February 2026 Research Update

Across this month's papers, most evidence came from case reports (single patients) plus two retrospective pediatric hospital studies. Overall, immune-targeting treatments (such as steroids, intravenous immunoglobulin, plasma exchange, and rituximab) were repeatedly followed by clinical improvement, but relapses were also described, meaning some people required repeated or maintenance immune therapy. [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

Several case reports described improvement after rituximab, particularly when earlier treatments were not enough or when disease returned. In anti-IgLON5 disease with major complications, the patient did not improve on steroids or intravenous immunoglobulin but improved after rituximab, alongside a drop in antibody levels and better sleep, swallowing, cognition, and seizure control. [1]

Two anti-GAD (glutamic acid decarboxylase) encephalitis case reports described relapse-prone illness where rituximab was used as part of longer-term control. One adult patient had complete resolution after steroids then rituximab, later had a relapse treated again, and had no relapses reported at 18 months while continuing maintenance rituximab. [2]

In a child with three episodes of anti-GAD encephalitis over about 15 months, first-line immune treatments helped initially but relapses occurred with new brain scan changes, and remission was reported after rituximab with no recurrence at last follow-up. This report also showed that antibody levels in blood and spinal fluid changed across relapses, but it is still one patient's course. [4]

Other case reports highlighted first-line immune treatments and how antibody tests may change with recovery. A 15-year-old with both anti-Drebrin and anti-mGluR2 antibodies improved with steroids plus intravenous immunoglobulin, had a relapse that improved with renewed immunotherapy, and later had antibodies become undetectable with good function at 6 months. [3]

Plasma exchange and steroids were also described as helpful in a probable antibody-negative case where brain metabolism imaging supported inflammation. After plasma exchange (six sessions) and high-dose steroids, a 68-year-old woman with catatonia and cognitive decline improved markedly, with cognitive testing rising from 12/30 to 25/30. [8]

For Hashimoto s encephalopathy, one pediatric case with manic and psychotic symptoms improved rapidly with pulse steroids and intravenous immunoglobulin and had no relapse reported after stopping them. Another adult case reported complete recovery without corticosteroids, managed with seizure medicines and thyroid hormone replacement, suggesting that spontaneous remission may occur in rare situations. [9] [10]

Two retrospective pediatric studies added broader context about hospital outcomes. In a 50-patient pediatric group receiving therapeutic plasma exchange for antibody-mediated brain and nerve inflammatory diseases, 46% recovered their prior functional status by discharge and 56% still had neurologic symptoms, with 6% mortality; steroid use was linked to improved survival in certain conditions including autoimmune encephalitis. [5]

In a 36-child pediatric intensive care study of new-onset super-refractory status epilepticus (NORSE), seizure control occurred in 80.6% and 72.2% survived to discharge, with functional improvement by 6 months. Earlier immunotherapy was linked to better survival in this setting, though only 5 cases were labeled autoimmune encephalitis in that cohort. [6]

A newer immune approach was described in a single older patient who could not tolerate steroids. In an 83-year-old with Hashimoto s encephalopathy, the FcRn antagonist efgartigimod-alpha led to noticeable symptom improvement within 3 days, but benefits lasted about one month before relapse, and repeated doses were needed before transitioning to azathioprine. [7]

For patients and caregivers, this month's research mainly reinforces a practical point: many reported care approaches aim to lower harmful immune activity (through steroids, intravenous immunoglobulin, plasma exchange, or longer-acting options like rituximab), and symptom improvement is often described, but relapses can still occur and may require repeated treatment. Because most papers were single-patient reports, it remains uncertain how often these responses happen, which subtypes benefit most, and how to balance benefits with side effects in different age groups and medical situations. The two pediatric hospital studies provide broader outcome snapshots during severe illness, but they do not prove which treatment caused recovery. [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

This group of papers suggests that no single test or scan pattern is perfect for diagnosing autoimmune encephalitis. Standard brain MRI sometimes showed helpful patterns, advanced imaging methods measured broader brain network changes, and spinal fluid or blood markers were explored as additional clues, including for estimating severity. Several results overlapped with other neurologic conditions, and multiple studies emphasized the need for larger or external validation. [11] [12] [13] [14] [15] [16] [17] [18]

In a moderate-sized case-control study of 115 people with definite anti-NMDA receptor encephalitis, just over half had MRI changes, most often in limbic and nearby brain regions, and about one quarter had thick lining (pachymeningeal) enhancement. These MRI patterns helped distinguish anti-NMDA receptor encephalitis from primary psychotic disorders, but they overlapped with antibody-negative autoimmune encephalitis, showing why imaging alone may not fully separate causes. [11]

Two exploratory imaging studies highlighted the back (occipital) part of the brain as a potentially important area in autoimmune encephalitis. One small adult study (16 people) reported that lower activity in the occipital lobe on a brain metabolism scan at baseline tracked with greater severity at diagnosis and worse function at 1 year. Another study using resting-state functional MRI and structural brain connections found stage-dependent disruptions in brain network efficiency, strongest in the acute phase, and machine-learning models separated patients from healthy controls and acute from recovery phases with high accuracy. [13] [12]

Several papers looked at tests beyond scans. In 114 antibody-positive patients, a combined blood measure that reflects inflammation and albumin levels (C-reactive protein-to-albumin ratio) was higher in people needing intensive care and correlated with initial severity and disability at discharge, performing better than either blood test alone for predicting intensive care need in that dataset. In a small pediatric study, soluble TREM2 (a marker linked to immune activity in the brain) was higher in spinal fluid and blood in children with anti-NMDA receptor encephalitis than in non-inflammatory controls, and higher spinal fluid levels tracked with worse severity and were lower in those with better outcomes. [14] [15]

Individual case reports reinforced how variable testing can look in real life. One person with anti-NMDA receptor antibodies had an MRI pattern (lentiform fork sign) often discussed in other conditions, despite normal kidney function, showing that unusual MRI appearances can still occur in autoimmune encephalitis. Another person with LGI1 antibody-positive encephalitis had electrographic seizures detected on EEG without obvious seizure events, suggesting that EEG patterns may provide early clues even when spinal fluid is normal. [16] [17]

A commentary on a suspected post-infection case stressed that incomplete testing can leave major uncertainty, including when MRI is done without contrast and when autoimmune encephalitis antibody tests are not reported. The authors argued for thorough evaluation for both infectious and autoimmune causes, highlighting how diagnosis can remain unclear when key data are missing. [18]

For patients and caregivers, these studies support a general message: diagnosis and severity assessment in autoimmune encephalitis often rely on combining information from brain scans, EEG, and spinal fluid or blood tests, because each tool can be abnormal in some people and normal or overlapping in others. Some results (like certain MRI patterns, occipital changes on advanced imaging, or newer lab markers) may be most useful as background research for now, since several studies were small, retrospective, or single-center and need confirmation before they can be relied on broadly. Uncertainty remains about how well these newer measures perform across different antibody types, ages, and hospitals, and how they compare head-to-head with standard testing in everyday care. [11] [12] [13] [14] [15] [16] [17] [18]

Across this group of studies, relapse and long-term effects after autoimmune encephalitis were influenced by both the type of immune response (which antibodies are present) and how severe the first illness was. Evidence ranged from a meta-analysis of observational studies to moderate-sized cohorts, plus one small cognitive follow-up study, so results describe patterns rather than guarantees for any one person. [19] [20] [21] [22] [23]

A systematic review and meta-analysis pooling 39 observational studies estimated an overall recurrence rate of 0.162, with anti-N-methyl-D-aspartate receptor (anti-NMDAR) encephalitis recurrence estimated at 0.148. In those studies, relapse was lower in people who received second-line immune treatment and in people with teratoma removal when a teratoma was present, and relapse did not rise further after the first year. The same review reported higher relapse risk in anti-LGI1 encephalitis and with older age, and it reported a relationship between relapse risk and shorter time to treatment in the included multivariable analyses (the direction reported in the paper was higher risk with shorter delay). [19]

In a multicenter retrospective cohort of 218 adults in China, having systemic or organ-specific autoantibodies (autoantibodies beyond the brain-targeting antibodies) was common and tracked with a harder initial hospital course and more relapses over 24 months. People who were positive for these additional autoantibodies had higher intensive care unit admission rates and worse function at discharge, and they had a higher 24-month relapse rate (29.8% vs 13.2%). After adjustment for other factors, this autoantibody pattern still predicted relapse risk (adjusted hazard ratio 2.270). [20]

Longer-term seizures were a key outcome in a prospective Austrian cohort that followed 149 adults with autoimmune or infectious encephalitis. Post-encephalitic epilepsy occurred in 12% overall and was more common after autoimmune encephalitis than after infectious encephalitis, and needing more seizure medicines during the initial hospitalization correlated with later epilepsy risk. The authors also created an Excel-based risk estimation tool, but they noted it still needs validation in other groups. [21]

In a single-center pediatric cohort of 84 children with teratoma-associated anti-NMDAR encephalitis, certain early features signaled a higher chance of poor function at 12 months. Independent predictors included central hypoventilation, having symptoms across 4 or more categories, multifocal brain MRI lesions, an extreme delta-brush pattern on electroencephalogram (EEG), and tumor removal more than 21 days after symptom onset; about 23.81% had poor outcomes at 12 months. Relapse occurred in 9.52% and death in 3.57% in this cohort. [22]

Cognition over time was addressed in a small retrospective study of 18 people with LGI1 antibody autoimmune encephalitis treated with immunotherapy. Average thinking scores improved (Montreal Cognitive Assessment increased from 20 to 24), but executive skills and delayed recall, while improved, could still be affected, especially in older patients. Early cognitive testing helped predict longer-term cognitive outcomes in this group. [23]

For patients and caregivers, these studies together suggest that follow-up often focuses on two separate but related concerns: whether symptoms return (relapse) and what longer-term effects remain (such as ongoing seizures or thinking and memory changes). The research also suggests that blood and spinal fluid antibody patterns, plus how severe the first hospitalization was, can be linked to higher or lower risks in groups of patients, but the studies cannot predict exactly what will happen to an individual. Some results (like the role of second-line immune treatment, teratoma removal, and additional non-brain autoantibodies) are relevant now as background for understanding relapse patterns, while tools that estimate an individual risk (like the epilepsy risk spreadsheet) still need testing in other hospitals and populations. [19] [20] [21] [22] [23]

Across these reports, autoimmune encephalitis showed up in several cancer-related situations: as a paraneoplastic problem (where the immune system reacts to a tumor and affects the brain), after immune checkpoint therapy, and alongside ovarian teratomas. Many patients improved after immune-focused treatments (such as steroids or intravenous immunoglobulin) and/or after treating or removing the tumor, but the evidence is limited because all papers describe single patients. [24] [25] [26] [27] [28] [29]

In small cell lung cancer, autoimmune limbic encephalitis was reported as an initial presentation, including a case linked to anti-AMPAR antibodies. This single case highlights that, in some people, brain inflammation symptoms can appear before a lung cancer diagnosis is recognized. [24]

Two reports described encephalitis developing after immune checkpoint therapy given for cancer: one after tislelizumab plus chemotherapy in small cell lung cancer with AMPA2 antibodies detected, and one after pembrolizumab plus chemotherapy in breast cancer with negative antibody testing. In both cases, the patients had substantial neurologic decline that improved after immune treatment with high-dose steroids, and in the AMPA2 case the antibody levels in blood and spinal fluid fell over 3 months. [25] [26]

A separate small cell lung cancer report described a person with subacute psychiatric and seizure symptoms suggestive of paraneoplastic encephalitis, but with negative testing for multiple common neuronal antibodies. Immune treatments gave only partial and temporary improvement, while chemotherapy after tissue confirmation of cancer was followed by improvement in both neurologic/psychiatric symptoms and signs of impending superior vena cava obstruction, with near-complete remission at 3 months. [27]

Two reports involved ovarian teratoma and autoimmune encephalitis in adolescents and young adults, including cases with negative common antibodies. In one 15-year-old girl, anti-NCAM1 antibodies were found in blood and spinal fluid, the teratoma tissue showed NCAM1 expression, and symptoms improved after immune treatment plus tumor removal; antibodies later disappeared, although a relapse was reported and improved with additional immune treatment. In an 18-year-old woman, early ovariectomy plus steroids and high-dose intravenous immunoglobulin was followed by clinical and MRI improvement despite negative antibody tests. [28] [29]

For patients and caregivers, these reports mainly provide background on how autoimmune encephalitis can intersect with cancer care in different ways: as an immune reaction related to an underlying tumor, as a side effect after immune checkpoint therapy, or in association with ovarian teratoma (sometimes with psychiatric symptoms). Because the evidence here is limited to single-patient case reports, it is not possible to estimate how often these situations occur or which tests or treatments work best in general, and it remains uncertain how to separate paraneoplastic encephalitis from immunotherapy-triggered encephalitis in some cases.

This set of studies mainly highlights that autoimmune encephalitis can show up with less common antibodies and may involve organs beyond the brain, such as the eyes, heart, and muscles. Most evidence here is exploratory (small observational work and case reports), but the overall pattern was that symptoms and scans often improved after immune-based treatments, while antibody levels sometimes fell, stayed detectable, or varied over time. [3] [30] [31] [32] [33] [34] [35]

Two single-patient reports described uncommon antibody patterns in autoimmune encephalitis and tracked symptoms, brain MRI changes, and antibody levels over time. In a 15-year-old boy with both anti-Drebrin and anti-mGluR2 antibodies, symptoms and MRI changes improved after steroid pulses and intravenous immunoglobulin (IVIG), and the antibodies later became undetectable, although a relapse occurred before improvement again. [3]

Another single-patient report linked limbic autoimmune encephalitis to anti-recoverin antibodies in a 33-year-old woman with seizures and cognitive problems, with clinical and MRI improvement over 2 years after steroids and IVIG. Her anti-recoverin antibody level stayed positive but decreased over time, showing that antibody tests may not always turn negative even when symptoms improve. [30]

One case report emphasized that autoimmune encephalitis can present with severe seizures even when early brain imaging looks normal. A 77-year-old man with CASPR2 antibodies improved after early treatment with methylprednisolone and IVIG, and antiseizure medicines could be reduced as seizures came under control. [31]

In a small retrospective study of 14 children with autoimmune encephalitis tied to antibodies against intracellular targets, seizures were the most common first symptom, and MRI was sometimes normal. After immunotherapy, antibody levels in blood and spinal fluid usually decreased or became negative, and most children recovered well, although some left the hospital with ongoing difficulties; one antibody type (Ri) often did not fully turn negative. [32]

Two reports focused on broader, "whole-body" immune involvement that included the brain. A 41-year-old woman with thymoma-associated myasthenia gravis had simultaneous inflammatory muscle disease, heart inflammation (myocarditis), and autoimmune encephalitis with anti-GAD65 antibodies; she improved after intensified immunotherapy and thymus removal, with heart swelling on cardiac MRI resolving while some brain MRI changes persisted. [33]

A separate single-patient report described a 17-year-old girl with systemic lupus erythematosus (SLE) who developed neuropsychiatric symptoms during apparent body-wide remission, with temporary worsening and then resolution of deep brain MRI lesions after treatment. This illustrates that lupus-related brain inflammation can fluctuate on imaging and may not match how active lupus looks elsewhere in the body. [34]

Finally, a small cross-sectional study explored whether anti-IgLON5 encephalitis could involve the retina. All six patients had a specific staining pattern in retinal tissue testing that disappeared after removing anti-IgLON5 antibodies, but detailed eye testing did not show a consistent pattern of retinal dysfunction, suggesting possible retinal binding without clear, measurable retinal damage in this group. [35]

For patients and caregivers, this group of papers is most useful as background on how varied autoimmune encephalitis can be: different antibodies may be involved, symptoms can include seizures and changes in thinking or behavior, and some people may also have eye findings or inflammation in organs like the heart or muscles. Because most data here come from small studies and single-patient reports, it remains uncertain how often these rare antibodies occur, how strongly they predict a specific course, and how well results generalize to others.

Across patient and caregiver accounts, autoimmune encephalitis was described as a condition with impacts that can continue well after the acute illness. The overall pattern included delays and uneven access to knowledgeable care, persistent symptoms that affected daily functioning, challenges rebuilding work and social life, and substantial ongoing caregiver burden. [36]

In this small qualitative study (10 people with prior autoimmune encephalitis and their carers), people described delayed diagnosis and difficulty accessing quality medical care, with gaps in clinician knowledge and communication breakdowns. Participants also described differences between rural services and tertiary centers, which sometimes shaped how supported they felt. [36]

Many described chronic symptoms that continued to limit day-to-day life, especially fatigue and cognitive difficulties such as memory and attention problems. Some also reported treatment-resistant seizures after encephalitis, contributing to reduced independence and social isolation. [36]

People and carers described reintegration as difficult, including returning to previous routines, work, and relationships. Caregivers reported significant strain from ongoing practical care needs and changes in family roles and relationships. [36]

This research mainly helps document and name common challenges that some people and families face after autoimmune encephalitis, including long-lasting fatigue, thinking changes, and caregiver stress, as well as barriers in care pathways. Because it was a small, self-reported study from one country, it cannot show how often these experiences happen or predict an individual persons recovery, but it adds patient-centered detail that may inform future support services and clinician education. [36]