1	Lessons Learned: Surrogates & Intermediate Outcomes David L. DeMets, Ph.D. Department of Biostatistics and Medical Informatics University of Wisconsin-Madison
2	Definition of Surrogate & Intermediate Outcome A physiologic or biological measurement that is used instead of a clinical outcome (e.g. live longer, feel good, function better) Surrogate may be used to Develop a new intervention Assess biological or mechanical activity Assess clinical efficacy
3	Appeal of Surrogates/Intermediate Outcomes Criticism of clinical trials Too long Too large Too expensive May allow for studies to be Smaller Shorter and faster Less expensive
4	Response Variables Cancer Death or disease recurrence Tumor shrinkage Cardiology Death, non fatal MI or hospitalization Blood pressure, lipid levels, PVCs, cardiac output
5	Response Variables AIDS Death or progression to AIDS T4 Cells, viral load Diabetes Death, visual impairment, kidney failure Microaneurysms, clearance
6	Use of Surrogates Phase I Trials Maximum Dose/Dose Response Phase II Trials Measures of Activity Phase III Trials Supporting Evidence/Secondary Outcomes e.g., Cholesterol Changes Primary Outcome?
7
8
9	Concerns About Surrogates 1. Relationship between surrogate and true endpoint may not be causal, but coincidental to a third factor 2. Other unfavorable effects of the drug 3. Effect on surrogate may correlate with one clinical endpoint, but not others
10	Intermediate/Surrogate Outcomes Reliance on intermediate outcome might lead to incorrect conclusion about benefit or harm Consider a few examples
11	Nocturnal Oxygen Therapy Trial (NOTT) Hypothesis Is continuous oxygen therapy better than nocturnal oxygen therapy in chronic obstructive lung disease patients? Surrogates Survival Design 203 patients Two-sided 0.05 Type I error Powered for several intermediate outcomes Randomized, multicenter
12	Possible NOTT Surrogates PaO₂ Hematocrit FEV₁ % Predicted FVC % Predicted Maximum Workload Heart Rate Mean Pulmonary Artery Pressure Cardiac Index Pulmonary Vascular Resistance Neuropsychiatric Impairment Quality of Life
13
14	Cardiac Arrhythmias Cardiac arrhythmias associated with sudden death Class of drugs developed to suppress arrhythmias Drugs approved for high risk patients “Off-label” use increased
15	Cardiac Arrhythmia Suppression Trial Hypothesis Does suppression of arrhythmia following an MI reduce incidence of: 1. Sudden death 2. Total mortality
16	Cardiac Arrhythmia Suppression Trial Three Drug Arms vs. Placebo Double blind placebo control One Sided (0.025 Type I Error) for Benefit Advisory One Sided (0.025) for Harm Sequential monitoring plan
17	Cardiac Arrhythmia Suppression Trial Early Termination in Two Drug Arms Drugs Placebo Sudden Death 33 9 Total Mortality 56 22
18	Chronic Heart Failure A major problem in heart disease Increased mortality, decreased quality of life Drugs developed to improve cardiac function Not known if survival or quality of life improved
19	PROFILE (Prospective Randomized Flosequinan Longevity Evaluation) Flosequinan is primarily a vasodilator Approved for CHF improved exercise tolerance reduced symptoms slight adverse mortality PROFILE Design randomized double blind multicenter Mortality outcome placebo vs. 75 mg vs. 100 mg Class III or IV CHF
20	PROFILE Results Flosequinan Dose 75 mg 100 mg Combined Total Flosequinan 40/206 201/964 241/1170 Mortality (19.4%) (20.9%) (20.6%) Placebo 43/238 138/937 181/1175 Mortality (18.1%) (14.7%) (15.4%) Relative Risk 1.05 1.48 1.39 P-value .83 .0004 .0009
21	SURROGATES Can delay use of effective treatments Example: Beta blockers in congestive heart failure; betablockers known to lower blood pressure slow heart rate Beta blocker drugs not used for heart failure for a decade or more
22	Beta-Blockers in Heart Failure Then four trials Metoprolol in MERIT [JAMA 283(10):1295-1302, 2000] Bisoprolol in CIBIS-II [Lancet 353: 9-13, 1999] Carvedilol in COPERNICUS [New Engl J Med 334(22):1651-58, 2001] Bucindolol in BEST [Am J Cardiol 1995;75:1220-3] Class II-IV heart failure, low ejection fraction patients Demonstrated beneficial effects
23	MERIT
24	Intermediate/Surrogate Outcomes Reliance on an intermediate outcome might lead to concluding harm when in fact intervention is beneficial
25	Diabetes Diabetes affects several organ systems (heart, kidney, eyes) Long duration causes visual impairment (diabetic retinopathy) Clinical Outcome Blindness Severe visual loss Surrogate Microaneurysm (retinal small vessel deformity filled with blood)
26	DCCT (Diabetes Complication and Control Trial) (NEJM, 1994) Hypothesis Does tight control of glucose reduce visual impairment compared to normal control? Design Tight control achieved by intense monitoring of an insulin pump Randomized multicenter trial 1441 diabetic patients Followed for average of 6 years
27	DCCT (Diabetes Complication and Control Trial) Results Early trends for microaneurysm were in negative direction, could perhaps have led to termination if the primary outcome Longer term follow-up showed definite reduction in visual impairment Trial terminated early for benefit
28	Concluding Remarks on Surrogates Surrogates play an important role in the development of Phase I, II, and pilot Phase III studies Treatments may affect more than one mechanism “Surrogates” do not reliably predict treatment on clinical outcome Problems seen in many disease areas Continued success in a given field is not even guaranteed
29	References Prentice RL: Surrogate endpoints in clinical trials: Definition and operational criteria. Statistics in Medicine 8:431-440, 1989 Temple RJ: A regulatory authority's opinion about surrogate endpoints. In: "Clinical Measurement in Drug Evaluation" (Ed. WS Nimmo, GT Tucker). John Wiley & Sons Ltd., 1996. Fleming TR and DeMets DL: Surrogate endpoints in clinical trials: Are we being misled? Annals of Int Med 125(7):605-613, 1996