414 Journal of Atherosclerosis and Thrombosis　Vol.21, No.5

Original Article

Vascular Complications and Coagulation-Related Changes in the Perioperative Period in Japanese Patients Undergoing Non-Cardiac Surgery

Chikako Aoyama-Mani1, 2, 3, Shoji Kawachi1, Yoshihiro Ogawa3 and Norihiro Kato2

1Department of Anesthesiology, National Center for Global Health and Medicine, Tokyo, Japan2 Department of Gene Diagnostics and Therapeutics, Research Institute, National Center for Global Health and Medicine, Tokyo, Japan

3 Department of Molecular Endocrinology and Metabolism, Graduate School of Medical and Dental Sciences, Tokyo Medical and Dental University, Tokyo, Japan

Aim: To properly assess the guidance for perioperative management, we undertook a clinical epide-miology study with the primary aim of evaluating the incidence of perioperative vascular complica-tions and their associated factors in a cohort of Japanese patients who underwent non-cardiac surgery in a tertiary medical care center.Methods: This observational study comprised two parts. In the first part, thrombotic and bleeding events and their risk factors in the perioperative period were evaluated in a total of 2,654 consecutive patients. In the second part, perioperative changes in coagulation-related factors, including the thrombin-antithrombin complex (TAT) and platelet aggregation activity, were serially characterized in 82 individuals randomly chosen from the consecutive patients. Results: The incidence of perioperative vascular complications was as follows: 1.0% for major bleed-ing, 0.21% for stroke and 0.21% for venous thromboembolism. No episodes of symptomatic myo-cardial infarction were identified in the studied population. Perioperative changes in coagulation-related factors were found to be complex and correlated in the mixed direction of pro- and anticoagu-lation. The TAT values showed prolonged (across postoperative days 1-5) and prominent (＞116% increase) perioperative activation of coagulation, whereas global coagulation parameters, such as the prothrombin time, showed a tendency of anticoagulation in the immediate postoperative period. Conclusions: Our data confirm the relatively low incidence of perioperative vascular complications in the general Japanese non-cardiac surgical population. Given the delicate balance between thrombotic and bleeding events, it is important to comprehensively understand the associations between the patient’s baseline risk factors and vascular complications for effective clinical management.

J Atheroscler Thromb, 2014; 21:414-434.

Key words: Perioperative period, Thrombosis, Bleeding, Hematologic change, Japanese

Introduction

Patients undergoing highly invasive surgery may

Address for correspondence: Norihiro Kato, Department of Gene Diagnostics and Therapeutics, Research Institute, National Center for Global Health and Medicine, 1-21-1 Toyama, Shinjuku-ku, Tokyo 162-8655, JapanE-mail: nokato@ri.ncgm.go.jpReceived: August 20, 2013Accepted for publication: November 1, 2013

develop thrombotic complications, such as myocardial infarction (MI), stroke, deep vein thrombosis (DVT) and pulmonary embolism (PE), in the perioperative period. Such thrombotic complications are infrequent, although they can result in significant morbidity and mortality following non-cardiac surgery, especially in patients with an advanced age and cardiovascular risk factors1). The incidence of complications depends on the type of surgery, comorbidities and outcomes. Major perioperative bleeding resides on the opposite

415Perioperative Vascular Complications

ans are consistent with those observed in Japanese remains sufficiently unanswered.

Aim

We undertook this study with the primary aim of evaluating the incidence of perioperative throm-botic and hemorrhagic complications and their associ-ated factors in a cohort of patients who underwent non-cardiac surgery in a tertiary medical care center. Additionally, we examined perioperative changes in coagulation-related factors systematically in order to integrate clinical and laboratory evidence of postoper-ative hypercoagulability.

Methods

Study PopulationThis observational study comprised two parts, as

depicted in Fig.1. In the first part, we assessed, by reviewing individual clinical charts, thrombotic and hemorrhagic events and their risk factors during the perioperative period in patients who underwent non-cardiac surgery. In the second part, we characterized perioperative coagulation changes in a total of 82 patients (80 patients for a procoagulation and throm-bosis factor analysis and 43 patients for a platelet function analysis) who were randomly chosen from the consecutive patients.

In the first part of the study, patients ≥ 20 years of age who underwent surgery under general anesthe-sia at the National Center for Global Health and Medicine (NCGM) Hospital were enrolled consecu-tively in two separate periods: from November 2009 to October 2010 (Period 1, during which 1,599 patients were prospectively enrolled) and from January 2003 to September 2003 (Period 2, during which 1,055 patients were retrospectively enrolled). In order to evaluate the potential risk of the perioperative dis-continuation of aspirin and clopidogrel in patients undergoing non-cardiac surgery, we investigated Period 2, in which the rate of prescription of anti-platelet agents for the secondary prevention of cardio-vascular disease was lower than that observed in Period 1 (Supplementary Table 2). This increased the chance of obtaining patients with some cardiovascular disease who had not taken antiplatelet agents for the secondary prevention of cardiovascular disease and who could be used as controls against those under anti-platelet therapy. Patients undergoing cardiac/vascular surgical procedures with systemic heparinization and those with underlying diseases accompanied by blood coagulation disorders were excluded from the analysis.

end of the hemostatic equilibrium, increasing the risk of reoperation, the overall length of stay and hospital costs2). A state of hypercoagulability has long been known to develop after surgery and plays a key role in the delicate balance between thrombotic and bleeding events in the perioperative period3-5). This issue has drawn considerable attention in relation to the use of preoperative antiplatelet therapy, in particular, the decision to stop the administration of aspirin and clopidogrel6, 7). With the aging population, a substan-tial portion of patients undergoing non-cardiac sur-gery have underlying cerebro- and cardiovascular dis-eases and frequently take aspirin for secondary preven-tion. In order to properly assess the guidance for peri-operative management, including the choice of surgi-cal technique and the location and intensity of post-operative care, it is necessary to first elucidate the inci-dence of vascular complications after surgery.

A number of epidemiological studies, largely in populations of European descent, have investigated the incidence of thrombotic complications during the perioperative period1, 8-12), i.e. MI and stroke due to arterial thrombosis and PE and DVT due to venous thromboembolism (VTE). The reported incidence of major perioperative cardiac events (including MI) is 3.9% in patients with or at risk of cardiac disease8) and 1.4% in relatively unselected patients1) undergo-ing non-cardiac surgery, although this rate appears to largely depend on the definition adopted9). The inci-dence of perioperative stroke has also been reported to be 0.08-0.7%10), depending on the type of surgical procedure and patient risk factors. The incidence of VTE has been reported to be 0.63% after non-cardiac surgery11), with the presence of racial or ethnic varia-tion13, 14). There are few data regarding the periopera-tive risk of bleeding in patients undergoing non-car-diac surgery; one study reported the incidence of major bleeding to be 5.4% in patients treated with orthopedic surgery12). In these epidemiological stud-ies, demographic factors associated with each vascular complication have been sought in order to identify at-risk patients and institute appropriate prophylactic measures in the perioperative period.

In addition to epidemiological studies, hemo-static and coagulant factors have been evaluated peri-operatively in patients undergoing non-cardiac surgery to identify biological markers reflecting thrombotic complications5, 15). Here, a number of studies have shown significant ethnic differences in baseline and perioperative coagulation profiles as well as the inci-dence of thrombotic events between Asian (including Japanese) and European populations16-20). Therefore, the question of whether previous findings in Europe-

416 Aoyama-Mani et al.

incidence of hemorrhagic complications was also assessed according to standardized bleeding definitions in the Randomized Evaluation of Long-Term Antico-agulant Therapy (RE-LY) Trial 22). The outcomes included stroke, ischemic heart disease, occlusion of other vessels for arterial thrombosis and PE or DVT due to VTE. The incidence of all vascular complica-tions was documented until 30 days after surgery.

Blood Coagulation TestsBlood samples were drawn from the peripheral

vein preoperatively on the day of surgery or a few days before surgery (baseline data) and in the morning on the first, third and fifth postoperative days (PODs 1, 3 and 5). Perioperative changes in blood coagulation were monitored according to the following parame-ters: prothrombin time (PT), activated partial throm-boplastin time (aPTT), antithrombin Ⅲ (AT), protein C activity and free protein S antigens, thrombin-anti-thrombin Ⅲ complex (TAT) and soluble fibrin mono-mer complex (SFMC).

The samples were drawn into tubes buffered with 3.13% sodium citrate. For the analysis of PT, aPTT and AT, the samples were subsequently spun at 3,500 rpm for five minutes at 5℃ to yield platelet-poor plasma. Aliquots of plasma were stored at 5℃ until the further analysis. Before the analysis, the samples were brought back to 37℃, and the PT and aPTT

In addition, because stroke has been shown to occur in ~30% of subarachnoid hemorrhage (SAH) patients, principally due to vasospasm21), SAH patients were excluded from the analysis of thrombosis.

Ethical approval for the study was acquired from the Ethics Committee of NCGM, and written informed consent was obtained from all patients examined in the second part of the study.

Data CollectionA series of data were collected from each patient’s

clinical chart for the analysis. The information included: age, sex, race, height, body weight, medical history, smoking habits and operation-related infor-mation, which further included the operative proce-dure, emergency or elective surgery, operative time, type of anesthesia, anesthesia time, amount of bleed-ing during the operation, the pre-/peri-operative use of antiplatelet and/or anticoagulant agents (amount, withdrawal period before surgery and timing of resumption after surgery), the use of heparin replace-ment (or bridge) therapy in the perioperative period and the incidence of perioperative thrombotic and hemorrhagic complications and their time of onset. In this study, antiplatelet and/or anticoagulant agents were discontinued for a certain length of time during the perioperative period in most cases, and a portion of patients received heparin replacement therapy. The

Study Design

Part 1: Review of the clinical chartsPart 2: Blood testsfor postoperative

changes

Initial study sample (n=2,654)

Period 1 (n=1,599)

Period 2 (n=1,055)

Procoagulation/thrombosis factor (n=80)

( ) ( )

Exclusion criteria:1) Age <20 years (n=113)2) Blood coagulation disorders (n=4)3) Undergoing cardiac/vascular surgery with

Platelet function (n=43)

Random selection(excl emergency surgery) ) g g g y

systemic heparinization (n=153)

Study sample for the evaluation of hemorrhagic events (n=2,384)

(excl. emergency surgery)

hemorrhagic events (n 2,384)

Exclusion criteria:4) Under perioperative antiplatelet Tx5) Subarachnoid hemorrhage (n=50)

Study sample for the evaluation of thrombotic events (n=2,326)

Fig. 1Fig.1. Schematic representation of the study design

Tx, therapy.

417Perioperative Vascular Complications

operative time and amount of bleeding.In all statistical tests, a value of P＜0.05 was con-

sidered to be significant.

Results

Incidence of Vascular Complications and Their Associated Factors

A total of 2,654 patients were consecutively enrolled from a variety of specialties in two study peri-ods: 1,599 subjects in Period 1 and 1,055 subjects in Period 2 (Fig.1). Between the two study periods, while several background characteristics (i.e. age, smoking status and dyslipidemia) and operative char-acteristics (i.e. the anesthesia time) differed, no signifi-cant differences were found in the incidence of peri-operative thrombotic or hemorrhagic complications (Supplementary Table 2). Therefore, in order to esti-mate the overall incidence of vascular complications and evaluate their associated factors, we combined the patients from the two study periods. Before the analy-sis of bleeding events, we excluded 270 individuals according to the exclusion criteria: 153 patients undergoing cardiac/vascular surgery with systemic heparinization, 113 patients ＜20 years of age and four patients with blood coagulation disorders. The number of patients according to specialty is shown in Supplementary Table 1. In the resultant panel of 2,384 individuals (Table 1 and Supplementary Table 2), 51 bleeding events (2.1%) -- 23 major bleeding events and 28 minor bleeding events -- were detected during the follow-up (up to POD 30) period (Supple-mentary Table 3). Bleeding events tended to occur in the early postoperative period; 75% (38 of 51) of bleeding events occurred on or before POD 7. The subjects treated with emergency surgery and/or the preoperative use of anticoagulant agents appeared to exhibit an increased risk of major bleeding (Table 1). When compared to the corresponding control groups, the relative risk of major bleeding associated with emergency surgery and the preoperative use of antico-agulant agents was 7.02 (95% CI, 3.1-15.8; P=9.7×10−6) and 21.0 (95% CI, 9.1-48.8; P=1.8×10−7), respectively, in the examined cohort. An advanced age, history of cerebro-/cardiovascular disease and hyper-tension were associated with major bleeding. The pre-operative use of anticoagulant agents was also associ-ated with minor bleeding (Table 1).

Before the analysis of thrombotic events, we fur-ther excluded eight individuals who did not discon-tinue antiplatelet agents in the perioperative period. With respect to thrombosis, 50 SAH patients were excluded from the analysis; 19 cerebral infarction cases

were measured using thromboplastin and rabbit brain cephalin (reference range for PT, 70-130%; aPTT, 22.0-37.0 sec.), respectively. The AT was measured using a kinetic, functional chromogenic anti-Xa method [reference range, 80-130% of normal human plasma (NHP)]. For the analysis of the protein C activity, free protein S antigens, TAT and SFMC, the samples were spun at 3,000 rpm for 10 minutes at 5℃ and stored at −20℃, then brought back to 37℃ before the analysis. The protein C activity and free protein S antigen level were measured using a chro-mogenic substrate assay and the latex agglutination method, respectively (reference range for Protein C, 70-140% of NHP; Protein S, 60-150% of NHP). The TAT was measured using an enzyme immunosorbent assay (ELISA) method (reference range, 1.0-4.1 μg/l). The SFMC was measured using a latex turbidimetric immuno assay (reference range, ≤ 6.1 μg/ml).

Platelet Aggregation TestThe platelet aggregation test was performed

using a laser-light scattering aggregometer PA-200 (Kowa Company Ltd., Tokyo, Japan). Blood samples were collected twice, once on the day of surgery or a few days before surgery (baseline data) and once in the morning on POD 1. The blood samples were carefully drawn with a 1:9 volume of 3.13% sodium citrate via venipuncture without arm squeezing, then spun at 500 rpm for 15 minutes to yield platelet-rich plasma. The aggregating agents were collagen (1.00 μg/ml), adenosine 5’-diphosphate (ADP) (2.00 μM) and epi-nephrine (10.00 μg/ml). The platelet aggregation activ-ity was evaluated with a maximum aggregation rate (%MAX)23).

StatisticsThe statistical tests were performed with the

IBM SPSS ver. 21 software program (IBM Corp. Armonk, NY, USA). Student’s t-test (for continuous variables) and the chi-squared test (for categorical vari-ables) were used to compare data for the two study groups. In order to assess the strength of the relation-ships between the tested variables, Spearman’s rank correlation coefficient was calculated along with its statistical significance.

In the assessment of risks for perioperative thrombotic complications, with regard to the periop-erative discontinuation of either of two antiplatelet agents (aspirin and clopidogrel), a multivariate regres-sion analysis and propensity score matching were per-formed to adjust for the following confounding fac-tors: age, sex, present status of smoking, disease sta-tuses of hypertension, diabetes and dyslipidemia,

418 Aoyama-Mani et al.

patients receiving prophylactic anticoagulant therapy (Supplementary Table 4), reflecting the prothrom-botic conditions of the underlying diseases. While the difference did not reach a significant level, presumably due to the small absolute number of vascular events in the present study, a history of cerebro-/cardiovascular disease and the preoperative use of antiplatelet agents tended to be prominent in the stroke cases.

Although not included in the analysis according to the exclusion criteria, three other patients were identified to have arterial thrombosis events among the 78 patients undergoing cardiac/vascular surgery with systemic heparinization but not extracorporeal circulation (see Supplementary Table 5). One death was observed among the patients with thrombotic events in the present study.

Management of Antiplatelet/Anticoagulant Agents in the Perioperative Period

In the initial study sample (n=2,654), 335 indi-viduals (13%) used antiplatelet agents preoperatively: 225 (of 335; 67%) took aspirin and 34 received the coadministration of ≥ 2 antiplatelet agents. The tim-

(38% of SAH) were detected, in agreement with the results of previous studies21). Five arterial thrombosis and five VTE events among 2,326 individuals (0.21% each) were identified during the follow-up period (Table 2). Arterial thrombosis progressed to stroke in all five cases and occurred on PODs 2 and 3, except in one case, while VTE involved both PE and DVT and occurred between PODs 8 and 30, except in one case. Significant (P＜0.05) associations were observed between arterial thrombosis and an advanced age and between VTE and the amount of bleeding during sur-gery (Table 1). These associations remained significant or nearly significant (P=0.006-0.07), even when the distinction between the study periods (Period 1 and Period 2) was adjusted for in the logistic regression model (data not shown). The Japanese guidelines for the prevention of VTE24) were formulated in 2004, between Period 1 and Period 2 in the present study; however, there were no significant differences in the incidence of VTE between the periods (Supplemen-tary Table 2). Furthermore, risk factors for VTE (i.e. age) did not appear to considerably influence its inci-dence, which tended to be elevated (0.7%) among the

Table 1. Clinical characteristics of participants according to the status of perioperative vascular complications

Total study sample†

Evaluation of thrombotic events‡ Evaluation of hemorrhagic events§

Arterial thrombosis

Venous throm-boembolism

No apparent thrombotic events

Minor bleeding

Major bleeding

No apparent hemorrhagic events

n (men/women)Age, yearsBody mass index, kg/m2

Case of emergency operation (%)Operation time, minAnesthesia time, minAmount of bleeding during operation, mlHistory of cerebro-/cardio-vascular disease (%)Hypertension (%)Diabetes mellitus (%)Dyslipidemia (%)Current smoking (%)Preoperative use of antiplatelet agents (%)Preoperative use of anticoagulant agents (%)Perioperative prophylactic anticoagulant therapy (%)

2384 (1221/1163)55.5±17.722.3±3.7

319 (13.4%)166.0±125.9227.2±137.0238.7±536.9217 (9.1%)738 (31.0%)264 (11.1%)322 (13.5%)629 (26.4%)253 (10.6%)

51 (2.1%)301 (12.6%)

5 (3/2)71.8±16.5＊

22.0±2.22 (40%)

186.4±74.4247.6±71.3342.4±331.2

2 (40%)3 (60%)1 (20%)1 (20%)1 (20%)2 (40%)1 (20%)0 (0%)

5 (1/4)52.2±22.521.7±4.2

0 (0%)356.0±272.0424.2±298.9787.6±882.0*

1 (20%)2 (40%)1 (20%)0 (0%)1 (20%)1 (20%)0 (0%)2 (40%)

2316 (1194/1122)55.3±17.722.3±3.7277 (12%)

162.2±123.9222.8±134.4233.3±521.4

203 (9%)701 (30%)258 (11%)312 (14%)612 (26%)235 (10%)49 (2.1%)

298 (13%)

28 (11/17)56.4±19.622.6±3.2

2 (7%)206.0±231.5269.0±250.1289.5±394.0

3 (11%)12 (43%)

4 (14%)7 (25%)9 (32%)4 (14%)3 (11%)＊

7 (25%)

23 (9/14)63.3±18.4＊

21.2±2.712 (50%)＊＊

158.6±99.7233.2±126.1418.5±586.7

6 (25%)＊

12 (50%)＊

4 (17%)4 (17%)3 (13%)3 (13%)7 (29%)＊＊

5 (21%)

2333 (1201/1132) 55.4±17.6 22.4±3.7269 (12%)

165.6±124.4 226.7±135.3 236.3±537.7

208 (9%) 714 (31%) 256 (11%) 311 (13%) 617 (27%) 246 (11%)

41 (2%) 289 (12%)

Values are means±SD unless otherwise indicated. †Exclusion criteria were: (1) patients aged ＜20 years, (2) those with some underlying diseases accompanied by blood coagulation disorder, and (3) those undergoing cardiovascular surgery with systemic heparinization. ‡Patients who did not discontinue antiplatelet agents in the perioperative period and those with subarachnoid hemorrhage (SAH) were further excluded from the evaluation of thrombotic events. §Major bleeding was defined by ≥ 1 of the RE-LY Trial criteria (hemoglobin reduction, ≥ 2.0 g/L; blood transfusion, ≥ 2 units; symptomatic bleed-ing in a critical area or organ). Otherwise, the events were defined as minor bleeding.＊P＜0.05; ＊＊P＜0.005 by χ2 test (categorical variables) or t-test (continuous variables), vs. the corresponding reference category (without apparent vascular events) in each evaluation.

419Perioperative Vascular Complications

After surgery, antiplatelet agents were resumed on PODs 4-11, on average. In the case of aspirin, the timing of preoperative withdrawal and perioperative discontinuation appeared to be widely distributed, with a mean of 9.0 and 16.4 days, respectively (Fig.2). Only 36 (of 335) patients received perioperative bridging anticoagulation with heparin.

Among 78 individuals under the preoperative administration of warfarin, the average timing of pre-operative withdrawal and postoperative resumption was 7.4 and 5.6 days, respectively, with 35 patients receiving perioperative bridging anticoagulation (Sup-plementary Table 6). Seven of 23 (30%) patients

ing of the preoperative withdrawal and postoperative resumption of antiplatelet agents is summarized in Supplementary Table 6. Overall, the timing of pre-operative withdrawal in this study tended to be later than that generally recommended in guidelines. For example, in the case of aspirin, the average period was 9.0 days before surgery, and 61% of the patients dis-continued the drug, in accordance with the recom-mended period (7-10 days) or longer. Several types of antiplatelet agents (i.e. cilostazol, sarpogrelate and dipyridamole) that are recommended to be discontin-ued two to three days before surgery tended to be dis-continued at an earlier time in this study.

Table 2. Patients with thrombotic complications in the perioperative and follow-up (up to POD 30) periods

Type of complication

Onset Age (y) SexTarget disease for operation

Method of surgery AnesthesiaPreoperative

antiplatelet/anticoagulant agent; timing of withdrawal

Perioperative anticoagulant therapy

Postoperative recovery

Transient ischemic

attack

POD 3 80 F Ascending colon cancer

Laparoscopic right hemicolectomy

EPI and GA Limaprost alfadex unknown dose; Pre.OD 10 (not resumed

postoperatively)

None Sitting position on POD 1

Stroke POD 2 67 M Cerebral hemorrhage

Craniotomy for removal of hematoma

GA Warfarin, bridge Tx (not resumed postoperatively)

None Lying on the bed (clouded

consciousness)

Stroke POD 3 84 F Subcortical bleeding

Craniotomy for removal of hematoma

GA None None Lying on the bed (clouded

consciousness)

Stroke POD 3 45 M Primary systemic amyloidosis

Bone marrow harvest GA None None Start walking on POD 1

Stroke POD 30 83 M Zygomatic fracture

Open reduction and internal fixation

GA Aspirin 100 mg/day; Pre.OD 11 (resumed on POD 10)

None Sitting position on POD 1

Deep vein thrombosis

POD 10 75 F Gastric carcinoma and cholelithiasis

Total gastrectomy and cholecystectomy

EPI and GA Ticlopidine hydrochloride 100 mg/day; Pre.OD 23 (not resumed postoperatively)

None Start walking on POD 2

Deep vein thrombosis

POD 20 76 F Subglottic stenosis

Tracheotomy GA None None Start walking on operation day

Pulmonary embolism

POD 1 28 F Premature rupture of the membrane

Cesarean section GA None Heparin 5,000 IU/day, s.c.; discontinued

on POD 1

Wheelchair on POD 2, Start walking on

POD 3

Pulmonary embolism

POD 8 48 F Pseudomyxoma peritonei

Cytoreductive surgery EPI and GA None Heparin 10,000 IU/day, started on POD 4

Start walking on POD 6

Pulmonary embolism

POD 10 34 M Cerebellar neoplasm

Tumor resection GA None None Wheelchair on POD 5

Abbreviations: EPI, epidural anesthesia; GA, general anesthesia; bridge Tx, bridging anticoagulation therapy; s.c., subcutaneus injection; i.v., intravenous injection; Pre.OD, preoperative day; POD, postoperative day. Two SAH patients suffered from venous thromboembolism, one each from deep vein thrombosis and pulmonary embolism; they were not included in the table, following the exclusion criteria (see Methods).

420 Aoyama-Mani et al.

The relative changes in the coagulation factors serially measured on PODs 1, 3 and 5 are shown in Fig.3. Downregulators of coagulation -- AT, the pro-tein C activity and free protein S antigens -- all decreased considerably in the perioperative period. Harmoniously, a marker of subclinical thrombosis, TAT, considerably increased, supporting a tendency toward hypercoagulability. In contrast, although mod-est in fold change, the global coagulation parameters -- PT-INR and aPTT -- exhibited significant (P＜0.001) increases, i.e., a tendency toward anticoagula-tion (Fig.3 and Supplementary Table 9), which appeared to be counterintuitive in terms of assumed hypercoagulation after surgery. Notably, the changes in blood coagulation factors were most prominent on POD 1, except for TAT, which remained significantly elevated by ＞2 fold across PODs 1-5.

The platelet aggregation activity was found to be reduced by 13-23% on POD 1 as a whole, using three aggregating agents -- collagen, ADP and epinephrine (Fig.4); this was also counterintuitive.

Next, we examined the correlations between perioperative changes (from baseline to the POD 1 data) in the tested coagulation-related factors (Fig.5). There were fair correlations between the variables reflecting similar functions, i.e. between anticoagula-tion indicators (PT-INR and aPTT), between down-regulators of coagulation (AT, the protein C activity and free protein S antigens) and between the platelet aggregation activity levels associated with collagen, ADP and epinephrine. In particular, we observed neg-ative correlations (r =−0.47-−0.71; P＜0.001) between the indicators of anticoagulation and down-regulators of coagulation, suggesting the possible pres-

who experienced major bleeding events were under the preoperative administration of warfarin (Supple-mentary Table 3).

The use of both antiplatelet and anticoagulant agents was considerably (P＜0.005) associated with risk factors for cardiovascular disease, such as male sex, an advanced age, history of cerebro-/cardiovascular disease and current smoking (Supplementary Table 7). It has been reported that the withdrawal of aspirin and clopidogrel, but not other antiplatelet agents, potentially induces a rebound or prothrombotic effect7). In the present study, we found only one patient who took either aspirin or clopidogrel preop-eratively among the stroke group (Table 2). Therefore, we were unable to robustly assess the potential risk of the perioperative interruption of these two antiplatelet agents (aspirin and clopidogrel) for perioperative thrombotic complications using a multivariate regres-sion analysis [risk ratio (RR) for perioperative stroke=1.8 (95% CI, 0.2-19.8; P=0.63)] or multivar-iate regression analysis with the propensity score for preoperative antiplatelet drug use [RR=3.0 (95% CI, 0.2-39.6; P=0.39)].

Perioperative Changes in Blood Coagulation and Platelet Aggregation Activity

To assess the clinical and laboratory evidence for hypercoagulability after surgery, we characterized the perioperative changes in pro- or anticoagulant factors in a subgroup of patients treated without the adminis-tration of antiplatelet/anticoagulant agents (Fig.1): 80 and 43 individuals examined for blood coagulation factors and the platelet function, respectively (Supple-mentary Table 8).

25

30 Perioperative discontinuationPreoperative withdrawal

15

20

25

ber o

f cas

esPreoperative withdrawal

0

5

10

Num

b

0 5 10 15 20 25 30 35 40 45Timing [days]

Fig. 2

Fig.2. Preoperative withdrawal and perioperative discontinuation of aspirin

The distribution of the preoperative withdrawal (within 30 days before surgery; bars in light blue) and perioperative discontinuation (within 30 days before and after surgery; bars in dark blue) is shown in the figure. Not all subjects in the former counting are included in the latter counting, partly because they did not resume the agent postoperatively (see Supplementary Table 6).

421Perioperative Vascular Complications

assess the clinical relevance, we examined the correla-tions between each of the coagulation-related factors and their potentially associated variables (Table 3). There were fair correlations between the collagen-induced platelet aggregation factors and the patient’s age and between TAT, PT-INR and protein C and the operative (and anesthesia) time. Scatter plots showed that perioperative changes in collagen-induced platelet aggregation were positive (or more than zero) in more than half of the individuals ＞70 years of age; that is, an increase in platelet aggregation appeared to be pro-nounced in the elderly (Supplementary Fig.1). The scatter plots also showed that the correlations between perioperative TAT changes on different PODs became less prominent as time passed after surgery (Supple-mentary Fig.2).

Discussion

Limited data are available for the incidence and risk factors of thrombotic and hemorrhagic complica-tions in patients undergoing non-cardiac surgery (Supplementary Table 10). In the present study, we collected outcome data for a total of 2,654 consecu-tive patients ≥ 20 years of age presenting for non-car-diac surgery at a single tertiary medical care center and

ence of counterbalance in the immediate postoperative period. The markers of a hypercoagulable state -- TAT and SFMC -- did not exhibit high correlations with coagulation-related factors other than collagen-induced platelet aggregation. Furthermore, in order to

***

e)

******

2 5ge

(log

-sca

le

2.0POD 1POD 3

2.5

1 5

Fold

cha

n

1.0

*********

******

POD 3POD 5

1.5

0.75 ******

*** ********

*

0.65

Fig. 3

Fig.3. Postoperative changes in procoagulation/thrombosis factors

Fold changes were calculated by dividing the median of individual measure-ments on PODs 1, 3 and 5 by the measurement obtained at baseline (preopera-tive). POD, postoperative day; TAT, thrombin-antithrombin Ⅲ complex; SFMC, soluble fibrin monomer complex; PT-INR, prothrombin time interna-tional normalized ratio; aPTT, activated partial thromboplastin time; AT, anti-thrombin Ⅲ. The vertical bars indicate the ranges from the 25th to 75th percen-tiles of the fold change for individual factors.＊P＜0.05, ＊＊P＜0.01, ＊＊＊P＜0.001, according to the paired t-test, vs. the baseline data.

Fig.4. Postoperative changes in platelet aggregation parame-ters, shown in a maximum aggregation rate (%MAX)

Blood test dates (upper row) and aggregating agents (lower row) are shown on the x-axis. Pre-ope, preoperative period; POD, post-operative day; ADP, adenosine 5’-diphosphate.＊P＜0.05, ＊＊P＜0.001, according to the paired t-test. The bars indicate the mean±SD.

70

80

90[%]

*** **

est

40

50

60

70

ggre

gatio

n Te

10

20

30

40

Per

cent

Ag

0Pre-ope POD 1 Pre-ope POD 1 Pre-ope POD 1

Collagen ADP Epinephrine

Fig. 4

422 Aoyama-Mani et al.

presumably accounting for the increased risk of thrombotic and hemorrhagic events. Of particular note is the fact that perioperative changes in platelet aggregation activity were significantly correlated with an individual patient’s age and showed opposite direc-tions on POD 1 at almost the turn of 70 years of age; i.e., these changes were increased in patients ≥ 70 years of age and decreased in those ＜70 years of age (Sup-plementary Fig.1). This phenomenon appears to account, at least in part, for the high incidence of peri-operative stroke observed in the elderly patients.

Among arterial thrombosis complications in the perioperative period, MI and stroke have drawn sub-

found the following results: 1. the incidence of peri-operative stroke was 0.21% and the overall incidence of VTE (involving PE and DVT) was 0.21%, both in accordance with previously reported rates19, 20); and 2. events of perioperative stroke tended to occur on or before POD 3, while those of VTE were distributed in a later phase of the postoperative period (POD 8 to POD 20). To our knowledge, this is the first study to report the recent incidence of VTE using a hospital-based cohort design in a Japanese population. Further-more, our data for a subgroup of patients (n=82 in total) revealed that a series of coagulation-related fac-tors changed dynamically in the perioperative period,

Table 3. Correlation of perioperative changes (from baseline to POD-1 data) in coagulation-related factors with potentially associated variables

Collagen-induced PA TAT PT-INR Protein C

Age, yearsBody mass index, kg/m2

Platelet count, ×104/μlAspartate aminotransferase, U/lOperation time, minAnesthesia time, minAmount of bleeding, ml

0.62＊＊

−0.020.120.34＊

0.090.070.07

0.05−0.15−0.15

0.30＊

0.33＊＊

0.33＊＊

−0.1

0.11−0.07

0.080.33＊＊

0.49＊＊

0.44＊＊

0.27＊

−0.28＊

−0.01 −0.05 −0.33＊

−0.41＊＊

−0.44＊＊ −0.15

The numbers of patients tested for the individual parameters were: 43 for PA, and 80 for TAT, PT-INR, and protein C. Only the results for a selected list of parameters are shown in the table, with consideration of inter-parameter correlations in Figure 5. Abbreviations: PA, platelet aggregation; TAT, thrombin-antithrombin Ⅲ complex; PT-INR, prothrombin time international normalized ratio. ＊P＜0.05; ＊＊P＜0.005 by Spearman’s rank correlation analysis.

Fig.5. Correlations between postoperative changes (from baseline to the POD 1 data) in the coag-ulation-related factors tested in this study

The interparameter correlations were examined in 41 patients tested for both procoagulation/thrombosis factors and the platelet function (see Supplementary Table 8). Modest (0.4 ≤|r|<0.7) and strong (0.7≤|r|) correlations are highlighted. For the parameters with a normal distribution, Pearson and Spearman’s corre-lation analyses did not show any significant differences in the correlation coefficients. TAT, thrombin-anti-thrombin Ⅲ complex; SFMC, soluble fibrin monomer complex; PT-INR, prothrombin time international normalized ratio; aPTT, activated partial thromboplastin time; AT, antithrombin Ⅲ; PA, platelet aggrega-tion; ADP, adenosine 5’-diphosphate.

TAT SFMC PT-INR aPTT AT Protein C Protein S PA_collagen PA_ADP PA_ epinephrine

TAT 7 8E 13 0 121 0 068 0 024 0 015 0 33 0 002 0 065 0 11TAT 7.8E-13 0.121 0.068 0.024 0.015 0.33 0.002 0.065 0.11SFMC 0.77 0.05 0.053 0.027 0.044 0.76 4.0E-04 0.007 0.007PT-INR 0.18 0.26 2.1E-11 1.6E-04 6.8E-08 4.8E-05 0.196 0.872 0.583aPTT 0.21 0.25 0.66 1.9E-04 2.2E-10 8.8E-08 0.023 0.506 0.188

PAT -0 3 -0 29 -0 48 -0 47 3 3E-06 3 0E-05 0 071 0 620 0 915 P-value

AT 0.3 0.29 0.48 0.47 3.3E 06 3.0E 05 0.071 0.620 0.915Protein C -0.32 -0.27 -0.64 -0.71 0.57 1.9E-08 0.152 0.856 0.493Protein S -0.13 -0.04 -0.51 -0.64 0.52 0.66 0.028 0.666 0.186PA_collagen 0.47 0.54 0.21 0.36 -0.29 -0.23 -0.35 6.6E-07 8.5E-08

PA_ADP 0.29 0.42 0.03 0.11 -0.08 -0.03 -0.07 0.68 4.9E-10PA_epinephrine 0.25 0.42 0.09 0.21 0.02 -0.11 -0.21 0.71 0.78

Spearman's rank Correlation Coefficient

Fig. 5

423Perioperative Vascular Complications

acute-phase reactions, such as tissue factor release from damaged vessels, inflammatory activation and com-promised fibrinolysis5). Despite clinical and laboratory evidence of postoperative hypercoagulability, there is no sufficient description regarding the time course of coagulation changes and their dynamic regulation in patients undergoing non-cardiac surgery. Several stud-ies have examined the ability to detect perioperative changes in various coagulation components for the purpose of assessing hypercoagulability5, 15). Such stud-ies have mostly focused on the intraoperative or immediate postoperative period in a small number (n＜30) of highly selected patients, i.e. those undergoing cardiac or vascular surgery or cholecystectomy15, 27, 28). One previous study of Europeans demonstrated sig-nificant hypercoagulability beginning on POD 2 after major surgery using a combination of coagulation tests in 51 patients5). In the present study, we newly report that perioperative changes in coagulation-related factors are complex and substantially correlated in the mixed direction of pro- and anticoagulation (Fig.5), as summarized in Supplementary Table 11. For instance, some coagulation components (e.g., the platelet aggregation activity) are suppressed en masse in the immediate postoperative period (POD 1; Fig. 4), while others appears to be significantly modified by a patient’s age (Supplementary Fig.1). In this line, it has been reported that the platelet aggregation activ-ity continues to increase in the later postoperative period (on POD 6)5). We also found that the routine coagulation parameters PT and aPTT exhibit statisti-cally significant (P＜0.001) perioperative changes, although modest in relative changes (~11% increase, compared to the preoperative data), in the direction of anticoagulation, whereas TAT shows prolonged (across PODs 1-5) and prominent (＞116% increase) periop-erative activation of coagulation (Fig.3 and Supple-mentary Table 9). Taken together, an integrated approach is essential for interpreting coagulation-related hematological data, with correction for an individual patient’s background and clinical manifes-tations.

Another issue of interest is the perioperative management of antiplatelet therapy. Recent data regarding the risk of discontinuing antiplatelet therapy in patients with coronary stents have highlighted this issue6, 7, 29). Among commonly-used antiplatelet drugs, aspirin and clopidogrel are particularly considered to induce a rebound or prothrombotic effect when dis-continued6, 30). As a result, the routine withdrawal of aspirin 7-10 days before surgery, which is principally aimed at reducing the risk of bleeding complications, has been questioned29). In addition, little evidence has

stantial attention due to their high mortality and seri-ous disability8-10). It has been reported that, while the incidence of MI and stroke in the general non-cardiac surgical population is relatively low (~1.4%)1, 10), the prevalence of these diseases is significantly increased in subgroups of patients with or at risk for cardiovascular disease. A previous Japanese study reported that the incidence of perioperative MI and stroke is 0.33% and 0.34%, respectively, based on a 16-year longitudinal observation period19). Although the present study con-firmed the low incidence of stroke (0.21%), the post-operative temporal distribution of the condition appears to differ between the two studies; i.e., the medial time to stroke was three days in this study (Table 2) compared to nine days in the previous study19). In agreement with the findings of previous studies10, 19), an advanced age (＞70 years) was found to constitute a principal risk factor for perioperative stroke (Table 1). On the other hand, we did not iden-tify any perioperative MI patients in the present study. This may be due to the possibility that we missed asymptomatic MI patients, considering the previously reported high incidence of such cases9), as discussed below.

VTE is known to occur relatively less frequently in Japanese patients than in other ethnic groups, although there have been few reports of the epidemi-ology of VTE apart from questionnaire surveys24) and literature searches25). Against this background, one study20) recently reported that the incidence of VTE is 0.24% (DVT 0.19% and PE 0.05%) using a diagno-sis-procedure combination database that involves administrative data for approximately 1 million surgi-cal patients from 260 hospitals obtained between April 2008 and March 2010. Despite the similar incidence of overall VTE (0.21% in the present study), the inci-dence of PE (0.13%) was equivalent to or slightly higher than the reported incidence (0.05% in the entire population and 0.23% in the high-frequency group undergoing gastroenterologic or orthopedic sur-gery)20). When we focus on high-risk (cancer patients ≥ 40 years of age) and intermediate-risk (elderly patients ≥ 60 years of age) groups for VTE, the inci-dence remains low (0.1-0.2%) in the present study (Supplementary Table 4). Notably, our data indicate that VTE events occur in both the early and relatively late phase (POD 8 or later) of the postoperative period. Despite the careful management of prophy-laxis in the clinical setting, these findings support the need for prolonged VTE prophylaxis in at-risk surgi-cal patients, as previously indicated26).

The hypercoagulable state is an issue of primary interest and can be induced by surgery itself due to

424 Aoyama-Mani et al.

perioperative discontinuation of aspirin and/or clopi-dogrel on the incidence of perioperative thrombotic complications. However, it is likely that these effects reflect the pre-existing risk for cardiovascular disease in patients already on antiplatelet therapy as well as the risk of a rebound phenomenon. Fourth, although dynamic perioperative changes in coagulation-related factors indicate an overall tendency of hypercoagula-bility, the present study was not designed to correlate such hemostatic changes with thrombotic complica-tions at the individual patient level. Further prospec-tive studies are needed to evaluate, in a longer postop-erative period, the predictive value of test variables for detecting the development of thrombotic complica-tions.

Conclusion

Our cohort data support the relatively low inci-dence of perioperative thrombotic complications in the general non-cardiac surgical population of Japa-nese descent. Our data also suggest the need for pro-longed VTE prophylaxis in at-risk surgical patients. We found that stroke and bleeding complications tend to occur in the early (until POD 7) postoperative period. Given the delicate balance between throm-botic and bleeding events during the corresponding period, it is important to understand the association between baseline risk factors and vascular complica-tions and to integrate such knowledge with serially measured coagulation-related hematological data for effective perioperative management, although solid evidence is difficult to obtain due to the substantial complexity of the patient conditions in the postopera-tive period.

Acknowledgments

We acknowledge the patients who participated in this study; without their kind cooperation, this study would not have been possible. We are also grateful to Drs. Takuro Shimbo, Noriko Tanaka and Ngoc Minh Pham, medical research coordinators and laboratory technicians at NCGM, for their assistance in the data analysis and management. C.A.-M. is also grateful for the staff members of the Department of Anesthesiol-ogy at NCGM. This study was supported by a Grant from the National Center for Global Health and Medicine (22-119).

Conflicts of Interest

The authors have no direct or indirect conflicts

been provided with respect to patients taking aspirin and/or clopidogrel for stable coronary artery disease in the setting of non-cardiac surgery. In our cohort data, aspirin and clopidogrel were routinely discontinued 8-9 days before surgery, on average (Fig.2 and Sup-plementary Table 6). Because only one patient under treatment of aspirin and/or clopidogrel developed perioperative stroke, we were unable to calculate the robust value, although we broadly estimated that an increased risk for discontinuation in the perioperative period of non-cardiac surgery was not immense; the risk ratio for perioperative stroke was roughly esti-mated to be 1.8 (95% CI, 0.2-19.8) according to a multivariate regression analysis. A multicenter, ran-domized controlled trial -- the ASPIRIN trial (Anti-platelet Strategies in the Perioperative Period in Patients at Risk of Ischaemic Events) -- is ongoing and expected to provide definitive guidelines for those tak-ing aspirin in the perioperative period of non-cardiac surgery7).

The present study is associated with several limi-tations. First, since the first part of the study was an observational study, the data are limited to variables measurable in the clinical setting and subject to under-estimation of medical comorbidities and the target endpoints. For example, the diagnosis of VTE was confirmed solely based on the findings of diagnostic imaging methods, which are not necessarily performed in cases in which the symptoms are very mild. In addi-tion, a recent study demonstrated that nearly two-thirds of patients with perioperative MI are asymp-tomatic and that obtaining a diagnosis based on the clinical chart will lead to underestimation of the inci-dence of the disease9). Nevertheless, our cohort data may have some advantages over the results of ques-tionnaire-based epidemiological studies involving a large number of hospitals in terms of consecutive sam-pling and the quantity and completeness of individual clinical data required for the analysis of associated risk factors. Second, perioperative bleeding is often diffi-cult to define. We used the RE-LY trial criteria for the definition of major and minor bleeding in surgical patients22). The incidence of bleeding varies based on the criteria and type of surgery, ranging from 0% (skin incision) to 75% (transurethral resection of the pros-tate)31). Similar to that observed for thrombotic com-plications, it is not feasible to perform a complete medical record review in all subjects in order to deter-mine the signs/symptoms of thrombosis. Third, since a vast majority of our patients taking aspirin and/or clopidogrel discontinued the medications before sur-gery (Supplementary Table 6), we may have arbi-trarily estimated the effects of the preoperative use and

425Perioperative Vascular Complications

14) Montagnana M, Favaloro EJ, Franchini M, Guidi GC, Lippi G: The role of ethnicity, age and gender in venous thromboembolism. J Thromb Thrombolysis, 2010; 29: 489-496

15) Schietroma M, Carlei F, Mownah A, Franchi L, Mazzotta C, Sozio A, Amicucci G: Changes in the blood coagula-tion, fibrinolysis, and cytokine profile during laparoscopic and open cholecystectomy. Surg Endosc, 2004; 18: 1090-1096

16) Bocaz JA, Barja P, Bonnar J, Daly L, Carrol A, Coutinho E, Goncalves M, Tsakok M, Koh S, Thomson JM, Poller L, Heady A, Holck S, Pinol A: Differences in coagulation and haemostatic parameters in normal women of child-bearing age from different ethnic groups and geographical locations. Task Force on Oral Contraceptives--WHO Spe-cial Programme of Research, Development and Research Training in Human Reproduction. Thromb Haemost, 1986; 55: 390-395

17) Singhal D, Smorodinsky E, Guo L: Differences in coagu-lation among Asians and Caucasians and the implication for reconstructive microsurgery. J Reconstr Microsurg, 2011; 27: 57-62

18) White RH, Keenan CR: Effects of race and ethnicity on the incidence of venous thromboembolism. Thromb Res, 2009; 123(Suppl 4): S11-S17

19) Kikura M, Oikawa F, Yamamoto K, Iwamoto T, Tanaka KA, Sato S, Landesberg G: Myocardial infarction and cerebrovascular accident following non-cardiac surgery: differences in postoperative temporal distribution and risk factors. J Thromb Haemost, 2008; 6: 742-748

20) Kunisawa S, Ikai H, Imanaka Y: Incidence and prevention of postoperative venous thromboembolism: are they meaningful quality indicators in Japanese health care set-tings? World J Surg, 2012; 36: 280-286

21) Vergouwen MD: Participants in the International Multi-Disciplinary Consensus Conference on the Critical Care Management of Subarachnoid Hemorrhage. Vasospasm versus delayed cerebral ischemia as an outcome event in clinical trials and observational studies. Neurocrit Care, 2011; 15: 308-311

22) Connolly SJ, Ezekowitz MD, Yusuf S, Eikelboom J, Old-gren J, Parekh A, Pogue J, Reilly PA, Themeles E, Varrone J, Wang S, Alings M, Xavier D, Zhu J, Diaz R, Lewis BS, Darius H, Diener HC, Joyner CD, Wallentin L: Dabiga-tran versus warfarin in patients with atrial fibrillation. N Engl J Med, 2009; 361: 1139-1151

23) Eto K, Takeshita S, Ochiai M, Ozaki Y, Sato T, Isshiki T: Platelet aggregation in acute coronary syndromes: use of a new aggregometer with laser light scattering to assess platelet aggregability. Cardiovasc Res, 1998; 40: 223-229

24) Editorial Committee on Japanese Guideline for Preven-tion of Venous Thromboembolism: Japanese Guideline for Prevention of Venous Thromboembolism. Medical Front International, Tokyo, 2004 (Japanese)

25) Sakon M, Kakkar AK, Ikeda M, Sekimoto M, Nakamori S, Yano M, Monden M: Current status of pulmonary embolism in general surgery in Japan. Surg Today, 2004; 34: 805-810

26) Huo MH, Muntz J: Extended thromboprophylaxis with low-molecular-weight heparins after hospital discharge in

of interest to declare.

References

1) Lee TH, Marcantonio ER, Mangione CM, Thomas EJ, Polanczyk CA, Cook EF, Sugarbaker DJ, Donaldson MC, Poss R, Ho KK, Ludwig LE, Pedan A, Goldman L: Deri-vation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation, 1999; 100: 1043-1049

2) Stokes ME, Ye X, Shah M, Mercaldi K, Reynolds MW, Rupnow MF, Hammond J: Impact of bleeding-related complications and/or blood product transfusions on hos-pital costs in inpatient surgical patients. BMC Health Serv Res, 2011; 11: 135

3) Aillaud MF, Juhan-Vague I, Alessi MC, Marecal M, Vin-son MF, Arnaud C, Arnaud C, Vaque P, Collen D: Increased PA-inhibitor levels in the postoperative period--no cause-effect relation with increased cortisol. Thromb Haemost, 1985; 54: 466-468

4) Nielsen VG, Asmis LM: Hypercoagulability in the periop-erative period. Best Pract Res Clin Anaesthesiol, 2010; 24: 133-144

5) Lison S, Weiss G, Spannagl M, Heindl B: Postoperative changes in procoagulant factors after major surgery. Blood Coagul Fibrinolysis, 2011; 22: 190-196

6) Jaffer AK: Perioperative management of warfarin and antiplatelet therapy. Cleve Clin J Med, 2009; 76(Suppl 4): S37-S44

7) O’Riordan JM, Margey RJ, Blake G, O’Connell PR: Antiplatelet agents in the perioperative period. Arch Surg, 2009; 144: 69-76

8) Devereaux PJ, Goldman L, Cook DJ, Gilbert K, Leslie K, Guyatt GH: Perioperative cardiac events in patients undergoing noncardiac surgery: a review of the magnitude of the problem, the pathophysiology of the events and methods to estimate and communicate risk. CMAJ, 2005; 173: 627-634

9) Devereaux PJ, Xavier D, Pogue J, Guyatt G, Sigamani A, Garutti I, Leslie K, Rao-Melacini P, Chrolavicius S, Yang H, Macdonald C, Avezum A, Lanthier L, Hu W, Yusuf S: Characteristics and short-term prognosis of perioperative myocardial infarction in patients undergoing noncardiac surgery: a cohort study. Ann Intern Med, 2011; 154: 523-528

10) Macellari F, Paciaroni M, Agnelli G, Caso V: Perioperative stroke risk in nonvascular surgery. Cerebrovasc Dis, 2012; 34: 175-181

11) Rogers SO Jr, Kilaru RK, Hosokawa P, Henderson WG, Zinner MJ, Khuri SF: Multivariable predictors of postop-erative venous thromboembolic events after general and vascular surgery: results from the patient safety in surgery study. J Am Coll Surg, 2007; 204: 1211-1221

12) Oberweis BS, Nukala S, Rosenberg A, Guo Y, Stuchin S, Radford MJ, Berger JS: Thrombotic and bleeding compli-cations after orthopedic surgery. Am Heart J, 2013; 165: 427-433

13) Bergqvist D: Geographical aspects of postoperative venous thromboembolism. J Thromb Haemost, 2005; 3: 26-27

426 Aoyama-Mani et al.

ative to antiplatelet therapy in patients with prior percutaneous coronary intervention. J Am Coll Cardiol, 2012; 60: 2005-2016

30) Lordkipanidzé M, Diodati JG, Pharand C: Possibility of a rebound phenomenon following antiplatelet therapy withdrawal: a look at the clinical and pharmacological evidence. Pharmacol Ther, 2009; 123: 178-186

31) Gerstein NS, Schulman PM, Gerstein WH, Petersen TR, Tawil I: Should more patients continue aspirin therapy perioperatively?: clinical impact of aspirin withdrawal syn-drome. Ann Surg, 2012; 255: 811-819

high-risk surgical and medical patients: a review. Clin Ther, 2009; 31: 1129-1141

27) Horne MK 3rd, Merryman PK, Cullinane AM, Nghiem K, Alexander HR: The impact of major surgery on blood coagulation factors and thrombin generation. Am J Hematol, 2007; 82: 815-820

28) Pleym H, Wahba A, Videm V, Asberg A, Lydersen S, Bjella L, Dale O, Stenseth R: Increased fibrinolysis and platelet activation in elderly patients undergoing coronary bypass surgery. Anesth Analg, 2006; 102: 660-667

29) Singla S, Sachdeva R, Uretsky BF: The risk of adverse car-diac and bleeding events following noncardiac surgery rel-

427Perioperative Vascular Complications

60

70

POD

3

60

70

POD

5

40

50

60

chan

ges

on P

ng/m

l] 40

50

60

chan

ges

on P

g/m

l]

10

20

30

osto

pera

tive

c [n

10

20

30

stop

erat

ive

c [ng

00 20 40 60

Po

Postoperative changes on POD 1[ng/ml]

00 20 40 60

Po

Postoperative changes on POD 1[ng/ml]

Supplementary Figure 2. Correlations among the perioperative changes in the TAT levels between [POD 3 - Pre ope] and [POD

[ng/ml] [ng/ml]

Correlations among the perioperative changes in the TAT levels between [POD 3 - Pre.ope] and [POD 1 - Pre.ope] and between [POD 5 - Pre.ope] and [POD 1 - Pre.ope].TAT, thrombin/antithrombin III complex.

Supplementary Fig.2. Correlations among the perioperative changes in the TAT levels between [POD 3 - Pre.ope] and [POD 1 - Pre.ope] and between [POD 5 - Pre.ope] and [POD 1 - Pre.ope]

TAT, thrombin-antithrombin Ⅲ complex.

10 20 30 40

PA [Δ

%]

-30-20 -10

0

en-in

duce

d P

-60 -50 -40 -30

0 50 100

Col

lag

0 50 100

Age [years]

0.6

0.7

0.6

0.7

0.3

0.4

0.5

T-IN

R

0.3

0.4

0.5

T-IN

R

0

0.1

0.2PT

0

0.1

0.2PT

-0.10 500 1000

Operative time [min]

-0.10 500 1000

Anesthesia time [min]

Supplementary Figure 1. Scatter plots for representative pairs of significant correlations. Perioperative changes (from baseline to the POD 1 data) are shown for collagen-induced platelet aggregation (PA) and PT-INR. PT-INR, prothrombin time international normalized ratio.

Supplementary Fig.1. Scatter plots for representative pairs of significant correlations

Perioperative changes (from baseline to the POD 1 data) are shown for collagen-induced platelet aggregation (PA) and PT-INR. PT-INR, prothrombin time international normalized ratio.

428 Aoyama-Mani et al.

Supplementary Table 1. Number of participants according to specialty

Specialty n (men/women) Age, years

General surgeryCardiac/vascular surgeryThoracic surgeryNeurosurgeryOrthopaedicsOphthalmologyOtorhinolaryngologyDermatologyUrologyGynaecologyPlastic surgeryOral surgeryObstetricsHaematologyRadiology

8000

162217214

3446

281

30041802512

1

(441/359) (0/0) (108/54) (116/101) (122/92) (2/1) (293/153) (1/1) (66/15) (0/300) (27/14) (37/43) (0/25) (7/5) (1/0)

62.1±15.6–

58.6±16.557.0±16.257.3±17.955.7±15.350.9±17.685.5±6.464.3±13.445.6±14.747.2±17.641.9±18.331.4±6.137.0±9.2

79.0

The values are presented as the mean±SD, unless otherwise indicated.

Supplementary Table 2. Clinical characteristics of all participants initially enrolled in the study

All data Period 1 Period 2

n (men / women)Age, yearsBody mass index, kg/m2

Emergency operation (%)Operation time, minAnesthesia time, minAmount of bleeding during operation, mlHistory of cerebro-/cardio-vascular disease (%)Hypertension (%)Diabetes mellitus (%)Dyslipidemia (%)Current smoking (%)Preoperative use of antiplatelet agents (%)†

Preoperative use of anticoagulant agents (%)Preoperative prophylactic anticoagulant therapy (%)Postoperative prophylactic anticoagulant therapy (%)Postoperative thrombotic complications (%)‡

Arterial thrombosis (%)Venous thrombosis (%)

Postoperative hemorrhagic complications (%)Major Bleeding (%)Minor Bleeding (%)

2384 (1221 / 1163)55.5±17.722.3±3.7

319 (13.4%)166.0±125.9227.2±137.0238.7±536.9217 (9.1%)738 (31.0%)264 (11.1%)322 (13.5%)629 (26.4%)253 (10.6%)

51 (2.1%)42 (1.8%)

267 (11.2%)10 (0.4%)

5 (0.2%)5 (0.2%)

51 (2.1%)23 (1.0%)28 (1.2%)

1451 (750 / 701)56.4±17.922.4±3.8

196 (13.5%)161.9±122.8220.7±133.5217.6±133.5147 (10.1%)479 (33.0%)169 (11.6%)226 (15.6%)415 (28.6%)172 (11.9%)

37 (2.5%)36 (2.5%)

174 (12.0%)7 (0.5%)5 (0.3%)2 (0.1%)

28 (1.9%)12 (0.8%)16 (1.1%)

933 (471 / 462)54.1±17.2＊

22.3±3.6123 (13.2%)172.2±130.5237.5±141.8＊

271.6±474.270 (7.5%)＊

259 (27.8%)＊

95 (10.2%)96 (10.3%)＊＊

214 (22.9%)＊＊

81 (8.7%)＊

14 (1.5%)6 (0.6%)

93 (10.0%)3 (0.3%)0 (0.0%)3 (0.3%)

23 (2.5%)11 (1.2%)12 (1.3%)

The values are presented as the mean±SD, unless otherwise indicated.†All types of antiplatelet agents were included (see Supplementary Table 6).＊P＜0.05, ＊＊P＜0.005 according to the χ2 test (categorical variables) or t-test (continuous variables); Period 1 vs. Period 2.‡Patients who did not discontinue antiplatelet agents in the perioperative period and those with subarachnoid hemorrhage (SAH) were fur-ther excluded from the evaluation of thrombotic events.

429Perioperative Vascular ComplicationsSu

pple

men

tary

Tab

le 3

.Pa

tien

ts w

ith

blee

ding

com

plic

atio

ns in

the

peri

oper

ativ

e an

d fo

llow

-up

(up

to P

OD

30)

per

iods

Type

of c

ompl

icatio

nRE

-LY

trial

defin

ition

Ons

etAg

eG

ende

rTa

rget

dise

ase f

or o

pera

tion

Met

hod

of su

rger

yEm

erge

ncy

Ope

ratio

nAn

esth

esia

Preo

pera

tive a

ntip

latele

t ag

ent

(tim

ing o

f with

draw

al)

Posto

pera

tive u

se o

f an

tiplat

elet a

gent

(ti

min

g of

resu

mpt

ion)

Preo

pera

tive

antic

oagu

lant a

gent

(ti

min

g of

with

draw

al)

Perio

pera

tive

brid

ge T

xPo

stope

rativ

e use

of

antic

oagu

lant a

gent

(ti

min

g of

resu

mpt

ion)

Alve

olar

hem

orrh

age

majo

rPO

D 1

083

MSu

bara

chno

id h

emor

rhag

eIn

tracr

anial

aneu

rysm

s sur

gery

No

GA

No

No

No

No

No

Hem

oper

itone

umm

ajor

POD

3†

62F

Ova

rian

tum

orAd

nexe

ctom

yN

oG

AAs

pirin

100

mg/

day

(Pre

.OD

6);

dipi

ryda

mol

e 30

0mg/

day (

Pre.O

D 1

2)

Aspi

rin 1

00 m

g/da

y (P

OD

5)

War

farin

4 m

g/da

y (P

re.O

D 3

)H

epar

in 1

5,00

0 IU

/day

, fro

m P

OD

1

to P

OD

19

War

farin

4 m

g/da

y (P

OD

3)

Hem

oper

itone

umm

ajor

Ope

ratio

n da

y32

FN

on-re

assu

ring f

etal

statu

sCa

esar

ean

sect

ion

Yes

GA

No

No

No

Hep

arin

5,0

00 IU

/da

y, s.c

. PO

D 1

No

Intra

cran

ial h

emor

rhag

em

ajor

POD

153

MSu

bara

chno

id h

emor

rhag

eIn

tracr

anial

aneu

rysm

s sur

gery

Yes

GA

No

No

No

No

No

Intra

cran

ial h

emor

rhag

em

ajor

POD

11

27F

Suba

rach

noid

hem

orrh

age

Intra

cran

ial an

eury

sms s

urge

ryYe

sG

AN

oN

oN

oN

oN

o

Intra

cran

ial h

emor

rhag

em

ajor

POD

843

FH

ydro

ceph

alus

Vent

ricul

o-pe

riton

eal

shun

t pro

cedu

reN

oG

AN

oN

oN

oN

oN

o

Intra

cran

ial h

emor

rhag

em

ajor

POD

267

MCe

rebr

al he

mor

rhag

eCr

anio

tom

y for

rem

oval

of h

emat

oma

Yes

GA

No

No

War

farin

, dos

e un

know

n, w

/o

with

draw

al

No

No

Intra

cran

ial h

emor

rhag

em

ajor

POD

2†

66M

Cere

bral

hem

orrh

age

Cran

ioto

my f

or re

mov

al of

hem

atom

aYe

sG

AAs

pirin

100

mg/

day,

w/o

with

draw

alAs

pirin

100

mg/

day

(PO

D 1

5)W

arfar

in 5

.5 m

g/da

y, w/

o wi

thdr

awal

Hep

arin

15,

000

IU/d

ay, f

rom

PO

D

1 to

PO

D 2

War

farin

5.5

mg/

day

(PO

D 1

5)

Intra

cran

ial h

emor

rhag

em

ajor

POD

184

FCe

rebr

al he

mor

rhag

eCr

anio

tom

y for

rem

oval

of h

emat

oma

Yes

GA

No

No

No

No

No

Mele

nam

ajor

POD

662

MFe

mor

al ne

ck fr

actu

reFe

mor

al he

ad p

rosth

etic

repl

acem

ent

No

GA

No

No

No

No

No

Mele

nam

ajor

POD

12‡

75F

Colo

nic s

tom

a pro

lapse

Colo

stom

yN

oG

AN

oAs

pirin

100

mg/

day

(PO

D 1

1)W

arfar

in 1

.75

mg/

day

(Pre

.OD

7)

No

War

farin

1.7

5 m

g/da

y (P

OD

2)

Mele

nam

ajor

POD

19‡

85F

Ilium

frac

ture

Ope

n re

duct

ion

and

inte

rnal

fixat

ion

No

GA

No

No

War

farin

2 m

g/da

y (P

re.O

D 1

2)N

oW

arfar

in 2

mg/

day

(PO

D 7

)

Mele

nam

ajor

POD

546

MPe

rirec

tal a

bsce

ssCo

losto

my

Yes

GA

No

No

No

No

No

Mele

nam

ajor

POD

14

87F

Inte

stina

l obs

truct

ion

due t

o str

angu

latio

n Ri

ght h

emico

lecto

my

Yes

EPI a

nd

GA

No

No

No

No

No

Subc

utan

eous

blee

ding

majo

rPO

D 2

47F

Ova

rian

tum

orSi

mpl

e hys

tere

ctom

y an

dadn

exec

tom

yN

oG

AN

oN

oN

oH

epar

in 5

,000

IU/

day,

s.c. P

OD

1N

o

Subc

utan

eous

blee

ding

majo

rPO

D 2

72F

Sigm

oid

colo

n pe

rfora

tion

Har

tman

n op

erat

ion

Yes

GA

No

No

No

No

No

Subc

utan

eous

blee

ding

majo

rO

pera

tion

day

62F

Non

-tube

rcul

ous

myc

obac

terio

sisTh

orac

osco

pic l

ung r

esec

tion

No

EPI a

nd

GA

No

No

No

No

No

Subc

utan

eous

blee

ding

majo

rPO

D 3

47F

Posto

pera

tive h

emor

rhag

eLa

paro

tom

y and

dra

inag

eYe

sG

AN

oN

oN

oN

oN

o

Subc

utan

eous

blee

ding

majo

rPO

D 1

83M

Sigm

oid

volvu

lus

Sigm

oide

ctom

yN

oEP

I and

G

AN

oN

oW

arfar

in 1

mg/

day

(Pre

.OD

10)

No

War

farin

1 m

g/da

y (P

OD

24)

Subc

utan

eous

blee

ding

majo

rPO

D 1

72M

Sigm

oid

colo

n pe

rfora

tion

Har

tman

n op

erat

ion

Yes

GA

No

No

No

No

No

Subc

utan

eous

blee

ding

majo

rPO

D 1

91F

Hip

join

t disl

ocat

ion

Ope

n re

duct

ion

Yes

GA

Aspi

rin 1

00 m

g/da

y, w/

o wi

thdr

awal

Aspi

rin 1

00 m

g/da

y (P

OD

15)

No

No

No

Subc

utan

eous

blee

ding

majo

rPO

D 2

68M

Lung

canc

erLe

ft lo

wer l

obec

tom

yN

oEP

I and

G

AN

oN

oW

arfar

in 5

mg/

day

(Pre

.OD

3)

Hep

arin

10,

000

IU/d

ay, f

rom

Pre

.O

P 5

to P

re.O

P 1

No

Subc

utan

eous

blee

ding

majo

rO

pera

tion

day

68F

Sigm

oid

colo

n ca

ncer

Sigm

oide

ctom

yN

oEP

I and

G

AN

oN

oN

oN

oN

o

Epist

axis

min

orPO

D 1

69F

Cyst

in th

e max

illar

y sin

usRa

dica

l ope

ratio

n of

m

axill

ary s

inus

No

GA

No

No

No

No

No

430 Aoyama-Mani et al.Ep

istax

ism

inor

Ope

ratio

n da

y25

MCh

roni

c sin

usiti

s Co

ncho

tom

yN

oG

AN

oN

oN

oN

oN

o

Hem

atur

iam

inor

POD

773

MBe

nign

rena

l tum

orPa

rtial

neph

rect

omy

No

GA

No

No

War

farin

8 m

g/da

y (P

re.O

D 1

1)H

epar

in 1

0,00

0 IU

/day

, fro

m P

re.

OP

11 to

PO

D 2

6

War

farin

8 m

g/da

y (P

OD

24)

Hem

atur

iam

inor

POD

5†

63M

Blad

der c

arcin

oma

Tran

sure

thra

l res

ectio

n of

bl

adde

r tum

orN

oG

AAs

pirin

100

mg/

day

(Pre

.OD

10)

Aspi

rin 1

00 m

g/da

y (P

OD

20)

War

farin

4.5

mg/

day

(Pre

.OD

6)

Hep

arin

20,

000

IU/d

ay, f

rom

Pre

.O

P 8

to P

OD

20

War

farin

4.5

mg/

day

(PO

D 2

0)

Hem

oper

itone

umm

inor

POD

26

45F

Fibr

omyo

ma i

n th

e ute

rus

Sim

ple h

yste

rect

omy

No

EPI a

nd

GA

No

No

No

Hep

arin

5,0

00 IU

/da

y, s.c

. PO

D 1

No

Intra

cran

ial h

emor

rhag

em

inor

POD

277

FN

orm

al pr

essu

re

hydr

ocep

halu

sVe

ntric

ulo-

perit

onea

l sh

unt p

roce

dure

No

GA

No

No

No

No

No

Intra

cran

ial h

emor

rhag

em

inor

POD

12

27F

Brain

tum

orTu

mor

rese

ctio

nN

oG

AN

oN

oN

oN

oN

o

Subc

utan

eous

blee

ding

min

orPO

D 6

70F

Panc

reat

ic ca

ncer

Panc

reat

icodu

oden

ecto

my

No

EPI a

nd

GA

No

No

No

No

No

Subc

utan

eous

blee

ding

min

orPO

D 2

127

FFi

brom

yom

a in

the u

teru

sEn

uclea

tic m

yom

ecto

my

No

GA

No

No

No

No

No

Subc

utan

eous

blee

ding

min

orPO

D 4

32M

Chro

nic t

onsil

litis

Tons

illec

tom

yN

oG

AN

oN

oN

oN

oN

o

Subc

utan

eous

blee

ding

min

orPO

D 1

74F

Incis

iona

l ven

tral h

erni

aH

erni

oplas

tyN

oG

AN

oN

oN

oN

oN

o

Subc

utan

eous

blee

ding

min

orPO

D 1

72F

Brea

st tu

mor

Mas

tect

omy

No

GA

No

No

No

No

No

Subc

utan

eous

blee

ding

min

orPO

D 1

073

MH

ilar b

ile d

uct c

arcin

oma

Righ

t hem

ihep

atec

tom

yN

oEP

I and

GA

No

No

No

No

No

Subc

utan

eous

blee

ding

min

orPO

D 4

69F

Brea

st ca

ncer

Mas

tect

omy

No

GA

No

No

No

No

No

Subc

utan

eous

blee

ding

min

orPO

D 1

046

FFi

brom

yom

a in

the u

teru

sSi

mpl

e hys

tere

ctom

yN

oEP

I and

G

AN

oN

oN

oH

epar

in 5

,000

IU/

day,

s.c. P

OD

1N

o

Subc

utan

eous

blee

ding

min

orPO

D 1

378

FSu

bara

chno

id h

emor

rhag

eIn

tracr

anial

aneu

rysm

s su

rger

yN

oG

AN

oN

oW

arfar

in 1

mg/

day

(Pre

.OD

3)

No

No

Subc

utan

eous

blee

ding

min

orPO

D 4

56F

Fibr

omyo

ma i

n th

e ute

rus

Sim

ple h

yste

rect

omy

No

EPI a

nd

GA

No

No

No

Hep

arin

5,0

00 IU

/da

y, s.c

. PO

D 1

No

Subc

utan

eous

blee

ding

min

orPO

D 1

78M

Chro

nic s

inus

itis

Endo

scop

ic en

dona

sal

sinus

surg

ery

No

GA

Ticlo

pidi

ne h

ydro

chlo

ride

100

mg/

day (

Pre.O

D 8

)U

nkno

wnN

oN

oN

o

Subc

utan

eous

blee

ding

min

orPO

D 4

22F

Chro

nic t

onsil

litis

Tons

illec

tom

yN

oG

AN

oN

oN

oN

oN

o

Subc

utan

eous

blee

ding

min

orPO

D 1

171

MEs

opha

geal

carc

inom

aEs

opha

gect

omy

No

EPI a

nd

GA

No

No

No

No

No

Subc

utan

eous

blee

ding

min

orO

pera

tion

day

82F

Spin

al co

rd tu

mor

Tum

or re

sect

ion

No

GA

No

No

No

No

No

Subc

utan

eous

blee

ding

min

orPO

D 6

46F

Fibr

omyo

ma i

n th

e ute

rus

Sim

ple h

yste

rect

omy

No

GA

No

No

No

Hep

arin

5,0

00 IU

/da

y, s.c

. PO

D 1

No

Subc

utan

eous

blee

ding

min

orPO

D 3

53F

Aden

omyo

sis u

teri

Sim

ple h

yste

rect

omy

No

EPI a

nd

GA

No

No

No

Hep

arin

5,0

00 IU

/da

y, s.c

. PO

D 1

No

Subc

utan

eous

blee

ding

min

orPO

D 1

57F

Brea

st ca

ncer

Mas

tect

omy

No

GA

No

No

No

No

No

Subc

utan

eous

blee

ding

min

orPO

D 2

44M

Acut

e app

endi

citis

Appe

ndec

tom

yYe

sEP

I and

G

AAs

pirin

81

mg/

day

(Pre

.OD

9)

No

No

No

No

Subc

utan

eous

blee

ding

min

orPO

D 6

75F

Supr

acon

dylar

frac

ture

of

hum

erus

Ope

n re

duct

ion

and

inte

rnal

fixat

ion

Yes

GA

Aspi

rin 1

00 m

g/da

y, w/

o wi

thdr

awal

Aspi

rin 1

00 m

g/da

y (P

OD

2)

No

No

No

Subc

utan

eous

blee

ding

min

orO

pera

tion

day

23M

Tem

poro

man

dibu

lar jo

int

diso

rder

Reco

nstru

ctive

surg

ery

of m

andi

ble

No

GA

No

No

No

No

No

Subc

utan

eous

blee

ding

min

orPO

D 5

51M

Rhin

itis h

yper

troph

ica

chro

nica

Endo

scop

ic en

dona

sal

sinus

surg

ery

No

GA

No

No

No

No

No

Subc

utan

eous

ble

edin

g: s

ubcu

tane

ous

blee

ding

in th

e w

ound

ed a

rea.

†B

ecau

se th

e pa

tien

t und

erw

ent r

epla

cem

ent o

f a m

echa

nica

l val

ve in

the

hear

t, an

tico

agul

ant t

hera

py c

ould

not

be

disc

onti

nued

des

pite

ble

edin

g.‡

Whi

le th

e re

sum

ed d

ose

of w

arfa

rin

was

iden

tica

l to

that

use

d in

the

preo

pera

tive

per

iod,

the

PT-I

NR

was

pro

long

ed, l

ikel

y re

sulti

ng in

the

occu

rren

ce o

f mel

ena.

Abb

revi

atio

ns: E

PI, e

pidu

ral a

nest

hesi

a; G

A, g

ener

al a

nest

hesi

a; b

ridg

e T

x, b

ridg

ing

anti

coag

ulat

ion;

s.c

., su

bcut

aneo

us in

ject

ion;

i.v.

, int

rave

nous

inje

ctio

n; P

re.O

D, p

reop

erat

ive

day;

PO

D, p

osto

pera

tive

day

.

431Perioperative Vascular Complications

Supplementary Table 5. Patients with thrombotic complications who were excluded from the analysis according to the exclusion criteria (cardio/vascular surgery with systemic heparinization)

Type of complication Onset Age (y) Sex Target disease for operation

Method of surgery

Anesthesia Preoperative antiplatelet/anticoagulant agent; timing of withdrawal

Postoperative use of anticoagulant agent (timing

of recommencement)

Postoperative recovery

Arterial thrombosis in the legs

Operation day

81 FAbdominal

aortic aneurysm

Endovascular aneurysm

repairGA

Sarpogrelate hydrochloride 200 mg/day; Pre.OD 12 (not

restarted postoperatively), Beraprost 180 mg/day; Pre.

OD 32 (restarted on POD 6)

NoneStart walking on POD 1

Occlusion of superior mesenteric artery

POD 3 78 FAngina pectoris

Coronary artery

bypass graftGA

Aspirin 100 mg/day; Pre.OD 1 (not restarted

postoperatively), Dipyridamole 150 mg/day; started on POD

2.

Warfarin 3 mg/day (POD 3)

Lying on the bed (clouded

consciousness)

Stroke POD 6 80 MThoracic aortic

aneurysm

Endovascular aneurysm

repairGA None None

Start walking on POD 2

Among the excluded cases, 75 patients underwent cardiac surgery with extracorporeal circulation; however, the incidence of stroke is not listed in the table.Abbreviations: GA, general anesthesia; Pre.OD, preoperative day; POD, postoperative day.

Supplementary Table 4. Incidence of venous thromboembolism (VTE) and participant characteristics according to the risk status

VariableOverall

Age stratified subgroup Cancer, age ≥ 40 y(High risk group)

Prophylactic anti-coagulant TxAge ＜40y 40y ≤ Age ＜60y Age ≥ 60y

Incidence of VTE (men/women), %n (men/women)Age, yearsBody mass index, kg/m2

Proportion of major surgery (%)Operative time ≥ 45 min (%)

Cancer (%)History of cerebro-/cardio-vascular disease (%)Hypertension (%)Diabetes mellitus (%)Dyslipidemia (%)Current smoking (%)Preoperative use of antiplatelet agents (%)Preoperative use of anticoagulant agents (%)Perioperative prophylactic anticoagulant therapy (%)

0.2 (0.1/0.4)2321 (1195/1126)

52.2±22.521.7±4.22073 (89.3)2060 (88.8)

782 (33.7)

204 (8.8)

703 (30.3)259 (11.0)312 (13.4)613 (26.4)

236 (10.2)

49 (2.1)†

300 (12.9)

0.4 (0.4/0.3)528 (237/291)

30.0±5.722.0±3.9447 (84.7)438 (83.0)

37 (7.0)

2 (0.4)

13 (2.5)6 (1.1)8 (1.5)

68 (12.9)

5 (0.9)

1 (0.2)

104 (19.7)

0.1 (0/0.3)704 (344/360)

50.0±5.922.8±3.9640 (90.9)639 (90.8)209 (29.7)

19 (2.7)

137 (19.5)55 (7.8)60 (8.5)

170 (24.1)

29 (4.1)

7 (1.0)

104 (14.8)

0.2 (0/0.4)1089 (614/475)

71.1±7.322.2±3.5986 (90.5)983 (90.3)536 (49.2)

183 (16.8)

553 (50.8)198 (18.2)244 (22.4)375 (34.4)

202 (18.5)

41 (3.8)

92 (8.4)

0.1 (0/0.3)745 (420/325)

66.0±11.222.2±3.6710 (95.3)706 (94.8)745 (100.0)

85 (11.4)

306 (41.1)125 (16.8)131 (17.6)260 (34.9)

102 (13.7)

11 (1.5)

63 (8.5)

0.8 (0/0.8)266 (8/258)46.9±16.022.2±3.9266 (100)262 (98.5)

51 (19.2)

9 (3.4)

48 (18.0)19 (7.1)30 (11.3)23 (8.6)

17 (6.4)

5 (1.9)

266 (100)

Major surgery was arbitrarily defined as abdominal surgery and/or an operative time of ≥ 45 min, referring to the Japanese guidelines for VTE (ref.24). The values are presented as the mean±SD, unless otherwise indicated. For each variable, the proportion in relation to the whole number of subjects in the corresponding group (or subgroup) was calculated and shown in the parentheses.†A patient with arterial thrombosis and VTE was excluded from the table.

432 Aoyama-Mani et al.

Supplementary Table 6. Preoperative withdrawal and postoperative resumption of antiplatelet/anticoagulant agents

n Preoperative withdrawal timing

(days) [n]‡

Postoperative resumption

timing (days) [n]§

＃of cases who did not resume drugs postoper-

atively

＃of cases who did not discon-

tinue drugs perioperatively

＃of cases whose

postoperative prescription is

unknown

＃of cases who changed drugs

Antiplatelet agent†

Aspirin (%)Clopidogrel bisulfate (%)Ticlopidine hydrochloride (%)Cilostazol (%)Sarpogrelate hydrochloride (%)Ethyl icosapentate (%)Beraprost sodium (%)Dipyridamole (%)Limaprost alphadex (%)

Anticoagulant agentWarfarin (%)

225153834

316103332

78

9.0±6.9 [173]8.2±6.7 [13]

10.8±8.2 [31]7.9±5.8 [27]8.7±4.2 [3]7.4±6.6 [12]7.7±5.2 [7]9.3±8.9 [23]9.1±7.7 [16]

7.4±6.1 [71]

7.4±6.6 [121]4.0±3.5 [4]9.2±5.8 [13]7.8±7.9 [12]

- [0]4.7±4.0 [3]6.8±8.5 [4]5.5±4.8 [14]

11.0±10.7 [4]

5.6±6.7 [53]

77101914

384

1115

23

(34%) (67%) (50%) (41%) (100%) (50%) (40%) (33%) (47%)

(30%)

912302256

1

(4%) (7%) (5%) (9%) (0%) (13%) (20%) (15%) (19%)

(1%)

1804503037

1

(8%) (0%) (11%) (15%) (0%) (19%) (0%) (9%) (22%)

(1%)

930110101

0

(4%) (20%) (0%) (3%) (33%) (0%) (10%) (0%) (3%)

(0%)

The values are presented as the mean±SD.†More than one antiplatelet agent was prescribed in 34 patients.‡Patients who withdrew agents within 30 preoperative days were included in the calculation.§Patients who were documented to resume antiplatelet/anticoagulant agents by POD 30 were included in the calculation. Patients who postopera-tively resumed antiplatelet agents different from the preoperative agents were excluded from the calculation.Abbreviations: POD, postoperative day.

Supplementary Table 7. Clinical characteristics of the participants according to the use of antiplatelet/anticoagulant agents

Antiplatelet agents No antiplatelet agents Anticoagulant agents No anticoagulant agents

n (men / women)Age, yearsBody mass index, kg/m2

Emergency operation (%)Operation time, minAnesthesia time, minAmount of bleeding during operation, mlHistory of cerebro-/cardio-vascular disease (%)Hypertension (%)Diabetes mellitus (%)Dyslipidemia (%)Current smoking (%)

253 (163/90)＊＊

70.2±10.4＊＊

22.8±3.3＊

28 (11%)187.9±132.9255.1±143.9＊＊

284.3±643.7132 (52%)＊＊

188 (74%)＊＊

72 (29%)＊＊

99 (39%)＊＊

122 (48%)＊＊

2131 (1059/1072)53.8±17.522.3±3.7291 (14%)

163.4±124.9223.9±135.9233.3±522.7

85 (4%)550 (26%)192 (9%)223 (11%)507 (24%)

51 (38/13)＊＊

69.9±10.9＊＊

22.9±3.913 (26%)＊

173.1±108.8248.8±130.5271.2±479.226 (51%)＊＊

36 (73%)＊＊

18 (35%)＊＊

18 (35%)＊＊

25 (49%)＊＊

2333 (1184/1149)55.1±17.722.3±3.7306 (13%)

162.6±124.8222.8±135.1236.9±540.7

191 (8%)701 (30%)246 (11%)304 (13%)604 (26%)

The values are presented as the mean±SD, unless otherwise indicated.Exclusion criteria: (1) age ＜20 years, (2) underlying diseases accompanied by blood coagulation disorders and (3) undergoing cardiovascular sur-gery with systemic heparinization. ＊P＜0.05; ＊＊P＜0.005 according to the χ2 test (categorical variables) or t-test (continuous variables), vs. the corresponding reference category (without agents).

433Perioperative Vascular Complications

Supplementary Table 8. Clinical characteristics of the participants in the blood tests for periopera-tive coagulation changes

Parameter Procoagulation/ thrombosis factor

Platelet function

n (men / women)Age, yearsBody mass index, kg/m2

Operation time, minAnesthesia time, minAmount of bleeding during operation, mlPlatelet count, ×104/μlAspartate aminotransferase, U/lHypertension (%)Diabetes mellitus (%)Dyslipidemia (%)Current smoking (%)

80 (56/24)60.43±16.0521.94±3.09257.2±125.5328.7±129.0345.0±544.7

25.9±7.728.4±24.831 (39%)14 (18%)15 (19%)17 (21%)

43 (32/11)57.51±16.1621.78±2.82284.9±154.9354.9±159.3441.3±841.8

25.8±6.128.3±25.111 (26%)

8 (19%)8 (14%)

10 (23%)

A total of 82 patients were examined for procoagulation/thrombosis factors and/or the platelet function (see Methods); i.e., 41 patients were examined for both types of blood tests, excluding cases of emergency surgery.The values are presented as the means±SD, unless otherwise indicated.The platelet count and aspartate aminotransferase level were measured in the preoperative period.

Supplementary Table 9. Time course of the measured coagulation factors

Preoperative POD 1 POD 3 POD 5

TAT, μg/lSFMC, μg/mlPT-INRaPTT, secAT, %Protein C, %Protein S, %

3.0 1.9

1.0326.7101120109

(2.0-4.4) (1.5-3.4) (0.98-1.08) (24.6-29.2) (96-113) (110-137) (96-120)

6.52.9

1.1329.6

859087

(4.3-11.2) (1.9-6.2) (1.07-1.23) (26.9-33.6) (75-95) (79-108) (80-108)

7.43.9

1.0827.5

88107

96

(5.0-13.3) (2.3-7.4) (1.02-1.12) (25.5-30.8) (75-100) (91-123) (80-113)

7.74.6

1.0927.6

98116102

(4.0-12.0) (2.8-7.3) (1.05-1.14) (25.3-29.6) (87-107) (97-133) (89-117)

The values are presented as the median (25th-75th percentiles). Abbreviations: POD, postoperative day; TAT, thrombin-antithrombin Ⅲ complex; SFMC, soluble fibrin monomer complex; PT-INR, prothrombin time international normalized ratio; aPTT, activated partial thromboplastin time; AT, antithrombin Ⅲ.

434 Aoyama-Mani et al.

Supplementary Table 10. Comparisons between the current and previously published studies of Japanese patients

Study Design Enrollment period

Type of surgery

Number of patients enrolled

Stroke＊ VTE

Incidence Postoperative distribution of vascular events

Risk factors

Incidence Fatality rate

Postoperative distribution of vascular events

Risk factors

Current Study Cohort study in a tertiary medical care

center

Nov. 2009-Oct. 2010Jan. 2003-Sep. 2003

Non-cardiac surgery

2,654 0.21% 2-3 days Advanced age

(＞70 yrs)

0.21%(DVT, 0.09%;

PE, 0.13%)

Not observed among VTE

cases included

8 days or later

Amount of bleeding during

operation

Kikura et al. 2008

Cohort study in two tertiary medical care

centers

Jan. 1991-Dec. 2006

Non-cardiac, non-carotid

surgery

36,634 0.34% Constant rate (median time to event is 9

days)

Advanced age

(＞70 yrs);Female

Not Reported

Kunisawa et al. 2012

DPC data from 260 hospitals in Japan

Apr. 2008-Mar. 2010

Surgery(all types)

1,016,496 Not Reported 0.24%(DVT, 0.19%;

PE, 0.05%)

DVT, 3.44%;

PE, 22.86%

NA NA

Sakon et al. 2004

Literature (e.g., PubMed)

search

1985-2002 (literature)

General surgery

35,369 Not Reported PE, 0.33% PE, 31% NA NA

＊In the current study, no episodes of symptomatic myocardial infarction were identified until 30 days after surgery, whereas the incidence of MI was reported to be 0.33% with the peak observed early after surgery.

Supplementary Table 11. Perioperative changes in the coagulation-related factors

Coagulation-related factor variablesChanges from preoperative day

POD 1 POD 3 POD 5

Standard coagulation test: PT, aPTT† ↓ ↓ ~→ →

Down-regulators of coagulation: AT, Pro.C, Pro.S ↑↑ ↑ →

Subclinical thrombosis marker: TAT, SFMC ↑↑ ↑↑ ↑↑

Fibrinogen‡ (→ ) (↑ ) (↑ )

Platelet aggregation test‡ ↓ (↓ ) (→ )

Fibrinolysis‡: PAI-Ⅰ (↑ ) (→ ) N/A

↑ : direction of coagulation; ↓ : direction of anticoagulation.†PT and PT-INR reflect the extrinsic pathway of coagulation, and aPTT reflects both the intrinsic and the common coagulation pathways.‡The arrows in the parentheses are based on the findings in the literature.Abbreviations: POD, postoperative day; PT, prothrombin time; aPTT, activated partial thromboplastin time; AT, antithrombin Ⅲ; Pro.C, protein C; Pro.S, protein S; TAT, thrombin-antithrombin Ⅲ complex; SFMC, soluble fibrin monomer complex; PAI-Ⅰ, plasminogen activator inhibitor Ⅰ.