BACKGROUND. Neuronal remodeling in human heart disease is not well understood. METHODS. Stellate ganglia from patients with cardiomyopathy (CMY) and refractory ventricular arrhythmias undergoing cardiac sympathetic denervation (n = 8), and from organ donors with normal hearts (n = 8) collected at the time of organ procurement were compared. Clinical data on all subjects were reviewed. Electron microscopy (EM), histologic, and immunohistochemical assessments of neurotransmitter profiles, glial activation and distribution, and lipofuscin deposition, a marker of oxidative stress, were quantified. RESULTS. In CMY specimens, lipofuscin deposits were larger, and present in more neurons (26.3% ± 6.3% vs. 16.7% ± 7.6%, P < 0.043), than age-matched controls. EM analysis revealed extensive mitochondrial degeneration in CMY specimens. T cell (CD3+) infiltration was identified in 60% of the CMY samples, with one case having large inflammatory nodules, while none were identified in controls. Myeloperoxidase-immunoreactive neutrophils were also identified at parenchymal sites distinct from inflammatory foci in CMY ganglia, but not in controls. The adrenergic phenotype of pathologic samples revealed a decrease in tyrosine hydroxylase staining intensity compared with controls. Evaluation of cholinergic phenotype by staining for the vesicular acetylcholine transporter revealed a low but comparable number of cholinergic neurons in ganglia from both groups and demonstrated that preganglionic cholinergic innervation was maintained in CMY ganglia. S100 staining (a glial cell marker) demonstrated no differences in glial distribution and relationship to neurons; however, glial activation demonstrated by glial fibrillary acidic protein (GFAP) staining was substantially increased in pathologic specimens compared with controls. CONCLUSIONS. Stellate ganglia from patients with CMY and arrhythmias demonstrate inflammation, neurochemical remodeling, oxidative stress, and satellite glial cell activation. These changes likely contribute to excessive and dysfunctional efferent sympathetic tone, and provide a rationale for sympathectomy as a treatment for arrhythmias in this population. FUNDING. This work was made possible by support from NIH grants HL125730 to OAA, GM107949 to DBH, and HL084261 and OT2OD023848 to KS.
Olujimi A. Ajijola, Donald B. Hoover, Thomas M. Simerly, T. Christopher Brown, Jane Yanagawa, Reshma M. Biniwale, Jay M. Lee, Ali Sadeghi, Negar Khanlou, Jeffrey L. Ardell, Kalyanam Shivkumar