Abstract
INTRODUCTION:
Bevacizumab is used in the treatment of recurrent glioblastoma, but there is limited evidence on the incidence of thromboembolic complications regarding this drug in real-world settings. We aimed to evaluate the risk of arterial thromboembolism (ATE) and venous thromboembolism (VTE) associated with use of bevacizumab among adults diagnosed with high-grade gliomas in a commercially-insured U.S. population.
METHODS:
We conducted a nested case-control study within a retrospective cohort of 2157 patients with high-grade gliomas undergoing craniotomy, radiation and concurrent temozolomide treatment using the Truven Health MarketScan Databases between 2009 and 2015. ATE (n=25) and VTE (n=99) cases and matched controls (n=170 and n=819 respectively) were identified using incidence density sampling with replacement. Multivariable conditional logistic regression models were used to estimate odds ratios (OR) and 95% confidence intervals (CI) for risk of ATE and VTE separately.
RESULTS:
A higher proportion of ATE cases received bevacizumab compared to controls (28% versus 17%, OR=1.51, 95% CI 0.54-4.24) but this excess in odds was not statistically significant. Similarly, bevacizumab was not significantly associated with VTE (13% vs. 9%, OR=1.40, 95% CI 0.71-2.75).
CONCLUSIONS:
We found no significant association between use of bevacizumab and thromboembolic events in patients with high-grade gliomas, although our study was limited by a small number of ATE events and the potential for complications from arterial thrombosis cannot be completely ruled out.
Keywords: Bevacizumab, glioma, thromboembolism, safety, arterial thromboembolism, venous thromboembolism
Introduction
High-grade gliomas are severe, rapidly progressive, and the most common form of primary brain malignancy in adults1. These cancers are further classified as anaplastic gliomas (grade III) and glioblastoma (grade IV) with the latter being a more aggressive form with a poor prognosis. Patients with glioblastoma have a 5-year survival rate ranging between 0.5% and 4.7% in the U.S.1-3. Standard care for incident high-grade glioma includes maximal surgical resection followed by radiotherapy and concurrent/adjuvant temozolomide, a treatment regimen first disseminated by Stupp et al. (hereafter referred to as the Stupp protocol)4,5. Nearly all glioblastoma cases recur even in instances of apparent complete removal of the tumor6. Bevacizumab, approved treatment for recurrent glioblastoma, is a recombinant humanized monoclonal antibody that inhibits the action of vascular endothelial growth factor (VEGF) thereby stunting angiogenesis and subsequently reducing tumor size7. This mechanism of action, however, may lead to an adverse event of thromboembolisms8,9.
In a study conducted using a large cancer registry data linked with Medicare claims, the reported incidence of arterial thromboembolism (ATE) varied by cancer site (1-year cumulative incidence 3.9-10.3%) with the highest risk observed in patients with lung cancer10. Evidence is limited on the risk of ATE experienced among high-grade glioma patients. Risk of venous thromboembolism (VTE) was higher among cancer patients in general including those with malignant gliomas11,12 compared to non-cancer patients but this also varies by type of cancer with the highest risk observed among patients with pancreatic cancer (incidence rate 8.2-19.2%)13.
Thromboembolic events in patients with cancer may often result in discontinuation of therapy and are associated with significant morbidity and high risk of mortality14. While the risk of ATE among cancer patients treated with bevacizumab is demonstrated to be increased across various cancer sites15, there are conflicting findings on treatment-related VTE associated with the use of bevacizumab with more varied association by cancer site (HR 0.67-2.08)16,17. A phase III clinical trial of bevacizumab (AVAglio) in 911 glioblastoma patients reported that those receiving bevacizumab had higher incidence of ATE but not VTE compared to placebo18,19. However, there are no known studies on the safety of bevacizumab with regard to risk of ATE or VTE among high-grade glioma patients in real world settings. Therefore, our study aimed to evaluate the risk of ATE and VTE associated with bevacizumab among high-grade glioma patients undergoing treatment per the standard Stupp protocol in a commercially-insured U.S. population.
Methods
We conducted a nested case-control study in a retrospective cohort of patients with high-grade gliomas using the Truven Health MarketScan Commercial and Medicare Supplemental Databases20 between January 1, 2009 and September 30, 2015. This data source contains the administrative health claims of millions of commercially insured enrollees and their dependents across the U.S. with information from inpatient and outpatient encounters, pharmacy dispensings, and health plan enrollment information.
Study population
We selected a retrospective cohort of patients receiving treatment for high-grade gliomas under the Stupp protocol4 using a validated algorithm for administrative claims databases5. To be eligible, patients were required to: 1) undergo incident craniotomy on or after January 1, 2010; 2) begin radiation therapy within 91 days of surgical resection; 3) initiate temozolomide within 91 days of craniotomy; and 4) be continuously enrolled during the one year prior to craniotomy. The date of the incident craniotomy was defined as the index date and 91 days after the index date was defined as the time of cohort entry for all subjects. The study sample was restricted to patients entering the cohort prior to September 30, 2015.
To ensure follow up from treatment initiation for high-grade gliomas, patients were excluded if they received any craniotomy, radiation, temozolomide or bevacizumab during the year prior to index craniotomy or experienced one of our outcomes of interest (ATE or VTE) during the cohort ascertainment period, which was between index and cohort entry dates. Finally, patients were required to have continuous health plan enrollment during the 12-month prior to index date and follow up periods, unless died.
Cases and controls
Cases of ATE and VTE were each identified in the overall cohort using a validated algorithm for administrative claims data21-23. To determine an ATE case, primary discharge codes in ICD9-CM for ischemic stroke (433.x1, 434.x1, and 436)22 and acute myocardial infarction (410.x0, 410.x1) were used23. Primary inpatient discharge codes in ICD9-CM for pulmonary embolism (415.11, 415.19) and venous thrombosis (453.1, 453.2, 453.40, 453.41, 453.42, 453.8, and 453.9) were used to determine occurrence of a VTE event21. For ATE and VTE separately, each case was matched to up to ten controls on sex, age, quarter-year of index time, and follow-up duration using incidence density sampling with replacement from the overall cohort24. Controls were at risk for the outcome of interest (i.e., ATE or VTE) at the time of case occurrence and survived at least as long as their referent case. Cases without any matched control were excluded from the analysis.
Exposure ascertainment
Exposure to bevacizumab was determined using the Healthcare Common Procedure Coding System (HCPCS) code J9035 during inpatient or outpatient encounters between index date and the event or corresponding matched control date. Exposure was characterized as having received any bevacizumab (yes vs. no) and recent use of bevacizumab (last bevacizumab infusion within 30 days; more than 30 days prior event or control censoring; or having never received bevacizumab) following the Stupp protocol.
Covariates
Information on demographic characteristics, VTE risk factors, comorbidities, healthcare utilization, previous ATE or VTE events, and medications (direct oral anticoagulants [DOAC], low molecular weight heparin [LMWH], warfarin, and antiplatelet) were collected during 1-year baseline before index date (date of craniotomy) using diagnosis and procedure codes. A cancer-specific comorbidity score developed by the National Cancer Institute (NCI) during the baseline was also collected using data from the baseline period25. Additionally, for analyses on VTE risk, we obtained information on prophylaxis (DOAC, LMWH, warfarin or antiplatelet) post-surgical resection.
Statistical analysis
We estimated the relative risk of ATE and VTE associated with bevacizumab in separate models using conditional logistic regression models to calculate odds ratios (OR) and 95% confidence intervals (CI)26. Due to matching, the crude models for both the risk of ATE and VTE were controlled for sex, age and follow-up time. We additionally adjusted for a priori selected covariates: baseline NCI comorbidity score in all models; and number of baseline VTE risk factors and VTE prophylaxis received in VTE analysis model. Covariates in the model were selected based on clinical significance such as well-established factors that have been associated with increased risk of VTE and/or ATE. We performed subgroup analyses where follow up was restricted to 1 year and 180 days after cohort entry to limit to events likely related to primary therapy (vs. reintervention).
Results
Our final study cohort included 2,157 patients undergoing the standard Stupp protocol for high-grade gliomas. Our nested case-control analysis included 25 ATE cases (Table 1), 99 VTE cases (Table 2), and their matched incidence density-sampled controls (n=170 for ATE; n=819 for VTE).
Table 1.
Characteristics of arterial thromboembolism cases and matched, incidence density sampled controls pre-surgery
| ATE Casesa (n=25) |
Controls (n=170) |
||||
|---|---|---|---|---|---|
| n | (%) | n | (%) | P | |
| Age, years | |||||
| Mean (SD) | 58.9 | (7.9) | 58.5 | (5.7) | 0.74 |
| 35-44 | 1 | (4.0) | 2 | (1.2) | 0.12 |
| 45-54 | 4 | (16.0) | 24 | (14.1) | |
| 55-64 | 16 | (64.0) | 134 | (78.8) | |
| 65+ | 4 | (16.0) | 10 | (5.9) | |
| Sex | |||||
| Men | 14 | (56.0) | 102 | (60.0) | 0.70 |
| Women | 11 | (44.0) | 68 | (40.0) | |
| Comorbidities at baseline | |||||
| Atrial fibrillation | 2 | (8.0) | 6 | (3.5) | 0.27 |
| Hypertension | 12 | (48.0) | 65 | (38.2) | 0.35 |
| COPD | 2 | (8.0) | 14 | (8.2) | 0.99 |
| Diabetes Mellitus | 5 | (20.0) | 18 | (10.6) | 0.19 |
| Hyperlipidemia | 6 | (24.0) | 61 | (35.9) | 0.24 |
| Congestive heart failure | 0 | (0.0) | 7 | (4.1) | 0.60 |
| Chronic renal disease | 0 | (0.0) | 3 | (1.8) | 0.99 |
| NCI comorbidity score | |||||
| 0 | 16 | (64.0) | 118 | (69.4) | 0.60 |
| 1 | 7 | (28.0) | 32 | (18.8) | |
| 2+ | 2 | (8.0) | 20 | (11.8) | |
| Previous VTE | 0 | (0.0) | 1 | (0.6) | 0.99 |
| Previous ATE | 1 | (4.0) | 2 | (1.2) | 0.34 |
| Bevacizumab post-craniotomy | |||||
| None | 18 | (72.0) | 141 | (82.9) | 0.17 |
| Any | 7 | (28.0) | 29 | (17.1) | |
| Started within 180 days | 4 | (16.0) | 10 | (5.9) | |
| Started after >180 days | 3 | (12.0) | 19 | (11.2) | |
Abbreviations: ATE, arterial thromboembolism; VTE, venous thromboembolism; COPD, chronic obstructive pulmonary disease; NCI, National Cancer Institute
To compare cases and controls we used Student’s t test for continuous variables and chi-square test for categorical variables (Fisher’s exact test with cells <5)
ATE cases consisted of 20 ischemic strokes and 5 acute myocardial infarction
Table 2.
Characteristics of venous thromboembolism cases and matched, incidence density sampled controls pre-surgery
| VTE Casesa (n=99) |
Controls (n=819) |
||||
|---|---|---|---|---|---|
| n | (%) | n | (%) | P | |
| Age, years | |||||
| Mean (SD) | 56.6 | (10.4) | 56.3 | (8.1) | 0.78 |
| 18-34 | 4 | (4.0) | 13 | (1.6) | 0.13 |
| 35-44 | 4 | (4.0) | 24 | (2.9) | |
| 45-54 | 32 | (32.3) | 283 | (34.6) | |
| 55-64 | 45 | (45.5) | 425 | (51.9) | |
| 65+ | 14 | (14.1) | 74 | (9.0) | |
| Sex | |||||
| Men | 66 | (66.7) | 537 | (65.6) | 0.83 |
| Women | 33 | (33.3) | 282 | (34.4) | |
| Comorbidities at baseline | |||||
| Atrial fibrillation | 2 | (2.0) | 15 | (1.8) | 0.70 |
| Hypertension | 30 | (30.3) | 294 | (35.9) | 0.27 |
| COPD | 7 | (7.1) | 28 | (3.4) | 0.09 |
| Diabetes Mellitus | 10 | (10.1) | 92 | (11.2) | 0.73 |
| Hyperlipidemia | 31 | (31.3) | 266 | (32.5) | 0.81 |
| Congestive heart failure | 2 | (2.0) | 13 | (1.6) | 0.67 |
| Chronic renal disease | 2 | (2.0) | 19 | (2.3) | 0.99 |
| NCI comorbidity score | |||||
| 0 | 63 | (63.6) | 570 | (69.6) | 0.12 |
| 1 | 27 | (27.3) | 153 | (18.7) | |
| 2+ | 9 | (9.1) | 96 | (11.7) | |
| Medication use at baseline | |||||
| DOAC | 1 | (1.0) | 7 | (0.9) | 0.60 |
| LMWH | 2 | (2.0) | 14 | (1.7) | 0.69 |
| Warfarin | 4 | (4.0) | 28 | (3.4) | 0.77 |
| Antiplatelet agents | 2 | (2.0) | 19 | (2.3) | 0.99 |
| VTE risk factors | |||||
| Trauma | 22 | (22.2) | 152 | (18.6) | 0.38 |
| Immobility | 2 | (2.0) | 35 | (4.3) | 0.42 |
| Central catheter use | 1 | (1.0) | 13 | (1.6) | 0.99 |
| Recent hospitalizationb | 35 | (35.4) | 254 | (31.0) | 0.38 |
| Oral contraceptives or hormone replacement therapy | 5 | (5.1) | 46 | (5.6) | 0.82 |
| Number of VTE risk factorsc | |||||
| 0-1 | 82 | (82.8) | 709 | (86.6) | 0.31 |
| 2+ | 17 | (17.2) | 110 | (13.4) | |
| Previous VTE | 2 | (2.0) | 7 | (0.9) | 0.25 |
| Previous ATE | 1 | (1.0) | 9 | (1.1) | 0.99 |
| Post-operative VTE prophylaxisd | 16 | (16.2) | 137 | (16.7) | 0.89 |
| Bevacizumab post-craniotomy | |||||
| None | 86 | (86.9) | 743 | (90.7) | 0.16 |
| Any | 13 | (13.1) | 76 | (9.3) | |
| Started within 180 days | 7 | (7.1) | 55 | (6.7) | |
| Started after >180 days | 6 | (6.1) | 21 | (2.6) | |
Abbreviations: VTE, venous thromboembolism; ATE, arterial thromboembolism; COPD, chronic obstructive pulmonary disease; NCI, National Cancer Institute; DOAC, direct oral anticoagulants; LMWH, low molecular weight heparin
To compare cases and controls we used Student’s t test for continuous variables and chi-square test for categorical variables (Fisher’s exact test with cells <5)
VTE cases consisted of 60 pulmonary embolism and 39 venous thrombosis
Hospitalization within 30-days prior to index date (date of craniotomy)
VTE risk factors: pregnancy, transplant, hip fracture, hypercoagulable state, obesity and risk factors listed above
Use of DOAC, LMWH, warfarin, or antiplatelet agents as post-operative VTE prophylaxis
A higher proportion of ATE cases had atrial fibrillation (8% vs. 4%), hypertension (48% vs. 38%), and diabetes (20% vs. 11%) during the baseline compared to controls; more ATE cases received bevacizumab during follow up compared to the controls (28% vs. 17%) and among those treated with bevacizumab a higher proportion of ATE cases received bevacizumab within 180 days after surgery compared to controls (57% vs. 34%) but these differences were not statistically significant.
In the VTE analysis, cases had a slightly higher proportion of baseline VTE risk factors (2 or more: 17% vs. 13%) compared to controls (see Table 2); more VTE cases received bevacizumab compared to controls (13% vs. 9%), and among those who received bevacizumab a lower proportion of VTE cases were treated with bevacizumab within 180 days of surgical tumor resection compared to controls (54% vs. 72%) but none of these differences were statistically significant.
Table 3 provides results from multivariable conditional logistic regression models for the association between bevacizumab and the odds of ATE and VTE. We found no significantly increased association between ATE and any use of bevacizumab (aOR 1.51, 95% CI 0.54-4.24) or recent use of bevacizumab within 30 days prior to the occurrence of an ATE (aOR 1.36, 95% CI 0.40-4.66). However, the confidence intervals for the odds of ATE were very wide due to the low number of overall events identified. Similarly, we found no significantly positive association of VTE with receipt of any bevacizumab use (aOR 1.40, 95% CI 0.71-2.75) or the most recent use of bevacizumab within 30 days prior VTE (aOR 1.40, 95% CI 0.65-3.01). There were no significant changes in the odds of ATE and VTE in subgroup analyses restricting the length of follow-up (Figure 1).
Table 3.
Multivariable, conditional logistic regression models of bevacizumab in relation to risk of arterial and venous thromboembolism
| Cases | Controls | Model 1a | Model 2b | |||||
|---|---|---|---|---|---|---|---|---|
| n | (%) | n | (%) | OR | (95%CI) | OR | (95% CI) | |
| ATE | ||||||||
| Bevacizumab use | ||||||||
| Never | 18 | (72.0) | 141 | (82.9) | 1.00 | Ref. | 1.00 | Ref. |
| Ever | 7 | (28.0) | 29 | (17.1) | 1.49 | (0.53, 4.17) | 1.51 | (0.54, 4.24) |
| Recent bevacizumab use | ||||||||
| Never | 18 | (72.0) | 141 | (82.9) | 1.00 | Ref. | 1.00 | Ref. |
| 1-30 days prior to event | 4 | (16.0) | 18 | (10.6) | 1.41 | (0.42, 4.78) | 1.36 | (0.40, 4.66) |
| >30 days prior to event | 3 | (12.0) | 11 | (6.5) | 1.63 | (0.40, 6.69) | 1.80 | (0.41, 7.89) |
| VTE | ||||||||
| Bevacizumab use | ||||||||
| Never | 86 | (86.9) | 743 | (90.7) | 1.00 | Ref. | 1.00 | Ref. |
| Ever | 13 | (13.1) | 76 | (9.3) | 1.38 | (0.71, 2.70) | 1.40 | (0.71, 2.75) |
| Recent bevacizumab use | ||||||||
| Never | 86 | (86.9) | 743 | (90.7) | 1.00 | Ref. | 1.00 | Ref. |
| 1-30 days prior to event | 9 | (9.1) | 59 | (7.2) | 1.41 | (0.66, 3.00) | 1.40 | (0.65, 3.01) |
| >30 days prior to event | 4 | (4.0) | 17 | (2.1) | 1.32 | (0.40, 4.33) | 1.40 | (0.41, 4.73) |
Matched on age, sex and at risk time by incidence density sampling
Additional multivariable adjustment for NCI comorbidity score in ATE analysis; and adjusted for NCI comorbidity score, number of VTE risk factors and receipt of VTE prophylaxis in VTE analysis
Figure 1.
Association between bevacizumab and ATE/VTE in subgroups
* Adjusted model included matching on age, sex and at-risk time by incidence density sampling. Models were additionally adjusted for NCI comorbidity score in ATE analysis; NCI comorbidity score, number of VTE risk factors and receipt of VTE prophylaxis in VTE analysis by multivariable modeling
Discussion
Bevacizumab was not significantly associated with thromboembolic risk in high-grade glioma patients in this study, which to our knowledge is the first study evaluating real-world thromboembolic safety of bevacizumab in this population. Bevacizumab’s action of inhibiting VEGF may hinder regeneration of endothelial cells and cause subendothelial collagen to be exposed, potentially leading to thrombosis27,28. Other than the differential underlying thrombotic risk of different cancer site, the mechanism of why VEGF inhibitors may have less risk of thrombosis on certain types of tumors compared to others remains unclear.
Our study had limitations. As with other nonrandomized studies, there is a potential for residual confounding due to unmeasured factors including information on clinical biomarkers such as D-dimer and body mass index (BMI), which are predictors of VTE in patients with cancer29. Although the cases and controls were matched on sex, age, cohort entry time and follow-up time, the small number of cases restricted the number of covariates to be included in the multivariable analyses. Thus, in the ATE analysis, only NCI comorbidity index was included in the multivariable-adjusted model. In the VTE analysis, number of VTE risk factors was included in the multivariable regression model rather than including all the individual risk factors. .
There is difficulty in defining high-grade glioma patients using administrative claims databases as there is no specific ICD-9CM code for high-grade gliomas. We utilized a previously validated algorithm based on procedure codes and medication dispensing records to identify high-grade glioma patients undergoing the Stupp protocol treatment5. Even so, it remains possible that high-grade glioma patients were misclassified. Another limitation is that as the Truven MarketScan Databases contain claims from commercially-insured health plan enrollees, the results may not be representative of others, such as those lacking health insurance or are covered by Medicaid.
In a randomized clinical trial of bevacizumab in patients with aggressive brain cancers, AVAglio, newly diagnosed glioblastoma patients treated with bevacizumab experienced higher incidence of ATE compared to placebo (5.9% vs. 1.6%; P=0.001)18. In the same trial, the rate of grade 3-5 ATE was significantly higher among bevacizumab group compared to placebo group (rate ratio=3.18, 95% CI 1.30-7.81)19. The frequency of VTE was similar between the bevacizumab and placebo groups (8.2% vs. 9.6%) and our finding of no significant increase in the odds of VTE associated with bevacizumab is consistent with the AVAglio trial. In our study, we had 80% power to determine a minimum detectable OR of 3.6 for ATE and 2.3 for VTE, both of which are higher than what was observed from the AVAglio trial. Therefore, there was a possibility for type 2 error in this study. While the risk estimate for ATE was modest and not statistically significant in our study contrary to the AVAglio study, we likely did not have the power to detect an association if it exists and thus cannot completely rule out the potential risk of ATE associated with bevacizumab. Another potential explanation for the discrepancy between the clinical trial and real-world outcome could be due to potential channeling bias where physicians channel patients away from bevacizumab for those considered at high risk of ATE thereby mitigating the risk of ATE seen among bevacizumab users in real-world.
Conclusion
Our findings from this population-based cohort study do not suggest significantly increased association between bevacizumab treatment and thromboembolic events among patients with high-grade gliomas. Further research is needed to confirm the thromboembolic safety of bevacizumab in a larger sample of high-grade glioma patients.
Acknowledgments
Authors of this work were supported by the UIC-AbbVie Fellowship in Pharmacovigilance and Patient Safety (Lee) and the UIC-AbbVie Fellowship in Health Economics and Outcomes Research (Zhou). This work was supported by grants from the National Institutes of Health, National Heart, Lung and Blood Institute through grant number R21HL140531 and National Center for Advancing Translational Sciences through grant number KL2TR002002. The funders had no role in the design, conduct and interpretation of results of this study. The findings are the responsibility of the authors and do not necessarily reflect the official views of the National Institutes of Health.
Footnotes
Prior presentation
Results from this study were presented, in part, at the 60th Annual Meeting of the American Society of Hematology in San Diego, California, December 1-4, 2018.
Conflict of interest
Pritesh R. Patel has consulted and received honoraria from Celgene, Amgen and Janssen. All other authors have no conflict of interest to disclose.
References
- 1.Ostrom QT, Bauchet L, Davis FG, Deltour I, Fisher JL, Langer CE, et al. The epidemiology of glioma in adults: a “state of the science” review. Neuro-Oncol. 2014. July 1;16(7):896–913. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ostrom QT, Gittleman H, Farah P, Ondracek A, Chen Y, Wolinsky Y, et al. CBTRUS Statistical Report: Primary Brain and Central Nervous System Tumors Diagnosed in the United States in 2006-2010. Neuro-Oncol. 2013. November 1;15(suppl_2):ii1–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Arrigo RT, Boakye M, Skirboll SL. Patterns of care and survival for glioblastoma patients in the Veterans population. J Neurooncol. 2012. February;106(3):627–35. [DOI] [PubMed] [Google Scholar]
- 4.Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn MJB, et al. Radiotherapy plus Concomitant and Adjuvant Temozolomide for Glioblastoma. N Engl J Med. 2005. March 10;352(10):987–96. [DOI] [PubMed] [Google Scholar]
- 5.Nathan JK, Brezzell AL, Kim MM, Leung D, Wilkinson DA, Hervey-Jumper SL. Early initiation of chemoradiation following index craniotomy is associated with decreased survival in high-grade glioma. J Neurooncol. 2017. November 1;135(2):325–33. [DOI] [PubMed] [Google Scholar]
- 6.Hou LC, Veeravagu A, Hsu AR, Tse VCK. Recurrent glioblastoma multiforme: a review of natural history and management options. Neurosurg Focus. 2006. April;20(4):E3. [DOI] [PubMed] [Google Scholar]
- 7.Cohen MH, Shen YL, Keegan P, Pazdur R. FDA Drug Approval Summary: Bevacizumab (Avastin®) as Treatment of Recurrent Glioblastoma Multiforme. The Oncologist. 2012. November;17(11):1482–1482. [DOI] [PubMed] [Google Scholar]
- 8.Elice F, Rodeghiero F, Falanga A, Rickles FR. Thrombosis associated with angiogenesis inhibitors. Best Pract Res Clin Haematol. 2009. March;22(1):115–28. [DOI] [PubMed] [Google Scholar]
- 9.Genentech, Inc. Avastin® (bevacizumab) Efficacy, Safety, Dosing & Prescribing Information [Internet]. [cited 2018 Nov 9]. Available from: https://www.avastin-hcp.com/
- 10.Navi BB, Reiner AS, Kamel H, Iadecola C, Okin PM, Elkind MSV, et al. Risk of Arterial Thromboembolism in Patients With Cancer. J Am Coll Cardiol. 2017. August;70(8):926–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Marras LC, Geerts WH, Perry JR. The risk of venous thromboembolism is increased throughout the course of malignant glioma. Cancer. 2000. August 1;89(3):640–6. [DOI] [PubMed] [Google Scholar]
- 12.Perry JR. Thromboembolic disease in patients with high-grade glioma. Neuro-Oncol. 2012. September 1;14(suppl 4):iv73–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Khorana AA, Dalal M, Lin J, Connolly GC. Incidence and predictors of venous thromboembolism (VTE) among ambulatory high-risk cancer patients undergoing chemotherapy in the United States: VTE in Ambulatory US Cancer Patients. Cancer. 2013. February 1;119(3):648–55. [DOI] [PubMed] [Google Scholar]
- 14.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. March;5(3):632–4. [DOI] [PubMed] [Google Scholar]
- 15.Schutz FAB, Je Y, Azzi GR, Nguyen PL, Choueiri TK. Bevacizumab increases the risk of arterial ischemia: a large study in cancer patients with a focus on different subgroup outcomes. Ann Oncol. 2011. June 1;22(6):1404–12. [DOI] [PubMed] [Google Scholar]
- 16.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. November 19;300(19):2277. [DOI] [PubMed] [Google Scholar]
- 17.Hurwitz HI, Saltz LB, Van Cutsem E, Cassidy J, Wiedemann J, Sirzén F, et al. Venous Thromboembolic Events With Chemotherapy Plus Bevacizumab: A Pooled Analysis of Patients in Randomized Phase II and III Studies. J Clin Oncol. 2011. May;29(13):1757–64. [DOI] [PubMed] [Google Scholar]
- 18.Chinot OL, Wick W, Mason W, Henriksson R, Saran F, Nishikawa R, et al. Bevacizumab plus Radiotherapy–Temozolomide for Newly Diagnosed Glioblastoma. N Engl J Med. 2014. February 19;370(8):709–22. [DOI] [PubMed] [Google Scholar]
- 19.Saran F, Chinot OL, Henriksson R, Mason W, Wick W, Cloughesy T, et al. Bevacizumab, temozolomide, and radiotherapy for newly diagnosed glioblastoma: comprehensive safety results during and after first-line therapy. Neuro-Oncol. 2016. July 1;18(7):991–1001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Hansen L The Truven Health MarketScan Databases for life sciences researchers. Truven Health Analytics; 2017.
- 21.White RH, Garcia M, Sadeghi B, Tancredi DJ, Zrelak P, Cuny J, et al. Evaluation of the predictive value of ICD-9-CM coded administrative data for venous thromboembolism in the United States. Thromb Res. 2010. July 1;126(1):61–7. [DOI] [PubMed] [Google Scholar]
- 22.Andrade SE, Harrold LR, Tjia J, Cutrona SL, Saczynski JS, Dodd KS, et al. A systematic review of validated methods for identifying cerebrovascular accident or transient ischemic attack using administrative data. Pharmacoepidemiol Drug Saf. 2012. January 1;21:100–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Cutrona SL, Toh D, Iyer A, Foy S, Cavagnaro E, Forrow S, et al. Mini-Sentinel Systematic Validation of Health Outcome of Interest: Acute Myocardial Infarction Cases Report. FDA Mini-Sentinel; 2011. May.
- 24.Richardson DB. An incidence density sampling program for nested case-control analyses. Occup Environ Med. 2004. December 1;61(12):e59–e59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.NCI Comorbidity Index Overview [Internet]. NIH National Cancer Institute; [cited 2018 Oct 22]. Available from: https://healthcaredelivery.cancer.gov/seermedicare/considerations/comorbidity.html [Google Scholar]
- 26.Rundle AG, Vineis P, Ahsan H. Design Options for Molecular Epidemiology Research within Cohort Studies. Cancer Epidemiol Biomark Amp Prev. 2005. August 1;14(8):1899. [DOI] [PubMed] [Google Scholar]
- 27.Kilickap S, Abali H, Celik I. Bevacizumab, Bleeding, Thrombosis, and Warfarin. J Clin Oncol. 2003. September 15;21(18):3542–3542. [DOI] [PubMed] [Google Scholar]
- 28.Zachary I Signaling mechanisms mediating vascular protective actions of vascular endothelial growth factor. Am J Physiol-Cell Physiol. 2001. June 1;280(6):C1375–86. [DOI] [PubMed] [Google Scholar]
- 29.Ay C, Dunkler D, Marosi C, Chiriac A-L, Vormittag R, Simanek R, et al. Prediction of venous thromboembolism in cancer patients. Blood. 2010. December 9;116(24):5377. [DOI] [PubMed] [Google Scholar]