Search

Artificial intelligence (AI) normal tissue contouring tools are now widely available and used in many treatment planning systems. However, few studies have prospectively demonstrated the value of such tools when employed in clinical workflows. We conducted a randomized controlled trial to assess the benefit of using an AI heart contouring algorithm to assist breast radiotherapy (RT) planning. A single institution, 2-arm randomized controlled crossover trial was undertaken between December 2021 and October 2023. A convolutional neural network for heart auto-contouring, which had median Dice 0.95 compared to gold standard contours and reduced contouring time by 50% in pre-clinical studies, was implemented as a scripted tool within a treatment planning system. Eligible patients had breast cancer planned for RT, with hearts contoured according to the randomization of the dosimetrist assigned to their plan. Dosimetrists were stratified by experience and randomized to (1) manual heart contouring first or (2) AI-assisted heart contouring first. AI-assisted contouring consisted of running the model within the treatment planning system and editing the contour as needed. After completing 5 cases on the initial contour strategy, dosimetrists crossed to the other strategy. Co-primary outcomes were feasibility and efficacy, defined as heart contouring time (measured by recording screen during contouring). Secondary endpoints included dosimetrist and treating physician contour assessments. The trial had 90% power to detect a 30% reduction in contour time with 2-sided type I error of 5%. One hundred eighteen patients enrolled; 60 patients' hearts were contoured manually and 58 with AI assistance. Eleven dosimetrists enrolled; 5 were randomized to manual first arm and 6 to AI-assisted first arm. There was no difference in manual vs. AI-assisted contour time overall (mean 277.3 ± 151.2 vs. 267.2 ± 199.0 seconds, P = 0.76), on the manual first arm only (P = 0.15), or on the AI first arm only (P = 0.67). There was no difference in contour time among only dosimetrists with ⩽2 yr experience (mean = 374.4 ± 174.2 vs. 403.3 ± 275.9 secs, P = 0.72), nor among dosimetrists with > 3 yr experience (mean = 241.9 ± 126.4 vs. 210.8 ± 121.9 secs, P = 0.25). Dosimetrists considered AI contours acceptable with minor/no modification in 13 of 47 (27.6%) cases and unacceptable in 34 of 47 (72.4%) cases; but considered the AI helpful in 35 of 47 (74.5%) cases and to improve subjective efficiency in 29 of 47 (61.7%) cases. Physicians, blinded to randomization, thought contours presented for review were unacceptable in 6 of 56 (10.7%) AI-assisted and 3 of 56 (5.4%) manual cases. Despite improving efficiency pre-clinically, AI assistance did not reduce heart contour time compared to standard manual contouring. Physicians were more likely to find contours unacceptable when AI-assistance was used. Pre-clinical findings may not translate into clinical benefits, emphasizing the critical need to evaluate new AI technologies under their intended use in real clinical workflows. [ABSTRACT FROM AUTHOR]

The FORMULA-509 trial randomized patients receiving salvage radiation between 6 months of GnRH agonist plus bicalutamide or 6 months of GnRH agonist plus AAP/Apa. Here we evaluate factors associated with testosterone recovery after trial treatment and the impact of testosterone recovery on patient-reported HRQoL. Testosterone was evaluated prior to treatment, at 3 months, at 6 months (end of systemic treatment), and at least every six months after treatment. Validated Expanded Prostate Cancer Index Composite (EPIC)-26 questionnaire was administered at baseline, end of treatment, and 1 year after completion of treatment. A change of 4 or more was considered a clinically meaningful change in the hormonal function domain. Log-rank test compared testosterone recovery between treatment groups. Wilcoxon rank sum and chi-square compared characteristics between those who did and did not recover testosterone. Multivariate logistic regression evaluated factors associated with lack of testosterone recovery. Chi-square evaluated frequency of clinically meaningful change in hormonal function. Three hundred forty-five patients were randomized (172 bicalutamide; 173 AAP/Apa). Median follow-up was 34 (6-53) months. There was no difference in testosterone recovery (>190 ng/dL) between treatment groups (p = 0.10). 52.6% [95% CI = 42.7-58.0] of patients recovered testosterone 6 months after treatment (bicalutamide: 51.2%; AAP/Apa 54.0%); 74.5% [69.8-79.2] 12 months (70.5%; 78.6%); 81.9% [77.7-86.1] 18 months (77.7%; 86.2%); and 88.8% [85.4-92.3] 30 months after treatment (85.8%; 91.9%). On univariate analysis, older age (median 67 vs 63 years; p<0.01), lower baseline testosterone (median 274 vs 379 ng/dL; p<0.01) and higher PSA at enrollment (median 0.3 vs 0.2; p = 0.02) were associated with lack of testosterone recovery 1 year after treatment. Race, Gleason score, and treatment arm were not associated with testosterone recovery. On multivariate analysis, age at enrollment, baseline testosterone, and PSA at enrollment remained significant (all p <0.05). In the subset of patients who completed EPIC-26 1 year after treatment (n = 219), those patients who did not have testosterone recovery were more likely to report a clinically meaningful decline in hormonal function relative to baseline than those who did achieve testosterone recovery, but it was not statistically significant (59% vs. 43% reporting clinically meaningful functional decline, p = 0.06). The addition of AAP/Apa did not lengthen time to testosterone recovery. Older men or those with lower baseline testosterone or higher PSA prior to treatment were less likely to recover testosterone 1 year after treatment. Patients who did not have testosterone recovery at 1 year were more likely to report decline in hormonal function. [ABSTRACT FROM AUTHOR]