This review yielded 143 original publications reporting CNS and PNS involvement by COVID-19, with the selected characteristics, alone or combined (see Figure 1).
The included studies were essentially observational and descriptive. From the total, 119 focus on clinical manifestations, 62 describe neuroimaging findings, 60 studies report cerebrospinal fluid results, 51 describe pathophysiologic mechanisms of CNS and PNS involvement by COVID-19, 28 report electrophysiology findings, and four describe neuropathology findings. Among all studies reviewed, we found a total of 10723 patients with a confirmed diagnosis of COVID-19 who displayed features compatible with neurological involvement. Among them, we found 1633 patients with specific nosological entities or clinical conditions affecting the central nervous system and 43 the peripheral nervous system (Table 1). The remaining 9047 patients reportedly presented with one or more neurological signs and symptoms not attributed to a specific clinical condition or nosological entity.
Table 1. Neurological conditions associated with COVID-19.
Clinical conditions associated with COVID-19 affecting the central nervous system
|
No. of patients / Proportion
|
Encephalopathy
|
990 (60·7%)
|
Unspecified stroke type
|
416 (25·5%)
|
Ischemic stroke
|
159 ( 9·7%)
|
Hemorrhagic stroke
|
40 (2·4%)
|
Encephalitis and meningoencephalitis
|
19 (1·2%)
|
Acute disseminated encephalomyelitis (ADEM)
|
4 (0·2%)
|
Venous sinus thrombosis
|
3 (0·2%)
|
Multiple sclerosis exacerbation
|
2 (0·1%)
|
Total
|
1633 (100%)
|
Clinical conditions associated with COVID-19 affecting the peripheral nervous system
|
No. of patients / Proportion
|
Guillain-Barre syndrome
|
22 (51·2%)
|
Other cranial nerve disorders
|
12 (27·9%)
|
Facial palsy (Bell syndrome)
|
5 (11·6%)
|
Miller-Fisher syndrome and polyneuritis cranialis
|
4 (9·3%)
|
Total
|
43 (100%)
|
A total of 8885 (86,3%) reports of neurological signs and symptoms were related to CNS (Table 2), while 1414 (13,7%) were related to PNS (Table 3).
Table 2. Distribution of signs and symptoms indicating central nervous system involvement in patients with COVID-19
Signs and symptoms indicating central nervous system involvement
|
No. of patients / Proportion
|
Diffuse compromise
|
8129 (91·5%)
|
Psychiatric symptoms (including anxiety disorders, mood disorders, psychosis, and insomnia)
|
4981
|
Headache
|
1805
|
Dizziness
|
527
|
Consciousness impairment
|
416
|
Delirium
|
340
|
Nausea/vomiting
|
16
|
Nuchal Rigidity
|
4
|
Non-specific combination of signs and symptoms
|
40
|
Focal deficit
|
410 (4·6%)
|
Extrapyramidal disorders
|
279
|
Corticospinal tract impairment
|
61
|
Ataxia
|
18
|
Dysarthria
|
13
|
Amnesia
|
12
|
Aphasia
|
7
|
Monoparesis
|
6
|
Central facial weakness
|
5
|
Myoclonus
|
5
|
Homonymous hemianopia
|
4
|
Seizures
|
346 (3·9%)
|
Non-specified seizures
|
324
|
Generalized seizures
|
9
|
Non-convulsive status epilepticus
|
6
|
Focal seizures
|
3
|
Seizure-like events (abnormal involuntary movements)
|
3
|
Non-epileptic convulsive syncope
|
1
|
Total patients with CNS signs and symptoms
|
8885 (100%)
|
Table 3. Distribution of signs and symptoms indicating peripheral nervous system involvement in patients with COVID-19.
Signs and symptoms indicating peripheral nervous system involvement
|
No. of patients / Proportion
|
Smell/taste impairment
|
746 (52·8%)
|
Anosmia + Ageusia
|
477
|
Anosmia/hyposmia
|
128
|
Ageusia/dysgeusia
|
141
|
Visual impairment
|
9 (0·6%)
|
Unspecified decreased visual acuity
|
8
|
Complete visual loss
|
1
|
Oculomotor impairment
|
14 (1%)
|
Ophthalmoparesis
|
7
|
Diplopia
|
3
|
Anisocoria
|
1
|
Bilateral mydriasis
|
1
|
Bilateral Abducens Palsy
|
1
|
Unilateral Abducens Palsy
|
1
|
Facial palsy
|
13 (0·9%)
|
Bilateral weakness/diplegia
|
7
|
Unilateral
|
6
|
Other cranial nerve impairment
|
32 (2·3%)
|
Glossopharyngeal neuralgia
|
9
|
Trigeminal neuralgia
|
8
|
Tinnitus
|
5
|
Decreased hearing
|
5
|
Vasoglossopharyngeal neuralgia
|
2
|
Dysphagia
|
2
|
Reduced tongue movements/tongue deviation
|
1
|
Peripheral neuropathies involving trunk and limbs
|
353 (24·9%)
|
Mixed neuropathy
|
247
|
Pure sensitive impairment
|
31
|
Paresthesia
|
30
|
Hypoesthesia
|
1
|
Pure motor impairment
|
40
|
Areflexia
|
14
|
Distal weakness
|
8
|
Tetraparesis
|
7
|
Gait difficulties/instability
|
6
|
Paraparesis
|
3
|
Tetraplegia
|
1
|
Paraplegia
|
1
|
Neuralgia
|
8
|
Limb neuralgia
|
7
|
Occipital neuralgia
|
1
|
Dysautonomia manifestations
|
27
|
Nerve roots and plexus disorders
|
145 (10·3%)
|
Myopathic involvement
|
102 (7·2%)
|
Total patients with PNS signs and symptoms
|
1414 (100%)
|
Figure 2 summarizes the main neuroimaging findings associated with COVID-19, while Table 4 represents the main CSF findings. Most of the included studies did not report mortality rates explicitly associated with neurological involvement by SARS-COV-2.
Table 4. Cerebrospinal fluid (CSF) findings in patients with COVID-19 and different neurological conditions
Cerebrospinal fluid findings in patients with central nervous system involvement by SARS-CoV2
|
|
Clinical manifestation
|
N° patients
|
Proteins (mg/dL)
|
Cells (cells/µL)
|
SARS-CoV-2 PCR
|
|
Average
|
Range
|
Average
|
Range
|
Positive
|
Negative
|
N/M
|
|
|
Encephalopathy
|
13
|
67.3
|
0 to 230
|
9.8
|
0 to 37
|
0
|
9
|
4
|
|
Encephalitis
|
7
|
83.1
|
19 to 200
|
40.6
|
0 to 115
|
2
|
5
|
0
|
|
Demyelinating lesions in brain and spine
|
4
|
47
|
32 to 62
|
1
|
1
|
1
|
1
|
2
|
|
Meningoencephalitis
|
3
|
463.5
|
461 to 466
|
17.3
|
0 to 21
|
1
|
2
|
0
|
|
Stroke
|
3
|
50
|
50
|
0
|
0
|
0
|
2
|
1
|
|
Seizures
|
3
|
66
|
66
|
3
|
1 to 5
|
0
|
3
|
0
|
|
Brain vasculopathies
|
2
|
292.5
|
78 to 507
|
6.5
|
0 to 13
|
0
|
2
|
0
|
|
Rhombencephalitis
|
1
|
42.3
|
42.3
|
0
|
0
|
0
|
0
|
1
|
|
Maniac episode
|
1
|
190
|
190
|
0
|
0
|
0
|
1
|
0
|
|
Total
|
37
|
|
4
|
25
|
8
|
|
Cerebrospinal fluid findings in patients with peripheral nervous system involvement by SARS-CoV2
|
|
Clinical manifestation
|
N° patients
|
Proteins (mg/dL)
|
Cells (cells/µL)
|
SARS-CoV-2 PCR
|
|
Average
|
Range
|
Average
|
Range
|
Positive
|
Negative
|
N/M
|
|
Guillain-Barré syndrome
|
17
|
105.9
|
40 to 193
|
2.4
|
0 to 9
|
0
|
11
|
6
|
|
Miller-Fisher syndrome
|
2
|
75
|
62 to 80
|
2.5
|
0 to 5
|
0
|
1
|
1
|
|
Polineuritis cranealis
|
1
|
62
|
62
|
2
|
2
|
0
|
1
|
0
|
|
Facial palsy
|
2
|
44
|
44
|
0
|
0
|
0
|
2
|
0
|
|
Total
|
22
|
|
0
|
15
|
7
|
|
CSF glucose levels were normal or slightly increased in all reports. Cell differential showed lymphocytic predominance in all reports. N/M: Not measured. CNS: Central nervous system. PCR: Polymerase chain reaction. CSF: Cerebrospinal fluid.
The reviewed studies reflect a significant prevalence of nervous system' involvement in patients with COVID-19. Neurological manifestations appear in a range of 22.5% to 36.4% of all COVID-19 patients among different studies [6, 11–14]. We classified them in diffuse and focal CNS signs and symptoms, seizures, cranial nerve impairment, encephalopathy, neuroinflammatory disorders, acute cerebrovascular disease, and peripheral neuropathies.
Diffuse CNS signs and symptoms
Fifty-three studies reported 8129 diffuse signs and symptoms of CNS involvement by COVID-19, including neuropsychiatric disorders (61.3%), headache (22.2%), dizziness (6.6%), consciousness impairment (5.2%), delirium (4.3%) nausea/vomiting (0.3%), and nuchal rigidity (0.1%). Psychiatric symptoms included anxiety, mood disorders, psychosis, insomnia, and others. These symptoms are described in depth in other studies [11, 15, 16] and are not the focus of this review. Headache is, indeed, one of the most common neurological manifestations of SARS-CoV-2 infection, with a variability range of 8 to 39% of cases [13, 17]. Headache can be a primary process in these patients or part of a broad spectrum of neurological syndromes such as meningitis, encephalitis, vasculitis, elevated intracranial pressure, and other clinical conditions associated with COVID-19's neuroinflammatory mechanisms and other underlying systemic causes [7]. Impairment of consciousness and arousal is another common neurological disturbance, documented in up to 37% of patients with COVID-19 as a manifestation of encephalopathy [6, 7]. Delirium was present in 20 to 65% of patients with SARS-CoV-2 infection[18]. It can be attributed directly to SARS-CoV-2 invasive mechanisms to the CNS, leading to a neuroinflammatory response or a multifactorial compromise secondary to sedative therapies, mechanical ventilation, and environmental factors, including social isolation [15]. On the other hand, delirium in critically ill patients with COVID-19 may be a prodromal symptom of infection and hypoxia secondary to severe respiratory failure [15] or an isolated manifestation of COVID-19 [19]. Delirium can also overlap an underlying cognitive impairment, which generates a baseline vulnerability state. However, the elevation of inflammatory markers indicates a concomitant immune response as a precipitant [18]. Furthermore, a history of delirium can increase the risk of post-intensive care syndrome, including cognitive impairment, mental state disorders (such as depression, anxiety, and post-traumatic stress disorder), and physical impairment after leaving the intensive care unit [15]. Dizziness is also a prevalent neurological manifestation, ranging between 7 and 9.4% of patients admitted to intensive care units [20] and 26.21% in general series [6, 17]. Nausea and vomiting are also common neurologic manifestations, with an estimated prevalence of 5% [4]. Their presence may be related to an impairment in the CNS structures related to emesis control in the dorsal vagal complex in the medulla oblongata caused by SARS-CoV-2 [21].
Focal CNS signs and symptoms
Twenty-eight studies reported 410 patients with focal neurological disturbances, including corticospinal and corticobulbar tract impairment, ataxia, dysarthria, amnesia, aphasia, retrochiasmatic visual field alterations, and extrapyramidal disorders. Most of these disturbances were associated with stroke in patients with COVID-19 and are discussed below in this article.
Seizures
Twenty-eight studies reported seizures in 346 patients with COVID-19. Most reports did not specify the seizure type in 324 patients, while a few studies documented generalized or focal seizures [22–30], focal or diffuse non-convulsive status epilepticus [31], seizure-like motor events [22], and non-epileptic convulsive syncope [32] (Table 2). Although COVID-19 patients may present seizures due to hypoxia, metabolic derangements, organ failure, or cerebral damage [33], SARS-CoV-2 systemic infection per se represents a minimal risk for seizures during acute illness [34]. In a retrospective multicentric study aiming to evaluate the incidence and risk of acute symptomatic seizures in 304 patients without a prior history of epilepsy, there were no new-onset seizures or status epilepticus during the acute phase of COVID-19 [34]. The association between seizures and the severity of COVID-19 remains a matter of debate with evidence in favor [23] and against [35]. There was a previous history of cognitive impairment, older age, and higher levels of creatine-kinase and C-reactive protein after admission for COVID-19 [35] for many patients with seizures. For patients with baseline epilepsy, SARS-Cov-2 infection may trigger seizures; therefore, it is ideal to anticipate breakthrough seizures and prescribe short-term antiseizure medications opportunely [7]. Continuous electroencephalography (EEG) monitoring in any patient with a critical medical condition who has changed in mental status facilitates the timely diagnosis of non-convulsive status epilepticus [14].
Cranial nerve impairment
Thirty studies reported cranial nerve impairments in patients with COVID-19. Markedly, 746 patients presented smell/taste impairment. Anosmia and dysgeusia/parageusia indicate early involvement of the PNS by SARS-CoV-2, allowing for early screening and isolation of suspected cases before the onset of respiratory symptoms. The neurotrophic properties of SARS-CoV-2 may facilitate access to the CNS through the olfactory nerve and explain why many patients have reported anosmia as a preceding symptom [36]. In COVID-19, the sudden olfactory loss is typically unrelated to nasal swelling or rhinitis [7, 37]. The prevalence of anosmia and ageusia ranges widely from 5% in a study of patients hospitalized in Wuhan to 88% of patients for a cohort study conducted in Germany [7, 12, 37]. Visual deficits reported in COVID-19 include hemianopia in patients with acute ischemic stroke [26, 38, 39] and optic neuritis with acute visual loss [35], associated with optic nerve contrast enhancement in magnetic resonance imaging (MRI). Oculomotor impairment was present in 14 patients, in isolation or as part of a Miller-Fisher syndrome [40, 41]. These patients presented with a compromise of the III, IV, and VI cranial nerves leading to ophthalmoparesis and diplopia. Uni or bilateral abducens' involvement associated with COVID-19 has been described [41, 42]. MRI studies confirmed nerve enhancement in some of them [43, 44]. Facial nerve compromise by SARS-CoV-2 can occur in isolation [45] or as part of peripheral neuropathy like Guillain-Barre syndrome (GBS) [46]. A group of patients presented with bilateral facial diplegia with unresponsive blink reflex or unilateral facial nerve palsy, around ten days of SARS-CoV-2 infection [45, 46]. Usual MRI findings in these patients included facial nerve contrast enhancement [43]. Similarly, in some patients with GBS and cranial nerve impairment, III, VI, VII, and VIII contrast enhancement in MRI was evident [44]. Finally, some authors reported compromise of low cranial nerves among patients with COVID-19, including dysphagia as part of GBS [47], isolated dysphagia [48], and hypoglossal deficit due to rhombencephalitis [49].
Encephalopathy
In this review, 990 patients in 19 studies presented features compatible with acute encephalopathy. Encephalopathy may appear as the predominant disorder at the initial presentation of COVID-19, although most cases rarely progress to severe encephalopathy [27]. Many patients with a clinical diagnosis of encephalopathy had no findings on brain imaging. Transmission electron microscopy studies performed postmortem in patients with acute encephalopathy revealed viral particles within cytoplasmic vacuoles of brain capillary endothelial cells in frontal lobe sections. Reverse transcription polymerase chain reaction (RT-PCR) testing of frozen tissue confirmed the presence of SARS-CoV-2 in the brain [50]. The frontal lobe compromise could explain the behavioral changes seen in some patients, and the viral particles in endothelial cells may support a hematogenous dissemination pathway on SARS-CoV-2 into the CNS. Four patients with confirmed COVID-19 presented acute hemorrhagic necrotizing encephalopathy, with findings associated with disrupting mechanisms of the blood-brain barrier that could be related to cytokine storm [25, 51, 52]. Four additional patients presented clinical and imaging features of posterior reversible encephalopathy syndrome (PRES), with acute onset of headache, altered mental status, seizures, and visual disturbances accompanied by fluctuations of blood pressure, with hemorrhagic complications [35, 53]. A multifactorial series of mechanisms related to SARS-CoV-2 infection, along with a breakdown of the blood-brain barrier, may contribute to PRES development in susceptible patients. EEG findings in several patients with acute encephalopathy in our review included diffuse or focal (frontal or frontotemporal) slow activity wave patterns and some rhythmic discharges [23, 27, 35, 54–57].
Neuroinflammatory disorders
In our review, 23 patients (in 17 studies) had confirmed CNS inflammatory lesions, including encephalitis, meningoencephalitis, and encephalomyelitis, with variable prevalence [5, 25, 35, 36]. In a cohort of 2660 hospitalized COVID-19 patients, six patients presented with encephalitis as the first and only disorder, two with fatal outcomes [5]. In another cohort of 841 patients, only one patient had confirmed encephalitis [35]. The CSF of patients with inflammatory lesions showed elevated proteins, with an average of 196.3 mg/dl (range of 19-466 mg/dl) and increased cellularity, with 28.95 cells/µL (range 0-115 cell/µL). Most of these patients had normal glucose levels on CSF, although four patients had slightly increased CSF/Serum glucose ratio [25, 26, 58]. The isolation of SARS-CoV-2 was possible only in three CSF samples [29, 59, 60]. Several authors reported that CSF cellularity was predominantly lymphocytic [26, 29, 58, 61], reaching 100% lymphocytes in one case [60]. Proinflammatory cytokines in CSF measured in six patients showed high levels of interleukin (IL) 6-8, IP-10, monocyte chemoattractant protein-1 (MCP-1), neurofilament light polypeptide (NFL), glial fibrillary acidic protein (GFAP), tumor necrosis factor-alpha (TNF-α), and B-2 microglobulin [55, 58, 61, 62]. CSF samples of 20 patients with CNS compromise, and four with PNS involvement, ruled out concomitant or alternative infections. Ischemic neuronal damage, demyelination, and viral RNA in and around the hippocampus allowed the confirmation of encephalopathy in a patient who underwent brain biopsy[63]. Other neuropathology findings included perivascular lymphocytes and focal leptomeningeal inflammation. Immunohistochemical analysis showed no cytoplasmic viral staining. Brain sections from five different patients expressed low levels of the virus, but positive tests could be explained by in situ virions or bloodstream viral RNA [64]. EEG studies in patients with neuroinflammatory disorders showed generalized slow-wave activity patterns. [60, 61, 65].
Acute cerebrovascular disease
Twenty studies reported 615 patients with acute stroke. Among COVID-19 patients, the reported prevalence of acute ischemic stroke ranged between 1.1 to 2,5%, increasing to 11 to 31% among those with neurological compromise. The mean time of stroke occurrence was ten days after the onset of COVID-19 symptoms, although other studies reported an occurrence as early as two days after clinical onset [66]. Thromboembolic events were the most common neuroimaging finding associated with COVID-19 in this review, including 184 events of acute ischemic strokes. Most of the cases had atypical characteristics, such as multiple arterial territories affected without an identified cardioembolic source, bilateral compromise, a high proportion of vertebrobasilar stroke, arterial dissection, and vasculitis [35, 67, 68]. Most patients presenting COVID-19 associated stroke had baseline cardiovascular risk conditions such as hypertension, diabetes mellitus, hyperlipidemia, smoking, or previous stroke history [13]. In neuropathology findings of selected cases, a few microscopic cortical infarcts were apparent, and the neocortex. Hippocampus and cerebellum had scattered necrotic neurons, indicating terminal hypoxic-anoxic injury. Regional infarcts, however, were not present in the brain, brainstem, or spinal cord. T cells were present surrounding blood vessels and lesions, contrasting with no B cell activation. These non-specific findings can be due to vascular impairment, inflammatory/demyelinating processes, or a combination of mechanisms with direct endothelial cells damage and excessive cytokines release. [69]. For hemorrhagic stroke, the estimated prevalence ranged between 0.4 to 2.4%. In this review, 40 patients presented hemorrhagic stroke, 66.5% of which were on anticoagulation therapy [70], and one patient reportedly had a ruptured aneurysm [71]. The fact that SARS-CoV-2 binds explicitly to ACE2 receptors, along with thrombocytopenia observed in severe cases, may lead to an increased risk of a cerebral hemorrhage [13]. Postmortem examinations of selected cases revealed mild brain swelling and foci of white matter hemorrhagic lesions central fibrin with associated extravasated red blood cells, with surrounding reactive gliosis. All lesions presented macrophages at the periphery, axonal damage, and myelin loss. In patients with stroke, CSF studies showed an average protein level of 50 mg/dL; SARS-CoV-2 in CSF was positive in only one case [24, 71]. EEG findings in patients with acute stroke were described only for two patients in one study, one of them demonstrated bilateral slowing of the background rhythm with sharp frontal waves, and the other one showed persistence of sharp slow waves, mainly over the left-hemispheric regions [24].
Peripheral neuropathies
Fifteen studies reported 22 patients infected with SARS-CoV-2 and GBS, while three studies reported three patients with Miller-Fisher syndrome and one patient with polyneuritis cranialis [40, 41, 72]. The prevalence of GBS accounted for approximately 0.5% of COVID-19 patients, who developed the clinical features between five to ten days after the acute onset of respiratory symptoms [45, 73]. Whittaker et al. suggested that GBS could be a significant neurological sequela of SARS-CoV-2 [12], whereas other authors imply that GBS may occur in patients with COVID-19 without any preceding respiratory or systemic symptoms [47]. Most case reports describe patients presenting with marked lower limb weakness over upper limbs and areflexia that occurred acutely or subacutely. Variable sensory abnormalities have also been described [12]. Nerve conduction and electromyography studies carried out in 16 patients with GBS and COVID-19 reported acute inflammatory demyelinating polyneuropathy in 12 cases [44, 47, 73–80], sensory-motor axonal neuropathy in 3 cases [73, 81], and acute motor axonal neuropathy in 1 case [7]. Three patients with GBS presented enhancement of the caudal nerve roots on spine MRI [72, 73]. Sixteen studies reported CSF findings in 22 patients with SARS-CoV-2-related PNS manifestations. The main findings were albuminocytological dissociation with an average of proteins of 96,6 mg/dL, and white cell counts ranging from 0 to 9 cells/µL, without other pathologic findings. SARS-CoV-2 tests were negative in all the CSF samples [40, 41, 44, 45, 73, 74, 78–80]. Concomitant Campylobacter jejuni infection and Lyme disease were also ruled out [47, 76].