Acute toxicity of image-guided hypofractionated radiotherapy for prostate cancer: Nonrandomized comparison with conventional fractionation

Urologic Oncology: Seminars and Original Investigations 29 (2011) 523–532

Original article

Acute toxicity of image-guided hypofractionated radiotherapy for prostate cancer: Nonrandomized comparison with conventional fractionation Barbara Alicja Jereczek-Fossa, M.D., Ph.D.a,e,*, Dario Zerini, M.D.a, Cristiana Fodor, M.Sc.a, Luigi Santoro, M.Sc.d, Raffaella Cambria, M.Sc.b, Cristina Garibaldi, M.Sc.b, Barbara Tagaste, B.Sc.a, Andrea Vavassori, M.D.a, Federica Cattani, M.Sc.b, Daniela Alterio, M.D.a, Federica Gherardi, M.D.a, Flavia Serafini, M.D.a, Bernardo Rocco, M.D.c, Gennaro Musi, M.D.c, Ottavio De Cobelli, M.D.c,e, Roberto Orecchia, M.D.a,e a

Department of Radiation Oncology, European Institute of Oncology, Milan, Italy b Department of Medical Physics, European Institute of Oncology, Milan, Italy c Department of Urology, European Institute of Oncology, Milan, Italy d Department of Epidemiology and Biostatistics, European Institute of Oncology, Milan, Italy e University of Milan, Milan, Italy Received 13 August 2009; received in revised form 3 October 2009; accepted 6 October 2009

Abstract Objectives: To compare acute toxicity of prostate cancer image-guided hypofractionated radiotherapy (hypo-IGRT) with conventional fractionation without image-guidance (non-IGRT). To test the hypothesis that the potentially injurious effect of hypofractionation can be counterbalanced by the reduced irradiated normal tissue volume using IGRT approach. Materials and methods: One hundred seventy-nine cT1-T2N0M0 prostate cancer patients were treated within the prospective study with 70.2 Gy/26 fractions (equivalent to 84 Gy/42 fractions, ␣/␤ 1.5 Gy) using IGRT (transabdominal ultrasound, ExacTrac X-Ray system, or cone-beam computer tomography). Their prospectively collected data were compared with data of 174 patients treated to 80 Gy/40 fractions with non-IGRT. The difference between hypo-IGRT and non-IGRT cohorts included fractionation (hypofractionation vs. conventional fractionation), margins (hypo-IGRT margins: 7 mm and 3 mm, for all but posterior margins; respectively; non-IGRT margins: 10 and 5 mm, for all but posterior margins, respectively), and use of image-guidance or not. Multivariate analysis was performed to define the tumor-, patient-, and treatment-related predictors for acute toxicity. Results: All patients completed the prescribed radiotherapy course. Acute toxicity in the hypo-IGRT cohort included rectal (G1: 29.1%; G2: 11.2%; G3: 1.1%) and urinary events (G1: 33.5%; G2: 39.1%; G3: 5%). Acute toxicity in the non-IGRT patients included rectal (G1: 16.1%; G2: 6.3%) and urinary events (G1: 36.2%; G2: 20.7%; G3: 0.6%). In 1 hypo-IGRT and 2 non-IGRT patients, radiotherapy was temporarily interrupted due to acute toxicity. The incidence of mild (G1-2) rectal and bladder complications was significantly higher for hypo-IGRT (P ⫽ 0.0014 and P ⬍ 0.0001, respectively). Multivariate analysis showed that hypo-IGRT (P ⫽ 0.001) and higher PSA (P ⫽ 0.046) are correlated with higher acute urinary toxicity. No independent factor was identified for acute rectal toxicity. No significant impact of IGRT system on acute toxicity was observed. Conclusions: The acute toxicity rates were low and similar in both study groups with some increase in mild acute urinary injury in the hypo-IGRT patients (most probably due to the under-reporting in the retrospectively analyzed non-IGRT cohort). The higher incidence of acute bowel reactions observed in hypo-IGRT group was not significant in the multivariate analysis. Further investigation is warranted in order to exclude the bias due to the nonrandomized character of the study. © 2011 Elsevier Inc. All rights reserved. Keywords: Prostate cancer; Image guided radiotherapy; Acute toxicity; Hypofractionation; Conventional fractionation

1. Introduction * Corresponding author. Tel.: ⫹00390257489037; fax: ⫹00390294379227. E-mail address: [email protected] (B.A. Jereczek-Fossa). 1078-1439/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.urolonc.2009.10.004

Hypofractionated radiation therapy for prostate cancer has become of increasing interest with the recognition of a poten-

524

B.A. Jereczek-Fossa et al. / Urologic Oncology: Seminars and Original Investigations 29 (2011) 523–532

tial improvement in therapeutic ratio with treatments delivered in larger-sized fractions [1,2]. Its introduction into the clinical practice has become possible due to the development of image guided radiotherapy (IGRT) with daily target localization allowing for reduction of safety margins and, as a consequence, decrease in normal tissue irradiation [3]. In our department, three-dimensional conformal (3DCRT) 2-dynamic arc radiotherapy to the dose of 80 Gy in 40 fractions has been used since 2003 for curative prostate cancer treatment, and the data on this technique, dosimetry, toxicity, and tumor control have been published elsewhere [4,5]. In 2006, 3 IGRT modalities were installed: ultrasound-based b-mode acquisition and targeting system (BAT; NOMOS, Cranberry Township, PA), stereo X-Ray imaging system coupled with the 6° of freedom robotic couch (ExacTrac; BrainLAB, Feldkirchen, Germany), and Cone-Beam computer tomography CT (On Board Imager CT; Varian Medical System, Palo Alto, CA). First, the pilot study on the modest increase in dose/fraction (72 Gy in 30 fractions of 2.4 Gy) using IGRT showed the feasibility and good acute toxicity profile of such approach [6]. The second dosimetric study comparing accuracy of daily BAT-based prostate localization with computer tomography localization and electronic portal imaging (EPI) showed accuracy of the BAT IGRT system [7]. Based on these preliminary data and on the report from the Fox Chase Cancer Center randomized study [8], in August 2006 we introduced in our department the hypofractionated IGRT (hypo-IGRT) for the organ-confined prostate cancer. The hypofractionation Fox Chase regimen equivalent to our standard fractionation schedule was prospectively chosen and included the dose of 70.2 Gy given in 26 fractions of 2.7 Gy. Assuming a prostate cancer ␣/␤ ratio of 1.5 Gy, this schedule is biologically equivalent to 84.24 Gy in 2.0 Gy fractions [8,9]. In this way, our standard schedule (80 Gy in 40 fractions) and the new hypofractionated regimen (70.2 Gy in 26 fractions) should have a similar effect on prostate cancer. Assuming acute responding tissue ␣/␤ ratio of 10.0 Gy, the new schedule is biologically equivalent to 74.29 Gy given in 2.0 Gy fractions, suggesting some protective effect on the acutely responding normal tissue, compared with the 80 Gy regimen. Therefore, the hypothesis that the potentially injurious effect of hypofractionation (in particular for late effects) can be counterbalanced by the reduced irradiated normal tissue volume using IGRT approach has become intriguing. Unfortunately, the short follow-up in the majority of phase III hypofractionation trials using IGRT precludes the credible comparison of late toxicity. At the moment, acute injury rates could represent a surrogate for overall treatment-related complications. There are several recent studies showing the correlation between acute and late toxicity in pelvic radiotherapy in general [10 –13], and in particular, in prostate cancer [14 – 17]. Therefore, early identification of treatment-, patient-, and tumor-related variables correlated with acute events

would facilitate to reduce acute and, in consequence, late treatment side effects. The aim of this report is to present the incidence and the predictors for acute toxicity of the hypo-IGRT using three different IGRT systems. The data on the localization errors (including comparison of 3 IGRT systems), tumor control, and late toxicity (including quality of life assessment) will be a subject of future reports. We compare here the acute toxicity of the hypo-IGRT with the acute toxicity observed in the consecutive 174 patients treated in our department between 2003 and 2006 with conventionally fractionated 3D-CRT to 80 Gy in 40 fractions using the same 2-dynamic arc technique without image guidance. This subpopulation made part of the 542 prostate cancer patients treated with 3D-CRT to the dose of 76 – 80 Gy [4]. We extracted the data of the patients treated to 80 Gy in 40 fractions in order to compare them in a nonrandomized manner with our hypo-IGRT schedule equivalent to ⬎80 Gy. Despite important methodological limitations (nonrandomized comparison), such analysis might be helpful to evaluate the new technology (IGRT) associated with new fractionation schedules.

2. Materials and methods 2.1. Inclusion criteria The inclusion criteria for hypo-IGRT were as follows: organ-confined prostate cancer (cT1-2 N0 M0), no concomitant inflammatory bowel disease, no collagenopathy, no hip prosthesis, no significant comorbidity (in particular, no anticoagulation treatment, no significant vasculopathy, or long-lasting diabetes mellitus), limited volume of the prostate and seminal vesicles [clinical target volume (CTV) ⬍ 100 cc], normal urodynamic study (no urinary obstruction, i.e., maximum flow ⬎ 10 ml/s), and written informed consent. Patients were stratified into low, intermediate, and high risk groups, according to the National Comprehensive Cancer Network Criteria (NCCN) [18]. Androgen deprivation (ADT) was permitted for intermediate and high risk patients. 2.2. Image guided radiotherapy Planning 3 mm slicing CT scan was performed, in the supine position with leg immobilization (Combifix– SinMed, The Netherlands). Patients were asked to have full bladder for the simulation CT and each treatment session in order to minimize daily variations in prostate location and to reduce the bladder irradiation (drinking 0.5 liters of water 1 hour before CT and treatment session). The patients were also asked to keep their rectum empty during the whole treatment period.

B.A. Jereczek-Fossa et al. / Urologic Oncology: Seminars and Original Investigations 29 (2011) 523–532

In the patients treated with ExacTrac-IGRT, 2 fiducial markers (Visicoil, Radiomed, Tyngsboro, MA) were inserted with the transrectal ultrasound guidance to the prostate 1 week before simulation. All patients were treated with 2 dynamic lateral 3D conformal arcs 100° wide (40°–140° and 220°–320°). The BrainScan (ver. 5.31; BrainLAB, Feldkirchen, Germany) and ERGO 3D-line treatment planning systems were employed. The dose of 70.2 Gy in 26 fractions was prescribed in the International Commission of Radiation Units (ICRU) point [19], equivalent to 84.24 Gy in 40 fractions, based on the linear quadratic model, ␣/␤ ratio of 1.5 Gy [8,9]. CTV included only the prostate in low risk patients and both prostate and seminal vesicles (at least proximal 10 mm) in the intermediate or high risk patients. In the latter group, seminal vesicles were excluded after 56.7 Gy in 21 fractions (equivalent to 68 Gy in fractions of 2 Gy, based on the linear quadratic model, ␣/␤ ratio of 1.5 Gy). The planning target volume (PTV) margins were reduced in the hypo-IGRT patients compared with the non-IGRT 3D-CRT employed in our department (conventional margins: 10 and 5 mm, for all but posterior margins, respectively) [4]. This reduction was decided in order to lower the potentially increased complication risk from hypofractionation and availability of IGRT. Hypo-IGRT PTV was created by adding a 7 mm margin to CTV in all directions, except for the posterior one where 3 mm margin was added. An additional 5 mm margin was added in all directions around the PTV to account for the beam penumbra. Organ at risk contouring has been described earlier [5]. Dose-volume histograms (DVH) for CTV, PTV, rectum, urinary bladder, and femoral heads were computed. DVH constraints based on our own analysis and other published data have been established for conventional 3D-CRT in our department [4,5,20 –22]. For hypo-IGRT, DVH constraints have been recalculated using normalized total dose using ␣/␤ ratio of 3 and 5 Gy for late responding tissues (rectum, urinary bladder, femoral heads). Based on this calculation, less than 30% of the rectal volume may receive the dose greater or equal to 61 Gy (86% of the total dose), less than 60% of the rectal volume may receive the dose greater or equal to 35 Gy (50% of the total dose). Less than 50% of the bladder volume may receive the dose greater or equal to 61 Gy (86% of the total dose) and less than 50% of femoral head may receive the dose grater or equal to 43 Gy (61% of the total dose). If the treatment plan did not fulfill the DVH constraints, various procedures were undertaken in order to obtain a satisfactory dose distribution (reduction of arc width by 10° anteriorly or posteriorly, reduction of the seminal vesicles volume included in the CTV, earlier exclusion of seminal vesicles). Daily target localization was performed using 1 of 3 modalities (the choice of the IGRT modality was done at the discretion of the physician, patient preference, and anatomy, imaging feasibility, etc.): ultrasound-based b-mode acquisi-

525

tion and targeting system (BAT; NOMOS, Cranberry Township, PA), stereo X-Ray imaging system coupled with the 6° of freedom robotic couch (ExacTrac; BrainLAB, Feldkirchen, Germany), and Cone-Beam computer tomography CT (On Board Imager CT; Varian Medical System, Palo Alto, CA). 2.3. Assessment of the toxicity and tumor control data During radiotherapy, the patients were seen by a radiation oncologist once a week (no preventive measures were given before radiotherapy and only symptomatic therapy was prescribed when side effects occurred). After the treatment, the patients were seen by a radiation oncologist every 6 months (PSA test was performed every 3 months). All hypo-IGRT patients were asked to fill the questionnaires on the quality of life (European Organization for Research and Treatment of Cancer, EORTC CQ30 and specific for prostate cancer PR25), International Prostate Symptom Score (IPSS), and International Index of Erectile Function-5 (IIEF-5) before starting the treatment, and 6, 12, and 24 months after the treatment completion. The results of quality of life assessment will be a subject of a future report when follow-up data are available. 2.4. Control group (non-IGRT) The data of the 174 consecutive prostate cancer patients treated at our department between October 2003 and December 2006 with conventionally fractionated 3D-CRT to 80 Gy in 40 fractions using the same 2-dynamic arc technique without image guidance were analyzed (non-IGRT). The radiotherapy technique was described elsewhere [4,5]. Briefly, simulation CT, planning, and toxicity assessment were similar to the IGRT patients, with different CT slicing (5 mm), CTV-PTV margins (see above), DVH constraints and fractionation schedule (see above). Generally speaking, the differences between 2 cohorts included usage or not of IGRT, fractionation (70.2 Gy/26 fractions equivalent to 84.24 Gy/42 fractions vs. 80 Gy/40 fractions), and CTV-PTV margins (7 and 3 mm vs. 10 and 5 mm, for hypo-IGRT and non-IGRT cohort, respectively). 2.5. Statistical analysis Radiation Therapy Oncology Group/European Organization for Research and Treatment of Cancer (RTOG/ EORTC) criteria were used to evaluate acute treatment toxicity [23]. Sexual dysfunction was not analyzed here (it will be analyzed for late events). We evaluated the occurrence of 2 separate radiotherapyrelated toxicities (rectal and urinary) within the whole population (hypo-IGRT and non-IGRT cases), and in the hypoIGRT cohort alone, as well. Multivariate logistic regression models [24] were planned in both populations to calculate the impact of various patient-, tumor-, and treatment-related

526

B.A. Jereczek-Fossa et al. / Urologic Oncology: Seminars and Original Investigations 29 (2011) 523–532

factors on acute rectal and urinary toxicity. The number of possible predictors to include in each multivariable model was related to the number of reported events. To guarantee robustness of estimates, 10 events per predictor is the optimal ratio [25]. As a result of the small number of rectal toxicities, no more than 3 parameters could be included in the multivariable model planned on the whole population, while no multivariable model at all was performed on the hypo-IGRT cohort. For urinary toxicity, on the contrary, given the great number of events, no limitation on the number of predictors was needed. The statistical significance of each variable from multivariate models was evaluated by the Wald ␹2 test. For acute urinary toxicity evaluated on all patients, the following variables were included: cohort (hypo-IGRT vs. nonIGRT), age (ⱕ65 vs. ⬎65 years), stage (T1 vs. T2⫹T3), initial PSA value (ⱕ10 vs. ⬎10 ng/ml), Gleason score (ⱕ6 vs. ⬎6), NCCN prognostic risk group, previous transurethral prostate resection (TURP: no vs. yes), ADT (concomitant no vs. yes), CTV (prostate alone vs. prostate and seminal vesicles). In the hypo-IGRT cohort we included also IGRT system used and concomitant diseases (no vs. yes). For acute rectal toxicity, the statistical model included: cohort (hypo-IGRT vs. non-IGRT), initial PSA value (ⱕ10 vs. ⬎10 ng/ml), and ADT (concomitant no vs. yes). Statistical analyses were performed using the Statistical Analysis System software, version 8.02 for Windows (SAS Institute, Cary, NC).

3. Results 3.1. Study population (hypo-IGRT and non-IGRT cohorts) One hundred seventy-nine consecutive prostate cancer patients treated with hypo-IGRT between August 2006 and October 2008 at the Department of Radiation Oncology, European Institute of Oncology, Milan, Italy, were included (Table 1). Non-IGRT group included 174 prostate cancer treated with conventionally fractionated radiotherapy and no image guidance. The patients included in the hypo-IGRT cohort had significantly higher age, lower Gleason score, PSA level and tumor stage, and lower NCCN risk category (Table 1). Fewer hypo-IGRT patients were treated with previous TURP and received ADT combined with radiotherapy (Table 1). All patients gave written informed consent. 3.2. Radiotherapy data including acute toxicity All patients completed the prescribed treatment, and only limited acute toxicity was observed (Table 2). There were only a few grades 3 and 4 events in both groups. Acute toxicity in the hypo-IGRT group included rectal (G1: 29.1%; G2: 11.2%; G3: 1.1%) and urinary events (G1: 33.5%; G2: 39.1%; G3: 5%). Acute toxicity in the non-

IGRT patients included rectal (G1: 16.1%; G2: 6.3%) and urinary events (G1: 36.2%; G2: 20.7%; G3: 0.6%). In only 1 hypo-IGRT and 2 non-IGRT patients radiotherapy was temporarily interrupted due to acute toxicity. Median duration of hypo-IGRT and non-IGRT was 38 and 61 days, respectively (Table 2). 3.3. Nonrandomized comparison of acute toxicity in hypo-IGRT and non-IGRT cohorts The incidence of mild (grades 1–2) acute rectal and bladder complications were significantly higher in the hypoIGRT group compared with the non-IGRT population (P ⫽ 0.0014 and P ⬍ 0.0001, respectively). No difference was observed in the number of temporary radiotherapy interruptions between groups (1 and 2 interruptions due to the acute toxicity in the hypo-IGRT and non-IGRT cohorts, respectively). Multivariate analysis showed that hypo-IGRT (P ⫽ 0.001) and higher PSA (P ⫽ 0.046) are correlated with higher acute urinary toxicity (Table 3). No independent factor was identified for acute rectal toxicity (Table 3). The higher incidence of acute bowel reactions observed in the hypo-IGRT group was not significant in the multivariate analysis. 3.4. Multivariate analysis in hypo-IGRT cohort The multivariate analysis limited to the hypo-IGRT group showed that higher PSA (P ⫽ 0.011) is correlated with higher acute urinary toxicity (Table 4). No independent factor was identified for acute rectal toxicity in the hypoIGRT cohort (Table 4). No significant impact of IGRT system on acute toxicity was observed (a trend for higher urinary injury with ExacTrac was found, P ⫽ 0.09). There was also a trend for lower urinary injury in the patients treated with concomitant ADT (P ⫽ 0.076).

4. Discussion Our study showed a small increase in the mild acute urinary reactions in the prostate cancer patients treated with hypofractionated IGRT compared with the patients treated with conventionally fractionated radiotherapy without image guidance. The higher incidence of mild acute bowel reactions observed in the hypo-IGRT group was not significant in multivariate analysis. Grades 3 and 4 acute toxicity rates were similar and low in both cohorts. Our series has a number of limitations. First, it has a nonrandomized character. Indeed, this is a common criticism of modern radiation technology assessment, and numerous investigators have been recently discussing this issue [26]. A new technology might improve treatment quality without a measurable impact on clinical outcome because of the low sensitivity and specificity of clinical endpoints [27]. Ideally, randomized studies would be war-

B.A. Jereczek-Fossa et al. / Urologic Oncology: Seminars and Original Investigations 29 (2011) 523–532

527

Table 1 Patient characteristics (n ⫽ 353 patients) Characteristics

BAT n ⫽ 118

Exactrac n ⫽ 32

CBCT n ⫽ 29

All hypo-IGRT n ⫽ 179

Non-IGRT n ⫽ 174

Age (years), median (range) ⱕ65 ⬎65 Stage T1 T2 T3 Gleason score, median (range) Initial PSA (ng/ml) median (range) NCCN risk group§ Low Intermediate High Previous TURP Concomitant diseases Diabetes Cardiopathy Hypertension Vascular disease Diverticulosis Other neoplasia Dyslipidemia Other diseases ADT added to RT Yes Duration (months), median (range) Type of ADT added to RT CAB Anti-androgen alone‡ LH-RH analog alone ADT as neoadjuvant to RT Yes Duration (months), median (range) ADT as concomitant to RT† Yes ADT as adjuvant to RT Yes Duration (months), median (range)

74 (45–86) 20 (16.9%) 98 (83.1%)

69 (46–79) 10 (31.3%) 22 (68.7%)

74 (56–82) 3 (10.3%) 26 (89.7%)

74 (45–86) 33 (18.4%) 146 (81.6%)

71 (47–84) 45 (25.9%) 129 (74.1%)

81 (68.6%) 37 (31.4%)

18 (56.3%) 14 (43.8%)

19 (65.5%) 10 (34.5%) 7 (6–9) 7.3 (3.1–44.6)

118 (65.9%) 61 (34.1%) 0 (–) 6 (3–9) 6.7 (0.6–82)

63 (36.2%) 80 (46.0%) 31 (17.8%) 7 (4–9) 11.2 (0.7–118)

P* 0.01 0.09 ⬍0.0001

6 (3–9) 6.6 (0.6–82)

6 (5–7) 7 (3.0–19.5)

0.027 ⬍0.0001 ⬍0.0001

57 (50.7%) 46 (40.4%) 11 (9.6%) 5 (4.2%) 91 (77.1%) 12 (10.2%) 30 (25.4%) 41 (34.7%) 12 (10.2%) 4 (3.4%) 8 (6.8%) 6 (5.1%) 55 (44.6%)

18 (56.3%) 14 (43.8%) 0 (–) 1 (3.1%) 23 (71.9%) 2 (6.3%) 8 (25.0%) 10 (31.3%) 3 (9.4%) 0 (–) 5 (15.6%) 2 (6.3%) 8 (25.0%)

8 (28.6%) 14 (50.0%) 6 (21.4%) 1 (3.4%) 25 (86.2%) 0 (–) 5 (17.2%) 9 (31.0%) 3 (10.3%) 1 (3.4%) 4 (13.8%) 1 (3.4%) 17 (58.6%)

83 (47.7%) 74 (42.5%) 17 (9.8%) 7 (3.9%) 139 (77.7%) 14 (7.8%) 43 (24.0%) 60 (33.5%) 18 (10.1%) 5 (2.8%) 17 (9.5%) 9 (5.0%) 80 (44.7%)

32 (18.4%) 68 (39.1%) 74 (42.5%) 18 (10.3%)

0.018

35 (29.7%) 5 (0–123)

5 (15.6%) 6 (3–17)

14 (48.3%) 7.5 (1–16)

54 (30.2%) 6 (0–123)

120 (69.0%) 10 (1–47)

⬍0.0001 0.001

14 (11.9%) 4 (3.4%) 17 (14.4%)

2 (6.3%) 1 (3.1%) 2 (6.3%)

6 (20.7%) 0 (–) 8 (27.6%)

22 (12.3%) 5 (2.8%) 27 (15.1%)

70 (40.2%) 11 (6.3%) 39 (22.4%)

⬍0.0001

34 (28.8%) 3 (0–124)

5 (15.6%) 3 (2–6)

14 (48.3%) 4.5 (1–15)

53 (29.6%) 3 (0–124)

120 (69.0%) 4 (0–31)

⬍0.0001 0.52

30 (25.4%)

5 (15.6%)

12 (41.4%)

47 (26.3%)

103 (59.2%)

⬍0.0001

13 (11.0%) 5.5 (1–10)

4 (12.5%) 2 (1–12)

6 (20.7%) 9 (1–11)

23 (12.8%) 5.5 (1–12)

68 (39.1%) 6 (0–37)

⬍0.0001 0.28

RT ⫽ radiotherapy; IGRT ⫽ image guided radiotherapy; hypo-IGRT ⫽ hypofractionated IGRT; BAT ⫽ b-mode ultrasound acquisition and targeting system; Exactrac ⫽ X-ray system; CBCT ⫽ cone beam computer tomography; ADT ⫽ androgen deprivation therapy; CAB ⫽ complete androgen blockade; LHRH ⫽ luteinizing hormone releasing factor; TURP ⫽ transurethral prostate resection; SD ⫽ standard deviation; NCCN ⫽ National Comprehensive Cancer Network [18]. * Comparison between all-hypo-IGRT vs. Non-IGR cohorts. † At least 1 day when both treatments were given simultaneously. ‡ Anti-androgens included mainly bicalutamide. § 5 Patients with missing data excluded from the percentages calculation.

ranted in order to compare 2 devices or treatment modalities (technologies), however, due to the mentioned low sensitivity and specificity of clinical endpoints, they would require enormous expense and human resources. In real life, we can assess the new radiation technology using rigorous clinical outcome analysis, keeping in mind all the limitations of a nonrandomized investigation. We believe that our paper analyzing the data of almost 400 consecutive prostate cancer patients treated over the last years (2003–2008), when technology evolved from the 3D conformal conventionally fractionated radiotherapy to im-

age-guided hypofractionated irradiation, presents a critical comparison of technologies. Further patient follow-up and analysis of tumor outcome will undoubtedly produce useful data regarding the clinical value of these technologies. In our series, the hypo-IGRT cohort includes patients treated within a prospective study where all the data were prospectively registered. The data on the non-IGRT cohort were retrospectively collected from the patient charts, leading most probably to some under-reporting of toxicity rates, in particular regarding mild, clinically irrelevant events. Therefore, several authors of retrospective analyses of pel-

528

B.A. Jereczek-Fossa et al. / Urologic Oncology: Seminars and Original Investigations 29 (2011) 523–532

Table 2 Radiotherapy technical data, toxicity, and patient compliance Characteristics Year of treatment 2003–2005 2006 2007 2008 RT duration (days), median (range) RT volume Prostate only Prostate ⫹ seminal ves. RT acute toxicity* Rectal No rectal toxicity G1 G2 G3 G4 Urinary No urinary toxicity G1 G2 G3 G4 RT temporarily interrupted Number of patients Duration (days), median (range) Reasons: Acute toxicity Personal reasons Other non-RT related disorders RT definitively interrupted Number of patients

BAT n ⫽ 118

Exactrac n ⫽ 32

CBCT n ⫽ 29

All hypo-IGRT n ⫽ 179

Non-IGRT n ⫽ 174

P†

0 (–) 9 (7.6%) 58 (49.2%) 51 (43.2%) 39 (35–53)

0 (–) 2 (6.3%) 13 (40.6%) 17 (53.1%) 37 (34–45)

0 (–) 2 (6.9%) 16 (55.2%) 11 (37.9%) 40 (36–54)

0 (–) 13 (7.3%) 87 (48.6%) 79 (44.1%) 38 (34–54)

109 (62.6%) 65 (37.4%) 0 (–) 0 (–) 61 (35–76)

57 (48.3%) 61 (51.7%)

19 (59.4%) 13 (40.6%)

11 (37.9%) 18 (62.1%)

87 (48.6%) 92 (51.4%)

47 (27.0%) 127 (73.0%)

75 (63.6%) 32 (27.1%) 9 (7.6%) 2 (1.7%) 0 (–)

14 (43.8%) 11 (34.4%) 7 (21.9%) 0 (–) 0 (–)

16 (55.2%) 9 (31.0%) 4 (13.8%) 0 (–) 0 (–)

105 (58.7%) 52 (29.1%) 20 (11.2%) 2 (1.1%) 0 (–)

135 (77.6%) 28 (16.1%) 11 (6.3%) 0 (–) 0 (–)

0.0014

28 (23.7%) 42 (35.6%) 40 (33.9%) 8 (6.8%) 0 (–)

9 (28.1%) 6 (18.8%) 17 (53.1%) 0 (–) 0 (–)

3 (10.3%) 12 (41.4%) 13 (44.8%) 1 (3.4%) 0 (–)

40 (22.3%) 60 (33.5%) 70 (39.1%) 9 (5.0%) 0 (–)

63 (36.2%) 72 (41.4%) 36 (20.7%) 1 (0.6%) 2 (1.1%)

⬍0.0001

1 (0.8%) 11 (11–11)

0 (–) —

1 (3.4%) 11 (11–11)

2 (1.1%) 11 (11–11)

4 (2.3%) 8 (2–10)

0.39 0.10

1 (0.8%) 0 (–) 0 (–)

0 (–) 0 (–) 0 (–)

0 (–) 0 (–) 1 (3.4%)

1 (0.6%) 0 (–) 1 (0.6%)

2 (1.1%) 1 (0.6%) 1 (0.6%)

0 (–)

0 (–)

0 (–)

0 (–)

0 (–)

⬍0.0001

⬍0.0001 ⬍0.0001

RT ⫽ radiotherapy; IGRT ⫽ image guided radiotherapy; hypo-IGRT ⫽ hypofractionated IGRT; BAT ⫽ b-mode ultrasound acquisition and targeting system; Exactrac ⫽ X-ray system; CBCT ⫽ cone beam computer tomography. * RTOG/EORTC criteria [23]. † Comparison between all-hypo-IGRT vs. non-IGRT cohorts.

vic radiotherapy-related toxicity report only severe events [28,29], or even neglect reporting any acute reactions due to insufficient information in medical records [30] or due to the self-limiting character of acute morbidity [31–33]. This difference in the data collection could at least partially explain the increase in the acute urinary toxicity observed in our study. The comparison of prospectively and retrospectively collected data has already demonstrated the increase in mild toxicity in pelvic malignancies cited in [13]. However, one cannot exclude that the hypofractionation has contributed to the slightly increased acute urinary toxicity in the hypo-IGRT cohort. Hypofractionated IGRT used in our series includes some factors that could increase the acute toxicity (higher dose/ fraction) and some variables correlated with the potential decrease of acute events (lower equivalent dose due to the lower total dose, reduced irradiated volume due to the margin reduction, and use of image guidance). Based on our report, one has to be careful to claim that the potentially injurious effect of hypofractionation was counterbalanced

by the reduced irradiated volume. Further radiobiological and clinical investigation including prospective toxicity registration based on the objective criteria is warranted in order to understand the mechanisms and predictors of acute normal tissue injury. In particular, using the acute toxicity as a surrogate of overall radiotherapy complications due to the mentioned earlier correlation between acute and late injury might be not applicable to hypofractionation studies (the correlation has been shown in the conventionally fractionation series). The analysis of urinary tract complications in prostate cancer patients is difficult due to the interference with preexisting dysfunction (correlated to the age, previous surgery, including transurethral prostate resection or prostatectomy, etc.) [34,35] and unclear influence of androgen deprivation [2,14]. Urinary toxicity is influenced by urinary symptoms before the start of radiotherapy [14,34,35]. Some of these pre-existing symptoms may be erroneously registered as acute and even late urinary toxicity. Unfortunately, we have the baseline evaluation of urinary symptoms avail-

B.A. Jereczek-Fossa et al. / Urologic Oncology: Seminars and Original Investigations 29 (2011) 523–532

529

Table 3 Univariate and multivariate analysis of potential predictors for grade ⱖ 2 acute rectal and urinary toxicity (N ⫽ 353) Total

Urinary toxicity

Rectal toxicity

Univariate Events All patients Cohort Non-IGRT (2.0 Gy) Hypo-IGRT (2.7 Gy) Age (years) ⱕ65 ⬎65 Stage T1 T2 T3 Initial PSA ⬍10 ng/ml ⱖ10 ng/ml Gleason score ⬍⫽6 ⬎6 mg/ml NCCN risk group Low Intermediate High Previous TURP No Yes Concomitant diseases No Yes Neoadjuvant ADT No Yes Concomitant ADT No Yes RT volume Prostate only Prostate ⫹ seminal ves.

N ⫽ 353

118 (33.4%)

n ⫽ 174 n ⫽ 179

39 (22.4%) 79 (44.1%)

Multivariate* P value (Wald ␹2)

Univariate

OR (95% CI)

P value (Wald ␹2)

Events

n ⫽ 181 n ⫽ 141 n ⫽ 31

68 (37.6%) 48 (34.0%) 2 (6.4%)

n ⫽ 218 n ⫽ 135

72 (33.0%) 46 (34.1%)

0.001

11 (6.3%) 22 (12.3%)

1 1.54 (0.84–2.81)

0.16

40 (34.8%) 56 (39.4%) 20 (22.0%)

n ⫽ 328 n ⫽ 25

113 (34.4%) 5 (20.0%)

n ⫽ 40 n ⫽ 139

15 (37.5%) 64 (46.0%)

n ⫽ 180 n ⫽ 173

73 (40.6%) 45 (26.0%)

n ⫽ 203 n ⫽ 150

80 (39.4%) 38 (25.3%)

0.62

0.36

0.087 1 1.93 (1.01–3.67)

0.046

25 (11.5%) 8 (5.9%) 0.39

1 1.04 (0.51–2.12)

0.92

15 (8.4%) 18 (11.1%) 0.80

1 1.67 (0.68–4.07) 1.02 (0.32–3.22) 0.14

0.26 0.98 0.24

12 (10.4%) 14 (9.9%) 7 (7.7%) 0.81 31 (9.4%) 2 (8.0%)

0.53 (0.18–1.53) 0.34

1 0.96

5 (12.5%) 17 (12.2%) 0.004

0.063 22 (12.2%) 11 (6.4%)

0.006

0.21 1 0.68 (0.37–1.24)

0.15 51 (38.1%) 67 (30.6%)

1 0.66 (0.27–1.61)

21 (11.6%) 11 (7.8%) 1 (3.2%)

0.02

n ⫽ 134 n ⫽ 219

0.24

0.14 1 0.88 (0.52–1.48)

0.40

n ⫽ 115 n ⫽ 142 n ⫽ 91

1 1.62 (0.72–3.63)

7 (9.0%) 26 (9.4%)

0.84

63 (35.2%) 53 (32.7%)

P value (Wald ␹2)

0.90

0.003

n ⫽ 179 n ⫽ 162

OR (95% CI)

0.058 1 2.47 (1.42–4.30)

0.099 20 (25.6%) 98 (35.6%)

P value (Wald ␹2)

33 (9.3%) ⬍0.0001

n ⫽ 78 n ⫽ 275

Multivariate†

0.068 24 (11.8%) 9 (6.0%)

0.22 1 0.64 (0.31–1.31)

1 0.64 (0.27–1.52)

0.31

0.58 14 (10.4%) 19 (8.7%)

IGRT ⫽ image guided radiotherapy; hypo-IGRT ⫽ hypofractionated IGRT (cohort); non-IGRT ⫽ non-image guided radiotherapy (cohort); NCCN ⫽ National Comprehensive Cancer Network [18]; TURP ⫽ transurethral prostate resection; ADT ⫽ androgen deprivation therapy; RT ⫽ radiotherapy; OR ⫽ overall risk; CI ⫽ confidence interval. * Logistic regression including cohort (hypo-IGRT vs. non-IGRT), age (ⱕ65 vs. ⬎65 years), stage (T1 vs. T2⫹T3), initial PSA value (ⱕ10 vs. ⬎10 ng/ml), Gleason score (ⱕ6 vs. ⬎6), NCCN prognostic risk group, previous TURP (no vs. yes), concomitant ADT (no vs. yes), CTV (prostate alone vs. prostate and seminal vesical). † Logistic regression including cohort (hypo-IGRT vs. non-IGRT), initial PSA value (ⱕ10 vs. ⬎10 ng/ml), ADT (concomitant no vs. yes)

able only for the hypo-IGRT patients, so its impact on acute toxicity cannot be analyzed in our study. In the era of dose escalation radiotherapy, urinary toxicity and, in particular, bladder neck injury (areas receiving close to 100% of the prescribed dose) becomes dose-limiting toxicity. However, the knowledge on the radiobiological features of this tissue is still limited [36,37]. Again, due to its partially retrospective character, our report does not include DVH analysis. Such analysis will be a subject of a future study on all hypo-IGRT patients and a subgroup of non-IGRT cohort. Several authors found only weak or no relationship between

acute toxicity and dose-volume parameters [8,38 – 44]. Pollack et al. [8] found that bladder volume at planning was independently associated with an increase in acute effects. In our analysis, higher PSA and no use of ADT were correlated to higher urinary toxicity, suggesting some impact of the prostate volume. Another limitation of our study is the use of the RTOG/ EORTC scoring system based on the subjective evaluation and treatment-requiring criteria (for example, dysuria requiring local anesthesia). Such endpoints may, however, be related to the management policy of the treating physicians

530

B.A. Jereczek-Fossa et al. / Urologic Oncology: Seminars and Original Investigations 29 (2011) 523–532

Table 4 Univariate and multivariate analysis of potential predictors for grade ⱖ 2 acute toxicity on hypo-IGRT cohort only Total

Urinary toxicity Univariate Events

All patients IGRT method BAT Exactrac CBCT Age (years) ⱕ65 ⬎65 Stage T1 T2 Initial PSA ⬍10 ng/ml ⱖ10 ng/ml Gleason Score ⱕ6 ⬎6 mg/mL NCCN risk group Low Intermediate High Previous TURP No Yes Concomitant Diseases No Yes Neoadjuvant ADT No Yes Concomitant ADT No Yes

N ⫽ 179

79 (44.1%)

n ⫽ 118 n ⫽ 32 n ⫽ 29

48 (40.7%) 17 (53.1%) 14 (48.3%)

n ⫽ 33 n ⫽ 146

12 (36.4%) 67 (45.9%)

n ⫽ 118 n ⫽ 61

Rectal toxicity Multivariate*

P value (Wald ␹2)

Univariate

OR (95% CI)

P value (Wald ␹2)

Events

Multivariate† P value (Wald ␹2)

OR (95% CI)

P value (Wald ␹2)

22 (9.3%) 1 2.07 (0.89–4.83) 1.27 (0.52–3.06)

11 (9.3%) 7 (21.9%) 4 (13.8%)

0.17

0.09 0.60

1 1.51 (0.64–3.58)

0.35

5 (15.1%) 17 (11.6%)

0.58

0.32

56 (47.5%) 23 (37.7%)

0.21

1 0.56 (0.28–1.13)

0.10

16 (13.6%) 6 (9.8%)

0.47

n ⫽ 137 n ⫽ 42

54 (39.4%) 25 (59.5%)

0.02

1 3.36 (1.32–8.52)

0.011

19 (13.9%) 3 (7.1%)

0.25

n ⫽ 99 n ⫽ 75

44 (44.4%) 33 (44.0%)

0.95

1 1.18 (0.38–3.63)

0.77

11 (11.1%) 11 (14.7%)

n ⫽ 83 n ⫽ 74 n ⫽ 17

35 (42.2%) 35 (47.3%) 7 (41.2%)

1 0.88 (0.26–3.00) 0.82 (0.14–4.80)

0.84 0.83

10 (12.0%) 9 (12.2%) 3 (17.6%)

n ⫽ 172 n⫽7

76 (44.2%) 3 (42.9%)

0.95

1 0.93 (0.17–5.02)

0.94

21 (12.2%) 1 (14.3%)

0.87

n ⫽ 40 n ⫽ 139

15 (37.5%) 64 (46.0%)

0.34

1 1.41 (0.64–3.12)

0.40

5 (12.5%) 17 (12.2%)

0.96

n ⫽ 126 n ⫽ 53

59 (46.8%) 20 (37.7%)

0.26

18 (14.3%) 4 (7.6%)

0.21

n ⫽ 132 n ⫽ 47

63 (47.7%) 16 (34.0%)

0.10

18 (13.6%) 4 (8.5%)

0.36

0.40

0.78

1 0.45 (0.19–1.09)

0.076

0.48

0.78

IGRT ⫽ image guided radiotherapy; hypo-IGRT ⫽ hypofractionated IGRT (cohort); BAT ⫽ b-mode ultrasound acquisition and targeting system; Exactrac ⫽ X-ray system; CBCT ⫽ cone beam computer tomography; NCCN ⫽ National Comprehensive Cancer Network [18]; TURP ⫽ transurethral prostate resection; ADT ⫽ androgen deprivation therapy; OR ⫽ overall risk; CI ⫽ confidence interval. * Logistic regression including IGRT system (BAT vs. Exactrac vs. CBCT), age (ⱕ65 vs. ⬎65 years), stage (T1 vs. T2⫹T3), initial PSA value (ⱕ10 vs. ⬎10 ng/ml), Gleason score (ⱕ6 vs. ⬎6), NCCN prognostic risk group, previous TURP (no vs. yes), concomitant diseases (no vs. yes), concomitant ADT (no vs. yes). † Due to the small number of rectal toxicities no multivariable model was performed on the hypo-IGRT cohort.

(subjective to the changes over the studied periods) and influenced by patient-related factors, such as age and comorbidities [33]. Definitely, one has to be cautious drawing conclusions from this type of study. It might allow a general assumption, but specific conclusions can be elusive. On the other hand, our series includes two large cohorts of patients treated in 1 center over the limited period, reflecting just “real life” transition from conventionally fractionated radiotherapy to the new technology for prostate cancer (availability of retrospective data for the old technique and prospective data collection for the new approach). The results of our study allow for the generation of the hypothesis that IGRT enables the introduction of safe “soft” hypofractionation, offering a

shorter overall treatment time with an advantage for patients who live far away from radiation oncology facilities, and providing potentially more economic healthcare. Importantly, our study, as well as the observations of other investigators, suggest that hypofractionation might be correlated with some increase in acute toxicity [1]. Such findings might have implications for daily practice in the centers starting to use hypofractionation with or without IGRT. We employed in our study 3 different IGRT systems, and the choice of the system was not randomized. Our analysis did not show any impact of the IGRT system on acute toxicity, however, the trend for higher urinary injury with ExacTrac was found. Obviously, the assessment of the

B.A. Jereczek-Fossa et al. / Urologic Oncology: Seminars and Original Investigations 29 (2011) 523–532

IGRT and hypofractionation for prostate cancer is complex and will require long-term evaluation, including tumor control, late toxicity data, and quality of life analysis. Hypofractionated schedules, apart from hypothetical radiobiological advantages, offer more economic healthcare and higher patient convenience due to the shorter overall treatment time (in our series, overall treatment time of hypofractionated IGRT was 62% of the duration of non-IGRT). This benefit must be a trade-off against possible increase in some toxicity as reported by our study. Interestingly, the Dutch investigators showed that most patients with localized prostate cancer prefer a less aggressive radiotherapy [45]. Many patients attach more weight to specific quality-of-life aspects (in particular rectal toxicity) than to improving survival. The Dutch study demonstrates that treatment preferences of patients with localized prostate cancer can and should be involved in radiotherapy decision-making [45]. Importantly, hypofractionated IGRT used in our series was correlated with some increase in urinary events, and no increase in rectal toxicity. At the moment, these preliminary findings are discussed with the patient candidates for IGRT, however, further data are necessary. Several reports on hypofractionated radiotherapy have been published recently, and some modern randomized trials are ongoing (discussed in [37]). In the nonrandomized comparison of hypofractionated schedule (without image guidance) with standard fractionation, hypofractionated regimens (3 Gy/fraction and 3.15 Gy/fraction) were correlated with increase in grade ⱖ 2 acute rectal and urinary toxicity [37]. In the phase III study from the Fox Chase Cancer Center, where the patients were randomized between standard (76 Gy in 38 fractions) and hypofractionated (the regimen employed in our series) BAT-based IMRT, a small but significant increase in acute bowel reactions was observed in the hypofractionated arm [8]. Similar to our report, also in the Fox Chase series, the PTV margins were smaller than in the standard arm, but such volume reduction was not enough to counterbalance the effect of higher dose/ fraction in that series.

5. Conclusions Our study showed that acute toxicity rates were low and similar in both study groups with some increase in mild acute urinary injury in the hypo-IGRT patients (most probably due to the under-reporting in the retrospective nonIGRT cohort). The higher incidence of acute bowel reactions observed in hypo-IGRT group was not significant in the multivariate analysis. Further investigation is warranted in order to exclude the bias due to nonrandomized character of the study and to test the hypothesis that the potentially injurious effect of hypofractionation can be counterbalanced by the reduced irradiated normal tissue volume using IGRT approach. We generate the hypothesis that IGRT allows for safe “soft” hypofractionation, offering a shorter overall

531

treatment time, and providing potentially more economical healthcare.

References [1] Ritter M. Rationale, conduct, and outcome using hypofractionated radiotherapy in prostate cancer. Semin Radiat Oncol 2008;18: 249 –56. [2] Jereczek-Fossa BA, Orecchia R. Evidence-based radiation oncology: Definitive, adjuvant, and salvage radiotherapy for nonmetastatic prostate cancer. Radiother Oncol 2007;84:197–215. [3] Kupelian PA, Langen KM, Willoughby TR, et al. Image-guided radiotherapy for localized prostate cancer: Treating a moving target. Semin Radiat Oncol 2008;18:58 – 66. [4] Jereczek-Fossa BA, Vavassori A, Fodor C, et al. Dose escalation for prostate cancer using the three-dimensional conformal dynamic arc technique: Analysis of 542 consecutive patients. Int J Radiat Oncol Biol Phys 2008;71:784 –94. [5] Jereczek-Fossa BA, Cattani F, D’Onofrio A, et al. Dose distribution in three-dimensional conformal radiotherapy for prostate cancer: Comparison of two treatment techniques (six coplanar fields and two dynamic arcs). Radiother Oncol 2006;81:294 –302. [6] Zerini D, Jereczek-Fossa BA, Trovò M, et al. Radioterapia 3-dimensionale conformazionale (3D-CRT) ipofrazionata nel carcinoma della prostata allo stadio iniziale con l’ausilio della localizzazione ecografica quotidiana: Dati preliminari. Ital Arch Urol Androl 2006;89:S199 [Abstract]. [7] Jereczek-Fossa BA, Cattani F, Garibaldi C, et al. Transabdominal ultrasonography, computed tomography and electronic portal imaging for three-dimensional conformal radiotherapy for prostate cancer. Strahlenther Onkol 2007;183:610 – 6. [8] Pollack A, Hanlon AL, Horwitz EM, et al. Dosimetry and preliminary acute toxicity in the first 100 men treated for prostate cancer on a randomized hypofractionation dose escalation trial. Int J Radiat Oncol Biol Phys 2006;64:518 –26. [9] Brenner DJ, Martinez AA, Edmundson GK, et al. Direct evidence that prostate tumors show high sensitivity to fractionation (low ␣/␤ ratio), similar to late-responding normal tissue. Int J Radiat Oncol Biol Phys 2002;52:6 –13. [10] Weiss E, Hirnle P, Arnold-Bofinger H, et al. Therapeutic outcome and relation of acute and late side effects in the adjuvant radiotherapy of endometrial carcinoma stages I and II. Radiother Oncol 1999;53: 37– 44. [11] Wang CJ, Leung SW, Chen HC, et al. The correlation of acute toxicity and late rectal injury in radiotherapy for cervical carcinoma: Evidence suggestive of consequential late effect (CQLE). Int J Radiat Oncol Biol Phys 1998;40:85–91. [12] Bourne RG, Kearsley JH, Grove WD, et al. The relationship between early and late gastrointestinal complications of radiation therapy for carcinoma the cervix. Int J Radiat Oncol Biol Phys 1983;9:1445–50. [13] Jereczek-Fossa BA, Jassem J, Badzio A. Relationship between acute and late normal tissue injury after postoperative radiotherapy in endometrial cancer. Int J Radiat Oncol Biol Phys 2002;52:476 – 82. [14] 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;37:3–11. [15] Zelefsky MJ, Cowen D, Fuks Z, et al. Long term tolerance of high dose three-dimensional conformal radiotherapy in patients with localized prostate carcinoma. Cancer 1999;85:2460 – 8. [16] Pearse M, Choo R, Danjoux C, et al. Prospective assessment of gastrointestinal and genitourinary toxicity of salvage radiotherapy for patients with prostate-specific antigen relapse or local recurrence after radical prostatectomy. Int J Radiat Oncol Biol Phys 2008;72:792– 8. [17] Jereczek-Fossa BA, Zerini D, Fodor C, et al. Correlation between acute and late toxicity in 973 prostate cancer patients treated with

532

[18]

[19]

[20]

[21]

[22]

[23]

[24] [25] [26] [27] [28]

[29]

[30]

[31]

B.A. Jereczek-Fossa et al. / Urologic Oncology: Seminars and Original Investigations 29 (2011) 523–532 radical or post-prostatectomy external beam irradiation. Int J Radiat Oncol Biol Phys, in press. Prostate cancer. In: National Comprehensive Cancer network Clinical Practice Guidelines in Oncology, version 1.2007. Jenkinstown, PA: National Comprehensive Cancer Network, March 2007. (http:// www.nccn.org/professionals/physician_gls/default.asp). ICRU. Report 62: Prescribing, recording and reporting photon beam therapy (supplement to ICRU report 50). Bethesda: International Commission on Radiation Units and Measurements, 1999. Boersma LJ, van den Brink M, Bruce AM, et al. Estimation of the incidence of late bladder and rectum complications after high-dose (70 –78 Gy) conformal radiotherapy prostate cancer. Int J Radiat Oncol Biol Phys 1998;41:83–92. Fiorino C, Sanguineti G, Cozzarini C, et al. Rectal dose-volume constraints in high-dose radiotherapy of localized prostate cancer. Int J Radiat Oncol Biol Phys 2003;57:953– 62. Greco C, Mazzetta C, Cattani F, et al. Finding dose-volume constraints to reduce late rectal toxicity following 3D-conformal radiotherapy (3D-CRT) of prostate cancer. Radiother Oncol 2003; 69:215–22. Cox JD, Stetz J, Pajak TF. Toxicity criteria of the Radiation Therapy Oncology Group (RTOG) and the European Organization for Research and Treatment of Cancer (EORTC). Int J Radiat Oncol Biol Phys 1995;31:1341– 6. Hosmer DW, Lemeshow S. Applied Logistic Regression, 2nd ed. New York (NY): John Wiley and Sons, 2000. Harrell FEJ. Regression modeling strategies. New York (NY): Springer-Verlag, 2001. Bentzen SM. Radiation oncology health technology assessment—the best is the enemy of the good. Nat Clin Pract Oncol 2008;5:563. Bentzen SM. Randomized controlled trials in health technology assessment: Overkill or overdue?. Radiother Oncol 2008;86:142–7. Jackson SM, Fairey RN, Kornelsen RO, et al. Clinical results in carcinoma of the cervix; radium compared to caesium using remote afterloading. Clin Radiol 1989;40:302– 6. Kilic D, Egehan I, Ozenirler S, et al. Double-blinded, randomized, placebo-controlled study to evaluate the effectiveness of sulphasalazine in preventing acute gastrointestinal complications due to radiotherapy. Radiother Oncol 2000;57:125–9. Lewandowski G, Torrisi J, Potkul RK, et al. Hysterectomy with extended surgical staging and radiotherapy versus hysterectomy alone and radiotherapy in stage I endometrial cancer. Gynecol Oncol 1990; 36:401– 4. Greven KM, Corn BW, Case D, et al. Which prognostic factors influence the outcome of patients with surgically staged endometrial cancer treated with adjuvant radiation? Int J Radiat Oncol Biol Phys 1997;39:413– 8.

[32] Small W, Mahadevan A, Roland P, et al. Whole-abdominal radiation in endometrial carcinoma: An analysis of toxicity, patterns of recurrence, and survival. Cancer 2000;6:394 – 400. [33] Bentzen SM, Overgaard J. Patient-to-patient variability in the expression of radiation-induced normal tissue injury. Sem Radiat Oncol 1999;4:68 – 80. [34] Liu M, Pickles T, Berthelet E, et al. Urinary incontinence in prostate cancer patients treated with external beam radiotherapy. Radiother Oncol 2005;74:197–201. [35] Sandhu AS, Zelefsky MJ, Lee HJ, et al. Long-term urinary toxicity after three-dimensional conformal radiotherapy for prostate cancer in patients with prior history of transurethral resection. Int J Radiat Oncol Biol Phys 2000;48:643–7. [36] Cahlon O, Hunt M, Zelefsky JM. Intensity-modulated radiation therapy: Supportive data for prostate cancer. Semin Radiat Oncol 2008; 18:48 –57. [37] Leborgne F, Fowler J. Acute toxicity after hypofractionated conformal radiotherapy for localized prostate cancer: Nonrandomized contemporary comparison with standard fractionation. Int J Radiat Oncol Biol Phys 2008;72:770 – 6. [38] 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;48:635– 42. [39] Teh BS, Mai WY, 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;49:705–12. [40] Nuyttens JJ, Milito S, Rust PF, et al. Dose-volume relationship for acute side effects during high dose conformal radiotherapy for prostate cancer. Radiother Oncol 2002;64:209 –14. [41] De Meerleer G, Vakaet L, Meersschout S, et al. Intensity modulated radiotherapy as primary treatment for prostate cancer: Acute toxicity in 114 patients. Int J Radiat Oncol Biol Phys 2004;60:777– 87. [42] Peeters ST, Heemsbergen WD, van Putten WL, et al. Acute and late complications after radiotherapy for prostate cancer: Results of a multicenter randomized trial comparing 68 Gy to 78 Gy. Int J Radiat Oncol Biol Phys 2005;61:1019 –34. [43] Pollack A, Zagars GK, Starkschall G, et al. Conventional vs. conformal radiotherapy for prostate cancer: Preliminary results of dosimetry and acute toxicity. Int J Radiat Oncol Biol Phys 1996; 34:555– 64. [44] Michalski JM, Purdy JA, Winter K, et al. Preliminary report of toxicity following 3D radiation therapy for prostate cancer on 3DOG/ RTOG 9406. Int J Radiat Oncol Biol Phys 2000;46:391– 402. [45] van Tol-Geerdink JJ, Stalmeier PF, van Lin EN, et al. Do patients with localized prostate cancer treatment really want more aggressive treatment? J Clin Oncol 2006;24:4581– 6.