Vinod, Shalini K. MBBS, MD, FRANZCR*; O’Connell, Dianne L. BMaths(Hons1), PhD†; Simonella, Leonardo BA, MPH(Hons)†; Delaney, Geoff P. MBBS, MD, FRANZCR*; Boyer, Michael MBBS, PhD‡; Peters, Matthew MBBS, PhD§; Miller, Danielle BSc, MPH∥; Supramaniam, Rajah BSc, MPH(Hons)†; McCawley, Leslie MPH†; Armstrong, Bruce AM, DPhil, FRACP, FAFPHM, FAA¶
Lung cancer is an important cause of morbidity and mortality in Australia. It is the fourth commonest cancer in Australian male and female patients, and the leading cause of cancer death.1,2 The 5-year relative survival is only 14%.2 Despite this, there are few data on how lung cancer is managed in Australia.
Hypothetical scenarios have demonstrated variability in treatment recommendations for lung cancer.3,4 Surveys of respiratory physicians and thoracic surgeons have found differences in survival estimates, treatment recommendations, and referral patterns.5 Differences have been documented in the treatment recommendations of Australian specialists compared with their international counterparts.6 “What happens to people with this disease depends largely on the personal beliefs of their doctors and the fate of the individual patient is determined by accidents of geography and patterns of referral that lead him or her to one doctor’s office rather than another.”7
Existing data show variability in practice and suboptimal utilization of treatment in Australia. In 1993, a Victorian study of lung cancer revealed a wide range of patient management and lack of adherence to published international guidelines.8 In 1996, pathologic confirmation of diagnoses ranged from 80 to 91% and treatment utilization varied from 55 to 78% across three health areas in New South Wales (NSW). Patients in an outer metropolitan/rural health area were significantly less likely to have treatment.9 The results from these studies are dated and a complete picture of lung cancer management was not available for NSW, the largest state in Australia. Also, it was difficult to determine from these studies the reasons why patients did not have treatment.
We conducted a population-based lung cancer patterns of care study in NSW to assess how current practice compared with that recommended in national guidelines and to identify factors associated with treatment given.
The study population comprised NSW residents diagnosed with primary lung cancer between November 1, 2001 and December 31, 2002. Mesothelioma, carcinoid tumors and sarcomas were excluded. Eligible cases were ascertained from notifications to the NSW Central Cancer Registry (CCR) which meets the data quality standards for inclusion in “Cancer Incidence in Five Continents” compiled by the International Agency for Cancer Research. Rates of histologic verification are high (87%) and notification by death certificate only are low (1.4%). Also, the mortality to incidence ratio of 37% is in good agreement with the 5-year relative survival of 63%.10
Where possible, patients were asked to give consent to obtain information about their lung cancer from their doctors. If the patient had died, specific ethics committee approval allowed treatment data to be obtained directly from doctors. This study was approved by the Ethics Committees of The Cancer Council NSW and Area Health Services where field collection was undertaken.
Doctors were sent questionnaires seeking information on the patient and treatment administered under their care relating to (1) patient demographics, clinical characteristics (weight loss, Eastern Cooperative Oncology Group [ECOG] status, number of primary tumors, site of primary tumor, morphology, tumor, node, metastasis stage for non-small cell lung cancers (NSCLC), small cell lung cancer (SCLC) stage, basis for staging, and comorbidities), investigations, referrals, and initial management; (2) surgery; (3) radiotherapy; and (4) chemotherapy. When required, a trained field officer was provided to extract information from medical records.
SCLC was staged using the Veteran’s Administration staging system.11 Tumor, node, metastasis staging was used for NSCLC and cancers with no pathologic confirmation (NPC).12 Performance status was based on the ECOG scale that rates patients on a five-point scale from being able to carry out normal activities (ECOG 0) to being bed bound (ECOG 4).13 The questionnaire included a description of the scale so that clinicians could assign a score. Where ECOG status was not stated in medical records, the field officer assigned a score based on available information in the records. Each patient’s place of residence was coded according to the rurality of the health area they lived in at diagnosis (Metropolitan, Other Urban, Rural). Doctors’ experience in relation to lung cancer was classified using a combination of medical specialty and lung cancer patient volume during the study period. Comorbidities were documented if the additional condition was likely to impact on the decision to receive cancer-specific treatment.
Data were entered into an Access database and analyzed using SAS statistical software. Cross-tabulations and logistic regression examined factors related to not receiving cancer-specific therapy. Variables statistically significantly related to not receiving treatment in the bivariate analysis together with variables considered to be clinically important were included in the logistic models. Kaplan-Meier survival curves summarized survival up to 2 years after diagnosis.
There were 2931 potentially eligible patients identified by the CCR for whom treatment information could be sought in 2126 (73%). Information was not sought when either the patient or the doctor could not be contacted (n = 198) or refused consent (n = 483), or the CCR notification was received after the end of data collection (n = 124). The study sample comprised 1812 patients (62% of potentially eligible patients) for whom diagnosis and treatment data were available. Reasons for lack of diagnosis and treatment data included doctor refused to provide data, medical record could not be found, or data received after close of the study. There were no significant differences in demographic and disease variables between the study sample and all CCR cases diagnosed during the study period (data not shown).
Median age at diagnosis was 71 years and similar in men and women and 66% of patients were men (Table 1). Less than half of all patients lived in a metropolitan health area (48%).
At initial diagnosis, the majority of patients had NSCLC (71%), 15% had SCLC, and there was NPC of cancer at diagnosis for 13% (Table 2). Of the latter, pathologic confirmation was not attempted by the diagnosing clinician in 44%, attempted but cancer tissue not obtained in 37%, and 16% of patients refused. Clinical evidence alone (clinical examination and imaging) was used to stage 71% of patients, whereas pathologic evidence (surgery and tissue confirmation of nodal or other metastases) was also used for 27%. Weight loss and comorbidities were recorded for 81% and 92% of patients, respectively. Of these 40% had weight loss (>10% in 3 months) at presentation and 73% had at least one comorbid condition likely to affect treatment, the most common being chronic airflow limitation (CAL) (38%), ischemic heart disease (19%), and diabetes (10%). The proportion with comorbidities was similar in male and female patients (75% versus 72%). Male patients had higher rates of CAL (40% versus 34%), but similar rates of ischemic heart disease (19% versus 16%) and diabetes (11% versus 9%). Overall 74% had good performance status (ECOG 0–2).
Respiratory physicians, cardiothoracic surgeons, medical oncologists, and radiation oncologists were considered to be lung cancer specialists. Eighty-nine percent of patients saw a lung cancer specialist at some time. Over half saw a respiratory physician (54%), with most consultations occurring at initial presentation (98%). Of those who saw a respiratory physician at initial presentation 90% were referred, 86% of these were referred to a cardiothoracic surgeon or an oncologist. Approximately two thirds (64%) of patients saw a treating lung cancer specialist at some time during initial treatment: radiation oncologist 47%, medical oncologist 45%, and cardiothoracic surgeon 23%.
The 11% of patients (n = 207) who did not see a lung cancer specialist were more likely to: be older (median age 74 years versus 70 years, p < 0.001); have poorer performance status (35% versus 17% with ECOG 3–4, p < 0.001); live in a rural health area (21% versus 9% and 8% for metropolitan and other urban, respectively; p < 0.001); have a higher proportion of unknown comorbidities (14% versus 7%, p < 0.001); have NPC lung cancer (37% versus 10%, p < 0.001); and have cancer of unknown stage (19% versus 6%, p < 0.001). There were no gender differences.
Pulmonary function tests were done in 47% and imaging in 98% of patients. Imaging included chest computed tomography in 90%, brain computed tomography in 29%, bone scan in 33%, and positron emission tomography (PET) scans in 17% of cases. Fifty percent underwent bronchoscopy and 4% had mediastinoscopy. A pathologic specimen was obtained in 91% of patients and the diagnosis of lung cancer was confirmed in 87%; lung cancer remained the preferred diagnosis for those in whom the specimen did not show cancer.
Cancer-specific therapy, defined as surgery, chemotherapy, or radiotherapy was given to 67% of patients and 33% received none of these after initial diagnosis. The utilization rate for surgery was 17%, radiotherapy 39%, and chemotherapy 30%. Treatment by pathology and stage are shown in Tables 3 through 5.
Approximately three-quarters of patients with NSCLC received treatment and 20% of stage I to 31% of stage IV did not receive treatment (Table 3). Surgery was the most common intervention for stage I disease. Surgery and radiotherapy alone were equally common treatments for patients with stage II disease. Patients with stage III or IV disease were most likely to receive radiotherapy alone.
Most patients with limited stage SCLC had chemotherapy and radiotherapy (43%), whereas those with extensive disease usually had chemotherapy alone (37%) (Table 4). Anticancer treatment was not given to 25% and 31% of limited and extensive stage SCLC patients, respectively. Stage I patients with NPC lung cancer were most likely to have surgery (42%) (Table 5). The remaining NPC patients were most likely to have no treatment, ranging from 65% for stage III to 71% for stage IV.
A total of 598 patients did not receive cancer-specific therapy. Factors statistically significantly and independently associated with no treatment were female gender, aged 70 years or more, weight loss, ECOG score ≥2, advanced or unknown stage, seeing a doctor with a low patient volume (<15 patients seen in the survey equating to one or fewer a month) or not seeing a lung cancer specialist, and residence in “other urban” area health services (neither metropolitan Sydney nor rural) (Table 6). Comorbidities and histologic type (data not shown) were not independently associated with not receiving treatment.
Of the 593 patients who had an initial presentation and diagnosis questionnaire, and received no cancer-specific therapy, approximately 95% were considered inoperable at initial diagnosis. Nevertheless, 71% were referred for further treatment and/or evaluation. After adjusting for significant prognostic factors, variables associated with no referral after initial diagnosis (data not shown) included increasing age (odds ratio (OR) = 1.04 per year of age, p = 0.001); resident in a rural health area relative to a metropolitan health area (OR = 1.69, p = 0.047) and diagnosis with unknown stage relative to locoregional spread (OR = 1.75, p = 0.04).
Of all the patients who did not receive cancer-specific therapy, 138 were potentially curable and may have benefited from curative treatment (Figure 1). These patients had good performance status (ECOG 0–2) and either NSCLC stage I, II, or III disease (n = 120), or limited SCLC (n = 18) at initial diagnosis. Over half of these patients were diagnosed by a lung cancer specialist (54%) and 80% were referred for treatment; but only 37% went on to see a treating clinician. Twenty percent of these patients refused treatment. One third of these refusals were among patients who were not referred. Overall in the study sample of 1812 patients, 5% (n = 99) decided against any treatment.
The median survival was 172 days or approximately 6 months (95% CI: 152–186 days) and the 2-year crude survival was 17%.
This study is the first comprehensive population-based lung cancer patterns of care study in NSW and the most recent study in Australia. The study attempted to capture the patient’s journey from diagnosis to initial treatment, and explore the reasons for the care given.
The demographic characteristics of the patients were representative of a typical Western lung cancer population with a preponderance of men and a median age at diagnosis of 71 years.8,14,15 Approximately three-quarters of the sample had an identifiable comorbid condition likely to impact on treatment, most usually CAL. This is higher than 66% of patients with serious concomitant illnesses reported in a Dutch lung cancer population.16
Pathologic testing was performed in 91% and confirmed the diagnosis in 87% of cases. This is similar to the 1993 Victorian study and a previous NSW study for 1993–1996 where pathologic confirmation was obtained in 88% and 91% respectively.8,17 These are higher than figures derived empirically in guidelines from Scotland and the United Kingdom.18,19 Only 9% did not undergo any pathology testing. This seems reasonable as pathology would not have influenced management for those with comorbidities that contraindicated biopsy, who may have been unfit for treatment or who refused further investigation and management. Similar to other lung cancer populations, approximately two thirds of all patients presented with locally advanced or metastatic disease.16
Lung cancer is managed by a multitude of health care providers as different and multiple treatment modalities may be used for each stage of disease. In this study, 89% consulted a lung cancer specialist. This is comparable to Scotland (87%) and higher than in Yorkshire (75%).15,20 Only 54% of patients consulted a respiratory physician when one would assume that they are the main specialists diagnosing lung cancer. This referral rate is identical to that in Yorkshire but lower than in Scotland where three-quarters of patients saw a respiratory physician. Referrals to surgeons were similar, 23% in NSW versus 19 to 21% in Yorkshire and Scotland, as were referrals to oncologists ranging from 45 to 47% in NSW versus 45 to 53% in Scotland and Yorkshire.
Scottish guidelines state that all patients with suspected lung cancer should be referred to a respiratory physician and UK COIN guidelines indicate that 95% of nonsurgical patients should be discussed in a multidisciplinary team.18,19 Assuming that specialist referral is required as a minimum for multidisciplinary care, given that 11% of patients did not see a lung cancer specialist of any type, neither of these benchmarks was reached in this study.
As described previously in South Western Sydney, Victoria, and South Australia, one third of patients did not receive any anticancer treatment after diagnosis.8,17,21 Since the former studies, readily accessible lung cancer guidelines have been published and there is now good evidence that radiotherapy and chemotherapy can improve survival and quality of life.22,23 Nevertheless, there have been no measurable changes in treatment of lung cancer in NSW. The rate of no treatment is lower than that reported in Scotland (43%) and Ireland (50%) but higher than in the United States where it varies from 15% for SCLC to 20% for NSCLC.14,15,24
Patient factors that were associated with lack of treatment included female gender, increasing age, and poorer performance status. Although poorer performance status may be a valid reason for not treating a patient, older age alone is not. Numerous studies have confirmed the benefits of treating the elderly and studies have identified treatments without excessive toxicity.25–27 Increased utilization of active treatment has resulted in significantly improved survival in the elderly with lung cancer.28 The association of female gender with no treatment has not been described in other studies and needs further exploration.15,17
Stage had a small impact on the no treatment rates for NSCLC and SCLC. This ranged from one in three patients with stage III and IV NSCLC and extensive SCLC to one in five patients with stage I NSCLC. One would expect the rate of no treatment to increase with stage as the intent of treatment changes from curative to palliative. It may be argued that immediate treatment for palliation is not required if the patient is asymptomatic. Nevertheless, this would be true for a minority of patients, as most present with symptoms and the proportion with symptoms increases with disease stage.24
The presence of comorbidity did not independently affect likelihood of treatment. This may have been because most patients had good performance status despite the presence of comorbidities. Similarly in a Dutch study, the presence of comorbid conditions influenced surgical resection rates in operable NSCLC but not treatment for locally advanced or metastatic NSCLC or SCLC.16
Patients with weight loss were less likely to have treatment. Most randomized trials of curative treatment exclude patients with greater than 5 or 10% weight loss as it is taken to be a marker of systemic disease, particularly in the pre-PET era, where patients experience no survival benefit from treatment. Nevertheless, there are conflicting data on whether this is an independent prognostic factor for survival and whether it should be used to make treatment decisions in isolation.29 This is particularly true now that more accurate staging investigations such as PET scans may be used to exclude patients with metastatic disease from curative treatment.
Specialist care has been associated with greater access to curative treatment and significantly better survival.15,30 In this study, health provider characteristics influenced treatment. Patients referred to a lung cancer specialist, especially one with higher caseload, were more likely to receive treatment. This may reflect a selection bias with patients thought not to be fit enough for treatment not being referred to treating specialists. Nevertheless, this assumes that the nonreferring doctor has sufficient experience to make that decision and this may not be valid as surveys of different specialty types have reported a wide variety of opinions on the management of lung cancer.5 Other studies have shown that specialists who saw fewer lung cancer patients per year were less likely to be aware of the benefits of nonsurgical treatment and to offer any treatment.4,5 Specialty type, caseload, year of training, and country of practice have all been found to influence various aspects of care.4,5
The study has some limitations. Treatment data were obtained for only 62% of eligible patients despite considerable efforts to gain doctor and patient consent and obtaining ethics approval to acquire treatment information directly from doctors of deceased cases. Although the study sample was unlikely to be materially biased, as it did not differ greatly from all lung cancer patients diagnosed during the study period on the demographic (age and sex) and disease (stage, morphology, method of diagnosis) characteristics, there were some statistically significant differences between participants and nonparticipants: nonparticipants were more likely to have early stage disease (p < 0.0001) and a histologically confirmed diagnosis (p < 0.0001) (data not shown). Although these differences could have a small effect on overall treatment patterns, we see no reason why they should affect the representativeness of stage-specific treatment patterns.
It might also be argued that the data are now 5 years old and that practice might have changed, especially with the publication of the Australian Lung Cancer Guidelines in 2004 with a corresponding information card for General Practitioners.29 Nevertheless, changes are slow to occur in practice as evidenced by the similar proportion of patients not being treated in 1993, 1996, and 2002.8,31 Moreover, the utilization rates for chemotherapy and radiotherapy showed only modest increases from 21% and 33%, respectively in 1996 to 25% and 37% in 2002, despite new evidence supporting the use of these modalities during that period. Although there is also new evidence supporting the use of adjuvant chemotherapy, it applies only to a relatively small patient group and would not be expected to change utilization rates of chemotherapy as much as would its widespread adoption in the palliative setting. Further data will need to be collected to determine how pervasive these changes have become. The findings of this study are therefore likely to be relevant to current practice.
Another possible limitation is this study’s focus on the individual doctors who patients consulted or were referred to. In a multidisciplinary setting, it is possible that some cases may have been presented, discussed, and treatment not recommended without the patient ever having been referred directly to a specialist. Our approach to data collection may have missed these cases.
A significant proportion of lung cancer patients in NSW did not receive any treatment initially. The majority of patients presented with locally advanced or metastatic disease in which treatment can improve a patient’s quality of life and prolong survival, although it may not be curative. Nevertheless, a proportion of patients with early stage, potentially curable NSCLC, did not receive treatment. In part, this was because of patient refusal of treatment; but health care provider characteristics also influenced the likelihood of not receiving treatment. Patients of specialists seeing few lung cancer patients and non-lung cancer specialists were more likely not to be treated. Lung cancer survival in NSW can only improve if treatment utilization increases. This requires education of patients about all treatment options and ensuring that patients are at least discussed with high volume lung cancer specialists or, preferably, receive experienced multidisciplinary team management.
The Cancer Council NSW funded the study. The NSW Central Cancer Registry identified eligible patients while it was operated under contract to the NSW Department of Health at The Cancer Council NSW. It is now operated by the Cancer Institute NSW. Some of the analysis was funded by the Cancer Institute NSW.
Sincere gratitude goes to members of the Study Advisory Group and the doctors and patients who participated in the study.
1. AIHW, AACR, NCSG. Cancer Incidence Projections, Australia 2002 to 2011. Canberra: Australian Institute of Health and Welfare (AIHW), Australasian Association of Cancer Registries (AACR), National Strategies Cancer Group (NSCG), 2005.
2. Australian Institute of Health and Welfare (AIHW), Australasian Association of Cancer Registries (AACR). Cancer Survival in Australia 2001. Part 1: National Summary Statistics. Canberra: Australian Institute of Health and Welfare, 2001.
3. Mackillop WJ, Ward GK, O’Sullivan B. The use of expert surrogates to evaluate clinical trials in non-small cell lung cancer. Br J Cancer
4. Townsley C, Pond GR, Peloza B, et al. Analysis of treatment practices for elderly cancer patients in Ontario, Canada. J Clin Oncol
5. Schroen AT, Detterbeck FC, Crawford R, Rivera MP, Socinski MA. Beliefs among pulmonologists and thoracic surgeons in the therapeutic approach to non-small cell lung cancer. Chest
6. van Houtte P, Ball DB, Danhier S, Scalliet P. Treatment indications and clinical target volume. In P van Houtte, J Klastersky, P Rocmans (Eds.), Progress and Perspectives in the Treatment of Lung Cancer. Berlin: Springer Verlag, 1999. Pp. 225–239.
7. Raby B, Pater J, Mackillop WJ. Does knowledge guide practice? Another look at the management of non-small cell lung cancer. J Clin Oncol
8. Richardson GE, Thursfield VJ, Giles GG. Reported management of lung cancer in Victoria in 1993: comparison with best practice. Anti-Cancer Council of Victoria Lung Cancer Study Group. Med J Aust
9. Vinod SK, Hui AC, Esmaili N, Hensley MJ, Barton MB. Comparison of patterns of care in lung cancer in three area health services in New South Wales, Australia. Intern Med J
10. Tracy EA, Chen S, Baker D, Bishop J, Jelfs P. Cancer in New South Wales: Incidence and Mortality 2004. Sydney: The Cancer Institute NSW, 2006.
11. Micke P, Faldum A, Metz T, et al. Staging small cell lung cancer: Veterans Administration Lung Study Group versus International Association for the Study of Lung Cancer—what limits limited disease? Lung Cancer
12. Mountain CF. Revisions in the International System for Staging Lung Cancer. Chest
13. Oken MM, Creech RH, Tormey DC, et al. Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol
14. Mahmud SM, Reilly M, Comber H. Patterns of initial management of lung cancer in the Republic of Ireland: a population-based observational study. Lung Cancer
15. Gregor A, Thomson CS, Brewster DH, et al. Management and survival of patients with lung cancer in Scotland diagnosed in 1995: results of a national population based study. Thorax
16. Janssen-Heijnen ML, Schipper RM, Razenberg PP, Crommelin MA, Coebergh JW. Prevalence of co-morbidity in lung cancer patients and its relationship with treatment: a population-based study. Lung Cancer
17. Vinod SK, Delaney GP, Bauman AE, Barton MB. Lung cancer patterns of care in south western Sydney, Australia. Thorax
18. Royal College Of Physicians. Management of Lung Cancer. A National Clinical Guideline. Edinburgh: Scottish Intercollegiate Guidelines Network (SIGN), 2005.
19. Anonymous. The Royal College of Radiologists Clinical Oncology Information Network. Guidelines on the non-surgical management of lung cancer. Clin Oncol (R Coll Radiol)
20. Connolly CK, Jones WG, Thorogood J, Head C, Muers MF. Investigation, treatment and prognosis of bronchial carcinoma in the Yorkshire Region of England 1976–1983. Br J Cancer
21. South Australian Cancer Registry. Epidemiology of Cancer in South Australia.
Adelaide: South Australian Department of Health, 2000.
22. Rowell NP, Williams CJ. Radical radiotherapy for stage I/II non-small cell lung cancer in patients not sufficiently fit for or declining surgery (medically inoperable): a systematic review. Thorax
23. Non-small Cell Lung Cancer Collaborative Group. Chemotherapy in non-small cell lung cancer: a meta-analysis using updated data on individual patients from 52 randomised clinical trials. Non-small Cell Lung Cancer Collaborative Group. BMJ
24. Chute CG, Greenberg ER, Baron J, Korson R, Baker J, Yates J. Presenting conditions of 1539 population-based lung cancer patients by cell type and stage in New Hampshire and Vermont. Cancer
25. Bernet F, Brodbeck R, Guenin MO, et al. Age does not influence early and late tumor-related outcome for bronchogenic carcinoma. Ann Thorac Surg
26. Quon H, Shepherd FA, Payne DG, et al. The influence of age on the delivery, tolerance, and efficacy of thoracic irradiation in the combined modality treatment of limited stage small cell lung cancer. Int J Radiat Oncol Biol Phys
27. Langer CJ, Manola J, Bernardo P, et al. Cisplatin-based therapy for elderly patients with advanced non-small-cell lung cancer: implications of Eastern Cooperative Oncology Group 5592, a randomized trial. J Natl Cancer Inst
28. Fry WA, Phillips JL, Menck HR. Ten-year survey of lung cancer treatment and survival in hospitals in the United States: a national cancer data base report. Cancer
29. The Cancer Council Australia. Clinical Practice Guidelines for the Prevention, Diagnosis and Management of Lung Cancer. Canberra: National Health and Medical Research Council (NHMRC), 2004.
30. Fergusson RJ, Thomson CS, Brewster DH, Brown PH, Milroy R. Lung cancer: the importance of seeing a respiratory physician. Eur Respir J
31. Simonella L, O’Connell DL, Vinod SK, et al. No improvement in lung cancer care: The management of lung cancer in 1996 and 2002 in New South Wales. Intern Med J 2008; In press.