Diabetic retinopathy. I. The assessment of severity and progress by comparison with a set of standard fundus photographs

Classification of diabetic retinopathy: Past, present and future

Diabetic retinopathy (DR) is a leading cause of visual impairment and blindness worldwide. Since DR was first recognized as an important complication of diabetes, there have been many attempts to accurately classify the severity and stages of disease. These historical classification systems evolved as understanding of disease pathophysiology improved, methods of imaging and assessing DR changed, and effective treatments were developed. Current DR classification systems are effective, and have been the basis of major research trials and clinical management guidelines for decades. However, with further new developments such as recognition of diabetic retinal neurodegeneration, new imaging platforms such as optical coherence tomography and ultra wide-field retinal imaging, artificial intelligence and new treatments, our current classification systems have significant limitations that need to be addressed. In this paper, we provide a historical review of different classification systems for DR, and discuss the limitations of our current classification systems in the context of new developments. We also review the implications of new developments in the field, to see how they might feature in a future, updated classification.

Retinal perfusion pressure and pulse pressure: clinical parameters predicting progression to sight-threatening diabetic retinopathy

Background and aims his paper investigates the roles of retinal perfusion pressure and pulse pressure in progression to sight-threatening diabetic retinopathy. Retinal perfusion pressure is dependent upon the systemic blood pressure and the intra-ocular pressure. We have examined the hypothesis that the early deleterious effect of hypertension on the retinal circulation which leads to perfusion pressure damage is due to a unique relationship between the intra-ocular pressure and the systemic blood pressure. Clinically this is manifest as hypertensive retinopathy and diabetic retinopathy.

Materials and methods From a prospective database, data on all new patients with background diabetic retinopathy referred during a three year period (n=104) were analysed for progression to sight-threatening diabetic retinopathy.

Results Retinal perfusion pressure and pulse pressure were found to be the strongest predictors of progression to sight-threatening diabetic retinopathy (Chi Square linear trend statistic, relative risk ratios).

Progression to maculopathy by quartilesFor retinal perfusion pressure, the relative risk ratios increased from 1.00 in the first quartile (39.7—50.1 mmHg), to x 5.8 in the second quartile (50.2—56.1 mmHg), to x 6.4 in the third quartile (56.2—60.0 mmHg), to x 9.1 in the fourth quartile (60.1—75.3 mmHg) (p=0.012). For pulse pressure the values were x 1.00, x 4.0, x 4.5, x 6.0 respectively (p=0.021). For systolic pressure the values were x 1.00, x 3.4, x 5.5, x 5.5 respectively (p=0.026). Diastolic pressure was not significant.

Progression to proliferative retinopathy by quartilesFor retinal perfusion pressure, the relative risk ratios increased from 1.00 in the first quartile to x 2.9 in the second quartile, x 4.2 in the third quartile to x 4.1 in the fourth quartile (p=0.002). For pulse pressure the values were x 1.00, x 1.7, x 3.8, x 5.5 respectively (p=0.002). For systolic pressure the values were x 1.00, x 0.5, x 2.9, x 3.8 respectively (p=0.008). Diastolic pressure was not significant.

Conclusion Retinal perfusion pressure and the pulse pressure are important predictors of progression to sight-threatening diabetic retinopathy.

Prevalence of retinopathy and proteinuria in type 1 diabetics in Iceland

The prevalence of retinopathy and proteinuria was assessed in 212 and 230 type 1 diabetics in Iceland respectively. They represent 78% and 84% of all such patients identified in the country. Retinopathy was present in 33.5%, background lesions only in 27.4% and more severe retinopathy in 6.1%. After 5-9 years of diabetes (mean 6.9) the prevalence of retinopathy was 18.8%, including 2.1% proliferative, and rose to 76.7% after 20 years or more of diabetes (mean 26.7), including 16.2% in the proliferative stage. Blindness was found in 2.4% of the patients. Diabetics diagnosed at the age of 0-19 years had a lower prevalence of retinopathy during their first 20 years of diabetes than those diagnosed later in life (p less than 0.05). Proteinuria was present in 14%, intermittent in 10% and continuous in 4%. After 10 years or more of diabetes (mean 19.3) the prevalence of intermittent and continuous proteinuria was 13.0% and 8.3%, respectively.

Prevalence of retinopathy in non insulin dependent diabetes mellitus at a diabetes centre in southern India

A cohort of 6792 NIDDM patients attending a diabetes centre at Madras in South India was screened using a combination of retinal photography and clinical examination by retinal specialists. A total of 2319 patients (34.1%) had evidence of retinopathy. This included 2090 patients (30.8%) with non-proliferative diabetic retinopathy including 435 patients (6.4%) with maculopathy and 229 patients (3.4%) with proliferative diabetic retinopathy. Multiple logistic regression analyses showed that duration of diabetes, glycosylated haemoglobin, type of treatment (insulin treatment versus non-insulin treatment), systolic and diastolic blood pressures and serum creatinine, showed a positive association with retinopathy while body mass index (BMI) showed an inverse association. The prevalence rates of retinopathy in Southern Indians are comparable to those seen in Europeans. However in view of the high prevalence of diabetes in the Indian sub-continent, diabetic retinopathy could become a formidable challenge in the future.

Vascular complications in long-term south Indian NIDDM of over 25 years' duration

The prevalence of vascular complications was assessed in 726 South Indian non-insulin dependent diabetes mellitus (NIDDM) patients with over 25 years' duration of diabetes. Retinopathy was detected in 52.0% of patients which included 41.7% with non-proliferative and 10.3% with proliferative diabetic retinopathy. Nephropathy was present in 12.7% and neuropathy in 69.8% of patients. While 32.8% of patients had ischaemic heart disease, the prevalence of peripheral vascular disease was only 15.4%. Multivariate logistic regression analyses showed that serum creatinine was associated with retinopathy, creatinine and post-prandial plasma glucose with nephropathy and post-prandial plasma glucose and age with neuropathy. This is one of the first reports on vascular complications in long-term diabetes from the Indian sub-continent.

A computerized system for localization of diabetic lesions from fundus images

A computerized system for the localization of diabetic lesions from fundus photographs is described. The coordinates for each lesion in relation to reference points are determined from fundus images projected on a digitizing table. The program accepts some variation in coordinates for a single lesion, still to be recorded as the same lesion in a series of photographs, and the magnitude of this variation is adjustable. The data obtained from the localization program are transferred to a register program yielding the following parameters: number of lesions, new lesions, lost lesions, variable lesions, time of presence for lost lesions. These data can be further exported into Microsoft Excel 4, for analysis of large patient materials. An intra-observer error of 0% and an inter-observer error of 1.1% was found in localization of lesions from fundus images. The program was insensitive to small variations in the fundus pictures, but comparisons of 30 degrees and 60 degrees photographs of the same fundus, or 20 degrees deviation of gaze, induced identification errors in 2.9% and 11.9% of localizations, respectively. With this system new parameters on diabetic fundus images can be obtained. The significance of these parameters in the progression of diabetic retinopathy will be analyzed in further studies.

Retinal blood flow in diabetic retinopathy

OBJECTIVES

(a) To report on the basic parameters of retinal blood flow in a population of diabetic patients with and without retinopathy and non-diabetic controls; (b) to formulate a haemodynamic model for the pathogenesis of diabetic retinopathy from this and other studies.

DESIGN

Laser-Doppler velocimetry and computerised image analysis to determine retinal blood flow in a large cross sectional study.

SETTING

Diabetic retinopathy outpatient clinic.

SUBJECTS

24 non-diabetic controls and 76 diabetic subjects were studied (63 patients with insulin dependent diabetes, 13 with non-insulin dependent diabetes). Of the diabetic subjects, 12 had no diabetic retinopathy, 27 had background retinopathy, 13 had pre-proliferative retinopathy, 12 had proliferative retinopathy, and 12 had had pan-retinal photocoagulation for proliferative retinopathy.

MAIN OUTCOME MEASURES

Retinal blood flow (microliters/min) and conductance (rate of flow per unit of perfusion pressure).

RESULTS

In comparison with non-diabetic controls (9.52 microliters/min) and diabetic patients with no diabetic retinopathy (9.12 microliters/min) retinal blood flow was significantly increased in all grades of untreated diabetic retinopathy (background 12.13 microliters/min, pre-proliferative 15.27 microliters/min, proliferative 13.88 microliters/min). There was a significant decrease in flow after pan-retinal photocoagulation in comparison with all the other groups studied (4.48 microliters/min). Conductance of the retinal circulation was higher in the untreated diabetic retinopathy groups. These results were independent of age, sex, type of diabetes, duration of diabetes, glycated haemoglobin concentration, blood glucose concentration, blood pressure, and intraocular pressure.

CONCLUSIONS

Retinal blood flow is significantly increased in diabetic retinopathy in comparison with non-diabetic controls and diabetic subjects with no retinopathy. This has implications for controlling hypertension and hyperglycaemia as a strategy in reducing morbidity from diabetic retinopathy.

A population-based study of the incidence of complications associated with type 2 diabetes in the elderly

One hundred and eighty-eight known Type 2 diabetic patients aged over 60 years identified by a geographically based survey of a population of 40,076 were followed for a median of 6 years to determine the incidence of various complications. There were 63 deaths and two patients were lost to follow-up. The presence of complications was determined using a structured questionnaire and clinical examination. Incidence rates of ischaemic heart disease, stroke, and peripheral vascular disease (PVD) were 56 (95% CI 41-75), 22 (13-35), and 146 (117-174) 1000-person-years-1 of follow-up, respectively. Rates of stroke and PVD rose significantly with age. Retinopathy occurred at a rate of 60 (42-83) 1000-person-years-1 and cataract at 29 (17-46) 1000-person-years-1 although visual acuity in survivors did not deteriorate overall, probably reflecting the high mortality associated with cataract. The rate of proteinuria (albumin concentration greater than 300 mg l-1) was 19 (9-34) 1000-person-years-1. Incidence rates were unrelated to sex or duration of diabetes. Diabetes is associated with a continuing incidence of complications into old age. Adequate facilities are required to assess and treat the resulting morbidity in a population with an increasing proportion of elderly people.

Screening for diabetic retinopathy

There is a need for greater educational emphasis, both at undergraduate and postgraduate level, on the detection of diabetic eye disease, in particular diabetic retinopathy. The early diagnosis of the sight-threatening lesions of proliferative retinopathy and maculopathy is a prerequisite for the prevention or reduction of the visual loss and blindness associated with this diabetic complication. It is also essential that patients are aware that diabetes can result in visual loss due to diabetic retinopathy. Patients should understand that diabetic retinopathy may be present without ophthalmic or diabetic symptoms and that its incidence increases with duration of diabetes, poor diabetes control, and hypertension. They must also be aware that, if detected early, retinopathy can be treated successfully and vision preserved. Early detection depends on regular eye examination involving both visual acuity assessment and ophthalmoscopy through dilated pupils by experienced personnel. A comprehensive programme of screening followed by prompt and adequate treatment would made a significant contribution to eradicating diabetic retinopathy as a cause of blindness.

Lens opacity and mortality in diabetes

A population-based cohort of 294 diabetic patients were examined for the presence of lens opacities and followed up for a median of 6 years. Only two patients were lost to follow-up but there were 73 deaths, 49 in the 108 patients with opacities and 24 in the 184 patients without (odds ratio 2.4, 95% Cl 1.5-3.9). Lens opacities were a powerful predictor of death and the effect was independent of other prognostic factors. The presence of cataract identifies a high risk sub-group of elderly diabetic patients.

Macular blood flow response to acute reduction of plasma glucose in diabetic patients measured by the blue light entoptic technique

The response of retinal blood flow to acute reduction in plasma glucose levels was studied in 20 poorly controlled type I diabetic patients. Perifoveal flow velocity was determined, using the blue-light entoptoscope, and arterial calibers measured, using a computer-aided digitizing system. Mean plasma glucose level was lowered from 17.7 +/- 4 to 7.0 +/- 1 mmol/l after a subcutaneous insulin infusion and measurements taken at both glucose levels. The autoregulatory change induced by breathing 60% oxygen at the two plasma glucose levels also was compared. Mean flow velocities were 0.54 +/- 0.28 mm/sec at a high plasma glucose level compared with 0.55 +/- 0.32 mm/sec at a low plasma glucose level, whereas hyperoxia reduced these by 16.58 and 16.71%, respectively. No significant difference in the responses of arterial diameters to hyperoxia between the two glucose levels was found. The authors conclude that acute reduction in plasma glucose level in this group of patients is not associated with significant changes in macular blood flow or in alteration in autoregulation.

Diabetes in the elderly: the Oxford Community Diabetes Study

Most diabetic patients are elderly but their clinical characteristics remain poorly defined. A population survey identified 259 known diabetic patients aged 60 years or more giving a prevalence of 3% in this age group. A total of 193 patients (75%) were interviewed and examined, 155 (80%) of whom had been diagnosed at under 70 years of age. Forty-two patients (22%) were insulin-treated but clinical characteristics suggested that at least 95% of all elderly patients had Type 2 diabetes. Blood glucose control was poor with median HbA1 9.7% (range 4.9-17.1%, normal reference range 5.0-7.5%), and 55% were either overweight or obese. There was a high morbidity from diabetes and other conditions: the prevalence of hypertension (untreated blood pressure of 160/95 mmHg or more or antihypertensive medication) was 52%, of stroke 5%, of nephropathy (urinary albumin concentration greater than or equal to 300 mg l-1) 3%, of lower limb amputations 4%, and of foot ulcers 7%. The prevalence of symmetrically impaired distal vibration perception was 23%, and 54% of patients either needed or were receiving chiropody. The prevalence of a corrected distant visual acuity of 6/12 or worse was 32% and of retinopathy of any degree was 26%. There was extensive co-morbidity which was not confined to a single subgroup of patients.

Long-term follow-up of retinal blood flow in diabetes using the blue light entoptic phenomenon

The blue light entopic phenomenon was used to measure retinal blood flow in 87 diabetics and 10 normal controls in a longitudinal study. The time interval between the initial and final studies was 21-39 months (mean 31.4). The retinal blood flow velocity did not change in the 10 subjects in the control group nor in those nine patients with no retinopathy, all but one of whom remained free of retinal lesions. In the 24 with background retinopathy the flow velocity fell significantly from 0.71 (SD 0.35) mm/s to 0.48 (0.13) mm/s (p less than 0.01). The velocity also decreased in the preproliferative group of 16 patients from 0.62 (0.39) mm/s to 0.41 (0.14) mm/s (p less than 0.05). The retinopathy status changed in 11 of this group. In those with proliferative retinopathy (now treated) and those treated previously there was no change in flow velocity.

Pancreatic B-cell function in relation to diabetic retinopathy in Asian Indian NIDDM patients

Pancreatic B-cell function in relation to diabetic retinopathy was studied in 195 NIDDM patients with long-standing diabetes. Background diabetic retinopathy (BDR) was present in 95 (48.7%) and proliferative retinopathy (PDR) in 17 (8.7%) of the subjects. There was no significant difference between the BDR, PDR, and non-retinopathy groups with respect to age, age at diagnosis of diabetes and HbA1 values. Mean duration of diabetes was higher in the PDR group (p less than 0.05). Serum C-peptide values showed no correlation with the presence of retinopathy or with the duration of diabetes. The C-peptide values were widely scattered in patients with BDR and PDR showing no association between pancreatic B-cell reserve and occurrence or severity of retinopathy in NIDDM patients. Thus, decreased pancreatic B-cell reserve does not appear to be a risk factor for diabetic retinopathy in NIDDM patients.

Clinical features of diabetes in the young as seen at a diabetes centre in south India

This study reports on the clinical pattern of 545 consecutive young diabetic patients with age at onset below 30 years attending a diabetes centre in Southern India. Three hundred and fourteen patients (57.7%) were classified as having non-insulin-dependent diabetes of the young (NIDDY), 119 (22%) as insulin-dependent diabetes (IDDM) and 28 (5%) as malnutrition-related diabetes (MRDM); 4% fibrocalculous pancreatic diabetes and 1% protein-deficient pancreatic diabetes. The remaining 84 patients could not be classified into any of the above categories. A positive family history of diabetes was more common in NIDDY compared to the other groups (P less than 0.001). While 40.3% of patients with IDDM had age at onset below 15 years, the other types of diabetes were rarely seen in patients younger than this. Body mass index (BMI) did not reliably indicate the MRDM forms of diabetes as 70% of patients with IDDM also had a BMI of less than 18, one of the criteria recommended for the diagnosis of MRDM. C-peptide levels in MRDM were intermediate between the IDDM and NIDDY groups. Microvascular complications were present in all the groups of young diabetics. The frequency was higher in NIDDY patients who also had a longer duration of diabetes. There was an increasing prevalence of complications with increasing duration of diabetes.

Three year prospective study of visual function and retinopathy in diabetics with improved glycaemic control. Diabetic Retinopathy Study Group St Thomas' Hospital

The visual function and degree of retinopathy was assessed, over a three-year period, in a cohort study of twenty-eight diabetics, in whom glycaemic control was improved by intensive monitoring and supervision of conventional therapies. With the exception of visual acuity and some tests of visual field sensitivity, there was no significant change in visual functions or retinopathy; with improved control of blood glucose, these two visual functions showed a small initial deterioration and subsequently returned towards starting values. Six subjects required laser photocoagulation for progressive peripheral neovascularisation (including two subjects with peripheral new vessels), the six having a significantly longer duration of diabetes, slightly worse measures of extra-foveal retinal functions and a significantly greater reduction in haemoglobin A1 concentration during the first six months of the study. In this study, the improvement of blood glucose control by intensive supervision of conventional therapy did not appear to be associated with the significant acute deterioration of visual function or retinopathy that has been reported with the strict diabetic control by multiple daily insulin injections or continuous subcutaneous insulin infusion.

Long-term follow-up of patients who underwent yttrium-90 pituitary implantation for treatment of proliferative diabetic retinopathy

Between 1960 and 1976 117 patients underwent pituitary implantation with yttrium-90 (90Y) for treatment of proliferative retinopathy at the Hammersmith Hospital, London. Mean age at operation was 35 +/- 11 years (mean +/- SD), and mean duration of diabetes 18.6 +/- 10.0 years. Mean insulin dosage prior to implant was 67.2 +/- 24 units, falling to 30.4 +/- 14.9 units post-implant. Thirty-two per cent of patients are still living, 60% are deceased and 8% are lost to follow-up. The 5-year survival rate was 82%. Of the causes of death, 21% died of infection, adrenal insufficiency or hypoglycaemia, 12% of renal failure, and 47% of myocardial or cerebral vascular disease. Ophthalmological follow-up was carried out on the 100 patients operated on between 1965 and 1976. The mean age of this group at implant was 35 +/- 10.5 years, and mean duration of diabetes 17.2 +/- 8.7 years. Visual acuity in the better eye at operation was 6/12 or better in 84% of patients, and this percentage remained similar at the time of the 5 and 10 year follow-up. Blindness (6/60 or worse) in both eyes was present in 12% of patients at the time of 5 and 10 year assessments. By 5 years new vessels on the disc had improved from a mean grading of 2.7 +/- 1.6 to 0.8 +/- 1.2 (p less than 0.001), and by 10 years there was no disc neovascularisation in any eye. There was a similar improvement in the grading of hard exudates, microaneurysms and haemorrhages, but there was an increase in fibrous retinitis proliferans.(ABSTRACT TRUNCATED AT 250 WORDS)

Lack of effect of gliclazide on early diabetic nephropathy and retinopathy: a two-year controlled study

This study has attempted to document a specific haemovascular action of gliclazide on the reversal of early diabetic microangiopathy. A prospective double-blind controlled study was performed over 2 years, comparing gliclazide versus placebo in insulin-treated and gliclazide versus glibenclamide in non-insulin-treated diabetic subjects, after a 1-year run-in period. Glycaemic control was not significantly different in gliclazide- and non-gliclazide-treated subjects before or after the commencement of active therapy. Following treatment with gliclazide in 17/32 insulin-treated and 8/17 non-insulin-treated subjects with Albustix-negative proteinuria, there was no difference in retinopathy score, total proteinuria or the renal clearance of creatinine, albumin, transferrin and immunoglobulin G. In the insulin-treated group, progression of retinopathy was observed in approximately one-third of subjects, but no parameter of proteinuria progressed over 2 years. Thus, this study did not detect a reversal of the parameters measured and does not support an action of gliclazide on diabetic microangiopathy, independent of its hypoglycaemic action.

Quantifying early diabetic retinopathy

A precise and accurate method of numerically quantifying diabetic retinopathy, on standardised retinal colour photographs, has been developed which allows small changes and trends to be monitored. Colour slides are projected onto a screen and features noted on an acetate sheet which provides a permanent record. Sector analysis showed microaneurysms and haemorrhages to occur most often at the temporal-to-macula area, exudates at the macula and cotton wool spots on the nasal side of the retina. Seventy percent of microaneurysms appeared in the previous year, irrespective of the severity of the retinopathy. In proportion to their usual relative prevalences, after normalisation of distribution, the various features can be combined to provide a single value, the Retinopathy Index. This provides an overall assessment of retinopathy which is suitable for comparing the progress of mild retinopathy in prospective studies.

Diabetic retinopathy. Pathogenesis and the role of retina-derived growth factor in angiogenesis

Diabetic retinopathy results from a combination of systemic and ocular abnormalities. Vasodilation, basement membrane pathology, microaneurysms, abnormal blood flow and tissue oxygenation, connective tissue abnormalities, and retinal ischemia are all components of early diabetic retinopathy. The pathogenesis of neovascularization is discussed with respect to the effects of vasodilation, vascular leakage, vitreous changes, and retinal ischemia. The evidence supporting Michaelson's hypothesis that a chemical messenger from the retina provides the stimulus for neovascularization is cited. The sequence of events involved in angiogenesis are cellular and basement membrane changes, endothelial cell migration, endothelial cell proliferation, and vessel formation. The experimental evidence in support of a role for retina-derived growth factor as a mediator of these cellular events is reviewed.

Natural history of peripheral neovascularisation in diabetic retinopathy

A study of 62 untreated eyes of 54 diabetic patients with peripheral neovascularisation is reported. The study shows that neovascularisation rarely remains confined to one sector of the retina; 18% of eyes had developed disc new vessels within one year, 27% had a vitreous haemorrhage within five years, and 10% developed either macular traction or a retinal detachment.

Photocoagulation for peripheral neovascularisation in diabetes

A study of 90 eyes in 72 diabetic patients with peripheral retinal neovascularisation treated by photocoagulation indicated that the subsequent development of further peripheral or disc new vessels is infrequent if panretinal photocoagulation is the initial treatment procedure. Focal or sectorial treatment may be considered only if a small area of neovascularisation is present initially and provided regular follow-up can be guaranteed.

Retinopathy is associated with higher glycaemia in maturity-onset type diabetes

In a group of 149 maturity-onset type diabetic patients followed from diagnosis, 55 (37%) had retinopathy on colour photography 7 years later. Those patients with retinopathy had significantly greater glycaemia, as shown by higher fasting plasma glucose levels at diagnosis, larger mean values for fasting glucose 1, 3 and 5 years later, and higher random glucose and haemoglobin A1c at ophthalmic review (p = 0.001, 0.002, 0.007 and 0.001, respectively). Substantial retinopathy, as measured by greater than 5 microaneurysms, also correlated significantly with each index of glycaemic control.

Photocoagulation for proliferative diabetic retinopathy: a randomised controlled clinical trial using the xenon-arc

The final results of a randomised controlled study of xenon-arc photocoagulation for proliferative retinopathy are reported, after all patients have been followed for at least 5 years and some for up to 7 years. One hundred and seven patients with two similarly affected eyes had one treated (chosen by a random procedure), while the other eye remained untreated and served as a control. Of the 107 patients, 77 completed the 5 year follow up, 13 died and 17 stopped attending for various reasons. Of the recorded coexistent medical abnormalities, only renal complications affected survival, none influenced visual outcome. Visual outcome was significantly better in the treated than in the control eyes at each yearly interval (0.001 less than p less than 0.05). The greatest difference was seen in those with disc vessels at entry. In this group, control eyes deteriorated by a mean of four lines on the Snellen chart, treated eyes by one line only. Six patients became legally blind in both eyes, four were blind in the treated eye only, but 28 control eyes were blind when treated eyes retained vision (p less than 0.001). Treated eyes which became blind had less treatment than those that retained vision. Of the 42 treated eyes with peripheral new vessels only at entry, 12 developed disc new vessels. These 12 had fewer burns than the 30 which did not develop disc new vessels. It is concluded that in proliferative retinopathy, treatment by photocoagulation is better than no treatment at all. Adequate treatment is required to maintain vision.

A comparative trial of xenon arc and argon laser photocoagulation in the treatment of proliferative diabetic retinopathy

We present the results of a randomised study comparing xenon arc and argon laser photocoagulation in the treatment of proliferative diabetic retinopathy. Our results show that when used in the way described below there is no significant difference between the effect of the 2 treatments on optic disc neovascularisation and visual acuity. We also compare the results of peripheral retinal ablation with those of panphotocoagulation in the argon and xenon groups. The results suggest that in some patients peripheral treatment may be sufficient to cause regression of disc neovascularisation.

Xenon arc photocoagulation for the treatment of diabetic maculopathy. Interim report of a multicentre controlled clinical study

The results of a randomised controlled clinical study in the treatment of diabetic maculopathy are reported. Ninety-nine patients with two similarly affected eyes had one--chosen by random procedure--treated, while the other was left as an untreated control. Visual acuity was maintained in the treated eyes, while it deteriorated in the untreated eyes, reaching over 2 lines difference after 4 years. The difference in deterioration was significantly less in treated than in untreated eyes at each assessment. After 4 years the difference between the 2 eyes was greatest in those with initially good vision. There was no difference in visual acuity in patients whose initial vision was less than 6/36. Only 9 treated but 25 control eyes lost vision to blindness (P less than 0.001). It is concluded that maculopathy is a treatable form of diabetic retinopathy if treated before severe visual loss occurs.

Diabetic retinopathy and pregnancy

Diabetic retinopathy was found to be present in 12 out of a group of 67 diabetic patients supervised by us during 92 pregnancies, and 3 further pregnant diabetics were referred to us because of retinopathy. The mean duration of diabetes was 13 years (range 3--25 years). Nine patients had minimal retinopathy, 2 had background retinopathy, and the remaining 4 proliferative retinopathy. The cases with minimal retinopathy showed no progression during pregnancy. In 1 patient with background retinopathy there was deterioration. Of the 4 patients with proliferative retinopathy 1 showed regression during the pregnancy, 2 showed advance and were treated with photocoagulation (these 2 patients now have normal vision), while the patient with extensive retinitis proliferans, with retinal detachment in both eyes and previous photocoagulation remained unchanged. The prognosis during pregnancy for patients with diabetic retinopathy is reasonable and has been improved by the advent of photocoagulation.

Diabetic retinopathy study. Assessment and comparison of retinal lesions for computer analysis

A method of documentation of a diabetic retinopathy (dR) is described. Panorama fundus photographs (PPhs) of each eye are taken every 3-6 months. The retinopathy is classified according to qualitative, quantitative, and topographic criteria. The retinal lesions are assessed according to a subjective grading scale. Data of direct readings and of comparisons between two PPhs of one eye are transferred to punch cards for storage and computer processing. In a double evaluation of 200 PPhs, the overall reliability of readings was found to be approximately 95%.

Role of heat-labile toxigenic Escherichia coli and Reovirus-like agent in diarrhoea in Boston children

61 Boston children aged five years or less with acute diarrhoea were studied for evidence of infection with Escherichia coli strains that produce heat-labile enterotoxin (L.T.) or with a reovirus-like agent associated with childhood gastroenteritis. This represented the first evaluation of the prevalence of disease produced by these two agents in the same population. E. coli, isolated from acute-phase stool specimens, were tested in adrenal-cell tissue-culture and adult-rabbit ileal-loop assays for L.T. Acute and convalescent phase sera, collected from 31 children, were tested by the adrenal-cell assay for anti-L.T. activity. None of the 61 children demonstrated evidence of infection with L.T.-positive E. coli. Paired sera from 31 of the children studied were also tested for evidence of recent infection with the reovirus-like agent by determining titres of immunofluorescent-staining antibody to the serologically related Nebraska calf diarrhoea virus. 11 of the children (35%) had evidence of recent infection. These results suggest that an important proportion of endemic acute diarrhoea of young children in Boston is caused by the reovirus-like agent, and that disease caused by L.T.-producing E. coli is uncommon.

Diabetic retinopathy

Retinopathy is present in 60%-80% of long-term diabetics, and 5%-10% of diabetics surviving 20 yr from the time of diagnosis will be blind, mostly from retinopathy, which is now the commonest cause of newly diagnosed blindness in the 30-65 yr age group. The mean survival time after a diagnosis of retinopathy is only 5 yr. The natural history of diabetic retinopathy is now being understood more clearly. Mild background retinopathy, characterised by microaneurysms or scattered hard exudates, may progress to maculopathy, with macular vascular pathology leading to exudates and edema at the macula; this is most common in older patients. It may also lead to proliferative retinopathy, which may progress slowly with new vessels and fibrous-tissue formation, or rapidly with widespread capillary closure and soft-exudate formation, extensive neovascularisation, hemorrhages, and blindness. The changes of diabetic retinopathy can be documented using retinal photography, and several grading methods have been devised that are useful for evaluation treatment. The cause of diabetic retinopathy is still unclear. Evidence for incriminating genetic factors, growth-hormone excess, and hypoxia associated with changes in blood flow and retinal metabolism are reviewed. Insulin responses and plasma triglyceride seem to be different in maturity-onset diabetics with retinopathy when compared with those in whom this complication is absent. Most physicians agree that good diabetic control may both lower the incidence of retinopathy and reduce the speed of its progress. While there is little evidence that drugs are ever of much value in this condition, the role of photocoagulation both by laser and xenon arc is becoming clearer with increased experience of these techniques, and the current situation is reviewed. Pituitary ablation is a very drastic method of treatment and should never be used as a desperate measure in a patient with advanced proliferative disease. It is, however, the treatment of choice in early florid retinopathy, when it proves the only chance to arrest the condition. Some new techniques of vitreous surgery are now being developed and the possible role of these in the management of patients with advanced proliferative disease is briefly reviewed.

Serum growth hormone levels and the response of diabetic retinopathy to pituitary ablation

Serum growth hormone levels were measured during insulin tolerance tests in 36 patients after yttrium-90 pituitary implantation for diabetic retinopathy. The response of the new blood vessels was more clearly related to loss of growth hormone function than was the improvement of retinal haemorrhages and microaneurysms. The overall response of the retinopathy was greatest when growth hormone function was lost.Since the loss of growth hormone function was related to the loss of other aspects of anterior pituitary function, a unique role of growth hormone in the response of diabetic retinopathy to pituitary ablation could not be established.