816 S3 CLINICAL TRIALS OF CONFORMAL TIlERAPY - PHYSICS ASPEcrs Alan Nahum, Joint Ecr artmentof Physics,Instituteof CancerResearch and RoyalMarsdenNHSTrust,SuttonSM25PT,UK [email: ALAN@ICR.AC.UKj If Conformal Therapy (CFRn hadn't existed, physicists would have invented it! So many of the concepts involved are physicist ones: 3-D dose calculation/planning, Beam's-Eye-View, Dose-Volume Histograms, Multileaf Collimators, Computer-Controlled Delivery, Megavoltage Imaging, Optimization, Inverse Planning, Tomotherapy, Biological Modeling, even Protons. All the above developments, many of them involving fairly expensive technology, are on trial. If we wishto be able to use and to continuedevelopment of thesephysicisttools in the future then it has to be demonstrated conclusively that CFRT results in improved clinical outcome. Physicists should therefore be in the front line of planning, executingand evaluatingClinical Trials of Conformal Therapy, by which I mean Randomized, Prospective Phase-III trials i.e. ones that have improved complication-free localcontrol/survival as their endpoint. A prospective, randomized trial to assessthe effectof reducingthe volume of irradiated normal tissue on acute side-effects in pelvic radiotherapy (93% prostateor bladderca.) has beencarriedout at our centre,on 266 patients.In both arms a 3-field,6 MV x-ray technique was used with identical dose prescriptions; in the conventional arm rectangularfields were employed whereas in the conformal arm the fieldswereshapedwith customized blocks drawn according to the Bearn's- Eye-View of the target volume. Substantial differences in normal-tissue volumes (rectum,bladderetc.) wereachieved: mean High-Dose Volume(= PTV)of 690 cm 3 for the conformal technique vs 940 cm 3 conventionally. Comprehensive quality-of- life questionnaire were completed beforethe start of treatment, weekly during and for 3 weeksafter the end of treatmentand then monthly for a further 2 months. A clear pattern of an increase during followed by a decrease after treatment in symptoms relating to bowel and bladder functions was observed for the patient group as a whole. However. a very extensive analysis has not revealed any significantdifferences in symptoms. nor in medication prescribed between the two arms. This result is not consistent with the findings of Soffenet al (1992) and by Vijayakumar et al (1993)butthesewere non-randomized studies. There are many lessons to be learned from the RMT trial e.g, DVH accuracy should be checked beforehand. a standardized symptom-scoring system should also be used (e.g. RTOG). We are about to begin a prospective, randomized trial of dose escalation (64 vs 74 Gy) for prostate cancer and are attempting to broadenthis into a Europe-wide multi-centre trial in whichthe emphasis will be on a sufficientdose difference between the arms rather than on prescribing a given techniquefor all participants. The futureof a great deal of physicsresearchhangs on the outcomeof such clinicaltrials. SS RADIOTHERAPY TO TARGETS WITH CONCAVE SURFACES: PLANNING AND CLINICAL EXECUTION BY INTENSITY MODULATION USING A STATIC-SEGMENTA- TION TECHNIQUE. K. De Jaeger, B. Van Duyse, C. De Wagter, L. Fortan, and W. De Neve, University Hospital Gent, Clinic for Radiotherapy and Nuclear Medicine, Gent, BELGIUM. When the target volume has concavities in its surface, it may be impossible to find "all-in" beam incidences that isolate the target volume from the tissues at risk. In order to obtain a 3D-conformal dose distribution, beam intensity modulation may be required. We have developed a heuristic planning procedure that allows to obtain such dose distributions that selectively exclude those tissues at risk indenting into the target. During the planning process, a number of beam incidences circumscribing the tissues at risk, are defined. Dose homogeneiza- tion at the target is obtained by calculation of intensities given to beam segments of specific, predetermined geometry. This specific geometry is required to maximize the area of each segment and than reduces the number of segments. Treatments were planned using the GRATIS® planning system developed by Sherouse and were executed using the MLC Philips multileaf collimator in a static, sequential mode. Generally, 6-8 beam incidences are required to obtain a 80-85 % dose homogeneity at the target. The number of segments is proportional to the volume of the target. Treatments executed required 30-50 segments. The planning, dosimetry and clinical execution of actually treated cases will be discussed. 54 INn2lSI1Y KlIXIlA'l'ED ARC 1llI!EAPY: A NaJ HE'11lOD FCR IELlVERIN:; cmF<EMAL 'IREAnlENI'S Cedric x. Yu. David A. Jaffray. -John w. am Alvaro Hartini!z Presently, several conformal radiotherapy techniques have been proposed. These include the use of multiple intensity modulated beams, and tomotherapy, where a treatment is delivered slice by slice using a temporally modulated, rotating fan beam. We have implemented a new technique, intensity modulated arc therapy, or IMAT, to deliver tomotherapy treatment plans by combining gantry rotation and dynamic multileaf collimation. Instead of using a slit fan beam to treat a slice of the patient at a time as in tomotherapy, we use the multileaf collimator shaped fields, which changes shape during gantry rotation, to deliver the dose to the treatment target. Arbitrary two-dimensional beam intensities at all the beam angles are first translated into multiple superimposing sub-fields of uniform intensity. The sub-fields at all angles form multiple arcs, which are delivered with dynamic multileaf collimation. A system capable of delivering the IMAT treatments has been implemented and tested. Clinical examples were delivered that illustrated the feasibility of IMAT. Beam delivery time is dependent on the number of arcs needed, which in tum depends on the complexity of the intensity distributions at all the beam angles. With our current implementation, a typical treatment of 250 monitor units takes less than 15 minutes beam time. The technique combines treatment optimization and three- dimensional intensity modulation as with tomotherapy but implemented with existing MLC technology. When a large number of beams are required for a conformal treatment, IMAT is more efficient than the approach using multiple intensity modulated beams .. We believe that IMAT win provide a practical mechanism to achieve complex, highly conformal dose distributions. 56 DYNAMIC AND QUASI-DYNAMIC MULTILEAFCOLLIMATION T. Bortfeld DKFZ, FS05, 1m Neuenheimer Feld 280, Heidelberg, Gennany Several recent investigations deal with the problem of how to pro- duce arbitrary two-dimensional x-ray fluence distributions by means of a multileaf collimator (MLC), an approach, which could be called multileaf modulation. The goal of this approach is to facilitate the delivery of compensated or intensity-modulated fields. The present work gives an overview of these developments. The hardware requirements on MLCs for this special application are specified. Most commercially available MLCs fulfill these requirement sufficiently, however, the MLC control software is generally not capable of controlling an MLC dynamically. There is also the question of how to verify the dynamic movement of the leaves. Some minimum requirements on a control software suitable for application in clinical practice are therefore specified. An alternative, the stepwise or "quasi-dynamic" movement of the MLC-leaves, is also discussed with respect to practicality. In this case the control is easier, but the demands on the stability of the accelerator for small dose deliveries are higher. Nevertheless, it can be expected that, for reasons of ease of control and verification, the quasi-dynamic technique will become the method of choice in the near future, while the slightly more effective fully dynamic technique will become available later in the future. In any case, multileaf modulation is an interesting and important alternative to the tomotherapy-concept.