Abstract
Background
Non-small cell lung cancer (NSCLC) is associated with a high risk of venous thromboembolism (VTE). However, data on specific risk factors for VTE in patients with advanced NSCLC remain limited.
Methods and Results
Using a Japanese nationwide administrative database, we analyzed 20,206 patients aged ≥18 years with advanced NSCLC who received first-line chemotherapy between January 2016 and January 2023. VTE events were identified through International Classification of Diseases, Tenth Revision codes and imaging studies. Risk factors were evaluated using Cox proportional hazards models with time-dependent covariates. The cumulative incidence of VTE was 4.2% and 6.1% at 365 and 730 days after the first date of chemotherapy for NSCLC, respectively. Several significant risk factors for VTE were identified, including female sex (hazard ratio [HR] 1.374; 95% confidence interval [CI] 1.157–1.631), higher body mass index (HR 1.029 per 1-kg/m2 increase; 95% CI 1.009–1.048), previous VTE (HR 2.707; 95% CI 1.907–3.843), platinum-based chemotherapy (HR 1.217; 95% CI 1.051–1.410), anti-vascular endothelial growth factor agent (HR 1.763; 95% CI 1.458–2.132), heart failure (HR 1.677; 95% CI 1.432–1.965), and stroke/transient ischemic attack (HR 1.296; 95% CI 1.055–1.593).
Conclusions
This large-scale study identified several significant risk factors for VTE in patients with advanced NSCLC. The findings suggest the need for risk-stratified monitoring and prophylactic strategies to reduce VTE-related complications in high-risk patients.
Key Words: Nationwide database, Non-small cell lung cancer, Risk factor, Venous thromboembolism
Central Figure.
Non-small cell lung cancer (NSCLC) accounts for approximately 85% of all lung cancer cases globally, making it the predominant form of lung cancer.1 The prognosis for advanced NSCLC is particularly poor, with a 5-year survival rate of <15%.2 As the incidence of NSCLC continues to rise, particularly in developed countries, managing its associated complications becomes increasingly critical. Venous thromboembolism (VTE), which includes deep vein thrombosis (DVT) and pulmonary embolism (PE), is a significant and potentially life-threatening complication in patients with cancer.3 The incidence of VTE is estimated to be 4- to 7-fold higher in patients with cancer than in the general population, contributing to increased morbidity and mortality.4 In patients with advanced NSCLC, the presence of VTE further complicates their already precarious health status, necessitating early identification and management of VTE risk factors. Identifying risk factors for VTE in patients with cancer is crucial for improving patient outcomes. Early identification of VTE risk factors allows for the implementation of tailored prophylactic measures, potentially reducing the incidence of VTE and improving survival rates. Despite the importance of this issue, studies focusing specifically on patients with advanced NSCLC are lacking, particularly studies using large-scale, comprehensive datasets.
The aim of this study was to determine the risk factors associated with VTE in patients diagnosed with advanced NSCLC using a large nationwide administrative database in Japan.
Methods
Data Source
This study used administrative claim data from advanced treatment centers in Japan using the database maintained by Medical Data Vision Co., Ltd. (MDV; Tokyo, Japan). The nationwide MDV database includes information on over 50 million patients, representing 30% of all Diagnostic Procedure Combination/Per Diem Payment System (DPC/PDPS) hospitals in Japan. This system determines provider compensation at a fixed daily rate based on diagnostic groups, combining diagnoses and procedures performed during hospitalization.5 Most large designated acute care hospitals in Japan use this system. The MDV dataset contains routinely collected clinical information, such as basic patient characteristics (age, sex, weight, and height), hospitalization dates, diagnoses based on International Classification of Diseases, Tenth Revision (ICD-10) codes, survival status, and detailed medical practices. The database integrates inpatient and outpatient data, enabling longitudinal follow-up of patients throughout their treatment course at participating institutions. The MDV database contains information for patients diagnosed with lung cancer (ICD-10 code C34) between April 1, 2008, and January 31, 2023, and has been used previously in several clinical studies investigating lung cancer in Japan.6,7
Study Population
The study included patients aged ≥18 years who received chemotherapy for newly diagnosed advanced NSCLC between January 2016 and January 2023. Pembrolizumab has been covered by insurance as first-line chemotherapy in patients with NSCLC since December 2016. To identify patients who received first-line chemotherapy following a lung cancer diagnosis and to exclude patients with VTE, we included patients with at least a 3-month interval from their database entry to the date of their first lung cancer diagnosis. The index date was defined as the first date of chemotherapy. To identify patients with advanced NSCLC, we excluded patients who received chemoradiation in the 7 days prior to the index date and those who received adjuvant chemotherapy in the 60 days prior to the index date, following methods similar to those used in a previous study.8 In addition, to accurately determine the incidence of VTE, patients diagnosed with VTE, using oral anticoagulants, or receiving parenteral anticoagulation in the 90 days prior to the index date were excluded from the study. Further details of the exclusion criteria are provided in Supplementary Methods 1.
Identification of VTE
The incidence of VTE was defined using a combination of diagnoses based on ICD-10 codes for DVT and PE, along with imaging modalities, as described previously.9 Imaging for DVT and PE included computed tomography, ultrasonography of the leg veins for DVT, and computed tomography and a ventilation-perfusion lung scan for PE.10 The date of VTE onset was defined as the date of the imaging modality if the diagnosis of DVT or PE was made in the same month or the following month after imaging. The codes used for the identification of VTE are summarized in Supplementary Table 1.
Demographic and Clinical Information
Data on the following demographic and clinical characteristics were obtained: age, sex, body mass index (BMI), smoking status, NSCLC histology, Barthel index,11 previous VTE, chemotherapy for NSCLC (platinum-based chemotherapy, anti-vascular endothelial growth factor [VEGF] agents, immune checkpoint inhibitors [ICIs], driver mutation-targeted treatments), underlying diseases (chronic obstructive pulmonary disease [COPD], heart failure, stroke or transient ischemic attack [TIA], varicose veins of the lower extremities, peptic ulcer disease, hypertension, dyslipidemia, diabetes, chronic kidney disease [CKD], liver dysfunction), medications (antiplatelet agents, non-steroidal anti-inflammatory drugs [NSAIDs], H2 receptor antagonists, proton pump inhibitors, erythropoiesis stimulating agents), central venous catheter, and the calendar date of treatment. A list of ICD-10 and Anatomical Therapeutic Chemical codes for the underlying diseases and medical treatments is presented in Supplementary Table 2. Details regarding the collection of demographic and clinical data are provided in Supplementary Methods 2.
Statistical Analysis
Demographic and clinical data are summarized as the median and interquartile range (IQR) for continuous variables and as numbers and percentage for categorical variables. The cumulative incidence of VTE in patients with advanced NSCLC was estimated using the Aalen–Johansen estimator, with death treated as a competing risk. We investigated potential risk factors for VTE using Cox proportional hazards regression with time-dependent covariates for patients with advanced NSCLC.12 Again, death was treated as a competing risk. Therefore, the estimated hazard ratios (HRs) correspond to cause-specific HRs, which is consistent with the objectives of this study.13 The model incorporated various potential risk factors selected based on clinical expertise and previously published literature on cancer-associated thrombosis, including time-fixed covariates (age, sex, BMI, smoking status, histology of NSCLC, Barthel index, previous VTE, and date of chemotherapy) and time-dependent covariates (the use of platinum-based chemotherapy, anti-VEGF agents, ICIs, and driver mutation-targeted treatments; COPD, heart failure, stroke or TIA, varicose veins of lower extremities, peptic ulcer disease, hypertension, dyslipidemia, diabetes, CKD, liver dysfunction; the use of antiplatelet agents, NSAIDs, H2 receptor antagonists, proton pump inhibitors, and erythropoiesis stimulating agents; and the presence of a central venous catheter). A counting-process data structure was applied to all time-dependent covariates. Specifically, the exposure indicator was coded as 0 (without exposure) for all person-time prior to the date on which the first exposure was documented. Starting on that same calendar date, because the data used in this study are recorded only at the daily level, the exposure indicator switched irreversibly to 1 (with exposure) and remained 1 (with exposure) thereafter. Furthermore, to mitigate reverse causation, new diagnoses on the same day as VTE onset and the use of new medical procedures were disregarded.14 We performed additional analyses with the effects of BMI and age modeled using non-linear functions to examine their relationship with the incidence of VTE. Furthermore, to explore potential effect modification by sex, we included interaction terms (BMI × sex and age × sex) in the model. To address missing covariate data, we implemented multiple imputation using predictive mean matching to create 10 imputed datasets.15 We used a 2-sided 5% significance level for all statistical inferences. All analyses were conducted using R version 4.4.2 (R Foundation for Statistical Computing, Vienna, Austria).16
Ethical Considerations
This study was performed in accordance with the Declaration of Helsinki. Institutional Review Board/Ethics Committee approval was not required for this study because the Japanese Ethical Guidelines for Medical and Biological Research Involving Human Subjects17 do not apply to research using only anonymized data. The data used in this study were anonymized before the start of the study by MDV, and written informed consent was not required because all records were fully deidentified at the source and could not be linked back to individual patients.
Results
Study Population
The study included data from 32,872 patients (aged ≥18 years) diagnosed with NSCLC who started first-line chemotherapy between December 2016 and January 2023. Of these patients, 20,206 met the study eligibility criteria for patients with advanced NSCLC (Figure 1). The demographic and clinical characteristics of these patients are presented in Table 1. BMI and age distributions were visualized using histograms (Supplementary Figure). The median age of patients included in the study was 73 years, and more than 66.7% were male. As detailed in Table 1, 1.3% of patients had a previous VTE, 50.5% of patients received platinum-based chemotherapy, 40.2% received ICIs, 17.0% received anti-VEGF agents, and 27.8% received driver mutation-targeted therapy.
Figure 1.
Study flowchart. NSCLC, non-small cell lung cancer; VTE, venous thromboembolism.
Table 1.
Demographic and Clinical Characteristics of Patients (n=20,206)
| Age (years) | 73.00 [68.00–79.00] |
| Sex | |
| Male | 13,479 (66.7) |
| Female | 6,727 (33.3) |
| BMI (kg/m2) | 21.95 [19.71–24.30] |
| Barthel index=100 | 14,600 (88.3) |
| Smoking status (current/past smoker) | 12,762 (66.6) |
| Histology of NSCLC | |
| Adenocarcinoma | 10,673 (80.5) |
| Squamous cell carcinoma | 2,585 (19.5) |
| Previous venous thromboembolism | 253 (1.3) |
| Chemotherapy for advanced NSCLC | |
| Platinum-based chemotherapy | 10,201 (50.5) |
| Anti-VEGF agents | 3,425 (17.0) |
| Immune checkpoint inhibitors | 8,120 (40.2) |
| Driver mutation-targeted treatments | 5,622 (27.8) |
| Medication | |
| Antiplatelet agents | 3,621 (17.9) |
| NSAIDs | 11,847 (58.6) |
| H2 receptor antagonists | 6,006 (29.7) |
| Proton pump inhibitors | 13,060 (64.6) |
| Erythropoiesis stimulating agents | 75 (0.4) |
| Medical procedure | |
| Central venous catheter | 1,866 (9.2) |
| Underlying diseases | |
| Chronic obstructive pulmonary disease | 8,498 (42.1) |
| Heart failure | 4,086 (20.2) |
| Stroke or transient ischemic attack | 2,716 (13.4) |
| Varicose veins of lower extremities | 135 (0.7) |
| Peptic ulcer disease | 6,851 (33.9) |
| Hypertension | 11,633 (57.6) |
| Dyslipidemia | 6,651 (32.9) |
| Diabetes | 8,406 (41.6) |
| Chronic kidney disease | 355 (1.8) |
| Liver dysfunction | 6,655 (32.9) |
| Calendar dates of treatment | |
| January 1, 2016–December 31, 2019 | 9,625 (47.6) |
| January 1, 2020–January 1, 2023 | 10,581 (52.4) |
Data are presented as the median [interquartile range] or n (%). For BMI, the Barthel index, smoking status, and NSCLC histology, data were missing for 15.0%, 18.1%, 5.1%, and 34.4% of patients, respectively. Time-dependent covariates (use of platinum-based chemotherapy, anti-VEGF agent, immune checkpoint inhibitor use, and driver mutation-targeted treatments; chronic obstructive pulmonary disease, heart failure, stroke or transient ischemic attack, varicose veins of the lower extremities, peptic ulcer disease, hypertension, dyslipidemia, diabetes, chronic kidney disease, and liver dysfunction; the use of antiplatelet agents, NSAIDs, H2 receptor antagonists, proton pump inhibitors, and erythropoiesis stimulating agents; and placement of a central venous catheter) were summarized using information from the entire follow-up period. BMI, body mass index; NSAIDs, non-steroidal anti-inflammatory drugs; NSCLC, non-small cell lung cancer; VEGF, vascular endothelial growth factor.
Incidence of VTE
The estimated cumulative incidence of VTE in patients with advanced NSCLC is shown in Figure 2. In patients with advanced NSCLC, the incidence of VTE was 4.2% within 365 days from the index date and 6.1% within 730 days from the index date.
Figure 2.
Cumulative incidence curve for venous thromboembolism (VTE). The shaded area represents the 95% confidence interval.
Risk Factors for VTE
Cox proportional hazards regression with time-dependent covariates identified several significant risk factors for VTE in patients with advanced NSCLC (Table 2). These risk factors included female sex (HR 1.339; 95% CI 1.157–1.631), higher BMI (HR for each 1-kg/m2 increase 1.029; 95% CI 1.009–1.048), previous VTE (HR 2.707; 95% CI 1.907–3.843), platinum-based chemotherapy (HR 1.217; 95% CI 1.051–1.410), anti-VEGF agents (HR 1.763; 95% CI 1.458–2.132), heart failure (HR 1.677; 95% CI 1.432–1.965), and stroke/TIA (HR 1.296; 95% CI 1.055–1.593).
Table 2.
Risk Factors for VTE
| With VTE (n=921) |
Without VTE (n=19,285) |
Multivariable Cox regression analysis | |||
|---|---|---|---|---|---|
| HR for VTE | 95% CI | P value | |||
| Age (per 10-year increase) | 73.00 [67.00–78.00] | 73.00 [68.00–79.00] | 0.982 | 0.911–1.058 | 0.628 |
| Female (reference: male) | 353 (38.3) | 6,374 (33.1) | 1.374 | 1.157–1.631 | <0.001 |
| BMI (per 1-kg/m2 increase) | 22.52 [20.15–24.91] | 21.93 [19.68–24.27] | 1.029 | 1.009–1.048 | 0.004 |
| Barthel index=100 (reference: <100) | 703 (91.4) | 13,897 (88.1) | 1.086 | 0.857–1.378 | 0.495 |
| Smoker (reference: non-smoker) | 592 (66.3) | 12,170 (66.6) | 1.184 | 0.990–1.417 | 0.065 |
| Histology of NSCLC | |||||
| Adenocarcinoma (reference: squamous) | 554 (83.4) | 10,119 (80.3) | 1.054 | 0.853–1.301 | 0.629 |
| Previous VTE | 33 (3.6) | 220 (1.1) | 2.707 | 1.907–3.843 | <0.001 |
| Chemotherapy for NSCLC | |||||
| Platinum-based chemotherapy | 415 (45.1) | 9,786 (50.7) | 1.217 | 1.051–1.410 | 0.009 |
| Anti-VEGF agents | 147 (16.0) | 3,278 (17.0) | 1.763 | 1.458–2.132 | <0.001 |
| Immune checkpoint inhibitors | 243 (26.4) | 7,877 (40.8) | 1.161 | 0.989–1.363 | 0.069 |
| Driver mutation-targeted treatments | 144 (15.6) | 5,478 (28.4) | 1.085 | 0.887–1.328 | 0.427 |
| Medication | |||||
| Antiplatelet agents | 142 (15.4) | 3,479 (18.0) | 0.885 | 0.722–1.085 | 0.240 |
| NSAIDs | 413 (44.8) | 11,434 (59.3) | 1.069 | 0.932–1.226 | 0.344 |
| H2 receptor antagonists | 222 (24.1) | 5,784 (30.0) | 0.994 | 0.848–1.165 | 0.943 |
| Proton pump inhibitors | 443 (48.1) | 12,617 (65.4) | 1.100 | 0.957–1.263 | 0.179 |
| Erythropoiesis stimulating agents | 3 (0.3) | 72 (0.4) | 1.254 | 0.387–4.061 | 0.705 |
| Medical procedure | |||||
| Central venous catheter | 34 (3.7) | 1,832 (9.5) | 1.293 | 0.913–1.830 | 0.148 |
| Underlying diseases | |||||
| Chronic obstructive pulmonary disease | 354 (38.4) | 8,144 (42.2) | 0.981 | 0.854–1.126 | 0.780 |
| Heart failure | 222 (24.1) | 3,864 (20.0) | 1.677 | 1.432–1.965 | <0.001 |
| Stroke or transient ischemic attack | 126 (13.7) | 2,590 (13.4) | 1.296 | 1.055–1.593 | 0.014 |
| Varicose veins of lower extremities | 9 (1.0) | 126 (0.7) | 1.521 | 0.786–2.940 | 0.213 |
| Peptic ulcer disease | 296 (32.1) | 6,555 (34.0) | 1.093 | 0.946–1.262 | 0.229 |
| Hypertension | 534 (58.0) | 11,099 (57.6) | 1.056 | 0.915–1.217 | 0.458 |
| Dyslipidemia | 319 (34.6) | 6,332 (32.8) | 1.032 | 0.888–1.198 | 0.685 |
| Diabetes | 376 (40.8) | 8,030 (41.6) | 1.021 | 0.888–1.175 | 0.768 |
| Chronic kidney disease | 9 (1.0) | 346 (1.8) | 0.673 | 0.341–1.330 | 0.255 |
| Liver dysfunction | 289 (31.4) | 6,366 (33.0) | 1.014 | 0.881–1.168 | 0.845 |
| Calendar dates of treatment | |||||
| 2020–2023 (reference: 2016–2019) | 453 (49.2) | 10,128 (52.5) | 1.185 | 1.033–1.348 | 0.015 |
Unless indicated otherwise, data are presented as the median [interquartile range] or n (%). CI, confidence interval; HR, hazard ratio; VTE, venous thromboembolism. Other abbreviations as in Table 1.
We examined the relationships between VTE risk and both BMI and age using non-linear functions (Figure 3A,B). This analysis revealed a modest positive association between the incidence of VTE and increasing BMI (Figure 3A), whereas the association with age showed a decline after age 70 years (Figure 3B). Analysis of sex-specific effects revealed that at a BMI of 20 kg/m2, the HR for female vs. male patients was 1.265 (95% CI 0.978–1.638); this increased to 1.910 (95% CI 1.337–2.730) at a BMI of 30 kg/m2, indicating a stronger association between obesity and VTE risk in female patients (Figure 3C). In addition, among patients aged 30–80 years, female patients consistently showed a higher incidence of VTE than male patients (Figure 3D).
Figure 3.
Relationship between the log hazard of venous thromboembolism (VTE) and (A,C) body mass index (BMI) and (B,D) age in the study population overall (A,B) and according to sex (C,D). All relationships were non-linear. The shaded areas represent the 95% confidence intervals.
Discussion
This nationwide study analyzed data from 20,206 patients with advanced NSCLC who received first-line chemotherapy between December 2016 and January 2023. Given that VTE, including DVT and PE, significantly impacts prognosis and the survival of cancer patients, we investigated its incidence and risk factors in this population. The cumulative incidence of VTE was 4.2% and 6.1% within 365 and 730 days from the index date, respectively, confirming the significant impact of VTE in patients with advanced NSCLC. Although the incidence of VTE in this study is consistent with that of other solid tumors, VTE rates vary considerably among cancer types, with higher rates for pancreatic, gastric, and brain cancers, and lower rates for breast and prostate cancers.18,19 Previous studies on NSCLC have been limited by small sample sizes,20,21 and most existing research has focused on general cancer populations, leaving a significant knowledge gap regarding NSCLC-specific risk factors and the influence of newer treatment modalities. Current guidelines recommend VTE prophylaxis for high-risk patients with cancer,22 but the effectiveness of VTE prophylaxis specifically for patients with NSCLC remains understudied. This study, using a nationwide Japanese administrative database, addresses this knowledge gap through comprehensive analysis.
Patient-related factors emerged as important predictors of VTE risk. The identification of female sex as a risk factor, particularly in conjunction with higher BMI and across all age groups over 30 years, suggests sex-specific mechanisms. The increased VTE risk in female patients with higher BMI may be attributed to estrogen’s known effects on thrombosis.23 Medical history significantly influenced VTE risk, with previous VTE emerging as the strongest predictor of subsequent VTE events, consistent with established evidence in cancer populations.24 This finding emphasizes the importance of a thorough evaluation of a patient’s thrombotic history during their initial assessment.
Treatment-related factors also played a crucial role in VTE risk.25 Anti-VEGF agents emerged as the strongest treatment-related risk factor (HR 1.763; 95% CI 1.458–2.132), which aligns with established evidence of their prothrombotic effects through endothelial dysfunction and altered hemostatic balance.26 Anti-VEGF agents, particularly bevacizumab, disrupt normal endothelial function through multiple mechanisms, including inhibition of VEGF-mediated endothelial repair, increased vascular permeability, and promotion of a prothrombotic endothelial phenotype.26,27 Platinum-based chemotherapy also significantly increased VTE risk (HR 1.217; 95% CI 1.051–1.410), consistent with prior literature demonstrating the thrombotic potential of these agents.28,29 Platinum compounds promote thrombosis through direct endothelial toxicity, activation of the coagulation cascade, and induction of tissue factor expression.30 The combination of platinum agents with anti-VEGF therapy, commonly used in advanced NSCLC, may have synergistic thrombotic effects, emphasizing the need for heightened VTE surveillance in patients receiving such combination regimens. Interestingly, ICI therapy showed only a marginal, non-significant association with VTE risk when adjusted for other treatment-related factors (HR 1.161; 95% CI 0.989–1.363; P=0.069). Although some studies have suggested that ICI therapy may influence VTE risk through enhanced inflammatory responses and endothelial activation,31 our comprehensive analysis indicates that anti-VEGF agents and platinum-based chemotherapy represent the predominant treatment-related thrombotic risk factors.
Cardiovascular comorbidities, including heart failure and stroke/TIA, were found to be significantly associated with VTE. These findings align with the Cancer-VTE Registry data24 and underscore the importance of comprehensive cardiovascular risk assessment in patients with NSCLC. The presence of these comorbidities should prompt consideration of more intensive VTE monitoring and prophylaxis strategies.
The clinical implications of our findings emphasize the importance of the early identification of high-risk patients and the implementation of risk-stratified VTE prophylaxis in patients with advanced NSCLC. Decisions regarding VTE prophylaxis should prioritize patients receiving anti-VEGF agents and platinum-based chemotherapy, particularly in combination regimens, when combined with patient-specific risk factors such as female sex, higher BMI, previous VTE history, or cardiovascular comorbidities. Early recognition of these risk factors allows for timely intervention, which may improve patient outcomes. Emerging data demonstrating the safety and efficacy of prophylactic anticoagulation in high-risk patients support the need for personalized risk assessment.32 This approach may reduce treatment interruptions and delays, potentially leading to improved survival rates. The identification of multiple independent risk factors suggests that a comprehensive risk assessment approach, considering both patient-specific and treatment-related factors, may be most effective in identifying high-risk patients who would benefit from prophylactic intervention. This individualized approach should also incorporate bleeding risk stratification to support safer anticoagulation decisions.33
Study Limitations
This study has some limitations. First, the observational design of this study may have introduced inherent biases. Specifically, because treatment selections were made at the discretion of individual clinicians rather than through randomization, this could have substantially influenced the results. Second, the reliance on administrative data may result in potential misclassification of diagnoses and treatments. Third, the study population is limited to patients treated in advanced treatment centers in Japan, which may limit the generalizability of the findings to other settings or populations. Fourth, our database lacked detailed clinical information that could influence VTE risk, including tumor staging, metastatic burden, Eastern Cooperative Oncology Group performance status, and lifestyle factors. Although we included histology, driver mutation therapy, and the Barthel Index as available surrogate measures, the absence of these clinical assessments may have resulted in residual confounding. Fifth, our definition of VTE included both symptomatic and incidental events, but we could not distinguish between them due to limitations of the administrative data. This may have introduced detection bias, because patients receiving certain treatments may undergo more frequent surveillance imaging. Sixth, our analysis lacked information on treatment-specific details, including chemotherapy dosing and scheduling, and concurrent radiotherapy during follow-up. Finally, the study period (2016–2023) encompasses the COVID-19 pandemic, during which pandemic-related factors may have influenced VTE risk but could not be adjusted for in our analysis.
Conclusions
This large-scale study, using a nationwide administrative database, identified several significant risk factors for VTE in patients with advanced NSCLC, including female sex, higher BMI, previous VTE, anti-VEGF therapy, platinum-based chemotherapy, heart failure, and stroke/TIA. The use of this large database enhances the generalizability of the results, providing valuable insights into how the management of patients with advanced NSCLC can be improved. To reduce VTE-related complications and potentially improve survival outcomes, clinicians should consider risk-stratified monitoring and appropriate prophylactic strategies in high-risk patients.
Disclosures
T.K. is an employee of Daiichi Sankyo Co., Ltd. Y.Y. has received lecture fees from Bayer Healthcare, Bristol-Myers Squibb, Pfizer, and Daiichi Sankyo, as well as grant support from Bayer Healthcare and Daiichi Sankyo. Y.I. was affiliated with Daiichi Sankyo until December 2024. A.S. has received lecture fees from AbbVie, AstraZeneca, Asahi Kasei Corporation, Astellas Pharma, Bayer Yakuhin, Bristol Myers Squibb, Chugai Pharmaceutical, Daiichi Sankyo, Eisai, Janssen Pharmaceutical, Kissei Pharmaceutical, Kyowa Kirin, Mallinckrodt Pharmaceuticals, Maruho, Merck Biopharma, Mitsubishi Tanabe Pharma Corporation, Nipro Corporation, Nippon Shinyaku, Novo Nordisk Pharma, Ono Pharmaceutical, Pfizer, Shionogi Pharma, Taisho Pharmaceutical, Takeda Pharmaceutical Company Limited, and Torii Pharmaceutical. No other disclosures are reported.
Author Contributions
T.K. formulated the study concept and design. T.K., Y.I., and R.K. performed the statistical analysis. T.K., Y.I., and Y.Y. wrote the manuscript. All authors contributed to the discussion, reviewed, edited, and approved the final manuscript, and agreed to submission.
IRB Information
Approval by an institutional review board/ethics committee was not required for this study because the Japanese Ethical Guidelines for Medical and Biological Research Involving Human Subjects do not apply to research using only anonymized data. The data used in this study were anonymized before the start of the study by Medical Data Vision Co., Ltd.; all records were fully deidentified at the source and could not be linked back to individual patients.
Supplementary Files
Supplementary Method 1. Supplementary Method 2. Supplementary Table 1. Supplementary Table 2. Supplementary Figure.
Acknowledgments
The authors acknowledge the use of artificial intelligence (AI) tools in the development of this manuscript. Specifically, the authors used OpenAI’s GPT-4 for assistance with writing and language editing. The authors take full responsibility for the content and conclusions of this paper.
Funding Statement
Sources of Funding: This study was supported by funding from the Department of Medical Statistics, Osaka Metropolitan University Graduate School of Medicine.
Data Availability
The data used in this study cannot be shared with external researchers according to the contract with Medical Data Vision Co., Ltd., which is the database provider.
References
- 1. Siegel RL, Miller KD, Jemal A.. Cancer statistics, 2020. CA Cancer J Clin 2020; 70: 7–30, doi:10.3322/caac.21590. [DOI] [PubMed] [Google Scholar]
- 2. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A.. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018; 68: 394–424, doi:10.3322/caac.21492. [DOI] [PubMed] [Google Scholar]
- 3. Khorana AA, Connolly GC.. Assessing risk of venous thromboembolism in the patient with cancer. J Clin Oncol 2009; 27: 4839–4847, doi:10.1200/JCO.2009.22.327. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Timp JF, Braekkan SK, Versteeg HH, Cannegieter SC.. Epidemiology of cancer-associated venous thrombosis. Blood 2013; 122: 1712–1723, doi:0.1182/blood-2013-04-460121. [DOI] [PubMed] [Google Scholar]
- 5. Ishii M.. DRG/PPS and DPC/PDPS as prospective payment systems. Japan Med Assoc J 2012; 55: 279–291. [PubMed] [Google Scholar]
- 6. Ihara Y, Sawa K, Imai T, Bito T, Shimomura Y, Kawai R, et al.. Immunotherapy and overall survival among patients with advanced non-small cell lung cancer and obesity. JAMA Netw Open 2024; 7: e2425363, doi:10.1001/jamanetworkopen.2024.25363. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Iwai C, Jo T, Konishi T, Fujita A, Michihata N, Matsui H, et al.. Thrombotic risk of platinum combination chemotherapy with and without immune checkpoint inhibitors for advanced non-small cell lung cancer: A nationwide inpatient database study. Cancer Immunol Immunother 2023; 72: 3581–3591, doi:10.1007/s00262-023-03508-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Jacob S, Rahbari K, Tegtmeyer K, Zhao J, Tran S, Helenowski I, et al.. Lung cancer survival in patients with autoimmune disease. JAMA Netw Open 2020; 3: e2029917, doi:10.1001/jamanetworkopen.2020.29917. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Yamashita Y, Fukasawa T, Takeda C, Takeuchi M, Ono K, Kawakami K.. Clinical characteristics and outcomes of patients with venous thromboembolism receiving edoxaban in the real world. Circ J 2024; 88: 371–379, doi:10.1253/circj.CJ-23-0818. [DOI] [PubMed] [Google Scholar]
- 10. Fukasawa T, Seki T, Nakashima M, Kawakami K.. Comparative effectiveness and safety of edoxaban, rivaroxaban, and apixaban in patients with venous thromboembolism: A cohort study. J Thromb Haemost 2022; 20: 2083–2097, doi:10.1111/jth.15799. [DOI] [PubMed] [Google Scholar]
- 11. Mahoney FI, Barthel DW.. Functional evaluation: The Barthel index. Md State Med J 1965; 14: 61–65. [PubMed] [Google Scholar]
- 12. Zhang Z, Reinikainen J, Adeleke KA, Pieterse ME, Groothuis-Oudshoorn CGM.. Time-varying covariates and coefficients in Cox regression models. Ann Transl Med 2018; 6: 121, doi:10.21037/atm.2018.02.12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Lau B, Cole SR, Gange SJ.. Competing risk regression models for epidemiologic data. Am J Epidemiol 2009; 170: 244–256, doi:10.1093/aje/kwp107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Suissa S, Dell’Aniello S.. Time-related biases in pharmacoepidemiology. Pharmacoepidemiol Drug Saf 2020; 29: 1101–1110, doi:10.1002/pds.5083. [DOI] [PubMed] [Google Scholar]
- 15. Sterne JA, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, et al.. Multiple imputation for missing data in epidemiological and clinical research: Potential and pitfalls. BMJ 2009; 338: b2393, doi:10.1136/bmj.b2393. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. R Core Team (2023).. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/ (accessed January 31, 2024).
- 17.Ministry of Education, Culture, Sports, Science and Technology, Ministry of Health, Labour and Welfare, Ministry of Economy, Trade and Industry (2021/2023). Ethical Guidelines for Medical and Biological Research Involving Human Subjects.
- 18. Khorana AA, Kuderer NM, Culakova E, Lyman GH, Francis CW.. Development and validation of a predictive model for chemotherapy-associated thrombosis. Blood 2008; 111: 4902–4907, doi:10.1182/blood-2007-10-11632. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Timp JF, Braekkan SK, Versteeg HH, Cannegieter SC.. Epidemiology of cancer-associated venous thrombosis. Blood 2013; 122: 1712–1723, doi:10.1182/blood-2013-04-460121. [DOI] [PubMed] [Google Scholar]
- 20. Hicks LK, Cheung MC, Ding K, Hasan B, Seymour L, Le Maître A, et al.. Venous thromboembolism and nonsmall cell lung cancer: A pooled analysis of National Cancer Institute of Canada Clinical Trials Group trials. Cancer 2009; 115: 5516–5525, doi:10.1002/cncr.24596. [DOI] [PubMed] [Google Scholar]
- 21. Lee YG, Kim I, Lee E, Bang SM, Kang CH, Kim YT, et al.. Risk factors and prognostic impact of venous thromboembolism in Asian patients with non-small cell lung cancer. Thromb Haemost 2014; 111: 1112–1120, doi:10.1160/TH13-11-0930. [DOI] [PubMed] [Google Scholar]
- 22. Farge D, Frere C, Connors JM, Ay C, Khorana AA, Munoz A, et al.. 2019 International clinical practice guidelines for the treatment and prophylaxis of venous thromboembolism in patients with cancer. Lancet Oncol 2019; 20: e566–e581, doi:10.1016/S1470-2045(19)30336-5. [DOI] [PubMed] [Google Scholar]
- 23. Cushman M, Kuller LH, Prentice R, Rodabough RJ, Psaty BM, Stafford RS, et al.. Estrogen plus progestin and risk of venous thrombosis. JAMA 2004; 292: 1573–1580, doi:10.1001/jama.292.13.1573. [DOI] [PubMed] [Google Scholar]
- 24. Ohashi Y, Ikeda M, Kunitoh H, Sasako M, Okusaka T, Mukai H, et al.. One-year incidence of venous thromboembolism, bleeding, and death in patients with solid tumors newly initiating cancer treatment: Results from the Cancer-VTE Registry. Thromb Res 2022; 213: 203–213, doi:10.1016/j.thromres.2021.09.012. [DOI] [PubMed] [Google Scholar]
- 25. Hara T, Sata M.. Cancer-associated thrombosis, with a focus on ischemic heart disease. Circ J 2026; 90: 371–375, doi:10.1253/circj.CJ-25-0627. [DOI] [PubMed] [Google Scholar]
- 26. Ay C, Pabinger I, Cohen AT.. Cancer-associated venous thromboembolism: Burden, mechanisms, and management. Thromb Haemost 2017; 117: 219–230, doi:10.1160/TH16-08-0615. [DOI] [PubMed] [Google Scholar]
- 27. Nalluri SR, Chu D, Keresztes R, Zhu X, Wu S.. Risk of venous thromboembolism with the angiogenesis inhibitor bevacizumab in cancer patients: A meta-analysis. JAMA 2008; 300: 2277–2285, doi:10.1001/jama.2008.656. [DOI] [PubMed] [Google Scholar]
- 28. Moore RA, Adel N, Riedel E, Bhutani M, Feldman DR, Tabbara NE, et al.. High incidence of thromboembolic events in patients treated with cisplatin-based chemotherapy: A large retrospective analysis. J Clin Oncol 2011; 29: 3466–3473, doi:10.1200/JCO.2011.35.5669. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Alexander M, Ball D, Solomon B, MacManus M, Manser R, Riedel B, et al.. Dynamic thromboembolic risk modelling to target appropriate preventative strategies for patients with non-small cell lung cancer. Cancers (Basel) 2019; 11: 50, doi:10.3390/cancers11010050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Lechner D, Kollars M, Gleiss A, Kyrle PA, Weltermann A.. Chemotherapy-induced thrombin generation via procoagulant endothelial microparticles is independent of tissue factor activity. J Thromb Haemost 2007; 5: 2445–2452, doi:10.1111/j.1538-7836.2007.02788.x. [DOI] [PubMed] [Google Scholar]
- 31. McCrae KR, Swaidani S, Diaz-Montero CM, Khorana AA.. Old is new again: Emergence of thromboembolic complications in cancer patients on immunotherapy. Thromb Res 2022; 213(Suppl 1): S51–S57, doi:10.1016/j.thromres.2022.01.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Key NS, Khorana AA, Kuderer NM, Bohlke K, Lee AYY, Arcelus JI, et al.. Venous thromboembolism prophylaxis and treatment in patients with cancer: ASCO clinical practice guideline update. J Clin Oncol 2020; 38: 496–520, doi:10.1200/JCO.19.01461. [DOI] [PubMed] [Google Scholar]
- 33. Ikeda S, Yamashita Y, Morimoto T, Ueno Y, Maemura K, Chatani R, et al.. Prediction score for major bleeding in patients with venous thromboembolism receiving direct oral anticoagulants: Insights from the COMMAND VTE Registry-2. Circ J 2025; 89: 1896–1905, doi:10.1253/circj.CJ-25-0186. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Method 1. Supplementary Method 2. Supplementary Table 1. Supplementary Table 2. Supplementary Figure.
Data Availability Statement
The data used in this study cannot be shared with external researchers according to the contract with Medical Data Vision Co., Ltd., which is the database provider.