Departments of Cardiology, Clinical Governance, and Cardiac Surgery, The Cardiothoracic Centre-Liverpool, Liverpool, United Kingdom

Preoperative White Blood Cell Count is Independently Associated With Higher Perioperative Cardiac Enzyme Release and Increased 1-Year Mortality After Coronary Artery Bypass Grafting Nick Newall, MRCP, Antony D. Grayson, BS, Aung Y. Oo, FRCS, Nicholas D. Palmer, MRCP, Walid C. Dihmis, FRCS, Abbas Rashid, FRCS, and Rodney H. Stables, DM, FRCP Departments of Cardiology, Clinical Governance, and Cardiac Surgery, The Cardiothoracic Centre-Liverpool, Liverpool, United Kingdom Background. Elevated preprocedural systemic markers of inflammation, including white blood cell count, have been associated with adverse clinical outcomes after percutaneous coronary intervention. The relationship between preoperative white blood cell count and clinical outcomes after coronary artery bypass grafting is less clear despite increasing evidence that neutrophils participate in reperfusion injury. We sought to determine the relationship between preoperative white blood cell count and in hospital major morbidity and 1-year survival after coronary artery bypass grafting. Methods. We prospectively studied 3,024 consecutive isolated coronary artery bypass graft procedures. Preoperative white blood cell count was determined by automated counter, perioperative cardiac enzyme release by the creatine kinase myocardial band isoenzyme, and all-cause mortality over the first postoperative year taken from a national death registry. Multivariate logistic regression and Cox proportional hazards analyses were performed. Results. Preoperative white blood cell count offered as a continuous variable and as five predetermined groups was independently associated with cardiac enzyme release three or more times the upper limit of the reference range (adjusted odds ratio 1.5 per 10 10 9 /L increase, 95% confidence interval: 1.2 to 2.0, p 0.002) and higher 1-year mortality (adjusted hazard ratio 1.6 per 10 10 9 /L increase, 95% confidence interval: 1.2 to 2.1, p < 0.001). Conclusions. Higher preoperative white blood cell count is independently associated with higher perioperative myonecrosis and 1-year mortality after coronary artery bypass grafting. (Ann Thorac Surg 2006;81:583 90) 2006 by The Society of Thoracic Surgeons Atherosclerosis has increasingly been recognized to be an inflammatory process [1]. Supporting this concept, systemic markers of inflammation have been associated with both increased risk of developing atherosclerotic disease among apparently healthy persons [2 4] and increased frequency of vascular events among patients with established atherosclerotic disease [5]. Moreover, preprocedural markers of inflammation have been associated with periprocedural myonecrosis and reduced medium and long-term survival after percutaneous coronary intervention in acute coronary syndromes [6], acute myocardial infarction [7, 8], and elective percutaneous coronary intervention [9, 10]. Microvascular obstruction by mixed white cell and Accepted for publication Aug 22, 2005. Address correspondence to Mr Grayson, Clinical Governance Department, The Cardiothoracic Centre-Liverpool, Thomas Dr, Liverpool L14 3PE, United Kingdom; e-mail: tony.grayson@ctc.nhs.uk. platelet aggregates have been implicated in the no-reflow phenomenon associated with procedure-related myocardial infarction after percutaneous coronary intervention [11]. Although cardiac surgery provokes a systemic inflammatory response, which can be profound, the relationship between preoperative inflammatory markers and clinical outcomes after coronary artery bypass grafting surgery (CABG) is less clear [12, 13]. The white blood cell (WBC) count is the most universally available inflammatory marker in patients undergoing surgery, and numerous reports have suggested a causative relationship between the most numerous white cell subpopulation, neutrophils, and reperfusion injury [14]. Putative mechanisms for this reperfusion injury include proteolytic and oxidative injury, enhanced platelet activation, microvascular plugging, complement activation, and enhanced accumulation within infarcted myocardium [14, 15]. 2006 by The Society of Thoracic Surgeons 0003-4975/06/$32.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2005.08.051

584 NEWALL ET AL Ann Thorac Surg PREOPERATIVE WBC LINK TO CABG OUTCOME 2006;81:583 90 We hypothesized that the preoperative WBC count may be associated with increased perioperative myocardial injury, other major postoperative morbidities and increased 1-year mortality after isolated CABG. Material and Methods Patient Population and Data Data were prospectively collected on a total of 3,024 consecutive patients undergoing isolated CABG surgery between January 1, 1999, and December 31, 2001, at the Cardiothoracic Centre-Liverpool. Data collection methods and definitions have been described in detail previously [16]. The study received approval from the Royal Liverpool and Broadgreen University Hospital Ethics Committee. Preoperative data were collected prospectively during the patient s admission as part of routine clinical practice and entered into a cardiac surgical registry. Variables collected included age, sex, body mass index, urgency of operation, prior cardiac surgery, diabetes, hypercholesterolaemia, hypertension, peripheral vascular disease, respiratory disease, renal dysfunction, degree of left main stem stenosis, extent of coronary disease, and left ventricular ejection fraction. White Blood Cell Count and Cardiac Enzymes The most recent preoperative WBC count including neutrophil count, and peak cardiac enzyme (CE) release for the first 24 hours postoperative were abstracted from a routinely recorded electronic clinical biochemistry archive, blind to survival and clinical data. The WBC counts were determined by an automated counter (Beckman Coulter Gen S, Fullerton, CA) in EDTA whole blood. The cardiac enzyme assays used were creatine kinase myocardial band (CK-MB) immunoassay (Roche [reference range, 5 to 24 U/L at 37 C]) and total CK (Roche [reference range, 190 U/L in men, 167 U/L in women at 37 C]). When more than one cardiac enzyme result existed, the highest value of CK-MB isoenzyme was selected in preference to the highest value of total CK. The cardiac enzyme results were then expressed as multiples of the upper limit of the reference range. Study Endpoints Outcome data included 1-year and in-hospital mortality, perioperative CE release, stroke, reexploration for bleeding, acute renal failure, deep sternal wound infection, and intra-aortic balloon pump (IABP) support. In-hospital mortality was defined as death within the same hospital admission regardless of cause. All postoperative patients transferred from the surgical center to another hospital were followed up to confirm their status at discharge. Patient records were linked to the National Strategic Tracing Service (NSTS), which records all-cause mortality in the United Kingdom to establish current vital status at 1 year of follow-up. Matching to the NSTS was based on patient name, National Health Service unique number, date of birth, sex, and postal code. Cardiac enzyme release three or more times the upper limit of the reference range (ULR) has been shown to be of prognostic significance after CABG [17] and was taken as the threshold of clinically significant perioperative CE release. Postoperative stroke was defined as a new focal neurologic deficit or comatose states, or both, occurring postoperatively that persisted longer than 24 hours after its onset. We excluded confused states, transient events, and intellectual impairment from our study to reduce subjective bias. Postoperative bleeding was defined as bleeding that required surgical reexploration after initial departure from the operating theater. Acute renal failure was defined as patients requiring new dialysis or hemofiltration postoperatively. Criteria for defining deep sternal wound infections were in accord with the published evidence-based guidelines by the Centers for Disease Control and Prevention [18]. Statistical Analysis Preoperative WBC count was treated in the first instance as a continuous variable and then to test for trend in outcomes, divided into five strata, and treated as categorical data. Continuous variables, due to nonnormality, are shown as median with 25th and 75th percentiles. Categorical data are shown as percentages. Univariate comparisons were made with Kruskal-Wallis tests and 2 tests for trend as appropriate. Deaths occurring as a function of time over the first postoperative year were described using the product limit methodology of Kaplan and Meier [19]. Cox proportional hazards analysis was used to examine the effect of preoperative WBC count on 1-year mortality and to calculate risk-adjusted Kaplan- Meier survival curves [20]. Multivariate logistic regression was used to assess the relationship between preoperative WBC count and postoperative CE release as a dichotomous outcome of more or less than 3 times ULR, and other in-hospital morbidity [21]. All analyses were risk adjusted according to American College of Cardiology/American Heart Association practice guidelines [22] for risk adjustment of CABG outcomes. This method was preferred over identification of those variables driving the outcome of interest and using these so-called risk factors for adjustment as a consequence of the low event rate and concerns regarding model over specification. The guideline-mandated adjustment includes age, sex, previous cardiac surgery, left ventricular ejection fraction, left main stem stenosis, number of major coronary arteries with stenosis greater than 70%, priority of surgery, peripheral vascular disease, diabetes, renal dysfunction, respiratory disease, hypertension, and hypercholesterolemia. We also analyzed the additional benefit that the preoperative WBC counts might have in conjunction with the additive version of European System for Cardiac Operative Risk Evaluation (EuroSCORE) [23] to enhance preoperative risk stratification for patients. This analysis is presented as odds ratios with 95% confidence intervals (CI) and area under the receiver operating characteristic (ROC) curves. The significant WBC count subgroups could then be assigned additive death points by rounding off the odds ratio to

Ann Thorac Surg NEWALL ET AL 2006;81:583 90 PREOPERATIVE WBC LINK TO CABG OUTCOME 585 Table 1. Patient and Disease Characteristics Across WBC Strata WBC 6 (n 484) WBC 6 8 (n 1,092) WBC 8 10 (n 905) WBC 10 12 (n 280) WBC 12 (n 263) p Value Age at operation (years) 65.1 (58.6 to 70.1) 64.3 (58.9 to 70.2) 63.9 (57.9 to 69.5) 63.9 (57.4 to 70.6) 64.1 (57.1 to 70.4) 0.49 Female sex (%) 16.1 20.2 20.1 25.0 19.4 0.071 BMI 30 kg/m 2 (%) 25.0 28.3 30.7 27.9 31.2 0.069 Diabetes mellitus (%) 13.0 17.4 19.6 19.6 18.3 0.018 Hypercholesterolemia (%) 80.0 83.8 84.7 87.9 82.9 0.058 Hypertension (%) 49.8 56.2 56.1 54.6 58.9 0.052 Renal dysfunction (%) 2.5 1.8 2.4 5.4 4.6 0.004 PVD (%) 10.5 11.8 15.4 21.4 14.4 0.001 Respiratory disease (%) 23.8 30.9 33.5 36.8 36.1 0.001 Ejection fraction 30% (%) 7.2 7.6 8.8 10.7 12.9 0.002 Three-vessel disease (%) 79.3 82.4 80.1 81.4 84.0 0.38 Left main stenosis (%) 15.7 20.2 21.4 19.6 23.6 0.019 Prior cardiac surgery (%) 3.5 1.8 3.0 2.1 1.5 0.33 Emergent procedure (%) 1.0 0.4 1.8 3.2 3.4 0.001 Age is shown as median with 25 th and 75 th percentiles: categorical variables are shown as a percentage. BMI body mass index; PVD peripheral vascular disease; WBC white blood cell count. the nearest 0.5. In all cases a p value less than 0.05 was considered significant. All statistical analyses were performed using SAS for Windows Version 8.2 (SAS Institute, Cary, North Carolina). Results All 3,024 patients had a preoperative WBC count recorded. Most of the patients last available preoperative WBC counts were performed the day before surgery (median, 1 day; 25th and 75th percentiles: 1 and 6 days; maximum, 19 days before surgery). The preoperative WBC count ranged from 3.3 to 64.8 10 9 /L, with a median value of 7.8 10 9 /L (25th and 75th percentiles: 6.5 and 9.3 10 9 /L). To identify a threshold of association, and to demonstrate a graded association between WBC count and adverse outcomes across the whole range of values, the preoperative WBC counts were divided into five prespecified groups. Four hundred and eighty-four patients (16.0%) had a WBC count of less than 6 10 9 /L, 1,092 (36.1%) had a WBC count between 6 and 8 10 9 /L, 905 (29.9%) had a WBC count between 8 and 10 10 9 /L, 280 (9.3%) had a WBC count between 10 and 12 10 9 /L, and 263 (8.7%) had a WBC count greater than 12 10 9 /L. The preoperative descriptive data and in-hospital outcome data were complete. The patient and disease characteristics for each of the five WBC count strata are shown in Table 1. Higher WBC counts were more frequently observed in patients with diabetes, renal dysfunction, peripheral vascular disease, left ventricular impairment, chronic respiratory disease, significant left main stem stenosis, and emergency cases. Crude in-hospital outcomes are shown in Table 2. Cardiac enzymes were recorded in 2,860 patients (94.6%). Four hundred and ninety-eight patients (16.5%) had CE three or more times the ULR (Roche CK-MB immunoassay 5 to 24 U/L). One-year follow-up for all-cause mortality was complete. There were 122 deaths (4.0%) within the first postoperative year. With WBC count treated as a continuous variable, initial analysis revealed that higher preoperative WBC counts were independently associated with CE release three or more times the ULR (adjusted odds ratio, 1.5 per 10 10 9 /L increase; 95% CI: 1.2 to 2.0; Table 2. Crude In-Hospital Mortality and Morbidity for WBC Strata WBC 6 (n 484) WBC 6 8 (n 1,092) WBC 8 10 (n 905) WBC 10 12 (n 280) WBC 12 (n 263) p Value In-hospital mortality (%) 0.8 1.6 1.8 4.6 5.7 0.001 IABP support (%) 2.5 2.2 2.5 3.6 10.3 0.001 Reexploration for bleeding (%) 2.7 2.1 2.5 2.1 6.5 0.013 Stroke (%) 1.6 1.8 2.9 3.6 0.8 0.55 Deep SWI (%) 0.6 0.6 0.7 1.4 1.5 0.104 Acute renal failure (%) 0.4 0.5 1.2 1.1 2.7 0.001 Cardiac enzyme release 3 ULN (%) 15.5 13.7 18.0 19.3 21.3 0.002 Categorical variables are shown as a percentage. IABP intra-aortic balloon pump; SWI sternal wound infection; ULN upper limit of normal; WBC white blood cell count.

586 NEWALL ET AL Ann Thorac Surg PREOPERATIVE WBC LINK TO CABG OUTCOME 2006;81:583 90 Table 3. Adjusted a In-Hospital Mortality and Morbidity for WBC Strata WBC 6 (n 484) WBC 6 8 (n 1,092) WBC 8 10 (n 905) WBC 10 12 (n 280) WBC 12 (n 263) p Value In-hospital mortality (%) 0.9 1.9 1.8 3.3 4.2 0.001 IABP support (%) 2.8 2.3 2.5 3.2 8.7 0.001 Reexploration for bleeding (%) 2.7 2.1 2.5 2.0 6.3 0.018 Stroke (%) 1.7 2.0 2.9 3.0 0.7 0.95 Deep SWI (%) 0.8 0.7 0.6 1.2 1.4 0.25 Acute renal failure (%) 0.4 0.5 1.2 0.8 2.5 0.003 Cardiac enzyme release 3 ULN (%) 16.5 13.9 17.9 17.9 20.9 0.012 a Adjusted for age, sex, previous cardiac surgery, left ventricular ejection fraction, left main stem stenosis, number of major coronary arteries with stenosis 70%, priority of surgery, peripheral vascular disease, diabetes, renal dysfunction, respiratory disease, hypertension, and hypercholesterolemia. Categorical variables are shown as a percentage. IABP intra-aortic balloon pump; SWI sternal wound infection; ULN upper limit of normal; WBC white blood cell count. p 0.002) and increased 1-year mortality (adjusted hazard ratio, 1.6 per 10 10 9 /L increase; 95% CI: 1.2 to 2.1; p 0.001). Other components of the full blood count, including hemoglobin, hematocrit, and platelet counts, were not associated with greater CE release or increased mortality. Adjusted in-hospital mortality, CE release, and other morbidity outcomes are shown in Table 3. Increased preoperative WBC counts were associated with higher in-hospital mortality (p 0.001), IABP support (p 0.001), reexploration for bleeding (p 0.018), renal failure requiring dialysis support (p 0.003), and CE release three or more times ULR (p 0.012). Crude 1-year survival curves for the WBC count strata are shown in Figure 1. One-year survival for WBC count of less than 6 10 9 /L was 97.3%, WBC count between 6 and 8 10 9 /L was 97.1%, WBC count between 8 and 10 10 9 /L was 96.1%, WBC count between 10 and 12 10 9 /L was 93.2%, and WBC count of greater than 12 10 9 /L was 91.3% (p 0.001). Adjusted 1-year survival curves for the WBC count strata are shown in Figure 2. Adjusted 1-year survival for WBC count of less than 6 10 9 /L was 98.4%, WBC count between 6 and 8 10 9 /L was 97.8%, WBC count between 8 and 10 10 9 /L was 97.1%, WBC count between 10 and 12 10 9 /L was 96.1%, and WBC count of greater than 12 10 9 /L was 94.8% (p 0.001). We repeated our analysis with preoperative neutrophil count (median, 4.8 10 9 /L; 25th and 75th percentiles: 3.8 and 6.0 10 9 /L) as a more refined risk marker for adverse outcomes. The analysis revealed that the neutrophil count largely explained the association between increased preoperative WBC counts and adverse outcomes. Higher preoperative neutrophil counts were independently associated with CE release three or more times the Fig 1. Observed 1-year survival for white blood cell count (WBC) strata. (Solid line WBC less than 6 10 9 /L; long dashed line WBC6to8 10 9 /L; short dashed line WBC8to10 10 9 /L; single-dotted dashed line WBC 10 to 12 10 9 /L; double-dotted dashed line WBC 12 10 9 /L or greater.) Fig 2. Adjusted 1-year survival for white blood cell count (WBC) strata, adjusted for age, sex, previous cardiac surgery, left ventricular ejection fraction, left main stem stenosis, number of major coronary arteries with stenosis greater than 70%, priority of surgery, peripheral vascular disease, diabetes, renal dysfunction, respiratory disease, hypertension, and hypercholesterolemia. (Solid line WBC less than 6 10 9 /L; long dashed line WBC6to8 10 9 /L; short dashed line WBC8to10 10 9 /L; single-dotted dashed line WBC 10 to 12 10 9 /L; double-dotted dashed line WBC 12 10 9 /L or greater.)

Ann Thorac Surg NEWALL ET AL 2006;81:583 90 PREOPERATIVE WBC LINK TO CABG OUTCOME 587 Fig 3. Area under the receiver operating characteristic (ROC) curve for additive European System for Cardiac Operative Risk Evaluation (EuroSCORE) only (top) and additive EuroSCORE and preoperative white blood cell count (bottom). ULR (adjusted odds ratio, 1.6 per 10 10 9 /L increase, 95% CI: 1.1 to 2.2, p 0.005) and increased 1-year mortality (adjusted hazard ratio, 1.8 per 10 10 9 /L increase, 95% CI: 1.1 to 2.5, p 0.014). To assess how the preoperative WBC counts might be applied with existing preoperative risk stratification for in-hospital mortality, we analyzed the area under the ROC curve for additive EuroSCORE only (ROC 0.74, Fig 3) and additive EuroSCORE with the preoperative WBC count (ROC 0.8, Fig 3). Table 4 shows the associated odds ratio for in-hospital mortality by preoperative WBC counts adjusted for additive EuroSCORE. Comment We have shown that higher preoperative white blood cell count is independently associated with greater perioperative cardiac enzyme release, more frequent postoperative renal failure, and higher in-hospital allcause mortality. These associations appear consistent across the observed range of WBC counts and are mostly driven by neutrophil count. These findings are consistent with, but cannot prove, a causal relationship between preoperative leukocytosis and important postoperative morbidity and mortality. We have also demonstrated a striking and gradual independent association between preoperative WBC counts and 1-year all-cause mortality, including WBC counts within the normal range. These associations are unlikely to be due to recognized preoperative infection, and suggest that the WBC count is not only associated with early perioperative major morbidity and mortality but also with further mortality over the first postoperative year. This reduction in survival during the first postoperative year, to some degree, can be explained by the preoperative WBC counts association with increased perioperative cardiac enzyme, as we have previously shown that higher cardiac enzyme release is associated with poorer 1-year outcome for patients [24]. There was an independent association between highest ( 12 10 9 /L) total WBC count and postoperative IABP use and reexploration for bleeding, but this association was not consistent across lower total WBC counts. The association with IABP support may be due to the higher rates of myocardial infarction and reexploration and subsequent need for hemodynamic support observed in patients with the highest WBC counts. Among the preoperative descriptive data, higher WBC counts were more frequently observed in patients with diabetes, preexisting renal dysfunction, peripheral vascular disease, and left ventricular impairment perhaps reflecting higher total atherosclerotic burden. The relationship between higher WBC count and chronic respiratory disease has been observed elsewhere and that higher WBC was observed among emergency cases is unsurprising. Despite these potentially confounding relationships, the association between higher WBC counts and adverse clinical outcomes persisted after multivariate analyses. We did not observe significant associations between preoperative WBC counts and other the important postoperative morbidities including deep sternal wound infection and stroke defined as new focal neurologic deficit. That sternal wound infection was not associated with preoperative WBC counts is unsurprising, as white-cell mediated reperfusion injury has not been implicated in either complication. The absence of association between stroke and preoperative WBC counts contrasts with earlier reports from Albert and associates [25] who found an independent association between WBC higher than 9 10 9 /L and perioperative cerebrovascular accidents. Dacey and associates [26] from the Northern New England Cardiovas- Table 4. Odds Ratio for In-Hospital Mortality by Preoperative WBC Counts Adjusted for Additive EuroSCORE Adjusted a Odds Ratio (95% CI) p Value Preoperative WBC 6 Reference Preoperative WBC 6 8 1.9 (0.7 5.6) 0.22 Preoperative WBC 8 10 1.9 (0.6 5.9) 0.24 Preoperative WBC 10 12 3.7 (1.1 11.8) 0.028 Preoperative WBC 12 4.3 (1.2 4.7) 0.013 a Adjusted for the additive EuroSCORE. CI confidence intervals; EuroSCORE European System for Cardiac Operative Risk Evaluation; WBC white blood cell.

588 NEWALL ET AL Ann Thorac Surg PREOPERATIVE WBC LINK TO CABG OUTCOME 2006;81:583 90 cular Disease Study Group also found an association between preoperative WBC count and stroke in 11,270 isolated CABG. Our findings may be due to very low numbers of recognized stroke and a relatively insensitive definition for stroke used in our study. Study Limitations There are several limitations that need to be considered in this study. Because we used data acquired during routine clinical practice, we did not standardize sample time for WBC count measurements, although 71% of patients had WBC counts measured within 48 hours of surgery; and we did not compare WBC counts with other preoperative inflammatory markers such as highly sensitive C-reactive protein, interleukin 6, and tumor necrosis factor-alpha that would have permitted prognostic comparison and may have refined the associations we observed between WBC counts and adverse clinical outcome. However, the WBC count remains the simplest and most readily available inflammatory marker and is more easily applied in clinical practice. Further study to establish the additional prognostic value of other inflammatory markers compared with WBC count is required. Although, data from an angiographic cohort from our own hospital have found that highly sensitive C-reactive protein and neutrophil count are modestly correlated, but that the neutrophil component of the WBC count is a superior prognostic marker for death or nonfatal myocardial infarction among patients with angiographically determined coronary disease [27]. The most important limitation to our study, as in all observational studies, is that the associations we have shown may be confounded by as yet unrecognized confounding variables. An adequately powered randomized controlled trial of white-cell specific intervention would be required to circumvent this weakness and confirm our observed associations as causal. A number of such interventions including systemic [28], blood cardioplegia [29], and salvaged blood [30] leukodepletion have been examined using surrogate endpoints, with no consistent conclusion with respect to clinical outcome. Recently less specific inhibitors of white-cell mediated injury have been suggested including perioperative adenosine [31], or aprotinin [32] use. Further largescale clinical trials to evaluate therapeutic attenuation of white cell mediated reperfusion injury are required to establish preoperative WBC count as a modifiable risk factor for clinical outcomes. A final limitation of the study is that we do not have data on the cause of death for the 122 deceased patients. Clinical Implications Our findings support and extend previous observational studies and are consistent with other publications associating inflammatory markers with adverse outcomes in coronary artery disease [5 10]. The present study confirms and refines the previously reported independent association between higher preoperative WBC count and adverse clinical outcomes including perioperative myocardial infarction, and subsequent increased mortality in patients undergoing CABG [26]. Because of this strong independent association with adverse outcome, the preoperative WBC count represents an important potentially modifiable preoperative risk factor. Preoperative elevation of WBC count should stimulate patient scrutiny for a reversible cause, and perhaps form a target for therapeutic intervention to attenuate neutrophil mediated reperfusion injury in future randomized controlled trials. The most immediate impact of this study on clinical practice, however, is to enhance clinical risk stratification by incorporating WBC count in current preoperative risk assessment tools for patients undergoing coronary artery bypass grafting. The EuroSCORE has not only been shown to be a reliable predictor of in-hospital mortality [33, 34], but has also been proven useful for predicting postoperative morbidity [35, 36] and long-term mortality [37]. Therefore, it is essential that preoperative WBC count be taken into account to allow for meaningful risk stratification and informed patient consent. From our data, it appears sensible to add three and a half additive points to the EuroSCORE value if the preoperative WBC count is between 10 and 12, and to add four and a half points if the count is greater than 12. Further work would be needed to assess whether the weighting assigned to preoperative WBC counts is applicable to other institutions and countries. We would like to acknowledge the cooperation given to us by Brian M. Fabri, John A. C. Chalmers, Neeraj K. Mediratta, Elaine M. Griffiths, and Trevor J. Hine. We would also like to thank Janet Deane, who maintains the quality and ensures completeness of data collected in our Cardiac Surgery Registry. References 1. Ross R. Atherosclerosis an inflammatory disease. N Engl J Med 1999;340:115 26. 2. Ridker PM, Cushman M, Stampfer MJ, et al. Inflammation, aspirin and the risk of cardiovascular disease in apparently healthy men. N Engl J Med 1997;336:973 9. 3. Ridker PM, Buring SE, Shih J, et al. Prospective study of C-reactive protein and the risk of future cardiovascular events among apparently healthy women. Circulation 1998; 98:731 3. 4. Danesh J, Wheeler JG, Hirschfield GM, et al. C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease. N Engl J Med 2004;350: 1387 97. 5. Haverkate F, Thompson SG, Pyke SDM, et al. Production of C reactive protein and risk of coronary events in stable and unstable angina. Lancet 1997;349:462 6. 6. Bhatt DP, Chew DP, Lincoff AM, et al. Effect of revascularisation on mortality associated with an elevated white blood cell count in acute coronary syndromes. Am J Cardiol 2003;92:136 40. 7. Barron HV, Cannon CP, Murphy SA, et al. Association between white blood cell count, epicardial blood flow, myocardial perfusion, and clinical outcomes in the setting of acute myocardial infarction. Circulation 2000;102:2329 34. 8. Pellizzon GG, Dixon SR, Stone GW, et al. Relation of admission white blood cell count to longterm outcomes after primary coronary angioplasty for acute myocardial infarction (the Stent PAMI trial). Am J Cardiol 2003;91:729 31.

Ann Thorac Surg NEWALL ET AL 2006;81:583 90 PREOPERATIVE WBC LINK TO CABG OUTCOME 589 9. Gurm HS, Bhatt DL, Gupta R, et al. Preprocedural white cell count and death after percutaneous coronary intervention. Am Heart J 2003;146:692 8. 10. Gurm HS, Bhatt DL, Lincoff AM, et al. Impact of preprocedural white blood cell count on long term mortality afer percutaneous coronary intervention: insights from the EPIC, EPILOG, and EPISTENT trials. Heart 2003;89:1200 4. 11. Siminiak T, Flores NA, Sheridan DJ, et al. Neutrophil interactions with endothelium and platelets: possible role in the development of cardiovascular injury. Eur Heart J 1995;16: 160 70. 12. Czerny M, Baumer H, Kilo J, et al. Inflammatory response and myocardial injury following coronary artery bypass grafting with or without cardiopulmonary bypass. Eur J Cardiothoracic Surg 2000;17:737 42. 13. Zahler S, Massoudy P, Hartl H, et al. Acute inflammatory responses to post ischaemic reperfusion during cardiopulmaonary bypass. Cardiovascular research 1999;41:722 30. 14. Hansen PR. Role of neutrophils in myocardial ischaemia and reperfusion. Circulation 1995;91:1872 85. 15. Dreyer WJ, Michael LH, West MS, et al. Neutrophil accumulation in ischaemic canine myocardium: insights into time course, distribution, and mechanism of localisation during early reperfusion. Circulation 1991;84:400 11. 16. Wynne-Jones K, Jackson M, Grotte G, et al, on behalf of the North West Regional Cardiac Surgery Audit Steering Group. Limitations of the Parsonnet score for measuring risk stratified mortality in the north west of England. Heart 2000;84:71 8. 17. Steuer J, Hörte LG, Lindahl B, et al. Impact or peri-operative myocardial injury on early and long-term outcome after coronary artery bypass grafting. Eur Heart J 2002;23:1219 27. 18. Mangram AJ, Horan TC, Pearson ML, et al, for the Hospital Infection Control Practices Advisory Committee. Guideline for prevention of surgical site infection, 1999. Infect Control Hosp Epidemiol 1999;20:247 78. 19. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:547 81. 20. Cox DR. Regression models and life-tables. J R Stat Soc 1972;34:187 220. 21. Hosmer D, Lemeshow S. Applied logistic regression. New York: John Wiley & Sons, 1989. 22. Eagle KA, Guyton RA, Davidoff R, et al. ACC/AHA guidelines for coronary artery bypass graft surgery. A report of the American College of Cardiology/American Heat Association Task Force on Practice Guidelines (committee to revise the 1991 guidelines for coronary artery bypass graft surgery). Circulation 1999;100:1464 80. 23. Nashef SA, Roques F, Michel P, et al. European System for Cardiac Operative Risk Evaluation (EuroSCORE). Eur J Cardiothorac Surg 1999;16:9 13. 24. Newall N, Palmer ND, Oo AY, et al. Intermediate and high peri-operative cardiac enzyme release are independently associated with increased one-year mortality following isolated coronary artery bypass surgery. Circulation 2003; 108(Suppl 4):656. 25. Albert AA, Beller CJ, Walter JA, et al. Preoperative high leukocyte count: a novel risk factor for stroke after cardiac surgery. Ann Thorac Surg 2003;75:1550 7. 26. Dacey LJ, DeSimone J, Braxton JH, et al. Preoperative white blood cell count and mortality and morbidity after coronary artery bypass grafting. Ann Thorac Surg 2003;76:760 4. 27. Newall N, Moore R, Browning P, et al. Higher preprocedural neutrophil count is independently associated with death or myocardial infarction in medically managed patients following cardiac catheterization. Circulation 2004;110(Suppl 3): 814. 28. Ortolano GA, Aldea GS, Lilly K, et al. A review of leukofiltration in cardiac surgery: the time course of reperfusion injury may facilitate study design of anti-inflammatory effects. Perfusion 2002;17:53 62. 29. Hiramatsu Y, Koishizawa T, Matsuzaki K, et al. Leukocytedepleted blood cardioplegia reduces cardiac troponin T release in patients undergoing coronary artery bypass grafting. Jpn J Thoracic cardiovasc Surg 2000;48:625 31. 30. Gu YJ, Devries AJ, Boonstra PW, et al. Leukocyte depletion results in improved lung function and reduced inflammatory response after cardiac surgery. J Cardiovasc Surg 1996;112: 494 500. 31. Vinten-Johansen J, Zhao ZQ, Corvera JS, et al. Adenosine in myocardial protection in on-pump and off-pump cardiac surgery. Ann Thorac Surg 2003;75(Suppl):691 9. 32. Bull DA, Maurer J. Aprotinin and preservation of myocardial function after ischaemia-reperfusion injury. Ann Thorac Surg 2003;75(Suppl):735 9. 33. Roques F, Nashef SA, Michel P, et al. Does EuroSCORE work in individual European countries? Eur J Cardiothorac Surg 2000;18:27 30. 34. Nashef SA, Roques F, Hammill BG, et al. Validation of the European System for Cardiac Operative Risk Evaluation (EuroSCORE) in North American Cardiac Surgery. Eur J Cardiothorac Surg 2002;22:101 5. 35. Toumpoulis IK, Anagnostopoulos CE, Swistel DG, DeRose JJ Jr. Does EuroSCORE predict length of stay and specific postoperative complications after cardiac surgery? Eur J Cardiothorac Surg 2005;27:128 33. 36. Nilsson J, Algotsson L, Hoglund P, Luhrs C, Brandt J. EuroSCORE predicts intensive care unit stay and costs of open heart surgery. Ann Thorac Surg 2004;78:1528 34. 37. Toumpoulis IK, Anagnostopoulos CE, DeRose JJ Jr, Swistel DG. European System for Cardiac Operative Risk Evaluation predicts long-term survival in patients with coronary artery bypass grafting. Eur J Cardiothorac Surg 2004;25:51 8. INVITED COMMENTARY Assessment of preoperative risk and risk modeling is an integral part of clinical practice. The current retrospective observational study recognizes an independent association between preoperative white cell count (WCC) and postoperative mortality during a 12-month period [1]. No causal relationship can be established because it is an observational study. The WCC is readily available, but it is a relatively crude inflammatory marker, and as the authors have stated, they did not measure other more specific inflammatory markers such as C-reactive protein, interleukin-6, and tumour necrosis-factor alpha. White cell count varies and a single sample may not be representative. Subgroup analysis revealed that neutrophil count largely explained the association. The WCC can be elevated in a wide range of comorbid conditions, some of which need to be identified and treated prior to cardiopulmonary bypass (eg, bacterial infection). The key question is the threshold at which these results affect clinical practice. Dacey and colleagues [2] showed a modest increase in mortality with preoperative WCCs between 6 and 12 10 9 /L with a sharp increase greater than that level. The current article [1] suggests that any WCC greater than 10 x 10 9 /L may be clinically important. There may be insufficient time to investigate an underlying cause of the WCC elevation in patients 2006 by The Society of Thoracic Surgeons 0003-4975/06/$32.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2005.08.041