PMNS has been defined as a complex neurological condition developing within 2 months after fully recovery from a malaria episode [4]. The first reports date back to 1966. Since then a number of case reports and case series of PMNS have been described. Incidence rates of PMNS has been estimated to range from 1.2 per 1000 cases in uncomplicated forms, up to 18 out of 1000 in severe forms. PMNS after P. falciparum infections are the most frequently reported [4]. Noteworthy, most of the PMNS presenting as ADEM occurred after non-falciparum malaria [6], but the reason for this association is not clear.
Clinical presentation
Clinical findings in “classical PMNS” are extremely heterogeneous, ranging from acute encephalitis in almost 80% of the patients, confusion and fever in over half of the cases, seizures (33%), language disturbance (about 30%), tremor (23%), myoclonus (11%), and psychiatric illness (17%) [5]. Neurocognitive decline, acalculia, and, rarely, autonomic system disorders, ophtalmoplegia and cranial nerve palsy have been described as well. DCA is most typically characterized by gait ataxia and isolated cerebellar syndrome, whereas in ADEM-like forms motor impairment in predominant [8, 12].
Diagnostic methods
Overall, instrumental tests seemingly did not represent valuable diagnostic tools in any of the described cases. In a minority of reports, MR showed nonspecific signal uptake in white matter. The most frequent abnormalities are localized in periventricular areas, basal ganglia, brain stem, and cerebellum. Spinal cord and optic nerve involvement may also occur albeit more rarely [13]. Commonly, the diagnosis of ADEM is suggested by MRI multifocal, bilateral, asymmetric, white matter abnormalities, with hyperintense appearance in T2 and FLAIR sequences [14]. Localization of MR lesions is thus unrelated to clinical presentation (e.g., ataxia does not necessarily correlates with cerebellar lesions) [8].
Pathogenesis
An immune-mediated process is the most corroborated hypothesis regarding PMNS pathogenesis. Different mechanisms have been proposed.
Auto-reacting T-cells could be triggered by a certain grade of molecular mimicry and nonspecific activation, leading to autoimmune response toward CNS antigens [6, 15]. This mechanism is also a well-known prerequisite of classical AIE, AIDP and ADEM.
In the wake of the immunologic-theory, same authors described a series of PMNS cases developing as classical autoimmune encephalitis (AIE) mediated by neuronal antibodies against ions channels and synapses. Alike AIE, PMNS has been associated to the production of N-methyld-aspartate-receptor -NMDAR, antibodies, anti-voltage-gated-potassium-channel (VGKC) antibodies, anti neuroanexin α3 antibodies. In these cases, the disappearance of autoantibodies and MRI lesions have been documented following steroid treatment cycles [16, 17].
Polyclonal B cell over-activation is another possible immunological mechanism. It has been demonstrated that Plasmodium-parasitized erythrocytes express several membrane microbial immunoglobulin binding proteins (IBPs), which persist over time following parasite eradication. Some IBPs, such as P. falciparum erythrocyte membrane protein 1 (PfEMP1) extensively bound to different circulating human immunoglobulins, thus leading to direct B-lymphocyte stimulation and subsequent secretion of different antibodies [18]. This theory might explain the frequent finding of elevated titres of IgG and IgM antibodies against multiple viruses during PMNS. Despite the exclusion of any concomitant infectious disease as underlying cause of the encephalopathy, the patient presented various degrees of IgG and IgM positivity against multiple viruses, low positivity for ANA, CSF lymphocytic pleocytosis and intrathecal immunoglobulin production.
The case study protagonist developed a severe form of PMNS characterized by a dramatic neuropsychiatric pattern with abnormal generalized immune-activation and severe blood brain barrier alteration. Indeed, clinical presentation, immunological findings, and rapid response to steroids endorsed the immune-mediated trigger theory.
Two alternative hypotheses regarding PMNS pathogenesis have been postulated so far: 1) transient ischaemia; 2) cytokine storm.
The ischaemic hypothesis suggests that parasitized red blood cells adhere to endothelia and reversibly obstruct the brain microvasculature. It has been proposed that the higher tendency to cytoadhere of P. falciparum might explain the higher prevalence of PMNS in this species [19]. Hsieh et al. reported a case of PMNS where brain SPECT revealed decreased radiolabelled agent incorporation in cerebral hemispheres, hence suggesting impairment of cerebral microcirculation [20, 21]. However, the absence of parasitaemia during PMNS and the time laps between malaria episodes and post-infectious syndrome remains unexplained. Furthermore, an ischaemic genesis is much more compatible with cerebral malaria than PMNS.
The cytokine storm hypothesis derived from a single in vivo study. de Silva HJ et al. conducted a prospective observational analysis in 12 patients with post-malaria DCA, reporting significantly higher levels of pro-inflammatory cytokines, such as tumour necrosis factor (TNF), interleukin 6 (IL-6) and interleukin 2 (IL-2), in both serum and CSF, and comparing PMNS affected patients to non PMNS affected ones (8 patients). Indeed, TNF levels in CSF have been linked to more severe and disabling forms of CM [22, 23]. It has been suggested that TNF might play a pathogenic role in CNS damage in CM, by possibly promoting parasite sequestration and endothelial activation. A similar role may be hypothesized also in PMNS. However, no prospective analysis has been conducted to date in post-malarial disorders.
The role of anti-malarial treatment
During the past decade, a series of studies suggested that the type of anti-malarial treatment could correlate to occurrence of PMNS. In particular, the administration of mefloquine and atovaquone-proguanil was recognized as a risk factor for PMNS [24, 25]. The quinoline anti-malarials (and especially mefloquine) have been associated to neurological disorders, both in form of psychiatric symptoms, and central anticholinergic syndrome [4]. Some of these manifestations overlap with PMNS.
However, to date no significant association of PMNS with anti-malarial treatment has been found.
Firstly, quinoline-related effects are self-limiting, and they rapidly fade. None of the anti-malarial drugs has been related to iatrogenic ADEM-like, AIP, or DCA-like toxicity. Furthermore, not all the patients presenting with PMNS received quinolines. Notably, most of the cases emerged days or weeks after quinoline-withdrawal.
Treatment and outcome
Overall, the majority of PMNS cases revert without specific treatment. Prognosis is generally good and no long-term sequelae have been described. Several off-label treatments have been proposed for most severe forms. High dose steroids have been administered in the majority of the reported cases. Overall, in this revision 30 patients (19.8%) received steroids, the majority of which in the classical PMNS group (15 cases) and the ADEM-like group (11 cases). The most frequently applied therapeutic schedule consisted in IV administration of either 1 g/day of methylprednisolone or equivalent dosage of prednisolone administered over 5–7 doses, and subsequent tapering over 4–6 weeks. Alternative dosages included: IV methylprednisolone at 100 mg/day for three doses, and tapered for 10 days; oral prednisolone at 60 mg/day for 4 days without tapering [26, 27]. The first schedule was chosen.
In almost all the reported cases, a rapid improvement of symptoms was observed upon steroids administration. Similarly, the condition of the patient improved rapidly after the first dose of corticosteroid and his neurological impairment fully reverted after 3 weeks. It has been suggested that steroids might hasten the resolution of PMNS through anti-inflammatory effect and immune-suppression, as to turn-off the auto-immune trigger. Notably, steroids have been associated to sensible reduction of serum and CSF concentrations of inflammatory cytokines [24]. Nonetheless, the impact of steroid treatment on outcome and sequelae has not been analysed in any of the current studies. To date, there is no evidence that steroids either ameliorates the prognosis, reduces the sequelae, or affect mortality rates in PMNS. Nevertheless, corticosteroids might be harmful in patients suffering of cerebral malaria, as they might increase the risk of seizures and gastrointestinal haemorrhages [28]. It is not clear whether steroids really changes the course of the disease, as some of these forms might have probably spontaneously revert without treatment.
Some authors promote the use of intravenous immunoglobulins (IV-Ig) in refractory cases as a second line therapy. Indeed, IV-Ig have been demonstrated to exert beneficial effects by inhibiting and reverting the cytoadherence of infected erythrocytes in vitro [29]. Stangel et al. were the first group to report the ability of polyclonal immunoglobulins in vitro to modulate nitric oxygen production and microglial function in vitro. It has been assumed that IV-Ig modulate the local immune response also in CNS [30]. Further modes of action of IV-Ig have been attributed to the ability of hyperimmune Ig clones to bind and neutralize circulating antibodies and activated B cells, as well as to modulate auto-immune response [31, 32]. Marchioni et al. described a case series of 5 patients affected by steroid-resistant post-infectious (non post-malaria) ADEM and myelitis which successfully reverted upon IV-Ig [33]. A two patient-case series by Ravaglia et al. described more extensively the IV-Ig role on specific functional systems and long-term sequelae, also suggesting a potential synergy between IV-Ig and steroids [34]. However, both the aforementioned case series included patients affected by post-infectious neurological complications, but none of them specifically included cases of PMNS (Fig. 1).
Plasmapheresis is another salvage therapy, which could be considered in most severe PMNS, in analogy with post-infectious encephalitis [26]. In this review, plasma exchange was successfully performed in one case of severe AIDP after P. falciparum malaria [35, 36].
In summary, there is lack of consensus regarding PMNS management. There is currently no available data comparing the “watchful waiting strategy”” with the immunomodulatory treatments. The actual efficacy of steroids, IV-Ig, and plasmapheresis is currently under debate.
A course of steroids should be attempted in severe non-self limiting forms (e.g. those with no improvement or progressive worsening in few days after the presentation). Second-line treatment with IV-Ig should be considered for steroid-refractory disease, especially as regards “classical” PMNS. A combination of steroids and IV-Ig is strongly recommended especially in those patients presenting with ADEM-like forms.
To date plasmapheresis has been considered a rescue strategy for progressive non-resolving disease. However, it could be attempted as a first-line treatment for AIDP-like forms.
