The effect of prone and supine treatment positions for the pre-operative treatment of rectal cancer on organ-at-risk sparing and setup reproducibility using volumetric modulated arc therapy

Globally and every year, over 1.4 million people are diagnosed with colorectal cancer [1, 2]. In 2015, it was the fourth most common malignancy (49.9 cases per 100,000) in Canada [3].

Preoperative radiotherapy or chemoradiation followed by total mesorectal excision (TME) is currently the standard of care for patients with locally advanced rectal cancer [4, 5]. The small bowel is the most relevant organ-at-risk (OAR) nearby typical rectal cancer target volumes. A standard prescription for preoperative radiotherapy is 50.4 Gy in 28 fractions, which may be high enough to elicit small bowel complications [6]. In an effort to avoid small bowel toxicities, many centers treat patients in the prone position propped up on belly boards where the abdomen falls through a hole in the board to allow the small bowel to drop anteriorly away from the target volume [7].

Volumetric modulated arc therapy (VMAT) is increasingly used to deliver radiation treatment for a variety of sites, including rectal cancer [8]. VMAT allows for rapid treatment delivery and highly conformal dose distributions compared to 3D conformal radiation therapy (3DCRT) or intensity modulated radiation therapy (IMRT). Previously, our center treated rectal patients using a 3-beam 3DCRT technique. Recently, we adopted VMAT for rectal cancer treatment, employing a ~180° posterior arc delivering conformal dose distribution whilst avoiding beam entry anteriorly through the small bowel and bladder. Treating with VMAT has necessitated the use of cone-beam CT (CBCT) in-room image guidance for patient setup verification to improve accuracy of dose delivery.

The prone orientation for rectal cancer patients is known to cause patient discomfort, especially for patients with a stoma. The combination of a belly board with the prone position is also known for setup errors [9]. Thus the supine treatment position is an attractive alternative. With the implementation of VMAT and CBCT, we assessed if the supine position can be safely used to treat rectal cancer. Our objectives were i) to compare small bowel and bladder doses and ii) to compare the setup reproducibility using CBCT for the prone and supine orientations.

To date, this is the first study to report dosimetric data for OARs and setup reproducibility for prone and supine orientations in the context of VMAT treatment for rectal cancer. This study was driven by our clinical need to determine if treating patients in the supine position is preferable to the prone position in light of our adoption of VMAT and CBCT-mediated image-guided RT for rectal cancer.

Use of the belly board and the prone position for rectal cancer was born of the need to reduce small bowel toxicity. This was prominent in the era of 3DCRT when prone positioning on a belly board and/or distending the bladder were used to move the small bowel away from the treatment volume [10]. The aim of reducing bowel toxicity comes with a trade-off of reduced patient reproducibility and stability. A previous study demonstrated that using a belly board increased anterio-posterior reproducibility displacements compared with no belly board in the prone position [9]. Positional changes would be expected when considering the smooth, curved structure of the belly board and differences in patient habitus. Our radiation therapists had reported that patients’ abdomens fall variably into the belly board cavity at every fraction, thus, we anticipated that the pitch direction would be more variable for the prone orientation. Our CBCT data confirms that a belly board introduces set-up errors in the pitch and roll rotational directions. We did not observe any differences in the yaw direction, as this is arguably the direction the therapists have the most control over with the aid of a superior straightening tattoo.

There are safety considerations that would warrant moving to the supine position. Figure 1c shows how high a patient needs to vault in order to mount a belly board, a potentially hazardous situation for elderly or frail patients. During the course of this work, there was a non-study patient that fell off one of the prone belly boards, triggering an incident investigation at our institution. The supine setup is more comfortable for patients and convenient for therapists on the treatment units [11].

Various studies have compared prone and supine orientations for rectal cancer treatment and reported the dosimetric impact on the small bowel. Dryzmala et al. have simulated rectal cancer patients in both the prone and supine positions and created clinical-grade 3DCRT plans for both orientations [12]. Nijkamp et al. tested two prone belly boards and the supine position on volunteers scanned in an MRI and planned treatments using 7-beam IMRT [13]. The treatment modalities discussed above are likely not optimal for avoiding small bowel dose, where three-field techniques are limited in their ability to sculpt dose away from anterior structures, and isotropically placed IMRT beams (e.g. 7-field techniques) place several entry beams through the small bowel on the way to the rectal target. The results from both studies broadly agree with each other in that the belly board significantly reduced the volume of irradiated small bowel in the low-intermediate dose range (Dryzmala: 5–15 Gy; Nijkamp: <35 Gy), while no significant differences were seen in the high dose range (Dryzmala: 20–45 Gy; Nijkamp: 50 Gy). Additional data from Froseth et al. demonstrated that at dose levels from 5 to 45 Gy (tested at 5 Gy increments) there was no statistically significant difference between irradiated volumes when comparing a prone setup study arm (N = 40) with a supine study arm (N = 43) using a 3DCRT planning strategy [14]. Our data also showed a similar trend to the Dryzmala and Nijkamp studies with a smaller volume of irradiated small bowel in the low dose range with prone versus supine patients. At the high dose range (V45 and V50.4) the volume of irradiated small bowel was lower in supine patients. It should be noted that none of our OAR dosimetric data showed statistically significant differences between the prone and supine positions. One possible reason for the lack of differences is the use of a posteriorly-applied VMAT half arc. Highly conformal dose distributions are possible (Fig. 1) with nearly none of the beam angles entering through the small bowel and bladder on the way to the target. Joye et al. produced a study using VMAT full arcs comparing the full bowel, small bowel, and bowel bag DVH metrics, and found statistically significant differences in the irradiated volumes for V15 and lower doses [15]; this suggests that our technique of using a VMAT posterior half arc may have an additional bowel sparing effect compared with a full arc geometry.

Acute small bowel toxicity is a concern for rectal cancer patients receiving preoperative radiotherapy. Baglan et al. (2002) demonstrated a strong relationship between the volume of small bowel receiving at least 15 Gy and the risk of toxicity [16]. Grade 0–1 toxicity was seen after irradiation with a V15 of less than 150 cm³ (grade 0–1 in 90%, grade 2 in 10% of the patients), and 70% of the patients with a V15 of more than 300 cm³ experienced grade 3 toxicity. Roeske et al. (2003) demonstrated an increased risk of acute grade 3 toxicity with a V45 greater than 200cm³ [17]. VMAT is an advanced method to spare OARs for any treatment site, particularly compared to 3DCRT techniques. The general consensus is that there are dosimetric benefits to VMAT with toxicity levels that are lower than for 3DCRT [18, 19]. In our study, the small bowel V15 was greater than 300cm³ for some patients treated in the prone and supine positions, while V45 was consistently less than 200cm³. Review of the patient treatment records did not reveal any grade 3 or greater acute GI toxicity in our study. However, a clear limitation of our study is the lack of prospective toxicity assessment. Nonetheless, based on the low volume of irradiated small bowel we expected minimal toxicity. Furthermore, a study from the University Medical Center at Göttingen, Germany demonstrated that VMAT substantially reduced high-grade acute toxicity compared with 3DCRT (5% versus 20% incidence), as well as late toxicity (6% versus 22% incidence) [8].

In addition to the prone orientation being less reproducible than supine, the belly board may unintentionally push the bowel into the treatment volume. In consideration of this, it is a reasonable supposition that the delivered dose distribution is markedly different than the plan dosimetry for many prone patients due to inaccuracies in the patient setup. The rotational setup error can also introduce considerable targeting error. A rudimentary calculation can demonstrate this effect. Several CBCTs showed > ± 4°rotational error in the pitch direction (Fig. 3). Rectal cancer PTV volumes can be 20 cm long in the superior-inferior direction. With the beam isocenter in the middle of the PTV, and a 4° error in pitch, this can result in the superior and inferior ends of the PTV being displaced by 7 mm, which would be deleterious for target coverage and (if rotated in a disadvantageous direction) small bowel sparing.

The results drawn from this study were based on rectal cancer patients undergoing preoperative radiotherapy and may not be applicable to rectal cancer patients undergoing postoperative adjuvant radiotherapy.

In conclusion, we found non-significant dose data differences in the bladder and small bowel between the prone and supine study arms, and greater rotational set-up errors with the prone position. As such, our results support treating all preoperative rectal cancer patients in the supine position. The improved conformality of VMAT aided by the superior setup imaging of CBCT further enables treatment in the supine position, which is more comfortable, reproducible, and safer.