N-of-1 investigational study of a novel antisense oligonucleotide drug in CHCHD10-related ALS shows early signs of efficacy

Summary

Background:

Pathogenic CHCHD10 variants cause a rare, dominantly inherited form of amyotrophic lateral sclerosis. Non-allele-selective CHCHD10 knockdown may mitigate a toxic gain-of-function mechanism. We evaluated nL-CHCHD-001 in one participant with the p.Arg15Leu variant.

Methods:

In this open-label N-of-1 study, 320 gapmer antisense oligonucleotides were screened and a lead candidate was selected using specificity and nonclinical safety criteria. Six intrathecal doses were administered over 12 months (three 50-mg doses followed by three 75-mg doses). Prespecified primary outcomes were 12-month changes in functional, cognitive, quality-of-life, respiratory, neurofilament light, and survival measures; secondary outcomes assessed safety and tolerability. Analyses were descriptive.

Findings:

No serious adverse events occurred. Post-dose headache and fatigue were mild to moderate, and cerebrospinal fluid safety results were unremarkable. Plasma neurofilament light decreased by approximately 50% from a mildly elevated pretreatment baseline and entered the laboratory reference range. The Amyotrophic Lateral Sclerosis Functional Rating Scale-Revised increased from 33 to 36, vital capacity from 48% to 55% predicted, cognitive and quality-of-life scores remained stable, and the participant was alive at 12 months.

Conclusions:

Individualized CHCHD10-directed treatment was feasible and well tolerated and was temporally associated with biomarker and clinical stability or improvement. This N-of-1 study demonstrates the feasibility of developing individualized ASO therapy for CHCHD10-related ALS and provides preliminary evidence of biomarker and clinical benefit. These findings support further evaluation of nL-CHCHD-001 and highlight NfL as a practical treatment-response biomarker for personalized therapeutics in ALS ClinicalTrials.gov: NCT06392126.

Graphical Abstract

eTOC Blurb

Cousin et al. report an individualized antisense oligonucleotide for CHCHD10-related ALS. In one participant, six intrathecal doses were well tolerated, plasma neurofilament light decreased into the reference range, and functional and respiratory measures were stable or modestly improved, supporting further evaluation of CHCHD10-directed therapy.

Introduction

Amyotrophic lateral sclerosis (ALS) can be caused by heterozygous pathogenic variants in the CHCHD10 gene and are associated with dominantly inherited disease.¹ Although phenotypes are varied, CHCHD10-associated disorders are often characterized by slowly progressive, late-onset motor neuron-disease.² Pathogenic variants in CHCHD10, which encodes a mitochondrial membrane protein, are reported to cause toxic gains of function that engender prominent TDP-43 pathology.³ Multiple lines of evidence support non–allele-selective knockdown of CHCHD10 as a rational and likely safe therapeutic strategy. Although early data allowed for a possible loss-of-function (LoF) contribution to p.R15L,^4,5 recent comprehensive studies using patient-derived and isogenic cell models, mouse knock-in lines, CSF proteomics, and neuropathology of spinal cord and motor cortex consistently indicate a predominantly toxic gain-of-function (GoF) mechanism with some hypomorphic features.² Population constraint metrics^6–8 (pLI = 0, LOEUF = 1.54, haploinsufficiency index = 77%) indicate that protein-truncating variants are tolerated at the gene level. In a recent UK Biobank analysis,² predicted LoF variants were more frequent in the general population and observed in 31 controls and 0 ALS cases, arguing strongly against LoF as a common highly penetrant cause. In vivo, heterozygous S55L (human S59L) knock-in mice develop fully penetrant, fatal mitochondrial cardiomyopathy with a robust mitochondrial integrated stress response (mtISR), whereas germline Chchd10 knockout alone yields normal survival without cardiomyopathy or mtISR; even single Chchd10 or Chchd2 knockouts do not significantly impair mitochondrial function, and only double Chchd2/Chchd10 knockouts show cardiomyopathy and mtISR.⁹ Consistently, ClinVar/ClinGen catalog multiple truncating CHCHD10 alleles without classifying any as pathogenic for ALS/FTD. Together, these data indicate that CHCHD10 haploinsufficiency is generally well tolerated, while disease is driven by toxic missense variants, making non–allele-selective CHCHD10 knockdown a mechanistically well-supported strategy.

Given that successful antisense oligonucleotide (ASO) knockdown has already been demonstrated for other ALS-associated genes, such as FUS and SOD1,^10,11 we sought to investigate the efficacy of an ASO targeting CHCHD10 in a patient with CHCHD10-related ALS. The risks associated with a first-in-human drug trial and the possibility of toxicity with overall reduction of CHCHD10 expression were considered, but compared to the natural history of the disease these risks were considered acceptable by the investigators, the patient, IRB and FDA regulators in line with FDA guidance for individual ASO development for severely debilitating or life-threatening conditions^12–15 and a subsequently proposed frame work.¹⁶

Results

The participant had slowly progressive ALS with onset in his right shoulder in 2020. His past medical history was notable for radiating left neck and arm pain with mild weakness in 2018, which completely resolved after a cervical 5-6-7 fusion. His phenotype can be described as O2pPLM3 using the recent OPM classification and he was Kings stage 2 at initiation of treatment. The patient’s father had developed ALS at age 72 and eventually succumbed to the disease at age 76. One of the father’s sisters developed amnestic dementia in her 80’s a few years before she died, but details are sparse. Otherwise, there was no family history of neurological disease. On initial examination, the patient had normal tone, atrophy of the left arm muscles including the deltoid, biceps, triceps finger flexors and intrinsic hand muscles, bilateral fasciculations in the pectoralis, deltoids, biceps, triceps and abdominal muscles and in the right FDI, moderate left slightly greater than right weakness of arm muscles with the weakest tested muscle being the left abductor pollicis brevis, and diffuse hyperreflexia including biceps, brachioradialis, triceps, quadriceps markedly brisk +2 (Mayo scale) and Achilles brisk with clonus +3 (Mayo scale). Babinski and Hoffman responses were not clearly present. Electrophysiological evaluation demonstrated acute and chronic denervation changes in essentially all muscles studied, without myopathic changes and a 7-Tesla brain MRI demonstrated abnormal iron deposition within the middle and deep layers of the primary motor cortex illustrating the upper motor neuron involvement.^28,29 He met Gold Coast Criteria for ALS and the older, revised El Escorial criteria for clinically definite ALS Genetic testing identified the R15L pathogenic variant in CHCHD10 without any other relevant findings, including variants of uncertain significance.

The patient received the first six doses (3× 50 mg and 3× 75mg) of nL-CHCHD-001 at the Mayo Clinic Florida over a year and treatment was well tolerated with unremarkable cerebrospinal fluid laboratory test results. Post-treatment plasma NfL concentrations declined by up to 50% (Table 1 and Figure 1). Untreated levels were mildly elevated 28.1 pg/ml (range 26.8–30.2 pg/ml; normal being up to 22.4 pg/ml) and at 3 month on treatment the average was 16.3 pg/ml (range 13.4–18.3 pg/ml). Scores on the ALSFRS-R improved 3 points since treatment initiation (Table 1). Vital Capacity remained stable from 48% pre-dose to 55% of predicted value at one year of treatment (Table 1). Based on the ALS Cognitive Behavioral Screen, cognition remained stable and within normal range (Table 1). While no serious adverse events have occurred, the patient did suffer a fall with a mild concussion shortly after the fourth dose. The patient consistently reports a mild headache and moderate fatigue the day after dosing, but no lingering or severe symptoms. All adverse events were mild to moderate graded CTCAE 1–2 (Table 2).

Table 1.

Longitudinal clinical and biomarker measures

Measure	Jun 2021	Feb 2022	Mar 2022	Jun 2022	Nov 2022	Mar 2023	Sep 2023	Jan 2024	Apr 2024*	May 2024	Jul 2024	Oct 2024	Jan 2025	Apr 2025
ALSFRS-R	47	46		44	39	40	36	31	33	33		33	35	36
Plasma NfL (pg/mL)						26.8	30.2		27.4	26.8	23	13.4	18.3	17.2
FVC (%)			98					68	48	60	51	57	53	55
CBS				18	15	18		18		18	19	20		19
ALSAQ-5									4	5	5	4	5	4

ALSFRS-R: 48 represents normal function. Neurofilament light: normal range < 22.4 pg/mL.

FVC (forced vital capacity): % predicted value. CBS (cognitive behavioral screen): normal range 15–20. ALS Assessment Questionnaire (ALSAQ) range 0 (best)-100 (worst).

^*First dose

Figure 1. Longitudinal clinical and biomarker measures before and after initiation of nL-CHCHD-001.

Neurofilament light levels pre- and post-initiation of therapy with the upper limit of normal shown as a dotted line. Forced vital capacity % of predicted, ALS Functional rating scale revised (ALSFRSR), and ALS cognitive behavioral screening test (CBS) results are also shown. Exact dates are used in the graph and only months are listed in the table.

Table 2.

Adverse events reported during the first 12 months of nL-CHCHD-001 treatment

Event	Timing/frequency	Maximum CTCAE grade	Serious?	Relatedness	Outcome
Headache	Day after dosing; recurrent	1 (mild)	No	Likely related to to procedure.	Resolved without lingering symptoms
Fatigue	Day after dosing; recurrent	2 (moderate)	No	Likely related to to procedure.	Resolved without lingering symptoms
Fall with mild concussion	Shortly after dose 4	2 (moderate)	No	Unrelated	Recovered without sequelae

Discussion

Personalized medicine approaches are being designed for rare genetically defined conditions including ALS.^17–19 Here, we describe our initial year of experience with a CHCHD10-targeting ASO patient with genetically confirmed CHCHD10-related ALS. To date, the ASO has been well-tolerated with a good safety profile and has demonstrated early signs of efficacy. The patient reports subjective improvement, and our primary response biomarker, Neurofilament light (NfL), has normalized. NfL was reduced as early as one month after 1st treatment, with concentrations decreasing more than 50% at the 4-month measurement. This supports the use of NfL as a treatment response marker that can be utilized in individual cases and not only in aggregate series. The use of NfL may pave a way for assessing individualized treatment response in ALS. NfL remains stable over time in ALS with most series showing no significant changes longitudinally.^20,21

Clinical outcome measures in the form of ALSFRS-R and FVC% support a small clinical improvement which would be consistent with the NfL data. This clinical and biomarker improvement would be rare in untreated ALS. Function in ALS generally declines and this is typically seen on the ALSFRS-R. Mean rates of decline in ALS at 0.89 point per month.²² No annual mean rates have been published for CHCHD10 R15L patients specifically, but scores of 30–38 were seen in 5 patients 2–5 years after onset. This patient had a ALSFRS-R score of 36 six years after disease onset consistent with this range. What constitutes a minimal clinically meaningful difference (MCID) on the ALSFRS-R measure is still being explored with recent studies suggesting 3.7 pts /3 months.²³ Cognition often is stable in ALS, but frontal lobe dysfunction can occur in CHCHD10-related disease. In this case there is no suggestion of cognitive decline with the patient continuing to practice as a physician during this period and maintaining stable ALS-CBS scores.²⁴ Vital capacity is a measure of breathing strength utilized routinely for clinical care and many ALS research projects, the decline is typically 2–2.9% per month and MCID has not been established.^25,26

The results of this case to date can be compared to the results with the SOD1 ASO tofersen. In the 6-month long VALOR phase 3 study, slowly progressive patients demonstrated a decrease in NfL by 40%, but NfL remained markedly elevated above normal and ALSFRS-R declined 1.4 points with FVC declining 4.6%.¹¹

The investigational study of nL-CHCHD-001 has subsequently been expanded to other sites through Silence ALS. Per the FDA approved protocol, we will increase the following doses to our target dose of 100 mg with the hope of even greater efficacy. The non-allele-specific nature of this ASO may make the product relevant for all patients with CHCHD10 ALS and, if effective in future trials, it would be the first ASO with benefit against a genetic form of ALS linked to TDP-43 pathology. The apparent treatment response coupled with longitudinal pretreatment biosamples will also allow us to explore the treatment response with markers of TDP-43 function, measures of mitochondrial function and transcriptomics in future experiments.

Limitations of the study

The unblinded n-of-1 context prevents more generalizable conclusions. Here we noted a minor relative increase (still within the normal range) in NfL concentration after dose four, potentially due to the concussion the patient suffered. NfL is a non-specific marker of neuroaxonal injury and is known to be increased after head trauma.²⁷ No additional sampling of NfL was conducted immediately after the mild event. Another limitation of the project is our inability to measure CHCHD10 protein and mRNA concentrations in CSF, future research may provide these.

Resource Availability

Lead contact

Further information and requests for resources should be directed to and will be fulfilled by the lead contact Björn Oskarsson (Oskarsson.Bjorn@mayo.edu).

Materials availability

The investigational antisense oligonucleotide (ASO), nL-CHCHD-001, was developed by the n-Lorem Foundation for individualized therapeutic use. Availability is subject to regulatory and institutional approval.

Data and code availability

• Requests for de-identified data supporting the findings of this study should be submitted to the Lead Contact as a written proposal describing the proposed analyses. Requests will be reviewed for scientific purpose, participant privacy, and compliance with institutional and regulatory requirements; a data-use agreement may be required before access is granted.

• This study did not generate original code.

• Any additional information required to reanalyze the data reported in this work paper is available from the Lead Contact upon request.

STAR Methods

EXPERIMENTAL MODEL AND PARTICIPANT DETAILS

Human participant

A male ALS patient was identified as having a CHCHD10 R15L (NM_213720.3:c.44G>T, p.(Arg15Leu)) pathogenic variant in February 2021. The family (TP) reached out to the n-Lorem Foundation and the variant was viewed as a possibly treatable genetic abnormality and a collaboration with was initiated with the patient establishing care at Mayo Clinic in May. ASO drug development was conducted by the n-Lorem Foundation with the patient participating under the Mayo Clinic approved IRB 21–006562 at Mayo Clinic in Jacksonville, Florida. The patient then received the investigational ASO treatment under the Mayo Clinic approved IRB 23–011476 as part of the N-of-1 trial (NCT06392126). He provided informed consent to participate in both studies.

The participant’s information on sex, age, gender and race was self-reported to be male, 51 years of age at time of treatment initiation, and Asian Indian, not Hispanic or Latino.

METHOD DETAILS

ASO design and selection

The n-Lorem Foundation designed 320 ASOs targeting CHCHD10 to promote the selective degradation of CHCHD10 pre-mRNA through recruitment of RNase H1 to the RNA-oligonucleotide heteroduplex.^30,31 These were 20 residue (20-mer) chimeric 2′-O-methoxyethyl/DNA modified oligonucleotides with five 2′-O-(2-methoxyethyl)-D-ribose (MOE) modifications on the 5′-and 3′- ends of the oligonucleotide and the central 10 nucleotides being deoxynucleotides (DNA) (MOE gapmer ASOs). One lead ASO (nL-CHCHD-001) was selected based on specificity and early nonclinical safety testing. An overview covering the n-Lorem Foundation ASO design, discovery, and preclinical development process was recently published by Crooke et al.³²

Drug Administration

nL-CHCHD-001 was administered intrathecally. The first two doses were administered four weeks apart, with subsequent doses given quarterly. The first three doses were 50mg, followed by three doses at 75mg.

Outcome measures

Prespecified primary outcomes were changes from baseline to 12 months in function, cognition, quality of life, respiratory function, plasma NfL,^33,34 and survival; secondary outcomes assessed safety and tolerability. Safety assessments included adverse event monitoring and cerebrospinal fluid safety laboratory testing. Plasma NfL concentrations were measured using commercial clinical CLIA certified testing using a single-molecule array (Simoa) assay on an HD-1 Analyzer (Quanterix, Billerica, MA, USA) by Mayo Clinic Laboratories, Rochester, MN, USA.³⁵ Test were run individually and shared with the patient.³⁶

The Amyotrophic Lateral Sclerosis Functional Rating Scale (ALSFRS-R) was used as a global measure of function. Ventilatory muscle strength was followed as forced vital capacity (FVC) expressed as percent predicted. The quality-of-life measure Amyotrophic Lateral Sclerosis Assessment Questionnaire 5 (ALSAQ-5) was collected as was the Amyotrophic Lateral Sclerosis Cognitive Behavioral Screen (ALS-CBS).^24,37 The same certified rater (JD) was used during the trial for ALSFRS-R, FVC and ALS-CBS to reduce inter rater variability. Historical data points have various raters.

QUANTIFICATION AND STATISTICAL ANALYSES

Analyses were descriptive, and no inferential statistical testing was performed.

ADDITIONAL RESOURCES

Clinical trial registry: ClinicalTrials.gov NCT06392126.

KEY RESOURCES TABLE (KRT)

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Critical commercial assays
Plasma neurofilament light assay (Simoa® NF-light™ Advantage Kit assay)	Quanterix, Billerica, MA, USA	Test ID: NFLP, Mayo Clinic Laboratories, Rochester, MN, USA35 Catalog #103180
Oligonucleotides
Antisense oligonucleotide nL-CHCHD-001	n-Lorem Foundation (nlorem.org)	nL-CHCHD-001

Highlights.

• An individualized ASO was developed for CHCHD10-related ALS

• Six intrathecal doses were well tolerated in an N-of-1 study

• Plasma neurofilament light decreased by approximately 50% into the normal reference range

• Clinical function and respiratory measures were stable or modestly improved

Context and Significance.

Amyotrophic lateral sclerosis (ALS) remains a fatal neurodegenerative disease with few effective treatments, despite growing recognition that a subset of patients have monogenic forms amenable to gene-targeted therapy. This study describes the first individualized antisense oligonucleotide developed for CHCHD10-associated ALS and demonstrates the feasibility of rapidly translating a genetic diagnosis into a patient-specific therapeutic intervention through collaboration among the participant, clinicians, and the n-Lorem Foundation. Although limited to a single participant, treatment was well tolerated and was accompanied by normalization of plasma neurofilament light concentrations and stabilization or modest improvement in clinical outcomes. These findings support further evaluation of CHCHD10-targeted therapy and highlight neurofilament light as a practical biomarker for individualized ALS therapeutics.