
==== Front
Jpn J Clin Oncol
Jpn J Clin Oncol
jjco
Japanese Journal of Clinical Oncology
0368-2811
1465-3621
Oxford University Press

38769814
10.1093/jjco/hyae064
hyae064
Original Article
AcademicSubjects/MED00300
D-dimer cut-off value for predicting venous thromboembolism at the initial diagnosis in Japanese patients with advanced lung cancer
https://orcid.org/0000-0003-2281-7778
Kawakado Keita Division of Medical Oncology and Respiratory Medicine, Department of Internal Medicine, Shimane University Faculty of Medicine, Izumo, Japan
Department of Respiratory Internal Medicine, National Hospital Organization Hamada Medical Center, Hamada, Japan

Tsubata Yukari Division of Medical Oncology and Respiratory Medicine, Department of Internal Medicine, Shimane University Faculty of Medicine, Izumo, Japan

Hotta Takamasa Division of Medical Oncology and Respiratory Medicine, Department of Internal Medicine, Shimane University Faculty of Medicine, Izumo, Japan

Yamasaki Masahiro Department of Respiratory Disease, Hiroshima Red Cross Hospital and Atomic-Bomb Survivors Hospital, Hiroshima, Japan

Ishikawa Nobuhisa Department of Respiratory Medicine, Hiroshima Prefectural Hospital, Hiroshima, Japan

https://orcid.org/0000-0003-3557-0049
Masuda Takeshi Department of Respiratory Medicine, Hiroshima University Hospital, Hiroshima, Japan

Kubota Tetsuya Department of Respiratory Medicine and Allergology, Kochi University Hospital, Nankoku, Japan

Kobayashi Kunihiko Department of Respiratory Medicine, Saitama Medical University International Medical Center, Saitama, Japan

Isobe Takeshi Division of Medical Oncology and Respiratory Medicine, Department of Internal Medicine, Shimane University Faculty of Medicine, Izumo, Japan

For reprints and all correspondence: Yukari Tsubata, Division of Medical Oncology and Respiratory Medicine, Department of Internal Medicine, Shimane University Faculty of Medicine, 89-1 Enya-cho, Izumo, Shimane 693-8501, Japan. E-mail: ytsubata@med.shimane-u.ac.jp
9 2024
20 5 2024
20 5 2024
54 9 10321036
11 1 2024
30 4 2024
04 5 2024
© The Author(s) 2024. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permission@oup.com.
2024
https://creativecommons.org/licenses/by-nc/4.0/ This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (https://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com

Abstract

Objective

Cancer is a well-known risk factor for venous thromboembolism. The D-dimer level is used to predict venous thromboembolism; however, reports on an appropriate D-dimer cut-off value in Japanese patients with advanced lung cancer are lacking. Therefore, this study aimed to calculate the D-dimer cut-off value for venous thromboembolism at the time of lung cancer diagnosis.

Methods

The Rising-venous thromboembolism/NEJ037 study was a multicenter, prospective observational study. Patients with lung cancer who were contraindicated for radical resection or radiation were enrolled and followed up for 2 years. In the present study (jRCT no. 061180025), a receiver operating characteristic curve for D-dimer levels was created using the dataset of the Rising-venous thromboembolism/NEJ037 study.

Results

The Rising-venous thromboembolism/NEJ037 study included a total of 1008 patients, of whom 976, whose D-dimer levels had been measured at the time of cancer diagnosis, were included in the present study. At the time of lung cancer diagnosis, 62 (6.3%) and 914 (93.7%) patients presented with and without venous thromboembolism, respectively. The D-dimer values ranged from 0.1 to 180.1 μg/ml and from 0.1 to 257.2 μg/ml in patients with and without venous thromboembolism, respectively. The receiver operating characteristic curve was discriminative with a cut-off value of 3.3 μg/ml and an area under the curve of 0.794 (sensitivity, 0.742; specificity, 0.782; 95% confidence interval, 0.725–0.863).

Conclusions

This is the first study to calculate the D-dimer cut-off value in Japanese patients with advanced lung cancer. Patients with D-dimer levels ≥3.3 μg/ml at the time of initial diagnosis may have coexisting venous thromboembolism.

Easily measurable biomarkers are important to identify asymptomatic venous thromboembolism (VTE). A D-dimer level ≥3.3 μg/ml at the time of diagnosis of lung cancer is predictive of coexisting VTE.

lung cancer
venous thrombosis
D-dimer
edoxaban
Daiichi Sankyo Co., Ltd LIXMD-15003
==== Body
pmcIntroduction

Venous thromboembolism (VTE) is a common medical complication in patients with cancer, and the risk of developing VTE is 4–20 times greater in patients with cancer than in those without cancer (1,2). Additionally, lung cancer is associated with a high risk of VTE (2). The incidence of VTE increases with chemotherapy (3,4). The Rising-VTE/NEJ037 was a multicenter prospective observational study on patients with lung cancer, in which 9.9% of patients developed VTE 2 years after enrollment; of these, 75% were asymptomatic (5). Therefore, it is important to identify cases that should be monitored closely, for which easily measurable biomarkers are important. However, evidence regarding biomarkers for screening VTE associated with cancer is scarce, and further development is expected in this area. D-dimer is a degradation product of cross-linked fibrin that appears in the blood after a blood clot is degraded by fibrinolysis (6). D-dimer levels are useful for the auxiliary diagnosis of VTE (7–9). Furthermore, patients with cancer have higher D-dimer levels than those without cancer, and a specific cut-off value is required for VTE diagnosis in patients with cancer (10). Therefore, we conducted a post-hoc analysis of the Rising-VTE/NEJ037 study dataset to construct a receiver operating characteristic (ROC) curve and calculated the D-dimer cut-off value for predicting VTE at the time of lung cancer diagnosis.

Materials and methods

The Rising-VTE/NEJ037 study was conducted in accordance with the principles of the Declaration of Helsinki and Good Clinical Practice Guidelines. The study protocol was approved by the Shimane University Institutional Review Board based on the Clinical Trials Act enacted in Japan in 2017 and registered with the Japan Registry of Clinical Trials (jRCTs061180025). Written informed consent was obtained from all patients (5).

Patients

The Rising-VTE/NEJ037 study was a multicenter prospective observational study involving Japanese patients with advanced small and non-small cell lung cancer. Among the patients enrolled in this study were participants whose D-dimer levels were measured at the time of initial diagnosis of lung cancer. The primary endpoints were the incidence of newly diagnosed VTE at 2 years and the incidence of VTE recurrence at 6 months in the group treated with edoxaban, which is a direct oral anticoagulant. The study enrollment period was from June 2016 to August 2018. The main enrollment criteria were age ≥20 years, lung cancer for which radical resection or radiation therapy was difficult, Eastern Cooperative Oncology Group Performance Status (ECOG PS) scores of 0–3, and expected survival ≥6 months. There were no exclusion criteria for patient enrollment (5). The present post-hoc study included patients whose D-dimer levels had been measured at the time of lung cancer diagnosis.

Definition of VTE

VTE was defined as a pulmonary embolism (PE) or deep-vein thrombosis (DVT) with a propensity for bilateral or central extension. DVT in only one lower leg was not included as a VTE in this study.

Diagnosis of VTE

In all cases enrolled in the Rising-VTE/NEJ037 study, contrast-enhanced computed tomography (CT) of the chest to the pelvic region and lower extremities, as well as venous echoes of the lower extremities, were used to diagnose VTE. The final diagnosis was made by two radiologists (5).

Statistical analyses

All statistical analyses were performed using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan) in R software (11). EZR is a modified version of the R Commander designed to add statistical functions frequently used in biostatistics. A ROC curve was constructed using the enrollment dataset of the Rising-VTE/NEJ037 study, and the D-dimer cut-off value for VTE at the time of initial diagnosis was calculated. In this study, the positive and negative predictive values were calculated when the cutoff value of D-dimer was set at 1.0 μg/ml, according to a previous study (12).

Results

Patients’ characteristics

The Rising-VTE/NEJ037 study included a total of 1021 patients, of whom 13 were excluded because of missing data. After excluding 32 cases in which D-dimer levels were not measured at the time of initial diagnosis, 976 cases were included in the present study (Fig. 1). The median age of the patients was 71 years (range, 30–94 years); most patients were male (690 patients, 70.7%) and had a good ECOG PS (0–1, 88.4%). The most common histological subtype of lung cancer was adenocarcinoma (623 patients, 63.8%), and most patients underwent chemotherapy (893 patients, 91.5%). The disease stage was assessed according to the Union for International Cancer Control, seventh edition, tumor-node-metastasis staging system for lung cancer (13); extensive disease (small-cell lung carcinoma) and stage IV disease accounted for ~75% of all cases (Table 1) (5).

Figure 1 Cluster classification for venous thromboembolism screening. ECOG PS, Eastern Cooperative Oncology Group Performance Status.

Table 1 Patient characteristics excluding patients whose D-dimer was not measured

All N = 976		N(%)	
Age	Median	71	
	Range	30–94	
Gender (%)	Male	690 (70.7)	
	Female	286 (29.3)	
ECOG PS (%)	0	389 (39.9)	
	1	473 (48.5)	
	2	73 (7.5)	
	3	41 (4.2)	
Histological type	Adenocarcinoma	623 (63.8)	
	Squamous cell carcinoma	181 (18.5)	
	Small cell carcinoma	133 (13.6)	
	Others	39 (4.1)	
Chemotherapy	Yes	893 (91.5)	
	No	79 (8.1)	
	Unknown	4 (0.4)	
Stage	Ib	1 (0.1)	
	IIa	4 (0.4)	
	IIb	4 (0.4)	
	IIIa	28 (2.9)	
	IIIb	71 (7.3)	
	IV	600 (61.5)	
	Postoperative recurrence	135 (13.8)	
	Limited disease (SCLC)	7 (0.7)	
	Extensive disease (SCLC)	126 (12.9)	
ECOG PS, Eastern Cooperative Oncology Group Performance Status; SCLC, small cell lung carcinoma.

D-dimer value and VTE incidence

VTE at the initial diagnosis was observed in 62 patients (6.3%). DVT was the most common type of VTE (32 cases), followed by DVT with PE (15 cases). Regarding VTE symptoms, 51 patients (82.3%) were asymptomatic (Fig. 2). D-dimer levels were compared between the VTE and non-VTE groups using t-tests. The median D-dimer values in the groups with and without VTE were 7.7 μg/ml (range, 0.1–180.1 μg/ml) and 1.4 μg/ml (range, 0.1–257.2 μg/ml), respectively. The D-dimer level was significantly higher in the VTE group (P < 0.001) (Fig. 3). A ROC curve was constructed for the D-dimer value in patients with VTE at the time of the initial diagnosis, and the cut-off value was 3.3 μg/ml (sensitivity, 0.742; specificity, 0.782). The area under the curve (AUC) was 0.794 (95% confidence interval [CI], 0.7247–0.8625), indicating discrimination ability (Fig. 4). Out of 245 cases with a D-dimer value ≥3.3 μg/ml, 46 (18.8%) had VTE. Similarly, out of 731 cases with a D-dimer value <3.3 μg/ml, 715 (97.8%) did not have VTE. The positive and negative predictive values were 0.188 and 0.978, respectively, indicating that a D-dimer cut-off value of 3.3 μg/ml was useful for excluding VTE (Table 2). When the cutoff value of D-dimer was set at 1.0 μg/ml, VTE was complicated in 56 patients (8.7%) and uncomplicated in 590 patients (91.3%) in the D-dimer ≥1.0 μg/ml group. In the D-dimer <1.0 μg/ml group, 6 patients (1.8%) had VTE and 324 patients (98.2%) did not have VTE. The positive and negative predictive values were 8.7% and 98.2%, respectively, and the negative predictive value was the same as when the cutoff value was 3.3 μg/ml, but the positive predictive value was lower (Table 3).

Figure 2 Types and classification of VTE. VTE, venous thromboembolism; DVT, deep-vein thrombosis; PE, pulmonary embolism.

Figure 3 Comparison of D-dimer values between groups with and without VTE. D-dimer values were compared between the groups with and without VTE using a t-test.

Figure 4 ROC curve for D-dimer value in patients with VTE at initial diagnosis. The cut-off value is 3.3 μg/ml (sensitivity, 0.742; specificity, 0.782; 95% CI, 0.7247–0.8625). The AUC is 0.794, indicating discrimination ability. ROC, receiver operating characteristic; VTE, venous thromboembolism; AUC, area under the curve.

Table 2 VTE incidence based on a D-dimer cutoff value of 3.3 μg/ml

	VTE (+)	VTE (−)	Total	
D-dimer ≧3.3 μg/ml	46	199	245	
D-dimer <3.3 μg/ml	16	715	731	
Total	62	914	976	
VTE, venous thromboembolism.

Table 3 VTE incidence based on a D-dimer cutoff value of 1.0 μg/ml

	VTE (+)	VTE (−)	Total	
D-dimer ≧ 1.0 μg/ml	56	590	646	
D-dimer < 1.0 μg/ml	6	324	330	
Total	62	914	976	

Discussion

The Rising-VTE/NEJ037 study is one of the largest prospective observational studies on VTE and hemorrhagic events in Japanese patients with advanced lung cancer. Our study is the first to use data from a prospective observational study to calculate the cut-off value for D-dimer levels for excluding VTE at the initial diagnosis in Japanese patients with advanced lung cancer.

In this study, we constructed a ROC curve to determine the cut-off D-dimer level for excluding VTE at the initial diagnosis in Japanese patients with advanced lung cancer. The cut-off value for D-dimer was 3.3 μg/ml, and the AUC of the ROC curve was 0.794, indicating that it had discrimination ability. The positive predictive value was 0.188, which was not high, but this was thought to be due to the fact that the incidence of VTE in the overall population was 62 out of 978 cases (6.3%), which was not high. However, the negative predictive value was very high (0.978), suggesting that it was useful for diagnosis based on exclusion.

Kearon et al. reported that in a group of patients with a low clinical pre-test probability of PE, the risk of PE was low when the D-dimer was <1.0 μg/ml. In that study, the number of patients with cancer was 187 out of 2017 (9.0%). Of the 187 participants, 142 (78.9%), 27 (14.4%) and 18 (9.6%) were categorized in the low, moderate and high pre-test probability groups, respectively. In that study, no patient in the low pre-test probability group with D-dimer levels <1.0 μg/ml developed PE (12). In this study, the cut-off value for D-dimer for predicting concurrent VTE at the time of initial diagnosis was 3.3 μg/ml, which was high compared to that in the aforementioned studies. Since our study included only patients with unresectable advanced lung cancer, it likely included several patients with a relatively high risk of VTE. We believe that the cut-off value for D-dimer may differ not only between patients with and without cancer but also between different types of cancer. Further studies are required to confirm these speculations.

Other prospective studies have investigated the prothrombin fragment (PF) 1 + 2 (PF1 + 2) in relation to incident VTE. An increase in PF1 + 2 was associated with a hazard ratio of 2.0–2.11 for VTE (14,15). In the Rising-VTE/NEJ037 study, the odds ratio for PF1 + 2 ≥ 325 pmol/L in Japanese patients with advanced lung cancer was 2.155 (95% CI, 1.313–3.535, P value = 0.002) (16). The choice of coagulation marker may also be important, and D-dimer is a key coagulation marker in Japan because it can be easily measured.

Fibrin/fibrinogen degradation products (FDP) are a fibrinolytic marker similar to D-dimer. In a single-center prospective observational study on VTE complications among 158 Taiwanese lung cancer patients, 27 (17.1%) had VTE. The study calculated cutoff values for FDP and D-dimer, and when the cutoff value for FDP was 2.84 mg/L, the AUC was 0.869, sensitivity was 0.963 and specificity was 0.457. Furthermore, when the cutoff value for D-dimer was 0.77 μg/ml, the AUC was 0.860, sensitivity was 0.926 and specificity was 0.403 (17). These findings differ from those in the present study, possibly due to differences in race and VTE complication rates, meaning that various cutoff values may be needed for each race and cancer type.

In the Rising-VTE/NEJ037 study, the median overall survival in the group with VTE treated with edoxaban and the group without VTE was 24.0 months (95% CI, 16.8–not estimable) and 19.2 months (95% CI, 16.8–21.6), respectively (P = 0.793) (5). The prognosis of patients with lung cancer with VTE did not differ from that of patients without VTE, provided that appropriate anticoagulation therapy was used. Because VTE is often asymptomatic, it is clinically important to establish a cut-off value for D-dimer levels.

Limitations

This study has some limitations. First, this study targeted only Japanese patients with advanced lung cancer; the cut-off values may differ according to race. Second, the calculation of the D-dimer cut-off value was a post hoc study using data from Rising-VTE/NEJ037. Third, although the results of this study have validity within the Rising-VTE/NEJ037 cohort, their external validity has not been proven. We are currently preparing to undertake a prospective observational study based on the results of this study to obtain external validity.

The cut-off value of the D-dimer level for predicting VTE at the time of initial diagnosis in Japanese patients with advanced lung cancer was 3.3 μg/ml. Aggressive investigations to detect VTE are important in patients with D-dimer levels ≥3.3 μg/ml at the time of initial diagnosis.

Acknowledgements

We thank all our patients, their families, and all site investigators for their cooperation. We also thank Dr Hiroyuki Kuroda and Dr Megumi Nakamura for constituting the Image Assessment Committee, and Dr Takashi Yoshioka and Dr Teruhisa Azuma for constituting the Safety Monitoring Committee.

Funding

This work was funded by Daiichi Sankyo Co., Ltd [LIXMD-15003]. The sponsor had no role in the study design, conduct of the study, data collection and analysis, decision to publish, or preparation of the manuscript.

Disclosures

Dr Tsubata was supported by Daiichi Sankyo Co., Ltd. All other authors have no relationships relevant to the contents of this paper to disclose.

Conflict of interest statement

None declared.
==== Refs
References

1. Key  NS, Khorana  AA, Kuderer  NM, et al.  Venous thromboembolism prophylaxis and treatment in patients with cancer: ASCO clinical practice guideline update. J Clin Oncol  2020;38 :496–520. 10.1200/JCO.19.01461.31381464
2. Khorana  AA, Francis  CW, Culakova  E, Kuderer  NM, Lyman  GH. Thromboembolism is a leading cause of death in cancer patients receiving outpatient chemotherapy. J Thromb Haemost  2007;5 :632–4. 10.1111/j.1538-7836.2007.02374.x.17319909
3. Moore  RA, Adel  N, Riedel  E, et al.  High incidence of thromboembolic events in patients treated with cisplatin-based chemotherapy: a large retrospective analysis. J Clin Oncol  2011;29 :3466–73. 10.1200/JCO.2011.35.5669.21810688
4. Fernandes  CJ, Morinaga  LTK, Alves  JL, et al.  Cancer-associated thrombosis: the when, how and why. Eur Respir Rev  2019;28 :180119. 10.1183/16000617.0119-2018.30918022
5. Tsubata  Y, Hotta  T, Hamai  K, et al.  Incidence of venous thromboembolism in advanced lung cancer and efficacy and safety of direct oral anticoagulants: a multicenter, prospective, observational study (Rising-VTE/NEJ037 study). Ther Adv Med Oncol  2022;14 :175883592211101. 10.1177/17588359221110171.
6. Hu  Q, Wu  Y, Tang  J, et al.  Expression of polycystins and fibrocystin on primary cilia of lung cells 1. Biochem Cell Biol  2014;92 :547–54. 10.1139/bcb-2014-0062.25367197
7. Kearon  C . Diagnosis of suspected venous thromboembolism. Hematology Am Soc Hematol Educ Program  2016;2016 :397–403. 10.1182/asheducation-2016.1.397.27913507
8. Raja  AS, Greenberg  JO, Qaseem  A, et al.  Evaluation of patients with suspected acute pulmonary embolism: best practice advice from the Clinical Guidelines Committee of the American College of Physicians. Ann Intern Med  2015;163 :701–11. 10.7326/M14-1772.26414967
9. Lim  W, Le Gal  G, Bates  SM, et al.  American Society of Hematology 2018 guidelines for management of venous thromboembolism: diagnosis of venous thromboembolism. Blood Adv  2018;2 :3226–56. 10.1182/bloodadvances.2018024828.30482764
10. Kolomansky  A, Hoffman  R, Sarig  G, Brenner  B, Haim  N. Prospective evaluation of patients hospitalized with venous thromboembolism: comparison between cancer and non-cancer patients. Isr Med Assoc J  2006;8 :848–52.17214102
11. Kanda  Y . Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transplant  2013;48 :452–8. 10.1038/bmt.2012.244.23208313
12. Goldstraw  P, Crowley  J, Chansky  K, et al.  The IASLC lung cancer staging project: proposals for the revision of the TNM stage groupings in the forthcoming (seventh) edition of the TNM classification of malignant tumours. J Thorac Oncol  2007;2 :706–14. 10.1097/JTO.0b013e31812f3c1a.17762336
13. Kearon  C, de  Wit  K, Parpia  S, et al.  Diagnosis of pulmonary embolism with d-dimer adjusted to clinical probability. N Engl J Med  2019;381 :2125–34. 10.1056/NEJMoa1909159.31774957
14. Ay  C, Vormittag  R, Dunkler  D, et al.  D-dimer and prothrombin fragment 1 + 2 predict venous thromboembolism in patients with cancer: results from the Vienna cancer and thrombosis study. J Clin Oncol  2009;27 :4124–9. 10.1200/JCO.2008.21.7752.19636003
15. Reitter  EM, Kaider  A, Ay  C, et al.  Longitudinal analysis of hemostasis biomarkers in cancer patients during antitumor treatment. J Thromb Haemost  2016;14 :294–305. 10.1111/jth.13218.26662117
16. Tsubata  Y, Hotta  T, Hamai  K, et al.  A new risk-assessment tool for venous thromboembolism in advanced lung cancer: a prospective, observational study. J Hematol Oncol  2022;15 :40. 10.1186/s13045-022-01259-7.35379309
17. Di  W, Xu  H, Ling  C, Xue  T. Early identification of lung cancer patients with venous thromboembolism: development and validation of a risk prediction model. Thromb J  2023;21 :95. 10.1186/s12959-023-00544-w.37710256
