Dystrophic calci fication and heterotopic ossi fication in fibrocartilaginous ti

Dale E.Fournier ,Patti K.Kiser,Ryan J.Beach,S.Jeffrey Dixon 2, and Cheryle A.Séguin2,

INTRODUCTION

Diffuse idiopathic skeletal hyperostosis(DISH)is a nonin flammatory spondyloarthropathy characterized by bony outgrowths or hyperostoses along the anterolateral aspect of the vertebral column,particularly in the thoracic region.1DISHis diagnosed by the radiographic detection of flowing mineral formation along four contiguous vertebral bodies,the preservation of intervertebral disc(IVD)height in the involved areas,and the absence of bony ankylosis of the vertebral facet and/or sacroiliac joints.2Although not included in the diagnostic criteria,DISH is often associated with the presence of extraspinal hyperostoses,commonly in the knee,ankle,hip,shoulder,and elbow joints.3Symptoms associated with DISH are variable,ranging from spine stiffness and decreased spinal ranges of motion(with or without back pain)to,in severe cases,dysphagia,spinal cord/nerve root compression,and vertebral fracture.4–6Notably,the clinical symptoms are poorly understood,and the radiographic diagnosis of DISH is limited to an advanced disease state.7

The prevalence of DISH in North America and Europe is estimated to be 15%–25%and 17%of the population over the age of 50,respectively.8–9Risk factors for DISH include ethnicity(e.g.,Caucasians),8sex(males＞ females),8–9advanced age,8,10and metabolic disorders(e.g.,obesity,diabetes mellitus).11Together,the lack of early detection of DISH,the rise in potential risk factors,and the unfamiliarity with DISH among medical professionals suggest that its prevalence is greater than previously reported.

The cause and biological pathways controlling the formation of ectopic mineral in DISH are unknown.Systemic metabolic changes related to obesity,diabetes mellitus,large waist circumference,hypertension,hyperinsulinemia,dyslipidemia,and hyperuricemia are associated with DISH.4,11–14Familial cases of DISH,15–16although rare,paired with the characterization of ectopic spine mineralization in animal models,17–20suggest the contribution of genetic factors in the etiology of DISH.A related disorder,ossi fication of the posterior longitudinal ligament,hasbeen reported frequently with DISHand is postulated to share a similar pathogenesis.21DISH is characterized by the right-sidedness of the mineral formation in the thoracic spine,which is thought to be a consequence of the mechanical pressure created by persistent aortic pulsations on the left serving to inhibit soft tissue mineralization.22Ultimately,our limited understanding of the pathobiology of DISH has resulted in the lack of early diagnostic indicators,prognostic factors,and disease-modifying treatments.23

The hallmark DISH characteristic of flowing ectopic spinal mineralization is based on radiographic description.2The histopathological characteristics of DISH are less well known.Initial reports described histopathological features of endochondral ossi fication of the anterior longitudinal ligament24as well as changes in the gross features of the IVD.21,25In contrast,recent studies in cadaveric tissues demonstrated that the anterior longitudinal ligament was morphologically normal in appearance and was displaced by ectopic mineral,but did not itself undergo aberrant mineralization.26Taken together,these findings underscore the need to identify and characterize the composition of the ectopic mineral formed in DISH,as well as the tissue types affected and associated cellular changes.The current study investigated these questions by combining radiological,histological,and physical analyses of human spine segments affected by DISH.

RESULTS

A previous study by our group characterized the morphometry and radiodensity of ectopic mineral associated with DISH in embalmed cadaveric human spines using microcomputed tomography(μCT)imaging.27Mineral formations at individual motion segments in the thoracic spine were differentiated based on morphology,categorized as vertical bands when the resultant bridging angle relative to the vertebrae was greater than 90°,as horizontal outgrowths when the resultant bridging angle was ＜90°,or as discontinuous-patchy when the mineral formed an incomplete bridge.27For the present study,we carried out detailed analyses of 15 individual motion segments from the thoracic spine of six donors with DISH from the previous study(one female and five males;median age 82 years,range 72–87)(Fig.1).27These spines were selected based on(i)having met the diagnostic criteria for DISH(with at least four contiguous segments affected,2previously assessed by two clinician observers)and(ii)each spine containing at least two of the three morphological presentations of ectopic mineralization described in our previous study(discontinuous-patchy,vertical,or horizontal).27

Histological features of spinal tissues

Twelve regions of the spine containing ectopic mineral structures at individual motion segments were isolated(Supplementary Fig.S1)to assess the histological appearance of affected tissues.Figure 2 is representative of a motion segment with the morphological features of“horizontal”and “discontinuous-patchy”mineralization;Fig.3 is representative of a motion segment with the morphological features of “horizontal”mineralization;and Fig.4 is representative of a motion segment with the morphological features of “vertical” mineralization.As a reference,Supplementary Fig.S2 is representative of a motion segment with no evidence of ectopic mineralizations from a spine that did not meet the diagnostic criteria for DISH.In Figs.2–4,digital radiographs are paired with low-magni fication views of the intact motion segment to correlate histological features with areas of tissue mineralization within decalci fied tissue sections.Speci fic areas of interest are presented at higher magni fication,with serial sections stained with hematoxylin and eosin to localize cells and evaluate extracellular matrix structure,Masson’s trichrome to detect the collagenous component of the extracellular matrix,and later visualized by polarized light microscopy to assess birefringence as an indicator of optical isotropy.

IVDs within the intact motion segments examined demonstrated consistent histological features (n=12:Figs.2–4).Although radiographic IVD height was maintained in keeping with Resnick’s criteria,2the nucleus pulposus was granular in appearance,indicative of mild-to-moderate degeneration.In all sections examined,varying degrees of IVD degeneration were observed.The characteristics of degeneration included the loss of a distinct transition between the nucleus pulposus and surrounding annulus fibrosus(Figs.2b and 4b)as well as moderate-tosevere disorganization of the lamellar structure of the annulus fibrosus,most often detected along the anterior aspect of the IVD(Fig.3b).The cartilage endplates showed irregularities,ranging from the loss of organization of the cartilaginous matrix to small clefts and the disruption of the subchondral bone(i.e.,Schmorl’s nodes,detected in 3/12 motion segments from 3/6 specimens;Figs.2b and 3b).The degenerative changes detected in the IVD were expected given the advanced age of donors(range 72–87 years).The anterior longitudinal ligament(when in-section)was either well preserved or contained discrete regions of mineralization(Fig.4).

Histological features of ectopic mineral structures

In all motion segments assessed,histological examination con firmed the radiographic selection criterion of a mineralized tissue bridge between the superior and inferior vertebrae across an IVD(Figs.2–4).In general,cellular indications of in flammation were absent from areas of ectopic mineralization,although small areas of scattered in flammatory cells(primarily macrophages,lymphocytes,and plasma cells)and neovascularization were noted.Tremendous heterogeneity was,however,noted in the histological features of the ectopic mineral structures associated with DISH.

First,the ectopic mineral bridges between motion segments associated with the radiographic “fl owing candle wax”appearance of DISH varied in length,thickness,volume,and the type of bone(woven vs.mature).In motion segments characterized byμCTas either vertical or horizontal morphological presentations,the ectopic bridges showed features of well-developed lamellar bone,consistent with heterotopic ossi fication(Figs.2c–e and 4f–h).The presence of organized osteons was con firmed by polarized light microscopy(Figs.2e and 4h).In some motion segments,these areas of heterotopic bone also contained bone marrow spaces(Fig.2c–e).Often noted were focal areas of fibrosis and primary woven bone adjacent to or within areas of mature heterotopic bone(Fig.2f–h).

Second,although motion segments with ectopic mineral presenting as horizontal outgrowths showed similar radiographic appearances(Figs.1,2a,and 3a),the large ectopic bone masswas not always continuous.Instead,areas of fibrocartilage were often detected in this region,extending from the IVD and positioned between the superior and inferior mineralized outgrowths(Fig.3b–k).Within these fibrocartilage regions were multifocal areas of granular degeneration,fi brosis,and ossi fication.These sites showed a transition from areas of fibrocartilage to nests of chondrocytes within cartilage to sites of woven bone,consistent with the process of endochondral ossi fication.

Last,isolated regions of amorphous calci fied material were consistently identi fied in motion segments corresponding to all the morphological presentations of ectopic mineralization associated with DISH(i.e.,discontinuous-patchy,vertical,and horizontal).These regions showed features consistent with dystrophic calci fication,which occurs in damaged soft tissues and is characterized by amorphous deposits of calcium phosphate.28–29Areas of dystrophic calci fication were detected by histology and were localized within(i)the fibrocartilaginous tissues located between the IVD and areas of heterotopic ossi fication;(ii)the annulus fibrosus(Fig.4c–e);or(iii)the anterior longitudinal ligament(Fig.4i–k).These areas of dystrophic calci fication were often variably stained and granular in appearance,which may indicate a mixture of calci fied material and degenerated fibrocartilage or differences in the organic constituents of the mineralized matrix.The origin of this material could not be determined.Polarized light microscopy of regions of dystrophic calci fication showed little birefringence,consistent with a lack of structural organization(Fig.4e,k).

Elemental composition and X-ray diffraction patterns of ectopic mineral

Given that areas of ectopic mineralization associated with DISH contained histological features consistent with both heterotopic ossi fication and dystrophic calci fication,we sought to determine and compare the mineral composition of these structures.First,regions of interest within four individual motion segments from two spines meeting the diagnostic criteria for DISH(Fig.1a,f)were analyzed by energy-dispersive X-ray spectroscopy to determine the elemental composition in areas of heterotopic ossi fication(n=5–10 regions/motion segment)and dystrophic calci fication(n=3–8 regions/motion segment).The calcium content was found to be signi ficantly greater in all sites of ectopic mineralization associated with DISH(heterotopic ossi fication and dystrophic calci fication)than in unaffected vertebral bone(indicated as cortical bone)(Table 1).Sites of dystrophic calci fication were associated with a greater phosphorus content compared to unaffected vertebral bone and showed greater calcium and phosphorus content compared with regions of heterotopic ossi fication.The calcium/phosphorus ratio was signi ficantly greater in all regions of ectopic mineralization than in unaffected vertebral bone.On the other hand,the calcium/phosphorus ratios were similar at sites of heterotopic ossi fication and dystrophic calci fication.Next,to assess the crystalline structure within these regions,tissues were analyzed by X-ray diffraction.Regions of interest within sites of dystrophic calci fication and heterotopic ossi fication showed an X-ray diffraction pattern matching that of calcium-de ficient hydroxyapatite[Ca8.8(PO4)6(OH)1.92,powder diffraction file:86-1201,International Centre for Diffraction Data,2018],as did regions within unaffected cortical bone(Fig.5).The theoretical calcium/phosphorus ratio for calcium-de ficient hydroxyapatite is1.47,close to the valuesfound by energy-dispersive X-ray spectroscopy(Table 1).

Structure and radiodensity analysis of ectopic mineral

Ectopic mineral structures associated with DISH in individual motion segments were further characterized by scanning electron microscopy andμCT(Fig.6).An examination of areas of dystrophic calci fication by scanning electron microscopy revealed distinct regions with a uniform,amorphous structure that consistently appeared fractured following tissue processing,distinct from the appearance of adjacent cortical bone(Fig.6b,c).Similar to findings from the histopathological analyses,these regions were localized to the annulus fibrosus of the IVD as well as the fibrocartilaginous structures adjacent to heterotopic ossi fications.Finally,μCT was used to generate pseudocolored radiodensity maps of the same regions assessed for physical characteristics.This analysis demonstrated remarkable variation in the radiodensity of mineralized structures(Fig.6).Within motion segments examined,foci of mineralized tissues corresponding to regions of dystrophic calci fication were detected,with radiodensities that exceeded those corresponding to regions of heterotopic ossi fication and unaffected vertebral bone.The presence of these regions of high radiodensity is consistent with the greater calcium and phosphorus content found in dystrophic calci fications using energy-dispersive X-ray spectroscopy.

Table 1.Calcium and phosphorus content as determined by energy-dispersive X-ray spectroscopy

Having identi fied hyperdense regions(with radiodensities exceeding those of normal cortical bone)as a feature of dystrophic calci fication,we used this as a criterion to evaluate the presence of dystrophic calci fication in spines not diagnosed with DISH and their association with mineralization bridging the IVD.UsingμCT,we first evaluated the presence or absence of hyperdense mineralized foci within the annulus fibrosus or anterior longitudinal ligament in 77 motion segments from spines that did not meet the diagnostic criteria for DISH and 94 motion segments from spines with DISH(Table 2).One spine with advanced bridging associated with ankylosing spondylitis was excluded from the analysis to reduce false positives.We showed that in spines that do not meet the diagnostic criteria for DISH,38 of 77 motion segments(49%)had hyperdense fociwithin the IVD,compared with 72 of 94 motion segments(77%)in spines with DISH.Although dystrophic calci fications are not a unique feature associated with DISH,they are more prevalent in motion segments from spines diagnosed with DISH than in those from spines not diagnosed with DISH(P=0.000 4).We also examined whether the presence of these hyperdense foci was associated with the presence of mineralization bridging the IVD(heterotopic ossi fication).In spines that did not meet the diagnostic criteria for DISH,12 of 38 motion segments with hyperdense foci(32%)also had mineralization bridging the IVD;in contrast,in spines with DISH,58 of 72 motion segments with hyperdense foci(80%)also had mineralization bridging the IVD.Statistical analysis demonstrated that the presence of hyperdense foci was associated with mineralization bridging the IVD in both groups.

This analysis is importantly limited by the diagnostic criteria for DISH used to categorize the specimens.Speci fically,the diagnosis of DISH requires four continuous motion segments with radiographic evidence of bridging mineralization,a threshold thought to be associated with an advanced stage of disease.In fact,some of the spines that did not meet the diagnostic criteria for DISH showed ectopic mineralization bridging more than one but less than four motion segments.To eliminate potential variability associated with the stage of DISH,we analyzed the association between the presence of hyperdense foci and mineralization bridging the IVD in all motion segments,independent of the classi fication as DISH(Table 3).Of the 171 motion segments assessed,110 showed hyperdense fociwithin the IVD.Of those,70(64%)had mineralization bridging the IVD,demonstrating astrong association between these two features(P＜0.000 1).

DISCUSSION

The features of ectopic mineral formation that de fine DISH have been primarily characterized through clinical imaging,with few previousreportscontaining detailed characterization of the tissuespeci fic changes within spinal motion segments.The present investigation addressed this gap by characterizing the pathological features of ectopic mineral structures associated with DISH in human cadaveric tissues using a combination of radiological,histological,and physical approaches.Our findings are the first to establish that ectopic mineralization in DISH results from both heterotopic ossi fication and dystrophic calci fication of spinal tissues,including the IVD and anterior longitudinal ligament.Based on these and previous studies from our group,27we postulate that the two forms of ectopic mineralization may re flect different disease processes or perhaps distinct stages in the pathogenesis of DISH.

Table 2.Prevalence of hyperdense mineralized fociin all motion segments with or without ectopic mineral bridging of the IVD as determined by μCT,in spines that meet or do not meet the diagnostic criteria for DISH

The classic radiographic feature of DISHisthe presence of flowing mineralization along the anterolateral aspect of the vertebral column,1but our previous report identi fied heterogeneity in the morphological appearance of ectopic mineral associated with DISH based onμCTanalysis.27This heterogeneity was likewise evident in the histological appearance of individual motion segments from spines with DISH based on variability noted in the structure,organization,and features of the mineralized bridges along the anterior aspect of the spine.For example,despite demonstrating a similar radiographic appearance,variability was detected in the featuresof mature lamellar bone(with or without bone marrow)and indicators of endochondral ossi fication.Importantly,our study also identi fied discrete regions of amorphous,acellular calci fied material within the outer lamellae of the annulus fibrosus,the anterior longitudinal ligament,and fibrocartilage extending from the IVD adjacent to anterolateral bony bridges.These regions showed histological features consistent with dystrophic calci fication and were detected regardless of the morphological classi fication of the ectopic mineral structure(i.e.,discontinuous-patchy,vertical,and horizontal presentations).Regions of dystrophic calci fication were histologically distinct from areas of heterotopic ossi fication and demonstrated differences in calcium and phosphorus content;on the other hand,both showed X-ray diffraction patterns matching that of calcium-de ficient hydroxyapatite.Consequently,DISH may not solely be a pathology of bone formation (heterotopic ossi fication)but also may include featuresof dystrophic calci fication.Notably,μCT analysis of areas of dystrophic calci fication demonstrated radiodensities exceeding those in unaffected vertebral bone and in areas of heterotopic ossi fication.These hyperdense foci of dystrophic calci fications were present in all three of the described morphological presentations of disease and were associated with the presence of ectopic mineralization bridging the IVD.While it is tempting to speculate that these distinct histopathological features of dystrophic calci fication may re flect different stages of disease,further studies are required to assess their relation to the spatiotemporal pattern of disease progression.

Important to the diagnosis of DISH is the exclusion of other spine pathologies,such as spondylosis(i.e.,IVDdegeneration)and ankylosing spondylitis.2Speci fically,in contrast to spondylosis,IVD height is preserved,and osteophytes are vertical and bridging in DISH,whereas in spondylosis,they are usually transverse,based on radiographic evaluation.30A pilot study conducted with spinal tissues from ten cadavers with DISHestablished that the degree of histopathological IVD degeneration and measures of disc height were comparable between age-and sex-matched specimens without DISH,suggesting a limited role for IVD degeneration in the pathogenesis of DISH.25The histopathological evaluation in the current study demonstrated mild-to-moderate IVD degeneration in all specimens examined,which may have been insuf ficient to affect radiographic disc height.Note that loss of IVD height would preclude the diagnosis of DISH based on radiographic criteria.2These findings are in keeping with previous studies demonstrating degenerative changes in the IVDs of people with DISH.31–32Taken together,mild-to-moderate IVD degeneration should be reconsidered as an accompanying clinical scenario in the formation of ectopic mineralin DISH,since it iscommon in this population.

Table 3. Prevalence of hyperdense mineralized foci in all motion segments with and without ectopic mineral bridging of the IVD as determined byμCT

Despite the radiographic preservation of IVD height(with or without mild-to-moderate degeneration),the current study demonstrated that morphological changes to the outer annulus fibrosus are frequent in DISH.Our findings agree with those of previous studies that identi fied fibrocartilage extending from the IVD between the anterolateral bony bridges associated with both the superior and inferior vertebrae,lying adjacent to the anterior longitudinal ligament.25,33In fact,Kuperus et al.proposed a scoring system to characterize the gross morphological shape of these fibrocartilaginous extensions from the IVD(de fined as regular,tapered,spatulate,or irregular29—several of which were noted in the current investigation).In our study,histopathological evaluation of these fibrocartilaginous regions revealed features of endochondral ossi fication,including nestsof chondrocytesand primary woven bone,as well as distinct areas of dystrophic calci fication.

We theorize that dystrophic calci fication may be the result of cellular changes or necrosis of resident fibrocartilage cells in response to trauma,degeneration,or age.In this regard,injury to soft tissues,such ascardiovascular and connective tissues,including tendons,can lead to local aberrant production or unmasking of molecules that serve as substrates to nucleate the initial formation of calcium phosphate crystalswithin the extracellular matrix,leading to dystrophic calci fication.34–35It is possible that dystrophic calci fications within the soft tissue structures of the spine(i.e.,annulus fibrosus and anterior longitudinal ligament)play a role in the pathogenesis of DISH.There are several potential mechanisms by which dystrophic calci fication could lead to heterotopic ossi fication.For example,it is known that calcium phosphate itself within soft tissue can induce ossi fication,36–38given the presence of appropriate precursor cells and microenvironmental conditions.Thus,dystrophic calci fication could directly lead to heterotopic ossi fication in spinal tissues.It is possible that within the spinal ligaments and periphery of the IVD,the right conditions exist to permit ossi fication.In contrast,within the deeper lamellae of the annulus fibrosus,dystrophic calci fications might persist without transitioning to heterotopic ossi fication due to the lack of appropriate osteogenic precursors or an osteoconductive microenvironment.A second potential mechanism is that dystrophic calci fication in spinal fibrocartilage leads to altered tissue biomechanics.Increased tissue stiffness,due to the presence of dystrophic calci fication,may induce the formation of osteophytes.In this regard,others have proposed that increased mechanical stress on vertebrae can lead to the formation of osteophytes.39It is conceivable that these osteophytes may then lead to bridging ossi fication of the spinalligamentsand the periphery of the IVD.This possibility isin keeping with previousstudiesproposing that DISHis associated with degenerative osteophyte formation,similar to that occurring in spondyloses and osteoarthritis.4To resolve the mechanism,we require further studies of the soft tissues of the spine at early stages of ectopic mineralization.

There are mixed reports regarding the involvement of the anterior longitudinal ligament in the pathogenesis of DISH.Some studies have reported mineralization of the anterior longitudinal ligament in DISH,21,24,40whereas others report that the anterior longitudinal ligament does not undergo mineralization1,21or that it is displaced by the formation of ectopic mineral structures.26In the current study,when present in the histological sections examined,the structure of the anterior longitudinal ligament was typically preserved.However,in a subset of specimens,we detected focal regions of dystrophic calci fication within the ligament,resembling those detected within the annulus fibrosus.Further studies are required to speci fically examine the spatiotemporal association between the calci fication of the IVD and the calci fication of the anterior longitudinal ligament in the context of DISH progression and severity.

The findings from the current study establish that current radiographic criteria for DISH capture heterogeneous features associated with both dystrophic calci fication and heterotopic ossi fication of fibrocartilaginous tissuesof the spine.Taken together,the current investigation infers that DISHpathogenesis involves two distinct types of ectopic mineralization based on(i)the morphological appearance of the mineral;(ii)their calcium and phosphorus content;(iii)their radiodensity;and(iv)the tissuesaffected.Similar to calci fic tendonitis,the detection of focal areas of dystrophic calci fication within fibrocartilaginous tissues of the spine may serve as a radiographic indication of early-stage DISH.Future work is required to validate the association of these radiographic and histological features in a larger cohort of specimens,to correlate histological featureswith disease symptomsand/or comorbidities,to identify biomarkers to differentiate these presentations in clinical studies,and to associate clinical symptomsand disease featureswith the progression of the disease.

MATERIALSAND METHODS

This study was conducted with intact embalmed human cadaveric spines(n=19:6 females and 13 males,median age 85 years,range 65–94)from the Schulich School of Medicine&Dentistry at The University of Western Ontario in accordance with the Anatomy Act of Ontario and guidelines of Western’s Committee for Cadaveric Use in Research(REB#10292018).The embalming process consisted of arterial distribution of embalming fluid,containing a mixture of ethanol,phenol,and formalin(Wessels&Associates:Troy,MI,USA),24 to 48 h postmortem.The removal of the head and neck musculature exposed the first cervical vertebrae,and complete resection was performed inferior to the twelfth thoracic vertebral body.The ribs were dissected 3–5 cm lateral to the costovertebral joints,and soft tissues associated with the spine were preserved except for the descending thoracic aorta,which was removed.Microcomputed tomography(μCT)imaging

Intact human vertebral columns were previously scanned byμCT as described27using a cone-beam X-ray imaging system(GELocus eXplore Ultra:London,CAN)at a peak voltage of 80 kVp and tube current of 50 mA.The 1 000 X-ray projections were reconstructed into a single three-dimensional volume with an isotropic voxel spacing of 154μm.Image volumes were rescaled into Houns field unitsusing an internal calibrator of air and water and corticalbone substitute(450-SB3,Gammex RMI:Middleton,WI,USA).TheμCT data were used to generate a series of images for each specimen that were previously assessed by two clinician observers to diagnose DISH using Resnick and Niwayama’s radiographic criteria.2Three-dimensional isosurface renderings and pseudocoloredμCTimages were exported from MicroView(Version 2.2,GE Healthcare:London,CAN;and Version 2.5.0,Parallax Innovations Inc.:Ilderton,CAN).Data fromμCTwere grouped into contingency tables and assessed by two-sided Fisher’s exact test(α =0.05).Histological evaluation

Six thoracic spines meeting the diagnostic criteria for DISH(one female and five males;median age 82 years,range 72–87)were dissected into individual motion segments by transverse cuts across the waist of the superior and inferior vertebrae to maintain the IVD and by oblique cuts to remove the posterior osseous features from the vertebral body(Supplemental Fig.S1a).Multiple motion segments from each specimen were selected for analyses based on the presence of distinct presentations of ectopic mineralization previously described by our group(i.e.,discontinuous-patchy,vertical,and horizontal).27A total of 15 motion segments were analyzed.

To isolate tissues of interest from the intact motion segments,a sagittal slice was made through the center of the ectopic mineralization associated with each motion segment using a rotating diamond blade saw(Supplemental Fig.S1b).In cases of rightsidedness,oblique cutswere made through the center of the ectopic mineralization(Supplemental Fig.S1c).From one of these halves,a 1-mm-thick slice of tissue was dissected for subsequent analysis of tissue composition.Each half of the isolated motion segments and the 1-mm-thick slices were imaged using a digital X-ray system(Planmeca ProX™:Helsinki,FIN)to localize areas of ectopic mineral.Based on the digital radiographs,one half was selected for histological analysis and decalci fied with Shandon™ TBD-2™ (catalog no.6764004,Thermo Scienti fic™:Nepean,CAN)for a duration of 14–21 days.Following decalci fication,the tissues were processed and sectioned at 5μm thickness using a Leica RM2255 microtome(Leica Biosystems Nußloch GmbH:Nußloch,DEU)and collected on 50×75 mm slides(Brain Research Laboratories:Newton,MA,USA).Serial sections were stained with hematoxylin and eosin to localize cells and to evaluate extracellular matrix structure,and Masson’s trichrome to detect the collagenous component of the extracellular matrix.Low-magni fication micrographs were captured using a Cytation™5 Cell Imaging Multi-Mode Reader(BioTek Instruments,Inc.:Winooski,VT,USA),and high-magni fication micrographs(including those under linear polarized light)were captured using an Olympus BX41 opticalmicroscope equipped with adigitalcamera(Olympus U-TVO.5XC-3,Olympus Canada Inc.:Toronto,CAN)and In finity Analyze software(Version 6.5.5,Lumenera Co.:Ottawa,CAN).Images were imported into Adobe®Photoshop®CC 2018(Version 19.1.7,Adobe Systems Inc.:San Jose,CA,USA)for figure construction.

Scanning electron microscopy,energy-dispersive X-ray

spectroscopy,and X-ray diffraction

The 1-mm-thick slices of tissue taken from the center of the ectopic mineralization were desiccated with Drierite(W.A.Hammond Drierite Co.:Xenia,OH,USA)for 30 days,photographed with a Canon EOS7D digital single-lens re flex camera,rinsed in 100%chloroform for 1h,and subsequently redried for 1–2 days before coating with a 10 nm layer of osmium.41Scanning electron microscopy was performed using a Zeiss 1540XB FIB/SEM instrument(Carl Zeiss:Oberkochen,DEU)at The University of Western Ontario Nanofabrication Facility.Energy-dispersive X-ray spectroscopy data were collected using an Oxford Instruments X-max50 analysis system and ICNA software(Oxford Instruments:Abingdon,UK).Multiple regions of interest within each area of ectopic mineral(10–16 regions per deposit)were analyzed,along with regions of interest in unaffected cortical bone within the same specimens(e.g.,posterior aspect of the vertebrae).The elemental results were expressed as atomic percentages.Data were imported into GraphPad Prism(Version 6.01:San Diego,CA,USA)for statistical analysis.Data were assessed for normality using the Shapiro-Wilk test and for outliers using Grubbs’two-sided test(α=0.05:one result from cortical bone was identi fied).Data were then assessed using one-way ANOVA(α =0.05)with Bonferroni’s multiple comparisons test.

The same tissueswere assessed for X-ray diffraction patterns using a Bruker D8 Discover diffractometer(Bruker Co.:Billerica,MA,USA)and Bruker AXSgeneral area detector diffraction system(Bruker AXS GmbH:Karlsruhe,DEU)at the Department of Earth Sciences,University of Western Ontario.42The nominal beam diameter for each measurement was 300μm.Regions of interest within the areas of ectopic mineral(5–7 regions per deposit)and unaffected cortical bone(1–2 regions)were selected for each motion segment using a microscope equipped with a charge-coupled device camera.Data were analyzed with Bruker’s DiffracPlus™ EVA software(Bruker Co.:Billerica,MA,USA)for comparison to powder diffraction files from the International Centre for Diffraction Data®(Newton Square,PA,USA)database as detailed in a previous protocol.42

ACKNOWLEDGEMENTS

First and foremost,we must acknowledge the do nors to our institution’s bequeathal program and their families for their altruism,without whom this research could not have been accomplished.Thisstudy wassupported by the Canadian Institutesof Health Research(grant number 115068).CASis supported by a New Investigator Award from CIHR and an Early Researcher Award from the Ontario Ministry of Research and Innovation.DEFwas supported in part by a Transdisciplinary Training Award from the Bone and Joint Institute at The University of Western Ontario.We also thank the following colleagues for their contributions:Drs David Holdsworth and Joseph Umoh(Robarts Preclinical Imaging Research Centre)forμCT scanning;Dr Marjorie Johnson and Haley Linklater for access to the Hasse Education in Anatomy and Research Technologiescadaveric laboratory;Kevin Walker for assistance with cadaver dissections;Dr Stephen Ferrier(Schulich Dentistry)for digital X-ray imaging;Dr Todd Simpson(Western Nanofabrication Facility)for SEM/EDX expertise;Dr Roberta Flemming and Alex Rupert(Western’s Department of Earth Sciences)and Nadia Sharma for XRD expertise;Drs Frank Beier and Michael Pest for access and assistance with imaging equipment;and Diana Quinonez,Linda Jackson,Tom Chrones,and Caroline O’Neil for assistance with histology.

AUTHORCONTRIBUTIONS

Conceived and/or designed the work:DEF,SJD,CAS.Acquired data:DEF,RJB.Important role in interpreting the results:DEF,RJB,PKK,SJD,CAS.Drafted or revised the manuscript:DEF,SJD,CAS.Approved the final version of the manuscript:all authors.

ADDITIONAL INFORMATION

The online version of this article (https://doi.org/10.1038/s41413-020-0091-6)contains supplementary material,which is available to authorized users.

Competing interests:The authors declare no competing interests.