2017 Volume 81 Issue 12 Pages 1936-1944
Background: The prognosis of asymptomatic deep vein thrombosis (DVT) is uncertain and there is no consensus on the necessity of detection and treatment.
Methods and Results: We retrospectively evaluated 300 patients with asymptomatic lower extremity DVT screened from 4,514 consecutive patients on ultrasound at Kyoto University Hospital between January 2010 and September 2015. The subjects had concomitant active cancer in 40%, unprovoked DVT in 59%, and distal DVT in 70%. The cumulative 5-year incidences of symptomatic recurrent venous thromboembolism (VTE); major bleeding; and all-cause death were 14.5%, 16.6%, and 34.1%, respectively. Among 232 patients (77%) with prolonged anticoagulant therapy, anticoagulants were discontinued in 48.4% at 1 year. Anticoagulant therapy was associated with a significantly higher incidence of major bleeding compared with the non-anticoagulant group (20.5% vs. 1.5%, P=0.01) with no significant effect on the incidence of VTE. In patients with active cancer, the favorable effect of anticoagulants relative to no anticoagulants for VTE was significant (HR, 0.22; 95% CI: 0.05–0.95).
Conclusions: Prolonged anticoagulants therapy was implemented in the majority of patients with asymptomatic DVT, but was associated with a significantly higher risk for major bleeding. On subgroup analysis in patients with active cancer, however, there appeared to be a benefit of prolonged anticoagulant therapy in decreasing the rate of symptomatic recurrent VTE.
Anticoagulation is the established management strategy for symptomatic venous thromboembolism (VTE) and for prevention of recurrence according to the current VTE guidelines.1–4 Recently, leg vein ultrasound has become common in patients at high risk for or with suspected lower extremity deep vein thrombosis (DVT).5,6 Because of its non-invasive nature and availability, ultrasound DVT is commonly performed in daily clinical practice in various situations such as minor symptoms, high D-dimer or during the perioperative period. The frequent use of ultrasound has led to an increased number of diagnoses of incidental asymptomatic DVT. Asymptomatic DVT, however, has not been adequately studied, and even the prognosis of asymptomatic DVT is still unclear. Furthermore, there is no general agreement on whether asymptomatic DVT needs to be diagnosed and treated.2
The aim of this study was therefore to evaluate the clinical characteristics, management strategies, and long-term outcomes of asymptomatic lower extremity DVT detected on ultrasound in the daily clinical practice of a single center.
The current study was an observational single-center retrospective cohort study enrolling consecutive patients with asymptomatic lower extremity DVT detected on ultrasound of leg veins at Kyoto University Hospital between January 2010 and September 2015.
Among the 4,514 consecutive patients who underwent ultrasound of leg veins during the study period, 552 patients were diagnosed as having DVT. The subject group consisted of 300 patients with asymptomatic lower extremity DVT with or without asymptomatic pulmonary embolism (PE), after excluding 111 patients with symptomatic PE with (n=37) or without (n=74) symptoms for DVT, 135 patients with symptomatic DVT, and 6 patients with inferior vena cava thrombus (Figure 1).
Study flow chart. Patients with symptomatic pulmonary embolism (PE) included those with (n=37) or without (n=74) symptoms for deep vein thrombosis (DVT).
The study protocol was approved by the ethics committee of Kyoto University Hospital. Because of retrospective enrollment, written informed consent was waived. This strategy is concordant with the guidelines for epidemiological studies issued by the Ministry of Health, Labor and Welfare in Japan.
Data for baseline characteristics and patient management were collected from hospital charts or hospital databases according to the pre-specified definitions. According to the previous reports, the symptoms of VTE were defined as follows: pain, swelling and tenderness of legs for DVT; and dyspnea, pleuritic chest pain, substernal chest pain, cough, fever, hemoptysis, and syncope for PE.7,8 Additionally, abnormalities in vital signs such as a decrease in arterial oxygen saturation and hypotension were also regarded as symptoms of PE. The presence or absence of symptoms was evaluated at the time of the ultrasound.
Active cancer was defined as treatment for cancer such as chemotherapy or radiotherapy, scheduling for cancer surgery, metastasis to other organs, and/or terminal cancer (expected life expectancy ≤6 months) at the time of the diagnosis of DVT. Proximal DVT was defined as venous thrombosis located in the popliteal, femoral, or iliac veins; and distal DVT was defined as venous thrombosis located in the calf veins including peroneal, posterior tibial, anterior tibial, and soleus muscle veins below the knee. Provoked DVT was defined as DVT caused by antecedent and transient environmental risk factors such as surgery, trauma, immobilization, pregnancy and puerperium, and unprovoked DVT was defined as DVT in which no identifiable provoking environmental factors were evident.9 Immobilization was defined as bedridden status >4 days.10 Concomitant asymptomatic PE was evaluated on contrast-enhanced computed tomography (CT) at diagnosis of DVT, although not all patients underwent contrast-enhanced CT.
Parenteral anticoagulants early after diagnosis were defined as parenteral anticoagulant therapy (unfractionated heparin or fondaparinux) ≤10 days after diagnosis of DVT, while prolonged anticoagulant therapy was defined as anticoagulant therapy (warfarin, direct oral anticoagulant, or unfractionated heparin) continued for >10 days after diagnosis of DVT.1
Collection of follow-up information was mainly conducted through review of inpatient and outpatient hospital charts, and additional follow-up information was collected through contact with patients or relatives by phone, with questions regarding vital status, occurrence of symptomatic recurrent VTE and major bleeding, and status of anticoagulants. Median follow-up duration for the surviving patients was 623 days (IQR, 348–1,171 days).
The outcome measures were symptomatic recurrent VTE; major bleeding; and all-cause death. Symptomatic recurrent VTE was defined as symptomatic DVT and/or symptomatic PE objectively confirmed on ultrasound, contrast-enhanced CT, ventilation-perfusion lung scintigraphy, pulmonary angiography, or contrast venography. Exacerbation of the thrombus without any symptoms was not regarded as symptomatic recurrent VTE, and VTE with symptoms accompanied by new or exacerbation of the thrombus was regarded as symptomatic recurrent VTE.
Major bleeding was defined as International Society on Thrombosis and Hemostasis (ISTH) major bleeding, which consisted of reduction in the hemoglobin level by ≥2 g/dL, transfusion ≥2 units of blood or symptomatic bleeding in a critical area or organ.11
Ultrasound was performed using the Aplio 500, Aplio XG, Xario XG (Toshiba Medical Systems Corporation), or ProSound α10 (Hitachi Healthcare). High-resolution 5.0–7.5-MHz linear transducers were used in all examinations. Lower extremity deep veins were examined in the supine position by moving the probe from the inferior vena cava to the calf veins at ankle level. Evaluation was carried out using both tomography and the compression method. Distal vein examination was also performed, at the discretion of technicians, in the sitting position with the legs hanging down when observation in the supine position was insufficient. The criterion for diagnosis of proximal DVT was the presence of thrombus and/or non-compressibility of the vein in the transverse plane on color Doppler. The criterion for distal DVT was the presence of thrombus and/or non-compressibility of the vein combined with the absence of venous flow after distal compression. Ultrasound was performed by 8 experienced technicians (T.Y., K.I., H.K., H.M., A.I., M. Yamamoto, M. Yamashita, and H.F.), and supervising cardiologists further reconfirmed the technician reports.
Categorical variables are presented as numbers and percentages. Continuous variables are presented as mean±SD or as median (IQR). Categorical variables were compared using the chi-squared test when appropriate; otherwise, Fisher’s exact test was used. Continuous variables were compared using Student’s t-test or Wilcoxon rank sum test based on data distribution. Cumulative incidences were estimated using the Kaplan-Meier method and differences were assessed with log-rank test. To examine the effect of prolonged anticoagulant therapy on clinical outcome, we divided the entire cohort into 2 groups according to the use of prolonged anticoagulants therapy: a prolonged anticoagulants group and a no prolonged anticoagulant group. Furthermore, we also performed analysis in the clinically relevant subgroups including active cancer status, provoked or unprovoked DVT, proximal DVT or distal DVT only, and high or low D-dimer. D-dimer levels were dichotomized using the median. The Cox proportional hazard model was used to estimate hazard ratio (HR) and 95% CI for the effect of anticoagulants relative to no anticoagulants on symptomatic recurrent VTE during follow-up on subgroup analysis. The small number of patients with symptomatic recurrent VTE did not allow construction of an appropriate multivariable model. Statistical analysis was conducted using JMP version 10.0.2 (SAS Institute, Cary, NC, USA). All statistical analysis was 2-tailed, and P<0.05 was considered statistically significant.
Among 300 patients with asymptomatic VTE, female patients accounted for 70% (210 patients; Table 1). Mean age was 69.1±13.7 years. Body weight and body mass index were 53.7±12.8 kg and 22.0±4.5 kg/m2, respectively.
Data given as n (%), mean±SD or median (IQR). AC, anticoagulant; BMI, body mass index; DVT, deep vein thrombosis; ICU, intensive care unit; IVC, inferior vena cava; MI, myocardial infarction; PCI, percutaneous coronary intervention; PE, pulmonary embolism; PT-INR, international normalized ratio of prothrombin time; UFH, unfractionated heparin; VTE, venous thromboembolism.
Patients with cancer and those with active cancer at the time of VTE consisted of 146 patients (49%) and 120 patients (40%), respectively. Unprovoked DVT was identified in 59% (176 patients). In 67 patients (22%), VTE was diagnosed during the perioperative period after surgery. Regarding the location of DVT, only distal DVT was present in 210 patients (70%). Concomitant asymptomatic PE was present in 68 patients (23%). Parenteral anticoagulants early after diagnosis were given to 152 patients (51%), whereas prolonged anticoagulant therapy was implemented in 232 patients (77%). A total of 59 patients (20%) received neither parenteral anticoagulant therapy early after diagnosis nor prolonged anticoagulant therapy (no anticoagulant therapy after diagnosis).
Cumulative 5-year incidences of symptomatic recurrent VTE; major bleeding; and all-cause death were 14.5%, 16.6%, and 34.1%, respectively (Figure 2). Most deaths were due to cancer (71%; Table S1). Among 232 patients with prolonged anticoagulant therapy, the cumulative incidence of anticoagulant discontinuation, which was 48.4% at 1 year, tended toward a plateau beyond 1 year (Figure 3).
Kaplan-Meier curves for cumulative incidence of symptomatic recurrent VTE; major bleeding; and all-cause death. VTE, deep vein thrombosis and/or pulmonary embolism.
Kaplan-Meier curve for discontinuation of anticoagulant in the prolonged anticoagulant therapy group.
Among 17 patients with symptomatic recurrent VTE during follow-up, 2 patients developed symptomatic PE (including 1 fatal PE). Active cancer was present in 8 of 17 patients (47%) with developing symptomatic VTE during follow-up (Table 2). Anticoagulants were prescribed in 8 of 17 patients at the time of development of symptomatic VTE (7 patients with warfarin and 1 patient with unfractionated heparin). Among 7 patients who received warfarin, the international normalized ratio of prothrombin time (PT-INR) was <1.5 in 5 patients. In 2 of 17 patients (12%) with development of symptomatic VTE during follow-up, anticoagulants had been interrupted due to invasive procedures before the development of symptomatic VTE.
Abbreviations as in Table 1.
Patients with prolonged anticoagulant therapy were younger and had higher body weight than those without (Table 1). The prevalence of active cancer was significantly higher in patients with anticoagulants than in those without (44% vs. 28%, P=0.02). All the patients with asymptomatic PE were treated with anticoagulants. Use of parenteral anticoagulants early after diagnosis was significantly more prevalent in patients with prolonged anticoagulant therapy than in those without (62% vs. 13%, P<0.001; Table 1).
There were no significant differences in the cumulative 5-year incidences of symptomatic recurrent VTE and all-cause death between the anticoagulant and no anticoagulant groups (11.8% vs. 22.9%, log-rank P=0.23; and 32.4% vs. 40.7%, log-rank P=0.46, respectively). The cumulative 5-year incidence of major bleeding, however, was significantly higher in the anticoagulant than the no anticoagulant group (20.5% vs. 1.5%, log-rank P=0.01; Figure 4).
Kaplan-Meier curves for cumulative incidence of symptomatic recurrent VTE; major bleeding; and all-cause death vs. presence of prolonged anticoagulant (AC) therapy. VTE, deep vein thrombosis and/or pulmonary embolism.
There were no significant differences in the cumulative 5-year incidences of symptomatic recurrent VTE with regard to active cancer status, DVT provocation status and location, or D-dimer level (Figure 5). After the discontinuation of anticoagulant therapy, there were also no significant differences in the incidence of symptomatic recurrent VTE in all the subgroups (Figure S1).
Kaplan-Meier curves for symptomatic recurrent VTE vs. active cancer status, deep vein thrombosis (DVT) provocation status, DVT location, and D-dimer level. VTE, DVT and/or pulmonary embolism.
In patients with active cancer, the effect of anticoagulant relative to no anticoagulant for symptomatic recurrent VTE was significant (HR, 0.22; 95% CI: 0.05–0.95, P=0.04), whereas it was not significant in patients without active cancer (HR, 1.23; 95% CI: 0.30–8.30, P=0.79; Table 3). In patients with active cancer, none of the patients in the no anticoagulant group had major bleeding events, but there was no statistically significant difference between the groups (log-rank P=0.10; Figure S2). The effects of anticoagulant for symptomatic recurrent VTE with regard to provoked or unprovoked DVT, proximal DVT or distal DVT only, and D-dimer level are given in Table 3.
Abbreviations as in Table 1.
The main findings of the current study are as follows: (1) in patients with asymptomatic DVT, distal DVT accounted for only 70%, and the cumulative 5-year incidence of symptomatic recurrent VTE was 14.5%; (2) prolonged anticoagulant therapy was implemented in 77% of patients with asymptomatic DVT, resulting in higher risk for major bleeding without significant effect on the incidence of symptomatic recurrent VTE during follow-up; and (3) in a subgroup of patients with active cancer, use of anticoagulants was associated with significantly lower risk for developing symptomatic VTE.
Distal DVT are often asymptomatic.12 In the current study, distal DVT was identified in 70% of patients with asymptomatic DVT. A high incidence (16.2%) of asymptomatic distal DVT was reported in a previous study in which ultrasound was performed for 154 consecutive patients admitted to an internal medicine department for acute medical illnesses.13 Also, a high incidence (23%) of asymptomatic DVT was reported in a previous study evaluating 174 consecutive hospitalized non-ambulatory neurosurgical patients.14 These results suggest that the extensive ultrasound screening for DVT in high-risk patients could lead to frequent detection of asymptomatic DVT, but it is unclear whether the intensive screening and detection of asymptomatic VTE is clinically useful. Regarding asymptomatic PE, the clinical outcomes in terms of prognosis and VTE recurrence in patients with asymptomatic PE were almost identical to those for symptomatic PE.15–17 In previous autopsy studies, incidental PE was suggested to be a frequent unrecognized cause of death in cancer patients.18 Based on these results, several clinical VTE guidelines have made weak recommendations for anticoagulant use for cancer-associated asymptomatic PE.1,2 There is a paucity, however, of previous reports on the natural history and the need for anticoagulants in asymptomatic DVT. Vaitkus et al reported higher mortality in asymptomatic proximal DVT patients as compared with those without DVT in their study of hospitalized patients with acute medical illness who were recruited into a randomized, prospective clinical trial of thromboprophylaxis with dalteparin (PREVENT).19 In the current study, the patients with asymptomatic DVT had moderate risk for developing symptomatic VTE (14.5% at 5 years), suggesting that some patients were at high risk for developing symptomatic VTE, even if DVT was initially asymptomatic.
To date, there have been no randomized clinical trials evaluating the efficacy and safety of anticoagulant therapy for asymptomatic DVT. In the current study, prolonged anticoagulant therapy was associated with a significantly increased risk for major bleeding, but without reduced risk for developing symptomatic VTE. On subgroup analysis of active cancer status, however, anticoagulant therapy was associated with a significantly lower risk for symptomatic VTE. Hypercoagulative status in cancer patients might be associated with high risk for symptomatic VTE, resulting in the potential benefit of anticoagulants in patients with active cancer. In contrast, cancer patients are generally considered to be at higher risk for bleeding due to bleeding from tumor, surgical procedure for cancer, and cancer-related thrombocytopenia.20 For decision-making with regard to indications for anticoagulant therapy in patients with active cancer, therefore, we should evaluate the net clinical benefits in individual patients, balancing the risk for developing symptomatic VTE and the risk for bleeding.
The current study had several limitations. First and most importantly, the present single-center study did not have adequate statistical power to estimate the true incidence of symptomatic VTE in the entire cohort as well as in the subgroups. Due to the small number of patients with event, we could not conduct appropriate multivariate analysis to adjust for the confounding factors. Also, analysis of the effect of prolonged anticoagulant therapy for symptomatic recurrent VTE was severely underpowered, particularly on subgroup analysis. Therefore, the favorable effect of anticoagulant therapy in the active cancer group should be regarded, at best, as hypothesis generating. Second, patient demographics, practice patterns for ultrasound screening of DVT, selection of patients for anticoagulant therapy as well as clinical outcomes in asymptomatic DVT patients in the current single-center study, might be different from those in other institutions. Therefore, a future large-scale multicenter study is needed to evaluate the prognosis of asymptomatic DVT. Third, the true rate of concomitant asymptomatic PE was unknown in patients with asymptomatic DVT, because not all patients underwent contrast-enhanced CT at diagnosis of DVT. Fourth, because use of low-molecular-weight (LMW) heparin for VTE was not covered by Japanese national insurance, warfarin was mostly used as the prolonged anticoagulant therapy for cancer patients, although LMW heparin is recommended over warfarin in the current clinical guideline.21 Fifth, thrombosis predisposition was not evaluated in most of the present patients, and we could not investigate the influence of thrombosis predisposition on clinical outcome. Sixth, ultrasound non-compressibility examination was not mandatory, and diagnosis according to the presence of thrombus could lead to false-positive diagnosis of DVT. Seventh, we could not evaluate time in therapeutic range for patients taking warfarin during follow-up, and therefore the effect of quality of warfarin control on outcome is unknown. Finally, the current study was retrospective, with the limitations inherent to observational study design. The indications for ultrasound depended on the discretion of the physicians in charge, which might lead to varying rate of asymptomatic DVT diagnosis across centers. Due to lack of specific recommendations in the current guidelines, there were no standard management strategies for asymptomatic DVT at the present institution. A prospective randomized trial is warranted to define the role of anticoagulants therapy for asymptomatic DVT.
In this study on asymptomatic lower extremity DVT, prolonged anticoagulant therapy was implemented in the majority of patients, but was associated with a significantly higher risk for major bleeding. On subgroup analysis, however, in patients with active cancer, there appeared to be a benefit of prolonged anticoagulant therapy in decreasing the rate of symptomatic recurrent VTE.
We thank the experienced ultrasound technicians T. Yoneda, K. Iwata, H. Kajita, H. Motoda, A. Ishii, M. Yamamoto, M. Yamashita, and H. Fukuyama.
The authors declare no conflict of interest.
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Supplementary File 1
Figure S1. Kaplan-Meier curves for cumulative incidence of symptomatic recurrent VTE vs. active cancer status, deep vein thrombosis (DVT) provocation status, DVT location, and D-dimer level after discontinuation of anticoagulant therapy.
Figure S2. Kaplan-Meier curves for cumulative incidence of major bleeding vs. prolonged anticoagulant (AC) status in patients with active cancer.
Table S1. Cause of death
Please find supplementary file(s);