Dosimetric impacts of endorectal balloon in CyberKnife stereotactic body radiation therapy (SBRT) for early‐stage prostate cancer

Abstract Purpose In SBRT for prostate cancer, higher fractional dose to the rectum is a major toxicity concern due to using smaller PTV margin and hypofractionation. We investigate the dosimetric impact on rectum using endorectal balloon (ERB) in prostate SBRT. Materials and Methods Twenty prostate cancer patients were included in a retrospective study, ten with ERB and 10 without ERB. Optimized SBRT plans were generated on CyberKnife MultiPlan for 5 × 7.25 Gy to PTV under RTOG‐0938 protocol for early‐stage prostate cancer. For the rectum and the anterior half rectum, mean dose and percentage of volumes receiving 50%, 80%, 90%, and 100% prescription dose were compared. Results Using ERB, mean dose to the rectum was 62 cGy (P = 0.001) lower per fraction, and 50 cGy (P = 0.024) lower per fraction for the anterior half rectum. The average V50%, V80%, V90%, and V100% were lower by 9.9% (P = 0.001), 5.3% (P = 0.0002), 3.4% (P = 0.0002), and 1.2% (P = 0.005) for the rectum, and lower by 10.4% (P = 0.009), 8.3% (P = 0.0004), 5.4% (P = 0.0003), and 2.1% (P = 0.003) for the anterior half rectum. Conclusions Significant reductions of dose to the rectum using ERB were observed. This may lead to improvement of the rectal toxicity profiles in prostate SBRT.

while minimizing rectal and bladder toxicities to a level that is comparable to those seen in conventional radiotherapy, including 3D-CRT (3D Conformal Radiation Therapy), IMRT (Intensity Modulated Radiation Therapy), and HDR (High Dose Rate Brachytherapy). With the follow-up data approaching 6 years to this date, prostate SBRT has now been considered as an alternative therapeutic option to the conventional radiotherapy for localized prostate cancer, either as a monotherapy for low-and intermediate-risk prostate cancer [13][14][15][16] or post-IMRT boost treatment for high-risk prostate cancer. 17 In SBRT for prostate cancer, due to the much higher dose per fraction and use of smaller PTV margins (2-3 mm posterior, 3-5 mm in all other directions) than those in 3D-CRT or IMRT, it is particularly critical to minimize the prostate motion and the exposure of rectum volumes to intermediate and high dose which are predictive factors for late rectal toxicity. [18][19][20] It has been shown that an air or water-filled endorectal balloon (ERB) can significantly reduce prostate motion [21][22][23] and displace the posterior portion of the rectal wall away from the intermediate-to-high dose regions in 3D-CRT and IMRT. This displacement can lead to significant rectal wall sparing and reducing rectal toxicity from prostate or post-prostatectomy radiation treatment, potentially to allow for further dose escalation to the prostate. [24][25][26][27][28][29][30] For prostate SBRT, late rectal toxicity data are very limited with maximum follow-up just under 6 years. [14][15][16] It can be anticipated that any systematic reduction of rectal dose in such hypofractionated prostate treatment may be beneficial, such as those potentially achievable using ERB to minimize exposing rectal volume to intermediate and high dose. However, to this date, there has been no specific study based on CyberKnife prostate SBRT experience on how using ERB may help to reduce rectum dose and improve rectal dosevolume profiles under the hypofractionated target dose specifications and OARs (organ-at-risk) constraints. In addition, use of ERB has not been included neither on protocols treating prostate alone for early-stage prostate cancer nor on protocols treating both the prostate and the proximal seminal vesicles for intermediate-risk and high-risk prostate cancer.
In this work, we performed a systematic treatment planning study on the potential dosimetric impacts of using ERB for prostate cancer patients who may receive SBRT treatments such as those described in RTOG-0938 "A Randomized Phase II Trial of Hypofractionated Radiotherapy for Favorable Risk Prostate Cancer". 31 2 | MATERIALS AND METHODS

2.A | Patient data and use of ERB
Twenty prostate cancer patient cases previously treated at two institutions were selected for a retrospective CyberKnife SBRT treatment planning study. Ten of the patients had CT simulation and treatment using ERB filled with 60-100 cc water (the ERB group). The other ten patients had CT simulation and treatment with empty rectum and no ERB (the noERB group). Figure 1 shows the endorectal balloon (ERB) used for patients included in this planning study.
All patients had at least three well-spaced fiducials implanted in the prostate for image-guided target volume localization. For patient in the ERB group, immediately before CT simulation scan, an indexed lumen ERB (RadiaDyne, LLC, Houston, TX, USA) was inserted into the rectum, filled with 60-100 cc water so that their density would be essentially homogeneous with that of the surrounding tissue. For patients in the noERB group, instructions were given for pre-CT and pretreatment bowel preparation, including dietary guidelines, use of anti-gas tablets, and administered enemas to ensure an empty rectum. An optimized SBRT plan was generated for each case using the multi-objective sequential optimization in CyberKnife MultiPlan TPS system to meet the five fraction (5 9 725 cGy) dose-specification and dose-volume constraints per RTOG-0938 for early-stage prostate cancer. These plans typically used 2-3 collimators of different sizes, and 100-200 noncoplanar and nonisocentric beams of 6 MV x ray. Target dose coverage was characterized by the new conformity index (nCI), heterogeneity index, and mean PTV dose. Plans were typically prescribed to 79%-85% isodose line (IDL) to ensure that at least 95% of the PTV was covered by the prescription dose. Dosimetric parameters for the rectum and the anterior half of the rectum were compared between the two groups, including the mean dose F I G . 1. The endorectal balloon (ERB) used for patients included in this planning study. and the percentages of volume receiving 50%, 80%, 90%, and 100% of the prescription dose.

2.C | Statistical analysis
Descriptive statistics were calculated for each dosimetric parameter in the ERB group and noERB group. Mean and standard deviation are reported along with the difference in mean value for each parameter and denoted as D. Independent samples t-tests were performed to examine the differences between the two groups.
The analyses were repeated after logarithmic (lg 10) transformation of the dosimetric data. All statistical analyses were performed using the SPSS software for Windows (version 20.0, SPSS Inc., Chicago, IL, USA). Differences with P-value <0.05 was considered statistically significant.   represent the percentage volumes exposed to x% of the prescription dose (36.25 Gy) for the rectum volumes and the anterior half rectum volumes.

| RESULTS
As shown in Table 2 29 reported that in both 3D-CRT and IMRT, using ERB can lead to significant dose reduction for rec- reductions seen in previous studies with 3D-CRT and IMRT. 27,29,30 This difference may be related to two important points emphasized in our study. First, in planning, we strictly followed the RTOG 0938 planning DVHs constraint guidelines, especially the DVH constraints for rectal volumes receiving high doses, V3806 < 1 cc (3806 cGy corresponding to 105% of the prescription dose of 3625 cGy) and V3440 < 3 cc (corresponding to 95% of the prescription dose). Our results showed similar V100% of both the rectum volume and the anterior rectum volume for the ERB group vs. the noERB group, as revealed in Fig. 4. Second, the use of non-coplanar and non-isocentric beams from CyberKnife physically offered a better rectal    and DVHs profiles of rectum, especially the anterior half of the rectum, should be minimized to ALARA (As Low As Reasonably Achievable) while respecting the relatively general objectives for rectum DVHs. 15 Overall, this study showed systematically better DVH profiles for rectum using ERB. Potentially, such dosimetric improvements may lead to better rectal toxicity profiles for SBRT treatment; yet, this hypothesis remains to be examined with clinical follow-up data when it becomes available.
It should be noted that this study is limited to SBRT planning for early-stage prostate cancer where the CTV is defined as prostate volume alone without including the seminal vesicles (SV). This may explain why in general we achieved overall much better rectum DVH profiles (as seen in Table 2 and Fig. 3) than those described by King et al. 15 This can be due to the fact that smaller section of the rectum is involved when CTV includes prostate alone. Patel et al. 30 reported significant rectal sparing using ERB in IMRT for five patients with and without inclusion of seminal vesicles. In comparison, they observed significant (about 10% more) rectal sparing in terms of reducing rectum volumes receiving intermediate-to-high dose of 55-70 Gy for plans with prostate alone in CTV than those with both prostate and SV in CTV. As a next step, we will extend our analysis to cases of intermediate risk and high risk by including the proximal section of the seminal vesicles into the CTV to characterize the potential dosimetry improvement using ERB.

| CONCLUSION S
In conclusion, significant reductions of dose to the rectum using ERB were observed in the intermediate and high-dose region from a retrospective planning study of CyberKnife prostate SBRT. This may be considered as a valuable technique for clinical implementation to improve the rectal toxicity profiles in prostate SBRT.