34 Interim Analysis of Treatment Plans in the FAST TrialY. M. Tsang*, K. Venables*, M. Sydenhamy, on behalf of theFAST Working Party*Mount Vernon Hospital, Northwood, UK; yICR-CTSU, Instituteof Cancer Research, Surrey, UK

Introduction: Faster radiotherapy for breast cancer patients (FAST)is a prospective randomized clinical trial evaluating hypofractio-nation as a treatment option for breast cancer. This qualityassurance (QA) study aims to gather and analyse the radiotherapytreatment plan data from the trial centres, ensuring all centres arecompliant with the trial protocol.Methods and materials: The trial centres were suggested tosubmit their radiotherapy plan data electronically in RTOG orDICOMRT formats. This data together with the informationgathered from radiotherapy QA forms and individual patient casereports, was analysed. The FAST protocol recommended the dosedistribution across the target volume be modified to ensure a dosehomogeneity within International Commission on Radiation Unitsand Measurements (ICRU) guidelines of e5% to +7%. Dose volumehistograms and maximum dose (Dmax) from the plans wereevaluated.Results: 140 plans were collected from 9 hospitals. 77% (108/140)of patients were treated using multi-leaf collimator segments, 16%(22/140) using breast compensators and 7% (10/140) usinga rectangular field in field technique. 89% (125/140) of plans hada Dmax of 107% or less complying with the protocol. To furtherassess the dose distributions, the volume receiving doses between105% and 107% of the prescribed dose was compared betweencentres. For some departments, a much larger proportion ofpatients received significant doses within this range (up to 77% ofplans from one centre). Although this complied with the FASTprotocol, the dose compensation methods used in these centresmay need to be refined if tighter homogeneity criteria are to beapplied.Conclusion: With the dose compensation methods defined in thisstudy, all participating centres were able to achieve dosehomogeneity as defined by the protocol.

35 Conventional and CT-based Volumes in IFRT forSupradiaphragmatic Malignant Lymphoma

P. Dı́ez, P. J. HoskinMount Vernon Hospital, Middlesex, UK

Purpose: The study arose within the Stanford V Trial to assessdifferences between conventional and CT-based treatment volumeand field definition in lymphoma patients undergoing involved-fieldsupradiaphragmatic radiotherapy.Patients and methods: All supradiaphragmatic lymphoma pa-tients CT-planned over the last 2 years were reviewed. Twentypatients were included, 10 cervical and 10 mediastinallymphomas, of whom 4 had hilar involvement. A CTV wasdelineated on CT and grown by 5 mm to create a PTV. This wasfurther expanded by 6 mm to account for penumbra, thusdirectly comparing it to the conventionally placed fields. Thesewere defined on digitally reconstructed radiographs and shield-ing of critical organs drawn. Both techniques were comparedtwo-dimensionally in terms of tumour coverage and shielding ofcritical structures. All critical organ doses were kept withintolerance when planning in 3D.Results: Comparing conventional with CT planning in themediastinum, part of the hilum was missed in all patients.Sections of the lower mediastinum were missed in 9/10 cases. Theupper mediastinum and lower supra-clavicular fossae (SCF) werenot adequately covered in 5/10 patients. 2D planning had greaterlung and heart shielding in most cases. Conventional planning ofthe cervical region undertreated the lower SCF medially in 8/10cases and the upper SCF and neck in 7/10 patients. However,there was a lower spinal cord dose. Laterally, the 2D field was toowide in the upper SCF in 8/10 patients and too long inferiorly inhalf the cases.

Conclusion: Using CT-based volume definition as the standard,conventional treatment fields result in inadequate coverage ofhilum and lower mediastinum in the thorax and medially in theneck and SCF in the cervical region. Conventional fields also tend toextend too far laterally in the upper SCF. CT-based planning allowsmore precise coverage of target volumes and accurate shielding ofcritical organs.

Wednesday 21 March 2007

Teaching Session C: Data Collection And Statistics InRadiotherapy : Size Matters

A. NormanRoyal Marsden NHS Trust, Surrey, UK

As with all clinical research, radiotherapy researchers should,with the help of a statistician, consider the data collected andstatistical aspects of the study at an early stage, i.e. before thestudy begins. Data collection should be prospective and meticu-lously planned and organised. The aim of this presentation will beto inform radiation oncologists of the statistical issues andtechniques involved in radiotherapy clinical research. Studydesign will be discussed, with particular reference to the factorsinfluencing the power of studies; i.e. size of effect vs size of thestudy (number of patients). Methodologies used in study design,data collection and analysis will also be presented with particularattention paid to late radiation induced side effects andradiotherapy planning data. The primary aims and endpoints ofradiotherapy studies are evolving, efficacy is still the mostimportant and pertinent endpoint in radical treatment studies,however there are a number of studies under way where thedelivery of the treatment is being modified and investigated, forexample with hypofractionated schedules. These studies mayhave multiple endpoints and require expert planning. The aimmay be to demonstrate non-inferiority of the new techniquecoupled with controls on acute and late toxicity measures. Theconcept of non-inferiority designs will be explained. The re-cording and analysis of late toxicities is complex, follow up shouldbe taken into account using right censored survival analysistechniques. Methods used to rule out defined levels of toxicitieswill be presented. The conduct of clinical research in Europe andin particular the UK is under increased scrutiny, radiationoncologists should be aware of their legal obligations under thenew GCP standards introduced by EU directive 2005/28 and wherea medicinal product is involved EU directive 2001/20. Theserequirements will be outlined.

Teaching Session: 4D RadiotherapyM. Van Herk

T12 The Role of Partial Breast Radiotherapy in Women with EarlyBreast Cancer

J. YarnoldRoyal Marsden Hospital, Downs Road, Sutton, Surrey, UK

The historical replacement of mastectomy by whole breastradiotherapy brought significant quality of life benefits to womenwith early breast cancer, the majority of whom are now cured. Inwomen treated by breast conservation surgery, whether or notfollowed by radiotherapy, at least three-quarters of ipsilateraltumour relapses present in the vicinity of the primary tumourwithin the first 5e10 years, most of which, on pathological criteria,represent local recurrences of the original primary. The majority ofipsilateral tumour relapses occurring outside the index quadrantand presenting after 10 years post-treatment have pathologicalfeatures consistent with new primary tumours. There is littleempirical evidence to suggest that high dose radiotherapy iseffective in preventing the development of breast cancer. Given

S16 CLINICAL ONCOLOGY