Patients’ basic clinical characteristics
Clinical features including age, body mass index (BMI), pregnancy and reproduction history, gestation, initial symptoms, time from symptom initiation to admission, and acute neurological symptoms are described in Table 1. The mean age of seven patients was 32.4±2.28 years (range 27-42 years), and mean BMI was 21.67 km/m2(range 19.5-23.4 km/m2). All seven patients were within 12 weeks of gestation, five patients (71.45%) were primiparas. N1 and N2 (28.57%) were conceived through assisted reproductive technology. They had not previously been on contraceptive pills, and did not have any personal or family history of venous thrombosis, autoimmune disease, or hematologic disease. The detailed information regarding the initial symptoms, acute neurological symptoms as well as the clinical assay and treatment are summarized as below.
Initial symptoms
All seven patients had been initially suffering from nausea and vomiting accompanied with headache. Five patients (71.45%) complained severe nausea and vomiting with 5 to 10 episodes of vomiting daily which unavoidably made normal eating impossible, leading to maternal weight loss, thus hyperemesis gravidarum (HG) was diagnosed and followed by treatment of rehydration in other hospitals. The other two patients complained of mild nausea and vomiting and chose to make minor adjustments on
daily diet at home. Visual analogue scale (VAS) was used to assess the headache severity of which 0–10 means the severity of pain of which “0” means no such symptoms while “10” indicates the most severe symptoms. The headache complained by the patients was mild at the beginning with an average score of 3.86, however, the headache was described as diffuse and aggravation in severity with an average from 3.86 to 7.16 over several days. Time from symptom initiation to admission was 5.7 days (range 3-10 days).
Acute Neurological symptoms
All seven patients had typical neurological symptoms when they came to our institute due to progressive headache and unresponsive nausea and vomiting. Acute symptoms included epileptic seizures (n=2, 28.6%), limb weakness (n=4, 57.1%), aphasia (n=5, 71.4%), blurred vision (n=2, 28.6%), disturbed consciousness (n=3, 42.9%). Glasgow coma scale (GCS) was used to estimate the neurological state of these patients. They scored with an average of 11.3 (range 3-15) on the Glasgow coma scale (GCS). There was one patient of N4 (14.2%) in severe injury with a GCS score lower than 8, and she was transferred to our emergency department with deep coma by other hospital. Two patients of N6 and N7 (n=2, 28.4%) was in moderate injury with a GCS score 9-12. Both of N6 and N7 were brought to our emergency department by family members, and N6 was reported with behavioral changes and progressing to a global aphasia, N7 presented with aphasia, confusion, and impaired balance. Four patients of N1, N2, N3 and N5 (n=4, 71.4%) were in mild injury with a GCS score 13-15. The clinical symptoms of the involved 7 patients were summarized in Table 1.
Laboratory assay
All 7 patients were performed with comprehensive laboratory assays as listed in Table 2. Liver and kidney function, thyroid function, echocardiography and electrocardiogram were all normal. A complete blood count, chemistry panel, sedimentation rate, and coagulation function are detected for these patients. Laboratory examination revealed a mild inflammatory response with an elevated WBC count of 14.39*109/L, D-dimer were abnormally elevated to 2.97 mg/L (range from 0.8 to 11.6 mg/L), but coagulation profiles were normal. Screening for thrombophilia including protein C, protein S and homocysteine showed that only protein S is slightly lower in N3 (43.8%, normal range 50-80%), and all the coagulation factors were within normal range. Autoimmune diseases with higher risk of hypercoagulation including anticardiolipin antibody, lupus anticoagulant (LA), antinuclear antibody and anti-double strain DNA were also screened and the results showed that only N5 presented with a slightly positive antinuclear antibody (ANA) with 1:160, the rest of the antibody were all negative, and all 7 patients were excluded from autoimmune related disease by rheumatologist. Additionally, N1 and N3 underwent lumbar puncture and elevated opening pressure was revealed but without abnormalities of specific cerebrospinal fluid (CSF).
Neuroimaging examination
All 7 patients underwent neuroimaging examinations by CT or/and MRI/MRV/MRA for CVT confirmation. Computed tomography scan (CT) was performed on 6 patients (85.7%) and revealed cerebral infarction (n=3, 42.9%), intracranial hemorrhage (n=3, 42.9%), subarachnoid hemorrhage (n=3, 42.9%) and encephaledema (n=3, 42.9%), also hyperdense lesion can be found occasionally in the venous sinus indicating thrombosis. MRI can differentiate CVT from tumor apoplexy, and magnetic resonance venography (MRV) can reveal sinus occlusion. Six patients (85.7%) were examed by MRV, which demonstrated CVT on all patients and multiple venous sinuses were involved in most of them: the mean involved venous sinuses was 3.7 (range 2-6), the most involved was transverse sinus (n=7, 100%), followed by sigmoid sinus (n=5, 71.4%), and the least involved was inferior sagittal sinus (n=1, 14.3%), followed by torcular Herophili (n=2, 28.6%). The imaging features of all seven patients were summarized in Table 3.
Digital subtraction angiography (DSA) was conducted for patients when there is a need for endovascular minimally invasive surgery. Three patients of N4, N6 and N7 (n=3, 42.9%) were further proceeded with DSA. DSA confirmed cortical veins and some deep venous structures which CT or MRI/MRV may not reveal, and great cerebral vein thrombosis (n=2, 28.6%) and internal cerebral vein thrombosis (n=1, 14.3%) were demonstrated by DSA.
N4 scored 3 (M1V1 E1) points on the GCS. Emergency cranial CT revealed cerebellar hemorrhage and brain stem edema. She was performed with DSA due to rapid neurological deterioration. DSA confirmed occlusion with great cerebral vein, straight sinus, left transverse sinus and sigmoid sinus, thus immediate interventional thrombolysis was performed, a microcatheter was delivered to the distal thrombosed sinus using a microwire from the jugular bulb and thrombolysis with heparin and urokinase was injected, unfortunately, the occluded sinuses were failed to be recanalized, and N4 deteriorated unavoidably and died within four days.
N6 scored 9 (M4V1 E4) points on the GCS. CT showed left frontal lobe and bilateral parietal lobe hemorrhage, subarachnoid hemorrhage with diffused brain swelling. MRI/MRV revealed almost all intracranial venous sinuses and deep internal cerebral veins were occluded. DSA confirmed a complete superficial and deep venous occlusion. As the representative of extremely serious patients, N6 imaging was showed in Figure 1. N6 experienced rapid consciousness deterioration and herniation signs of left pupil dilatation, although endovascular thrombolysis was attempted and emergency decompressive craniotomy was performed, the brain was extremely extensive swollen with diffused thrombosis of the superficial veins. She died four days post-operation.
N7 scored 11 (M6 V1 E4) points on the GCS. CT venography showed hyper density in left transverse sinus, inferior sagittal sinus, straight sinus and sinus confluence. MRI/MRV showed that massive veins including straight sinus, sinus confluence, great cerebral vein, internal cerebral vein and basilar vein were almost completely occluded. DSA indicated that sinus and deep venous were occluded but basal vein and deep vein have been compensated by sylvian vein reflux, thus endovascular thrombolysis or surgical procedure was not further considered.
CVT anticoagulant treatment
CVT management should be comprehensive including treatment and control of underlying primary disease, thrombolytic and anticoagulant therapy, intracranial pressure reduction, symptom relief and prevention of epilepsy complication. Six patients (85.7%) received dehydration therapy to depress intracranial pressure. Three patients (42.9%) received antiepileptic treatment. All 7 patients were initiated anticoagulation treatment with body-weight-adjusted subcutaneous low-molecular weight heparin (LMWH) instantly within 12 hours of admission as soon as CVT was defined and obvious intracerebral hemorrhage was excluded. When the patient's convulsions were controlled or complaints improved which was consistent with the neuroimaging alteration indicated by partial recanalization on repeat CT or MRI/MRV, LMWH was substituted with oral warfarin in N1, N2, N3 and N5. Warfarin dose was strictly adjusted based on International Normalized Ratio (INR) with a target of 2.0 to 3.0. The most recent patient N7 was treated with LMWH for two weeks followed by oral rivaroxaban (20 mg daily) when her consciousness gradually improved. The period of LMWH bridge to warfarin or rivaroxaban averaged 13.2 days.
Pregnancy outcome and Prognosis
Of the 7 patients, N4 and N6 died within 5 days after admission due to rapid deterioration, other 5 patients progressed well with significant improvement and discharged finally. All these 5 patients requested induced abortion, thus when they were stabilized by evaluation of multidisciplinary team, therapeutic justifiable abortion were performed with a suspension of anticoagulation for 12h, and the average days from the admission to abortion was 6.5 days (range 4-13 days). The mean hospitalization of these 5 patients were 21.4 days (range 18-22 days). N2 and N7 experienced almost complete recovery without discomfort, N1 and N5 presented slight visual ghosting, and N3 discharged with one limb slight weakness. The treatment and outcome of the seven patients were summarized in Table 4.
All these discharged 5 patients were remained anticoagulation for one year and close regular follow-up. N1 and N3 patients continued taking antiepileptic drugs and showed good recovery at one-year follow up. At 6-month follow-up, five patients showed no neurologic or functional deficits. At 6 or 12-month follow-up, MRI/MRV revealed sinuses were recanalized in N2, N3 and N5 as showed in Figure 2.
Literature Review
In order to accumulate and summarize published case reports or case series regarding CVT in early pregnancy, a systemic literature search was conducted using Medline, Embase and Google Scholar using the relevant keywords: “early pregnancy”, “first trimester pregnancy”, “cerebral venous thrombosis (CVT)”, “intracranial venous thrombosis” as well as combinations of these terms. Studies that involved symptoms and management approach on the CVT presenting in early pregnancy were included in our pooled analysis, and totally fifteen patients were included as showed in Table 5 [3-16].
The first case was reported by Robert in 1957. She was misdiagnosed as nervous breakdown at the beginning, died six hours after admission and was found CVT after autopsy. The first 3 cases reported were diagnosed as CVT after postmortem autopsy (two cases in 1957 and one case in 1978). With the advent and development of imaging modalities, all the following 12 cases were accurately diagnosed with neuroimaging examination and appropriately treated. The mean age of fifteen patients was 28.9±2.28 years (range 20-42 years). Nine patients (60%) had unremarkable medical history, the other six patients (40%) had past history of seizure, migraine, antiphospholipid syndrome, oral contraceptive pills, heterozygous factor V Leiden mutation, or CVT. All 15 patients represented typical neurological symptoms, headache is the most common symptom (n=12, 80%), followed by seizures (n=6, 40%), consciousness disorder (n=3, 20%), aphasia (n=2, 13.3%) and hemiparesis (n=2, 13.3). It is noteworthy that NVP (n=5, 33.3%) and HG (n=3, 20%) was also complained in some of those patients. Eleven patients (73.3%) received therapeutic anticoagulation, which included systemic heparin, intravenous heparin, subcutaneous injection of LMWH, warfarin, acetylsalicylic acid and eptifibatide. One patient (6.7%) underwent direct catheter thrombolysis and thrombectomy, two patients (13.3%) experienced decompressive craniectomy. Eight patients (53.3%) experienced a complete recovery, the other four patients (26.7%) showed neurological dysfunction such as one limb weakness, a slight visual field defect, mild neurological deficits and hemiplegia and aphasia. Six patients (40%)maintained a delivery until the third trimester, two of them underwent planned Cesarean section and the newborn babies were completely healthy.