Proton relative biological effectiveness (RBE) for survival in mice after thoracic irradiation with fractionated doses

Int. J. Radiation Oncology Biol. Phys., Vol. 47, No. 4, pp. 1051–1058, 2000 Copyright © 2000 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/00/$–see front matter

PII S0360-3016(00)00535-6

BIOLOGY CONTRIBUTION

PROTON RELATIVE BIOLOGICAL EFFECTIVENESS (RBE) FOR SURVIVAL IN MICE AFTER THORACIC IRRADIATION WITH FRACTIONATED DOSES JOHN GUEULETTE, PH.D.,* LOTHAR BO¨ HM, PH.D.,‡ JACOBUS P. SLABBERT, PH.D.,† BLANCHE M. DE COSTER, B.SC.,* GERALD S. RUTHERFOORD, M.D.,‡ ARNOUT RUIFROK, PH.D.,§ MICHELLE OCTAVE-PRIGNOT, M.SC.,* PETER J. BINNS, PH.D.,† A. NICHOLAAS SCHREUDER, M.SC.,† JULIAN E. SYMONS, B.SC.,† PIERRE SCALLIET, M.D., PH.D.,* AND DAN T. L. JONES, PH.D.† *Universite´ Catholique de Louvain, Brussels, Belgium; †National Accelerator Centre, Faure, Republic of South Africa; ‡University of Stellenbosch, Tygerberg Hospital, Tygerberg, Republic of South Africa; and §University of Texas M.D. Anderson Cancer Center, Houston, Texas Purpose: This study aims at providing relative biological effectiveness (RBE) data under reference conditions accounting for the determination of the “clinical RBE” of protons. Methods and Materials: RBE (ref. 60Co ␥-rays) of the 200 MeV clinical proton beam produced at the National Accelerator Centre (South Africa) was determined for lung tolerance assessed by survival after selective irradiation of the thorax in mice. Irradiations were performed in 1, 3, or 10 fractions separated by 12 h. Proton irradiations were performed at the middle of a 7-cm spread out Bragg peak (SOBP). Control ␥ irradiations were randomized with proton irradiations and performed simultaneously. A total of 1008 mice was used, of which 96 were assessed for histopathology. Results: RBEs derived from LD50 ratios were found not to vary significantly with fractionation (corresponding dose range, ⬃2–20 Gy). They, however, tend to increase with time and reach (mean of the RBEs for 1, 3 and 10 fractions) 1.00, 1.08, 1.14, and 1.25 for LD50 at 180, 210, 240, and 270 days, respectively (confidence interval approximately 20%). ␣/␤ ratios for protons and ␥ are very similar and average 2.3 (0.6 – 4.8) for the different endpoints. Additional irradiations in 10 fractions at the end of the SOBP were found slightly more effective (⬃6%) than at the middle of the SOBP. A control experiment for intestinal crypt regeneration in mice was randomized with the lung experiment and yielded an RBE of 1.14 ⴞ 0.03, i.e., the same value as obtained previously, which vouches for the reliability of the experimental procedure. Conclusion: There is no need to raise the clinical RBE of protons in consideration of the late tolerance of healthy tissues in the extent that RBE for lung tolerance was found not to vary with fractionation nor to differ significantly from those of the majority of early- and late-responding tissues. © 2000 Elsevier Science Inc. Radiotherapy, Proton beams, Relative biological effectiveness (RBE), Late responding tissues.

INTRODUCTION There are two reasons for determining the relative biological effectiveness (RBE) of clinical hadron beams. The first is that the clinical results of hadrontherapy have to be compared with those of alternative techniques using conventional radiations. This requires the knowledge of the “gamma equivalent doses” that are equal to the physical doses (in Gy) multiplied by the clinical RBE (see below) of the concerned radiation. The second reason is characteristic of heavy ions and (to a lesser extent) of protons and lies in the fact that the biological effectiveness of these particles varies with depth. This, which makes that an isodose in a given tissue does not necessarily correspond to an isoeffect, obliges the radiation oncologist to give new attention to the

treatment plans and to evaluate the ballistic of an irradiation in connection with the possible RBE variations. The goal of the present experiment is to provide radiobiological information accounting for the determination of the “clinical” RBE of protons. Clinical RBE is an operational concept that has been initially used in the United States for neutrontherapy. It is understood as the ratio of the absorbed dose that would have been given in a photon treatment and the neutron dose that is actually prescribed at a given neutrontherapy facility (and for a given tumor localization) (1). The same concept applies to protontherapy, even if proton RBEs are much lower than neutron RBEs, but too large to be neglected taking into account that the dose accuracy needed in radiotherapy is 3.5% (2). Clinical RBE is not in the strict sense an RBE. Due to the

Reprint requests to: Dr. J. Gueulette, Universite´ Catholique de Louvain, Cliniques Universitaires St-Luc, 54, Avenue Hippocrate, 1200 Bruxelles, Belgium. E-mail: [email protected] Acknowledgments—This work was supported by the “Fonds de la

Recherche Scientifique Me´dicale,” convention FRSM 9.4602.94 and by a grant of the National Accelerator Centre of Faure (South Africa). Accepted for publication 25 February 2000. 1051

1052

I. J. Radiation Oncology

●

Biology

●

Physics

wide variation of RBE with the biological system and criterion, dose, and experimental conditions, it is necessary to select reference conditions for RBE specification when transferring or exchanging clinical information. The following conditions appear to be relevant: (1) a dose level of 2 Gy (␥ equivalent) per fraction, and (2) a biological system and endpoint representative of the RBE for the overall late tolerance of normal tissues (3). The RBE obtained in such reference conditions is not yet the clinical RBE. It is called “reference” RBE, which is a radiobiological approach implying that a single RBE value can be defined for the overall late tolerance of normal tissues in patients (which assumption is supported by the fact that ␣/␤ ratios for late tolerance of different normal tissues are similar). As for the “clinical” RBE, it is the reference RBE weighted by empirical factors derived by clinical experience (4). However, these factors, which mainly allowed for the difference in beam penetration between neutrons and photons, are practically useless for protons whose ballistic performance are per se better than the one of photons. Consequently, the clinical RBE of protons tends to become close to the reference RBE. The present RBE study, which involves a late responding tissue and fractionated irradiations (low dose per fraction), meets the requirements for determining a reference RBE. Therefore, it takes on a particular interest, the more so as in vivo proton RBE data for such conditions are up to now not available (5).

METHODS AND MATERIALS Radiation beams The proton irradiations were performed in the clinical 200-MeV proton beam produced at the National Accelerator Centre (NAC) in Faure (South Africa). The protons are accelerated in a separated-sector cyclotron combined with a solid pole injector cyclotron. The beam has a fixed horizontal direction and is laterally spread and flattened using double-scatter/occluding-ring system. The maximum field diameter is 10 cm. The distal 90 –10% fall-off occurs in 5.5 mm; the 90 –10% penumbra of the central axis profiles is 2.8 mm. The Bragg peak region of the depth dose distribution is extended (modulated beam) by means of a rotating stepper-absorber (modulating wheel). Reference ␥ irradiations were performed with a clinical telecobaltherapy machine located close by the proton irradiation room.

Biological system Survival at 180 –270 days after selective irradiation of the thorax in mice has been chosen as the biological criterion. It is well known that an irradiation of the lung results in pneumonitis 3– 6 months after irradiation, which leads the animal to die when the irradiation dose is adequate (6 –7). This in vivo system has been used by different authors as a model for evaluating the late tolerance of healthy tissues (8 –10).

Volume 47, Number 4, 2000

Animals A total of 1008 13-week-old male and female Balb/c mice were used. They were produced in the laboratory of the Medical Research Council (MRC) in Tygerberg (South Africa). After irradiation, the animals were returned to the same laboratory, which took care of their maintenance and their follow-up for a 12-month period. Death rates were scored every day, and the dead bodies were fixed in formalin for ulterior postmortem analyses. Experimental procedure One week before irradiation, the mice were transported to NAC and accommodated in batches of 6 animals (84 cages for males and 84 for females). Due to the fact that the animals would stay alive for a long period, special attention has been paid to minimize the “cage effect.” Therefore, the mice were individually assigned to a given treatment modality through a computerized randomization procedure. Each cage contained mice devoted to different radiation qualities and/or different fractionation schemes and/or different dose levels (see below). Three of the six mice were colored in yellow and each pair of white/yellow mice had either left, right, or no ear cut, which made it possible to distinguish them from each other. The whole of the irradiations was spread over a 4-week period, the 1st and 2nd week being successively devoted to photon and proton irradiation of males (first subexperiment) and the 3rd and 4th week to irradiation of females (second subexperiment). Irradiation condition For both 200-MeV protons and reference 60Co ␥-rays, the radiation doses were delivered in a single, 3 or 10 equal fractions separated by time intervals of 12 h. A total of 6 dose– effect relationships were thus established, each of them including 9 groups of 16 animals (8 males and 8 females) exposed to increasing doses in the range expected to yield a death rate between 10% and 90% at 180 days (total of 144 mice per dose– effect relationship). The animals were selectively irradiated on the thorax by placing them transversally in a 3-cm high oblong field (Fig. 1). They were not anesthetized and kept in elongated position by means of special wooden clothes pegs, in each jaw of which a groove was made according to the diameter of the legs of the animals and covered with adhesive plaster to give some elasticity (11). The animals were centered by means of a laser system in reference to the caudal part of the ears which was chosen as delimiting the upper part of the thoracic field. They were positioned at the depth of the peak dose for photons and at the middle of a 7-cm SOBP for protons. For the latter beam, additional 10-fraction irradiations were performed at the end of the 7-cm SOBP (2.5-cm displacement of the mid-abdomen down stream the beam, in comparison with the middle of the SOBP). The dose homogeneity was ⫾ 5% and the dose rate was ⬃0.7 Gy/min for both protons and ␥-rays. After irradiation, the mice were returned to their respective cages and were provided with food and water ad libitum. Absorbed doses were specified in water

Proton RBE for survival in mice after thoracic irradiation

●

J. GUEULETTE et al.

1053

Fig. 1. Irradiation setup (the beam is coming from the rear of the picture). The mice are squeezed between 2 Lucite plates 1.5-cm apart (the fore Lucite plate has been removed for the clarity of the picture). The dotted lines show the limits of the thoracic field (3-cm high), the animals being centred in the field in reference to their ears. Grooves are made in the jaws of the clothes pegs according to the diameter of the legs of the animals and are covered with adhesive plaster. Positioning of the mice at the specified depth (middle or end of a 7-cm SOBP for protons and depth of the peak dose for 60Co ␥-rays) is made by adding appropriate layers of Lucite between the collimator and the jig. For 60Co ␥-rays, a 10-cm thick Lucite plate is placed behind the jig for back scattering.

and measured according to the ICRU protocol for ␥ irradiations (12) and according to the Code of Practice for Clinical Proton Dosimetry for protons (13). Histopathology To check the cause of death, 96 mice out of the 1008 were assessed for histopathology. These animals had died in the time window between Day 90 and Day 270, while their majority had died between Day 120 and Day 240 (see Fig. 2). Approximately, one-half of the mice had received 60Co ␥-ray irradiations and the other one-half 200-MeV proton irradiations. The results were pooled, and no distinction was made between radiation quality, nor between fractionation schedule or dose level. Control experiment (crypt regeneration) A check of the whole experimental procedure was performed using the intestinal crypt regeneration assay. The idea was to reproduce for this system the same RBE values as those we obtained on different occasions in the same beam (14), which would be a warrant for the consistency of the irradiation conditions (e.g., dosimetry). This control experiment included a total of 80 13-week-old males and females balb/c mice from the same genetic stock as the mice used for the lung experiment. Irradiation procedure (which is analogous to the one for lung except for the fact that the animals are irradiated to the whole body) has been given extensively elsewhere (15).

RESULTS Kinetics (thoracic irradiation) Animal mortality as a function of time after selective irradiation of the thorax is presented in the six panels of Fig. 2, for 60Co ␥-rays and 200 MeV protons, and for irradiations in 1, 3, or 10 fractions. For each treatment, 9 groups of 16 animals were irradiated with increasing doses (per fraction) ranged between the limits indicated on the figure. Proton irradiations were performed at the middle of a 7-cm SOBP. On the whole, only a limited number of animals died before the 3rd month, the highest death rate being observed between this period and the 8th month. The kinetics are the same for both radiations and do not depend on fractionation. On the other hand, one can see that the best spread of survival rates among the different doses is reached around the 180th day, which is the classical delay for evaluating lung tolerance in mice (6). However, dose–survival relationships have been drawn for later endpoints as well (see below). Dose–survival relationships (thoracic irradiation) Dose–response curves derived from the former data are presented in Fig. 3. Although it made no substantial difference, the survival rates have been corrected for intercurrent mortalities occurring between the first and the 30th day after irradiation (i.e., taking as 100% the survival rate at 30 days). Probit analysis was used to fit the relationship between survival and dose assuming that, for a given number of

1054

I. J. Radiation Oncology

●

Biology

●

Physics

Fig. 2. Survival as a function of time after selective irradiation of the thorax in mice with 60Co ␥-rays (left hand side panels) or protons (right hand side panels). Irradiations were given in a single, 3 or 10 equal fractions separated by a time interval of 12 h (upper, mid or lower panels, respectively). The solid and dotted lines correspond to 9 increasing doses in the ranges indicated on the panels. For protons, animals were positioned in the middle of a 7 cm SOBP. Each panel includes 144 mice, i.e., 16 animals per dose (15 for some doses). The arrows indicate the time delays for which dose–survival curves (see Fig. 3) have been derived.

fractions, the straight lines (in probit/linear coordinates) for ␥ and protons are parallel. The corresponding LD50s and their 95% confidence intervals are presented in Table 1 together with the RBEs for each fractionation modality and time delay. Due to the asymmetry of the LD50 confidence intervals, the RBE confidence intervals had to be estimated through an approximate method of calculation. They should be of the order of ⫾ 15–20% for survival at 180 and 210 days and of the order of ⫾ 30 –35% for survival at 240 and 270 days. The data indicate that the possibility for the RBE to increase with the number of fractions has to be rejected for all endpoints. Nevertheless, if the RBEs for single fractions appear to increase markedly with time, the RBEs for 3 and 10 fractions are keeping up. As can be seen in Fig. 3, there is a clear distinction between the dose–response curves for 1, 3, and 10 fractions, which persists for all the endpoints. The corresponding ␣/␤ ratios were estimated from the direct analysis of the “quantal radiation response data” of Thames et al. (16) and are presented in Table 2. In order to evaluate the influence of the position in the SOBP on the RBE, additional irradiations in 10 fractions were performed placing the animals at the end of the SOBP. As can be seen in Fig. 4, irradiations at the end of the SOBP (closed lozenges, new data) exhibit a tendency to be more

Volume 47, Number 4, 2000

Fig. 3. Dose–response curve for survival at different times derived from the data presented in Fig. 2. Open and closed symbols correspond to proton and 60Co ␥-ray irradiations; circles, squares, and triangles correspond to irradiations with a single, 3 or 10 fractions, respectively. Probit analysis was used to fit the relationship between survival and dose. The corresponding LD50 and RBEs are presented in Table 1; ␣/␤ ratios are presented in Table 2.

effective than irradiations at the middle of the SOBP (open triangles, same data as in Fig. 3), which observation applies for all the endpoints. The corresponding RBE increase averages ⬃6%. Histopathology The histopathologic analysis of the 96 considered samples revealed the classical symptoms of an irradiation-induced pneumonitis (17), such as, e.g., interalveolar macrophages and interstitial edema. They revealed also, e.g., obliterations of alveolar spaces with connective tissue and interstitial fibrosis, which can be attributed to a late chronic phase (18). Complete analysis of the histopathology data will be published elsewhere. Control experiment (intestinal crypt regeneration) Dose– effect relationships for intestinal crypt regeneration 3.5 days after 60Co ␥-rays or 200-MeV protons are presented in Fig. 5. Each point represents the mean number of regenerated crypts per circumference for 8 mice (4 females and 4 males). They were fitted by straight lines (in log-linear coordinates) according to the method that has been specially worked out for that purpose (19). The RBE of

Proton RBE for survival in mice after thoracic irradiation

●

J. GUEULETTE et al.

1055

Table 1. DL50 after thoracic irradiation in mice and relative biological effectiveness (RBE) of 200 MeV protons (irradiation in the middle of the SOBP) relative to 60Co ␥ rays DL50/Gy Time after irradiation

Number of fractions

180 days

1 3 10

210 days

1 3 10

240 days

1 3 10

270 days

1 3 10

60

Co

12.0 (10.8–13.1) 17.9 (15.2–20.0) 27.7 (19.8–31.1) 10.9 (9.3–12.2) 16.3 (13.2–18.4) 26.1 (21.0–19.1) 9.6 (7.7–11.1) 14.2 (9.2–16.7) 23.7 (11.7–27.7) 8.6 (6.4–10.1) 12.4 (3.7–15.7) 21.1 (7.8–25.4)

Protons

RBE*

11.5 (10.3–12.7) 16.4 (13.8–18.5) 32.2 (28.0–46.3) 9.1 (7.3–10.4) 15.2 (12.4–17.3) 27.2 (24.2–31.0) 7.3 (4.8–8.9) 13.8 (9.7–16.2) 22.5 (14.2–26.3) 5.5 (2.5–7.4) 11.7 (3.8–14.8) 20.6 (11.4–24.1)

1.04 1.09 0.86 1.20 1.07 0.96 1.33

Fig. 4. Comparison of dose-survival curves for proton irradiations (10 fractions) at the middle of the SOBP (open triangles and dotted lines, same data as in Fig. 3) and the end of the SOBP (closed lozenges and solid lines, new data).

1.02 1.05 1.55 1.06 1.02

* Confidence interval approximately ⫾ 20% (see text).

protons derived from the ratio of the doses corresponding to 20 regenerated crypts is equal to 1.14 ⫾ 0.05, i.e., the same value as those we obtained previously in the same beam (14).

same results as those we obtained previously vouches for the reliability of the experimental procedure and dosimetry. The major objective of this study was to determine whether or not proton RBE would increase when the dose decreases. This problem has been studied for a long time, especially by the team at Harvard University (160-MeV synchrocyclotron), who determined proton RBE in an animal tumour system and in murine small intestine and skin using fractionated dose schedules up to 20 fractions (21– 24). The general conclusion from these studies was that proton RBE as compared with 60Co ␥-rays does not vary

DISCUSSION Survival at 180 –240 days after selective irradiation of the thorax meets the requirements for determining a “reference” RBE, because mortality in this period is due to compromised lung function resulting from radiation pneumonitis and progressive fibrosis (6, 20). The clinical picture has been described very often and is corroborated by our histopathology analysis. Nevertheless, the fact that a simultaneous study of intestinal crypt regeneration yielded the Table 2. ␣/␤ for 200 MeV protons (irradiation at the middle of a 7-cm SOBP) and 60Co ␥ rays, for mice survival after thoracic irradiation

␣/␤ (Gy) Endpoint (days) 180 210 240 270

␥

Protons

2.4 (0.6–4.7) 2.4 (0.7–4.4) 2.5 (0.6–5.1) 2.2 (0.3–4.8)

2.4 (0.6–4.7) 2.3 (0.7–4.3) 2.3 (0.6–4.8) 2.0 (0.3–4.4)

Fig. 5. Number of regenerated crypts per circumference as a function of dose after irradiation with protons and 60Co ␥-rays. Proton irradiations were performed at the middle of a 7-cm SOBP. RBE is equal to 1.14 ⫾ 0.05 at the level, of 20 regenerated crypts per circumference (corresponding ␥ dose equals 14.7 Gy).

1056

I. J. Radiation Oncology

●

Biology

●

Physics

substantially with dose or fractionation, however, the absolute value of the RBE could vary according to the tissue or experimental endpoint. Our results corroborate the former conclusions and provide new evidence that proton RBE does not vary significantly with fractionation or, which is the same, with dose in the range of approximately 2–20 Gy, neither for the in vivo biological systems used in the Harvard experiments nor for lung tolerance assessed by LD50 used in the present experiment. Indeed, if the relatively large uncertainty in the accuracy of the RBEs presented in this study forbids disclosure of small RBE differences, the data are, however, fairly strong evidence for an absence of systematic increase of the RBE with fractionation. The moreso, as the values obtained for a single fraction are even higher than for 3 and 10 fractions (except for 180 days). Another evidence of the invariance of RBE with fractionation is brought by the fact that another type of statistical analysis (based on the analysis of the shape of the dose–response curves (25, 26) concluded that the dose-modifying factor (DMF) between ␥ and protons is the same for all data of a single endpoint (time delay). This, which means that RBE does not depend on dose or fractionation, allowed estimation of the following RBE values for the different endpoints: 0.98 ⫾ 0.07, 1,03 ⫾ 0.07, 1,07 ⫾ 0.11, and 1,09 ⫾ 0.09 for 180, 210, 240, and 270 days, respectively. These values are somewhat smaller than the mean values of the RBEs for 1, 3, and 10 fractions presented in Table 1 (mean values of the RBEs ⫽ 1.00, 1.08, 1.14, and 1.25 for 180, 210, 240, and 270 days, respectively) but are not statistically different. The moreso if we exclude the RBEs for 1 fraction that, most probably for statistical reasons, are beyond measure in comparison with 3 and 10 fractions. The lack of variation of RBE with fractionation is consistent with the fact that proton RBEs average relatively small values. As a matter of fact, a substantial increase of the RBE with fractionation would have required that the repair capacities after ␥ and proton irradiations be substantially different. This is unlikely because the ␥ and proton microdosimetric spectra are still close to each other and cover a LET range mostly under the flat part of the biological weighting function (27, 28). However, survival curves of cells responsible for late tissue reactions are known to have large repair capacity: ␣/␤ ratios derived from our experiments were found to be 2.4 Gy (for both proton and ␥ irradiations), which is fully consistent with a large repair capacity and in agreement with existing literature (9, 29). Therefore it cannot be ruled out that RBE may increase for doses per fraction smaller than 2 Gy. This will have to be investigated and should be kept in mind in case of hyper fractionated irradiations. Concerning the variation of the RBE with time, both types of statistical analysis indicated a tendency for the RBE to increase with this parameter. Considering all data and taking account of the reservations we made earlier in this discussion about the RBE values for single fractions, one can estimate that the increase of the RBE between 180 and

Volume 47, Number 4, 2000

270 days is ⬃10%. Variation of the RBE with time was already observed by Travis et al. (9) who, studying the repair capacity of lung after X-rays and 4-MeV neutrons, observed a slight but non significant increase of the neutron RBE for LD50 between 180 and 400 days. It is not possible to determine the relative proportions of pneumonitis- and fibrosis-related mortalities and to “correct” the death rates for two separate RBEs. Moreover, mortalities at time delays greater than 180 days may be due to increasing dominance of other irradiation-induced abnormalities, such as lymphomatosis and leukemia spread, which are very common in mice. In other respects, the number of doses and dose sizes employed for the different fractionation schemes was necessarily limited and centered on the LD50/180. This explains the progressive imbalance of the dose–response relationships and the increased dispersion of the RBEs with time, which prevents the making of further suppositions about the reason for the RBE increasing with this parameter. Therefore, a pragmatic attitude would consist in considering lung toxicity as a whole and assuming that the RBE for the “overall” late tolerance of lung could be represented by a single value. An estimation of it can be obtained by averaging the whole of the RBEs presented in Table 1 for all fractionations and endpoints (mean RBE, 1.12). In spite of the fact that this value is necessarily provided with large uncertainty, it is interesting to note that it is of the same order of magnitude or even smaller than the RBEs reported for the majority of the in vivo systems (21–24). Concerning the increase of the RBE at the end of the SOBP, this fact was already observed by us for crypt regeneration in mice (14) and by the vast majority of authors using cells in vitro (30 –34). From a theoretical point of view, an increase of the RBE at the end of the SOBP could be expected because the LET is lowest at the beginning of the SOBP (where many particles are passing through to reach greater depth) and highest at the end of the SOBP (where only stopping particles are present). Different authors using cells in vitro reported an 8 –15% increase of the RBE throughout the SOBP of medium-energy proton beam (⬃70 MeV) (30, 32), while Belli et al. (34) studying proton energies of a few MeV (i.e., energies comparable with those in the descending part of the SOBP) reported very high RBE values (in the range of 3–7), largely beyond what is obtained in the internal part of the SOBP. In the present case, the target volume (i.e., the lung) is spread over the thickness of the mouse (thickness ⫽ 1.5 cm) so that the measured RBE is actually an average for the corresponding LET interval. In another aspect, the increase of LET with depth is not linear but accelerates suddenly when the protons get near the end of their range (35). Therefore, the fact of moving the mice from the middle of the SOBP and adjusting them close to its end (approximately 0.5 cm apart from the distal edge) is likely to result in a smaller RBE increase in comparison with what could be expected from the LET values at the (true) middle and (true) end of the SOBP. This, together with the uncertainty in the accuracy of the data, explains that the RBE increase we observe (⬃6%)

Proton RBE for survival in mice after thoracic irradiation

is relatively small in comparison with the in vitro data. Nevertheless it has to be remembered that our results refer to irradiations in 10 fractions, which regimen appears thus not to amplify the radiosensitivity differences. CONCLUSIONS The conclusions can be summarized as follows: 1. Proton RBE (ref. 60Co ␥-rays) for lung tolerance assessed by survival at 180 –270 days after selective irradiation of the thorax in mice does not vary with fractionation or, which is the same, with dose in the range of ⬃2–20 Gy.

●

J. GUEULETTE et al.

1057

2. RBE slightly increases (⬃10%) with time between 180 and 270 days. RBEs for all fractionations and endpoints average 1.12. 3. The increase of the proton RBE from the middle to the end of the SOBP for lung irradiations in 10 fractions is not greater than what is reported for other in vivo or in vitro systems with single doses. 4. The classical value of the clinical RBE of protons is 1.10. There is no need to raise this value in consideration of the late tolerance of healthy tissues in the extent that RBE for lung tolerance in mice was found not to vary with fractionation nor to differ significantly from those of the majority of early and late responding tissues.

REFERENCES 1. Wambersie A, Batterman JJ. Practical problems related to RBE in neutrontherapy. In: Ka¨rcher KH, Kogelnik HD, Szepesi T, editors. Progress in radio-oncology III. Vienna: ICRO (International Club for Radio-Oncology); 1987 p. 155–162. 2. Mijnheer BJ, Batterman JJ, Wambersie A. What degree of accuracy is required and can be achieved in photon and neutron therapy? Radiother Oncol 1987;8:237–252. 3. Wambersie A, Menzel HG. Specification of absorbed dose and radiation quality in heavy particle therapy (a review). Radiat Prot Dosimetry 1997;70:517–527. 4. Menzel HZ, Wambersie A, Pihet P. The clinical RBE and microdosimetric characterization of radiation quality in fast neutron therapy. Acta Oncol 1994;33:251–259. 5. Gerweck L, Kozin S. Relative biological effectiveness of proton beams in clinical therapy. Radiother Oncol 1999;50: 135–142. 6. Travis EL, Harley RA, Fenn JO, et al. Pathologic changes in the lung following single and multi-fraction irradiation. Int J Radiat Oncol Biol Phys 1977;2:475– 490. 7. Siemann DW, Hill RP, Penney DP. Early and late pulmonary toxicity in mice evaluated 180 and 420 days following localized lung irradiation. Radiat Research 1982;89:396 – 407. 8. Field SB, Hornsey S. Damage to mouse lung with neutrons and X-rays. Eur J Cancer 1974;10:621– 627. 9. Parkins CS, Fowler JF, Maughan RL, et al. Repair in mouse lung for up to 20 fractions of X rays or neutrons. Br J Radiol 1985;58:225–241. 10. Gre´goire V, Beauduin M, Gueulette J, et al. Radiobiological intercomparison of p(45)⫹Be and p(65)⫹Be neutron beams for lung tolerance in mice after single and fractionated irradiations. Radiat Res 1993;133:27–32. 11. Wambersie A, Dutreix J, Gueulette J, et al. Early recovery for intestinal stem cells, as a function of dose per fraction, evaluated by survival rate after fractionated irradiation of the abdomen in mice. Radiat Res 1974;58:498 –515. 12. International Commission on Radiation Units and Measurements (ICRU). Determination of absorbed dose in patient irradiated by beams of X or ␥ rays in radiotherapy procedures (ICRU Report 24). Bethesda, MD: ICRU Publications; 1976. 13. Vynckier S, Bonnett DE, Jones DTL. Code of practice for clinical proton dosimetry. Radiother Oncol 1991;20:53– 63. 14. Gueulette J, Bo¨hm L, De Coster BM, et al. RBE variation as a function of depth in the 200-MeV proton beam produced at

15.

16.

17.

18. 19.

20. 21.

22.

23.

24. 25. 26. 27.

28.

29.

the National Accelerator Centre in Faure (South Africa). Radiother Oncol 1997;42:303–309. Gueulette J. Preclinical radiobiological experiments. In: Linz U, editor. Ion beam in tumor therapy. Weinheim: Chapman and Hall; 1995. p. 73– 82. Thames HD, Rozell ME, Tucker SL, et al. Direct analysis of quantal radiation response data. Int J Radiat Biol 1986;49: 999 –1009. Morgan CX, Breit SN. Radiation and the lung: A reevaluation of the mechanism mediating pulmonary injury. Int J Radiat Oncol Biol Phys 1995;31:361–369. Murray JC, Parkins CS. Collagen metabolism in mouse lung after X-irradiation. Radiat Res 1987;111:498 –510. Gueulette J, Menzel HG, Pihet P, et al. Specification of radiation quality in fast neutron therapy: microdosimetric and radiobiological approach. In: Engenhart-Cabillic R, Wambersie A, editors. Recent results in cancer research. Heidelberg: Springer-Verlag; 1998. p. 31–53. Field B, Hornsey S. Damage to mouse lung with neutrons and X-rays. Eur J Cancer 1974;10:621– 627. Tepper J, Verhey L, Goitein M, et al. In vivo determinations of RBE in a high energy modulated proton beam using normal tissue reactions and fractionated dose schedules. Int J Radiat Oncol Biol Phys 1977;2:1115–1122. Urano M, Goitein M, Verhey L, et al. Relative biological effectiveness of a high energy modulated proton beam using a spontaneous murine tumor in vivo. Int J Radiat Oncol Biol Phys 1980;6:1187–1193. Urano M, Verhey LJ, Goitein M, et al. Relative biological effectiveness of modulated proton beams in various murine tissues. Int J Radiat Oncol Biol Phys 1984;10:509 –514. Munzenrider JE, Austin-Seymour M, Blitzer PJ, et al. Proton therapy at Harvard. Strahlentherapie 1985;161:756 –763. Thames HD, Rasmussen SL. A test for dose-modifying factors. Radiat Res 1978;76:308 –324. Porter EH. The statistics of dose/cure relationships for irradiated tumours. Part I. Br J Radiol 1980;53:210 –227. Pihet P, Menzel HG, Schmidt R, et al. Biological weighting function for RBE specification of neutrontherapy beams. Intercomparison of 9 European centres. Radiat Prot Dosim 1990;31:437– 442. Loncol T, Cosgrove V, Denis JM, et al. Radiobiological effectiveness of radiation beams with broad LET spectra: microdosimetric analysis using biological weighting functions. Radiat Prot Dosimetry 1994;52:347–352. van Rongen E, Tan CHT, Durham SK. Late functional, biochemical and histological changes in the rat lung after frac-

1058

I. J. Radiation Oncology

●

Biology

●

Physics

tionated irradiation to the whole thorax. Radiother Oncol 1987;10:231–246. 30. Courdi A, Brassart N, Herault J, et al. The depth-dependent radiation response of human melanoma cells exposed to 65 MeV protons. Br J Radiol 1994;67:800 – 804. 31. Blomquist E, Russell KR, Stenerlow B, et al. Relative biological effectiveness of intermediate energy protons. Comparisons with 60Co gamma-radiation using two cell lines. Radiother Oncol 1993;28:44 –51. 32. Wouters BG, Lam GK, Oelfke U, et al. Measurements of relative biological effectiveness of the 70 MeV proton beam at

Volume 47, Number 4, 2000

TRIUMF using Chinese hamster V79 cells and the highprecision cell sorter assay. Radiat Res 1996;146:159 –70. 33. Tang JT, Inoue T, Inoue T, et al. Comparison of radiobiological effective depths in 65-MeV modulated proton beams. Br J Cancer 1997;76:220 –225. 34. Belli M, Cera F, Cherubini R, et al. Inactivation and mutation induction in V79 cells by low energy protons: re-evaluation of the results at the LNL facility. Int J Radiat Oncol Biol Phys 1993;63:331–337. 35. Larsson B. Pre-therapeutic physical experiments with high energy protons. Br J Radiol 1961;34:143.