Correlation Analysis Of EGFR Gene Phenotypes,Clinical Features And Serum Levels Of Tumor Markers In NSCLC

Objectives:1.A multi-center retrospective study for the patients who had be diagnosed with non-small-cell lung cancer in Kunming General Hospital of Chengdu Military Region, Tumor Hospital of Yunnan Province and the First People’s Hospital of Yunnan Province, aim to analyze the EGFR genetic phenotypes distribution.2.By EGFR gene mutation detection in the non-small cell lung cancer patients, to analyze the relationship between EGFR genetic phenotypes and the characteristics of clinical, pathological and imaging and to conclude the general rules of EGFR gene mutation.3.By studying the correlation of serum tumor markers before treatment and EGFR gene mutation, purpose is to verify whether tumor biomarkers can work as a predictor of EGFR mutations, particularly to CEA and SCC.4.To analyze the advantages and disadvantages between PCR-direct sequencing and amplification refractory mutation system in EGFR gene detection. To discuss the influences of different biopsy positions and methods on detection result.Methods:Collecting 259 patients who had be diagnosed with non-small-cell lung cancer respectively from Kunming General Hospital of Chengdu Military Region, Tumor Hospital of Yunnan Province and the First People’s Hospital of Yunnan Province in January 2011 to December 2015, To Carry out the multi-center retrospective survey and analysis. Survey content include: ①The distribution of EGFR genetic phenotypes; ②General patient characteristics:age, gender, nationality; ③The clinical and pathological features:PS score, smoking history, drinking history, clinical staging, lesion location, the tumor maximum diameter, lung infection and effusion condition, pathological type;④The level of serum tumor markers of patients before treatment: CEA, SCC, Cyfra21-1, CA125, CA199, CA242, NSE, CA72-4; ⑤EGFR gene detection kit company and detection method; ⑥Biopsy positions and methods; Finally, make statistical analysis with SPSS 22.0 software.Results:1.The result of distribution of EGFR genetic phenotypes:(1) In the whole group, a total patients of EGFR wild type were 138 cases, a total patients of EGFR mutation type were 121 cases (included 4 cases with specific mutation type data lost), EGFR mutation rate 46.7%; (2) A total patients of single exon mutation were 105 cases, accounting for 88.5%of the patients with EGFR gene mutation type.(3) A total patients of multiple exons mutations were 12 cases, accounting for 9.9%of the patients with EGFR gene mutation type.(4) Common EGFR mutation types were exon 19 and exon21 mutations, accounting for 41.3%and 28.9%of the patients with EGFR gene mutation type, respectively, meanwhile the other EGFR gene mutation types were accounting for 29.8%.2.General characteristics:(1) The median age of the whole group is 55 years old, whether mutantion or wild type, that is in accord with normal distribution form of unimodal with the k-s analysis. NSCLC usually occurs in arrange from 50 to 60 years old. The average age of the patients with EGFR gene wild type is 54 years old, meanwhile the average age of the patients with EGFR gene mutation type is 55 years old. There is no correlation between age distribution and EGFR mutations.(2)In the whole group, ratio between male to female (M/F) is 1.4:1, M/F ratio of mutation type is 1.0:1, M/F ratio of wild type is 1.9:1; the EGFR mutation rate of female is 55.6%, while, the EGFR mutation rate of male is 40.4%; women are more likely to appear EGFR gene mutation.(3) There are 247 cases from Han ethnic and 12 cases from minority nationality, including each 3 cases respectively from Bai and Buyi, each 3 cases respectively from Hani, Hui, Lahu, Li, Miao, Yi. In general terms, Han acounts for 95.4%, minorities account for 4.6%; The mutation rates of Han and minority are 47.0%and 41.7%, respectively, the difference is not significant.3. Clinical and pathological features:(1) The ratio of the patients whether has a smoking history or not is 1:1.7. EGFR mutation rate of the patients with smoking history is 38.1%, EGFR mutation rate of the patients without smoking history is 51.9%, smoking history is associated with EGFR mutations (X2=4.58, P=4.58). (2) The ratio of the patients whether drinking history or not is 1:4.3; EGFR mutation rate of the patients with drinking history is 49.5%, EGFR mutation rate of the patients without drinking history is 34.7%, patients without drinking habits is inclined to have EGFR gene mutation, but correlation is not significant (X2=3.51, P=0.06). (3) Lesion location distribution:There are 117 cases occur in left lung (45.2%),138 cases occur in right lung(53.3%) and 4 cases with multiple double-pulmonary cancer (1.5%). And, there are 119 cases occur in upper lobe of lung(46.3%) and 107 cases occur in middle and lower lobe (41.3%).(4) In the EGFR wild type group, there are 111 cases with imaging records and 27 cases with data missing; In EGFR mutation type group, there are 93 cases with imaging records and 28 cases with data missing. According to the distribution curve, the tumor maximum diameter in EGFR wild type and EGFR mutation type are 3.9 ±3.3 cm and 3.0 ±1.6 cm, respectively. Skewness and kurtosis of EGFR mutation type is less than 1, which infers that Dmax in EGFR mutation type is closer to normal distribution. However, skewness of EGFR wild type is 3.40±which infers that Dmax distribution is right-skewed, that means the majority of Dmax in EGFR wild type is larger than the median diameter; Analysis with chi-square test, the results indicate that the smaller Dmax has correlation with EGFR mutation type (X2=85.38, P=0.017). (5) There are 5 pathological types in the whole group, including 224 cases of adenocarcinoma,17 cases of squamous carcinoma and 2 cases of large cell lung cancer,11 cases of mixed carcinoma and 5 cases of other types, accounting for 86.5%,6.6%,0.8%,4.2%and 1.9%, respectively; EGFR gene mutation rate of all pathological types is 46.7%. Statification analysis shows that EGFR gene mutation rate of squamous carcinoma, adenocarcinoma and other pathological types is 5.9%,50.0%and 44.4%, respectively. Adenocarcinoma is inclined to have a higher EGFR mutation rate (X2=13.40, P<0.001). (6) According to Chi-square test, no correlation has found between EGFR mutation with clinical stage, PS score, lung infection and pleural effusion (P>0.05)4.Serum tumor marker:(1) NSCLC patients’serum CEA level was detected before treatment. Serum CEA level is divided into five degrees with 5ng/ml、20ng/ml、 50ng/ml、200ng/ml. The positive rate of EGFR gene detection among five degrees is respectively 31.7%,57.6%,50.0%,68.8%and 100%. Patients with serum CEA level range from 0 to 5ng/ml account for 52.3%of the wild type; Patients with serum CEA level range from 5.01 to 20.00ng/ml account for 41.2%of the mutation type. Chi-square test shows that the correlation between with high CEA expression and EGFR gene mutation is statistically significant (X2=22.18, PO.0001). (2) Serum SCC level of 99 cases is divided into normal group and high group by 1.5ng/ml level, which EGFR mutation rate is 50.6%and 27.8%, respectively. Serum SCC level is lower in the patients with EGFR mutation type than in the patients with EGFR wild type, and the correlation is not significant(X2=3.09, P=0.079). (3) No correlations have been found between EGFR gene phenotypes and other tumor markers, such as Cyfra21-1, CA125, NSE, CA242, CA199 and CA724 (P>0.05)5.EGDR gene detection method and kit company:(1) There are 34 cases using direct sequencing detection method and 225 cases using ARMS detection method. ARMS detection kit come from 4 companies, including 131 cases using kit from ACCB company,63 cases using kit from Top gene company and 29 cases using kit from Permed company and 2 cases using kit from biotecan company. (2)Exclude 2 cases using kit from biotecan, EGFR gene mutation positive rate using kit from ACCB company, Top gene company and Permed company is 47.0%,46.0%and 51.7%, respectively.There is no significant difference by the chi-square (X2=0.274, P=0.872). (3) The positive rates of EGFR gene mutation with direct sequencing detection method and ARMS detection method are 45.5%and 47.1%, there is no significant difference(X2=0.32, P=0.860). (4) There are 225 cases detected by ARMS method, including 119 cases of wild type and 106 cases of mutation type. There are 34 cases detected by direct sequencing detection method, including 19 cases of wild type and 15 cases of mutation type.Direct sequencing method tends to have a higher detection positive rate than ARMS method in rare mutation types and multisite mutation(X2=9.83, P=0.04). However, ARMS method are mainly concentrated in exon 19 and exon 21 mutations (35.1%).6.Sample sampling site and method:(1) In the whole group of 259 cases’ samples, 5 cases derived from blood or pleural effusion are excluded. There are 220 cases’ samples from primary tumor and 34 cases’ sample from metastases, accounting for 86.6% and 13.4%, respectively. EGFR gene mutation rates derived from primary tumors and metastases are 49.1% and 38.2%, respectively. However, there is no significant difference (P>0.05). (2) There are three different biopsy methods, such as 186 cases from surgery,31 cases from puncture and 35 cases from bronch oscopy respectively, the 5 cases from other method (3 cases from pleural effusion and 2 cases from peripheral blood), EGFR gene mutation positive rates of three methods above are 47.8%,41.9% and 51.4%, respectively. No significant difference has be found from different biopsy methods (P> 0.05).7.Logistics analysis of the correlations between some factors with EGFR mutations:(1) Binary regression analysis results indicate:EGFR mutations have significant correlation with higher level of CEA, female, no smoking history, adenocarcinoma and smaller tumor Dmax, which can be seen as predictors of EGFR mutations (P< 0.05). (2) The multivariate Logistic regression analysis shows that less than 3cm tumor diameter [P=0.012,OR(95%CI)=5.50(1.451-20.858)] and higher levels of serum CEA [(P=0.001,OR(95%CI)=6.926(2.217-21.640)] are independently associated with EGFR mutation.Conclusions:1.EGFR gene mutation rate is 46.7% in whole group, including 86.8% from single exons mutations,9.9% from multi-exon mutation and 3.3% from unknown mutation. 19 exon mutation type and 21 exon mutation type are the most common types of EGFR gene mutation.2.Female, adenocarcinoma, non-smoker and smaller tumor Dmax are closely correlation with EGFR mutations. No drinking history is associated with EGFR mutations, but the correlation is not significant.The higher level of serum CEA before treatment can be seen as a predicator of EGFR gene mutation.3.Direct sequencing method has advantage in rare and multi-exons mutation, EGFR detection positive rate from ARMS method mainly concentrate in 19 exon and 21 exon mutation.4.The multivariate Logistic regression analysis shows that less than 3cm tumor diameter and higher than 5ng/ml serum CEA can be seen as predicators of EGFR mutation in non-small-cell lung cancer.