Axonal loss in normal-appearing white matter in a patient with acute MS

Case report.

The patient was a 43-year-old man with the clinical onset of relapsing-remitting MS (RR-MS) 9 months before death from respiratory arrest, presumably as a result of abrupt loss of automatic breathing while asleep (Ondine’s curse). Briefly, symptom onset occurred in September 1997 with the development of gait ataxia and oscillopsia over 1 week. These symptoms spontaneously improved before the development of diplopia and mild quadriparesis in December 1997. He received treatment with corticosteroids and was referred to the Cleveland Clinic Foundation. Diagnostic evaluation revealed cranial MRI changes typical of early MS, the presence of spinal fluid oligoclonal bands, vertical nystagmus, mild spastic quadriparesis (right > left), severely weak hand intrinsic muscles, a spastic–ataxic gait, minimal sensory changes, and the ability to walk unassisted. Shortly after this initial evaluation, he developed orthopnea and paroxysmal nocturnal dyspnea associated with increased oscillopsia and diplopia, and worsening gait. During admission for evaluation and treatment, he became progressively tachypneic and required tracheal intubation on March 16, 1998. Following a 7-day course of IV methylprednisolone (1,000 mg/day), he improved and was extubated on March 23, 1998. Following discharge from the hospital, he was again walking unassisted, but his condition worsened as the dosage of oral steroids was decreased. This resulted in readmission on May 8, 1998 during the evening shift. Shortly after transfer to the hospital floor, he was found unresponsive in asystole. Despite resuscitation, he was declared brain dead on May 11, 1998. CT imaging on May 11, 1998 revealed loss of all gray–white junctions, bilateral basal ganglion infarctions, bilateral superior cerebellar and midbrain infarction, and obstructive hydrocephalus—all findings consistent with severe anoxic brain injury.

Methods.

The protocol was approved by the Cleveland Clinic Institutional Review Board. Informed consent was obtained from the patient’s family to perform research on CNS tissue postmortem. The spinal cord and brainstem were removed through a rapid autopsy procedure, cut into 2-cm segments, and immediately fixed in 4% paraformaldehyde (PFA). Postmortem time was 4 hours, given the fact that after the terminal respiratory arrest causing anoxic brain injury, the patient was on respiratory support for approximately 72 hours before death. A 52-year-old man and a 63-year-old woman without neurologic disease served as controls.

The fixed tissue was cryoprotected in glycerol and frozen. Free-floating sections (30-μm-thick) were cut on a sliding-freezing microtome (Microm, Heidelberg, Germany) and immunostained by the avidin–biotin complex procedure and diaminobenzidine, as described previously.2,3⇓ Myelin basic protein (MBP; SMI-94), phosphorylated neurofilament (NF; SMI-31), and nonphosphorylated NF (SMI-32) antibodies were obtained from Sternberger Monoclonals (Baltimore, MD). Major histocompatibility complex (MHC) class II antibodies, a microglia/macrophage marker, were obtained from DAKO (Carpinteria, CA). For confocal microscopy, sections were incubated with two primary antibodies, incubated with fluorescein-conjugated and Texas red–conjugated secondary antibodies (Jackson Laboratories, West Grove, PA), and examined in a confocal microscope (Leica TCS-NT, Heidelberg, Germany).

Axons were quantified as described previously.3 Triangulated areas in MS and control cross-sections (segments C6–7) that included ventral 28.5-degree sectors and dorsal 9-degree sectors were analyzed bilaterally (figure 1F). Axonal numbers and density were obtained using NIH Image (version 1.61, NIH, Bethesda, MD) and Adobe PhotoShop software (version 5.0, San Jose, CA).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Phagocytic macrophages and activated microglia identify areas of fiber degeneration in the cervical spinal cord of the MS case. (A) In the ventral column (VC) of a control subject, major histocompatibility complex (MHC) class II antibodies stained resident microglia cells. (B) Large, rounded MHC class II–positive cells with short processes, characteristic of phagocytic macrophages, were prominent in the most affected ventral MS column. (C) The dorsal columns (DC) of the same section as in (B) contained MHC class II–positive resident microglia. (D and E) Neurofilament stained axons in the affected ventral column of the patient with MS (E) and a control subject (D). (F) Axons were quantified in four triangulated cervical (segments C6–7) spinal cord white matter areas using a computer-based morphometric system. (G) Total axon number in the four areas indicated in (F). Scale bars = 50 μm (A to C) and 25 μm (D and E).

For electron microscopy, 2-mm-thick ventral cervical spinal cord segments (C6–7) were fixed with 2.5% glutaraldehyde and 4% PFA in 0.08 M Sorenson’s buffer, treated with 1% osmium tetroxide (4 hours), dehydrated, and embedded in Epon. Ultrathin sections were counterstained with uranyl acetate and lead citrate and examined by electron microscopy.

Results.

Empty myelin profiles in NAWM.

Double labeling for myelin and NF or MHC class II, combined with confocal three-dimensional reconstruction, revealed degeneration of fibers in the cervical ventral column. In cross-sections, empty myelin profiles and myelin bodies occurred among intact myelinated axons (figure 2A). Activated microglia cells surrounded many of these myelin structures (see figure 2B). Primary demyelination was not detected. In longitudinal sections, aberrant myelin profiles were resolved as discontinuous myelin ovoids lacking NF-positive axons (see figures 2, C and D). In the lateral and dorsal columns, or in spinal cords from control patients, all myelin sheaths were structurally intact and contained NF-stained axons.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Descending spinal cord axons distal to the brain stem lesion undergo degeneration. Confocal images of the ventral cervical column stained for myelin (green), neurofilaments (A, C, and D; red) or the activated microglia/macrophage marker major histocompatibility complex (MHC) class II (B; red). In cross-sections (A), numerous myelin profiles lacking axons (arrowheads) were detected among intact myelinated axons. Some appeared as “empty” myelin circumferences (lower right profile), while others were composed of collapsed myelin membranes. Signs of primary demyelination were not detected. Activated class II cells (B; red; arrows) often contained phagocytosed myelin (green). Overt stripping of myelin from internodes was not observed in this region. In longitudinal sections (C), fibers undergoing degeneration appeared as rows of myelin ovoids (green; arrows) between intact myelinated axons. At higher magnification (D), the ovoids contained disrupted myelin membranes and lacked axonal staining (red). All images represent projections of three to five contiguous confocal slices. Scale bars = 25 μm.

Electron microscopy extended the confocal microscopy data. Myelin ovoids lacking discernible axons and collapsed myelin figures containing membranous debris were observed in the ventral column (figure 3, A and B). There were no ultrastructural signs of primary demyelination.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3. Electron microscopy confirms axonal degeneration in the ventral cervical column of the MS case. In both cross-section (A) and longitudinal section (B), myelin ovoids lacking axons (arrowheads) were detected among intact myelinated axons (Ax). In longitudinal sections (B), these ovoids often lay in rows. Scale bars = 2.5 μm.

Discussion.

These results provide direct evidence that neuronal or axonal injury at early stages of MS can cause significant axonal loss in NAWM distal to the lesion. In this case, myelin was relatively preserved despite 22% axonal loss. Ultimately, CNS myelin can remain for years and contribute to a grossly normal appearance of the affected white matter. Importantly, the data suggest that irreversible axonal loss with considerable retention of slowly degenerating myelin is one histopathologic correlate to NAWM abnormalities observed in patients with MS by MRI and MRS techniques.

In active MS lesions, neuronal or axonal injury is likely to be caused by inflammatory substances such as proteolytic enzymes, cytokines, oxidative products, and free radicals produced by activated immune and glial cells.18 Transected CNS axons will undergo rapid degeneration distal to the site of injury.9,10⇓ Although MS lesions were not found in the spinal cord postmortem, the ventral column containing descending corticospinal, vestibulospinal, and reticulospinal tracts exhibited a 22% axonal loss. The dorsal columns however, containing ascending sensory fibers, had normal axonal numbers and morphology. This patient experienced respiratory failure 2 months before death caused by a brain stem lesion after a short history of RR-MS. Postmortem, an active MS lesion was observed ventrally at the cervicomedullary junction. This lesion could account anatomically for the development of downbeat nystagmus, spastic quadriparesis, and the patient’s terminal event as a result of Ondine’s curse. This lesion could also account for the loss of axons distally. In addition, multiple periventricular abnormalities observed on MRI scans 3 months before death may have caused loss of corticospinal axons. Because the morphology of adjacent fibers in the ventral column and in other areas of the spinal cord was extremely well preserved, we consider it unlikely that the pathologic changes are the result of autolysis during the 72 hours on respiratory support between the terminal respiratory arrest and death. Together, the data support specific dropout of descending spinal cord axons caused by proximal axonal damage in acute brain stem or periventricular MS lesions.

Although quantitative MRI techniques have detected abnormalities in NAWM in patients with MS, the histopathologic basis for these changes has remained speculative.19 Recently, Simon et al.17 concluded that the development of T2-hyperintensities observed by MRI along the corticospinal tract in five patients, subsequent and inferior to isolated T2-hyperintense brain lesions, represented Wallerian degeneration. The data in the current study support this interpretation. It is also possible that mobile lipid signals, detected in MS brains by MRS at short echo times,20 represent myelin breakdown products associated with slowly degenerating myelin profiles when occurring in NAWM.21,22⇓ It is well established that disintegrating CNS myelin, in contrast to axons, can persist for a long time after proximal fiber transection.9,10⇓ The slow myelin breakdown in NAWM, compared with active MS lesions, may reflect lack of inflammation or relatively low blood–brain barrier permeability to monocytes in the former location.9 In contrast to transected peripheral nerves, where Schwann cells appear to actively participate in myelin breakdown,23 oligodendrocytes associated with multiple myelin sheaths may not respond to partial axonal loss. Because degeneration of CNS fibers causes only subtle changes in gross white matter appearance and routine histologic examination by Luxol fast blue, axonal loss in NAWM may be severely underestimated in autopsies of acute MS. In contrast, axonal loss in NAWM from patients with SP-MS can vary between 19 and 57%.12-14⇓⇓ Because NAWM constitutes the greatest proportion of white matter in patients with MS19 and because NAA levels in NAWM show strong correlation with disability over time,6,7,11⇓⇓ total white matter axonal status may be a more precise surrogate marker of disease progression than the presence and characteristics of individual lesions.5

Another interesting question relates to the immunologic consequences of axonal degeneration in NAWM. Conceivably, the process of slow myelin degradation may lead to a secondary inflammatory response, with additional cycles of autosensitization. The process of epitope spreading has been well documented in animals with experimental autoimmune encephalomyelitis and patients with MS,24 but the mechanisms have not been clarified. The process of slow myelin degradation secondary to axonal degeneration may drive epitope spreading and perpetuate the autoimmune cascade, resulting in recurrent multicentric inflammatory CNS lesions.

Although axonal loss correlates with reduced NAA levels in chronic MS lesions,3 decreased tissue NAA may reflect several cellular events such as altered neuronal activity or local metabolic changes due to CNS inflammation.3-5,16⇓⇓⇓ Interestingly, magnetization transfer imaging abnormalities in NAWM can precede the formation of new lesions in MS.25 Methods that distinguish axonal degeneration distal to sites of transection from other changes in NAWM are therefore urgently needed because they may have important impact on the understanding of MS pathogenesis.