Differential motor neuron involvement in progressive muscular atrophy: a comparative study with amyotrophic lateral sclerosis

Abstract

Objective: Progressive muscular atrophy (PMA) is a clinical diagnosis characterised by progressive lower motor neuron (LMN) symptoms/signs with sporadic adult onset. It is unclear whether PMA is simply a clinical phenotype of amyotrophic lateral sclerosis (ALS) in which upper motor neuron (UMN) signs are undetectable. To elucidate the clinicopathological features of patients with clinically diagnosed PMA, we studied consecutive autopsied cases.

Design: Retrospective, observational.

Setting: Autopsied patients.

Participants: We compared clinicopathological profiles of clinically diagnosed PMA and ALS using 107 consecutive autopsied patients. For clinical analysis, 14 and 103 patients were included in clinical PMA and ALS groups, respectively. For neuropathological evaluation, 13 patients with clinical PMA and 29 patients with clinical ALS were included.

Primary outcome measures: Clinical features, UMN and LMN degeneration, axonal density in the corticospinal tract (CST) and immunohistochemical profiles.

Results: Clinically, no significant difference between the prognosis of clinical PMA and ALS groups was shown. Neuropathologically, 84.6% of patients with clinical PMA displayed UMN and LMN degeneration. In the remaining 15.4% of patients with clinical PMA, neuropathological parameters that we defined as UMN degeneration were all negative or in the normal range. In contrast, all patients with clinical ALS displayed a combination of UMN and LMN system degeneration. CST axon densities were diverse in the clinical PMA group, ranging from low values to the normal range, but consistently lower in the clinical ALS group. Immunohistochemically, 85% of patients with clinical PMA displayed 43-kDa TAR DNA-binding protein (TDP-43) pathology, while 15% displayed fused-in-sarcoma (FUS)-positive basophilic inclusion bodies. All of the patients with clinical ALS displayed TDP-43 pathology.

Conclusions: PMA has three neuropathological background patterns. A combination of UMN and LMN degeneration with TDP-43 pathology, consistent with ALS, is the major pathological profile. The remaining patterns have LMN degeneration with TDP-43 pathology without UMN degeneration, or a combination of UMN and LMN degeneration with FUS-positive basophilic inclusion body disease.

Figure 1

Measures of degeneration in the upper motor neuron system. (A–D) Loss of Betz cells in the primary motor cortex: stage (−), the Betz cells were spared in number and gliosis was absent (A); stage (+), mild neuronophagia and gliosis were noted (B) and stage (++), marked neuronophagia and glial proliferation were observed (C). (D–K) Aggregation of CD68 macrophages in the primary motor cortex (D–G) and the corticospinal tract in the lateral column of the spinal cord (H–K): stage (−), the aggregates were absent (D and H); stage (+), the aggregates were occasionally present (E and I); stage (++), the aggregates were present at a number of 1–5/ ×100 field (F and J) and stage (+++), the aggregates were diffusely observed (G and K). (L–O) Myelin pallor in the corticospinal tracts (CST) of the lateral column of the spinal cord: stage (−), myelin pallor was not detected (L); stage (+), myelin pallor was slightly notable (M); stage (++), myelin pallor was moderate (N) and stage (+++), the CST was entirely pale. (A–C) H&E staining, (D–K) anti-CD68 immunohistochemistry and (L–O) Klüver-Barrera staining. Scale bars: (A–G) 100 μm, (H–K) 50 μm and (L–O) 3 mm.

Figure 2

Quantitative analysis of the axonal fibres in the corticospinal tract. (A) Phosphorylated neurofilament (pNF)-positive fibres were automatically binarised using Luzex AP software. The density of pNF-positive axons (particles/10 000 μm²) was automatically calculated using averaged data from five fields (×400). The histogram of axonal sizes revealed that the percentages of axons that were more than 1 μm in diameter were smaller in ALS and PMA than in controls. (B) The large axonal fibres more than 1 μm in diameter were automatically recognised, binarised and counted using the software to successfully evaluate the axonal density. (C) There were significant differences in the densities of axons that were more than 1 μm in diameter between all pairs of clinical groups: p=0.001 (*) between the clinical amyotrophic lateral sclerosis (ALS) and clinical progressive muscular atrophy (PMA) groups, p=0.001 (*) between the clinical PMA and control groups and p<0.001 (**) between the clinical ALS and control groups. All patients diagnosed with clinical ALS exhibited lower values than the controls. In contrast, the results of the clinical PMA group were widely diverse, ranging from low values to values within the normal range.

Figure 3

Summary of the neuropathological findings in the included patients. The stages of the pathological changes correspond to those in figure 1. Pathological changes between the clinical groups were compared using Pearson's χ² test. ALS, amyotrophic lateral sclerosis; FUS, antifused-in-sarcoma; GCI, glial cytoplasmic inclusions; KB, Klüver-Barrera staining; NCI, neuronal cytoplasmic inclusion; PMA, progressive muscular atrophy; TDP-43, 43 kDa TAR DNA-binding protein; TPPV, tracheostomy positive-pressure ventilation.

Figure 4

Neuropathological profiles of the patients in the clinical progressive muscular atrophy group. (A–J) correspond to patient 2. The corticospinal tracts (CST) did not display myelin pallor (A), loss of large axonal fibres (B), or aggregation of macrophages (C). Additionally, in the primary motor cortex, neither the loss of Betz cells (D) nor aggregation of macrophages (E) was detected. The upper layers of the primary motor cortex rarely contained phosphorylated 43 kDa TAR DNA-binding protein (pTDP-43)-positive neuronal (F) and glial (G) inclusions. The spinal anterior horn displayed severe neuronal loss (H), pTDP-43-positive skein-like inclusions (I) and Bunina bodies (J). (K–M) correspond to patient 12. The CST displayed myelin pallor (K) and the depletion of large axonal fibres (L). Neuronophagia was often found in the primary motor cortex (M, arrows). (N and O) correspond to patient 6. The spinal motor neurons contained basophilic inclusion bodies (N) that were positive for antifused-in-sarcoma (FUS) based on immunohistochemistry (O). (A and K) Klüver-Barrera staining, (B) antiphosphorylated neurofilament immunohistochemistry, (C and E) anti-CD68 immunohistochemistry, (D, H, J and M) H&E staining, (F, G and I) anti-pTDP-43 immunohistochemistry and (O) anti-FUS immunohistochemistry. Scale bars: (A and K) 3 mm, (D and E) 100 μm, (C, H and M) 50 μm, (B) 20 μm and (F, G, I, J, N and O) 10 μm.