The Ignite app detects early cognitive impairment and tracks disease severity in genetic frontotemporal dementia

Technical Summary: The Ignite App for Detecting Early Cognitive Impairment in Genetic Frontotemporal Dementia

Problem Statement
Frontotemporal dementia (FTD) is a heterogeneous neurodegenerative disorder with a significant familial component, primarily driven by mutations in C9orf72, GRN, and MAPT genes. As disease-modifying therapies emerge, clinical trials face a critical bottleneck: the lack of sensitive, scalable outcome measures capable of detecting subtle cognitive changes during the presymptomatic and prodromal phases. While biomarkers like neurofilament light chain (NfL) and structural MRI show abnormalities decades before symptom onset, traditional pen-and-paper neuropsychological assessments often fail to detect cognitive decline until the years immediately preceding clinical conversion. Furthermore, existing digital tools for FTD are limited, and the disorder's unique clinical heterogeneity and prominent behavioral symptoms pose specific challenges for psychometric sensitivity and task design. There is an urgent need for digital cognitive tools that are sensitive to early presymptomatic changes, reflect gene-specific impairment profiles, and are suitable for remote administration in multi-site trials.

Methodology
This study evaluated the performance of "Ignite," an iPad-based cognitive assessment app, within the Genetic FTD Initiative (GENFI) cohort.

• Participants: The study included 569 participants from 24 sites, comprising mutation carriers (stratified by disease stage: asymptomatic CDR=0, prodromal CDR=0.5, mildly symptomatic CDR=1, and moderately/severely symptomatic CDR≥2) and non-carrier controls (N=200).
• Intervention: Participants completed the Ignite app, which consists of 12 interactive tasks designed to assess six cognitive domains: processing speed, executive function, social cognition, visuospatial skills, arithmetic, and semantic knowledge. Tasks included speed-accuracy trade-off measures (e.g., Think Back, Colour Mix, Path Finder) and specific domain tests (e.g., Face Match for social cognition, Sum Up for calculation).
• Comparators: Ignite scores were compared against a standard GENFI neuropsychological battery (including tests like the D-KEFS, Trail Making Test, and Facial Emotion Recognition Test) and normative data derived from over 2,000 healthy controls.
• Analysis:
  • Scores were converted to age, sex, and education-adjusted z-scores.
  • Cognitive composites were calculated for each domain.
  • Linear regressions (adjusting for demographics) compared Ignite scores between non-carriers and mutation carriers across disease stages and genetic groups (C9orf72, MAPT, GRN).
  • Receiver Operating Characteristic (ROC) analyses were conducted to compare the discriminative ability (AUC) of Ignite versus standard GENFI neuropsychology composites for distinguishing symptomatic (CDR≥1) and prodromal (CDR=0.5) carriers from non-carriers.

Key Contributions

1. Validation of a Digital Tool: The study provides empirical evidence that Ignite is a sensitive digital cognitive assessment capable of detecting early cognitive impairment in genetic FTD.
2. Gene-Specific Sensitivity: The research demonstrates that Ignite can capture distinct, gene-specific cognitive profiles. It identified that C9orf72 carriers show the earliest and broadest deficits, MAPT carriers exhibit early semantic and social cognition impairments, and GRN carriers show prominent executive and processing speed deficits at symptomatic stages.
3. Stage-Dependent Detection: The study maps the sensitivity of digital tasks across the disease spectrum, showing that Ignite detects impairments in asymptomatic carriers (CDR=0) in specific domains (executive function, social cognition, calculation) where standard measures may be less sensitive.
4. Comparative Performance: The study establishes that Ignite performs comparably to established clinical neuropsychological batteries in symptomatic stages and demonstrates modestly superior discrimination in the prodromal phase.

Results

• Sensitivity to Disease Stage: Ignite detected significant cognitive impairment (p<0.001) in all symptomatic carriers (CDR≥1) across all six cognitive domains compared to non-carriers. Prodromal carriers (CDR=0.5) showed impairments in all domains except social cognition. Asymptomatic carriers (CDR=0) demonstrated reduced performance in executive function, social cognition, and calculation.
• Genetic Stratification:
  • C9orf72 carriers showed significant deficits across all domains even at the asymptomatic stage.
  • MAPT carriers showed early impairments in social cognition and semantic knowledge at the asymptomatic stage, with broader deficits emerging later.
  • GRN carriers showed significant differences in executive function and semantic knowledge at prodromal and mildly symptomatic stages, but no significant differences were detected at the asymptomatic stage.
• Discriminative Accuracy (ROC Analysis):
  • Symptomatic Carriers: Ignite and GENFI neuropsychology showed similarly high discrimination (AUCs 0.91–0.92 for Ignite vs. 0.89–0.93 for GENFI) across executive function, processing speed, and social cognition.
  • Prodromal Carriers: Discrimination was lower for both measures but consistently higher for Ignite across all domains (AUCs 0.58–0.70 for Ignite vs. 0.55–0.64 for GENFI).

Significance and Claims
The authors conclude that Ignite is a sensitive digital cognitive assessment capable of detecting early cognitive changes in genetic FTD. The study claims that Ignite demonstrates improved discrimination compared to standard neuropsychological measures at milder disease severity (specifically the prodromal stage). The authors assert that the app's sensitivity to disease stage and gene-specific profiles supports its potential utility as a scalable, remote outcome measure for future FTD clinical trials.

The paper notes that while Ignite is promising, it currently lacks an episodic memory task and performance validity measures, limitations that are being addressed in a revised version (Ignite+). The authors emphasize that the cross-sectional nature of the data limits inferences about within-person change over time, and longitudinal studies are required to fully validate Ignite's utility for tracking disease progression. The study positions Ignite not as a replacement for clinical diagnosis, but as a tool to complement standard assessments by providing more granular, sensitive measures of cognitive change in the earliest stages of the disease.