Prognostic signicance of NDRG2 combined with EGFR-sensitizing mutations in lung adenocarcinoma

Background: N-myc downstream-regulated gene 2 (NDRG2) plays a substantial role in lung adenocarcinoma (LUAD). Epidermal growth factor receptor (EGFR)-sensitizing mutation could signicantly improve prognosis in patients with LUAD. Here, we aimed to elucidate the prognostic value of NDRG2 combined with EGFR-sensitizing mutations in patients with LUAD. Methods: Lung parenchyma specimens obtained during surgery or CT-guide percutaneous lung puncture biopsy for the NDRG2 protein and EGFR genomic testing were obtained. Associations between NDRG2/EGFR and clinicopathological characteristics of patients with LUAD were extracted from the Tianjin First Central Hospital in China between June 2013 and June 2014. Results: The expression of NDRG2 was signicantly decreased in patients with LUAD. Expressions of NDRG2 and EGFR-sensitizing mutations showed positive correlations with survival. Expression of NDRG2 and EGFR-sensitizing mutations were associated with the longer overall survival (OS), disease-free survival (DFS) and progression-free survival (PFS). Advanced stages were signicantly associated with low expression of NDRG2. In multivariate analysis, compared with other patients, NDRG2 (+)/EGFR (+) was independently associated with prolonged OS and PFS. Conclusion: NDRG2 combined with EGFR-sensitizing mutations might be valuable markers to evaluate the prognosis of LUAD patients. t-test. The relationship between NDRG2 expression levels and factors analyzed using the test. EGFR and with marker

Normal tissues were removed from at least 5 cm away from the edge of the tumors. All tumor samples obtained during surgery or CT-guide percutaneous lung puncture biopsy underwent genomic testing to detect EGFR mutations.
EGFR mutation analysis DNA from formalin-xed, para n-embedded tumor tissue and matched blood samples was extracted. Comprehensive genomic pro ling was performed by next generation sequencing (NGS) with a 37 or 450 cancer related gene panel covering the whole EGFR gene at a mean coverage depth of > 800x. The genomic alterations including single base substitution and deletions were assessed.

Immunohistochemistry
In the present study, formalin-xed para n-embedded tissue sections (thickness, 4-µm) were used for detecting the expression of NDRG2. Tissue sections were dewaxed, rehydrated, antigen-retrieved, and cooled to room temperature. The sections were incubated with mouse monoclonal anti-NDRG2 antibody (1:500 dilution, Santa Cruz Biotechnology, Santa Cruz, CA, USA) at 4ºC overnight, rinsed with phosphate-buffered saline (PBS), and incubated with horseradish peroxidase (HRP)-labeled goat anti-mouse secondary antibody for 60 min. NDRG2 expression was revealed using 3,3′diaminobenzidine (DAB) as the chromogen. Negative control was performed by replacing the primary antibody with normal mouse serum. The brown or yellow staining was identi ed as a positive expression. The total staining score of 0-12 was considered in a semi-quantitative manner and strati ed as follows: negative (-, range of score: 0-1), weak (+, range of score: 2-4), moderate (++, range of score: 5-8), or strong (+++, range of score: 9-12). The tumor specimens were divided into the low-expression group (range of score: 0-4) and the high-expression group (range of score: 5-12) [12].

Western blot analysis
Total protein concentration was measured by the bicinchoninic acid (BCA) assay kit. The proteins were separated by 10% sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene uoride (PVDF) membranes. The membranes were incubated with mouse anti-human NDRG2 antibody (1:500, Santa Cruz Biotechnology, Santa Cruz, CA, USA) and GAPDH antibody (1:1000, Beyotime Institute of Biotechnology, Haimen, China) at 4°C overnight after being blocked with 5% non-fat milk for 1 h. The membranes were washed and incubated with HRP-conjugated secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA). The blots were visualized using an enhanced chemiluminescence kit (Amersham Pharmacia Biotech, Arlington Heights, IL, USA) according to the manufacturer's instructions. Each experiment was performed in triplicate.

Quantitative real-time PCR
The total RNA was extracted from the fresh tissues using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions.
The rst chain of cDNA was synthesized by reverse transcription with TaqMan® Reverse Transcription Reagents (Applied Biosystems, Grand Island, NY). GAPDH was used as internal control. The sequences of the primers were: NDRG2, forward-5'-ATG GCG GAG CTG CAG GAG GTC-3', and reverse-5'-AAC AAG GGC CAT TCA ACA GGA GAC-3'; GAPDH, forward-5'-GCC TCA AGA TCA GCA AT-3' and reverse-5'-AGG TCC ACC ACT GAC ACG TT-3'. RT-PCR reaction was performed using the CFX96 Touch PCR system (Bio-Rad). The relative mRNA expression of NDRG2 was calculated by the -2 ∆∆Ct method. Each experiment was undertaken in triplicate.

Statistical analysis
Statistical analysis was performed with SPSS software (version 18.0; SPSS, Chicago, IL, USA). Continuous data were presented as means ± standard deviations (SD). Differences between categorical variables were investigated by the chi-square test. The differences in means between groups were analyzed using the Student's t-test. The relationship between NDRG2 expression levels and clinicopathological factors was analyzed using the Wilcoxon-Mann-Whitney test.
The patient's survival curves were evaluated using the Kaplan-Meier method, and PFS, DFS and OS were compared using the log-rank test as well as univariate Cox proportional-hazards regression model. To adjust for the clinically important factors, multivariate Cox proportional-hazards regression model was utilized. All statistical tests were two-sided, and P < 0.05 was considered statistically signi cant.

Results
Clinical characteristics of the patients In total, 89 patients with lung adenocarcinoma were included in this retrospective study. The median age of the study population was 65.6 years old (range, 38-86) and 52 (58.4 %) were male. NDRG2 expression was positive in 36 (40.4 %) patients. EGFR mutations (denoted as EGFR (+)) were detected in 29 (32.6 %) patients. 69 (77.5 %) patients received postoperative adjuvant chemotherapy. The baseline characteristics of the study population and their associations with NDRG2 and EGFR are summarized in Table 1.
Driver mutations in the LUAD The EGFR mutation patterns are shown in Table 2. The most frequent mutations found in EGFR in turn were the exon 19 deletion (51.7 %), followed by the activation-mutation of exon 21 (L858R) (41.4 % of the cases).
Expression of NDRG2 and EGFR mutations correlates with clinicopathological characteristics of LUAD The expression and distribution of NDRG2 in 34 cases of lung adenocarcinoma were examined by IHC (Figure 1 A-C). In total, 23 (67.6 %) cases were highly positive for NDRG2 (denoted as NDRG2 (+)) while 11 (32.4 %) cases were negative-to-weakly positive (denoted as NDRG2 (-)). 34 cases of lung adenocarcinoma were also examined by Western blotting and qRT-PCR revealed that the expression of NDRG2 at the protein (Figure 1 D) and mRNA (Figure 1 E) levels in LUAD was signi cantly lower compared with adjacent non-carcinoma tissues.
To evaluate the expression of NDRG2 in different stages of LUAD, we divided LUAD samples into four groups according to the disease stage. The expression of NDRG2 decreased as the TNM stage increases ( Figure 1 F-G).
Statistically signi cant correlations were found between the level of NDRG2 expression and the presence of vascular invasion (P < 0.001), and TNM stage (P < 0.001). Meanwhile, statistically signi cant correlations were found between the wild-type of EGFR (denoted as EGFR (-)) and the poor degree of histological differentiation (P = 0.001), the presence of vascular invasion (P<0.001), TNM stage (P < 0.001) and smoking history (P = 0.003) ( Table  1).

Prognostic signi cance of NDRG2 and EGFR
The Kaplan-Meier analysis for postoperative OS and DFS showed that patients with preserved NDRG2 and mutant EGFR had longer OS, and better DFS than patients with reduced NDRG2 and wild-type EGFR ( Figure 2). The results also showed that iodine-125 radioactive seeds brachytherapy for advanced LUAD with low expression level of NDRG2 and wild-type EGFR led to signi cantly poor OS and PFS ( Figure 3).
In multivariate analysis, NDRG2 and EGFR may be two prognostic factors for OS of LUAD patients (Table 4).
Kaplan-Meier analysis ( Figure 4) showed that OS was lower in the others group than in the NDRG2 (+)/EGFR (+) group; the DFS and PFS were also signi cantly poorer in the former.

Discussion
To improve the prediction of lung cancer survival, several tumor markers have been assessed and extensively used [13,14], but each marker has its own speci city and sensitivity, which might lead to limitations in prognostic ability. The combined detection of tumor markers may be of great importance for improving the prediction of lung cancer survival.
MYC in uences growth, proliferation, differentiation, and apoptosis of cancer cells through regulating the expression of numerous genes [15]. In addition, MYC governs events associated with tumor progression, including genetic stability, migration, and angiogenesis [16]. Two human cDNAs, encoding NDRG3 and NDRG4, are homologous to NDRG1. These two genes, together with NDRG1 and a previously deposited cDNA (designated NDRG2), constitute the NDRG gene family, which is identi ed as a novel type of myc-repressed genes and believed to play important roles in diverse biological processes of human cancer [17]. Previous studies reported that NDRG2 was associated with human lung cancer, and the decreased expression of NDRG2 was correlated with a worse outcome of lung cancer patients [12,18]. Similarly, we showed signi cantly decreased NDRG2 levels in patients with lung adenocarcinoma. NDRG2 levels showed opposite correlated with cancer progression on the basis of the TNM staging system. Advanced disease stage was signi cantly associated with low NDRG2, suggesting that NDRG2 can be a promising tumor marker of lung adenocarcinoma re ecting systemic tumor burden. We also con rmed that NDRG2 positivity was correlated with EGFR mutations.
EGFR mutations are detectable in approximately 10-35% of lung adenocarcinoma [19]. Several EGFR-targeting tyrosine kinase inhibitors (TKIs) have demonstrated higher objective response rates (ORR) and PFS over chemotherapy in patients with advanced disease, making rst-line EGFR TKIs the treatment of choice for this subtype of adenocarcinoma [20,21,22]. Catherine Labbé et al. demonstrated that patients with dual TP53/EGFR mutations, especially missense mutations, had marginally lower response rates and shorter PFS when treated with EGFR TKI therapy [23]. Feng Wang et al. also showed that mutant TP53 is a poor prognostic factor in LUAD patients, and the prognosis of TP53/EGFR co-mutation is worse [24]. They also demonstrated that, there was no signi cant difference in OS bene t when patients with EGFR mutations were compared with those with EGFR wild type. However, our data did show a signi cant difference in OS of patients with EGFR mutations. Ji Young Park et al. also noted that patients with EGFR-positive lung adenocarcinoma had longer OS than those with EGFR-negative malignancy [6].
In this current study, we rst observed the correlation between NDRG2 and EGFR in LUAD patients and the prediction of LUAD prognosis based on the expression of NDRG2 and mutant EGFR. Our results also showed that NDRG2 expression and EGFR mutation correlated with vascular invasion, adjuvant chemotherapy and TNM stage of LUAD patients. MYC and mutant EGFR have been identi ed as potential biomarkers that can predict the e cacy of targeted therapy [25,26,27]. We reclassi ed the patients into two groups using the marker combinations of NDRG2 and EGFR to further investigate the prognostic impact of NDRG2 and EGFR and found that patients with NDRG2 (+)/EGFR (+) had a signi cantly better OS and PFS than those with other marker combinations.
This study has several limitations. First, it was a retrospective study performed at a single center. Thus, a wide multicentric study is probably required to con rm our results. Second, the sample size was very limited. The present study, in itself, cannot be used to change clinical practice and to propose the use of a new test. Those results have to be validated using a large sample size. Hence, with increasing numbers, it should be possible to evaluate the true effect of NDRG2/EGFR as both a prognostic and predictive variable.

Conclusions
In summary, this study reported the different expression levels of NDRG2 in patients with LUAD. In addition, for the rst time, the relationship between NDRG2/EGFR and clinicopathological characteristics of patients with LUAD, especially prognosis status, was investigated. NDRG2/EGFR can be used as a novel prognostic biomarker for patients with LUAD. The authors declare no support from any organizations for the submitted work. The design of the study, the analyses and the writing of the manuscript were solely the responsibility of the authors.

Ethics approval and consent to participate
This study was conducted in accordance with the Declaration of Helsinki and was con rmed by the Ethics Committee of Tianjin First Central Hospital of Nankai University (Tianjin, China; approval no. 2018N054KY). All patients signed the written informed consent forms.

Consent for publication
Not applicable

Availability of data and materials
The data of the current research are available from the corresponding author on a reasonable request.

Competing interests
The authors declare that there are no con icts of interest.

Funding
This study was nancially supported by the Science & Technology Program of Tianjin First Central Hospital, Nankai University (Grant Nos. CM201803 and CF201807) in the design of the study, collection, analysis, and interpretation of data.
Authors' contributions BY, GSW, TJ, and WDZ conceived of the study. LJ, HGZ, TX, and XHL performed data analysis for experiments. BY, HGZ, LJ, and FY. drafted the nal version of the manuscript and gure legends. BY, XPL, XHL, and LZ revised the gures, added critical content to the discussion, and were responsible for revising all portions of the submitted portion of the manuscript. TX and HX performed the experiments using lung adenocarcinoma and control tissue. All contributors meet the criteria for authorship. All of the authors read and approved the nal manuscript.  Table 3. Univariate Cox regression analysis for overall survival, disease-free survival and progression-free survival.

Figure 1
The expression level of NDRG2 in LUAD patients and in different stages. Immunohistochemistry showed the expression level of NDRG2 in adjacent normal tissues ( Figure 1A), LUAD tissues ( Figure 1B), and negative control ( Figure 1C) (200× magni cation). The expression level of NDRG2 was determined by western blot assay ( Figure 1D) and qRT-PCR ( Figure 1E). It was signi cantly downregulated in patients with LUAD tissues compared with that in adjacent normal tissues at both protein and mRNA levels.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.