Peripheral Nerve Pathology in Antibody Positive Paraneoplastic Neurological Syndromes
Andre Granger1, Tina Rajnauth1, Divyanshu Dubey1, John Mills1, Michelle Mauermann1, Sarah Berini1, P. James B. Dyck1, Christopher Klein1
1Mayo Clinic
Objective:
To describe peripheral nerve histopathology in neuropathy patients with paraneoplastic neurological syndromes (PNS) supported by autoantibody testing.
Background:
Rare autopsy cases in PNS suggest an association with shared antigens of misdirected cytotoxic T-cells against cancers and neural tissues.  Neuropathy may be the first manifestation but reports of peripheral nerve pathology are limited. We sought to review the pathologic findings within nerve biopsies of patients having neuropathies linked to cancer and high-risk paraneoplastic antibody seropositivity.
Design/Methods:
We retrospectively reviewed the clinical data and nerve biopsy reports of patients presenting with neuropathy and paraneoplastic syndromes reviewing their cancer diagnosis and peripheral onsets.  
Results:
Twelve sural nerves and one radial nerve were biopsied from patients with the following antibodies: Amphiphysin (n=6); Antineuronal nuclear antibodies (ANNA)-1 (n=2); ANNA-2 (n=2); Collapsin response mediator protein-5 (CRMP-5) (n=2); and Purkinje cytoplasmic antibody (PCA1) (n=1). Antibody testing occurred typically after biopsy or concurrently, and all patients had subacute onsets. Patients had overlapping involvements: length-dependent neuropathies (n=5), cranial neuropathies (n=1), brachial plexopathy (n=1), multiple mononeuropathies (n=2), and lumbosacral radiculoplexus involvement (n=8). Time to diagnosis was prolonged; commonly after one year. Malignancies were diverse including breast, ovarian, bladder, non-small cell lung cancer, small cell lung cancer, and coexisting colon and breast cancer. Fiber density was reduced in all, with multifocality frequently seen. Perineurial injury was absent, but sub-perineurial edema was common (n=4). Significant inflammation was absent in all but 5 cases, where small to moderate-sized CD45 collections were seen with vessel wall involvement with clinical lumbosacral plexus neuropathy (Amphiphysin (n=2), ANNA-2, and CRMP5) or multiple mononeuropathies (PCA-1).
Conclusions:
The absence of inflammatory infiltrates with reduced fiber density in distal sensory nerve biopsies is common in paraneoplastic neuropathies. A proximal inflammatory process is theorized. Microvessel inflammatory infiltrates with vessel wall involvement also occurs especially among lumbosacral plexopathy phenotypes. 
10.1212/WNL.0000000000203395