REFERENCES 15. Sims DO. Histiocytosis X. Follow-up of 43 cases. Arch Dis Chi/d l977;52:433— 1. Chu T, D'Angio GJ, Favara B, Ladisch 5, Nesbit M, Pritchard J. Histiocytosis 440. syndromes in children. Lancet 1987;l:208—209. 16. Cline Mi, Golde DW. A review and re-evaluation of the histiocytic disorders. Am J 2. Resnick D. Lipidoses, Histiocytoses and hyperlipoproteinemias. In: Diagnosis of bone Med 1973;55:49—60. andjoint disorders, 2nd ed. Philadelphia:W.B. Saunders;1988:2404—2459. 17. Sartoris DJ, Parker BR. Histiocytosis X: rate and pattern of resolution of osseous 3. Gilday DL, Ash JM, Reilly BJ. Radionuclide skeletal survey for pediatric neoplasms. lesions. Radio/ogy 1984;152:679—684. Radio/ogy1977;l23:399—406. 18. Martinez-Lage J, Poza M, Cartagena J, vincente J, Biec F, De Las Heras M. Solitary @ 4. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating eosinophilic granuloma of the pediatric skull and spine. Chi/ds Nerv Syst l99l;7: characteristic (ROC) curve. Radiology 1982;l43:29—36. 448—451. 5. Hanley JA, McNeil BJ. A method of comparing the areas under receiver operating 19. Schaub T, Ash JM, Gilday DL. Radionuclide imaging in histiocytosis X. Pediatr characteristic curves derived from the same cases. Radio/ogy 1983;148:839—843. Radio/ I987; 17:397—404. 6. The InternationalCommission ofRadiological Protection. ICRP Publication 34, Protection 20. Parker BR, Pinckney L, Etcubanas E. Relative efficacy of radiographic and radionu ofthe patient in diagnosticradiology.Oxford, England:PergamonPress; 1982:53-77. clide bone surveys in the detection ofthe skeletal lesions ofHistiocytosis X. Radio/ogy 7. Weber DA, Makler PT, Watson EE, Coffey JL. Radiation absorbed dose from l980;134:377—380. 9@―Tc-IabeIedbone imaging agents. J Nuc/ Med 1989;30:1 117—1122. 21. Antonmattei 5, Tetalman MR. Lloyd TV. The multi-scan appearance of eosinophilic 8. Kereiakis J, Rosenstein M, Bocaraton F. Handbook of radiation doses in nuclear granuloma. C/in Nuc/ Med l979;4:53—55. medicine and diagnostic x-ray. Colorado: CRC Press; 1980:165. 22. Kumar R, Balachandran S. Relative roles of radionuclide scanning and radiographic 9. Berry DH, Becton DL. Natural history of Histiocytosis X. Haem/Oncol CIin N Am imaging in eosinophilic granuloma. C/in Nuc/ Med 1980;5:538—542. 1987; 1:23—35. 23. 5iddiqui AR, Tashijan JH, Lazarus K, Wellman HN, Baehner RL. Nuclear medicine 10. Favara BE, McCarthy RC, Mierau GW. Histiocytosis X. Hum Pathol 1983;14:663-676. I I. Avery ME, McAfee JG, Guild HG. The course and prognosis of reticuloendotheliosis studies in evaluation of skeletal lesions in children with Histiocytosis X. Radio/ogy (eosinophilic granuloma, Schuller-Christian disease and Letterer-Siwe disease). A 1981; 140:787—789. study of forty cases. Am J Med 1957;22:636—652. 24. Westra 5J, vanwoerden H, Postma A, Elema JD, Piers DA. Radionuclide bone 12. Gianotti F, Caputo R. Histiocytic syndromes: a review. J Am Acad Dermato/ scintigraphy in patients with Histiocytosis X. Eur J Nuci Med 1983;8:303—306. 1985;13:383—404. 25. Fezoulidis I, Wickenhauser J, Schurawitzki H, Gritzmann N. Histiozytosis X: 13. Ccci A, De Terlizzi M, Colella R, et al. Langerhans cell histiocytosis in children: szintigraphische und rontgenologische Befunde. Rontgen-Bl l987;40:234—239. results from the Italian co-operative AIEOP-CNR-H.X'83 study. Med Pediatr Onco/ 26. Crone-Munzebrock W, Brassow F. A comparison of radiographic and bone scan 1993;2 1:259—264. findings in Histiocytosis X. Sk.e/eta/ Radio/ 1983;9:I70—173. 14. Enriquez P, Dahlin DC, Hayles AB, Henderson ED. Histiocytosis X: a clinical study. 27. George JC, Buckwalter KA, Cohen MD, Edwards MK, Smith RR. Langerhans cell Mayo C/in Proc 1967;42:88—99. histiocytosis of bone: MR imaging. Pediatr Radio/ 1994;24:29—32. Radionuclide Venography and Its Functional Analysis in Superior Vena Cava Syndrome

Asif Mujtaba Mahmud, Toyoharu Isawa, Takeo Teshima, Tomio Hirano, Yoshiki Anazawa, Makoto Mild and Toshihiro Nukiwa Department ofRespiratory Medicine, Institute ofDevelopment, Aging and Cancer (formerly Research Institutefor Chest Diseases and Cancer), Tohoku University, Sendai, Japan Superiorvenacava(SVC)syndromewasfirstdescribedby In addition to imaging, radionuclide venography maybe used for the William Hunter in 1757 (1 ). Presently, bronchogenic carcinoma functional assessment of superior vena cava (SVC) syndrome by is the leading cause and accounts for 67%—82%of all cases of applying the indices of transit time (IT), time of half-peak count (11-I) SVC obstruction (2). Initially, SVC syndrome was considered and peak count ratio (PC ratio). Methods Ten healthy subjects to be an oncologic emergency (3), but later reports have refuted (Group N) and 107 patients with SVC syndrome (64 symptomatic this notion (4—6). Emergency or not, it bears considerable and 43 asymptomatic) were studied. Images were visually assessed forcollateralsorjugularvenous refiuxand values of the indices were importance for both the clinician and patient. calculated. Results The 107 patients were subclassffied into three A careful diagnostic work-up is an essential prerequisite to groups according to the images obtained. Collateral circulation was successful management of this syndrome (4,6). Radionuclide seen in 37 patients (Group C). In 20 patients, jugular venous refiux venography is a convenient, noninvasive procedure which was observed (Group J). Fifty patients who showed neither collat forms an important component of such a work-up (2). In the erals nor refluxwere included in Group P. Incomparison to Group N present study, we intended to establish certain criteria on [3.6 ±0.56 sec (sern)],TI values were signfficantly higher (p < 0.05) radionuclide venography to aid in the diagnosis and post forGroupJ (7.13 ±1.16 see) and GroupC (7.00 ±0.87 see). Values therapeutic evaluation of SVC syndrome and to study hemody ofTh were significantly prolonged (p < 0.05)for Group J (23.6 ±4.8 namic characteristics in the central veins, consequent to stenosis see) and Group C (18.8 ±2.2 see) in comparison to Group N (9.2 ± and/or obstruction of the SVC. 1.5 sac). PC ratio values were higher in all patient groups in comparison to Group N (3.4 ±0.57). Conclusion: These indices are potentially useful in the initialdiagnosis and post-therapeutic evalu METHODS ation of SVC syndrome. Inthe absence ofother causes, appearance ofjugular venous refiux may be considered a sign of SVC syndrome. Between February 1986 and July 1994, radionuclide venography Key Words radionuclide venography superior vena cava syn was performed on 117 subjects. Ten healthy volunteer men (aged drome; jugular venous refiux; fUnCtiOnalindices 31—57yr; mean age 43 yr) constituted the control Group N. The J NucI Med 1996;37:1460-1464 remaining 107 were patients with SVC syndrome (97 men, 10 women; age 33—81yr; mean age 62.6 yr). Selection criteria included presence of clinically obvious SVC syndrome consisting of swelling of the face, neck and/or the upper extremities (symp Received May 30, 1995; revision accepted Nov. 3, 1995. tomatic in 64 patients) or only radiological signs such as medias Forcorrespondenceor reprintscontact As@MujtabaMahmud,Departhentof Respiratory Med@ine,Institute of Deve1opment@Agrig and Cancer, Tohoku University, tinal widening, suprahilar or hilar lesions and tumor location in 4-1 SeWyOUmachi,Aoba-ku, Sendsi 980, Japan. upper lung zones, rendering them susceptible to SVC syndrome

1460 THEJOURNALOFNUCLEARMEDICINE•Vol. 37 •No. 9 •September 1996 (asymptomatic in 43 patients). More than three of four had broncho Time of half-peak count is: genic carcinoma. The diagnostic profile is shown in Table 1. THasCV

Data Acquisition THLSCV, Boluses of 111 MBq (3 mCi) 9@Tc-MAA, less than 0.5 ml in volume, were injected simultaneously into both median cubital where TH is the time ofhalf-peak count in the subclavian veins and veins ofthe subjects lying in the supine position with their arms by is the time difference between the points of half PC on the curve their sides and holding the neck in the neutral position, i.e., the (Fig. 1). back and occiput touching the examination table. They were asked Peak count ratio is: to continue normal tidal breathing, avoiding Valsalva's maneuver PC ratio = MaX{PCRSCV, PCLSCV}/PCSvC. (7). The gamma camera was placed anteriorly over the neck and the upper chest. Data were stored in frame mode at 1 frame/sec for Peak count ratio was calculated by dividing the PC in the RSCV 2 mm at 64 x 64 matrices. This was followed by lung perfusion or LSCV, whichever was greater, by the PC in the SVC. imaging in anterior, posterior, right and left lateral views in sitting Statistical significance was determined with Bonferoni's multi position. ple-comparison procedure (8), paired or unpaired t-test; a p value less than 0.05 was considered statistically significant. Data Analysis Images were carefully assessed visually for the presence of RESULTS collaterals or internal jugular venous reflux. The stored data then Visual Assessment and Grouping were processed by a computer as follows. After image addition, In Group N, no jugular venous reflux or collaterals were seen four regions of interest (ROIs), depending on the anatomical size of in the venograms ofthe 10 normal subjects (Fig. 2A). In Group the vessel, were drawn with one each on the right and left P, 50 patients (46%) who showed neither collaterals nor jugular subclavian veins (RSCV and LSCV) (6 X 10 matrix size), the reflux were included (Fig. 2B). superior vena cava (SVC) (4 X 12 matrix size) and the right atrium Twenty patients (19%) showing only jugular venous reflux (RA) (6 x 9 matrix size) (Fig. 1). The size of each ROI were classified as Group J (Fig. 2C). Of these, 11 had no corresponds to the normal dimensions of the respective vessel. A clinical signs of SVC syndrome but had suggestive radiological time-activity curve was plotted for each ROl (Fig. 1). The curves signs. The remaining nine were symptomatic. for the subclavian veins showed single peaks. On the other hand, Thirty-seven patients (35%) showing various collaterals, as two peaks were usually visible on the curves for the SVC and the listed in Table 2, were categorized as Group C (Fig. 2D). In this RA corresponding to the entry ofradionuclide from the RSCV and group, seven patients showed no clinical manifestations of SVC the LSCV. The points on the curves were denoted as PC = peak syndrome. Certain common patterns of collaterals are shown in count, PT = peak time, PC/2 = half-peak count and TH = time of Figure 3. half-peak count. The subscript RSCV, LSCV or SVC denotes the vessel from which the curve was derived. R and L denote peaks Quantitative Analysis produced by the bolus from the right and left subclavian veins. The Paired t-test analysis revealed no statistically significant first peak is marked R because peak time in the RSCV was earlier, difference between the TT(R) and TT(L) and the THRSCVand as compared with the peak time in the LSCV, in the majority of THLSCVin all four groups. Therefore, the higher of the two cases. In patients with earlier peaks on the left side, the first peak values was used for comparison since it was considered more was accordingly considered to be caused by the left subclavian indicative of the hemodynamic derangement. Since we de bolus. Appearance of a single peak indicated concurrent entry of pended on the appearance of clearcut upward and downward boluses from the subclavian veins. slopes on the time-activity curves of the subclavian veins and the appearance of two peaks on the curve for the RA, we could PTRJ@(R) peak time in the right atrium, calculate TH or TT from one side only in a few cases having unclear curves. In addition, some cases with bizarre curves had produced by the bolus from RSCV to be omitted, which explains the discrepancy between the case number (n) presented on the graph and the total number for each PTI@J@(L) peak time in the right atrium, group. In the case of the PC ratio, a single value was obtained produced by the bolus from LSCV. from each subject. Mean values of indices for each group were compared among As explained earlier, the converse was considered in patients the four groups (Bonferoni's method) and among the asymp with earlier left subclavian vein peaks. The following three indices tomatic and symptomatic subgroups within each group (un were defined and their values were calculated for each subject, paired t-test). No statistical significance was seen in the latter according to: case. Transit time is: Transit Time As shown in Figure 4, an upward trend of U was observed, TT(R) = PTRA(R) PTRSCV indicating that it is prolonged as the disease state progresses. No further prolongation was observed in Group C. In comparison to TT(L) = PTRJ@(L) PTLSCV, Group N (3.6 ±0.56 sec), TT was significantly higher in Group J (7.13 ±1.16 sec; p < 0.05) and Group C (7.00 ±0.87 sec; where TT(R) denotes transit time of the bolus from the RSCV to p < 0.05). Group P differed significantly from Groups J (j < the RA and is calculated by subtracting the peak time in the RSCV 0.05) and C (p < 0.01). from the time of the peak produced in the RA by the right subclavian bolus. Similarly, TT(L) means transit time from the Time of Halt-Peak Count LSCV to the RA and is calculated by using corresponding values As illustrated in Figure 5, the TH in the subclavian veins was on the left side. highest in Group J (23.6 ±4.8 sec) which differed significantly

RADIONUCLIDE VENOGRAPHY IN SVC SYNDROME ‘Mahmud et al. 1461 TABLE I Diagnostic Profile of Patients of patientsBronchogenicD_Number carcinoma(88)Squamous-cell26Adenocarcinoma25Small-ceO22Large-cell15Mediastinal

tumors13Metastatic cancer3Miscellaneous3Total107

from Groups N (9.2 ±1.5 sec; p < 0.05) and P (13.2 ±1.4 sec; @ FiGURE2. Venographicimages of four subject groups. Group N. (B) p < 0.05). Group P. (C)Group J, note visu@ation of internaljugularveins bilaterally Peak Count Ratio (arrows)indicatingretrogradeflowas a consequenceof impalredflow throughthe SVC.(D)Group C, bilateralinternalthoracic veins are seen as The PC ratio showed an upward trend (Fig. 6) which was collaterals. similar to the tendency observed for iT and TH. Although insignificant statistically, values were higher for all patient mous-cell carcinoma being the leading cell type (2). There were groups in comparison to Group N (3.4 ±0.57). There was little no patients with valvular heart disease, overt heart failure or difference between Group J (6.8 ±2.55) and Group C (6.78 ± thoracic venous catheters. 0.97). Collatarals DISCUSSION In our series, 37 patient (35%) showed collaterals, whereas Radionuclide venography is a useful, accurate and noninva Colta.rt et al. (11) encountered collaterals in 50%. The most sive method of imaging the venous system (9). In addition to common collateral pathways were the lateral thoracic, internal imaging, it can be used to study the functional aspects of SVC thoracic and intercostal veins. We used the term cervical venous syndrome (10). Consistent with other large series, more than network to denote the vertebral and highest intercostals as 80% of our patients had bronchogenic carcinoma, with squa defined by Muramatsu (12). Twenty percent ofthe patients with collaterals were asymptomatic and, therefore, the presence of collaterals served as a confirmatory sign of SVC syndrome. SCHEMATICDIAGRAM Gradual development of the syndrome in cases with partial OF ROl ANDTIMEACTIVITYCURVES stenosis of the SVC may lead to such a phenomenon (13). couu'r Jugular Venous Reflux Jugular venous reflux has been observed by several authors in the course of dynamic brain flow imaging (14—16).Rao et al. (16) mentioned certain technical factors as possible causes of the reflux. Retrograde flow in the central veins of subjects with valveless cervical venous systems or cervical venous incompe tence has also been implicated (1 7). It has also been described as a sign ofelevated systemic venous pressure due to congestive heart failure (18). According to Peart and Driedger, appearance of jugular reflux warrants further investigation of the medias tinum (19). Miyamae encountered it, secondary to central vein COUNTS obstruction, during radionuclide venography (10). Irrespective of the cause, retrograde flow seems to be the principal factor in the pathogenesis ofthe jugular reflux, even in cases with absent or incompetent cervical valves. In our series, reflux was

TABLE 2 Frequency of Collateral Pathways of patientsI@ateralCollateralNumber vein15(40.5%)Internalthoracic vein15(40.5%)Intercostalthoracic vein15(40.5%)Jugularvenousarch10 (27%)Cervical venous8(21%)networkExternal FIGURE1. Diagramof four ROls and relevanttime-activitycurves. PC = peak count; PT = peak time;TH = time of half-peakcount; RSCV= right vein7(19%)Azygosvein6(16%)Superiorjugular subclavianvein;LSCV= leftsubclavianvein;SVC= superiorvena cava RA = right atrium; PTRA(R) P@tifl1@ of bolus coming from right subclavian intercostalvein4(10%) vein;PT@ = peak time of bolus comingfromleftsubclavianvein.

1462 THEJOURNALOFNUCLEARMEDICINE•Vol. 37 •No. 9 . September1996 LAT.THORACICV. AZYGOSV. JUGULARARCH

‘v@

INT.THORACICV. CERVICALVENOUSNETWORKEXT.JUGULARV. FiGURE3. Frequentpatterns of collateralcirculation.Upperrowfromleftto nght lateralthoracic vein,azygos vein,jugulararch. Lowerrowfrom leftto right internalthoradc vein,ceMcalvenousnetwork(12),extemsijugularvein. FiGURE5@Comparison of time of half peak count (TH)between normal associated with increased values of TT, TH and PC ratio in the subjects and patientgroups. *p< 0.05. Errorbars show 1 s.e.m. presence of either clinical or only radiological signs of SVC syndrome. In addition, the absence of heart failure and avoid The TH is a parameter which is directly proportional to the ance of technical factors indicated SVC stenosis as the cause of persistence ofthe radionuclide in the subclavian vein and serves as retrograde flow leading to the appearance of jugular venous a sensitive index of slow blood flow in the central veins. In Figure reflux. In Group J, 11 patients (>50%) did not have clinically 5, an upward trend suggestsincreasingly sluggishflow through the overt SVC syndrome. The presence of jugular venous reflux in central veins with the highest values noted for Group J. Interest these patients could suggest the existence of SVC syndrome. It, ingly, values ofTH register a fall in Group C, owing to availability therefore, served as a valuable and early sign of SVC obstruction. ofcollateral pathways. The presence ofincreased TH per se should not be considered diagnostic of SVC syndrome as has been Functional Indices previously reported (21). For diagnostic purposes, it should be The appearance of two peaks in the curves from the SVC and considered in combination with other findings such as jugular the RA may be attributed to a slight discrepancy in timing of venous reflux or increased transit time. It may serve as a useful injections between the patient's two arms. parameter in follow-up studies. The PC ratio also showed an Transit time values for normal subjects (3.6 ±0.56 sec) are upward trend, similar to U and TH. This common upward concordant with those ofVincken et al. (2t1). We considered TT tendency represents the disturbance in blood flow. from the subclavian veins to the right atrium as most sensitive Presence ofa statistically significant difference ofTT and TH and specific for flow disturbance in the central veins, since it between normal subjects and different groups of patients represents IT across our focus of interest (the central veins) and indicates that different hemodynamics are at work in each is independent of local factors, such as arm edema and position group. Although, no statistical difference was seen for the PC and axillary node swelling. Also, it is not influenced by the ratio among the four groups of subjects, we consider the ratio to discrepancy in timing of injection. Other authors have calcu be potentially useful for follow-up studies. Since none of the lated transit time of the bolus from the antecubital veins to the indices showed any statistical difference between edematous right heart (10) and its arrival time in the subclavian veins and and nonedematous patients within each group, we may con the SVC from the arm (9). Local factors may exert an influence dude that, hemodynamically, they have similar derangements. on these values. Figure 4 depicts an upward trend of TT among the four groups but no further prolongation is observed on opening of collateral pathways. CONCLUSION Our analysis of the time-activity curves in correlation with the images indicate that:

N P J C FIGURE4@Comparisonoftransittime(11)betweennormalsubjectsand FiGUREa c@ompansonof peak count ratio (PC ratio) between normal patient groups. *:p < 0.05; ‘@:p < 0.01. Errorbars show 1 s.e.m. subjects and patient groups. Errorbars show 1 s.e.m.

RADIONUCLIDE VENOGRAPHY IN SVC SYNDROME •Mahmud et al. 1463 1. On radionuclide venography, the presence ofjugular venous 5. Sculier JP, Evans WK, Feld R, et al. Superior vena caval obstruction syndrome in small cell lung cancer. Cancer 1986;57:847—85l. reflux may serve as an early sign of SVC syndrome, 6. Yellin A, Rosen A, Riechert N, Lieberman Y. Superior vena cava syndrome. The provided other causes of the reflux are ruled out. myth—thefacts. Am Rev Respir Dis 1990;141:1114—1118. 2. Prolonged transit time (TT) from subclavian veins to the 7. Friedman BH, Lovegrove FTA, Wagner HN. An unusual variant in cerebral circulation right atrium is proposed as a useful indicator of slow studies. J Nuci Med 1974;15:363—364. 8. Godfrey K. Comparing the means ofseveral groups. In: Bailar JC 111,Mosteller F, ed. passage of blood through the central veins. Medica/ uses ofstatistics. Waltham, MA: NEJM Books; 1986;205—234. 3. Increased values of time of half-peak count (TH) indicate 9. Van Houtte P, FrOhlingJ. Radionuclide venography in the evaluation of superior vena sluggish blood flow through central veins. cava syndrome. C/in Nuc/ Med 1981;6:l77—183. 10. Miyamae T. Interpretation of ‘@“Tcsuperior vena cavograms and results of studies in 4. The peak count ratio (PC ratio) may be considereda 92patients.Radio/ogy1973;l08:339—352. valuable indicator of hemodynamic disturbance in the I 1. Coltart R5, Wraight EP. The value of radionuclide venography in superior vena cava SVC syndrome. obstruction. C/in Radio/ l985;36:4l5—418. 12. Muramatsu T, Miyamae T, Doi Y. Collateral pathways observed by radionuclide 5. These indices are helpful in detectinghemodynamic changes superior cavography in 70 patients with superior vena caval obstruction. C/in Nuci Med in patients, irrespective of the presence or absence of clini 199l;16:332—336. cally manifest edema of the upper half of the body. 13. Yedlicka JW, Schultz K, Moncada R, et al. CT findings in superior vena cava obstruction. Semin Roentgen 1989;24:84—90. 6. These indices are potentially useful parameters in fol 14. Yeh E, Pohlman GP, Reutz PP, Meade RC. Jugular venous reflux in cerebral low-up studies of the same patient. radionuclide angiography. Radio/ogy 1976;l 18:730—732. 15. Ogawa 1K, So 5K, Gerberg E, et al. Jugular—dural sinuses—jugular reflux in dynamic brain flow imaging as a sign ofunilateral innominate vein obstruction. JNuc/ ACKNOWLEDGMENT Med 1977;18:39—41. We thank Ms. Yuko Ogata for providing technical assistance 16. Rao BK, Polcyn RE, Lieberman LM. Influence of respiratory maneuvers on jugular during the study. venous reflux. C/in Nuci Med 198 l;6:23—26. 17. Hayt DB, Perez LA. Cervical venous reflux in dynamic brain scintigraphy. J Nuc/ Med 1976;17:9—12. REFERENCES 18. Blood flow studies@In: Baum 5, Vincent NR, Lyons KP, Wu Gurkin SC, eds. Atlas 1. Hunter W. The history ofan aneurysm ofthe aorta with some remarks on aneurysm in ofnuclear medicine imaging. New York, NY: Appleton-Centuiy-Cmfts; 1981:291-314. general. Med Observ Inq l757;I:323—357. 19. Peart RA, Driedger AA. Effect of obstructed mediastinal venous return on dynamic 2. Nieto AF, Doty DB. Superior vena cava obstruction, clinical syndrome, etiology and brain blood flow studies: case report. J Nuc/ Med 1975;16:622—625. treatment. Curr Prob/ Surg I986;10:442—484. 20. Vmcken W, Rods P, Sonstabo R, et al. Effect of neck position during radionuclide 3. Lokich ii, Goodman R. Superior vena cava syndrome. Clinical management. JAMA superior cavography. C/in Nuc/ Med l983;8:424—426. l975;23l:58—61. 21. Podoloff DA, Kim EE. Evaluation of sensitivity and specificity of upper extremity 4. Ahman FR. A reassessment of the clinical implications the superior vena cava radionuclide venography in cancer patients with indwelling central venous catheters. syndrome. J C/in Onco/ 1984;2:96 1—969. C/in NucI Med 1992;17:457—462. Iodine-123-MIBG Imaging of Neuroblastoma: Utility of SPECT and Delayed Imaging

Vittoria Rufini, Gregg A. Fisher, Barry L. Shulkin, James C. Sisson and Brahm Shapiro Division ofNuclear Medicine, Department oflnternal Medicine, University ofMichigan Medical Center, Ann Arbor, Michigan; and Division ofNuclear Medicine, Sacred Heart Catholic University, Rome, Italy

more lesionsthan 24-hr imaging,but it may also miss lesions with Possible incremental diagnostic benefits of SPECT and delayed rapklwashout. SPECTimagingsignificantlyincreasesthe numberof planarimaging with r@MIBG in neuroblastoma have not yet been lesionsdetected and better definesanatomiclocationoftumors. fullyestablished. Methods Whole-body delayed planar r@oMIBG imagingat 48 hrand SPECT imagingof the chest-abdomen orother Key Words iodine-123-MIBG; neuroblastoma planar imaging; de suspected sites obtained at 24 hr were compared with routine layedimaging;SPECT planar imaging at 24 hr in 83 studies of 29 children with neuroblas J Nuci Med 1996 37:1464-1468 tome. The sensitivity for each of the [1231]MIBGimaging methods was calculated on a study-by-study and on a lesion-by-lesion basis. Results: Frfty-one planar imaging studies were performed in 20 INletaiodobenzylguanidine (MIBG) labeled with either ‘@‘Ior patients with evidence of disease which was detected in 48 studies 1231 has been used for nearly 15 yr in the scintigraphic by 24-hr imaging (94.1% sensitivity) and in 44 studies by 48-hr evaluation ofneuroendocrine tumors, particularly pheocbromo imaging (86.3% sensitivity). On a lesion-by-lesion basis, sensitivity cytomas and neuroblastomas (1—3).In patients with neuroblas was 88.8% forthe 24-hr scan, 86.7% forthe 48-hr scan and 922% toma, a sensitivity of 77%—96%and a high specificity (approx for a combination of the two (p = ns). Forty-three SPECT studies were performed in 20 patients with evidence of disease in the field of imately 100%) has been reported for both radiopharmaceuticals @iewof the SPECT camera Dlaease was detected in 40 SPECT (3—6).Due to its physical properties (159 keV photon energy, studies(93% sensitivity),in38 planarscans at 24 hr(84.4%sensitMty) 13-hr half-life and paucity of particulate emission of 1231)and and in 37 planarscans at 48 hr (86.0% sensitivity).On a lesion-by the high activity that can be administered, MIBG labeled with lesionbasis, sensitivitywas 83.6% forthe 24-hr planarscan, 86.1% 123! has superior imaging characteristics. These characteristics for the 48-hr planar scan, 882% for a combination of the two planar of [‘231]MIBG,and its favorable dosimetry, even in higher scans and 97.9% for SPECT (p < 0.001 compared with planar). The administered doses, make its use preferable in children (7—8). anatomk@locations of tumors were clearer on SPECT in 15 studies. Conclusion: Delayed48-hrplanarscanningmay occasionallydepkt Iodine-123-MIBG imaging is most commonly performed at 24 hr after tracer administration.However, the optimal timing Received Jun. 26, 1995; revision accepted Dec. 27, 1995. for scintigraphy has not been definitively established; a recent For correspondence or reprints contact: V. Rufini,MD, Divisionof Nuclear Medkina, Sacred Heart Catholic University, PoliclinicoGemelli, Largo A. Gemelli 8, Rome 00168, report by Paltiel et al. (9) points out the improvement of itsiy. thoraco-abdominal lesion visualization in 48-hr images.

1464 THEJOURNALOFNUCLEARMEDICINE. Vol. 37 •No. 9 •September 1996