Pulmonologists have generally taken a nihilistic view of chemotherapy for the treatment of advanced lung cancer. In a study of beliefs among pulmonologists in the therapeutic treatment of lung cancer, Schroen and colleagues found that only one third of those surveyed believed that chemotherapy conferred a survival benefit for patients with stage IV non-small cell lung cancer and 35% said they would refer patients with metastatic cancer directly to a hospice without referral to medical oncology.¹ A survey of British physicians also found that beliefs about the benefits of chemotherapy did not correspond with current medical knowledge and portrayed a more negative impression than the literature would indicate.² There may be legitimate explanations for this difference in perception about outcomes for this patient population. As a group, respiratory physicians may not believe that the small survival benefit achieved with chemotherapy is enough to offset the toxicity associated with treatment. Perhaps we are biased against chemotherapy because we have seen the worst of the worst cases admitted to intensive care units with neutropenic fever and sepsis after receiving chemotherapy. There is also a perception that the additional expense associated with chemotherapy does not outweigh the benefit, particularly as it relates to quality of life.

The Big Lung Trial published in this issue of Thorax answers the questions about survival benefit, toxicity, quality of life, and cost effectiveness.³ This is the largest and probably the last trial that will compare supportive care and chemotherapy in the treatment of advanced lung cancer. The trial was well designed, well executed, and was powered to answer these important questions. The findings are generalisable to patients with lung cancer throughout the UK with advanced lung cancer. The criteria for enrolment allowed physicians to treat patients with a number of different chemotherapeutic regimens and accepted patients with a poorer performance status, a subgroup for which we have little information because they have been excluded from many previous trials.

There are three major findings in this study. Firstly, it confirmed the median survival benefit of slightly more than 2 months reported in a previous meta-analysis of chemotherapy for advanced lung cancer.⁴ More important to patients, however, may be the comparison of 1 and 2 year survival; 29% of those treated with chemotherapy were alive at 1 year compared with 20% in the supportive care arm, and 2 year survival was doubled in the chemotherapy arm from 5% to 10%. Secondly, while quality of life did not appear to improve substantially in patients treated with chemotherapy, it did not deteriorate—a finding that may come as a surprise to some. The third major finding of this study is that, while the cost of providing chemotherapy was higher than supportive care, the increased expense was offset by the survival gain and is quite reasonable when compared with other healthcare interventions.

Should these findings persuade pulmonologists to refer patients routinely for chemotherapy for advanced or metastatic lung cancer? Clearly the benefit remains small and the toxicity remains real (in this study 4% of patients died from treatment related toxicity). Carney⁵ has suggested that we have reached a plateau in the benefits that will arise from chemotherapy, and the lack of meaningful improvement in cure rates with a myriad of different chemotherapy regimens over the last 30 years would seem to confirm this.

When such a decision has to be made, where the benefit is small and toxicity is present, one possible approach is to assess patients’ preferences. The literature in this area reveals that patients have a wide variation in their preferences for chemotherapy for advanced lung cancer. In one study patients varied from accepting chemotherapy for 1 week of survival benefit to not accepting treatment for survival benefit of more than 2 years.⁶ Patients had logical—if not at times improbable—reasons for their preferences. The patient who chose treatment for 1 week of survival benefit stated that lung cancer could be cured during that week and he would not want to miss the opportunity for cure. Another would not accept treatment even for a 2 year survival benefit, stating that he had lived a good life and did not want anything to interfere with whatever time he had left. Still another would accept treatment for a 1 month survival gain because he had a child getting married in that month and wanted to survive long enough to reach that important milestone for his family.⁶ It is impossible to predict which patients will choose what treatment before these conversations. Perhaps the best approach to this problem is a shared decision making model where the physician provides the patient with full disclosure regarding the risks and benefits of chemotherapy including discussions of survival, quality of life, toxicity, and patient expectations. This study provides the best information to date regarding the data necessary for good decision making. Upon completion of these discussions the decision for treatment should reside with the patient with input from his or her physician. In this context, patients can decide whether the glass is half empty or half full. Either way, the decision will have been the right one for them.