The British Journal of Radiology, 82 (2009),

Transcription

1 The British Journal of Radiology, 82 (2009), A comparison of prone three-dimensional conformal radiotherapy with supine intensity-modulated radiotherapy for prostate cancer: which technique is more effective for rectal sparing? 1,2 T KATO, MS, 3 Y OBATA, MD, 1,2 N KADOYA, MS and 4 N FUWA, MD 1 Department of Radiological and Medical Sciences, Nagoya University Graduate School of Medicine, Nagoya, 2 Department of Medical Physics, Southern Tohoku Research Institute for Neuroscience, Koriyama, 3 School of Health Sciences, Nagoya University, Nagoya and 4 Department of Radiation Oncology, Southern Tohoku Research Institute for Neuroscience, Koriyama, Japan ABSTRACT. The purpose of this study was to assess the potential dose reductions to the rectum with three-dimensional conformal radiotherapy in the prone position (prone 3D-CRT) compared with intensity-modulated radiotherapy in the supine position (supine IMRT) for prostate cancer. 17 prostate cancer patients underwent treatment planning CT scans in the supine and prone positions. Prone 3D-CRT and supine IMRT plans were constructed for each patient and compared in terms of the volume of rectum exposed to the V90 (volume of rectum receiving at least 90% of the prescription dose) as the high dose region. It was confirmed that supine IMRT was significantly superior to prone 3D-CRT (p50.023). Although, in some cases, the distance between the seminal vesicles and the rectum could change by more than 20 mm in the transition from supine to prone, the change in distance was,5 mm in many other cases. While prone 3D-CRT resulted in significant improvements in some patients in terms of rectal sparing, the degree of the effect may be dependent on a patient s anatomy and physical condition in prone 3D-CRT compared with supine IMRT. If the cases in which prone 3D-CRT was more effective in rectal dose reduction could be extracted using some anatomical predictor before treatment planning, prone 3D-CRT may be appropriate in such a case. We consider that prone 3D-CRT still warrants further investigation because of its advantages in terms of simplicity, cost-effectiveness and labour saving; continued research to find an appropriate anatomical predictor is required. Received 11 June 2008 Revised 31 August 2008 Accepted 13 October 2008 DOI: /bjr/ The British Institute of Radiology Advanced radiotherapy techniques available today, such as intensity-modulated radiotherapy (IMRT) and image-guided radiotherapy (IGRT), are remarkably effective [1, 2]. These technologies allow considerably higher doses of radiation to be delivered to the target with relatively lower doses to the organs at risk (OARs) and surrounding normal tissues. However, delivering adequate doses of radiotherapy to targets positioned near OARs remains a challenging task because the tolerance dose of OARs is an important dose-limiting factor. Often, the target is not well separated physically from OARs. Theoretically, the best way to deliver adequate doses to the target with relatively lower doses to the OARs is to increase the distance between the target and the OARs. If this can be achieved, it is comparatively easy to deliver higher doses to the target without the need for the expensive equipment required by IMRT and IGRT, leading to better treatment outcomes. Both invasive and non-invasive methods of increasing the distance between the target and OARs have been Address correspondence to: Takahiro Kato, Department of Radiological and Medical Sciences, Nagoya University Graduate School of Medicine, Daikominami, Higashiku, Nagoya , Japan. takahiro.kato@mt.strins.or.jp proposed. Invasive methods that have been assessed include intraoperative radiotherapy, ovarian transposition to conserve ovarian function during pelvic radiotherapy, and inserting a pelvic tissue expander to displace the small bowel from the radiation portals [3, 4]. However, these invasive techniques can be applied in only a very small number of patients. A widely tested non-invasive technique is positioning the patient during treatment such that internal organs are not in the path of the radiotherapy dose. This method is particularly useful in radiotherapy of the thoracic and abdominal regions because the organs in these regions are sufficiently mobile that their position and shape can be changed simply by changing body position. In general, patients prefer the supine position during external beam radiotherapy because of its greater comfort and stability, but assuming a prone position is the technique that has been assessed most frequently as a means of increasing the distance between the target and OARs. The efficacy of the prone position has been assessed for the treatment of various diseases, such as oesophageal, breast, prostate and gynaecological cancers amongst others [5 9]. In the case of external beam radiotherapy for prostate cancer, the rectum appears to be the most clinically relevant and dose-limiting organ, and the 654 The British Journal of Radiology, August 2009

2 Prone 3-D CRT vs supine IMRT for prostate cancer probability of late rectal complications has been reported to increase when larger volumes are irradiated to target dose levels [10, 11]. Some investigators have suggested a dosimetric advantage in having the patient in the prone position because of the increased distance between the prostate gland and the anterior rectal wall [7, 8]. McLaughlin et al [8] evaluated the relative impact of patient position and beam arrangement on the dose received by normal tissues in three-dimensional conformal radiotherapy (3-D CRT) of prostate cancer. They found that the prone position was consistently superior to the supine position, independent of beam arrangement, when the rectal dose was evaluated. They concluded that patient position has a greater impact on rectal sparing than beam arrangement. However, as the adopted irradiation techniques in their study were limited to conventional 3-D CRT, it is unclear whether similar results can be obtained with IMRT. In general, most facilities use the supine position during external beam radiotherapy for prostate cancer. If 3-D CRT in the prone position (prone 3-D CRT) is superior to IMRT in the supine position (supine IMRT) in terms of rectal sparing, it is advantageous from the point of view of cost-effectiveness and labour saving because IMRT requires special equipment, and pre-treatment verification of IMRT is known to be very time-consuming [12]. In this study, we quantitatively assessed the potential dose reductions to the rectum with prone 3-D CRT and compared this with supine IMRT for prostate cancer. Methods and materials Patients and structure delineation 17 consecutive patients with clinically localised prostate cancer (stage T1c T3N0M0) underwent treatment planning CT scans in the supine and prone positions. Patients were scanned using an Asteion CT scanner (Toshiba Medical Systems, Tokyo, Japan) at 3 mm intervals from 10 mm inferior to the level of the ischial tuberosities to the apex of the bladder or to the bottom of the sacroiliac joint, whichever was greater in both treatment positions. As a patient immobilisation device, VacLok (MedTec, Orange City, IA) was used to immobilise the patients legs in both treatment positions. Patients legs were slightly bent in the supine position but were extended in the prone position. Patients were asked to void their rectum and bladder to minimise daily variations in the shape and anatomical location of the prostate gland before the treatment planning CT scanning performed. The supine scan was performed initially with the prone scan following immediately. The CT scans were transferred to the Xio radiotherapy treatment planning system (Computerized Medical Systems, St Louis, MO), where the clinical target volume (CTV) and OARs, such as the rectum, bladder and femoral heads, were delineated by a single physician to minimise potential bias associated with multiple planners when comparing treatment plans. The CTV consisted of the prostate gland plus the seminal vesicles. The planning target volume (PTV) included the CTV plus an 8 mm safety margin, except at the prostate gland rectum interface, where a 5 mm margin was used to decrease the risk of rectal toxicity. Assuming that the degree of geometric uncertainty for both treatment positions was the same as reported by Stroom et al [13], identical PTV margin sizes were adopted in each position. The external wall of the rectum was contoured. The craniocaudal rectal extension was defined as the first CT slice above the anal verge (caudal border), and the cranial limit was defined as the first slice below the sigmoid flexure. The bladder was contoured in its entirety. Treatment planning The 3-D CRT and IMRT plans were generated by the Xio TPS using the beam data of Clinac 21EX (Varian, Palo Alto, CA). Three treatment plans (supine 3-D CRT, prone 3-D CRT and supine IMRT) were generated for each patient. A total dose of 76 Gy in 2 Gy daily fractions was prescribed for 95% of the PTV volume with 10 MV X-rays in all plans. For 3-D CRT plans, we adopted a six-field coplanar technique described by Zelefsky et al [7]. This technique consists of lateral opposed fields and four oblique fields at 35 from the coronal plane; the plan was to deliver 50% of the dose laterally and 50% from the four equally weighted oblique fields. This technique has been used often in the USA and spares not only the rectum but also the bladder [14, 15]. For the IMRT plans, a seven-field coplanar with segmental multileaf collimation delivery technique was adopted. Plans were optimised using an inverse planning module, which uses a conjugate gradient optimisation algorithm that permits real-time modification of the optimisation parameters and encourages user interactivity to minimise the overall optimisation time. The superposition algorithm was adopted as the computational algorithm. A minimum beamlet size of cm was used with 20 intensity levels throughout. Our plan acceptance criteria for prostate cancer IMRT were as follows. A minimum of 95% of the PTV must receive 100% of the prescription dose (76 Gy). Rectal constraints must be met such that no more than 20% of the rectum may receive.65 Gy and no more than 40% may receive.40 Gy. The 50% isodose line needed to fall within the rectal contour as much as possible on any individual CT slice and the 90% isodose line was not to exceed half of the diameter of the rectal contour on any slice. For the bladder, no more than 25% of the bladder was to receive.65 Gy and no more than 50% of the bladder was to receive.40 Gy. However, it should be noted that in many cases, these bladder constraints were not possible because the bladder was voided prior to the simulation CT. Analysis Because it is still controversial whether prone 3-D CRT is significantly superior to supine 3-D CRT in terms of rectal sparing, we initially compared the rectal dose received by each patient using both techniques. Rectal V90 (volume of normal tissue structure receiving at least 90% of the prescription dose) was compared with the high-dose area of irradiation for both treatment plans using dose volume histograms (DVH). Next, we compared prone 3-D CRT and supine IMRT in the same manner. The bladder was also evaluated as one of the The British Journal of Radiology, August

3 T Kato, Y Obata, N Kadoya and N Fuwa OARs in external beam radiotherapy for prostate cancer. As comparative indices, not only V90 but also mean dose (D mean ), D 05, D 10, D 20 and D 30 (D n ; dose received by n% of the organ volume) for the rectum, the maximum dose (D max ) and the homogeneity index (HI) for the PTV were calculated for each plan. HI is defined as the ratio of D max to D min for the PTV and represented the dose uniformity in the PTV. The statistical difference was determined by a two-sided paired t-test. Differences with p,0.05 were considered significant. Results The accuracy of structural delineation directly influenced the results of this comparative study, so confirmation is very important. No significant differences in volumes were noticed except for the bladder (Table. 1). A possible reason for this difference is discussed below. The rectal V90 values in each patient are shown in Figure 1. The differences in rectal V90 between supine 3- D CRT and prone 3-D CRT were significant (p ). Prone 3-D CRT was superior to supine 3-D CRT in terms of rectal sparing in 14 patients, and among these patients prone 3-D CRT was approximately equal to supine IMRT in six patients and vastly superior to supine IMRT in two patients. Figure 2 shows CT images obtained in the supine and prone positions in one of these two patients, patient no. 14. They show that placing the patient in the prone position caused the CTV to be displaced away from the rectum. In this case, the seminal vesicles were displaced away from the rectum by more than 20 mm by the transition from the supine to the prone position. In contrast, prone 3-D CRT was inferior to supine 3-D CRT in three patients. In these cases, the distance between the CTV and rectum decreased when the patients, in whom the prostate gland adhered to the bladder when they were in the supine position, were changed to the prone position (Figure 3). The values of the bladder V90 in each patient are given in Figure 4. No significant difference in bladder V90 was noted between supine 3-D CRT and prone 3-D CRT (p50.055). As the main subject of this study, the results of rectal and bladder V90 during prone 3-D CRT with supine IMRT are given Figure 5 and Figure 6, respectively. They show that supine IMRT was significantly superior to prone 3-D CRT in terms of rectal sparing (p50.023). We also found that supine IMRT tended to be superior to prone 3-D CRT in bladder sparing, but the difference was not significant (p50.096). Dose volume metrics for the PTV and rectum, along with the results of comparisons using two-sided paired t-tests, are summarised in Table 2. The D 10 and D 20 for the rectum were significantly smaller for supine IMRT than prone 3-D CRT, but the D 05 was approximately equal for supine IMRT and prone 3-D CRT. In contrast, dose homogeneity in the PTV for supine IMRT was significantly inferior to that for prone 3-D CRT. The dose distribution of the central axis plane is demonstrated in Figure 7 for a typical example. In this case, a 50% and 43% reduction in rectal V90 was achieved in supine IMRT and prone 3-D CRT, respectively, compared with supine 3-D CRT. We confirmed that the CTV is slightly displaced away from the rectum by placing the patient in the prone position. Although in some cases the distance between the CTV and the rectum could be changed by more than 20 mm in the transition from the supine to the prone position, the change in distance was about 5 mm in many other cases. Discussion In this study we evaluated the relative impact of prone 3-D CRT and supine IMRT on doses to normal structures, and in particular the rectum. We found that prone 3-D CRT was associated with a better dose distribution to the rectum than supine 3-D CRT, as reported in previous studies [7, 8]; however, in most cases, prone 3-D CRT was not superior to supine IMRT. In addition, we confirmed that prone 3-D CRT is not always superior to supine 3-D CRT in terms of rectal sparing. In the case of patients in whom the prostate gland adheres to the bladder in the supine position, the distance between the CTV and the rectum can be reduced by changing to the prone position, when the bladder hangs down as a result of gravity. In this study, three patients fell into this category, and the prone position should not be adopted in cases similar to these. On the other hand, prone 3-D CRT was approximately equal or superior to supine IMRT in terms of rectal sparing in eight cases. If it were possible to predict the patients in whom prone 3-D CRT would be more effective than supine IMRT in reducing rectal dose reduction, e.g. by using some anatomical predictor before treatment planning, prone 3-D CRT would be a valid option in such cases. Kitamura et al [16] reported that internal organ motion is less frequent in the supine position than in the prone position. Thus, adopting the prone position during, for example, treatment of prostate cancer could lead to other problems, such as inaccurate treatment delivery [16 18]. Nevertheless, it is worth considering adoption of the prone position. As a means of improving the accuracy of treatment delivery in the prone position, a belly board device, which can reduce table top pressure, or IGRT may be effective [16, 20]. However, Stroom et al [13] reported that the two treatment positions are associated with the same degree of geometrical uncertainty; thus, further investigations are needed to ascertain whether the prone position Table 1. Structure volume information for the two treatment positions (n 5 17) Supine Prone p-value CTV volume (cm 3 ) ( ) ( ) Rectal volume (cm 3 ) ( ) ( ) Bladder volume (cm 3 ) ( ) ( ) CTV, clinical target volume. 656 The British Journal of Radiology, August 2009

4 Prone 3-D CRT vs supine IMRT for prostate cancer Figure 1. Volume of the rectum receiving at least 90% of the prescription dose (V90) during supine three-dimensional conformal radiotherapy (3-D CRT), prone 3-D CRT and supine intensity-modulated radiotherapy (IMRT) in each patient. results in less accurate treatment delivery. In this study, the CT scanned range was not sufficient to determine how to predict those patients who would achieve the greatest rectal dose reduction by assuming a prone (a) (a) (b) Figure 2. (a) CT image of patient no.14 in the supine treatment position. (b) CT image of the same patient in the prone treatment position. Both images are taken through the same level of the seminal vesicles. We confirmed that the clinical target volume is displaced away from the rectum in the prone position. The outlined structures shown in each figure are defined as follows: red, clinical target volume (prostate gland plus seminal vesicles); green, rectum; yellow, bladder. (b) Figure 3. Surface rendering of bladder (yellow), rectum (green) and clinical target volume (red) in a lateral threedimensional view demonstrating bladder and rectal position in (a) the supine and (b) the prone position in the same patient (patient no. 6). The British Journal of Radiology, August

5 T Kato, Y Obata, N Kadoya and N Fuwa Figure 4. Volume of the bladder receiving at least 90% of the prescription dose (V90) during supine three-dimensional conformal radiotherapy (3-D CRT), prone 3-D CRT and supine intensity-modulated radiotherapy (IMRT) in each patient. position. Weber et al [19] analysed anatomical values, such as body mass index, body thickness and seminal vesicle volumes, but found that none of these values influenced dose distribution to the rectum or the bladder in the transition between supine and prone positions. We suspect that the quantity of intrapelvic fat is a factor and could be useable as a predictor (as discussed later). We intend to continue using MRI to try to determine if any anatomical factors can be used to predict the appropriate treatment position, i.e. supine or prone. The results for bladder sparing showed much smaller differences between prone 3-D CRT and supine 3-D CRT than were apparent in rectal studies. A significant difference was observed in bladder volume between the supine and prone positions. This observation may be a result of increased bladder filling in the transition from supine to prone positions. However, the magnitude of the increased bladder filling effect was relatively small because the median time interval between the supine and prone planning CT scans was 5 min; it was assumed that the effect of increasing bladder filling did not affect the results of this comparative study. The results for bladder sparing in supine IMRT were slightly better than those for prone 3-D CRT but no significant difference was observed. A possible reason may be that bladder constraints could not be achieved in the IMRT planning process in many cases because of voiding of the bladder prior to the simulation CT. There are various opinions as to the cause of the change in the distance between the CTV and rectum when the patient changes from the supine to the prone position. Zelefsky et al [7] suggested that the treatment position has the greatest effect on the seminal vesicle positioning, affecting the dose to the upper portions of the rectal wall. They also reported that in some cases, although the mobility of the rectum may be restricted as Figure 5. Comparison of supine intensity-modulated radiotherapy (IMRT) and prone three-dimensional conformal radiotherapy (3D CRT) in a mean volume of rectal V90. Error bar indicates one standard deviation. Figure 6. Comparison of supine intensity-modulated radiotherapy (IMRT) and prone three-dimensional conformal radiotherapy (3D CRT) in the mean volume of the bladder receiving at least 90% (V90). Error bar indicates one standard deviation. 658 The British Journal of Radiology, August 2009

6 Prone 3-D CRT vs supine IMRT for prostate cancer Table 2. Dose volume metrics of supine IMRT and prone 3-D CRT for the PTV and rectum Volume Metric Supine IMRT Prone 3-D CRT Average Range (min max) Average Range (min max) p-value PTV D max (cgy) , D mean (cgy) HI Rectum D mean (cgy) D 05 (cgy) D 10 (cgy) D 20 (cgy) D 30 (cgy) D CRT, 3D-conformal radiotherapy; D max, maximum dose; D mean, mean dose; D n, dose received by n% of the organ volume; HI, homogeneity index; IMRT, intensity-modulated radiotherapy; PTV, planning target volume. a result of its attachment ot muscles and ligaments, the prostate gland may be more mobile and may tend to shift anteriorly when the patient is in the prone position. However, they also reported that anterior shifts of the prostate gland in the prone position could not be determined from the CT information available. McLaughlin et al [8] reported that a large fraction of rectal sparing was due to rectal retraction. They thought that the mechanism was related to passive displacement of the rectum by the abdominal contents in the prone position. Although we could not clarify the mechanism directly, we believe that the upper portion of the rectum has relatively high mobility because the puborectalis, which pulls the lower portion of rectum in an anterior direction, lies extremely low in the pelvis, and the prone position tends to displace the upper portion of the rectum posteriorly as a result of the increased pressure on the abdominal contents from the table top. Close study of CT images from individual patients showed that the quantity of fat surrounding the upper portion of the rectum tended to be greater in prone CT images than in corresponding supine CT images. This suggests that the significant space created between the CTV and rectum resulting from a change in treatment position from supine to prone could be the result of displacement of intrapelvic fat in a posterior direction. In addition, it seems that anterior displacement of the seminal vesicles in the prone position contributes slightly to the rectal dose reduction. However, in a few cases the prostate gland was displaced in the prone position and the degree of displacement was small. Bentel et al [20] analysed four treatment positions to assess the impact of pressure from the table top and patient position on the relationship of the prostate gland, rectum and bladder to the bony pelvis [20]. They concluded that variations in organ location with position change may in fact be due to the presence or absence of the normal force of the table top on the buttocks/lower abdomen. As mentioned above, we speculated that the main causes of rectal dose reduction in prone 3-D CRT compared with supine 3-D CRT were posterior retraction of the rectum due to table top pressure and anterior displacement of the seminal vesicles due to gravity. IMRT is increasingly being used for the treatment of prostate cancer, enabling the dose to the CTV to be increased while minimising the dose to the surrounding OARs, namely the rectum and bladder. Increased radiation dose levels for patients with clinically localised prostate cancer have now become an established standard practice. The dose response has been determined for the dose range Gy, particularly in the intermediate-risk group [21]. Improved disease control through dose escalation should not be at the expense of unacceptable toxicity. IMRT dose escalation for prostate cancer must satisfy this condition. To do so, a high degree of geometric accuracy and better conformity in dose distribution are indispensable. IMRT and IGRT are expected to achieve these results; however, these methods require expensive, specialised equipment. Furthermore, the pre-treatment quality assurance procedure in IMRT is known to be very time-consuming [12]. To be considered, given the added time for planning, treatment delivery, and cost, it is important to quantify the actual benefit of using IMRT. In addition, IMRT requires more monitor units to deliver the prescribed dose than does 3-D CRT. This results in increased leakage and scatter of the dose outside the intended treatment volume, which may lead to a higher risk of radiation-induced secondary malignancies [22]. Although IMRT technology is still in a state of ongoing development, we must give heed to these problems and understand that IMRT may not be suitable for clinical application in all cases. IMRT is known to produce slight dose non-homogeneity in the PTV and hot spots in surrounding normal tissues owing to the trade-off between maintaining dose coverage of the PTV and minimising the dose to the surrounding normal tissues. The clinical value of a more homogeneous dose distribution is unclear. However, smaller hot spots in the PTV could translate into lower urethral doses and potentially a lower risk of bladder neck or urethral symptoms. Hot spots in the PTV, rectum and bladder in current supine IMRT plans are extremely small, so this issue may not be considered clinically significant. We hypothesised that the best way to deliver adequate doses to the CTV with relatively lower doses to the rectum, and avoiding the various problems of IMRT described above, would be by increasing the distance between the CTV and the rectum, for which prone 3-D CRT is considered a promising method. In addition, it is considered to be potentially easier to implement prone 3- D CRT in routine practice and has advantages in terms of cost reduction and labour saving compared with the IMRT technique. McLaughlin et al [8] reported that The British Journal of Radiology, August

7 T Kato, Y Obata, N Kadoya and N Fuwa (a) (b) (c) Figure 7. Isodose distribution of central axis plane for (a) supine three-dimensional conformal radiotherapy (3-D CRT), (b) prone 3-D CRT and (c) supine intensity-modulated radiotherapy in patient no. 3. The dose in each figure is relative to the prescribed dose of 76 Gy (i.e. 100% 5 76 Gy). The isodose levels shown in each figure are red, 100%; green, 90%; blue, 80%; yellow, 50%; light blue, 20%. patient position had a greater impact on rectal sparing than beam arrangement in 3-D CRT. However, our study has demonstrated that this is not the case in IMRT. Ultimately, IMRT in the prone position (prone IMRT) may be the most effective method to reduce rectal dose. However, even if the same treatment position was applied for IMRT, the results of planning studies may be dependent on institutional specific protocols for the CTV treated, patient immobilisation, and rectum and bladder filling. Thus, planning studies must be carried out for each institution. Conclusions This study demonstrated that supine IMRT had a greater impact on rectal sparing than prone 3-D CRT. Although prone 3-D CRT appeared to be one of the effective techniques to reduce rectum dose, the size of the effect may depend on the patient s anatomy and physical condition in prone 3-D CRT compared with supine IMRT. If the patients in whom prone 3-D CRT would result in greater rectal dose reduction than supine IMRT could be identified using some anatomical predictor before treatment planning, then prone 3-D CRT could be a promising technique. We believe that prone 3-D CRT still warrants further investigation because of its advantages in terms of simplicity, costeffectiveness and labour saving. Continued research to find an appropriate anatomical predictor is also required. References 1. Cahlon O, Hunt M, Zelefsky MJ. Intensity-modulated radiation therapy: supportive data for prostate cancer. Semin Radiat Oncol 2008;18: Dawson LA, Jaffray DA. Advances in image-guided radiation therapy. J Clin Oncol 2007;25: Hodel K, Rich WM, Austin P, DiSaia PJ. The role of ovarian transposition in conservation of ovarian function in radical hysterectomy followed by pelvic radiation. Gynecol Oncol 1982;13: Herbert SH, Solin LJ, Hoffman JP, Schultz DJ, Curran WJ, Lanciano RM, et al. Volumetric analysis of small bowel displacement from radiation portals with the use of a pelvic tissue expander. Int J Radiat Oncol Biol Phys 1993;25: Corn BW, Coia LR, Chu JH, Hwang CC, Stafford PM, Hanks GE. Significance of prone positioning in planning treatment for esophageal cancer. Int J Radiat Oncol Biol Phys 1991;21: Merchant TE, McCormick B. Prone position breast irradiation. Int J Radiat Oncol Biol Phys 1994;30: Zelefsky MJ, Happersett L, Leibel SA, Burman CM, Schwartz L, Dicker AP, et al. The effect of treatment positioning on normal tissue dose in patients with prostate cancer treated with three-dimensional conformal radiotherapy. Int J Radiat Oncol Biol Phys 1997;37: McLaughlin PW, Wygoda A, Sahijdak W, Sandler HM, Marsh L, Roberson P, et al. The effect of patient position and treatment technique in conformal treatment of prostate cancer. Int J Radiat Oncol Biol Phys 1999;45: Pinkawa M, Gagel B, Demirel C, Schmachtenberg A, Asadpour B, Eble MJ. Dose-volume histogram evaluation of prone and supine patient position in external beam radiotherapy for cervical and endometrial cancer. Radiother Oncol 2003;69: Huang EH, Pollack A, Levy L, Starkschall G, Dong L, Rosen I, et al. Late rectal toxicity: dose-volume effects of conformal radiotherapy for prostate cancer. Int J Radiat Oncol Biol Phys 2002;54: Fiorino C, Cozzarini C, Vavassori V, Sanguineti G, Bianchi C, Cattaneo GM, et al. Relationships between DVHs and late rectal bleeding after radiotherapy for prostate cancer: analysis of a large group of patients pooled from three institutions. Radiother Oncol 2002;64: Ezzell GA, Galvin JM, Low D, Palta JR, Rosen I, Sharpe MB. Guidance document on delivery, treatment planning, clinical implementation of IMRT: report of the IMRT subcommittee of the AAPM radiation therapy committee. Med Phys 2003;30: The British Journal of Radiology, August 2009

8 Prone 3-D CRT vs supine IMRT for prostate cancer 13. Stroom JC, Koper PC, Korevaaar GA, Os MV, Janssen M, Boer HC, et al. Internal organ motion in prostate cancer patients in prone and supine treatment position. Radiother Oncol 1999;51: Khoo VS, Bedford JL, Webb S, Dearnaley DP. Class solutions for conformal external beam prostate radiotherapy. Int J Radiat Oncol Biol Phys 2003;55: Fiorino C, Reni M, Cattaneo GM, Bolognesi A, Calandrino R. Comparing 3-, 4- and 6-fields techniques for conformal irradiation of prostate and seminal vesicles using dosevolume histograms. Radiother Oncol 1997;44: Kitamura K, Shirato H, Seppenwoolde Y, Onimaru R, Oda M, Fujita K, et al. Three-dimensional intrafractional movement of prostate measured during real-time tumor-tracking radiotherapy in supine and prone treatment position. Int J Radiat Oncol Biol Phys 2002;53: Dawson LA, Litzenberg DW, Brock KK, Sanda M, Sullivan M, Sandler HM, et al. A comparison of ventilatory prostate movement in four treatment positions. Int J Radiat Oncol Biol Phys 2000;48: Bayley AJ, Catton CN, Haycocks T, Kelly V, Alasti H, Bristow R, et al. A randomized trial of supine vs. prone positioning in patients undergoing escalated dose conformal radiotherapy for prostate cancer. Radiother Oncol 2004;70: Weber DC, Nouet P, Rouzaud M, Miralbell R. Patient positioning in prostate radiotherapy: is prone better than supine? Int J Radiat Oncol Biol Phys 2000;47: Bentel GC, Munley MT, Marks LB, Anscher MS. The effect of pressure from the table top and patient position on pelvic organ location in patients with prostate cancer. Int J Radiat Oncol Biol Phys 2000;47: Pollack A, Smith LG, von Eschenbach AC. External beam radiotherapy dose response characteristics of 1127 men with prostate cancer treated in the PSA era. Int J Radiat Oncol Biol Phys 2000;48: Kry SF, Salehpour M, Followill DS, Stovall M, Kuban DA, White RA. The calculated risk of fatal secondary malignancies from intensity-modulated radiation therapy. Int J Radiat Oncol Biol Phys 2005;62: The British Journal of Radiology, August