The dangerous gamble of heparinization within two weeks of nonoperative traumatic acute subdural hematoma in patients with increased stroke risk: A case seriesS McClelland1, SJ Mackey2, SS Kim1
1 Department of Neurosurgery, Lahey Clinic, Burlington, Massachusetts, United States
2 Department of General Surgery, Lahey Clinic, Burlington, Massachusetts, United States
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0022-3859.132344
Source of Support: None, Conflict of Interest: None
Background: In traumatic acute subdural hematoma (aSDH) management, systemic anticoagulation is contraindicated, particularly during the first 2 weeks. We present two cases of patients with nonoperative aSDH whose stroke risk led to heparinization within 2 weeks of the initial hemorrhage and examine their outcomes to illustrate the risks and benefits associated with systemic anticoagulation. Materials and Methods: Two elderly males, on warfarin at baseline who developed traumatic nonoperative aSDH were heparinized within 2 weeks of aSDH onset. Results: One patient showed a decreased SDH volume on Day 19. The second patient developed sudden onset headache with fixed/dilated pupils on Day 5. In this patient, a CT scan of the brain revealed marked enlargement of the aSDH from 0.9 to 2.4 cm with midline shift of 1.5 cm, and uncal herniation that was incompatible with life. Conclusion: Heparinization within two weeks of aSDH may cause SDH enlargement resulting in rapidly fatal neurologic deterioration. Further study is needed to more definitively address this issue.
Keywords: Anticoagulation, conservative management, heparin, neurosurgery, stroke risk, traumatic acute subdural hematoma, warfarin
Traumatic acute subdural hematoma (aSDH) can present as either a life-threatening injury requiring operative intervention or a condition not requiring surgery, depending on the size, midline shift, and neurologic examination associated with the aSDH.  Due to the untoward consequences should an aSDH acutely enlarge, systemic anticoagulation is strongly contraindicated, particularly during the first 2 weeks while the clot is still in an acute phase.
On occasion, patients with nonoperative aSDH are considered candidates for early anticoagulation due to medical comorbidities that place them at an increased risk of stroke. We present two cases of patients with nonoperative aSDH whose stroke risk led to heparinization within 2 weeks of the initial hemorrhage and examine their outcomes to illustrate the risks and benefits associated with systemic anticoagulation.
A retrospective chart review of all trauma patients at Lahey Clinic from January 2012 to March 2013 was conducted in order to identify patients presenting with aSDH and anticoagulant medication. Two patients who presented in this fashion underwent restarting of anticoagulation within 2 weeks of the initial hemorrhage; the clinical course of each patient is presented in the following sections.
The first patient was a 75-year-old woman with a pertinent past medical history of atrial fibrillation, atrial flutter, mechanical St. Jude mitral valve replacement for which she was on Coumadin, as well as a biventricular implantable cardiac defibrillator. She tripped over her cat at home, resulting in a mechanical fall with head trauma. This resulted in an acute chronic right temporoparietal subdural hematoma with maximum thickness of 0.9 cm, no midline shift [Figure 1], and a Glasgow Coma Scale (GCS)score of 14. Although her Coumadin was initially discontinued [international normalized ratio (INR)on admission was 3.1], given the risks of thrombus associated with her mechanical valve, the decision was made jointly between the cardiology and neurosurgery services to systemically heparinize her in order to bridge her back to Coumadin. Consequently, she was heparinized on day 2 following the onset of the SDH. Serial head CT scanning during the next 2 days revealed no change in the size of her SDH and no midline shift, and she was transitioned to Coumadin two days later.
She was subsequently discharged from the hospital without neurologic or cardiac incident, and her GCS remained 14. Her outpatient head CT 19 days after the SDH revealed a slight decrease in SDH size with no midline shift [Figure 1]. She suffered no thromboembolic events following her SDH.
The second patient was an 85-year-old man with pertinent medical history of atrial fibrillation for which he was on Coumadin. He suffered a syncopal fall, resulting in an acute left hemispheric subdural hematoma with maximum thickness of 0.9 cm at the temporal region with no associated midline shift [Figure 2]a. His GCS was 15, and he was neurologically intact with an INR of 1.9. Consequently, he was managed nonoperatively, his Coumadin was discontinued, and his INR was reversed to less than 1.4. He was discharged by the trauma service to rehabilitation four days later, with scheduled outpatient neurosurgery follow-up in 2 weeks with a repeat head CT. During his time in rehab, he developed acute left facial weakness with left upper extremity weakness, and he was consequently readmitted to the neurology service, where workup revealed a right internal carotid artery occlusion (embolic source) with the left aSDH remaining stable in size without midline shift [Figure 2]a. Due to the symptomatic nature of his carotid occlusion, the neurology service decided to start the patient on systemic heparin nine days following the onset of the aSDH; he received no other antithrombotic agent (ie, aspirin, clopidogrel, dipyridamole). His GCS remained 15 during this time.
Five days following heparinization in the setting of a PTT of 113, the patient developed sudden onset of headache with rapidly declining mental status. Head CT [Figure 2]b revealed marked enlargement of the aSDH from 0.9 to 2.4 cm with midline shift of 1.5 cm, after which neurosurgery was consulted. By this time, the patient had deteriorated to a GCS of 4 with bilaterally fixed and dilated pupils and decerebrate posturing in all extremities consistent with uncal herniation. Given his age and neurologic status, the family elected to proceed with comfort measures only. The patient expired shortly thereafter.
Acute subdural hematoma is a potentially lethal condition, with a mortality rate in comatose patients of approximately 60%; this number is even greater in patients older than 65 years. , Standard guidelines dictate that an acute subdural hematoma should be operatively evacuated if either its maximum thickness is greater than 1 cm and/or is associated with greater than 0.5 cm of midline shift.  For SDH not meeting operative criteria, optimal therapy includes normalization of coagulation parameters, close neurologic monitoring, and a repeat head CT within 12-24 h of the initial CT scan in order to assess whether the SDH is enlarging either in size or in associated midline shift. In patients with nonoperative SDH, if systemic anticoagulation is desired secondary to medical comorbidities, it is usually held at least 2 weeks following the SDH due to the risk of precipitating acute SDH expansion. Although the two-week guideline is not evidence-based, it is derived from the observation that anticoagulation increases the likelihood of traumatic SDH development fourfold in men, and 13-fold in women.  Furthermore, a recent study investigating therapeutic anticoagulation for patients with traumatic intracranial hemorrhage refrained on average from initiating anticoagulation until 13 days after initial trauma; in fact, only 13/22 (59%) of patients with traumatic SDH and thromboembolic disease were even started on anticoagulation.  Unfortunately, this case series did not stratify outcomes according to the type of intracranial hemorrhage (SDH, epidural hematoma, subarachnoid hemorrhage, contusion); consequently the study is unable to provide an analysis of traumatic SDH outcomes following anticoagulation.  This series examines just two instances where systemic anticoagulation was initiated sooner than 2 weeks following the aSDH, and follows the outcomes in each instance.
In the first patient, a 75-year-old woman, her mechanical heart valve encouraged the cardiology service to recommend her restarting anticoagulation nearly immediately following her SDH. This was due to the reported risk of prosthetic valve thrombosis in the absence of anticoagulation of 0.3%-1.8% per patient-year, with a mortality of 10% in patients who develop prosthetic valve thrombosis.  Furthermore, the incidence of thromboembolic complications in mechanical heart valves (including death, stroke, and peripheral ischemia requiring surgery) is 4% per patient-year without anticoagulation, compared with 1% per patient-year with anticoagulation. ,, The second patient, an 85-year-old man with atrial fibrillation, suffered symptomatic ICA occlusion approximately 1week after his anticoagulation was discontinued. Given his atrial fibrillation, which has adjusted annual thromboembolic rate of 2.3% without coumadin and 1.3% with coumadin,  the neurology service elected to restart systemic anticoagulation in order to transition the patient back to coumadin, despite realizing the potential for worsening the patient's acute SDH given the reported increase in annual intracranial hemorrhage rate from 0.3% to 0.6% secondary to Coumadin in atrial fibrillation patients without any history of intracranial hemorrhage. 
Consequently, both patients were started on systemic heparin therapy well before such therapy would be deemed compatible with optimal conservative management of their acute subdural hematomas. Although neither suffered any thromboembolic events following heparinization and one patient clinically did very well, the other did suffer an acute SDH enlargement with rapid neurologic deterioration resulting in death. Both patients received frequent neurologic monitoring (q1-2 h) involving dedicated neurologic nursing once heparin was initiated. Also worthy of mention is that the INR of our second patient was normalized following admission, whereasthat of our first patient was not; this may have contributed to the development of stroke in the second patient. Furthermore, while patient one had serial head CTs during the two days of heparinization, the second patient only had a head CT after he neurologically deteriorated at day 5. Althoughin theory an enlargement of the SDH could have been seen had he been scanned within the first two days of heparinization, the neurology service made the decision in advance not to perform a head CT unless his neurologic examination deteriorated while on heparin.
A previous report involving seven patients with mechanical heart valves and concomitant subdural hematoma secondary to anticoagulant use found that of the three meeting conservative SDH management criteria, the mean duration anticoagulation was withheld was 21.3 days, and no patient suffered thromboembolic complications.  In each patient, anticoagulation was restarted only after SDH volume reduction/resolution had been confirmed radiographically by repeat head CT. Given these data, and the potential for catastrophic SDH enlargement (as suffered by our second patient), in retrospect we would have recommended that both patients in this series remain free of systemic anticoagulation for at least 2 weeks, preferably for 3-4 weeks. A recent study examining warfarin for atrial fibrillation patients revealed warfarin to increase the 30-day mortality of intracranial hemorrhage (33% of which were SDH) by 62%, while reducing the 30-day mortality from ischemic stroke by 36%. 
The incidence of acute SDH enlargement within the first 2 weeks following systemic heparin therapy is not known; it is the authors' hope that this study provides a glimpse into the catastrophic neurologic potential of anticoagulation so soon after an acute SDH, even if the SDH does not meet operative neurosurgical criteria. It is our belief that no patient with traumatic acute SDH should receive anticoagulation within the first 2 weeks of injury, given the relatively small risk of developing catastrophic ischemic disease compared with the risk of developing catastrophic enlargement of the SDH, although there is yet to be definitive evidence to either support or oppose this view. It is important to note that the danger of anticoagulation reported in this articleapplies to systemic anticoagulation, and not to subcutaneous anticoagulation for deep vein thrombosis (DVT) prophylaxis, which despite conflicting reports regarding the prophylactic benefit at 24 h after presentation, has been shown to significantly reduce DVT when administered within 48 h of presentation without worsening neurologic outcome. ,
In conclusion, heparinization within 2 weeks of traumatic acute subdural hematoma is an exceedingly risky endeavor, with the consequences of acute enlargement of the subdural hematoma resulting in rapid neurologic deterioration incompatible with life. The risk of stroke in patients such as those in this series (mechanical valve, internal carotid artery occlusion secondary to atrial fibrillation) over a 2-week period must be carefully and rigorously evaluated versus the risk of catastrophic enlargement of an existing subdural hematoma secondary to systemic heparinization.
[Figure 1], [Figure 2]