DUS indicates duplex ultrasonography; IPC, intermittent pneumatic compression; LMWH, low-molecular-weight heparin sodium; O&C, open and close; Op, operative; and PROTECTOR, Optimal Prophylactic Method for Venous Thromboembolism After Gastrectomy in Korean Patients.
DUS indicates duplex ultrasonography; IPC, intermittent pneumatic compression; LMWH, low-molecular-weight heparin sodium; Op, operative; and POD, postoperative day.
eTable 1. Baseline Thrombophilic Factors of the Patients
eTable 2. Incidence of Symptomatic and Asymptomatic VTE in Patients After Gastrectomy (Intent-to-Treat Analysis)
eTable 3. Clinical Features of Patients with DVT After Gastrectomy
eTable 4. Relative Risks of VTE Among Patients After Gastrectomy
eTable 5. Incidence of Postoperative Bleeding Complications After Gastrectomy (Intent-to-Treat Analysis)
eTable 6. Risk Factors Associated with Bleeding in the IPC Plus Enoxaparin Group
eTable 7. Characteristics and Types of the Bleeding Cases
Customize your JAMA Network experience by selecting one or more topics from the list below.
Jung YJ, Seo HS, Park CH, et al. Venous Thromboembolism Incidence and Prophylaxis Use After Gastrectomy Among Korean Patients With Gastric Adenocarcinoma: The PROTECTOR Randomized Clinical Trial. JAMA Surg. 2018;153(10):939–946. doi:10.1001/jamasurg.2018.2081
Is it safe and effective to routinely use the conventional prophylaxis for venous thromboembolism in Asian patients after gastric cancer surgery?
This randomized clinical trial of 666 Korean patients with gastric adenocarcinoma found that, overall, the incidence of venous thromboembolism was 2.1% and the incidence of bleeding complications was significantly higher among patients assigned to receive intermittent pneumatic compression combined with low-molecular-weight heparin sodium.
Use of intermittent pneumatic compression alone is inferior to the use of intermittent pneumatic compression combined with low-molecular-weight heparin in preventing postoperative venous thromboembolism; because pharmacological prophylaxis is associated with bleeding, further study is needed to stratify the patients at high risk for developing venous thromboembolism.
The guidelines by the National Comprehensive Cancer Network and the American Society for Clinical Oncology recommend the routine use of thromboprophylaxis for patients with gastric adenocarcinoma. However, many physicians in Asian countries use venous thromboembolism (VTE) prophylaxis much less often because of the perceived lower VTE incidence in this population.
To evaluate the incidence of postgastrectomy VTE in Korean patients with gastric adenocarcinoma, and to identify the complications and evaluate the efficacy and safety of VTE prevention methods.
Design, Setting, and Participants
The Optimal Prophylactic Method for Venous Thromboembolism After Gastrectomy in Korean Patients (PROTECTOR) randomized clinical trial was conducted between August 1, 2011, and March 31, 2015. Patients with histologically confirmed gastric adenocarcinoma presenting to a single center (Seoul St Mary’s Hospital in Seoul, South Korea) were enrolled.Patients were randomized to either an intermittent pneumatic compression (IPC)–only group or an IPC+low-molecular-weight (LMW) heparin sodium group. The data were analyzed on intention-to-treat and per protocol bases. Data analysis was performed from April 1, 2016, to October 30, 2017.
Main Outcomes and Measures
Venous thromboembolism incidence was the primary outcome. Postoperative complications, particularly those associated with VTE prophylaxis methods, were the secondary end point.
Of the 682 patients enrolled and randomized, 447 (65.5%) were male and 245 (34.5%) were female, with a mean (SD) age of 57.67 (12.94) years. Among the 666 patients included in the analysis, the overall incidence of VTE was 2.1%. The incidence of VTE was statistically significantly higher in the IPC-only group compared with the IPC+LMW heparin group (3.6%; 95% CI, 2.05%-6.14% vs 0.6%; 95% CI, 0.17%-2.18%; P = .008). Among the 14 patients (2.1%) with VTE, 13 were asymptomatic and received a deep vein thrombosis diagnosis, whereas 1 patient received a symptomatic pulmonary thromboembolism diagnosis. The overall incidence of bleeding complications was 5.1%. The incidence of bleeding complications was significantly higher in the IPC+LMW heparin group compared with the IPC-only group (9.1% vs 1.2%; P < .001). No cases of VTE–associated mortality were noted.
Conclusions and Relevance
Use of IPC alone is inferior to the use of IPC+LMW heparin in preventing postoperative VTE. Because LMW heparin is associated with a high bleeding risk, further study is needed to stratify the patients at high risk for perioperative development of VTE.
ClinicalTrials.gov Identifier: NCT01448746
Venous thromboembolism (VTE) is one of the leading causes of in-hospital morbidity,1 and cancer is an independent major risk factor in developing VTE. Conversely, VTE is the second leading cause of death for patients with medically and surgically treated cancer.2 Based on these facts, the guidelines from the National Comprehensive Cancer Network and the American Society for Clinical Oncology recommend pharmacological thromboprophylaxis, unless contraindicated, for patients who have undergone surgery. Although mechanical VTE prophylaxis methods can be used, they are not recommended as monotherapy for VTE prevention.3,4 The ninth edition of the American College of Chest Physicians Evidence-based Clinical Practice Guidelines recommends, after abdominal or pelvic surgery to treat cancer, 4-week pharmacological prophylaxis for patients who are at high risk for VTE but not otherwise at high risk for major bleeding complications.5
Despite this favorable evidence, the application of these guideline recommendations to Asian patients is uncommon. Studies conducted in Asian countries have reported much lower incidence of VTE compared with rates found in US or European studies.6-10 Furthermore, low-molecular-weight (LMW) heparin sodium is associated with bleeding complications in Asian patients.11,12 Therefore, many surgeons in Asian countries do not use VTE prophylaxis. However, to date no adequate, well-designed study involving an Asian population has been performed.
Our hypothesis was that mechanical VTE prophylaxis methods would not be inferior to LMW heparin in combination with intermittent pneumatic compression (IPC). We designed this study to determine the incidence of VTE among Korean patients with gastric adenocarcinoma and the efficacy and safety of LMW heparin and IPC interventions.
From August 1, 2011, to March 3, 2015, patients with histologically confirmed gastric adenocarcinoma (n = 682) were enrolled in the Optimal Prophylactic Method for Venous Thromboembolism After Gastrectomy in Korean Patients (PROTECTOR) randomized clinical trial conducted at Seoul St Mary’s Hospital in Seoul, South Korea (ClinicalTrials.gov Identifier: NCT01448746). Patients were excluded for the following reasons: history of deep vein thrombosis (DVT) or pulmonary thromboembolism, preoperative prolonged immobilization, disease with a high bleeding risk, allergy to heparin or history of heparin-induced thrombocytopenia, varicose veins or chronic venous insufficiency, previous chemotherapy or radiotherapy, current anticoagulation therapy, body mass index (calculated as weight in kilograms divided by height in meters squared) of 18.5 or lower, and pregnancy or an intention to become pregnant. The PROTECTOR trial was approved by the institutional review board of Seoul St Mary’s Hospital. Written informed consent was obtained from all patients prior to gastrectomy. Data analysis was performed from April 1, 2016, to October 30, 2017. See Supplement 1 for the trial protocol.
Enrolled patients were assigned either to the IPC-only group, using a compression system (Kendall SCD Express Sequential Compression System; Medtronic) or the IPC+LMW heparin group, using LMW heparin (Clexane; Sanofi-Aventis Korea) (Figure 1). Patients and investigators were not blinded to the treatment allocation.
For all patients, IPC use was initiated preoperatively and continued until discharge from the hospital. Patients were monitored by their medical team for compliance and encouraged to use IPCs at all times unless ambulating. Low-molecular-weight heparin was administered after surgery at 24-hour intervals at a daily dose of 40 mg beginning on the day of surgery. The drug was administered via deep subcutaneous injection delivered alternately to the left and right arms. Abdominal wall injection was not preferred because of the proximity to the drain site and associated incisional pain. All patients underwent duplex ultrasonography (DUS) routinely on postoperative day 4. All examinations were performed using the 2-point compression method, from the proximal venous system of the common and superficial femoral veins to below the popliteal vein. Patients with VTE were treated with anticoagulants (Xarelto; Janssen Pharmaceuticals).
The primary objective was to determine the total number of VTEs in each group within 30 days of surgery. The secondary objectives were to identify postoperative complications and evaluate the safety of VTE prevention methods. Major and minor bleeding after surgery were defined according to the definition of the Subcommittee on Control of Anticoagulation of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis.13 Thus, major bleeding for patients who underwent surgery was defined as fatal bleeding and/or extrasurgical-site bleeding causing a decrease in hemoglobin level of 2 g/dL or greater (to convert to grams per liter, multiply by 10.0), bleeding requiring transfusion of 2 or more units of whole blood or red blood cells, or surgical-site bleeding requiring a second intervention.
A primary noninferiority analysis, based on a 2-sided 95% CI, was computed on the basis of the difference between the 2 groups with a power of 80%. To prove noninferiority, the upper limit of the 95% CI was set within a preestablished noninferiority margin of 2%, and a baseline event rate of 1% was assumed for the standard treatment group. We determined that 341 patients were required per group, assuming a 10% dropout rate; therefore, a total of 682 patients were enrolled. The data were analyzed on intention-to-treat and per protocol bases. Because per protocol analysis is more conservative in noninferiority tests, if the results were conflicting, we emphasized the result of the per protocol analysis over that of the intention-to-treat analysis.14 The Wald method (without continuity correction) was used to compute the 95% CI. Pearson χ2 test was used to assess the differences in categorical variables, and Wilcoxon rank sum test was used for the continuous variables. All P values were 2-sided, and P < .05 was indicative of statistical significance. All statistical analyses were performed using SAS software, version 9.4 (SAS Institute, Inc).
Of the 682 patients enrolled and randomized, 447 (65.5%) were male and 245 (34.5%) were female, with a mean (SD) age of 57.67 (12.94) years. Sixteen patients (2.3%) were excluded from the per protocol analysis: 1 experienced heparin-induced thrombocytopenia, 8 failed curative surgery, 3 had concurrent cancer diagnoses, and 4 withdrew from the study (Figure 2). Therefore, a total of 666 patients (97.7%) were evaluated in the per protocol analysis (336 in the IPC-only group and 330 in the IPC+LMW heparin group); the groups were well balanced after randomization, except in the surgical approach used (Table 1). In the IPC+LMW heparin group alone, more patients underwent laparoscopic surgery than open gastrectomy (219 [66.7%] vs 11 [33.6%]; P = .01). None of the preoperative laboratory findings or baseline thrombophilic factors showed significant differences between the 2 groups (eTable 1 in Supplement 2).
In total, 14 patients were determined to have VTEs, resulting in an overall incidence of 2.1% (per protocol analysis [Table 2] and intention-to-treat analysis [eTable 2 in Supplement 2]). Of those 14 patients, 12 (3.6%) were from the IPC-only group and 2 (0.6%) were from the IPC+LMW heparin group. The incidence of VTE was statistically significantly higher in the IPC-only group compared with the IPC+LMW heparin group (3.6%; 95% CI, 2.05%-6.14% vs 0.6%; 95% CI, 0.17%-2.18%; P = .008). The VTE incidence rate was 3.5% (intention-to-treat analysis) and 3.6% (per protocol analysis) in the IPC-only group, and the rate was 0.6% (intention-to-treat analysis) and 0.6% (per protocol analysis) in the IPC+LMW heparin group. The upper limit of the 97.5% (2-tailed 95% CI) on the VTE difference for the intention-to-treat population was 5.1%, a value exceeding the preestablished noninferiority margin of 2%. Similar results were obtained in the per protocol population, in which the upper limit was 5.1% (above the 2% noninferiority margin), suggesting that the VTE incidence associated with IPC-only intervention is inferior to the incidence associated with IPC+LMW heparin treatment.
Among the 14 patients with VTE, 13 (1.9%) were asymptomatic and received a diagnosis of DVT when screening DUS was performed. Twelve of the 13 patients had single-focus DVTs—isolated calf-vein thromboses—and 1 patient had multiple lesions in the soleal and peroneal veins (eTable 3 in Supplement 2). None of the 13 patients had symptoms of DVT, such as pain or edema in the extremities. Four of the 13 patients were treated with anticoagulation therapy for the first 3 months following DVT diagnosis, and further treatment was determined by follow-up sonography, according to recommendations in the 2014 American College of Chest Physicians Evidence-based Clinical Practice Guidelines. One patient (0.3%) of the 332 individuals evaluated with DUS in the IPC-only group demonstrated dyspnea; this patient subsequently underwent a computed tomographic scan and received a pulmonary thromboembolism diagnosis.
Substantial risk factors for developing VTE were identified in the patient population of the PROTECTOR trial, including age older than 65 years (relative rate, 3.0; 95% CI, 1.06-8.76; P = .04) and stage III disease on final pathology (relative rate, 3.52; 95% CI, 1.06-11.69; P < .04) (eTable 4 in Supplement 2). Because the surgical approach (open surgery or laparoscopy) was not stratified, we analyzed VTE incidence by the surgical approach. The VTE frequency of patients who underwent open gastrectomy was 1.7% (7 of 411), and the frequency of patients who underwent laparoscopic surgery was 2.7% (7 of 255), which were not substantially different between the 2 groups (P = .41).
The incidence of bleeding complications was significantly higher in the IPC+LMW heparin group compared with the IPC-only group (9.1% vs 1.2%; P < .001). Both major and minor bleeding events occurred more frequently in the IPC+LMW heparin group compared with the IPC-only group (per protocol analysis [Table 3] and intention-to-treat analysis [eTable 5 in Supplement 2]). Two patients from the IPC+LMW heparin group experienced minor bleeding. Of these 2 patients, 1 had incisional bleeding without any change in hemoglobin level, and the other showed sanguineous drainage fluid color and was treated conservatively with discontinued LMW heparin. To minimize bias in our reporting of bleeding complications, we performed a separate analysis for the 330 patients in the IPC+LMW heparin group which revealed that male sex, history of smoking, open surgical approach, and extended resection were associated with bleeding complications (eTable 6 in Supplement 2).
We analyzed each case of bleeding to determine whether the episode was associated with IPC+LMW heparin use or was merely a technical error (eTable 7 in Supplement 2). Among the 34 patients with bleeding complications, 5 patients from the IPC+LMW heparin group had bleeding vessels that were definitively controlled by reoperation. The remaining 29 patients showed sanguineous drainage, which was treated by stopping use of LMW heparin and applying conservative care. After excluding the 5 cases resolved by reoperation, the bleeding complication rate was significantly higher in the IPC+LMW heparin group compared with the IPC-only group (25 vs 4); therefore, we conclude that IPC+LMW heparin is statistically significantly associated with bleeding in these patients.
We recorded no notable adverse effect (such as bleeding or infection) associated with LMW heparin injection. No mortality was attributable to VTE or any surgical complication within 30 days after surgery.
In Western countries, routine prophylaxis for VTE is considered necessary to minimize the risk of sudden death associated with pulmonary thromboembolism. Many clinicians in Asian countries, however, infrequently use routine prophylaxis because they believe VTE is rare among their patients. The reasons for these geographic or ethnicity-based differences in VTE incidence are unclear, although some researchers have implicated increased hypercoagulability resulting from the differences in specific gene-associated factors.15,16 The incidence of VTE in the Korean population is lower than that in white populations, but several studies in other Asian countries have reported rapid increases in incidence.17-19 Most treatment guidelines are based on findings by research performed predominantly in the West; thus, the actual incidence of VTE after abdominal surgery in Asia remains unclear. Therefore, we conducted the PROTECTOR trial to determine the incidence of VTE among Asian patients with gastric adenocarcinoma who underwent radical surgery.
The PROTECTOR trial demonstrated that VTE developed in 14 of 666 patients with gastric adenocarcinoma—an overall incidence of 2.1%. This incidence rate is lower than the rates reported in other studies of advanced gastroesophageal cancer during medical treatment, some of which were higher than 13%.20-24 Because the incidence of symptomatic DVT has historically been low at our institution, to determine the incidence of both symptomatic and asymptomatic DVT, we performed DUS on postoperative day 4 in all patients irrespective of symptoms. The incidence of DVT was statistically significantly higher in the IPC-only group. However, none of the patients with DVT had any symptoms or signs, and in the IPC-only group, only 1 patient had symptomatic pulmonary thromboembolism. Accordingly, the incidence of symptomatic VTE was only 0.2%, without any mortality. The actual number of VTEs was substantially different, but all DVTs were detected only on DUS, which means that DVT might not have been diagnosed unless DUS had been performed.
Several explanations may be advanced for the low VTE incidence rate in this study compared with those reported among Western populations. First, the cancer stages may differ. Earlier studies in Western settings that reported higher DVT incidences included patients with more advanced cancer. In this trial, more than 70% of patients had a diagnosis of stage I tumors. Minimally invasive surgery is frequently performed to treat early cancers. Minimally invasive surgery is associated with shorter operative times, faster recovery times, and earlier ambulation, all of which may lower the risk of VTE. However, whether laparoscopic surgery is associated with a low incidence of VTE remains a controversial question. Some investigators have considered that, during surgery featuring a pneumoperitoneum, venous return may be reduced, increasing the risk of VTE.25,26 Several studies have reported that VTE incidence was lower among patients who underwent laparoscopic colorectal surgery.27,28 In this trial, patients who underwent open surgery tended to experience more VTEs compared with those who underwent laparoscopic surgery, but the difference was not statistically significant (2.7% vs 1.7%; P = .41). In addition, open surgery was not a statistically significant risk factor for VTE, but it was for bleeding complications (eTables 4 and 6 in Supplement 2). Therefore, we cannot conclude that open surgery is associated with a higher incidence of VTE; further study is required.
The second reason for the observed difference in VTE incidence rate may be the dissimilarity in body mass index between Asian and Western populations. In a previous study of patients with gastric cancer treated in Korea and the United States, the body mass index of Korean patients compared with US patients was statistically significantly lower (22.6; 95% CI, 13.7-35.9 vs 25.4; 95% CI, 14.1-47.6; P = .001).29 The mean (SD) body mass index of patients in this trial was 23.7 (1.86), lower than that of Westerners. Therefore, a routine single dose of 40 mg of LMW heparin may not only efficiently prevent VTE but also increase the bleeding tendency after gastrectomy.
Reported outcomes related to the safety of heparins are strikingly different between the East and West. Several prospective randomized clinical trials and meta-analyses from the West have shown that both unfractionated heparin and LMW heparin are effective in preventing VTE but only minimally increase the incidence of major bleeding complications.30-35 However, a Korean study showed that the risk of postoperative bleeding was higher in patients who used LMW heparin.11 Our interim and final analyses had similar findings: Patients who were administered LMW heparin had a higher rate of bleeding, even after we excluded cases that might be associated with technical failure.36 Intermittent pneumatic compression enhances venous drainage, prevents venous stasis, improves blood flow velocity in the femoral vein by 50% to 250%, and enhances fibrinolytic activity without increasing the risk of bleeding.37
We hypothesized, on the basis of these findings, that mechanical prophylaxis without pharmacological methods would be sufficient to prevent VTE in Korean patients. Contrary to our expectation, however, we found that IPC-only intervention was inferior to IPC+LMW heparin treatment. However, a prophylactic dose of 40-mg LMW heparin might be excessive for Asian patients and may lead to postoperative bleeding complications. Further study into an optimal LMW heparin dose (such as 20 mg vs 40 mg) for patients with gastric adenocarcinoma who underwent gastrectomy is required, as is stratification of the risks for VTE development and bleeding complications, to identify patients requiring medical prophylaxis against VTE.
The PROTECTOR trial had several limitations. First, the timing of thrombotic events in 13 patients was unclear. We justified the routine performance of DUS on postoperative day 4 on the low incidence of symptomatic DVT, but to ensure cost-effectiveness, we did not perform DUS both preoperatively and postoperatively. Instead, we attempted to avoid including patients with preoperative DVT by specifically excluding those with a history of VTE, with vascular insufficiency, or with an immobilizing condition requiring a wheelchair. We compared the patients’ clinical characteristics to evaluate the risk factors for thrombosis according to the Caprini score (score range: 4-12, with the highest score, which is higher than 5, indicating a high risk for DVT), as well as their baseline thrombophilic factors, but the results did not substantially differ according to the occurrence of VTE (eTable 4 in Supplement 2); thus, the possibility of a preexisting thrombus cannot be completely ruled out.
Second, the American College of Chest Physicians Evidence-based Clinical Practice Guidelines for the treatment of patients with isolated calf-vein thromboses were changed in 2012; therefore, there was treatment heterogeneity among patients enrolled in the trial before and after the guidelines change.31 Third, the surgical approach (open surgery vs laparoscopy) was not stratified preoperatively because more than 70% of patients had early-stage cancer and were treated laparoscopically. Hence, we cannot discuss any possible association between the surgical approach and VTE development. A multicenter trial that includes more patients with advanced cancer who underwent open surgery is required. Fourth, we did not blind the clinicians to the treatment group allocations. However, the same perioperative clinical pathway, including diet, laboratory study, and fluid therapy, was used for all patients with gastric cancer. Decisions to perform interventions, such as transfusions, reoperations, or endoscopic procedures, were based on vital signs or a change in blood test results. Analysis of bleeding cases revealed that among 34 patients, 31 had major bleeding (according to our definition) and 3 had minor bleeding (according to the judgment of a clinician). Therefore, the likelihood of treatment bias resulting from lack of blinding is minimal.
Prophylaxis using IPC+LMW heparin was more effective than IPC alone in minimizing the incidence of VTE among Korean patients who had undergone gastrectomy. However, given the extremely low incidence of symptomatic VTE and the relatively higher incidence of bleeding complications among these patients, routine administration of IPC+LMW heparin for VTE prophylaxis is not recommended. To our knowledge, the PROTECTOR trial is the first prospective randomized clinical trial to determine the optimal VTE prophylaxis method in Korea and to investigate the incidence of VTE among patients with gastric adenocarcinoma. Further study is needed to identify patients who may gain the maximum benefit from pharmacological thromboprophylaxis.
Accepted for Publication: April 16, 2018.
Corresponding Author: Kyo Young Song, MD, PhD, Division of Gastrointestinal Surgery, Department of Surgery, Seoul St Mary’s Hospital, College of Medicine, The Catholic University of Korea, 222 Banpo-daero, Seocho-gu, Seoul 137-701, Republic of Korea (email@example.com).
Published Online: July 18, 2018. doi:10.1001/jamasurg.2018.2081
Author Contributions: Dr Song had full access to all of the data in this study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Jung, Park, Kim, Song.
Acquisition, analysis, or interpretation of data: Jung, Seo, Jeon, Yim, Song.
Drafting of the manuscript: Jung, Jeon, Song.
Critical revision of the manuscript for important intellectual content: Jung, Seo, Park, Kim, Yim, Song.
Statistical analysis: Jung, Seo, Yim.
Obtained funding: Song.
Administrative, technical, or material support: Jung, Kim, Song.
Study supervision: Seo, Park, Jeon, Song.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was funded by Medtronic.
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: The authors appreciate the statistics-related consultation provided by the Department of Biostatistics, Clinical Research Coordinating Center, The Catholic University of Korea. We thank the independent scientific steering committee for writing the protocol and Medtronic for providing input before protocol revision.