- Email: [email protected]
Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial
Radiotherapy and Oncology xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
Original article
Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial James W. Denham a,⇑, Allison Steigler a, David Joseph b, David S. Lamb c, Nigel A. Spry b, Gillian Duchesne d, Chris Atkinson e, John Matthews f, Sandra Turner g, Lizbeth Kenny h, Keen-Hun Tai d, Nirdosh Kumar Gogna i, Suki Gill b, Hendrick Tan b, Rachel Kearvell b, Judy Murray c, Martin Ebert b, Annette Haworth d, Angel Kennedy d, Brett Delahunt j, Christopher Oldmeadow a,k, Elizabeth G. Holliday a,k, John Attia a,k a School of Medicine and Public Health, University of Newcastle; b Sir Charles Gairdner Hospital, Perth, Australia; c Department of Pathology and Molecular Medicine, University of Otago, Wellington, New Zealand; d Peter MacCallum Cancer Centre, Melbourne, Australia; e St Georges Cancer Care Centre, Christchurch; f Auckland Hospital, New Zealand; g Westmead Hospital, Sydney; h Royal Brisbane and Women’s Hospital; i Mater Radiation Oncology Centre, Princess Alexandra Hospital, Brisbane, Australia; j Wellington School of Medicine and Health Sciences, University of Otago, New Zealand; k Hunter Medical Research Institute, Newcastle, Australia
a r t i c l e
i n f o
Article history: Received 5 March 2015 Received in revised form 20 May 2015 Accepted 22 May 2015 Available online xxxx Keywords: Prostate cancer Dose escalation High dose rate brachytherapy boost Androgen suppression
a b s t r a c t Background: The relative effects of radiation dose escalation (RDE) and androgen suppression (AS) duration on local prostatic progression (LP) remain unclear. Methods: We addressed this in the TROG 03.04 RADAR trial by incorporating a RDE programme by stratification at randomisation. Men were allocated 6 or 18 months AS ± 18 months zoledronate (Z). The main endpoint was a composite of clinically diagnosed LP or PSA progression with a PSA doubling time P6 months. Fine and Gray competing risk modelling with adjustment for site clustering produced cumulative incidence estimates at 6.5 years for each RDE group. Results: Composite LP declined coherently in the 66, 70 and 74 Gy external beam dosing groups and was lowest in the high dose rate brachytherapy boost (HDRB) group. At 6.5 years, adjusted cumulative incidences were 22%, 15%, 13% and 7% respectively. Compared to 6 months AS, 18 months AS also significantly reduced LP (p < 0.001). Post-radiation urethral strictures were documented in 45 subjects and increased incrementally in the dosing groups. Crude incidences were 0.8%, 0.9%, 3.8% and 12.7% respectively. Conclusion: RDE and increasing AS independently reduce LP and increase urethral strictures. The risks and benefits to the individual must be balanced when selecting radiation dose and AS duration. Ó 2015 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology xxx (2015) xxx–xxx
Opinion remains divided on how much androgen suppression (AS) is necessary for men with locally advanced prostate cancer (LAPC) if they are treated with escalated radiation doses. Cancer centres in Australia and New Zealand were being re-equipped with the technologies necessary to enable radiation doses to be escalated when the RADAR randomised controlled trial for LAPC was being designed in the early 2000s. Using a 2 2 factorial design this trial sought to determine whether an additional year of AS or 18 months of zoledronate (Z) or both improved outcomes in men receiving six months of neo-adjuvant AS and radiotherapy (RT). Through stratification for increasing radiation dose, a structured dose escalation programme was built into the trial
design to enable centres to escalate doses and, as a result, to address the relative benefits of an additional year of AS, and 2 Gy fractional equivalent radiation dose increases of up to 20%. Dose escalation options included 66, 70 and 74 Gy in 2 Gy fractions using external beam (EBRT) alone, and 46 Gy delivered externally followed by a high dose rate brachytherapy boost of 19.5 Gy in 3 fractions over 24 h (HDRB). In this report we also address the relative benefits of the HDRB boost as a means of achieving dose escalation and examine the associated risk of urethral strictures. Methods Patients and treatment
⇑ Corresponding author at: University of Newcastle, Locked Bag 1, Hunter Region Mail Centre, NSW 2310, Australia. E-mail address: [email protected] (J.W. Denham).
Men with histologically confirmed adenocarcinoma of the prostate without lymph node or systemic metastases with primary tumours T stage 2b and above, or T stage 2a, Gleason score
http://dx.doi.org/10.1016/j.radonc.2015.05.016 0167-8140/Ó 2015 Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Denham JW et al. Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.05.016
2
Radiation dose escalation
(GS) P 7 and baseline PSA levels P10 ng/mL, were eligible to participate after providing informed consent. Details of the trial were presented in Lancet Oncology in 2012 [1]. All men received 6 months of leuprorelin (22.5 mg i.m. 3 monthly) commencing at randomisation, 5 months before RT to the prostate and seminal vesicles, excluding pelvic lymph nodes. Men in the control arm received no further treatment (i.e. short term AS [STAS]). Men in the other androgen suppression only treatment arm received an additional 12 months of leuprorelin (22.5 mg i.m. 3 monthly) (i.e. intermediate term AS [ITAS]). Men allocated to the two bisphosphonate treatment arms received Z 4 mgs i.v. every 3 months for 18 months starting at randomisation alone (STAS + Z) or with an additional 12 months of leuprorelin (ITAS + Z). In this report we describe the influence of radiation dose escalation ± the addition of 12 months of leuprorelin on local progression (LP). A regulated radiation dose escalation programme was achieved by requiring participating centres to select their preferred dosing options from a pre-determined range of doses and techniques. The dosing options were 66, 70 and 74 Gy using 2 Gy fractional increments to the ICRU point using external beam alone (EBRT only), and 46 Gy in 2 Gy fractions to the ICRU point using external beams followed by a high dose rate brachytherapy (HDRB) boost dose of 19.5 Gy using 3 fractions of 6.5 Gy. Brachytherapy dose was prescribed to the isodose encompassing the prostate gland and any identified extracapsular extensions. Dose constraints were 120% to the prostatic urethra and 70% to the anterior rectal wall. No more than 50% of the prescription volume should receive 150% of the prescribed dose. Full details of the methodology employed for dose escalation, derivation of radiation target volumes, dose volume histogram constraints and set up accuracy requirements are provided in our earlier report [2] and in the relevant portions of the RADAR protocol (reproduced in the Web Appendix of that report). Assignment to dose and technique was not a random process in the trial; however the stratification scheme employed ensured that radiation dose and technique used were balanced across trial arms. Consistency in EBRT dose delivery was assessed via dosimetric audit of centres [3–7] and adherence to trial planning protocol via expert review of each treatment plan [8]. Follow up After treatment, all patients were followed up in clinic every 3 months for 18 months, then 6 monthly up to 5 years post randomisation and then annually for a further 5 years. At each visit PSA levels, patient-reported and clinician-assessed outcomes were collected and clinical examinations including digital rectal examination were performed. Closeout date was 28 February 2014, 6.5 years after the last subject was randomised. Endpoints The primary endpoint of this study report is ‘‘composite’’ local progression. Secondary endpoints are ‘‘maximum likelihood’’ and ‘‘minimum likelihood’’ local progressions. Because there is no consensus in the field we have derived three potential surrogate endpoints for local progression (LP). Our ‘‘maximum likelihood estimate’’ (MaxLE) of LP endpoint was the widely used potential surrogate for LP, PSA progression. We regard it as a MaxLE because it can be triggered by distant progression in the presence or absence of synchronous or metachronous LP. We used the Phoenix method [9] (i.e., a PSA rise of 2 ng/mL above the post-treatment nadir). Our ‘‘minimum likelihood estimate’’ (MinLE) was clinically diagnosed LP i.e. a clinically obvious cancerous progression in or directly adjacent to the prostate recognised by digital rectal examination and/or by imaging techniques.
‘‘Composite LP’’ was based on a composite variable comprising either a clinically recognised LP (i.e. MinLE) or a PSA progression characterised by a PSA doubling time of P6 months. An event was deemed to occur at LP or 12 months after Phoenix failure. PSA doubling time was estimated using PSA values from Phoenix failure up to 12 months thereafter, or else from the PSA prior to Phoenix failure if insufficient values were available (e.g. due to early secondary therapy (STI)). Time to event was measured from randomisation.
Analyses In our main endpoints analyses we identified an interaction between the use of Z and GS of the primary tumour for all distant progression related endpoints [10]. Our first step in this analysis therefore was to determine whether this interaction influenced our potential LP surrogate endpoints. For the clinically diagnosed and composite LP endpoints no such interaction was detected. However for our PSA progression potential LP surrogate this interaction was present, presumably due to its high level of contamination by distant progression. We therefore used the entire dataset of 1051 evaluable cases to determine cumulative incidence estimates of the MinLE and composite LP endpoints, and the dataset of the 529 evaluable cases in the two trial arms where zoledronate was not administered for the cumulative incidence estimates of the MaxLE LP. Two sets of models were used to derive cumulative incidence failure plots for the three potential surrogate LP endpoints. The primary exposure was radiotherapy dose, described by a categorical variable with four levels: 66 Gy, 70 Gy, 74 Gy and HDRB boost. In addition to the primary exposure, covariates assessed in multivariate models included: treatment arm (STAS, STAS + Z, ITAS, ITAS + Z); patient age at randomisation; tumour stage (T2 vs T3/T4); GS (67 vs >7); and baseline PSA (620 vs >20). For the three outcomes, the corresponding competing risks (CR) were detailed as below: 1. MaxLE LP: CR = secondary therapeutic intervention (STI); and death. 2. MinLE LP: CR = distant progression alone >2 months before LP; STI; and death. 3. Composite LP: CR = distant progression alone >2 months before LP; PSA doubling time <6 months or >100 months after PSA progression; STI; and death. The effect of the exposure and covariates on the probability of LP in the presence of competing risks was modelled via multivariate proportional hazards models using the method of Fine and Gray [11]. For each outcome, two types of models were fitted, as alternative methods to account for potential clustering (correlation) of event times within recruitment sites. The first model used robust standard error estimation (population-averaged model) by specifying centre as a cluster-level variable. The second model included frailty as a random effect (conditional, or mixed model), i.e. intra-site correlation was assumed to result from a site-level random effect, or frailty. This frailty was assumed to have a multiplicative effect on the hazard function and to be generated by a gamma distribution with mean 1 and variance h, with the variance of the site-level frailties (h) being estimated. However if h was not significantly different from zero in models for all three LP endpoints, intra-site correlation was ignored and the population-averaged model was presented. Cumulative incidence functions were estimated and plotted for each radiotherapy dose, adjusted for all other covariates. Adjusted cumulative incidence estimates at 6.5 years after randomisation were used to derive dose response. Sub-hazard ratios and
Please cite this article in press as: Denham JW et al. Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.05.016
3
J.W. Denham et al. / Radiotherapy and Oncology xxx (2015) xxx–xxx
cumulative incidences for 66, 74 and HDRB were estimated using 70 Gy as the reference level. A dose–response curve comparison between subjects treated on the 6 months AS and 18 months AS trial arms was derived for the LP composite endpoint using separate models of these two treatment groups. Multiple logistic regression was used to estimate risk of post-radiation urethral stricture by radiation dosing group, trial arm and patient factors at randomisation, including age, body mass index and pre-existing lower urinary tract disorders. All analyses were performed on an intention-to-treat basis and a two-sided p value of <0.05 considered statistically significant for the primary and secondary endpoints. All statistical analyses were programmed using Stata v13.0. Results Of the 1071 trial subjects, 1051 subjects who received radiotherapy as specified in the trial protocol, were eligible for inclusion in these analyses. Baseline characteristics and trial arm assignment of the radiation dose escalation groups are presented in Table 1. Tumour characteristics were balanced evenly across trial arms and EBRT (only) dosing groups. Subjects selected for HDRB boost were younger and had higher T stage and GS tumours. Median follow up duration from randomisation was 7.4 years. Results have been reported at 6.5 years minimum follow up from randomisation, which is 5 years from the end of all trial protocol treatments. No interaction between the use of Z and GS for the composite LP endpoint was detected. This event occurred in 189 of 1051 patients (18%). Of these, 44 (23.3%) had distant progression without evidence of clinically recognised LP at any time point and their
Table 1 Baseline characteristics and trial arm assignment of the radiation dose escalation groups. 66 Gy (n = 125)
70 Gy (n = 427)
74 Gy (n = 262)
HDRB (n = 237)
Age Median IQR
68.6 (63.3–72.8)
69.5 (64.2–73.3)
69.9 (64.5–74.2)
66.3 (60.6–71.4)
Trial arm
No.
%
No.
%
No.
%
No.
%
STAS STAS + Z ITAS ITAS + Z
30 30 32 33
24.0 24.0 25.6 26.4
111 108 106 102
26.0 25.3 24.8 23.9
68 65 64 65
25.9 24.8 24.4 24.8
57 57 61 62
24.1 24.1 25.7 26.2
PSA <10 10–19.99 20–39.99 P40
38 52 24 11
30.4 41.6 19.2 8.8
120 179 89 39
28.1 41.9 20.8 9.1
68 97 61 36
25.9 37.0 23.3 13.7
60 92 57 28
25.3 38.8 24.1 11.8
Tstage T2 T3,4
94 31
75.2 24.8
311 116
72.8 27.2
179 83
68.3 31.7
85 152
35.9 64.1
Gleason score <7 7 >7
18 73 34
14.4 58.4 27.2
48 261 118
11.2 61.1 27.6
29 143 90
11.1 54.6 34.4
4 112 121
1.7 47.3 51.1
D’Amico risk groups Intermediate 32 High 93
25.6 74.4
98 329
23.0 77.1
58 204
22.1 77.9
19 218
8.0 92.0
Body mass index <25 P25 and <30 P30 Missing
25.6 44.0 26.4 4.0
97 209 110 11
22.7 48.9 25.8 2.6
59 127 64 12
22.5 48.5 24.4 4.6
52 123 60 2
21.9 51.9 25.3 0.8
Pre-existing lower urinary disorders and treatment No 122 97.6 409 95.8 247 Yes 3 2.4 18 4.2 15
94.3 5.7
228 9
96.2 3.8
32 55 33 5
composite LP may therefore be a false positive (i.e. represent potential contamination). Table 2 shows the results of the Fine and Gray model. Compared to the 427 subjects in the 70 Gy reference group, the 125 subjects who received 66 Gy had a significantly greater hazard of composite LP [SHR = 1.51 (95% CI: 1.15– 1.97) p = 0.003]. The 237 subjects who received the HDRB boost had a significantly lower hazard [SHR = 0.43 (95% CI: 0.33–0.56) p < 0.001]. The 262 subjects receiving 74 Gy had a lower but non-significant hazard [SHR = 0.82 (95% CI: 0.57–1.19) p = 0.29). The dose–response relationship for the composite LP endpoint is depicted in Fig. 1. At 6.5 years post-randomisation, adjusted cumulative incidences for 66 Gy, 70 Gy, 74 Gy and HDRB boost were 22%, 15%, 13% and 7% respectively. After adjusting for radiotherapy dose, the 525 subjects who received the additional year of AS in the ITAS arms also had a significantly lower hazard of the LP composite than the 526 who received 6 months AS on the STAS trial arms [SHR = 0.59 (95% CI: 0.49–0.72) p < 0.001]. The dose–response relationships for these two subgroups are shown in Fig. 2. Other patient risk factors predictive of the LP composite were decreasing age (SHR = 0.96 per year increase, p < 0.001), GS 8–10 (SHR = 1.60, p = 0.004) and PSA > 20 (SHR = 1.78, p < 0.001). No interaction was detected for MinLE of LP (clinically diagnosed LP) either. Only 42 of 1051 subjects (4.2%) experienced this endpoint. Compared to 70 Gy, the hazard for the 66 Gy group was significantly increased [SHR = 3.46 (95% CI: 1.85–6.50) p = 0.001] and for the HDRB group was significantly reduced [SHR = 0.46 (95% CI: 0.22–0.97) p = 0.041]. Subjects receiving an additional 12 months of AS in the ITAS arms had a reduced hazard of MinLE LP when compared to subjects who received 6 months AS on the STAS arms which did not reach significance [SHR = 0.53 (95% CI: 0.24–1.24) p = 0.145]. We have reported previously that PSA progression was influenced by a significant interaction between the use of Z and GS on the multiplicative scale [ratio of SHRs = 0.38 (95% CI: 0.17– 0.84) p = 0.016] [10]. Analysis of the MaxLE LP (PSA progression) endpoint was therefore confined to the 529 subjects not receiving Z on the STAS and ITAS arms. MaxLE LPs (PSA progressions) were documented in 182 (34.4%) subjects, of which 90 were due to distant progressions in the absence of clinical evidence of LP at any timepoint. The contamination rate due to these potential false positive MaxLE LPs was 48.9%. Compared to the 70 Gy reference group, the HDRB group had a significant reduction in MaxLE LPs [SHR = 0.67 (95% CI: 0.50–0.89) p = 0.005]. Compared to the 6 month AS group, 18 months AS significantly reduced MaxLE LP [SHR = 0.75 (95% CI: 0.60–0.94) p = 0.012]. The dose–response relationships for the minLE and maxLE LP endpoints are displayed in Fig. 1.
Table 2 Fine and Gray model of the influence of radiation dose, AS duration, age at randomisation, T stage, Gleason score and initial PSA on the cumulative incidence of the local progression composite endpoint.
Radiation dose 66 vs 70 Gy 74 vs 70 Gy HDRB vs 70 Gy 18mths AS vs 6 mths AS Age (years) Stage T3,4 vs T2 GS > 7 vs GS 6 7 PSA > 20 vs PSA 6 20
Sub hazard ratio
95% confidence intervals
pValue
1.51 0.82 0.43 0.59
1.15–1.97 0.57–1.19 0.33–0.56 0.49–0.72
0.003 0.29 <0.001 <0.001
0.96 1.15 1.60 1.78
0.95–0.98 0.83–1.61 1.16–2.21 1.46–2.16
<0.001 0.40 0.004 <0.001
Please cite this article in press as: Denham JW et al. Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.05.016
4
Radiation dose escalation
urethra was involved in 60% of cases. Multiple treatment procedures were required in at least 34% of strictures, and at least 59% have not been fully resolved by treatment. In only 34% of cases has a full resolution been documented to date. Significant risk factors for post-radiation strictures were HDRB boost (OR: 8.8, p < 0.001) and BMI P 30 (OR: 3.1, p = 0.021). A reduced risk was associated with the trial arm 6 months AS and 18 months Z (OR: 0.26, p = 0.043) but more data are needed to confirm the latter.
Discussion
Fig. 1. Cumulative incidence estimates in the four radiation dose escalation groups at 6.5 years post randomisation for three potential surrogate endpoints for local (prostatic) progression: 1 – PSA progression (the maximum likelihood estimate 5); 2 – local progression diagnosed clinically (the minimum likelihood estimate 4); and 3 – local progression or PSA doubling time P6 months (a reasonable likelihood estimate h). See the electronic supplement for comment regarding the 74 Gy dose group.
A total of 47 urethral strictures were reported of which 45 occurred post-radiation. Their features are described in Tables 3 and 4. Post-radiation stricture occurred much later in men receiving EBRT and those receiving only 6 months AS. Obstructive symptoms predominated with nocturia being most common (45%). Dysuria was a less common feature (22%). While HDRB strictures tended to cause frequency (50%), EBRT strictures tended to cause urgency (45%) and incontinence (30%). Many strictures were quite long and affected several sections of the urethra. The bulbar
LP is a difficult endpoint to assess. Serial MRI scans represent the most promising option [12], but this option is likely to remain beyond the means of most centres around the world for some time. It was therefore highly instructive to look at the behaviour of three different surrogate endpoints. We had anticipated some contamination of our PSA based potential LP surrogates by distant progression occurring in the absence of clinical evidence of LP because 80.3% of subjects in the study group had high risk ‘‘primary’’ cancers which often produce distant progressions. However the 48.9% potential contamination rate for the MaxLE of LP (PSA progression) endpoint was both unexpected and alarmingly high for such a commonly used potential surrogate endpoint. Fortunately contamination by distant progression is likely to be lower in studies involving men with low and intermediate risk cancers such as the on-going dose escalation trials [13–18]. In spite of this source of concern, our radiation dose–response relationships for all three endpoints for LP are gratifyingly coherent. Our composite LP endpoint was the most reliable of our potential surrogates due to its lower level of contamination by distant progression and its independence from the Z/GS interaction that has influenced all of our distant progression endpoints. This in turn enabled composite LP to be evaluated in our entire dataset. In summary, regardless of the surrogate used, dose escalation very clearly reduces LP. However reductions in distant progression or prostate cancer deaths have not been identified yet because we have had to restrict our analyses of these endpoints to the two AS arms due to the interaction between zoledronate and Gleason score. Moreover metastases originating as a result of failure to control the primary tumour may not have grown sufficiently to be detected yet [6].
Adjusted cumulative incidence (%) Composite local progression
35 30 25 20 15 10 5 0 66
70
74
HDRB
Radiation dose escalation group (Gy) 6 months AS
18 months AS
Fig. 2. A dose–response comparison between radiation dose escalation subgroups of men treated with 6 months neo-adjuvant AS versus 6 months neo-adjuvant AS followed by 12 months adjuvant AS (i.e. 18 months in total) at 6.5 years post randomisation. The error bars represent standard errors. See the electronic supplement for comment regarding the 74 Gy group.
Please cite this article in press as: Denham JW et al. Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.05.016
5
J.W. Denham et al. / Radiotherapy and Oncology xxx (2015) xxx–xxx Table 3 Clinical features of the 47 urethral strictures documented, procedures undertaken to relieve them, and outcomes to date. N
%
2 0 45 47
4.3 0.0 95.7 100.0
Median
Range
Years after EBRT (n = 12)
3.6
Years after HDRB (n = 28) Missing
1.2 3
(0.6– 8.6) (0–4.8) 6.7
Onset of symptoms Before radiotherapy During radiotherapy After radiotherapy Total Timing of diagnosis (post radiotherapy strictures only)
Site of stricture (NB multiple sites involved in some cases) Prostatic 7 Membranous 17 Bulbar 33 Missing 4
14.9 36.2 70.2 8.5
Severity of stricture Minor, easily dilated Moderate, urethrotomy Tight or impassable Missing
18 8 12 9
38.3 17.0 25.5 19.1
Number of procedures necessary One Two Multiple Missing
20 9 16 2
42.5 19.2 34.0 4.3
Types of procedures Dilatation (±catheterisation) Internal urethrotomy (±catheterisation) Urethroplasty TUR Bladder neck incision Other procedures Missing
19 23 1 0 0 2 2
40.4 48.9 2.1 0.0 0.0 4.3 4.3
Outcomes of procedures Complete resolution Asymptomatic but continuing self-catheterisation Partial resolution and may require further procedures Unresolvable permanent in dwelling catheter Unknown Missing
16 9 16 3 2 1
34.0 19.2 34.0 6.4 4.3 2.1
noted that urethral strictures had been diagnosed in men receiving the HDRB boost. In this report we have documented the development of urethral strictures in 47 subjects. It is important to recognise that our data are limited by the fact that this treatment related complication was not anticipated a priori, and as a result stricture data were not collected prospectively. Stricturing is almost certainly more frequent than we have reported. Men treated with EBRT alone tended to have their strictures later than those receiving the HDRB boost. Since stricturing is less likely to be investigated and treated in elderly patients, the dose–response relationship may be less clear cut than our data suggest. Catheter slippage has been implicated as a cause for stricture following HDRB, however our quality control data suggest that unintended urethral coverage by EBRT may also have been an important contributor. EBRT treatment was planned using CT alone which is well recognised to lead to coverage of the membranous and sometimes bulbar portions of the urethra because the prostatic apex is not well visualised using CT. It is important to note that a similar result emerged from the Canadian ASCENDE-RT⁄ trial for men with unfavourable risk prostate cancer recently reported orally by Morris and colleagues. [22] In this randomised trial all 400 subjects received 8 months neo-adjuvant AS and pelvic EBRT to 46 Gy using 23 2 Gy fractions. In the control arm a conformal EBRT boost of 32 Gy using 16 Gy fractions followed. In the experimental arm a low dose rate brachytherapy (LDRB) boost prescribed to a minimum peripheral dose of 116 Gy occurred afterwards. The hazard of PSA progression was significantly reduced in the LDRB arm (0.45 [95% CI: 0.29– 0.77], p = 0.002) but Grade 3 urinary toxicity was significantly increased in the brachytherapy boost arm (p < 0.0001). [23] Urethral strictures were responsible for approximately half of the toxic events in the LDRB trial arm and, of course, catheter slippage would not have been a contributing factor to their development. It is hoped that recent technical innovations, including MRI fusion in the planning of treatment and fiducial marker guidance, will successfully prevent unnecessary coverage of the urethra and its vascular supply and, in doing so, reduce urethral strictures. The price of reducing LP using an additional year of AS was not negligible either. ‘‘Hormone treatment related symptoms’’ as defined in the EORTC PR25 instrument remained significantly
Table 4 Risk factors for post-radiation urethral strictures.
Therefore we intend to revisit this issue in 2017 when 10 year follow up data will be available for analysis. Our data also provide one of the first demonstrations that an additional year of AS can shift the radiation dose–response relationship favourably. In fact it reduced the hazard of our composite LP endpoint by approximately 40%, after adjusting for radiation dose. In this portion of the dose–response curve this equates to approximately 8 Gy using 2 Gy fractions. It must be pointed out that there were too few participants in the RADAR trial with intermediate risk cancers to know whether this degree of benefit is experienced by them. Some authorities will therefore continue to be doubtful that AS is necessary for men with intermediate risk cancers [19–21]. It is also gratifying to note that the benefits of radiation dose escalation and one additional year of AS have been achieved at the cost of modest increases in treatment related morbidity. In our 2014 main endpoints report [10] we had the opportunity to provide an update on our rectal and urinary dysfunction symptomatology report published in this journal in 2012 [2] as well as quality of life indices reported in the same year [1]. It was pleasing to report that at 6.5 years of follow-up, differences between dosing groups remained absent. However this does not mean that the price of radiation was negligible in the RADAR trial. In our 2012 report we
Dosing group 66 Gy 70 Gy 74 Gy HDRB Total
Total 125 425 262 237 1049
Stricture 1 4 10 30 45
% 0.8 0.9 3.8 12.7 4.3
Risk factors for stricture (multivariate logistic regression model) Factor OR 95% confidence intervals
pValue
Radiation dose EBRT HDRB boost
1.00 8.82
– 4.52–17.18
– <0.001
Treatment arm 6 mths AS 6 mths AS + 18 mths Z 18 mths AS 18 mths AS + 18 mths Z
1.00 0.26 1.28 1.60
– 0.07–0.96 0.56–2.95 0.70–3.63
– 0.043 0.56 0.26
Age (years)
1.04
0.99–1.09
0.08
BMI <25 P25 and <30 P30
1 1.47 3.05
0.59–3.65 1.19–7.82
0.40 0.021
2.52
0.79–8.06
0.12
Pre-existing lower urinary disorders and treatment
Please cite this article in press as: Denham JW et al. Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.05.016
6
Radiation dose escalation
worse in subjects receiving an additional year of AS for almost 3 years after randomisation. Our RADAR Survivorship Substudy, which will measure the legacies of protocol treatment 10 years after randomisation, will be helpful in determining what long-term price will be paid by our subjects for radiation and AS induced morbidity. Our appraisal of the value of a HDRB boost as a means of achieving dose escalation is limited by the fact that radiation dose was not allocated to dosing group by randomisation. Consequently younger subjects with high stage and grade tumours at centres with HDRB equipment were more likely to receive the HDRB boost. Although the HDRB boost emerged as the most successful means of preventing all three measures of LP, the adverse selection biases make it impossible to be sure what the HDRB regimen is equivalent to using a 2 Gy per fraction EBRT regimen, and therefore where it fits on the 2 Gy per fraction dose–response curve. A well designed randomised controlled trial such as the British trial run by Hoskin et al. is necessary to provide an appropriate estimate [24]. In the meantime the less accurate estimates may come from the ASCENDE-RT⁄ trial (Morris et al.) and the carefully matched comparison of outcomes between men treated with the HDRB boost regimen that was subsequently used in the RADAR trial, and men treated by EBRT alone using 74 Gy in 37 daily 2 Gy fractions at the Peter McCallum Cancer Institute, Melbourne [25]. In conclusion, the RADAR trial provides evidence that in men with high risk LAPC both radiation dose escalation and longer AS reduce LP regardless of how it is defined. The RADAR trial data also indicate that the advantages of both approaches are offset by increased morbidity. Optimal outcomes will be achieved by judicious combinations that take into account the patient’s life expectancy and fitness to tolerate the expected complications of treatment. Trial number The TROG 03.04 Trial is registered with the National Institutes of Health Clinical Trials Registry, number NCT00193856 and has approval from Hunter New England Human Research Ethics Committee (Trial ID. 03/06/11/3.02). Role of funding source The Goodfellow Foundation, Auckland (New Zealand); Cancer Standards Institute of New Zealand; National Health and Medical Research Council of Australia (NHMRC) (Project Application ID 300705 and 455521); Novartis Pharmaceuticals Australia; Abbott Pharmaceuticals Australia. Conflict of interest Nigel Spry: Consultant or advisory relationship (compensated) with Abbott Pharmaceuticals, Honoraria from Abbott Pharmaceuticals; Research Funding from Abbott Pharmaceuticals. All other authors have no conflicts of interest to disclose. Acknowledgements We thank Rosemary Bradford once again for her expert preparation of the manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.radonc.2015.05. 016.
References [1] Denham JW, Wilcox C, Joseph D, Spry NA, Lamb DS, Tai KH, et al. Quality of life in men with locally advanced prostate cancer treated with leuprorelin and radiotherapy with or without zoledronic acid (TROG 03.04 RADAR): secondary endpoints from a randomised phase 3 factorial trial. Lancet Oncol 2012;13:1260–70. [2] Denham JW, Wilcox C, Lamb DS, Spry NA, Duchesne G, Atkinson C, et al. Rectal and urinary dysfunction in the TROG 03.04 RADAR trial for locally advanced prostate cancer. Radiother Oncol 2012;105:184–92. [3] Ebert MA, Harrison KM, Cornes D, Howlett SJ, Joseph DJ, Kron T, et al. Comprehensive Australasian multicentre dosimetric intercomparison: issues, logistics and recommendations. J Med Imaging Radiat Oncol 2009;53:119–31. [4] Ebert MA, Harrison KM, Howlett SJ, Cornes D, Bulsara M, Hamilton CS, et al. Dosimetric intercomparison for multicentre clinical trials using a patientbased anatomic pelvic phantom. Med Phys 2011;38:5167–75. [5] Harrison KM, Ebert MA, Kron T, Howlett SJ, Cornes D, Hamilton CS, et al. Design, manufacture, and evaluation of an anthropomorphic pelvic phantom purpose-built for radiotherapy dosimetric intercomparison. Med Phys 2011;38:5330–7. [6] Denham JW, Steigler A, Tai KH, Joseph D, Matthews J, Atkinson C, et al. Paradoxical metastatic progression following 3 months of neo-adjuvant androgen suppression in the TROG 96.01 trial for men with locally advanced prostate cancer. Radiother Oncol 2013;107:123–8. [7] Kearvell R, Haworth A, Ebert MA, Murray J, Hooton B, Richardson S, et al. Quality improvements in prostate radiotherapy: outcomes and impact of comprehensive quality assurance during the TROG 03.04 ‘RADAR’ trial. J Med Imaging Radiat Oncol 2013;57:247–57. [8] Ebert MA, Haworth A, Kearvell R, Hooton B, Coleman R, Spry N, et al. Detailed review and analysis of complex radiotherapy clinical trial planning data: evaluation and initial experience with the SWAN software system. Radiother Oncol 2008;86:200–10. [9] Roach III M, Hanks G, Thames Jr H, Schellhammer PF, Shipley WU, Sokol GH, et al. Defining biochemical failure following radiotherapy with or without hormonal therapy in men with clinically localized prostate cancer: recommendations of the RTOG-ASTRO Phoenix Consensus Conference. Int J Radiat Oncol Biol Phys 2006;65:965–74. [10] Denham JW, Joseph D, Lamb DS, Spry NA, Duchesne G, Matthews J, et al. Shortterm androgen suppression and radiotherapy versus intermediate-term androgen suppression and radiotherapy, with or without zoledronic acid, in men with locally advanced prostate cancer (TROG 03.04 RADAR): an openlabel, randomised, phase 3 factorial trial. Lancet Oncol 2014;15:1076–89. [11] Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Statist Assoc 1999;94:496–509. [12] Haider MA, Chung P, Sweet J, Toi A, Jhaveri K, Menard C, et al. Dynamic contract-enhanced magnetic resonance imaging for localization of recurrent prostate cancer after external beam radiotherapy. Int J Radiat Oncol Biol Phys 2008;70:425–30. [13] Kuban DA, Levy LB, Cheung MR, Lee AK, Choi S, Frank S, et al. Long-term failure patterns and survival in a randomized dose-escalation trial for prostate cancer. Who dies of disease? Int J Radiat Oncol Biol Phys 2011;79:1310–7. [14] Zietman AL, DeSilvio M, Slater JD, Rossi C, Miller DW, Adams JA, et al. Comparison of convention-dose vs high-dose conformal radiation therapy in clinically localized adenocarcinoma of the prostate. JAMA 2005;294: 1233–9. [15] Beckendorf V, Guerif S, Le Prise E, Cosset JM, LeFloch O, Chauvet B, et al. The GETUG 70 Gy vs. 80 Gy randomized trial for localized prostate cancer: feasibility and acute toxicity. Int J Radiat Oncol Biol Phys 2004;60:1056–65. [16] Peeters ST, Heemsbergen WD, Koper PC, van Putten WL, Slot A, Dielwart MF, et al. Dose-response in radiotherapy for localized prostate cancer: results of the Dutch multicenter randomized phase III trial comparing 68 Gy of radiotherapy with 78 Gy. J Clin Oncol 2006;24:1990–6. [17] Dearnaley DP, Sydes MR, Graham JD, Aird EG, Bottomley D, Cowan RA, et al. Escalated-dose versus standard-dose conformal radiotherapy in prostate cancer: first results from the MRC RT01 randomised controlled trial. Lancet Oncol 2007;8:475–87. [18] Shipley WU, Verhey LJ, Munzenrider JE, Suit HD, Urie MM, McManus PL, et al. Advanced prostate cancer: the results of a randomized comparative trial of high dose irradiation boosting with conformal protons compared with conventional dose irradiation using photons alone. Int J Radiat Oncol Biol Phys 1995;32:3–12. [19] Zelefsky MJ, Pei X, Chou JF, Schechter M, Kollmeier M, Cox B, et al. Dose escalation for prostate cancer radiotherapy: predictors of long-term biochemical tumour control and distant metastases-free survival outcomes. Eur Urol 2011;60:1133–9. [20] Al-Mamgani A, Heemsbergen WD, Levendag PC, Lebesque JV. Subgroup analysis of patients with localized prostate cancer treated within the Dutchrandomized dose escalation trial. Radiother Oncol 2010;96:13–8. [21] Smith MJ, Akhtar NH, Tagawa ST. The current role of androgen deprivation in patients undergoing dose-escalated external beam radiation therapy for clinically localized prostate cancer. Prostate Cancer. 2012;2012, Article ID 280278. [22] Morris W.J., Tyldesley S., Pai H.H., Halperin R., McKenzie M.R., Duncan G., et al. ASCENDE-RT⁄: A multicenter, randomized trial of dose-escalated external beam radiation therapy (EBRT-B) versus low-dose-rate brachytherapy (LDR-B)
Please cite this article in press as: Denham JW et al. Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.05.016
J.W. Denham et al. / Radiotherapy and Oncology xxx (2015) xxx–xxx for men with unfavorable-risk localized prostate cancer. J Clin Oncol 33. 2015(suppl 7; abstr 3). [23] Rodda S., Tyldesley S., Morris W. GU and GI toxicity in ASCENDE-RT⁄: a multicentre randomized trial of dose-escalated radiation for prostate cancer. 3rd ESTRO Forum (Abstract PD-0047); Barcelona 2015. [24] Hoskin PJ, Rojas AM, Bownes PJ, Lowe GJ, Ostler PJ, Bryant L. Randomised trial of external beam radiotherapy alone or combined with high-dose-rate
7
brachytherapy boost for localised prostate cancer. Radiother Oncol 2012;103:217–22. [25] Khor R, Duchesne G, Tai KH, Foroudi F, Chander S, Van Dyk S, et al. Direct 2arm comparison shows benefit of high-dose-rate brachytherapy boost vs external beam radiation therapy alone for prostate cancer. Int J Radiat Oncol Biol Phys 2013;85:679–85.
Please cite this article in press as: Denham JW et al. Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.05.016
Contents lists available at ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
Original article
Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial James W. Denham a,⇑, Allison Steigler a, David Joseph b, David S. Lamb c, Nigel A. Spry b, Gillian Duchesne d, Chris Atkinson e, John Matthews f, Sandra Turner g, Lizbeth Kenny h, Keen-Hun Tai d, Nirdosh Kumar Gogna i, Suki Gill b, Hendrick Tan b, Rachel Kearvell b, Judy Murray c, Martin Ebert b, Annette Haworth d, Angel Kennedy d, Brett Delahunt j, Christopher Oldmeadow a,k, Elizabeth G. Holliday a,k, John Attia a,k a School of Medicine and Public Health, University of Newcastle; b Sir Charles Gairdner Hospital, Perth, Australia; c Department of Pathology and Molecular Medicine, University of Otago, Wellington, New Zealand; d Peter MacCallum Cancer Centre, Melbourne, Australia; e St Georges Cancer Care Centre, Christchurch; f Auckland Hospital, New Zealand; g Westmead Hospital, Sydney; h Royal Brisbane and Women’s Hospital; i Mater Radiation Oncology Centre, Princess Alexandra Hospital, Brisbane, Australia; j Wellington School of Medicine and Health Sciences, University of Otago, New Zealand; k Hunter Medical Research Institute, Newcastle, Australia
a r t i c l e
i n f o
Article history: Received 5 March 2015 Received in revised form 20 May 2015 Accepted 22 May 2015 Available online xxxx Keywords: Prostate cancer Dose escalation High dose rate brachytherapy boost Androgen suppression
a b s t r a c t Background: The relative effects of radiation dose escalation (RDE) and androgen suppression (AS) duration on local prostatic progression (LP) remain unclear. Methods: We addressed this in the TROG 03.04 RADAR trial by incorporating a RDE programme by stratification at randomisation. Men were allocated 6 or 18 months AS ± 18 months zoledronate (Z). The main endpoint was a composite of clinically diagnosed LP or PSA progression with a PSA doubling time P6 months. Fine and Gray competing risk modelling with adjustment for site clustering produced cumulative incidence estimates at 6.5 years for each RDE group. Results: Composite LP declined coherently in the 66, 70 and 74 Gy external beam dosing groups and was lowest in the high dose rate brachytherapy boost (HDRB) group. At 6.5 years, adjusted cumulative incidences were 22%, 15%, 13% and 7% respectively. Compared to 6 months AS, 18 months AS also significantly reduced LP (p < 0.001). Post-radiation urethral strictures were documented in 45 subjects and increased incrementally in the dosing groups. Crude incidences were 0.8%, 0.9%, 3.8% and 12.7% respectively. Conclusion: RDE and increasing AS independently reduce LP and increase urethral strictures. The risks and benefits to the individual must be balanced when selecting radiation dose and AS duration. Ó 2015 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology xxx (2015) xxx–xxx
Opinion remains divided on how much androgen suppression (AS) is necessary for men with locally advanced prostate cancer (LAPC) if they are treated with escalated radiation doses. Cancer centres in Australia and New Zealand were being re-equipped with the technologies necessary to enable radiation doses to be escalated when the RADAR randomised controlled trial for LAPC was being designed in the early 2000s. Using a 2 2 factorial design this trial sought to determine whether an additional year of AS or 18 months of zoledronate (Z) or both improved outcomes in men receiving six months of neo-adjuvant AS and radiotherapy (RT). Through stratification for increasing radiation dose, a structured dose escalation programme was built into the trial
design to enable centres to escalate doses and, as a result, to address the relative benefits of an additional year of AS, and 2 Gy fractional equivalent radiation dose increases of up to 20%. Dose escalation options included 66, 70 and 74 Gy in 2 Gy fractions using external beam (EBRT) alone, and 46 Gy delivered externally followed by a high dose rate brachytherapy boost of 19.5 Gy in 3 fractions over 24 h (HDRB). In this report we also address the relative benefits of the HDRB boost as a means of achieving dose escalation and examine the associated risk of urethral strictures. Methods Patients and treatment
⇑ Corresponding author at: University of Newcastle, Locked Bag 1, Hunter Region Mail Centre, NSW 2310, Australia. E-mail address: [email protected] (J.W. Denham).
Men with histologically confirmed adenocarcinoma of the prostate without lymph node or systemic metastases with primary tumours T stage 2b and above, or T stage 2a, Gleason score
http://dx.doi.org/10.1016/j.radonc.2015.05.016 0167-8140/Ó 2015 Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Denham JW et al. Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.05.016
2
Radiation dose escalation
(GS) P 7 and baseline PSA levels P10 ng/mL, were eligible to participate after providing informed consent. Details of the trial were presented in Lancet Oncology in 2012 [1]. All men received 6 months of leuprorelin (22.5 mg i.m. 3 monthly) commencing at randomisation, 5 months before RT to the prostate and seminal vesicles, excluding pelvic lymph nodes. Men in the control arm received no further treatment (i.e. short term AS [STAS]). Men in the other androgen suppression only treatment arm received an additional 12 months of leuprorelin (22.5 mg i.m. 3 monthly) (i.e. intermediate term AS [ITAS]). Men allocated to the two bisphosphonate treatment arms received Z 4 mgs i.v. every 3 months for 18 months starting at randomisation alone (STAS + Z) or with an additional 12 months of leuprorelin (ITAS + Z). In this report we describe the influence of radiation dose escalation ± the addition of 12 months of leuprorelin on local progression (LP). A regulated radiation dose escalation programme was achieved by requiring participating centres to select their preferred dosing options from a pre-determined range of doses and techniques. The dosing options were 66, 70 and 74 Gy using 2 Gy fractional increments to the ICRU point using external beam alone (EBRT only), and 46 Gy in 2 Gy fractions to the ICRU point using external beams followed by a high dose rate brachytherapy (HDRB) boost dose of 19.5 Gy using 3 fractions of 6.5 Gy. Brachytherapy dose was prescribed to the isodose encompassing the prostate gland and any identified extracapsular extensions. Dose constraints were 120% to the prostatic urethra and 70% to the anterior rectal wall. No more than 50% of the prescription volume should receive 150% of the prescribed dose. Full details of the methodology employed for dose escalation, derivation of radiation target volumes, dose volume histogram constraints and set up accuracy requirements are provided in our earlier report [2] and in the relevant portions of the RADAR protocol (reproduced in the Web Appendix of that report). Assignment to dose and technique was not a random process in the trial; however the stratification scheme employed ensured that radiation dose and technique used were balanced across trial arms. Consistency in EBRT dose delivery was assessed via dosimetric audit of centres [3–7] and adherence to trial planning protocol via expert review of each treatment plan [8]. Follow up After treatment, all patients were followed up in clinic every 3 months for 18 months, then 6 monthly up to 5 years post randomisation and then annually for a further 5 years. At each visit PSA levels, patient-reported and clinician-assessed outcomes were collected and clinical examinations including digital rectal examination were performed. Closeout date was 28 February 2014, 6.5 years after the last subject was randomised. Endpoints The primary endpoint of this study report is ‘‘composite’’ local progression. Secondary endpoints are ‘‘maximum likelihood’’ and ‘‘minimum likelihood’’ local progressions. Because there is no consensus in the field we have derived three potential surrogate endpoints for local progression (LP). Our ‘‘maximum likelihood estimate’’ (MaxLE) of LP endpoint was the widely used potential surrogate for LP, PSA progression. We regard it as a MaxLE because it can be triggered by distant progression in the presence or absence of synchronous or metachronous LP. We used the Phoenix method [9] (i.e., a PSA rise of 2 ng/mL above the post-treatment nadir). Our ‘‘minimum likelihood estimate’’ (MinLE) was clinically diagnosed LP i.e. a clinically obvious cancerous progression in or directly adjacent to the prostate recognised by digital rectal examination and/or by imaging techniques.
‘‘Composite LP’’ was based on a composite variable comprising either a clinically recognised LP (i.e. MinLE) or a PSA progression characterised by a PSA doubling time of P6 months. An event was deemed to occur at LP or 12 months after Phoenix failure. PSA doubling time was estimated using PSA values from Phoenix failure up to 12 months thereafter, or else from the PSA prior to Phoenix failure if insufficient values were available (e.g. due to early secondary therapy (STI)). Time to event was measured from randomisation.
Analyses In our main endpoints analyses we identified an interaction between the use of Z and GS of the primary tumour for all distant progression related endpoints [10]. Our first step in this analysis therefore was to determine whether this interaction influenced our potential LP surrogate endpoints. For the clinically diagnosed and composite LP endpoints no such interaction was detected. However for our PSA progression potential LP surrogate this interaction was present, presumably due to its high level of contamination by distant progression. We therefore used the entire dataset of 1051 evaluable cases to determine cumulative incidence estimates of the MinLE and composite LP endpoints, and the dataset of the 529 evaluable cases in the two trial arms where zoledronate was not administered for the cumulative incidence estimates of the MaxLE LP. Two sets of models were used to derive cumulative incidence failure plots for the three potential surrogate LP endpoints. The primary exposure was radiotherapy dose, described by a categorical variable with four levels: 66 Gy, 70 Gy, 74 Gy and HDRB boost. In addition to the primary exposure, covariates assessed in multivariate models included: treatment arm (STAS, STAS + Z, ITAS, ITAS + Z); patient age at randomisation; tumour stage (T2 vs T3/T4); GS (67 vs >7); and baseline PSA (620 vs >20). For the three outcomes, the corresponding competing risks (CR) were detailed as below: 1. MaxLE LP: CR = secondary therapeutic intervention (STI); and death. 2. MinLE LP: CR = distant progression alone >2 months before LP; STI; and death. 3. Composite LP: CR = distant progression alone >2 months before LP; PSA doubling time <6 months or >100 months after PSA progression; STI; and death. The effect of the exposure and covariates on the probability of LP in the presence of competing risks was modelled via multivariate proportional hazards models using the method of Fine and Gray [11]. For each outcome, two types of models were fitted, as alternative methods to account for potential clustering (correlation) of event times within recruitment sites. The first model used robust standard error estimation (population-averaged model) by specifying centre as a cluster-level variable. The second model included frailty as a random effect (conditional, or mixed model), i.e. intra-site correlation was assumed to result from a site-level random effect, or frailty. This frailty was assumed to have a multiplicative effect on the hazard function and to be generated by a gamma distribution with mean 1 and variance h, with the variance of the site-level frailties (h) being estimated. However if h was not significantly different from zero in models for all three LP endpoints, intra-site correlation was ignored and the population-averaged model was presented. Cumulative incidence functions were estimated and plotted for each radiotherapy dose, adjusted for all other covariates. Adjusted cumulative incidence estimates at 6.5 years after randomisation were used to derive dose response. Sub-hazard ratios and
Please cite this article in press as: Denham JW et al. Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.05.016
3
J.W. Denham et al. / Radiotherapy and Oncology xxx (2015) xxx–xxx
cumulative incidences for 66, 74 and HDRB were estimated using 70 Gy as the reference level. A dose–response curve comparison between subjects treated on the 6 months AS and 18 months AS trial arms was derived for the LP composite endpoint using separate models of these two treatment groups. Multiple logistic regression was used to estimate risk of post-radiation urethral stricture by radiation dosing group, trial arm and patient factors at randomisation, including age, body mass index and pre-existing lower urinary tract disorders. All analyses were performed on an intention-to-treat basis and a two-sided p value of <0.05 considered statistically significant for the primary and secondary endpoints. All statistical analyses were programmed using Stata v13.0. Results Of the 1071 trial subjects, 1051 subjects who received radiotherapy as specified in the trial protocol, were eligible for inclusion in these analyses. Baseline characteristics and trial arm assignment of the radiation dose escalation groups are presented in Table 1. Tumour characteristics were balanced evenly across trial arms and EBRT (only) dosing groups. Subjects selected for HDRB boost were younger and had higher T stage and GS tumours. Median follow up duration from randomisation was 7.4 years. Results have been reported at 6.5 years minimum follow up from randomisation, which is 5 years from the end of all trial protocol treatments. No interaction between the use of Z and GS for the composite LP endpoint was detected. This event occurred in 189 of 1051 patients (18%). Of these, 44 (23.3%) had distant progression without evidence of clinically recognised LP at any time point and their
Table 1 Baseline characteristics and trial arm assignment of the radiation dose escalation groups. 66 Gy (n = 125)
70 Gy (n = 427)
74 Gy (n = 262)
HDRB (n = 237)
Age Median IQR
68.6 (63.3–72.8)
69.5 (64.2–73.3)
69.9 (64.5–74.2)
66.3 (60.6–71.4)
Trial arm
No.
%
No.
%
No.
%
No.
%
STAS STAS + Z ITAS ITAS + Z
30 30 32 33
24.0 24.0 25.6 26.4
111 108 106 102
26.0 25.3 24.8 23.9
68 65 64 65
25.9 24.8 24.4 24.8
57 57 61 62
24.1 24.1 25.7 26.2
PSA <10 10–19.99 20–39.99 P40
38 52 24 11
30.4 41.6 19.2 8.8
120 179 89 39
28.1 41.9 20.8 9.1
68 97 61 36
25.9 37.0 23.3 13.7
60 92 57 28
25.3 38.8 24.1 11.8
Tstage T2 T3,4
94 31
75.2 24.8
311 116
72.8 27.2
179 83
68.3 31.7
85 152
35.9 64.1
Gleason score <7 7 >7
18 73 34
14.4 58.4 27.2
48 261 118
11.2 61.1 27.6
29 143 90
11.1 54.6 34.4
4 112 121
1.7 47.3 51.1
D’Amico risk groups Intermediate 32 High 93
25.6 74.4
98 329
23.0 77.1
58 204
22.1 77.9
19 218
8.0 92.0
Body mass index <25 P25 and <30 P30 Missing
25.6 44.0 26.4 4.0
97 209 110 11
22.7 48.9 25.8 2.6
59 127 64 12
22.5 48.5 24.4 4.6
52 123 60 2
21.9 51.9 25.3 0.8
Pre-existing lower urinary disorders and treatment No 122 97.6 409 95.8 247 Yes 3 2.4 18 4.2 15
94.3 5.7
228 9
96.2 3.8
32 55 33 5
composite LP may therefore be a false positive (i.e. represent potential contamination). Table 2 shows the results of the Fine and Gray model. Compared to the 427 subjects in the 70 Gy reference group, the 125 subjects who received 66 Gy had a significantly greater hazard of composite LP [SHR = 1.51 (95% CI: 1.15– 1.97) p = 0.003]. The 237 subjects who received the HDRB boost had a significantly lower hazard [SHR = 0.43 (95% CI: 0.33–0.56) p < 0.001]. The 262 subjects receiving 74 Gy had a lower but non-significant hazard [SHR = 0.82 (95% CI: 0.57–1.19) p = 0.29). The dose–response relationship for the composite LP endpoint is depicted in Fig. 1. At 6.5 years post-randomisation, adjusted cumulative incidences for 66 Gy, 70 Gy, 74 Gy and HDRB boost were 22%, 15%, 13% and 7% respectively. After adjusting for radiotherapy dose, the 525 subjects who received the additional year of AS in the ITAS arms also had a significantly lower hazard of the LP composite than the 526 who received 6 months AS on the STAS trial arms [SHR = 0.59 (95% CI: 0.49–0.72) p < 0.001]. The dose–response relationships for these two subgroups are shown in Fig. 2. Other patient risk factors predictive of the LP composite were decreasing age (SHR = 0.96 per year increase, p < 0.001), GS 8–10 (SHR = 1.60, p = 0.004) and PSA > 20 (SHR = 1.78, p < 0.001). No interaction was detected for MinLE of LP (clinically diagnosed LP) either. Only 42 of 1051 subjects (4.2%) experienced this endpoint. Compared to 70 Gy, the hazard for the 66 Gy group was significantly increased [SHR = 3.46 (95% CI: 1.85–6.50) p = 0.001] and for the HDRB group was significantly reduced [SHR = 0.46 (95% CI: 0.22–0.97) p = 0.041]. Subjects receiving an additional 12 months of AS in the ITAS arms had a reduced hazard of MinLE LP when compared to subjects who received 6 months AS on the STAS arms which did not reach significance [SHR = 0.53 (95% CI: 0.24–1.24) p = 0.145]. We have reported previously that PSA progression was influenced by a significant interaction between the use of Z and GS on the multiplicative scale [ratio of SHRs = 0.38 (95% CI: 0.17– 0.84) p = 0.016] [10]. Analysis of the MaxLE LP (PSA progression) endpoint was therefore confined to the 529 subjects not receiving Z on the STAS and ITAS arms. MaxLE LPs (PSA progressions) were documented in 182 (34.4%) subjects, of which 90 were due to distant progressions in the absence of clinical evidence of LP at any timepoint. The contamination rate due to these potential false positive MaxLE LPs was 48.9%. Compared to the 70 Gy reference group, the HDRB group had a significant reduction in MaxLE LPs [SHR = 0.67 (95% CI: 0.50–0.89) p = 0.005]. Compared to the 6 month AS group, 18 months AS significantly reduced MaxLE LP [SHR = 0.75 (95% CI: 0.60–0.94) p = 0.012]. The dose–response relationships for the minLE and maxLE LP endpoints are displayed in Fig. 1.
Table 2 Fine and Gray model of the influence of radiation dose, AS duration, age at randomisation, T stage, Gleason score and initial PSA on the cumulative incidence of the local progression composite endpoint.
Radiation dose 66 vs 70 Gy 74 vs 70 Gy HDRB vs 70 Gy 18mths AS vs 6 mths AS Age (years) Stage T3,4 vs T2 GS > 7 vs GS 6 7 PSA > 20 vs PSA 6 20
Sub hazard ratio
95% confidence intervals
pValue
1.51 0.82 0.43 0.59
1.15–1.97 0.57–1.19 0.33–0.56 0.49–0.72
0.003 0.29 <0.001 <0.001
0.96 1.15 1.60 1.78
0.95–0.98 0.83–1.61 1.16–2.21 1.46–2.16
<0.001 0.40 0.004 <0.001
Please cite this article in press as: Denham JW et al. Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.05.016
4
Radiation dose escalation
urethra was involved in 60% of cases. Multiple treatment procedures were required in at least 34% of strictures, and at least 59% have not been fully resolved by treatment. In only 34% of cases has a full resolution been documented to date. Significant risk factors for post-radiation strictures were HDRB boost (OR: 8.8, p < 0.001) and BMI P 30 (OR: 3.1, p = 0.021). A reduced risk was associated with the trial arm 6 months AS and 18 months Z (OR: 0.26, p = 0.043) but more data are needed to confirm the latter.
Discussion
Fig. 1. Cumulative incidence estimates in the four radiation dose escalation groups at 6.5 years post randomisation for three potential surrogate endpoints for local (prostatic) progression: 1 – PSA progression (the maximum likelihood estimate 5); 2 – local progression diagnosed clinically (the minimum likelihood estimate 4); and 3 – local progression or PSA doubling time P6 months (a reasonable likelihood estimate h). See the electronic supplement for comment regarding the 74 Gy dose group.
A total of 47 urethral strictures were reported of which 45 occurred post-radiation. Their features are described in Tables 3 and 4. Post-radiation stricture occurred much later in men receiving EBRT and those receiving only 6 months AS. Obstructive symptoms predominated with nocturia being most common (45%). Dysuria was a less common feature (22%). While HDRB strictures tended to cause frequency (50%), EBRT strictures tended to cause urgency (45%) and incontinence (30%). Many strictures were quite long and affected several sections of the urethra. The bulbar
LP is a difficult endpoint to assess. Serial MRI scans represent the most promising option [12], but this option is likely to remain beyond the means of most centres around the world for some time. It was therefore highly instructive to look at the behaviour of three different surrogate endpoints. We had anticipated some contamination of our PSA based potential LP surrogates by distant progression occurring in the absence of clinical evidence of LP because 80.3% of subjects in the study group had high risk ‘‘primary’’ cancers which often produce distant progressions. However the 48.9% potential contamination rate for the MaxLE of LP (PSA progression) endpoint was both unexpected and alarmingly high for such a commonly used potential surrogate endpoint. Fortunately contamination by distant progression is likely to be lower in studies involving men with low and intermediate risk cancers such as the on-going dose escalation trials [13–18]. In spite of this source of concern, our radiation dose–response relationships for all three endpoints for LP are gratifyingly coherent. Our composite LP endpoint was the most reliable of our potential surrogates due to its lower level of contamination by distant progression and its independence from the Z/GS interaction that has influenced all of our distant progression endpoints. This in turn enabled composite LP to be evaluated in our entire dataset. In summary, regardless of the surrogate used, dose escalation very clearly reduces LP. However reductions in distant progression or prostate cancer deaths have not been identified yet because we have had to restrict our analyses of these endpoints to the two AS arms due to the interaction between zoledronate and Gleason score. Moreover metastases originating as a result of failure to control the primary tumour may not have grown sufficiently to be detected yet [6].
Adjusted cumulative incidence (%) Composite local progression
35 30 25 20 15 10 5 0 66
70
74
HDRB
Radiation dose escalation group (Gy) 6 months AS
18 months AS
Fig. 2. A dose–response comparison between radiation dose escalation subgroups of men treated with 6 months neo-adjuvant AS versus 6 months neo-adjuvant AS followed by 12 months adjuvant AS (i.e. 18 months in total) at 6.5 years post randomisation. The error bars represent standard errors. See the electronic supplement for comment regarding the 74 Gy group.
Please cite this article in press as: Denham JW et al. Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.05.016
5
J.W. Denham et al. / Radiotherapy and Oncology xxx (2015) xxx–xxx Table 3 Clinical features of the 47 urethral strictures documented, procedures undertaken to relieve them, and outcomes to date. N
%
2 0 45 47
4.3 0.0 95.7 100.0
Median
Range
Years after EBRT (n = 12)
3.6
Years after HDRB (n = 28) Missing
1.2 3
(0.6– 8.6) (0–4.8) 6.7
Onset of symptoms Before radiotherapy During radiotherapy After radiotherapy Total Timing of diagnosis (post radiotherapy strictures only)
Site of stricture (NB multiple sites involved in some cases) Prostatic 7 Membranous 17 Bulbar 33 Missing 4
14.9 36.2 70.2 8.5
Severity of stricture Minor, easily dilated Moderate, urethrotomy Tight or impassable Missing
18 8 12 9
38.3 17.0 25.5 19.1
Number of procedures necessary One Two Multiple Missing
20 9 16 2
42.5 19.2 34.0 4.3
Types of procedures Dilatation (±catheterisation) Internal urethrotomy (±catheterisation) Urethroplasty TUR Bladder neck incision Other procedures Missing
19 23 1 0 0 2 2
40.4 48.9 2.1 0.0 0.0 4.3 4.3
Outcomes of procedures Complete resolution Asymptomatic but continuing self-catheterisation Partial resolution and may require further procedures Unresolvable permanent in dwelling catheter Unknown Missing
16 9 16 3 2 1
34.0 19.2 34.0 6.4 4.3 2.1
noted that urethral strictures had been diagnosed in men receiving the HDRB boost. In this report we have documented the development of urethral strictures in 47 subjects. It is important to recognise that our data are limited by the fact that this treatment related complication was not anticipated a priori, and as a result stricture data were not collected prospectively. Stricturing is almost certainly more frequent than we have reported. Men treated with EBRT alone tended to have their strictures later than those receiving the HDRB boost. Since stricturing is less likely to be investigated and treated in elderly patients, the dose–response relationship may be less clear cut than our data suggest. Catheter slippage has been implicated as a cause for stricture following HDRB, however our quality control data suggest that unintended urethral coverage by EBRT may also have been an important contributor. EBRT treatment was planned using CT alone which is well recognised to lead to coverage of the membranous and sometimes bulbar portions of the urethra because the prostatic apex is not well visualised using CT. It is important to note that a similar result emerged from the Canadian ASCENDE-RT⁄ trial for men with unfavourable risk prostate cancer recently reported orally by Morris and colleagues. [22] In this randomised trial all 400 subjects received 8 months neo-adjuvant AS and pelvic EBRT to 46 Gy using 23 2 Gy fractions. In the control arm a conformal EBRT boost of 32 Gy using 16 Gy fractions followed. In the experimental arm a low dose rate brachytherapy (LDRB) boost prescribed to a minimum peripheral dose of 116 Gy occurred afterwards. The hazard of PSA progression was significantly reduced in the LDRB arm (0.45 [95% CI: 0.29– 0.77], p = 0.002) but Grade 3 urinary toxicity was significantly increased in the brachytherapy boost arm (p < 0.0001). [23] Urethral strictures were responsible for approximately half of the toxic events in the LDRB trial arm and, of course, catheter slippage would not have been a contributing factor to their development. It is hoped that recent technical innovations, including MRI fusion in the planning of treatment and fiducial marker guidance, will successfully prevent unnecessary coverage of the urethra and its vascular supply and, in doing so, reduce urethral strictures. The price of reducing LP using an additional year of AS was not negligible either. ‘‘Hormone treatment related symptoms’’ as defined in the EORTC PR25 instrument remained significantly
Table 4 Risk factors for post-radiation urethral strictures.
Therefore we intend to revisit this issue in 2017 when 10 year follow up data will be available for analysis. Our data also provide one of the first demonstrations that an additional year of AS can shift the radiation dose–response relationship favourably. In fact it reduced the hazard of our composite LP endpoint by approximately 40%, after adjusting for radiation dose. In this portion of the dose–response curve this equates to approximately 8 Gy using 2 Gy fractions. It must be pointed out that there were too few participants in the RADAR trial with intermediate risk cancers to know whether this degree of benefit is experienced by them. Some authorities will therefore continue to be doubtful that AS is necessary for men with intermediate risk cancers [19–21]. It is also gratifying to note that the benefits of radiation dose escalation and one additional year of AS have been achieved at the cost of modest increases in treatment related morbidity. In our 2014 main endpoints report [10] we had the opportunity to provide an update on our rectal and urinary dysfunction symptomatology report published in this journal in 2012 [2] as well as quality of life indices reported in the same year [1]. It was pleasing to report that at 6.5 years of follow-up, differences between dosing groups remained absent. However this does not mean that the price of radiation was negligible in the RADAR trial. In our 2012 report we
Dosing group 66 Gy 70 Gy 74 Gy HDRB Total
Total 125 425 262 237 1049
Stricture 1 4 10 30 45
% 0.8 0.9 3.8 12.7 4.3
Risk factors for stricture (multivariate logistic regression model) Factor OR 95% confidence intervals
pValue
Radiation dose EBRT HDRB boost
1.00 8.82
– 4.52–17.18
– <0.001
Treatment arm 6 mths AS 6 mths AS + 18 mths Z 18 mths AS 18 mths AS + 18 mths Z
1.00 0.26 1.28 1.60
– 0.07–0.96 0.56–2.95 0.70–3.63
– 0.043 0.56 0.26
Age (years)
1.04
0.99–1.09
0.08
BMI <25 P25 and <30 P30
1 1.47 3.05
0.59–3.65 1.19–7.82
0.40 0.021
2.52
0.79–8.06
0.12
Pre-existing lower urinary disorders and treatment
Please cite this article in press as: Denham JW et al. Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.05.016
6
Radiation dose escalation
worse in subjects receiving an additional year of AS for almost 3 years after randomisation. Our RADAR Survivorship Substudy, which will measure the legacies of protocol treatment 10 years after randomisation, will be helpful in determining what long-term price will be paid by our subjects for radiation and AS induced morbidity. Our appraisal of the value of a HDRB boost as a means of achieving dose escalation is limited by the fact that radiation dose was not allocated to dosing group by randomisation. Consequently younger subjects with high stage and grade tumours at centres with HDRB equipment were more likely to receive the HDRB boost. Although the HDRB boost emerged as the most successful means of preventing all three measures of LP, the adverse selection biases make it impossible to be sure what the HDRB regimen is equivalent to using a 2 Gy per fraction EBRT regimen, and therefore where it fits on the 2 Gy per fraction dose–response curve. A well designed randomised controlled trial such as the British trial run by Hoskin et al. is necessary to provide an appropriate estimate [24]. In the meantime the less accurate estimates may come from the ASCENDE-RT⁄ trial (Morris et al.) and the carefully matched comparison of outcomes between men treated with the HDRB boost regimen that was subsequently used in the RADAR trial, and men treated by EBRT alone using 74 Gy in 37 daily 2 Gy fractions at the Peter McCallum Cancer Institute, Melbourne [25]. In conclusion, the RADAR trial provides evidence that in men with high risk LAPC both radiation dose escalation and longer AS reduce LP regardless of how it is defined. The RADAR trial data also indicate that the advantages of both approaches are offset by increased morbidity. Optimal outcomes will be achieved by judicious combinations that take into account the patient’s life expectancy and fitness to tolerate the expected complications of treatment. Trial number The TROG 03.04 Trial is registered with the National Institutes of Health Clinical Trials Registry, number NCT00193856 and has approval from Hunter New England Human Research Ethics Committee (Trial ID. 03/06/11/3.02). Role of funding source The Goodfellow Foundation, Auckland (New Zealand); Cancer Standards Institute of New Zealand; National Health and Medical Research Council of Australia (NHMRC) (Project Application ID 300705 and 455521); Novartis Pharmaceuticals Australia; Abbott Pharmaceuticals Australia. Conflict of interest Nigel Spry: Consultant or advisory relationship (compensated) with Abbott Pharmaceuticals, Honoraria from Abbott Pharmaceuticals; Research Funding from Abbott Pharmaceuticals. All other authors have no conflicts of interest to disclose. Acknowledgements We thank Rosemary Bradford once again for her expert preparation of the manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.radonc.2015.05. 016.
References [1] Denham JW, Wilcox C, Joseph D, Spry NA, Lamb DS, Tai KH, et al. Quality of life in men with locally advanced prostate cancer treated with leuprorelin and radiotherapy with or without zoledronic acid (TROG 03.04 RADAR): secondary endpoints from a randomised phase 3 factorial trial. Lancet Oncol 2012;13:1260–70. [2] Denham JW, Wilcox C, Lamb DS, Spry NA, Duchesne G, Atkinson C, et al. Rectal and urinary dysfunction in the TROG 03.04 RADAR trial for locally advanced prostate cancer. Radiother Oncol 2012;105:184–92. [3] Ebert MA, Harrison KM, Cornes D, Howlett SJ, Joseph DJ, Kron T, et al. Comprehensive Australasian multicentre dosimetric intercomparison: issues, logistics and recommendations. J Med Imaging Radiat Oncol 2009;53:119–31. [4] Ebert MA, Harrison KM, Howlett SJ, Cornes D, Bulsara M, Hamilton CS, et al. Dosimetric intercomparison for multicentre clinical trials using a patientbased anatomic pelvic phantom. Med Phys 2011;38:5167–75. [5] Harrison KM, Ebert MA, Kron T, Howlett SJ, Cornes D, Hamilton CS, et al. Design, manufacture, and evaluation of an anthropomorphic pelvic phantom purpose-built for radiotherapy dosimetric intercomparison. Med Phys 2011;38:5330–7. [6] Denham JW, Steigler A, Tai KH, Joseph D, Matthews J, Atkinson C, et al. Paradoxical metastatic progression following 3 months of neo-adjuvant androgen suppression in the TROG 96.01 trial for men with locally advanced prostate cancer. Radiother Oncol 2013;107:123–8. [7] Kearvell R, Haworth A, Ebert MA, Murray J, Hooton B, Richardson S, et al. Quality improvements in prostate radiotherapy: outcomes and impact of comprehensive quality assurance during the TROG 03.04 ‘RADAR’ trial. J Med Imaging Radiat Oncol 2013;57:247–57. [8] Ebert MA, Haworth A, Kearvell R, Hooton B, Coleman R, Spry N, et al. Detailed review and analysis of complex radiotherapy clinical trial planning data: evaluation and initial experience with the SWAN software system. Radiother Oncol 2008;86:200–10. [9] Roach III M, Hanks G, Thames Jr H, Schellhammer PF, Shipley WU, Sokol GH, et al. Defining biochemical failure following radiotherapy with or without hormonal therapy in men with clinically localized prostate cancer: recommendations of the RTOG-ASTRO Phoenix Consensus Conference. Int J Radiat Oncol Biol Phys 2006;65:965–74. [10] Denham JW, Joseph D, Lamb DS, Spry NA, Duchesne G, Matthews J, et al. Shortterm androgen suppression and radiotherapy versus intermediate-term androgen suppression and radiotherapy, with or without zoledronic acid, in men with locally advanced prostate cancer (TROG 03.04 RADAR): an openlabel, randomised, phase 3 factorial trial. Lancet Oncol 2014;15:1076–89. [11] Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Statist Assoc 1999;94:496–509. [12] Haider MA, Chung P, Sweet J, Toi A, Jhaveri K, Menard C, et al. Dynamic contract-enhanced magnetic resonance imaging for localization of recurrent prostate cancer after external beam radiotherapy. Int J Radiat Oncol Biol Phys 2008;70:425–30. [13] Kuban DA, Levy LB, Cheung MR, Lee AK, Choi S, Frank S, et al. Long-term failure patterns and survival in a randomized dose-escalation trial for prostate cancer. Who dies of disease? Int J Radiat Oncol Biol Phys 2011;79:1310–7. [14] Zietman AL, DeSilvio M, Slater JD, Rossi C, Miller DW, Adams JA, et al. Comparison of convention-dose vs high-dose conformal radiation therapy in clinically localized adenocarcinoma of the prostate. JAMA 2005;294: 1233–9. [15] Beckendorf V, Guerif S, Le Prise E, Cosset JM, LeFloch O, Chauvet B, et al. The GETUG 70 Gy vs. 80 Gy randomized trial for localized prostate cancer: feasibility and acute toxicity. Int J Radiat Oncol Biol Phys 2004;60:1056–65. [16] Peeters ST, Heemsbergen WD, Koper PC, van Putten WL, Slot A, Dielwart MF, et al. Dose-response in radiotherapy for localized prostate cancer: results of the Dutch multicenter randomized phase III trial comparing 68 Gy of radiotherapy with 78 Gy. J Clin Oncol 2006;24:1990–6. [17] Dearnaley DP, Sydes MR, Graham JD, Aird EG, Bottomley D, Cowan RA, et al. Escalated-dose versus standard-dose conformal radiotherapy in prostate cancer: first results from the MRC RT01 randomised controlled trial. Lancet Oncol 2007;8:475–87. [18] Shipley WU, Verhey LJ, Munzenrider JE, Suit HD, Urie MM, McManus PL, et al. Advanced prostate cancer: the results of a randomized comparative trial of high dose irradiation boosting with conformal protons compared with conventional dose irradiation using photons alone. Int J Radiat Oncol Biol Phys 1995;32:3–12. [19] Zelefsky MJ, Pei X, Chou JF, Schechter M, Kollmeier M, Cox B, et al. Dose escalation for prostate cancer radiotherapy: predictors of long-term biochemical tumour control and distant metastases-free survival outcomes. Eur Urol 2011;60:1133–9. [20] Al-Mamgani A, Heemsbergen WD, Levendag PC, Lebesque JV. Subgroup analysis of patients with localized prostate cancer treated within the Dutchrandomized dose escalation trial. Radiother Oncol 2010;96:13–8. [21] Smith MJ, Akhtar NH, Tagawa ST. The current role of androgen deprivation in patients undergoing dose-escalated external beam radiation therapy for clinically localized prostate cancer. Prostate Cancer. 2012;2012, Article ID 280278. [22] Morris W.J., Tyldesley S., Pai H.H., Halperin R., McKenzie M.R., Duncan G., et al. ASCENDE-RT⁄: A multicenter, randomized trial of dose-escalated external beam radiation therapy (EBRT-B) versus low-dose-rate brachytherapy (LDR-B)
Please cite this article in press as: Denham JW et al. Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.05.016
J.W. Denham et al. / Radiotherapy and Oncology xxx (2015) xxx–xxx for men with unfavorable-risk localized prostate cancer. J Clin Oncol 33. 2015(suppl 7; abstr 3). [23] Rodda S., Tyldesley S., Morris W. GU and GI toxicity in ASCENDE-RT⁄: a multicentre randomized trial of dose-escalated radiation for prostate cancer. 3rd ESTRO Forum (Abstract PD-0047); Barcelona 2015. [24] Hoskin PJ, Rojas AM, Bownes PJ, Lowe GJ, Ostler PJ, Bryant L. Randomised trial of external beam radiotherapy alone or combined with high-dose-rate
7
brachytherapy boost for localised prostate cancer. Radiother Oncol 2012;103:217–22. [25] Khor R, Duchesne G, Tai KH, Foroudi F, Chander S, Van Dyk S, et al. Direct 2arm comparison shows benefit of high-dose-rate brachytherapy boost vs external beam radiation therapy alone for prostate cancer. Int J Radiat Oncol Biol Phys 2013;85:679–85.
Please cite this article in press as: Denham JW et al. Radiation dose escalation or longer androgen suppression for locally advanced prostate cancer? Data from the TROG 03.04 RADAR trial. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.05.016