Article

D-dimer levels in VTE patients with distal and proximal clots

a b s t r a c t

Objectives: There is growing evidence that venous thromboembolism patients with distal clots (distal calf deep vein thrombosis [DVT] and sub-segmental pulmonary embolism [PE]) may not routinely benefit from anticoagulation. We compared the D-Dimer levels in VTE patients with distal and Proximal clots.

Methods: We conducted a multinational, prospective observational study of low-to-intermediate risk adult pa- tients presenting to the emergency department (ED) with suspected VTE. Patients were classified as distal (calf DVT or sub-segmental PE) or proximal (proximal DVT or non-sub-segmental PE) clot groups and compared with univariate and multivariate analyses.

Results: Of 1752 patients with suspected DVT, 1561 (89.1%) had no DVT, 78 (4.4%) had a distal calf DVT, and 113 (6.4%) had a proximal DVT. DVT patients with proximal clots had higher D-dimer levels (3760 vs. 1670 mg/dL) than with distal clots. Sensitivity and negative predictive value (NPV) for proximal DVT at an optimal D-dimer cutoff of 5770 mg/dL were 40.7% and 52.1% respectively. Of 1834 patients with suspected PE, 1726 (94.1%) had no PE, 7 (0.4%) had isolated sub-segmental PE, and 101 (5.5%) had non-sub-segmental PE. PE patients with prox- imal clots had higher D-dimer levels (4170 vs. 2520 mg/dL) than those with distal clots. Sensitivity and NPV for proximal PE at an optimal D-dimer cutoff of 3499 mg/dL were 57.4% and 10.4% respectively.

Conclusions: VTE patients with proximal clots had higher D-dimer levels than patients with distal clots. However, D-dimer levels cannot be used alone to discriminate between VTE patients with distal or proximal clots.

(C) 2018

Introduction

Venous thromboembolism has a disease spectrum from asymptomatic distal Deep vein thrombosis to fatal pulmonary embolism (PE) [1]. Because VTE is potentially fatal and can easily be missed, efforts to improve identification of VTE have included the intro- duction of biomarkers (such as D-dimer) as well as advanced imaging technologies such as multi-detector computed tomographic pulmonary angiography (CTPA) and venous duplex ultrasound (US). This has led to an increasing number of patients being diagnosed with VTE [2,3] and the identification of more distal clots (such as isolated subsegmental PE or calf-vein DVT) [4] that may have been missed with earlier, less- advanced imaging.

? Supported by Siemens Healthcare Diagnostics, Newark, DE.

* Corresponding author at: Department of Emergency Medicine, HSC-L4-080, 8350 SUNY, Stony Brook, NY 11794, United States.

E-mail address: [email protected]. (A.J. Singer).

There is considerable controversy whether to anticoagulate patients with isolated, distal calf DVT or subsegmental PE. Studies suggest that many subsegmental PE and distal DVT are, in fact, false positives [4,5]. Even if true positives, diagnosing subsegmental PE in many cases may be irrelevant [6,7], exposing patients to radiation, potentially harmful intravenous Contrast agents and the risks of bleeding when anticoagulation is not truly indicated [7]. A retrospective study includ- ing 60 patients with subsegmental PE and no proximal DVT who were not anticoagulated reported no episode of recurrent VTE [7]. In contrast, another study found that the risk of recurrent VTE in patients with subsegmental PE who were anticoagulated was similar to that of patients with more proximal PE and higher than those with no PE at all [8]. Recent guidelines issued by the American College of Chest Physicians address these issues. They suggest that patients with isolated calf DVT and subsegmental PE without cancer do not require anticoagulation [9]. Regardless, efforts to identify patients with distal clots using clinical information or the results of D-dimer assays have the potential to reduce the associated unnecessary risks and costs of imaging and anticoagulation in these low-risk patients.

https://doi.org/10.1016/j.ajem.2018.04.040

0735-6757/(C) 2018

The first goal of the current study was to determine the incidence of distal calf DVT and subsegmental PE in a large contemporary cohort of emergency department (ED) and outpatient clinic patients with suspected VTE. The second goal was to determine whether patient characteristics and D-dimer levels could discriminate between patients with distal and more proximal VTE. We hypothesized that D-dimer levels would be significantly lower in patients with isolated calf DVT or subsegmental PE than in patients with more proximal DVT and PE.

Methods

Study design and setting

We conducted a multi-center, international, prospective, observa- tional study of adult ED patients with suspected PE or DVT. The centers included academic and non-academic emergency departments from the United States (n = 18) and Europe (n = 6). The study was part of another study evaluating a novel D-dimer assay that was not used for clinical purposes. The study was sponsored by Siemens Healthcare Diagnostics (Newark, DE), which had no role in data analysis or inter- pretation. The study was approved by all institutional review boards and all patients gave written, informed consent.

Patients

We included a convenience sample of patients ages 18 years and older who presented to an ED or outpatient clinic with suspected PE or DVT who were referred for objective testing and were able to give written informed consent. We only included patients with low-to- intermediate risk of VTE based on Wells criteria (<=6 for PE and b 2 for DVT). For suspected PE, objective testing was based on a CTPA. For suspected DVT, objective testing was based on venous US. Patients receiving anticoagulants for N24 h prior to enrollment were excluded as well as pregnant women and patients at high-risk of VTE (i.e., Wells PE score N 6, Wells DVT score >= 2).

Measures

We used a structured electronic data collection form to record baseline patient demographics and clinical characteristics including age, sex, race, clinical D-dimer assays, Imaging results, three-month outcome and final diagnosis.

Outcomes

All patients underwent a diagnostic criterion standard evaluation for VTE. A PE was diagnosed during the index visit if the patient had a filling defect in a pulmonary artery on CTPA. DVT was diagnosed if the patient had a venous ultrasound demonstrating a non-compressible clot in their leg. Clots were considered to be proximal if the PE was in a seg- mental or larger pulmonary artery or if the DVT was proximal to the calf trifurcation (i.e., in or above the popliteal vessels). All diagnoses were confirmed by a three-member adjudication committee.

Data analysis

Dichotomous data are presented as numbers and percentages frequency of occurrence and compared using the Chi-square or Fisher’s exact tests. Continuous data are presented as means with standard deviations (SD) or medians with inter-quartile ranges (IQR) and com- pared with the Mann Whitney U test or analysis of variance. Multivari- able adjustment was performed using linear regression for continuous data and logistic regression for dichotomous data. Receiver operating characteristic (ROC) analyses and area under the curve (AUC) calcula- tions were performed to determine Optimal D-dimer thresholds using the Youden index for discriminating between distal and proximal clots

for PE and DVT separately. Statistical analyses were conducted using SAS version 9.4 (SAS Institute, Inc., Cary, NC).

Results

General study population

We enrolled at total of 3586 patients across the study sites. Of these, there were 1752 patients (48.9%) with suspected DVT. Their mean (SD) age was 53.1 (+-16.2) years; 1042 (59.5%) were female and 1172 (66.9%) were white. There were also 1834 patients (51.1%) with suspected PE. Their mean (SD) age was 47.7 (+-15.8) years, 1158 (63.1%) were female and 1081 (58.9%) were white. VTE was confirmed in 191 patients (11.1%) with suspected DVT and in 108 patients (5.9%) with suspected PE.

Patients with suspected DVT

Of 1752 patients with suspected DVT, 1561 (89.1%) had no DVT, 78 (4.4%) had a distal calf DVT, and 113 (6.4%) had a proximal DVT. A com- parison of patients by presence or absence of clot, and location of clot is presented in Table 1. The results of multivariable analysis are presented in Table 2.

DVT patients with a proximal clot were more commonly male (62.0% vs. 48.7%; OR 1.85 [95% CI, 1.33-2.59]), more likely to have a

High clinical suspicion of DVT (81.4% vs. 66.7%; OR 1.82 [95% CI, 1.02-3.24]) and had higher D-dimer levels (3760 vs. 1670 mg/dL, P b 0.001) than DVT patients with distal clots. Patients with no DVT were younger (52.8 vs. 55.0 years, P = 0.02), more likely female (61.4% vs. 43.5%, P b 0.001), less likely to have calf swelling (14.0% vs. 24.1%, P b 0.001), less likely to have a history of cancer (3.8% vs. 7.3%, P = 0.02), and had lower D-dimer levels than those with any clot (520 mg/dL vs. 2570 mg/dL, P b 0.001).

The optimal D-dimer cut-off value for discriminating between prox- imal and distal DVT was 5770 mg/dL with an area under the curve of

0.68 (95% CI, 0.61 to 0.76, Fig. 1). The test characteristics using this threshold are presented in Table 3.

Patients with suspected PE

Of 1834 patients with suspected PE, 1726 (94.1%) had no PE, 7 (0.4%) had isolated sub-segmental PE, and 101 (5.5%) had non-sub-segmental PE. Thus, 6.4% of all PE were isolated subsegmental. A comparison of pa- tient characteristics by presence or absence of clot, and location of clot is presented in Table 4. The results of multivariable analysis are presented in Table 5.

PE patients with proximal clots were younger (56.5 vs. 67.3 years, P b 0.001), more likely to have a Well’s score >= 4 (47.0% vs. 14.3%, OR 4.82 [95% CI, 3.03-7.67]), and had higher D-dimer levels (4170 vs.

Table 1

Univariate associations between predictor variables and deep vein thrombosis.

No DVTa Distal DVT Proximal DVT P value

Mean (SD) age

52.78 (16.31)

55.29 (14.95)

56.17 (15.5)

0.048

Females, no. (%)

Race, no. (%)

959 (61.43)

40 (51.28)

43 (38.05)

b0.001

White

1040 (66.62)

49 (62.82)

83 (73.45)

0.171

Black

427 (27.35)

24 (30.77)

24 (21.24)

Hispanic

73 (4.68)

2 (2.56)

4 (3.54)

Asian

8 (0.51)

2 (2.56)

2 (1.77)

Other

13 (0.83)

1 (1.28)

0 (0)

Prior cancer, no. (%) 59 (3.78)

5 (6.41)

9 (7.96)

0.059

Bedridden, no. (%) 145 (9.29)

10 (12.82)

11 (9.73)

0.580

Calf swelling, No. (%) 218 (13.97)

12 (15.38)

34 (30.09)

b0.001

Calf tenderness, no. (%) 607 (38.89)

41 (52.56)

41 (36.28)

0.043

Median (IQR) D-dimer 520 mg/dL

1670 mg/dL

3760 mg/dL

b0.001

a DVT = deep vein thrombosis.

Table 2

Multivariate association between predictor variables and proximal DVTa.

Variable Odds ratio 95% CI P-value

Table 3

Diagnostic test characteristics using the optimal D-dimer cutoff point (5770 mg/dL) for discriminating between distal and proxi- mal deep vein thrombosis.

a DVT = deep vein thrombosis.

PPV 90.2% (79.0-95.6%)

NPV 52.1% (43.9-60.3%)

Age (per one year increase)

Male sex

1.00

1.85

0.99-1.01

1.33-2.59

0.9457

0.0003

Estimate and 95% CI

white race

0.85

0.57-1.28

0.4439

Sensitivity

40.7% (32.1-49.9%)

High clinical suspicion for DVT

1.82

1.02-3.24

0.0003

Specificity

93.6% (85.8-97.2%)

2520 mg/dL), than PE patients with distal clots. While there was a sig- nificant difference among patients with no PE, distal PE, and proximal PE, the difference in D-dimer levels between those with a distal and proximal clot was not statistically significant (P = 0.21). The difference in D-dimer levels between patients with no PE or a distal PE was significant (P = 0.004). The presence of tachycardia, signs of DVT, and hemoptysis did not differ among the groups.

Patients with no PE were younger (47.1 vs. 60.0 years, P b 0.001), more likely female (63.9% vs. 50.9%, P = 0.007), less likely to have a recent history of immobilization (9.4% vs. 20.4%, P b 0.001), less likely to have a previous VTE (8.5% vs. 23.1%, P b 0.001), and had lower D-dimer levels (440 mg/dL vs. 4110 mg/dL, P b 0.001) than patients with any clots (both proximal and distal).

The optimal D-dimer cut-off value for discriminating between prox- imal and distal PE was 3499 mg/dL with an area under the curve of 0.64 (95% CI, 0.43 to 0.86, Fig. 2). The test characteristics using this threshold are presented in Table 6.

Discussion

The results of our prospective, observational international study in- dicate that distal calf DVT and especially sub-segmental PE are relatively rare in low-to-intermediate risk adult patients presenting to the ED or outpatient setting with suspected VTE. The incidence of subsegmental PE in our investigation is consistent with a prior VTE study [6]. Our results also show that D-dimer levels in patients with distal clots are significantly lower than those in patients with proximal clots while being significantly higher than in those patients without any clots. Our result also shows that, in general, traditional predictors of having any

Fig. 1. Area under the curve for dep vein thrombosis.

clot are more common in those with proximal clots than distal clots. However, the low sensitivities and negative predictive values of the D-dimer when using the thresholds suggested by the Youden indices are not accurate enough to identify patients with distal clots. This question is especially important if one considers that some patients with distal clots may not require anticoagulation at all based on the evidence and recent guidelines [9]. Despite being incorporated into the ACCP Guidelines there remains equipoise among the thrombosis community due to the absence of a large randomized clinical trial inves- tigating the topic [10].

The overall rate of subsegmental PE in our study was low (about 6% of all PE). Other studies suggest that improvements in CTPA have led to a doubling in the rate of diagnosing subsegmental PE from 5 to 10% [6,11,12]. Some studies have reported rates of subsegmental PE on CTPA as high as 15-25% [13,14]. As a result, some experts have voiced concern that the increased incidence of PE in recent years (together with no increase in mortality) is the result of identification of small distal clots by more sensitive CT scanners. However, the fact that the overwhelming majority of PEs in our study are proximal reduces this concern. It is possible that in our study, the radiologists were more attuned to the interrater unreliability of CTPA for small subsegmental PE [14] and were aware of the new Chest guidelines of not treating isolated subsegmental PE [9]. As a result, there were less likely to call an isolated, possible small filling defect, as positive.

Prior studies have also evaluated the association of D-dimer levels and Clot burden or location. Ghanima et al. studied 100 consecutive outpatients diagnosed with PE [15]. There was a significant association between log D-dimer and between log RV/LV ratio (r = 0.45), pulmo- nary artery obstruction index (r = 0.50), and PaO2 (r = 0.40). There was also a significant association between the D-dimer level and the most proximal level of PE (P b 0.005). Galle et al. studied 104 hemody- namically stable patients with PE [16]. Median plasma D-dimer levels were higher in patients with N50% Perfusion defects compared to those with b30% defects (7950 mg/dL vs. 2731 mg/dL, P = 0.001).

Table 4

Univariate associations between predictor variables and pulmonary embolism.

No PEa

Distal PE

Proximal PE

P valueb

Mean (SD) age

47.1 (15.57)

67.3 (10.31)

56.5 (16.6)

b0.001

Females, no. (%)

1103 (63.9%)

3 (42.86%)

52 (51.49%)

0.023

Race, no. (%)

White, no. (%)

1007 (58.34)

3 (42.86)

71 (70.3)

0.093

Black

525 (30.42)

3 (42.86)

25 (24.75)

Hispanic

141 (8.17)

0 (0)

4 (3.96)

Asian

18 (1.04)

0 (0)

0 (0)

Other

35 (2.03)

1 (14.29)

1 (0.99)

Wells score N 4, no. (%)

244 (14.14)

1 (14.29)

47 (47)

b0.001

Signs of DVT, no. (%)

132 (7.65)

0 (0)

13 (12.87)

0.124

Tachycardia, no. (%)

486 (28.17)

3 (42.86)

32 (31.68)

0.523

Immobilization, no. (%)

162 (9.39)

1 (14.29)

21 (20.79)

b0.001

Previous DVT/PE, no. (%)

147 (8.52)

0 (0)

25 (24.75)

b0.001

Hemoptysis, no. (%)

47 (2.72)

0 (0)

2 (1.98)

0.820

Cancer, no. (%)

115 (6.67)

4 (57.14)

12 (11.88)

b0.001

Median (IQR) D-dimer

440 mg/dL

2520 mg/dL

4170 mg/dL

b0.001

a PE = pulmonary embolism.

b The P value indicates whether there is a difference among the 3 groups: no clot, distal clot or proximal clot.

Table 5

Multivariate association between predictors variables and proximal pulmonary embolism.

Variable Odds ratio 95% CI P-value

Table 6

Diagnostic test characteristics using the optimal D-dimer cutoff point (3499 mg/dL) for discriminating between distal and proxi- mal pulmonary embolism.

Age (per 1 year increase)

Female sex

0.97

1.38

0.96-0.99

0.87-2.20

0.0001

0.1686

Estimate and 95% CI

White race

0.81

0.49-1.33

0.3993

Sensitivity

57.4% (47.7-66.6%)

Wells score N 4

4.82

3.03-7.67

b0.0001

Specificity

71.4% (34.9-91.5%)

PPV

96.7% (88.7-99.0%)

NPV

10.4% (4.6-22.2%)

In another cohort of 674 consecutive patients with confirmed PE, high D-dimers (N3000 mg/dL) were associated with mortality (OR 7.29). They were also associated with central location of emboli, active malignancy and age [17]. Aujesky et al. studied 366 patients with PE diagnosed at four emergency departments [18]. Patients who died had higher median D-dimer levels than those who survived (4578 mg/dL vs. 2946 mg/dL, P = 0.005). Patients with D-dimers below 1500 mg/dL had very low mortality and the area under the ROC curve of D-dimer for predicting overall mortality was 0.69 (95% CI, 0.58-0.80). Hochuli et al. assessed D-dimers in 254 patients with suspected PE [19]. The prevalence of PE increased with D-dimer levels and was 7% at D-dimer levels of 500-1000 mg/dL, reaching N90% at D-dimer levels exceeding 9000 mg/dL (P b 0.001). High D-dimer levels (above 4000 mg/dL) were also associated with significantly higher clot burden in pulmonary arteries and thrombus in the main pulmonary artery. Similar results were found for distal versus proximal DVT.

Strengths and limitations

Our study has several strengths and some limitations. We employed a prospective, real-world design, with pre-test probability assessment and blood sampling performed prior to the clinical diagnosis or exclu- sion of VTE. Our study was performed in a large number of emergency departments in various regions throughout the United States and Europe, and should therefore be generalizable to a broad population of emergency patients. Last, the large number of subjects enrolled resulted in tight confidence intervals for our primary analysis and permitted several sub-analyses.

A major limitation of our study is that not all patients underwent diagnostic imaging for VTE. Some patients had VTE excluded with a

Fig. 2. Area under the curve for pulmonary embolism.

combination of Low pre-test probability, a negative clinical D-dimer and follow-up. This is consistent with previous studies and the standard of care [20]. Therefore, it would not be ethical to expose patients with negative D-dimer results to unnecessary imaging. Some patients were lost to follow-up. These patients were considered not to have clinically significant VTE based on the results of their workup during the index visit, but we cannot exclude the possibility that some had VTE during the follow-up period. However, standard clinical care permits the exclusion of VTE in low and intermediate probability patients based on negative testing in the emergency department, and our results are consistent with this approach. A final limitation is that most patients did not have both imaging of their legs and their lungs. As a result, there were very few patients in whom a concomitant DVT and PE were confirmed. Thus, we could not perform a separate analysis to determine whether D-dimer levels can be used to identify patients with clots (distal or proximal) in one location (leg or lung) who were more likely to have another clot (distal or proximal) in the other location (lung or leg).

Conclusions

The overall incidence of distal calf DVT and isolated sub-segmental PE is relatively low in adult patients presenting to the ED or outpatient clinic with suspected VTE of low-to-intermediate probability. VTE patients with proximal clots were more likely to have a high clinical suspicion for VTE and had higher D-dimer levels than patients with more distal clots. However, due to poor sensitivity and NPV, D-dimer levels cannot be used alone to discriminate between VTE patients with distal or proximal clots.

References

  1. Di Nisio M, Buller HR. Deep vein thrombosis and pulmonary embolism. Lancet 2016;

    388:3060-73.

    Wiener RS, Schwartz LM, Woloshin S. Time trends in pulmonary embolism in the United States: evidence of overdiagnosis. Arch Intern Med 2011;171:831-7.

  2. Dentali F, Ageno W, Pomero F, Fenoglio L, Squizzato A, Bonzini M. Time trends and case fatality rate of in-hospital treated pulmonary embolism during 11 years of observation in Northwestern Italy. Thromb Haemost 2016;115:399-405.
  3. Pena E, Kimpton M, Dennie C, Peterson R, Le Gal G, Carrier M. Differences in interpre- tation of Computed tomography pulmonary angiography diagnosis of subsegmental thrombosis in patients with suspected pulmonary embolism. J Thromb Haemost 2012;10:496-8.
  4. Righini M, Paris S, Le Gal G, Laroche JP, Perrier A, Bounameaux H. Clinical relevance of distal deep vein thrombosis. Review of the literature. Thromb Haemost 2006;95: 56-64.
  5. Carrier M, Righini M, Wells PS, et al. Subsegmental pulmonary embolism diagnosed by computed tomography: incidence and clinical implications. A systematic review and meta-analysis of the management outcome studies. J Thromb Haemost 2010;8: 1716-22.
  6. Carrier M, Righini M, Le Gal G. Symptomatic subsegmental pulmonary embolism: what is the next step? J Thromb Haemost 2012;10:1489-90.
  7. den Exter PL, van Es J, Klok FA, et al. Risk profile and clinical outcome of symptomatic subsegmental acute pulmonary embolism. Blood 2013;122:1144-9.
  8. Kearon C, Elie AA, Ornelas J, et al. antithrombotic therapy for VTE disease. CHEST guideline and expert panel report. Chest 2016;149:315-52.
  9. Yoo HH, Queluz TH, El Dib R. anticoagulant treatment for subsegmental pulmonary embolism. Cochrane Database Syst Rev 2016 Jan 12;1:CD010222. https://doi.org/10. 1002/14651858.CD010222.pub3 (Review, PubMed PMID: 26756331).
  10. Weiner RS, Schwartz LM, Woloshin S. When is a test too good: how CT pulmonary angiograms find pulmonary emboli that do not need to be found. BMJ 2013;347: f3368.
  11. Stein PD, Goodman LR, Hull RD, Dalen JE, Matta F. Diagnosis and management of iso- lated subsegmental pulmonary embolism: review and assessment of the options. Clin Appl Thromb Hemost 2012;18:20-6.
  12. Nazaroglu H, Ozmen CA, Akay HO, Kilinc I, Bilici A. 64-MDCT pulmonary angiogra- phy and CT venography in the diagnosis of Thromboembolic disease. AJR Am J Roentgenol 2009;192:654-61.
  13. Miller Jr WT, Marinari LA, Barbosa Jr E, Litt HI, Schmitt JE, Mahne A, et al. Small pul- monary artery defects are not reliable indicators of pulmonary embolism. Ann Am Thorac Soc 2015;12:1022-9.
  14. Ghanima W, Abdelnoor M, Holmen LO, Nielssen BE, Ross A, Sandset PM. D-dimer level is associated with the extent of pulmonary embolism. Thromb Res 2007;12:281-8.
  15. Galle C, Papazyan JP, Miron MJ, Slosman D, Bounameaux H, Perrier A. Prediction of pulmonary embolism extent by clinical findings, D-dimer level and deep vein thrombosis shown by ultrasound. Thromb Haemost 2001;86:1146-60.
  16. Klok FA, Djurabi RK, Nijkeuter M, et al. High D-dimer level is associated with in- creased 15-d and 3 months mortality through a more central location of pulmonary emboli and serious comorbidity. Br J Haematol 2007;140:218-22.
  17. Aujesky D, Roy PM, Guy M, Cornuz J, Sanchex O, Parrier A. Prognostic value of D-dimer in patients with pulmonary embolism. Thromb Haemost 2006;96:478-82.
  18. Hochuli M, Duewell S, Frauchiger B. Quantitative d-dimer levels and the extent of venous thromboembolism in CT angiography and lower limb ultrasonography. Vasa 2007;36:267-74.
  19. Kearon C, Ginsberg JS, Douketis J, Turpie AG, Bates SM, Lee AY, et al. An evaluation of D-dimer in the diagnosis of pulmonary embolism: a randomized trial. Ann Intern Med 2006;144:812-21.

Leave a Reply

Your email address will not be published. Required fields are marked *