Purpose:
To compare the late toxicity, quality of life (QOL) and biochemical disease control between prostate stereotactic ablative radiotherapy (SABR) delivered in 2- and 5-fractions for the treatment of intermediate risk prostate cancer. A secondary exploratory analysis is to evaluate, in an offline fashion, the potential utility and feasability of incorporating a novel artificial intelligence-driven dose guidance/adaptive positioning algothim into the image-guidance workflow process for prostate SABR.
Background:
In recent years prostate SABR has emerged as a standard treatment option for low and intermediate risk prostate cancer. Prostate SABR is a precise an accurate form of external beam radiotheapy, typically delivered in 5-7 fractions. The most common dose-fractionation for prostate SABR involves 40Gy to the prostate gland in 5 fractions over 2 weeks.
Multiple randomized trials have established the non-inferiority of prostate SABR in 5-7 fractions compared to conventinal or moderately hypofractionated radiotherapy. In the setting of HDR brachytherapy, 27Gy delivered in 2 fractions is highly convenient and is associated with very high rates of disease control with low toxicity. Two-fraction prostate SABR has been studied in a number of single arm studies, with enouraging results. Cross-study comparison of 2- and 5-fraction SABR suggests the two-fraction regimen may have less bowel and sexual side effects with compariable disease control. Currrently, randomized head-to-head compariosn betweeen 2- and 5-fraction prostate SABR is limited, so it is unknown how the two compare with regard to toxicity, QOL and disease control. Two-fraction SABR is certainly desirable given the potential for improved patient convenience and reduced resource utilization compared to the 5-fraction regimen, and the fact that SABR is less invasive than HDR brachytherapy.
Because of the large dose per fracton of 2-fraction SABR, techniques to minimize toxicity risks are desirable. In this regard, a rectal spacer such as Space OAR Hydrogel is associated with improved toxicity and QOL with prostate radiotherapy and is standard at our institution. Also, real-time prostate tracking is associated with reduced urinary toxicity. Incorporation of these technologies can minimize the risk of treamtent, whether delivered in 2- or 5-fractions.
Also, theortically, improvements in image-guidance that allow for real time dose assessment and adaptive positioning could be used to improve treatment delivery and reduce toxicity. LIMBUS/RADFORMATION have developed a novel AI-driven dose-guidance application that allows for real time assessment of organ at risk and target dose at the time of cone-beam CT at the treatment unit, and can also provide data for adaptive positioning to optimize dose delivery. Study of this technology in the offline setting can provide insight as to the potential utility of the application and how to fit it into prostate SABR worklow, provding the first step before the future online testing of the product.
Given the data above, the present protocol proposes a randomized phase II trial of prostate SABR with the use of a rectal spacer and real-time tracking, comparing 27Gy in 2 weekly fractions and 40Gy in 5 fractions given every other day over 2 weeks, with offline evaluation of an AI-driven dose guidance/adaptive positioning application.
Trial Design:
100 patients with intermediate risk prostate cancer will be randomized 1:1 to prostate SABR with a dose-fractionation of either, (A) 27Gy in 2 fractions over 2 weeks, or (B) 40Gy in 5 fractions over 2 weeks. All patietns will undego fiducial marker and rectal spacer insertion prior to treatment. CT simulation and MR imaging are necessary for treatment planning. Treatments will proceed using a VMAT-based planning and delivery technique. Online image-guidance will involve CBCT and prostate tracking with triggered imaging. The RADFORMATION/LIMBUS dose guidance/adaptive positioning will be applied offline for data collection purposes only.
Patients will be assessed at baseline and in followup with regard to PSA, testosterone, provider-reported toxcity (RTOG and CTCAEv5 scales), and patient-reported outcomes (EPIC-26 bowel/bladder/sexual domains). Subject follow-up will contiue for 5 years after completion of SABR.
Six months of androgen-deprivation therapy is permitted at the discretion of the treating physican. Participants may not be on cytoxic chemotherapy while on SABR. For example, subjects taking methotrexate for benign disease, such as rheumatoid arthritis, must stop the medication during radiotherapy.
Schedule of Assessements:
* Complete history and physical exam including digital rectal examination (DRE), along with assessment of ECOG performance status must be performed within the 16 weeks preceding the start of radiotherapy.
* Pathologic confirmation of adenocarcinoma of the prostate within 18 months of randomization is required.
* CT abdomen/pelvis within 6 months of randomization (may be omitted for NCCN low risk participants).
* Bone scan within 6 months of randomization (may be omitted for NCCN low risk participants).
* Estimation of prostate volume within 6 months of randomization - this can be estimated by imaging (CT, MRI, US) or by DRE. Prostate volume must be estimated to be less than 100cc.
* PSA and testosterone measured within 90 days prior to randomization.
* Baseline symptom assessment using the CTCAE version 5.0 and RTOG/SOMA bladder and bowel toxicity scoring should be performed within the 16 weeks preceding start of radiotherapy
* Baseline participant questionnaires, including the International Prostate Symptom Score (IPSS) and the Expanded Prostate Index Composite-26 Short form (EPIC-26) should be performed within the 16 weeks preceding start of radiotherapy.
Formal Stopping Rules:
The primary objective of this study is to determine and compare the late grade 2+ toxicity at 4 years between arms, thus accrual will continue until the total sample of 100 participants is obtained. However, the trial may be suspended/terminated under certain events as outlined below:
If there are 3 or more major complications (grade 3 or higher toxicity) in the first 11 participants in the 2-fraction arm, or 7 or more during accrual of the 2-fraction arm, then the lower limit of the 95% confidence interval of the true complication rate would be at least 6%, assuming a binomial distribution, and strong consideration will be given to suspend or terminate the trial. In this regard, the Data Monitoring and Safety Committee would review the toxicities and make a recommendation on suspension or termination of the trial.
Endpoints for evaluation:
Primary endpoint: Freedom from grade 2 or higher late toxicity (combined GU + GI) at four years post-treatment.
Secondary endpoints:
i. Freedom from grade 2 or higher acute toxicity (combined GU + GI) at 6 months post-treatment.
ii. Percentage of participants in each arm with a minimally-important change in EPIC-26 bladder, bowel and sexual QOL domains at 6 months (early) and 4 years (late).
iii. Percentage of participants in each arm with moderate to severe urinary symptoms (IPSS≥8) at 6 months (early) and 4 years (late).
iv. 5-year biochemical control in each arm evaluated using the Phoenix (nadir + 2) definition of biochemical failure.
v. Mean dose reduction to the organs at risk (OARs) using RADFORMATION dose-guided adaptive positioning, offline.
vi. Percentage of fractions in which the parallel, offline use of the RADFORMATION dose-guidance algorithm could be utilized without exceeding the time constraints of a standard treatment slot.
Sample Size Calculation:
The study will use a screening-type phase II design. With α of 0.2 the sample size of 100 subjects randomized 1:1 to each arm provides 80% power to detect a 20% difference in cumulative late grade 2 toxicity (combined GI + GU) at 4 years.
Planned Study Period:
Based on the rate of enrollment on the BC Cancer Prostate SABR registry and the previous ASSERT trial, the estimated accrual period is 3 years. Subsequently, 5 years of follow-up is required after the last participant is randomized. Therefore, the total study duration is estimated at 8 years after commencement of accrual.
