May 2026 Research Update

Together, these papers highlight two practical challenges in tuberous sclerosis complex (TSC): seizures may be missed without brain-wave testing and seizure patterns can change over time, which can complicate planning. In parallel, evidence from randomized trials across several epilepsy syndromes (including TSC) supports cannabidiol (CBD) as an add-on medicine that can reduce seizure frequency on average, although effects varied by syndrome and by how CBD was delivered. [1] [2] [3]

In neonates with definite TSC and seizures confirmed on electroencephalogram (EEG), more than half had seizures seen on EEG without obvious outward signs, and about a third had neonatal status epilepticus (a prolonged or repeated seizure state). Seizures could begin very early, including before full-term age in preterm infants, and most had interictal EEG abnormalities (abnormalities between seizures). [1]

In 45 children with TSC being evaluated with presurgical EEGs, seizure-onset locations on EEG were consistent over time in about one quarter, while most children showed changes or multiple seizure-onset areas as time went on. Abnormal EEG signals between seizures often did not match where seizures started, with discordance reported in most patients. [2]

A systematic review and meta-analysis of seven randomized, placebo-controlled trials (1154 participants across several drug-resistant epilepsy groups, including TSC) found that add-on CBD reduced seizure frequency overall. The strongest effect in the combined trial evidence was reported with highly purified oral CBD at 20 mg/kg/day, while liposomal CBD showed some benefit and transdermal CBD showed limited short-term effect. Side effects were generally described as manageable, but sleepiness and higher liver enzymes occurred more often in some people, especially when CBD was taken with certain other seizure medicines. [3]

For patients and caregivers, these studies mainly reinforce that seizure counting based only on what is seen may miss events in some situations (especially in newborns), and that seizure patterns on EEG can be variable over time in children with TSC, which is relevant background for discussions about monitoring and, when considered, surgical evaluation. The CBD results summarize trial-level evidence that CBD can lower seizure frequency on average as an add-on therapy, but they also show that effects depend on the specific epilepsy group and the formulation, and that some side effects (including liver enzyme changes and sleepiness) remain important uncertainties for individual patients.

This month, the papers focused on how brain development and behavior can unfold very differently across people with tuberous sclerosis complex (TSC), and how researchers are trying to measure these changes over time. One small, exploratory study suggested that a combination of a brain white-matter scan and early behavior assessments around age 2 years could predict an autism diagnosis at age 3 years. [4]

In an exploratory prediction study, using diffusion-based brain scans at 24 months together with standardized autism-related behavior measures at 24 months performed about as well as, or slightly better than, including earlier 12-month scan data when predicting autism spectrum disorder diagnosis at 36 months. The authors described the work as exploratory, and the sample was small and clinically varied, which limits how widely the results may apply right now. [4]

Two case reports illustrated very different individual paths. One child treated in infancy for infantile spasms and a type of brain tumor linked to TSC had cognitive, language, and motor scores in the average range through 36 months, with early low scores in daily living and overall adaptive skills that were not seen at later follow-ups. [5]

Another case report described a 17-year-old who developed hard-to-treat psychosis after surgery to remove a seizure-related brain lesion, and whose symptoms resolved after a course of electroconvulsive therapy; no neurological complications were reported in this single patient. Because this is one person, it cannot tell how often this outcome occurs or who might respond similarly. [6]

A separate publication laid out plans for a large, long-term study to map trajectories of TSC-associated neuropsychiatric difficulties, look for predictors of different symptom paths, and examine caregiver well-being as a possible factor linked to symptom severity over time. It also described adapting and piloting a brief online caregiver well-being program, but it did not yet report results. [7]

For patients and caregivers, these papers mainly add context rather than changing care now. The autism prediction work and the planned long-term tracking study are about building tools to better anticipate developmental and behavior needs over time, but both are still exploratory and need confirmation in broader, real-world settings. The case reports highlight that outcomes can range from typical early developmental testing to severe psychiatric symptoms in adolescence, and single-person reports cannot predict what will happen for any other individual. [4] [7] [5] [6]

Across these papers, heart rhabdomyomas (benign heart tumors) in tuberous sclerosis complex (TSC) most often became smaller, either naturally over time or after treatment with medicines that affect the mTOR pathway. At the same time, some children had irregular heart rhythms, and the reports highlight that outcomes can range from stable follow-up to serious illness. The overall evidence is limited because much of it comes from retrospective studies and single-patient case reports. [8] [9] [10]

In a single-center Brazilian study of 69 people with TSC (42 with cardiac tumors, mostly rhabdomyomas), incomplete shrinkage was the most common course, seen in 76.2%, while 16.7% had complete disappearance on heart ultrasound follow-up. Irregular heart rhythms were reported in 21.4%, and one person had reduced heart pumping function (left ventricular ejection fraction under 50%). One death occurred after surgical treatment for a rhabdomyoma, while overall mortality in the full TSC group was low. [8]

In an 11-infant retrospective cohort with symptomatic or large rhabdomyomas treated with sirolimus, tumor size dropped by 32% at 1 month and 77% by the final follow-up, and 64% had complete disappearance. Most rhythm problems resolved, except in one infant who already had congenital complete heart block. After stopping sirolimus, 3 infants (27%) had tumor regrowth, and all shrank again when sirolimus was restarted, showing that regrowth after treatment can happen in some children. [9]

One term newborn with multiple rhabdomyomas received early low-dose everolimus with pharmacist-led drug-level monitoring, and heart ultrasound showed marked tumor regression within 4 weeks, with no treatment-related adverse events reported and stable status at 6 months. Because this was one patient and follow-up was short, it cannot show how often similar results happen or what long-term outcomes look like. [10]

A separate autopsy-based report described a preterm newborn who was not known to have TSC (she had no skin signs of TSC, and genetic testing was declined because of cost), with a very large left-ventricular rhabdomyoma found before birth, severe breathing failure, and pulmonary hypoplasia (underdeveloped lungs), with death by day 3. The autopsy clarified the diagnosis and the extent of lung underdevelopment, while the authors noted that limited imaging and lack of genetic testing in a resource-limited setting made it hard to fully evaluate a possible TSC connection during life. [11]

For families affected by TSC, these studies add context that heart rhabdomyomas often get smaller over time and may shrink quickly when mTOR-pathway medicines (such as sirolimus or everolimus) are used in selected infants, but rhythm problems can occur and some babies can be very sick depending on tumor size and location. Questions that remain include how often tumors regrow after stopping medicine, how long treatment should continue when it is used, and what long-term outcomes look like, since the evidence here includes small cohorts and individual case reports.

This month, two reports described using catheter-based procedures to block blood flow to angiomyolipomas (AMLs) in the kidney or in the retroperitoneum (the space behind the abdomen). Overall, the reports pointed in the same direction: embolization often reduced tumor size and aimed to lower bleeding or rupture risk, while kidney function was generally preserved, though side effects and repeat procedures were sometimes needed. The evidence includes one retrospective study and one single-patient case report, so results may not apply to everyone. [12] [13]

In a retrospective single-center study of 44 people (46 tumors) with very large kidney AMLs (10 cm or larger), selective trans-arterial embolization led to a measurable average decrease in tumor size over follow-up. Kidney function numbers stayed similar before and after the procedure, suggesting the kidneys were often preserved in this group. The authors reported an overall success rate of 84.7%, while 13.6% of patients needed another intervention. [12]

The same study reported side effects that mattered for recovery: post-embolization syndrome occurred in 43.1% and allergic reactions in 6.8%. One patient required kidney removal due to a perinephric abscess, showing that serious complications can occur even when the overall goal is kidney-sparing treatment. [12]

A separate case report described one 83-year-old man with tuberous sclerosis complex (TSC) who had a large angiomyolipoma located outside the kidney in the retroperitoneum. After embolization of three feeding arteries, imaging at 12 months showed the tumor had decreased in size and an aneurysm within it was no longer seen, with no major complications reported. Because this is one person, it mainly shows what is possible rather than what is typical. [13]

For people with TSC who have kidney angiomyolipomas, or the rarer situation of an angiomyolipoma outside the kidney, these reports add to the picture that embolization is commonly used to reduce tumor size and manage bleeding risk while trying to protect kidney function. At the same time, the study shows that short-term side effects are common and that some patients may need additional procedures, and rare serious complications can happen. What remains uncertain from these papers is how well these results generalize across different ages, tumor locations, and longer follow-up, especially for extra-renal retroperitoneal angiomyolipomas where evidence is currently limited to individual cases.

Across these papers, two messages stood out: how LAM can be more confidently confirmed in tissue samples, and what long-term treatment with sirolimus looked like in a very small real-world example. The diagnostic work emphasized markers that were easier to interpret and more consistently present, especially in subtle disease. The treatment report showed lung breathing tests staying fairly stable over time, while day-to-day functional capacity still declined, highlighting that different measures of health can move in different directions. [14] [15]

In an immunohistochemistry study of lung and lymph node LAM tissue, loss of the TSC2 protein was seen in all cases, and cathepsin K staining was strongly positive in all cases, including subtle lung LAM. In contrast, HMB-45 and beta-catenin were more variable or harder to interpret, particularly when lesions were subtle. The authors suggested TSC2 and cathepsin K may be especially helpful for supporting a tissue diagnosis, with TSC2 also serving as confirmation of a genetic pathway linked to mTOR-inhibitor responsiveness. [14]

A case report of identical twin sisters with tuberous sclerosis complex (TSC) and LAM described 10 years of sirolimus treatment with no clinically significant side effects reported. Over time, forced expiratory volume in 1 second (FEV1) declined slowly for both twins, forced vital capacity (FVC) stayed relatively stable, and gas transfer (DLCO) improved from baseline in both. However, their 6-minute walk distance declined, and one twin needed supplemental oxygen earlier, showing that exercise ability and oxygen needs may change even when some lung test results look stable. [15]

The twin report also highlighted that people with the same underlying TSC2 mutation can still experience different symptoms and body-wide effects over time. One twin had greater weight loss, fluctuating sirolimus concentrations, menstrual irregularity, and an earlier need for oxygen therapy, despite the twins' identical genetics and broadly similar lung involvement. Because this is only two patients, it cannot show how common these patterns are. [15]

For patients and caregivers, this set of studies mainly adds background context rather than changing care right now. If a biopsy is done, the marker study suggests some tissue tests (TSC2 loss and cathepsin K) may be more consistently positive than others, especially when LAM changes are subtle, but it does not address how often biopsies are needed or when. The long-term sirolimus case report may be reassuring about stable breathing test trends in two people over a decade, while also underscoring uncertainty: exercise capacity and oxygen needs can still worsen and individuals can differ even with similar genetics, so larger studies are still needed to know how typical these experiences are. [14] [15]

This months papers together point to a complex picture of tuberous sclerosis complex (TSC) biology. In a large adult population study, many people carrying clearly harmful TSC1 or TSC2 gene changes did not have a TSC diagnosis or recorded TSC-related health issues, suggesting incomplete penetrance and possible under-recognition in routine records. In contrast, lab and tissue studies showed that TSC1/TSC2 can strongly reshape cell signaling and cell behavior, but the effects differed by tissue type and setting. [16] [17] [18] [19] [20]

In over 900,000 adults from UK Biobank and All of Us, researchers identified 61 adults with high-confidence pathogenic or likely pathogenic TSC1/TSC2 variants and available phenotype data. Only 20 of 61 (33%) had a recorded TSC diagnosis, and 33 of 61 (54%) had no recorded TSC diagnosis, manifestations, or suggestive medication history, which the authors interpreted as substantial incomplete penetrance and/or missed recognition in health records. They also saw cohort differences, including different apparent penetrance estimates for TSC2 between UK Biobank and All of Us, which could partly reflect who enrolls in each cohort (for example, healthy volunteer bias). [16]

In human brain tissue, detailed protein and phosphorylation mapping suggested that mTORC1 (a growth-control pathway linked to TSC) was not detectably activated in cortical tubers, but was strongly activated in subependymal giant cell astrocytomas (SEGAs). SEGA samples also showed large-scale shifts in proteins related to ribosomes, inflammation, and especially RNA handling and mRNA splicing, and the splicing changes overlapped with patterns described in cancer biology. This supports the idea that different TSC-related brain lesions may be driven by different molecular programs, even within the same condition. [17]

Two exploratory lab studies broadened the range of body systems where TSC genes may play a role. In fruit fly larvae, reducing Tsc1 or Tsc2 in a blood-forming organ increased blood-cell proliferation and altered differentiation, with disrupted maintenance of progenitor cells and more cell death, involving TOR and Notch signaling (and JNK signaling for Tsc2). In human lung endothelial cells, TSC2 knockdown reduced activation of NF-kappaB signaling and lowered inflammatory markers like ICAM-1, VCAM-1, and IL-6 after stimulation, and the authors reported this effect was independent of mTORC1. [18] [19]

One preclinical cancer study designed a TSC2 inhibitor (AcTor) that activates mTORC1, aiming to strengthen the activity of the proteasome inhibitor ixazomib in acute myeloid leukemia (AML). In AML cell lines, patient samples, and mouse models, the AcTor and ixazomib combination showed stronger leukemia cell killing than either alone, with signs of mitochondrial failure and reduced leukemia burden in mice, while the authors noted remaining questions such as pharmacokinetics and the fact that AcTor alone could accelerate disease in models. Although this work was not about TSC, it highlights that directly changing TSC2 activity can have powerful, context-dependent effects on cell survival. [20]

For patients and caregivers, these studies mainly add context rather than changing care now. The population-genetics work suggests that carrying a harmful TSC1/TSC2 variant does not always match what is recorded in medical charts for adults, which matters for how results from broad genetic screening might be interpreted, but it does not explain why outcomes vary from person to person. The brain tissue study reinforces that different TSC-related brain lesions (tubers versus SEGAs) may differ in key signaling pathways, which may help guide future research into why some lesions grow and others do not. The cell and animal studies expand what scientists are testing about TSC genes (inflammation signaling and blood-cell development), but these are early-stage and may not reflect what happens in people with TSC.