Heavy-ion tumor therapy: Physical and radiobiological benefits

High-energy beams of charged nuclear particles (protons and heavier ions) offer significant advantages for the treatment of deep-seated local tumors in comparison to conventional megavolt photon therapy. Their physical depth-dose distribution in tissue is characterized by a small entrance dose and a distinct maximum (Bragg peak) near the end of range with a sharp fall-off at the distal edge. Taking full advantage of the well-defined range and the small lateral beam spread, modern scanning beam systems allow delivery of the dose with millimeter precision. In addition, projectiles heavier than protons such as carbon ions exhibit an enhanced biological effectiveness in the Bragg peak region caused by the dense ionization of individual particle tracks resulting in reduced cellular repair. This makes them particularly attractive for the treatment of radio-resistant tumors localized near organs at risk. While tumor therapy with protons is a well-established treatment modality with more than 60 000 patients treated worldwide, the application of heavy ions is so far restricted to a few facilities only. Nevertheless, results of clinical phase I-II trials provide evidence that carbon-ion radiotherapy might be beneficial in several tumor entities. This article reviews the progress in heavy-ion therapy, including physical and technical developments, radiobiological studies and models, as well as radiooncological studies. As a result of the promising clinical results obtained with carbon-ion beams in the past ten years at the Heavy Ion Medical Accelerator facility (Japan) and in a pilot project at GSI Darmstadt (Germany), the plans for new clinical centers for heavy-ion or combined proton and heavy-ion therapy have recently received a substantial boost.

Figure 1

(Color online) Depth-dose profiles of ${}^{60}\mathrm{Co}$ $\gamma$ radiation, megavolt photons, and ${}^{12}\mathrm{C}$ ions in water.

Figure 2

(Color online) Specific energy loss $dE∕dx$ of ${}^{12}\mathrm{C}$ ions and protons in water. The range of ${}^{12}\mathrm{C}$ ions in water corresponding to their specific energy is indicated at the top.

Figure 3

(Color online) Mean range of heavy ions in water.

Figure 4

(Color online) Measured Bragg peaks of protons and ${}^{12}\mathrm{C}$ ions having the same mean range in water (251).

Figure 5

(Color online) Calculated beam spread for ${}^{12}\mathrm{C}$ ions and protons in a typical treatment beam line. It was assumed that an intially parallel particle beam ($5\mathrm{mm}$ full width at half maximum) passes through the nozzle (including a thin vacuum window and beam monitors) and enters a water absorber (patient) at $1\mathrm{m}$ distance from nozzle exit. At small depth (i.e., small particle energies) the width is mainly determined by scattering in the nozzle, while at higher energies the scattering in the water absorber dominates. Carbon ions show a much smaller spread than protons at the same penetration depth. Figure courtesy of U. Weber, Rhön-Klinikum AG.

Figure 6

(Color online) Illustration of the abrasion-ablation model of peripheral collisions at high energies according to 264. Adapted from 97.

Figure 7

(Color online) Buildup of secondary fragments produced by $400\mathrm{MeV}∕\mathrm{u}$ ${}^{12}\mathrm{C}$ ions stopping in water. From 101.

Figure 8

(Color online) Measured Bragg curves of ${}^{12}\mathrm{C}$ ions stopping in water. From 251.

Figure 9

(Color online) Bragg curve for $670\mathrm{MeV}∕\mathrm{u}$ ${}^{20}\mathrm{Ne}$ ions in water measured at GSI (circles) and calculated contributions of primary ions, secondary and tertiary fragments. Adapted from 272.

Figure 10

(Color online) Sketch of a fully passive beam shaping system. The initially narrow beam is broadened by a scattering system and adapted to the target volume by various passive beam shaping devices. Adaption of the dose field to the distal contour of the target volume is achieved by a compensator, but results in unwanted normal-tissue dose in the proximal part (indicated by the doubly hatched area). Figure courtesy of U. Weber, Rhön-Klinikum AG.

Figure 11

Stacking of subvolumes using a dynamic range shifter and collimator. Adapted from 57.

Figure 12

Layout of the $200\mathrm{MeV}$ proton spot-scanning system at PSI. From 228.

Figure 13

(Color online) Principle of the intensity-controlled magnetic scanning system at GSI. Left: The target volume is irradiated by moving a pencil-like ion beam ($80–430\mathrm{MeV}∕\mathrm{u}$ ${}^{12}\mathrm{C}$) with fast magnets over thin slices in depth. The required beam parameters are supplied on a pulse-to-pulse operation by the synchrotron (SIS) control system. From 99. Right: Beam’s-eye view of slices for a typical patient treatment plan. In each panel one slice is shown. The actually irradiated slice is seen in the magnified panel with the raster point positions indicated as open circles. The superimposed dots show the beam center positions measured online by a multiwire chamber. The spot size of the beam is larger than the circles and overlaps many positions.

Figure 14

(Color online) Common gantry design (left) and cork-screw gantry (right) for passive proton therapy systems. Adapted from 227.

Figure 15

(Color online) Sectional view of the heavy-ion gantry at HIT Heidelberg. Figure courtesy of MT Mechatronics GmbH (Mainz, Germany).

Figure 16

(Color online) Illustration of the spatial shift of a target volume in the thorax caused by breathing. The dark and white contour lines indicate the target positions at the time of maximum inhalation and exhalation, respectively. Figure courtesy of E. Rietzel and C. Bert.

Figure 17

(Color online) Principle of respiratory-gated irradiation. Adapted from 199.

Figure 18

(Color online) Film exposures for stationary, moving (uncompensated), and motion-compensated (tracking) irradiations. The corresponding horizontal dose profiles are shown in the lower right panel. Adapted from 25.

Figure 19

(Color online) Biologically effective dose distribution optimized with the treatment planning system TRiP (162) for a skull base tumor treated at GSI Darmstadt. With three fields an excellent sparing of critical organs (brain stem and optical nerves) is achieved. Figure courtesy of O. Jäkel.

Figure 20

(Color online) Principle of in situ range verification by autoactivation and PET imaging. A small fraction $(<1\mathrm{\%})$ of the incident ${}^{12}\mathrm{C}$ ions is transformed into radioactive ${}^{11}\mathrm{C}$ in high-energy fragmentation reactions. Their stopping depth is nearly the same as for the primary ions and can be monitored with a PET camera.

Figure 21

Autoactivation of thick PMMA targets by beams of $260\mathrm{MeV}∕\mathrm{u}$ carbon ions (top) and $140\mathrm{MeV}$ protons (bottom). The solid line shows the depth profile of the measured ${\beta }^{+}$-activity. For comparison the depth-dose profile of the primary beam is shown by a dotted line. From 217.

Figure 22

(Color online) Example of in-beam PET monitoring showing the irradiation of a skull base tumor at GSI. Left: Planned dose distribution superimposed onto the CT image. The target volume and the brain stem as an organ at risk are highlighted. Middle: Predicted ${\beta }^{+}$-activity distribution calculated from the treatment plan and time course of the irradiation. Right: Measured ${\beta }^{+}$-activity distribution. By comparison with the prediction it was verified that the carbon ions were stopped before the brain stem. Adapted from 73.

Figure 23

(Color online) Track structure of a $1\mathrm{MeV}∕\mathrm{u}$ carbon ion simulated by the Monte Carlo code TRAX. Lines represent the paths of individual secondary electrons. Figure courtesy of Michael Krämer.

Figure 24

(Color online) Illustration of the different microscopic dose distribution by x rays and carbon ions with different specific energies. The average macroscopic dose is $2\mathrm{Gy}$ in all cases. From 259.

Figure 25

(Color online) Determination of RBE for cell inactivation for 10% and 1% survival level for a typical heavy ion irradiation.

Figure 26

Dependence of ${\mathrm{RBE}}_{\alpha }={\alpha }_{\text{ion}}∕\alpha$ on LET and particle type, where $\alpha$ and ${\alpha }_{\text{ion}}$ are the linear part of the survival curve for photons and ions, respectively. Data are redrawn from 19 and 86. From 259.

Figure 27

(Color online) RBE dependence on the radiation sensitivity of cell lines. Left: Dependence of initial ${\mathrm{RBE}}_{\alpha }$ on the three cell lines CHO, V79, and XRS after carbon irradiation. Right: Survival curves of the same cell lines demonstrate the different radiosensitivities after x ray irradiation. Experimental data are from 301.

Figure 28

(Color online) RBE values based on the dose for 50% complication probability for the radiation tolerance of the rat spinal cord after 1, 2, 6, and 18 fractions of carbon ions. Data for the spread-out Bragg peak $(125\mathrm{keV}∕\mu \mathrm{m})$ and the plateau $(13\mathrm{keV}∕\mu \mathrm{m})$ are adapted from 140.

Figure 29

(Color online) Influence of the oxygen level on cell survival of human kidney T-1 cells for carbon ions with different LET. Lines are based on experimental data by 34.

Figure 30

Distribution of the photon-equivalent dose along a $60\mathrm{mm}$ spread-out Bragg peak used at HIMAC. Experimental data for several different cell lines are depicted. From 134.

Figure 31

LET dependence for RBE of 10% survival for HSG and Hela cells after irradiation by $290\mathrm{MeV}∕\mathrm{u}$ carbon ions using a ridge filter to extend the Bragg peak. The dashed line depicts the corresponding value of the NIRS neutron beam for HSG cells. From 134.

Figure 32

Schematics to demonstrate the determination of the absolute clinical dose for carbon-ion treatments. The RBE of 3.0 belongs to the neutron-equivalent beam, whereas the clinical RBE is given in the SOBP center. From 134.

Figure 33

Schematics of the local effect model. From 69.

Figure 34

(Color online) Validation of the treatment planning system with CHO cells in a therapylike scenario with an organ at risk close to the tumor volume (represented by the H-shaped structure). Upper left: Measured survival data. Upper right: Calculated survival. Lower row: Survival along cuts through the irradiated volume. From 165.

Figure 35

(Color online) Skin reaction after irradiation with x rays and carbon ions. The scoring level indicates the severity of the skin reaction [for details see 309]. Using the predictions of LEM, the doses for carbon-ion irradiation were adjust to result in the same biological effect as the photon fields applied simultaneously. From 260; adapted from 309.

Figure 36

(Color online) Axial MRI scan of a skull base chordoma prior to RT (left) and tumor regression $6\text{weeks}$ after irradiation (right).