Rett syndrome (RTT) is a rare neurodevelopmental disorder predominantly affecting females (1, 2). Prevalence of RTT is estimated at 7.1 per 100,000 females (2) and is a leading cause of intellectual disability in young girls (3). Individuals with RTT characteristically have normal neuropsychic development for the first 6–18 months of life and then experience a progressive deterioration of acquired behavioral and motor skills (4). RTT impacts multiple physiological domains including growth, development, gastrointestinal, respiratory, musculoskeletal, and neurologic function (5, 6). The diagnosis of RTT is made clinically, though there are strong associations with mutations in the methyl-CpG-binding protein 2 gene [MECP2; (7)].

A common neurologic comorbidity associated with RTT syndrome is epilepsy, affecting 50% to 90% of individuals with RTT (7, 8). Seizures tend to present between the ages of 3 and 20 (9) with peak frequency reported in ages 7–12 (10). Incidence of epilepsy is higher in individuals with RTT who experience more severe developmental disabilities (11). Seizure types in RTT vary and include focal onset, complex partial, atypical absence, atonic, and generalized tonic-myoclonic (9, 12). Generalized seizures are correlated with treatment-resistance while seizure presentation after age 5 predicts future control (13). Genetic variants of RTT can be associated with different epilepsy presentation. Individuals with cyclin-dependent kinase-like 5 (CDKL5) variants generally present with epilepsy before age 1 while individuals with Forkhead Box G1 (FOXG1) variants generally present with epilepsy at 2 years of age with drug resistant seizures (13). The association of genetic variants of MECP2 mutations with epilepsy ranges from an estimated 48%-100% (12). Only 20% of individuals with epilepsy in RTT achieve seizure control (9).

Families and caregivers are negatively burdened by epilepsy, especially among those who experience generalized, prolonged, cyanotic, or refractory seizures (14, 15). Parents of individuals with RTT and epilepsy can experience fear of seizures as well as difficulty in finding home care for their children (14).

Real-world evidence regarding epilepsy in RTT, frequency, severity of seizures, and their impact upon healthcare resource utilization (HCRU) is limited. Claims-based analysis frequently underreports symptoms and clinical events using diagnosis codes only. Thus, having access to clinical chart progress notes can provide important information regarding seizure frequency.

Currently, there is limited published real-world data providing detailed information about seizures among RTT syndrome. To better understand the prevalence, seizure severity (severe vs. non-severe), frequency (daily, weekly, monthly), and HCRU related to epilepsy among individuals with RTT syndrome, we completed a retrospective database study with structured (diagnosis codes) and unstructured data (clinical progress notes) among individuals with RTT receiving care at a RTT Center of Excellence in the United States.

A total of 126 individuals were identified with a RTT ICD-10 code. Of these, 28 individuals were excluded based on the absence of confirmed RTT in the clinical progress note. A total of 98 individuals met study eligibility criteria and were included in the final analyses (Figure 1). There were 71 (72%) RTT individuals with epilepsy, of which 33 (46.5%) had documentation of severe epilepsy (Figure 1). Of the 71 individuals with epilepsy, 29 (40.8%) had focal epilepsy, 23 (32.4%) had generalized epilepsy, and 19 (26.8%) were unspecified with regards to focal or generalized epilepsy. Methods of epilepsy identification are shown in Supplementary Table 2, where epilepsy was identified using encounter diagnosis (68.4%), clinical notes (52.0%), and epilepsy medication (28.6%). Methods of identification of individuals with severe epilepsy are shown in Supplementary Figure 1 and include clinical progress notes alone (3 of 33), ICD-10 codes alone (19 of 33), and both clinical progress notes and ICD-10 codes (11 of 33). Of the 98 individuals with RTT, majority were female (n = 91, 92.9%), and most were of white race (n = 60, 61.2%); 30 (30.6%) did not have race documented (Table 1). Prevalence of select comorbid conditions included dysphagia (40.8%), constipation (37.8%), gastroesophageal reflux disorder [GERD] (37.8%), scoliosis (36.7%), vomiting (30.6%), sleep apnea (30.6%), cardiac arrhythmias (21.4%), and QT prolongation (12.2%) (Table 1).

There were 59 RTT individuals with epilepsy who had 12 months of post-index follow-up of clinical history following the index epilepsy diagnosis. Table 2 displays the prevalence of providers visited and medications prescribed during the follow-up period among the 59 RTT individuals. The most commonly prescribed AEDs were levetiracetam (45.8%) followed by clonazepam (37.3%), oxcarbazepine (30.5%) (Table 2).

Figure 2 displays seizure frequency among overall epilepsy RTT individuals. There were 40 RTT individuals with documentation from clinical notes categorizing seizure frequency. Overall, one quarter (25.0%) of RTT individuals with epilepsy experienced daily seizures. Frequent seizures were more common among individuals with severe epilepsy relative to those with non-severe epilepsy. All individuals with severe epilepsy experienced seizures daily (53.8%), weekly (23.1%), or monthly (23.1%); while over half (51.9%) of individuals with non-severe epilepsy experienced no seizures during 12 months post-index. Among RTT individuals with non-severe epilepsy, 11.1% experienced daily seizures, 14.8% experienced weekly seizures, and 22.2% had monthly seizures.

Comparing all-cause HCRU during the 12-month follow up period (Figure 3), the prevalence of ED visits was significantly higher among those with severe epilepsy vs. those with non-severe epilepsy (27% vs. 7%, p < 0.05). Inpatient admissions were also significantly more prevalent among those with severe epilepsy vs. non-severe epilepsy (37% vs. 14%, p < 0.05). There were no significant differences in the prevalence of outpatient visits across the two groups. In the year following diagnosis, epilepsy-specific inpatient admissions were documented for 12 (20%) and ED visits for 4 (7%) of the 59 individuals with epilepsy (Figure 4). Compared to individuals with non-severe epilepsy, individuals with severe epilepsy were significantly more likely to have an epilepsy-specific inpatient admissions (33% vs. 7%; p < 0.05). Though not statistically significant (NS), individuals with severe epilepsy were also more likely to have ≥1 ED visits (10% vs. 3%, NS).

Figure 5 displays the results of the 12-month pre/post-index analyses comparing all-cause HCRU before and after epilepsy diagnosis. Among the 71 individuals with documented epilepsy, a total of 29 (41%) had clinical history during the 12 months pre-index and 12 months post-index. There was an increase of inpatient admissions post-epilepsy diagnosis during the 12 months post-index vs. 7% during the 12 months pre-index (p < 0.05). The trend was similar for outpatient and ED visits, but not statistically significant; during the baseline 12 month prior to initial diagnosis compared to the follow up period, higher percentages of RTT individuals had ≥1 ED visits during the 12 months post-index compared to pre-index, 14% vs. 7% (NS), respectively; and higher outpatient visits during post-index compared to pre-index, 79% vs. 66 % (NS), respectively.

To date, there are no published studies examining epilepsy outcomes and related HRCU in individuals with RTT using a combination of structured EHR data and unstructured data abstracted from clinical progress notes from a RTT COE. In our study population, epilepsy was common (72%), with close to half (46%) experiencing prolonged, cyanotic, or intractable seizures. In this study, the rate of daily seizures (25%) among individuals with epilepsy was higher compared to previously reported studies of 11% (8). Daily seizures are associated with increased caregiver burden and lower emotional quality of life (15). While the elevated frequency of seizures among the study population may be a result of selection bias due to a more profoundly affected RTT population that seeks care at a COE or a limitation of the available seizure frequency data, seizure frequency reported by parents of children with epilepsy tend to underestimate actual seizure frequency (16).

This study highlights the risk and burden to patients and families associated with epilepsy, especially those with refractory, cyanotic, or prolonged seizures. In this study, individuals with these types of severe seizures were 3.9 times more likely to visit an ED and 2.5 times more likely to be admitted to a hospital in the year post initial diagnosis. Clinically, the consequences of refractory seizures include increased risk for neuropsychiatric complications (e.g., ADHD, depression) and decreased life expectancy (17). Individuals who experience prolonged seizures and/or status epilepticus are increased risk for pneumonitis and aspiration pneumonia, respiratory failure, neurological deficits (hemiparesis, cognitive decline, dysphasia), fractures, decreased growth, and increased mortality (3, 6, 18–20). It is not surprising that individuals with RTT who also experience these more serious forms of epilepsy are more likely to require emergent and inpatient care. However, seizures appear to contribute to caregiver burden even in non-severe forms (21).

The most common causes of mortality in individuals with RTT are respiratory-related including lower respiratory tract infection, aspiration/asphyxiation, and respiratory failure (22). There is a strong link between aspiration and subsequent respiratory failure in individuals with RTT (23). Individuals with RTT have multiple comorbidities that put them at risk for aspiration and subsequent respiratory complications including dysphagia, GERD, scoliosis as well as epilepsy (23). Gastrointestinal comorbidities, also common in this RTT cohort, can also be a source of inpatient admissions and ED visits (24, 25). The increased healthcare utilization in individuals with RTT and epilepsy found in this study is likely not an isolated phenomenon but rather the result of interactions between multiple comorbid conditions.

While the diagnosis of RTT is made clinically, there are strong associations with genetic mutations and clinical presentation including seizure onset and severity. Individuals with CDKL5 and FOXG1 mutations appear to have earlier onset of seizures and those with large MECP2 gene deletions have later onset (11, 26). Individuals without identifiable genetic mutations appear to be at lower risk for epilepsy (11). While beyond the scope of this study, future research is warranted to explore the interaction between RTT genetic variant, epilepsy severity, healthcare utilization and clinical burden. While AEDs are effective therapy, either as monotherapy or in combination, for many individuals who have seizures in RTT, 20–40% experience drug-resistant epilepsy (27). While almost all the most common AEDs for long-term seizure control reported in our study (levetiracetam, oxcarbazepine, lamotrigine, valproic acid, and topiramate) are recommended in one or more clinical guidelines for epilepsy in RTT, with the exception of oxcarbazepine, specific recommendations regarding first-line agents or care pathways are lacking (28). Due to the non-localized nature of epilepsy in RTT, unlike individuals with other types epilepsy (i.e., focal onset epilepsy), surgical treatments are generally not effective for drug-resistant epilepsy in RTT, though corpus callosotomy has been used successfully in select cases (27, 29, 30).

The additional burden placed upon families with RTT who also have seizures, especially among those with refractory and/or frequent, prolonged seizures, is substantial and can result in caregiver anxiety, depression, and decreased quality of life (31, 32). Seizures are the second most reported concern by caregivers in RTT, second only to lack of effective communication. Parental concerns related to seizures tend to be minor among children under the age of 5, grow until they peak at 15–20 years of age, and slowly decline after age 20 (33). Increasing seizure frequency in RTT is directly associated with increased caregiver burden (34). AED treatment can also add burden to patients and families. Use of multiple AEDs is associated with increased caregiver burden (32), and many individuals experience significant side effects including behavioral disturbances, anorexia, nephrolithiasis, and sedation (3). Interventions such as music therapy, reported to reduce caregiver stress and reduce seizure frequency, can be an important component of a RTT treatment plan (35). Thus, to control or stabilize seizures, especially refractory seizures in RTT, have the potential to reduce unwanted healthcare utilization and profoundly improve the quality of life for individuals with RTT and their families.

There are several limitations to address. This study was performed based upon data from a single health system whose coding and documentation practices may differ from other health providers. This study was also limited to clinical encounters and documentation related to care received at VUMC and may not represent the full healthcare experience. In addition, the population of VUMC individuals with RTT is relatively small and has substantial variability in available data, limiting the ability of this study to detect significant differences between cases and comparators. VUMC is a RTT COE and their population of RTT individuals may differ from other health systems. The RTT population seeking care at a RTT COE may represent a cohort of individuals with more severe RTT with more disabling disease and more symptomatic epilepsy. As a result, the results presented in the study may not represent the experience of individuals with RTT cared for at other clinical settings. Additionally, the sample sizes of the cohorts and sub cohorts in this study are small, and as a result, limits the generalizability of our findings. For example, the results of the pre- post-longitudinal analysis, limited to a subpopulation of 29 individuals, while suggestive of increased utilization associated with incident epilepsy, should be interpreted with caution and validated with a larger sample size. High-quality, rich clinical datasets for rare conditions such as RTT can be associated with sample limitations. While the authors sought to maximize the value of the available data using multiple study designs, analyses of larger populations of individuals with RTT are warranted.

RTT is also a heterogeneous condition with typical and atypical variants, different genetic mutations, and numerous comorbidities which can impact seizure activity and healthcare utilization. Exploration of these clinical and genetic features is beyond the scope of this study and deserves further analysis (6, 23). With regards to the accuracy of parental report of seizure activity, manifestations of RTT such as breath holding, staring, and unusual movements can be mistaken for seizures if not appropriately correlated with EEG-video monitoring (9). Conversely, parents of children with epilepsy commonly under-report seizure activity, both when parents are able and not able to directly observe their child (16). Confirmation of seizure frequency with EEG-video monitoring, however, was beyond the scope of this initiative.

There are strengths to mention such as real-world evidence studies of individuals with RTT are limited and this study of epilepsy in individuals with RTT is the first to assess associated HCRU at a RTT COE. The combination of structured and unstructured data provides greater understanding of seizure frequency and severity that structured data alone.

Among the RTT COE data, over 70% of individuals with RTT had comorbid epilepsy and 25% experienced daily seizures. Individuals with RTT and refractory or prolonged seizures were more likely to experience inpatient admissions and ED visits. It is important to recognize that individuals with RTT have high concomitant rates of epilepsy, and to alleviate additional burden among caregivers and patients, the goal to stabilize the frequency of seizures is essential; especially for those with severe epilepsy.