Framework for Radiation Oncology Department-wide Evaluation and Implementation of Commercial Artificial Intelligence Autocontouring.

PURPOSE: Artificial intelligence (AI)-based autocontouring in radiation oncology has potential benefits such as standardization and time savings. However, commercial AI solutions require careful evaluation before clinical integration. We developed a multidimensional evaluation method to test pretrained AI-based automated contouring solutions across a network of clinics. METHODS AND MATERIALS: Curated data included 121 patient planning computed tomography (CT) scans with a total of 859 clinically approved contours used for treatment from 4 clinics. Regions of interest (ROIs) were generated with 3 commercial AI-based automated contouring software solutions (AI1, AI2, AI3) spanning the following disease sites: brain, head and neck (H&N), thorax, abdomen, and pelvis. Quantitative agreement between AI-generated and clinical contours was measured by Dice similarity coefficient (DSC) and Hausdorff distance (HD). Qualitative assessment was performed by multiple experts scoring blinded AI-contours using a Likert scale. Workflow and usability surveying was also conducted. RESULTS: AI1, AI2, and AI3 contours had high quantitative agreement in 27.8%, 32.8%, and 34.1% of cases (DSC >0.9), performing well in pelvis (median DSC = 0.86/0.88/0.91) and thorax (median DSC = 0.91/0.89/0.91). All 3 solutions had low quantitative agreement in 7.4%, 8.8%, and 6.1% of cases (DSC <0.5), performing worse in brain (median DSC = 0.65/0.78/0.75) and H&N (median DSC = 0.76/0.80/0.81). Qualitatively, AI1 and AI2 contours were acceptable (rated 1-2) with at most minor edits in 70.7% and 74.6% of ROIs (2906 ratings), higher for abdomen (AI1: 79.2%) and thorax (AI2: 90.2%), and lower for H&N (29.0/35.6%). An end-user survey showed strong user preference for full automation and mixed preferences for accuracy versus total number of structures generated. CONCLUSIONS: Our evaluation method provided a comprehensive analysis of both quantitative and qualitative measures of commercially available pretrained AI autocontouring algorithms. The evaluation framework served as a roadmap for clinical integration that aligned with user workflow preference.

The Judicious Use of Stereotactic Ablative Radiotherapy in the Primary Management of Localized Renal Cell Carcinoma.

Localized renal cell carcinoma is primarily managed surgically, but this disease commonly presents in highly comorbid patients who are poor operative candidates. Less invasive techniques, such as cryoablation and radiofrequency ablation, are effective, but require percutaneous or laparoscopic access, while generally being limited to cT1a tumors without proximity to the renal pelvis or ureter. Active surveillance is another management option for small renal masses, but many patients desire treatment or are poor candidates for active surveillance. For poor surgical candidates, a growing body of evidence supports stereotactic ablative radiotherapy (SABR) as a safe and effective non-invasive treatment modality. For example, a recent multi-institution individual patient data meta-analysis of 190 patients managed with SABR estimated a 5.5% five-year cumulative incidence of local failure with one patient experiencing grade 4 toxicity, and no other grade ≥3 toxic events. Here, we discuss the recent developments in SABR for the management of localized renal cell carcinoma, highlighting key concepts of appropriate patient selection, treatment design, treatment delivery, and response assessment.

A Prospective Study of a Resorbable Intravesical Fiducial Marker for Bladder Cancer Radiation Therapy.

Purpose: We conducted a prospective pilot study to evaluate safety and feasibility of TraceIT, a resorbable radiopaque hydrogel, to improve image guidance for bladder cancer radiation therapy (RT). Methods and Materials: Patients with muscle invasive bladder cancer receiving definitive RT were eligible. TraceIT was injected intravesically around the tumor bed during maximal transurethral resection of bladder tumor. The primary endpoint was the difference between radiation treatment planning margin on daily cone beam computed tomography based on alignment to TraceIT versus standard-of-care pelvic bone anatomy. The Van Herk margin formula was used to determine the optimal planning target volume margin. TraceIT visibility, recurrence rates, and survival were estimated by Kaplan-Meier method. Toxicity was measured by Common Terminology Criteria for Adverse Events version 4.03. Results: The trial was fully accrued and 15 patients were analyzed. TraceIT was injected in 4 sites/patient (range, 4-6). Overall, 94% (95% confidence interval [CI], 90%-98%) of injection sites were radiographically visible at RT initiation versus 71% (95% CI, 62%-81%) at RT completion. The median duration of radiographic visibility for injection sites was 106 days (95% CI, 104-113). Most patients were treated with a standard split-course approach with initial pelvic radiation fields, then midcourse repeat transurethral resection of bladder tumor followed by bladder tumor bed boost fields, and 14/15 received concurrent chemotherapy. Alignment to fiducials could allow for reduced planning target volume margins (0.67 vs 1.56 cm) for the initial phase of RT, but not for the boost (1.01 vs 0.96 cm). This allowed for improved target coverage (D95% 80%-83% to 91%-94%) for 2 patients retrospectively planned with both volumetric-modulated arc therapy and 3-dimensional conformal RT. At median follow-up of 22 months, no acute or late complications attributable to TraceIT placement occurred. No patients required salvage cystectomy. Conclusions: TraceIT intravesical fiducial placement is safe and feasible and may facilitate tumor bed delineation and targeting in patients undergoing RT for localized muscle invasive bladder cancer. Improved image guided treatment may facilitate strategies to improve local control and minimize toxicity.

Oncogenic Genomic Alterations, Clinical Phenotypes, and Outcomes in Metastatic Castration-Sensitive Prostate Cancer.

PURPOSE: The genomic underpinning of clinical phenotypes and outcomes in metastatic castration-sensitive prostate cancer is unclear. EXPERIMENTAL DESIGN: In patients with metastatic castration-sensitive prostate cancer at a tertiary referral center, clinical-grade targeted tumor sequencing was performed to quantify tumor DNA copy number alterations and alterations in predefined oncogenic signaling pathways. Disease volume was classified as high volume (≥4 bone metastases or visceral metastases) versus low volume. RESULTS: Among 424 patients (88% white), 213 (50%) had high-volume disease and 211 (50%) had low-volume disease, 275 (65%) had de novo metastatic disease, and 149 (35%) had metastatic recurrence of nonmetastatic disease. Rates of castration resistance [adjusted hazard ratio, 1.84; 95% confidence interval (CI), 1.40-2.41] and death (adjusted hazard ratio, 3.71; 95% CI, 2.28-6.02) were higher in high-volume disease. Tumors from high-volume disease had more copy number alterations. The NOTCH, cell cycle, and epigenetic modifier pathways were the highest-ranking pathways enriched in high-volume disease. De novo metastatic disease differed from metastatic recurrences in the prevalence of CDK12 alterations but had similar prognosis. Rates of castration resistance differed 1.5-fold to 5-fold according to alterations in AR, SPOP (inverse), and TP53, and the cell cycle, WNT (inverse), and MYC pathways, adjusting for disease volume and other genomic pathways. Overall survival rates differed 2-fold to 4-fold according to AR, SPOP (inverse), WNT (inverse), and cell-cycle alterations. PI3K pathway alterations were not associated with prognosis once adjusted for other factors. CONCLUSIONS: This study identified genomic features associated with prognosis in metastatic castration-sensitive disease that may aid in molecular classification and treatment selection.

Dose-Escalated Intensity Modulated Radiation Therapy for Prostate Cancer: 15-Year Outcomes Data.

PURPOSE: To report 15-year outcomes for dose-escalated intensity modulated radiation therapy (IMRT) for localized prostate cancer (PC) by evaluating biochemical relapse, distant metastases, cancer-specific survival, and long-term toxicity. METHODS AND MATERIALS: A database search was conducted for the first cohort of patients treated at this institution with 81 or 86.4 Gy between 1996 and 1998 using IMRT. Toxicity data were scored according to the Common Terminology Criteria for Adverse Events version 3.0. Median follow-up was 11.6 years (range, 5-21 years). RESULTS: In the study, 301 patients were treated with 81 Gy (n = 269, 89%) or 86.4 Gy (n = 32, 11%). Patients were analyzed by National Comprehensive Cancer Network risk group, with 29% low risk (LR), 49% intermediate risk (IR), and 22% high risk (HR). Late grade 3 gastrointestinal (GI) toxicity was seen in 3 patients (1.0%). No grade 4 GI toxicity events occurred. Median time from radiation therapy to late grade 3 GI toxicity was 2.9 years. One event occurred after 10 years. Late grade 3 and 4 genitourinary (GU) toxicity was seen in 6 (2.0%) and 1 (0.3%) patient, respectively. Median time to late grade 3+ GU toxicity was 5.5 years. Two events occurred after 10 years. In addition, 38 (12.6%) developed second primary malignancies (SPMs), 8 of which were in-field malignancies. Median time from radiation therapy to all SPM and in-field SPM was 10 years. The 15-year relapse-free survival was 76%, 65%, and 55% in the LR, IR, and HR groups, respectively. Distant metastases-free survival was 88%, 75%, and 63% for LR, IR, and HR patients, respectively. PC-specific mortality was 1.9%, 7.1%, and 12.2% for LR, IR, and HR patients. CONCLUSIONS: This report represents the longest follow-up data set to our knowledge of patients treated with high-dose IMRT for PC. Our findings indicate that it is well tolerated with 1.0% and 2.3% incidence of long-term grade 3+ GI and GU toxicity, respectively. The cohort had excellent PC-specific survival.

Long-term Impact of Androgen-deprivation Therapy on Cardiovascular Morbidity After Radiotherapy for Clinically Localized Prostate Cancer.

OBJECTIVE: To characterize the impact of androgen-deprivation therapy (ADT) on the incidence of cardiovascular events (CE) in prostate cancer patients treated with radiotherapy (RT). MATERIALS AND METHODS: There were 2211 patients with localized prostate cancer treated with RT from 1988 to 2008 at our institution. There were 991 patients (44.8%) who received ADT at the time of RT for a median of 6.1 months. Salvage ADT was initiated prior to CE in 365 men (16.5%) at a median of 5.5 years (range: 0.6 to 18.4 years) after RT and continued for a median of 4.3 years. A nomogram was constructed to predict the 10-year risk of CE "post-RT" (i.e., after RT). RESULTS: Patients receiving ADT at the time of RT exhibited significantly higher 10-year incidence of CE (19.6%, 95% CI 17.0%-22.6%) than those not receiving ADT (14.3%, 95% CI 12.2%-16.7%, P = .005). On multivariate analysis, both ADT at the time of RT (P = .007) and the time of salvage (P = .0004) were associated with increased CE risk, as were advanced age (P = .02), smoking (P = .0007), history of diabetes (P = .0007), and history of CE before RT (P < .0001). A nomogram using patient age, smoking status, history of pre-RT CE, history of diabetes, and ADT use at the time of RT predicted the rate of 10-year CE with a C-index of 0.81 (95% CI, 0.72-0.88). CONCLUSION: While ADT is often an essential part of prostate cancer treatment, patients should be counseled regarding increased risks of CE and prophylactic efforts should be considered to mitigate that risk.