Persistent Brainstem Dysfunction in Long-COVID: A Hypothesis.

Long-COVID is a postviral illness that can affect survivors of COVID-19, regardless of initial disease severity or age. Symptoms of long-COVID include fatigue, dyspnea, gastrointestinal and cardiac problems, cognitive impairments, myalgia, and others. While the possible causes of long-COVID include long-term tissue damage, viral persistence, and chronic inflammation, the review proposes, perhaps for the first time, that persistent brainstem dysfunction may also be involved. This hypothesis can be split into two parts. The first is the brainstem tropism and damage in COVID-19. As the brainstem has a relatively high expression of ACE2 receptor compared with other brain regions, SARS-CoV-2 may exhibit tropism therein. Evidence also exists that neuropilin-1, a co-receptor of SARS-CoV-2, may be expressed in the brainstem. Indeed, autopsy studies have found SARS-CoV-2 RNA and proteins in the brainstem. The brainstem is also highly prone to damage from pathological immune or vascular activation, which has also been observed in autopsy of COVID-19 cases. The second part concerns functions of the brainstem that overlap with symptoms of long-COVID. The brainstem contains numerous distinct nuclei and subparts that regulate the respiratory, cardiovascular, gastrointestinal, and neurological processes, which can be linked to long-COVID. As neurons do not readily regenerate, brainstem dysfunction may be long-lasting and, thus, is long-COVID. Indeed, brainstem dysfunction has been implicated in other similar disorders, such as chronic pain and migraine and myalgic encephalomyelitis or chronic fatigue syndrome.

Introduction

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is arguably one of the most highly infectious viruses faced by humankind. Nearing the end of 2020, the coronavirus disease 2019 (COVID-19) pandemic, caused by SARS-CoV-2, has exceeded 70 million cases with nearly two million deaths worldwide (www.worldometers.info/coronavirus/). About 81% of COVID-19 cases are asymptomatic and mild, which takes about 2 weeks to recover. In contrast, severe pneumonia and critical multiorgan failure happen to 14% and 5% of cases, respectively, that last for 3–6 weeks.[1,2] However, a sizable proportion of COVID-19 survivors do not fully recover and suffer a postviral syndrome known as long-COVID (or long-haul COVID-19), despite being released from the hospital and tested negative for SARS-CoV-2.[3,4]

While the official medical definition or diagnostic criteria remains unconfirmed, the current description of long-COVID is the ongoing symptoms of fatigue, dyspnea, headache, cognitive impairments, cough, joint and chest pains, mood alterations, smell and taste dysfunction, and myalgia that persist for at least 4 weeks after symptom onset or hospital discharge.[5−8] Studies conducting follow-ups on survivors of COVID-19 have reported that 30–80% of them will develop symptoms of long-COVID lasting for 1–6 months. Such studies have also found that long-COVID can happen to anyone, even in young adults, students, children, and those who had only mild COVID-19 that needed neither respiratory support nor hospital care.[9−18]

This raises an intriguing question: Why does COVID-19, even when mild, turn into long-COVID in about 30–80% of the time? Common explanations thus far are residual tissue damage, viral persistence, and chronic inflammation that remain unresolved from acute COVID-19.[3,5,6,19,20] However, one more potential cause of long-COVID, which has not been discussed in detail elsewhere, could be the brainstem tropism of SARS-CoV-2 and the resulting persistent, low-grade brainstem dysfunction.

SARS-CoV-2 Tropism for the Brainstem

It is widely known that SARS-CoV-2 is a neurotropic virus with the capacity to infect and replicate in neuronal cell cultures, brain organoids, and murine brains.[21−23] This is in line with the high rate of neuropsychiatric or neurological symptoms (e.g., cognitive and mood impairments, headache, smell and taste alterations, fatigue, and myalgia) among cases of COVID-19 and other pathogenic human coronaviruses.[24−26] Even mild COVID-19 cases exhibit these neurological signs, indicating that brain involvement may occur in the early phase of COVID-19.[27] Moreover, SARS-CoV-2 RNA has been found in the brain in 30–40% of cases during autopsy of deceased COVID-19 patients.[28−30] However, other autopsy studies found no histological alterations or SARS-CoV-2 materials in the brain of COVID-19 victims.[31−33] This suggests that SARS-CoV-2 neurotropism or brain invasion can happen but not in every case.

More interestingly, autopsy studies examining specific brain regions have found evidence of brainstem involvement in COVID-19, such as inflammatory responses, neurodegeneration, and viral invasion (Table ). For example, in six victims of COVID-19, brain autopsies have identified pronounced neurodegeneration in the brainstem and its neural connections in every case examined. Importantly, this study ruled out hypoxia as the cause of brainstem damage as other brain areas more susceptible to hypoxia were not significantly injured. Thus, either pathological immune responses or SARS-CoV-2 invasion of the brainstem was suspected.[34] Another autopsy study has isolated 32 brain sections from 16 victims of COVID-19 and found concentrated SARS-CoV-2 RNA (>5 copies/mm3) in three sections from the olfactory nerves and the brainstem’s medulla.[35] More convincingly, in another autopsy study of deceased COVID-19 patients, SARS-CoV-2 RNA and proteins (nucleocapsid or spike) were detected in 50% and 40% of brainstem samples, respectively.[36] Similarly, another autopsy study has found SARS-CoV-2 RNA and spike proteins in the olfactory mucosal-neuronal junction and brainstem’s medulla in 67% and 19% of samples, respectively.[37] In sum, these autopsy studies have provided evidence for SARS-CoV-2 tropism from the olfactory system into the brainstem.

Table 1

Autopsy Studies Investigating Specific Brain Regions, with Emphasis on the Brainstem or Medulla Oblongata

study	brain samples from deceased COVID-19 patients	notable autopsy findings
(38)	Whole brain from 1 patient (male; 73 years; New York, U.S.).	Shrunken neurons, infarcts, and inflammation present in several brain regions, including the medulla.
		Positive SARS-CoV-2 RNA in olfactory bulb and cerebellum but not medulla.
(67)	Whole brain from 5 patients (42–84 years; 2 males and 3 females; Washington, U.S.).	Hemorrhages present in the brainstem of 1 patient.
(68)	Olfactory nerve, gyrus rectus, and brainstem’s medulla from 1 patient (male; 54 years; Milan, Italy).	Numerous viral-like particles and tissue damage were present in olfactory nerves, gyrus rectus, and medulla.
(39)	Whole brain from 7 patients (6 males and 1 female; 54–96 years; Basel, Switzerland).	Pronounced inflammation (microglial and astrocytic activations) in the olfactory bulb, pons, brainstem, and medulla in 2 representative patients.
		Positive SARS-CoV-2 RNA in the olfactory bulb and optic nerve of 4 and 2 patients, respectively. Negative SARS-CoV-2 RNA in brainstem and cerebellum in all cases.
(40)	Whole brain from 10 patients (7 males and 3 females; 51–74 years; Bologna, Italy).	Positive SARS-CoV-2 RNA in the olfactory nerve and brain of 1 patient.
		Pronounced microthrombi in the basal ganglia and brainstem of all patients.
(69)	Whole brain from 2 patients (2 males; 70–79 years; Zurich, Switzerland).	Pronounced inflammation (leukocyte infiltration) and microthrombosis in olfactory epithelium and nerves, and the basal ganglia in both patients.
(36)	Whole brain from 43 patients (27 males and 16 females; 51–94 years; Hamburg, Germany).	Pronounced inflammation (leukocyte infiltration and microglial and astrocytic activations) in most cases that were most evident in the cerebellum and medulla.
		Positive SARS-CoV-2 RNA in the frontal lobe (9 out of 23 patients) and brainstem (4 out of 8 patients).
		Positive SARS-CoV-2 nucleocapsid or spike protein in the medulla and its neural connections (glossopharyngeal or vagal cranial nerves) in 16 out of 40 patients.
(37)	Olfactory mucosa and specific brain regions from 33 patients (22 males and 11 females; 67–79 years; Berlin, Germany).	Positive SARS-CoV-2 RNA in olfactory mucosa (20 out of 30 patients), cerebellum (3 out of 24 patients), and medulla (6 out of 31) samples.
		Positive SARS-CoV-2 spike S1 protein in olfactory mucosa–neuronal junction.
		Pronounced inflammation (leukocyte infiltration and microglial activation; 13 out of 25 patients) and microthrombi (6 out of 33 patients) in the brain.
(42)	Whole brain from 9 patients from a sample of 21 patients (16 males and 5 females; 41–78 years; Amsterdam, Netherlands).	Pronounced inflammation (leukocyte infiltration and microglial and astrocytic activations) was most evident in the olfactory bulb and medulla in all cases.
(35)	20 brain sections from various brain regions from 2 patients (2 males; 50–71 years; Massachusetts; U.S.).	Ambiguous viral infection (<5 copies/mm3 of SARS-CoV-2 RNA) in 9 sections, of which 4 sections belonged to the medulla.
		Negative viral infection in the remaining 11 sections.
	32 brain sections (16 from the frontal lobe and olfactory nerve and 16 from the brainstem’s medulla) from 16 patients (12 males and 4 females; 48–90 years; Massachusetts; U.S).	Positive viral infection (>5 copies/mm3 of SARS-CoV-2 RNA) in 6 sections (3 from the frontal lobe and olfactory nerve and 3 from the medulla).
		Ambiguous and negative viral infection in 20 and 6 sections, respectively.
(34)	Whole brain from 6 patients (4 males and 2 females; 58–82 years; Munich, Germany).	Perivascular and interstitial encephalitis and neurodegeneration of brainstem’s solitary, dorsal raphe nucleus, and neural connections (i.e., dorsal motor nuclei of the vagus nerve, olfactory and trigeminal nerves, and fasciculus longitudinalis medialis) in all samples.
		Hypoxia-prone regions such as the neocortex, hippocampus, and cerebellum were not extensively damaged.

However, in other instances, brainstem neuropathology was observed despite the absence of SARS-CoV-2 RNA in brain autopsy of deceased COVID-19 patients.[38−40] This suggests that COVID-19 may induce brainstem damage through methods besides SARS-CoV-2 invasion, such as pathological immune or vascular activation. For example, leukocyte infiltration, activation of resident microglia and astrocytes, and microthrombosis have been observed in brain autopsies of COVID-19 victims, particularly at the brainstem.[37,39,41,42] Indeed, the brainstem is known to be highly vulnerable to acute and chronic damage from various sources, including inflammation, trauma, metabolic alterations, or vascular injury.[43] Taken together, SARS-CoV-2 may damage the brainstem through viral invasion, inflammation, and vascular activation[44] (Table , Figure ).

Figure 1

Mechanisms of SARS-CoV-2-induced brainstem dysfunction. For one, SARS-CoV-2 may invade the brainstem directly via the surface expression of angiotensin-converting enzyme 2 (ACE2) and possibly neuropilin-1 (NRP-1) receptors. SARS-CoV-2 or COVID-19 may also initiate neuroinflammation via microglial and astrocytic activation and leukocyte infiltration through the blood–brain barrier. These would lead to the production of inflammatory mediators, which damage the brainstem and vascular cells. The resulting microthrombosis from vascular injury may further potentiate neuroinflammation and brainstem dysfunction.

Angiotensin-converting enzyme 2 (ACE2) is the receptor SARS-CoV-2 uses to infect cells.[45,46] The brainstem also expresses ACE2, which may explain the SARS-CoV-2 tropism therein. Indeed, ACE2 mRNA and proteins have been found in neurons and astrocytes of the murine brainstem.[47−49] A study analyzing tissues collected from 21 distinct human brain regions has discovered high ACE2 expression in the cerebral cortex, amygdala, and brainstem. Importantly, the brainstem’s pons and medulla have the highest expression of ACE2.[50] More recently, neuropilin-1 has been confirmed as a co-receptor that facilitates SARS-CoV-2 infection into cells, especially in the olfactory epithelium where ACE2 expression is relatively low.[51,52] Interestingly, neuropilin-1 is also expressed in the brainstem of developing animal brains.[53,54] Thus, the mature brainstem may also express neuropilin-1 that promotes SARS-CoV-2 infection.

Prior pathogenic severe acute respiratory syndrome (SARS) and Middle East respiratory syndrome (MERS) coronaviruses have also shown a predilection to target and damage the brainstem in murine experiments.[55−58] Intriguingly, a proportion of SARS and MERS survivors also face long-lasting postviral illnesses that last for years.[24,59−61] Therefore, it is not a surprise that SARS-CoV-2 evolved a similar tropism feature and the capacity to induce postviral long-COVID. The putative pathways by which SARS-CoV-2 invades the brainstem involve trans-synaptic transfer via peripheral, olfactory, or cranial nerves and blood–brain-barrier (BBB) penetration from systemic circulation.[62−66]

Brainstem Functions Overlap with Long-COVID Symptoms

Others have also raised the idea of SARS-CoV-2 tropism for the brainstem as a possible cause of respiratory failure in COVID-19. This is because the brain’s cardiorespiratory center is located in the brainstem.[70−75] However, the brainstem consisting of the medulla, pons, and midbrain also plays other roles besides respiratory regulation, such as maintaining cardiovascular, gastrointestinal, and neurological processes (Figure ). Since both peripheral and central neuronal injury may be permanent and regeneration is extremely rare and slow,[76−78] brainstem injury may be long-lasting as well. Interestingly, functions of the brainstem and symptoms of long-COVID overlap to a great extent (Figure ).

Figure 2

Overview of the brainstem dysfunction hypothesis in long-COVID. Note that nuclei and subparts of the brainstem’s medulla, pons, and midbrain are not drawn to scale and may not reflect the exact neuroanatomical structures. Abbreviations used are the following: ACE2, angiotensin-converting enzyme 2; COVID-19, coronavirus disease 2019; NRP-1, neuropilin-1; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

For one, up to 30–50% of COVID-19 survivors experience persistent dyspnea and cough for 2–3 months.[13,15,79−81] Chest pain, heart palpitations, and tachycardia are also common symptoms of long-COVID that occur in about 20–40% of survivors.[13,82] Notably, the brain’s respiratory and cardiovascular neuron circuits are closely intertwined in the brainstem.[83] The ventral respiratory column (VRC) in the medulla harbors the Bötzinger and pre-Bötzinger complexes, which control rhythmic breathing during expiration and inspiration, respectively.[84,85] The brainstem also contains the pontine respiratory group (PRG), where the parabrachial and Kölliker Fuse nuclei reside and control transition between expiration and inspiration.[86] In addition, the brainstem has the caudal ventrolateral medulla (CVLM) that contains neurons responsible for regulating heart rhythms.[87,88] Hence, persistent brainstem dysfunction at the level of VRC, PRG, and CVLM may explain the cardiorespiratory-related symptoms of long-COVID.

About 25–30% of COVID-19 survivors may also experience gastrointestinal symptoms, such as nausea, vomiting, diarrhea, and abdominal pain, for 2–3 months after hospital discharge.[82,89] Interestingly, the nucleus tract solitaries (NTS) and dorsal motor nucleus of the vagus (DMV) of the brainstem have afferent and efferent neuronal projections, respectively, with the gastrointestinal tract to regulate motility and secretion.[90,91] Dysfunction of these brainstem areas may, therefore, lead to diarrhea, vomiting, and abdominal pain. While receiving afferent neurons from the gastrointestinal tract, the NTS also has neural connections to higher-order brain areas (e.g., amygdala, insula, and anterior cingulate cortex), which may convey nausea sensation.[92,93]

Long-COVID entails a myriad of neurological symptoms that may happen to 20–70% of COVID-19 survivors, such as fatigue, myalgia, insomnia, headache, depression, anxiety, smell and taste alterations, and cognitive impairments.[10−13,15] The brainstem has gustatory neurons within the NTS that relay neural information from taste buds into the gustatory cortex.[94] While dysfunction of this area may explain the altered taste perception in long-COVID, the associated smell malfunction may be attributed to olfactory bulb damage commonly observed in COVID-19 autopsy (Table ). Moreover, the reticular activation system (RAS) in the brainstem controls the sleep–wake cycle and executive attention.[43] The brainstem also harbors the raphe nuclei and locus coeruleus, the brain’s primary source of serotonergic and noradrenergic neurons, respectively.[95,96] The ventral tegmental area and substantia nigra also reside in the brainstem’s midbrain, which supplies dopaminergic neurons to the higher brain regions.[97,98] These neurotransmitters stemming from the brainstem have been implicated in a broad range of neurological disorders, including depression, anxiety, sleep and cognitive impairments, headache, fatigue, myalgia, and pain perception.[99−102] Therefore, SARS-CoV-2 invasion into the brainstem may disrupt neurotransmitter systems in the brain, resulting in diverse neurological symptoms.

Persistent Low-Grade Brainstem Dysfunction in Long-COVID

Taken together, it can be hypothesized that brainstem dysfunction may be involved in the pathology of long-COVID. However, one major caveat to this hypothesis is the ambiguity in the magnitude of brainstem dysfunction. Disruption to the brainstem, particularly at the respiratory pacemaker pre-Bötzinger complex, has been proved to be lethal, resulting in neurogenic respiratory failure and death.[43,71,103] Moreover, primary brainstem injury from traumatic brain impact has a high mortality rate.[104] It has also been proposed that sudden infant death syndrome (SIDS) results from failure of the medullary serotoninergic system in the brainstem.[105,106] Other severe and life-threatening clinical outcomes associated with brainstem damage include dysautonomia, consciousness disorders (e.g., coma and delirium), and brain death.[43,107] Hence, it may be difficult to conceive the involvement of brainstem dysfunction in chronic or persistent conditions such as long-COVID.

However, brainstem dysfunction has also been implicated in the pathophysiology of chronic disorders. For one, brain imaging studies have identified altered diffusivity and functional connectivity in the brainstem among participants with chronic musculoskeletal and neuropathic pain.[108−110] The involvement of brainstem dysfunction in chronic migraine or headache in humans has also been widely demonstrated.[111−114] Notably, long-COVID resembles and is closely associated with myalgic encephalomyelitis or chronic fatigue syndrome (ME/CFS), characterized by fatigue, myalgia, and cognitive and sleep impairments.[115,116] Interestingly, brain imaging research has found that symptom severity of ME/CFS associates and correlates with brainstem dysfunction, particularly at the RAS.[117−121] Thus, brainstem dysfunction may result in fatal or persistent disease, of which the latter may include long-COVID (Figure ).

Figure 3

Overview of the disorders related to brainstem dysfunction, which can be fatal or persistent. Note that darker shades of red/orange indicate more severe conditions.

Concluding Remarks

This review emphasized that the brainstem may be an overlooked aspect of long-COVID. In acute COVID-19, brainstem damage has been widely documented in autopsy studies. A few autopsy reports have even found SARS-CoV-2 genes and proteins in the brainstem, indicative of viral tropism and invasion. Indeed, the brainstem has a relatively high expression of ACE2 receptor, and possibly neuropilin-1, that SARS-CoV-2 exploits for cell infection. As neurological manifestations appear even in mild cases of COVID-19, the brainstem could be affected in the early disease phase. Since neurons rarely regenerate, the brainstem damage from COVID-19 may be long-lasting. Following this, the respiratory, cardiovascular, gastrointestinal, and neurological functions of the brainstem may get compromised indefinitely. Interestingly, these systems are also suspected to have malfunctioned in long-COVID. Indeed, brainstem dysfunction has also been implicated in other chronic disorders, such as chronic pain and migraine and ME/CFS. Therefore, it can be hypothesized that long-COVID may stem from persistent brainstem dysfunction.

In line with this, only one study to date has discovered persistent structural brain changes (e.g., cingulate gyrus, olfactory cortex, and hippocampus) among long-COVID survivors that were discharged from the hospital 3 months ago; unfortunately, this study did not examine the brainstem due to technological limitations.[11] In conclusion, future studies involving long-COVID may be interested in probing the brainstem dysfunction hypothesis further, such as performing brain imaging, neurophysiological assessments, and reflex (e.g., cough, pupillary light, or corneal reflexes) and auditory response tests to detect any brainstem abnormalities.[43,64,121]