A PRACTICAL METHOD TO ACHIEVE PROSTATE GLAND IMMOBILIZATION AND TARGET VERIFICATION FOR DAILY TREATMENT

Transcription

1 PII S (01) Int. J. Radiation Oncology Biol. Phys., Vol. 51, No. 5, pp , 2001 Copyright 2001 Elsevier Science Inc. Printed in the USA. All rights reserved /01/$ see front matter PHYSICS CONTRIBUTION A PRACTICAL METHOD TO ACHIEVE PROSTATE GLAND IMMOBILIZATION AND TARGET VERIFICATION FOR DAILY TREATMENT ANTHONY V. D AMICO, M.D., PH.D.,* JUDI MANOLA, PH.D., MARIAN LOFFREDO, R.N., O.C.N.,* LYNN LOPES, R.N.,* KRISTOPHER NISSEN, B.S., R.T.T.,* DESMOND A. O FARRELL, C.M.D.,* LEAH GORDON, B.A.,* CLARE M. TEMPANY, M.D., AND ROBERT A. CORMACK, PH.D.* Departments of *Radiation Oncology, Biostatistics, and Radiology, Brigham and Women s Hospital and Dana Farber Cancer Institute, Boston, MA Purpose: A practical method to achieve prostate immobilization and daily target localization for external beam radiation treatment is described. Methods and Materials: Ten patients who underwent prostate brachytherapy using permanent radioactive source placement were selected for study. To quantify prostate motion both with and without the presence of a specially designed inflatable intrarectal balloon, the computerized tomography based coordinates of all intraprostatic radioactive sources were compared over 3 consecutive measurements at 1-min intervals. Results: The placement and inflation of the intrarectal balloon were well tolerated by all patients. The mean (range) displacement of the prostate gland when the intrarectal balloon was present vs. absent was 1.3 (0 2.2) mm vs. 1.8 (0 9.1) mm (p 0.03) at 2 min respectively. The maximum displacement in any direction (anterior posterior, superior inferior, or right left) when the intrarectal balloon was inflated vs. absent was reduced to <1 mm from 4 mm. Conclusions: Both prostate gland immobilization and target verification are possible using a specially designed inflatable intrarectal balloon. Using this device, the posterior margin necessary on the lateral fields to ensure dosimetric coverage of the entire prostate gland could be safely reduced to 5 mm and treatment could be set up and verified using a lateral portal image Elsevier Science Inc. Prostate cancer, Immobilization, Radiation therapy, Dose escalation. INTRODUCTION Reprint requests to: Anthony V. D Amico, M.D., Ph.D., Brigham and Women s Hospital, Department of Radiation Therapy, Harvard Medical School, 75 Francis Street, L-2 Level, Boston, MA Tel: (617) ; Fax: (617) ; The major dose-limiting late toxicity following external beam radiation therapy (RT) in the treatment of patients with clinically localized prostate cancer is radiation proctitis (1). This toxicity is related to the total radiation dose prescribed and the volume of the anterior rectal wall receiving a high ( 70 Gy) radiation dose (2). Today a major issue limiting radiation oncologists attempts to reduce the volume of the anterior rectal wall receiving a high radiation dose is the intrinsic motion of the prostate gland of up to 5 mm in the anterior to posterior direction caused by rectal peristalsis (3). Given this motion, radiation oncologists generally will add a margin to the radiation field to ensure that the entire prostate gland receives the prescription dose. This margin is typically on the order of mm (3 5 mm to account for prostate motion, 3 5 mm for patient setup error, and 3 5 mm to allow for the dose to reach 100% of the prescription [i.e., buildup]). Current practitioners of three-dimensional (3D) conformal dose escalation will place a smaller margin on the posterior aspect of the prostate when dose escalating beyond 70 Gy to avoid significant rectal toxicity (4 6). However, this smaller margin also causes a potential risk of not dose escalating in the region of the posterior aspect of the prostate gland (i.e., peripheral zone) where most prostate cancers are located. Realizing that a preliminary report (7) of a randomized dose escalation radiation trial reports a benefit in prostate-specific antigen (PSA) outcome at 5 years for patients predominately with a PSA ng/ml receiving an isocenter dose of 78 Gy as compared to 70 Gy, the potential for long-term improvement in disease-specific survival exists. Therefore, a method permitting prostate immobilization and target localization could allow for the delivery of higher radiation doses to the entire prostate gland while minimizing dose to the rectum. Such a technique could lead to an increase in local control while minimizing radiation-induced proctitis by permitting the use of smaller posterior margins. An inflatable intrarectal balloon specially designed to adamico@jcrt.harvard.edu Received Apr 17, 2001, and in revised form Jul 23, Accepted for publication Sep 4,

2 1432 I. J. Radiation Oncology Biology Physics Volume 51, Number 5, 2001 Fig. 1. (a) Anterior (below) and lateral (above) views of the deflated intrarectal balloon. (b) Anterior (below) and lateral (above) views of the intrarectal balloon inflated with 60 cc of air. The blue line along the anterior surface of the device corresponds to the anterior surface of the balloon designed to conform to the posterior prostate surface. This blue line should be positioned at 12 o clock during insertion of the balloon. conform to the shape of the prostatic rectal interface has been in use for over a decade for imaging the prostate gland using magnetic resonance imaging (MRI). We had this device altered by removing the internal imaging apparatus necessary for MRI but preserving the unique feature that allows it to conform to the prostatic rectal interface when inflated. Selecting patients for this study who had undergone magnetic resonance (MR)-guided prostate brachytherapy permitted the tracking of prostate motion using the intraprostatic sources as fiducial markers in a time-lapse study of computerized tomography (CT). This study was performed both in the presence and absence of the intrarectal balloon to quantify prostate motion in both settings. The intrarectal balloon was also visualized using portal imaging on the treatment unit and CT at the treatment planning session. Axial measurements from the anterior surface of the intrarectal balloon to the sacrum and symphysis pubis were made using both the left lateral portal and simulation image. This set of measurements (simulation vs. portal image) permitted an assessment of the correspondence of the anterior surface of the intrarectal balloon on the portal image and the anterior surface of the intrarectal balloon on the simulation image. METHODS AND MATERIALS Patient selection and the intrarectal balloon Ten consecutive low-risk patients treated with MRguided prostate brachytherapy between 12/00 and 1/01 participated in this study. Patients who had received prior hormonal therapy were excluded. The intrarectal balloon was supplied by Medrad (Indianola, PA) and was constructed by removing the internal imaging apparatus necessary for MR imaging from the preexisting endorectal MR coil. This provided a deflated balloon whose outside diameter was 15 mm, approximating the size of a rectal suppository. The deflated and inflated (60 cc air) device is shown in Figs. 1a and 1b respectively.

3 Method for prostate gland immobilization and target verification A. V. D AMICO et al Fig. 2. (a) Left lateral portal image with the intrarectal balloon inflated with 60 cc of air and inserted to a depth of 12 cm from the anal verge. The axial distance from the anterior sacrum to the anterior aspect of the intrarectal balloon is labeled A. The distance labeled B extends from the anterior aspect of the intrarectal balloon to the posterior aspect of the symphysis pubis. (b) Left lateral digitally reconstructed radiograph with the intrarectal balloon inflated with 60 cc of air and inserted to a depth of 12 cm from the anal verge. The CT-defined prostate and seminal vesicles are shown. Imaging protocol to assess prostate motion After completing postimplant filming on the General Electric (GE) CT simulator (Milwaukee, WI), three sets of CT images (2 mm slice thickness) were obtained at 1-min time intervals, starting above the seminal vesicles and finishing at the level of the penile bulb with the patient in the supine position. The intrarectal balloon was then inserted to a depth of 12 cm. The depth was measured from the tip of the intrarectal balloon to the anal verge. The 12-cm depth ensured that the balloon was juxtaposed to the entire posterior aspect of the prostate. Sixty cc of air were used routinely in each case to maximally expand the rectal surface away from the prostatic rectal interface. With the intrarectal balloon in place and inflated, CT images (2 mm slice thickness) were again obtained at 1-min intervals starting above the seminal vesicles and finishing at the level of the penile bulb for each case. Imaging protocol to assess treatment setup A single patient had a left lateral portal image obtained on the treatment unit and taken using Eastern Kodak Co. ECL film (Rochester, New York) and 15-MV photons (Fig. 2a) to assess the visualization of the intrarectal balloon using portal imaging. This film was compared to the left lateral simulation film shown in Fig. 2b in which the intrarectal balloon is inserted to the same depth (12 cm) and inflated with the same volume of air (60 cc). The intrarectal balloon and the prostate gland were contoured using the GE CT simulator and projected onto the digitally reconstructed radiograph (DRR) (Fig. 2b). The distance from the anterior border of the air visualized on the left lateral portal film and the DRR to the posterior aspect of the symphysis pubis and the anterior border of the sacrum were obtained at 2-mm intervals in the axial plane. Measurement of the thickness of the prostatic rectal interface Using the axial slices obtained at 2.0-mm intervals from the GE CT simulator from the CT-defined prostate base to apex, a measurement was made from the posterior aspect of the prostate gland to the anterior aspect of the intrarectal balloon. This process was repeated for each patient to assess a mean and range of the prostatic rectal interface distances beginning at the prostatic apex and extending to the base. Statistical methods to assess prostate gland motion The coordinates of the radioactive sources were obtained on each CT slice using an automated source-finding algorithm (8) developed for postimplant dosimetry at time 0, 1, and 2 min. The algorithm, developed at our institution by RAC, utilized a threshold analysis. Specifically, clusters of CT voxels whose voxel value was significantly greater than the voxel value for bone were defined as the sources and served as the fiducial markers. Artifacts such as intraprostatic calcium deposits were of no consequence using this approach because there displacement was also calculated as part of the prostate motion calculation. The displacement of the sources was calculated by comparing the source loca-

4 1434 I. J. Radiation Oncology Biology Physics Volume 51, Number 5, 2001 tions in each of the 3 consecutive scans. Specifically, a source in the baseline scan (time 0) was compared with a source location in the two later scans using a nearest neighbor algorithm. This procedure was repeated for each case for patients with and without the intrarectal balloon. It was assumed that the motion of the intraprostatic sources served as a surrogate for the motion of the prostate gland. The error associated with the measurement of the motion of the sources using this approach was calculated by imaging a series of 10 sources 10 times and was found to be approximately 1 mm. A Wilcoxon rank sum test was used to compare the differences in the mean displacement of the intraprostatic radioactive sources with and without the intrarectal balloon for each time interval (0 1 and 0 2 min). Statistical methods to assess treatment setup Using both the left lateral portal image and the left lateral DRR, measurements were made from the anterior surface of the intrarectal air column and both the posterior aspect of the symphysis pubis and the anterior aspect of the sacrum. As displayed in Fig. 2a, measurements were performed at 2-mm intervals in an axial plane. For the case where the sacrum was used as one border, the measurements extended from the superior to inferior aspect of the acetabulum. In the case where the symphysis pubis was used as one border, the measurements extended from the superior to inferior aspect of the symphysis pubis. All measurements were adjusted for the magnification factor used during the acquisition of the simulation and portal image. The method of Bland and Altman (9) was used to define the limits of agreement of the two methods of measurement (portal image, DRR). Spearman rank correlation coefficients were used to describe the strength of the relationship. RESULTS Prostate motion with and without the intrarectal balloon There was a significant difference (p 0.03) in the mean displacement of the prostate gland between 0 and 2 min evaluated for patients with and without the intrarectal balloon. However, this difference was not significant between 0 and 1 min (p 0.39). Specifically, the mean (range) displacement of the prostate gland when the intrarectal balloon was present vs. absent was 1 (0 2) mm vs. 1.3 (0 4.3) mm at 1 min and 1.3 (0 2.2) mm vs. 1.8 (0 9.1) mm at 2 min, respectively. Of particular importance for the purpose of external beam RT field design, the maximum displacement in any direction (anterior posterior, superior inferior, or right left) when the intrarectal balloon was inflated vs. absent was reduced to 1 mm from 4 mm respectively during the 2-min time interval evaluated. Table 1 illustrates the proportion of patients whose maximal displacement of the prostate gland was 1, 2, 3, 4, 5, or 10 mm or less in a given direction (anterior posterior, superior inferior, or right left) stratified by the presence or absence of the intrarectal balloon during the 2-min time interval. Table 1. The proportion of patients whose maximal displacement of the center of the prostate gland was 1, 2, 3, 4, 5, or 10 mm or less stratified by the presence or absence of the intrarectal balloon during the 2-min time interval Overall maximal displacement* Assessment of treatment setup Using both the left lateral portal image and the DRR, measurements from the posterior aspect of the symphysis pubis and the anterior aspect of the sacrum to the anterior aspect of the intrarectal balloon were made in an axial plane. These measurements were acquired at 2-mm intervals on both the DRR and portal image from the superior to inferior aspect of the symphysis pubis and adjusted for any magnification factor. The pairs of measurements obtained from the DRR and portal image are listed in Table 2 and were compared using the methodology of Bland and Altman (9). Using this metric, the maximal difference in measurements in the anterior posterior dimension from all axial levels examined was less than 0.8 mm (Spearman s correlation coefficient 0.97 and p ) consistent with the maximum prostate motion in the anterior posterior direc- Vertical distance from the superior aspect of the symphysis pubis or acetabulum With intrarectal balloon Table 2. Measurements in the axial plane using DRR and port films Anterior sacrum3anterior rectal balloon Without intrarectal Balloon 1 0% 0% 2 90% 0% 3 100% 10% 4 100% 50% 5 100% 80% % 100% * The maximum displacement in any direction (anterior posterior, superior inferior, or right left) when the intrarectal balloon was inflated versus absent was reduced to 1 mm from 4 mm respectively during the 2-min time interval evaluated. Anterior rectal balloon3posterior symphysis pubis DRR Port DRR Port

5 Method for prostate gland immobilization and target verification A. V. D AMICO et al tion noted when the intrarectal balloon is inflated. Therefore, the anterior border of the intrarectal balloon as visualized on the left lateral portal image can identify the mucosal surface of the anterior rectal wall to within the limits of our measurement, which was approximately 1 2 mm. Measurement of the thickness of the prostatic rectal interface The mean (range) thickness of the prostatic rectal interface for all 10 study patients was 4.1 ( ) mm. The measurement of the prostatic rectal interface was consistent for each of the 10 study patients when the intrarectal balloon was inflated or absent. Therefore, there did not appear to be any measurable compression of the anterior rectal wall as a result of the insertion and inflation of the intrarectal balloon. In addition, there was little variation ( mm) in the thickness of the prostatic rectal interface when obtaining measurements at the prostatic apex, mid-gland, or base for a given patient. DISCUSSION A recent study of the gastrointestinal (GI) complications resulting from 3D conformal external beam radiation therapy delivered as part of a randomized dose escalation trial (isocenter dose of 78 vs. 70 Gy) has been reported (2). The investigators found that the Grade 2 or higher GI complications were 37% vs. 13% (p 0.05) when the dose volume histogram (DVH) analysis revealed that 25% vs. 25% of the rectum respectively received a dose of 70 Gy or greater. Moreover, all Grade 3 GI complications were observed only for patients receiving 70 Gy or higher to 30% of the rectum. Therefore rectal toxicity was clearly associated with the rectal volume receiving doses in excess of 70 Gy. In that dose escalation study a 4-field box technique was used for the initial 46 Gy and a 6-field cone down employed for the remaining dose. The margin from the edge of the contoured target to the block edge ranged from 1.25 to 1.50 cm anterior and inferior and from 0.75 to 1.0 cm posterior and superior. Knowing that radiation-induced rectal toxicity is associated with the rectal volume exceeding 70 Gy, posterior margins of 5 mm or less have been attempted when performing dose escalation (4 6) to minimize the risk of injury to the rectal wall. While the conservative posterior margin has been shown to permit dose escalation without marked increases in rectal toxicity (4 6), known prostate motion in the anterior posterior direction due to rectal peristalsis can lead to less than full dose delivery when posterior margins are reduced. A single randomized dose escalation trial (7) has been performed to date; in that study, the posterior margin utilized ensured that despite prostate motion the entire prostate was treated daily. That study showed both improved PSA control and a trend toward improved distant disease-free survival at 5 years. Therefore, it may be important to ensure full prostate gland coverage in order to have the potential to improve cancer control when dose escalating in the management of prostate cancer. However, to limit the volume of rectum receiving 70 Gy or higher while simultaneously escalating the RT dose to the entire prostate gland, both prostate gland immobilization and localization are necessary during the dose escalation portion of the therapy. Therefore, in this study a practical method to immobilize and localize the prostate gland was investigated with the eventual goal of being able to exploit the potential for improved local control using dose escalation while maintaining rectal integrity. The results found that using the intrarectal balloon, the maximal prostate gland motion in the anterior posterior direction decreased from 4 mm to 1 mm measured during the time interval necessary to deliver two lateral radiation treatment fields. In addition, localization of the mucosal surface of the anterior rectal wall was shown to be within the maximal anterior posterior displacement of the prostate gland ( 1.0 mm) when the intrarectal balloon was inflated. Finally, the prostatic rectal interface was found to be consistent within a given patient from the prostatic apex to base but ranged from 3 to 5 mm between patients. Therefore using an online portal image, the anterior surface of intrarectal balloon could define the anterior rectal wall to within 1 mm. This would ensure that a posterior field margin set 1 mm posterior to the anterior surface of the balloon would deliver full dose to the entire prostate gland because 3 to 5 mm of tissue exist in the prostatic rectal interface to permit dosimetric buildup. An important potential application of this methodology is in the case where androgen suppression therapy (AST) is given concurrently with external beam RT. It is known that during the course of RT the prostate gland will decrease in volume under the effect of the AST, bringing more rectal mucosa into the high-dose RT volume and thereby increasing the risk of rectal toxicity. In fact, Schultheiss and colleagues (10) have shown that the use of neoadjuvant and concurrent AST and RT was a significant predictor of rectal bleeding on multivariable analysis after accounting for RT dose. Therefore, having a method that could adjust the posterior border of the lateral treatment fields to the decreasing prostate gland volume during the course of RT would help to reduce rectal volume being treated to above 70 Gy while still ensuring dosimetric coverage of the entire prostate gland. Moreover, encouraging data from the Radiation Therapy Oncology Group (RTOG) 9406 dose escalation study (11) reported only a 1% rate of Grade 3 late rectal toxicity using 3D conformal technique using a minimum prostate dose of 73.8 Gy with 5-mm margins posteriorly and no prostate immobilization device. Those data suggest that with prostate immobilization and the use of a 5-mm posterior margin, rates of Grade 3 late rectal toxicity should be low at doses up to 73.8 Gy. Other methods of prostate localization exist and have been reported. In particular, daily prostate localization using transabdominal ultrasound localization has revealed that prostate motion occurs and is significant. Lattanzi and col-

6 1436 I. J. Radiation Oncology Biology Physics Volume 51, Number 5, 2001 leagues (12) have reported that the isocenter field misalignment between the baseline CT and ultrasound performed during the final cone-down treatments was significant. Specifically 51% of the anterior to posterior misalignments were 5 mm and 21% 10 mm. They concluded that daily clinical isocenter misalignments may be greater than previously reported (3), limiting our ability to optimize 3D treatment planning and delivery because of the need to expand the width of the fields to ensure prostate gland coverage. Although the daily transabdominal ultrasound localization system can account for motion that occurs on a day-to-day basis, it does not account for motion that may occur during the delivery of RT. This study would suggest that motion on the order of 5 mm is uncommon during the course of an RT treatment but can be seen in up to 20% of the cases. Therefore, a need to localize and immobilize the prostate gland will be necessary if the goal is to ensure that the entire prostate gland receives the prescribed dose each day. Others (13, 14) have begun to explore using an intrarectal balloon for prostate immobilization during the delivery of dose escalated radiation therapy (mean dose to the prostate 75.8 Gy) and have reported minimal acute toxicity using the RTOG grading scale (11% Grade 1, 6% Grade 2, 0% Grade 3 or higher). In that study (13), 45 of the 100 patients received AST before, during, and/or after RT. As a result, during the course of RT both the prostate gland volume and position may have changed under the influence of the AST. Despite this changing prostate volume, acute toxicity was low, possibly due to the ability to adjust the field daily using the intrarectal balloon as a landmark for the anterior rectal wall. Further follow-up will determine whether these low rates of acute toxicity will translate into low rates of late rectal bleeding. In conclusion, both prostate gland immobilization and localization are possible using a specially designed inflatable intrarectal balloon. Such a device permits the use of smaller posterior margins while still ensuring that the entire prostate gland receives the prescription dose. This method may also allow for adjustments in the lateral field border for patients whose prostate gland volume is decreasing under the effect of hormonal therapy. A prospective trial is now under way to quantify the GI and genitourinary toxicity using both physician-assessed (15) and anonymous patient reported (16) instruments for patients undergoing 3D conformal RT to a dose of 75.6 Gy to the clinical target volume with AST using the intrarectal balloon daily for treatment setup during the cone-down portion of the therapy. REFERENCES 1. Teshima T, Hanks GE, Hanlon AL, et al. Rectal bleeding after conformal 3D treatment of prostate cancer: Time to occurrence, response to treatment and duration of the morbidity. Int J Radiat Oncol Biol Phys 1997;39: Storey MR, Pollack A, Zagars G, et al. Complications from radiotherapy dose escalation in prostate cancer: Preliminary results of a randomized trial. Int J Radiat Oncol Biol Phys 2000;48: Beard CJ, Kijewski P, Bussiere M, et al. Analysis of prostate and seminal vesicle motion: Implications for treatment planning. Int J Radiat Oncol Biol Phys 1996;34: Hanks GE, Hanlon AL, Schultheiss TE, et al. Dose escalation with 3D conformal treatment: Five-year outcomes, treatment optimization, and future directions. Int J Radiat Oncol Biol Phys 1998;41: Zelefsky MJ, Leibel SA, Gaudin PB, et al. Dose escalation with three-dimensional conformal radiation therapy affects outcome in prostate cancer. Int J Radiat Oncol Biol Phys 1998;41: Michalski JM, Purdy JA, Winter K, et al. Preliminary report of toxicity following 3D conformal radiation therapy in prostate cancer on 3DOG/RTOG Int J Radiat Oncol Biol Phys 2000;46: Pollack A, Zagars GK, Smith LG, et al. Preliminary results of randomized dose escalation study comparing 70 Gy to 78 Gy for the treatment of prostate cancer. J Clin Oncol 2000;18: Cormack RA, Tempany CM, D Amico AV. Optimizing target coverage by dosimetric feedback during prostate brachytherapy. Int J Radiat Oncol Biol Phys 2000;48: Bland JM, Altmn DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;8: Schultheiss TE, Lee WR, Hunt MA, et al. Late GI and GU complications in the treatment of prostate cancer. Int J Radiat Oncol Biol Phys 1997;37: Michalski JM, Winter K, Purdy JA, Wilder R, Perez CA, Roach M, Parliament M, Pollack A, Markoe A, Harms WB, Sandler H, Cox JD. Update of toxicity following 3D radiation therapy for prostate cancer on RTOG Int J Radiat Oncol Biol Phys 2000;48: Lattanzi J, McNeeley S, Donnelly S, et al. Ultrasound-based stereotactic guidance in prostate cancer: Quantification of organ motion, and setup errors in external beam radiation therapy. Comput Aided Surg 2000;5: Teh BS, Mai W, Uhl BM, et al. Intensity-modulated radiation therapy (IMRT) for prostate cancer with the use of a rectal balloon for prostate immobilization: Acute toxicity and dose volume analysis. Int J Radiat Oncol Biol Phys 2001;49: McGary JE, Grant WH III. A clinical evaluation of set-up errors for a prostate immobilization system. J Appl Clin Med Phys 2000;1: Hanlon AL, Schultheiss TE, Hunt MA, et al. Chronic rectal toxicity after high-dose conformal treatment of prostate cancer warrants modification of existing morbidity scales. Int J Radiat Oncol Biol Phys 1997;38: Talcott JA, Reiker P, Propert KJ, et al. Patient reported impotence and incontinence after nerve-sparing radical prostatectomy. J Natl Cancer Inst 1997;89: