Total-Hip Arthroplasty: Periprosthetic Indium-.

111-Labeled LeukocyteActivityand Complementary

Technetium-.99m-Sulfur

Colloid Imaging in Suspected Infection Christopher

J. Palestro, Chun K. Kim, Alfred J. Swyer, James D. Capozzi,

Robert W. Solomon,

and

Stanley J. Goldsmith Departments ofPhysics-Nuclear Medicine and Orthopedic Surgery, Mount Sinai School ofMedicine, Mount Sinai Medical Center, New York, New York ses. Neither pattern nor intensity of periprosthetic

ad

Indium-i i 1-labeled leukocyte images of 92 cemented to tal-hip arthroplasties were correlated with final diagnoses. Prostheses were divided into four zones: head (including

tivity reliably separated infected from uninfected ar

acetabulum),trochanter,shaft, and tip. The presence(or

are similar (24), and because osteomyelitis typically stimulates leukocyte accumulation and inhibits sulfur

absence) and intensity of activity in each zone was noted,

throplasties.

Because the physiologic

distributions

of

99mTcsulfur colloid and labeled leukocytes in marrow

and comparedto the correspondingcontralateralzone. colloid accumulation (25), combined labeled leuko Thoughpresentin all 23 infectedarthroplasties,peripros cyte/sulfur colloid imaging was performed in an effort thetic activity was also presentin 77% of uninfectedax to differentiate physiologic marrow accumulation from throplasties,and was greaterthan the contralateralzone infectious accumulation oflabeled leukocytes. 5i % of the time. When analyzed by zone, head zone We report the patterns of periprosthetic labeled leu activity was the best criterion for infection (87% sensitivity, 94% specifIcity,92% accuracy).Fiftyof the arthroplasties kocyte activity in 92 cemented total-hip arthroplasties, as well as the results of combined ‘ ‘ ‘In-labeled leuko were studied with combined labeled leukocyte/suttur col

bid imaging.Usingincongruenceof imagesasthecriterion cytefl9mTc@sulfur colloid imaging of 50 of these arthro for infection, the sensitivity, specificity, and accuracy of

plasties.

the studywere 100%,97%, and98%, respectively.While variablepenprostheticactivity makes labeledleukocyte MATERIALS AND METHODS imaging alone unreliable for diagnosing hip arthroplasty

infection,the additionof sulfurcolloidimagingresultsin a highlyaccuratediagnosticprocedure.

Patient Population Seventy-two consecutive patients with 92 (68 primary, 24 revision) cemented total-hip arthroplasties were retrospec tively reviewed. Fifty-two patients had unilateral arthroplasties

J NucIMed 1990;31:1950—1955

and 20 had bilateral arthroplasties. There were 26 males and 46 females with a mean age of 62 yr (range: 20—87yr). The

he noninvasive diagnosis of orthopedic implant infection is challenging. Radionuclide procedures used

include technetium-99m (99mTc) bone, gallium-67(67Ga)citrate,and indium-l 11-(‘ ‘ ‘In) labeled leukocyte (WBC) scintigraphy(1—23). Our initial experiences using ‘ ‘ ‘In-labeled leukocyte

mean time between arthroplasty (range: 1 wk—17 yr).

and scintigraphy was 4.6 yr

Arthropbasties were considered infected ifoperative cultures grew out organisms (n= 19) or if gross purubence was evident at surgery (n=4). Arthropbasties were considered uninfected if operative smears revealed no leukocytes and intraoperative cultures were reported as no growth (n=33). Arthropbasties not operated on were considered uninfected if: (a) they were

scintigraphy for diagnosis ofthis entity were unsatisfac tory. While absence of periprosthetic leukocyte activity

the asymptomatic arthroplasty in a patient whose contralaterab

was seen only in uninfected prostheses, periprosthetic

the asymptomatic arthropbasty in a patient undergoing beu

activity was present in infected and uninfected prosthe

kocyte imaging for reasons unrelated to that arthroplasty (n= 10), or (c) they were a painful arthroplasty that responded to conservative (anti-inflammatory, but not antibiotic) ther apy (n=6). No patients with clinically uninfected arthroplas

ReCeivedFeb. 20, 1990;revisionaccepted May15, 1990. For reprintscontact: ChristopherJ. Palestro, MD.Box 1141. Dept. of Physics-Nuclear Medione, Mt. Sin@ Medical Center 1 Gustave L. Levy

Place, NewYork,NY10029.

1950

hip arthroplasty

was symptomatic

(n=20), or (b) they were

ties developed symptoms referable to their arthropbasties dur ing a minimum follow-up period of6 mo after imaging.

TheJournalof NuclearMedicine• Vol. 31 • No. 12 • December1990

Scintigraphy

For the 50 arthroplasties studies with labeled beukocytes

Leukocyte scintigraphy was performed 24 hr after injection of -‘-18.5 MBq (500 MCi) of autobogous mixed leukocytes labeled with ‘ ‘ ‘In-oxine, according to the method of Thakur et al. (26). Six-minute anterior and posterior images of the arthropbasties were obtained on a barge field of view gamma camera equipped with a medium-energy parallel-hole colli mator using 20% windows centered over the 174 keY and 246

keV photopeaks of ‘ ‘ ‘In. The last 50 arthroplasties were studied with @mTc@sulfur colboid imaging immediately after the labeled beukocyte study. Patients were injected with 370 MBq ( 10 mCi) of@mTc@subfur

colboid, and 6 mm anterior and posterior static images were obtained 1—2 hr later on a gamma camera equipped with a low-energy, high-resolution, parallel-hole collimator, using a 10% window centered on the 140 keV photopeak of 99mTc. Prior to 99mTcsulfur colboid injection, to confirm that the contribution of ‘ ‘ ‘In photons to the @mTc image was minimal, a 6-mm anterior view of the region of interest was obtained using a 10% window centered around the 140 keV photopeak of 99mTc and the standard intensity setting for routine sulfur colloid marrow imaging. Because no discernible contribution of ‘ ‘ ‘In photons to the @mTc image was seen in any ofthe 20

and sulfur cobboid, images were classified as congruent if the

distribution ofthe two radiotracers was spatially identical, and incongruent ifbabeled leukocyte activity was observed without

corresponding activity on the sulfur colboid images. Four criteria for diagnosing infection were selected and compared:

Criterion 1: Any periprosthetic activity. Criterion 2: Periprosthetic activity greater than the contra lateral side. Criterion 3: Labeled beukocyte activity in the head zone,

regardless of activity in any other zone. (Ret rospective analysis of activity by zone mdi cated that this was the best criterion for diag nosing infection with labeled leukocyte imag ing alone). Criterion 4: Incongruent labeled leukocyte/sulfur colboid images.

The results reported represent the consensus of independ ent, blinded, random readings by two experienced nuclear physicians, who had no knowledge of the final diagnoses.

arthroplasties so studied, this is no longer routinely performed.

RESULTS Image Evaluationand DiagnosticCriteria Each arthropbasty was divided into four zones: head (in cluding acetabubum), trochanter, shaft, and tip (Fig. 1). The

presence or absence of labeled beukocyte activity in each zone was noted. When present, the intensity of this activity was compared to the intensity of activity in the corresponding

contrabateral zone. The pattern of periprosthetic labeled leu kocyte activity around each arthroplasty was also noted. Chi square statistics were used to analyze the significance of the relationship between periprosthetic labeled leukocyte activity (n=76) by zone and the presence or absence of infection.

Twenty-three arthroplasties were infected, all opera tively confirmed. Thirteen different organisms were identified: 6 gram-positive, 6 gram-negative and 1 fun gal. Sixty-nine were uninfected, 33 operatively con firmed, and 36 by one of the previously described clinical criteria. Eleven different patterns of peripros thetic leukocyte activity ranging from no uptake in any

zone to uptake in all four zones were identified (Fig. 2). Sixteen (17%) arthroplasties demonstrated no peri prosthetic labeled leukocyte activity: none ofthese were

infected. Seventy-six (83%) arthroplasties demonstrated vary ing degrees of periprosthetic

labeled leukocyte

activity:

all 23 infected and 53 uninfected arthroplasties. Peri

Patterns of Periprosthetic Leukocyte Activity in 92 Hip Replacements 0 Uninfected 14

U,@ Infected

13

Shaft

N

Tip FIGURE1 Diagrammatic representation of the four zones into which

eacharthroplastywas divided.

Total-Hip Arthroplasty • Palestro et al

H

1,

Tp

H.Tr Tr.S Tr.Tp H+Tr H,1

+s •ip•Ip.S.Tp

FIGURE2 Elevenpatternsof periprostheticlabeledleukocyteactivity wereidentifiedamongthe92arthroplasties reviewed.N = no penprostheticactivity, H = head and acetabulum,Tr = tro

chanter,S = shaft,andTp = tip.

1951

TABLE 1 Comparisonof Criteriafor Diagnosisof the InfectedHipArthroplasty AccuracyAny Presenceof

Infected

Uninfected

Sensitivity

Specfficity

activity23/2353/69100%23%42%Activity > contralaterai zone1 5/2327/6965%61%62%Head

activity20/234/6987%94%92%Incongruent zone leukocyte/sulfurcolloid10/1

01/401

00%97%98%

prosthetic activity was greater than corresponding con tralateral zone activity in 15 infected and 27 uninfected

prostheses. The most significant finding observed following analysis by zone, of76 arthroplasties with periprosthetic

orthopedic between

implant

infection has been reported to be

60—80% (8,12,15,20,36).

Indium-l 11-labeled leukocyte scintigraphy has also been evaluated, alone, and in combination

with other

leukocyte activity, was that infection was infrequently

radiotracers and has been compared to 99mTcbone and 67Gascintigraphy,with a wide range of results reported

found in the absence ofhead zone activity (3/49), while

(10—23).

20 of 24 arthroplasties with head zone activity were infected. The difference was highly significant (x2=43.2,

p40.OOl). When 16 arthroplasties without any uptake are included, Criterion 3 yielded a sensitivity and spec ificity of 87% and 94%, respectively.

Generally,

when

used

alone,

interpretation

of this

study has been based upon either the intensity of peri prosthetic labeled leukocyte activity in comparison to various system

reference points of periprosthetic

(12,15,1 7—20) or a grading uptake (14,16,22). The re

Of 50 arthroplastiesstudied with labeled leukocyte/ sulfur colloid imaging there were 10 infected and 40

ported sensitivity of the procedure

uninfected

45% to 100% (Table 2). The low specificity is especially

prostheses. Images were incongruent

in all

10 infected prostheses and 1 uninfected prosthesis. Im ages were congruent in 39 uninfected arthroplasties.

A comparison ofthe four criteriafor diagnosis of hip arthroplasty

infection is summarized

in Table 1.

using these criteria

has rangedfrom 50%to 100%,and the specificity, from critical because of the low prevalence of arthroplasty infection in general. In our own series, with the exception of head zone activity, neither the presence of periprosthetic activity

nor the intensity of such activity accurately identified DISCUSSION TABLE 2

Pain occurs in @—20% of hip arthroplasties, and is usually secondary to loosening with or without infection

LabeledNo. Diagnosis of Orthopedic Implant Infection wit

(27). While infection occurs in 1%-4% of primary hip arthroplasties (28—31), infection rates of 32%—38%fol

ofGroup/Reference Criteria implants SensitivitySpecificity

lowing revision arthroplasty have been reported (31, 32). Therefore, accurate preoperative diagnosis and treatment of an occult infection is essential for a suc

LeukocyteImagingAloneh

Reference 16'150100%90%Reference 14'140100%92%Palestroetalt192100%23%Referencel221686%100%Refe

cessful revision arthroplasty.

Criteria used to diagnose the infected orthopedic prosthesis, including clinical history, physical exami nation, erythrocyte sedimentation rate, peripheral leu kocyte count, aspiration and culture, and plain radiog raphy, are insensitive or nonspecific or both, and their usefulness

is limited

(21,22,33).

Technetium-99m bone scintigraphy, sensitive for de tecting osseous inflammation, suffers from a lack of specificity. It is especially unreliable in the assessment

of hip prostheses during the first 12 mo after implan tation (1,4,5,9,12,1 7,22,34,35).

Sequential

@mTc bone

and 67Ga scintigraphy has been used to improve the accuracy of diagnosing prosthetic infection. Gallium, however, localizes in septic and aseptic inflammation,

and in any region of increased osseous activity. The overall accuracy ofsequential bone/gallium imaging for

1952

2024095%90%Reference 212291 00%50%Reference

535073%95%Palestro 1 alt39265%61%Reference et 43888%93%Reference 183,

2259888%73%Reference 1161 650%100%Palestro et alt79287%94% Criteria:(1) Anyperiprostheticactivity;(2) Periprostheticactiv ity > surroundingboneactivity;(3) Periprostheticactivity> con tralateralside; (4) Periprostheticactivity> symphysispubis on anterior v@w or > sacroiliac joint on posterior view; (5) Penprostheticactivity> normalmarrowactivity;(6) Criteria not specified;(7) Headzoneactivity. . Studies

performed

t Current

series.

with

111In-Iabeled

granulocytes.

TheJournalof NuclearMedicine• Vol. 3i • No. i 2 • December1990

TABLE 3 Diagnosis of Orthopedic Implant Infection by Dual-Tracer

Imaging Group/

No.of

Reference

Tracers Criteria implants Sensitivity Specificity

Reference 19A12485%85%Reference 21A12988%95%Reference 0B23092%100%Reference 1 23B254100%93%Palestro et alB2501

92%, and a specificityof 100%,while Fink-Bennett et al. (23) reported a sensitivity of 100%, and a specificity

of 93%. Using this technique we obtained a sensitivity of 100% and a specificity of 97%, which are in close accordwith theresultsof thesetwo investigators. Corn bining the results of these two series with our data, the sensitivity,

specificity, and accuracy of the procedure

are 96%, 97%, and 97%, respectively, superior to la beled leukocyte imaging alone or in combination with routine bone scintigraphy.

00%97%

This superiority is, we believe, related to the distri A: labeledleukocyteand boneimaging;B: labeledleukocyte and sulfur colloid imaging;1: incongruentbone and Ieukocyte butions ofeach ofthese tracers. Technetium-99m-MDP accumulates in bone, while labeled leukocytes and sul images;2: incongruentsulfurcolloidandleukocyteimages. . Current

fur colloid accumulate in marrow. Conditions that af

series.

fect bone may or may not exert similar effects on marrow,

the infected arthroplasty (Figs. 3—4)(Table 2). We at tribute this to the variable patterns of periprosthetic labeled leukocyte activity encountered in both infected and uninfected arthroplasties. Although the normal distribution of marrow in adults up to 70 yr old gener

and vice versa.

and marrow, even in the reported that incongruent

Therefore

incongruent

bone

i.e., labeled leukocyte, images could result absence of infection. Oswald et al. (39) labeled leukocyte and bone images were 15% of the time, even in the absence of

ally includes the axial skeleton, humeral heads, femoral

infection. Both labeled leukocyte and sulfur colloid images reflect radiotracer accumulation in the reticuloendothe

trochanters,

hal cells of the marrow, although it is unclear whether

and frequently the femoral shafts, there is

considerable individual variation. The implantation of a prosthetic device may produce additional variations. Changes in the normal distribution of marrow through

marrow activity present on leukocyte images reflects labeled leukocytes alone or in combination with free

surgical

is otherwise

manipulation

have, in fact, been demonstrated

in animals (37,38). Improved accuracy for diagnosing osseous infection using labeled leukocyte plus 99mTc@boneimaging (13,

19,21), and labeled leukocyte plus 99mTc@sulfur colloid marrow

imaging

(10,23)

has been observed

(Table

3).

The reported sensitivity and specificity of leukocyte/ bone imaging have ranged from 70% to 88%, and 75% to 95%, respectively. Using combined

aging, Mulamba

labeled leukocyte/sulfur

et a!. (10) reported

colloid im

a sensitivity

of

I I ‘In.

Alterations

in

the

normal)

distribution

produce

of

similar,

marrow

(which

or congruent,

images on both of these studies (24). Infection, how ever, may actually exert opposite effects on leukocytes

and sulfur colloid. While stimulating leukocyte accu mutation infection of bone has been reported to de crease sulfur colloid accumulation (25). These inverse effects result in incongruent images, and permit dis

crimination of infection from unusual, but not abnor mat, periprosthetic marrow distribution. In contrast to previous reports (10,23), we performed sulfur colloid imaging after, rather than before, the

FIGURE3 A 60-yr-old female with a left total-hip arthroplastyimplanted6 yr prior to im aging. (A) Anterior 24-hr labeled leu

kocyte image demonstrates peripros

A%@..:'

B

thetic leukocyte activity in the trochan

tenc,shaft,and tip regions.Although slightly heterogeneous,this activity is approximately the same intensity as the surrounding marrow activity, and less intense then marrow activity on the contralateralside. (B) Anterior sul fur colloid image performed approxi mately 1 hr after (A) revealsnearlyab sent marrow activity in the trochanteric andshaft regions.The study was inter preted as incongruentleukocyte/sulfur

colloidimages,consistentwith infec tion. An infected arthroplasty was re

movedat surgery.

Total-Hip Arthroplasty • Palestro et al

1953

FIGURE4

B

A

An 81-yr-old female with a right total hip arthroplastyinserted 15 yr prior to

imaging.(A)Anterior24-hrlabeledIeu kocyte image demonstrates intense periprostheticactivity surroundingthe trochanteric,shaft, and tip zones.This

activityis moreintensethanthecorre spendingcontralateralregion. (B) Sul

fur colloidimagerevealsdistributionof radiotracer similar to that in (A). This study was interpreted as congruent leukocyte/sulfur colloid images, with

out evidenceof infection.A loose,but uninfectedprosthesiswasremoved.

labeled leukocyte study. The advantage of this is that those arthroplasties which demonstrate no peripros thetic labeled leukocyte activity (17% in our series) need not be studied with sulfur coltoid imaging.

One false-positive labeled leukocyte/sulfur colloid study occurred. While we have no definite explanation for this occurrence, it is possible that overlying soft tissue inflammation

was the cause.

REFERENCES 1. Mclnerney DP, Hyde ID. Technetium-99m-pyrophosphate scanning in the assessment ofthe painful hip prosthesis. C/in Radiol1978;29:513—517. 2. Williamson BRJ, McLaughlin RE, Wang GJ, Miller CW,

Teates CD, Bray ST. Radionuclide bone imagingas a means of differentiating loosening and infection in patients with a painful hipprosthesis. Radiology 1979;133:723—725. 3. Weiss PE, Mall JC, Hoffer PB, Murray WR, Rodrigo JJ,

Genant HK. 99mTcmethyleflediphosphonate bone imaging

In conclusion, periprosthetic activity on labeled leu kocyte images is extremely variable, both in pattern of distribution and intensity of uptake, frequently making interpretation of these images by themselves difficult

4.

(Fig. 5). Although recognition of a significant relation between the presence of activity in the head zone and

5.

arthroplasty

infection

improves

the accuracy

of the

labeled leukocyte study when interpreted alone, the addition of sulfur coltoid imaging improves discrimi nation of physiologic labeled teukocyte accumulation in marrow from accumulation due to infection, result ing in a superior (98% accuracy in our series) test for

the diagnosis of the infected hip arthroplasty. ACKNOWLEDGMENT The authors thank Ms. Elsa Ortiz for the preparation of this manuscript.

?

I HipArthroplasty Infection

6.

7.

in the evaluation of total hip prostheses. Radiology 1979; 133:727—729. Rosenthabl L, Lisbona R, Hernandez M, Hadjipavbou A. 99mTcpp and 67Gaimaging following insertion of orthopedic devices. Radiology 1979;133:717—721. Williams F, McCall 1W, Park WM, O'Connor BT, Morris V. Gablium-67 scanning in the painful total hip replacement. C/inRadiol1981;32:431—439. Rushton N, Coakley AJ, Tudor J, Wraight EP. The value of technetium and gallium scanning in assessing pain after total hip replacement. J Bone Joint Surg[BrJ 1982; 64-B:3l3—3l8. Utz JA, Galvin EG, Lull Ri. Natural history of technetium

99m-MDP bone scan in asymptomatic total hip prostheses.J Nuc/Med 1982;23:28—29. 8. Merkeb KD, Brown L, Fitzgerald RH, Jr. Sequential techne tium-99m-HMDP/gablium-67-citrate imaging for the evalua tion of infection in the painful prosthesis. J NucI Med 1986; 27:1413—14 17. 9. Aliabadi P, Tumeh SS, Weissman BN, McNeil BJ. Cemented total hip prosthesis: radiographic and scintigraphic evaluation. Radiology 1989;173:203—206. 10. Mulamba L'AH, Ferrant A, Leners N, deNayer P. Rombouts

JJ, Vincent A. Indium-b 11-leukocytescanning in the evabu ation of painful hip arthropbasty. Acta Orthop Scand 1983;

54:695—697. him LabeledLet@kocyte Imaging Penprosthetic Activity? No

@

Yes

SulfurCOiledImaging congwem Images No Infection

Incongwent

Images

I Infection

FIGURE5 Interpretative algorithmfor labeledleukocyteimagingin the diagnosisof the infectedhip arthroplasty.

1954

11. McKilbop JH, McKay I, Cuthbert OF, Fogelman I, Gray HW, Sturrock RD. Scintigraphic evaluation of the painful pros thetic joint: a comparison of galbium-67-citrate and indium 11b-labeled leukocyte imaging. C/in Radiol 1984; 35:239— 241. 12. Merkel KD, Brown ML, Dewanjee MK, Fitzgerald RH, Jr. Comparison of indium-babeled-leukocyte imaging with se quential technetium-gallium scanning in the diagnosis of bow grade muscuboskeletal sepsis. J Bone Joint Surg fAmJ 1985; 67-A:465—476. 13. Al-Sheikh W, Sfakianakis ON, Mnaymneh W, et al. Sub acute and chronic bone infections: diagnosis using In- 111, Ga-67, and Tc-99m-MDP bone scintigraphy and radiography. Radiology 1985;155:501—506. 14. Pnng DJ, Henderson RG, Rivett AG, Krausz T, Coombs

TheJournalof NuclearMedicine• Vol. 3i • No. i 2 • Decemberi 990

RRH, LavenderJP. Autologousgranubocytescanningof pain 15.

16.

I7.

18.

19.

20.

21.

22.

23.

ful prosthetic joints. J Bone Joint Surg fBrJ 1986; 68-B:647— 652. 27. Mountford PJ, Hall FM, Wells CP. Coakbey AJ. @mTc@MDP, 67Ga-citrate and ‘ ‘ ‘In-leukocytesfor detecting prosthetic hip 28. infection. Nuc/Med Comm 1986;7:113—120. Pring Di. Henderson RG, Keshavarzian A, et al. Indium granulocyte scanning in the painful prosthetic joint. AiR 1986;146:167—172. 29. Prchal CL, Kahen HL, Blend Mi, Barmada R. Detection of muscuboskeletal infection with the indium-l 11-leukocyte scan. Orthopedics 1987;10:1253—1 257. 30. Ouzounian Ti, Thompson L, Grogan TJ, Webber MM, Am stutz HC. Evaluation of musculoskeletabsepsis with indium I 1b-white blood cell imaging. C/in Orthop 1987; 221:304— 311. 31. Wukich DK, Abreu SH, Callaghan ii, et ab. Diagnosis of infection by preoperative scintigraphy with indium-labeled white blood cells. J Bone Joint Surg fAm] 1987; 69-A: 1353— 32. 1360. Gomez-Luzuriaga MA, Gaban V. ViblariM. Scintigraphy with Tc, Ga, and In in painful total hip prostheses. ml Orthop 33. 1988: 12:163—167. Johnson JA, Christie Mi, SandIer MP, Parks PF Jr. Homra 34. L, Kaye ii. Detection ofoccubt infection following total joint arthroplasty using sequential technetium-99m-HDP bone 35. scintigraphy and indium-b I b-WBC imaging. J Nuc/ Med 1988:29:1347—b 353. Magnuson iE, Brown ML, Hauser MF, Berquist TH, Fitzger 36. aid RH ir, Klee GG. In-i 1b-labeled leukocyte scintigraphy in suspected orthopedic prosthesis infection: comparison with other imaging modalities. Radiology 1988; 168:235—239. Fink-Bennett D, Stanisavbjevic S. Blake D, Weber K, Weir J, Mayne B. Improved accuracy for detecting an infected hip arthroplasty (HA): sequential technetium-99m-sulfur colboid 37. (TSC)/indium-l 1l(In-l 11) WBC imaging [Abstractj. J Nuc/ Med 1988: 29:P887.

24. Palestro Ci, Charalel i, Vablabhajosuba S.Greenberg M,Gold

AW. Indium-b 1I-labeled autologous leukocytes in man. J Nuc/Med 1977;18:1014—1021. Hams WH. Total joint replacement. N Eng/ J Med 1977; 297:650—651. Charnley J, Eftekar N. Postoperative infection in total pros thetic replacement arthroplasty of the hip joint: with special

reference to the bacterial content of the air of the operating room.BriSurg 1969;56:641—649. Andrews Hi, Arden OP. Hart GM, Owen iW. Deep infection after total hip replacement. J Bone Joint Surg [Br] 1981; 63B:53—57. Lidwell OM, Lowbury EJL, Whyte W, Blowers R, Stanley Si, Lowe D. Effect of ultra clean air in operating room on deep sepsis in the joint after total hip or knee replacement: a

randomized study. BrMedJ 1982;285:10—14. Maderazo EG, iudson S. Pastemak H. Late infections of total joint prostheses. A review and recommendations

for preven

tion. C/inOrthop1988;229:131—142. Hunter GA, Welsh RP, Cameron HU, Bailey WH. The results of revision of total hip arthropbasty. J Bone Joint Surg [Br] 1979;6i-B:4b9—42b. Weissman BN. The radiology of total joint replacement. OrthopCoinNorthAm 1983;14:171—191. Skeletal Scintigraphy. In: Gabasko CSB, ed. Orthopaedics. Edinburgh: Churchill Livingstone; 1984;b35—162. Utz JA, Lull Ri, Galvin EG. Asymptomatic total hip pros thesis: natural history determined using Tc-99m-MDP bone scans. Radiology 1986;161:509—5 12. Merkel KD, Fitzgerald RH Jr. Brown ML. Scintigraphic

examination of total hip arthroplasty: comparison of indium with technetium-gallium in the loose and infected canine arthropbasty.In: Welch RB, ed. The hip. Proceedings of the twelfth open scientific meeting of the Hip Society. Atlanta; 1984:163—192. Van Dyke D, Shkurkin C, Price D, Yano Y, Anger HO. Differences in distribution of erythropoeitic and reticuboen dothebial marrow in hematologic disease. Blood 1967; 30:364— 374.

smith Si. ln-WBC as a bone marrow imaging agent [Abstract].

38. Custer RP. An at/as ofihe blood & bone marrow.2nd edition.

J Nuel Med 1987;27:P574.

Philadelphia: WB Saunders; 1974:33—42. 39. Oswald SG, Van Nostrand D, Savory CG, Callaghan JJ. Three-phase bone scan and indium white blood cell scintig raphy following porous-coated hip arthropbasty: a prospective studyoftheprosthetic tip. JNuc/Med 1989;30:1321—1331.

25. Feigin DS, Strauss HW, James AE. Detection of osteomyebitis by bone marrow scanning [Abstract]. J NucI Med 1974; b5:P490. 26. Thakur ML, Lavender JP, Arnot RN, Silverstein Di, Segab

Editorial

DiagnosingProstheticJointInfection For the past 17 years, we have been studying the problem of how to best diagnose infection in a pros thetic joint by nuclear imaging. We

have studied prosthetic joints with various combinations

of technetium

ReceivedAug. 29, 1990; accepted Aug. 29, 1990. Forreprintscontact: NaomiAlazrakl,MD,do Director, Divisionof Nuclear Medicine,Emory Lk@iversity Hospital.1364 CliftonRoad, NE, At larna,GA30322.

bone scans, gallium, indium leuko

need to review our past history in

cytes, indium

and newer

imaging infected prosthetic joints,

infection imaging formulations in cluding indium-labeled gamma

including the hows and whys of the various radiopharmaceuticals used.

globulin and 99mTdHMPAO Pales tro et at. in this issue of JNM pre sents the case for the combination of II‘In-labeled leukocytes and

Between1973and1979,Reinget al. and Bauer et al. (1,2) reported results of trials of bone scans and

chloride,

99mTc@olloidbone marrow imag ing. To better understand why this has been so problematic a clinical diagnostic

Prosthetic Joint InfectionImaging • Alazraki

imaging

challenge,

we

gallium scans in patients suspected

of having infection or loosening of prostheses.Conclusions often stated that a normal bone scan excluded the possibility of need for surgical

1955