Neuronal Ceroid Lipofuscinosis Type 2 (CLN2) and Other Neuronal Ceroid Lipofuscinoses

CLN2 disease due to tripeptidyl peptidase 1 (TPP1) deficiency, with the spectrum of the other CLN diseases (CLN1, CLN3, CLN4, CLN5, CLN6, CLN7, CLN8, CLN10, CLN11, CLN13 and CLN14).

Literature and regulatory information updated: 6 October 2026.

0. Initial Reference Table

1. Introduction

The neuronal ceroid lipofuscinoses form a group of inherited neurodegenerative disorders that collectively constitute the most frequent genetic cause of progressive childhood neurodegeneration. Their common denominator is intralysosomal accumulation of autofluorescent material, ceroid lipofuscin, particularly in neurons of the brain and retina, and a clinical triad of psychomotor regression, epilepsy and visual loss culminating in premature death. “Neuronal ceroid lipofuscinosis” is the term commonly used in Europe, whereas “Batten disease” became widespread in the United States and came to designate the entire group; the current recommendation is to use “Batten disease” as the umbrella term and name each condition by its gene, for example “CLN2 disease” [1–3].

The history of the group is a succession of reclassifications. For decades NCLs were included among the “familial amaurotic idiocies” and interpreted as lipidoses alongside GM2 gangliosidosis. The review by Zeman and Dyken, published in 1969, consolidated the concept of neuronal ceroid lipofuscinosis as a condition distinct from amaurotic idiocy, and the classical eponyms — Santavuori–Haltia for the infantile form, Jansky–Bielschowsky for late-infantile disease, Spielmeyer–Vogt (or Batten) for juvenile disease and Kufs for adult disease — became associated with age-defined forms [2, 20, 21]. In 1973, Santavuori, Haltia, Rapola and Raitta described 15 Finnish patients with the infantile form, integrating clinical, electroencephalographic, ophthalmological and neuropathological findings [22]. In the 1980s and 1990s, storage material was shown to be predominantly protein — mitochondrial ATP synthase subunit c or saposins A and D — definitively overturning the interpretation as a lipidosis [21].

The molecular era began in 1995 with identification of variants in the palmitoyl-protein thioesterase 1 gene (PPT1) in all 42 Finnish patients with the infantile form and identification of CLN3 by the International Batten Disease Consortium [23, 24]. In 1997, Sleat and colleagues used mannose-6-phosphate modification of newly synthesised lysosomal enzymes as an affinity marker and found a single protein absent in patients with classical late-infantile disease: a pepstatin-insensitive lysosomal peptidase, now termed tripeptidyl peptidase 1, whose activity was deficient in autopsy tissue and whose gene carried variants in patients [25]. The historical review by Haltia and Goebel summarises almost two centuries of research and identification, from 1995 onwards, of 13 genes and more than 360 variants, underpinning the current genetic classification [26].

Identification of the CLN2 enzyme deficiency opened a therapeutic avenue unavailable for the other forms: enzyme replacement targeted to the central nervous system. Cerliponase alfa, a recombinant enzyme administered directly into the cerebral ventricles, was authorised in the United States and European Union in 2017 following an open-label study in which motor and language decline was lower than in historical controls [1, 11, 27]. CLN2 disease thus became an example of what changes when treatment becomes available for a childhood neurodegenerative disorder: the interval between the first seizure and diagnosis ceases to be merely descriptive and becomes a determinant of prognosis, because treatment does not reverse established functional loss [6, 28].

This article addresses CLN2 disease as the central condition and integrates, within the sources consulted, the other CLN diseases (CLN1, CLN3, CLN5, CLN6, CLN7 and CLN8) and the group of neuronal ceroid lipofuscinoses as a whole. Rare adult forms (CLN4, CLN11, CLN13), the congenital form (CLN10) and CLN14 are described to complete the differential diagnosis and counselling.

2. Molecular Genetics and Pathophysiology

2.1. A group defined by pathology rather than a metabolic pathway

NCLs share a cellular phenotype — lysosomal accumulation of autofluorescent lipopigment, positive for periodic acid–Schiff and Sudan black B, resistant to lipid solvents — and selective neuronal loss accompanied by gliosis and secondary white matter injury [29]. They do not, however, share a biochemical pathway: proteins encoded by CLN genes include soluble lysosomal enzymes, lysosomal membrane proteins, endoplasmic reticulum proteins and vesicle-associated cytosolic proteins. Stored material is also not specific to each disease: saposins A and D predominate in infantile forms (CLN1, CLN10) and mitochondrial ATP synthase subunit c in late-infantile and juvenile forms, probably as secondary markers rather than primary substrates [2, 21]. Deposit ultrastructure varies between forms: granular osmiophilic deposits (GRODs) in early forms (CLN1, CLN10), curvilinear profiles in classical CLN2 and CLN7, fingerprint profiles in CLN3 and mixed combinations in the others [1, 2].

The number of genes depends on the source and criterion. The Saudubray chapter lists 13 genes, including ATP13A2 (CLN12) and KCTD7 (CLN14), with the proviso that they rarely produce NCL phenotypes; the 2025 GeneReview considers 12 genes confirmed, records CLN12 as described in a single family and notes that CLN9 was withdrawn because the initial association with CLN5 proved incorrect and no gene is assigned to it [1, 2]. This difference does not alter diagnostic practice — panels usually include all these genes — but explains variation in numbers cited in the literature.

Table 2.1 — Neuronal ceroid lipofuscinosis genes, proteins, locations and phenotypes

Sources: Saudubray Table 40.2 and GeneReviews Table 1 [1, 2].

2.2. TPP1 and the pathophysiology of CLN2 disease

TPP1 is located at 11p15.4 and encodes a lysosomal serine protease of the sedolisin family [5, 6]. The enzyme is synthesised as an inactive proenzyme, receives mannose-6-phosphate residues, is targeted to the lysosome by the cation-independent mannose-6-phosphate receptor and autoactivates as pH becomes acidic; the mature 46 kDa form removes tripeptides from the amino terminus of small polypeptides produced by other lysosomal proteases and has weak endopeptidase activity [6, 10]. Physiological substrates are not precisely defined; TPP1 has been linked to macroautophagy and tumour necrosis factor alpha-induced apoptosis [2]. Its expression is developmentally regulated and peaks between ages 2 and 4, precisely when the classical form manifests [3].

Without functional TPP1, undegraded peptides accumulate as ceroid lipofuscin with curvilinear profiles, notably ATP synthase subunit c; massive glial activation and neuronal loss follow, with early cerebellar atrophy and subsequently generalised cerebral atrophy and retinal degeneration [3, 6]. Subunit c accumulation also occurs in other NCLs and other lysosomal disorders, suggesting that it is not the primary defect; shared cascades involving altered lipid trafficking, autophagy, calcium homeostasis and oxidative stress have been proposed [6]. The same logic explains treatment: cerliponase alfa is a recombinant proenzyme taken up by central nervous system cells through the mannose-6-phosphate receptor, translocated to the lysosome and activated there [10].

2.3. TPP1 variants, nomenclature and genotype–phenotype correlation

The update by Gardner and colleagues assembled 131 unique variants in 389 patients (717 alleles). Two variants — c.509-1G>C, an intron 5 splice acceptor site variant (historically IVS5-1G>C), and c.622C>T, p.(Arg208Ter) (historically R208X) — occur in 60 % of patients and represent 50 % of alleles; at least 86 variants (66 %) are private to one family, and 45 % of patients with both alleles known are homozygous [6, 30]. Saudubray reports approximately 150 variants, mostly private [2]. Variants are distributed throughout the protein; most cause loss of activity, but few have been confirmed biochemically, and only approximately 30 % were in ClinVar with a clinical classification at publication [6].

Table 2.2 — TPP1 variants cited in this article (MANE transcript NM_000391.4; protein NP_000382.3)

Sources: Gardner 2019 (NM_000391.3 numbering, matching NM_000391.4 at these positions, verified using VariantValidator), Guelbert 2024 and Biglari 2026 [6, 7, 31, 32].

The atypical phenotype, with later onset and slower progression, accounts for approximately 13 % of registered patients (8 % juvenile, 3 % with SCAR7 and less than 1 % with spastic paraplegia or congenital disease) [6]. These patients' genotypes predict residual TPP1 activity rather than total absence; in other cases activity may be absent in certain cell types but persist in leucocytes, affecting assay interpretation [3]. In general, residual activity, rather than variant type, tracks age at onset and rate of decline across the spectrum from classical to protracted disease [1, 2]. Autosomal recessive spinocerebellar ataxia 7 (OMIM 609270) is an allelic disorder: slowly progressive ataxia into old age, without epilepsy or visual abnormalities [2].

2.4. Other CLN diseases: mechanisms and recurrent variants

In CLN1, PPT1 removes acyl groups, particularly palmitate, attached through thioester bonds to cysteines in S-acylated proteins, allowing their degradation; it contributes to maintenance of the synaptic vesicle pool, and CSPα, altered in CLN4, is probably one of its substrates [2]. Cathepsin D (CLN10) is an aspartic protease involved in apoptosis, autophagy and GABAergic neuron biogenesis; cathepsin F (CLN13) is a cysteine protease cleaving LIMP-2; progranulin (CLN11) participates in inflammation and lysosomal lipid homeostasis [2]. CLN5, whose function was debated for years, is identified in GeneReviews by its current protein designation, CLN5 bis(monoacylglycero)phosphate synthase [1]. Endoplasmic reticulum proteins CLN6 and CLN8 form a complex recruiting lysosomal enzymes and promoting transfer to the Golgi; MFSD8 (CLN7) has been described as an endolysosomal chloride channel and its loss alters mTORC1 signalling; ATP13A2 (CLN12) is an ATPase with a neuroprotective effect; KCTD7 (CLN14) regulates potassium conductance and autophagy–lysosome pathways [2, 33]. In CLN3, battenin influences ionic balance, endolysosomal trafficking, autophagy and mitochondrial function; disease results from loss of function, with the degree of loss correlating with severity and inversely with lymphocyte vacuolation [14].

Table 2.3 — Recurrent or founder variants in other CLN diseases (MANE transcripts)

Sources: Saudubray, CLN3 GeneReviews, 2026 CLN3 consensus, ClinVar and VariantValidator [2, 7, 14, 17, 34].

Two technical points prevent misinterpretation. First, historical numbering of the Finnish CLN5 variant (p.Tyr392Ter) differs from the current MANE transcript, in which ClinVar describes it as NM_006493.4:c.1028_1029del; reports should state the transcript and mention both when literature uses the older designation [7, 34]. Second, the common CLN3 deletion is approximately 1 kb and may escape exome sequencing, short-read genome sequencing and chromosomal microarrays of usual resolution; GeneReviews estimates that sequence analysis identifies 20–30 % of pathogenic alleles and targeted deletion/duplication analysis 70–80 % [14].

Inheritance is autosomal recessive in all forms except CLN4: DNAJC5 variants act heterozygously, with high penetrance, and explain only some families with dominant Kufs disease [1]. GRN illustrates the importance of allele dosage: heterozygosity is a frequent cause of adult frontotemporal lobar degeneration, whereas homozygosity causes CLN11, with rapid visual loss around age 20, myoclonus, ataxia and early cognitive deterioration, and an intermediate spectrum has already been proposed [1, 2].

3. Clinical Spectrum and Manifestations by System

3.1. Classical CLN2 disease: natural history timeline

In classical late-infantile disease, development is usually normal or marked only by language delay until age 2–4. The combined DEM-CHILD (n = 74) and Weill Cornell (n = 66) cohort of untreated patients with confirmed genotypes documented a median age of 35 months (interquartile range 24.0–38.5) at first symptom, 37 months (35.0–42.0) at first seizure and 54 months (47.5–60.0) at diagnosis. In DEM-CHILD, first symptoms were seizures (70 %), language difficulties (57 %), motor difficulties (41 %), behavioural change (16 %) and dementia (9 %); because a patient could have more than one initial symptom, percentages do not total 100 % [8]. In an Australian series of 13 patients, 11 (85 %) had language delay before the first seizure, and initial seizures were focal, generalised tonic–clonic, absence or febrile [35].

From the first seizure, functional loss is rapid and predictable. In 41 DEM-CHILD patients assessed longitudinally throughout their course, the motor–language score (0–6 points) fell by a mean 1.81 points per year (95 % CI 1.50–2.12), progressing from normal to absent function in approximately 30 months; median time from first symptom to death was 7.8 years [8]. GeneReviews describes loss of ambulation between ages 4 and 5 through myoclonus and ataxia; loss of most gross motor function around age 6, when seizures become difficult to control and gastrostomy is often needed; increasing dependence between ages 7 and 10, with complete visual loss; and untreated life expectancy between 6 and 12 years [1]. Exploratory survival analysis of the historical control used by the EMA estimated median age at death at 10.4 years (95 % CI 9.5–12.5) [11]. Saudubray places blindness at age 5–6; GeneReviews places complete visual loss between ages 7 and 10. The difference reflects the criterion (onset of visual loss or complete blindness) and sample; the 2021 guideline emphasises that visual loss in CLN2 follows other manifestations and is not an early diagnostic sign [1–3].

Table 3.1 — Manifestations of classical CLN2 disease by system

3.2. Atypical CLN2 disease and SCAR7

The atypical phenotype starts between preschool age and adulthood, often presents with seizures or ataxia, cerebellar atrophy or visual loss, and progresses more slowly [1]. The international guideline describes a slower course in which behavioural and movement disorders and ataxia may predominate, sometimes without epilepsy and with preserved vision [3]. In a series of 14 treated atypical patients, median age at first symptom was 5.9 years and at diagnosis 10.8 years; initial symptoms were language abnormalities (43 %), seizures (29 %), ataxia with language abnormalities (21 %) and isolated ataxia (7 %) [40]. In the Latin American series of 36 patients, language regression dominated the classical presentation (90 %) and seizures the atypical presentation (87 %), with median onset ages of 3 and 7 years, respectively [31]. In SCAR7, ataxia occurs without epilepsy or visual abnormalities, progressing slowly into old age; the guideline recommends including TPP1 assays in the investigation of unexplained spinocerebellar ataxias [2, 3].

3.3. Other CLN diseases

Table 3.2 — Clinical spectrum of the other CLN diseases

Sources: NCL and CLN3 GeneReviews, Saudubray, Simonati 2017 (CLN5), Kayani 2024 (CLN7), KCTD7 GeneReviews [1, 2, 14, 41–43].

Several features warrant clinical emphasis. CLN1 has the earliest deterioration and most severe imaging abnormalities, with pronounced cerebral atrophy and an electroencephalogram becoming isoelectric [2]. In the DEM-CHILD series of 15 patients with CLN5, learning and cognitive difficulties were the most frequent initial symptoms, and severe variants were associated with faster decline [41]. In a cohort of 97 patients with CLN6 tested in one laboratory, 86 had late-infantile onset, eight juvenile and three adult onset [44]. CLN6 variants are the main cause of recessive Kufs type A disease, differing from late-infantile disease through much later onset and absence of visual involvement [45]. In a series of eight children with CLN7, language problems preceded or accompanied the first seizure in every case [42].

CLN3 has extraneurological involvement requiring specific surveillance. Cardiac abnormalities — sinus bradycardia, QT interval prolongation and left ventricular hypertrophy — usually become evident after age 20 but have been described from age 14, and Saudubray notes the risk of death associated with bradycardia [1, 2, 14]. Non-epileptic autonomic episodes (“paroxysmal sympathetic storms”) may occur in adolescence, with anxious behaviour and fear [14]. The 2026 consensus mentions hyperandrogenism, early menarche and signs of anovulation, with only 63 % expert agreement [17]. An atypical protracted phenotype has also been described with rapid visual loss, seizures 20 years later, hypertrophic cardiomyopathy, muscle autophagic vacuoles and mild cognitive impairment after a 40-year course [2].

The French PNDS adds semiological detail to this description. In CLN1 (onset between 3 and 18 months), regression with hypotonia, loss of head control, irritability and stereotypies is accompanied by early blindness due to optic atrophy and macular degeneration; spectroscopy first shows elevated choline with low N-acetylaspartate and, later, disappearance of all peaks, while the EEG evolves to an almost isoelectric trace (“vanishing EEG”); in CLN10 there is congenital, sometimes prenatal, microcephaly and myoclonic epileptic encephalopathy from the neonatal period [18]. In the more slowly progressive late-infantile forms (CLN5 and CLN6 with onset between 4 and 6 years; CLN7 and CLN8 between 4 and 10 years), the first complaints may be educational or motor, myoclonic epilepsy does not dominate, and loss of walking occurs around age 10 in CLN7 and CLN8; in CLN3, with onset between 4 and 9 years through visual loss, the ERG becomes extinguished and MRI shows thalamic and basal ganglia hypointensity on T1 and T2; CLN12 (ATP13A2) is associated with a levodopa-responsive parkinsonian syndrome [18]. Among the adult forms, which the PNDS defines by onset from age 12, it distinguishes Kufs disease type A (progressive myoclonic epilepsy without retinopathy; CLN4 or CLN6 loci), type B (motor and cognitive signs without epilepsy; CLN6, CLN11, CLN13) and Parry disease (CLN4, autosomal dominant); MRI is non-specific, the EEG has strong orienting value (posterior abnormalities with photic driving at about 3 Hz), plasma progranulin is reduced in CLN11 and, in the absence of a genetic diagnosis, axillary or rectal skin biopsy with electron microscopy remains useful [18]. According to the PNDS, the retinal dystrophy of NCL begins in the inner retina and is usually of rod–cone type, with night blindness (sometimes interpreted as “fear of the dark”), followed by reduced acuity and photophobia; the fundus may be normal in the early stages, so multimodal imaging (autofluorescence and spectral-domain OCT with sections outside the macula, feasible in clinic from age 3) and ISCEV-standard ERG (in clinic before age 1 and after age 4) are systematic in any child with night blindness [18].

4. Clinical Recognition and Warning Signs

Early recognition is difficult because each initial sign is common in paediatrics: language delay, a first febrile or afebrile seizure, clumsy gait. The distinguishing features of CLN2 disease are their combination and sequence. The international guideline recommends CLN2 testing for any child with significant speech delay or regression, clumsiness and epilepsy of unknown cause before age 4, emphasising detection following unprovoked seizures or unsteadiness in children with psychomotor delay or decline [3]. The Brazilian consensus gives the same rule: suspect the condition in any child aged 2–4 with language delay and epilepsy [15]. Experts analysing language in CLN2 draw attention to age at first words and first phrases and language plateauing, which may precede seizures [36].

GeneReviews extends the warning to the entire group: consider an NCL in a child with developmental delay, particularly speech and language delay, accompanied by progressive visual loss, epilepsy, ataxia or regression. A patient with epilepsy and regression or ataxia should be investigated for neurodegenerative disease because these findings, although potentially attributed to antiseizure medication adverse effects, are not common side effects [1]. The Oxford Desk Reference includes NCLs in the differential diagnosis of developmental regression, seizures with developmental delay, retinal dystrophies, nystagmus associated with seizures and regression, childhood ataxia and presentations mistaken for cerebral palsy [46–51]. In the approach to inborn errors of metabolism, late-infantile and juvenile forms are among the most frequent neurometabolic disorders with prominent seizures in childhood and adolescence, and NCLs account for approximately half of storage disorders among inborn errors causing neurological regression [37].

Table 4.1 — Warning signs and most likely CLN condition

The case in section 10 illustrates the gap between these signs and diagnosis: language delay from age 12–18 months, seizures at age 3.2 years with paroxysmal activity activated by low-frequency photic stimulation, and confirmation only at age 4.5–4.9 [53]. In a French population-based study (2017–2023), median age at diagnosis was 5 years, mean diagnostic delay from the first coded symptom was 19.1 months and 84.3 % of children were admitted through the emergency department [13].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Criteria for confirmation of CLN2 disease

The recommended gold standard is demonstration of deficient TPP1 activity (in leucocytes, fibroblasts or dried blood spots) and identification of pathogenic variants in both TPP1 alleles. When both analyses cannot be performed, demonstration of deficient TPP1 activity in leucocytes or fibroblasts, or detection of two pathogenic variants in trans, is diagnostic [3, 9]. In 2015, experts recommended TPP1 activity among first-line tests for suspected NCL, together with PPT1 activity to exclude CLN1, and encouraged epilepsy panels when investigating unexplained late-infantile seizures [9]. The 2025 GeneReview describes evolving practice: molecular testing has often become first-line, with enzyme assays used mainly afterwards to clarify the pathogenicity of variants of uncertain significance [1]. Both sequences are accepted; the key is not to let one pending result delay the other.

5.2. Enzyme assays: matrices and limitations

Fluorimetric TPP1 assays in dried blood spots were evaluated at a reference centre over 12 years: of 3,882 requests, 50 samples with subnormal activity and adequate quality had clinical and molecular data, and all 50 patients had confirmed CLN2, including seven atypical patients; the centre knew of no patient with a normal test subsequently diagnosed with CLN2 [54]. This supports positive predictive value in that series but does not provide a formal sensitivity estimate, because false negatives would be detected only if reported to the laboratory. The matrix matters: serum TPP1 exists only in the neutral form and is not deficient in CLN2 patients, so serum is unsuitable for diagnosis, unlike PPT1, which may be assayed in plasma or serum in CLN1; heterozygotes have intermediate acidic activity, preventing the assay from reliably excluding carrier status [55]. In atypical patients, activity may be absent in certain cell types but persist in leucocytes [3].

5.3. Molecular testing and reporting

Strategies include multigene panels (often first choice and more accessible), exome or genome sequencing; panels vary between laboratories in genes and methods (sequencing and deletion/duplication analysis) [1]. The report should state the gene, versioned transcript (for example TPP1 NM_000391.4 or CLN3 NM_001042432.2), DNA and protein HGVS nomenclature, zygosity, ACMG/AMP classification and technical limitations [7, 14]. Two variants of uncertain significance, or one pathogenic variant and one of uncertain significance, neither establish nor exclude diagnosis [14]. In CLN2, enzyme activity is the functional test allowing reclassification of uncertain variants; in CLN3, where there is no enzyme to assay, quantification of lymphocyte vacuolation and intracellular CD107a flow cytometry serve this purpose, with the proviso that isolated retinal disease may have normal values [14]. Confirmation that two variants are in trans requires parental testing; apparent homozygosity may reflect deletion of one or more exons on the other allele or uniparental isodisomy [1].

Table 5.1 — NCL tests: what they detect and do not detect

5.4. Neurophysiology, imaging and ophthalmology

A photoparoxysmal response to low-frequency intermittent photic stimulation (1–2 Hz), with eyes open, is considered characteristic of classical CLN2, particularly with speech delay or ataxia; standard protocols use higher frequencies and may miss it [2, 3]. Giant somatosensory evoked potentials occur in some progressive myoclonic epilepsies, including late-infantile disease; in advanced stages, EEG amplitude decreases and cortical visual evoked potentials increase [3, 37]. Specificity and sensitivity of this sign are not quantified in the sources consulted, and an Argentine series of 19 children with genetically confirmed NCL — six with CLN2 — found no photoparoxysmal response in any patient, so its absence does not exclude diagnosis [52]. In CLN1, the first electroencephalographic sign is loss of reactivity to eye opening and closure, followed by loss of sleep spindles and a rapidly isoelectric trace [2, 37].

Magnetic resonance imaging shows early cerebellar atrophy in CLN2, sometimes preceding subtle posterior cortical atrophy; central review of the Australian series found cerebellar or cerebral atrophy in every patient [3, 35]. Cerebral atrophy is particularly severe in CLN1 [2]. The electroretinogram is generally reduced at presentation and rapidly extinguishes; full-field electroretinography may help when genotyping is unavailable, and optical coherence tomography quantifies macular degeneration [2, 3]. In CLN3, the initial pattern is an electronegative rod electroretinogram, with early loss of the outer nuclear layer and ellipsoid zone [14].

5.5. Newborn and selective screening

The 2021 guideline considers that early childhood diagnosis would benefit from newborn screening (85 % agreement, evidence level D) [3]. A Japanese pilot study has assayed PPT1 and TPP1 in more than 1,000 newborn dried blood spots [55], and a Turkish selective screening study in children aged 2–6 with seizures and non-specific neurological findings is registered on ClinicalTrials.gov (NCT06128226; 750 participants planned) [12]. In the United Kingdom, according to the 2026 NICE appraisal, there is no national routine CLN2 screening programme, so diagnosis occurs only after symptoms unless there is an affected sibling; the Generation Study research project is ongoing, with no certainty that it will lead to routine genomic newborn screening, work is underway to introduce a dried blood spot newborn screening method into the NHS, and the 10-year health plan for England states an ambition to make genomic sequencing universal at birth [56]. Clinical experts consulted by the committee reported that increased awareness has brought diagnoses forward — one had seen no diagnosis with a motor–language score below 5 in the previous two years, whereas another expert's centre data showed 28.5 % of children starting with score 6, 28.5 % with 5 and 42 % with 4 — but that without newborn screening diagnoses at more advanced stages will continue [56]. Inclusion of CLN2 in a national newborn screening panel was not found in the sources consulted; this absence reflects the sources and does not prove non-existence.

5.6. Practical algorithm

1. Child aged 2–4 with language delay and an unprovoked seizure, or regression and ataxia: request TPP1 and PPT1 simultaneously (dried blood spots or leucocytes) and an epilepsy/NCL panel; EEG with 1–2 Hz photic stimulation; magnetic resonance imaging [3, 9].
2. Deficient TPP1: CLN2 diagnosis; complete genotyping (counselling, prenatal diagnosis, atypical phenotype) and refer immediately to a centre experienced in NCL [3].
3. Panel with two pathogenic TPP1 variants: diagnosis; enzyme assay if a variant of uncertain significance is present [1].
4. Child aged 4–7 with rapid visual loss: ophthalmological assessment with electroretinography, smear for vacuolation and molecular testing with targeted CLN3 deletion analysis [14, 17].
5. Negative panel with convincing phenotype: reanalyse using exome or genome sequencing, request CNV analysis and electron microscopy and store DNA; a negative test does not exclude NCL [1, 46].

In France, the biological diagnosis is made in a reference medical biology laboratory, with simultaneous assay of PPT1 and TPP1 in leucocytes or dried blood spots (rarely fibroblasts), followed by sequencing of the corresponding gene, or by a high-throughput panel of the 13 NCL genes (PPT1, TPP1, CLN3, DNAJC5, CLN5, CLN6, MFSD8, CLN8, CTSD, GRN, ATP13A2, CTSF, KCTD7); progranulin is measured in plasma or cerebrospinal fluid, and cathepsin D and F assays are not routine practice; the PNDS stresses that the enzymatic and molecular diagnosis of CLN2 must be rapid so as not to delay access to cerliponase [18]. The PNDS requests brain MRI, EEG with prolonged low-frequency intermittent photic stimulation (0.5–2 Hz, to be stated on the request because it is not performed routinely) and ERG in any child with epilepsy with myoclonus, and points out that absence of photic driving does not exclude the diagnosis; the window from 2 to 4 years, with epilepsy and learning delay, is the period of greatest suspicion for CLN2 [18].

6. Differential Diagnosis

The differential depends on the dominant presentation. In regression with seizures, non-genetic causes (severe deprivation, psychiatric disease, tumours, infections) and poorly controlled epilepsies in which regression resolves with seizure control should be considered before concluding neurodegeneration [46]. Progressive myoclonic epilepsy is defined by the triad of stimulus-sensitive myoclonus, generalised seizures and neurological deterioration; NCLs share this presentation with Unverricht–Lundborg disease, Lafora disease, sialidosis, myoclonic epilepsy with ragged-red fibres, dentatorubral-pallidoluysian atrophy, Gaucher disease type 3, spinal muscular atrophy with progressive myoclonic epilepsy due to acid ceramidase deficiency and LIMP-2 deficiency [37, 47].

Table 6.1 — Differential diagnosis

In adults, the French PNDS centres the differential diagnosis on young-adult dementias without leukodystrophy or focal lesions: inflammatory or autoimmune causes (dysimmune encephalitis, neurosarcoidosis, lupus), metabolic causes (hepatic or renal failure, hyperammonaemia), deficiency states (Wernicke encephalopathy, Korsakoff syndrome, vitamin B12 deficiency, homocystinurias), infectious causes (HIV, syphilis, Lyme disease, prions) and genetic causes (Huntington disease, Wilson disease, neuroferritinopathies, genetic forms of Alzheimer disease and frontotemporal dementia, Niemann–Pick disease type C); in children, mitochondrial diseases, other lysosomal diseases and Rett syndrome [18].

7. Treatments and Risk Management Strategies

7.1. Principles and organisation of care

All children with suspected CLN2 disease should be referred to a centre experienced in NCL; the first consultation should be conducted as soon as possible after diagnosis by a physician experienced in treating the disease, and care should be coordinated by a paediatric neurologist or rare disease specialist supported by a local multidisciplinary team (89–96 % agreement with the guideline statements) [3]. Management follows paediatric palliative care principles: initially, the priority is preserving function; as disease progresses, the goal shifts towards quality of life for the child and family [3, 57]. Psychological support, contact with patient associations, genetic counselling and bereavement support form part of the plan from diagnosis [3]. In the 2026 NICE appraisal, a patient organisation described most parents as full-time carers needing financial support, home adaptations, social and mental health support and personal assistance, with physical and psychological consequences for parents and unaffected siblings, and families reporting isolation and anticipatory grief; experts added that supportive care varies between regions [56].

7.2. Cerliponase alfa (Brineura®): mechanism, formulation and dosing

Cerliponase alfa is recombinant human TPP1 produced in Chinese hamster ovary cells: 544 amino acids in proenzyme form (approximately 59 kDa), 368 in the mature enzyme, with five N-glycosylation sites including phosphorylated mannose, allowing uptake through the mannose-6-phosphate receptor [10]. As it does not cross the blood–brain barrier in useful quantities, it is administered directly into cerebrospinal fluid through a surgically implanted intraventricular reservoir and catheter. Pharmacokinetically, plasma exposure was 300 to 1,000 times lower than cerebrospinal fluid exposure, without accumulation on fortnightly dosing and without a relationship between the magnitude of cerebrospinal fluid exposure and efficacy, supporting the 300 mg dose [58].

The formulation is a 150 mg/5 mL (30 mg/mL) solution for infusion, packaged with two cerliponase alfa vials and one 5 mL vial of intraventricular electrolyte solution (in the EU, “flushing solution”) [10, 11]. Cerliponase infusion is followed by electrolyte solution infusion at 2.5 mL/h (1.25 mL/h below 6 months in the US label), with a total duration of 2 to 4.5 hours depending on dose and volume [10, 11].

Table 7.1 — Cerliponase alfa dose by age (identical in US and EU labels)

Sources: FDA prescribing information (revised July 2024), EMA summary of product characteristics and Saudubray drug table [10, 11, 59]. The maximum approved dose is 300 mg every 2 weeks, from age 2 and, between ages 1 and 2, from the fifth dose onwards [10, 11].

Administration conditions appear in both labels. Premedication with an antihistamine, with or without an antipyretic (and, in the US label, with or without a corticosteroid), is recommended 30 to 60 minutes before infusion. Before each infusion, the scalp is inspected for signs of device leakage, failure or infection and patency is checked; the FDA label requires cerebrospinal fluid sampling for cell count and culture before each infusion, and the EMA label requires routine sample submission to detect subclinical infections. Vital signs are monitored before, during and after infusion; ECG should be monitored during infusion in patients with bradycardia, conduction disorders or structural heart disease, and other patients should have a 12-lead ECG every 6 months. The first dose should be administered at least 5 to 7 days after implantation (FDA), and the reservoir should be replaced before 4 years of punctures because material degradation has been observed [10, 11]. If a dose is missed, treatment resumes as soon as possible, subsequently maintaining the 2-week interval [10].

There are relevant jurisdictional differences. The European label allows a 50 % dose reduction and/or a lower infusion rate in patients unable to tolerate infusion and recommends stopping or slowing infusion for signs of intracranial hypertension (headache, nausea, vomiting, reduced consciousness), particularly below age 3; the US label allows re-administration after anaphylaxis at approximately half the rate at which it occurred [10, 11]. In the USA, the medicine is not recommended below 37 weeks' postmenstrual age or 2.5 kg; in the EU, dosing below age 2 was estimated from brain mass and there are no clinical data below age 1, so decisions should be based on individual benefit–risk balance, explicitly indicating treatment should start as early as possible [10, 11]. Contraindications also differ: the USA contraindicates the medicine with local or central nervous system infection, acute device complications or a ventriculoperitoneal shunt; the EU contraindicates it with life-threatening anaphylaxis if rechallenge has failed, a ventriculoperitoneal shunt and acute signs of device leakage, failure or infection, and advises caution in obstructive hydrocephalus [10, 11].

7.3. Regulatory status and access

Table 7.2 — Regulatory status of cerliponase alfa

Sources: [1, 3, 10, 11].

Authorisation does not guarantee access. In Brazil, cerliponase alfa has been approved since 2018, but the Brazilian consensus describes limited access to molecular testing, dependent on industry support, and public health system difficulties in diagnosing and treating rare diseases [15]. In an Iranian series of 20 patients, none received enzyme replacement therapy for financial reasons [32]. In England, National Institute for Health and Care Excellence (NICE) highly specialised technologies guidance HST34, published on 18 February 2026 and replacing HST12, does not recommend cerliponase alfa within its marketing authorisation for treating CLN2 disease [56]. This recommendation does not affect treatment funded through the managed access agreement (MAA) before publication of final guidance: children already receiving treatment, or starting before the managed access period ends, may continue until they and the English National Health Service (NHS) professional consider stopping appropriate, through a joint decision by the professional, child or young person and parents or carers [56]. HST12 recommended cerliponase alfa only under MAA conditions, including a data collection agreement to reduce clinical uncertainty; the medicine was available through that route but was never routinely available in the NHS [56].

The committee concluded that data from studies 190-201/202 and 190-203 and the MAA cohort, encompassing all eligible patients in England and compared with matched patients from retrospective natural history study 190-901, show that cerliponase alfa slows progression, and recognised it as transformative and innovative; residual uncertainty concerned long-term effects, variability of benefit and non-neurological outcomes (section 7.4) [56]. The comparator was supportive care, including medicines for seizures, dystonia and myoclonus, speech and language therapy and physiotherapy [56]. The company's stated list price is £20,107 per 300 mg pack (two 150 mg vials), excluding VAT, corresponding to £522,782 per patient per year at 300 mg every 2 weeks above age 2; the commercial discount and all cost-effectiveness results are confidential [56]. For these technologies, an incremental cost-effectiveness ratio (ICER) below £100,000 per quality-adjusted life year (QALY) is normally considered an effective use of resources, with a weight of 1 to 3 potentially applied when gains are 10 to 30 QALYs [56]. The committee accepted greater uncertainty because this was a rare disease, a predominantly paediatric population and an innovative, complex technology, applied the weight corresponding to its assumptions (confidential, below the maximum) plus 0.2 for uncertainty, and still found the most likely estimates with the discount substantially above the acceptable range; they would also remain above it under the additional MAA commercial arrangement, which, being time-limited, was not in force during appraisal [56].

Table 7.2b — Main assumptions discussed in NICE appraisal HST34

Source: NICE HST34, sections 3.5–3.20 [56].

The committee considered stage-based starting rules and rejected them because they would exclude children diagnosed late, deciding to consider all newly diagnosed patients regardless of score [56]. Consultation objections — indirect age and disability discrimination, and inequity between children, including siblings, according to whether diagnosis preceded or followed the end of managed access — were recorded; the committee considered access under the agreement distinct from routine funding and stated that when a managed access treatment is not recommended at final review, it ceases to be available in the NHS for those who have not yet started [56]. Consequently, the same molecule authorised in the United States and European Union (Table 7.2) is not routinely available through the NHS in England for children diagnosed after managed access ends; this is an economic appraisal decision, questioning neither efficacy nor authorisation, with a rationale — price, threshold and model assumptions — specific to that system [10, 11, 56]. No information on access in Portugal was found in the sources consulted.

7.4. Efficacy evidence

Table 7.3 — Cerliponase alfa studies

Study design affects interpretation. There was no randomised contemporaneous control group: the comparator was a historical cohort, with risk of unmeasured differences in selection, supportive care and assessment, which matching and adjustment for age, baseline score and genotype mitigate but do not eliminate [10, 60]. The FDA considered comparability between the study and natural history cohort impossible to establish for the language domain, so the approved effect is restricted to loss of ambulation [10]. A decline rate of 0.27–0.46 points per 48 weeks during treatment means slowing rather than universal stabilisation; in the extension, 21 of 24 patients (88 %) experienced at least one serious adverse event [27, 60, 61]. A single-arm meta-analysis of three observational studies (55 patients) found significant heterogeneity and cannot replace controlled trials [63]. Cerliponase does not prevent visual loss: children treated intracerebroventricularly continue losing vision, prompting investigation of intravitreal delivery [64, 65]. In a prospective neuroimaging study of 27 treated patients (170 assessments), mean cortical thickness decreased by 4.2 % annually, compared with 12.5 % in natural history controls, and reduction correlated with clinical scales [66]. The 2026 NICE appraisal, including data from patients treated in the NHS in England under the managed access agreement, reached the same interpretation: the external assessment group considered slowing of progression conclusively demonstrated but noted lack of data beyond 6 years' follow-up, likely dependence of benefit on age and stage at initiation, scarcity of non-neurological outcome data and some evidence of seizure prevention and reduced severity, with uncertain quality-of-life impact; agreement data may have underestimated benefit because of delays in diagnosis, treatment and access to other interventions during the COVID-19 pandemic [56]. According to the company, none of the children younger than 3 with baseline score 6 in study 190-203 changed score during 6 years' follow-up; clinical experts considered that only presymptomatic children would be likely to retain that score for 6 years, and that observed improvements during treatment mostly reflect recovery after intercurrent illness in children near stage boundaries, with substantial motor and language improvement unlikely [56]. Regarding vision, the committee received no evidence that intracerebroventricular cerliponase delays visual loss — which, according to consulted experts, starts around age 5 and is nearly complete around age 9, although some children lose vision more slowly — and noted that intravitreal administration is outside the authorisation [56]. The guideline recommends initiating long-term therapy in classical patients with potential benefit and suggests it in non-classical patients following confirmed diagnosis and family–team agreement (level C, 84 % agreement) [3].

7.5. Safety and adverse effect management

Table 7.4 — Adverse reactions in studies 1 and 2 through 96 weeks (n = 24) and study 3 (n = 14)

Source: FDA prescribing information [10].

Hypersensitivity reactions occurred during infusion or in the following 24 hours despite premedication, predominantly as fever with vomiting, pleocytosis or irritability, a pattern not corresponding to classical immune-mediated hypersensitivity; they resolved with time or antipyretics, antihistamines or corticosteroids, without treatment withdrawal [10]. Postmarketing anaphylaxis cases treated with adrenaline have been reported, as well as one anaphylactic reaction in a child younger than 3 in study 3; among 38 patients considered in the European summary of product characteristics, hypersensitivity occurred in 50 %, severe (grade 3) in six [10, 11]. Antidrug antibodies were detected in cerebrospinal fluid in 25 % and serum in 79 % of patients, without neutralisation, association with hypersensitivity or loss of response [3]. Management includes reducing the rate, temporarily interrupting infusion and administering antihistamines or antipyretics; after anaphylaxis, re-administration requires a resuscitation team and facilities and an initial rate reduced by approximately half [10].

Device-associated infections are the main route-specific risk. In trials, nine participants developed intracerebroventricular device-related infections, with nine episodes in six patients requiring device replacement [60]. Bacterial meningitis requiring antibiotics and device removal has been reported postmarketing [10]. At one US centre, 11 of 16 patients (69 %) had at least one positive culture and 64 of 1,401 cultures (4.56 %) were positive, predominantly Cutibacterium acnes in liquid medium only; the authors opted to observe asymptomatic patients with low-virulence skin flora without intervention, an institutional practice differing from trial management (antibiotics and device replacement) and not a label recommendation [11, 67]. Analysis of spontaneous reports in the European EudraVigilance database confirmed the known profile — fever, device infection, vomiting, seizures, pleocytosis, irritability, ventriculitis and respiratory abnormalities [68].

Device implantation requires general anaesthesia by an experienced paediatric neurosurgeon in a group with increased anaesthetic risk from muscle atrophy, seizures, upper airway obstruction, hypothermia and cardiac abnormalities [3]. To reduce sedation and access problems, some centres connect the ventricular catheter to a subcutaneous chest port: in a British series of 17 children, nurses considered access and fixation easier, access success was higher and device survival and complications were similar to cranial devices; a US series of seven patients and a 2026 technical note describe the technique [69–71]. This configuration differs from reservoirs specified in the FDA label and should be considered specialist-centre practice [10].

There are no pregnancy or lactation data; animal reproduction studies have not been performed [10]. The guideline does not define universal starting and stopping criteria because they vary between health systems; it suggests stopping for another life-limiting disease or severe uncontrollable infusion reactions and recommends agreeing individual goals and reassessment criteria with families before starting [3]. In the English managed access agreement, discontinuation depended on several criteria, including age at initiation, treatment duration, motor–language score decline persisting for 3 or more infusions and carer-reported reduced quality of life; the NICE committee considered that, in practice, treatment should not stop solely because the score reaches 0, but continue until the family, carers and NHS professional agree stopping is appropriate, and that if cerliponase is used as a palliative option at that stage, other options should be preferred [56]. Experts and patient representatives emphasised that stopping decisions should not be based solely on motor–language score and are individual [56]. Haematopoietic stem cell transplantation has no demonstrated benefit in CLN2 and is not recommended [3].

The French PNDS describes infusion practice in France, where cerliponase alfa has been available since 2017: 300 mg as a 4-hour infusion every 15 days in a day hospital, with lower doses below age 2, in an experienced centre; the vials, which are frozen, are left at room temperature for at least 1 hour, are neither shaken nor diluted, and the infusion is given at 2.5 mL/h through a 0.2 µm filter, with a pressure alarm at 250 mmHg and a final flush with 4 mL of intraventricular electrolyte solution; before each infusion, 0.5–1 mL of cerebrospinal fluid is withdrawn through the reservoir (biochemistry, cytology, culture and, if needed, 16S and 18S RNA PCR) and is not reinjected; the first infusions are followed by 4 hours of observation [18]. The PNDS sample protocol includes 2 hours of fasting, topical anaesthetic cream and premedication (antihistamine, antipyretic and, in the example of a 15 kg child, rectal nalbuphine and melatonin), and grades reactions as mild (halve the infusion rate and restore it after 15–30 minutes), moderate (stop, give an intravenous antihistamine, and resume at half the rate if symptoms resolve) and severe (immediate discontinuation, antihistamine, methylprednisolone 1 mg/kg, nebulised salbutamol, intravenous adrenaline 0.01 mg/kg), with blood count, CRP, total IgE and serum for antibodies collected within the first hour; delayed reactions without an infectious cause justify premedication with a systemic corticosteroid 1 hour beforehand [18]. Repeated infusions in themselves increase cerebrospinal fluid leucocytes and protein, which complicates the diagnosis of reservoir infection, sometimes insidious; when infection is suspected, the PNDS advocates empirical antibiotic therapy targeting staphylococci and proposes a fever protocol: without a focus, blood tests (blood count, CRP, procalcitonin, electrolytes, coagulation, blood cultures, urine dipstick, viral PCR), retrieval of the latest cerebrospinal fluid culture and no immediate antibiotic; with meningeal signs, behavioural change, an unusual seizure, altered consciousness or an inflammatory syndrome, reservoir puncture by an experienced physician and, if organisms, meningeal signs or an inflammatory syndrome are present, cefotaxime 50 mg/kg every 6 hours and vancomycin 15 mg/kg every 6 hours [18]. The vancomycin dose in this protocol (60 mg/kg/day) exceeds the usual paediatric dose in the US label (10 mg/kg every 6 hours, 40 mg/kg/day), which serves only as a comparison [72]. Regarding indications, the PNDS recommends starting treatment as early as possible, does not recommend it once the disease has progressed to profound multiple disability, notably after loss of walking, and calls for regular multidisciplinary review of disease progression, infusion tolerance (catheter secured by a bulky dressing, 4 hours of immobility, fortnightly travel) and reservoir complications (infection, displacement, valve damage with fluid leakage, calcifications) [18].

7.6. Intravitreal retinopathy treatment (off-label use)

Intravitreal cerliponase alfa is not included in any label and is therefore off-label or investigational. A prospective controlled compassionate-use study showed safety of 0.2 mg intravitreally every 8 weeks and reduced macular volume loss in patients still in the degenerative phase [38]. In a child homozygous for a variant associated with severe retinopathy, treated intracerebroventricularly from 40 months and with intravitreal injections every 4 weeks in one eye from 60 months, visual acuity in the treated eye was 0.2 after 24 months, versus less than 0.02 in the untreated eye, which developed end-stage retinopathy; no intraocular adverse effects occurred [65]. A US centre described a pathway for same-day intraventricular infusion and intravitreal injections (0.2 mg in 0.05 mL artificial cerebrospinal fluid every 4 weeks, under anaesthesia), using vial overfill [64]. Randomised masked phase 1/2 trial NCT05152914 (24–72 months, five participants planned) completed primary outcome collection in March 2026, without results posted in the registry at access [12]. These doses come from series and an institutional protocol rather than a label.

7.7. Symptomatic treatment common to NCLs

Epilepsy. Seizures are polymorphic and become drug-resistant; complete seizure freedom is often unrealistic in advanced disease but may be possible initially [1, 3]. Broad-spectrum antiseizure medicines are usually preferred — valproate, benzodiazepines (clobazam, clonazepam), lamotrigine, zonisamide, levetiracetam or brivaracetam — without data favouring any one [1]. Sodium channel blockers (phenytoin, fosphenytoin, carbamazepine, oxcarbazepine and also lamotrigine) should be used cautiously because they may aggravate myoclonic and generalised seizures; carbamazepine and phenytoin should be avoided in CLN2 [1–3]. There is an apparent divergence: Saudubray considers lamotrigine usually effective, and GeneReviews includes it both among preferred medicines and sodium blockers requiring caution; in practice, choice should be individual and reassessed if myoclonus worsens [1, 2]. Topiramate may worsen language difficulties; valproate, the most commonly used, has rarely been associated in late stages with worsening dystonia, hyperthermia and hyperCKaemia; cannabidiol has been described as better tolerated than tetrahydrocannabinol in NCL; the ketogenic diet may be effective for refractory seizures, with monitoring for constipation, renal stones and growth impairment [1, 3]. An emergency plan for home and school should include buccal or intranasal midazolam, diazepam or lorazepam, and polypharmacy (10–12 medicines in some children) requires attention to interactions [3]. These medicines follow standard paediatric dosing references, not reproduced here because doses are absent from disease-specific sources.

Movement disorders and pain. Movement disorders include non-epileptic myoclonus, dystonia, spasticity, chorea, athetosis and tremor; pain should first be treated or excluded because it worsens motor manifestations, and sources include constipation, reflux, urinary retention, skin lesions, dental problems, hip dislocation and osteoporotic fractures [3]. The guideline cites benzodiazepines, phenobarbital, levetiracetam, pregabalin, valproate, lamotrigine and zonisamide; baclofen, tizanidine and diazepam relieve spasticity and, with trihexyphenidyl and levodopa/carbidopa, may improve dystonia; botulinum toxin is an option for focal spasticity and dystonia; transdermal fentanyl has shown some efficacy for central pain [1, 3]. Positioning, weighted blankets, physiotherapy and warmth help [3].

Nutrition, secretions and respiration. Tube feeding should be considered with increased choking risk, inability to meet nutritional requirements, videofluoroscopically confirmed silent aspiration or repeated aspiration pneumonias; gastrostomy is used in late stages [2, 3]. Sialorrhoea may require stepped anticholinergic therapy or botulinum toxin to the salivary glands. Complete vaccination of children and family members, airway clearance (mucolytics, high-frequency chest oscillation, mechanical insufflation–exsufflation, bronchodilators) and regular mobilisation are recommended [3].

Sleep, behaviour, communication and schooling. Most children with CLN2 have sleep disturbance; polysomnography is warranted for snoring, and melatonin is used with still controversial efficacy [3]. Hearing is preserved, so schooling and social interaction should be maintained; augmentative communication (objects of reference, gestures) should start early and a speech and language therapist should be involved for every patient [3]. Palliative care should begin as early as possible, with end-of-life planning, prevention of pressure ulcers and aspiration pneumonia and bereavement support [1, 3].

Table 7.7a — Symptomatic treatment according to the French PNDS (2022) [18]

7.8. Features specific to other CLN diseases

No disease-modifying therapy is approved for any other NCL; cerliponase alfa replaces TPP1 and has no indication in CLN1, CLN3 or other forms [1, 10]. The international CLN1 consensus prioritises a family-centred multidisciplinary model and considers phenotype identification (infantile, late-infantile, juvenile or adult) essential to anticipate course and needs [16]. In CLN3, the 2026 Delphi consensus (53 statements, 11 domains, AGREE II score 6.4 out of 7) recommends referral to an NCL-experienced centre, a multidisciplinary team and contact with associations; it considers generalised tonic–clonic seizures most frequent and reasonably controlled in most patients with one or two antiseizure medicines, without progressive drug resistance; it notes evidence for dopaminergic medicines in parkinsonism and warns that antidepressants and antipsychotics, useful at low or moderate doses for behavioural symptoms, worsen parkinsonism [17]. The CLN3 GeneReview mentions risperidone and pipamperone (the latter not marketed in the USA) for neurobehavioural manifestations and reduced environmental stress as an effective measure [14].

Several interventions have been tested without benefit: cysteamine bitartrate combined with N-acetylcysteine in CLN1 did not significantly alter disease course; mycophenolate mofetil in a placebo-controlled CLN3 trial showed no clear clinical benefit; neural stem cell transplantation provided no relevant benefit [2, 12].

7.9. Ongoing trials and updates from the last 12 months

Table 7.5 — Registered research programmes (ClinicalTrials.gov status on 6 October 2026)

Source: [12].

The evidence hierarchy must be respected: cerliponase alfa studies were published in peer-reviewed journals but without randomised controls; gene therapy results in other CLN diseases, where available, take the form of registry data or announcements, without a peer-reviewed publication identified in the sources consulted; preclinical programmes are not clinical options [12, 28]. Developments identified in the last 12 months were: the CLN3 GeneReview (September 2026) and CLN3 Delphi consensus (March 2026); the French national incidence study; the neuroimaging study during cerliponase treatment; chest port series; NICE HST34 guidance (February 2026), not recommending cerliponase alfa for new NHS patients in England after managed access ends, a health technology appraisal rather than regulatory decision; registration of a new CLN6 gene therapy trial; and withdrawal of the phase 3 PLX-200 trial in CLN3 [12–14, 17, 56, 66, 69, 71]. No new authorisation or indication extension was identified in the FDA and EMA labels consulted, and absence of other data in this search, which excluded general search engines because of access limitations, does not prove non-existence.

7.10. French PNDS and other guidance: divergences

Table 7.10 — Divergences between the French PNDS (2022) and the other guidance cited [1, 3, 10, 18]

8. Reproductive Options and Family Counselling

In autosomal recessive forms, parents of a patient are considered heterozygotes; molecular confirmation of both parents' status is recommended to validate variant phase and accurately calculate recurrence risk. If the variant is found in only one parent, with biological parentage confirmed, consider a de novo variant in the patient or parental postzygotic mosaicism; if the patient appears homozygous, exclude a deletion of one or more exons undetected by sequencing and uniparental isodisomy [1]. At conception, each sibling has a 25 % probability of being affected, 50 % of being an asymptomatic carrier and 25 % of being unaffected and not a carrier; carriers do not develop the disease [1]. Children of an affected individual — not described in classical or protracted CLN3 but possible in isolated retinal disease — will be obligate carriers unless the other parent is also a carrier [1, 14]. Each sibling of the parents has a 50 % probability of being a carrier [1]. Intrafamilial variability exists even between siblings homozygous for the common CLN3 deletion and is greater in mild forms [14].

In autosomal dominant CLN4, most patients have an affected parent; each child has a 50 % probability of inheriting the DNAJC5 variant, penetrance is high and, if the variant is not found in parental leucocyte DNA, sibling recurrence risk is estimated at approximately 1 % because of possible germline mosaicism [1].

Carrier testing in relatives requires known familial variants. It should be offered to reproductive partners of carriers, particularly with shared ancestry, because founder variants occur in CLN3, CLN5, CLN8 and PPT1 in Finland, CLN6 in Costa Rica, CLN6 and TPP1 in Newfoundland and/or Labrador and MFSD8 in the Roma population [1]. Enzyme assays do not replace molecular testing for CLN2 carrier detection because heterozygotes have intermediate activity [55]. The case in section 10 shows that rare recessive diseases also occur in non-consanguineous couples and that carrier screening may have value beyond founder populations [53].

Once variants are identified, prenatal diagnosis and preimplantation genetic testing (PGT-M) are possible. CLN2 prenatal diagnosis was historically performed using electron microscopy of foetal lymphocytes and, from the early 2000s, enzyme and molecular analysis of amniocytes and chorionic villi [3]; Saudubray notes availability of enzyme and/or molecular diagnosis and PGT in some countries [2]. The ideal time for risk assessment and discussion of these options is before pregnancy, and counselling should appropriately be offered to young adults at risk of carrier status [1]. Views on these techniques vary between professionals and families, and the decision is personal [1].

Testing apparently healthy siblings warrants particular attention. The Oxford Desk Reference notes that presymptomatic childhood testing requires specialist psychological support and discussion of consent, but these issues become simpler when a therapeutic intervention exists [46]. In CLN2, availability of cerliponase alfa and presymptomatic treatment data in siblings make early sibling testing a decision with direct therapeutic consequences, although treatment before symptoms remains outside the main study populations and should be discussed at a specialist centre [11, 62]. Storage of DNA and, if possible, a patient fibroblast line is recommended, particularly while diagnosis remains unresolved [46].

The French PNDS adds that prenatal diagnosis is performed by testing for the parental variants in chorionic villi (10–12 weeks of gestation) or fresh amniocytes (15–16 weeks), once segregation of the alleles in both parents has been confirmed, and that preimplantation genetic testing, possible for all NCLs, involves long waiting times and a success rate of about 30%; in the adult forms, siblings are tested in a presymptomatic genetics clinic, and in the dominant form an apparently healthy child is tested only on reaching adulthood, unless signs appear; carrier testing of the partner of a carrier is mainly of interest in consanguineous couples [18].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Frequency: what the numbers measure

NCLs collectively are the most frequent inherited childhood neurodegenerative disorders, with estimated incidences ranging from 0.6 to 14 per 100,000 live births depending on population and criterion (clinical or genetic) [1, 73]. The numbers are not interchangeable. The CLN2 guideline cites 6–8 cases per 100,000 live births but cautions that incidence and prevalence are poorly documented and geographically variable [3]. The French population-based study, using 2017–2023 hospital data, identified 51 newly diagnosed children with CLN2 over 7 years, or 0.99 per 100,000 live births; it measured coded diagnoses rather than incidence at birth, and 13 children (25.5 %) had died, at a median age of 9 years [13]. For the United Kingdom, the NICE appraisal states that exact prevalence and incidence are unknown and estimates approximately 3 to 6 diagnoses annually and 30 to 50 living children with the disease, presented as estimates without an explicit population denominator [56]. Orphanet records worldwide CLN2 point prevalence in the less than 1 per 1,000,000 category (mean value 0.07 per 100,000), and validated NCL group values of 2.0 per 100,000 in Europe (point prevalence), 4.0 per 100,000 in Norway (birth prevalence) and, in Italy, 0.12 per 100,000 (point prevalence), 0.98 per 100,000 per year (incidence) and 0.56 per 100,000 (birth prevalence) [4]. Because untreated survival is short, point prevalence is much lower than incidence at birth.

Using ExAC data (approximately 61,000 exomes), carrier frequency depended on ancestry, with the highest value (1 in 75) for PPT1 in Finland; adjusted for US ethnic diversity, PPT1, TPP1 and CLN3 carrier frequencies were the highest among NCLs, approximately 1 in 500 each, consistent with estimates based on living patient counts [73]. The same study identified variants annotated as pathogenic in public databases but with frequencies incompatible with clinical data, a relevant warning for panel interpretation [73]. The Oxford Desk Reference gives an overall juvenile Batten disease birth rate of approximately 1 in 20,000 (heterogeneity: 21 % CLN1, 7 % CLN2 and 72 % CLN3) and approximately 1 in 30,000 for CLN3, corresponding to a carrier frequency of approximately 1 in 85 [74]; GeneReviews cites an incidence of approximately 1 in 100,000 for classical disease, ranging from 0.02 per 100,000 in the Czech Republic to 4.8 per 100,000 in Finland [14].

The French PNDS estimates the overall incidence of NCL in Europe and the United States at between 1 in 12,500 and 1 in 100,000 births and identifies late-infantile CLN2 as the most frequent form in France [18].

9.2. Geographical distribution and founder effects

Saudubray describes late-infantile disease as most common in southern Europe and juvenile disease as most common in Anglo-Saxon countries [2]. In Finland, CLN1 incidence is 1 in 20,000, and founder variants occur in PPT1 (p.Arg122Trp), CLN3 (1 kb deletion, 70 % of alleles and carrier frequency 1 in 70), CLN5 (Fin major) and CLN8 (Northern epilepsy, p.Arg24Gly) [1, 2, 14]. In CLN2, the two most frequent variants concentrate in Europe and North America; c.851G>T is a Newfoundland founder variant, c.1525C>T predominates in the Middle East, c.827A>T has been recorded only in Argentina and Chile, and c.640C>T, p.(Gln214Ter), only in China and Italy [6]. CLN6 is more frequent in patients from southern Europe, the Indian subcontinent and South America and has founders in Costa Rica and Newfoundland; CLN7 was initially described in Turkish patients and late-infantile CLN8 in consanguineous Turkish families; the CLN6 variant c.257A>G occurred in five of 23 unrelated Egyptian patients [1, 2, 44]. In Iran, 20 patients from 18 consanguineous families revealed a founder effect for TPP1 c.729C>G, with a common ancestor estimated approximately 450 years ago [32].

In an Italian cohort of 152 children with molecularly diagnosed NCL, median age at onset was 4 years and at death 17 years; earlier onset was associated with shorter survival (6 years), and lysosomal enzyme deficiency forms had earlier onset and shorter survival [75].

9.3. Portugal, Brazil and Latin America

The TPP1 variant update records c.887-10A>G, associated with an atypical juvenile phenotype, only in patients from Argentina, Chile, Colombia, Portugal and Spain [6]. No Portuguese incidence data were found in the sources consulted. In Brazil, the consensus of nine paediatric neurologists (92 questions voted on) emphasises clinical suspicion at age 2–4, limited access to molecular diagnosis and approval of cerliponase alfa since 2018 [15]. The largest treated Latin American series (36 patients from five countries) had a higher proportion of atypical phenotypes than the literature, a high proportion of intronic variants and long intervals to diagnosis — median 4 years between symptoms and treatment in classical forms and 7.5 years in atypical forms [31]. In an Argentine series of 19 children, CLN7 was the most frequent subtype (36.8 %), followed by CLN2 (31.6 %) and CLN5 (10.5 %), contrasting with other series and reinforcing the need for local data [52]. These hospital series do not measure population prevalence.

9.4. Reference networks and patient associations

In Europe, NCLs fall within the European reference network for hereditary metabolic disorders MetabERN, whose lysosomal storage disease subnetwork covers the group, although the network portal does not explicitly name NCLs [19]. European rare epilepsy centres publishing CLN2 data identify themselves as ERN EpiCARE members [13, 76]. Associations and registries cited by GeneReviews include the BDSRA Foundation (United States, with centres of excellence), Batten Disease Family Association (United Kingdom), BDSRA Australia, BDSRA Canada, Beat Batten (Netherlands, predominantly CLN3), Courageous Parents Network, the BDSRA natural history registry and international DEM-CHILD registry (NCT04613089); an Italian NCL registry also exists (NCT06844877) [1, 12, 14].

In France, patients with NCL qualify for long-term condition (ALD) no. 17, with full coverage of disease-related care, and the PNDS details the social pathway: a personalised schooling plan with a support assistant, an individualised reception plan for seizures, early medico-social action centres before age 6, medico-social establishments (IME, IEM, EEAP) and home-based special education services, parental leave and allowances (up to 28 months of parental presence leave), education and disability allowances, recognition of disabled worker status and residential facilities for dependent adults, transport to centres more than 150 km away with prior approval, an invalidity card with a disability rate of 80% or more, and respite stays [18]. The Vaincre les Maladies Lysosomales association has a group dedicated to NCL; transition should be anticipated with a joint paediatric and adult clinic, and palliative care, distinct from end-of-life support, should begin early, with a specific-situation form addressed to emergency departments and the emergency medical service that defines which treatments should be given or withheld [18].

10. Illustrative Clinical Case

10.1. Late diagnosis of classical CLN2 in a Polish child (Bryzik et al., 2025)

A boy born in 2010 to young, healthy, non-consanguineous parents, with a healthy older sister and half-sister and unremarkable family history, was delivered at 41 weeks with an Apgar score of 10, weight 4 kg and head circumference 33 cm; the neonatal period was complicated by congenital pneumonia [53]. Tongue protrusion had been present from birth, assessed without findings by a speech and language therapist and neurologist. Motor development was appropriate — sitting at 6 months, crawling at 7 and walking independently at 10 — but language was delayed: first words at 12 months and approximately six simple words at 18 months, without progress until neurological symptoms began. He had frequent respiratory infections, was diagnosed with asthma at age 2 and was suspected of hyperactivity by his parents [53].

At age 3.2, focal seizures with secondary generalisation developed, with imbalance, loss of consciousness, tonic extension, head deviation and oral automatisms. Computed tomography was normal. Sleep EEG after the second seizure showed central and mid-temporal paroxysmal discharges up to 580 µV, periodically generalised and activated by low-frequency photic stimulation (Figure 1 of the original paper). Valproate was started after the third seizure, followed by levetiracetam; MRI in October 2013 showed only dilatation of the occipital horn of the right lateral ventricle, and karyotyping and microdeletion testing were normal [53]. At age 3.8, subtle pyramidal lower limb paresis and incoordination were present; prolonged tonic and tonic–clonic seizures followed, with a transient four-week improvement on topiramate and upper limb tremor. Psychological assessment at age 4 showed mild intellectual disability, and MRI at age 4.1 revealed mild widening of cerebellar sulci and the cisterna magna [53].

NCL was suspected at age 4.5 because of ataxia, persistent seizures and generalised discharges with myoclonus activated by photic stimulation; fundoscopy was still normal. Leucocyte TPP1 activity was 0.61 U/mg protein/h (reference 54 ± 18.2), and at age 4.9 TPP1 sequencing showed homozygosity for the substitution introducing a termination codon at position 208, namely c.622C>T, p.(Arg208Ter) [7, 53]. Topiramate and levetiracetam were replaced by lamotrigine. At age 5, progressive cerebral atrophy, visual deterioration with optic atrophy and diffuse slow EEG activity at 1–3 Hz were present (Figure 2). Seizures became refractory, with nearly continuous myoclonus and desaturation; at 5 years 11 months, he was admitted to intensive care, where gastrostomy was performed and pharmacological coma induced without improvement. His final regimen included valproate, ethosuximide, phenobarbital, topiramate, clonazepam and haloperidol. Palliative care followed, initially inpatient and subsequently at home, with repeated admissions for respiratory failure, tracheostomy at 8 years 7 months and death at home at 9 years 3 months. The full timeline is summarised in Figure 3 of the paper [53]. His mother was offered genetic counselling and prenatal diagnosis, which she declined; his sister was not a carrier [53].

10.2. Critical interpretation of the case

The case combines warning signs described in section 4: language delay from the second year, first seizure at age 3, ataxia and activation of discharges by low-frequency photic stimulation more than a year before NCL was suspected [3, 53]. Diagnosis at age 4.5–4.9 is consistent with the median age of 54 months in international cohorts and 5 years in France [8, 13], and the course — death at 9 years 3 months — corresponds to untreated natural history [1, 11]. The authors did not treat the child with cerliponase alfa because disease was advanced when the medicine became available in their setting. Some statements in the paper's discussion diverge from primary sources and should not be reproduced as facts: the 2017 European authorisation was not restricted to children aged 3 and over, and BMN 190 is the development designation of cerliponase alfa itself, not a gene therapy [11, 12]. The case does not establish routine treatment approaches; it illustrates the therapeutic window and usefulness of TPP1 and PPT1 testing for combined language delay and epilepsy before age 4 [3, 9].

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis without and with treatment

Untreated classical CLN2 progresses from normal function to total loss of ambulation and language within approximately 30 months of onset, with death between ages 6 and 12 (estimated median 10.4 years in the European historical control) [1, 8, 11]. During cerliponase alfa treatment, none of the 23 extension patients died during a mean 272 weeks of treatment, median age at last assessment was 10.3 years (7.8–13.1) and median time to total loss of ambulation and language was not reached [11, 60]. Treatment does not reverse established loss or prevent retinal degeneration; prognosis depends primarily on function at initiation, explaining emphasis on early diagnosis and presymptomatic treatment, still under evaluation [11, 28, 62]. Atypical forms have longer survival and reported response is stabilisation in most patients, with still short follow-up [40]. In other NCLs, age at onset is the best indicator of life expectancy, and lysosomal enzyme deficiency forms have earlier onset and shorter survival on average [75]. Long-term treated life expectancy and the phenotype of children treated presymptomatically are not established [28].

11.2. Follow-up plan

Table 11.1 — Recommended follow-up

Divergence in cardiac intervals requires an explicit decision: in CLN3, abnormalities have been described from age 14, so starting surveillance at age 12 (CLN3 GeneReviews) is most conservative; the Delphi consensus accepts age 15, and the general GeneReview proposes annual ECG from age 18 for all NCLs [1, 14, 17]. For patients receiving cerliponase alfa, label requirements for semiannual ECG take precedence [10, 11]. In the NICE appraisal, the external assessment group assumed from managed access data that 3 % of patients had clinically significant cardiac abnormalities at baseline, increasing to 27 % at 3.5 years; one expert stated ECG had previously been performed at every infusion and that annual ECG reviewed by a cardiologist would probably be useful, acknowledging insufficient data, and the committee retained the summary of product characteristics frequency (ECG at every infusion with a history of bradycardia, conduction disorder or structural heart disease; semiannually with normal cardiac function) [56]. Follow-up should occur more than once annually, with a treating physician or specialist team [1].

12. Summary and Final Considerations

Neuronal ceroid lipofuscinoses are a group of at least 12 disorders with different genes, proteins and mechanisms, united by lysosomal ceroid lipofuscin accumulation and a combination of regression, epilepsy, visual loss and premature death. Current classification is genetic, and age at onset and variant type explain much of the within-gene variability [1, 2].

CLN2 disease has a recognisable presentation — language delay followed by seizures and ataxia between ages 2 and 4, with a 1–2 Hz photoparoxysmal response and cerebellar atrophy — and rapid laboratory diagnosis through TPP1 assays in dried blood spots or leucocytes, completed by TPP1 testing [3, 9]. Nevertheless, diagnosis still occurs on average around age 4.5–5, after some function has been lost [8, 13].

Cerliponase alfa, administered intracerebroventricularly every 2 weeks, slows motor and language decline compared with historical controls and appears to delay symptom onset when started presymptomatically; it does not treat retinopathy and requires an implanted device, with risks of infection and hypersensitivity [10, 11, 27, 60]. US and European labels share age-specific doses but differ in indication wording, contraindications and scope for dose adjustment [10, 11]. Authorisation does not equal access: in England, NICE (HST34, 2026) recognised efficacy but did not recommend cerliponase alfa because cost-effectiveness was far above the highly specialised technology threshold, maintaining treatment only for children who started during managed access [56].

For other CLN diseases, treatment is symptomatic and multidisciplinary, with specific CLN1 and CLN3 consensus statements and early-phase gene therapy trials [12, 16, 17]. The quality elements most consistent across sources are referral to experienced centres, careful antiseizure medicine selection, cardiac surveillance in CLN3, early palliative care and genetic counselling with access to carrier testing, prenatal diagnosis and PGT-M [1, 3, 14].

13. Bibliographic References

1. Malik K, Santucci K, Sremba L, Steenari M, Miele A, Demarest S, Whiteman I. Neuronal Ceroid Lipofuscinoses Overview. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Published on 10 Oct 2001; fully updated on 29 May 2025. NBK1428.
2. Vanier MT, Caillaud C, Levade T. Disorders of Sphingolipid Synthesis, Sphingolipidoses, Niemann-Pick Disease Type C and Neuronal Ceroid Lipofuscinoses. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 40, section 40.5 Neuronal Ceroid Lipofuscinoses. doi:10.1007/978-3-662-63123-2_40.
3. Mole SE, Schulz A, Badoe E, Berkovic SF, de Los Reyes EC, Dulz S, et al. Guidelines on the diagnosis, clinical assessments, treatment and management for CLN2 disease patients. Orphanet J Rare Dis. 2021;16:185. doi:10.1186/s13023-021-01813-5. PMID:33882967.
4. Orphanet. CLN2 disease (ORPHA:228349), Neuronal ceroid lipofuscinosis (ORPHA:216), CLN1 disease (ORPHA:228329) and CLN3 disease (ORPHA:228346): epidemiological data [Internet]. Orphadata API. api.orphadata.com/rd-epidemiology. Accessed on 6 Oct 2026.
5. HUGO Gene Nomenclature Committee. TPP1 — tripeptidyl peptidase 1 (HGNC:2073) [Internet]. genenames.org. Accessed on 6 Oct 2026.
6. Gardner E, Bailey M, Schulz A, Aristorena M, Miller N, Mole SE. Mutation update: Review of TPP1 gene variants associated with neuronal ceroid lipofuscinosis CLN2 disease. Hum Mutat. 2019;40:1924-38. doi:10.1002/humu.23860. PMID:31283065.
7. VariantValidator. Variant validation on GRCh38 with MANE Select transcripts (NM_000391.4, NM_001042432.2, NM_000310.4, NM_006493.4, NM_018941.4) [Internet]. rest.variantvalidator.org. Accessed on 6 Oct 2026.
8. Nickel M, Simonati A, Jacoby D, Lezius S, Kilian D, Van de Graaf B, et al. Disease characteristics and progression in patients with late-infantile neuronal ceroid lipofuscinosis type 2 (CLN2) disease: an observational cohort study. Lancet Child Adolesc Health. 2018;2:582-90. doi:10.1016/S2352-4642(18)30179-2. PMID:30119717.
9. Fietz M, AlSayed M, Burke D, Cohen-Pfeffer J, Cooper JD, Dvořáková L, et al. Diagnosis of neuronal ceroid lipofuscinosis type 2 (CLN2 disease): Expert recommendations for early detection and laboratory diagnosis. Mol Genet Metab. 2016;119:160-7. doi:10.1016/j.ymgme.2016.07.011. PMID:27553878.
10. BioMarin Pharmaceutical Inc. BRINEURA (cerliponase alfa) injection, for intraventricular use. Prescribing Information. July 2024 revision (BLA 761052; SPL version 30). U.S. Food and Drug Administration/DailyMed. DailyMed setid 485ab3c5-0359-4878-872f-2fe1e7df4047. Accessed on 5 Oct 2026.
11. European Medicines Agency. Brineura 150 mg solution for infusion: EPAR — Product Information (summary of product characteristics). First authorised on 30 May 2017; renewed on 28 Mar 2022. EMA. Accessed on 6 Oct 2026.
12. ClinicalTrials.gov. Records NCT01907087, NCT02485899, NCT02678689, NCT05152914, NCT05791864, NCT03770572, NCT05174039, NCT04637282, NCT07740512, NCT05228145, NCT02725580, NCT07582484, NCT04737460, NCT04476862, NCT06128226, NCT04613089, NCT06844877, NCT01399047 and NCT00337636 [Internet]. Bethesda (MD): National Library of Medicine. clinicaltrials.gov. Accessed through API v2 on 6 Oct 2026.
13. Auvin S, Hamandjian MA, Karam P, Hoebeke C. Incidence and timing of diagnosis of neuronal ceroid lipofuscinosis type 2 (CLN2): A nationwide study using the French hospital discharge database. Eur J Paediatr Neurol. 2026;60:44-9. doi:10.1016/j.ejpn.2025.12.001. PMID:41354011.
14. Van Hasselt PM, Van Genderen MM, Kuper WFE. CLN3-Related Disorder. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Published on 3 Sep 2026. NBK624593.
15. Sampaio LPB, Manreza MLG, Pessoa A, Gurgel-Giannetti J, Coan AC, Júnior HVL, et al. Clinical management and diagnosis of CLN2 disease: consensus of the Brazilian experts group. Arq Neuropsiquiatr. 2023;81:284-95. doi:10.1055/s-0043-1761434. PMID:37059438.
16. Augustine EF, Adams HR, de Los Reyes E, Drago K, Frazier M, Guelbert N, et al. Management of CLN1 Disease: International Clinical Consensus. Pediatr Neurol. 2021;120:38-51. doi:10.1016/j.pediatrneurol.2021.04.002. PMID:34000449.
17. Mink JW, Adams HR, Ahrens-Nicklas R, Andersen BN, Augustine E, Boustany RM, et al. Recommendations for the diagnosis and management of CLN3 disease (Batten disease) using the Delphi consensus methodology. Orphanet J Rare Dis. 2026;21:155. doi:10.1186/s13023-026-04298-2. PMID:41808212.
18. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Céroïde-lipofuscinoses neuronales. Filière de santé maladies rares G2M; centres de référence des maladies héréditaires du métabolisme. Saint-Denis La Plaine: HAS; November 2022 (published on 12 Dec 2022). Available from: https://www.has-sante.fr/jcms/p_3390370
19. European Reference Network for Hereditary Metabolic Disorders (MetabERN). Subnetworks: Lysosomal storage disorders (LSD) [Internet]. metab.ern-net.eu. Accessed on 6 Oct 2026.
20. Zeman W, Dyken P. Neuronal ceroid-lipofuscinosis (Batten's disease): relationship to amaurotic family idiocy? Pediatrics. 1969;44:570-83. PMID:5346636.
21. Haltia M. The neuronal ceroid-lipofuscinoses: from past to present. Biochim Biophys Acta. 2006;1762:850-6. doi:10.1016/j.bbadis.2006.06.010. PMID:16908122.
22. Santavuori P, Haltia M, Rapola J, Raitta C. Infantile type of so-called neuronal ceroid-lipofuscinosis. 1. A clinical study of 15 patients. J Neurol Sci. 1973;18:257-67. doi:10.1016/0022-510x(73)90075-0. PMID:4698309.
23. Vesa J, Hellsten E, Verkruyse LA, Camp LA, Rapola J, Santavuori P, et al. Mutations in the palmitoyl protein thioesterase gene causing infantile neuronal ceroid lipofuscinosis. Nature. 1995;376:584-7. doi:10.1038/376584a0. PMID:7637805.
24. The International Batten Disease Consortium. Isolation of a novel gene underlying Batten disease, CLN3. Cell. 1995;82:949-57. doi:10.1016/0092-8674(95)90274-0. PMID:7553855.
25. Sleat DE, Donnelly RJ, Lackland H, Liu CG, Sohar I, Pullarkat RK, et al. Association of mutations in a lysosomal protein with classical late-infantile neuronal ceroid lipofuscinosis. Science. 1997;277:1802-5. doi:10.1126/science.277.5333.1802. PMID:9295267.
26. Haltia M, Goebel HH. The neuronal ceroid-lipofuscinoses: a historical introduction. Biochim Biophys Acta. 2013;1832:1795-800. doi:10.1016/j.bbadis.2012.08.012. PMID:22959893.
27. Schulz A, Ajayi T, Specchio N, de Los Reyes E, Gissen P, Ballon D, et al. Study of Intraventricular Cerliponase Alfa for CLN2 Disease. N Engl J Med. 2018;378:1898-907. doi:10.1056/NEJMoa1712649. PMID:29688815.
28. Shock M, Nigro E, Donner EJ, Whitney R. CLN2 Disease: Current Understandings, Challenges, and Future Directions. J Child Neurol. 2026;41:118-34. doi:10.1177/08830738251374539. PMID:40966012.
29. Haltia M. The neuronal ceroid-lipofuscinoses. J Neuropathol Exp Neurol. 2003;62:1-13. doi:10.1093/jnen/62.1.1. PMID:12528813.
30. Steinfeld R, Heim P, von Gregory H, Meyer K, Ullrich K, Goebel HH, et al. Late infantile neuronal ceroid lipofuscinosis: quantitative description of the clinical course in patients with CLN2 mutations. Am J Med Genet. 2002;112:347-54. doi:10.1002/ajmg.10660. PMID:12376936.
31. Guelbert N, Espitia Segura OM, Amoretti C, Arteaga Arteaga A, Atanacio NG, Bazan Natacha S, et al. Classic and Atypical Late Infantile Neuronal Ceroid Lipofuscinosis in Latin America: Clinical and Genetic Aspects, and Treatment Outcome with Cerliponase Alfa. Mol Genet Metab Rep. 2024;38:101060 (corrigendum in 2024;41:101081). doi:10.1016/j.ymgmr.2024.101060. PMID:38469103.
32. Biglari S, Rezaei H, Asadollahzadeh E, Nikuei P, Sahraian MA, Sohanforooshan Moghaddam A, et al. Bi-allelic TPP1 variants in neuronal ceroid lipofuscinosis 2: clinical findings from an Iranian cohort of 20 patients, founder effect, and in silico analyses. Hum Genet. 2026;145:17. doi:10.1007/s00439-025-02812-3. PMID:41627539.
33. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 34, section 34.2.3.11 ATP13A2 Deficiency.
34. National Center for Biotechnology Information. ClinVar: VCV000002564, NM_006493.4(CLN5):c.1028_1029del (OMIM 608102.0001), pathogenic [Internet]. ClinVar. Accessed on 6 Oct 2026.
35. Johnson AM, Mandelstam S, Andrews I, Boysen K, Yaplito-Lee J, Fietz M, et al. Neuronal ceroid lipofuscinosis type 2: an Australian case series. J Paediatr Child Health. 2020;56:1210-8. doi:10.1111/jpc.14890. PMID:32329550.
36. Nickel M, Gissen P, Greenaway R, Cappelletti S, Hamborg C, Ragni B, et al. Language Delay in Patients with CLN2 Disease: Could It Support Earlier Diagnosis? Neuropediatrics. 2023;54:402-6. doi:10.1055/s-0043-1770143. PMID:37329878.
37. Saudubray JM, García-Cazorla Á. Clinical Approach to Inborn Errors of Metabolism in Paediatrics. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 1, sections 1.5.2.2, 1.5.2.6 and 1.5.8. doi:10.1007/978-3-662-63123-2_1.
38. Priglinger C, Courage C, Maier EM. Enzyme Replacement Therapy in CLN2-Associated Retinopathy. Klin Monbl Augenheilkd. 2025;242:213-8. doi:10.1055/a-2528-7886. PMID:40127655.
39. Rietdorf K, Coode EE, Schulz A, Wibbeler E, Bootman MD, Ostergaard JR. Cardiac pathology in neuronal ceroid lipofuscinoses (NCL): More than a mere co-morbidity. Biochim Biophys Acta Mol Basis Dis. 2020;1866:165643. doi:10.1016/j.bbadis.2019.165643. PMID:31863828.
40. Wibbeler E, Wang R, Reyes EL, Specchio N, Gissen P, Guelbert N, et al. Cerliponase Alfa for the Treatment of Atypical Phenotypes of CLN2 Disease: A Retrospective Case Series. J Child Neurol. 2021;36:468-74. doi:10.1177/0883073820977997. PMID:33356800.
41. Simonati A, Williams RE, Nardocci N, Laine M, Battini R, Schulz A, et al. Phenotype and natural history of variant late infantile ceroid-lipofuscinosis 5. Dev Med Child Neurol. 2017;59:815-21. doi:10.1111/dmcn.13473. PMID:28542837.
42. Kayani S, BordesEdgar V, Lowden A, Nettesheim ER, Dahshi H, Messahel S, et al. Neuronal ceroid lipofuscinoses type 7 (CLN7): a case series reporting cross sectional and retrospective clinical data to evaluate validity of standardized tools to assess disease progression, quality of life, and adaptive skills. Orphanet J Rare Dis. 2024;19:468. doi:10.1186/s13023-024-03448-8. PMID:39702211.
43. Yoganathan S, Jain P, Costain G, Whitney R. KCTD7-Related Progressive Myoclonic Epilepsy. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Published on 20 Nov 2025. NBK619245.
44. Rus CM, Weissensteiner T, Pereira C, Susnea I, Danquah BD, Morales Torres G, et al. Clinical and genetic characterization of a cohort of 97 CLN6 patients tested at a single center. Orphanet J Rare Dis. 2022;17:179. doi:10.1186/s13023-022-02288-8. PMID:35505348.
45. Arsov T, Smith KR, Damiano J, Franceschetti S, Canafoglia L, Bromhead CJ, et al. Kufs disease, the major adult form of neuronal ceroid lipofuscinosis, caused by mutations in CLN6. Am J Hum Genet. 2011;88:566-73. doi:10.1016/j.ajhg.2011.04.004. PMID:21549341.
46. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Developmental regression, p. 128–130. Local chapter.
47. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Seizures with developmental delay/intellectual disability, p. 308–310. Local chapter.
48. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Retinal receptor dystrophies, p. 302–304. Local chapter.
49. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Nystagmus, p. 252. Local chapter.
50. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Ataxic child, p. 74–75. Local chapter.
51. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Cerebral palsy, p. 96. Local chapter.
52. Touzon MS, Reyes Valenzuela G, Loos M, Gonzalez B, Gauto A, Obregón Gomez R, et al. Expanding the genotypic landscape of the neuronal ceroid lipofuscinoses: clinical and molecular findings from a tertiary-care center in Argentina. Seizure. 2026;141:26-33. doi:10.1016/j.seizure.2026.06.015. PMID:42413193.
53. Bryzik A, Larysz D, Larysz P, Karpierz JI, Paprocka J. Case Report: The window that closed too soon: lessons from a late CLN2 diagnosis and death of a 9-year-old boy. Front Genet. 2025;16:1622185. doi:10.3389/fgene.2025.1622185. PMID:40686560; PMCID:PMC12270889.
54. Lukacs Z, Nickel M, Murko S, Nieves Cobos P, Schulz A, Santer R, et al. Validity of a rapid and simple fluorometric tripeptidyl peptidase 1 (TPP1) assay using dried blood specimens to diagnose CLN2 disease. Clin Chim Acta. 2019;492:69-71. doi:10.1016/j.cca.2019.02.010. PMID:30771299.
55. Itagaki R, Endo M, Yanagisawa H, Hossain MA, Akiyama K, Yaginuma K, et al. Characteristics of PPT1 and TPP1 enzymes in neuronal ceroid lipofuscinosis (NCL) 1 and 2 by dried blood spots (DBS) and leukocytes and their application to newborn screening. Mol Genet Metab. 2018;124:64-70. doi:10.1016/j.ymgme.2018.03.007. PMID:29599076.
56. National Institute for Health and Care Excellence. Cerliponase alfa for treating neuronal ceroid lipofuscinosis type 2. Highly Specialised Technologies Guidance HST34 (replaces HST12). London: NICE; 18 Feb 2026. ISBN 978-1-4731-9275-1. NCBI Bookshelf NBK621308. PMID:41849646. (https://www.ncbi.nlm.nih.gov/books/NBK621308/). Accessed on 7 Oct 2026.
57. Williams RE, Adams HR, Blohm M, Cohen-Pfeffer JL, de Los Reyes E, Denecke J, et al. Management Strategies for CLN2 Disease. Pediatr Neurol. 2017;69:102-12. doi:10.1016/j.pediatrneurol.2017.01.034. PMID:28335910.
58. Kim A, Grover A, Hammon K, de Hart G, Slasor P, Cherukuri A, et al. Clinical Pharmacokinetics and Pharmacodynamics of Cerliponase Alfa, Enzyme Replacement Therapy for CLN2 Disease by Intracerebroventricular Administration. Clin Transl Sci. 2021;14:635-44. doi:10.1111/cts.12925. PMID:33202105.
59. Morris AAM, Jones S. Medications Used in the Treatment of Inborn Errors of Metabolism. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 45, Table 45.1. doi:10.1007/978-3-662-63123-2_45.
60. Schulz A, Specchio N, de Los Reyes E, Gissen P, Nickel M, Trivisano M, et al. Safety and efficacy of cerliponase alfa in children with neuronal ceroid lipofuscinosis type 2 (CLN2 disease): an open-label extension study. Lancet Neurol. 2024;23:60-70. doi:10.1016/S1474-4422(23)00384-8. PMID:38101904.
61. Schulz A, Schwering C, Wibbeler E, Westermann LM, Hagenah L, Lezius S, et al. Real-world clinical outcomes of patients with CLN2 disease treated with cerliponase alfa. Front Neurol. 2025;16:1516026. doi:10.3389/fneur.2025.1516026. PMID:40162009.
62. Schaefers J, van der Giessen LJ, Klees C, Jacobs EH, Sieverdink S, Dremmen MHG, et al. Presymptomatic treatment of classic late-infantile neuronal ceroid lipofuscinosis with cerliponase alfa. Orphanet J Rare Dis. 2021;16:221. doi:10.1186/s13023-021-01858-6. PMID:33990214.
63. Oliveira JA, Saddique MN, Qadri M, Shahid F, de Jesus ACFS, Sarmento FP. Treatment of Neuronal Ceroid Lipofuscinosis Type 2 with Cerliponase Alfa: A Systematic Review and Single-Arm Meta-Analysis of Two Studies. J Child Neurol. 2026;41:430-6. doi:10.1177/08830738251389180. PMID:41160491.
64. Rogers DL, Blind JE, Kienzle T, De Los Reyes E, Mendel TA, Jordan CO. Same-Day Approach for Combined Intravitreal and Intracerebroventricular Enzyme Replacement Therapy to Prevent Retinal Disease Progression in Children With Neuronal Ceroid Lipofuscinosis Type 2. Pediatr Neurol. 2025;170:1-3. doi:10.1016/j.pediatrneurol.2025.06.007. PMID:40578219.
65. Priglinger CS, Courage C, Lotz-Havla AS, Gerhardt M, Ehrt O, Kurz M, et al. Intravitreal Enzyme Replacement Therapy Slows Retinopathy in Late Infantile Ceroid Lipofuscinosis Type 2. Neuropediatrics. 2025;56:142-6. doi:10.1055/a-2510-5592. PMID:39776429.
66. Petersen M, Westermann LM, Hagenah L, Nickel M, Schwering C, Wibbeler E, et al. Mapping Clinical Progression to Brain Atrophy in CLN2 Patients Under Cerliponase Alfa Treatment: A Prospective Neuroimaging Study. J Inherit Metab Dis. 2026;49:e70202. doi:10.1002/jimd.70202. PMID:42175674.
67. Fote GM, Schafenacker A, Singh J, Sherlock D, Romero B, Himstead A, et al. Management of positive cerebrospinal fluid cultures from intraventricular reservoirs of neuronal ceroid lipofuscinosis type 2 patients: one institution's experience. J Neurosurg Pediatr. 2025;36:649-56. doi:10.3171/2025.4.PEDS24452. PMID:40845388.
68. Ammendolia I, Sframeli M, Esposito E, Cardia L, Noto A, Currò M, et al. Adverse Reactions to the Orphan Drug Cerliponase Alfa in the Treatment of Neurolipofuscinosis Type 2 (CLN2). Pharmaceuticals (Basel). 2024;17:1513. doi:10.3390/ph17111513. PMID:39598424.
69. Read J, Donald A, Rhead S, Acquaah L, Chan G, Heap F, et al. Chest-sited intraventricular access devices for cerliponase alfa infusion in Batten disease at a single tertiary United Kingdom pediatric center. J Neurosurg Pediatr. 2025;37:87-94. doi:10.3171/2025.7.PEDS25222. PMID:41172362.
70. Boop S, Nistal D, Barrios-Anderson A, Cherny WB, Chang IJ, Shelkowitz E, et al. Novel surgical approach for intraventricular cerliponase alfa enzyme replacement therapy via central venous access device (CVAD) port in neuronal ceroid lipofuscinosis type 2 (CLN2) disease. Childs Nerv Syst. 2025;41:172. doi:10.1007/s00381-025-06822-4. PMID:40295329.
71. Gopalka M, Manocha S, Romanski K, Hoeman E, Boop S, Lam SK. Conversion of a ventricular rickham reservoir to a subcutaneous chest port for repeated intraventricular enzyme replacement therapy: technical note and surgical considerations. Childs Nerv Syst. 2026;42:368. doi:10.1007/s00381-026-07463-x. PMID:42760386.
72. Gland Pharma Limited. Vancomycin hydrochloride for injection: prescribing information, ANDA 205694; SPL version of 30/01/2026. openFDA. Accessed on 6 Nov 2026.
73. Sleat DE, Gedvilaite E, Zhang Y, Lobel P, Xing J. Analysis of large-scale whole exome sequencing data to determine the prevalence of genetically-distinct forms of neuronal ceroid lipofuscinosis. Gene. 2016;593:284-91. doi:10.1016/j.gene.2016.08.031. PMID:27553520.
74. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Carrier frequency and testing in autosomal recessive disorders, p. 810. Local chapter.
75. Simonati A, Pezzini F, Nardocci N; CLNet Consortium; Santorelli FM. Age at onset and gene variants predict lifespan and disease duration in childhood neuronal ceroid lipofuscinoses. Dev Med Child Neurol. 2026;68:276-86. doi:10.1111/dmcn.16416. PMID:40708162.
76. Striano P. CLN2 disease: why early diagnosis matters more than ever. Eur J Paediatr Neurol. 2026;60:A4-A6. doi:10.1016/j.ejpn.2026.03.001. PMID:41826146.