July 2026 Research Update

This month added a clearer picture of how trofinetide is being used in everyday care for very young children, focusing on how long families stay on it rather than whether it improves symptoms. In insurance and specialty-pharmacy claims data, most children ages 2 to 4 who started trofinetide remained persistent during follow-up, while a smaller group stopped earlier, and a minority of those later restarted. [1]

Children who did not remain persistent had more signs of multisystem complexity recorded before starting trofinetide, including more swallowing problems and more use of feeding tubes, and more breathing irregularities. The study could not explain why people stopped (for example, side effects, perceived benefit, or access), because claims data do not include those reasons, but it helps describe which families may face more complicated treatment journeys in real-world settings. [1]

For patients and caregivers, this evidence mainly helps set expectations about real-world treatment patterns rather than outcomes. It suggests that many young children who start trofinetide stay on it for months, but that treatment interruptions can happen, especially in children who already have more feeding and breathing-related medical needs recorded in their charts. What remains uncertain is the reason for stopping or restarting, and whether persistence reflects tolerability, access, perceived benefit, or some mix of factors. [1]

This month, human studies moved the field toward more practical, consistent brain-based measurements in Rett syndrome. One study showed that automated EEG cleaning can keep far more usable signal than simply throwing out noisy segments, making shorter recordings more workable. Another connected everyday sensory differences to less consistent timing of brain responses, pointing to a measurable brain signal that may relate to sensory symptoms. [2] [3]

In a multisite EEG analysis, a correction-based automated processing pipeline retained about 95% of the recording on average, compared with about 28% when using a rejection-based approach. The study also estimated that stable resting EEG spectral features could be obtained after roughly 76 to 136 seconds of artifact-free data, and that a 3-minute raw EEG could still preserve relationships with a clinical severity measure when processed this way. [2]

In a separate study focused on sensory function, females with Rett syndrome showed both higher hypo-responsivity and higher hyper-responsivity on caregiver reports, meaning under- and over-reacting can co-occur. Their brain responses to visual and auditory stimuli showed lower inter-trial phase coherence (a measure of how consistently the brain response lines up in time across repeated trials), and lower consistency was linked to smaller evoked responses and more severe visual-related symptoms. [3]

For families, these findings mainly matter as progress toward better research tools and potentially more feasible testing visits, rather than new treatments. The EEG pipeline work suggests that shorter recordings may still yield stable measurements if processed with methods designed to correct common artifacts, which could help studies include more participants who cannot tolerate long sessions. The sensory evoked-potential work helps clarify that sensory challenges may reflect, in part, less reliable timing of brain activity, but it is still early research and not yet a test used to guide day-to-day care. [2] [3]

This month was heavy on early-stage research, with several studies trying to solve a core Rett challenge from different angles: how to safely adjust MeCP2-related biology and how to measure system-wide effects beyond the brain. The results were generally encouraging as proof-of-concept in models, but they also highlighted practical hurdles like dosing control, delivery into the brain, and the need to validate proposed mechanisms in more human-relevant systems. [4] [5] [6]

Two animal studies focused directly on changing MeCP2 levels in the brain. In a Mecp2-knockout mouse model, systemic AAV9 gene delivery combined with focused ultrasound to open the blood-brain barrier increased the fraction of brain cells expressing Mecp2 and was reported to improve survival, growth, breathing function, and activity, while also raising a safety concern about possible MeCP2 overexpression. [4]

In a mouse model of MECP2 duplication, a single dose of divalent siRNA (a gene-silencing approach) reduced Mecp2 expression for at least 4 months and was reported to rescue early mortality and some behavioral impairments. Together with the gene-delivery study, this highlights how Rett-related research is working in both directions: raising MeCP2 where it is missing and lowering it where it is too high, while still needing careful control of the final MeCP2 amount. [5]

Several studies explored underlying biology that could eventually guide new treatments or explain symptoms. Analysis of human Rett brain tissue identified heat shock factor 1 (HSF1), a stress-response pathway, as strongly altered, and in mouse models, increasing HSF1 worsened outcomes while inhibiting HSF1 improved Rett-like measures such as motor learning and general health. This points to stress-response signaling as a potential research target, but the intervention results are in mice rather than people. [6]

A longitudinal nonhuman primate model suggested Rett-related changes may involve the gut and the heart early in life, not only the nervous system. MeCP2-mutant monkeys showed delayed gut microbiome maturation, lower levels of certain short-chain fatty acids in stool at 9 months, signs of intestinal inflammation, and longer QTc on ECG without clear evidence of heart muscle injury markers, which together raises testable questions about body-wide involvement over development. [7]

Other lab studies added building blocks for future mechanistic work. A vagus nerve stimulation study in Mecp2+/- rats found higher activation thresholds and weaker locus coeruleus responses, but also showed those responses could be adjusted by changing stimulation parameters. This creates baseline information for neuromodulation research, but it does not yet show how VNS would affect symptoms in people with Rett syndrome. [8]

Cell and computational studies addressed related questions about gene regulation and drug mechanisms. CRISPRtune, a MeCP2-dCas9 tool, enabled intermediate and tunable gene repression in human cell lines, and Rett-associated MeCP2 mutations weakened this repression, helping clarify which parts of MeCP2 matter for gene control. Separately, a computer modeling study proposed that trofinetide could plausibly interact with multiple brain targets, which offers hypotheses to test in experiments rather than evidence of how it works in patients. [9] [10]

Finally, two studies broadened the genetic and molecular landscape around Rett-like neurodevelopmental conditions. A mouse study manipulating two MeCP2 phosphorylation sites reported that the combined effect was not simply additive and appeared more biased toward excitatory (glutamatergic) neurons, linking cell-type-specific molecular changes to behavior in that model. An edited cell-line study of a CDKL5 missense change showed reduced CDKL5 levels and more cell death signals in a non-neuronal system, providing a starting point for future work in neuronal models. [11] [12]

A separate consensus project proposed a practical definition of developmental regression in childhood as a skill loss lasting at least 4 weeks in at least 1 of 6 areas such as communication, motor skills, or self-help. While not Rett-specific, it provides shared language that could make future research studies more comparable when they describe regression, which is relevant to many neurodevelopmental conditions. [13]

For patients and caregivers, these studies mostly matter as early research that shapes what might be testable later, rather than changes that affect care right now. The MeCP2-raising and MeCP2-lowering animal results show that both directions can improve disease features in models, but they also underline that controlling the amount of MeCP2 is a major safety and design challenge. The gut-heart findings in primates and the stress-response pathway work in human tissue help clarify that Rett biology may involve whole-body systems and cellular stress, but it remains uncertain how directly these model findings translate to individual people or to specific symptoms. [4] [5] [7] [6]

This month included a single-patient report that expands the list of genetic changes that can look partly like Rett syndrome. The report described a female with a de novo deletion at 9q34.11 affecting SET, confirmed with detailed genetic testing, and patient-derived cells showed markedly reduced SET protein. [14]

Clinically, the patient had no functional speech earlier in life but later gained limited functional communication in adulthood, in the context of early speech therapy. As a single case, it cannot show what is typical, but it supports the idea that SET haploinsufficiency can be linked to neurodevelopmental features that overlap with Rett presentations. [14]

For families, this case mainly adds understanding that Rett-like symptoms can sometimes arise from genetic changes outside MECP2, and that careful genetic testing can clarify the underlying cause in some individuals. It does not establish a common pathway for most people with Rett syndrome, and it does not test a treatment, but it may matter for research into overlapping neurodevelopmental conditions and how specific genes affect communication over time. [14]