In-beam PET for radiotherapy monitoring

Wolfgang Enghardt1, Fine Fiedler, Daniela Möckel, Katia Parodi2

1Technische Universität Dresden, OncoRay – Radiation Research in Oncology
2Heidelberger Ionenstrahl-Therapiezentrum, Heidelberg, Germany

Radiation therapy is one of the cornerstones of modern cancer treatment. With increasing tendency more than 50 % of tumor patients are irradiated, either as the exclusive form of treatment or in combination with other modalities, like surgery or chemotherapy. The central challenge of radiotherapy is to destroy the tumor completely, while saving the surrounding healthy tissue. In some delicate therapeutic cases, e.g. for compact, deep-seated, radioresistant tumors growing in close vicinity to organs at risk, these objectives cannot be reached by the state of the art radiotherapy technology which rests upon hard photon or electron beams delivered by compact electron linear accelerators. Therefore, proton and light ion (e.g. carbon) beams become more and more important due to their favourable physical and radiobiological properties. To translate this potential into clinical results, new technologies in generating, forming and monitoring ion beams are required. In-beam positron emission tomography (PET), one of these innovative technologies, has been developed and transferred to clinical application by FZD at the experimental carbon ion therapy facility located at the Gesellschaft für Schwerionenforschung (GSI) Darmstadt in collaboration with the University Hospital and the German Cancer Research Center Heidelberg.

Collaborating with Siemens AG, Medical Solutions, we aim at transferring FZD’s unique knowledge of in-beam PET into the development of a clinical in-beam PET scanner for global ion therapy facilities which are planned or already under construction. This know-how comprises detector, signal processing and data acquisition technology for the critical operation of PET detectors at high-energy ion beams [1, 2], tomographic reconstruction algorithms optimized for in-beam PET and experience in clinical application [3].

Making invisible beams visible

The technological basis of in-beam PET is a double head positron camera [1] integrated into the therapy unit (fig. 1). During therapeutic irradiation this device detects the annihilation gamma-rays following the decay of minor amounts of positron emitting nuclei (predominantly 11C and 15O), which are produced via nuclear reactions between the impinging ions and the atomic nuclei of the tissue (fig. 2). Sophisticated algorithms of tomographic reconstruction deliver the spatial distribution of positron emitters in vivo. They are related to the dose distribution by means of a precise Monte Carlo simulation [4] of the production of positron emitters and the detection of annihilation gamma-rays (fig. 3). When these are compared with the measured PET images, deviations between the planned and actually delivered dose distributions can be revealed, quantified [3] and compensated for in the further course of the fractionated treatment. Deviations are caused by ion range modifications due to minor patient positioning errors in combination with large tissue density gradients or due to changes of the tissue density distribution within the irradiated volume (e.g. radiation induced tumor shrinking) during the three weeks of fractionated radiotherapy. At the carbon ion therapy facility of GSI more than 350 cancer patients, most of them with tumors in the head and neck region, have been treated since 1997 (134 in the years between 2004 and 2006). All these treatments were monitored by means of in-beam PET for quality assurance.In-beam PET phantom studies with protons at FZD/GSI [5] triggered research on post-radiation PET/CT imaging at Massachusetts General Hospital, Boston (fig. 4). First clinical results [6] confirm the predicted positive impact.

Enghardt_fig1 Enghardt_fig.2
Fig. 1: Double head positron camera developed by FZD at the treatment site of GSI Darmstadt. The horizontal carbon ion beam leaves the beam pipe visible through a 20´20 cm2 window in the centre of the picture. To provide sufficient space for patient positioning, the PET scanner can be moved on rails parallel to the beam between the measuring position displayed and the parking position upbeam. Fig. 2: Depth distributions of calculated dose (blue, dashed) and measured+-activity (red, solid) induced by beams of protons as well as 3He, 12C and 16O ions in thick targets of polymethyl methacrylate. The prominent maxima in the cases of 12C and 16O are formed by positron radioactive projectile fragments, whereas the pedestals as well as the distributions generated by 1H and 3He are due to target fragments.

For radiobiological reasons not only proton and carbon beams are highly desirable for therapy, but also a large variety of ion species showing atomic numbers between 1 (hydrogen) and 8 (oxygen). Novel ion therapy accelerators (e.g. at the Heidelberger Ionenstrahl-Therapiezentrum) are capable of delivering beams of all these ions with therapeutically relevant energy values. Since in-beam PET offers the unique possibility to measure particle ranges in-vivo, it allows sensitive testing of the physical beam model underlying the dose calculation algorithms for treatment planning. This is highly relevant for commissioning new ion species for therapy. Therefore, the GSI therapy facility is closely collaborating with the Heidelberger Ionenstrahl-Therapiezentrum and the European Organization for Nuclear Research/CERN, using its in-beam PET scanner to measure physical data which are necessary for extending in-beam PET to other ion beams like helium and oxygen with very high precision (fig. 2).

Enghardt_fig.3_links Enghardt_Fig.3_mitte Enghardt_fig.3_rechts
Fig. 3: Clinical application of in-beam PET at the carbon ion therapy facility at GSI Darmstadt. As an example, the irradiation of a chondrosarcoma of the skull base with a lateral portal coming from the left side of the patient, i.e. right side in the picture, (maximal dose: 0.63 Gy) is displayed. As indicated by the dose distribution superimposed onto the computed tomogram (left), the carbon ions must not penetrate the brain stem as an organ at risk. The comparison of the predicted (middle) with the measured (right) b+-activity distributions shows that this was fulfilled during the treatment. The isodose and isoactivity lines are decoded in rainbow colours and denote 5, 15 … 95 % of the maxima.

Back to the photons

The successful application of PET for monitoring radiation therapy with ions motivated investigations on the feasibility of the method for hard photon beams. This is studied within the EC FP6 integrated project “BioCare – Molecular Imaging for Biologically Optimized Cancer Therapy” in collaboration with the Karolinska Institute, Stockholm, the Mathematics-Physics Department at Stockholm University, the Soltan Institute for Nuclear Studies, Otwock-Swierk, Poland, and CERN in Geneva. The aim of this research is to prove the feasibility of combining an in-beam PET scanner with an extremely compact 50 MeV electron accelerator delivering a pencil-like beam of ultrahard bremsstrahlung photons. In this case positron emitters are generated by (gamma, n) photonuclear reactions in the tissue at photon energy values above 20 MeV. The in-beam PET-related experimental research of the BioCare project is carried out at 20 until 40 MV bremsstrahlung beams delivered by the Radiation Source ELBE. By means of Monte Carlo calculations [7] it has already been shown that for electron beam energies beyond 30 MeV the induced dose-related activity density is comparable with that obtained during irradiation with carbon ions. This has been successfully confirmed in the second step [8], in which the 11C activity generated in plastic phantoms during photon irradiation at ELBE was measured and quantified by means of the human PET scanner of the Institute of Radiopharmacy at FZD (fig. 5). Just recently bremsstrahlung induced positron emitters (11C and 15O) have been imaged in-beam for the first time worldwide by means of a small limited angle positron camera installed at the ELBE beam. The encouraging result was that dosimetry as well as the control of patient positioning on the basis of in-beam PET appears feasible (fig. 5).

Enghardt_fig.4_rechts Enghardt_fig.4_links
Fig. 4: Monte Carlo calculated (left) and measured (right) activity distribution after proton irradiation of a clivus chordoma patient at Massachusetts General Hospital, Boston. Images by courtesy of K. Parodi and T. Bortfeld, to be published soon in [6].

Fig. 5: Two-dimensional b+-activity distributions generated by ultrahard bremsstrahlung (coming from the left) in an inhomogeneous phantom consisting of 2 cm thick slabs of polymethyl methacrylate (PMMA), polyethylene (PE) and different tissue-equivalent materials (lung, adipose and bone). Upper image: in-beam PET result measured during 34 MV photon irradiation showing the distribution of 11C (half-life: 20 min) and 15O (half-life: 2 min) from the photonuclear reactions 12C(gamma, n) and 16O(gamma, n), respectively. Lower image: result of the PET measurement at the human scanner of the Institute of Radiopharmacy. This measurement started 30 min after finishing the 30 MV photon irradiation. Therefore, it only shows the 11C distribution (15O has nearly completely decayed during 30 min) leading to a reduced contrast between PE, PMMA, adipose tissue and bone compared to the in-beam image.

Spin-off for education

Based on the expertise on PET instrumentation available at the Institute of Radiation Physics a PET scanner has been installed for education (fig. 6). This device is aimed at practical training in the general principles of tomography, in physics and mathematics of PET imaging and, furthermore, in the fundamentals of multi-parameter measurements in nuclear, radiation and particle physics. The scanner is used for hands-on training of students of medical radiation sciences at the Postgraduate School of the Center for Radiation Research in Oncology “OncoRay” – located at the Medical Faculty of Technische Universität Dresden – as well as of undergraduate physics students.

Fig. 6: Left: photograph of the PET-scanner for education. Right: detailed picture with (1) detector system, (2) object table, (3) crate with components for signal processing, (4) step motor controller, (5) standard PC for data acquisition and tomographic reconstruction.

[1] Charged hadron tumour therapy monitoring by means of PET, W. Enghardt, P. Crespo, F. Fiedler, R. Hinz, K. Parodi, J. Pawelke, F. Pönisch, Nuclear Instruments and Methods in Physics Research A525, 284 (2004)

[2] On the detector arrangement for in-beam PET for hadron therapy monitoring, P. Crespo, G. Shakirin, W. Enghardt, Physics in Medicine and Biology 51, 2143 (2006)

[3] Dose quantification from in-beam positron emission tomography, W. Enghardt, K. Parodi, P. Crespo, F. Fiedler, J. Pawelke, F. Pönisch, Radiotherapy and Oncology 73, S96 (2004)

[4] The modelling of positron emitter production and PET imaging during carbon ion therapy, F. Pönisch, K. Parodi, B.G. Hasch, W. Enghardt, Physics in Medicine and Biology 49, 5217 (2004)

[5] Experimental study on the feasibility of in-beam PET for accurate monitoring of proton therapy, K. Parodi, F. Pönisch and W. Enghardt, IEEE Transactions on Nuclear Science 52, 778 (2005)

[6] Patient study on in-vivo verification of beam delivery and range using PET/CT imaging after proton therapy, K. Parodi, H. Paganetti, H. Shih, S. Michaud, J. Loeffler, T. DeLaney, N. Liebsch, J. Munzenrider, A. Fischman, A. Knopf and T. Bortfeld, International Journal of Radiation Oncology, Biology, Physics, in press

[7] In-beam PET at high-energy photon beams: a feasibility study, H. Müller, W. Enghardt, Physics in Medicine and Biology 51, 1779 (2006)

[8] Quantification of β+ activity generated by hard photons by means of PET, D. Möckel, H. Müller, J. Pawelke, M. Sommer, E. Will, W. Enghardt, Physics in Medicine and Biology 52, 2515 (2007)

Project partners of the Institute of Radiation Physics involved in the first project:

  • Gesellschaft für Schwerionenforschung Darmstadt, Germany
  • Universitätsklinikum Heidelberg, Radiologische Klinik, Heidelberg, Germany
  • Deutsches Krebsforschungszentrum Heidelberg, Germany
  • Heidelberger Ionenstrahl-Therapiezentrum, Heidelberg, Germany
  • European Organization for Nuclear Research – CERN, Geneva, Switzerland
  • Siemens AG, Medical Solutions, Particle Therapy, Erlangen, Germany

Project partners of the Institute of Radiation Physics involved in the second project:

  • Karolinska Institute, Department of Oncology-Pathology, Stockholm, Sweden
  • Stockholm University, Mathematics-Physics Department, Stockholm, Sweden
  • Soltan Institute for Nuclear Studies, Otwock-Swierk, Poland
  • European Organization for Nuclear Research – CERN, Geneva, Switzerland
  • Technische Universität Dresden, Institut für Kern- und Teilchenphysik, Dresden, Germany
  • Forschungszentrum Dresden-Rossendorf, Institute of Radiopharmacy, Dresden, Germany

Project partners of the Institute of Radiation Physics involved in the third project:

  • Technische Universität Dresden, Institut für Kern- und Teilchenphysik, Dresden, Germany
  • Center for Radiation Research in Oncology – OncoRay, Technische Universität Dresden, Germany