- Email: [email protected]
From clinical observations of intensity-modulated radiotherapy to dedicated in vitro designs
Mutation Research 704 (2010) 200–205
Contents lists available at ScienceDirect
Mutation Research/Reviews in Mutation Research journal homepage: www.elsevier.com/locate/reviewsmr Community address: www.elsevier.com/locate/mutres
Review
From clinical observations of intensity-modulated radiotherapy to dedicated in vitro designs Ste´phanie Blockhuys a,b,*, Barbara Vanhoecke a, Carlos De Wagter b, Marc Bracke a, Wilfried De Neve b a b
Lab. Experimental Cancer Research, Ghent University Hospital, De Pintelaan 185, B-9000 Ghent, Belgium Department for Radiation Oncology and Experimental Cancer Research, Ghent University Hospital, De Pintelaan 185, B-9000 Ghent, Belgium
A R T I C L E I N F O
A B S T R A C T
Article history: Received 15 September 2009 Received in revised form 12 February 2010 Accepted 16 February 2010 Available online 21 February 2010
In this review, an overview of intensity-modulated radiotherapy (IMRT) and related high precision radiation techniques is presented. In addition, the related radiobiological issues are discussed. Hereby, we try to point to the potential differences in radiobiological effect between popular intensitymodulated radiotherapy and related techniques (IMRT+) and conventional or three-dimensional radiotherapy (3D-RT). Further, an overview of the existing in vitro and in vivo radiobiological models to investigate the effect of spatially and/or temporally fractionated dose distributions, as applied in IMRT+, on the biological outcome is given. More in detail, our radiobiological models will be presented. Additionally, we will discuss the (dis)advantages of the presented models, and give some consideration to improve the existing radiobiological models in terms of set-up and clinical relevance. ß 2010 Elsevier B.V. All rights reserved.
Keywords: Intensity-modulated radiotherapy Fractionated dose distribution Radiobiology Dosimetry Models Bystander effect
Contents 1.
2.
3. 4. 5.
Intensity-modulated radiotherapy (IMRT) and related radiobiological issues . . . . . . . . . . . . . . . 1.1. Static beam IMRT techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2. Single-arc techniques and IMAT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3. Tomotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4. CyberKnife. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Radiobiological models for the investigation of the effect of STFDD on the biological outcome . 2.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. The radiobiological models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1. The radiophysical table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2. The radiobiological table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Abbreviations: 3D-CRT, conventional or three-dimensional radiotherapy; 3D-RT, three-dimensional radiotherapy; C-IMRT, compensator IMRT; cRT, conventional radiotherapy; CV, crystal violet; DMSO, dimethylsulfoxide; IMAT, intensitymodulated arc therapy; IMRT, intensity-modulated radiotherapy; IMRT+, intensity-modulated radiotherapy and related techniques; MLC, multileaf collimators; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide; MN, micronuclei; OD, optical density; S-Arc, single arc techniques; SLDR, sub-lethal damage repair; STFDD, spatio-temporal fractionated dose distribution; SW-IMRT, sliding window IMRT; SS-IMRT, step-and-shoot IMRT. * Corresponding author at: Department of Radiation Oncology and Experimental Cancer Research, Ghent University Hospital, De Pintelaan 185 (1P7), B-9000 Ghent, East-Flanders, Belgium. Tel.: +32 93323063. E-mail address: [email protected] (S. Blockhuys). 1383-5742/$ – see front matter ß 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.mrrev.2010.02.003
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
200 201 201 202 202 202 202 202 202 203 203 204 205 205
1. Intensity-modulated radiotherapy (IMRT) and related radiobiological issues The technology to deliver radiation therapy has been enormously changed during the last decade. IMRT and related high precision radiation techniques created the possibility to generate dose distributions that can be tailored to fit tumors of a complex geometrical shape while avoiding nearby or even surrounded radiosensitive normal tissues. The delivery of these techniques disregards the basic paradigm of earlier technology called conventional or three-dimensional radiotherapy (3D-CRT) which
S. Blockhuys et al. / Mutation Research 704 (2010) 200–205
used beams that covered the entire target with an intensity that was either uniform or linearly increasing in one direction across the beam’s cross-section. Using these earlier techniques, beams were delivered in quick succession with each beam contributing significant dose to the entire target and the total delivery procedure taking less than 4–10 min. An unifying principle of IMRT and related techniques, further called IMRT+, is the sequential use of numerous small beams which irradiate each only part of the target but temporally combine to generate radiation dose clouds that cover the entire target. Consequently, the target tissue is exposed to a spatio-temporal fractionated dose distribution (STFDD). IMRT+ differs substantially in the way they deliver the spatial as well as the temporal component of the STFDD. The spatial component ranges from a few tenths of dose depositions to parts of the target (e.g. IMRT close-in step-and-shoot techniques) to many thousands (e.g. serial or helical tomotherapy techniques) to deliver a single radiation fraction. Likewise, to deliver the fraction dose to each part of the target, the temporal component ranges from a sequence very short interruptions of dose deposition over a few minutes (e.g. mono-arc techniques like VMAT or RapidArc) to thousands of interruptions— with some of substantial length—over more than an hour (e.g. CyberKnife when used for large targets). The determination of the radiobiologically relevant features of the resulting 4D dose distribution – time being the 4th dimension – is extremely challenging but necessary. This task is different from the early attempt to quantitatively characterize the spatial beam modulation from a delivery point of view as undertaken by Webb in 2003 [1]. Indeed, the degree of absorbed dose modulation – rather than the degree of applied beam modulation – should be taken as the basis for making radiobiological inferences. At least 2 contemporary evolutions further complicate the radiobiological interpretation of various IMRT+ approaches: techniques to irradiate moving targets (gating, freezing, and tracking) and the increasing popularity of hypo-fractionation. STFDDs differ enormously between the various IMRT+. For a homogeneous 2 Gy fraction, the biological effect of some IMRT+ may closely resemble that of the 3D-CRT. For other techniques substantial differences can be expected. Assuming that in vitro observations listed in Table 1 are relevant in the clinic, we hereby try to point to the potential differences in radiobiological effect between popular IMRT+ and 3D-CRT.
201
roughly comparable to 3D-RT regarding the time needed for delivering a 2 Gy fraction; typically less than 10 min. Little or no evidence exists that cell survival is significantly affected by variations of temporal fractionation schemes within a maximum of 10 min overall delivery time. C-IMRT allows creating extremely steep dose gradients. In such cases spatial fractionation by single beams will be very different between compensator IMRT and 3DCRT but gradients will average out by integrating multiple beams. Whether bystander effects across temporally spaced opposed gradients will average out is unknown. Overall, C-IMRT can be expected to provide mostly similar cell survival as 3D-CRT after delivery of a homogeneous 2 Gy fraction. The use of physical compensators went rapidly out of fashion with the advent of computer controlled multileaf collimators (MLC) in the mid-90s. One MLC technique mostly uses a narrow window consisting of all of the leaves that cover the length of the target and sweeps this opening across its volume. The exact size of the opening for each leaf pair is modulated according to the desired dose at each point in the target volume [3]. This sliding window method is often implemented in a continuous dynamic fashion to speed up the dose delivery. Earlier implementations required reduced dose rates of the linear accelerator. Together with the relative inefficiency of sliding window as compared to compensator IMRT, overall delivery times increased by a factor of 2–3 times. Biologically, 2-Gy delivery times exceeding 15–20 min raise questions of temporal fractionation. This would translate to local average dose rates lower than 14 cGy/min for which increased cell survival can be expected. Another popular IMRT approach used superimposed static fields from each beam direction that are sequentially delivered. The beam is halted during the transition from one field to the next when the MLC leaves accomplish the required motion [4]. The delivery is performed as step-and-shoot sequences with beam-off during the leaf motion (step) and beamon (shoot) with fixed collimator aperture. Regarding spatial and temporal fractionation, the same reasoning as for sliding window can be made although the final dose distribution of step-and-shoot IMRT (SS-IMRT) tends to be less smooth. Over the last decade, delivery efficiency has increased. Modern implementations of sliding window IMRT (SW-IMRT) and SS-IMRT are efficient and probably biologically comparable to C-IMRT regarding their temporal component. 1.2. Single-arc techniques and IMAT
1.1. Static beam IMRT techniques IMRT can be delivered in a very robust way by the use of physical compensators that are put in the beam line [2]. IMRT using compensators will require entering the treatment room to change compensators between beams. Entering the room inbetween three-dimensional radiotherapy (3D-RT) beams will also be needed in case of non-coplanar set-up or manually inserted wedges. Overall, multi-beam compensator IMRT (C-IMRT) is
Intensity-modulated arc therapy (IMAT) by means of standard MLC was invented by Yu in the early 90s [5]. Recently, IMATrelated single-arc techniques were developed, their common property being the delivery of a 2-Gy fraction within one or a few minutes, much faster than static beam IMRT techniques and even faster than multi-beam 3D-CRT. Although spatial and temporal modulation occurs at high mean frequency the biological effect of mono-arc techniques may closely resemble that occurring after
Table 1 Current radiation therapy modalities with number of corresponding radiobiological papers found in Web of Science.
CyberKnife Tomotherapy IMAT S-Arc SW-IMRT SS-IMRT C-IMRT 3D-RT cRT
Search keywords phrase in Web of Science
No. of papers
Topic = (CyberKnife and radiobiology) Topic = (tomotherapy) AND Topic = (radiobiology) Topic = (IMAT or VMAT or RapidArc) AND Topic = (radiobiology) Topic = (arc AND radiobiology) + manual inspection Topic = (IMRT and radiobiology and sliding window) Topic = (IMRT and radiobiology) + manual inspection Topic = (IMRT and radiobiology) + manual inspection Topic = (conformal and radiotherapy and radiobiology) NOT Topic = (IMRT) Topic = (radiotherapy and radiobiology) NOT Topic = (IMRT) NOT Topic = (conformal)
3 3 1 0 1 1 0 25 768
202
S. Blockhuys et al. / Mutation Research 704 (2010) 200–205
high dose continuous irradiation. The hypothesis of (slightly) improved anti-tumor efficacy by short delivery times could be investigated.
2. Radiobiological models for the investigation of the effect of STFDD on the biological outcome 2.1. Introduction
1.3. Tomotherapy Tomotherapy techniques employ a radiation source rotating around the patient that – using a fast binary slit collimator – deposits adjacent slices (serial tomotherapy) or a single helical slice (helical tomotherapy) of dose across the target. Within the slice, spatial and temporal modulation is enormous [6]. Each elementary volume inside the target may be hit by hundreds of small dose deposits although within a time-span that is as short as for mono-arc techniques. If we look at each rotation separately, the biological effect could be considered equal to mono-arc techniques. Different from mono-arc techniques is the propagation of a high-gradient slice edge across the target volume over a time period that may exceed 10 min for large targets. To the best of our knowledge, the effect of a propagating gradient has never been modeled in vitro or in vivo.
In radiotherapy, the dose–response curves established for uniform dose distribution are still used to determine the biological response for irradiation with spatially and temporally fractionated dose distributions. The latter comes from the commonly accepted hypothesis that the radiobiological effect depends on the individual cell radiation sensitivity, and is not influenced by, for example, a response of the surrounding cells [8]. Nowadays, the validity of this hypothesis is seriously under discussion in the context of spatially and temporally fractionated irradiation. Indeed, a number of in vitro and in vivo studies show a different biological outcome for irradiation with a spatially [9–13] and/or temporally fractionated dose [14–22] suggesting that cells surrounding the target tissue might affect the radiobiological outcome [23]. Therefore, there is a need for more reliable radiobiological models to accurately determine the biological outcome of spatio-temporal fractionated irradiation.
1.4. CyberKnife 2.2. The radiobiological models The Accuray CyberKnife uses a linear accelerator with circular collimators mounted on a robotic arm. CyberKnife implements a ‘‘nodal’’ delivery method of IMRT. Nodes are small spheres of radiation which are positioned throughout the target volume to create the desired dose distribution. Each of the numerous nodes is irradiated from many directions in space. ‘Walking’ through all the nodes may take more than 1 h for large targets. Irradiating one node almost invariably contributes dose to other nodes. Enormous numbers of small dose deposits to small volumes over long treatment times result in extreme forms of STFDDs. The risk for increased cell survival by very low overall dose rate is substantial for a 2 Gy fraction. However, CyberKnife is typically used to deliver fraction sizes which are much larger than 2 Gy. The contribution of tumor-associated host tissue to therapeutic response may become very important – if not dominant – for large fraction sizes [7]. Techniques to deal with moving targets include gating, freezing and tracking. Gating, freezing as well as the CyberKnife implementation of tracking cause additional interrupts in delivery. These interrupts potentially worsen the temporal component of STFDD with regard to tumor cell survival. When studying the spatio-temporal modulation characteristics of the different radiotherapy modalities, we can conclude the following: (1) more spatial fractionation in ‘Tomotherapy, Cybernife and IMAT’, in comparison with ‘3D-RT and conventional radiotherapy (cRT)’, (2) more temporal fractionation for ‘Cybernife, IMRT and 3D-RT’, in comparison with ‘cRT, single-arc techniques (S-Arc)’, IMAT and Tomotherapy. Further, in Table 1 we have compiled a list containing the total number of radiobiological studies conducted for the different radiation therapy modalities using the Web of Science1 (Thomson-ISI). cRT has the highest number of papers (768) followed by 3D-RT (25), Cybernife (3) and Tomotherapy (3). IMAT, SW-IMRT and SS-IMRT are reported in only one radiobiology paper. Both SArc and C-IMRT are not reported in a radiobiological paper. We can conclude that the recent techniques with more spatio-temporal modulation are not frequently reported in radiobiological papers. So, there is clearly a cumulated time lag between the technology and clinically oriented studies and the radiobiologically oriented work. Consequently, more radiobiological research investigating treatment modalities with spatio-temporal fractionation is urgently required to optimize their therapeutic potential and efficiency.
We compiled in two complementary tables the published radiobiological models that have been developed and used so far to investigate the effect of spatially and temporally fractionated irradiation on the radiobiological response (see Tables 2A and 2B). Table 2A, namely ‘the radiophysical table’, contains the parameters radiation beam and dose characteristics, fractionation (spatial and/ or temporal modulation) and dosimeter. Table 2B, namely ‘the radiobiological table’, contains the parameters cell line, assay (endpoint) and results. 2.2.1. The radiophysical table 2.2.1.1. Experimental set-up (spatial and/or temporal fractionation). For the evaluation of the spatially fractionated dose distribution, two model systems have been used. First, Mackonis et al. [13] and Bijl et al. [9] investigated the effect of segmented dose distributions, namely the partial fields ‘quarter’ and ‘stripped’ field, and the radiation set-ups ‘split-field’, ‘symmetrical’- and ‘asymmetrical bath and shower’ respectively, whereas others applied a linear dose gradient [10,11,24]. For example, we developed 3 different in vitro set-ups [24]. According to our Setup 1, a cell monolayer is irradiated with 3 small radiation fields creating steep dose gradients at the border of the fields. Set-up 1 is closely related with the in vivo situation where tumors are irradiated to spatially and temporally fractionated high doses, resulting in steep dose gradients within and at the border of the irradiated volumes. According to Set-ups 2 and 3, the cell monolayers in a T25 flask and a 96-well plate, respectively, are irradiated with a linear smooth dose gradient along the length of the recipient. To evaluate the effect of the temporal aspect of IMRT+, the following parameters have been examined: (1) total dose delivery time [12,14,19,21], (2) intra-fraction dose rate [20], (3) dose per fraction [17], (4) inter-fraction time [15,18,22], and (5) fraction number [16]. 2.2.1.2. Dosimetry. Since the calculation and planning methods applied in clinic for these kinds of experimental set-ups result in less accurate dose measurements, dosimetry is strongly recommended [11,13,24]. For our set-ups, we performed dosimetry with the radiochromic EBT film. This dosimeter allows dose measurement in cell geometry, for Set-ups 1 and 2, more precisely by
S. Blockhuys et al. / Mutation Research 704 (2010) 200–205
203
Table 2A Radiobiological models developed to investigate the biological effect of spatio-temporal modulated irradiation: the radiophysical table.
Chronological presentation of the radiobiological models developed so far to investigate the effects of irradiation with spatio-temporal fractionated dose distributions on the biological outcome. Parameters: radiation beam and dose characteristics, fractionation type (spatial and/or temporal modulation) and dosimeter.
replacement of the cell monolayer by stack of film and polystyrene sheets. For Model 3, a fitting piece of film was positioned immediately under the bottom surface of the 96-well plate [25]. The use of film dosimetry with EBT radiochromic film has different advantages. (1) Film versus ionization chamber: film allows the determination of the dose distribution over a certain area, in contrast to the measurement at a single point with the ionization chamber [11,12,15,16,20,21,26]. (2) Radiochromic [13,24] versus radiographic film [11,13]: the radiochromic film is relatively light insensitive in comparison to the radiographic film, and is more suitable to cut into the desired form. Consequently, radiochromic film dosimetry can be done in the cell geometry. Additionally, radiochromic film is nearly water-equivalent and features a flat energy response for high-energy X-rays. Dosimetry has also been done in the cell geometry albeit with the parallel plate ionization chamber [9,10] and with the Fricke dosimeter [14].
the MTT assay and conclude that the MTT assay is a relevant assay to investigate the effect of spatial fractionated irradiation. The MTT assay is a common tool in radiobiology to assess radiosensitivity. It is a sensitive, quantitative and reliable assay that measures the conversion of the soluble yellow MTT substrate into insoluble dark blue formazan salt by cellular dehydrogenases dependent on the metabolic activity [27]. Consequently, the formazan optical density (OD) is equivalent with the metabolic activity, which is determined by cells proliferation and/or cell metabolic activity. In contrast to the classical protocol, as described by Bromley et al. [11], whereby in the final step formazan crystals are dissolved in DMSO, we applied an adapted MTT protocol, whereby the final solubilization step is not necessary for quantification [24]. Indeed, DMSO solubilization erases all in situ information about the spatial distribution of cell growth. 3. Results
2.2.2. The radiobiological table 2.2.2.1. Assays and endpoints. So far, different biological assays, namely crystal violet (CV) staining, clonogenic assay, micronuclei (MN) visualisation, g-H2AX staining, cell cycle analysis, and 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT) assay, have been applied to determine the biological effect of spatio-temporal fractionated irradiation. These assays are used to investigate different endpoints, namely cell survival, cell growth, cell proliferation, cell cycle inhibition, DNA damage and metabolic activity. Until now, only the CV staining and the clonogenic assay have been used to investigate the effect of spatial fractionated dose distribution [10–13]. In contrast to previous models, we applied
From the radiobiological results with spatially and/or temporally fractionated dose, we may conclude that there is an additional effect on the radiobiological outcome as predicted from the current treatment planning systems. For spatially fractionated dose distribution, the results from in vitro experiments indicate that a bystander effect, i.e. biological effect due to intercellular communication in between cells irradiated with different radiation doses, may be responsible for an additional effect [9–13,24]. A recent letter by Ross and Klassen [28] recently commented on the research-paper of Mackonis et al. [13], more precisely on their determination of two new types of radiation-induced bystander effects (II and III), which would not be supported by their presented data. Additionally, radiation-induced
204
S. Blockhuys et al. / Mutation Research 704 (2010) 200–205
Table 2B Radiobiological models developed to investigate the biological effect of spatio-temporal modulated irradiation: the radiobiological table.
Chronological presentation of the radiobiological models developed so far to investigate the effects of irradiation with spatio-temporal fractionated dose distributions on the biological outcome. This table reports cell line, assay (endpoint) and results. Note that the in vivo studies are indicated with an asterisk added to the respective study number, and the studies describing spatial modulation are marked in gray.
bystander effects are not always detected. Groesser et al. [29] published that possibly the epigenetic status of the specific cell line or the precise culture conditions and medium supplements, such as serum, may be critical for inducing bystander effects. Our results suggest that the MTT assay can also be used to reveal bystander effects [24]. We observed an increased metabolic activity for intermediate doses with the MTT assay for irradiation according to Set-ups 1–3. In contrast, no enhanced colour intensity was observed at the intermediate dose region when using CV staining, indicating that the increased metabolic activity is not due to an increased cell number but may be due to an increased cell metabolic activity. Further, we observed differences in peak expression time and related intermediate dose region between the different models. Since Set-ups 2 and 3 only differ in the possibility of intercellular communication, the latter may be due to bystander effects. Results from in vitro models for temporally fractionated dose distribution indicate that: (1) the irradiation time must be kept as short as possible, since an increase in total delivery time for a fixed total dose results in an increase of cell survival, which is at least partially due to sub-lethal damage repair (SLDR). Hereby, multiple parameters can result in an increased radiation-delivery time, namely dose rate [14,20], fraction number [16,19,26] and beaminterruption time [15,19]. (2) Some radiation doses can lead to hypo- and hyper-sensitive biological responses. For example when a 2 Gy dose is divided into multiple partial fractions, fractions less
than 0.5 Gy (S) can lead to hyper-sensitive and those higher that 0.5 or 1 Gy (L) can lead to hypo-sensitive biological response. Consequently, the sequence (L–S or S–L) in which the fractions are delivered must be taken into account [17]. In parallel with the in vitro experiments, the in vivo experiments suggest an additional effect of intercellular communication and of SLDR delay for spatially [9] and temporally fractioned irradiation [18,22], respectively. In contrast to the in vitro experiments for temporal fractionation, the in vivo studies suggest that the increase in survival due to extended delivery time for the different in vitro studies may be diminished due to decrease of SLDR when reoxygenation takes place. However, since the magnitude and velocity of reoxygenation of human tumors are unpredictable, the clinical effect of intermittent irradiation cannot be derived from the radiobiological outcome for continuous irradiation. These few in vivo models suggest that a short radiation-time is required considering the possibility of slow or insufficient reoxygenation of the tumor. 4. Discussion In general, we can conclude from the above-presented results, that spatially and temporally fractionated irradiation can result in additional effects on the biological outcome, namely multiple effects of intercellular communication (bystander effects) and of radiation dose delivery time (enhanced SLDR as a result of
S. Blockhuys et al. / Mutation Research 704 (2010) 200–205
increased total dose delivery time), respectively. These additional effects have to be taken into account to result in a good clinical prediction. We suggest an optimization of existing in vitro models. Therefore, following aspects should be taken into account: (1) for dosimetry, dosimeters have to be employed which allow the determination of the dose distribution in the cell geometry. This includes the replacement of the cells by a tissue-equivalent material in the in vitro set-up, in order to assess the spatial features of the dose distribution and avoid the inaccuracies in dose measurement due to a different set-up. (2) For radiobiology, more reliable experimental set-ups should be designed to approach the in vivo (radiotherapeutic) setting. Therefore, special attention should be paid to radiation dose, STFDD and clinical equipment. (3) In order to take into account the additional effects in clinical settings, it is important to assess multiple endpoints. Apart from cell viability also other endpoints should be considered, like cellular motility and invasion [30]. Therefore, it is interesting to apply for example the MTT assay, which measures metabolic activity and might give the opportunity to detect cells with enhanced motility and invasive capacities. Additionally, for temporal fractionated irradiation cell reoxygenation should be investigated since SLDR may be counterbalanced by reoxygenation. 5. Conclusion From a clinical point of view, we can conclude that radiation techniques with spatio-temporal dose modulation have become popular in radiotherapy. One of the most important advantage includes the delivery of a highly controlled dose distribution conformal to the tumor creating the possibility to raise the dose to the tumor with less toxicity to the surrounding healthy tissues. However, radiobiological research has shown that the biological outcome of spatio-temporal fractionated irradiation cannot be fully explained using the existing dose–response data. Consequently, further research is needed to establish more reliable radiobiological models to predict the clinical outcome of spatiotemporally fractionated radiation therapy. Conflict of interest The authors declare that there are no conflicts of interest. References [1] S. Webb, Use of a quantitative index of beam modulation to characterize dose conformality: illustration by a comparison of full beamlet IMRT, few-segment IMRT (fsIMRT) and conformal unmodulated radiotherapy, Phys. Med. Biol. 48 (2003) 2051–2062. [2] G.W. Sherouse, In regard to Intensity-Modulated Radiotherapy Collaborative Working Group, IJROBP 51 (2001) 880–914; G.W. Sherouse, Int. J. Radiat. Oncol. Biol. Phys. 53 (2002) 1088–1089. [3] D.J. Convery, M.E. Rosenbloom, The generation of intensity modulated fields for comformal radiotherapy by dynamic collimator, Phys. Med. Biol. 37 (1992) 1359– 1374. [4] J.M. Galvin, X.G. Chen, R.M. Smith, Combining multileaf fields to modulate fluence distributions, Int. J. Rad. Oncol. Biol. Phys. 27 (1993) 697–705.
205
[5] C.X. Yu, Intensity-modulated arc therapy with dynamic multileaf collimation: an alternative to tomotherapy, Phys. Med. Biol. 40 (1995) 1435–1449. [6] R.T. Flynn, M.W. Kissick, M.P. Mehta, G.H. Olivera, R. Jeraj, T.R. Mackie, The impact of linac output variations on dose distributions in helical tomotherapy, Phys. Med. Biol. 53 (2008) 417–430. [7] Z. Fuks, R. Kolesnick, Engaging the vascular component of the tumor response, Cancer Cell 8 (2005) 89–91. [8] E.J. Hall, Radiobiology for the Radiologist, Lippincott Williams and Wilkins, Philadelphia, PA, 2000. [9] H.P. Bijl, P. van Luijk, R.P. Coppes, J.M. Schippers, A.W.T. Konings, A.J. van der Kogel, Unexpected changes of rat cervical spinal cord tolerance caused by inhomogeneous dose distributions, Int. J. Radiat. Oncol. Biol. Phys. 57 (2003) 274–281. [10] N. Suchowerska, M. Ebert, M. Zhang, M. Jackson, In vitro response of tumour cells to non-uniform irradiation, Phys. Med. Biol. 50 (2005) 3041–3051. [11] R. Bromley, R. Davey, L. Oliver, R. Harvie, C. Baldock, A preliminary investigation of cell growth after irradiation using a modulated X-ray intensity pattern, Phys. Med. Biol. 51 (2006) 3639–3651. [12] V. Moiseenko, C. Duzenli, R.E. Durand, In vitro study of cell survival following dynamic MLC intensity-modulated radiation therapy dose delivery, Med. Phys. 34 (2007) 1514–1520. [13] E. Mackonis, N. Suchowerska, M. Zhang, M. Ebert, D. McKenzie, M. Jackson, Cellular response to modulated radiation fields, Phys. Med. Biol. 52 (2007) 5469–5482. [14] X. Mu, P. Lo¨froth, M. Karlsson, B. Zackrisson, The effect of fraction time in intensity modulated radiotherapy: theoretical and experimental evaluation of an optimisation problem, Radiother. Oncol. 68 (2003) 181–187. [15] Y. Shibamoto, M. Ito, C. Sugie, H. Ogino, M. Hara, Recovery from sublethal damage during intermittent exposures in cultured tumor cells: implications for dose modification in radiosurgery and IMRT, Int. J. Radiat. Oncol. Biol. Phys. 59 (2004) 1484–1490. [16] H. Ogino, Y. Shibamoto, C. Sugie, M. Ito, Biological effects of intermittent radiation in cultured tumor cells: influence of fraction number and dose per fraction, J. Radiat. Res. 46 (2005) 401–406. [17] P. Lin, A. Wu, Not all 2 Gray radiation prescriptions are equivalent: cytotoxic effect depends on delivery sequences of partial fractionated doses, Int. J. Radiat. Oncol. Biol. Phys. 63 (2005) 536–544. [18] C. Sugie, Y. Shibamoto, M. Ito, H. Ogino, A. Mitamoto, N. Fukaya, H. Niimi, T. Hashizume, Radiobiologic effect of intermittent radiation exposure in murine tumors, Int. J. Radiat. Oncol. Biol. Phys. 64 (2006) 619–624. [19] V. Moiseenko, J. Banath, C. Duzenli, P. Olive, Effect of prolongation radiation delivery time on retention of gammaH2AX, Radiat. Oncol. 3 (2008). [20] J. Bewes, N. Suchowerska, M. Jackson, M. Zhang, D. McKenzie, The radiobiological effect of intra-fraction dose-rate modulation in intensity modulated radiation therapy (IMRT), Phys. Med. Biol. 53 (2008) 3567–3578. [21] P. Keall, M. Chang, S. Benedict, H. Thames, S. Vedam, P. Lin, Investigating the temporal effects of respiratory-gated and intensity-modulated radiotherapy treatment delivery on in vitro survival: an experimental and theoretical study, Int. J. Radiat. Oncol. Biol. Phys. 71 (2008) 1547–1552. [22] N. Tomita, Y. Shibamoto, M. Ito, H. Ogino, C. Sugie, S. Ayakawa, H. Iwata, Biological effect of intermittent radiation exposure in vivo: recovery from sublethal damage versus reoxygenation, Radiother. Oncol. 86 (2008) 369–374. [23] W.F. Morgan, M.B. Sowa, Non-targeted bystander effects induced by ionizing radiation, Mutat. Res. 616 (2007) 159–164. [24] S. Blockhuys, B. Vanhoecke, L. Paelinck, M. Bracke, C. De Wagter, Development of in vitro models for investigating spatially fractionated irradiation: physics and biological results, Phys. Med. Biol. 54 (2009) 1565–1578. [25] N. Tomic, M. Gosselin, J.F. Wan, U. Saragovi, E.B. Podgorsak, M. Evans, S. Devic, Verification of cell irradiation, Phys. Med. Biol. 50 (2007) 3121–3131. [26] F. Sterzing, M. Mu¨nter, M. Scha¨fer, P. Haering, B. Rhein, C. Thilmann, J. Debus, Radiobiological investigation of dose-rate effects in intensity-modulated radiation therapy, Strahlenther. Onkol. 181 (2005) 42–48. [27] T. Slater, Studies on a succinate-neotetrazolium reductase system of rat liver: II. Points of coupling with the respiratory chain, Biochem. Biophys. Acta 77 (1963) 365–382. [28] C.K. Ross, N.V. Klassen, Comments on ‘Cellular response to modulated radiation fields’, Phys. Med. Biol. 54 (2009) L11–L13. [29] T. Groesser, B. Cooper, B. Rydberg, Lack of bystander effects from high-LET radiation for early cytogenic end points, Radiat. Res. 170 (2008) 794–802. [30] I. Madani, W. De Neve, M. Mareel, Does ionizing radiation stimulate cancer invasion and metastasis, Bull. Cancer 95 (2008) 292–300.
Contents lists available at ScienceDirect
Mutation Research/Reviews in Mutation Research journal homepage: www.elsevier.com/locate/reviewsmr Community address: www.elsevier.com/locate/mutres
Review
From clinical observations of intensity-modulated radiotherapy to dedicated in vitro designs Ste´phanie Blockhuys a,b,*, Barbara Vanhoecke a, Carlos De Wagter b, Marc Bracke a, Wilfried De Neve b a b
Lab. Experimental Cancer Research, Ghent University Hospital, De Pintelaan 185, B-9000 Ghent, Belgium Department for Radiation Oncology and Experimental Cancer Research, Ghent University Hospital, De Pintelaan 185, B-9000 Ghent, Belgium
A R T I C L E I N F O
A B S T R A C T
Article history: Received 15 September 2009 Received in revised form 12 February 2010 Accepted 16 February 2010 Available online 21 February 2010
In this review, an overview of intensity-modulated radiotherapy (IMRT) and related high precision radiation techniques is presented. In addition, the related radiobiological issues are discussed. Hereby, we try to point to the potential differences in radiobiological effect between popular intensitymodulated radiotherapy and related techniques (IMRT+) and conventional or three-dimensional radiotherapy (3D-RT). Further, an overview of the existing in vitro and in vivo radiobiological models to investigate the effect of spatially and/or temporally fractionated dose distributions, as applied in IMRT+, on the biological outcome is given. More in detail, our radiobiological models will be presented. Additionally, we will discuss the (dis)advantages of the presented models, and give some consideration to improve the existing radiobiological models in terms of set-up and clinical relevance. ß 2010 Elsevier B.V. All rights reserved.
Keywords: Intensity-modulated radiotherapy Fractionated dose distribution Radiobiology Dosimetry Models Bystander effect
Contents 1.
2.
3. 4. 5.
Intensity-modulated radiotherapy (IMRT) and related radiobiological issues . . . . . . . . . . . . . . . 1.1. Static beam IMRT techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2. Single-arc techniques and IMAT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3. Tomotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4. CyberKnife. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Radiobiological models for the investigation of the effect of STFDD on the biological outcome . 2.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. The radiobiological models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1. The radiophysical table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2. The radiobiological table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Abbreviations: 3D-CRT, conventional or three-dimensional radiotherapy; 3D-RT, three-dimensional radiotherapy; C-IMRT, compensator IMRT; cRT, conventional radiotherapy; CV, crystal violet; DMSO, dimethylsulfoxide; IMAT, intensitymodulated arc therapy; IMRT, intensity-modulated radiotherapy; IMRT+, intensity-modulated radiotherapy and related techniques; MLC, multileaf collimators; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide; MN, micronuclei; OD, optical density; S-Arc, single arc techniques; SLDR, sub-lethal damage repair; STFDD, spatio-temporal fractionated dose distribution; SW-IMRT, sliding window IMRT; SS-IMRT, step-and-shoot IMRT. * Corresponding author at: Department of Radiation Oncology and Experimental Cancer Research, Ghent University Hospital, De Pintelaan 185 (1P7), B-9000 Ghent, East-Flanders, Belgium. Tel.: +32 93323063. E-mail address: [email protected] (S. Blockhuys). 1383-5742/$ – see front matter ß 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.mrrev.2010.02.003
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
200 201 201 202 202 202 202 202 202 203 203 204 205 205
1. Intensity-modulated radiotherapy (IMRT) and related radiobiological issues The technology to deliver radiation therapy has been enormously changed during the last decade. IMRT and related high precision radiation techniques created the possibility to generate dose distributions that can be tailored to fit tumors of a complex geometrical shape while avoiding nearby or even surrounded radiosensitive normal tissues. The delivery of these techniques disregards the basic paradigm of earlier technology called conventional or three-dimensional radiotherapy (3D-CRT) which
S. Blockhuys et al. / Mutation Research 704 (2010) 200–205
used beams that covered the entire target with an intensity that was either uniform or linearly increasing in one direction across the beam’s cross-section. Using these earlier techniques, beams were delivered in quick succession with each beam contributing significant dose to the entire target and the total delivery procedure taking less than 4–10 min. An unifying principle of IMRT and related techniques, further called IMRT+, is the sequential use of numerous small beams which irradiate each only part of the target but temporally combine to generate radiation dose clouds that cover the entire target. Consequently, the target tissue is exposed to a spatio-temporal fractionated dose distribution (STFDD). IMRT+ differs substantially in the way they deliver the spatial as well as the temporal component of the STFDD. The spatial component ranges from a few tenths of dose depositions to parts of the target (e.g. IMRT close-in step-and-shoot techniques) to many thousands (e.g. serial or helical tomotherapy techniques) to deliver a single radiation fraction. Likewise, to deliver the fraction dose to each part of the target, the temporal component ranges from a sequence very short interruptions of dose deposition over a few minutes (e.g. mono-arc techniques like VMAT or RapidArc) to thousands of interruptions— with some of substantial length—over more than an hour (e.g. CyberKnife when used for large targets). The determination of the radiobiologically relevant features of the resulting 4D dose distribution – time being the 4th dimension – is extremely challenging but necessary. This task is different from the early attempt to quantitatively characterize the spatial beam modulation from a delivery point of view as undertaken by Webb in 2003 [1]. Indeed, the degree of absorbed dose modulation – rather than the degree of applied beam modulation – should be taken as the basis for making radiobiological inferences. At least 2 contemporary evolutions further complicate the radiobiological interpretation of various IMRT+ approaches: techniques to irradiate moving targets (gating, freezing, and tracking) and the increasing popularity of hypo-fractionation. STFDDs differ enormously between the various IMRT+. For a homogeneous 2 Gy fraction, the biological effect of some IMRT+ may closely resemble that of the 3D-CRT. For other techniques substantial differences can be expected. Assuming that in vitro observations listed in Table 1 are relevant in the clinic, we hereby try to point to the potential differences in radiobiological effect between popular IMRT+ and 3D-CRT.
201
roughly comparable to 3D-RT regarding the time needed for delivering a 2 Gy fraction; typically less than 10 min. Little or no evidence exists that cell survival is significantly affected by variations of temporal fractionation schemes within a maximum of 10 min overall delivery time. C-IMRT allows creating extremely steep dose gradients. In such cases spatial fractionation by single beams will be very different between compensator IMRT and 3DCRT but gradients will average out by integrating multiple beams. Whether bystander effects across temporally spaced opposed gradients will average out is unknown. Overall, C-IMRT can be expected to provide mostly similar cell survival as 3D-CRT after delivery of a homogeneous 2 Gy fraction. The use of physical compensators went rapidly out of fashion with the advent of computer controlled multileaf collimators (MLC) in the mid-90s. One MLC technique mostly uses a narrow window consisting of all of the leaves that cover the length of the target and sweeps this opening across its volume. The exact size of the opening for each leaf pair is modulated according to the desired dose at each point in the target volume [3]. This sliding window method is often implemented in a continuous dynamic fashion to speed up the dose delivery. Earlier implementations required reduced dose rates of the linear accelerator. Together with the relative inefficiency of sliding window as compared to compensator IMRT, overall delivery times increased by a factor of 2–3 times. Biologically, 2-Gy delivery times exceeding 15–20 min raise questions of temporal fractionation. This would translate to local average dose rates lower than 14 cGy/min for which increased cell survival can be expected. Another popular IMRT approach used superimposed static fields from each beam direction that are sequentially delivered. The beam is halted during the transition from one field to the next when the MLC leaves accomplish the required motion [4]. The delivery is performed as step-and-shoot sequences with beam-off during the leaf motion (step) and beamon (shoot) with fixed collimator aperture. Regarding spatial and temporal fractionation, the same reasoning as for sliding window can be made although the final dose distribution of step-and-shoot IMRT (SS-IMRT) tends to be less smooth. Over the last decade, delivery efficiency has increased. Modern implementations of sliding window IMRT (SW-IMRT) and SS-IMRT are efficient and probably biologically comparable to C-IMRT regarding their temporal component. 1.2. Single-arc techniques and IMAT
1.1. Static beam IMRT techniques IMRT can be delivered in a very robust way by the use of physical compensators that are put in the beam line [2]. IMRT using compensators will require entering the treatment room to change compensators between beams. Entering the room inbetween three-dimensional radiotherapy (3D-RT) beams will also be needed in case of non-coplanar set-up or manually inserted wedges. Overall, multi-beam compensator IMRT (C-IMRT) is
Intensity-modulated arc therapy (IMAT) by means of standard MLC was invented by Yu in the early 90s [5]. Recently, IMATrelated single-arc techniques were developed, their common property being the delivery of a 2-Gy fraction within one or a few minutes, much faster than static beam IMRT techniques and even faster than multi-beam 3D-CRT. Although spatial and temporal modulation occurs at high mean frequency the biological effect of mono-arc techniques may closely resemble that occurring after
Table 1 Current radiation therapy modalities with number of corresponding radiobiological papers found in Web of Science.
CyberKnife Tomotherapy IMAT S-Arc SW-IMRT SS-IMRT C-IMRT 3D-RT cRT
Search keywords phrase in Web of Science
No. of papers
Topic = (CyberKnife and radiobiology) Topic = (tomotherapy) AND Topic = (radiobiology) Topic = (IMAT or VMAT or RapidArc) AND Topic = (radiobiology) Topic = (arc AND radiobiology) + manual inspection Topic = (IMRT and radiobiology and sliding window) Topic = (IMRT and radiobiology) + manual inspection Topic = (IMRT and radiobiology) + manual inspection Topic = (conformal and radiotherapy and radiobiology) NOT Topic = (IMRT) Topic = (radiotherapy and radiobiology) NOT Topic = (IMRT) NOT Topic = (conformal)
3 3 1 0 1 1 0 25 768
202
S. Blockhuys et al. / Mutation Research 704 (2010) 200–205
high dose continuous irradiation. The hypothesis of (slightly) improved anti-tumor efficacy by short delivery times could be investigated.
2. Radiobiological models for the investigation of the effect of STFDD on the biological outcome 2.1. Introduction
1.3. Tomotherapy Tomotherapy techniques employ a radiation source rotating around the patient that – using a fast binary slit collimator – deposits adjacent slices (serial tomotherapy) or a single helical slice (helical tomotherapy) of dose across the target. Within the slice, spatial and temporal modulation is enormous [6]. Each elementary volume inside the target may be hit by hundreds of small dose deposits although within a time-span that is as short as for mono-arc techniques. If we look at each rotation separately, the biological effect could be considered equal to mono-arc techniques. Different from mono-arc techniques is the propagation of a high-gradient slice edge across the target volume over a time period that may exceed 10 min for large targets. To the best of our knowledge, the effect of a propagating gradient has never been modeled in vitro or in vivo.
In radiotherapy, the dose–response curves established for uniform dose distribution are still used to determine the biological response for irradiation with spatially and temporally fractionated dose distributions. The latter comes from the commonly accepted hypothesis that the radiobiological effect depends on the individual cell radiation sensitivity, and is not influenced by, for example, a response of the surrounding cells [8]. Nowadays, the validity of this hypothesis is seriously under discussion in the context of spatially and temporally fractionated irradiation. Indeed, a number of in vitro and in vivo studies show a different biological outcome for irradiation with a spatially [9–13] and/or temporally fractionated dose [14–22] suggesting that cells surrounding the target tissue might affect the radiobiological outcome [23]. Therefore, there is a need for more reliable radiobiological models to accurately determine the biological outcome of spatio-temporal fractionated irradiation.
1.4. CyberKnife 2.2. The radiobiological models The Accuray CyberKnife uses a linear accelerator with circular collimators mounted on a robotic arm. CyberKnife implements a ‘‘nodal’’ delivery method of IMRT. Nodes are small spheres of radiation which are positioned throughout the target volume to create the desired dose distribution. Each of the numerous nodes is irradiated from many directions in space. ‘Walking’ through all the nodes may take more than 1 h for large targets. Irradiating one node almost invariably contributes dose to other nodes. Enormous numbers of small dose deposits to small volumes over long treatment times result in extreme forms of STFDDs. The risk for increased cell survival by very low overall dose rate is substantial for a 2 Gy fraction. However, CyberKnife is typically used to deliver fraction sizes which are much larger than 2 Gy. The contribution of tumor-associated host tissue to therapeutic response may become very important – if not dominant – for large fraction sizes [7]. Techniques to deal with moving targets include gating, freezing and tracking. Gating, freezing as well as the CyberKnife implementation of tracking cause additional interrupts in delivery. These interrupts potentially worsen the temporal component of STFDD with regard to tumor cell survival. When studying the spatio-temporal modulation characteristics of the different radiotherapy modalities, we can conclude the following: (1) more spatial fractionation in ‘Tomotherapy, Cybernife and IMAT’, in comparison with ‘3D-RT and conventional radiotherapy (cRT)’, (2) more temporal fractionation for ‘Cybernife, IMRT and 3D-RT’, in comparison with ‘cRT, single-arc techniques (S-Arc)’, IMAT and Tomotherapy. Further, in Table 1 we have compiled a list containing the total number of radiobiological studies conducted for the different radiation therapy modalities using the Web of Science1 (Thomson-ISI). cRT has the highest number of papers (768) followed by 3D-RT (25), Cybernife (3) and Tomotherapy (3). IMAT, SW-IMRT and SS-IMRT are reported in only one radiobiology paper. Both SArc and C-IMRT are not reported in a radiobiological paper. We can conclude that the recent techniques with more spatio-temporal modulation are not frequently reported in radiobiological papers. So, there is clearly a cumulated time lag between the technology and clinically oriented studies and the radiobiologically oriented work. Consequently, more radiobiological research investigating treatment modalities with spatio-temporal fractionation is urgently required to optimize their therapeutic potential and efficiency.
We compiled in two complementary tables the published radiobiological models that have been developed and used so far to investigate the effect of spatially and temporally fractionated irradiation on the radiobiological response (see Tables 2A and 2B). Table 2A, namely ‘the radiophysical table’, contains the parameters radiation beam and dose characteristics, fractionation (spatial and/ or temporal modulation) and dosimeter. Table 2B, namely ‘the radiobiological table’, contains the parameters cell line, assay (endpoint) and results. 2.2.1. The radiophysical table 2.2.1.1. Experimental set-up (spatial and/or temporal fractionation). For the evaluation of the spatially fractionated dose distribution, two model systems have been used. First, Mackonis et al. [13] and Bijl et al. [9] investigated the effect of segmented dose distributions, namely the partial fields ‘quarter’ and ‘stripped’ field, and the radiation set-ups ‘split-field’, ‘symmetrical’- and ‘asymmetrical bath and shower’ respectively, whereas others applied a linear dose gradient [10,11,24]. For example, we developed 3 different in vitro set-ups [24]. According to our Setup 1, a cell monolayer is irradiated with 3 small radiation fields creating steep dose gradients at the border of the fields. Set-up 1 is closely related with the in vivo situation where tumors are irradiated to spatially and temporally fractionated high doses, resulting in steep dose gradients within and at the border of the irradiated volumes. According to Set-ups 2 and 3, the cell monolayers in a T25 flask and a 96-well plate, respectively, are irradiated with a linear smooth dose gradient along the length of the recipient. To evaluate the effect of the temporal aspect of IMRT+, the following parameters have been examined: (1) total dose delivery time [12,14,19,21], (2) intra-fraction dose rate [20], (3) dose per fraction [17], (4) inter-fraction time [15,18,22], and (5) fraction number [16]. 2.2.1.2. Dosimetry. Since the calculation and planning methods applied in clinic for these kinds of experimental set-ups result in less accurate dose measurements, dosimetry is strongly recommended [11,13,24]. For our set-ups, we performed dosimetry with the radiochromic EBT film. This dosimeter allows dose measurement in cell geometry, for Set-ups 1 and 2, more precisely by
S. Blockhuys et al. / Mutation Research 704 (2010) 200–205
203
Table 2A Radiobiological models developed to investigate the biological effect of spatio-temporal modulated irradiation: the radiophysical table.
Chronological presentation of the radiobiological models developed so far to investigate the effects of irradiation with spatio-temporal fractionated dose distributions on the biological outcome. Parameters: radiation beam and dose characteristics, fractionation type (spatial and/or temporal modulation) and dosimeter.
replacement of the cell monolayer by stack of film and polystyrene sheets. For Model 3, a fitting piece of film was positioned immediately under the bottom surface of the 96-well plate [25]. The use of film dosimetry with EBT radiochromic film has different advantages. (1) Film versus ionization chamber: film allows the determination of the dose distribution over a certain area, in contrast to the measurement at a single point with the ionization chamber [11,12,15,16,20,21,26]. (2) Radiochromic [13,24] versus radiographic film [11,13]: the radiochromic film is relatively light insensitive in comparison to the radiographic film, and is more suitable to cut into the desired form. Consequently, radiochromic film dosimetry can be done in the cell geometry. Additionally, radiochromic film is nearly water-equivalent and features a flat energy response for high-energy X-rays. Dosimetry has also been done in the cell geometry albeit with the parallel plate ionization chamber [9,10] and with the Fricke dosimeter [14].
the MTT assay and conclude that the MTT assay is a relevant assay to investigate the effect of spatial fractionated irradiation. The MTT assay is a common tool in radiobiology to assess radiosensitivity. It is a sensitive, quantitative and reliable assay that measures the conversion of the soluble yellow MTT substrate into insoluble dark blue formazan salt by cellular dehydrogenases dependent on the metabolic activity [27]. Consequently, the formazan optical density (OD) is equivalent with the metabolic activity, which is determined by cells proliferation and/or cell metabolic activity. In contrast to the classical protocol, as described by Bromley et al. [11], whereby in the final step formazan crystals are dissolved in DMSO, we applied an adapted MTT protocol, whereby the final solubilization step is not necessary for quantification [24]. Indeed, DMSO solubilization erases all in situ information about the spatial distribution of cell growth. 3. Results
2.2.2. The radiobiological table 2.2.2.1. Assays and endpoints. So far, different biological assays, namely crystal violet (CV) staining, clonogenic assay, micronuclei (MN) visualisation, g-H2AX staining, cell cycle analysis, and 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT) assay, have been applied to determine the biological effect of spatio-temporal fractionated irradiation. These assays are used to investigate different endpoints, namely cell survival, cell growth, cell proliferation, cell cycle inhibition, DNA damage and metabolic activity. Until now, only the CV staining and the clonogenic assay have been used to investigate the effect of spatial fractionated dose distribution [10–13]. In contrast to previous models, we applied
From the radiobiological results with spatially and/or temporally fractionated dose, we may conclude that there is an additional effect on the radiobiological outcome as predicted from the current treatment planning systems. For spatially fractionated dose distribution, the results from in vitro experiments indicate that a bystander effect, i.e. biological effect due to intercellular communication in between cells irradiated with different radiation doses, may be responsible for an additional effect [9–13,24]. A recent letter by Ross and Klassen [28] recently commented on the research-paper of Mackonis et al. [13], more precisely on their determination of two new types of radiation-induced bystander effects (II and III), which would not be supported by their presented data. Additionally, radiation-induced
204
S. Blockhuys et al. / Mutation Research 704 (2010) 200–205
Table 2B Radiobiological models developed to investigate the biological effect of spatio-temporal modulated irradiation: the radiobiological table.
Chronological presentation of the radiobiological models developed so far to investigate the effects of irradiation with spatio-temporal fractionated dose distributions on the biological outcome. This table reports cell line, assay (endpoint) and results. Note that the in vivo studies are indicated with an asterisk added to the respective study number, and the studies describing spatial modulation are marked in gray.
bystander effects are not always detected. Groesser et al. [29] published that possibly the epigenetic status of the specific cell line or the precise culture conditions and medium supplements, such as serum, may be critical for inducing bystander effects. Our results suggest that the MTT assay can also be used to reveal bystander effects [24]. We observed an increased metabolic activity for intermediate doses with the MTT assay for irradiation according to Set-ups 1–3. In contrast, no enhanced colour intensity was observed at the intermediate dose region when using CV staining, indicating that the increased metabolic activity is not due to an increased cell number but may be due to an increased cell metabolic activity. Further, we observed differences in peak expression time and related intermediate dose region between the different models. Since Set-ups 2 and 3 only differ in the possibility of intercellular communication, the latter may be due to bystander effects. Results from in vitro models for temporally fractionated dose distribution indicate that: (1) the irradiation time must be kept as short as possible, since an increase in total delivery time for a fixed total dose results in an increase of cell survival, which is at least partially due to sub-lethal damage repair (SLDR). Hereby, multiple parameters can result in an increased radiation-delivery time, namely dose rate [14,20], fraction number [16,19,26] and beaminterruption time [15,19]. (2) Some radiation doses can lead to hypo- and hyper-sensitive biological responses. For example when a 2 Gy dose is divided into multiple partial fractions, fractions less
than 0.5 Gy (S) can lead to hyper-sensitive and those higher that 0.5 or 1 Gy (L) can lead to hypo-sensitive biological response. Consequently, the sequence (L–S or S–L) in which the fractions are delivered must be taken into account [17]. In parallel with the in vitro experiments, the in vivo experiments suggest an additional effect of intercellular communication and of SLDR delay for spatially [9] and temporally fractioned irradiation [18,22], respectively. In contrast to the in vitro experiments for temporal fractionation, the in vivo studies suggest that the increase in survival due to extended delivery time for the different in vitro studies may be diminished due to decrease of SLDR when reoxygenation takes place. However, since the magnitude and velocity of reoxygenation of human tumors are unpredictable, the clinical effect of intermittent irradiation cannot be derived from the radiobiological outcome for continuous irradiation. These few in vivo models suggest that a short radiation-time is required considering the possibility of slow or insufficient reoxygenation of the tumor. 4. Discussion In general, we can conclude from the above-presented results, that spatially and temporally fractionated irradiation can result in additional effects on the biological outcome, namely multiple effects of intercellular communication (bystander effects) and of radiation dose delivery time (enhanced SLDR as a result of
S. Blockhuys et al. / Mutation Research 704 (2010) 200–205
increased total dose delivery time), respectively. These additional effects have to be taken into account to result in a good clinical prediction. We suggest an optimization of existing in vitro models. Therefore, following aspects should be taken into account: (1) for dosimetry, dosimeters have to be employed which allow the determination of the dose distribution in the cell geometry. This includes the replacement of the cells by a tissue-equivalent material in the in vitro set-up, in order to assess the spatial features of the dose distribution and avoid the inaccuracies in dose measurement due to a different set-up. (2) For radiobiology, more reliable experimental set-ups should be designed to approach the in vivo (radiotherapeutic) setting. Therefore, special attention should be paid to radiation dose, STFDD and clinical equipment. (3) In order to take into account the additional effects in clinical settings, it is important to assess multiple endpoints. Apart from cell viability also other endpoints should be considered, like cellular motility and invasion [30]. Therefore, it is interesting to apply for example the MTT assay, which measures metabolic activity and might give the opportunity to detect cells with enhanced motility and invasive capacities. Additionally, for temporal fractionated irradiation cell reoxygenation should be investigated since SLDR may be counterbalanced by reoxygenation. 5. Conclusion From a clinical point of view, we can conclude that radiation techniques with spatio-temporal dose modulation have become popular in radiotherapy. One of the most important advantage includes the delivery of a highly controlled dose distribution conformal to the tumor creating the possibility to raise the dose to the tumor with less toxicity to the surrounding healthy tissues. However, radiobiological research has shown that the biological outcome of spatio-temporal fractionated irradiation cannot be fully explained using the existing dose–response data. Consequently, further research is needed to establish more reliable radiobiological models to predict the clinical outcome of spatiotemporally fractionated radiation therapy. Conflict of interest The authors declare that there are no conflicts of interest. References [1] S. Webb, Use of a quantitative index of beam modulation to characterize dose conformality: illustration by a comparison of full beamlet IMRT, few-segment IMRT (fsIMRT) and conformal unmodulated radiotherapy, Phys. Med. Biol. 48 (2003) 2051–2062. [2] G.W. Sherouse, In regard to Intensity-Modulated Radiotherapy Collaborative Working Group, IJROBP 51 (2001) 880–914; G.W. Sherouse, Int. J. Radiat. Oncol. Biol. Phys. 53 (2002) 1088–1089. [3] D.J. Convery, M.E. Rosenbloom, The generation of intensity modulated fields for comformal radiotherapy by dynamic collimator, Phys. Med. Biol. 37 (1992) 1359– 1374. [4] J.M. Galvin, X.G. Chen, R.M. Smith, Combining multileaf fields to modulate fluence distributions, Int. J. Rad. Oncol. Biol. Phys. 27 (1993) 697–705.
205
[5] C.X. Yu, Intensity-modulated arc therapy with dynamic multileaf collimation: an alternative to tomotherapy, Phys. Med. Biol. 40 (1995) 1435–1449. [6] R.T. Flynn, M.W. Kissick, M.P. Mehta, G.H. Olivera, R. Jeraj, T.R. Mackie, The impact of linac output variations on dose distributions in helical tomotherapy, Phys. Med. Biol. 53 (2008) 417–430. [7] Z. Fuks, R. Kolesnick, Engaging the vascular component of the tumor response, Cancer Cell 8 (2005) 89–91. [8] E.J. Hall, Radiobiology for the Radiologist, Lippincott Williams and Wilkins, Philadelphia, PA, 2000. [9] H.P. Bijl, P. van Luijk, R.P. Coppes, J.M. Schippers, A.W.T. Konings, A.J. van der Kogel, Unexpected changes of rat cervical spinal cord tolerance caused by inhomogeneous dose distributions, Int. J. Radiat. Oncol. Biol. Phys. 57 (2003) 274–281. [10] N. Suchowerska, M. Ebert, M. Zhang, M. Jackson, In vitro response of tumour cells to non-uniform irradiation, Phys. Med. Biol. 50 (2005) 3041–3051. [11] R. Bromley, R. Davey, L. Oliver, R. Harvie, C. Baldock, A preliminary investigation of cell growth after irradiation using a modulated X-ray intensity pattern, Phys. Med. Biol. 51 (2006) 3639–3651. [12] V. Moiseenko, C. Duzenli, R.E. Durand, In vitro study of cell survival following dynamic MLC intensity-modulated radiation therapy dose delivery, Med. Phys. 34 (2007) 1514–1520. [13] E. Mackonis, N. Suchowerska, M. Zhang, M. Ebert, D. McKenzie, M. Jackson, Cellular response to modulated radiation fields, Phys. Med. Biol. 52 (2007) 5469–5482. [14] X. Mu, P. Lo¨froth, M. Karlsson, B. Zackrisson, The effect of fraction time in intensity modulated radiotherapy: theoretical and experimental evaluation of an optimisation problem, Radiother. Oncol. 68 (2003) 181–187. [15] Y. Shibamoto, M. Ito, C. Sugie, H. Ogino, M. Hara, Recovery from sublethal damage during intermittent exposures in cultured tumor cells: implications for dose modification in radiosurgery and IMRT, Int. J. Radiat. Oncol. Biol. Phys. 59 (2004) 1484–1490. [16] H. Ogino, Y. Shibamoto, C. Sugie, M. Ito, Biological effects of intermittent radiation in cultured tumor cells: influence of fraction number and dose per fraction, J. Radiat. Res. 46 (2005) 401–406. [17] P. Lin, A. Wu, Not all 2 Gray radiation prescriptions are equivalent: cytotoxic effect depends on delivery sequences of partial fractionated doses, Int. J. Radiat. Oncol. Biol. Phys. 63 (2005) 536–544. [18] C. Sugie, Y. Shibamoto, M. Ito, H. Ogino, A. Mitamoto, N. Fukaya, H. Niimi, T. Hashizume, Radiobiologic effect of intermittent radiation exposure in murine tumors, Int. J. Radiat. Oncol. Biol. Phys. 64 (2006) 619–624. [19] V. Moiseenko, J. Banath, C. Duzenli, P. Olive, Effect of prolongation radiation delivery time on retention of gammaH2AX, Radiat. Oncol. 3 (2008). [20] J. Bewes, N. Suchowerska, M. Jackson, M. Zhang, D. McKenzie, The radiobiological effect of intra-fraction dose-rate modulation in intensity modulated radiation therapy (IMRT), Phys. Med. Biol. 53 (2008) 3567–3578. [21] P. Keall, M. Chang, S. Benedict, H. Thames, S. Vedam, P. Lin, Investigating the temporal effects of respiratory-gated and intensity-modulated radiotherapy treatment delivery on in vitro survival: an experimental and theoretical study, Int. J. Radiat. Oncol. Biol. Phys. 71 (2008) 1547–1552. [22] N. Tomita, Y. Shibamoto, M. Ito, H. Ogino, C. Sugie, S. Ayakawa, H. Iwata, Biological effect of intermittent radiation exposure in vivo: recovery from sublethal damage versus reoxygenation, Radiother. Oncol. 86 (2008) 369–374. [23] W.F. Morgan, M.B. Sowa, Non-targeted bystander effects induced by ionizing radiation, Mutat. Res. 616 (2007) 159–164. [24] S. Blockhuys, B. Vanhoecke, L. Paelinck, M. Bracke, C. De Wagter, Development of in vitro models for investigating spatially fractionated irradiation: physics and biological results, Phys. Med. Biol. 54 (2009) 1565–1578. [25] N. Tomic, M. Gosselin, J.F. Wan, U. Saragovi, E.B. Podgorsak, M. Evans, S. Devic, Verification of cell irradiation, Phys. Med. Biol. 50 (2007) 3121–3131. [26] F. Sterzing, M. Mu¨nter, M. Scha¨fer, P. Haering, B. Rhein, C. Thilmann, J. Debus, Radiobiological investigation of dose-rate effects in intensity-modulated radiation therapy, Strahlenther. Onkol. 181 (2005) 42–48. [27] T. Slater, Studies on a succinate-neotetrazolium reductase system of rat liver: II. Points of coupling with the respiratory chain, Biochem. Biophys. Acta 77 (1963) 365–382. [28] C.K. Ross, N.V. Klassen, Comments on ‘Cellular response to modulated radiation fields’, Phys. Med. Biol. 54 (2009) L11–L13. [29] T. Groesser, B. Cooper, B. Rydberg, Lack of bystander effects from high-LET radiation for early cytogenic end points, Radiat. Res. 170 (2008) 794–802. [30] I. Madani, W. De Neve, M. Mareel, Does ionizing radiation stimulate cancer invasion and metastasis, Bull. Cancer 95 (2008) 292–300.