Köhler A. Zur Röntgentiefentherapie mit Massendosen. Münchener medizinische Wochenschrift 56: 2314-2316, 1909. (Slatkin DN and Laissue JA. English translation. May 2005)
The author presented a lecture with a similar title at the Belgian Röntgen Society in Antwerp on May 30, 1909. It concerned the theory of a method designed to deliver doses to deep tumors that are about ten to twenty times greater than have been hitherto possible without damaging the overlying skin. The significance of the method for the treatment of deep malignancies should be obvious. The principle, which has been described in detail in "Fortschritten auf dem Gebiet der Röntgenstrahlen, Bd. 14, Heft 1" (Advances in the Applications of Roentgen Rays, Vol. 14, Part 1), could not be simpler; it is summarized here in a couple of sentences:
An X-ray tube with a very large cathode-ray hot-anode target area -- 4 to 8 times larger than usual -- produces very sharp shadows of a metal grid if the latter is positioned close to a photographic plate or fluoroscopic screen. However, if one distances the grid 4 to 5 centimeters from the plate and brings the tube to within a few centimeters of the grid, the shadows disappear and the plate is uniformly exposed, leaving no trace of the grid.
The implications of this for therapy are as follows. If one places a grid right on the skin overlying a several cm-deep malignancy and if an X- ray tube with a large target area is positioned only a few centimeters from the grid, the skin will be shielded underneath the metal and only irradiated through the grid's spaces, but the tumor will be irradiated evenly. If one applies this method to deliver a massive dose, for example tenfold greater than one full erythema dose, the resulting multiple foci of burned or necrotic skin - whatever the case may be - will be healed within a few weeks by surrounding surviving skin cells. Without the grid, the same dose would cause an extensive, continuous skin burn that would never heal, or would only heal with scarring many years later. X-ray tubes with such large targets must be made to order, as they make fuzzy radiographs and therefore can be used only for therapy. However, they are simpler to manufacture than are X-ray tubes with small, sharply focused target spots. This report also reflects ideas that I developed from the discussions that followed my Antwerp lecture. Whatever may come of the proposed method, the physical basis for it is not in doubt.
It should be stressed that misunderstanding of the method may result, for example, in poor placement of the grid. The grid must be on the skin, preferably pressed close to it. Moreover, for fundamental reasons indicated below, the most critical aspect of the system is a thin filter separating the grid from the skin. Only with that filter can cells beneath the metal strands remain entirely viable, enabling rapid healing of necrosis in their vicinity. An important question arises: In general, is it necessary to insert a thin filter (for example, of leather or aluminium) under the grid? Probably, one should reply confidently in the affirmative, in particular because skin cells under the metal are directly irradiated by the grid's secondary radiations. Although the smaller the atomic weights of absorbing substances, the softer and less penetrating are their secondary radiations, even secondaries from elements of high atomic weight such as lead or iron are so feeble that they can scarcely penetrate a thin sheet of paper. Nonetheless, secondary radiations from such metal strands, if generated by the equivalent of 10 to 20 primary-radiation erythema doses, could deliver enough dose immediately under the strands to damage skin cells not shielded by our thin-filter technique. Therefore it is recommended that one insert a thin filter between grid and skin to absorb all but a small, higher-energy portion of those secondaries. Furthermore, we gain another advantage by using a thin, light filter: it absorbs the preferentially skin-damaging soft portion of primary X rays that bypass the strands without absorbing a significant portion of those primaries of intermediate hardness that also act therapeutically at depth.
The shape and thickness of the metal strands of the grid and the size of the grid spaces were also considered during those discussions. It was recommended that the metal strands be triangular or rectangular - not circular - in cross section. The reasons for that are so clear that they need not be specified -- indeed so obvious that the author has already had such grids manufactured.
In metallurgy it is feasible to make grids of platinum or lead (those metals would appear to be the best) --- although even ordinary window screens might suffice. A filter made like the platinum upper screen of a Bunsen burner would have metal filaments too fine to shield the skin effectively, so I would prefer a different material. To be pressed against the skin, a grid should be as stiff as possible - another reason for choosing a different material. G. Schwarz** has recently shown that susceptibility of the skin to X-ray damage is reduced if one inhibits its metabolism, for example by direct pressure; the stiffer the grid, the more effectively could pressure be applied. Since thick platinum wires are very expensive, one should prefer a grid of iron wires. The somewhat smaller X- ray absorption coefficient of iron than platinum would be inconsequential.
Theoretically, it might be preferred that the metal strands of the grid not overlap; the grid could then be stamped from a metal sheet. In practice, however, not only should overlapping strands of metal cause no harm, they could even save areas of unshielded skin after the latter receive erythema doses intense enough to damage epidermal cells beneath non- overlapping strands. This provides all the more reason for not discounting the potential of islands of epidermis beneath overlapping strands of metal to remain intact after such irradiation to heal adjacent zones of skin necrosis. It is left for further investigation to judge what gauges of iron wire would be best for a grid. In the meantime, I have ordered a grid of 1 millimeter-wide wire with 2.5 x 2.5 square millimeter spaces between them from the firm of Reiniger, Gebbert, and Schall. So that grid and filter remain as close as possible to the skin --- on the trunk, minimally perturbed by breathing --- it is best if they are pressed so hard against the skin as to blanch it. This can be done, for example, using the cylindrical glass housing of an X-ray collimator.
Makers of Röntgen tubes all know that it is much less difficult to make tubes with a broad, rather than a precise, narrow cathode-ray target spot. This author was pleased to hear from a tube-maker during our conversation in Antwerp that fabrication of tubes with broad target spots was really a straightforward matter, involving no challenging steps. Moreover, I learned from an electricity technologist there that such tubes enable higher power to be used for therapy, so that if one doubled the size of the cathode, there could result a proportionate increase in the radiation dose rate --- which could also damage the tube. Alternatively, it has long been known that a tube can have several cathodes, which would be particularly advantageous for our method. If such a tube had a wide anode with three or four separate, broad target spots, its X-ray output would be intensified, which would likewise speed up our work - an appreciable advantage.
Several points should be mentioned about the stress on the tube. Even if the distance from tube to skin were minimized, it would still take 1 to 2 hours to deliver 10 to 15 full erythema doses. One would wish that to be quicker, which would overload the tube somewhat; an overloaded tube would of course soon be weakened to the point of uselessness. With its vacuum compromised, such a tube could not deliver an exact dose -- and one would not want a 10- to 15-fold increase of dose to be inexact by more than one- half of a full erythema dose. A solution would be to wire several tubes in series and position them securely in their housings; with collimators close to grid and filter, it would be unimportant if individual tubes were not perfectly aligned toward the same target. Alternatively, one could implement the author's deep-irradiation technique by using several readily available Röntgen tubes. Each tube, having a small target spot, would be aligned and centered individually in a common housing. Other improvements might be to operate four of the ordinary tubes at the same distance from the skin . and so on. It goes without saying that this method, which can be simple, should not be made more complex needlessly. For example, the deeper the irradiation target, the less is it required that the target spot be extra-large, so even an ordinary Röntgen tube could be used confidently to use the grid technique for a 10 centimenter-deep lung tumor.
An objection to the method was voiced during the aforementioned discussions which, at first, appeared reasonable: the grid itself cuts down considerably on a dose intended to be delivered at depth. For example, about half of the incident radiation on a grid with 1 millimeter-wide wires and 2.5 millimeter-square holes would be blocked. To compensate for that, the unshielded zones of skin would have to receive double the dose they would receive without the grid. Strictly speaking, the facts underlying that objection are true, but it should not be accepted unreservedly. With a grid, unshielded zones of skin can receive 10 to 20 times the dose deliverable to unshielded skin because strands of shielded cells surround each hole; without a grid, such doses could not be delivered. Even with thinner wires and much larger holes, rapid healing of many small zones of skin would follow. Should a problem of non-healing skin arise, it could be addressed by constructing tubes with several special, large cathodes or with several small cathodes per tube.
We now turn to matters of biology. It is known that necrotic or ulcerated zones of skin are prone to infection, which would lead to destruction of zones of remaining viable skin cells and subsequent confluence of those zones to form an extensive, non-healing skin ulcer. Since that could indeed occur, the radiologist must thoroughly disinfect the irradiated zone and, immediately after the irradiation, bandage it to keep it so until it heals. Should that prove insufficient to prevent an infection the grid could be made of thicker wires, but the grid method should not fail on that account in any case.
The restriction that a grid irradiation cannot be repeated is quite beside the point. If one can increase a dose 20-fold using this method, one will have achieved very much more than what has been achieved to date by delivering a maximum tolerable dose in fractions. Moreover, there are a number of regions in the body (breast, bones of the extremities, and such) that can be irradiated from 3 or 4 different directions. Granted, a zone of the skin once stressed by this grid technique cannot be so stressed ever again, but since this technique permits delivery of an enormous total dose, perhaps 40 times greater than has been possible to date, one should be encouraged by its development. That said, however, the radiotherapist should remain vigilant --- aware of the importance of sufficient shielding of the surrounding tissues, even of the whole body, from such large doses of X rays.
Animal experiments have been initiated, but for a number of reasons I don't have much to say in their favour. I have decided that my primary obligation is to apply this method as soon as possible to treat malignancies that are likely to kill a patient within several months.
In summary, I concede that this grid method carries some disadvantages with its advantages, as do all kinds of therapy. Among its advantages are: delivery of massive doses, relatively short exposure times, and (even with filters and compression) straightforward implementation. Those compensate for its few disadvantages, the most important being the risk of losing potentially healing strands of cells to infection --- the likelihood of which is lessened by using aseptic techniques.
**Gottwald Schwarz (1880-1959)
Alban Köhler. "Une nouvelle méthode permettant de faire agir, dans la profondeur des tissues, de hautes doses de rayons Roentgen et un moyen nouveau de protection contre les radiodermites" (A new method to deliver high doses of Roentgen rays to deep tissues and to mitigate radiation dermatitis.) Annales d'Électrobiologie et de Radiologie 10, 661-664, 1909. (Slatkin DN and Laissue JA. English translation. December 2006)
Previously, radiologists who wanted to deliver a high dose of Roentgen rays to tissues several centimeters deep faced a Hobson's choice. Delivering a high dose of unfiltered Roentgen rays to deep tissues caused intractable skin ulcers. Although filtration of the rays enabled deeper penetration, it deprived them of almost all their therapeutic efficacy - to such an extent that, in a good many cases, cell division seemed to be stimulated instead of inhibited or stopped by them.
Recently, various studies have been directed toward alleviating these problems by irradiating from a great distance, but they are not yet ready for evaluation; certainly, it should be understood that any such technique could, at best, only partially improve the efficacy of these radiations. However, were it possible to deliver doses to deep tissues 10-15 times greater than are achievable at present without exposing patients to the concomitant risk of intractable skin ulcers, that would obviously supersede existing methods, especially if greater simplicity and convenience counted among its advantages.
I now submit the following method for consideration; I have been led to it by experiments that, so far, have not been designed to test its therapeutic efficacy but rather to investigate analogous setups in roentgenography.
If one puts a 1- to 2-millimeter-mesh grid of metal wires on a photographic plate and exposes the assembly to X rays from a cathode-ray tube [CRT], of course one gets a distinct photographic image of the grid. However, if one distances that grid about 20 centimeters from the plate, parallel to it, then positions the CRT close behind the former, one gets only a very hazy grid image; if one then proceeds to distance the grid 40 or 50 centimeters from the plate, one no longer obtains any grid image - not even the least indication of it; the radiosensitive layer of the plate is uniformly darkened as if there were nothing to block X rays between the CRT and the plate.
The reason for this effect is well known by now: Investigations of the physical properties of Roentgen rays show, in brief, that the effect does not involve refraction or diffraction, but simply depends on the design of the CRT's focus. In ordinary CRTs, the focal zone on which the convergent cathode rays impact is, of course, not a geometric point but rather a circle of greater or lesser diameter, usually 1 to 5 mm. (One should note, however that so-called precision CRTs have much smaller focal zones.) Thus, to obtain equal, constant darkening of the plate, i.e., disappearance of any trace of the grid, for a given distance from CRT to metal grid, the closer the grid to the plate, the larger should be the CRT's focal zone.
For therapeutic applications, the corollaries of those simple, well- known observations are as follows: First, a small focal zone, so important in roentgenography, is of course unnecessary for Roentgen-ray therapy. Suppose one had an appropriate CRT with a large focal zone, say 1.0 - 1.5 cm in diameter. Suppose also that one selected a metal grid - either a grid of ordinary iron wire or, better, a mesh of lead or, best, of platinum wire - and applied it directly to the skin or to a thin leather filter on the skin; the CRT is positioned several centimeters from the grid. If we so irradiate the surface of the grid, we will achieve constant, uniform irradiation of tissue at a certain tissue depth - as homogeneous as if there were no metal grid between CRT and deep tissue.
Moreover, having that grid on the skin surface enables the delivery of enormous doses of X rays to deep tissues in comparison with those possible to date, without risk of non-healing ulcers or of nearly incurable radiodermatitis. Whereas each cell is homogeneously irradiated in deep tissue, for example in a deep tumor, X rays penetrate the skin only between the grid elements, so cells directly under the metal grid elements remain, in effect, intact.
With large doses, the skin between the grid elements is certainly subject to erythema and necrosis, but those lesions are spontaneously repaired within a few weeks because individual zones of necrosis are surrounded on each side by a narrow slice of healthy, well vascularized tissue. An analogous method is already established in surgery where, for example, one treats an angioma by thermocautery; as is well known, the many eschars that result are healed rapidly. The conditions for healing such deep eschars are not as favorable after grid-mediated radiotherapy as after thermocautery. Nevertheless, the principles of those techniques are analogous. In any case, it is much better to reduce radiodermatitis by using a metallic grid than to cause large, deep ulcers by exceptionally high X-ray doses delivered without a grid.
The question of knowing how many times more intense the irradiation may be with than without a metallic grid is difficult to answer; it depends, to a certain extent, on what kind of metal is used to make the strands of the grid and on their thickness. Thus, as many as one hundred erythema doses* might be needed to elicit necrosis of skin cells shielded by such strands of metal, and perhaps as many as fifty erythema doses might be needed to cause any visible lesions at all in zones of skin so shielded. Using approximately fifteen-fold greater erythema doses to unshielded than to shielded skin cells, the skin recovers so well that radiation damage may be discounted entirely. Although this leaves metal-shielded tissues apparently unaffected by the irradiation, it is not absolutely certain that such doses will allow necroses in exposed zones to heal readily. Based on my preliminary studies, it seems wise in practice to use not more than ten erythema doses. More detailed experiments will establish exactly how extendable that dose limit may be.
I have not yet had a chance to practice this technique. However, since the procedure I have outlined here is based on simple and precise physical and biological phenomena, there is no doubt that its practical value will be recognized some day. By then, animal studies will have shown whether doses that are twenty-fold or even fifty-fold greater than those allowed currently can cause skin damage too severe to be justifiable by the benefits of their efficacy for deep lesions. For the first clinical trials, one should be able to enlist patients in whom a laparotomy will have already revealed an inoperable malignant tumor.
This method appears to have no disadvantage other than causing minuscule zones of necrosis, already described. On the other hand, it has the following advantages in addition to those already mentioned: Fundamentally, the method depends on irradiating at a very short distance, which will allow much shorter exposure times than those used at present. For example, for a distance of 5 cm between the glass wall of a 15 centimeter-diameter CRT and the skin surface (i.e., 7.5 + 5 cm from focal zone to skin), one gets a 16-fold stronger X-ray effect than would be obtained during the same exposure time with the CRT distanced 50 cm from the skin.
Another advantage of my method is that it allows X-ray-energy filters to be used just as well as in older methods. Any kind of filter may be placed underneath the metal grid1; leather filters are recommended in particular. It would be advantageous to apply some pressure to the grid against the filter, either with a diaphragm or with the leaded-glass cylindrical tubes used for the radiotherapy. Finally, one can augment the method's effectiveness by irradiating from various directions. In this way, one can treat a deep lesion, for example a sarcoma of the thigh, with exceptionally high doses by directing X rays toward four different sides of it in succession.
Another aspect is that the metal grid is of value in protection against skin burns from irradiation of superficial lesions, particularly for inexperienced technologists not yet certain of the doses they are delivering. Using the grid, an overdose that causes an ulcer or a zone of necrosis should be healed very rapidly. Using this method, irradiations of lesions in patients with blood dyscrasias should also become less dangerous. In therapy of cutaneous lesions, however, it should be kept in mind that zones of skin underneath metal strands will not be treated. It is not certain that cells exposed to X rays traversing the spaces of the grid, having received intense radiation, might have a palliative effect on unexposed cells; judging from certain well-known phenomena, that would not be impossible, but experiments will decide the question. Neither should it be forgotten, in using the method for benign lesions of the face, that one should carefully avoid causing a gridlike pattern of pigmentation in exposed areas.
1 It is especially advisable to use filters to block secondary radiations originating in the metal strands of the grid.
*Translators' note: In 1909, one 'erythema dose' was considered to be the minimum exposure to X rays that resulted in skin erythema. One 'erythema dose' is now considered to have represented a dose of from 2 to 5 gray, depending on the types of X-ray tubes and radiation filters used at a hospital, on the different susceptibilities to erythema of patients typically treated there, and on the different kinds and concentrations of pigment in areas of their skin exposed to X rays.
*PROJECT NUMBER : 91-10
PROJECT TITLE: FEASIBILTY STUDY FOR MICROBEAM RADIATION THERAPY WITH 30-90 keV X-RAYS FROM THE NSLS X17 BEAMLINE
PRINCIPAL INVESTIGATOR: Slatkin, Daniel N.
LDRD FUNDING: FISCAL YEAR 1991 AMOUNT $60,742
PROJECT DESCRIPTION:
This LDRD proposes to carry out feasibility studies in the mouse for the use of monochromatic X-rays from the X17 superconducting wiggler at the NSLS for Microbeam Radiation Therapy (MRT). The studies to be conducted under this LDRD will be: a) Design and assemble a set of computer-driven beam slits to provide square microbeams of 20-1000 micrometer dimensions; b) design and assemble a set of positioning stands for the animal apparatus to position the animal in the beam and move it across the beam in steps of 50-100 micrometers; c) Carry out experiments in which the mouse cerebrum, cerebellum, and eye are irradiated unilaterally with a single monochromatic microbeam at energies of about 30, 60, and 90 keV, at beam dimensions of 25, 50, 100, and 1000 micrometers, and with doses of 20,000, 100,000, and 400,000 rad; d) Evaluate mouse neuropathology and ophthalmic pathology as a function of these irradiation parameters; and e) Repeat the studies indicated in sections (c) and (d) for selected energies, beam dimensions, and radiation doses, with a square pattern of 25 equally spaced parallel pencil beams with a center-to-center beam spacing in the 50- to 150-micrometer range. The goal of the latter study will be to discover the critical interbeam separation at which the probablilty for delayed CNS radiation necrosis is expected to increase sharply with decreasing separation.
TECHNICAL PROGRESS AND RESULTS - FISCAL YEAR 1991:
Synchrotron radiation from the wiggler insertion device of the X17B beamline at the National Synchrotron Light source (NSLS) is practically non-divergent and has enormous flux at photon energies under 100 keV. This means that it is suitable for development of a new type of radiosurgery called 'microbeam radiation therapy' (MRT). As was recently predicted from studies by Curtis et al. using 22 MeV deuteron microbeams at Brookhaven National Laboratory (BNL) three decades ago, and from recent Monte Carlo photon transport and secondary-electron transport computations, a spatially fractionated bundle of parallel, 25 µm-wide, 4 mm-high, 40-55 keV microbeams (delivered at ≈250 gray per second) yields no cerebrovascular damage in the rat one month after irradiation, even at peak microbeam doses of 2500 or 5000 gray. A 1 mm x 1 mm pencil beam of the same quality of radiation delivered at the same dose rate without such spatial fractionation produces colliquative necrosis of the rat cerebrum in its path within two weeks after a dose of only 156 gray.
It is predicted that multiple crossfired bundles of appropriately- dimensioned parallel, planar, synchrotron x-ray microbeams in the 50-150 keV range will ablate tissues in the target zone of a human brain where the bundles overlap, but will spare tissues where they do not overlap when peak/valley doses are as much as 100 gray/5 gray. Such MRT will help to renew radiotherapy in pediatric neuro-oncology and will treat tumors in patients of any age with greatly diminished probability of collateral damage to non-targeted CNS vital structures such as medulla and spinal cord.
MICROBEAM RADIATION THERAPY: Microscopically fractionated 40-55 keV synchrotron x-rays produce no cerebrovascular damage in the rat at doses up to 5000 gray. For MRT, this means that crossfired bundles of planar microbeams will ablate targeted CNS tissues and spare non-targeted tissues when a peak microbeam dose in the ~10 -100 gray range is delivered in several milliseconds.
*pages 36-37 of: Laboratory Directed Research and Development Program Annual Report. G. J. Ogeka
To the Editor, Developmental Medicine & Child Neurology:Charles Kennedy and Avraham Dilmanian1 assessed our review of prospects for microbeam radiation therapy [MRT] in paediatric neuro-oncology2 authoritatively. They also pointed out that unwanted side effects of combination therapy are often "…wrongly blamed solely on the radiation therapy." Not mentioned, but we suspect inferable, is that such blame may have spurred more effective efforts to develop better drugs than better radiotherapy.
Kennedy and Dilmanian rightly observed that there are significant limitations to the depth of targets that can be treated with microbeams. Although we stated ' … the falloff of dose with increasing depth in tissue is considerably more gradual in the 'valleys', which are the sources of reparative cells, than in the 'peaks,' where most of the acute necrobiosis occurs. … ' (p 579)2, we failed to draw attention to the relevance of the former to MRT treatment planning. Also, the attention Kennedy and Dilmanian drew to the greatly attenuated doses at the microbeams' entrances to a piglet's cerebellum 3.4 cm beneath the skin, i.e., 44% of their peak doses to the skin (p 579, Fig 3; p 578, Fig 1)2, but not to more gradual attenuation of valley doses between microbeams in deeper tissues, might have inadvertently overemphasized limitations. Our rough interpolations from Table III of Siegbahn et al.3 yield 50% attenuation depths of ≈5 cm for valley doses versus only ≈3 cm for peak doses in those piglets' heads. Given evidence suggesting that microsegments of blood vessels in mature and immature normal tissues, and in some tumour tissues also4, may recover from considerable radiation damage, valley doses could be viewed with reason just as important as peak doses in planning MRT for paediatric radio-oncology, if not more so if a multidirectional irradiation protocol (p 580)2 were adopted.
A minor point: pre-clinical work in marmosets, although mentioned in an earlier version of the manuscript, was not suggested in our publication.
Respectfully, Daniel Slatkin, Hans Blattmann, HansPeter Wagner, Michael Glotzer, and Jean Laissue
- Dilmanian A and Kennedy C. Prospects for Microbeam Radiation Therapy of Brain Tumors in Children: Commentary. Dev Med Child Neurol 49: 566, 2007
- Laissue JA et al. Prospects for Microbeam Radiation Therapy of Brain Tumors in Children to Reduce Neurological Sequelae. Dev Med Child Neurol 49: 577-581, 2007
- Siegbahn EA et al. (2006) Determination of dosimetrical quantities used in microbeam radiation therapy (MRT) with Monte Carlo simulations. Med Phys 33: 3248 - 3259, 2006
- Serduc R (2006) Effets de la radiothérapie par microfaisceaux synchrotron sur la microvascularisation cérébrale saine et tumorale chez la souris. PhD Thesis, December 5. Université Joseph Fourier, Grenoble, France. (In French)
Microbeam radiation therapy (MRT): Milestones - Clinical prospects
JA Laissue 1, H Blattmann 1 , MA Grotzer 2 , B Kaser-Hotz 3, DN Slatkin 4, HP Wagner 5
1 Institut für Pathologie, Universität Bern, CH-3010 Bern, Switzerland 2 Universitäts-Kinderkliniken, CH-8032 Zürich, Switzerland 3 Vetsuisse Fakultät, Universität Zürich, CH-8057 Zürich, Switzerland 4 Brookhaven National Laboratory, Upton, New York 11973-5000, USA 5 Universitäts-Kinderkliniken des Inselspitals, CH-3010 Bern, Switzerland
Key words: synchrotron X-ray microbeam radiotherapy - MRT - radiosurgery - pediatric neuro-oncology - CNS tumors
Rationale and objectives
Collateral damage to vital normal tissues during radiotherapy can be reduced by using three-dimensional treatment planning and external sources of ionizing radiation. Nevertheless, pediatric oncologists try to postpone or forgo 4 any radiosurgery or radiotherapy, especially for children under three years old because irradiating a child’s CNS entails a substantial risk of dysfunctional central nervous system (CNS) development 1, 2, 3. In radiosurgery, spatially accurate and highly conformal beams of radiation are targeted toward a well-delineated tumor in a single session 5. High-dose radiosurgery using multiple millimeters-wide beams of X rays was first described in 1909 6. In modern radiosurgery 7, multiple millimetres-wide beams of linac-generated X rays, or of gamma rays, converge in the target. Might MRT, a radiosurgery mediated by multiple microscopically thin planar beams of synchrotron-generated X rays, yield larger therapeutic indices for CNS tumors than other forms of radiosurgery or radiotherapy?
Methods
MRT, a spatially fractionated radiotherapy, uses an array of microscopically thin, nearly parallel synchrotron-generated X-ray beams 8, 9. Peak radiation doses are up to fifty times higher than in other radiosurgeries. Unlike conventional radiotherapy, for which the effect of changing an irradiation parameter, e.g., the dose fractionation schedule, is predictable, methods to predict the effect of varying an MRT parameter are only beginning to be developed. Among MRT parameters are array width and height, slit width, spacing of microbeams, energy spectrum, changes in tissue dose microdistribution with tissue depth and, possibly in the future, the schedule selected for temporal fractionation of multidirectional MRT.
Results
In animal experiments, MRT has shown unprecedented sparing of normal radiosensitive tissues as well as preferential damage to malignant tumor tissues growing into and around such normal tissues in laboratory animals 10, 11, 12-15, 16-18, 19, 21, 22, 23-25, 26-28. MRT research at the National Synchrotron Light Source (NSLS), Upton, New York, and at the European Synchrotron Radiation Facility (ESRF), Grenoble, France, has shown that single-fraction, unidirectional MRT yields a larger therapeutic index than does single-fraction unidirectional broad beam irradiation for the intracerebral rat 9L gliosarcoma (9L GS) 13, 15-17, 23-26 and for the transplanted subcutaneous murine mammary carcinoma EMT-6 13-15, as does bidirectional (orthogonally cross-fired) MRT for the subcutaneously transplanted, aggressively invasive, extraordinarily radioresistant murine squamous cell carcinoma VII20.
Since postponing radiotherapy may jeopardize survival of some children with brain tumors 29, MRT has been undergoing and undergoes experimental assessment in living animals because it is believed to be potentially useful for inhibiting children’s brain tumors while sparing nearby normal CNS tissues, which should reduce the burden of malignant cells and, therefore, enhance the effectiveness of ancillary therapies 30. The relative sparing by X-ray microbeams of normal tissues of vertebrates - particularly of their normal central nervous system tissues - has been documented in suckling and adult rats 18, 24, 26, 28, duck embryos 12, and weanling piglets 19. These preclinical results, although encouraging, are not yet sufficient to justify a Phase I (safety) trial of MRT for human patients because they are all based on small animal models, except for a set of studies at the ESRF that used the normal piglet cerebellum 16. All other normal-tissue microbeam tolerance studies at the NSLS and ESRF have used fruit-fly pupae 21, rabbits, rats 17, 18, 24-26, gerbils, mice 14, 20, 22, 27, duck embryos 12, and chick embryos 10, 11.
Conclusions
The 6 GeV ESRF ring is the only source of synchrotron radiation in Europe generating intense X ray microbeams for experimental MRT, having a broad energy spectrum of photons peaking in the 80 – 120 keV range and beam intensities high enough, potentially, to deliver an absorbed physical radiation dose to deep targets in large animals, small children, and adult humans; MRT requires the delivery of several hectogray doses within a fraction of a second, deep to the skin. Regulatory and logistic requirements for implementation of clinical MRT will be stringent. The impetus for investigating the potentially unique advantages of MRT for certain human diseases has been recognized and is sustained by the recent consensus of an ESRF scientific advisory panel of sufficient diversity and broad expertise in its membership to merit serious consideration by the ESRF directorate. Accordingly, we propose that ID17 be used to implement a large-animal veterinary MRT study for veterinary radio-oncology. In that way, a wider community of clinical veterinarians and physicians will be able to assess outcomes from MRT in relation to those from existing radiotherapies for similar lesions in large animals.
References
1. Wagner HP. Cancer in childhood and supportive care. Support Care Cancer 1999; 7: 293-294. 2. Mulhern RK, Merchant TE, Gajjar A, Reddick WE, Kun LE. Late neurocognitive sequelae in survivors of brain tumours in childhood. Lancet Oncol 2004; 5: 399-408. 3. Ribi K, Relly C, Landolt MA, Alber FD, Boltshauser E, Grotzer MA. Outcome of medulloblastoma in children: long term complications and quality of life. Neuropediatrics 2005; 36: 357-365. 4. Rutkowski S, Bode U, Deinlein F, Ottensmeier H, Warmuth-Metz M, Soerensen N, Graf N, Emser A, Pietsch T, Wolff JE, Kortmann RD, Kuehl J. Treatment of early childhood medulloblastoma by postoperative chemotherapy alone. N Engl J Med 2005; 352: 978-986. 5. Adler JR Jr, Colombo F, Heilbrun MP, Winston K. Toward an expanded view of radiosurgery. Neurosurgery 2004; 55: 1374-1376. 6. Köhler A. Une nouvelle méthode permettant de faire agir, dans la profondeur des tissus, de hautes doses de rayons Roentgen et un moyen nouveau de protection contre les radiodermites. Annales d'Électrobiologie et de Radiologie 1909; 10: 661-664. 7. Kondziolka D, Lunsford LD, Loeffler JS, Friedman WA Radiosurgery and radiotherapy: observations and clarifications. J Neurosurg 2004; 101: 585-589. 8. Slatkin DN, Spanne P, Dilmanian FA. Sandborg M: Microbeam radiation therapy. Med Phys 1992; 19: 1395-1400. 9. Laissue J, Spanne PO, Dilmanian FA, Gebbers J-O, Slatkin DN: Zell- und Gewebeläsionen nach räumlich fraktionierter Mikro-Bestrahlung des ZNS mit Synchrotron-Photonen. Schweiz Med Wochenschr 1992; 122: 1627. 10. Blattmann H, Burkard W, Djonov V, Di Michiel M, Brauer E, Stepanek J, Bravin A, Gebbers JO, Laissue JA. Microbeam irradiation of the chorio-allantoic membrane (CAM) of chicken embryo. Strahlentherapie und Onkologie 2002; 178 (Suppl. June 1): 118 11. Blattmann H, Gebbers J-O, Bräuer-Krisch E, Bravin A, Le Duc G, Burkard W, Di Michiel M, Djonov V, Slatkin DN, Stepanek J, Laissue JA. Applications of synchrotron X-rays to radiotherapy. Nucl Instr Meth Physics Res A 2005; 548: 17-22. 12. Dilmanian FA, Morris GM, Le Duc G, Huang X, Ren B, Bacarian T, Allen JC, Kalef-Ezra J, Orion I, Rosen EM, Sandhu, T, Sathe P, Wu XY, Zhong Z, Shivaparasad HL. Response of avian embryonic brain to spatially segmented xray microbeams. Cell Mol Biol 2001; 47: 485-493. 13. Dilmanian FA, Button TM, Le Duc G, Zhong N, Peña LA, Smith JA, Martinez SR, Bacarian T, Tammam J, Ren B, Farmer PM, Kalef-Ezra J, Micca PL, Nawrocky MM, Niederer JA, Recksiek FP, Fuchs A, Rosen EM. Response of rat intracranial 9L gliosarcoma to microbeam radiation therapy. Neuro-Oncol 2002; 4: 26-38. 14. Dilmanian FA, Morris GM, Zhong N, Bacarian T, Hainfeld JF, Kalef-Ezra J, Brewington LJ, Tammam J, Rosen EM. (2003) Murine EMT-6 carcinoma: high therapeutic efficacy of microbeam radiation therapy. Radiat Res 159: 632-641. 15. Dilmanian FA, Qu Y, Liu S, Cool CD, Gilbert J, Hainfeld JF, Kruse CA, Laterra J, Lenihan D, Nawrocky MM, Pappas G, Sze C-I, Yuasa T, Zhong N, Zhong Z, McDonald JW. X-ray microbeams: Tumor therapy and central nervous system research. Nucl Instr Meth Physics Res A 2005; 548: 30-37. 16. Laissue JA, Spanne P, Dilmanian FA, Nawrocky MM, Gebbers J-O, Slatkin DN, Joel DD: Mikrobestrahlung von Gliosarkomen der Ratte: Zell- und Gewebeläsionen (Microbeam irradiation of rat gliosarcomas: Cell and tissue lesions). Schweiz med Wochenschr 1995; 125:1887. 17. Laissue JA, Geiser G, Spanne PO, Dilmanian FA, Gebbers JO, Geiser M, Wu XY, Makar MS, Micca PL, Nawrocky MM, Joel DD, Slatkin DN. Neuropathology of ablation of rat gliosarcomas and contiguous brain tissues using a microplanar beam of synchrotron-wiggler-generated X rays. Int J Cancer 1998; 78: 654-660. 18. Laissue JA, Lyubimova N, Wagner HP, Archer DW, Slatkin DN, Di Michiel M, Nemoz C, Renier M, Brauer E, Spanne PO, Gebbers JO, Dixon K, Blattmann H. Microbeam radiation therapy. Proc SPIE 1999; 3770: 38-45. . 19. Laissue JA, Blattmann H, Di Michiel M, Slatkin DN, Lyubimova N, Guzman R, Zimmermann W, Birrer S, Bley T, Kircher P, Stettler R, Fatzer R, Jaggy A, Smilowitz HM, Brauer E, Bravin A, Le Duc G, Nemoz C, Renier M, Thomlinson W, Stepanek J, Wagner HP. The weanling piglet cerebellum: a surrogate for tolerance to MRT (microbeam radiation therapy) in pediatric neuro-oncology. Proc SPIE 2001; 4508: 65-73. 20. Miura M, Blattmann H, Bräuer-Krisch E, Bravin A, Hanson AL, Nawrocky MM, Micca PL, Slatkin DN, Laissue JA. Radiosurgical palliation of aggressive murine SCCVII squamous cell carcinomas using synchrotron-generated X-ray microbeams. Br J Radiol 2006; 79: 71-75. 21. Schweizer PM, Spanne P, Di Michiel M, Jauch U, Blattmann H, Laissue JA: Tissue lesions caused by microplanar beams of synchrotron-generated x-rays in Drosophila melanogaster. Int J Radiat Biol 2000; 76 (4): 567-574. 22. Serduc R, Vérant P, Vial J-C, Farion R, Rocas L, Rémy C, Fadlallah T, Bräuer E, Bravin A, Laissue J, Blattmann H, van der Sanden B. In vivo two-photon microscopy study of short term effects of microbeam irradiation on normal mouse brain microvasculature . Int J Radiat Oncol Biol Phys; 2006; 64 (5):1519-1527. 23. Slatkin DN, Dilmanian FA, Nawrocky MM, Spanne P, Gebbers J-O, Archer DW, Laissue JA. Design of a multislit, variable width collimator for microplanar beam radiotherapy. Rev Sci Instrum 1995; 66:1459-1460. 24. Slatkin DN, Spanne P, Dilmanian FA, Gebbers JO, Laissue JA. Subacute neuropathological effects of microplanar beams of x-rays from a synchrotron wiggler. Proc Natl Acad Sci USA 1995; 92: 8783-8787. . 25. Smilowitz HM, Blattmann H, Bräuer-Krisch E, Bravin A, Di Michiel M, Gebbers J-O, Hanson AL, Lyubimova N, Slatkin DN, Stepanek J, Laissue JA. Synergy of gene-mediated immunoprophylaxis and microbeam radiation therapy (MRT) for advanced intracerebral rat 9L gliosarcomas. J Neurooncol 2006;78: 135-143. 26. Regnard P, Le Duc G, Bräuer-Krisch E, Troprès I, Siegbahn EA, Kusak A, Clair C, Bernard H, Dallery D, Laissue JA , Bravin A: Irradiation of intracerebral 9L gliosarcoma by a single array of microplanar X-ray beams from a synchrotron: balance between curing and sparing. Phys Med Biol 2008; 53: 861-878. 27. Serduc R, van de Looij Y, Francony G, Verdonck O, van der Sanden B, Laissue J, Farion R, Bräuer-Krisch E, Siegbahn EA, Bravin A, Prezado Y, Segebarth C, Rémy C, Lahrech H. Characterization and quantification of cerebral edema induced by synchrotron x-ray microbeam therapy. Phys Med Biol 2008; 53: 1153-1166.
MRT-planning (similarities and differences with and between planning for therapy with photons, hadrons and MR)
Authors: Blattmann Hans1, Kaser-Hotz Barbara2, Laissue Jean A.1, Rohrer Bley Carla2, Stepanek Jiri1, Bräuer-Krisch Elke3, Bravin Alberto3, Le Duc Géraldine3, Siegbahn Erik3, Hanson Albert L.4, Miura Michiko4, Slatkin Daniel N. 4
Affiliations: 1Institute of Pathology, University of Bern, Switzerland, 2Freie Universität Berlin, and Animal Oncology and Imaging Center, Switzerland, 3Medical Beamline, European Synchrotron Radiation Facility, Grenoble, France 4Brookhaven National Laboratory, Upton, New York; USA Keywords: MRT, treatment planning, dosimetry.
Rationale and objectives:
Using established radiosurgical techniques, acute and delayed radiobiological responses of tissues are predictably dependent on the quality, the rate, and the density of ionisation energy imparted to them (i.e., radiation 'doses' and 'dose' rates). Although different kinds of tissue display various early and late responses to a given irradiation, whether the imparted dose be uniform or not, normal tissue reactions are predicted reliably for clinical radio-oncology by analysing overlays of dose distributions on serial tomographic images of the irradiated volume.
Treatment planning for MRT will be more complex. Microscopically contiguous cells in the same tissue may be exposed to drastically different doses. Unfamiliar tissue responses may be elicited by various microscopic dose distributions, depending on the organization of the tissue, its milieu intérieur, and its microvasculature. It is natural that we concentrate on the most obvious factors first. Radiation dose-response curves for cells grown in vitro have yielded relative biological effectivenesses (RBEs) applicable to broad-beam radiation therapies, including hadron and pion therapy. Since microscopically contiguous cells in the same tissue may be exposed to drastically different doses from microbeams, RBE data for MRT are needed more for vascularized normal vital animal tissues in vivo than for colonies of non-vascularized cells in vitro. Established treatment planning systems have to be checked on a regular basis by physical dosimetry. This is done in homogeneous phantoms, preferably with calibrated ionization chambers. While this technique is well established for photon, electron and proton irradiations, for hadron therapy absolute calibration, demanding and imprecise, remains elusive. For MRT, dose calibration is even more difficult, as precise physical microdosimetry is still under development. Even Monte-Carlo (MC) dose computations (1), so far the method of choice for MRT, should be provided with large error bars. Accordingly, an MRT treatment planning program for large animals will be developed in an iterative process.
Methods:
The most important single parameter responsible for tissue response to microbeams is believed to be the ‘valley’ dose. The valley dose will be calculated by MC calculation for microbeam exposures in a geometrical phantom, built up by solid tissue substitutes. The resulting dose distribution will be measured with radio-chromic films and at some locations with ionization chambers, to establish the relationship between calculated and measured dose.
For treatment of animal patients the valley dose is calculated in a simplified geometry, but taking into account volumes of significantly higher or lower x-ray absorption as bone or air. The calculated doses will be checked in a parallel plate phantom setup imitating the volume to be irradiated. If the starting dose for the dose escalation program is selected conservatively normal tissue damage will be avoided.
Results / Conclusion: The proposed course of action aims at making full use of the existing results from rodent and pig irradiations (2,3) and at the same time moving from the predominantly used exposures of around 28 µm beam width and 200 µm beam separation of small irradiation volumes to 50 µm beam width and 400 µm separation in significantly larger volumes.
References:
- Stepanek J et al, Physics study of microbeam radiation therapy with PSI-version of Monte Carlo code GEANT as a new computational tool. Med Phys 2000; 27: 1664-1675.
- Laissue JA et al, Microbeam radiation therapy. Proceedings of SPIE: 1999; 3770: 38-45.
- Laissue JA et al, The weanling piglet cerebellum: a surrogate for tolerance to MRT (microbeam radiation therapy) in pediatric neuro-oncology. Proceedings of SPIE 2001; 4508: 65-73.
High tolerance of the rat spinal cord to microplanar irradiation.
Laissue JA, Blattmann H, Bräuer-Krisch E, Bravin A, Dalléry D, Hanson AL, Hopewell JW, Kaser-Hotz B, Keyriläinen J, Laissue P, Le Duc G, Slatkin DN, Siegbahn E, Miura M.
Abstract of the presentation by Jean A. Laissue at the 'Syrad' workshop: New prospects for brain tumour radiotherapy: Synchrotron light and Microbeam Radiation Therapy, Grenoble, France, June 2-4, 2008. http://www.esrf.eu/events/conferences/Syrad/DetailedProgramme
Rationale and objectives
Several radio-oncologists familiar with animal studies of microbeam radiation therapy (MRT) have suggested that this technique might enable the palliation of central nervous system malignancies in infants and young children that cannot, at present, be safely palliated by existing radiation therapies1. The main objective of the present study was to determine the radiation dose required to induce delayed foreleg myeloparesis in 50% of initially normal rats (ED50) following multiple transverse irradiations of an ~1 cm-long segment of their cervico-thoracic spinal cords.
Methods
Microplanar irradiation: The segment including C6 to T2 of the spinal cord of anesthetized, prone, young adult male rats (SPF Fischer, body weight 220-260g) was irradiated at the MRT facility of ID 17 (ESRF, Grenoble) laterally, anatomically from right to left, by a ~10.6 mm-wide array consisting of 52 microplanar beams ~35 µm-wide, 20 mm-high, spaced at ~210 µm intervals. The entrance doses were approximately 330, 470, 660, and ~940 Gy. There were 8 to 10 rats per dose group; 4 rats were sham-irradiated controls.
Seamless synchrotron X-ray beam: Young female rats were also irradiated with synchrotron X rays, in the same facility, in the same prone position, but by a seamless collimated beam, 1.35 mm-wide and 2.5 cm-high, with a 16 mm-thick Al filtration. The antero-lateral border of the radiation field was marked by a vertical line drawn through a virtual point situated at ~2 cm posteriorly to the incisura intertragica. The entrance doses were approximately: 80, 160, 200, 250, 310 and 620 Gy.
The rats were closely monitored for impaired foreleg function by a colleague (DD) experienced in animal husbandry. Rats displaying signs of paresis/paralysis of both forelegs were killed using a standard regulated procedure. All remaining other rats were killed ~1 year after irradiation and the cervico-thoracic spinal cords processed for histopathology.
Results
Paralysis was accompanied by a loss in body weight in the 940 Gy and 660 Gy dose groups (microplanar irradiation) , with all rats in these two subgroups killed by day 56 and day 60 after irradiation, respectively. With the exception of two rats in the 470 Gy dose group that had to be killed at 289 and 311 days (~41 and ~44 weeks) after irradiation, none of those and none in the ~330 Gy group ever showed foreleg dysfunction. In the latter two groups changes in body weight were similar to those of control rats. After synchrotron Xray irradiation in the seamless mode, no rat survived without paralysis beyond 2, 3, 11 or 156 days (~22 weeks) after entrance doses of ~620, 310, 250 or 200 Gy, respectively. Sagittal histological sections of the spinal cord of rats killed from 52 to 311 days after microplanar irradiation displayed areas of white matter necrosis, often associated with fibrinoid vascular necrosis, all within the microbeam array. The striped microplanar irradiation paths were roughly 40 µm wide, separated by on-center distances of ~190 µm.
Conclusions
A rough preliminary visual interpolation of the data yielded an entrance dose ED50 of ~530 Gy for foreleg paralysis after 52 simultaneous parallel microplanar synchrotron X-ray beam irradiations (MRT) over an ~10.6 mm-long cervico-thoracic cord segment, about three times the entrance dose ED50, ~180 Gy after one 1.35 mm-wide seamless synchrotron X-ray beam irradiation. Conversely, seamless or broad-beam irradiation of an 8 mm-long cervical cord segment with a single dose of 250kV X- rays yielded an entrance dose ED50 of ~30 Gy in >30 weeks; the ED50 increased to ~51 Gy when the irradiated segment was only 4 mm-long 2.
Marked histopathologic lesions have been found in the spinal cord of all rats irradiated using the MRT mode and killed following the development of clinical signs of paralysis. The presence of conspicuous vascular lesions such as necrosis, recent and older hemorrhage, involving not only tissue slices irradiated with peak doses, is compatible with the notion of a primarily endothelial pathogenesis of central nervous system (CNS) damage caused by ionizing radiation that may result in a CNS radiation syndrome 3, or in white matter necrosis 4. The post-irradiation latency of clinical signs was ~ 8 weeks for both high dose groups, i.e., 940 and 660 Gy, and ~ 10 months for two rats in the 470 Gy subgroup. This indicates the need for caution against the early assessment of radiation damage or its absence, particularly in the CNS. Thus, no damage was observed to the brain vasculature within one month after a 1000 Gy entrance irradiation dose using the MRT mode 4. Dilmanian et al 5 exposed the spinal cord of few rats, transversely, to 400 Gy from four 0.68 mm-wide microbeams, spaced 4 mm apart on center. None of these rats showed paralysis or behavioral changes during their 7 month observation period. The dose delivered to tissue slices between the microplanar beams, the “valley dose”, is likely to be of prime importance for the extent of tissue damage, as a spatially fractionated array should biologically result in a broad beam effect when the valley dose exceeds threshold values for normal tissue damage. Conceivably, the valley dose might also alter the ED50 of the peak dose.
The reasons for the remarkable relative tolerance of the CNS for MRT-type irradiations may be related to a small-volume effect due to the presence of a huge interface between highly irradiated tissue slices and underirradiated interjacent contiguous tissues. Even four spatially fractionated macroscopic “minibeams”, 0.68 mm thick, widely (1.36 mm) spaced on center, with an entrance dose of 170 Gy, impacting transversely on a normal rat spinal cord, may be tolerated for 7 months5. Such findings suggest potential applicabilities of microbeams – and/or perhaps of minibeams – to clinical palliation of tumors.
Treatment of spontaneous tumors in pet animals as part of the development of a new radiation treatment modality.
Kaser-Hotz B, Blattmann H, Laissue JA, Stepanek J, Bräuer-Krisch E, Bravin A, Le Duc G, Siegbahn E, Dilmanian A, Hanson AL, Miura M, Slatkin DN.
Abstract of the presentation by Barbara Kaser-Hotz at the 'Syrad' workshop: New prospects for brain tumour radiotherapy: Synchrotron light and Microbeam Radiation Therapy, Grenoble, France, June 2-4, 2008. http://www.esrf.eu/events/conferences/Syrad/DetailedProgramme
Rationale and objectives
Treatment of spontaneous (autochthonous) benign and malignant tumors in animal patients is considered a major step between experiments on laboratory animals and humans. The dimensional and physiological characteristics of spontaneous tumors of dogs and cats have more similarity to many human malignancies than implanted tumors of mice and rats. The biological response of tumors and normal tissues is dependent on the volume irradiated.
For MRT the radiation quality, i.e. peak to valley dose ratio (PVDR) is also volume dependent. Obviously, treatment of pet animals involves a more heterogeneous treatment population and smaller numbers of animals can be included into a study. However, the model is more realistic and a closer follow up done by the owners can be done. An important aspect for the implementation of a new radio-oncology modality is the testing of all practical procedures, from treatment planning to follow up care.
Methods
Animal patients eligible are: Animals with a) small, superficial skin or subcutaneous tumors b) superficial benign or malignant tumors of the central nervous system d) other neoplasms to be considered individually. Six animals per group should suffice for a preliminary evaluation of normal tissue tolerance and sensitivity to MRT. A dose escalation will be performed, starting at a conservative dose, expected to produce no significant side effects. After an observation period of at least 6 months the dose will be escalated in small steps to determine the optimal dose.
Results
In the past, the treatment of dog patients with protons at Paul Scherer Institute, Villigen has contributed to the establishment of routine human patient treatment procedures. It has further shown that the spot scanning technique for protons, developed at PSI, was safe and did not lead to any unexpected biological response. Conclusion
The treatment of spontaneous animal tumors can be to the benefit of the animal treated and at the same time give valuable information for a safe start of a human patient program.
MRT-planning (similarities and differences with and between planning for therapy with photons, hadrons and MR).
Blattmann H, Kaser-Hotz B, Laissue JA, Rohrer Bley C, Stepanek J, Bräuer-Krisch E, Bravin A, Le Duc G, Siegbahn E, Hanson AL, Miura M, Slatkin DN.
Rationale and objectives:
Using established radiosurgical techniques, acute and delayed radiobiological responses of tissues are predictably dependent on the quality, the rate, and the density of ionisation energy imparted to them (i.e., radiation 'doses' and 'dose' rates). Although different kinds of tissue display various early and late responses to a given irradiation, whether the imparted dose be uniform or not, normal tissue reactions are predicted reliably for clinical radio-oncology by analysing overlays of dose distributions on serial tomographic images of the irradiated volume.
Treatment planning for MRT will be more complex. Microscopically contiguous cells in the same tissue may be exposed to drastically different doses. Unfamiliar tissue responses may be elicited by various microscopic dose distributions, depending on the organization of the tissue, its milieu intérieur, and its microvasculature. It is natural that we concentrate on the most obvious factors first. Radiation dose-response curves for cells grown in vitro have yielded relative biological effectivenesses (RBEs) applicable to broad-beam radiation therapies, including hadron and pion therapy. Since microscopically contiguous cells in the same tissue may be exposed to drastically different doses from microbeams, RBE data for MRT are needed more for vascularized normal vital animal tissues in vivo than for colonies of non-vascularized cells in vitro.
Established treatment planning systems have to be checked on a regular basis by physical dosimetry. This is done in homogeneous phantoms, preferably with calibrated ionization chambers. While this technique is well established for photon, electron and proton irradiations, for hadron therapy absolute calibration, demanding and imprecise, remains elusive. For MRT, dose calibration is even more difficult, as precise physical microdosimetry is still under development. Even Monte-Carlo (MC) dose computations (1), so far the method of choice for MRT, should be provided with large error bars. Accordingly, an MRT treatment planning program for large animals will be developed in an iterative process.
Methods:
The most important single parameter responsible for tissue response to microbeams is believed to be the ‘valley’ dose. The valley dose will be calculated by MC calculation for microbeam exposures in a geometrical phantom, built up by solid tissue substitutes. The resulting dose distribution will be measured with radio-chromic films and at some locations with ionization chambers, to establish the relationship between calculated and measured dose.
For treatment of animal patients the valley dose is calculated in a simplified geometry, but taking into account volumes of significantly higher or lower x-ray absorption as bone or air. The calculated doses will be checked in a parallel plate phantom setup imitating the volume to be irradiated. If the starting dose for the dose escalation program is selected conservatively normal tissue damage will be avoided.
Results / Conclusion:
The proposed course of action aims at making full use of the existing results from rodent and pig irradiations (2,3) and at the same time moving from the predominantly used exposures of around 28 µm beam width and 200 µm beam separation of small irradiation volumes to 50 µm beam width and 400 µm separation in significantly larger volumes.
References:
- Stepanek J et al, Physics study of microbeam radiation therapy with PSI-version of Monte Carlo code GEANT as a new computational tool. Med Phys 2000; 27: 1664-1675.
- Laissue JA et al, Microbeam radiation therapy. Proceedings of SPIE: 1999; 3770: 38-45.
- Laissue JA et al,The weanling piglet cerebellum: a surrogate for tolerance toMRT (microbeam radiation therapy) in pediatric neuro-oncology.Proceedingsof SPIE 2001; 4508: 65-73.