Effectiveness of Rectal Balloon as an Internal Immobilization Device for Proton Therapy of Prostate Cancer

Reviewer: Christine Hill, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: May 24, 2008

Presenter: Andrew Lee, M.D., M.P.H.
Presenter's Affiliation: Division of Radiation Oncology, The University of Texas M.D. Anderson Cancer Center
Type of Session: Scientific


  • Radiotherapy for prostate cancer depends largely on the technical ability to deliver adequate dose to the prostate without severe bladder and rectal toxicity. The fraction of rectum irradiated has previously been demonstrated to impact gastrointestinal and genitourinary acute and late effects (Kuban, 2008).
  • Delivery of dose to the prostate with minimization of dose to the bladder and rectum is impacted by prostate motion, as well as immobilization of the prostate which may allow smaller volume expansion margins to be employed. 
  • When passively scattered proton beams are used to treat prostate cancer, daily alignment of bony structures is generally performed in order to align the patient with the beam and compensator. Minimization of prostate motion allows use of small margins; however, accurate localization is imperative in avoidance of inadequate dosing to the prostate volume itself when small margins are used.
  • A water-filled, endorectal balloon may be employed as an internal prostate immobilization device in efforts to minimize intra- and interfraction motion. The balloon may also serve to displace the posterior rectum away from the treatment field, and to decrease air within the rectum.
  • This study was undertaken in order to evaluate the reproducibility of anatomic position with use of the endorectal balloon, as well as the uncertainties in prostate position and shape relative to bony anatomy.

Materials and Methods

  • Each patient analyzed underwent two or three separate CT scans:
    • On the day of CT simulation, each patient underwent enema treatment. Thereafter, the Medrad endorectal balloon was placed and filled with 65 milliliters of water. Patients were positioned supine with a knee/ foot cradle for pelvic immobilization. A wedge cut-out was present at midline of the immobilization device to allow for positioning of the balloon. CT scan was then performed.
    • Following this, after removal of the endorectal balloon, patients were permitted to move off of the table. After several minutes, the balloon was replaced and a second CT scan was performed.
    • A subset of 35 patients received a third CT scan that was performed midway through radiotherapy, again with placement of the rectal balloon (no enema was performed).
    • The endorectal balloon was placed with a similar technique during daily treatments, without enema pre-treatment.
  • Of the scans performed for each patient, one was identified by the physician as the planning CT scan. The planning CT scan was registered to each repeat CT scan using CT-assisted targeting (CAT), an in-house developed three-dimensional image registration method.
    • Alignments were performed twice – once with the prostate as the alignment target, and once with the pelvic bones as the target.
    • The difference between the prostate position and the pelvic bone position relative to the planning CT was defined as the net prostate motion relative to bony anatomy.
  • The overlap between the shape of the prostate on the planning CT scan and its shape on the repeat CT scan(s) was calculated as the volume overlap index (VOI).
    • A VOI of one represented perfect agreement between the two contoured volumes, while VOI of zero indicated no spatial overlap. 


  • Positional variations:
    • Positional variations of the prostate relative to bony anatomy between the two CT simulations performed on the same day were small, with mean variations being 1.1 mm [range -1.0 – 3.0 mm, one standard deviation (1SD) 1.1mm] in the anterior-posterior position, -0.8 mm (range -3.4 – 2.0 mm, 1SD 1.3 mm) in the superior-inferior positions, and -0.1 mm (range -0.1 – 1.0 mm, 1SD 0.5 mm) in the right-left position.
    • Positional variations of the prostate were greater in repeat CT scans performed during treatment, with mean variations being 0.1 mm (range -3.9 – 3.9 mm, 1SD 1.9 mm) in the anterior-posterior position, -0.7 mm (range -4.2 – 5.0 mm, 1SD 1.9 mm) in the superior-inferior positions, and -0.1 mm (range -2.9 – 1.0 mm, 1SD 0.7 mm) in the right-left position.
    • The largest variation in both comparisons was in the superior-inferior positions.
  • Prostate shape deformation:
    • The mean VOI for the prostate was 0.93 (range 0.72 – 0.98).
    • The mean VOI for the rectum was 0.87 (range 0.51 – 0.98).
    • The prostate-rectum interface volume was noted to agree well even in situations of low VOI for the rectum.

Author's Conclusions

  • The authors conclude that only small prostate positional variations and deformations occur with use of rectal balloon.
  • The small amount of positional variability, as well as the stable prostate-rectum interface identified within this study supports the use of relatively small margins, particularly with regards to anterior-posterior position. Smaller posterior margins would be expected to decrease rectal toxicity.
  • The authors note that newer balloons may offer greater patient comfort and improved prostate immobilization, and should be investigated in the future.

Clinical/Scientific Implications

  • Regardless of radiotherapy technique employed, knowledge of prostate position on a fraction by fraction basis is essential to appropriate dose delivery. At centers around the world, prostate localization and/ or immobilization are performed in a variety of ways.
  • The position of the prostate is dependent, in part, on degrees of rectal distention and bladder filling. Prior data have demonstrated that degree of rectal distention at the time of CT simulation may impact biochemical-relapse free survival (de Crevoisier, 2005). Additionally, other groups have shown that the fraction of rectum receiving radiotherapeutic dose may directly affect the acute and late rectal toxicities encountered (Kuban, 2008). These data, as well as our knowledge of pelvic anatomy, support the use of prostate immobilization with a rectal device, such as the endorectal balloon described in this presentation.
  • The data presented here demonstrate that prostate position is relatively stable with use of an endorectal balloon, and that prostate distortion may be limited by the presence of this device. Even so, it should be noted that the range of prostate motion in the superior-inferior position ranged from -4.2 to 5.0 mm on the CT scans performed during radiotherapy. These scans should be representative of anatomic variation on a daily basis during treatment (scans during radiotherapy were performed without enema, and with the endorectal balloon placed in the same manner with which it was placed each day). This degree of motion is certainly not irrelevant. The volume expansions currently utilized by the presenting authors were not discussed during the presentation; however, this degree of prostate motion should be taken into account during treatment planning for each patient.
  • Given the precise and conformal nature of proton beam irradiation, decreased prostate movement and distortion would be expected to decrease volume expansion margins, and to potentially decrease toxicity to the rectum and bladder while allowing adequate dosing to the prostate itself.
  • Having said this, the technical difficulty of placing an endorectal balloon on a daily basis, as well as the patient discomfort associated with this procedure, should be weighed against the benefit of the use of this device.
  • Several other methods are available for daily prostate localization, including cone-beam CT scan and Zmed ultrasonography. Additionally, the four dimensional localization system, Calypso, employed by some centers may be used to direct planned radiotherapy according to prostate position on a daily basis. Each of these systems is, of course, associated with its own risks and benefits.
  • As the use of proton beam radiotherapy becomes widespread, many of the debates which have existed in prostate cancer treatment for many years with regards to photon radiotherapy will certainly continue. Among these is discussion of the optimal method(s) for prostate immobilization and/ or localization. This topic may in fact gain importance for patients being treated with proton radiotherapy due to the increased conformality as well as the potential to “overshoot” the Bragg peak, potentially delivering increased dose to normal structures, should anatomic variation be unknown or uncertain when proton beam radiotherapy is employed.
  • These authors demonstrate interesting data regarding use of an endorectal balloon for prostate immobilization, and this technique certainly has potential to benefit patients undergoing proton beam irradiation. Comparison with other techniques for prostate localization/ immobilization, as well as long-term clinical data, will certainly further elucidate the benefits of use of endorectal ballon devices in the future.