Post-acute sequelae of COVID infection and cerebral venous outflow disorders: Overlapping symptoms and mechanisms?

Abstract

Neurological long Covid (NLC) is a major post-acute sequela of SARS-CoV-2 infection, affecting up to 10% of infected patients. The clinical presentation of patients with NLC is varied, but general NLC symptoms have been noted to closely mimic symptoms of cerebral venous outflow disorders (CVD). Here we review key literature and discuss evidence supporting this comparison. We also aimed to describe the similarity between CVD symptomatology and neuro-NLC symptoms from two perspectives: a Twitter-distributed survey for long covid sufferers to estimate nature and frequency of neurological symptoms, and through a small cohort of patients with long covid who underwent CVD work up per our standard workflow. Over 700 patients responded, and we argue that there is a close symptom overlap with those of CVD. CVD workup in a series of 6 patients with neurological long COVID symptoms showed jugular vein stenosis by CT venography and varying degrees of increased intracranial pressure. Finally, we discuss the potential pathogenic association between vascular inflammation, associated with COVID-19 infection, venous outflow congestion, and its potential involvement in NLC.

Introduction

The recognition of transverse sinus stenosis as a pathophysiologic driver of cerebral venous hypertension in a large subset of patients with idiopathic intracranial hypertension (IIH) has led to a dramatic shift in the attention surrounding the pathophysiology and treatment of IIH. ¹ There has been a marked increase in scientific research being published on cerebral venous anatomy, physiology, and mechanisms of disease due to cerebral venography and venous sinus stenting now being performed regularly as treatment for this condition. This has led to an increasing attention on associated cerebral venous outflow disorders (CVD), ¹ which can be considered as a set of IIH-spectrum conditions. Recently, the Society of Neurointerventional Surgery formalized this interest by developing the cerebral venous and cerebrospinal fluid (CSF) Disorders Section as a means of promoting education, research, and guidance on these conditions.

The zoonotic novel β-coronavirus produced a human pandemic disease by early 2020, that was labeled as the Coronavirus Disease 2019 (COVID-19) by the World Health Organization. Following the acute infection, about 10% of patients are left with symptoms that are collectively termed post-acute sequelae of SARS-CoV-2 infection (PASC), or Long Covid (LC). ² Neurological features constitute a major component of LC (so-called neurological “long” COVID or NLC). ³ Interestingly, NLC symptoms have been noted to closely mimic symptoms of cerebral venous outflow disorders (CVD). In this paper, we highlight the results of an informal twitter survey of over 700 long COVID patients that show close symptom overlap with those of CVD and describe a series of 6 patients with neurological long COVID symptoms who were investigated specifically for cerebral venous outflow disorders. We also discuss potential pathogenic association between vascular inflammation, associated with COVID-19 infection, and venous outflow congestion.

Background

IIH and cerebral venous hypertension

IIH is now well recognized to be a disease of cerebral venous outflow obstruction. ¹ The close relationship between CSF pressures and venous sinus pressures from animal studies appears to be conserved in humans, such that elevations in intracranial venous pressures result in direct elevations in CSF pressures.^4,5 Studies evaluating cerebral, cervical, and right atrial venous pressures obtained during supine venography in a large collection of patients with suspected IIH have suggested that there is a progressive decrement in venous pressures from the superior sagittal sinus (SSS) to the heart (the so-called brain-heart gradient). ⁶ Additionally, significant venous stenosis with trans-stenosis pressure gradients is present in upwards of 80%–90% of IIH patients, leading to even larger brain-heart venous gradients that drive CSF pressures higher. The mechanism through which these stenoses develop is poorly understood; however, venous sinus stenosis appears dynamic in nature in many instances, occurring secondary to a positive feedback loop at susceptible sites. ⁷ Due to relatively limited intracranial outflow collaterals, significant intracranial venous sinus stenosis may result in dramatic elevations in intracranial pressure (ICP), often leading to severe symptoms and other manifestations of high ICP (papilledema, visual impairment, optic hydrops, empty sella).

Recognition of venous sinus stenosis resulting in dramatic elevations in intracranial venous pressures through the development of a pressure gradient with resultant high ICP and the subsequent, immediate causal reduction in ICP once the gradient is eliminated with stenting, provides strong support for this relationship.^8,9 These observations, among others, have led to calls for changing the name of IIH to more accurately describe intracranial venous hypertension, within the spectrum of cerebral venous outflow disorders, as the primary pathophysiologic mechanism of the disease. ¹⁰

Cerebral venous outflow disorders as IIH-spectrum conditions

The development of venous stenosis at any site between the superior sagittal sinus (SSS) and the heart has the theoretical potential of impairing venous outflow, leading to upstream venous congestion or hypertension. Recently, a number of potential causative sites of extracranial venous flow impairment have been identified. ¹¹ The most commonly described site of outflow impairment is in the rostral internal jugular vein (IJV) near the transverse process of C1 and the styloid process, often referred to as styloidogenic jugular stenosis or, more colloquially, as “Eagle's syndrome.”.^12,13 Other patients have been identified with stenosis of the caudal IJV from carotid or sternocleidomastoid compression, the brachiocephalic vein, or the superior vena cava. These may collectively be referred to as cerebral venous outflow disorders (CVD). Importantly, extracranial sites of compression share distinct and important differences compared to intracranial sites of stenosis, such as the transverse sinus. First, extracranial veins are subject to mechanical stresses and rotational forces, unlike the intracranial venous sinuses. There is consistent ultrasound data demonstrating that the IJV is at its largest when supine and becomes progressively more narrowed as the patients sits upright; in the upright position, the IJV is 10% of its supine caliber, with paravertebral veins assuming the majority of the venous drainage. ¹⁴ In the case of jugular venous compression, the positional mechanical effects of neck rotation, flexion, and extension on jugular caliber results in symptom changes that are improved or worsened with changes in head position. Secondly, due to the extensive suboccipital venous collateralization channels that are often present to circumnavigate jugular sites of stenosis, associated trans-stenosis pressure gradients are often considerably smaller than those seen intracranially when measured supine during venography. This therefore results in upstream venous congestion often in the absence of hypertension, such that ICP may be normal or marginally elevated (and therefore papilledema is usually not present). Additionally, due to venous congestion occurring at sites that are not just intracranial, non-neurologic symptoms often manifest in other sites. For instance, IJV stenosis commonly causes venous engorgement in the neck, leading to swallowing symptoms or neck fullness; brachiocephalic stenosis may result in chest pain, ipsilateral arm pain and limb engorgement.

Due to these differences, patients with CVD conditions have classic symptoms that mimic IIH closely, but not completely (Figure 1). Many patients with CVD complain of brain fog, headaches, venous pulsatile tinnitus, and barometric pressure sensitivity. In two recent reviews of styloidogenic IJV stenosis, headaches, tinnitus, memory problems, hearing loss, blurred vision, neck discomfort, and sleep disorders occurred frequently.^12,13 These classic symptoms tend to be well conserved among patients with different sites of outflow stenosis in these conditions and are overwhelmingly the largest factors in reducing quality of life and functional capacity in patients.^12,13

Figure 1.

IIH and CVD have similar and overlapping symptoms (purple), with papilledema and high opening pressures generally absent in isolated CVD, and IIH generally not featuring jaw pain, swallowing issues, anterior neck fullness or pain.

COVID and IIH-spectrum conditions and intracranial hypertension

There is increasing data suggesting a link between viral infections and the development of IIH-spectrum conditions. Intracranial hypertension with papilledema is a known sequelae of viral meningoencephalitis but tends to be self-limited in scope, often requiring only short-term medical treatment. ¹⁵ A number of neurological manifestations have been widely reported in patients with acute COVID infection. Surveys have suggested the majority of COVID patients (∼75%) develop headache^] with 20% developing tinnitus and disorders of equilibrium.^16,17 Several pediatric case reports of intracranial hypertension following COVID infection have been reported, including temporary papilledema and cranial nerve palsy, as well as case reports in adults.^18–22 In a series of patients with acute COVID infection who underwent lumbar puncture, about 20% with headache demonstrated opening pressures (OP) greater than 20 mmHg and about 10% with OP greater than 25 mmHg. Two-thirds of patients demonstrated headache resolution by roughly 2 months, with the remaining individuals having ongoing symptoms. ²³

Neurological long COVID symptoms and overlap with CVD

While some discrepancy exists regarding the definition of ‘Long COVID,’ the WHO proposes a consensus agreement defining this entity as occurring “in individuals with a history of probable or confirmed SARS-CoV-2 infection, usually 3 months from the onset of COVID-19 with symptoms that last for at least 2 months and cannot be explained by an alternative diagnosis.” ²⁴ Based upon a review of medical records of over 230,000 patients following COVID infection, 34% of patients were diagnosed with either neurological or psychiatric disorders in the 6 months following infection, with 13% of these being their first diagnosis. ²⁵ The most commonly reported neurological symptoms exhibited in patients with NLC are fatigue, brain fog, headache, sleep disorders, cognitive impairment, mood disorders, dizziness, and dysautonomia. ²⁶ The symptoms exhibited by patients with CVD seem to mimic NLC quite closely.

Aims and objectives

We aimed to describe the similarity between cerebral venous outflow disorders (CVD) symptomatology and neuro-long covid (NLC) symptoms from two perspectives: a small cohort of patients with long covid who underwent CVD work up per our standard workflow, and a Twitter-distributed survey for long covid sufferers to estimate nature and frequency of neurological symptoms.

Materials and methods

We identified a retrospective cohort of six patients who were evaluated for typical CVD symptoms noted 3 or more months following acute COVID infection. A positive COVID test at the time of viral symptoms was considered as acute COVID infection. Typical CVD symptoms were considered as presence of pulsatile tinnitus, head pressure, brain fog, and visual obscurations. IRB approval was obtained. Evaluation began with CT venography, which disclosed varying degrees of internal jugular vein stenosis on one or both sides. A subset underwent catheter venography and pressure measurements from the torcula/distal superior sagittal sinus, across the internal jugular vein and down to the right atrium. Opening pressure was documented (in cm of water) when lumbar puncture was performed. Demographic features and venography results were noted (Table 1).

Table 1.

NLC cohort investigated for CVD.

Patient No.		1	2	3	4	5	6
Age (years)		Late 30s	Early 40s	Late 30s	Early 40s	Early 20s	Early 60s
Sex		M	F	F	M	M	M
Clinical symptoms*		Y	Y	Y	Y	Y	Y
Provocation of symptoms		Extension	Flexion/Extension	Flexion	Flexion	None	Flexion/L lateral Flexion
Evidence of CSF Leak		Y	N	N	Y	N	N
OP (mmH20)		18	16	-	14.5	39	-
Pressure	Torcular	13	-	-	13	22	10
	Right Atrium	5	-	-	6	13	6
	VBHG	8	-	-	7	9	4
	Gradient on provocation	15	-	-	13	No Change	13
Pressure gradient	At L C1	2	-	-	3	5	3
	At R C1	3	-	-	7	4	2
Stenosis Grade	RIJ	Moderate	Mild	Moderate	Mild	Critical	Moderate
	LIJ	Critical	Critical	Mild	None	Severe	Severe
mRS		3	3	2	1	4	2

We performed an informal, institutional review board-approved survey distributed through Twitter via the #longcovid hashtag to Long COVID patients (Table 2; Supplementary materials). The components of the survey and the questionnaire are listed in Supplementary Material 1. Descriptive statistics were calculated for the responses

Table 2.

Informal Twitter survey of the current neurological symptoms of long COVID among 755 respondents who provided answers. Full results can be found in supplementary materials. The most prevalent answer is marked in bold.

Symptom	0 (Not present)	1 (Mild: interferes slightly with activities of daily living but still able to function normally)	2 (Moderate: interferes with ability to work or carry out normal activities but still able to do so)	3 (Severe/ debilitating: interferes with ability to work or function every day)	Total Responses
Head pressure/ headache	110 (14.6%)	226 (29.9%)	257 (34.0%)	162 (21.5%)	755
Brain fog/difficulty thinking/memory impairment	38 (5.0%)	162 (21.5%)	307 (40.7%)	248 (32.9%)	755
Visual blurriness	234 (31.0%)	301 (39.9%)	185 (24.5%)	34 (4.5%)	754
Double vision	575 (76.8%)	107 (14.3%)	54 (7.2%)	13 (1.7%)	749
Dizziness/vertigo	179 (23.7%)	286 (37.9%)	188 (24.9%)	102 (13.5%)	755
Whooshing/ringing in ears/tinnitus	224 (29.7%)	270 (35.8%)	192 (25.4%)	69 (9.1%)	755
Ear pressure/pain	398 (52.9%)	241 (32.1%)	94 (12.5%)	19 (2.5%)	752
Pain/fullness in throat	339 (44.9%)	277 (36.7%)	111 (14.7%)	28 (3.7%)	755

Results

Retrospective cohort

Of six patients in the series, mean age was 40 years (20–60 years) and one-third were female (n = 2). All reported pulsatile tinnitus, head pressure, brain fog, and visual obscurations. Symptom onset following acute COVID infection ranged from 3 to 9 months. Positional symptom provocations when flexing, extending, or rotating the neck were present in 5/6 (83.3%). All patients identified in the case series were found to have narrowing of at least one internal jugular vein on CT venography. 5/6 (83.3%) patients had bilateral stenosis (Figure 2). Four patients underwent catheter venography and pressure recordings (Table 1). 4/6 had severe or critical stenosis on one side; 1 patient in this group had bilateral severe/critical stenoses. Prominent suboccipital collateral veins were noted in all patients. Modified Rankin score (mRS) at the time of clinic evaluation ranged from 1 to 4 (mean and median 2.5). Opening pressure (OP) was obtained on lumbar puncture in four patients and ranged from 14.5 to 39 cm of water (mean. 21.4); the highest OP was seen in the patient with bilateral severe/critical stenoses (Table 1). All patients were managed conservatively, including acetazolamide for elevated OP. Venous manometry in patients with cerebral venous congestion is a major element of evaluation and diagnosis. While this was not possible in our cohort, future analysis of neuro long covid patients may benefit from detailed venous analysis, given overlap in symptoms. ²⁷

Figure 2.

Example patient with NLC with bilateral IJ stenosis. There is right IJ stenosis (red arrow) in neutral position with 3 mmHg gradient (left panel) which progresses to near-occlusion with a 13 mmHg gradient with neck flexion provocation (middle panel). The left IJ demonstrates similar stenosis with 2 mmHg gradient in neutral position that worsened to 15 mmHg with provocation (yellow arrow, right panel).

We have also encountered IIH-spectrum disorders related to vaccination. A male patient in his 30's with known IIH that had previously undergone SSS and right transverse sinus stenting for severe visual loss and headache developed recurrence of symptoms 1 month after COVID vaccination (Figure 3). Repeat venography followed by LP demonstrated dissociation of CSF pressure from venous pressure with no further venous stenosis or gradients present. The patient improved following a 4-week course of oral steroids, resulting in symptom resolution. A female in her 30's developed new-onset severe papilledema and headache roughly 4 weeks following COVID vaccination that has not improved over several months and was found to have an OP of 41 cm of water with imaging demonstrating diffuse venous thrombosis including the superior sagittal sinus and dominant transverse sinus. Ultimately CSF shunting was required to salvage vision.

Figure 3.

A patient that had undergone SSS and right TS stenting for severe visual loss returned with recurrent headache and papilledema and was found to have a 24 mmHg gradient above his previously implanted stent construct (yellow arrows), with an OP of 27 cm of water (A and B). He had complete symptom resolution following additional upstream stenting. However, 1 month after COVID vaccination he developed severe recurrent symptoms and venography demonstrated no further stenosis yet LP performed immediately following venography revealed an OP of 32 (C and D). His symptoms resolved completely with a one-month course of oral steroids.

Twitter survey

A total of 757 respondents completed the survey questionnaire (Table 2). About 637 (84.1%) individuals reported a clear association between covid infection and symptom onset. About 93 (14.6%) of these subjects had CVD symptoms before covid infection and had noticeable exacerbation after the infection. Therefore, the overwhelming majority report to having developed symptoms after covid infection. The most common severe/debilitating symptoms were ‘brain fog’ and headaches/pressure (over half of respondents who graded symptoms as severe). Interestingly, tinnitus as a debilitating symptom was reported in a small proportion only. A very small percentage of respondents reported resolution of symptoms at time of survey (0.2%) and roughly one-third (32.8%) had improvement to the point of having residual symptoms only (Supplementary Materials; Survey Results). Hence the vast majority had continuing significant symptoms.

Possible inflammatory mechanisms of COVID precipitating CVD

The authors perceive four potential mechanisms in which COVID infection may potentiate CVD or IIH-spectrum symptoms. Recent work has demonstrated that the SARS-CoV-2 Spike protein results in endothelial dysfunction including blood brain barrier dysfunction and vascular leaks, which may be regarded as an intrinsic injury. ²⁸ The endothelial glycocalyx layer in blood vessels, which greatly contribute to the integrity of the vascular barrier and has lubricating, anti-thrombotic, and anti-inflammatory properties, is also disrupted by the Spike protein of COVID, potentially exposing a ‘naked’ endothelium to damage. ²⁹ Importantly, studies have demonstrated that SARS-COV-2 spike protein and glycocalyx digestion products can be detected in patient serum for months post infection.^30,31 Neuroinflammation is an active area of research in post COVID sequalae,^32,33 but the precise mechanisms remain unclear. Endothelial injury likely changes the coefficient of friction of the endothelium which may increase global vascular resistance to flow. This mechanism could also result in decompensation of the venous outflow system. Venous stasis and congestion are also associated with increased neuroinflammation^34,35 and COVID-19 related neuroinflammation has been associated with worse outcomes in neurovascular emergencies, such as thrombectomy following large vessel occlusion. ³⁶ Therefore, a future area of inquiry may be the sequalae of vascular injury due to COVID resulting in impaired vascular function triggering neuroinflammation as well as direct cytotoxic effects from the infection. Another hypothesis centers around lymphatic inflammation which may impede venous drainage via extrinsic compression. The IJV has a dense network of lymphatics surrounding it sheath on the medial, lateral, and anterior sides. It is possible that an acute inflammatory process may potentially convert a ‘compensated’ patient with asymptomatic but baseline IJV outflow stenosis to an uncompensated condition manifesting as NLC symptoms. At time of operative resection of the styloid process to decompress the IJV, enlarged lymph nodes may occasionally be found encroaching upon the IJV as a source of additional extramural compression. Thirdly, COVID-19 has been strongly associated with thrombosis,^37,38 which in turn has been associated with IIH.^39–41 Venous sinus thrombosis was observed at a higher frequency during the first phase of Covid 19, a condition treatable by mechanical thrombectomy.^42,43 While the mechanism for increased thrombosis is unknown, it may relate to COVID induced neuroinflammation and endothelial dysfunction. Thrombosis and microthrombi in the venous drainage system may be present undetectable on imaging but may ultimately increase venous outflow resistance via occluding small or large channels. Lastly, some authors have opined that the secondary effects on intracranial pressures are potentially related to impairment of CSF glymphatic flow with COVID infection, as the glymphatic pathways are predominantly located along the cribriform plate around the olfactory nerves and there is a known injury to supporting olfactory cells and lymphatic endothelium with infection (causing the characteristic anosmia associated with COVID).

Therapeutic targets

Overlapping mechanisms shared by venous intracranial hypertension and NLC suggest that treatments presently used for IIH or CVD may be useful in NLC. Presently, the authors of this commentary use blood thinners ⁴⁴ and carbonic anhydrase inhibitors for both their IIH and NLC patients with patient reported improvement in their cognition and headaches (unpublished). Steroids may also be of benefit in select patients. Medical and surgical enhancement of venous outflow may also become a reasonable consideration in patients with severe and chronic neurological symptoms, most notably surgical or endovascular relief of jugular compression in those with evidence of outflow impairment with symptom provocation occurring with corresponding anatomical outflow impairment during venography.

Conclusion

Symptoms of NLC are often severe, closely mimic those of CVD, and there appears to be a relationship between the development of IIH-spectrum conditions following COVID infection. We hypothesize that CVD may be precipitated by the inflammation resulting from COVID exposure and include potential mechanisms for this process. Going forward, population-based evaluation of the vascular anatomy of patients with and without NLC should be considered. Identification of overlapping vascular pathologies may guide future medical and surgical therapies and yield insights into other IIH-spectrum related disease states.