Impact of gender and mutational differences in hormone receptor expressing non-small cell lung cancer

Background The incidence of lung cancer in the US has been decreasing but a bigger decline has been observed in men despite similar declines in tobacco use between men and women. Multiple theories have been proposed, including exposure to exogenous estrogens. Our study seeks to understand the relationship between hormone receptors (HR), gender, and the genomic landscape of non-small lung cancer (NSCLC). Methods 3,256 NSCLC tumor samples submitted for molecular profiling between 2013-2018 were retrospectively identified and assessed for HR expression. Hormone receptor (HR+) was defined as ≥ 1% nuclear staining of estrogen receptor-alpha (ER-a) or progesterone receptor (PR) by immunohistochemistry. DNA sequencing by NGS included cases sequenced by the Illumina MiSeq hot spot 47 gene panel (n=2753) and Illumina NextSeq 592 gene panel (n=503). An adjusted p-value (q-value) <0.05 was determined significant. Results HR+ was identified in 18.3% of NSCLC. HR+ occurred more commonly in women compared to men (19.6% vs 11.4%, p <0.0001, q <0.0001). EGFR mutations occurred more commonly in HR+ NSCLC than HR- NSCLC (20.2% vs. 14.6%, p = 0.002, q=0.007). Overall, men with EGFR mutations were affected by HR status with a higher prevalence in HR+ NSCLC while such differences were not seen in women. However, in women ages ≤45, there was a trend towards greater prevalence HR+ NSCLC (25.25% vs. 11.32%, q= 0.0942) and 10/25 (40.0%) of HR+ cases in young women were found to be EGFR mutated. KRAS mutations and ALK+ IHC expression occurred more in HR+ NSCLC whereas TP53 mutations occurred more in HR- NSCLC. Conclusions Women were more likely to have HR+ NSCLC than men and EGFR and KRAS mutations occurred more commonly in HR+ NSCLC. Additional studies with more strict inclusion criteria for HR+ are warranted to see if there is benefit to targeting HR in these subgroups.


Background
Lung cancer is the most common cause of cancer-related deaths in the United States; cigarette smoking is a major risk factor (1).The general incidence of lung cancer has been decreasing in both men and women largely due to decrease in the incidence of smoking, but there has been a minimal decline in the incidence of lung cancer in women (2,3).Although smoking behaviors are similar between men and women today, historically men had higher prevalence of smoking than women resulting in higher incidence rates of lung cancer.As smoking declined in men, their rates of lung cancer declined precipitously.However, women have not experienced a decline of the same magnitude (4).Some theories postulate that women have a higher sensitivity to adverse biological effects of smoking including a prevalence of tumor protein p53 (TP53)/Kirsten rat sarcoma viral oncogene homolog (KRAS) co-mutations, higher levels of polycyclic aromatic hydrocarbons (PAH)-DNA adducts at any given level of smoking, and higher CYP1A1 expression (which encodes an enzyme used in the metabolism of PAHs) (5, 6).Women may have a higher exposure to passive smoking, which is a known risk factor for lung cancer (7).In addition, adenocarcinoma is more common in women especially never-smokers and the risk of adenocarcinoma decreases more slowly than other histologies (8).Consistent with this observation, women have more EGFR mutations than men (9).Other non-smoking risk factors for lung cancer include passive smoking (10), viral infections such as human papilloma virus (HPV) (11), low body mass index (BMI) (12), diet (significantly lower grain and carbohydrate consumption in patients with epidermal growth factor receptor (EGFR) mutations) (13), socioeconomic status (14), and exposures to arsenic, asbestos and radon (15).
The effect of estrogen on lung cancer pathogenesis is complex and not well understood, and current data linking the effect of estrogen and hormone replacement therapy (HRT) on the incidence of lung cancer is conflicting.Estrogen has two major receptors implicated in carcinogenesis of non-small cell lung cancer (NSCLC): estrogen receptor alpha (ER-a) and estrogen receptor beta (ER-b), both with high affinity for estradiol (16).It has been shown that ER-a, in the presence of estrogen, activates transcription, whereas ER-b inhibits transcription in presence of estrogen (16).Regarding the mechanism of estrogen and the carcinogenesis of lung cancer, some studies show that blocking ER can inhibit proliferation of NSCLC in mice while others have shown estrogen can reduce inflammatory cytokines which reduce the risk of NSCLC.Some have proposed that there may be a protective role with HRT in smokers due to the antiinflammatory properties of estrogen by neutralizing the extra inflammation induced by smoking (14).
The role of progesterone receptors (PR) in the pathogenesis of NSCLC is unclear and there have been mixed results on the prognostic implications.Ishibashi et al. first looked at PR in NSCLC and showed that PR+ NSCLC was inversely associated with tumor node metastasis (TNM) stage and histology with better clinical outcomes in patients with PR+ status (17).A later study further showed that PR expression in tumor-surrounding stromal cells is associated with improved disease-specific survival and positive PR expression in tumor epithelial cells is associated with poor disease-specific survival in females (18).Yet, Raso et al. did not show any correlation between PR and patient clinicopathologic characteristics, which included histology, gender, tobacco history, and staging (19).We sought in our retrospective study to understand the relationship between HR status, gender, and the genomic landscape in NSCLC.

Tumor samples
The study included NSCLC tumor samples submitted to Caris Life Sciences (Phoenix, AZ) for analysis.This study was conducted in accordance with guidelines of the Declaration of Helsinki, Belmont report, and U.S. Common rule.In keeping with 45 CFR 46.101(b) (4), this study was performed utilizing retrospective, deidentified clinical data.Therefore, this study was considered IRB exempt and patient consent was not required.

Immunohistochemistry
Immunohistochemistry (IHC) was performed on formalin-fixed paraffin-embedded (FFPE) sections of glass slides.Slides were stained using automated staining techniques, per the manufacturer's instructions, and were optimized and validated per CLIA/CAO and ISO requirements.HR-positive (HR+) status was defined as ≥ 1+ and ≥ 1% nuclear staining of ER-a (SP1, Ventana) and/or PR (IE2, Ventana) by immunohistochemistry. ALK IHC status was determined using the Ventana ALK CDx Assay (D5F3,Ventana); ALK positivity was defined as 3+ in >1% of cells (20).

Next-generation sequencing
NGS was performed on genomic DNA isolated from FFPE tumor samples using the NextSeq platform (Illumina, Inc., San Diego, CA).Cases were either sequenced by the Illumina MiSeq hot spot 47 gene panel (n=2753) and Illumina NextSeq 592 gene panel (n=503).For tumors tested with MiSeq, specific regions of the genome were amplified using the Illumina TruSeq Amplicon Cancer Hotspot panel (21,22).For NextSeq, a custom-designed SureSelect XT assay was used to enrich 592 whole-gene targets (Agilent Technologies, Santa Clara, CA) (23).All variants were detected with > 99% confidence based on allele frequency and amplicon coverage, with an average sequencing depth of coverage of > 500 and an analytic sensitivity of 5%.Prior to molecular testing, tumor enrichment was achieved by harvesting targeted tissue using manual microdissection techniques.Genetic variants were interpreted by molecular geneticists and categorized as "pathogenic," "presumed pathogenic," "pathogenic variant", "variant of unknown significance," "presumed benign" or "benign" according to the American College of Medical Genetics and Genomics (ACMG) standards."Pathogenic", "presumed pathogenic", and "pathogenic variants" were counted as mutations whereas "benign", "presumed benign", and "variants of unknown significance" were excluded.Pan wild type tumors were defined as tumors that did not contain a "pathogenic," "presumed pathogenic," or "pathogenic variant" mutation.
Tumor mutational burden TMB was measured (592 genes and 1.4 megabases [MB] sequenced per tumor) by counting all non-synonymous missense mutations found per tumor that had not been previously described as germline alterations.TMB analysis was available only for those tumors that were tested with the Illumnia NextSeq 592 gene panel NGS testing.

Statistical analyses
Standard descriptive statistics were used for this retrospective analysis.For dichotomous outcomes, Fisher's exact test was performed.For comparison of TMB, student's t-test was performed.Given the nature of multiple comparisons, p-values with multiple comparisons were further corrected using the Benjamini-Hochberg method and an adjusted p-value (q-value) of <0.05 was considered a significant difference.However, due to the exploratory nature of the investigation, multivariate analysis was not performed, only univariate analysis was performed.Statistical analyses were conducted using R (version 3.5.0)and Prism Graphpad (version 10.0.0).1).

Baseline characteristics
In terms of prevalence of mutations, TP53 was most commonly seen in 51.35% of tumor samples followed by KRAS (26.44%) and EGFR mutations (15.48%).In females, TP53 mutations were seen in 47.88% of patients followed by KRAS mutations in 30.91% of patients and EGFR mutations in 20.71% of patients.In males, TP53 mutations were seen in 54.82% of patients followed by KRAS mutations in 21.98% of patients, and then EGFR mutations in 10.25% of patients.(Table 1).
The overall mean TMB was 11.01 mutations/Mb among the 503 patients with TMB tested; the mean TMB was 10.58 mutations/Mb in females (n= 243) and 11.43 mutations/Mb in males (n=260).(Table 1).
We did not see a significant difference in HR+ and HR-NSCLC in receptor tyrosine-protein kinase erbB-2 (ERBB2) and v-raf murine sarcoma oncogene homolog B1 (BRAF) mutations in our population (Table 5, Figures 2D, E).Percentage of EGFR mutated cases in all HR+ cases versus HR-cases, in all HR+ female cases versus HR-male cases, in all HR+ male cases versus HR-male cases.**q<0.001.

Discussion
In our study, we found that a higher percentage of women have lung cancers that are hormone receptor positive and that among hormone receptor positive NSCLC patients, women had a significantly greater prevalence of ER-a positivity.Preclinical studies have examined ER and EGFR simultaneously and have found that estrogen through its receptor can stimulate lung cancer cell proliferation, resistance to cell death, angiogenesis, and metastasis (24).Epidemiological evidence is lacking.In a study in the Women's Health Initiative (WHI) there was no statistically significant association between HRT and the incidence of NSCLC (25).However, no similar investigation has been conducted with hormone based oral contraceptives.Clearly more observational research is needed to conclusively address this question.
We found that HR+ was associated with increased prevalence of EGFR mutations in NSCLC patients age≥65 and in males overall, which is interesting as patients with EGFR mutant lung cancer typically have a lower median age than the average age of U.S. lung cancer patients and seen more in females (26).Both estrogen signaling and EGFR signaling can promote proliferation by inducing tumor angiogenesis through vascular endothelial growth factor (VEGF) secretion and other growth factors (27).EGFR signaling activation increases the expression and activity of aromatase in NSCLC cells and estrogen can induce epidermal growth factor (EGF) production and activate EGFR signaling (24).Studies have shown a correlation between both ER-a and ER-b expression and the presence of EGFR mutations (19, 28).Further studies are needed to better understand the role of estrogen in older men, but it has been shown that among male patients with advanced NSCLC, those with high serum levels of free b-estradiol had significantly worse survival than those with lower b-estradiol so hormone therapy targeting b-estradiol may have benefit in older men (29).
Meanwhile in women, we did not see significance difference in prevalence of EGFR mutations in HR+ versus HR-cases and in women age ≥ 65 years.However, we saw a noticeably higher percentage of females ≤ 45 years with EGFR mutations with an even higher percentage (10/25, 40.0%) seen in HR+ NSCLC.In addition, females ≤ 45 years trended towards having higher prevalence of HR+ NSCLC compared to males.Young lung cancer patients have a different profile, as many young lung cancer patients are never smokers, have actionable mutations (most common being ALK and EGFR), and have predominantly adenocarcinoma histology (30,31).Comparing between young women and men, the lung cancer incidence in young women has been more rapid than the incidence in young men with much of this driven by increases in adenocarcinoma incidence rates in women (32).Much of the reasoning for this remains unclear but strong family genetics may play a role in lifetime nonsmoking women being more suspectable to lung cancer (33).Plus, research has shown that female sex, age of diagnosis ≤ 60, and those with a family history of cancer had lower DNA repair capacity so further understanding of DNA repair genes beyond BRCA may identify targets driving these increases (34).
Our study reflects these patterns, but also shows that HR+ NSCLC and HR+/EGFR mutated NSCLC are more common in young women.Comparisons between premenopausal and postmenopausal NSCLC women have shown that that adenocarcinoma is more prevalent in premenopausal women (35).In premenopausal women, estrogens are produced by their ovaries through ER-a and thus targeting ER-a may help aid in the treatment of lung cancer in young women (28).
Various hormonal markers and their association with NSCLC clinical outcomes have been previously investigated.High levels of circulating estrogen have been associated with worse survival both in women and men (36).Overexpression of aromatase leads to poor survival in postmenopausal women with NSCLC (37).ER-b overexpression has been shown to be a predictive factor of poor survival in women particularly when co-expressed with aromatase (38,39).As our data shows higher prevalence of HR+ with EGFR mutations in older age NSCLC patients, future studies directed towards response to TKIs based on aromatase levels and specific ER receptor expression is warranted.Also, since the time period of our study, there have been new novel treatments in patients with EGFR Exon 20 insertions and given that 8 of our 9 HR+ EGFR Exon 20 insertion cases were female, it may be worth investigating the role of ER+ specifically with EGFR Exon 20 insertions (40,41).
Our study also showed that TP53 mutations were negatively associated with the presence of hormone receptors.This could be in part because the estrogen receptor positive tumors were more likely to be EGFR mutated and this subtype is less commonly associated with TP53 mutations.Smoking has been associated with TP53 mutations and not with EGFR associated cancers (42).However, EGFR and TP53 co-mutations were seen at a similar prevalence in HR+ and HR-NSCLC, but they were more prevalent in females in our HR+ NSCLC population.TP53 co-mutation with EGFR has conferred worse overall survival to first line EGFR TKI use in real world settings and our gender disparity findings in HR+ NSCLC suggest that further investigation is needed (43).
Multiple driver mechanisms and the impact of co-mutations has become increasingly recognized in NSCLC.We showed a significant prevalence in ALK IHC positive NSCLC in HR+ NSCLC compared to HR-NSCLC; this combination has not been studied much in lung cancer and may be worth further investigation first by evaluating HR+ in ALK fusion NSCLC.Our study also showed a significant increase in KRAS mutations in HR+ NSCLC yet a significant decrease in KRAS TP53 co-mutations in HR+ NSCLC.Almotlak et al. showed in ER-b/KRAS mutant mice models that the combination of an ER-b blocker, fulvestrant, with a pan-HER tyrosine kinase inhibitor dacomitinib had a synergistic antitumor effect in treating ER-b positive lung cancer.Furthermore, they showed that sequential immunotherapy improved treatment response, suggesting that this combination may provide a novel approach for HR+ KRAS mutated NSCLC (44,45).On further analysis incorporating TMB analysis, we saw a trend towards lower TMB in HR+ NSCLC but we saw a trend towards higher TMB in females in HR+ NSCLC and that HR+ KRAS mutant females specifically had a significantly higher TMB in comparison to males.KRAS G12C mutations, which have therapeutic implications, are more seen in women with a younger median age and less of a smoking history (5, 46).As we see that KRAS mutant women in HR + NSCLC have significantly higher TMB but not in HR-NSCLC compared to men, there may be additional benefit incorporating hormone therapy in this subset.Future studies also evaluating KRAS mutant subtype, as never smokers are more likely to have G>A transition mutations, along with PD-L1 scores and STK11/ KEAP1 mutations maybe beneficial in better understanding higher incidence of KRAS mutations in HR+ NSCLC.
With regards to the therapeutic implications of our findings, there have been several studies of anti-estrogen therapy in lung cancer particularly looking at EGFR mutated NSCLC.Garon et al. conducted a phase II study looking at erlotinib with fulvestrant in advanced stage NSCLC and did not find a significant difference in overall response rate (ORR), progression free survival (PFS), or overall survival (OS) (47).Meanwhile another randomized phase II trial investigating EGFR-TKI naïve postmenopausal women with advanced lung cancer combining gefitinib with fulvestrant showed tolerability but did not show PFS benefit (48).However, it should be noted that these two studies did not limit enrollment to patients who were HR+ nor limit enrollment to patients with EGFR mutations.
Our study showed that men had a greater prevalence of PR+ overall.Little has been studied regarding anti-progesterone therapy in NSCLC, however, a recent preclinical study showed that PR contains a polyproline domain (PPD) that inhibits NSCLC cell proliferation and has a synergistic effect when given in combination with EGFR TKIs while another preclinical study demonstrated that progesterone can inhibit lung adenocarcinoma cell growth via membrane progesterone receptor alpha (49,50).Further work targeting progesterone receptors should be considered particularly given that 23% of our HR+ NSCLC cases that were ER-a-/PR+.
The strength of our study was the large cohort of 3,256 NSCLC patients available for testing, compared to most other studies with much smaller sample sizes (17,19,38,51).A limitation was that our markers were not directly comparable to other studies.For IHC of ER-a and PR in our study, we used Sp1 transcription factor (SP1) and Calnexin antibody (IE2) respectively and looked at nuclear staining.Other studies have used mouse monoclonal PAI-1 antibody (1D5), anti-estrogen alpha receptor (6F11), or rabbit polyclonal estrogen alpha receptor (HC-20) antibody clone or have not specified when checking for ER-a positivity and mouse anti-progesterone receptor (MAB429) has also been used to evaluate for PR+ (17,19,38,51).Consequently, there have been large variation in detection rates; for example, a review of studies looking at ER-a positivity in NSCLC showed detection rates ranging from 0-97% (51).Another limitation was that our IHC panel only examined ER-a but not ER-b.Like ER-a, studies looking at ER-b have used different antibody clones (Tau antibody (H-150), antiestrogen receptor beta antibody (14C8), and estrogen receptor beta 1 antibody (PPG5/10)) with varying percentages of detection from 19-98% looking at expression both in the nucleus and cytoplasm (51).Finally, our dataset did not have information on KRAS mutation subtypes, PD-L1, or information on STK11/KEAP1 which may be useful in better understanding the differences in the higher prevalence of KRAS mutations in HR+ NSCLC in both genders and HR+ KRAS mutant females having a significantly higher TMB than males.Thus, gauging absolute percentages of HR+ between studies should be cautioned and future studies should standardize the IHC being used to evaluate HR+ in NSCLC.
Further clinical trials in the future evaluating HR+ NSCLC with specific mutations should have more specific inclusion criteria regarding hormone positivity.For example, a Phase I trial recently investigating a combination treatment with aromatase inhibitor exemestane and a carboplatin-based therapy for postmenopausal women with advanced NSCLC showed a significant correlation between overall response rate with level of positive aromatase IHC expression (52).Also, in both NSCLC and breast cancer, there have been new novel agents since the previous phase II studies were completed.There are new TKIs not only in EGFR but for ALK rearrangements and in KRAS G12C (46,53,54).A new class of selective estrogen receptor degraders has shown promise in ER+/HER-2-breast cancer; a recent phase 3 trial on elacestrant showed significant benefit in patients with ESR1 mutation versus standard of care and another phase 2 on camizestrant demonstrated superior PFS when compared to fulvestrant (55,56).As a greater majority of NSCLC has ER-a expression, these new class of endocrine therapies in breast cancer focusing on ESR1 mutations may hold promise in future studies in HR+ NSCLC (57).Thus, additional clinical trials with more selective inclusion parameters and investigation of new TKIs and estrogen modulator combinations should be investigated in HR+ NSCLC.outside of the submitted work.JJN reports personal fees from AstraZeneca, Naveris, AADi, Bioatla, Mindmed, stock in Epic Sciences, Cansera, Quantgene, Indee P/L, and research funding from Merck and Genentech outside of the submitted work.
The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

TABLE 1 Baseline
Characteristics of study population.

TABLE 2 (
A) Hormone receptor status total and percentage in NSCLC by gender and age.(B) Estrogen receptor/progesterone receptor status total and percentage by gender.

TABLE 3 (
A) Hormone receptor status and EGFR mutations in NSCLC total and by gender (B) Estrogen receptor/progesterone receptor status total and percentage by gender.

TABLE 4
Hormone receptor status and EGFR subtypes by total and by gender.

TABLE 5
Hormone receptor status and KRAS, TP53, ALK IHC+, BRAF, and ERBB2 mutations by total and by gender.

TABLE 6
Prevalence of mutations in patients age ≤45 by gender and by hormone receptor status.

TABLE 7 Tumor
Mutational Burden by gender and in KRAS mutant cases.