1. Preoperative Cerebral Oxygen Saturation and ClinicalOutcomes in Cardiac SurgeryMatthias Heringlake, M.D.,* Christof Garbers, Cand. Med.,† Jan-Hendrik Kabler, Cand. Med.,† ¨Ingrid Anderson, Cand. Med.,† Hermann Heinze, M.D.,‡ Julika Schon, M.D.,‡ ¨Klaus-Ulrich Berger, M.D.,‡ Leif Dibbelt, M.D.,§ Hans-Hinrich Sievers, M.D.,࿣ Thorsten Hanke, M.D.#ABSTRACT What We Know about This Topic • Intraoperative cerebral oxygen saturation (ScO2) monitoringBackground: The current study was designed to determine has been used to assess the adequacy of cerebral oxygenthe relation between preoperative cerebral oxygen saturation delivery to demand.(ScO2), variables of cardiopulmonary function, mortality, andmorbidity in a heterogeneous cohort of cardiac surgery patients.Methods: In this study, 1,178 consecutive patients scheduled What This Article Tells Us That Is Newfor on-pump surgery were prospectively studied. Preoperative • Preoperative ScO2 concentrations are reflective of baselineScO2, demographics, N-terminal pro-B-type natriuretic pep- severity of cardiopulmonary dysfunction, associated withtide, high-sensitive troponin T, clinical outcomes, and 30-day short- and long-term mortality and morbidity, and may add toand 1-yr mortality were recorded. preoperative risk stratification in patients undergoing cardiac surgery.Results: Median additive EuroSCORE was 5 (range: 0 –19).Thirty-day and 1-yr mortality and major morbidity (at leasttwo major complications and/or a high-dependency unit stay (␳: 0.19), EuroSCORE (␶: 0.20), and left ventricular ejec-of at least 10 days) were 3.5%, 7.7%, and 13.3%, respec- tion fraction class (␶: 0.12). Thirty-day nonsurvivors had atively. Median minimal preoperative oxygen supplemented lower ScO2min-ox than survivors (median 58% [95% CI,ScO2 (ScO2min-ox) was 64% (range: 15–92%). ScO2min-ox was 50.7– 62%] vs. 64% [95% CI, 64 – 65%]; P Ͻ 0.0001).correlated (all: P value Ͻ0.0001) with N-terminal pro-B- Receiver-operating curve analysis of ScO2min-ox and 30-daytype natriuretic peptide (␳: Ϫ0.35), high-sensitive troponin mortality revealed an area-under-the-curve of 0.71 (95% CI,T (␳: Ϫ0.28), hematocrit (␳: 0.34), glomerular filtration rate 0.68 – 0.73%; P Ͻ 0.0001) in the total cohort and an area- under-the-curve of 0.77 (95% CI, 0.69 – 0.86%; P Ͻ 0.0001) * Professor, Deputy Director, Cardiac Anesthesia Unit, Department in patients with a EuroSCORE more than 10. Logistic regres-of Anesthesiology, University of Lubeck, Lubeck, Germany, † Medical ¨ ¨ sion based on different EuroSCORE categories (0 –2; 3–5,Student, Department of Anesthesiology, University of Lubeck, ‡ Dep- ¨uty Director, Department of Anesthesiology, University of Lubeck, ¨ 6 –10, Ͼ10), ScO2min-ox, and duration of cardiopulmonary by-§ Deputy Director, Department of Clinical Chemistry, University of pass showed that a ScO2min-ox equal or less than 50% is anLubeck, ࿣ Professor, Director, Department of Cardiac and Thoracic ¨ independent risk factor for 30-day and 1-yr mortality.Vascular Surgery, University of Lubeck, # Deputy Director, Department ¨of Cardiac and Thoracic Vascular Surgery, University of Lubeck. ¨ Conclusions: Preoperative ScO2 levels are reflective of the Received from the Department of Anesthesiology, University of severity of cardiopulmonary dysfunction, associated with short-Lubeck, Lubeck, Germany. Submitted for publication April 5, 2010. ¨ ¨ and long-term mortality and morbidity, and may add to preop-Accepted for publication August 31, 2010. Support was provided erative risk stratification in patients undergoing cardiac surgery.solely from institutional and/or departmental sources. Covidien(Mannheim, Germany) provided technical support through anINVOS4100 monitor to be used outside the operating room. Thehormone analyses were performed by Roche Diagnostics (Gren-zach-Wyhlen, Germany). The author M. Heringlake receives hono-raria for lectures from Covidien (Neustadt, Germany) and scientific C EREBRAL oxygen saturation (ScO2) monitoring by near-infrared spectroscopy is increasingly used for as- sessing the adequacy of cerebral oxygen delivery to demandsupport from Roche Diagnostics (Mannheim, Germany). The work in patients undergoing cardiac and noncardiac surgery.1–3 Ahas been presented in part at the autumn meeting of the WorkingGroup in Cardiac Anesthesia of the German Association of Anaes- relative decrease of intraoperative ScO2 to less than 80% ofthesiology and Intensive Care, November 12–14, 2009, in Ulm, the preoperative baseline or to absolute levels lower thanGermany, and at the 25th Annual Meeting of the European Associ- 50% have been associated with postoperative cognitive dys-ation of Cardiothoracic Anesthesiologists, June 9 –11, 2010, in Edin-burgh, Scotland. Address correspondence to Dr. Heringlake: Department of An- ᭛ This article is featured in “This Month in Anesthesiology.”esthesiology, University of Lubeck, Lubeck, Ratzeburger Allee 160, ¨ ¨ Please see this issue of ANESTHESIOLOGY, page 9A.D-23538 Lubeck, Germany. email@example.com. This article may ¨be accessed for personal use at no charge through the Journal Web site, ᭜ This article is accompanied by an Editorial View. Please see:www.anesthesiology.org. Murkin JM: Cerebral oximetry: Monitoring the brain as theCopyright © 2010, the American Society of Anesthesiologists, Inc. Lippincott index organ. ANESTHESIOLOGY 2011; 114:12–3.Williams & Wilkins. Anesthesiology 2011; 114: 58 – 69Anesthesiology, V 114 • No 1 58 January 2011
2. PERIOPERATIVE MEDICINEfunction,1,4 a higher rate of stroke or coma,5 increased post- 48 patients’ surgery was cancelled, and no ScO2 measure-operative morbidity from other nonneurological causes,6 and ments were performed in 2 patients, leaving 1,178 that coulda prolonged stay in the high-dependency unit or the hospi- be analyzed. As a result of missing either room or oxygen-tal.7 Sparse data are available about the relation between supplemented ScO2 concentrations or hormone values, 194preoperative ScO2 levels and clinical outcomes in cardiac sur- datasets were partially incomplete. Complete datasets, in-gery patients. Murkin et al. have shown that patients present- cluding hormones, were obtained from 984 patients.ing with major organ morbiditiy and mortality after coronaryartery bypass grafting (CABG) had lower baseline and nadir Determination of ScO2ScO2 concentrations than those that did not.6 Based on these ScO2 was determined by near-infrared spectroscopy with anfindings and our own clinical observations that cardiac sur- INVOS௡ 4100 or 5100 cerebral oximeter (Somanetics,gery patients presenting with decreased preoperative ScO2 Troy, MI). With the exception of emergency patients and(i.e., lower than 50% absolute) often developed complica- intubated patients, who were studied upon arrival in thetions and needed prolonged therapy in the intensive care operating room, all patients were examined preoperativelyunit, we hypothesized that preoperative ScO2 is reflective of on the cardiac surgery ward. In this case, the patients werethe global cardiopulmonary function and may be used for fully awake and had received neither sedating drugs nor opi-risk stratification. The current study was designed to ex- oids. Emergency and intubated patients received sedating orplore the association of preoperative ScO2 concentrations analgesic drugs, as clinically appropriate.and established risk factors in patients scheduled for on- On the ward, ScO2 was first determined in the resting state inpump cardiac surgery and to determine the relation be- the semirecumbent position, while the patients breathedtween these variables and clinical outcomes (i.e., mortality room air, and the lowest value from the left or right hemi-and morbidity). sphere was used for determination of (ScO2min-room). Trans- cutaneous arterial oxygen saturation (SaO2), mean arterialMaterials and Methods blood pressure, and heart rate were determined concomi-Study Design tantly. Thereafter, ScO2 measurements were repeated whenThe study was designed as a prospective cohort study. Pri- the patients were breathing oxygen-enriched air until no fur-mary objectives were to determine anamnestic and cardiovas- ther increase in SaO2 was observed. The lowest ScO2 valuecular factors influencing preoperative ScO2 and to explore the during oxygen supplementation recorded in a patient, eitherrelation between preoperative ScO2 and clinical outcome (30- from the left or the right hemisphere, was determined asday and 1-yr mortality and morbidity). ScO2min-ox. In 69 patients, no ScO2 measurements with oxy- gen were available; in these patients, ScO2min-room values werePatients used for further analysis.After approval by the local ethical committee (Ethikkommis-sion der Universitat zu Lubeck, Lubeck, Germany), all pa- ¨ ¨ ¨ Standard Risk Factors and Clinical Outcomestients scheduled for cardiac surgery with cardiopulmonary Besides demographic variables serving as potential risk factors,bypass (CPB) at the University of Lubeck from January 1, ¨ the following specific cardiovascular risk factors were assessed:2008, to December 31, 2008, were screened for participationin this prospective, observational trial. Exclusion criteria ● Left ventricular ejection fraction (LVEF)were less than 18 yr of age and planned off-pump surgery. ● New York Heart Association gradeWritten informed consent was obtained from all elective and ● Additive EuroSCOREurgent patients as well as emergency patients capable of com- ● Estimated glomerular filtration ratemunication. In case of sedated or intubated patients sched- LVEF was derived from the preoperative diagnostics (le-uled for emergency surgery, consent was obtained from the vocardiography or echocardiography) and graded as: se-next of kin and reconfirmed after recovery. verely reduced, LVEF less than 30%; moderately reduced,Justification of Sample Size 30 – 49%; and normal, Ͼ50%. The additive EuroSCORETo our knowledge, the association between preoperative was calculated according to Roques et al.8 and analyzed asScO2, cardiovascular and surgical risk factors, and clinical a continuous variable. Estimated glomerular filtration rateoutcome has not been formally analyzed previously in was calculated from preoperative plasma creatinine by the ab-larger studies. The current study may thus be regarded as breviated Modifications of Diet in Renal Disease equation.9a pilot trial. The sample size was adjusted to comparable Clinical outcomes (30-day mortality, major complica-studies analyzing the association between preoperative tions, and duration of treatment in the high-dependencyvariables and outcome. unit) were derived from the prospectively sampled cardiac surgery database. All-cause 1-yr mortality was determinedRecruitment from the hospital database, by contacting the patient’s pri-During the study period, 1,230 patients fulfilled the inclu- mary physician, the patient, or the patient’s next of kin. Thesion criteria. Two patients refused to participate in the study, follow-up rate was 98.3%.Anesthesiology 2011; 114:58 – 69 59 Heringlake et al.
3. Preoperative Cerebral Oxygen Saturation and OutcomeDetermination of High-sensitivity Troponin T and 7; combined valve surgery, 127; major thoracic vascular sur-N-terminal Pro-B-type Natriuretic Peptide gery, 60; other on-pump procedures, 27).Arterial blood samples for determination of N-terminal pro-B-type natriuretic peptide (NTproBNP) and high-sensitive Statistical Analysestroponin T (hsTNT) were obtained immediately before in- Analyses were performed in the total cohort and in a group ofduction of anesthesia. EDTA plasma was separated and 102 high-risk patients with an additive EuroSCORE morestored at Ϫ80° for further analysis. than 10.12 No adjustments for multiple testing were made. hsTNT was determined by the electrochemiluminescence Data were analyzed by MedCalc 11.3.3 statistical soft-method as described recently (ECLIA; Elecsys 2010 analyzer, ware package (MedCalc Software bvba, Mariakerke, Bel-Roche Diagnostics, Grenzach-Wyhlen, Germany).10 The gium). Because most of the relevant data were not normallylower detection limit of this assay was 3 pg/ml. The interassay distributed after Kolmogornov–Smirnov testing, all data, ifcoefficient of variation was 5.4% at 28 pg/ml and 7.1% at not stated otherwise, are given as median and 25% of 75%2,350 pg/ml. quartiles. Statistical significance was assessed at the 5% level NTproBNP was determined by an electrochemilumines- (P Ͻ 0.05 is statistically significant).cence immunoassay (Elecsys proBNP sandwich immunoas- To allow better comparisons with the additive Euro-say; Roche Diagnostics) on Elecsys 2010.11 The interassay SCORE, analyses regarding mortality were primarily calcu-coefficient of variation was 9.3% at 130 pg/ml and 14.4% at lated for 30-day mortality. One-yr mortality was used for3,890 pg/ml. The lower detection limit was 5 pg/ml. construction of the Kaplan–Meier survival curves and logis- tic regression analyses. Postoperative morbidity was defined as a combined end-Anesthesia point of (A) more than 1 point in the major adverse eventsGeneral anesthesia was induced with etomidate and sufen- and complications score (MaCS) according to Schon et al.7 ¨tanil and maintained with sevoflurane and remifentanyl, (need of renal replacement therapy, reintubation, strokebefore and after CPB, and with propofol and remifenta- [Rankin disability score13 Ͼ1], and low-cardiac output syn-nyl, during CPB. If necessary, midazolam was added to drome) or (B) need for high-dependence unit (intensive careachieve the desired anesthesia depth. Beyond standard unit plus intermediate care) treatment of at least 10 days.anesthesia monitoring, including electrocardiogram, in- Comparisons between groups for univariate predictors ofvasive arterial blood pressure, and central venous pressure outcome were performed by a two-sided chi-square test withmonitoring, all patients were intraoperatively routinely Yates correction for categorical variables and Mann–Whit-equipped bihemispherically with ScO2 sensors (INVOS௡ ney test or Kruskal–Wallis test, as appropriate, for continu-Cerebral Oximeter 5100). ous variables. Correlation analyses were performed by the Spearman rank correlation test for continuous variables or Kendall’sCPB Management ␶ for scaled variables, as appropriate. Cutoff values forBefore CPB, all patients received 400 IU/kg heparin. Surgery mortality and morbidity were derived from receiver-oper-was performed in moderate hypothermia using antegrade ating curve (ROC) analyses. The lowest cutoff values forblood cardioplegia. Blood flow during CPB was adjusted to ScO2min-ox derived from the ROC analysis was used forachieve a mean arterial blood pressure between 50 –70 construction of a Kaplan–Meier survival curve with log-mmHg, a mixed venous oxygen saturation—measured at the rank test statistics.inflow of the CBP circuit— higher than 70%, and relative Backward logistic regression analyses were performed byScO2 concentrations higher than 80% of the preoperative entering additive EuroSCORE, ScO2min-ox, and CPB dura-baseline determined in the operating room when breathing tion in the model. For the latter variables, the cutoff valuesroom air. To achieve this goal, norepinephrine and nitroglyc- derived from the ROC analyses were used. EuroSCORE waserin were applied as a bolus or a continuous infusion. He- added as a graded variable with the different risk categoriesmatocrit was adjusted between 26% and 29%. Acid-base (0 –2; 3–5, 6 –10, and Ͼ10) graded as 1– 4.balance was performed after ␣-stat blood gas principles. After Analyses were performed in the total cohort of patientsweaning from CPB, protamine was applied accordingly. with and without entering the cutoff concentrations of hsTNT and NTproBNP. With respect to the low event rateSurgical Procedures for 30-day mortality, additional analyses were performed us-Of the 1,178 patients, 800 (67.9%) underwent CABG sur- ing 1-yr mortality as the dependent variable.gery (isolated CABG, 571; CABG plus aortic valve surgery,151; CABG plus other valve surgery, 62; CABG plus ascend- Resultsing aortic surgery, 4). Of the total number of patients, 378 Demographics and Surgical Course(32.1%) underwent surgery other than CABG (aortic valve Of the total number of patients, 70.5% were men. The me-surgery, 174; mitral valve surgery, 37; tricuspid valve surgery, dian age was 68 yr (59 –74 yr), weight was 81 kg (72–91 kg),Anesthesiology 2011; 114:58 – 69 60 Heringlake et al.
4. PERIOPERATIVE MEDICINEheight was 172 cm (166 –178 cm), and body mass index was ScO2min-ox derived from the measurements during oxygen27 (24 –32). The median American Society of Anesthesiol- supplementation in 1,109 patients and from room air mea-ogy classification was 3 (3–3), and median additive Euro- surements in 69 patients was 64 (55/69). A histogram ofSCORE was 5 (3– 8). Of the total number of patients, 484 ScO2min-ox values in the total cohort is given in figure 1.(41%) were graded as New York Heart Association grade In high-risk patients with an additive EuroSCOREIII/IV, and 694 (59%) were in New York Heart Association more than 10, SaO2 with room air was 96% (94 –97%).grade I/II. Mean arterial blood pressure and heart rate were 80 Cardiovascular risk factors, preoperative medication, mmHg (80 –100 mmHg) and 75 beats/min (66 –91 beats/baseline hemodynamics, hematocrit, estimated glomeru- min), respectively. ScO2min-room was 54% (49 – 60%), andlar filtration rate, hsTNT, and NTproBNP are presented ScO2min-ox was 60% (52– 65%). All oximetry results in thein table 1. high-risk group were significantly different from the total cohort. A recorded 1,058 patients (90%) were in elective surgicalstatus, with 120 (10%) of these cases classified as urgent oremergency surgery. A recorded 635 patients (54%) suffered Factors Influencing ScO2from angina pectoris, 215 (18%) had a main left stem coronary Table 2 shows that ScO2min-ox was correlated with variousstenosis, 295 (25%) had a history of myocardial infarction, 108 demographic and physiologic variables, especially age, gen-(9%) were scheduled for redo surgery, and 33 patients (2%) had der, body mass index, American Society of Anesthesiology grade, additive EuroSCORE, LVEF, glomerular filtrationan active endocarditis. Before surgery, 61 (5%) patients needed rate, and hematocrit.circulatory support (i.e., inotropic and vasopressor support or Analysis of ScO2min-ox in different EuroSCORE groupsintraaortic counterpulsation). Main variables of the surgical showed significant differences between the groups with highercourse and outcomes are presented in table 1. Thirty-day and ScO2 concentrations in the decreased additive EuroSCORE1-yr mortality and major morbidity (at least two major compli- groups (fig. 2). Comparably, patients with a decreased LVEF, acations and/or a high-dependency unit stay of at least 10 days) higher NTproBNP, and a higher hsTNT had lower ScO2min-oxwere 3.5%, 7.7%, and 13.3%, respectively. than the healthier groups of patients (fig. 2). The subgroup of high-risk patients with a EuroSCOREgreater than 10 consisted of 102 patients with a median ScO2 and Clinical Outcomesadditive EuroSCORE of 12 (11–14). The patients in this Univariate analyses of patients with postoperative mortalitygroup were older (74 yr [67–78 yr]), weighed less (75 kg and morbidity revealed that patients with an adverse out-[67–97 kg]), and were smaller (170 cm [160 –178 cm]; all: come had decreased preoperative ScO2 levels, a higher addi-P Ͻ 0.05) compared with the total cohort. tive EuroSCORE, a decreased preoperative hematocrit, and The surgical mix was comparable with the total group. In a longer duration of surgery and CPB (table 1).this group, 60.8% of patients underwent CABG surgery and ROC analysis of ScO2min-ox and 30-day mortality revealed39.2% were categorized as NO-CABG. Of these high-risk an area under the curve (AUC) of 0.71 (95% CI, 0.68 – 0.73;patients, 57.8% were graded as urgent or emergent. Dura- P Ͻ 0.0001) and a cutoff value of Յ51% (sensitivity 41.5%,tion of surgery was 308 min (261– 410 min), CPB and aortic specificity 93.6%) in the total cohort and an AUC of 0.77 (95%cross-clamp time were 150 (116 –200 min) and 111 min CI, 0.69 – 0.86; P Ͻ 0.0001) and a cutoff value of Յ53% (sen-(78 –145 min), respectively. Thirty-day and 1-yr mortality sitivity 75.0%, specificity 79.1%) in patients with a Euro-were 15.7% and 27.5%. Of these patients, 40.2% fulfilled SCORE more than 10. The respective results for the combinedthe criteria of the combined morbidity endpoint. NT- morbidity endpoint (a MaCS of at least two and/or a high-proBNP concentrations were 2,739 pg/ml (1,074 –29,346 dependency unit time of at least 10 days) revealed an AUC ofpg/ml) and hsTNT concentrations were 72.5 ng/ml (17.2– 0.68 (95% CI, 0.65– 0.70; P Ͻ 0.0001) in the total cohort. The376 ng/ml). With the exception of the surgical mix, all de- cutoff value was a ScO2min-ox of Յ60% with a sensitivity ofpicted variables were significantly (P Ͻ 0.05) different from 56.1% and a specificity of 71.4%.the total cohort. To rule out a possible confounding effect of intubation and controlled ventilation in emergency patients, the ROC analyses were also performed in the subgroup ofPreoperative Cerebral Oxygen and Arterial Saturation elective patients. The analysis of ScO2min-ox and 30-dayScO2 values available with room air were 1,123, and ScO2 values mortality revealed an AUC of 0.66 (95% CI, 0.63– 0.69)when breathing oxygen-enriched air were 1,109. The ScO2min- with a cutoff value of Յ53%, a sensitivity of 40.9%, and aroom in the total cohort when patients were breathing room air specificity of 90.7% (P ϭ 0.018). The comparable analy-with a median SaO2 of 97% (95–98%) was left 62% (57– 67%) sis in the patients with emergency status revealed an AUCand right 62% (56 – 67%). The SaO2 when breathing oxygen- of 0.74 (95% CI, 0.68 – 0.8) with a cutoff ScO2min-ox ofenriched air was 100% (100 –100%), and ScO2 with oxygen was Յ50%, a sensitivity of 47.4%, and a specificity of 91.5%left 66% (61–71%) and right 66% (61–71%). (P ϭ 0.0002).Anesthesiology 2011; 114:58 – 69 61 Heringlake et al.
5. Preoperative Cerebral Oxygen Saturation and OutcomeTable 1. Patient Characteristics, Surgical Course, Mortality, and Morbidity in 1,178 Patients Undergoing On-pumpCardiac Surgery 30-day Mortality Combined Morbidity Endpoint Patient Characteristics Total Cohort Alive Deceased Significance Not Fulfilled Fulfilled SignificanceNumber of cases 1,178 1,137 (96.5%) 41 (3.5%) 1,021 (86.7%) 157 (13.3%)Demographic data Female 348 (29.5%) 330 (29%) 18 (44%) P ϭ 0.061 290 (28.4%) 58 (36.9%) P ϭ 0.037 Male 830 (70.5%) 807 (71%) 23 (56%) 731 (71.6%) 99 (63.1%) Age (yr) 68 (59/74) 68 (59/74) 74 (70/78) P Ͻ 0.001 67 (58/74) 72 (67/77) P Ͻ 0.001 BMI (kg/m2) 27 (24/32) 27 (25/31) 26 (23/29) P ϭ 0.098 27 (25/30) 26 (24/31) P ϭ 0.013History Hyperlipidemia 699 (59.3%) 680 (59.8%) 19 (46.3%) P ϭ 0.163 608 (59.5%) 91 (57.9%) P ϭ 0.764 Arterial hypertension 995 (84.5%) 961 (84.5%) 34 (82.9%) P ϭ 0.902 855 (83.7%) 140 (89.2%) P ϭ 0.131 Diabetes mellitus 346 (29.4%) 330 (29.0%) 14 (34.1%) P ϭ 0.229 283 (27.7%) 63 (40.1%) P ϭ 0.002 Carotid stenosis 20 (1.7%) 18 (1.6%) 2 (4.9%) P ϭ 0.323 17 (1.7%) 3 (1.9%) P ϭ 0.913 Peripheral artery 80 (6.8%) 76 (6.7%) 4 (9.8%) P ϭ 0.651 61 (6.0%) 19 (12.1%) P ϭ 0.008 disease Stroke* 41 (3.5%) 38 (3.3%) 3 (7.3%) P ϭ 0.352 35 (3.4%) 6 (3.8%) P ϭ 0.986Preoperative therapy ␤-Blockers 685 (58.1%) 662 (58.2%) 23 (56.1%) P ϭ 0.648 596 (58.4%) 89 (56.7%) P ϭ 0.588 Amiodarone 22 (1.9%) 21 (1.8%) 1 (2.4%) P ϭ 0.815 20 (1.9%) 1 (0.6%) P ϭ 0.434 Antiarrhythmics 18 (1.5%) 17 (1.5%) 1 (2.4%) P ϭ 0.935 18 (1.8%) 0 (0%) P ϭ 0.207 Diuretics 542 (46.0%) 521 (45.8%) 21 (51.2%) P ϭ 0.194 459 (44.9%) 83 (52.9%) P ϭ 0.009 RAAS antagonists 753 (63.9%) 733 (64.5%) 20 (48.7%) P ϭ 0.259 658 (64.4%) 95 (60.5%) P ϭ 0.968 Nitrates po 138 (11.7%) 133 (11.7%) 5 (12.2%) P ϭ 0.915 116 (11.4%) 22 (14.0%) P ϭ 0.249 Nitrates iv 67 (5.7%) 64 (5.6%) 3 (7.3%) P ϭ 0.858 50 (4.9%) 17 (10.8%) P ϭ 0.003 Digitalis 102 (8.6%) 96 (8.4%) 6 (14.6%) P ϭ 0.163 83 (8.1%) 19 (12.1%) P ϭ 0.072 Lipid-lowering drugs 585 (49.7%) 568 (49.9%) 17 (41.4%) P ϭ 0.805 525 (51.4%) 60 (38.2%) P ϭ 0.017 Heparin/LMWH 364 (30.9%) 348 (30.6%) 16 (39%) P ϭ 0.116 307 (30.1%) 57 (36.3%) P ϭ 0.027 Warfarin 75 (6.4%) 72 (6.3%) 3 (7.3%) P ϭ 0.908 60 (5.9%) 15 (9.6%) P ϭ 0.065 Clopidogrel 147 (12.5%) 143 (12.6%) 4 (9.8%) P ϭ 0.972 128 (12.5%) 19 (12.1%) P ϭ 0.981 Aspirin 623 (52.9%) 613 (53.9%) 19 (46.3%) P ϭ 0.976 561 (54.9%) 71 (45.2%) P ϭ 0.161 Clonidine 12 (1.0%) 12 (1.0%) 0 (0%) P ϭ 0.832 11 (1.1%) 1 (0.6%) P ϭ 0.997 Calcium antagonists 223 (18.9%) 219 (19.3%) 4 (9.8%) P ϭ 0.292 198 (19.4%) 25 (15.9%) P ϭ 0.581 Direct vasodilatators 40 (3.4%) 39 (3.4%) 1 (2.4%) P ϭ 0.812 38 (3.7%) 2 (1.3%) P ϭ 0.221 Risk stratification NYHA III/IV 484 (41.1%) 461 (40.5%) 23 (56.1%) P ϭ 0.051 398 (39.0%) 86 (54.7%) P Ͻ 0.001 NYHA I/II 694 (58.9%) 676 (59.5%) 18 (43.9%) 618 (60.5%) 69 (43.9%) LVEF Ͻ30% 54 (4.6%) 46 (4.0%) 8 (19.5%) P Ͻ 0.001 39 (3.8%) 15 (9.6%) P ϭ 0.003 LVEF Ͻ50% 277 (23.5%) 261 (22.9%) 16 (39%) P ϭ 0.022 244 (23.9%) 53 (33.7%) P ϭ 0.002 ASA 3 (3/3) 3 (3/3) 3 (3/4) P Ͻ 0.001 3 (3/3) 3 (3/4) P Ͻ 0.001 Additive EuroSCORE 5 (3/8) 5 (3/7) 10 (7/12) P Ͻ 0.001 5 (3/7) 8 (6/11) P Ͻ 0.001 Angina pectoris 635 (53.9%) 617 (54.3%) 18 (43.9%) P ϭ 0.248 92.3 88 (56.0%) P ϭ 0.630 Previous MI 295 (25.0%) 280 (24.6%) 15 (36.6%) P ϭ 0.098 243 (23.8%) 52 (33.1%) P ϭ 0.012 Previous cardiac 108 (9.2%) 103 (9.0%) 5 (12.2%) P ϭ 0.683 86 (8.4%) 22 (14.0%) P ϭ 0.035 surgery Left main stem stenosis 215 (18.2%) 211 (18.5%) 4 (9.7%) P ϭ 0.219 190 (18.6%) 25 (15.9%) P ϭ 0.505 Active endocarditis 33 (2.8%) 31 (2.7%) 2 (4.9%) P ϭ 0.735 21 (2.0%) 12 (7.6%) P Ͻ 0.001 Circulatory support† 61 (5.2%) 49 (4.3%) 12 (29.3%) P Ͻ 0.001 38 (3.7%) 23 (14.6%) P Ͻ 0.001Type of surgery CABG 800 (67.9%) 772 (67.9%) 28 (68.3%) P ϭ 0.888 692 (67.8%) 110 (70.1%) P ϭ 0.631 No CABG 378 (32.1%) 365 (33.0%) 13 (31.7%) 329 (32.2%) 47 (29.9%)Surgical priority Elective 1,058 (89.8%) 1,032 (90.7%) 26 (63.4%) P Ͻ 0.001 943 (92.4%) 115 (73.2%) P Ͻ 0.001 Ͻ12 h and emergency 120 (10.2%) 105 (9.3%) 15 (36.6%) 78 (7.6%) 42 (26.8%)Physiological status MAP (mmHg) 95 (87/103) 95 (87/103) 94 (80/100) P ϭ 0.276 95 (87/103) 93 (83/103) P ϭ 0.058 Heart rate (/min) 69 (63/81) 69 (62/80) 84 (75/93) P Ͻ 0.001 69 (61.5/78) 76 (65/87) P Ͻ 0.001 SaO2 initial (%) 97 (95/98) 97 (95/98) 96 (95/97) P ϭ 0.058 97 (95/98) 96 (95/97) P Ͻ 0.001ScO2 room air n ϭ 1,123 Right (%) 62 (57/67) 62 (57/67) 55 (47/62) P Ͻ 0.001 62 (58/67) 57 (52/62) P Ͻ 0.001 Left (%) 62 (56/67) 62 (57/67) 55 (49/65) P ϭ 0.002 62 (57/67) 57 (51/62) P Ͻ 0.001 Minimal left or right (%) 60 (55/64) 60 (55/65) 54 (47/61) P Ͻ 0.001 61 (56/65) 56 (49/60) P Ͻ 0.001ScO2 with oxygen n ϭ 1,109 Right (%) 66 (61/71) 66 (61/71) 60 (50/68) P Ͻ 0.001 66 (62/72) 62 (56/67) P Ͻ 0.001 Left (%) 66 (61/71) 66 (61/71) 61 (51/67) P ϭ 0.001 67 (62/72) 61 (55/67) P Ͻ 0.001Minimal ScO2 with oxygen n ϭ 1,178 (ScO2min-ox)‡ScO2min-ox (left or right) (%) 65 (59/69) 64 (59/69) 58 (47/64ox) P Ͻ 0.001 65 (60/70) 60 (53/65) P Ͻ 0.001 (continued)Anesthesiology 2011; 114:58 – 69 62 Heringlake et al.
6. PERIOPERATIVE MEDICINETable 1. Continued 30-day Mortality Combined Morbidity Endpoint Patient Characteristics Total Cohort Alive Deceased Significance Not Fulfilled Fulfilled SignificanceLaboratory analyses NTproBNP (pg/ml) 501 (154/1,471) 463 (151/1,309) 2,330 (1,567/9,474) P Ͻ 0.001 427 (137/1,143) 1,459 (419/4,882) P Ͻ 0.001 n ϭ 1,011 hsTNT (ng/ml) n ϭ 985 14 (7/33) 13 (7/30) 54 (30/582) P Ͻ 0.001 12 (7/26) 35 (15/124) P Ͻ 0.001 Plasma creatinin (mg/dl) 0.93 (0.80/1.13) 0.92 (0.78/1.12) 1.06 (0.83/1.31) P ϭ 0.029 0.92 (0.78/1.01) 1.02 (0.80/1.33) P Ͻ 0.001 n ϭ 1,175 eGFR§ (ml/min/1.73 m2) 80 (63/99) 81 (64/99) 66 (51/82) P ϭ 0.002 82 (66/100) 67 (50/87) P Ͻ 0.001 n ϭ 1,175 Hemoglobin (g/l) 135 (124/146) 135 (123/146) 129 (104/146) P ϭ 0.017 136 (125/146) 129 (115/141) P Ͻ 0.001 n ϭ 1,132 Hematocrit (%) 39 (36/42) 39 (36/42) 37 (31/42) P ϭ 0.033 39 (36/42) 37 (34/42) P Ͻ 0.001 n ϭ 1,132Intraoperative course Duration surgery (min) 249 (209/300) 246 (208/295) 342 (267/426) P Ͻ 0.001 243 (207/290) 306 (241/383) P Ͻ 0.001 Duration CPB (min) 109 (88/141) 108 (88/139) 169 (117/226) P Ͻ 0.001 107 (87/136) 138 (105/191) P Ͻ 0.001 Aortic cross-clamp time 85 (68/110) 85 (67/109) 103 (84/137) P ϭ 0.001 84 (67/107) 102 (78/138) P Ͻ 0.001 (min)Data are given as median and (25/75) percentiles for continuous variables and as percentage for categorical variables. The combinedmorbidity endpoint was defined as Ն2 complications (need of new renal replacement therapy, reintubation, stroke (RANKIN grade Ͼ1),and low-cardiac output syndrome) and/or need for a high-dependency unit treatment of Ն10 days.* History of stroke defined as former ischemic cerebral infarction or hemorrhagic stroke with a Rankin grade Ͼ1; † circulatory supportwith intravenous inotropes, vasopressors, or intraaortic counterpulsation; ‡ minimal ScO2 either with or without oxygen; § eGFRcalculated by the Modifications of Diet in Renal Disease equation.ASA ϭ American Society of Anesthesiology grading; BMI ϭ body mass index; CABG ϭ coronary artery bypass grafting; CPB ϭcardiopulmonary bypass; eGFR ϭ estimated glomerular filtration rate; hsTNT ϭ high-sensitivity troponin T; LMWH ϭ low-molecularweight heparine; LVEF ϭ left ventricular ejection fraction; MAP ϭ mean arterial blood pressure; MI ϭ myocardial infarction; NT-proBNP ϭ N-terminal pro-B-type natriuretic peptide; NYHA ϭ New York Heart Association grading; RAAS antagonists ϭ renin-angiotensin-aldosterone system; SaO2 ϭ arterial oxygen saturation; ScO2 ϭ cerebral oxygen saturation. To determine whether the capability of increasing ScO2 In the high-risk group with an additive EuroSCOREconcentrations during oxygen supplementation in patients more than 10, ScO2min-ox had a better accuracy for 30-daywith a low ScO2min-room had prognostic relevance, patients mortality than the EuroSCORE (EuroSCORE, 0.54 [95%presenting with a ScO2min-room of Յ51% and showing an CI, 0.43– 0.64%]; ScO2min-ox, 0.79 [CI 95%, 0.7– 0.86%]; ␦increase to ScO2min-ox concentrations higher than 51% dur- AUC 0.25, P ϭ 0.0044) (fig. 3). The predictive capacity foring oxygen application were compared with those patients morbidity in the high-risk group was not different betweenthat remained below 51% ScO2min-ox. Of 142 patients, 81 the EuroSCORE and the ScO2min-ox (EuroSCORE, 0.61;with a ScO2min-room of Յ51% responded to oxygen by in- [95% CI, 0.51– 0.71%]; ScO2min-ox: 0.64 [CI 95%, 0.57–creasing ScO2min-ox to concentrations higher than 51%; in 0.76%]; ␦ AUC 0.06, P ϭ n.s.).61 patients, ScO2min-ox remained lower or equal than 51%.Mortality and morbidity were significantly higher in the Predictive Accuracy of hsTNT and NTproBNPnonresponders than in the responders (30-day mortality: ROC analyses of hsTNT and 30-day mortality revealed an16.4% vs. 2.5%, P ϭ 0.0081, combined morbidity end- AUC of 0.8 (95% CI, 0.77– 0.82%; P Ͻ 0.001) and an optimalpoint; a MaCS of at least two and/or a high-dependency unit cutoff value of 28 pg/ml with a sensitivity of 81.8% and a spec-time of at least 10 days: 37.7% vs. 14.8%, P ϭ 0.0033). ificity of 73.8% in the total cohort. The respective results in the high-risk group (EuroSCORE more than 10) were: AUC of 0.71 (95% CI: 0.59 – 0.81%), P ϭ 0.002, a cutoff of 37.1Predictive Accuracy of ScO2 versus Additive EuroSCORE pg/ml, a sensitivity of 100%, and a specificity of 50%.Comparative ROC analyses of additive EuroSCORE with re- ROC analysis of the predictive capacity of hsTNT for thespect to 30-day mortality and the combined morbidity end- combined morbidity endpoint showed an AUC of 0.72point in the total cohort revealed an AUC of 0.82 (95% CI, (95% CI, 0.69 – 0.75%), P Ͻ 0.001, and an optimal cutoff0.8 – 0.84%) for mortality and an AUC of 0.77 (95% CI: 0.74 – of 21.8 pg/ml with a sensitivity of 62.7%, and a specificity of0.79%) for morbidity. A comparison with ScO2min showed that 71.5%. The respective analysis in the high-risk group failedthe EuroSCORE had a better accuracy in predicting 30-day to reach statistical significance.mortality (difference between areas, 0.11; P ϭ 0.015; fig. 3) and ROC analysis of NTproBNP and 30-day mortalitymorbidity (difference between areas, 0.09; P Ͻ 0.001) in the showed an AUC of 0.78 (95% CI, 0.76 – 0.81%), P Ͻ 0.001,total cohort. Comparable results were obtained in patients with and a cutoff concentration of 1,743 pg/ml with a sensitivity ofan additive EuroSCORE below 10 (data not shown). 75.8% and a specificity of 79.9% in the total cohort. The ROCAnesthesiology 2011; 114:58 – 69 63 Heringlake et al.
7. Preoperative Cerebral Oxygen Saturation and Outcome Comparative analyses on the accuracy of NTproBNP, hsTNT, and ScO2min-ox were performed in the subgroup of patients with complete hormone analysis and revealed no significant differences in the accuracy for predicting 30-day and 1-yr mortality in the total cohort (data not shown). Logistic Regression Backward logistic regression analyses for 30-day and 1-yr mor- tality was performed using a model with the EuroSCORE graded in different risk categories (0 –2, 3–5, 6 –10, Ͼ10) and the lowest cutoff value derived from the different ROC analyses for ScO2min-ox (Յ50%). As a marker for the severity of the sur- gical insult, CPB time was added to the model. The cutoff valueFig. 1. Histogram of the lowest preoperative cerebral oxygensaturation measured during oxygen insufflation (ScO2min-ox) in for 30-day mortality derived from ROC analysis for CPB timea cohort of 1,178 cardiac surgery patients. was more than 114 min with an AUC of 0.76, a sensitivity of 82.9%, and a specificity of 57.2% (P Ͻ 0.0001). To allow easieranalysis of NTproBNP and mortality in the high-risk group was comparisons in future studies, the CPB time cutoff was set tonot significant. ROC analysis of NTproBNP and morbidity more than 120 min. The 30-day mortality rates in the differentrevealed an AUC of 0.7 (95% CI, 0.67– 0.73%), P Ͻ 0.001, EuroSCORE categories were: EuroSCORE 0 –2, 0.3%; Euro-and a cutoff concentration of 1,061 pg/ml with a sensitivity of SCORE 3–5, 1.3%; EuroSCORE 6 –10, 5.2%; EuroSCORE more than 10, 16.7%.61.2% and a specificity of 74.4%. With respect to the fact that preoperative determination of NTproBNP and hsTNT cannot be considered routine inTable 2. Relationship among Preoperative Cerebral many European heart centers, calculations were either per-Oxygen Saturation, Demographics, and Variables formed with or without inclusion of NTproBNP andRepresentative of Cardiovascular Status or Risk hsTNT. The results of the logistic regression analyses are given in table 3, showing that a ScO2min-ox equal or less than Spearman’s Rank 50% is an independent predictor of 30-day mortality, if con- Test/Coefficient ventional risk factors are used, and of 30-day, as well as 1-yr, Variable of Contingence Significance mortality, if NTproBNP and hsTNT are entered into theAge Ϫ0.2 P Ͻ 0.0001 model. The category EuroSCORE 3–5 was not included inGender 0.09* P ϭ 0.0026 the regression models because of the low event rate (5 deathsBody mass index 0.19 P Ͻ 0.0001 per 401 patients).Surgical priority 0.032* P ϭ 0.267 urgent/emergencyCABG vs. NO-CABG 0.02* P ϭ 0.521 Kaplan–Meier Survival AnalysesPreoperative circulatory 0.04* P ϭ 0.225 The effects of ScO2 on long-term mortality using a ScO2min-ox support (IABP, Յ50% as a cutoff value was analyzed by Kaplan–Meier sta- Inotropes, Vasopressors) tistics. Log-rank test revealed a significant effect in the totalHeart rate Ϫ0.03 P ϭ 0.342LVEF Ͻ30% 0.09* P ϭ 0.0016 cohort as well as in the high-risk group with a EuroSCORECreatinine (plasma) Ϫ0.12 P ϭ 0.0001 more than 10 (fig. 4).eGFR 0.19 P Ͻ 0.0001NTproBNP Ϫ0.35 P Ͻ 0.0001hsTNT Ϫ0.28 P Ͻ 0.0001 DiscussionHemoglobin 0.37 P Ͻ 0.0001 Increasing evidence suggests that decreased ScO2 during sur-Hematocrit 0.34 P Ͻ 0.0001Preoperative heparin/ 0.07* P ϭ 0.022 gery is associated with unwarranted neurologic effects and a LMWH* complicated postoperative course.1,3 Most studies in this Kendall’s ␶ field have only focused on the relative changes of ScO2 duringASA Ϫ0.10 P Ͻ 0.0001 cardiac6,14,15 and noncardiac surgery.16,17 To what extentAdditive EuroSCORE Ϫ0.20 P Ͻ 0.0001 decreased preoperative ScO2 concentrations are related to* Coefficient of contingence: chi-square test. postoperative complications and mortality has not been for-ASA ϭ American Society of Anesthesiology grading; CABG ϭ mally analyzed in a large cohort of patients.coronary artery bypass grafting; eGFR ϭ estimated glomerular The findings of the current study, showing that preoperativefiltration rate; hsTNT ϭ high-sensitivity troponin T; IABP ϭ in- ScO2 levels are closely related to relevant measures of cardiopul-traaortic ballon pump; LMWH ϭ lowmolecular weight heparine;LVEF ϭ left ventricular ejection fraction; NTproBNP ϭ N-terminal monary function, postoperative morbidity, and short- and long-pro-B-type natriuretic peptide. term mortality, may help to better understand the role of peri-Anesthesiology 2011; 114:58 – 69 64 Heringlake et al.
8. PERIOPERATIVE MEDICINEA BC D Fig. 3. Receiver-operating curve analyses of minimal preop- erative cerebral oxygen saturation during oxygen insufflation (ScO2min-ox, broken line) and additive EuroSCORE (unbroken line) for 30-day mortality in the total cohort of 1,178 patients (A) and in 102 high-risk patients with a EuroSCORE more than 10 (B) showing a significantly better predictive accuracy of the EuroSCORE in the total cohort (P ϭ 0.015) and of ScO2min-ox in the high-risk group (P ϭ 0.0044).Fig. 2. Preoperative oxygen-supplemented cerebral oxygensaturation (ScO2min-ox) in different additive EuroSCORE risk had lower ScO2 concentrations than healthy control pa-groups (A), according to preoperative left ventricular ejectionfraction (LVEF) (B), in different quartiles of the plasma con- tients and that ScO2 increased during heart failure thera-centration of high-sensitivity troponin T (hsTNT) (C), and N- py.24 Koike et al. showed that the decrease in ScO2 duringterminal pro-B-type natriuretic peptide (NTproBNP) (D). * Sig- exercise in patients with coronary artery disease had prog-nificant difference (P Ͻ 0.05) compared with the respective nostic effect and was associated with an unfavorablegroups (Kruskal–Wallis test). Data are given as median and outcome in these patients.25 Paquet et al. observed95% confidence interval of the median. that baseline ScO2 was related to LVEF in patients under- going cardiac surgery and that ScO2 had an acceptableoperative changes in ScO2 and may have implications for accuracy to identify left ventricular dysfunction in theseimproving risk stratification in cardiac surgery. patients.26 From a technical point of view, ScO2, determined by near- The findings of the current study, showing significantinfrared spectroscopy, is reflective of tissue oxygen saturation correlations between NTproBNP, hsTNT, LVEF, and ScO2in a small sample of the frontal cortex, and, with respect to in a large patient sample, support the concept that ScO2 isthe distribution of arterial and venous blood, weighted by the influenced by myocardial function and the performance ofvenous compartment.1,18 Various experimental and clinical the cardiopulmonary system.studies have clearly shown that ScO2 determined by near- This may have clinical and scientific implications. First, mostinfrared spectroscopy is directly related to cerebral blood guidelines on ScO2 monitoring for guiding hemodynamics andflow and jugular venous oxygen saturation (SvO2).19,20 CPB management during cardiac surgery suggest the adjust- Current evidence supports that ScO2 also indirectly re- ment of therapy according to relative changes from the preop-flects the adequacy of the systemic circulation and cardiopul- erative baseline (i.e., to maintain relative ScO2 concentrationsmonary function. In line with other studies comparing ScO2 higher than 80% of baseline).27 Based on our findings, thisand mixed SvO2 in children,21 we were able to show a close concept may be appropriate for patients starting with a ScO2 incorrelation between ScO2 and SvO2 in extubated patients a healthy range but may be inadequate for patients presentingafter cardiac surgery.** Comparably, despite the absolute lev- with decreased preoperative ScO2 as an expression of severe car-els of ScO2 and SvO2 that were only moderately correlated, diopulmonary dysfunction. Second, these findings offer scien-Dullenkopf et al. showed that trends in ScO2 reflect changes tific perspectives for using ScO2 as a noninvasive technology forin ScO2 in intubated patients immediately after cardiac sur- guiding therapy in patients with heart failure as already sug-gery.22 Skhirtladze et al. observed that patients with se- gested by Madsen et al. more than 10 yr ago.24verely reduced LVEF being scheduled for implantation of The accuracy of ScO2 for predicting 30-day mortality ina defibrillator cardioverter system had lower ScO2 concen- the total cohort was lower than the additive EuroSCORE buttrations than patients with a preserved LVEF.23 Madsen et comparable with established (NTproBNP)28,29 and emerg-al. showed that patients presenting with acute heart failure ing (hsTNT)10,30 humoral markers of postoperative risk in ** Poster presentation by Julika Schon, M.D., Department of ¨ cardiac surgery. As derived from the ROC and the Kaplan–Anesthesiology, University of Lubeck, Lubeck, Germany at the au- ¨ ¨ Meier analyses, the accuracy of ScO2 for predicting short- andtumn meeting of the Working Group in Cardiac Anesthesia of theGerman Association of Anesthesiology and Intensive Care Medi- long-term prognosis in high-risk patients is clearly superiorcine, November 12–14, 2009, Ulm, Germany. to the EuroSCORE. It is noteworthy that patients with aAnesthesiology 2011; 114:58 – 69 65 Heringlake et al.
9. Preoperative Cerebral Oxygen Saturation and OutcomeTable 3. Results of Logistic Regression Models for 30-day and 1-year Mortality 30-day Mortality Excluding 30-day Mortality Including 1-yr Mortality Including Variables NTproBNP and hsTNT NTproBNP and hsTNT NTproBNP and hsTNTNumber of patients n ϭ 1,178 n ϭ 984 n ϭ 984Number of events n ϭ 41 n ϭ 33 N ϭ 72EuroSCORE 0–2 Reference Reference ReferenceEuroSCORE 3–5 Not included in the model Not included in the model Not included in the modelEuroSCORE 6–10 4.9 (1.7–13.2) P ϭ 0.0015 3.3 (1.1–10.2) P ϭ 0.034 2.1 (1.1–4.2) P ϭ 0.031EuroSCORE Ͼ10 12.0 (4.5–31.9) P Ͻ 0.0001 6.4 (2.1–19.8) P ϭ 0.006 5.7 (2.7–11.6) P Ͻ 0.0001ScO2 min-ox Յ50% 4.6 (2.2–9.8) P ϭ 0.0001 2.5 (1.0–6.0) P ϭ 0.044 2.6 (1.2–4.2) P ϭ 0.0077Duration of CPB* 2.0 (0.97–4.4) P ϭ 0.06 Not included in the model 1.7 (0.97–2.9) P ϭ 0.062 Ͼ120 minhsTNT Ͼ28 pg/ml Not applicable 3.6 (1.4–9.2) P ϭ 0.008 1.8 (1.0–3.3) P ϭ 0.042NTproBNP Not applicable 3.4 (1.3–8.6) P ϭ 0.01 2.2 (1.2–4.1) P ϭ 0.01 Ͼ1,743 pg/mlOverall model fit: 73.5 P Ͻ 0.0001 76.8 P Ͻ 0.0001 107.4 P Ͻ 0.0001 Chi-square SignificanceResults of the backward logistic regression are given as odds ratio and 95% confidence interval in parentheses. The models werecalculated using additive EuroSCORE (graded in four categories of different mortality risks: 0 –2; 3–5; 6 –10; Ͼ10) and the cutoff valuesderived from receiver-operating curve analyses of the current study.* The cutoff for 30-day mortality derived from the receiver-operating curve analysis for duration of CPB (114 min) was set to 120 minfor better comparability.CPB ϭ cardiopulmonary bypass; hsTNT ϭ high-sensitivity troponin T; NTproBNP ϭ N-terminal pro-B-type natriureti-peptide; ScO2min-ox ϭminimal cerebral oxygen saturation derived when breathing oxygen enriched air.EuroSCORE more than 10 and a ScO2min-ox equal or less these models, comparable with the EuroSCORE, have beenthan 50% had a 1-yr mortality that was twice as high as in developed and validated in patients undergoing CABG sur-high-risk patients with preserved—more than 50%— ScO2 gery, and the predictive accuracy for patients undergoingconcentrations. This is a remarkable finding because risk other types of surgery is often limited. In contrast, the Societystratification in high-risk cardiac surgery patients by Euro- of Thoracic Surgeons score is available in different versionsSCORE, as frequently performed in most European heart for various kinds of cardiac procedures. This score has ancenters, is limited by the fact that the additive and the logistic excellent predictive capacity but needs a large number ofEuroSCORE overestimate mortality in this group of pa- variables to be entered.34 –36 Consequently, at least in Eu-tients.31,32 Consequently, our findings, if reproduced pro- rope, it is less frequently used than the EuroSCORE.spectively, may have relevant implications for improving risk Focusing more on physiologic than historical variables,stratification in high-risk cardiac surgery patients. Ranucci et al. have shown that a simple score consisting of the To rule out a possible confounding effect of arterial hyp- variables age, creatinine, and left ventricular ejection fraction hasoxia on the results of ScO2 monitoring, we measured ScO2 an accuracy for 30-day mortality that is comparable withwith room air and during oxygen insufflation. The difference the additive and the logistic EuroSCORE.37 This suggestsbetween the median results with both measures was 4% and that risk stratification based on age and physiologic vari-comparable with the 3% difference in transcutaneous oxygen ables is feasible and may at least be as predictive as a classicsaturation achieved with application. Numerically, the accu- risk score.racy for predicting 30-day mortality was not different if the ScO2 Our findings, regarding the discriminatory capacity of NT-values derived with room air or during oxygen supplementation proBNP for predicting 30-day mortality, are in line with previ-were used (data not shown). However, an analysis of whether ous observations in cardiac28,29 as well as noncardiac patients.29the capability to increase ScO2 during oxygen supplementation The accuracy of NTproBNP was comparable with the additivein patients with a low ScO2 during room air had prognostic EuroSCORE and ScO2min. The latter also holds true forrelevance revealed that nonresponders (i.e., patients that failed hsTNT, an observation that is remarkable in so far as this pre-to increase ScO2 beyond the cutoff value) had a significantly clinical test has not been used in patients undergoing cardiachigher mortality and morbidity than those showing an in- surgery before. It is also of note that the hsTNT cutoff forcrease beyond the ScO2 cutoff. This suggests that if ScO2 will 30-day mortality was highly comparable with the one derivedbe used for risk stratification, this parameter should ideally be from a large trial in medical patients with heart failure,10 furtherdetermined during oxygen supplementation. suggesting that hsTNT serves not only as a marker of myocar- Besides the EuroSCORE, various other risk stratification dial ischemia but, comparable with NTproBNP, as a measure ofmodels are used in cardiac surgery.33 Currently, most of global cardiopulmonary dysfunction.Anesthesiology 2011; 114:58 – 69 66 Heringlake et al.
10. PERIOPERATIVE MEDICINE A BFig. 4. One-yr Kaplan–Meier survival curves of patients undergoing on-pump cardiac surgery with (red) or without (blue) apreoperative cerebral oxygen saturation (ScO2min-ox) during oxygen insufflation equal or less than 50% absolute in the totalcohort of 1,178 patients (A) and in a subgroup of 102 patients with an additive EuroSCORE more than 10 (B). Log-rank testindicates that groups differ, both the total cohort and the subgroup with a EuroSCORE more than 10.Limitations used the ScO2min-room values from the missing 69 patients forThe current study has limitations. calculation of ScO2min-ox to allow calculations on the maxi- mal number of patients with adverse events.1. The short-time mortality in the study population was 4. The modeling strategy of the current study was designed low. Thus, the results presented, despite being statistically significant, are prone to potential error and require pro- to explore associations, and many of the analyses are based spective confirmation. However, our findings are sub- on post hoc and subgroup analyses of the primary dataset. stantiated by the fact that when analyzing 1-yr mortality, Consequently, the results of the current study require the number of patients with an adverse outcome was prospective replication in an independent sample. more than twice as high and remained statistically signif- icant not only when focusing on the total cohort but also Conclusion on the subsets of patients with different surgical priority.2. The ScO2 monitors used in the current study (INVOS௡ In conclusion, the current study shows that preoperative 4100 and 5100) only have a Federal Drug Administration ScO2 levels determined by near-infrared spectroscopy are re- approval for trend monitoring of ScO2 and not for abso- lated to objective measures of cardiopulmonary function and lute values. Recently, another ScO2 monitoring system that low preoperative ScO2 concentrations are associated with (Fore-Sight௡; CAS Medical Systems, Branford, CT) has an adverse perioperative course. A ScO2min-ox equal or less been introduced to the market that, by using four instead than 50% is an independent predictor of short- and long- of three wavelengths for determination of oxygenated and term mortality in patients undergoing on-pump cardiac sur- deoxygenated hemoglobin and laser light instead of a gery and might serve as a refined marker for preoperative risk light emitting diode, has been suggested to be capable of stratification in cardiac surgery patients. measuring absolute ScO2 concentrations.3 No compara- The members of the cardiac anesthesia unit and the personnel of the tive data of ScO2 concentrations derived by these different Department of Clinical Chemistry (University of Lubeck, Lubeck, ¨ ¨ monitors are available, thus it remains currently specula- Germany) are thanked for their help in data and blood sample tive whether one system is more precise than the other. acquisition. However, it is likely that any measure improving the re- liability of the near-infrared spectroscopy technology will also improve the accuracy of ScO2 for risk stratification. References3. Unfortunately, ScO2 concentrations with oxygen were only 1. Murkin JM, Arango M: Near-infrared spectroscopy as an in- available for 1,109 patients and were missing in some pa- dex of brain and tissue oxygenation. Br J Anaesth 2009; 103(suppl 1):i3–13 tients that have died. However, with respect to the observa- 2. Casati A, Spreafico E, Putzu M, Fanelli G: New technology for tion that patients with a poor prognosis showed only minor noninvasive brain monitoring: Continuous cerebral oxime- increases of ScO2 during oxygen supplementation, we have try. Minerva Anestesiol 2006; 72:605–25Anesthesiology 2011; 114:58 – 69 67 Heringlake et al.
11. Preoperative Cerebral Oxygen Saturation and Outcome 3. Fischer GW: Recent advances in application of cerebral 18. Edmonds HL Jr, Ganzel BL, Austin EH 3rd: Cerebral oximetry oximetry in adult cardiovascular surgery. Semin Cardiotho- for cardiac and vascular surgery. Semin Cardiothorac Vasc rac Vasc Anesth 2008; 12:60 –9 Anesth 2004;8:147– 66 4. Slater JP, Guarino T, Stack J, Vinod K, Bustami RT, Brown JM 19. Kurth CD, Levy WJ, McCann J: Near-infrared spectroscopy 3rd, Rodriguez AL, Magovern CJ, Zaubler T, Freundlich K, cerebral oxygen saturation thresholds for hypoxia-ischemia Parr GV: Cerebral oxygen desaturation predicts cognitive in piglets. J Cereb Blood Flow Metab 2002; 22:335– 41 decline and longer hospital stay after cardiac surgery. Ann 20. Kim MB, Ward DS, Cartwright CR, Kolano J, Chlebowski S, Thorac Surg 2009; 87:36 – 44 Henson LC: Estimation of jugular venous O2 saturation from 5. Olsson C, Thelin S: Regional cerebral saturation monitoring cerebral oximetry or arterial O2 saturation during isocapnic with near-infrared spectroscopy during selective antegrade hypoxia. J Clin Monit Comput 2000; 16:191–9 cerebral perfusion: Diagnostic performance and relationship 21. Nagdyman N, Ewert P, Peters B, Miera O, Fleck T, Berger F: to postoperative stroke. J Thorac Cardiovasc Surg 2006; Comparison of different near-infrared spectroscopic cerebral 131:371–9 oxygenation indices with central venous and jugular venous 6. Murkin JM, Adams SJ, Novick RJ, Quantz M, Bainbridge D, oxygenation saturation in children. Paediatr Anaesth 2008; Iglesias I, Cleland A, Schaefer B, Irwin B, Fox S: Monitoring 18:160 – 6 brain oxygen saturation during coronary bypass surgery: A randomized, prospective study. Anesth Analg 2007; 104: 22. Dullenkopf A, Baulig W, Weiss M, Schmid ER: Cerebral near- 51– 8 infrared spectroscopy in adult patients after cardiac surgery is not useful for monitoring absolute values but may reflect 7. Schon J, Serien V, Hanke T, Bechtel M, Heinze H, Groesdonk ¨ trends in venous oxygenation under clinical conditions. HV, Sedemund-Adib B, Berger KU, Eleftheriadis S, Heringlake J Cardiothorac Vasc Anesth 2007; 21:535–9 M: Cerebral oxygen saturation monitoring in on-pump car- diac surgery: A 1-year experience. Appl Cardiopulm Patho- 23. Skhirtladze K, Birkenberg B, Mora B, Moritz A, Ince I, Ank- physiol 2009; 13:243–52 ersmit HJ, Steinlechner B, Szeinlechner B, Dworschak M: Cerebral desaturation during cardiac arrest: Its relation to 8. Roques F, Nashef SA, Michel P, Gauducheau E, de Vincentiis arrest duration and left ventricular pump function. Crit Care C, Baudet E, Cortina J, David M, Faichney A, Gabrielle F, Med 2009; 37:471–5 Gams E, Harjula A, Jones MT, Pintor PP, Salamon R, Thulin L: Risk factors and outcome in European cardiac surgery: Anal- 24. Madsen PL, Nielsen HB, Christiansen P: Well-being and ce- ysis of the EuroSCORE multinational database of 19030 pa- rebral oxygen saturation during acute heart failure in hu- tients. Eur J Cardiothorac Surg 1999; 15:816 –22 mans. Clin Physiol 2000; 20:158 – 64 9. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D: A 25. Koike A, Nagayama O, Hoshimoto-Iwamoto M, Suzuki T, more accurate method to estimate glomerular filtration rate Tajima A, Uejima T, Aizawa T: Clinical significance of cere- from serum creatinine: A new prediction equation. Modifi- bral oxygenation during exercise in patients with coronary cation of Diet in Renal Disease Study Group. Ann Intern Med artery disease. Circ J 2008; 72:1852– 8 1999; 130:461–70 26. Paquet C, Deschamps A, Denault AY, Couture P, Carrier M,10. Latini R, Masson S, Anand IS, Missov E, Carlson M, Vago T, Babin D, Levesque S, Piquette D, Lambert J, Tardif JC: Base- Angelici L, Barlera S, Parrinello G, Maggioni AP, Tognoni G, line regional cerebral oxygen saturation correlates with left Cohn JN, Val-HeFT Investigators: Prognostic value of very ventricular systolic and diastolic function. J Cardiothorac low plasma concentrations of troponin T in patients with Vasc Anesth 2008; 22:840 – 6 stable chronic heart failure. Circulation 2007; 116:1242–9 27. Denault A, Deschamps A, Murkin JM: A proposed algorithm11. Heringlake M, Kox T, Poeling J, Klaus S, Hanke T, Franz N, for the intraoperative use of cerebral near-infrared spectros- Eberhardt F, Heinze H, Armbruster FP, Bahlmann L: The copy. Semin Cardiothorac Vasc Anesth 2007; 11:274 – 81 effects of physical exercise on plasma levels of relaxin, 28. Yeh HM, Lau HP, Lin JM, Sun WZ, Wang MJ, Lai LP: Preop- NTproANP, and NTproBNP in patients with ischemic heart erative plasma N-terminal pro-brain natriuretic peptide as a disease. Eur J Med Res 2009; 14:106 –12 marker of cardiac risk in patients undergoing elective non-12. Shanmugam G, West M, Berg G: Additive and logistic Euro- cardiac surgery. Br J Surg 2005; 92:1041–5 SCORE performance in high risk patients. Interact Cardio- 29. Cuthbertson BH, Croal BL, Rae D, Gibson PH, McNeilly JD, vasc Thorac Surg 2005; 4:299 –303 Jeffrey RR, Smith WC, Prescott GJ, Buchan KG, El-Shafei H,13. Rankin J: Cerebral vascular accidents in patients over the age Gibson GA, Hillis GS: N-terminal pro-B-type natriuretic pep- of 60. II. Prognosis. Scott Med J 1957; 2:200 –15 tide levels and early outcome after cardiac surgery: A pro-14. Yao FS, Tseng CC, Ho CY, Levin SK, Illner P: Cerebral oxygen spective cohort study. Br J Anaesth 2009; 103:647–53 desaturation is associated with early postoperative neuropsy- 30. Omland T, de Lemos JA, Sabatine MS, Christophi CA, Rice chological dysfunction in patients undergoing cardiac sur- MM, Jablonski KA, Tjora S, Domanski MJ, Gersh BJ, Rouleau gery. J Cardiothorac Vasc Anesth 2004; 18:552– 8 JL, Pfeffer MA, Braunwald E, Prevention of Events with15. Edmonds HL Jr: Detection and treatment of cerebral hypoxia Angiotensin Converting Enzyme Inhibition (PEACE) Trial In- key to avoiding intraoperative brain injuries. J Clin Monit vestigators: A sensitive cardiac troponin T assay in stable Comput 2000;16:69 –74 coronary artery disease. N Engl J Med 2009; 361:2538 – 4716. Casati A, Fanelli G, Pietropaoli P, Proietti R, Tufano R, Dane- 31. Parolari A, Pesce LL, Trezzi M, Cavallotti L, Kassem S, Loardi C, lli G, Fierro G, Fierro G, De Cosmo G, Servillo G: Continuous Pacini D, Tremoli E, Alamanni F: EuroSCORE performance in valve monitoring of cerebral oxygen saturation in elderly patients surgery: A meta-analysis. Ann Thorac Surg 2010; 89:787–93 undergoing major abdominal surgery minimizes brain expo- sure to potential hypoxia. Anesth Analg 2005; 101:740 –7 32. Gogbashian A, Sedrakyan A, Treasure T: EuroSCORE: A sys- tematic review of international performance. Eur J Cardio-17. Casati A, Fanelli G, Pietropaoli P, Proietti R, Tufano R, Mon- thorac Surg 2004; 25:695–700 tanini S, Collaborative Italian Study Group on Anaesthesia in Elderly Patients, Danelli G, Nuzzi M, Mentegazzi F, Torri G, 33. Ranucci M, Castelvecchio S, Menicanti L, Frigiola A, Pelissero Martani C, Spreafico E, Fierro G, Pugliese F, De Cosmo G, G: Accuracy, calibration and clinical performance of the Aceto P, Servillo G, Monaco F: Monitoring cerebral oxygen EuroSCORE: Can we reduce the number of variables? Eur saturation in elderly patients undergoing general abdominal J Cardiothorac Surg 2010; 37:724 –9 surgery: A prospective cohort study. Eur J Anaesthesiol 2007; 34. Shahian DM, O’Brien SM, Filardo G, Ferraris VA, Haan CK, 24:59 – 65 Rich JB, Normand SL, DeLong ER, Shewan CM, DokholyanAnesthesiology 2011; 114:58 – 69 68 Heringlake et al.
12. PERIOPERATIVE MEDICINE RS, Peterson ED, Edwards FH, Anderson RP, Society of Tho- 36. Shahian DM, O’Brien SM, Filardo G, Ferraris VA, Haan CK, racic Surgeons Quality Measurement Task Force: The Society Rich JB, Normand SL, DeLong ER, Shewan CM, Dokholyan of Thoracic Surgeons 2008 cardiac surgery risk models: Part RS, Peterson ED, Edwards FH, Anderson RP, Society of Tho- 1– coronary artery bypass grafting surgery. Ann Thorac Surg racic Surgeons Quality Measurement Task Force: The Society 2009; 88:S2–22 of Thoracic Surgeons 2008 cardiac surgery risk models: Part35. O’Brien SM, Shahian DM, Filardo G, Ferraris VA, Haan CK, 3–valve plus coronary artery bypass grafting surgery. Ann Rich JB, Normand SL, DeLong ER, Shewan CM, Dokholyan RS, Thorac Surg 2009; 88:S43– 62 Peterson ED, Edwards FH, Anderson RP, Society of Thoracic 37. Ranucci M, Castelvecchio S, Menicanti L, Frigiola A, Pelissero Surgeons Quality Measurement Task Force: The Society of Tho- G: Risk of assessing mortality risk in elective cardiac opera- racic Surgeons 2008 cardiac surgery risk models: Part 2–iso- tions: Age, creatinine, ejection fraction, and the law of par- lated valve surgery. Ann Thorac Surg 2009; 88:S23– 42 simony. Circulation 2009; 119:3053– 61 ANESTHESIOLOGY REFLECTIONSOehmig’s Monument to Ether, Public Gardens Unveiled in 1868 on the Boston Public Gardens, the granite “Monument to Ether” avoided taking sides as to which co-patentee, dentist W. T. G. Morton or physician C. T. Jackson, actually deserved credit for bringing ether anesthesia to humankind. As the Ether Controversy raged, Dr. Oliver Wendell Holmes (1805–1894) mused that the granite was a “monument to ether— or either” of the quarreling co-paten- tees. In 1992 Monument to Ether, Public Gardens was painted (left) by Keith S. Oehmig of Brunswick, Maine, and then placed in the American Society of Anesthesiologists’ newly opened WLM Gallery (center). Oehmig captured in oil that the “ordered, defined structure of the monument and footpaths contrast with the soft, amorphous roses and trees of the Garden.” Note that the nebulously painted “Good Samaritan” figure atop the monument (right) was that of a medieval Moorish physician—a figure chosen purposefully to avoid siding with any one person or profession as to “who discovered anesthesia.” (Copyright © the American Society of Anesthesiologists, Inc. This image also appears in the Anesthesi- ology Reflections online collection available at www.anesthesiology.org.) George S. Bause, M.D., M.P.H., Honorary Curator, ASA’s Wood Library-Museum of Anesthesi- ology, Park Ridge, Illinois, and Clinical Associate Professor, Case Western Reserve University, Cleveland, Ohio. UJYC@aol.com.Anesthesiology 2011; 114:58 – 69 69 Heringlake et al.