Cerebral Venous Thrombosis in COVID-19
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.344623
Source of Support: None, Conflict of Interest: None
Keywords: Cerebral venous thrombosis, COVID-19, CVT, SARS-CoV-2
Cerebral venous thrombosis (CVT) is an uncommon cause of stroke. The causes for CVT in a developing country such as India are protean, including infections. The widespread availability of healthcare and the use of potent antibiotics in those with infections have reduced the bacterial causes of CVT. Although viruses are known to induce thrombosis by direct effects (endotheliitis) and through their activation of the inflammatory response, they are uncommonly associated with CVT.
The coronavirus disease 2019 (COVID-19) pandemic caused by the highly infectious severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2 virus) is primarily a respiratory illness but is known to induce thrombosis. Many of the thrombotic complications are known to occur in those with severe COVID-19 and are associated with a poor outcome. Arterial strokes,,, venous thrombosis in the limbs and pulmonary thromboembolism in COVID-19 are well described., However, there is limited literature on CVT associated with COVID-19.,,,,
In this study, we present the data of 18 patients with CVT associated with COVID-19 with an objective to describe the characteristics of these patients with special reference to the onset of CVT in relation to COVID- 19, differences in CVT features in those with severe, and non-severe COVID-19, and outcomes of these patients. We also compared their characteristics and outcome with non-COVID CVT patients admitted during the same period.
Study Design and Participants: This was a retrospective, multicenter observational study. Data from six centers in the city of Mumbai, Western India, were collected. During the study period from April 4 to October 15, 2020, consecutive patients more than 18 years of age who had radiological evidence of CVT manifesting as focal neurological dysfunction, seizure, or alteration of sensorium with a positive real-time reverse transcriptase-polymerase chain reaction (RT-PCR) for SARS-CoV-2 virus were included. The study protocol was approved by the Fortis Hospital Institutional Academic Ethics Committee. Data were anonymized and waiver of written consent for patients was granted.
Data Collection: All patients were tested on day one for SARS-CoV-2 infection by RT-PCR on presentation to the hospital and only those positive were included in the study. Data were collected from each patient for the parameters detailed in Annexure 1.
The primary outcome was discharged from the hospital or death. Those who were discharged were graded for their disability using the modified Rankin scale (mRS) as a secondary outcome. The mRS is a validated disability score coded from 0 (no symptoms at all) through 5 (severe disability) and 6 (death). On discharge, if the patient was asymptomatic or able to carry out daily activities independently despite a mild disability (mRS ≤2), it was classified as a good outcome; mRS >2 was classified as a poor outcome.
We also compared the parameters in those with severe (those fulfilling the criteria for severe and critical COVID-19 as per the World Health Organization [WHO] guidelines) and non-severe COVID-19.
We also compared patients who had CVT but were negative for the SARS-CoV-2 virus admitted during the same study period in the participating centers. Patient demographics, risk factors, imaging features, laboratory parameters, and outcomes were compared between the two groups.
Statistical Analysis: Qualitative data are represented in form of frequency and percentage. Association between qualitative variables was assessed using the Chi-square test, with continuity correction for all 2 × 2 tables, and by Fisher's exact test for all 2 × 2 tables, where the Chi-square test was not valid due to small counts. In the presence of small counts in tables with more than two rows and/or columns, adjacent rows and/or columns, data were pooled and the Chi-square test was reapplied. Continuity correction was applied for all 2 × 2 tables after pooling of data. Fisher's exact test was applied for all 2 × 2 tables where the P value of continuity correction was not valid due to small counts, despite pooling of data. Appropriate statistical software, including but not restricted to MS Excel, PSPP version 1.0.1, was used for statistical analysis. Graphical representation was done in the MS Excel package included in Microsoft Office 365. An alpha value (P-value) of <0.05 was used as the cut-off for statistical significance.
Eighteen patients were included during the study period with a diagnosis of CVT and a positive RT-PCR for the SARS-CoV-2 virus. There was an equal number of males and females in the study. The mean age of the patients was 40.55 ± 12.28 years (range: 27–67 years). The majority of the patients (16) were less than 50 years of age; two patients were above the age of 60 years.
Clinical presentation: Fourteen patients presented with a neurological symptom (seizure or focal neurological deficit or altered sensorium) without the classical symptoms of COVID-19 (fever or respiratory or gastrointestinal). The median time from last normal to presentation with a neurological deficit in these 14 patients was one day. Only four patients had fever at presentation (as a manifestation of COVID-19). Of these, in the two patients with altered sensorium and seizure, fever was present for 1 to 2 days prior to the onset of the neurological symptom; two other patients with focal neurological deficits were detected to have a fever when examined in the emergency room. The neurological dysfunction occurred in these four patients on the day of admission and was the reason for seeking medical care. So, all patients in our study presented with neurological symptoms.
Focal neurological deficit, as a manifestation of the CVT, was seen in 16 patients; in the remaining 2 patients, the seizure was the sole presentation of CVT. Overall, 12 patients had seizures; seven had generalized tonic–clonic seizures, three had partial seizures with awareness, and the remaining two had partial seizures without awareness. Six patients had altered sensorium. Headache was a common symptom in the majority of the patients (15).
An associated abnormality of smell was observed in three patients and an abnormality of taste in four patients. Two patients had hypertension and diabetes; one had chronic renal failure. There was no evidence of meningitis, myalgia, oral contraceptive pill use, or underlying malignancy in any of the patients. Two of these patients regularly consumed alcohol. One patient had ulcerative colitis and a history of left middle cerebral artery infarct. None of these patients had thrombotic complications elsewhere (neither limb thrombosis nor pulmonary embolism) during their stay in the hospital.
Imaging of CVT: The MRI of the brain with venogram revealed superficial venous thrombosis in 16 patients, one had deep CVT, and one patient had combined (superficial and deep) CVT. Six patients had hemorrhagic lesions on MRI. One patient had a right corona radiata infarct along with the superficial CVT.
HRCT of the chest was done in 15 patients between days 1 and 6 of admission to the hospital. It was abnormal in eight patients.
Laboratory parameters observed in these patients are enumerated in [Table 1].
The mean D dimer was 2.2 mg/L (range: 0.2–9.4), mean CRP was 25.6 mg/L (range: 4.9–208) and mean IL-6 was 19.3 pg/mL (range: 2.5–56.6). A mild rise in homocysteine (15–30 μmol/L) was observed in four patients (mean: 16 μmol/L).
Severe versus non-severe COVID-19: Five patients went on to develop severe COVID-19. The observations on comparing the two groups are detailed in [Table 2].
Course of CVT and outcome: All patients were started on the therapeutic dose of low-molecular-weight heparin along with the standard medications for COVID-19. Steroids for treatment of COVID-19 were used in six patients and remdesivir in one patient. Five patients needed ventilation. Two patients died. One of the patients who died developed a left frontal bleed after developing a severe coagulopathy characterized by a platelet count < 50,000 with raised D dimer and CRP 5 days after the admission.
Twelve patients had a good outcome. Various features in those with good and poor outcomes are detailed in [Table 3].
Characteristics of 43 patients with CVT but negative for the SARS-CoV-2 virus who were admitted during the same study period were compared with these COVID-19-infected CVT patients. [Table 4] depicts the differences in parameters between these two groups.
This study is one of the largest series of patients with COVID-19 and associated CVT.
The association of COVID-19 with thrombosis is well-documented now. Pulmonary thromboembolism, and arterial strokes,, are seen in those with COVID-19. Most of these thrombotic complications tend to occur in those with severe COVID-19, and late in the disease course. In our study, 14 patients presented with features of CVT (focal neurological dysfunction or seizure, with or without associated alteration of sensorium) without fever or any of the other typical presenting features of COVID-19 (cough, breathlessness, or diarrhea). In these patients, the diagnosis of COVID-19 was made only because of the mandatory testing for COVID as needed by the hospital policy prior to admission. This contrasts with recent studies where only 7.7% to 14.3% of patients with CVT did not reveal the typical symptoms of COVID-19. The remaining four patients in our study presented with CVT and fever (as a manifestation of COVID-19); the duration of fever was for 1 to 2 days prior to the neurological presentation.
Thirteen patients in our study had non-severe COVID-19. This is in concordance with the presence of CVT even in those with mild-to-moderate COVID-19 in recent studies., Hence, in our cohort, the CVT occurred even in those with asymptomatic or non-severe COVID-19 and early in the disease course.
In the current pandemic, many hospitals rely on HRCT chest to triage patients into suspected COVID-19. In our study, HRCT chest was normal in seven patients (46.7%) (done up to 6 days of admission). The fact that COVID-19 can present as a CVT without other typical features of the disease, even on HRCT of the chest, is important to recognize in the current pandemic to prevent the nosocomial spread of the virus by those patients who present with CVT as the sole presenting feature of COVID-19.
In our study, most of the patients (16) with CVT were less than 50 years of age as has been reported in two other studies., However, another study found CVT in COVID-19 to be commoner in those greater than 50 years of age. Thus, probably, CVT in COVID-19 is seen in all age groups.
The most common presentation of patients in our study was with a focal neurological dysfunction (16), followed by headaches (15), seizures (12), and alteration of sensorium (6). These symptoms and signs, alone or in combination, should make one suspect the possibility of an underlying CVT in those with COVID-19.
The majority of our patients (16) had thrombosis of the superficial venous system. Six patients had hemorrhagic infarct on MRI, which has been seen in other studies also. The site of CVT did not influence the eventual outcome in our study contrary to the data from another series of COVID-19 and CVT.
Abnormality of LFT and raised CRP were markers of the multi-system inflammatory response in those with severe COVID-19. Both the patients who succumbed were in the severe COVID-19 group.
Twelve patients had good outcomes. Poor outcomes were seen in those patients who presented with a worse GCS on admission, abnormal HRCT chest or had severe COVID-19 and needed invasive ventilation.
COVID-19 versus non-COVID-19 CVT: Patients in both the groups were young (less than 50 years), which is in contrast to other studies that observed CVT in elderly COVID-19 patients. Those with COVID-19 and CVT presented with a focal neurological deficits, compared to non-COVID-19 CVT patients. The majority of patients in our study had a reasonable workup for an alternative explanation for their thrombophilic state. Our non-COVID-19 CVT patients were more likely to have at least one risk factor as compared to those with COVID-19-infected CVT (76.4% vs. 44.44%; P value = 0.01), suggesting a possible association between the occurrence of COVID-19 and CVT. The mechanism by which COVID-19 can cause CVT is probably multifactorial: direct infection of endothelial cells, increased concentration of coagulation factors, acquired antiphospholipid antibodies, platelet activation, and decrease in the concentration of anticoagulant proteins.
Recent reports suggest the role of lupus anticoagulant as part of the antiphospholipid antibody (APLA) syndrome as one of the mechanisms by which thrombosis occurs in those with COVID-19. We could test for the APLA syndrome in only seven of our patients, and it was negative in all of them.
LFT was abnormal in those with COVID-19 CVT, suggestive of multi-system involvement due to COVID-19 or the inflammatory response. Mortality and disability outcomes; however, these were not significantly different between the two groups.
Limitations of our study: The number of patients in our study (n = 18) was small. It is possible some patients with CVT may have been missed as it may have occurred in those with severe illness and in the intensive care unit (ICU), where imaging would not have been possible Due to logistic issues during the pandemic, certain laboratory data were incomplete in some subjects.
Our study suggests a possible association between COVID-19 and CVT. CVT can be the presenting manifestation of an underlying COVID-19, which occur early, and even in those with mild disease. Patients with a worse GCS on admission, abnormal HRCT chest, severe COVID-19, and the need for invasive ventilation had a poor outcome.
A waiver of consent was approved by the Ethics Committee. The patient's identity has been adequately anonymized. We acknowledge that the journal will not be responsible for any medico-legal issues arising out of issues related to the patient's identity.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
Annexure 1: Parameters of patients
Age and sex of the patients; presenting complaint of the patient: whether neurological or non-neurological (fever, breathlessness, cough, or diarrhea); time of onset of the neurological complaint prior to admission; in those with non-neurological presentation, the time interval to neurological dysfunction was noted; other associated features were noted such as the presence of headaches, myalgia, abnormality of smell and taste; comorbidities and risk factors for CVT were noted: diabetes mellitus (DM), hypertension (HT), chronic renal failure (CRF), immunosuppression, underlying malignancy, chronic alcoholism, tobacco abuse, past history of transient ischemic attack/stroke, and oral contraceptive pill (OC) use; Glasgow coma scale (GCS) on admission—this was further categorized as mild (14–15), moderate (9–13), and worse GCS (3–8); National Institute for Health Stroke Scale/Score (NIHSS) during peak deficit for those who presented with a focal neurological dysfunction (stroke)—they were classified as mild (1–4), moderate (5–15), moderate–severe (16–20), and severe strokes (21–42) based on the NIHSS scores; those with seizures were categorized as generalized tonic–clonic, focal aware and focal impaired awareness; imaging data for lung involvement (HRCT chest); MRI of the brain with venogram was done in all patients to confirm the CVT and subclassify as superficial (sagittal sinuses, transverse, sigmoid, jugular, and cortical veins), deep (internal cerebral vein, great cerebral vein of Galen, basal vein of Rosenthal, and transcerebral venous system) and combined (combination of superficial and deep vein involvement); both, hemorrhagic infarct and intra-parenchymal hemorrhage due to CVT were considered as a hemorrhagic lesion; other relevant laboratory data: haemoglobin, platelet counts, liver function tests, anti-nuclear antibody (ANA), homocysteine levels, anti-phospholipid antibody panel (lupus anti-coagulant, anti-cardiolipin antibody, and anti-β2 glycoprotein 1 antibody), electroencephalogram (EEG), creatine kinase (CK) levels; D dimer and C-reactive protein (CRP) levels closest to the occurrence of the neurological manifestation and the maximum level of interleukin 6 (IL- 6) were collated. Hemoglobin levels <13 g/dL or >16.5 g/dL in men and < 12 g/dL or >16 g/dL in non-pregnant women, platelet count <1,50,000 μL, homocysteine levels >15 μmol/L, CK > 200 U/L, D dimer >0.5 mg/L, CRP > 10 mg/L, and IL-6 >7 pg/mL were taken as abnormal. Liver function test (LFT) was considered abnormal if SGOT was > 60 IU/L and SGPT was >50 IU/L. Treatment given to these patients for COVID-19 was also noted including if they needed ventilation (invasive or non-invasive). Standard medication changed as the pandemic evolved. In the initial months, it was a combination of hydroxychloroquine and azithromycin; currently it is ivermectin and doxycycline. The use of special medications (steroids, remdesivir, and tocilizumab) for COVID-19 was noted.
Patients were defined as severe COVID-19 when they developed any of the following criteria during the course of their hospital stay (those fulfiling the criteria for severe and critical disease as per the WHO guidelines were considered to have severe COVID-19 in this study): respiratory distress (≥30 breaths/min), oxygen saturation ≤93% at rest, PaO2/FiO2 ≤300 mm Hg (l mm Hg = 0.133 kPa), chest imaging shows obvious lesion progression by >50% within 24 to 48 h, respiratory failure requiring mechanical ventilation, shock, and with other organ failure that required ICU care.
[Table 1], [Table 2], [Table 3], [Table 4]