David
M. Yousem,
Proliferative of White
MD
#{149} William
Diabetes
terms:
Brain,
mellitus,
complications, Radiology
1991;
MR studies, 10.1214. #{149} Diabetes mellitus, 2245.62 #{149} Retina, 2245.62 9*2.724
179:229-230
MD
#{149} Robert
Retinopathy: Matter Lesions
One form of retinopathy associated with diabetes is a proliferative small vessel process thought to be mediated by biochemical, hemodynamic, and endocrinologic factors. The authors conducted a prospective study to determine whether patients with diabetes who had proliferative retinopathy had evidence of intracranial microangiopathy visible at magnetic resonance (MR) imaging. Twenty-five patients under 40 years of age with proliferative retinopathy and insulin-dependent diabetes mellitus and 10 agematched control subjects were studied with MR imaging. Axial images were reviewed by two neuroradiologists for the presence of white matter foci of high signal intensity. No patients demonstrated evidence of these foci. There was no evidence of ischemic foci in any of the patients (all patients were neurologically asymptomatic). The vasculopathy associated with proliferative retinopathy does not appear to affect the intracranial circulation to the extent detectable with MR imaging. The presence of white matter foci of high signal intensity or ischemic changes in the brains of insulin-dependent diabetic patients under 40 years of age should not be attributed to diabetic vasculopathy. Other causes should be considered. Index
S. Tasman,
D
I. Grossman,
Absence at MR Imaging’
mcllitus in its many forms occurs in 1%-2% of people in the Western world. One of its most devastating complications is blindness, caused by a proliferative vasculopathy of the retina (1). Diabetes accounts for 10% and 20% of blindness in middleaged men and women, respectively; 2% of all patients with diabetes in this age group arc legally blind. In the United States, 5,000 people a year become blind due to diabetic retinopathy (2). IABETES
Diabetic
retinopathy
begins
in a non-
proliferative (“background”) form in which microancurysms, retinal hcmommhagcs, and exudates predominate. Prolifcrativc
retinopathy
occurs
after
sev-
cral years of glucose intolerance and consists of neovascularization of the disk and retina, which may lead to vitmcous hemorrhage and retinal detachment (1-3). Magnetic
resonance
(MR)
imaging
has been shown to be very sensitive to small vessel angiopathy in such discases as systemic lupus erythematosus, Sj#{246}gmcn syndrome, and moyamoya disease. Such vasculitides demonstrate small areas of high signal intensity in the white matter on bong repetition time (TR)/echo time (TE) images. Similam findings in the elderly population arc often attributed to small vessel antemiosclcrotic
ischcmic
foci
(4-7).
We
sought to determine whether patients with proliferative diabetic retinopathy who were under 40 years of age, an age range in which arteriosclerotic white matter infarcts are generally not encountered, had more white matter foci of high signal intensity than did agematched nondiabetic volunteers. Because of the greaten frequency of hypertension and age-rebated artemiosclerotic changes in patients older than 40 years, only those patients 40 years old
I From the Neuroradiology Section, Department of Radiology, Pennsylvania, 3400 Spruce St. Philadelphia, PA 19104 (D.M.Y., the Wills Eye Hospital, Philadelphia (W.S.T.). Received October vemben 21; revision received December 10; accepted December D.M.Y. 2 9* indicates generalized vein and artery involvement. RSNA, 1991
MD
Hospital of the University of RIG.), and the Retina Service of 5, 1990; revision requested No12. Address reprint requests to
or younger
were
included
in the study
group.
SUBJECTS
AND
Twenty-five
proliferative
diabetic
nondiabetic,
age-matched
derwent
METHODS
nonpregnant
MR
patients
with
retinopathy
and
volunteers
10
un-
imaging.
In the diabetic group, there were eight men and 17 women aged 18-39 years (mean, 31 years). The patients had required insulin (and had had diabetes) for 6-28 years (mean, 19.3 years; standard deviation, 5.6 years; median, 20 years). The mean total daily insulin dose that the patients were receiving at the time of their MR imaging examination was 60.5 U (obtamed for 24 of the 25 patients; standard deviation,
35.5
U).
Neovascularization
of
the retina was present in all patients. All but one patient had undergone photocoagulation (laser therapy) on the eyes. Five patients were hypertensive and were meceiving medications to reduce blood pressure. The mean number of years these patients had been hypertensive was 5.4 years (range, 2-10 years). Serum creatinine bevels were obtained from the mcdical records of 1 1 patients. This level was greater than 1.5 mg/dL (132.6 Mmol/L) in only one patient. This patient had a creatmine level of 20.0 mg/dL (1,768 zmol/L) and was undergoing dialysis at the time of MR examination. Among the healthy volunteers, there were four men and six women, aged 2236 years (mean, 32 years). None were hypcrtensive.
Informed consent was obtained from every subject. The study had been approved by the investigational review boards at both the Hospital of the University
of Pennsylvania
and
the
Wills
Eye
Hospital. Imaging was performed with a 1.5-T Signa unit (GE Medical Systems, Mibwaukee). A localizing Ti-weighted spin-echo sequence (TR msec/TE msec 600/20) was initially used, followed by a long TR/ TE spin-echo sequence (3,000/30-35, 90) with flow compensation techniques (first order gradient moment nulling), one sig.
Abbreviations: tion
TE
echo
time,
TR
=
repeti-
time.
229
nal averaged, inferior spatial presaturation pulses, a 192 X 256 matrix, and 5mm-thick sections with a 2.5-mm section gap. The long TRITE images were reviewed by two neuroradiologists-onc familiar with and one unaware of the goal of the study. The prevalence of foci of high signal intensity in the periventricular white matter, centrum semiovale, and basal ganglia was determined from the long TR/TE images. RESULTS No
diabetic
patients
demonstrated
foci of high signal intensity in the white matter or deep gray matter. The images were read as normal by both the blinded radiologist and the ncunoradiologist dedicated to the study, who was specifically looking for deep and penventricular white matter changes. No control subject had white matter foci of high signal intensity. Using the Manteb-Hacnszeb test for the odds ratio (8), at beast eight of the diabetic patients would have had to have shown white matter foci of high signal intensity to produce a statisticalby significant (P < .05) difference between the control group and the patient group. DISCUSSION In the elderly population, the prcsence of white matter foci of high signal intensity
on
long
TRITE
MR
images
is
not uncommon. White matter foci of high signal intensity due to disease must be distinguished from a normal variant, the dilated perivascular (Vimchow-Robin) spaces, which are isointense with ccrebnospinal fluid on the proton-density
(3,000/30-35)
images.
By contrast, white matter foci of high signal intensity due to disease arc usually hypenntcnse to ccmebmospina! fluid on the proton-density images. These besions have been attributed to small yessd ischemic foci with gliosis (4-7,9). At pathologic examination, the arterioles show hyaline degeneration and microathemomata with surrounding atrophic pemivascular demyelination and/or myelin pallor presumably because of chronic ischemia on a hypoperfusion basis (7,10). These abnormal foci in the white matter have been meported to be present in as many as 30%44% of all patients 60 years old or older on long TRITE images. This rate is even higher (56%-78%) in patients with risk factors for atherosclerotic disease (6,11). In the population under the age of 50 years, diseases that produce white matten foci of high intensity include multiplc sclerosis, sarcoidosis, Lyme disease, vasculitis, and acute disseminated encephalomyclitis. cnibe
230
this
The
sign
#{149} Radiology
tendency
to diabetic
to as-
vascubopathy
in younger patients who have a long history of insulin-dependent diabetes mellitus derives from the knowledge that diabetes is a risk factor for stroke in the elderly (4-6,11). Our study suggests that this assumption is incorrect. One must search for another cause of the white matter lesions in the patient population 40 years old and younger. The proliferative vasculopathy that characterizes
diabetic
retinopathy
is
within the vascular distribution of the central retinal artery, a branch of the ophthalmic artery from the internal carotid system. Vascular disease identified in the intracranial circulation of the internal carotid artery would suggest an intrinsic vascular or systemic abnormality.
Alternatively,
one
could
hypothesize that the end organ, the retma, is the primary cause of or combines with the feeding vessel in the pathogenesis of diabetic retinopathy. Diabetic retinopathy includes lesions such as microaneurysms, hemorrhages, and infarcts. According to Memimce (2), the first insult occurs at the capillary penicytes, which control perfusion to the retina. Neovascularization, however, is the hallmark of proliferative retinopathy, and this occurs at the capillary and artcmiolar bevel. In the probifcrative stages, there arc capillary occlusions, neovasculanization, fibmovascular proliferation, and fibrous contraction leading to traction retinal detachments. Since vascular occlusions, infarcts, hemorrhages, pemivascular fibrosis/ gliosis, and Charcot-Bouchard ancurysms occur intracranially in patients with diabetes, the theory that the vasculopathy is a systemic disorder is rcasonablc.
There have been numerous theories as to the cause of diabetic retinopathy. In a recent review article by Mcmimec (2), three theories of the pathogcncsis of the disorder were described. The biochemical postulate emphasizes the preeminent mole of hyperglycemia in causing basement membrane damage and, hence, cndotheliab cell incornpetencc with production of prostagbandins that promote vascular thrombosis or hemorrhage. Glycosylated hemogbobin (known to be increased in patients with severe diabetic retinopathy) can induce other thmombogenic chemicals or can thicken the cndothelial basement membrane, activating the cornplernent system or causing local bloodflow variations (12). The second theory stresses the role of increased blood viscosity and platelet agglutination in diabetes, which meduces flow in capillaries and causes thrombosis. Proliferative neovascularization follows as an attempt to revascularize thrornbosed areas of the retina (2). The third component of the pathogenesis of diabetic retinopathy cmpha-
sizes
elevated
growth
hormone
and
insulin-like growth factor I levels, which have been shown to be elevated in diabetic patients with retinopathy. Growth hormone can induce thromboses, and injured, anoxic retinal cells may produce the insulin-like growth factor I, which can induce vascular neogenesis (2,13). The presence of white matter foci of high signal intensity in the brain of insulin-dependcnt diabetic patients would have suggested an intrinsic vascuban abnormality or would have supported the biochemical or hemodynamic theories of retinopathy pathogenesis. These theories hypothesize a more systemic basis of the disease. As regional cerebral blood flow decreases with age, the effect of diabetic small vessel disease would presumably have a greater role in inducing small vessel ischemic change in the white matter. The abscnce of these foci when there have been significant proliferative retinal changes, as seen in this study, testifies to the primary role of the substrate (ie, the retina) in the formation of proliferative diabetic retinopathy. The changes do not appear to be present elsewhere in the carotid system at MR imaging, a modality that is very sensitive to small vessel disease in the brain. U Acknowledgment: have been possible tance
provided
This
project
without
by Laura
could
not
the enormous
assis-
Bourke.
References 1.
Lynn athy.
2.
Menimee
3.
4.
5.
6.
7. 8. 9.
JR. Snyder WB. Vaiser A. Diabetic New York: Grune & Stratton, 1974. TJ.
Diabetic
retinopathy:
19:263-265. Drayen BP. mal findings.
Imaging of the aging brain. I. NorRadiology 1988; 166:785-796.
Woolson RF. Statistical methods for the analysis of biomedical data. New York: Wiley. 1987; 418422. Bradley WG. Waluch V. Brandt-Zawadzki M. YadWycoff
RR.
Patchy
matter lesions in the elderly: tion during NMR imaging. aging
11. 12.
13.
of
perspectives. N EngI J Med 1990; 322:978-983. Jerneld B. Algvere P. Relationship of duration and onset of diabetes to prevalence of diabetic retinopathy. Am J Ophthal 1986; 15:431-437. Brandt-Zawadzki M, Fein C. Van Dyke C. Kiernan R, Davenport L. deGroot J. MR imaging of the aging brain: patchy white-matter lesions and dementia. AJNR 1985; 6:675-682. Awad IA, Spetzler RF, Hodak JA, Awad CA. Carey R. Incidental subcontical lesions identified on magnetic resonance imaging in the elderly. I. Correlation with age and cerebrovascular risk factons. Stroke 1986; 17:1084-1089. Lechner K, Schmidt R, Bertha G, Justick E, Offanbacher H, Schneider C. Nuclear magnetic resonance image white matter lesions and risk factors for stroke in normal individuals. Stroke 1988;
icy RA.
10.
retinop-
a synthesis
1984;
periventnicular white common obsenvaNoninvasive Med Im-
1:35-41.
Kirkpatrick JB, Heyman LA. White-matter Icsions in MR imaging of clinically healthy brains of elderly subjects: possible pathologic basis. Radioiogy Gerard
1987;
162:509-511.
C, Weisberg L. MRI periventnicular Icsions in adults. Neurology 1986; 36:998-1001. Klein R, Klein BEK, Moss SE. et al. Glycosylated hemoglobin predicts the incidence and prognession of diabetic retinopathy. JAMA 1988; 260:2864-2871. Merimee TJ, Zapf
J, Foresch ER. Insulin-like studies in diabetics with and without retinopathy. N Engl J Med 1983; 309:527-530. growth
factors:
April
1991
M. Yousem,
Proliferative of White
MD
#{149} William
Diabetes
terms:
Brain,
mellitus,
complications, Radiology
1991;
MR studies, 10.1214. #{149} Diabetes mellitus, 2245.62 #{149} Retina, 2245.62 9*2.724
179:229-230
MD
#{149} Robert
Retinopathy: Matter Lesions
One form of retinopathy associated with diabetes is a proliferative small vessel process thought to be mediated by biochemical, hemodynamic, and endocrinologic factors. The authors conducted a prospective study to determine whether patients with diabetes who had proliferative retinopathy had evidence of intracranial microangiopathy visible at magnetic resonance (MR) imaging. Twenty-five patients under 40 years of age with proliferative retinopathy and insulin-dependent diabetes mellitus and 10 agematched control subjects were studied with MR imaging. Axial images were reviewed by two neuroradiologists for the presence of white matter foci of high signal intensity. No patients demonstrated evidence of these foci. There was no evidence of ischemic foci in any of the patients (all patients were neurologically asymptomatic). The vasculopathy associated with proliferative retinopathy does not appear to affect the intracranial circulation to the extent detectable with MR imaging. The presence of white matter foci of high signal intensity or ischemic changes in the brains of insulin-dependent diabetic patients under 40 years of age should not be attributed to diabetic vasculopathy. Other causes should be considered. Index
S. Tasman,
D
I. Grossman,
Absence at MR Imaging’
mcllitus in its many forms occurs in 1%-2% of people in the Western world. One of its most devastating complications is blindness, caused by a proliferative vasculopathy of the retina (1). Diabetes accounts for 10% and 20% of blindness in middleaged men and women, respectively; 2% of all patients with diabetes in this age group arc legally blind. In the United States, 5,000 people a year become blind due to diabetic retinopathy (2). IABETES
Diabetic
retinopathy
begins
in a non-
proliferative (“background”) form in which microancurysms, retinal hcmommhagcs, and exudates predominate. Prolifcrativc
retinopathy
occurs
after
sev-
cral years of glucose intolerance and consists of neovascularization of the disk and retina, which may lead to vitmcous hemorrhage and retinal detachment (1-3). Magnetic
resonance
(MR)
imaging
has been shown to be very sensitive to small vessel angiopathy in such discases as systemic lupus erythematosus, Sj#{246}gmcn syndrome, and moyamoya disease. Such vasculitides demonstrate small areas of high signal intensity in the white matter on bong repetition time (TR)/echo time (TE) images. Similam findings in the elderly population arc often attributed to small vessel antemiosclcrotic
ischcmic
foci
(4-7).
We
sought to determine whether patients with proliferative diabetic retinopathy who were under 40 years of age, an age range in which arteriosclerotic white matter infarcts are generally not encountered, had more white matter foci of high signal intensity than did agematched nondiabetic volunteers. Because of the greaten frequency of hypertension and age-rebated artemiosclerotic changes in patients older than 40 years, only those patients 40 years old
I From the Neuroradiology Section, Department of Radiology, Pennsylvania, 3400 Spruce St. Philadelphia, PA 19104 (D.M.Y., the Wills Eye Hospital, Philadelphia (W.S.T.). Received October vemben 21; revision received December 10; accepted December D.M.Y. 2 9* indicates generalized vein and artery involvement. RSNA, 1991
MD
Hospital of the University of RIG.), and the Retina Service of 5, 1990; revision requested No12. Address reprint requests to
or younger
were
included
in the study
group.
SUBJECTS
AND
Twenty-five
proliferative
diabetic
nondiabetic,
age-matched
derwent
METHODS
nonpregnant
MR
patients
with
retinopathy
and
volunteers
10
un-
imaging.
In the diabetic group, there were eight men and 17 women aged 18-39 years (mean, 31 years). The patients had required insulin (and had had diabetes) for 6-28 years (mean, 19.3 years; standard deviation, 5.6 years; median, 20 years). The mean total daily insulin dose that the patients were receiving at the time of their MR imaging examination was 60.5 U (obtamed for 24 of the 25 patients; standard deviation,
35.5
U).
Neovascularization
of
the retina was present in all patients. All but one patient had undergone photocoagulation (laser therapy) on the eyes. Five patients were hypertensive and were meceiving medications to reduce blood pressure. The mean number of years these patients had been hypertensive was 5.4 years (range, 2-10 years). Serum creatinine bevels were obtained from the mcdical records of 1 1 patients. This level was greater than 1.5 mg/dL (132.6 Mmol/L) in only one patient. This patient had a creatmine level of 20.0 mg/dL (1,768 zmol/L) and was undergoing dialysis at the time of MR examination. Among the healthy volunteers, there were four men and six women, aged 2236 years (mean, 32 years). None were hypcrtensive.
Informed consent was obtained from every subject. The study had been approved by the investigational review boards at both the Hospital of the University
of Pennsylvania
and
the
Wills
Eye
Hospital. Imaging was performed with a 1.5-T Signa unit (GE Medical Systems, Mibwaukee). A localizing Ti-weighted spin-echo sequence (TR msec/TE msec 600/20) was initially used, followed by a long TR/ TE spin-echo sequence (3,000/30-35, 90) with flow compensation techniques (first order gradient moment nulling), one sig.
Abbreviations: tion
TE
echo
time,
TR
=
repeti-
time.
229
nal averaged, inferior spatial presaturation pulses, a 192 X 256 matrix, and 5mm-thick sections with a 2.5-mm section gap. The long TRITE images were reviewed by two neuroradiologists-onc familiar with and one unaware of the goal of the study. The prevalence of foci of high signal intensity in the periventricular white matter, centrum semiovale, and basal ganglia was determined from the long TR/TE images. RESULTS No
diabetic
patients
demonstrated
foci of high signal intensity in the white matter or deep gray matter. The images were read as normal by both the blinded radiologist and the ncunoradiologist dedicated to the study, who was specifically looking for deep and penventricular white matter changes. No control subject had white matter foci of high signal intensity. Using the Manteb-Hacnszeb test for the odds ratio (8), at beast eight of the diabetic patients would have had to have shown white matter foci of high signal intensity to produce a statisticalby significant (P < .05) difference between the control group and the patient group. DISCUSSION In the elderly population, the prcsence of white matter foci of high signal intensity
on
long
TRITE
MR
images
is
not uncommon. White matter foci of high signal intensity due to disease must be distinguished from a normal variant, the dilated perivascular (Vimchow-Robin) spaces, which are isointense with ccrebnospinal fluid on the proton-density
(3,000/30-35)
images.
By contrast, white matter foci of high signal intensity due to disease arc usually hypenntcnse to ccmebmospina! fluid on the proton-density images. These besions have been attributed to small yessd ischemic foci with gliosis (4-7,9). At pathologic examination, the arterioles show hyaline degeneration and microathemomata with surrounding atrophic pemivascular demyelination and/or myelin pallor presumably because of chronic ischemia on a hypoperfusion basis (7,10). These abnormal foci in the white matter have been meported to be present in as many as 30%44% of all patients 60 years old or older on long TRITE images. This rate is even higher (56%-78%) in patients with risk factors for atherosclerotic disease (6,11). In the population under the age of 50 years, diseases that produce white matten foci of high intensity include multiplc sclerosis, sarcoidosis, Lyme disease, vasculitis, and acute disseminated encephalomyclitis. cnibe
230
this
The
sign
#{149} Radiology
tendency
to diabetic
to as-
vascubopathy
in younger patients who have a long history of insulin-dependent diabetes mellitus derives from the knowledge that diabetes is a risk factor for stroke in the elderly (4-6,11). Our study suggests that this assumption is incorrect. One must search for another cause of the white matter lesions in the patient population 40 years old and younger. The proliferative vasculopathy that characterizes
diabetic
retinopathy
is
within the vascular distribution of the central retinal artery, a branch of the ophthalmic artery from the internal carotid system. Vascular disease identified in the intracranial circulation of the internal carotid artery would suggest an intrinsic vascular or systemic abnormality.
Alternatively,
one
could
hypothesize that the end organ, the retma, is the primary cause of or combines with the feeding vessel in the pathogenesis of diabetic retinopathy. Diabetic retinopathy includes lesions such as microaneurysms, hemorrhages, and infarcts. According to Memimce (2), the first insult occurs at the capillary penicytes, which control perfusion to the retina. Neovascularization, however, is the hallmark of proliferative retinopathy, and this occurs at the capillary and artcmiolar bevel. In the probifcrative stages, there arc capillary occlusions, neovasculanization, fibmovascular proliferation, and fibrous contraction leading to traction retinal detachments. Since vascular occlusions, infarcts, hemorrhages, pemivascular fibrosis/ gliosis, and Charcot-Bouchard ancurysms occur intracranially in patients with diabetes, the theory that the vasculopathy is a systemic disorder is rcasonablc.
There have been numerous theories as to the cause of diabetic retinopathy. In a recent review article by Mcmimec (2), three theories of the pathogcncsis of the disorder were described. The biochemical postulate emphasizes the preeminent mole of hyperglycemia in causing basement membrane damage and, hence, cndotheliab cell incornpetencc with production of prostagbandins that promote vascular thrombosis or hemorrhage. Glycosylated hemogbobin (known to be increased in patients with severe diabetic retinopathy) can induce other thmombogenic chemicals or can thicken the cndothelial basement membrane, activating the cornplernent system or causing local bloodflow variations (12). The second theory stresses the role of increased blood viscosity and platelet agglutination in diabetes, which meduces flow in capillaries and causes thrombosis. Proliferative neovascularization follows as an attempt to revascularize thrornbosed areas of the retina (2). The third component of the pathogenesis of diabetic retinopathy cmpha-
sizes
elevated
growth
hormone
and
insulin-like growth factor I levels, which have been shown to be elevated in diabetic patients with retinopathy. Growth hormone can induce thromboses, and injured, anoxic retinal cells may produce the insulin-like growth factor I, which can induce vascular neogenesis (2,13). The presence of white matter foci of high signal intensity in the brain of insulin-dependcnt diabetic patients would have suggested an intrinsic vascuban abnormality or would have supported the biochemical or hemodynamic theories of retinopathy pathogenesis. These theories hypothesize a more systemic basis of the disease. As regional cerebral blood flow decreases with age, the effect of diabetic small vessel disease would presumably have a greater role in inducing small vessel ischemic change in the white matter. The abscnce of these foci when there have been significant proliferative retinal changes, as seen in this study, testifies to the primary role of the substrate (ie, the retina) in the formation of proliferative diabetic retinopathy. The changes do not appear to be present elsewhere in the carotid system at MR imaging, a modality that is very sensitive to small vessel disease in the brain. U Acknowledgment: have been possible tance
provided
This
project
without
by Laura
could
not
the enormous
assis-
Bourke.
References 1.
Lynn athy.
2.
Menimee
3.
4.
5.
6.
7. 8. 9.
JR. Snyder WB. Vaiser A. Diabetic New York: Grune & Stratton, 1974. TJ.
Diabetic
retinopathy:
19:263-265. Drayen BP. mal findings.
Imaging of the aging brain. I. NorRadiology 1988; 166:785-796.
Woolson RF. Statistical methods for the analysis of biomedical data. New York: Wiley. 1987; 418422. Bradley WG. Waluch V. Brandt-Zawadzki M. YadWycoff
RR.
Patchy
matter lesions in the elderly: tion during NMR imaging. aging
11. 12.
13.
of
perspectives. N EngI J Med 1990; 322:978-983. Jerneld B. Algvere P. Relationship of duration and onset of diabetes to prevalence of diabetic retinopathy. Am J Ophthal 1986; 15:431-437. Brandt-Zawadzki M, Fein C. Van Dyke C. Kiernan R, Davenport L. deGroot J. MR imaging of the aging brain: patchy white-matter lesions and dementia. AJNR 1985; 6:675-682. Awad IA, Spetzler RF, Hodak JA, Awad CA. Carey R. Incidental subcontical lesions identified on magnetic resonance imaging in the elderly. I. Correlation with age and cerebrovascular risk factons. Stroke 1986; 17:1084-1089. Lechner K, Schmidt R, Bertha G, Justick E, Offanbacher H, Schneider C. Nuclear magnetic resonance image white matter lesions and risk factors for stroke in normal individuals. Stroke 1988;
icy RA.
10.
retinop-
a synthesis
1984;
periventnicular white common obsenvaNoninvasive Med Im-
1:35-41.
Kirkpatrick JB, Heyman LA. White-matter Icsions in MR imaging of clinically healthy brains of elderly subjects: possible pathologic basis. Radioiogy Gerard
1987;
162:509-511.
C, Weisberg L. MRI periventnicular Icsions in adults. Neurology 1986; 36:998-1001. Klein R, Klein BEK, Moss SE. et al. Glycosylated hemoglobin predicts the incidence and prognession of diabetic retinopathy. JAMA 1988; 260:2864-2871. Merimee TJ, Zapf
J, Foresch ER. Insulin-like studies in diabetics with and without retinopathy. N Engl J Med 1983; 309:527-530. growth
factors:
April
1991
