1. The UK Prospective Diabetes Study

2. UK Prospective Diabetes Study
multi-centre
randomised controlled trial
of different therapies
of Type 2 diabetes

3. UKPDS : need for a long-term study
complications of Type 2 diabetes develop over decades
Protocol written
Recruitment
End of study Sept. 1997
Clinical Centres 23
Type 2 diabetic patients 5102
Person years follow-up 53,000
Funding £23 million

4. UK Prospective Diabetes Study Centres
Aberdeen Lilian Murchison Manchester Andrew Boulton
Belfast City Randal Hayes Northampton Charles Fox
Belfast Royal David Hadden Norwich Richard Greenwood
Birmingham David Wright Oxford Robert Turner
Carshalton Steve Hyer Rury Holman
Memo Spathis Peterborough Jonathan Roland
Derby Ian Peacock Salford Tim Dornan
Dundee Ray Newton Scarborough Phil Brown
Roland Jung St George’s Nigel Oakley
Exeter Kenneth McLeod Stevenage Les Borthwick
John Tooke Stoke on Trent John Scarpello
Hammersmith Anne Dornhorst Lionel Alexander
Eva Kohner Torbay Richard Paisey
Ipswich John Day Whittington John Yudkin
Leicester Felix Burden

5. Co-ordinating Staff Chief Investigators : Robert Turner, Rury Holman
Statisticians : Irene Stratton, Carole Cull
Ziyah Mehta, Heather McElroy
Modeller : Richard Stevens
Epidemiologists : Andrew Neil, Amanda Adler
Diabetologists : David Matthews, Valeria Frighi
Biochemists : Susan Manley, Iain Ross
Administrators : Philip Bassett, Suzy Oakes
Retinopathy Grading Centre : Eva Kohner, Steve Aldington
Health Economics : Alastair Gray, Maria Raikou
Grant Applications : Ivy Samuel, Caroline Wood
Computing Support : Ian Kennedy, John Veness
And many others

6. Acknowledgements patients physicians nurses dietitians
retinal photographers
Retinopathy Grading : Hammersmith Hospital
Biochemistry : Diabetes Research Laboratories
ECG Grading : Guy’s Hospital

7. Major Funding Bodies UK Medical Research Council
British Diabetic Association
UK Department of Health USA National Institutes of Health (NEI, NIDDK)
British Heart Foundation Wellcome Trust
Novo Nordisk Bayer Lilly
Bristol Myers Squibb Lipha Hoechst
Farmitalia Carlo Erba

8. UK Prospective Diabetes Study
Glucose Control Study

9. Blood Glucose Control Study : Aims
to determine whether
improved glucose control of Type 2 diabetes will prevent clinical complications
therapy with
sulphonylurea - first or second generation
insulin
metformin
has any specific advantage or disadvantage

10. Patient Characteristics
5102 newly diagnosed Type 2 diabetic patients
age years mean 53 y
gender male : female 59 : 41%
ethnic group Caucasian 82% Asian 10%
Afro-caribbean 8%
Body Mass Index mean 28 kg/m2
fasting plasma glucose (fpg) median mmol/L
HbA1c median %
hypertensive 39%

11. UK Prospective Diabetes Study
follow-up of patients to major fatal and non-fatal clinical endpoints
recording of surrogate endpoints : clinical and biochemical markers e.g. urine albumin retinal photographs visual acuity
intention to treat analysis

12. Randomisation
14%
fpg : fasting plasma glucose (mmol/L)

13. UK Prospective Diabetes Study
Does an intensive glucose control policy reduce the risk of complications of diabetes?

14. Randomisation of Treatment Policies
342 allocated to metformin
Conventional Policy 30% (n=1138)
Intensive Policy 70% (n=2729)
Sulphonylurea n=1573
Insulin n=1156
Main Randomisation n=4209 (82%)
3867

15. Treatment Policies in 3867 patients
Conventional Policy
n = 1138
initially with diet alone
aim for near normal weight best fasting plasma glucose < 15 mmol/L asymptomatic
when marked hyperglycaemia develops allocate to non-intensive pharmacological therapy

16. Treatment Policies in 3867 patients
Intensive Policy with sulphonylurea or insulin
n = 2729
aim for fasting plasma glucose < 6 mmol/L asymptomatic
when marked hyperglycaemia develops on sulphonylurea add metformin move to insulin therapy on insulin, transfer to complex regimens

17. Actual Therapy

18. HbA1c
cross-sectional, median values

19. Change in Body Weight
cross-sectional, mean values

20. Hypoglycaemic Episodes
self-reported at each clinic visit
assessed by clinician to determine severity
graded as
minor : treated by patient alone
major : requiring third party assistance
grade of most severe episode recorded
all major episodes audited from clinical records

21. Hypoglycaemic episodes per annum
Actual Therapy analysis

22. Any Diabetes Related Endpoint
1401 of 3867 patients (36%)
First occurrence of any one of:
diabetes related death
non fatal myocardial infarction, heart failure or angina
non fatal stroke
amputation
renal failure
retinal photocoagulation or vitreous haemorrhage
cataract extraction or blind in one eye

23. Any Diabetes Related Endpoint (cumulative )
1401 of 3867 patients (36%)

24. Diabetes Related Deaths
414 of 3867 patients (11%)
Any of:
fatal myocardial infarction or sudden death
fatal stroke
death from peripheral vascular disease
death from renal disease
death from hyper/hypoglycaemia

25. Diabetes Related Deaths (cumulative)
414 of 3867 patients (11%)

26. Microvascular Endpoints (cumulative)
renal failure or death, vitreous haemorrhage or photocoagulation
346 of 3867 patients (9%)

27. Myocardial Infarction (cumulative)
fatal or non fatal myocardial infarction, sudden death
573 of 3867 patients (15%)

28. Aggregate Clinical Endpoints

29. Progression of Retinopathy
Two step change in Early Treatment Diabetic Retinopathy Study (ETDRS) scale

30. Microalbuminuria
Urine albumin >50 mg/L

31. Glucose Control Study Summary
The intensive glucose control policy maintained a lower HbA1c by mean 0.9 % over a median follow up of 10 years from diagnosis of type 2 diabetes with reduction in risk of:
12% for any diabetes related endpoint p=0.029
25% for microvascular endpoints p=0.0099
16% for myocardial infarction p=0.052
24% for cataract extraction p=0.046
21% for retinopathy at twelve years p=0.015
33% for albuminuria at twelve years p=

32. Conclusion
The UKPDS has shown that intensive blood glucose control reduces the risk of diabetic complications, the greatest effect being on microvascular complications

33. UK Prospective Diabetes Study
Does insulin or sulphonylurea therapy have specific advantages or disadvantages?

34. Sulphonylurea Therapy
advantages
known to improve glycaemic control
stimulates endogenous insulin production
disadvantages
in the heart sulphonylurea mimics ATP
and may prevent vasodilation in ischaemia
1st generation agents may increase arrhythmia

35. Insulin Therapy advantages
well-used therapy to improve glycaemic control
may be essential for many patients
disadvantages
need for injections
risk of weight gain and hypoglycaemia
raised insulin levels may promote atherosclerosis

36. Randomisation comparison between three intensive therapies
compare each with conventional policy

37. HbA1c
cohort, median data

38. change in weight
cohort, mean data

39. Hypoglycaemic episodes per annum
Actual Therapy analysis

40. Blood Pressure
cohort, mean data

41. Any diabetes-related endpoints
C v G v I p = 0.36

42. Myocardial Infarction
C v G v I p = 0.66

43. Progression of Retinopathy : 2 step change
favours intensive
favours conventional

44. Sulphonylurea or Insulin : Summary 1
all three therapies were similarly effective in reducing HbA1c
all three therapies had equivalent risk reduction for major clinical outcomes compared with conventional policy
in those allocated to chlorpropamide there was equivalent reduction of risk of microalbuminuria but no reduction of risk of progression of retinopathy

45. Sulphonylurea or insulin : Summary 2
Sulphonylurea therapy
no evidence of deleterious effect on myocardial infarction, sudden death or diabetes related deaths
Insulin therapy
no evidence for more atheroma-related disease

46. UK Prospective Diabetes Study
Does metformin in overweight diabetic patients have any advantages or disadvantages?

47. Introduction
the UKPDS has shown that an intensive glucose control policy using sulphonylurea or insulin therapy is effective in reducing the risk of complications in both overweight and normal weight patients
overweight (>120% Ideal Body Weight) UKPDS patients could be randomised to an intensive glucose control policy with metformin instead of diet, sulphonylurea or insulin

48. Insulin or Sulphonylurea 951
Randomisation
Main Randomisation 4209
Non overweight 2505
Overweight 1704
Conventional Policy 411
Intensive Policy 1293
Insulin or Sulphonylurea 951
Metformin 342

49. Patient Characteristics
overweight patients > 120% ideal body weight after three months’ diet therapy
age mean 53 years
gender male / female 46% / 54%
ethnic groups Caucasian 86%
Asian 6%
Afro-caribbean 8%
Body Mass Index mean 31 kg/m2
fasting plasma glucose median 8.1 mmol/L
HbA1c mean 7.2 %

50. HbA1c
overweight patients
cohort, median values

51. Change in Weight
overweight patients
cohort, mean values

52. Hypoglycaemic episodes per annum
overweight patients
Actual Therapy analysis

53. Any diabetes related endpoint
overweight patients
M v C p=0.0023
M v I p=0.0034

54. Diabetes related deaths
overweight patients
M v C p=0.017
M v I p=0.11

55. Myocardial Infarction
overweight patients
M v C p=0.010
M v I p=0.12

56. Microvascular endpoints
overweight patients
M v C p=0.19
M v I p=0.39

57. Metformin Comparisons
overweight patients
RR (95% CI)
favours metformin
favours conventional

58. Metformin Comparisons
overweight patients
RR (95% CI)
favours metformin or intensive
favours conventional

59. Sulphonylurea plus Metformin
patients primarily randomised to intensive therapy with sulphonylurea were not given additional metformin until their fpg was >15 mmol/L or they developed hyperglycaemic symptoms
in view of the progressive hyperglycaemia in these patients, a protocol modification was made to secondarily randomise the subset of patients who were on maximum sulphonylurea therapy and had fpg >6 mmol/L to earlier addition of metformin

60. Aim
the aim of this secondary randomisation was to assess the degree to which glycaemic control might be improved by early combination therapy with metformin
in view of the interesting results in the primary metformin study a secondary analysis was undertaken to examine any endpoints that had occurred

61. Aggregate Endpoints *
* interpret with caution in view of small numbers : 26 deaths on sulphonylurea plus metformin versus 14 deaths on sulphonylurea alone

62. Metformin in Overweight Patients
compared with conventional policy 32% risk reduction in any diabetes-related endpoints p= % risk reduction in diabetes-related deaths p= % risk reduction in all cause mortality p= % risk reduction in myocardial infarction p=0.01

63. Metformin : Summary
the addition of metformin in patients already treated with sulphonylurea requires further study
on balance, metformin treatment would appear to be advantageous as primary pharmacological therapy in diet-treated overweight patients

64. Blood Pressure Control Study
UK Prospective Diabetes Study
Blood Pressure Control Study

65. Blood Pressure Control Study : Aims
to determine whether
tight blood pressure control policy can reduce morbidity and mortality in Type 2 diabetic patients
ACE inhibitor (captopril) or Beta blocker (atenolol) is advantageous in reducing the risk of development of clinical complications

66. Inclusion criteria
patients NOT on anti-hypertensive therapy systolic >160 and/or diastolic > 90 mmHg
patients already ON anti-hypertensive therapy systolic >150 and/or diastolic > 85 mmHg
excluded if:
required strict blood pressure control; severe illness; contraindication to study medication or declined informed consent

67. Patient Characteristics
1148 Type 2 diabetic patients
age 56 years
gender male / female 55% / 45%
ethnic groups Caucasian 87%
Asian 6%
Afro-caribbean 7%
Body Mass Index 29 kg/m2
HbA1c %
systolic / diastolic blood pressure 160 / 94 mmHg
urine albumin > 50 mg/l 18%

68. Randomisation

69. Blood Pressure : Tight vs Less Tight Control
cohort, median values
Less tight control Tight control

70. Mean Blood Pressure mmHg baseline mean over 9 years
Less tight control 160 / / 87
Tight control 161 / / 82
difference 1 / 0 10 / 5
p n.s. <0.0001
ACE inhibitor 159 / / 83
Beta blocker 159 / / 81
difference 0 / 0 1 / 1
p n.s n.s. / p=0.02

71. Therapy requirement

72. Any diabetes-related endpoints
risk reduction 24% p=0.0046

73. Diabetes-related deaths
risk reduction 32% p=0.019

74. Myocardial Infarction
risk reduction 21% p=0.13

75. Stroke
risk reduction 44% p=0.013

76. Microvascular endpoints
risk reduction 37% p=0.0092

77. Heart Failure
risk reduction 56% p=0.0043

78. Progression of Retinopathy : 2 step change
243
461
207
411
152
300 20 40 60
% patients 23 37 28 51 34
3 years
6 years
9 years
p=0.38
p=0.019
p=0.004
Years from randomisation
numbers above bars are % affected

79. Deterioration of Vision : 3 lines on ETDRS chart
293
575
257
523
180
332 10 20 30 7 5 9 8 19
3 years
6 years
9 years
p=0.40
p=0.47
p=0.004
% patients
Years from randomisation
numbers above bars are % affected

80. Urine Albumin >50 mg/L
317
618
274
543
166
299 10 20 30 40 24 18 29 33
3 years
6 years
9 years
p=0.008
p=0.052
p=0.33
% patients
Years from randomisation
numbers above bars are % affected

81. Blood Pressure Control Study
in 1148 Type 2 diabetic patients a tight blood pressure control policy which achieved blood pressure of 144 / 82 mmHg gave reduced risk for
any diabetes-related endpoint 24% p=0.0046
diabetes-related deaths 32% p=0.019
stroke 44% p=0.013
microvascular disease 37% p=0.0092
heart failure 56% p=0.0043
retinopathy progression 34% p=0.0038
deterioration of vision 47% p=0.0036

82. UK Prospective Diabetes Study
Do ACE inhibitors or Beta Blockers have any specific advantages or disadvantages?

83. Blood Pressure : ACE inhibitor vs Beta blocker
cohort, median values
Less tight control ACE inhibitor Beta blocker

84. Reasons for non-compliance

85. Any Diabetes Related Endpoint (cumulative)
429 of 1148 patients (37%)

86. Diabetes Related Deaths (cumulative)
144 of 1148 patients (13%)

87. Microvascular Endpoints (cumulative)
renal failure or death, vitreous haemorrhage or photocoagulation
122 of 1148 patients (11%)

88. Aggregate Clinical Endpoints

89. Surrogate endpoints Relative Risk & 99% CI favours ACE inhibitor
favours Beta blocker

90. Conclusion
ACE inhibitors and Beta blockers were equally effective in lowering mean blood pressure in hypertensive patients with type 2 diabetes and in reducing the risk of:
any diabetes related endpoint
diabetes related deaths
microvascular endpoints

91. Potential implications for clinical care of diabetic patients
UK Prospective Diabetes Study
Potential implications for clinical care of diabetic patients

92. UK Prospective Diabetes Study
An intensive glucose control policy HbA1c 7.0 % vs 7.9 %
reduces risk of
any diabetes-related endpoints 12% p=0.030
microvascular endpoints 25% p=0.010
myocardial infarction 16% p=0.052
A tight blood pressure control policy 144 / 82 vs 154 / 87 mmHg reduces risk of
any diabetes-related endpoint 24% p=0.005
microvascular endpoint 37% p=0.009
stroke 44% p=0.013

93. Choice of Therapies diabetes :
each of the available therapies studied can be used
in overweight, diet-treated patients, metformin may be advantageous
hypertension :
Beta blockers and ACE inhibitors each provide protection

94. Which goals of therapy? current guidelines suggest HbA1c <7%
the risk of diabetic complications was reduced in the UKPDS trial which achieved a median HbA1c 7.0% in the intensive glucose control group
this HbA1c level is in accord with current guidelines but is difficult to accomplish in some patients
epidemiological analysis suggests that any reduction of hyperglycaemia would be advantageous

95. Which goals of therapy?
current guidelines suggest blood pressure <140 / 85 mmHg or <130 / 85 mmHg
the risk of diabetic complications was reduced in the UKPDS blood pressure control trial which achieved a mean blood pressure 144 / 82 mmHg in the tight control group
this result is in accord with current guidelines, which are also supported by the epidemiological analysis

96. Polypharmacy glycaemia
combinations of agents with different actions will be needed
more patients will require insulin
blood pressure
many patients will need 3 or more different types of agents

97. Differences between Therapies
sulphonylurea, insulin and metformin are each effective in reducing the risk of any diabetes related endpoints and microvascular endpoints
no evidence of increased risk of complications for any single therapy
ACE inhibitors and Beta blockers are each effective in reducing the risk of macrovascular and microvascular endpoints
no evidence that either is specifically advantageous

98. UK Prospective Diabetes Study
The UKPDS has shown conclusively that :
intensive therapy to reduce glycaemia is worthwhile as it reduces risk of complications
tight blood pressure control is worthwhile as it reduces risk of complications
there are no major differences between the therapies tested
reduction in risk of complications of diabetes is a realisable goal

99. Beneficial Effects of Intensive Therapy
The UKPDS has shown that
more intensive monitoring
more intensive use of existing therapies
which improves
blood glucose control
blood pressure control
can reduce the risk of diabetic complications

100. UK Prospective Diabetes Study
papers presenting major results of the study
UKPDS 33: Lancet (1998) 352,
UKPDS 34: Lancet (1998) 352,
UKPDS 38: BMJ (1998) 317,
UKPDS 39: BMJ (1998) 317,