Rectal Ulcers and Rectoprostatic Fistulas after (125)I Low Dose Rate Prostate Brachytherapy.

PURPOSE: Radiation induced rectal ulcers and fistulas are rare but significant complications of low dose rate prostate brachytherapy for localized prostate cancer. We describe the incidence of ulcers and fistulas, and associated risk factors. MATERIALS AND METHODS: We reviewed the records of 4,690 patients with localized prostate cancer who were treated with low dose rate (125)I prostate brachytherapy to a dose of 144 Gy with or without 6 months of androgen deprivation therapy. Patient, disease, comorbidity, treatment, dosimetric and posttreatment intervention factors were analyzed for an association with ulcer or fistula formation. RESULTS: At a median followup of 53 months 21 cases were identified, including 15 rectal ulcer cases, of which 6 progressed to fistulas, and an additional 6 cases of fistulas with no prior documented ulcers. Overall 9 rectal ulcer cases (0.19%) and 12 fistula cases (0.26%) were identified. In 8 of 15 patients ulcers healed with conservative management. No fistulas healed without surgical management. Two patients with fistulas died. Eight patients diagnosed with rectal ulcers subsequently underwent rectal biopsies, after which fistulas developed in 3. One patient with a de novo fistula underwent a preceding biopsy. Urinary interventions such as transurethral resection of the prostate were performed after brachytherapy in 5 of 12 patients with fistulas compared to 0 of 9 with ulcers alone. Argon plasma coagulation of the rectum for hematochezia was performed after brachytherapy in 3 of 12 patients with fistulas. CONCLUSIONS: Rates of post-brachytherapy rectal ulcers and fistulas are low as previously described. Post-brachytherapy interventions such as rectal biopsy, argon coagulation and urinary intervention may increase the risk of fistulas.

The challenges facing deep learning-based catheter localization for ultrasound guided high-dose-rate prostate brachytherapy.

BACKGROUND: Automated catheter localization for ultrasound guided high-dose-rate prostate brachytherapy faces challenges relating to imaging noise and artifacts. To date, catheter reconstruction during the clinical procedure is performed manually. Deep learning has been successfully applied to a wide variety of complex tasks and has the potential to tackle the unique challenges associated with multiple catheter localization on ultrasound. Such a task is well suited for automation, with the potential to improve productivity and reliability. PURPOSE: We developed a deep learning model for automated catheter reconstruction and investigated potential factors influencing model performance. The model was designed to integrate into a clinical workflow, with a proposed reconstruction confidence metric to aid in planner verification. METHODS: Datasets from 242 patients treated from 2016 to 2020 were collected retrospectively. The anonymized dataset comprises 31,000 transverse images reconstructed from 3D sagittal ultrasound acquisitions and 3500 implanted catheters manually localized by the planner. Each catheter was retrospectively ranked based on the severity of imaging artifacts affecting reconstruction difficulty. The U-NET deep learning architecture was trained to localize implanted catheters on transverse images. A fivefold cross-validation method was used, allowing for evaluation over the entire dataset. The postprocessing software combined the predictions with patient-specific implant information to reconstructed catheters in 3D space, uniquely matched to the implanted grid positions. A reconstruction confidence metric was calculated based on the number and probability of localized predictions per catheter. For each patient, deep learning prediction and postprocessing reconstruction were completed in under 2 min on a nonperformance PC. RESULTS: Overall, 80% of catheter reconstructions were accurate, within 2 mm along 90% of the length. The catheter tip was often not detected and required extrapolation during reconstruction. The reconstruction accuracy was 89% for the easiest catheter ranking and decreased to 13% for the highest difficulty ranking, when the aid of live ultrasound would have been recommended. Even when limited to the easiest ranked catheters, the reconstruction accuracy decreased at distal grid positions, down to 50%. Individual implantation style was found to influence the frequency of severe artifacts, slightly impacting the model accuracy. A reconstruction confidence metric identified the difficult catheters, removed the observed individual variation, and increased the overall accuracy to 91% while excluding 27% of the reconstructions. CONCLUSIONS: The deep learning model localized implanted catheters over a large clinical dataset, with overall promising results. The model faced challenges due to ultrasound artifacts and image degradation distal to the probe, underlining the continued importance of maintaining image quality and minimizing artifacts. A potential workflow for integration into the clinical procedure was demonstrated, including the use of a confidence metric to predict low accuracy reconstructions. Comparison between models evaluated on different datasets should also consider underlying differences, such as the frequency and severity of imaging artifacts.

Lung SBRT credentialing in the Canadian OCOG-LUSTRE randomized trial.

Purpose: To report on the Stereotactic Body Radiation Therapy (SBRT) credentialing experience during the Phase III Ontario Clinical Oncology Group (OCOG) LUSTRE trial for stage I non-small cell lung cancer. Methods: Three credentialing requirements were required in this process: (a) An institutional technical survey; (b) IROC (Imaging and Radiation Oncology Core) thoracic phantom end-to-end test; and (c) Contouring and completion of standardized test cases using SBRT for one central and one peripheral lung cancer, compared against the host institution as the standard. The main hypotheses were that unacceptable variation would exist particularly in OAR definition across all centres, and that institutions with limited experience in SBRT would be more likely to violate per-protocol guidelines. Results: Fifteen Canadian centres participated of which 8 were new, and 7 were previously established (≥2 years SBRT experience), and all successfully completed surveys and IROC phantom testing. Of 30 SBRT test plans, 10 required replanning due to major deviations, with no differences in violations between new and established centres (p = 0.61). Mean contouring errors were highest for brachial plexus in the central (C) case (12.55 ± 6.62 mm), and vessels in the peripheral (P) case (13.01 ± 12.55 mm), with the proximal bronchial tree (PBT) (2.82 ± 0.78 C, 3.27 ± 1.06 P) as another variable structure. Mean dice coefficients were lowest for plexus (0.37 ± 0.2 C, 0.37 ± 0.14 P), PBT (0.77 ± 0.06 C, 0.75 ± 0.09 P), vessels (0.69 ± 0.29 C, 0.64 ± 0.31 P), and esophagus (0.74 ± 0.04 C, 0.76 ± 0.04 P). All plans passed per-protocol planning target volume (PTV) coverage and maximum/volumetric organs-at-risk constraints, although variations existed in dose gradients within and outside the target. Conclusions: Clear differences exist in both contouring and planning with lung SBRT, regardless of centre experience. Such an exercise is important for studies that rely on high precision radiotherapy, and to ensure that implications on trial quality and outcomes are as optimal as possible.

Hypofractionated Nodal Radiation Therapy for Breast Cancer Was Not Associated With Increased Patient-Reported Arm or Brachial Plexopathy Symptoms.

PURPOSE: To determine whether nodal radiation therapy (RT) for breast cancer using modest hypofractionation (HF) with 2.25 to 2.5 Gy per fraction (fx) was associated with increased patient-reported arm symptoms, compared with conventional fractionation (CF) ≤2 Gy/fx. METHODS AND MATERIALS: Two cancer registries were used to identify subjects who received computed tomography-planned nodal RT for pT1-3, pN0-2, M0 breast cancer, from 2007 to 2010 at 2 cancer institutions. After ethics approval, patients were mailed an explanatory letter and the Self-reported Arm Symptom Scale, a validated instrument with 8 questions about arm symptoms and 5 related to activities of daily living. Clinicopathologic characteristics and Self-reported Arm Symptom Scale scores were compared between HF/CF cohorts using nonparametric analysis, χ2 analysis, and multivariate ordinal regression. RESULTS: Of 1759 patients, 800 (45.5%) returned a completed survey. A total of 708 eligible cases formed the study cohort. Of these, 406 (57%) received HFRT (40 Gy/16 fx, 45 Gy/20 fx), and 302 (43%) received CFRT (45-50 Gy/25 fx, 50.4 Gy/28 fx). Median time interval after RT was 5.7 years. Forty-three percent and 75% of patients received breast-conserving surgery and chemotherapy, respectively. Twenty-two percent received breast boost RT, independent of fractionation. Median age at diagnosis was 59 years (HF) and 53 years (CF) (P<.001). The mean numbers of excised (n=12) and involved (n=3) nodes were similar between fractionation cohorts (P=.44), as were the mean sums of responses in arm symptoms (P=.17) and activities of daily living (P=.85). Patients receiving HF reported lower rates of shoulder stiffness (P=.04), trouble moving the arm (P=.02), and difficulty reaching overhead (P<.01) compared with the CF cohort. There was no difference in self-reported arm swelling or symptoms related to brachial plexopathy. CONCLUSIONS: Nodal RT with hypofractionation was not associated with increased patient-reported arm symptoms or functional deficits compared with CF. Subjects treated with CF reported more disability in certain aspects of arm/shoulder function. These data support shorter fractionation utilization when regional nodes are within the therapeutic target.

Single-nucleotide polymorphisms studied for associations with urinary toxicity from (125)I prostate brachytherapy implants.

PURPOSE: To identify clinical, dosimetric, and genetic factors that are associated with late urinary toxicity after a (125)I prostate brachytherapy implant. METHODS AND MATERIALS: Genomic DNA from 296 men treated with (125)I prostate brachytherapy monotherapy was extracted from saliva samples for this study. A retrospective database was compiled including clinical, dosimetric, and toxicity data for this cohort of patients. Fourteen candidate single-nucleotide polymorphism (SNPs) from 13 genes (TP53, ERCC2, GSTP1, NOS, TGFß1, MSH6, RAD51, ATM, LIG4, XRCC1, XRCC3, GSTA1, and SOD2) were tested in this cohort for correlations with toxicity. RESULTS: This study identified 217 men with at least 2 years of followup. Of these, 39 patients developed Grade ≥2 late urinary complications with a transurethral resection of prostate, urethral stricture, gross hematuria, or a sustained increase in their International Prostate Symptom Score. The only clinical or dosimetric factor that was associated with late urinary toxicity was age (p = 0.02). None of the 14 SNPs tested in this study were associated with late urinary toxicity in the univariate analysis. CONCLUSIONS: This study identified age as the only variable being associated with late urinary toxicity. However, the small sample size and the candidate gene approach used in this study mean that further investigations are essential. Genome-wide association studies are emerging as the preferred approach for future radiogenomic studies to overcome the limitations from a candidate gene approach.