Neurological manifestations in hospitalized COVID-19 patients: a cross-sectional study

Abstract

Background

Accumulating evidence on the neurological sequelae of COVID-19 is a serious concern, with patients possibly being at risk of permanent debilitation if not managed appropriately. We aimed to determine the prevalence and pattern of neurological manifestations and diagnostic and therapeutic findings among hospitalized COVID-19 patients consulted with the neurology service for neurological disorders. We conducted a retrospective, observational study at the Golestan Hospital of Ahvaz, Iran, between March 20, 2020, and March 19, 2021. Patients' demographic, clinical, paraclinical, and therapeutic characteristics were extracted from medical records and then subjected to statistical analysis.

Results

Overall, 6.7% (157/2340) of COVID-19 patients at Golestan Hospital had a neurological disorder. Most of the patients (90/157) were men, and the mean age of patients was 62.91 ± 91 years. A total of 56.68% of patients (89/157) were SARS-CoV-2 RT-PCR positive. The mean chest CT severity score was 8.26 ± 4.4, ranging from 1 to 19. The most common neurologic disorders were cerebrovascular disease (72.6%), encephalopathy (8.9%), and Guillain–Barre syndrome (6.4%). The CSF SARS-CoV-2 PCR test was positive in one patient with Guillain–Barre syndrome. The in-hospital mortality rate was 43.9%. Definite COVID-19, ICU admission, history of stroke and dementia, and comorbidities were associated with an increased mortality risk in these patients.

Conclusions

Patients with COVID-19 can present with serious neurological disorders such as cerebrovascular disease and impaired consciousness, even without typical COVID-19 symptoms. Close monitoring for neurological symptoms may help improve prognosis in hospitalized COVID-19 patients.

Introduction

Almost seven million people have succumbed to coronavirus disease 2019 (COVID-19), representing the largest infectious disease crisis since the 1918 influenza pandemic [1, 2]. In Iran, there have been 7,597,982 documented cases and 145,571 documented deaths from COVID-19 to date [3]. Although the virus primarily targets the respiratory tract, there is widespread evidence of neurological manifestations [4]. Up to a third of people with COVID-19 show at least one neurological manifestation. Headache and dizziness are the most common; more serious manifestations include encephalopathy, meningoencephalitis, stroke, peripheral nervous system disorders, and myopathy. Neurotropism, direct central nervous system (CNS) invasion, and postinfectious neurological complications have been suggested as causes of these manifestations [5].

Accumulating evidence on the neurological sequelae of COVID-19 is a serious concern, with patients possibly being at risk of permanent debilitation if not managed appropriately [6]. Conversely, patients with a previous diagnosis of chronic neurological diseases may be at increased risk of serious COVID-19 [7]. Also, the probability of COVID-19 continuation is high due to the emergence of new viral mutations [8, 9]. In light of these factors, there is an urgent need to study the neurological effects of COVID-19. Hence, this study aims to identify the prevalence and the patterns of neurological manifestations among patients with COVID-19 and determine the predictors of mortality and debilitating neurological complications.

Methods

Study design

We conducted a retrospective, cross-sectional study of all patients hospitalized between 20 March 2020 and 19 March 2021 at the Department of Neurology of Golestan Hospital (Ahvaz, Iran) diagnosed with COVID-19 and neurological symptoms.

Definitions

Patients were classified as suspected, probable, or definite COVID-19 patients. Suspected cases met the related clinical and epidemiological criteria. Probable cases included suspected patients with a history of contact with a probable or definite patient, a cluster of patients with at least one definite case reported among them, suspected patients with imaging findings in favor of COVID-19, such as multi-lobular or bilateral infiltration, especially peripheral infiltration on chest CT scan or chest radiographic or/and ground glass on CT scan of the lung, patients with acute loss of sense of smell and taste, and death in a patient suspected of COVID-19 that is not justified by other reasons. Definite cases had a positive PCR result for SARS-CoV-2 infection, regardless of symptoms and signs [10, 11].

The neurological syndromes were categorized as cerebrovascular diseases, encephalopathy (impairment of consciousness without localized signs, presenting as confusion, lethargic status, delirium, or coma), meningitis, and meningoencephalitis (encephalopathy with inflammatory CSF), epilepsy, demyelinating disease, and peripheral nervous system syndromes (involving muscles, neuromuscular junction, nerves or roots).

The chest computed tomography (CT) severity score was calculated to describe the extent of pulmonary involvement; this method was based on the degree of involvement of the lung lobes as 0% (0 points), 1–25% (1 point), 26–50% (2 points), 51–75% (3 points), and 76–100% (4 points). The CT severity score was quantified by summing the five lobe indices (range 0–20) [2].

Treatment

Patients for COVID-19 were treated based on national COVID-19 treatment guidelines with medications like corticosteroids, hydroxychloroquine, atazanavir/ritonavir, remdesivir, and ReciGen® [12].

Data collection

All patients' information including demographic characteristics (age, sex), neurological presentations, clinical signs of respiratory infection, the interval between onset of neurological manifestations and clinical signs of respiratory infection, paraclinical and experimental findings, history of previous neurological disease, history of comorbidities, history of immunosuppression and corticosteroids (> 20 mg of prednisolone for > 2 weeks or cumulative dose of > 600 mg of prednisolone), therapeutic measures, and in-hospital outcomes (in-hospital mortality, discharge, and length of stay) were extracted from the medical records. Finally, the outcome of COVID-19 was also expressed as mortality or discharge from the hospital.

Statistical analysis

Statistical analysis was performed using SPSS version 21 software. The Shapiro–Wilk test was used to determine the normality of the distributions. Simple descriptive analysis of data was summarized as frequencies (%) and means ± standard deviations (SD) for normally distributed variables or medians with interquartile ranges (IQR) for non-normally distributed parameters. Normally distributed parameters were compared by the student t-test and others by the Mann–Whitney U test. Categorical variables were compared by chi-squared or Fisher's exact tests, where appropriate. A P-value of < 0.05 was considered significant.

Results

General findings

A total of 2,340 adult patients were admitted to different wards of Golestan Hospital with an initial diagnosis of COVID-19 during the study period; 157 (6.7%) presented with a serious neurological disorder and were  consulted with or admitted by the neurology service. Of these, 90 (57.3%) were men. The mean age was 62.91 ± 91 years (median: 64; range: 19–99 years). Most of the patients (65%) were older than 60. The mean duration of hospitalization was 10.44 ± 8.43 days. Dyspnea (37.6%) and cough (24.2%) were the most respiratory symptoms. Hypertension and diabetes mellitus were the most common comorbidities (71.3% and 36.9%, respectively). Most patients (68.8%) had no pre-existing neurological diseases; nevertheless, ischemic stroke and myasthenia gravis were seen in 20.4% and 3.8% of patients, respectively. Forty-nine percent of patients (77/157) had a neurologic manifestation several days after COVID-19 symptoms. Most patients presented with a focal neurological deficit (48.4%) or loss of consciousness (26.8%). Table 1 presents a summary of the characteristics of the study participants.

Table 1 Demographic, clinical, and laboratory findings among hospitalized COVID-19 patients with neurologic disorders according to hospital outcome

Figure 1 depicts the frequency of different neurological disorders in our patients. The most common were cerebrovascular disease (114/157, 72.6%), peripheral nervous system disorder (10.8%), and encephalopathy (8.9%). No patient had myelitis or myelopathy related to SARS-CoV-2 infection (Fig. 1).

Fig. 1

Frequency of neurological disorders among hospitalized COVID-19 patients with neurological disorders

Table 2 summarizes the patients' neuroimaging, lung imaging, electromyography/nerve conduction velocity, cerebrospinal fluid analysis, and medications. Evidence of encephalitis (n = 1) and leptomeningeal enhancement (n = 1) were occasionally seen on brain MRI, and two patients had cerebral venous sinus thrombosis (CVST) in the superior sagittal sinus and transverse sinus (Table 2).

Table 2 Summary of patients’ neuroimaging, lung imaging, electromyography/nerve conduction velocity, cerebrospinal fluid analysis, and medications

The in-hospital mortality rate was 43.9% (69 patients), compared with 31.2% (730/2340) of all hospitalized COVID-19 patients. A definite COVID-19 diagnosis, hospitalization in the COVID-19 ICU or internal medicine wards, previous history of stroke and dementia, and co-morbidities were significantly more common in those who died than in those who survived (P < 0.05). Moreover, leukocytosis, thrombocytopenia, anemia, uremia, positive troponin, and high titer of creatine phosphokinase (CPK) and D-dimer were associated with an increased risk of mortality in these patients (P < 0.05).

Ischemic stroke

Among 114 patients with cerebrovascular disease, an ischemic stroke was most common (74.5%; n = 85) and was seen in patients with a median age of 61 years, ranging from 28–99 years. Most (39%) were males. The mortality rate was 42.5% in these patients. Seventeen patients (20%) had COVID-19 symptoms simultaneously with stroke, while 13 patients (15.3%) developed COVID-19 symptoms during hospitalization after a stroke was diagnosed. The mean chest CT severity score was 8.51 ± 4.49. A focal neurological deficit was the most neurologic presentation (75%), while hypertension (75%) was the predominant risk factor. Most ischemic infarcts (56.5%) were in the middle cerebral artery territory. All patients with ischemic stroke received prophylactic anticoagulants for deep vein thrombosis subcutaneously, though 11 developed hemorrhagic changes during their hospital course. A total of 12 patients out of 85 (14%) with ischemic stroke received a brief course of low-dose corticosteroids, out of whom 5 died (41%). This mortality proportion was similar to that among ischemic stroke patients who did not receive corticosteroids (42%; 31 out of 73 patients; P = 0.5) (Table 3).

Table 3 Characteristics of COVID-19 patients with acute ischemic stroke (n = 85)

Intracranial hemorrhage

Intracranial hemorrhage (ICH) was the less prevalent type (25.5%; n = 29) of cerebrovascular disease. Most cases featured intraparenchymal or intraventricular hemorrhage (55.17%), predominantly in the basal ganglia. Like ischemic stroke, most patients (51.7%) were males, and the median age was 61 years (range 36–83). Most (58.5%) had COVID-19 symptoms several days before hospitalization. Mortality occurred in 15 out of 29 patients with an intracranial hemorrhage (51.5%). In two patients diagnosed with CVST, treatment with intravenous anticoagulants was prescribed. Patients diagnosed with ICH received deep vein thrombosis (DVT) prophylaxis in the form of subcutaneous anticoagulants. Considering the lower mortality proportion in patients with ischemic stroke, the prognosis was better in ischemic stroke patients than in hemorrhagic ones (Table 4).

Table 4 Characteristics of COVID-19 patients with intracranial hemorrhage (n = 29)

Encephalopathy

Among fourteen patients with encephalopathy, six had a definite COVID-19 diagnosis. The median chest CT severity score was 9, and 85.5% received invasive mechanical ventilation. Brain CT showed generalized brain edema in two patients. The blood culture was positive for Klebsiella in one patient and for Pseudomonas in another. No leukocytosis was seen in the CSF analysis. The mortality rate in these patients was 78.5%.

Seizure

A total of 9 patients were admitted with the initial manifestation of seizure, among which four were diagnosed first with COVID-19. None had a prior history of seizures or epilepsy. Among them, two patients had encephalitis (both died), three had an ICH, one had a CVST (died), and three were diagnosed with an epileptic syndrome (one died). Excluding status epilepticus, all other patients with seizures responded well to a single appropriate antiepileptic drug. Two patients with status epilepticus, unresponsive to adequate doses of two or three antiepileptic drugs, were treated with intravenous midazolam, resulting in a satisfactory response. It seems that in 5 patients, the seizures were related to COVID-19, and in 4 patients, they were related to a cerebrovascular accident. A total of 4 patients died (44.4%) despite proper treatment.

Guillain–Barré syndrome (GBS)

Ten patients (6.4%) with a mean age of 49.10 ± 15.22 years (range 25 to 69) were diagnosed with GBS. Most were women (60%). Eight had a definite COVID-19 diagnosis. Two patients were in poor condition and required intubation. All of them had COVID-19 symptoms several days before neurological presentation. The protein concentration in CSF was high, with a median of 119 mg/dl, and there was no evidence of CSF leukocytosis. The CSF analysis was positive for SARS-CoV-2 (PCR) with elevated protein and no leukocytosis in one GBS patient—a 69-year-old woman with severe COVID-19 (chest CT score = 18). This patient eventually died. Six patients received plasma exchange therapy, three had contraindications (e.g., instability) and received IVIG instead, and one had only mild symptoms and was monitored without immunotherapy. Four patients received low-dose corticosteroids and antiviral drugs simultaneously for COVID-19. Among the four patients who received low-dose corticosteroids, one patient died. The overall mortality proportion among patients with GBS was 20%.

Myasthenia gravis (MG)/myopathy

Seven patients developed an exacerbation of MG (n = 6) or myopathy (n = 1). Three had a definite COVID-19 diagnosis; they had respiratory symptoms preceding muscle weakness, and only one patient developed severe respiratory involvement and died. All patients with MG exacerbation had a positive history of MG, but the patient with irritative myopathy had no history of myopathies or rheumatologic disorders. Regarding MG, two patients had a severe clinical course and required immunotherapy—one received plasma exchange, and the other received IVIG as plasma exchange was contraindicated. One patient received only rituximab. Three patients had a relatively mild clinical course and improved with treatment for COVID-19 (including antiretrovirals). One patient (1/7; 14%) died.

Encephalitis

Four patients were diagnosed with encephalitis, out of which two were definitively diagnosed with COVID-19. The patients exhibited cognitive impairment and fever, requiring hospitalization. Only one patient required mechanical ventilation, and all patients with encephalitis received a triple-drug regimen for encephalitis along with simultaneous antiviral treatment for COVID-19. Unfortunately, one patient died.

Discussion

The COVID-19 pandemic has affected people from all walks of life, burdening patients, healthcare professionals, and governments in many countries [13]. Despite numerous reports of neurological deficits in patients with COVID-19 worldwide, the exact occurrence of these manifestations remains unknown. Patients with severe COVID-19 tend to have more severe neurological manifestations, possibly from cerebral hypoxia due to respiratory failure [7, 13]. We found that 6.7% of hospitalized COVID-19 patients had a serious neurologic disorder; the most common ones were ischemic stroke (54.1%), intracranial hemorrhage (18.5%), encephalopathy (8.9%), Guillain–Barre syndrome (6.4%) and MG exacerbation (3.8%).

The prevalence of neurologic manifestations in hospitalized COVID-19 patients was estimated at 36.4% in China and 57.4% in Europe [5, 7, 13, 14]. In a previous study from Iran, 4.8% (211/4372) of hospitalized COVID-19 patients had a neurological disorder; the most frequent CNS manifestations were headache (40.3%), reduced consciousness (36%), and focal neurologic symptoms (18%) [2]. In New York City, a slightly higher prevalence of 13.5% (606/4491) was reported [15]. The reported prevalence rates are heterogeneous depending on the methods and definitions for detecting neurological manifestations. For instance, the prevalence recorded in the present study (6.7%) is lower than most of the mentioned studies, probably due to our selection of serious neurological disorders requiring consultation with or admission by the neurology service.

Neurologic symptoms are sometimes the initial presentation of COVID-19 [8]; this was seen in 30% (47/157) of cases in the present study. However, most patients developed neurological manifestations after several days of onset of clinical COVID-19 symptoms. Liotta et al. reported neurologic manifestations at COVID-19 onset in 42.2%, at hospitalization in 62.7%, and at any time during the disease course in 82.3%. The most frequent neurologic manifestations were myalgias (44.8%), headaches (37.7%), and encephalopathy (31.8%); strokes, movement disorders, motor and sensory deficits, ataxia, and seizures were uncommon [16]. Thirty-one percent of patients in our study had a previous history of neurological disease—mainly stroke, dementia, or myasthenia gravis. In the Herman et al. study, the rate of previous neurological illnesses varied from 0 to 40%, with a pooled percentage of 8%; cerebrovascular disease was the main comorbidity (16%), similar to our study [17]. Patients with a preceding neurological disease may be at increased risk of serious COVID-19, so special attention must be directed toward preventing COVID-19 in this population [7].

According to the current study, CNS manifestations were the most prevalent serious neurological manifestation in hospitalized COVID-19 patients, with ischemic stroke being the most common (54.1%), followed by intracranial hemorrhage (18.4%) and encephalopathy (8.9%). Overall, 72.6% of patients had cerebrovascular disease, representing the most common group of neurological disorders. Stroke was reported at 1.7% in Spain [14], 3.5% to 5.7% in China [5], and 62% in a report from the UK [18]. Studart-Neto and colleagues reported a prevalence of encephalopathy (44.4%) and stroke (16.7%) in Brazil in hospitalized COVID-19 patients requiring neurology consultations [19]. A similar study by Brucki et al. in Latin America reported stroke as the most common neurologic disorder (47.6%), followed by encephalopathy (27%) [20]. In New York City, the main neurologic disorder was toxic/metabolic encephalopathy (51%), followed by stroke (14%), seizure (12%), and hypoxic/ischemic brain injury (11%) [15]. Although our findings are generally compatible with the literature, cerebrovascular disease might have been more prevalent in our study because our center is the referral center for stroke in Ahwaz, and because the internal medicine team sometimes manages encephalopathies without a neurological consultation. Other potential factors for the difference in the prevalence and frequency of neurological manifestations include SARS-CoV-2 strain variations and genetic factors, particularly polymorphisms in the angiotensin-converting enzyme 2 (ACE 2) gene [21, 22].

Possible mechanisms behind impaired consciousness and encephalopathy include infections, parenchymal lesions, electrolyte imbalance, hypoxia, toxic and metabolic encephalopathies, and non-convulsive epilepsy [13]. One suggested explanation for the increased ischemic and vascular events in COVID-19 patients is the reduction of ACE2 by the SARS-CoV-2 virus, which may cause an imbalance of the renin-angiotensin system (RAS), eventuating in endothelial cell dysfunction and ischemic accidents. ACE2 dysregulation by SARS-CoV-2 may also lead to vasoconstriction and dysfunction of the brain, resulting in increased blood pressure, which can trigger bleeding via the rupture of arterial walls [23]. SARS-CoV-2 also ruptures existing atherosclerotic plaques via the inflammatory response involving CRP, interleukin (IL-7), interleukin 6 (IL-6), and other inflammatory markers [24]. Cardiac manifestations and arrhythmic complications of COVID-19 can also contribute to ischemic events [24], as can coagulation disorders, which have been seen in patients with severe COVID-19 [25].

The overall prevalence of peripheral nervous system manifestations was less than about 10.8% in our study. GBS was diagnosed in 6.4% of patients, with a median age of 44 (25 to 69) years. Most developed neurological symptoms several days after COVID-19. Frontera et al. reported three patients with GBS within 2–4 weeks of documented SARS-CoV-2 infection; all three had a negative RT-PCR SARS-CoV-2 in the CSF [15]. While GBS is thought to occur via postinfectious or parainfectious mechanisms, we found a rare case of GBS with a positive CSF SARS-CoV-2 PCR test in a 69-year-old woman with severe COVID-19. This rare finding, coupled with evidence of ischemia on imaging, suggests the possibility of direct CNS invasion by SARS-CoV-2, which should be explored further in future studies.

In the current study, 50 (32%) patients were in poor condition upon hospitalization, and 81 (52%) patients received invasive mechanical ventilation during hospitalization. Our in-hospital mortality rate was 44%. A definite COVID-19 diagnosis, ICU admission, previous history of stroke and dementia and associated morbidity, uremia, anemia, thrombocytopenia, leukocytosis, positive troponin, and high titers of CPK and D-dimer were associated with increased risk of mortality. In a similar study, Frontera et al. found higher in-hospital mortality rates and lower discharge rates among patients with COVID-19 with a neurologic disorder than those without a neurologic disorder. History of medical and neurologic diseases, the severity of illness (invasive mechanical ventilation and SOFA scores), and acute renal failure differences in COVID-19-specific medication administration (therapeutic anticoagulation, hydroxychloroquine, corticosteroid, lopinavir/ritonavir use) were significantly associated with in-hospital death in those with a neurologic disorder (35%) than in those without a neurologic disorder (19%) [15]. Moreover, Yaghi et al. found that patients with laboratory-confirmed COVID-19 and ischemic stroke had higher mortality rates than contemporary and historical patients with ischemic stroke who did not have COVID-19 [26]. Hence, our findings align well with the literature.

There are limitations to this study that should be highlighted. First, although our sample size was large, the study's retrospective nature resulted in many patients being lost due to a lack of requesting a neurology consultation. Second, the current study included all patients diagnosed with COVID-19 based on clinical suspicion (with and without laboratory confirmation), meaning that undocumented diagnoses or unconfirmed SARS-CoV-2 infection may lead to inaccurate estimates of the prevalence rate. Third, some patients may not have provided an accurate history of neurological symptoms due to the severity of pulmonary symptoms, which can underestimate the prevalence. Finally, in most cases, it was impossible to determine whether COVID-19 caused new neurological symptoms or worsened pre-existing ones due to insufficient information in the medical records. Well-designed prospective studies can overcome these limitations.

Conclusion

In our study, over 5% of hospitalized COVID-19 patients developed life-threatening neurological complications. A definite COVID-19 diagnosis, a previous history of neurological disease, and comorbidities predicted mortality in these patients. Without proper care and treatment, COVID-19 patients with neurological disorders may become permanently disabled, burdening individuals, families, societies, and healthcare systems. As a result, we urgently need to understand and respond to the latent yet potentially debilitating neurological consequences of COVID-19 in the acute and chronic stages of the disease.