A case was classified as SMLA if two or more GGO nodules in peripheral lung were histologically diagnosed as AD with lepidic feature. The criteria of Martini and Melamed were used to confirm that the multiple lesions were not metastatic [9]. To avoid the influence of ovarian estrogen secretion, patients younger than 55 years old were excluded from this study. The patient who received preoperative chemotherapy or oral or inhaled steroid therapy were also excluded. Thirty SMLA cases were selected from the consecutive patients who underwent surgical resection in the Department of Thoracic Surgery of Kumamoto University Hospital from 2012 to 2015. Among these 30 patients, two GGO nodules were clinically diagnosed as lung ADs in 15 cases, three of more nodules were recognized in15. In five cases (16.7%), AAH was detected within the resected specimens. The pathological diagnosis of representative nodules (the most invasive or largest among the resected nodules), in accordance with the new classification system recently proposed by Travis et al. [10] were AD in situ, non-mucinous in 8 cases, minimally invasive AD, non-mucinous in 6 cases, invasive AD, in 16 cases. The control group comprised 79 postmenopausal female patients with single-tumor lung AD who received curative surgical resection during 2013 to 2015. Their subtypes of adenocarcinoma were AD in situ, non-mucinous in 14 cases, minimally invasive AD, non-mucinous in 6 cases and invasive AD, in 59 cases. All of the analyzed patients were Japanese. The study protocol for genetic analysis of the blood samples and resected specimens was approved by the Ethics Committee of Kumamoto University Hospital in May 2011. All patients provided written informed consent for their data and materials to be used for research purpose.

Non-cancerous peripheral lung tissue was harvested in 5mm square sections and frozen immediately in liquid nitrogen. Tissue concentrations of E1 and E2 were measured by LC–MS/MS analysis, in ASKA Pharma Medical (Kawasaki, Japan), as described previously [8, 11]. Briefly, the frozen tissue was homogenized in 1 mL of 0.1M KH2PO4 solution by Ultra-Turrax homogenizer. The steroids were extracted by 4 mL of methyl tert-butyl ether (MTBE) from the suspension and eluted with 1 mL of methanol/distilled water/pyridine mixture (90:10:1, v/v/v). The sample was allowed to react with 50 μL of pentafluoropyridine, 40 μL 1M sodium hydroxide solution, 100 μL acetonitrile and 20 μL ethanol for 20 min at room temperature. The residue was dissolved in 0.5 mL distilled water and allowed to react with 0.14 mL of the reaction reagent mixture [1 mg of 2-methyl-1-hydradino-pyridine in 2.6 mL acetonitrile and 0.2 mL of acetonitrile/trifluoro acetic acid mixture (99:1, v/v)] for 1 h at room temperature. The residue was applied to an Oasis WCX cartridge and was derivatized E1 [E1-3-tetrafluoropyridil-17-(1’-methylpyridinium-2’)-hydrazone and dissolved in 0.1 mL acetonitrile/ distilled water mixture (7:3, v/v) and 20 μL of the solution was subjected to an LC-MS/MS. The derivatized E2 fraction was allowed to react with 50 μL of the reaction reagent mixture (80 mg of 2-methyl-6-nitrobenzoic anhydride, 20 mg of 4-dimethylaminopyridine, and 40 mg of fusaric acid in 1 ml of acetonitrile) and 10 μl of triethylamine for 30 min. at room temperature. After the reaction. The cartridge was eluted in 2.5 mL of ethyl acetate /hexane mixture (9:11, v/v). The residue was dissolved in 0.1 mL acetonitrile/ distilled water mixtuer (4:1, v/v) and 20 μL of the solution was subjected to an LC-MS/MS. An API-5000 triple stage quadrupole mass spectrometer (SCIEX, Framingham, MA, USA) equipped with a positive ESI source and a Shimadzu HPLC system (SCL-10Avp system controller, LC-20AD pump, SIL-HTc column oven, CTO-20A auto-sampler, Shimazu GLC, Tokyo, Japan) was used. A Kinetex C18 column (2.6 μm, 150×3 μmm. Shimazu) was used at 50°C. The lower limit of quantitation of E1 and E2 was 0.05 pg/tube and 0.03 pg/tube, respectively.

Genomic DNA was extracted from whole blood using QiaAmp DNA Blood Mini Kit (QIAGEN, Tokyo, Japan) according to the supplier’s instructions. SNPs were identified by using a 5’ nuclease assay with a TaqMan MGB probe. The reactions were performed in an ABI 7500HT (Applied Biosystems, Foster City, CA, USA). The SNP primers and TaqMan MGB probes were provided by the Assay on-Demand service (Applied Biosystems). The 5’ nuclease assays were performed with 10ng genomic DNA, 1x Taqman Genotyping Master Mix (Applied Biosystems), and 1x primer/probe mix according to the manufacturer’s instructions.

To confirm the effects of the SNP in the CYP19A1, the expression levels of CYP19A1 gene in noncancerous lung were quantified by real-time RT-PCR. The FastPure RNA Kit (Takara Bio, Shiga, Japan) was used to isolate total RNA from the frozen tissue. Levels of CYP19A1 mRNA were measured by real-time RT-PCR. Total RNA (1μg) was reversed transcribed using the SuperScript III First-Strand System (Life Technologies Inc., Rockville, MD, USA). Quantitative RT-PCR reactions were performed on the ABI ViiA^™ 7 instrument using TaqMan^® Universal Master Mix and gene-specific primer mixes (both from ABI): CY19A1 (Hs00903413_m1), The Ct values for each gene were normalized to the housekeeping gene GAPDH (Hs02758991_g1), and the fold change in the transcript level was caluculated using the ΔΔCt method. Expression levels were calculated relative to those from the lung AD cell line H358.

Data are expressed as median (minimum–maximum) value for continuous data and as numbers and percentages for categorical data. All statistical analyses were performed by using the SPSS program for Windows (version 18 statistical software; Texas Instruments, IL, USA). The differences of estrogen concentrations and relative expression of CYP19A1 mRNA were evaluated by using the Mann-Whitney test, and categorical data were compared using the chi-square test.

The characteristics of all patients in the cases and the control groups are shown in Table 1. There were no significant differences between the groups with regard to age, smoking status or a familial history of lung cancer. EGFR mutations were detected in at least one nodule in 17 of the 30 cases of SMLA (56.7%), which was not significantly different to that of Control cases (59.5%). Heterozygous and homozygous carriers of a minor allele (A) of SNP rs3764221 of CYP19A1 were significantly frequent among patients with multiple ADs than control group, but the difference was not significant (P = 0.03).

The levels of E1 and E2 in each cases showed a linear relationship (r = 0.839, p<0.001). The distributions of tissue estrogen concentration in the noncancerous peripheral lung of patients of SMLA cases and the control patients are shown in Fig 1. The median concentration (minimum–maximum) of E1 and E2 in the noncancerous lung of SMLA cases were 45.2 (7.0–216.9) pg/g and 2.2 (0.4–8.6) pg/g, respectively. The concentration of E1 and E2 in the noncancerous lung of control cases were 29.0 (0.4–183.8) pg/g and 1.2 (0.0–9.5) pg/g, respectively. The concentration of E1 and E2 were significantly higher in SMLA cases than that of control cases (P = 0.004 and P = 0.02, respectively). The estrogen concentrations were not significantly associated with age, smoking status, pathological stage, familial history of lung cancer or EGFR mutation status. (S1 Table) The distributions of tissue estrogen concentration of all analyzed lung adenocarcinoma patients according to the allele type of rs3764221 of CYP19A1 gene is shown in Fig 1. The median concentration (minimum–maximum) of E1 and E2 in the noncancerous lung of homozygous carriers of major allele of SNP rs3764221(GG) was 26.3 (0.3–139.5) and 1.2 (0.0–6.7), respectively. That of heterozygous and homozygous carriers of minor allele (AA/AG) were and 40.1 (7.0–216.9) and 1.8 (0.0–9.5), respectively. The minor allele of SNP rs3764221 (A) were significantly associated with higher concentration of E1 and E2 (Fig 2: P = .002 and P = 0.01, respectively).

The distributions of CYP19A1 expression in the noncancerous tissue of patients with lung cancer are shown in Fig 3. The median value (minimum–maximum) of relative quantities of CYP19A1 expression in homozygous carriers of major allele of SNP rs3764221(GG) and heterozygous and homozygous carriers of minor allele (AA/AG) were 1.2(0–5.2) and 1.8 (0–9.1), respectively. The minor allele of SNP rs3764221 (adenine) was significantly associated with higher expression of CYP19A1 mRNA (P = 0.03).