Contemporary management of non-small cell lung cancer in Australia

Author:	D Ball
Details:	Radiation oncologist, Peter MacCallum Cancer Institute

Historical background

In 1955, Bromley and Szur reported on 66 patients with a pathologic diagnosis of lung cancer who had been treated with high dose radiotherapy and who had then proceeded to surgery1. In 24 (46.7%) no tumour was identified in the resected specimen, and in a further 14 (22.5%) tumour was said to be present but degenerate. Eleven patients had oat cell carcinoma and since it is not stated how many of these were disease-free at operation the result cannot be taken to apply exclusively to non-oat cell carcinoma, although at the very least 13 (19.6% of 53) of the complete responders must have had non-small cell lung cancer. Hence we have known for almost half a century that radiotherapy is an extremely active treatment for non-small cell lung cancer, yet a series of randomised trials involving patients with inoperable lung cancer and conducted after Bromley and Szur’s report showed little or no survival advantage associated with the use of radiotherapy in comparison with best supportive care2-4.

How do we explain this paradox? There are three possible explanations.

1. It could be that all patients develop metastatic disease, and the effect of radiotherapy confined to the primary site, no matter how powerful, cannot alter the natural history of the disease, which is determined by secondary deposits outside the thorax. That is certainly true of small cell lung cancer, for which systemic chemotherapy has been shown to be superior to locoregional radiotherapy alone5. But if it were also true for non-small cell lung cancer, then surgery would likewise be ineffective, which clearly it is not, at least for early stage disease6.

2. It could be that any benefit resulting from local disease eradication is counterbalanced by the toxicity of treatment. This certainly appeared to be the case in a recently reported French randomised trial of postoperative radiotherapy for completely resected non-small cell lung cancer, in which the potential advantages of superior local control in the radiotherapy arm were completely offset by treatment related deaths7. These complications appeared to be due to excessive dose, excessive dose per fraction, and bad technique.

3. The wide eligibility criteria and outmoded staging procedures used in these older trials resulted in prognostically heterogenous populations. As a result, the benefits of radiotherapy may have been masked by the inevitably poor outcomes in that proportion of patients who were understaged or had poor performance status. This appears to be the most likely explanation for the inconclusive nature of these early studies.
Unfortunately, an effect of these disappointing trial results was the almost complete extinction of radiotherapy research in lung cancer, at least in the UK, for several decades, during which an attitude of nihilism prevailed and was propagated by a generation of clinical teachers. That this attitude was still prevalent in Victoria in the early 1990s was evident in a survey of clinicians who cared for lung cancer patients during 19938. In this study, 12% of patients did not have a histologic diagnosis, and a large proportion (25%) received no specific anti-tumour therapy.

The 1990s

The last decade has seen dramatic changes in the role of radiotherapy in patients with lung cancer. Previously used almost entirely as a palliative treatment, radiotherapy is now recognised as an important part of the multimodality treatment of both small cell and non-small cell lung cancer, with high level evidence supporting its use in selected patients as a means of prolonging their lives. This quiet revolution has occurred as a result of a number of factors, including better patient selection, better treatment planning and delivery, the use of combined chemotherapy and radiotherapy, and recognition of the importance of overall treatment time. We will review these developments one by one.

Patient selection

Performance status

It is now generally accepted that performance status is the most important prognostic factor in lung cancer9, yet there is no reference to it in one of the most influential older trials2. Inclusion of patients with poor performance status whose survival is likely to be only a few months in randomised trials introduces background noise that can obscure treatment benefits which are evident only in fitter patients. Most contemporary treatment protocols employing a radical approach restricting eligibility to patients whose performance status is ECOG 0 or 1.

Stage

Anatomical extent of disease is also an important prognostic factor, and it is self-evident that patients who have metastatic disease will obtain only limited benefit from a treatment that is confined to the primary site. When patterns of relapse were analysed in an Australian trial for patients thought to have locoregional non-small cell lung cancer, 27% failed initially at distant sites10. It now appears that a proportion of these patients with metastatic disease can be identified by the use of F-18 fluorodeoxyglucose (FDG) PET scanning as a staging procedure, thus sparing them potentially futile and toxic treatment. It could be anticipated that the exclusion of these patients would produce a prognostically more favourable group of patients through the process of stage migration. In our own institution, we have compared the survival of two cohorts of patients with non-small cell lung cancer, one staged before FDG PET became available, and the other in whom FDG PET was used routinely as a staging tool11. Although the treatments used in both groups were similar, the use of FDG PET was associated with a marked increase in median survival from 16 to 31 months (P = 0.01). This suggests that future studies of novel radiotherapy-based treatments for non-small cell lung cancer might most usefully be confined to patients staged with FDG PET.

Tumour volume

The TNM staging system reflects to varying degrees both disease extent and location in relation to surgical resectability. When surgery was the only effective form of therapy, this was understandable, but now that there is strong evidence that non-surgical treatments can also improve survival of patients with non-small cell lung cancer, the usefulness of the current system is under scrutiny. From a radiobiological point of view, the volume of the primary tumour is probably more relevant than its location. Two recent retrospective studies of patients treated with radiotherapy found no influence of stage on survival, but the effect of tumour volume was highly significant12,13. This issue is currently being investigated in an Australasian context in a prospective study of the Trans Tasman Radiation Oncology Group (TROG protocol 9905).

Improved treatment planning and delivery

The widespread application of computer technology to treatment planning and delivery has been one of the most spectacular developments in radiation oncology in the last decade. It is now possible to define the target (for irradiation) using fused CT and FDG PET images, and to establish the anatomical relationship in three dimensions (3D) of the target to critical normal structures with unprecedented precision. The 3D view provided by the planning computer has led to the introduction of highly conformal techniques in which the region of high dose is shaped as closely as possible to conform to the shape of the target, thus minimising dose to surrounding normal tissues (figure one). The dose distribution can be calculated in 3D and this enables objective analysis of dose volume relationships in critical organs, rather than the traditional method in which the radiation oncologist eyeballs a simulator film to determine whether or not a treatment volume is “safe” (figure two). This has allowed dose escalation without complication to levels above 100 Gy14. Whether this will lead to better local control and longer survival remains to be seen. However, it would seem reasonable to expect that improved precision will lead to a lower incidence of serious toxicity, such that the problems observed in the French postoperative study7 do not occur again.

View larger version 0 Bytes:
[in this window]
[in a new window]

Figure 1

View larger version 0 Bytes:
[in this window]
[in a new window]

Figure 2

Combined radiotherapy and chemotherapy

The result of the meta-analysis published in the British Medical Journal15 in 1995, which revealed a 13% reduction in risk of death associated with the use of platinum-based chemotherapy in combination with radical radiotherapy, is well known, but several points are worth reiterating. The first is that in most of the studies, the chemotherapy was given before radiotherapy, so the result does not reflect the potential benefit of concomitant radiotherapy and chemotherapy. Second, the benefit was not confined to patients with stage III disease, but was also observed in stage I and II patients, although because of smaller numbers the effect was not statistically significant.

There are good theoretical reasons why concomitant radiotherapy and platinum-based chemotherapy should be more effective than the two treatments given sequentially. There are at least three randomised trials which have confirmed an advantage for the concomitant versus sequential approach16-18, and the use of both treatments together should now be regarded as the standard of care.

Overall treatment time - CHART

The idea that some cancers repopulate at an accelerated rate during radiotherapy was first proposed by Withers19, and it is now widely accepted that to avoid this potential cause of treatment failure, overall treatment times should be kept short. This may be one reason why concomitant chemoradiation has proven more effective than sequential treatment. In a landmark study, Saunders and colleagues demonstrated that by giving a course of radiotherapy over 12 days instead of 42, local control could be improved, which in turn translated into a statistically significant survival advantage20. The acronym given to the regimen developed by Saunders was CHART (continuous hyperfractionated accelerated radiotherapy), but in order to complete treatment in such a short time, it was necessary to treat patients three times a day, seven days a week. This has proven almost impossible to implement in practice, and as we have demonstrated, similar benefits might be achievable with chemoradiotherapy, without the inconvenience of CHART21. Nevertheless, the results of the CHART trial have confirmed that it is possible to prolong survival with a non-surgical treatment, and that this is a result of improved local control, an observation that proved so elusive to the early investigators.

Radiotherapy for non-small cell lung cancer in Australia in 2002

The remarkable advances that have occurred in the last decade in the radiotherapeutic management of non-small cell lung cancer give cause for more optimism than was the case in 1992, but significant problems remain. We can deliver treatment with much more precision than ever before, but this requires accurate delineation of the target. In our own department, we have shown significant variation in tumour volume delineation by different radiation oncologists22 – we do not know how much of a problem this is in other departments, and whether it is associated with a risk of geographic miss. What account is taken during the planning procedure of tumour movement during the respiratory and cardiac cycles – the so-called 4th dimension?

We now have a much stronger evidence base on which to make recommendations to our patients about the most effective treatment, but will our resources allow us to implement best evidence? At a joint meeting of the Medical Oncology Group of Australia and the Faculty of Radiation Oncology, held in the Barossa Valley in August 2002, the participants were asked how they would manage a hypothetical case of inoperable non-small cell lung cancer. The majority recommended a combination of chemotherapy and radiotherapy, with 30 opting for induction, rather than concomitant chemotherapy, even though the available evidence favours the concomitant approach. Of the 30, only two stated they would recommend that approach by choice, the remainder indicated that it was a strategy forced on them by the long waiting times for radiotherapy. It is to be hoped that as the next decade evolves, the acquisition of new evidence for the best management of lung cancer will be matched by our ability to deliver it.