Abstract
Background
Children undergoing a cardiac surgical procedure are at increased risk for acute kidney injury (AKI). Novel biomarkers are needed to improve risk stratification of AKI after cardiac surgery.
Methods
We enrolled children aged 1 month to 18 years old from July 2007 to December 2010 undergoing cardiopulmonary bypass. Three United States Food and Drug Administration-approved plasma biomarkers of cardiac stretch, N-terminal pro B-type natriuretic peptide (NTproBNP), inflammation (ST2), or fibrosis (galectin-3), were measured preoperatively and postoperatively within 6 hours of cardiac surgery. All analyses were stratified by age (<2 or ≥2 years old) to account for changing biomarker distributions during childhood and due to a significant interaction between biomarker and age for galectin-3 and NTproBNP (P < .05).
Results
Postoperatively, AKI, defined by a doubling of baseline serum creatinine, was diagnosed in 51 of 194 children <2 years and in 28 of 201 children ≥2 years. After multivariable adjustment, for children <2 years, none of the biomarkers were independently associated with AKI, whereas for children ≥2 years, the highest tertile of preoperative galectin-3 and NTproBNP as well as the postoperative galectin-3 and ST2 were associated with AKI.
Conclusions
Preoperative plasma galectin-3 and NTproBNP and the first postoperative galectin-3 and ST2 levels were independently associated with AKI in children ≥2 years old. The performance of cardiac biomarkers after cardiac surgical procedure is affected by age, and research is required to develop biomarkers for children <2 years old.
A cute kidney injury (AKI) occurs postoperatively in up to 50% of children undergoing cardiac operations.1 Postoperative AKI is associated with increased inpatient mortality, length of ventilation, and hospital length of stay.1,2 Identifying the children who are at highest risk of AKI could guide clinical care, resource allocation, and clinical trial enrollment to prevent AKI.3
Biomarkers of tubular injury, myocardial injury, and inflammation have been shown to improve risk stratification of AKI in children.4–7 Plasma cardiac biomarkers may improve clinical risk prediction of AKI by noninvasively assessing myocardial injury, stretch, and fibrotic remodeling. Plasma N-terminal pro B-type natriuretic peptide (NTproBNP) is a developmentally regulated cardiac gene that is reexpressed by the ventricles of the heart in response to increased mechanical load and myocardial stretch. Plasma NTproBNP is used routinely in clinical practice for the diagnosis of acute and chronic heart failure. Preoperative and postoperative NTproBNP improves risk prediction of postoperative AKI in adult cardiac surgery cohorts. However, the association of NTproBNP with AKI was not confirmed in smaller pediatric cardiac surgery cohorts.6,8
ST2 is a secreted scavenger receptor for the ligand interleukin 33 and a member of the interleukin 1 receptor family. ST2 expression indicates cardiac strain, stretch, and inflammation, providing important prognostic information in multiple types of cardiac disease. Galectin-3, a β-galactosidase–binding lectin expressed ubiquitously, plays a role in cell-to-cell adhesion, macrophage activation, and in the pathophysiology of fibrosis in the kidney, liver, and heart.
To date, no data have been published on measurements of ST2 or galectin-3 in children undergoing cardiac surgery.9 ST2 and galectin-3 may indicate the severity of congenital heart disease or the degree of intraoperative cardiac injury, both of which are known risk factors of postoperative AKI. These biomarkers are not cardiac specific and may also reflect chronic noncardiac inflammation and fibrosis, pathologic processes that may increase risk for AKI.10 Plasma ST2, galectin-3, and NTproBNP are biomarkers approved by the United States Food and Drug Administration to assess the severity and prognosis of chronic heart failure.
We investigated the utility of the plasma cardiac injury biomarkers ST2, galectin-3, and NTproBNP for risk stratification of AKI in pediatric participants of the Translational Research Investigating Biomarker End-points in AKI (TRIBE-AKI) prospective multicenter pediatric cardiac surgery cohort. We hypothesized that blood biomarkers of cardiac stretch, function, and fibrosis will help to risk stratify for postoperative AKI after cardiac surgery.
Patients and Methods
Study Cohort
This was a prospective cohort study of children with congenital heart disease who were enrolled from July 2007 to December 2010 at Cincinnati Children’s Hospital, Montreal Children’s Hospital, and Yale New Haven Children’s Hospital. The study included children aged between 1 month and 18 years who were scheduled to be placed on cardiopulmonary bypass (CPB) during cardiac surgery. We excluded neonates, patients with a history of kidney transplantation, or prior receipt of dialysis. Written informed consent was obtained from all parents or legal guardians, along with assent when appropriate. The Institutional Review Board of each center approved the study.
Sample Collection
Patients were recruited during preoperative evaluations within 1 month of surgical procedure. Blood and urine specimens were collected before surgical procedure or with anesthesia induction. Postoperatively, blood was collected within 6 hours of intensive care unit (ICU) arrival (day 1 sample) and then daily for up to 5 days. Biospecimens were centrifuged, and urine supernatant and plasma were aliquoted. Aliquots of urine and plasma were stored at −80°C until biomarker measurement.
Data Collection
We used the Risk Adjustment for Congenital Heart Surgery-1 (RACHS-1) consensus-based scoring system to categorize the complexity of surgical procedure. Higher scores signify more complex surgical procedure.
Outcome Definitions
Postoperative AKI was defined as the development of at least stage II AKI, based on the AKI Network, as at least a 100% or greater rise from baseline serum creatinine during hospitalization after cardiac surgery. Age-adjusted percentiles for estimated glomerular filtration rate (eGFR) values were obtained based on previously published data on normal pediatric kidney function.11 Preoperative and postoperative serum creatinine levels were measured in the same clinical laboratory for each patient at all sites.
Biomarker Measurements
All plasma samples were continuously stored at −80°C with emergency back-up power and continuous electronic monitoring. Plasma biomarkers were measured preoperatively and postoperatively within 6 hours of ICU arrival (referred to as day 1 sample or first postoperative measurement). ST2, galectin-3, and NTproBNP concentrations were measured using the Meso Scale Discovery platform (Meso Scale Discovery, Gaithersburg, MD) at a median of 7 years after collection. To study the imprecision and variability of the biomarker measurements, we determined the intraassay and interassay coefficients of variation (CV). Intraassay CVs test the variability of biomarker measurements performed in the same sample on the same assay plate. Interassay CVs test the variability of biomarker measurements performed in the same sample on different assay plates and are often used to measure long-term imprecision. When measuring biomarkers with multiplex assays, as we did in this study, CVs of less than 15% are generally targeted and CVs of less than 5% are considered excellent. All intraassay CVs were less than 10%. Interassay CVs for galectin-3, NTproBNP, and ST2 were 19.3%, 21.5%, and 12.7%, respectively. ST2, galectin-3, and NTproBNP were selected as candidate cardiac biomarkers for risk stratification of kidney outcomes because they are all Food and Drug Administration-approved cardiac biomarkers. Plasma neutrophil gelatinase–associated lipocalin (NGAL) was measured using an assay (ARCHITECT; Abbott Diagnostics, Abbott Park, IL) with intraassay and interassay CVs of 5% and 8%, respectively.12 Interleukin 8 (IL-8) was measured using the MesoScale Discovery multiplex platform with intraassay and interassay CVs of less than 10%.5,7
Statistical Analysis
Continuous variables were compared with the 2-sample t test or the Wilcoxon rank sum test and dichotomous variables with the χ2 test or the Fisher exact test. Mixed-effects models were used to determine the effect of storage time on biomarker levels (Supplemental Table 1 and Supplemental Figures 1a, 1b, 1c).13 Because 95% of AKI occurred in the first 2 postoperative days, we evaluated associations between preoperative and first postoperative biomarker measurements with AKI.1 From our prior work on other biomarkers in this cohort and the understanding that kidney function continues to mature up to 2 years of age, we suspected significant differences in biomarker concentrations for children <2 vs ≥2 years old.5,12 We used the continuous natural log-transformed level of the biomarker in models looking at the interactions between biomarker levels and age <2 vs ≥2 years old, because this provided the most power for the evaluation of effect modification. We found that children <2 vs ≥2 years old had different strengths of association between preoperative galectin-3 and preoperative NTproBNP with AKI (P < .05). As such, we stratified analyses by age <2 vs ≥2 years old. Tertile threshold cut points were defined in the entire cohort and then applied to each age subgroup. We used multivariable logistic regression analysis to evaluate the association between tertiles of each preoperative and first postoperative biomarkers with AKI, determining adjusted odds ratios (aOR) with 95% confidence intervals (CIs).
The preoperative biomarkers logistic regression models and the preoperative clinic model included age, sex, race, RACHS-1, preoperative GFR percentile for age, and study site. Postoperative biomarker analyses logistic regression models and the postoperative clinical model also included CPB time.
The area under the receiver operating characteristic curve (AUC) was calculated from multivariable logistic regression models (C statistic) to determine the ability of the cardiac biomarkers to discriminate between patients with vs without AKI. Because plasma NGAL and IL-8 were shown to be the best performing plasma biomarkers of tubular injury and inflammation in the TRIBE-AKI pediatric cohort, we paired each cardiac biomarker with plasma NGAL or IL-8.5,12 We determined the AUC for the pair of cardiac and inflammatory biomarkers or a cardiac and kidney injury biomarker to see whether a pair of biomarkers could increase discrimination for AKI at specific time points. SAS 9.4 software (SAS Institute Inc, Cary, NC) was used for analyses.
Results
Of the 395 children who were enrolled in the cohort, 194 were <2 years old and 201 were ≥2 years old. AKI was diagnosed postoperatively in 79 children, (20%) of whom 51 were <2 years old and 28 were ≥2 years old (Supplemental Figure 2). Table 1 summarizes the demographic and perioperative characteristics of participants with and without AKI stratified by age. Children with AKI were more likely to be younger, had longer CPB times, and had longer ICU and hospital lengths of stay.
Table 1.
Baseline and Postoperative Characteristics by Acute Kidney Injury Status
| Age <2 Years Old (n = 194) |
Age ≥2 Years Old (n = 201) |
|||
|---|---|---|---|---|
| Characteristics | No AKI (n = 143) | AKI (n = 51) | No AKI (n = 173) | AKI (n = 28) |
| Age at operation, y | 0.5 (0.3, 0.7) | 0.4 (0.3, 0.7) | 5.3 (3.9, 9.3) | 4.2 (3.1, 8.0) |
| Female sex | 63 (44) | 26 (51) | 77 (45) | 15 (54) |
| White | 117 (82) | 39 (76) | 144 (83) | 25 (89) |
| CPB time >120 min | 39 (27)a | 27 (53)a | 42 (24) | 14 (50) |
| RACHS category ≥3 | 55 (38) | 19 (37) | 92 (53) | 17 (61) |
| eGFR pre-op percentile | 84 (31, 88)a | 87 (72, 99)a | 54 (24, 79)a | 84.5 (42, 97)a |
| Pre-op serum creatinine, mg/dL | 0.3 (0.3, 0.4)a | 0.3 (0.2, 0.3)a | 0.5 (0.4, 0.6)a | 0.3 (0.25, 0.5)a |
| Nonelective operation | 127 (89) | 41 (80) | 165 (95) | 27 (96) |
| Cross-clamp time, min | 54 (27, 75)a | 74 (46, 97)a | 26 (0, 49.5) | 0 (0, 45.5) |
| Length of stay | ||||
| Intensive care unit, d | 3 (2, 4)a | 5 (3, 8)a | 1 (1, 2)a | 3 (1, 7)a |
| Hospital, d | 5 (4, 9)a | 10 (7, 16)a | 4 (3, 7)a | 7 (5, 14.5)a |
| Days on ventilator, No. | 1 (1, 2)a | 2 (2, 5)a | 1 (0, 1)a | 2 (1, 4)a |
P< .05 when comparing AKI vs no AKI.
AKI, acute kidney injury; CPB, cardiopulmonary bypass; eGFR, estimated glomerular filtration rate; RACHS, Risk Adjustment for Congenital Heart Surgery.
Data are presented as the median (interquartile range) or as n (%).
Preoperative and Postoperative Biomarker Results by AKI Status
Figure 1 displays biomarker levels by AKI status at both preoperative and postoperative time points stratified by age <2 years and ≥2 years old. The first postoperative plasma ST2, galectin-3, and NTproBNP levels were significantly higher than the preoperative levels in both age groups.
Figure 1.
Biomarker distribution preoperatively (pre-op) and early postoperatively (post-op) of acute kidney injury (AKI) status stratified by age <2 years (left) and ≥2 years (right). (A) ST2, (B) galectin-3 (Gal3), and (C) N-terminal pro B-type natriuretic peptide (NTproBNP). The horizontal line in the middle of each box indicates the median; the top and bottom borders of the box mark the 75th and 25th percentiles, respectively, and the whiskers mark minimum and maximum of all the data.
In children <2 years, there was no statistically significant difference in preoperative biomarker levels in children with AKI vs without AKI (Supplemental Table 2). The first postoperative ST2 was significantly higher in children with AKI. In children ≥2 years old, the preoperative and first postoperative galectin-3 and NTproBNP concentrations were significantly higher in patients with AKI vs without AKI. For ST2, only the first postoperative levels were significantly higher in patients with AKI compared with those without AKI.
Association Between Preoperative Biomarkers and AKI
For children aged <2 years, there was no significant association between preoperative cardiac biomarkers and AKI, after adjusting for age, sex, race, study site, RACHS-1 category, and preoperative eGFR percentile (Table 2 and Supplemental Table 3). Among children aged ≥2 years old, the third tertile of preoperative galectin-3 and NTproBNP was significantly associated with a 4- and 6-fold higher odds (galectin-3: aOR, 4.2 [95% CI, 1.0–17.5]; NTproBNP: aOR, 6.7 [95% CI, 1.8–24.6], relative to the first tertile), respectively, of AKI after multivariable adjustment.
Table 2.
Association of Cardiac Biomarkers With Acute Kidney Injury Stratified by Agea
| Age <2 Years Old (n = 194) |
Age ≥2 Years Old (n = 201) |
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Biomarker | Time Point | Tertile | AKI events, No. (%) | Unadjusted Odds Ratio (95% CI) | P Value | Adjusted Oddsa Ratio (95% CI) | P Value | AKI events, No. (%) | Unadjusted Odds Ratio (95% CI) | P Value | Adjusted Oddsa Ratio (95% CI) | P Value |
| ST2 | Pre-op | 1 | 11 (22.45) | 1.0 [Reference] | 1.0 [Reference] | ... | 4 (6.67) | 1.0 [Reference] | ... | 1.0 [Reference] | ... | |
| 2 | 14 (30.43) | 1.51 (0.6–3.79) | .378 | 1.22 (0.45–3.27) | .694 | 13 (20.31) | 3.57 (1.09–11.65) | .035 | 3.41 (0.97–11.99) | .056 | ||
| 3 | 14 (22.95) | 1.03 (0.42–2.53) | .95 | 0.8 (0.3–2.15) | .66 | 7 (14.29) | 2.33 (0.64–8.49) | .199 | 2.06 (0.52–8.19) | .306 | ||
| First post-op | 1 | 6 (14.63) | 1.0 [Reference] | ... | 1.0 [Reference] | ... | 3 (4.17) | 1.0 [Reference] | ... | 1.0 [Reference] | ... | |
| 2 | 18 (29.51) | 2.44 (0.88–6.81) | .088 | 1.75 (0.56–5.48) | .838 | 8 (15.38) | 4.18 (1.05–16.61) | .042 | 2.31 (0.52–10.27) | .722 | ||
| 3 | 22 (36.07) | 3.29 (1.2–9.05) | .021 | 1.94 (0.61–6.15) | .952 | 14 (26.92) | 8.47 (2.29–31.34) | .001 | 4.85 (1.12–21) | .001 | ||
| Galectin-3 | Pre-op | 1 | 10 (19.23) | 1.0 [Reference] | ... | 1.0 [Reference] | ... | 3 (5.26) | 1.0 [Reference] | ... | 1.0 [Reference] | ... |
| 2 | 15 (32.61) | 2.03 (0.81–5.13) | .133 | 1.76 (0.66–4.72) | .26 | 10 (15.63) | 3.33 (0.87–12.78) | .079 | 2.59 (0.64–10.53) | .184 | ||
| 3 | 14 (24.14) | 1.34 (0.54–3.34) | .535 | 1.18 (0.45–3.06) | .74 | 11 (21.15) | 4.83 (1.27–18.44) | .021 | 4.23 (1.02–17.54) | .047 | ||
| First post-op | 1 | 12 (26.09) | 1.0 [Reference] | ... | 1.0 [Reference] | ... | 6 (8.96) | 1.0 [Reference] | ... | 1.0 [Reference] | ... | |
| 2 | 12 (21.82) | 0.79 (0.32–1.98) | .616 | 0.84 (0.28–2.49) | .468 | 6 (10.34) | 1.17 (0.36–3.86) | .793 | 0.72 (0.19–2.82) | .73 | ||
| 3 | 22 (35.48) | 1.56 (0.67–3.61) | .30 | 2.65 (0.95–7.42) | .133 | 13 (25.49) | 3.48 (1.22–9.92) | .02 | 6.81 (1.91–24.23) | .019 | ||
| NTproBNP | Pre-op | 1 | 8 (25.81) | 1.0 [Reference] | ... | 1.0 [Reference] | ... | 5 (6.41) | 1.0 [Reference] | ... | 1.0 [Reference] | ... |
| 2 | 10 (22.73) | 0.85 (0.29–2.46) | .759 | 0.78 (0.26–2.39) | .665 | 10 (15.15) | 2.61 (0.84–8.06) | .096 | 1.71 (0.51–5.67) | .382 | ||
| 3 | 21 (25.93) | 1.01 (0.39–2.59) | .99 | 0.71 (0.26–1.96) | .513 | 9 (31.03) | 6.57 (1.98–21.81) | .002 | 6.72 (1.84–24.6) | .004 | ||
| First post-op | 1 | 2 (11.11) | 1.0 [Reference] | ... | 1.0 [Reference] | ... | 9 (9.47) | 1.0 [Reference] | ... | 1.0 [Reference] | ... | |
| 2 | 17 (28.81) | 3.24 (0.67–15.63) | .144 | 2.6 (0.48–14.16) | .366 | 9 (16.67) | 1.91 (0.71–5.15) | .201 | 1.01 (0.33–3.14) | .71 | ||
| 3 | 27 (31.4) | 3.66 (0.79–17.06) | .098 | 2.28 (0.43–12) | .486 | 7 (25.93) | 3.34 (1.11–10.06) | .032 | 1.61 (0.43–6.05) | .702 | ||
Pre-op values are adjusted for age, sex, race, Risk Adjustment for Congenital Heart Surgery (RACHS) category ≥3, pre-op estimated glomerular filtration rate percentile, and site. First post-op values are adjusted for age, sex, race, cardiopulmonary bypass time >120 minutes, RACHS category ≥3, pre-op estimated glomerular filtration rate percentile, and site. Interaction P values between biomarker and age: ST2 pre-op, P = .088, post-op P = .651; galectin-3 pre-op P = .192, post-op P = .894; NTproBNP pre-op P = .065, post-op P = .808.
AKI, acute kidney injury; CI, confidence interval; No., number; NTproBNP, N-terminal pro B-type natriuretic peptide.
Bold values are statistically significant.
Association Between Postoperative Biomarkers and AKI
In children <2 years old, none of the biomarkers was associated with AKI after adjusting for age, sex, race, study site, RACHS-1 category, preoperative eGFR percentile, and CPB time (Table 2). In children ≥2 years old, the third tertile of the first postoperative ST2 and galectin-3 were each associated with AKI compared with the first tertile (ST2: aOR, 4.9 [95% CI, 1.1–21.0]; galectin-3: aOR, 6.8 [95% CI, 1.9–24.2], respectively). The results with first postoperative ST2 and galectin-3 were also observed in the full cohort, as the third tertile of first postoperative ST2 and galectin-3 were each associated with AKI compared with the first tertile (ST2: aOR, 2.8 [95% CI, 1.1–7.1]; galectin-3: aOR, 3.8 [95% CI, 1.7–8.5], respectively) (Supplemental Table 4).
Discrimination of Biomarkers for AKI
Table 3 and Supplemental Table 5 report the AUCs of the clinical model alone and the clinical model including an AKI biomarker. We also examined the combination of a cardiac biomarker at each time point with the best performing plasma biomarker of tubular injury (plasma NGAL) or inflammation (plasma IL-8) and determined their discrimination for AKI (Table 3). In children <2 years, the cardiac biomarkers did not increase discrimination for AKI. In children ≥2 years, the addition of both preoperative plasma NTproBNP and plasma NGAL to the clinical model increased the clinical model AUC from 0.75 to 0.84. The first postoperative plasma galectin-3 combined with IL-8 increased the AUC of the clinical model from 0.86 to 0.95. Finally, the addition of both the first postoperative plasma ST2 and NGAL increased the clinical model AUC from 0.86 to 0.95.
Table 3.
Discrimination of Acute Kidney Injury by Biomarkers for Age ≥2 Yearsa
| AUC (SE) |
|||||||
|---|---|---|---|---|---|---|---|
| Biomarker (Log Transformed) | Time Point | Biomarker | Biomarker + Clinical Model | Biomarker + Plasma NGAL | Biomarker + Clinical Model + Plasma NGAL | Biomarker + Plasma Il-8 | Biomarker + Clinical Model + Plasma IL-8 |
| ST2 | Pre-op | 0.60 (0.06) | 0.74 (0.06) | 0.68 (0.07) | 0.78 (0.09) | 0.73 (0.06) | 0.79 (0.05) |
| First post-op | 0.77 (0.06) | 0.88 (0.03) | 0.87 (0.05) | 0.95 (0.03) | 0.77 (0.06) | 0.94 (0.03) | |
| Galectin-3 | Pre-op | 0.66 (0.06) | 0.78 (0.05) | 0.64 (0.09) | 0.79 (0.09) | 0.72 (0.05) | 0.80 (0.05) |
| First post-op | 0.63 (0.06) | 0.88 (0.03) | 0.73 (0.09) | 0.93 (0.04) | 0.72 (0.07) | 0.95 (0.03) | |
| NTproBNP | Pre-op | 0.65(0.06) | 0.77 (0.06) | 0.77 (0.07) | 0.84 (0.06) | 0.73 (0.06) | 0.80 (0.06) |
| First post-op | 0.67(0.06) | 0.86 (0.04) | 0.77 (0.08) | 0.92 (0.05) | 0.73 (0.07) | 0.94 (0.03) | |
Clinical model includes age, sex, race, Risk Adjustment for Congenital Heart Surgery (RACHS) category ≥3, pre-op estimated glomerular filtration rate percentile, site, and cardiopulmonary bypass time >120 minutes at post-op time points. Clinical model AUC (SE) for pre-op is 0.75 (0.06), post-op is 0.86 (0.04) for age ≥2. Clinical model + NGAL AUC (SE) for pre-op is 0.70 (0.08), post-op is 0.89 (0.05) for age ≥2. Clinical model + IL-8 AUC (SE) for pre-op is 0.79 (0.05), postop is 0.94 (0.03) for age ≥2. Clinical model + NGAL + IL-8 AUC (SE) for pre-op is 0.80 (0.08), post-op is 0.96 (0.02) for age ≥2.
AUC, area under the receiver operating characteristic curve; IL, interleukin; NGAL, neutrophil gelatinase–associated lipocalin; NTproBNP, N-terminal pro B-type natriuretic peptide.
Comment
In this multicenter prospective study of children undergoing cardiac surgery, we demonstrated that the preoperative measurements of galectin-3 and NTproBNP and the first postoperative measurements of galectin-3 and ST2 were associated with AKI only in children ≥2 years. The preoperative and first postoperative measurements of galectin-3 were associated with AKI in children ≥2 years, such that the third tertile had a 4- and 7-fold increased risk of AKI, respectively. In addition, the combination of postoperative plasma ST2 and NGAL increased discrimination for AKI beyond a clinical risk model.
Importantly, previous research on cardiac biomarkers in children did not stratify their analyses by age and did not account for the varying cardiac biomarker distributions throughout childhood.14 Considering age in pediatric research is essential because biomarker reference ranges in children dramatically change with age.5,15–17 This was consistent with our finding of a significant interaction between biomarker and age present for galectin-3 and NTproBNP. Previous studies have shown that biomarkers in the youngest children have weaker associations with AKI as well as distinct threshold cutoffs.18 We found that the cardiac biomarkers had weaker discrimination for AKI in children <2 years. The worse biomarker performance in children aged <2 years may be due to the variability introduced by varying rates of organ development and maturing kidney function.
Elevated levels of galectin-3 have been found to be associated with mortality in adults with acute and chronic heart failure.19,20 We found that in children ≥2 years old, higher levels of preoperative and the first postoperative galectin-3 were independently associated with AKI. We did not identify an association between galectin-3 and AKI in children <2 years old. The strong association between the preoperative and first postoperative galectin-3 and AKI suggests that galectin-3 may be used for either preoperative or early postoperative AKI risk stratification after cardiac surgery. As such, galectin-3, as a biomarker of kidney and myocardial fibrosis, may be identifying those at highest risk for kidney and heart dysfunction after cardiac surgery.
ST2 is an independent predictor of mortality in adults with heart failure and acute myocardial infarction.21,22 Preoperative ST2 levels have been shown to be predictive of postoperative AKI and mortality after coronary artery bypass graft surgery in adults.21 Although ST2 has not been measured before in children undergoing cardiac operations, Hauser and colleagues23 demonstrated that ST2 had poor performance for diagnosing heart failure in children.23 In children ≥2 years old, a higher first postoperative ST2 concentration was associated with an increased risk of AKI. The first postoperative ST2 had superior performance compared with NTproBNP for discrimination of AKI. It is also notable that postoperative ST2 had similar discrimination for AKI compared with biomarkers of tubular injury, such as urine NGAL or IL-18, shown in prior research.12,17 When ST2 was combined with plasma NGAL, a biomarker of tubular injury, the pair displayed increased discrimination for AKI. In our study, the first postoperative ST2 was strongly associated with AKI and is likely serving as a marker of postoperative cardiac strain, inflammation, and complexity of surgical procedure.
Limitations
This study has a number of limitations. First, we stratified our results by age, and therefore, the number of events in each age category was reduced. A larger number of events in each category would have improved our ability to assess the performance of these cardiac biomarkers. However, multiple studies have shown that biomarker performance and threshold cutoffs are very different throughout childhood. As such, age must be considered when developing new biomarkers in children.
Another limitation is that we did not have a sufficient sample size to include subgroup analyses by type of congenital heart disease. Cardiac biomarkers of stretch and fibrosis may have distinct performance in specific types of congenital heart disease.
We were unable to study the association between the biomarkers and the outcome of mortality because there were only 8 deaths in this cohort.
Although intraassay CVs were all less than 10%, the interassay CVs for galectin-3 and NTproBNP were higher, at 19.3% and 21.5%, respectively. The higher interassay CVs seen with these biomarker measurements highlight measurement imprecision as a consideration when studying novel biomarkers. Before being used in a clinical setting, these assays will need to be further optimized.
Finally, we excluded neonates from this study. Future studies should study cardiac biomarker performance in children <2 years old, including neonates who undergo cardiac surgery, because they often have severe cardiac disease and are at high risk for AKI.
The strengths of the study include its prospective multicenter design and the known time of AKI.
Conclusion
We have demonstrated that ST2, galectin-3, and NTproBNP rise after CPB in children. In addition, in children ≥2 years old, we have presented a novel finding, in that galectin-3 and NTproBNP can preoperatively predict AKI and may be a useful biomarker to plan the timing of surgical procedure. For children with a high risk of developing AKI after an elective cardiac operation, close monitoring of serum creatinine and urine output as well as avoidance of nephrotoxic medications is warranted. We also found that galectin-3 and ST2 measured within 6 hours of surgical procedure in children ≥2 years can be used for risk stratification of AKI. These findings suggest that preoperative and postoperative cardiac biomarkers may be useful to identify children at high risk for AKI and may help with the design of AKI prevention clinical trials. NTproBNP, in particular, is readily available in most clinical settings and could easily be measured before cardiac surgery for AKI risk stratification. Future studies should validate the use of these biomarkers and biomarker pairs in pediatric cardiac surgery and investigate the utility of these biomarkers to assess therapeutic response.
Supplementary Material
Acknowledgments
Dr Greenberg is supported by the National Institutes of Health (NIH) National Institute of Diabetes and Digestive and Kidney Diseases career development grant K08-DK-110536. Dr Parikh is supported by NIH National Heart, Lung and Blood Institute grant R01-HL-085757 to fund the TRIBE-AKI Consortium to study novel biomarkers of AKI in cardiac surgery. Drs Parikh and Devarajan are members of the Assessment, Serial Evaluation, and Subsequent Sequelae of Acute Kidney Injury (ASSESS-AKI) Consortium supported by NIH National Institute of Diabetes and Digestive and Kidney Diseases grant U01-DK-082185. Dr Devarajan reports being a coinventor on the neutrophil gelatinase–associated lipocalin patent.
Footnotes
The Supplemental Tables and Supplemental Figures can be viewed in the online version of this article [https://doi.org/10.1016/j.athoracsur.2020.03.010] on https://www.annalsthoracicsurgery.org.
References
- 1.Li S, Krawczeski CD, Zappitelli M, et al. Incidence, risk factors, and outcomes of acute kidney injury after pediatric cardiac surgery: a prospective multicenter study. Crit Care Med. 2011;39:1493–1499. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Sutherland SM, Ji J, Sheikhi FH, et al. AKI in hospitalized children: epidemiology and clinical associations in a national cohort. Clin J Am Soc Nephrol. 2013;8:1661–1669. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Basu RK, Chawla LS, Wheeler DS, Goldstein SL. Renal angina: an emerging paradigm to identify children at risk for acute kidney injury. Pediatr Nephrol. 2012;27:1067–1078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Greenberg JH, Whitlock R, Zhang WR, et al. Interleukin-6 and interleukin-10 as acute kidney injury biomarkers in pediatric cardiac surgery. Pediatr Nephrol. 2015;30:1519–1527. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.de Fontnouvelle CA, Greenberg JH, Thiessen-Philbrook HR, et al. Interleukin-8 and tumor necrosis factor predict acute kidney injury after pediatric cardiac surgery. Ann Thorac Surg. 2017;104:2072–2079. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Bucholz EM, Whitlock RP, Zappitelli M, et al. Cardiac biomarkers and acute kidney injury after cardiac surgery. Pediatrics. 2015;135:e945–e956. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Greenberg JH, Zappitelli M, Jia Y, et al. Biomarkers of AKI progression after pediatric cardiac surgery. J Am Soc Nephrol. 2018;29:1549–1556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Hornik CP, Krawczeski CD, Zappitelli M, et al. Serum brain natriuretic peptide and risk of acute kidney injury after cardiac operations in children. Ann Thorac Surg. 2014;97:2142–2147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Meeusen JW, Johnson JN, Gray A, et al. Soluble ST2 and galectin-3 in pediatric patients without heart failure. Clin Biochem. 2015;48:1337–1340. [DOI] [PubMed] [Google Scholar]
- 10.Alam ML, Katz R, Bellovich KA, et al. Soluble ST2 and galectin-3 and progression of CKD. Kidney Int Rep. 2019;4: 103–111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Piepsz A, Tondeur M, Ham H. Revisiting normal (51)Cr-ethylenediaminetetraacetic acid clearance values in children. Eur J Nucl Med Mol Imaging. 2006;33:1477–1482. [DOI] [PubMed] [Google Scholar]
- 12.Parikh CR, Devarajan P, Zappitelli M, et al. Postoperative biomarkers predict acute kidney injury and poor outcomes after pediatric cardiac surgery. J Am Soc Nephrol. 2011;22: 1737–1747. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Liu KD, Siew ED, Reeves WB, et al. Storage time and urine biomarker levels in the ASSESS-AKI Study. PLoS One. 2016;11:e0164832. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Cantinotti M, Walters HL, Crocetti M, Marotta M, Murzi B, Clerico A. BNP in children with congenital cardiac disease: is there now sufficient evidence for its routine use? Cardiol Young. 2015;25:424–437. [DOI] [PubMed] [Google Scholar]
- 15.Bennett MR, Nehus E, Haffner C, Ma Q, Devarajan P. Pediatric reference ranges for acute kidney injury biomarkers. Pediatr Nephrol. 2015;30:677–685. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Parikh CR, Coca SG, Thiessen-Philbrook H, et al. Postoperative biomarkers predict acute kidney injury and poor outcomes after adult cardiac surgery. J Am Soc Nephrol. 2011;22:1748–1757. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Parikh CR, Thiessen-Philbrook H, Garg AX, et al. Performance of kidney injury molecule-1 and liver fatty acid-binding protein and combined biomarkers of AKI after cardiac surgery. Clin J Am Soc Nephrol. 2013;8:1079–1088. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Greenberg JH, Parikh CR. Biomarkers for diagnosis and prognosis of AKI in children: one size does not fit all. Clin J Am Soc Nephrol. 2017;12:1551–1557. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.van Kimmenade RR, Januzzi JL Jr, Ellinor PT, et al. Utility of amino-terminal pro-brain natriuretic peptide, galectin-3, and apelin for the evaluation of patients with acute heart failure. J Am Coll Cardiol. 2006;48:1217–1224. [DOI] [PubMed] [Google Scholar]
- 20.Lok DJ, Van Der Meer P, de la Porte PW, et al. Prognostic value of galectin-3, a novel marker of fibrosis in patients with chronic heart failure: data from the DEAL-HF study. Clin Res Cardiol. 2010;99:323–328. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Shimpo M, Morrow DA, Weinberg EO, et al. Serum levels of the interleukin-1 receptor family member ST2 predict mortality and clinical outcome in acute myocardial infarction. Circulation. 2004;109:2186–2190. [DOI] [PubMed] [Google Scholar]
- 22.Weinberg EO, Shimpo M, Hurwitz S, Tominaga S, Rouleau JL, Lee RT. Identification of serum soluble ST2 receptor as a novel heart failure biomarker. Circulation. 2003;107:721–726. [DOI] [PubMed] [Google Scholar]
- 23.Hauser JA, Demyanets S, Rusai K, et al. Diagnostic performance and reference values of novel biomarkers of paediatric heart failure. Heart. 2016;102:1633–1639. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.