Standardized assessment of seizures in patients with juvenile neuronal ceroid lipofuscinosis

METHOD

RESULTS

A total of 98 individuals with a clinical diagnosis of JNCL were evaluated over a 12-year period. A small proportion of participants were excluded because of lack of genetic confirmation (n=5) or missing UBDRS seizure subscale data (n=7). Data from 86 individuals (44 males, 42 females) were analyzed in this cohort, with an average of 3.1 assessments per participant (range 1–11y, SD 2y 10mo; Table I). The sample was predominantly non-Hispanic white in ethnicity and race which was anticipated given the known higher prevalence of NCL disorders in Scandinavian and northern European populations. The majority of participants were homozygous for the common CLN3 mutation, an approximately 1 kbp deletion that includes exons 7 and 8. Remaining participants were either compound heterozygotes for the common deletion and another unique mutation, compound heterozygous for two unique mutations, or homozygous for one unique mutation (Table II). Because of the variety in genotypes for those with unique mutations, genotypes were separated into two categories – common deletion homozygotes and other genotypes. Across all assessments, the age of participants ranged from 4 years 10 months to 31 years 1 month.

Unless otherwise stated, data are based on the most recent assessment completed. Of all 86 participants, 86% (n=74) reported ever having a seizure. Of these 74, 97% (n=72) reported having had seizures within the year before the assessment. The mean age at seizure onset was 9 years 7 months (SD 2y 10mo, range 10mo–20y). Seizure onset for one participant was at 10 months, the next youngest age at seizure onset was 6 years.

A variety of seizure types were reported, most commonly generalized tonic-clonic (78%, n=67), but also complex partial or absence, myoclonic, atonic, and simple partial seizures (Table III). Of the 72 individuals with seizures reported at the most recent assessment, there were an average 1.7 (SD 0.9) different seizure types. Almost half (49%, n=36) reported only a single seizure type. Fewer participants demonstrated multiple seizure types: two (31%, n=23); three (12%, n=9), or four (5%, n=4). Two participants had experienced seizures in the past, but did not experience any seizures during the interval evaluated for the most recent assessment (past one year). With the exceptions of myoclonic seizures and simple partial seizures which occurred at higher frequency, seizures occurred on average less than once every three months (Table III). The post-ictal recovery time from generalized tonic-clonic seizures was approximately one hour on average.

In the 6 months preceding the most recent assessment, seizure-related injury ‘sometimes’ occurred in 13% (n=11) of the sample. Higher injury frequency (frequent, always) was not observed. Also in the 6 months before, most participants did not require emergency medical intervention for seizures: 74% (n=64) did not require any intervention; 13% (n=11) required first aid at home; for 7% (n=6) a paramedic was called; and 6% (n=5) were seen in an Emergency Department. Inpatient hospital care for seizure treatment in the 6 months before occurred in only 4% (n=3). In the 1 month before, 17% (n=15) required adjustment of anti-epileptic medications for improved seizure control. The median UBDRS total seizure subscale score was 14.8, and the median CGI seizure was 3 (mild severity). There was no significant difference in total seizure subscale score according to sex (male median=13.9, female median=13.0, U=855.5, p=0.56) or genotype group (common deletion homozygotes median=14.8, other genotypes median=11.1, U=593.5, p=0.06).

Medication history was available for 73 of 74 participants reporting seizures. At the most recent evaluation, 80.8% (n=59) used one or two anti-epileptic medications (Table IV). Valproate (49.3%, n=36) and levetiracetam (41.1%, n=30) were the most commonly used medications. Other anti-epileptic medications in use included: zonisamide (n=11); lamotrigine (n=10); clonazepam (n=9); carbamazepine (n=7); topiramate (n=6); oxcarbazepine (n=5); phenobarbital (n=3); phenytoin and lorazepam (n=2, each); acetazolamide, rufinamide, and clorazepate, (n=1, each). No individuals reported use of the ketogenic diet or vagal nerve stimulation. Among participants with seizures, the number of anti-epileptic medications showed a weak positive correlation with seizure subscale score (r=0.35, p=0.002).

At the most recent assessment, the UBDRS total seizure subscale score was highly positively correlated with CGI seizure (r=0.80, p<0.001). Including individuals with no history of seizure, the seizure subscale score showed moderate positive correlation with CGI overall (r=0.49, p<0.001). When evaluating only participants with positive seizure history, there was no correlation between total seizure subscale score and CGI overall (r=0.22, p=0.08).

In evaluation of the most recent cross-sectional data of participants with seizures, seizure severity showed weak positive correlation with age (r=0.33, p=0.005). Using all available data with multiple visits per participant, we found an overall mean increase per year in seizure severity of 0.97 points (95% confidence interval [CI]: 0.61–1.33, p<0.001). Males had a higher rate of change per year (1.17 points, 95% CI: 0.71–1.63, p<0.001) as compared to females (0.61 points, 95% CI: 0.04–1.19, p=0.04), but the difference was not significant (p=0.14). Annual rates of change in seizure severity did not differ (p=0.63) between common deletion homozygotes (1.03 points, 95% CI: 0.56–1.50, p<0.001) and those with other genotypes (0.85 points, 95% CI: 0.30–1.41, p=0.003).

In evaluation of the most recent Physical Impairment subscale assessment, we found an average myoclonus score of 0.18 (range 0–3). Only 12 of 98 participants (12%) demonstrated myoclonus observed by the trained examiner, typically of mild to moderate severity.

DISCUSSION

In our JNCL cohort, the majority experienced seizures, starting at approximately 10 years of age. In contrast to the broader literature^4–6 regarding a common association between refractory myoclonic epilepsy and NCLs, JNCL is most commonly characterized by generalized-tonic-clonic seizures that are infrequent, controlled with one to two medications, and which demonstrate slight progression in severity over time. Some of the literature before included individuals for whom the JNCL diagnosis was based on clinical characteristics or pathophysiology. This is now recognized as less reliable, as these features can overlap with other NCL forms. In addition, there is evidence of pathologically consistent NCL cases which are negative upon genetic assessment of the known genes. Following identification of the CLN3 gene in 1995¹ and clinical availability of genetic testing in the late 1990s, gene-based diagnosis has become standard for the NCLs.

Myoclonic seizures, although they may be present and prominent in other NCLs, are not the hallmark of JNCL. We did not find a high prevalence of diagnosed myoclonic seizures or of observed myoclonus on examination. There is a single case series where three of eight individuals identified as having JNCL had myoclonus.⁹ However, it is notable that in this series the diagnoses of JNCL were made based on age at onset without confirmation by DNA testing. This study evaluated all NCLs, but grouped patients into infantile, juvenile, and adult classes. A late infantile categorization was not included, which is now recognized as a distinct disorder.⁹ Absence of myoclonic seizures clinically should not dissuade consideration of the diagnosis of JNCL.

When comparing our JNCL cohort to published literature on other NCLs, seizures in infantile NCL (INCL) and late infantile NCL (LINCL) occur earlier in the disease course, at 3 to 4 years of age, and most affected individuals have multiple seizure types that are refractory to treatment.^15–17 In contrast, our data suggest that almost half of patients with JNCL demonstrate only a single seizure type. Furthermore, myoclonic seizures are highly characteristic of LINCL. In Northern epilepsy syndrome (NES, CLN8), seizure onset is at a mean 6 years 9 months and is characterized by generalized tonic-clonic seizures increasing in severity over time, followed by decline or stabilization of seizures in early adulthood, and vision loss in adulthood.¹⁸ Seizures are typically the presenting symptom in NES, in contrast to the early childhood vision loss in JNCL as a presenting symptom.^2,19 In the variant LINCL form of disease resulting from mutations in the same CLN8 gene, affected individuals have severe refractory epilepsy with prominent myoclonus in early childhood.²⁰

This study’s cohort received care from clinicians across the United States, and valproate and levetiracetam were the most commonly used medications. As ours is an observational study, not an interventional trial, we are unable to discern the specific treatment rationale, whether this pattern exists because of prescribing preference on the part of clinicians, or if these results indicate a treatment regimen to which individuals with JNCL are particularly responsive.

Observational studies, particularly in rare diseases, are subject to recruitment bias. Our ascertainment methods are largely dependent on the ability of the patient and caregivers to travel, either to our Center at the University of Rochester, or to annual BDSRA meetings. It is possible that our assessment of seizures late in life is limited, although the age of our sample extends to 31 years, close to or exceeding the average age of death for most individuals affected by JNCL.²¹ These recruitment methods may also bias toward families with financial means for travel. However, in addition to Rochester visits, we also conduct research visits onsite during BDSRA meetings, which is held in a different United States city each year, limiting the latter factor over time. Our study is further limited by reliance upon parent report for categorization of seizure semiology and medication indication. It is possible that parental report is biased toward obvious seizure types with prominent motor manifestations. In an effort to refine the accuracy of parental report, for categorization of seizure type, parents were specifically asked about each potential seizure type with a description of the typical semiology and inquiry as to how the seizures were categorized by the child’s treating physician. This is particularly salient for complex partial and absence seizures, as well as for the differentiation of myoclonic seizures from myoclonus. Because of this inability to distinguish cortical myoclonus from other types of myoclonus using EEG, we also evaluated myoclonus based on direct clinician observation as well, which did not alter our overall conclusion that myoclonic movements, which include myoclonic seizures, are infrequent in JNCL. We considered evaluation of electrophysiological data to confirm diagnoses, however, the quality of EEG reports and availability of raw data for review from center to center was variable and thus this could not be accomplished. For all medications, the specific indication was sought: benzodiazepines were excluded as seizure medications unless expressly reported as non-rescue seizure medications.

Understanding the range of seizure profiles in the NCLs should factor into decisions regarding treatment. For JNCL, where a variety of seizure types may be observed, selection of a broad spectrum agent such as valproate, levetiracetam, or lamotrigine may be prudent. Whereas for LINCL, given the high prevalence of myoclonic seizures, consideration of strategies that will treat myoclonic seizures and avoidance of medications that potentially exacerbate myoclonic seizures, such as phenytoin, carbamazepine, and possibly lamotrigine, are paramount. Clinical distinction among NCL forms is also important for directing the diagnostic workup, which may be prolonged if not recognized early. This is useful in providing families with anticipatory guidance regarding expected disease course and prognosis. These distinctions are also important as it relates to therapeutic development and measurement of disease. Finally, improved understanding of the clinical course of each NCL, including the predominant seizure type, may yield further insights into the neurobiology of these diseases, specifically regarding selective vulnerability of different brain regions impacted over the course of the disease.