August 4, 2016

expert reaction to study looking at air pollution and survival in people with lung cancer

The effect of air pollution on mortality in patients with lung cancer is explored in a paper published in the journal Thorax with the researcher team reporting that exposure to air pollution after lung cancer diagnosis can shorten survival time.

Dr Krishnan Bhaskaran, Senior Lecturer in Statistical Epidemiology & Wellcome Trust/Royal Society Sir Henry Dale Fellow, London School of Hygiene and Tropical Medicine, said:

“It is now quite well established that air pollution is a cause of lung cancer. This study takes an interesting new angle, by looking at the impact of pollution on the prognosis of people who have already been diagnosed with lung cancer. The investigators brought together data on more than 350,000 people with lung cancer in California, and obtained a measure of air pollution exposure by linking their residential address to nearby data from pollution monitors. They then looked at how pollution exposure related to survival time after diagnosis. Their results suggest that higher levels of several commonly measured pollutants are associated with shorter survival.

“The methods the investigators used do not lead to a perfectly individualised measure of pollution exposure because they ignore how much time people spent at their home address, how much time they spent outdoors, and so on. However, the approach taken is common for large studies of pollution and health, where more accurate methods, like giving everyone an individual wearable pollution monitor, would be prohibitively expensive. So it is probably the best that could have been done in this study.

“A possible factor explaining the results is that living in an area of high pollution is often linked to socioeconomic factors, because these areas are often cheaper to live in; put simply, people who have a choice do not tend to choose to live in areas with dirty air. Higher levels of deprivation are also known to be related to poorer cancer survival. The authors have tried to account for this, by taking into consideration measures of education and socioeconomic status, but these measures may not be perfect.

“There is one other potentially serious problem with the methods. Exposure to pollution was averaged over the full period from each person’s diagnosis to their death or the end of the study. But this is problematic: suppose that pollution levels were similar everywhere, but reducing over time. Then the people who happened to live longest would end up with a lower average pollution exposure than the people who died quickly, spuriously making pollution look like a cause of quicker death.

“Despite these limitations, given the strong and established links between air pollution and lung cancer, it is quite plausible that pollution would also have an impact on survival from lung cancer. It will be interesting to see whether other studies in other parts of the world can replicate this finding. If true, the effects observed in this study add yet more weight to the case for concerted international efforts to improve air quality.”

Prof. Michael Peake, Honorary Consultant and Professor of Respiratory Medicine, University of Leicester, said:

“Air pollution is well known to be a factor in the causation of lung cancer, but this study is the first to demonstrate a significant impact of air pollution on survival in patients who have already been diagnosed with lung cancer.

“As the authors recognise, it has a number of limitations but the level of details regarding the lung cancer patient population they studied is impressive. It is always difficult in such studies to ascribe a direct cause and effect relationship between such a diffuse phenomenon as air pollution and a disease such as lung cancer where a there are inevitably a number of potential confounding factors. However, I doubt if a better study of the same issue is likely to emerge, certainly in the foreseeable future.

“I believe the findings are of particular importance because the largest impact on survival was in the group of patients with early stage cancers, many of whom are potentially curable. For this group of patients, the authors showed that in those patients with the highest level of exposure to air pollutants, the proportion who remained alive at 5 years was 30% compared to 50% of those with the lowest level of exposure.

“Overall lung cancer survival remains poor and there are great efforts to promote its earlier diagnosis, such as increasing public awareness of the symptoms of lung cancer and the introduction of screening programmes. This work suggests that high levels of air pollution are likely to significantly reduce the impact of such efforts on the numbers of people who eventually die of lung cancer, even if detected early. It adds significant weight to the urgent need for more strenuous efforts to reduce air pollution.”

Prof. Paul Pharoah, Professor of Cancer Epidemiology, University of Cambridge, said:

“This study highlights the importance of air pollution regulations to populations, but the findings are of limited interest to individual people. The paper demonstrates a clear association between estimated air pollution exposure and reduced survival in people with one type of lung cancer, but it has several limitations including that smoking and other lifestyle factors were not taken into account, and no measurements of individual exposure to air pollution were taken, rather exposure was estimated from pollution monitoring stations at locations near where patients lived.

“This paper reports a complex analysis investigating the effects of regional air pollution on the survival of lung cancer patients after diagnosis. There have been many studies over the years investigating the role of air pollution and the risk of developing cancer. Indeed, the earliest studies of Sir Richard Doll and Sir Austin Bradford Hill that showed a link between smoking and lung cancer in the 1950s had originally been set up to investigate the role of exposure to traffic-related pollution in lung cancer. However, only two previous studies have investigated the impact of air pollution on survival after a cancer diagnosis.

“Research into the health effects of air pollution is difficult mainly because of the difficulty of measuring the “exposure” of individuals to air pollution. This study, as noted in the accompanying editorial, has been carried out to a high standard and set the bar high for this type of work.

“Nevertheless, the measure used for exposure to air pollution is inevitably imprecise. The authors used an estimation of the average monthly concentration of air pollution as the exposure to air pollution in over 350,000 individuals with lung cancer diagnosed in California from 1988-2009. The air pollutant estimation was based on the residential address of the patient and the reported level of pollutants from nearby pollution monitoring stations. Individuals were then grouped into four categories (quartiles) according to their level of exposure. No measurements of exposure of individual patients were taken.

“The authors found a modest association between the level of exposure to three pollutants – nitrogen dioxide, small particulate matter and very small particulate matter – and the risk of dying after a diagnosis of lung cancer. The association was stronger for individuals diagnosed with early stage disease (limited spread at the time of diagnosis), with no association in individuals with advanced disease at the time of diagnosis.

“For example, 50 per cent of individuals with localised lung cancer survived for 68 months or more if they were in the lowest exposure category (lowest one quarter) of exposure to very small particulate matter, compared to 29 months for those in the highest category. Another way of showing this difference is that 51 per cent of those in the lowest very small particulate category survived for at least five years, compared to 31 per cent in the highest exposure category.

“Similar associations were observed for the other two pollutants. The association tended to be stronger for one subtype of lung cancer – adenomcarcinoma – than for the other common subtype, squamous cell carcinoma.

“No association was found between exposure to ozone and survival after diagnosis.

“The observed association is quite clear, but association does not necessarily mean causation. The study has some limitations. In particular there was no information on some important possible confounders, smoking in particular. The authors argue that smoking is unlikely to be a confounder because it is not likely to be associated with exposure to air pollution, but evidence for this statement is lacking.

“This work emphasises the importance of air pollution regulations, but the findings are of limited interest to individuals.”

‘Air pollution affects lung cancer survival’ by Sandrah P Eckel et al. published in Thorax on Thursday 4 August 2016. 

Declared interests

Dr Krishnan Bhaskaran: “Nothing to declare.”

Prof. Michael Peake: “I do not have any conflicts of interest.”

Prof. Paul Pharoah: “I have no conflicts of interest to declare.”