Katherine R. Tuttle, MD, FASN, FACP

1
View from the NKF-KDOQI Diabetes and Chronic
Kidney Disease Work Group
Albuminuria as a Surrogate Outcome in Diabetic
Kidney Disease Pitfalls and Opportunities

• Katherine R. Tuttle, MD, FASN, FACP
• Medical and Scientific Director
• Providence Medical Research Center
• Clinical Professor of Medicine
• Division of Nephrology
• University of Washington School of Medicine
• Spokane and Seattle, Washington
• USA

2
Historical Perspective on Microalbuminuria as a
Predictor of Clinical Outcomes in Diabetes

• Early marker of diabetic kidney disease (DKD) in
  type 1 diabetes
• Predictor of cardiovascular disease (CVD)
  mortality in type 2 diabetes
• Death rate increased 100-150
• Most deaths were due to CVD causes

Mogensen CE. N Engl J Med 1984310356-60
3
Natural History of Diabetic Kidney Disease
Onset of Hyperglycemia DIABETES
High GFR
Low GFR
Normal GFR
Glomerulosclerosis and Tubulointerstitial Fibrosis
Cellular Injury
Rising Blood Pressure
Hypertension
Microalbuminuria
Macroalbuminuria
Rising Blood Creatinine
End-Stage Kidney Disease
Cardiovascular Death
Diabetes
2
5
10
20
30
Years
4
Annual Rates of Kidney Disease Progression and
Death in Type 2 Diabetes (UKPDS)
No Kidney Disease
D E A T H
1.4 (1.3 to 1.5)
2.0 (1.9 to 2.2)
Microalbuminuria
0.1 (0.1 to 0.2)
0.1 (0.0 to 0.1
3.0 (2.6 to 3.4)
2.8 (2.5 to 3.2)
Macroalbuminuria
4.6 (3.6 to 5.7)
2.3 (1.5 to 3.0)
0.3 (0.1 to 0.4)
Elevated blood creatinine level or kidney
replacement therapy
19.2 (14.0 to 24.4)
Adler AI et al. Kidney Int 200361225-232
5
Risks of CVD Death, MI, and Stroke by Quartiles
of Albuminuria in Diabetes (LIFE)
4
Unadjusted hazard ratio
Adjusted hazard ratio
Hazard Ratio ( 95 CI)
3
2
1
0
lt1
1-3
3-12
gt12
Adjusted for LVH, Framingham risk, treatment
Baseline Quartiles of Albuminuria (mg/mmol)
Ibsen H et al. Diabetes Care 200629595-600
6
Structural Correlate Albuminuria and Severity
of Angiographic Coronary Artery Disease
50

40
31

Urinary Albumin to Creatinine Ratio (mg/g)
25
30
13
20
10
10
0
Absent
Mild
Moderate
Severe
Angiographic Severity Score
Tuttle KR et al. Am J Kidney Dis 199934918-925
7
Relationship of Albuminuria and Angiographic
Coronary Artery Disease by Diabetes Status

70
49
60
50
Urinary Albumin to Creatinine Ratio (mg/g)
23
40

22
30
20
9
10
0
Present
Absent
Present
Absent
Type 2 Diabetic Patients
Non-Diabetic Patients
Tuttle KR et al. Am J Kidney Dis 199934918-925
8
Degree of Overt Proteinuria Predicts Stroke and
CVD Event Rates in Type 2 Diabetes
A U-Prot lt 150 mg/L
B U-Prot 150300 mg/L
C U-Prot gt 300 mg/L
1
40
0.9
p lt0.001
30
A
0.8
Survival Free of CVD Mortality
B
20
0.7
Incidence ()
0.6
C
10
Overall between-group plt0.001
0.5
0
0
Stroke
Coronary events
0
10
20
30
40
50
60
70
80
90
Months
Miettinen H et al. Stroke 1996272033-2039
9
Pitfalls of Albuminuria as a Surrogate Outcome
Measurement, Analysis, Interpretation

• Intra-patient variability in albuminuria
  measurement is often large.
• Urinary albumin excretion can fluctuate
  considerably from day-to-day, a particular
  problem at the low-end range.
• Analytic approaches for albuminuria are not
  standardized.
• Relationships between albuminuria and glomerular
  structure are inconsistent.
• Increased levels of urinary albumin are not
  always present in DKD.
• Connection of albuminuria to systemic vascular
  disease is indirect.

10
How Does the Kidney Reflect Status of the
Circulation-at-Large? Glomerular Structure
Capillary Loop
Mesangial Cell
Podocyte
Endothelial Cell
Afferent Arteriole
Efferent Arteriole
Juxtaglomerular Apparatus
11
Albuminuria Response to ACE Inhibition Predicts
Endothelial and Non-Endothelial-Dependent
Vascular Reactivity in Diabetes
18
16.6
16
14
11.4
10.8
12
With vs. without Microalbuminuria
(MA) plt0.001 p0.011
10
Vasodilatory response ()
8
6
4.2
4
2
0
1
2
3
4
With MA
Without MA
With MA
Without MA
FMD
NDD
Flow-mediated dilation FMD Nitroglycerine-depende
nt dilation NDD
Jawa A. et al. J Clin Endo Metab 20069131-35
12
Pitfalls of Albuminuria as a Surrogate Outcome
Clinical Utility

• Transition between albuminuria categories
  (normo-, micro-, macro-) is not a clinical
  endpoint.
• Data relating albuminuria to chronic kidney
  disease (CKD) endpoints are limited to
  observational analyses primarily from studies of
  renin angiotensin system (RAS) inhibition in
  patients with type 2 diabetes and
  macroalbuminuria.
• Masking phenomenon?
• Applicability to other populations (type 1
  diabetes, earlier and later CKD stages, normal-
  or low-level albuminuria) or treatments (novel
  therapies)?
• Albuminuria per se has not been a treatment
  target in phase 3 trials.
• Blood pressure with RAS inhibition
• Glycemic control

13
Death, CKD, and CVD Events by Microalbuminuria
Status in Type 2 Diabetes (multi-factorial
approach)
40
Non-reduction
30
20
Cumulative incidence ()
10
gt50 reduction
0
0
2
4
6
8
10
Time (years)
Araki S et al. Diabetes 2007561727-1730
14
CVD Death, MI, and Stroke by Time-Varying
Albuminuria in Type 2 Diabetes (LIFE)
0.36
lt1 mg/mmol (n274, 408, 311) 1-3 mg/mmol (n255,
239, 250) 3-12 mg/mmol (n267, 230, 213) gt12
mg/mmol (n267, 174, 175)
0.24
Proportional Endpoint Rate
0.12
Baseline, years 2 and 4
0.00
10
20
30
40
50
60
70
Month
Ibsen H et al. Diabetes Care 200629595-600
15
Pitfalls of Albuminuria as a Surrogate Outcome
Missing other Prospects?

• Failure to reduce albuminuria/proteinuria does
  not necessarily preclude therapeutic benefit.
• Primary reliance on this marker could lead to
  missed prospects for other effective therapies
  that work through different pathways or
  mechanisms.

16
Reduced Protein Diet Decreased ESRD andDeath in
Type 1 Diabetic Kidney Disease
30
Usual Protein Diet (1.02 g/kg/d)
20
Cumulative Incidence of ESRD or Death ()
10
Reduced Protein Diet (0.89 g/kg/d)
0
0
1
3
2
4
Follow-up Time (Years)

• Stage 2 CKD (inferred)
• 90 on ACEI, good BP control
• No difference in albuminuria
• Independent of risk factors, CVD

Hansen HP et al. Kidney Int 200262220-228
17
Opportunities for Albuminuria as a Surrogate
Outcome Confirm Treatment Target

• Interventions that reduce albuminuria are
  promising as potential therapies for preventing
  or reducing complications of CKD and associated
  CVD.
• Observational associations raise a strong
  hypothesis that albuminuria reduction produces
  clinical benefits.
• An alternate explanation is that albuminuria
  reduction marks patients who are more responsive
  to treatment.
• Clinical trials of therapies targeting
  albuminuria reduction with clinical endpoints as
  primary outcomes are necessary to confirm
  efficacy and safety.

18
Opportunities for Albuminuria as a Surrogate
Outcome Identify Novel Therapies

• Novel therapies for DKD are urgently needed to
  reduce this devastating complication of the
  worldwide diabetes epidemic.
• PKC inhibitors, AGE inhibitors, anti-fibrotic
  agents
• Albuminuria, as well as emerging biomarkers,
  should be useful for screening potentially
  effective therapies.

19
Biomarker Discovery Key Clinical Points

• Biological plausibility
• Adjudicated clinical endpoints
• Doubling of blood creatinine, dialysis, kidney
  transplant, MI, stroke, death
• Verification by test performance characteristics
• True positive rate, false positive rate
• ROC curve analysis
• Generalizable to population of interest
• Validated in different clinical groups

20
Process for Connecting Protein Biomarker
Discovery with Rigorous Clinical Validation
Rifai N et al. Nature Biotechnol 200624971-983