February 2026 Research Update

This month added several real-world signals about treatments used when tuberous sclerosis complex (TSC) causes urgent or hard-to-control problems, especially epilepsy and heart tumors. Across these reports and small cohorts, mTOR inhibitors (sirolimus or everolimus) were linked with shrinkage of certain TSC-related tumors, while an add-on anti-seizure medicine (cenobamate) was linked with better seizure control for many participants and improved day-to-day measures like quality of life. The evidence is still early because much of it comes from single-center cohorts and individual patients, but it gives a clearer picture of what researchers are trying in higher-need situations. [1] [2] [3] [4]

For drug-resistant epilepsy in young adults with TSC, a 12-month observational study reported that adding cenobamate was followed by fewer seizures for many participants, along with improved executive function scores, better quality of life scores, and lower parent stress scores. Overall IQ and adaptive skills scores stayed about the same over the year, which fits with the idea that changes were more noticeable in daily functioning than in broad cognition in this group. Side effects were mostly mild and happened early in treatment. [1]

For fetal or infant heart rhabdomyomas (heart tumors often seen in TSC), one retrospective pregnancy cohort suggested that prenatal sirolimus given for more than 7 days was linked with tumor shrinkage and improvement in some heart complications, while shorter courses did not show the same pattern. The same report noted rebound tumor growth when sirolimus was stopped before birth, which makes timing and duration an important open question in research. A separate preterm-infant case also reported rapid tumor regression with sirolimus when surgery was not an option, although treatment was interrupted by sepsis during the newborn course. [2] [3]

Several individual reports reinforced that everolimus (an mTOR inhibitor) can be paired with a tumor's genetic findings or a tumor-related complication. One man with no personal or family history of TSC, whose metastatic epithelioid angiomyolipoma (a rare kidney-related tumor type) carried two TSC2 changes found only in the tumor, had marked shrinking of metastases on everolimus for 24 months before later progression. Another person with cardiac rhabdomyomas causing atrioventricular block had tumor shrinkage and the heart rhythm problem resolved after everolimus, with no regrowth reported after stopping. [5] [4]

Some studies focused on procedures rather than medications, showing how teams try to improve seizure surgery or vascular surgery decision-making in complex settings. A small case series described using intraoperative ultrasound plus depth-electrode recordings to guide epilepsy lesion removal, with most patients seizure-free at follow-up in this selected group that included one person with a TSC-related lesion. A separate infant case report of aortic aneurysm repair highlighted how clotting tendency (seen on a clotting test) and heart imaging during surgery were used to guide management when complications developed. [6] [7]

For patients and caregivers, the main update is that February's studies added more examples of how targeted mTOR inhibitors and newer anti-seizure approaches are being used in higher-need situations in TSC, with some encouraging outcomes reported. This does not prove what will happen for most people, because many findings came from small groups or single patients, and treatment duration and side effects varied. Still, these reports help clarify which TSC problems researchers are actively trying to treat (drug-resistant seizures and fetal/infant heart tumors) and what kinds of improvements have been observed. The biggest remaining uncertainties are which patients benefit most, how long treatment needs to be continued to avoid rebound growth, and how results hold up in larger and more diverse groups. [1] [2] [3] [4] [6] [7]

This month, multiple studies focused on detecting seizures earlier or predicting seizure-related outcomes in TSC, using long EEG monitoring and computer analysis of brain imaging. Taken together, they point in the same direction: seizures or seizure-driving brain areas may be missed without specialized monitoring and careful review, and prediction tools are being built to support decision-making. These tools showed strong performance inside the tested groups, but they still need broader testing to know how well they work across different hospitals and patient populations. [8] [9] [10]

In neonates diagnosed around birth, long-term video-EEG monitoring found that most had seizures that showed up on EEG but were not obvious clinically, and these happened within the first 10 days of life in this small series. All babies in this report had TSC2 variants, and most had difficult-to-treat seizures requiring multiple anti-seizure medicines. This adds weight to the idea that the earliest seizures in TSC can be "silent" without EEG monitoring. [8]

For people with intractable epilepsy undergoing epilepsy surgery, a multicenter machine-learning study combined clinical features with imaging patterns from CT, MRI, and PET to identify which brain areas were likely driving seizures and to predict long-term surgical outcomes. The best combined model performed well across several cohorts, and the authors also reported that a web-based version improved accuracy for less experienced clinicians in their testing setup. This work suggests that combining multiple types of information may be more useful than relying on a single scan or single data source. [9]

Another imaging-based machine-learning study in 95 children with TSC-related epilepsy used two common MRI sequences to predict whether seizures would be controlled after 1 year of anti-seizure medicines. The dual-sequence model outperformed a baseline model within this dataset, supporting the idea that MRI patterns may contain clues about future seizure control. The study did not report external validation, so it is not yet clear how well the model would generalize outside the original center. [10]

For families, these studies mainly add understanding rather than changing anything immediately, because the prediction tools are still research and not universal parts of care. The clearest near-term message is that seizures in very young babies with TSC can sometimes be detected only on EEG, which helps explain why some centers use prolonged monitoring in high-risk infants. For older children and adults, the imaging-based tools are part of a bigger push to better locate seizure-driving brain areas and to better anticipate long-term outcomes after surgery or medication, but these tools still need wider testing before they can be considered reliable across settings. What remains uncertain is how these approaches perform in more varied real-world populations, including those with different imaging protocols and different types of epilepsy severity. [8] [9] [10]

This month reinforced how widely TSC can vary, from common features like epilepsy and skin findings to less common but important heart, kidney, lung, and vascular issues. Larger observational work helped quantify what shows up most often in a defined region, while a separate hospital-based series highlighted that many childhood heart tumors linked to TSC shrink over time. Case reports added examples of rare presentations and severe multi-organ disease, which do not describe what most people experience but help map the outer edges of what can happen. [11] [12] [13] [14] [15]

A regional survey in Japan collected information on 232 people with TSC and estimated a prevalence of 3.1 per 100,000 in that area. It reported that cortical tubers and epilepsy were very common, and that autism spectrum disorder and subependymal giant cell astrocytoma were also frequent. The survey also highlighted differences by facility type, including missing brain MRI data in institutions for handicapped children, which can affect how completely features are documented. [11]

A 20-year single-center series of 33 children with intracardiac masses found rhabdomyomas were the most common and were all linked to TSC in this cohort. Many rhabdomyomas regressed spontaneously during follow-up, and surgery was uncommon. This supports the idea that in many children, these tumors can lessen over time, although the study also reported that outcomes depend on the overall clinical situation and management choices. [12]

Kidney involvement was highlighted through case-based reports that emphasized genetics and the range of severity. A 13-year-old with a combined TSC2/PKD1 deletion (PKDTS) had a ruptured epithelioid angiomyolipoma that was treated with partial nephrectomy with stable kidney function at 24 months. A multigenerational family with a specific TSC2 nonsense variant was described as having a severe kidney pattern with large angiomyolipomas, cystic disease, and progression to kidney failure in some members, with variable outcomes reported among those who used everolimus. [13] [14]

A single case illustrated severe multi-system disease that included pulmonary lymphangioleiomyomatosis (LAM) and kidney angiomyolipomas with bleeding, with death reported about 6 months after discharge after the patient declined mTOR inhibitor therapy. Because this is one person with fragmented prior history, it cannot show how often this course occurs, but it does underline that some people can have high-burden disease across several organs. [15]

For patients and caregivers, these studies help put individual experiences into context: many TSC features are common across populations, but the mix and severity can differ widely from person to person. The survey and the heart-mass series may help families understand which problems are seen often (like epilepsy) and which sometimes improve with time (like many rhabdomyomas), while also showing that incomplete testing can leave gaps in a person's documented history. The kidney and multi-organ case reports mainly matter as awareness-building research, showing that certain genetic patterns can come with heavier kidney disease and that severe multi-organ involvement can occur. What remains uncertain is how strongly specific gene changes predict outcomes for an individual, because the strongest kidney-genetics examples here came from a single family or single cases. [11] [12] [13] [14] [15]

A major research focus this month was improving how we interpret changes in the TSC2 gene, especially variants of uncertain significance (VUS). The main pattern across these studies is that functional laboratory tests can sort many unclear variants into more meaningful buckets, and single-patient deep-dive work can reveal possible mechanisms behind severe symptoms. These approaches are not the same as a clinical diagnosis by themselves, but they aim to reduce uncertainty when genetic test results are hard to interpret. [16] [17]

One high-throughput laboratory study developed two functional tests for TSC2 missense variants: one measuring how much TSC2 protein is made, and another measuring how strongly mTOR signaling is turned on (using a pS6 readout). When the authors applied these tests, they reported reclassifying many previously uncertain variants in a clinical cohort, mostly toward likely benign, with a smaller number moving toward pathogenic or likely pathogenic. This helps explain how future genetic reports might include stronger evidence about whether a specific TSC2 change is likely to matter. [16]

A separate single-child report paired a de novo TSC2 frameshift variant with laboratory experiments showing that the altered TSC2 protein was less stable and broke down faster, while also interacting with a different set of cell proteins than typical TSC2. The authors highlighted pathways related to RNA handling and mitochondrial quality control as possible areas affected by this variant. While this does not prove cause-and-effect in patients, it adds a concrete example of how a particular TSC2 change could disrupt cell regulation beyond the usual mTOR pathway framing. [17]

For families dealing with genetic testing results, these studies mainly matter for the future because they aim to reduce the frustrating category of "uncertain" TSC2 findings. The high-throughput assays suggest that laboratory-based evidence could help clarify whether a TSC2 variant is more likely harmless or more likely disease-related, but these methods are not yet described here as routine clinical tools everywhere. The single-patient functional report shows how researchers can connect a specific variant to measurable changes in cells, which may eventually help explain why symptoms differ across people. What remains uncertain is how well these lab results predict a person's clinical course, since the assays use limited cell types and do not capture every kind of gene effect. [16] [17]

February's biology studies added several pieces to the bigger picture of how TSC-related pathways may shape seizures, sleep, and immune activity in the brain. Across mouse models and human tissue or cell models, a common thread was altered regulation of mTOR signaling and related cell stress pathways, but the immediate patient impact remains indirect because these are not treatment trials. Still, they offer possible explanations for why TSC can involve seizures and sleep disruption. [18] [19]

A mouse study of TSC-related epilepsy reported that seizures were linked with disrupted sleep, especially reduced REM sleep, and that orexin (a wakefulness-related brain signal) was higher in mice with seizures. When the orexin blocker suvorexant was tested in this model, it improved the REM sleep reduction but did not significantly change seizures. This separates two experiences that often overlap for patients, suggesting that sleep disruption and seizure frequency may not always move together in the same way. [18]

A translational microglia study using human TSC brain tissue and patient-derived microglia-like cells reported a pattern of calcium signaling disruption, higher mitochondrial activity, and increased phagocytosis, along with gene-expression signatures consistent with a hyperresponsive microglial state. In plain terms, this suggests some immune-like brain cells may be more easily activated in TSC, which could matter for how brain networks change over time. Because these were lab and tissue studies, it is not yet clear how these signals map onto specific symptoms in an individual person. [19]

A separate study, in autism mouse models without TSC, connected the TSC2 protein to autism-related behavior. It reported that nitric oxide-related chemical tagging of TSC2 (S-nitrosylation) led to lower TSC2 levels and higher mTOR activity in autism models and in human ASD samples, and that preventing this process improved ASD-like behaviors in mice. [20]

For patients and caregivers, these studies mainly add background understanding rather than near-term changes in care. They help clarify that TSC-related biology can influence sleep, seizures, and brain immune activity through specific signaling pathways, which may partly explain why symptoms can cluster and why they can be hard to predict. The sleep-seizure mouse study is a reminder that improving sleep-related brain signals might not automatically reduce seizures, at least in that model. The biggest uncertainty is how directly these mouse and lab findings translate to people with TSC, and which of these pathways will turn into practical tests or treatments in future research. [18] [19] [20]

This month included several case reports that do not define typical TSC, but illustrate unusual events, rare vascular problems, and conditions that can look like TSC on imaging. As a group, these cases emphasize that TSC can sometimes involve vascular and bleeding risks, and that tissue confirmation and careful follow-up can be important when imaging patterns are confusing. These reports are best read as "what can happen" examples rather than predictions for others. [21] [22] [23]

A pediatric case described a pulmonary artery trunk aneurysm that enlarged over time with rising pulmonary artery pressure, leading to surgical repair at age 10. Another case described a teenager with TSC who developed a small epidural hematoma after minor trauma that rapidly expanded within hours, followed by a seizure and emergency surgery, with later normal EEG and MRI reported. Together, these cases point to vascular and bleeding-related complications that can occur in TSC, even though they are not common enough to be quantified from these reports. [21] [22]

One report described multifocal micronodular pneumocyte hyperplasia (MMPH), a lung condition that can be associated with TSC, in a person whose genetic testing was negative for TSC1 and TSC2. The imaging showed many ground-glass nodules, including some near the pleura that could mimic other lung disease patterns, and biopsy confirmed MMPH with stability over 1.5 years. This case highlights that some imaging appearances overlap across diseases and that biopsy can sometimes be needed to sort them out. [23]

For patients, these cases mainly serve as awareness-building research about uncommon situations rather than new expectations. They help clarify that TSC can rarely involve significant blood vessel changes or unusual bleeding courses after injury, and they show that some lung findings that resemble TSC-related changes can also appear without detectable TSC gene variants. The uncertainty is high because each report describes only 1 person, and we cannot estimate how likely these events are from case reports alone. What these cases add is a clearer map of diagnostic and monitoring challenges that can come up in individual situations. [21] [22] [23]