Venous thromboembolism occurs in approximately one-third of patients with glioma, but the decision to treat patients with anticoagulation is complicated by the risk for spontaneous intracranial hemorrhage (ICH). The incidence of ICH in patients with glioma receiving anticoagulation is not known, making it difficult to determine the risk–benefit profile of anticoagulation in this setting.
In a matched, retrospective cohort study published in Blood, Charlene Mantia, MD, from the Department of Medicine at Beth Israel Deaconess Medical Center at Harvard Medical School, and co-authors assessed the safety of anticoagulation in patients with high-grade glioma, finding that those receiving enoxaparin were three times more likely to develop a major ICH, compared with those who did not receive anticoagulation (14.7% vs. 2.5%; hazard ratio [HR] = 3.37; 95% CI 1.02-11.14; p=0.036).
"We were surprised that enoxaparin in patients with glioma was associated with a significantly increased risk of major ICH," co-author Jeffrey I. Zwicker, MD, told ASH Clinical News. "We believe caution is warranted when considering therapeutic anticoagulation in this setting."
The researchers collected data from electronic medical records at Beth Israel Deaconess Medical Center between 2000 and 2016. Patients were included if they had a World Health Organization grade III or IV glioma, had at least two radiographic brain images on file, and had received care at the facility for at least 2 months. Their outcomes were compared with a control cohort (matched for age at diagnosis, year of diagnosis, and sex) of patients with glioma who did not receive anticoagulation.
A total of 133 patients with primary brain tumors (the most common of which were glioblastoma [84%], anaplastic oligodendroglioma [10%], and anaplastic astrocytoma [5%]) were enrolled; 50 received enoxaparin and 83 comprised the control cohort.
Most patients (76%) received enoxaparin 1 mg/kg; two patients (4%) received 1.5 mg/kg, eight (16%) received less than the standard therapeutic dose, and two (4%) did not have dose information available.
A total of 61 cases of ICH (confirmed via radiology reports reviewed by a neuro-oncologist) were recorded in both arms of the study.
In addition to the higher risk of major ICH at 1 year, enoxaparin was also associated with higher rates of measurable ICH (defined as ICH with a bleed volume ≥1 mL), compared with controls: 18.8 percent versus 7.8 percent (HR=2.16; 95% CI; 0.99-4.74; p=0.05).
The researchers also noted that the rates of major ICH did not significantly differ when they adjusted for type of glioma, age, hypertension, glioma treatment, and administration of anti-angiogenic agents.
To further evaluate the association between enoxaparin and development of ICH, the investigators considered enoxaparin administration as a time-varying covariate, comparing time on and time off enoxaparin. During exposure to enoxaparin, the risk of major ICH increased 13-fold (HR=13.29; 95% CI 3.33-52.85; p<0.0001).
Median overall survival (OS) was similar in the enoxaparin and control cohorts (18.7 months vs. 17.1 months; p=0.82). However, among all patients who experienced a major ICH, OS was significantly shorter for those taking enoxaparin, compared with controls (3.3 months vs. 10.2 months; p=0.012). This association remained significant after the researchers adjusted for time from glioma diagnosis until hemorrhage, as well as age at diagnosis (HR=3.74; 95% CI 1.24-11.30; p=0.02). "The median survival was shortened by nearly 70 percent [with enoxaparin]," the authors wrote, adding that this was "a concerning finding."
There are no risk scores specifically designed to predict ICH with therapeutic anticoagulation in patients with glioma, so the investigators used the PANWARDS risk score – which was developed to predict ICH with anticoagulation in non-cancer cohorts – to determine potential risk in this population.
With PANWARDS, risk is determined by the following variables:
- platelet count
- albumin
- history of congestive heart failure
- warfarin
- age
- race
- diastolic blood pressure
- history of stroke or transient ischemic attack
All major ICHs occurred in those with PANWARDS risk scores of ≥25 (out of a maximum of 99 points), with a corresponding sensitivity of 100 percent (95% CI 63-100) and specificity of 40 percent (95% CI 25-56). The cumulative incidence of major ICH at 1 year was 23 percent (95% CI 9.91-39.41) in patients with PANWARDS scores ≥25, compared with 0 percent for those with lower scores (p=0.03). "The PANWARDS prediction score may prove useful to identify those glioma patients at highest risk of suffering from a major intracranial hemorrhage," Dr. Zwicker said.
The study is limited by its single-center, retrospective design and small patient population. The study was not randomized, and thus may not have been able to control for all factors determining why some patients received anticoagulation therapy, while others did not.
"Additional studies are needed to confirm the prognostic value of the PANWARDS scale and to establish the efficacy of alternative therapeutic approaches in those patients considered at high risk of developing an ICH," Dr. Zwicker noted.
Reference
Mantia C, Uhlmann EJ, Puligandla M, et al. Predicting the higher rate of intracranial hemorrhage in glioma patients receiving therapeutic enoxaparin. Blood. 2017 May 3. [Epub ahead of print]