- Email: [email protected]
MRI-based Treatment Planning for Prostate Brachytherapy
I. J. Radiation Oncology d Biology d Physics
S372
2376
Volume 78, Number 3, Supplement, 2010
Interfraction Seminal Vesicle Motion and Target Margin Assessment for Fiducial-Guided Intensity Modulated Radiotherapy for Prostate Cancer
M. H. Stenmark, K. A. Vineberg, D. W. Litzenberg, D. A. Hamstra, M. Feng University of Michigan, Ann Arbor, MI Purpose/Objective(s): To determine if differential interfraction motion necessitates specific planning target volume (PTV) margins for the seminal vesicles (SV) during fiducial-guided radiotherapy of the prostate. Materials/Methods: 20 patients were enrolled in an IRB-approved prospective study, which included fiducial implantation and 3 sets of CTs during treatment. SV motion relative to the prostate was analyzed by CT at 60 interfraction time points for these patients, assuming perfect set-up to the prostate with daily guidance. Following mutual information image registration of the prostate for each instance, the simulation SVs were expanded in 1 mm increments until the expansion fully covered the union of SV volumes for each patient to evaluate the geometric extent of motion. To evaluate the dosimetric consequences of variable SV positions, two intensity-modulated radiotherapy (IMRT) plans were generated for each patient including, 1) the prostate and proximal (1cm) SV (PSV), and 2) the prostate and full SV (FSV), both with a standard 5 mm PTV margin for optimization. Treatment planning was performed by a single operator using a multi-criteria hierarchical optimization method. Dosimetric coverage of both the proximal and full SVs was assessed at each instance and compared with the initial treatment plan. Results: Positional variability was greater for the FSV as compared to the PSV. A 5 mm PSV-PTV margin yielded 95% geometric coverage (C95%) in 95% of cases; median coverage of 99% (range 93-100%, SD 2%). However, a 5 mm FSV-PTV margin only resulted in C95% SV in 55% of cases; median coverage of 96% (83%-100%, SD 5%). To achieve C95% for the FSV in .95% of cases an 8 mm margin was required (median 99%, 95-100%, SD 2%). To obtain C99% in .95% of cases, the PSV required a 7mm expansion, while the FSV required a 10 mm expansion. Despite the lack of full geometric coverage of all PSV instances with a 5mm PTV expansion, dosimetric coverage was adequate. For plans optimized to the prostate and PSV, motion did not impact the volume of PSV covered by the 100% isodose line (V100%), mean 95% (SD 3%), or the 95% isodose line (V95%), mean 99% (SD 1%). For plans optimized to the FSV, dosimetric coverage was compromised in some patients, as V95% was .95% in only 70% of cases. However, overall, mean V100% was 89% (range 65%-98%, SD 8%) and mean V95% was 96%, (range 89%-100%, SD 4%). Conclusions: The SVs move differentially from the prostate and exhibit greater variation with increasing distance from the prostate. For plans targeting just the prostate and PSVs, 5mm PTV expansions are adequate. However, despite daily localization of the prostate, larger PTV margins are required for cases where the intent is to cover the FSVs. Supported by NIH/NCI (1 R21 CA110485-01A1). Author Disclosure: M.H. Stenmark, None; K.A. Vineberg, None; D.W. Litzenberg, None; D.A. Hamstra, None; M. Feng, None.
2377
The Use of MRI Imaging in IMRT Treatment Planning for Prostate Cancer
S. Vora, A. C. Silva, W. W. Wong, J. M. Collins, S. E. Schild Mayo Clinic Arizona, Phoenix, AZ Purpose/Objective(s): Dose escalation improves biochemical control rates based on phase III trials. However, higher doses have been associated with increased risk of bladder and rectal toxicity. Improvements in MRI imaging allow better definition of suspicious areas within the prostate. The purpose of this study is to assess the clinical utility of incorporating MRI imaging and its’ impact in treatment planning. Materials/Methods: 25 patients with clinically localized prostate adenocarcinoma underwent 1.5 T 8-channel surface coil MRI of the prostate using high resolution axial, coronal, and sagittal T2 weighted series, axial T1 weighted series, contrast enhanced dynamic and delayed T1 weighted fast gradient-echo series and a diffusion weighted sequence. No endorectal coil was used. The MRI was performed following placement of internal fiducial markers and CT simulation. The MRI images were reviewed for the following factors: early arterial enhancement, restricted diffusion, and T2 hypointensity. If patients had $2 factors, the area was identified as suspicious. MRI images were fused with the planning CT scan. The entire prostate+3mm margin (PTV) was planned to receive 77.4 Gy/43 fractions (1.8 Gy daily dose) and MRI suspicious areas (GTV) received concomitant boost to 83 Gy/43 fractions (1.93 Gy daily dose). Rectal/Bladder tolerance doses were maintained to keep D30\70 Gy, D10 \75 Gy, D1.8cc\81 Gy. If the constraints could not be met or if imaging was inconclusive, the GTV dose was either reduced to 81 Gy or no boost was delivered. Results: Of 25 patients, 22 (88%) patients had MRI imaging that contained suspicious areas consistent with malignancy. IMRT treatment plans generated .99% coverage of the PTV and .97% coverage of the GTV. 15 pts (68%) received boost dose of 83 Gy and the remaining 7 (32%) received 81 Gy. Review of all MRI imaging demonstrated that placing the internal fiducial markers prior to the MRI caused artifacts which can alter the GTV. Nine patients (36%) had evidence of post-fiducial placement artifact with hemorrhage most commonly observed. Conclusions: We found MRI imaging helpful in IMRT treatment planning, allowing us to define a tumor volume for dose escalation. Longer follow-up will be needed to assess biochemical control and toxicity. We now perform MRI imaging prior to internal fiducial marker placement to reduce artifacts and improve the accuracy when defining the GTV. Author Disclosure: S. Vora, None; A.C. Silva, None; W.W. Wong, None; J.M. Collins, None; S.E. Schild, None.
2378
MRI-based Treatment Planning for Prostate Brachytherapy
J. M. Albert, M. Zhang, T. L. Bruno, R. J. Kudchadker, D. A. Swanson, S. J. Frank M. D. Anderson Cancer Center, Houston, TX Purpose/Objective(s): Transrectal ultrasound (TRUS) is the standard imaging modality for brachytherapy treatment planning. However, MRI provides better anatomical delineation of the prostate. With more precise anatomic detail, MRI may confer advantages during treatment planning. Endorectal coil MRI (ecMRI), often used for staging, provides even greater anatomical resolution
Proceedings of the 52nd Annual ASTRO Meeting than standard MRI (sMRI). We aim to compare prostate volume, dosimetric parameters, and activity between plans based on sMRI and ecMRI versus TRUS. Materials/Methods: Patients were treated with I-125 brachytherapy in a prospective phase II trial at M.D. Anderson Cancer Center. TRUS-based pre-planning was utilized, and all patients received a pre-treatment ecMRI and 30-day post-implant sMRI. Blinded to TRUS-based plans, we re-planned treatments on T2-weighted sMRI and ecMRI for 20 consecutive patients. Varian’s VariSeed 8.0 was used for planning. Contours were approved by two reviewers and plans were jointly developed by a medical dosimetrist and radiation oncologist using a standard modified peripheral loading technique. All treatment plans satisfied the following dosimetric parameters: PTV V100. 95%, V150\ 60%, V200\ 20%, R100\ 1cc, and U200\ 5%. Prostate volume/dimensions, activity to volume ratio, and dosimetry were compared using a paired t-test and Wilcoxon signed rank test as appropriate. Results: Compared to TRUS, mean ecMRI prostate volume was smaller (29.5 vs. 32.5 cc, p = 0.001). This corresponded with an increased medial-lateral diameter (5.01 vs. 4.65 cm, p \ 0.001) and a decreased anterior-posterior diameter (2.69 vs. 3.06 cm, p\0.001), suggesting that the endorectal coil caused substantial anatomical distortion. Interestingly, there was a decreased craniocaudal length on both types of MRI compared to TRUS (TRUS: 4.23 cm; ecMRI: 3.71 cm, p \ 0.001; sMRI: 3.55 cm, p \ 0.001). This suggests that TRUS may overestimate prostate length, resulting in seed implantation in the genitourinary diaphragm or penile bulb, a hypothesis which was confirmed by review of post-implant MRIs. There was a significant decrease in activity per volume when using sMRI compared to TRUS (0.901 vs. 0.974 mCi/cc, p \ 0.001). Despite using less activity, sMRI plans had slightly improved PTV coverage (V100: 97.3% vs. 96.2%, p \ 0.001). sMRI-based plans also had improved dose homogeneity (V150: 47.4% vs. 53.8%, p \ 0.001 and V200: 16.6% vs. 19.2%, p \ 0.001). R100 was less than 1 cc and U200 was less than 0.07 cc on all plans. Conclusions: MRI more accurately defines prostate anatomy than TRUS. ecMRI is useful for staging, but should not be used for treatment planning due to anatomical distortion of the prostate. sMRI allows improved prostate delineation compared to TRUS, which results in less activity per volume required to achieve adequate PTV coverage and improved homogeneity. Author Disclosure: J.M. Albert, None; M. Zhang, None; T.L. Bruno, None; R.J. Kudchadker, None; D.A. Swanson, None; S.J. Frank, None.
2379
Initial Clinical Experience with Daily Cone Beam CT for Target Localization in Prostate Cancer Patients following Robotic Prostatectomy
J. P. Einck, D. R. Simpson, S. K. Nath, R. A. Sethi, J. Z. Wang, A. J. Mundt, A. P. Sandhu University of California San Diego, La Jolla, CA Purpose/Objective(s): Intensity-modulated radiotherapy (IMRT) in post-prostatectomy patients (pts) has increased the demand for more accurate treatment delivery using image-guidance. Previous studies have reported on kilovoltage (kV) planar techniques using radio-opaque clips as fiducial markers for prostate bed (PB) localization, however many pts who have undergone robotic radical prostatectomy (RP) do not have not radio-opaque clips. With the increasing use of robotic RP, we began using kV cone-beam CT (CBCT). The purpose of this study is to review our clinical experience using this approach and to compare our results with non-robotic RP pts in which daily kV planar imaging was performed using the Trilogy on-board imaging (OBI) system. Materials/Methods: Fifty IMRT pts treated following RP were analyzed: 27 conventional RP pts with daily kV planar imaging and 23 robotic RP pts (without radio-opaque clips) with daily CBCT. The PB was localized with CBCT by aligning soft tissue on CBCT and simulation CT, whereas with OBI it was aligned by matching clip positions to those on the DRR. Shifts were recorded in anterior-posterior (AP), superior-inferior (SI), and left-right (LR) axes. Total error (TE) was defined as the change in PB position on the OBI or CBCT compared to the DRR or simulation CT. PB motion (PBM) was the PB position change relative to bone. Set-up error (SE) was the set-up inaccuracy using tattoos compared to bony alignment. SE was calculated as TE minus PBM. Differences in TE, SE, and PBM were analyzed using mixed models analysis. Acute toxicity was graded using RTOG criteria. Toxicity frequency differences were analyzed using Fisher’s exact test. Results: Total error was measured in 752 CBCTs and 725 OBI image pairs. PBM and SE were measured in 585 CBCTs and 384 OBI image pairs. The mean (±SD) TE was greater with OBI compared to CBCT in the AP (4.8 ± 4.4 vs. 3.2 ± 4.4 mm, p\0.01), SI (3.8 ± 3.7 vs. 1.9 ± 3.7, p\0.01), and LR (3.9 ± 3.5 vs. 2.8 ± 3.5, p\0.01). PBM was greater with OBI in the AP (2.6 ± 1.6 vs. 0.9 ± 1.6mm, p \0.01), SI (2.4 ± 1.5 vs. 0.4 ± 1.5, p \ 0.01), and LR (1.0 ± 0.9 vs. 0.4 ± 0.9, p \0.01). SE was also greater in the AP (5.2 ± 3.8 vs. 3.7 ± 3.8, p \ 0.01), SI (4.9 ± 4.1 vs. 1.9 ± 4.1, p \ 0.01), and LR (3.8 ± 3.4 vs. 2.8 ± 3.4, p \ 0.01). Frequencies of acute $ grade 2 GI (7 vs. 13%, p = 0.7) and GU (11 vs. 9%, p = 1.0) were similar for OBI and CBCT pts. No . grade 2 toxicities were seen. Conclusions: These results suggest that although the magnitudes of TE, PBM, SE were larger with OBI compared to CBCT, the levels of acute toxicity were acceptable and comparable between the two. The reasons for the differences are unclear, but we postulate that discernment of the PB on the CBCT is difficult. Further studies are needed to evaluate the benefits and risks of CBCT in this setting. Author Disclosure: J.P. Einck, None; D.R. Simpson, None; S.K. Nath, None; R.A. Sethi, None; J.Z. Wang, None; A.J. Mundt, None; A.P. Sandhu, None.
2380
Early Outcomes following Proton Therapy for Prostate Cancer in Men 55 Years Old and Younger
B. S. Hoppe1, R. Henderson1, R. C. Nichols1, R. B. Marcus1, W. M. Mendenhall1, J. Costa2, C. Williams2, Z. Su1, Z. Li1, N. P. Mendenhall1 1
University of Florida Proton Therapy Institute, Jacksonville, FL, 2University of Florida Shands Hospital, Jacksonville, FL
Purpose/Objective(s): Controversy exists over the optimal management of young men with prostate cancer. We report early outcomes in men # 55 years old treated with proton therapy. Materials/Methods: From August 15, 2006, to September 1, 2009, 98 men # 55 years old with low- (N = 57), intermediate- (N = 29), and high- (N = 12) risk prostate cancer were enrolled on University of Florida IRB-approved protocols and received proton therapy to between 78 and 82 CGE at 2 CGE per fraction or 70 to 72.5 CGE at 2.5 CGE per fraction. Androgen deprivation (AD)
S373
S372
2376
Volume 78, Number 3, Supplement, 2010
Interfraction Seminal Vesicle Motion and Target Margin Assessment for Fiducial-Guided Intensity Modulated Radiotherapy for Prostate Cancer
M. H. Stenmark, K. A. Vineberg, D. W. Litzenberg, D. A. Hamstra, M. Feng University of Michigan, Ann Arbor, MI Purpose/Objective(s): To determine if differential interfraction motion necessitates specific planning target volume (PTV) margins for the seminal vesicles (SV) during fiducial-guided radiotherapy of the prostate. Materials/Methods: 20 patients were enrolled in an IRB-approved prospective study, which included fiducial implantation and 3 sets of CTs during treatment. SV motion relative to the prostate was analyzed by CT at 60 interfraction time points for these patients, assuming perfect set-up to the prostate with daily guidance. Following mutual information image registration of the prostate for each instance, the simulation SVs were expanded in 1 mm increments until the expansion fully covered the union of SV volumes for each patient to evaluate the geometric extent of motion. To evaluate the dosimetric consequences of variable SV positions, two intensity-modulated radiotherapy (IMRT) plans were generated for each patient including, 1) the prostate and proximal (1cm) SV (PSV), and 2) the prostate and full SV (FSV), both with a standard 5 mm PTV margin for optimization. Treatment planning was performed by a single operator using a multi-criteria hierarchical optimization method. Dosimetric coverage of both the proximal and full SVs was assessed at each instance and compared with the initial treatment plan. Results: Positional variability was greater for the FSV as compared to the PSV. A 5 mm PSV-PTV margin yielded 95% geometric coverage (C95%) in 95% of cases; median coverage of 99% (range 93-100%, SD 2%). However, a 5 mm FSV-PTV margin only resulted in C95% SV in 55% of cases; median coverage of 96% (83%-100%, SD 5%). To achieve C95% for the FSV in .95% of cases an 8 mm margin was required (median 99%, 95-100%, SD 2%). To obtain C99% in .95% of cases, the PSV required a 7mm expansion, while the FSV required a 10 mm expansion. Despite the lack of full geometric coverage of all PSV instances with a 5mm PTV expansion, dosimetric coverage was adequate. For plans optimized to the prostate and PSV, motion did not impact the volume of PSV covered by the 100% isodose line (V100%), mean 95% (SD 3%), or the 95% isodose line (V95%), mean 99% (SD 1%). For plans optimized to the FSV, dosimetric coverage was compromised in some patients, as V95% was .95% in only 70% of cases. However, overall, mean V100% was 89% (range 65%-98%, SD 8%) and mean V95% was 96%, (range 89%-100%, SD 4%). Conclusions: The SVs move differentially from the prostate and exhibit greater variation with increasing distance from the prostate. For plans targeting just the prostate and PSVs, 5mm PTV expansions are adequate. However, despite daily localization of the prostate, larger PTV margins are required for cases where the intent is to cover the FSVs. Supported by NIH/NCI (1 R21 CA110485-01A1). Author Disclosure: M.H. Stenmark, None; K.A. Vineberg, None; D.W. Litzenberg, None; D.A. Hamstra, None; M. Feng, None.
2377
The Use of MRI Imaging in IMRT Treatment Planning for Prostate Cancer
S. Vora, A. C. Silva, W. W. Wong, J. M. Collins, S. E. Schild Mayo Clinic Arizona, Phoenix, AZ Purpose/Objective(s): Dose escalation improves biochemical control rates based on phase III trials. However, higher doses have been associated with increased risk of bladder and rectal toxicity. Improvements in MRI imaging allow better definition of suspicious areas within the prostate. The purpose of this study is to assess the clinical utility of incorporating MRI imaging and its’ impact in treatment planning. Materials/Methods: 25 patients with clinically localized prostate adenocarcinoma underwent 1.5 T 8-channel surface coil MRI of the prostate using high resolution axial, coronal, and sagittal T2 weighted series, axial T1 weighted series, contrast enhanced dynamic and delayed T1 weighted fast gradient-echo series and a diffusion weighted sequence. No endorectal coil was used. The MRI was performed following placement of internal fiducial markers and CT simulation. The MRI images were reviewed for the following factors: early arterial enhancement, restricted diffusion, and T2 hypointensity. If patients had $2 factors, the area was identified as suspicious. MRI images were fused with the planning CT scan. The entire prostate+3mm margin (PTV) was planned to receive 77.4 Gy/43 fractions (1.8 Gy daily dose) and MRI suspicious areas (GTV) received concomitant boost to 83 Gy/43 fractions (1.93 Gy daily dose). Rectal/Bladder tolerance doses were maintained to keep D30\70 Gy, D10 \75 Gy, D1.8cc\81 Gy. If the constraints could not be met or if imaging was inconclusive, the GTV dose was either reduced to 81 Gy or no boost was delivered. Results: Of 25 patients, 22 (88%) patients had MRI imaging that contained suspicious areas consistent with malignancy. IMRT treatment plans generated .99% coverage of the PTV and .97% coverage of the GTV. 15 pts (68%) received boost dose of 83 Gy and the remaining 7 (32%) received 81 Gy. Review of all MRI imaging demonstrated that placing the internal fiducial markers prior to the MRI caused artifacts which can alter the GTV. Nine patients (36%) had evidence of post-fiducial placement artifact with hemorrhage most commonly observed. Conclusions: We found MRI imaging helpful in IMRT treatment planning, allowing us to define a tumor volume for dose escalation. Longer follow-up will be needed to assess biochemical control and toxicity. We now perform MRI imaging prior to internal fiducial marker placement to reduce artifacts and improve the accuracy when defining the GTV. Author Disclosure: S. Vora, None; A.C. Silva, None; W.W. Wong, None; J.M. Collins, None; S.E. Schild, None.
2378
MRI-based Treatment Planning for Prostate Brachytherapy
J. M. Albert, M. Zhang, T. L. Bruno, R. J. Kudchadker, D. A. Swanson, S. J. Frank M. D. Anderson Cancer Center, Houston, TX Purpose/Objective(s): Transrectal ultrasound (TRUS) is the standard imaging modality for brachytherapy treatment planning. However, MRI provides better anatomical delineation of the prostate. With more precise anatomic detail, MRI may confer advantages during treatment planning. Endorectal coil MRI (ecMRI), often used for staging, provides even greater anatomical resolution
Proceedings of the 52nd Annual ASTRO Meeting than standard MRI (sMRI). We aim to compare prostate volume, dosimetric parameters, and activity between plans based on sMRI and ecMRI versus TRUS. Materials/Methods: Patients were treated with I-125 brachytherapy in a prospective phase II trial at M.D. Anderson Cancer Center. TRUS-based pre-planning was utilized, and all patients received a pre-treatment ecMRI and 30-day post-implant sMRI. Blinded to TRUS-based plans, we re-planned treatments on T2-weighted sMRI and ecMRI for 20 consecutive patients. Varian’s VariSeed 8.0 was used for planning. Contours were approved by two reviewers and plans were jointly developed by a medical dosimetrist and radiation oncologist using a standard modified peripheral loading technique. All treatment plans satisfied the following dosimetric parameters: PTV V100. 95%, V150\ 60%, V200\ 20%, R100\ 1cc, and U200\ 5%. Prostate volume/dimensions, activity to volume ratio, and dosimetry were compared using a paired t-test and Wilcoxon signed rank test as appropriate. Results: Compared to TRUS, mean ecMRI prostate volume was smaller (29.5 vs. 32.5 cc, p = 0.001). This corresponded with an increased medial-lateral diameter (5.01 vs. 4.65 cm, p \ 0.001) and a decreased anterior-posterior diameter (2.69 vs. 3.06 cm, p\0.001), suggesting that the endorectal coil caused substantial anatomical distortion. Interestingly, there was a decreased craniocaudal length on both types of MRI compared to TRUS (TRUS: 4.23 cm; ecMRI: 3.71 cm, p \ 0.001; sMRI: 3.55 cm, p \ 0.001). This suggests that TRUS may overestimate prostate length, resulting in seed implantation in the genitourinary diaphragm or penile bulb, a hypothesis which was confirmed by review of post-implant MRIs. There was a significant decrease in activity per volume when using sMRI compared to TRUS (0.901 vs. 0.974 mCi/cc, p \ 0.001). Despite using less activity, sMRI plans had slightly improved PTV coverage (V100: 97.3% vs. 96.2%, p \ 0.001). sMRI-based plans also had improved dose homogeneity (V150: 47.4% vs. 53.8%, p \ 0.001 and V200: 16.6% vs. 19.2%, p \ 0.001). R100 was less than 1 cc and U200 was less than 0.07 cc on all plans. Conclusions: MRI more accurately defines prostate anatomy than TRUS. ecMRI is useful for staging, but should not be used for treatment planning due to anatomical distortion of the prostate. sMRI allows improved prostate delineation compared to TRUS, which results in less activity per volume required to achieve adequate PTV coverage and improved homogeneity. Author Disclosure: J.M. Albert, None; M. Zhang, None; T.L. Bruno, None; R.J. Kudchadker, None; D.A. Swanson, None; S.J. Frank, None.
2379
Initial Clinical Experience with Daily Cone Beam CT for Target Localization in Prostate Cancer Patients following Robotic Prostatectomy
J. P. Einck, D. R. Simpson, S. K. Nath, R. A. Sethi, J. Z. Wang, A. J. Mundt, A. P. Sandhu University of California San Diego, La Jolla, CA Purpose/Objective(s): Intensity-modulated radiotherapy (IMRT) in post-prostatectomy patients (pts) has increased the demand for more accurate treatment delivery using image-guidance. Previous studies have reported on kilovoltage (kV) planar techniques using radio-opaque clips as fiducial markers for prostate bed (PB) localization, however many pts who have undergone robotic radical prostatectomy (RP) do not have not radio-opaque clips. With the increasing use of robotic RP, we began using kV cone-beam CT (CBCT). The purpose of this study is to review our clinical experience using this approach and to compare our results with non-robotic RP pts in which daily kV planar imaging was performed using the Trilogy on-board imaging (OBI) system. Materials/Methods: Fifty IMRT pts treated following RP were analyzed: 27 conventional RP pts with daily kV planar imaging and 23 robotic RP pts (without radio-opaque clips) with daily CBCT. The PB was localized with CBCT by aligning soft tissue on CBCT and simulation CT, whereas with OBI it was aligned by matching clip positions to those on the DRR. Shifts were recorded in anterior-posterior (AP), superior-inferior (SI), and left-right (LR) axes. Total error (TE) was defined as the change in PB position on the OBI or CBCT compared to the DRR or simulation CT. PB motion (PBM) was the PB position change relative to bone. Set-up error (SE) was the set-up inaccuracy using tattoos compared to bony alignment. SE was calculated as TE minus PBM. Differences in TE, SE, and PBM were analyzed using mixed models analysis. Acute toxicity was graded using RTOG criteria. Toxicity frequency differences were analyzed using Fisher’s exact test. Results: Total error was measured in 752 CBCTs and 725 OBI image pairs. PBM and SE were measured in 585 CBCTs and 384 OBI image pairs. The mean (±SD) TE was greater with OBI compared to CBCT in the AP (4.8 ± 4.4 vs. 3.2 ± 4.4 mm, p\0.01), SI (3.8 ± 3.7 vs. 1.9 ± 3.7, p\0.01), and LR (3.9 ± 3.5 vs. 2.8 ± 3.5, p\0.01). PBM was greater with OBI in the AP (2.6 ± 1.6 vs. 0.9 ± 1.6mm, p \0.01), SI (2.4 ± 1.5 vs. 0.4 ± 1.5, p \ 0.01), and LR (1.0 ± 0.9 vs. 0.4 ± 0.9, p \0.01). SE was also greater in the AP (5.2 ± 3.8 vs. 3.7 ± 3.8, p \ 0.01), SI (4.9 ± 4.1 vs. 1.9 ± 4.1, p \ 0.01), and LR (3.8 ± 3.4 vs. 2.8 ± 3.4, p \ 0.01). Frequencies of acute $ grade 2 GI (7 vs. 13%, p = 0.7) and GU (11 vs. 9%, p = 1.0) were similar for OBI and CBCT pts. No . grade 2 toxicities were seen. Conclusions: These results suggest that although the magnitudes of TE, PBM, SE were larger with OBI compared to CBCT, the levels of acute toxicity were acceptable and comparable between the two. The reasons for the differences are unclear, but we postulate that discernment of the PB on the CBCT is difficult. Further studies are needed to evaluate the benefits and risks of CBCT in this setting. Author Disclosure: J.P. Einck, None; D.R. Simpson, None; S.K. Nath, None; R.A. Sethi, None; J.Z. Wang, None; A.J. Mundt, None; A.P. Sandhu, None.
2380
Early Outcomes following Proton Therapy for Prostate Cancer in Men 55 Years Old and Younger
B. S. Hoppe1, R. Henderson1, R. C. Nichols1, R. B. Marcus1, W. M. Mendenhall1, J. Costa2, C. Williams2, Z. Su1, Z. Li1, N. P. Mendenhall1 1
University of Florida Proton Therapy Institute, Jacksonville, FL, 2University of Florida Shands Hospital, Jacksonville, FL
Purpose/Objective(s): Controversy exists over the optimal management of young men with prostate cancer. We report early outcomes in men # 55 years old treated with proton therapy. Materials/Methods: From August 15, 2006, to September 1, 2009, 98 men # 55 years old with low- (N = 57), intermediate- (N = 29), and high- (N = 12) risk prostate cancer were enrolled on University of Florida IRB-approved protocols and received proton therapy to between 78 and 82 CGE at 2 CGE per fraction or 70 to 72.5 CGE at 2.5 CGE per fraction. Androgen deprivation (AD)
S373