Fax: (011) 61-2-9828-5299
Estimating the optimal external-beam radiotherapy utilization rate for genitourinary malignancies
Article first published online: 20 DEC 2004
Copyright © 2004 American Cancer Society
Volume 103, Issue 3, pages 462–473, 1 February 2005
How to Cite
Delaney, G., Jacob, S. and Barton, M. (2005), Estimating the optimal external-beam radiotherapy utilization rate for genitourinary malignancies. Cancer, 103: 462–473. doi: 10.1002/cncr.20789
- Issue published online: 20 JAN 2005
- Article first published online: 20 DEC 2004
- Manuscript Accepted: 7 OCT 2004
- Manuscript Revised: 7 SEP 2004
- Manuscript Received: 18 MAY 2004
- Australian Commonwealth Department of Health and Ageing
- bladder cancer;
- genitourinary cancer;
- renal cancer;
- prostate cancer;
- testicular cancer;
- radiotherapy utilization
Benchmark radiotherapy utilization rates for genitourinary malignancies are largely unknown, despite the finding that genitourinary cancers comprise approximately 19% of all registered malignancies in Australia.
To develop an evidence-based benchmark of the optimal proportion of patients with genitourinary malignancies who should receive at least one course of radiotherapy at some time during their illness, the authors studied treatment guidelines and treatment reviews regarding genitourinary malignancies. Optimal radiotherapy utilization trees were constructed to show the clinical attributes that indicated possible benefit from radiotherapy based on evidence. Epidemiologic incidence data for each of these clinical attributes were obtained to calculate the optimal proportion of all patients with genitourinary cancer for whom radiotherapy was considered appropriate.
The proportion of patients with genitourinary malignancies for whom radiotherapy was indicated at some point in their illness, according to the best available evidence, was estimated to be 27% of patients with renal cancer, 58% of patients with bladder cancer, 60% of patients with prostate cancer, and 49% of patients with testicular cancer. The occurrence of ureteric and penile cancers among patients was too rare, and, therefore, these patients were not included in the current study.
There was a large discrepancy between actual radiotherapy utilization and the evidence-based optimal rate. The authors recommended strategies to implement the evidence-based guidelines. Evidence-based benchmarks for radiotherapy utilization rates such as the ones described in the current study were important in the evaluation of the appropriate use of radiotherapy. Cancer 2005. © 2004 American Cancer Society.
A benchmark of the adequacy of radiotherapy service provision is the proportion of patients with new diagnoses of cancer that receive at least one course of radiotherapy during the course of their illness (the radiotherapy utilization rate). Estimating an optimal radiotherapy utilization rate provides valuable information for the planning of adequate radiotherapy facilities, as well as a benchmark for actual practice.
Optimal or ideal radiotherapy utilization rates for breast cancer,1, 2 lung cancer,3, 4 colorectal cancer,5 melanoma,6 and prostate cancer7 have been reported previously. Some of these studies highlighted a large gap between what is considered optimal and the actual radiotherapy utilization rates observed in clinical practice. We present the results for genitourinary malignancies and compare the optimal radiotherapy utilization rate with actual radiotherapy utilization rates in Australia.
The current article presents data from a much larger study that was conducted to estimate the overall radiotherapy utilization rate for all registered cancers in Australia. The study included all cancers that are registered with Australian state cancer registries and have an incidence of ≥ 1%. Hence, kidney cancer, bladder cancer, prostate cancer, and testicular cancer are included under genitourinary cancer, but ureteric and penile cancers were not included as these cancers each have an incidence of < 1% of all registered cancers.
MATERIALS AND METHODS
Indications for Radiotherapy
We defined an indication for external-beam radiotherapy as a clinical situation in which radiotherapy is the treatment of choice on the basis of evidence that radiotherapy has a better clinical outcome than other treatment modalities (including no treatment), and in which the patient is an appropriate candidate for radiotherapy in terms of performance status indicators and the presence or absence of comorbidities. The superiority of radiotherapy compared with other treatment options was based on survival, local control, or toxicity profile.
As the aim of the study purely was to assess the optimal utilization rate for external-beam radiotherapy, brachytherapy was not studied.
International guidelines for the management of genitourinary malignancies published before August 2003 were identified. Guidelines for the treatment of renal cancer have been issued by the National Cancer Institute's (NCI) Physician Data Query (PDQ),8 the British Columbia Cancer Agency (BCCA),9 and the National Comprehensive Cancer Network (NCCN).10 Treatment guidelines available for bladder cancer include the NCI PDQ,11 the BCCA guidelines,12 the American Urological Association bladder cancer treatment guidelines (which did not discuss the role of radiotherapy at all),13 and the NCCN guidelines.14 Guidelines for the management of testicular cancer included the NCI PDQ guidelines,15 the BCCA guidelines,16 the Royal College of Radiologists Clinical Oncology Information Network (COIN) guidelines,17 the NCCN guidelines,18 the European Association of Urologists guidelines,19 and the guidelines issued by the German Testicular Cancer Study Group.20 Clinical practice guidelines for the management of prostate cancer have been published by the NCI (PDQ statement on prostate cancer),21 the NCCN,22 the Royal College of Radiologists/British Association of Urological Surgeons COIN,23 and the American Society of Therapeutic Radiation Oncology (ASTRO).24 There are no national-level Australian guidelines for the management of genitourinary cancer.
The level of evidence that supported each recommendation for radiotherapy use was classified using the Australian National Health and Medical Research Council hierarchy of levels of evidence.25
Using the best evidence available, a list of indications for radiotherapy in the management of genitourinary malignancies was created (see Table 1) and an optimal radiotherapy utilization tree was generated for each cancer. In the decision tree, each branch point represents an attribute (e.g., tumor stage or whether clear surgical margins are present) that affects a radiotherapy treatment decision. Each terminal branch of the tree showed whether radiotherapy was recommended for individuals with those particular clinical attributes.
|Outcome number in optimal radiotherapy utilization tree||Level of evidencea||Reference no.||Proportion of all patients with that cancer|
|Proportion of all patients with renal cancer in whom radiotherapy is recommended 0.28 (28%)|
|Proportion of all patients with bladder cancer in whom radiotherapy is recommended 0.58 (58%)|
|Proportion of all patients with testicular cancer in whom radiotherapy is recommended 0.49 (49%)|
|Proportion of all patients with prostate cancer in whom radiotherapy is recommended 0.60 (60%)|
An indication for radiotherapy may occur in the initial stages of treatment or may be delayed (for example in patients who develop local or distant disease recurrence, and who previously have not required radiotherapy as part of the original management). Patients requiring external-beam radiotherapy were counted only once, even if they had multiple indications for radiotherapy at different stages of their illness. This allows comparison of the optimal rate with the actual radiotherapy utilization rate (defined as the number of patients treated by radiotherapy for the first time divided by the incidence of specific cancers during a defined period).
We collected epidemiologic data on the proportion of patients who had attributes for which radiotherapy may be indicated and ranked the relative quality of epidemiologic data using the hierarchy shown in Table 2. The source with the highest quality ranking was used to determine the incidence of each radiotherapy indication. Australian epidemiologic data (from Australian national and state cancer registries) were used whenever possible because the results from the current study will be used specifically to plan future radiotherapy facilities in Australia. The trees and the epidemiologic data were sent for external review to a multidisciplinary panel of cancer experts (from surgical, medical oncology, radiation oncology, palliative care, and nursing backgrounds).
|Population/subpopulation of interest||Proportion represented by this population||Quality of informationa||Reference no.|
|B10||0.0–0.47||μ, β||Sensitivity analysis|
|T15||0.25||λ||69–72 (pooled data)|
|P5||0.0–0.22||γ||36;79–93 (pooled data)|
|P15||0.35||γ||36;79–93 (pooled data)|
Optimal Radiotherapy Utilization Rates
The optimal utilization rate was calculated by determining the incidence for each indication for radiotherapy. Summing these incidences gives the total optimal utilization rate for each genitourinary malignancy studied. This optimal utilization rate was then compared by histology and stage with current utilization data obtained from patterns of care studies.
Combined, renal cancer, bladder cancer, prostate cancer, and testicular cancer represent 19% of all registered cancers in Australia.26 The optimal radiotherapy utilization trees for these cancers are shown in Figures 1–4. Each branch of the tree signifies an attribute that impacts a management decision (e.g., tumor stage and whether clear surgical margins are present). Above each branch in the decision tree is a description of the specific attribute that has led to the treatment decision. The number below each branch signifies the proportions of the attribute based on epidemiologic data.
When considering all of the genitourinary cancers combined, there were 83 possible outcomes or end branches for these radiotherapy treatment trees. We identified 50 outcomes or end branches where radiotherapy would be considered (7 for renal cancer, 13 for bladder cancer, 15 for prostate cancer, and 15 for testicular cancer). Table 2 identifies the clinical situations in which radiotherapy is recommended, the level of evidence, and the guideline or source of evidence for the recommendation. The outcome numbers in Table 1 correspond to terminal nodes or outcome positions in the tree. The last column of Table 1 represents the incidence of each clinical indication for radiotherapy as a proportion of all patients with that specific genitourinary malignancy. This is calculated by multiplying the incidences for each branch leading to this terminal branch. The sum of this column is the optimal radiotherapy utilization rate for each genitourinary malignancy. This was calculated to be 58% for bladder cancer, 49% for testicular cancer, 27% for renal cancer, and 60% for prostate cancer. Table 2 shows the epidemiologic data corresponding to each branch point, and the source of that data as well as the hierarchical level of the data obtained.
Whenever there was uncertainty in either a treatment guideline recommendation or in the accuracy of the epidemiologic incidence data, sensitivity analysis was conducted to assess the impact of this uncertainty on the optimal radiotherapy utilization rate. This allows an assessment of the robustness of the overall estimate when faced with uncertainty. For each treatment recommendation or data uncertainty, variables were inserted into the tree so that sensitivity analysis could be performed. Monte Carlo simulations were conducted to assess the impact of varying the values of each of these variables simultaneously on the overall estimate.
There were 7 variables for which 2 or more epidemiologic sources of data, equivalent in terms of data source quality, differed in their incidence values by > 10%. They include 1) the proportion of patients who develop metastatic disease after nephrectomy for renal cancer (range, 23–58%)27, 28; 2) the proportion of patients with metastatic renal cancer who develop brain metastases (range, 7–19%)28, 29; 3) the proportion of patients with metastatic bladder cancer who develop brain metastases (range, 1–12%)30, 31; 4) the proportion of patients with metastatic bladder cancer who develop bone metastases (range, 17–43%)30, 32; 5) the proportion of patients with early prostate cancer who are treated conservatively who later require radiotherapy for disease progression (range, 7–26%);33, 34 and 6) the proportion of patients who are treated with definitive surgery for prostate cancer who later develop distant disease recurrence (range, 4–15%).35, 36
In addition, there were several instances where the various treatment guidelines differed in their recommendation as to when (or whether) radiotherapy should be used. These included the following: 1) whether patients with renal cancer with positive margins after nephrectomy should receive radiotherapy (radiotherapy is not recommended in this tree but sensitivity analysis included the possibility that radiotherapy is used); 2) whether radiotherapy should be recommended in the treatment of an isolated local disease recurrence after nephrectomy for renal cancer (sensitivity analysis was conducted to assess the effect of both options on the utilization rate); 3) whether patients with metastatic renal cancer who have symptoms of primary disease should be treated with radiotherapy or surgery (sensitivity analysis conducted); 4) whether the role of radiotherapy is superior to cystectomy in the management of patients with Stage II–III bladder cancer (as there is no definitive randomized evidence showing that either treatment modality is superior to the other, both options have been modeled in the radiotherapy utilization tree); 5) whether radiotherapy should be indicated for patients with bulky Stage II or Stage III seminoma who have residual masses after chemotherapy (not recommended in the tree but sensitivity analysis was conducted); 6) whether radiotherapy should be indicated for patients with Stage IV seminoma who have persistent disease after chemotherapy (not recommended in the tree but sensitivity analysis was conducted to assess the effect of both options on the final utilization rate); 7) there is no conclusive evidence to prove the superiority of any one of the treatment options (i.e., surgery or radiotherapy or observation) in the management of early prostate cancer (the guidelines do not make specific treatment recommendations but state that each of these modalities could be considered reasonable). Hence, instead of optimal figures, actual radiotherapy utilization data have been used in the tree. For example, the proportion of patients with early prostate cancer [T1N0M0 and T2N0M0] who undergo surgery was set at a range of 10–70%, modeling the two extremes of practice); 8) whether radiotherapy should be indicated for patients with positive margins after radical prostatectomy (radiotherapy is recommended in the tree but sensitivity analysis has been conducted to assess the effect of both options on the final utilization rate).
Monte Carlo simulations for renal cancer, bladder cancer, and prostate cancer were performed. For data uncertainties where various trial data sets were used, the available epidemiologic data were used to calculate beta distributions using Fast*pro software (version 1.8; Academic Press, San Diego, CA). For most conventional Bayesian calculations of differing data sets, it is usually assumed that these data follow a beta distribution.37 These distribution calculations were based on the sample size and the proportion data quoted in the original article. Variations in the radiotherapy treatment indication due to guideline uncertainty could not be modeled and, therefore, the preferred practice as recommended by the majority of treatment guidelines was used in the Monte Carlo analysis. Using this approach, the optimal radiotherapy utilization rate for each of the cancers where Monte Carlo analysis was performed was 25.0% (95% confidence interval [CI], 8.9–%46.1%) for kidney cancer, 57.0% (95%CI, 53.9–%60.6%) for bladder cancer, and 61.3% (95%CI, 58.0–%65.4%) for prostate cancer based on 10,000 simulations. In testicular cancer, the incidence data did not vary significantly among studies and therefore Monte Carlo analysis was not conducted.
Monte Carlo analysis only can be performed when the distribution of the variable data can be calculated. However, where there is uncertainty about a treatment recommendation, then the two possible extremes may be modeled by univariate analysis where the variables associated with the treatment recommendation are varied in value between two extremes. This assesses the effect of the uncertainty in treatment recommendation on each of the estimates of optimal radiotherapy utilization. Univariate analysis showed that despite quite wide variations in some of the variables, the overall impact on the radiotherapy utilization estimate for each of the genitourinary cancers is relatively small. The range of possible radiotherapy utilization rates for each of the cancers based on the univariate tornado analyses are 25–35% for kidney cancer, 57–67% for prostate cancer, 44–58% for bladder cancer, and 49–50% for testicular cancer.
In the current study, we found that, optimally, 58% of patients with bladder cancer, 49% of patients with testicular cancer, 27% of patients with renal cancer, and 60% of patients with prostate cancer should receive radiotherapy, according to evidence-based treatment guidelines for the various genitourinary malignancies. However, there are several areas in the management of genitourinary malignancies that remain uncertain and this leads to wide confidence limits on this estimate when Monte Carlo analysis is performed. Further study and refinement of the management decisions for genitourinary malignancies need to be made before a more accurate estimate can be calculated. In addition, better epidemiologic data would also assist with making the estimate more accurate.
This model of optimal radiotherapy utilization can be modified readily if required (e.g., if there are changes in cancer incidence, stage distribution, or treatment recommendations) using the TreeAge Data software (Tree Age Software, Inc., Williamstown, MA) that was used to construct the trees. For prostate cancer, epidemiologic data that have become available since the conclusion of our study have shown a change in stage distribution as a result of increased prostate-specific antigen screening. We have published the results of our original study. However, the model can be modified easily in the future to take into account such changes.
Comparison to Actual Practice
Data regarding the actual radiotherapy utilization rates for genitourinary cancers are available from the United States (Surveillance, Epidemiology and End Results [SEER] program,),38 Australia (South Australia),39 Sweden,40 and the United Kingdom (Northern and Yorkshire Cancer Registry and Information Service).41 These were compared with the recommended optimal rates by tumor site in Table 3.
|Disease||Optimal radiotherapy utilization rate (%)||Actual radiotherapy utilization rates|
|SEER (%)38||South Australian Hospital-Based Cancer Registry (%)39||Sweden (%)40||Northern and Yorkshire Cancer Registry (%)41|
Table 3 shows that actual utilization rates fall far short of the optimal recommended rates for all of the tumor sites (for bladder cancer, the actual utilization rate from SEER is less than one-tenth of the recommended rate). For bladder and renal cancers, the actual radiotherapy utilization rates derived from SEER have decreased over time. One surprising result was that the actual rate of radiotherapy utilization for renal cancer was reported to be 63% in Sweden, which is well above the other databases and also above the optimal rate calculated in the current study (27%). The Swedish radiotherapy utilization report does not provide sufficient detail for more extensive review of this discrepancy.40
Further study is required to determine the reasons why such low utilization rates are prevalent for some of these cancers when compared with the ideal situation. Possible explanations include inadequate access to treatment resources, reduced referral rates because of a lack of agreement concerning the value of radiotherapy, a lack of sufficiently good quality evidence to support radiotherapy for some referring physicians, a bias away from adjuvant radiotherapy, patient and physician concern regarding radiotherapy toxicity, and a lack of agreement among referring physicians that the guideline recommendations are appropriate.
We recommend that the optimal genitourinary radiotherapy utilization rates, as described in the current study, should be validated by comparing the optimal rate with actual radiotherapy utilization rates in areas that have adequate radiotherapy resources to assess discrepancies between the model and actual practice. If the evidence for the guideline recommendations is considered to be adequate by the committees responsible for writing national guidelines, then strategies to improve the evidence-based use of radiotherapy in genitourinary malignancies are recommended. The lack of Level I and II evidence to support the use of radiotherapy for various indications in genitourinary cancer also needs to be addressed in future trials. Further epidemiologic studies regarding the use of radiotherapy for genitourinary malignancies are required.
The authors thank Associate Professor Gill Duchesne, Dr. Sandra Turner, Dr. Joseph Bucci, Dr. Keen Hun Tai, Dr. Martin Berry, Associate Professor Alan Rodger, Associate Professor David Currow, and the members of the steering committee of the Australian National Cancer Control Initiative for their helpful comments regarding the study design and decision trees. They also thank Associate Professor David Roder for providing valuable epidemiologic data.
- 8National Cancer Institute. PDQ cancer information summaries: treatment of renal cell cancer. Available from URL: www.nci.nih.gov [accessed August 29, 2002].
- 9BC Cancer Agency. Cancer management guidelines: renal cancer. Available from URL: http://www.bccancer.bc.ca [accessed August 29, 2002].
- 10National Comprehensive Cancer Network. Clinical practice guidelines in oncology—v.1.2003—kidney cancer. Available from URL: www.nccn.org [accessed August 5, 2003].
- 11National Cancer Institute. PDQ cancer information summaries: treatment of bladder cancer. Available from URL: www.nci.nih.gov [accessed August 29, 2002].
- 12BC Cancer Agency. Cancer management guidelines: bladder cancer. Available from URL: http://www.bccancer.bc.ca [accessed August 29, 2002].
- 14National Comprehensive Cancer Network. Practice guidelines in oncology—v.1.2002—bladder cancer. Available from URL: www.nccn.org [accessed October 31, 2002].
- 15National Cancer Institute. PDQ cancer information summaries: treatment of testicular cancer. Available from URL: www.nci.nih.gov [accessed August 29, 2002].
- 16BC Cancer Agency. Cancer management guidelines: management of testicular cancer. Available from URL: http://www.bccancer.bc.ca [accessed August 28, 2002].
- 17Royal College of Radiologists COIN (Clinical Oncology Information Network) Scottish Intercollegiate Guidelines Network. Guidelines on the management of adult testicular germ cell tumours. Available from URL: http://www.rcr.ac.uk/testicular.htm [accessed August 29, 2002].
- 18National Comprehensive Cancer Network. NCCN practice guidelines for testicular cancer. Oncology. 1998; 12: 417–462.
- 21National Cancer Institute. PDQ cancer information summaries: treatment of prostate cancer. Available from URL: www.nci.nih.gov [accessed October 16, 2001].
- 22National Comprehensive Cancer Network. Clinical practice guidelines in oncology—v.1.2002—prostate cancer. Available from URL: www.nccn.org [accessed August 28, 2002].
- 23Royal College of Radiologists COIN (Clinical Oncology Information Network) and British Association of Urological surgeons. Guidelines on the management of prostate cancer. Available from URL: http://www.rcr.ac.uk/prostate.html [accessed October 16, 2001].
- 24American Society for Therapeutic Radiology and Oncology Consensus Panel. Consensus statements on radiation therapy of prostate cancer: guidelines for prostate re-biopsy after radiation and for radiation therapy with rising prostate-specific antigen levels after radical prostatectomy. J Clin Oncol. 1999; 17: 1155–1163.
- 25National Health and Medical Research Council. Guide to the development, implementation and evaluation of clinical practice guidelines. Appendix B. Canberra: National Health and Medical Research Council, 1998: 56.
- 26Australian Institute of Health and Welfare (AIHW) and Australasian Association of Cancer Registries (AACR). Cancer in Australia 1998. CAN 12. 2001. Cancer Series No 17. Canberra: AIHW.
- 37Bayesian statistical modelling. New York: John Wiley & Sons, 2002..
- 38Cancer Statistics Branch, National Cancer Institute. Surveillance, Epidemiology and End Results (SEER) program public-use CD-ROM (1973-1997). Bethesda: National Cancer Institute, 2000.
- 39SA Cancer Registry. Epidemiology of cancer in South Australia (Cancer Series No. 22). Adelaide: South Australian Cancer Registry, 2000.
- 41Northern and Yorkshire Cancer Registry and Information Service (NYCRIS). Northern and Yorkshire Cancer Networks. A report on incidence and management for the main sites of cancer 1999. Available from URL: http://www.nycris.org.uk/ [accessed November 25, 2003].
- 66International Germ Cell Cancer Collaborative Group. International germ cell consensus classification: a prognostic factor-based staging system for metastatic germ cell cancers. J Clin Oncol. 1997; 15: 594–603.