In vitro assessment of Lipiodol-targeted radiotherapy for liver and colorectal cancer cell lines

Sir We read with great interest the article by Al-Mufti et al (19 (Br J Cancer79: 1665–1671). However, we found the concept which their study design was based needs to be clarified. The toxic action of a chemotherapeutic drug on cancer cells is diffe from that of a therapeutic radiopharmaceutical. For a chemot peutic drug such as a methylating agent to produce its cytotox its molecules need to enter the cancer cells before it can a their DNA. That is why P-glycoprotein, which causes efflux of drug molecule out of the cells, leads to drug resistance. O other hand, the radiation-emiting isotope contained in a radio maceutical produces its cytotoxic effect by bombarding the D of the cancer cells with beta particles or depositing radia energy to the cells in some other manner. This can be achiev long as the cancer cells are within the range at which the rad can penetrate. For iodine-131 ( 131I) the beta-particle it emits carries t majority of the radiation energy. The mean penetration of its particle in soft tissues is 0.4 mm, which is equivalent to 16– cell diameters. By the inverse-square law, the closer the r pharmaceutical molecule to the cancer cell, the greater will b amount of energy deposited, but it is not necessary for the r pharmaceutical molecules to enter the cancer cells before cy icity can be produced. In this study, the authors wante demonstrate that 131I produces its preferential cytotoxic effec only when it is inside the malignant cell (intracellular rad therapy, in the words of the authors) and Lipiodol was essenti transporting the radionuclide into the cell. They tried to show the Lipiodol cannot enter the cancer cells in the absence of 131I and so produces no cytotoxic effects. The same happens for 131I alone without Lipiodol. As stated by the authors, the iodine ( 127I) content of Lipiodol is 38–40% w/v (Al-Mufti et al, 1999), whereas < 4 μg of sodium iodide is present in every 25 mCi of 131I solution we used to label 1 ml of Lipiodol (Lau et al, 1999). Therefore percentage of 127I in Lipiodol being exchanged by 131I during the radiolabelling is negligible. The conversion of 127I-Lipiodol to 131ILipiodol should not have any change in the physiochemical p erties and therefore the cells should have recognized bot non-radioactive and the radioactive Lipiodol as identi Nevertheless, the authors reported very different electron mi raphy patterns of the two kinds of Lipiodol in both the malign and benign cells (Al-Mufti et al, 1999). Cold Lipiodol appeared the form of cytoplasmic membrane-bound vesicles, whereas 131Ilipiodol was detected inside dead cancer cells and viable b cells which are claimed to be healthy by the authors. Lipiodol (density 1.28 g ml –1) is denser than aqueous cultu medium and the two liquids are immiscible. From our experie even in the presence of Urografin as an emulsifying agent emulsion state could not be prolonged for more than 1 h (unl was continuously agitated) and the Lipiodol still separated ou ot ay, n

used to label 1 ml of Lipiodol (Lau et al, 1999). Therefore the percentage of 127 I in Lipiodol being exchanged by 131 I during the radiolabelling is negligible. The conversion of 127 I-Lipiodol to 131 I-Lipiodol should not have any change in the physiochemical properties and therefore the cells should have recognized both the non-radioactive and the radioactive Lipiodol as identical. Nevertheless, the authors reported very different electron micrography patterns of the two kinds of Lipiodol in both the malignant and benign cells (Al-Mufti et al, 1999). Cold Lipiodol appeared in the form of cytoplasmic membrane-bound vesicles, whereas 131 Ilipiodol was detected inside dead cancer cells and viable benign cells which are claimed to be healthy by the authors.
Lipiodol (density 1.28 g ml -1 ) is denser than aqueous culture medium and the two liquids are immiscible. From our experience, even in the presence of Urografin as an emulsifying agent, the emulsion state could not be prolonged for more than 1 h (unless it was continuously agitated) and the Lipiodol still separated out and settled at the bottom on standing. Hence their experimental observations might be interpreted as follows. On standing inside the incubator for over 6 h, the 131 I-Lipiodol component of the emulsion has separated and sunk. It came into intimate contact with the monolayer of cells at the bottom of the well. The local radioactivity concentrations in the vicinity of the cells were 1.0, 2.0 and 4.0 µCi of 131 I in 4 µl of 131 I-Lipiodol (equivalent to 250, 500 and 1000 µCi ml -1 for low, medium and high dose). Thus a high radiation dose was delivered to the cells. When the cells are killed or damaged by the radiation, the cells lost their integrity and the cell membrane became permeable to the passage of 131 I-Lipiodol which ended up inside the cells. The endothelial cells should have a lower population of dividing cells which are more susceptible to radiation damage and so the effects on the benign cells were seen to be sub-lethal while the malignant cells were killed. However, the integrity of the endothelial cell might still have been affected in some way and so the 131 I-Lipiodol could gain its passage into some of these benign cells.
Such radiation effects were not observed with aqueous NaI ( 131 I) solution. The aqueous NaI ( 131 I) solution, being completely miscible with the culture medium, will be distributed uniformly throughout the 100 µl of medium and therefore the effective radioactivity concentrations in the vicinity of the cells was really 40 µCi ml -1 . So even the low dose 131 I-Lipiodol was providing a 6.25 times higher local radioactivity concentration of 131 I than in the aqueous NaI ( 131 I) solution. It is therefore logical for the authors to observe no cytotoxic effect in the NaI ( 131 I) solution despite the total radioactivity in the well was four times that of the low dose 131 I-Lipiodol. We can assure the authors of a similar cell killing effect as observed with 131 I-Lipiodol if they try to increase the NaI ( 131 I) concentration to 250 µCi ml -1 .
From our own experience in treating 26 patients with hepatocellular carcinoma using 131 I-Lipiodol (Leung et al, 1994) and the results of the other 12 clinical series of hepatic cancer treated with the same agent that have recently been reviewed by us (Ho et al, 1998) we agree with the authors that the response is highly variable and may partly depend on local pharmacokinetics. The other important factor of course is the radio-sensitivity of the tumour cells. As we have pointed out in our review article (Ho et al, 1998) for a lesion to be completely destroyed by 131 I-Lipiodol, it needs to uptake the agent in such a way that every cancer cell lies within 2.4 mm from the radioactive oil, which is the maximum penetration of the beta-radiation from 131 I. 131 I-Lipiodol could hardly concentrate in hypovascular or necrotic tumours and therefore their responses are poor.
In conclusion, 131 I-Lipiodol bound to cytoplasmic membrane of the cancer cell as vesicles of lipids is already at a sufficiently close distance to kill the cancer cells. It is not necessary for the 131 I-Lipiodol molecules to enter the cancer cell before it can produce its cytotoxicity.

Letter to the Editor
In vitro assessment of Lipiodol-targeted radiotherapy for liver and colorectal cancer cell lines