Lung cancer: let’s try for prevention and cure

In this issue of the Journal, Mitchell and colleagues report a follow-up survey from Victoria on the management of lung cancer, 10 years after an initial survey in 1993.1 The issues of therapeutic nihilism,2 variation in management and poor survival remain, but it is encouraging that there are improvements. Moreover, the follow-up was until 2003 so that 5-year results after that time could be reported, and further advances have undoubtedly occurred in the subsequent decade.

In the Mitchell et al study, the crude 5-year survival rate for patients with non-small cell lung cancer (NSCLC) was 10.8%, but the important finding was that for those with NSCLC treated surgically with curative intent, it was 43.2%. Even those with locally advanced disease treated with radiotherapy or chemoradiotherapy had a 5-year survival rate of 19%. This emphasises the importance of offering treatment to patients with early-stage disease and, in this study, many patients’ lung cancers were detected as an incidental finding during thoracic imaging for other reasons. It is problematic that 7.4% of patients with early-stage disease were not treated.

It is encouraging that 80% of patients with NSCLC had positron emission tomography (PET) scans before being treated with curative intent to ensure the absence of occult mediastinal or metastatic disease before proceeding, but this should be universal standard practice.3 The importance of keeping abreast of changing practice was also evident in the less frequent use of surgical staging of the mediastinum over the 10 years, in favour of endobronchial ultrasonography and biopsy.4

The survival of patients with small cell lung cancer, which metastasises earlier in its natural history, depended, as would be expected, on stage, performance status and the use of chemotherapy. Survival rates have changed little.

In patients with metastatic disease, treatment has limited efficacy and, for patients with poor performance status and severe comorbid conditions, palliative care may be the best option given the balance between the side effects of chemotherapy and the modest gain in disease-free survival time. In future, the severity of the side effects of systemic therapy is likely to decrease if targeted therapies (such as the epidermal growth factor receptor inhibitors [EGFRs]) with more tolerable toxicities replace conventional therapies.5 Some of these agents are given as tablets and have a different spectrum of side effects than conventional chemotherapy, including, for example, skin toxicity. Furthermore, patients can be selected for these therapies by whether they express the target or not. It was the EGFR antagonists that showed the importance of identifying the correct target. It was not simply EGFR-positive cancers that responded but rather those with a specific enabling mutation in the region of the EGFR gene.6

The finding that 99% of patients were referred to a specialist shows an awareness of the complexity of lung cancer management. The surprising finding was that patients whose management was discussed at a multidisciplinary meeting (MDM) were more likely to be treated and had significantly longer survival. Although the group of patients being presented at the MDM had better prognostic features and earlier stage cancers, presentation at an MDM was an independent prognostic factor in multivariate and landmark analyses. Although it is often taken as self-evident that MDMs are efficient and the collaborative discussion likely to be useful, the objective evidence of improved outcomes in patients with lung cancer is an important addition to the literature.7 It would be hoped that increased participation in MDMs may also increase the likelihood of participation in clinical trials. It would be reasonable to aim for 10% clinical trial participation for patients with cancers with emerging treatments; this is well in excess of the currently reported 3.9%.

Although no difference in survival was found between urban and rural centres overall in this analysis, other studies have shown that rural and remote patients have worse outcomes and more difficulty in accessing diagnostic scans.8

What guidance does the report by Mitchell and colleagues give for continued improvement in outcomes? First, as the authors mention, a registry is important to prospectively record cases, and it would be most valuable if the data collected included the stage at presentation to correlate with treatments and outcomes. Taking this a step further, however, is the absolute necessity of electronic medical records and a digital database where correlations between patient demographics, cancer subtype and stage, and outcomes could be made across thousands of patients. In future, with increasing numbers of subtypes of lung cancer being identified by their genetic makeup, we will no longer be able to test every treatment strategy in randomised clinical trials, but will need to analyse big datasets to identify the best treatment approaches for each tumour subtype.

The report by Mitchell and colleagues highlights key areas where clinical practice varied considerably, even within one state. Some patients were offered treatment and some were not; some had a pathologically confirmed diagnosis and others only a clinical diagnosis; some had their cancers staged with PET imaging before definitive treatment while others did not. This demonstrates the need for national guidelines that can be easily disseminated and updated. Cancer Council Australia has just produced lung cancer treatment guidelines on a MediaWiki platform and written by an expert group based on systematic reviews of the literature.9 This platform allows these guidelines to be continually updated as new evidence becomes available, and their being linked with education modules. It is hoped that this will help standardise practice. With the evidence favouring more complex multimodal treatment in various stages of disease and with the advent of targeted therapies, access to the most current evidence-based treatment recommendations will be essential.

Smoking tobacco is still considered the major cause of lung cancer. Mitchell and colleagues postulate that it was decreased tobacco use that resulted in the later onset of the disease recorded over the decade.10 Women smoked less than men, but lung cancer mortality in men is now decreasing while it is still increasing to become the major cause of cancer mortality in women, because their peak smoking prevalence occurred later. Women also have a higher rate of non-tobacco related lung cancer, classically adenocarcinomas (near 20% compared with 10% in men). Nonetheless, it is clear that the most effective method of changing the outcomes of lung disease still lies with preventing the disease by tobacco control. The Australian Government has announced increased tobacco taxation annually for the next 4 years which, along with minimising advertising and perhaps decreasing retail density, makes the target of reducing the prevalence of smoking among adults in Australia to 10% by 2018 attainable.

A more controversial suggestion is that of screening high-risk groups of smokers by means of low-radiation-dose computed tomography (CT) scanning. A recent systematic review of lung cancer screening trials has suggested that although a survival advantage can be demonstrated, this must be balanced against the harms.11 To reduce the mortality rate from lung cancer in subsequent decades, prevention by tobacco control is still the key. The risk-to-benefit ratio for CT screening of smokers at high risk still makes this problematic and prevention programs are arguably currently a better use of resources.

Outcomes from lung cancer would be improved if best-practice diagnosis and treatment were standardised with the use of continually updated national guidelines. Continued funding of research to characterise subtypes of tumours that may respond to targeted therapies holds promise for improved outcomes.