presence of multiple new abnormalities without evidence for a benign etiology that can be consideredcharacteristic of metastatic involvement (11, 12). Whilethe prevalence of metastases in solitary bone scan abformalities has been examined by several groups (13-18), a rigorous investigation ofthe association betweenthe number and sites of scan abnormalities and thelikelihood of metastatic disease has not been described.

We retrospectivelyidentifiedall new abnormalitieson the serial skeletal scintigrams of a group of breastcancer patients followed on a regular basis after en

trance into an adjuvant chemotherapy protocol (19,20).From available scintigraphic, radiologic, and clinicalfollow-up, the presence or absence ofmetastatic diseaseat each new abnormal site was ascertained. The prevalence of bone metastases as a function of the numberof new scan abnormalities was then determined, andthe anatomic distributions and correlative radiologicfindings for benign and malignant lesions were cornpared.

MATERIALSANDMETHODSPatient Population

Six hundred and sixty-one breast carcinoma patients, allclinicalStageI, II, or III at diagnosis(21), wereentered intoan adjuvantchemotherapyprotocol at our institution between1974 and 1985 (19,20). Requirements for protocol entryincludeda bone scan with no evidenceof metastaticdiseaseprior to initiation ofchemotherapy. Recommended follow-upat conclusion of chemotherapy included a bone scan every 6mo for the first three years and annually thereafter. Twohundred fifty-eightprotocolparticipantsidentifiedon nuclearmedicine records as having had at least two bone scans at ourinstitution form the initial patient population examined forthis report.

ImageReviewA total of 1,741 bone scans performedon the 258 patients

were available for review. All scans were performed 2—4hrfollowingi.v. injection of 555—925MBq (15—25mCi) of

Review of 1,441 bone scans performedon 242 breastcancer patients without known skeletal metastases iden

tified 239 scans with new abnormalities. Findings on 54 ofthese239 scans(23%) representedbonemetastases.Theproportionofscansreflectingmetastases,groupedbythenumberof newabnormalities,was:(1)20/182(11%);(2)9/26 (35%); (3) 4/9 (45%); (4) 1/2 (50%); 5 —20/20(100%). When metastatic disease presented as a bonescanwith1-4newabnormalities,thespinewasthemostcommonsiteof involvement(18 of34(53%)),followedbytheskull(5/34;15%),extremitiesandsternum(each4/34;12%). Rib lesions were the most common new findings onscanswith<5 newabnormalities(seenon76of 219 scans(35%)) but only infrequently represented metastases (n =2). Consideringas indicativeof malignancyonly, thosebonescanswhichdemonstratedeither(a) 5 new abnormalities,(b) initialradiographiccorrelationsuggestiveofmetastases,or (c)thoracicspinelesionswithnormalcot'-relativeradiographs,the presenceof skeletalmetastaticdiseasecouldbepredictedwithasensitivityof0.80andaspecificityof0.94.

J NuclMed 1990;31:387-392

one scintigraphy is a sensitive method for detectingthe skeletal metastases of breast carcinoma, often providing the first evidence ofdistant malignant spread (1—4). However, a limitation of the technique is its nonspecificity (5—10), particularly when used in routinefollow-up ofpatients without known metastaticdisease.Any individual bone scan abnormality can be due toeither a benign or malignant process, and it is only the

ReceivedMay8, 1989;revisionaccepted Nov.1, 1989.For reprintscontact: WiUtamD. Kaplan,MD,Divisionof NuclearMedi

one, Dana-FarberCancer Institute, 44 Binney St., Boston, MA 02115.* Current Address: Nuclear Medicine Section, VA MediCal Center, Se

attle, WA.t Current Address: Department of RadiOlOgy, Roswell Park MemOrial

Institute,BUffalO,NY.

387BoneScanAbnormalitiesand SkeletalMetatases•Jacobsonet al

Association Between Number and Sites of NewBone Scan Abnormalities and Presence ofSkeletAlMetastases in Patients With BreastCancerArnold F. Jacobson,* Paul C. Stomper,t Maxine S. Jochelson, Diane M. Ascoli, I. Craig Henderson, andWilliam D. Kaplan

Division ofNuclear Medicine, Department ofRadiology, and Department ofMedicine, Dana-Farber Cancer Institute,Harvard Medical School, Boston, Massachusetts

by on April 11, 2018. For personal use only. jnm.snmjournals.org Downloaded from

Review scan images and imagesfrommost recentpreviousscan

technetium-99m- (99mTc)labeled diphosphonate compounds.Threebaselinestudiesperformedin 1974employeda rectilinear scanner.Gamma camera studiesemployedlargefield-ofview cameras with low-energy, high-resolution collimators.From 1974 to 1982, scans were acquired using a moving tableformat (22 cm/mm) to produce whole-body images in theanterior and posterior projections, with selected stationaryviews obtained of abnormal sites. Studies performed after1982 utilized overlapping stationary views exclusively. In virtuallyall cases,the entire skeletalsystemwasimaged.

A flow chart summarizing the imaging and clinical reviewprocessesis presentedin Figure 1. Briefly,a nuclearmedicinephysician (AFJ) read the written reports for each patient'sscintigraphicstudies,and guided by these reports, examinedthe following scans:

1. The first scan for which the report mentioned a newabnormality.

2. The study which immediately preceded the one in whichthe new abnormality was described, to determine if the“new―abnormality could be retrospectively identified.

3. Allsubsequentscansforaminimumof18mo.

For scans with four or fewer new lesions, the reviewer recordedthe anatomic location ofeach new abnormality and the lesionintensity relative to adjacent bone, scored as 1+ (minimal tomildly increased), 2+ (moderately increased), or 3+ (markedlyincreased).For scanswith fiveor more new lesions,only thetotal number of new findingswas recorded;detailed tabulations of these lesions were performed only if suspected malignancy was not confirmed on subsequent data review. Casesin which the reviewer disagreed with the written report regarding the presence of an abnormality, its anatomic location, orits relative intensity, were submitted to a second nuclearmedicine physician (WDK) for examination, and a final interpretation was achieved by consensus.

IBonescanreportreadbyreviewer

I Anyabnormalitiesmentioned?

IRecordlocation,intensity,and date eachnew abnormalitycould firstbe identified

Review all available correlativeradiologicalstudiesforeach scan lesion,and recordfindingsattimescanwasfirstabnormal

Review all clinicaldata, and recorddate(s) andsite(s)offirstlocaland/ordistantrecurrenceofbreast cancer or second malignancy

The reports of all radiologic examinations (plain films,fluoroscopy,trispiral, and computed tomography (CT)) ofareas abnormal on bone scan were reviewed. Initial radiologiccorrelation was categorized as either: 1) negative (RN); 2)benign (RB); 3) suggestive of metastasis (RS); 4) consistentwith metastasis (RM); or 5) no correlative study performedwithin 30 days of the bone scan (RO). Radiographic studiesoriginally reported as showing a finding of uncertain significance were blindly reread by a radiologist (PCS or MSJ) andclassifiedaseitherRN, RB,RS,or RM.Radiographsoriginallyreported as RS and RN or RB films ofscan lesions eventuallyproven to be bone metastases, were also reexamined to confirm the initial interpretations.

RecurrenceStatusClinical records through June 30, 1987 were reviewed to

identify those patients who:

1. Developed local or distant nonosseous recurrence of

breast cancer.2. Had a documented second malignancy.3. Developednewmalignantinvolvementwhoseetiology

(new primary versus breast or other metastasis) wasuncertain.

Bone was considered a site of initial distant relapse ifmetastaseswere either detected there first or were identifiedwithin 30 days following confirmation of a nonosseous site ofrecurrence. An individual scan lesion was established as abone metastasis by either:

1. Radiographic or tomographic images showing a typicalblastic or lytic lesion and changes on follow-upradiographicand/or scintigraphicstudiesconsistentwith progression or healing in response to therapy.

2. Progression on a follow-up bone scan, with an increasein lesion intensity and appearance of at least two additionalnewlesionswithoutevidencefora benignetiology.

3. Positivetissuediagnosisatautopsy.

The earliest scintigraphic abnormality, which on follow-upproved to be a metastasis, was recorded as the initial site ofbone involvement, even if verification of this finding did notoccur until after other nonosseous metastatic sites had beendocumented. A scintigraphic abnormality which remainedstable or decreased in intensity on at least one follow-up scanwas considered nonmalignant if the patient had no objectiveevidence of recurrence in bone during the 18 mo after theoriginalscandate, or at autopsy.

DataTabulationFollowing scan and clinical review, 16 patients were ex

cluded from analysis. These included 12 patients with recurrent breast cancer and evidence of metastatic disease prior toor on their first scan at our institution, and 4 who did nothave two scans prior to developing either a second malignancyor new neoplasm of uncertain histologic type.

For the remaining 242 patients, data were tabulated fromall scans of those patients without distant recurrence, andfrom scans up to the time of first distant recurrence (osseousor nonosseous) ofbreast cancer or a second malignancy in theother patients. The following findings were excluded:

Proceedwithreviewofnextscan report

FIGURE1Flowchartsummarizingthe reviewprocessusedto identifyandcharacterizenewbonescintigraphicabnormalities.

388 The Journalof NuclearMedicine•Vol. 31 •No. 4 •April1990

by on April 11, 2018. For personal use only. jnm.snmjournals.org Downloaded from

ClinicalstatusNumberof

patientsDistant

recurrenceBoneinvolvementatrecurrence55No

boneinvolvementatrecurrence27Localrecurrenceonly12Second

malignancywithoutmetastases15Newneoplasticinvolvementofuncertainhis8tologic

typeNorecurrence125Total242

5 6 7 8 9 10 11 12

Number of Scans13 14

1. All lesions on the earliest available scan for each patient.2. Two or more adjacent, aligned rib lesions in a pattern

indicative of trauma.3. Polyarticularincreaseduptakein thejointsofthecx

tremities.4. Diffuse or focal increased uptake in the appendicular

skeleton at sites of documented trauma.5. Increased uptake in the maxilla or mandible indicative

of dental pathology.6. Abnormalitieswithout sufficientfollow-upto establish

their etiologies.

Each scan abnormality was classified in terms of anatomiclocation, initial scintigraphic intensity, and results of correlative radiographs. Nine anatomic regions were defined: skull,cervicalspine,thoracicspine,lumbarspine,sternum,shouldergirdle (scapula and clavicle), ribs, pelvis, and extremities. Chisquare statistics were used for comparisons between differentgroupsofscan findings.Sensitivityand specificityfordetectionof metastases using selected criteria were calculated.

RESULTS

A total of 1,441 scans met the inclusion criteria foranalysis, 1,032 from the 152 patients without distantrecurrence (Table 1) (range 2 to 18 scans/patient, mean6.8) (Fig. 2A), and 409, including the first scans withevidence of metastatic breast cancer, in the other 90patients (range 2 to 14 scans/patient, mean 4.5) (Fig.2B). Two hundred thirty-nine of the 1,199 follow-upscans (20%) had new abnormalities, involving 117 patients (48% of the study population).

Fifty-five patients had bone as a site of initial distantrecurrence (Table 1), with bone metastases first identifled on a follow-up scan in 54 patients and on radiographs at a time no scan was performed in one patient.

Metastatic disease was confirmed by: positive radiographs with subsequent radiographic and/or scinti

graphic progression (n = 50); scintigraphic progression(n = 4); autopsy (n = 1). The number of new abnormalities on the 54 initial positive scans ranged from 1to 24, with the prevalence ofmetastases increasing from11% (20/182) for solitary new abnormalities to 100%(20/20) for scans with five or more new lesions (Table2). The interscan interval during which bone metastasesfirst appeared was the same regardless ofthe number ofnew lesions, with means of 10.1 (±4.5) (1 s.d.) , 10.1(±6. 1), and 10.6 (±6.4) mo for scans with 1, 2, and

:3 new abnormalities, respectively.The anatomicdistributionfor the 182 new solitary

bone scan abnormalities is shown in Table 3. Benignsolitary lesions were most common in the ribs (n = 58),while solitary metastases were most frequent in thethoracic spine (n = 8). There was no statistical difference between the prevalence of metastases for lesionsof different scintigraphic intensity (1-i-: 10/1 11 (9%);2+: 9/66 ( 14%); 3+: 1/5 (20%) (p = ns)).

Initial radiologic correlation was available for 162 ofthe 182 solitary lesions (Table 3). The majority of bothbenign and malignant lesions were RN, 59% and 53%,

TABLE IFollow-Upof 242 BreastCancerPatients

20

15

20U)Cw

U)a.

0

a)

Ez

U)Ca)

U)a.

0

a).aEz

15

10

5

0

10'

5@

A BFIGURE2Number of serial bone scans reviewed in 242 breast cancer patients. (A) 152 patients without distant recurrence and (B) 90patientswith either distant recurrenceof breast cancer (n = 82) or new neoplasmof uncertainhistologictype (n = 8).

2342 3 4 5 6 7 8 9 101112131415161718

Number of Scans

389Bone Scan Abnormalities and Skeletal Metatases •Jacobson et al

by on April 11, 2018. For personal use only. jnm.snmjournals.org Downloaded from

TABLE2Occurrenceof Metastasesas a Functionof NumberofNewAbnormalitieson BoneScintigraphyNumberof newNumberbonescan Numberofwithabnormalities

occurrencesmetastases1

182 20(11%)226 9(35%)3

9 4(45%)42 1(50%)544(100%)622(100%)>6

14 14(100%)Total

239 54(23%).

Percentage with metastases in parentheses.

many bone scan abnormalities has caused the utility ofthe technique to be questioned in the routine follow-upof clinically disease-free patients (6, 7,10,22—24).Unexplained new scan abnormalities, namely those withnormal correlative radiographs, represent a diagnosticdilemma for the clinician, who must decide how aggressively to pursue the possibility of new distant relapse. Although some early bone metastases, particularly in the spine, can be detected by CT before theyare evident on plain radiographs (25), bone biopsy maybe the only method capable of establishing the etiologyof a scan finding in other cases (26—29).

In the present study, a bone scan with five or morenew abnormalities always reflected the presence of metastatic disease, but this pattern was seen in only 20 of54 patients (37%) with scintigraphicevidence of initialdistant recurrence in bone. Twenty other patients (37%)had only a solitary new abnormality as the first indication of recurrence in bone, a scan finding eight timesmore likely to be benign than malignant (162/20 =8. 1). In contrast to solitary scan findings, there was abetter than one in three likelihood that a scan with 2—4 new lesionsreflectedbone metastases.The low likelihood of malignancy for solitary new abnormalitiesallowed attainment of a high specificity (0.94) simplyby considering as malignant only scans with 5 newabnormalities, lesions with positive (RM or RS) radiographs, and RN lesions in the thoracic spine, the latterincluded because of the increased frequency of suchfindings proving to be metastases (Table 4). While thesensitivity of 0.80 achieved using these criteria wouldbe unacceptably low, this example does demonstratethat the specificity of bone scintigraphy can be improved through careful examination of patterns of scanfindings.

Only new abnormalities which (a) developed in theinterval between two consecutive bone scans, and (b)did not represent obvious trauma or degenerative disease were included in this review; this allowed us tofocus on findings of the greatest clinical relevance. Thepresence or absence of malignancy at a scan positivesite wasjudged based upon imaging and clinical followup of 18 mo or longer (except in instances of patientdeath). Although biopsy or autopsy confirmation oflesion etiology was rarely available, almost all patientswith bone metastases showed obvious progression towidely disseminated disease. Most nonmalignant scanabnormalities either resolved or had eventual radiologicdemonstration of a benign cause.

Considering scans with less than five new metastatic lesions to reflect early distant recurrence, themost common site of involvement for this stage ofdisease was the spine. This is consistent with the cornmon involvement of the spine in breast cancer patientswith widely disseminated bone metastases (11,30—32).Although rib abnormalities were the most common

respectively. One metastasis (in the thoracic spine) hadinitial radiographs interpreted as RB, while one riblesion reported as RM and one extremity and three ribabnormalities interpreted as RS proved to be benign onfollow-up.

The sites of the 24 confirmed metastases on bonescans with 2—4new abnormalities were: thoracic spine,7; pelvis,4; extremities,4; lumbarspine,3; skull,3;sternum, 2; ribs, 1. Radiographic results for these metastaseswere: RN: 13(54%); RM: 5 (21%); RS: 4(17%);RB: 1 (4%);RO:1 (4%). Of the 10 other abnormalitieson these 14 scans, 2 were benign, and for the other 8,follow-up was insufficient to establish an etiology. Overall, only 6 of the 14 scans (43%) had at least one RMor RS lesion.

Ofthe 34 bone scans with metastatic disease presenting as <5 new abnormalities, 53% showed malignantinvolvement of the spine, 38% in the thoracic spine(Table 4). There were no RS or RM findings for anylesions on 21 of these scans (62%).

Sensitivity and specificity for identification of metastases on bone scans with new abnormalities were 0.37(20/54) and 1.00 ( I85/ 185), respectively, if only scanswith five or more lesions were considered to reflect thepresence of metastases. Inclusion of initial radiologiccorrelation (thus excluding the 20 solitary new abnormalities (three metastases) for which initial radiographswere not obtained), with RS lesions considered as metastases, resulted in sensitivity of 0.65 (33/5 1) andspecificity of 0.97 (163/168). As RN lesions in thethoracic spine proved to be metastases on 8 of 13 scans(62%) (Tables 3 and 4), categorizing all such lesions asmetastases resulted in sensitivity of 0.80 (4 1/5 1) andspecificity ofO.94 (158/168).

DISCUSSION

Bone scintigraphy is the most sensitive method foridentifying skeletal metastates in breast carcinoma (1—4,12). However, uncertainty concerning the etiology of

390 The Journalof NuclearMedicine•Vol. 31 •No. 4 •April1990

by on April 11, 2018. For personal use only. jnm.snmjournals.org Downloaded from

ScanfindingsInitialcorrelativeradiographtSite

Intensity No.RN RBRSRM RO

TABLE4Sitesof MetastaticInvolvementin 34 PatientswithInitialScanPresentationof LessThanFiveNewAbnormalitiesNumber

of scans NumberofscansAnatomicsite withmetastase&t withonlyRN lesions

. Percentage of scans in parentheses.

t Column total exceeds 34 because of metastates in more than

one anatomicsite on four scans:thoracic+ lumbarspine (2);thoracicspine+ skull(1);lumbarspine+ skull(1).

* RN: initial correlative radiograph negative.

TABLE3AnatomicDistributionandInitialRadiographicCorrelationfor 182 SolitaryNewBoneScanAbnormalities

Skull1 +2+8

(1)6 (1)4 5(1)3 00 00 01

(1)1Sternum1+

2+3(1)

2(1)2 10 00 01(1)

1(1)00Ribs1+

2+3+31

25 (1)322

16(1)27

611

201

000

10Shoulder

girdle1+ 2+11 75 63 00 00 031Cervical

spine1+ 2+6 5 (1)1 04 40 00 1 (1)10Thoracic

spine1+

2+3+1

8 (4)4 (3)1 (1)7

(3)1 (1)07

1 (1)00

1 (1)1(1)0

004

(1)0

0Lumbar

spine1+

2+12 (2)6 (1)5

(1)1 (1)6 41

(1)00 001Pelvis

Extremities1+

2+3+1+2+10

(1)51

12 (1)7(1)5

307 (1)13

20320

00010

0001(1)2(1)

012

2Totals1

+2+3+1

11 (10)66 (9)

5 (1)182 (20)58

(5)34(4)

@___

94 (9)36

19(1)

56(1)2

(1)4(1)i_w7 (3)2

(1)3(3)05 (4)13

(3)61

20(3)

. Number of metastases in parentheses.

t Key: RN = normal; RB = benign; RS = suggestive of metastasis; RM = consistent with metastasis; and RO = not done.

new findings on all scans with 1-4 new lesions (76/2 19;35%), they represented metastatic disease on only twoscans.

Bone scan abnormalities are commonly evaluated inconjunction with correlative radiographs (4,10,25,33,34). Although such correlation can be helpful evenwhen the pattern ofscan abnormalities strongly suggestseither a benign or malignant explanation, its majorimpact is on cases where the scan shows a finding ofuncertain clinical relevance such as a solitary new lesion. Our finding that 58% of such abnormalities wereRN demonstrates that correlative x-rays alone cannotadequately deal with the problem of unexplained newscintigraphic lesions, and underscores the value ofknowing the relative likelihood of a lesion being ametastasis based upon its skeletal location.

The results of the present review indicate that thenumber, sites, and radiologic appearance of new bone

Thoracicspine1 3(38%)8Lumbarspine6 (18%)3Skull5

(15%)2Sternum4(12%)2Extremities4(12%)2Pelvis3

(9%)1Ribs2(6%)1Cervical

spine1 (3%)0

BoneScanAbnormalitiesandSkeletalMetatases•Jacobsonet al 391

by on April 11, 2018. For personal use only. jnm.snmjournals.org Downloaded from

scan abnormalities in patients with breast cancer canbe helpful in predicting the likelihood that skeletalmetastases are present. Although staging bone scanshave a low yield for identifying unsuspected metastaticdisease, particularly for clinical Stage I and II patients(10,22,23,35,36), they do provide baseline information

which allows the determination of whether abnormalities seen on subsequent scans are new, thereby aidingin the prediction oflesion etiology. While bone scintigraphy in conjunction with selected radiographs remainsthe most efficacious means for identifying early skeletalmetastases, isolated new RN lesions in sites such as thethoracic and lumbar spine warrant further investigationby means such as CT (9,25), MRI (37,38), or bonebiopsy (27,29).

REFERENCES

1. Roberts JO, Gravelle IH, Baum M, et al. Evaluation ofradiography and isotope scintigraphy for detecting skeletalmetastases in breast cancer. Lancet 1975; 1:237—239.

2. Osmond III JD, Pendergrass HP, Potsaid MS. Accuracy of99mTc@iphosphonatebone scans and roentgenograms in thedetection of prostate, breast, and lung carcinoma metastases.AJR 1975; 124:972—979.

3. Citrin DL, Bessent RO, Grieg WR, et al. The application ofthe @mTcphosphate bone scan to the study of breast cancer.BrfSurg 1975;62:201—204.

4. Gold RH, Bassett LW. Radionuclide evaluation of skeletalmetastases: practical considerations. Skeletal Radiol 1986;15:1—9.

5. Perez DJ, Milan J, Ford HT, et al. Detection of breastcarcinoma metastases in bone: relative merits of x-rays andskeletal scintigraphy. Lancet 1983; 2:613—616.

6. RobertsMM, HaywardJL. Bone scanningand earlybreastcancer@five-year follow-up [Letter]. Lancet 1983; 1:997—998.

7. Burkett FE, Scanlon EF, Garces RM, et al. The value of bonescans in the management of patients with carcinoma of thebreast. Surg GynecObstet 1979; 149:523—525.

8. Rossing N, Munck 0, Nielsen SP, Ct al. What do early bonescans tell about breast cancer patients? Eur J Cancer ClinOncol 1982; 18:629—636.

9. Feig SA. Imaging techniques and guidelines for evaluationand follow-upof breast cancer patients. CRC Cr11Rev DiagImaging1987;27:1—16.

10. Kunkler IH, Merrick MV, Rodger A. Bone scintigraphy inbreast cancer a nine-year follow-up. C/in Radiol 1985;36:279—282.

11. Wilson MA, Calhoun FW. The distribution ofskeletal metastases in breast and pulmonary cancer concise communication. iNuc/Med 1981; 22:594—597.

12. Arnstein NB, Harbert JC, Byrne PJ. Efficacy of bone andliver scanning in breast cancer patients treated with adjuvantchemotherapy. Cancer 1984;54:2243—2247.

13. Corcoran RJ, Thrall JH, Kyle RW, Ctal. Solitary abnormalities in bone scans ofpatients with extraosseous malignancies.Radiology1976;121:663—667.

14. Shirazi PH, Rayudu GVS, Fordham EW. Review of solitaryF18 bone scan lesions.Radiology 1974; 112:369—372.

15. Rappaport AH, Hoffer PB, Genant HK. Unifocal bone findings by scintigraphy—clinical significance in patients withknown primary cancer. WestJ Med 1976; 129:188—192.

16. Tumeh SS, Beadle 0, Kaplan WD. Clinical significanceof

solitary rib lesions in patients with extraskeletal malignancy.JNuclMed 1985;26:1140—1143.

17. Thivolle P, Mathieu L, Mathieu P, et al. Significance ofsolitary spine abnormalities on technetium-99m bone imaging.ClinNuclMed 1988;13:527—528.

18. Kwai AH, Stomper PC, Kaplan WD. Clinical significance ofisolated scintigraphic sternal lesions in patients with breastcancer. JNuclMed 1988;29:324—328.

19. Henderson IC, Gelman RS, Harris JR. et al. Duration oftherapy in adjuvant chemotherapy trials. NC! Monogr 1986;1:95—98.

20. Griem KL, Henderson IC, Gelman R, et al. The 5-year resultsof a randomized trial of adjuvant radiation therapy afterchemotherapy in breast cancer patients treated with mastectomy.JClin Oncol 1987; 5:1546—1555.

2 1. American Joint Committee on Cancer. Breast. In: Manualfor staging ofcancer. Philadelphia:J.B. Lippincott Co.; 1983:129—130.

22. ButzelaarRMJM, Van Dongen JA, De GraafPW, et al. Bonescintigraphy in patients with operable breast cancer stages Iand II: finalconclusionafter five-yearfollow-up.Eur J CancerC/inOncol1984;20:877—880.

23. Wickerham L, Fisher B, Cronin W, et al. The efficacy of bonescanning in the follow-up of patients with operable breastcancer. Breast Cancer Res Treat 1984;4:303—307.

24. Khandekar JD. Role of routine bone scans in operable breastcancer: an opposing view. Cancer Treat Reports 1979;63:1241—1244.

25. Harbin WP. Metastatic disease and the nonspecific bone scan:value of spinal computed tomography. Radiology 1982;145:105—107.

26. Mink JH, Weitz I, Kagan R, et al. Bone scan-positive andradiograph- and CT-negative vertebral lesion in a womanwith locally advanced breast cancer. AJR 1987; 148:341—343.

27. Shih W-J, DeLand HI, Domstad PA, et al. Open rib biopsyguided by radionuclide technique. Ann Thorac Surg 1984;38:59—62.

28. Mink J. Percutaneous bone biopsy in the patient with knownor suspected osseous metastases. Radiology 1986; 161:191—194.

29. Little AG, DeMeesterTR, Kirchner PT, et al. Guided biopsiesofabnormalities on nuclear bone scans.J Thorac CardiovascSurg 1983; 85:396—403.

30. Lee Y-T M. Patterns of metastasis and natural course ofbreast carcinoma. Cancer Met Rev 1985; 4:153—172.

31. Galasko CSB. The anatomy and pathways of skeletal metastases. In: Bone metastasis. Boston: G. K. Hall Medical Publishers; 1981: 49—63.

32. Del Castillo R, Audisio T, Kuschnir E, et al. Diagnostic valueofthe different methods for evaluation of bone metastasis inbreast carcinoma. EurfGynaec Oncol 1987; 6:128—133.

33. Derimanov SO. ‘For'or ‘against'bone scintigraphyof patientswith breast cancer. NucIMed Commun 1987;8:79—86.

34. KunklerIH, MerrickMV. The valueof non-stagingskeletalscintigraphy in breast cancer. C/in Radiol 1986; 37:56 1—562.

35. Inoue Y, Nishi T, Hirose T, et al. Bone scanning in patientswith breast carcinoma. Should it be a routine examination inevery stagegroup? Fortschr Rontgenstr 1985; 143:676—678.

36. Wilson OS, Rich MA, Brennan MJ. Evaluation of bone scanin preoperative clinical staging of breast cancer. Arch Surg1980;115:415—418.

37. Krauss B, Youssry T, Nagele M. Fast sequences in the detection of malignant bone marrow infiltration of the spine [Abstract]. Radiology 1987; 165:274.

38. Smoker WRK, Godersky JC, Knutzon RK, et al. The role ofMR imaging in evaluating metastatic spinal disease. AiR1987;149:1241—1248.

392 The Journalof Nudear Medicine•Vol. 31 •No. 4 •April1990

by on April 11, 2018. For personal use only. jnm.snmjournals.org Downloaded from

1990;31:387-392.J Nucl Med.   Arnold F. Jacobson, Paul C. Stomper, Maxine S. Jochelson, Diane M. Ascoli, I. Craig Henderson and William D. Kaplan  of Skeletal Metastases in Patients With Breast CancerAssociation Between Number and Sites of New Bone Scan Abnormalities and Presence

http://jnm.snmjournals.org/content/31/4/387This article and updated information are available at:

  http://jnm.snmjournals.org/site/subscriptions/online.xhtml

Information about subscriptions to JNM can be found at:  

http://jnm.snmjournals.org/site/misc/permission.xhtmlInformation about reproducing figures, tables, or other portions of this article can be found online at:

(Print ISSN: 0161-5505, Online ISSN: 2159-662X)1850 Samuel Morse Drive, Reston, VA 20190.SNMMI | Society of Nuclear Medicine and Molecular Imaging

is published monthly.The Journal of Nuclear Medicine

© Copyright 1990 SNMMI; all rights reserved.

by on April 11, 2018. For personal use only. jnm.snmjournals.org Downloaded from