May 2026 Research Update

Taken together, these papers highlight two sides of getting the diagnosis right: (1) autoimmune encephalitis can be a common, treatable explanation for certain fast-changing brain syndromes, and (2) several non-autoimmune conditions can look very similar at first. The overall message is that diagnosis often depends on combining symptoms and exam patterns with targeted testing for immune causes, infections, and other medical conditions. [1] [2] [3] [4]

In a prospective Dutch cohort of 147 adults with rapidly progressive dementia, autoimmune encephalitis was the largest diagnostic group, seen in 58 of 147 people (39%). Seizures at first presentation were more common when autoimmune encephalitis was the cause, and some seizures were subtle focal events, which can make recognition harder without careful history-taking and evaluation. [1]

In a prospective study of 85 children with encephalitis in intensive care, spinal fluid anti-NMDAR (N-methyl-D-aspartate receptor) antibodies identified a group with more psychiatric symptoms, seizures, abnormal movements, and speech problems, and a distinctive brainwave pattern was more common in this group. The authors reported high diagnostic performance for the spinal fluid antibody test in separating anti-NMDAR encephalitis from other encephalitis cases in this setting. [2]

Several papers underscored that infections can overlap with, or imitate, autoimmune encephalitis. A structured narrative review of case reports described neurosyphilis presenting like limbic encephalitis and noted that some reported patients also had neuronal surface antibodies, raising the possibility of mixed infection-and-immune pictures in a minority. [3]

A single-patient case report described acute intermittent porphyria presenting with seizures and confusion that initially resembled autoimmune encephalitis, but autoimmune tests were negative and biochemical testing confirmed porphyria. The patient improved after receiving intravenous hemin, showing how a non-autoimmune, non-infectious disorder can still produce an encephalitis-like episode. [4]

Two exploratory biomarker studies looked at situations where standard testing may not fully explain symptoms. In a pilot study of 35 people with atypical, treatment-resistant psychiatric illness, several spinal fluid inflammatory markers differed from control groups, and 6 patients had their diagnoses revised and improved after immunosuppressive therapy, suggesting that a subset may have immune-driven brain inflammation that is not obvious on routine testing. [5]

In a translational study of 387 people suspected of autoimmune encephalitis who were negative for known antibodies, researchers found antibodies against pyruvate dehydrogenase complex (PDHc) in 12 patients and showed these antibodies could reduce PDH activity in lab experiments. This work suggests a possible additional antibody target in some antibody-negative cases, but it remains early-stage and does not yet establish how often it causes disease or how it should be used clinically. [6]

For patients and caregivers, this group of studies mainly provides background on why diagnosis can take time and why results sometimes change as more information comes in. It supports that autoimmune encephalitis can be a common explanation in certain fast-progressing adult presentations and that specific spinal fluid antibody tests can strongly support particular diagnoses in children, while also showing that infections (like syphilis) and metabolic conditions (like porphyria) can closely resemble autoimmune encephalitis. New biomarker work may help explain some antibody-negative or primarily psychiatric presentations in the future, but these tools are still exploratory and not yet proven for routine decision-making.

Across these studies, brain PET scans were used as an extra test when standard studies do not clearly show inflammation, especially when magnetic resonance imaging (MRI) looks normal. In a moderate-sized retrospective study and smaller exploratory reports, PET signals in the brain lined up with how severe symptoms were and sometimes changed alongside improvement, but the best way to use PET (and which tracer to use) is still being worked out. [7] [8] [9]

In 60 people with possible autoimmune encephalitis whose MRI did not show clear abnormalities, one type of PET scan aimed at measuring inflammation showed higher-than-normal uptake in 65% of patients. Higher PET activity at baseline matched worse disability and symptom scores, and on repeat scans PET activity fell in step with clinical improvement; higher baseline activity was also linked to greater later brain volume loss. [7]

A separate feasibility study tested whether shorter PET scanning could still capture inflammation-related signals. In 20 autoimmune encephalitis patients (plus comparison groups), a late, shorter imaging window around 30 to 50 minutes after injection best matched the longer, more complex full scan and separated abnormal areas from healthy tissue. [8]

One case report described a 54-year-old woman with suspected COVID-19-associated autoimmune encephalitis whose MRI and spinal fluid tests were normal, while a different PET scan (measuring brain metabolism) showed low activity in frontal and temporal areas and changes in the cerebellum consistent with an immune-related encephalitis pattern. The patient improved after immunotherapy and was reported to have fully recovered by 3 months, but this is a single-person report and may not apply to others. [9]

For patients and caregivers facing ongoing symptoms but unclear routine testing, these papers suggest that PET scans can sometimes show brain changes even when MRI does not, and that some PET measures may reflect how active the illness is over time. Right now, this mainly helps explain why a clinician might consider PET as an add-on test in select situations; it does not mean PET will be abnormal in everyone, and different PET tracers and scan lengths are still being evaluated.

These papers focused on when brain MRI can provide important clues that point toward a specific cause of cortical autoimmune encephalitis. Overall, they highlight that certain cortical MRI patterns can support recognition of myelin oligodendrocyte glycoprotein (MOG) antibody-associated disease, which may look different from antibody-negative or other antibody-positive forms. The evidence includes one moderate-sized retrospective cohort and small case reports, so results may vary across settings. [10] [11]

In a single-center retrospective cohort of 120 adults with autoimmune cortical encephalitis, MOG-related cortical encephalitis made up 18.3% of cases and was described as commonly seizure-dominant with unilateral (one-sided) cortical MRI lesions. Compared with antibody-negative and non-MOG antibody-positive groups, the MOG group had less disability and fewer recurrences, while peak disability score was the strongest predictor of a poorer prognosis across the cohort. [10]

Two case reports of adults with MOG antibody-associated cerebral cortical encephalitis described diagnostic challenges but a shared MRI pattern: cortical fluid-attenuated inversion recovery (FLAIR) brightness with nearby subcortical darkness. The authors emphasized that noticing this MRI appearance was pivotal to making the correct diagnosis in these individual cases. [11]

For people whose symptoms center on seizures or suspected cortical encephalitis, these studies provide context that MRI pattern recognition can sometimes help narrow down the type of autoimmune condition involved, particularly MOG-related disease. This is most relevant when clinicians are trying to interpret MRI results alongside antibody testing, because MRI appearance and antibody status did not look the same across groups. What remains uncertain is how consistently these MRI patterns apply in broader, multi-hospital populations and how often similar-looking MRI changes occur in other conditions.

Across this group of reports, treatments aimed at lowering harmful antibodies were repeatedly linked to clinical improvement in people whose autoimmune encephalitis was not improving with earlier immune treatments. The evidence is early, mostly from case reports and small observational studies, so it is still uncertain which patients benefit most and how durable benefits are long-term. [12] [13] [14] [15]

In a small retrospective cohort of 15 people given efgartigimod, disability and autoimmune encephalitis symptom scores improved over 4 weeks, and blood immunoglobulin G (IgG) levels dropped after the first infusion and continued to fall during treatment. No serious adverse events were reported in this study. [14]

A three-patient case series used sequential treatment with efgartigimod followed by rituximab, with rapid improvement after efgartigimod and then sustained remission after rituximab consolidation during 13 to 18 months of follow-up. These three people had different antibody types (anti-GAD65, anti-LGI1, and anti-NMDAR), and antibody levels in blood and spinal fluid decreased alongside clinical improvement. [12]

A single-patient report in anti-LGI1 autoimmune encephalitis described improvement in tremor and memory after adding efgartigimod when steroid pulses had not helped, along with a drop in anti-LGI1 antibody levels and less visible brain changes on MRI. No adverse events were reported during 3 months of follow-up. [13]

Another single-patient report described a 54-year-old man with probable autoimmune encephalitis but negative antibody testing who worsened despite high-dose steroids and intravenous immunoglobulin (IVIg), then showed progressive neurological improvement after four weekly injections of telitacicept. No clinically significant adverse events were described. [15]

For patients and caregivers, these papers mainly add background about newer ways researchers are trying to reduce the immune system's antibody activity when standard approaches have not been enough. Because most data here come from small, uncontrolled studies (including single-patient reports), it remains uncertain how often these results would apply to other people, how long benefits last, and what the full safety profile looks like across broader autoimmune encephalitis populations. The strongest take-home message for now is that antibody-lowering strategies are being actively studied across different autoimmune encephalitis situations, including antibody-positive and antibody-negative cases. [12] [13] [14] [15]

Two papers focused on escalation of immune treatment and the challenge of relapse, especially in less common autoimmune encephalitis situations. Together, they suggest that some subtypes can respond well to immune treatment, but relapse may depend on specific clinical features and test results rather than improving being guaranteed. [16] [17]

A systematic review of DPPX encephalitis reported that 49 of 66 patients with treatment data had favorable clinical improvement. In subgroup analyses, a good response to first-line immunotherapy predicted favorable outcome in patients without exaggerated startle (hyperekplexia), but not in those with hyperekplexia. [16]

In the same review, relapse was more likely when spinal fluid (CSF) abnormalities were present and when hyperekplexia was present. The authors also noted that some patients recovered completely with first-line immunotherapy and a short steroid course, while others required long-term immunotherapy. [16]

A single-patient case report described a 17-year-old girl with autoimmune polyendocrine syndrome type 1 (APS-1) who had recurrent autoimmune encephalitis linked to GAD65 antibodies and severe illness requiring intensive care. She improved markedly within 48 hours after starting plasma exchange and high-dose steroids. Rituximab was given later, and seizure control and discharge by hospital day 12 were reported. [17]

For patients, this theme is most relevant as context if someone has a rarer autoimmune encephalitis subtype (such as DPPX encephalitis) or autoimmune encephalitis occurring alongside a broader immune condition (such as APS-1). The review suggests that certain symptoms (exaggerated startle) and spinal fluid changes may matter when thinking about relapse risk in DPPX encephalitis, but it cannot predict what will happen for any one person. The APS-1 case shows that intensive immune treatment escalation may coincide with rapid improvement in a single patient, but one case cannot establish what would happen for others or which part of combination treatment mattered most. [16] [17]

Across these papers, severe seizures in the intensive care unit (status epilepticus) were described as a spectrum where some monitoring patterns may signal a more difficult course, especially when autoimmune encephalitis is the cause. Evidence ranges from one larger retrospective ICU study to small case reports and a small pediatric series, so cause-and-effect and best next steps remain uncertain. Overall, the reports highlight how brainwave monitoring and imaging can reveal important differences between patients, while treatment responses can vary widely. [18] [19] [20]

In a retrospective neuro-ICU study of 138 people with status epilepticus, about 12% had a cyclic seizure pattern on continuous electroencephalogram (EEG) monitoring, meaning short seizures recurring about every 5 to 10 minutes without obvious outward signs. This group was younger on average and more often had an acute cause, with autoimmune encephalitis reported as the most common cause. Patients with cyclic seizures had much higher seizure counts, roughly double the ICU length of stay, worse functional outcomes, and a 30% death rate, suggesting this EEG pattern may mark progression toward very hard-to-stop seizures. [18]

A single case report described a 33-year-old woman with anti-NMDA receptor encephalitis and super-refractory status epilepticus (seizures continuing despite multiple treatments). After many antiseizure and immune therapies, her status epilepticus stopped after adding medium-chain triglyceride (MCT)-enriched nutrition alongside perampanel, and she was later extubated and transferred to rehabilitation. Because this is one patient who also received other immune treatments, it is not clear how much the nutrition change contributed or whether others would respond the same way. [20]

In a small retrospective series of four previously healthy children who had SARS-CoV-2 infection and later developed immune-mediated brain and spinal cord disorders, one of the diagnoses was anti-NMDA receptor encephalitis and another was febrile infection-related epilepsy syndrome (FIRES), which can include severe seizures. Brain MRI patterns differed by disorder, and one child showed imaging changes that progressed from early inflammation to longer-term injury. All received immune-modulating treatments, but outcomes varied, showing that even within immune-related conditions, recovery and lasting effects can differ from child to child. [19]

For patients and caregivers, these studies mainly provide context for why ICU teams often rely on continuous EEG and repeated brain imaging during severe seizures: certain EEG patterns (like cyclic seizures) may signal a higher-risk course, particularly in autoimmune encephalitis, and may help explain longer ICU stays and recovery challenges. The treatment-related evidence here is early: one adult case suggests a possible role for pairing a specific seizure medicine with MCT-enriched nutrition, but it cannot show this will help most people. The pediatric series underscores that immune-related brain disorders after infections can look different on MRI and have variable outcomes, and it remains uncertain which findings best predict long-term recovery in an individual patient. [18] [20] [19]

Across this group, the main pattern was that autoimmune encephalitis sometimes occurred together with an underlying tumor, and improvement was described after treating both the immune attack and the tumor. Several reports highlighted that early brain scans, spinal fluid tests, or antibody panels may be unrevealing, yet a tumor was still present or was found later. The strongest limitation is that most evidence here is from individual case reports, so results may not apply to everyone. [21] [22] [23] [24]

In three reports of anti-NMDA receptor encephalitis in women aged 21 to 33, ovarian teratomas were involved, and in two of them recovery followed a combination of immunotherapy and surgery. One patient had complete recovery within a month after removal of a left ovarian teratoma plus steroids and plasmapheresis, although a catheter-related infection stopped plasmapheresis early. [21] [22] [23]

Another patient with an immature ovarian teratoma had diagnostic challenges because early brain MRI and spinal fluid results were normal, but anti-NMDA receptor antibodies were later found in spinal fluid. After tumor removal and immunotherapy (including rituximab), she had complete recovery and remained seizure-free at six months. [22]

A separate anti-NMDA receptor case was dominated by psychiatric symptoms, improved substantially after plasmapheresis when steroids helped only a little, and a mature ovarian teratoma was detected two years later after initially negative tumor evaluation. This report illustrates that a related tumor may not be identified at the first evaluation in some people. [23]

Tumor links were also described with other antibody types and cancers. A man with thymoma-associated anti-AMPAR encephalitis improved with multiple immune therapies and later developed myasthenia gravis; he then reached sustained remission after thymectomy combined with additional immune treatment, but his course included serious complications such as a myasthenic crisis requiring intensive care. [24]

A separate case did not meet criteria for autoimmune encephalitis: an older woman had recurrent seizure-like disorientation and amnesia with negative antibody testing and normal MRI, spinal fluid, and EEG, yet her symptoms fully resolved after surgical removal of a lung adenocarcinoma and did not return during 10 months of follow-up. The authors suggested this could represent tumor-related brain network dysfunction even without detectable antibodies, while noting that cause-and-effect cannot be proven from one case. [25]

In a larger nationwide Japanese screening study (930 people evaluated for suspected autoimmune brain and balance disorders), anti-KLHL11 antibodies were found in four patients with cerebellar ataxia and/or limbic encephalitis. None had testicular tumors, and responses were mixed, with improvement described in one case with immunotherapy alone and in another after treatment of a non-testicular tumor. [26]

For patients and caregivers, this set of papers mainly reinforces a background point: some autoimmune encephalitis syndromes can be connected to tumors in different organs (ovary, thymus), and some individuals improved after the tumor was treated along with immune therapy. It also shows that standard early tests may look normal in some cases, and that not all antibody-linked syndromes have the same tumor patterns (for example, anti-KLHL11 cases here did not involve testicular tumors). Because most papers are single-patient reports, it remains uncertain how often these patterns occur and which people are most likely to benefit from specific tumor-focused approaches.

Across this group, reports described autoimmune encephalitis happening alongside other autoimmune conditions (especially thyroid disease), after cancer immunotherapy, or with other body-wide complications like low sodium. Most evidence here comes from single-patient case reports, so patterns are suggestive rather than definitive. Overall, the shared message was that autoimmune brain inflammation can sometimes look like infection or dementia, and may involve other organs at the same time. [27] [28] [29] [30] [31] [32] [33]

Several papers focused on thyroid-related encephalopathy, describing subacute confusion and thinking changes with elevated thyroid antibodies and marked improvement after immune therapy. Two adult women improved substantially after steroid-based treatment (one also received intravenous immunoglobulin, and one also had thyroid hormone replacement). [27]

A separate pediatric case connected thyroid autoimmunity and encephalopathy with a rare inherited skin condition (Nagashima-type palmoplantar keratoderma) due to SERPINB7 gene variants, alongside elevated inflammatory signaling proteins in blood. The child improved neurologically and had normalized thyroid tests with seizure medicine and thyroid hormone replacement, while the skin thickening persisted. [28]

Autoimmune encephalitis also overlapped with other autoimmune targets and body complications in individual cases. One older man with anti-LGI1 autoimmune encephalitis also had autoimmune gastritis and presented with hard-to-correct low sodium; after steroid pulse therapy, both sodium and neurologic symptoms improved and remained stable at 8 months. [30]

Some case reports highlighted diagnostic confusion with other neurologic conditions. One woman initially labeled as having a dementia syndrome was later found to have anti-GAD65 autoimmune encephalitis; plasma exchange was followed by remission of psychosis and improved movement symptoms, with partial cognitive improvement. [29]

Immune-triggered and antibody-negative situations were also described. A woman developed a central nervous system demyelinating illness in the antibody-negative NMOSD/MOGAD spectrum after nivolumab and ipilimumab (immune checkpoint inhibitors) and had sustained clinical and scan stability for 30 months after corticosteroids followed by intravenous immunoglobulin and rituximab. [31]

One report emphasized that symptoms can start outside the brain, which can make early recognition harder: one patient developed anti-NMDAR encephalitis after a gastrointestinal illness and improved after steroids and plasma exchange despite weak antibody positivity and infection-like lab signals. [32]

A separate case involved a related but different condition: a man with progressive encephalomyelitis with rigidity and myoclonus (PERM, a rare stiff-person spectrum disorder; no antibodies were found) developed vertebral and pelvic fractures during hospitalization, along with abnormal bone turnover blood markers and delayed healing. After intrathecal baclofen, spasticity improved and bone turnover markers and bone density improved, with signs of fracture healing. [34]

Finally, one report illustrated ongoing uncertainty when antibody tests remain negative and imaging overlaps with other diagnoses. A man evaluated for suspected GABA-A receptor autoimmune encephalitis had extensive negative infectious and autoimmune testing; imaging initially supported a diagnosis of posterior reversible encephalopathy syndrome, but years later he developed mood changes, psychosis, and seizures despite negative repeat autoimmune testing. [33]

For patients and caregivers, these reports mainly serve as background about how autoimmune encephalitis can sometimes occur with other autoimmune diseases (especially thyroid-related), after certain cancer immunotherapies, or with complications affecting sodium levels, digestion, or skin. Because most papers were single cases or very small series, they cannot show how common these overlaps are or predict what will happen for any one person. What remains uncertain is which symptoms reliably signal an autoimmune overlap, how often antibody tests stay negative, and which complications (like fractures in severe stiffness syndromes) are directly caused by immune activity versus illness severity or treatment effects.

Two exploratory lab profiling studies looked for patterns in body fluids that might help explain autoimmune encephalitis biology. Both pointed toward broad immune inflammation signals, with repeated emphasis on IL-6-related signaling, and one also highlighted complement activation and altered cell cleanup pathways. These studies can suggest mechanisms to study, but they do not by themselves diagnose autoimmune encephalitis or show that changing a single pathway will improve symptoms. [35] [36]

In a study of 65 people with autoimmune encephalitis compared with age-matched controls, spinal fluid showed higher counts of immune cells and many differences in proteins and small molecules. The altered molecules clustered in immune-inflammatory pathways as well as lipid and carbohydrate metabolism, and an IL6-STAT3 pathway was highlighted as a possible target for future research. [35]

In LGI1 autoimmune encephalitis, plasma protein profiling showed widespread changes compared with healthy controls, including signals linked to the complement cascade, while autophagy-related pathways were downregulated. Separately, gene-activity analysis in monocytes highlighted IL-6 pathway dysregulation. The study did not find significant protein differences between the acute and chronic phases. [36]

For patients, these studies are mainly early-stage research that helps map what kinds of immune activity may be happening during autoimmune encephalitis. They may support future work on better tests or more targeted therapies, but they do not change current day-to-day care on their own. Key uncertainties include whether these fluid patterns are specific to autoimmune encephalitis (versus other brain inflammation) and whether they track with symptoms, recovery, or relapse risk.