between plans (p�0.12). Maximum PTV doses were significantly different (mean 6022 vs. 6616, p�0.001 for homogenous vs.heterogeneity-corrected plans, respectively). The change in volume of tissue treated to prescription dose was highly variable,dependent on tumor size, location, beam arrangement, and beam weighting. Mean prescription volume 32.44 cm3 homogenouscalculation vs. 37.67 cm3 (p�0.13) heterogeneity-corrected calculation. Mean lung dose and V20 were significantly higher(p�0.0001 and 0.01 respectively) for heterogeneity-corrected plans.

Conclusions: Using SBRT, heterogeneity-corrected methods produce significantly different PTV dose coverage and mean lungdoses. SBRT demonstrates a markedly sigmoidal dose-response (Wulf et al. IJROBP, 2005) suggesting small differences indose may result in different control and complication rates. Any comparison of published series should take heterogeneitycorrections into account. The magnitude of differences between individual plans can be even greater than what is shown for theentire series. Our institutional practice will now change to use plans calculated with heterogeneity corrections.

Author Disclosure: J.J. Urbanic, None; W.H. Hinson, None; V.W. Stieber, Elekta, G. Other; J.M. Butler, None; W.T. Kearns,None; C.J. Hampton, None; A.W. Blackstock, None.

2484 The Study on Correlation Between F-18 FDG PET/CT Standard Uptake Value and Clinical TargetVolume Definition in Radiotherapy for Non-Small Cell Lung Cancer

J. M. Yu1, X. Meng1, L. G. Xing1, D. B. Mu2, G. R. Yang3, Z. Fu3, X. D. Sun1, L. Kong1, W. X. Zhong2, A. Q. Han3

1Department of Radiotherapy, Shandong Tumor Hospital, Jinan, China, 2Department of Pathology, Shandong TumorHospital, Jinan, China, 3Positron Emission Tomography Center, Shandong Tumor Hospital, Jinan, China

Purpose/Objective(s): The definition of the clinical target volume (CTV) is the major difficulty in three-dimensional conformalradiotherapy (3DCRT) and intensity-modulated radiation therapy (IMRT) planning for non-small cell lung cancer (NSCLC).Minimal data are available for the influence of 18F-fluorodeoxyglucose Positron Emission Tomography (FDG PET) on definingCTV. The purpose of this study is to find the correlation between standard uptake values (SUV) of FDG PET/CT and the lineardistance of microscopic extension (MicExt), in order to probe into the impact of functional imaging on CTV definition.

Materials/Methods: From 2002 to 2004, 29 patients with operable NSCLC (19 with squamous carcinomas and 10 withadenocarcinomas) were included in the protocol. All patients had whole body FDG PET/CT scans prior to surgery. Scans wereperformed in a GE Discovery LS PET/ CT scanner after intravenous injection of 9–14 mCi 18F-FDG. A tracer uptake phaselasting about 60 minutes was implemented. Mean SUV (SUVmean) and maximal SUV (SUVmax) that were obtained byselecting a small region of interest (ROI) were used in the analysis respectively. A single pathologist (NSG) reviewed all slidesfrom each case. The tumor was transected postoperatively and the grossly visualized perimeter was outlined through lightmicroscopy with HE stained slices. It was measured the MicExt distance of carcinomars beyond perimeter at three dimensions.The maximal MicExt distance of each case was recorded. The SPSS software package version 10.0 was used for the statisticaldata analysis. The maximal MicExt distances were correlated with SUVmax or SUVmean by means of Pearson correlationcoefficient test respectively. P�0.05 was considered to indicate a significant difference.

Results: The average values of the SUVmax and SUVmean for all cases were 11.53 (5.22�30.00) and 8.05 (4.76�12.51)respectively. The average of SUV max and SUV mean for squamous carcinomas were 10.81(5.64�17.00) and 8.73(4.76�12.51). The average of SUV max and SUV mean for adenocarcinomas were 13.12 (5.22�30.00) and 6.25 (4.81�8.07).To all cases, the maximal MicExt distance ranged from 0.50mm to 8.90mm (3.22mm on the average). The average of themaximal MicExt distance for squamous carcinomas and adenocarcinomas were 4.09mm (0.50mm�8.90mm) and 1.59mm(0.50mm�3.00mm) respectively. The maximal MicExt diatance of all cases had the positive correlation with the SUV mean(r�0.82,P�0.05), but the negative correlation with the SUV max (r�0.43,P�0.05). To those cases with squamous carcinoma,the maximal MicExt distance also had the positive correlation with the SUV mean (r�0.77,P�0.05), but not with the SUV max(r�0.27,P�0.05). However, to those cases with adenocarcinomas, both the SUVmean and the SUV max had negativecorrelation with the maximal MicExt distance (r�0.86,P�0.05; r�0.94,P�0.05).

Conclusions: Conventional planning for NSCLC 3DCRT or IMRT does not customize the CTV definition to account for thepotential risk of MicExt of tumor in each individual patient. This may be result in significant reduction of target margin.Functional imaging such as 18FDG PET/CT may provide meaningful message for CTV definition.

Author Disclosure: J.M. Yu, None; X. Meng, None; L.G. Xing, None; D.B. Mu, None; G.R. Yang, None; Z. Fu, None; X.D.Sun, None; L. Kong, None; W.X. Zhong, None; A.Q. Han, None.

2485 Preliminary Report of Image-Guided Hypofractionated Stereotactic Body Radiotherapy to Treat PatientsWith Medically Inoperable Stage I or Isolated Peripheral Lung Recurrent Non-Small Cell Lung Cancer

J. Y. Chang, P. Balter, Z. Liao, J. D. Melenda, T. Guerrero, Y. Borghero, T. E. Tutt, J. A. Roth, J. D. Cox, R. Komaki

M.D. Anderson Cancer Center, Houston, TX

Purpose/Objective(s): The conventional treatment of medically inoperable stage I non-small cell lung cancer (NSCLC) isdefinitive radiotherapy to a dose of 60–66 Gy with 2 Gy/fraction (biologically effective dose, 79.2 Gy). However, the 2-yearlocal control ranges from 30% to 70%, and the long-term survival rate is about 30%. There is no standard treatment for isolatedrecurrent NSCLC. We evaluated the therapeutic efficacy and treatment-related toxicity of image-guided hypofractionatedstereotactic body radiotherapy (SBRT) in patients with medically inoperable stage I NSCLC or isolated peripheral lungrecurrent NSCLC.

Materials/Methods: Subjects had pathologically confirmed stage I (n�37) or isolated peripheral lung recurrent (n�22)NSCLC. Disease was staged with chest CT, PET, and brain MRI. Four-dimensional (4-D) CT images were obtained with a GEsimulator and a Varian RPM system. Internal gross target volume was delineated using maximal intensity projection createdby combining the data from the multiple 4-D CT datasets at different breath phases. Clinical target volume was internal grosstarget volume plus an 8-mm margin, and a 3-mm setup uncertainty margin was added to determine the planning target volume.

S480 I. J. Radiation Oncology ● Biology ● Physics Volume 66, Number 3, Supplement, 2006

Daily CT on-rail simulation was conducted during each fraction of radiotherapy. The prescribed dose was 50 Gy to planningtarget volume at daily 12.5 Gy/fraction (daily SBRT for 4 contiguous days). Patients were followed every 3 months for 2 yearswith chest CT. PET scan was recommended 3–5 months after SBRT.

Results: As of March 2006, 42 patients had been followed with median follow-up of 10 months (range, 3–18 months). Theoverall progression-free survival rate at the treated site was 100%. For stage Ia (T1N0M0) disease (n�22), the completeresponse (CR) rate was 66.7% and the partial response (PR) rate was 28.8%; the rate of stable disease (SD) was 4.5%. The CRrate was 100% if PET was used for post-SBRT evaluation (n�11). One patient developed mediastinal lymph node metastasisand distant metastasis. For stage Ib (T2N0M0) disease (n�3), CR, PR or SD was achieved in these three patients respectively.One patient developed mediastinal lymph node metastasis. For isolated peripheral lung recurrent disease (n�17; 15 had a tumorless than 3 cm), the CR rate was 72.7% and the PR rate was 27.3%; distant metastasis and mediastinal metastasis weredeveloped in two patients each. There was no grade II or higher radiation pneumonitis in patients with stage I disease. 17.6%of patients who had recurrent lung cancer and who had been previously treated by conventional radiotherapy or surgicalresection had worse dyspnea after SBRT. No esophagitis was noted. 9.5% of patients developed grade II dermatitis at the treatedsite that appears related to the dose (�35 Gy) and volume of the skin involved. All patients tolerated SBRT well without anysymptoms during the SBRT.

Conclusions: 4-D CT-based treatment planning and daily on-board image-guided SBRT was convenient, well tolerated withexcellent local progression-free survival in patients with stage I or isolated peripheral lung recurrent NSCLC. A regimen of 50Gy in four fractions (biologically effective dose, 112.5 Gy) appears adequate for local control using our image-guided SBRTapproach. Longer follow up is needed.

Author Disclosure: J.Y. Chang, None; P. Balter, None; Z. Liao, None; J.D. Melenda, None; T. Guerrero, None; Y. Borghero,None; T.E. Tutt, None; J.A. Roth, None; J.D. Cox, None; R. Komaki, None.

2486 Preserving Functional Lung Using Perfusion Imaging and Intensity-Modulated Radiotherapy forAdvanced-Stage Non-Small-Cell Lung Cancer

Y. Shioyama1, S. Jang2, H. H. Liu2, T. Guerrero1, X. Wang1, I. W. Gayed3, W. D. Erwin4, Z. Liao1, J. Y. Chang1,R. Komaki1

1Department of Radiation Oncology, MD Anderson Cancer Center, Houston, TX, 2Department of Radiation Physics, MDAnderson Cancer Center, Houston, TX, 3Department of Nuclear Medicine, MD Anderson Cancer Center, Houston, TX,4Department of Imaging Physics, MD Anderson Cancer Center, Houston, TX

Background: Patients with non-small-cell lung cancer (NSCLC) often have inhomogeneous lung perfusion and functiondistributions. Thus, dose distributions to normal lung can be tailored accordingly in minimizing radiation damage to healthyfunction lung and subsequent pulmonary toxicity.

Purpose/Objective(s): To investigate the degree of dose-volume reduction of irradiated functional lung using single photonemission computed tomography (SPECT) and intensity-modulated radiotherapy (IMRT) for advanced-stage NSCLC.

Materials/Methods: Ten patients with Stage III-IV and recurrent NSCLC who underwent definitive radiotherapy at ourinstitution were included in this study. Prior to radiotherapy, lung perfusion images were acquired using 99mTc-labeledmicro-aggregated albumin and SPECT. The SPECT images were fused with the simulation CT and were used to evaluateperfusion distributions of lung. Functional lung at the 50 and 90 percentile perfusion levels (“F50 lung” and “F90 lung”) weresegmented on the SPECT-CT images. IMRT plans were designed to deliver 63Gy to 95% of the planning target volume (PTV)using nine equidistant coplanar 6-MV beams. Inverse planning was performed to minimize the volumes of normal lung, heart,esophagus, and spinal cord irradiated above their tolerance doses in a regular IMRT plan using CT images only (“Anatomical-Plan”). A corresponding IMRT plan using the SPECT-CT images (“Functional-Plan”) was designed to minimize thedose-volumes of F50 and F90 lung in addition to the other objectives in the regular plan. Dose distributions and dosimetricparameters for the PTV and critical structures in both plans were computed and compared.

Results: In the Functional-Plan, the median absolute reductions of the mean lung dose for total, F50 and F90 lung were 1.1,2.2 and 4.2Gy, respectively, compared with that in the Anatomical-Plan. The median reductions in the percentage of volumeirradiated � 5Gy, � 10Gy and � 20Gy were 6.4%, 2.7% and 4.0% for the total lung, 7.9%, 5.7%, and 6.4% for F50 lung, and13.4%, 12.2% and 7.0% for F90 lung. These differences between the Functional-Plan and Anatomical-Plan were statisticallysignificant. In contrast, the heterogeneity index for PTV and dose-volume of heart in the Functional-Plan were comparable tothat in the Anatomical-Plan.

Conclusions: IMRT planning using SPECT perfusion images reduced the volumes of not only the perfused functional lung butalso those of total lung. Functional-image based IMRT planning may be effective in preserving normal lung function for locallyadvanced-stage NSCLC patients.

Author Disclosure: Y. Shioyama, None; S. Jang, None; H.H. Liu, None; T. Guerrero, None; X. Wang, None; I.W. Gayed, None;W.D. Erwin, None; Z. Liao, None; J.Y. Chang, None; R. Komaki, None.

Reduction of the dose-volume of the irradiated lung in Functional-Plan compared to Anatomical-Plan

Parameter Total Lung F50 F90

Mean dose (Gy) 1.1 (0.1-1.8) 2.2 (0.7-3.1) 4.2 (0.7-9.0)V5 (%) 6.4 (1.2-8.9) 7.9 (4.5-15.0) 13.4 (1.7-33.9)V10 (%) 2.7 (-0.4-11.5) 5.7 (2.1-17.2) 12.2 (0.5-27.1)V20 (%) 4.0 (-0.7-6.4) 6.4 (0.7-12.2) 7.0 (0.4-25.9)

Abbreviations: V5, V10 and V20: irradiated volume more than 5 Gy, 10 Gy and 20 Gy; F50 and F90: 50 and 90 percentile perfusion area inlung. Data presented as the median, with the range in parenthesis.

S481Proceedings of the 48th Annual ASTRO Meeting