Brain (1996), 119, 1449-1460

Pathology of acute and chronic ischaemicneuropathy in atherosclerotic peripheral vasculardiseaseHitoshi Nukada,1 Andre M van Rij,2 Stephen G. K. Packer2 and P. Denise McMorran1

Departments of Medicine and 2Surgery, University ofOtago Medical School, Dunedin, New Zealand

Correspondence to: H. Nukada, Department of Medicine,University of Otago Medical School, PO Box 913, GreatKing Street, Dunedin, New Zealand

SummaryThe peripheral nerve pathology in ischaemic limbs withatherosclerotic peripheral vascular diseases (PVD) is difficultto ascertain because of the limited number of reports. Inaddition, it has been debated whether chronic ischaemiaper se could cause morphological abnormalities in peripheralnerves. In this prospective study, we examined pathologicalfindings in the sural, saphenous, deep peroneal, superficialperoneal and tibial nerves, taken from seven acutely andnine chronically ischaemic amputated legs in which ischaemiawas due to non-diabetic severe PVD. For morphologicalcomparison, nerves were also taken from amputated legswithout ischaemic disease and those in which PVD wasassociated with diabetes. In acutely ischaemic nerves,pathological changes were dependent upon the duration ofischaemia. Axonal degeneration of both myelinated andunmyelinated nerve fibres (MFs and UMFs) with occluded

vessels was -prominent, if acute ischaemia was present for>24 h. Focal lesions, a hallmark of acute ischaemicneuropathy, were seen in both acute and chronic PVD nerves.Chronic PVD nerves also revealed considerable variationsin the density of MFs between the fascicles of individualnerves and between the nerves of individual subjects;demyelination and remyelination, endoneurial oedemaparticularly at the subperineurial region, swollen endothelialcells, various but infrequent axonal changes, and relativepreservation of UMFs were also seen. All pathologicalchanges found in acute and chronic PVD nerves, except fora high rate of demyelinated and remyelinated nerve fibres,have been described in experimental models of acuteischaemic/reperfusion injury. Demyelination could be inducedby chronic ischaemia. Thus, pathological alterations inchronic ischaemic neuropathy may be due to the combinedeffects of acute ischaemia/reperfusion and chronic hypoxia.

Keywords: vascular disease; ischaemic neuropathology; ischaemic limb; nerve pathology

Abbreviations: MF = myelinated nerve fibre; PVD = peripheral vascular disease; UMF =• unmyelinated nerve fibre

IntroductionThe term 'ischaemic neuritis' or 'ischaemic neuropathy' isused to describe peripheral nerve lesions associated with andcaused by atherosclerotic occlusive PVD, which is a commonproblem among the elderly (Hutchinson, 1970). The recentEdinburgh Artery Study reported that, in addition to 4.5% ofthe general population aged between 55 and 74 havingintermittent claudication, almost twice that number (8.0%)had abnormal vascular laboratory tests (Ruckley, 1991).Although neuropathic abnormalities are often seen inischaemic limbs with PVD, Asbury (1970) stated that 'perhapsthe most salient feature of this particular type of peripheralnerve disorder resides in the fact that it is almost always

© Oxford University Press 1996

overlooked because of more obvious effects of peripheralvascular insufficiency upon skin and muscle'. The peripheralnerve pathology with PVD is difficult to ascertain becauseof incomplete data (Daube and Dyck, 1984; Chalk andDyck, 1993). It is also uncertain whether chronic ischaemia,insufficient to cause infarction, could induce structuralchanges in peripheral nerves.

Mufson (1952) demonstrated that 37% of 145 PVD subjectshad neuropathic impairment in ischaemic legs. Sensorysymptoms and signs were the most common findings, butreflex changes and muscle weakness were also noted.Hutchinson and Liversedge (1956) suggested that the presence

1450 H. Nukada et al.

of neuropathy was directly related to the severity of PVD,although the most striking feature was the mildness of theneurological abnormalities in spite of advanced PVD. Thehighest frequency of neuropathic deficit in PVD was reportedby Eames and Lange (1967) who found such abnormalitiesin 88% of 32 subjects with severely ischaemic limbs, althoughsome neurological findings may be part of the general ageingprocess. A high frequency of sensory dominant neuropathicsigns (74%) in ischaemic limbs was also reported by Chopraand Hurwitz (1969a). Abnormalities in nerve conduction andEMG have been well recognized in chronic PVD subjects(Miglietta and Lowenthal, 1962; Miglietta, 1966; Chopra andHurwitz, 1969a; D'Amour et al., 1987; Hunter et al., 1988;England et al., 1992, 1995) and in acute PVD (Lachance andDaube, 1991). These abnormalities include slowed conductionvelocity and low amplitude in common peroneal, posteriortibial and sural nerves, and evidence of distal denervation.

In spite of relatively few pathological studies in PVDnerves, various morphological abnormalities have beendescribed such as segmental demyelination and remyelina-tion, axonal degeneration and a loss of nerve fibres. Segmentaldemyelination is probably the most common feature, whileaxonal degeneration is dominant in some cases. Previousreports on nerve pathology in PVD subjects are summarizedin Table 1. Priestley (1931, 1932) was the first to describemorphological changes of peripheral nerves in amputatedischaemic legs with non-diabetic PVD in the Englishliterature. The extent of nerve pathology was directlyproportional to the degree of pathologically observedarteriosclerotic changes in vessels. In cases of thromboangiitisobliterans, patchy nerve fibre degeneration involving certainfascicle or fibres was observed in peroneal and tibial nervesof 17 amputated legs (Barker, 1938).

With the development of a single teased fibre techniqueand electron-microscopic analysis, evidence of a considerableamount of segmental demyelination and remyelinationbecame apparent in chronic PVD nerves (Chopra and Hurwitz,1967; Eames and Lange, 1967). Eames and Lange (1967)also observed (i) a certain amount of axonal degeneration,(ii) that most of unmyelinated nerve fibres (UMFs) appearednormal, (iii) basement membrane thickening and proliferatedendothelial cells in endoneurial capillaries, and (iv)thrombosed epineurial arterioles. In contrast to the previousstudies, Chopra and Hurwitz (1969ft) reported no obviousloss of myelinated nerve fibres (MFs) in chronic PVD nerves.Farinon et al. (1984) found varying severity of pathologicalchanges, although four out of seven subjects were diabetic.Similar heterogeneous pathology in type and severity inchronic PVD nerves was described by Vital et al. (1986). Insuperficial peroneal nerves, they demonstrated that (i) theloss of MFs varied from case to case, (ii) axonal degenerationwas prevalent, (iii) segmental demyelination was notprominent, and (iv) the severity of UMF involvement differedamong cases. These results suggested that variouspathological abnormalities could be found in chronic PVD

nerve, although the factors affecting the variability of nervepathology remain uncertain.

Aims of this prospective study were to demonstratemorphological changes among different peripheral nerves insevere acute and chronic PVD subjects, and to compare thepathological finding in acute PVD nerves with those inchronic PVD. We examined pathological features in sural,saphenous, deep peroneal, superficial peroneal, and tibialnerves, taken from acutely and chronically ischaemicamputated limbs due to non-diabetic PVD. Nerve pathologyin these PVD legs was compared with those in non-ischaemiclegs and in legs in which ischaemia was due to chronic PVDassociated with diabetes. All nerve specimens were obtainedfrom amputated legs.

Material and methodsSubjectsTwenty-two subjects, 14 males and eight females admittedto the Surgery Department, Dunedin Hospital for below kneeamputation between 1989 and 1993 were studied. Seven(mean age, 67.0 years) suffered from severe acute ischaemiaresulting from non-diabetic atherosclerotic occlusive disease(hereafter 'acute PVD'), and nine (mean age, 72.3 years) hadthreatened ischaemic limb due to severe chronic non-diabeticatherosclerotic occlusive disease ('chronic PVD'). Clinicaland laboratory data in acute and chronic PVD subjects aresummarized in Tables 2 and 3. Three acute PVDs werepreceded by the surgery of abdominal aorta aneurysm (Cases1, 2 and 4), and four other had embolism (Cases 3, 5, 6 and7). None of the acute PVD subjects had a history ofintermittent claudication. In acute PVD subjects, there wasno signs of peripheral nerve involvement except one (Case7) who showed mild sensory impairment up to ankle in theleg to be amputated. Control nerves were obtained from threesubjects of a similar age range suffering from sepsis (age:mean 66.3 years, range 62-71 years; hereafter 'non-ischaemiccontrol'). None of these control subjects had a history ofheavy alcoholic intake, diabetes or any other disease whichmight cause vasculopathy or peripheral neuropathy. Forcomparison, nerves in three subjects with chronic PVDassociated with diabetes were also taken (age: mean 74.0years, range 61-83 years; 'diabetic chronic PVD'). Diabeticchronic PVD subjects had clinical and electrophysiologicalevidence of severe polyneuropathy associated with ulcers intheir legs. Two of them had no rest pain. This study wasapproved by the Ethics Committee of the Southern RegionalHealth Authority.

All subjects in chronic PVD presented with complaintsof pain at rest associated with skin necrosis or ulcer intheir legs to be amputated, and had a history of multiplevascular reconstructive interventions in their ischaemic limbsat some time before the below knee amputation. Thesesurgical procedures include endarterectomy, femoro-iliacthrombectomy, femoro-popliteal bypass and aorto-femoral

Pathology in peripheral vascular disease nerves 1451

Table 1 Pathological findings of peripheral nerves in non-diabetic peripheral arterial occlusive diseases

Authors Cases/course Nerves examined* Major pathological findings

Priestly (1931, 1932) 6 chronic

Gairns et at. (1960) 1 chronicGarven et al. (1962) 2 chronic

Eames and Lange (1967) 8 chronic

Chopra and Hurwitz (1967, 1969) 6 chronic

Farinon et al. (1984) 7 chronic5

Vital et al. (1988) 12 chronic

Hunter et al. (1988) 1 chronic

Rodriguez-Sanchez et al. (1991) 8 chronic

Nukadaera/. (1996) 7 acute9 chronic

Posterior tibial nerve at four levels*

Digital nerve of the great toePosterior tibial, plantar, and digitalnerves at several levels

Sural nerve

Sural nerve

Sciatic and tibial nerves

Superficial peroneal nerves

Posterior tibial and sural nerves

Sural nerve

Tibial, sural, saphenous, superficialperoneal and deep peroneal nerves

Degeneration* and fibrosis, moreprominent distallySevere loss of MFsProgressive distal loss of MFs withmarked reduction in digital nerves,vascular swellingDemyelination and remyelination >axonal degeneration, obvious loss ofMFs, vascular swelling, preservation ofUMFsDemyelination and remyelination >axonal degeneration, no loss of MFsDemyelination and remyelination and/oraxonal degeneration, various loss ofMFs, axonal atrophy, preservation ofUMFsAxonal degeneration > demyelination,various loss of MFs, focal lesion, UMFinvolvementMarked loss of MFs, microvascularocclusionVarious types and severities ofpathology, demyelination andremyelination or axonal degeneration,UMF abnormalitiesAcute PVD: focal lesions, axonaldegeneration, occluded vesselsChronic PVD: focal and multifocallesions, various MF densities,remyelination > demyelination >axonal degeneration, endoneurialoedema, vascular swelling, preservationof UMFs

MF = myelinated nerve fibre; UMF = unmyelinated nerve fibre; PVD = peripheral vascular disease. *A11 nerves were taken at the timeof leg amputation, except five nerve biopsy cases by Chopra and Hurwitz. fFrom the popliteal space to the internal malleolus. *The term'degeneration' was applied to nerve fibres with true Wallerian degeneration or fascicle with a definite loss of Mfs. ^Including fourdiabetic subjects.

Table 2 Clinical and laboratory data on seven amputated legs with non-diabetic acute PVD

Caseno.

1234567

Age (years)/sex

82/F79/M69/F79/M62/M58/F55/M

Side

LRLL and RRRR

Duration*

24 h2 days4 days5 days1 month1 month6 months

Rest pain

NoNoYesNoYesYesYes

Aetiology

Post-AAAPost-AAAEmbolismPost-AAAEmbolismEmbolismThrombosis

Site of occlusion

Superficial femoralSuperficial femoralSuperficial femoralPopliteal and tibialExternal iliac and superficial femoralDiffuse, mainly distal smallSuperficial femoral and popliteal

AAA = abdominal aorta aneurysm. *Duration after the onset of ischaemia.

bypass graft. All chronic PVD subjects revealed unilateralor asymmetrical sensory impairment to all modalities inthe legs to be amputated; at regions of deep peroneal orsaphenous nerves in five legs, and stocking distribution upto ankle or lower-calf levels in four. Mild weakness ofdorsiflexion of toes and foot (4/5 MRC grading) on theamputated side was observed in three subjects. Nerveconduction studies of common peroneal and sural nerves

evaluated prior to amputation in seven chronic PVD subjectsshowed a variety of mild to moderate abnormal results, e.g.slowed conduction velocity and reduced amplitude (Table 3).

Pathological assessmentsSural, saphenous, deep peroneal, superficial peroneal, andtibial nerves at the lower-calf to ankle levels were taken

1452 H. Nukada et al.

Table 3 Clinical and laboratory data on nine amputated legs with non-diabetic chronic PVD

Caseno.

89

10111213

141516

Age (years)/sex

83/F73/F72/M75/M65/M65/F

75/M64/M79/M

Side

RRLRLL

LRL

Duration*(years)

63532

12

1.536

Restpain

YesYesYesYesYesYes

YesYesYes

ABIat rest

0.3_t0.400.420.280.22

_t_t0.14

Site of occlusion

PoplitealPopliteal and all vessels in calfSuperficial femoralSuperficial femoralSuperficial femoral and poplitealInternal iliac, superficial femoraland poplitealPoplitealAorta and superficial femoralSuperficial femoral

Sural nerve conduction

INCV+

4-Amplitude^NTlAmplitude§ and lNCVf

-lAmplitude§ and -lNCVf

-iNCV1

•iNCV1

4-NCVtNT

ABI = ankle brachial index; NT = not tested. *History of intermittent claudication; slowed nerve conduction velocity; *unmeasurable;reduced amplitude.

immediately after amputation. Nerve processing for ' '. ^ ppathological evaluation has been detailed previously (Nukadaet al., 1989). In brief, nerve specimens for plastic sectionswere fixed with 2.5% glutaraldehyde in 0.025 M cacodylatebuffer, pH 7.4, at room temperature for 24 h. Nerves werecut into smaller pieces (1-2 mm in length), then postfixedin 2% osmium tetroxide for 24 h, dehydrated in serialalcohols and propylene oxide, and embedded in Araldite.Transverse sections, 1.0 |im in thickness, were cut andstained with methylene blue, thionin and acridine orange,and phenylenediamine. Ultra-thin sections were stained withuranyl acetate followed by lead citrate. Nerve segments forteased fibre analysis were fixed for 30 min in glutaraldehyde,and then postfixed in 2% osmium tetroxide for 8 h. Aminimum of 100 single fibres were prepared and graded(Dyck et al., 1993). Morphometric measurements of MFs ontransverse semi-thin sections were performed directly fromslides with the use of a SAMBA 2005 computerized imageanalysis system (TITN, Grenoble, France) (Wright andNukada, 1994). The coefficient of variation and the indexof dispersion of MF density (mm"2) were employed forrecognition of variability of MF density within and betweenfascicles (Dyck et al., 1984). Data were expressed asmean±SD, and P values <0.05 were considered significantusing the two-tailed Student's t test.

Fig. 1 Transverse section of the superficial peroneal nerve takenfrom Case 4 who had acute PVD showing panfascicular axonaldegeneration with occluded vessels in endo-, peri- and epineurialregions. Plastic-embedded I um thick section, thionin andacridine orange staining. Bar = 100 urn.

myelin were also noted in these three cases, although lessnoticeable when compared with those in chronic PVD nerves.

ResultsAcute PVD nervesPathological changes in acutely ischaemic nerves weredependent on the duration of ischaemia. Teased fibres inCase 1 whose nerves were taken after 24 h of ischaemiarevealed <5% of MFs with axonal degeneration. Axonaldegeneration of both MFs and UMFs was conspicuous ifacute ischaemia was present for 48 h. Panfascicular or centralfascicular fibre degeneration with occluded vessels wasprominent in Cases 2, 3 and 4 (Fig. 1). Nerves taken at 1-6months after the onset of acute limb ischaemia showedvariable MF densities with regenerated fibres (Cases 5, 6 and7). Endoneurial oedema and MFs with abnormally thin

Chronic PVD nervesFocal lesions such as selective damage at central orsubperineurial fascicular regions were observed in all tibialnerves and most of the sural, saphenous and superficialperoneal nerves (Figs 2—4). Central fascicular lesions weremore prominent in large fascicles than smaller fascicles,whereas subperineurial lesions were observed in both largeand small fascicles. These focal lesions were seen either tobe due to loss of nerve fibres (Fig. 2) or were filled withregenerating fibres (Fig. 3). The variability of MF numberswas obvious between the fascicles of the individual nerves(Fig. 2), and between the nerves of the individual subjects(Fig. 5). A complete loss of MFs was found only in onesaphenous nerve taken from Case 10 (Fig. 5A), while MF

Pathology in peripheral vascular disease nerves 1453

Fig. 2 Pathological appearance of the tibial nerve at the lower-thigh level taken from Case 11, who hadchronic PVD, showing the variation of the density of myelinated nerve fibres. A central fascicularlesion is seen in the left fascicle, and the number of myelinated nerve fibres is relatively preserved inthe right lower fascicle when compared with other fascicles. Note endoneurial oedema, at thesubperineurial region in particular. This tibial nerve contains 21 fascicles. Plastic-embedded 1 |im thicksection, methylene blue stain. Bar = 200 (im.

Fig. 3 Central fascicular lesion filled with regenerating nerve fibres in the tibial nerve from Case 10who had chronic PVD. Note subperineurial oedema and myelinated nerve fibres with vesicular myelinswelling. Plastic-embedded 1 ,um thick section, methyler.e blue stain. Bar ~ 100 )im.

1454 H. Nukada et al.

\

Fig. 4 Subperineurial lesion associated with subperineurial oedema in the saphenous nerve from Case 12 who had chronic PVD. Plastic-embedded 1 urn thick section, thionin and acridine orange staining. Bar = 30 um.

densities of other nerves in this subject were only moderatelyreduced (Fig. 5B).

Nerve fibres with disproportionately thin myelin relativeto axon area were commonly found (Fig. 6). In the teasednerve study, the frequency of remyelinated fibres wassignificantly greater in chronic PVD nerves than in non-ischaemic controls for all the nerves examined except thesaphenous nerve (Table 4). Demyelinated nerve fibres werealso observed (Fig. 6C) and teased fibre analysis demonstratedthat demyelination was more frequently seen in chronic PVDnerves than in controls, but statistically significant differencewas reached only in the sural nerve (Table 4). Myelinatednerve fibres with myelin infolding and clefts were oftenobserved particularly in the fascicles with severe reductionin the number of MFs.

Subperineurial oedema was another prominent feature(Figs 2-4), although not observed in all the fasciclesevaluated. Nerves, which showed reduced amplitude of nerveaction potential electrophysiologically (Table 3), revealedparticularly severe endoneurial oedema. Mean transversefascicular areas of each nerve were not significantly differentbetween chronic PVD nerves and controls (6.1 ±0.9 mm2 and5.3±1.2 mm2 for tibial nerves, respectively). Intramyelinicoedema was also frequently found in chronic PVD nerves(Figs 3 and 5B). Nerve fibres showing axonal changes wereless commonly seen, and the teased fibre study confirmedthe infrequency of axonal degeneration (Table 4). However,darkly stained swollen axons or attenuated axons wereobserved in chronic PVD nerves (Fig. 7). Unmyelinatednerve fibres appeared to be normal, even in nerves revealinga definite loss of MFs, except the saphenous nerve, which

showed a complete loss of MFs, had degenerated UMFs.Endoneurial vessels exhibited swollen endothelial cells.Results of vascular morphometry will be reported separately.

Mean MF densities (mm"2) in chronic PVD nerves weremoderately but significantly reduced when compared withthose in non-ischaemic control nerves (Table 5). The indexof dispersion and the coefficient of variation of MF densitiesbetween the sampled frames within the fascicle and betweenthe fascicles within the same nerve were significantly greaterin chronic PVD nerves than in non-ischaemic controls: fortibial nerves, mean coefficient of variation and index ofdispersion of MF density among the frames was 45.5 ± 10.6%in chronic PVD and 30.6±6.2 in controls {P < 0.01) and2.5±0.7 in chronic PVD and 1.4±0.5 in controls (P < 0.02),respectively; and similar tends were noted for the coefficientof variation and the index of dispersion of MF density amongthe fascicles.

Control nervesNon-ischaemic controlsNeither focal nor multifocal pathology was seen. MFswith disproportionately thin myelin were seen and teasedfibre study demonstrated that ~20-30% of MFs revealedremyelination, presumably due to part of the general ageingprocess in peripheral nerve. Axonal degeneration or nerveoedema was not commonly found.

Diabetic chronic PVDNerves in diabetic chronic PVD exhibited a marked diffuseloss of nerve fibres in all the nerves examined. In particular,

Pathology in peripheral vascular disease nerves 1455

Fig. 5 Transverse sections of the saphenous (A) and sural (B) nerves from Case 10 with chronic PVD showing distinct contrast ofmyelinated nerve fibre density between the two nerves. The saphenous nerve (A) reveals a complete loss of myelinated nerve fibres,whereas the sural nerve (B) exhibits only moderate reduction in the number of myelinated nerve fibres. Note a myelinated nerve fibrewith intramyelinic oedema in the sural nerve (arrow). Plastic-embedded section, thionin and acridine orange staining (A) and methyleneblue stain (B). Bars = 30 |im.

nerves taken from two subjects who did not complain ofpain at rest revealed a complete loss of both MFs and UMFs.As expected, severe vascular changes such as basementmembrane reduplication were observed.

DiscussionPathological nerve changes in acute PVD are dependent onthe duration of ischaemia. When pathological changes becomeapparent, acute PVD nerves showed panfascicular or focallesions with degeneration of both MFs and UMFs. Majorpathological findings in chronic PVD nerves are focal lesionssuch as selective damage of the central fascicular orsubperineurial region, considerable variability between MFdensities for individual fascicles of one nerve or individualnerves of one subject, demyelination and remyelination,endoneurial oedema, endothelial swelling and relativepreservation of UMFs. Axonal changes were less frequentlyfound in chronic PVD nerves. These pathological changes,except for a high rate of demyelination and remyelination,

have been documented in experimental acute ischaemicneuropathy (Nukada and Dyck, 1984; Benstead et al., 1990;Nukada and McMorran, 1994), suggesting that mostmorphological alterations in chronic PVD nerves may becaused by acute ischaemia/reperfusion.

Experimentally, acute nerve ischaemia induces a localizedlesion of axonal degeneration especially of nerve fibresat the central fascicular region ('central fascicular fibredegeneration' or 'ischaemic core') (Korthals and Wisniewski,1975; Hess et al., 1979; Parry and Brown, 1981; Nukadaand Dyck, 1984). In the microembolization model of acuteischaemic neuropathy, an 'ischaemic core' was seen in largefascicles such as sciatic or tibial nerves, but not in peronealor sural nerves which contain smaller fascicles (Nukada andDyck, 1984). In PVD nerves an 'ischaemic core' was foundalso in large fascicles rather than in small fascicles. Centralfascicular vulnerability to ischaemia, therefore, could dependon the fascicular size. Localized ischaemic damage atthe subperineurial region, contrary to central fascicularfibre degeneration, is described as 'subperineurial fibre

1456 H. Nukada et al.

O.-co

•> oO o OtfC

OC

G

o

Fig. 6 (A) The sural nerves from Case 13, who had chronic PVD, showing myelinated nerve fibres with disproportionately thin myelinlayers relative to axon area. Plastic-embedded 1 |im thick section, thionin and acridine orange staining. Bar = 20 (im. (B and C)Electron micrographs of the peroneal nerve from Case 13 illustrating thinly myelinated nerve fibres (B) and a demyelinated fibre (C).Lead citrate and uranyl acetate staining. Bars = 5 \xm.

degeneration' (Nukada and Dyck, 1984; Korthals andKorthals, 1990; Myers et al, 1991; Nukada et al., 1993).These focal lesions cause an increase in variability of MFdensity which is confirmed by assessing the coefficient ofvariation and the index of dispersion of MF density. Severalweeks after the onset of ischaemia, regenerating nerve fibresbecome apparent in the localized lesion (Nukada and Dyck,1986; Korthals and Korthals, 1990). These spatial

distributions of acute ischaemic lesions have been describedin human vasculitic (Dyck et al., 1972), diabetic (Sugimuraand Dyck, 1982; Dyck et al, 1986a, b; Johnson et al., 1986;Said et al, 1994) and chronic multifocal demyelinatingneuropathies (Nukada et al., 1989).

If acute ischaemia is implicated in the pathological changeof chronic PVD nerves, reperfusion can also play a role inits development. Ischaemic injury could be exaggerated

Pathology in peripheral vascular disease nerves 1457

Table 4 Frequency of various pathological conditions in single teased fibres from nerves taken from non-diabetic chronicPVD and control (%, mean±SD)

Nerves

SuralSaphenousSuperficial peronealDeep peronealTibial

Normal

Chronic PVD

48.8±6.5f

48.4±5.3f

51.2±13.558.3±10.1*54.8±6.8f

Controls

78.2±10.270.2±4.367.2±9.777.3+10.177.6±3.6

Demyelination with orwithout remyelination*

Chronic PVD

9.2±4.0§

12.2+8.74.2±5.93.4±4.76.9±3.0

Controls

2.5±3.43.1+2.87.8±4.01.6±1.43.1±2.7

Remyelination*

Chronic PVD

37.0±10.6*26.8±6.642.6±10.5*32.9±8.5*35.7±9.7*

Controls

18.9+8.525.8±5.222.7±3.417.3+4.718.8±4.5

Axonal degeneration*

Chronic PVD

4.9+8.812.6+15.32.1+3.05.6+7.82.8+3.9

Controls

0.4±0.70.9±1.52.2±3.93.8±4.71.6±1.9

*Dyck's grading: C+D = demyelination with/without remyelination; F = remyelination; E+H = axonal degeneration (Dyck et al,1993); ip < 0.001; *0.01 < P< 0.025; §0.025 < P <0.05.

with occluded lumen and thickening of basement membrane(Garven et al., 1962; Eames and Lange, 1967). It is of notethat some vessels in the present study exhibited reduplicationof basement membrane without evidence of diabetes.

Chronic PVD subjects have a history of intermittentclaudication and of multiple vascular surgical interventionswhich could induce acute ischaemic/reperfusion injury toperipheral nerve. Intermittent claudication has been shownto activate neutrophils, increase leucocyte-endothelialadhesion and vascular permeability, and decrease antioxidantlevels (Hickey et al, 1993; Edwards et al, 1994; Khairaet al, 1995). Ischaemic/reperfusion injury associated withvascular reconstruction has been evaluated clinically andexperimentally, although all the studies have focused onsystemic or muscle involvement rather than peripheral nerve(Beyersdorf et al, 1989; Hoch et al, 1991; Adiseshiah et al,1992; Sternbergh and Adelman, 1992). Post-ischaemic rathindlimb exhibited a transient, early burst of tumour necrosisfactor (Sternbergh et al, 1994). Chervu et al (1989)demonstrated that electrophysiologically, peripheral nerve ismore susceptible to ischaemia and short reperfusion intervalsthan skeletal muscle. In ischaemic limbs with PVD, reper-fusion may also occur through collateral channels. Thus,transient ischaemic events including multiple vascularreconstructive interventions may represent a relativelycommon insult in chronic PVD nerves.

Myelinated fibres with disproportionately thin myelinlayers relative to axon area are commonly observed in chronicPVD nerves. Teased nerve fibre study confirmed a significantincrease in the number of remyelinated fibres in chronic PVDnerves. A high rate of demyelination and remyelination hasbeen documented previously in PVD nerves (Table 1).Although demyelination has been reported in the models ofacute ischaemic neuropathy, such a high frequency of primarydemyelination and remyelination has never been induced byan episode of acute nerve ischaemia/reperfusion (Hess et al.,1979; Nukada and Dyck, 1987; Nukada et al, 1993; Nukadaand McMorran, 1994). Demyelination in chronic PVD nerves,therefore, may be due to chronic hypoxia. In contrast toacute ischaemic neuropathy, it has been debated whetherchronic ischaemia, insufficient to cause infarction, inducespathological changes in peripheral nerve. Demyelinated and

Fig. 7 Transverse sections of the sural nerve from Case 10showing darkly stained swollen axon (arrow), attenuated or emptyaxons, and demyelinated nerve fibre. Inset, high magnification ofattenuated axon and demyelinated axon from the boxed region.Plastic-embedded 1 urn thick section, methylene blue stain.Bars = 30 \xm and 10 urn (inset).

Table 5 Density of myelinated nerve fibres in variousnerves taken from non-diabetic chronic PVD (numbermm'2, mean±SD)

Nerves Chronic PVD Controls

SuralSaphenousSuperficial peronealDeep peronealTibial

3550±656*2I61±1595*3254±741f

2893 ±807*3450±998f

5316±15985327 ±14725233±9346181 =t 15106414±1521

*0.025 < P < 0.05; f0.01 < P < 0.025; tP < 0.001.

by subsequent reperfusion, and the peripheral nerve is noexception (Day et al., 1989; Schmelzer et al., 1989). Weobserved endothelial swelling, endoneurial and intramyelinicoedema and demyelination in ischaemic/reperfused rathindlimb nerves (Nukada and McMorran, 1994); chronicPVD nerves exhibited these same morphological features.Endoneurial vessels in chronic PVD have been investigatedby other investigators showing swollen endothelial cells

1458 H. Nukada et al.

remyelinated nerve fibres have been observed frequentlyin diabetic subjects whose nerves are chronically hypoxic(Thomas and Lascelles, 1965, 1966; Ohnishi et. al., 1983;Newrickefa/., 1986; Thomas and Tomlinson, 1993). Systemichypoxia secondary to chronic obstructive pulmonary diseaseis associated with neuropathy resulting in slowed nerveconduction and demyelination (Malik et al, 1990). Sladkyet al. (1991) developed the model of chronic ischaemicneuropathy using a femoral arterio-venous shunt in rats. After10 months of 50-70% reduction in nerve blood flow, noevidence of segmental demyelination, axonal degenerationor fibre loss was found in sciatic and tibial nerves. Pathologicalchanges were confined to nodes of Ranvier associated withslowed nerve conduction. Conventional experimental modelsof chronic ischaemic neuropathy such as galactosaemia,(although the concomitant metabolic derangement may alsobe injurious), suggested that chronic hypoxia might causedemyelination (Powell and Myers, 1983; Myers and Powell,1984; Lower al., 1985).

Axonal changes found in chronic PVD nerves, e.g. darklystained swollen axons and attenuated axons, have beendescribed in acute nerve ischaemia, particularly at the borderzone of an 'ischaemic core' (Korthals et al., 1978; Nukadaand Dyck, 1984, 1987). Darkly stained swollen axons withdisproportionately thin myelin have been reported insuperficial peroneal nerves taken from amputated legs inarteritic diabetic patients (Vital et al., 1983). Axonal changeswere more prominent in acute PVD nerves when comparedwith those in chronic PVD nerves. Experimentally, theseverity rather than the duration of ischaemia is moresignificant in inducing structural changes in peripheral nerve(Sladky et al., 1991; Nukada et al., 1993). Our chronic PVDsubjects with a longer history of intermittent claudication didnot reveal any different pathology in type and severity whencompared with those with a shorter clinical history. Indeed,Eames and Lange (1967) stressed that severity of ischaemiarather than the duration of ischaemia could play a key rolein the development of neuropathy in chronic PVD.

What is the underlying mechanism of a complete MF lossin one saphenous nerve? This saphenous nerve revealedextensive degeneration of UMFs, whereas other chronic PVDnerves showed relative preservation of UMFs. Saphenousneuralgia has been described as a complication of vascularreconstructive surgery, possibly due to direct trauma to thenerve during operation (Adar et al., 1979; R0der et al, 1984).The femoral nerve could also be damaged by direct traumaduring the insertion of the arterial catheter. Thus, thissaphenous nerve lesion is most likely due to direct mechanicalinjury. An asymptomatic lesion of femoral nerve, concomitantwith sciatic neuropathy, has recently been reported as one ofcomplications of vascular surgery (McManis, 1994).

Are there any differences in morphological vulnerabilityto ischaemia between motor and sensory nerves? This issue,at least in humans, is not resolved. Relative sparing of UMFsin chronic PVD, in contrast to a severe loss of both MFsand UMFs in diabetic PVD nerve, is most likely related to

their greater morphological resistance to chronic ischaemia.In the model of acute ischaemic neuropathy, UMFs and smallMFs were found to be less vulnerable than large MFsmorphologically (Fujimura et al., 1991). Post-ischaemicparaesthesia has been well documented and sensory axonsare more prone than motor axons to developing ectopicactivity during and after release of nerve ischaemia(Merrington and Nathan, 1949; Poole, 1956). Bostock et al.(1994) has recently confirmed more marked ischaemicsusceptibility in sensory axons than in motor axons. Acomparative study of motor and sensory nerve conductionin PVD subjects with intermittent claudication showed asignificant reduction in amplitude of the sensory actionpotential in the lateral popliteal nerve, but no significantchanges in any other nerve conduction parameters, suggestingthat sensory nerve may be more susceptible to ischaemiathan motor nerves (Chopra and Hurwitz, 1969a). However,the underlying mechanisms of physiological susceptibility toischaemia could be different from those of morphologicalischaemic vulnerability (Nukada, 1993).

AcknowledgementsThis study was supported by the Health Research Councilof New Zealand, the New Zealand Lottery Grants Board andthe New Zealand Neurological Foundation.

ReferencesAdar R, Meyer E, Zweig A. Saphenous neuralgia. A complicationof vascular reconstructions below the inguinal ligament. Ann Surg1979; 190: 609-13.

Adiseshiah M, Round JM, Jones DA. Reperfusion injury in skeletalmuscle: a prospective study in patients with acute limb ischaemiaand claudicants treated by revascularization [see comments]. Br JSurg 1992; 79: 1026-9. Comment in: Br J Surg 1993; 80: 401-2.

Asbury AK. Ischemic disorders of peripheral nerve. In: VinkenPJ, Bruyn GW, editors. Handbook of clinical neurology. Vol. 8.Amsterdam: North-Holland, 1970: 154-64.

Barker NW. Lesions of peripheral nerves in thromboangiitisobliterans: a clinicopathologic study. Arch IntMed 1938; 62: 271-84.

Benstead TJ, Sangalang VE, Dyck PJ. Acute endothelial swellingis induced in endoneurial microvessels by ischemia. J Neurol Sci1990; 99: 37^*9.

Beyersdorf F, Matheis G, Kriiger S, Hanselmann A, FreislebenHG, Zimmer G, et al. Avoiding reperfusion injury after limbrevascularization: experimental observations and recommendationsfor clinical application. [Review]. J Vase Surg 1989; 9: 757-66.

Bostock H, Burke D, Hales JP. Differences in behaviour of sensoryand motor axons following release of ischaemia. Brain 1994: 117:225-34.

Chalk CH, Dyck PJ. Ischemic neuropathy. In: Dyck PJ, ThomasPK, editors. Peripheral neuropathy. 3rd ed. Philadelphia: W. B.Saunders, 1993: 980-9.

Pathology in peripheral vascular disease nerves 1459

Chervu A, Moore WS, Homsher E, Quinones-Baldrich WJ.Differential recovery of skeletal muscle and peripheral nervefunction after ischemia and reperfusion [published erratum appearsin J Surg Res 1989; 47: 470]. J Surg Res 1989; 47: 12-19.

Chopra JS, Hurwitz LJ. Internodal length of sural nerve fibres inchronic occlusive vascular disease. J Neurol Neurosurg Psychiatry1967; 30: 207-14.

Chopra JS, Hurwitz LJ. A comparative study of peripheral nerveconduction in diabetes and non-diabetic chronic occlusive peripheralvascular disease. Brain 1969a; 92: 83-96.

Chopra JS, Hurwitz LJ. Sural nerve myelinated fibre density andsize in diabetics. J Neurol Neurosurg Psychiatry 1969b; 32: 149-54.

D'Amour ML, Lebrun LH, Rabbat A, Trudel J, Daneault N.Peripheral neurological complications of aortoiliac vascular disease.Can J Neurol Sci 1987; 14: 127-30.

Daube JR, Dyck PJ. Neuropathy due to peripheral vascular diseases.In: Dyck PJ, Thomas PK, Lambert EH, Bunge R, editors. Peripheralneuropathy. 2nd ed. Philadelphia: W. B. Saunders, 1984: 1458-78.

Day TJ, Schmelzer JD, Low PA. Aortic occlusion and reperfusionand conduction, blood flow, and the blood-nerve barrier of ratsciatic nerve. Exp Neurol 1989; 103: 173-8.

Dyck PJ, Conn DL, Okazaki H. Necrotizing angiopathic neuropathy:three-dimensional morphology of fiber degeneration related to sitesof occluded vessels. Mayo Clin Proc 1972; 47: 461-75.

Dyck PJ, Karnes J, O'Brien P, Nukada H, Lais A, Low P.Spatial pattern of nerve fiber abnormality indicative of pathologicmechanism. Am J Pathol 1984; 117: 225-38.

Dyck PJ, Karnes JL, O'Brien P, Okazaki H, Lais A, Engelstad J.The spatial distribution of fiber loss in diabetic polyneuropathysuggests ischemia. Ann Neurol 1986a; 19: 440-9.

Dyck PJ, Lais A, Karnes JL, O'Brien P, Rizza R. Fiber loss isprimary and multifocal in sural nerves in diabetic polyneuropathy.Ann Neurol 1986b; 19: 425-39.

Dyck PJ, Giannini C, Lais A. Pathologic alterations of nerves. In:Dyck PJ, Thomas PK, editors. Peripheral neuropathy. 3rd ed.Philadelphia: W. B. Saunders, 1993: 514-95.

Eames RA, Lange LS. Clinical and pathological study of ischaemicneuropathy. J Neurol Neurosurg Psychiatry 1967; 30: 215-26.

Edwards AT, Blann AD, Suarez-Mendez VJ, Lardi AM, McCollumCN. Systemic responses in patients with intermittent claudicationafter treadmill exercise. Br J Surg 1994; 81: 1738-41.

England JD, Regensteiner JG, Ringel SP, Carry MR, Hiatt WR.Muscle denervation in peripheral arterial disease. Neurology 1992;42: 994-9.

England JD, Ferguson MA, Hiatt WR, Regensteiner JG. Progressionof neuropathy in peripheral arterial disease. Muscle Nerve 1995;18: 380-7.

Farinon AM, Marbini A, Gemignani F, Govoni E, Bragaglia MM,Sianesi M, et al. Skeletal muscle and peripheral nerve changescaused by chronic arterial insufficiency—significance and clinicalcorrelations—histological, histochemical and ultrastructural study.Clin Neuropathol 1984; 3: 240-52.

Fujimura H, Lacroix C, Said G. Vulnerability of nerve fibres toischaemia: a quantitative light and electron microscope study. Brain1991; 114: 1929-42.

Gairns FW, Garven HSD, Smith G. The digital nerves and the nerveendings in progressive obliterative vascular disease of the leg. ScotMed J 1960; 5: 382-91.

Garven HSD, Gairns FW, Smith G. The nerve fibre populations ofthe nerves of the leg in chronic occlusive arterial disease in man.Scot Med J 1962; 7: 250-65.

Hess K, Eames RA, Darveniza P, Gilliatt RW. Acute ischaemicneuropathy in the rabbit. J Neurol Sci 1979; 44: 19—43.

Hickey NC, Hudlicka O, Gosling P, Shearman CP, Simms MH.Intermittent claudication incites systemic neutrophil activation andincreased vascular permeability. Br J Surg 1993; 80: 181—4.

Hoch JR, Stevens RP, Keller MP, Silver D. Recovery ofneuromuscular function during reperfusion of the ischemicextremity: effect of mannitol and superoxide dismutase. Surgery1991; 110: 656-63.

Hunter GC, Song GW, Nayak NN, Zapotowski D, GuernseyJM. Peripheral nerve conduction abnormalities in lower extremityischemia: the effects of revascularization. J Surg Res 1988; 45:96-103.

Hutchinson EC. Ischaemic neuropathy and peripheral vasculardisease. In: Vinken PJ, Bruyn GW, editors. Handbook of clinicalneurology, Vol. 8. Amsterdam: North-Holland, 1970: 149-53.

Hutchinson EC, Liversedge LA. Neuropathy in peripheral vasculardisease: its bearing on diabetic neuropathy. Quart J Med 1956; 25:267-74.

Johnson PC, Doll SC, Cromey DW. Pathogenesis of diabeticneuropathy. Ann Neurol 1986; 19: 450-7.

Khaira HS, Maxwell SRJ, Shearman CP. Antioxidant consumptionduring exercise in intermittent claudication. Br J Surg 1995: 82;1660-2

Korthals JK, Korthals MA. Distribution of nerve lesions in serotonin-induced acute ischemic neuropathy. Acta Neuropathol (Berl) 1990;81: 20-4.

Korthals JK, Wisniewski HM. Peripheral nerve ischemia: part 1.Experimental model. J Neurol Sci 1975; 24: 65-76.

Korthals JK, Korthals MA, Wisniewski HM. Peripheral nerveischemia: part 2. Accumulation of organelles. Ann Neurol 1978; 4:487-98.

Lachance DH, Daube JR. Acute peripheral arterial occlusion:electrophysiologic study of 32 cases. Muscle Nerve 1991; 14: 633-9.

Low PA, Nukada H, Schmelzer JD, Tuck RR, Dyck PJ. Endoneurialoxygen tension and radial topography in nerve edema. Brain Res1985; 341: 147-54.

Malik RA, Masson EA, Sharma AK, Lye RH, Ah-See AK, ComptonAM, et al. Hypoxic neuropathy: relevance to human diabeticneuropathy. Diabetologia 1990; 33: 311-18.

McManis PG. Sciatic nerve lesions during cardiac surgery.Neurology 1994; 44: 684-7.

1460 H. Nukada et al.

Merrington WR, Nathan PW. A study of post-ischaemicparaesthesiae. J Neurol Neurosurg Psychiatry 1949; 12: 1-18.

Miglietta O. Nerve motor fiber characteristics in chronic ischemia.Arch Neurol 1966; 14: 448-53.

Miglietta O, Lowenthal M. Nerve conduction velocity and refractoryperiod in peripheral vascular disease. J Appl Physiol 1962; 17:837^0.

Mufson I. Diagnosis and treatment of neural complications ofperipheral arterial obliterative disease. Angiology 1952; 3: 392-6.

Myers RR, Powell HC. Galactose neuropathy: impact of chronicendoneurial edema on nerve blood flow. Ann Neurol 1984; 16:587-94.

Myers RR, Heckman HM, Galbraith JA, Powell HC. Subperineurialdemyelination associated with reduced nerve blood flow and oxygentension after epineurial vascular stripping. Lab Invest 1991; 65:41-50.

Newrick PG, Wilson AJ, Jakubowski J, Boulton AJM, Ward JD.Sural nerve oxygen tension in diabetes. Br Med J 1986; 293: 1053-^.

Nukada H. The susceptibility of rat diabetic nerve to ischemia:increased or decreased? J Neurol Sci 1993; 119: 162-8.

Nukada H, Dyck PJ. Microsphere embolization of nerve capillariesand fiber degeneration. Am J Pathol 1984; 115: 275-87.

Nukada H, Dyck PJ. Neovascularization after ischemic nerve injury.Exp Neurol 1986; 92: 391-7.

Nukada H, Dyck PJ. Acute ischemia causes axonal stasis, swelling,attenuation, and secondary demyelination. Ann Neurol 1987; 22:311-18.

Nukada H, McMorran PD. Perivascular demyelination andintramyelinic oedema in reperfusion nerve injury. J Anat 1994; 185:259-66.

Nukada H, Pollock M, Haas LF. Is ischemia implicated in chronicmultifocal demyelinating neuropathy? [see comments]. Neurology1989; 39: 106-10. Comment in: Neurology 1989; 39: 1270.

Nukada H, Powell HC, Myers RR. Spatial distribution of nerveinjury after occlusion of individual major vessels in rat sciaticnerves. J Neuropathol Exp Neurol 1993; 52: 452-9.

Ohnishi A, Harada M, Tateishi J, Ogata J, Kawanami S. Segmentaldemyelination and remyelination in lumbar spinal roots of patientsdying with diabetes mellitus. Ann Neurol 1983; 13: 541-8.

Parry GJ, Brown MJ. Arachidonate-induced experimental nerveinfarction. J Neurol Sci 1981; 50: 123-33.

Poole EW. Ischaemic and post-ischaemic paraesthesiae inpolyneuritis: normal responses in the upper limb with specialreference to the effect of age. J Neurol Neurosurg Psychiatry 1956;19: 148-54.

Powell HC, Myers RR. Schwann cell changes and demyelinationin chronic galactose neuropathy. Muscle Nerve 1983: 6: 218-27.

Priestley JB. The histopathology of peripheral nerves removed fromextremities amputated for arteriosclerotic gangrene. Proc Staff MeetMayoClin 1931; 6: 517-18.

Priestley JB. Histopathologic characteristics of peripheral nerves in

amputated extremities of patients with arteriosclerosis. J Nerv MentDis 1932; 75: 137-43.

R0der OC, Kamper A, J0rgensen SJ. Incidence of saphenousneuralgia in arterial surgery. Acta Chir Scand 1984; 150: 2 3 ^ .

Rodriguez-Sanchez C, Medina Sanchez M, Malik RA, Ah-See AK,Sharma AK. Morphological abnormalities in the sural nerve frompatients with peripheral vascular disease. Histol Histopathol 1991;6: 63-71.

Ruckley CV. Symptomatic and asymptomatic disease. In: FowkesFGR, editor. Epidemiology of peripheral vascular disease. London:Springer-Verlag, 1991: 97-108.

Said G, Goulon-Goeau C, Lacroix C, Moulonguet A. Nerve biopsyfindings in different patterns of proximal diabetic neuropathy. AnnNeurol 1994; 35: 559-69.

Schmelzer JD, Zochodne DW, Low PA. Ischemic and reperfusioninjury of rat peripheral nerve. Proc Natl Acad Sci USA 1989; 86:1639-42.

Sladky JT, Tschoepe RL, Greenberg JH, Brown MJ. Peripheralneuropathy after chronic endoneurial ischemia. Ann Neurol 1991;29: 272-8.

Sternbergh WC 3d, Adelman B. The temporal relationship betweenendothelial cell dysfunction and skeletal muscle damage afterischemia and reperfusion. J Vase Surg 1992; 16: 30-9.

Sternbergh WC 3d, Tuttle TM, Makhoul RG, Bear HD, Sobel M,Fowler AA 3d. Postischemic extremities exhibit immediate releaseof tumor necrosis factor. J Vase Surg 1994; 20: 474—81.

Sugimura K, Dyck PJ. Multifocal fiber loss in proximal sciaticnerve in symmetric distal diabetic neuropathy. J Neurol Sci 1982;53: 501-9.

Thomas PK, Lascelles RG. Schwann-cell abnormalities in diabeticneuropathy. Lancet 1965; 1: 1355-7.

Thomas PK, Lascelles RG. The pathology of diabetic neuropathy.Quart J Med 1966; 35: 489-509.

Thomas PK, Tomlinson DR. Diabetic and hypoglycemic neuropathy.In: Dyck PJ, Thomas PK, editors. Peripheral neuropathy. 3rd ed.Philadelphia: W. B. Saunders, 1993: 1219-50.

Vital C, Brechenmacher C, Sense JM, Bellance R, Vital A, DartiguesJF, et al. Ultrastructural study of peripheral nerve in arteritic diabeticpatients. Acta Neuropathol (Berl) 1983; 61: 225-31.

Vital A, Vital C, Brechenmacher C, Serise JM, Callen S, NicolauH, et al. Quantitative, histological and ultrastructural studies ofperipheral nerve in arteriosclerotic non-diabetic patients. ClinNeuropathol 1986; 5: 224-9.

Wright A, Nukada H. Sciatic nerve morphology and morphometryin mature rats with streptozocin-induced diabetes. Acta Neuropathol(Berl) 1994; 88: 571-8.

Received January 6, 1996. Revised March 15, 1996.Accepted April 19, 1996