Imaging Biomarkers: utilisation for the purposes of registration EMEA-EFPIA Workshop on Biomarkers 15 December 2006
Vascular Imaging Technologies Carotid Ultrasound-IMT IVUS-PAV QCA-% stenosis 2
ICH E 9 Guidance for Industry (II.2.2.6) statistical principles for clinical trials Evidence to establish surrogacy Biological Plausibility e.g., pathophysiology suggests a relationship between the surrogate and a clinical outcome Epidemiologic Data prognostic value of the surrogate for clinical outcome Clinical Trial Evidence treatment effects on the surrogate correspond to clinical outcomes QCA CIMT IVUS ICH, E9, Step 4 5 Feb 1998 3
Quantitative Coronary Angiography (QCA) First accepted imaging surrogate of clinical efficacy Set the standard for the field when placebo control groups were still acceptable Measures lumen diameter (mm) or percent stenosis 4
QCA progression and clinical events Imaging measurements and clinical events with cholesterol lowering therapies QCA meta-analyses (Rossouw) Lower odds of Less progression = coronary events Lifestyle Resins Statins Combx All Drug Surgery All 0.0 0.5 1.0 1.5 2.0 Odds Ratio of Progression Lifestyle Resins Statins Combx All Drug Surgery All 0.0 0.5 1.0 1.5 2.0 Odds Ratio of Clinical Event Rossouw J E. Am J Cardiol. 1995;76:86C-92C. 5
Limitations QCA Limitations QCA can only measure luminal atherosclerosis (i.e., advanced disease that impinges the lumen) Most events occur in non-stenotic vessels % Patients with MI (n = 195) 80 <50% 50-70% >70% 70 60 50 40 30 20 10 0 Diameter Stenosis Smith. Circulation. 1996;93:2205-2211 (data from four studies) Digital QCA loses spatial resolution - Not sufficient resolution for active control studies 6
Advantages Vascular Ultrasound Coronary and carotid ultrasound permit direct visualisation of vascular disease Detects latent disease not visible to angiography atheroma De Franco AC, Nissen SE, Am J Cardiol 2001;88:7M-20M 7
Carotid Ultrasound Measures of intima-media thickness 8
Evidence that Carotid IMT Predicts Clinical Disease Epidemiological Data Rotterdam Study CLAS (Cholesterol Lowering Atherosclerosis Study) posttreatment long-term outcomes follow-up CHS (Cardiovascular Health Study) Study in CIMT risk of MI risk of Stroke Rotterdam 0.163 mm SD 1.43 (OR) 1.41 (OR) CLAS 0.03 mm/yr 2.2 (RR) CHS 0.2 mm SD 1.46 (OR) 1.43 (OR) Bots ML, Hoes AW, Koudstaal PJ, et al. Circulation 1997;96:1432-1437 (Rotterdam) Hodis HN, Mack WJ, LaBree L, et al Ann Int Med 1998;128:262-269 (CLAS) O Leary DH, Polak JF, Kronman RA, et al N Eng J Med 1999;340:14-22 (CHS) 9
Cardiovascular Health Study: Baseline CIMT Predictive of Stroke and MI 4,476 subject without CHD at baseline Median follow-up 6.2 yrs RR of stroke or MI for highest vs. lowest CIMT quintile was 3.87 100 Mean CIMT* Cumulative Event-free Rate (%) 95 90 85 80 75 1st Quintile 2nd Quintile 3rd Quintile 4th Quintile 5th Quintile 0.89 mm 0.96 mm 1.13 mm 1.36 mm 1.50 mm 70 0 1 2 3 4 5 6 7 Years O Leary et al, N Engl J Med. 1999;340:14-22. 10
Evidence that changes in CIMT Predict Clinical Disease: Lipid intervention studies: meta-analysis Clinical trials involving HMG-CoA reductase inhibitors that reported both Carotid IMT and Cardiovascular Event Outcomes Clinical Trial (N*) Statin Relative Impact on IMT Progression of Primary Outcome (mm/yr): Mean [95% CI] (Reported p-value) Relative Impact on Reported Cardiovascular Endpoints: Odds Ratio [95% CI] Abstracted CVD Event Odds Ratio ACAPS (25) (N=919) Lovastatin -0.015 [-0.023, -0.007] (p=0.001) CVD Death, MI, Stroke 0.34 [0.12, 0.69] KAPS (26) (N=447) Pravastatin -0.014 [-0.022, -0.006] (p=0.005) CVD Death, MI, Stroke 0.57 [0.22, 1.47] PLAC-II (47) (N=151) Pravastatin -0.009 [-0.031, 0.013] (p=0.44) Clinical Coronary Events 0.37 [0.11, 1.24] CAIUS (48) (N=305) Pravastatin -0.014 [-0.021, -0.005] (p=0.0007) CVD Death, MI 1.02 [0.14, 7.33] REGRESS (28) (N=255) Pravastatin -0.030 [-0.056, -0.004] (p=0.002) Clinical Events 0.51 [0.24, 1.07] BCAPS (49) (N=793) Fluvastatin -0.008 [-0.013, -0.003] (p=0.002) CVD Death, MI, Stroke 0.64 [-0.24, 1.66] FAST (50) (N=164) Pravastatin Significant Benefit (p <0.001) CVD Death, MI 0.32 [0.10, 1.06] Pooled Estimate -0.012 [-0.016, -0.007]** 0.48 [0.30, 0.78] *Arms used in meta-analysis; ** Excludes FAST. Espeland MA et al. Current Controlled Trials in Cardiovasc Med. 2005;6:3-6. 11
Change in CIMT in Blood Pressure Lowering Trials: meta-analysis Trial n Baseline IMT (µm) Change/y (µm) Difference (µm/y, 95% CIs) Control:Active Control:Active Migdalis 20:20 680:701 PART2 309:308 790:800 SECURE 2.5 mg 227:232 1146:1148 SECURE 10 mg 227:234 1146:1160 Hosomi 50:48 700:700 PREVEND 323:319 770:770 All ACEIs 929:1161 Heterogeneity:X 2 =18.1, P=0.003 BCAPS 390:393 893:912 ELVA 44:35 897:894 All BBs 434:428 Heterogeneity: X 2 =3.7, P=0.05 Control:Active 103:30 5:8 22:18 22:14 20:10 11:8 8:7 12: 12 6 ( 12 to 0.4) P=0.07 10 ( 33 to 13) P=0.41 PREVENT 186:191 1258:1259 11: 4 All trials 1549:1780 Heterogeneity: X 2 =25.9, P=0.001 7 ( 12 to 2) P=0.01 100 50 0 50 100 Favors Active Treatment Favors Control Wang et al. Stroke 2006;37:1933-40 12
CIMT as surrogate for CV events: Summary Of LDL and BP Trials IMT change produces parallel estimates of risk and benefit. Changes in events are deductible from observed changes in the surrogate Statin RCTs BP RCTs CIMT effect (mm/yr) -0.012 [ 0.016, 0.007] -0.007 [ 0.012, 0.002] CV event effect 0.48 [0.30, 0.78]* 0.71 [0.55, 0.92]** * Espeland et al. Curr Control Trials Cardiovasc Med 2005;6:3 ** Wang et al. Stroke 2006;37:1933-40 RCT = randomized control trial 13
Atherosclerosis Imaging: QCA and IMT Current cardiovascular imaging supported USA and/or CAN Labeling 1 Compound Imaging type: endpoints Supporting Studies Imaging Indication Lovastatin (Mevacor, 1995) QCA: MLD, % stenosis QCA: CCAIT, MARS, FATS slow progression of CAD CIMT CIMT: ACAPS Pravastatin (Pravachol, 1996) QCA: MLD CIMT QCA: PLAC I; REGRESS slow progression of CAD CIMT: PLAC II, REGRESS, KAPS Simvastatin (Zocor, 1996) QCA: MLD QCA: MAAS slow progression of coronary atherosclerosis; reduce new lesions and total occlusions (Canada) Fluvastatin (Lescol, 1997) QCA: MLD QCA: LCAS slow progression of CAD Niacin (Niaspan, 1997) + resin QCA: global change score, % stenosis QCA: CLAS, FATS slow progression or promote regression of CAD 1. lovastatin NDA 19-643/Supp #034,1995; pravastatin NDA 19-898/Supp # 013,1996; fluvastatin NDA 02-026/Supp #012,1997; Niaspan NDA 20-381, 1997. 14
Coronary Ultrasound: Evidence that IVUS Predicts Clinical Disease Epidemiological Data CHD patients (Von Birgelen et al) Correlation with change in left main coronary artery atheroma cross sectional area (CSA) with Risk factors Risk scores Clinical events IVUS-PAV Von Birgelen C, Hartmann M, Mintz G, et al. Circulation. 2004;110:1579-1585. 15
Coronary Ultrasound: Evidence that IVUS Predicts Clinical Disease Plaque & Media Changes / Year (%) Total Cholesterol 204 mg/dl (n=31) Total Cholesterol <204 mg/dl (n=25) 6.1 ± 16.5 16.6 ± 19.5 p <0.05 LDL-Cholesterol 125 mg/dl (n=28) LDL-Cholesterol <125 mg/dl (n=28) 3.8 ± 15.2 19.8 ± 19.1 p <0.001 HDL-Cholesterol 46 mg/dl (n=28) HDL-Cholesterol >46 mg/dl (n=28) 2.4 ± 14.2 21.2 ± 18.5 p <0.001 Total-C/HDL-C-Ratio 4 (n=39) Total-C/HDL-C-Ratio <4 (n=17) 3.7 ± 15.5 15.3 ± 19.3 p <0.02 Triglycerides 118 mg/dl (n=28) Triglycerides <118 mg/dl (n=28) 9.4 ± 14.6 14.2 ± 22.4 p=0.4 Systolic Blood Pressure 140 mmhg (n=34) Systolic Blood Pressure <140 mmhg (n=22) 10.2 ± 19.3 14.3 ± 18.4 p=0.4 Age 59 Years (n=30) Age <59 Years (n=26) 13.2 ± 19.5 10.6 ± 18.7 p=0.6 Family History (n=10) No Family History (n=46) 10.9 ± 17.8 15.8 ± 23.9 p=0.6 Diabetes Mellitus (n=10) No Diabetes Mellitus (n=46) 10.4 ± 19.5 18.3 ± 15.2 p=0.2 Smoker (n=16) No Smoker (n=40) 6.6 ± 17.0 24.7 ± 17.5 p <0.002 0 5 10 15 20 25 30 Von Birgelen C, Hartmann M, Mintz G, et al. Circulation. 2004;110:1579-1585. 16
Mean LDL Cholesterol Levels and Median Change in Percent Atheroma Volume for Several Intravascular Ultrasound Trials ACTIVATE Placebo Adapted from Nissen, S. E. et al. JAMA 2006;0:295.13.jpc60002-10. 17
Indirect Evidence that CIMT and IVUS Predict Clinical Events: Pravastatin/Atorvastatin Studies Mean CIMT +/ 2 SEM 0.70 0.68 0.66 0.64 0.62 0.60 0.58 0.56 0.54 ARBITER (CIMT) Pravastatin Atorvastatin Baseline 6 Months 12 Months Change in TAV (%) 4 3 2 1 0-1 REVERSAL (IVUS) 2.7* Pravastatin 40 mg Significant atherosclerotic progression from baseline P=0.02-0.4 Atorvastatin 80 mg No significant change from baseline; atherosclerotic progression was stopped 30 PROVE-IT Death or Major Cardiovascular Event (%) No. at Risk Pravastatin Atorvastatin 25 20 15 10 5 0 40 mg of Pravastatin 80 mg of Atorvastatin P=0.005 0 3 6 9 12 15 18 21 24 27 30 Months of Follow-up 2063 2099 1688 1736 1536 1591 1423 1485 810 842 138 133 18
Combination therapies have the potential to reduce cardiovascular risk further than mono-therapy An example Risk of Coronary Heart Disease and Lipid Levels 3 Relative Risk In 4 Years 2 1 0 25 0.7 Increased Risk 100 55 85 1.4 2.2 2.6 HDL-Cholesterol (mg/dl or mmol/l) Framingham Study; Am J Cardiol. 2000. 220 160 5.7 4.1 LDL-Cholesterol (mg/dl or mmol/l) Increased Risk 19
Clinical Development Ongoing Vascular Imaging Trials: a case study Scientific objective Demonstrate the incremental benefit of novel therapy combined with LDL-C lowering therapy over standard of care alone in slowing the progression of atherosclerosis QCA CIMT IVUS Biological Plausibility e.g., pathophysiology suggests a relationship between the surrogate and a clinical outcome Epidemiologic Data prognostic value of the surrogate for clinical outcome ICH, E9, 5 Feb 1998 Clinical Trial Evidence treatment effects on the surrogate correspond to clinical outcomes? 20
Atherosclerosis Imaging 2006 Questions Does the EMEA consider that carotid ultrasound now meets the criteria of ICH E9 for a surrogate variable, in consideration of the wealth of pathophysiologic, epidemiologic, and clinical trial data published since the guidance was issued? If not, why not? Does the EMEA concur that atherosclerosis is a systemic disorder and that IVUS measures of wall thickness represent a higher resolution image of IMT than that detected by carotid ultrasound? If not, why not? 21
Atherosclerosis Imaging 2006 Questions What are the EMEA comments on the fact that vessel wall thickness is an integrated marker of cardiovascular risk and that IVUS, as an example does not attempt to assess vulnerable plaque, but as is the case with carotid IMT, is an assessment of the patient s potential for future CV events? 22
Backup 23
Example of a Phase 3 Imaging Program Coronary IVUS Subject Population: Angio. proven CAD (> 20% stenosis) Eligible for statin treatment Core labs for central reading multi-country, multi-center Coronary IVUS B-mode US Carotid IMT Subject Populations: Heterozygous FH mixed hyperlipidemia, TG > 150 eligible for statin treatment Coronary IVUS B-mode US / 6 months S C R E E N I N G statin dose titration target: LDL-C to CV risk goal novel agent and statin statin alone* 24 months D-B/ randomized Tx * Same as statin dose at the end of the titration period. Target population technology Number of subjects Endpoint CHD IVUS ~ 1000 Nominal PAV Mixed dyslipidemia Carotid ultrasound ~ 800 IMT (mm)/year Heterozygous FH Carotid ultrasound ~ 800 IMT (mm)/year