DR was first postulated in 1953, when retinal neovascularization in a diabetic patient was found to regress after postpartum pituitary necrosis (12). Subsequent controlled clinical trials showed that pituitary ablation could improve DR (13). Many mitogenic effects of GH are mediated by IGF-1 (14). Merimee et al. (15) found increased serum IGF-1 levels in patients with rapidly accelerating DR, and Chantelau et al. (16,17) reported a relationship between upregulation of IGF-1 and progression of retinopathy in Mauriac’s syndrome and in type 1 diabetes.

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

From the Department of 1Ophthalmology, University Hospital, Malmo¨, Sweden; the 2Department of Endocrinology, University Hospital, Malmo¨, Sweden; and the 3Department of Clinical Chemistry, University Hospital, Malmo¨, Sweden. Address correspondence and reprint requests to Marianne Henricsson, Department of Ophthalmology, S-25187 Helsingborg, Sweden. E-mail: [email protected]. Received for publication 13 June 2001 and accepted in revised form 16 October 2001. Abbreviations: DR, diabetic retinopathy; GH, growth hormone; PDR, proliferative diabetic retinopathy. A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion factors for many substances.

DIABETES CARE, VOLUME 25, NUMBER 2, FEBRUARY 2002

Patients A total of 45 consecutive, orally treated type 2 diabetic patients ⬍75 years of age were included in the study when they, due to unsatisfactory metabolic control, were referred to the Diabetic Day Care Unit at the Department of Endocrinology in Malmo¨, Sweden, to be changed to insulin therapy (Table 1). Patients with severe concomitant diseases preventing follow-up, patients with severe nonproliferative retinopathy or proliferative retinopathy, and patients with cataracts, making retinal photography impossible, were excluded from the study. The patients were examined at baseline, before the introduction of insulin therapy, and at follow-up after 1, 3, 6, 12, 24, and 36 months. Fundus photography including visual acuity testing was performed, and blood samples for HbA1c and IGF-1 were collected during all examinations. At baseline, a simple neurological examination was performed; tendon reflexes and vibratory sensory thresholds were assessed using a biothesiometer at the ankle region (Bio-thesiometer; BioMedical Instrument, Newbury, OH). Patients were asked about smoking and current antihypertensive treatment. Blood pressure was measured in each patient in the supine position after 5 min rest. Hypertension was considered 381

Progression of retinopathy Table 1—Clinical characteristics at entry of patients with progression of retinopathy 3 levels

n Age at first eye examination Sex Men Women Duration of diabetes at entry (years) Weight (kg) BMI (kg/m2) C-peptide Smoking Hypertension Degree of retinopathy at entry None (level 10/10) Mild nonproliferative DR (level 21, 10–31, 31) Moderate nonproliferative DR (level ⬍41, 41–51, ⬍51) Macular edema or treatment Visual acuity, best eye (Monoyer, median, range) Albuminuria at entry (ⱖ20 mikrog min⫺1) Peripheral pulses absent Achilles reflexes absent Patellar reflexes absent Vibration sense (worse leg, median range)

Progression ⱕ2 levels

Progression ⱖ3 levels

P

31 (74) 61.7 ⫾ 8.0

11 (26) 63.1 ⫾ 5.6

0.5

14 (45) 17 (55) 8.8 ⫾ 5.2 80.5 (13.3) 27.6 (3.9) 0.78 (0.43) 5 (16) 21 (68)

8 (73) 3 (27) 11.9 ⫾ 5.3 83.8 (16.0) 25.3 (4.1) 0.72 (0.42) 2 (18) 5 (45)

17 (55) 12 (39)

4 (36)* 4 (36)

2 (6)

3 (27)

0.1 0.54 0.11 0.72 0.9 0.3 0.2

2 (6) 1.0 (0.5–1.0)

3 (27) 1.0 (0.5–1.0)

0.1 0.8

10 (32) 4 (13) 9 (29) 3 (10) 20 (10–50)

4 (36) 2 (18) 6 (55) 2 (18) 32 (11–35)

0.9 0.8 0.2 0.7 0.11

Data are n (%) or mean ⫾ SD, unless otherwise indicated. * Percentages do not add up to 100 because of rounding.

present if the patient was taking antihypertensive drugs or if blood pressure was ⬎160/95 mmHg. Urinary albumin excretion rate was measured at baseline and after 12, 24, and 36 months. The study was approved by the Ethics Committee at the University of Lund, Sweden. Informed consent was obtained from all participating patients. Analytical methods HbA1c was measured by ion exchange chromatography (19); the reference interval was ⬍5.3%. Urinary albumin was measured by rate immunonephelometry (Beckman ARRAY instrument; Beckman Instruments, Fullerton, CA). Microalbuminuria was defined as an albumin excretion of 20 –200 ␮g/min. Fasting plasma C-peptide was measured by radioimmunoassay (20). IGF-1 was measured by an immunometric assay kit (IGF-1 IRMA; Nichols Institute Diagnostics, San Juan Capistrano, CA) according to the instructions of the manufacturer. The intra- and 382

interassay coefficients of variation were 5 and 15%, respectively, at the level 60 ␮g/l. The detection limit was 6 ␮g/l. Fundus photography and visual acuity Stereo color fundus photographs of the seven standard fields were taken during pharmacological mydriasis, at an angle of 30°, with a Topcon TRC-50 VT fundus camera (Tokyo Optical, Tokyo). The alternative classification of the Wisconsin Epidemiologic Study of Diabetic Retinopathy was used to define the retinopathy level (21,22). This classification provides an overall retinopathy scale: level 10 represents no retinopathy, levels 21–51 represent nonproliferative diabetic retinopathy of increasing severity, and levels 60⫹ represent all forms of PDR with and without laser treatment. The patient’s retinopathy level was derived by giving the eye with the higher level a greater weight. This scheme provides an 11-step scale: 10, 21/10, 21/21, 31/⬍31,

31/31, 41/⬍41, 41/41, 51/⬍51, 51/51, 60⫹/⬍60⫹, 60⫹/60⫹ (22). Thereafter, the retinopathy levels were divided into four groups; no diabetic retinopathy (level 10), mild retinopathy (levels 21/10 – 31/31), moderate nonproliferative retinopathy (levels 41/⬍41–51/⬍51), and severe nonproliferative or proliferative retinopathy (levels 51/51– 60⫹/ 60⫹). Macular edema was defined as the presence of hard exudates and/or retinal thickening within one disc diameter of the center of the macula in at least one eye. Macular treatment was defined as focal laser treatment in at least one eye. Proliferative retinopathy and macular edema were treated according to established guidelines. Photographs were viewed against light boards using Donaldson’s stereo viewer (magnification ⫻5; G. J. Davco, Holbrook, MA) and assessed by an independent grader with extensive experience with this classification. The grader had no knowledge of the patients’ clinical data. No patients were excluded from the study because of media opacities. For the testing of visual acuity, the Monoyer charts (0.1–1.0 decimal scale) were used. The visual acuity of the best eye was reported. Main outcome measures The group with progression ⱖ3 levels was compared with the group with progression ⱕ2 levels. A progression of ⱖ3 levels was considered a clinically significant deterioration. Change in severity of retinopathy was assessed for the association with HbA1c and IGF-1. Statistical methods The statistical analyses were performed with the statistical software SPSS for Windows (SPSS, Chicago, IL). Friedman and Wilcoxon signed-rank tests were used to analyze changes between repeated measurements. The area under the curve was estimated for each patient and variable that was measured serially (23). When analyzing differences between variables in two independent groups, Student’s t test, Mann-Whitney U test, ␹2 test, and Fisher’s exact test were used. Multiple regression analysis was used to assess the influence of HbA1c, IGF-1, and retinopa-


Henricsson and Associates

Table 2—Distribution of retinopathy in patients at baseline and follow-up after 3 years At baseline

n At follow-up (3 years) No DR (level 10) Mild nonproliferative DR (levels 21, 10–31, 31) Moderate nonproliferative DR (levels 41, ⬍41–51, ⬍51) Severe nonproliferative DR or PDR (levels 51, 51–60⫹, 60⫹)

No DR

Mild nonproliferative DR

Moderate nonproliferative DR

21

16

5

11 10

1 9

0 0

—

5

1

—

1

4

with most severe progression (0.6 [0.5– 1.0]) during the 3 years of observation (P ⫽ 0.02). The insulin doses per kilogram body weight were equal in both groups (measured at 24 months). The frequency of albuminuria (ⱖ20 ␮g/min) was higher at 36 months than at baseline (P ⫽ 0.04), but there was no difference between the groups (Table 3). Four patients underwent cataract surgery during follow-up (no differences between groups), and one patient with progression ⱖ3 levels underwent vitrectomy of both eyes because of recurrent vitreous hemorrhages.

Data are n.

thy at baseline on progression ⱖ3 levels and was performed with all variables entered in a single step. Significance was considered for P ⱕ 0.05. Results were given with 95% CI; data are presented as means ⫾ SD. RESULTS — Three patients died during the 3-year follow-up; 42 patients completed the study and remained on insulin therapy after 3 years. At baseline, 21 of 42 patients (50%) showed no retinopathy, whereas among the remaining 21 patients, 16 had mild retinopathy and 5 had moderate retinopathy (Tables 1 and 2). After 3 years, the level of retinopathy had progressed in 20 of 42 patients (48%). The progression affected all levels of retinopathy, i.e., from no retinopathy to moderate nonproliferative diabetic retinopathy (Table 2). At 36 months, three patients had progressed to PDR and eight patients (five at inclusion) had been treated for macular edema. When comparing baseline values with average values during follow-up in the whole patient group (area under the curve divided by 36 months), HbA1c and IGF-1 were significantly reduced (Fig. 1) (P ⬍ 0.01). The decrease in HbA1c levels occurred rapidly (within 3 months) after the initiation of insulin therapy, whereas IGF-1 levels decreased gradually (maximal after 36 months). The level of IGF-1 was not related to age. The group of patients with progression ⱖ3 levels (n ⫽ 11) and the group of patients with progression ⱕ2 levels (n ⫽ 31) had significant reduction in HbA1c values, with no significant difference between the groups, whereas the IGF-1 levels were signifi-

cantly reduced only in the group with less progression (P ⬍ 0.01) (Table 3). In agreement, a higher average IGF-1 (per 10 units) at 36 months was significantly associated with retinopathy progression ⱖ3 levels in a logistic regression analysis (Table 4). When average HbA 1c and IGF-1 were entered into the analysis, this association disappeared. There were no relationships between the degree of retinopathy or other clinical features at entry and the progression by ⱖ3 levels (Table 1). Macular edema developed in two patients in the group with little or no progression and in one patient in the group with progression ⱖ3 levels (NS). Visual acuity of the best eye was similar in both groups at baseline (1.0 [95% CI 0.5–1.0]) but deteriorated significantly in the group

CONCLUSIONS — Several of the patients in our study had progression of retinopathy. Of the 42 patients with mild to moderate retinopathy, four patients required panretinal laser treatment and one patient underwent vitrectomy. In the previous 2-year follow-up of the current patient group, we found a relationship between decreased HbA1c level and retinopathy progression (9). Such a relationship was not seen after 3 years. A lack of decrement in IGF-1 3 years after the start of insulin therapy was associated with progression of retinopathy. Therefore, it seems that an initial improvement in glycemic control may have favored the early progression of retinopathy, and that factors associated with high IGF-1 concentrations may be related to later-occurring progression. The higher IGF-1 value may

Table 3—HbA1c and IGF-1 at entry and follow-up

n HbA1c at entry (%) HbA1c at 12 months (%) HbA1c at 24 months (%) HbA1c at 36 months (%) Average HbA1c* IGF-1 at entry (␮g/l) IGF-1 at 12 months (␮g/l) IGF-1 at 24 months (␮g/l) IGF-1 at 36 months (␮g/l) Average IGF-1* Albuminuria at entry (ⱕ20 mikrog min⫺1) Albuminuria at 24 months Albuminuria at 36 months Insulin doses per kilogram weight (at 24 months)

Progression ⱕ2 levels

Progression ⱖ3 levels

P

31 (81) 9.8 ⫾ 1.3 7.9 ⫾ 1.0 8.0 ⫾ 1.1 8.0 ⫾ 0.9 7.9 ⫾ 0.8 139 ⫾ 49 108 ⫾ 31 99 ⫾ 34 106 ⫾ 29 117 ⫾ 35 10/31 (32) 11/29 (38) 11/28 (39) 0.49 ⫾ 0.15

11 (19) 10.0 ⫾ 1.0 7.4 ⫾ 1.5 7.7 ⫾ 1.7 7.8 ⫾ 1.4 7.4 ⫾ 1.2 167 ⫾ 59 133 ⫾ 64 116 ⫾ 52 128 ⫾ 47 124 ⫾ 47 4/11 (36) 4/10 (40) 2/6 (33) 0.44 ⫾ 0.15

0.65 0.21 0.60 0.73 0.59 0.13 0.35 0.26 0.003 0.07 0.9 0.70 0.36 0.36

Data are means ⫾ SD or n (%). *Area under the curve during the study period (36 months).


383

Progression of retinopathy

Table 4—Logistic regression analysis of relating independent variables of progression >3 levels Relative risk Age (years) Duration of diabetes (years) Mild moderate nonproliferative DR at entry/no DR HbA1c at entry (per %) HbA1c at 36 months (per %) IGF-1 at entry (per 10 units) IGF-1 at 36 months (per 10 units)

Figure 1—HbA1c (A) and IGF-1 (B) in the two progression groups: ⱖ3 levels and ⱕ2 levels (mean, 95% CI). f, ⱖ3 levels, 䡺, ⱕ2 levels. HbA1c improved within 3 months after initiation of insulin therapy in both groups; IGF-1 decreased gradually and significantly (P ⬍ 0.01) only in the group of patients without progression (ⱕ2 levels) of retinopathy. * P ⫽ 0.003 for differences in IGF-1 between the two groups 36 months after initiation of insulin therapy.

be a marker of the improved glycemic control. In keeping with our observation of decrements in IGF-1 levels during insulin therapy, IGF-1 levels are lower in insulin-treated type 2 diabetic patients than in those on other therapies (24). A 384

decrease in IGF-1 in type 2 diabetic patients may depend on reduced endogenous insulin production secondary to systemic insulin injections (14). This hypothesis could not be tested in our study; endogenous insulin secretion as C-

95% CI

P

1.0 1.05

0.85–1.17 0.99 0.63–1.26 0.61

3.3

0.32–34.9 0.32

1.1

0.4–2.9

0.81

1.9

0.90–4.17 0.09

0.9

0.74–1.18 0.58

1.6

1.09–2.31 0.02

peptide was only measured before the start of insulin therapy. There is experimental evidence of an influence of GH and IGF-1 on progression of retinopathy. Retinal microvascular cells have receptors for IGF-1, and IGF-1 induces the release of plasminogen activator and increases DNA synthesis in retinal endothelium (25). Inhibitory action of the GH suppressor ocreotide on human retinal endothelial cells’ proliferation and DNA synthesis has been shown in vitro (26). According to Smith et al. (27), an IGF-1 receptor antagonist suppresses retinal neovasularization. An interaction between IGF-1 and the IGF-1 receptor is necessary for induction of neovascularization by vascular endothelial growth factor (28). In conclusion, in type 2 diabetic patients treated with oral agents, within a few months after initiation of insulin therapy, glycemic control, as reflected by HbA1c, improved and persisted for at least 3 years. There was, however, a tendency toward increased glycemia after 2–3 years. In contrast to HbA1c, IGF-1 was continuously reduced; the lowest values were recorded 3 years after initiation of insulin therapy. We found a relationship between progression of retinopathy and a higher IGF-1 value at 3 years. Our study is small but raises questions about the relationship between retinopathy, hyperglycemia, insulin therapy, and IGF-1. Acknowledgments — This study was supported by the Gorthon and Zoe´ ga Foundations


Henricsson and Associates

(Helsingborg), the Ja¨ rnhardt Foundation, the Swedish Diabetes Association, the Lundstro¨ m Foundation, the Novo-Nordic Foundation, and research funds of Malmo¨ University Hospital, the County of Malmo¨ hus, and the Swedish Medical Research Council (Grant 7507). We thank Jonas Ranstam, PhD, for statistical supervision; Eva Steffert for grading of photographs; and Janet Parmvi for valuable assistance. References 1. Kohner EM: Diabetic retinopathy. BMJ 307:1195–1993, 1993 2. The Diabetes Control and Complications Trial Research Group: The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 329:977–986, 1993 3. Reichard P, Nilsson B-Y, Rosenqvist U: The effect of long term intensified insulin treatment on the development of microvascular complications of diabetes mellitus. N Engl J Med 329:304 –309, 1993 4. U.K. Prospective Diabetes Study (UKPDS) Group: Intensive blood-glucose control with sulfonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 352:837– 853, 1998 5. Dahl-Jorgensen K, Brinchmann-Hansen O, Hanssen KF, Sandvik L, Aagenaes O, Aker Diabetes Group: Rapid tightening of blood glucose control leads to transient deterioration of retinopathy in insulin-dependent diabetes mellitus: the Oslo study. BMJ 290:811– 815, 1985 6. Lauritzen T, Frost-Larsen K, Larsen HW, Deckert T: The Steno Study Group: twoyear experience with continuous subcutaneous insulin infusion in relation to retinopathy and neuropathy. Diabetes 34 (Suppl. 3):74 –79, 1985 7. Kroc Collaborative Study Group: Diabetic retinopathy after two years of intensified insulin treatment. JAMA 260:37– 41, 1988 8. The Diabetes Control and Complications Trial Research Group: Early worsening of diabetic retinopathy in the Diabetes Con-

9.

10.

11.

12.

13. 14. 15.

16.

17.

18.

19.


trol and Complications Trial. Arch Ophthalmol 116:874 – 886, 1998 Henricsson M, Berntorp K, Berntorp E, Fernlund P, Sundkvist G: Progression of retinopathy after improved metabolic control in type II diabetic patients: relation to IGF-1 and hemostatic variables. Diabetes Care 22:1944 –1949, 1999 Clauson P, Linnarsson R, Gottsa¨ ter A, Sundkvist G, Grill V: Relationships between diabetes duration, metabolic control and ␤-cell function in a representative population of type 2 diabetic patients. Diabet Med 11:794 – 801, 1994 Sundkvist G, Lilja B, Alme´ r LO: Absent elevations in growth hormone, factor VIII related antigen, and plasminogen activator activity during exercise in diabetic patients resistant to retinopathy. Diabetes Res 7:25–30, 1988 Poulsen J: The Houssay phenomen in man: recovery from retinopathy in a case of diabetes with Simmond’s disease. Diabetes 2:7–12, 1953 Lundbaek K: Growth hormone’s role in diabetic microangiopathy. Diabetes 25 (Suppl. 2):845– 849, 1976 LeRoith D: Insulin-like growth factors. N Engl J Med 336:633– 640, 1997 Merimee TJ, Zapf J, Froesch ER: Insulinlike growth factors: studies in diabetics with and without retinopathy. N Engl J Med 309:527–530, 1983 Chantelau E, Seppel T, Althaus C, Scho¨ nau E: Elevation of serum IGF-1 rather than sex steroids precedes proliferative retinopathy in Mauriac’s syndrome. Horm Res 48:131–134, 1997 Chantelau E: Evidence that upregulation of serum IGF-1 concentration can trigger acceleration of diabetic retinopathy. Br J Ophthalmol 82:725–730, 1998 Grant MB, Mames RN, Fitzgerald C, Hazariwala KM, Cooper-DeHoff R, Caballero S, Estes KS: The efficacy of octreotide in the therapy of severe nonproliferative and early proliferative diabetic retinopathy. Diabetes Care 23:504 –509, 2000 Jeppsson JO, Jerntorp P, Sundkvist G, Englund H, Nylund V: Measurement of hemoglobin A1c by a new liquid-chromatographic assay: methodology, clinical utility, and relation to glucose tolerance evaluated. Clin Chem 32:1867–1872, 1986

20. Gottsa¨ ter A, Landin-Olsson M, Fernlund P, Gullberg B, Lernmark Å, Sundkvist G: Pancreatic beta-cell function evaluated by intravenous glucose and glucagon stimulation: a comparison between insulin and C-peptide to measure insulin secretion. Scand J Clin Lab Invest 52:631– 639, 1992 21. Klein R, Klein BEK, Magli YL, Brothers RJ, Meuer SM, Davis MD: An alternative method of grading diabetic retinopathy. Ophthalmology 93:1183–1187, 1986 22. Klein R, Klein BEK, Moss SE, Davis MD, DeMets DL: The Wisconsin Epidemiologic Study of Diabetic Retinopathy IX: four-year incidence and progression of diabetic retinopathy when age at diagnosis is less than 30 years. Arch Ophthalmol 107: 237–243, 1989 23. Matthews JNS, Altman DG, Campbell MJ, Royston P: Analysis of serial measurements in medical research. BMJ 300:230 – 235, 1990 24. Clauson PG, Brismar K, Hall K, Linnarsson, Grill V: Insulin-like growth factor-1 and insulin-like growth factor binding protein-1 in a representative population of type 2 diabetic patients in Sweden. Scand J Clin Lab Invest 58:353–360, 1998 25. King GL, Goodman AD, Buzney S, Moses A, Kahn CR: Receptors and growth promoting effects of insulin and IGF-1 on cells from bovine retinal capillaries and aorta. J Clin Invest 75:1028 –1036, 1985 26. Grant MB, Caballero S, Millard WJ: Inhibition of IGF-1 and b-FGF stimulated growth of human retinal endothelial cells by the somastotatin anlaogue, ocreotide: a potential treatment for ocular neovascularization. Regul Pept 48: 267–278, 1993 27. Smith LE, Kopchick JJ, Chen W, Knapp J, Kinose F, Daley D, Foley E, Smith RG, Schaeffer JM: Essential role of growth hormone in ischemia-induced retinal neovascularization. Science 276:1076 –1709, 1997 28. Smith LEH, Shen W, Perruzzi C, Soker S, Kinose F, Xu X, Robinson G, Driver S, Bishoff J, Zhang B, Schaeffer JM, Senger DR: Regulation of vascular endothelial growth factor-dependent retinal neo-vascularization by insulin-like growth factor-1 receptor. Nature Med 5:1390 –1395, 1999

385