ORIGINAL ARTICLE 

Annals of Nuclear Medicine Vol. 14, No. 6, 477-483, 2000 

Nontumorous decrease in Tc-99m GSA accumulation 

Shiro AKAKI, Susumu KANAZAWA, Masatoshi TSUNODA, Yoshihiro OKUMURA, Izumi TOGAMI, Masahiro KURODA, Yoshihiro TAKEDA, and Yoshio HIRAKI 

Department of Radiology. Okayama University Medical School 

Nontumorous decreasc in 99mTc-GSA accumulation has not been well covered in the literature. Understanding of this phenomenon is, however, essential for accurate evaluation of regional hepatic function. Scintigrams (transaxial SPECT) of 269 patients who underwent 99mTc-GSA liver scintigraphy were reviewed for the presence of nontumorous decreases in 99mTc-GSA accumulation. Nontumorous decreases in 99mTc-GSA accumulation were seen in 32 of 269 patients (12%). In 16 of the 32 patients (6%), nontumorous decreases in 99mTc-GSA accumulation corresponded to regional decrease in portal venous flow. The causes of such decrease in portal venous flow were portal thrombus of hepatocellular carcinomas in eight patients, portal venous stenosis or occlusion by hilar cholangiocarcinomas in five patients, inter alia. In eight patients (3%), the regions with decreased 99mTc-GSA accumulation correlated with massive hepatic necrosis in fulminant hepatitis, scar in hepatitis, or confluent fibrosis in cirrhotic liver. In two paticnts (0.7%) with hilar cholangiocarcinomas, the possible causes of lobar decrease in 99mTc-GSA accumulation were thought to he lobar decrease in portal venous flow, lobar biliary stasis, or both. In four patients (1.5%), the exact causes of nontumorous decrease in 99mTc-GSA accumulation could not be determined. 

Key words: liver, technetium-99m GSA, portal venous flow decrease 


INTRODUCTION 

Technetium-99m diethylenetriamine-penta-acetic acidgalactosyl human serum albumin (99mTc-GSA) Is an analog ligand to asialoglycoprotein receptor which is used widely for radionuclide imaging of the liver. Uptake of 99mTc-GSA depends on the presence of normally-functioning hepatocytes.1-3 Most hepatic neoplasms apart from some tumor-like conditions such as focal nodular hyperplasia4 show an accumulation defect on hepatocytceoriented scintigraphy.5 We previously studied 99mTc-GSA liver scintigraphy in cases with hepatic regional attenuation/signal intensity differences on CT/MRI.6,7 The study revealed that nontumorous decreases in 99mTC-GSA accumulation could occur in cases of decrease in regional portal venous flow and confluent hepatic fibrosis, 

 Received May 1, 2000, revision accepted October 16, 2000. 
 For reprint contact: Shiro Akaki. M.D., Department of Radiology, Okayama University Medical School. 2-5-1 Shikatacho, Okayama 700-8558, JAPAN. 
 E-mail: akaki@cc.okayama-u.ac.jp 

etc. These decreases were not situated in the tumor itself, but situated in the hepatic parenchyma. A case of localized decrease in 99mTc-GSA accumulation due to segmental biliary obstruction was also reported,8 The incidence of these nontumorous decrcases in 99mTc-GSA accumulation and the relative percentages of the various causes are, however, unknown. The aim of this study is to evaluate the possible conditions causing such nontumorous decreases in 99mTc-GSA accumulation, to clarify thc incidences of them, and to calculate the relative percentages of the various causes. 


MATERIALS AND METHODS 

Imaging Techniques 
The patients were examined in the supine position with a dual-head rotating gamma camera interfaced to a minicompuler (GCA72OOA/DI: Toshiha, Tokyo, Japan). A parallel-hole, low-energy, high-resolution collimator was used. Immediately after intravenous admlnistration of 185 MBq of 99mTc-GSA (Nihon Medi-Physics, Nishinomiya, Japan), dynamic images were acquired in 15-sec frames for 30 min. Next, single-photon emission computed tomography (SPECT) was performed after 35-45 min by the acquisition or 60 projectional images over 360deg. in a 128x128 matrix. Spatial resolution was calculated as 10 mm. The data obtained were reconstructed for transaxial sections 6.9 mm thick. Coronal and sagittal reconstruction was also performed in some more recent cases. A Butterworth filter was used for pre-reconstruction, and final reconstruction was performed with a Ramp filter. A Chang attenuation correction was applied. 

Evaluation 
Four hundred and thirty 99mTc-GSA liver scintlgraphies were done at our institution over about four years to evaluate hepatic functional reserve. In this series, 63 patients underwent repeated studies (twice in 46 patients, three times in 13, four times in three, and five times in one), so that 345 patients underwent 99mTc-GSA liver scintigraphy at least once during this period. Of these 345 patients, 76 were excluded from this study because their 99mTc-GSA liver scintigrams and/or CT/MRI were not available and therefore could not be evaluated. Two radiologists experienced in radionuclide and liver imaging evaluated the images of the remaining 269 patients. These were 182 men and 87 women aged 19-86 years old. (mean: 57 years). Of these 269 patients, 130 had hepatocellular carcinomas, 46 underwent 99mTc-GSA liver scintigraphy to evaluate live, cirrhosis, and 19 had cholangiocarcinomas. 
 The readers first reviewed the CT/MRI of the 269 patients and recognized the contour of the liver, and masses when such existed. They then reviewed 99mTc-GSA SPECT for the presence of nontumorous decreases in 99mTc-GSA accumulation.6 In patients with repeated 99mTc-GSA scans, abnormality was defined to exist when nontumorous decrease in 99mTc-GSA accumulation was seen in at least one study. The site, size and relationship with hepatic masses were recorded. Causes of these nontumorous decreases in 99mTc-GSA accumulation were clarified as far as possible by reviewing CT/MRI and/or angiograms and by checking surgical records. The incidence of each cause was calculated and variously compared. 


RESULTS 

Nontumorous decreases in 99mTc-GSA accumulation were seen in 32 of 269 patients (12%). The distribution of the decreases were lobar in 13 (2 in the right lobe and 11 in the left lobe), segmental in 6, subsegmental or less in 7, and non-segmental in 6 patients. In 16 patients (6%), nontumorous decreases in 99mTc-GSA accumulation corresponded to regional decrease in portal venous flow (Fig. 1). Portal vein stenosis or occlusion was seen on angiography in 12 of the 16 patients and confirmed on surgery in two patients. In one patient with idiopathic portal hypertension, portal vein stenosis was not seen, but portal perfusion decrease was seen in the periphery of the liver.7 In two patients, portal vein stenosis was recognized on postcontrast CT. In one patient, portal vein stenosis was confirmed by contrast-enhanced MR portography. 
 The causes of such a decrease in portal venous flow were portal thrombus of hepatocellular carcinomas in eight patients, portal venous stenosis or occlusion by hilar cholangiocarcinomas in five, portal venous stenosis by hepatic abscess in one, portal venous thrombosis in one, and there was one patient with idiopathic portal hypertension.7 Most regions with a decrease in portal venous flow were hypodense on precontrast CT, hypointense on T1-weighted MR images, and hyperintense on T2-weighted images. Most of them had regional enhancement in the arterial phase of dynamic CT/MRI. 
 In eight patients (3%), the regions with decreased 99mTc-GSA accumulation were wedge-shaped or band-like in the periphery of the liver. They were hypoattenuating on precontrast CT, hypointense on T1-weighted MR images, and hyperintense on T2-weighted images. Of these eight patients. six had suffered from advanced liver cirrhosis (Fig. 2). one had developed fulminant hepatitis (Fig. 3), and one had suffered from autoimmune hepatitis. In one patient with advanced cirrhosis, confluent fibrosis was confirmed during surgery for hepatocellular carcinoma. 
 In two patients (0.7%) with hilar cholangiocarcinomas, lobar decrease in 99mTc-GSA accumulation was seen. Ipsilateral portal vein obstruction was seen on angiography. On CT, the bile duct was dilated in both hepatic lobes but predominant in the lobe with decreased 99mTc-GSA accumulation. In one patient (0.4%), with intrahepatic stones, bile duct dilatation was seen in the left lobe. In one patient (0.4%) with hepatocellular carcinoma, 99mTc-GSA accumulation was decreased in the right lobe, and both the right portal vein and the right hepatic vein were occluded. In this case, the cause was considered to be a decrease in portal venous flow or occlusion of the hepatic vein or both (Fig. 4). Figures 1-4 show representative cases of each cause. In four patients (1.5%), the causes of nontumorous decrease in 99mTc-GSA accumulation could not be determined. Of these four, one patient with a decrease in 99mTc-GSA accumulation in the left lobe had hepatocellular carcinomas, but there was no evidence of a decrease in the portal venous flow in the left lobe. Two patients had hepatocellular carcinomas and metastatic tumors, respectively. Decreases in 99mTc-GSA accumulation were seen in the peripheral hepatic parenchyma to the tumors. Portal vein or hepatic vein stenosis or occlusion by the tumors was the suspected cause. This, however, could not be confirmed. One patient with liver cirrhosis had a decrease in 99mTc-GSA accumulation in the left lobe and right anterior segment. The cause of this decrease could not be determined. 


DISCUSSION 

Nontumorous decrease in 99mTc-GSA accumulation has been poorly covered in the literature. Understanding of this phenomenon is, however, important for evaluating regional hepatic reserve. In order to properly evaluate the phenomenon, all serial cases, including those with homogeneous 99mTc-GSA accumulation needed to be reviewed. This we did for all available cases undergoing 99mTc-GSA liver scintigraphy over a period of about four years. In this study, the incidence of nontumorous decrease in 99mTc-GSA accumulation proved to be surprisingly high (12%). These nontumorous decreases in 99mTc-GSA accumulation can often be misdiagnosed for hepatic masses on the scintigrams alone. Moreover, such inhomogeneity of 99mTc-GSA accumulalion directly suggests inhomogeneity of regional hepatic function and this indicates some possible problems when dealing with hepatic diseases. 
 In the first place, these inhomogeneities may cause a discrepancy between hepatic reserve evaluation by 99mTc-GSA and by other anatomical methods such as CT volumetry.9 CT cannot express regional function or functional density of the liver, because it estimates regions of poor function as normal parenchymal volume and may overestimate these regions Moreover, to know these functional inhomogeneities of the liver is essential in planning therapy for hepatic disease. In cases with biliary stasis, percutaneous biliary decompression or stent insertion should be performed preferentially to save a region or lobe rich in 99mTc-GSA accumulation which retains better function than other regions or lobes. In planning surgery too, such functional inhomogeneity needs careful consideration. Lobectomy which resects a lobe with poor function and ipsilateral to the mass may have few serious consequences. When extended lobectomy is planned, on the other hand, one should be careful because it resects part of a contralateral lobe which still functions well. 
 In this study, decrease in portal venous flow was defined as existing when stenosis or obstruction of a portal branch was seen on angiograms or enhanced CT. Biliary stasis was defined as existing when dilatation of the intrahepatic bile duct was recognized on CT/MR images at that time or before percutaneous drainage. When dilatation of the intrahepatic bile duct is obviously predominant in a region, it has been established that biliary stasis in the region was more severe than in other regions, so it became a candidate cause for a relative decrease in 99mTc-GSA accumulation. In two cases, the causes of a decrease in 99mTc-GSA accumulation were thought to be a lobar decrease in portal venous flow, lobar biliary stasis, or both. As for the other pathological conditions, each typical finding on CT/MRI previously reported10-14 was referred to. 
 A decrease in portal venous flow accounted for half of the causes of nontumorous decrease in 99mTc-GSA accumulation. In evaluating regional attenuation/signal intensity differences on CT/MRI, we previously reported regional decrease in 99mTc-GSA accumulation due to a decrease the portal venous flow.6,7 This was thought to correlate with the phenomenon where hepatic parenchyma with decreased portal venous flow becomes hypoattenuating on precontrast CT, hypointense on T1-weighted MR images, and hyperintense on T2-weighted images. Edema, hepatocyte depletion, and fibrosis often occur in regions of decreased portal venous flow. As a result, hepatic attenuation/intensity abnormalities occur on CT/MRI. Normally-functioning hepatocytes may decrease due to these histological changes and 99mTc-GSA accumulation in the region may also fall.6 We previously evaluated cases where both lobar decrease in portal venous flow and lobar biliary stasis existed. Hilar cholangiocarcinomas ooften involve both the portal vein and the bile duct. In such cholangiocarcinomas, lobar decrease in 99mTc-GSA accumulation correlated better with the decrease in portal venous flow than with the severity of biliary stasis.15 These phenomena resembled the lobar atrophy following obstruction of the ipsilateral portal vein from hilar cholangiocarcinoma which was reported by Takayasu et al.16 and Hann et al.17 
 The interrelationship of massive hepatic necrosis in fulminant hepatitis,10 scar in hepatitis of insidious onset and/or without definite coma,11 and confluent hepatitis in cirrhotic liver12,13 has not been properly clarified. They do, however, appear similar on pre- and postcontrast CT and MRI.11 In this study, we dealt with them as a single pathological condition. In these regions, decreases in 99mTc-GSA accumulation were seen, and making up one fourth of all causes of nontumorous decrease in 99mTc-GSA accumulation, they were the second most frequent cause (after portal venous flow decrease). The number of hepatocytes tends to manifest a sharp decrease in these necrotic or scar-like lesions. In the same manner as on CT/MRI, these lesions can be misdiagnosed as neoplasms. Knowledge of CT/MRI findings typical of these conditions and any clinical history of fulminant hepatitis or long-term cirrhosis is essential. In the case of fulminant hepatitis, a decrease in 99mTc-GSA accumulation was seen in the region of massive necrosis at a relatively early stage when precontrast CT was unable to show any abnormality. This appears to suggest that 99mTc-GSA liver scintigraphy is better able to predict at an earlier stage those regions which might develop massive hepatic necrosis. 
 In one case 99mTc-GSA accumulation was decreased in the left lobe and the cause was thought to be lobar biliary stasis. In the patient with hepatocellular carcinoma, 99mTc-GSA accumulation was decreased in the right lobe, and both the right portal vein and the right hepatic vein were occluded. The cause of the decrease was considered to be right portal or right hepatic vein occlusion or both. Differences in the regions on CT which correspond with the right portal vein and right hepatic vein were well described by Murata et al.18 In our case, however, deformity of the liver by hepatocellular carcinoma and the limited spatial resolution of SPECT made it difficult to distinguish between the two regions, and closer determination of the possible cause proved impossible. 
 Several limitations were perceived with this study. First, we failed to review 76 of 345 patients due to the unavailability of CT/MRI or 99mTc-GSA scintigrams. This did not seem to produce any significant bias, however, in patient population or proportion. During the study, we confined ourselves to introducing the basic principle of hepatic functional inhomogeneity expressed by 99mTc-GSA. More accurate prediction of postoperative hepatic reserve and any safe extension of hepatectomy indication with regard to these phenomena need to be further evaluated and discussed. 
 In conclusion, we reviewed 99mTc-GSA scintigrams of a series of 269 patients and studied the incidence and causes of nontumorous decrease in 99mTc-GSA accumulation, a surprisingly not uncommon occurrence. Half of these decreases correlated with regional decrease in the portal venous flow. 


REFERENCES 

1. Kawa S, Tanaka Y. A quantitative model of technetium-99m-DTPA-galactosyl-HSA for the assessment of hepatic binding receptor. J Nucl Med 32: 2233-2240, 1991. 
2. Koizumi K, Uchiyama G, Arai T, Ainoda T, Yoda Y. A new liver functional study using Tc-99m DTPA-galactosyl human serum albumin: evaluation of the validity of several functional parameters. Ann Nucl Med 6: 83-87, 1992. 
3. Kudo M, Todo A, Ikekubo K, Hino M. Receptor index via asialoglycoprotein receptor imaging: correlation with chronic hepatocellular damage. Am J Gastroenterol 87: 865-870, 1992. 
4. Arbab AS, Koizumi K, Hiraike S, Arai T, Mera K, Ohtomo K, et al. Discordant uptake of three different agents for liver scintigraphy in a patient with focal nodular hyperplasia. Clin Nucl Med 20: 937-939, 1995. 
5. Stadalnic RC, Vera DR, Woodle ES, Trudeau WL, Porter BA, Ward RE, et al. Technetium-99m NGA functional hepatic imaging: preliminary clinical experience. J Nucl Med 26: 1233-1242, 1985. 
6. Akaki S, Mitsumori A, Kanazawa S, Togami I, Takeda Y, Joja I, et al. Technetium-99m-DTPA-galactosyl human serum albumin liver scintigraphy evaluation of regional CT/MRI attenuation/signal intensity differences. J Nucl Med 39: 529-532, 1998. 
7. Akaki S, Mitsumori A, Kanazawa S, Takeda Y. Joja I, Hiraki Y, et al. Reduced radioactivity in the periphery of the liver in a patient with idiopathic portal hypertension. Clin Nucl Med 22: 369-371, 1997. 
8. Inoue Y, Machida K, Honda N, Mamiya T, Takahashi T. Kamano T, et al. Impaired hepatic function in segmental biliary obstruction demonstrated with a receptor-binding radiotracer Ann Nucl Med 8: 209-212, 1994. 
9. Mitsumori A, Nagaya I, Kimoto S, Akaki S, Togami I, Takeda Y, et al. Preoperative evaluation of hepatic functional reserve following hepatectomy by technetium-99m galactosyl human serum albumin liver scintigraphy and computed tomography. Eur J Nucl Med 25: 1377-1382, 1998. 
10. Murakami T, Baron RL, Peterson MS. Liver necrosis and regeneration after fulminant hepatitis: pathologic correlation with CT and MR findings. Radiology 198: 239-242 1996. 
11. Itai Y, Ohtomo K, Kokubo T, Minami M, Yoshida H. CT and MR imaging of postnecrotic liver scars. J Comput Assist Tomogr 12: 971-975, 1988. 
12. Ohtomo K, Baron RL, Dodd GD III, Federle MP, Miller WJ, Campbell WL, et al. Confluent hepatic fibrosis in advanced cirrhosis: appearance at CT. Radiology 188: 31-35, 1993. 
13. Ohtomo K, Baron RL, Dodd GD III, Federle MP, Ohtomo Y, Confer SR. Confluent hepatic fibrosis in advanced cirrhosis: evaluation with MR imaging. Radiology 189: 871-874, 1993. 
14. Itai Y, Sekiyama K, Ahmadi T, Obuchi M, Yoshida M. Fulminant hepatic failure: observation with scrial CT. Radiology 202: 379-382. 1997. 
15. Akaki S, Mitsumori A, Kanazawa S, Togami I, Takeda Y, Hiraki Y. Lobar decrease in 99mTc-GSA accumulation in hilar cholangiocarcinoma. J Nucl Med 40: 394-398, 1999. 
16. Takaryasu K, Muramatsu Y, Shima Y, Moriyama N, Yamada T, Makuuchi M. Hepatic lobar atrophy following obstruction of the ipsilateral portal vein from hilar cholangiocarcinoma. Radiology 160: 389-393, 1986. 
17. Hann LE, Getrajdman GI, Brown KT, Bach AM, Teitcher JB, Fong Y, et al. Hepatic lobar atrophy: association with ipsilateral portal vein obstruction. AJR 167: 1017-1021, 1996. 
18. Murata S, Itai Y, Asato M, Kobayashi H, Nakajima K, Eguchi N, et al. Effect of temporary occlusion of the hepatic vein on dual blood supply in the liver: evaluation with spiral CT. Radiology 197: 351-356, 1995.