CBCT based online daily adaptive treatment improves target coverage with limited benefit for OAR protection in SBRT prostate cancer

Abstract

Background

The precise application of stereotactic body radiation therapy (SBRT) for prostate cancer is challenged by prostate motion and anatomical variations in surrounding organs such as the bladder and rectum. Online adaptive radiotherapy (oART) using cone beam computed tomography (CBCT) has been shown to improve target coverage while limiting doses to organs-at-risk (OARs).

Methods

We retrospectively analyzed data from 30 prostate cancer patients who were treated with CBCT-based oART simultaneous integrated boost SBRT in five fractions, divided into two dose groups: lower dose group (LDG) receiving 30 Gy to seminal vesicles and prostate, and 35 Gy to prostate, and higher dose group (HDG) receiving 30 Gy to the vesicles and prostate, and 37.5 Gy boost to the prostate. Target volumes and OARs were contoured and evaluated per fraction. For each fraction, SCH (Scheduled plan) and ADP (Adaptive plan) plans were generated. Eight target volumes and fifteen OAR DVH (Dose Volume Histogram) parameters were collected and analyzed. Statistical comparisons were performed between differences in target coverages and doses to OARs in SCH and ADP plans, and between the LDG and HDG groups.

Results

Target coverages were significantly improved in ADP plans in both cohorts and across the entire cohort (p < 0.001). SCH plans often failed to achieve the goal for PTV prostate. The median differences between target volume parameters of SCH and ADP plans were consistent, the adaptive plans were always superior regardless of the prescribed dose. Significant deterioration in ADP was found across the entire cohort in 7 OAR volumes. Dose reductions in OARs due to adaptation were less consistent. Significant differences were observed in bladder, rectum, sigmoid, femur, and penile bulb parameters in the sub-group analysis. The LDG-ADP plans resulted in slightly higher values of rectum_D0.04cm3 and rectum_D50% than LDG-SCH, whereas in the HDG-ADP plans significantly lower doses for the rectum _D0.04cm3, rectum_D1cm3, bladder wall_V18.3 Gy, bladder_D40% and higher value for bladder_V37Gy, bladder wall_D15cm3, penile bulb_{D5 cm3}, penile bulb_D3cm3, penile bulb_D5cm3, sigmoid colon_D20cm3 were achieved compared to the HDG-SCH plans. The magnitude of target coverage improvement was similar in the LDG and HDG groups, although higher prescribed doses were related to increased OAR doses in both SCH and ADP plans, particularly in the HDG group.

Conclusions

CBCT-based online adaptive SBRT for prostate cancer significantly improves target coverage compared to the scheduled plans. Although the benefits of OARs dose vary, adaptive planning offers significant dosimetric advantages, especially at higher prescribed doses. These findings support the integration of oART into prostate SBRT to optimize the treatment’s efficacy and safety.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s13014-026-02888-9.

Introduction

In the modern era of radiotherapy, there are numerous options for the effective treatment of patients with early, organ-confined prostate cancer (PCa). This patient cohort could be treated with brachytherapy and external beam radiotherapy (EBRT) [1]. The stereotactic body radiation therapy (SBRT) is a commonly and widely used EBRT technique to treat low- and intermediate- risk PCa. Using this high-end technology is non-inferior, compared to the conventional radiotherapy approach in terms of tumour control and toxicity, confirmed in many clinical trials, for example in PACE B study, and in other prospective studies [2–6]. It is well known that the prostate gland and seminal vesicles are not independently moving organs, and their volume does not change rapidly. However, movements and forces caused by the surrounding organs, such as bladder, rectum, and intestines, can displace the prostate gland or the seminal vesicles, changing their shape [7–10]. Since we deliver high doses per fraction with high dose gradients and high conformity during SBRT treatments, high precision is required. Online adaptive radiotherapy (oART) is the technique that takes into account for daily positions of the organs and volume changes. It is proved, that oART is a well-applied radiotherapy approach to treat PCa patients [11–18]. In many cases we can achieve better dose coverage in terms of clinical target volume (CTV) and planning target volume (PTV), meanwhile the organs at risk doses could be lower than in non-adaptive technique. Waters et al. already confirmed that treating PCa patients with cone beam CT (CBCT) based online adaptive therapy, has a dosimetric benefit, based on data of 7 patients [13]. The clinical relevance of the improved dose coverage that can be achieved through adaptation may be a higher therapeutic dose delivered to the tumor-infiltrated organ, thereby presumably improving the tumor-killing effect, which may have an impact on local control and possibly on expected survival. In addition, to take into account the position of an OAR every day, we can achieve the same or better organ protection while increasing the target volume dose, so the number and severity of possible side effects are not expected to change. These results are mostly from dosimetric studies, so they will only really be clinically relevant when the long-term follow-up results come out. In our study, we retrospectively analysed 30 PCa patients who underwent oART with simultaneous integrated boost (SIB) SBRT treatments. We created two subgroups, based on the prescribed dose. Dose-volume parameters of organs at risk (OARs) and target volumes were evaluated on a fractional basis. Our goal was to investigate the benefit of oART SIB SBRT in PCa in a larger cohort of patients. In addition, we collected 23 different dosimetric parameters, and compared them in different dose cohorts. We hypothesized that we would find significant differences between scheduled and the daily online generated adapted radiotherapy plans, and prove the dosimetric benefit, in case of oART for SIB SBRT treatments.

Methods

Patient selection and groups

Thirty patients with low, intermediate and high risk PCa cancer (clinically N0M0) were selected for our evaluation. All of the patients were treated in 5 fractions with CBCT-based oART technique. The treated volumes were the prostate and the seminal vesicle (prostate and proximal seminal vesicle in one volume, called pvs), as weel the prostate itself (pros), treated simultaneously. Two groups, with fifteen patients in each, were created according to the prescribed dose. In the lower dose group (LDG), the seminal vesicles and prostate (prostate+seminal vesicles, pvs), were treated with 30 Gy, and simultaneously the prostate (prostate, pros) with 35 Gy. In the higher dose group (HDG), the seminal vesicles and prostate (pvs), were treated with 30 Gy, and simultaneously the prostate (pros) with 37.5 Gy. Based on our institutional protocol, patients aged between 70 and 75 years were treated with 37.5 Gy to the prostate, and patients older than 75 years with 35 Gy to the prostate. Note that in this study, high risk cases were also treated with local SBRT, since according to our institutional protocol, high-risk patients, localized PCa with clinical N0M0 stage above 70 years of age are not treated regionally, only locally to avoid toxicity. Therefore, none of the patients had received elective irradiation of lymph node region. Data from 30 patients were analysed, 25 of whom received hormone therapy. The average PSA level was 13.4 ng/ml (min. 0.15, max. 66.03), and the average Gleason score was 7 (min. 6, max. 9).

Treatment planning and contouring

A Varian Ethos (Varian, Palo Alto, CA, USA) linear accelerator was used to treat the patients. This machine is suitable for performing CBCT-based oART treatments. The radiotherapy plans were generated in the dedicated Ethos treatment planning system V1.0. Intensity modulated radiotherapy (IMRT) plans with the sliding window technique were created. Plans made with 9 or 12 fields were selected depending on which plan resulted in the better reference plan. The linear accelerator is a 6 MV monoenergetic system, without a flattening filter. In the planning system, the minimum calculation grid size was 2.5 mm, and Acuros XB dose calculation algorithm was used. A template containing pre-defined clinical goals was used for all plans to create consistent plans. Beyond that, these goals helped to evaluate the adapted sessions. All of the plans were created with SIB technique. All reference plans and daily adapted plans were checked using Mobius 3D (Varian, Palo Alto, CA, USA) secondary dose calculation software, fraction by fraction.

CTV prostate volume (CTVpros) was contoured as the prostate. The CTVpvs volume was contoured as CTVpros plus the proximal seminal vesicles (1 cm of the vesicles was included in intermediate risk, and 2 cm of vesicles were included in high risk, non T3b cases, and all the vesicles were included in T3b cases where the tumour invaded the vesicles), and the prostate with a 5 mm margin, covering the microscopic spread, except towards the rectum where no margin was used. A 5 mm isotropic margin around the CTVs was used to create PTVpros and PTVpvs. Bladder, bladder wall, rectum, sigmoid colon, penile bulb, right and left femoral heads were contoured as OARs. These volumes are called influencer structures. The bladder wall was created with a 3 mm contour subtraction from the bladder contour. No rectal spacer was used.

In an oART session, the AI generated OAR volumes were evaluated and modified by a radiotherapy technologist and a physician. The original target volumes were deformed by the recontoured influencer structures, and these volumes were adjusted by a physician, fraction by fraction. There was a possibility to accept the reference plan, which is called the scheduled plan (SCH). In that case, the reference plan is re-calculated based on the daily anatomy using the original beam configurations and monitor units. The adapted plan (ADP) is a re-optimized plan that takes all anatomical changes into account.

The acceptance of the daily plan was performed by a physician who assessed the DVH, the traffic light system, and the isodose lines. The traffic light system is a decision support system, where the predefined goals are presented visually with green, yellow, or red colours indicating their fulfillment. The green means that the value is within tolerance, the yellow means it is within tolerance with minor deviation, and red means it is out of tolerance.

The goals were used uniformly for every patient and predefined as a template. Only three planning physicists were involved in the planning, and every plan underwent a multiple stages of review. This ensured the consistent quality of the plans. We also used standardized priorities, so the reference plans and the daily online adapted plans were consistent.

Our goals are listed in Table 1 of Supplementary Material.

Table 1.

Median values with interquartile ranges of the dose-volume parameters of target volumes for SCH and ADP plans of all patients. The difference is the median of the fractional differences

	CTVprosV100%(%)	CTVpvs V100%(%)	PTVpros V100% (%)	PTVpros V95% (%)	PTVpros D0.01cm3 (%)	PTVpros Dmean (%)	PTVpvs V100% (%)	PTVpvs V95% (%)
SCH	99.4 (97.9–100.0)	100 (99.7–100.0)	89.9 (84.3–93.2)	97 (93.7 99.1)	113.0 (110.9-115.1)	104.8 (103.6-105.8)	95.6 (92.1–97.4)	97.6 (95.2–98.8)
ADP	100 (99.8–100.0)	100 (100.0-100.0)	95.1 (92.2–97.3)	99.4 (97.4–99.9)	114.0 (112.4-116.5)	106.2 (105.0-107.3)	99.1 (98.0-99.6)	100 (99.7–100.0)
Difference (ADP-SCH)	-0.4 (-1.7-0.0)	0.0 (-0.3-0.0)	-5.4 (-10.0-( -1.2))	-1.7 (-4.2-( -0.1))	-1.0 (-3.5-0.7)	-1.6 (-3.0-(-0.3))	-3.3 (-6.3-(-1.0))	-2.2 (-4.7-( -0.8))
P value	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001

SCH: scheduled plan in which the dose is re-calculated on daily anatomy without adaptation, ADP: adapted plan in which the dose is re-optimised on daily anatomy

Dosimetric analysis

The treatment data were retrospectively analyzed. Parameters of eight different target volumes (TVs) and fifteen OARs were collected from five fractions of thirty patients.

The TV and OAR parameters are shown in Table 1 of Supplementary Material. We compared the SCH and ADP data per patient, also for the whole cohort, and separately for the LDG and HDG groups. We calculated the differences between SCH-ADP and HDG-LDG for each fraction and compared the medians of these values. We also calculated the median of the SCH, ADP, HDG, and LDG values.

Statistical analysis

We used the Shapiro-Wilk test to determine the normality of the data distribution. To compare non-normally distributed data, we used the Wilcoxon signed-rank test for paired data, and the Mann-Whitney test for unpaired data. We applied Student’s t test for normally distributed data. We compared the SCH and ADP plans using paired tests, while the LDG and HDG groups were compared using unpaired tests. Bonferroni corrections were not applied. The difference was considered significant if the p-value was < 0.05. Prism 8.0.1 (GraphPad Software Inc., Boston, MA, USA) was used for statistical calculation.

Results

All 30 patients were successfully treated in 5 fractions without any interruption, and modification of the fraction number. Out of 150 treatment sessions, SCH plan was chosen on only three occasions (2%). After consulting with the treating physician, he/she chose the SCH plan at his/her discretion because it provided better OAR protection. Evaluating the differences between target coverages in the whole cohort, an advantage was found with higher values in favour of the ADP plans (Table 1).

Considering all target volume parameters, a higher dose to the tumour was delivered in the ADP plan. For the PTVpros_V100% parameter, SCH plans failed to meet the minimum requirement of V100% > 90%.

Figure 1 shows that the parameters analysed for the prostate have a much larger variance compared to the parameters for the PVS when analyzed over the whole cohort in every fraction. The left figure shows the percentage volumes of four different PTVpros parameters in every fraction, and the right figure shows the volumes of four different PTVpvs parameters in every fraction. In both volumes, deviations between the ADP plans are smaller than those between the SCH plans.

Fig. 1.

Target volume coverages of the PTV_pros (left), and PTV_pvs (right) in the function of fractions. The black line shows the goal of V100%

In the cases of the bladder wall_D15cm3, rectum_D50%, bladder_D40%, penile bulb_D5cm3, penile bulb_D3cm3, sigmoid colon_D0.04cm3, and sigmoid colon_D20cm3, significantly higher values were observed in the ADP plans, and lower for sigmoid colon_D0.04cm3, and Bladder wall_V18.3 Gy than in the SCH plans (Table 2).

Table 2.

Median values with interquartile range of the dose-volume parameters for OARs in the SCH and ADP plans in all patients. The difference is the median of the fractional differences

	Bladder wall V18.3 Gy (cm3)	Bladder wall D15cm3 (Gy)	Bladder wall D0.04cm3 (Gy)	Rectum D1cm3 (Gy)	Rectum D0.04cm3 (Gy)	Rectum D50% (Gy)	Rectum D20% (Gy)	Bladder D40% (Gy)
SCH	15.6 (11.9–19.7)	2.7 (2.0-4.4)	7.7 (7.4–8.2)	6.9 (6.7–7.2)	7.3 (7.1–7.6)	3.1 (2.9–3.4)	4.6 (4.2-5.0)	2.3 (1.5–3.2)
ADP	14.4 (12.4–16.6)	2.9 (1.9–3.9)	7.8 (7.4–8.2)	6.9 (6.6–7.1)	7.3 (7.1–7.5)	3.2 (3.0-3.5)	4.7 (4.3-5.0)	2.3 (1.6-3.0)
difference (ADP-SCH)	-0.3 (-0.8-2.7)	0.0 (0.0-0.3)	0.0 (-0.2-0.1)	0.0 (-0.1-0.2)	0.0 (-0.2-0.3)	0.1 (-0.3-0.1)	0.0 (-0.3-0.2)	0.1 (-0.1-0.4)
P value	0.032	0.025	0.238	0.107	0.209	0.002	0.320	0.005

	Bladder V37Gy (cm3)	Penile bulb D5cm3 (Gy)	Penile bulb D3cm3 (Gy)	Sigmoid colon D0.04cm3 (Gy)	Sigmoid colon D20cm3 (Gy)	Femur h. left D10cm3 (Gy)	Femur h. right D10cm3 (Gy)
SCH	1.6 (0.1–4.5)	0.8 (0.5–2.4)	0.5 (0.4–0.8)	0.5 (0.3–1.5)	0.2 (0.2–0.4)	2.7 (2.3–3.1)	2.7 (2.4-3.0)
ADP	1.8 (0.0-5.4)	1.2 (0.6–2.3)	0.6 (0.4–0.8)	0.5 (0.4–2.1)	0.2 (0.2–0.4)	2.7 (2.3–3.1)	2.6 (2.3-3.0)
difference (ADP-SCH)	0.0 (-0.4-1.4)	0.1 (-0.6-0.1)	0.0 (-0.1-0.0)	0.0 (-0.1-0.0)	0.0 (0.0–0.0)	0.0 (-0.1-0.2)	-0.1 (-0.1-0.2)
P value	0.185	0.005	0.011	0.007	0.025	0.203	< 0.001

SCH: scheduled plan in which the dose is re-calculated on daily anatomy without adaptation, ADP: adapted plan in which the dose is re-optimised on daily anatomy

We also performed analyses on paired SCH and ADP plans. The differences between the two groups are presented in Table 3. The median values of parameters for target volumes in SCH and ADP plans were grouped by the prescribed dose. We found that all ADP plans showed better results by the same magnitude compared to SCH, regardless of the prescribed dose (ADP-SCH difference). It can be seen from the data that all of the prostate target volume parameters show significantly lower values in the HDG group, compared to LDG, except for the D0.01cm³ parameter, which was higher.

Table 3.

Median values with interquartile ranges of dose-volume parameters for target volumes in patients in the higher dose group (HDG) and lower dose group (LDG) with the corresponding SCH and ADP dose parameters. The differences are calculated from medians of the fractional differences for HDG and LDG, as well as ADP and SCH groups

	Type of plan	HDG	LDG	difference (HDG-LDG)	P value
CTVprosV100% (%)	SCH	98.6 (97.3–99.6)	99.9 (99.0-100)	-0.5 (-2.1-0.1)	< 0.001
	ADP	99.8 (99.6–99.9)	100.0 (100–100)	-0.2 (-0.4-0.0)	< 0.001
difference (ADP-SCH)		1.0 (0.0-2.7)	0.1 (0.0–1.0)
P value		< 0.001	< 0.001
CTVpvsV100% (%)	SCH	100 (99.7–100)	100 (99.6–100)	0.0 (-0.2-0.3)	0.523
	ADP	100 (100–100)	100 (100–100)	0.0 (0.0–0.0)	0.242
difference (ADP-SCH)		0.0 (0.0-0.3)	0.0 (0.0-0.4)
P value		< 0.001	< 0.001
PTVprosV100% (%)	SCH	87.9 (82.3–91.9)	91.1 (87.0-94.3)	-2.2 (-9.5 2.1)	0.005
	ADP	93.9 (90.9–95.4)	97.1 (94.3–98.6)	-3.2 (-6.2-(-1.2))	< 0.001
difference (ADP-SCH)		5.5 (-0.1-9.1)	5.3 (1.5–10.1)
P value		< 0.001	< 0.001
PTVprosV95% (%)	SCH	95.1 (91.7–97.6)	98.5 (96.9–99.4)	-2.7 (-6.4-0.6)	< 0.001
	ADP	97.5 (95.9–98.6)	99.9 (99.7–100)	-2.4 (-3.5-(-0.4))	< 0.001
difference (ADP-SCH)		2.5 (-0.1-4.7)	1.4 (0.3–3.1)
P value		< 0.001	< 0.001
PTVprosD0.01cm3 (%)	SCH	114.0 (112.8-115.1)	111.6 (110.3-114.7)	1.9 (-1.9-4.6)	0.003
	ADP	114.8 (113.2-116.7)	113.2 (111.4–116)	1.4 (-0.5-3.8)	0.003
difference (ADP-SCH)		1.0 (-0.7-3.1)	1.1 (-0.9-3.7)
P value		< 0.001	< 0.001
PTVprosDmean (%)	SCH	105.1 (103.5-106.5)	104.7 (103.8-105.4)	0.4 (-0.8-1.6)	0.200
	ADP	106.5 (105.4-107.5)	105.9 (104.8-107.2)	0.4 (-1.1-2.1)	0.196
difference (ADP-SCH)		1.6 (0.6–2.5)	1.6 (0-3.4)
P value		< 0.001	< 0.001
PTVpvsV100% (%)	SCH	95.6 (92.4–97.5)	95.7 (92.0-97.3)	0.6 (-4.3-4.5)	0.863
	ADP	99.2 (98.0-99.5)	98.8 (97.9–99.8)	-0.2 (-1.1-1.3)	0.704
difference (ADP-SCH)		3.3 (1.0-6.8)	3.3 (1.3–6.1)
P value		< 0.001	< 0.001
PTVpvsV95% (%)	SCH	97.4 (94.4–98.7)	97.8 (95.7–98.9)	-0.3 (-3.1-2.4)	0.306
	ADP	99.9 (99.6–100)	99.9 (99.8–100)	0.0 (-0.4-0.1)	0.331
difference (ADP-SCH)		1.9 (0.6–4.8)	2.3 (1.0-4.1)
P value		< 0.001	< 0.001

SCH: scheduled plan in which the dose is re-calculated on daily anatomy without adaptation, ADP: adapted plan in which the dose is re-optimised on daily anatomy

For 14 OAR parameters, we found that the HDG had significantly higher dose parameters in both SCH and ADP plans (Table 4). The rectum_D1cm3 parameter was an exception, because the ADP value was lower in the HDG group (see Table 4, difference (HDG-LDG); -0.22). A relatively large difference was found between the LDG and HDG groups for the bladder_V37Gy parameter. This difference was observed in SCH and in ADP plans, too. In the case of the HDG, even the ADP plans slightly exceeded the limit of our protocol.

Table 4.

Median values with interquartile ranges of dose-volume parameters for target volumes in patients in the higher dose group (HDG) and lower dose group (LDG) with the corresponding SCH and ADP dose parameters. The differences are calculated from medians of the fractional differences for HDG and LDG, as well as ADP and SCH groups

	Type of plan	HDG	LDG	difference (HDG-LDG)	P value
Bladder wallV18.3 Gy (cm3)	SCH	14.5 (12.2–17.5)	17.5 (14.1–20.3)	-4.0 (-7.5-4.8)	0.613
	ADP	13.5 (11.2–15.1)	17.3 (13.6–19.6)	-4.0 (-8.3-2.3)	0.639
difference (ADP-SCH)		-0.7 (-2.0-1.6)	0.1 (-0.6-1.1)
P value		0.004	0.573
Bladder wallD15cm3 (Gy)	SCH	2.5 (1.9–4.8)	2.8 (2.0-4.1)	0.2 (-1.8-1.7)	0.436
	ADP	2.8 (1.8–4.5)	2.9 (2.0-3.7)	0.1 (-1.1-1.3)	0.591
difference (ADP-SCH)		-0.1 (-0.4-0.1)	-0.1 (-0.5-0.2)
P value		0.010	0.456
Bladder wallD0.04cm3(Gy)	SCH	8.2 (7.9–8.4)	7.4 (7.3–7.5)	0.8 (0.6-1.0)	< 0.001
	ADP	8.2 (7.9–8.4)	7.4 (7.3–7.5)	0.7 (0.4–0.9)	< 0.001
difference (ADP-SCH)		0.0 (-0.2-0.1)	0.1 (-0.1-0.2)
P value		0.495	0.008
RectumD1cm3 (Gy)	SCH	7.0 (6.6–7.3)	6.9 (6.7-7.0)	0.0 (-0.4-0.4)	0.338
	ADP	6.7 (6.3–7.1)	7.0 (6.8–7.1)	-0.3 (-0.5-0.1)	< 0.001
difference (ADP-SCH)		-0.1 (-0.5-0.0)	0.00 (-0.1-0.2)
P value		0.001	0.116
RectumD0.04cm3 (Gy)	SCH	7.6 (7.3–7.8)	7.2 (7.0-7.4)	0.3 (-0.1-0.7)	0.899
	ADP	7.4 (7.2–7.5)	7.3 (7.1–7.5)	0.0 (-0.2-0.3)	0.493
difference (ADP-SCH)		-0.2 (-0.4-0.0)	0.1 (-0.1-0.2)
P value		0.003	0.008
RectumD50% (Gy)	SCH	3.2 (3.0-3.7)	3.0 (2.8–3.2)	0.2 (-0.1-0.7)	< 0.001
	ADP	3.3 (3.0-3.7)	3.1 (3.0-3.3)	0.1 (-0.1-0.5)	0.018
difference (ADP-SCH)		0.0 (-0.2-0.3)	0.1 (-0.1-0.3)
P value		0.329	< 0.001
RectumD20% (Gy)	SCH	4.7 (4.4–5.2)	4.5 (4.2–4.9)	0.3 (-0.5-0.8)	0.023
	ADP	4.7 (4.3–5.2)	4.7 (4.3–4.9)	0.1 (-0.3-0.4)	0.247
difference (ADP-SCH)		0.0 (-0.3-0.3)	0.0 (-0.1-0.3)
P value		0.784	0.054
BladderD40% (Gy)	SCH	2.8 (1.8–3.5)	2.1 (1.3–2.7)	0.8 (-0.7-1.8)	< 0.001
	ADP	2.4 (1.8–3.2)	2.1 (1.5–2.7)	0.5 (-0.5-1.1)	0.023
difference (ADP-SCH)		-0.3 (-0.7-0.0)	0.0 (-0.2-0.2)
P value		< 0.001	0.358
BladderV37 Gy (cm3)	SCH	4.6 (2.8–11.3)	0.1 (0.0-0.4)	4.3 (2.4–9.3)	< 0.001
	ADP	5.3 (2.9–6.9)	0.1 (0.0-0.4)	4.9 (2.3–6.5)	< 0.001
difference (ADP-SCH)		-1.3 (-3.4-1.0)	0.0 (-0.1-0.2)
P value		0.007	0.092
Penile bulbD5cm3 (Gy)	SCH	0.8 (0.5–2.3)	0.7 (0.5–2.7)	0.0 (-0.8-0.5)	0.244
	ADP	1.3 (0.7–2.6)	0.9 (0.5–2.1)	-0.1 (-0.5-0.9)	0.080
difference (ADP-SCH)		0.0 (0.0-0.5)	0.0 (-0.2-0.5)
P value		0.004	0.211
Penile bulbD3cm3 (Gy)	SCH	0.5 (0.4–0.7)	0.4 (0.4-1.0)	0.0 (-0.4-0.2)	0.068
	ADP	0.6 (0.5–0.8)	0.5 (0.4–0.8)	0.1 (-0.3-0.3)	0.007
difference (ADP-SCH)		0.0 (0.0-0.2)	0.0 (-0.1-0.1)
P value		0.014	0.240
Sigmoid colonD0.04cm3 (Gy)	SCH	0.7 (0.4–2.9)	0.4 (0.3–0.7)	0.0 (-0.4-1.2)	0.015
	ADP	0.7 (0.4–3.2)	0.4 (0.3–0.9)	0.0 (-0.4-1.8)	0.008
difference (ADP-SCH)		0.0 (0.0-0.1)	0.0 (0.0-0.1)
P value		0.057	0.059
Sigmoid colonD20cm3 (Gy)	SCH	0.3 (0.2–0.4)	0.2 (0.2–0.3)	0.1 (-0.1-0.2)	0.004
	ADP	0.3 (0.2–0.4)	0.2 (0.1–0.3)	0.0 (-0.1-0.2)	0.007
difference (ADP-SCH)		0.0 (0.0–0.0)	0.0 (0.0–0.0)
P value		0.035	0.347
Femur h. leftD10cm3 (Gy)	SCH	2.8 (2.3–3.3)	2.7 (2.3–2.9)	-0.2 (-0.6-0.9)	0.271
	ADP	2.8 (2.4–3.1)	2.6 (2.3–2.9)	0.1 (-0.4-0.8)	0.097
difference (ADP-SCH)		0.0 (-0.1-0.2)	-0.1 (-0.2-0.1)
P value		0.762	0.021
Femur h. rightD10cm3 (Gy)	SCH	2.8 (2.4–3.2)	2.7 (2.5–2.9)	0.0 (-0.5-0.5)	0.402
	ADP	2.7 (2.3–3.2)	2.6 (2.3–2.9)	0.1 (-0.5-0.7)	0.291
difference (ADP-SCH)		0.0 (-0.2-0.1)	-0.1 (-0.2-0.0)
P value		0.084	< 0.001

SCH: scheduled plan in which the dose is re-calculated on daily anatomy without adaptation, ADP: adapted plan in which the dose is re-optimised on daily anatomy

In the case where the two groups were analysed separately, we found that in all cases the ADP plans achieved significantly better target coverage than the SCH plans (Table 3). In terms of OARs, in the LDG group, the bladder wall_D0.04cm3, rectum_D0.04cm3, rectum_D50%, parameters were significantly higher, and the left and right femur h._{D10 cm3} were lower in the ADP plans than in the SCH plans (Table 4).

In the HDG group, the value of the bladder wall_V18.3 Gy, rectum_D1cm3, rectum_D0.04cm3, bladder_D40% were significantly lower, but bladder wall_D15cm3, bladder_V37Gy, penile bulb_{D5 cm3}, penile bulb_D3cm3, and sigmoid colon_D20cm3 were higher.

In Fig. 2, we show 3 different evaluated parameters, where a significant difference was found between HDG and LDG.

Fig. 2.

Box and whiskers plot of the dose-volume parameters of OARs where significant differences were observed. The boxes represent the medians and interquartile ranges (75th and 25th percentiles). SCH: scheduled, ADP: adaptive, LDG: lower dose group, HDG: higher dose group. The black lines show the institutional OAR goals

Discussion

In this work, using a large number of patients and fractions, we demonstrated that online adaptation can be beneficial for patients with prostate tumour undergoing SBRT radiotherapy. In our study, we found a clear and significant advantage of adaptation in all eight TV parameters for the entire cohort. The V100% parameter of the PTVpros must be at least 90%, but in the SCH plans, we found a median value of 89.9%, which is less than required in our institutional and international protocols [19]. However, in the ADP plans, in all cases we were able to meet the acceptance criterion, similar to Waters et al. [13]. Fisher and colleagues also achieved similarly good target volume coverage, but in their study, they analyzed data from only 6 patients, and evaluated the D98% values, whereas we used the V100% and V95% in our study. However, they found differences in D98% of PTV with significant improvement with adaptation (p < 0.0001; median D98% 91.8–96.5% [14]. With regard to the OARs, we found that, for 7 of the parameters examined, the ADP group had significantly higher values, and for two parameters, lower than the SCH group, when the entire cohort was analyzed. We found this result surprising because it did not support our hypothesis. It should be noted that the differences were minimal. Compared with the findings of Fisher and colleagues, no clear advantage was demonstrated in terms of organ protection either, rectum_V40Gy median dose values were 24% vs. 19.8% in SCH vs. ADP, bladder V40Gy median dose values were 31% vs. 32.6% in SCH vs. ADP (both p < 0.001). It should be noted that fewer parameters of organs were evaluated (V40Gy and mean dose for rectum, V40Gy and mean dose for bladder, D0.1cm³ for bowel), and the power of significance was reduced due to the low number of patients [14]. In contrast, Waters et al. [13] found that there was a small but statistically significant difference in the rectum D0.03cm³ parameter for ADP plans, and a minimal dose increase for the bladder parameter when the ADP plans were considered for the whole cohort.

We observed that, regardless of the prescribed dose, all ADP plans showed a similar improvement in prostate target volume parameters compared to the SCH plans. This is possible because we used similar clinical goals for optimization. We observed that the target volume parameters for prostate were significantly lower in SCH and ADP plans in the HDG group compared to the LDG group. This can be explained by the fact that we had to reduce the dose coverage of the target volumes in the HDG group if the OAR dose limitations were prioritized (but still within the target acceptance criteria) to keep the dose to surrounding critical organs within predetermined dose limits. In favour of ADP, the largest deviation was found in PTVpros_V100% (HDG: 93.9 (90.9–95.4) and LDG: 97.1 (94.3–98.6)). The PTVpros_D0.01cm3 was an exception, it was higher in the HDG group. This is because, with higher dose treatment, in order to achieve the same OARs protection as for lower dose treatment, the TPS has to create a higher dose maximum to achieve a larger dose fall off.

When subgroups with lower and higher prescribed doses were analysed separately, there was no clear improvement in favour of the ADP in terms of OARs. In the LDG, the bladder wall_D0.04cm3, rectum_{D0.04 cm3}, rectum_D50%, parameters were significantly higher, and the left and right femur h._D10cm3 were lower in ADP plans than in SCH plans. In reviewing these plans, we found that, based on the clinical goals used in the reference plan, the system prioritized target volume coverage over the OAR protection during the re-optimization of the ADP. Thus, coverage of the target volume was of paramount importance, but at the same time, the acceptance criteria for the organs at risk in the ADP also had to be met. In contrast, in the HDG group, significantly lower dose values were obtained in ADP plans for the bladder wall_V18.3 Gy, rectum_D1cm3, rectum_D0.04cm3, bladder_D40%, but for the bladder wall_D15cm3, bladder_V37Gy, penile bulb_D5cm3, penile bulb_D3cm3, sigmoid colon_D20cm3 the doses were higher than in the SCH plans. Because the decimal places had to be rounded, the slight differences between the values may not always be clearly visible in the tables, in some cases. In this group, the prescribed dose was 7.5 Gy per fraction. This dose level has proven to be high enough in nearly all treatments to allow for a much greater emphasis on protecting the OARs. Consequently, the optimization criterion for the rectum and bladder in the reference plans, and thus also in the re-optimized plans, resulted in lower doses for the two organs to be protected. With regards to the bladder wall_D15cm3 parameter, we found that this parameter is generally an easily achievable clinical goal, in contrast to the requirements for the maximum dose for the rectum, so we allowed a higher dose at the expense of the bladder wall_D15cm3 parameter, to meet the maximum requirements. For similar reasons, we allowed higher doses for adaptive plans with other organs. In terms of bladder_V37Gy, we found that the larger volume of the bladder receives a higher dose during adaptation, but this may be because in the HDG group a larger proportion of the vesicles had to be included in the target volume, so that, taking the daily anatomical movement into account this was essential for adequate coverage of the vesicles. In summary, by prioritizing the goals of the targets, the planning system is able to maintain the dose to organs at risk even when there is potential overlap between the target volume and organs at risk.

When comparing data for OARs in the LDG and HDG groups, we found significantly lower median dose values in the LDG group in both the SCH and ADP plans for 14 parameters. These results can be explained by the fact that the prescribed dose was lower in the LDG group. In the HDG group, we found that the rectum_D1cm3 parameter was significantly lower in the ADP plans, than in the LDG group. This can be explained by the fact that the rectum is a critical organ, and its wall is in close contact with the prostate. This makes it very difficult to meet the dose limits when a higher prescribed dose is applied. Therefore, when planning, we must prioritize the protection of this organ, even at the expense of the prostate.

When high-dose online adaptive prostate treatments were implemented, we expected that the maximum dose to the rectum would decrease in all cases with adaptive plans, regardless of the prescribed doses. However, the analyzed data did not support our initial hypothesis, on the other hand, theydemonstrated that this technique can be safely used with regard to rectal dose. Based on the entire cohort, we observed more significant deterioration in the ADP group; however, at sub-group analysis, we were able to achieve a significantly lower dose for maximal rectum dose in the HDG group. Furthermore, it has been confirmed that this technique can be safely used with regard to the rectal dose.

There are several scientific papers in the literature that have demonstrated the combination of target volume advantage and OAR protection in oART [14, 20–24]. However, in most cases, the number of patients was low, but the fraction number was high. The question is whether we can apply our results to larger patient population. It is worth mentioning the work of Prezios et al. [25], who performed dosimetric analysis on prostate cancer patients treated with conventional dose fractionation with CBCT-based adaptive treatment. Their results were similar to ours, with clear advantages in terms of OAR protection.

To the best of our knowledge, this is the largest study to examine the dosimetric efficacy of the CBCT-based online adaptive stereotactic therapy in a cohort of prostate cancer patients, and the only study to examine such a large number of dosimetric parameters of target volumes and organs at risk, comparing two different prescribed dose groups.

After evaluating the very early and early side effects of our first treated 30 patients, the next step for our group will be to investigate how we can decrease the rate of side effects or increase the local control with oART in prostate cancer compared to other external beam radiotherapy techniques. However, based on the excellent results of phase 3 SBRT studies in the international literature [26, 27], we do not expect large differences in toxicity in OARs. Rather, we expect that slightly better local control can be achieved through improved dose coverage of the target volumes.

It is evident that we have excellent coverage of CTVs in all cases, but we must not forget that the adaptation time is long (approx. 15 min), and the treatment time is approx. 4 min. We take a verification image before starting the irradiation, but residual errors may still remain, not to mention anatomical changes during the treatment. We therefore believe that, given the current situation the examination of dose to the PTV cannot be completely disregarded. However, in light of the excellent PTV coverage observed in this study, it is reasonable to consider reducing the CTV-PTV margin. The rationale for reducing OAR doses will be confirmed or refuted when the first clinical results become available with a comparison of toxicities across a large numbers of patients.

We also plan to analyse variations in fraction doses on an individual patient basis and conducti research on the possibility of reducing the PTV margin. This work is ongoing, and the manuscript is under preparation.

Conclusions

In summary, we have successfully introduced an online adaptive stereotactic simultaneous integrated boost technique with online adaptive treatment for prostate cancer patients at our institute. After analyzing the treatment data of 30 patients, we found that the target coverage parameters had improved significantly. In one of the evaluated volumes, only the daily adapted treatment was clinically acceptable.

In contrast, an analysis of the entire patient population revealed no clear benefit of adaptation with regard to the organs at risk However, this finding shows that the target volume coverage can be improved while keeping the dose to OARs at acceptable level, regardless of the prescribed dose. In a subgroup analysis, we found significant differences in several organs at risk. In most cases, the difference occurred when the prescribed dose was higher, and the treatment planning system had to work harder to meet the dose requirements.

Limitation

A limitation of our study is that the daily adapted contouring was not always performed by the same physician, so the results may be influenced by intraobserver variability. In addition, there was also slight difference between the two dose groups in the contouring protocol regarding the target volumes, depending on the extent of the tumour, which may also have had a minimal effect on the dosimetric results. We cannot ignore the fact that further anatomical changes may occur during the irradiation after the adaptation phase [17, 23, 28]. The average irradiation time is approximately 4 min, during which time online verification is not possible.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ADP

Daily adapted plan

AI

Artificial intelligence

CBCT

Cone Beam Computed Tomography

CT

Computer tomography

CTV

Clinical target volume

GTV

Gross tumor volume

HDG

Higher dose group

IMRT

Intensity-modulated radiotherapy

LDG

Lower dose group

LINAC

Linear Accelerator

OAR

Organ-at-risk

oART

Online adaptive radiotherapy

PCa

Prostate cancer

PTV

Planning target volume

PVS

Prostata plus vesicles seminal

SBRT

Stereotactic body radiation therapy

SCH

Scheduled plan

SIB

Simultaneous integrated boost

TPS

Treatment planning system

Author contributions

Á.G. collected and processed the data’s, and wrote the manuscript main text, T. M was the supervisor and wrote the manuscript in consultation with P. Á, CS. P, Z. N.T.

Funding

Open access funding provided by Semmelweis University. There was no any funding for used for this project.

Data availability

Measurement data can be disclosed on request.

Declarations

Ethical approval

This retrospective study was approved by the Regional and Local Institutional Ethical Committee, and was conducted following the principles of the Helsinki Declaration. Document reference number: OOI/Ált06091-2/2025.

Consent for publication

All authors contributed to the publication of the scientific work.

Competing interests

The authors declare no competing interests.