Nontraumatic extradural hematoma in sickle cell anemia: A rare neurological complication not to be missed

Abstract

A 20-year-old black African-Caribbean male (patient 1), known to have sickle cell anemia (SCA) and on hydroxyurea therapy, presented to our hospital with a one day history of chest and lower back pain. He had a history of multiple previous vaso-occlusive crises requiring simple analgesia. At the time of admission, he was afebrile and hemodynamically stable with no abnormalities detected on physical examination. Laboratory values on admission were: hemoglobin (Hb) 117 g/L (usual baseline Hb 110 g/L), hematocrit 37%, white cell count 14.0 × 109/L, reticulocyte count 216 × 109/L, platelets 303 × 109/L, serum creatinine 49 µmol/L, and total serum bilirubin 75 µmol/L. Sickle cell disease (SCD) results from the recessive inheritance of a mutant beta globin gene in which valine is substituted for glutamic acid at position 6 of the beta-globin chain, resulting in the formation of sickle hemoglobin (HbS). Our patient had sickle cell anemia (SCA), the homozygous form (HbSS) and presented with symptoms suggestive of vaso-occlusion, a hallmark feature of SCD. Vaso-occlusion is caused by the deformed sickle shaped red cells blocking microvasculature leading to tissue infarction and acute inflammation, which results in a painful “crisis” for the patient. At this stage, it appeared that our patient was suffering from a simple unprecipitated vaso-occlusive crisis. Patient 1 was commenced on supportive therapy with simple analgesia, oral fluids, and low molecular weight heparin (LMWH) for prevention of venous thromboembolism. The next day he spiked a fever and was started on intravenous antibiotics because of consolidation on the chest X-ray, consistent with acute chest syndrome. On the same day, he reported a lump on his head over the left parietal area, which was fluctuant and initially measured 2 × 4 cm. The patient denied any headache or trauma. Over the next 24 hr, he deteriorated with ongoing fevers and worsening pain requiring escalation to an opiate based “patient controlled analgesia” pain relief. His laboratory parameters also changed significantly: Hb dropped acutely to 54 g/L with a hematocrit of 17%, and reticulocyte count of 69.9 × 109/L; his platelet count dropped to 75 × 109/L with normal coagulation (PT, APTT, fibrinogen). His C-reactive protein (CRP) level reached a peak of 408 mg/L. The hemolysis markers became markedly elevated: lactate dehydrogenase (LDH) reaching a peak of 3,675 IU/L and bilirubin 200 µmol/L (conjugated bilirubin 29 µmol/L). Three units of packed red blood cells were transfused improving his Hb to 84 g/L. The LMWH was stopped in view of the thrombocytopenia. Twelve hours after the acute fall in Hb, it was noted that his scalp lump had increased in size to 6 × 8 cm and the patient then complained of paraesthesia involving the left side of his chin together with a mild frontal headache. Another hallmark feature of SCD is the hemolytic anemia due to the shortened life-span of the irreversibly sickled RBCs. Patient 1's findings were consistent with an acute hemolytic episode but the massively elevated LDH combined with an acute drop in the platelet count, were findings not commonly encountered in typical vaso-occlusive crises, where the increase in hemolysis is mild to moderate and mild thrombocytosis, more typical. The drop in platelet count raised the possibility of thrombotic thrombocytopenic purpura but the ADAMTS13 level was normal at 80% (normal range 60–123%). Post transfusion, his Hb remained stable with HbS between 45 and 47%, and his hemolytic parameters improved, reticulocytes returned to baseline at 167 × 109/L. His respiratory problems slowly improved during the admission. The rapidly increasing scalp lump was investigated with an ultrasound scan which demonstrated a hematoma and this unusual finding combined with the facial paraesthesia prompted intracranial imaging. Some of the most devastating complications of sickle cell disease involve the cerebrovascular system. Cerebral lesions of ischemic origin account for up to 75% of the neurological manifestations, the remaining 25% are hemorrhagic, mainly intracerebral and subarachnoid hematomas 1. There are very few case reports describing the occurrence of spontaneous (nontraumatic) extradural hematomas as a complication of sickle cell disease. Patient 1 underwent an urgent head computed tomography (CT) scan followed by a magnetic resonance imaging (MRI), which demonstrated: (1) an 11 mm deep extradural hematoma overlying the left frontal lobe (see Fig. 1A,B) with deformation of the subjacent brain but without midline shift; (2) the scalp lesion to be a subgaleal hematoma overlying the left frontal and parietal bones; (3) signal changes in the parietal bone (high T1 and T2, with associated increased diffusion weighted signal) considered to represent bone infarction; and (4) further bone infarction within the left mandibular marrow with a hematoma adjacent to the left condylar/subcondylar region and sigmoid notch which was thought to account for the chin paraesthesia. Throughout his inpatient stay, he remained fully conscious (Glasgow Coma Score, GCS, of 15), alert and oriented, the only focal neurology being reduced sensation in the left chin area. After neurosurgical review, he was managed conservatively with a plan for repeat imaging. At one week, a CT scan demonstrated appropriate evolution and a decrease in the size of the extradural hematoma (depth reduced from 14 to 7 mm), however, no improvement in his subgaleal hematoma. At six weeks, a CT scan showed resolution of both his subgaleal and extra-dural hematomas. A 7-year-old girl (patient 2) with known SCA presented acutely to her local emergency department. Previously, she had two acute episodes requiring blood transfusion, parvovirus-associated anemia, and acute chest syndrome. Her routine transcranial Doppler screening 2 years previously, had shown normal stroke risk for HbSS and she was not taking hydroxyurea. Her parents had found her unrousable when they went to wake her in the morning after she had complained of feeling hot with generalized pain the night before. She had gone to bed with her symptoms relieved by paracetamol, and she had not complained of a headache. There was no history of trauma. On presentation to her local department, her GCS was three and she had a fixed and dilated right pupil. Her laboratory findings showed Hb 86 g/L, hematocrit 25%, white cell count (WBC) 10.6 × 109/L, reticulocyte count 80 × 109/L, platelets 216 × 109/L, serum creatinine 49 µmol/L, total serum bilirubin 21 µmol/L, and CRP 75 mg/L. The main differential diagnosis for this presentation was an acute cerebrovascular accident (CVA). CVAs are a catastrophic complication of sickle cell disease and a leading cause of death in both children and adults. CVAs occur most commonly in the HbSS genotype with an incidence of 0.61 per 100 patient-years. The Cooperative Study of Sickle Cell Disease demonstrated that the incidence of infarctive CVA was highest in children and older patients and the incidence of hemorrhagic stroke was highest in patients aged 20–29 years. Hemorrhagic stroke was associated with a low steady-state Hb and a high WBC. Subarachnoid hemorrhages are less common and extradural bleeds were rare 2. Patient 2 was intubated and CT imaging demonstrated a large right-sided temporal extradural hematoma with significant midline shift of 10 mm (Fig. 2A). She was transferred to our hospital for urgent neurosurgery, and underwent an emergency right-sided craniotomy. A large extradural hematoma was evacuated and dural arterial bleeding controlled. She required urgent exchange blood transfusion and was managed postoperatively in the pediatric intensive care unit. Three days post surgery, she regained consciousness and was extubated. Initially, she had a left-sided hemiparesis and right-sided third nerve palsy. She had one partial seizure affecting the left hand and was started on phenytoin although her EEG was normal. There are 12 case reports describing the occurrence of spontaneous (nontraumatic) extradural hematomas as a complication of sickle cell disease, the first in 1987 3-13. The majority of the previous reports identify skull bone infarction in the same area as the extradural hematoma and therefore most cases have suggested that the extradural hematoma is a complication of periosteal elevation secondary to bone infarction with disruption of the cortical bone margin, and bleeding into the extradural space 3, 5, 8, 9, 11. In half of these cases, a subgaleal hematoma was also present at the time of diagnosis. An MRI was performed to look for an underlying etiology for the extradural hematoma. There was no evidence of an arterial-venous malformation or any evidence of trauma to the skull. MRI interpretation of the bone on the right side was difficult at that stage because of the changes due to the craniotomy. However, she also had a small untreated left-sided extradural bleed with evidence of overlying skull bone infarction (Fig. 2B). She made excellent progress with rehabilitation and was discharged after one month and returned to school. She was transfused regularly for one year post-event to maintain the HbS < 30% and then started on oral hydroxyurea therapy. Her residual neurological problems one year postevent, include an improving right-sided divergent squint causing intermittent diplopia and some selective cognitive impairment; she no longer requires phenytoin. Patient 3 was a 12 year old boy with SCA, undergoing investigation for autoimmune hepatitis and possible pulmonary hypertension. He had normal transcranial Doppler velocities. He underwent a transjugular liver biopsy and one week later was admitted to his local hospital with abdominal pain, deranged clotting (abnormal prothrombin time and APTT, probably related to the autoimmune hepatitis) and confusion. Laboratory values on admission to King's College Hospital were: Hb 63 g/L, hematocrit 18%, WBC 10.0 × 109/L, reticulocyte count 412 × 109/L, platelets 335 × 109/L, serum creatinine 43 µmol/L, total serum bilirubin 77 µmol/L, and LDH 1,040 IU/L. He was treated with opiate analgesia and antibiotics but deteriorated rapidly, with increasing confusion, a decreasing level of consciousness and death. Post-mortem examination showed a large extradural hematoma, with herniation of the brain, and evidence of infarction of the skull overlying the hematoma. These cases illustrate spontaneous extradural hematomas in relation to acute skull infarction in SCA. One of the major questions raised by these cases is the underlying etiology of such an event. All patients had SCA which is known to have a higher incidence of bone infarction and cerebrovascular events 2. The majority of the previous reports identify skull bone infarction in the same area as the extradural hematoma and the commonest interpretation is that extradural hematomas are a complication of bone infarction disrupting the cortical bone, causing periosteal elevation and subsequent bleeding into the extradural space 3, 5, 8, 9, 11. Another interpretation suggested spontaneous rupture of epidural vessels in the vicinity of the infarcted bone. More recently, Dahdaleh offered an alternative mechanism suggesting that patients with SCD have abnormal skull anatomy due to chronic medullary hematopoiesis. In response to an acute anemia, there is rapid hemopoietic tissue proliferation and expansion, which could disrupt the skull cortex and precipitate extravasation of blood and marrow into the subgaleal and epidural spaces 6. Although patient 1 did have evidence of infarcted bone on MRI, he did not complain of a significant headache prior to this event. The hematoma developed contemporaneously with the acute drop in Hb level, which was accompanied by elevated LDH levels between 3,500 and 3,700 IU/L, thrombocytopenia and reticulocytopenia. The skull bone marrow infarction is likely to contribute to the super-elevated LDH and reticulocytopenia. The extradural hematoma in Patient 1 may be secondary to acute expansion of hemopoietic tissue but it could also have developed as a periosteal reaction to the infarcted bone. Patient 2 also had evidence of infarcted skull bone in the affected area but she did not demonstrate a large hematocrit drop so the etiology is more compatible as a complication of periosteal elevation. Patient 3, similarly, had evidence of skull infarction on postmortem. Acute bone infarction due to vaso-occlusion probably causes bleeding and hematoma formation in most cases, although the consequences of this are likely to be fairly insignificant in the vertebrae and long bones including femur, humerus, radius, and ulna, which are surrounded by soft tissue. The skull is very thin and hematomas are likely to be much more obvious, particularly if they occur in relation to the inner surface of the skull and compress the brain. Two of the patients described here also had abnormal blood clotting, one due to the use of heparin and the other related to liver disease. The phenomenon of skull infarction and intracranial bleeding described here is analogous to the complications of orbital vaso-occlusion in SCA, in which orbital hematomas occur and result in proptosis with sight-threatening complications 14. These cases are significant as they highlight the difference in presentation of the same underlying disease process. Case 1 had a good outcome, although Case 2 resulted in cerebral damage and Case 3 caused death. Acute skull infarction is a rare cause of neurological symptoms in SCA, although clinicians should be alert that these complications can occur even when there is no associated headache or focal neurology signs. Acute skull infarctions are often accompanied by spontaneous extradural bleeds. The suggested mechanisms include either disruption to the skull bone structure from acute expansion of hemopoietic tissue, or associated periosteal reaction and bleeding. The cases highlight specific issues for clinicians: first, intracranial complications can occur even when there is no associated headache or neurology; second, an acute drop in hematocrit should alert the clinician to subsequent complications; and third, the presence of subgaleal hematomas should raise suspicion of an underlying associated extradural hematoma and clinicians should have a low threshold for undertaking brain imaging in these cases.