Medical applications

penetration depths and decrease for R&D investment, and this should skin exposure. Alternatively, electro­ lead to smaller hospital space re­ Therapy therapy can be used with different quirements, lower operating costs, energies for lower and variable and elimination of external data s life expectancy steadily im­ penetration depths - approximately handling, resulting in simpler and A proves, the incidence of cancer 0.5 cm per MeV. In this way surface more cost effective clinical proce­ also increases as the population tissue may be treated without affect­ dures. ages. Up to 50% of cancer patients ing deeper and more critical anatomi­ undergo radiotherapy - a form of cal regions. From D Tronc, General Electric protracted biological surgery by This type of linac, 1 to 2 metres Medical Systems, Paris selective sterilization of malignant long, is mounted parallel to the cells. patient with a bending magnet to direct the beam to the radiotherapy system, which includes the target, Neutron therapy X-ray radiotherapy thick movable collimator jaws, a beam field equalizer, dose rate and tandard radiotherapy uses X- he most common form of radio optical field simulation and energy S rays or electrons which have low T therapy is X-ray therapy, where a controls. LET (linear energy transfer); in beam of photons or their parent There are over 2000 accelerator- contrast, particles such as neutrons electrons break down hydrogen based X-ray treatment units world­ with high LET have different bonds within the body's cells and wide. Western countries have up to radiobiological responses. In the late remove certain DNA information two units per million population, 1960s, clinical trials by Mary Catterall necessary for cell multiplication. whereas in developing countries at the Hammersmith Hospital in This process can eradicate malignant such as Bangladesh, the density is London indicated that fast neutron cells leading to complete recovery, to only one per 100 million. radiation had clinical advantages for the remission of some cancers, or at Several major medical equipment certain malignant tumours. least to a degree of pain relief. companies manufacture X-ray Following these early clinical trials, The radiotherapy instrument is therapy systems - General Electric, several cyclotron facilities were built usually an electron linac, and the Mitsubishi, Philips, Siemens and in the 1980s for fast neutron therapy, electrons are used either directly in Varian. In this crowded marketplace for example at the University of 'electrotherapy' for some 10% of where the useful lifespan of ma­ Washington, Seattle, and at UCLA. patients, or the electrons bombard a chines exceeds 10 years, purchase Most of these newer machines use conversion target creating a broad prices are less than $1 million per extracted cyclotron proton beams in beam of high energy photons or unit. the range 42 to 66 MeV with beam 'penetration X-rays'. X-ray therapy remains the most intensities of 15 to 60 microamps. The simplest machine consists of common and cheapest form of The proton beams are transported to several accelerating sections at accelerator therapy. Ongoing techni­ dedicated therapy rooms, where around 3 GHz, accelerating electrons cal developments aim to achieve neutrons are produced from beryllium to 6 MeV; a cooled tungsten target is better matching of dose delivery to targets. used to produce a 4 Gray/min X-ray tumour volume; multileaf collimators Second-generation clinical trials field which can be collimated into a shape the X-ray field to the biomedi­ showed that accurate neutron beam rectangular shape at the patient cal target, and portal imaging from delivery to the tumour site is more position. This tiny linac is mounted behind the patient can control posi­ critical than for photon therapy. inside a rotating isocentric gantry tioning and dose delivery. In order to achieve precise beam above the patient who must remain Combined compact X-ray sources geometries, the extracted proton perfectly still. Several convergent are being developed with both beams have to be transported beams can also be used to increase treatment and realtime dosimetry through a gantry which can rotate the delivered dose. control, incorporating CT scanning around the patient and deliver beams More sophisticated accelerators into one single device. Integrated from any angle; also the neutron operate at up to 18 MeV to increase diagnosis and therapy is the direction beam outline ("field shape") must be

CERN Courier, July/August 1995 7 Medical applications

In the late 1960s, clinical trials showed that fast neutron radiation had clinical advantages for the treatment of certain cancers. In modern neutron therapy centres, protons from a cyclotron bombard a beryllium target to generate a collimated beam of neutrons. (Photo University of Washington, Seattle)

with other users, particularly since beam-time for an actual therapy run is only a few minutes but involves considerable set-up periods between runs; radioisotope or PET production may be performed concurrently with neutron therapy. At present there are about a dozen neutron therapy facilities worldwide, but due to limitations of fixed beams, low energies or low intensities, only a few of these can deliver the precision beams required . Clinical results with neutron therapy are very encouraging for certain specific tumour sites; for salivary gland tumours neutron therapy is considered the treatment of choice. Other promising areas of clinical research are advanced adeno­ carcinoma of the prostrate, some adjusted to extremely irregular scanners and radiation simulators head and neck tumours, some lung shapes using a flexible collimation must be made available. Neutron cancers and sarcomas of bone and system. therapy is usually performed only in soft tissue. A therapy procedure has to be radiation oncology departments of For future neutron therapy systems appropriately organized, with physi­ major medical centres. there is a choice between two basic cians, radiotherapists, nurses, However neutron therapy is much designs: medical physicists and other staff in more expensive than conventional - a small cyclotron with internal attendance; other specialized photon or electron therapy. Often the target, mounted directly on the equipment, such as CT or MRI cyclotron operation has to be shared therapy gantry in a very compact machine dedicated to neutron therapy (such as that developed for Harper Hospital in Detroit using a superconducting deuteron cyclotron at 48.5 MeV); or - a cyclotron in a separate vault with a beamline feeding one or more therapy rooms with rotating isocentric gantries. This type of facility has been built commercially, e.g. a 60 MeV system by Scanditronix. Like X-ray therapy, future develop­ ments in this field depend on reduc­ ing the cost of therapy systems and providing total systems suitable for hospital environments with limited technical and engineering resources.

Layout of a neutron therapy centre

CERN Courier, July/August 1995

Medical applications

To be clinically effective, energies of several ion cyclotron for cancer therapy. This fixed hundred MeV are required for proton therapy. energy isochronous cyclotron's magnet system Pioneering projects had to work with complex, is optimized for high magnetic field but is still inadequate equipment originally intended for small enough to be installed in a hospital; it nuclear physics research, but recently a can deliver beams of up to 1.5 microamps for number of specialist organizations and treating certain categories of tumours. commercial companies have been working on dedicated systems for proton therapy. (From IBA, Louvain-la-Neuve, Belgium) This is an artist's view of a 235 MeV negative

As with other therapy methods, the accelerator is only one component, and it is important that manufacturers are able to offer integrated medical service systems

From Rudi Riesler, University of Washington, Seattle, USA

Proton therapy

deal radiotherapy deposits a large amount of energy in the tumour volume, and none in the surrounding healthy tissues. Proton therapy comes closer to this goal because of a greater concentration of dose, well defined proton ranges and points of energy release which are precisely known - the Bragg peak1. In the past, the development of deliver beams of up to 1.5 microamps, cyclotron, the beamline and the clinical proton therapy has been but with a hardware limitation to restrict gantry optics; hampered by complexity, size, and the maximum possible dose; - a safety management system, cost. To be clinically effective, ener­ - variable energy beam (235 to independent of the control system, gies of several hundred MeV are 70 MeV ) with energy spread and using hardwired interlocks and required; these were previously emittance verification; independent programmable logic unavailable for hospital installations, - a beam transport and switching controllers; and pioneering institutions had to system to connect the exit of the - a robotic patient positioning work with complex, inadequate energy selection system to the system, with monitoring equipment equipment originally intended for entrances of a number of gantries completely surrounding the patient. nuclear physics research. and fixed beamlines. Along the A few companies have proposed Recently a number of specialist beam transport system, the beam other systems, which may differ in organizations and commercial compa­ characteristics are monitored with concept designs for the gantries, nies have been working on dedicated non-interceptive multiwire ionization nozzles, patient positioners or safety systems for proton therapy. One, IBA chambers for automatic tuning; and control systems. The proton of Belgium, has equipment for in- - gantries fitted with nozzles and therapy system at the Loma Linda house hospital operation which beamline elements for beam control; University Medical Centre, California, encompasses a complete therapy both beam scattering and beam is based on a proton synchrotron and centre, delivered as a turnkey pack­ wobbling techniques are available for was built by the Fermi National age and incorporating a compact, shaping the beam; Accelerator Laboratory, the Loma automated, higher energy cyclotron - a control system including an Linda University, the Lawrence with isocentric gantries. Their system "accelerator control unit" with inde­ Berkeley Laboratory and Science will be installed at Massachusetts pendent and networked "therapy Applications International Corpora­ General Hospital, Boston. control stations". Through this net­ tion (SAIC) of San Diego. The proton therapy system com­ work, each of the therapy control prises: systems can also take over the From Y. Jongen, IBA S.A., Louvain, - a 235 MeV isochronous cyclotron to computer-based unit controlling the Belgium

10 CERN Courier, July/August 1995