Variation in Radiotherapy Referral and Treatment for High-risk Pathological Features after Radical Prostatectomy: Results from a Population-based Study

Clinical Oncology xxx (2017) 1e10 Contents lists available at ScienceDirect

Clinical Oncology journal homepage: www.clinicaloncologyonline.net

Original Article

Variation in Radiotherapy Referral and Treatment for High-risk Pathological Features after Radical Prostatectomy: Results from a Population-based Study C.J. Jin *y, T.P. Hanna *z, E.F. Cook y, Q. Miao z, M.D. Brundage *z * Department

of Oncology, Cancer Center of Southeastern Ontario at Kingston General Hospital, Queen’s University, Kingston, Ontario, Canada y Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA z Division of Cancer Care and Epidemiology, Queen’s University Cancer Research Institute, Kingston, Ontario, Canada Received 14 June 2017; received in revised form 21 September 2017; accepted 25 September 2017

Abstract Aims: Guidelines recommend the discussion of adjuvant radiotherapy post-prostatectomy for prostate cancer patients with high-risk pathology to consider all of their treatment options. We determine whether patterns of radiotherapy referral and treatment post-prostatectomy reflect guideline-based use in a contemporary prostatectomy cohort. Materials and methods: Electronic treatment records were linked to Ontario’s cancer registry. Multivariable regression was used to evaluate clinical and health systems factors associated with referral and the use of adjuvant radiotherapy within 6 months post-prostatectomy. Results: Among 2663 patients treated with prostatectomy between 1 January 2012 and 30 November 2012, 1261 (47%) were found to have adverse pathology and 492 were referred to radiation oncology 6 months post-prostatectomy, of whom 51% received adjuvant radiotherapy. Multivariable analysis showed that patients were more likely to be referred to radiation oncology from a low-volume surgical facility (50 versus >50 radical prostatectomy cases, odds ratio 2.50 [1.80e3.48]), if they lived farther from a radiotherapy centre (>50 km versus <10 km, odds ratio 1.73 [1.22e2.46]), if they were seen by radiation oncology preoperatively (odds ratio 1.95 [1.51e2.52]), or if they had adverse pathology: high T-category (pT3b/T4 versus pT2, odds ratio 17.87 [12.14e26.30]; pT3a versus pT2, odds ratio 5.24 [3.95e6.97]), positive margins (non-apex positive versus negative, odds ratio 4.20 [3.19e5.53]; apex only positive versus negative, odds ratio 2.60 [1.71e3.94]) and high Gleason score (8e10 versus 6, odds ratio 11.32 [5.37e23.84]; 7 versus 6, odds ratio 4.18 [2.16e8.10]). Wide geographic variation in radiotherapy referral rates persisted (range 6e66%; P < 0.0001). After radiotherapy referral, only high T-category (pT3b/T4 versus pT2, odds ratio 5.37 [3.01e9.60]; pT3a versus pT2, odds ratio 2.72 [1.59e4.65]) and non-apex positive margins (odds ratio 2.81 [1.86e4.23]) remained significantly predictive of treatment. Conclusions: Variations in referral for a discussion of radiotherapy post-prostatectomy are not mainly explained by patient characteristics. After seeing radiation oncology, treatment decisions correlated most strongly with pathological findings. Understanding the reasons for the tremendous non-clinical variations in care is needed to ensure access to potentially curative radiotherapy post-prostatectomy for high-risk prostate cancer patients. Ó 2017 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.

Key words: Access and evaluation; health care quality; physician’s practice patterns; prostatectomy; prostatic neoplasms; radiation oncology

Introduction The decision on many cancer treatments is challenging and value-laden, and patients need information to meaningfully participate in decisions about their health care. A survey of Ontario cancer patients identified the need for

information as one of the greatest problems they encountered [1]. Particularly for men with a diagnosis of high-risk prostate cancer, recent trends show an increase in the use of radical prostatectomy [2]. For patients found to have highrisk pathological features, including pT3 disease and/or positive margins, 60e70% will develop biochemical

Author for correspondence: M.D. Brundage, Department of Oncology, Cancer Centre of Southeastern Ontario at Kingston General Hospital, Burr 2 Kingston General Hospital, 25 King Street W, Kingston, Ontario, K7L 5P9 Canada. Tel: þ1-613-544-2631 ext 4144; Fax: þ1-613-548-1355. E-mail address: [email protected] (M.D. Brundage). https://doi.org/10.1016/j.clon.2017.10.009 0936-6555/Ó 2017 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Jin CJ, et al., Variation in Radiotherapy Referral and Treatment for High-risk Pathological Features after Radical Prostatectomy: Results from a Population-based Study, Clinical Oncology (2017), https://doi.org/10.1016/j.clon.2017.10.009

2

C.J. Jin et al. / Clinical Oncology xxx (2017) 1e10

recurrence and face a complex decision with respect to the timing and type of management post-radical prostatectomy [3e6]. Adjuvant radiotherapy (ART) has been shown to reduce the risk of biochemical recurrence and, in some cases, improve survival for men with high-risk pathology postradical prostatectomy in three large randomised trials [3,6,7]. However, controversy surrounding the trade-offs with toxicity exists, and delayed referral for salvage radiotherapy is under evaluation as an alternative strategy [8e11]. Considering these factors, patients with high-risk pathological features, specifically pT3 disease and/or positive margins, are recommended by guidelines to have an informed discussion of the potential benefits and risks of ART as compared with alternatives [12e17]. The American Urological Association (AUA) and American Society for Radiation Oncology (ASTRO) recommend a patient-centred, multidisciplinary discussion and guidelines from the Genitourinary Radiation Oncologists of Canada (GUROC) and Cancer Care Ontario (CCO) specifically recommend radiation oncologist consultation for this discussion [13,14,17]. Despite the need for ART discussion raised by international consensus guidelines, accumulating epidemiological data suggest that ART rates remain low, and are correlated with institutional characteristics, including surgical volume and academic versus non-academic affiliation [18e24]. The considerable geographic variation in the use of ART raises the question of whether patients are provided with equal opportunities in getting the information that they need to make fully informed decisions on curative treatments. This has important implications for principles of patient autonomy and justice [25]. Although clinical practice guidelines are systematically developed and are intended to streamline care and reduce unwarranted variation, the dissemination of evidence and guidelines often has variable effects on practice [26]. Management issues surrounding adverse pathology post-radical prostatectomy will affect a large number of North Americans this year. Past analyses of processes of care, including referral to radiation oncology, leading to radiotherapy use in prostate cancer, are limited. Given that in the Canadian setting early radiation oncology referral is clearly recommended for the discussion of treatment options in practice guidelines, we set out to evaluate health system performance regarding radiotherapy referral and subsequent treatment post-prostatectomy. We sought to investigate medical and non-medical factors associated with practice patterns in a large, population-based study, and the degree to which practice patterns reflected guideline-recommended access to patient-centred care. Given the lack of definitive comparison between adjuvant versus salvage radiotherapy, as well as the competing time pressures and burden of disclosure faced by urologists, we hypothesised that sizeable variations in referral rates to radiation oncology and subsequent ART use would exist and that non-clinical factors would strongly predict access to a radiotherapy opinion.

Materials and Methods Study Design This retrospective cohort study included all prostate cancer patients treated with radical prostatectomy between 1 January 2012 and 30 November 2012 in Ontario. Data Sources and Linkage Ontario has a population of 13.8 million people, and a publicly funded universal health insurance plan. The provincial cancer agency, CCO, is responsible for co-ordinating provincial cancer centres, which are the only providers of radiotherapy. Radical prostatectomy is carried out in a wider range of public hospitals. The Ontario Cancer Registry is a population-based registry operated by CCO that collects demographic and clinical information on >95% of cancer cases diagnosed in Ontario [27]. Records on incident prostate cancer cases were used. Registry data were linked to hospital separation data identifying radical prostatectomy cases (Canadian Institute of Health Information); pathology data (uniformly electronically compiled by CCO since 2012); radiotherapy visit and treatment data (routinely electronically compiled by CCO from each radiotherapy centre); and for socioeconomic status, neighbourhood median income quintile data (Statistics Canada). Values in the results were suppressed for groups of less than five patients, as per privacy agreements with CCO. Definitions of Radiation Oncology Consultation and Radiotherapy Use The primary outcomes of this study were whether patients were seen by radiation oncology early for consideration of ART, and receipt of ART. We defined radiation oncology consultation post-radical prostatectomy as the first radiation oncology visit 6 months post-radical prostatectomy. ART was defined as curative-intent radiotherapy initiation 6 months, as previously described [9,17,28]. Pathology data were routinely electronically compiled since 2012 and radiotherapy records were complete to 30 May 2013, allowing us to report on radiotherapy use 6 months post-radical prostatectomy for cases with an index surgery date from 1 January 2012 up to 30 November 2012. Definitions of Explanatory Variables Patients were considered to have received a preoperative radiation oncology consultation if they were seen within 12 months prior to radical prostatectomy [29]. Surgical margin positive cases were stratified by site of positive margin to account for vague boundaries at the prostatic apex. When tumour was found at the inked apical margin and margin-confined elsewhere, margins were classified as ‘apex positive only’. Other cases were categorised as ‘at least 1 non-apex margin positive’.

Please cite this article in press as: Jin CJ, et al., Variation in Radiotherapy Referral and Treatment for High-risk Pathological Features after Radical Prostatectomy: Results from a Population-based Study, Clinical Oncology (2017), https://doi.org/10.1016/j.clon.2017.10.009

C.J. Jin et al. / Clinical Oncology xxx (2017) 1e10

Radical prostatectomy hospitals were categorised into tertiles by radical prostatectomy volume. The highest tertile facilities were grouped as ‘high’ volume and others as ‘low’ volume. We used a lower limit of 50 radical prostatectomy cases over the study period to define ‘high’ volume facilities, consistent with the literature [23,30,31]. Radical prostatectomy hospitals were additionally classified by affiliation with a radiotherapy facility: hospitals with no radiotherapy facility, multisite hospitals with radiotherapy at another site only and hospitals with radiotherapy on-site [32]. Cancer centre region was determined for each radical prostatectomy hospital. This was defined as the most responsible radiotherapy centre at which most patients treated with radical prostatectomy at that hospital received radiotherapy. If <70% received radiotherapy at the same centre, the closest cancer centre to the radical prostatectomy hospital was assigned. The distance to the most responsible radiotherapy centre was calculated from the linear distance between the geometric centre of a patient’s residential postal code and the assigned cancer centre. Statistical Analysis We carried out univariate and multivariable logistic regression to evaluate rates of radiation oncology referral and subsequent ART use over the study period. We repeated the analyses using stepwise elimination models where P < 0.15 and results were not meaningfully different. Exploration of confounding and effect modification was carried out using stratified contingency tables and logistic regression. Associations were considered statistically significant when P < 0.05; all tests were two-tailed. A statistical analysis was carried out using SAS statistical software (version 9.4; SAS Institute Inc, Cary, NC, USA).

Table 1 Characteristics of 2663 patients with prostate cancer treated with radical prostatectomy in Ontario from January to November 2012 Characteristic Clinical Age at radical prostatectomy (years) <50 50e64 65 Surgical margin Negative Positive Apex only At least 1 non-apex Unknown Pathological T classification T2 T3a T3b/T4 Unknown Pathological N classification N0/Nx N1 Gleason score 6 7 8e10 Unknown Systems Radiation oncology visit pre-radical prostatectomy No Yes Radical prostatectomy hospital surgical volume Low (50 radical prostatectomy cases) High (>50 radical prostatectomy cases) Cancer centre region* Smallest Middle values

Results Table 1 shows the characteristics of the 2663 prostate cancer patients in this cohort treated with radical prostatectomy in Ontario. Following radical prostatectomy, 1261 (47%) had at least one high-risk pathological feature: pT3a disease (25%), pT3b/T4 disease (10%) or positive margins (7% apex only; 20% non-apical margin involved). Among this group of patients with adverse pathology, 492 (39%) were seen by radiation oncology 6 months post-radical prostatectomy, 36% with T3a disease, 68% with T3b/T4 disease, 27% with apex only positive margin and 48% with non-apical margin involved. Among 492 high-risk pathology patients seen by radiation oncology, 51% subsequently received ART (Figure 1). Post-prostatectomy management patterns for patients with adverse pathology varied widely among Ontario cancer centre regions (Figure 2). Among this subcohort of men with at least one pathological risk factor (T3/T4 disease and/

3

Largest Availability of radiotherapy at radical prostatectomy hospital No radiotherapy on site Radiotherapy at 1 but not all sites Radiotherapy on-site Distance from home to nearest radiotherapy centre <10 km 10e50 km >50 km Socioeconomic status quintile 1 (lowest) 2 3 4 5 (highest)

n (% column)

116 (4%) 1647 (62%) 900 (34%) 1691 (63%) 197 (7%) 545 (20%) 230 (9%) 1519 (57%) 659 (25%) 269 (10%) 216 (8%) 2585 (97%) 78 (3%) 330 (12%) 1892 (71%) 212 (8%) 229 (9%)

1818 (68%) 845 (32%)

780 (29%) 1883 (71%) 32 (1%) 83 (3%), 175 (7%) 449 (17%)

1319 (50%) 778 (29%) 566 (21%)

774 (29%) 1162 (44%) 727 (27%) 419 485 525 569 665

(16%) (18%) (20%) (21%) (25%)

* Full data on all 14 cancer centre regions could not be shown due to privacy policies in the data sharing agreement.

Please cite this article in press as: Jin CJ, et al., Variation in Radiotherapy Referral and Treatment for High-risk Pathological Features after Radical Prostatectomy: Results from a Population-based Study, Clinical Oncology (2017), https://doi.org/10.1016/j.clon.2017.10.009

4

C.J. Jin et al. / Clinical Oncology xxx (2017) 1e10

Fig 1. Referral to radiation oncology (RO) within 6 months and use of adjuvant radiotherapy (ART) among 2663 patients with prostate cancer treated with radical prostatectomy (RP) in Ontario from January to November 2012. *High-risk pathology refers to cases with at pT3/4 disease and/or positive margins.

or positive margins), radiation oncology referral rates ranged from 11% to 92% and ART use ranged from 5% to 38%. Radiation oncology referral rates before surgery ranged from 15% to 59%. Table 2 summarises the multivariable logistic regression findings, showing higher rates of radiation oncology referral for patients with adverse pathological features, including high stage (pT3b/T4 versus pT2, odds ratio 17.87, 95% confidence interval e12.14 to 26.30; pT3a versus pT2, odds ratio

5.24, 95% confidence interval e3.95 to 6.97, P < 0.0001) and high Gleason score (8e10 versus 6, odds ratio 11.32, 95% confidence interval e5.37 to 23.84; 7 versus 6, odds ratio 4.18, 95% confidence interval e2.16 to 8.10, P ¼ 0.0001). Margin positive cases were more likely to be referred than margin negative cases, and having at least one non-apex margin positive (odds ratio 4.20, 95% confidence interval e3.19 to 5.53; P < 0.0001) was more strongly associated with referral than was an apex only positive margin (odds ratio 2.60, 95% confidence interval e1.71 to 3.94; P < 0.0001). Cancer centre region was also predictive of referral rates (odds ratios 0.15e6.95; P < 0.0001). A third of patients previously received a radiation oncology consultation before surgery and this was correlated with radiation oncology referral postoperatively (odds ratio 1.95, 95% confidence interval e1.51 to 2.52; P < 0.0001). Furthermore, patients receiving radical prostatectomy in low-volume surgical hospitals (odds ratio 2.50, 95% confidence interval e1.80 to 3.48; P < 0.0001) and residing further away from cancer centres (>50 km versus <10 km, odds ratio 1.73 95% confidence interval e1.22 to 2.46, P ¼ 0.009) were more likely to be referred. The characteristics of referred patients are further described in Table S1. Table 3 summarises the use of ART among patients seen by radiation oncology. A large proportion of referred patients with adverse features were not treated; nearly 40% of patients did not receive ART following referral even for nonapical margin positive or pT3b/T4 disease. After radiotherapy referral, only high stage (pT3b/T4 versus pT2, odds ratio 5.37, 95% confidence interval e3.01 to 9.60; pT3a versus pT2, odds ratio 2.72, 95% confidence interval 1.59 to 4.65, P < 0.0001) and positive margins (non-apex positive versus negative, odds ratio 2.81, 95% confidence interval e1.86 to 4.23, P < 0.0001) remained significantly predictive

Fig 2. Variation in radiation oncology consultation and adjuvant radiotherapy usage within 6 months post-prostatectomy for patients with highrisk pathological features among the 14 cancer centre regions in Ontario. Radiotherapy referral is shown as dark grey bars, with treatment as pale grey bars. *5% received radiotherapy. Value suppressed according to the privacy agreement with Cancer Care Ontario. Please cite this article in press as: Jin CJ, et al., Variation in Radiotherapy Referral and Treatment for High-risk Pathological Features after Radical Prostatectomy: Results from a Population-based Study, Clinical Oncology (2017), https://doi.org/10.1016/j.clon.2017.10.009

C.J. Jin et al. / Clinical Oncology xxx (2017) 1e10

5

Table 2 Factors associated with radiation oncology referral within 6 months among 2663 patients with prostate cancer treated with radical prostatectomy in Ontario in 2012 Characteristic

Clinical Age at radical prostatectomy (years) <50 (n ¼ 116) 50e64 (n ¼ 1647) 65 (n ¼ 900) Surgical margin Negative (n ¼ 1691) Positive Apex only (n ¼ 197) At least 1 non-apex (n ¼ 545) Unknown (n ¼ 230) Pathological T classification T2 (n ¼ 1519) T3a (n ¼ 659) T3b/T4 (n ¼ 269) Unknown (n ¼ 216) Pathological N classification N0/Nx (n ¼ 2585) N1 (n ¼ 78) Gleason score 6 (n ¼ 330) 7 (n ¼1892) 8e10 (n ¼ 212) Unknown (n ¼ 229) Systems Radiation oncology visit pre-radical prostatectomy No (n ¼ 1818) Yes (n ¼ 845) Radical prostatectomy hospital surgical volume Low (50 radical prostatectomy cases) (n ¼780) High (>50 radical prostatectomy cases) (n ¼ 1883) Cancer centre region* Lowest referral rate Middle referral rates Highest referral rate Availability of radiotherapy at radical prostatectomy hospital No radiotherapy on-site (n ¼ 1319) Radiotherapy at 1 but not all sites (n ¼ 778) Radiotherapy on-site (n ¼ 566) Distance from home to nearest radiotherapy centre <10 km (n ¼ 774) 10e50 km (n ¼ 1162) >50 km (n ¼ 727) Socioeconomic status quintile 1 (lowest) (n ¼ 419) 2 (n ¼ 485) 3 (n ¼ 525) 4 (n ¼ 569) 5 (highest) (n ¼ 665) *

Factors associated with seeing radiation oncology within 6 months % seen by radiation oncology (n ¼ 2663)

Odds ratio (95% confidence interval)

18% 21% 22%

Reference 0.71 (0.37e1.33) 0.70 (0.37e1.35)

13%

Reference

27% 48% 19%

2.60 (1.71e3.94) 4.20 (3.19e5.53) 4.42 (1.31e14.90)

8% 36% 68% 18%

Reference 5.24 (3.95e6.97) 17.87 (12.14e26.30) 0.61 (0.11e3.46)

21% 55%

Reference 0.88 (0.48e1.61)

3% 21% 59% 19%

Reference 4.18 (2.16e8.10) 11.32 (5.37e23.84) 5.54 (1.33e23.15)

18% 31%

Reference 1.95 (1.51e2.52)

31% 18%

2.50 (1.80e3.48) Reference

6% 25%, 25% 66%

0.15 0.68 0.80 6.95

21% 16% 32%

Reference 1.28 (0.87e1.90) 1.08 (0.72e1.64)

19% 23% 23%

Reference 1.35 (0.10e1.82) 1.73 (1.22e2.46)

21% 20% 22% 21% 24%

Reference 0.72 (0.48e1.09) 0.96 (0.64e1.42) 0.79 (0.52e1.18) 1.12 (0.76e1.66)

P value

0.457

<0.0001

<0.0001

0.681

0.0001

<0.0001

<0.0001

<0.0001 (0.06e0.39) (0.26e1.75), (0.38e1.68) (2.28e21.16) 0.455

0.009

0.128

Full data on all 14 cancer centre regions could not be shown due to privacy policies in the data sharing agreement.

of ART use post-radiation oncology referral. The magnitude of variation in ART by cancer centre region was less than for radiation oncology referral and statistically insignificant (odds ratios 0.28e1.92, P ¼ 0.33). Health systems factors

significantly correlated with whether a patient saw radiation oncology (surgery volume, distance) were not correlated with subsequent treatment choice. Patient age, N-category, affiliation of radical prostatectomy hospital

Please cite this article in press as: Jin CJ, et al., Variation in Radiotherapy Referral and Treatment for High-risk Pathological Features after Radical Prostatectomy: Results from a Population-based Study, Clinical Oncology (2017), https://doi.org/10.1016/j.clon.2017.10.009

6

C.J. Jin et al. / Clinical Oncology xxx (2017) 1e10

Table 3 Factors associated with adjuvant radiotherapy (ART) use among 581 patients with prostate cancer seen by radiation oncology post-radical prostatectomy in Ontario in 2012 Characteristic

Clinical Age at radical prostatectomy (years) <50 (n ¼ 21) 50e64 (n ¼ 354) 65 (n ¼ 206) Surgical margin Negative (n ¼ 221) Positive Apex only (n ¼ 54) At least 1 non-apex (n ¼ 262) Unknown (n ¼ 44) Pathological T classification T2 (n ¼ 124) T3a (n ¼ 235) T3b/T4 (n ¼ 183) Unknown (n ¼ 39) Pathological N classification N0/Nx (n ¼ 538) N1 (n ¼ 43) Gleason score 6* 7 (n ¼ 403) 8e10 (n ¼ 124) Unknown* Systems Radiation oncology visit pre-radical prostatectomy No (n ¼ 321) Yes (n ¼ 260) Radical prostatectomy hospital surgical volume Low (50 radical prostatectomy cases) (n ¼241) High (>50 radical prostatectomy cases) (n ¼ 340) Cancer centre regiony Lowest ART rate Middle ART rates

Factors associated with ART use among cases seen by radiation oncology % receiving ART (n ¼ 581)

Odds ratio (95% confidence interval)

38% 47% 47%

Reference 0.84 (0.28e2.48) 0.71 (0.23e2.17)

36%

Reference

33% 59% 48%

1.32 (0.66e2.63) 2.81 (1.86e4.23) 2.19 (0.31e15.51)

23% 47% 63% 46%

Reference 2.72 (1.59e4.65) 5.37 (3.01e9.60) 1.27 (0.05e30.59)

46% 58%

Reference 1.05 (0.51e2.17)

27% 44% 57% 49%

Reference 0.73 (0.16e3.24) 1.18 (0.26e5.49) 1.53 (0.09e25.72)

48% 45%

Reference 1.04 (0.70e1.56)

43% 49%

0.89 (0.52e1.50) Reference

19% 45%, 45%

0.31 0.99 1.19 1.68

P value

0.668

<0.0001

<0.0001

0.900

0.210

0.842

0.649

0.332

Highest ART rate 59% Availability of radiotherapy at radical prostatectomy hospital No radiotherapy on-site (n ¼ 274) 45% Radiotherapy at 1 but not all sites (n ¼ 127) 56% Radiotherapy on-site (n ¼ 180) 43% Distance from home to nearest radiotherapy centre <10 km (n ¼ 144) 47% 10e50 km (n ¼ 268) 47% >50 km (n ¼ 169) 46% Socioeconomic status quintile 1 (lowest) (n ¼ 87) 44% 2 (n ¼ 98) 60% 3 (n ¼ 119) 44% 4 (n ¼ 120) 44% 5 (highest) (n ¼ 157) 44%

(0.06e1.55) (0.25e3.92), (0.43e3.29) (0.53e5.31) 0.766

Reference 1.25 (0.65e2.37) 1.14 (0.63e2.08) 0.671 Reference 1.24 (0.77e1.98) 1.17 (0.67e2.04) 0.143 Reference 1.71 (0.88e3.33) 0.93 (0.50e1.73) 0.78 (0.40e1.49) 0.90 (0.48e1.66)

* Full data could not be shown due to privacy policies in the data sharing agreement. Total n ¼ 54 for Gleason score 6 and unknown categories. y Full data on all 14 cancer centre regions could not be shown due to privacy policies in the data sharing agreement.

Please cite this article in press as: Jin CJ, et al., Variation in Radiotherapy Referral and Treatment for High-risk Pathological Features after Radical Prostatectomy: Results from a Population-based Study, Clinical Oncology (2017), https://doi.org/10.1016/j.clon.2017.10.009

C.J. Jin et al. / Clinical Oncology xxx (2017) 1e10

with a radiotherapy facility and socioeconomic status were not associated with referral or treatment on multivariate analysis. Analyses were repeated using backwards elimination models (Tables S2, S3), as well as adjusting for two-way interactions between pN classification and both facility volume (P ¼ 0.04) and distance to cancer centre (P ¼ 0.06) (data not shown) and results were not substantially different.

Discussion Our findings show that overall post-prostatectomy referral rates remain low and wide variations in referral and treatment patterns persist, despite level 1 evidence and Canadian [12e14], European [15,16] and US [17] guidelines recommending early consideration of treatment options for adverse pathology post-prostatectomy. We observed that 47% of radical prostatectomy patients had high-risk features and may benefit from ART. However, 61% of these patients did not receive an early referral for a discussion of the radiotherapy options. Of those referred, half received ART. There was much less regional variation in ART utilisation once referred, compared with variation in radiation oncology referral. Our findings of higher ART rates in pT3/T4 and margin positive disease are to be expected and are consistent with other reports and guidelines [18,20e23]. Although adverse pathology was appropriately associated with higher referral and treatment rates, non-medical factors were important determinants of whether patients were referred. Marked variation in radiation oncology referral rates across regions persisted after adjusting for case-mix and other known factors. Once referred, only adverse pathology significantly predicted for ART utilisation among radiation oncologists. Explanations for this inter-centre variability are speculative, but presumably reflect the controversies surrounding ART versus salvage radiotherapy and the absence of mature clinical trial results directly comparing these alternatives. Additionally, lack of a consistent survival benefit was seen in updated randomised trials [3,33]. Although guidelines recommend discussion of ART to all prostatectomy patients with adverse pathology, to our knowledge there is no criterion-based benchmark rate of ART use. The protracted natural history of prostate cancer, proliferation of treatment options, perceived impact of radiotherapy toxicity and patient preferences, all probably contribute to variability in clinician and patient decision-making. Our results clearly illustrate the diversity of practice in radiation oncology referral and ART rates among high-risk patients across regions, despite national guidelines recommending radiation oncology referral for these patients. The largest variations are in radiation oncology referral and there are smaller variations in subsequent ART use among referred patients. Although the decision to refer (or not) to radiation oncology affects patient management, much of the observed regional variability in Figure 2 remains

7

unexplained. In centre D, low referral and very low subsequent treatment rates raise questions as to whether physicians are less likely to refer to radiation oncology when referred patients rarely receive ART. In centre N, a minority of high-risk men are referred, but the vast majority are then treated, raising concern whether participatory decisionmaking among physicians and patients is routinely occurring. Notably, patients treated at high-volume surgical facilities were substantially less likely to be referred to radiation oncology, and this difference persisted after adjusting for pathological and health systems factors. US-based studies have also identified lower ART use among high-volume surgical facilities [23] and teaching/research centres [22]. Whether this reflects institutional biases or factors inherent to high-volume facilities that reduce the propensity to refer for ART is unknown and requires further study. Local radiotherapy infrastructure availability in terms of affiliation of radical prostatectomy hospital with a cancer centre and living closer to a radiotherapy centre did not improve radiation oncology referral rates. Paradoxically, the effect of distance from the radiotherapy facility was inversely correlated with referral, which has been reported in the literature [19]. One explanation is that patients living furthest from radiotherapy facilities generally have more limited access to specialised medical care, increasing the propensity for referral. There were also wide regional variations in preoperative referral rates. The independent correlation between postoperative radiation oncology referral and whether patients received a preoperative radiation oncology consultation is intriguing and perhaps speaks to the level of collaboration between urology and radiation oncology or decisions by radiation oncology to reassess patient status post-radical prostatectomy. We considered other factors that may have contributed to the referral and treatment patterns observed. One such factor is the use of multidisciplinary case conferences (MCCs), which are increasingly used to facilitate peer review. Hospitals are required by CCO to hold a MCC for disease sites where >35 patients are treated yearly [34,35] and adherence to MCC standards overall is >50% for genitourinary cancers across Ontario during the study period, but varies by region. Differential use of MCCs across regions may account for a proportion of non-referred cases. These rounds may affect referral rates by helping to establish a culture of multidisciplinary care or may also replace the referral of an individual patient by providing a forum for discussion of a case that otherwise might have been referred. The latter, however, is less likely, as only a minority of cases are tabled at rounds, even when the MCC standards are adhered to. The primary relevant clinical trial running in the province during the study period was RADICALS [9]. Five of 14 cancer centre regions participated and only 25 patients in total were randomised to adjuvant or delayed radiotherapy in Ontario. Therefore, clinical trials were unlikely to have significantly influenced post-prostatectomy patterns of care.

Please cite this article in press as: Jin CJ, et al., Variation in Radiotherapy Referral and Treatment for High-risk Pathological Features after Radical Prostatectomy: Results from a Population-based Study, Clinical Oncology (2017), https://doi.org/10.1016/j.clon.2017.10.009

8

C.J. Jin et al. / Clinical Oncology xxx (2017) 1e10

Interpretation of our study results requires consideration of study limitations. First, the study was confined to a single Canadian province and may not be generalisable to other communities. Notably, the proportion of prostatectomies with high-risk features in our study was the same as the population-based rate found in British Columbia (47%) [18]. The reported ART rate in British Columbia was low (only 2% between 2005 and 2007), although this study was carried out shortly after the first trials favouring ART over observation were published, and there is variability in study definitions of ART. Furthermore, the interpretation of findings is limited by information available in retrospective administrative data. We lack data on the use of androgen ablation and previous surveys of urologists and oncologists have indicated a wide variation in the use of hormone therapy in the postprostatectomy patient population [36,37]. We also cannot ascertain patient preferences. Additionally, information regarding patient recovery post-radical prostatectomy was unavailable. Urologists may be reluctant to refer patients who continue to experience urinary incontinence postradical prostatectomy because of concerns about impaired recovery. Also, the lack of postoperative prostate-specific antigen data may have resulted in misclassification of ART versus salvage radiotherapy; radiotherapy within 6 months of radical prostatectomy probably encompasses a proportion of cases treated for recurrence or disease progression. However, this bias towards over-representation of ART may be compensated by cases with delays in the start of intended ART beyond 6 months post-radical prostatectomy. We also carried out a sensitivity analysis of radiotherapy within 4 and 8 months to describe ART; as the definition of ART is broadened to increase sensitivity, its specificity decreases, but cumulative radiotherapy utilisation trends were consistent. Despite these methodological limitations, the robustness of our study is based on large, comprehensive, reliable registries for the entire population of interest, thus minimising referral bias. Furthermore, reasons for low uptake of ART post-prostatectomy are not well described, and the contribution of radiation oncology referral to practice variation has not been fully accounted for. We evaluated two steps in the care pathway post-radical prostatectomy: referral to radiation oncology and subsequent use of ART. We identified substantial regional variation in referral and treatment rates, which remains unexplained by measurable factors. Practice is not always driven by diseaserelated factors, but may be subject to other forces, hypothesised to include clinical intuition, institutional conventions, patient preference and market forces. Presumption of equivalence for ART and early salvage radiotherapy must be viewed critically until comparative randomised trials are completed. Issues best explored qualitatively in future studies include urologists’ rationale for their practice patterns to better understand factors influencing referral choices of individual physicians and regions. For example, lower referral rates from high-volume surgical centres may reflect that some urologists feel comfortable having this discussion of adjuvant versus

salvage radiotherapy with their patients. Qualitative studies ascertaining the views of urologists on early radiation oncology referral, as well as the reasons for many men not receiving ART after referral, would be of interest. Although the oncology-oriented urologist is knowledgeable and the patient places an essential trust in his urologist, for a patient faced with navigating a highly personal decision in unfamiliar, complex territory, multidisciplinary team involvement is often indicated. Evidence suggests that patients generally want more information than they feel they receive [38e40]. Physicians have legal and ethical obligations to ensure that patients understand information about their options to meaningfully participate in treatment decisions [41]. Beyond these obligations, valuing patients’ information needs is beneficial because it increases patient satisfaction with treatment outcomes, reduces treatment decision-related regret and decreases anxiety and depression [42e44]. Specifically for prostate cancer patients, satisfaction with information received on therapy is correlated with quality of life post-treatment [45,46]. However, patients’ information needs are highly heterogeneous [47e50]. Furthermore, there is heterogeneity in how specialists counsel patients and the treatment options they favour [39,51]. As a result, patient decision-making is highly correlated with the professional background of the consulting specialist [52e54]. Therefore, consultation from an individual trusted physician alone may not be sufficient. Studies have shown that a multidisciplinary team results in better covering of patients’ information needs [55e57]. Notably, in the post-prostatectomy setting, ASTRO/AUA guidelines specifically recommend multidisciplinary involvement and Canadian GUROC and CCO guidelines recommend radiation oncology involvement [13,14,17]. Thus, conducting guideline-based referrals to ensure that patients are counselled on treatment options and the available clinical evidence is patient-centred, ethically sound and good medical practice. By serving as consistent advocates for patients and conducting patient-focused referrals, providers enhance trust by showing commitment to patient understanding and value for the patients’ definition of ‘benefit’. Fostering multidisciplinary management strategies would help support guideline-directed practice and may reduce practice-level variability, ensuring timely, equitable access to patient-centred care. An additional contributor to variations may be the lack of feedback available to providers. The ability of providers to compare their practices with guideline recommendations and their peers could help referring and treating physicians make appropriate decisions to optimise patient-centred access to care.

Conclusions Nearly half of patients have high-risk features postradical prostatectomy and may benefit from a discussion of the treatment options according to level I evidence and international consensus guidelines. We found that

Please cite this article in press as: Jin CJ, et al., Variation in Radiotherapy Referral and Treatment for High-risk Pathological Features after Radical Prostatectomy: Results from a Population-based Study, Clinical Oncology (2017), https://doi.org/10.1016/j.clon.2017.10.009

C.J. Jin et al. / Clinical Oncology xxx (2017) 1e10

radiotherapy referral and utilisation rates remain low despite national guidelines recommending referral. The tremendous variation in observed patterns of care is inconsistent with guideline recommendations and suggests that decision-making for prostate cancer patients with adverse pathology is subject to considerable discretion and uncertainty. This may have important implications for patient participation in decision-making across centres. Efforts should be made to understand and reduce inter-centre practice variation post-prostatectomy to ensure optimal access to patient-centred care.

Acknowledgements This study was supported by an Abbvie-Canadian Association of Radiation Oncology Award. Funding contribution supported a portion of our statistician’s salary. T.P. Hanna’s work was supported in part by the Ontario Institute for Cancer Research (#IA-035) through funding provided by the Government of Ontario.

Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.clon.2017.10.009.

References [1] Canadian Cancer Society. Final report on the needs of people living with cancer across Canada. Canadian Cancer Society; 1992. [2] Cooperberg MR, Carroll PR. Trends in management for patients with localized prostate cancer, 1990e2013. JAMA 2015; 314:80e82. [3] Bolla M, van Poppel H, Tombal B, Vekemans K, Da Pozzo L, de Reijke TM, et al. Postoperative radiotherapy after radical prostatectomy for high-risk prostate cancer: long-term results of a randomised controlled trial (EORTC trial 22911). Lancet 2012;380:2018e2027. €rkel S, [4] Wiegel T, Bottke D, Steiner U, Siegmann A, Golz R, Sto et al. Phase III postoperative adjuvant radiotherapy after radical prostatectomy compared with radical prostatectomy alone in pT3 prostate cancer with postoperative undetectable prostate-specific antigen: ARO 96-02/AUO AP 09/95. J Clin Oncol 2009;27:2924e2930. [5] Thompson IM, Tangen CM, Paradelo J, Lucia MS, Miller G, Troyer D, et al. Adjuvant radiotherapy for pathologically advanced prostate cancer: a randomized clinical trial. JAMA 2006;296:2329e2335. [6] Thompson IM, Tangen CM, Paradelo J, Lucia MS, Miller G, Troyer D, et al. Adjuvant radiotherapy for pathological T3N0M0 prostate cancer significantly reduces risk of metastases and improves survival: long-term followup of a randomized clinical trial. J Urol 2009;181:956e962. [7] Bolla M, van Poppel H, Collette L, van Cangh P, Vekemans K, Da Pozzo L, et al. Postoperative radiotherapy after radical prostatectomy: a randomised controlled trial (EORTC trial 22911). Lancet 2005;366:572e578. [8] King CR. Adjuvant versus salvage radiotherapy for high-risk prostate cancer patients. Semin Radiat Oncol 2013;23: 215e221.

9

[9] Parker C, Sydes MR, Catton C, Kynaston H, Logue J, Murphy C, et al. Radiotherapy and androgen deprivation in combination after local surgery (RADICALS): a new Medical Research Council/National Cancer Institute of Canada phase III trial of adjuvant treatment after radical prostatectomy. BJU Int 2007; 99:1376e1379. [10] King CR. The timing of salvage radiotherapy after radical prostatectomy: a systematic review. Int J Radiat Oncol Biol Phys 2012;84:104e111. [11] Pearse M, Fraser-Browne C, Davis ID, Duchesne GM, Fisher R, Frydenberg M, et al. A phase III trial to investigate the timing of radiotherapy for prostate cancer with high-risk features: background and rationale of the Radiotherapy e Adjuvant Versus Early Salvage (RAVES) trial. BJU Int 2014;113(Suppl. 2): 7e12. [12] Pickles T, Morgan S, Morton G, Souhami L, Warde P, Lukka H. Adjuvant radiotherapy following radical prostatectomy: Genito-Urinary Radiation Oncologists of Canada Consensus Statement. Can Urol Assoc J 2008;2:95e99. [13] Sia M, Pickles T, Morton G, Souhami L, Lukka H, Warde P. Salvage radiotherapy following biochemical relapse after radical prostatectomy: proceedings of the Genito-Urinary Radiation Oncologists of Canada consensus meeting. Can Urol Assoc J 2008;2:500e507. [14] Morgan SC, Waldron TS, Eapen L, Mayhew LA, Winquist E, Lukka H, et al. Adjuvant radiotherapy following radical prostatectomy for pathologic T3 or margin-positive prostate cancer: a systematic review and meta-analysis. Radiother Oncol 2008;88:1e9. [15] Heidenreich A, Bastian PJ, Bellmunt J, Bolla M, Joniau S, van der Kwast T, et al. EAU guidelines on prostate cancer. Part 1: screening, diagnosis, and local treatment with curative intent-update 2013. Eur Urol 2014;65:124e137. [16] Horwich A, Parker C, de Reijke T, Kataja V, ESMO Guidelines Working Group. Prostate cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 2013;24(Suppl. 6):vi106evi114. [17] Thompson IM, Valicenti RK, Albertsen P, Davis BJ, Goldenberg SL, Hahn C, et al. Adjuvant and salvage radiotherapy after prostatectomy: AUA/ASTRO Guideline. J Urol 2013;190:441e449. [18] Tyldesley S, Peacock M, Morris JW, So A, Kim-Sing C, Quirt J, et al. The need for, and utilization of prostate-bed radiotherapy after radical prostatectomy for patients with prostate cancer in British Columbia. Can Urol Assoc J 2012;6: 89e94. [19] Quon H, Suderman D, Guilbert K, Lambert P, Bucher O, Ong A, et al. Population-based referrals for adjuvant radiotherapy after radical prostatectomy in men with prostate cancer: impact of randomized trials. Clin Genitourin Cancer 2014;12: e1ee5. [20] Taggar A, Alghamdi A, Tilly D, Kostaras X, Kerba M, Sia M. Assessing guideline adherence: referral patterns of postprostatectomy patients to radiation oncologists. Radiother Oncol 2015;116:S14. [21] Wagner AE, Ghia AJ, Shrieve DC, Kokeny KE, Lowrance WT, Tward JD. Adjuvant radiotherapy after prostatectomy: have randomized clinical trials had any impact? J Radiat Oncol 2014;3:387. [22] Sineshaw HM, Gray PJ, Efstathiou JA, Jemal A. Declining use of radiotherapy for adverse features after radical prostatectomy: results from the National Cancer Data Base. Eur Urol 2015;68: 768e774. [23] Kalbasi A, Swisher-McClure S, Mitra N, Sunderland R, Smaldone MC, Uzzo RG, et al. Low rates of adjuvant radiation

Please cite this article in press as: Jin CJ, et al., Variation in Radiotherapy Referral and Treatment for High-risk Pathological Features after Radical Prostatectomy: Results from a Population-based Study, Clinical Oncology (2017), https://doi.org/10.1016/j.clon.2017.10.009

10

[24]

[25] [26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

C.J. Jin et al. / Clinical Oncology xxx (2017) 1e10 in patients with nonmetastatic prostate cancer with high-risk pathologic features. Cancer 2014;120:3089e3096. Zimmermann M, Taussky D, Delouya G, Alenizi AM, Zorn KC. Radiation therapy after radical prostatectomy: a single-centre radiation oncology experience in trends of referral and treatment practices. Can Urol Assoc J 2015;9:E608eE612. Faden RR, King NMP, Beauchamp TL. A history and theory of informed consent. New York: Oxford University Press; 1986. Lomas J, Anderson GM, Domnick-Pierre K, Vayda E, Enkin MW, Hannah WJ. Do practice guidelines guide practice? The effect of a consensus statement on the practice of physicians. N Engl J Med 1989;321:1306e1311. Robles SC, Marrett LD, Clarke EA, Risch HA. An application of capture-recapture methods to the estimation of completeness of cancer registration. J Clin Epidemiol 1988;41:495e501. Abdollah F, Suardi N, Cozzarini C, Gallina A, Capitanio U, Bianchi M, et al. Selecting the optimal candidate for adjuvant radiotherapy after radical prostatectomy for prostate cancer: a long-term survival analysis. Eur Urol 2013;63:998e1008. Jin CJ, Brundage MD, Cook EF, Miao Q, Hanna TP. Quality of radiation therapy referral and utilisation post-prostatectomy: a population-based study of time trends. Clin Oncol 2016; 28(12):783e789. Begg CB, Riedel ER, Bach PB, Kattan MW, Schrag D, Warren JL, et al. Variations in morbidity after radical prostatectomy. N Engl J Med 2002;346:1138e1144. Hu JC, Gold KF, Pashos CL, Mehta SS, Litwin MS. Role of surgeon volume in radical prostatectomy outcomes. J Clin Oncol 2003;21:401e405. Mackillop WJ, Kong W, Brundage M, Hanna TP, ZhangSalomons J, McLaughlin P, et al. A comparison of evidencebased estimates and empirical benchmarks of the appropriate rate of use of radiation therapy in ontario. Int J Radiat Oncol Biol Phys 2015;91:1099e1107. Wiegel T, Bartkowiak D, Bottke D, Bronner C, Steiner U, Siegmann A, et al. Adjuvant radiotherapy versus wait-and-see after radical prostatectomy: 10-year follow-up of the ARO 9602/AUO AP 09/95 trial. Eur Urol 2014;66:243e250. Cancer Care Ontario. Multidisciplinary cancer conferences (MCCs): transforming the delivery of cancer care, 19 June 2015 update. Available at: https://www.cancercare.on.ca/ common/pages/UserFile.aspx?fileIdZ13572. Cancer Care Ontario. Multidisciplinary cancer conferences, 18 February 2010 update. Available at: https://www.cancercare. on.ca/cms/one.aspx?portalIdZ1377&pageId Z8256. Morris SL, Parker C, Huddart R, Horwich A, Dearnaley D. Current opinion on adjuvant and salvage treatment after radical prostatectomy. Clin Oncol 2004;16:277e282. Lee LW, Clarke NW, Ramani VA, Cowan RA, Wylie JP, Logue JP. Adjuvant and salvage treatment after radical prostatectomy: current practice in the UK. Prostate Cancer Prostatic Dis 2005; 8:229e234. Cegala DJ, Bahnson RR, Clinton SK, David P, Gong MC, Monk JP, et al. Information seeking and satisfaction with physicianpatient communication among prostate cancer survivors. Health Commun 2008;23:62e69. Snow SL, Panton RL, Butler LJ, Wilke DR, Rutledge RDH, Bell DG, et al. Incomplete and inconsistent information provided to men making decisions for treatment of early-stage prostate cancer. Urology 2007;69:941e945. Davison BJ, Goldenberg SL, Gleave ME, Degner LF. Provision of individualized information to men and their partners to facilitate treatment decision making in prostate cancer. Oncol Nurs Forum 2003;30:107e114.

[41] Appelbaum PS, Lidz CW, Meisel A. Informed consent: legal theory and clinical practice. New York: Oxford University Press; 1987. [42] Mills ME, Sullivan K. The importance of information giving for patients newly diagnosed with cancer: a review of the literature. J Clin Nurs 1999;8:631e642. [43] D’haese S, Vinh-Hung V, Bijdekerke P, Spinnoy M, de Beukeleer M, Lochie N, et al. The effect of timing of the provision of information on anxiety and satisfaction of cancer patients receiving radiotherapy. J Cancer Educ 2000;15: 223e227. [44] Thomas R, Daly M, Perryman B, Stockton D. Forewarned is forearmed e benefits of preparatory information on video cassette for patients receiving chemotherapy or radiotherapy e a randomised controlled trial. Eur J Cancer 2000;36:1536e1543. [45] Davies NJ, Kinman G, Thomas RJ, Bailey T. Information satisfaction in breast and prostate cancer patients: implications for quality of life. Psychooncology 2008;17:1048e1052. [46] Gattellari M, Butow PN, Tattersall MH. Sharing decisions in cancer care. Soc Sci Med 2001;52:1865e1878. [47] Ruesch P, Schaffert R, Fischer S, Feldman-Stewart D, Ruszat R, €rri P, et al. Information needs of early-stage prostate Spo cancer patients: within- and between-group agreement of patients and health professionals. Support Care Cancer 2014; 22:999e1007. [48] Feldman-Stewart D, Brundage MD, Hayter C, Groome P, Nickel JC, Downes H, et al. What prostate cancer patients should know: variation in professionals’ opinions. Radiother Oncol 1998;49:111e123. [49] Feldman-Stewart D, Brundage MD, Hayter C, Groome P, Nickel JC, Downes H, et al. What questions do patients with curable prostate cancer want answered? Med Decis Making 2000;20:7e19. [50] Feldman-Stewart D, Capirci C, Brennenstuhl S, Tong C, Abacioglu U, Gawkowska-Suwinska M, et al. Information needs of early-stage prostate cancer patients: a comparison of nine countries. Radiother Oncol 2010;94:328e333. [51] Zeliadt SB, Ramsey SD, Penson DF, Hall IJ, Ekwueme DU, Stroud L, et al. Why do men choose one treatment over another?: a review of patient decision making for localized prostate cancer. Cancer 2006;106:1865e1874. [52] Fowler Jr FJ, McNaughton Collins M, Albertsen PC, Zietman A, Elliott DB, Barry MJ. Comparison of recommendations by urologists and radiation oncologists for treatment of clinically localized prostate cancer. JAMA 2000;283:3217e3222. [53] Jang TL, Bekelman JE, Liu Y, Bach PB, Basch EM, Elkin EB, et al. Physician visits prior to treatment for clinically localized prostate cancer. Arch Intern Med 2010;170:440e450. [54] Sommers BD, Beard CJ, D’Amico AV, Kaplan I, Richie JP, Zeckhauser RJ. Predictors of patient preferences and treatment choices for localized prostate cancer. Cancer 2008;113: 2058e2067. [55] Gomella LG, Lin J, Hoffman-Censits J, Dugan P, Guiles F, Lallas CD, et al. Enhancing prostate cancer care through the multidisciplinary clinic approach: a 15-year experience. J Oncol Pract 2010;6:e5ee10. [56] Valdagni R, Albers P, Bangma C, Drudge-Coates L, Magnani T, Moynihan C, et al. The requirements of a specialist Prostate Cancer Unit: a discussion paper from the European School of Oncology. Eur J Cancer 2011;47:1e7. [57] Strebel RT, Sulser T, Schmid HP, Gillessen S, Fehr M, Huber U, et al. Multidisciplinary care in patients with prostate cancer: room for improvement. Support Care Cancer 2013;21: 2327e2333.

Please cite this article in press as: Jin CJ, et al., Variation in Radiotherapy Referral and Treatment for High-risk Pathological Features after Radical Prostatectomy: Results from a Population-based Study, Clinical Oncology (2017), https://doi.org/10.1016/j.clon.2017.10.009