oculomotor nerve regeneration after aneurysm surgery
TRANSCRIPT
OCULOMOTOR NERVE REGENERATION AFTER ANEURYSM SURGERY
JAMES M. KERNS, P H . D . , D O N A L D R. SMITH, M.D., FRANK S. JANNOTTA, M.D., AND M E L V I N G. ALPER, M.D.
Washington, D.C.
Intracranial aneurysms may produce oculomotor nerve palsy as the initial clinical symptom1 - 4 with or without associated subarachnoid hemorrhage. The most common site of origin for such aneurysms is the internal carotid artery at the origin of the posterior communicating artery, whereas a location at the posterior cerebral artery is less frequent.1,5 We describe herein a patient who had oculomotor nerve palsy secondary to a posterior cerebral artery aneurysm that did not rupture. Partial recovery occurred after ligation of this aneurysm despite the 12-month duration of oculomotor nerve palsy before surgery. Comparison of light and electron microscopic measurements indicated regeneration of the nerve. The number, density and diameter of axons and their myelin sheaths in the regenerated nerve were compared to the normal oculomotor nerve. Such morphometric evaluations of human peripheral nerves may provide insight into the nature and mechanisms involved in normal myelination,6,7 as well as in peripheral nerve degeneration and regeneration.8,9 Others have reported briefly on this study.10,11
C A S E REPORT
A 57-year-old woman had an acute onset of severe retro-orbital headaches that were accompanied by complete blepharoptosis and diplopia on the right
From the Departments of Anatomy (Dr. Kerns), Neurological Surgery (Dr. Smith), Pathology (Dr. Jannotta), and Ophthalmology (Dr. Alper), George Washington University Medical Center. This study was supported in part by BRSG Grant RR 5359 from the Biomedieal Research Program, Division of Research Resources, National Institutes of Health (Dr. Kerns).
Reprint requests to James M. Kerns, Ph.D, Department of Anatomy, George Washington University Medical Center, 2300 Eye St., N.W., Washington, DC 20037.
side. Treatment for diabetes was instituted. Glaucoma had been under treatment for several years.
After ten months without improvement, the patient was referred to a neuro-ophthalmologist. Examination revealed absence of headaches, but a complete right oculomotor nerve palsy, both internal and external, and slight exophthalmos. Visual acuity after elevation of the blepharoptic eyelid was normal and visual fields were full in both eyes. Intraocular pressure under treatment with pilocar-pine 4% four times a day was normal. A mild hemiparesis was apparent by slight-weakness in the left arm hyperreflexia in both the left arm and leg, and a left Babinski response. Computed tomography (Fig. 1) and angiography (Fig. 2) indicated an aneurysm of the right posterior cerebral artery that projected directly into the right cerebral peduncle.
Craniotomy was performed and the aneurysm was found to be firm with a broad origin from the posterior cerebral artery at the junction with the
Fig. 1 (Kerns and associates). Computed tomo-gram revealing a dense mass in the region of the right midbrain (arrow).
AMERICAN JOURNAL OF OPHTHALMOLOGY 87:22.5-233, 1979 225
226 A M E R I C A N JOURNAL O F O P H T H A L M O L O G Y FEBRUARY, 1979
Fig. 2 (Kerns and associates). Angiogram is base view demonstrating the large aneurysm projecting posteriorly from the right posterior cerebral artery into the region of the cerebral peduncle.
posterior communicating artery. It was successfully ligated with a large suture after clipping of the posterior communicating artery. The right oculomotor nerve, passing around the mass of aneurysm, was extremely thin, pinkish, and appeared to be grossly demyelinated. After surgery, the neurologic status was unchanged except for the permanent addition of a complete left homonymous hemianopia. The pre-operative hemiparesis gradually improved and the patient was ambulatory.
By two months postoperatively, there was elevation of the right upper eyelid. Oculomotor function slowly returned until one year after craniotomy when there was only slight blepharoptosis of the right upper eyelid in the primary position of gaze. Because the patient had glaucoma and took pilocar-pine, this medication was discontinued 72 hours before examination of the pupillary reflexes. The right pupil was slightly larger than the left, but both reacted to light and accommodation. On abduction, the pupillary fissure of the right side narrowed and on adduction the fissure widened. On down gaze to the right, the right upper eyelid retracted and the pupil dilated. On up gaze, the function of the right superior rectus and right inferior oblique muscles was deficient.
At 21 months postoperatively, the patient returned with severe back pain caused by metastatic disease of the lumbar spine. The patient deteriorated rapidly and died two months later of widespread metastatic adenocarcinoma.
Autopsy showed a primary adenocarcinoma of the gall bladder with metatases to liver, lungs, lymph
nodes, bones, and adrenal glands. The immediate cause of death was bronehopneumonia. The fixed brain weighed 1,200 g and showed symmetrical cerebral hemisphere with normal gyral configurations and unremarkable leptomeninges.
The saccular aneurysm measured 1.0 x 1.1 x 0.7 cm and projected superoposteriorly into the inter-peduncular fossa, where it indented the medial half of the right cerebral peduncle, and was bound to adjacent leptomeninges by fibrous adhesions (Fig. 3). The right oculomotor nerve was closely applied to the aneurysm and was displaced inferior-ly by the aneurysm within the interpeduncular fossa; in its extra-axial course beyond the aneurysm, the right oculomotor nerve was distinctly smaller than the left oculomotor nerve, measuring 1.8 mm in greatest diameter as opposed to 2.5 mm for the left oculomotor nerve.
The aneurysm and adjacent posterior cerebral artery were both filled with pale fibrous tissue microscopically consistent with an organized thrombus. A metal surgical clip occluded the right posterior cerebral artery.
The right basis pedunculi showed indentation, compression, and erosion with gliosis in its medial
Fig. 3 (Kerns and associates). Artist's depiction of the aneurysm and its anatomic relationships. Anterior is to the top. Note the marked compression of the right cerebral peduncle and the relationship of the right oculomotor nerve.
VOL. 87, NO. 2 OCULOMOTOR NERVE REGENERATION 227
half, corresponding to the aneurysmal mass. The right corticospinal tracts in pons and medulla were slightly atrophied. The right lateral geniculate body, optic radiations, and portions of the right parahip-pocampal and medial occipitotemporal gyres, including calcarine cortex, were atrophic and softened with the microscopic appearance of a semicystic healed infarct.
R E S U L T S
Primary fixation of the brain at autopsy was by immersion in neutral buffered formalin 24 hours after death. The methods of preparation and analysis are otherwise similar to those described by others.8,9 The right (regenerated) and left (normal) oculomotor nerves were fixed further in a weak aldehyde mixture in cacodylate buffer, osmicated, dehydrated, embedded transversely in Epon-Araldite, and sectioned at 1 JUL.
Both nerves were arbitrarily divided into inner core and outer rim portions of equal area (Fig. 4) to test for differential nerve compression effects, and photographs were randomly taken at x530 magnification from each of these areas. Nerve counts and measurements were made from four randomly selected grids (0.05 mm2) on each of these four areas
Fig. 4 (Kerns and associates). Comparison of normal oculomotor nerve (left) with the regenerated nerve (right), showing a reduction in the transverse area. The heavy line separates the inner core region from the outer rim in each nerve (toluidine blue stain on 1 u. section, x 37).
(Fig. 5). Fiber counts included all profiles that were identifiable as myelinated axons within the grid area. An estimate of the total number of fibers was derived from these grid counts and the overall area measurements. The sample included about 10% of the total estimated number. Morphometric measurements were made on the first 50 axons having distinct mye-lin sheaths and nearly spherical profiles from each grid, giving a total of 400 fibers for each nerve. The axon profile (d) was estimated from the perpendicular axes.12
The myelin sheath thickness was measured at the most uniform point, doubled, added to the diameter (d) to give the overall diameter (D), and the g-ratio (d/D) was then calculated.
The light microscopic appearance of the oculomotor nerve on the normal side was identical in the outer rim and inner core regions (Fig. 4, left and Fig. 5, bottom). The effects of postmortem autolysis were evident, but the morphologic structure of most myelinated axons was sufficient for detailed morphometric comparisons (Table 1). Compared to the normal nerve, the overall area of the regenerated nerve decreased about 62%. The mean density of myelinated axons was slightly increased because of compression in the rim (19%, P<.10, Fig. 4, right, and Fig. 5, top left) and even more in the core (57%, P<.001, Fig. 5), top right, but the estimated net effect was an overall reduction of about 49% in the total number of fibers (Table 1).
A cluster of small myelinated and unmyelinated fibers was present in the peripheral rim of both the normal and regenerated nerve. The mean diameter of the myelinated axons on the regenerated side was 23% (P<.001) less than the normal side (Fig. 5 and Table 2) as determined by two-factor analysis of variance. However, further statistical analysis (Newman-Keuls multiple range test) revealed no significant difference between
228 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY, 1979
Fig. 5 (Kerns and associates). Light microscopic appearance of the outer rim (top left) and inner core (top right) regions on the regenerated oculomotor nerve compared to the rim (bottom left) and core (bottom right) regions on the normal side. Note the reduction in diameter and myelin sheath thickness on the regenerated side (toluidine blue stain on 1 a. section, x.350).
the rim and core regions. Similar results of the regenerated nerve (Table 2) was were obtained for the g-ratio (d/D), with increased significantly by 12% to 0.73 the control nerve having a mean g-ratio (P<.001), but still within the normal of about 0.64. The overall mean g-ratio range of mean values.6 A higher mean
TABLE 1 QUANTITATIVE CHANGES IN OCULOMOTOR NERVE
Factor
Area (mm2) Density*
(No. of fibers/ 0.05 mm2)
Estimated No. of fibers
Rim
1.9 303 ± 7
11,500
Normal
Core
1.9 309+4
11,700
Whole Nerve
3.8
23,200
Rim
0.7 360±22
5,000
Regenerated
Core
0.7 486±2
6,800
Whole Nerve
1.4
11,800
(-62%)
(-49%)
*Mean ± standard error; any two means not spanned by the same underscore are significantly different (P-C.001), n=4.
VOL. 87, NO. 2 OCULOMOTOR NERVE REGENERATION 229
TABLE 2
QUANTITATIVE CHANGES IN MYELINATED AXONS
Factor
Myelinated axon diameter ((i)*
g-Ratio (d/D)*
Rim
9.6±0.21
0.64 ±.006
Normal Core
9.1 + 0.21
0.66±.005
Regenerated Rim
7.0±0.14
0.73±.005
Core
7.3±0.15
0.73 + .005
(-23%)
(+12%)
*Mean + standard error; any two means not spanned by the same underscore are significantly different (P<.001), n=200.
g-ratio on the regenerated side indicates that the reformed myelin sheath was proportionately thinner than normal.
There was a negative correlation (r = -0 .53 , P<.001) beteween the total fiber diameter and the g-ratio in the control nerve (Fig. 6), but the points scattered around a nearly straight regression line (Y = .787 - .014X), with a slightly negative slope (a). A positive correlation (r = 0.53, P < .001) was present in the regenerated nerve. When these latter values were plotted (Fig. 6), the points fit a straight regression line (Y = .600 + .018X) with a slightly positive slope (b). The slopes of the two lines were significantly different (P<.001). The variability
Fig. 6 (Kerns and associates). Plot of the g-ratio (d/D) vs the total diameter (D) for the normal (a) and regenerated (b) oculomotor nerves, with the corresponding means at the arrows on each regression line. The shaded histograms depict the fiber distribution for each nerve. Data for each nerve based on 400 fibers.
of g-ratio values was slightly less for the regenerated nerve (Fig. 6). Both nerves showed a wide range of individual g-ratio values (0.45 to 0.85) so the advantage of greater precision by electron microscopy became less important. Also, variation in the g-ratio along individual fibers (n = 7) measured on adjacent semithin sections was only 2%. These data were evaluated by electron microscopy from a smaller sample with similar results. However, a detailed ultrastructural analysis was not performed because of marginal tissue preservation. Numerous unmyelinated axons and an increased amount of connective tissue were evident in the regenerated nerve (Fig. 7).
DISCUSSION
The prognosis for recovery of oculomotor nerve function is generally excellent if aneurysmal pressure is surgically relieved within the first few days.3 '13,14 In these patients, the lesion is most likely a physiologic conduction block caused by compression rather than axonal disruption. When the duration of paresis lasts weeks or months, regeneration usually requires two to three months.5 ,13 Wallerian degeneration may have occurred and actual regrowth of the axons is required. Usually regeneration involves an increase in the number of distal axons because of sprouting, 5,1S instead of a decrease as shown in our case.
230 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY, 1979
i •
A
**-! ' W N»
v I.Op
Fig. 7 (Kerns and associates). Ultrastructural appearance of the rim region in the regenerated nerve. A single myelinated axon with a thin myelin sheath is surrounded by many unmyelinated axons and an increased connective tissue endoneurium (electron microscopy, x 10,000).
Delayed regeneration is usually accompanied by a high incidence of aberrant regeneration as described by Walsh16 and others.17 '18 Others 3-19 have stated that some fiber misdirection always occurs if the nerve lesion is caused by aneurysm or trauma and is of more than six weeks' duration. It is therefore not unusual that our patient showed signs of aberrant regeneration, first apparent two months after surgery and continuing until her death nearly two years later. In the characteristic pseudo-Graefe sign, the eyelid elevates when the affected eye attempts inward or downward gaze, because some axons originally directed to the medial or inferior rectus end up in the elevator
muscles. Vertical movements were poorly performed, probably because of simultaneous contraction of the superior and inferior recti by a random blend of the axons directed to these muscles.16
There was a significant return of oculomotor nerve function, as indicated by eyelid motility, absence of diplopia, and a normal pupillary response. This was unexpected because of the one-year duration of the preoperative oculomotor nerve palsy, the attenuation of the nerve at surgery, and the marked structural deficits in the regenerated nerve. Lack of correlation between structure and function was noted by Hyland and Barnett20 in a similar study. The misdirection syndrome has
VOL. 87, NO. 2 OCULOMOTOR NERVE REGENERATION 231
been experimentally produced in monkeys with detailed clinical stages of recovery,18 but few studies include any histology of the regenerated oculomotor nerve fibers.20,21
An estimated number of 23,200 axons in the normal oculomotor nerve in our study agrees with previous estimates of 24,000 by others.22,23 However, Traut-man24 gave a lower number of axons, a smaller transverse area and a larger mean axonal diameter compared to our values. Most investigators describe the normal human oculomotor nerve as consisting of large extrafusal fibers 10 to 16 JJL, with a smaller population of both extrafusal and parasympathetic fibers at 1 to 6 u..23-26
This distinct bimodal distribution was not apparent in our material. The peripheral location of a cluster of small presumably parasympathetic fibers26-28 was confirmed, although they were not included in the random sample for morphometric analysis. Clinically, pressure caused by aneurysm usually affects these smaller pupillary fibers in the periphery.28,29 The importance of considering regional variations in fiber size and density within a given nerve has been previously noted.30
In our study, a significant rim or core difference, or both, was found only with fiber density.
In our study, there seemed to be some random sparing of fibers of both large and small diameters, but this is difficult to determine in the late recovery stage.31
The extent of regeneration is considered more complete one year after crush or compression lesions as compared to nerve section, because the physical alignment of the distal segment is retained.32 Most investigators believe the recovery is incomplete even after crush lesions, but the greatest reduction in fiber diameter is in the most distal portion of the nerve.33 Our estimated reduction in fiber diameter by 23% at the site of injury may be conservative.
Numerous studies have shown that my-elin sheath thickness is directly related to the axon diameter.6,7 The relationship is often expressed as a g-ratio of the axon diameter to the overall diameter, and measured values of 0.6 to 0.7 correspond to theoretical predictions.34 However, the exact nature of the relationship remains unclear. Most assume a curvilinear correlation, with smaller axons having relatively thicker myelin sheaths. Others have reported a horizontal line signifying a constant g-ration, a straight line with either a positive or negative slope, or random scatter.6,35 Our data of the normal oculomotor nerve suggest a fairly constant g-ratio as the total axon diameter increases.
The slope of the g-ratio in the regenerated nerve shows a significant positive shift and the mean value is higher compared to the normal nerve. In contrast, Schroder9 showed a negative shift in the g-ratio. Friede35 found that the g-ratio remained fairly constant at about 0.7 following ligation of the sciatic nerve in young rats. The slopes were the same except at the site of ligation and results were similar after release of the ligation.
The g-ratio variability in the regenerated nerve is less, as compared to the normal nerve, probably because the myelin sheaths have been newly formed. However, Schroder9 and Friede35 found just the opposite. Variability in the g-ratio may be related to section thickness, photographic and measurement errors, fixation, species differences, and the peripheral nerve examined.6,8 '9
There is some indication that the smallest axons have a lower g-ratio, making the plot curvilinear. A greater proportion of these axons in the regenerated oculomotor nerve may correspond to atrophic nerve fibers9,36 that are either dying or have made inappropriate connections with the eye muscles in aberrant regeneration. Most fibers in the regenerated nerve have
232 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY, 1979
a thinner myelin sheath, and this, rather than the decreased axon diameter, causes the reduction in conduction velocity.34
Perhaps in our patient the large reduction in area was partially compensated by an increase in fiber density, and the loss of myelinated axons by retention of many unmyelinated fibers.
SUMMARY
A 57-year-old woman had symptoms of oculomotor nerve palsy first appearing one year before successful surgical liga-tion of a saccular aneurysm arising from the right posterior cerebral artery. During the subsequent postoperative period of two years, oculomotor nerve functions improved as the result of regeneration. Extensive morphometric evaluation of the regenerated nerve was compared to the normal side at the light microscopic level. The affected nerve showed a reduction in the transverse area (62%), estimated number of fibers (49%), and mean diameter of myelinated axons (23%). The normal g-ratio of axon to total diameter was almost constant at 0.64, but on the regenerated side it increased to 0.73. An increase in unmyelinated axons and connective tissue endoneurium was evident at the ultrastructural level. The significance of these marked quantitative changes was compared to the partial return of oculomotor nerve function.
R E F E R E N C E S
1. Harris, P., and Udvarhelyi, G. B.: Aneurysms arising at the internal carotid-posterior communicating junction. J. Neurosurg. 14:180, 1957.
2. Cogan, D. G., and Mount, H. T. J.: Intracranial aneurvsms causing ophthalmoplegia. Arch. Oph-thalmol. 70:757, 1963.
3. Hepler, R. S., and Cantu, R. C.: Aneurysms and third nerve palsies. Ocular status of survivors. Arch. Ophthalmol. 77:604, 1967.
4. Miller, N. R.: Solitarv oculomotor nerve palsy in childhood. Am. J. Ophthalmol. 83:106, 1977.
5. Walsh, F. B., and Hoyt, W. F.: Clinical Neuro-ophthalmologv, vol. 1, 3rd ed. Baltimore, Williams and Wilkins Co., 1969, pp. 252, 1764.
6. Bischoff, A., and Thomas, P. K.: Microscopic
anatomy of myelinated nerve fibers. In Dyck, P. J., Thomas, P. K., and Lambert, E. H. (eds.): Peripheral Neuropathy, vol. 1. Philadelphia, W. B. Saunders, 1975, p. 104.
7. Waxman, S. G.: Integrative properties and design principles of axons. Int. Rev. Neurobiol. 18:1, 1975.
8. Dyck, P. J., Gutrecht, J. A., Bastron, J. A., Karnes, W. E., and Dale, A. J. D.: Histologic and teased-fiber measurements of sural nerve in disorders of lower motor and primarv sensorv neurons. Mayo Clin. Proc. 43:81, 1968.
9. Schroder, J. M.: Altered ratio between axon diameter and myelin sheath thickness in regenerated nerve fibers. Brain Res. 45:49, 1972.
10. Kerns, J. M., Smith, D. R., and Jannotta, F. S.: A morphometric evaluation of oculomotor nerve regeneration following aneurysm surgery, abstracted. Anat. Rec. 190:443, 1978.
11. Alper, M. G., Davis, D. O., and Pressman, B. D.: Use of computerized axial tomography (EMI scanner) in diagnosis of exophthalmos. Trans. Am. Acad. Ophthalmol. Otolaryngol. 79:151, 1975.
12. Mayhew, T. M., and Momoh, C. K.: Contribution to the quantitative analysis of neuronal parameters. The effects of biased sampling procedures on estimates of neuronal volume, surface area and packing density. J. Comp. Neurol. 148:217, 1973.
13. Botterell, E. H., Lloyd, L. A., and Hoffman, H. J.: Oculomotor palsy due to supraclinoid internal carotid artery berry aneurysm. A long-term study of the results of surgical treatments on the recovery of third-nerve function. Am. J. Ophthalmol. 54:609, 1962.
14. Drake, C. G.: Further experience with surgical treatment of aneurysms of basilar artery. J. Neurosurg. 29:372, 1968.
15. Cajal, S. R.: Degeneration and Regeneration of the Nervous System, vol. 1. New York, Hafner Publishing Co., 1959, p. 174.
16. Walsh, F. B.: Third nerve regeneration, a clinical evaluation. Br. J. Ophthalmol. 41:577,1957.
17. Ford, F. R., and Woodhall, B.: Phenomena due to misdirection of regenerating fibers of cranial, spinal and autonomic nerves. Clinical observations. Arch. Surg. 36:480, 1938.
18. Bender, M. B., and Fulton, J. F.: Factors in functional recovery following section of the oculomotor nerve in monkevs. J. Neurol. Psvchiatr. 2:285, 1939.
19. Cogan, D. G.: Neurology of the Ocular Muscles, 2nd ed. Springfield, Charles C Thomas, 1956, p. 1.
20. Hyland, H. H., and Barnett, H. J. M.: The pathogenesis of cranial nerve palsies associated with intracranial aneurysms. Proc. R. Soc. Med. 47:141, 1954.
21. Levin, P. M.: Intracranial aneurysms. Clinic-opathologic considerations of oculomotor nerve regeneration and intracerebral hemorrhage. Arch. Neurol. Psvchiatr. 67:771, 1952.
22. Bjorkman, A., and Wohlfart, G.: Faseranalys der Nn. oculomotorius, trochlearis und abducens des Menschen und des N. abducens verschiedener Tiere. Z. Mikrosk. Anat. Forsch. 39:631, 1936.
VOL. 87, NO. 2 O C U L O M O T O R N E R V E R E G E N E R A T I O N 233
23. Zaki, W.: Le nerf pathetique chez 1'homme. Etude relative a son origine, son trajet intracerebral et a sa structure. Arch. Anat. Histol. Embrvol. 45: 105, 1960.
24. Trautman, J. C : Diseases of the third, fourth and sixth cranial nerves. In Dyck, P. J., Thomas, P. K., and Lambert, E. H. (eds.): Peripheral Neuropathy, vol. 1. Philadelphia, W. B. Saunders, 1975, p. 513.
25. Barratt, J. O. W.: Observations on the structure of the third, fourth and sixth cranial nerves. J. Anat. Phvsiol. 35:214, 1901.
26. Kerr, F. W. L., and Hollowell, O. W.: Location of pupillomotor and accommodation fibers in the oculomotor nerve. Experimental observations on paralytic mydriasis. J. Neurol. Neurosurg. Psvchiatr. 27:473, 1964.
27. Gaskell, W. H.: On the relation between the structure, function, distribution and origin of the cranial nerves; together with a theory of the origin of the nervous system of vertebrata. J. Physiol. (Lond.) 10:153, 1889.
28. Sunderland, S., and Hughes, E. S. R.: The pupillo-constrictor pathway and the nerves to the ocular muscles in man. Brain 69:301, 1946.
29. Kasoff, I., and Kelly, D. L., Jr.: Pupillary sparing in oculomotor palsy from internal carotid aneurysm. J. Neurosurg. 42:713, 1975.
30. Donovan, A.: The nerve fibre composition of the cat optic nerve. J. Anat. 101:1, 1967.
31. Ford, F. R., Walsh, F. B., and King, A.: Clinical observations on the pupillary phenomena resulting from regeneration of the third nerve, with especial reference to the Argyll Robertson pupil. Bull. Johns Hopkins Hosp. 68:309, 1941.
32. Gutmann, E., and Sanders, F. K.: Recovery of fibre numbers and diameters in the regeneration of peripheral nerves. J. Physiol. (Lond.) 101:489,1943.
33. Cragg, B. G., and Thomas, P. K.: The conduction velocity of regenerated peripheral nerve fibres. J. Physiol. (Lond.) 171:164, 1964.
34. Rushton, W. A. H.: A theory of the effects of fibre size in medullated nerve. J. Phvsiol. (Lond.) 115:101,1951.
35. Friede, R. L.: Control of myelin formation by axon caliber (with a model of the control mechanism). J. Comp. Neurol. 144:233, 1972.
36. Friede, R. L., and Martinez, A. J.: Analysis of axon-sheath relations during early Wallerian degeneration. Brain Res. 19:199, 1970.