Clinical outcomes of HbA1c standardisation

Clinical outcomes of HbA1c standardisation Eric S. Kilpatrick Division Chief, Clinical Chemistry Sidra Medicine, Doha, Qatar Professor of Pathology & Laboratory Medicine, Weill Cornell Medicine-Qatar Honorary Professor in Clinical Biochemistry, Hull York Medical School, UK

HbA1c standardisation and clinical outcomes What are the clinical outcomes associated with HbA1c? How has standardisation improved the precision of outcomes? Could a HbA1c unit change worsen outcomes? Is there more to life than just HbA1c?

What are the clinical outcomes associated with HbA1c? The Diabetes Control and Complications Trial

The DCCT 1,441 patients with type 1 diabetes Half the patients assigned to conventional treatment one or two insulin injections per day aim to remain asymptomatic Half the patients assigned to intensive treatment 3 or more injections per day aim for preprandial BG 3.9 to 6.7mmol/L Followed for a mean of 6.5 (0-9) years DCCT Group 1993;329:977-86

DCCT: Glucose in treatment groups

HbA1c (%) DCCT: HbA1c in treatment groups 11 10 9 8 Intensive Conventional 7 6 5 0 1 2 3 4 5 6 7 8 9 Yrear of Study

DCCT: Results Intensive treatment was associated with: 76% reduction in risk of retinopathy 39% reduction in risk of microalbuminuria 54% reduction in risk of albuminuria 60% reduction in risk of neuropathy DCCT Group 1993;329:977-86

DCCT: Risk of Microvascular Complications

DCCT: Risk of Severe Hypoglycaemia

UKPDS: Risk of Macro and Microvascular Complications UKPDS Group. BMJ 2000;321:405-

CVD after the DCCT 42% DCCT/EDIC Group. N Engl J Med 2005;353:2643-53

Playing the odds If I had a child develop type 1 diabetes tomorrow What are the chances they will remain complication-free for the rest of their life?

% Risk of Retinopathy progression in next year DCCT: risk of retinopathy progression 18 16 14 12 10 8 6 4 2 0 5 6 7 8 9 10 11 12 HbA1c (%) 40 50 60 70 80 90 100 HbA1c (mmol/mol)

Probability of retinopathy progression Cumulative risk 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 6 7 8 9 10 HbA1c (%) 50 60 70 80 HbA1c (mmol/mol) 1 year

Probability of retinopathy progression Cumulative risk 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 6 7 8 9 10 HbA1c (%) 50 60 70 80 HbA1c (mmol/mol) 1 year 5 years

Probability of retinopathy progression Cumulative risk 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 6 7 8 9 10 HbA1c (%) 50 60 70 80 HbA1c (mmol/mol) 1 year 5 years 10 years

Probability of retinopathy progression Cumulative risk 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 6 7 8 9 10 HbA1c (%) 50 60 70 80 HbA1c (mmol/mol) 1 year 5 years 10 years 20 years

Probability of retinopathy progression Cumulative risk 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 6 7 8 9 10 HbA1c (%) 50 60 70 80 HbA1c (mmol/mol) 1 year 5 years 10 years 20 years 30 years

Probability of retinopathy progression Cumulative risk 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 6 7 8 9 10 HbA1c (%) 50 60 70 80 HbA1c (mmol/mol) 1 year 5 years 10 years 20 years 30 years 40 years

Probability of retinopathy progression Cumulative risk 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 6 7 8 9 10 HbA1c (%) 50 60 70 80 HbA1c (mmol/mol) 1 year 5 years 10 years 20 years 30 years 40 years 50 years

Probability of retinopathy progression Playing the odds 19 years old 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 6 7 8 9 10 11 12 HbA1c (%) 40 50 60 70 80 90 100 108 HbA1c (mmol/mol) 1 year 5 years

50:50 odds of developing retinopathy HbA1c % mmol/mol 12 108 11 97 10 86 9 75 8 64 7 53 6 42 Age (years) 24 26 30 36 48 73 129

Sep-96 Jan-97 May-97 Aug-97 Nov-97 Mar-98 Jul-98 Nov-98 Mar-99 Juy-99 Jan-00 May-00 Sep-00 Jan-01 May-01 Sep-01 Mar-02 Jul-02 Nov-02 May-03 Sep-03 Jan-04 May-04 Sep-04 Jan-05 May-05 Overall CV (%) % DCCT aligned methods 12 10 8 NGSP programme -Dec'96 Menarini Standardisation meeting - Sep '98 Interlaboratory variation HbA1c UK Consensus statement - Jan '00 IFCC standardisation adopted by manufacturers Jan '04 Sample 1 Sample 2 CV Trendline 100 90 80 70 60 6 50 4 40 30 2 20 10 0 0 Date

50:50 odds of developing retinopathy HbA1c % mmol/mol 12 108 11 97 10 86 9 75 8 64 7 53 6 42 Age (years) 24 26 30 36 48 73 129

50:50 odds of developing retinopathy HbA1c % mmol/mol 12 108 11 97 10 86 9 75 8 64 7 53 6 42 Age 8% CV 24 26 30 36 48 73 129

50:50 odds of developing retinopathy HbA1c % mmol/mol 12 108 11 97 10 86 9 75 8 64 7 53 6 42 Age 8% CV 24 (22 29) 26 (23 34) 30 (25 42) 36 (28 56) 48 (34 85) 73 (46 138) 129 (73 277)

50:50 odds of developing retinopathy HbA1c % mmol/mol 12 108 11 97 10 86 9 75 8 64 7 53 6 42 Age 8% CV 24 (22 29) 26 (23 34) 30 (25 42) 36 (28 56) 48 (34 85) 73 (46 138) 129 (73 277) 3% CV

50:50 odds of developing retinopathy HbA1c % mmol/mol 12 108 11 97 10 86 9 75 8 64 7 53 6 42 Age 8% CV 3% CV 24 (22 29) (23 27) 26 (23 34) (24 28) 30 (25 42) (27 33) 36 (28 56) (32 41) 48 (34 85) (41 58) 73 (46 138) (61 90) 129 (73 277) (101 171)

What about using HbA1c for diagnosis of type 2 diabetes?

WHO, 2011 www.who.int/cardiovascular_diseases/report-hba1c_2011_edited.pdf

Imprecision and diagnosis At 48mmol/mol (6.5%) threshold: 8% CV between-labcorresponds to 40 to 56mmol/mol (5.8% to 7.3%)

UKNEQAS HbA1c Aug 2014 SI (IFCC) value 46mmol/mol (6.3%) CV=5.6% 5.6% 6.9%

Getting used to SI numbers DCCT IFCC (mmol/mol) 4% 20 5% 31 6% 42 7% 53 8% 64 9% 75 10% 86

Effect of method change on clinician HbA1c targets Hanas R. Diabetes Care 2002; 25: 2110-2111

Getting used to SI numbers DCCT IFCC (mmol/mol) 4% 20 5% 31 6% 42 7% 53 8% 64 9% 75 10% 86

Kilpatrick s Kludge minus 2 minus 2 DCCT 7% Eric Kilpatrick, ACB mailbase 30/1/08

Kilpatrick s Kludge minus 2 minus 2 DCCT -2 7% 5 Eric Kilpatrick, ACB mailbase 30/1/08

Kilpatrick s Kludge minus 2 minus 2 DCCT -2-2 7% 5 3 Eric Kilpatrick, ACB mailbase 30/1/08

Kilpatrick s Kludge minus 2 minus 2 DCCT -2-2 IFCC 7% 5 3 53mmol/mol Eric Kilpatrick, ACB mailbase 30/1/08

DCCT 4% 5% 6% 7% 8% 9% 10% Kilpatrick s Kludge

DCCT -2 4% 2 5% 3 6% 4 7% 5 8% 6 9% 7 10% 8 Kilpatrick s Kludge

DCCT -2-2 4% 2 0 5% 3 1 6% 4 2 7% 5 3 8% 6 4 9% 7 5 10% 8 6 Kilpatrick s Kludge

DCCT -2-2 4% 20 5% 31 6% 42 7% 53 8% 64 9% 75 10% 86 Kilpatrick s Kludge

Kilpatrick s Kludge DCCT IFCC (mmol/mol) 4% 20 5% 31 6% 42 7% 53 8% 64 9% 75 10% 86

Transition to SI HbA1c units in the UK Pre-1 st June 2009 Only NGSP (%) units reported 1 st June 2009-1 st October 2011 Dual reporting of NGSP and SI units Post-1 st October 2011 Only SI (mmol/mol) reported

Aims To establish if HbA1c values in diabetes patients changed in the year after SI-only reporting was introduced in Hull, UK To determine if the HbA1c following a raised (>8%/64mmol/mol) result was different after the unit change

Monthly mean HbA1c (%) HbA1c before and after unit change

HbA1c before and after unit change Year before unit Year after unit change 2010-11 change 2011-12 All samples n 21880 22841 p-value HbA1c (%) 7.5 (6.6,8.7) 7.5 (6.5,8.7) 0.34 (mmol/mol) 58 (49,72) 58(48,72) All data expressed as median (25 th, 75 th centiles)

Change in HbA1c following initial value>8% Year before unit change 2010-11 Year after unit change 2011-12 n 4316 4396 p value HbA1c change (%) (mmol/mol) -0.2(-0.9,0.3) -2(10,3) -0.2(-0.8,0.3) -2 (9,3) 0.44 Days between HbA1c samples 99(64,147) 98(64,147) 0.45 Difference between 2 successive DCCT/SI values (before unit change) and 2 successive SI-only values (after unit change)

Why no difference??healthcare staff adapted better than anticipated?related to good educational material and/or the dual reporting period?having to take care because of the complete change in numbers?converting back to NGSP

Conclusions A move to SI HbA1c reporting did not lead to: Any marked short-term deterioration in glycemia A different HbA1c outcome in patients with initial poor glucose control.

HbA1c (%) DCCT: HbA1c in treatment groups 11 10 9 8 Intensive Conventional 7 6 5 0 1 2 3 4 5 6 7 8 9 Yrear of Study

HbA1c and retinopathy risk in the DCCT HbA1c explained 96% of the difference in retinopathy risk between the treatment groups BUT HbA1c did not explain 96% of the retinopathy risk Diabetes 2008;57:995-1001

Understanding the DCCT Glycaemic exposure (A1C + duration of diabetes) explains 11% of the variation of DR risk in the DCCT 11 89 Diabetes 2008;57:995-1001

Understanding the DCCT 89 so that other factors may presumably explain the remaining 89% of the variation in risk among subjects independent of A1C. Diabetes 2008;57:995-1001

What might explain the 89% of risk??genetic?environmental?other known vascular risk factors e.g. hypertension, hyperlipidaemia

HbA1c variability and complication risk in the DCCT Diabetes Care 31:2198 2202, 2008

HbA1c variability and complication risk Type 1 diabetes Type 2 diabetes in the majority of studies, [HbA1c] variability was more predictive of adverse outcomes than mean HbA1c. Diabetes Care 2015;38:2354 2369

HbA1c variability and complication risk Can we run people with high mean HbA1c values as long as they are stable? Would this mean they would be at the same risk of microvascular complications but less risk of hypoglycaemia? Might HbA1c variability become a new diagnostic target to aim for?

Summary HbA1c has been shown to be a marker of vascular complication risk in type 1 and 2 diabetes Standardisation has allowed a large improvement in the clinical precision of the test, especially over longer periods of time Evidence to date is that changing HbA1c units as part of standardisation does not necessarily have a detrimental effect on patient care Glycaemia is only one factor in the development of diabetes complications