Division Medical Physics (E0400) - dkfz.de · with IMRT based on our planning system KonRad....

240

Research ProgramRadiological Diagnostics and Therapy

DivisionDivision Medical Physics

DKFZ 2001: Research Report 1999/2000

Scientists:PD Dr. Thomas BortfeldDr. Rolf BendlDr. Jürgen Dams (-5/99)Dr. Barbara Dobler(10/2000-)Dr. Stefan Gölz (3/99-)Dr. Karl Heinz GrosserDr. Thomas HaaseSilvia HandlosPD Dr. Günther HartmannDr. Peter Heeg (12/99-)Dr. Bernd HesseDr. Jörg Hilsbecher (-8/99)Angelika HössDr. Oliver JäkelDr. Christian KargerDr. M.A. Keller-Reichenbecher (-4/00)Dr. Gernot Kuhr (02/00-)Dr. Gunnilla Küster (4/99-9/00)Dr. Sabine LevegrünDr. Andreas Mahr (09/99-)Dr. Uwe OelfkeDr. Mike Partridge (8/00-)Dr. Karsten Pfeiffer (01/00-01/01)Dr. Wilfried Schneider (-08/00)Dr. Lothar Spies (11/99-08/00)Dr. Lan Ton

Postgraduate studentsSimone Barthold Klaus BorkensteinBarbara Dobler (-10/00) Matthias EbertBurkhard Groh Rüdiger Hofmann (11/99-)Werner Korb (08/00-) Gernot Kuhr (-02/00)Gunnilla Küster (-4/99) Christian Lappe (-03/99)Thorsten Liebler (02/00-) Andreas Lüttgau (08/99-)Andreas Mahr (-09/99) Simeon NillLuciana Pavel (12/99-) Karsten Pfeiffer (-01/00)Markus Rheinwald (-12/99) Maria SchererLothar Spies (-11/99) Marc Schneberger (4/00-)Christian Thieke (10/00-)

Graduate studentsMatthias Ebert (04/97-02/98)Florian Föhlisch (8/00-)Andreas Kessen (-06/00) Thorsten Liebler (-01/00)Andreas Lüttgau (-05/99) Marc Schneberger (02/99-02/00) Andreas Schöninger (08/00)Christian Scholz (01/00-) Stefan Wesarg (01/00-)

AssistantsUlrike Bauder-Wüst Karin BeinertJohann Böhm (-06/00) Dietbert BucherJohann Cieslok Klaus Dietz (-06/00)Abolghasem Etemati (12/00-) Gernot EchnerManfred Geißler Udo HaffnerHermann Hamleh Gerd JakobHermann Kohler (-03/00) Siegbert LukschRoland Moschel (-08/00) Wilfried MüllerOtto Pastyr Heinrich Rühle

Susanne Schmitt (05/99-) Steffen SeeberVolker Stamm Wolfram Stroh

TraineesViktor Vöhringer (09/99-)

Civil servantsWolfgang Becken (11/99-08/00)Thomas Claßen (-07/99)Felix Kick (10/00-)Sascha Mechler (10/00-)Christian Pfeifer (11/99-08/00)Kim Pramme (-06/99)

Failure of local tumor control is still a problem in about20% of all cancer patients. Due to this fact, there is an ur-gent need for the optimization of existing and developmentof new and more effective treatment techniques for local-ized tumors. Research at the division of Medical Physics isfocussed on new conformal radiotherapy techniques withphotons, electrons and hadrons. Major achievements ofthe last two decades were the introduction of 3D treatmentplanning, Stereotactic Radiosurgery, 3D Conformal Preci-sion Radiotherapy, inverse treatment planning, IntensityModulated Radiotherapy and Radiotherapy with Car-bon-12 Ions.Besides the ongoing projects in Hadron-therapy, our futurework will be concentrated on establishing mathematicaland biological models of tumour and normal tissue re-sponse and on the consideration of dynamic changes oftarget volumes (PTVs) and organs at risk (OARs) undertherapy, either caused by therapeutic response, by organmovements or by patient repositioning. The approach weare going to develop is called Adaptive Cone BeamTherapy. It will combine conformal dose delivery with on-line imaging of 3D anatomy and on-line monitoring of 3Ddose distributions.Based on the experience and technology from radio-therapy developments, we are also going to develop andinvestigate new methods in Neurosurgery, especially usingshort pulsed lasers and photo dynamic therapy (PDT).The skills of the division cover Computer science, MedicalRadiation Physics, Mathematics, as well as Mechanicaland Electronic Engineering. It is one of the major advan-tage of our division that prototypes of software and hard-ware developments can directly be transferred within thesame building into clinical application in close connectionand cooperation with the Clinical Cooperation Unit Onco-logical Radiotherapy (J.Debus). Thus the division is alsoactive in testing and evaluating the new techniques aswell as in establishing adequate QA programs.

Division Medical Physics (E0400)Head: Prof. Dr. Wolfgang Schlegel

241




Physical Models (E0401)T. Bortfeld, U. Oelfke, B.-M. Hesse, S. Gölz,W. Schneider, M. Partridge, H. Szymanowski,G. Küster, L. Spies, M. Ebert, B. Groh, S. Nill,L. Pavel, C. Thieke, A. Kessen, S. Wesarg, C. ScholzIn collaboration with: Dr. K.-H. Grosser, Dr. R. Bendl, A. Höss,Prof. G. Hartmann, Dr. O. Jäkel, DKFZ, Abt. Medizinische Physik;PD Dr. Dr. J. Debus, DKFZ, Klinische KooperationseinheitStrahlentherapeutische Onkologie; B. Rhein, P. Häring, DKFZ,Zentraler Strahlenschutz / Dosimetrie; Dr. C. Schulze, Dr. J. Stein,MRC Systems, Heidelberg; Prof. S. Webb, Dr. P. Evans, RoyalMarsden Hospital, Sutton, UK; J. Hughes, Siemens OncologyCare Systems (OCS), Concord, USA; Dr. A. Lomax, Paul ScherrerInstitute (PSI), Villigen, Switzerland; Dr. B. Mijnheer, The Nether-lands Cancer Institute (NKI), Radiotherapy Department,Amsterdam, The Netherlands Cancer; Prof. H. Kooy, Massachu-setts General Hospital, Dept. of Radiation Oncology, Boston,USA; Prof. F. Nüsslin, M. Alber, Universität Tübingen, Institut fürMedizinische Physik; Dr. D. Jaffray, Department of Radiation On-cology, William Beaumont Hospital, Royal Oak, MI USA; Prof. M.Kröning, Dr. M. Maisl, Fraunhofer Institut für ZerstörungsfreiePrüfverfahren (IZFP), Saarbrücken; Prof. A. K. Louis, Institut fürAngewandte Mathematik, Universität des Saarlandes,Saarbrücken; Prof. H. W. Hamacher, PD Dr. K.-H. Küfer, Institutfür Techno- und Wirtschaftsmathematik, Kaiserslautern; Prof. R.Männer, PD Dr. J. Hesser, Universität Mannheim, Lehrstuhl fürInformatik V

Our objective is the improvement of tumor therapy throughthe development and application of mathematical andphysical models. We focus on the optimization of radio-therapy, particularly on so-called intensity-modulated ra-diotherapy (IMRT) [1,6,12]. Building upon our efforts in thelast few years it has been possible to treat more than 140patients at DKFZ and at the Radiological University Clinicswith IMRT based on our planning system KonRad.Thereby the patients could be treated with higher doses inthe tumor target volume and/or with better sparing of thesurrounding normal tissues. Worldwide over 1000 patientshave been successfully treated with IMRT on the basis ofour optimization programs.

In one project of our working group we investigate the ap-plication of the IMRT concept to irradiation with chargedparticles, in particular with protons [5,15] (IMPT, intensity-modulated proton therapy) and with heavier ions. Themethods of inverse therapy planning for photons havebeen transferred to charged particles and been extendedby one dimension in order to optimize not only the intensi-ties but also the energies of the used particles [9,10]. Par-allel to the theoretical investigations of new irradiationtechniques with charged particles (e.g. “Distal Edge Track-ing”) for simple model cases, we have developed andimplemented a new modular planning platform for theIMPT. At present the new multi-modality components ofthe KonRad program are being evaluated (in particular theoptimization module and the dose calculation). This coversseparate tests of the optimization and the dose calculationusing phantoms, as well as planning studies for typicalclinical cases using various irradiation techniques. The ap-plication of the new IMPT planning procedures and furtherdevelopments, e.g. regarding biological characteristics ofcharged particles, are examined in co-operation with theleading centers in the field of particle therapy (NTPC Bos-ton, PSI Villigen (Switzerland) and GSI Darmstadt).

a) b)

Fig. 1: Comparison of a photon IMRT dose distribution (a) with aproton IMPT distribution (b) for a prostate case. IMRT allows toachieve good conformation of the high-dose region to the tumortarget volume (red line), but a significant amount of normal tissueis treated with medium dose values (yellow and green areas).IMPT yields an excellent conformal dose distribution at all doselevels.

The development of a multi-criterion optimization conceptfor radiotherapy is planned. The probability of local controland possible complications are evaluated separately andoptimized for individual radiation-sensitive “organs-at-risk”.First estimates suggest that, by taking this approach, bet-ter results can be obtained than with past optimization pro-cedures, in which the entire irradiation plan was evaluatedsimultaneously.

Another important aspect of our work is Monte Carlo simu-lation for radiation therapy. This comprises on the onehand the problem of dose calculation in patients in com-plex irradiation situations [13]. On the other hand, MonteCarlo is used for the simulation of treatment machines andequipment [8]. For this purpose the GEANT program fromCERN has been implemented.

Equally important as achieving a highly conformal dose tothe tumor is the verification side of therapy, since smallchanges in the region of interest might have an impact ontumor control and normal tissue complication probabilities.Our investigations have shown that anatomical sectionalviews of the patient produced with the therapy beam(MVCT) can contain all the necessary information for acomplete verification. For these investigations we useamorphous silicon (a:Si) flat panel detectors. The verifica-tion process consists of two parts: CT imaging and thedose reconstruction (transit dosimetry) based on these CTimages. Problems arising from mechanical instabilities ofthe detector attached to the rotating linac gantry havebeen solved using an off-line correction [16]. A furtherproblem in transit dosimetry is scattered radiation pro-duced as the beam traverses the patient, which overlaysand degrades the image. By means of Monte Carlo calcu-lations we have shown that scattered photons, which aremainly in the low energy region, can give rise to a dispro-portionately large signal in a:Si flat panel detectors[11,14,17,18]. Studies have been carried out to optimisethe composition of the converter plate (metal/scintillator)such that the detector becomes less sensitive to these lowenergies.

A further aspect of our imaging research has been the de-velopment and implementation of a 3D algorithm for imagereconstruction [2,3] using cone beam geometry. Thus thedistribution of the electron density over the entire treated

242




volume can be reconstructed with only one rotation of thegantry. At present it is being investigated whether recon-structions are also possible from a reduced range of gan-try angles allowing for faster examinations with a smallerdose to the patient. In collaboration with the University ofMannheim a prototype of a dedicated reconstruction hard-ware is designed to achieve online 3D reconstruction. Theobjective of this online imaging technique is to detect er-rors in the positioning of the patient and organ move-ments. Such knowledge of organ motion and positioningerrors are also currently being studied and the expecteddosimetric effects quantified. This work forms the basis ofa plan to include these errors in the optimization processand to produce strategies for adaptive radiotherapy.

In an industrial collaboration the efficiency of each singlestep of the IMRT procedure was investigated to make atransfer of the IMRT technique into the clinical practice ofmany hospitals possible [4]. To achieve this goal a database was set up to investigate the possible use of classsolutions for specific clinical situations. Furthermore, thetime effort for therapy planning, verification and documen-tation was reduced by improving existing techniques aswell as implementing new devices and tools. As a laststep, workshops and site visits were held to train physi-cians, physicists and technicians in the IMRT method.Grants: Tumorzentrum Heidelberg/Mannheim: 1 position (BAT Ib);Industrial collaboration Siemens: 1 position (BAT IIa) ;Strategiefonds Helmholtzgemeinschaft: 2 Post-Doc positions (BATIIa); DFG: 1 position (BAT IIa), 2 Ph.D. positions

Selected Publications (* = external co-author)[1] T. Bortfeld, Optimized planning using physical objectives andconstraints. Seminars in Radiation Oncology, vol. 9, pp. 20-34,1999.[2] T. Bortfeld and U. Oelfke, CT-reconstruction from fan data us-ing parallel backprojection. Med. Phys. vol. 26, pp. 2036, 1999.[3] T. Bortfeld and U. Oelfke, Fast and exact 2D image reconstruc-tion by means of Chebyshev decomposition and backprojection.Phys. Med. Biol. vol. 44, pp. 1105-1120, 1999.[4] M.-A. Keller-Reichenbecher, T. Bortfeld, S. Levegrün, J. Stein,K. Preiser, and W. Schlegel. Intensity modulation with the “stepand shoot” technique using a commercial mlc: planning study Int.J. Radiat. Oncol. Biol. Phys. vol. 45, pp. 1315-1324, 1999.[5] A. J. Lomax*, T. Bortfeld, G. Goitein*, J. Debus, C. Dykstra, P.Tercier*, P. A. Coucke*, and R. O. Mirimanoff*, A treatment plan-ning inter-comparison of proton and intensity modulated photonradiotherapy. Radiother. Oncol. vol. 51, pp. 257-271, 1999.[6] U. Oelfke and T. Bortfeld, Inverse planning for x-ray rotationtherapy: a general solution of the Inverse Problem. Phys. Med.Biol. vol. 44, pp. 1089-1104, 1999.[7] T. Bortfeld, U. Oelfke, and S. Nill, What is the optimum leafwidth of a multileaf collimator? Med. Phys. vol. 27, pp. 2494-2502,2000.[8] G. Küster and T. Bortfeld, Applicability of a multi-hole collima-tor for scanned photon beams: A Monte Carlo study. In: Proc. ofthe XIIIth International Conference on The Use of Computers inRadiation Therapy, eds. W. Schlegel and T. Bortfeld. Heidelberg:Springer, 2000.pp. 179-181.[9] S. Nill, U. Oelfke, and T. Bortfeld. A new planning tool for IMRTtreatment: Implementation and first application for proton beams.In: Proc. of the XIIIth International Conference on The Use ofComputers in Radiation Therapy, eds. W. Schlegel and T.Bortfeld. Heidelberg: Springer, 2000.pp. 326-328.

[10] U. Oelfke and T. Bortfeld, Intensity modulated radiotherapywith charged particle beams: Studies of inverse treatment plan-ning for rotation therapy. Med. Phys. vol. 27, pp. 1246-1257,2000.[11] M. Partridge, B. Groh, L. Spies, B. M. Hesse, and T. Bortfeld,A Study of the Spectral Response of Portal Imaging Detectors.IEEE Medical Imaging, Lyon, October 15-20 (2000).[12] W. Schlegel and T. Bortfeld, The Use of Computers in Radia-tion Therapy eds. W. Schlegel and T. Bortfeld (eds.) pp. 1-604,2000. Springer. Heidelberg.[13] W. Schneider, T. Bortfeld, and W. Schlegel, Correlation be-tween CT numbers and tissue parameters needed for MonteCarlo simulations of clinical dose distributions. Phys. Med. Biol.vol. 45, pp. 459-478, 2000.[14] L. Spies, P. M. Evans*, M. Partridge, V. N. Hansen*, and T.Bortfeld, Direct measurement and analytical modeling of scatter inportal imaging. Med. Phys. vol. 27, pp. 462-471, 2000.[15] A. Zurlo*, A. Lomax*, A. Höss, T. Bortfeld, M. Russo*, G.Goitein*, V. Valentini*, L. Marucci*, R. Capparella*, and A.Loasses*, The role of proton therapy in the treatment of large irra-diation volumes: a comparative planning study of pancreatic andbiliary tumors. Int. J. Radiat. Oncol. Biol. Phys. vol. 48, pp. 277-288, 2000.[16] M. Ebert, B. A. Groh, M. Partridge, B. M. Hesse ,and T.Bortfeld, 3D image guidance in radiotherapy: a feasibility study.SPIE International Symposium on Medical Imaging, San Diego,CA, Feb. 17-22 (2001).[17] L. Spies, M. Ebert, B. A. Groh, B. M. Hesse, and T. Bortfeld,Correction of scatter in megavaltage cone-beam CT. Phys. Med.Biol. 46 821-33 (2001).[18] L. Spies and T. Bortfeld, Analytical scatter kernels for portalimaging at 6 MV, Med. Phys 2001 (in press)

Biological Models (E0402)S. Levegrün, K. Borkenstein, A. Mahr, Lan TonIn cooperation with: PD Dr. Dr. J. Debus, Dr. P. Peschke, KlinischeKooperationseinheit Strahlentherapeutische Onkologie, FSRadiologische Diagnostik und Therapie, DKFZ; Dr. A. Jackson,Dr. C.C. Ling, Department of Medical Physics, Memorial Sloan-Kettering Cancer Center, New York, USA; Dr. M.J. Zelefsky, Dr. Z.Fuks, Dr. S. Leibel, Department of Radiation Oncology, MemorialSloan-Kettering Cancer Center, New York, USA; Dr. M.L. Bahner,Abteilung Onkologische Diagnostik und Therapie, FSRadiologische Diagnostik und Therapie, DKFZ

The aim of any tumor therapy with curative intent consistsin the eradication of the tumor without causing severe sideeffects. In radiation therapy, the dose required to achievelocal tumor control is often compromised by the increasingrisk of complications in surrounding critical organs andhealthy tissues. To optimize radiation therapy, the treat-ment plan leading to the best compromise between hightumor control probability and low risks of adverse effectsmust be found for each individual patient. In the process ofradiotherapy treatment planning, radiation oncologistshave to assess the likely biological effect of a plannedphysical dose distribution, which is a complex task. Atpresent, treatment decisions are entirely based on clinicalexperience of tolerance doses in various normal tissuesand empirical knowledge of volume effects (i.e. the obser-vation that in many organs complication probabilities in-crease when increasing fractional volumes of the organsare irradiated). Therefore, efforts are being made to de-velop new tools to quantitatively evaluate treatment plans.To quantify the biological effect of a delivered dose distri-

243




bution, biological models have been proposed that attemptto describe the response of tumors and normal tissues toirradiation. The ultimate goal of these models is to reliablypredict tumor control probability (TCP) and normal tissuecomplication probabilities (NTCP), based on the physicaldose distribution and radiobiological parameters. Theprojects of the working group focus on the investigation ofvolume effects, the quantification of dose-response rela-tions and the evaluation of current biological models astools for treatment planning.

The aim of one investigation of the working group was toevaluate the usefulness of existing models to predict nor-mal tissue complication probabilities after radiosurgery ofpatients with cerebral arteriovenous malformations (AVM).The study concentrated on radiation-induced tissuechanges on follow-up neuroimaging (e.g. edema, blood-brain-barrier breakdown, necrosis). Outcomes for 286 AVMpatients who received stereotactic Linac radiosurgery atDKFZ were retrospectively analyzed and compared to thepredictions of existing models [1-3]. In addition, dose-re-sponse relations were investigated. The comparisons indi-cated that risk prediction models may allow the identifica-tion of patients with a high probability of adverse effects.However, further evaluations of the merit of current modelsas tools to guide treatment planning for lesions in the brainare required.

In cooperation with the Departments of Medical Physcisand Radiation Oncology at Memorial Sloan-Kettering Can-cer Center (MSKCC), New York, late radiation effects inthe rectum [4-6] and local tumor control [7,8] after three-di-mensional conformal radiation therapy of prostate cancerpatients were retrospectively analyzed. Since December1988, more than 1200 patients with stage T1c-T3 clinicallylocalized adenocarcinoma of the prostate were enrolled ina phase I dose-escalation trial at MSKCC. Patients weretreated with conventional photon beam 3D-CRT and inten-sity-modulated treatment techniques to prescribed dosesbetween 64.8 and 86 Gy. To assess local tumor control, asubset of the patients underwent a post-treatment prostatebiopsy at least 2,5 years after end of radiation therapy.Variables derived from the dose distribution of individualpatients and tumor-related prognostic factors that corre-lated with biopsy outcome were identified. The usefulnessof several existing TCP models as predictors of biopsy out-come was evaluated and a multivariate model to predictbiopsy outcome was derived. The results may help toidentify patients, based on their individual pre-treatmentprognostic factors, that would benefit most from dose-es-calation and to guide dose prescription.

In 1999, the working group started a new project, the mod-eling and computer simulation of tumor growth and of tu-mor response to various treatment modalities [9]. The goalof the project is to integrate simulated tumor control prob-abilities into the treatment planning process to assist ra-diation oncologists in their decision making. Creating real-istic models requires a thorough understanding of the biol-ogy of tumor cell growth and tumor response to differenttreatments such as radiation therapy. For this reason, themodeling is done in close cooperation with the Radiobiol-

ogy Group of the Division of Radiation Oncology (KlinischeKooperationseinheit Strahlentherapeutische Onkologie) atDKFZ. The developed model takes into account individualtumor cells. This allows to include a variety of cell biologi-cal parameters. Tumor growth is modeled using basicmechanisms for cell growth and death: Cell cycle phasesand their duration, control mechanisms for inactive cellsand apoptosis, oxygen supply and resorption of necroticand apoptotic cells. At the beginning of tumor growth themicrovasculature is modeled via a homogeneous capillarydistribution. Hypoxic areas in a tumor’s interior will induceangiogenesis. Modeling of tumor response to radiationtherapy is based on the linear-quadratic model. The radi-osensitivity of each cell depends on its oxygen supplyand its cell cycle phase. Complete repair is assumed be-tween fractions. Computer simulation is performed on a3D rigid cubic lattice. Growth is initiated by a single tumorcell dividing after one cell cycle time TC. Randomprocesses such as cell displacement after cell division andDNA-damage are simulated by Monte-Carlo techniques.The TCs and sensitivities of the cells are normallydistributed. Tumor growth and tumor response to radiationcan be simulated for macroscopic tumors with a size of upto 12 mm (see Figures 1 and 2).

Another area of research has been the development andevaluation of a computer system to determine tumor vol-umes within clinically required accuracy as a tool to sup-port therapy decisions, therapy monitoring and the assess-ment of tumor response to treatment [10,11]. In a detailedstudy, different available volumetry systems based onsemiautomatic image segmentation algorithms were com-pared. The best algorithm was re-implemented into a newvolumetry system. The delineated data are then analyzedand compared to values in a local knowledge-base. Thisknowledge-base consists of data gathered from extensivephantom studies. Depending on the size and contrast ofthe tumor, an uncertainty factor is deduced from the know-ledge-base. This factor is combined with a user-dependentuncertainty factor derived through regular delineations oftest data by every user. The combination of these two fac-tors allows to assess the accuracy of the delineation.Hence, the new system will not only provide absolute vol-ume information but also detailed information about the re-liability of the result. Eventually, this system will lead to ahigher clinical acceptance of volumetric information andprovide the basis for the clinical implementation of tumorvolumetry to monitor tumor response to therapy and to in-dividualize treatments.

Publications (* = external co-author)[1] Ton L, Levegrün S, Debus J, Spahn U*, Keller-ReichenbecherMA, Swiderski S*: Modelle zur Abschätzung der Strahlenwir-kungen bei AVM-Patienten. In: Medizinische Physik 99: Hrsg.: H.Gfirtner H, Passau: DGMP (1999) 177-178.[2] Ton L, Levegrün S, Debus J, Swiderski S, Schlegel W: Estima-tion of Complication Probabilities After Radiosurgery of AVM Pa-tients Using a Biological Model. In: Procs XIIIth International Con-ference on the Use of Computers in Radiation Therapy, SchlegelW, Bortfeld T, eds. Heidelberg: Springer, (2000) 246-248.

244




[3] Swiderski S, Zuna I, Thilmann C, Ton L, Bahner ML,Wannenmacher M, Debus J: Einzeitbestrahlung arterio-venöserMalformationen (AVM) im Gehirn: Ergebnisse des DKFZ (1984-1997) und strahlentherapeutisches Scoring-System. Strahlenther.Onkol. 176 Sondernr. 1 (2000) 25.[4] Skwarchuk MW*, Jackson A*, Zelefsky MJ*, Venkatraman ES*,Cowen DM*, Levegrün S, Burman CM*, Fuks Z*, Leibel SA*, LingCC*: Late Rectal Toxicity after Conformal Radiotherapy of Pros-tate Cancer. (I): Multivariate Analysis and Dose-Response. Int JRadiat Oncol Biol Phys 47 (2000) 103-113.[5] Jackson A*, Skwarchuk MW*, Zelefsky MJ*, Cowen DM*,Venkatraman ES*, Levegrün S, Burman CM*, Kutcher GJ*, FuksZ*, Leibel SA*, Ling CC*. Late Rectal Bleeding After ConformalRadiotherapy of Prostate Cancer. (II): Volume Effects and Dose-Volume Histograms. Int J Radiat Oncol Biol Phys 49(3) (2001)685-698.[6] Jackson A*, Skwarchuk MW*, Levegrün S: Volume Effects inExternal Beam Treatments of Prostate Cancer. In: Procs 5th Inter-national Symposium on 3D Conformal Therapy and Brachy-therapy, Amols H, Zelefsky M, eds. New York: Memorial Sloan-Kettering Cancer Center, (2000) 197-199.[7] Levegrün S, Jackson A*, Zelefsky MJ*, Venkatraman ES*,Skwarchuk MW*, Schlegel W, Fuks Z*, Leibel SA*, Ling CC*:Analysis of Biopsy Outcome After Three-Dimensional ConformalRadiation Therapy of Prostate Cancer Using Dose DistributionVariables and Tumor Control Probability Models. Int J RadiatOncol Biol Phys 47(5) (2000) 1245-1260.[8] Levegrün S, Jackson A*, Zelefsky, MJ*, Skwarchuk MW*,Venkatraman ES*, Schlegel W*, Fuks Z*, Leibel SA*, Ling CC*:Pitfalls in Deducing Radiobiological Parameters for Tumors fromClinical Data: A Study of Biopsy Outcome After Three-dimensionalConformal Radiation Therapy (3D-CRT) of Prostate Cancer. Int JRadiat Oncol Biol Phys 48 (3), Suppl. (2000) 136-137.[9] Borkenstein K, Levegrün S, Peschke P, Schlegel W:Modellierung und Computersimulation von Tumorwachstum undTumorantwort auf Strahlentherapie. In: Medizinische Physik 2000,Kneschaurek P, ed. München: DGMP, (2000) 111-112.[10] Mahr A, Levegrün S, Bahner ML, Kress J, Zuna I, SchlegelW: Usability of Semiautomatic Algorithms for Tumor Volume De-termination. Investigative Radiology 34 (1999) 143-150.

Fig 1: Planes through a spherical tumor with 25 Mio. cells beforeradiation. TC=2d. There are necrotic centers in the tumor‘s interior(black), some of which have already been resorbed (white). Thistumor model does not include angiogenesis.

Fig 2: Planes through the tumor of Fig. 1 after 2 weeks of radia-tion with 10 fractions of 2 Gy each. Most necrotic and apoptoticcells have been resorbed.

[11] Mahr A, Bahner ML, Levegrün S, Schlegel W: A New Ap-proach for Improved Tumor Volumetry. In: Procs XIIIth Interna-tional Conference on the Use of Computers in Radiation Therapy,Schlegel W, Bortfeld T, eds. Heidelberg: Springer, (2000) 113-115.

Therapy Planning – Development (E0403)R. Bendl, J. Dams, B. Dobler, M. Kieber, A. Lüttgau,M.A. Keller-Reichenbecher, K. PfeifferIn cooperation with: PD Dr. J. Debus, G. Hartmann, A. Höss, O.Jäkel, A. Lorenz, I. Zuna, DKFZ; Dr. M. van Kampen, Dr. G.Sroka-Pérez, Radiologische Universitätsklinik Heidelberg; Dr.Bonsanto, Dr. A. Staubert, Neurochirurgische UniversitätsklinikHeidelberg; Dr. K. Welker, Dr. K. Zink, M. Scholz, KlinikumMoabit, Berlin; Prof. Dr. V. Sturm, Dr. R. Lehrke, Dr. K. Luyken,Dr. H. Treuer, Klinik für stereotaktische und funktionelleNeurochirurgie, Universität Köln; Prof. Dr. J. F. Bille, Institut fürAngewandte Physik, Universität Heidelberg; Dr. M. Götz, Dr. S.Fischer, MRC Systems, Heidelberg; Dr. H. Fuchs, Dr. H. Kluge,Dr. C. Rethfeld, Hahn-Meitner-Institut, Berlin; Dr. Nausner, Dr.Bechrakis, Universitätsklinikum Benjamin-Franklin, Berlin

The task of the group „Therapy Planning - Development“ isthe investigation, development and implementation ofcomputer assisted tools, which can improve the planning,simulation and evaluation of minimal-invasive treatmenttechniques in oncology. In this way, clinicians should beenabled to plan, test and optimise their treatment strate-gies pre-therapeutically.

There is no doubt, that a pre-therapeutical treatmentoptimisation can increase the overall treatment results:better local tumour control, less side effects and a reduc-tion of treatment time in surgical interventions, faster re-covery of the patient and consequently a reduction oftreatment costs. Due to the close cooperation of physi-cians, physicists and computer scientists our institute isthe ideal environment for developing computer assistedplanning and simulation methods with clinical relevance.One superior intention of all developments is a carefulelaboration of promising methods in a way, that their ben-

245




efits can be demonstrated directly in real patient treatment.The former activities were focused on methods for three-dimensional conformal radiotherapy planning. Becausethere are a lot of similarities in computer assisted tools forplanning and optimising other minimal-invasive treatmenttechniques, we try to extend the available experiences andto generalise new developments in a way that other thera-peutical approaches may derive benefits, too.

The working areas of the group include all links of thetherapeutical chain that can be supported by computer as-sisted tools. Special emphasis is put on image processing,registration of multi-modal image sequences, segmenta-tion, three-dimensional modelling, representation and vi-sual presentation of anatomy, visual simulation of treat-ment approaches and virtual reality, presentation of nu-merical simulation results and appropriate evaluationtools, knowledge based systems for therapy planning andmethods and tools for therapy monitoring. We have con-centrated our activities on the following topics:1. Registration of multi-modal images, segmentation and

representation of anatomical structures2. Knowledge based radiotherapy planning3. Stereotactic laser neurosurgery4. Planning system for the proton therapy of eye tumours5. VIRTUOS, VIRTUal RadiOtherapy Simulator6. New technologies for transferring therapeutical knowl-

edgeThe main strategy can be divided into two sub-goals. Thefirst one is the continuous further development of the exist-ing radiotherapy planning software as a support to theClinical Cooperation Unit „Radiation Oncology“ (PD Dr.Debus, E0500), to enable this group to continue and ex-tend their scientific work. They need a reliable state of theart planning system, to which new functionalities can eas-ily be added. In addition to activities 1, 2 and 5 the furtherdevelopment of the system for its use in the “Heavy IonsTherapy Project” (Dr. Jäkel. GSI/DA-Project E0409)[2] isone of the important tasks.

In activity 2 we have investigated a new approach whichwill not only speed up the generation of three-dimensionaltreatment plans markedly, but will allow the systematic col-lection of well proven treatment strategies [3]. Togetherwith Dr. Welker, Klinikum Berlin-Moabit, radiotherapeuticaltreatment strategies have been collected and will be inte-grated into the knowledge based system in future. Thisway the system will allow a continuous exchange of thera-peutical knowledge between different research andtherapy centres. In activity 6 we develop methods andstrategies which should permit easy distribution of ap-proved treatment strategies via an internet-based informa-tion system.

Besides the development of new methods a central objec-tive of our efforts is to enable other clinicians to use 3Dtherapy planning for routine patient treatment and tospread out planning tools and promising therapy strate-gies. Therefore, commercial cooperations are used to dis-tribute the developed planning tools to other clinics and re-

search centres. In one cooperation our VIRTUOS programwas made ready for the market and is now distributed bySTRYKER-Leibinger, Freiburg.

The second sub-goal is the extension and adaptation ofthe developed planning strategies and tools to new mini-mal-invasive therapy concepts. On one side this means ageneralization of planning concepts. On the other sidethese activities should result in dedicated planning sys-tems, which should enable clinicians to plan, simulate, op-timize and evaluate interventions pre-therapeutically.

In activity 4 we develop together with Dr. H. Fuchs and Dr.H. Kluge, Hahn-Meitner-Institut Berlin a planning systemfor therapy of eye tumours with protons (Fig.: 1). Protontherapy of eye tumours has shown good clinical results(96% local tumour control). It is carried out in somespecialised institutes all around the world. At HMI Berlinthe first therapy facility for that kind of treatment in Ger-many was established in 1998. For planning therapy nor-mally the program EYEPLAN is used (developed by M.Goitein in the early eighties). Since three-dimensional im-age modalities were not widely used at that time, planningand dose calculation is based on a simple spherical modelof the eye. This results in uncertainties, which must becompensated by relatively large security margins. If thedistance of tumour and sensitive structures (optical nerve,macula) is smaller than 3 mm, conventional treatmentscan lead to a loss of functionality of that structures. Theaim of the project is therefore, to establish a precise modelof the eye (by using modern image modalities, CT, MRIand fundus photographs) and this way the precision oftherapy should be increased, to reduce possible side ef-fects (loss of sight). At DKFZ we have developed a real-time dose calculation algorithm, which presents the ex-pected dose distribution synchronously to adjustments ofthe irradiation direction [4]. This way planning time shouldbe reduced noticeable. To increase the precision of pre-calculated dose distributions an additional pencil beam al-gorithm will be integrated by the partners in Berlin.

Fig. 1: Mapping of a fundus photograph on the three-dimensionalmodel of the eye, which was established based on a sequence ofCT images. Visible structures: Cornea, Lens Optical Nerve, Tu-mour and Clips, which were sewn on the sclera for controlling thetreatment position

246




An additional promising approach is investigated in activity3 (stereotactic laser neurosurgery) together with Prof.Bille, Institute for Applied Physics, University Heidelbergand Prof. Sturm Clinic für stereotactic Neurosurgery, Uni-versity Cologne and MRC Systems Heidelberg. By meansof an ultra short pulsed laser deep-seated brain tumoursshould be resected without diffuse thermal damage of thesurrounding tissue [1]. The application of such interven-tions will depend on the availability of suitable planningand monitoring tools. Therefore we develop a dedicatedplanning system for that new kind of therapy. Meanwhile afirst prototype was installed at the clinical cooperation part-ners. In collaboration with the future users that prototypewas evaluated and is now adjusted to the clinical needs.To monitor the application of the short-pulsed laser differ-ent intra-operative imaging modalities were investigated[5]. Since interventional MR scanners are not widely usedup to now, we have investigated how 3D ultrasound canbe used as intra-operative imaging modality [5]. The devel-opment of intra-operative monitoring facilities will have im-portant influence on other surgical treatment concepts,too.Nearly all activities of the group were financed substantially by theDFG (german research community) the Deutsche Krebshilfe andfundings supplied by industrial cooperation partners.

Publications (* = external co-author)[1] Bendl R, Dams J, Suhm N, Lorenz A, Bille JF, Schlegel W: APlanning System for Stereotactic Laser-Neurosurgery. In: LemkeHU, Inamura K, Jaffe CC, Felix R (eds.) Computer Assisted Radi-ology. Proceedings of the International Symposium CAR 97,Elsevier Science: 778 - 783[2] Jäkel O, Krämer M, Bendl R, Hartmann G, *Kraft G:Bestrahlungsplanung für Schwerionen. In: Voigtmann L undGeyer P (eds.) Med. Physik 1998, DGMP, Dresden: 123 - 124[3] Keller-Reichenbecher MA, *van Kampen M, Bendl R, *Sroka-Pérez G, Debus J, Schlegel W A Knowledge-Based System forRapid Generation of Alternative Preoptimized Plans for ConformalRadiotherapy Planning. In: Leavitt D and Starkschall G (eds.) Pro-ceedings of the XIIth International Conference on the Use ofComputers in Radiation Therapy, Salt Lake City (1997): 298 -301[4] K. Pfeiffer, R. Bendl: A real time dose calculation andvisualisation for the proton therapy of ocular tumours. Phys. Med.Biol. 46, issue 3 (2001)[5] Suhm N, Dams J, van Leyen K, Lorenz A, Bendl R: Limitationsfor Three-Dimensional Ultrasound Imaging Through a Bore-HoleTrepanation. Ultrasound in Medicine and Biology 24 5 (1998),:663-671

Therapy Planning - Application (E0404)A. HössIn cooperation with: Prof. Dr. N. Ayache, INRIA, Sophia Antipolis,France; Prof. Dr. M. Bamberg, Dept. for Radiotherapy, Universityof Tübingen; Dr. M.L. Bahner, Div. of Oncological Diagnostics andTherapy, DKFZ; Prof. Dr. H. Blattmann, Paul Scherrer Institute,Villigen, Switzerland; Dr. J. Bohsung, Dept. of Radiotherapy,Charité, Berlin; PD Dr. T. Bortfeld, Dr. R. Bendl, Prof. Dr. G.Hartmann, Div. of Medical Physics, DKFZ; PD Dr. Dr. J. Debus etal., Clinical Cooperation Unit Radiation Oncology, DKFZ; Prof. Dr.R. Felix, PD Dr. P. Wust, Virchow-Klinikum, Berlin; Dr. D.T.L.Jones, Dr. A.N. Schreuder, National Accelerator Centre, Faure,South Africa; Prof. Dr. G. Kraft, Gesellschaft fürSchwerionenforschung mbH, Darmstadt; MRC Systems GmbH,Heidelberg; Prof. Dr. G. Nemeth, Dr. O. Esik, National Institute ofOncology, Budapest, Hungary; Nucletron B.V., Veenendaal, The

Netherlands; Prof. Dr. F. Nüsslin, Dept. of Medical Physics, Uni-versity of Tübingen; B. Rhein, P. Häring, Prof. Dr. L. Schad, Dep.of Biophysics and Medical Radiation Physics, DKFZ; Prof. Dr. J.Richter, Clinics for Radiation Therapy, University of Würzburg; Dr.S. Scheib, Klinik Im Park, Zurich, Switzerland; Prof. Dr. R.Schmidt, Dept. of Radiotherapy, University Hospital Eppendorf,Hamburg; Dr. U. Schneider, Clinics for Radiooncology andNuclear Medicine, Zurich, Switzerland; Stryker Leibinger GmbH,Freiburg; Prof. Dr. V. Sturm, Dept. for Stereotactic & FunctionalNeurosurgery, University of Cologne; Prof. Dr. M. Wannenmacher,Radiologic University Hospital, Heidelberg; Prof. Dr. S. Webb, Dr.J. Bedford, Royal Marsden Hospital, Sutton, UK

The task of the working group is the integration, pre-clini-cal testing, support, maintenance and quality assurance ofsoftware modules and packages designed for 3D treat-ment planning which have been developed by members ofthe Division of Medical Physics and are afterwards put atthe disposal of the Clinical Cooperation Unit Radiation On-cology (E0500) for clinical testing. The VOXELPLANproject, initially funded by the Deutsche Krebshilfe, hasdemonstrated impressively the scientific and clinical poten-tial of the software packages developed by the Division ofMedical Physics. The unique research environment givenin the Research Program Radiological Diagnostics andTherapy enables an interdisciplinary close cooperation ofphysicians, physicists and computer scientists resulting ina direct benefit to thousands of cancer patients treatedwith 3D conformal radiotherapy and intensity modulatedradiotherapy in clinical trials at DKFZ [1-6,8] and at thesites of its cooperation partners [7,9]. In addition to thismajor clinical activity the software packages supported bythe group serve as carrier systems for research activitiesof members of the Division of Medical Physics as well asof their national and international scientific cooperationpartners.

Until recently, there have been few regulations in Europeconcerning medical equipment, and there still is a lot ofuncertainty on behalf of „manufacturers“ of „medical de-vices“ like DKFZ, if and how to employ the new Medizin-produktegesetz and its ordinances derived from the Euro-pean Medical Devices Directive (93/42/EWG). As themedical devices built by the Division of Medical Physicsare not „placed on the market“, but only „put into service“within the Clinical Cooperation Unit Radiation Oncologythey do not need to bear the CE mark indicating that theyhave been subjected to a conformity assessment proce-dure. However, the obligation remains to keep the risk tothe patient justifiable compared with the potential benefit,and to carry out quality assurance procedures complyingto state-of-the-art safety standards and applicable law. Asthere is also an obligation to notify the authorities aboutthe manufacturing, servicing or marketing of medical de-vices - and to report any incidences - the competent au-thorities were contacted and a preliminary certificate of ex-emption for the further clinical application of radiotherapyequipment manufactured in-house was obtained. Due toongoing efforts by the legislator to facilitate compliancewith the new regulations by enacting and modifying spe-cific laws and ordinances, the working group is followingup the state of affairs and fulfilling the associated qualifica-tions (e.g. compilation of medical devices books).

247




During the period under report a completely revised, UNIXbased version of VOXELPLAN was implemented, testedand finally released for clinical evaluation - in conjunctionwith the IMRT software package KonRad - in close coop-eration with the working groups “Therapy Planning – De-velopment (E0403)”, “Physical Models (E0401)” and “Bio-physics and Radiotherapy Physics (E0408)”. This new ver-sion compensates any known shortcomings of its prede-cessors and contains a wide range of new features andfunctionalities to simplify and speed up treatment planningprocedures and to facilitate the introduction and evaluationof new irradiation techniques. Supplemental quality assur-ance tools have been implemented to check the correct-ness and reliability of the planning systems as well as theconsistency and integrity of the patient data and devicesparameters which are all decisive for the quality of the re-sulting treatment plans. These tools can be employed tomonitor the quality of such systems, and to direct qualityassurance activities. However, although the developmentand application of dedicated treatment planning softwareis a major clinical activity of the Division of Medical Phys-ics, the efforts to shift routine work to commercially avail-able systems were further intensified due to the additionalexpenses - in terms of manpower and other resources -caused by the obligation to comply with regulations duringthe permanent operation of non-CE certified systems, andespecially when newly developed software is put into ser-vice. The objective is to ensure the availability of the in-house systems for research purposes – provided that theresources remain constant - by discharging them from asmany routine tasks as possible. The working group there-fore also covers the commissioning, operation, servicingand quality assurance of the CE certified treatment plan-ning systems available at DKFZ.

Publications (* = external co-author)[1] Debus, J.; Kocagoncu, K.O.*; Höss, A.; Wenz, F.; Wannen-macher, M.: Fractionated stereotactic radiotherapy (FSRT) for op-tic glioma. Int J Radiat Oncol Biol Phys 44 (2): 243–248 (1999)[2] Herfarth, K.K.; Debus, J.; Lohr, F.; Bahner, M.L.; Fritz, P.;Höss, A.; Schlegel W.; Wannenmacher, M.F.: Extracranial stereo-tactic radiation therapy: set-up accuracy of patients treated forliver metastases. Int J Radiat Oncol Biol Phys 46 (2): 329–335(2000)[3] Lohr, F.; Pirzkall, A.; Debus, J; Rhein, B.; Höss, A.; Schlegel,W.; Wannenmacher, M.: Conformal three-dimensional photon ra-diotherapy for paranasal sinus tumors. Radiother Oncol 56 (2):227- 231 (2000)[4] Panten, T.*; Höss, A.; Bohsung, J.*; Becker, G.*; Sroka-Perez,G.*: Time Requirements in Conformal Radiotherapy TreatmentPlanning. Radiother Oncol 51 (3): 211–214 (1999)[5] Pirzkall, A.; Lohr, F.; Rhein, B.; Höss, A.; Schlegel, W.;Wannenmacher, M.; Debus, J.: Conformal radiotherapy of chal-lenging paraspinal tumors using a multiple arc segment therapy.Int J Radiat Oncol Biol Phys 48 (4): 1197–1204 (2000)[6] Pirzkall, A.; Carol, M.*; Lohr, F.; Höss, A.; Wannenmacher, M.;Debus, J.: Comparison of intensity-modulated radiotherapy withconventional conformal radiotherapy for complex-shaped tumors.Int J Radiat Oncol Biol Phys 48 (5): 1371–1380 (2000)[7] Schreuder, A.N.*; Jones, D.T.L.*; Symons, J.E.*; de Kock,E.A.*; Hough, J.K.*; Wilson, J.*; Vernimmen, F.J.*; Schlegel, W.;Höss, A.; Lee, M.*: The NAC proton treatment planning system.Strahlenther Onkol 175 (Suppl 2): 10–12 (1999)

[8] Schulz-Ertner, D.; Debus, J.; Lohr, F.; Frank, C.; Höss, A.;Wannenmacher, M.: Fractionated stereotactic conformal radiationtherapy of brain stem gliomas: outcome and prognostic factors.Radiother Oncol 57 (2): 215–223 (2000)[9] Zurlo, A.*; Lomax, A.*; Höss, A.; Bortfeld, T.; Russo, M.*;Goitein, G.*; Valentini, V.*; Marucci, L*; Capparella, R.*; Loasses,A.*: The role of proton therapy in the treatment of large irradiationvolumes: a comparative planning study of pancreatic and biliarytumors. Int J Radiat Oncol Biol Phys 48 (1): 277–288 (2000)

Hardware Developments and TechnicalSystems (E0405/E0406)W. Schlegel, S. Barthold, G. Echner, K.-H. Grosser,W. Korb, G. Kuhr, T. Liebler O. Pastyr, M. Scherer,M. SchnebergerIn cooperation with: Howmedica Leibinger GmbH, Freiburg; Dr. A.Hamilton, University of Arizona Health Sciences Center, Tucson,USA; Priv. Doz.. Dr. Dr. J. Debus, Clinical Cooperation Unit Radia-tion Oncology, DKFZ; Dr. K.-H. Höver, Prof. Dr. Dr. W.Semmler,Div. of Biophysics and Medical Radiation Physics, DKFZ; MRCSystems GmbH, Heidelberg; Prof. Dr. J. Richter, Clinics for Radia-tion Therapy, University of Würzburg; Prof. Dr. V. Sturm, Dept. forStereotactic & Functional Neurosurgery, University of Cologne;Prof. Dr. Dr. M. Wannenmacher, Radiologic University Hospital,Heidelberg; Zett Mess Technik GmbH St. Augustin; SiemensMedical Systems, Inc. Oncology Care Systems Group, Concord,USA.

The aim of radiation therapy is to eradicate a tumor with-out causing significant damage to contiguous normal tis-sue, especially to organs at risk. These demands definethe guidelines for two important concepts in radiotherapy:Spatial conformation of the radiation dose to the targetand fractionation of the treatment.

Spatial conformation of the dose to the target allows theapplication of high doses to the tumor volume. The corre-sponding irradiation techniques require on the one handadequate field shaping devices as multileaf collimators(MLC) and on the other hand accurate patient setup, im-mobilization and, in the ease of fractionated radiotherapy,exact repositioning. This is particularly true, if radiosensi-tive organs are situated close to the target.

The task of the two working groups is the development ofnew techniques and hardware components for radiationtherapy. Main issues are the improvement of patient setupaccuracy and motion control during treatment, the devel-opment and implementation of electro-optical devices, im-age processing tools for treatment validation and verifica-tion and the development of electro-mechanical devicesfor field shaping conformal radiotherapy.

Fixation, immobilization and positioningIn radiotherapy or radiosurgery, safety margins have to beincluded in the target volume to account for patient mis-alignment during setup, involuntary patient movementsduring therapy and organ motion, that can cause rathercomplex possibilities of movements of the target point.Minimising this safety margin allows a more conformtherapy with decreased field sizes and the delivery ofhigher doses to the tumor.

248




Whereas the bony structure of the skull realizes a fairlygood immobilization for intracranial structures, the situa-tion for extracranial targets is much worse. Especially forthoraco-abdominal targets, periodical movements, organmotion and deformation compromises the aim of a highlyaccurate and reproducible positioning of the patient.

Due to individualized motion patterns, a generalized con-sideration in the treatment and planning process is not ad-equate. Thus the actual work focusses on the fields of im-mobilization and fixation as well as optical tracking andmonitoring of the patient position.

For extracranial targets, a stereotactical fixation devicehas been developed in cooperation with HowmedicaLeibinger in Freiburg and the Health Sciences Center inTucson, University of Arizona, USA. It is designed espe-cially for treatment demanding exceptional high spatial ac-curacy, e.g. lesions in the immediate vicinity of the spinalcord. Figure 1 shows the extracranial fixation device with amounted localization system for CT imaging. For single-fraction irradiation of intracranial lesions, stereotactic im-mobilization techniques have been developed. Thesetechniques are generally based on invasive stereotacticframes, that define reference coordinate systems for imag-ing, therapy planning and patient positioning. The position-ing and immobilization accuracies for these invasive framebased techniques are better than 1 mm. For fractionatedradiotherapy of head and neck lesions, non invasive headholder systems are frequently used for patient setup andimmobilization.

The problem of fixation accuracy of patients in headmasks is discussed in the literature. Depending on themask material and the molding techniques, the achievableaccuracy for immobilization and repositioning is in a rangeof several mm . This spatial uncertainty does not allow afractionated irradiation scheme for lesions in the immedi-

ate vicinity of critical structures like the brainstem, opticnerves etc. Therefore, non-invasive immobilization and po-sitioning devices with a high spatial accuracy and patienttolerability have to be developed. Furthermore, these sys-tems should not prolongate the treatment time and thecomplexity of manual interactivity.

For high precision fractionated radiotherapy in the headand neck region, an integrated videogrammetry basedsystem (PPSU) has been developed [1,2,3]. Two cali-brated CCD-cameras generate continuously stereoscopicimages of external markers. The markers are attached to adento-maxillary fixation system which fits perfectly theteeth of the patients upper jaw. The stereoscopic imagesare processed and allow the calculation of the markers po-sition in space. Under the assumption of rigid body me-chanics between markers and target point, the tracking ofthe markers position reveals the corresponding targetpoint position. For a default position of the target point, i.e.linac isocenter, the actual deviation can be monitored instandard video frequency (25 frames per second).

The described patient positioning sensor unit (PPSU) canserve multiple purposes. The measurement system can beused to determine and display the position of the targetpoint relative to the isocenter during patient setup. Further-more it can be used to detect and display target point dis-placements caused by patient motions during therapy.User defined limits in all 6 degrees of freedom can serveas interrupt criterion for the therapist.

The optional positioning system serves as an automatichigh precision device for patient setup during subsequentirradiations in fractionated therapy. Therefore the mea-sured data of target point displacement are input data forthe movement of the tabletop. Beyond this ability duringthe setup procedure, the positioning system can be usedfor online compensation of the motions detected during

Figure 1: Extrac-ranial fixation de-vice for the trunkof body.

249




therapy. The overall positioning accuracy of the completesystem was determined to be 0.6 mm.

This high positioning accuracy allows for a significant de-crease of the commonly added safety margins. In conse-quence, the volume of the tumor-surrounding healthy tis-sue, that will be irradiated as an effect of the safety mar-gins, can substantially be reduced.

An additional important application of video-based posi-tioning and tracking systems are breast irradiations. Theuse of intensity modulated radiotherapy techniques is apromising approach to improve radiotherapy of breast can-cer. But it is very important to position the breast irradiatedas precise as possible performing new precise irradiationtechniques. Female breast is a deformable organ and it isnot allowed to put thick material on breast skin because ofthe danger of irradiation damage of the skin. Therefore it isa challenge to position the female breast in a precise andreproducible way during the treatment course.

In the framework of a breast cancer project we are investi-gating the feasibility of a video-based positioning systemfor breast cancer treatments.

The overall goal of these activities is to provide the toolsfor adaptive radiotherapy, which is the application of navi-gation concepts to conformal radiotherapy.

Conformal irradiation techniquesBesides accurate immobilization and positioning of the pa-tient, it is important to tailor the delivered dose distributionwith respect to the target volume and the organs at risk.Field shaping in combination with intensity modulation al-lows the delivery of arbitrary dose distributions. This con-formation of the tumor is realized by using compensatorsor, with more flexibility, by using multileaf collimators. TheMLC consists of 40 pairs of tungsten leafs, driven by 80DC motors. With a leaf width at isocenter of 1.6 mm, themaximum field size is 73 x 64 mm. With a maximum leafovertravel of 24 mm, a max. leaf speed of 15 mm/sec anda micro-controller based leaf position verification with a ac-curacy < 0.5 mm, the MLCs specifications meets the de-mands for dynamic field shape and intensity modulated ra-diotherapy (IMRT). The clinical Implementation of the MLCat DKFZ is completed now. Furthermore the developmentwas commercialized by MRC, Heidelberg and the MLC isnow on market available.

In 1998 we started with IMRT patient treatments as one ofthe first centers in Europe using the build-in MLC of ourtreatment facility. Because of the complexity of this treat-ments, individual treatment verifications are mandatory.For that reason several dosimetric phantoms have beendeveloped. An important feature for clinical feasibility andclinical practice of a new treatment procedure is the timerequired. Therefore a software has been developed to ac-celerate the process of verification and evaluation of indi-vidual patient treatments, which was one of the most time-consuming process in treatment planning and preparation.Another way to reduce treatment time is to change thedose fractionation scheme slightly [4].

Furthermore it is planned to implement IMRT using theDKFZ Micro MLC.

This is the next logical step on the way to our goal toimplement patient case specific IMRT-tools in order to im-prove radiotherapy treatments

CoRA – A new Cobalt Radiotherapy ArrangementLinear accelaerators (Linac) are the standard irradiationtreatment units in clinical routine for both radiotherapy andradiosurgery. The available infrastructure in developingcountries together with technical and economical argu-ments allow in most cases only the use of isotope units forradiotherapy. On the other side there is a natural interestin modern conformal treatment techniques in these coun-tries. However, the technology of todays gamma tele-therapy units and modern treatment methods are mutuallyexclusive in most cases. Isotope units can not be utilizedin combination with multileaf collimators (MLC) due to theirlarge source sizes and the mechanical conditions of theseunits do not allow stereotactic radiosurgery. The Gamma-Knife and the Rotating Gamma Unit exclusively designedfor radiosurgical treatments are only applicable for craniallessons. So in developing countries the postulation for amultifunctional treatment unit is existent based also on fi-nancial limitations.

Therefore we developed a new concept of an irradiationsystem which combines the advantages of the GammaKnife and the Linac [5]. It allows the delivery of fraction-ated radiotherapy, usage of MLC for irregular field shap-ing, simultaneous irradiation form different directions, geo-metrical accuracy etc.. The unit is designed as anisocentric device. Stereotactic treatment techniques willalso be possible. The source holders including the sourcesare fixed on a swivel-mounted arc in such a way that theirbeam axes intersect at the isocenter of the unit (see fig. 2).The system gives also the possibility to attach beam limit-ing devices like MLC. For the first time CoRA allows the ir-radiation of more than one irregular field at the same timefor all target positions.

Figure 2: Schematic picture of CoRA indicates the components.

It was found that the System CoRA and the CoRA arctechnique will be suitable for treatments of cranial targetvolumes. In comparison with the multiple arc technique nodrawback is noticeable due to the reduced number of arcs.The CoRA arc technique has the advantage of simulta-

250




neous beam application, e.g. the ability for an optimizationof the treatment time. In the future CoRA perhaps will pro-vide a treatment modality suitable for both radiotherapyand radiosurgery in developing countries.

Development of a new stereotactic system forneurosurgeryStereotactic techniques allow to place surgical instruments(i.e.biopsy needles) or radioactive implants accurately andreproducibly at a certain target point and to move this in-struments along a selected trajectory with minimal damageto overlying structures.

In conventional stereotactic neurosurgery a base ring ismounted to the patient to fix and immobilize the head ofthe patient and to define a coordinate system to registerdiagnostical images. Further it is used to fix the target sys-tem which facilitates the accurate placement of the instru-ments.

Nevertheless there are a few drawbacks of these conven-tional systems. Not all trajectories which are useful from aclinical point of view are technically feasible with such con-ventional systems. Furthermore conventional arc systemsare not adaptable to the whole body stereotactical neuro-surgery. These systems are restricted to the area to thehead.

To overcome this disadvantages computer controlled ro-bots have been developed or adapted for clinical use.Meanwhile such robots are commercially available. A ma-jor drawback of robotic systems are the high costs. Besidethis the absolute accuracy of robotic systems is not suffi-cient for stereotactic neurosurgery.

In cooperation with Prof. Dr. V. Sturm, Dept. for Stereotac-tic & Functional Neurosurgery, University of Cologne andan industrial partner (Zettmess GmbH, St. Augustin), weare developing a new stereotactical system for neurosur-gery based on a commercial 6 axes 3D measuring system.Such a system is similar to a passive manipulator arm (seefig. 3). The aim of the project is to adapt that system to theclinical constraints and to provide a prototype for clinicalstudies.

The newly developed system will be much more cost-ef-fective as robotic systems and more flexible as conven-tional arc systems. Furthermore it is also suitable for wholebody stereotaxy.

Because certain constraints in stereotactical neurosurgeryare similar to the constraints in stereotactical radiosurgeryour experience in the field of stereotactical radiosurgery istransferable to the new project.

Patient couch systemIn cooperation with Prof. Dr. V. Sturm, Dept. for Stereotac-tic & Functional Neurosurgery, University of Cologne an“operating room couch-system” was developed so that thepatient remains on a special couch which can be movedfrom every diagnostic instrument (CT, MRT, x-ray, etc) tothe treatment units (operating room, linac, etc) with thehelp of trolleys. Adapter plates were constructed to fit thepatient couch to the different couches of the various ma-

chines. A newly developed operating room table allows tomove the patient from a basic height of appr. 70 cm to amax. height of appr. 125 cm. The advantage of this table isthat the surgeon can operate the patient from the occipitalside while the surgeon is sitting on his chair and the pa-tient is lying on the couch in the supine position. Specialmaterials such as kevlar (patient couch) make it possibleto treat the patient e.g. inside the MRT.

The first patient treatments on the new couch system willstart in Köln in May 2001.

Publications (* = external co-author)[1] G. Kuhr, C. Lappe, W. Schlegel: Patient Positioning SensoringUnit (PPSU) - Patientenpositionierung für die konformierendeStrahlentherapie im Kopf und Halsbereich, TagungsbandMeditinische Physik ’98, Dresden, (1998) 237-238.[2] G. Kuhr, C. Lappe, W. Schlegel: Patient Positioning SensorUnit (PPSU) for stereotactically guided fractionated radiotherapyESTRO 17" Annual Meeting, Edinburgh 1998.[3] G.Kuhr, Ein optisches Mess-System zurPatientenpositionierung in der Präzisionsstrahlentherapie vonTumoren, PhD-Thesis, Ruprecht-Karls-Universität Heidelberg ,1999[4] K.-H. Grosser, Reduction of Treatment Time in IMRT Step andShoot Irradiations by Means of a Changed Fractionation Scheme,In: The Use of Computers in Radiotherapy, Editors: W.Schlegel,T.Bortfeld, (2000) p. 308-310.[5] S. Barthold, W. Schlegel, O. Pastyr, G. Echner: Investigationson medical aspects of CoRA –A new Cobalt Radiotherapy Ar-rangement (feasibility study), In: The Use of Computers in Radio-therapy, Editors: W.Schlegel, T.Bortfeld, (2000), 347-449.

Photodynamic Therapy (PDT) and Diagnosis(PDD) (E0407)T. Haase, M. Rheinwald, U. Bauder-Wüst.In cooperation with: Dr. J. Gahlen, Dept. of Surgery, University ofHeidelberg; Dr. P. Kremer, Dept. of Neurosurgery, University ofHeidelberg; Dr. Dr. A. Kübler, Dept. of Oral and Maxillofacial Sur-gery, University of Heidelberg; Dr. H. Sinn, Dr. A. Wunder, DKFZ,Division of Radiochemistry and Radiopharmacology; Dr. U.Zillmann, DKFZ, Central Animal Experiment Facility

Photodynamic Therapy is a novel therapeutic modality forthe treatment of tumors. A locally or systemically adminis-tered light-sensitive substance (photosensitizer or fluores-cent dye) selectively accumulates in tumor tissue. Whenexcited by light with an appropriate wavelength, the photo-sensitizer molecules emit fluorescence light which makesthe tumor visible or delineates tumor margins (diagnosis),and accordingly, singlet oxygen is produced by energytransfer processes, which kills the treated tumor cells(therapy).

In cooperation with different departments of the Heidel-berg University clinic, new photosensitizers and applica-tions are investigated.

Main subjects of interest are the investigation of thephotophysical and pharmacokinetic properties of newly de-veloped macromolecular photosensitizers and fluorescentdyes, the optimization of therapy and diagnosis and thetechnical support of the clinical partners during the imple-mentation of photodynamic therapy in the clinic.

251




Characterization of PDT-relevant properties of thephotosensitizer mTHPC coupled to polyethyleneglycolTo further enhance the pharmacokinetic properties of thephotosensitizer mTHPC, a derivatization technique tocouple mTHPC covalently to polyethylene glycol (PEG)has been developed by the Division of Radiochemistry andRadiopharmacology of the DKFZ.

For this new substance, the preservation of absorption andfluorescence properties could be found. In animal experi-ments, mTHPC-PEG showed to have an increased circula-tion half-life (in our model t = 20 h) compared to freemTHPC, as well as an enhanced tumor uptake. After 72 h,20 % of the initially applied dose of radioactively labeledmTHPC-PEG could be detected in the tumor (fig. 1). Thiswas a 40-fold concentration of sensitizer compared tomuscle and a 10-fold concentration compared to skin.

However, cell culture experiments and therapy experi-ments in animal models showed a decreased photody-namic efficiency of mTHPC-PEG compared to mTHPC.The molecule linking mTHPC with PEG could be identifiedas responsible for inferior cell uptake of mTHPC-PEG.

A new substance with a different linker molecule was suc-cessfully developed and is currently investigated in cellculture and animal experiments. Preliminary results indi-cate that its efficiency is comparable to that of the low mo-lecular mTHPC.

Photodynamic therapy with mTHPC in oral andmaxillofacial surgeryNon-melanomatous skin tumors are one of the most fre-quent tumors in the white population and mainly causedby cumulative exposure to solar ultraviolet B radiation. Onaccount of this, about 80 percent of all non-melanomatousskin tumours are located on the arms or the head andneck. Standard treatment for most tumours is surgical re-section, with often only a moderate cosmetic outcome.

In cooperation with the Dept. of Oral and Maxillofacial Sur-gery, University of Heidelberg, the effect of photodynamictherapy on primary non-melanomatous skin tumors of thehead and neck (squamous cell cancer, basal cell cancer,actinic keratosis, Bowen´s disease) was tested in a pro-spective clinical trial. In this study meta-Tetrahydroxy-phenylchlorin (mTHPC / Foscan�), a systemic photosensi-tizer of the second generation, was applied. Patients wereinjected 0.15 mg/kg or 0.10 mg/kg mTHPC 96 hours priorto laser light exposure. Light was delivered via fibres by anargon-dye laser at 652 nm, 100 mW/cm2 and a light doseof 5 – 20 J/cm2.

25 patients with a total of more than 150 non-melanoma-tous skin tumors and a mean follow up of 15 months(ranging 6 to 26 months) were treated. Within several daystumor necrosis appeared followed by wound healing within4 to 8 weeks, leaving only minor scars behind (fig. 2). Sev-enty tumors showed a complete response with an excel-lent cosmetic outcome and only three basal cell cancersresponded with partial success. No adverse eventsoccured. The therapy was supported by a high degree ofpatient satisfaction.

By choosing the correct drug and light dosage, a selectivetumor necrosis can be obtained. PDT using mTHPCseems to be a promising new and safe treatment modalityfor the treatment of primary non-melanomatous skintumours of the head and neck which can substitute surgi-cal therapy, offering an even better cosmetic outcome.

In another phase II study 22 patients with primary squa-mous cell cancer of the head and neck were treated withphotodynamic therapy. Again the response rates werecomparable to those of standard surgical therapy, whilethe cosmetic and functional outcome was much better.

Fig. 1: Scintigraphy of a rat with Walker-Carcino-Sarkoma on the right hind leg. 72 h after application 20 % of the initially applied doseof radiolabled mTHPC-PEG accumulates in the tumor region.

252




Laser-induced fluorescence detection of gliomausing aminofluorescein-labeled serum albumin(AFL-HSA)Surgical treatment of malignant gliomas is limited due tothe difficult identification of the tumor margins even underthe operating microscope and the diffuse migration of ma-lignant cells into the surrounding brain. Although evencomplete removal of the solid tumor mass cannot cure thedisease, patients still benefit in terms of quality of life andsurvival time. In most cases, residual tumor and tumor re-currence develop from these residual macroscopic parts ofthe tumor margins, which can be unequivocally detectedby early postoperative and following-up MR images.

In cooperatin with the Dept. of Neurosurgery, University ofHeidelberg, the use of aminofluorescein coupled to humanserum albumin (HSA) as a fluorescent tumor marker for in-traoperative visualization of the tumor margins was investi-gated.

In cell and animal experiments it could be shown that sig-nificant amounts of the applied dose of AFL-HSA weretaken up by proliferating cells. In a C6 glioma rat modelthe fluorescent dye could be used to identify the tumormargin, using laser light to excite fluorescence.

Up to now, three patients have been successfully treatedusing AFL-HSA as intraoperative tumor marker (fig. 3). Pa-tients were injected with AFL-HSA four to five days prior tosurgery. During the operation the fluorescein-contrastedtumor was well visible when inspected with the naked eyeafter activation with laser light and was clearly delineated

against the surrounding brain. A clinical study investigatingthe use of AFL-HSA for neurosurgery is in preparation.

Publications (* = external co-author)[1] A. Kübler*, T. Haase, M. Rheinwald, T. Barth*, J. Mühling*:Treatment of oral leukoplakia by topical application of 5-Aminolevulinic acid., Int J Oral Maxillofacial Surg 27, 466-469(1998)[2] A. Kübler*, T. Haase, P. Kremer*, M. Rheinwald, T.Barth*, S.Kunze*, J. Mühling*: Ein Lasersystem für die klinischeAnwendung der Photodynamischen Therapie und Diagnostik,Lasermedizin 14:8-14 (1998)[3] A. Kübler*, T. Haase, P. Kremer*, M. Rheinwald, S. Kunze*, J.Mühling*: An argon-dye laser system for Photodynamic Therapyand Diagnosis, Neurological Research, Vol. 21, 103-107, (1999)[4] Kubler AC*; Haase T; Staff C**; Kahle B*; Rheinwald M;Muhling J*: Photodynamic therapy of primary nonmelanomatousskin tumours of the head and neck. Lasers in Surgery and Medi-cine. 1999; 25(1): 60-8[5] Gahlen J*, Pietschmann M*, Laubach H-H*, Prosst R*,Rheinwald M, Haase T, Herfarth Ch*: Minimal invasiveMöglichkeiten zur Fluoreszenz-optischen Detektionintraperitonealer Metastasen. Lasermedizin. 1999, 14: 86-90[6] Gahlen J*, Prosst RL*, Pietschmann M*, Rheinwald M, HaaseT, Herfarth C*: Spectrometry supports fluorescence staginglaparoscopy after intraperitoneal aminolaevulinic acid lavage forgastrointestinal tumours. Journal of Photochemistry and Photobi-ology B. 1999 52 (1-3): 131-135[7] Kremer P*, Wunder A, Sinn H, Haase T, Rheinwald M,Zillmann U, Albert FK*, Kunze S*: Laser-induced fluorescence de-tection of malignant gliomas using fluorescein-labeled serum al-bumin: experimental and preliminary clinical results. NeurologicalResearch. 2000, 22 (5): 481-489

Fig. 2: Basal cell cancer on the nose before photodynamic therapy (left) and 3 months after treatment (right)

Fig. 3: Glioma intraoperative 72 h after injection of AFL-HSA. . Left: white light, right: fluorescence after laser light excitation.

253




Biophysics and radiation therapy physics(E0408)G.H. Hartmann, F. Fölisch, D. Haspel, R. Hofmann,O. Jäkel, C. Karger

There are three projects that were carried out on a largerscale in cooperation with other groups within and outsidethe Division of Medical Physics: the medical-physical is-sues to the German Heavy Ion Therapy Project (seeE0409), the project aiming at an implementation of a totalquality management system to all radiotherapy activitieswithin the division, and a project to develop motor drivenmulti-leaf collimators and to introduce them into clinicalpractice. Two other activities are located totally within theworking group: the development and application of arraydetectors for 2D and 3D dosimetry, and a research projectto study long time radiation damage after radiosurgery inan animal model.

1. Cooperation projects1.1 Implementation of a total quality managementsystem (TQM) for the radiotherapy activities ofthe divisionMany aspects of a quality management system of softwarerelated items in radiotherapy planning have already beenaddressed under E0401. In addition to that, a TQM mustaddress each single part of the chain of procedures in-volved in a course of radiotherapy, starting from imagedata acquisition, then going through creating the patientmodel (including patient related coordinates), radiotherapyplanning, patient positioning and finally coming to thebeam delivery itself. A formal structure has been estab-lished to accomplish this task, and this structure has beensuccessfully applied in the area of radiosurgery and heavyion therapy. It remains to extend this concept to any radio-therapy activities within the division.

1.2 Introduction of motor driven multi-leafcollimators into clinical practiceThe introduction of multi-leaf-collimators (MLC) in radio-therapy has substantially initiated new methods with ahigh tumor conformation of the delivered radiation dose.The development of adequate MLC devices (manual ormotor driven) is one of our central activities aiming at animprovement of limiting the dose distribution to the target

volume only. Several processes have to be performed andcoordinated to be able to finally apply MLCs at patientssuch as the mechanical development, control electronics,test of dosimetrical properties, implementation in a treat-ment planning system. Finally the needs to get the ap-proval of authorities have to be fulfilled. This project is aim-ing at an application of our MLCs under routine clinicalconditions up to the end of this year.

2. Development of array detectors for 2D and 3DdosimetryThe dose distribution achieved by the new radiationtherapy techniques (conformal radiotherapy, IMRT, rasterscan particle therapy) are complex and three dimensionalin shape. The general problem associated with that is toverify the correct (i.e. the pre-planned) generation of suchdose distributions. Adequate 3D measuring techniques aretherefore required. Our concept is to accomplish a 3Dmeasurement by a high number of small detectors distrib-uted in an appropriate way in space such as segmentedarray detectors using liquid detector material (liquid ioniza-tion chamber). We currently also develop an electronic ac-quisition system for that of up to thousand simultaneouslymeasuring channels and suitable phantoms.

3. Long time radiation damage after radiosurgeryin an animal modelHigh single doses are particularly critical with respect tocreate a necrosis in the brain. The aim of this researchproject is to quantitatively and systematically determinelong time radiation damage as observable by MR investi-gations after small sized dose volumes in the rat brain. Forthis purpose a radiosurgery irradiation technique was de-veloped applicable to the rat. In a pilot experiment irradia-tions were performed with the 3 mm collimator applyingdose values from 20 - 100 Gy to the left part of the brain.Changes of MR signals were observed over a time periodof 18 months. Following this, a preliminary dose effectcurve as well as a value for D50 were determined (Fig. 1).The histological evaluation revealed that the changeswere limited to the small sized dose volume. Based on thispilot experiment, the main irradiation experiment followedusing 110 animals to obtain more significant data. Theseresults were expected for this year aiming at a referencefor further experiments with modified irradiation conditions.As an example, an irradiation experiment was performedapplying carbon ions instead of X-rays from the linear ac-celerator. Again, results of this experiment with respect tothe associated relative biological effectiveness were ex-pected within this year.

Publications (* = external co-author)[1] Henke K; Hartmann GH; Peschke P; Hahn EW: Stereotacticradiosurgery of the rat dunning R3327-AT1 prostate tumor. Inter-national Journal of Radiation Oncology Biology and Physics, 36(1996) 385-391.[2] Karger CP; Hartmann GH; Peschke P; Debus J; Hoffmann U;Brix G; Hahn EW; Lorenz WJ: Dose-response relationship for latefunctional changes in the rat brain after radiosurgery evaluated bymagnetic resonance imaging. International Journal of RadiationOncology Biology and Physics, 39 (1997) 1163-1172.

Fig. 1. Results from the pilot experiment: long time radiation effectafter radiosurgery in the rat brain

254




[3] Münter M. W., Karger C. P., Schneider H.-M., Peschke P.,Debus J.: Delayed vascular injury after single high dose irradiationin the rat brain: Histological, immunohistochemical andangiographic studies. Radiology 212 (1999) 475-482[4] Reinacher-PC Blum-C Gass-P Karger-CP Debus-J: Quantifica-tion of Microglial Late Reaction to Stereotactic Irradiation of theRat-Brain Using Computer-Aided Image-Analysis. ExperimentalNeurology 160 (1999) 117-123[5] Jäkel O, Hartmann GH, Heeg P, Jacob C, Karger CP, KrämerM, Kress J, Krießbach A, Lappe C, Debus J: ErsteSchwerionenbestrahlung von Patienten in Europa:Medizinphysikalische Aspekte. Zeitschrift für Medizinische Physik9, 88-95, 1999[6] Hartmann GH, Jäkel O, Heeg P, Karger CP, Krießbach A: De-termination of water absorbed dose in a carbon ion beam usingthimble ionization chambers. Phys Med Biol 44, 1193-1206, 1999[7] Heeg P., Hartmann G.H., Jäkel O., Karger C.P., *Kraft G.:Quality assurance at the heavy-ion therapy facility at GSI.Strahlentherapie und Onkologie 175 (Suppl. II), 36-38, 1999[8] Jäkel O., Krämer M., Hartmann G.H., Heeg P., Karger C.P.,*Kraft G.: Treatment planning for the heavy-ion facility at GSI.Strahlentherapie und Onkologie 175 (Suppl. II), 15-17, 1999[9] Karger C. P., Jäkel O., Hartmann G., H Heeg P.: A System forThree-Dimensional Dosimetric Verification of Treatment Plans inIntensity-Modulated Radiotherapy with Heavy Ions. Med. Phys.26: (10) 2125-2132, 1999[10] Karger C. P., Hartmann G. H., Jäkel O., Heeg P.: QualityManagement of Medical Physics Issues at the German Heavy IonTherapy Project. Med. Phys. 2000 (in press)[11] Jäkel O., Hartmann G. H., Heeg P. Schardt D.: Effective pointof measurement of cylindrical ionization chambers for heavycharged particles. Phys. Med. Biol. 45 (2000) 599–607[12] Karger C.P., Hartmann G.H.: Determination of tolerance doseuncertainties and optimal design of dose response experimentswith small animal numbers. Strahlentherapie und Onkologie 177(2001) 37-42[13] Jäkel O., Jacob C., Schardt D., Karger C.P., Hartmann G.H.:Relation between carbon ion ranges and x-ray CT numbers fortissue equivalent phantom materials. Medical Physics 2001 (ac-cepted).[14] Karger C.P., Hartmann G.H., Heeg P., Jäkel O.: A method fordetermining the alignment accuracy of the treatment table axis atan isocentric irradiation facility. Physics in Medicine and Biology46 (2001) N19-N26[15] Karger C.P., Jäkel O., Debus J., Kuhn S., Hartmann G.H.:Three- dimensional accuracy and interfractional reproducibility ofpatient fixation and positioning using a stereotactic head masksystem. International Journal of Radiation Oncology, Biology,Physics 49, 1223-1234, 2001

Heavy Ion Therapy Project (E0409)O. Jäkel,. P. Heeg, C. Karger, U. OelfkeIn cooperation with: PD J. Debus, Clinical Cooperation UnitStrahlentherapeutische Onkologie, DKFZ; Prof. M.Wannenmacher, Radiological University Hospital Heidelberg; Prof.G. Kraft, Department for Biophysics of the Gesellschaft fürSchwerionenforschung (GSI), Darmstadt

The aim of the heavy ion therapy project is to investigate,if the special physical and biological properties of heavyions can be translated into an increased local control ofmalign tumors. The special feature of heavy ions in radio-therapy is, that they have a limited range in tissue that canbe controlled by their energy and that their radiobiologicaleffect in tissue is strongly enhanced, as compared to con-ventional X-rays.

In a joint research project of the radiological universityhospital Heidelberg, the German heavy ion research labo-ratory (GSI) and the DKFZ, a therapy facility for carbonions was established at GSI. Between December 1997and November 2000 in total 78 patients mostly with maligntumors of the base of skull were treated with carbon ionsin various clinical studies. The project group at DKFZ is re-sponsible for the medical physics aspects. This includesclincal dosimetry, treatment planning, patient positioningand quality assurance for these issues. Besides improve-ments and developments of new techniques for the clinicalroutine at GSI, the project group is also involved in thepreparation and design of a clinical heavy ion center at theuniversity hospital Heidelberg within a strategy fund.

The ongoing projects in the field of dosimetry are the pre-cise measurement of various correction factors within thedosimetry protocol to improve the precision of dosemeasurement and the optimization of verifications of thepatient specific treatment fields, to allow an efficientcontrol of dose distributions prior to the begin of therapy.This also includes modified techniques that are necessaryfor the use with a rotating gantry.

In the field of treatment planing, the planning software isconstantly being developed further, to allow a more effi-cient planning process. This is necessary to cope with themuch higher number of patients to be treated at a clinicaltherapy facility. The modifications of the optimization algo-rithms aim at a further improvement of dose conformation.The existing software will also be used to investigate theprincipal possibilities and limits of this new therapy whenapplied to new tumor localizations and indications.

For a more flexible patient positioning, a treatment chair isbeing developed, to gain an additional degree of freedomin the treatment angles. Furthermore, the X-ray-units forposition verification of the treated patients are evaluated toretrospectively assess the accuracy of positioning andtheir implications for the treatment. The currently used po-sitioning methods for the head and neck region are beingmodified for future applications in the trunk of the body.

The quality assurance procedures are being developedfurther in all areas. They allow to check the precision ofthe techniques used in the various fields and to constantly

255




maintain a high level of quality by the use of constancychecks. New test methods are especially important for theofficial approval of new techniques by the authorities (e.g.irradiations in the trunk of the body or use of the treatmentchair).Additional financial support: Strategy funds of the HelmholtzGesellschaft, HGF (Federal ministry for education and research,BMBF): 3 Postdocs and capital investments; BiomedizinischerForschungsverbund: capital investments.

Publications (* = external co-author)[1] O. Jäkel, M. Krämer*, C.P. Karger, J. Debus “Treatment plan-ning for heavy ion radio-therapy: clinical implementation and ap-plication” Phys. Med. Biol. (2001) in press.[2] O. Jäkel, C. Jacob, D. Schardt*, C.P. Karger. G.H. Hartmann,“Relation between carbon ions ranges and X-ray CT numbers”Med. Phys. 28(4) (2001) in press.[3] Karger C. P., Jäkel O., Debus J., Kuhn S., Hartmann G. H.:“Threedimensional accuracy and interfractional reproducibility ofpatient fixation and positioning using a stereotactic head masksystem”. Int. J. Radiat. Oncol. Biol. Phys. 49/5 (2001) 1223-1234.[4] Karger C. P., Hartmann G. H., Heeg P., Jäkel O.: “A method fordetermining the alignment accuracy of the treatment table axis atan isocentric irradiation facility”. Phys. Med. Biol. 46 (2000) N19-N26.[5] O. Jäkel, G.H. Hartmann, C.P. Karger, P. Heeg: “Quality assur-ance for a treatment planning system in scanned ion beamtherapy”. Med. Phys. 27 (2000) 1588-1600.[6] O. Jäkel, J. Debus: “Selection of beam angles for radiotherapyof skull base tumours using charged particles”. Phys. Med. Biol.45 (2000) 1229-1241.[7] O. Jäkel, G.H. Hartmann, P. Heeg, D. Schardt*: “Effectivepoint of measurement of cylindrical ionization chambers for heavycharged particles.” Phys. Med. Biol. 45 (2000) 599-607.[8] C.P. Karger, G.H. Hartmann, O. Jäkel, P. Heeg: “Quality Man-agement of Medical Physics Issues at the German Heavy IonTherapy Project. Medical Physics 27 (2000) 725-736.[9] J. Debus, T. Haberer*, D. Schulz-Ertner, O. Jäkel, F. Wenz*,W. Enghardt*, W. Schlegel, G. Kraft*, M. Wannenmacher*“Bestrahlung von Schädelbasistumoren mit Kohlenstoffionen beider GSI – Erste klinische Ergebnisse und zukünftigePerspektiven”, Strahlenther. Onkol. 176 (2000) 211-216.[10] M. Krämer*, O.Jäkel, T. Haberer*, G. Kraft*, D. Schardt*, U.Weber* “Treatment planning for heavy ion radiotherapy: physicalbeam model and dose optimization” Phys. Med. Biol. 45 (2000)3299-3317.[11] O. Jäkel, G. Hartmann, P. Heeg, C. Jacob, C. Karger, M.Krämer*, J. Kress, A. Krießbach, C. Lappe, J. Debus: “ErsteSchwerionenbestrahlung von Patienten in Europa: Medizin-physikalische Aspekte”; Z. Med. Phys. 9 (1999) 88-95.[12] Hartmann G. H., Jäkel O., Heeg P., Karger C. P., KrießbachA.: “Determination of water absorbed dose in a carbon ion beamusing thimble ionization chambers.” Phys. Med. Biol. 44 (1999)1193-1206.[13] C.P. Karger, O. Jäkel, G.H. Hartmann, P. Heeg: “A System forThree-Dimensional Dosimetric Verification of Treatment Plans inIntensity-Modulated Radiotherapy with Heavy Ions”; Med. Phys.26 (1999) 2125-2132.

Division Medical Physics (E0400) - dkfz.de · with IMRT based on our planning system KonRad....

Documents

Transcript of Division Medical Physics (E0400) - dkfz.de · with IMRT based on our planning system KonRad....