Search

Background: Deep learning auto-segmentation (DLAS) models have been adopted in the clinic; however, they suffer from performance deterioration owing to the clinical practice variability. Some commercial DLAS software provide an incremental retraining function that enables users to train a custom model using their institutional data to account for clinical practice variability., Purpose: This study was performed to evaluate and implement the commercial DLAS software with the incremental retraining function for definitive treatment of patients with prostate cancer in a multi-user environment., Methods: CT-based target organs and organs-at-risk (OAR) delineation of 215 prostate cancer patients were utilized. The performance of three commercial DLAS software built-in models was validated with 20 patients. A retrained custom model was developed using 100 patients and evaluated on the remaining data (n = 115). Dice similarity coefficient (DSC), Hausdorff distance (HD), mean surface distance (MSD), and surface DSC (SDSC) were utilized for quantitative evaluation. A multi-rater qualitative evaluation was blindly performed with a five-level scale. Visual inspection was performed in consensus and non-consensus unacceptable cases to identify the failure modes., Results: Three commercial DLAS vendor built-in models achieved sub-optimal performance in 20 patients. The retrained custom model had a mean DSC of 0.82 for prostate, 0.48 for seminal vesicles (SV), and 0.92 for rectum, respectively. This represents a significant improvement over the built-in model with DSC of 0.73, 0.37, and 0.81 for the corresponding structures. Compared to the acceptance rate of 96.5% and consensus unacceptable rate (i.e., both reviewers rated as unacceptable) of 3.5% achieved by manual contours, the custom model achieved a 91.3% acceptance rate and 8.7% consensus unacceptable rate. The failure modes of retrained custom model were attributed to the following: cystogram (n = 2), hip prosthesis (n = 2), low dose rate brachytherapy seeds (n = 2), air in endorectal balloon(n = 1), non-iodinated spacer (n = 2), and giant bladder(n = 1)., Conclusion: The commercial DLAS software with the incremental retraining function was validated and clinically adopted for prostate patients in a multi-user environment. AI-based auto-delineation of the prostate and OARs is shown to achieve improved physician acceptance, overall clinical utility, and accuracy., (© 2023 American Association of Physicists in Medicine.)

Purpose: To evaluate late gastrointestinal (GI) and genitourinary (GU) toxicity of moderately hypofractionated intensity modulated proton therapy (IMPT) targeting the prostate and pelvic lymph nodes., Methods and Materials: A target accrual of 56 patients with high-risk or unfavorable intermediate risk prostate cancer were enrolled into a prospective study (ClinicalTrials.gov: NCT02874014) of moderately hypofractionated IMPT. IMPT with pencil beam scanning was used to deliver 6750 and 4500 cGy relative biological effectiveness in 25 daily fractions simultaneously to the prostate and pelvic lymph nodes, respectively. All received androgen deprivation therapy. Late GI and GU toxicity was prospectively assessed using Common Terminology Criteria for Adverse Events version 4.0, at baseline, weekly during radiation therapy, 3-month postradiation therapy, and then every 6 months. Actuarial rates of late GI and GU toxicity were estimated using Kaplan-Meier method., Results: Median age was 75.5 years. Fifty-four patients were available for late toxicity evaluation. Median follow-up was 43.9 months (range, 16-66). The actuarial rate of late grade ≥2 GI toxicity at both 2 and 3 years was 7.4% (95% confidence interval [CI], 0.2%-14.2%). The actuarial rate of late grade 3 GI toxicity at both 2 and 3 years was 1.9% (95% CI, 0%-5.4%). One patient experienced grade 3 GI toxicity with proctitis. The actuarial rate of late grade ≥2 GU toxicity was 20.5% (95% CI, 8.9%-30.6%) at 2 years, and 29.2 % (95% CI, 15.5%-40.7%) at 3 years. None had grade 3 GU toxicity. The presence of baseline GU symptoms was associated with a higher likelihood of experiencing late grade 2 GU toxicity., Conclusions: A moderately hypofractionated IMPT targeting the prostate and regional pelvic lymph nodes was generally well tolerated. Patients with pre-existing GU symptoms had a higher rate of late grade 2 GU toxicity. A phase 3 study is needed to assess any therapeutic gain of IMPT, in comparison with photon-based radiation therapy., (Copyright © 2022 Elsevier Inc. All rights reserved.)

Purpose: Determine prostate intrafraction motion with Cine-magnetic resonance imaging (MRI) and deformable registration., Methods: A total of 68 cine-MRI studies were done in 17 different series with 4 scans per series in 7 patients. In without rectal balloon (WORB) scans, 100 mL of water was infused in the rectum. Each series consisted of supine and prone, with a rectal balloon (WRB) and WORB. Each scan was performed over 4 minutes. Automatic deformable registration software developed by View Ray, Inc., Cleveland, Ohio was employed to segment the prostate for each cine-MRI image. A time-based analysis was done for the different positions and the use of the rectal balloon., Results: The variation/standard deviation of the prostate position during 240 seconds was: supine WRB: 0.55 mm, WORB: 1.2 mm, and prone WRB: 1.48 mm, WORB: 2.15 mm (P < 0.001). A strong relationship was observed between time and prostate motion. For the initial 120 s the standard deviation was smaller than for the second 120 s supine WRB 0.54 mm versus 1.37 mm; supine WORB 0.61 mm versus 1.70 mm; prone WRB 0.85 mm versus 1.85 mm; and prone WORB 1.60 mm versus 2.56 mm. The probabilities for prostate staying within +/-2 mm to its initial position are: 94.8% supine WRB; 91.5% supine WORB; 92.3% prone WRB; 79.2% prone WORB., Conclusions: Intrafraction prostate motion was found dependent on time, patient position, and the use of a rectal balloon. Relatively stable positions can be obtained for 4 minutes or less especially in the supine position with a rectal balloon.

Purpose: The ability to determine the accuracy of the final prostate position within a determined action level threshold for image-guided proton therapy is unclear., Materials and Methods: Three thousand one hundred ten images for 20 consecutive patients treated in 1 of our 3 proton prostate protocols from February to May of 2007 were analyzed. Daily kV images and patient repositioning were performed employing an action-level threshold (ALT) of > or = 2.5 mm for each beam. Isocentric orthogonal x-rays were obtained, and prostate position was defined via 3 gold markers for each patient in the 3 axes., Results: To achieve and confirm our action level threshold, an average of 2 x-rays sets (median 2; range, 0-4) was taken daily for each patient. Based on our ALT, we made no corrections in 8.7% (range, 0%-54%), 1 correction in 82% (41%-98%), and 2 to 3 corrections in 9% (0-27%). No patient needed 4 or more corrections. All patients were treated with a confirmed error of < 2.5 mm for every beam delivered. After all corrections, the mean and standard deviations were: anterior-posterior (z): 0.003 +/- 0.094 cm; superior-inferior (y): 0.028 +/- 0.073 cm; and right-left (x) -0.013 +/- 0.08 cm., Conclusion: It is feasible to limit all final prostate positions to less than 2.5 mm employing an action level image-guided radiation therapy (IGRT) process. The residual errors after corrections were very small.

Purpose: To determine the influence of magnetic-resonance-imaging (MRI)-vs. computed-tomography (CT)-based prostate and normal structure delineation on the dose to the target and organs at risk during proton therapy., Methods and Materials: Fourteen patients were simulated in the supine position using both CT and T2 MRI. The prostate, rectum, and bladder were delineated on both imaging modalities. The planning target volume (PTV) was generated from the delineated prostates with a 5-mm axial and 8-mm superior and inferior margin. Two plans were generated and analyzed for each patient: an MRI plan based on the MRI-delineated PTV, and a CT plan based on the CT-delineated PTV. Doses of 78 Gy equivalents (GE) were prescribed to the PTV., Results: Doses to normal structures were lower when MRI was used to delineate the rectum and bladder compared with CT: bladder V50 was 15.3% lower (p = 0.04), and rectum V50 was 23.9% lower (p = 0.003). Poor agreement on the definition of the prostate apex was seen between CT and MRI (p = 0.007). The CT-defined prostate apex was within 2 mm of the apex on MRI only 35.7% of the time. Coverage of the MRI-delineated PTV was significantly decreased with the CT-based plan: the minimum dose to the PTV was reduced by 43% (p < 0.001), and the PTV V99% was reduced by 11% (p < 0.001)., Conclusions: Using MRI to delineate the prostate results in more accurate target definition and a smaller target volume compared with CT, allowing for improved target coverage and decreased doses to critical normal structures.

Purpose: To determine the target coverage for proton therapy with and without image guidance and daily prebeam reorientation., Methods and Materials: A total of 207 prostate positions were analyzed for 9 prostate cancer patients treated using our low-risk prostate proton therapy protocol (University of Florida Proton Therapy Institute 001). The planning target volume was defined as the prostate plus a 5-mm axial and 8-mm superoinferior extension. The prostate was repositioned using 5- and 10-mm shifts (anteriorly, inferiorly, posteriorly, and superiorly) and for Points A-D using a combination of 10-mm multidimensional movements (anteriorly or inferiorly; posteriorly or superiorly; and left or right). The beams were then realigned using the new prostate position. The prescription dose was 78 Gray equivalent (GE) to 95% of the planning target volume., Results: For small movements in the anterior, inferior, and posterior directions within the planning target volume (< or =5 mm), treatment realignment demonstrated small, but significant, improvements in the clinical target volume (CTV) coverage to the prescribed dose (78 GE). The anterior and posterior shifts also significantly increased the minimal CTV dose (Delta +1.59 GE). For prostate 10-mm movements in the inferior, posterior, and superior directions, the beam realignment produced larger and significant improvements for both the CTV V(78) (Delta +6.4%) and the CTV minimal dose (Delta +8.22 GE). For the compounded 10-mm multidimensional shifts, realignment significantly improved the CTV V(78) (Delta +11.8%) and CTV minimal dose (Delta +23.6 GE). After realignment, the CTV minimal dose was >76.6 GE (>98%) for all points (A-D)., Conclusion: Proton beam realignment after target shift will enhance CTV coverage for different prostate positions.

Purpose: To determine the impact of prostate motion on dose coverage in proton therapy., Methods and Materials: A total of 120 prostate positions were analyzed on 10 treatment plans for 10 prostate patients treated using our low-risk proton therapy prostate protocol (University of Florida Proton Therapy Institute 001). Computed tomography and magnetic resonance imaging T(2)-weighted turbo spin-echo scans were registered for all cases. The planning target volume included the prostate with a 5-mm axial and 8-mm superoinferior expansion. The prostate was repositioned using 5- and 10-mm one-dimensional vectors and 10-mm multidimensional vectors (Points A-D). The beam was realigned for the 5- and 10-mm displacements. The prescription dose was 78 Gy equivalent (GE)., Results: The mean percentage of rectum receiving 70 Gy (V(70)) was 7.9%, the bladder V(70) was 14.0%, and the femoral head/neck V(50) was 0.1%, and the mean pelvic dose was 4.6 GE. The percentage of prostate receiving 78 Gy (V(78)) with the 5-mm movements changed by -0.2% (range, 0.006-0.5%, p > 0.7). However, the prostate V(78) after a 10-mm displacement changed significantly (p < 0.003) with different movements: 3.4% (superior), -5.6% (inferior), and -10.2% (posterior). The corresponding minimal doses were also reduced: 4.5 GE, -4.7 GE, and -11.7 GE (p < or = 0.003). For displacement points A-D, the clinical target volume V(78) coverage had a large and significant reduction of 17.4% (range, 13.5-17.4%, p < 0.001) in V(78) coverage of the clinical target volume. The minimal prostate dose was reduced 33% (25.8 GE), on average, for Points A-D. The prostate minimal dose improved from 69.3 GE to 78.2 GE (p < 0.001) with realignment for 10-mm movements., Conclusion: The good dose coverage and low normal doses achieved for the initial plan was maintained with movements of < or = 5 mm. Beam realignment improved coverage for 10-mm displacements.

Purpose: The contrast in dose distribution between proton radiotherapy (RT) and intensity-modulated RT (IMRT) is unclear, particularly in regard to critical structures such as the rectum and bladder., Methods and Materials: Between August and November 2006, the first 10 consecutive patients treated in our Phase II low-risk prostate proton protocol (University of Florida Proton Therapy Institute protocol 0001) were reviewed. The double-scatter proton beam plans used in treatment were analyzed for various dosimetric endpoints. For all plans, each beam dose distribution, angle, smearing, and aperture margin were optimized. IMRT plans were created for all patients and simultaneously analyzed. The IMRT plans were optimized through multiple volume objectives, beam weighting, and individual leaf movement. The patients were treated to 78 Gray-equivalents (GE) in 2-GE fractions with a biologically equivalent dose of 1.1., Results: All rectal and rectal wall volumes treated to 10-80 GE (percentage of volume receiving 10-80 GE [V(10)-V(80)]) were significantly lower with proton therapy (p < 0.05). The rectal V(50) was reduced from 31.3% +/- 4.1% with IMRT to 14.6% +/- 3.0% with proton therapy for a relative improvement of 53.4% and an absolute benefit of 16.7% (p < 0.001). The mean rectal dose decreased 59% with proton therapy (p < 0.001). For the bladder and bladder wall, proton therapy produced significantly smaller volumes treated to doses of 10-35 GE (p < 0.05) with a nonsignificant advantage demonstrated for the volume receiving < or =60 GE. The bladder V(30) was reduced with proton therapy for a relative improvement of 35.3% and an absolute benefit of 15.1% (p = 0.02). The mean bladder dose decreased 35% with proton therapy (p = 0.002)., Conclusion: Compared with IMRT, proton therapy reduced the dose to the dose-limiting normal structures while maintaining excellent planning target volume coverage.

Purpose: To describe dose-volume values with the use of water alone vs. a rectal balloon (RB) for the treatment of prostate cancer with proton therapy., Materials and Methods: We analyzed 30 proton plans for 15 patients who underwent CT and MRI scans with an RB or water alone. Simulation was performed with a modified MRI endorectal coil and an RB with 100 mL of water or water alone. Doses of 78-82 gray equivalents were prescribed to the planning target volume. The two groups were compared for three structures: rectum, rectal wall (RW), and rectal wall 7 cm (RW7) at the level of the planning target volume., Results: Rectum and RW volumes radiated to low, intermediate, and high doses were small: rectum V10, 33.7%; V50, 17.3%; and V70, 10.2%; RW V10, 32.4%; V50, 20.4%; and V70, 14.6%. The RB effectively increased the rectal volume for all cases (139.8 +/- 44.9 mL vs. 217.7 +/- 32.2 mL (p < 0.001). The RB also decreased the volume of the rectum radiated to doses V10-V65 (p < or = 0.05); RW for V10-V50; and RW7 for V10-V35. An absolute rectum V50 improvement >5% was seen for the RB in 5 of 15 cases, for a benefit of 9.2% +/- 2.3% compared with 2.4% +/- 1.3% for the remaining 10 cases (p < 0.001). Similar benefit was seen for the rectal wall. No benefit was seen for doses > or =70 gray equivalents for the rectum, RW, or RW7. No benefit of < or =1% was seen with an RB in 46% for the rectum V70 and in 40% for the rectal wall V70., Conclusions: Rectum and rectal wall doses with proton radiation were low whether using water or an RB. Selected patients will have a small but significant advantage with an RB; however, water alone was well tolerated and will be an alternative for most patients.

Purpose: This study was designed to characterize the relationship between the observed rate of postradiation genitourinary (GU) complications and the prostate gland size., Methods and Materials: Two hundred seventy-three patients received conformal neutron and photon irradiation to the prostate seminal vesicles. Data on post-treatment urinary morbidity were collected and examined in relationship to a number of clinical and technical factors., Results: With a median follow-up of 30 months (range 7-61), the 4-year rate of Grade 2 or higher GU complications was 21%. On univariate analysis, the risk of complications was significantly associated with prostate size and neutron beam arrangement. On multivariate analysis, only the prostate size was significantly associated with the risk of GU morbidity. Patients with a preradiation prostate volume more than 74 cc had a two and a half fold increase in the risk of complications compared to patients with smaller glands., Conclusion: Patients with an enlarged prostate have a significantly higher risk of chronic GU complications. Although these data were obtained for patients receiving combined neutron and photon irradiation, it is likely that these data would also be applicable for those patients receiving photon irradiation as well. These observations may add an additional rationale for the study of preirradiation hormonal treatment.