Prognostic impact of EGFR mutation in non-small-cell lung cancer patients with family history of lung cancer

Background A family history can be a valuable tool in the era of precision medicine. Although a few studies have described an association of family history of lung cancer with EGFR activating mutation, their impact on survival of lung cancer patients is unclear. Methods The study included consecutive 829 non-small-cell lung cancer patients who received analysis of EGFR mutation in a prospective lung cancer cohort. Family history of lung cancer was obtained by face-to-face interviews at the time of diagnosis. An association of EGFR activating mutation with a family history of lung cancer in first-degree relatives was evaluated with multivariate logistic regression analysis, and its association with survival was estimated with Cox’s proportional hazards model. Results Seventy five (9.0%) patients had family history of lung cancer. The EGFR mutation was commonly observed in patients with positive family history compared to those with no family history (46.7% v 31.3%, χ2 p = 0.007). The family history was significantly associated with the EGFR mutation (aOR and 95% CI: 2.01 and 1.18–3.60, p = 0.011). Patients with the positive family history survived longer compared to those without (MST, 17.9 v 13.0 months, log-rank p = 0.037). The presence of the EGFR mutation was associated with better survival in patients without the family history (aHR and 95% CI: 0.72 and 0.57–0.90, p = 0.005). However, this prognostic impact was not observed in patients with the positive family history (aHR and 95% CI: 1.01 and 0.50–2.36, p = 0.832). Conclusions In comparison to patients without the family history, EGFR activating mutation was common, and it did not affect prognosis in patients with positive family history.


Introduction
Lung cancer is the most common cause of cancer-related death worldwide [1]. Non-small-cell lung cancer (NSCLC) accounts for 80-85% of lung cancers. Histology-based cytotoxic chemotherapy and molecular targeted therapy are a mainstay in treating NSCLC patients, recently incorporating immunotherapy as well [2]. In spite of these advances, the prognosis for patients remains grave.
In the era of precision medicine, the importance of accurate characterization of clinical phenotype as well as genetic diversity is increasing. In this regard, family histories of lung cancer have been thought to add information in predicting risk for lung cancer and prevalence of mutations on epidermal growth factor receptor (EGFR) gene. Having a family history of lung cancer from first-degree relatives confers increased risk for lung cancer [3][4][5][6]. Activating mutations on EGFR gene are more commonly observed in the patients with the family history compared to those without the history [7][8][9]. The T790M or V843I mutation on EGFR gene was reported to be inherited, so the family history may have an effect on the susceptibility of EGFR tyrosine kinase inhibitors (TKIs) [10,11]. However, the effect of the family history on patient survival has been investigated in a few studies where the association was inconsistent [12,13]. Moreover, the prognostic impact of EGFR mutations by the family history has not been examined.
Therefore, we conducted a study to investigate the association of family history of lung cancer with EGFR activating mutation and the prognostic impact in a prospective lung cancer cohort.

Study population
This study initially considered 902 consecutive NSCLC patients who were histologically diagnosed in Inha University Hospital from January 2006 through January 2014 and received EGFR mutational testing. To maintain quality of information, patients whose family histories were not obtained (n = 12), who failed to test EGFR mutational status due to poor quality of DNA (n = 11), or who were transferred to other hospitals just after diagnosis (n = 50) were all excluded. Finally, 829 patients were included in this study. All information about the clinical characteristics of the patients was obtained from the Inha Lung Cancer Cohort, where clinical information was prospectively collected [14]. This study was approved by the institutional review board of Inha University hospital, and informed consent by patients was waived.

EGFR mutation analysis
Mutational testing was performed by ISU ABXIS Co. Ltd (Seoul, South Korea), an independent commercial laboratory before November 2011, and then by the Department of Pathology at the Inha University Hospital. Mutational analysis was performed to detect mutations in exons 18-21 of the EGFR gene by directional sequencing of polymerase chain reaction fragments amplified with DNA from paraffin-embedded tissue obtained at the time of the diagnosis.

Clinical variables and survival measurement
Information on age, sex, smoking habits, Eastern Cooperative Oncology Group (EGOG) performance status, body mass index, and histology were obtained at the time of the diagnosis. A family history of lung cancer was obtained by face-to-face interviews by a research nurse (HJK). Presence of the family history was defined when any first-degree relatives of the patients had a history of lung cancer. First-degree relatives included a parent, sibling or offspring of the patients. The stages of all patients were estimated according to the 7th edition of TNM classification [15]. Treatments given to the patients were classified into treatment and no treatment. Palliative treatment including radiation therapy to metastatic site or supportive care was classified as no treatment.
Overall survival was measured as an outcome and estimated from the time of diagnosis until death as a result of all causes. In all, 541 patients died. The date of death was obtained by medical records of our hospital or by contacting relatives of the patients. We could not obtain survival of six patients who were lost at follow-up after hospital discharge.

Statistical methods
Distributions of clinical variables between patients by family history were compared using χ 2 tests. The association of family history with EGFR activating mutation was evaluated through multivariate logistic regression testing after adjusting by other clinical variables. The association of family history on overall survival was estimated using the Kaplan-Meier method and log-rank testing. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated using the Cox's proportional hazard model. Null hypotheses of no difference were rejected if p-values were less than .05, or, equivalently, if the 95% CIs of risk point estimates excluded 1. All analyses were performed using IBM's SPSS statistical software package (SPSS Inc. Released 2009. PASW Statistics for Windows, Version 18.0. Chicago: SPSS Inc.). Table 1 displays the clinical characteristics of a total of 829 NSCLC patients by presence of family history of lung cancer in first-degree relatives. Seventy five (9%) patients had a positive family history. The median age of the patients was 66.5 years. There were 535 men (64.5%), 553 ever-smokers (66.9%) and 537 adenocarcinomas (64.8%). The EGFR mutation was found in 271 (37.2%) patients: 132 (48.7%) patients carried the L858R mutation on exon 21; 112 carried the exon 19 deletion (41.3%); and others numbered 27 (10.0%). Of the patients with the EGFR mutation, 162 (59.8%) patients received EGFR-TKIs treatment where 113 (20.2%) patients not having the mutation received the EGFR-TKIs treatment (data not shown). Use of EGFR-TKIs was not different between patients with or without family history of lung cancer (χ 2 p = 0.446).

Association of family history of lung cancer and EGFR activating mutation
Distribution of the L858R mutation and exon 19 deletion did not differ between patients by family history of lung cancer (χ 2 p = 0.448). The EGFR mutation was commonly observed in patients with family history of lung cancer compared to those not having the family history (χ 2 p = 0.007). After adjusting for age, sex, smoking habits, the ECOG performance status, histology, and stages, we found that positive family history was significantly associated with presence of EGFR activing mutation (aOR and 95% CI: 2.01 and 1.18-3.60, p = 0.011) ( Table 2).

Impact of EGFR activating mutation on overall survival of patients by family history of lung cancer
Patients with family history of lung cancer survived longer compared to those without the family history (MST, 17.9 v 13.0 months, log-rank p = 0.037) (Fig 1). However, this impact disappeared when adjusting for age, sex, smoking habits, the ECOG performance status, histology, stages, and treatment (aHR and 95% CI, 0,97 and 0.70-1.34, p = 0.840). As expected, patients with EGFR activating mutation survived significantly longer compared to those having the wild type EGFR gene (MST, 13.3 v 30.9 months, log-rank p < 0.001) (data not shown). When the patients were classified into two groups by family history, EGFR activating mutation was significantly associated with survival in the patients with no family history (aHR and 95% CI: 0.72 and 0.57-0.90, p = 0.005), while this was not observed in patients with positive family history (aHR and 95% CI: 1.01 and 0.50-2.36, p = 0.832) ( Table 3).

Discussion
For the first time, we demonstrate that EGFR mutations are prognostic in NSCLC patients without family history of lung cancer in first-degree relatives but not in patients with a positive family history in this study.  In a few studies on prevalence of EGFR mutations or prognostic impact of family history, information on family history was obtained from reviewing medical records retrospectively, and the prevalence varied from 5.1 to 26.4 percent [7,9,12,13,16] (Table 4).
Family history was prospectively ascertained through face-to-face interviews with patients or their family by a well-trained registered nurse in this study. Family history of lung cancer was asked only in first-degree relatives because recall bias can make family history of seconddegree or more distant relatives inaccurate. This study shows that the prevalence of family history of lung cancer in first-degree relatives is 9%. In addition, it is indicative of the proportion of EGFR activating mutations in the patients with the family history.
Two studies have reported the proportion of EGFR activating mutations by family history [7,9]. This study confirms that family history of lung cancer in first-degree relatives is significantly associated with the presence of EGFR activating mutation. Why patients with family history of lung cancer have a higher prevalence of EGFR activating mutation has been poorly understood. However, accumulating genetic evidence can explain familial aggregation in a variety of cancers, including lung cancer. In NSCLC patients with family history of lung cancer, familial aggregations were noted in genetic variants located in the chromosomal region 15q24-25.1 [17].
As for the effect of family history of lung cancer on survival of the patients, it has been examined in two different cohorts where the prognostic impact was contrasted, and EGFR activating mutations were not considered [12,13]. This study showed an intermediate result in that patients with a positive family history survive longer compared to patients with no family history in univariate analysis. However, the prognostic impact disappeared after adjustments with variables. Nevertheless, it is the strength of this study that the clinical information including family history is carefully scrutinized to make sure this association.
EGFR activating mutations are associated with prognosis in patients with lung cancer [18]. However, its impact is not observed in patients with the family history when patients are divided by the family history in this study. Germ-line mutations of EGFR gene such as T790M or V843I can begin to explain familial aggregation of lung cancer patients; its presence is important in developing resistance to EGFR-TKIs [10,11,[19][20][21][22]. Prevalence of germ-line T790M mutation in patients with EGFR activating mutation varies with sensitivity of detection methods so the range is 0.4-3% using Sanger sequencing but it increases up to 34.2% with more sensitive methods such as next-generation sequencing (NGS) [23]. Although it is not clear whether using high-sensitive techniques to detect low copy number of T790M mutations is related to clinical outcome in NSCLC patients, EGFR activating mutation without T790M mutation is associated with longer survival than those with T790M mutation [23]. Given that EGFR activating mutations were more likely in patients having the germ-line EGFR mutations-related to poor prognosis [10,11,21], this coincidence could provide a plausible explanation on the prognostic impact of EGFR activation mutations by the family history in this study. There are some limitations to this study. First, direct squencing was used to detect mutations on EGFR gene in this study. The patients were diagnosed from January 2006 through January 2014, when direct sequencing was the standard method. It is a less sensitive method in detecting the mutations than NGS analysis so that prevalence of the germ-line mutation is too low to analyze their effect. Therefore, it is a challenge to demonstrate familial aggregation of the germ-line mutation or their association with survival by the interaction of family history and EGFR activating mutations. Nevertheless, the extrapolation from results of past studies may offer a rational explanation [19,23,24]. Second, overall survival was estimated, but progression-free survival was not in this retrospective study. Overall survival is a gold standard for prognostic impact. Progression-free survival is the best option for evaluating the effect of EGFR-TKIs. However, the reliability of the results is always a problem in a retrospective study.
In conclusion, this is the first study to demonstrate that the presence of EGFR activating mutation is associated with better overall survival in NSCLC patients without family history of lung cancer. However, the prognostic impact disappears in patients with a positive family history. Thus, family history needs to be considered when treating with EGFR-TKIs.