Fast Neutron Beams In Radiotherapy : influence of energy and clinical implications

Les indications cliniques de neutrons rapides, délivrés dans des conditions techniques similaires à celles des photos se sont dégagées à partir d’essais cliniques pluricentriques randomisés. Leurs résultats ont permis de conclure aux avantages de la neutronthérapie, notamment dans le traitement local du cancer avancé des glandes salivaires et de celui de la prostate.
Les neutrons se différencient des photons par leur haute densité d’ionisation ou, ce qui revient au même, par leur valeur élevée de Transfert Linéique d’Energie (TEL). Si dans la gamme des faibles TEL propres aux rayonnements conventionnels (photons, électrons), on observe que peu de variation de la réponse biologique en fonction de la qualité des faisceaux, il n’en est pas de même dans la gamme des hauts TEL où la réponse est susceptible de varier significativement selon les énergies. Il en résulte que les doses de neutrons devront être pondérées par un paramètre biologique qui tient compte des différences d’efficacité en fonction de l’énergie effective des neutrons. C’est dans ce contexte que s’inscrit notre étude.
En utilisant deux systèmes biologiques, l’un, Vicia faba, système végétal répondant à des faibles doses uniques, l’autre, le jéjunum de la souris, répondant à des doses uniques élevées, une différence de l’effet biologique exprimée par l’Efficacité Biologique Relative (EBR) atteint 50% dans une large gamme d’énergie de neutrons utilisées en clinique. La variation de l’accroissement de l’EBR en fonction de l’énergie décroissante des neutrons est plus importante avec le système recourant à des faibles doses uniques. Nous avons confirmé cette observation dans une longue expérience in vivo avec le poumon de la souris, tissu à haute capacité de réparation, en utilisant une procédure de multifractionnement et comparant des neutrons d’énergie voisine. Une différence de l’EBR de 20% a été objectivée pour les doses par fraction de l’ordre de celles utilisées en thérapie, alors qu’avec une dose unique pour ce même tissu pulmonaire la différence n’était que de 5%. L’Efficacité Biologique Relative d’un faisceau de qualité donnée par rapport à un faisceau de référence se définit par un rapport de doses de sorte que toute variation de l’EBR se traduit par une variation similaire des doses. Il est donc indispensable dans la conduite d’une étude pluricentrique impliquant des faisceaux de neutrons d’énergie différente de doubler les comparaisons dosimétriques avec une comparaison radiobiologique

The place of radiotherapy on the arsenal of cancer treatment is nowadays well established. It is based on the use of high energy photons produced by cobalt-60 units or linear accelerators and of electrons produced from the latter. Its indications and limitations have been defined during the last decades.
In addition to the developments that made it possible to overcome most of the technical and physical impediments of dose delivery, other improvement sin radiotherapy have emerged. The radiobiological observations accumulated during the last three decades allowed to distinguish the different according to the type of tissues and the irradiation modalities. Neutrontherapy has been developed in this framework.
Improvement of radiotherapeutic treatments is first of all based on seeking better physical and technical conditions to achieve optimal and precise dose distributions (physical precision). On the other hand it is based on searching for treatment modalities which for a same absorbed dose would enhance the radiation effects on the tumour and reduce the effects on normal tissues (differential effect).
1. Physical selectivity
Physical precision results from the combination of technology and physics. For instance, a combination of new imaging techniques (i.e. nuclear magnetic resonance and computed tomography …) with a radiation therapy simulator enables a better delimitation of target volume in the patient. In addition, devices especially designed to immobilize the patient in treatment position guarantee the reproducibility of the set-up. In this respect one can take advantage of dosimetry and computerizing to innovate in planning complex treatments. So, the conformal therapy is an example of external beam radiotherapy in which the high-dose volume is made to conform closely to the target volume [ Tait, Nahum, 1990]. The ultimate development of it is the radiosurgery which using a stereotaxic frame to immobilize the patient’s head can treat very small intracranial tumours (less than 1 cm³ to 10 cm³) [Larson et al., 1990].
Another approach is the use of proton beams by which an optimum dose distribution can be achieved (Bragg peak). The proton depth dose curves can be flat at ≈ 100% over the depth of interest while the dose behind the target volume is reduced to virtually zera [Suit, 1990]. Obviously, the use of protons requires the same technical conditions as those needed for conformal radiotherapy. However, the improvement of the physical precision cannot resolve the problem of geographically mixed malignant and benign cells. Therefore, the search for other modalities aiming at destroying more selectively tumoral cells than normal cells is another approach to improve the therapeutic gain.
2. Biological selectivity
It is well known that the radiation response of tissues, at doses relevant to radiotherapy, is related to cell death, defined as the irreversible loss of reproductive capacity. In this regard the problem of differential effect can be analyzed through the cell survival curves which express the variation of cell lethality as a function of dose. Different mathematic models have been proposed to characterize these cell survival curves.
The linear-quadratic model is nowadays the most widely used; it described easily the shape of survival curves of mammalian cells exposed to photon radiation [Tubiana et al, 1986; Thames, Hendry, 1987]. In the linear-quadratic model, the expression for the cell survival curve is S= e-(αD+βD²) where S is the fraction of cells surviving at a dose D and α and β are constants representing direct lethal events and sublethal events respectively. The ratio α/β corresponds to the dose at which the two components of cell killing are equal; it is an index of their relative importance. The ratio is high when the survival curve is almost exponential and smaller when the survival curve exhibits a shoulder in its initial part. The ration α/β is higher for normal tissues responsible for early effects and tumours than for healthy tissues responsible for late effects; which indicates that the shape (i.e. the shoulder) of the survival curves for both types of biological tissues is different.
the first mean to obtain an advantages differential effect has been the fractionation of the dose. The differential effect obtained in this way results from a difference in repair capacity between some healthy and some tumoral tissues, as reflected by the difference in the shapes of the initial parts (at the level of 2 Gy) of the survival curves. Cell killings result either from direct lethal events or from accumulation of repairable sublethal damages. The repair of sublethal damages takes place immediately after irradiation and is competed within a few hours (Elkind recovery) [Elkind, Sutton, 1960]. The relative proportion of both lethal mechanisms (which determines the initial part of the curve) will depend on several factors as well related to the intrinsic radio sensitivity as to the radiation modalities (e.g. radiation quality, dose level …). Regarding photon irradiations, the classical treatment regimen (2 Gy / fraction over 5-6 weeks is justified by the fact that the survival curves in this dose range general exhibit the largest advantageous differences inc ell killing rate between tumoral and normal cell lines. On the other hand, the influence of tumour proliferation in an overall time of 5-6 weeks was assumed until now to be negligible although more recently this has been questioned.
The optimization of the differential effect resulting from dose fractionation as been sought in “hyperfractionation” which provides a better sparing of some healthy tissues such as those responsible for late effects (e.g. lung, kidney, spinal cord …); which enables to increase the total dose [Thames, Hendry, 1987; Withers, Horiot, 1988]. Practically, this technique consists of larger number of smaller fractions of more or less 1.2 Gy, distributed into an overall time nearly the same as with classical fraction sizes of 2 Gy per day [Horiot et al., 1988]. As the time of proliferation onset can be some days after the start of treatment in some situations [Fowler, 1986; Peters, Ang, 1988], another modality if hyperfractionation consists of reducing the overall time in order to circumvent tumour proliferation during therapy course (“accelerated fractionation”) [Dische, 1990].
As the presence of malignant hypoxic cells has been recognized as limiting the tumour cure by radiotherapy, another way to enhance the differential effect can be found in the combination of radiation with hypoxic cell sensitizer drugs. However, the clinical use of the misonidazole leads to severe neurotoxic effects [Hall, 1987]. But analogues show reduced uptake un neural tissue and appear to less neurotoxic. Phase III studies are still in progress or data are not yet sufficiently mature for evaluation.
Hyperthermia combined with radiation might be another way [Field, 1989]. The vasculature and blood supply in tumours is generally different and less well-organized than that of normal tissues. As a matter of fact, some tumours (or regions within tumour) will have poorer cooling capacity than normal tissues and may thus be more sensitive to the heating. Moreover, hyperthermia acting to poorly vascularized parts of tumours and thus to less oxygenated cells can further enhance the radiation effect. Although first results obtained for superficial tumours are encouraging, the heating of deep-seated tumours is still clinically unachievable [Tubiana et al, 1986].
A fascination research way fir improving the differential effect is the use of other radiation qualities [Raju, 1990]. In this regard, particles such as carbon, neon or argon (so called “heavy particles” by the radiation oncologist) and neutrons have specific radiobiological properties related to their ionization density. Therefore those having a high ionization density (e.g. neutron, carbon, neon, argon) are called “high LET ‘Linear Energy Transfer) radiations” in contrast with conventional “low LET radiations” (such as gamma, electron, proton, helium. Linear Energy Transfer (LET) property will be presented more extensively subsequently.
Our work takes part in the radiobiological study of “high LET radiations” and more especially of clinical neutron beams having different energies