Liver, 1990: 10, 72-78 Key words: hepatocellular carcinoma; 1-131 Lipiodol; Lipi- odol.

Distribution of Lipiodol in hepatocellular carcinoma

CHANIL PARK, SO0 IM CHOI, HOGUEN KIM, HYUNG SIK YOO' AND YO0 BOCK LEE

Department of Pathology and 'Radiology, Yonsei University College of Medicine, Seoul, Korea

ABSTRACT ~ Intrahepatic distribution of Lipiodol and 1-131 Lipiodol infused via the hepatic arteries was evaluated in six patients with HCC who had undergone hepatic lobectomy or segmentectomy. CT scan and gamma camera radiograph confirmed that the oily contrast material or 1-131 radioactivity accumulated selectively in the tumor over a long period. One to two thirds of the tumor mass appeared necrotic, although the extent tended to be larger in the case of radioactive Lipiodol infusion. The tumor cells contained numerous lipid globules within the cytoplasm. Also, oil red 0 stain demonstrated that the individual tumor cells had non-globular lipid on their surface. In conclusion, Lipiodol leaks out of the vascular spaces to attach to the cancer cell membrane as a non-globular lipid as well as to enter the cancer cells as a globular lipid. This phenomenon specific to cancer cells suggests a biochemical membrane change which may have occurred during carcinogenesis, causing alteration of mem- brane transport and cell death.

Accepted for publication I0 May 1989

Lipiodol is an iodinated ethyl ester of the poppy- seed oil fatty acid. Since it contains 37-3870 iodine by weight, it has been used as a contrast material. Once injected into the hepatic arteries of patients with primary hepatocellular carcinoma (HCC), Lipiodol concentrates preferentially within the HCC and is retained for longer periods in the tumor than in the non-neoplastic tissue (14 ) . Using this property, the identification and locali- zation of HCC can be achieved very precisely. Recent studies have attempted to utilize Lipiodol as a carrier vehicle for radioactive iodine (410) by substituting the 1-127 of Lipiodol with 1-131, or for chemotherapeutic agents (1 1-14).

Although there are reports dealing with the biodistribution of intra-arterially administered Li- piodol in animals (15, 16) and in man with or without HCC (4), the mechanism by which Lipi- odol accumulates selectively in HCC over a long

period has not been completely explained. The aim of this study was to clarify the histological distribution pattern of Lipiodol in HCC, which might yield reasons for both the selective concen- tration and the long-term retention of Lipiodol in HCC.

Patients and methods Patients The material consisted of the hepatic tissues of six pa- tients with primary HCC, who underwent hepatic lobec- tomy or segmentectomy. The diagnosis of HCC was made through clinical manifestations, serum AFP, and various imaging examinations including angiography, ul- trasonography and CT scan. Two to 12 weeks prior to surgery, all patients had Lipiodol or 1-131 Lipiodol in- fused into the hepatic artery or its branch via the femoral artery in a dose of 4 to 5 ml. The arteries selected for infusion, the intervals between Lipiodol or 1-131

DISTRIBUTION OF LlPlODOL IN HCC 73

Table 1 Protocol for Lipiodol and 1-131 Lipiodol infusion applied to the patients with HCC

Dose of

Case Age/sex Selected artery Interval Lipiodol I- 13 1 I 37/M Rt. anterosuperior segmental branch 3 wks 5 c c 2 34/M Posteroinferior segmental branch 12 wks 5 c c 3 60/M Hepatic artery 2wks 4 c c 5mCi 4 58/M Posteroinferior segmental branch 4 wks 5 cc 10mCi 5 46/M Posteroinferior segmental branch 4 wks 5 c c 10mCi 6 51/M Rt. hepatic artery 12 wks 4 c c 30mCi

Interval: Interval between infusion and operation; Rt: Right.

Lipiodol infusion to surgery, and the doses of Lipiodol and 1-131 radioactivity are summarized in Table 1.

An overview of the major clinical features of the pa- tients was carried out. Some patients underwent CT scans before Lipiodol or 1-1 3 I Lipiodol administration and just prior to surgery. In cases which were given I- 131 Lipiodol, the images were also obtained by a gamma camera to examine the intrahepatic distribution of the 1-13] radioactivity.

Pathologic study The resected hepatic tissues were cut into I-cm-thick slices. One slice was frozen immediately and the others were immersed in 10% neutral formalin. Every slice which passed through the center of the global mass was

formol-fixed blocks for hematoxylin-eosin stain. Areas of necrosis were marked directly onto the glass slides as accurately as possible. The areas of the necrotic portion were mapped and the total areas of viable and necrotic tumor tissues of the slice were measured. Sections were obtained from the frozen slice for oil red 0 stain. Light microscopic examination was performed with particular concern for the distribution of Lipiodol in HCC and in cirrhotic nodules.

Results The Lipiodol was retained in the tumor for a long time in all cases. This was confirmed by CT scan, even in the cases which were followed up for 3 months postinfusion (Fig. 1). CT scan after Lipi- cut into 3 x 1 cm blocks, after the exac; topography of .

the slice with the tumor had been sketched On trans- parent paper for mapping. Sections were made of the

o d d Or 1-131 Lipiodol infUSiOn delineated the tumor nodules as areas of markedly high density

Fig. 1. CT scans taken before (left) and 1 month following (right) Lipiodol infusion. Retention of Lipiodol is confined to the mass (right), which is smaller than that of preinfusion (left).

74 PARKETAL.

Fig. 2. A gamma camera radiograph taken 32 days after 1-131 Lipiodol infusion (right). The radioactivity is confined to the area corresponding to the tumor seen on the CT scan (left).

because of its selective accumulation. Gamma camera radiograph obtained after 1-1 3 1 Lipiodol infusion demonstrated the radioactivity confined exactly to the areas corresponding to the tumor (Fig. 2).

The patients were all men with ages ranging from 34 to 60 years. Serum HBsAg was positive in all patients. Table 2 summarizes the interval changes of AFP and tumor size and the percent necrosis of the tumor. The serum AFP was elev- ated initially in two of the six patients (Cases 2 and 6). It was reduced markedly following infusion of Lipiodol or I- 13 1 Lipiodol and further decreased to the upper normal level after surgical resection of the tumor.

All tumors in the patients were of the expansive type without discrete satellite nodules. They meas-

ured 2.0 to 4.5 cm in the largest diameter at oper- ation. When compared to the tumor sizes meas- ured on CT scan taken at the time of infusion, they had decreased more or less during the inter- val periods, except in Cases 4 and 5. The changes in tumor size of the cases with Lipiodol infusion were not different from those with 1-131 Lipiodol infusion. The remaining hepatic tissues were cir- rhotic in all cases.

Histologic findings A very large proportion, one third to two thirds, of these small HCCs were found to be necrotic. The extent of tumor necrosis was larger in the cases with 1-131 Lipiodol infusion than those with Lipiodol infusion, but did not correlate well with

Table 2 Interval changes of serum AFP and tumor size following infusion of Lipiodol or 1-131 Lipiodol and the extent of necrosis

Serum AFP (ng) Tumor size (cm) % Necrosis Case Pre-inf./Post-inf./Post-op. at inf./at op. of tumor

~ ~~

1 Negative 4 x 4 / 3 x 2 38 2 200/27/10; 3 4.5 x 4.514.0 x 3.5 34 3 Negative 2.5 x 1.512.0 x 1.5 0 4 Negative 2.0 x 1.512.5 x 2.0 65 5 Negative 4.0 x 4.014.5 x 4.0 63 6 12,800/3,200; 5418; 3 3.8 x 3.512.0 x 0.8 65

Inf.: infusion; Op.: operation.

DISTRIBUTION OF LlPlODOL IN HCC 75

the intervals between infusion and surgical resec- tion. Many of the necrotic areas were gradually replaced by granulation tissue (Fig. 3), and some seemed to be shrunken by fibrous tissue prolifer- ation.

The HCCs had a giant trabecular pattern with or without areas of pseudoacinar structures or solid sheets of tumor cells. The tumor cells had numerous lipid globules within the cytoplasm (Fig. 4). In some HCC lobules, particularly in Case 6, the tumor cells had degenerated, the cyto- plasm was cleared, sometimes with an acidophilic glassy substance pushed to the side of the cell. Only a few lipid globules were present in cirrhotic nodules and in the connective tissue septa. In these areas, the lipid globules were larger than those seen in the tumor cells and occasionally led to

nodules, which compressed the nearby nodules or revealed dysplastic nuclear changes, had more intracellular lipid globules than the other nodules

granulomas o f a foreign body type, Some cirrhotic Fig. 4. HCC oftrabiXXllar type showing numerous intra- cytoplasmic lipid vacuoles. H-E x 200.

(Fig. 5). In addition to the positive staining of the intra-

cytoplasmic globular lipid, oil red 0 for lipid dem- onstrated that the individual tumor cells were en- veloped by a non-globular lipid, giving the HCC lobules a honeycomb pattern (Fig. 6, 7), which

sections. No such finding was seen in the non- neoplastic hepatocytes of the cirrhotic nodules. The sinusoids and blood vessels, on the other hand, were patent and almost free of globular or non-globular lipid.

was unexpected in the hematoxylin-eosin stained

Fig. 3. A necrotic focus surrounded by granulation tis- sue. Note the extracellular lipid globules of various sizes. H-E ~ 1 0 0 .

Fig. 5. Parts of two cirrhotic nodules showing lipid vacu- oles only in the left one, where nuclear dysplasia is apparent. H-E x 200.

76 PARKETAL.

Discussion Iodinized oil, such as Lipiodol and Ethiodol, has been known to be retained selectively in HCC for a long time following intraarterial hepatic infusion (14) , and has been successfully used for the detec- tion of small-sized HCC and daughter nodules ( I , 3) . In this study, Lipiodol was retained in the tumor as long as 3 months postinfusion. Gamma camera radiographs and/or CT scans taken im- mediately following infusion or at the time of surgery revealed that the Lipiodol or 1-1 3 1 radio- activity was selectively accumulated within HCCs. These findings agree with previous reports and indicate that radioiodinated Lipiodol is also re- tained in HCC, like non-radioiodinated Lipiodol. Using this property, Lipiodol is now in use as a carrier vehicle of lipophilic anticancer drugs or of radioactive iodine (4-14, 17, 18).

The clinical results appeared to be quite promis- ing, particularly in expanding type HCC. In this type of HCC, more than 70% of the cases showed good clinical responses. This is supported by the present cases in which a dramatic decrease of serum AFP was noted as well as a decrease of tumor size with remarkable necrosis even in the small HCCs. It remains to be explained why, in this study, non-radioiodinated Lipiodol also had

a therapeutic effect, as shown in the changes of serum AFP and tumor necrosis.

In spite of this promising result of Lipiodol in the diagnosis and treatment of HCC, the mechan- isms of the selective retention of contrast material within the tumor and of the intracellular accumu- lation that might be required for the selective drug delivery into the tumor cells are still obscure.

The siphoning effect that is created by the increased blood flow in such hypervascular tu- mors as HCC (3) and the increased permeability of tumor vessels by a permeability enhancing factor which might be released from tumor cells (1 9) have been proposed to explain the selective concentration of Lipiodol in the tumor. The tor- tuous and irregular neoplastic vessels (1) and the delayed wash-out of Lipiodol from HCC due to the underdevelopment of local lymphatics (2) have been considered to be responsible for the prolonged retention. The decreased number of Kupffer cells in HCC (20) may affect the clear- ance of Lipiodol.

On the basis of the finding in this study, that Lipiodol attached to the tumor cell membrane as a non-globular lipid, the authors propose another attractive mechanism for the prolonged retention. Considering that the blood vessels and sinusoids

Fig. 6 . A fat-stained section showing dark linear lipid along the tumor cell membrane and globular lipid in the cytoplasm. Oil red 0 x 200.

Fig. 7. A high power view of HCC. Note the non- globular lipid along the cell membrane, giving a honey- comb appearance. Oil red 0 x 400.

DISTRIBUTION OF LlPlODOL IN HCC 77

were almost free from stainable lipid in this inves- tigation, Lipiodol seems to be washed away rap- idly by the blood stream. When Lipiodol leaks out of the blood vessels and sinusoids of the tumor, it will be retained for a long time due to the underdeveloped lymphatics of the tumor. The membrane attachment of Lipiodol should occur from the beginning of its leaking out. The attach- ment of Lipiodol along the cell membrane of HCC should play a role in its further retention, and should compromise the membrane transport of tumor cells, altering the intracellular environment, followed by the degeneration and death of cancer cells. This could be the reason why the injection of non-radioiodinated Lipiodol also caused tumor necrosis in the present cases. These findings sug- gest that the liver cell membrane must have been transformed biochemically to become lipophilic during cancerous changes.

Since Lipiodol is a macromolecule, it is very unlikely that Lipiodol enters the tumor cells of HCC. There must be certain modifications of the Lipiodol molecule at the cancer cell membrane, or changes in the cancer cell membrane making itself liable to pinocytosis of the macromolecule. Since there is no proof that the intracellular lipid droplets are derived from Lipiodol, another possi- bility arises: they may be a product of altered metabolism of hepatocytes during cancerous transformation. The mechanism of the intracellu- lar entrance of Lipiodol to non-neoplatic hepato- cytes has yet to be elucidated.

References 1.

2.

3.

4.

5.

NAKAKUMA K, TASHIRO S, HIRAOKA K, OGATA K, OOTSUKA K. Hepatocellular carcinoma and meta- static cancer detected by iodized oil. Radiology 1985:

YUMOTO Y, JINNO K, TOKUYAMA K et al. Hepatocel- Mar carcinoma detected by iodized oil. Radiology

OHISHI H, UCHIDA H, YOSHIMURA H et al. Hepato- cellular carcinoma detected by iodized oil: use of anticancer agents. Radiology 1985: 154: 25-29. MADSEN M T, PARK C H, THAKUR M L. Dosimetry of iodine-131 Ethiodol in the treatment of hepa- toma. J Nucl Med 1988: 29: 1083-1044. BRETAGNE J F, BOURGUET P, RAOUL J L et al. Biodis- tribution of 131 labelled iodized oil after injection into the hepatic artery: preliminary results in pa-

154: 15-17.

1985: 154: 19-24.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

tients with hepatocellular carcinoma. Dig Dis Sci 1986: 31: 406s. KOBAYASHI H, HIDAKA Y, KAJIYA Y et al. Treatment of hepatocellular carcinoma by transarterial injec- tion of anticancer agents in iodized oil suspension or of radioactive iodized oil solution. Acfa Radio1 Diagn 1986 27: 139-147. PARK C H, SUH J H, Yo0 H S, LEE J T, KIM D I. Evaluation of intrahepatic 1-131 Ethiodol on a pa- tient with hepatocellular carcinoma - therapeutic feasibility. CIin Nucl Med 1986: 11: 514-517. Yo0 H S, SUH J H, LEE J T et al. Therapeutic trial of intrahepatic 1-131 Lipiodol on patients with hepatocellular carcinoma. Korean J Gastroenterol

BRETAGNE J F, RAOUL J L, BOURGUET P et al. Hepa- tic artery injection of I-131-labeled Lipiodol. Part [I. Preliminary results of therapeutic use in patients with hepatocellular carcinoma and liver metastases. Radiology 1988: 168: 547-550. RAOUL J L, BOURCUET P, BRETAGNE J F et al. Hepa- tic artery injection of I-131-labeled Lipiodol. Part I. Biodistribution study results in patients with hepa- tocellular carcinoma and liver metastasis. Radiology

NAKAKUMA K, TASHIRO S, HIRAOKA T, et al. Studies on anticancer treatment with an oily anticancer drug injected into the ligated feeding hepatic artery for liver cancer. Cancer 1983: 52: 2193-2200. KANEMATSU T, INOKUCHI K, SUGIMACHI K et al. Selective effects of lipiodolized antitumor agents. J Surg Oncol 1984: 25: 218-226. KONNO T, MAEDA H, IWAI K et al. Selective target- ing of anticancer drug and simultaneous image en- hancement in solid tumors by arterially adminis- tered lipid contrast medium. Cancer 1984: 54:

MAKI S, KONNO T, MAEDA H. Image enhancement in computerized tomography for sensitive diagnosis of liver cancer and semiquantitation of tumor selec- tive drug targeting with oily contrast medium. Can- cer 1985: %. 751-757. IWAI K, MAEDA H, KONNO Y. Use of oily contrast medium for selective drug trapping to tumor: en- hanced therapeutic effect and x-ray image. Cancer Res 1984: 44: 2115-2121. MILLER D L, O’LEARY T J, GIRTON M. Distribution of iodized oil within the liver after hepatic arterial injection. Radiology 1987: 162: 849-852. FUKUSHIMA S, KAWAGUCHI T, NISHIDA M et al. Selective anticancer effects of 3’,5’-Dioctanoyl-5- fluoro-2’-deoxyuridine, a lipophilic prodrug of 5’- Fluoro-2’-deoxyuridine, dissolved in an oily lym- phographic agent on hepatic cancer of rabbit bear- ing VX-2 tumor. Cancer Res 1987: 47: 193& 1934. Yo0 H S, PARK C H, SUH J H et al. Hepatocellular carcinoma: treatment with radioiodinated fatty acid ester. Yonsei Med J 1988: 29: 166-175. SENGER D R, GALU S J, DVORAK A M, PERRUZZI

1986: 18: 217-223.

1988: 168: 541-545.

2367-2374.

78 PARKETAL.

C A, HARVEY V S , DVORAK H F. Tumor cells secrete a vascular permeability factor that promotes ac- cumulation of ascitic fluid. Science 1983: 219: Deparfmenf of Pathology 983-985.

hepatocellular carcinoma. Korean J Pafhol (in Seoul press). Korea

Address: Chanil Park, M D

Yonsei University College of Medicine C.PO. Box 8044 20. PARK Y N, AHN H J, PARK C I. Kupffer cells in