Neuronal Ceroid Lipofuscinosis (NCL) disorders are the most common neurodegenerative diseases in childhood, and are reported as the leading cause of childhood dementia worldwide (1, 2). The higher prevalence of selected forms of NCL in restricted geographic areas is historical and might reflect early progress in molecular diagnosis in some countries (3). Epidemiological data indicates an incidence of about 1/1,000,000 (4), and the estimated total incidence ranges from 0.01 to 9 per 100 000 live births (5, 6).

Ceroid-lipofuscinosis, neuronal 8 (CLN8) belongs to the NCL disorders and predominantly affects the central nervous system, leading to progressive cognitive decline, motor impairment, and visual deterioration. Initial descriptions of CLN8 delineated two clearly distinct phenotypes. The late infantile Turkish variant (vLINCL) represents the most severe form, typically initiating between 2 and 7 years (7, 8). Affected patients develop myoclonic epilepsy and ataxia, accompanied by developmental regression leading to the loss of the ability to walk and talk. Cognitive capacity progressively declines in this form and affected individuals rarely survive beyond late childhood or early adolescence (7). Northern epilepsy or progressive epilepsy with mental retardation (EPMR) is the less severe form of the disease, characterized by recurrent seizures (9). It usually does not present myoclonus or visual failure, unlike vLINCL. The usual onset occurs between 5 and 10 years. As the disease advances, affected individuals develop ataxia, other motor dysfunctions, and a gradual decline in cognitive abilities (9, 10). EPMR has a much slower course, and patients usually live longer than those with vLINCL (11).

The genetic basis of CLN8 involves loss-of-function mutations in the CLN8 gene, located on chromosome 8p23.3, which encodes a transmembrane endoplasmic reticulum protein. The function of the CLN8 protein has yet to be entirely elucidated, but it is required for the endoplasmic reticulum-to-Golgi transfer of lysosomal enzymes. CLN8 deficiency leads to depletion of soluble enzymes in the lysosome, thus impairing lysosome biogenesis and leading to a lysosomal storage disorder (12). It has been demonstrated that CLN8 protein forms a complex with the product of CLN6, another gene whose loss of function is associated with NCL, necessary to recruit lysosomal enzymes and promote their Golgi transfer (13). Apparently, the knockdown of CLN8 led to an increase in the size of the Golgi apparatus, the number of mobile vesicles, and the velocity of endo-lysosomes, alongside significant lysosomal alkalinization in CLN8-deficient cells (14). Also, these findings (14) indicate that CLN8 deficiency is involved in atrophy, shortening, and degeneration of the neural dendritic tree. These suggest that the abnormalities induced by CLN8 deficiency in the basal endo-lysosomal system underlie morphological changes in neurons that ultimately contribute to the characteristic neurodegeneration observed in this NCL.

Despite classically defined age windows, an increasing number of patients demonstrate variable progression and onset age, even within the same family (15–17). This clinical variation is typically attributed to patients' genetic background [i.e., modifier genes (15)] and the severity of causal mutations. Recently, mutations in unrelated genes have been considered modifiers of gene expression, and interactions between mutated genes and modifiers can lead to clinical variations and observed phenotypic heterogeneity (1). Rare cases with CLN8 pathogenic variants report congenital presentations or symptoms onset in the first year of life, as well as presentations deviating from previous NCL paradigms, such as the absence of myoclonic seizures or visual sensory loss (10, 18, 19). These reports further suggest a continuous spectrum of phenotypes, a phenotypic heterogeneity associated with CLN8 instead of a clear distinction between EPMR and vLINCL (20, 21).

Here, we present a congenital case of CLN8 with a very slow disease progression confirmed by pathologic subcellular markers. Whole genome sequencing (WGS) revealed the diagnosis in the context of a national genomics academic program.

The patient is a ten-year-old Uruguayan male, son of non-consanguineous parents, with no relevant family or perinatal history. In the neonatal stage, he presented sucking disorders, which led to malnutrition in the first trimester of life. He presented developmental compromise from an early age, achieving cephalic support at nine months and independent standing at three years of age, without ever having acquired independent walking or oral language. He never presented a loss of acquired maturational behaviors. At ten years of age, he presents a severe intellectual deficit and autism spectrum disorder (ASD), characterized by limited communicative intention and visual contact with frequent manual stereotypies (flapping).

At three years of life, he started with epilepsy in the form of asymmetric and alternating hemibody focal tonic seizures, occasionally evolving to generalized clonic seizures, always of short duration. He never presented with myoclonic seizures. Since the onset of seizures, he has always presented several seizures per week, with a poor response to multiple antiepileptic drugs (valproic acid, levetiracetam, phenobarbital, clobazam, cannabis).

Physical examination showed a normal head circumference without dysmorphic features, poor eye contact, and frequent manual stereotypies. He presented nonparetic hypotonia with normal osteotendinous reflexes and a plantar cutaneous reflex in bilateral flexion.

A first brain magnetic resonance imaging (MRI) was performed at two years of age, which was normal. An additional brain MRI was done at nine years of age, which showed cerebellar and mild cerebral atrophy (Figure 1). A basic metabolic study was performed, which included normal urine organic acids and normal blood amino acids. Transferrin isoelectrophoresis was normal. Visual and auditory evoked potential tests at the age of 3 years were normal. Multiple electroencephalograms (EEG) have been performed since age three, showing a poorly organized and slow background rhythm with focal activity in different topographies. No EEG with low-frequency photic stimulation was conducted.

Karyotyping and array comparative genomic hybridization were normal.

He did not receive any other pharmacological treatments. Since infancy, he has been receiving speech and occupational therapy, with limited progress in cognitive domains. The child attends a particular education school.

Although the first clinical manifestations occurred congenitally and the child has achieved few developmental milestones throughout his life, there was no progression of the condition beyond the appearance of his epilepsy, behaving almost statically.

We present a new case of NCL caused by two previously unreported likely pathogenic mutations in the CLN8 gene. The patient had an atypical presentation, being congenital but with a very slow progression of the disease, an observation that supports a relatively broad spectrum of presentation of NCL types.

Epilepsy is common to all forms of NCL (1). While generalized epileptic seizures [Generalized tonic-clonic seizure (GTCS), myoclonic, atonic, or absence] are present in all known cases of CLN8, focal seizures, as reported in our case, have been infrequently documented. It is also noteworthy that spontaneous myoclonus was absent in our case. Myoclonus is a typical feature of NCL, although there are reports of cases lacking this type of crisis. As observed in our case, the EEG shows a progressive slowing of the background rhythm, along with interictal discharges, including slow spike-wave or polyspike-wave complexes (20). A characteristic feature of the EEG in NCL patients is the paroxysmal response to spike-wave, evoked by intermittent low-frequency photic stimulation (1–3 Hz) (1, 36). However, different types of NCL show varying susceptibility to photoparoxysmal response (2). Unfortunately, performing a low-frequency photic stimulation EEG in our case was impossible.

The progression and severity of deterioration vary widely and generally parallel to the overall course of the disease (20). In the present case, it has been congenital and very slowly progressive, simulating a non-progressive course. Behavioral phenotypes with ASD features have also been reported before (2, 36). Heterozygous variations in CLN8 have been proposed to confer increased susceptibility to ASD (37).

Brain MRI may appear normal in the early stages of the disease. Cerebellar atrophy has become a consistent feature over time and is present in our patient at age 9, affecting both the vermis and cerebellar hemispheres as previously described (2, 36). Supratentorial cerebral atrophy is often observed but could be a late feature, as in our patient (20). A relevant aspect of this clinical case is two MRIs separated seven years from each other, in which progressive brain atrophy is not observed. As we already mentioned, we did not observe neurological clinical progressivity either.

Treatment is symptomatic and multidisciplinary, focusing on providing the best quality of life (1, 38). Some pharmacological treatments are in preclinical studies: AMPA receptor modulator therapy (ZK-187638), retigabine, insulin-like growth factor 1, as well as gene therapy (vectors-Amicus Therapeutics' AAV9-CLN8) (11, 16). Stem cell therapy in mouse and dog models of CLN8 led to CLN8 expression in the spinal cord but not in the CNS (11). Most of these therapies in development are potentially disease-modifying, meaning they may delay or even halt disease progression, but few are likely to reverse the disease, even partially. Therefore, early diagnosis and treatment will become increasingly crucial as these damage-limiting interventions become available (16). Reports on life expectancy and mortality in NCL are scarce. Variability within each form, even in CLN8, does not allow precise life expectancy predictions in individual patients. Over the last two decades, generally longer survival has been observed in patients with many forms of NCL, regardless of mutation severity. Such findings can largely be attributed to improved care for these patients (1).

Ultrastructural examination through skin biopsy remains useful for confirming genetically undiagnosed atypical forms (2). In this case, it was motivated by the variants detected in the CLN8 gene after WGS analysis. Results showed typical findings, consisting of abnormal accumulation of macromolecular material.

To date, ClinVar reports eight missense, 14 frameshifts, 21 nonsense, and one splice site CLN8 mutations classified as pathogenic (https://www.ncbi.nlm.nih.gov/clinvar/, accessed on 26th February 2024). Studies have also been published reporting those mutations associated specifically with CLN8 disorder (accessible through the NCL resource https://www.ucl.ac.uk/ncl-disease/). According to the numbers obtained from ClinVar, missense mutations appear to be an unusual or under-detected cause in CLN8-related disorders.

A close inspection of the substitutions also supports pathogenicity considerations of the patient's variants. First, Pro260 is a strictly conserved residue, probably due to a stabilizing role of the C-terminus of an alpha helix in its cytoplasmic face, just after the transmembrane segment (39–41). The Pro260Leu substitution may imply a conformational distortion of the protein, affecting its dimerization (42) and hence its downstream actions, which include exporting protein products to the lysosome (12). Second, the effect of the Val261Ala mutation may be less dramatic, yet changing the start of the 261VDWNF265 motif, an export signal from the endoplasmic reticulum to the Golgi (43). In line with a recessive inheritance pattern and the fact that the parents are healthy, we surmise that both variants compromise the protein function. Cascading effects at the molecular and cellular level would result from expressed pools of proteins carrying one or the other mutation.

The discovery of these two missense variants in a new case of NCL expands the repertoire of reported variants associated with this disorder, providing insights into the intricate nature of phenotypic presentations in CLN8. Functional studies providing evidence of a loss of function effect of the two variants are needed, and the impact of specific missense mutations in the product of CLN8 deserves further investigation. Defining genetic (and eventually epigenetic) variation that impacts CLN8 function will be essential to understanding the increasingly complex relationships between NCL genotype and phenotype and advancing treatment options (15).