Despite a reported decline in numbers especially since the introduction of the conjugate Haemophilus influenzae type b vaccine [23, 24], one third of children with bacterial meningitis still die and over 25 % are left with severe sequelae [21, 25–27]. The main pathophysiological mechanism is meningeal inflammation. Bacterial cell wall antigens activate an acute inflammatory process in the sub-arachnoid space producing oxidant stress and inflammatory injury to neurons [28–30]. Infection may extend into the labyrinth and cause sensori-neural deafness [31] or along penetrating vessels causing a vasculitis. Involvement of the vascular intima may obliterate vessels and cause necrosis of brain tissue. Infarcts may be limited to one vessel or involve large cortical areas resulting into hemiparesis or quadriparesis and subsequently, epilepsy and learning disability [32–35]. Acute hydrocephalus and brain swelling worsen intracranial hypertension, which reduces cerebral perfusion further leading to ischemic injury [35, 36].

There are several patient management challenges that affect the treatment of bacterial meningitis in Africa. These include misdiagnosis (in many malaria endemic areas, the default position for many clinicians is to treat a febrile illness in a child with an anti malarial), delayed presentation and therefore late initiation of specific antibiotic therapy and a growing problem of multiple drug resistance with indiscriminate antibiotic use. Thus, although intravenous ceftriaxone, 100 mg/kg/day is first line therapy in many centres, there are reports of drug resistance and treatment failures yet, alternative treatments are few. Furthermore, adjuvant therapy with steroids which has been shown to improve outcome elsewhere has not shown similar benefits in Africa probably because of the high burden of HIV/AIDS or exposure to antibiotics prior to presentation [37]. Upon recovery, follow up care and rehabilitation is almost non-existent in most centres. Thus, the long-term outcome of bacterial meningitis in the region remains poor.

Malaria is a major cause of ill health, neurodisability and death in tropical countries. Most transmission occurs in sub-Saharan Africa and in South East Asia. In 2010, the World Health Organization estimated that there were 216 million clinical cases of malaria and 655,000 deaths. Most of the deaths occurred in Africa among children younger than 5 years. Here, the incidence of disease is lower and less severe in the older children and adults most probably, as a result of the building immunity [38]. Recent reports suggest that the incidence of malaria is on the decline in many endemic countries [39, 40].

Cerebral malaria is a diffuse encephalopathy characterized by parasite sequestration in post capillary venules as a consequence of adherence of infected erythrocytes to the endothelial cell lining [41]. Diagnosis is made in a patient with peripheral malaria parasitemia, unrousable coma and no other cause to explain the coma [41]. Brain injury probably results from hypoxic or inflammatory injury initially affecting the cerebral micro-vascular endothelium [42]. The sequestered mass reduces blood flow and alters blood-brain barrier function but without significant leakage of plasma proteins into perivascular spaces [43]. Changes in blood brain barrier function may be a result of minute disruptions in endothelial junctions [43, 44]. Blood brain barrier dysfunction may contribute to intracranial hypertension [45, 46]. However, increased cerebral blood volume following sequestration and increased blood flow in seizures, hyperthermia and anemia may better explain the increased intracranial pressure [47, 48]. A critical reduction in metabolite supply may occur but significant neural necrosis is unlikely since with anti malaria treatment, coma is reversible. However, during periods of increased metabolic demand (e.g. during seizures), the risk of neural injury is high and is worse in patients with hypoglycemia [49] or when blood flow is further compromised by intracranial hypertension [46]. Many with severe intracranial hypertension die or survive with sequelae [50]. Thus, sequelae have been associated with prolonged and repeated seizures, deep and prolonged coma, intracranial hypertension and hypoglycemia [51].

In cerebral malaria, 15–20 % die and 25 % have persistent neurologic, cognitive and behavioral impairments or epilepsy [41, 52, 53]. These impairments affect the child’s development and quality of life, placing an economic and social burden on families that often have limited resources. Supportive therapy may improve outcome, but trials of adjunct therapies have been disappointing.

Viral infections of the central nervous system are an important cause of hospitalization and death in children in sub-Saharan Africa. Etiology is diverse. In a study of 513 Malawian children with suspected central nervous system viral infections from 2002 to 2004, at least one virus was detected in 133 children (26 %) of whom, 43 (33 %) died. Twelve different viruses were detected. Adenovirus was the most common infection, affecting 42 children. Mumps, human herpes virus 6, cytomegalovirus, herpes simplex virus 1 and enterovirus were also important. These infections were as important even among children whose coma was attributably solely to cerebral malaria [54]. Forty five (9 %) of the 513 children had both plasmodium falciparum parasitemia and viral infection, including 27 (35 %) of 78 diagnosed clinically with cerebral malaria. Children with dual infection were more likely to have seizures than those with malaria parasitemia alone, viral infection only, or neither. Seventeen children (38 %) of the 45 children with dual infection died, compared with 26/88 (30 %) with viral infection only, 17/118 (14 %) of parasitemia only, and 34/262 (13 %) with neither. This and a second study in Kenya [55] suggested that interaction between viral central nervous system infections and malaria parasitemia could increase disease severity.

Rabies is an important problem in urban centres. The clinical manifestations may be unusual and diagnosis may be made at post-mortem. In the Malawi series, 3/26 (11.5 %) of 26 fatal cases were mistakenly originally attributed to be cerebral malaria [56].

Brain injury in viral encephalitis is caused by either direct (primary) or indirect (post-infectious) involvement of brain tissue. In primary disease, viruses replicate in neurons, glia or macrophages, both in white and grey mater. In herpes encephalitis, there is necrosis of brain tissue with hemorrhage and softening and in severe cases, extensive loss of neurons. Mortality is high and among surviving patients, long-term moderate to severe sequelae develop in up to 35 %. These include hemi-paresis, epilepsy, aphasia and behavioral problems [57–59].

Of the three encephalopathies, the diagnosis and management of viral encephalitis is least developed in Africa. Laboratory capacity (e.g. for PCR diagnosis) is poor and very few centers have the diagnostic capacity to offer appropriate management. Thus, the provision of specific treatment even for etiologies for which specific drug therapies are available, such as herpes simplex encephalitis, still has a long way to go.

In 2012, the World Health Organization estimated that over 90 % of the approximately 3.4 million children with human immunodeficiency virus worldwide were in sub-Saharan Africa. Most paediatric HIV disease is vertically acquired through mother-to-child transmission and many patients are not on antiretroviral therapy (ART) [60]. The neurological complications of HIV-1 can be divided into those due to direct HIV-1 infection of the brain (primary disorders such as HIV encephalopathy) or due to indirect complications (opportunistic infections and malignancies). In addition, children infected with HIV are also at risk of other CNS disorders unrelated to the underlying infection.

HIV encephalopathy is the hallmark of untreated primary HIV infection of the brain. It presents with a triad of acquired microcephaly, neurodevelopmental delay and progressive motor dysfunction that are not attributable to other biological and environmental risk factors. This is a clinical consequence of early HIV invasion of the developing fetal and infant brain and characteristically presents in infancy and in toddlers. The commonest neurologic findings are a global developmental delay or regression together with pyramidal tract signs and seizures. The presentation differs according to age and mode of HIV-1 infection. Infants and younger children manifest the most severe and global dysfunction, a smaller head circumference, and lower birth weight, while older children have more specific signs. In one series, the mean head circumference was below the third percentile in 40 % of HIV-infected compared with 22 % of uninfected children [61]. The encephalopathy can be progressive with loss of acquired skills or stagnation/plateau in attaining developmental milestones or static in which case the children acquires new skills and abilities more slowly than normal. Their standardised test scores are below average, but stable. HIV infected children may also develop other mental and cognitive deficits evident at different ages [62]. Behavioral problems are more common in preschool and school aged children while focal seizures are more common in older children and pre-adolescents. Rapid progression of the disease carries a grave prognosis [63].

Other manifestations of HIV nervous system disease include stroke, myelopathies, peripheral neuropathy, myopathy, epilepsy, CNS lymphomas and neuropsychiatric difficulties including Attention Deficit Hyperactivity Disorders, ADHD. Stroke is the most common cause of focal neurological deficits. This may be a direct result of HIV infection, may be secondary to opportunistic infections or HIV arteriopathy. Adolescents with long standing disease may also develop accelerated atherosclerotic cerebrovascular disease. Thus, management of stroke in these children should include low-dose aspirin administration [64]. Epilepsy may be a direct consequence of HIV infection or may be an acquired pathology. Because of its limited drug to drug interactions with most antiretroviral medications, sodium valproate is suggested as the first line anti-epileptic drug. The second line drug is lamotrigine. Structural cord lesions must be excluded in children presenting with paraparesis. Peripheral neuropathy and myopathies are often compounded by antiretroviral therapy. Other manifestations of central nervous system involvement and brain injury in HIV/AIDS include progressive multifocal leukoencephalopathy (PML) [65].

Brain imaging in HIV encephalopathy shows cortical cerebral atrophy with dilatation of the lateral ventricles and calcification of the basal ganglia and, peri-ventricular involvement of the white matter [63]. These changes may predate neurological deterioration. Patients may show clinical, cognitive and functional improvement or plateauing of symptoms with initiation of highly active antiretroviral therapy (HAART), although abnormal neurological signs and gross motor difficulties persist [66]. The non-nucleoside reverse transcriptase inhibitors such as nevirapine have the best potential for treatment of CNS disease because of CNS penetration [67]. Differences in the effects on cognitive function in different settings may be a result of infection by different HIV clades [68]. Micronutrient deficiencies may worsen outcomes [69]. The long term preventive strategy should be interventions to prevent HIV/AIDS.

In 2007, the World Health Organization estimated that there were 9.27 million new clinical cases of mycobacterium tuberculosis (139/100,000) and 13.7 million prevalent cases (206/100,000). Most patients have pulmonary tuberculosis (TB). Central nervous system involvement, one of the most devastating clinical manifestations of TB, accounts for 5–10 % of extrapulmonary cases, and for 1 % of all TB cases. The HIV/AIDS epidemic has ballooned the problem. Today, 1.4 million new cases of TB occur in HIV infected individuals annually [70] and TB also accounts for 23 % of AIDS related deaths.

Childhood TBM usually develops within 3 months of primary TB infection [71]. A family history of TB is common. Most patients present with fever. Neurological features range from lethargy, stiff neck, seizures, agitation or coma. Cranial nerve palsies are common and may be the presenting manifestation. Fundoscopy may show papilloedema and choroid tubercles especially in patients in who TBM is associated with miliary tuberculosis [72]. Clinically, the degree of neurological involvement and extent of brain injury correlates with stages of the contemporary modification [73] of the Medical Research Council Staging of TBM. These stages are:

Cerebrovascular complications typically involve regions supplied by the middle cerebral artery and its perforating branches. These lesions may be a consequence of local inflammatory exudates which may also trap the cranial nerves. Infiltrative, proliferative and necrotising vessel pathologies may lead to luminal thrombosis and single or multiple infarcts. The brain CT scan can help diagnose TBM, and is useful in decisions regarding surgical interventions for hydrocephalus. Choroid plexus enhancement with ventricular enlargement on imaging is highly suggestive of TBM. The MRI shows diffuse, thick, meningeal enhancement. Cerebral infarcts can be seen in nearly 30 % of cases [74].

The first-line treatment regiment for TBM is a combination of daily Isoniazid, Rifampicin, Pyrazinamide and Ethambutol. The most important determinant of outcome is the stage at which treatment is started. Other determinants include age, malnutrition, hydrocephalus and focal neurological deficits. Mortality and morbidity is low in stage I disease but in stage III almost 50 % of patients die. Survivors manifest a variety of neurological sequelae [75].

Other forms of central nervous system involvement include intracranial TB such as TB encephalopathy, vasculopathy, tuberculoma and brain abscess or spinal forms such as Pott’s disease, non-osseous spinal tuberculoma and spinal meningitis.

Nodding syndrome is a poorly understood but devastating chronic brain disorder affecting thousands of individuals in the eastern African countries of South Sudan, Uganda and Tanzania [76]. There are reports of possible cases in west (Liberia) and central Africa (Burundi and the Democratic Republic of Congo). In Tanzania, nodding syndrome has affected a relatively smaller number of individuals. The picture in South Sudan and northern Uganda is that of an epidemic. The World Health Organization recently classified nodding syndrome under the neglected tropical diseases [80]. Clinical studies suggest that it may be an epileptic encephalopathy and probably symptomatic generalised epilepsy disorder [76–79]. The syndrome affects previously normally developing children. Symptoms develop between the ages of 3–15 years [78, 80]. Head nodding, the pathognomonic feature is an atonic seizure [78]. Over the years, this is complicated by frequent tonic clonic, myoclonic and atypical absence seizures, declining cognitive and motor function, psychiatric disorders, wasting, growth failure and physical deformities leading to severe disability and in some cases death. These complications develop through five clinical stages of a prodrome, head-nodding, multiple seizures and severe disability. These time intervals may potentially provide windows for intervention to arrest progression.

The background electroencephalogram (EEG) is characterised by generalised slow wave activity and multiple inter-ictal epileptiform discharges. Ictal EEG activity consists of mostly generalised spike and spike and wave discharges. Brain imaging shows generalised cerebral and cerebellar cortical atrophy [76–78].

A number of toxic, nutritional, infectious, para-infectious and environmental causes have been studied but to date, there have been no leads except an epidemiologic association with O. volvulus (reviewed in [81]). Testing for 19 virus families has been negative. The affected age group and duration of symptoms makes prion disease unlikely. The EEG and brain MRI too are uncharacteristic [76, 77]. Clustering of cases within specific locations and within families suggest a common exposure [77]. The affected communities are not known for consanguineous marriages and no specific epilepsy genes were identified on exon sequencing in two children [81]. In south Sudan, 76 % of cases and 47.4 % controls had O. volvulus microfilaria on skin snip testing [82], while in Uganda, 94.9 % cases and 48.8 % controls tested positive for O. volvulus specific antibodies [83]. However, O. volvulus is endemic in many parts of Africa, Latin America and Asia, yet nodding syndrome has only been reported in a few areas. Also, it is unclear how the parasites would cause brain injury as there is no evidence of breach of blood brain barrier [76]. Alternative mechanisms other than direct parenchymal injury are likely.

Using proteomics, American investigators at the Centers of Disease Control and the National Institutes of Health have demonstrated antibodies against leiomodin-1 (a muscle protein also expressed in neurons) in 11/19 (58 %) cases. Part of this protein shares 83 % sequence similarity with a conserved region of O. volvulus tropomyosin. The antibodies were neuro-toxic in mice brain suggesting that neuropathology in nodding syndrome may be caused by cross-reacting antibodies (Tory Johnson et al., unpublished). Separately, we studied serum samples of 31 patients and 11 sibling controls for antibodies against the neuron surface protein – voltage-gated potassium channel (VGKC) complex proteins and against the intracellular glutamic acid decarboxylase (GAD); 15/31(48.3 %) cases and 1/11(9.1 %) controls had antibodies against the VGKC but none tested positive for antibodies against the intracellular GAD, (Richard Idro, unpublished). These pilot studies await confirmation. Other than O. volvulus, another source of cross-reacting antibodies may be host response to variant Wolbachia species which are essential symbiotic bacteria of filarial worms. Again, further studies are awaited.