Language impairments and CNS infections: a review

ABSTRACT

Introduction

Infectious agents such as viruses, bacteria, parasites, fungi, and arthropods can affect the meninges and/or the brain and trigger/relate to difficulties with language tasks. We reviewed the literature on language impairments in people with Central Nervous System (CNS) infections.

Aims

To review published reports of CNS infections that trigger/relate to difficulties with language tasks, to highlight the most commonly infectious agents, and to provide further grounds to study this topic.

Methods

PubMed, Medline, Google Scholar, and Web of Science records were examined. Search strings with 70 infection terms and the words aphasia, fluency, language, naming, reading, and writing were used. The reference section and the citations of each study were examined. The type of infection, infectious agent, number of patients, neurological damage, language tests administered, and main findings were noted for each article meeting the inclusion criteria.

Results

We found 41 articles published in the last 45 years. Half of the reports were single case (series) and the other half group studies. Infections with 18 agents were detected as triggering/related to difficulties with language tasks. Bacterial meningitis, Herpes Simplex Encephalitis (HSV), and HIV were most commonly reported. Reports of English-speaking patients were most common (46%) followed by other languages spoken in the Western world. Damage to specific cortico-subcortical areas is indicated in half of the studies. Most reports presented cases with diffuse or focal bilateral damage affecting the frontal or temporal lobes.

Discussion and conclusion

The current literature is heterogeneous regarding assessment times and language protocols. There is a tendency towards testing language production and administering tests that assess severe language damage. HSV reports use the most in-depth language protocols, probably because of the specific impairments triggered by this disease due to damage to the temporal lobes. Further work may focus on understanding mild language impairments; less explored brain infections, language domains, and modalities (e.g., written/spoken comprehension); risk factors, and impact of co-occurring conditions; and how CNS infections damage the brain (e.g., subcortically). Potential issues such as the presence of delirium, stress/anxiety, and publication biases regarding prevalent infections in the Western world seem relevant to consider.

KEYWORDS:

• Aphasia
• encephalitis
• herpes
• language
• meningitis

Introduction

Infections of the Central Nervous System (CNS) are one of the main causes of acquired language impairments, alongside stroke, tumours, epilepsy, traumatic brain injury, and neurodegeneration (Goodglass, 1993; Hegde & Freed, 2016; Whitworth et al., 2014). An infection is the invasion of an organism by disease-causing agents, their growth, and the response of host tissues to the infectious agents and the toxins that they produce (Gavazzi & Krause, 2002; Mook-Kanamori et al., 2011; Riddell & Shuman, 2012; Whitley, 1990). Different agents can cause an infection, including viruses, bacteria, parasites, fungi, and arthropods. They do so via three main routes: (i) carried by blood/fluids, (ii) through a skull fracture, intravenous injection, surgery, or (iii) via adjacent structures such as the middle ear or the sinuses (Sarrazin et al., 2012). Some infections can also be transmitted through organ transplants (Basavaraju et al., 2014). Many infections are preventable or treatable (Tyler & Ropper, 2018; Van de Beek et al., 2006). However, some may cause brain damage such as swelling, stroke, vasculitis, mass, and other problems such as epilepsy or blood perfusion issues (Schmidt et al., 2006; Sokolov et al., 2015).

Infections of CNS can be focal or diffuse: focal lesions correspond to abscesses, cysts, and granulomas; while diffuse lesions are labelled under meningitis, encephalitis, meningoencephalitis, myelitis, myeloencephalitis, and radiculomyelitis (Akhaddar, 2017). In this review, we focus on CNS infections that trigger/relate to language difficulties (Benke et al., 1995; Grahn et al., 2013; Salih et al., 1991). In this regard, two types of infection of the arachnoid matter or the pia mater/brain parenchyma are commonly reported, namely, meningitis and encephalitis.

Meningitis is the infection of the meninges, that is, the three-tissue membrane surrounding the encephalon and the spinal cord (Van de Beek et al., 2004; Weisfelt et al., 2006). Meningitis can be bacterial (e.g., streptococcus pneumoniae, neisseria meningitis) or viral (e.g., enterovirus, varicella zoster). It typically presents with headaches, neck stiffness, fever, and change in mental status (Bijlsma et al., 2016). The precise diagnosis of meningitis includes a lumbar puncture, where cerebrospinal fluid is extracted and examined (Costerus et al., 2018), neurological assessment, and neuroimaging (Durand et al., 1993; Kastrup et al., 2005). Importantly, meningitis per se does not result in neurocognitive dysfunction, as it does not directly affect the nervous tissue (Decimo et al., 2012). Neurocognitive disorders in people with meningitis are often due to indirect effects, such as fluid retention, brain swelling (i.e., oedema), or an inflammatory reaction of the underlying brain tissue (Hoogman et al., 2007; Kirmi et al., 2009; Kloek et al., 2020, 2021).

Encephalitis refers to an infection of the brain parenchyma. Patients with meningitis often have symptoms and signs that can be referred to the involvement of the brain parenchyma. However, meningitis and encephalitis must be seen as a continuum, as the natural course of some infectious agents is to affect both the meninges and the brain if left untreated (Mook-Kanamori et al., 2011). Encephalitis can be bacterial (e.g., S. pneumoniae, Tropheryma whippelii), viral (e.g., herpes simplex virus, human immunodeficiency virus), and parasitical (e.g., falciparum malaria). Encephalitis can also be caused by autoimmune disorders, but these are beyond the scope of the current review. Infectious encephalitis typically triggers fever, seizures, and focal neurologic deficits (Khatib et al., 2017; Krbková et al., 2015; Sittinger et al., 2002). Diagnosis is similar to that of meningitis, in that it may include a lumbar puncture, neurological and radiological examinations (Tyler & Ropper, 2018).

Despite previous reviews of neuropsychological impairments due to CNS infection (Hokkanen & Launes, 2000; Hoogman et al., 2007; Kloek et al., 2020; Lucas et al., 2016; Nau & Schmidt, 2007), we were unable to find a recent overview with a focus on language difficulties. Such an article could be used as an updated reference for research and teaching (cf. Carter et al., 2003). Therefore, here we report the results of a review of the literature on CNS infections and difficulties with language tasks.

It is not our intention to cite all the reports of CNS infections over a specific period. Rather, we wish to review the different types of infection that have been reported as triggering and/or relating to difficulties in language tasks, with special emphasis on the most commonly reported. Additionally, we wish to stress the interest of the study of CNS infections to enhance our knowledge of how language is processed in the brain, while facilitating research and the referral of people to specialized language services. Such a review simply cannot be exhaustive, as the various aetiologies we will report on may relate to disorders that go beyond language (e.g., cysticercosis may result in cysts that occupy any subcortical area of the hemisphere). Furthermore, the review is exposed to publication bias, as some infectious agents have different incidence/prevalence in different areas of the world (Kenyon et al., 2014; Leonhard et al., 2021; Murthy et al., 2014), potentially underestimating the occurrence of language impairments in agent endemic areas where studies of aphasia may not be well represented (Code et al., 2016; Code & Petheram, 2011).

Methods

PubMed, Medline, Google Scholar, and Web of Science records were searched for studies of CNS infection and language both in adult and paediatric populations. The following 70 search terms were used: acanthamoeba, acquired immunodeficiency syndrome (AIDS), amoebic, arbovirus, aspergillus, babesiosis, bacterial, balamuthia, balamuthia mandrillaris, bartonella henselae, COVID, COVID-19, cryptococcosis, cryptococcus neoformans, cyst, cysticercosis, cytomegalovirus, dengue, encephalitis, enterovirus, Epstein-barr, falciparum malaria, chagas, haemophilus influenzae, hepatitis, hepatitis c, hepatitis e, herpes simplex, human granulocytic ehrlichiosis (HGE), human immunodeficiency virus (HIV), flavivirus, fungal, influenza, language, leprosy, limbic, lymphoma, malaria, measles, meningitis, meningococcal, meningoencephalitis, molluscum contagiosum, mucormycosis, mycoplasma pneumoniae, myelitis, myeloencephalitis, myxovirus, naegleria fowleri, nocardia, pneumococcal, pyogenic, rabies, rubella, SARS-CoV-2, Spanish flu, streptococcus intermedius, syphilis, toxoplasma, toxoplasmosis, tuberculoma, tuberculosis, trypanosomiasis, varicella zoster, varicella, viral, Whipple, Whipple’s disease, zika, zygomycosis. These terms were used in connection with the words: aphasia, comprehension, difficulties, fluency, language, naming, production, reading, syntax, and writing.

We examined the reference sections of each study and their citations to identify further references. For each relevant study, we considered the following aspects: type and nature of CNS infection (i.e., meningitis, encephalitis, meningoencephalitis); infectious agent (e.g., bacterial, viral, parasitical; and the specific agent); number of people reported on the study; neurological damage reported, if any; language tests used; and main findings (e.g., comparisons between the population with CNS infection at stake and a group of age-matched controls and/or another group of people with CNS infection but different symptoms). Studies labelled under meningoencephalitis indicate cases where the natural course of the infectious agent affects both the meninges and the brain.

We excluded (i) studies mentioning aphasia, dysphasia, or a language impairment without specifying the tasks administered in the assessment protocol; (ii) reports of language impairments in people with different types of infection collapsed into the same group, or in people with unrelated neurological pathologies (e.g., a stroke in 2007 and bartonella henselae [cat-scratch disease] infection years later, Marienfeld et al., 2010); (iii) reports of people with urinary tract infections and autoimmune limbic encephalitis, because language difficulties could be due to delirium or confusional state (Constantinides et al., 2018; Evoli et al., 2012; Fong et al., 2015; Kishi et al., 2010); (iv) reports of people with proteinaceous infectious particle (i.e., prion) diseases, encephalitis lethargica, acute disseminated encephalomyelitis (ADEM), sarcoidosis, and of people with Anti-N-methyl-d-aspartate receptor encephalitis (Anti-NMDAR), as the mechanisms underlying these conditions are sporadic or autoimmune, hence, not necessarily or very rarely related to CNS invasion (Dale et al., 2004; Geschwind et al., 2008; Hayashi et al., 1995; Hébert et al., 2018; McGinnis, 2011; Yoshikawa & Oda, 1999); and (v) studies of people with lymphoma and language impairments, in which an infectious cause was ruled out or left unspecified (Donix et al., 2007; Mori et al., 1989; Simon et al., 2015).

Results

We found 41 articles meeting the inclusion criteria for this review. All articles were published over the last 45 years, with the exception of two articles published in 1893 and 1901. A total of 18 infectious agents were detected as triggering/related to difficulties with language tasks. Six infectious agents were found under meningitis: three bacterial (i.e., meningococcus, pneumococcus, and haemophilus influenzae), and three viral (i.e., enteroviruses, myxovirus, and varicella zoster). Seven infectious agents were found under encephalitis: three bacterial (i.e., treponema pallidum, streptococcus intermedius, tropheryma whippelii), two viral (i.e., herpes simplex virus, SARS-CoV-2), and two parasitical (i.e., falciparum malaria, taenia solium). Finally, five infectious agents were found under meningoencephalitis: two bacterial (i.e., borrelia burgdoferi, mycobacterium tuberculosis), two viral (i.e., HIV, enterovirus), and one fungal (i.e., aspergillus fumigatus).

Half of the articles were single case (series) studies (21 of the 41 total papers, 51%) and the other half group studies (20/41, 49%). Some CNS infections were more frequently reported as single case studies (e.g., herpes simplex encephalitis, tuberculosis, cysticercosis), others as group studies (e.g., bacterial meningitis, HIV/AIDS). Reports of English-speaking patients were the most common (19/41, 46%) followed by Spanish (5/41, 12%), German-speakers, and Italian-speakers (4/41, 9%); Dutch, Greek, Korean (2/41, 5%); and Gujarati-French, Luganda, and Mijikenda (1/41, 3%). The only study that assessed bilingual speakers in their two languages was conducted in Canada, on a Gujarati-French speaker (Paradis & Goldblum, 1989).

Neurological damage to specific cortico-subcortical areas is specified in 54% (22/41) of the studies. Among these studies, 84% (18/22) indicate damage to specific cortical areas, 1% (2/22) only to subcortical areas, and 1% (2/22) to both cortical and subcortical areas. While brain infections can affect more than one brain region, the cortical areas most commonly involved are the left temporal lobe (6), the temporal lobes bilaterally (5), the left frontal lobe (6), and the left parietal lobe (4). Other areas mentioned are the insula (2) and the right temporal lobe (2). At the subcortical level, the thalamus (2), hippocampus (1), and arcuate fasciculus (1) are mentioned.

In many publications, more than one language task was administered. Full aphasia batteries were used in six reports and caregiver questionnaires in two. The total number of tasks (counting repeated tasks) for the reports we considered amounted to 72. Naming tasks were used most commonly (24/72, of which 18 were object naming with pictures, 4 object naming to definition, 1 naming famous faces, 1 naming by sound), followed by letter fluency (13/72) and category fluency (12/72). Other less frequently used tasks were repetition (6/72, of which 4 were word repetition, 1 sentence, and 1 syllable repetition), spontaneous speech (5/72), word reading (4/72, of which 2 reading words, and 2 reading irregular words), sentence comprehension (1/72), comprehension of spoken instructions (1/72), lexical decision (1/72), understanding of prosody/metaphors (1/72), semantic judgment (1/72), reciting the alphabet (1/72), verbal concept formation (1/72), and sentence completion with verbs in the past (1/72).

We present a summary of the results in Table 1.

Table 1. Studies reporting language impairments due to brain infection.

Type	Agent	Reference	Participant(s)	Neurological damage	Language tests	Main findings
Meningitis	Bacterial (meningococcus, pneumococcus, and haemophilus influenzae)	Pentland et al. (2000)	30 English-speaking children (4 pneumococcal, 26 haemophilus influenzae; 9–11 years old, assessed ~2 years after time of illness)	Not specified	Linguistic knowledge (CELF-3; Semel et al., 1995); Discourse ability (TLC-E; Wiig & Secord, 1989)	Below normal scores in linguistic knowledge, particularly, sentence assembly; and below normal scores in discourse ability, particularly understanding ambiguous sentences, making inferences, recreating sentences, and figurative language
	Bacterial (meningococcus, pneumococcus)	De Beek et al. (2002)	76 Dutch-speaking adults (25 meningococcal, 26 pneumococcal, 25 controls; 16–65 years old; assessed ~1 year after time of illness)	Not specified	Reading irregular words (DART; Schmand & Lindeboom, 1992); Category fluency (animals and jobs); Letter fluency (3 letters, not specified)	No differences between groups in reading irregular groups; people with pneumococcal meningitis scored lower than people with meningococcal and controls in category and letter fluency
	Bacterial (meningococcus, pneumococcus)	Weisfelt et al. (2006)	137 Dutch-speaking adults (49 meningococcal, 38 pneumococcal; 16–65 years old; 50 controls; assessed ~3 months after time of illness)	Not specified	Reading irregular words (DART; Schmand & Lindeboom, 1992); Category fluency (animals and jobs); Letter fluency (3 letters, not specified); Object naming (BNT; Kaplan et al., 1983)	No differences between groups in reading irregular words, category/letter fluency, or object naming
	Bacterial (meningococcus, pneumococcus) and viral (not specified)	Schmidt et al. (2006)	149 German-speaking adults (59 bacterial meningitis, 59 viral meningitis, 30 controls; 30–60 years old, assessed ~6 years after time of illness)	Brain swelling (24%) and ischaemic lesions (10%), only in people with bacterial meningitis	Letter fluency (P, M, K; Aschenbrenner et al., 2000); Category fluency (animals, foods, jobs; Aschenbrenner et al., 2000); Comprehension of spoken instructions (Aachen Aphasia test, Token Test; Huber et al., 1983)	No differences between groups in letter fluency. People with bacterial and viral meningitis scored lower than controls in category fluency, and people with bacterial meningitis scored lower than controls in comprehension of spoken instructions
	Viral (enterovirus, myxovirus, varicella zoster, unidentified)	Sittinger et al. (2002)	42 German-speaking adults (21 viral meningitis; 21 controls; 22–53 years old, assessed ~2 years after time of illness)	5 of 21 people with viral meningitis showed parenchyma involvement	Reciting the alphabet (WAIS-III; Dahl, 1986); Lexical decision (in each item 1 word and 4 non-words are presented – e.g., nesa, naso, nose, nosa; MWT-B, Lehrl, 1995); Verbal concept formation (identify conceptual similarities between two words, WAIS-II; Dahl, 1986)	No differences between people with viral meningitis and controls in any of the tests
Encephalitis	Bacterial (Streptococcus intermedius)	Khaja et al. (2017)	1 English-speaking female (72-year-old, assessed “few days” after complaints)	Left temporal lobe	Naming of a real object “stethoscope”, forward and backward spelling of the word “world”, qualitative spontaneous speech interview	Responded “football” when asked to name the real object “stethoscope”; able to spell forward; one error spelling backward. Spontaneous speech fluent and coherent
	Bacterial (Streptococcus anginosus)	Jimenez et al. (2018)	1 English-speaking female (46-year-old, assessed on hospital admission)	Left thalamic mass, pressure on left lateral ventricle, third ventricle, and hypothalamus	Naming and repetition	Impaired naming and repetition
Encephalitis	Bacterial (treponema pallidum)	Ahn et al. (2012)	1 Korean-speaking female (47-year-old, high school graduate, assessed 9 months after diagnosis)	Bilateral medial temporal lobes and insula. Also, meningioma in the left cerebellum	Animal naming (Hachinski et al., 2006); letter fluency (/k/, /ng/, /s/)	10 items in action naming, 7 words in letter fluency. After a 6-week rehabilitation, 9 items action naming, 9 word in letter fluency
	Bacterial (treponema pallidum)	Ioannidis et al. (2014)	1 Greek-speaking male (52-year-old, assessed 6 months after first complaints, second assessment 6 months after first consultation)	Bilateral perisylvian cortical atrophy, more prominent in the left hemisphere. Theta-wave electroencephalographic activity in the left frontotemporal area	Object naming (5 animals, 5 objects; ACE-R, Konstantinopoulou et al., 2011)	Word-finding difficulties in the first assessment, resolved 6 months after
Encephalitis	Bacterial (tropheryma whippelii)	Benito-León et al. (2008)	1 Spanish-speaking male (63-year-old, assessed at hospital admission and 2 years after).	Mild bilateral frontotemporal atrophy. Single-photon emission computer tomography revealed decreased regional cerebral blood flow in frontal and temporal areas (greater decrease in left hemisphere)	Category fluency (animals, fruits)	Scores below the norm in category fluency (7 animals, 5 fruits). Within-normal scores 2 years after treatment.
	Bacterial (tropheryma whippelii)	Manzel et al. (2000)	1 English-speaking male (41-year-old, assessed 3–4 times for a 7-month period”)	Mesial temporal lobes and basal forebrain bilaterally, striatum, pallidum and hypothalamus. Thalamus relatively spared	Two comprehensive aphasia battery (Boston Diagnostic Aphasia Examination, BDAE; Goodglass & Kaplan, 1983; Multilingual Aphasia Examination, Benton et al., 1983), object naming (Boston naming test, BNT, Kaplan et al., 1983), letter fluency (“C”, “F”, “L”).	Letter fluency was below the norm, only in the first assessment. The other language tasks were within the norm
Encephalitis	Viral (SARS-CoV-2)	Almeria et al. (2020)	35 Spanish-speaking adults (24–60 years old, assessed 10–35 days from hospital discharge)	Not mentioned	Picture naming (Boston naming test, BNT), letter fluency (p), category fluency (animals) from Neuronorma project (Peña-Casanova et al., 2009).	Below normal scores depending on clinical complaints. People with headache showed below normal scores in letter fluency. People with loss of taste had below normal scores in picture naming. People who required oxygen therapy during hospitalization fared worse in category fluency.
Encephalitis	Viral (herpes simplex virus)	Baratelli et al. (2015)	1 Italian-speaking female (63-year-old, assessed one week, four weeks, and four months post onset)	Left insula and anterior parahippocampal region	Picture naming, naming on verbal definition, naming of famous faces, word-picture matching, verbal semantic questionnaire	Longer naming latencies in early assessments. Below-normal performance in all tasks with the exception of word-picture matching and verbal semantic questionnaire. No significant differences between biological entities and artifacts
	Viral (herpes simplex virus)	Barbarotto et al. (1996)	7 Italian-speaking adults (17–68 years old, assessed “few-months” to 4 years post-onset)	Temporal damage prevailing on the left in all but one participant who presented right temporal damage	Picture naming with biological categories: animals, fruits, vegetables; and artifacts: tools, furniture, vehicles. (Laiacona et al., 1993)	Four of seven participants were more impaired at naming biological categories. Frequency, familiarity, and name agreement influenced performance.
	Viral (herpes simplex virus)	E. De Renzi and Lucchelli (1994)	11 Italian-speaking adults (1 female with HSV, 49-year-old, assessed 4, 7, and 15 months after first hospital admission, and 10 age-matched controls)	Bilateral antero-medial temporal lobes, more marked on the left	Picture naming, naming to definition, picture identification (pointing to the correct item named, 8–10 possibilities); word-picture matching; naming by sound; describing perceptual differences; picture naming of objects/animals and non-objects/animals; recall the colour of an object/animal when reading its name; drawing objects/animals; sorting cards. Seven experiments with different lists of items/tests	Poor performance on animals and flowers, moderate on fruits, within normal range for objects
	Viral (herpes simplex virus)	Hanley et al. (1989)	24 English-speaking adults (1 male with HSV 60-year-old, assessed 15 years after first diagnosis and “some months” later, and two aged-matched control groups of 11 and 12 people)	Right superior temporal gyrus, anterior end of longitudinal fasciculus. Possible left temporal lobe damage	Picture naming; recognition of category members from definitions; recognizing famous faces and names (rate familiarity of the items, state person’s occupation, name the person); recognizing voices of famous people (Meudell et al., 1980); forced-choice tests with famous faces and names	More difficulty with biological entities than artifacts, with the exception of animals which is not different than artifacts. Difficulties with famous faces stable over a 4-month period (familiarity is rated low, few correct answers on occupations and name of the person). Poor at recognizing voices of famous people and forced-choice tests
	Viral (herpes simplex virus)	Lacidogna et al. (2020)	1 Italian-speaking female (26-year-old, assessed 16 months post-onset)	Marked left temporal and moderate bilateral orbital and mesial frontal, sparing inferior frontal gyrus (perfusion deficit)	Comprehensive aphasia battery (BADA, Miceli et al., 1994); picture naming (nouns, verbs, biological entities; artifacts); word-picture matching with auditory and written presentation (nouns, verbs; biological entities; artifacts)	More difficulty naming and comprehending nouns relative to verbs; no significant differences between biological entities and artifacts
	Viral (herpes simplex virus)	Tyler et al. (2002)	4 English-speaking adults (4 people with HSV, 35–68 years old, assessed 4–20 years post herpes onset)	Left temporal lobe (temporal and medial temporal cortex, temporal pole), two people had extensive right temporal damage	Past tense elicitation paradigm (i.e., two-sentence context, the second sentence needs to be completed – “My nose sometimes ‘bleeds’. Last night it … ”)	Poorer performance on irregular than regular past tense forms
	Viral (herpes simplex virus)	Warrington and Shallice (1984)	4 English-speaking adults (23–60 years old, assessed 1–8 months after first hospital admission)	Bilateral temporal lobes	Picture naming/ description; auditory naming/description. Seven experiments with different lists of items	More difficulty with biological entities than artifacts
Encephalitis	Parasitical (falciparum malaria)	Bangirana et al. (2016)	227 Luganda-speaking children (1.5–4 years old; 80 cerebral malaria, 86 severe malaria anaemia, 61 controls. Children with cerebral malaria and severe malaria anaemia, assessed 1 week after discharge or at enrolment, and then 6 and 12 months after)	Not specified	Interactive tasks of expressive and receptive language (Mullen Scales of early learning; Mullen, 1995). Testing performed in English and in Luganda when children did not understand English	Children with cerebral malaria and severe malaria anaemia were more impaired than children the control group. Children with cerebral malaria were more impaired than children with severe malaria anaemia
	Parasitical (falciparum malaria)	Carter et al. (2006)	487 Mijikenda-speaking children (6–9 years old; 152 cerebral malaria, 156 malaria and complicated seizures, 179 controls. Children were assessed 64–71 months in median after admission)	Not specified	Receptive grammar (Bishop, 1983); receptive vocabulary (Dunn & Dunn, 1981); syntax (Armstrong & Ainley, 1989); lexical-semantics (Crystal, 1992); pragmatics (Dewart & Summers, 1995); phonology and word finding (German, 1986). Mijikenda adaptations of standardized English tests	Children with cerebral malaria were more impaired than children with malaria and complicated seizures and the control group. Children with cerebral malaria were particularly impaired in receptive vocabulary and lexical semantics. Children with malaria and complicated seizures were particularly impaired in phonology and pragmatics
Encephalitis	Parasitical (taenia solium)	Otero et al. (1989)	1 Spanish-speaking male (7-year-old, informally assessed 8 and 4 months before admission, formally assessed on admission, and 6 months after treatment)	Left sylvian fissure	Informal testing: teacher and parental report; formal testing: intelligence scale (WISC, Wechsler Intelligence Scale for Children, version not mentioned)	Informal testing 8 months before treatment indicates poor story comprehension, telegraphic speech marked with phonological paraphasias (i.e., “llobeta” instead of “botella” [bottle], “manzan” instead of “manzana” [apple]) and other phonological or morphological omission errors (i.e., “él va casa” instead of “él va a la casa” [he goes home]; 4 months before treatment lack of response to oral commands and near mutism. Formal testing indicated a “mild delay in response and sentence formulation with decreased comprehension of spoken language” (verbal IQ = 89, performance IQ = 87). Improvement 6 months after treatment (no details are mentioned), but speech therapy was recommended.
	Parasitical (taenia solium)	Paradis and Goldblum (1989)	1 Gujarati-French-Malagasy-speaking male (25-year-old, assessed before surgery, two weeks, and 8 months after)	Right precentral gyrus	Comprehensive aphasia battery (Boiler & Hécaen, 1979); French-Guajarati and French-Malagasy versions of the Bilingual aphasia test (Paradis & Libben, 1987)	No problems with French across assessments. Difficulties with Guajarati only early after surgery (i.e., verbal fluency, word-finding difficulties, and borrowing French words in spontaneous speech). Difficulties with Malagasy only 8 months after surgery (i.e., syntactic comprehension, verbal fluency, word-finding difficulties)
Meningo- Encephalitis	Bacterial (mycobacterium tuberculosis)	Booth and Curtis (1893)	1 English-speaking male (24-year-old, assessed before and after surgery)	Left frontal lobe	Spontaneous speech	Mild word-finding difficulties, intact auditory comprehension. No difficulties after surgery
	Bacterial (mycobacterium tuberculosis)	Garcia-Grimshaw et al. (2018)	1 Spanish-speaking female (24-year-old, assessed upon hospital admission)	Left parietal lobe, cerebellum, and arcuate fasciculus	Naming, following simple commands, reading out loud, writing, repetition of simple sentences (the tests are not specified)	Problems with repetition and some paraphasias (type unspecified) in spontaneous speech.
	Bacterial (mycobacterium tuberculosis)	Hinsdale (1901)	1 English-speaking male (35-year-old, assessed 2 years after second pneumonia – first pneumonia 14 years before, and re-assessed 1 year after)	Left inferior parietal lobe	Spontaneous speech, naming real objects (i.e., pencil, coin), reading the alphabet, word reading (nouns).	More difficulties in the first assessment than one year after. Difficulties with lexical retrieval upon naming real objects (not semantics, as responded correctly to semantic-related questions, e.g., for a pencil, is this a knife? is this something to write with?); difficulties reading the alphabet, and reading words (semantic and phonological paraphasias)
Meningo-encephalitis	Bacterial (borrelia burgdoferi)	Benke et al. (1995)	40 German-speaking adults (20 people with Lyme disease; 35–72 years; assessed ~4 years on average after infection [4–100 months]; 20 age-matched controls)	Facial neve palsy (4 people), multiple cranial nerve involvement (1 person), left-sided weakness (1 person), minor nonspecific abnormalities (14 people)	Letter fluency (1 task, letter “s”); rapid syllable repetition (Mateer & Kimura, 1977)	Both tasks are significantly impaired, relatively to aged-matched control
	Bacterial (borrelia burdoferi)	Krupp et al. (1991)	25 English-speaking adults (15 people with Lyme disease; 30–40 years; assessed ~7 months after infection; 10 age-matched controls)	8/15 people with Lyme disease did MRI. 6 had normal scans. 2 people presented “one or more small areas” of white matter abnormalities.	Letter fluency (“C”, “F”, “L”)	Letter fluency significantly impaired, relatively to age-matched controls
	Bacterial (borrelia burgdoferi)	Sokolov et al. (2015)	1 Swiss German-speaking female (16-year-old, assessed 3 and 24 hours after symptom onset)	Left parietotemporal junction (perfusion deficit, no structural brain damage or vasculitis as seen with MRI)	Spontaneous speech, repetition	Reduced speech production and comprehension, phonological paraphasias preserved repetition. Symptoms resolved 24 hours after antibacterial and antiviral treatment.
	Bacterial (borrelia burgdoferi)	Svetina et al. (1999)	44 English-speaking adults (30–40 years-old; time of assessment after infection not mentioned, but 91% where in a chronic stage; 43 age-matched controls)	Not mentioned	Object naming (BNT, Kaplan et al., 1983), letter fluency (“C”, “F”, “L”), category fluency (“animals”, “fruits”, “vegetables”); subjective reports	55% of people with Lyme disease (vs 14% of people in the control group) reported subjective word-finding difficulties; people with Lyme disease were more impaired in object naming; no differences in the total number of words in fluency tasks.
Meningo-encephalitis	Viral (HIV)	Abusamra et al. (2016)	20 Spanish-speaking adults (26–57 years-old, assessment time not mentioned)	Not mentioned	Comprehension and repetition of conversational and narrative prosody, interpretation of metaphors and indirect speech tasks, fluency tasks (letter and semantic), semantic judgment (Ferreres et al., 2007)	Measures of narrative discourse, fluency (verbal and semantic) and metaphor comprehension were the most impaired.
	Viral (HIV)	Becker et al. (1997)	267 English-speaking adults (200 people with HIV, and 67 people without HIV, 20–40 years-old, assessment time not mentioned)	Not mentioned	Category fluency (i.e., animals) and letter fluency (i.e., “F”, “A”, “S”)	People with HIV without dementia fared lower on category fluency and letter fluency relative to a control sample
	Viral (HIV)	Kambanaros et al. (2019)	80 Greek-speaking adults (40 people with HIV, and 40 people without HIV, 19–59 years old, assessment time?	Not mentioned	Sentence repetition (8 different structures), category fluency (i.e., objects, animals, fruits); letter fluency (i.e., X, A, Σ)	People with HIV fared lower on sentence repetition, category and letter fluency relative to a control sample
	Viral (HIV)	P. McCabe et al. (2002)	5 English-speaking adults (34–67 years-old, assessment time not mentioned).	Not mentioned	Comprehensive language test (Western Aphasia Battery, Kertesz, 1982), Shortened Token Test (De Renzi & Faglioni, 1978), Right hemisphere language battery (Bryan, 1995), Communicative Effectiveness Index (Lomas et al., 1989), Pragmatic Protocol (Prutting & Kittchner, 1987), object naming (BNT, Kaplan et al., 1983), spontaneous speech (i.e., Cinderella story, event recount), category and letter fluency	Normal scores on comprehensive language testing in spite of word finding difficulties in spontaneous speech, object naming, spelling, reduced word fluency, and disturbances in pragmatic skills (i.e., metaphor interpretation)
	Viral (HIV)	P. J. McCabe et al. (2008)	3 English-speaking males (36-year-old with AIDS and dementia; 34-year-old with AIDS and no dementia; 34-year old healthy control; assessed a first time, 6 months and 13 months after)	Not mentioned	Comprehensive language test (Western Aphasia Battery, Kertesz, 1982), Shortened Token Test (De Renzi & Faglioni, 1978), Right hemisphere language battery (Bryan, 1995), Communicative Effectiveness Index (Lomas et al., 1989), Pragmatic Protocol (Prutting & Kittchner, 1987), object naming (BNT, Kaplan et al., 1983), spontaneous speech (i.e., Cinderella story, event recount), category (i.e., animals) and letter fluency (i.e., /s/)	The person with AIDS and dementia had a pragmatic language disorder which was particularly marked during spontaneous speech (e.g., self-focused, verbose, unable to maintain topic). Also, he had some word finding difficulties, circumlocutions but largely. Slight differences across assessment times, attributed to non-linguistic factors such as attention or fatigue
	Viral (HIV)	Millikin et al. (2004)	217 English-speaking adults with HIV (28 people characterized as asymptomatic; 58 mildly symptomatic; 131 met criteria for AIDS; 30–50 years-old, assessment time not mentioned).	Not mentioned	Letter fluency (“F”, “A”, “S”) and category fluency (i.e., animals).	Relative to a normative sample and to people with no to mild symptoms, people with severe symptoms fared worse in letter fluency but not in category fluency. Furthermore, people with severe symptoms produced fewer switches.
	Viral (HIV)	Sheppard et al. (2017)	128 English-speaking adults (50–56 years-old, 40 with HIV, 48 HIV-seronegative controls matched for age, and 40 HIV seronegative controls aged ≥65)	Not mentioned	Object naming (BNT, Goodglass et al., 2001)	No differences between groups in object naming.
	Viral (HIV)	White et al. (1997)	39 English-speaking adults (30–50 years-old, 9 with HIV and dementia, 15 HIV-seropositive without dementia, 15 HIV-seronegative controls)	Not mentioned	Letter fluency (F, A, S; Benton et al., 1983); category fluency (animals, fruits, vegetables; Monsch et al., 1992); object naming (BNT, Kaplan et al., 1983)	People with HIV and dementia scored lower than the other two groups in letter fluency. No differences between groups in object naming or category fluency
	Viral (HIV)	Woods et al. (2004)	102 English-speaking adults (25–50 years-old, 21 with HIV and dementia, 51 with HIV, 30 healthy controls, assessment time not mentioned)	Not mentioned	Letter fluency (F, A, S; Benton et al., 1983)	People with HIV and dementia produced fewer correct words, switches, and more errors than people with HIV without dementia and healthy controls. No differences in cluster size were attested
Meningo-encephalitis	Viral (enterovirus)	Kim et al. (2012)	1 Korean-speaking female (23-year-old, assessed 4 days after hospital admission)	Left frontal lobe	Comprehension of complex sentences, word repetition and object naming	Impaired scores 4 days after hospital admission, no impairments 3 month after pharmaceutical treatment
Meningo-encephalitis	Fungal (aspergillus fumigatus)	Lin et al. (2014)	1 English-speaking male (32-years-old, assessed 55, 57, 68 days after hospital admission)	Left middle cerebral artery territory	Reading, conversation, repetition	Severe language problems early after infract due to air emboli, resolving over time with presence of paraphasias (type unspecified) and repetition, no comprehension difficulties.

*ACE-R = Addenbrooke’s Cognitive Examination‐Revised (Konstantinopoulou et al., 2011); AIDS = Acquired immune deficiency syndrome; BADA = Battery for Analysis of Aphasic Deficits (Miceli et al., 1994); BNT = Boston naming test (Kaplan et al., 1983); CELF-3 = Clinical evaluation of language fundamentals – 3rd edition (Semel et al., 1995); DART = Dutch Adult Reading Test (Schmand & Lindeboom, 1992); HSV = Herpes simples virus; MWT-B – Multiple choice vocabulary test (Lehrl, 1995); NA = Not applicable; SI = Streptococcus intermedius; HIV = Human immunodeficiency virus; TLC-E = Test for language competence – expanded edition (Wiig & Secord, 1989); WAIS = Wechsler Adult Intelligence Scale (Dahl, 1986).

Discussion

In this article, we provide an overview of CNS infections that trigger and/or relate to language impairments. The 41 reports we included represent a variety of disease-specific studies published over the last 45 years. The relatively recent publication date of these articles does not reflect an increased incidence of CNS infections (Bentivoglio et al., 2011). Rather, it is in line with advances in the fields of medicine and cognitive science, and a more widespread interest in language disorders and aphasiology (Prins & Bastiaanse, 2006). More importantly, language impairments are most commonly reported in three populations, namely: (i) children and adults with bacterial meningitis (four articles), (ii) adults with herpes simplex virus encephalitis (seven articles), and (iii) adults with meningoencephalitis due to HIV (nine articles). Parasitical encephalitis related to Falciparum malaria is discussed in only two articles, but in a very large number of cases (227 and 487, respectively).

We start off by discussing frequently described CNS infections, along with less common conditions. We highlight the language impairments reported most commonly, how these infections trigger or relate to brain damage (diffuse, focal, multifocal, preferred site of damage), the ways in which some CNS infections may help interpret brain-behaviour relations, and the anatomo-functional similarities between CNS infections and other neurological disorders. Finally, we discuss the use of CNS infections to study language in the brain.

1. Bacterial meningitis (S. pneumoniae, N. meningitidis, and H. influenzae)

It has an estimated incidence of one to five per 100,000 adults per year in developed countries and is estimated to be ten times higher in less developed countries (Bijlsma et al., 2016; Weisfelt et al., 2006). Different pathogens can cause bacterial meningitis, including S. pneumoniae, N. meningitidis, and H. influenzae (De Beek et al., 2002; Pentland et al., 2000). People with bacterial meningitis can present with fever, cold limbs, neck stiffness, and altered mental status (Bijlsma et al., 2016). The mortality rate (20–40%) and the incidence of neurological sequelae are high (15–40%, Hoogman et al., 2007; Weisfelt et al., 2006). Lesions due to bacterial meningitis tend to be diffuse (Akhaddar, 2017).

Adults with bacterial meningitis perform worse than healthy controls on tests of intelligence, attention, executive functioning, and reaction speed (De Beek et al., 2002; Kloek et al., 2020; Weisfelt et al., 2006). Three language tasks have been reported as being problematic for this population, namely, (i) category fluency (in particular, animal fluency; i.e., say as many animals as possible in one minute), (ii) object naming (i.e., to a picture of a horse, the participant is asked to respond “horse”), and (iii) the Token Test, that assesses comprehension of spoken instructions, requiring a good understanding of lexical items (geometric shapes, colours – Huber et al., 1983; De Renzi & Vignolo, 1962).

Bacterial meningitis results in more severe language problems and more extensive brain damage than viral meningitis (Schmidt et al., 2006; Sittinger et al., 2002). Children with bacterial meningitis (9–11 years) fare poorly in tasks that require good knowledge of language use/pragmatics (processing ambiguous sentences, making inferences, and figurative language) and not in tasks that require establishing semantic relationships, recalling sentences or language production/comprehension skills (Pentland et al., 2000). Bacterial meningitis in children may cause a language delay, but not a language impairment, as children tend to catch up and perform normally over time (Pentland et al., 2000). Hearing loss is observed in 7–9% of these children (Koomen et al., 2003).

We are not aware of studies on applied language or discourse ability in adults with bacterial meningitis, nor of reports of naming or fluency tasks in children with the same type of CNS infection. Hence, it is hard to judge whether children and adults with bacterial meningitis are similarly impaired. Further work is also needed to understand whether the language problems reported in people with bacterial meningitis are strictly linguistic and/or to what extent they result from impairments of other cognitive functions.

2. Herpes simplex virus encephalitis (HSV-1)

Herpes simplex virus encephalitis (HSV) is caused by the Herpes Simplex Virus 1 (HSV-1) which is the variety that causes orolabial lesions (cf. HSV-2 causing genital lesions). People with HSV encephalitis are commonly infected before adulthood. The virus crosses mucous membranes and travels to the trigeminal ganglion, where it remains latent, thus avoiding the host immune system. Through mechanisms that are not fully understood, the virus periodically reactivates and most often results in viral shedding or the common cold sore (Koyuncu et al., 2013). In some cases, reactivation results in brain infection and inflammation, leading to death in approximately 70% of untreated cases and in ~10-25% of treated cases (Modi et al., 2017; Whitley & Roizman, 2001).

HSV has been extensively discussed in studies on aphasia (for review, Gainotti, 2018). Warrington and Shallice (1984) were the first to link HSV infection with lexical-semantic deficits, and particularly with knowledge of biological entities vs non-biological items or artefacts. In HSV encephalitis both temporal lobes are frequently affected, and damage involves the temporal pole and often extends to the insular cortex (Baratelli et al., 2015; E. De Renzi & Lucchelli, 1994; Lacidogna et al., 2020). Many studies assessed participants with HSV encephalitis in several sessions, using standardized tools and tailor-made tasks to explore differences between different biological items and artifacts (Barbarotto et al., 2005), as well as language modalities (Baratelli et al., 2015; E. De Renzi & Lucchelli, 1994). The level of neurocognitive detail of HSV case reports and the systematic comparison with control groups is quite unique in comparison with reports of other brain infections.

3. HIV meningoencephalitis (human immunodeficiency virus)

When untreated, the human immunodeficiency virus (HIV) can weaken the immune system (i.e., acquired immune deficiency syndrome, AIDS) and make it unable to fight other opportunistic infections, such as pneumonia or tuberculosis (Piot & Quinn, 2013). HIV presents as a serious epidemic, as more than 35 million people have died of AIDS-related causes and ~39 million people are living with the disease worldwide (Bekker et al., 2018). Furthermore, HIV/AIDS is one of the leading causes of death in low-income economies and there are specific populations at risk, also in middle-income and higher income countries (Baral et al., 2012; Kral et al., 2003; Rodger et al., 2019). In recent years, antiretroviral therapy has helped diminish the mortality rate, prevent new HIV infections, and improve the quality of life of people infected with HIV (Bekker et al., 2018). However, people with HIV under antiretroviral therapy can present with subtle neurocognitive impairments (Rubin & Maki, 2019).

HIV-infected individuals may present with difficulties finding words in spontaneous speech, fluency tasks, and sentence repetition (Abusamra et al., 2012; P. J. McCabe et al., 2008; Kambanaros et al., 2019; P. McCabe et al., 2002; Millikin et al., 2004). In comparison with healthy controls, people with HIV fare worse in fluency tasks (Becker et al., 1997; Millikin et al., 2004; White et al., 1997; Woods et al., 2004) and in sentence repetition (Kambanaros et al., 2019). Interestingly, object naming does not tend to be impaired, even when performance is compared with that of a significantly older group of healthy controls (Sheppard et al., 2017; cf. P. McCabe et al., 2002).

Damage to subcortical frontal-basal ganglia regions supporting rule-guided search strategies has been proposed to account for these impairments (cf. lexical-semantic stores in temporal structures, Aylward et al., 1993; Clifford et al., 2017; Hestad et al., 1993). The argument relates to other disorders resulting from subcortical damage, such as aphasias due to periventricular white-matter damage and to striato-capsular lesions (Kuljic‐Obradovic, 2003) and vascular dementia (Graham et al., 2004). Damage to these subcortical circuits may affect the syntax-discourse interface, for example, repetition of complex syntactic sentences (Kambanaros et al., 2019). Finally, other authors suggested that such damage could also relate to reduced attention, that putatively plays a greater role in fluency tasks than in object naming (Sheppard et al., 2017).

Two other factors may be considered: (1) about 20% of people with HIV develop dementia, thereby proving to be more impaired than age-matched and older healthy individuals or showing premature/accelerated impairment (P. McCabe et al., 2002; Sheppard et al., 2017); and (2) people with HIV are immunosuppressed and therefore exposed to opportunistic infections. Lüscher and Horber (1992) present the case of a 27-year left-handed man with HIV who had toxoplasma encephalitis (parasite: toxoplasma gondii). This person had a lesion in the “left posterior white matter” and exhibited difficulties reading but not writing. He could write in German, read letters and simple words, but “would confabulate if asked to read whole sentences of words consisting of three or more syllables”. In a later report, a right-handed man with HIV and a mass lesion in the left lateral ventricle due to Toxoplasma infection fared poorly in language comprehension, naming and repetition, and presented with unspecified paraphasias in spontaneous speech (Özkaya et al., 2006).

4. Other types of CNS infections

4.1 Bacterial encephalitis: treponema pallidum

Treponema pallidum causes syphilis. This disease was very common until the discovery of penicillin in the late 1920s and its widespread availability in the 1940s (Fleming, 1929; Kent & Romanelli, 2008). The disease is now re-emerging (e.g., five cases per 100,000 in the US in 2012; 20-fold increase from 1989 to 1998 in China), reportedly in connection with specific populations with/or at higher risk of HIV infection (Bhai & Lyons, 2015; Zheng et al., 2011).

Syphilis is sexually transmitted through abrasions on skin/mucous membranes. It can affect any organ including the nervous system at any stage of the disease (i.e., primary, secondary, latent, and tertiary). Neurological signs can emerge at any stage of the disease and most commonly relate to meningitis, particularly in people with HIV (Bhai & Lyons, 2015). Historically, 12–25% of syphilis cases were diagnosed in psychiatric institutions, as neurosyphilis tends to be associated with grandiose ideas, delusions, seizures, personality changes, and varieties of dementia (Hook III & Marra, 1992). These symptoms emerge in late stages of tertiary syphilis (~10 years or more after infection) or so-called “general paresis” in which stroke (among other neurological disorders) can occur due to meningovascular invasion. In Zheng et al. (2011), 64 of 75 people with general paresis due to syphilis (85.3%) scored below norm in the Mini-Mental State Exam (range 3–19, mean = 14.7, SD = 3.2). In the same study, 107 of 116 people with syphilis (92.2%) were found to have brain atrophy, most frequently in the frontal and temporal lobes. Small vascular or demyelinating subcortical lesions were also observed.

Specific to language impairments, we found recent reports of two 50-yr old males with syphilis. One is presented with reduced verbal fluency and word finding difficulties (Ioannidis et al., 2014), the other with non-fluent speech and problems in both language production and comprehension (Budhram et al., 2017).

4.2 Bacterial encephalitis: streptococcus intermedius

Streptococcus intermedius can cause a brain abscess that begins focally as a localized area of cerebritis, to develop later into a purulent mass surrounded by a vascularized capsule (Melo & Raff, 1978). It is rare in developed countries (e.g., 1 in 100.000 or ~2000 cases per year in the USA); but poses significant problems in developing countries (Mamelak et al., 1995). Abscesses can appear anywhere in the brain (Muzumdar et al., 2011). However, the source of CNS infection typically relates to a typical site of abscess: middle-ear (temporal lobe, cerebellum); paranasal sinus (frontal lobe, deep temporal lobe, subdural empyema); hematogenous-metastatic (parietal, frontal, or temporal lobes); trauma and postoperative (contiguous with site of trauma, surgery).

Khaja et al. (2017) report on a 72-year-old female with a left temporal lobe abscess. This patient had naming problems, including the production of unrelated paraphasias in speech (e.g., “football” to “stethoscope”), and made errors in spelling backwards, in the context of fluent and coherent spontaneous speech. This profile indicates difficulties in lexical-semantic processing, which are common in people with other types of temporal lobe lesions (Binder et al., 2011; Haglund et al., 1994).

Jimenez et al. (2018) report on the case of a 45-year-old female with a left thalamic abscess. The patient had problems naming and with repetition upon hospital admission. Unfortunately, the language profile one month after treatment is not reported. Two other people with brain abscess (one in the occipital lobe, one multifocal) presented with confusion and aphasia without comprehension difficulties, as shown by the ability to follow simple verbal commands (Melo & Raff, 1978).

Finally, Nass et al. (1998) report on a 3-year-old girl with an abscess of undetermined cause involving the left subcortical and cortical angular gyrus, arcuate fasciculus, and posterior corpus callosum. The abscess was detected after the surgical removal of a carcinoma in the choroid plexus of the left ventricle. Due to the presence of a previous pathology (i.e., tumour), this study did not meet the inclusion criteria. However, the language report indicates that the child had reduced spontaneous speech, mild dysarthria and apraxia, semantic and phonological paraphasias during naming, and errors in repetition and reading letters that improved in follow-up assessments.

4.3 Bacterial encephalitis: thropheryma whippelii

Thropheryma whippelii can cause a rare illness named Whipple’s disease. This disease is characterized by digestive symptoms and weight loss. In 10–43% of cases, it also affects the CNS. When this happens, most symptoms are cognitive and psychiatric (for reviews, Compain et al., 2013; Gerard et al., 2002).

Manzel et al. (2000) report on a 41-year-old individual with damage in the mesial temporal lobes and midbrain. This subject scored below norm on category fluency. The same individual had normal scores in object naming and auditory comprehension and had no problem communicating. In this case, the difficulties in the category fluency task could be attributed to problems with executive functions rather than language. This interpretation is supported by the fact that the patient fared below normal in the Trial-Making Test (part B) and the Wisconsin Card-Sorting Test (Heaton et al., 1993). In another report, the patient fared below normal in category fluency and his family felt that he had difficulties in everyday communication. Scores returned to normal 2 years after treatment (Benito-León et al., 2008). Finally, Gerard et al. (2002) reviewed 12 cases of people with Whipple disease, of whom two were indicated as having “aphasia”. Unfortunately, no other information was included.

4.4 Viral encephalitis: SARS-CoV-2

Severe Acute Respiratory Syndrome CoronaVirus 2 (SARS-CoV-2) is the aetiologic agent of the novel coronavirus disease (COVID-19). Common symptoms of this disease include fever, fatigue, dry cough, and difficulty breathing leading to respiratory failure/pneumonia. Other symptoms include headache, loss of smell/taste, and diarrhoea (Mao et al., 2020; Rogers et al., 2020).

While the impact of COVID-19 on the CNS is still under scrutiny, some of the reported symptoms may be caused by CNS infection, by the response of the autoimmune system to the disease, or be secondary to pharmacological treatment (Baig et al., 2020). Anecdotally, this disease has been related to stroke or hypoxia in brain areas related to language, such as the temporal lobe and the hippocampus (Moriguchi et al., 2020; Al Saiegh et al., 2020; Troyer et al., 2020).

We found one report of 35 people aged 24–60 with COVID-19 and no other diseases, who fared poorly on tests assessing attention, memory, executive functions, and language (Almeria et al., 2020). Interestingly, the authors indicated that below-normal scores depended on clinical complaints or symptoms. Specific to language, people with headache scored below the norm in letter fluency; people with loss of taste had pathological scores in picture naming; and people who required oxygen therapy during hospitalization fared worse in category fluency.

4.5 Parasitical encephalitis: plasmodium falciparum

Malaria is a mosquito-borne disease that can be detected by a blood exam (Caraballo & King, 2014). Malaria has a profound public health impact in Africa, with nearly half a million people developing it every year, 100.000 deaths, and survivors enduring neurological/physical consequences (Postels et al., 2020). Malaria is caused by five different parasites of the Plasmodium group. Most deaths are attributed to plasmodium falciparum (Nadjm & Behrens, 2012). When it involves the brain, malaria can cause cognitive disorders (Idro et al., 2010).

Specific to language, children with malaria are more impaired than age-and-education-matched controls in a variety of production and comprehension tasks. Interestingly, Carter et al. (2006) argue that the language impairment (i) is often related to lesions in the language areas with a median age of circa 2 years; and (ii) may be the triggering cause of lower scores in other cognitive tasks. The same study also presented two individual case reports. In one case, poor performance on language tests could be due to illiteracy. Nonetheless, the second case attested that malaria can result in language impairments even in an educated child. The child presented problems possibly in post-lexical processes and in pragmatics (i.e., substituting phonemes in production, understanding sarcasm and humour), in the context of spared lexical-semantic or morphosyntactic aspects of language (i.e., “he was able to converse with the assessors in age-appropriate manner”). Two additional papers that did not meet the inclusion criteria also reported language impairments in people with malaria (Idro et al., 2010; Salih et al., 1991).

4.6 Parasitical encephalitis: taenia solium and others

The larval cysts of the tapeworm Taenia solium can infest the brain causing cysticercosis (also, neurocystercosis). People get cysticercosis by swallowing eggs lying in the faeces of a person who has an intestinal tapeworm (Brunet et al., 2015). These larval forms traverse the stomach wall and are carried to the brain via the bloodstream (Zhao et al., 2015). The disease is uncommon, and can occur more frequently in households where a person has been infected by an intestinal taenia. It has a higher prevalence in low-income countries (Bharani et al., 2001; Otero et al., 1989). Neuroimaging findings frequently show multiple areas of damage in the cerebral hemispheres, sometimes also including the basal ganglia, brain stem, and cerebellum (Zhao et al., 2015).

Paradis and Goldblum (1989) presented the case of a trilingual 25-year-old male with a cyst in the right precentral gyrus. This individual showed no problems in French across assessments, only problems in Gujarati early after surgery, and only problems in Malagasy at an 8-month follow-up. This case is interesting but problematic because it is not clear why different languages should be most impaired at different time points, or why damage to the right precentral gyrus should lead to such difficulties. Interestingly, this region featured as relevant in a quantitative meta-analysis of neuroimaging studies of language switching, along with other brain areas (Luk, Green, Abutalebi, & Gradi, 2012).

Otero et al. (1989) reported on a 7-year-old Spanish-speaking child with a cyst in the left sylvian fissure. The child presented comprehension problems and produced paraphasias, i.e., “llobeta” instead of “botella” [bottle]) and morphological errors in spontaneous speech (i.e., “él va casa” instead of “él va a la casa” [he goes home]). Also, we found three other single cases of people with language impairments in cysticercosis. These later reports did not meet the inclusion criteria. However, one of these reports describes a child with language impairments and epilepsy (Bharani et al., 2001). The two other studies report on women in their 40s with a cyst in the left sylvian fissure (Sawhney et al., 1998) and the left temporal lobe (Brunet et al., 2015), respectively.

Finally, N. fowleri is often mentioned in the context of primary amebic (meningo)encephalitis. It has an acute and fulminant prognosis, resulting in death very rapidly after exposure (i.e., 3–7 days). It occurs most often in children and young healthy adults with a recent history of freshwater swimming. Also, Acanthamoeba and B. mantrillaris are often mentioned in the context of granulomatous amoebic encephalitis. This type of infection has a longer clinical course and is often seen in people who are immunosuppressed, for example, due to HIV/AIDS or therapy for organ transplant (Ma et al., 1990; Trabelsi et al., 2012). Specific to language, infections with these parasites have been associated with difficulties in naming and reading (Baral & Vaidya, 2018; Jung et al., 2004; Keane et al., 2020; Salameh et al., 2015). These reports do not mention the language tests used for the functional diagnosis. Still, it is worth stressing that damage always involves the left temporal lobe – which is an area widely reported as relevant for language production (Binder et al., 2011; Haglund et al., 1994).

4.7 Bacterial meningoencephalitis: Mycobacterium tuberculosis

Mycobacterium tuberculosis (MTB) causes tuberculosis. The disease typically affects the lungs and can travel to the brain and meninges via the lymphatic system. Focal lesions in the posterior fossa are commonly reported in individuals younger than 20 years of age. Lesions in the frontal and parietal lobes are also reported in older individuals (Bernaerts et al., 2003; Nicolls et al., 2005). Tuberculosis is spread through the air (e.g., cough, saliva) and is diagnosed via a chest x-ray, microscopic examination of body fluids, or tuberculin skin tests (Pai et al., 2016). Tuberculosis is one of the top ten causes of death and the leading cause of death from a single infectious agent. It is curable and preventable (World Health Organization, 2020).

We found two reports of people with focal lesions due to tuberculosis (i.e., tuberculomas) affecting the left frontal and the left inferior parietal lobe, respectively (Booth & Curtis, 1893; Hinsdale, 1901). The authors reported on two men in their 20–30s with mild word-finding difficulties in spontaneous speech. One of the patients also had difficulty naming objects and reading words (Hinsdale, 1901). The paper by Booth and Curtis (1893) is beautifully described in a historical medical account by Shafi (2015).

A more recent paper describes a 24-year-old woman with lesions in the left parietal lobe, cerebellum, and arcuate fasciculus (Garcia-Grimshaw et al., 2018). She had no difficulty in comprehension or writing. However, she could not repeat simple sentences and produced occasional paraphasias in spontaneous speech.

Other papers describe participants with tuberculosis and language impairment (Raghunath et al., 2016). Some of these patients also suffer from HIV (Manea et al., 2015). There are also reports of people with tuberculous limbic encephalitis, an autoimmune disorder commonly seen in association with systemic autoimmune or infectious diseases (Daouda et al., 2017). However, these latter papers do not focus on language.

4.8 Bacterial meningoencephalitis: Borrelia burgdoferi

The bite of a tick carrying the bacterium Borrelia burgdoferi can cause Lyme disease. Some authors make a distinction between Lyme borreliosis (or neuroborreliosis) and Lyme disease, where the former refers to “neurologic symptoms with objective evidence of CNS infection”, while the latter refers to “persistent symptoms … despite standard treatment” (Westervelt & McCaffrey, 2002). We will not make this distinction and consider both acute and chronic cases, also because it is unclear if these terms are used interchangeably in the literature. Lyme disease reaches the CNS via the peripheral nerves. Neuroimaging findings tend to be nonspecific and predominantly include the frontal cortex and the arcuate fasciculus (Hildenbrand et al., 2009).

The disease has a high prevalence in North America and Europe (10 to 100.000 people) and the risk of infection is greater between June and October, when the ticks feed the most. In 3–15% of infected cases, Lyme disease affects cranial or spinal nerve roots (i.e., polyradiculitis), the meninges, and sometimes the spinal cord and the brain (Mygland et al., 2010; Rauer et al., 2020; Wormser et al., 2006).

People with Lyme disease have significantly more word-finding difficulties than age-matched controls, as indicated by both subjective reports and an object naming task (Svetina et al., 1999). Difficulties in fluency tasks have also been reported, but inconsistently (Krupp et al., 1991; cf. Sventina et al., 1999). Also, they may show hypoperfusion in the left parietotemporal junction (not necessarily due to structural brain damage or vasculitis) and present with difficulties in language production and comprehension, which usually resolve 24 hours after antibacterial and antiviral treatment (Sokolov, 2015). Nonetheless, problems in fluency and articulation can persist up to 4 years after infection (Benke et al., 1995). It is unclear whether any of these problems may also relate to difficulties in other cognitive domains such as verbal memory, mental flexibility, and associative functions (Krupp et al., 1991; for a review Westervelt & McCaffrey, 2002).

Other tick-borne infections such as babesiosis can induce language impairments (Buttar & Szema, 2019). Further details on the test materials used to assess language and replication of findings seem necessary.

4.8 Viral meningoencephalitis: enteroviruses

Enteroviruses are pathogens in the picornavirus family that are acquired by faecal-oral contamination and more rarely via respiratory droplets. They include viruses like poliovirus, coxsackievirus, and echovirus (Valcour et al., 2008). Enterovirus infection entails a favourable prognosis and does not affect language. For example, in Sittinger et al. (2002), 21 people with viral meningitis did not differ from age-matched controls in automated speech tasks (i.e., reciting the alphabet), lexical decision, and in identifying similarities between two words.

In more severe cases, enterovirus infection may induce meningoencephalitis and other severe symptoms, particularly in people with weak immune systems, such as new-borns and adults under immunosuppressive treatment (Cree et al., 2003; Modlin et al., 1991; Servais et al., 2010). For example, Mantri and Shah (2016) described a 28-year-old immunodepressed woman who showed aphasia and other signs mimicking dementia. Unfortunately, no details on test materials are provided.

Also, there exist reports of immunocompetent individuals. Kim et al. (2012) described a 23-year-old female with a left frontal lobe lesion who, 4 days after hospital admission, presented pathological scores in comprehension of complex sentences, word repetition, and object naming (the specific tests are not specified). The disorder was no longer present 3 months after drug treatment. Two additional studies report on immunocompetent males in their 70s, with left temporal lobe damage and language difficulties before treatment. In Enders et al. (2002), the participant is described as having “Broca’s aphasia” without further detail; in Valcour et al. (2008), the participant had “rambling and tangential speech”, echolalia, decreased comprehension, and hypophonic speech, mimicking dementia. Unfortunately, also in these cases the materials used for language assessment are not mentioned.

4.9 Fungal meningoencephalitis: aspergillus fumigatus

Aspergillus fumigatus is the most common type of fungus (mould) to affect the brain (Mishra et al., 2019). This fungus is part of the species aspergillus. We regularly inhale its spores. However, infection is typically reported in individuals with immunodeficiency (Antinori et al., 2013; Segal, 2009). This fungus involves the CNS only if it is widely disseminated, for example, if the infection is not properly treated. In fact, the brain is rarely the only site of infection (Hamill, 2009). When the CNS is affected, the infection may lead to seizures or focal lesions, such as a stroke or abscess (Memória Jr et al., 2020; Segal, 2009). Different patterns of cortico-subcortical involvement have been identified, including multiple lesions and damage to the orbitofrontal cortex (Ashdown et al., 1994; Kumar et al., 2018).

Specific to language, Lin et al. (2014) reported on a 32-year-old male whose immune reactions had been reduced in order to perform a stem cell transplant. Almost two months after hospital admission, the patient had a stroke in the left middle cerebral artery territory due to a pulmonary embolism caused by aspergillus. The patient had deficits in conversation, reading, and repetition, most of which resolved two weeks after the infarct. We found three other reports of people with aspergillus infection with aphasia, one following surgery for meningioma (Galassi et al., 1978) and two in immunocompetent patients who were found to have focal lesions in the left temporal (deShazo et al., 1997) and in the left frontal lobe (Memória Jr et al., 2020).

5. CNS infections and the study of language

Different issues emerge when reviewing studies of the impact of CNS infections on language. We will now discuss three main points: (1) the variability of assessment time points, (2) the underspecification and heterogeneity of test administered, and (3) the case of HSV as an example to follow. Finally, we will discuss other issues that may be worth considering in future studies.

5.1 Variability of assessment points

Some patients were evaluated in the acute stage (Garcia-Grimshaw et al., 2018), others shortly after treatment (Sokolov et al., 2015), yet others at 4–6 years from disease onset or treatment discontinuation (Barbarotto et al., 1996; Schmidt et al., 2006). The timing of language evaluation is relevant, as in the acute stage difficulties may not be due to the disruption of language-specific systems, rather, to delirium or confusion that reverses after treatment (Constantinides et al., 2018; Evoli et al., 2012; Fong et al., 2015; Kishi et al., 2010). Another issue is that the clinical neurocognitive condition in the acute stage may result from both the direct, agent-specific neural damage and from transient, aspecific reactions within the CNS such as oedema (Van de Beek et al., 2006). This heterogeneity makes it difficult to evaluate the incidence of language/cognitive deficits over time, the nature of brain damage (e.g., focal, multifocal, diffuse), the sensitivity to treatment, to mention but a few.

5.2 Underspecification and heterogeneity of tests administered

In relation to test materials, we found the expected differences between the tests used to assess children vs adults, between the assessment of literate vs illiterate participants (Castro-Caldas et al., 1995; Reis & Castro-Caldas, 1997), as well as across-language differences. In addition, batteries developed for the evaluation of stroke-related aphasia were used (Aachen Aphasia test, Huber et al., 1983; Battery for the Analysis of Deficits in people with Aphasia; Miceli et al., 1994; Bilingual aphasia test, Paradis & Libben, 1987; Boston Diagnostic Aphasia Examination, Goodglass & Kaplan, 1983; Western Aphasia Battery, Kertesz, 1982), as well tasks commonly used in other neuropsychological studies, such as category/letter fluency (Benton et al., 1983; Monsch et al., 1992), and object naming (Boston Naming Test, Kaplan et al., 1983). Spontaneous speech and tasks assessing language comprehension and grammatical processing (semantic judgment tasks, Ferreres et al., 2007; Token Test; Huber et al., 1983), and reading/writing (Schmand & Lindeboom, 1992) were less frequently used.

The lack of materials to evaluate people with CNS infections is certainly problematic. Language materials to assess people with post-stroke aphasia are designed to detect severe language impairments, as they require participants to say the name of frequent and familiar objects such as “book”, “jacket”, or “tiger” (Miceli et al., 1994). However, the impairments that we described in people with CNS infections may be better described by looking at specific stimulus types (biological vs non-biological items, like in the case of HSV) or when administering tasks like spontaneous speech, where participants include words in specific grammatical and metalinguistic contexts. In addition, the neuropsychological workups we reviewed were not necessarily designed to distinguish between language-specific and domain-general deficits. Hence, test batteries may include not only screeners for language disorders, but also screeners for delirium, confusion, diffuse cognitive damage, etc. (Van Eijk et al., 2009).

Furthermore, the tendency to consider language production rather than other tasks/skills does not necessarily reflect a greater incidence of difficulties with language production over, say, comprehension. Rather, it relates to a tendency to favour production tasks in clinical practice (Rofes et al., 2017a; Sevcik, 2006). In fact, when comprehension tasks were administered, language impairments were also unveiled (Abusamra et al., 2012; Lacidogna et al., 2020; Schmidt et al., 2006).

In future developments, it may be relevant to administer tasks that tap into comprehension, reading and writing, as well as less constrained language tasks, such as spontaneous speech. Also, to assess whether the language impairments reported in people with CNS infections are qualitatively and quantitatively similar to those observed in people with stroke, and/or if people with different kinds of CNS infections also suffer from different language impairments. Work similar to current studies comparing post-stroke aphasia and primary progressive aphasia may be relevant in this regard (Grossman & Irwin, 2018; Kordouli et al., 2018). Furthermore, we may also apply best practices for assessments at the acute stage, such as the use of short bedside tests that tap into production and comprehension processes (Flamand-Roze et al., 2011).

5.3 The study of HSV encephalitis

HSV encephalitis is an excellent model for the study of the representation of language in the brain, given its predilection for the temporal lobes. In this respect, it differs from other types of CNS infections such as cysticercosis or brain abscesses, which in principle can affect any brain area (Muzumdar et al., 2011; Otero et al., 1989). Furthermore, the administration of very detailed picture naming tasks to people with this condition allowed studying differences between biological categories (animals, fruits, vegetables) and artifacts (tools, furniture, vehicles) (Barbarotto et al., 1996; E. De Renzi & Lucchelli, 1994; Laiacona et al., 1993). Such studies fuelled a discussion on the possible reasons underlying naming failures at much greater depth than allowed by more general tests, like the Boston Naming test (Kaplan et al., 1983).

A relevant question that arises when considering the performance of people with HSV is whether different levels of difficulty between and within biological categories and artifacts relate to impairments accessing word meanings (e.g., a dog is a type of pet, the best friend of man) or word labels (e.g., the concept “dog” is referred to as “dog” in English, “hond” in Dutch, “cane” in Italian). This issue has been investigated using tasks other picture naming, such as lexical decision, word-picture matching, spontaneous speech, or fluency tasks. For example, Baratelli et al. (2015) report on a person with HSV encephalitis with comparable performance on biological entities and artifacts and poor scores in picture naming, naming to verbal definition and naming famous faces, while performing within the norm in word-picture matching and in a verbal semantic questionnaire. Similar results were reported by Lacidogna et al. (2020). Adding to this evidence, Barbarotto et al. (1996) reported on participants with HSV encephalitis who showed the different levels of difficulty between biological entities and artifacts, and in whom properties of the words (i.e., frequency, familiarity) influenced performance.

Beyond HSV, other types of CNS infection may be valuable to study the role of subcortical structures. For example, Whipple’s disease can affect the temporal lobes, but also subcortical structures such as the striatum, pallidum, and hypothalamus (Manzel et al., 2000). Also, people with HIV have subcortical damage, particularly in frontal-basal ganglia circuits (Aylward et al., 1993; Clifford et al., 2017). Subcortical structures can be studied with single electrodes, deep brain stimulation, and direct electrical stimulation (Duffau, 2020; Rofes et al., 2017b; Zanini et al., 2009). However, these invasive procedures require a multidisciplinary team, a surgical theatre, or testing in rather special and time-constrained circumstances during hospitalization (Rofes et al., 2019). In contrast, people who had CNS infections can undergo multiple baselines, long assessments, and follow-ups (Hanley et al., 1989; P. J. McCabe et al., 2008; Schmidt et al., 2006). These aspects certainly allow a more extensive and intricate use of language assessment tasks in this population, while providing excellent data to understand how language is processed in the brain.

6. Other issues to consider in future work

A few articles that appeared in our search referred to language impairments as “aphasia” or “dysphasia” without specifying the tasks or procedures used to provide this diagnostic. Some of these articles discussed infectious agents and/or diseases such as cryptococcus (Searls et al., 2009), dengue (Soares et al., 2006), granulomatosis (McKeith, 1985; Sahn & Sahn, 1976), hepatitis (Córdoba et al., 2003; Dalton et al., 2017), mucormycosis/zygomycosis (Galetta et al., 1990; Hazama et al., 2017; Kerezoudis et al., 2019), nocardia (Matulionyte et al., 2004; Yamamoto et al., 2017), and pyogenic ventriculitis (Lee, 1977).

Other articles do not provide evidence regarding brain damage. While this may not be as feasible in specific settings or visible even after MR scanning (Córdoba et al., 2003; Potchen et al., 2013), understanding whether some brain areas are more commonly affected by an infectious agent seems relevant to draw objective relations between language impairments and brain damage.

We noticed an incipient number of studies on autoimmune limbic encephalitis. We did not include such studies, as the condition is generally associated with the presence of a tumour elsewhere in the body and not to CNS infection (Tüzün & Dalmau, 2007). However, this condition resembles some CNS infections we mentioned, and difficulties in naming and fluency tasks are often reported (Constantinides et al., 2018; Evoli et al., 2012; Kishi et al., 2010).

Viruses are living agents that mutate and may affect the brain in manners difficult to predict. For example, the influenza virus of 1918 (Spanish flu, H1N1) mutated and became less aggressive, only to “re-appear” under a different subtype in 1957 (Asian flu, H2N2), and again in 1968 and 1979 (H3N2) (Tauenberger, & Morens, 2006). Also, drug and medical practices become more effective over time. These two factors add variability to cross-sectional studies where patients with “the same virus” may be recruited over separate periods of time.

Identifying people at risk and the impact of co-occurring conditions is a relevant because CNS do not necessarily have a greater incidence in older populations or in people with cardiovascular risk factors (Idro et al., 2010; cf. Armstrong et al., 2019), and because specific populations may be at higher risk for one or more concomitant infections (Ashdown et al., 1994; Özkaya et al., 2006; Rodger et al., 2019).

A question that remains open is whether our results are influenced by a publication bias. The majority of studies that qualified for this review focus on languages “native” to the western world. The only exceptions are the reports on Falciparum malariae, that included speakers of Luganda (Bangirana et al., 2016) and Mijikenda (Carter et al., 2006), a report of a Korean speaker with meningoncephalitis (Kim et al., 2012), and a report of a Gujarati French and Malagasy speaker with cysticercosis, who lived in Canada (Paradis & Goldblum, 1989).

Finally, even though CNS infections and complications thereof are relatively rare, mostly owing to improved therapies (Sarrazin et al., 2012), updates on the neurocognitive sequelae of these diseases are necessary, especially considering the upsurge of new infections such as the new coronavirus disease 2019 (COVID-19), which we briefly reported on in this review.

Conclusions

We reviewed 41 articles and 18 infectious agents that relate to/trigger difficulties with language tasks. Certain types CNS infections were more commonly reported as single cases, others as a group. Lesions affecting the left hemisphere and in particular the temporal lobe, where most common, but in many studies details on brain damage were not provided. Herpes simplex encephalitis is studied in most detail. Other CNS infections can be studied with a similar approach, by applying knowledge gained from studies of people with other neurological disorders. These studies represent a non-invasive methodology for the study of language in the brain.

Disclosure of interest

No potential conflict of interest was reported by the author(s).

Acknowledgments

We thank Joy Zhuo Ding and Vera Hermanns for comments in early stages of writing this manuscript. We also thank Sjoukje van der Werf for advice regarding the literature search and Marta Almeria for kindly replying to questions on her work.

Additional information

Funding

No specific funding was obtained for this research.