ASCORE An uptodate cardiovascular risk score for hypertensive

ASCORE : An up-to-date cardiovascular risk score for hypertensive patients reflecting. contemporary clinical practices developed. using the ASCOT trial database. D Prieto-Merino 1, AK Gupta 2, J Dobson 2, CL Chang 2, SE Mastorantonakis 2, PS Sever 2, B Dahlof 3, H Wedel 4. S Pocock 1, NR Poulter 2 on behalf of the ASCOT investigators London School of Hygiene and Tropical Medicine 2 ICCH, Imperial College London 3 Sahlgrenska University Hospital, Sweden 4 Nordic School of Public Health, Goteborg, Sweden 1

Introduction • Guidelines recommend total CV risk estimation for CV patients 1 • Various risk assessment tools available Framingham 2 QRisk 2 ASSIGN PROCAM NHANES INDANA 3 SCORE • Need for a new score for patients with multiple CV risk factors 1 Mancia G et al. J Hypertens 2007; 25: 1751 -62 2 Anderson KM et al. Am Heart J 1991; 121: 293 -8 3 Pocock SJ et al BMJ 2001; 323: 75 -81

ASCOT Study design • 160/100 mm. Hg untreated, 140/ 90 mm. Hg treated • Age 40 -79 years • No previous MI or current clinical CHD • 3 or more CV risk factors Sever PS et al. J Hypertens 2001; 19: 1139 -47

ASCOT Study design 19, 257 hypertensive patients atenolol ± bendroflumethiazide ASCOT-BPLA 5. 5 years follow-up amlodipine ± perindopril PROBE design 10, 305 patients TC ≤ 6. 5 mmol/L (250 mg/d. L) atorvastatin 10 mg Double-blind Sever PS et al. J Hypertens 2001; 19: 1139 -47 ASCOT-LLA 3. 3 years follow-up placebo

ASCORE Study design • Study population: ASCOT patients excluding those with history of CV disease at baseline • Primary composite endpoint: first CV event (Death, MI or Stroke) • Statistical analysis: Cox proportional hazard models with backward stepwise variable selection

ASCORE Study design Baseline variables defined a priori: – – – Sex Age Ethnicity Education Smoking Diabetes Prev. antihypert. treatment Prev. lipid lowering treatment Aspirin use SBP DBP Heart rate – – – Height Weight BMI Microalbuminuria / proteinuria Renal dysfunction Total cholesterol HDL Fasting LDL Fasting TGs Fasting glucose Creatinine GFR

ASCORE ASCOT Results 19, 257 -3302 excluded ASCORE 15, 955 Endpoints at 5 years 1, 114

ASCORE and ASCORE-S

Model calibration Predicted by ASCORE . 10 Predicted by ASCORE-S Observed (KM) . 15 Risk of having the event in 5 years . 15 Observed (KM) ASCORE . 10 P-value of Chi 2 = 0, 22 . 05 ASCORE-S P-value of Chi 2 = 0, 49 . 05 . 00 0 1 2 3 4 5 6 7 Deciles of predicted risk 8 9 . 00 0 1 2 3 4 5 6 7 8 Deciles of predicted risk 9

0. 75 0. 50 area: 0. 664 ASCORE-S area: 0. 648 0. 25 ASCORE P-value for comparison < 0. 0001 0. 00 Sensitivity 1. 00 ROC curves and AUC 0. 00 0. 25 0. 50 0. 75 1 -Specificity 1. 00

External validation INDANA (INdividual Data ANalysis of Antihypertensive intervention trials) 53000 patients 10 antihypertensive trials 15 -20 years ago Interventions Primarily ß-blockers and diuretics vs placebo

External validation • Subset of 13335 patients – with no CV history – aged 40 -80 y, not randomized to placebo – Complete data on CV endpoints – Complete data for all variables of ASCORE-S

Risk of CV event in 5 years . 2 Average risk predicted with ASCORE-S Observed risk in INDANA database . 15 . 1 . 05 0 0 1 2 3 4 5 6 Patients by deciles of risk 7 8 9

Mean change after 5 years 20 SBP T. C. 1 10 0. 5 0 0 -10 -0. 5 -20 -1 ASCOT -30 -40 -1. 5 -2 INDANA -50 -2. 5 -60 -3 <150 150 -170 170 -190 >190 SBP at baseline <4 4 -6 6 -8 >8 T. Cholesterol at baseline Dahlof B et al. Lancet 2005; 366: 895 -906, Sever PS et al. Lancet 2003; 361: 1149 -58

Average risk predicted with ASCORE-S Risk of CV event in 5 years . 2 Expected risk in INDANA if treatment was as in ASCOT Observed risk in INDANA database . 15 . 1 . 05 0 0 1 2 3 4 5 6 Patients by deciles of risk 7 8 9

1 -0. 9985 exp(0. 1*total risk score)

1 -0. 9985 exp(0. 1*total risk score)

Score conversion to 5 year risk Total Risk score ASCORE-S 0 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 0. 15% 0. 41% 0. 50% 0. 61% 0. 74% 0. 90% 1. 10% 1. 35% 1. 64% 2. 00% 2. 44% 2. 97% 3. 62% 4. 40% 5. 35% 6. 49% 7. 87% 9. 53% 11. 51% 13. 87% 16. 67% 19. 97% 0. 34% 0. 92% 1. 12% 1. 37% 1. 67% 2. 04% 2. 49% 3. 03% 3. 68% 4. 48% 5. 45% 6. 61% 8. 02% 9. 70% 11. 72% 14. 12% 16. 97% 20. 32% 24. 23% 28. 74% 33. 89% 39. 68%

ASCORE Summary • ASCORE and ASCORE-S models and easy to use score charts • 5 year risk prediction in hypertensive patients under treatment • Internal and External validation (ASCORE-S) • Consistency with contemporary treatment guidelines • Established predictors in an up-to-date model