• Open Access

Histological type of Thorotrast-induced liver tumors associated with the translocation of deposited radionuclides

Authors


5To whom correspondence should be addressed.
E-mail: fukumoto@idac.tohoku.ac.jp

Abstract

Exposure to internally deposited radionuclides is known to induce malignant tumors of various histological types. Thorotrast, a colloidal suspension of radioactive Thorium dioxide (232ThO2) that emits alpha-particles, was used as a radiographic contrast during World War II. Thorotrast is known to induce liver tumors, particularly intrahepatic cholangiocarcinoma (ICC) and angiosarcoma (AS), decades after injection. Therefore, patients injected with Thorotrast comprise a suitable study group to understand biological effects of internal ionizing radiation injury. Autoradiography and X-ray fluorescence spectrometry (XRF) were carried out on non-tumorous liver sections from Thorotrast-induced ICC (T-ICC) and Thorotrast-induced AS (T-AS). Autoradiography revealed that the slope of the regression line of the number of alpha tracks for the amount of deposited Thorium (232Th) was higher in non-tumorous parts of the liver with T-ICC than those with T-AS. XRF showed that the intensity ratio of Radium (Ra) to Thorium (Th) in non-tumorous liver tissue with T-ICC was significantly higher than that with T-AS. Furthermore, the mean 228Ra/232Th radioactivity ratio at the time of death calculated was also significantly higher in T-ICC cases than in T-AS cases. These suggest that the metabolic behavior of radionuclides such as relocation and excretion, as well as the content of deposited radionuclides, is a major factor in determining the histological type of Thorotrast-induced liver tumors. (Cancer Sci 2009)

Thorotrast is the trade name of a 25% colloidal suspension of radioactive thorium dioxide (ThO2) that naturally and predominantly emits alpha-particles. Thorotrast was used as a radiographic contrast agent from the 1930s to the 1950s.(1) Intravascularly injected, Thorotrast remains in the reticuloendothelial system over the lifetime of the person. In particular, more than 60% of total Thorotrast is located in the liver and those organs which are chronically irradiated by alpha-particles.(2)

Several decades after injection, Thorotrast has been known to induce liver cancers. Among the primary liver tumors in the Japanese population hepatocellular carcinoma (HCC) is the most frequent and the second most frequent is intrahepatic cholangiocarcinoma (ICC).(3) The incidence of hepatic angiosarcoma (AS) is negligible in patients who have not been injected with Thorotrast (non-Thorotrast patients). Whereas in patients injected with Thorotrast (Thorotrast patients), the most frequently induced liver tumor is ICC and the second-most frequently induced is AS. Compared with non-Thorotrast (non-T) patients with liver malignancies, odds ratios of ICC and AS in Thorotrast patients are 64.4 and 1658, respectively.(4) Our previous study revealed that Thorotrast-induced ICC (T-ICC) may originate from a common stem cell progenitor of hepatocytes and bile duct epithelial cells.(5) Furthermore, a major portion of HCC cases in Japan are associated with hepatitis virus C or B. Therefore, T-ICC and Thorotrast-induced AS (T-AS) are the most suitable cases for the analysis of the mechanisms of Thorotrast-induced liver malignant tumors.

The goal of this study was to elucidate the crucial difference to determine which liver tumor is induced by Thorotrast, T-ICC or T-AS. In order to resolve the problem, we compared the non-tumorous parts of the liver with T-ICC and T-AS. We measured the local dose of alpha-particles emitted from Thorotrast using autoradiography, and the amount of radionuclides using X-ray fluorescence spectrometry (XRF), which is suitable for nondestructive multi-elemental analysis of heavy elements such as rare-earth elements.

Materials and Methods

Materials.  We analyzed data on autopsy cases of Thorotrast patients with ICC (42 cases) and AS (22 cases) including age at administration, incubation period, and dose rates of the liver and the spleen (Table 1, Fig. 1). The data were retrieved from the Database on Thorotrast Patients in Japan (http://www.idac.tohoku.ac.jp/db/thorotrast/index%20english.html). The amount of deposited 232Th in the liver was calculated from the mean deposited amount of Actinium (daughter radionuclide of 232Th) found in several parts of the liver and was measured using a germanium counter.(6) Japanese Thorotrast patients comprise a reasonably homogeneous population, consisting of young male soldiers in their 20s or 30s at the time of administration, and they had received annual medical examinations. Therefore, we assumed that the incubation period almost matched the survival period after administration. We histologically classified tumors based on the World Health Organization classification criteria.(7) At least three pathologists diagnosed all the cases independently and reached collective consensus if initial independent diagnoses were different. We selected five cases of each of T-ICC and T-AS for further analyses according to the deposited amount of 232Th such as small, medium, and large amounts to minimize deposition dependent bias (Table 2).

Table 1.   Profile of autopsy cases with Thorotrast-induced liver tumors
 T-ICCT-AS
  1. Mean ± SD. †Ratio of dose rate of the liver to the spleen. T-AS, Thorotrast-induced angiosarcoma; T-ICC, Thorotrast-induced intrahepatic cholangiocarcinoma.

Number of cases4222
Age at administration (years)23.5 ± 3.626.1 ± 5.4
Incubation period (years)37.0 ± 7.738.5 ± 5.7
Liver/spleen (dose rate)†0.45 ± 0.370.40 ± 0.27
Figure 1.

 Scatter graph of dose rate-incubation period for Thorotrast-induced intrahepatic cholangiocarcinoma (T-ICC) and angiosarcoma (T-AS). Both dose rate and incubation period are not significantly different between patients with T-ICC (42 cases) and those with T-AS (22 cases). (△) Cases with T-ICC. y = −0.24 × x + 43.6, r = 0.54, P < 0.01 (dotted line). (bsl00001) Cases with T-AS. y = −0.23 × x + 43.8, r = 0.58, P < 0.01 (solid line).

Table 2.   Five cases of each of Thorotrast-induced ICC and AS in autoradiography and X-ray fluorescence analysis
Case of T-ICC (Autopsy number)1 (1192)2 (1096)3 (1205)4 (1218)5 (1228) 
 Age at administration (years)233825252327 ± 6.3 (25)
 Incubation periods (years)423542414341 ± 3.2 (42)
 Time since death (years)232922232124 ± 3.1 (23)
 Amount of Th deposits (mg/wet tissue)7.41.71.20.90.52.3 ± 2.9 (1.2)
 Liver/spleen (dose rate)†0.230.120.420.151.10.40 ± 0.36 (0.23)
Cases of T-AS (Autopsy number)1 (1148)2 (1213)3 (1092)4 (1181)5 (1212) 
  1. Data are presented as the mean ± SD (median). Autopsy number is presented according to the Database of Thorotrast Patents in Japan. †Ratio of dose rate of the liver to the spleen. T-AS, Thorotrast-induced angiosarcoma; T-ICC, Thorotrast-induced intrahepatic cholangiocarcinoma.

 Age at administration (years)232332202324 ± 4.5 (23)
 Incubation periods (years)393735444239 ± 3.6 (39)
 Time since death (years)252231232225 ± 3.8 (23)
 Amount of Th deposits (mg/wet tissue)4.03.11.11.10.52.0 ± 1.5 (1.1)
 Liver/spleen (dose rate)†1.40.050.10.150.090.36 ± 0.52 (0.1)

This study was approved by the Ethical Committee of Tohoku University, Faculty of Medicine.

Autoradiography.  Autoradiography was performed on 4-μm thick sections of liver tissue by dipping in liquid emulsion (Kodak NTB2; Kodak, Rochester, NY, USA) and exposing for 7 months. Image files of 15 different regions randomly selected from non-tumorous tissues of each case were recorded under a microscope using a × 40 objective lens (PLFL 40X; Olympus, Tokyo, Japan). The number of alpha tracks was counted. We manually traced the circumference of each Thorotrast conglomerate and measured its area using the software ‘length and area measuring’ (http://hp.vector.co.jp/authors/VA004392/Download.htm). Then, the density of alpha tracks per 100 μm2 of Thorotrast conglomerate was calculated.

X-ray fluorescence spectrometry (XRF).  In order to measure the content of trace elements in non-tumorous parts of the liver, we applied XRF to tissue sections adjacent to the sections used for autoradiography. Briefly, tissue sections were deparaffinized and analyzed by the method previously described.(8) XRF was carried out at a synchrotron radiation facility, Photon Factory (Ibaraki, Japan) using a BL4A bending magnet beamline; the beam was monochromatized at photon energies of 17.4 keV by silicon crystals and had a beam size of 0.5 × 1.4 mm on the sample. The fluorescent X-rays were measured with a Si (Li) detector and multichannel pulse height analyzer.(9) As negative controls for Thorotrast injection, liver tissues from two cases of non-T ICC (64- and 80-year-old males) were used. These control cases were histologically confirmed as being without cirrhosis or other noticeable changes other than ICC.

Statistical analysis.  Data were statistically analyzed by Student’s t-test. In the case of P < 0.05, the difference was considered statistically significant.

Role of the funding source.  The funders of the study had no role in study design, data collection, data analysis, data interpretation, or the writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.

Results

Both the age at administration and incubation period were not significantly different between patients with T-ICC and those with T-AS. The ratio of dose rate of the liver to the spleen (liver/spleen [dose rate]) of T-ICC was larger than that of T-AS, but was not significant (Table 1). The incubation period of both T-ICC and T-AS tended to decrease according to the increase of the dose rate in the liver (Fig. 1). The difference between regression slopes of T-ICC cases and T-AS cases was not significant (t-test).

Autoradiography was performed to assess the number of alpha tracks in non-tumorous parts of the liver. Since T-ICC and T-AS are characteristic of radiation-induced liver tumors and not associated with hepatitis virus, we analyzed five cases of each of T-ICC and T-AS (Table 2). The incubation period of years of T-ICC was 41 ± 3.2 and that of T-AS was 39 ± 3.6. All the demographic data presented were not significant between T-ICC and T-AS cases. As shown in Figure 2(A), deposited Thorotrast was mainly seen in the portal area as gray granules and alpha tracks radiated from the central Thorotrast conglomerate, indicating that alpha-particles were emitted from the deposited Thorotrast. The density of alpha tracks in the liver with T-ICC and T-AS was positively correlated with the deposited amount of 232Th, (Fig. 2B). Furthermore, the difference between the regression slopes of T-ICC cases and T-AS cases was significant (P < 0.05, t-test).

Figure 2.

 (A) Autoradiography of non-tumor part of the liver from Thorotrast patients (representative images). Alpha tracks were emitted from deposited Thorotrast in non-tumorous parts of the liver. The density of alpha tracks in the liver of Thorotrast-induced intrahepatic cholangiocarcinoma (T-ICC)-1 (b) is higher than that of Thorotrast-induced angiosarcoma (T-AS)-1 (d). (a) Non-tumor part adjacent to the portal area of T-ICC1. (b) Higher magnification of the rectangle of a. (c) Non-tumor part adjacent to the portal area of T-AS1. (d) Higher magnification of the rectangle of c. Exposure time, 7 months. (B) The number of alpha tracks per 100 μm2 of the surface of deposited Thorotrast. White triangle and dotted line, the liver with T-ICC (y = 1.72 × x, r = 0.78, P < 0.01). Black square and solid line, cases with T-AS (y = 1.26 × x, r = 0.59, P < 0.01). The difference of the regression slopes (r) of two groups was significant (P < 0.05) by t-test.

To elucidate the reason as to why the density of alpha tracks was different between the liver with T-ICC and that with T-AS, XRF was carried out with the section adjacent to that used for autoradiography. Representative charts of XRF of the liver tissue with T-ICC and that with control non-T ICC are shown in Figure 3(A). Peaks specific to the Th decay series such as peaks of Th at 12.97 keV (Lα) and Radium (Ra, a daughter radionuclide of 232Th) at 14.84 keV (Lβ) were distinctively observed in non-tumorous parts of liver tissues from both T-ICC and T-AS cases. We calculated the intensity ratio of Ra to Th (Ra/Th) from these XRF data. Ra/Th was significantly higher in T-ICC cases than in T-AS cases (P < 0.05) (Fig. 3B). This result suggests that radioactive substances other than 232Th, daughter radionuclides of the 232Th series, contribute to the density of alpha tracks.

Figure 3.

 (A) X-ray fluorescence spectrometry (XRF) of non-tumorous part of the liver. (a) Thorotrast-induced intrahepatic cholangiocarcinoma (T-ICC)-1; (b) non-Thorotrast case. The peaks of Th (Lα), 12.97 keV (black arrow) and Ra (Lβ), 14.84 keV (grey arrow) are not detectable in the non-Thorotrast case. (B) The intensity ratio of Radium (daughter radionuclide) to Thorium. Mean ± SD. *P < 0.05, statistically significant; Mann–Whitney U-test.

Discussion

The scatter graph of dose rate-incubation period indicated that the incubation period tended to decrease according to the increase of the dose rate of the liver from both T-ICC cases and T-AS cases. Thorotrast injected and not excreted from the body accumulates mainly in the liver and the spleen. Therefore, liver/spleen (dose rate) partly reflects excretion/relocation of Thorotrast from the liver.(10) The incubation periods of T-ICC and T-AS were not significantly different. Although the difference was not significant, liver/spleen (dose rate) of T-ICC was higher than that of T-AS.

The results of autoradiography revealed that radionuclides are deposited within a limited area, irrespective of the amount of radionuclides present. These results indicate that the greater the amount of radionuclides that are introduced into the body, the greater the amount of radionuclides that are concentrated within a limited area at the microscopic level.(11) As shown in Figure 2, the number of alpha-tracks was apparently higher in T-ICC cases than in T-AS cases, although the particle size of Thorotrast conglomerates was not different. Since the size of Thorotrast conglomerates was not different between T-ICC and T-AS cases, contribution of self-absorption of alpha-particles to the dose could be negligible in the present study. The slope of the regression line for the density of alpha tracks in relation to the deposited amount of 232Th was higher in the liver with T-ICC than that with T-AS, indicating that the content of radionuclides is different between T-ICC cases and T-AS cases. Therefore, we carried out XRF to determine what sort of radionuclides are deposited in the liver with T-ICC and with T-AS, respectively. The thorium decay series is composed of seven steps with alpha-particle emissions (Fig. 4). It is possible that only the decay step from 232Th to 228Ra was not in equilibrium status at the time of XRF analysis because all the Thorotrast patients in this study passed away more than 20 years ago. XRF revealed that the intensity ratio of Ra/Th in the liver with T-ICC was significantly higher than that with T-AS. Since the amount of 232Th and 228Ra is much higher than 228Th and 224Ra, respectively, the intensity ratio of Ra/Th calculated from XRF data almost represents that of 228Ra/232Th. The intensity ratio of 228Ra/232Th was different among cases examined, suggesting that the decay step from 232Th to 228Ra in deposited Thorotrast in the liver had not reached the equilibrium status even 20 years after death. The non-equilibrium state is attributable to the difference of metabolism between patient groups with T-ICC and those with T-AS. All XRF analyses in this study were performed more than 20 years after Thorotrast patients died. Ra/Th ratio was the highest in case T-ICC5. Furthermore, both the point at death and that at XRF analysis of T-ICC5 almost reached a plateau (Fig. 5). We therefore assumed that T-ICC5 already reached the equilibrium status at the time of XRF analysis. Assuming that the purity of 232Th in injected Thorotrast was the same among the cases examined, we calculated the 228Ra/232Th radioactivity ratio and how each case at XRF analysis departed from radiation equilibrium at the time of death. Figure 5 shows the curve of the 228Ra/232Th radioactivity ratio. Each case was marked at the time of XRF analysis in this study and at the time of death. Two studies have reported that the 228Ra/232Th radioactivity ratio in the liver of Thorotrast patients is 0.39 (0.25–0.6) and 0.38, respectively.(6,12) However, those studies did not take tumor histology into account. In this study, the mean 228Ra/232Th radioactivity ratio at death was significantly higher in T-ICC cases (0.74) than in T-AS cases (0.31) (Table 3). For elucidating radiation carcinogenic mechanisms, it is necessary to measure the deposited amount of each radionuclide separately depending on the histological classification of tumors. Thorotrast injected into rats remains in the body for a prolonged period, whereas daughter radionuclides such as Ra are primarily eliminated from the body or relocated to the bone and to the spleen, and the excretion rate of Thorium is much slower than its daughter nuclides.(13) In addition, the mass ratio of 232Th and 228Ra is 4.1 × 0.110 at radiation equilibrium. These results indicate that Thorotrast conglomerates which we can observe under a microscope only prove in situ existence of 232Th but do not reflect its amount. The area or volume of Thorotrast conglomerates does not reflect the total dose including daughter radionuclides of 232Th. Each calculation based on autoradiography and XRF revealed that the dose from Thorotrast conglomerates in the liver with T-ICC was higher than that with T-AS. It is indicated that overall metabolic behavior for daughter radionuclides such as relocation and excretion is different between patients in the T-ICC and T-AS patient groups.

Figure 4.

 Thorium decay series. Alpha-particles are emitted at seven steps of the series.

Figure 5.

228Ra/232Th radioactivity ratio in non-tumorous part of the liver at X-ray fluorescence spectrometry (XRF) and at the death. Gray arrow, Thorotrast-induced intrahepatic cholangiocarcinoma (T-ICC) at XRF analysis. Gray arrowhead, T-ICC at the death. Black arrow, Thorotrast-induced angiosarcoma (T-AS) at XRF analysis. Black arrowhead, T-AS at the death. Number, case number. Curve: inline image. Under radiation equilibrium of 232Th and 228Ra, inline image. inline image: Disintegration constant of 232Th. inline image: Disintegration constant of 228Ra.

Table 3.   Radioactivity ratio of 228Ra/232Th at present time and at the death
Case of T-ICC (Autopsy number)1 (1192)2 (1096)3 (1205)4 (1218)5 (1228) 
 Present time0.990.980.960.9810.98 ± 0.16 (0.98)
 At the death0.940.490.540.79>0.940.74 ± 0.21 (0.79)
Cases of T-AS (Autopsy number)1 (1148)2 (1213)3 (1092)4 (1181)5 (1212) 
  1. Data are presented as the mean ± SD (median). Autopsy number is presented according to the Database of Thorotrast Patents in Japan. T-AS, Thorotrast-induced angiosarcoma; T-ICC, Thorotrast-induced intrahepatic cholangiocarcinoma.

 Present time0.960.970.940.930.960.95 ± 0.15 (0.96)
 At the death0.280.6300.110.540.31 ± 0.27 (0.28)

As discussed in our previous study, three major factors are considered to be responsible for the long incubation time of Thorotrast-induced liver malignancies, several decades after ingestion: uneven distribution of radionuclides, limited range of irradiation, and dynamic movement of tumor precursor cells.(9) We also hypothesized that target cells in the liver of Thorotrast patients susceptible to malignant transformation may undergo one event and may then migrate outside of the range of alpha-particles, thereby avoiding immediate induction of successive additional events that would lead to cell death or neoplastic changes.(11)

In this study, we further postulate a mechanism to determine the histology of Thorotrast-induced liver tumors, ICC or AS. In case of T-ICC, the target cell toward ICC, that is, a progenitor cell of hepatocytes and bile duct epithelial cells, is located adjacent to the portal area.(14) Compared with T-AS, a relatively greater quantity of radionuclides are accumulated in the portal area of T-ICC cases because of the lower metabolic activity in the excretion of daughter radionuclides of 232Th from the liver. In the results, hepatocytes or their progenitor cells in the liver with T-ICC are irradiated more than vascular endothelial cells. In case of T-AS, the target cell toward AS is a vascular endothelial cell or its progenitor cell. Patients with T-AS have relatively higher metabolic activity in the excretion of Ra from the liver and the route of excretion; that is, endothelial cells are more severely irradiated than parenchymal cells.

In conclusion, this study suggests that the histological type of liver tumors induced by Thorotrast is associated with the patent’s metabolic activity of the liver, such as relocation and excretion of the daughter nuclides of 232Th.

Acknowledgments

We thank Dr Yutaka Tomita (Keio University, Department of Biosciences and Informatics) for advice about statistical processing. This work was supported in part by a grant from the Ministry of Health, Labour and Welfare of Japan.

Disclosure Statement

The authors have no conflict of interest.

Ancillary