Risk Factors and Clinical Significance of D-Dimer in the Development of Postoperative Venous Thrombosis in Patients with Lung Tumor.

BACKGROUND: The incidence of venous thromboembolism (VTE) is higher in patients with lung cancer. The aim of this study was to investigate the risk factors associated with postoperative VTE and explore the VTE predication capacity of D-dimer kinetics. PATIENTS AND METHODS: Six hundred patients who had lung tumor surgery were analyzed retrospectively between January 2018 and August 2019, and venous ultrasound imaging and D-dimer examination before and after surgery were recommended to all operative patients. Of these 600 patients, 523 patients had venous thromboembolism after surgery, and 77 patients had not found. The general clinical data, postoperative prophylactic anticoagulant therapy, early systemic thromboprophylaxis, 50% increment of D-dimer, 100% increment of D-dimer, and perioperative (preoperative and days 1, 3, and 5 after surgery) D-dimer levels were collected. Logistic regression analysis was used to analyze the independent risk factors of postoperative VTE.
RESULTS: VTE developed in 77 (12.8%) patients. In a univariate analysis, age, surgical approach, tumor size, histology, postoperative preventive anticoagulation, postoperative limb compression therapy, postoperative hemostasis, duration of operation, early systemic thromboprophylaxis, 100% increment of D-dimer, preoperative and postoperative D-dimer level, intraoperative blood loss, and time spent in the hospital were significantly different between the thrombus group and nonthrombus group (P < 0.05). Multivariate analysis showed that age >60 years (P = 0.006) and D-dimer level on 5 days after surgery (P = 0.000) were significant independent risk factors for VTE. Postoperative D-dimer was significantly higher than the preoperative level (P < 0.001). Postoperative D-dimer level was significantly different between benign and malignant tumor groups (P < 0.05) and between the thrombus group and nonthrombus group (P < 0.001). Preventive anticoagulation and limb compression therapy starting from the first day after surgery was statistically significant between the thrombus group and the nonthrombus group (P < 0.05).
CONCLUSION: Continuous detection of D-dimer level after pulmonary tumor surgery combined with thrombotic-related risk factors can better evaluate the occurrence of VTE. Preventive anticoagulant therapy and limb compression therapy starting from the first day after surgery can effectively reduce the incidence of VTE.

Introduction

Venous thromboembolism (VTE) mainly includes deep vein thrombosis (DVT) and pulmonary embolism (PE). Twenty percent of VTE deaths are associated with cancer, and patients with cancer and VTE have 3 times the risk of death as those without VTE.1,2 The incidence of VTE in lung cancer is 7% to 13%, which is the highest incidence of VTE in malignant tumors.3–5 In addition, surgery and chemotherapy can directly increase the risk of VTE.6,7 Plasma D-dimer is a hydrolytic product of fibrin monomers. Elevated D-dimer level often indicates hypercoagulability and secondary fibrinolysis hyperactivity in the blood. Therefore, D-dimer level can be used as an additional test for VTE. However, the blood of patients with lung cancer is often in a hypercoagulable state and activates a partial coagulation cascade, which results in a naturally high D-dimer level;8 therefore, defining the D-dimer threshold as 500 ng/mL may not accurately assess the risk of VTE. In addition, domestic and foreign guidelines for VTE suggest that anticoagulant therapy such as low-molecular-weight heparin (LMWH) should be recommended for patients with diagnosed or suspected tumors without anticoagulant contraindications.9,10 Nevertheless, there was no consistent conclusion on the type, dose, and starting time of preventive anticoagulant therapy for patients after surgery to excise lung tumors. Hence, this study aimed to explore the changes in D-dimer level in patients undergoing lung surgery and how to manage preventive anticoagulant therapy after lung tumor surgery.

Patients and Methods

Clinical Data

This study retrospectively analyzed the data of 675 patients who had lung tumor surgery during the period from January 2018 to August 2019 at the Department of Lung Cancer Surgery, Tianjin Medical University General Hospital. Patients without a history of venous thrombosis, chemotherapy, and radiotherapy were included in the study. Patients with the following history were excluded from the study: (1) anticoagulant drugs within 1 month before entering the hospital; (2) blood transfusion; (3) surgery or trauma within 3 months of admission; or (4) VTE. Seventy-five patients were excluded because they did not meet the inclusion criteria. Finally, 600 eligible patients were included (Figure 1). There were 135 cases of benign tumor, including 48 cases of inflammatory pseudotumor (35.6%), 25 cases of tuberculoma (18.5%), 10 cases of hamartoma (7.4%), and 52 cases of other benign tumors (38.5%). There were 465 patients with malignant tumor, including 314 (67.5%) with adenocarcinoma, 98 (21.1%) with squamous cell carcinoma, 15 (3.2%) with small cell carcinoma, 7 (1.5%) with large cell carcinoma, and 31 (6.7%) with other malignant tumors. Table 1 shows the patient’s clinical data. The collected information included sex, age, smoking history, hypertension history, diabetes history, surgical approach, duration of operation, intraoperative blood loss, time spent in the hospital, histology, tumor size, pathological stages, postoperative preventive anticoagulation (as-needed subcutaneous injections of nadroparin calcium starting the first day after surgery), postoperative limb compression therapy (venous thrombosis prevention device for double lower limb compression massage for 15 minutes from the first day after surgery, 2 times a day), postoperative hemostasis, early systemic thromboprophylaxis (prophylactic anticoagulation and limb compression therapy began on the first day after surgery), 50% and 100% increment of D-dimer (the D-dimer level on the first day after surgery increased by 50% and 100% compared with the preoperative period), and D-dimer levels measured preoperatively and on days 1, 3, and 5 after surgery.

Figure 1

Study flow diagram. VTE, venous thromboembolism; MCVT, muscular calf vein thrombosis; DVT, deep venous thrombosis.

Table 1

Clinical Characteristics of Patients

Category	n (%)
Age (y)
≤60	237(39.5)
>60	363(60.5)
Sex
Male	339(56.5)
Female	261(43.5)
Smoking history
Ever	258(43.0)
Never	342(57.0)
Hypertension history
Yes	210(35.0)
No	390(65.0)
Diabetes history
Yes	77(12.8)
No	523(87.2)
Surgical approach
Thoracoscopy	435(72.5)
Thoracotomy	165(27.5)
Tumor size (cm)
<3	379(63.2)
≥3	221(36.8)
Histology
Benign tumor	135(22.5)
Malignant tumor	465(77.5)
Postoperative anticoagulation
Yes	220(36.7)
No	380(63.3)
Limb compression therapy
Yes	489(81.5)
No	111(18.5)
Postoperative hemostasis
Yes	91(15.2)
No	509(84.8)
Duration of operation (h)
<3	342(57.0)
≥3	258(43.0)
Pathological stages
I+II	320(69.6)
III+IV	140(30.4)
VTE
Yes	77(12.8)
No	523(87.2)

Abbreviation: VTE, venous thromboembolism.

Diagnosis of VTE

Venous ultrasound imaging was used to confirm DVT events and was recommended for all operative patients. PE events were confirmed by computed tomography pulmonary angiogram. This study was conducted in accordance with the Declaration of Helsinki. The study protocol was reviewed and approved by the institutional review boards of the participating institutions.

Statistical Analysis

Statistical analyses were conducted with SPSS version 22.0. Continuous variables were presented as mean and standard deviation or median and percentile 25 (p25) and percentile 75 (p75), and categorical variables as absolute and relative percentages. Categorical variables were analyzed with the chi-squared test. Continuous variables were compared using the t test, the nonparametric Mann–Whitney U-test, or the Friedman test. The Bonferroni method was used to correct for multiple comparisons. The independent risk factors of postoperative VTE were analyzed by binary logistic regression. The level of statistical significance was set at P < 0.05.

Results

Clinical Features and Risk Factors of VTE

Among the 600 patients, 77 patients had lower extremity venous thrombosis, including 73 patients (12.3%) with lower extremity muscular calf vein thrombosis (MCVT) and 4 patients (0.6%) with proximal DVT. There were not any death cases due to venous thrombosis of the lower extremities or PE.

Tables 2 and 3 showed that sex, smoking history, hypertension, diabetes mellitus, 50% increment of D-dimer, and pathological stages have no significant difference (P > 0.05), but the age, surgical approach, tumor size, histology, postoperative anticoagulation, limb compression therapy, postoperative hemostasis, duration of operation, preoperative and days 1, 3, and 5 D-dimer level after surgery, intraoperative blood loss, early systemic thromboprophylaxis, 100% increment of D-dimer, and time spent in the hospital between the thrombus group and nonthrombus group have a significant difference (P < 0.05). In order to eliminate the confounding factors, the risk factors with statistical significance in a single-factor analysis were analyzed by binary logistic regression analysis and the results showed that age and D-dimer on day 5 after surgery were independent risk factors for postoperative lower extremity venous thrombosis (P < 0.05) (Table 4). The incidence of lower extremity venous thrombosis at age ≤ 60 years was 0.388 times that at age > 60 years (odds ratio [OR]=0.388; P = 0.006), and the risk of venous thrombosis in the lower extremity was doubled for every increased 1 ng/mL of D-dimer number on day 5 after surgery (OR=1.001; P = 0.000).

Table 2

Univariate Analysis of Categorical Data of Lower Extremity Venous Thrombosis in Postoperative Patients

Risk Factors		VTE(+)(n=77)	VTE(−)(n=523)	χ2	P
Gender	Male(%)	39(11.5)	300(88.5)	1.230	0.267
	Female(%)	38(14.6)	223(85.4)
Age(yr)	≤60(%)	15(6.3)	222(93.7)	14.815	0.000*
	>60(%)	62(17.1)	301(82.9)
Smoking history	No(%)	47(13.7)	295(86.3)	0.588	0.443
	Yes(%)	30(11.6)	228(88.4)
Hypertension history	No(%)	43(11.0)	347(89.0)	3.255	0.071
	Yes(%)	34(16.2)	176(83.8)
Diabetes history	No(%)	65(12.4)	458(87.6)	0.598	0.439
	Yes(%)	12(15.6)	65(84.4)
Surgical approach	Thoracoscopy(%)	43(9.9)	392(90.1)	12.291	0.000*
	Thoracotomy(%)	34(20.6)	131(79.4)
Tumor size	<3cm(%)	38(10.0)	341(90.0)	7.247	0.007*
	≥3cm(%)	39(17.6)	182(82.4)
Histology	Benign tumor(%)	10(7.4)	125(92.6)	4.584	0.032*
	Malignant tumor(%)	67(14.4)	398(85.6)
Postoperative anticoagulationa	No(%)	58(15.3)	322(84.7)	5.470	0.019*
	Yes(%)	19(8.6)	201(91.4)
Limb compression therapy	No(%)	23(20.7)	88(79.3)	7.574	0.006*
	Yes(%)	54(11.0)	435(89.0)
Postoperative hemostasis	No(%)	56(11.0)	453(89.0)	10.062	0.002*
	Yes(%)	21(23.1)	70(76.9)
Duration of operation	<3h(%)	32(9.4)	310(90.6)	8.594	0.003*
	≥3h(%)	45(17.4)	213(82.6)
Pathological stages	I+II(%)	47(14.7)	273(85.3)	0.269	0.604
	III+IV(%)	18(12.9)	122(87.1)
Early systemic thromboprophylaxisb	No(%)	61(14.8)	352(85.2)	4.443	0.035*
	Yes(%)	16(8.6)	171(91.4)
50% increment of D-dimerc	No(%)	20(11.9)	148(88.1)	0.180	0.671
	Yes(%)	57(13.2)	375(86.8)
100% increment of D-dimerd	No(%)	27(9.3)	262(90.7)	6.074	0.014*
	Yes(%)	50(16.1)	261(83.9)

Notes: aAs-needed subcutaneous injections of nadroparin calcium starting the first day after surgery. bProphylactic anticoagulation and limb compression combined therapy began on the first day after surgery. cThe D-dimer level on the first day after surgery increased by 50% compared with the preoperative period. dThe D-dimer level on the first day after surgery increased by 100% compared with the preoperative period. *Indicates statistical significance, P<.05.

Abbreviation: VTE, venous thromboembolism.

Table 3

Univariate Analysis of Measurement Data of Lower Extremity Venous Thrombosis in Postoperative Patients

Risk Factors		VTE (+)	VTE (−)	P
D-dimer (ng/mL)	Before surgery	482(342,944)	361(256,573)	0.000*
	Day 1 after surgery	1717(922,2675)	771(539,1105)	0.000*
	Day 3 after surgery	1649(1024,2544)	680(485,1036)	0.000*
	Day 5 after surgery	2293(1287,3811)	946(684,1348)	0.000*
Blood loss (mL)		50(30,100)	50(20,50)	0.000*
Time in hospital (d)		16(15,23)	15(13,19)	0.001*

Note: *Indicates statistical significance, P<.05.

Abbreviation: VTE, venous thromboembolism.

Table 4

Two-Class Logistic Regression Analysis of Lower Extremity Venous Thrombosis in Postoperative Patients

Risk Factors	B	Standard Error	Wald	P	OR	95% CI
						Lower	Upper
Age	−0.948	0.346	7.511	0.006*	0.388	0.197	0.763
D-dimer level day 5 after surgery	0.001	0	29.358	0.000*	1.001	1.001	1.001

Note: *Indicates statistical significance, P<.05.

Abbreviations: OR, odds ratio. CI, confidence interval.

Characteristics of D-Dimer During the Perioperative Period

The level of D-dimer after surgery was significantly higher than that before surgery, and it was significantly different not only in the benign tumor group but also in the malignant tumor group (P < 0.001) (Figure 2). The median preoperative D-dimer level was 240 ng/mL in the benign tumor group and 387 ng/mL in the malignant tumor group, and there was no statistical difference between the 2 groups (P = 0.097). However, the difference in postoperative D-dimer level between the benign and malignant group was statistically significant (P < 0.05) (Figure 3). The comparison of D-dimer level between the preoperative and postoperative different time points was statistically significant not only in the thrombus group but also in the nonthrombus group (P < 0.001), and the difference in D-dimer level between the thrombus group and the nonthrombus group was also statistically significant (Figure 4).

Figure 2

Comparison of D-dimer level at different time points in the benign tumor group and malignant tumor group. ***P<.001, the difference between preoperative and postoperative D-dimer levels is statistically significant in the benign tumor group. ###P<.001, the difference in D-dimer level is statistically significant in the malignant tumor group.

Figure 3

Comparison of D-dimer level at different time points between benign tumor group and malignant tumor group. NS, the difference was not statistically significant. *Indicates statistical significance, P<.05. **Indicates statistical significance, P<.01.

Figure 4

Comparison of D-dimer level at different time points in the thrombus group and nonthrombus group. ***P<.001, the difference between preoperative and postoperative D-dimer levels is statistically significant in the thrombus group and nonthrombus group; ###P<.001, the difference in D-dimer level is statistically significant between the thrombus group and nonthrombus group.

Prophylactic Management of VTE

In the analysis of prophylactic treatment of VTE, subcutaneous injection with different doses of nadroparin calcium was started at different times. It was found that the preventive anticoagulant injection started on the first day after the operation showed a significant difference between the thrombus group and the nonthrombus group (P < 0.05). However, when nadroparin calcium was given as preventive anticoagulant therapy on the second day after the operation, no statistical significance was found between the thrombus group and the nonthrombus group, even if the patient was given regular (0.4 mL/day) subcutaneous injections or at a larger dose and frequency (Table 5). Besides, our study showed that there were significant differences between early systemic thromboprophylaxis and regular anticoagulation day 1 after surgery (Nadroparin calcium was injected regularly at least 0.4 mL daily.), limb compression therapy, and non-treatment in the comparison between thrombus group and nonthrombus group (P=.037) ().

Table 5

Analysis of Postoperative Venous Thrombosis of Lower Extremity Under Different Anticoagulation Conditions

Anticoagulant Condition		VTE (+)	VTE (−)	P
Anticoagulation day 1 after surgerya	Yes(%)	19(8.6)	201(91.4)	0.019*
	No(%)	58(15.3)	322(84.7)
Anticoagulation day 1 after surgeryb	Yes(%)	17(8.6)	181(91.4)	0.029*
	No(%)	60(14.9)	342(85.1)
Anticoagulation day 2 after surgeryc	Yes(%)	30(11.2)	239(88.8)	0.267
	No(%)	47(14.2)	284(85.8)
Anticoagulation day 2 after surgeryd	Yes(%)	25(10.8)	206(89.2)	0.244
	No(%)	52(14.1)	317(85.9)

Notes: aAs-needed subcutaneous injections of nadroparin calcium starting the first day after surgery. bFrom the first day after surgery, nadroparin calcium was injected subcutaneously on a regular basis (at least 0.4 mL daily). cAs-needed injections of nadroparin calcium starting the second day after surgery. dFrom the second day after surgery, nadroparin calcium was injected subcutaneously on a regular basis (at least 0.4 mL daily). *Indicates statistical significance, P<.05.

Abbreviation: VTE, venous thromboembolism.

Discussion

VTE is one of the leading causes of morbidity and mortality in patients with cancer. Approximately one-fifth of patients with VTE are diagnosed with malignant tumors simultaneously,1,10,11 and VTE is the most common cause of death after cancer surgery.12 Compared with other types of tumors, the risk of VTE in lung cancer is very high. If we consider the high incidence of lung cancer, lung cancer may be one of the cancers with the highest incidence of cancer-related thrombosis.13 The overall incidence of VTE after surgery in this study was 12.8%, of which benign and malignant tumors were 7.4% and 14.4%, respectively. This is consistent with previous studies (Table 6) and it again shows that the incidence of postoperative VTE is higher than other tumors and highlights the importance of postoperative anticoagulation.

Table 6

Incidence of VTE After Surgery in Different Disease Categories

Study	Total Surgical Patients	Disease Categories	Incidence of Postoperative VTE
Cui et al14	339	Pulmonary malignancy	11.5%
Wang et al15	249	Pulmonary malignancy	14.5%
Tian et al16	52	Pulmonary malignancy	23.1%
	59	Pulmonary benign tumor	10.2%
Song et al17	147	Pulmonary malignancy	15.0%
	115	Pulmonary benign disease	7.0%
Song et al18	285	Pulmonary malignancy	16.4%
		Pulmonary benign tumor	7.5%
Shi et al19	1133	Urological malignant tumor	2.4%
Yoshioka et al20	72	Malignant spine tumor	4.2%
		Benign spine surgery	4.2%
Yang et al21	3645	Colorectal cancer	0.9%
Kim et al22	375	Gastric Cancer	2.4%
Dar et al23	88	Abdominal malignancy	9.0%
Hennessey et al24	93,663	Head and neck cancer	2.0%
Kimmell et al25	3098	Intracranial tumor	5.0%
		Benign intracranial disease	1.9%

Abbreviation: VTE, venous thromboembolism.

Prior studies show that platinum-containing chemotherapeutic drugs,26,27 interleukin 6, tumor necrosis factor,28 antiangiogenic agents (bevacizumab),29 platelet count,30 age, tumor pathology type (adenocarcinoma),31 time spend in the hospital,32 obesity, and surgery can increase the risk of VTE.33,34 Other studies show that the incidence of VTE is the highest within 1 month after operation, after which the risk is gradually reduced.35,36 The risk of venous thrombosis is 4.2% in patients undergoing lung cancer surgery before discharge, and 77% of patients with VTE are diagnosed before discharge.36,37 Nevertheless, there are few studies on the incidence of postoperative thrombus in patients undergoing lung tumor surgery, which is one of the advantages of this study. We found that the incidence of VTE in patients with thoracotomy, operation time > 3 hours, more blood loss during surgery, and longer time spent in the hospital was higher than that in the control group. Moreover, age and D-dimer on day 5 after surgery were independent risk factors for lower limb venous thrombosis. We found that the incidence of postoperative thrombus in patients with age > 60 years (17.1%) was significantly higher than that in patients with age ≤ 60 years (6.3%). Several studies also show that age is a risk factor for thrombosis.38,39 In addition, the D-dimer level on day 5 after surgery was significantly higher than the other time points, and the most significant difference was noted between the thrombus group and the nonthrombus group. Therefore, it is necessary to perform a timely check of coagulation function on day 5 after surgery.

D-dimer is positive for the prediction of thrombosis, but previous studies show that the current threshold does not properly assess the occurrence of thrombosis, and it is recommended to adjust the threshold to 3500 ng/mL or 2000 ng/mL.40–42 D-dimer level increases first and then decreases in the first 10 days after surgery.40,42 We found that the median postoperative D-dimer level in the thrombus group, the nonthrombus group, the benign tumor group, and the malignant tumor group were all > 500 ng/mL. Therefore, whether to define the critical level of D-dimer as 500 ng/mL after tumor surgery needs further consideration and research. In addition, when comparing the doubled D-dimer level on the first day after surgery with pre-surgery, the occurrence of venous thrombosis should be alerted because our data showed that the incidence of thrombosis was significantly increased. Postoperative plasma D-dimer was significantly different not only in the thrombus and nonthrombus groups but also in the benign and malignant tumor groups. Therefore, we suggest that regular review of D-dimer levels is necessary for postoperative patients, especially in the first week after surgery.

Because of the high incidence of lung cancer and VTE, we need to consider whether the preventive anticoagulant therapy should be adjusted accordingly. Currently, preventive anticoagulant therapy after lung cancer surgery is mainly based on clinical consensus and lack of relevant evidence. In addition, postoperative patients have a potential risk of bleeding. Therefore, how to balance the risk of thrombosis and bleeding after surgery needs further study. Previous studies have found that about 7% of PE is caused by intermuscular veins, and suggest that patients who have symptomatic MCVT need to be treated with an anticoagulant therapeutic dose lasting at least 15 to 30 days, and it is necessary to prolong the anticoagulation time to prevent the thrombus from extending to the deep vein when related risk factors are present.43 Previous study reveals that the majority of postoperative VTE is lower extremity intermuscular venous thromboses, which is consistent with this study (94.8%), and shows that up to 15% of the distal VTEs in the first 2 weeks after surgery subsequently progress and affect the proximal vein.42 Thrombosis prevention programs include anticoagulant drugs, mechanical measures (intermittent compression therapy and elastic stockings), and early active activities. These programs can effectively reduce the incidence of VTE after lung surgery and our results have confirmed some of them.44 Further, LMWH for the preventive treatment of VTE does not significantly increase risk of hemorrhage.45 Our findings are in line with the American guidelines which recommend early initiation of systemic thromboprophylaxis for at least 7 to 10 days after tumor surgery. Moreover, their findings suggest an added benefit of mechanical thromboprophylaxis in this population.46 Besides, early systemic thromboprophylaxis may be better than regular anticoagulation day 1 after surgery, limb compression therapy in preventing thrombosis.

Hence, prophylactic anticoagulant therapy and limb compression therapy from the first day after surgery, D-dimer examination on day 5 after surgery, and thrombus-related risk factor analysis can well evaluate the risk of thrombus and help clinicians decide whether to continue anticoagulant therapy.

Study Limitations

The results of this study were encouraging, but the limitations were not negligible. The study did not have regular follow-up for VTE progression and overall survival of patients postoperatively. Because of the low overall incidence of VTE in the population, comparative studies of postoperative anticoagulation for other types of tumors should be performed in subsequent studies. In addition, the neutrophil-lymphocyte ratio is considered to be a potential biomarker for predicting efficacy and prognosis of anticoagulant therapy in patients with lung cancer,47 so relevant research can be conducted in future studies.

Conclusions

In conclusion, VTE after pulmonary tumor surgery should cause widespread concern and attention. In addition, D-dimer level and color Doppler ultrasonography of lower extremity veins should be reviewed regularly. After the anticoagulation contraindications are excluded, prophylactic anticoagulant therapy and mechanical thromboprophylaxis should be given as soon as possible according to the patient’s condition. When lower extremity intermuscular venous thrombosis occurs, more aggressive treatment should be taken to protect the health of the patient from progression of this condition.