Clinical and experimental studies of potentially pathogenic brain-directed autoantibodies: current knowledge and future directions

Abstract

The field of neuronal surface-directed antibody-mediated diseases of the central nervous system has dramatically expanded in the last few years and now forms an important cluster of treatable neurological conditions. In this review, we focus on three areas. First, we review the demographics, clinical features and treatment responses of these conditions. Second, we consider their pathophysiology and compare autoantibody mechanisms and their effects to genetic or pharmacological disruptions of the target antigens. Third, we discuss areas of controversy within the field, propose possible resolutions, and explore new directions for neuronal surface antibody-mediated diseases.

Fig. 1

a Trends in NMDAR-antibody encephalitis. Demographics of published cases (series containing >3 patients) with NMDAR (N-methyl-d-aspartate receptor)-antibody encephalitis. Note the slightly decreasing median age (black line) and increasing male and falling female representation (green and red, respectively). Tumour frequencies (blue line) are falling, mainly due to the recent publications of many paediatric cases. Figure adapted from Irani et al. [31]. b The effect of immunotherapy on mortality, the percentage with a good recovery (modified rankin score 0–2) and relapse-free recovery at 24 months. Data derived from Titulaer et al. [5]. c Key features of a representative patient with faciobrachial dystonic seizures (FBDS). Note the increasing seizure frequency (red line), poor response to anti-epileptic drugs (AEDs), time of onset of cognitive impairment (quantified by fall in Addenbrooke’s cognitive examination-Revised score (ACE-R, green line)) and of hyponatraemia (orange line). IT results in dramatic improvement in all features. LGI1 leucine-rich glioma-inactivated, VGKC voltage-gated potassium channel—complex antibody titres are shown in purple and black, respectively

Fig. 2

The contrasting probable natural histories of three antibody-related encephalitidies. Key things to note are the relapsing course NMDAR (N-methyl-d-aspartate receptor)-antibody encephalitis, often with a good long-term outcome. The LGI1 (leucine-rich glioma-inactivated 1) or CASPR2 (contactin-associated protein 2)-associated encephalitis has a tendency to be more monophasic often with residual memory and functional deficits. GAD (glutamic acid decarboxylase)-antibody-associated LE has an insidious onset and tends to adopt a more chronic course with ongoing seizures and memory deficits

Fig. 3

Illustration of the VGKC-complexes: the association of Kv1s and CASPR2 (contactin-associated protein), LGI1 (leucine-rich glioma-inactivated and other components of the complexes. a Neuronal subcellular domains including the axon initial segment, presynaptic terminal, node of Ranvier (NoR), paranode (PN), juxtaparanode (JXTPN) and internode (IN). b Juxtaparanode: Kv1 channels (blue, alpha subunits = rectangle, beta subunit = circle), CASPR2 (pink oval), contactin-2 (black diamond), MAGUKs (membrane-associated guanylate-kinases) (semicircles), protein 4.1B/spectrins/ankyrins (green/blue triangles). c Synaptic Kv1 organisation. Kv1 s (blue, such as Kv1.1), LGI1 (red) and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors (AMPAR) and ADAM22/23 (a disintegrin and metalloproteinase 22/23) (brown) anchored at post-synaptic membranes

Fig. 4

Potential pathogenic mechanisms of neuronal surface-directed antibodies (NSAbs). a Internalisation of receptors has been demonstrated in vitro using NMDAR (N-methyl-d-aspartate receptor), AMPAR (a-amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptor) and GABA_AR (γ-aminobutyric acid A receptor)-antibodies. Here the LGI1–ADAM22 interaction is shown as a possible unit for co-internalisation. b Antibody-mediated complement fixation and complement-mediated membrane receptor disruption as seen with antibodies against AQP4 (aquaporin-4). c Direct alteration of ion-channel molecular function is an alternative mechanism