ASPO Abstracts
Examination of Targetable Mutations by Smoking Status
Category: Early Detection & Risk Prediction
Conference Year: 2020
Abstract Body:
Purpose: The purpose of this report was to develop predictive models examining
smoking status, along with demographic and clinical factors, including age, race, gender,
stage and vital status, and their association with several lung cancer specific mutations,
such as EGFR, ALK, ROS1. KRAS and RET.
Methods: The data source for this study was The Cancer Genome Atlas (TCGA)
database, which resides at the National Cancer Institute and provides publicly available
genomic data for many cancers. Data from 522 histologically confirmed lung cancer
cases (adenocarcinoma) were analyzed using multiple logistic regression to examine
associations between genetic alterations (ALK, ROS1, EGFR, KRAS and RET) by
smoking status (current, former, never) and gender, race, age at diagnosis and stage.
Results: Of 522 cases, over half (54%) were female, 86% were white, the mean age was
65 years, and 61% were former smokers, 24% were current smokers and 15% reportedly
had never smoked. Adjusted logistic regression models revealed almost a 5-fold
increased odds of EGFR for nonsmokers versus both current and former smokers
(p<0.001), and almost a 2-fold increased odds for males (p<0.05). White race showed a
2.5 higher odds of ALK mutation, which approached significance (p-0.07), ), and current
and former smokers showed a 4- and 3- fold increased odds of EGFR, respectively
(p<0.05). Current smokers showed a 6.5 higher odds of ROS1 mutation versus
nonsmokers (p=0.004). Similarly, current and former smokers had a 3- and 5-fold
increased odds of KRAS mutation versus nonsmokers (p<0.05). No significant
differences were found for the RET mutation.
Conclusions: This study showed diverse patterns of association between smoking and
lung cancer related mutations. Never smokers had a higher odds of EGFR, which is
consistent with past findings, but current and former smokers showed a strong increased
odds for most of the other mutations. White race showed potential associations with ALK
only and males only showed associations with EGFR. Overall, these results shed new
light on smoking as a possible predictive factor for genetic alterations, which has
implications for treatment and warrants further research. Future research with larger,
more diverse populations is needed to further ref
Keywords: lung cancer genetic alterations smoking