Few cases of ADEM developing after bacterial meningitis have been reported so far, and its diagnosis may be difficult. This diagnosis should be suspected in case of lack of clinical improvement, recurrence or onset of new neurological symptoms in patients receiving adequate antibiotic and supportive therapy (33). In some cases, after an initial improvement, patients can develop a sudden and unexplained encephalopathic symptoms (27, 29), as described in our patient. In other cases, patients may show persistence of altered mental status and fever or even severe neurological deterioration, requiring mechanical ventilation (19, 26, 30, 31). In these situations, the criterion of “encephalopathy”, necessary for the clinical suspicion of ADEM, may be difficult to assess, as an altered consciousness can be partly explained by fever or by the underlying CNS infection. Lumbar puncture and brain and spinal cord MRI are fundamental tools for differential diagnosis. MRI scan allows to assess the number, location and extension of the lesions; when performed in the acute phase of ADEM, it tipically shows diffuse, poorly demarcated large lesions involving predominantly the cerebral white matter. Deep grey matter lesions in the thalami or basal ganglia can be also present (9, 30, 31). CSF examination is useful to differentiate ADEM from other inflammatory or autoimmune diseases, and to determine the presence of auto-antibodies (e.g. anti-myelin oligodendrocyte glycoprotein, MOG) which can be indicative for recurrence (3, 34–39).
Due to the limited number of cases and diagnostic difficulties, it is unclear if meningoencephalitis progressing to ADEM is a truly rare event or just underreported. Moreover, the causative microorganism can be identified only in a small number of patients, making it difficult to identify risk factors and pathogenetic mechanisms (33, 40, 41). Streptococcus pneumoniae has been rarely described, and the development of related demyelinating CNS disease is reported only after serious infections, like meningitis. ADEM is traditionally considered a post-infectious disorder, as the time before the onset of neurological symptoms is usually sufficiently long for the development of an adaptive immune response (2–4 weeks). In this case, molecular mimicry between causative pathogens and host cells causes T-cell activation (42). In experimental animal models of autoimmune allergic encephalitis (EAE) induced by S. pneumoniae, damage was mediated by toll-like receptor 2 (TLR2), which can specifically recognize teichoic and lipoteichoic acid (LTA), major constituents of the cell wall of this bacterium. Pneumolysin is also an agonist of TRL4. Also tumour necrosis factor-alpha (TNFα) and interleukin-6 (IL-6) have been indicated as important immunostimulatory mediators, activating leukocytes and microglial cells, and which can enter the brain tissue in regions without a tight blood-brain barrier (43, 44).
In our paediatric patient, such as in other cases reported in the literature (26, 28, 30), the onset of neurological symptoms occurred within a shorter interval (few days). This rather accounts for a para-infectious mechanism, as latency is too rapid to allow an autoimmune reaction after a primary exposure to a microorganism. Direct invasion and damage of the CNS by S. pneumoniae is supposed in this case, as already described for other pathogens (45). No post-mortem studies are available, thought, to support this hypothesis, while brain biopsy was performed only in one patient (26), without identification of any microbal material. Disruption of the blood-brain barrier by the pathogen can allow CNS-confined autoantigens to leak into the systemic circulation, causing breakdown of tolerance and consequent induction of self-reactive T-cell activation (46). This mechanism could explain why the short latency of onset of neurological symptoms usually occurs after severe pneumococcal infection, while milder infections require a longer time to activate an immune response, as described in the EAE murine model (44).
In this light, risk factors for the development of severe S. pneumoniae infection have to be considered. In our pediatric case, a bone breach due to a previous trauma was regarded as the gateway for infection, and the lack of an adequate vaccination schedule did not prevent the girl to develop a serious course of the disease. Also post-surgical breaches, inadequately treated upper-airway infections, immunosuppression and splenectomy (26) have been reported in patients who developed ADEM after bacterial meningitis. In the few pediatric cases described, inadequate or absent immunization against S. pneumoniae was reported (30, 31). Though ADEM is primarily a pediatric disease (1), vaccinations campaigns and the widespread diffusion of pneumococcal immunization in children could be the reasons, in this case, of the lower incidence of neurological complications in this age group. Conversely, even if ADEM can also be considered a post-vaccinal complication, only one case has been described following S. pneumoniae vaccination in an adult patient (47). These evidences should further reinforce the need for vaccinations, at all ages.
First-line treatment of ADEM is widely based on intravenous high-dose methylprednisolone, eventually followed by a slow oral tapering with prednisone over 4–6 weeks. IVIG are considered in case of failure, as a second line treatment (48, 49). All reported cases of ADEM following pneumococcal meningitis were treated with pulse doses of steroids for 3–10 days (Table 1), which proved to be effective. Another patient, following pneumococcal meningitis, was diagnosed with acute transverse myelitis associated with evidence of widespread white matter brain lesions, which though did not fulfil the diagnostic criteria for ADEM; he was treated with high-dose steroids, with prompt recovery (50). In only one pediatric case steroids and IVIG were administered at the same time (31). The occurrence of demyelination and worsening neurological symptoms few days after a bacterial meningitis, though, may create some uncertainty on the use of high-dose steroids. In our case, IVIG were chosen a first-line therapy to avoid further immunosuppression in a patient whose treatment with antibiotics had just started and with a probable yet inadequate clearance of the bacterial agent from the CNS. Because of re-exacerbation of symptoms at the eleventh day after admission, she was indeed treated with high-dose steroids, with complete recovery. This worsening did not fulfil the criteria for multiphasic ADEM, as it occurred before than three months after the onset (9), but it was considered due to a lack of proper response to the first-line treatment. A similar treatment course was described by Jorens in 2005 in an adult woman (51). In this case, two days after diagnosis of pneumococcal meningitis, patchy areas of increased signal intensity on T2-weighted images were noticed in both white and grey matter, interpreted as focal areas of ischaemia with cytotoxic oedema, secondary to necrotising vasculitis and thrombosis. First-line treatment with IVIG was ineffective, so high-dose corticosteroids were administered. Even though this patient did not properly fulfil ADEM criteria, myelin disruption was thought to be consequent to parenchymal insult and immunomodulatory therapy was therefore effective.
ADEM following pneumococcal meningitis is rare, and differential diagnosis can be challenging, especially in children. Further studies are needed to clarify the pathogenetic mechanisms and to evaluate the proper therapeutic approach, which have to be administered promptly in order to avoid long-term sequelae.