Identifying Esophagectomy Patients at Risk for Predischarge Versus Postdischarge Venous Thromboembolism

GENERAL THORACIC

Identifying Esophagectomy Patients at Risk for Predischarge Versus Postdischarge Venous Thromboembolism Jeremiah T. Martin, MBBCh, FRCSI, Angela L. Mahan, MD, Victor A. Ferraris, MD, Sibu P. Saha, MD, Timothy W. Mullett, MD, Joseph B. Zwischenberger, MD, and Ching-Wei D. Tzeng, MD Department of Surgery, Division of Cardiothoracic Surgery, University of Kentucky, Lexington, Kentucky

Background. Current guidelines recommend postoperative venous thromboembolism (VTE) chemoprophylaxis for moderate-risk patients (3% rate or greater) and extended-duration chemoprophylaxis for high-risk patients (6% or greater). Large-scale studies of and recommendations for esophagectomy patients are lacking. This study was designed to evaluate the timing, rates, and predictors of postesophagectomy VTE. Methods. Patients undergoing esophagectomies for cancer were identified from the 2005 to 2012 American College of Surgeons National Surgical Quality Improvement database. Timing and rates of VTE (deep venous thrombosis or pulmonary embolism, or both) were calculated. Events were stratified as predischarge or postdischarge. Perioperative factors associated with 30day rates of predischarge and postdischarge VTE were analyzed. Results. Of 3,208 patients analyzed, the surgical approach was Ivor-Lewis (n [ 1,131, 35.3%), transhiatal (n [ 945, 29.5%), three-field (n [ 587, 18.3%), thoracoabdominal (n [ 364, 11.3%), and nongastric conduit reconstruction (n [ 181, 5.6%). Rates were 2.0%

pulmonary embolism, 3.7% deep venous thrombosis, and 5.1% VTE. Overall median length of stay was 11 days (versus 19 days, p < 0.001, if predischarge VTE). Predischarge VTE occurred on median day 9, whereas postdischarge VTE occurred on day 19 (p < 0.001). Only 17% of VTE occurred after discharge. Multivariate analysis identified being male (odds ratio [OR] 2.09, p [ 0.018), white race (OR 1.93, p [ 0.004), prolonged ventilation (OR 3.24, p < 0.001), and other major complications (OR 1.90, p [ 0.005) as independent predictors of predischarge VTE. Older age (OR 1.06 per year, p [ 0.006) and major complications (OR 3.14, p [ 0.004) were independently associated with postdischarge VTE. Conclusions. Postesophagectomy VTE occurs in a clinically significant proportion of esophageal cancer patients with identifiable risk factors for predischarge and postdischarge events. Elderly patients and patients with major complications are most likely to benefit from extended-duration chemoprophylaxis.

V

perioperative VTE chemoprophylaxis from the time of surgery to the date of discharge and even extendedduration chemoprophylaxis for select patients to prevent this complication, which is a clinically significant cause of decreased quality of life [11], delayed adjuvant therapy [12], and even death after the index hospitalization [9]. High-risk operations are associated with increased risk of postdischarge VTE, and distinct sets of predictors for predischarge and postdischarge VTE can be used to identify patients who might benefit from extendedduration chemoprophylaxis [6, 13, 14]. Guidelines exist for extended chemoprophylaxis, including those from the National Comprehensive Cancer Network (NCCN). However, these have not seen uniform clinical application [1, 4, 7, 15, 16]. Additionally, these guidelines apply to broad categories of high-risk surgical patients, and none has been specifically developed to guide the care of patients undergoing esophagectomy. For esophagectomy, traditional surgical training has always taught the need

enous thromboembolism (VTE) remains a significant cause of potentially preventable morbidity and mortality among surgical patients, especially those with active cancer [1–3]. The incidence of VTE varies with type of cancer and surgical procedure [4–6], and despite decreasing hospital lengths of stay with modern postoperative care [7], it continues to be major clinical issue after discharge [4, 8, 9]. Methods of prophylaxis against VTE may be classified as either mechanical (eg, sequential compression device, compression stockings) or chemical (eg, unfractionated and low molecular weight heparin), known as chemoprophylaxis. Numerous national and international consensus groups [10] advocate

Accepted for publication April 6, 2015. Presented at the Fifty-first Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 24–28, 2015. Address correspondence to Dr Martin, 740 S Limestone St, A301, Lexington, KY 40509; e-mail: [email protected].

Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier

(Ann Thorac Surg 2015;100:932–8) Ó 2015 by The Society of Thoracic Surgeons

0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2015.04.042

Ann Thorac Surg 2015;100:932–8

933

Statistical Analysis Descriptive statistics included mean and standard deviation for normally distributed continuous variables, median and interquartile range for nonparametric continuous variables, and counts and percentages for categoric variables. Univariate comparisons were performed using Student’s t test, analysis of variance, c2 test, and Fisher’s exact test, as appropriate. Logistic regression was used to measure the association between the occurrence and timing of VTE and other clinical variables. Significant variables were entered into a final logistic model to determine independent associations with predischarge and postdischarge VTE. Statistical analyses were performed using SAS 9.3 (SAS Institute, Cary NC). All tests were two-sided, and statistical significance was defined as p less than 0.05.

Results Patients and Methods Patients and Data Collection All patients undergoing esophagectomy (Current Procedural Terminology [CPT] 43107 to 43124) for cancer (International Classification of Diseases 150.x or 151.x) were identified in the 2005 to 2012 American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) participant use file. Surgical approach was further classified as transhiatal (CPT 43107), three-field (CPT 43122), Ivor-Lewis (CPT 43117, 43121), thoracoabdominal (CPT 43122), and use of any nongastric conduit (CPT 43108, 43113, 43116, 43118, 43123, 43124). All available NSQIP preoperative, intraoperative, and postoperative variables were examined to identify significant differences and associated risk factors between the groups.

Definitions Preoperative risk factors, intraoperative variables, and postoperative metrics have been previously defined [17]. Venous thromboembolism was a composite variable defined as the occurrence of clinically detected deep venous thrombosis or pulmonary embolus. Patients documented to have both deep venous thrombosis and pulmonary embolus were labeled only once for VTE. The incidence of VTE was then measured from the recorded 30-day NSQIP outcomes. Timing of VTE was further categorized relative to discharge date and classified as predischarge versus postdischarge events. Postoperative (non-VTE) major morbidity was defined as the occurrence of any one or combination of the following NSQIP-derived variables: pneumonia, reintubation, failure to wean from ventilator, renal insufficiency, renal failure, cardiac arrest, myocardial infarction, coma, stroke, sepsis, septic shock, return to the operating room, fascial dehiscence, or organ space infection. Based on NSQIP limitations, postesophagectomy mortality was defined as death within 30 postoperative days or death during the original index admission.

Patients and Esophagectomies Between 2005 and 2012, 3,208 patients underwent esophagectomy; they had a median age of 64 years (Table 1). The distribution of surgical approach was as follows: Ivor-Lewis (n ¼ 1131, 35.3%), transhiatal (945, 29.5%), three-field (587, 18.3%), thoracoabdominal (364, 11.3%), and nongastric conduit reconstruction (181, 5.6%). Overall, 79.3% of patients were American Society of Anesthesiologists class 3 or more, and 25.8% were smokers. The major morbidity rate, excluding VTE, for patients undergoing esophagectomy was 35.7% and that was associated with an overall 3.1% (99 of 3,208) 30-day mortality rate (Table 1).

Occurrence and Timing of VTE Overall event rates were 2.0% for pulmonary embolus, 3.7% for deep venous thrombosis, and 5.1% for VTE. Of these, 17% of VTE occurred after discharge (Fig 1). Median length of stay was 11 days for all patients versus 19 days (p < 0.001) for patients with predischarge VTE and versus 12 days (p ¼ 0.359) for patients with postdischarge VTE. Overall median day of VTE was postoperative day 11. Predischarge VTE occurred on median day 9, whereas postdischarge VTE occurred on median day 19 (p < 0.001; Fig 2).

Factors Associated With Predischarge Versus Postdischarge VTE Higher American Society of Anesthesiologists class, diabetes mellitus, preoperative dyspnea, cardiovascular history, hypertension requiring medications, and preoperative hematocrit less than 39% were identified as preoperative risk factors associated with the development of predischarge VTE on univariate analysis (Table 1). The only preoperative risk factor associated with the development of postdischarge VTE was older age. Choice of surgical approach was not associated with predischarge or postdischarge VTE (p ¼ 0.368). The occurrence of postesophagectomy major complications, however, including organ space infection (as a surrogate for leak), pneumonia, failure to wean, reintubation,

GENERAL THORACIC

for awareness of postoperative major complications such as anastomotic leak, aspiration, and respiratory failure, whereas VTE receives less attention as a complication when consenting patients for surgery. To influence national and international practice patterns, the authors felt the need to establish the clinical importance of VTE within the population of patients with cancer undergoing esophagectomy and to make recommendations to decrease VTE event rates. The hypothesis of this study was that the postoperative VTE rate after esophagectomy was clinically significant and that certain risk factors could be identified to help surgeons predict patients who were at risk for predischarge versus postdischarge events. To test this hypothesis, this study was designed to define the rates and timing of postesophagectomy VTE events and to analyze the clinical variables associated with predischarge versus postdischarge VTE.

MARTIN ET AL IDENTIFYING RISK FOR VTE AFTER ESOPHAGECTOMY

GENERAL THORACIC

94.92% 2,165 (71.1) 64.0 (57.0–71.0) 456 (15.0) 2405 (79.0) 388 (12.7) 604 (19.8) 795 (26.1) 103 (3.4) 53 (1.7) 26.9 (23.7–30.8) 225 (7.4) 323 (10.6) 293 (9.6) 1608 (52.8) 1527 (50.1) 1857 (61.0) 375 (12.3) 684 (22.5) 266 (8.7) 69 (2.3) 173 (5.7) 55 (1.8) 440 (14.4) 393 (12.9) 386 (12.7) 56 (1.8) 31 (1.0) 304 (10.0) 196 (6.4) 360 (11.8) 379 (12.4) 11 (8–16) 92 (3.0) 1,042 (34.2)

Pre-D VTE (n ¼ 135) 109 66.0 21 116 25 29 29 5 4 27.7 7 24 20 87 75 95 18 37 16 4 18 2 38 55 49 6 4 26 28 29 33 19 5 86

4.21% (80.7) (59.0–72.0) (15.6) (85.9) (18.5) (21.5) (21.5) (3.7) (3.0) (24.0–32.0) (5.2) (17.8) (14.8) (64.4) (55.6) (70.4) (13.3) (27.4) (11.9) (3.0) (13.3) (1.5) (28.1) (40.7) (36.3) (4.4) (3.0) (19.3) (20.7) (21.5) (24.4) (12–31) (3.7) (63.7)

Post-D VTE (n ¼ 28) 22 70.0 9 23 3 7 5 1 0 26.8 4 5 2 15 18 18 4 8 2 0 3 1 8 8 6 0 0 4 4 2 8 11.5 2 18

0.87% (78.6) (63.5–77.0) (32.1) (82.1) (10.7) (25.0) (17.9) (3.6) (0.0) (22.3–30.9) (14.3) (17.9) (7.1) (53.6) (64.3) (64.3) (14.3) (28.6) (7.1) (0.0) (10.7) (3.6) (28.6) (28.6) (21.4) (0.0) (0.0) (14.3) (14.3) (7.1) (28.6) (8–15) (7.1) (64.3)

Total (n ¼ 3,208) 2296 64.0 486 2544 416 640 829 109 57 27.0 236 352 315 1710 1620 1970 397 729 284 73 194 58 486 456 441 62 35 334 228 391 420 11 99 1,146

(71.6) (57.0–71.0) (15.1) (79.3) (13.0) (20.0) (25.8) (3.4) (1.8) (23.7–30.9) (7.4) (11.0) (9.8) (53.3) (50.5) (61.4) (12.4) (22.7) (8.9) (2.3) (6.0) (1.8) (15.1) (14.2) (13.7) (1.9) (1.1) (10.4) (7.1) (12.2) (13.1) (8–16) (3.1) (35.7)

p Value Pre-D vs None

p Value Post-D vs None

p Value Pre-D vs Post-D

0.015 0.059 0.857 0.051 0.028 0.639 0.230 0.815 0.304 0.167 0.335 0.009 0.047 0.008 0.758 0.044 0.874 0.246 0.213 0.551 <0.001 >0.999 <0.001 <0.001 <0.001 0.046 0.059 <0.001 <0.001 <0.001 <0.001 <0.001 0.605 <0.001

0.385 0.004 0.028 0.683 >0.999 0.496 0.322 >0.999 >0.999 0.885 0.151 0.214 >0.999 0.936 0.166 0.871 0.766 0.434 >0.999 >0.999 0.214 0.404 0.053 0.023 0.159 >0.999 >0.999 0.519 0.105 0.766 0.019 0.359 0.210 <0.001

0.793 0.030 0.039 0.568 0.266 0.683 0.668 >0.999 >0.999 0.488 0.097 >0.999 0.373 0.279 0.243 0.453 0.761 0.825 0.741 >0.999 >0.999 0.434 0.964 0.229 0.130 0.591 >0.999 0.537 0.434 0.078 0.647 <0.001 0.345 0.95

Values are n (%) or median (interquartile range). ASA ¼ American Society of Anesthesiologists; COPD ¼ chronic obstructive pulmonary disease; preoperative; VTE ¼ venous thromboembolism.

MI ¼ myocardial infarction;

Post-D ¼ postdischarge;

Pre-D ¼ predischarge;

Preop ¼

Ann Thorac Surg 2015;100:932–8

Percentage Race, white Age, years  75 years ASA  3 Diabetes mellitus 10% weight loss in 6 months Smoker Alcohol consumption Lack of functional independence Body mass index, kg/m2 History of COPD Dyspnea Cardiac history Hypertension Preop serum albumin < 4 Preop hematocrit < 39 Preop chemotherapy Preop radiation therapy Wound infection, superficial Wound infection, deep Organ space infection Dehiscence Pneumonia Failure to wean Reintubation Cardiac arrest Postoperative MI Sepsis Septic shock Transfusion Return to operating room Length of stay 30-Day mortality 30-Day morbidity (except VTE)

No VTE (n ¼ 3,045)

MARTIN ET AL IDENTIFYING RISK FOR VTE AFTER ESOPHAGECTOMY

Factors Associated With VTE

934

Table 1. Factors Associated With Occurrence of Predischarge and Postdischarge Venous Thromboembolism

Ann Thorac Surg 2015;100:932–8

MARTIN ET AL IDENTIFYING RISK FOR VTE AFTER ESOPHAGECTOMY

935

Comment

Fig 1. Seventeen percent (17%) of postesophagectomy venous thrombolism (VTE) events occurred after discharge. Blue portion of bars indicates early; orange indicates late. (DVT ¼ deep venous thrombosis; PE ¼ pulmonary embolism.)

cardiac arrest, sepsis, shock, transfusion, and return to the operating room, were associated with predischarge VTE. Failure to wean and return to the operating room were associated with postdischarge VTE.

Multivariate Analysis Multivariate analysis identified being male (odds ratio [OR] 2.09, p ¼ 0.018), white race (OR 1.93, p ¼ 0.004),

This study was designed to measure the incidence and timing of postesophagectomy VTE and to identify specific risk factors associated with predischarge versus postdischarge events. As such, this is the first study to examine the distinct risk factors for predischarge versus postdischarge VTE in esophageal cancer patients treated with esophagectomy. With its standardized reporting measures and sheer scale, the ACS-NSQIP database was well suited for this task. Among 3,208 patients in this study, the incidence of VTE was 5%, with 17% of these events occurring after discharge. Although certain risk factors such as sex, race, and older age are not modifiable, there were specific risk factors, most notably ventilator dependence and other major complications, that are potential targets for quality improvement to prevent VTE. Furthermore, older patients and patients who had major complications continued to be at risk for VTE even after leaving the hospital. In addition to the prothrombotic state associated with cancer itself [2, 3], the extent of major surgery is a known risk factor for the development of VTE [5, 17]. From previous NSQIP data, rates of postoperative VTE ranged from 0.28% for breast surgery to 7.3% for esophagectomy, which was the highest rate among major operations in that study [18]. The American College of Chest Physicians (ACCP) strongly recommends that patients undergoing major abdominal surgery for gastrointestinal cancers receive VTE chemoprophylaxis [16]. The specific type and duration depends on the balance of bleeding risk and VTE risk (moderate risk defined as approximately 3%, versus high, approximately 6%). The benefit of extendedduration VTE chemoprophylaxis has been well documented [1, 19, 20]. The issue, therefore, is to determine which patients truly benefit from extended-duration Table 2. Multivariable Factors Associated With Development of Predischarge Versus Postdischarge Venous Thromboembolism Factors Associated With VTE Predischarge VTE Male White race Failure to wean Other postoperative morbidity Postdischarge VTE Age, per year Postoperative morbidity

Fig 2. (A) Histogram showing distribution of discharge day compared with (B) distribution of occurrence of venous thrombolism (VTE).

CI ¼ confidence interval; thromboembolism.

OR

2.086 1.927 3.235 1.899

95% CI

p Value

(1.136–3.83) 0.0177 (1.237–3.001) 0.0037 (2.067–5.062) <0.0001 (1.219–2.958) 0.0046

1.059 (1.016–1.104) 3.142 (1.443–6.837)

OR ¼ odds ratio;

0.0062 0.0039

VTE ¼ venous

GENERAL THORACIC

prolonged ventilation (OR 3.24, p < 0.001), and other major complications (OR 1.90, p ¼ 0.005) as being independently associated with the development of predischarge VTE. Older age (OR 1.06 per year, p ¼ 0.006) and major complications (OR 3.14, p ¼ 0.004) were independently associated with postdischarge VTE (Table 2).

GENERAL THORACIC

936

MARTIN ET AL IDENTIFYING RISK FOR VTE AFTER ESOPHAGECTOMY

chemoprophylaxis, because there are cost and patient comfort issues with postdischarge treatment. Multimodality therapy, which is the standard of care, may also increase the VTE risk for esophageal cancer patients being properly treated within a multidisciplinary treatment plan [21]. One recent study demonstrated that modern trimodality (chemoradiation plus surgery) esophageal therapy is associated with a measurable prothrombotic state, persistent for as long as 6 months after surgery [22]. In that study, neoadjuvant therapy was associated with activation of factor VIII, proinflammatory nuclear factor-kb, prothrombin, and the thrombinantithrombin complex. Likewise, surgery was associated with dysregulation of the coagulation pathways for as long as 6 months. Although the surgical literature overwhelmingly supports VTE chemoprophylaxis for inpatients, there is significant variation in opinion and practice regarding specific therapy and whether to continue after discharge. Guidelines and recommendations do exist for extendedduration chemoprophylaxis, but these are inconsistently followed nationally and internationally [7, 13, 15, 16]. For example, the 32-country observational Epidemiological International Day for the Evaluation of Patients at Risk for VTE in the Acute Hospital Care Setting (ENDORSE) study found that only 58.5% of surgical patients received ACCP-recommended VTE prophylaxis [23]. From these data and other studies, multiple groups have recommended that hospitals institute formal strategies to ensure better VTE chemoprophylaxis compliance [24]. Despite decades of data advocating VTE chemoprophylaxis [25], with esophagectomy and many high-risk gastrointestinal operations, there remains a plethora of traditional beliefs regarding surgeon’s choices of timing, dosing, and duration of chemoprophylaxis, but a dearth of objective data on patient risk factors to convince surgeons [17]. This is where our study sought to provide data-driven proof that esophagectomy patients were at high risk for VTE (5.1% in 30 days) and that certain demographic factors (male, white race, older age) plus prolonged ventilator dependence or other major complications significantly increased the likelihood of a postoperative VTE beyond even the baseline risk. The first aim of this study was to delineate the typical timing of predischarge versus postdischarge VTE after esophagectomy and to define the exact rates within the current national practice patterns. It is interesting to note the relationship between day of discharge and the occurrence of VTE. The median day of discharge was 11, and median day for occurrence of predischarge VTE was 9 (Fig 2). However, the median day of occurrence of postdischarge VTE was 19, and the histogram of occurrence demonstrates that VTE can occur at any point within the 30-day NSQIP data collection window (and likely beyond at a decreasing rate). That illustrates the point that a small but relevant proportion of postesophagectomy patients remain at risk of VTE even after safely walking out of the hospital. The second aim of the study was to identify the clinical risk factors that surgeons could use to identify patients

Ann Thorac Surg 2015;100:932–8

who are at risk for predischarge versus postdischarge VTE. Based on these risk factors, patients could be prescribed limited-duration (inhospital) versus extendedduration (postdischarge) VTE chemoprophylaxis. Current recommendations from the ACCP [16], and mirrored by the NCCN [15], recommend that high-risk surgical patients (VTE risk approximately 6%) without concerns for postoperative bleeding should be prescribed 28 days of chemoprophylaxis after surgery. Esophagectomy patients in this national cohort fit between moderate risk (approximately 3% risk) and high risk. Because postoperative hemorrhage is not among the usual complications from esophagectomy and because hemorrhage is rare with prophylactic doses of chemoprophylaxis [26], it would seem logical that all esophagectomy patients should receive chemoprophylaxis from the time of surgery as the overall risk of VTE is 5.1% for the first 30 days. Although it is important to define the predictors of predischarge VTE, these factors should not affect practice patterns because all esophagectomy patients should receive inpatient VTE chemoprophylaxis according to guidelines. There is a provision in the ACCP guidelines to consider the financial burden of extended chemoprophylaxis for all patients, as outlined in a clause that some patients may be managed with limited-duration therapy. However, the postdischarge VTE risk factors that this study has identified (older age and major complications) are very important in the debate about who might selectively benefit from extended-duration chemoprophylaxis. For clinicians who remain skeptical about the need for standardized 28-day chemoprophylaxis for all esophagectomy patients, they should at least selectively prescribe it. Based on this study’s findings, patients who had postoperative major complications and patients who are elderly should be considered for extended-duration chemoprophylaxis. To contain costs, it may be argued that patients with routine recovery and without the risk factors in the multivariate model might be managed with inhospital limited-duration chemoprophylaxis. It would also be reasonable, however, to prescribe it for 28 postoperative days, starting at the time of surgery, considering the practice among other high-risk surgical patients with gastrointestinal cancers. This study has several potential limitations. It is a retrospective study of a national database and is limited by the inherent selection bias in such an analysis. Although the NSQIP data are robust, the data collection is limited to 30 postoperative days. Cancer surgery outcomes studies are moving toward 90-day measures. A major limitation of the current data is that NSQIP lacks data regarding the specific therapy, dosage, and duration of chemoprophylaxis, and if it was even used. That must be borne in mind when interpreting the results of the multivariate analysis and limits one’s ability to conclude that the risk factors identified were true predictors rather than statistically significant associations. In addition, there are currently no specific codes for minimally invasive esophagectomies. The unlisted esophageal procedure code (43499) is inconsistently used as an add-on with the corresponding open esophagectomy code. Therefore,

Ann Thorac Surg 2015;100:932–8

Disclaimer for participant use file research: The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) and the hospitals participating in the ACS-NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or for the conclusions derived by the authors.

References 1. Bergqvist D, Agnelli G, Cohen AT, et al. Duration of prophylaxis against venous thromboembolism with enoxaparin after surgery for cancer. N Engl J Med 2002;346:975–80. 2. Blom JW, Doggen CJ, Osanto S, Rosendaal FR. Malignancies, prothrombotic mutations, and the risk of venous thrombosis. JAMA 2005;293:715–22. 3. Blom JW, Vanderschoot JP, Oostindier MJ, Osanto S, van der Meer FJ, Rosendaal FR. Incidence of venous thrombosis in a large cohort of 66,329 cancer patients: results of a record linkage study. J Thromb Haemost 2006;4:529–35. 4. Merkow RP, Bilimoria KY, McCarter MD, et al. Postdischarge venous thromboembolism after cancer surgery: extending the case for extended prophylaxis. Ann Surg 2011;254:131–7. 5. Mukherjee D, Lidor AO, Chu KM, Gearhart SL, Haut ER, Chang DC. Postoperative venous thromboembolism rates vary significantly after different types of major abdominal operations. J Gastrointest Surg 2008;12:2015–22. 6. Tzeng CW, Curley SA, Vauthey JN, Aloia TA. Distinct predictors of pre- versus post-discharge venous thromboembolism after hepatectomy: analysis of 7621 NSQIP patients. HPB (Oxford) 2013;15:773–80. 7. Findlay JM, Gillies RS, Millo J, Sgromo B, Marshall RE, Maynard ND. Enhanced recovery for esophagectomy: a systematic review and evidence-based guidelines. Ann Surg 2014;259:413–31. 8. Spencer FA, Lessard D, Emery C, Reed G, Goldberg RJ. Venous thromboembolism in the outpatient setting. Arch Intern Med 2007;167:1471–5. 9. Agnelli G, Bolis G, Capussotti L, et al. A clinical outcome-based prospective study on venous thromboembolism after cancer surgery: the @RISTOS project. Ann Surg 2006;243:89–95.

937

10. Farge D, Debourdeau P, Beckers M, et al. International clinical practice guidelines for the treatment and prophylaxis of venous thromboembolism in patients with cancer. J Thromb Haemost 2013;11:56–70. 11. van Es J, den Exter PL, Kaptein AA, et al. Quality of life after pulmonary embolism as assessed with SF-36 and PEMBQOL. Thromb Res 2013;132:500–5. 12. Merkow RP, Bilimoria KY, Tomlinson JS, et al. Postoperative complications reduce adjuvant chemotherapy use in resectable pancreatic cancer. Ann Surg 2014;260:372–7. 13. Trinh VQ, Karakiewicz PI, Sammon J, et al. Venous thromboembolism after major cancer surgery: temporal trends and patterns of care. JAMA Surg 2014;149:43–9. 14. Tzeng CW, Katz MH, Lee JE, et al. Predicting the risks of venous thromboembolism versus post-pancreatectomy haemorrhage: analysis of 13,771 NSQIP patients. HPB (Oxford) 2014;16:373–83. 15. NCCN clinical practice guidelines in oncology. Venous thromboembolic disease, version 2, 2014. Available at: http:// www.nccn.org. Accessed January 12, 2015. 16. Gould MK, Garcia DA, Wren SM, et al. Prevention of VTE in nonorthopedic surgical patients: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest 2012;141(Suppl):e227–77. 17. Tzeng CW, Katz MH, Fleming JB, et al. Risk of venous thromboembolism outweighs post-hepatectomy bleeding complications: analysis of 5651 National Surgical Quality Improvement Program patients. HPB (Oxford) 2012;14: 506–13. 18. De Martino RR, Goodney PP, Spangler EL, et al. Variation in thromboembolic complications among patients undergoing commonly performed cancer operations. J Vasc Surg 2012;55: 1035–40.e1034. 19. Kakkar VV, Balibrea JL, Martinez-Gonzalez J, Prandoni P. Extended prophylaxis with bemiparin for the prevention of venous thromboembolism after abdominal or pelvic surgery for cancer: the CANBESURE randomized study. J Thromb Haemost 2010;8:1223–9. 20. Rasmussen MS, Jorgensen LN, Wille-Jorgensen P, et al. Prolonged prophylaxis with dalteparin to prevent late thromboembolic complications in patients undergoing major abdominal surgery: a multicenter randomized open-label study. J Thromb Haemost 2006;4:2384–90. 21. Rollins KE, Peters CJ, Safranek PM, Ford H, Baglin TP, Hardwick RH. Venous thromboembolism in oesophagogastric carcinoma: incidence of symptomatic and asymptomatic events following chemotherapy and surgery. Eur J Surg Oncol 2011;37:1072–7. 22. Byrne M, Reynolds JV, O’Donnell JS, et al. Long-term activation of the pro-coagulant response after neoadjuvant chemoradiation and major cancer surgery. Br J Cancer 2010;102: 73–9. 23. Cohen AT, Tapson VF, Bergmann JF, et al. Venous thromboembolism risk and prophylaxis in the acute hospital care setting (ENDORSE study): a multinational cross-sectional study. Lancet 2008;371:387–94. 24. Tooher R, Middleton P, Pham C, et al. A systematic review of strategies to improve prophylaxis for venous thromboembolism in hospitals. Ann Surg 2005;241:397–415. 25. Collins R, Scrimgeour A, Yusuf S, Peto R. Reduction in fatal pulmonary embolism and venous thrombosis by perioperative administration of subcutaneous heparin. Overview of results of randomized trials in general, orthopedic, and urologic surgery. N Engl J Med 1988;318: 1162–73. 26. Leonardi MJ, McGory ML, Ko CY. A systematic review of deep venous thrombosis prophylaxis in cancer patients: implications for improving quality. Ann Surg Oncol 2007;14: 929–36.

GENERAL THORACIC

it is not possible to establish the trend of minimally invasive surgery in this NSQIP cohort and the potential impact on the incidence of VTE. Finally, the reported event rates may not represent the true natural history of postesophagectomy VTE, because of the mixed practice pattern of chemoprophylaxis in the United States. However, the scale of this cohort allowed detailed analysis of what could not be studied with a typical single-institution retrospective cohort. As this is the first study to identify esophagectomy patients who are at high risk for VTE and might benefit from 4 weeks of chemoprophylaxis, further prospective studies, including cost analyses, may be warranted before uniformly recommending extended chemoprophylaxis after esophagectomy. In conclusion, postesophagectomy VTE occurs in a clinically significant proportion of esophageal cancer patients with identifiable risk factors for predischarge and postdischarge events. Elderly patients and patients with major complications are potentially the most likely to benefit from extended-duration chemoprophylaxis after esophagectomy.

MARTIN ET AL IDENTIFYING RISK FOR VTE AFTER ESOPHAGECTOMY

938

MARTIN ET AL IDENTIFYING RISK FOR VTE AFTER ESOPHAGECTOMY

Ann Thorac Surg 2015;100:932–8

GENERAL THORACIC

DISCUSSION DR THOMAS J. WATSON (Rochester, NY): Great talk. This is a very important topic. It is sobering to see how many late venous thromboembolisms (VTE) there are. These are scary data of which we must be aware. At our institution over the past 10 to 15 years, our strategy for treating esophageal cancer has evolved. We used to go straight to esophagectomy for most patients, then we went to perioperative chemotherapy as per the MAGIC (Medical Research Council Adjuvant Gastric Infusional Chemotherapy) trial, and more recently we have gone to preoperative chemoradiation per the CROSS trial (Neoadjuvant chemoradiation followed by surgery versus surgery alone for patients with adenocarcinoma or squamous cell carcinoma of the esophagus). It has been our observation that once we started adding chemotherapy to the regimen, the incidence of VTE went way up; we were seeing a lot of thromboembolic complications in this group of cancer patients getting chemotherapy. I was wondering whether you were able to stratify your data based on those getting perioperative chemotherapy and those not. Was chemotherapy a risk factor for late VTE? DR MARTIN: That is a great question. Unfortunately, the National Surgical Quality Improvement Program really does not have data about chemotherapy and radiation therapy. Earlier parts of the dataset started to look at that, but it is no longer a mandatory variable, so we do not have that. It may be something that we can unearth from The Society of Thoracic Surgeons database. I do know as I was reviewing the literature for this that there was a study done in Ireland of 36 patients prospectively where they measured proinflammatory and prothrombotic markers. That was where it was found that, purely the induction therapy saw a rise in these markers, and surgery, a further rise, and this effect is persistent for 6 months. All of these things do combine into increasing our risk, and, as we see, this is some of the highest risk in the database. DR KALPAJ R. PAREKH (Iowa City, IA): I enjoyed your presentation. One question that I had, going one step further than Dr Watson, is that we are seeing a lot of pulmonary embolisms (PE) in our preoperative, when the patients come back for their restaging computed tomography (CT) scan after completion of

neoadjuvant therapy. Unfortunately, not all payers are paying for a CT scan with intravenous contrast after completion of neoadjuvant therapy. We pick up a fair incidence of PE, and I wonder if some of these PE are happening even preoperatively, before the patient even goes for surgery, if they don’t get a CT scan with contrast. DR MARTIN: Again, I think that is a great question, one that we cannot answer from this dataset, but the National Comprehensive Cancer Network does have guidelines for a lot of these highrisk patients who should be getting extended prophylaxis even outside of our surgical population. Again, that is a great question. Thank you. DR KATIE S. NASON (Pittsburgh, PA): One of your conclusions from these data is that elderly patients and patients with complicated postoperative courses should be treated with chemoprophylaxis after discharge to impact on a VTE rate that is less than 1%. In an elderly patient, more likely to fall, recovering from an esophagectomy, how do you reconcile the risks of the chemoprophylaxis against what is really a very low risk of VTE? Thirty-four percent of patients have a PE, which is really the complication that you would most dread. How do you reconcile that? What would you do with your 85-year-old patient going home? DR MARTIN: The absolute risk rates are low, and therefore the numbers needed to treat are high. However, the consequence of a VTE occurrence, which is potentially preventable, is significant. That is why the American College of Chest Physicians guidelines for the high-risk cancer patients stops just short of making this a mandatory recommendation for all patients. Looking at the general surgery patients where the risks are even lower, there is an inclusion that it should be strongly considered for those patients. More important than age, in my opinion, is the complicated postoperative course. If I have a patient who has a bump in the road, stays on the ventilator, has pneumonia or otherwise, now those are the patients that I am going to be thinking about sending home on extended prophylaxis. This is something that we certainly need to pay more attention to as a group and consider generating guidelines.