Article Text
PDF
Abstract
We quantified the proportion of diagnoses of pulmonary fibrosis (PF) among 25 136 people with lung cancer and 250 583 matched controls and compared the natural history of lung cancer in people with and without PF. Diagnoses of PF were more common in people with lung cancer than those without (1.5% vs 0.8%, OR 1.97; 95% CI 1.77 to 2.21). Within people with PF, squamous cell carcinoma was more (22.9% vs 19.1%), and adenocarcinoma was less common (18.0% vs 21.3%). People with PF were less likely to have stage 4 disease at diagnosis (OR 0.43, 95% CI 0.28 to 0.65) but their survival was worse.
- Lung Cancer
- Clinical Epidemiology
- Idiopathic pulmonary fibrosis
- Histology/Cytology
Data availability statement
Data may be obtained from a third party and are not publicly available. Data are available through CPRD and NCRAS.
Statistics from Altmetric.com
Background
The risk of lung cancer in people with idiopathic pulmonary fibrosis (IPF) has been reported to be increased, but data on the impact of a diagnosis of IPF on cancer histology and stage at diagnosis are conflicting.1 Both conditions are associated with a poor prognosis, and when the two conditions occur together the outlook may be particularly bleak.2 3 The primary objective of this research was to quantify the association between diagnoses of lung cancer and PF including IPF, and to determine whether a simultaneous diagnosis of PF was associated with differences in cancer histology, stage at diagnosis and/or survival. To answer these questions, we have used real-world data from the Clinical Practice Research Datalink (CPRD) linked to the National Cancer Registration dataset (NCRD).4
Methods
This population-based matched case–control study used data from our previously described Selection of Eligible people for Lung cancer screening using EleCTronic primary care data (SELECT) study.4 Briefly, this study included incident cases of lung cancer identified in a linked CPRD/NCRD dataset from 2000 to 2015, and up to 10 age-matched and sex-matched controls (online supplement 1). We identified people with a diagnosis of PF using the criteria established by Morgan et al 5 to classify the PF medcodes as ‘narrow’ (IPF) and ‘broad, based on the clinical code (online supplement 2).
Supplemental material
In our initial case–control analysis, we compared the odds of a diagnosis of IPF in cancer cases and matched controls using conditional logistic regression, adjusting for smoking, Charlson Comorbidity Index (CCI) and body mass index (BMI).
Within our cancer cases, we compared patient demographics, stage and histology between people with and without a simultaneous diagnosis of PF using logistic regression, and adjusting for age, sex, smoking, CCI and BMI. We compared survival between people with and without PF by plotting a Kaplan-Meier graph and using Cox regression to calculate HRs before and after adjusting for age, sex, smoking status, CCI, BMI, histology and stage.
In our initial analyses, we used the broad definition of PF, but we also performed a sensitivity analysis using the ‘narrow’ IPF category.
Results
We included 25 136 people with lung cancer and 250 583 controls, of whom 388 (1.5%) and 1984 (0.8%) also had PF (S3– online supplemental table S1). This gave an OR of 1.97 (95% CI 1.77 to 2.21) which increased after adjusting for smoking habit, BMI and CCI (3.10, 95% CI 2.76 to 3.49) (S3– online supplemental table S2).
Within people with cancer, the mean age at diagnosis was 69.8 years for people with PF and 65.8 years for people without PF. People with PF were more likely to be current or ex-smokers (table 1). After adjusting for age, sex, smoking habit, BMI and CCI, people with IPF were more likely to have a diagnosis of squamous cell carcinoma and were less likely to present with stage IV disease (table 1).
Demographic, clinical and histological features of the lung cancer cohort along with ORs for association between PF and those features at the lung cancer diagnosis using logistic regression
The mean follow-up time for our cases of cancer was 1.05±1.75 years, and during this time, 368 people with PF and 23 621 people without PF died. The people with PF had a lower survival compared with those with cancer only (figure 1A)—adjusted HR was 1.85 (95% CI 1.47 to 2.33) (table 2). Increasing age, male sex and small cell histology were associated with a worse prognosis (table 2). We did not observe any improvement in survival during the study period (S3– online supplemental tables S3,S4).
(A) Kaplan-Meier curves showing comparative analysis of follow-up time for the presence or absence of idiopathic pulmonary fibrosis (PF) in lung cancer (LC) using the restricted study population excluding those with broad PF. (B) Comparative analysis of follow-up time for the presence or absence of PF in LC using Kaplan-Meier curve.
Unadjusted and mutually adjusted HRs in patients with a lung cancer diagnosis
In our sensitivity analysis, the pattern of results was similar though there were insufficient numbers to analyse stage at presentation. The impact of a diagnosis of IPF on cancer survival was if anything more marked (figure 1B).
The results of the sensitivity analysis and an additional multiple imputation+propensity score-matched analysis are reported in S3– online supplemental tables S5–S11.
Discussion and conclusion
In our case–control study, we found that 1.54% of people with lung cancer had an additional diagnosis of PF, and this was double the proportion found in general population-matched controls. The patients with cancer with coexisting PF were older, more likely to be male and current or ex-smokers than patients without PF. Squamous cell carcinoma was more common within people with PF. They were also less likely to be diagnosed with advanced disease, but despite this their survival was worse.
The main strengths of our study are the large size of our database, the high validity of the lung cancer diagnoses and death recording, as these are derived from the NCRD and Office for National Statistics data, respectively. The main potential weakness of our study is the validity of PF diagnoses, particularly as we have used the broader definition.5 It is reassuring, however, that our results were similar when we used the narrower more specific codes, although this did lead to a reduction in statistical power. Furthermore, our results may be partly affected by ascertainment bias, as people with lung cancer have more CT scans which enable the PF diagnosis, and, conversely, people with PF routinely undergo interval CT scans, so this could explain why the cancer is detected at an early stage. Finally, the lack of some clinical information, such as performance status, may represent another limitation.
The increased risk of developing lung cancer in PF has been documented before.6 7 However, to our knowledge, the proportion of PF diagnoses in a lung cancer cohort in comparison to that of the general population has not been reported.
Previous studies of lung cancer histology in people with IPF, compared with histology in people with cancer only, suggested an increase of squamous cell carcinoma in people with IPF.1 3 6 We found the same pattern in our study.8
Our finding that a diagnosis of PF worsens the prognosis of lung cancer has been reported before.8 A further challenge is selecting the right treatment for people with both diagnoses.
In conclusion, 1.54% (95% CI 1.39% to 1.71%) of patients with lung cancer have a coexisting PF, doubling general population and this worsens prognosis. As around 50 000 lung cancer cases are reported each year in the UK, we expect that ca. 750 of them will have a concomitant/underlying PF. This should prompt caring clinicians to look for PF in patients with a newly diagnosed lung cancer.
Data availability statement
Data may be obtained from a third party and are not publicly available. Data are available through CPRD and NCRAS.
Ethics statements
Patient consent for publication
Ethics approval
Approval for use of data for this project was granted by the CPRD Independent Scientific Advisory Committee (ISAC) (Protocol numbers 18_223 and 20_014R).
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
FG and JK are joint first authors.
X @DRBLUNGS
Contributors RH, JK and FG took responsibility for the accuracy and the integrity of the manuscript. This study was originally conceived by RH. FG and JK performed the analyses and drafted the manuscript. All authors were involved in data interpretation, critically revised the manuscript and approved the final draft. RH is guarantor for the work.
Funding This research was funded by Cancer Research UK C35238/A26388. This research is linked to the CanTest Collaborative, which is funded by Cancer Research UK (C8640/A23385).
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.