- Email: [email protected]
Hypofractionated Online Cone-beam CT-guided Intensity Modulated Radiation Therapy (hypoCBCT-IMRT) for Localized Prostate Cancer: Acute Toxicity Profile Compared to Adaptive Offline Image-guided IMRT (ART-IMRT)
Proceedings of the 52nd Annual ASTRO Meeting relatively uniform, high quality intraprostatic and extracapsular dose distributions. It is unlikely that further local dose intensification would have altered the biochemical control of the failed patients in this study. Author Disclosure: W.M. Butler, None; G.S. Merrick, None; N. Bittner, None; Z.A. Allen, None; B. White, None; A. Adamovich, None; K.E. Wallner, None.
2369
Measurement of Stranded Source Migration in Prostate Brachytherapy
S. E. Daniel, C. Rabbani, R. Taylor, P. Chuba St. John Macomb Oakland Hospital, Warren, MI Purpose/Objective(s): The use of radioactive seeds embedded in absorbable vicryl suture material has emerged as one preferred method for prostate cancer brachytherapy. While there are several practical advantages to the use of stranded sources, it is unclear how their displacement affects dosimetry. Our objective was to use CT imaging to determine strand displacement between day zero and day 30 scans and to assess the dosimetric consequences of strand displacement. Materials/Methods: Between 3/06 and 12/09 there were 87 prostate brachytherapy patients with day zero and day 30 post-plan imaging. Implants included 62 cases using I-125, 11 cases using Cs-131, and 14 cases using Pd-103. There were 9 to 28 strands per implant, and implants included between 3 and 28 loose seeds. On average, there were 72 seeds per implant (range of 48 to 105). Substantial strand displacement was identified by inspection. Migrated strands were identified on day 0 and day 30 scans and the distance of displacement was measured. Displacement was considered only in strands that had migrated away from the implant center. Dosimetric parameters (D90, D80, V200, V150, V100, urethra maximum, urethra mean, rectum maximum, rectum mean) were evaluated using CMS. Results: Of the 1942 strands placed, 19 strands were found to have substantial migration. These displacements occurred in 18 of the 87 cases. The distance of strand displacement ranged from 0.5 cm to 2.5 cm, with an average of 1.5 cm. Of the displaced strands, 10 strands moved superior to the prostate and 9 strands moved inferior to the prostate. Between 1 and 4 seeds were lost in 27 cases. No considerable differences were found in post-implant day zero versus day 30 dosimetry. Comparisons between day zero and day 30 changes in dosimetry in seed displacement cases versus control cases also revealed no considerable dosimetric differences. Conclusions: Strand movement may be expected to occur in approximately 0.0098 percent of stranded sources and 20.69 percent of cases. Little dosimetric consequence is expected. Author Disclosure: S.E. Daniel, None; C. Rabbani, None; R. Taylor, None; P. Chuba, None.
2370
One Year PSA Value as Strong Predictor of Biochemical Outcome after Definitive Prostate Seed Brachytherapy
W. Ding1, L. Yang1, J. Lee1, J. Tay2 1
California Pacific Medical Center, San Francisco, CA, 2St. Mary’s Regional Medical Center, Reno, NV
Purpose/Objective(s): Extensive studies of pre-treatment PSA have resulted in predictive tools with good concordance indices. Post-treatment PSA (ptPSA) has not been analyzed as closely, especially in patients who have received brachytherapy seed implants. Using ptPSA levels in a prediction model incorporates one of the most important elements of outcome, the actual treatment itself. The purpose of this study is to correlate the role of PSA value at one year to PSA relapse-free survival (PRFS) following definitive prostate seed brachytherapy. Materials/Methods: Records of 204 hormone-naı¨ve patients with localized adenocarcinoma of the prostate with $ 4 years of follow-up treated at St. Mary’s Regional Medical Center in Reno, NV and Carson Tahoe Regional Medical Center in Carson City, NV between 1998-2003 using I-125 or Pd-103 seed brachytherapy were retrospectively analyzed. Treatment planning was done according to a pre-planned, modified peripheral loading technique. All implants were planned by the same dosimetrist and physicist, and all the implants were performed by the same radiation oncologist. 185 of 204 patients had PSA records at 12 months after implant. The following variables were included in the statistical analysis: Age, initial pretreatment PSA (iPSA), Gleason score, T-stage, NCCN risk group (RG), perineural invasion (PNI), use of external beam boost, V(100), and ptPSA levels at 12 months with cut points at #1, 1.01-2.00, 2.01-3.00, and .3.00 ng/ml. Survival analysis was done using the Kaplan-Meier method. Test for significance was performed using the log-rank test. Multivariate analysis was done using the Cox-proportional hazard model. Results: Median follow-up was 80 months. Median age was 69. Numbers of patients stratified by NCCN low, intermediate and high RG were 110:65:10 respectively. The prescription dose was 145 Gy and 125 Gy for I-125 and Pd-103 seeds respectively. The median V(100) was 89%. In univariate analysis, age (p = 0.021), iPSA (p = 0.003), RG (p = 0.029), and ptPSA levels at 12 months (p \ 0.001) all significantly predicted PRFS. PNI was borderline significant (p = 0.086). The 5-year PRFS at ptPSA levels #1, 1.01-2.00, 2.01-3.00, and .3.00 ng/ml were 98.5%, 85.7%, 61.5%, and 22.2% respectively. The 10-year PRFS at ptPSA levels #1 and 1.01-2.00 ng/ml were 90.5% and 85.7% respectively. In multivariate analysis, both ptPSA and PNI were significant independent predictors of PRFS. Hazard rations (HR) for ptPSA levels at #1, 1.01-2.00, 2.01-3.00, and .3.00 ng/ml at 12 months were 1, 4.96, 27.57, and 65.10 respectively, and PNI had a HR of 6.1 (p = 0.009). Conclusions: Post-treatment PSA value at 12 months is a strong predictive variable for PRFS after definitive prostate brachytherapy seed implant. Author Disclosure: W. Ding, None; L. Yang, None; J. Lee, None; J. Tay, None.
2371
Hypofractionated Online Cone-beam CT-guided Intensity Modulated Radiation Therapy (hypoCBCT-IMRT) for Localized Prostate Cancer: Acute Toxicity Profile Compared to Adaptive Offline Image-guided IMRT (ART-IMRT)
M. Ghilezan, D. Yan, L. Kestin, D. Brabbins, M. Wallace, C. Mitchell, A. Casey, D. Krauss, F. Vicini, A. Martinez William Beaumont Hospital, Royal Oak, MI
S369
I. J. Radiation Oncology d Biology d Physics
S370
Volume 78, Number 3, Supplement, 2010
Purpose/Objective(s): To report the acute toxicity of hypoCBCT-IMRT compared to ART-IMRT for localized prostate cancer. Materials/Methods: From 2007-2010, 38 patients with biopsy-proven, non-metastatic prostate cancer were treated with HIC-approved online hypoCBCT-IMRT protocol. Patients were classified according to prognostic factors (Gleason score, PSA and clinical stage) in low risk (group I) and intermediate/high risk (group II). Patients were prospectively treated to a median minimal peripheral dose of 64 Gy in 20 fractions of 3.2 Gy to the prostate for group I and prostate + seminal vesicles for group II. Prior to RT planning, three 4 cm Visicoil markers were placed transperineally in the prostate using transrectal ultrasound at 3, 7, and 11 o’clock positions under local anesthesia for online image-guidance. MRI was used for target/anatomy contouring. GTV to CTV margins were based on risk of extra-prostatic disease extension and were in the range of 0-4mm (only in the posterolateral direction). CTV+3mm=PTV. Daily CBCT were obtained prior to each treatment. Online shifts were performed following registration based on Visicoil position and soft tissue anatomy. Toxicity was scored using the Common Toxicity Criteria, version 3. Results: 35 patients with a minimum follow-up of 3 months were analyzed. Median follow-up was 1 year (range 0.3-3). The comparison group consisted of 1037 patients treated with ART-IMRT (minimal peripheral dose range of 75.6-82 Gy). Acute GU and GI toxicities were comparable between the groups. However, in terms of GU frequency/urgency 36% and 3% of ART patients experienced grade 2 and 3 toxicity whereas only 20% and 0% of hypoCBCT-IMRT patients had grade 2 and 3 toxicity (p = 0.02). 16% of ART patients complained of rectal pain/tenesmus grade 2 while only 3% of hypoCBCT-IMRT patients had similar symptoms (p = 0.002). Although not statistically significant, there were 12% of ART patients with grade 1 rectal bleeding as opposed to only 3% (1 patient) in the hypoCBCT-IMRT group. Conclusions: The 4-week hypoCBCT-IMRT regimen is well tolerated with low acute toxicity rates. The significant decrease in urinary frequency/urgency and rectal pain/tenesmus is likely due to better targeting and use of minimal margins enabled by online CBCT image guidance that appears to be superior to standard offline ART-IMRT. Author Disclosure: M. Ghilezan, None; D. Yan, None; L. Kestin, None; D. Brabbins, None; M. Wallace, None; C. Mitchell, None; A. Casey, None; D. Krauss, None; F. Vicini, None; A. Martinez, None.
2372
The Role of Intra- or Periprostatic Calcifications in Image-guided Radiotherapy for Prostate Cancer
S. A. Hanna, G. N. Marta, W. F. P. Neves-Junior, C. M. K. Haddad, J. F. da Silva Hospital Sirio Libanes, Sao Paulo 01308050, Brazil Purpose/Objective(s): Image-guided radiotherapy (IGRT) allows a more precise localization of the prostate, thus minimizing errors due to organ motion and set-up in the treatment of prostate cancer (PC). Using megavoltage cone-beam computed tomography (MVCBCT), references such as bones, the prostate itself or implanted fiducials may help in correcting patient positioning immediately before each RT fraction. However, the use of fiducials requires an invasive procedure and is associated with increased costs. Intra- or periprostatic calcifications (IPC) occur in 30-40% of patients and can be viewed both in the planning CT (PCT) and in MVCBCT images. The objectives of this study were to assess the role of IPC as natural fiducials for IGRT in PC. Materials/Methods: Between June 2007 and February 2010 data on 10 patients treated with IGRT for PC using implanted fiducials were retrospectively analyzed. In all cases IPC were visible in both PCT and MVCBCT images. IPC were classified according to their position (central when inside the prostate and peripheral when within the PTV margins). In each case we compared the daily isocenter deviations in three directions (lateral, longitudinal and vertical) from baseline positioning using fiducials and using IPC. These patients had median 36 (range, 4-38) daily values for all six deviations. Differences between the daily deviations for these two different references were compared, using all available patient data, with Wilcoxon rank sum test. Results: Two hundred eighty-seven MVCBCT/PCT paired images were available, each one having six deviations (three coordinates using two references, fiducials and IPC). IPC were central in 8 cases and peripheral in 2, and their median number was 2 (range, 1-5). The mean daily deviation (in mm) in the lateral, longitudinal, and vertical coordinates were respectively 0.55 ± 3.11, 0.56 ± 3.45, and -0.54 ± 4.03 when the fiducials were used, and 0.72 ± 3.22, 0.63 ± 3.58, and -0.69 ± 4.26 when IPC were used. The p values for differences between paired comparisons (fiducials vs. IPC) were 0.003, 0.653, and 0.078, for the lateral, longitudinal, and vertical coordinates, respectively. When only cases with central IPC were analyzed, the corresponding p values were 0.195, 0.455, and 0.296. Corresponding p values for cases with peripheral locations were 0.006, 0.406, and 0.122. Conclusions: These data suggest that centrally located IPC may be used as natural fiducials for treatment positioning during IGRT for PC. Author Disclosure: S.A. Hanna, None; G.N. Marta, None; W.F.P. Neves-Junior, None; C.M.K. Haddad, None; J.F. da Silva, None.
2373
Individualized Predictive Treatment Margins to Account for Prostate Motion during Treatment using Real-time Intra-fraction Tracking
M. Howard1, M. Khan2, L. Miller3, P. Xia2 1
Parkridge Medical Center, Chattanooga, TN, 2Cleveland Clinic, Cleveland, OH, 3University of Tennessee, Knoxville, TN
Purpose/Objective(s): To develop a mathematical model to predict individualized treatment margins for each patient based on patient specific prostate motion during treatment, and to compare this model with two conventional fixed margins of 6 mm in all directions except for 4 mm posteriorly and 2 mm uniform margins in all directions. Materials/Methods: Seven patients underwent real-time prostate localization and tracking during treatment using the Calypso system. A new mathematical model was created to fit the data derived from the first five fractions for each patient [model building data set]. The model was tested for its ability to predict future prostate motion for each patient by using data in the remaining sessions [validation set]. Once validated, the model was used to generate patient specific treatment margins and to compare these margins with the conventional fixed margin approaches. Results: The average percentage of time that the prostate moved outside the treatment margins using 6/4 mm margins in all patients was 0.07 ± 0.06, 3.69 ± 4.53, and 1.55 ± 2.08 in the right-left (x), anterior-posterior (y), and superior-inferior directions (z),
2369
Measurement of Stranded Source Migration in Prostate Brachytherapy
S. E. Daniel, C. Rabbani, R. Taylor, P. Chuba St. John Macomb Oakland Hospital, Warren, MI Purpose/Objective(s): The use of radioactive seeds embedded in absorbable vicryl suture material has emerged as one preferred method for prostate cancer brachytherapy. While there are several practical advantages to the use of stranded sources, it is unclear how their displacement affects dosimetry. Our objective was to use CT imaging to determine strand displacement between day zero and day 30 scans and to assess the dosimetric consequences of strand displacement. Materials/Methods: Between 3/06 and 12/09 there were 87 prostate brachytherapy patients with day zero and day 30 post-plan imaging. Implants included 62 cases using I-125, 11 cases using Cs-131, and 14 cases using Pd-103. There were 9 to 28 strands per implant, and implants included between 3 and 28 loose seeds. On average, there were 72 seeds per implant (range of 48 to 105). Substantial strand displacement was identified by inspection. Migrated strands were identified on day 0 and day 30 scans and the distance of displacement was measured. Displacement was considered only in strands that had migrated away from the implant center. Dosimetric parameters (D90, D80, V200, V150, V100, urethra maximum, urethra mean, rectum maximum, rectum mean) were evaluated using CMS. Results: Of the 1942 strands placed, 19 strands were found to have substantial migration. These displacements occurred in 18 of the 87 cases. The distance of strand displacement ranged from 0.5 cm to 2.5 cm, with an average of 1.5 cm. Of the displaced strands, 10 strands moved superior to the prostate and 9 strands moved inferior to the prostate. Between 1 and 4 seeds were lost in 27 cases. No considerable differences were found in post-implant day zero versus day 30 dosimetry. Comparisons between day zero and day 30 changes in dosimetry in seed displacement cases versus control cases also revealed no considerable dosimetric differences. Conclusions: Strand movement may be expected to occur in approximately 0.0098 percent of stranded sources and 20.69 percent of cases. Little dosimetric consequence is expected. Author Disclosure: S.E. Daniel, None; C. Rabbani, None; R. Taylor, None; P. Chuba, None.
2370
One Year PSA Value as Strong Predictor of Biochemical Outcome after Definitive Prostate Seed Brachytherapy
W. Ding1, L. Yang1, J. Lee1, J. Tay2 1
California Pacific Medical Center, San Francisco, CA, 2St. Mary’s Regional Medical Center, Reno, NV
Purpose/Objective(s): Extensive studies of pre-treatment PSA have resulted in predictive tools with good concordance indices. Post-treatment PSA (ptPSA) has not been analyzed as closely, especially in patients who have received brachytherapy seed implants. Using ptPSA levels in a prediction model incorporates one of the most important elements of outcome, the actual treatment itself. The purpose of this study is to correlate the role of PSA value at one year to PSA relapse-free survival (PRFS) following definitive prostate seed brachytherapy. Materials/Methods: Records of 204 hormone-naı¨ve patients with localized adenocarcinoma of the prostate with $ 4 years of follow-up treated at St. Mary’s Regional Medical Center in Reno, NV and Carson Tahoe Regional Medical Center in Carson City, NV between 1998-2003 using I-125 or Pd-103 seed brachytherapy were retrospectively analyzed. Treatment planning was done according to a pre-planned, modified peripheral loading technique. All implants were planned by the same dosimetrist and physicist, and all the implants were performed by the same radiation oncologist. 185 of 204 patients had PSA records at 12 months after implant. The following variables were included in the statistical analysis: Age, initial pretreatment PSA (iPSA), Gleason score, T-stage, NCCN risk group (RG), perineural invasion (PNI), use of external beam boost, V(100), and ptPSA levels at 12 months with cut points at #1, 1.01-2.00, 2.01-3.00, and .3.00 ng/ml. Survival analysis was done using the Kaplan-Meier method. Test for significance was performed using the log-rank test. Multivariate analysis was done using the Cox-proportional hazard model. Results: Median follow-up was 80 months. Median age was 69. Numbers of patients stratified by NCCN low, intermediate and high RG were 110:65:10 respectively. The prescription dose was 145 Gy and 125 Gy for I-125 and Pd-103 seeds respectively. The median V(100) was 89%. In univariate analysis, age (p = 0.021), iPSA (p = 0.003), RG (p = 0.029), and ptPSA levels at 12 months (p \ 0.001) all significantly predicted PRFS. PNI was borderline significant (p = 0.086). The 5-year PRFS at ptPSA levels #1, 1.01-2.00, 2.01-3.00, and .3.00 ng/ml were 98.5%, 85.7%, 61.5%, and 22.2% respectively. The 10-year PRFS at ptPSA levels #1 and 1.01-2.00 ng/ml were 90.5% and 85.7% respectively. In multivariate analysis, both ptPSA and PNI were significant independent predictors of PRFS. Hazard rations (HR) for ptPSA levels at #1, 1.01-2.00, 2.01-3.00, and .3.00 ng/ml at 12 months were 1, 4.96, 27.57, and 65.10 respectively, and PNI had a HR of 6.1 (p = 0.009). Conclusions: Post-treatment PSA value at 12 months is a strong predictive variable for PRFS after definitive prostate brachytherapy seed implant. Author Disclosure: W. Ding, None; L. Yang, None; J. Lee, None; J. Tay, None.
2371
Hypofractionated Online Cone-beam CT-guided Intensity Modulated Radiation Therapy (hypoCBCT-IMRT) for Localized Prostate Cancer: Acute Toxicity Profile Compared to Adaptive Offline Image-guided IMRT (ART-IMRT)
M. Ghilezan, D. Yan, L. Kestin, D. Brabbins, M. Wallace, C. Mitchell, A. Casey, D. Krauss, F. Vicini, A. Martinez William Beaumont Hospital, Royal Oak, MI
S369
I. J. Radiation Oncology d Biology d Physics
S370
Volume 78, Number 3, Supplement, 2010
Purpose/Objective(s): To report the acute toxicity of hypoCBCT-IMRT compared to ART-IMRT for localized prostate cancer. Materials/Methods: From 2007-2010, 38 patients with biopsy-proven, non-metastatic prostate cancer were treated with HIC-approved online hypoCBCT-IMRT protocol. Patients were classified according to prognostic factors (Gleason score, PSA and clinical stage) in low risk (group I) and intermediate/high risk (group II). Patients were prospectively treated to a median minimal peripheral dose of 64 Gy in 20 fractions of 3.2 Gy to the prostate for group I and prostate + seminal vesicles for group II. Prior to RT planning, three 4 cm Visicoil markers were placed transperineally in the prostate using transrectal ultrasound at 3, 7, and 11 o’clock positions under local anesthesia for online image-guidance. MRI was used for target/anatomy contouring. GTV to CTV margins were based on risk of extra-prostatic disease extension and were in the range of 0-4mm (only in the posterolateral direction). CTV+3mm=PTV. Daily CBCT were obtained prior to each treatment. Online shifts were performed following registration based on Visicoil position and soft tissue anatomy. Toxicity was scored using the Common Toxicity Criteria, version 3. Results: 35 patients with a minimum follow-up of 3 months were analyzed. Median follow-up was 1 year (range 0.3-3). The comparison group consisted of 1037 patients treated with ART-IMRT (minimal peripheral dose range of 75.6-82 Gy). Acute GU and GI toxicities were comparable between the groups. However, in terms of GU frequency/urgency 36% and 3% of ART patients experienced grade 2 and 3 toxicity whereas only 20% and 0% of hypoCBCT-IMRT patients had grade 2 and 3 toxicity (p = 0.02). 16% of ART patients complained of rectal pain/tenesmus grade 2 while only 3% of hypoCBCT-IMRT patients had similar symptoms (p = 0.002). Although not statistically significant, there were 12% of ART patients with grade 1 rectal bleeding as opposed to only 3% (1 patient) in the hypoCBCT-IMRT group. Conclusions: The 4-week hypoCBCT-IMRT regimen is well tolerated with low acute toxicity rates. The significant decrease in urinary frequency/urgency and rectal pain/tenesmus is likely due to better targeting and use of minimal margins enabled by online CBCT image guidance that appears to be superior to standard offline ART-IMRT. Author Disclosure: M. Ghilezan, None; D. Yan, None; L. Kestin, None; D. Brabbins, None; M. Wallace, None; C. Mitchell, None; A. Casey, None; D. Krauss, None; F. Vicini, None; A. Martinez, None.
2372
The Role of Intra- or Periprostatic Calcifications in Image-guided Radiotherapy for Prostate Cancer
S. A. Hanna, G. N. Marta, W. F. P. Neves-Junior, C. M. K. Haddad, J. F. da Silva Hospital Sirio Libanes, Sao Paulo 01308050, Brazil Purpose/Objective(s): Image-guided radiotherapy (IGRT) allows a more precise localization of the prostate, thus minimizing errors due to organ motion and set-up in the treatment of prostate cancer (PC). Using megavoltage cone-beam computed tomography (MVCBCT), references such as bones, the prostate itself or implanted fiducials may help in correcting patient positioning immediately before each RT fraction. However, the use of fiducials requires an invasive procedure and is associated with increased costs. Intra- or periprostatic calcifications (IPC) occur in 30-40% of patients and can be viewed both in the planning CT (PCT) and in MVCBCT images. The objectives of this study were to assess the role of IPC as natural fiducials for IGRT in PC. Materials/Methods: Between June 2007 and February 2010 data on 10 patients treated with IGRT for PC using implanted fiducials were retrospectively analyzed. In all cases IPC were visible in both PCT and MVCBCT images. IPC were classified according to their position (central when inside the prostate and peripheral when within the PTV margins). In each case we compared the daily isocenter deviations in three directions (lateral, longitudinal and vertical) from baseline positioning using fiducials and using IPC. These patients had median 36 (range, 4-38) daily values for all six deviations. Differences between the daily deviations for these two different references were compared, using all available patient data, with Wilcoxon rank sum test. Results: Two hundred eighty-seven MVCBCT/PCT paired images were available, each one having six deviations (three coordinates using two references, fiducials and IPC). IPC were central in 8 cases and peripheral in 2, and their median number was 2 (range, 1-5). The mean daily deviation (in mm) in the lateral, longitudinal, and vertical coordinates were respectively 0.55 ± 3.11, 0.56 ± 3.45, and -0.54 ± 4.03 when the fiducials were used, and 0.72 ± 3.22, 0.63 ± 3.58, and -0.69 ± 4.26 when IPC were used. The p values for differences between paired comparisons (fiducials vs. IPC) were 0.003, 0.653, and 0.078, for the lateral, longitudinal, and vertical coordinates, respectively. When only cases with central IPC were analyzed, the corresponding p values were 0.195, 0.455, and 0.296. Corresponding p values for cases with peripheral locations were 0.006, 0.406, and 0.122. Conclusions: These data suggest that centrally located IPC may be used as natural fiducials for treatment positioning during IGRT for PC. Author Disclosure: S.A. Hanna, None; G.N. Marta, None; W.F.P. Neves-Junior, None; C.M.K. Haddad, None; J.F. da Silva, None.
2373
Individualized Predictive Treatment Margins to Account for Prostate Motion during Treatment using Real-time Intra-fraction Tracking
M. Howard1, M. Khan2, L. Miller3, P. Xia2 1
Parkridge Medical Center, Chattanooga, TN, 2Cleveland Clinic, Cleveland, OH, 3University of Tennessee, Knoxville, TN
Purpose/Objective(s): To develop a mathematical model to predict individualized treatment margins for each patient based on patient specific prostate motion during treatment, and to compare this model with two conventional fixed margins of 6 mm in all directions except for 4 mm posteriorly and 2 mm uniform margins in all directions. Materials/Methods: Seven patients underwent real-time prostate localization and tracking during treatment using the Calypso system. A new mathematical model was created to fit the data derived from the first five fractions for each patient [model building data set]. The model was tested for its ability to predict future prostate motion for each patient by using data in the remaining sessions [validation set]. Once validated, the model was used to generate patient specific treatment margins and to compare these margins with the conventional fixed margin approaches. Results: The average percentage of time that the prostate moved outside the treatment margins using 6/4 mm margins in all patients was 0.07 ± 0.06, 3.69 ± 4.53, and 1.55 ± 2.08 in the right-left (x), anterior-posterior (y), and superior-inferior directions (z),