Analysis of metastases in non-small cell lung cancer patients with epidermal growth factor receptor mutation.

BACKGROUND: Most lung cancer patients are diagnosed at an advanced stage with metastases. There was no population-based data on metastases in non-small cell lung cancer (NSCLC) patients with epidermal growth factor receptor (EGFR) mutation. This study focused on the metastases in NSCLC patients with EGFR mutation.
METHODS: In our research, we retrospectively studied 365 NSCLC patients with EGFR mutation (EGFR positive-mutant group) were not resistant to first-generation EGFR TKIs and 316 NSCLC patients with T790M mutation (T790M-mutant group) who were resistant to first-generation EGFR TKIs. In the study, we also investigated sex, smoking status, age at diagnosis, histology, T, N, and M stage, and mutation status. In addition, we analyzed metastatic sites in stage IV patients.
RESULTS: Among the EGFR positive-mutant group, 248 (67.95%) patients were stage IV disease. Among them, 41 patients had brain metastases, 86 patients had bone metastases, 16 patients had liver metastases, 168 patients had intrapulmonary metastases, and 39 patients had metastases in other sites. Among the T790M-mutant group, 277 (87.66%) patients were stage IV disease. Among them, 158 patients had brain metastases, 82 patients had bone metastases, 241 patients had liver metastases, 53 patients had intrapulmonary metastases, and 229 patients had metastases in other sites. We also found that lung cancer patients in the T790M-mutant group had higher incidences of the brain (P<0.001), bone (P<0.001), liver (P=0.001), and intrapulmonary metastases (P<0.001). Moreover, wherever the metastatic site was, the metastasis time all centrally distributed in the first two months after diagnosis.
CONCLUSIONS: For patients with metastatic lung cancer, most metastases happened before diagnosis, which indicated that metastases related to driving mutations, such as EGFR positive mutation or T790M mutation, but not to the survival time. Lung cancer patients with T790M mutation were more likely to metastasize before the diagnosis. 2021 Annals of Translational Medicine. All rights reserved.

Introduction

The mortality rate of lung cancer increases year by year (1,2). The annual diagnosis rate of new cases is approximately 1.6 million all over the world (3,4). About 85% of lung cancer patients are diagnosed as non-small cell lung cancer (NSCLC) (5). A total of 47.3% of NSCLC patients presented with metastases at the time of diagnosis (6). Distant metastasis is the leading cause of most cancer deaths. NSCLC metastasizes to the brain (47%), bone (36%), liver (22%), adrenal glands (15%), thoracic cavity (11%), distant lymph nodes (10%), and other organs (less than 5%), which leads to shorter survival (7,8). Metastasis seems to be a random process, which has not well qualified. Some researchers put forward an opinion that the oncogenic drivers, such as epidermal growth factor receptor (EGFR), anaplastic lymphoma kinase (ALK), and ROS1 proto-oncogene receptor tyrosine kinase, may induce the metastases (9-12). EGFR tyrosine kinase inhibitors (TKIs) are effective in treating NSCLC with EGFR mutation. However, some patients may have the mutation that substitutes methionine for threonine at amino acid position 790 (T790M) after being treated with first-generation TKI (13). T790M mutation inhibits first-generation TKI to its binding site, and the resistance to first-generation TKI arises. Some reviews have illustrated the relationship between T790M and the development of resistance to first-generation TKI (14,15). Whether patients with T790M mutation are more likely to have metastases remains unknown. In this study, we analyzed the time distribution of lung cancer distant metastases and the correlation between EGFR mutation, T790M mutation and lung cancer metastases.

We present the following article in accordance with the MDAR checklist (available at http://dx.doi.org/10.21037/atm-20-2925).

Methods

We collected the data retrospectively from the clinical records of patients with lung cancer and metastases diagnosed at the Oncology Department of Shanghai Pulmonary Hospital. After admission, performed systemic bone image, brain MRI, abdominal MRI, or color Doppler ultrasonography, and chest computed tomography (CT) on patients diagnosed with lung cancer every six to eight weeks in case of metastasis. We enrolled a total of 681 lung cancer patients with EGFR mutant who had provided their written consent in the study. Among them, 316 patients from Feb 2001 to Dec 2016 were enrolled. After treating with first-generation EGFR TKIs, such as erlotinib and gefitinib, we observed T790M mutation in these patients. We also enrolled another 365 patients diagnosed from June 2018 to May 2019 who were all treated with the first-generation EGFR TKIs. However, unlike the last 316 patients, these patients did not get resistance to the first-generation EGFR TKIs when we analyzed the data. Their metastatic sites included the brain, liver, bone, contralateral lung, pleural, pleural effusion, adrenal gland, pericardium, abdominal cavity, subcutaneous tissue, and lymph nodes in the cervical, retroperitoneal, and inguinal regions. For analysis, besides the brain, the liver, and the bone, we divided the rest of the sites into two parts: intrapulmonary metastases (contralateral lung, pleural, pleural effusion) and other sites metastases (adrenal gland, pericardium, abdominal cavity, subcutaneous tissue, and lymph nodes). Some patients suffered from two or more metastases and were respectively analyzed in each metastatic site. We collected information from the diagnostic imaging, pathology reports, physician notes, and treatment information to identify the confirmed diagnosis date and the metastasis date. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The research was approved by Ethics Committee of Shanghai Pulmonary Hospital. The ethics reference number was NO K18-203.

Statistical analysis

We tableted data according to the information and used a histogram to visualize the distribution of the time between the confirmed diagnosis date and the metastasis date (metastasis time). Statistical analysis performed using SPSS 22.0 (SPSS Inc., Chicago, IL, USA). Enumeration data expressed as number were analyzed by Chi-square test. Statistical significance was considered as P value less than 0.05.

Results

In the T790M-mutant group, most patients were adenocarcinoma (88.6%), and 119 (37.7%) patients had brain metastases, 195 (61.7%) patients had bone metastases, 36 (11.4%) patients had liver metastases, 224 (70.9%) patients had intrapulmonary metastases, and 48 (15.2%) patients had other metastatic sites.

In the EGFR positive-mutant group, 54.2% of them were adenocarcinoma, and 41 (11.2%) patients had brain metastases, 86 (23.6%) patients had bone metastases, 16 (4.4%) patients had liver metastases, 169 (46.3%) patients had intrapulmonary metastases, and 39 (10.7%) patients had other sites metastases. The pathological results of 23.8% patients only suggested NSCLC, since some patients were just tested by cytology rather than immunohistochemistry. Patient characteristics were summarized in and . (The sum of patients with different metastatic sites exceeds the total number of patients because patients with different metastases were counted for each site independently.)

Table 1

Patient Characteristics of T790M mutant group

Patient characteristics	Total	Metastases				Brain metastases				Bone metastases				Liver metastases				Intrapulmonary metastases				Other sites metastases
		Yes, n (%)	No, n (%)	P		Yes, n (%)	No, n (%)	P		Yes, n (%)	No, n (%)	P		Yes, n (%)	No, n (%)	P		Yes, n (%)	No, n (%)	P		Yes, n (%)	No, n (%)	P
Total	316	277 (87.7)	39 (12.3)			119 (37.7)	197 (62.3)			121 (38.3)	121 (38.3)			36 (11.4)	280 (88.6)			224 (70.9)	92 (29.1)			48 (15.2)	268 (84.8)
Age				0.002				0.004				0.022				0.011				0.435				0.013
<70	246	223 (90.7)	23 (9.3)			103 (41.9)	143 (58.1)			86 (35.0)	86 (35.0)			34 (13.8)	212 (86.2)			177 (72.0)	69 (28.0)			44 (17.9)	202 (82.1)
≥70	70	54 (77.1)	16 (22.9)			16 (22.9)	54 (77.1)			35 (50.0)	35 (50.0)			2 (2.9)	68 (97.1)			47 (67.1)	23 (32.9)			4 (5.7)	66 (94.3)
Sex				0.128				0.583				0.17				0.077				0.141				0.726
Male	131	116 (88.5)	15 (11.5)			47 (35.9)	84 (64.1)			56 (42.7)	56 (42.7)			10 (7.6)	121 (92.4)			87 (66.4)	44 (33.6)			21 (16.0)	110 (84.0)
Female	185	161 (87.0)	24 (13.0)			72 (38.9)	113 (61.1)			65 (35.1)	65 (35.1)			26 (14.1)	159 (85.9)			137 (74.1)	48 (25.9)			27 (14.6)	158 (85.4)
Smoking				0.000				0.457				0.002				0.251				0.082				0.136
Smoked	57	42 (73.7)	15 (26.3)			19 (33.3)	38 (66.7)			32 (56.1)	32 (56.1)			4 (7.0)	53 (93.0)			35 (61.4)	22 (38.6)			5 (8.8)	52 (91.2)
Non-smoked	259	235 (90.7)	24 (9.3)			100 (38.6)	159 (61.4)			89 (34.4)	89 (34.4)			32 (12.4)	227 (87.6)			189 (73.0)	70 (27.0)			43 (16.6)	216 (83.4)
Pathology				0.611				0.032				0.696				0.327				0.315				0.348
Squamous	7	6 (85.7)	1 (14.3)				7 (100.0)			3 (42.9)	3 (42.9)				7 (100.0)			4 (57.1)	3 (42.9)			2 (28.6)	5 (71.4)
Non-squamous	309	271 (91.2)	26 (8.8)			119 (40.1)	178 (59.9)			106 (35.7)	106 (35.7)			36 (12.1)	261 (87.9)			220 (74.1)	77 (25.9)			46 (15.5)	251 (84.5)
T				0.988				0.344				0.305				0.315				0.294				0.847
1	20	19 (95.0)	1 (5.0)			6 (30.0)	14 (70.0)			9 (45.0)	9 (45.0)			1 (5.0)	19 (95.0)			13 (65.0)	7 (35.0)			3 (15.0)	17 (85.0)
2−4	270	249 (92.2)	21 (7.8)			110 (40.7)	160 (59.3)			91 (33.7)	91 (33.7)			34 (12.6)	236 (87.4)			204 (75.6)	66 (24.4)			45 (16.7)	225 (83.3)
N				0.007				0.009				0.000				0.764				0.243				0.14
0	29	23 (79.3)	6 (20.7)			5 (17.2)	24 (82.8)			19 (65.5)	19 (65.5)			3 (10.3)	26 (89.7)			19 (65.5)	10 (34.5)			2 (6.9)	27 (93.1)
1−3	261	244 (93.5)	17 (6.5)			110 (42.1)	151 (57.9)			82 (31.4)	82 (31.4)			32 (12.3)	229 (87.7)			197 (75.5)	64 (24.5)			46 (17.6)	215 (82.4)
Stage				0.000				0.158				0.018				0.52				0.003				0.447
I−II	3		3 (100.0)				3 (100.0)			3 (100.0)	3 (100.0)				3 (100.0)				3 (100.0)				3 (100.0)
III−V	297	277 (93.3)	20 (6.7)			119 (40.1)	178 (59.9)			102 (34.3)	102 (34.3)			36 (12.1)	261 (87.9)			224 (75.4)	73 (24.6)			48 (16.2)	249 (83.8)
EGFR 19 DEL				0.036				0.927				0.565				0.108				0.489				0.027
Negative	145	121 (83.4)	24 (16.6)			55 (37.9)	90 (62.1)			58 (40.0)	58 (40.0)			12 (8.3)	133 (91.7)			100 (69.0)	45 (31.0)			15 (10.3)	130 (89.7)
Positive	171	156 (91.2)	15 (8.8)			64 (37.4)	107 (62.6)			63 (36.8)	63 (36.8)			24 (14.0)	147 (86.0)			124 (72.5)	47 (27.5)			33 (19.3)	138 (80.7)
EGFR L858R				0.224				0.273				0.215				0.471				0.928				0.055
Negative	198	177 (89.4)	21 (10.6)			70 (35.4)	128 (64.6)			117 (59.1)	81 (40.9)			25 (12.6)	173 (87.4)			140 (70.7)	58 (29.3)			36 (18.2)	162 (81.8)
Positive	118	100 (84.7)	18 (15.3)			49 (41.5)	69 (58.5)			78 (66.1)	40 (33.9)			11 (9.3)	107 (90.7)			84 (71.2)	34 (28.8)			12 (10.2)	106 (89.8)
EGFR T790M
Negative	0
Positive	316	277 (87.7)	39 (12.3)			119 (37.7)	197 (62.3)			121 (38.3)	121 (38.3)			36 (11.4)	280 (88.6)			224 (70.9)	92 (29.1)			48 (15.2)	268 (84.8)

EGFR, epidermal growth factor receptor; T790M, a mutation that substitutes methionine for threonine at amino acid position 790; 19 DEL, exon 19 deletion; L858R, substitutions of leucine for arginine in exon 21.

Table 2

Patient Characteristics of EGFR positive mutant group

Patient characteristics	Total	Metastases				Brain metastases				Bone metastases				Liver metastases				Intrapulmonary metastases				Other sites metastases
		Yes, n (%)	No, n (%)	P		Yes, n (%)	No, n (%)	P		Yes, n (%)	No, n (%)	P		Yes, n (%)	No, n (%)	P		Yes, n (%)	No, n (%)	P		Yes, n (%)	No, n (%)	P
Total	365	248 (69.5)	109 (30.5)			41 (11.2)	324 (88.8)			86 (23.6)	279 (76.4)			16 (4.4)	349 (95.6)			169 (46.3)	196 (53.7)			39 (10.7)	326 (89.3)
Age				0.l56				0.104				0.005				0.354				0.725				0.86
<70	229	165 (72.1)	64 (27.9)			31 (13.5)	198 (86.5)			66 (28.8)	163 (71.2)			12 (5.2)	217 (94.8)			110 (48.0)	119 (52.0)			26 (11.4)	203 (88.6)
≥70	128	83 (64.8)	45 (35.2)			10 (7.8)	118 (92.2)			20 (15.6)	108 (84.4)			4 (3.1)	124 (96.9)			59 (46.1)	69 (53.9)			13 (10.2)	115 (89.8)
Sex				0.019				0.436				0.017				0.05				0.066				0.14
Male	237	155 (65.4)	82 (34.6)			25 (10.5)	212 (89.5)			48 (20.3)	189 (79.7)			7 (3.0)	230 (97)			104 (43.9)	133 (56.1)			30 (12.7)	207 (87.3)
Female	120	93 (77.5)	27 (22.5)			16 (13.3)	104 (86.7)			38 (31.7)	82 (68.3)			9 (7.5)	111 (92.5)			65 (54.2)	55 (45.8)			9 (7.5)	111 (92.5)
Smoking				0.098				0.481				0.008				0.454				0.478				0.259
Smoked	166	108 (65.1)	58 (34.9)			17 (10.2)	149 (89.8)			29 (17.5)	137 (82.5)			6 (3.6)	160 (96.4)			75 (45.2)	91 (54.8)			21 (12.7)	145 (87.3)
Non-smoked	190	139 (73.2)	51 (26.8)			24 (12.6)	166 (87.4)			56 (29.5)	134 (70.5)			10 (5.3)	180 (94.7)			93 (48.9)	97 (51.1)			17 (8.9)	173 (91.1)
Pathology				0.000				0.128				0.002				0.033				0.292				0.335
Squamous	78	41 (53.2)	36 (46.8)			5 (6.4)	73 (93.6)			8 (10.3)	70 (89.7)				78 (100.0)			32 (41.0)	46 (59.0)			6 (7.7)	72 (92.3)
Non- squamous	287	207 (73.9)	73 (26.1)			36 (12.5)	251 (87.5)			78 (27.2)	209 (72.8)			16 (5.6)	271 (94.4)			137 (47.7)	150 (52.3)			33 (11.5)	254 (88.5)
T				0.000				0.103				0.462				0.959				0.000				0.124
1	46	21 (45.7)	25 (54.3)			2 (4.3)	44 (95.7)			9 (19.6)	37 (80.4)			2 (4.3)	44 (95.7)			10 (21.7)	36 (78.3)			2 (4.3)	44 (95.7)
2−4	310	226 (72.9)	84 (27.1)			39 (12.6)	271 (87.4)			76 (24.5)	234 (75.5)			14 (4.5)	296 (95.5)			159 (51.3)	151 (48.7)			37 (11.9)	273 (88.1)
N				0.001				0.237				0.002				0.11				0.469				0.038
0	47	23 (48.9)	24 (51.1)			3 (6.4)	44 (93.6)			3 (6.4)	44 (93.6)				47 (100.0)			20 (42.6)	27 (57.4)			1 (2.1)	46 (97.9)
1−3	309	225 (72.8)	84 (27.2)			38 (12.3)	271 (87.7)			83 (26.9)	226 (73.1)			16 (5.2)	293 (94.8)			149 (48.2)	160 (51.8)			38 (12.3)	271 (87.7)
Stage				0.000				0.039				0.001				0.214				0.000				0.045
I−II	30	1 (3.3)	29 (96.7)				30 (100.0)				30 (100.0)				30 (100.0)			1 (3.3)	29 (96.7)				30 (100.0)
III−IV	326	247 (75.8)	79 (24.2)			41 (12.6)	285 (87.4)			86 (26.4)	240 (73.6)			16 (4.9)	310 (95.1)			168 (51.5)	158 (48.5)			39 (12.0)	287 (88.0)
EGFR 19 DEL				0.000				0.019				0.179				0.536				0.000				0.713
Negative	298	188 (64.8)	102 (35.2)			28 (9.4)	270 (90.6)			66 (22.1)	232 (77.9)			14 (4.7)	284 (95.3)			124 (41.6)	174 (58.4)			31 (10.4)	267 (89.6)
Positive	67	60 (89.6)	7 (10.4)			13 (19.4)	54 (80.6)			20 (29.9)	47 (70.1)			2 (3.0)	65 (97.0)			45 (67.2)	22 (32.8)			8 (11.9)	59 (88.1)
EGFR T790M																								−
Negative	365	248 (69.5)	109 (30.5)			41 (11.2)	324 (88.8)			86 (23.6)	279 (76.4)			16 (4.4)	349 (95.6)			169 (46.3)	196 (53.7)			39 (10.7)	326 (89.3)
Positive	0

EGFR, epidermal growth factor receptor; T790M, a mutation that substitutes methionine for threonine at amino acid position 790; 19 DEL, exon 19 deletion; L858R, substitutions of leucine for arginine in exon 21.

We obtained the metastasis time in months of the T790M-mutant group by calculating the length of time between the confirmed diagnosis date and the metastasis date. We made histograms to analyze better the distribution of metastasis time of different sites (). In patients with metastases, no matter the metastatic site was brain, bone, liver, intrapulmonary, or other sites, the metastases more likely happened before the diagnosis, suggesting that in patients with metastatic lung cancer, most metastases were detected at first diagnosis.

Figure 1

The distribution of metastasis time of different sites in T790M group. The metastases most happened before the diagnosis, suggesting that in patients with metastatic lung cancer, most metastases were detected at first diagnosis.

Nearly 49.6% of the patients with brain metastasis were found to have metastasis before or in the first two months after diagnosis, far more than the number of patients whose metastasis time was distributed at another periods. For bone, liver, inside the chest and outside the chest, the proportion of patients whose metastasis time spread before or in the first two months after diagnosis was 60.5%, 42.2%, 61.6%, and 58.3%, respectively, which was far more than those at another period.

To find whether T790M is related to lung cancer metastases, we used the chi-square test to determine whether there was a difference between the T790M-mutant group and EGFR positive-mutant group.

There were 87.7% of patients in the T790M-mutant group and 69.7% of patients in the EGFR positive-mutant group with metastasis. There was a statistically significant difference between the two groups in metastasis (P<0.001) ().

Table 3

Analysis of the association between metastases and T790M

	T790M mutant group	EGFR positive mutant group	χ2	p
No metastases, n (%)	39 (12.3)	108 (30.3)	31.721	<0.001
With metastases, n (%)	277 (87.7)	248 (69.7)
No brain metastases, n (%)	197 (62.3)	324 (88.8)	65.798	<0.001
With brain metastases, n (%)	119 (37.7)	41 (11.2)
No bone metastases, n (%)	121 (38.3)	279 (76.4)	101.692	<0.001
With bone metastases, n (%)	195 (61.7)	86 (23.6)
No liver metastases, n (%)	280 (88.6)	349 (95.6)	11.797	0.001
With liver metastases, n (%)	36 (11.4)	16 (4.4)
No intrapulmonary metastases, n (%)	92 (29.1)	196 (53.7)	41.944	<0.001
With intrapulmonary metastases, n (%)	224 (70.9)	169 (46.3)
No other sites metastases, n (%)	268 (84.8)	326 (89.3)	3.085	0.079
With other sites metastases, n (%)	48 (15.2)	39 (10.7)

EGFR, epidermal growth factor receptor; T790M, a mutation that substitutes methionine for threonine at amino acid position 790.

There were 37.7% of patients in the T790M-mutant group and 11.2% of patients in the EGFR positive-mutant group with brain metastasis. There was a statistically significant difference between the two groups in brain metastasis (P<0.001; ).

There were 61.7% of patients in the T790M-mutant group and 31.6% of patients in the EGFR positive-mutant group with bone metastasis. There was a statistically significant difference between the two groups in bone metastasis (P<0.001) ().

There were 11.4% of patients in the T790M-mutant group and 4.4% of patients in the EGFR positive-mutant group with liver metastasis. There was a statistically significant difference between the two groups in liver metastasis (P=0.001; ).

There were 70.9% of patients in the T790M-mutant group and 46.3% of patients in the EGFR positive-mutant group with intrapulmonary metastasis. There was a statistically significant difference between the two groups in intrapulmonary metastasis (P<0.001; ).

Discussion

The metastasis and the resistance against treatment make lung cancer the leading cause of cancer-related death. EGFR is one of the most essential driver genes in both pathological and cancerous processes (2,16). When EGFR binds to a ligand, the downstream signal is activated, mediating proliferation, migration, invasion and suppression of apoptosis (17). The overactivation and mutation of EGFR signaling were proved to be related to poor prognosis in lung cancer (18,19).

Patients with different characteristics are prone to different metastatic sites. Carcinoembryonic antigen (20), size of the tumor, nodal stage, adenocarcinoma (21), presence of bone metastases (22), and EGFR mutation (23), might be the predictive factors for brain metastases. Besides, high serum level of hepatoma-derived growth factor (HDGF) might correlate to bone metastasis (24). Researchers showed that histology, age at diagnosis, and sex influenced on the pattern of metastasis. Women, younger patients, and SCLC patients were more likely to have metastases. Patients with liver or bone metastases had a shorter survival time (7). It was reported that lung adenocarcinoma patients with EGFR mutations were more likely to have distant metastases (25). T790M mutation is a common EGFR mutation in patients with resistance to first-generation EGFR-TKIs. However, the correlation between T790M and metastases remains unclear. Our research found that lung cancer patients with T790M mutation were more likely to have metastases, especially brain metastases, bone metastases, liver metastases, and intrapulmonary metastases. However, the mechanism was still unclear, and further investigation was indispensable, which might be necessary for finding new methods to restrict the development of metastases.

EGFR exon 19 deletion and L858R are driver mutations in NSCLC. EGFR TKIs are effective in treating EGFR positive lung cancer. Comparing with conventional cytotoxic chemotherapy, target therapy improved the overall survival (OS) and reduced the side effects of treatments in lung cancer patients with EFGR mutation (26). The median OS of advanced NSCLC patients treated with combination chemotherapy was 8 to 12 months, and the median progression-free survival (PFS) was 5 to 6 months (27-29). While treated with target therapy, the median OS was 20 to 30 months, and the median PFS was 10 to 14 months (10,30-34). Unfortunately, within a median period of 10–14 months, acquired resistance to first- and second-generation EGFR TKIs happened (35). The occurrence of EGFR T790M mutation in exon 20 was the most common mechanism of EGFR TKI resistance (36-38) Third-generation EGFR TKIs showed significant efficacy in preclinical studies for patients with T790M mutation (39-42). The use of target therapy had extended lung cancer patient life. The metastasis was an essential factor leading to poor prognosis of lung cancer patients with metastasis. Did the metastasis of lung cancer result from the longer lifetime of patients or the drive mutation? The results of our research might answer. From our results, wherever cancer metastasized, the time of metastasis largely concentrated before diagnosis or in the first two months after diagnosis. Even patient lives were prolonged, the metastases were happened before diagnosis or in the first two months after diagnosis. Although these patients all accepted the targeted therapy, the metastasis time seemed not to be related to the longer lifetime. Based on our research, we found that NSCLC patients with T790M had a higher incidence of metastases.

Despite these significant findings, there were also limitations in our study. First, our data were not large enough. Second, this was a retrospective study. Third, the mechanism of the correlation between T790M and metastases still needs more in-depth exploration.

Conclusions

For T790M-mutant patients with metastatic lung cancer, most metastases were detected before diagnosis or in the first two months after diagnosis, which certified that the metastases not related to the prolonged lifetime of patients or the use of target therapy. Moreover, NSCLC patients with T790M mutation had a higher incidence of metastases. We should conduct further studies to explore the mechanism of the correlation between T790M and metastases.

The article’s supplementary files as