Search

A new high-energy neutron therapy facility has been installed at the Douglas Cyclotron Centre, Clatterbridge Hospital, Merseyside, in order to extend the clinical trials of fast neutrons initiated by the Medical Research Council. The neutron beam is produced by bombarding a beryllium target with 62 MeV protons. The target is isocentrically mounted with the potential for 360 degrees rotation and has a fully variable collimator. This gives a range of rectilinear field sizes from 5 cm x 5 cm to 30 cm x 30 cm. Basic neutron beam data including output, field flatness, penumbra and depth-dose data have been measured. For a 10 cm x 10 cm field, the 50% depth dose occurs at 16.2 cm in water and the output is 1.63 cGy microA-1 min-1 at the depth of dose maximum. The effectiveness of the target shielding and the neutron-induced radioactivity in the treatment head have also been measured. It is concluded that the equipment meets both the design specifications and also fully satisfies criticisms of earlier neutron therapy equipment. A full radiation survey of the centre was also carried out and it was found that radiation levels are low and present no significant hazard to staff.

Measurements have been made of beam profiles at 10 cm deep for the neutron beam of mean energy 7.5 MeV produced by the MRC cyclotron at Hammersmith Hospital. Neutron and gamma-ray doses in water were measured with a tissue-equivalent chamber and a Geiger-Müller counter. Far outside the beam the radiation consists mostly of gamma rays penetrating the shielding. Perfect shielding would reduce this but the profile of biologically effective dose would still be worse than that from conventional X- and gamma-ray generators owing to the high RBE of small doses of neutrons. The main hazard to the patient is probably that of carcinogenesis. An approximate assessment is made of this hazard which is compared with that estimated in previous publications.

We have measured the biological equivalence of the Clatterbridge neutron therapy beam [p(62)-Be] and the Hammersmith neutron therapy beam [d(16)-Be] using the mouse intestinal crypt assay. The ratio (NDR) of Clatterbridge neutron (n + γ) dose relative to Hammersmith neutron dose (n + γ) was found to be 1.2–1.13 over a dose/fraction range of 1.8–9 Gy at 2 cm deep in a Perspex phantom. It is shown that the effectiveness of the Clatterbridge beam was reduced with penetration into the phantom because of hardening of the beam to a maximum reduction of 11% at 12 cm deep in the phantom. The hardening of the beam with depth of penetration will need to be taken into account by clinicians in assessing the tumour dose and tissue tolerance. Relative biological effectiveness values for the Clatterbridge and Hammersmith neutron beams were also measured. All neutron doses for both Hammersmith and Clatterbridge beams are total doses (n + γ) which comply with the European protocol for neutron dosimetry and includ...

Chinese hamster V79 cells have been used to assess changes in RBE of the p(62)Be neutron beam at the Clatterbridge Hospital with depth in a phantom and with use of a hydrogenous filter. The cells were exposed at depths of 2 and 12 cm and at a depth of 2 cm with a hydrogenous filter. Two groups of experimenters each conducted two experiments. The ratios of relative biological effectiveness (RBE) at a depth of 12 cm to that at 2 cm were found by the two groups to be 0.99 ± 0.04 and 0.96 ± 0.02 (standard errors). The effect of a polythene filter 4.5 cm thick was measured at a depth of 2 cm and the ratio of RBE with and without the filter was found by both groups to be 0.99 ± 0.02. All the experiments suggest that there may be small effects of beam hardening by depth and filtration but these results are in marked contrast with those obtained using an in vivo system.

A carbon calorimeter is described which has been used to measure the energy locally absorbed from beams of X radiation generated at 4, 7·5, 14 and 20 MV and from γ rays from 60Co. The calorimeter is of the isothermal twin type and does not need a thermostatic bath. The calorimetric measurements have been compared with ionization in a carbon chamber of known volume. The calorimeter and chamber were designed to be as similar as possible, so that they measure, to the same degree of approximation, the energy absorption at a point in a block of carbon and the ionization in a small cavity at the same point. With cobalt γ rays the measurements give W = 34·0 ± 0·2 eV per ion pair in air. The ionization in the carbon chamber has been compared with that in a Baldwin-Farmer chamber exposed under particular conditions in a Perspex block, and a calibration for the Baldwin-Farmer chamber in terms of absorbed dose in water has been derived.

The relative effectiveness of X radiation generated at 8 million volts and γ radiation from radium is compared in patients having treatment for cancer of the anterior part of the mouth. The onset of the mucosal reaction has been used as the indicator of biological effect. Comparison of the doses received on the mucosa on the day of onset of the reaction suggests that the RBE of 8 MV X radiation as compared with γ rays is slightly less than unity. Factors which may affect the reaction other than the energy of the radiation are discussed.

Neutrograms of patients are being taken during the course of radiotherapy with the fast neutron beam from the MRC cyclotron. A combination of industrial X ray film with radiographic salt screens is used so that the cassette can be left in position during most (or all) of a treatment. Fogging by scattered neutrons and γ rays is reduced by use of a patient-film separation of 20 cm. Bone is practically invisible on neutrograms which show only the contrast between tissue and air spaces. It is doubtful, however, whether this fact would enable any anatomical information to be visualized which could not be shown with X rays using a suitable kV and projection. Fundamental problems in the diagnostic use of neutron radiography are the insensitive imaging techniques, leading to high dosage to the patient, and increased mottle in neutron pictures.

The neutrons are generated by accelerating a beam of deuterons to 15 MeV in the M.R.C. 45 in. Cyclotron (Gallop, Vonberg, Post, Powell, Sharp and Waterton, 1957). A fraction of the beam is extracted from between the dees and falls on an external target consisting of a thick sheet of beryllium (the thickness of the target was 290 mg/cm2, the range of the deuterons being 180 mg/cm2). From here neutrons radiate in all directions, being most intense in the forward direction. The M.R.C. cyclotron was designed to give a large beam current and to run with the greatest possible degree of reliability. It performs well, giving about 50 μA at the external target for long periods of time. Figure 1 is a general view of the machine as it was several years ago, before the external target had been added. Figure 2 is a plan view of the cyclotron chamber. Neutrons are shown emerging from the target and crossing the space where the radiobiological experiments were performed. The cyclotron team is at present modifying the ma...

Ehrlich mouse ascites cells have been used as a biological dosimeter to check physical methods of measurement of central axis depth doses of a beam of fast neutrons. The cells were irradiated in suspension in Perspex tubes, using air or nitrogen bubbling. Irradiations with fast neutrons under aerobic conditions were carried out at four depths in a tissue-equivalent phantom and in the absence of the phantom. The relative intensities of the radiation under these conditions agreed well with measurements using a special ionisation chamber, provided allowance was made for accompanying γ radiation. The oxygen enhancement ratio was found to be 2·78 with 250 kV X rays and 1·80 with the fast neutron beam. No signicant change of OER was observed for irradiations at 8·7 cm deep in the phantom.

Calculated values of integral doses delivered during radiotherapy at 8 MeV and with radiation of H.V.L. = 2 mm of Cu are presented. These values have been checked experimentally. Meredith and Neary's method is found to give reliable values of integral dose at H.V.L. = 2 mm of Cu. The experimental measurements were made with a celluloid model of the human body, the model being composed entirely of ionization chambers. A description of the model is given, together with an analysis of the derivation of integral doses from measured ionization currents. The integral doses delivered during typical treatments at the two qualities are compared. It is found that in practice integral doses are not always greatly reduced at 8 MeV, owing to the differing field arrangements used.

Fast neutrons were used therapeutically in California between 1937 and 1943, and the disadvantage of the long-term skin damage was mentioned in the introductory paper. Robert Stone and his collaborators treated 249 patients with the fast neutron beams from the 37 in. and 60 in. Berkeley cyclotrons. Accounts of the work and the clinical results obtained have already been published (Stone, Lawrence and Aebersold, 1940; Stone and Larkin, 1942; Stone, 1948) and only the salient facts will be summarised here. All the patients treated, except one, suffered from advanced cancer and were considered incurable by any conventional technique. The exception, a patient with a laryngeal tumour, was suitable for treatment by conventional radiation, but himself elected to have neutron therapy; he remained free of disease for at least six years. This work was not a controlled comparative trial of fast neutrons and X or γ rays. The primary tumours were of varied histology and originated in many different anatomical sites, t...

Experiments have been carried out to compare the skin reactions in pigs produced by various regimes of fractionated irradiation with X rays, extending up to 28 days. A total of 27 X-ray fields were applied to two pigs. The results show that the major factor which determines the skin response is the size (or number) of the fractions, i.e. the shape of the dose—effect curve. This factor is found to explain approximately five-sixths of the increase of total dose required to produce a given reaction when the dose is divided into five fractions. The remaining one-sixth of the increase is due to tissue repair or repopulation occurring between five and 28 days. The implication in radiotherapy is as follows. If fractionated treatments using fewer and larger fractions are given over the same overall time as a “daily” regime, it is suggested that care must be taken to reduce the total dose below that used in the daily regime, in order to avoid greater skin damage.

The 8 MeV linear accelerator of the Medical Research Council, which is installed at Hammersmith Hospital, London, is described. An account is given of the physical measurements which have been made to enable the best use to be made of the machine for X-ray therapy. These measurements include work on protection, beam flattening, output calibration and isodose curves. The machine normally gives a stabilised dose-rate of 100 rads/min at 2 cm deep in a water phantom with its surface at 1 m from the target. Four hundred patients were treated between September 1953 and December 1954. A report is made on the operation of the machine, which has been very satisfactory.

There is some uncertainty about the oxygen enhancement ratio (OER) for mammalian and plant cells irradiated with high-energy neutrons. With neutrons generated by 80 MeV deuterons on Be and 101 MeV protons on Be widely varying results have been reported (Hall et al., 1977; Hall and Durand, 1980; Harrison et al., 1976) and there is at present no information on OER for neutrons of higher energy. To obtain more information with high-energy neutrons we have measured the OER with mouse Ehrlich ascites cells irradiated with neutrons generated by 160 MeV protons on Be from the synchrocyclotron at AERE, Harwell.

Mice have been irradiated with fast neutrons (mean energy 6 MeV) under normal and hypoxic conditions. Hypoxia was induced by brief periods of nitrogen breathing. Observations of the thymus weight two days later gave an oxygen enhancement ratio of 1·7 ± 0·2. Earlier irradiations with 8 MeV electrons gave an OER of 2·7 ± 0·3. This gives a gain factor of l·6 ± 0·2. The RBE relative to 8 MeV electrons is 2·35 in air and 3·82 in nitrogen.

The physical and radiobiological properties of two neutron beams have been compared. The beams were generated by deuterons of 16 MeV at Hammersmith Hospital and 42 MeV at Harwell, in both cases falling on a Be/Cu target. The dose-rate and depth-dose characteristics at the higher energy were found to be superior to those at the lower energy. Collimation and shielding at the higher energy are facilitated by the greater degree of forward-peaking and by the fact that a higher dose-rate allows longer collimators to be used. Attenuation in iron was found to be similar at the two energies. The radiobiological properties of the two neutron beams are very similar. There is a difference of about 20 per cent in RBE for effects on mammalian tissues for doses between 300 and 2,000 rad of neutrons. The OER and the sparing effect of two large fractions are the same for the two beams.

Survival of reproductive capacity of murine leukaemia P-388 cells was assayed in vivo after the cells had been irradiated in vitro under aerobic or hypoxic conditions with collimated beams of X rays or 16 MeV D-Be fast neutrons at various depths in tissue-equivalent phantom material. The response to X-irradiation was the same in the absence of the phantom and at 8·7 cm depth. The response to fast neutrons under aerobic conditions was unchanged from 0 to 23 cm depth within the phantom. However, under hypoxic conditions, the dose-response curve for fast neutrons became significantly steeper with increasing depth in the phantom. The OER decreased from 2·0 in the absence of the phantom to 1·5 at 15 cm deep.

The fast neutron beam produced by the Medical Research Council's cyclotron at Hammersmith Hospital, London, has been in use for radiotherapy since 1966 and for radiobiology for several years previous to that date. Since the inception of the fast neutron beam at Hammersmith we have been concerned particularly with maintaining reproducible and internally consistent dosimetry, with the knowledge that the absolute accuracy with which the accepted unit of absorbed dose, the rad, could be realized was poor.

This machine was the first linear accelerator designed for medical use and the first to be installed in a hospital. The first patient was treated in 1953 and the last in 1969. During this period the machine was used for research during the evenings and latterly it was available for research all day long. It was switched off for the last time in February 1984. In addition to the X-ray beam, electrons were used for treatment and for research, where the high dose rate and large field size were particularly useful. An account is given of some of the highlights of the clinical and research programmes.

The reproductive capacity of isologous P–388 leukaemic cells has been determined in vivo after irradiation with fast neutrons from the M.R.C. cyclotron, Hammersmith Hospital. When tumour-bearing donor animals are irradiated at 10 cm depth in tissue-equivalent phantom material, using a relatively narrow fast neutron beam, the results are indistinguishable from results when donor animals are irradiated in air using a wide neutron beam. Implications of this finding for radiotherapy of human tumours are discussed.

The proportion of animals surviving four days after whole body neutron irradiation is not altered by giving the radiation in two fractions separated by four to six hours. This result is compared with previously reported data after 250 kV X irradiation when it was found that a dose higher by 400–600 rads was necessary to produce the same killing effect if the radiation was given in two fractions rather than a single dose. This results in a larger RBE (250 kV/neutron) for the fractioned radiation than for single-dose irradiation and any increase in the number of fractions would increase the RBE further. These results may be relevant to the use of neutrons for therapy.

A beam of fast neutrons, of mean energy about 6 MeV, has been in use for radiotherapy since August, 1966. Patients are treated in a room next to the cyclotron chamber, the beryllium target being located in the wall separating the two rooms. Treatments are given at 120 cm FSD, and at this distance the average dose-rate is about 50 rads/minute. A set of wooden collimators, which are easily changed by hand, gives a variety of field sizes up to 20 × 20 cm. A protective cone round the collimator reduces the stray radiation to an acceptable level. Depth dose characteristics are intermediate between those of 200 kVp X rays and 60Co γ rays. The dose from γ radiation amounts only to a few per cent of the dose from fast neutrons and has been measured with photographic films and thermoluminescent dosimeters. Thermal and epithermal neutrons make only small contributions to the absorbed dose. The beam has been developed sufficiently for a clinical trial to be made. The penetration of the beam is adequate for ...

Fifteen fields on one pig were irradiated with 1, 2, 2, 3 and 5 fractions in 1, 5, 2, 3 and 5 days respectively. Skin reactions were recorded up to 80 or 100 days, so that the doses which produced equal reactions could be assessed. Three degrees of damage were analysed for early, medium and late reactions. As an example, the doses to produce one level of reaction over 30–85 days were respectively 2,050, 2,560–2,750, 2,750, 3,360 and 4,060 rads for the fractions listed above. The corresponding values of “DQ” = (Dn — D1)/(n —; 1) were high, 500–700 rads. The values of (Dn − D1)/(n − 1) became smaller with larger numbers of fractions. Two pigs were irradiated to investigate the “reciprocal vicinity” effect. It was found to be small for the field arrangements used, but the sensitivity of pig skin increased in the dorso-ventral direction and corrections were sometimes necessary. Two pigs were irradiated with pairs of equal doses separated by 0 to 24 hours. A recovery pattern was observed similar to that found ...

It is a common experience that B.D.11 ionization chambers, when used for a routine protection service, are liable to suffer from spurious charge losses. These are usually put down to causes such as the presence of hairs or excessive humidity, although we have found that ionization chambers stored in saturated air for several hours do not in fact lose charge as a result; but we have found that chambers do lose charge when subjected to mechanical shock.

The skin of pigs has been used to study the effect of fractionated irradiation with fast neutrons of mean energy 6 MeV. Six, nine and 15 fractions have been used over a period of three weeks, and two fractions divided by short intervals of time up to 24 hours. The effects of fractionated treatments with X rays are also summarised. The RBE of the neutron beam rises steadily from 2·3 for a single dose to 3·6 for 15 fractions. Observations up to five years after treatment have failed to show any sign of increased late reactions after neutrons. There is less recovery during a fractionated course of treatment with fast neutrons than with X rays, the relative amount of recovery being 0·6. Very little recovery was observed during the first seven hours after treatment with neutrons. Reactions occurring after fractionated treatments with X rays show that the number of fractions is more important than the time over which the treatment is extended, up to 28 days, but that prolongation is nevertheless a sign...

The use of an 8 MeV linear accelerator for the treatment of patients with electrons is described. Dose distributions in water show a zone extending from the surface to a depth of 2·4 cm in which the irradiation is uniform to within 20 per cent, with a steep decline at greater depths. It is considered that effective treatment can be given only to a depth of 2·4 cm. The types of case which can be treated with 8 MeV electrons are discussed and a few treatments which have been given are described.

Tissue equivalent phantoms have been used to estimate the dose to various parts of the foetus and to the maternal ovaries during obstetric radiography at 100 to 120 kVp. These phantoms were made by taking plaster casts of pregnant women at term. An estimate has been made, for each projection, of the dose delivered to the foetal gonads in the average case. An estimate has also been made of the value of the dose averaged over the whole foetus, for each projection. The inlet projection is found to give rise to by far the highest radiation levels. A short investigation was made of the various factors which influence the doses delivered in radiography and some recommendations are given on methods of keeping the doses to a minimum.

The International System of Units (SI) was adopted internationally in 1960. In 1968 the Royal Society of Medicine organized a conference of medical editors to discuss the introduction of SI units, but their recommendations did not appear until 1972 (Ellis, 1972). Now another, more comprehensive book on the subject has appeared (Armstrong Lowe, 1975). Nevertheless, few journals have made a complete change to the new system. Why the long delay? In the medical field no doubt writers have held back until the hospitals had changed over internally to the new system.

In a recent letter Smith and Shah (1970) raised doubts about the significance of differences reported by Bates (1969) in the recurrence-free rates for stage III carcinoma of the cervix treated by radiotherapy with and without hyperbaric oxygen. Without wishing to dispute their point that a real difference in results was not conclusively demonstrated, a point which agrees with the conclusions of Bates (1969), I think the statistical problems involved can usefully be looked at from a slightly different point of view.

Some photographs are shown which give a direct demonstration of the shape of the curve connecting the forward emission of electrons due to the γ rays of radium and the atomic number of the substance used as the secondary emitter. They also indicate the variation of ionization at various depths in tissue-like material.

The reader of the paper by Pointon et al (1985) might well infer from the discussion section that the Hammersmith trials included carcinoma of the bladder. In fact, we have treated no patients with bladder cancer because the penetration of our neutron beam was inadequate and its penumbra too wide. The beam of 15 MeV neutrons used by Dr. Pointon had similar penetration and an even worse penumbra. In addition, we had the problem of a beam fixed in a horizontal direction.

The table below gives an analysis by subject of the contributions to the journal during 1976. The journal covers a wide range of subjects, as indicated on the front cover. Papers extending across the frontiers between subjects are a particularly noteworthy feature, reflecting the varied membership of The British Institute of Radiology which includes both medical and non-medical members on an equal footing. This, however, makes it difficult to allocate the papers in an objective way. In the table, papers put in interdisciplinary categories are those in which substantial parts relate to more than one category.

In 1976 we wrote to report a revision of neutron dosimetry at Hammersmith (Parnell & Bewley, 1976). The effect of this was that the old standard dose of 1440 rad (cGy) was changed by 8% and became 1560 rad (cGy). Dosimetry at Edinburgh was based on the new Hammersmith values. Since 1976, various protocols for neutron dosimetry have been published. It is now customary in Europe to use the European Protocol (Broerse et al, 1981). In this protocol the various factors which are used in neutron dosimetry were assigned rather different values from those previously used at Hammersmith and Edinburgh. When dosimetric measurements were made using the new protocol, the values of neutron dose, DN, turned out to be greater than the value used for clinical work at Edinburgh by 3–4% and at Hammersmith by 5–6%.