- Email: [email protected]
Patterns of palliative radiotherapy in Canada
Clinical Oncology(1993) 5:92-97 © 1993The RoyalCollegeof Radiologists
Clinical Oncology
Original Article Patterns of Palliative R a d i o t h e r a p y in C a n a d a G. Duncan 1, W. Duncan 2 and E. J. Maher 3 1Division of Radiation Oncology, BCCA, 600 West 10th Avenue, Vancouver, BC V5Z 4E6, Canada, ZWestern General Hospital, Edinburgh, Scotland and 3Regional Centre for Radiotherapy and Oncology, Mount Vernon Hospital, Northwood, Middlesex, UK
Abstract. A questionnaire study was performed to determine patterns of radiotherapy practice across Canada and the relationship of age, work-load, resources, intent and aims of therapy on departmental policy; 99 returns were analysed. Three case scenarios were given: Brain metastases were treated typically without a simulation as a parallel pair with palliative intent to 2000 cGy in 5 fractions; for locally advanced lung cancer most patients were simulated, but 40% we deemed 'radical' and there was a bimodal pattern of dose fractionation (the age of the radiation oncologist appeared to influence this and geographical differences were also notable); bony metastases from breast cancer were typically simulated as a single field using 2000-2500 cGy in 5 fractions and always considered palliative. Work-load and resources are likely to influence patterns of care. For example, over 20% felt that treatment choice was suboptimal, largely due to lack of radiation technologists and megavoltage equipment. The implications of these determinants on palliative care are discussed. Objective criteria and definitions are required to allow satisfactory workload measurements and comparisons to be made. Keywords: Bone metastases; Brain metastases; Locally advanced lung cancer; Palliation; Patterns of care; Radiotherapy; Resources
INTRODUCTION A wide range of attitudes and beliefs, as well as the availability of resources, influence the strategies for the treatment of patients with advanced cancer. Surveys of treatment strategies using three common clinical scenarios have been performed in Europe, the USA and Canada; these have been published individually and in a comparative study elsewhere [1]. This article gives the results of the survey to determine treatment strategies for advanced cancer in Canada and discusses the impact of radiation Correspondence and offprint requests to: Dr G. Duncan, Division
of Radiation Oncology, BCCA, 600 West 10th Avenue, Vancouver, BC, V5Z 4E6, Canada.
oncologists' beliefs, demographics and resources in influencing treatment choices.
MATERIALS AND METHODS At the time of this survey (1990) there were 181 practising radiation oncologists in Canada, of whom 142 were members of the Canadian Association of Radiation Oncologists (CARO). A questionnaire was sent to the 120 CARO members whose full addresses were known. One hundred and six replies were received, 97 of which were complete and suitable for detailed analysis. This represents responses from over 50% of all Canadian radiation oncologists. Respondents were asked to 'tick' a box alongside a stated response. In addition, information was sought on the respondents' age, palliative workload, funding, attitudes towards optimal management and resources. A detailed description of resources and work-load was also provided by the head of radiation oncology at each cancer centre and matched to other radiation oncologists replying from that centre. Department heads were not, however, asked whether treatment strategies were considered to be optimal or not. Statistical comparisons were made using contingency tables, generally with a Chi-square test, using Yates's correction when required. In situations where group sizes were less than 20 or events less than 5, Fisher's exact test was employed. The questionnaire has been described in more detail elsewhere [1]. There were three clinical case scenarios: 1. Patient A, a 64-year-old man, previously treated for limited stage small cell lung cancer (post-chemotherapy and mediastinal radiotherapy) develops multiple cerebral metastases with headaches and vomiting, and is responding to steroids. 2. Patient B, a 59-year-old man with a 4 cm hilar mass and positive mediastinoscopy is complaining of haemoptysis. 3. Patient C, a 64-year-old woman treated 2 years previously for T2Nx breast cancer develops postmastectomy non-radicular pain in the lower thoracic
Patterns of Palliative Radiotherapy in Canada
93
spine from a bone metastasis confirmed radiographically; she is on tamoxifen. All radiation oncologists who questionnaire specified:
completed
Table 1. Attitude and general management of three patients A (brain)
the
1. Radiation treatment techniques, dose and fractionation. 2. The perceived aims in treating the patient (for example to improve survival, relieve symptoms, give hope, prevent symptoms) and the intent (radical/ palliative). 3. The perceived prognosis both for the percentage considered likely to be cured and the estimated survival time. 4. A n y additional care that the radiation oncologist would r e c o m m e n d at that time. 5. A c o m m e n t on who was most likely to be primarily involved in terminal care.
RESULTS
(%)
Treat Yes No Simulate Yes No Plan Opposed Single Multiple Treatment intent Radical Palliative Aims Extend life Prevent symptoms Relieve symptoms Give hope % likely to be cured None <10% 11-20 %
Overview/Demographics The 97 analysed replies were received from 23 Canadian radiation oncology centres. Twelve department heads, who worked at nine centres with 44 other radiation oncologists, completed the section on resources and palliative care work-load. This allowed 44 questionnaires to be analysed against data describing the radiotherapeutic resources and demographic criteria of these nine centres. The replies of 83 radiation oncologists were analysed by their opinion of whether the chosen treatment strategy was perceived to be optimal or not. Responses from d e p a r t m e n t heads were excluded from this analysis. All other analysis of the m a n a g e m e n t of the three case scenarios were based on the data provided in 97 returned questionnaires. All respondents were staff (consultant) radiation oncologists. It is noted that 63% were under 46 years of age. In Canada, radiation oncology services are almost exclusively government funded. The overall palliative service work-load was stated to be less than 50% of the overall work-load of the departments. The attitude to and general m a n a g e m e n t of the three cases in Canada is outlined in Table 1. Some general points about m a n a g e m e n t m a y be made. Few radiation oncologists chose not to advise radiotherapy. Simulation was not often p e r f o r m e d for the treatment of brain metastases. The radiotherapy of locally advanced lung cancer was d e e m e d 'radical' in 40% of responses. The most c o m m o n objective in all three case scenarios was to relieve symptoms and the second commonest was to prevent symptoms. Whilst radiation oncologists considered that the patient with locally advanced lung cancer had the highest likelihood of being cured (although less than 10%), the life expectancy predicted was only between 7 and 24 months. The use of c h e m o t h e r a p y as an additional treatment was u n c o m m o n but most notable in metastatic breast cancer to bone in about 20% of respondents. A uniform pattern was described of terminal
Don't know Life expectancy <6 months 7-12 months 1-2 years 2-5 years >5 years Don't know Give other treatment Chemotherapy Other None Terminal management Self Medical oncologist Family doctor Other, NOS
B (lung)
(%)
C (bone)
(%)
99.0 1.0
100.0 -
91.8 8.2
35.4 64.6
99.0 1.0
98.0 1.1
100.0 -
63.9 36.1
3.3 94.5 2.2
100.0
40.2 69.1
100.0
21.9 64.6 78.1 24.0
37.1 57.7 85.6 32.0
1,1 47.3 100.0 15.4
88.7 11.3
36.1 57.7
90.6 3.1
-
6.2
2.1
-
-
4.2
80.4 19.6 -
1.0 46.4 45.4 4.1 3.1
6.5 35.5 45.1 2.1 10.8
2.1 18.6 79.3
4.1 11.3 84.6
19.8 28.1 52.1
31.0 10.0 62.0 5.0
37.0 2.0 60.0 14.0
38.0 10.0 54.0 9.0
care being p e r f o r m e d predominantly by the family doctor. The oncologist was also highly likely to be integrally involved in this c o m p o n e n t of management.
P a t i e n t A: S m a l l C e l l C a n c e r M e t a s t a t i c to Brain
Ra~o~er~y Almost two-thirds of the radiation oncologists surveyed chose not to simulate the treatment fields and setup for the patient with brain metastases. In all instances a parallel pair of fields was chosen. The objective of treatment was always regarded as palliative. The majority of respondents (68%) employed 20 Gy in 5 fractions over a week; 30 G y in 10 fractions was also frequently prescribed (26%). T h e r e was, however, a large variation in dose fraction schedules ranging from 10 Gy in a single fraction to 40 Gy delivered in 15 fractions. Radiation oncologists who believed that the patient had a shorter than average life expectancy were more likely to select a lower dose and shorter fractionation schedule (P = 0.049). A lower dose and
G. Duncan, W. Duncan and E. J. Maher
94
shorter fractionation schedule was also chosen by those who treated a higher than average number of palliative care patients each year (P = 0.015).
Table 2. Dose-~actionationfor patient B
Dose(cGy)(%)
Fractions(%)
Doseper~action
(cGy)(%)
Aims Over three-quarters of radiation oncologists faced with this clinical scenario aimed to relieve symptoms and almost two-thirds to prevent symptoms. Those in departments with less than the average palliative service work-load were less likely to aim to improve survival with treatment (P = 0.045).
2000 (25.8) 3000(22.7) 4000-4500 (14,4) 5000-5500 (12.4) 6000 (24,7)
5 (25.8) 10 (24.7) 15 (9,3) 20-22 (15.5) 25-30 (24.7)
Age of radiation oncologist (years)
No. fractions
Prognosis Almost 90% of radiation oncologists considered that no patient with brain metastases from small cell lung cancer would be cured. Over 80% felt that the life expectancy would be less than 6 months and none felt it would be longer than 12 months.
24-45 >46
5-9
10-19
20-29
30
23 2
16 17
5 11
17 6
(P
No. fractions
Additional Care Almost 80% of respondents felt that no additional treatment should be advised. Terminal care was likely to be carried out by the family doctor in 62% and in 31% by the oncologist. Radiation oncologists were more likely to be involved in the terminal care of these patients in departments to which were referred a higher than average number of new patients each year (P = 0.018).
Patient B: Locally Advanced Lung Cancer Radiotherapy Almost all of these patients were planned on a treatment simulator. Most radiation oncologists (64%) used a parallel opposed pair of fields, but 36% stated they would employ multiple fields. The use of a parallel pair of fields was significantly correlated with a higher departmental work-load (P = 0.037). The selection of multiple fields was correlated with the prescription of higher doses and a larger number of fractions, and also with a radical intent and a perceived better life expectancy. Overall there was a bimodal distribution of selected dose fractionation schedules. The low dose group selected either 20 Gy in 5 fractions or 30 Gy in 10 fractions, whilst the high dose group typically prescribed 60 Gy in 30 fractions. The overall range of dose was from 20 Gy in 5 fractions to 60 Gy in 30 daily fractions. It was observed that the bimodal pattern of prescribing was only implemented by the radiation oncologists under 46 years of age, whereas those above 46 years of age tended to select more intermediate dose fractionation schedules (P < 0.001) (Table 2). There also appeared to be a regional difference in the choice of dose fractionation schedules and whether the treatment was designated as palliative (Table 3). It was noted that the selected dose fractionation was significantly determined by departmental policy (P = 0.023). A total of 20% of radiation oncologists treating
200 (27,8) 200-300 (44.3) >/400 (27.8)
<20 Radical Palliative
2 56
20--29
30
13 4
20 4
(P
Table 3. Regional differences in designation of treatment for
patient B as palliative (P<0,0005) Area
%
Atlantic Quebec Ontario Prairies BC
62.5 50.0 70.8 55.6 92.3
patient B chose a radical treatment, usually prescribing a dose given in 30 fractions over 6 weeks, but 56% chose a palliative treatment, typically using less than 20 daily fractions. For the same dose fractionation some deemed it radical and others palliative. This was true for doses >--4000 cGy up to 6000 cGy (Table 4), although above 4000 cGy the majority favoured radical treatment for this scenario. It is also interesting to note that three replies were stated to be both palliative and radical; this highlights the difficulties inherent in the loose 'definitions' of these descriptions of treatment 'intent'. Radiation oncologists who felt that they were primarily aiming to relieve symptoms selected a Table 4. Relationship of dose to treatment intent in patient B
Dose (cGy)
Radical (No.)
Palliative (No.)
2000 3000 4000 4500 5000 5500 6000
0 0 3 3 8 1 21
25 22 8 1 4 0 4
Patterns of PalliativeRadiotherapyin Canada lower dose/shorter ffactionation schedule (P = 0.013), whereas those who felt that they were aiming to improve survival selected higher dose/longer fractionation schedules (P < 0.000001). It was observed that the selected dose fractionation was directly related to the availability of megavoltage equipment and to the departmental work-load.
Aims Overall, 85% of radiation oncologists aimed to relieve symptoms and 58% to prevent symptoms. It was noted that those who did not aim to improve survival were more likely to aim to alleviate symptoms. Those who resolved to prevent symptoms were unlikely also to aim to give hope to the patient. The aim only to relieve symptoms was associated with a poorer perceived prognosis and correlated with a smaller percentage cure probability (P < 0.001) and a shorter life expectancy (P = 0.00066).
Prognosis Departmental policy was signficantly correlated with the expected percentage of patients likely to be cured (P = 0.041). Younger radiation oncologists tended to predict a shorter expected survival (P = 0.02).
95
Aims All radiation oncologists aimed to relieve symptoms for this patient with metastatic breast cancer. Those who aimed to prevent symptoms also tried to give hope (P = 0.007) and predicted a better life expectancy (P = 0.009). They were also more likely to feel that there were inadequate numbers of inpatient beds for the management of these patients (P = 0.032). Radiation oncologists who aimed to give hope to the patients were also more likely to recommend additional treatment, the difference being highly significant (P = 0.00007). They were also more likely to feel that current practice was suboptimal (P = 0.034) because of a lack of simulators (P = 0.017) and beds (P = 0.038).
Prognosis It was observed that 91% of radiation oncologists felt that this patient had no likelihood of being cured. In spite of this, there was a wide range of predicted life expectancies.
Additional Care
It was observed that additional treatment was more likely to be offered by radiation oncologists who would not expect to be involved in terminal care (P = 0.048). Furthermore, the possibility of the radiation oncologist being involved with terminal care was directly related to the departmental work-load (P = 0.018), those in larger departments being less likely to be involved.
About 20% of respondents suggested that chemotherapy would be appropriate for additional management. Interestingly, 28% considered that other forms of management should be sought. It was stated that terminal care would be performed by the family doctor in 54% of cases; 38% of radiation oncologists said that they would be predominantly involved with terminal care. It was interesting to note that there was a greater probability that the radiation oncologist would be involved with terminal care if the department was large (P = 0.023) and if a greater annual number than average of new patients was seen (P = 0.036).
Patient C: Breast Cancer Metastatic to Bone
Correlations Between Case Scenarios
Radiotherapy
Radiotherapy
Additional Care
The majority of radiation oncologists felt that the radiotherapy for this patient should be simulated. It was unaminously agreed that treatment was palliative in intent and that a single field should be used. Just over half (57%) prescribed 20 Gy in 5 fractions and occassionally 25 Gy in 5 fractions (over I week). The schedule prescribed ranged from 8 Gy in a single fraction to 30 Gy in 10 fractions. Surprisingly, only 15% used single fractions in spite of trial evidence supporting this policy. The RTOG study of palliation of bone metastases indicated that low dose, short course schedules were as effective as high dose, protracted programmes [2], although a re-analysis by Blitzer challenged this conclusion [3]. Data from Price et al. [4] and Cole [5] support the shorter regimen. The use of a single field (as opposed to more complex field arrangement) as well as the use of a larger number of fractions was directly related to the aim of preventing symptoms (P = 0.045, P = 0.02, respectively).
Radiation oncologists who chose a lower dose and fewer fraction schedules tended to do so for all three cases, but there was a particularly strong correlation when patients A and B were compared (P ~ 0).
Aims There was consistency in the objectives to prevent symptoms or give hope in all three case scenarios (Table 1). By comparison, there was no significant correlation between cases when the stated aim was to improve survival or relieve symptoms.
Prognosis Radiation oncologists who considered that few scenario A patients were likely to be cured were shown to feel similarly for scenario C (P = 0.0055). The anticipated life expectancy for patient C was significantly longer than that for patients A or B, and, in addition, a greater variation of life expectancy was stated for patient C (Table 1).
96
Additional Care Those who advised the use of additional treatment, other than chemotherapy, for patient B also considered that similar treatments should be advised for patient C (P = 0.0005).
Resource Impact Departmental heads were purposely not asked to comment if they considered the treatment routinely prescribed in their department was considered to be optimal or not. Sixty-five (78.4%) of staff radiation oncologists felt that the treatment they prescribed was satisfactory for these patients, but 18 (21.6%) were concerned that their treatment policies were not ideal. In order of decreasing concern they complained of: lack of technologists (16), lack of megavoltage equipment (11), departmental policies (6), shortage of beds (5), and inadequate numbers of simulators (4). It was clear that these perceived deficiencies of resources are inter-related. However, there was no significant correlation between a perception of having suboptimal resources and the actual resources available in those departments. This may be due to the small numbers of respondents in each category.
DISCUSSION Twenty-two percent of Canadian radiation oncologists consulted had concerns that treatment was compromised by lack of resources such as technologists or megavoltage equipment. It was clear that radiation oncology departments with a heavy palliative work-load tended to employ shorter dose fractionation schedules. The lack of linear accelerators and cobalt therapy machines was also related to the selection of lower dose fractionation schedules. The major determinant, however, of palliative treatment should be the lowest dose/shortest fractionation schedule which will provide optimal control of symptoms with minimal acute morbidity. Late morbidity is unlikely to be a problem, but tolerance doses of radiation should never be exceeded. The simplest technique should always be preferred if the same result can be achieved. The RTOG studies, for example, suggested that, for most patients with brain metastases, 2000 cGy in 5 fractions was as effective as 4 weeks of higher dose regimens [6,7]. Analysis of the Canadian radiotherapy practice for locally advanced lung cancer shows that radiation oncologists under the age of 46 select either a very short dose fractionation schedule or a 6-week programme, whereas those over 46 years of age chose relatively evenly from 5 up to 30 fractions. This may be accounted for by changing work-loads with age [8], initial training, subsequent work experience, or becoming more work-load conscious, or by altered perceptions of management with increasing age. The interprovincial differences in treatment for locally advanced lung cancer may be due in part to age,
G. Duncan, W. Duncan and E. J. Maber
departmental policy, or training. Regional differences in the use of single fractions versus multiple fractions have been noted in a UK survey of treatment for bone metastases [8] as likely to be predominantly a result of training. In this Canadian survey, however, no interprovincial difference in fractionation for bone metastases was seen. There is a lack of consensus in the management of non-small cell lung cancer [9,10]; policy may be based on what doctors would choose for themselves [10]. Sixty percent of radiation oncologists in Canada treat locally advanced lung cancer with less than 20 fractions, compared with 60% in the USA choosing a 30 fraction treatment schedule. These differences have considerable economic consequences and deserve further study. Patterns of practice established in the UK and the USA serve as a useful baseline [8,11]. In Canada 9% of respondents felt that the radiotherapy was both radical and palliative. In several instances the same dose fractionation was deemed radical by one oncologist but palliative by another. This highlights the confusion and subjectivity of present 'definitions'. In order to establish cost effective treatment strategies, more precise definitions are essential. 'Radical' is usually employed to refer to a higher dose/longer fractionation treatment and the potential consequences. Definitions may be organ tolerance based or empirical as suggested in Table 5. 'Curative', (not 'radical') therapy aims to improve survival and is a function of a tumour's inherent radiocurability. 'Palliative' should refer solely to the aims of relieving or preventing symptoms (or perhaps to give hope). Thus, curative therapy aims to improve survival, and palliative to improve quality of survival. We believe it would be helpful in addressing critical work,load and resource questions internationally if definitions were agreed. Table 5. Definitions of 'curative' and 'palliative' treatment
Curative
Palliative
Increase survival ( + / - improving symptoms) 'Quantity of Life'
Improve symptoms only 'Quality of Life'
The goal in palliation must remain to return the patient to a comfortable, functional life if possible, in an expeditious manner, whilst considering the economic consequences [12]. Studies on palliation, therefore, must define costs as well as measuring quality of life [13]. The answers regarding optimal management can only be derived from randomized clinical trials. Hopefully more such studies in palliative care will be conducted.
References 1. Maher EJ, Coia LR, Duncan GG, et al. Treatment strategies in advanced and metastatic cancer: Differences in attitude between the USA, Canada and Europe. Int J Radiat Oncol Biol Phys 1992;23:23944.
97
Patterns of Palliative Radiotherapy in Canada 2. Tong D, Gillick L, Hendrickson FR. The palliation of symptomatic osseous metastases: Final results of the study by the Radiation Therapy Oncology Group. Cancer 1982;50:893-9. 3. Blitzer PH. Re-analysisoftheRTOGstudyofthepalliationof symptomatic osseous metastasis. Cancer 1985;55:1468-72. 4. Price P, Hoskin PJ, Easton D, et al. Prospective randomized trial of single and multifraction radiotherapy schedules in the treatment of painful bony metastases. Radiat Oncol 1986;6:247-55. 5. Cole DJ. A randomized trial of a single treatment versus conventional fractionation in the palliative radiotherapy of painful bone metastases. Clin Oncol 1989;1:5%62. 6. Borgelt B, Gelber R, Kramer S, et al. The palliation of brain metastases: Final results of the first two studies by the Radiation Therapy Oncology Group. Int J Radiat Oncol Biol Phys 1980;6:1-9. 7. Kurtz JM, Gelber R, Brady LW, et al. The palliation of brain metastases in a favourable patient population: A randomized
8. 9. 10. 11. 12. 13.
clinical trial by the Radiation Therapy Oncology Group. Int J Radiat Oncol Biol Phys 1981;7:891-5. Crellin AM, Marks A, Maher EJ. Why don't British radiotherapists give single fractions of radiotherapy for bone metastases? Clin Oncol 1989;1:63-6. MacKillop WJ, O'Sullivan B, Ward GK. Non-small cell lung cancer: How oncologists want to be treated. Int J Radiat Oncol Biol Phys 1987;13:929-34. Palmer MJ, O'Sullivan B, Steele R, et al. Controversies in the management of non-small cell lung cancer: The results of an expert surrogate study. Radiat Oncol 1990;19:17-28. Coia LR, Hanks GE, Martz K, et al. Practice patterns of palliative care for the United States, 1984-1985. Int J Radiat Oncol Biol Phys 1988;14:1261-9. Hendrickson FR. Strategyofpalliative treatment. IntJRadiat Oncol Biol Phys 1982;8:155-6. Bowen OR. Shattuck lecture: What is quality care? N Engl J Med 1987;316:1578-9.
Received for publication August 1992 Accepted following revision December 1992
Clinical Oncology
Original Article Patterns of Palliative R a d i o t h e r a p y in C a n a d a G. Duncan 1, W. Duncan 2 and E. J. Maher 3 1Division of Radiation Oncology, BCCA, 600 West 10th Avenue, Vancouver, BC V5Z 4E6, Canada, ZWestern General Hospital, Edinburgh, Scotland and 3Regional Centre for Radiotherapy and Oncology, Mount Vernon Hospital, Northwood, Middlesex, UK
Abstract. A questionnaire study was performed to determine patterns of radiotherapy practice across Canada and the relationship of age, work-load, resources, intent and aims of therapy on departmental policy; 99 returns were analysed. Three case scenarios were given: Brain metastases were treated typically without a simulation as a parallel pair with palliative intent to 2000 cGy in 5 fractions; for locally advanced lung cancer most patients were simulated, but 40% we deemed 'radical' and there was a bimodal pattern of dose fractionation (the age of the radiation oncologist appeared to influence this and geographical differences were also notable); bony metastases from breast cancer were typically simulated as a single field using 2000-2500 cGy in 5 fractions and always considered palliative. Work-load and resources are likely to influence patterns of care. For example, over 20% felt that treatment choice was suboptimal, largely due to lack of radiation technologists and megavoltage equipment. The implications of these determinants on palliative care are discussed. Objective criteria and definitions are required to allow satisfactory workload measurements and comparisons to be made. Keywords: Bone metastases; Brain metastases; Locally advanced lung cancer; Palliation; Patterns of care; Radiotherapy; Resources
INTRODUCTION A wide range of attitudes and beliefs, as well as the availability of resources, influence the strategies for the treatment of patients with advanced cancer. Surveys of treatment strategies using three common clinical scenarios have been performed in Europe, the USA and Canada; these have been published individually and in a comparative study elsewhere [1]. This article gives the results of the survey to determine treatment strategies for advanced cancer in Canada and discusses the impact of radiation Correspondence and offprint requests to: Dr G. Duncan, Division
of Radiation Oncology, BCCA, 600 West 10th Avenue, Vancouver, BC, V5Z 4E6, Canada.
oncologists' beliefs, demographics and resources in influencing treatment choices.
MATERIALS AND METHODS At the time of this survey (1990) there were 181 practising radiation oncologists in Canada, of whom 142 were members of the Canadian Association of Radiation Oncologists (CARO). A questionnaire was sent to the 120 CARO members whose full addresses were known. One hundred and six replies were received, 97 of which were complete and suitable for detailed analysis. This represents responses from over 50% of all Canadian radiation oncologists. Respondents were asked to 'tick' a box alongside a stated response. In addition, information was sought on the respondents' age, palliative workload, funding, attitudes towards optimal management and resources. A detailed description of resources and work-load was also provided by the head of radiation oncology at each cancer centre and matched to other radiation oncologists replying from that centre. Department heads were not, however, asked whether treatment strategies were considered to be optimal or not. Statistical comparisons were made using contingency tables, generally with a Chi-square test, using Yates's correction when required. In situations where group sizes were less than 20 or events less than 5, Fisher's exact test was employed. The questionnaire has been described in more detail elsewhere [1]. There were three clinical case scenarios: 1. Patient A, a 64-year-old man, previously treated for limited stage small cell lung cancer (post-chemotherapy and mediastinal radiotherapy) develops multiple cerebral metastases with headaches and vomiting, and is responding to steroids. 2. Patient B, a 59-year-old man with a 4 cm hilar mass and positive mediastinoscopy is complaining of haemoptysis. 3. Patient C, a 64-year-old woman treated 2 years previously for T2Nx breast cancer develops postmastectomy non-radicular pain in the lower thoracic
Patterns of Palliative Radiotherapy in Canada
93
spine from a bone metastasis confirmed radiographically; she is on tamoxifen. All radiation oncologists who questionnaire specified:
completed
Table 1. Attitude and general management of three patients A (brain)
the
1. Radiation treatment techniques, dose and fractionation. 2. The perceived aims in treating the patient (for example to improve survival, relieve symptoms, give hope, prevent symptoms) and the intent (radical/ palliative). 3. The perceived prognosis both for the percentage considered likely to be cured and the estimated survival time. 4. A n y additional care that the radiation oncologist would r e c o m m e n d at that time. 5. A c o m m e n t on who was most likely to be primarily involved in terminal care.
RESULTS
(%)
Treat Yes No Simulate Yes No Plan Opposed Single Multiple Treatment intent Radical Palliative Aims Extend life Prevent symptoms Relieve symptoms Give hope % likely to be cured None <10% 11-20 %
Overview/Demographics The 97 analysed replies were received from 23 Canadian radiation oncology centres. Twelve department heads, who worked at nine centres with 44 other radiation oncologists, completed the section on resources and palliative care work-load. This allowed 44 questionnaires to be analysed against data describing the radiotherapeutic resources and demographic criteria of these nine centres. The replies of 83 radiation oncologists were analysed by their opinion of whether the chosen treatment strategy was perceived to be optimal or not. Responses from d e p a r t m e n t heads were excluded from this analysis. All other analysis of the m a n a g e m e n t of the three case scenarios were based on the data provided in 97 returned questionnaires. All respondents were staff (consultant) radiation oncologists. It is noted that 63% were under 46 years of age. In Canada, radiation oncology services are almost exclusively government funded. The overall palliative service work-load was stated to be less than 50% of the overall work-load of the departments. The attitude to and general m a n a g e m e n t of the three cases in Canada is outlined in Table 1. Some general points about m a n a g e m e n t m a y be made. Few radiation oncologists chose not to advise radiotherapy. Simulation was not often p e r f o r m e d for the treatment of brain metastases. The radiotherapy of locally advanced lung cancer was d e e m e d 'radical' in 40% of responses. The most c o m m o n objective in all three case scenarios was to relieve symptoms and the second commonest was to prevent symptoms. Whilst radiation oncologists considered that the patient with locally advanced lung cancer had the highest likelihood of being cured (although less than 10%), the life expectancy predicted was only between 7 and 24 months. The use of c h e m o t h e r a p y as an additional treatment was u n c o m m o n but most notable in metastatic breast cancer to bone in about 20% of respondents. A uniform pattern was described of terminal
Don't know Life expectancy <6 months 7-12 months 1-2 years 2-5 years >5 years Don't know Give other treatment Chemotherapy Other None Terminal management Self Medical oncologist Family doctor Other, NOS
B (lung)
(%)
C (bone)
(%)
99.0 1.0
100.0 -
91.8 8.2
35.4 64.6
99.0 1.0
98.0 1.1
100.0 -
63.9 36.1
3.3 94.5 2.2
100.0
40.2 69.1
100.0
21.9 64.6 78.1 24.0
37.1 57.7 85.6 32.0
1,1 47.3 100.0 15.4
88.7 11.3
36.1 57.7
90.6 3.1
-
6.2
2.1
-
-
4.2
80.4 19.6 -
1.0 46.4 45.4 4.1 3.1
6.5 35.5 45.1 2.1 10.8
2.1 18.6 79.3
4.1 11.3 84.6
19.8 28.1 52.1
31.0 10.0 62.0 5.0
37.0 2.0 60.0 14.0
38.0 10.0 54.0 9.0
care being p e r f o r m e d predominantly by the family doctor. The oncologist was also highly likely to be integrally involved in this c o m p o n e n t of management.
P a t i e n t A: S m a l l C e l l C a n c e r M e t a s t a t i c to Brain
Ra~o~er~y Almost two-thirds of the radiation oncologists surveyed chose not to simulate the treatment fields and setup for the patient with brain metastases. In all instances a parallel pair of fields was chosen. The objective of treatment was always regarded as palliative. The majority of respondents (68%) employed 20 Gy in 5 fractions over a week; 30 G y in 10 fractions was also frequently prescribed (26%). T h e r e was, however, a large variation in dose fraction schedules ranging from 10 Gy in a single fraction to 40 Gy delivered in 15 fractions. Radiation oncologists who believed that the patient had a shorter than average life expectancy were more likely to select a lower dose and shorter fractionation schedule (P = 0.049). A lower dose and
G. Duncan, W. Duncan and E. J. Maher
94
shorter fractionation schedule was also chosen by those who treated a higher than average number of palliative care patients each year (P = 0.015).
Table 2. Dose-~actionationfor patient B
Dose(cGy)(%)
Fractions(%)
Doseper~action
(cGy)(%)
Aims Over three-quarters of radiation oncologists faced with this clinical scenario aimed to relieve symptoms and almost two-thirds to prevent symptoms. Those in departments with less than the average palliative service work-load were less likely to aim to improve survival with treatment (P = 0.045).
2000 (25.8) 3000(22.7) 4000-4500 (14,4) 5000-5500 (12.4) 6000 (24,7)
5 (25.8) 10 (24.7) 15 (9,3) 20-22 (15.5) 25-30 (24.7)
Age of radiation oncologist (years)
No. fractions
Prognosis Almost 90% of radiation oncologists considered that no patient with brain metastases from small cell lung cancer would be cured. Over 80% felt that the life expectancy would be less than 6 months and none felt it would be longer than 12 months.
24-45 >46
5-9
10-19
20-29
30
23 2
16 17
5 11
17 6
(P
No. fractions
Additional Care Almost 80% of respondents felt that no additional treatment should be advised. Terminal care was likely to be carried out by the family doctor in 62% and in 31% by the oncologist. Radiation oncologists were more likely to be involved in the terminal care of these patients in departments to which were referred a higher than average number of new patients each year (P = 0.018).
Patient B: Locally Advanced Lung Cancer Radiotherapy Almost all of these patients were planned on a treatment simulator. Most radiation oncologists (64%) used a parallel opposed pair of fields, but 36% stated they would employ multiple fields. The use of a parallel pair of fields was significantly correlated with a higher departmental work-load (P = 0.037). The selection of multiple fields was correlated with the prescription of higher doses and a larger number of fractions, and also with a radical intent and a perceived better life expectancy. Overall there was a bimodal distribution of selected dose fractionation schedules. The low dose group selected either 20 Gy in 5 fractions or 30 Gy in 10 fractions, whilst the high dose group typically prescribed 60 Gy in 30 fractions. The overall range of dose was from 20 Gy in 5 fractions to 60 Gy in 30 daily fractions. It was observed that the bimodal pattern of prescribing was only implemented by the radiation oncologists under 46 years of age, whereas those above 46 years of age tended to select more intermediate dose fractionation schedules (P < 0.001) (Table 2). There also appeared to be a regional difference in the choice of dose fractionation schedules and whether the treatment was designated as palliative (Table 3). It was noted that the selected dose fractionation was significantly determined by departmental policy (P = 0.023). A total of 20% of radiation oncologists treating
200 (27,8) 200-300 (44.3) >/400 (27.8)
<20 Radical Palliative
2 56
20--29
30
13 4
20 4
(P
Table 3. Regional differences in designation of treatment for
patient B as palliative (P<0,0005) Area
%
Atlantic Quebec Ontario Prairies BC
62.5 50.0 70.8 55.6 92.3
patient B chose a radical treatment, usually prescribing a dose given in 30 fractions over 6 weeks, but 56% chose a palliative treatment, typically using less than 20 daily fractions. For the same dose fractionation some deemed it radical and others palliative. This was true for doses >--4000 cGy up to 6000 cGy (Table 4), although above 4000 cGy the majority favoured radical treatment for this scenario. It is also interesting to note that three replies were stated to be both palliative and radical; this highlights the difficulties inherent in the loose 'definitions' of these descriptions of treatment 'intent'. Radiation oncologists who felt that they were primarily aiming to relieve symptoms selected a Table 4. Relationship of dose to treatment intent in patient B
Dose (cGy)
Radical (No.)
Palliative (No.)
2000 3000 4000 4500 5000 5500 6000
0 0 3 3 8 1 21
25 22 8 1 4 0 4
Patterns of PalliativeRadiotherapyin Canada lower dose/shorter ffactionation schedule (P = 0.013), whereas those who felt that they were aiming to improve survival selected higher dose/longer fractionation schedules (P < 0.000001). It was observed that the selected dose fractionation was directly related to the availability of megavoltage equipment and to the departmental work-load.
Aims Overall, 85% of radiation oncologists aimed to relieve symptoms and 58% to prevent symptoms. It was noted that those who did not aim to improve survival were more likely to aim to alleviate symptoms. Those who resolved to prevent symptoms were unlikely also to aim to give hope to the patient. The aim only to relieve symptoms was associated with a poorer perceived prognosis and correlated with a smaller percentage cure probability (P < 0.001) and a shorter life expectancy (P = 0.00066).
Prognosis Departmental policy was signficantly correlated with the expected percentage of patients likely to be cured (P = 0.041). Younger radiation oncologists tended to predict a shorter expected survival (P = 0.02).
95
Aims All radiation oncologists aimed to relieve symptoms for this patient with metastatic breast cancer. Those who aimed to prevent symptoms also tried to give hope (P = 0.007) and predicted a better life expectancy (P = 0.009). They were also more likely to feel that there were inadequate numbers of inpatient beds for the management of these patients (P = 0.032). Radiation oncologists who aimed to give hope to the patients were also more likely to recommend additional treatment, the difference being highly significant (P = 0.00007). They were also more likely to feel that current practice was suboptimal (P = 0.034) because of a lack of simulators (P = 0.017) and beds (P = 0.038).
Prognosis It was observed that 91% of radiation oncologists felt that this patient had no likelihood of being cured. In spite of this, there was a wide range of predicted life expectancies.
Additional Care
It was observed that additional treatment was more likely to be offered by radiation oncologists who would not expect to be involved in terminal care (P = 0.048). Furthermore, the possibility of the radiation oncologist being involved with terminal care was directly related to the departmental work-load (P = 0.018), those in larger departments being less likely to be involved.
About 20% of respondents suggested that chemotherapy would be appropriate for additional management. Interestingly, 28% considered that other forms of management should be sought. It was stated that terminal care would be performed by the family doctor in 54% of cases; 38% of radiation oncologists said that they would be predominantly involved with terminal care. It was interesting to note that there was a greater probability that the radiation oncologist would be involved with terminal care if the department was large (P = 0.023) and if a greater annual number than average of new patients was seen (P = 0.036).
Patient C: Breast Cancer Metastatic to Bone
Correlations Between Case Scenarios
Radiotherapy
Radiotherapy
Additional Care
The majority of radiation oncologists felt that the radiotherapy for this patient should be simulated. It was unaminously agreed that treatment was palliative in intent and that a single field should be used. Just over half (57%) prescribed 20 Gy in 5 fractions and occassionally 25 Gy in 5 fractions (over I week). The schedule prescribed ranged from 8 Gy in a single fraction to 30 Gy in 10 fractions. Surprisingly, only 15% used single fractions in spite of trial evidence supporting this policy. The RTOG study of palliation of bone metastases indicated that low dose, short course schedules were as effective as high dose, protracted programmes [2], although a re-analysis by Blitzer challenged this conclusion [3]. Data from Price et al. [4] and Cole [5] support the shorter regimen. The use of a single field (as opposed to more complex field arrangement) as well as the use of a larger number of fractions was directly related to the aim of preventing symptoms (P = 0.045, P = 0.02, respectively).
Radiation oncologists who chose a lower dose and fewer fraction schedules tended to do so for all three cases, but there was a particularly strong correlation when patients A and B were compared (P ~ 0).
Aims There was consistency in the objectives to prevent symptoms or give hope in all three case scenarios (Table 1). By comparison, there was no significant correlation between cases when the stated aim was to improve survival or relieve symptoms.
Prognosis Radiation oncologists who considered that few scenario A patients were likely to be cured were shown to feel similarly for scenario C (P = 0.0055). The anticipated life expectancy for patient C was significantly longer than that for patients A or B, and, in addition, a greater variation of life expectancy was stated for patient C (Table 1).
96
Additional Care Those who advised the use of additional treatment, other than chemotherapy, for patient B also considered that similar treatments should be advised for patient C (P = 0.0005).
Resource Impact Departmental heads were purposely not asked to comment if they considered the treatment routinely prescribed in their department was considered to be optimal or not. Sixty-five (78.4%) of staff radiation oncologists felt that the treatment they prescribed was satisfactory for these patients, but 18 (21.6%) were concerned that their treatment policies were not ideal. In order of decreasing concern they complained of: lack of technologists (16), lack of megavoltage equipment (11), departmental policies (6), shortage of beds (5), and inadequate numbers of simulators (4). It was clear that these perceived deficiencies of resources are inter-related. However, there was no significant correlation between a perception of having suboptimal resources and the actual resources available in those departments. This may be due to the small numbers of respondents in each category.
DISCUSSION Twenty-two percent of Canadian radiation oncologists consulted had concerns that treatment was compromised by lack of resources such as technologists or megavoltage equipment. It was clear that radiation oncology departments with a heavy palliative work-load tended to employ shorter dose fractionation schedules. The lack of linear accelerators and cobalt therapy machines was also related to the selection of lower dose fractionation schedules. The major determinant, however, of palliative treatment should be the lowest dose/shortest fractionation schedule which will provide optimal control of symptoms with minimal acute morbidity. Late morbidity is unlikely to be a problem, but tolerance doses of radiation should never be exceeded. The simplest technique should always be preferred if the same result can be achieved. The RTOG studies, for example, suggested that, for most patients with brain metastases, 2000 cGy in 5 fractions was as effective as 4 weeks of higher dose regimens [6,7]. Analysis of the Canadian radiotherapy practice for locally advanced lung cancer shows that radiation oncologists under the age of 46 select either a very short dose fractionation schedule or a 6-week programme, whereas those over 46 years of age chose relatively evenly from 5 up to 30 fractions. This may be accounted for by changing work-loads with age [8], initial training, subsequent work experience, or becoming more work-load conscious, or by altered perceptions of management with increasing age. The interprovincial differences in treatment for locally advanced lung cancer may be due in part to age,
G. Duncan, W. Duncan and E. J. Maber
departmental policy, or training. Regional differences in the use of single fractions versus multiple fractions have been noted in a UK survey of treatment for bone metastases [8] as likely to be predominantly a result of training. In this Canadian survey, however, no interprovincial difference in fractionation for bone metastases was seen. There is a lack of consensus in the management of non-small cell lung cancer [9,10]; policy may be based on what doctors would choose for themselves [10]. Sixty percent of radiation oncologists in Canada treat locally advanced lung cancer with less than 20 fractions, compared with 60% in the USA choosing a 30 fraction treatment schedule. These differences have considerable economic consequences and deserve further study. Patterns of practice established in the UK and the USA serve as a useful baseline [8,11]. In Canada 9% of respondents felt that the radiotherapy was both radical and palliative. In several instances the same dose fractionation was deemed radical by one oncologist but palliative by another. This highlights the confusion and subjectivity of present 'definitions'. In order to establish cost effective treatment strategies, more precise definitions are essential. 'Radical' is usually employed to refer to a higher dose/longer fractionation treatment and the potential consequences. Definitions may be organ tolerance based or empirical as suggested in Table 5. 'Curative', (not 'radical') therapy aims to improve survival and is a function of a tumour's inherent radiocurability. 'Palliative' should refer solely to the aims of relieving or preventing symptoms (or perhaps to give hope). Thus, curative therapy aims to improve survival, and palliative to improve quality of survival. We believe it would be helpful in addressing critical work,load and resource questions internationally if definitions were agreed. Table 5. Definitions of 'curative' and 'palliative' treatment
Curative
Palliative
Increase survival ( + / - improving symptoms) 'Quantity of Life'
Improve symptoms only 'Quality of Life'
The goal in palliation must remain to return the patient to a comfortable, functional life if possible, in an expeditious manner, whilst considering the economic consequences [12]. Studies on palliation, therefore, must define costs as well as measuring quality of life [13]. The answers regarding optimal management can only be derived from randomized clinical trials. Hopefully more such studies in palliative care will be conducted.
References 1. Maher EJ, Coia LR, Duncan GG, et al. Treatment strategies in advanced and metastatic cancer: Differences in attitude between the USA, Canada and Europe. Int J Radiat Oncol Biol Phys 1992;23:23944.
97
Patterns of Palliative Radiotherapy in Canada 2. Tong D, Gillick L, Hendrickson FR. The palliation of symptomatic osseous metastases: Final results of the study by the Radiation Therapy Oncology Group. Cancer 1982;50:893-9. 3. Blitzer PH. Re-analysisoftheRTOGstudyofthepalliationof symptomatic osseous metastasis. Cancer 1985;55:1468-72. 4. Price P, Hoskin PJ, Easton D, et al. Prospective randomized trial of single and multifraction radiotherapy schedules in the treatment of painful bony metastases. Radiat Oncol 1986;6:247-55. 5. Cole DJ. A randomized trial of a single treatment versus conventional fractionation in the palliative radiotherapy of painful bone metastases. Clin Oncol 1989;1:5%62. 6. Borgelt B, Gelber R, Kramer S, et al. The palliation of brain metastases: Final results of the first two studies by the Radiation Therapy Oncology Group. Int J Radiat Oncol Biol Phys 1980;6:1-9. 7. Kurtz JM, Gelber R, Brady LW, et al. The palliation of brain metastases in a favourable patient population: A randomized
8. 9. 10. 11. 12. 13.
clinical trial by the Radiation Therapy Oncology Group. Int J Radiat Oncol Biol Phys 1981;7:891-5. Crellin AM, Marks A, Maher EJ. Why don't British radiotherapists give single fractions of radiotherapy for bone metastases? Clin Oncol 1989;1:63-6. MacKillop WJ, O'Sullivan B, Ward GK. Non-small cell lung cancer: How oncologists want to be treated. Int J Radiat Oncol Biol Phys 1987;13:929-34. Palmer MJ, O'Sullivan B, Steele R, et al. Controversies in the management of non-small cell lung cancer: The results of an expert surrogate study. Radiat Oncol 1990;19:17-28. Coia LR, Hanks GE, Martz K, et al. Practice patterns of palliative care for the United States, 1984-1985. Int J Radiat Oncol Biol Phys 1988;14:1261-9. Hendrickson FR. Strategyofpalliative treatment. IntJRadiat Oncol Biol Phys 1982;8:155-6. Bowen OR. Shattuck lecture: What is quality care? N Engl J Med 1987;316:1578-9.
Received for publication August 1992 Accepted following revision December 1992