Diagnosis of partial and total physeal arrest by bone single-photon emission computed tomography.

Abstract

Bone single-photon emission computed tomography (SPECT), capable of creating maps of the distribution of osteoblastic activity in every spatial plane of a physis, should provide images of diagnostic value in the case of patients suffering from growth arrests (epiphysiodeses). Seventy-five bone SPECT scans were obtained in 64 children suspected to have developed physeal arrests. The transaxial slices of the physis, in the case of partial epiphysiodeses: (a) indicated the percentage of the remaining normal physis, (b) located the bony bridge within the physis and (c) showed the slowdown of the growth of the remaining normal physis induced by the bony bridge in some children. Misdiagnosis occurred in six patients. For total epiphysiodeses, the radionuclide diagnosis was confirmed in 20 of 21 patients. Radionuclide, x-ray and MRI examinations in the study of growth disturbances were found to be complementary.

Full text

Page 1

greater the potential for these complications.
The surgicaltreatmentof epiphysiodeses is based on the
capacity of regenerationof the remainingnormal physis
(4). Two procedures are possible in the case of partial
epiphysiodeses:(1) when less than half of the cartilaginous
surface remains normal,the epiphysiodesis
and the eventual associated axial deformity corrected; (2)
when more than half of the cartilaginoussurface remains
normal, the bony bridge is resected and the defect filled
with inert material (physiolysis) (Fig. 3). In these cases,
when growth is reestablished, new bone is formed below
theinertmaterialwhichis thenpushedupwithinthedi
aphysis (Fig. 4).
An important part of the preoperative
epiphysiodeses is:(a)thelocationof thebony bridgewithin
the physis and (b) the estimation of the percentage of the
remaining normal physis. For this purpose, mapping the
distributionof the ossified and remainingnormalphyseal
areas is necessary. At present, the rough distribution of the
bony bridgewithin a physis is determinedby radiological
tomography(6).Coronaland sagittal
region of interest (ROl) are obtained;
bony bridgearemeasuredon each cut andtransferredonto
a cross-sectioned diagrammaticrepresentationor map of
thephysis(Fig.5). More preciseanatomic information is
obtained by computed tomodensitometry, which locates
trabecular bone with semi-coronal slices, and by magnetic
resonance imaging (MRI), whichlocatesnormal cartilage
with coronal and sagittal slices (7—10). The three-dimen
sional radiology is still experimental.
The examinations on the distributionof bone mineral
content andnormalcartilagewithin aphysis by x-rays and
Mifi canbecomplemented by theradionucideexamina
tions that provide informationon the distributionof the
osteoblastic activity within the same region. The propor
tionality between the uptake of an osteotropic radiophar
maceuticalby a skeletal ROl and the osteoblastic activity
within this region is widely documented (11—22). There
fore, bone single-photon emission computed tomography
is completed
planningof partial
tomograms
the dimensions
of the
of the
gWethe rnqec@ation
girls(5).
ofgrowthforevery physis atalages, SepeI@fOrbOyS and
Bonesingle-photon emission computed tomography (SPECT),
capable of creating maps of the distributionof osteoblasticac
tMtyineveryspatialplaneofaphysis, shouldprovide images of
diagnostic value in the case of patients suffering from growth
arrests(epiphysiodeses). Seventy-five boneSPECT scanswere
obtained in64thildren suspected tohavedeveloped physeal
arrests. Thetransaxial slicesofthephysis,inthecaseofpartial
epiphyslodeses: (a)indicated thepercentage ofthe remaining
normalphyals, (b)located thebonyb@dge withinthephysisand
(C)showed the slowdown of the growth of the remaining normal
pt@s induced bythebonybridgeinsomechildren. Misdiagno
sisoccurred insb patients. Fortotalepiphyslodeses, therad,o
nudidediagnosis wasconfirmed in20of21patients. Radlonu
dide, x-ray and MRI examinabons in the study of growth
disturbances werefoundto becomplementaiy.
J Nuci Med 1993; 34:1410—1415
children, some injuries and/or infections of the physes
may be complicated by growth arrestsresultingfrompro
mature ossificationof a part or whole of the surface be
tween the epiphysis and the metaphysis (partialand total
epiphysiodesisrespectively) (1). The clinical aspects of the
epiphysiodesesdepend upon the percentage of the physeal
surface ossified (2). When the entire area is ossified,
length discrepancy develops. In the case ofpartial epiphys
iodeses, gradients of growth exist within the physis and an
axial deformity of limb develops (Fig. 1). Then, the ossified
region, usually named the “bony bridge,― often clearly
visible on the x-ray documents (Fig. 2),blocks out progres
sively the growth of the remaining normal physis, thereby
reducing its metabolic activity (Hueter and Volkmann's
law) (3). The younger the child andthe greaterthe expect
ancy of growth of the physis at the time of injury@,the
a limb
Received Dec. 3,1992; revIsion accepted Apr.27,1993.
Forconespondenceorrepdn@ con@ M@hsl Woland, Service deMédecine
Nuclésire,HOpltsl Sslrit-Mtnlne, 184,rueduFajbourg Sslnt-Mtoine, F 75012
Paris, Fraioe.
•Eveiy ph@s contilbutes a fixed percentage
pie,thed@tsl temomi physls cortibutes 70%ofthekngltucSnal
temtt,and theproodmalphysls 30%; thetbsl prnidmsl physls contilbutes
thelon@kislgrowthofthebbieandthedistsl
@di@hysesl growth. For exnoi
growth ofthe
55% of
physis45%. Inthehtsiature, tables
1410
TheJournalof NudearMedicine •Vol.34•No.9•September 1993
Diagnosis of Partial and Total Physeal Arrest by
Bone Single-Photon Emission Computed
Tomography
Michel Wioland and Valerie Bonnerot
Se,vice de Médecine Nudéaire,HOpital Saint-Antoine and Service de RadiOlOgie Pediat,iqUe, HOpital des Enfant-Malades,
Paris, France

Page 2

I
FIGURE 3.
radiograph during treatment of
partialepiphysiodesis of radial
distalphysis(samepatientas
In Figure2).The bonybridge
was resectedand the defect
filled with inert cement.
lntraoperative
Sixty-fourchildren2 to 15yr of ageat theirfirstexamination
wereinvestigated (meanage:8.9yr).Sixchildren wereinvesti
gatedtwice and one threetimes. Threechildrenexhibiteda
growth arrest at two different sites, simultaneously
orat2yrof interval (onepatient).
The regionsexaminedwere the shoulders(2 patients),the
wrists (2 patients), the hips (2 patients), the knees (48 patients) and
the ankles(11patients).The originof the epiphysiodesis was
traumatic(29 patients),infectious
patients),vascular(purpurafulminans)in2patients,cystic,dys
plasic, genetic (NOONAN syndrome) and iatrogenic in 1 case
each.Theoriginremained unknown inthreecases.Fourchildren
affected by Blount's disease (deformity of the internal extremity
of the tibialproximalphysisin childrenmostlyfromthe West
Indiesand Scandinavia)were examinedto diagnosean eventual
epiphysiodesis of thetibialproximal physis.
Among15children suffering fromacompletegrowtharrest,8
remain untreated to date because skeletalmaturity is awaited
before a final therapeutic decision is taken. The others were
(two patients)
(17 patients),therapeutic (2
FIGURE
physlolysis (same patient sam
Figures2 and 3). Growthof
radialdistalphysisreappeared
andthe bk@ck of inertcement
iswithin thediaphysis.
4. Successful
F@V
1. Radlographotkneesofchfldsuffeiingfromaxlalde
formity ofleftlowerextremity (partial epiphysiodesis).
ofexternal halfofleftproximal tibialphyseal bar.
I@
Notethinning
(SPEC!'),capable of creating mapsof thedistributionof
osteoblastic activity in eveiy
should clearly provide images of diagnostic value in the
case of patients suffering from growth arrests. The objec
tive of the present work was to evaluate the capability,
limits and pitfalls of bone SPECF in the diagnosis of the
various types of epiphysiodeses.
spatial plane of an ROl,
FIGURE 2.
partial epiphysiodesis
distal physis. The ‘bOny
bridge―isclearlyvisibleasa
whiteareaatjundionbetween
internal and medial thirds of
— bar.
POsttraumatiC
ofradial
1411
BoneSPECTinPhyseal Arrest•WiolandandBonnerot
MATERIALSAND METHODS
Patients

Page 3

;:
@fUft
@/1
@1 I
,@
-rr
+@
4÷
&-*@t@t1:
L&
4':
:@
‘S
@-
@TEffiOR
354 45 555555 7 71 5 551
POSTE@OR
1S@0
@TE@OR
A
zz@
flo@
2,•
210
@z0s ___
200@ —.
1.11‘15
‘10
_J ‘SI
‘So
-
‘itt
@.
@
B
225
220
21 S
210
205
@200
_J@
@
@
‘So
FIGURE 5.
tationof ph@
tomogram In child sufferingfrom partial
eplphy&odeals.(B) Physealbar in the an
teroposteriorprojectiontransferredto map
as a thick,strai@itNtis.(C)M anteroposte
rkxand@mographk@
growtharrestpbttedaccordlngto loca
tion. (D)Anal cross-sectioned mapof phy
seal bar in diatalfemur of child (reprinted
w@1 permlealon fromCarisonandWenger,
ret6).
(A) Diagrammatic represen
baron an anteroposterior
@iavelsinvc@ved
MITERIOR
221
220
21 5
210
200
‘10
:@ 205
M@@I5
<
_.I ‘55
‘So
g
35445 5555S5 7 755 555
POSTERK@R
1S@0
C
D35SSISSSSSSPP5SSSIISIO
@ffiOR
treated by diaphyseal lengthening(23), with orwithout osteotomy
forcorrectionof theeventual,associatedaxialdeformity (4pa
tients), by therapeutic epiphysiodesis
patient)andby singleosteotomy(2patients).Threechildrena!
fectedbyacompletegrowtharrestofone tibialdistalphysiswere
treated by therapeutic epiphysiodesis of the corresponding fibular
physis.
The childrensufferingfroman axialdeformityof limb(32
patients) weretreatedbyphysiolysis(14patients), completion of
the epiphysiodesiswith or without correctionof the axial defor
mity (8 patients) and by single osteotomy (5 patients). Five chil
thenremainuntreated so far.
l@ Data
Dimethylaminodiphosphonate
maceuticalchosenbecause the data availableshow that the le
sion-to-normal bone uptake ratio is increased by 25%, compared
to thatof theotherdiphosphonates currentlyin use (24).Two
hours after the intravenous injection of 7 MBq of Kg-i, planar
scintigraphy of the ROl was carried out with a gamma camera
GeneralElectricStar 300.Then, 64tomographicprojectionsof
theROl(20to 40sec each)wereacquired ina 128x 128pixel
matrix. The first 16patients were studied with a tomoscintigraph
Elscint Apex 406 (circular rotation ofthe circular detector 400mm
in diameter).The otherpatientswere examinedwith a tomo
graphicgammacameraGeneralElectricStar400(ellipticalrota
tion of the circular detector 400mm in diameter). It took —40 mm
tocollectdataforeachpatient.
Transaxial, saglual and frontal slices (1 pixel thick) were cre
ated by filtered backprojection (Butterworth cut-off frequency
between0.8to 1.2cycle/cm;order: 10—12)
ing thickness of the physeal region investigated is 3.2 mm. For
some patients, additions of two transaxial slices were carried out
toobtainbettersmoothing ofthedata.Limblength discrepancies,
ofthe contralateralphysis(1
was the osteotropicradiophar
(25).Thecorrespond
frequent in such children, complicate interpretation of the data
because the transaxial slices of one physis do not involve the
contralateral one: separate magnification of the left and right
halves of the scintigraphic data allowed a simultaneous display of
therightandlefthomonymous physesonthefinalimages.Such
treatmentwaspossiblewiththeGeneralElectricStar300system
only.
Analysis of the Slices
Epiphysiodeses are rare (one new case per year for 1 million
individuals). More than 100patients have been examined to date,
butdatacollectedintheearlystagesof gainingexperienceand
datafromchildrenstillundermedicalcontrolhavenotbeenin
cludedinthestudy.A singlephysicianinterpreted thedata Ex
perience in reading the images improved during the 10-yr period
needed to collect the data. The criteria taken into account were
theshapeofthephysisandthepresenceofanabnormality inthe
uptake of the radiopharmaceutical. To avoid creation of artifac
tual images, attention was paid to the differences in thickness
and/or the uptake of the bone-seeker between the periphery and
thecentreofthephysis(21).Inthecaseofpartialepiphysiodeses,
theconfiguration andtheextentof thebonybridgeweredeter
mined. The area of the remaining normal physes was estimated
visuallyand expressed as beingbelow
physeal area. The scintigraphicdata were correlatedwith the
clinicalfollowup and type oftreatment applied, which established
thediagnosisretrospectively, aswell as withtheintraoperative
observation when physiolyses were carried out.
or abovehalf of the entire
RESULTS
Scintigraphic Aspects
NonnalAspects.
(26,27).Due to the lack of information
The anatomy ofphyses varies with age
of tomoscinti
1412
TheJournal ofNudear Medicine •Vol.34•No.9• September 1993

Page 4

____________
@,. die bonybridge,
uptake inthemedial partofthephysle.
n
F)l @@levL@1easaIineofdecreased
FiGURE 6. Transaidal tomosdntmgrams oftemoraldistalphyses
of child sufferingfrom partiale@physiodesle (Summationof the
transa@dalslicesoftheregkn ofinterest).Thebonybridgeisdevel
opedIntheexternalhailofthephysis.Theuptakeof radiopharma
ceuticalby the remainingnormalphysisis lowerthan that of the
contralateral normalphyals(FlusterandVolkmann'slaw;seetext@.
Blount's Disease. Both planar and tomoscintigraphic
aminations showed the typical deformity of the tibial prox
imal physes (Fig. 9).
cx
Efficiency of Bone SPEd
Nor,nallndividuals.
epiphysiodesis
occurred(nine examinations).
Total Ej$physiodeses. The clinical evolution confirmed
the radionucide diagnosis in 20 of the 21 cases studied
(Table 1). Once, when the diagnosis of partial epiphysiode
sis was given, a complete epiphysiodesis developed. For
three childrenexhibitinga limb length discrepancy,
uptakeof the radioactive bone-seeker
physis was below that of the contralateral
No diagnosis of partial or complete
was given when no disturbanceof growth
the
by the suspected
side with an
DT
graphic slices of the various physes at all ages of childhood
andadolescence, we comparedtomoscintigraphicslices of
the physis of interest to those of the contralateralside.
TotalEp4thysiodeses. In allcases, adecreased uptake of
the radioactivebone-seeker by the ossifiedphysiswas
observed on planar bone scans and on tomoscintigraphic
slices. The SPECF studies showed thatthe decreased up
takewas homogeneous inthe cases of therapeuticepiphys
iodesis as well as in some of the diseases of infectious and
traumaticorigin. In the other cases, the decreased
was marginally heterogeneous.
Partial Ep@physiodeses.In the case of partial epiphys
iodeses, the tranSaXial slices mapped the distribution of
osteoblastic activity within the physis (Fig. 6). These maps
located the bony bridge and indicated the percentage of the
remainingnormal physealarea. The sagiftal and frontal
slices showed the direction of the bony bridge in three
dimensions(Fig. 7). The planar bone scan (Fig. 8) did not
contain such informationon the location, dimensions
obliqueness ofthe bony bridge; it showed only a decreased
uptake of the externalhalf of the femoral distal physis,
suggesting the diagnosis of partialepiphysiodesis.
In 32 children, the radioactivity
below that of the remaining normal areas (Figs. 6, 7), and
greaterin 2 others. Moreover, comparisonof the radioac
tivity of the remainingnormal
the normal contralateral one (Figs. 6, 7), showed that the
radioactivity of the remaining normal physis was similar to
thatof the contralateralphysis in 20children,andlower in
10 others. In 15 patients, the area of the remaining normal
physis was above halfofthe entire physeal area, and below
this value in 14 others.
uptake
and
of the bony bridge was
area of thephysisto thatof
C,
FiGURE 8.
halfof leftfemoraldistalphyals(samechildas in Agures6 and7).
Notedisappearance of physiological increased uptakeinexternal
halfofthe physisandthe physealdeformity.
Planarbonescanofpartialepiphyskxlesis ofexternal
BoneSPECTinPhyseal Arrest•WiolandandBonnerot
1413

Page 5

Scintigraphic
Cirilcal diagnosisCorrect
9
Erroneousdiagnosis
—
False
positiveFalse
—
negative
—
Normal
(9examinations)
Totalepiphysiodeses
(21examinatIons)
Partialepiphyslodeses
(33examinatIons)
MISCeIIaneOUS
(12exw*iatlons)
20
12
•The
disease,physealdeformitiesandgeneralizedslowdownofthe growthof
k@ig bOnes.
group “miscellaneous―correspondsto cMdren with BIount@s
taken up by any other part of the skeleton. Therefore, the
tomographicprojectionswere acquiredina 128x 128pixel
matrix instead of the usual 64 x 64. Consequently, the
spatial resolution of the maps was twice thinner than that
commonly obtained in nuclear orthopedics.
The limits of demarcation between the normal and
pathologicalareas of an altered physis are always uncer
tam. No method of imagingdelineates the impaired physeal
region withprecision. Radiology remains uncertain inthe
interpretation ofthe progressive thinning of the physeal bar
in the neighborhoodof the pathologicalarea. Mill has the
same limitation in considering
ance of the cartilaginous signal. The direct observation of
the bony bridge,feasible duringphysiolysis, cannot deter
mine the exact limits of the impaired region either, since
surgeons remove tissues until normal cartilage is clearly
visible. Comparison between the great dimensions of the
block of inert cement put in the place of a bony bridge
(Figs. 3, 4), and the dimensions of the bony bridge itself
(Fig. 2) illustrates the uncertainty
ossified region within a physis partially altered. Therefore,
bone SPEC!' should be tested about its capability to pro
vide answers for the following questions relevant to ther
apy: (a) Is the disturbanceof physeal
complete? (b) Is the ossified region central or peripheral?
(c)Isthenonfunctional areagreaterorsmallerthanhalfof
the physeal area?
This study showed that in the case of partial epiphys
iodeses, bone SPECTprovidedmapsof the distributionof
osteoblastic activity within the physis of interest. These
maps located thebonybridge andallowed theestimation of
the percentage of the remainingnormal physeal area. In
some children, the decreased
mal physeal area showed the slowdown of the growth of
this area induced by the bony bridge (Hueter
the progressive disappear
in the delineationof an
growthpartial or
uptake of the remaining nor
and Volk
mann's
However, only unilateral disturbances ofskeletal growth
maybe establishedwith certaintyinthe absence of anatlas
ofnormal tomoscintigraphic slicesofthedifferent physes.
Thedimensionsof thelesion shouldbe greaterthan1cm to
be detectable. Fortunately, this situation is the most fre
quent since the physes with the greatest physiological im
portance are “near the knees and far from the elbows,―
and have the greatest dimensions. The dimensions of the
physes also should be greater than 1 cm. Three children
referred for investigation of the elbows were examined
with planar scintigraphy only.
Variations in the uptake of the radioactive bone-seeker
by the various partsof a physis partiallyossified (20) may
relate to redistributionsof osteogenesis within the physis
following its progressive deformity. These variationsmay
correspond also to the successive steps of maturation of
the bony bridge, according to the following hypothesis:
first, the impaired physeal area, probably constituted with
tissue inthe ossificationprocess, would concentrateradio
activitygreaterthan the adjacent
uptakeof the traceris especially high in the calcified car
law).
cartilaginousregions (the
[‘PC' IT
FIGURE 9.
Blounrsdisease.Theproximaltiblalphysishasatypicaldeformity.
Planar bone scan of knees of child affected by
uptake ratio between 0.8 and 0.9. This suggested the diag
nosis of idiopathic generalized slowdown of growth.
PartialEp@physiodeses.The clinical evolution confirmed
the radionucide diagnosis in 24 of the 33 patients studied.
The conclusions were false-negative in five patients and
erroneousin one other patient: a complete epiphysiodesis
developed, whereas the diagnosis of partial epiphysiodesis
wassuggested(the samecase
four children suffering from an axial deformity, a physeal
deformity without abnormalities in the uptake of the radio
active bone-seeker was observed.
as reported previously).In
DISCUSSION
Bone SPECF was accepted well by children although 40
min were needed to cariy out the examination.
activity taken up by the physes is much larger than that
The radio
TABLE I
Clinicaland ScintigraphicDiagnosesConsideredfor Children
Examined inStudy
5
1414
The Journalof NuclearMethane• Vol. 34 •No. 9 •September1993