Radiotherapy and Oncology

Transcription

1 Radiotherapy and Oncology 99 (2011) Contents lists available at ScienceDirect Radiotherapy and Oncology journal homepage: PET treatment planning FET PET for malignant glioma treatment planning Maximilian Niyazi a, Julia Geisler b, Axel Siefert a, Silke Birgit Schwarz a, Ute Ganswindt a, Sylvia Garny a, Oliver Schnell c, Bogdana Suchorska c, Friedrich-Wilhelm Kreth c, Jörg-Christian Tonn c, Peter Bartenstein b, Christian la Fougère b, Claus Belka a, a Department of Radiation Oncology; b Department of Nuclear Medicine; and c Department of Neurosurgery, Ludwig-Maximilians-University Munich, München, Germany article info abstract Article history: Received 25 September 2010 Received in revised form 28 February 2011 Accepted 10 March 2011 Keywords: FET PET Planning study Malignant glioma BTV Target delineation Background and purpose: The aim of this study was to compare MRI-based morphological gross tumour volumes (GTVs) to biological tumour volumes (BTVs), defined by the pathological radiotracer uptake in positron emission tomography (PET) imaging with 18 F-fluoroethyltyrosine (FET), subsequently clinical target volumes (CTVs) and finally planning target volumes (PTVs) for radiotherapy planning of glioblastoma. Patients and methods: Seventeen patients with glioblastoma were included into a retrospective protocol. Treatment-planning was performed using clinical target volume (CTV = BTV + 20 mm or CTV = GTV + 20 mm + inclusion of the edema) and planning target volume (PTV = CTV + 5 mm). Image fusion and target volume delineation were performed with OTP-Masterplan Ò. Initial gross tumour volume (GTV) definition was based on MRI data only or FET PET data only (BTV), secondarily both data sets were used to define a common CTV. Results: FET based BTVs (median 43.9 cm 3 ) were larger than corresponding GTVs (median 34.1 cm 3, p = 0.028), in 11 of 17 cases there were major differences between GTV/BTV. To evaluate the conformity of both planning methods, the index (CTV MRT \ CTV FET )/(CTV MRT [ CTV FET ) was quantified which was significantly different from 1 (0.73 ± 0.03, p < 0.001). Conclusion: With FET PET-CT planning, the size and geometrical location of GTVs/BTVs differed in a majority of patients. It remains open whether FET PET-based target definition has a relevant clinical impact for treatment planning. Ó 2011 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 99 (2011) Glioblastoma is the most common and aggressive primary brain tumour [1]. Malignant gliomas account for approximately 70% of new cases of malignant primary brain tumours that are diagnosed in adults. Radiotherapy (RT) is the mainstay of treatment for malignant glioma, even in case of recurrence [2]. The addition of radiotherapy to surgery increases survival among patients with glioblastoma from a range of 3 4 months to a range of 7 12 months [3,4]. Corresponding author. Address: Department of Radiation Oncology, Ludwig-Maximilians-University Munich, Marchioninistr. 15, München, Germany. addresses: maximilian.niyazi@med.uni-muenchen.de (M. Niyazi), julia. geisler@med.uni-muenchen.de (J. Geisler), axel.siefert@med.uni-muenchen.de (A. Siefert), silke.schwarz@med.uni-muenchen.de (S.B. Schwarz), ute.ganswindt@ med.uni-muenchen.de (U. Ganswindt), sylvia.garny@med.uni-muenchen.de (S. Garny), oliver.schnell@med.uni-muenchen.de (O. Schnell), bogdana.suchorska@ med.uni-muenchen.de (B. Suchorska), friedrich-wilhelm.kreth@med. uni-muenchen.de (F.-W. Kreth), joerg.christian.tonn@med.uni-muenchen.de (J.-C. Tonn), peter.bartenstein@med.uni-muenchen.de (P. Bartenstein), christian. lafougere@med.uni-muenchen.de (C. la Fougère), claus.belka@med.uni-muenchen. de (C. Belka). Although imaging techniques and multimodal treatment strategies have improved since the mid-1980s, little impact has been made on the ultimate prognosis of glioblastoma. Standard treatment for glioblastoma was significantly improved only following the results of a large phase III trial conducted by the EORTC and the National Cancer Institute of Canada (NCIC) [4]. The EORTC NCIC CE3 trial randomized 573 patients with newly-diagnosed glioblastoma to RT alone or to RT in combination with the oral alkylating agent temozolomide (TMZ). Concurrent and adjuvant administration of TMZ improved two-year survival of patients from 11.2% to 27.3%, three-year survival from 4.4% to 16.0% and five-year survival from 1.9% to 9.8% [4,5]. Successful 3D-conformal radiation therapy necessitates an accurate and reliable definition of the extent of viable tumour. Even if conventional imaging techniques like MRI and CT provide excellent anatomical details, additional information from molecular imaging techniques like positron emission tomography (PET) are currently discussed to have a major impact on radiation oncology [6 9], because they enable to visualise biologic pathways in vivo. Therefore, the combination of morphological and /$ - see front matter Ó 2011 Elsevier Ireland Ltd. All rights reserved. doi: /j.radonc

2 M. Niyazi et al. / Radiotherapy and Oncology 99 (2011) molecular imaging techniques for radiation planning purposes may provide precise anatomical and biological target delineation to guide radiation dose delivery. As such, biologically image-guided radiation therapy (RT), coupled to the current anatomical imaging technology, will deliver optimally radiation, with a high-degree of geometrical precision [10] and biological conformity. Different PET-radiotracers have been introduced as suitable PET tracer in glioma [11]. Among these radiolabelled amino acids were shown to be particularly attractive for imaging brain tumours because of the high uptake in tumour tissue and low uptake in normal brain that results in higher tumour-to-normal-tissue contrast [12], especially when compared to the most widely used radiotracer 18 F-fluorodeoxyglucose (FDG), which comes with high tracer uptake due to the high glucose metabolism in the cerebral cortex. The use of amino acids is based upon the observation of generally increased amino acid transport in malignant transformation [13]. In this study the fluorinated amino acid analogue 18 F-fluoroethyltyrosine (FET) was used which was shown to enter the cell by specific amino acids transporters, but is not incorporated into proteins [14]. The high in vivo stability of FET, its more appropriate half-life when compared to 11 C-labelled radiotracers like 11 C- methionine (MET) (110 min vs. 20 min), its fast brain and tumour uptake kinetics, its low accumulation in non-tumour tissue and its ease of synthesis make it an optimal amino acid tracer for cerebral tumours. Since literature data indicate that both FET and MET have a similar performance regarding biopsy location, tumour delineation, and detection of recurrence [15], FET PET was suggested to be appropriate for accurate RT-planning [16,17]. We retrospectively analysed a series of 17 glioblastoma patients in whom CT/MRI based treatment planning was complemented by FET PET. The purpose of this study was to quantify the BTVs and GTVs, their corresponding CTVs/PTVs and to assess their volumetric and geometric relationships. Patients and methods Patients Seventeen patients with glioblastoma received diagnostic MRI, RT planning CT and additional FET PET prior to 3D conformal RT between 7/2004 and 7/2009 in our institution. Median age at treatment was 66.2 years (range years). The male/female ratio was 13/4, 5 of the 17 patients underwent surgical tumour resection, stereotactic biopsy was performed in the remaining 12 cases. Only WHO grade IV glioma was included into this retrospective study. Tumours were mainly located temporal (N = 10), the other locations were parietal (N = 4), frontal (N = 2), occipital (N = 2), cerebellar (N = 1) and thalamic (N = 1); 11 tumours were located in the left hemisphere. Baseline characteristics of the patients are summarised in Table 1. FET PET The standardised FET PET-protocol has been described and validated elsewhere [18]. In brief, FET PET scans were obtained with a Siemens ECAT EXACT HR + scanner. To obtain standardised metabolic conditions, patients fasted for a minimum of 6 h before performing the PET scan. The scanner acquires 63 contiguous transaxial planes, simultaneously covering 15.5 cm of axial field of view. After a 15 min transmission scan ( 68 Ge-sources), approximately 180 MBq 18 F-FET was injected intravenously and emission data were acquired for 40 min. Images were reconstructed by filtered back projection using a Hann filter and were corrected for scatter and attenuation. For further evaluation, data were transferred to a HERMES work station (Hermes Medical Solutions, Sweden) where standard quantification procedures were performed. Table 1 Patient characteristics. All glioblastoma patients (N = 17) received radiotherapy with 60 Gy as an adjuvant or primary therapy. Characteristic Patients (N = 17) Sex Male 13 (76.5%) Female 4 (23.5%) Median age [yr] 66.2 (45 79) MGMT methylation status Methylated 6 (35.3%) Not methylated 5 (29.4%) Unknown 6 (35.3%) Surgery Yes 5 (29.4%) No 12 (70.6%) Location of the tumour Frontal 2 Parietal 4 Occipital 2 Cerebellar 1 Temporal 10 Thalamic 1 Left/right 11/6 Active tumour volume was semi-automatically defined with the software Hermes Hybrid PDR on coregistered FET PET/MRI using a fixed threshold SUV max /background ratio of 1.5. This procedure resulted in a three-dimensional volume of interest (3D-VOI) which represents the FET-positive tumour tissue and which was afterwards transferred to the OTP-Masterplan Ò package (Theranostic GmbH, Solingen, Germany) for further data processing. Treatment planning and target volume definition 3D-conformal RT treatment planning was primarily based on diagnostic MRI data and was secondarily evaluated solely based on additional FET PET information. In a third step, MRI was complemented by the information from FET PET. RT planning was performed on a 3D-data set generated from 3 mm CT scans in treatment position. For immobilization of the head an individual thermoplastic head mask fixation was used. Image fusion of diagnostic MRI, RT planning CT and PET as well as target volume delineation was done with the OTP-Masterplan Ò package (Theranostic GmbH, Solingen, Germany). The CT planning images in mask fixation were fused with the diagnostic MRI images coregistered with PET using the automatic matching algorithm stored in the OTP- Masterplan Ò system. For gross tumour volume (GTV) delineation the initial macroscopic tumour volume definition was based on (preoperative) MRI findings and RT planning CT information only (referred to as GTV). Subsequently the PET positive tumour lesions were defined by the same radiation oncologist (biological tumour volume BTV) based on the volumes delineated by nuclear medicine experts. Both GTV and BTV were counterchecked by another radiation oncologist (see Fig. 1). A 2 cm margin was chosen for both gross tumour volumes. The MRI based CTV (clinical tumour volume) was delineated including the edema and excluding bone and other anatomical barriers [19]. Finally the CTV was expanded with an overall safety margin of 5 mm to the PTV (planning target volume). Quantitative analysis of tumour volumes For quantification of target volume changes based on the PET findings both tumour (GTV-MRI/CT and BTV) and intersection volumes between the GTV-MRI/CT and BTV (intersection-gtv-mri/ CT/PET) were evaluated. The same procedure was performed for the CTVs. For both modalities we computed the conformity index:

3 46 FET PET planning in malignant glioma Table 2 Tumour volumes. Measurement of different tumour volumes in 17 glioblastoma patients. BTV [cc] BTV + 2 [cc] GTV [cc] CTV [cc] Fig. 1. Target volume concept showing BTV and a margin of 2 cm. The generated CTV is extended by 5 mm to create the PTV. Additionally, the definition and meaning of the conformity index is shown. This ratio is calculated by dividing intersection volume by conjunction volume. the ratio between intersection volume and the conjunction of both volumes (Fig. 1). In order to report these volume parameters for the whole patient collective pure descriptive statistics (median, range) were used. Statistical analysis We performed all analyses using the Statistical Package for Social Sciences (SPSS, Ver. 18.0, SPSS Inc., Chicago, IL). For descriptive analyses of patients characteristics and volumes sizes we used percents and median score. The GTV, BTV and CTV delineation methods were compared using the Wilcoxon signed-rank test as numerical data were not normally distributed (paired tests); no corrections for multiple comparisons were performed. A two-tailed p-value of less than 0.05 was considered to indicate statistical significance. Results The GTV-MRI/CT included the macroscopic tumour visible in the planning CT and contrast-enhanced T1-weighted MRI. All glioblastomas could be delineated on MRI and CT. Median GTV-MRI/CT was 34.1 cubic centimetre (cc, range cc). There were several FET PET positive lesions (e.g., large vessels) without any further suspicious tumour detection in an additional CT or MRI examination. Target volume definition based on FET PET (BTV) included the tumour volume with an intense tracer uptake, all 17 glioblastomas showed a high tumour-to-background contrast. For the target volume definition the windowing of the FET PET was determined by a fixed SUV max /background threshold ratio of 1.5. The median BTV was 43.9 cc (range cc). In general, FET PET data led to additional information concerning the viable tumour extension. Overall the BTV was larger than the GTV-MRI in 10 patients, smaller in three patients and almost the same in four patients (62.5 cc deviation). Considering median values, there was a trend towards larger BTVs compared to corresponding GTVs (43.9 cc vs cc, p = 0.028; Table 2). Concerning CTVs, 12 major changes occurred; all CTVs were different, in 11 cases the CTV-FET was larger, in six cases vice versa. Median CTVs were not significantly different: cc vs cc, p = Median 43.9 Median Median 34.1 Median Though the volume differences did not reach statistical significance (probably due to high inter-case variance), the conformity index was significantly different from one. This means that the geometrical structure or localisation of CTV-MRI and CTV-FET are significantly different (0.73 ± 0.03, p < 0.001, t-test) an information that is more reliable than pure equality of volume parameters. This number was calculated as ratio between intersection and conjunction volume. A ratio of 0 means that there is no overlap the volumes are completely different (though they might have the same volume). Contrarily, a ratio of one would be a perfect overlap. Furthermore, the union of CTV-MRI and CTV-FET was created as a solely mathematical conjunction. Similarly, a final CTV was created using both the information of MRI and FET; it was called CTV final. The median CTV final was significantly larger than the initial CTV-MRI (286.8 cc vs cc, p < 0.001). Interestingly, the CTV final was significantly smaller than the corresponding median conjunction volume CTV union (286.8 cc vs cc, p = 0.002). The PTV was created from the CTV with an overall safety margin of 5 mm. Median PTV-MRI was cc (range, cc), median PTV-FET was cc (range, cc), median PTV final was cc (range, cc). Statistical comparisons revealed that PTV-MRI was significantly smaller than PTV final (p < 0.001). Concerning the first site of tumour progression, follow-up imaging of 12 patients was available. Regarding patients with MRIbased treatment planning, first progression was outside the GTV- MRI in three cases, inside the GTV in six cases and marginal (inand outside the GTV) in three cases. With FET PET planning, first progression was outside the BTV in two cases, inside the BTV in six cases and marginal (in- and outside the BTV) in four cases. Discussion A wide range of publications has documented the value of external beam radiation for the treatment of glioblastoma. According to the ICRU definition, CTV should include all region of possible microscopic spread. Several series have shown undisputedly that tumour infiltration, proven with stereotactic biopsies, was identified in areas congruent with abnormal signal on MRI images [20,21]. In a biopsy-controlled glioma study, FET PET improved the tumour extension delineation by the combined use of FET PET and MRI or CT, in comparison with conventional imaging alone [22]. We

4 M. Niyazi et al. / Radiotherapy and Oncology 99 (2011) are presently left with the question of how to optimally integrate these various imaging modalities for tumour delineation. Despite its high sensitivity (>90%) and specificity (>80%) for diagnosis of glioma [18], the positive predictive value of FET PET (84%) did not permit the replacement of stereotactic biopsy in patients with an MRI-based suspicion of a glioma recurrence or progression [23]. However, FET PET was shown to be a powerful tool to improve the differential diagnosis in these patients [24] because it reliably distinguishes between post-therapeutic benign lesions and tumour recurrence after initial treatment of low- and high-grade gliomas which makes it a valuable diagnostic tool for pseudoprogression [25,26]. FET uptake kinetics is an important factor in this distinction [27]. This is very important in patients with glioma undergoing multimodal treatment or various forms of irradiation, since conventional follow-up with MRI is insufficient to distinguish between benign side effects of therapy and tumour recurrence. However, the role of FET PET for radiation therapy planning is not adequately assessed yet. In our approach, BTVs turned out to have a larger extension than classical GTVs defined by MRI. This may partly be explained by FET enhancement of small viable tumour islands which may not be clearly identified by MRI. This hypothesis is corroborated by the fact that the conformity index was significantly different from one a hint on different coverage of distinct target volumes. Interestingly, in one case the relative recurrence pattern between BTV and GTV differed substantially as the location was completely outside the GTV but mainly inside the BTV (marginal recurrence). As this was only one case, this does not allow for final conclusions, but warrants further investigations. An important consideration in this context is the open question if there is a SUV threshold to define the BTV. In our study for the target volume definition a fixed threshold value for SUV max /background was used (see Fig. 2) though up to now clear evidence for a certain SUV cut-off in delineation of glioblastomas is missing. This was compared to a more heuristic method: windowing of the FET PET was determined by the matching between the PETpositive areas and the viewable tumour margins determined by CT/MRI. The physiological signal of large blood vessels was masked out via windowing. The median target volumes were not significantly different from our systematic approach (BTVs: 43.9 vs cc, p = 0.554; CTVs: vs cc, p = 0.981; PTVs: vs cc, p = 0.906). The group of Vees et al. has investigated various strategies for FET PET high-grade tumour delineation using various functional image segmentation algorithms [28], but it remains to be determined which segmentation technique is the most appropriate for glioma delineation. The core problem to find an exact relationship between signal intensity and vital tumour tissue has not been solved to date. With reference to the FET PET planning study by Weber et al. [17] and a study with another amino acid tracer (3-O-methyl-6- [F-18]fluoro-L-DOPA, OMFD) by Alheit et al. [9], our study uses a different thresholding technique and compares it to a heuristic method. BTV and GTV did not differ significantly in the study of Weber et al. and there was no direct comparison of planning methods when FET PET or MRI was applied alone. On the one hand, our study compared FET PET planning only, MRI planning only and combined planning against each other, on the other hand MRI planning was ICRU conformal. As a further new aspect a recurrence pattern analysis was performed. Concerning the clinical impact, several questions have to be answered in the future. With conventional planning, it seems reasonable to study the recurrence pattern more systematically and with a higher number of cases [29]; if FET uptake translates into a higher rate of recurrence locations, the viability has a direct impact on tumour control probability. Furthermore, in this case it may be advantageous to administer a simultaneous integrated boost (SIB) to the BTV/FET-positive regions [30,31]. Conclusions With FET PET-CT planning, the size and geometrical location of GTVs and BTVs differed in a majority of patients. FET PET information complements image data from MRI and CT. Therefore, we would recommend the use of the FET as additional information for GTV definition in all cases with macroscopic tumour (either in the primary situation or as a residual after surgery). Further evaluation with a larger number of patients seems to be justified and long-term follow-up is needed to evaluate the clinical impact. Conflict of interest statement The authors declare that they have no competing interests. Authors contributions C.B., A.S. & M.N. planned, coordinated and conducted the study. S.G. provided information on physical treatment planning. J.G., C.L.F. & P.B. performed PET imaging and provided the standard for FET-contouring. M.N. & A.S. analysed the PET and MRI imaging data as well as the treatment planning data. M.N., U.G., A.S., S.B.S., B.S., O.S. & C.B. prepared the manuscript. Medical care was covered by J.C.T., F.W.K. & C.B. All authors read and approved the final manuscript. Acknowledgement We thank Dr. Markus Diemling from Hermes Medical Solutions for the technical assistance and for providing the Hybrid PDR software. References Fig. 2. The macroscopic tumour on MRI (two different slices) and the corresponding PET VOI with a threshold of SUV max /background = 1.5. [1] Louis DN, Ohgaki H, Wiestler OD, et al. The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol 2007;114: [2] Niyazi M, Siefert A, Schwarz SB, et al. Therapeutic options for recurrent malignant glioma. Radiother Oncol 2011;98:1 14. [3] Walker MD, Alexander Jr E, Hunt WE, et al. Evaluation of BCNU and/or radiotherapy in the treatment of anaplastic gliomas. A cooperative clinical trial. J Neurosurg 1978;49: [4] Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 2005;352: [5] Stupp R, Hegi ME, Mason WP, et al. Effects of radiotherapy with concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the EORTC NCIC trial. Lancet Oncol 2009;10: [6] Grosu AL, Piert M, Weber WA, et al. Positron emission tomography for radiation treatment planning. Strahlenther Onkol 2005;181:

5 48 FET PET planning in malignant glioma [7] Gregoire V, Chiti A. PET in radiotherapy planning: particularly exquisite test or pending and experimental tool? Radiother Oncol 2010;96: [8] Chiti A, Kirienko M, Gregoire V. Clinical use of PET-CT data for radiotherapy planning: what are we looking for? Radiother Oncol 2010;96: [9] Alheit H, Oehme L, Winkler C, et al. Radiation treatment planning in brain tumours potential impact of 3-O-methyl-6-F-18 fluoro-l-dopa and PET. Nuklearmedizin Nucl Med 2008;47: [10] Wagner D, Christiansen H, Wolff H, Vorwerk H. Radiotherapy of malignant gliomas: comparison of volumetric single arc technique (RapidArc), dynamic intensity-modulated technique and 3D conformal technique. Radiother Oncol 2009;93: [11] Grosu AL, Weber WA. PET for radiation treatment planning of brain tumours. Radiother Oncol 2010;96: [12] Chen W. Clinical applications of PET in brain tumors. J Nucl Med 2007;48: [13] Isselbacher KJ. Sugar and amino acid transport by cells in culture differences between normal and malignant cells. N Engl J Med 1972;286: [14] Langen KJ, Hamacher K, Weckesser M, et al. O-(2-F-18 Fluoroethyl)-L-tyrosine: uptake mechanisms and clinical applications. Nucl Med Biol 2006;33: [15] Moulin-Romsee G, D Hondt E, de Groot T, et al. Non-invasive grading of brain tumours using dynamic amino acid PET imaging: does it work for 11 C- methionine? Eur J Nucl Med Mol Imaging 2007;34: [16] Gross MW, Weber WA, Feldmann HJ, Bartenstein P, Schwaiger M, Molls M. The value of F-18-fluorodeoxyglucose PET for the 3-D radiation treatment planning of malignant gliomas. Int J Radiat Oncol Biol Phys 1998;41: [17] Weber DC, Zilli T, Buchegger F, et al. [(18)F]Fluoroethyltyrosine positron emission tomography-guided radiotherapy for high-grade glioma. Radiat Oncol 2008;3:44. [18] Pauleit D, Floeth F, Hamacher K, et al. O-(2-F-18 fluoroethyl)-l-tyrosine PET combined with MRI improves the diagnostic assessment of cerebral gliomas. Brain 2005;128: [19] Kantor G, Loiseau H. Analysis of target volumes for gliomas. Cancer Radiother 2005;9: [20] Watanabe M, Tanaka R, Takeda N. Magnetic resonance imaging and histopathology of cerebral gliomas. Neuroradiology 1992;34: [21] Kelly PJ, Daumas-Duport C, Kispert DB, Kall BA, Scheithauer BW, Illig JJ. Imaging-based stereotaxic serial biopsies in untreated intracranial glial neoplasms. J Neurosurg 1987;66: [22] Langen KJ, Floeth FW, Stoffels G, Hamacher K, Coenen HH, Pauleit D. Improved diagnostics of cerebral gliomas using FET PET. Z Med Phys 2007;17: [23] Mehrkens JH, Popperl G, Rachinger W, et al. The positive predictive value of O- (2-[18F]fluoroethyl)-L-tyrosine (FET) PET in the diagnosis of a glioma recurrence after multimodal treatment. J Neurooncol 2008;88: [24] Rachinger W, Goetz C, Popperl G, et al. Positron emission tomography with O- (2-[18F]fluoroethyl)-L-tyrosine versus magnetic resonance imaging in the diagnosis of recurrent gliomas. Neurosurgery 2005;57: discussion [25] Bertolini F, Bagni B, Valentini A, et al. 18F-FET PET in early detection of relapse/ progression in high-grade gliomas. Neuro-Oncol 2008;10: [26] Brandes AA, Tosoni A, Spagnolli F, et al. Disease progression or pseudoprogression after concomitant radiochemotherapy treatment: pitfalls in neurooncology. Neuro-Oncol 2008;10: [27] Popperl G, Kreth FW, Herms J, et al. Analysis of 18F-FET PET for grading of recurrent gliomas: is evaluation of uptake kinetics superior to standard methods? J Nucl Med 2006;47: [28] Vees H, Senthamizhchelvan S, Miralbell R, Weber DC, Ratib O, Zaidi H. Assessment of various strategies for F-18-FET PET-guided delineation of target volumes in high-grade glioma patients. Eur J Nucl Med Mol Imaging 2009;36: [29] Weber DC, Casanova N, Zilli T, et al. Recurrence pattern after [(18)F]fluoroethyltyrosine-positron emission tomography-guided radiotherapy for high-grade glioma: a prospective study. Radiother Oncol 2009;93: [30] Douglas JG, Stelzer KJ, Mankoff DA, et al. F-18-Fluorodeoxyglucose positron emission tomography for targeting radiation dose escalation for patients with glioblastoma multiforme: clinical outcomes and patterns of failure. Int J Radiat Oncol Biol Phys 2006;64: [31] Piroth MD, Pinkawa M, Holy R, et al. Integrated-boost IMRT or 3-D-CRT using FET PET based auto-contoured target volume delineation for glioblastoma multiforme a dosimetric comparison. Radiat Oncol 2009;4.