human joint and intervertebral - annals of the rheumatic

Annals of the Rheumatic Diseases 1995; 54: 182-188

Human lumbar apophyseal joint damage andintervertebral disc degeneration

MW Swanepoel, L M Adams, J E Smeathers

AbstractObjectives-To record the extent andlocation of lumbar apophyseal cartilagedamage, and to ascertain if the extent ofdamage is correlated with the grade ofdisc degeneration, age, or both.Methods-The extent and location offibrillated areas of the apophysealcartilage ofthe joint surfaces of29 lumbarmotion segments were examined usingcomputer aided image processing ofIndian ink stained areas, and degener-ation ofthe associated intervertebral discsgraded using the method ofNachemson.Results-It was found that these jointsshowed a greater extent and prevalence ofcartilage fibrillation than the knee, hip orankle, with significant damage in speci-mens younger than 30 years. Damagewas predominantly located peripherally,superiorly, and posteriorly in the concavesuperior apophyseal surfaces, and waspredominantly peripheral and posteriorin the inferior surfaces, with a tendencyto be located inferiorly. There was aweak correlation between apophysealjoint damage and the intervertebraldisc degenerative grade, but this wasinconclusive, as both increased with age.Conclusions-The pattern of damageexhibited by superior joint surfaces ismost probably caused by tension oncollagenous joint capsule fibres whichinsert into the surfaces posteriorly, soproducing an area of fibrocartilageunsuited to loadbearing. Tension on suchfibres would be greatest during spinalflexion. The pattern of damage of theinferior surfaces lends some support to thehypothesis that their apices impact thelaminae ofthe lumbar vertebra inferior tothem, consequent upon the degenerationand narrowing of the associated inter-vertebral disc. The predominantlyperipheral location of fibrillation of bothsuperior and inferior surfaces may beassociated with inadequate mechanicalconditioning ofmarginal joint areas. Discdegeneration cannot be the initial cause ofapophyseal fibrillation in most specimens.The study indicates a need for regularspinal exercise, starting at a young age.

(Ann Rheum Dis 1995; 54: 182-188)

The apophyseal joints have been suggested tocause pain; possible sources include pinchingof synovial folds between the joint surfaces,'-3

excessive stretching of the joint capsule,4impingement of the tips of the inferiorapophyses of a lumbar vertebra againstthe laminae of the neighbouring inferiorvertebra,4 5and compression of the spinalnerve roots as a result of stenosis of the inter-vertebral foramina caused by apophyseal osteo-phytosis.4 Clinical evidence exists to supportthe hypothesis of an apophyseal origin forpain,6-9 and nerves likely to be nociceptive infunction have been identified in capsulartissue. 10-3

Putti"4 noted extreme asymmetry, osteo-arthrosis, and ankylosis of apophyseal joints.Apophyseal joint orientation has beensuggested to influence intervertebral discpathology.'5 16 Ghormley5 believed that apo-physeal osteoarthrosis could cause pain. Puttiand Logroscino'7 ascribed apophyseal osteo-arthrosis to irregularities of joint shape andmotion. Marginal apophyseal fractures havebeen noted.5 18-20 Ghormley5 and Mitchell20noted the difficulty of diagnosing suchfractures.

Butler et al,2' who examined intervertebraldiscs by magnetic resonance imaging (MRI)and the associated apophyseal joints withcomputed tomography (CT), concluded thatdegeneration of the intervertebral discinvariably precedes that of the apophysealjoints. This conclusion may be wrong if MRIwas better at detecting disc degeneration thanCT was at detecting apophyseal osteoarthrosis-apophyseal joint damage has been foundindependently of degeneration of the inter-vertebral disc and body, in both mediaevalskeletons22 and modem specimens.23 24A study of the extent and location of

apophyseal cartilage damage should revealmuch about its pathogenesis and interactionwith intervertebral disc degeneration. Thispaper presents such an investigation-onecomponent of a study which also examinedthe compressive modulus and thickness oflumbar apophyseal cartilage, and the jointmorphology. It aimed to ascertain if the extentof apophyseal cartilage damage is correlatedwith disc grade, age, or both, and to record thedistribution of damage to the apophyseal jointsurfaces.

Material and methodsThirty one upper lumbar motion segments(table 1) were dissected and the lumbarapophyseal surfaces stained with Indian ink,following the method ofMeachim.25 A drawingand photographic record were made of eachjoint surface. The intervertebral discs were

School ofMechanicalEngineering,University oftheWitwatersrand,Wits 2050,Transvaal,South AfricaMW SwanepoelRheumatology andRehabilitationResearch Unit,School ofMedicine,University ofLeeds,Leeds,United KingdomJ E SmeathersThe Centre for HumanBiology,Worseley BuildingL M AdamsCorrespondence to:Dr Swanepoel.Accepted for publication6 October 1994

182

on May 11, 2022 by guest. P

rotected by copyright.http://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.54.3.182 on 1 March 1995. D

ownloaded from

http://ard.bmj.com/

Human lumbar apophysealjoint damage and intervertebral disc degeneration

sectioned in the transverse plane, photo-graphed, and their degenerative state gradedusing the method ofNachemson.26The photographic slides taken of the joint

surfaces during dissection were projected ontoa screen and tracings made of the outlines ofthe entire joint surfaces. Those areas stainedwith Indian ink, indicating fibrillation, were

marked and cross-hatched. Damage was

considered to be any disruption of the cartilageworse than slight flecking. This was subjective,but damaged areas were selected by thesame individual for all the joint surfaces.Transparent acetate sheets were placed over

the tracings and two black silhouettes made ofeach joint surface; the first being of the entirejoint surface and the second of the damagedareas alone. The silhouettes were marked toindicate anatomical orientation.The silhouettes were placed sequentially

under a low magnification dissectingmicroscope (Type 162065, Wild, Heerbrugg,Switzerland). The image was manuallyadjusted by varying the intensity of thetransmitted light. A closed circuit televisioncamera (HV-720K, Hitachi Denshi, Japan)transmitted the images to an image processingunit (Image Manager, Sight Systems Ltd,Newbury, Berks, UK) attached to a desktopcomputer (Elonex PC386 SXM/16, Elonexplc, 2 Apsley Way, London NW2 7LF). Theimage was reproduced on a CRT display(Philips CM 8833/05G) in 256 shades fromblack to white, with a resolution of 512 X 512pixels. The focused image was stored for digitalmanipulation and analysis. The narrow rangeof the black silhouettes allowed the contrast tobe altered digitally to remove artefacts.The silhouettes of each joint surface and its

damaged areas were grabbed sequentially, so

that the extent and distribution of damagecould be measured. The area, geometriccentre, mean distance at which surface area

pixels were located from this centre, maximumanteroposterior length and maximum supero-inferior length of the entire joint surface were

calculated. The image of the damaged areas

was then superimposed on the image of theentire surface, with the same magnification andalignment. The program then: (1) Calculatedthe ratio of the damaged area to the entire jointarea. (2) Found the geometric centre of thecombined damaged areas. (3) Ascertainedthe position of the geometric centre of thedamaged areas with respect to the centre of theentire joint area in the anteroposterior andsuperoinferior directions. (4) Divided thedistance of the centre of the damaged areas

from the centre of the joint surface by the meandistance of the whole area from the centre ofthe joint surface. If the ratio was greater than1 0, most damage was peripheral, and if lessthan one, it was central.These measurements were compared not

only between themselves, but also with the ageof the specimens and the grade of degenerativechanges of the intervertebral discs. SpecimensQ and R were excluded from the analysis of theextent and location of apophyseal fibrillationbecause of complete intervertebral ankylosis.

For purposes of the correlation between discgrade and apophyseal fibrillation, the calcifiedintervertebral discs of these specimens were

recorded as grade 4+, and the apophysealankylosis as 100% damage.

If the extent of damage generally increasedwith the measure of its peripheral distributionand specimen age, then it could be concludedthat damage starts centrally, spreading out-wards. Similarly, it could be determined ifdamage occurred predominantly in a particularpart of the joint surfaces.The displacement of the geometric centre of

the damaged area from the centre of the entirejoint surface in the anteroposterior (AP) andsuperoinferior (SI) directions was expressed as

ratios of the maximum AP width and SI lengthof the joint surface, respectively. The positionof the geometric centre of the damaged area

could be expressed anatomically, for exampleas being located anteriorly and posteriorly withrespect to the centre of the joint. Theautomated calculations eliminated subjectivity,and the use of ratios allowed comparisonsbetween joint surfaces regardless of differencesin joint size and in the magnification incurredin photography and image grabbing.

ResultsTable 1 lists details of the disc specimens.Specimens Ql and Rl exhibited completeintervertebral ankylosis, with advancedalthough incomplete ankylosis of theapophyseal joints. They were included incomparisons of intervertebral disc andapophyseal joint degradation with each otherand with age, but were excluded fromsubsequent analyses.The percentage of area damaged was

calculated as a mean value for the four

Table 1 List of31 specimens studied

Specimen Cause ofdeath

Code Type Sex Age

B1 Li + L2 M 78 Lung carcinomaCI LI +L2 M 41 Cardiac arrestDI Li +L2 M 55 Airway obstructionEl LI + L2 F 46 Cardioresp. failureFl L2 + L3 M 39 Cardiac arrestGi LI +L2 F 35 UnknownHI L2 + L3 M 21 RTAII L3 + L4? M 20 Head injuryJi L3 + L4 M 23 UnknownKi Ll + L2 M 19 UnknownLI Li +L2 M 54 Cardiac arrestMi Ll +L2 M 27 NeoplasmNi Li + L2 M 54 Cardiac arrest01 LI + L2 M 59 Cardiac arrestPi LI + L2 ? 60 Cardiac arrestQi LI + L2 + L3 M 70 Bladder carcinomaRI T12 + Li + L2 F 60? Cardiac arrestSi LI + L2 M 31 Head injuryTI LI + L2 M 34 Cardiac arrestU1 LI + L2 M 35 Cardiac arrestVI Li + L2 M 35 Heart transplantWI LI + L2 M 35 Cardiac arrestXi LI +L2 M 30 Mitral valveYi Li +L2 M 16 Cardiac arrestA2 Li +L2 F 20 Cardiac arrestB2 L2 + L3 M 22 RTAC2 LI +L2 F 36 Cardioresp. failureD2 LI +L2 F 33 HeadinjuryE2 LI +L2 M 25 Multiple injuriesF2 Li + L2 M 17 Head injury, RTAG2 Li + L2 M 35 Cardioresp. failure

Mean age 38 (SD 17) years. Heart transplant = failed hearttransplant; RTA = road traffic accident.

183



j.com/

Ann R

heum D


ownloaded from

http://ard.bmj.com/

Swanepoel, Adams, Smeathers

Table 2 Percentage ofjoint areas damaged

Specimen Superior Superior Inferior Inferior Mean aUlcode left right left right areas

B 1 38 50 40 23 38C1 47 54 39 71 53Dl 6 8 20 39 18El 66 51 71 75 66Fl 32 57 13 19 30GI 52 57 53 51 53Hi 45 76 32 41 49I1 72 16 19 13 30Ji 68 38 40 27 43KI 35 20 23 29 27Ll 28 37 11 63 35Ml 28 29 29 43 32Ni 70 30 100 31 5801 41 72 NA NA 57P1 33 72 100 78 71S1 75 12 25 72 46Ti 54 51 35 52 48Ul 52 41 36 40 42Vi 76 41 51 36 51WI 79 29 86 72 67Xi 46 58 38 50 48Y1 68 73 40 45 57A2 46 46 45 27 41B2 36 47 12 31 32C2 36 45 36 50 42D2 46 62 83 50 60E2 80 52 41 19 48F2 32 29 45 36 36G2 8 10 77 51 37Mean (SD) 48 (20) 44 (19) 44 (25) 44 (18) 45 (13)

NA = Data not available.

apophyseal joint surfaces of each lumbarmotion segment. Table 2 lists the percentageof joint areas found to be at least slightlyfibrillated: each type of surface, whethersuperior, inferior, left or right, had a meanextent of damage of 44-48%. The extent ofdamage of contralateral inferior, convex jointsurfaces was significantly correlated (meansurface damage on the left 44-2 (SD 24-7)%;on the right 43-9 (17-9)%; r2 = 016, p < 005),but this was not true of any other surfaces.There was a weak correlation between theextent of apophyseal joint damage and donorage (fig 1), and a strong correlation betweenthe degenerative grade of the intervertebral

c)

2- [I]ol

y=0.047x +0.425C.1

r= 0.630

0 10 20 30 40 50 60 70 80Age (yr)

Figure 2 Variation of the state of the intervertebral discswith age. Discs graded on a scale of 1 = intact, to 4 = mostdegenerate.

discs and age (fig 2), so that it is unsurprisingthat the extent of lumbar apophyseal jointdamage was correlated with the grade of theintervertebral discs (fig 3).

Table 3 summarises the locations of centresof joint damage. The centres of the damagedareas were found to be posterior to the centresof the concave superior joint surfaces, althoughthis was no more than a tendency in the case

100 -

100 - 0 0

90] % damage = 0 5 age + 30r2= 0.20

80 -

70 -

0cc 60 -

a) 50-C)Ecu 40-a0c

4)30-

20 -

10 -

U

U

UU

90 % damage = 26 + 10 x disc grader2 = 0.29

80 -

00-

a)0)E

~0a)

O C0 ca)

U1 00

50 -

40 -

30 -

20 -U

10 -

10 20 30 40 50 60 70 80

U

U

0

U

U

3 U

0

U

3 42Age (yr) Disc grade

Figure 1 Variation of the extent of apophysealjoint Figure 3 Variation of the extent of apophysealjointdamage with age (mean damaged area of the fourjoint fibrillation with disc grade (1 = intact; 4 = mostsurfaces ofeach specimen). degenerate).

184

0



j.com/

Ann R

heum D


ownloaded from

http://ard.bmj.com/


Table 3 Anteroposterior, superoinferior and central-peripheral locations of the cjoint damage

Joint suiface Ratio (SD) Location t

Anteroposterior location (Ratio values: 0 = Central; +1 = posterior; -1 = anterior)All (n = 114) 0 034 (0 092) Posterior 3-92Superior (n = 58) 0-046 (0-098) Posterior 3-57Inferior (n = 56) 0-021 (0-086) Posterior 1-83

Superoinferior location (Ratio values: 0 = Central; +1 = inferior; -1 = superior)All (n = 114) -0003 (0-108) Superior 0-30Superior (n = 58) -0-028 (0-102) Superior 2-09Inferior (n = 56) 0-024 (0- 108) Inferior 1-66

Central-peripheral location (Ratio values: 0 = Central; 1-5 = peripheral; 1 0 = no biasAll(n=114) 1 100(0-134) Peripheral 797Superior (n = 58) 1-086 (0-139) Peripheral 4-71Inferior (n = 56) 1-114 (0-130) Peripheral 6-56

of the inferior surfaces. Damage was biasedsuperiorly in the superior surfaces, and theinferior bias of damage of the inferior surfacesverged on significance. Damage occurred morefrequently in the peripheral areas of all surfacesthan one would expect by chance. The centralto peripheral distribution of the centres ofdamaged areas with respect to the centres ofthe joint surface areas is presented in figure 4.There was no correlation between the

anteroposterior location of the centres ofdamage of both contralateral and articulating

100 -

90 -

80 -

0- 70-

0)

CD60-

Q 50-

a -

t 40-

in

c

o 30-

20-

10

O °5 Etfl0

00

0 E r

E5

0s e

°°lo lb PE15COrO00

_ . I

0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5

Coefficients of centralityFigure 4 Variation of the central-peripheral location ofjoint damage with its extent. Coefficients ofcentrality:0 = Central; 1 0 = no bias; 1 5 = peripheral.

Table 4 Significant correlations between the centrality ofdamage of differentjoint surfaces

Joint Displacement ratio t test Correlationrfde Left Right ?r p

All 1-10 (0-12) 1-10(0-15) NS 026 <0001Superior 1-09 (0 11) 1-09 (0-16) NS 0-24 <0 01Inferior 1-12 (0-12) 1-11 (0-14) NS 0 31 <001

Mean values (SD). Ratio values: 0 = Entirely central; 1 = nobias; 1 5 = entirely peripheral.

surfaces, but there was a correlation betweenthe superoinferior location of the centres ofdamage of the contralateral inferior jointsurfaces (mean displacement ratio on the left-0X031 (SD 0X107); on the right -0-014(0-103) (negative values indicating inferiorlocation); r2==035, p<O0001). The centralto peripheral damage distributions of thecontralateral (concave) superior joint surfaceswere correlated, as were those of thecontralateral convex inferior joint surface pairs,but this was not true of the matchedarticulating surfaces (table 4). The extent ofapophyseal cartilage damage was not cor-related with the estimated cadaveric height ormass, nor with the location of the centre ofjoint surface damage.

DiscussionCOMPUTER BASED IMAGE ANALYSIS

Computer based image analysis of Indian inkstained cartilage has not been previouslyattempted, thus some explanation of themethod is due. Originally, slides wereexamined directly with the image processingunit through the 256 shade closed circuittelevision camera (CCTV). The contrast of theimage was adjusted digitally to highlightfibrillation before the image underwent'thresholding'; fibrillated areas stained darkerthan a predetermined shade were selected to beblack, while the surrounding areas were set towhite. Thus a binary image of the fibrillatedcartilage was produced. Selection of damagedareas was unpredictable, because fibrillationwas recorded uniquely within the grey scale ofeach slide, thus nullifying selection algorithms.This was due to:(1) The slides being taken using a macrolenswith manual aperture setting during differentweather conditions, seasons, and times of day.Hence the illumination of the slides differedalthough they were all taken indoors with acompensatory flash.(2) The perimeter of stained areas decreasingwith an increase in the total light transmittedthrough non-stained areas. This effect wasexacerbated by the size range of joints. (Thebrightness of the image had a 'washout' effecton stained areas. This problem was inherent inthe use of a CCTV.)

After experimentation it was found that theanalogue capability of the human eye enjoyedan advantage over the limited digital resolutionof the computer. Hence the slides wereprojected, tracings made of the fibrillated areas(on which grades of damage were noted forfuture reference), and black silhouettes madefrom the tracings. These silhouettes madeimage capture in a 256 grey shade scale almostirrelevant. However, contrast stretching andthresholding, which can only be performed ongrey shade scales, were retained to removeartefacts caused by optical imperfections.

It was concluded that digital image analysisis excellent for calculating lengths, areas, andorientations, and is useful for comparingphotographic images taken under tightlycontrolled lighting, but is a poor second best

185 on M

ay 11, 2022 by guest. Protected by copyright.

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.54.3.182 on 1 M

arch 1995. Dow

nloaded from

http://ard.bmj.com/


to human judgment of shaded areas undernon-uniform conditions.

JOINT DAMAGE AND DISC DEGENERATION

Is lumbar apophyseal joint damage alwaysconsequent upon intervertebral disc de-generation, as suggested by Butler et al?21The apophyseal cartilage surfaces were

fibrillated over at least 20% of their stirfaces at20 years old, as determined by Indian inkstaining (fig 1), while the intervertebral discs atthe same age exhibited minimal degeneration(fig 2), so it is most unlikely that the former isinvariably associated with the latter; the weakcorrelation between apophyseal joint damageand intervertebral disc degeneration (fig 3), isinconclusive. Unfortunately, the grading ofintervertebral disc degeneration followingthe method of Nachemson26 proved to beinadequate for assessing the interplay betweenintervertebral disc and apophyseal jointdegeneration. A biochemical analysis of thedisc material, discograms, and radiographs ofthe vertebral endplates would have increasedthe objectivity and usefulness of the study.

Reduction of the intervertebral disc space

has been implicated in the pathogenesis ofapophyseal osteoarthrosis.2' 24 27 It is uncertainthat fissuring and fibrosis of discs are correlatedwith their thickness and mechanical properties.Smeathers and Joanes28 measured the inter-vertebral disc space of seven upper lumbarmotion segments from lateral radiographs andgraded degeneration on a scale of 1-4. Thenegative correlation between the disc gradeand intervertebral space of -0-664 indicatesthat disc spaces narrow with degeneration, butthe probability of this being a chance trend is0 1. Proof that disc space narrowing causes

apophyseal osteoarthrosis would demand invivo radiographic measurements of the same

individuals over many years.Apophyseal joint cartilage was fibrillated

even when the discs were intact (fig 3). Thisdoes not contradict studies linking apophysealosteoarthrosis to intervertebral disc de-generation, such as that of Butler et al,2' whodefined osteoarthrosis to include visible loss ofthe joint space, and that of Lewin,23 whodefined osteoarthrosis to include lesions of thesubchondral bone. Several authorities22-24 are

convinced that intervertebral disc degenerationcannot explain all apophyseal joint de-generation. Moreover, radiography of theapophyseal joints does not disclose cartilagedamage if the joint space is preserved. 7

Malmivaara et a129 concluded that apophysealosteoarthrosis is unrelated to disc degenerationor vertebral body osteophytosis at the T10-LIlevels. Figures 1, 2, and 3 may indicate boththat apophyseal joint cartilage softens anddeteriorates with age, and that joint damageincreases as a result of narrowing of theintervertebral discs. There is, however, still a

wide variation in joint damage associated withgrade 3 discs.

Baker et a130 noted that spontaneousarthrodesis/ankylosis of the apophyseal jointsfrequently occurs if the intervertebral bodies of

the corresponding motion segments are fused,and that the first changes are fibrous ingrowthfrom the joint edges over the cartilage, loss ofproteogylcan synthesis, and the developmentof a fibrous layer at the tidemark. Theapophyseal joints of two specimens withcalcified intervertebral discs (Ql and R1)allowed us to confirm that arthrodesis starts atthe joint margins.Dynamic periodic loading stimulates the

proteoglycan synthesis underpinning thehealth of articular cartilage, while prolonged orlight loading has the opposite effect.3' 32 It hasalso been shown that lumbar apophyseal jointsatrophy when the spine is distracted usingHarrington instrumentation;33 thus it is likelythat a lack of spinal exercise because of asedentary lifestyle would cause 'abnormal'softening of the lumbar apophyseal jointcartilage. The cephalo-caudal alignment of theupper lumbar apophyseal joints means thatthey experience significant loading only duringspinal flexion and torsion; moreover the spinalflexion experienced by sedentary workers is ofa postural rather than dynamic nature. Thehypothesis that lumbar apophyseal cartilage isunusually soft is borne out by its compressivestiffness, which can be as stiff as well exercisedareas of human knee cartilage, but is onaverage only a third of this stiffness.34 Thiswould predispose it to mechanical disruptionby infrequent high loads in the mannersuggested by Seedhom et al.35The finding that the cross-sectional area of

the narrowest part of the vertebral bodies iscorrelated with the estimated cadaveric height,but not with the estimated cadaveric mass,coupled with the predominant cause of death,cardiac arrest, suggests that the donors westudied had not been exercising adequately. Ifthe specimens had come from fit people, it isprobable that a correlation between bodyweight and the lumbar vertebral cross-sectionalarea would exist, as height and mass wouldthen be correlated.The locations of damage as indicated by

image processing must be carefully interpreted.Local sites of damage may not have beenapparent because their influence on the overallcentres and distributions of damage may havebeen balanced by other local areas of damageelsewhere on the joint surfaces. Thus it is likelythat damage to the inferior apices of theinferior apophyses would have been significantin this study but for the inclusion of randomsites of localised damage within the pro-gramming algorithm.The tips of the inferior apophyses exhibited

irregular cartilage margins and occasional grossfibrillation. Narrowing of the intervertebraldisc space causes the tip of the inferiorapophyses to press against the laminae of theinferior vertebra of each lumbar motionsegment, especially in lumbar extension.27 Itmay be expected that the large tip pressuresthen cause irreversible damage, the productionof debris and subsequent erosion of the jointsurfaces by wear, together with osteophyticgrowth. There were specimens suggestive ofthis mechanism.

186 on M


http://ard.bmj.com

/A

nn Rheum


arch 1995. Dow

nloaded from

http://ard.bmj.com/


The posterior and peripheral staining of thesuperior concave surfaces was associated bothwith overt fibrillation and with the fibro-cartilaginous transition zone between thecapsule and the articular cartilage (a zone alsonoted by Bogduk and Twomey36). Thecollagen fibres align with the tension exertedthrough the capsule, leading to fine splitsbetween and parallel to the fibres. This wouldcause an initial loss of the load bearingproperties of the surface, as the retention ofwater responsible for the resilience of thecartilage cannot be maintained. Localdisruption of the articular surface would ensue,leading to the posterior fibrillation typical ofsuperior joint surfaces. Although this was a'normal' state, it cannot be a healthy one. Itmay be caused by, and would certainly beaggravated by, prolonged lumbar flexion,which subjects the apophyseal joint capsule totraction. Prolonged flexion would typically beassociated with an incorrect seating posture,with loss of the lumbar lordosis.

However, this is not the only possible reasonfor posterior fibrillation of joint surfaces. Theapophyseal joints taper slightly inwardsinferiorly.37 When the narrower inferior partsof the convex inferior surfaces are drawnsuperiorly relative to the superior surfacesduring lumbar flexion, the inferior surfaces arefree to move so that their more coronal anteriorparts are loaded. If, however, the surfaces donot contact each other anteriorly because ofjoint incongruency, then the anterior shearloading incurred between the lumbar vertebraeduring flexion must be borne by the perisagittalposterior sections of the apophyseal joints. Theperisagittal orientation will cause highcompressive stresses between the articulatingposterior joint surfaces in order to carry theanteriorly directed force-this may bedescribed as 'wedging'. Maintenance ofwedging pressures resulting from prolongedlumbar flexion would lead to contact betweenthe collagen fibres of the articulating surfaces,and hence posterior fibrillation of theapophyseal joints.The apophyseal surfaces of specimen LI

curved medially and anteriorly so much thattheir most medial borders pointed posteriorlyand medially away from the coronal plane.This could be ascribed not only to thesubchondral bone shape, but also to thecartilage which was wrapped around the edgesof the inferior surfaces to form 'bumpers'('fenders'), as described by Bogduk andTwomey.36 Bogduk and Twomey thought thatstressing of the edges of articular processesduring torsion induces bumpers to form, whileHadley2 felt that recurrent impaction of theapices of the inferior apophyses could lead tobumpers forming inferiorly.The cumulative incidence of apophyseal

joint injury should increase with age, as theduration of the exposure to risk increases. Twoinjuries were noted. A crack emanatedupwards from the posterior and inferior marginof the superior surface of specimen HI, and aninferior apophysis of specimen I1 exhibited aposterior marginal fracture ascribable to

torsion. Both were old injuries. Surprisingly,the surfaces were less fibrillated than average.These cases, together with the high incidenceof cartilage fibrillation in the absence ofapophyseal fractures, prove that trauma playsa minor role.Gouge marks on the surface of one superior

(concave) joint surface suggested that theboney tip of the inferior apophysis of the jointhad been lifted from its inferior position, andpressed into the cartilage surface. This wouldhave occurred as a result of combined lumbarflexion and torsion.

In view ofthe extent and incidence oflumbarapophyseal joint fibrillation, and the softness ofthe cartilage, it is suggested that the apophysesare deliberately loaded by twisting the trunkaround the cephalo-caudal axis in the standingposture several times each day. This activityshould ensure adequate mechanical stimulusfor healthy conditioning of the cartilage.Further, it is suggested that prolonged lumbarflexion is avoided to minimalise large staticstresses caused by wedging of more sagittallyorientated apophyseal joints, and to reducetension on the insertions of the joint capsulesinto the concave superior joint surfacesposteriorly.

ConclusionsAll the apophyseal joints in this study rangingin age from 16 to 78 years showed signs ofcartilage fibrillation. The prevalence offibrillation in the lumbar apophyseal joints isgreater than that of the knee, hip or ankle. Onaverage 45% of the total joint surface wasdamaged.Damage of the superior, concave apophyseal

joints was predominantly peripheral, superior,and posterior, that of the inferior apophysealjoints was predominantly peripheral andposterior.The pattern of damage suggests that it is

initiated in areas of cartilage which are soft,either as a result of a lack of adequate loading,or because of their location close to theinsertion of collagenous joint capsule fibres.There is a weak correlation between the

amount of apophyseal joint damage andthe intervertebral disc degenerative grade(r2 = 029), both of which increase with age.

1 Keller G. Die Arthrose der Wirbelgelenken in ihre Bezie-hung zum Ruckenschmerz. Z Orthop Ihre Grenzgeb 1959;91: 538-50.

2 Hadley L A. Anatomico-roentgenographic studies of thepostero-spmal articulations. AmJ Roent 196 1; 86: 270-6.

3 Kos J, Wolf J. Les menisques intervertebraux et leur rolepossible dans les blocages vertebraux. Ann Med Phys1972; 15: 203-17.

4 Hadley L A. Subluxation of the apophyseal articulationswith boney impingement as a cause of back pain. Am JRoent 1935; 33: 209-13.

5 Ghormley R K. Low back pain with special reference to thearticular facets with presentation of an operativeprocedure. JAMA 1933; 101: 1773.

6 Rees W E S. Multiple bilateral subcutaneous rhizolysis ofsegmental nerves in the treatment of intervertebral discsyndrome. Ann Gen Pract 1971; 26: 126-7.

7 Toakley J G. Subcutaneous lumbar "rhizolysis"-anassessment of 200 cases. Med3JAust 1973; 2: 490-2.

8 Oudenhoven R C. Articular rhizotomy. Surg Neurol 1974;2: 275-8.

9 Shealy C N. The role of the spinal facets in back and sciaticpain. Headache 1974; 14: 101-4.

187



j.com/

Ann R

heum D


ownloaded from

http://ard.bmj.com/


10 Pedersen H E, Blunck C F J, Gardner E. The anatomy ofthe lumbosacral posterior rami and meningeal branchesof spinal nerves. J Bone Joint SurgAm 1956; 38: 377-91.

11 Bradley K C. The anatomy of backache. Aust NZ _T Surg1974; 44: 227-32.

12 Bogduk N, Long D M. The anatomy of the so-called"articular nerves" and their relationship to facet dener-vation in the treatment of low-back pain. Y Neurosurg1979; 51: 172-7.

13 Nade S, Bell E, Wyke B D. The innervation of the lumbarspinal joint and its significance. 7 Bone Joint Surg Br 1980;62:255.

14 Putti V. New conceptions in the pathogenesis of sciatic pain.Lancet 1927; 2: 53-60.

15 Goldthwait J E. The lumbosacral articulation: Anexplanation of many cases of lumbago, sciatica andparaplegia. Boston Med SurgJr 1911; 64: 365-72.

16 Farfan H F, Huberdeau R M, Dubow H I. Lumbar inter-vertebral disc degeneration. The influence of geometricalfeatures on the pattern of disc degeneration-a postmortem study. J Bone J7oint SurgAm 1972; 54: 492-510.

17 Putti V, Logroscino D. Anatomia dell'Artritismo VertebraleApofisario. Chir Organi Mov 1937/8; 23: 317-53.

18 Burk W. Ueber einen Bruch des Gelenkfortsatzes des 5.Lendenwirbels. BeitrKlin Chir 1908; 48: 558.

19 Koch K. Die isolierten Gelenkfortsatzbruche derLendenwirbelsaule. Deutsche Z Chir 1923; 180: 339.

20 Mitchell C L. Isolated fractures of the articular processes oflumbar vertebrae. _JBone Joint Surg 1933; 15: 608.

21 Butler D, Trafimow J H, Andersson G B J, McNeill T W,Huckman M S. Discs degenerate before facets. Spine1990; 15: 111-3.

22 Ingelmark B E. Function of, and pathological changes in thespinal joints, IV. Acta Anat Suppl 1959; 36: 12-57.

23 Lewin T. Osteoarthritis in lumbar synovial joints. A morpho-logic study. Acta Orthop Scand Suppl 1964; 73: 1-112.

24 Ziv I, Maroudas C, Robin G C, Maroudas A. Therelationship between the degeneration of the humanintervertebral disc and facet joints of the same lumbarsegment-a physico-chemical study. 36th AnnualMeeting of the Orthopedic Research Society, NewOrleans, Louisiana February 5, 1990.

25 Meachim G. Light microscopy of Indian ink preparationsof fibrillated cartilage. Ann Rheum Dis 1972; 31: 457-64.

26 Nachemson A L. Lumbar intervertebral pressure: Experi-mental studies on post-mortem material. Acta OrthopScand Suppl 1960; 43: 42-3.

27 Dunlop R B, Adams A, HuttonW C. Disc space narrowingand the lumbar facet joints. J Bone J7oint Surg Br 1984;66: 706-10.

28 Smeathers J E, Joanes D N. Dynamic compressiveproperties of human lumbar intervertebral joints: A com-parison between fresh and thawed specimens. J Biomech1988;21:425-33.

29 Malmivaara A, Videman T, Kuosma E, Troup J D G. Facetjoint orientation, facet and costovertebral joint osteo-arthrosis, disc degeneration, vertebral body osteophytosisand Schmorl's nodes in the thoracolumbar region ofcadaveric spines. Spine 1987; 12: 458-63.

30 Baker W C, Thomas G T, Kirkady-Willis W H. Changesin the cartilage of the posterior intervertebral joints afteranterior fusion. JBoneJoint SurgBr 1969; 51: 736-46.

31 Parkkinen J J, Lammi M J, Ikonen J, Helminen H J,Tammi M. The influence of cyclic hydrostatic pressure oncultured articular cartilage and chondrocytes. AbstractC 18, Joint Destruction in Arthritis and Osteoarthrosis19th Symposium of the ESOA, Noordwijkerhout,Netherlands, 24-27 May 1992.

32 Urban J, Hall A. Physical modifiers of cartilage metabolism.In: Kuettner K, et al, eds. (Workshop Conference HoechstWerk Kalle-Albert, Wiesbaden 1991.) Articular cartilageand osteoarthritis, ch. 27. New York: Raven Press, 1992;393-406.

33 Kahanovitz N, Bullough P, Jacobs R. The effects of internalfixation without arthrodesis on human facet cartilage.Orthop Trans 1983; 7: 14.

34 Swanepoel M W, Smeathers J E, Adams L M. The stiffnessof human apophyseal cartilage as an indicator of jointloading. Proc Inst Mech Eng [H] 1994; 208: 33-43.

35 Seedhom B B, Takeda T, Tsubuku M, Wright V.Mechanical factors and patello-femoral osteoarthritis.Ann Rheum Dis 1977; 38: 307-16.

36 Bogduk N, Twomey L T. Movements of the lumbar spine.In: Clinical anatomy of the lumbar spine, 2nd edn, ch. 7.Edinburgh: Churchill Livingstone, 1991; 65-81.

37 Pearcy M J, Hindle R J. Axial rotation of lumbar inter-vertebral joints in forward flexion. Proc Inst Mech Eng [H]1991; 205: 205-9.

188 on M


http://ard.bmj.com

/A

nn Rheum


arch 1995. Dow

nloaded from

http://ard.bmj.com/

human joint and intervertebral - annals of the rheumatic

Documents