Download Radionuclide Perfusion Images of Tumor Microembolism: Three Case Reports and more Study notes Radiology in PDF only on Docsity!
TumorMicroembolism:PulmonaryPerfusionPattern
RObertCrane, Thomas G. Rudd,and David Dail
The Mason Clinic, and University of Washington, Seattle, Washington
ThreepatientshadradionuclldelungperfusionImagesthat we believeto betyp
IcaI of tumor mlcroembolism.AutopsyconfirmatIonwas available In two of these
cases.A briefreviewof pulmonarytumormlcroembollsmIsIncluded.
J Nuci Med 25: 877—880,
Wehaverecentlystudiedthreepatientswithknownmetastat
ic carcinoma who presented with dyspnea and clinical suspicion of pulmonary thromboembolism.The radionuclide(RN) perfusion images showed multiple subsegmental peripheral defects that we
believeto betypicaloftumormicroemboli.Wereportthesecases
and briefly review previously published reports.
CASE A
Four yearsearliera 44-yr-oldfemalehad a modifiedradical
mastectomyfor ductal carcinomawith positiveaxillarynodes.
Chemotherapywasgiven.Oneyearlater,metastaseswerefound
in the liver,spine,and skull.Bilateraloophorectomywasperformed and radiation therapy given. There was no history ofheart failure, obstructive lung disease, thromboembolic disease, or leg tender ness. Seven weeks before admission she began to have dyspnea, particularly with exertion, that became significantly worse two days prior to admission, without cough, sputum, chest pain, or hemoptysis. There were a few inspiratory rales bilaterally. Arterial blood yielded pO251 mm Hg, pH 7.52, and @,CO222. Chest ra diograph showed slight prominence of central pulmonary vasul
ature (Fig. 1, upper left). RN lung perfusionimagesshowed
multiple subsegmental peripheral defects (Fig. 1): ventilation images were normal. A gated RN ventricular-function study showed marked right-ventricular enlargement and poor contrac tility; the LV was normal. She was digitalized but suffered a car diac arrest 4 days after admission.
At autopsytherewerenogrosspulmonaryemboli,discreteareas
of tumor spread, or pneumonia. Microscopy revealedtumor emboli in 80% of the muscular pulmonary arteries, with acute fi bromyxomatous intimal proliferation significantly narrowing the
vesselsand at timesobliteratingthem (Fig. 2). Mildright-yen
tricular dilatation supported the clinical impression of acute cor pulmonale. No pulmonary lymphangitic involvement was found.
Received May 18, 1983; revision accepted Apr. 4, 1984. For reprints contact: Thomas G. Rudd, MD, HMC Radiology ZA-65, 325 Ninth Ave., Seattle, WA 98104.
CASE B
Five years previously a 49-year-old female developed lower extremity paralysis due to epidural tumor at 12,13, metastatic from a left breast intraductal carcinoma. She was stable until the recent development of weakness and dyspnea, which had increased over 3 days.
On admissionshewasslightlycyanotic,in respiratorydistress,
and demonstrated bibasilar rales. Arterial blood yielded pO2 47,
pH 7.36,and,,C0219.A chestradiographshowedelevationofthe
right hemidiaphragmand minimalatelectasisat rightbase(Fig.
3, upper left). Radionuclide lung perfusion showed multiple small peripheral subsegmental defects (Fig. 3). Right-heart catheter ization was performed, revealing elevated pressures in right yen tricular and main pulmonary artery (RV 50/15 mm Hg, PA 50/25, wedge10).Duringtheprocedure,cardiacarrestoccurred and resuscitative efforts failed.
At autopsy,nogrossemboliwerefoundinthemainpulmonary
arteries or branch arteries, but there were numerous hemorrhagic, wedge-shaped areas in the periphery ofthe lungs,consistent with subacute to acute pulmonary infarction. Microscopy revealed extensive pulmonary arterial obstruction in vessels ofall sizes, fully 80% involvedwith unorganized, masses of tumor cells. In other
vesselsthere was fibrinousorganizationabout the tumor cells.
Lymphatic involvement with tumor was <5%. The right ventricle and main pulmonary artery showed no signs of hypertrophy or enlargement, consistent with an acute process.
CASE C
A 40-yr-oldfemalehada modifiedradicalmastectomyforin
filtrating ductal carcinoma, with one of 27 axillary nodes involved. Chemotherapy was given, but less than a year later metastases
occurredin the choroidof the lefteye,a cervicalnode,and right
hilum, and subsequently to the brain and bone marrow. She presented with a 3-wk history of dyspnea at rest without fever,chills,chest pain, or peripheraledema. Bloodgaseswere pO 54, pH 7.46, @,CO234. Chest radiograph showedan enlarged right hilum (Fig. 4). Radionuclide lung perfusion showed multiple bi lateral peripheral defects (Fig. 4). Ventilation images were normal.
Volume 25, Number 8 877
CRANE, RUDD, AND DAIL
L
FIG. 1. P@ A. Upperleft PA chest radiograph shows slight central pulmonary-arteryfull ness. Upper right Posterior perfusion image (reversedto match radlo@'aph).Lower left, right:RPO and LPO perfusion images sw muftlple subseg mental defects.
p A A
RPO (^) L P
Bronchoscopywas aborted withoutbiopsybut the right endo
bronchial architecture looked normal. Her condition could have been due to thrombus or tumor emboli. She received chemotherapy for a month, but had mild respira tory distress. On readmissionfor chemotherapy 6 weeks later, chest radiograph showed multiple new circular densities throughout both lungs, superimposed upon vague diffuse parenchymal densities. The patient died at home shortly thereafter. No autopsy was performed.
DISCUSSION
Patientspresentingwithdyspneaandnormalchestradiographs
are frequently evaluated with RN lungperfusion imaging for po tential pulmonary thromboembolism. Unlike the picture in thromboembolism,however,our patients' scintigramsare distinctly different: the defects are smaller (subsegmental), more numerous, tend to be peripherally located, and are distributed more or less evenly throughout both lungs, the P0 images being the most striking. Despite several days or even weeks of symptoms, there was little or no abnormality by chest radiograph. This combination of findings is distinctly atypical for thromboembolism (I ,2).
The mechanismthat producesthis unusual distributionof
perfusiondefectsisnotwellunderstood.Eatonproducedexperi
mental microembolism with I-mm polystyrene beads in dogs, and demonstrated preferential distribution of ischemic changes to the lung periphery. He postulated that small (micro) emboli are more likely to lodge in peripheral branches that arise in line rather than in side branches that come off at right angles (3). Sostman and associates recently reported two patients with similar findings and suggested the term “segmentalcontour pat tern―to describe the scintigraphic appearance of the perfusion images (4). They postulated that interstitial or parenchymal dis
easemustaccompanythevascularmicroembolismtoproducethis
perfusion pattern. In our autopsy cases, however, one had a corn plete lack oflymphangitic interstitial involvement, while the other had <5% involvement, contrasting with the 80% pulmonary ar
terial involvement. We believe the pattern to be primarilythe result
of the vascularemboliand peripheralarterialocclusion.
While we believe the described perfusion pattern to be typical of tumor microembolism, other conditions could produce a similar scintigraphic appearance. It could conceivably be produced by diffuse pulmonary arterial vasculitis, primary pulmonary arterial hypertension, fat or oil embolism, recurrent bland or septic em bolism from indwelling venous catheters and lines, and intravenous drug abuse. The associated clinical and radiologic findings should allow relatively easy differentiation of these conditions from tumor microembolism. Partially resolved or missed thromboembolism must also be considered in the differential, but the pattern of multiple, small subsegmental perfusion defects due to thrombo embolismhas been reportedonly rarely and, in fact, such a pattern is considered good evidence against thromboembolism (1,5). The incidence of tumor microembolism is not well documented,
@@.. -. @.... ( @. -
@@ .‘ •‘@@;
@@@ @“r@, ,.“;‘@.1@‘
@@@ ,-@;@4—. 1@'@@
@@@ @@k;'.@..il@' •. ..
@@@@@@@ ;“..@ ‘: •‘ -.. (. -..@-‘
@@@@ @@@@ @. -@ ‘T,@-
.j. ;_
FiG.2. Lungsectionshowstumorembolusinmuscularpulmonary
artery (black arrow), which has intimal proliferation (white arrow) significantlycompromisingthe lumen.Intimalproliferationhas occludedothervessels(openarrow).Notelack of lymphaticin
volvement.
878 THE JOURNAL OF NUCLEAR MEDICINE
—--
CRANE, RUDD, AND DAIL
ported by Sostman had similar lung perfusion images, both of which were histologically demonstrated to have pulmonary vas cular and lymphatic/interstitial involvement with tumor (4). Of the four cases reported by Green, three had both extensive lym phatic metastases and arterial tumor emboli (7). The one case in which perfusion images were shown had a pattern similar to our images. Pendergrassreporteda patient with gastric carcinoma who presented with increasing dyspnea (8). Chest radiographs showed changes compatible with lymphangitic involvement. Perfusion scintigraphy demonstrated multiple peripheral defects similar to those we have shown. Postmortem microscopy revealed extensive vascular occlusion by tumor emboli and associated thrombus, as well as extensive lymphatic involvement. Sadoff reported three breast-cancer patients who became dyspneic and had normal chest radiographs (9). Lung perfusion imaging revealed multiple defects, which resolved with chemotherapy. The findings in our patients and in the cases mentioned above suggest that tumor microembolism is the primary process pro ducing the perfusion changes, and that lymphangitic spread may represent a “complication―of tumor microembolism. As Green has suggested, the term “lymphangiticcarcinomatosis― may be inappropriate in many instances (7). Subacute cor pulmonale was the probable immediate cause of death in Patients A and B. Patient B had elevated pressures in the right ventricle and main pulmonary artery, without postmortem findings of right-heart enlargement or hypertrophy. Patient A had a dilated, poorly contracting right ventricle by gated RN yen triculography, but autopsy found no chronic changes of right ventricularor main pulmonaryartery enlargement or hypertrophy. Estimates have been made that about 2 months of increased pressures are required to produce myocardial hypertrophy; its absence in our patients argues that the tumor microembolism process was subacute to acute. In summary, we believe that the characteristic perfusion imaging findings in these patients are due to pulmonary arterial tumor microembolism rather than the frequently associated pul monary lymphangitic involvement. This unusual pattern in patients with a historyof neoplasm should warn the interpreterthat he may
not be dealing with simple thromboembolism but rather with tumor microembolism, a condition with significantly different prognostic and therapeutic implications.
REFERENCES
I. MCNEIL BJ: A diagnostic strategy using ventilation-perfusion studies in patients suspect for pulmonary embolism. J Nuci Med 17:613—616, 1976
2. MosEs DC, SILVERTM, BOOKSTEINii: Thecomplemen tary roles of chest radiography, lung scanning and selective pulmonary angiography in the diagnosis of pulmonary embo lism. Circulation 49:179—188, 3. EATON SB, JAMES AE, POTSAIDMS, et al: Scintigraphic findings in pulmonary microembolism. Am J Roentgenol 106:778-786, 1969
- SOSTMAN HO, BROWN M, TOOLE A, Ct al: Perfusion scan in pulmonary vascular/lymphangitic carcinomatosis: The segmental contour pattern. Am J Roentgenol 137:1072-1074, 1981 5. BIELLO DR, MATTAR AG, MCKNIGHT RC, et al: Ventila tion-perfusion studies in suspected pulmonary embolism. Am JRoentgenol133:1033—1037, 1979
6. KANERD,HAWKINSHK, MILLERJA,etal: Microscopic
pulmonary tumor emboli associated with dyspnea. Cancer 36:1473—1482, 1975
7. GREENN, SWANSONL, KERNW, et al: Lymphangitic
carcinomatosis: lung scan abnormalities. J Nuc/ Med 17: 258-260, 1976
8. PENDERGRASS HP, NEEL HS, CLEMENT PB, et al: Lung perfusion pattern associated with widespread occlusion of the pulmonary vessels and lymphatics. Radiology 105:615-616, 1972 9. SADOFF L, GROSSMAN J, WEINER H: Lymphangitic pul monary metastases secondary to breast cancer with normal chest x-rays and abnormal perfusion lung scans. Oncology 31:164—171,
The AnnualMeetingof the MissouriValleyChapterwill be heldOctober12—14,1984at the MarriottHotelin DesMoines, Iowa.The programwill featurecurrent informationon a varietyof topics such as Indium-ill imaging,SPECT,lympho
scintigraphy,andfuturedirectionsin nuclearmedicine.AMACategory1creditforphysiciansandVOICEcreditfortech
nologists have been applied for.
The ScientificProgramCommitteewelcomesthe submissionof abstractsof original contributionsin nuclearmedicine from members and nonmembers of the Society of Nuclear Medicine. Contributed papers will be presented on Saturday afternoon, October 13.
The Chapterwill presenttwo awardsfor best papersduring the meeting.The younginvestigatorwho presentsthe best paperwill receivethe RichardE. PetersonYoungInvestigatorsAward.Thetechnologistwho presentsthe bestpaperwill receive 50% of his or her expenses to attend the 1985 SNM Annual Meeting to present this paper.
For more information or to submit 200 word abstracts please contact:
Alexander Ervanian, M.D. Dept.of NuclearMedicine IowaMethodistMedicalCenter 1200PleasantStreet Des Moines,Iowa
Deadlinefor receiptof abstractsis September15,1984.
880 THE^ JOURNAL^ OF^ NUCLEAR^ MEDICINE
MissouriValleyChapter
Societyof NuclearMedicine
“NewHorizonsin NuclearMedicine―
October 12—14, 1984 MarriottHotel Des Moines, Iowa
