Radiation therapy for prostate cancer has traditionally been delivered in small daily doses over many weeks -- a schedule known as conventional fractionation. Research over the past two decades has demonstrated that prostate cancer cells are particularly sensitive to high doses per fraction, meaning larger single doses kill cancer cells more effectively relative to normal tissue than small daily doses do.
This biological property has driven the development of ultra-hypofractionated radiotherapy (UHF RT), where the entire radiation course is delivered in just five sessions rather than the 20-40 sessions required conventionally. Each session delivers a substantially higher dose -- typically 7-8 Gray per fraction compared to 1.8-2 Gray conventionally -- exploiting this dose sensitivity to achieve equivalent or better tumor control while dramatically shortening treatment duration.
The technical challenge with UHF RT is precision: when each fraction carries so much dose, any geometric miss that irradiates healthy tissue is magnified proportionally. Historical UHF RT delivery systems included robotic platforms like CyberKnife and conventional linear accelerators with image guidance. Both require fiducial markers implanted in the prostate and extensive preparation protocols for bladder and rectum filling consistency.
A newer approach, online adaptive radiotherapy (OART), addresses the geometric precision challenge differently. Rather than relying on rigid preparation protocols to achieve reproducible anatomy each day, OART images the patient immediately before each session and generates a new, patient-specific treatment plan that accounts for that day's actual anatomy. This daily anatomical adaptation allows treatment margins to be reduced without sacrificing target coverage.
The Ethos system is a commercially available radiation therapy platform specifically designed for online adaptive radiotherapy. It integrates a cone beam computed tomography (CBCT) scanner into the treatment gantry, allowing high-quality 3D anatomical imaging of the patient while they lie on the treatment couch, immediately before each radiation delivery session.
What distinguishes OART from conventional image-guided radiotherapy (IGRT) is the response to that daily imaging. In IGRT, the daily scan is used to align the patient to a fixed pre-planned treatment. In OART, the daily scan triggers a complete re-optimization of the treatment plan based on that day's anatomy -- accounting for daily changes in prostate position, bladder filling, and rectal gas that can shift organs by centimeters between sessions.
Ethos automates much of this adaptation through AI-based auto-contouring of the key structures, including the prostate, seminal vesicles, rectum, and bladder, from the daily CBCT. A radiation oncologist reviews and approves these contours, then the system generates a new plan optimized to today's anatomy using the same clinical dose goals defined during pre-planning. The re-optimized plan is ready for delivery within the same appointment slot.
The entire workflow from imaging through plan generation to quality verification and delivery must be completed efficiently to keep appointment times manageable. A key feature of the Ethos system is its intelligent optimization engine, which uses Fourier transform dose calculation to generate fluence-optimized plans rapidly. This study specifically evaluated whether this workflow could be completed within a clinically practical timeframe for a five-fraction UHF RT prostate treatment.
The patient was a 66-year-old man presenting with elevated PSA levels. Biopsy confirmed prostatic acinar adenocarcinoma with a Gleason score of 3+4=7, indicating intermediate-risk disease -- cancer with a moderately aggressive growth pattern. Three of 14 cores from the biopsy were positive, and PSA was 4.5 ng/mL. No lymph node involvement or distant metastases were identified.
MRI revealed an enlarged prostate of 35 cubic centimeters with multiple suspicious lesions in both the peripheral and anterior zones. The patient had an excellent performance status (WHO 0), indicating no functional limitations from the disease. A multidisciplinary team reviewed the case and recommended UHF RT as the treatment approach.
The treatment prescription used a simultaneous integrated boost (SIB) technique, delivering two different dose levels within the same session. The gross tumor volume (GTV, the prostate itself) received 40 Gray in five fractions (8 Gray per session). The planning target volume (PTV, which adds margins around the prostate and includes the proximal seminal vesicles) received 36.25 Gray in five fractions (7.25 Gray per session). Treatments were delivered twice weekly.
This SIB approach allows the highest-risk tissue -- the visible tumor -- to receive a higher dose while adjacent structures receive a slightly lower dose, all within the same plan. SIB requires sophisticated dose modulation, which is an inherent capability of modern MLC-equipped linear accelerator systems but is challenging to implement on certain delivery platforms.
Pre-planning began with a 1 mm slice-thickness planning CT and T2-weighted and diffusion-weighted MRI acquired for contouring purposes. The MRI was rigidly registered to the CT so that high-soft-tissue-contrast MRI images could be used to guide contouring. Key organs at risk (OARs) including the bladder, rectum, bowel, penile bulb, femoral heads, and neurovascular bundle were carefully delineated along with a prostate urethra structure to protect urinary function.
The planning target volume was created with a 5 mm margin around the clinical target volume (CTV) in all directions except posteriorly, where only a 3 mm margin was used to protect the rectum. A three-arc volumetric modulated arc therapy (VMAT) plan was selected as the delivery technique, offering efficient and conformal dose delivery. Clinical goals specifying desired dose constraints for each structure were defined and prioritized, guiding the optimization algorithm.
For each of the five treatment sessions, the patient followed a standardized bladder preparation protocol. After positioning, CBCT imaging was performed and the Ethos system automatically contoured the prostate, seminal vesicles, rectum, bladder, and bowel. The radiation oncologist reviewed and approved these contours, and the system then generated both an IGRT plan (repositioned original plan) and an adapted plan (fully re-optimized) for comparison and selection.
Before delivery, the adaptive plan underwent secondary dose calculation verification using the Mobius3D system, which independently recalculated the dose using a different algorithm and applied a 3%/3 mm gamma pass criteria. After approval, a second verification CBCT confirmed final patient position. Throughout the entire session, the AlignRT surface monitoring system tracked patient position in real time to detect any movement between the initial imaging and dose delivery.
The adaptive plans consistently outperformed the scheduled IGRT plans in every one of the five treatment sessions. Average GTV coverage was 85% with adaptive planning versus 72% with IGRT. PTV coverage averaged 97% with adaptive plans compared to 84% with IGRT. These are clinically meaningful differences -- a PTV coverage of 84% means that 16% of the target volume is receiving less than the prescribed dose, creating a risk of underdosing cancer.
Organ-at-risk sparing also improved with adaptive planning. Maximum doses to the bladder and rectum were 5% and 8% lower respectively with adaptive plans compared to IGRT. Rectal doses at multiple threshold volumes were consistently lower with adaptive planning -- for example, the volume of rectum receiving 29 Gray or more was reduced from 35% with IGRT to 29% with adaptive plans. Penile bulb sparing showed particularly strong improvement, with the volume receiving high dose reduced from 14.7% (IGRT) to 4% (adaptive).
Plan delivery quality was confirmed to be excellent. The Mobius3D gamma pass rate exceeded 99% for all five adaptive sessions. Average 3D couch shifts after the final verification CBCT were less than 1 mm, confirming that the adaptive workflow reliably achieved accurate patient positioning. Post-delivery log file analysis confirmed that the delivered dose matched the planned dose with over 99% gamma pass rates.
The entire OART workflow -- from CBCT imaging through plan generation, verification, and delivery -- was completed in 25 to 30 minutes per session, with an overall average of 27 minutes. This compares favorably to MRI-guided online adaptive approaches, which require an average of 38 minutes per session, while achieving similar or superior dosimetric outcomes.
A parallel dosimetric comparison was performed by replanning the same patient case on the CyberKnife (CK) system, a robotic non-coplanar delivery platform commonly used for prostate SBRT. This comparison covered only pre-planning, as the CyberKnife currently does not support adaptive replanning -- each session delivers a fixed plan regardless of that day's anatomy.
Ethos plans showed slightly better target coverage -- GTV coverage (V97%) averaged 91.3% with Ethos versus 86% with CyberKnife. For SIB specifically, the Ethos system's multileaf collimator (MLC) allowed effective simultaneous dose modulation between the GTV and PTV levels, while CyberKnife's limited dose modulation capability made true SIB implementation more challenging.
The CyberKnife's advantage was lower low-dose spillage to distant structures. Its non-coplanar beam arrangements reduced the volume of rectum receiving low doses -- for example, rectal V18Gy was 22.2% for CK versus 28.8% for Ethos. However, these non-coplanar beams also exposed a larger bowel volume to low dose (58% vs 38% for Ethos), representing a different dosimetric trade-off rather than a clear advantage.
A major practical distinction was treatment efficiency. CyberKnife required 15,253 monitor units (MUs) to deliver the treatment compared to only 2,639 MUs for Ethos -- a six-fold difference. Higher MU delivery means longer treatment times and greater beam-on time, with potential implications for intrafraction patient motion and total radiation exit dose. This efficiency advantage of the Ethos system compounds its adaptive planning benefits.
This case report demonstrates that CBCT-based online adaptive radiotherapy using the Ethos system is a practical and dosimetrically effective platform for delivering ultra-hypofractionated prostate radiotherapy. The adaptive plans consistently outperformed non-adaptive IGRT plans in both target coverage and organ sparing across all five treatment sessions.
The OART approach eliminates the need for invasive fiducial marker placement -- small gold seeds implanted into the prostate that are required by most IGRT-based UHF RT systems for daily positioning. It also reduces the stringent bladder and rectal preparation protocols required by fixed-plan systems, since daily adaptation compensates for any filling variations. This combination reduces procedural burden for patients without sacrificing precision.
The 27-minute average session time is clinically relevant. Online adaptive workflows must be efficient to integrate into busy radiotherapy departments. The Ethos VMAT approach achieved this efficiency while delivering superior dosimetric quality -- comparing favorably to MRI-guided adaptive systems that require approximately 38 minutes per session and may require more complex infrastructure.
The authors acknowledge that this is a single case report, and broader evidence from larger patient cohorts is needed to confirm clinical outcomes. Future research should examine not only dosimetric performance but also patient-reported toxicities, disease control rates, and quality-of-life outcomes in patients treated with CBCT-based online adaptive UHF RT compared to conventional and other adaptive delivery approaches.