- Email: [email protected]
Is Preoperative Endothelial Dysfunction a Potentially Modifiable Risk Factor for Renal Injury Associated With Noncardiac Surgery?
Is Preoperative Endothelial Dysfunction a Potentially Modifiable Risk Factor for Renal Injury Associated With Noncardiac Surgery? David R. McIlroy, MBBS, MClinEpi, FANZCA,*† Matthew T.V. Chan, MBBS, FANZCA,‡ Sophie K. Wallace, BSc, MPH,* Arvinder Grover, MBBS, FANZCA,* Emily G.Y. Koo, MBBS, FANZCA,‡ Jiajia Ma, MBBS,‡§ Joel A. Symons, MBBS, FANZCA,* and Paul S. Myles, MBBS, MPH, MD, FCARCSI, FANZCA, FRCA* Objectives: To determine whether preoperative endothelial dysfunction provides risk stratification for perioperative renal injury in patients undergoing noncardiac surgery. The relationship between perioperative renal injury and myocardial injury after noncardiac surgery (MINS) was explored secondarily. Design: An observational study. Setting: Two academic medical centers. Participants: A total of 218 patients scheduled to undergo intermediate or high-risk noncardiac surgery. Interventions: None. Measurements and Main Results: Endothelial dysfunction was identified preoperatively by a Reactive HyperemiaPeripheral Arterial Tonometry (RH-PAT) index. Renal injury was defined by peak delta serum creatinine (ΔSCr) or creatinine-based kidney disease: Improving global outcomes acute kidney injury (AKI) criteria within 7 days postoperatively. MINS was defined by peak troponin Z0.04 lg/L within 3 days postoperatively. AKI occurred in 22 patients (10.1%). Median RH-PAT index within the study cohort was 1.64 (range 1.03-4.96) and did not differ between patients with or without AKI. When adjusted for covariates, there
was no association between RH-PAT index and either AKI or peak ΔSCr. MINS occurred in 32 patients (14.7%) and was associated independently with the outcome of AKI (odds ratio [OR], 3.7; 95% confidence interval [CI], 1.2-10.8; p ¼ 0.02) and peak ΔSCr (β-regression coefficient 23; 95% CI, 9-37; p ¼ 0.002). Co-occurrence of AKI and MINS portended a marked increase in 30-day mortality (OR, 43; 95% CI, 6-305; p ¼ 0.001) and delayed time to discharge (hazard ratio, 0.27; 95% CI, 0.13-0.54; p ¼ 0.001). Conclusions: For patients undergoing noncardiac surgery, preoperative endothelial function assessed by noninvasive peripheral arterial tonometry was not associated with perioperative AKI. Perioperative renal injury was associated strongly with MINS, and this may represent a mechanism by which AKI increases adverse outcomes. Crown Copyright & 2015 Published by Elsevier Inc. All rights reserved.
P
options highlights the urgent need to identify potentially modifiable risk factors with the hope of reducing this burden of injury. Identification of end-organ interactions through which AKI may promote perioperative adverse events also might offer a potential target for therapeutic intervention. The vascular endothelium is an active participant in various forms of vascular disease, capable of releasing soluble factors, including nitric oxide, to induce vasodilation while also reducing the tendency toward platelet aggregation, white blood cell adhesion, and proliferation of vascular smooth muscle.5–8 Endothelial dysfunction in the coronary vasculature consistently has been demonstrated to predict adverse cardiac events.8–10 Although the etiology of AKI likely is multifactorial, endothelial dysfunction is thought to play an important mediator role in various models.11,12 However, it is unknown whether the presence of preoperative endothelial dysfunction represents an important and potentially modifiable risk factor for AKI in the perioperative context. Endothelial dysfunction has been reported in 433% of patients with or at high risk for coronary artery disease,13 and several interventions, including the use of statins, polyphenols, various antioxidant compounds, and even continuous positive airway pressure devices for patients with obstructive sleep apnea, have been suggested as potentially able to reverse such dysfunction.14–16 No diagnostic test specific for renal endothelial dysfunction has been described. However, the EndoPAT-2000 (Itamar Medical Ltd., Caesarea, Israel) is an automated noninvasive device capable of detecting endothelial dysfunction using peripheral arterial tonometry in response to reactive hyperemia (Reactive Hyperemia-Peripheral Arterial Tonometry [RH-PAT]
erioperative acute kidney injury (AKI) is a major health care burden associated with increased complications and mortality.1,2 However, mechanisms leading to adverse outcomes, particularly with lesser degrees of renal injury, remain obscure. Although AKI occurring after cardiac surgery has been the focus of extensive research, the risk factors for AKI occurring after noncardiac surgery are less well defined. Recent studies confirm that AKI after noncardiac surgery occurs commonly, with reported incidences ranging from 6% to 33%,1–4 and the lack of established preventive or therapeutic
From the *Department of Anaesthesia & Perioperative Medicine, Alfred Hospital and Monash University, Melbourne, Australia; †Department of Anesthesiology, College of Physicians and Surgeons of Columbia University, New York, NY; ‡Department of Anaesthesia & Intensive Care, Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, Hong Kong; and §Department of Anaesthesiology, Beijing Tiantan Hospital, Capital Medical University, Beijing, China. This work was jointly supported by funding from the Australian and New Zealand College of Anaesthestists (Project Grant 08/016); and the Health and Health Services Research Fund (07080421) and General Research Fund (461409), Research Grant Council of Hong Kong. There was no commercial funding for this study. Address reprint requests to: David R. McIlroy MBBS, MClinEpi, FANZCA, Department of Anaesthesia & Perioperative Medicine, Alfred Hospital, 55 Commercial Road, Melbourne, Victoria, 3004 Australia. E-mail: [email protected] Crown Copyright © 2015 Published by Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2015.05.116 1220
KEY WORDS: myocardial infarction, complications, renal complications, endothelium complications, noncardiac surgery
Journal of Cardiothoracic and Vascular Anesthesia, Vol 29, No 5 (October), 2015: pp 1220–1228
1221
PREOPERATIVE ENDOTHELIAL DYSFUNCTION
index).17 It has been approved by the FDA and has received CE mark. Although validated against endothelial dysfunction in the coronary vasculature, the generalized nature of endothelial dysfunction makes it reasonable to expect that the EndoPAT2000 also may reflect dysfunction of the renovascular endothelium. The authors have demonstrated the prognostic utility of the EndoPAT-2000 for identifying patients at risk for myocardial injury after noncardiac surgery (MINS),18 and the rapid and noninvasive nature of this testing makes it potentially applicable as a preoperative screening tool. Preoperative identification of AKI risk using an RH-PAT index would support endothelial dysfunction as a preoperative risk factor for AKI while also providing a potential target to facilitate meaningful risk reduction through preoperative optimization. Adding to data from an existing cohort previously reported with respect to myocardial injury,18 the authors sought to identify risk factors for perioperative AKI after noncardiac surgery, focusing on the prognostic utility of preoperative, noninvasively measured endothelial dysfunction while secondarily exploring the association between AKI and myocardial injury after surgery. METHODS
The study was carried out in 2 academic medical centers (Alfred Hospital, Melbourne, Australia, and Prince of Wales Hospital, Hong Kong, China). After obtaining approval from the research and ethics committees of each participating institution (Ethics Committee Project numbers 08/07, CRE2008.444 and CRE-2013.267) and written informed consent from all participants, the authors conducted an observational study in patients undergoing noncardiac surgical procedures. Patients aged 440 years and scheduled to undergo nonemergent surgery identified as intermediate or high risk for postoperative cardiac complications using American College of Cardiology/American Heart Association guidelines19 were enrolled. Patients requiring preoperative renal replacement therapy, identified as undergoing partial or complete nephrectomy, or with no postoperative serum creatinine (SCr) measurement within 7 days of surgery were excluded from analysis. In addition to routine preoperative evaluation and any additional testing performed at the discretion of the treating physician, all study patients underwent noninvasive endothelial function assessment using the automated EndoPAT-2000 device. The principles of peripheral arterial tonometry (PAT) in response to reactive hyperemia previously have been described in detail.20 In summary, proprietary technology is used to noninvasively measure the magnitude and dynamics of arteriolar tone changes in peripheral arterial beds. Similar to conventional pulse oximetry, a thimble-shaped pneumatic probe is placed on the tip of 1 finger on each hand where the volume of blood in the fingertip with each arterial pulsation is photoplethysmographically detected. After a brief period to establish baseline, a blood pressure cuff is inflated to suprasystemic pressure on 1 arm for approximately 5 minutes. After cuff deflation, the hyperemic response in the ipsilateral finger is evaluated, measuring the ratio of the pulsewave amplitude in this period to baseline pulsewave amplitude. This ratio is further normalized to the signal simultaneously obtained from
the contralateral arm, accounting for potential effects of systemic changes in vascular tone. Calculations are performed by the device software, providing a bedside quantitative assessment of peripheral endothelial function with the option for manual adjustment of calculation points as deemed appropriate. A previous study reported an RH-PAT index o1.35 to be a useful indicator of coronary artery endothelial dysfunction,17 although an RH-PAT index r1.22 previously has been identified and reported to have prognostic utility for perioperative MINS within the current study cohort.18 In addition to baseline demographics, data were collected on various comorbidities, including factors thought to potentially affect endothelial function, including hypercholesterolemia, use of 3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors, diabetes, and smoking status. The most recent SCr recorded up to and including the day before surgery was used as baseline, with preoperative estimated glomerular filtration rate (eGFR) calculated using the Chronic Kidney Disease Epidemiology Collaboration equation.21 Peak daily postoperative SCr was recorded on each of postoperative days 1 to 7 as available, with frequency of measurement and all clinical decisions at the discretion of the treating medical team. Serum troponin was measured daily for the first 3 postoperative days as previously described, with MINS defined as any troponin Z0.04 mg/L.18 Coprimary endpoints were the maximum delta serum creatinine (ΔSCr) occurring within 7 days postoperatively, as well as the incidence of AKI defined by recent creatinine-based Kidney Disease: Improving Global Outcomes guidelines.22 Additional outcomes included 30-day mortality and time to hospital discharge. Statistical Analysis Statistical analyses were performed using Stata 12 (StataCorp, College Station, TX). Continuous variables are presented as mean and standard deviation (SD) or median and interquartile range (IQR). Categorical variables are presented as counts and proportions. Logistic regression and Cox proportional hazards first determined the association between both AKI and maximum ΔSCr, with 30-day mortality and time to hospital discharge, confirming the clinical significance of these endpoints. Patients who died before discharge were assigned the longest observed time to discharge within the study cohort, ensuring that early in-hospital mortality did not spuriously reflect a favorable outcome. Univariate and multivariate logistic regression then determined the relationship between preoperative measures of endothelial dysfunction and postoperative AKI, adjusting for potentially confounding variables selected on the basis of existing data and biologic plausibility while limiting the number of variables included in any given model to avoid overfitting. A multivariate linear regression model identified factors associated with maximum ΔSCr within 7 days postoperatively. Variables were included in this starting model only if significant for the outcome of maximum ΔSCr at p o 0.10 on univariate analysis. Manual, backward, stepwise elimination proceeded, eliminating the least significant variable at each step. This procedure continued until likelihood testing confirmed either a parsimonious model or that all remaining explanatory variables were significant statistically.
1222
MCILROY ET AL
As a secondary analysis, postoperative MINS was substituted or added to these models as an explanatory variable, exploring the relationship between perioperative myocardial injury and renal injury. The authors further determined the association between AKI and MINS using MINS as the outcome of interest. Finally, the prognostic utility of these 2 surrogate outcomes were evaluated, both alone and in combination, for 30-day mortality and time to hospital discharge. Results are presented as odds ratios (ORs), hazard ratios (HRs), β-regression coefficients, and 95% confidence intervals (CI), with p o 0.05 considered significant. In this context, HR refers to the likelihood of having been discharged from hospital on any given day after surgery compared with a stated reference group, such that HR o1 indicates delayed time to discharge. In view of the post hoc nature of the analysis, a formal sample size calculation and power analysis were not undertaken. RESULTS
Between July 2007 and February 2010, the authors enrolled 249 patients into the study. After withdrawals and exclusions, 218 patients remained for analysis (Fig 1). Mean (SD) preoperative SCr was 89 (31) μmol/L, corresponding to an eGFR of 74 (21) mL/min/1.73m2. Mean (SD) peak ΔSCr within 7 days postoperatively was 9 (40) μmol/L. AKI occurred in 22 (10.1%) patients, with 14, 6, and 2 cases first identified on postoperative days 1, 2, and 3, respectively. Patients who developed AKI were older, had poorer preoperative renal function, received more intravenous fluid, and were more likely to receive packed red blood cells (RBCs) during surgery than patients who did not develop AKI (Table 1). AKI was associated with a marked increase in 30-day mortality (OR, 14.3; 95% CI, 3.0-68.9; p ¼ 0.001) and delayed time to hospital discharge (HR, 0.39; 95% CI, 0.24-0.62; p ¼ 0.001). Similarly, when assessed as a continuous variable, maximum ΔSCr within 7 days postoperatively also was associated with increased 30-day mortality (β-regression coefficient 1.02; 95% CI, 1.01-1.04; p ¼ 0.001) and delayed time to hospital discharge (HR, 0.9; 95% CI, 0.99-1.00; p ¼ 0.001). MINS occurred in 32 patients (14.7%) and was associated with an increase in both 30-day mortality (OR, 8.7; 95% CI, 1.9-41.0; p ¼ 0.01) and delayed time to hospital discharge (HR, 0.52; 95% CI, 0.35-0.76; p ¼ 0.001). RH-PAT Index and AKI Median RH-PAT index within the study cohort was 1.64 (range 1.03-4.96) and did not differ between patients with or without AKI. After adjusting for age, type of surgery, and intraoperative administration of packed RBCs, there remained no association between RH-PAT index and AKI. Despite sequentially substituting 2 separate binary thresholds for RHPAT index, both previously suggested as useful for identifying clinically relevant endothelial dysfunction,13,14 there remained no association between RH-PAT index measures of endothelial dysfunction and AKI. Further sensitivity analyses substituting preoperative eGFR in place of age, or a composite term to reflect surgical extent in place of packed RBC transfusion, in the multivariate model did not alter materially the relationship
between RH-PAT index and AKI, and no interaction was found between study site and RH-PAT index. In contrast, substituting an elevated troponin within 3 days postoperatively (MINS) into the multivariate model in place of RH-PAT, revealed a significant independent association between this index of myocardial injury and AKI (Table 2). RH-PAT Index and Maximal ΔSCr RH-PAT index was not associated with maximum ΔSCr over 7 days postoperatively regardless of whether RH-PAT index was analyzed as a continuous or binary variable (Table 3). However, the constructed multivariate model included multiple outliers with potential for influence, together with evidence of specification errors that may reflect either important missing variables or inclusion of irrelevant variables. When added to this model, MINS was associated independently with maximal ΔSCr (β-regression coefficient 23; 95% CI, 9-37; p ¼ 0.002), improving the adjusted R2 of the model from 0.20 to 0.24. Including only smaller postoperative troponin elevations (o0.3 mg/L) to reflect less severe myocardial injury did not meaningfully alter this relationship. In an exploratory analysis using MINS as the outcome of interest, both the occurrence of AKI (OR, 5.2; 95% CI, 2.013.5; p ¼ 0.001) and maximal ΔSCr (OR, 1.02 for every 1 μmol/L increase; 95% CI, 1.01-1.03; p ¼ 0.001) were associated strongly with MINS, although this was somewhat attenuated after adjusting for both age and a Revised (Lee) Cardiac Risk Index23 score (Table 4). There was a weak, albeit highly significant, correlation between maximal ΔSCr over 7 days postoperatively and peak postoperative troponin within 3 days postoperatively (Spearman’s ρ, 0.25; p ¼ 0.0002). AKI, MINS, and Outcome The presence of either AKI or MINS, occurring discordantly from the other, was associated with a tendency toward increased odds for 30-day mortality and significantly delayed time to hospital discharge (Table 5). However, the cooccurrence of both AKI and MINS was associated with a dramatic increase in odds for mortality (OR, 43; 95% CI, 6305; p ¼ 0.001) and delayed time to discharge (HR, 0.27; 95% CI, 0.13-0.54; p ¼ 0.001). Despite adjusting for age, type of surgery, and several markers of extent of surgery, the combination of AKI and MINS remained a potent predictor of increased time to discharge, appearing synergistically prognostic compared with either event in isolation. DISCUSSION
Key Findings Preoperative endothelial dysfunction identified by noninvasive PAT was not associated with AKI or the peak change in SCr measured over 7 days postoperatively in patients undergoing nonemergent noncardiac surgery. Perioperative AKI was associated strongly with MINS, although the temporal sequence and a mechanistic link remain uncertain. Nevertheless, the co-occurrence of these 2 events provides powerful prognostic information for adverse outcomes, identifying a group of patients in whom further study is required urgently to
1223
PREOPERATIVE ENDOTHELIAL DYSFUNCTION
Hong Kong
Melbourne
169 patients enrolled
80 patients enrolled
Consent withdrawn:1 patient
Consent withdrawn: 4 patients
Surgery cancelled: 1 patient Missed by investigators at time of surgery: 1 patient Surgery delayed beyond completion of study: 1 patient
165 patients completed study
76 patients completed study
No RH-PAT index available: 3 patients*
Preoperative RRT: 1 patient No postoperative SCr from POD1-7: 7 patients
157 patients with data for analysis
73 patients with data for analysis
12 patients undergoing partial or complete nephrectomy excluded
218 patients for final analysis
Fig 1. Study flow chart, by site. *Due to poor signal quality in 2 cases. In the third case, a postocclusion time shorter than the required minimum was recorded, without sufficient time to repeat the test prior to surgery. Abbreviations: POD postoperative day; RH-PAT, Reactive Hyperemia-Peripheral Arterial Tonometry; RRT, renal replacement therapy; SCr, serum creatinine.
understand and potentially mitigate the heightened perioperative risk associated with this form of cardiorenal syndrome. Interpreting the Lack of Association Between PAT and AKI A number of possible explanations should be considered when interpreting the lack of association between a PATderived measure of endothelial dysfunction and perioperative renal injury. The active role of endothelial dysfunction in vascular disease is well established,6–8 and multiple studies have indicated its role as a risk factor for adverse cardiovascular events, including in the perioperative period.9,10,18,24 Within the renal vasculature, endothelial cells are believed to play an important role in regulating blood flow to local tissue
beds, with endothelial injury promoting the adherence of inflammatory cells, increased permeability and edema, and reduced blood flow.25 Endothelial expression of molecules such as intercellular adhesion molecule (ICAM)-1 further upregulate the proinflammatory state,12 whereas dysfunction of the renal endothelium also is associated with a reduction in endothelial-dependent vasodilator activity.11 However, despite reported dysfunction of the renovascular endothelium in various models of AKI,26–28 this typically has been observed in the context of established or evolving injury, where it likely represents a mediator of AKI rather than a pre-existing and potentially modifiable risk factor. The EndoPAT-2000 has been validated as a marker of coronary endothelial dysfunction,17 a diagnosis confirmed by invasive coronary angiography with quantification of the response
1224
MCILROY ET AL
Table 1. Baseline Characteristics and Risk Factors AKI* No (n ¼ 196)
Baseline demographics Age Male sex BMI Clinical risk factors: Preoperative Hypertension Hypercholesterolemia Diabetes mellitus Current smoker NSAIDs within 2 days before surgery Diuretics HMG Co-A reductase inhibitors Aspirin within 5 days before surgery ACE inhibitors/ARBs Revised cardiac risk index Surgery-specific risk† Low Intermediate High Type of surgery General Orthopedic Vascular Urology/gynecology Major ENT/Plastics RH-PAT index‡ RH-PAT index o1.35 RH-PAT index r1.22 Risk factors: Intraoperative Intraoperative IV fluid§ (mL) Surgical duration (min) Packed RBCs intraoperatively (mL) Any packed RBCs intraoperatively, n (%) Biochemistry Preoperative SCr (μmol/L) Preoperative eGFR (mL/min/1.73m-2) Maximum delta SCr║ (μmol/L) Maximum postoperative SCr║ (μmol/L) Outcomes 30-days mortality Postoperative hospital LOS, days (IQR)
Yes (n ¼ 22)
p Value
68 (10) 124 (63.3) 25.3 (5.2)
75 (8) 18 (81.8) 25.3 (4.4)
0.001 0.08 0.99
158 100 74 23 6 32 58 52 72 1 6 114 76 109 6 41 36 4 1.61 42 14
(80.6) (51.0) (37.8) (11.7) (3.1) (16.3) (29.6) (26.5) (36.7) (1-2) (3.1) (58.1) (38.8) (55.6) (3.1) (20.9) (18.4) (2.0) (1.38-2.07) (21.4) (7.1)
21 13 11 7 0 3 13 8 9 2 0 9 13 6 0 11 5 0 1.70 3 3
(95.5) (59.1) (50.0) (31.8) (0.0) (13.6) (59.1) (36.4) (40.9) (1-3) (0.0) (40.9) (59.1) (27.3) (0.0) (50.0) (22.7) (0.0) (1.50-1.89) (13.6) (13.6)
0.14 0.47 0.26 0.02 0.99 0.99 0.01 0.33 0.70 0.005 0.21
2000 219 0 23
(1500-3000) (102) (0-0) (11.7)
3000 268 0 9
(2000-3500) (96) (0-500) (40.9)
0.01 0.03 0.0002 0.001
(92-112) (16) (36-113) (126-237)
0.001 0.001 0.0001 0.0001
82 75 1 80
(67-98) (21) (–8 to 9) (68-99)
3 (1.5) 8 (6-12)
98 61 49 153
4 (18.2) 11 (6-26)
0.03
0.73 0.58 0.39
0.002 0.13
Abbreviations: ACE, angiotensin-converting enzyme; AKI, acute kidney injury; ARB, angiotensin-receptor blocker; BMI, body mass index; eGFR, estimated glomerular filtration rate; ENT, ear, nose, throat; HMG Co-A, 3-hydroxy-3-methylglutaryl-coenzyme A; IQR, interquartile range; IV, intravenous; LOS, length of stay; NSAID, nonsteroidal anti-inflammatory drug; RH-PAT, Reactive Hyperemia–Peripheral Arterial Tonometry; RBC, red blood cell; SCr, serum creatinine. *AKI defined by Kidney Disease: Improving Global Outcomes Kidney Disease: Improving Global Outcomes creatinine-based criteria over 7 days postoperatively. †Surgery-specific risk assigned by the treating clinician with reference to American College of Cardiology/American Heart Association Guidelines on Perioperative Cardiovascular Evaluation and Care for Noncardiac Surgery. ‡Calculated by the EndoPAT device. §Defined as the sum of intraperative crystalloid and colloid administration. ║Defined within 7 days postoperatively.
to graded increases in direct acetylcholine administration. In contrast, no such gold standard exists to confirm the presence of endothelial dysfunction in the renal vasculature, and although it is held widely to be a generalized process occurring throughout the vasculature, the EndoPAT-2000 provides only an indirect measure and has not been evaluated specifically or validated to identify endothelial dysfunction within the renal vasculature. Moreover, a limited association between metrics of PAT and
flow-mediated vasodilation (FMD) when compared with each other29–31 suggests that quantification of endothelial dysfunction is complex and may not be captured adequately by a single measurement technique. Despite the automated and relative operator-independent nature of the EndoPAT device, it also remains uncertain to what extent variations in other baseline characteristics and risk factors, differences in patient activity before testing and environmental conditions at the time of testing, or unmeasured
1225
PREOPERATIVE ENDOTHELIAL DYSFUNCTION
Table 2. Unadjusted and Adjusted Odds Ratios for Outcome of AKI* After Noncardiac Surgery ‡
RH-PAT index RH-PAT index r1.22 RH-PAT index o1.35 Type of surgery General Orthopedic Vascular Urology/gynecology Major ENT/plastics Any intraoperative pRBCs Age (y) o65 65-74 Z75 Preoperative eGFR (mL/min/1.73 m-2) Preoperative eGFR o60 mL/min/1.73 m-2 Preoperative SCr (μmol/L) Female sex BMI Hypertension Hypercholesterolemia Diabetes mellitus Current smoker Aspirin (past 5 days) NSAID (past 2 days) ACEI or ARB Diuretic HMG CoA-reductase inhibitor Surgery specific risk§: Low Intermediate High Intraoperative pRBCs (per 100 mL) Study site Surgical duration (per 10 min) Intraoperative IV fluid║ (per 100 mL) Extensive surgery¶ Peak troponin Z0.04 mg/L by POD 3 Peak troponin Z0.04 mg/L and o0.30 mg/L by POD 3
Univariate OR (95% CI)
p Value
Multivariate† OR (95% CI)
p Value
1.0 (0.5-2.0) 2.1 (0.5-7.8) 0.6 (0.2-2.1) 1.00 (reference) N/A 4.9 (1.7-14.0) 2.5 (0.7-8.8) N/A 5.2 (2.0-13.5) 1.00 (reference) 4.2 (0.9-20.7) 7.7 (1.7-35.9) 0.97 (0.95-0.99) 2.0 (0.8-5.0) 1.01 (1.00-1.03) 0.4 (0.1-1.2) 1.00 (0.9-1.1) 5.1 (0.7-38.7) 1.4 (0.6-3.4) 1.6 (0.7-4.0) 3.5 (1.3-9.5) 1.6 (0.6-4.0) N/A 1.2 (0.5-2.9) 0.8 (0.2-2.9) 3.4 (1.4-8.5) N/A 1.00 (ref) 2.2 (0.9-5.3) 1.1 (1.0-1.3) 0.7 (0.3-1.6) 1.04 (1.00-1.08) 1.03 (1.00-1.06) 2.7 (1.1-6.6) 5.2 (2.0-13.5) 4.2 (1.3-13.2)
0.98 0.29 0.40 — — 0.003 0.15 — 0.001 — 0.07 0.01 0.003 0.13 0.02 0.09 0.99 0.12 0.47 0.27 0.01 0.33
1.0 (0.5-2.0) 1.6 (0.4-7.7) 0.5 (0.1-1.8) 1.00 (reference) — 3.5 (1.1-11.2) 2.2 (0.6-8.1) — 3.9 (1.4-11.1) 1.00 (reference) 4.7 (0.9-24.1) 5.0 (1.0-25.4) 0.98 (0.95-1.0)
0.94 0.53 0.26 — — 0.03 0.23 — 0.01
3.7 (1.2-10.8) 3.9 (1.0-15.1)
0.02 0.05
0.06 0.05 0.05
0.70 0.75 0.01 0.09 0.04 0.35 0.04 0.02 0.03 0.001 0.02
Abbreviations: ACEI, angiotensin-converting enzyme inhibitor; AKI, acute kidney injury; ARB, angiotensin-receptor blocker; BMI, body mass index; eGFR, estimated glomerular filtration rate; ENT, ear, nose, throat; HMG Co-A, 3-hydroxy-3-methylglutaryl-coenzyme A; IV, intravenous; NSAID, nonsteroidal anti-inflammatory drug; POD, postoperative day; pRBC, packed red blood cell; RH-PAT, Reactive Hyperemia–Peripheral Arterial Tonometry Index; SCr, serum creatinine. *Defined according to creatinine-based Kidney Disease: Improving Global Outcomes Kidney Disease: Improving Global Outcomes criteria. †Multivariate analysis included RH-PAT index adjusted for categorical variables of age, type of surgery, and intraoperative administration of pRBCs. Sensitivity analysis included substituting binary values for RH-PAT index according to previously published thresholds, substituting eGFR for age, and extensive surgery for administration of pRBCs. Finally, troponin terms were substituted into the model in place of RH-PAT index. For all models assessed, Hosmer-Lemeshow test 4 0.05. ‡Calculated by the EndoPAT device. §Surgery-specific risk assigned by the treating clinician with reference to American College of Cardiology/American Heart Association Guidelines on Perioperative Cardiovascular Evaluation and Care for Noncardiac Surgery. ║Defined as the sum of intraoperative crystalloid and colloid administration. ¶Defined as the composite of intravenous fluid 45000 mL, surgical duration 45 hours or any transfusion of pRBCs intraoperatively as a marker of surgical extent.
operator variability where manual adjustment of parameters for analysis was employed may have influenced results. However, the same device used in the same cohort recently demonstrated prognostic utility for myocardial injury after surgery.18 Although most of the data for the current study were collected prospectively, the frequency of measurement of SCr was at the discretion of the treating clinicians and may have introduced an unmeasurable degree of ascertainment bias. The
limited sample size and total number of events in the current study cohort raised the possibility of a type-II error and precludes a more detailed analysis of factors associated with AKI. However, despite the acknowledged limitations of this multivariate modeling process, the current study was not intended to build a robust and comprehensive model for the prediction of AKI after noncardiac surgery, and no meaningful conclusion should be drawn from these data about a potential
1226
MCILROY ET AL
Table 3. Unadjusted and Adjusted Regression Coefficients for the Outcome of Maximum ΔSCr* After Surgery Unadjusted β-regression Coefficient (95% CI)
Age (y) o65 65-74 Z75 Female sex BMI Hypertension Hypercholesterolemia Diabetes mellitus Current smoker Aspirin (within 5 days preoperatively) NSAID (within 2 days postoperatively) ACEI/ARB Diuretic HMG co-A reductase inhibitor Type of surgery General Orthopedic Vascular Urology/gynecology Major ENT/plastics Surgery-specific risk‡: Low Intermed High Intraoperative pRBCs (per 100 mL) Any intraoperative pRBCs Site RH-PAT index§ RH-PAT index r1.22 RH-PAT index o1.35 Surgical duration (per 10 min) Intraoperative IV fluid║ (per 100 mL) Preoperative SCr (mmol/L) Preoperative eGFR (mL/min/1.73 m-2) Preoperative eGFR o60 mL/min/1.73 m-2 Peak troponin Z0.04 mg/L by POD 3 Peak troponin Z0.04 mg/L and o0.30 mg/L-1 by POD 3
0.0 6 24 –10 1 14 6 3 24 3 –7 7 0 18
(reference) (–7 to 18) (11-37) (–22 to 1) (0-2) (0-28) (–4 to 17) (–8 to 14) (9-40) (–9 to 15) (–41 to 26) (–4 to 18) (–15 to 15) (7-29)
0.0 3 25 8 2 0 10 16 4 37 –8 0 20 1 1 1 0.2 –0.4 12 39 35
(reference) (–30 to 36) (12-38) (–6 to 22) (–37 to 42) (reference) (–23 to 43) (–17 to 50) (2-6) (22-51) (–20 to 3) (–9 to 9) (0-40) (–13 to 14) (0-1) (0-1) (0.0-0.4) (–0.6 to –0.1) (0-24) (25-53) (19-51)
Adjusted β-regression p Value
0.37 0.001 0.07 0.28 0.05 0.24 0.62 0.002 0.61 0.66 0.22 0.98 0.002
Coefficient† (95% CI)
p Value
0.0 (reference) 7 (–5 to 18) 21 (8-33)
0.28 0.001
25 (10-39)
0.001
17 (6-27)
0.002
31 (17-45)
0.001
23 (9-37) 24 (8-40)
0.002 0.003
0.86 0.001 0.26 0.90
0.56 0.34 0.001 0.001 0.16 0.99 0.06 0.94 0.02 0.003 0.01 0.004 0.04 0.001 0.001
Abbreviations: ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin-receptor blocker; BMI, body mass index; eGFR, estimated glomerular filtration rate; ENT, ear, nose, throat; IV, intravenous; HMG Co-A, 3-hydroxy-3-methylglutaryl-coenzyme A; IV, intravenous; NSAID, nonsteroidal anti-inflammatory drug; POD, postoperative day; pRBC, packed red blood cell; RH–PAT, Reactive Hyperemia–Peripheral Arterial Tonometry Index; SCr, serum creatinine. *Defined over 7 days postoperatively. †RH-PAT o1.22 was eliminated from the multivariate model. Subsequent addition of an elevated troponin into the remaining model increased the adjusted R2 from 0.20 to 0.24 with likelihood testing suggesting a better model (p ¼ 0.002). ‡Surgery-specific risk assigned by the treating clinician with reference to American College of Cardiology/American Heart Association Guidelines on Perioperative Cardiovascular Evaluation and Care for Noncardiac Surgery. §Calculated by the EndoPAT device. ║Defined as the sum of intraoperative crystalloid and colloid administration.
Table 4. Unadjusted and Adjusted Odds Ratios for the Outcome of MINS* After Noncardiac Surgery ‡
Any AKI Maximal ΔSCr§
Unadjusted OR (95% CI)
p Value
Adjusted† OR (95% CI)
p Value
5.2 (2.0-13.5) 1.02 (1.01-1.03)
0.001 0.001
3.1 (1.1-8.7) 1.02 (1.00-1.03)
0.03 0.01
Abbreviations: AKI, acute kidney injury; MINS, myocardial injury after noncardiac surgery; POD, postoperative day; SCr, serum creatinine. *Defined as troponin Z0.04 mg/L by POD 3. †Adjusted for categorical variables of age and Revised Cardiac Risk (Lee) Index Z2. ‡Defined by creatinine-based Kidney Disease: Improving Global Outcomes Kidney Disease: Improving Global Outcomes guidelines. §Odds ratio is for every 1 unit increase in maximal ΔSCr (mmol/L).
1227
PREOPERATIVE ENDOTHELIAL DYSFUNCTION
Table 5. Outcomes Associated With Postoperative AKI* and MINS† Status Odds ratio (95% CI) Outcome: 30-days mortality
All patients
AKI–/MINS–
AKIþ/MINS–
AKI–/MINSþ
AKIþ/MINSþ
1.0 (reference)
7.1 (0.6-84)
3.9 (0.3-45)
43 (6-305)
AKI–/MINS–
AKIþ/MINS–
AKI–/MINSþ
AKIþ/MINSþ
1.00 (reference) 1.00 (reference)
0.46 (0.25-0.82) 0.53 (0.28-0.99)
0.62 (0.40-0.97) 0.51 (0.32-0.81)
0.27 (0.13-0.54) 0.28 (0.13-0.60)
Hazard ratio (95% CI) Outcome: Time to hospital discharge Unadjusted HR Adjusted HR‡
Abbreviations: AKI, acute kidney injury; HR, hazard ratio; MINS, myocardial injury after noncardiac surgery. *Defined by creatinine-based Kidney Disease: Improving Global Outcomes Kidney Disease: Improving Global criteria over 7 days postoperatively. †Defined by peak postoperative troponin Z0.04 mg/L within 3 postoperative days. ‡Adjusted for surgical duration, total intraoperative volume of clear intravenous fluid, and categorical variables of age, type of surgery, and any intraoperative transfusion of packed red blood cells.
relationship among other risk factors, such as exposure to perioperative nonsteroidal anti-inflammatory drugs or angiotensinconverting enzyme I and renal injury. Rather, the data presented here provide important evidence against endothelial dysfunction measured by noninvasive PAT as a preoperative risk factor for AKI. Renal Injury and MINS The association among biomarkers of MINS and renal injury in the perioperative context is a novel finding. Although previous studies have demonstrated an increase in both short- and longerterm mortality when AKI complicates acute myocardial infarction (MI),32–34 an admission diagnosis of MI in these studies may reflect larger myocardial injury than typically identified in the perioperative period by routine troponin measurement (MINS). Although the pathophysiology of cardiorenal syndrome (CRS) remains poorly understood, its clinical significance increasingly is recognized. Well-described subtypes include an acute deterioration in cardiac function leading to AKI (CRS type I) or an acute deterioration in renal function leading to an acute cardiac disorder (CRS type III), highlighting the potential bidirectional nature of the syndrome.35 The pathophysiology of this bidirectional relationship has been described in terms of gross clinical interactions, such as AKI leading to congestive heart failure through oliguria, fluid overload, and uremia, or MI and heart failure leading to AKI through a process of reduced cardiac output and activation of the renin-angiotensin and sympathetic nervous systems, with subsequent reductions in renal blood flow and GFR.36 However, at a cellular level, experimental renal ischemia-reperfusion injury clearly induces a proinflammatory state with upregulation of various cytokines.12 Elevated levels of interluekin-1, tumor necrosis factor-α, and ICAM, as well as increased myeloperoxidase activity, all have been identified within the myocardium after isolated renal insult. In fact, even brief periods of renal ischemia insufficient to produce azotemia can induce cardiac apoptosis that appears to be mediated, at least
in part, by tumor necrosis factor -α.37 Similar humoral mechanisms of organ crosstalk have been postulated to promote AKI after cardiac injury.36 A mechanistic evaluation of the relationship between MINS and renal injury was beyond the scope of the current study, and a number of uncertainties remain. The pharmacokinetic effect of acute changes in renal function on circulating troponin levels was not well characterized, and the frequent mild elevations in postoperative troponin seen in the current study might simply reflect reduced renal clearance of this molecule secondary to reduced GFR. However, the increase in odds for serious adverse events when AKI and MINS occurred together compared with either event occurring in isolation argued against a simple pharmacokinetic explanation for elevated troponin levels, supporting instead an important biologic phenomenon with serious adverse consequences. Additionally, the different pharmacokinetics of troponin and SCr made it impossible to know whether cardiac or renal injury consistently occurred first in the current study, making it impossible to distinguish between type I or III CRS or a separate pathophysiologic process simultaneously inducing injury in both organ systems (type-V CRS). However, regardless of mechanism, the co-occurrence of AKI and MINS in the current study was associated with a marked increase in odds for 30-day mortality and delayed time to discharge from the hospital, highlighting the clinical importance of this injury pattern. In conclusion, no relationship was found between a preoperative noninvasive measure of endothelial dysfunction and perioperative renal injury after noncardiac surgery. In contrast, and despite their uncertain temporal relationship, a perioperative rise in troponin was associated strongly with postoperative AKI, and their co-occurrence portended a dramatic increase in odds for 30-day mortality. Further research is required urgently to better understand this important cardiorenal relationship, identifying potential targets that might allow mitigation of this amplified risk.
REFERENCES 1. Walsh M, Garg AX, Devereaux PJ, et al: The association between perioperative hemoglobin and acute kidney injury in patients having noncardiac surgery. Anesth Analg 117:924-931, 2013
2. Bihorac A, Brennan M, Ozrazgat-Baslanti T, et al: National surgical quality improvement program underestimates the risk associated with mild and moderate postoperative acute kidney injury. Crit Care Med 41:2570-2583, 2013
1228
3. Story DA, Leslie K, Myles PS, et al: Complications and mortality in older surgical patients in Australia and New Zealand (the REASON study): A multicentre, prospective, observational study. Anaesthesia 65: 1022-1030, 2010 4. Garg AX, Kurz A, Sessler DI, et al: Perioperative aspirin and clonidine and risk of acute kidney injury: A randomized clinical trial. JAMA 312:2254-2264, 2014 5. Anderson TJ: Assessment and treatment of endothelial dysfunction in humans. J Am Coll Cardiol 34:631-638, 1999 6. Bonetti PO, Lerman LO, Lerman A: Endothelial dysfunction: A marker of atherosclerotic risk. Arterioscler Thromb Vasc Biol 23: 168-175, 2003 7. Gokce N: Clinical assessment of endothelial function: Ready for prime time? Circ Cardiovasc Imaging 4:348-350, 2011 8. Suwaidi JA, Hamasaki S, Higano ST, et al: Long-term follow-up of patients with mild coronary artery disease and endothelial dysfunction. Circulation 101:948-954, 2000 9. Schachinger V, Britten MB, Zeiher AM: Prognostic impact of coronary vasodilator dysfunction on adverse long-term outcome of coronary heart disease. Circulation 101:1899-1906, 2000 10. Halcox JP, Schenke WH, Zalos G, et al: Prognostic value of coronary vascular endothelial dysfunction. Circulation 106:653-658, 2002 11. Basile DP: The endothelial cell in ischemic acute kidney injury: Implications for acute and chronic function. Kidney Int 72:151-156, 2007 12. Bonventre JV, Yang L: Cellular pathophysiology of ischemic acute kidney injury. J Clin Invest 121:4210-4221, 2011 13. Toggweiler S, Schoenenberger A, Urbanek N, et al: The prevalence of endothelial dysfunction in patients with and without coronary artery disease. Clin Cardiol 33:746-752, 2010 14. Radenkovic M, Stojanovic M, Potpara T, et al: Therapeutic approach in the improvement of endothelial dysfunction: The current state of the art. Biomed Res Int Article ID 252158, 2013 15. Jelic S, Padeletti M, Kawut SM, et al: Inflammation, oxidative stress, and repair capacity of the vascular endothelium in obstructive sleep apnea. Circulation 117:2270-2278, 2008 16. Reriani MK, Dunlay SM, Gupta B, et al: Effects of statins on coronary and peripheral endothelial function in humans: A systematic review and meta-analysis of randomized controlled trials. Eur J Cardiovasc Prev Rehabil 18:704-716, 2011 17. Bonetti PO, Pumper GM, Higano ST, et al: Noninvasive identification of patients with early coronary atherosclerosis by assessment of digital reactive hyperemia. J Am Coll Cardiol 44:2137-2141, 2004 18. McIlroy DR, Chan MT, Wallace SK, et al: Automated preoperative assessment of endothelial dysfunction and risk stratification for perioperative myocardial injury in patients undergoing non-cardiac surgery. Br J Anaesth 112:47-56, 2014 19. Eagle KA, Berger PB, Calkins H, et al: ACC/AHA guideline update for perioperative cardiovascular evaluation for noncardiac surgery—executive summary: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 39:542-553, 2002 20. Kuvin JT, Patel AR, Sliney KA, et al: Assessment of peripheral vascular endothelial function with finger arterial pulse wave amplitude. Am Heart J 146:168-174, 2003
MCILROY ET AL
21. Levey AS, Stevens LA, Schmid CH, et al: A new equation to estimate glomerular filtration rate. Ann Intern Med 150:604-612, 2009 22. Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group. KDIGO clinical practice guideline for acute kidney injury. Kidney Int 120:179-184, 2012 23. Lee TH, Marcantonio ER, Mangione CM, et al: Derivation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation 100:1043-1049, 1999 24. Gokce N, Keaney JF Jr, Hunter LM, et al: Risk stratification for postoperative cardiovascular events via noninvasive assessment of endothelial function: A prospective study. Circulation 105: 1567-1572, 2002 25. Molitoris BA: Therapeutic translation in acute kidney injury: The epithelial/endothelial axis. J Clin Invest 124:2355-2363, 2014 26. Brodsky SV, Goligorsky MS: Endothelium under stress: Local and systemic messages. Semin Nephrol 32:192-198, 2012 27. Kwon O, Hong SM, Ramesh G: Diminished NO generation by injured endothelium and loss of macula densa nNOS may contribute to sustained acute kidney injury after ischemia-reperfusion. Am J Physiol Renal Physiol 296:F25-F33, 2009 28. Sendeski MM, Persson AB, Liu ZZ, et al: Iodinated contrast media cause endothelial damage leading to vasoconstriction of human and rat vasa recta. Am J Physiol Renal Physiol 303:F1592-F1598, 2012 29. Martin BJ, Gurtu V, Chan S, et al: The relationship between peripheral arterial tonometry and classic measures of endothelial function. Vasc Med 18:13-18, 2013 30. Lee CR, Bass A, Ellis K, et al: Relation between digital peripheral arterial tonometry and brachial artery ultrasound measures of vascular function in patients with coronary artery disease and in healthy volunteers. Am J Cardiol 109:651-657, 2012 31. Hamburg NM, Palmisano J, Larson MG, et al: Relation of brachial and digital measures of vascular function in the community: The Framingham heart study. Hypertension 57:390-396, 2011 32. Goldberg A, Hammerman H, Petcherski S, et al: In hospital and 1-year mortality of patients who develop worsening renal function following acute ST-elevation myocardial infarction. Am Heart J 150: 330-337, 2005 33. Newsome BB, Warnock DG, McClellan WM, et al: Longterm risk of mortality and end-stage renal disease among the elderly after small increases in serum creatinine level during hospitalization for acute myocardial infarction. Arch Intern Med 168: 609-616, 2008 34. Parikh CR, Coca SG, Wang Y, et al: Long-term prognosis of acute kidney injury after acute myocardial infarction. Arch Intern Med 168:987-995, 2008 35. McCullough PA, Haapio M, Mankad S, et al: Prevention of cardio-renal syndromes: Workgroup statements from the 7th ADQI Consensus Conference. Nephrol Dial Transplant 25:1777-1784, 2010 36. Ronco C, Cicoira M, McCullough PA: Cardiorenal syndrome type 1: Pathophysiological crosstalk leading to combined heart and kidney dysfunction in the setting of acutely decompensated heart failure. J Am Coll Cardiol 60:1031-1042, 2012 37. Kelly KJ: Distant effects of experimental renal ischemia/reperfusion injury. J Am Soc Nephrol 14:1549-1558, 2003
was no association between RH-PAT index and either AKI or peak ΔSCr. MINS occurred in 32 patients (14.7%) and was associated independently with the outcome of AKI (odds ratio [OR], 3.7; 95% confidence interval [CI], 1.2-10.8; p ¼ 0.02) and peak ΔSCr (β-regression coefficient 23; 95% CI, 9-37; p ¼ 0.002). Co-occurrence of AKI and MINS portended a marked increase in 30-day mortality (OR, 43; 95% CI, 6-305; p ¼ 0.001) and delayed time to discharge (hazard ratio, 0.27; 95% CI, 0.13-0.54; p ¼ 0.001). Conclusions: For patients undergoing noncardiac surgery, preoperative endothelial function assessed by noninvasive peripheral arterial tonometry was not associated with perioperative AKI. Perioperative renal injury was associated strongly with MINS, and this may represent a mechanism by which AKI increases adverse outcomes. Crown Copyright & 2015 Published by Elsevier Inc. All rights reserved.
P
options highlights the urgent need to identify potentially modifiable risk factors with the hope of reducing this burden of injury. Identification of end-organ interactions through which AKI may promote perioperative adverse events also might offer a potential target for therapeutic intervention. The vascular endothelium is an active participant in various forms of vascular disease, capable of releasing soluble factors, including nitric oxide, to induce vasodilation while also reducing the tendency toward platelet aggregation, white blood cell adhesion, and proliferation of vascular smooth muscle.5–8 Endothelial dysfunction in the coronary vasculature consistently has been demonstrated to predict adverse cardiac events.8–10 Although the etiology of AKI likely is multifactorial, endothelial dysfunction is thought to play an important mediator role in various models.11,12 However, it is unknown whether the presence of preoperative endothelial dysfunction represents an important and potentially modifiable risk factor for AKI in the perioperative context. Endothelial dysfunction has been reported in 433% of patients with or at high risk for coronary artery disease,13 and several interventions, including the use of statins, polyphenols, various antioxidant compounds, and even continuous positive airway pressure devices for patients with obstructive sleep apnea, have been suggested as potentially able to reverse such dysfunction.14–16 No diagnostic test specific for renal endothelial dysfunction has been described. However, the EndoPAT-2000 (Itamar Medical Ltd., Caesarea, Israel) is an automated noninvasive device capable of detecting endothelial dysfunction using peripheral arterial tonometry in response to reactive hyperemia (Reactive Hyperemia-Peripheral Arterial Tonometry [RH-PAT]
erioperative acute kidney injury (AKI) is a major health care burden associated with increased complications and mortality.1,2 However, mechanisms leading to adverse outcomes, particularly with lesser degrees of renal injury, remain obscure. Although AKI occurring after cardiac surgery has been the focus of extensive research, the risk factors for AKI occurring after noncardiac surgery are less well defined. Recent studies confirm that AKI after noncardiac surgery occurs commonly, with reported incidences ranging from 6% to 33%,1–4 and the lack of established preventive or therapeutic
From the *Department of Anaesthesia & Perioperative Medicine, Alfred Hospital and Monash University, Melbourne, Australia; †Department of Anesthesiology, College of Physicians and Surgeons of Columbia University, New York, NY; ‡Department of Anaesthesia & Intensive Care, Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, Hong Kong; and §Department of Anaesthesiology, Beijing Tiantan Hospital, Capital Medical University, Beijing, China. This work was jointly supported by funding from the Australian and New Zealand College of Anaesthestists (Project Grant 08/016); and the Health and Health Services Research Fund (07080421) and General Research Fund (461409), Research Grant Council of Hong Kong. There was no commercial funding for this study. Address reprint requests to: David R. McIlroy MBBS, MClinEpi, FANZCA, Department of Anaesthesia & Perioperative Medicine, Alfred Hospital, 55 Commercial Road, Melbourne, Victoria, 3004 Australia. E-mail: [email protected] Crown Copyright © 2015 Published by Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2015.05.116 1220
KEY WORDS: myocardial infarction, complications, renal complications, endothelium complications, noncardiac surgery
Journal of Cardiothoracic and Vascular Anesthesia, Vol 29, No 5 (October), 2015: pp 1220–1228
1221
PREOPERATIVE ENDOTHELIAL DYSFUNCTION
index).17 It has been approved by the FDA and has received CE mark. Although validated against endothelial dysfunction in the coronary vasculature, the generalized nature of endothelial dysfunction makes it reasonable to expect that the EndoPAT2000 also may reflect dysfunction of the renovascular endothelium. The authors have demonstrated the prognostic utility of the EndoPAT-2000 for identifying patients at risk for myocardial injury after noncardiac surgery (MINS),18 and the rapid and noninvasive nature of this testing makes it potentially applicable as a preoperative screening tool. Preoperative identification of AKI risk using an RH-PAT index would support endothelial dysfunction as a preoperative risk factor for AKI while also providing a potential target to facilitate meaningful risk reduction through preoperative optimization. Adding to data from an existing cohort previously reported with respect to myocardial injury,18 the authors sought to identify risk factors for perioperative AKI after noncardiac surgery, focusing on the prognostic utility of preoperative, noninvasively measured endothelial dysfunction while secondarily exploring the association between AKI and myocardial injury after surgery. METHODS
The study was carried out in 2 academic medical centers (Alfred Hospital, Melbourne, Australia, and Prince of Wales Hospital, Hong Kong, China). After obtaining approval from the research and ethics committees of each participating institution (Ethics Committee Project numbers 08/07, CRE2008.444 and CRE-2013.267) and written informed consent from all participants, the authors conducted an observational study in patients undergoing noncardiac surgical procedures. Patients aged 440 years and scheduled to undergo nonemergent surgery identified as intermediate or high risk for postoperative cardiac complications using American College of Cardiology/American Heart Association guidelines19 were enrolled. Patients requiring preoperative renal replacement therapy, identified as undergoing partial or complete nephrectomy, or with no postoperative serum creatinine (SCr) measurement within 7 days of surgery were excluded from analysis. In addition to routine preoperative evaluation and any additional testing performed at the discretion of the treating physician, all study patients underwent noninvasive endothelial function assessment using the automated EndoPAT-2000 device. The principles of peripheral arterial tonometry (PAT) in response to reactive hyperemia previously have been described in detail.20 In summary, proprietary technology is used to noninvasively measure the magnitude and dynamics of arteriolar tone changes in peripheral arterial beds. Similar to conventional pulse oximetry, a thimble-shaped pneumatic probe is placed on the tip of 1 finger on each hand where the volume of blood in the fingertip with each arterial pulsation is photoplethysmographically detected. After a brief period to establish baseline, a blood pressure cuff is inflated to suprasystemic pressure on 1 arm for approximately 5 minutes. After cuff deflation, the hyperemic response in the ipsilateral finger is evaluated, measuring the ratio of the pulsewave amplitude in this period to baseline pulsewave amplitude. This ratio is further normalized to the signal simultaneously obtained from
the contralateral arm, accounting for potential effects of systemic changes in vascular tone. Calculations are performed by the device software, providing a bedside quantitative assessment of peripheral endothelial function with the option for manual adjustment of calculation points as deemed appropriate. A previous study reported an RH-PAT index o1.35 to be a useful indicator of coronary artery endothelial dysfunction,17 although an RH-PAT index r1.22 previously has been identified and reported to have prognostic utility for perioperative MINS within the current study cohort.18 In addition to baseline demographics, data were collected on various comorbidities, including factors thought to potentially affect endothelial function, including hypercholesterolemia, use of 3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors, diabetes, and smoking status. The most recent SCr recorded up to and including the day before surgery was used as baseline, with preoperative estimated glomerular filtration rate (eGFR) calculated using the Chronic Kidney Disease Epidemiology Collaboration equation.21 Peak daily postoperative SCr was recorded on each of postoperative days 1 to 7 as available, with frequency of measurement and all clinical decisions at the discretion of the treating medical team. Serum troponin was measured daily for the first 3 postoperative days as previously described, with MINS defined as any troponin Z0.04 mg/L.18 Coprimary endpoints were the maximum delta serum creatinine (ΔSCr) occurring within 7 days postoperatively, as well as the incidence of AKI defined by recent creatinine-based Kidney Disease: Improving Global Outcomes guidelines.22 Additional outcomes included 30-day mortality and time to hospital discharge. Statistical Analysis Statistical analyses were performed using Stata 12 (StataCorp, College Station, TX). Continuous variables are presented as mean and standard deviation (SD) or median and interquartile range (IQR). Categorical variables are presented as counts and proportions. Logistic regression and Cox proportional hazards first determined the association between both AKI and maximum ΔSCr, with 30-day mortality and time to hospital discharge, confirming the clinical significance of these endpoints. Patients who died before discharge were assigned the longest observed time to discharge within the study cohort, ensuring that early in-hospital mortality did not spuriously reflect a favorable outcome. Univariate and multivariate logistic regression then determined the relationship between preoperative measures of endothelial dysfunction and postoperative AKI, adjusting for potentially confounding variables selected on the basis of existing data and biologic plausibility while limiting the number of variables included in any given model to avoid overfitting. A multivariate linear regression model identified factors associated with maximum ΔSCr within 7 days postoperatively. Variables were included in this starting model only if significant for the outcome of maximum ΔSCr at p o 0.10 on univariate analysis. Manual, backward, stepwise elimination proceeded, eliminating the least significant variable at each step. This procedure continued until likelihood testing confirmed either a parsimonious model or that all remaining explanatory variables were significant statistically.
1222
MCILROY ET AL
As a secondary analysis, postoperative MINS was substituted or added to these models as an explanatory variable, exploring the relationship between perioperative myocardial injury and renal injury. The authors further determined the association between AKI and MINS using MINS as the outcome of interest. Finally, the prognostic utility of these 2 surrogate outcomes were evaluated, both alone and in combination, for 30-day mortality and time to hospital discharge. Results are presented as odds ratios (ORs), hazard ratios (HRs), β-regression coefficients, and 95% confidence intervals (CI), with p o 0.05 considered significant. In this context, HR refers to the likelihood of having been discharged from hospital on any given day after surgery compared with a stated reference group, such that HR o1 indicates delayed time to discharge. In view of the post hoc nature of the analysis, a formal sample size calculation and power analysis were not undertaken. RESULTS
Between July 2007 and February 2010, the authors enrolled 249 patients into the study. After withdrawals and exclusions, 218 patients remained for analysis (Fig 1). Mean (SD) preoperative SCr was 89 (31) μmol/L, corresponding to an eGFR of 74 (21) mL/min/1.73m2. Mean (SD) peak ΔSCr within 7 days postoperatively was 9 (40) μmol/L. AKI occurred in 22 (10.1%) patients, with 14, 6, and 2 cases first identified on postoperative days 1, 2, and 3, respectively. Patients who developed AKI were older, had poorer preoperative renal function, received more intravenous fluid, and were more likely to receive packed red blood cells (RBCs) during surgery than patients who did not develop AKI (Table 1). AKI was associated with a marked increase in 30-day mortality (OR, 14.3; 95% CI, 3.0-68.9; p ¼ 0.001) and delayed time to hospital discharge (HR, 0.39; 95% CI, 0.24-0.62; p ¼ 0.001). Similarly, when assessed as a continuous variable, maximum ΔSCr within 7 days postoperatively also was associated with increased 30-day mortality (β-regression coefficient 1.02; 95% CI, 1.01-1.04; p ¼ 0.001) and delayed time to hospital discharge (HR, 0.9; 95% CI, 0.99-1.00; p ¼ 0.001). MINS occurred in 32 patients (14.7%) and was associated with an increase in both 30-day mortality (OR, 8.7; 95% CI, 1.9-41.0; p ¼ 0.01) and delayed time to hospital discharge (HR, 0.52; 95% CI, 0.35-0.76; p ¼ 0.001). RH-PAT Index and AKI Median RH-PAT index within the study cohort was 1.64 (range 1.03-4.96) and did not differ between patients with or without AKI. After adjusting for age, type of surgery, and intraoperative administration of packed RBCs, there remained no association between RH-PAT index and AKI. Despite sequentially substituting 2 separate binary thresholds for RHPAT index, both previously suggested as useful for identifying clinically relevant endothelial dysfunction,13,14 there remained no association between RH-PAT index measures of endothelial dysfunction and AKI. Further sensitivity analyses substituting preoperative eGFR in place of age, or a composite term to reflect surgical extent in place of packed RBC transfusion, in the multivariate model did not alter materially the relationship
between RH-PAT index and AKI, and no interaction was found between study site and RH-PAT index. In contrast, substituting an elevated troponin within 3 days postoperatively (MINS) into the multivariate model in place of RH-PAT, revealed a significant independent association between this index of myocardial injury and AKI (Table 2). RH-PAT Index and Maximal ΔSCr RH-PAT index was not associated with maximum ΔSCr over 7 days postoperatively regardless of whether RH-PAT index was analyzed as a continuous or binary variable (Table 3). However, the constructed multivariate model included multiple outliers with potential for influence, together with evidence of specification errors that may reflect either important missing variables or inclusion of irrelevant variables. When added to this model, MINS was associated independently with maximal ΔSCr (β-regression coefficient 23; 95% CI, 9-37; p ¼ 0.002), improving the adjusted R2 of the model from 0.20 to 0.24. Including only smaller postoperative troponin elevations (o0.3 mg/L) to reflect less severe myocardial injury did not meaningfully alter this relationship. In an exploratory analysis using MINS as the outcome of interest, both the occurrence of AKI (OR, 5.2; 95% CI, 2.013.5; p ¼ 0.001) and maximal ΔSCr (OR, 1.02 for every 1 μmol/L increase; 95% CI, 1.01-1.03; p ¼ 0.001) were associated strongly with MINS, although this was somewhat attenuated after adjusting for both age and a Revised (Lee) Cardiac Risk Index23 score (Table 4). There was a weak, albeit highly significant, correlation between maximal ΔSCr over 7 days postoperatively and peak postoperative troponin within 3 days postoperatively (Spearman’s ρ, 0.25; p ¼ 0.0002). AKI, MINS, and Outcome The presence of either AKI or MINS, occurring discordantly from the other, was associated with a tendency toward increased odds for 30-day mortality and significantly delayed time to hospital discharge (Table 5). However, the cooccurrence of both AKI and MINS was associated with a dramatic increase in odds for mortality (OR, 43; 95% CI, 6305; p ¼ 0.001) and delayed time to discharge (HR, 0.27; 95% CI, 0.13-0.54; p ¼ 0.001). Despite adjusting for age, type of surgery, and several markers of extent of surgery, the combination of AKI and MINS remained a potent predictor of increased time to discharge, appearing synergistically prognostic compared with either event in isolation. DISCUSSION
Key Findings Preoperative endothelial dysfunction identified by noninvasive PAT was not associated with AKI or the peak change in SCr measured over 7 days postoperatively in patients undergoing nonemergent noncardiac surgery. Perioperative AKI was associated strongly with MINS, although the temporal sequence and a mechanistic link remain uncertain. Nevertheless, the co-occurrence of these 2 events provides powerful prognostic information for adverse outcomes, identifying a group of patients in whom further study is required urgently to
1223
PREOPERATIVE ENDOTHELIAL DYSFUNCTION
Hong Kong
Melbourne
169 patients enrolled
80 patients enrolled
Consent withdrawn:1 patient
Consent withdrawn: 4 patients
Surgery cancelled: 1 patient Missed by investigators at time of surgery: 1 patient Surgery delayed beyond completion of study: 1 patient
165 patients completed study
76 patients completed study
No RH-PAT index available: 3 patients*
Preoperative RRT: 1 patient No postoperative SCr from POD1-7: 7 patients
157 patients with data for analysis
73 patients with data for analysis
12 patients undergoing partial or complete nephrectomy excluded
218 patients for final analysis
Fig 1. Study flow chart, by site. *Due to poor signal quality in 2 cases. In the third case, a postocclusion time shorter than the required minimum was recorded, without sufficient time to repeat the test prior to surgery. Abbreviations: POD postoperative day; RH-PAT, Reactive Hyperemia-Peripheral Arterial Tonometry; RRT, renal replacement therapy; SCr, serum creatinine.
understand and potentially mitigate the heightened perioperative risk associated with this form of cardiorenal syndrome. Interpreting the Lack of Association Between PAT and AKI A number of possible explanations should be considered when interpreting the lack of association between a PATderived measure of endothelial dysfunction and perioperative renal injury. The active role of endothelial dysfunction in vascular disease is well established,6–8 and multiple studies have indicated its role as a risk factor for adverse cardiovascular events, including in the perioperative period.9,10,18,24 Within the renal vasculature, endothelial cells are believed to play an important role in regulating blood flow to local tissue
beds, with endothelial injury promoting the adherence of inflammatory cells, increased permeability and edema, and reduced blood flow.25 Endothelial expression of molecules such as intercellular adhesion molecule (ICAM)-1 further upregulate the proinflammatory state,12 whereas dysfunction of the renal endothelium also is associated with a reduction in endothelial-dependent vasodilator activity.11 However, despite reported dysfunction of the renovascular endothelium in various models of AKI,26–28 this typically has been observed in the context of established or evolving injury, where it likely represents a mediator of AKI rather than a pre-existing and potentially modifiable risk factor. The EndoPAT-2000 has been validated as a marker of coronary endothelial dysfunction,17 a diagnosis confirmed by invasive coronary angiography with quantification of the response
1224
MCILROY ET AL
Table 1. Baseline Characteristics and Risk Factors AKI* No (n ¼ 196)
Baseline demographics Age Male sex BMI Clinical risk factors: Preoperative Hypertension Hypercholesterolemia Diabetes mellitus Current smoker NSAIDs within 2 days before surgery Diuretics HMG Co-A reductase inhibitors Aspirin within 5 days before surgery ACE inhibitors/ARBs Revised cardiac risk index Surgery-specific risk† Low Intermediate High Type of surgery General Orthopedic Vascular Urology/gynecology Major ENT/Plastics RH-PAT index‡ RH-PAT index o1.35 RH-PAT index r1.22 Risk factors: Intraoperative Intraoperative IV fluid§ (mL) Surgical duration (min) Packed RBCs intraoperatively (mL) Any packed RBCs intraoperatively, n (%) Biochemistry Preoperative SCr (μmol/L) Preoperative eGFR (mL/min/1.73m-2) Maximum delta SCr║ (μmol/L) Maximum postoperative SCr║ (μmol/L) Outcomes 30-days mortality Postoperative hospital LOS, days (IQR)
Yes (n ¼ 22)
p Value
68 (10) 124 (63.3) 25.3 (5.2)
75 (8) 18 (81.8) 25.3 (4.4)
0.001 0.08 0.99
158 100 74 23 6 32 58 52 72 1 6 114 76 109 6 41 36 4 1.61 42 14
(80.6) (51.0) (37.8) (11.7) (3.1) (16.3) (29.6) (26.5) (36.7) (1-2) (3.1) (58.1) (38.8) (55.6) (3.1) (20.9) (18.4) (2.0) (1.38-2.07) (21.4) (7.1)
21 13 11 7 0 3 13 8 9 2 0 9 13 6 0 11 5 0 1.70 3 3
(95.5) (59.1) (50.0) (31.8) (0.0) (13.6) (59.1) (36.4) (40.9) (1-3) (0.0) (40.9) (59.1) (27.3) (0.0) (50.0) (22.7) (0.0) (1.50-1.89) (13.6) (13.6)
0.14 0.47 0.26 0.02 0.99 0.99 0.01 0.33 0.70 0.005 0.21
2000 219 0 23
(1500-3000) (102) (0-0) (11.7)
3000 268 0 9
(2000-3500) (96) (0-500) (40.9)
0.01 0.03 0.0002 0.001
(92-112) (16) (36-113) (126-237)
0.001 0.001 0.0001 0.0001
82 75 1 80
(67-98) (21) (–8 to 9) (68-99)
3 (1.5) 8 (6-12)
98 61 49 153
4 (18.2) 11 (6-26)
0.03
0.73 0.58 0.39
0.002 0.13
Abbreviations: ACE, angiotensin-converting enzyme; AKI, acute kidney injury; ARB, angiotensin-receptor blocker; BMI, body mass index; eGFR, estimated glomerular filtration rate; ENT, ear, nose, throat; HMG Co-A, 3-hydroxy-3-methylglutaryl-coenzyme A; IQR, interquartile range; IV, intravenous; LOS, length of stay; NSAID, nonsteroidal anti-inflammatory drug; RH-PAT, Reactive Hyperemia–Peripheral Arterial Tonometry; RBC, red blood cell; SCr, serum creatinine. *AKI defined by Kidney Disease: Improving Global Outcomes Kidney Disease: Improving Global Outcomes creatinine-based criteria over 7 days postoperatively. †Surgery-specific risk assigned by the treating clinician with reference to American College of Cardiology/American Heart Association Guidelines on Perioperative Cardiovascular Evaluation and Care for Noncardiac Surgery. ‡Calculated by the EndoPAT device. §Defined as the sum of intraperative crystalloid and colloid administration. ║Defined within 7 days postoperatively.
to graded increases in direct acetylcholine administration. In contrast, no such gold standard exists to confirm the presence of endothelial dysfunction in the renal vasculature, and although it is held widely to be a generalized process occurring throughout the vasculature, the EndoPAT-2000 provides only an indirect measure and has not been evaluated specifically or validated to identify endothelial dysfunction within the renal vasculature. Moreover, a limited association between metrics of PAT and
flow-mediated vasodilation (FMD) when compared with each other29–31 suggests that quantification of endothelial dysfunction is complex and may not be captured adequately by a single measurement technique. Despite the automated and relative operator-independent nature of the EndoPAT device, it also remains uncertain to what extent variations in other baseline characteristics and risk factors, differences in patient activity before testing and environmental conditions at the time of testing, or unmeasured
1225
PREOPERATIVE ENDOTHELIAL DYSFUNCTION
Table 2. Unadjusted and Adjusted Odds Ratios for Outcome of AKI* After Noncardiac Surgery ‡
RH-PAT index RH-PAT index r1.22 RH-PAT index o1.35 Type of surgery General Orthopedic Vascular Urology/gynecology Major ENT/plastics Any intraoperative pRBCs Age (y) o65 65-74 Z75 Preoperative eGFR (mL/min/1.73 m-2) Preoperative eGFR o60 mL/min/1.73 m-2 Preoperative SCr (μmol/L) Female sex BMI Hypertension Hypercholesterolemia Diabetes mellitus Current smoker Aspirin (past 5 days) NSAID (past 2 days) ACEI or ARB Diuretic HMG CoA-reductase inhibitor Surgery specific risk§: Low Intermediate High Intraoperative pRBCs (per 100 mL) Study site Surgical duration (per 10 min) Intraoperative IV fluid║ (per 100 mL) Extensive surgery¶ Peak troponin Z0.04 mg/L by POD 3 Peak troponin Z0.04 mg/L and o0.30 mg/L by POD 3
Univariate OR (95% CI)
p Value
Multivariate† OR (95% CI)
p Value
1.0 (0.5-2.0) 2.1 (0.5-7.8) 0.6 (0.2-2.1) 1.00 (reference) N/A 4.9 (1.7-14.0) 2.5 (0.7-8.8) N/A 5.2 (2.0-13.5) 1.00 (reference) 4.2 (0.9-20.7) 7.7 (1.7-35.9) 0.97 (0.95-0.99) 2.0 (0.8-5.0) 1.01 (1.00-1.03) 0.4 (0.1-1.2) 1.00 (0.9-1.1) 5.1 (0.7-38.7) 1.4 (0.6-3.4) 1.6 (0.7-4.0) 3.5 (1.3-9.5) 1.6 (0.6-4.0) N/A 1.2 (0.5-2.9) 0.8 (0.2-2.9) 3.4 (1.4-8.5) N/A 1.00 (ref) 2.2 (0.9-5.3) 1.1 (1.0-1.3) 0.7 (0.3-1.6) 1.04 (1.00-1.08) 1.03 (1.00-1.06) 2.7 (1.1-6.6) 5.2 (2.0-13.5) 4.2 (1.3-13.2)
0.98 0.29 0.40 — — 0.003 0.15 — 0.001 — 0.07 0.01 0.003 0.13 0.02 0.09 0.99 0.12 0.47 0.27 0.01 0.33
1.0 (0.5-2.0) 1.6 (0.4-7.7) 0.5 (0.1-1.8) 1.00 (reference) — 3.5 (1.1-11.2) 2.2 (0.6-8.1) — 3.9 (1.4-11.1) 1.00 (reference) 4.7 (0.9-24.1) 5.0 (1.0-25.4) 0.98 (0.95-1.0)
0.94 0.53 0.26 — — 0.03 0.23 — 0.01
3.7 (1.2-10.8) 3.9 (1.0-15.1)
0.02 0.05
0.06 0.05 0.05
0.70 0.75 0.01 0.09 0.04 0.35 0.04 0.02 0.03 0.001 0.02
Abbreviations: ACEI, angiotensin-converting enzyme inhibitor; AKI, acute kidney injury; ARB, angiotensin-receptor blocker; BMI, body mass index; eGFR, estimated glomerular filtration rate; ENT, ear, nose, throat; HMG Co-A, 3-hydroxy-3-methylglutaryl-coenzyme A; IV, intravenous; NSAID, nonsteroidal anti-inflammatory drug; POD, postoperative day; pRBC, packed red blood cell; RH-PAT, Reactive Hyperemia–Peripheral Arterial Tonometry Index; SCr, serum creatinine. *Defined according to creatinine-based Kidney Disease: Improving Global Outcomes Kidney Disease: Improving Global Outcomes criteria. †Multivariate analysis included RH-PAT index adjusted for categorical variables of age, type of surgery, and intraoperative administration of pRBCs. Sensitivity analysis included substituting binary values for RH-PAT index according to previously published thresholds, substituting eGFR for age, and extensive surgery for administration of pRBCs. Finally, troponin terms were substituted into the model in place of RH-PAT index. For all models assessed, Hosmer-Lemeshow test 4 0.05. ‡Calculated by the EndoPAT device. §Surgery-specific risk assigned by the treating clinician with reference to American College of Cardiology/American Heart Association Guidelines on Perioperative Cardiovascular Evaluation and Care for Noncardiac Surgery. ║Defined as the sum of intraoperative crystalloid and colloid administration. ¶Defined as the composite of intravenous fluid 45000 mL, surgical duration 45 hours or any transfusion of pRBCs intraoperatively as a marker of surgical extent.
operator variability where manual adjustment of parameters for analysis was employed may have influenced results. However, the same device used in the same cohort recently demonstrated prognostic utility for myocardial injury after surgery.18 Although most of the data for the current study were collected prospectively, the frequency of measurement of SCr was at the discretion of the treating clinicians and may have introduced an unmeasurable degree of ascertainment bias. The
limited sample size and total number of events in the current study cohort raised the possibility of a type-II error and precludes a more detailed analysis of factors associated with AKI. However, despite the acknowledged limitations of this multivariate modeling process, the current study was not intended to build a robust and comprehensive model for the prediction of AKI after noncardiac surgery, and no meaningful conclusion should be drawn from these data about a potential
1226
MCILROY ET AL
Table 3. Unadjusted and Adjusted Regression Coefficients for the Outcome of Maximum ΔSCr* After Surgery Unadjusted β-regression Coefficient (95% CI)
Age (y) o65 65-74 Z75 Female sex BMI Hypertension Hypercholesterolemia Diabetes mellitus Current smoker Aspirin (within 5 days preoperatively) NSAID (within 2 days postoperatively) ACEI/ARB Diuretic HMG co-A reductase inhibitor Type of surgery General Orthopedic Vascular Urology/gynecology Major ENT/plastics Surgery-specific risk‡: Low Intermed High Intraoperative pRBCs (per 100 mL) Any intraoperative pRBCs Site RH-PAT index§ RH-PAT index r1.22 RH-PAT index o1.35 Surgical duration (per 10 min) Intraoperative IV fluid║ (per 100 mL) Preoperative SCr (mmol/L) Preoperative eGFR (mL/min/1.73 m-2) Preoperative eGFR o60 mL/min/1.73 m-2 Peak troponin Z0.04 mg/L by POD 3 Peak troponin Z0.04 mg/L and o0.30 mg/L-1 by POD 3
0.0 6 24 –10 1 14 6 3 24 3 –7 7 0 18
(reference) (–7 to 18) (11-37) (–22 to 1) (0-2) (0-28) (–4 to 17) (–8 to 14) (9-40) (–9 to 15) (–41 to 26) (–4 to 18) (–15 to 15) (7-29)
0.0 3 25 8 2 0 10 16 4 37 –8 0 20 1 1 1 0.2 –0.4 12 39 35
(reference) (–30 to 36) (12-38) (–6 to 22) (–37 to 42) (reference) (–23 to 43) (–17 to 50) (2-6) (22-51) (–20 to 3) (–9 to 9) (0-40) (–13 to 14) (0-1) (0-1) (0.0-0.4) (–0.6 to –0.1) (0-24) (25-53) (19-51)
Adjusted β-regression p Value
0.37 0.001 0.07 0.28 0.05 0.24 0.62 0.002 0.61 0.66 0.22 0.98 0.002
Coefficient† (95% CI)
p Value
0.0 (reference) 7 (–5 to 18) 21 (8-33)
0.28 0.001
25 (10-39)
0.001
17 (6-27)
0.002
31 (17-45)
0.001
23 (9-37) 24 (8-40)
0.002 0.003
0.86 0.001 0.26 0.90
0.56 0.34 0.001 0.001 0.16 0.99 0.06 0.94 0.02 0.003 0.01 0.004 0.04 0.001 0.001
Abbreviations: ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin-receptor blocker; BMI, body mass index; eGFR, estimated glomerular filtration rate; ENT, ear, nose, throat; IV, intravenous; HMG Co-A, 3-hydroxy-3-methylglutaryl-coenzyme A; IV, intravenous; NSAID, nonsteroidal anti-inflammatory drug; POD, postoperative day; pRBC, packed red blood cell; RH–PAT, Reactive Hyperemia–Peripheral Arterial Tonometry Index; SCr, serum creatinine. *Defined over 7 days postoperatively. †RH-PAT o1.22 was eliminated from the multivariate model. Subsequent addition of an elevated troponin into the remaining model increased the adjusted R2 from 0.20 to 0.24 with likelihood testing suggesting a better model (p ¼ 0.002). ‡Surgery-specific risk assigned by the treating clinician with reference to American College of Cardiology/American Heart Association Guidelines on Perioperative Cardiovascular Evaluation and Care for Noncardiac Surgery. §Calculated by the EndoPAT device. ║Defined as the sum of intraoperative crystalloid and colloid administration.
Table 4. Unadjusted and Adjusted Odds Ratios for the Outcome of MINS* After Noncardiac Surgery ‡
Any AKI Maximal ΔSCr§
Unadjusted OR (95% CI)
p Value
Adjusted† OR (95% CI)
p Value
5.2 (2.0-13.5) 1.02 (1.01-1.03)
0.001 0.001
3.1 (1.1-8.7) 1.02 (1.00-1.03)
0.03 0.01
Abbreviations: AKI, acute kidney injury; MINS, myocardial injury after noncardiac surgery; POD, postoperative day; SCr, serum creatinine. *Defined as troponin Z0.04 mg/L by POD 3. †Adjusted for categorical variables of age and Revised Cardiac Risk (Lee) Index Z2. ‡Defined by creatinine-based Kidney Disease: Improving Global Outcomes Kidney Disease: Improving Global Outcomes guidelines. §Odds ratio is for every 1 unit increase in maximal ΔSCr (mmol/L).
1227
PREOPERATIVE ENDOTHELIAL DYSFUNCTION
Table 5. Outcomes Associated With Postoperative AKI* and MINS† Status Odds ratio (95% CI) Outcome: 30-days mortality
All patients
AKI–/MINS–
AKIþ/MINS–
AKI–/MINSþ
AKIþ/MINSþ
1.0 (reference)
7.1 (0.6-84)
3.9 (0.3-45)
43 (6-305)
AKI–/MINS–
AKIþ/MINS–
AKI–/MINSþ
AKIþ/MINSþ
1.00 (reference) 1.00 (reference)
0.46 (0.25-0.82) 0.53 (0.28-0.99)
0.62 (0.40-0.97) 0.51 (0.32-0.81)
0.27 (0.13-0.54) 0.28 (0.13-0.60)
Hazard ratio (95% CI) Outcome: Time to hospital discharge Unadjusted HR Adjusted HR‡
Abbreviations: AKI, acute kidney injury; HR, hazard ratio; MINS, myocardial injury after noncardiac surgery. *Defined by creatinine-based Kidney Disease: Improving Global Outcomes Kidney Disease: Improving Global criteria over 7 days postoperatively. †Defined by peak postoperative troponin Z0.04 mg/L within 3 postoperative days. ‡Adjusted for surgical duration, total intraoperative volume of clear intravenous fluid, and categorical variables of age, type of surgery, and any intraoperative transfusion of packed red blood cells.
relationship among other risk factors, such as exposure to perioperative nonsteroidal anti-inflammatory drugs or angiotensinconverting enzyme I and renal injury. Rather, the data presented here provide important evidence against endothelial dysfunction measured by noninvasive PAT as a preoperative risk factor for AKI. Renal Injury and MINS The association among biomarkers of MINS and renal injury in the perioperative context is a novel finding. Although previous studies have demonstrated an increase in both short- and longerterm mortality when AKI complicates acute myocardial infarction (MI),32–34 an admission diagnosis of MI in these studies may reflect larger myocardial injury than typically identified in the perioperative period by routine troponin measurement (MINS). Although the pathophysiology of cardiorenal syndrome (CRS) remains poorly understood, its clinical significance increasingly is recognized. Well-described subtypes include an acute deterioration in cardiac function leading to AKI (CRS type I) or an acute deterioration in renal function leading to an acute cardiac disorder (CRS type III), highlighting the potential bidirectional nature of the syndrome.35 The pathophysiology of this bidirectional relationship has been described in terms of gross clinical interactions, such as AKI leading to congestive heart failure through oliguria, fluid overload, and uremia, or MI and heart failure leading to AKI through a process of reduced cardiac output and activation of the renin-angiotensin and sympathetic nervous systems, with subsequent reductions in renal blood flow and GFR.36 However, at a cellular level, experimental renal ischemia-reperfusion injury clearly induces a proinflammatory state with upregulation of various cytokines.12 Elevated levels of interluekin-1, tumor necrosis factor-α, and ICAM, as well as increased myeloperoxidase activity, all have been identified within the myocardium after isolated renal insult. In fact, even brief periods of renal ischemia insufficient to produce azotemia can induce cardiac apoptosis that appears to be mediated, at least
in part, by tumor necrosis factor -α.37 Similar humoral mechanisms of organ crosstalk have been postulated to promote AKI after cardiac injury.36 A mechanistic evaluation of the relationship between MINS and renal injury was beyond the scope of the current study, and a number of uncertainties remain. The pharmacokinetic effect of acute changes in renal function on circulating troponin levels was not well characterized, and the frequent mild elevations in postoperative troponin seen in the current study might simply reflect reduced renal clearance of this molecule secondary to reduced GFR. However, the increase in odds for serious adverse events when AKI and MINS occurred together compared with either event occurring in isolation argued against a simple pharmacokinetic explanation for elevated troponin levels, supporting instead an important biologic phenomenon with serious adverse consequences. Additionally, the different pharmacokinetics of troponin and SCr made it impossible to know whether cardiac or renal injury consistently occurred first in the current study, making it impossible to distinguish between type I or III CRS or a separate pathophysiologic process simultaneously inducing injury in both organ systems (type-V CRS). However, regardless of mechanism, the co-occurrence of AKI and MINS in the current study was associated with a marked increase in odds for 30-day mortality and delayed time to discharge from the hospital, highlighting the clinical importance of this injury pattern. In conclusion, no relationship was found between a preoperative noninvasive measure of endothelial dysfunction and perioperative renal injury after noncardiac surgery. In contrast, and despite their uncertain temporal relationship, a perioperative rise in troponin was associated strongly with postoperative AKI, and their co-occurrence portended a dramatic increase in odds for 30-day mortality. Further research is required urgently to better understand this important cardiorenal relationship, identifying potential targets that might allow mitigation of this amplified risk.
REFERENCES 1. Walsh M, Garg AX, Devereaux PJ, et al: The association between perioperative hemoglobin and acute kidney injury in patients having noncardiac surgery. Anesth Analg 117:924-931, 2013
2. Bihorac A, Brennan M, Ozrazgat-Baslanti T, et al: National surgical quality improvement program underestimates the risk associated with mild and moderate postoperative acute kidney injury. Crit Care Med 41:2570-2583, 2013
1228
3. Story DA, Leslie K, Myles PS, et al: Complications and mortality in older surgical patients in Australia and New Zealand (the REASON study): A multicentre, prospective, observational study. Anaesthesia 65: 1022-1030, 2010 4. Garg AX, Kurz A, Sessler DI, et al: Perioperative aspirin and clonidine and risk of acute kidney injury: A randomized clinical trial. JAMA 312:2254-2264, 2014 5. Anderson TJ: Assessment and treatment of endothelial dysfunction in humans. J Am Coll Cardiol 34:631-638, 1999 6. Bonetti PO, Lerman LO, Lerman A: Endothelial dysfunction: A marker of atherosclerotic risk. Arterioscler Thromb Vasc Biol 23: 168-175, 2003 7. Gokce N: Clinical assessment of endothelial function: Ready for prime time? Circ Cardiovasc Imaging 4:348-350, 2011 8. Suwaidi JA, Hamasaki S, Higano ST, et al: Long-term follow-up of patients with mild coronary artery disease and endothelial dysfunction. Circulation 101:948-954, 2000 9. Schachinger V, Britten MB, Zeiher AM: Prognostic impact of coronary vasodilator dysfunction on adverse long-term outcome of coronary heart disease. Circulation 101:1899-1906, 2000 10. Halcox JP, Schenke WH, Zalos G, et al: Prognostic value of coronary vascular endothelial dysfunction. Circulation 106:653-658, 2002 11. Basile DP: The endothelial cell in ischemic acute kidney injury: Implications for acute and chronic function. Kidney Int 72:151-156, 2007 12. Bonventre JV, Yang L: Cellular pathophysiology of ischemic acute kidney injury. J Clin Invest 121:4210-4221, 2011 13. Toggweiler S, Schoenenberger A, Urbanek N, et al: The prevalence of endothelial dysfunction in patients with and without coronary artery disease. Clin Cardiol 33:746-752, 2010 14. Radenkovic M, Stojanovic M, Potpara T, et al: Therapeutic approach in the improvement of endothelial dysfunction: The current state of the art. Biomed Res Int Article ID 252158, 2013 15. Jelic S, Padeletti M, Kawut SM, et al: Inflammation, oxidative stress, and repair capacity of the vascular endothelium in obstructive sleep apnea. Circulation 117:2270-2278, 2008 16. Reriani MK, Dunlay SM, Gupta B, et al: Effects of statins on coronary and peripheral endothelial function in humans: A systematic review and meta-analysis of randomized controlled trials. Eur J Cardiovasc Prev Rehabil 18:704-716, 2011 17. Bonetti PO, Pumper GM, Higano ST, et al: Noninvasive identification of patients with early coronary atherosclerosis by assessment of digital reactive hyperemia. J Am Coll Cardiol 44:2137-2141, 2004 18. McIlroy DR, Chan MT, Wallace SK, et al: Automated preoperative assessment of endothelial dysfunction and risk stratification for perioperative myocardial injury in patients undergoing non-cardiac surgery. Br J Anaesth 112:47-56, 2014 19. Eagle KA, Berger PB, Calkins H, et al: ACC/AHA guideline update for perioperative cardiovascular evaluation for noncardiac surgery—executive summary: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 39:542-553, 2002 20. Kuvin JT, Patel AR, Sliney KA, et al: Assessment of peripheral vascular endothelial function with finger arterial pulse wave amplitude. Am Heart J 146:168-174, 2003
MCILROY ET AL
21. Levey AS, Stevens LA, Schmid CH, et al: A new equation to estimate glomerular filtration rate. Ann Intern Med 150:604-612, 2009 22. Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group. KDIGO clinical practice guideline for acute kidney injury. Kidney Int 120:179-184, 2012 23. Lee TH, Marcantonio ER, Mangione CM, et al: Derivation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation 100:1043-1049, 1999 24. Gokce N, Keaney JF Jr, Hunter LM, et al: Risk stratification for postoperative cardiovascular events via noninvasive assessment of endothelial function: A prospective study. Circulation 105: 1567-1572, 2002 25. Molitoris BA: Therapeutic translation in acute kidney injury: The epithelial/endothelial axis. J Clin Invest 124:2355-2363, 2014 26. Brodsky SV, Goligorsky MS: Endothelium under stress: Local and systemic messages. Semin Nephrol 32:192-198, 2012 27. Kwon O, Hong SM, Ramesh G: Diminished NO generation by injured endothelium and loss of macula densa nNOS may contribute to sustained acute kidney injury after ischemia-reperfusion. Am J Physiol Renal Physiol 296:F25-F33, 2009 28. Sendeski MM, Persson AB, Liu ZZ, et al: Iodinated contrast media cause endothelial damage leading to vasoconstriction of human and rat vasa recta. Am J Physiol Renal Physiol 303:F1592-F1598, 2012 29. Martin BJ, Gurtu V, Chan S, et al: The relationship between peripheral arterial tonometry and classic measures of endothelial function. Vasc Med 18:13-18, 2013 30. Lee CR, Bass A, Ellis K, et al: Relation between digital peripheral arterial tonometry and brachial artery ultrasound measures of vascular function in patients with coronary artery disease and in healthy volunteers. Am J Cardiol 109:651-657, 2012 31. Hamburg NM, Palmisano J, Larson MG, et al: Relation of brachial and digital measures of vascular function in the community: The Framingham heart study. Hypertension 57:390-396, 2011 32. Goldberg A, Hammerman H, Petcherski S, et al: In hospital and 1-year mortality of patients who develop worsening renal function following acute ST-elevation myocardial infarction. Am Heart J 150: 330-337, 2005 33. Newsome BB, Warnock DG, McClellan WM, et al: Longterm risk of mortality and end-stage renal disease among the elderly after small increases in serum creatinine level during hospitalization for acute myocardial infarction. Arch Intern Med 168: 609-616, 2008 34. Parikh CR, Coca SG, Wang Y, et al: Long-term prognosis of acute kidney injury after acute myocardial infarction. Arch Intern Med 168:987-995, 2008 35. McCullough PA, Haapio M, Mankad S, et al: Prevention of cardio-renal syndromes: Workgroup statements from the 7th ADQI Consensus Conference. Nephrol Dial Transplant 25:1777-1784, 2010 36. Ronco C, Cicoira M, McCullough PA: Cardiorenal syndrome type 1: Pathophysiological crosstalk leading to combined heart and kidney dysfunction in the setting of acutely decompensated heart failure. J Am Coll Cardiol 60:1031-1042, 2012 37. Kelly KJ: Distant effects of experimental renal ischemia/reperfusion injury. J Am Soc Nephrol 14:1549-1558, 2003