- Email: [email protected]
Hypericin as a Marker for Determination of Tissue Viability After Intratumoral Ethanol Injection in a Murine Liver Tumor Model
Laboratory Investigation
Hypericin as a Marker for Determination of Tissue Viability After Intratumoral Ethanol Injection in a Murine Liver Tumor Model1 Marie Van de Putte, Huaijun Wang, Feng Chen, Peter A.M. de Witte, Yicheng Ni
Rationale and Objectives. In this preclinical proof-of-principle study, the necrosis avid agent hypericin was investigated as a potential early indicator for therapeutic response after ethanol-mediated chemical ablation in murine liver tumors. Materials and Methods. Seven mice bearing intrahepatic radiation-induced fibrosarcoma-1 tumors were intravenously injected with hypericin 1 hour before (n ⫽ 3) or 24 hours after (n ⫽ 4) intratumoral ethanol injection. Mice were euthanized 24 hours after hypericin injection and, taking advantage of the fluorescent property of the compound, the excised livers were investigated qualitatively and quantitatively by means of fluoromacroscopic and fluoromicroscopic examinations, colocalized with conventional histomorphology. Results. Significant differences in hypericin fluorescence were found in necrosis, viable tumor and normal liver tissue in decreasing order (P ⬍ .05) (ie, in necrosis, mean fluorescence densities were about 4.5 times higher than in viable tumor and approximately 14 times higher than in normal liver). When hypericin was injected 1 hour before, maximal blood concentrations were achieved at the time of ethanol treatment, so that on ablation an outstanding extravasation took place in the entire necrotic area in comparison with accumulation of hypericin only at the peripheral zone of necrosis when it was injected 24 hours after ablation. Conclusions. Hypericin specifically enhanced the imaging contrast between necrotic and viable tissues and nonspecifically distinguished viable tumor from normal liver. Injection of hypericin shortly before ablation is more favorable than after ablation, because it circumvents difficulties with no-entry zones for hypericin and requires shorter intervals between ethanol ablation and imaging. Key Words. Necrosis; NACA; hypericin; ethanol injection; liver neoplasm. ©
AUR, 2008
Percutaneous ethanol injection (PEI) is a minimally invasive treatment, which has been clinically applied for a
Acad Radiol 2008; 15:107–113 1
From the Laboratory for Pharmaceutical Biology, Faculty of Pharmaceutical Sciences (M.V.d.P., P.A.M.d.W.) and Department of Radiology, University Hospital, Faculty of Medicine (H.W., F.C., Y.N.), K.U. Leuven, Herestraat 49, B-3000, Leuven, Belgium. Received July 13, 2007; accepted August 21, 2007. Supported by grants awarded by Fonds voor Wetenschappleijk Onderzoek-Vlaanderen (FWO Vlaanderen), Geconcerteerde Onderzoeksactie (GOA) of the Flemish Government, FWP Impulsfinanciering project, and OT project. Address correspondence to: Y.N. e-mail: Yicheng.Ni@ med.kuleuven.be
© AUR, 2008 doi:10.1016/j.acra.2007.08.008
couple of decades as an effective alternative to surgery for patients with small hepatocellular carcinoma. Instead of excising the entire tumor from the patient as in standard surgery, PEI serves as a therapy of chemical ablation to instantaneously kill the tumor in situ by intratumoral injection of absolute ethanol. Complete necrosis of lesions smaller than 3 cm in diameter has been achieved (1). However, as a result of difficulties in obtaining homogeneous intratumoral alcohol distribution, incomplete therapeutic effect may occur because of residual viable neoplastic tissue in tumors larger than 3 cm (2,3). This may lead to tumor recurrence particularly along the periphery of the treated area (4,5).
107
VAN DE PUTTE ET AL
A troublesome flaw related to PEI treatment is the difficulty to confirm complete necrosis early after treatment. Early detection of a residual or locally recurrent tumor has a major effect on the prognosis and adjuvant treatment, whereas late diagnosis is associated with peripheral regrowth and remote metastasis making retreatment difficult (6). Because biopsies and serum tumor markers are only of limited usefulness for assessing the response to PEI, evaluation of the therapeutic effect of PEI is based mainly on findings at imaging studies (7). However, the visual effects created with all currently applied in vivo imaging techniques (eg, magnetic resonance imaging [MRI], computed tomography, ultrasound [US] imaging) are nonspecific, indirect, and inaccurate. On contrast enhanced MRI, a thin and regular enhancing halo, from an inflammatory reaction, is often observed at the periphery of the hypointense necrotic lesion (4,8). Because this strongly enhanced halo may mimic residual tumor, interpretation of enhancing areas at the periphery of the treated lesion can be difficult. Only at later follow-up imaging studies, a few months after treatment, the halo disappears (9). The use of color Doppler US and contrastenhanced gray-scale US imaging is only recommended to monitor the response of the tumor during the course of PEI treatment from a substantial risk for false-negative results in postprocedural imaging (9). Therefore, with the current diagnostic techniques, it is difficult to make a clear-cut distinction between the ablated dead tissues and unablated viable tissues early after PEI treatment. Necrosis avid contrast agents (NACAs) have recently been introduced as potential markers for in vivo detection of necrosis. NACAs, suitable for MRI, may significantly improve the postprocedural imaging after necrosis inducing treatments, because such compounds specifically label necrotic tissues in vivo (10,11). High-intensity signals on delayed NACA-enhanced images could therefore be considered as an indicator for necrosis and could be useful for the stratification of tumors in nonnecrotic, necrotic, and severe necrotic regions (11). This information on the degree of necrosis could be valuable not only in postprocedural imaging, but also in the diagnosis and prognosis of bulky necrotic tumors. Hypericin is a naturally occurring chemical derived from the plant genus Hypericum (12,13). The compound has been used as a fluorescent photosensitizer in the experimental and clinical detection of cancer and antitumoral photodynamic therapy (14,15). More recently, hypericin was also recognized as a non–porphyrin necrosis avid agent because the compound specifically accumulates
108
Academic Radiology, Vol 15, No 1, January 2008
in nonviable tissues such as necrotic and irreversibly damaged ischemic tissues. For instance, after systemic injection, 10- to 18-fold higher concentrations of radiolabeled hypericin ([123I]-iodohypericin) were found in infarcted liver and myocardium, respectively, than in normal surrounding tissues, indicating an extraordinary high target to non–target ratio (16). A similar affinity was found for bulky subcutaneous tumors consisting of large areas of spontaneous necrosis, which, in comparison to viable tumor 19-fold higher activities of [123I]-iodohypericin, were recorded (Van de Putte M, unpublished data 2007). In view of these results, we believe that hypericin or its derivatives may hold promise as contrast enhancers applied with modalities of optical, nuclear, and magnetic resonance imaging for early assessment of antineoplastic response following necrosis-inducing treatments such as PEI. The aim of the present proof-of-principle study was to evaluate the necrosis-avid characteristics of hypericin in a murine model with intrahepatic implantation of a radiationinduced fibrosarcoma (RIF-1) after intratumoral ethanol injection. Because of the excellent fluorescent properties of hypericin, we were able to obtain detailed macroscopic and microscopic fluorescence images of the ethanol-induced lesions, which were further colocalized with conventional histopathology to identify coagulative necrotic zones, peripheral untreated tumor, and normal liver tissue.
MATERIALS AND METHODS Animals and Tumor System This animal experiment was in compliance with the current institutional regulations for use and care of laboratory animals. Subcutaneously implanted RIF-1 in mice is one of the most widely used tumor models in cancer research for its biologic stability, minimal immunogenicity, low metastatic potential, and responsiveness to various therapeutic interventions (17). Similar to the recently reported rat model with liver implantation of rhabdomyosarcoma (18), we created intrahepatic growth of RIF-1 in mice to mimic liver metastasis as a more clinically relevant tumor model for studying the pharmacologic property of hypericin and its potential use in minimally invasive tumor therapies such as PEI. Nine female C3H/km mice (Charles River Laboratories, France) weighing 21–25 g were anesthetized with intraperitoneal injection of Nembutal (Sanofi Sante Animale, Brussels, Belgium) at 40 mg/kg and midline lapa-
Academic Radiology, Vol 15, No 1, January 2008
HYPERICIN: A VIABILITY MARKER AFTER PEI
Figure 1. Monitoring tumor growth (arrow) of radiation-induced fibrosarcoma (RIF-1) in mice with T2-weighted magnetic resonance imaging at 7, 9, 11, and 13 days after intrahepatic implantation. Tumor size about 5 mm (d) in diameter was considered appropriate for the treatment with intratumoral ethanol injection.
rotomized as tumor recipients. One cubic millimeter of tumor tissue, freshly harvested from a donor mouse with subcutaneous growth of RIF-1, was implanted in the left lateral liver lobe of the recipient mouse, which was allowed to recover after closure of the abdomen with layered sutures. The RIF-1 tumor growth in the liver was monitored routinely using MRI at a 1.5-Tesla clinical scanner (Sonata, Siemens, Germany) coupled with a small surface loop coil (MRI Devices Corporation, Waukesha, WI) under isoflurane gas anesthesia (IMS, Harvard Apparatus, Holliston, MA). About 14 days after implantation, a hyperintense or slightly hypointense liver mass of 0.5 cm in diameter on T2- or T1-weighted images, respectively, was regarded adequate for receiving ablation therapy (Fig 1). One mouse died of hypothermia during liver tumor implantation and the other was found without tumor growth, both were eliminated from the study. Intratumoral Ethanol Injection Under the same anesthesia regime as that for liver tumor implantation, tumor ablation procedures were performed with laparotomy instead of percutaneous intervention to better steer the extent of ablation. Absolute ethanol (99.9%, 200 L) was gently injected through an insulin syringe into part of liver RIF-1 tumor, whereas the proximal liver lobe was compressed with two fingers for 3 minutes to prevent spilling of the ethanol liquid into systemic circulation. The ethanol-affected tumor and liver tissue became immediately pale because of chemical denaturation, whereas the appearances of viable tumor and liver tissues remained unchanged. The abdominal incision in all mice was closed after the treatments. Hypericin Preparation and Administration Hypericin was synthesized from emodin anthraquinone according to Falk et al (19). Before intravenous injection,
hypericin was dissolved in a mixture of 25% dimethylsulfoxide, 25% polyethylene glycol 400, and water (2 mg/mL). The solution was injected in a tail vein at a dose of 10 mg/kg either 1 hour before (n ⫽ 3) or 24 h after (n ⫽ 4) ethanol treatment of the liver tumor. Fluoromacroscopic and Fluoromicroscopic Examinations Mice were euthanized by an intraperitoneally injected overdose of phenobarbital 24 hours after hypericin injection. Livers were excised and photographed under tungsten light and UV365 light to assess gross distribution of fluorescent hypericin in treated tumor and surrounding liver. Subsequently, treated tumors surrounded by normal liver were dissected, immediately mounted in medium (Tissue Tek embedding medium, Miles Inc, Elkhart, IN) and immersed in liquid nitrogen. Different cryostat sections (5-m slices) were taken from each tumor and examined by fluorescence microscopy (Axioskop 2 plus) equipped with a light-sensitive, charge-coupled device digital camera (AxioCam HR, Carl Zeiss, Göttingen, Germany) to acquire fluorescence images. To visualize hypericin, the Zeiss filter set 14 (excitation: BP 510 –560 nm, emission: LP 590 nm) was used. Eventually, tissue slices were stained with hematoxylin and eosin (H&E) for conventional light microscopy and photomicrography. Imaging Quantitative Analyses Quantitative results of fluoromicroscopic images were obtained by manually drawing eight regions of interest per tumor (n ⫽ 7) in the areas of ablated and viable tissues. Before that, a distinction between ablated tissue, viable tumor, and normal liver was made based on the histologic examination of the H&E-stained slices. Mean fluorescence densities, corrected for background, were
109
VAN DE PUTTE ET AL
Academic Radiology, Vol 15, No 1, January 2008
Figure 2. Macroscopic photographs taken under ultraviolet light and corresponding photographs under tungsten light of livers bearing ethanol-treated RIF-1 tumors. Mice were intravenously injected with 10 mg/kg hypericin at 24 hours after (a,b) or 1 hour before (c,d) ethanol treatment. All mice were sacrificed at 24 hours after hypericin injection and excised livers were digitally photographed. Ablated tumor (*), viable tumor (o), normal liver (x), ablated liver (**). Scale bar ⫽ 0.5 cm.
obtained by means of a KS imaging software system (Carl Zeiss, Vision, Hallbergmoos, Germany). Ratios of fluorescence densities in the different regions were then calculated for each condition. Statistical analysis was performed using Prism 4.00, GraphPad Software (San Diego, CA). A two-way analysis of variance with Bonferroni test was performed and p-values of P ⬍ .05 were considered statistically significant.
RESULTS On macroscopic photographs taken under ultraviolet light (Fig 2), ablated tumor, residual viable tumor, and normal liver showed different levels of fluorescence intensity in a decreasing order. The effect was prominent either when mice received hypericin after the ethanol treatment of the tumor mass (Fig 2a,b), or when hypericin was given before (Fig 2c,d). During the 24 hours after hypericin injection, the normal liver was predominantly cleared from hypericin in contrast to viable tumor tissue, which showed moderately elevated hypericin fluorescence intensity at 24 hours after injection. Fluoromicroscopic images revealed a distinct fluorescence intensity between necrotic tissue, residual tumor and normal liver (Fig 3a,c,d). Necrotic areas displayed the highest fluorescence intensity, followed by viable tumor
110
with moderate fluorescence and normal liver with the lowest intensity. No-entry zones for hypericin, corresponding with large necrotic areas, were regularly seen when hypericin was injected 24 hours after ethanol treatment (Fig 3b). Only the border zone between the ablated tissue and normal liver or viable tumor showed high uptake of hypericin, resulting in a fluorescent rim surrounding the no-entry zone. On H&E stains, this no-entry zone consisted mainly of dehydrated tissue with signs of severe cell damage and loss of tissue structure. The rim surrounding the no-entry zone had a typical appearance of eosinophilic coagulation tissue. Ratios of mean fluorescence densities in different areas of ablated tissue, viable tumor, and normal liver are shown in Table 1. Overall, ratios indicate three different levels of fluorescence density in necrotic tissue, viable tumor, and normal liver in decreasing order (P ⬍ .05). Mean fluorescence densities were not significantly different when hypericin was injected 24 hours after or 1 hour before ethanol treatment (P ⬎ .05).
DISCUSSION In the context of the evaluation of necrosis avid agents for use as short-term therapeutic response indicators after necrosis-inducing treatments such as PEI, we conducted
Academic Radiology, Vol 15, No 1, January 2008
HYPERICIN: A VIABILITY MARKER AFTER PEI
Figure 3. Fluorescence photomicrographs and corresponding hematoxylin and eosin stains of 5-m slices of livers bearing ethanol-treated RIF-1 tumors. Mice were intravenously injected with 10 mg/kg hypericin at 24 hours after (a,b) or 1 hour before (c,d) ethanol treatment. All mice were sacrificed at 24 hours after hypericin injection. Ablated tissue (NE), dehydrated ablated tissue (NE*), viable tumor (T), normal liver (L). Scale bar ⫽ 100 m.
111
VAN DE PUTTE ET AL
Academic Radiology, Vol 15, No 1, January 2008
Table 1 Ratios of Mean Fluorescence Densities in Different Regions of Fluoromicroscopic Images Taken From 5-m Tumor Sections ⌬T
NE Tissue Viable T
NE Tissue Normal L
Viable T Normal L
24 h 1h
5.53 ⫾ 2.29 (ns) 3.54 ⫾ 1.12
15.20 ⫾ 8.89 (ns) 13.13 ⫾ 5.82
3.84 ⫾ 1.24 (ns) 3.05 ⫾ 1.86
Tumors were excised from mice that received 10 mg/kg hypericin either 24 h after (⌬ T ⫽ 24 h), or 1 h before ethanol treatment (⌬ T ⫽ 1 h). Statistical analysis, comparing (⌬ T ⫽ 24 h) with (⌬ T ⫽ 1 h) was performed using Prism 4.00, GraphPad Software (San Diego, CA). A two-way analysis of variance with Bonferroni test was performed and P values of P ⬍ .05 were considered statistically significant. NE: necrotic; T: tumor; L: liver; ns: non-significant.
this preclinical study exploiting the necrosis-avid property of hypericin. For this purpose, we designed a murine RIF-1 liver tumor model by implanting a small tissue chip of a subcutaneously grown tumor in the liver. Unlike other animal studies that recruit subcutaneously implanted tumors, this murine hepatic tumor model may better mimic clinical conditions including visceral blood perfusion, parenchymal tumor growth, and relevant microenvironment, hence providing more clinically pertinent information for cancer diagnosis and therapy. Other advantages of the method employed include successful tumor growth in 84% of cases, rapid tumor growth in about 2 weeks, survival rate higher than 90% until the end of the experiment, minimally metastatic potential, and uncomplicated implantation procedure. In the present study, microscopic images revealed significant differences in hypericin fluorescence in three areas of interest (ie, necrosis, viable tumor, and normal liver tissue). Necrotic areas displayed an outstanding fluorescence signal, followed by moderate fluorescence intensity in viable tumor and low intensity in normal liver. Thus the three different tissue components could be visually stratified, even as early as 24 hours after intratumoral ethanol injection, simulating the results obtained with a NACA on MRI in a rat model treated with radiofrequency ablation (11). In a recent study exploring the mechanism of hypericin’s avidity for necrosis, we found that hypericin specifically stained necrotic tissue through compound-dependent interactions with necrotic cellular debris, whereas viable tissues were nonspecifically labeled to a lesser extent (Van de Putte M et al, submitted, 2007, to be cross-referenced upon publication). The moderate but persistent accumulation of hypericin in viable tumor 24 hours after injection can therefore be explained mainly
112
by the nonspecific uptake of hypericin by the tumor cells or stroma and a lack of biliary excretion in the tumor, similar to the phenomenon reported for MRI hepatobiliary contrast agents (20). In normal liver, hypericin was substantially cleared within 24 hours from functioning hepatic metabolic activities and biliary excretion. In this study, hypericin was intravenously administered both before and after intratumoral injection of ethanol. Both conditions achieved similar levels of fluorescence densities in different tissues (ie, high uptake of hypericin in necrotic areas) compared with moderate and low uptake in respectively viable tumor and normal liver. When hypericin was injected 1 hour before, maximal blood concentrations were achieved at the time of ethanol treatment (21), so that on ablation an outstanding extravasation took place in the entire necrotic area. Conversely, when hypericin was injected 24 hours after treatment, no-entry zones for hypericin were regularly seen. This phenomenon can be explained by the dehydration effect of alcohol in tissue. Ethanol acts by diffusing within the cells, which causes immediate dehydration of cytoplasmic proteins with consequent coagulation necrosis followed by fibrosis. This phenomenon may be described as in vivo “chemical denaturation.” In addition, necrosis of endothelial cells and platelet aggregation results in thrombosis of blood vessels followed by ischemia (5,22). As a result of the vascular shutdown, large lesions are not easily accessible to hypericin and subsequent diffusion throughout the entire necrotic lesion is hampered. When hypericin is injected after ethanol treatment, the compound will enter the necrotic zone through residual functioning blood vessels in the vicinity, followed by concentration gradient guided diffusion into the necrotic area, causing an evolving rim of intense fluorescence. Moreover, as suggested in an earlier study by Ni and coworkers (10), in vivo tissue degradation is required before interaction with NACAs (eg, hypericin) can take place. In contrast to traditional coagulation necrosis, in which enzymatic degradation plays a predominant role, instantaneous dehydration, and denaturation of cytosolic proteins prevents enzymatic degradation from occurring in tissue debris. Yet tissue degradation after ethanol-induced dehydration or denaturation requires a time-consuming process of days to weeks, in which the center of the ethanol-induced coagulative necrosis will be infiltrated by immune cells. After digestion of the dead tissue by enzymatic reactions, the necrotic debris will then be absorbed and replaced by granulation tissue.
Academic Radiology, Vol 15, No 1, January 2008
Before we can determine the efficacy of hypericin in detecting residual disease in a clinical setting and possibly observe an improved antitumoral outcome after necrosisinducing antineoplastic treatments such as PEI, a number of issues have to be settled. A first drawback of the compound hypericin is its potential skin phototoxicity. A direct infusion of hypericin into zones of necrosis within minutes before or after a PEI procedure, in combination with a transcatheter interventional procedure could offer advantages including increased local concentration while reducing its dosage, less systemic exposure and, improved patient convenience. A second drawback is the requirement of invasive procedures for fluorescence imaging of lesions in visceral organs. Applying a laparoscopic approach, which has been increasingly adopted in clinical practice, may reduce its invasiveness. Last, real noninvasive imaging can only be realized by tagging hypericin with a marker for a particular imaging modality, for instance, a gadolinium chelate for MRI. However, potential phototoxicity of hypericin derivatives at millimolar doses may limit the use for contrast enhanced MRI and it remains to be investigated whether produced signals are strong enough for the stratification of tumors in nonnecrotic, necrotic, and severe necrotic regions. Alternatively, hypericin may be tagged with a radionuclide for nuclear imaging. Positron emission tomography using hypericin labeled with iodine-124 could circumvent phototoxicity since only nanomolar amount of the compound is administered. Effective imaging using positron emission tomography appears to be feasible if the relatively low positron abundance of iodine-124 is compensated by the high ratio of tracer uptake in necrosis to surrounding tissues. In conclusion, the results of the present study suggest that hypericin offers significant potential in the early assessment of response following PEI. Hypericin features the property to specifically enhance the imaging contrast between necrotic and viable tissues and to nonspecifically distinguish viable tumor from normal liver. Hypericin injection before treatment appears more favorable than after treatment, since this approach circumvents difficulties with no-entry zones for hypericin and requires shorter intervals between ethanol treatment and imaging. We believe that our results have shown sufficient promise to justify continued research. Labeling hypericin to extend its use to MRI could be a first step in this direction.
HYPERICIN: A VIABILITY MARKER AFTER PEI
REFERENCES 1. Steele G, Ravikumar T. Resection of hepatic metastases from colorectal cancer. Biologic perspective. Ann Surg 1989; 210:127–138. 2. Livraghi T, Festi D, Monti F, et al. US-guided percutaneous alcohol injection of small hepatic and abdominal tumors. Radiology 1986; 161: 309 –312. 3. Shiina S, Yasuda H, Muto H, et al. Percutaneous ethanol injection in the treatment of liver neoplasms. Am J Roentgenol 1987; 149:949 –952. 4. Bartolozzi C, Lencioni R, Caramella D, et al. Hepatocellular carcinoma: CT and MR features after transcatheter arterial embolization and percutaneous ethanol injection. Radiology 1994; 191:123–128. 5. Lin X, Lin L. Local injection therapy for hepatocellular carcinoma. Hepatobil Pancreat Dis Int 2006; 5:16 –21. 6. Vogl T, Eichler K, Zangos S, et al. Hepatocellular carcinoma: role of imaging diagnostics in detection, intervention and follow-up. Rofo 2002; 174:1358 –1368. 7. Bartolozzi C, Lencioni R. Ethanol injection for the treatment of hepatic tumours. Eur Radiol 1996; 6:682– 696. 8. Nagel H, Bernardino M. Contrast-enhanced MR imaging of hepatic lesions treated with percutaneous ethanol ablation therapy. Radiology 1993; 189:265–270. 9. Cioni D, Lencioni R, Bartolozzi C. Percutaneous ablation of liver malignancies: imaging evaluation of treatment response. Eur J Ultrasound 2001; 13:73–93. 10. Ni Y, Bormans G, Chen F, et al. Necrosis avid contrast agents: functional similarity versus structural diversity. Invest Radiol 2005; 40:526 – 535. 11. Ni Y, Chen F, Mulier S, et al. Magnetic resonance imaging after radiofrequency ablation in a rodent model of liver tumor: tissue characterization using a novel necrosis avid contrast agent. Eur Radiol 2006; 16: 1031–1040. 12. Giese A. Hypericism. Photochem Photobiol Rev 1980; 5:229 –255. 13. Kartnig T, Gobel I, Heydel B. Production of hypericin, pseudohypericin and flavonoids in cell cultures of various Hypericum species and their chemotypes. Planta Med 1996; 62:51–53. 14. D’Hallewin M, De Witte P, Waelkens E, et al. Fluorescence detection of flat bladder carcinoma in situ after intravesical instillation of hypericin. J Urol 2000; 164:349 –351 15. Kamuhabwa A, Agostinis P, Ahmed B, et al. Hypericin as a potential phototherapeutic agent in superficial transitional cell carcinoma of the bladder. Photochem Photobiol Sci 2004; 3:772–780. 16. Ni Y, Huyghe D, Verbeke K, et al. First preclinical evaluation of mono[123I]iodohypericin as a necrosis-avid tracer agent. Eur J Nucl Med Mol Imaging 2006; 33:595– 601. 17. Twentyman P, Brown J, Gray J, et al. A new mouse tumor model system (RIF-1) for comparison of end-point studies. J Natl Cancer Inst 1980; 64:595– 604. 18. Chen F, Sun X, De Keyzer F, et al. Liver tumor model with implanted rhabdomyosarcoma in rats: MR imaging, microangiography, and histopathologic analysis. Radiology 2006; 239:554 –562. 19. Falk H, Meyer J, Oberreiter M. A convenient semisynthetic route to hypericin. Monatsh Chem 1993; 124:339 –341. 20. Ni Y, Marchal G. Enhanced magnetic resonance imaging for tissue characterization of liver abnormalities with hepatobiliary contrast agents: an overview of preclinical animal experiments. Topics in Magn Reson Imaging 1998; 9:183–195. 21. Chen B, Xu Y, Roskams T, Delaey E, et al. Efficacy of antitumoral photodynamic therapy with hypericin: relationship between biodistribution and photodynamic effects in the RIF-1 mouse tumor model. Int J Cancer 2001; 93:275–282. 22. Guan Y, Sun L, Zhou X, et al. Hepatocellular carcinoma treated with interventional procedures: CT and MRI follow-up. World J Gastroenterol 2004; 10:3543–3548.
113
Hypericin as a Marker for Determination of Tissue Viability After Intratumoral Ethanol Injection in a Murine Liver Tumor Model1 Marie Van de Putte, Huaijun Wang, Feng Chen, Peter A.M. de Witte, Yicheng Ni
Rationale and Objectives. In this preclinical proof-of-principle study, the necrosis avid agent hypericin was investigated as a potential early indicator for therapeutic response after ethanol-mediated chemical ablation in murine liver tumors. Materials and Methods. Seven mice bearing intrahepatic radiation-induced fibrosarcoma-1 tumors were intravenously injected with hypericin 1 hour before (n ⫽ 3) or 24 hours after (n ⫽ 4) intratumoral ethanol injection. Mice were euthanized 24 hours after hypericin injection and, taking advantage of the fluorescent property of the compound, the excised livers were investigated qualitatively and quantitatively by means of fluoromacroscopic and fluoromicroscopic examinations, colocalized with conventional histomorphology. Results. Significant differences in hypericin fluorescence were found in necrosis, viable tumor and normal liver tissue in decreasing order (P ⬍ .05) (ie, in necrosis, mean fluorescence densities were about 4.5 times higher than in viable tumor and approximately 14 times higher than in normal liver). When hypericin was injected 1 hour before, maximal blood concentrations were achieved at the time of ethanol treatment, so that on ablation an outstanding extravasation took place in the entire necrotic area in comparison with accumulation of hypericin only at the peripheral zone of necrosis when it was injected 24 hours after ablation. Conclusions. Hypericin specifically enhanced the imaging contrast between necrotic and viable tissues and nonspecifically distinguished viable tumor from normal liver. Injection of hypericin shortly before ablation is more favorable than after ablation, because it circumvents difficulties with no-entry zones for hypericin and requires shorter intervals between ethanol ablation and imaging. Key Words. Necrosis; NACA; hypericin; ethanol injection; liver neoplasm. ©
AUR, 2008
Percutaneous ethanol injection (PEI) is a minimally invasive treatment, which has been clinically applied for a
Acad Radiol 2008; 15:107–113 1
From the Laboratory for Pharmaceutical Biology, Faculty of Pharmaceutical Sciences (M.V.d.P., P.A.M.d.W.) and Department of Radiology, University Hospital, Faculty of Medicine (H.W., F.C., Y.N.), K.U. Leuven, Herestraat 49, B-3000, Leuven, Belgium. Received July 13, 2007; accepted August 21, 2007. Supported by grants awarded by Fonds voor Wetenschappleijk Onderzoek-Vlaanderen (FWO Vlaanderen), Geconcerteerde Onderzoeksactie (GOA) of the Flemish Government, FWP Impulsfinanciering project, and OT project. Address correspondence to: Y.N. e-mail: Yicheng.Ni@ med.kuleuven.be
© AUR, 2008 doi:10.1016/j.acra.2007.08.008
couple of decades as an effective alternative to surgery for patients with small hepatocellular carcinoma. Instead of excising the entire tumor from the patient as in standard surgery, PEI serves as a therapy of chemical ablation to instantaneously kill the tumor in situ by intratumoral injection of absolute ethanol. Complete necrosis of lesions smaller than 3 cm in diameter has been achieved (1). However, as a result of difficulties in obtaining homogeneous intratumoral alcohol distribution, incomplete therapeutic effect may occur because of residual viable neoplastic tissue in tumors larger than 3 cm (2,3). This may lead to tumor recurrence particularly along the periphery of the treated area (4,5).
107
VAN DE PUTTE ET AL
A troublesome flaw related to PEI treatment is the difficulty to confirm complete necrosis early after treatment. Early detection of a residual or locally recurrent tumor has a major effect on the prognosis and adjuvant treatment, whereas late diagnosis is associated with peripheral regrowth and remote metastasis making retreatment difficult (6). Because biopsies and serum tumor markers are only of limited usefulness for assessing the response to PEI, evaluation of the therapeutic effect of PEI is based mainly on findings at imaging studies (7). However, the visual effects created with all currently applied in vivo imaging techniques (eg, magnetic resonance imaging [MRI], computed tomography, ultrasound [US] imaging) are nonspecific, indirect, and inaccurate. On contrast enhanced MRI, a thin and regular enhancing halo, from an inflammatory reaction, is often observed at the periphery of the hypointense necrotic lesion (4,8). Because this strongly enhanced halo may mimic residual tumor, interpretation of enhancing areas at the periphery of the treated lesion can be difficult. Only at later follow-up imaging studies, a few months after treatment, the halo disappears (9). The use of color Doppler US and contrastenhanced gray-scale US imaging is only recommended to monitor the response of the tumor during the course of PEI treatment from a substantial risk for false-negative results in postprocedural imaging (9). Therefore, with the current diagnostic techniques, it is difficult to make a clear-cut distinction between the ablated dead tissues and unablated viable tissues early after PEI treatment. Necrosis avid contrast agents (NACAs) have recently been introduced as potential markers for in vivo detection of necrosis. NACAs, suitable for MRI, may significantly improve the postprocedural imaging after necrosis inducing treatments, because such compounds specifically label necrotic tissues in vivo (10,11). High-intensity signals on delayed NACA-enhanced images could therefore be considered as an indicator for necrosis and could be useful for the stratification of tumors in nonnecrotic, necrotic, and severe necrotic regions (11). This information on the degree of necrosis could be valuable not only in postprocedural imaging, but also in the diagnosis and prognosis of bulky necrotic tumors. Hypericin is a naturally occurring chemical derived from the plant genus Hypericum (12,13). The compound has been used as a fluorescent photosensitizer in the experimental and clinical detection of cancer and antitumoral photodynamic therapy (14,15). More recently, hypericin was also recognized as a non–porphyrin necrosis avid agent because the compound specifically accumulates
108
Academic Radiology, Vol 15, No 1, January 2008
in nonviable tissues such as necrotic and irreversibly damaged ischemic tissues. For instance, after systemic injection, 10- to 18-fold higher concentrations of radiolabeled hypericin ([123I]-iodohypericin) were found in infarcted liver and myocardium, respectively, than in normal surrounding tissues, indicating an extraordinary high target to non–target ratio (16). A similar affinity was found for bulky subcutaneous tumors consisting of large areas of spontaneous necrosis, which, in comparison to viable tumor 19-fold higher activities of [123I]-iodohypericin, were recorded (Van de Putte M, unpublished data 2007). In view of these results, we believe that hypericin or its derivatives may hold promise as contrast enhancers applied with modalities of optical, nuclear, and magnetic resonance imaging for early assessment of antineoplastic response following necrosis-inducing treatments such as PEI. The aim of the present proof-of-principle study was to evaluate the necrosis-avid characteristics of hypericin in a murine model with intrahepatic implantation of a radiationinduced fibrosarcoma (RIF-1) after intratumoral ethanol injection. Because of the excellent fluorescent properties of hypericin, we were able to obtain detailed macroscopic and microscopic fluorescence images of the ethanol-induced lesions, which were further colocalized with conventional histopathology to identify coagulative necrotic zones, peripheral untreated tumor, and normal liver tissue.
MATERIALS AND METHODS Animals and Tumor System This animal experiment was in compliance with the current institutional regulations for use and care of laboratory animals. Subcutaneously implanted RIF-1 in mice is one of the most widely used tumor models in cancer research for its biologic stability, minimal immunogenicity, low metastatic potential, and responsiveness to various therapeutic interventions (17). Similar to the recently reported rat model with liver implantation of rhabdomyosarcoma (18), we created intrahepatic growth of RIF-1 in mice to mimic liver metastasis as a more clinically relevant tumor model for studying the pharmacologic property of hypericin and its potential use in minimally invasive tumor therapies such as PEI. Nine female C3H/km mice (Charles River Laboratories, France) weighing 21–25 g were anesthetized with intraperitoneal injection of Nembutal (Sanofi Sante Animale, Brussels, Belgium) at 40 mg/kg and midline lapa-
Academic Radiology, Vol 15, No 1, January 2008
HYPERICIN: A VIABILITY MARKER AFTER PEI
Figure 1. Monitoring tumor growth (arrow) of radiation-induced fibrosarcoma (RIF-1) in mice with T2-weighted magnetic resonance imaging at 7, 9, 11, and 13 days after intrahepatic implantation. Tumor size about 5 mm (d) in diameter was considered appropriate for the treatment with intratumoral ethanol injection.
rotomized as tumor recipients. One cubic millimeter of tumor tissue, freshly harvested from a donor mouse with subcutaneous growth of RIF-1, was implanted in the left lateral liver lobe of the recipient mouse, which was allowed to recover after closure of the abdomen with layered sutures. The RIF-1 tumor growth in the liver was monitored routinely using MRI at a 1.5-Tesla clinical scanner (Sonata, Siemens, Germany) coupled with a small surface loop coil (MRI Devices Corporation, Waukesha, WI) under isoflurane gas anesthesia (IMS, Harvard Apparatus, Holliston, MA). About 14 days after implantation, a hyperintense or slightly hypointense liver mass of 0.5 cm in diameter on T2- or T1-weighted images, respectively, was regarded adequate for receiving ablation therapy (Fig 1). One mouse died of hypothermia during liver tumor implantation and the other was found without tumor growth, both were eliminated from the study. Intratumoral Ethanol Injection Under the same anesthesia regime as that for liver tumor implantation, tumor ablation procedures were performed with laparotomy instead of percutaneous intervention to better steer the extent of ablation. Absolute ethanol (99.9%, 200 L) was gently injected through an insulin syringe into part of liver RIF-1 tumor, whereas the proximal liver lobe was compressed with two fingers for 3 minutes to prevent spilling of the ethanol liquid into systemic circulation. The ethanol-affected tumor and liver tissue became immediately pale because of chemical denaturation, whereas the appearances of viable tumor and liver tissues remained unchanged. The abdominal incision in all mice was closed after the treatments. Hypericin Preparation and Administration Hypericin was synthesized from emodin anthraquinone according to Falk et al (19). Before intravenous injection,
hypericin was dissolved in a mixture of 25% dimethylsulfoxide, 25% polyethylene glycol 400, and water (2 mg/mL). The solution was injected in a tail vein at a dose of 10 mg/kg either 1 hour before (n ⫽ 3) or 24 h after (n ⫽ 4) ethanol treatment of the liver tumor. Fluoromacroscopic and Fluoromicroscopic Examinations Mice were euthanized by an intraperitoneally injected overdose of phenobarbital 24 hours after hypericin injection. Livers were excised and photographed under tungsten light and UV365 light to assess gross distribution of fluorescent hypericin in treated tumor and surrounding liver. Subsequently, treated tumors surrounded by normal liver were dissected, immediately mounted in medium (Tissue Tek embedding medium, Miles Inc, Elkhart, IN) and immersed in liquid nitrogen. Different cryostat sections (5-m slices) were taken from each tumor and examined by fluorescence microscopy (Axioskop 2 plus) equipped with a light-sensitive, charge-coupled device digital camera (AxioCam HR, Carl Zeiss, Göttingen, Germany) to acquire fluorescence images. To visualize hypericin, the Zeiss filter set 14 (excitation: BP 510 –560 nm, emission: LP 590 nm) was used. Eventually, tissue slices were stained with hematoxylin and eosin (H&E) for conventional light microscopy and photomicrography. Imaging Quantitative Analyses Quantitative results of fluoromicroscopic images were obtained by manually drawing eight regions of interest per tumor (n ⫽ 7) in the areas of ablated and viable tissues. Before that, a distinction between ablated tissue, viable tumor, and normal liver was made based on the histologic examination of the H&E-stained slices. Mean fluorescence densities, corrected for background, were
109
VAN DE PUTTE ET AL
Academic Radiology, Vol 15, No 1, January 2008
Figure 2. Macroscopic photographs taken under ultraviolet light and corresponding photographs under tungsten light of livers bearing ethanol-treated RIF-1 tumors. Mice were intravenously injected with 10 mg/kg hypericin at 24 hours after (a,b) or 1 hour before (c,d) ethanol treatment. All mice were sacrificed at 24 hours after hypericin injection and excised livers were digitally photographed. Ablated tumor (*), viable tumor (o), normal liver (x), ablated liver (**). Scale bar ⫽ 0.5 cm.
obtained by means of a KS imaging software system (Carl Zeiss, Vision, Hallbergmoos, Germany). Ratios of fluorescence densities in the different regions were then calculated for each condition. Statistical analysis was performed using Prism 4.00, GraphPad Software (San Diego, CA). A two-way analysis of variance with Bonferroni test was performed and p-values of P ⬍ .05 were considered statistically significant.
RESULTS On macroscopic photographs taken under ultraviolet light (Fig 2), ablated tumor, residual viable tumor, and normal liver showed different levels of fluorescence intensity in a decreasing order. The effect was prominent either when mice received hypericin after the ethanol treatment of the tumor mass (Fig 2a,b), or when hypericin was given before (Fig 2c,d). During the 24 hours after hypericin injection, the normal liver was predominantly cleared from hypericin in contrast to viable tumor tissue, which showed moderately elevated hypericin fluorescence intensity at 24 hours after injection. Fluoromicroscopic images revealed a distinct fluorescence intensity between necrotic tissue, residual tumor and normal liver (Fig 3a,c,d). Necrotic areas displayed the highest fluorescence intensity, followed by viable tumor
110
with moderate fluorescence and normal liver with the lowest intensity. No-entry zones for hypericin, corresponding with large necrotic areas, were regularly seen when hypericin was injected 24 hours after ethanol treatment (Fig 3b). Only the border zone between the ablated tissue and normal liver or viable tumor showed high uptake of hypericin, resulting in a fluorescent rim surrounding the no-entry zone. On H&E stains, this no-entry zone consisted mainly of dehydrated tissue with signs of severe cell damage and loss of tissue structure. The rim surrounding the no-entry zone had a typical appearance of eosinophilic coagulation tissue. Ratios of mean fluorescence densities in different areas of ablated tissue, viable tumor, and normal liver are shown in Table 1. Overall, ratios indicate three different levels of fluorescence density in necrotic tissue, viable tumor, and normal liver in decreasing order (P ⬍ .05). Mean fluorescence densities were not significantly different when hypericin was injected 24 hours after or 1 hour before ethanol treatment (P ⬎ .05).
DISCUSSION In the context of the evaluation of necrosis avid agents for use as short-term therapeutic response indicators after necrosis-inducing treatments such as PEI, we conducted
Academic Radiology, Vol 15, No 1, January 2008
HYPERICIN: A VIABILITY MARKER AFTER PEI
Figure 3. Fluorescence photomicrographs and corresponding hematoxylin and eosin stains of 5-m slices of livers bearing ethanol-treated RIF-1 tumors. Mice were intravenously injected with 10 mg/kg hypericin at 24 hours after (a,b) or 1 hour before (c,d) ethanol treatment. All mice were sacrificed at 24 hours after hypericin injection. Ablated tissue (NE), dehydrated ablated tissue (NE*), viable tumor (T), normal liver (L). Scale bar ⫽ 100 m.
111
VAN DE PUTTE ET AL
Academic Radiology, Vol 15, No 1, January 2008
Table 1 Ratios of Mean Fluorescence Densities in Different Regions of Fluoromicroscopic Images Taken From 5-m Tumor Sections ⌬T
NE Tissue Viable T
NE Tissue Normal L
Viable T Normal L
24 h 1h
5.53 ⫾ 2.29 (ns) 3.54 ⫾ 1.12
15.20 ⫾ 8.89 (ns) 13.13 ⫾ 5.82
3.84 ⫾ 1.24 (ns) 3.05 ⫾ 1.86
Tumors were excised from mice that received 10 mg/kg hypericin either 24 h after (⌬ T ⫽ 24 h), or 1 h before ethanol treatment (⌬ T ⫽ 1 h). Statistical analysis, comparing (⌬ T ⫽ 24 h) with (⌬ T ⫽ 1 h) was performed using Prism 4.00, GraphPad Software (San Diego, CA). A two-way analysis of variance with Bonferroni test was performed and P values of P ⬍ .05 were considered statistically significant. NE: necrotic; T: tumor; L: liver; ns: non-significant.
this preclinical study exploiting the necrosis-avid property of hypericin. For this purpose, we designed a murine RIF-1 liver tumor model by implanting a small tissue chip of a subcutaneously grown tumor in the liver. Unlike other animal studies that recruit subcutaneously implanted tumors, this murine hepatic tumor model may better mimic clinical conditions including visceral blood perfusion, parenchymal tumor growth, and relevant microenvironment, hence providing more clinically pertinent information for cancer diagnosis and therapy. Other advantages of the method employed include successful tumor growth in 84% of cases, rapid tumor growth in about 2 weeks, survival rate higher than 90% until the end of the experiment, minimally metastatic potential, and uncomplicated implantation procedure. In the present study, microscopic images revealed significant differences in hypericin fluorescence in three areas of interest (ie, necrosis, viable tumor, and normal liver tissue). Necrotic areas displayed an outstanding fluorescence signal, followed by moderate fluorescence intensity in viable tumor and low intensity in normal liver. Thus the three different tissue components could be visually stratified, even as early as 24 hours after intratumoral ethanol injection, simulating the results obtained with a NACA on MRI in a rat model treated with radiofrequency ablation (11). In a recent study exploring the mechanism of hypericin’s avidity for necrosis, we found that hypericin specifically stained necrotic tissue through compound-dependent interactions with necrotic cellular debris, whereas viable tissues were nonspecifically labeled to a lesser extent (Van de Putte M et al, submitted, 2007, to be cross-referenced upon publication). The moderate but persistent accumulation of hypericin in viable tumor 24 hours after injection can therefore be explained mainly
112
by the nonspecific uptake of hypericin by the tumor cells or stroma and a lack of biliary excretion in the tumor, similar to the phenomenon reported for MRI hepatobiliary contrast agents (20). In normal liver, hypericin was substantially cleared within 24 hours from functioning hepatic metabolic activities and biliary excretion. In this study, hypericin was intravenously administered both before and after intratumoral injection of ethanol. Both conditions achieved similar levels of fluorescence densities in different tissues (ie, high uptake of hypericin in necrotic areas) compared with moderate and low uptake in respectively viable tumor and normal liver. When hypericin was injected 1 hour before, maximal blood concentrations were achieved at the time of ethanol treatment (21), so that on ablation an outstanding extravasation took place in the entire necrotic area. Conversely, when hypericin was injected 24 hours after treatment, no-entry zones for hypericin were regularly seen. This phenomenon can be explained by the dehydration effect of alcohol in tissue. Ethanol acts by diffusing within the cells, which causes immediate dehydration of cytoplasmic proteins with consequent coagulation necrosis followed by fibrosis. This phenomenon may be described as in vivo “chemical denaturation.” In addition, necrosis of endothelial cells and platelet aggregation results in thrombosis of blood vessels followed by ischemia (5,22). As a result of the vascular shutdown, large lesions are not easily accessible to hypericin and subsequent diffusion throughout the entire necrotic lesion is hampered. When hypericin is injected after ethanol treatment, the compound will enter the necrotic zone through residual functioning blood vessels in the vicinity, followed by concentration gradient guided diffusion into the necrotic area, causing an evolving rim of intense fluorescence. Moreover, as suggested in an earlier study by Ni and coworkers (10), in vivo tissue degradation is required before interaction with NACAs (eg, hypericin) can take place. In contrast to traditional coagulation necrosis, in which enzymatic degradation plays a predominant role, instantaneous dehydration, and denaturation of cytosolic proteins prevents enzymatic degradation from occurring in tissue debris. Yet tissue degradation after ethanol-induced dehydration or denaturation requires a time-consuming process of days to weeks, in which the center of the ethanol-induced coagulative necrosis will be infiltrated by immune cells. After digestion of the dead tissue by enzymatic reactions, the necrotic debris will then be absorbed and replaced by granulation tissue.
Academic Radiology, Vol 15, No 1, January 2008
Before we can determine the efficacy of hypericin in detecting residual disease in a clinical setting and possibly observe an improved antitumoral outcome after necrosisinducing antineoplastic treatments such as PEI, a number of issues have to be settled. A first drawback of the compound hypericin is its potential skin phototoxicity. A direct infusion of hypericin into zones of necrosis within minutes before or after a PEI procedure, in combination with a transcatheter interventional procedure could offer advantages including increased local concentration while reducing its dosage, less systemic exposure and, improved patient convenience. A second drawback is the requirement of invasive procedures for fluorescence imaging of lesions in visceral organs. Applying a laparoscopic approach, which has been increasingly adopted in clinical practice, may reduce its invasiveness. Last, real noninvasive imaging can only be realized by tagging hypericin with a marker for a particular imaging modality, for instance, a gadolinium chelate for MRI. However, potential phototoxicity of hypericin derivatives at millimolar doses may limit the use for contrast enhanced MRI and it remains to be investigated whether produced signals are strong enough for the stratification of tumors in nonnecrotic, necrotic, and severe necrotic regions. Alternatively, hypericin may be tagged with a radionuclide for nuclear imaging. Positron emission tomography using hypericin labeled with iodine-124 could circumvent phototoxicity since only nanomolar amount of the compound is administered. Effective imaging using positron emission tomography appears to be feasible if the relatively low positron abundance of iodine-124 is compensated by the high ratio of tracer uptake in necrosis to surrounding tissues. In conclusion, the results of the present study suggest that hypericin offers significant potential in the early assessment of response following PEI. Hypericin features the property to specifically enhance the imaging contrast between necrotic and viable tissues and to nonspecifically distinguish viable tumor from normal liver. Hypericin injection before treatment appears more favorable than after treatment, since this approach circumvents difficulties with no-entry zones for hypericin and requires shorter intervals between ethanol treatment and imaging. We believe that our results have shown sufficient promise to justify continued research. Labeling hypericin to extend its use to MRI could be a first step in this direction.
HYPERICIN: A VIABILITY MARKER AFTER PEI
REFERENCES 1. Steele G, Ravikumar T. Resection of hepatic metastases from colorectal cancer. Biologic perspective. Ann Surg 1989; 210:127–138. 2. Livraghi T, Festi D, Monti F, et al. US-guided percutaneous alcohol injection of small hepatic and abdominal tumors. Radiology 1986; 161: 309 –312. 3. Shiina S, Yasuda H, Muto H, et al. Percutaneous ethanol injection in the treatment of liver neoplasms. Am J Roentgenol 1987; 149:949 –952. 4. Bartolozzi C, Lencioni R, Caramella D, et al. Hepatocellular carcinoma: CT and MR features after transcatheter arterial embolization and percutaneous ethanol injection. Radiology 1994; 191:123–128. 5. Lin X, Lin L. Local injection therapy for hepatocellular carcinoma. Hepatobil Pancreat Dis Int 2006; 5:16 –21. 6. Vogl T, Eichler K, Zangos S, et al. Hepatocellular carcinoma: role of imaging diagnostics in detection, intervention and follow-up. Rofo 2002; 174:1358 –1368. 7. Bartolozzi C, Lencioni R. Ethanol injection for the treatment of hepatic tumours. Eur Radiol 1996; 6:682– 696. 8. Nagel H, Bernardino M. Contrast-enhanced MR imaging of hepatic lesions treated with percutaneous ethanol ablation therapy. Radiology 1993; 189:265–270. 9. Cioni D, Lencioni R, Bartolozzi C. Percutaneous ablation of liver malignancies: imaging evaluation of treatment response. Eur J Ultrasound 2001; 13:73–93. 10. Ni Y, Bormans G, Chen F, et al. Necrosis avid contrast agents: functional similarity versus structural diversity. Invest Radiol 2005; 40:526 – 535. 11. Ni Y, Chen F, Mulier S, et al. Magnetic resonance imaging after radiofrequency ablation in a rodent model of liver tumor: tissue characterization using a novel necrosis avid contrast agent. Eur Radiol 2006; 16: 1031–1040. 12. Giese A. Hypericism. Photochem Photobiol Rev 1980; 5:229 –255. 13. Kartnig T, Gobel I, Heydel B. Production of hypericin, pseudohypericin and flavonoids in cell cultures of various Hypericum species and their chemotypes. Planta Med 1996; 62:51–53. 14. D’Hallewin M, De Witte P, Waelkens E, et al. Fluorescence detection of flat bladder carcinoma in situ after intravesical instillation of hypericin. J Urol 2000; 164:349 –351 15. Kamuhabwa A, Agostinis P, Ahmed B, et al. Hypericin as a potential phototherapeutic agent in superficial transitional cell carcinoma of the bladder. Photochem Photobiol Sci 2004; 3:772–780. 16. Ni Y, Huyghe D, Verbeke K, et al. First preclinical evaluation of mono[123I]iodohypericin as a necrosis-avid tracer agent. Eur J Nucl Med Mol Imaging 2006; 33:595– 601. 17. Twentyman P, Brown J, Gray J, et al. A new mouse tumor model system (RIF-1) for comparison of end-point studies. J Natl Cancer Inst 1980; 64:595– 604. 18. Chen F, Sun X, De Keyzer F, et al. Liver tumor model with implanted rhabdomyosarcoma in rats: MR imaging, microangiography, and histopathologic analysis. Radiology 2006; 239:554 –562. 19. Falk H, Meyer J, Oberreiter M. A convenient semisynthetic route to hypericin. Monatsh Chem 1993; 124:339 –341. 20. Ni Y, Marchal G. Enhanced magnetic resonance imaging for tissue characterization of liver abnormalities with hepatobiliary contrast agents: an overview of preclinical animal experiments. Topics in Magn Reson Imaging 1998; 9:183–195. 21. Chen B, Xu Y, Roskams T, Delaey E, et al. Efficacy of antitumoral photodynamic therapy with hypericin: relationship between biodistribution and photodynamic effects in the RIF-1 mouse tumor model. Int J Cancer 2001; 93:275–282. 22. Guan Y, Sun L, Zhou X, et al. Hepatocellular carcinoma treated with interventional procedures: CT and MRI follow-up. World J Gastroenterol 2004; 10:3543–3548.
113