Radiobiology with heavy charged particles: a historical review.

Radiobiological studies using heavy charged particles followed closely the development of accelerators to produce beams of ever-increasing energy, driven primarily by the aspirations of physicists and chemists interested in the structure of matter. An impressive share of this development took place at Berkeley, beginning with the invention of the cyclotron by Ernest Lawrence in 1930. There followed a series of cyclotrons, synchrotrons and linear accelerators, culminating in the BEVALAC, which provided the first source of very heavy ions (helium to argon) to be used clinically, beginning in 1975. Other early entrants (1950's-1960's) in the clinical use of heavy ion beams (protons only) included Uppsala, Harvard/MGH and several facilities in the USSR. During the 1970's negative pi-meson (pion) beams for clinical use were developed in the US (LAMPF), Switzerland (SIN/PSI) and Canada (TRIUMF). Although the first accelerator built primarily for medical use, the Crocker Medical Cyclotron, was completed at Berkeley in 1939 (it was used primarily to produce neutron beams) it was not until 1990 that the next clearly dedicated medical heavy ion facility went into operation: the 3-gantry proton synchrotron at Loma Linda. There are several reasons for this long hiatus: the long time required to complete clinical trials; the need to develop more economic and flexible accelerators and beam handling systems; the early discouraging clinical results obtained with neutron beams at Berkeley in the 1940's, before the dose response differences for early and late effects were fully understood. During the last decade or so there has been a rapid increase in the number of proton beam facilities; heavier ion beams are so far available only at HIMAC in Japan and GSI in Germany. Earlier studies with radioactive alpha-particle sources and plant cells had already shown, by the early 1930's that high LET radiations were biologically more effective than X-rays in producing damage in eukaryotes. The increased penetration of high energy particles from accelerators made it possible to carry out in vivo radiobiological studies in animals, and the publication by Puck of the first radiation survival response for cultured mammalian cells in 1956, provided another valuable tool for radiobiological studies. One of the earliest systematic studies of the dependence of RBE (relative biological effectiveness) and OER (oxygen enhancement ratio) on LET (linear energy transfer) was that by Barendsen in the early 1960's; he irradiated cultured human kidney cells with deuterium and alpha-particles, and showed that RBE reached a maximum at an LET of 100-200 keV/micrometer, the same LET at which the OER decreased to approximately 1.0. More recent studies (Belli, Folkard, etc.) show that the RBE 'peaks' at a LET which is particle-dependent (for protons, RBE maximum is at approximately 30 keV/micrometer), indicating that LET alone does not adequately define the microscopic energy deposition and its influence on biological effect. One of the complications with heavy ion and pion beams is the increase in RBE with depth in the stopping region. Cultured cell techniques were developed to accurately map these RBE changes, which were investigated at each of the heavy ion and pion facilities, allowing physical dose profiles to be shaped to compensate for the change in biological effectiveness. With the heavier ions, RBE is also dependent on dose and on the dose fractionation scheme used. In vivo systems are the most suitable for such measurements and a variety of normal tissue and tumour end-points has been employed for such studies. A review of the published RBE values for proton beams, 1975-1997, shows very good consistency between the various centres, with average in vivo and average in vitro values falling in the range 1.11-1.18. In this article we have, due to space limitations, only been able to review a representative fraction of the extensive literature on heavy ion radiobiology. We have arbitrarily limited our discussion to mammalian systems, except for a few very early experiments of historical interest.