Sequence Variants in Small CAG Repeat Expansions of the <i>HTT</i> Gene and Disease Onset and Progression in Huntington Disease.

Abstract

Huntington disease is an autosomal dominant neurologic disorder caused by an unstable cytosine-adenine-guanine (CAG) expansion (>35 CAG) in the <i>HTT</i> gene. The CAG repeat length is the major determinant of disease onset and penetrance. In addition, sequence variants within the <i>HTT</i> repeat, not detectable by currently used fragment size-based assays in diagnostic laboratories, are potent modifiers of disease onset and penetrance, particularly in carriers of small repeat expansions (<40 CAG). However, differences regarding severity, progression, and phenotype are poorly characterized. We aim to investigate how DNA sequence variants within the <i>HTT</i> repeat influence disease progression in carriers with small and moderate repeat expansions (CAG_36-42) expansions beyond onset. We included carriers with uninterrupted CAG_36-42 repeats. We used either clonal Sanger sequencing (in Vancouver) or short read sequencing (in Paris) to detect sequence variants. We compared age at onset (AO) and the ratio between reported and the Langbehn-predicted AO depending on the DNA sequence. We assessed the longitudinal progression of cognitive, motor, and functional scales over disease duration using linear mixed models and compared progression slopes according to the DNA sequence. We analyzed 328 carriers with uninterrupted CAG_36-42 repeats. Age at motor onset was 8 years earlier in CAG-CCG LOI patients than in patients with the canonical sequence (53.5 ± 11.1 years vs 61.8 ± 10.9, <i>p</i> < 0.0001) and approximately 13 years earlier than predicted age at motor onset by the Langbehn formula (12.9 ± 6.6, <i>p</i> < 0.001). Motor progression and cognitive decline were significantly faster in patients with a loss of the CAA and CCA interruptions (CAG-CCG LOI) compared with those harboring the canonical sequence. In addition, we identified 1 novel variant (CAG LOI-LO CCG) in 5 patients, leading to underestimation of 3 CAG repeats. In this large cohort, including DNA sequence and phenotypical data, the LOI variant showed a significant modifying effect on AO, motor, and cognitive disease progression. These findings, along with the identification of a novel variant, have important implications for genetic testing and counseling, especially for individuals with expansions close to cutoff ranges. In addition, they underscore the need to integrate the DNA sequence in the diagnostic process and revisit current onset prediction models.