ORIGINAL ARTICLE 
Annals of Nuclear Medicine Vol. 8, No. 2, 115-123, 1994 
Intra-arterial infusion of N-isopropyl-p[123I]iodoamphetamine for assessing effective blood supply to pulmonary and hepatic neoplasms 

Chihoko MIYAZAKI 

Department of Diagnostic Radiology, Sapporo City General Hospital 

The biodistribution and pharmacokinetics of intra-arterially administered N-isopropyl-p[123I] iodoamphetamine ( 123I-IMP) were prospectively evaluated in 38 patients with histologically proven pulmonary or hepatic tumors. Intra-arterially infused 123I-IMP was distributed initially in peripheral tissues in which the blood supply was maintained. Its concentration in malignant neoplasms was demonstrated to be higher than in normal tissues. In pulmonary cancer, the tumor uptake of the administered dose without a tissue attenuation correction (% uptake) of 123I-IMP at 1-2 min after injection was 14.7 +- 5.7% (s.d.). The tumor to normal tissue ratio was 2.1 +- 0.7 in hepatocellular carcinoma and 1.4 +- 0.7 in metastatic tumors. The biodistribution of 123I-IMP was also compared to that of 99mTc-macroaggregated albumin (99mTc-MAA) in 9 cases of hepatic cancer. The distribution of 123I-IMP resembled that of 99mTC-MAA in 5 cases and was different in 4 cases. 123I-IMP was more concentrated in the tumor than 99mTc-MAA.  
Intra-arterial infusion scintigraphy with 123I-IMP seems to provide information on effective blood supply to neoplasms which are targeted in interventional radiology. 

Key words: tumor blood flow, intra-arterial infusion, N-isopropyl-p[123I]iodoamphetamine ( 123I-IMP), interventional radiology 

INTRODUCTION 
INTRA-ARTERIAL INFUSION of an anti-tumor chemotherapeutic agent has proven to be effective in the management of primary and metastatic malignant neoplasms. Inter-ventional radiology has come into wide use in this intra-arterial approach. In order to derive the maximal effectiveness by chemotherapy to the target masses, we have to consider several factors: the optimal dose, pharmacokinetics, distribution, and vascular supply of the drug administered intra-arterially. The best way to evaluate intra-cellular drug uptake may be to monitor the biodistribution and pharmacokinetics of the drug labeled with a radionuclide. Positron labeled BCNU and cisplatin have been proposed for the clinical assessment of chemotherapy. 1-3 However, these agents are not available for routine studies. 
Radionuclide studies with 99mTc-macroaggregated al-

Received June 7, 1993, revision accepted December 20, 1993. 
For reprint contact: Chihoko Miyazaki, M.D., Department of Diagnostic Radiology, Sapporo City General Hospital, W-9, N-1, Chuo-ku, Sapporo 060, JAPAN. 

bumin (99mTc-MAA), microspheres and 81mKr have been employed for quantifying drug delivery to the target tumor tissue and normal tissue. 4-8 99mTC-MAA perfusion scintigraphy plays a significant role in determining blood perfusion because of its easy availability and appropriate particle size for the first pass capillary blockade. Correlative studies between tumor response and drug distribution via direct measurement of tumor and liver drug levels with 3H-fluorodeoxyuridine (FUdR) and 99mTc-MAA have been reported. 9-11 Also, the prediction of tumor response with both 99mTc-MAA and 99mTc-sulfur colloid (SC) has been re ported. 12, 13 Although 99mTc-MAA perfusion scintigraphy allows accurate assessment of first-pass blood perfusion to the target masses and surrounding tissues, it is limited as an effective indicator of drug delivery to the extra-cellular space. 
N-isopropyl-p[123I]amphetamine (123I-IMP) was found to be a lipophilic substance and it was cleared from the blood with a greater than 90% extraction fraction on the first pass through the brain when injected into the internal carotid artery. 14,15 Winchell et al. proposed that the initial distribution of 123I-IMP might be useful for imaging relative regional brain perfusion in humans. 16,17 Kuhl et al., showed that the value of regional cerebral blood flow obtained by the distribution of 123I-IMP in the brain corresponded closely to the value obtained by the radioactive labeled microsphere method. 18 It is postulated that 123I-IMP passes through the blood brain barrier and that 123I-IMP binds to non-specific binding sites, either inside or outside the cell. 19,20 Considering these reports on 123I-IMP as a freely diffusible and highly extracted tracer for regional cerebral blood flow measurement, the author assumed that the first arterial distribution of 123I-IMP must reflect the lung and the liver blood perfusion including the neoplasms as well as brain tissues. 
The scintigraphic studies concerning the intra-arterial infusion of 123I-IMP in patients with pulmonary and hepatic neoplasms were designed to investigate the distribution of 123I-IMP in the lungs and liver to determine the tumor blood flow. In this paper, the first pass distribution, uptake, and change as a function of time in 123I-IMP in the lungs and the liver were analyzed. Tumor stains on contrast angiograms were compared to those on 123I-IMP images, and 123I-IMP and 99mTc-MAA scintigrams were compared in patients with hepatic tumors with respect to the distribution and the intensity of radioactive agents. 

MATERIALS AND METHODS 
Study population The population studied consisted of 10 patients with pulmonary carcinoma and 28 patients with hepatic tumors (Table 1). All tumors were histologically proven. The patients' ages ranged from 38 to 79 years old. A total of 12 bronchial arterial infusion (BAI) studies with 123I-IMP were performed in the 10 patients with pulmonary carcinoma immediately after percutaneous bronchial arteriography and interventional chemotherapy. In 12 cases of metastatic hepatic tumors and 2 cases of hepatocellular carcinoma (HCC), hepatic arterial infusion (HAI) studies with 123I-IMP were performed after percutaneous hepatic arteriography and interventional chemotherapy in which the catheter was placed in the celiac, common hepatic, proper hepatic, right or left hepatic artery. In 5 cases of HCC, 123I-IMP scintigraphy was carried out before hepatic arterial embolization. HAI studies with both 123I-IMP and 99mTc-MAA were conducted in the other 9 cases which had undergone operative placement of an hepatic chemotherapy catheter to metastatic tumors. Informed consent was obtained from all patients before the procedures. 

Radionuclide study 
Thirty-seven MBq of 123I-IMP/2 ml was injected slowly through the catheter which was inserted into the bronchial or hepatic arteries and was followed by an infusion of saline solution in 29 patients. In the 9 patients with surgical arterial catheter placements, 37 MBq of 123I-IMP/2 ml and 185 MBq of 99mTc-MAA/2 ml were introduced into the hepatic artery through the side port or a subcutaneous reservoir. The 99mTc-MAA study was done within one week after the 123I-IMP scintigraphy. The studies with 123I-IMP and 99mTc-MAA were carried out in the supine position with a large-field-of-view gamma camera with a parallel-hole low-energy collimator. Sequential images were obtained every 3 minutes after administration of the radiopharmaceuticals. Data were collected every 30 seconds for 60 minutes by means of a Scintipac 2400 computer. 
Before injection, the dose in the syringe was calibrated by the gamma camera under standardized geometry (at 20 cm) for 1 minute. After completion of the procedure, for both the syringe and the catheter, which were pulled out, the remaining radioisotopes were counted in the 29 patients. 

Data processing 
The data were processed according to the following protocol: the region of interest (ROI) of irregular shape was assigned along the increased nodular uptake in the lung or the liver and along the whole lung or liver parenchyma, based on the findings of the chest roentgenogram, CT, liver scintigram and angiogram. Square ROIs (5 x 5 pixels) were assigned to the normal lung and right lobe of the liver in BAI studies and to the right lung and heart in HAI studies. Time activity curves were generated from all the assigned ROIs. 
Percentage uptakes by the whole lung, lung tumor and whole liver of the administered dose were calculated as: [counts of whole lung, lung tumor, or whole liver divided by injection counts] x 100%. They were obtained from 1-2 minute-images after the administration of 123I-IMP.
Tissue attenuation correction for emitted gamma-rays was not carried out. The percentage uptake of the whole liver tumor was difficult to calculate because several lesions were multicentric. Instead, the tumor/liver (T/L) ratio was calculated as counts of the tumor divided by counts of the liver parenchyma by means of the same sized ROI. 
The washout rates from tumors and normal tissue at 30 and 60 minutes for counts of 1-2 minutes were calculated with corrections for background radioactivity and physical decay. Under curve analysis, a ROI of a contralateral normal lung, or a heart, was regarded as a ROI of background radioactivity in the BAI and HAI studies, respectively. 
The tumor delineation of 123I-IMP infused into the hepatic artery was compared with the tumor stain in the conventional contrast angiography. 
Statistical analysis was performed by Student's t-test. 

RESULTS 
Pulmonary neoplasms 
123I-IMP infused into the bronchial artery was distributed not only in the pulmonary cancer and lymph nodes, but also in normal tissue along the bronchial and intercostal arterial distribution and the mediastinum (Fig. 1). 
In the investigation of the percentage uptake of the whole lung and tumor in 11 cases of primary lung carcinoma, the mean value for the whole lung was 40.6 +- 9.2%, and that for the tumor was 14.7 +- 5.7%. In one case of metastatic breast carcinoma approximately 2 cm in diameter, the percentage uptake of the whole lung was 86.3% and no 123I-IMP accumulation in the tumor could be detected. 
The mean washout rates for 123I-IMP in 10 tumors, 8 lymph nodes, and 8 normal pulmonary areas were 39.6 +- 10.8, 45.1 +- 11.0 and 29.4 +- 11.8%, respectively, at 30 minutes; and 55.6 +- 14.2, 62.8 +- 11.0 and 41.4 +- 10.6% at 60 minutes (Fig. 2). There were statistically significant differences in the washout rate between lymph nodes and normal tissues at 30 and 60 minutes, and also between tumors and normal tissues at 60 minutes (p < 0.05, 0.01 , and 0.05 respectively). 
Hepatic neoplasms 
123I-IMP was accumulated in the liver parenchyma, gall bladder and tumor. For the cases in which the catheter tip was set at the celiac artery, accumulation of 123I-IMP was also seen in the spleen and gastrointestinal tract. 123I-IMP uptake in the liver parenchyma and tumor decreased with time. The splenic uptake of 123I-IMP decreased with time, and 123I-IMP in the spleen moved into the liver via the portal vein. 
With respect to the comparison of 123I-IMP studies with 99mTc-MAA studies in 9 patients, the distribution of 123I-IMP was the same as that of 99mTc-MAA in 5 patients and it was different from that of 99mTc-MAA in the rest. However, in the former cases, the intensity of 123I-IMP activity in the tumor was higher than that of 99mTc-MAA in 2 cases (Fig. 3). In the latter cases, the distribution of 123I-IMP in the peripheral liver parenchyma was more homogeneous and was better visualized than that of 99mTc-MAA. 
The mean percentage uptake in the whole liver was 40.2 +- 18.2% in 7 cases in which the catheter tip was set in the right hepatic artery. In 6 cases in the proper hepatic artery, in 3 cases in the common hepatic artery, and in 3 cases in the celiac artery, the mean percentage uptake in the whole liver was 51.2 +- 7.2%, 41.8 +- 15.9% and 16.1 +- 2.5%, respectively. 
The mean T/L ratio in 5 out of 7 HCCs was 2.11 +- 0.69 at 2 minutes, 1.80 +- 0.78 at 30 minutes and 1.55 +- 0.67 at 60 minutes. In the 2 cases excluded from the evaluation, a ROI could not be assigned because the gamma camera was incorrectly positioned. The mean T/L ratio of 6 out of 10 cases of metastatic liver tumors was 1.37 +- 0.70 at 2 minutes and 1.08 +- 0.65 and 1.09 +- 0.67 at 30 and 60 minutes (Table 2). There was no statistically significant difference between the mean T/L ratio of the HCC and the metastatic tumor at each time period. In 2 of the 4 cases excluded, the tumors were too small to assign a ROI and in the other 2 cases tumors could not be detected. 
The mean washout rate at 30 minutes after the administration of 123I-IMP was 0.7 +- 6.7% from normal liver parenchyma in 9 cases, 15.2 +- 11.6% from liver cirrhosis in 5 cases, 23.2 +- 7.8% from HCC in 5 cases, and 12.2 +- 7.6% from metastatic tumors in 6 cases. The mean washout rate at 60 minutes was 4.2 +- 8.3%, 21.6 +- 15.3%, 37.6 +- 11.9% and 17.7 +- 9.3%, respectively (Fig. 4). The mean washout rate from liver cirrhosis was faster than that from normal liver parenchyma (p < 0.05), and the mean washout rate from HCC was faster than from metastatic tumors at both 30 and 60 minutes (p < 0.05). 
The tumor uptake of 123I-IMP in the initial images was compared with the tumor stain in contrast angiograms in 19 patients. Two cases were positive for 123I-IMP and negative for contrast angiograms, whereas 3 cases were negative for 123I-IMP and positive for contrast angiogram. In the other 14 cases, the findings of 123I-IMP scintigraphy were similar to those of the contrast angiography. 

DISCUSSION 
123I-IMP infused into the bronchial or hepatic arteries accumulated in the normal and malignant tissues supplied by those arteries, and the uptake of 123I-IMP in the tissue decreased with time. Comparing the behavior of 99mTc-MAA and 123I-IMP infused into the hepatic artery in 9 patients, 4 cases showed different distribution patterns and 2 cases showed different intensities of radionuclide activity in the tumor. Among the former cases, there were 2 cases in which 99mTc-MAA was accumulated only in the region of the catheter tip, but 123I-IMP was distributed in both the region of the catheter tip and the liver parenchyma, then 123I-IMP in the catheter tip moved into the liver parenchyma and the tumor with time. Partial or complete hepatic arterial thrombosis and catheter thrombosis have been reported as the major causes of unsatisfactory perfusion after surgical placement for hepatric chemotherapy. 21-23 From this point of view, there is the possibility that stenosis of the catheter tip due to a thrombus or arterial thrombosis adjacent to the catheter might have been overestimated in the 99mTc-MAA study. In the other 2 cases, the left lobe of the liver was not visualized in 99mTc-MAA scintigraphy but it was seen in 123I-IMP scintigraphy. The occurrence of laminar flow in the hepatic artery, which may result in poor mixing of the infusate and the blood, might explain these findings. 22-25 The distribution of 99mTc-MAA might be influenced by arterial flow and inadequate mixing in the blood because of its larger molecular weight than that of 123I-IMP. 
The mean percentage uptake of the whole lungs and tumor in primary pulmonary cancer patients, to the total administration dose of 123I-IMP, was 40.6 +- 9.2% and 14.7 +- 5.7%. That of the whole liver was about 50% or less. One reason for the low percentage uptake might be calculating without tissue attenuation correction. In addition, the following reasons are also possible: anastomosis of the bronchial artery and pulmonary artery via the capillary bed around the bronchial wall in normal lung, a bronchial-pulmonary shunt as a pre-capillary shunt around the tumor, and a hepatic arterial-venous or arterial-portal shunt. 23,26,27 In these contexts, the percentage uptake of 99mTc-MAA might be overestimated because capillary blockade may happen before an arterial-venous shunt.
From the results of this study, the mean T/L ratio of HCC to liver parenchyma, at 2 minutes after the administration of 123I-IMP into the hepatic artery, was 2.1 +- 0.7. The case in Figure 5, in which the catheter tip was placed in the right hepatic artery, showed a T/L ratio of 2.26 at 2 minutes. Based on this result, it was surmised that the concentration of chemotherapeutic agent accumulated in the HCC might be more than two times that of the surrounding liver parenchyma. Nakajo et al. reported that the T/L count ratios in HCC were 1.9 to 3.7 by SPECT 2 hours after the arterial infusion of iodine-131-labeled lipiodol and these count ratios corresponded to 7.5-21 in the T/L concentration ratio. These count and concentration ratios were interconnected by an experimental study by SPECT using the liver, tumor and body phantoms. 28 On the other hand, Raoul et al, noted that the T/L ratio in 23 cases of HCC was 4.3 +- 2.6 by planar scintigrams 24 hours after the arterial infusion of iodine-131-labeled lipiodol ultra-fluid. 29 In patients with hepatic metastases, the mean T/L ratio was 1.4 +- 0.7 at 2 minutes after the administration of 123I-IMP. Sigurdson et al. mentioned that the T/L ratio in the cases of colorectal hepatic metastases was 0.78 +- 2.10 obtained by planar scintigrams with 99mTc-MAA and that after hepatic arterial infusion of 1 micro Ci/kg 3H-fluorodeoxyuridine, the mean tumor uptake of the drug was 12.4 +- 12.2 nmol/g tissue, compared with a mean hepatic level of 23.9 +- 11.4 nmol/g. 11 In both colorectal and carcinoid tumors in the liver, Gyves et al. demonstrated that the T/L ratio was approximately 3.0 by SPECT with 99mTc-MAA 30. 
The washout rate from pulmonary cancers and lymph nodes was faster than that from the normal lung tissue. The slower washout from the normal lung tissue may be due to gradual accumulation and slow release of 123I-IMP in the lung after systemic circulation of 123I-IMP. A high lung uptake of 123I-IMP has been reported and 123I-IMP is bound to the receptor of the endothelial cell membrane in the pulmonary capillaries. 31,32 In the study of pulmonary cancer with intravenous administration of 123I-IMP, impaired uptake of 123I-IMP in the lesion was demonstrated. 33-35 The washout rate from liver cirrhosis was faster than normal liver parenchyma. This is assumed to be due to a decrease in the cellularity of the liver parenchyma in liver cirrhosis. And the washout rate from HCC was faster than from metastatic cancer. This difference may be due to high blood perfusion in HCC. Also, the effect of the recirculation of 123I-IMP, the effect of amphetamine derivatives, or the environment of the cells including H must be taken into consideration. 
In the comparison between tumor stain in hepatic contrast angiogram and the tumor uptake of 123I-IMP, the 123I-IMP study showed better visualization of the tumors compared than in contrast angiography in 2 of 19 cases (Fig. 6). Those two cases were one metastatic cancer from the colon and one regenerative nodule of liver cirrhosis. Similar findings are reported in another 99mTc-MAA study. 36 On the other hand, in 3 cases (2 HCCs and one metastatic cancer from the colon) , no accumulation of 123I-IMP was detected in spite of a positive tumor stain in contrast angiogram. This is because there was minimum peripheral tumor stain with a central hypovascular area in an angiogram in one case, and the position for scintigraphy was not suitable in the other case. 
123I-IMP accumulated in the hepatic and pulmonary tumors as well as those of parenchyma through the catheter which was placed in the bronchial or hepatic arteries. 123I-IMP can be used as an intra-arterial imaging agent to estimate the arterial blood supply to the organs with neoplasms and possibly intracellular distribution of chemotherapeutic agents when they are intra-arterially administered. 

ACKNOWLEDGMENTS 
I would like to thank M. Furudate, M.D. and K. Itoh, M.D, from Hokhaido University and D. Spigos. M.D. from Ohio State University Hospital (Columbus, Ohio) for their valuable comments, and H. Suzuki, M. Tokita, and S. Kinouchi of the Nuclear Medicine Section of Sapporo City General Hospital for their help and technical support. 

REFERENCES 
1. Diksic M, Sako K, Feindel W, Kato A, Yamamoto YL, Farrokhzad S, et al. Pharmacokinetics of positron-labeled 1,3-Bis(2-chloroethyl)nitrosourea in human brain tumors using positron emission tomography. Cancer Res 44: 3120-3124, 1984. 
2. Tyler JL. Yamamoto YL, Diksic M, Theron J, Villemure JG, Worthington C, et al. Pharmacokinetics of superselective intra-arterial and intravenous [11C]BCNU evaluated by PET. J Nucl Med 27: 775-780, 1986. 
3. Ginos JZ, Cooper AL, Dhawan V. Lai JCK, Strother SC, Alcock N, et al. [13N]Cisplatin PET to assess pharmacokinetics of intra-arterial versus intravenous chemotherapy for malignant brain tumors. J Nucl Med 28: 1844-1852, 1987. 
4. Kim EE, Haynie T. Role of nuclear medicine in chemotherapy of malignant lesions. Semin Nucl Med 15: 12-20, 1985. 
5. Yang PJ, Thrall JH, Ensminger WD, Niederhuber JE, Gyves JW, Tuscan M, et al. Perfusion Scintigraphy (Tc-99m MAA) during surgery for placement of chemotherapy catheter in hepatic artery: concise communication. J Nucl Med 23: 1066-1069, 1982. 
6. Bledin AG, Kantarjian HM, Kim EE, Wallace S, Chuang VP, Patt YZ, et al. 99mTc-labeled macroaggregated albumin in intrahepatic arterial chemotherapy. Am J Roentgenol 139: 711-715, 1982. 
7. Rodari A, Bonfanti G, Garbagnati F, Marolda R, Milella M, Buraggi GL. Microsphere angiography in hapatic artery infusion for cancer. Eur J Nucl Med 6: 473-476, 1981. 
8. Sasaki Y, Imaoka S, Hasegawa Y, Nakano S. Ishikawa O, Ohigashi H, et al. Distribution of arterial blood flow in human hepatic cancer during chemotherapy-examination by short-lived 81mKr. Surgery 97: 409-413, 1985. 
9. Sigurdson ER, Ridge JA, Daly JM. Fluorodeoxyuridine uptake by human colorectal hepatic metastases after hepatic artery infusion. Surgery 100: 285-291, 1986. 
10. Ridge JA, Sigurdson ER, Daly JM. Distribution of fiuorodeoxyuridine uptake in the liver and colorectal hepatic metastases of human beings after arterial infusion. Surg Gynecol Obstet 164: 319-323, 1987. 
11. Sigurdson ER, Ridge JA, Kemeny N, Daly JM. Tumor and liver drug uptake following hepatic artery and portal vein infusion. J Clin Oncol 5: 1836-1840, 1987. 
12. Kaplan WD, Ensminger WD, Come SE. Smith EH, D'Orsi CJ, Levin DC, et al. Radionuclide angiography to predict patient response to hepatic artery chemotherapy. Cancer Treat Rep 64: 1217-1222, 1980. 
13. Daly JM, Butler J, Kemeny N, Yeh SDJ, Ridge JA. Botet J, et al. Predicting tumor response in patients with colorectal hepatic metastases. Ann Surg 202: 384-393, 1985. 
14. Lassen NA, Henriksen L, Holm S, Barry DL, Paulson OB, Vorstrup S, et al. Cerebral blood flow tomography : Xenon-133 compared with isopropylamphetamine Iodine 123 concise communication. J Nucl Med 24: 17-21, 1983. 
15. Drayer B, Jaszczak R. Friedman A, Albright R, Kung H, Greer K, et al. In vivo quantitation of regional cerebral blood flow in glioma and cerebral infarction: validation of the HIPDm-SPECT method. AJNR 4: 572-576, 1983. 
16. Winchell HS, Baldwin RM, Lin TH. Development of I-123-labeled amines for brain studies: Localization of I-123 iodophenylalkyl amines in rat brain. J Nucl Med 21 : 940-946, 1980. 
17. Winchell HS, Horst WD, Braun L, Oldendorf WH, Hattner R, Parker H, et al. N-isopropyl-[123I]p-iodoamphetamine: Single-pass brain uptake and washout; Binding to brain synaptosomes; and localization in dog and monkey brain. J Nucl Med 21 : 947-952, 1980. 
18. Kuhl DE, Barrio JR, Huang SC, Selin C. Ackermann RF, Lear JJ, et al. Quantifying local cerebral blood flow by N-isopropyl-p-[123I]iodoamphetamine (IMP) tomography. J Nucl Med 23: 196-203, 1982. 
19. Holman BL, Lee RGL, Hill TC, Lovett RD, Lister-James J. A comparison of two cerebral perfusion tracers. N-isopropyl I-123 p-iodoamphetamine and I-123 HIPDM, in the human. J Nucl Med 25: 25-30, 1984. 
20. Moretti JL. Cinotti L, Cesaro P, Defer G, Joulin Y, Sergent A, et al. Amines for brain tomoscintigraphy. Nucl Med Commun 8: 581-595, 1987. 
21. Stagg RJ, Lewis BJ, Friedman MA. Ignoffo RJ, Hohn DC. Hepatic arterial chemotherapy for colorectal cancer metastatic to the liver. Ann Intern Med 100: 736-743, 1984. 
22. Andrews JC, Williams DM, Cho KJ, Knol JA, Wahl RL, Ensminger WD. Unsatisfactory hepatic perfusion after placement of an implanted pump and catheter system: angiographic correlation. Radiology 173: 779-781, 1989. 
23. Cho KJ, Andrews JC, Williams DM, Doenz F, Guy GE. Hepatic arterial chemotherapy: role of angiography. Radiology 173: 783-791, 1989. 
24. Charnsangavej C, Carrasco CH, Wallace S, Richli W, Haynie TP. Hepatic arterial flow distribution with hepatic neoplasms: significance in infusion chemotherapy. Radiology 165: 71-73, 1987. 
25. Lutz RJ, Miller DL. Mixing studies during hepatic arterial infusion in an in vitro model. Cancer 62: 1066-1073, 1988. 
26. Bookstein JJ. Cho KJ. Davis GB, Dail D. Arterioportal communications: observations and hypotheses concerning transsinusoidal and transvasal types. Radiology 142: 581-590, 1982. 
27. Fishman AP, Fisher AB, Geiger SR. The respiratory system. Handbook of Physiology. Vol, 1. Sec. 3. Bethesda: American physiological society, 1985. 
28. Nakajo M, Kobayashi H, Shimabukuro K, Shirono K, Sakata H, Taguchi M, et al. Biodistribution and in vivo kinetics of iodine-131 lipiodol infused via the hepatic artery of patients with hepatic cancer. J Nucl Med 29: 1066-1077, 1988. 
29. Raoul JL, Bourguet P, Bretagne JF, Duvauferrier R, Coornaert S, Darnault P, et al. Hepatic artery injection of I-131-labeled lipiodol. Radiology 168: 541-545, 1988. 
30. Gyves JW, Ziessman HA, Ensminger WD, Thrall JH, Niederhuber JE, Keyes JW. Definition of hepatic tumor microcirculation by single photon emission computerized tomography (SPECT). J Nucl Med 25: 972-977, 1984. 
31. Rahimian J, Glass EC, Touya JJ, Akber SF, Graham LS, Bennett LR. Measurement of metabolic extraction of tracers in the lung using a multiple indicator dilution technique. J Nucl Med 25: 31-37, 1984. 
32. Touya JJ, Rahimian J, Corbus HF, Grubbs DE, Savala KM, Glass EC, et al. The lung as a metabolic organ. Semin Nucl Med 16: 296-305, 1986. 
33. Nakajo M, Shimada J. Shimozono M. Uchiyama N. Hiraki Y, Shinohara S. Serial lung imaging with 123I-IMP in localized pulmonary lesions. KAKU IGAKU (Jpn J Nucl Med) 25: 441-450, 1988. 
34. Suga K, Matsumoto T, Nakanishi T, Yokoyama T, Tanaka K, Nakamura H, et al. Clinical study on the mechanism of abnormal accumulation in lung scanning with N-isopropyl-p-123I-iodoamphetamine (123I-IMP). KAKU IGAKU (Jpn J Nucl Med) 25: 625-631, 1988. 
35. Kosuda S, Kawahara S, Tamura K, Ishikawa N, Ono A, Kubo A, et al. N-isopropyl-p-[I-123]iodoamphetamine lung imaging in a patient with chronic pulmonary thromboembolism. Clin Nucl Med 14: 756-758, 1989. 
36. Herba MJ, Illescas FF, Thirlwell MP, Boos GJ, Rosenthall L, Atri M, et al. Hepatic malignancies: improved treatment with intraarterial Y-90. Radiology 169: 311-314, 1988.