Volumetric‐based image guidance is superior to marker‐based alignments for stereotactic body radiotherapy of prostate cancer

Abstract Purposes The aim of this study was to evaluate a dual marker‐based and soft‐tissue based image guidance for inter‐fractional corrections in stereotactic body radiotherapy (SBRT) of prostate cancer. Methods/Materials We reviewed 18 patients treated with SBRT for prostate cancer. An endorectal balloon was inserted at simulation and each treatment. Planning margins were 3 mm/0 mm posteriorly. Prior to each treatment, a dual image guidance protocol was applied to align three makers using stereoscopic x ray images and then to the soft tissue using kilo‐voltage cone beam CT (kV‐CBCT). After treatment, prostate (CTV), rectal wall, and bladder were delineated on each kV‐CBCT, and delivered dose was recalculated. Dosimetric endpoints were analyzed, including V36.25 Gy for prostate, and D0.03 cc for bladder and rectal wall. Results Following initial marker alignment, additional translational shifts were applied to 22 of 84 fractions after kV‐CBCT. Among the 22 fractions, ten fractions exceeded 3 mm shifts in any direction, including one in the left‐right direction, four in the superior‐inferior direction, and five in the anterior‐posterior direction. With and without the additional kV‐CBCT shifts, the average V36.25 Gy of the prostate for the 22 fractions was 97.6 ± 2.6% with the kV x ray image alone, and was 98.1 ± 2.4% after applying the additional kV‐CBCT shifts. The improvement was borderline statistical significance using Wilcoxon signed‐rank test (P = 0.007). D0.03 cc was 45.8 ± 6.3 Gy vs. 45.1 ± 4.9 Gy for the rectal wall; and 49.5 ± 8.6 Gy vs. 49.3 ± 7.9 Gy for the bladder before and after applying kV‐CBCT shifts. Conclusions Marker‐based alignment alone is not sufficient. Additional adjustments are needed for some patients based kV‐CBCT.


| INTRODUCTION
Stereotactic body radiation therapy (SBRT) is a promising treatment regimen for localized prostate cancer because of the low a/b ratio of the prostate adenocarcinoma. 1 Several recent SBRT studies showed local control rates and toxicity profiles were comparable to the conventional dose regimen of 2 Gy per fraction (2 Gy/Fx). 2,3 It has been reported that with a prescription dose of 7.25 Gy/Fx or higher in SBRT treatment planning, margins are usually 5 mm or smaller to spare the surrounding organs at risk such as rectum, urethra, and bladder. 1 The high-dose conformity of SBRT plans imposes a stringent requirement to the management of inter-and intra-fractional variations.
Intra-fractional motion can be managed with use of an endorectal balloon. [4][5][6][7][8] However, it may introduce a large prostate inter-fraction motion, rotation, and deformation. Jones et al. reported that 69% of fractions required insertion adjustments of the endorectal balloon to reduce prostate rotation and deformation. 6 They recommended to acquire two Cone-Beam Computed Tomography (CBCTs) for each patient: one after insertion and the other after adjustment.
With daily Image-guided Radiation Therapy (IGRT) to correct translational setup error and inter-fractional motion of the prostate, uncorrected prostate rotation became a predominant impact on the dose delivered to the prostate. With 2-, 3-, and 5-mm Planning Target Volume (PTV) margins, Amro et al. showed that only 39%, 65%, and 84% of 26 patients had adequate dose coverage to the prostate without rotation correction, respectively. 9 With a tight planning margin, we implemented a two-step (dual) IGRT protocol for patients receiving prostate SBRT treatment with endorectal balloon. The dual IGRT protocol involved with a markerbased kV x ray guidance and six dimensional (6D) corrections using a robotic couch, and a soft-tissue based kV-CBCT with 3D translational corrections. Using a dosimetric analysis, the aim of this study was to evaluate whether the dual marker-based and soft-tissue based IGRT protocol is sufficient to achieve adequate Clinical Target Volume (CTV) coverage.

| MATERIALS AND METHODS
Eighteen patients treated with five-fraction SBRT for prostate cancer were included in this study. Patients were implanted with three gold fiducial markers at apex, left and right base of their prostate glands.
A total of 84 daily dose fractions were included in the dosimetric analysis (six fractions were not included due to incomplete imaging of the anatomical volume). In 22 of 84 fractions, additional translational corrections guided by kV-CBCT were applied after the kV x ray guided 6D shifts. These 22 fractions were separately analyzed to evaluate dosimetric differences between with and without the additional kV-CBCT shifts after the 6D kV x ray guided shifts.
To minimize the position variations of the internal organs, the patients were instructed to maintain a full bladder and empty rectum before the simulation and treatments. Immediately before the acquisition of the simulation CT and the daily imaging, an endorectal balloon was inserted and filled with 80-100 cc air to immobilize the prostate. Patients were setup in the supine position in a vac-lok bag (Civco Medical Solution, Coralville, IA, USA). After an initial laser alignment to the patients' skin tattoos, daily two-step IGRT protocol was performed. First, a 6D shift was applied using a robotic couch  Descriptive statistics were used for data analysis. The results were expressed as mean AE standard deviation. Wilcoxon signed rank test was used to compare the daily dosimetric endpoints between using the kV x ray guided shifts alone and with the additional kV-CBCT guided shifts. Statistical significance was assigned at P < 0.05.   (Arrows indicate the misalignments).

| RESULTS
with our results, although we used two-step alignments instead of two independent alignments. Several studies 7,8 indicated that utilizing an endorectal balloon can effectively immobilize prostate and restrict rectal filling. The use of endorectal balloon, however, may introduce a large inter-fraction prostate motion and deformation. 4 Neither translational nor rotational correction can resolve issues of organ deformation and organ shape changes. Therefore, online adaptive replanning strategies [13][14][15][16] may provide an optimal solution while maintaining tight planning margins. Without online adaptive planning, alternative strategy is to minimize inter-fractional prostate deformation and rotation. Instead of inserting the rectal balloon, one can use the transperineal ultrasound to monitor intra-prostate motion. Instead of using implanted markers as the surrogate of the prostate, one can directly align to the prostate soft tissue.
One limitation of this study is no consideration of the intra-fractional variation, although other studies have shown the effectiveness of prostate immobilization with the use of rectal balloon. 4,7,8 A recent prostate SBRT study also showed that the compensation of intra-fractional motion on a beam-by-beam basis has very little impact on the final dose parameters. 16 Therefore, we believe that the prostate motion between the two IGRT acquisition time delta is negligible.
It has been reported that alignment to markers often results in a larger rotation than with soft-tissue alignment. 17

CONFLI CTS OF INTEREST
The authors have no relevant conflicts of interest to disclose.