Table 3.

Sample size for hypothetical cardiopulmonary bypass time trial using various biomarkers as pharmacodynamic end points

	Event rate in control arm (CPB time > 120 min)	Minimal detectable relative risk	Required sample size	Reclassification % in control arm. AKI by biomarker versus clinical AKIb
				AKI > 50% or dialysis	AKI > 100% or dialysis
Clinical AKI defined by serum creatinine
AKI > 50% or dialysis	26%	0.5	320	…	…
AKI > 100% or dialysis	7%	0.57	1945	…	…
AKIN stage 1 or higher	49%	0.59	204	…	…
AKI defined by elevated biomarker levelsa
IL-18	26%	0.38	204	33%	23%
uNGAL	33%	0.24	95	38%	30%
KIM-1	24%	0.54	425	27%	24%
L-FABP	28%	0.25	119	33%	27%
pNGAL	26%	0.5	320	30%	21%
Urine albumin	18%	0.78	2736	25%	15%
IL-18 & uNGAL	21%	0.19	140	30%	18%

AKI, acute kidney injury; CPB, cardiopulmonary bypass; IL-18, interleukin-18; KIM-1, kidney injury molecule-1; L-FABP, liver-type fatty acid–binding protein; pNGAL, plasma neutrophil gelatinase–associated lipocalin; uNGAL, urine NGAL.

^aAKI based on biomarkers was defined as first postoperative biomarker value in the fifth quintile. Cut points for fifth quintile: IL-18, 60 pg/ml; uNGAL, 102 ng/ml; KIM-1, 1.19 ng/ml; L-FABP, 175 ng/ml; pNGAL, 293 ng/ml. Sample size calculation based on 2-group test with equal numbers in each group. Two-sided alpha of 5%, statistical power 0.8 for the reported relative risk.

^bReclassification percentage was calculated in the control arm as the percentage of individuals with improved reclassification when defining AKI by elevated biomarker levels compared to clinical AKI defined by elevation in serum creatinine. Improvements in reclassification occurred in the following 2 circumstances: (i) there was no clinical AKI by serum creatinine, but AKI by elevated biomarker levels, or (ii) there was clinical AKI but no AKI by biomarker elevation (see shaded cells in Supplementary Table S2).