PROJECT SUMMARY/ABSTRACT
Heterozygous deletion of human Peripheral Myelin Protein 22 (PMP22) gene results in hereditary neuropathy
with liability to pressure palsies (HNPP). HNPP is characterized pathologically by focal myelin thickenings known
as “tomacula” in peripheral nerves. Patients with HNPP typically present with transient focal sensory loss and
muscle weakness that may be evoked by mild mechanical compressions that do not affect healthy humans. Our
preclinical studies testing small molecule compound in HNPP animal model have been shown to arrest disease
pathology. However, translation of this potential therapy to clinical use demand specific validated outcome
measures to monitor the effects of treatment longitudinally in patients with HNPP. Therefore, the goal of this
study is to develop monitoring biomarkers that may serve as either primary or secondary outcome measurements
in HNPP clinical trials. The study will investigate two measures: quantitative magnetic resonance imaging (qMRI)
and human skin biopsy. The degree of axonal loss is usually correlated with disease severity in neurological
disorders. Therefore, measures of axonal loss are often reliable biomarkers for disease progression. Traditionally,
pathology in peripheral nerve diseases has been evaluated by sural nerve biopsy, an invasive procedure that
requires surgical removal of the nerve, so sequential sural nerve biopsies are not possible for longitudinal studies.
Human skin biopsies are minimally invasive and can be repeated many times in the same subject. Preliminary
studies have demonstrated that pathologies in dermal axon and myelin can be quantified automatically through
deep learning. This study will use the established deep learning-based model to automate dermal nerve
morphometrics in skin biopsy. On the other hand, using qMRI to assess muscular fat fraction (FF) in legs, a
marker of muscle denervation that indirectly reflects axonal loss, we found that FF is increased in patients with
HNPP. This study will determine whether FF quantified in individual muscle can be used to track progression of
HNPP with better responsiveness. Furthermore, this study will develop a deep learning-based method to
automate the laborious individual muscle FF quantification with qMRI. Although intramuscular fat accumulation
is the end pathology resulting from axonal loss in HNPP, direct assessment of the diseased nerve is still needed.
This study will use the strategically acquired gradient echo (STAGE) imaging technique to quantify peripheral
nerves, including magnetization transfer ratio, proton density, longitudinal and effective transverse relaxation
times. Due to the transient, multi-focal features of clinical presentation in patients with HNPP, existing function
outcome measures, such as the Charcot-Marie-Tooth neuropathy score or nerve conduction studies, may not
capture those episodes in a timely manner. However, axonal loss accumulated over time determines the final
disabilities in patients with HNPP. Our overall hypothesis is that quantitative data from skin biopsy together with
qMRI can reliably measure axonal loss in patients with HNPP.