Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Multicenter Study
. 2016 Feb;80(2):243-9.
doi: 10.1097/TA.0000000000000912.

TIMP2•IGFBP7 Biomarker Panel Accurately Predicts Acute Kidney Injury in High-Risk Surgical Patients

Free PMC article
Multicenter Study

TIMP2•IGFBP7 Biomarker Panel Accurately Predicts Acute Kidney Injury in High-Risk Surgical Patients

Kyle J Gunnerson et al. J Trauma Acute Care Surg. .
Free PMC article


Background: Acute kidney injury (AKI) is an important complication in surgical patients. Existing biomarkers and clinical prediction models underestimate the risk for developing AKI. We recently reported data from two trials of 728 and 408 critically ill adult patients in whom urinary TIMP2•IGFBP7 (NephroCheck, Astute Medical) was used to identify patients at risk of developing AKI. Here we report a preplanned analysis of surgical patients from both trials to assess whether urinary tissue inhibitor of metalloproteinase 2 (TIMP-2) and insulin-like growth factor-binding protein 7 (IGFBP7) accurately identify surgical patients at risk of developing AKI.

Study design: We enrolled adult surgical patients at risk for AKI who were admitted to one of 39 intensive care units across Europe and North America. The primary end point was moderate-severe AKI (equivalent to KDIGO [Kidney Disease Improving Global Outcomes] stages 2-3) within 12 hours of enrollment. Biomarker performance was assessed using the area under the receiver operating characteristic curve, integrated discrimination improvement, and category-free net reclassification improvement.

Results: A total of 375 patients were included in the final analysis of whom 35 (9%) developed moderate-severe AKI within 12 hours. The area under the receiver operating characteristic curve for [TIMP-2]•[IGFBP7] alone was 0.84 (95% confidence interval, 0.76-0.90; p < 0.0001). Biomarker performance was robust in sensitivity analysis across predefined subgroups (urgency and type of surgery).

Conclusion: For postoperative surgical intensive care unit patients, a single urinary TIMP2•IGFBP7 test accurately identified patients at risk for developing AKI within the ensuing 12 hours and its inclusion in clinical risk prediction models significantly enhances their performance.

Level of evidence: Prognostic study, level I.


Figure 1
Figure 1
Study design and number of patients in cohorts. AKI was defined as KDIGO AKI stage 2 or 3 for Sapphire and was determined by clinical adjudication for Topaz (based on KDIGO stage 2–3 AKI).
Figure 2
Figure 2
TIMP2•IGFBP7 biomarker performance in surgical subsets. (A) TIMP2•IGFBP7 levels by surgical subgroup and AKI status within 12 hours and (B) AUC by surgical subgroup. Two mutually exclusive subgroup pairs are shown: Noncardiothoracic versus Cardiothoracic and Elective versus Emergent. Boxes and whiskers show interquartile ranges and total observed ranges, censored at 1.5 times the interquartile ranges. Patients with AKI had significantly higher levels of TIMP2•IGFBP7 than patients without AKI for all surgical patients and within each subgroup shown (Wilcoxon rank-sum test adjusted p < 0.001 in all cases). All AUC values were approximately 0.8 or greater and significantly greater than 0.5 (p < 0.001).
Figure 3
Figure 3
ROC curves and odds ratios from a multivariate clinical model alone and the model with TIMP2•IGFBP7. Stepwise model selection was used to derive the clinical model starting from all variables from Table 1 with p < 0.1 for the end point. All patients with a TIMP2•IGFBP7 value and data for all clinical variables were included (n = 353). The AUC-ROC increases (one-sided p = 0.008) from 0.77 (0.69–0.86) to 0.88 (0.83–0.94) when TIMP2•IGFBP7 is added to the model. *Log10 transform of APACHE III and TIMP2•IGFBP7 were used in the models. Log2 transform of serum creatinine was used because log10 is not a clinically relevant scale for serum creatinine. Blood for serum creatinine testing was collected at the time of urine collection for TIMP2•IGFBP7 testing. §History of cirrhosis or hepatic failure. The inverse of body mass index was used in the model. All continuous variables were standardized by subtracting the mean and then dividing by 2 standard deviations.

Similar articles

See all similar articles

Cited by 19 articles

See all "Cited by" articles


    1. Bellomo R, Kellum JA, Ronco C. Acute kidney injury. Lancet. 2012; 380 (9843): 756– 766. - PubMed
    1. Kellum JA, Bellomo R, Ronco C. Kidney attack. JAMA. 2012; 307 (21): 2265– 2266. - PubMed
    1. Bihorac A, Brennan M, Ozrazgat-Baslanti T, Bozorgmehri S, Efron PA, Moore FA, Segal MS, Hobson CE. National surgical quality improvement program underestimates the risk associated with mild and moderate postoperative acute kidney injury. Crit Care Med. 2013; 41 (11): 2570– 2583. - PMC - PubMed
    1. Hobson C, Ozrazgat-Baslanti T, Kuxhausen A, Thottakkara P, Efron PA, Moore FA, Moldawer LL, Segal MS, Bihorac A. Cost and mortality associated with postoperative acute kidney injury. Ann Surg. 2015; 261: 1207– 1214. - PMC - PubMed
    1. Bihorac A, Yavas S, Subbiah S, Hobson CE, Schold JD, Gabrielli A, Layon AJ, Segal MS. Long-term risk of mortality and acute kidney injury during hospitalization after major surgery. Ann Surg. 2009; 249 (5): 851– 858. - PubMed

Publication types