Phenotype-genotype correlation in multiple primary lung cancer patients in China

Due to recent advances in high-resolution detection technology, multiple primary lung cancer (MPLC) is becoming an increasingly common diagnosis. However, the genotype-phenotype correlations in MPLC patients have not yet been assessed. In this study, we analyzed the clinical and pathological data for 129 consecutive MPLC patients who received curative surgery at the Tongji University Shanghai Pulmonary Hospital, China. We have screened 129 patients in the present study and found mutations in EGFR, BRAF, ROS1 and KRAS genes, as well as the rearrangement of the EML4-ALK gene in 113 patients. The mean patient age was 59.9 (25–78) years old and 41 patients were males (31.8%). Among the total patients, 123 (95.4%) had two primary lesions, 5 (3.9%) had three primary lesions, and 1 (0.8%) had four primary lesions. In 38.8% of the patients, all lesions were located on only one side of the body. Most of the detected mutations (98 patients) were in the EGFR gene. The patients exhibited significant differences in the EGFR mutation, age at diagnosis, and foci location.

Multiple primary lung cancer (MPLC) refers to the synchronous or metachronous appearance of more than one primary lung cancer in a single patient. MPLC was first reported in 1924 1 . In recent years, the rate of diagnosis worldwide has rapidly increased due to the improvement in radiological diagnostic procedures such as high-resolution multi-slice spiral computed tomography (CT) and positron emission tomography (PET) [2][3][4][5] . However, preoperative diagnosis can be problematic because it is often difficult to differentiate MPLC from lung cancer metastasis based on radiological and histopathological criteria alone.
Identification of oncologic biomarkers by molecular genetic analysis may aid in the accurate diagnosis of MPLC. Potential oncologic biomarkers for analysis include allelic loss of heterozygosity (LOH) markers and specific gene mutations [5][6][7][8][9][10][11][12] . Recent advances in tumor molecular biology have resulted in the identification of several candidate markers such as P53 13 and the EGFR gene 14 that can be used for MPLC diagnosis. In addition, several genes have been confirmed to be driver genes in non-small cell lung cancer (NSCLC), including anaplastic lymphoma receptor tyrosine kinase (ALK), v-akt murine thymoma viral oncogene homolog 1 (AKT1), B-Raf proto-oncogene, serine/threonine kinase (BRAF), v-erb-b2 avian erythroblastic leukemia viral oncogene homolog 2 (ERBB2), Kirsten rat sarcoma viral oncogene homolog (KRAS) and phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit alpha (PIK3CA) [15][16][17][18][19][20] . However, current knowledge of the molecular characteristics of MPLC is insufficient to allow for validation of the accuracy and sensitivity of molecular diagnosis. In this study, we screened for mutations in several oncogenic driver genes in a cohort of Chinese MPLC patients and analyzed the data to correlate our genetic findings with the clinical phenotypes.     Patients and specimen collection. From May 2011 to January 2015, 129 consecutive NSCLC patients with MLPC received surgical resection of their lung lesions at the Shanghai Pulmonary Hospital. Post-operative histological analyses confirmed the presence of primary lung cancer in all patients. Fresh primary tumor tissues containing more than 50% tumor cells were collected from all patients during surgery and were used for subsequent gene mutation analyses. Demographic and clinical data were collected from the patient notes and computer records, including age, gender, the tumor size and location, the TNM stage and the histological type.
Candidate gene mutation analysis. Genomic DNA and the total RNA were immediately extracted from fresh tissues using a commercial QIAamp DNA Tissue Kit and RNeasy Kit (Qiagen, Germany), respectively. Mutations in the EGFR, BRAF, ROS1 and KRAS genes as well as the rearrangement of EML4-ALK were detected with Amoy Diagnostics kits (Xiamen, China) utilizing proprietary real-time PCR technology to detect mutations in the target genes.
Statistical analysis. Statistical analysis was performed using SPSS software (Version 18.0, Chicago, IL). The Chi-square tests and one-way ANOVA were used to detect associations between the clinical characteristics and studied genes. A P value of < 0.05 was regarded as being significant.

Results
Demographic and Clinical characteristics. This cohort study included 41 male and 88 female patients,.
The mean age was 59.9 (25-83) years old (Fig. 1). All the 5 youngest patients (< 40 years) were females. A total of 299 foci were removed during the surgery. All lesions were detected simultaneously in each patient (Fig. 2). In total, 123 patients (95.35%) had two lesions, 5 patients (3.88%) had three lesions and 1 patient (0.78%) had four lesions. In 50 patients (38.8%), all the tumors were located on the same side. Most of the tumors were detected at an early stage, including 26 foci (8.70%) at the Tis stage (Fig. 3). Multiple lesions with an identical histological type were recorded in 88 patients, whereas multiple lesions with different histological types were detected in the other 41 patients.

Phenotype-genotype correlation in MPLC patients. After determining the MPLC patient genotypes,
we analyzed the phenotype-genotype correlations among these individuals. Significant differences were detected between the presence of EGFR mutations and the age at diagnosis and foci location (Table 7). Rare EGFR mutations (G719X, S768I, T790M and L861Q) were detected in 18 female patients. These rare mutations were accompanied by common mutations (L858R, exon 19del and exon 20ins) in multiple lesions. Among the lesions with L858R and a rare mutation, 6.25% were mucinous adenocarcinomas, whereas 62.5% were invasive adenocarcinomas. Further, 31.25% of these lesions were located in left upper lobe. Notably, none of the young patients had these rare mutations.
In spite of these difficulties, it remains critical to distinguish between primary and metastatic lung cancer because the long-term survival of these two types of cancer is significantly different. Metastatic lung cancers, especially in patients with multiple metastases, should not be resected as the prognosis is extremely poor. On the other hand, most MPLC patients have separate lesions that are each individually at an early stage. The 5-year survival rate for synchronous MPLC after curative surgery could be as high as 75.8% 24 . Therefore, the proper diagnsis of MPLC is key to the appropriate treatment of patients.
Molecular genetic diagnosis was introduced into the diagnostic criteria for ACCP in 2013; however, the lack of effective molecular markers has limited its clinical application in practice. Hence, there is a pressing need for comprehensive phenotypic and genotypic analyses of MPLC patients.
Multiple factors can contribute to the development of MPLC, including genetic factors, smoking and exposure to environmental pollutants. Previous studies 3 have shown that MPLC is more common in 50-to 70-year-old male smokers; most tumors in these patients have been reported to be squamous carcinoma-squamous  carcinoma, squamous carcinoma-adenocarcinoma or adenocarcinoma-adenocarcinoma. In contrast with these data, our findings suggests that MPLC is more common in females, although they confirm that 50-to 70-year-old patients represent the highest-risk age group.
In our study, we screened for mutations in the EGFR, BRAF, ROS1 and KRAS genes, as well as the rearrangement of the EML4-ALK gene.
Since the direct sequencing of nucleotides is a time-consuming and labor-intensive process 25,26 , we used a real-time PCR-based method to analyze these mutations. Both previous studies and our experience suggested that a real-time PCR-based method would detect the mutations of interest with the same efficiency as direct sequencing [27][28][29][30] . Only 13.95% of the patients screened were found to be free from any mutations in the genes analyzed in the present study. The frequency of EGFR gene mutations (76.1% in the female patients and 61.0% in the male patients) was higher in our study than that reported in Chinese lung cancer patients with a single focus 31,32 .   Furthermore, only 30 (32.6%) of the MPLC patients had identical gene mutations. These data suggest that more than half of second primary lung cancers result from different mechanisms compared with primary cancers. In addition, a previous study indicated that the risk of second primary lung cancers in patients treated with surgical resection for stage I NSCLC is 1.99 per 100 patient-years 33 . Based on these results, we conclude that each MPLC lesion should be independently diagnosed and treated by surgical resection. Our results suggested that rare mutations in the EGFR gene contributed significantly to the diversity of mutations detected in female patients. Although these data provide evidence of genotype-phenotype correlations in MPLC patients, further study involving more patients will be needed to confirm these conclusions.