188Re-HDD/lipiodol therapy for hepatocellular carcinoma: an activity escalation study
- First Online:
- Cite this article as:
- Lambert, B., Bacher, K., Defreyne, L. et al. Eur J Nucl Med Mol Imaging (2006) 33: 344. doi:10.1007/s00259-005-1954-1
- 96 Views
The aim of this study was to investigate the feasibility of administering increasing activities of 188Re-4-hexadecyl-1-2,9,9-tetramethyl-4,7-diaza-1,10-decanethiol/lipiodol (188Re-HDD/lipiodol) for the treatment of hepatocellular carcinoma (HCC) in patients with well-compensated cirrhosis.
The activity levels were increased by 1.1 GBq/step after a 6-week interval without unacceptable adverse events in at least five consecutive patients. Absorbed doses to the various organs were calculated according to the MIRD formalism, based on three gamma-scintigraphic studies. Response was assessed by means of MRI and alpha-fetoprotein (AFP) monitoring.
Thirty-five treatments were carried out in 28 patients. Activities from 4.8 to 7.0 GBq 188Re-HDD/lipiodol were administered via a transfemoral catheter. The mean absorbed dose to the liver (including tumour) was 7.6±2.2, 9.8±4.9 and 15.2±4.9 Gy for the 4.8-, 5.9- and 7.0-GBq groups, respectively. Treatment was well tolerated at all activity levels. Further escalation of the administered activity was not feasible owing to limitations related to the radiolabelling procedure. Response assessment on MRI showed partial response, stable disease and disease progression in 1, 28 and 2 assessable treatments, respectively. In 8 of 17 treatment sessions with an initially elevated AFP, a reduction ranging from 19% to 97% was observed 6 weeks later.
Following the intra-arterial administration of 4.8–7.0 GBq 188Re-HDD/lipiodol in patients with HCC and well-compensated liver cirrhosis, no severe adverse events occurred. Further escalation was not feasible owing to limitations in the radiolabelling procedure.
KeywordsHepatocellular carcinomaLipiodolRhenium-188Radionuclide therapyHDD
Hepatocellular carcinoma (HCC) is the most common type of primary liver cancer. HCC is among the most prevalent causes of cancer-related deaths, in particular in Southeast Asia, sub-Saharan Africa and Japan . In the West, its incidence keeps rising [2, 3]. The tumour often presents late, and since more than 80% of patients who present with HCC are suffering from underlying cirrhosis, therapeutic possibilities are limited. Surgery, by either hepatectomy or liver transplantation, is the mainstay of curative treatment. If the degree of liver dysfunction and the tumour load are taken into consideration, the vast majority of patients are not eligible for surgery . Local strategies such as percutaneous alcohol injection and radiofrequency ablation are often applied in inoperable patients presenting with a limited tumour load [5, 6]. For multifocal HCC, transarterial chemo-embolisation is generally accepted, and survival advantages have been identified in patients with well-compensated cirrhosis . Encouraging results have been reported using 131I-lipiodol, but concerns regarding radioprotection have restricted the use of 131I [8, 9]. The long physical half-life of 131I, i.e. 8 days, and its high energetic gamma ray of 364 keV necessitate prolonged hospitalisation for radioprotection purposes. 131I-lipiodol therapy is generally well tolerated, but the use of activities exceeding 2.22 GBq is restricted by the above-mentioned issue of radioprotection.
Rhenium-188 (188Re) has favourable characteristics for radionuclide therapy and, considering the limited success of 131I-lipiodol for treatment of relatively large tumours, the switch towards a radionuclide with a higher energy of the beta emission (2,120 keV and 1,960 keV for 188Re versus 606 keV for 131I) might yield improved response rates . 188Re emits a gamma ray of 155 keV at an abundance of 15%, allowing gamma camera imaging, and it has a relatively short physical half-life of 17 h, reducing radiation protection problems. Additionally, the radionuclide 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 .
The first clinical results using 188Re-4-hexadecyl-1-2,9,9-tetramethyl-4,7-diaza-1,10-decanethiol/lipiodol (188Re-HDD/lipiodol) were reported by Sundram et al. [12, 13]. Whereas Sundram and co-workers applied activities varying from 1.8 to 9.8 GBq, depending on the dose estimations following the administration of a scout dose, Lambert et al. used a fixed activity of 3.6 GBq. Tolerance was excellent in patients with well-compensated cirrhosis (Child-Pugh class A), and according to the dosimetric estimations a further escalation of the activity seemed feasible . The aim of this study was to administer increasing activities of 188Re-HDD/lipiodol and to assess the urinary elimination, organ dosimetry and toxicity. A secondary end point was response assessment.
Materials and methods
Synthesis and quality control of the radioconjugate
188W/188Re generators were purchased from the IRE (Institut des Radio-Eléments, Fleurus, Belgium). Lyophilised kits containing a HDD–chelator complex (4-hexadecyl-1-2,9,9-tetramethyl-4,7-diaza-1,10-decanethiol) were provided by the Seoul National University Hospital and 188Re-HDD/lipiodol was synthesised as described earlier by Jeong et al.[15, 16]. The concentrated eluate (6 ml) from a commercially available 188W/188Re generator was heated with the HDD/SnCl2 kit at 100°C for 1 h to produce 188Re-HDD complex [16, 17]. Three millilitres of lipiodol was added and mixed on a vortex to extract the 188Re-HDD into the lipiodol. After centrifugation at 4,000 g the 188Re-HDD/lipiodol fraction was separated. Finally, the 188Re-HDD/lipiodol layer was washed with a 0.9% NaCl solution. Quality control was performed according to the method described by Jeong et al. by instant thin-layer chromatography (ITLC-SG plates, Gelman Sciences, mobile phase: 0.9% NaCl and acetone) .
Groups of at least five patients were treated with increasing activities of 188Re- HDD/lipiodol, starting at 4.8 GBq. Escalation of the administered activity was implemented in increments of 1.1 (±0.55) GBq, following an interval of 6 weeks without the occurrence of an unacceptable event in at least five consecutive patients. If a case of unacceptable toxicity was recorded in the first five patients, a minimum of ten patients had to be treated at this activity level without the occurrence of an additional unacceptable adverse event. However, if a second case of unacceptable toxicity was observed, the study was stopped at that activity level.
An adverse event was considered unacceptable if it was grade 4 or 5 according to the CTC scale (Common Terminology Criteria for Adverse Events, Version 3.0. Cancer Therapy Evaluation Program. June 10, 2003. http://ctep.cancer.gov) and at least “probably related to the investigational agent”. However, grade 4 laboratory aberrations related to liver dysfunction could be considered acceptable, depending on the baseline status of particular cases. On the other hand, grade 3 toxicity could be deemed unacceptable if it was prolonged or had a clear negative impact on the patient’s quality of life compared with the baseline status. Any increase of more than 2 points on the Child-Pugh classification for liver dysfunction lasting until week 6 was regarded as unacceptable, unless tumour progression occurred.
Patients without evidence of disease progression on imaging were eligible for additional treatment sessions at 12-week intervals. An individual patient always remained within the same activity level if the inclusion criteria were still met.
The diagnosis of HCC had to be established by means of biopsy or conventional imaging techniques, such as computed tomography (CT), magnetic resonance imaging (MRI) or arteriography, possibly in conjunction with an elevated alpha-fetoprotein (AFP) level exceeding 400 ng/ml . For lesions smaller than 2 cm, a consensus had to be reached by two independent radiologists. Exclusion criteria were: eligibility for liver resection, pregnancy and breast feeding, age <18 years, Child B or C status according to the modified Child-Pugh score, poor general condition (Karnofsky <70%), white blood cell count <2×103/μl, forced expiratory volume in 1 s (FEV-1) or lung diffusion capacity for carbon monoxide (DLCO) <40% of the predicted value, abnormal baseline CT scan of the chest, an additional active primary malignancy other than HCC, symptomatic extrahepatic metastasis and potentially toxic anticancer treatment in the preceding 6 weeks . Additional contra-indications for arteriography and isolation procedure were: serum creatinine level >2 mg/dl, prothrombin time <50% of the normal value, platelet count <50×103/μl, inability to perform self-care, encephalopathy and incontinence. The study was approved by the institutional ethics committee and informed written consent was obtained from all patients.
Under local anaesthesia, a 5-French catheter was inserted transfemorally and introduced into the proper hepatic artery. After obtaining a diagnostic hepatic arteriogram, approximately 4 ml of 188Re-HDD/lipiodol was injected slowly into the proper hepatic artery under fluoroscopic control. Whenever aberrant arterial supply was present, the radioconjugate was injected selectively into the right and/or left hepatic artery separately. The volume of 4 ml was divided over the different hepatic arteries proportional to the volume of liver parenchyma supplied by each artery. To reduce uptake of free perrhenate in the thyroid or gastric mucosa, patients received 1 g sodium perchlorate for 3 days.
Assessment of urinary elimination and organ dosimetry
Following 188Re-HDD/lipiodol administration, patients were hospitalised in a dedicated radionuclide therapy room for 3 days. During the hospitalisation, patients collected their urine. Urine samples were analysed in a NaI(Tl) 3”×3” gamma well counter calibrated for 188Re (Cobra II , Perkin Elmer, USA). Time-activity curves were generated and fitted mono-exponentially (SPSS software version 10.0, Chicago, USA).
Whole-body scintigraphic studies were acquired at (±SD) 3.7±1.1, 24.5±3.0 and 49.9±3.5 h post administration, using a double- or triple-headed gamma camera (respectively AXIS and IRIX, Philips, Eindhoven, The Netherlands) equipped with medium-energy parallel-hole collimators. Scan speed varied from 30 to 10 cm/min depending on the time elapsed since administration. The imaging window was set at 155 keV (20%). For quantification purposes, a syringe containing a known activity of 188Re (mean activity ±SD: 16.0±3.0 MBq) was included in the whole-body scan. Regions of interests (ROIs) were drawn around the syringe, the total body, the liver (including tumour), the lungs and a background region on the first scan. The background corrected geometric mean of the total counts in the ROIs was used to calculate the total amount of activity in these regions, using the known activity in the syringe and experimental factors determined on an anthropomorphic phantom (Alderson Heart/Thorax SPECT phantom) for the conversion of the syringe activity into organ activity. In the latter conversion factors, the attenuation and scatter effects in this standard phantom were taken into account. The overall uncertainty using this methodology for the activity calculation was less than 18%.
Mono-exponential time-activity curves were generated for the total body, the liver and lungs using SPSS 10.0 software. Source organ residence times were determined from integration of the time-activity curves. Absorbed doses to the various organs were calculated according to the Medical Internal Radiation Dose (MIRD) formalism, using the MIRDOSE 3.1 (Oak Ridge Associated Universities, USA) software package [19, 20]. The patient’s dose rate was regularly measured at 1 m distance from the liver region using a survey meter.
Toxicity and response assessment
Laboratory testing of red and white blood cell counts, platelets, liver function and renal function was performed shortly before treatment, following 48 h and 2 and 6 weeks post injection. Clinical evaluation of toxicity was performed daily during hospitalisation and 2 and 6 weeks later. Lung function testing was performed before treatment and 6 weeks later. Toxicity measurements were scored by means of the CTC scale. Laboratory findings were compared by means of Friedman and Wilcoxon statistical testing and the significance level was set at P<0.05 (SPSS version 10.0). Radiological response was assessed by means of the RECIST (Response Evaluation Criteria in Solid Tumours) response criteria on CT scans or MR images acquired 6 weeks following treatment . AFP levels were measured at baseline, week 2 and week 6.
Between January 2004 and February 2005, 28 patients agreed to participate in the presented study. Nine patients underwent 14 treatment sessions with 4.8 GBq 188Re-HDD/lipiodol. In one patient his second administration was performed with only half of the planned activity owing to a technical problem during the radiolabelling procedure. Eleven administrations with 5.9 GBq of the radioconjugate were carried out in nine patients, and ten patients underwent a single treatment session with 7.0 GBq 188Re-HDD/lipiodol. Radiochemical purity was higher than 95% in all cases. The total radiochemical yield (n=34) was 49±5.4%.
Number of patients
The radioconjugate was injected in the proper hepatic artery, in both the right and the left branch and in only the right branch in 15, 11 and 2 patients, respectively. In the latter two cases only the right lobe was affected and treatment of the left lobe was not feasible owing to aberrant vascular anatomy.
Urinary elimination and organ dosimetry
Normal organ dosimetry: mean absorbed dose estimations (Gy) after treatment with 4.6±0.3, 5.8±0.3 and 6.8±0.2 GBq 188Re-HDD/lipiodol
4.6±0.3 GBq 188Re-HDD/lipiodol (n=12)
5.8±0.3 GBq 188Re-HDD/lipiodol (n=8)
6.8±0.2 GBq 188Re-HDD/lipiodol (n=7)
The dose rate at 1 m dropped below 20 μSv/h within the first 48 h following administration in all patients.
Adverse clinical events, scored according to the CTC scale version 3.0
Grade pre→post therapy
No. of therapies
Related to contrast agent
Causal relationship unclear, no signs of infection, normal CT scan of chest and lung function test
Cough, dyspnoea, fever
Twice in same patient, symptomatic relief with corticosteroids
Arthritis and hyperuricaemia
History of gout
Administered activity too low, history of variceal bleeding 
No clinical liver toxicity was encountered. In eight sessions the Child-Pugh score was increased by a single point 6 weeks later. Half of these cases were in patients treated at the highest activity level. In one case an increase by 2 points was observed, but this patient suffered a rapidly rising AFP and an increase in tumour dimensions. At discharge, statistically significant increases in bilirubin and AST levels were recorded for all activity levels. Overall, an increase in bilirubin and AST by one CTC grade was observed in 12 and six therapies, respectively, and these events did not occur more frequently at higher activity levels. Changes in liver parameters for the various activity levels are depicted in Fig. 1g–l.
Three patients were admitted to hospital for gastro-intestinal bleeding (grade 3). Scintigraphy did not reveal extrahepatic uptake of the radioconjugate in these patients. CT scans performed without contrast 2 weeks following treatment did not show extrahepatic lipiodol. Taking into account their individual histories of variceal bleeding, the cause of these events was not considered likely to be related to the radionuclide therapy.
One case of disease progression occurred in the patient whose second administration consisted of only half of the planned activity owing to a problematic radiolabelling procedure.
In 8 of 17 treatment sessions with an initially elevated AFP, a reduction (median 47%, range 19–97%) was observed 6 weeks later. Three patients had stable tumour markers whereas a rise in AFP (median 63%, 31–2,012% increase) was documented in six therapies.
In the present study increasing activities of 188Re-HDD/lipiodol were administered. When technically feasible, we aimed to treat the entire liver since there is a high risk of developing metachronous HCC in the non-treated segments . Targeting the whole liver was technically feasible in all but two patients. As regards the labelling procedure, no major technical problems were encountered up to 7.0 GBq except in one treatment.
Instant thin-layer chromatography showed that the activity in the urinary samples was perrhenate . If compared to our earlier experiences, the use of increasing activities from 3.6 up to 7.0 GBq 188Re-HDD/lipiodol did not yield an altered relative urinary excretion (p value: 0.30). If the shorter physical half-life of 188Re is taken into account, the urinary excretion compares favourably with the observations using 131I-lipiodol [25, 26]. For all activity levels, organ dose estimates were well below the threshold levels for adverse radiation-induced effects.
In the initial clinical experiences, Sundram et al. used 1.8–9.8 GBq 188Re-HDD/lipiodol, depending on the dose estimations following the administration of a scout dose, and the activity was administered as close to the tumour-feeding artery as possible. In that study, about 35 out of 70 patients did not experience any adverse event. The most frequent adverse events consisted of mild anorexia, right hypochondrial discomfort and low-grade fever . Although we aimed at treating the whole liver, we made similar observations. Irrespective of the relation with the radionuclide therapy, in 19 out of 32 assessable treatment sessions a symptom was reported by the patient. In five other treatments, only laboratory changes were recorded. In eight sessions, no changes were observed compared with baseline. Sundram et al. reported two cases of pleural effusion and these were attributed to radiation-induced pneumonitis . In the present study, one patient suffered cough and dyspnoea which exacerbated following her second treatment and required oral corticosteroid treatment. Similarities with a case previously described in the pilot study conducted earlier at our institution are striking. The absorbed cumulative lung doses for the two patients were estimated to be 10.4 and 11.5 Gy, and hence too low to explain this evolution. Although it was not clear whether the occurrence of pulmonary symptoms was related to the 188Re-HDD/lipiodol therapy, we advocate that in future particular attention be paid to patients developing pulmonary complaints.
The preliminary response rates indicate that the vast majority of patients had stable disease according to the RECIST criteria. However, these results are difficult to interpret since most patients were treated early after diagnosis and hence no evidence of rapidly progressive disease could be established at inclusion. Moreover, extensive tumour necrosis may not always be paralleled by a reduction in the diameter of the lesion . In two out of four subsequently transplanted patients in this series, more than 99% of tumour necrosis was present in the explanted liver although assessment by imaging suggested disease stabilisation.
Due to the relatively long half-life and high energy of the gamma emission of 131I, 131I-lipiodol treatment is associated with a hospital stay of up to 7 days in certain European countries. Besides the psychological burden for the patient, this isolation procedure is an extra cost. The need for rigorous radioprotective measures has been a major drawback to the administration of 131I-lipiodol activities exceeding 2.2 GBq. Moreover, for patients listed for liver transplantation, prior treatment with 131I-lipiodol excludes them from the waiting list for several weeks . In the present study, the patient’s dose rate measured at 1 m dropped below the local limit of 20 μSv/h within 48 h following administration of 188Re-HDD/lipiodol, allowing significantly shorter hospitalisation. Subsequent transplantation was allowed after an interval of only 1 week. The shorter stay in the shielded therapy room and the use of an on-site 188W/188Re generator improved the flexibility in treatment planning.
The labelling yield of 188Re-HDD ranges between 50% and 70%, and the synthesis of activities exceeding 7.0 GBq routinely therefore poses a problem. A number of authors have focussed on the development of new derivatives that may be produced at higher yields. These include bis-(diethyldithiocarbamato) nitrido 188Re lipiodol (188ReN-DEDC) and 188Re-(S2CPh)(S3CPh)2 lipiodol (188Re-SSS lipiodol). Feasibility of intra-arterial 188ReN-DEDC treatment, either alone or in combination with trans-arterial chemo-embolisation, was shown in a series of nine patients. However, some unexplained spleen and bone marrow uptake was observed 20 h post injection and one patient suffered grade 4 myelosuppression. Data on 188Re-SSS lipiodol in HCC patients are currently lacking [28, 29].
Future work should focus on the optimisation of the radiolabelling procedure of 188Re-HDD/lipiodol. Attempts should be made to elucidate various issues concerning the long-term effects of cumulative treatment on the liver and lung tissue. In addition, efforts should be made in these studies to perform SPECT tumour dosimetry.
Activities ranging from 4.8 to 7.0 GBq 188Re-HDD/lipiodol were administered to 28 patients in 35 treatment sessions. Urinary excretion eliminated 41.7% (±9.7) of the injected activity within 46 h (±9.6) after administration and did not differ significantly between the activity levels. The mean absorbed dose to the liver including the tumour was 7.6, 9.8 and 15.2 Gy for the 4.8-, 5.9- and 7.0-GBq patient groups, respectively, whereas the mean lung doses was 5.3, 6.8 and 8.9 Gy, respectively. The mean whole-body doses for the different activity levels were 0.6, 0.7 and 0.9 Gy, respectively. Treatment was well tolerated at all activity levels, without the occurrence of unacceptable adverse events. Further escalation of the administered activities was limited by technical reasons related to the radiolabelling procedure.
The authors would like to thank the IRE (Institut des Radio-Eléments, Fleurus, Belgium) for their technical assistance. The work was supported by a grant of the Bijzonder Onderzoeksfonds (Ghent University, No 011D9501).