| Background and object:Approximately 2 million cardiac surgeries are conducted each year around the world.AKI is a frequent complication of cardiac surgery.In those patients with severe AKI who double their serum creatinine or need acute dialysis,there is a two-to five-fold higher risk of death.In addition to increased short term mortality,there is growing evidence that patients who survive an episode of AKI will have an increased risk of developing progressive CKD or even ESRD.Evidence is also mounting that the severity and frequency of AKI are major risk factors for subsequent development of CKD,suggesting that severe AKI may closely correlate with poor longterm outcome as well.The failure of previous clinical trials has,in part,been attributed to delays in diagnosis of AKI on the basis of rise in serum creatinine,which occurs several days after the initial injury.In this context,early detection of patients at high risk of severe AKI,before a detectable rise in serum creatinine,would afford clinicians a critical time window to halt or reverse the ongoing renal injury.The matrix metalloproteinases(MMPs)are a family of zinc-containing enzymes with proteolytic activity against a wide range of extracellular proteins.MMP-7 is one of the smallest secreted MMPs,predominantly localized in renal tubular epithelium,and can be easily excreted into urine.Little or no MMP-7 expression is detected in the normal kidney.However,its expression is markedly induced in human and animal models of CKD,which is primary controlled transcriptionally by ?-catenin,the principal downstream mediator of canonical Wnt signaling.We have previously shown that urinary matrix metalloproteinase-7(uMMP-7)levels faithfully reflect renal Wnt/?-catenin activity,and the signal pathway is activated in AKI induced by ischemia-reperfusion injury or renal toxicity.We therefore hypothesized that uMMP-7 level might be used as a noninvasive biomarker for early prediction of AKI after cardiac surgery.To test and validate the hypothesis,we conducted a prospective,multicenter,twostage cohort study in 721 patients undergoing cardiac surgery to evaluate the performance of postoperative measures of uMMP-7 compared with five other reported biomarkers for predicting severe AKI in patients after cardiac surgery.In the first stage,323 children receiving cardiac surgery were recruited from three academic medical centers.In the second stage,398 adults were enrolled at six centers.Methods and materials:This is a prospective,multicenter,two-stage cohort study.The stage 1 study(child cohort)was conducted in three academic medical centers across two cities from September of 2013 to September of 2014.Sample collection for the stage 2 study(adult cohort)was performed in six centers across five cities from February of 2014 to December of 2015.The patients in both study stages were enrolled according to the same inclusion and exclusion criteria,except for age.Eligible participants were patients receiving elective cardiac surgery aged 1 month to 18 years old(in stage 1)or 19-80 years old(in stage 2).Exclusion criteria included exposure to nephrotoxin(i.e.,contrast media,aminoglycoside antibiotics,vancomycin,and nonsteroidal anti-inflammatory drugs except aspirin)within 4 weeks before surgery,preexisting advanced CKD(chronic dialysis,renal transplantation,or preoperative eGFR 30 ml/min per 1.73 m2),and urinary tract infection or obstruction.We collected spot urine and blood samples before operation and at frequent intervals for 5 days after surgery.Urine samples were collected every 2 hours for the first 12 hours after CPB,and then every 12 hours for the first 5 days after surgery.Blood samples were obtained at 2 and 12 hours after surgery,and then every 24 hours for 5 days after surgery.When the CPB time exceeded 2 hours,the first postoperative urine sample was obtained at the end of CPB,and this sample was regarded as the 2-hour sample.The urine samples were centrifuged at 3000 g for 10 minutes and the supernatants were stored at-80?.Serum creatinine was measured to determined changes in renal function.Urine and blood samples collected from the participating hospitals were shipped by commercial cold chain transportation.All of the biomarkers were measured in a central laboratory using a standard protocol,and all of the samples were labeled using study identification numbers without personal identifi ers or clinical conditions.Urinary and plasma MMP-7,uAGT,uNGAL,u[TIMP-2]*[IGFBP7]and uIL-18 levels were measured by ELISA Kit.The NEPHROCHECK for TIMP-2 and IGFBP-7 measuring approved by the US FDA was not available in China.All biomarkers were measured in triplicate.Investigator calculated intra-and interassay variability ranged between 2%-6%and 5%-9%on the basis of blinded replicate samples from study patients.Urinary albumin was measured using an automatic analyzor.Urinary and serum creatinine was measured using an automatic biochemical analyzer.All of the urinary biomarkers were normalized for urinary creatinine and ex-pressed as micrograms per gram of creatinine or nanograms per gram of creatinine.Outcome Definitions:The primary outcome was the development of severe AKI,defined as the peak level of serum creatinine after surgery being double that of the preoperative level,or as receiving acute dialysis(consistent with KDIGO stage 2 or 3 AKI).Mild AKI was defined as an increase in serum creatinine by 26.5 mmol/L(0.3 mg/dl)within 48 hours or a 50%increase in serum creatinine from the preoperative level within 7 days(KDIGO criteria).We did not use urine output criteria(>0.5 ml/kg per hour for 6 hours)for AKI diagnosis because an indwelling urinary catheter was not present in the majority of participants.The secondary outcome was the development of composite events,including severe AKI,acute dialysis,and inhospital death.We also used length of stay in the ICU and in hospital as the inhospital outcomes.Statistical Analyses:SPSS 17.0 software was used for all analyses.To compare continuous variables,we used the ANOVA test or the Kruskal-Wallis test.To compare categorical variables,we used the chi-squared test.For an overall comparison of change over time between groups,a repeated measures ANOVA was performed,with particular focus on the in-teraction between group and time.We categorized the uMMP-7 quintile essentially as a continuous variable and then performed logistic regression on created variables.We determined the adjusted ORs of severe AKI with multiple logistic regression analysis with random intercepts for each center.The selection of covariates for adjusted models was on the basis of known risk factors that predict severe AKI in children after cardiac surgery,including age,preoperative eGFR,preoperative UACR,risk adjustment for congenital heart surgery 1 score ?3,and CPB time>120 minutes.In adults undergoing cardiac surgery,selected risk factors for severe AKI included age,sex,diabetes,hypertension,preoperative eGFR,preoperative UACR,preoperative serum albumin,and CPB time>120 minutes.In these analyses,the mean of the uMMP-7 level within the first 6 hours post CPB was modeled both as a categorical variable(categorized into quintiles)and a continuous variable(log transformed).To compare the performance of uMMP-7 and other biomarkers at different cutoff value,an AUC was generated.To evaluate the utility of the biomarkers on risk reclassifi cation,we determined the category free NRI and IDI,as previously described.Results1.Cohort DescriptionThere were 721 patients receiving cardiac surgery included in this study.In stage 1,323 children aged between 1 month and 18 years were enrolled from three academic medical centers,and in stage 2,398 adults were included from six centers.All surgeries were elective and used cardiopulmonary bypass(CPB).One hundred and twenty six children(39.0%)and 164 adults(41.2%)developed AKI according to Kidney Disease Improving Global Outcomes(KDIGO)criteria.Among these patients,53(16.4%)children and 50(12.6%)adults developed severe AKI,defined as doubling of serum creatinine or a need for dialysis(consistent with KDIGO stage 2 or 3).Severe AKI occurred at a median of 24(interquartile range[IQR],24-48)hours after surgery in both children and adults.Children who developed severe AKI were younger,had longer CPB times,and higher surgery complexity scores.Adult patients with severe AKI were older,more likely to have a lower preoperative eGFR,a history of CKD or hypertension,and longer CPB times.The pediatric cohort had no preexisting comorbilities,such as diabetes and hypertension,and less preoperative medications compared with the adult cohort.2.uMMP-7 as a Predictor for Development of Severe AKI after Cardiac SurgeryUMMP-7 levels were measured every 2 hours for the first 12 hours after CPB and then every 12 hours for 5 days.Preoperative uMMP-7 levels were not significantly different between patients who did and did not develop AKI in both children and adults.However,the average levels of uMMP-7 within the first 6 hours were significantly higher in patients who sub-sequently developed severe AKI compared with those who developed mild AKI or had no AKI.The uMMP-7 level pattern in those with severe AKI was characterized by a peak within 6 hours after surgery and remained significantly elevated until 48 hours in both cohorts,whereas the peak in serum creatinine rise occurred after 24 hours of surgery.UMMP-7 increased in all individuals after cardiac surgery,but the magnitude of these changes over time(within the first 48 hours)differed significantly across groups(interaction P 0.01 from repeated measures ANOVA in both children and adults).In the adult cohortwithAKI,elevation in uMMP-7 level was noted in both subgroups with and without preexisting CKD,but the level was significantly higher in those with prior CKD compared with those without.The best cutoff value of uMMP-7 for predicting severe AKI was significantly higher in patients with prior CKD(16.9 ?g/g creatinine)compared with those with preserved eGFR before surgery(7.6 ?g/g creatinine)Interestingly,patients who had higher uMMP-7 levels within the first 6 hours post-CPB tended to have unrecovered renal dysfunction at discharge.Unlike uMMP-7,plasma MMP-7 levels did not change significantly in patients undergoing cardiac surgery with or without AKI.There was no significant difference in plasma MMP-7 levels between children and adult uMMP-7 levels were associated with incident severe AKI in both pediatric and adult cohorts.After adjustment for center and for clinical variables(including age,sex,preoperative confounders,preoperative renal function,preoperative urinary albumin-to-creatinine ratio(UACR),CPB time,and surgery com plexityscores),higher quintiles of the postoperative uMMP-7 levels within the first 6 hours were independently associated with increasing risk of severe AKI in both children and adults.The highest quintile of mean uMMP-7 within the first 6 hours was associated with increased odds for severe AKI by 36-fold in children and 17 fold in adults compared with the lowest quintile Elevation of uMMP-7 level also predicted development of AKI(KDIGO stage 1-3)in both children and adults.When themean of the uMMP-7 levelwithin the first 6 hours post-CPB was analyzed as a continuous variable,higher uMMP-7 level was also associated with the development of severe AKI(children:adjusted odds ratio[OR]per SD,5.4;95%confidence interval[95%CI],3.1 to 9.3;P 0.001,adults:adjusted OR per SD,3.1;95%CI,2.0 to 4.8;P 0.001)in amultivar amultivariable model.When duration of AKI was evaluated as an outcome,higher uMMP-7 level was also associated with persistent AKI(AKI persisted>3 days)(children:adjusted OR per SD,3.9;95%CI,2.0 to 7.7;P 0.001,adults:adjusted OR per SD,3.7;95%CI,2.3 to 5.7;P 0.001).3.uMMP-7 as a Predictor for Outcomesafter Cardiac SurgeryOf 323 children in the stage 1 study,four(1.2%)received acute dialysis and nine(2.8%)died during hospitalization.In the stage 2 study,five(1.3%)adults receiveddialysis and eight(2.0%)adults died beforedischarge.The composite outcomes(severeAKI,acute dialysis,or inhospital death)were observed in 56(17.3%)patients in the pediatric cohort and 54(13.6%)patientsin adult cohort.The median lengths of intensive care unit(ICU)staywere 2 days(IQR,1-7)in children and 4 days(IQR,2-14)in adults.Median lengths of hospitalstay were 17 days(IQR,12-25)in childrenand 20 days(IQR,16-28)in adults.Aftermultivariable adjustment,higher uMMP-7 levels soon after surgery were significantly associated with increased risk of compositeevents and a longer stay in the ICU and in hospital in both children and adults(adjusted P for trend 0.001 in children and 0.001 in adults;).4.Comparing the Performance of uMMP-7 with Other BiomarkersFor predicting severe AKI,the area under the receiver operating characteristic curve(AUC)of uMMP-7 within 6 hours postsurgery were 0.81 in children and 0.76 in adults.The AUC of uMMP-7 for predicting severe AKI was the greatest among the tested biomarkers,such as urinary IL-18(uIL-18),urinary angiotensinogen(uAGT),urinary neutrophil gelatinase associated lipocalin(uNGAL),urinary tissue inhibitor of metalloproteinase-2*IGF-binding protein-7(u[TIMP-2]*[IGFBP7]),UACR,and the clinical risk factor model.Moreover,uMMP-7 outperformed these tested biomarkers in predicting composite events in children and adults with AUCs of 0.80 and 0.77,respectively.For predicting KDIGO stage 1-3 AKI,the AUCs of uMMP-7were 0.75 in children and 0.72 in adults,which also outperformed the other tested biomarkers(Supplemental.We also assessed the best time point at 2,4,and 6 hours for each biomarker and compared them.We found that the AUC of uMMP-7 remained the greatest compared with the other tested biomarkers.We further evaluated the performance of a combination of urinary biomarkers within the first 6 hours after surgery.The combination of uMMP-7 with uIL-18 yielded the largest AUCfor predicting severe AKI(0.83 in children and 0.79 in adults)and composite events(0.82 in children and 0.79 in adults).When these two biomarkers were added to the clinical riskmodel,the performance for predicting severeAKI(AUCof 0.91in children and 0.85 in adults)and composite events(AUC of 0.88 in children and 0.84 in adults)were further improved.5.Effect of uMMP-7 on Risk Reclassification of Severe AKIand Composite EventsAddition of uMMP-7,uIL-18,uAGT,uNGAL,UACR,and u[TIMP-2]*[IGFBP7]to the clinical risk model significantly improved risk reclassification over the clinical model alone,with uMMP-7 displaying the largest category-free net reclassificationimprovement(NRI)and integrated discrimination improvement(IDI)for both severe AKI and composite events in children and adults.ConclusionIn summary,uMMP-7 is a novel promising predictor for development of severe AKI and poor inhospital outcomes in patients after cardiac surgery.If further confirmed,early measurement of uMMP-7 level would help clinicians to identify those at high risk of severe postoperative AKI and help them to plan and initiate the appropriate management strategies for patients undergoing cardiac surgery. |
PDF Full Text Request