Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Peak C-Reactive Protein Level Predicts Long-Term Outcomes in Type B Acute Aortic Dissection Kenichi Sakakura, Norifumi Kubo, Junya Ako, Hiroshi Wada, Naoki Fujiwara, Hiroshi Funayama, Nahoko Ikeda, Tomohiro Nakamura, Yoshitaka Sugawara, Takanori Yasu, Masanobu Kawakami, Shin-ichi Momomura Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017 Abstract—Acute aortic dissection (AAD) is associated with an inflammatory reaction, as evidenced by elevated inflammatory markers, including C-reactive protein (CRP). The association between the peak CRP level and long-term outcomes in type B AAD has not been systematically investigated. The purpose of this study was to investigate whether the peak CRP level during admission predicts long-term outcomes in type B AAD. We conducted a clinical follow-up study of type B AAD. We divided the study population into 4 groups according to the tertiles of peak CRP levels (T1: 0.60 to 9.37 mg/dL; T2: 9.61 to 14.87 mg/dL; T3: 14.90 to 32.60 mg/dL; and unavailable peak CRP group). Multivariate Cox regression analysis was applied to investigate whether the tertiles of peak CRP predict adverse events even after adjusting for other variables. A total of 232 type B AAD patients were included in this analysis. The median follow-up period was 50 months. CRP reached its peak on day 4.5⫾1.7. Mean peak CRP values in T1, T2, and T3 were 6.4⫾2.4, 12.0⫾1.5, and 19.5⫾4.0 mg/dL, respectively. There were 65 events (39 deaths and 26 aortic events) during the follow-up. T3 and T2 (versus T1) were strong predictors of adverse events (T3: hazard ratio: 6.02 [95% CI: 2.44 to 14.87], P⫽0.0001; T2: hazard ratio: 3.25 [95% CI: 1.37 to 7.71], P⫽0.01) after controlling for all of the confounding factors. In conclusion, peak CRP is a strong predictor for adverse long-term events in patients with type B AAD. (Hypertension. 2010;55:422-429.) Key Words: C-reactive protein 䡲 peak CRP 䡲 initial CRP 䡲 type B acute aortic dissection 䡲 long-term outcomes T ype B acute aortic dissection (AAD) is often successfully managed by medical therapy during the acute phase, with a lower in-hospital mortality rate compared with type A AAD.1 However, the long-term prognosis of type B AAD is associated with both higher morbidity and mortality.2–5 In an attempt to identify high-risk patients, several predictors of longterm adverse events in type B AAD have been reported.4,6–9 These predictors include false lumen closure status and maximum aortic diameter, which can be evaluated using 2D imaging studies, such as computed tomography (CT). However, the anatomy of the dissection is often complex, making it difficult to estimate the severity of disease by 2D imaging markers alone. AAD is associated with an inflammatory reaction,10 –12 as evidenced by a significant elevation in inflammatory markers, including C-reactive protein (CRP).13–16 Although CRP levels during hospital admission show significant temporal variations, the peak CRP level has been reported to be a useful marker to estimate the whole severity of acute illness or to predict adverse events.17,18 However, the association between peak CRP and long-term outcomes in type B AAD has not been systematically investigated. Therefore, the purpose of this study was to investigate whether the peak CRP level predicts long-term outcomes in type B AAD. Methods Patients and Follow-Up We identified type B AAD patients from hospital admission records between December 1989 and December 2008. The inclusion criteria were type B AAD presenting within 14 days of symptom onset and CT with contrast confirming a dissected descending aorta containing both a true and false lumen. The exclusion criteria were an in-hospital death in the index admission and loss to follow-up. Follow-up was performed via office visit, letter, or telephone contact. The day of the index discharge was determined as the beginning of follow-up. The study primary end points were all-cause death and aortic events, such as surgery of the thoracic aorta (ascending, arch, or descending), recurrence of an aortic dissection, and an aortic rupture. Because other follow-up studies in type B AAD had used all-cause mortality as the end point, we adopted all-cause mortality as the end point.3,9,19,20 However, we also collected information about the cause of each death, which allowed us to perform a secondary analysis using cardiovascular death and aortic events as the secondary end points. Patients were followed until meeting a study primary end point (death or an aortic event) or until the study end date (April 2009). Patients who did not experi- Received September 17, 2009; first decision October 8, 2009; revision accepted November 30, 2009. From the Division of Cardiovascular Medicine, Department of Integrated Medicine I, Jichi Medical University Saitama Medical Center, Omiya, Saitama, Japan. Correspondence to Norifumi Kubo, Division of Cardiovascular Medicine, Department of Integrated Medicine I, Jichi Medical University Saitama Medical Center, Amanuma 1-847, Omiya, Saitama 330-8503, Japan. E-mail [email protected] © 2010 American Heart Association, Inc. Hypertension is available at http://hyper.ahajournals.org DOI: 10.1161/HYPERTENSIONAHA.109.143131 422 Sakakura et al ence an outcome of interest were censored at the last known date of contact. This study was performed in accordance with the Helsinki Declaration and was approved by the internal review board. All of the study patients had previously granted permission for use of their medical charts for research purposes. Definition of Peak CRP Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017 Circulating CRP was mostly measured daily or every other day until it reached its initial peak in our institution. Initial peak CRP level was defined as the peak CRP. From December 1989 to November 2003, CRP was measured by turbidimetric immunoassay (Iatron CRP-TIA, Iatron) using an autoanalyzer (JCA-RX20, JEOL Ltd). From November 2003 to December 2008, CRP was measured by latex agglutination nephelometry (NanopiaCRP, Sekisui Medical) using an autoanalyzer (JCA-BM2250, JEOL Ltd). When the admission or predischarge CRP level was higher than the rest of the CRP levels, we were unable to determine the peak CRP level. We divided the patients who had peak CRP levels into 3 groups according to the tertiles of peak CRP, with T1 being the lowest and T3 being the highest (CRP level; T1: 0.60 to 9.37 mg/dL; T2: 9.61 to 14.87 mg/dL; T3: 14.90 to 32.60 mg/dL). The patients who did not have peak CRP levels were included in an unavailable peak CRP group (UG). The initial CRP level, which was measured after admission, was defined as the initial CRP. Definition of Other Clinical Criteria Clinical criteria were defined as described here. Hypertension was defined as systolic blood pressure ⬎140 mm Hg, diastolic blood pressure ⬎90 mm Hg, or medical treatment for hypertension before admission. Hyperlipidemia was total cholesterol level ⬎220 mg/dL, low-density lipoprotein cholesterol level ⬎140 mg/dL, or treatment for hyperlipidemia before admission. Diabetes mellitus was hemoglobin A1c level ⬎6.5% or treatment for diabetes mellitus before admission. Estimated glomerular filtration rate was calculated from the serum creatinine level, age, weight, and sex, according to the Cockroft-Gault formula,21 and estimated glomerular filtration rate ⱕ60 mL/min was considered impaired renal function. CT scan with intravenous contrast was performed ⱖ1 time in all of the patients. The initial examination was used for the statistical analysis. Partial enhancement of the false lumen was defined as a nonthrombosed type, and an intramural hematoma of the aorta was included as a thrombosed type. Maximum aortic diameter of the dissected aorta was also measured in all of the patients. Statistical Analysis Data are presented as frequencies and percentages for categorical variables and mean⫾SD for continuous variables. Patient characteristics are compared between the groups divided by the tertiles of peak CRP. Parametrical data were compared using a 1-way ANOVA, whereas nonparametrical data were compared using the Kruskal-Wallis test. Categorical data were compared using the 2 test. The Kaplan–Meier curves stratified according to the tertiles of peak CRP were constructed. Multivariate Cox regression analysis was applied to investigate whether the tertiles of peak CRP predict adverse events even after adjusting for other variables using 3 models. In model 1, we selected independent variables by a statistical selection. We used significantly different characteristics among the 4 groups as the independent variables. In model 2, we selected independent variables by a clinical selection. We adopted known factors such as age, sex, maximum aortic diameter, false lumen closure status, the extent of dissection, statin use, antihypertensive medications at discharge, and impaired renal function as independent variables.3,4,6 –9,22 The difference between CRP measurements was also adopted in this model. In model 3, we selected independent variables by a combination of statistical selection and clinical selection. We adopted all of the independent variables used in model 1 and Peak CRP in Type B Aortic Dissection 423 model 2. All of the variables were simultaneously adjusted in 1 step. Hazard ratios (HRs) and the 95% CI were calculated. We constructed a conventional receiver operating characteristic curve to analyze peak CRP levels to determine the cutoff points that yielded the highest combined sensitivity and specificity with respect to distinguishing patients with adverse events from those without such events. A P⬍0.05 was considered statistically significant. All of the analyses were performed using statistical software, SPSS 13.0/Windows (SPSS Inc). Results Between December 1989 and December 2008, there were 263 patients with type B AAD admitted to Saitama Medical Center. Ten patients were excluded from the study because of the lack of critical information, such as a CT. Nine patients died during the index admission. Twelve patients were lost to follow-up. Thus, a total of 232 type B AAD patients were included in this analysis. The median follow-up for the 232 patients was 50 months. Although peak CRP levels were available in 200 patients, peak CRP levels were unavailable in 32 patients (admission CRP levels higher than the rest of the CRP levels [n⫽22]; predischarge CRP levels higher than the rest of the CRP levels [n⫽8]; serial measurements of CRP data not obtained [n⫽2]). The 200 patients who had peak CRP levels were divided into 3 groups according to the tertiles of peak CRP, with T1 being the lowest and T3 being the highest. The 32 patients who did not have peak CRP levels were included in a UG. A patient flowchart is shown in Figure S1 (available in the online Data Supplement at http://hyper.ahajournals.org). In 88% of the patients who had an available peak CRP (176 of 200), CRP reached its peak between day 3 and day 6 (day: 4.5⫾1.7). The correlation coefficient between initial CRP levels and peak CRP levels was 0.23 (Spearman rank coefficient: P⫽0.001). The clinical characteristics among all, T1, T2, T3, and UG are shown in Table 1. Mean peak CRP values in the total population, T1, T2, and T3 groups were 12.6⫾6.1, 6.4⫾2.4, 12.0⫾1.5, and 19.5⫾4.0 mg/dL, respectively. Initial CRP levels, time from onset to admission, being overweight (body mass index ⬎25 kg/m2), current smoking, impaired renal function at admission, white blood cell counts, heart rate at discharge, and diuretics as an antihypertensive medication at discharge were significantly different among groups. The Kaplan–Meier curves stratified according to the tertiles of peak CRP and unavailable peak CRP are shown in the Figure. Log-rank testing revealed a significant increase in adverse events in the T3 group and UG as compared with the T1 group (P⫽0.0001 for T3 versus T1; P⫽0.0004 for UG versus T1). There were 65 events (39 deaths and 26 aortic events) during the follow-up period (Table S1). Multivariate Cox regression analysis was performed in 3 ways. In model 1, the tertiles of peak CRP and significant confounding factors, such as tertiles of initial CRP, time from onset to admission, being overweight, current smoking, impaired renal function at admission, white blood cell counts, heart rate at discharge, and diuretics as antihypertensive medication at discharge, were adopted as independent variables (Table 2). UG, T3, and T2 (versus T1) were strong predictors of adverse events (UG: HR: 7.07 [95% CI: 2.60 to 424 Table 1. Hypertension February 2010 Patients Characteristics Compared Among Tertiles of Peak CRP Patient Characteristics Peak CRP range, mg/dL No. Median follow-up period, mo Peak CRP, n, mg/dL Frequency of CRP measurements, times All Lowest Tertile (T1) Middle Tertile (T2) Highest Tertile (T3) Unavailable Peak (UG) P 0.60 to 32.60 0.60 to 9.37 9.61 to 14.87 14.90 to 32.60 … … 232 67 67 66 32 … 50 71 50 38 50 … 6.4⫾2.4, 67 12.0⫾1.5, 67 19.5⫾4.0, 66 12.6⫾6.1, 200 12⫾7 12⫾7 13⫾8 15⫾7 … 7⫾5 … ⬍0.0001 Methods of CRP measurements Turbidimetric immunoassay, n (%) 128/232 (55.2) 40/67 (59.7) 34/67 (50.7) 34/66 (51.5) 20/32 (62.5) Latex agglutination nephelometry, n (%) 104/232 (44.8) 27/67 (40.3) 33/67 (49.3) 32/66 (48.5) 12/32 (37.5) … 2.7⫾4.8, 232 1.2⫾2.0, 67 1.6⫾3.1, 67 2.9⫾4.6, 66 8.1⫾7.6, 32 ⬍0.0001 Age, n, y 64.1⫾11.9, 232 63.4⫾11.9, 67 62.4⫾12.1, 67 65.4⫾12.0, 66 66.3⫾11.0, 32 0.31 Male sex, n (%) 165/232 (71.1) 45/67 (67.2) 48/67 (71.6) 53/66 (80.3) 19/32 (59.4) Time from onset to admission, d 1.1⫾2.3 0.6⫾1.1 0.5⫾0.9 0.5⫾0.9 4.7⫾4.3 Initial CRP, n, mg/dL Overweight (BMI ⬎25 kg/m2), n (%) 0.54 0.15 ⬍0.0001 Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017 78/230 (33.9) 13/67 (19.4) 29/67 (43.3) 26/66 (39.4) 10/30 (33.3) 0.02 167/232 (72.0) 47/67 (70.1) 45/67 (67.2) 51/66 (77.3) 24/32 (75.0) 0.58 Hyperlipidemia, n (%) 74/232 (31.9) 22/67 (32.8) 17/67 (25.4) 27/66 (40.9) 8/32 (25.0) 0.21 Diabetes mellitus, n (%) 23/232 (9.9) 7/67 (10.4) 4/67 (6.0) 7/66 (10.6) 5/32 (15.6) 0.5 Current smoking, n (%) 111/232 (47.8) 25/67 (37.3) 41/67 (61.2) 32/66 (48.5) 13/32 (40.6) 0.04 Asthma, n (%) 8/232 (3.4) 2/67 (3.0) 1/67 (1.5) 4/66 (6.1) 1/32 (3.1) 0.54 Marfan syndrome, n (%) 6/232 (2.6) 3/67 (4.5) 0/67 (0) 2/66 (3.0) 1/32 (3.1) 0.42 Atrial fibrillation, n (%) 10/227 (4.4) 2/66 (3.0) 1/66 (1.5) 5/64 (7.8) 2/31 (6.5) 0.3 Previous MI or angina pectoris, n (%) 19/232 (8.2) 5/67 (7.5) 3/67 (4.5) 8/66 (12.1) 3/32 (9.4) 0.44 0/67 (0) 0/67 (0) 3/66 (4.5) 0/32 (0) 0.05 Hypertension, n (%) Previous aortic dissection, n (%) 3/232 (1.3) Aortoiliac aneurysm, n (%) 36/232 (15.5) 8/67 (11.9) 9/67 (13.4) 15/66 (22.7) 4/32 (12.5) 0.29 Creatinine at admission, n, mg/dL 1.0⫾1.2, 232 1.1⫾1.7, 67 1.0⫾1.4, 67 1.1⫾0.7, 66 0.8⫾0.3, 32 0.11 Creatinine at discharge, n, mg/dL 1.1⫾1.4, 225 1.0⫾1.0, 64 1.1⫾1.5, 67 1.4⫾1.9, 65 0.9⫾0.3, 29 0.07 Impaired renal function at admission, n (%) 63/232 (27.2) 13/67 (19.4) 14/67 (20.9) 27/66 (40.9) 9/32 (28.1) 0.02 Impaired renal function at discharge, n (%) 73/225 (32.4) 18/64 (28.1) 17/67 (25.4) 28/65 (43.1) 10/29 (34.5) 0.14 White blood cell counts (⫻103/mm3), n 12.0⫾3.9, 232 10.5⫾3.7, 67 12.7⫾3.7, 67 13.4⫾4.0, 66 10.5⫾3.4, 32 ⬍0.0001 Surgery at initial hospitalization, n (%)* 17/232 (7.3) 3/67 (4.5) 3/67 (4.5) 3/32 (9.4) 0.26 8/66 (12.1) CT findings Nonthrombosed type, n (%) 102/232 (44.0) 28/67 (41.8) 25/67 (37.3) 36/66 (54.5) 13/32 (40.6) 0.21 Maximum aorta diameter, n, mm 39.0⫾8.0, 232 37.7⫾7.3, 67 38.2⫾6.8, 67 40.1⫾8.9, 66 40.9⫾9.3, 32 0.31 45/232 (19.4) 16/67 (23.9) 14/67 (20.9) 10/66 (15.2) 5/32 (15.6) 0.57 DeBakey IIIa (did not across diaphragm), n/N (%) Systolic BP at admission, n, mm Hg 156⫾29, 232 156⫾25, 67 159⫾29, 67 155⫾33, 66 153⫾25, 32 0.65 Diastolic BP at admission, n, mm Hg 84⫾18, 232 83⫾17, 67 88⫾17, 67 82⫾21, 66 85⫾14, 32 0.11 Heart rate at admission, n, per min 82⫾16, 232 80⫾17, 67 81⫾18, 67 83⫾14, 66 87⫾17, 32 0.26 Systolic BP at discharge, n, mm Hg 121⫾16, 232 119⫾15, 67 120⫾17, 67 122⫾19, 66 123⫾12, 32 0.53 Diastolic BP at discharge, n, mm Hg 70⫾10, 232 70⫾9, 67 71⫾11, 67 68⫾9, 66 72⫾9, 32 0.13 Heart rate at discharge, n, per min 67⫾12, 232 66⫾11, 67 63⫾10, 67 68⫾13, 66 73⫾15, 32 0.01 Calcium channel blockers, n (%) 217/232 (93.5) 62/67 (92.5) 63/67 (94.0) 63/66 (95.5) 29/32 (90.6) 0.8 -Blockers, n (%) Antihypertensive medication at discharge 188/232 (81.0) 54/67 (80.6) 57/67 (85.1) 52/66 (78.8) 25/32 (78.1) 0.77 ACE inhibitors, n/N (%) 66/232 (28.4) 23/67 (34.3) 19/67 (28.4) 15/66 (22.7) 9/32 (28.1) 0.53 ARBs, n/N (%) 60/232 (25.9) 16/67 (23.9) 22/67 (32.8) 16/66 (24.2) 6/32 (18.8) 0.43 ␣-Blockers, n/N (%) 38/232 (16.4) 8/67 (11.9) 12/67 (17.9) 16/66 (24.2) 2/32 (6.3) 0.09 Diuretics, n/N (%) 40/232 (17.2) 9/67 (13.4) 8/67 (11.9) 19/66 (28.8) 4/32 (12.5) 0.03 28/232 (12.1) 4/67 (6.0) 7/67 (10.4) 13/66 (19.7) 4/32 (12.5) 0.11 Statins at discharge, n/N (%) Data are expressed as mean⫾SD or percentage unless otherwise specified. 2 test was used for categorical variables. One-way ANOVA or Kruskal-Wallis test was used for continuous variables. BMI indicates body mass index; BP, blood pressure; ACE inhibitors, angiotensin-converting enzyme inhibitors; ARBs, angiotensin receptor blockers; MI, myocardial infarction. *No patient had fenestrating or stenting at initial hospitalization. Sakakura et al Peak CRP in Type B Aortic Dissection 425 Event-free survival rate 1.00 Lowest Tertile (T1) of Peak CRP 0.75 Middle Tertile (T2) of Peak CRP 0.50 P values Overall, <0.0001 T1 vs. T2, 0.09 T1 vs. T3, 0.0001 T1 vs. UG, 0.0004 T2 vs. T3, 0.01 T2 vs. UG, 0.06 T3 vs. UG, 0.63 0.25 Unavailable peak CRP (UG) Highest Tertile (T3) of Peak CRP Figure. Kaplan–Meier curve stratified according to the tertiles of peak CRP. The P value was calculated by the log-rank test. Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017 0.00 No. at risk 0 1000 2000 3000 Days since Follow-up Lowest Tertile (T1) 67 48 35 23 Middle Tertile (T2) 67 45 28 10 Highest Tertile (T3) 66 36 19 6 Unavailable peak (UG) 32 24 13 8 19.26], P⫽0.0001; T3: HR: 5.13 [95% CI: 2.26 to 11.62], P⬍0.0001; T2: HR: 2.59 [95% CI: 1.13 to 5.94], P⫽0.02) even after controlling for all of the confounding factors. In this model, we also divided patients into 3 groups according Table 2. Multivariate Cox Regression Analysis Predicting Death or Event: Model 1 Variables HR 95% CI P Middle tertile (T2) of peak CRP (vs lowest tertile 关T1兴) 2.59 1.13 to 5.94 0.02 Highest tertile (T3) of peak CRP (vs lowest tertile 关T1兴) 5.13 2.26 to 11.62 ⬍0.0001 Unavailable peak CRP group (UG) (vs lowest tertile 关T1兴) 7.07 2.60 to 19.26 0.0001 Middle tertile of initial CRP (vs lowest tertile of initial CRP) 1.37 0.69 to 2.73 0.37 Highest tertile of initial CRP (vs lowest tertile of initial CRP) 1.14 0.57 to 2.30 0.71 Time from onset to admission 0.88 0.76 to 1.02 0.10 Overweight 0.92 0.51 to 1.65 0.78 Current smoking 1.09 0.63 to 1.90 0.76 Impaired renal function at admission (eGFR ⬍60 mL/min) 1.22 0.67 to 2.21 0.51 White blood cell counts (⫻103/mm3) 0.95 0.88 to 1.02 0.14 Heart rate at discharge (per 1/min incremental) 1.02 1.00 to 1.04 0.10 Diuretics 1.69 0.90 to 3.17 0.10 Peak CRP Initial CRP eGFR indicates estimated glomerular filtration rate. In this model, all of the significant confounding factors among tertiles of peak CRP (P⬍0.05 in Table 1) are adopted as independent variables. All of the variables are adjusted in 1 step. to the tertiles of initial CRP levels. However, neither the highest tertile of initial CRP (versus the lowest tertile of initial CRP group: HR: 1.14 [95% CI: 0.57 to 2.30]; P⫽0.71) nor the middle tertile of initial CRP (versus the lowest tertile of initial CRP group: HR: 1.37 [95% CI: 0.69 to 2.73]; P⫽0.37) was a significant predictor of adverse events. In model 2, the tertiles of peak CRP and known variables such as age, sex, maximum aortic diameter, false lumen closure status, the extent of dissection (DeBakey IIIa or IIIb), impaired renal function, statins, calcium channel blockers, -blockers, and angiotensinconverting enzyme inhibitors at discharge were included as independent variables. The 2 different methods of CRP measurement at our institution were also included as independent variables. UG (HR: 4.28 [95% CI: 1.78 to 10.32]; P⫽0.001), T3 (HR: 3.99 [95% CI: 1.78 to 8.99]; P⫽0.0008), and T2 (HR: 2.42 [95% CI: 1.04 to 5.61]; P⫽0.04) were associated with adverse events in long-term follow-up (Table 3). In model 3, we adopted all of the independent variables used in model 1 and model 2. UG (HR: 7.45 [95% CI: 2.20 to 25.28]; P⫽0.001), T3 (HR: 6.02 [95% CI: 2.44 to 14.87]; P⫽0.0001), and T2 (HR: 3.25 [95% CI: 1.37 to 7.71]; P⫽0.01) were associated with adverse events (Table 4). We performed a secondary analysis using cardiovascular deaths and aortic events as a secondary end point. In 65 total events, there were 11 deaths that were less associated with aortic dissection (cancer [n⫽4], pneumonia [n⫽3], emphysema [n⫽1], renal failure [n⫽1], nephritic syndrome [n⫽1], and hemorrhagic shock [n⫽1]). After excluding these deaths, multivariate Cox regression analysis revealed that the tertiles of peak CRP were still significantly associated with adverse events even after controlling for all of the variables used in model 3 (T2 versus T1: HR: 2.64 [95% CI: 1.00 to 6.98], P⫽0.05; T3 versus T1: HR: 426 Hypertension February 2010 Table 3. Multivariate Cox Regression Analysis Predicting Death or Event: Model 2 Variables HR 95% CI Middle tertile (T2) of peak CRP (vs lowest tertile 关T1兴) 2.42 Highest tertile (T3) of peak CRP (vs lowest tertile 关T1兴) Table 4. Multivariate Cox Regression Analysis Predicting Death or Event: Model 3 P Variables HR 95% CI 1.04 to 5.61 0.04 Peak CRP 3.99 1.78 to 8.99 0.0008 Unavailable peak CRP group (UG) (vs lowest tertile 关T1兴) 4.28 1.78 to 10.32 Age (per 10-y incremental) 1.09 Male sex Maximum aorta diameter (per 1-mm incremental) P Middle tertile (T2) of peak CRP (vs lowest tertile 关T1兴) 3.25 1.37 to 7.71 0.01 Highest tertile (T3) of peak CRP (vs lowest tertile 关T1兴) 6.02 2.44 to 14.87 0.0001 0.001 2.20 to 25.28 0.001 0.50 Unavailable peak CRP group (UG) (vs lowest tertile 关T1兴) 7.45 0.85 to 1.39 1.78 0.87 to 3.61 0.11 1.01 0.98 to 1.05 0.41 Middle tertile of initial CRP (vs lowest tertile of initial CRP) 1.48 0.71 to 3.09 0.29 Highest tertile of initial CRP (vs lowest tertile of initial CRP) 0.84 0.38 to 1.87 0.67 Time from onset to admission 0.91 0.78 to 1.07 0.25 Overweight 0.80 0.41 to 1.53 0.49 Initial CRP Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017 Nonthrombosed type 1.24 0.69 to 2.22 0.47 DeBakey IIIa (vs IIIb) 0.90 0.40 to 2.04 0.80 Impaired renal function at discharge (eGFR ⬍60 mL/min) 2.27 1.26 to 4.11 0.01 Turbidimetric immunoassay (vs latex agglutination nephelometry) 0.36 0.15 to 0.84 0.02 Current smoking 0.77 0.41 to 1.44 0.41 0.18 to 1.07 0.07 1.11 0.44 to 2.82 0.83 Impaired renal function at admission (eGFR ⬍60 mL/min) 0.44 Statins at discharge CCB at discharge 0.46 0.18 to 1.17 0.10 White blood cell counts (⫻103/mm3) 0.95 0.87 to 1.03 0.20 Heart rate at discharge (per 1/min incremental) 1.02 0.99 to 1.04 0.25 Diuretics 1.14 0.52 to 2.51 0.75 -Blockers at discharge 1.01 0.52 to 1.96 0.98 ACE inhibitors at discharge 0.50 0.25 to 0.97 0.04 ACE indicates angiotensin-converting enzyme; CCB, calcium channel blocker; eGFR, estimated glomerular filtration rate. In this model, age, sex, the difference of CRP measurements, and previously reported risk factors, such as maximum aorta diameter, nonthrombosed type, the extent of dissection, impaired renal function, statins, and antihypertensive medications are adopted as independent variables. All of the variables are adjusted in 1 step. 5.28 [95% CI: 1.94 to 14.41], P⫽0.001; UG versus T1: HR: 6.03 [95% CI: 1.66 to 21.90], P⫽0.006). We constructed receiver operating characteristic curves and calculated statistics for the area under the curve. The area under the curve was 0.60 (95% CI: 0.51 to 0.69; P⫽0.03). A peak CRP value of 9.5 mg/dL, which is the cutoff value between T1 and T2, has a 76% sensitivity and a 37% specificity for the occurrence of adverse events. A peak CRP value of 14.9 mg/dL, which is the cutoff value between T2 and T3, has a 46% sensitivity and a 71% specificity for the occurrence of adverse events. Because the timeline of the study is fairly long, the time effect might affect patient outcomes. To investigate the time effect, we divided the 232-patient study group in half and compared outcomes between the former 116 patients (from December 1989 to November 2002) and the latter 116 patients (from November 2002 to December 2008) by logrank test. Event-free survival at 3 years was not different (91% in former patients versus 88% in latter patients; P⫽0.12). The prescription rate of calcium channel blockers, -blockers, ␣-blockers, and diuretics at discharge was not significantly different between the former and the latter patients. Although the prescription rate of angiotensinconverting enzyme inhibitors was higher in the former patients (39.7% versus 17.2%; P⫽0.0002), that of angiotensin receptor blocker was higher in the latter patients (11.2% versus 40.5%; P⬍0.0001). Age (per 10-y incremental) 1.08 0.80 to 1.46 0.62 Male sex 2.06 0.96 to 4.43 0.06 Maximum aorta diameter (per 1-mm incremental) 1.03 1.00 to 1.07 0.07 Nonthrombosed type 1.31 0.68 to 2.55 0.42 DeBakey IIIa (vs IIIb) 0.92 0.39 to 2.18 0.85 Impaired renal function at discharge (eGFR ⬍60 mL/min) 3.98 1.86 to 8.50 0.0004 Turbidimetric immunoassay (vs latex agglutination nephelometry) 0.36 0.15 to 0.87 0.02 Statins at discharge 0.99 0.35 to 2.83 0.99 CCB at discharge 0.68 0.21 to 2.18 0.51 -Blockers at discharge 1.18 0.57 to 2.44 0.65 ACE inhibitors at discharge 0.45 0.21 to 0.94 0.03 ACE indicates angiotensin-converting enzyme; CCB, calcium channel blocker; eGFR, estimated glomerular filtration rate. In this model, all of the variables used in model 1 and model 2 are adopted as independent variables. All of the variables are adjusted in 1 step. Discussion We investigated whether the peak CRP level has any prognostic value in predicting long-term outcomes in 232 type B AAD patients. Although initial CRP levels were not associated with adverse events, peak CPR levels were significantly associated with adverse events. Peak CRP levels were a better marker than initial CRP levels in the risk stratification of type B AAD patients. Because it takes several days to reach a peak CRP, initial CRP levels might not reflect the whole severity of aortic dissection. Our results showed that the peak CRP level was a strong predictor of long-term outcomes in type B AAD. Although type A AAD is a life-threatening disease and needs emergent surgery, type B AAD is considered to be a relatively benign condition, with few cases requiring emer- Sakakura et al Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017 gent surgery during hospital admission. However, long-term outcomes of type B AAD are not necessarily better than that of type A AAD.2,3 Type B AAD is associated with a high mortality rate (⬇20% at 3 years3) and a high morbidity rate (⬇30% morbidity at 2 years) in the long term.4,5 Therefore, identifying high-risk patients is of special clinical importance. Earlier studies have identified long-term predictors of morbidity and mortality in type B AAD, including female sex, a history of previous aortic aneurysm, a history of atherosclerosis, and impaired renal function.3,19,20,22 In particular, multiple groups have reported that the maximum aortic diameter (ⱖ40 mm) and false lumen closure status are long-term predictors in type B AAD.4,6 –9 Some of these predictors, including sex, history of previous aneurysm, and history of atherosclerosis, may be more representative of a patient’s high-risk clinical background rather than the nature of the aortic dissection itself. On the other hand, false lumen closure status and the maximum aortic diameter may represent the nature of the aortic dissection for each individual patient. These parameters are relatively simply derived from 2D CT. False lumen closure status is generally classified as 1 of 2 patterns (thrombosed and nonthrombosed) or 3 patterns (thrombosed, nonthrombosed, and partially thrombosed).4,6,8,9 Maximum aortic diameter is calculated from 1 cross-sectional image.4,6,8 However, the anatomy of the aortic dissection is often complex and is not necessarily consistent, even during the initial hospital stay.23–25 It may be difficult to estimate the whole nature of the AAD by these simple imaging parameters alone. The possible explanation for why an elevated peak CRP level was associated with long-term adverse events may be several fold. In general, the peak CRP level is an excellent marker for the severity of a variety of acute illnesses. For instance, an elevated peak CRP level during hospitalization for an acute myocardial infarction has been reported to be an independent predictor of cardiac rupture, left ventricular aneurysmal formation, 1-year cardiac death, and left ventricular remodeling.17,18,26,27 Peak CRPs in acute myocardial infarction might reflect the myocardial damage and severity of myocardial infarction itself in the acute phase. In our study, peak CRP levels are presumably reflective of the severity of AAD in the acute phase. Our results also suggest the possibility that the severity expressed by the peak CRP has a lasting impact on long-term clinical events. Another possible explanation is that the peak CRP may represent the extent of the inflammatory reaction in the dissected aortic wall and may also reflect the damage to the lesion. Recent studies have revealed the close relationship between local inflammation in the aortic wall and the aortic dissection.10,12,28 Kuehl et al10 have demonstrated the association of inflammation with aortic dissection using positron emission tomography. He et al28 have immunopathologically demonstrated that T lymphocytes and macrophages are common features in medial degeneration in the aortic dissection. These studies suggest that inflammation is present in the dissected aortic wall. The severely damaged aortic wall may more easily expand, may be more prone to redissection, and may be at higher risk of rupture during the chronic phase versus a less severely damaged aortic wall. Peak CRP in Type B Aortic Dissection 427 The major part of CRP is synthesized by hepatocytes, driven by interleukin 6 with synergistic enhancement of interleukin 1 or tumor necrosis factor.29,30 CRP can also be produced locally in atherosclerotic lesions.31 Although CRP is generally considered to be an inflammatory marker or an atherosclerotic marker,32–34 CRP also plays a role as a mediator of atherosclerosis. CRP directly influences several phases of atherosclerosis via complement activation, apoptosis, vascular cell activation, monocyte recruitment, lipid accumulation, and thrombosis.31,35 Therefore, exaggerated CRP itself might exert harmful effects that promote the atherosclerosis of the dissected aortic wall. Of the 232 study patients, there were 32 patients whose peak CRP was unavailable. Interestingly, the outcomes of these patients were comparable to the highest tertile of the peak CRP group. One possible explanation is the difference in “time from onset to admission.” The mean time from onset to admission in these 32 patients was 4.7⫾4.3 days, which was longer than the other 3 groups. Some of these patients might not have received adequate medical therapy, such as aggressive blood pressure lowering in the days after the onset of dissection. Another possible explanation is a selection bias before admission to our medical center. Because our medical center is a tertiary referral hospital, patients who had developed complications over the course of several days may have subsequently been referred to our hospital. Peak CRP as a long-term predictor has several advantages over other predictors. First, the predictive power is potent with a high HR (3.99 to 6.02 adjusted for all of the confounders) in the highest tertile of peak CRP levels. Even the middle tertile of peak CRP has a moderate HR (2.42 to 3.25). In addition, CRP is a relatively simple marker widely available for clinical use. Study Limitations Although our study population is fairly large and the follow-up period is relatively long, this single-center retrospective study design poses a risk for patient selection bias. Because the timeline of the study period is fairly long, the change in the standard of care might have been a confounder of this study. Because CRP is a nonspecific inflammatory marker, it reflects not only the aortic dissection itself but also secondary processes, such as pneumonia. In addition, initial management, such as blood pressure control or pain control, might also affect the CRP. Although the CRP level was measured daily or every other day in most patients, it was measured less frequently in some patients. The methods to measure CRP have also changed from turbidimetric immunoassay to latex agglutination nephelometry during the study period. Although these 2 methods have been reported to be highly associated (r⫽0.9916; mean difference 0.19 mg/L; limits of agreement: ⫺0.36 to 0.74 mg/L),36 latex agglutination nephelometry was a more accurate method. However, the accuracy of the latex agglutination nephelometry is more important in detecting chronic inflammation in the general population and does not have as much significance during acute illness, whereas CRP levels are naturally much higher. Therefore, we do not think that changing the 428 Hypertension February 2010 Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017 method of CRP measurement would have a significant effect on our results. Because the time from onset to admission was different in each patient, the time from onset to initial CRP measurement after admission was also different in each patient. This time difference might limit the value of initial CRP. Our main diagnostic tool for the AAD has been CT with contrast. Therefore, it is often difficult to differentiate intramural hematoma from thrombosed type AAD. In addition, we could not evaluate the presence of rapid aortic expansion as a study end point, because each interval between follow-up CT analyses was not consistent. However, our institution recommended elective surgery to patients who had rapid aortic expansion. Therefore, it is possible that we were able to capture rapid aortic expansion in a portion of patients. Finally, because the surgeons had access to laboratory values reflecting an increased inflammatory status during the index admission, there is a possibility that such information affected their decision to operate during long-term follow-up. 7. 8. 9. 10. 11. Perspectives In our retrospective study, the peak CRP is a strong predictor for adverse long-term events in patients with type B AAD, but this should be validated by a prospective study. Peak CRP would be a better prognostic marker than an initial CRP. Obtaining an initial CRP level at 1 time point might be insufficient to identify high-risk patients. Our results may support the effort to find a peak CRP level for the risk stratification of type B AAD. Careful clinical follow-up may be warranted in those patients who have a high peak CRP during their index admission. 12. 13. 14. 15. 16. Disclosures None. 17. References 1. Hagan PG, Nienaber CA, Isselbacher EM, Bruckman D, Karavite DJ, Russman PL, Evangelista A, Fattori R, Suzuki T, Oh JK, Moore AG, Malouf JF, Pape LA, Gaca C, Sechtem U, Lenferink S, Deutsch HJ, Diedrichs H, Marcos y Robles J, Llovet A, Gilon D, Das SK, Armstrong WF, Deeb GM, Eagle KA. The International Registry of Acute Aortic Dissection (IRAD): new insights into an old disease. JAMA. 2000;283: 897–903. 2. Tsai TT, Evangelista A, Nienaber CA, Trimarchi S, Sechtem U, Fattori R, Myrmel T, Pape L, Cooper JV, Smith DE, Fang J, Isselbacher E, Eagle KA. Long-term survival in patients presenting with type A acute aortic dissection: insights from the International Registry of Acute Aortic Dissection (IRAD). Circulation. 2006;114:I350 –356. 3. Tsai TT, Fattori R, Trimarchi S, Isselbacher E, Myrmel T, Evangelista A, Hutchison S, Sechtem U, Cooper JV, Smith DE, Pape L, Froehlich J, Raghupathy A, Januzzi JL, Eagle KA, Nienaber CA. Long-term survival in patients presenting with type B acute aortic dissection: insights from the International Registry of Acute Aortic Dissection. Circulation. 2006; 114:2226 –2231. 4. Akutsu K, Nejima J, Kiuchi K, Sasaki K, Ochi M, Tanaka K, Takano T. Effects of the patent false lumen on the long-term outcome of type B acute aortic dissection. Eur J Cardiothorac Surg. 2004;26:359 –366. 5. Kodama K, Nishigami K, Sakamoto T, Sawamura T, Hirayama T, Misumi H, Nakao K. Tight heart rate control reduces secondary adverse events in patients with type B acute aortic dissection. Circulation. 2008; 118:S167–170. 6. Takahashi J, Wakamatsu Y, Okude J, Kanaoka T, Sanefuji Y, Gohda T, Sasaki S, Matsui Y. Maximum aortic diameter as a simple predictor of 18. 19. 20. 21. 22. 23. 24. acute type B aortic dissection. Ann Thorac Cardiovasc Surg. 2008;14: 303–310. Hata M, Sezai A, Niino T, Yoda M, Wakui S, Unosawa S, Umeda T, Shimura K, Osaka S, Furukawa N, Kimura H, Minami K. Prognosis for patients with type B acute aortic dissection: risk analysis of early death and requirement for elective surgery. Circ J. 2007;71:1279 –1282. Onitsuka S, Akashi H, Tayama K, Okazaki T, Ishihara K, Hiromatsu S, Aoyagi S. Long-term outcome and prognostic predictors of medically treated acute type B aortic dissections. Ann Thorac Surg. 2004;78: 1268 –1273. Tsai TT, Evangelista A, Nienaber CA, Myrmel T, Meinhardt G, Cooper JV, Smith DE, Suzuki T, Fattori R, Llovet A, Froehlich J, Hutchison S, Distante A, Sundt T, Beckman J, Januzzi JL Jr, Isselbacher EM, Eagle KA. Partial thrombosis of the false lumen in patients with acute type B aortic dissection. N Engl J Med. 2007;357:349 –359. Kuehl H, Eggebrecht H, Boes T, Antoch G, Rosenbaum S, Ladd S, Bockisch A, Barkhausen J, Erbel R. Detection of inflammation in patients with acute aortic syndrome: comparison of FDG-PET/CT imaging and serological markers of inflammation. Heart. 2008;94: 1472–1477. Suzuki T, Mehta RH, Ince H, Nagai R, Sakomura Y, Weber F, Sumiyoshi T, Bossone E, Trimarchi S, Cooper JV, Smith DE, Isselbacher EM, Eagle KA, Nienaber CA. Clinical profiles and outcomes of acute type B aortic dissection in the current era: lessons from the International Registry of Aortic Dissection (IRAD). Circulation. 2003;108:II312–II317. Luo F, Zhou XL, Li JJ, Hui RT. Inflammatory response is associated with aortic dissection. Ageing Res Rev. 2009;8:31–35. Schillinger M, Domanovits H, Bayegan K, Holzenbein T, Grabenwoger M, Thoenissen J, Roggla M, Mullner M. C-reactive protein and mortality in patients with acute aortic disease. Intensive Care Med. 2002;28: 740 –745. Sugano Y, Anzai T, Yoshikawa T, Satoh T, Iwanaga S, Hayashi T, Maekawa Y, Shimizu H, Yozu R, Ogawa S. Serum C-reactive protein elevation predicts poor clinical outcome in patients with distal type acute aortic dissection: association with the occurrence of oxygenation impairment. Int J Cardiol. 2005;102:39 – 45. Komukai K, Shibata T, Mochizuki S. C-reactive protein is related to impaired oxygenation in patients with acute aortic dissection. Int Heart J. 2005;46:795–799. Makita S, Ohira A, Tachieda R, Itoh S, Moriai Y, Yoshioka K, Niinuma H, Nakamura M, Hiramori K. Behavior of C-reactive protein levels in medically treated aortic dissection and intramural hematoma. Am J Cardiol. 2000;86:242–244. Anzai T, Yoshikawa T, Shiraki H, Asakura Y, Akaishi M, Mitamura H, Ogawa S. C-reactive protein as a predictor of infarct expansion and cardiac rupture after a first Q-wave acute myocardial infarction. Circulation. 1997; 96:778–784. Takahashi T, Anzai T, Yoshikawa T, Maekawa Y, Asakura Y, Satoh T, Mitamura H, Ogawa S. Serum C-reactive protein elevation in left ventricular remodeling after acute myocardial infarction–role of neurohormones and cytokine. Int J Cardiol. 2003;88:257–265. Estrera AL, Miller CC, III, Safi HJ, Goodrick JS, Keyhani A, Porat EE, Achouh PE, Meada R, Azizzadeh A, Dhareshwar J, Allaham A. Outcomes of medical management of acute type B aortic dissection. Circulation. 2006;114:I384 –I389. Tsai TT, Isselbacher EM, Trimarchi S, Bossone E, Pape L, Januzzi JL, Evangelista A, Oh JK, Llovet A, Beckman J, Cooper JV, Smith DE, Froehlich JB, Fattori R, Eagle KA, Nienaber CA. Acute type B aortic dissection: does aortic arch involvement affect management and outcomes? Insights from the International Registry of Acute Aortic Dissection (IRAD). Circulation. 2007;116:I150 –156. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron. 1976;16:31– 41. Sakakura K, Kubo N, Ako J, Fujiwara N, Funayama H, Ikeda N, Nakamura T, Sugawara Y, Yasu T, Kawakami M, Momomura S. Determinants of long-term mortality in patients with type B acute aortic dissection. Am J Hypertens. 2009;22:371–377. Liu Q, Lu JP, Wang F, Wang L, Tian JM. Three-dimensional contrastenhanced MR angiography of aortic dissection: a pictorial essay. Radiographics. 2007;27:1311–1321. Willoteaux S, Lions C, Gaxotte V, Negaiwi Z, Beregi JP. Imaging of aortic dissection by helical computed tomography (CT). Eur Radiol. 2004;14:1999 –2008. Sakakura et al 25. Akutsu K, Yokoyama S, Hata N, Shinada T, Mizuno K. Immediate regression of thrombosed false lumen in ascending aorta of retrograde type A aortic dissection. Ann Thorac Surg. 2009;87:e49. 26. Ueda S, Ikeda U, Yamamoto K, Takahashi M, Nishinaga M, Nago N, Shimada K. C-reactive protein as a predictor of cardiac rupture after acute myocardial infarction. Am Heart J. 1996;131:857– 860. 27. Orn S, Manhenke C, Ueland T, Damas JK, Mollnes TE, Edvardsen T, Aukrust P, Dickstein K. C-reactive protein, infarct size, microvascular obstruction, and left-ventricular remodelling following acute myocardial infarction. Eur Heart J. 2009;30:1180 –1186. 28. He R, Guo DC, Estrera AL, Safi HJ, Huynh TT, Yin Z, Cao SN, Lin J, Kurian T, Buja LM, Geng YJ, Milewicz DM. Characterization of the inflammatory and apoptotic cells in the aortas of patients with ascending thoracic aortic aneurysms and dissections. J Thorac Cardiovasc Surg. 2006;131:671– 678. 29. Li SP, Goldman ND. Regulation of human C-reactive protein gene expression by two synergistic IL-6 responsive elements. Biochemistry. 1996;35:9060 –9068. 30. Volanakis JE. Human C-reactive protein: expression, structure, and function. Mol Immunol. 2001;38:189 –197. Peak CRP in Type B Aortic Dissection 429 31. Paffen E, DeMaat MP. C-reactive protein in atherosclerosis: a causal factor? Cardiovasc Res. 2006;71:30 –39. 32. Schiele F, Meneveau N, Seronde MF, Chopard R, Descotes-Genon V, Dutheil J, Bassand JP. C-reactive protein improves risk prediction in patients with acute coronary syndromes. Eur Heart J. In press. 33. Woodward M, Welsh P, Rumley A, Tunstall-Pedoe H, Lowe GD. Do inflammatory biomarkers add to the discrimination of cardiovascular disease after allowing for social deprivation? Results from a 10 year cohort study in Glasgow, Scotland. Eur Heart J. In press. 34. Clarke R, Emberson JR, Breeze E, Casas JP, Parish S, Hingorani AD, Fletcher A, Collins R, Smeeth L. Biomarkers of inflammation predict both vascular and non-vascular mortality in older men. Eur Heart J. 2008;29:800 – 809. 35. Bisoendial RJ, Kastelein JJ, Stroes ES. C-reactive protein and atherogenesis: from fatty streak to clinical event. Atherosclerosis. 2007;195: e10 – e18. 36. Lolekha PH, Chittamma A, Roberts WL, Sritara P, Cheepudomwit S, Suriyawongpaisal P. Comparative study of two automated highsensitivity C-reactive protein methods in a large population. Clin Biochem. 2005;38:31–35. Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017 Peak C-Reactive Protein Level Predicts Long-Term Outcomes in Type B Acute Aortic Dissection Kenichi Sakakura, Norifumi Kubo, Junya Ako, Hiroshi Wada, Naoki Fujiwara, Hiroshi Funayama, Nahoko Ikeda, Tomohiro Nakamura, Yoshitaka Sugawara, Takanori Yasu, Masanobu Kawakami and Shin-ichi Momomura Downloaded from http://hyper.ahajournals.org/ by guest on June 18, 2017 Hypertension. 2010;55:422-429; originally published online December 28, 2009; doi: 10.1161/HYPERTENSIONAHA.109.143131 Hypertension is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright © 2009 American Heart Association, Inc. All rights reserved. Print ISSN: 0194-911X. Online ISSN: 1524-4563 The online version of this article, along with updated information and services, is located on the World Wide Web at: http://hyper.ahajournals.org/content/55/2/422 Data Supplement (unedited) at: http://hyper.ahajournals.org/content/suppl/2010/01/04/HYPERTENSIONAHA.109.143131.DC1 Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Hypertension can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: http://www.lww.com/reprints Subscriptions: Information about subscribing to Hypertension is online at: http://hyper.ahajournals.org//subscriptions/ Online supplement Peak C-Reactive Protein Level Predicts Long-term Outcomes in Type B Acute Aortic Dissection Kenichi Sakakura, Norifumi Kubo, Junya Ako, Hiroshi Wada, Naoki Fujiwara, Hiroshi Funayama, Nahoko Ikeda, Tomohiro Nakamura, Yoshitaka Sugawara, Takanori Yasu, Masanobu Kawakami, and Shin-ichi Momomura. Division of Cardiovascular Medicine, Department of Integrated Medicine I, Jichi Medical University Saitama Medical Center. Running title; Peak CRP in type B aortic dissection Address correspondence to Norifumi Kubo, M.D. Division of Cardiovascular Medicine, Department of Integrated Medicine I, Jichi Medical University Saitama Medical Center, Amanuma 1-847,Omiya, Saitama, 330-8503 Tel: +81-48-647-2111, FAX: +81-48-648-5188 E-mail: [email protected] Figure S1 Flow chart of study populations. 263 patients with type B AAD was admitted 10 patients excluded due to lack of critical data such as CT. 253 patients was included in the study cohort 9 patients died during index admission 244 patients discharged from index admission 12 patients were lost to follow-up 232 patients was final study cohort Peak CRP was available? Yes No 200 patients were divided according to tertile of peak CRP Lowest teritle 0.60-9.37 mg/dl 67 patients as T1 Middle tertile 9.61-14.87 mg/dl 67 patients as T2 Highest tertile 14.90-32.60 mg/dl Unavailable peak 66 patients as T3 32 patients as unavailable group (UG) Table S1. Each event among tertiles of peak CRP Lowest tertile of peak CRP (T1) Total event 12 ( death 9 and event 3) Middle tertile of peak CRP (T2) N Deaths Total event 15 (death 7 and event 8) Highest tertile of peak CRP (T3) N Deaths Total event 23 ( death 13 and event 10) Unavailable peak CRP gourp (UG) N Deaths Total events 15 (death 10 and event 5) N Deaths Sudden death 4 Sudden death 2 Sudden death 2 Sudden death 1 Complication of aortic dissection 2 Renal failure 1 Complication of aortic dissection 3 Rupture of aortic aneurysm 2 Recurrence of aortic dissection 1 Cerebral hemorrhage 1 Rupture of abdominal aneurysm 1 Recurrence of aortic dissection 1 Pneumonia 1 Cancer 3 AMI 1 AMI 2 Cancer 1 Cerebral infarction 1 Heart failure 2 Nephrotic syndrome 1 Cerebral hemorrhage 1 Pneumonia 2 Hemorrhagic Shock 1 Emphysema 1 Unknown cause 1 Events Replacement of descending thoracic aorta Events 3 Events Events Recurrence of aortic dissection 5 Replacement of total arch 2 Stent graft to thoracic aneurysm 1 Replacement of descending thoracic aorta 1 Replacement of ascending and arch 2 Replacement of descending thoracic aorta 2 Replacement of total arch 1 Replacement of distal arch 1 Surgery of aortic dissection 1 Surgery of modified Bentall procedure 1 Replacement of arch 1 Recurrence of aortic dissection 1 Replacement of thoraco abdominal aorta 2 Rupture of thoracic aneurysm 1 Recurrence of aortic dissection 1