Abstract
One approach to treatment of primary hepatocellular carcinoma (HCC) is intraarterial injection of 131I-lipiodol. Although clinical results have been positive, the therapy can be improved by using 188Re instead of 131I as the radionuclide. 188Re is a high-energy β-emitter, has a shorter half-life than 131I, and has only low-intensity γ-rays in its decay. The present study compared the cytotoxic effect of the radionuclide therapy in HCC patients treated with 131I-lipiodol and 188Re-4-hexadecyl 2,2,9,9-tetramethyl-4,7-diaza-1,10-decanethiol (HDD)/lipiodol. To this end, dicentric chromosomes (DCs) were scored in metaphase spreads of peripheral blood cultures. The equivalent total-body dose was deduced from the DC yields using an in vitro dose-response curve. Methods: Twenty 131I-lipiodol treatments and 11 188Re-HDD/lipiodol treatments were performed on, respectively, 16 and 7 patients with inoperable HCC. Patients received a mean activity of 1.89 GBq of 131I-lipiodol or 3.56 GBq of 188Re-HDD/lipiodol into the liver artery by catheterization. For each patient, a blood sample was taken during the week before therapy. A blood sample was also taken 7 and 14 d after administration for the patients treated with 131I-lipiodol and 1 or 2 d after administration for the patients treated with 188Re-HDD/lipiodol. Results: The mean DC yield of 188Re-HDD/lipiodol therapy (0.087 DCs per cell) was significantly lower than that of 131I-lipiodol therapy (0.144 DCs per cell) for the administered activities. Corresponding equivalent total-body doses were 1.04 Gy for 188Re-HDD/lipiodol and 1.46 Gy for 131I-lipiodol. Data analysis showed that, in comparison with 131I-lipidol, 188Re-HDD/lipiodol yielded a smaller cytotoxic effect and a lower radiation exposure for an expected higher tumor-killing effect. Conclusion: 188Re is a valuable alternative for 131I in the treatment of HCC with radiolabeled lipiodol, and a dose escalation study for 188Re-HDD/lipiodol therapy is warranted.
Primary hepatocellular carcinoma (HCC) is the most common primary liver malignancy and among the 10 most common tumors in the world. Chronic infection with the hepatitis B or C virus appears to be the most important risk factor for HCC. Patients with HCC have a poor prognosis, with a 5-y survival rate of less than 5%. Survival chances are best when liver transplantation or surgical resection is possible, but these therapies are applicable to only a few patients. For most patients, only palliative options remain. These include percutaneous ablation therapy (local) and intraarterial chemotherapy with or without subsequent embolization (locoregional) (1).
Metabolic radiation therapy using arterial administration of 131I-lipiodol has shown effective clinical results and good tolerance by patients. For patients with portal vein thrombosis, 131I-lipiodol has proven to be effective, and among patients treated surgically, 131I-lipiodol is the only auxiliary treatment proven effective at reducing recurrence (2,3). The treatment can also be used in a curative setting, with 131I-lipiodol being given neoadjuvantly before liver transplantation (4).
Despite the encouraging results obtained with 131I-lipiodol, the therapy—and especially the radionuclide—has important limitations. 131I is characterized by a high γ-ray emission (365 keV, 81%), which allows the imaging but is, together with the 8-d physical half-life of 131I, also responsible for long hospitalizations and limitation of the administered activity (5). On the other hand, the maximal β-energy of 131I is only 606 keV (89%) (6), allowing only rather small tumors to be irradiated efficiently. Because of the high-energy γ-radiation of 131I, distant tumor locations can still be irradiated by the cross-fire effect. However, Monte Carlo simulations show that the contribution of this cross-fire effect is only 10% in large tumors. 188Re has several physical characteristics that favor its replacing 131I in palliative therapy. The radionuclide has a relatively short physical half-life of 17 h and a maximal β-energy of 2.1 MeV (72%), with a 15% γ-component of 155 keV in its decay (6). The high-energy 188Re β-emission can destroy cells in a radius of several millimeters, and the 155-keV γ-rays allow γ-camera imaging. Furthermore, 188Re is eluted from a 188W/188Re generator, which has a long useful shelf-life of several months and provides a good yield of carrier-free 188Re routinely (7). 188Re is coupled indirectly to lipiodol using 4-hexadecyl 2,2,9,9-tetramethyl-4,7-diaza-1,10-decanethiol (HDD) as a chelating agent (8). Recently, promising clinical results were published for an International Atomic Energy Agency multicenter study using 188Re-HDD/lipiodol (9). However, 131I-lipiodol and 188Re-HDD/lipiodol were not compared.
Cytogenetic tests play an important role in the detection of biologic effects in patients exposed to ionizing radiation. Chromosomal aberrations, especially dicentric chromosomes (DCs), induced by ionizing radiation in human lymphocytes can be used as indicators of radiation exposure. Biologic dosimetry based on the analysis of DCs has been used since the mid 1960s. The aberrations scored in lymphocytes can be interpreted in terms of absorbed dose by reference to a dose-response calibration curve (10).
To compare the cytotoxic effect of 188Re-HDD/lipiodol radionuclide therapy with that of 131I-lipiodol therapy in the framework of a phase I study, we studied the frequency of DCs in a cohort of HCC patients given 188Re or 131I therapy. The equivalent total-body dose (ETBD) was estimated using an in vitro dose-response curve.
MATERIALS AND METHODS
188Re-HDD/Lipiodol Preparation
The HDD lyophilized kits were obtained from Seoul National University Hospital (8). Briefly, the concentrated eluate from the 188W/188Re generator is heated with HDD/SnCl2 in a boiling water bath for 1 h to produce 188Re-HDD complex. Lipiodol is then added, stirred in a vortex mixer, and centrifuged to extract the 188Re-HDD into the lipiodol. The 188Re-HDD/lipiodol radioconjugate is stable for at least 4 h (radiochemical purity > 95%).
Study Population
Between February 2002 and July 2003, 20 131I-lipiodol treatments (IL group, n = 20), and 11 188Re-HDD/lipiodol treatments (ReL group, n = 11) were administered in the Ghent University Hospital to, respectively, 16 and 7 patients with inoperable HCC. The patients in the IL group and in the ReL group were admitted the day before treatment and the day of treatment, respectively, for intravenous prehydration and initiation of thyroid blocking with either potassium iodide capsules (100 mg/d for 2 wk, IL group) or sodium perchlorate drops (500 mg for 5 d, ReL group). The patients received an activity of 1.89 GBq (SD, 0.15) of 131I-lipiodol (Lipiocis; Schering CIS BIO International) or 3.56 GBq (SD, 0.17) of 188Re-HDD/lipiodol into the liver artery by catheterization. The 131I-lipiodol and 188Re-HDD/lipiodol treatment programs were approved by the Ethical Committee of our hospital. Fifteen patients of the IL group and 4 patients of the ReL group received a single treatment. In each group, 2 patients received 2 consecutive treat-ments and 1 patient received 3 consecutive treatments over about half a year. An overview of the data is given in Table 1.
Overview of the Demographic Data and Results
Sample Collection
From each patient in the study, a heparinized blood sample was taken during the week before therapy. From each patient treated with 131I-lipiodol, a blood sample was obtained 7 and 14 d after administration. Taking into account the shorter half-life of 188Re, a blood sample from each patient treated with this radionuclide was obtained 1 d (21 h; SD, 3) and 2 d (51 h; SD, 2) after administration.
Lymphocyte Culture
From each sample, 2 blood cultures were made by the addition of 0.5 mL of whole blood to 4.5 mL of RPMI-1640 medium supplemented with 10% heat-inactivated fetal calf serum, 1% antibiotics (penicillin and streptomycin), and 1% l-glutamine. The lymphocytes were stimulated to divide with 1% phytohemagglutinin. The cultures were incubated at 37°C for 48 h. Three hours before arrest of the cultures, Colcemid (0.2 μg/mL; Alexis Biochemicals) was added to block the cells at metaphase. The cells were harvested by centrifugation of the samples, and the cell pellets were resuspended in 0.075 mol/L KCl for 15 min at 37°C. After the hypotonic shock, the cells were fixed 3 times in cold methanol:acetic acid (3:1). Finally, cells were dropped on clean slides and stained with 6% Romanovsky’s Giemsa solution. Three hundred well-spread metaphases were analyzed for the presence of DCs.
ETBD
The ETBD is the absorbed dose of ionizing radiation, which, if received homogeneously by the whole body, would produce the same yield of DCs as observed in the patients. The ETBD was derived from the increase of the DC yield in the blood samples of each patient, 1 and 2 wk (131I) or 1 and 2 d (188Re) after administration of the activity. The ETBD was calculated from an in vitro dose-response curve, Y = αD + βD2, with Y the DC yield observed and D the dose. To determine the in vitro dose-response curve, blood samples of 10 HCC patients were, before phytohemagglutinin stimulation, irradiated in a 37°C water bath with 60Co γ-rays with doses of 0.5, 1, 2, and 3 Gy at 1 Gy/min or sham irradiated.
Total-Body Scintigraphy and MIRD Dosimetry
For each patient, a set of biplanar anteroposterior total-body scintigraphic images was recorded after the therapy. The IL patients and the ReL patients received, respectively, 2 (7 and 14 d after therapy) and 4 (22, 30, 54, and 76 h after therapy) posttherapy scans. A syringe containing a known activity of 131I or 188Re in a polymethylmethacrylate phantom was scanned along with the patient. All scans were obtained with a IRIX camera (Philips) fitted with high-energy and medium-energy parallel-hole collimators for the IL and ReL patients, respectively.
For 6 of the 20 131I-lipiodol treatments and 2 of the 11 188Re-HDD/lipiodol treatments, a complete set of posttherapy scintigraphy scans was not available. Hence, the total-body dose could be calculated for only 14 131I-lipiodol and 9 188Re-HDD/lipiodol treatments.
On the HERMES system (Nuclear Diagnostics), irregular regions of interest were drawn around the syringe, the total body, the liver (including tumor), the lungs, and a background region on thefirst scan. The regions of interest were mirrored to the posterior image and copied to each subsequent scan. The background-corrected geometric mean of the total counts in the regions of interest was used to calculate the total amount of activity in the total body, the liver, and the lungs, using the known activity in the syringe. Then, the cumulative activity of the total body, the liver, and the lungs was calculated from the area under the time-activity curve and was represented by a single exponential fit drawn through the data points of all consecutive total-body scintigraphy scans. Absorbed doses to the total body, the liver, and the lungs were calculated according to the MIRD formula, using the MIRDDOSE3.0 software package (Oak Ridge Associated Universities) (11). Because of the heterogeneous dose distribution in the body and the inhomogeneous distribution of blood throughout the body (10% in the liver; 6% in the lungs) (12), the total-body absorbed dose cannot be used as a physical estimate of biologic ETBD. Therefore, the absorbed dose to the blood was calculated as the weighted sum of the doses to the liver, lungs, and remainder of the body, with the percentage of the blood pool in these compartments as a weighted factor.
An example of total-body scans for 131I-lipiodol and 188Re-HDD/lipiodol therapy for the same patient is given in Figure 1. The 188Re-HDD/lipiodol therapy was administered several months after the 131I-lipiodol therapy.
Total-body scans for 131I-lipiodol (A) and 188Re-HDD/lipiodol (B) for the same patient. These scans were obtained 14 d and 54 h, respectively, after administration of the activity.
Statistical Analysis
Linear-quadratic best fits were calculated using SPSS 10.0 software (SPSS Inc.). Statistical analysis was performed using MedCalc, version 4.0 (http://medcalc.med-ia.net/). Differences between 2 populations were investigated using a 2-tailed Mann-Whitney test. The χ2 test was used to compare the proportion of radiosensitive individuals in the patient and control populations.
RESULTS
In Vitro Dose Response
Figure 2 compares the in vitro induced DC yields for the HCC patients and the standard in vitro dose-response curve for a healthy population. Most HCC patients showed, for the entire studied dose range, more DCs per cell than did the healthy controls. At the 2-Gy dose, the differences between the HCC patients and the controls were statistically significant (P = 0.023, Mann-Whitney). When the 90th percentile of the healthy-control distribution was used as a cutoff, 60% of the HCC patients had elevated values of DCs at 2 Gy of in vitro irradiation (P = 0.06, χ2; Fig. 3).
Yield of DCs in lymphocytes of HCC patients after in vitro irradiation, vs. the standard dose-response curve for a healthy population.
Distribution of radiation-induced DCs after 2 Gy of in vitro irradiation in healthy donors and HCC patients. The vertical line represents the cutoff between radiosensitive and nonradiosensitive individuals based on the 90th percentile of the control population.
DCs: 188Re-HDD/Lipiodol Versus 131I-Lipiodol Therapy
An overview of the results is given in Table 1. Missing values are due to failure of the cell culture, severe health problems in patients, or patients who left the study protocol. The mean DC yield, after background correction, observed in patients treated with 131I-lipidol was 0.085 DCs per cell (SD, 0.068) 1 wk after therapy, compared with 0.065 DCs per cell (SD, 0.029) after 1 d in patients treated with 188Re-HDD/lipiodol. Statistical significance (P = 0.038, Mann-Whitney) between the 131I and the 188Re groups was reached at the second time point: 0.144 DCs per cell (SD, 0.075) for 131I-lipiodol after 2 wk and 0.087 DCs per cell (SD, 0.040) for 188Re-HDD/lipiodol after 2 d.
Three patients of the IL group and 3 of the ReL group received multiple subsequent treatments. The DC yield before therapy, 1 and 2 wk after therapy for 131I, and 1 and 2 d after therapy for 188Re are plotted against time in Figure 4. The figure shows that the increase in DC yield 2 wk (131I) or 2 d (188Re) after administration of the therapeutic activity at least partially recovered by the time of the subsequent therapy. The mean lymphocyte half-life calculated from these results was 7.8 mo.
Time plot of DC yield in patients undergoing multiple subsequent therapies with 131I-lipiodol (A) or 188Re-HDD/lipiodol (B).
Dose Estimations
The in vitro dose-response curve (α = 0.048; β = 0.031) for HCC patients was used to estimate the ETBD delivered by both therapies. The mean and the SD on the mean (SDM) are presented in Figure 5. At the second time point, the 131I-lipiodol therapy delivered an ETBD of 1.46 Gy (SD, 0.54), whereas the 188Re-HDD/lipiodol therapy was responsible for 1.04 Gy (SD, 0.36; P = 0.038, Mann-Whitney).
Comparison of the ETBD between 131I-lipiodol and 188Re-HDD/lipiodol at 1 and 2 wk after administration for 131I-lipiodol and at 1 and 2 d after administration for 188Re-HDD/lipiodol. Error bars represent the SD on the mean (SDM).
Similar biodistributions were found for both 131I-lipiodol and 188Re-HDD/lipiodol. The total-body biologic half-life obtained from the total-body scans was 9.2 d (SD, 1.4) for 188Re and 10.6 d (SD, 2.3) for 131I (P > 0.5, Mann-Whitney).
Two weeks after 131I-lipiodol therapy, the total-body, liver, and lung mean doses calculated from the total-body scans using the MIRD formalism were 0.67 Gy (SD, 0.12), 11.0 Gy (SD, 4.0), and 6.5 Gy (SD, 2.1), respectively. Two days after 188Re-HDD/lipiodol administration, absorbed doses of 0.46 Gy (SD, 0.07), 6.8 Gy (SD, 2.1), and 4.4 Gy (SD, 1.0) were calculated for the total body, the liver, and the lungs, respectively. The absorbed dose to the blood, calculated from the doses to the liver, lungs, and remainder of the body using the percentage of blood in these compartments, was 1.56 Gy (SD, 0.34) 2 wk after 131I-lipiodol therapy and 1.04 Gy (SD, 0.18) 2 d after 188Re-HDD/lipiodol administration.
Blood Count Data
In the framework of the phase I study, information on blood cell counts was available for patients treated with 188Re-HDD/lipiodol. The white blood cell (WBC) and thrombocyte counts versus time after administration are depicted in Figure 6. Except for patient ReL6, no major drop in WBC count was observed. The high increase in the WBC count of patient ReL1 2 wk after the first therapy was explained by enteritis. The low values in patient ReL2 were due to the HCC. In general, the WBC count tends to decrease slightly with time. Most patients (9/11) started therapy with thrombocyte counts less than 150 × 103/μL. Except for patient ReL3, no major fluctuations in thrombocyte counts were noted. The high WBC decrease noted in patient ReL6 was less pronounced in the thrombocyte count.
Plot of the WBC (A) and thrombocyte (B) counts vs. time after 188Re-HDD/lipiodol administration.
DISCUSSION
During the last few years, 131I-lipiodol as a treatment for HCC has attracted much attention because of the promising results that have been achieved. The 131I radioisotope has, however, significant constraints. An ideal radiotherapeutic agent should have good stability, high-energy β-radiation, and low-energy γ-emission, giving a low radiation burden to nontarget organs but sufficient to allow imaging. In addition, good availability and low cost are important requirements (13). The 8-d physical half-life and the high γ-ray emission of 131I make radioprotection difficult in 131I-lipiodol therapy and lead to isolation of the patients for up to 7 d after therapy in certain European countries, to comply with national guidelines. Moreover, the use of higher administered activities of 131I-lipiodol is limited by the radiation protection issues associated with the high-energy γ-radiation of 131I. Among the alternative isotopes to avoid these drawbacks, 188Re is an important candidate. The isotope has a short physical half-life of 17 h and a lower γ-ray emission, which decreases the isolation to a maximum of 2 d. Furthermore, 188Re has a higher β-emission and can be eluted from an in-house 188W/188Re generator, recently available on demand. The use of an 188W/188Re generator system is cost-effective, since these generators have a long shelf-life, resulting in a very low cost per dose (13). Radiopharmaceuticals labeled with 188Re have previously been used in the treatment of painful bone metastases, in the pretransplant treatment of leukemia patients, and in the prevention of coronary restenosis (14–17).
In this phase I study, the cytotoxic effects of 188Re-HDD/lipiodol and 131I-lipiodol therapy were compared. To assess cytotoxicity, the standard technique of DC scoring in lymphocytes was used. Doses were estimated from the DC yields on the basis of the dose-response curve derived from blood samples of HCC patients. With respect to in vitro irradiation, HCC patients turned out to be more sensitive than a healthy population. This observation confirmed other authors’ data indicating that a significant fraction of cancer patients show enhanced in vitro radiosensitivity (18,19). The use of patient-specific dose-response curves for total-body dose estimation is therefore important. 188Re-HDD/lipiodol treatment induced significantly fewer DCs than did 131I-lipiodol therapy. Consequently, the ETBD delivered by the 188Re-HDD/lipiodol therapy was lower than that delivered by the 131I-lipiodol therapy for administered activities of 3.56 GBq and 1.89 GBq, respectively. After the chosen time points, 51 h for 188Re and 14 d for 131I, and with an effective half-life of 15.7 h for 188Re and 4.6 d for 131I, the percentage of the total cumulative activity considered was 89% and 88%, respectively. The total patient dose due to the 188Re-HDD/lipiodol therapy was thus significantly lower than that due to the 131I-lipiodol therapy for the activities administered.
A MIRDOSE calculation of the self-dose S values for 5-cm-diameter spheric nodules representing the tumor, combined with the effective half-life values obtained, shows that 188Re-HDD/lipiodol and 131I-lipiodol may be expected to have the same biologic effect on the tumor when the administered 188Re activity is 60% higher than the 131I activity. In this calculation, the biologic effect of the difference in dose rate between both isotopes resulting in the same total tumor dose was taken into account based on the isoeffect curve for different dose rates used in brachytherapy (20). In our phase I trial, 3.56 GBq of 188Re-HDD/lipiodol and 1.89 GBq of 131I-lipiodol were administered. For these activity values, the tumor radiation response is expected to be higher for the 188Re-HDD/lipiodol therapy. Nevertheless, total-body doses for 188Re-HDD/lipiodol were significantly lower, as can be explained by the lower γ-radiation component of 188Re. Similar results were obtained by physical dosimetry using the MIRD formalism (188Re: 0.42 Gy; 131I: 0.67 Gy). The fact that the doses of the MIRD calculation are about half the biologically estimated doses (188Re: 1.04 Gy; 131I: 1.46 Gy) can be explained by the inhomogeneous distribution of blood throughout the body. In fact, in lipiodol therapy, significant activity accumulates in the liver and lungs. Because of the high absorbed doses in these organs and the large percentage of the blood pool inside, the total-body absorbed dose cannot be used as a physical estimate of the biologic ETBD. However, the absorbed dose to the blood, obtained from physical dosimetry, is in perfect agreement with the results from biologic dosimetry.
For total-body exposures of about 1 Gy, a slight decrease of approximately 20% in the WBC count and 30% in the platelet count is expected (21). Our patient group treated with 188Re-HDD/lipiodol showed a decrease in this range. On the basis of the ECOG Common Toxicity Criteria scale (version 2.0, revised 1999), this decrease is not alarming. Two patients reach a scale 3 for the thrombocyte counts, but these patients already had very low thrombocyte counts before therapy.
CONCLUSION
The present study showed that, compared with 131I-lipidol therapy, 188Re-HDD/lipiodol therapy yields a significantly lower cytotoxic effect and lower radiation exposure for an expected higher tumor-killing effect. 188Re is an excellent alternative for 131I in the internal radiation therapy of HCC with lipiodol. 188Re allows higher administered doses, reduces radiation-protection problems, and improves patients’ quality of life by shortening hospitalizations. The application of this new therapeutic agent for HCC will be investigated further in a dose-escalation study.
Acknowledgments
This work was supported by grant 01114501 from the Bijzonder Onderzoeksfonds (Gent University). We thank all the patients and control donors who participated. Special thanks are due to Virginie de Gelder for her help with the cell cultures and the scorings.
Footnotes
Received Sep. 24, 2003; revision accepted Dec. 12, 2003.
For correspondence or reprints contact: Kim De Ruyck, MSc, Department of Anatomy, Embryology, Histology, and Medical Physics, Ghent University, Proeftuinstraat 86, B-9000 Gent, Belgium.
E-mail: kim.deruyck{at}UGent.be
