- Email: [email protected]
Relative Biological Effectiveness of Carbon Ions for Local Tumor Control of a Radioresistant Prostate Carcinoma in the Rat
Int. J. Radiation Oncology Biol. Phys., Vol. 79, No. 1, pp. 239–246, 2011 Copyright Ó 2011 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/$ - see front matter
doi:10.1016/j.ijrobp.2010.07.1976
BIOLOGY CONTRIBUTION
RELATIVE BIOLOGICAL EFFECTIVENESS OF CARBON IONS FOR LOCAL TUMOR CONTROL OF A RADIORESISTANT PROSTATE CARCINOMA IN THE RAT PETER PESCHKE, PH.D.,* CHRISTIAN P. KARGER, PH.D.,y MICHAEL SCHOLZ, PH.D.,z JU¨RGEN DEBUS, M.D., PH.D.,x AND PETER E. HUBER, M.D., PH.D.*x *Clinical Cooperation Unit Radiation Oncology and yDepartment of Medical Physics in Radiation Oncology, German Cancer Research Center, Heidelberg, Germany; zDepartment of Biophysics, Helmholtz Center for Heavy Ion Research, Darmstadt, Germany; and x Department of Clinical Radiology, University of Heidelberg, Heidelberg, Germany Purpose: To study the relative biological effectiveness (RBE) of carbon ion beams relative to X-rays for local tumor control in a syngeneic rat prostate tumor (Dunning subline R3327-AT1). Methods and Materials: A total of 198 animals with tumors in the distal thigh were treated with increasing single and split doses of either 12C ions or photons using a 20-mm spread-out Bragg peak. Endpoints of the study were local control (no tumor recurrence within 300 days) and volumetric changes after irradiation. The resulting values for D50 (dose at 50% tumor control probability) were used to determine RBE values. Results: The D50 values for single doses were 32.9 ± 0.9 Gy for 12C ions and 75.7 ± 1.6 Gy for photons. The respective values for split doses were 38.0 ± 2.3 Gy and 90.6 ± 2.3 Gy. The corresponding RBE values were 2.30 ± 0.08 for single and 2.38 ± 0.16 for split doses. The most prominent side effects were dry and moist desquamation of the skin, which disappeared within weeks. Conclusion: The study confirmed the effectiveness of carbon ion therapy for severely radioresistant tumors. For 1- and 2-fraction photon and 12C ion radiation, we have established individual D50 values for local tumor control as well as related RBE values. Ó 2011 Elsevier Inc. Carbon ion radiotherapy, Dose–response, Dunning prostate tumor R3327, Local tumor control, Relative biological effectiveness.
INTRODUCTION
DNA repair, leading to enhanced cell inactivation, which is assumed to be less dependent on genetic disposition, as well as on oxygen and cell cycle status (17–20). In carbon ion radiotherapy, dose prescriptions refer to biological effective doses, and for their calculation tissuespecific RBEs have to be assigned to each point. To estimate the RBE for normal and neoplastic tissues at each point, a radiobiological model (21) was integrated into the treatment planning system (TPS). Unfortunately, RBE is a complex quantity, depending on physical parameters such as particle type, dose per fraction, and linear energy transfer (LET), as well as on biological factors like cell or tissue type and the selected biological endpoint. Because the RBE for clinical situations involves rather large uncertainties, biological systems were used to evaluate RBE relationships at various depths and their dependence on dose per fraction. Most of
Worldwide, the interest in carbon ion radiotherapy is increasing. After pioneering clinical studies at the Lawrence Berkley Laboratory (University of California) for several ion types (1, 2), a clinical research program was initiated at the National Institute of Radiologic Sciences in Chiba, Japan in 1994 (3–8). In 1997, carbon ion radiotherapy started at the Gesellschaft fu¨r Schwerionenforschung (Darmstadt, Germany) and reported promising results (9–13). Carbon ions exhibit an inverted depth–dose profile, which allows for precise irradiation of deep-seated targets (14, 15). In addition to these physical advantages, carbon ions are characterized by an enhanced relative biological effectiveness (RBE), which is higher in the Bragg peak than in the beam entrance region (16). Mechanistically, carbon ions cause clustered DNA damage, which aggravates
Acknowledgments—The authors thank Alexandra Tietz and Angela Funk (German Cancer Research Center) and Thilo Elsa¨sser (Gesellschaft fu¨r Schwerionenforschung [GSI] Helmholtz Center for Heavy Ion Research) for their excellent technical support; E.W. Hahn for his valuable advice and on-site assistance with study design and setup of the radiation procedure; and GSI for providing excellent experimental conditions with the highly sophisticated beam scanning technique.
Reprint requests to: Christian Karger, Ph.D., Department of Medical Physics in Radiation Oncology (E040), German Cancer Research Center, Im Neuenheimer Feld 280, 69120 Heidelberg, Germany. Tel: (+49) 6221-56-38965; Fax: (+49) 6221-56-4631; E-mail: [email protected] Supported in part by Bundesministerium fu¨r Bildung und Forschung (KVSF, 03NUK004A,C, and 06HD156). P.P. and C.P.K. contributed equally to this work. Conflict of interest: none. 239
I. J. Radiation Oncology d Biology d Physics
240
our present knowledge about tolerance doses and dose dependence of RBEs was derived from experiments with mouse skin (22) and jejunum (23), as well as from rat brain and spinal cord (24–27). Carbon ion effects in experimental solid tumors are addressed by only a few studies using human tumor cell line xenotransplants or syngeneic mouse models with tumor growth inhibition as biological endpoint (28–31). The present study investigates the response of wellcharacterized tumor sublines to carbon ion therapy. We have selected a syngeneic tumor system consisting of several cell lines, which represent the spectrum of androgen responsiveness, tumorigenicity, and metastatic ability seen during the progression of human prostate cancer (32). Growth delay and the probability of local tumor control in relation to locoregional side effects were selected as clinically significant biological endpoints. METHODS AND MATERIALS Tumor model The anaplastic subline AT1 of the Dunning prostate adenocarcinoma R3327 was subcutaneously implanted in the distal thigh of young adult male Copenhagen rats (180 g). This syngenic tumor
Volume 79, Number 1, 2011
is considered an excellent model system for human prostate tumor and has been extensively characterized in terms of ultrastructure, hormones, response to therapy, and immunogenicity (33). Tumor material was kept identical for all experiments by transplanting fresh tissue from tumors grown from a cryopreserved stock, maintained as a first passage of the original tumor tissue kindly supplied to us by J. T. Isaacs (Johns Hopkins University, Baltimore, MD). Ploidy status, histology, and the growth rate of control tumors were determined as measures of quality assurance. All experiments were approved by the governmental review committee on animal care, and animals were kept under standard laboratory conditions at the German Cancer Research Center. During all irradiation procedures, rats were deeply anesthetized by i.p. injection of Ketaminehydrochloride (125 mg/kg) mixed with Xylazinehydrochloride (20 mg/kg).
Experimental setup Average diameter of the tumors before treatment was 9 mm (range, 7.0–10.5 mm). Tumors larger than 10.5 mm in one direction were excluded from the study. Irradiations were performed with a single horizontal beam that hit the tumor from the rear direction. Rats were placed on a specially designed device allowing for rigid and reproducible fixation of the tumor relative to a milled marker on the device (Fig. 1a and b), and the device was aligned on the treatment table with crossed laser beams. In a setup experiment, the
Fig. 1. (a) Setup of the immobilization device for irradiation (here for photons) for the local treatment of s.c. growing tumors (b). Irradation was performed horizontally from the rear direction. (c) The clinical target volume (red) was extended horizontally by a 3-mm and vertically by a 2-mm margin to obtain the planning target volume (PTV) (blue). The resulting PTV extension was 16 14 mm2. (d, e) Sagittal views of the dose distributions in water together with the clinical target volume (PTV not shown) for (d) photons (95%, 90%, 80%, 50%, and 20% isodoses) and (e) carbon ions (95%, 90%, 80%, 70%, 50%, and 20%).
RBE of carbon ions for local tumor control d P. PESCHKE et al.
reproducibility of the tumor position was determined from X-ray images to be 1 mm (1 SD). The planning target volume (PTV) was defined as a cube (clinical target volume, CTV) containing the tumor (<10 mm in diameter) plus a 3-mm margin in the horizontal and a 2-mm margin in the vertical direction accounting for setup variability and variations in tumor size (Fig. 1c). The larger margin in the horizontal direction was selected because of the lack of a clear borderline between tumor and underlying muscle. As a result, the extension of the PTV was 16 14 mm2.
Irradiation technique Photon irradiation was delivered by a linear accelerator (Siemens, Erlangen, Germany) using a single 6-MV horizontal beam shaped by a cylindrical tungsten collimator with 12-mm aperture. Calculated diameter of the 90% and 80% isodoses at isocenter was 15 and 17 mm, respectively (Fig. 1d; TPS: STP [StrykerLeibinger, Freiburg, Germany]). Distance between source and center of the tumor was 99 cm. To achieve a homogeneous depth–dose profile in the tumor, a 10- and 15-mm polymethylmethacrylate (PMMA) bolus was placed directly in front of and behind the tumor, respectively. Dose rate under reference conditions was 2.5–3 Gy/min. The treatment field was calibrated using a pinpoint ion chamber (TM31009; PTW, Freiburg, Germany) placed in the middle of a 10-mm-thick plane-parallel build-up cap fixed at the tumor position within the device. Carbon ion irradiation was performed at the medical beam line at Gesellschaft fu¨r Schwerionenforschung using the intensitycontrolled raster scanning system (34). Irradiations were performed with a homogeneous 20-mm spread-out Bragg peak (SOBP) (scan steps in x, y, z: 1 mm; full width at half maximum of pencil beam: 5 mm) optimized in terms of absorbed dose (TPS: TRiP [35]). The SOBP ranged from 41.5 to 61.5 mm water-equivalent depth (corresponding to initial beam energies between 140 and 171 MeV/u defined by the synchrotron), and the center of the tumor was placed in the middle of the SOBP (Fig. 2). Mean dose average LET in the tumor was 75 keV/mm (range, 64–96 keV/mm). To adjust the range of the ions, a 39.9-mm PMMA bolus was placed directly in front of the tumor. In addition, a 15-mm PMMA bolus was placed directly behind the tumor. Diameter of the calculated 90% (80%) isodose was 16.5 mm (18.5 mm) in the lateral and 14.6 mm (16.8 mm) in the vertical direction (Fig. 1e). The calculated dose in the middle
of the SOBP was verified in a water phantom using a pinpoint ion chamber (TM31009; PTW). A deviation of 0.5% was found.
Dose levels, fractionation scheme, and follow-up A total of 198 rats were irradiated with single and split doses of either photons or carbon ions using a series of increasing dose levels (Table 1). Irradiations with 2 fractions were performed on consecutive days. Sham-treated animals served as controls. During the first cohort of the split-dose experiment for carbon ions, a technical problem with the beam delivery software led to a break of 48 rather than 24 h. Because this unintentionally prolonged break might lead to a modified radiation response, the experiment was repeated with a regular 24-h break. The experiment with the accidental 48-h break was nevertheless evaluated to study the impact of a prolonged break related to treatment outcome. The primary endpoint of this study was local tumor control, defined as no indication of tumor recurrence within an observation time of 300 days after irradiation. In addition, the volume of the tumors was measured routinely three times weekly (Monday, Wednesday, Friday), and growth curves were established. The tumor volume was calculated by the formula 4/3 p r3, with r being the mean of two orthogonal tumor radii. Because the growth patterns of locally controlled and locally noncontrolled tumors are quite different, averaged growth curves were established separately for locally controlled and locally noncontrolled tumors for the mixed dose groups.
Statistics For each dose level, the incident rates x/n (x out of n tumors are controlled) were determined, and the logistic dose response model PðDÞ ¼
eb0 þb1 D 1 þ eb0 þb1 D
(1a)
Table 1. Dose levels and number of animals evaluated for each experiment Experiment 1 fraction Photons Carbon ions 2 fractions Photons Carbon ions Carbon ions (48-h break)z Controls Total
Fig. 2. Depth–dose and dose average linear energy transfer (LET) profile used for the carbon ion irradiations. The position of the tumor within the spread-out Bragg peak is indicated.
241
Dose levels (Gy)
Animals evaluated
30, 40, 50, 60, 70*, 75, 77.5*, 80y, 85, 90 9, 12.5, 16, 19.5, 23, 26.5, 30*, 33.5, 37*, 40.5
37
35y, 40, 50, 60, 70, 80, 85, 90, 95, 100* 8, 14, 20, 26, 32*, 38, 44, 50, 56 26, 32, 38, 44, 50
39
5 sham-treated tumors per experimentx
20
42
37 15
190
In total, the study contained 198 animals, from which 190 could be evaluated. * One animal was excluded from analysis because of metastatic activity. y One animal died after therapy because of unknown reasons. z An additional cohort was treated with a 48 h break between doses caused by technical reasons and was evaluated separately (see text). x The 5 controls of the 2-fraction experiment with carbon ions also served as controls for the experiment with the 48-h break.
242
I. J. Radiation Oncology d Biology d Physics
was adjusted separately for each radiation modality and fractionation schedule using the maximum likelihood fitting procedure of the statistic package Statistica (36). D is the total dose, and b0 and b1 are the adjusted fit parameters from which the tolerance dose D50 (dose at 50% tumor control probability) can be calculated: b0 D50 ¼ : b1
(1b)
The RBE was then calculated as the ratio of the D50 values for photons and carbon ions, respectively. In addition, 90% confidence limits were calculated for the RBE values using Fieller’s theorem (37).
RESULTS Spontaneous death occurred in 2 of 198 animals. Cause of death remained unclear even after pathologic inspection. Additional 6 animals (3%) were removed from the study because of metastatic activity in lymph nodes adjacent to the tumor. Lymph node migration was found in all irradiated groups, with a tendency to higher radiation doses. The observed low spontaneous metastatic ability is well known for the Dunning R3327-AT sublines (38), and thus no bias
Volume 79, Number 1, 2011
of observed incidences is to be expected. Finally, a total of 190 animals could be evaluated.
Growth delay Tumor growth delay was found to be strongly dependent on dose and on whether tumors were locally controlled (Fig. 3a–d). Independent of local control status, all growth curves revealed a continuous increase in tumor volume within the first 3 weeks after irradiation, reaching three to four times the initial tumor volume. Dependent on dose, this was followed either by further growth or by shrinkage of the tumor. The decrease in volume was followed by dose-dependent regrowth or local control of tumors. For both irradiation modalities, the maximum time until tumor regrowth occurred was 95 days after irradiation. Further dose escalation led to treatment groups, with increasing fractions of local tumor control for both radiation qualities. In the order of increasing radiation doses, the first single-dose group in which all tumors were controlled was found at 37 Gy for carbon ions and 85 Gy for photons. For 2 fractions,
Fig. 3. Growth response of the tumor to either (a, b) single (a, b) or (c, d) split (c, d) doses of photons (a, c) and carbon ions (b, d) (mean 1 SD). Mean values are calculated separately for tumors without (nLC, closed symbols) and with local control (LC, open symbols). For better visibility, only curves for selected dose levels are shown. Because no further growth occurred, the x-axis was truncated at 180 days.
RBE of carbon ions for local tumor control d P. PESCHKE et al.
243
Side effects Local hair loss within the field of irradiation was observed in all animals. Dry and moist desquamation of the skin (#1.5+ according to the normal damage scale previously published [39]) occurred as the most prominent side effect, which disappeared within 4–6 weeks after therapy. Radiation-induced side effects in the normal skin around the tumor were worse in animals in which proliferative activity was not controlled, presumably because skin recovery was constraint during increased volumetric tumor expansion. At the end of the observation period of 300 days, animals with locally controlled tumors exhibited slight skin erythema and mild dystrophy of the thigh muscle. Interestingly, a fibrotic nodule remained perceptible in most animals with local tumor control at the end of the observation period, which is known to be an ordinary process after curative radiation doses. DISCUSSION
Fig. 4. Dose–response curves for (a) 1 and (b) 2 fractions of photons and carbon ions. The uncertainty (1 SD) of D50 is indicated.
the respective doses were 50 Gy for carbon ions and 100 Gy for photons. ED50 values, fractionation effects, and RBEs Figure 4a and b display the adjusted dose–response curves for single and split doses of carbon ions and photons, respectively. The respective values for D50 and RBE are given in Table 2. The ratio of the doses for 1 vs. 2 fractions to achieve a D50 isoeffect for carbon ions and photons was 1.16 and 1.20, respectively. When the break between two fractions of carbon ions was extended from 24 to 48 h, D50 decreased from 38.0 2.3 Gy to 32.9 2.6 Gy (Fig. 5).
Although heavy ions exhibit favorable physical and biological properties, the role of carbon ion radiotherapy has still to be determined. For that purpose, two therapy centers in Japan (Heavy Ion Medical Accelerator, [HIMAC] in Chiba [8] and Hyogo Ion Beam Medical Center [40]) and one in Germany (Heidelberger Ionenstrahl-Therapiezentrum [HIT]) (41, 42) have been established to date, and several tumor sites, including prostate carcinoma (3–5), uveal melanoma, and hepatocellular carcinoma, as well as skull base, lung, and prostate tumors, are under evalua (43). Beside these clinical studies, systematic biological investigations are necessary to characterize the effectiveness of this new beam quality and to help integrating carbon ions into the arsenal of radiation treatment strategies. Whereas our previous studies investigated mainly carbon ion–related damage to normal central nervous system tissues (26, 27), the present experiments focused on the response of an experimental tumor system. The selected prostate tumor subline R3327-AT1 represents a hormone-independent anaplastic carcinoma with a labeling index of 7.0% 0.5% and 8.8% 3.7% measured with flow cytometry and histology, respectively, a S-phase duration of 8 h, a potential doubling time of 4.7 days, and a cell loss factor of 15% (44). Tumor oxygen tension as examined by 19F nuclear magnetic resonance spin-lattice relaxation rate of perfluorocarbon emulsion revealed a size-dependent development of generalized
Table 2. D50 and RBE values measured in this study D50 SE (90% CL) (Gy) Study
Photons
Carbon ions
RBE SE (90% CL)
1 fraction 2 fractions
75.7 1.6 (69.9–78.6) 90.6 2.3 (85.6–95.4)
32.9 0.9 (30.8–34.9) 38.0 2.3 (33.7–42.6)
2.30 0.08 (2.17–2.44) 2.38 0.16 (2.15–2.67)
Abbreviations: D50 = dose at 50% tumor control probability; RBE = relative biological effectiveness; CL = confidence limit; LET = linear energy transfer, SE = standard error. Endpoint: tumor control at 300 days, LET = 75 (64–96) keV/mm.
244
I. J. Radiation Oncology d Biology d Physics
Volume 79, Number 1, 2011
Acceptable side effects to normal tissue within the treatment field enabled the extension of the study to the preferred biological endpoint of local tumor control by dose escalation. This endpoint depends only on the inactivation of clonogenic tumor cells and is regarded as the most important assay for curative effects in radiotherapy (52). In addition, D50 for local tumor control is a robust endpoint parameter, especially for measuring RBE values.
Fig. 5. Dose–response curves for 2 fractions of carbon ions applied either within 24 h or 48 h. The uncertainty (1 SD) of D50 is indicated. Note that the 48-h experiment was not intended and resulted from problems with the beam delivery software (see text). The photon curve for the regular split-dose regimen with a 24-h break is shown in Fig. 4b.
central hypoxia (45). Nevertheless, the degree of hypoxia is not sufficient to induce central necrosis, as is often seen in the more rapidly proliferating rodent tumors. Chronic hypoxia is one of the reasons why this syngeneic tumor subline exhibited a high radiation resistance, as documented in previous animal studies with various radiation modalities (46–48). This resistance is also known from cell experiments with X-rays, in which a values of 0.144 and 0.068 Gy1 were measured for oxic and hypoxic conditions, respectively, leading to an oxygen enhancement ratio of 2.35 (49). The selected tumor subline has a low metastatic potential. In our experiment we had 6 cases of lymph node metastasis. Because of this very small number, the dependence of these events on dose (and potentially LET) could not be reliably analyzed. For this, a dedicated experimental design is required using a tumor model with a higher metastatic potential. Such a study was carried out by Tamaki et al. (50) for the murine squamous cell carcinoma NR-S1. This study showed that small doses of photons or carbon ions inhibit the incidence of lung metastasis similarly. In addition, the transcriptional profile of the metastasis was independent of the application of radiotherapy per se, as well as the radiotherapy type. Biological endpoints In accordance with previous studies, growth delay was found to increase with increasing dose (50, 51), reflecting an increasing number of inactivated clonogenic cells. Extreme differences in growth delay within the three noncontrolled tumors were the reason for the large heterogeneity in the group treated with a single dose of 33.5 Gy carbon ions (Fig. 3b). A similar heterogeneity was also seen for a few other dose/modality groups near the respective D50 group (not displayed for visibility reasons). A definitive explanation for this behavior, however, could not be identified.
Split doses for carbon ions For the split-dose experiment with 12C ions, the dose– response curve was shifted to slightly lower doses, when the time interval between fractions was increased from 24 h to 48 h. Because the respective D50 values differ by less than 2 combined SDs, this shift is considered not statistically significant. It is therefore concluded that the prolonged time between fractions does not lead to significantly increased repair and proliferation. Rather, the shift may have occurred by chance. An alternative explanation might be improved reoxygenation for the 48-h experiment, given that a study by Ando et al. (53) showed accelerated reoxygenation of a murine fibrosarcoma after carbon ion radiation. While NFSa tumors reoxygenated 4 days after 30 Gy of X-ray irradiation, reoxygenation occurred as early as 1 day after 16 Gy of carbon ions measured by paired growth delay assay. These results have been confirmed in a second study, in which oxygen partial pressure profiles were measured in the same tumor system using coaxial high spatial resolution oxygen microelectrodes (54). RBE and fractionation ratio Our results clearly show that physical doses necessary to locally control this anaplastic prostate carcinoma are considerably lower for carbon ions than for photons. Additionally, only a very small fractionation effect for both beam qualities was found. The D50 isoeffect ratio of 1.20 for 2 vs. 1 fractions of photons is not markedly different from that of 1.16 for carbon ions. Our RBE values of 2.30 0.08 (1 fraction) and 2.38 0.16 (2 fractions) are in line with the results published by others. From in vitro data on the same cell line irradiated with carbon ions at different LETs (49), the RBE for 10% clonogenic cell survival can be linearly interpolated to be 2.2 at 75 keV/mm, which is close to our in vivo RBE value for the single-fraction experiment. A large series of studies with tumor cell lines has shown that high LET radiations are more effective than low LET radiations, such as X-rays or g rays, for the endpoint cell kill. The most comprehensive collection was published by Suzuki et al. (55), summarizing RBE values for 10% clonogenic cell survival in 16 different tumor cell lines. The RBEs obtained for a dose-averaged LET of approximately 77 keV/mm ranged between 2.00 and 3.01. In an in vivo study, an RBE of 2.02 was determined for 290 MeV/u 12C beam with an LET of 70 keV/mm in a human esophageal cancer (ESO-2) xenotransplant. The RBE was calculated from the slope of the dose–response curve for
RBE of carbon ions for local tumor control d P. PESCHKE et al.
tumor growth suppression at 4 weeks after irradiation with X-rays or carbon ion beams (28). The only study determining RBEs for animal tumors treated with fractionated doses of 290 MeV/u carbon ions was performed with the NFSa fibrosarcoma model in mice. In this study, a clear relationship between RBE and LET was documented. For 14 and 20 keV/mm, RBEs were 1.4 and independent of the number of fractions, whereas those for 44 and 74 keV/mm increased from 1.8 to 2.3 and from 2.4 to 3.0, respectively, when the number of fractions changed from 1 to 4. Raising the number of fractions from 4 to 6 was not associated with a further increase in RBE (31). The single-dose RBE of 2.4 at 74 keV/mm is well comparable to the value obtained in our study; however, our study revealed a much smaller increase of RBE when going from singe to split doses. This might be explained by the different applied endpoints. Whereas Koike et al. (31) define isoeffectivity (and hence RBE) by equal tumor growth delay for subtherapeutic doses, our RBEs are based on local control at therapeutic doses (i.e., at the level of D50). The RBEs determined from equal tumor growth delay at 1 and 2 fractions might therefore not correspond to the same tumor control probability at much higher doses, which is the most important endpoint in clinical application. A direct comparison of the RBEs between these two studies is therefore difficult. The reason for the differential response of the Dunning tumor subline R3327-AT1 with respect to the two radiation qualities is presently not clearly understood. The underlying
245
molecular mechanisms are multifactorial, and studies are in progress by many investigators (for review see Blakely and Chang [56]). Microarray analysis has been applied to identify genes involved in regulatory pathways that result in the development of resistance to therapy. Nojiri et al. (51), for example, investigated radiation-induced changes in gene expression related to angiogenesis, which influence responsiveness of tumors by the modulation of microenvironmental conditions. In some cases, these radiation-induced changes even seem to depend on the radiation type. For future preclinical assessment of carbon ion efficacy, such information is not only the prerequisite to define tumor models with respect to well-characterized features of radiation resistance, such as dysregulated cell death pathways and upregulated survival factors, but this knowledge may also have a predictive value for both patient selection and therapy outcome.
CONCLUSION The present study confirmed the effectiveness of carbon ion radiotherapy for severely radioresistant tumors and investigated an in vivo tumor model using local control as a clinically most relevant endpoint. Additional experiments with different fraction numbers and tumor sublines as well as comparison with normal tissue response are necessary to draw clinically relevant conclusions.
REFERENCES 1. Castro JR. Results of heavy ion radiotherapy. Radiat Environ Biophys 1995;34:45–48. 2. Castro JR, Linstadt DE, Bahary JP, et al. Experience in charged particle irradia of tumors of the skull base: 1977-1992. Int J Radiat Oncol Biol Phys 1994;29:647–655. 3. Akakura K, Tsujii H, Morita S, et al. Phase I/II clinical trials of carbon ion therapy for prostate cancer. Prostate 2004;58:252–258. 4. Ishikawa H, Tsuji H, Kamada T, et al. Carbon ion radiation therapy for prostate cancer: Results of a prospective phase II study. Radiother Oncol 2006;81:57–64. 5. Tsuji H, Yanagi T, Ishikawa H, et al. Hypofractionated radiotherapy with carbon ion beams for prostate cancer. Int J Radiat Oncol Biol Phys 2005;63:1153–1160. 6. Kanai T, Endo M, Minohara S, et al. Biophysical characteristics of HIMAC clinical irradiation system for heavy-ion radiation therapy. Int J Radiat Oncol Biol Phys 1999;44:201–210. 7. Kato H, Tsujii H, Miyamoto T, et al. Results of the first prospective study of carbon ion radiotherapy for hepatocellular carcinoma with liver cirrhosis. Int J Radiat Oncol Biol Phys 2004;59:1468–1476. 8. Tsujii H, Mizoe J, Kamada T, et al. Clinical results of carbon ion radiotherapy at NIRS. J Radiat Res (Tokyo) 2007;48(Suppl. A):A1–A13. 9. Schulz-Ertner D, Nikoghosyan A, Didinger B, et al. Therapy strategies for locally advanced adenoid cystic carcinomas using modern radiation therapy techniques. Cancer 2005;104: 338–344. 10. Schulz-Ertner D, Nikoghosyan A, Hof H, et al. Carbon ion radiotherapy of skull base chondrosarcomas. Int J Radiat Oncol Biol Phys 2007;67:171–177.
11. Schulz-Ertner D, Karger CP, Feuerhake A, et al. Effectiveness of carbon ion radiotherapy in the treatment of skull-base chordomas. Int J Radiat Oncol Biol Phys 2007;68:449–457. 12. Schulz-Ertner D. The clinical experience with particle therapy in adults. Cancer J 2009;15:306–311. 13. Combs SE, Nikoghosyan A, Jaekel O, et al. Carbon ion radiotherapy for pediatric patients and young adults treated for tumors of the skull base. Cancer 2009;115:1348–1355. 14. Kraft G. Radiotherapy with heavy ions: Radiobiology, clinical indications and experience at GSI Darmstadt. Tumori 1998;84: 200–204. 15. Karger CP, Ja¨kel O. Current status and new developments in ion therapy. Strahlenther Onkol 2007;183:295–300. 16. Blakely EA, Kronenberg A. Heavy-ion radiobiology: New approaches to delineate mechanisms underlying enhanced biological effectiveness. Radiat Res 1998;150:S126–S145. 17. Tsuboi K, Tsuchida Y, Nose T, et al. Cytotoxic effect of accelerated carbon beams on glioblastoma cell lines with p53 mutation: Clonogenic survival and cell-cycle analysis. Int J Radiat Biol 1998;74:71–79. 18. Weyrather WK, Ritter S, Scholz M, et al. RBE for carbon tracksegment irradiation in cell lines of differing repair capacity. Int J Radiat Biol 1999;75:1357–1364. 19. Staab A, Zukowski D, Walenta S, et al. Response of Chinese hamster v79 multicellular spheroids exposed to high-energy carbon ions. Radiat Res 2004;161:219–227. 20. Fournier C, Wiese C, Taucher-Scholz G. Accumulation of the cell cycle regulators TP53 and CDKN1A (p21) in human fibroblasts after exposure to low- and high-LET radiation. Radiat Res 2004;161:675–684.
246
I. J. Radiation Oncology d Biology d Physics
21. Scholz M, Kellerer AM, Kraft-Weyrather W, et al. Computation of cell survival in heavy ion beams for therapy. The model and its approximation. Radiat Environ Biophys 1997;36:59–66. 22. Ando K, Koike S, Uzawa A, et al. Repair of skin damage during fractionated irradiation with gamma rays and low-LET carbon ions. J Radiat Res (Tokyo) 2006;47:167–174. 23. Fukutsu K, Kanai T, Furusawa Y, et al. Response of mouse intestine after single and fractionated irradiation with accelerated carbon ions with a spread-out Bragg peak. Radiat Res 1997; 148:168–174. 24. Leith JT, McDonald M, Powers-Risius P, et al. Response of rat spinal cord to single and fractionated doses of accelerated heavy ions. Radiat Res 1982;89:176–193. 25. Karger CP, Debus J, Peschke P, et al. Dose-response curves for late functional changes in the normal rat brain after single carbon-on doses evaluated by magnetic resonance imaging: Influence of follow-up time and calculation of relative biological effectiveness. Radiat Res 2002;158:545–555. 26. Karger CP, Peschke P, Sanchez-Brandelik R, et al. Radiation tolerance of the rat spinal cord after 6 and 18 fractions of photons and carbon ions: Experimental results and clinical implications. Int J Radiat Oncol Biol Phys 2006;66:1488–1497. 27. Debus J, Scholz M, Haberer T, et al. Radiation tolerance of the rat spinal cord after single and split doses of photons and carbon ions. Radiat Res 2003;160:536–542. 28. Takahashi A, Yano T, Matsumoto H, et al. Effects of accelerated carbon-ions on growth inhibition of transplantable human esophageal cancer in nude mice. Cancer Lett 1998;122:181–186. 29. Kitabayashi H, Shimada H, Yamada S, et al. Synergistic growth suppression induced in esophageal squamous cell carcinoma cells by combined treatment with docetaxel and heavy carbon-ion beam irradiation. Oncol Rep 2006;15:913–918. 30. Asakawa I, Yoshimura H, Takahashi A, et al. Radiationinduced growth inhibition in transplanted human tongue carcinomas with different p53 gene status. Anticancer Res 2002;22: 2037–2043. 31. Koike S, Ando K, Oohira C, et al. Relative biological effectiveness of 290 MeV/u carbon ions for the growth delay of a radioresistant murine fibrosarcoma. J Radiat Res (Tokyo) 2002;43: 247–255. 32. Tennant TR, Kim H, Sokoloff M, et al. The Dunning model. Prostate 2000;43:295–302. 33. Isaacs JT, Heston WD, Weissman RM, et al. Animal models of the hormone-sensitive and -insensitive prostatic adenocarcinomas, Dunning R-3327-H, R-3327-HI, and R-3327-AT. Cancer Res 1978;38:4353–4359. 34. Haberer T, Becher W, Schardt D, et al. Magnetic scanning system for heavy ion therapy. Nucl Instrum Meth A 1993;A330: 296–305. 35. Kra¨mer M, Ja¨kel O, Haberer T, et al. Treatment planning for heavy-ion radiotherapy: Physical beam model and dose optimization. Phys Med Biol 2000;45:3299–3317. 36. StatSoft Inc. Statistica for Windows 7.1. Tulsa, OK: StatSoft; 2005. 37. Finney DJ. Statistical method in biological assay. 3rd ed. London: Charles Griffin; 1978. 38. Isaacs JT, Isaacs WB, Feitz WF, et al. Establishment and characterization of seven Dunning rat prostatic cancer cell lines and their use in developing methods for predicting metastatic abilities of prostatic cancers. Prostate 1986;9:261–281.
Volume 79, Number 1, 2011
39. Ohta T, Iwakawa M, Oohira C, et al. Fractionated irradiation augments inter-strain variation of skin reactions among three strains of mice. J Radiat Res (Tokyo) 2004;45:515–519. 40. Hishikawa Y, Oda Y, Mayahara H, et al. Status of the clinical work at Hyogo. Radiother Oncol 2004;73(Suppl. 2):S38–S40. 41. Haberer T, Debus J, Eickhoff H, et al. The Heidelberg Ion Therapy Center. Radiother Oncol 2004;73(Suppl. 2):S186– S190. 42. Combs SE, Jakel O, Haberer T, et al. Particle therapy at the Heidelberg Ion Therapy Center (HIT)—integrated researchdriven university-hospital-based radiation oncology service in Heidelberg, Germany. Radiother Oncol 2010;95:41–44. 43. Schulz-Ertner D, Tsujii H. Particle radiation therapy using proton and heavier ion beams. J Clin Oncol 2007;25:953–964. 44. Lohr F, Wenz F, Flentje M, et al. Measurement of the proliferative activity of three different sublines of the Dunning rat prostate tumor R3327. Strahlenther Onkol 1993;169:438–445. 45. Mason RP, Antich PP, Babcock EE, et al. Non-invasive determination of tumor oxygen tension and local variation with growth. Int J Radiat Oncol Biol Phys 1994;29:95–103. 46. Peschke P, Hahn EW, Wenz F, et al. Differential sensitivity of three sublines of the rat Dunning prostate tumor system R3327 to radiation and/or local tumor hyperthermia. Radiat Res 1998; 150:423–430. 47. Harms W, Peschke P, Weber KJ, et al. Dose-dependent differential effects of low and pulsed dose-rate brachytherapy in a radioresistant syngenic rat prostate tumour model. Int J Radiat Biol 2002;78:617–623. 48. Kaminski JM, Hanlon AL, Joon DL, et al. Effect of sequencing of androgen deprivation and radiotherapy on prostate cancer growth. Int J Radiat Oncol Biol Phys 2003;57:24–28. 49. von Neubeck C. Radiobiological experiments for carbon ion prostate cancer therapy: Interplay of normal and tumor cells in co-culture and measurement of the oxygen enhancement ratio [Ph.D. thesis]. Darmstadt: Faculty of Biology at the Technical University Darmstadt; 2009. 50. Tamaki T, Iwakawa M, Ohno T, et al. Application of carbonion beams or gamma-rays on primary tumors does not change the expression profiles of metastatic tumors in an in vivo murine model. Int J Radiat Oncol Biol Phys 2009;74:210–218. 51. Nojiri K, Iwakawa M, Ichikawa Y, et al. The proangiogenic factor ephrin-A1 is up-regulated in radioresistant murine tumor by irradiation. Exp Biol Med (Maywood) 2009;234:112–122. 52. Suit HD, Sedlacek R, Thames HD. Radiation-dose response assays of tumor control. In: Kallmann RF, editor. Rodent tumor models in experimental cancer therapy. New York: Pergamon Press; 1987. p. 138–148. 53. Ando K, Koike S, Ohira C, et al. Accelerated reoxygenation of a murine fibrosarcoma after carbon-ion radiation. Int J Radiat Biol 1999;75:505–512. 54. Fukawa T, Takematsu K, Oka K, et al. Differences in pO2 peaks of a murine fibrosarcoma between carbon-ion and X-ray irradiation. J Radiat Res (Tokyo) 2004;45:303–308. 55. Suzuki M, Kase Y, Yamaguchi H, et al. Relative biological effectiveness for cell-killing effect on various human cell lines irradiated with heavy-ion medical accelerator in Chiba (HIMAC) carbon-ion beams. Int J Radiat Oncol Biol Phys 2000; 48:241–250. 56. Blakely EA, Chang PY. Biology of charged particles. Cancer J 2009;15:271–284.
doi:10.1016/j.ijrobp.2010.07.1976
BIOLOGY CONTRIBUTION
RELATIVE BIOLOGICAL EFFECTIVENESS OF CARBON IONS FOR LOCAL TUMOR CONTROL OF A RADIORESISTANT PROSTATE CARCINOMA IN THE RAT PETER PESCHKE, PH.D.,* CHRISTIAN P. KARGER, PH.D.,y MICHAEL SCHOLZ, PH.D.,z JU¨RGEN DEBUS, M.D., PH.D.,x AND PETER E. HUBER, M.D., PH.D.*x *Clinical Cooperation Unit Radiation Oncology and yDepartment of Medical Physics in Radiation Oncology, German Cancer Research Center, Heidelberg, Germany; zDepartment of Biophysics, Helmholtz Center for Heavy Ion Research, Darmstadt, Germany; and x Department of Clinical Radiology, University of Heidelberg, Heidelberg, Germany Purpose: To study the relative biological effectiveness (RBE) of carbon ion beams relative to X-rays for local tumor control in a syngeneic rat prostate tumor (Dunning subline R3327-AT1). Methods and Materials: A total of 198 animals with tumors in the distal thigh were treated with increasing single and split doses of either 12C ions or photons using a 20-mm spread-out Bragg peak. Endpoints of the study were local control (no tumor recurrence within 300 days) and volumetric changes after irradiation. The resulting values for D50 (dose at 50% tumor control probability) were used to determine RBE values. Results: The D50 values for single doses were 32.9 ± 0.9 Gy for 12C ions and 75.7 ± 1.6 Gy for photons. The respective values for split doses were 38.0 ± 2.3 Gy and 90.6 ± 2.3 Gy. The corresponding RBE values were 2.30 ± 0.08 for single and 2.38 ± 0.16 for split doses. The most prominent side effects were dry and moist desquamation of the skin, which disappeared within weeks. Conclusion: The study confirmed the effectiveness of carbon ion therapy for severely radioresistant tumors. For 1- and 2-fraction photon and 12C ion radiation, we have established individual D50 values for local tumor control as well as related RBE values. Ó 2011 Elsevier Inc. Carbon ion radiotherapy, Dose–response, Dunning prostate tumor R3327, Local tumor control, Relative biological effectiveness.
INTRODUCTION
DNA repair, leading to enhanced cell inactivation, which is assumed to be less dependent on genetic disposition, as well as on oxygen and cell cycle status (17–20). In carbon ion radiotherapy, dose prescriptions refer to biological effective doses, and for their calculation tissuespecific RBEs have to be assigned to each point. To estimate the RBE for normal and neoplastic tissues at each point, a radiobiological model (21) was integrated into the treatment planning system (TPS). Unfortunately, RBE is a complex quantity, depending on physical parameters such as particle type, dose per fraction, and linear energy transfer (LET), as well as on biological factors like cell or tissue type and the selected biological endpoint. Because the RBE for clinical situations involves rather large uncertainties, biological systems were used to evaluate RBE relationships at various depths and their dependence on dose per fraction. Most of
Worldwide, the interest in carbon ion radiotherapy is increasing. After pioneering clinical studies at the Lawrence Berkley Laboratory (University of California) for several ion types (1, 2), a clinical research program was initiated at the National Institute of Radiologic Sciences in Chiba, Japan in 1994 (3–8). In 1997, carbon ion radiotherapy started at the Gesellschaft fu¨r Schwerionenforschung (Darmstadt, Germany) and reported promising results (9–13). Carbon ions exhibit an inverted depth–dose profile, which allows for precise irradiation of deep-seated targets (14, 15). In addition to these physical advantages, carbon ions are characterized by an enhanced relative biological effectiveness (RBE), which is higher in the Bragg peak than in the beam entrance region (16). Mechanistically, carbon ions cause clustered DNA damage, which aggravates
Acknowledgments—The authors thank Alexandra Tietz and Angela Funk (German Cancer Research Center) and Thilo Elsa¨sser (Gesellschaft fu¨r Schwerionenforschung [GSI] Helmholtz Center for Heavy Ion Research) for their excellent technical support; E.W. Hahn for his valuable advice and on-site assistance with study design and setup of the radiation procedure; and GSI for providing excellent experimental conditions with the highly sophisticated beam scanning technique.
Reprint requests to: Christian Karger, Ph.D., Department of Medical Physics in Radiation Oncology (E040), German Cancer Research Center, Im Neuenheimer Feld 280, 69120 Heidelberg, Germany. Tel: (+49) 6221-56-38965; Fax: (+49) 6221-56-4631; E-mail: [email protected] Supported in part by Bundesministerium fu¨r Bildung und Forschung (KVSF, 03NUK004A,C, and 06HD156). P.P. and C.P.K. contributed equally to this work. Conflict of interest: none. 239
I. J. Radiation Oncology d Biology d Physics
240
our present knowledge about tolerance doses and dose dependence of RBEs was derived from experiments with mouse skin (22) and jejunum (23), as well as from rat brain and spinal cord (24–27). Carbon ion effects in experimental solid tumors are addressed by only a few studies using human tumor cell line xenotransplants or syngeneic mouse models with tumor growth inhibition as biological endpoint (28–31). The present study investigates the response of wellcharacterized tumor sublines to carbon ion therapy. We have selected a syngeneic tumor system consisting of several cell lines, which represent the spectrum of androgen responsiveness, tumorigenicity, and metastatic ability seen during the progression of human prostate cancer (32). Growth delay and the probability of local tumor control in relation to locoregional side effects were selected as clinically significant biological endpoints. METHODS AND MATERIALS Tumor model The anaplastic subline AT1 of the Dunning prostate adenocarcinoma R3327 was subcutaneously implanted in the distal thigh of young adult male Copenhagen rats (180 g). This syngenic tumor
Volume 79, Number 1, 2011
is considered an excellent model system for human prostate tumor and has been extensively characterized in terms of ultrastructure, hormones, response to therapy, and immunogenicity (33). Tumor material was kept identical for all experiments by transplanting fresh tissue from tumors grown from a cryopreserved stock, maintained as a first passage of the original tumor tissue kindly supplied to us by J. T. Isaacs (Johns Hopkins University, Baltimore, MD). Ploidy status, histology, and the growth rate of control tumors were determined as measures of quality assurance. All experiments were approved by the governmental review committee on animal care, and animals were kept under standard laboratory conditions at the German Cancer Research Center. During all irradiation procedures, rats were deeply anesthetized by i.p. injection of Ketaminehydrochloride (125 mg/kg) mixed with Xylazinehydrochloride (20 mg/kg).
Experimental setup Average diameter of the tumors before treatment was 9 mm (range, 7.0–10.5 mm). Tumors larger than 10.5 mm in one direction were excluded from the study. Irradiations were performed with a single horizontal beam that hit the tumor from the rear direction. Rats were placed on a specially designed device allowing for rigid and reproducible fixation of the tumor relative to a milled marker on the device (Fig. 1a and b), and the device was aligned on the treatment table with crossed laser beams. In a setup experiment, the
Fig. 1. (a) Setup of the immobilization device for irradiation (here for photons) for the local treatment of s.c. growing tumors (b). Irradation was performed horizontally from the rear direction. (c) The clinical target volume (red) was extended horizontally by a 3-mm and vertically by a 2-mm margin to obtain the planning target volume (PTV) (blue). The resulting PTV extension was 16 14 mm2. (d, e) Sagittal views of the dose distributions in water together with the clinical target volume (PTV not shown) for (d) photons (95%, 90%, 80%, 50%, and 20% isodoses) and (e) carbon ions (95%, 90%, 80%, 70%, 50%, and 20%).
RBE of carbon ions for local tumor control d P. PESCHKE et al.
reproducibility of the tumor position was determined from X-ray images to be 1 mm (1 SD). The planning target volume (PTV) was defined as a cube (clinical target volume, CTV) containing the tumor (<10 mm in diameter) plus a 3-mm margin in the horizontal and a 2-mm margin in the vertical direction accounting for setup variability and variations in tumor size (Fig. 1c). The larger margin in the horizontal direction was selected because of the lack of a clear borderline between tumor and underlying muscle. As a result, the extension of the PTV was 16 14 mm2.
Irradiation technique Photon irradiation was delivered by a linear accelerator (Siemens, Erlangen, Germany) using a single 6-MV horizontal beam shaped by a cylindrical tungsten collimator with 12-mm aperture. Calculated diameter of the 90% and 80% isodoses at isocenter was 15 and 17 mm, respectively (Fig. 1d; TPS: STP [StrykerLeibinger, Freiburg, Germany]). Distance between source and center of the tumor was 99 cm. To achieve a homogeneous depth–dose profile in the tumor, a 10- and 15-mm polymethylmethacrylate (PMMA) bolus was placed directly in front of and behind the tumor, respectively. Dose rate under reference conditions was 2.5–3 Gy/min. The treatment field was calibrated using a pinpoint ion chamber (TM31009; PTW, Freiburg, Germany) placed in the middle of a 10-mm-thick plane-parallel build-up cap fixed at the tumor position within the device. Carbon ion irradiation was performed at the medical beam line at Gesellschaft fu¨r Schwerionenforschung using the intensitycontrolled raster scanning system (34). Irradiations were performed with a homogeneous 20-mm spread-out Bragg peak (SOBP) (scan steps in x, y, z: 1 mm; full width at half maximum of pencil beam: 5 mm) optimized in terms of absorbed dose (TPS: TRiP [35]). The SOBP ranged from 41.5 to 61.5 mm water-equivalent depth (corresponding to initial beam energies between 140 and 171 MeV/u defined by the synchrotron), and the center of the tumor was placed in the middle of the SOBP (Fig. 2). Mean dose average LET in the tumor was 75 keV/mm (range, 64–96 keV/mm). To adjust the range of the ions, a 39.9-mm PMMA bolus was placed directly in front of the tumor. In addition, a 15-mm PMMA bolus was placed directly behind the tumor. Diameter of the calculated 90% (80%) isodose was 16.5 mm (18.5 mm) in the lateral and 14.6 mm (16.8 mm) in the vertical direction (Fig. 1e). The calculated dose in the middle
of the SOBP was verified in a water phantom using a pinpoint ion chamber (TM31009; PTW). A deviation of 0.5% was found.
Dose levels, fractionation scheme, and follow-up A total of 198 rats were irradiated with single and split doses of either photons or carbon ions using a series of increasing dose levels (Table 1). Irradiations with 2 fractions were performed on consecutive days. Sham-treated animals served as controls. During the first cohort of the split-dose experiment for carbon ions, a technical problem with the beam delivery software led to a break of 48 rather than 24 h. Because this unintentionally prolonged break might lead to a modified radiation response, the experiment was repeated with a regular 24-h break. The experiment with the accidental 48-h break was nevertheless evaluated to study the impact of a prolonged break related to treatment outcome. The primary endpoint of this study was local tumor control, defined as no indication of tumor recurrence within an observation time of 300 days after irradiation. In addition, the volume of the tumors was measured routinely three times weekly (Monday, Wednesday, Friday), and growth curves were established. The tumor volume was calculated by the formula 4/3 p r3, with r being the mean of two orthogonal tumor radii. Because the growth patterns of locally controlled and locally noncontrolled tumors are quite different, averaged growth curves were established separately for locally controlled and locally noncontrolled tumors for the mixed dose groups.
Statistics For each dose level, the incident rates x/n (x out of n tumors are controlled) were determined, and the logistic dose response model PðDÞ ¼
eb0 þb1 D 1 þ eb0 þb1 D
(1a)
Table 1. Dose levels and number of animals evaluated for each experiment Experiment 1 fraction Photons Carbon ions 2 fractions Photons Carbon ions Carbon ions (48-h break)z Controls Total
Fig. 2. Depth–dose and dose average linear energy transfer (LET) profile used for the carbon ion irradiations. The position of the tumor within the spread-out Bragg peak is indicated.
241
Dose levels (Gy)
Animals evaluated
30, 40, 50, 60, 70*, 75, 77.5*, 80y, 85, 90 9, 12.5, 16, 19.5, 23, 26.5, 30*, 33.5, 37*, 40.5
37
35y, 40, 50, 60, 70, 80, 85, 90, 95, 100* 8, 14, 20, 26, 32*, 38, 44, 50, 56 26, 32, 38, 44, 50
39
5 sham-treated tumors per experimentx
20
42
37 15
190
In total, the study contained 198 animals, from which 190 could be evaluated. * One animal was excluded from analysis because of metastatic activity. y One animal died after therapy because of unknown reasons. z An additional cohort was treated with a 48 h break between doses caused by technical reasons and was evaluated separately (see text). x The 5 controls of the 2-fraction experiment with carbon ions also served as controls for the experiment with the 48-h break.
242
I. J. Radiation Oncology d Biology d Physics
was adjusted separately for each radiation modality and fractionation schedule using the maximum likelihood fitting procedure of the statistic package Statistica (36). D is the total dose, and b0 and b1 are the adjusted fit parameters from which the tolerance dose D50 (dose at 50% tumor control probability) can be calculated: b0 D50 ¼ : b1
(1b)
The RBE was then calculated as the ratio of the D50 values for photons and carbon ions, respectively. In addition, 90% confidence limits were calculated for the RBE values using Fieller’s theorem (37).
RESULTS Spontaneous death occurred in 2 of 198 animals. Cause of death remained unclear even after pathologic inspection. Additional 6 animals (3%) were removed from the study because of metastatic activity in lymph nodes adjacent to the tumor. Lymph node migration was found in all irradiated groups, with a tendency to higher radiation doses. The observed low spontaneous metastatic ability is well known for the Dunning R3327-AT sublines (38), and thus no bias
Volume 79, Number 1, 2011
of observed incidences is to be expected. Finally, a total of 190 animals could be evaluated.
Growth delay Tumor growth delay was found to be strongly dependent on dose and on whether tumors were locally controlled (Fig. 3a–d). Independent of local control status, all growth curves revealed a continuous increase in tumor volume within the first 3 weeks after irradiation, reaching three to four times the initial tumor volume. Dependent on dose, this was followed either by further growth or by shrinkage of the tumor. The decrease in volume was followed by dose-dependent regrowth or local control of tumors. For both irradiation modalities, the maximum time until tumor regrowth occurred was 95 days after irradiation. Further dose escalation led to treatment groups, with increasing fractions of local tumor control for both radiation qualities. In the order of increasing radiation doses, the first single-dose group in which all tumors were controlled was found at 37 Gy for carbon ions and 85 Gy for photons. For 2 fractions,
Fig. 3. Growth response of the tumor to either (a, b) single (a, b) or (c, d) split (c, d) doses of photons (a, c) and carbon ions (b, d) (mean 1 SD). Mean values are calculated separately for tumors without (nLC, closed symbols) and with local control (LC, open symbols). For better visibility, only curves for selected dose levels are shown. Because no further growth occurred, the x-axis was truncated at 180 days.
RBE of carbon ions for local tumor control d P. PESCHKE et al.
243
Side effects Local hair loss within the field of irradiation was observed in all animals. Dry and moist desquamation of the skin (#1.5+ according to the normal damage scale previously published [39]) occurred as the most prominent side effect, which disappeared within 4–6 weeks after therapy. Radiation-induced side effects in the normal skin around the tumor were worse in animals in which proliferative activity was not controlled, presumably because skin recovery was constraint during increased volumetric tumor expansion. At the end of the observation period of 300 days, animals with locally controlled tumors exhibited slight skin erythema and mild dystrophy of the thigh muscle. Interestingly, a fibrotic nodule remained perceptible in most animals with local tumor control at the end of the observation period, which is known to be an ordinary process after curative radiation doses. DISCUSSION
Fig. 4. Dose–response curves for (a) 1 and (b) 2 fractions of photons and carbon ions. The uncertainty (1 SD) of D50 is indicated.
the respective doses were 50 Gy for carbon ions and 100 Gy for photons. ED50 values, fractionation effects, and RBEs Figure 4a and b display the adjusted dose–response curves for single and split doses of carbon ions and photons, respectively. The respective values for D50 and RBE are given in Table 2. The ratio of the doses for 1 vs. 2 fractions to achieve a D50 isoeffect for carbon ions and photons was 1.16 and 1.20, respectively. When the break between two fractions of carbon ions was extended from 24 to 48 h, D50 decreased from 38.0 2.3 Gy to 32.9 2.6 Gy (Fig. 5).
Although heavy ions exhibit favorable physical and biological properties, the role of carbon ion radiotherapy has still to be determined. For that purpose, two therapy centers in Japan (Heavy Ion Medical Accelerator, [HIMAC] in Chiba [8] and Hyogo Ion Beam Medical Center [40]) and one in Germany (Heidelberger Ionenstrahl-Therapiezentrum [HIT]) (41, 42) have been established to date, and several tumor sites, including prostate carcinoma (3–5), uveal melanoma, and hepatocellular carcinoma, as well as skull base, lung, and prostate tumors, are under evalua (43). Beside these clinical studies, systematic biological investigations are necessary to characterize the effectiveness of this new beam quality and to help integrating carbon ions into the arsenal of radiation treatment strategies. Whereas our previous studies investigated mainly carbon ion–related damage to normal central nervous system tissues (26, 27), the present experiments focused on the response of an experimental tumor system. The selected prostate tumor subline R3327-AT1 represents a hormone-independent anaplastic carcinoma with a labeling index of 7.0% 0.5% and 8.8% 3.7% measured with flow cytometry and histology, respectively, a S-phase duration of 8 h, a potential doubling time of 4.7 days, and a cell loss factor of 15% (44). Tumor oxygen tension as examined by 19F nuclear magnetic resonance spin-lattice relaxation rate of perfluorocarbon emulsion revealed a size-dependent development of generalized
Table 2. D50 and RBE values measured in this study D50 SE (90% CL) (Gy) Study
Photons
Carbon ions
RBE SE (90% CL)
1 fraction 2 fractions
75.7 1.6 (69.9–78.6) 90.6 2.3 (85.6–95.4)
32.9 0.9 (30.8–34.9) 38.0 2.3 (33.7–42.6)
2.30 0.08 (2.17–2.44) 2.38 0.16 (2.15–2.67)
Abbreviations: D50 = dose at 50% tumor control probability; RBE = relative biological effectiveness; CL = confidence limit; LET = linear energy transfer, SE = standard error. Endpoint: tumor control at 300 days, LET = 75 (64–96) keV/mm.
244
I. J. Radiation Oncology d Biology d Physics
Volume 79, Number 1, 2011
Acceptable side effects to normal tissue within the treatment field enabled the extension of the study to the preferred biological endpoint of local tumor control by dose escalation. This endpoint depends only on the inactivation of clonogenic tumor cells and is regarded as the most important assay for curative effects in radiotherapy (52). In addition, D50 for local tumor control is a robust endpoint parameter, especially for measuring RBE values.
Fig. 5. Dose–response curves for 2 fractions of carbon ions applied either within 24 h or 48 h. The uncertainty (1 SD) of D50 is indicated. Note that the 48-h experiment was not intended and resulted from problems with the beam delivery software (see text). The photon curve for the regular split-dose regimen with a 24-h break is shown in Fig. 4b.
central hypoxia (45). Nevertheless, the degree of hypoxia is not sufficient to induce central necrosis, as is often seen in the more rapidly proliferating rodent tumors. Chronic hypoxia is one of the reasons why this syngeneic tumor subline exhibited a high radiation resistance, as documented in previous animal studies with various radiation modalities (46–48). This resistance is also known from cell experiments with X-rays, in which a values of 0.144 and 0.068 Gy1 were measured for oxic and hypoxic conditions, respectively, leading to an oxygen enhancement ratio of 2.35 (49). The selected tumor subline has a low metastatic potential. In our experiment we had 6 cases of lymph node metastasis. Because of this very small number, the dependence of these events on dose (and potentially LET) could not be reliably analyzed. For this, a dedicated experimental design is required using a tumor model with a higher metastatic potential. Such a study was carried out by Tamaki et al. (50) for the murine squamous cell carcinoma NR-S1. This study showed that small doses of photons or carbon ions inhibit the incidence of lung metastasis similarly. In addition, the transcriptional profile of the metastasis was independent of the application of radiotherapy per se, as well as the radiotherapy type. Biological endpoints In accordance with previous studies, growth delay was found to increase with increasing dose (50, 51), reflecting an increasing number of inactivated clonogenic cells. Extreme differences in growth delay within the three noncontrolled tumors were the reason for the large heterogeneity in the group treated with a single dose of 33.5 Gy carbon ions (Fig. 3b). A similar heterogeneity was also seen for a few other dose/modality groups near the respective D50 group (not displayed for visibility reasons). A definitive explanation for this behavior, however, could not be identified.
Split doses for carbon ions For the split-dose experiment with 12C ions, the dose– response curve was shifted to slightly lower doses, when the time interval between fractions was increased from 24 h to 48 h. Because the respective D50 values differ by less than 2 combined SDs, this shift is considered not statistically significant. It is therefore concluded that the prolonged time between fractions does not lead to significantly increased repair and proliferation. Rather, the shift may have occurred by chance. An alternative explanation might be improved reoxygenation for the 48-h experiment, given that a study by Ando et al. (53) showed accelerated reoxygenation of a murine fibrosarcoma after carbon ion radiation. While NFSa tumors reoxygenated 4 days after 30 Gy of X-ray irradiation, reoxygenation occurred as early as 1 day after 16 Gy of carbon ions measured by paired growth delay assay. These results have been confirmed in a second study, in which oxygen partial pressure profiles were measured in the same tumor system using coaxial high spatial resolution oxygen microelectrodes (54). RBE and fractionation ratio Our results clearly show that physical doses necessary to locally control this anaplastic prostate carcinoma are considerably lower for carbon ions than for photons. Additionally, only a very small fractionation effect for both beam qualities was found. The D50 isoeffect ratio of 1.20 for 2 vs. 1 fractions of photons is not markedly different from that of 1.16 for carbon ions. Our RBE values of 2.30 0.08 (1 fraction) and 2.38 0.16 (2 fractions) are in line with the results published by others. From in vitro data on the same cell line irradiated with carbon ions at different LETs (49), the RBE for 10% clonogenic cell survival can be linearly interpolated to be 2.2 at 75 keV/mm, which is close to our in vivo RBE value for the single-fraction experiment. A large series of studies with tumor cell lines has shown that high LET radiations are more effective than low LET radiations, such as X-rays or g rays, for the endpoint cell kill. The most comprehensive collection was published by Suzuki et al. (55), summarizing RBE values for 10% clonogenic cell survival in 16 different tumor cell lines. The RBEs obtained for a dose-averaged LET of approximately 77 keV/mm ranged between 2.00 and 3.01. In an in vivo study, an RBE of 2.02 was determined for 290 MeV/u 12C beam with an LET of 70 keV/mm in a human esophageal cancer (ESO-2) xenotransplant. The RBE was calculated from the slope of the dose–response curve for
RBE of carbon ions for local tumor control d P. PESCHKE et al.
tumor growth suppression at 4 weeks after irradiation with X-rays or carbon ion beams (28). The only study determining RBEs for animal tumors treated with fractionated doses of 290 MeV/u carbon ions was performed with the NFSa fibrosarcoma model in mice. In this study, a clear relationship between RBE and LET was documented. For 14 and 20 keV/mm, RBEs were 1.4 and independent of the number of fractions, whereas those for 44 and 74 keV/mm increased from 1.8 to 2.3 and from 2.4 to 3.0, respectively, when the number of fractions changed from 1 to 4. Raising the number of fractions from 4 to 6 was not associated with a further increase in RBE (31). The single-dose RBE of 2.4 at 74 keV/mm is well comparable to the value obtained in our study; however, our study revealed a much smaller increase of RBE when going from singe to split doses. This might be explained by the different applied endpoints. Whereas Koike et al. (31) define isoeffectivity (and hence RBE) by equal tumor growth delay for subtherapeutic doses, our RBEs are based on local control at therapeutic doses (i.e., at the level of D50). The RBEs determined from equal tumor growth delay at 1 and 2 fractions might therefore not correspond to the same tumor control probability at much higher doses, which is the most important endpoint in clinical application. A direct comparison of the RBEs between these two studies is therefore difficult. The reason for the differential response of the Dunning tumor subline R3327-AT1 with respect to the two radiation qualities is presently not clearly understood. The underlying
245
molecular mechanisms are multifactorial, and studies are in progress by many investigators (for review see Blakely and Chang [56]). Microarray analysis has been applied to identify genes involved in regulatory pathways that result in the development of resistance to therapy. Nojiri et al. (51), for example, investigated radiation-induced changes in gene expression related to angiogenesis, which influence responsiveness of tumors by the modulation of microenvironmental conditions. In some cases, these radiation-induced changes even seem to depend on the radiation type. For future preclinical assessment of carbon ion efficacy, such information is not only the prerequisite to define tumor models with respect to well-characterized features of radiation resistance, such as dysregulated cell death pathways and upregulated survival factors, but this knowledge may also have a predictive value for both patient selection and therapy outcome.
CONCLUSION The present study confirmed the effectiveness of carbon ion radiotherapy for severely radioresistant tumors and investigated an in vivo tumor model using local control as a clinically most relevant endpoint. Additional experiments with different fraction numbers and tumor sublines as well as comparison with normal tissue response are necessary to draw clinically relevant conclusions.
REFERENCES 1. Castro JR. Results of heavy ion radiotherapy. Radiat Environ Biophys 1995;34:45–48. 2. Castro JR, Linstadt DE, Bahary JP, et al. Experience in charged particle irradia of tumors of the skull base: 1977-1992. Int J Radiat Oncol Biol Phys 1994;29:647–655. 3. Akakura K, Tsujii H, Morita S, et al. Phase I/II clinical trials of carbon ion therapy for prostate cancer. Prostate 2004;58:252–258. 4. Ishikawa H, Tsuji H, Kamada T, et al. Carbon ion radiation therapy for prostate cancer: Results of a prospective phase II study. Radiother Oncol 2006;81:57–64. 5. Tsuji H, Yanagi T, Ishikawa H, et al. Hypofractionated radiotherapy with carbon ion beams for prostate cancer. Int J Radiat Oncol Biol Phys 2005;63:1153–1160. 6. Kanai T, Endo M, Minohara S, et al. Biophysical characteristics of HIMAC clinical irradiation system for heavy-ion radiation therapy. Int J Radiat Oncol Biol Phys 1999;44:201–210. 7. Kato H, Tsujii H, Miyamoto T, et al. Results of the first prospective study of carbon ion radiotherapy for hepatocellular carcinoma with liver cirrhosis. Int J Radiat Oncol Biol Phys 2004;59:1468–1476. 8. Tsujii H, Mizoe J, Kamada T, et al. Clinical results of carbon ion radiotherapy at NIRS. J Radiat Res (Tokyo) 2007;48(Suppl. A):A1–A13. 9. Schulz-Ertner D, Nikoghosyan A, Didinger B, et al. Therapy strategies for locally advanced adenoid cystic carcinomas using modern radiation therapy techniques. Cancer 2005;104: 338–344. 10. Schulz-Ertner D, Nikoghosyan A, Hof H, et al. Carbon ion radiotherapy of skull base chondrosarcomas. Int J Radiat Oncol Biol Phys 2007;67:171–177.
11. Schulz-Ertner D, Karger CP, Feuerhake A, et al. Effectiveness of carbon ion radiotherapy in the treatment of skull-base chordomas. Int J Radiat Oncol Biol Phys 2007;68:449–457. 12. Schulz-Ertner D. The clinical experience with particle therapy in adults. Cancer J 2009;15:306–311. 13. Combs SE, Nikoghosyan A, Jaekel O, et al. Carbon ion radiotherapy for pediatric patients and young adults treated for tumors of the skull base. Cancer 2009;115:1348–1355. 14. Kraft G. Radiotherapy with heavy ions: Radiobiology, clinical indications and experience at GSI Darmstadt. Tumori 1998;84: 200–204. 15. Karger CP, Ja¨kel O. Current status and new developments in ion therapy. Strahlenther Onkol 2007;183:295–300. 16. Blakely EA, Kronenberg A. Heavy-ion radiobiology: New approaches to delineate mechanisms underlying enhanced biological effectiveness. Radiat Res 1998;150:S126–S145. 17. Tsuboi K, Tsuchida Y, Nose T, et al. Cytotoxic effect of accelerated carbon beams on glioblastoma cell lines with p53 mutation: Clonogenic survival and cell-cycle analysis. Int J Radiat Biol 1998;74:71–79. 18. Weyrather WK, Ritter S, Scholz M, et al. RBE for carbon tracksegment irradiation in cell lines of differing repair capacity. Int J Radiat Biol 1999;75:1357–1364. 19. Staab A, Zukowski D, Walenta S, et al. Response of Chinese hamster v79 multicellular spheroids exposed to high-energy carbon ions. Radiat Res 2004;161:219–227. 20. Fournier C, Wiese C, Taucher-Scholz G. Accumulation of the cell cycle regulators TP53 and CDKN1A (p21) in human fibroblasts after exposure to low- and high-LET radiation. Radiat Res 2004;161:675–684.
246
I. J. Radiation Oncology d Biology d Physics
21. Scholz M, Kellerer AM, Kraft-Weyrather W, et al. Computation of cell survival in heavy ion beams for therapy. The model and its approximation. Radiat Environ Biophys 1997;36:59–66. 22. Ando K, Koike S, Uzawa A, et al. Repair of skin damage during fractionated irradiation with gamma rays and low-LET carbon ions. J Radiat Res (Tokyo) 2006;47:167–174. 23. Fukutsu K, Kanai T, Furusawa Y, et al. Response of mouse intestine after single and fractionated irradiation with accelerated carbon ions with a spread-out Bragg peak. Radiat Res 1997; 148:168–174. 24. Leith JT, McDonald M, Powers-Risius P, et al. Response of rat spinal cord to single and fractionated doses of accelerated heavy ions. Radiat Res 1982;89:176–193. 25. Karger CP, Debus J, Peschke P, et al. Dose-response curves for late functional changes in the normal rat brain after single carbon-on doses evaluated by magnetic resonance imaging: Influence of follow-up time and calculation of relative biological effectiveness. Radiat Res 2002;158:545–555. 26. Karger CP, Peschke P, Sanchez-Brandelik R, et al. Radiation tolerance of the rat spinal cord after 6 and 18 fractions of photons and carbon ions: Experimental results and clinical implications. Int J Radiat Oncol Biol Phys 2006;66:1488–1497. 27. Debus J, Scholz M, Haberer T, et al. Radiation tolerance of the rat spinal cord after single and split doses of photons and carbon ions. Radiat Res 2003;160:536–542. 28. Takahashi A, Yano T, Matsumoto H, et al. Effects of accelerated carbon-ions on growth inhibition of transplantable human esophageal cancer in nude mice. Cancer Lett 1998;122:181–186. 29. Kitabayashi H, Shimada H, Yamada S, et al. Synergistic growth suppression induced in esophageal squamous cell carcinoma cells by combined treatment with docetaxel and heavy carbon-ion beam irradiation. Oncol Rep 2006;15:913–918. 30. Asakawa I, Yoshimura H, Takahashi A, et al. Radiationinduced growth inhibition in transplanted human tongue carcinomas with different p53 gene status. Anticancer Res 2002;22: 2037–2043. 31. Koike S, Ando K, Oohira C, et al. Relative biological effectiveness of 290 MeV/u carbon ions for the growth delay of a radioresistant murine fibrosarcoma. J Radiat Res (Tokyo) 2002;43: 247–255. 32. Tennant TR, Kim H, Sokoloff M, et al. The Dunning model. Prostate 2000;43:295–302. 33. Isaacs JT, Heston WD, Weissman RM, et al. Animal models of the hormone-sensitive and -insensitive prostatic adenocarcinomas, Dunning R-3327-H, R-3327-HI, and R-3327-AT. Cancer Res 1978;38:4353–4359. 34. Haberer T, Becher W, Schardt D, et al. Magnetic scanning system for heavy ion therapy. Nucl Instrum Meth A 1993;A330: 296–305. 35. Kra¨mer M, Ja¨kel O, Haberer T, et al. Treatment planning for heavy-ion radiotherapy: Physical beam model and dose optimization. Phys Med Biol 2000;45:3299–3317. 36. StatSoft Inc. Statistica for Windows 7.1. Tulsa, OK: StatSoft; 2005. 37. Finney DJ. Statistical method in biological assay. 3rd ed. London: Charles Griffin; 1978. 38. Isaacs JT, Isaacs WB, Feitz WF, et al. Establishment and characterization of seven Dunning rat prostatic cancer cell lines and their use in developing methods for predicting metastatic abilities of prostatic cancers. Prostate 1986;9:261–281.
Volume 79, Number 1, 2011
39. Ohta T, Iwakawa M, Oohira C, et al. Fractionated irradiation augments inter-strain variation of skin reactions among three strains of mice. J Radiat Res (Tokyo) 2004;45:515–519. 40. Hishikawa Y, Oda Y, Mayahara H, et al. Status of the clinical work at Hyogo. Radiother Oncol 2004;73(Suppl. 2):S38–S40. 41. Haberer T, Debus J, Eickhoff H, et al. The Heidelberg Ion Therapy Center. Radiother Oncol 2004;73(Suppl. 2):S186– S190. 42. Combs SE, Jakel O, Haberer T, et al. Particle therapy at the Heidelberg Ion Therapy Center (HIT)—integrated researchdriven university-hospital-based radiation oncology service in Heidelberg, Germany. Radiother Oncol 2010;95:41–44. 43. Schulz-Ertner D, Tsujii H. Particle radiation therapy using proton and heavier ion beams. J Clin Oncol 2007;25:953–964. 44. Lohr F, Wenz F, Flentje M, et al. Measurement of the proliferative activity of three different sublines of the Dunning rat prostate tumor R3327. Strahlenther Onkol 1993;169:438–445. 45. Mason RP, Antich PP, Babcock EE, et al. Non-invasive determination of tumor oxygen tension and local variation with growth. Int J Radiat Oncol Biol Phys 1994;29:95–103. 46. Peschke P, Hahn EW, Wenz F, et al. Differential sensitivity of three sublines of the rat Dunning prostate tumor system R3327 to radiation and/or local tumor hyperthermia. Radiat Res 1998; 150:423–430. 47. Harms W, Peschke P, Weber KJ, et al. Dose-dependent differential effects of low and pulsed dose-rate brachytherapy in a radioresistant syngenic rat prostate tumour model. Int J Radiat Biol 2002;78:617–623. 48. Kaminski JM, Hanlon AL, Joon DL, et al. Effect of sequencing of androgen deprivation and radiotherapy on prostate cancer growth. Int J Radiat Oncol Biol Phys 2003;57:24–28. 49. von Neubeck C. Radiobiological experiments for carbon ion prostate cancer therapy: Interplay of normal and tumor cells in co-culture and measurement of the oxygen enhancement ratio [Ph.D. thesis]. Darmstadt: Faculty of Biology at the Technical University Darmstadt; 2009. 50. Tamaki T, Iwakawa M, Ohno T, et al. Application of carbonion beams or gamma-rays on primary tumors does not change the expression profiles of metastatic tumors in an in vivo murine model. Int J Radiat Oncol Biol Phys 2009;74:210–218. 51. Nojiri K, Iwakawa M, Ichikawa Y, et al. The proangiogenic factor ephrin-A1 is up-regulated in radioresistant murine tumor by irradiation. Exp Biol Med (Maywood) 2009;234:112–122. 52. Suit HD, Sedlacek R, Thames HD. Radiation-dose response assays of tumor control. In: Kallmann RF, editor. Rodent tumor models in experimental cancer therapy. New York: Pergamon Press; 1987. p. 138–148. 53. Ando K, Koike S, Ohira C, et al. Accelerated reoxygenation of a murine fibrosarcoma after carbon-ion radiation. Int J Radiat Biol 1999;75:505–512. 54. Fukawa T, Takematsu K, Oka K, et al. Differences in pO2 peaks of a murine fibrosarcoma between carbon-ion and X-ray irradiation. J Radiat Res (Tokyo) 2004;45:303–308. 55. Suzuki M, Kase Y, Yamaguchi H, et al. Relative biological effectiveness for cell-killing effect on various human cell lines irradiated with heavy-ion medical accelerator in Chiba (HIMAC) carbon-ion beams. Int J Radiat Oncol Biol Phys 2000; 48:241–250. 56. Blakely EA, Chang PY. Biology of charged particles. Cancer J 2009;15:271–284.