Classification of Non-Small Cell Lung Cancer Using Significance Analysis of Microarray-Gene Set Reduction Algorithm

Non-small cell lung cancer (NSCLC) has a poor prognosis and improved therapies are needed. Expression of EphA2 is increased in NSCLC metastases. In this study, we investigated EphA2 mutations in NSCLC and examined molecular pathways involved in NSCLC. Tumor and cell line DNA was sequenced. One EphA2 mutation was modeled by expression in BEAS2B cells, and functional and biochemical studies were conducted. A G391R mutation was detected in H2170 and 2/28 squamous cell carcinoma patient samples. EphA2 G391R caused constitutive activation of EphA2 with increased phosphorylation of Src, cortactin, and p130(Cas). Wild-type (WT) and G391R cells had 20 and 40% increased invasiveness; this was attenuated with knockdown of Src, cortactin, or p130(Cas). WT and G391R cells demonstrated a 70% increase in focal adhesion area. Mammalian target of rapamycin (mTOR) phosphorylation was increased in G391R cells with increased survival (55%) compared with WT (30%) and had increased sensitivity to rapamycin. A recurrent EphA2 mutation is present in lung squamous cell carcinoma and increases tumor invasion and survival through activation of focal adhesions and actin cytoskeletal regulatory proteins as well as mTOR. Further study of EphA2 as a therapeutic target is warranted.

Understanding of intratumor heterogeneity (ITH) among different non-small cell lung cancer (NSCLC) subtypes is necessary. Whether circulating tumor DNA (ctDNA) profile could represent these ITH is still an open question. We performed 181 multi-region tumor tissues sequencing and matched ctDNA sequencing from 32 operative NSCLC to compare ITH among different NSCLC subtypes, including EGFR-mutant lung adenocarcinoma (LUAD), KRAS-mutant LUAD, EGFR&KRAS-wild-type LUAD, and lung squamous cell carcinoma (LUSC), and examine potential value of ctDNA for ITH analysis. ITH is evaluated by ITH index (ITHi). If the somatic genetic alteration is shared by all the tissue regions, it is defined as trunk mutation. Otherwise, it is called branch mutation. The ITHi will be higher, if the tumor has less trunk mutations. We found EGFR-mutant LUAD showed significantly higher ITHi than KRAS-mutant LUAD/wild-type LUAD (P = 0.03) and numerically higher ITH than LUSC. For trunk mutations, driver mutations were identified at a higher proportion than passenger mutations (60% vs. 40%, P = 0.0023) in overall, especially in EGFR-mutant LUAD (86% vs. 14%, P = 0.0004), while it was opposite in KRAS-mutant LUAD (40% vs. 60%, P = 0.18). For branch mutations, the proportions of driver mutations and passenger mutations were similar for each NSCLC subtype. ctDNA analysis showed unsatisfactory detections of tumor-derived trunk and branch mutations (43% vs. 23%, P = 4.53e-6) among all NSCLC subtypes. In summary, EGFR-mutant LUAD has the highest ITH than other NSCLC subtypes, offering further understanding of tumorigenesis mechanisms among different NSCLC subtypes. Besides, ctDNA maybe not an appropriate method to reflect ITH.

BackgroundReduced forced expiratory volume in one second (FEV1) has been linked to non small cell lung cancer (NSCLC). However, it is unclear whether all or only certain histological subtypes of NSCLC are associated with reduced FEV1. Moreover, there is little information on whether gender modifies this relationship. Using a large tissue registry, we sought to determine the relationship between FEV1 and subtypes of NSCLC and determine whether this relationship is modified by gender.MethodsWe used data from patients who underwent tumor resection for NSCLC at a teaching hospital in Vancouver and had various pre-operative clinical measurements including FEV1. We divided the cohort into quartiles of predicted FEV1 and using both logistic and linear regression modeling techniques determined whether FEV1 was related to the occurrence of adeno or squamous cell carcinoma in men and women.ResultsThere were 610 patients in the study (36% females). On average, women were more likely to have adenocarcinoma than were men (72% of all cases of NSCLC in women versus 40% in men; p < 0.001). In women, there was no significant relationship between FEV1 and the risk of any histological subtypes of NSCLC. In men, however, there was an inverse relationship between the risk of adenocarcinoma and FEV1 such that the lowest quartile of FEV1 was 47% less likely to have adenocarcinoma compared with the highest FEV1 quartile (adjusted odds ratio, 0.52; 0.28 to 0.98; p for trend, 0.028). The reverse was observed for squamous cell carcinoma.ConclusionIn individuals undergoing lung resection for NSCLC, the risk of adenocarcinoma and squamous cell carcinoma of the lung varies as a function of FEV1, independent of smoking intensity in men but not in women.Clinical ImplicationsThese data indicate that women are much more susceptible to adenocarcinoma than are men especially when they have normal or near normal lung function. It may thus be useful to conduct periodic surveillance chest radiographs in asymptomatic female smokers (or ex-smokers) to ascertain peripheral nodules or masses before distant metastases occur since adenocarcinomas tend to metastasize earlier in the disease course than squamous cell carcinomas.

Significant differences in the aberrant methylation of genes exist among various histological types of non-small cell lung cancer (NSCLC), which includes adenocarcinoma (AC) and squamous cell carcinoma (SCC). Different chemotherapeutic regimens should be administered to the two NSCLC subtypes due to their unique genetic and epigenetic profiles. The purpose of this meta-analysis was to generate a list of differentially methylated genes between AC and SCC. Our meta-analysis encompassed 151 studies on 108 genes among 12946 AC and 10243 SCC patients. Our results showed two hypomethylated genes (CDKN2A and MGMT) and three hypermethylated genes (CDH13, RUNX3 and APC) in ACs compared with SCCs. In addition, our results showed that the pooled specificity and sensitivity values of CDH13 and APC were higher than those of CDKN2A, MGMT and RUNX3. Our findings might provide an alternative method to distinguish between the two NSCLC subtypes.

Lung cancer (LC) is a leading cause of cancer-associated mortalities worldwide. Adenocarcinoma (AC) and squamous cell carcinoma (SCC) account for ~70% of all cases of LC. Since AC and SCC are two distinct diseases, their corresponding prognostic genes associated with patient survival time are expected to be different. To date, only a few studies have distinguished patients with good prognosis from those with poor prognosis for each specific subtype. In the present study, the Cox filter model, a feature selection algorithm that identifies subtype-specific prognostic genes to incorporate pathway information and eliminate redundant genes, was adopted. By applying the proposed model to data on non-small cell lung cancer (NSCLC), it was demonstrated that both redundant gene elimination and search space restriction can improve the predictive capacity and the model stability of resulting prognostic gene signatures. To conclude, a pre-filtering procedure that incorporates pathway information for screening likely irrelevant genes prior to complex downstream analysis is recommended. Furthermore, a feature selection algorithm that considers redundant gene elimination may be preferable to one without such a consideration.

Histological subtypes of non-small cell lung cancer (NSCLC) are crucial for systematic treatment decisions. However, the current studies which used noninvasive radiomic methods to classify NSCLC histology subtypes mainly focused on two main subtypes: squamous cell carcinoma (SCC) and adenocarcinoma (ADC), while multi-subtype classifications that included the other two subtypes of NSCLC: large cell carcinoma (LCC) and not otherwise specified (NOS), were very few in the previous studies. The aim of this work was to establish a multi-subtype classification model for the four main subtypes of NSCLC and improve the classification performance and generalization ability compared with previous studies. In this work, we extracted 1029 features from regions of interest in computed tomography (CT) images of 349 patients from two different datasets using radiomic methods. Based on "three-in-one" concept, we proposed a model called SLS wrapping three algorithms, synthetic minority oversampling technique, ℓ2,1-norm minimization, and support vector machines, into one hybrid technique to classify the four main subtypes of NSCLC: SCC, ADC, LCC, and NOS, which could cover the whole range of NSCLC. We analyzed the 247 features obtained by dimension reduction, and found that the extracted features from three methods: first order statistics, gray level co-occurrence matrix, and gray level size zone matrix, were more conducive to the classification of NSCLC subtypes. The proposed SLS model achieved an average classification accuracy of 0.89 on the training set (95% confidence interval [CI]: 0.846 to 0.912) and a classification accuracy of 0.86 on the test set (95% CI: 0.779 to 0.941). The experiment results showed that the subtypes of NSCLC could be well classified by radiomic method. Our SLS model can accurately classify and diagnose the four subtypes of NSCLC based on CT images, and thus it has the potential to be used in the clinical practice to provide valuable information for lung cancer treatment and further promote the personalized medicine.

Introduction: Recent advances in the treatment of non-small-cell lung cancer (NSCLC) with new agents require accurate histological subtyping at diagnosis in order to avoid the higher risk of an adverse response and obtain a maximum therapeutic response. However, interobserver variability, tumor heterogeneity and the degree of differentiation may influence the decision concerning a pathological diagnosis of NSCLC. We therefore in this study attempted to identify specific microRNAs as standardized biomarkers with high sensitivity and specificity for distinguishing between squamous cell carcinoma (SCC) and adenocarcinoma (AC). Methods: Quantitative real time polymerase chain reaction (qRT-PCR) based microRNA array was performed to identify microRNAs differentially expressed between SCC and AC using 86 resected NSCLC samples as well as adjacent normal tissues. The results were confirmed by independent qRT-PCR assays with the same test samples and an additional 88 validation samples and the usefulness of the identified microRNAs as biomarkers to distinguish between SCC and AC was evaluated. Results: Three microRNAs (hsa-miR-196b, hsa-miR-205 and hsa-miR-375) were identified. Discriminant analysis combining all 3 microRNAs appeared to distinguish SCC from AC accurately in the test samples and the validation samples, showing sensitivity and specificity of 76% and 80%, and 85% and 83%, respectively. Conclusions: Hsa-miR-196b, hsa-miR-205 and hsa-miR-375 were identified as biomarkers capable of distinguishing between lung SCC and lung AC. These newly identified microRNAs may prove to be highly valuable molecular markers for the classification of NSCLC histologic subtypes, and could potentially contribute to the pathogenesis of each subtype of NSCLC. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 4956. doi:10.1158/1538-7445.AM2011-4956

Derlin-1 Is Overexpressed in Non-Small Cell Lung Cancer and Promotes Cancer Cell Invasion via EGFR-ERK–Mediated Up-Regulation of MMP-2 and MMP-9

Objectives: Non-small cell lung cancer (NSCLC), the most prevalent type of lung cancer, includes subtypes such as adenocarcinoma (ADC) and squamous cell carcinoma (SCC), which require distinct management approaches. Accurately differentiating NSCLC subtypes based on diagnostic imaging remains challenging. However, the extraction of radiomic features—such as first-order statistics (FOS), second-order statistics (SOS), and fractal dimension texture analysis (FDTA) features—from magnetic resonance (MR) images supports the development of quantitative NSCLC assessments. Methods: This study aims to evaluate whether the integration of FDTA features with FOS and SOS texture features in MR image analysis improves machine learning classification of NSCLC into ADC and SCC subtypes. The study was conducted on 274 MR images, comprising ADC (n = 122) and SCC (n = 152) cases. From the segmented MR images, 93 texture features were extracted. The random forest algorithm was used to identify informative features from both FOS/SOS and combined FOS/SOS/FDTA datasets. Subsequently, the k-nearest neighbors (kNN) algorithm was applied to classify MR images as ADC or SCC. Results: The highest performance (accuracy = 0.78, precision = 0.81, AUC = 0.89) was achieved using 37 texture features selected from the combined FOS/SOS/FDTA dataset. Conclusions: Incorporating fractal descriptors into the texture-based classification of lung MR images enhances the differentiation of NSCLC subtypes.

Non-small cell lung cancer (NSCLC) is one of the leading causes of cancer-related death, including two major pathologic subtypes, adenocarcinoma (AD) and squamous cell carcinoma (SCC). There are significant differences in the aberrant gene profiling and molecular characteristics among LUAD and LUSC, which lead to the different chemotherapeutic regimens administered to the two NSCLC subtypes. Therefore, unique genetic or epigenetic accurately distinguishing these two subtypes is critical for the diagnosis and treatment of patients. In this study, we tried to develop and validate novel long non-coding RNA (lncRNA)-based signatures to facilitate the diagnosis and individualized treatment of LUAD and LUSC. To analyze the distinct lncRNA profiling in LUAD and LUSC, the paired tissue samples of RNA sequencing or microarray data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) was used. We found that lncRNA expression pattern is different in LUAD and LUSC, and three up-regulated and down-regulated lncRNAs expressions in LUAD and LUSC tissues were validated. In addition, knockdown of the up-regulated lncRNA AFAP1-AS1 and LINC00511 impaired LUAD cell proliferation, while knockdown of PVT1 inhibited LUSC cell growth. Furthermore, four 6-lncRNAs signature expression patterns were found to be significantly associated with LUAD and LUSC patient overall and progression-free survival. Taken together, the different lncRNAs signature might accurately distinguish LUAD and LUSC and predict clinical outcome for these two different subtypes of NSCLC.

Non-small cell lung cancer (NSCLC) remains the most frequent malignancy worldwide, and lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) represent two major subtypes. LINC00628 has been demonstrated to promote LUAD progression; however, its clinical role in NSCLC remains elusive. The aim of the present study was to analyze the expression of long intergenic non-protein coding RNA 628 (LINC00628) in NSCLC, including in the LUAD and LUSC subtypes. In addition, its roles in NSCLC development and prognosis were also examined. Data from The Cancer Genome Atlas (TCGA) database were first used to assess the expression and prognostic potential in both LUAD and LUSC, then LINC00628 expression in 128 NSCLC tissues was measured using reverse transcription-quantitative PCR. A receiver operating characteristic curve was used to assess the ability of LINC00628 to discriminate between patients with LUAD and LUSC. Kaplan-Meier curves were used to analyze the relationship between LINC00628 expression and the overall survival of patients. Cox regression analysis was used to explore the potential prognostic factors that might be independently associated with NSCLC overall survival. Both in silico and tissue analysis data indicated that the expression of LINC00628 was significantly upregulated in NSCLC tissue compared with matched normal controls (P<0.001). LINC00628 expression levels were also significantly higher in LUAD cases than in patients with LUSC (P<0.001). In addition, LINC00628 could discriminate LUAD from LUSC cases. The expression of LINC00628 was significantly associated with tumor size (P=0.013), histological type (P=0.009), lymph node metastasis (P=0.021) and TNM stage (P=0.008). Survival analysis based on data from both TCGA and patients included in the present study identified an association between LINC00628 and overall survival in LUAD, but this relationship was not observed in LUSC for TCGA data. Cox regression analysis demonstrated that high LINC00628 expression was associated with poor overall survival in patients with LUAD (P=0.001), but not in patients with LUSC (P=0.088). In conclusion, LINC00628 expression was upregulated in NSCLC and associated with patient prognosis. Patients with LUAD had higher LINC00628 expression levels than those with LUSC, and increased LINC00628 served as an independent prognostic factor in LUAD, but not LUSC.

The aim of the study was to investigate how alterations in the PI3K pathway correlate with non-small cell lung cancer subtypes squamous cell carcinoma (SSC) and adenocarcinoma (ADCA). We analyzed copy number variation and protein expression of INPP4B, protein expression of pAkt, PDPK1, and PTEN and mutational status of PIK3CA and PTEN in 180 cases. Nineteen% displayed loss of INPP4B copy, whereas 47% lacked expression, both showing correlation with SCC. Elevated pAkt expression was seen in 63% of all cases, also correlating to SCC. PDPK1 was expressed in 70%, more in male than female patients. Regarding PTEN, 50% displayed loss of expression, of which seven were identified with mutations in the phosphatase domain. We detected nine cases (5%) of PIK3CA mutations, all identified as the E545K hot spot mutation in the helical domain, all except one in SCC. When analyzing all PI3K pathway components together, we show that patients with at least one alteration in the PI3K pathway are twice as likely to have SCC, than ADCA. Interestingly, we also found a strong correlation between high pAkt expression and PTEN expression. As comparison, we also analyzed mitogen-activated protein kinase (MAPK) pathway genes, where we identified fifteen KRAS mutations (8%) and one BRAF mutation (1%), significantly associated to ADCA. No association was found to the Gly972Arg polymorphism of IRS-1, involved in activation of both PI3K and MAPK pathways. In conclusion, we show here that several components of the PI3K pathway, alone and in combination, are correlated to development of SCC of the lung.

To investigate and analyze the expression of fascin and CK14 in multiple histological types of cancer and to explore the potential value of the two proteins as markers in diagnosis and differential diagnosis of various cancer types. Tissue microarray containing esophageal squamous cell carcinoma (SCC), lung SCC, larynx SCC, uterine cervical SCC, SCC of external genital organs, lung adenocarcinoma, gastric adenocarcinoma, colorectal adenocarcinoma, heptocellular carcinoma, pancreatic ductal adenocarcinoma, breast infiltrating ductal carcinoma, thyroid papillary carcinoma, uterine endometrioid adenocarcinoma, ovarian serous adenocarcinoma and renal clear cell carcinoma, 30 cases each, as well as corresponding normal controls was constructed. The expression of fascin and CK14 among different types of carcinoma and corresponding normal controls was detected by immunohistochemistry. In normal esophagus, bronchus, larynx, uterine cervix and skin, fascin was mainly expressed in the basal cells or reserve cells, but the expression was diffuse in esophageal SCC, lung SCC, larynx SCC, uterine cervical SCC and SCC of external genital organs, with a positive rate of 90.0%, 90.0%, 96.7%, 78.6% and 89.7%, respectively. In the normal tissue of other organs, except breast and uterine endometrium, fascin was negative. In lung adenocarcinoma, gastric adenocarcinoma, colorectal adenocarcinoma, hepatocellular carcinoma, pancreatic ductal adenocarcinoma, breast infiltrating dutal adenocarcinoma, thyroid papillary carcinoma, uterine endometrioid adenocarcinoma, ovarian serous adenocarcinoma and renal clear cell carcinoma, the positive rates were 38.0%, 23.3%, 14.3%, 10.3%, 73.3%, 13.3%, 6.7%, 60.0%, 66.7% and 10.0%, respectively. The difference between fascin expression in SCC and in other histological types was statistically significant (P < 0.001). CK14 was mainly expressed in the basal cells, reserve cells or myoepithelia of normal tissues. The positive rates of CK14 were 76.7%, 36.7%, 83.3%, 60.7% and 96.3% in esophageal SCC, lung SCC, larynx SCC, uterine cervical SCC and SCC of external genital organs, respectively. It was weak and focal in lung adenocarcinoma, gastric adenocarcinoma, colorectal adenocarcinoma, hepatocellular carcinoma, pancreatic ductal adenocarcinoma, breast infiltrating dutal adenocarcinoma, thyroid papillary carcinoma, uterine endometrioid adenocarcinoma, ovarian serous adenocarcinoma, and renal clear cell carcinoma, with a positive rate of 13.3%, 13.3%, 20.7%, 41.4%, 46.7%, 6.7%, 40.0%, 13.3%, 20.0% and 6.7%, respectively. The difference between CK14 expression in SCC and in other histological types was statistically significant (P < 0.001). The difference between co-expression of fascin/CK14 in SCC and in other histological types was also statistically significant (P < 0.001). Fascin and CK14 are highly expressed in SCC, compared with other histological types of carcinoma. Combination of fascin and CK14 should be a valuable marker in diagnosis and differential diagnosis of carcinoma.

Polymorphisms in ERCC1 and ERCC2/XPD genes and carcinogen DNA adducts in human lung.

Carcinogenesis and tumor proliferation are characterized by a complex interaction of cancer cells with the tumor microenvironment. In particular, a tumor-promoting effect can be assumed for the stroma and its fibroblasts. An influence of the immune system on non small cell lung cancer (NSCLC) is now also suspected. In our study, we examined 309 sections of squamous cell carcinoma (SCC), a subtype of NSCLC. We determined the cell densities and areas of the different tissues in SCC using the software QuPath. Spearman rank correlation showed a significant positive correlation between the different tumor cell densities and stromal cell densities, and between tumor cell densities and immune cell densities. Overall survival curves by the Kaplan–Meier method revealed a prominent negative curve in cases of low immune cell density. Based on our results, we can assume a positive influence of the tumor microenvironment, especially the stromal cells, on tumor proliferation in SCC. We have also revealed that low density of immune cells is prognostically unfavorable.