Clinical value of combined SHOX2 and RASSF1A methylation in lung cancer diagnosis across tissue and liquid biopsy samples: a systematic review and meta-analysis

Abstract

Lung cancer remains the leading cause of cancer-related mortality worldwide, with poor prognosis in advanced stage diseases. Although early diagnosis has the potential to improve patient outcomes, current diagnostic methods remain suboptimal, highlighting the need for accurate molecular biomarkers. We systematically reviewed 49 studies to evaluate the diagnostic performance of SHOX2 methylation, RASSF1A methylation, and their combined panel for lung cancer detection. The combined SHOX2/RASSF1A methylation panel demonstrated a pooled sensitivity of 77.8% (95% CI: 72.3%–82.5%) and specificity of 89.0% (95% CI: 86.6%–91.1%), with an HSROC area under the curve (AUC) of 0.916. SHOX2 methylation alone yielded a sensitivity of 69.4% and specificity of 91.7%, whereas RASSF1A methylation showed lower sensitivity (45.7%) but the highest specificity (93.8%). Pairwise comparisons demonstrated that the combined panel significantly improved sensitivity compared with either SHOX2 or RASSF1A alone while maintaining specificity comparable to SHOX2, although lower than that of RASSF1A. Subgroup analyses showed that assay method and pathological subtype contributed to differences in pooled sensitivity, whereas leave-one-out sensitivity analyses confirmed the robustness of the pooled estimates. In conclusion, the combined SHOX2/RASSF1A methylation panel provides a more balanced diagnostic performance than either biomarker alone and represents a promising adjunctive approach for lung cancer detection. Future studies should focus on standardizing detection methods, integrating these biomarkers with other diagnostic modalities, and evaluating their diagnostic performance in early-stage lung cancer to further enhance their clinical utility.

1 Introduction

Lung cancer remains the leading cause of cancer-related deaths worldwide, accounting for over two million new cases and 1.8 million deaths annually (1–3). Despite advancements in treatment strategies, including targeted therapies and immunotherapy, the prognosis for lung cancer, especially in advanced stages, remains dismal, with 5-year survival rates as low as 15% for stage III and 5% for stage IV disease (4–6). The two main types of lung cancer, non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC) differ in their biological behaviors and therapeutic responses, with NSCLC further classified into subtypes such as squamous cell carcinoma (SCC) and adenocarcinoma (AC) (7, 8).

Early detection of lung cancer is critical for improving survival outcomes, yet most cases are diagnosed at an advanced stage due to the lack of specific symptoms and the limitations of conventional diagnostic methods, such as imaging and invasive biopsies (9–12). In this context, minimally invasive diagnostic tools with higher sensitivity and specificity are urgently needed. Recent advances in the understanding of epigenetics have highlighted the role of DNA methylation as a promising biomarker for cancer detection (13, 14). DNA methylation plays a pivotal role in tumorigenesis, with cancer cells exhibiting either global hypomethylation or specific hypermethylation on promoter regions of tumor suppressor genes (15–17). Promoter hypermethylation is particularly relevant as it frequently silences genes critical for cell differentiation and fate, making it a promising target for biomarker development (18, 19).

To provide background on the development of DNA methylation biomarkers for lung cancer, we summarized 42 previously reported genes, including RASSF1A, SHOX2, CDKN2A, HOXA1, and others (Supplementary Material 1), which exhibit aberrant promoter methylation detectable in non-invasive samples such as blood, sputum, and pleural effusion, enabling non-invasive diagnostic approaches (20). Among the most studied methylation biomarkers, RASSF1A and SHOX2 genes have shown significant diagnostic potential (21). Hypermethylation of RASSF1A leads to silencing of its tumor-suppressive function, while SHOX2 methylation has been strongly associated with malignancy (22, 23).

Despite promising findings, the diagnostic efficacy of RASSF1A and SHOX2 methylation remains inconclusive, with reported sensitivity and specificity varying across studies. Therefore, a systematic synthesis of available evidence is needed to evaluate their diagnostic performance comprehensively. This study aims to conduct a systematic review and meta-analysis to assess the diagnostic accuracy of RASSF1A and SHOX2 methylation in lung cancer, focusing on key metrics such as sensitivity, specificity, and diagnostic odds ratio. By analyzing data from 47 studies, this meta-analysis seeks to clarify their utility as biomarkers and guide future research and clinical implementation.

2 Materials and methods

2.1 Search strategy

To identify relevant studies investigating the diagnostic efficacy of SHOX2, RASSF1A, and the combination of SHOX2 and RASSF1A methylation in lung cancer, a comprehensive and systematic search was conducted. The databases searched included the Cochrane Library, PubMed, Web of Science, and Embase, covering publications up to July 2026. Medical Subject Headings (MeSH) terms and free-text keywords related to lung cancer and DNA methylation were utilized, including terms such as “lung neoplasms,” “DNA methylation,” “SHOX2,” and “RASSF1A.” The search strategy was adapted to each database to ensure optimal coverage, focusing on peer-reviewed studies published in English.

2.2 Selection criteria

The study selection process adhered to predefined inclusion and exclusion criteria. Human studies employing case-control, cohort, or cross-sectional designs that evaluated the diagnostic performance of SHOX2 methylation, RASSF1A methylation, or their combined methylation panel for lung cancer were included. Eligible studies were required to include both patients with lung cancer and an appropriate control group, provide sufficient data to construct a 2 × 2 contingency table, and be published as original full-text articles.

Studies were excluded if they were conducted on animal models or in vitro systems, as these differ significantly in biology and pathology from human lung cancer. Reviews, conference abstracts, editorials, case reports, letters, comments, and non-English publications were also excluded. Studies involving duplicate or overlapping patient populations or those that did not evaluate the diagnostic performance of the biomarkers of interest were also excluded. The study selection process followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, and the detailed inclusion and exclusion process was summarized using a PRISMA flow diagram.

2.3 Data extraction

Eligibility of the identified studies was independently assessed by two reviewers based on the predefined inclusion and exclusion criteria. Discrepancies between the reviewers were resolved through discussion or consultation with a third reviewer. Data extraction was performed independently by three reviewers (T.W., JY.W., and H.X.) using a standardized data collection form. Extracted information included study characteristics, participant demographics, methodological details, and key outcome measures such as sensitivity, specificity, and other diagnostic performance metrics.

For each independent diagnostic dataset, the numbers of true positives (TP), false positives (FP), false negatives (FN), and true negatives (TN) were extracted directly from the original publication whenever available. When these values were not explicitly reported, they were derived from the reported sample sizes and diagnostic accuracy measures only when the corresponding 2 × 2 contingency table could be reconstructed uniquely. If the reported information permitted more than one possible combination of TP, FP, FN, and TN, the data were considered non-reconstructable and were not included in the quantitative synthesis.

Independent cohorts reported within the same publication were treated as separate diagnostic datasets. However, data derived from the same patient population were included only once within each biomarker analysis to avoid double-counting and violation of statistical independence. When multiple diagnostic results based on different sample types or assay methods were reported for the same patient population, only one dataset was retained according to a predefined sample-type hierarchy: tissue > bronchoalveolar lavage fluid > pleural effusion > bronchial aspirate > plasma > serum. This hierarchy prioritized specimens considered to provide a more direct representation of tumor-derived methylation signals.

2.4 Quality assessment

The methodological quality and risk of bias of the included studies were independently assessed by two reviewers using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool. QUADAS-2 evaluates four key domains: patient selection, index test, reference standard, and flow and timing. Each domain was assessed for risk of bias, while the first three domains were additionally evaluated for concerns regarding applicability. Any discrepancies between the reviewers were resolved through discussion or consultation with a third reviewer. The results of the quality assessment were summarized using standard QUADAS-2 risk-of-bias and applicability judgments to evaluate the overall methodological quality of the included studies.

2.5 Meta-analysis

Diagnostic test accuracy meta-analysis was performed to evaluate the diagnostic performance of SHOX2 methylation, RASSF1A methylation, and the combined SHOX2/RASSF1A methylation panel for distinguishing patients with lung cancer from controls.

Pooled diagnostic accuracy was estimated using a bivariate random-effects model (Reitsma model), which jointly models sensitivity and specificity while accounting for their within-study correlation and between-study heterogeneity. Based on the fitted bivariate model, hierarchical summary receiver operating characteristic (HSROC) curves were generated to summarize the overall diagnostic performance of each biomarker panel. Summary estimates of sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR), and the area under the HSROC curve (AUC) with corresponding 95% confidence intervals (CIs) were calculated.

Potential threshold effects were evaluated by calculating the Spearman correlation coefficient between the logit-transformed sensitivity and the logit-transformed false-positive rate (1 − specificity). A significant positive correlation was considered indicative of a threshold effect. Between-study heterogeneity was explored descriptively using Cochran’s Q and I² statistics for sensitivity, specificity, PLR, NLR, and DOR to facilitate comparison with previous diagnostic meta-analyses, while the primary pooled estimates were obtained from the bivariate random-effects model. Prespecified subgroup analyses were conducted according to sample type, assay method, sample size, ethnicity, and pathological subtype to explore potential sources of heterogeneity. Publication bias was assessed using Deeks’ funnel plot asymmetry test, which is recommended for diagnostic accuracy meta-analyses. Leave-one-out sensitivity analyses were performed to evaluate the robustness of the pooled estimates.

2.6 Statistical analysis

All statistical analyses were performed using R (version 4.6.0; R Foundation for Statistical Computing, Vienna, Austria). Diagnostic test accuracy meta-analyses were conducted using the mada package. Pairwise differences in pooled sensitivity and specificity among the combined SHOX2/RASSF1A panel, SHOX2 alone, and RASSF1A alone were evaluated using Wald tests for the corresponding biomarker-group coefficients in the joint bivariate meta-regression model. All statistical tests were two-sided. A P value <0.05 was considered statistically significant.

3 Results

3.1 Publication search and included studies

The study selection process for this meta-analysis is summarized in a PRISMA flow chart (Figure 1). A total of 1117 studies were retrieved from PubMed, Web of Science, Cochrane Library, and Embase. After removing 377 duplicate records, 740 studies were screened. Of these, 452 were excluded based on titles and abstracts, leaving 288 studies for full-text review. Ultimately, 239 studies were excluded for the following reasons: 63 were of incorrect study types, 64 were not conducted on humans, 29 were conference abstracts, 23 lacked sufficient data for outcome calculation, 8 full texts articles were unavailable, and 52 were review articles. Forty-nine studies met the inclusion criteria and were included in this meta-analysis (24–45), (46–72). The QUADAS-2 assessment showed that most studies had a low risk of bias in the flow and timing domain, whereas the patient selection and reference standard domains were predominantly rated as unclear. The index test domain showed the highest proportion of high-risk assessments. Concerns regarding applicability were generally low across all domains. Detailed QUADAS-2 assessments are presented in Supplementary Material 2.

FIGURE 1

3.2 Characteristics of included studies

The detailed characteristics of the 49 included studies are presented in Table 1. These studies, published between 2004 and 2025, included various study designs, participant demographics, sample types, and methodologies. Most studies (45/49) used a case-control design, while four employed a cohort design. The number of cases ranged from 9 to 585, and the number of controls ranged from 9 to 445. Sample types included FFPE tissues, bronchoalveolar lavage fluid (BALF), pleural effusion, bronchial epithelial cells, blood, bronchial aspirates, lung tissues, bronchial washings, and sputum. Because some studies evaluated multiple specimen types, the summed frequencies across specimen types exceeded the total number of included studies.

3.3 Meta-analysis results of SHOX2 and RASSF1A methylation

The combined diagnostic performance of SHOX2 and RASSF1A methylation for lung cancer is shown in Figure 2. Based on 18 studies (53–70), the pooled sensitivity was 0.778 (0.723–0.825), the pooled specificity was 0.890 (0.866–0.911), the PLR was 7.102 (5.565–9.063), the NLR was 0.249 (0.196–0.318), the DOR was 28.483 (18.095–44.835), and the area under the HSROC curve was 0.916 (0.872–0.924). Forest plots of pooled sensitivity and specificity are presented in Figures 2A,B, respectively, while the HSROC curve is shown in Figure 2C. Spearman correlation analysis demonstrated no significant threshold effect among the included studies (Spearman’s ρ = −0.201, P = 0.382; Figure 2D), suggesting that heterogeneity was unlikely to be attributable to differences in diagnostic thresholds.

FIGURE 2

3.4 Meta-analysis results of SHOX2 methylation

The diagnostic performance of SHOX2 methylation was evaluated in 21 studies (48–52, 54, 55, 57–59, 61–69, 71, 72). The pooled sensitivity was 0.694 (0.630–0.752), the pooled specificity was 0.917 (0.900–0.931), the PLR was 8.363 (6.584–10.623), the NLR was 0.334 (0.271–0.411), the DOR was 25.071 (16.468–38.166), and the area under the HSROC curve was 0.925 (0.890–0.930). The corresponding forest plots and HSROC curve are presented in Figures 3A–C. No significant threshold effect was identified (Spearman’s ρ = −0.144, P = 0.502; Figure 3D).

FIGURE 3

3.5 Meta-analysis results of RASSF1A methylation

The diagnostic performance of RASSF1A methylation was evaluated in 38 studies (24–47, 54, 55, 57–59, 61–69). The pooled sensitivity was 0.457 (0.416–0.498), the pooled specificity was 0.938 (0.918–0.953), the PLR was 7.338 (5.624–9.573), the NLR was 0.579 (0.540–0.622), the DOR was 12.666 (9.474–16.934), and the area under the HSROC curve was 0.789 (0.721–0.868). Figures 4A–C presents the corresponding forest plots and HSROC curve. No significant threshold effect was observed (Spearman’s ρ = 0.138, P = 0.388; Figure 4D).

FIGURE 4

3.6 Pairwise comparison of diagnostic performance

Compared with SHOX2 alone, the combined SHOX2/RASSF1A assay showed higher sensitivity (P = 0.019), whereas specificity did not differ significantly (P = 0.272). Compared with RASSF1A alone, the combined assay showed higher sensitivity (P < 0.001) and lower specificity (P = 0.015). SHOX2 also showed higher sensitivity than RASSF1A (P < 0.001), while their specificities did not differ significantly (P = 0.183) (Table 2).

3.7 Evaluation of heterogeneity and publication bias

Heterogeneity statistics are summarized in Supplementary Table S1. Significant heterogeneity was observed for pooled sensitivity in the combined SHOX2/RASSF1A, SHOX2, and RASSF1A analyses (all P < 0.001; I² = 87.29%, 91.96%, and 83.32%, respectively). In contrast, specificity showed low heterogeneity for the combined SHOX2/RASSF1A (I² = 26.98%, P = 0.063) and SHOX2 analyses (I² = 19.14%, P = 0.312), whereas moderate heterogeneity was observed for RASSF1A (I² = 66.69%, P < 0.001). Deeks’ funnel plot asymmetry test demonstrated significant publication bias for the combined SHOX2/RASSF1A, SHOX2, and RASSF1A analyses (all P < 0.001; Figures 5A–C).

FIGURE 5

3.8 Subgroup analysis

To further explore potential sources of between-study heterogeneity, prespecified subgroup analyses were performed according to ethnicity, sample type, sample size, assay method, and pathological subtype. For the combined SHOX2/RASSF1A analysis, no significant differences in pooled sensitivity or specificity were observed across sample types or sample size categories (all P > 0.05) (Table 3). However, pooled sensitivity differed significantly according to assay method (P = 0.005), with QMSP and RT-PCR showing higher sensitivities than MSP, whereas specificity did not differ significantly among assay methods (P = 0.730). Studies including multiple pathological subtypes also showed higher sensitivity than those restricted to a single subtype (P = 0.001), with a significant difference in specificity (P = 0.028). For SHOX2 (Table 4), pooled sensitivity varied significantly according to sample type (P = 0.001) and pathological subtype (P < 0.001), whereas no significant subgroup differences were observed for ethnicity, sample size, assay method, or specificity (all P > 0.05). For RASSF1A (Table 5), assay method significantly influenced pooled sensitivity (P = 0.043), while no significant subgroup differences were identified for ethnicity, sample type, sample size, pathological subtype, or specificity (all P > 0.05), although ethnicity showed a borderline association with sensitivity (P = 0.052).

3.9 Leave-one-out sensitivity analysis

Leave-one-out sensitivity analyses were performed by sequentially excluding each study to evaluate the robustness of the pooled diagnostic estimates. For combined SHOX2/RASSF1A methylation, pooled sensitivity ranged from 0.766 to 0.787 and pooled specificity from 0.887 to 0.896 (Supplementary Table S2). For SHOX2 methylation, pooled sensitivity ranged from 0.678 to 0.706 and pooled specificity from 0.912 to 0.921 (Supplementary Table S3). For RASSF1A methylation, pooled sensitivity ranged from 0.447 to 0.463 and pooled specificity from 0.936 to 0.940 (Supplementary Table S4). Overall, only minimal changes in the pooled estimates were observed after sequential omission of individual studies, indicating that no single study had a substantial influence on the overall diagnostic performance and confirming the robustness of the findings.

4 Discussion

This systematic review and meta-analysis comprehensively evaluated the diagnostic performance of SHOX2 methylation, RASSF1A methylation, and their combined panel for lung cancer detection based on 49 eligible studies. The combined SHOX2/RASSF1A methylation panel demonstrated a pooled sensitivity of 0.778 (0.723–0.825) and specificity of 0.890 (0.866–0.911), with an HSROC AUC of 0.916, indicating excellent overall diagnostic performance. In comparison, SHOX2 methylation alone achieved higher sensitivity than RASSF1A methylation (0.694 vs. 0.457), whereas RASSF1A demonstrated the highest specificity (0.938). Pairwise comparisons further showed that the combined panel significantly improved sensitivity compared with either SHOX2 or RASSF1A alone while maintaining specificity comparable to SHOX2, although lower than that of RASSF1A. These findings suggest that combining SHOX2 and RASSF1A provides a more balanced diagnostic strategy than either individual biomarker and supports its potential role as an adjunctive biomarker for lung cancer detection.

DNA methylation is a stable epigenetic alteration that occurs early in tumorigenesis and can be detected in blood, sputum, and other body fluids (12). These properties make it an attractive biomarker for cancer detection. Our findings highlight the critical role of DNA methylation biomarkers, particularly SHOX2 and RASSF1A, in lung cancer diagnosis. SHOX2 and RASSF1A are well-established tumor suppressor genes implicated in the pathogenesis of various cancers, including lung cancer. SHOX2, located at chromosome 3q25.32, is a key regulator of organ development during embryogenesis and is frequently hypermethylated in lung cancer, resulting in gene silencing (22). Similarly, RASSF1A is a tumor suppressor gene commonly hypermethylated in lung, kidney, and other solid tumors, contributing to tumor initiation and progression (23). The combined methylation panel demonstrated improved diagnostic accuracy, highlighting the complementary diagnostic potential of SHOX2 and RASSF1A. The clinical implications of our findings are significant. SHOX2 and RASSF1A methylation has the potential to serve as a non-invasive diagnostic tool for lung cancer, particularly in patients where invasive biopsy procedures are not feasible or yield insufficient material. Moreover, its balanced sensitivity and specificity suggest that it can minimize false-positive diagnoses, reducing unnecessary follow-up procedures and anxiety for patients.

Our results align with earlier meta-analyses that reported the diagnostic value of SHOX2 methylation in lung cancer detection. Zhao et al. demonstrated a sensitivity of 70% and specificity of 96% for SHOX2 methylation, which is comparable to our findings. However, our study expands on these findings by incorporating RASSF1A methylation and exploring the combined diagnostic efficacy of both biomarkers. Interestingly, while SHOX2 demonstrated higher sensitivity, RASSF1A displayed superior specificity. The combination of these biomarkers achieved an optimal balance between sensitivity and specificity, making it more suitable for clinical applications where both parameters are critical. Unlike previous meta-analyses that evaluated individual methylation markers separately, the present study directly compared pooled sensitivity and specificity among combined SHOX2/RASSF1A, SHOX2 alone, and RASSF1A alone using a joint bivariate meta-regression framework.

The combined analysis of SHOX2 and RASSF1A methylation represents a promising molecular diagnostic approach for lung cancer and may have potential utility in the early detection setting. However, because most included studies evaluated patients across mixed disease stages and did not report stage-specific diagnostic accuracy, the present meta-analysis could not assess performance specifically in early-stage disease. Its non-invasive nature nevertheless supports further evaluation in high-risk populations, such as smokers and individuals with a family history of lung cancer, particularly as an adjunct to existing screening strategies. Additionally, the ability to detect methylation in various sample types, including blood and sputum, enhances its feasibility for widespread clinical implementation. Future efforts should focus on standardizing methylation detection methods to ensure reproducibility and reliability across laboratories. Moreover, integrating methylation biomarkers with imaging modalities or other molecular markers could further improve diagnostic accuracy and guide personalized treatment strategies.

Although numerous DNA methylation biomarkers have been reported for lung cancer diagnosis, only SHOX2, RASSF1A, and their combined panel were supported by a sufficient number of independent studies with extractable diagnostic accuracy data to permit robust quantitative synthesis. Therefore, the present meta-analysis focused on these biomarkers. Despite prespecified subgroup analyses, substantial residual heterogeneity remained. One potential contributor to the residual heterogeneity is the variability in control populations across studies, which ranged from healthy individuals to patients with benign pulmonary diseases, atypical adenomatous hyperplasia, benign pleural effusion, and other non-malignant conditions. Studies using only healthy controls may overestimate diagnostic specificity compared with those including patients with benign pulmonary diseases or other malignancies, which more closely resemble real-world clinical diagnostic settings. Because many studies used mixed control populations or reported insufficient clinical details, subgroup analyses according to control type were not feasible. In addition, variations in methylation assays, laboratory protocols, and positivity definitions across studies may also have contributed to residual heterogeneity, although no significant threshold effect was detected. Together, these factors should be considered when interpreting the pooled diagnostic performance of SHOX2 and RASSF1A methylation. Future studies should adopt standardized methylation assays and prespecified positivity thresholds to improve comparability across studies and reduce between-study heterogeneity.

Despite these encouraging findings, several limitations should be acknowledged. First, most included studies were retrospective case-control studies, which may have overestimated diagnostic accuracy because of spectrum and selection bias. Second, substantial between-study heterogeneity remained despite subgroup analyses, likely reflecting differences in specimen type, assay methodology, and study populations. Third, significant publication bias was detected by Deeks’ funnel plot asymmetry test, suggesting that the pooled diagnostic performance may have been overestimated and should therefore be interpreted with appropriate caution. Finally, because most included studies enrolled patients across mixed disease stages and histological subtypes, stage-specific and histology-specific diagnostic performance could not be reliably evaluated. Consequently, the present findings reflect the overall diagnostic performance of these biomarkers across heterogeneous lung cancer populations rather than their accuracy in specific clinical settings. Future large-scale prospective studies using standardized methylation assays and uniform reporting of stage- and histology-specific diagnostic outcomes are warranted. In addition, cost-effectiveness analyses will be important before these biomarkers can be routinely implemented in clinical practice.

5 Conclusion

In summary, compared with either biomarker alone, the combined SHOX2/RASSF1A methylation panel significantly improved pooled sensitivity while maintaining high diagnostic specificity, supporting a more balanced overall diagnostic performance for lung cancer detection. These findings support its potential role as an adjunctive diagnostic approach, particularly for non-tissue or minimally invasive specimens. Nevertheless, large-scale prospective studies using standardized assay protocols are warranted before routine clinical implementation. Further studies should also evaluate stage-specific diagnostic performance, particularly in patients with stage I/II lung cancer, to better define the role of SHOX2/RASSF1A methylation in early detection.

References

• 1.

  BrayFLaversanneMSungHFerlayJSiegelRLSoerjomataramIet alGlobal cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: A Cancer Journal Clinicians (2024) 74(3):229–63. 10.3322/caac.21834

  • CrossRef
  • Google Scholar

• 2.

  TheL. Lung cancer treatment: 20 years of progress. Lancet (2024) 403(10445):2663. 10.1016/S0140-6736(24)01299-6

  • CrossRef
  • Google Scholar

• 3.

  MeyerMLPetersSMokTLamSYangPCAggarwalCet alLung cancer research and treatment: global perspectives and strategic calls to action. Ann Oncol (2024) 35(12):1088–104. 10.1016/j.annonc.2024.10.006

  • CrossRef
  • Google Scholar

• 4.

  HendriksLELRemonJFaivre-FinnCGarassinoMCHeymachJVKerrKMet alNon-small-cell lung cancer. Nat Rev Dis Primers (2024) 10(1):71. 10.1038/s41572-024-00551-9

  • CrossRef
  • Google Scholar

• 5.

  ChengLAlexanderREMaclennanGTCummingsOWMontironiRLopez-BeltranAet alMolecular pathology of lung cancer: key to personalized medicine. Mod Pathol (2012) 25(3):347–69. 10.1038/modpathol.2011.215

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 6.

  ChengLZhangSAlexanderRYaoYMacLennanGTPanCx.et alThe landscape of EGFR pathways and personalized management of non-small-cell lung cancer. Future Oncol (2011) 7(4):519–41. 10.2217/fon.11.25

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 7.

  LeiterAVeluswamyRRWisniveskyJP. The global burden of lung cancer: current status and future trends. Nat Rev Clin Oncol (2023) 20(9):624–39. 10.1038/s41571-023-00798-3

  • CrossRef
  • Google Scholar

• 8.

  ThaiAASolomonBJSequistLVGainorJFHeistRS. Lung cancer. Lancet (2021) 398(10299):535–54. 10.1016/s0140-6736(21)00312-3

  • CrossRef
  • Google Scholar

• 9.

  MazzonePJBachPBCareyJSchonewolfCABognarKAhluwaliaMSet alClinical validation of a cell-free DNA fragmentome assay for augmentation of lung cancer early detection. Cancer Discov (2024) 14(11):2224–42. 10.1158/2159-8290.CD-24-0519

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 10.

  TammemägiMCDarlingGESchmidtHWalkerMJLangerDLeungYWet alRisk-based lung cancer screening performance in a universal healthcare setting. Nat Med (2024) 30(4):1054–64. 10.1038/s41591-024-02904-z

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 11.

  WangCShaoJHeYWuJLiuXYangLet alData-driven risk stratification and precision management of pulmonary nodules detected on chest computed tomography. Nat Med (2024) 30(11):3184–95. 10.1038/s41591-024-03211-3

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 12.

  DavidsonDDChengL. Perspectives of lung cancer control and molecular prevention. Future Oncol (2019) 15(31):3527–30. 10.2217/fon-2019-0523

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 13.

  SchreiberhuberLBarrettJEWangJRedlEHerzogCVavourakisCDet alCervical cancer screening using DNA methylation triage in a real-world population. Nat Med (2024) 30(8):2251–7. 10.1038/s41591-024-03014-6

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 14.

  EzegboguMWilkinsonEReidGRodgerEJBrockwayBRussell-CampTet alCell-free DNA methylation in the clinical management of lung cancer. Trends Mol Med (2024) 30(5):499–515. 10.1016/j.molmed.2024.03.007

  • CrossRef
  • Google Scholar

• 15.

  CountsJLGoodmanJI. Alterations in DNA methylation may play a variety of roles in carcinogenesis. Cell (1995) 83(1):13–5. 10.1016/0092-8674(95)90228-7

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 16.

  LiPLiuSDuLMohseniGZhangYWangC. Liquid biopsies based on DNA methylation as biomarkers for the detection and prognosis of lung cancer. Clin Epigenetics (2022) 14(1):118. 10.1186/s13148-022-01337-0

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 17.

  SmithZDHetzelSMeissnerA. DNA methylation in mammalian development and disease. Nat Rev Genet (2025) 26(1):7–30. 10.1038/s41576-024-00760-8

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 18.

  AgarwalNJhaAK. DNA hypermethylation of tumor suppressor genes among oral squamous cell carcinoma patients: a prominent diagnostic biomarker. Mol Biol Rep (2025) 52(1):1–13. 10.1007/s11033-024-10144-0

  • CrossRef
  • Google Scholar

• 19.

  RamaziSDadzadiMSahafnejadZAllahverdiA. Epigenetic regulation in lung cancer. MedComm (2023) 4(6):e401. 10.1002/mco2.401

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 20.

  HulbertAJusue-TorresIStarkAChenCRodgersKLeeBet alEarly detection of lung cancer using DNA promoter hypermethylation in plasma and sputum. Clin Cancer Res (2017) 23(8):1998–2005. 10.1158/1078-0432.CCR-16-1371

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 21.

  WeiBWuFXingWSunHYanCZhaoCet alA panel of DNA methylation biomarkers for detection and improving diagnostic efficiency of lung cancer. Sci Rep (2021) 11(1):16782. 10.1038/s41598-021-96242-6

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 22.

  SchneiderKUDietrichDFleischhackerMLeschberGMerkJSchäperFet alCorrelation of SHOX2 gene amplification and DNA methylation in lung cancer tumors. BMC Cancer (2011) 11:102. 10.1186/1471-2407-11-102

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 23.

  DammannRSchagdarsurenginUStrunnikovaMRastetterMSeidelCLiuLet alEpigenetic inactivation of the Ras-association domain family 1 (RASSF1A) gene and its function in human carcinogenesis. Histol Histopathol (2003) 18(2):665–77. 10.14670/HH-18.665

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 24.

  YuZ-HWangYCChenLBSongYLiuCXiaXYet alAnalysis of RASSF1A promoter hypermethylation in serum DNA of non-small cell lung cancer. Zhonghua Zhong Liu Za Zhi [Chinese Journal Oncology] (2008) 30(4):284–7. Available online at: https://europepmc.org/article/med/18788633.

  • Pubmed Abstract
  • Google Scholar

• 25.

  WangYYuZWangTZhangJHongLChenL. Identification of epigenetic aberrant promoter methylation of RASSF1A in serum DNA and its clinicopathological significance in lung cancer. Lung Cancer (2007) 56(2):289–94. 10.1016/j.lungcan.2006.12.007

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 26.

  SmetannikovaNAEvdokimovAANetesovaNAAbdurashitovMAAkishevAGDubininEVet alApplication of GLAD-PCR assay for study on DNA methylation in regulatory regions of some tumor-suppressor genes in lung cancer. Zhongguo Fei Ai Za Zhi = Chin Journal Lung Cancer (2019) 22(9):551–61. 10.3779/j.issn.1009-3419.2019.09.01

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 27.

  ShivapurkarNStastnyVSuzukiMWistubaIILiLZhengYet alApplication of a methylation gene panel by quantitative PCR for lung cancers. Cancer Lett (2007) 247(1):56–71. 10.1016/j.canlet.2006.03.020

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 28.

  ShahNUDAliMNGanaiBAMudassarSKhanMSKourJet alAssociation of promoter methylation of RASSF1A and KRAS mutations in non-small cell lung carcinoma in Kashmiri population (India). Heliyon (2020) 6 (2). 10.1016/j.heliyon.2020.e03488

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 29.

  SchmiemannVBöckingAKazimirekMOnofreASCGabbertHEKappesRet alMethylation assay for the diagnosis of lung cancer on bronchial aspirates:: a cohort study. Clin Cancer Res (2005) 11(21):7728–34. 10.1158/1078-0432.CCR-05-0999

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 30.

  RykovaEYSkvortsovaTELaktionovPPTamkovichSNBryzgunovaOEStarikovAVet alInvestigation of tumor-derived extracellular DNA in blood of cancer patients by methylation-specific PCR. Nucleosides Nucleotides and Nucleic Acids (2004) 23(6-7):855–9. 10.1081/NCN-200026031

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 31.

  RoncaratiRLupiniLMiottoESaccentiEMascettiSMorandiLet alMolecular testing on bronchial washings for the diagnosis and predictive assessment of lung cancer. Mol Oncol (2020) 14(9):2163–75. 10.1002/1878-0261.12713

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 32.

  PonomaryovaAARykovaEYCherdyntsevaNVSkvortsovaTEDobrodeevAYZav'yalovAAet alPotentialities of aberrantly methylated circulating DNA for diagnostics and post-treatment follow-up of lung cancer patients. Lung Cancer (2013) 81(3):397–403. 10.1016/j.lungcan.2013.05.016

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 33.

  PengZShanCWangH. Value of promoter methylation of RASSF1A, p16, and DAPK genes in induced sputum in diagnosing lung cancers. Zhong Nan da Xue Xue Bao Yi Xue Ban = J Cent South Univ Med Sci (2010) 35(3):247–53. 10.3969/j.issn.1672-7347.2010.03.010

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 34.

  Pastuszak-LewandoskaDKordiakJMigdalska-SękMCzarneckaKHAntczakAGórskiPet alQuantitative analysis of mRNA expression levels and DNA methylation profiles of three neighboring genes: FUS1, NPRL2/G21 and RASSF1A in non-small cell lung cancer patients. Respir Res (2015) 16(1):76. 10.1186/s12931-015-0230-6

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 35.

  NunesSPDinizFMoreira-BarbosaCConstâncioVSilvaAVOliveiraJet alSubtyping lung cancer using DNA methylation in liquid biopsies. J Clin Med (2019) 8(9). 10.3390/jcm8091500

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 36.

  NawazIQiuXWuHLiYFanYHuLFet alDevelopment of a multiplex methylation specific PCR suitable for (early) detection of non-small cell lung cancer. Epigenetics (2014) 9(8):1138–48. 10.4161/epi.29499

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 37.

  MohammedFBaydaa Abed HusseinAAhmedT. Evaluation of methylation Panel in the promoter region of p16INK4a, RASSF1A, and MGMT as a biomarker in sputum for lung cancer. Arch Razi Inst (2022) 77(3):1075–81. 10.22092/ARI.2022.357985.2131

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 38.

  MaYBaiYMaoHHongQYangDZhangHet alA panel of promoter methylation markers for invasive and noninvasive early detection of NSCLC using a quantum dots-based FRET approach. Biosens and Bioelectron (2016) 85:641–8. 10.1016/j.bios.2016.05.067

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 39.

  LiuG-z.WuY-m.YangJ-y. Significance of combined detection of plasma RASSF1A and p16 gene methylation in diagnosis of non-small cell lung cancers. Zhonghua Zhong Liu Za Zhi [Chinese Journal Oncology] (2007) 29(8):613–4. Available online at: https://rs.yiigle.com/cmaid/36312.

  • Pubmed Abstract
  • Google Scholar

• 40.

  KimHKwonYMKimJSLeeHParkJHShimYMet alTumor-specific methylation in bronchial lavage for the early detection of non-small-cell lung cancer. J Clin Oncol (2004) 22(12):2363–70. 10.1200/JCO.2004.10.077

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 41.

  HubersAJvan der DriftMAPrinsenCFMWitteBIWangYShivapurkarNet alMethylation analysis in spontaneous sputum for lung cancer diagnosis. Lung Cancer (2014) 84(2):127–33. 10.1016/j.lungcan.2014.01.019

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 42.

  HubersAJHeidemanDAMHerderGJMBurgersSASterkPJKunstPWet alProlonged sampling of spontaneous sputum improves sensitivity of hypermethylation analysis for lung cancer. J Clin Pathol (2012) 65(6):541–5. 10.1136/jclinpath-2012-200712

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 43.

  HubersAJHeidemanDAMDuinSWitteBIde KoningHJGroenHJMet alDNA hypermethylation analysis in sputum of asymptomatic subjects at risk for lung cancer participating in the NELSON trial: argument for maximum screening interval of 2years. J Clin Pathol (2017) 70(3):250–4. 10.1136/jclinpath-2016-203734

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 44.

  HubersAJHeidemanDAMBurgersSAHerderGJMSterkPJRhodiusRJet alDNA hypermethylation analysis in sputum for the diagnosis of lung cancer: training validation set approach. Br J Cancer (2015) 112(6):1105–13. 10.1038/bjc.2014.636

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 45.

  GroteHJSchmiemannVGeddertHBockingAKappesRGabbertHEet alMethylation of RAS association domain family protein 1A as a biomarker of lung cancer. Cancer Cytopathology (2006) 108(2):129–34. 10.1002/cncr.21717

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 46.

  GaoLXieEYuTChenDZhangLZhangBet alMethylated APC and RASSF1A in multiple specimens contribute to the differential diagnosis of patients with undetermined solitary pulmonary nodules. J Thorac Dis (2015) 7(3):422–32. 10.3978/j.issn.2072-1439.2015.01.24

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 47.

  ConstâncioVNunesSPMoreira-BarbosaCFreitasROliveiraJPousaIet alEarly detection of the major male cancer types in blood-based liquid biopsies using a DNA methylation panel. Clin Epigenetics (2019) 11(1):175. 10.1186/s13148-019-0779-x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 48.

  XuZWangYWangLXiongJWangHCuiFet alThe performance of the SHOX2/PTGER4 methylation assay is influenced by cancer stage, age, type and differentiation. Biomark Med (2020) 14(5):341–51. 10.2217/bmm-2019-0325

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 49.

  Vo TTLVNguyenTNNguyenTTDuong PhamATVuongDLTaVTet alSHOX2 methylation in Vietnamese patients with lung cancer. Mol Biol Rep (2022) 49(5):3413–21. 10.1007/s11033-022-07172-z

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 50.

  SchmidtBLiebenbergVDietrichDSchlegelTKneipCSeegebarthAet alSHOX2 DNA methylation is a biomarker for the diagnosis of lung cancer based on bronchial aspirates. Bmc Cancer (2010) 10:600. 10.1186/1471-2407-10-600

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 51.

  KneipCSchmidtBSeegebarthAWeickmannSFleischhackerMLiebenbergVet alSHOX2 DNA methylation is a biomarker for the diagnosis of lung cancer in plasma. J Thorac Oncol (2011) 6(10):1632–8. 10.1097/JTO.0b013e318220ef9a

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 52.

  HuangWHuangHZhangSWangXOuyangJLinZet alA novel diagnosis method based on methylation analysis of SHOX2 and serum biomarker for early stage lung cancer. Cancer Control (2020) 27(1):1073274820969703. 10.1177/1073274820969703

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 53.

  ZengSHeXTanYHeZHuangY. Clinical value of SHOX2 and RASSF1A methylation combination in BALF samples for the diagnosis of lung cancer. J Environ Pathol Toxicol Oncol (2026) 45(2):15–29. 10.1615/jenvironpatholtoxicoloncol.2025060381

  • CrossRef
  • Google Scholar

• 54.

  GuPWangYShiCChenYQianJNiuYet alEnhancing the precision of auxiliary diagnosis for lung cancer through use of SHOX2 and RASSF1A methylation status in lung biopsy and lymph node biopsy specimens. Translational Lung Cancer Res (2025) 14(3):897–911. 10.21037/tlcr-2024-1082

  • CrossRef
  • Google Scholar

• 55.

  ChenYZhangXLiPChenZLiuZChenSet alDetection of SHOX2 and RASSF1A methylation for early-stage lung adenocarcinoma. Mol Clin Oncol (2025) 23(6):1–6. 10.3892/mco.2025.2907

  • CrossRef
  • Google Scholar

• 56.

  ZhongQWangYLiangCWeiFSheB. Combined methylation of SHOX2 and RASSF1A genes in diagnosing malignant pleural effusion. Discov Med (2023) 35(178):845–52. 10.24976/discov.med.202335178.79

  • CrossRef
  • Google Scholar

• 57.

  ZhaoJLuYRenXBianTFengJSunHet alAssociation of the SHOX2 and RASSF1A methylation levels with the pathological evolution of early-stage lung adenocarcinoma. BMC Cancer (2024) 24(1):687. 10.1186/s12885-024-12452-x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 58.

  ZhangNLiuZLiKXingXLongCLiuFet alDNA methylation analysis of the SHOX2 and RASSF1A panel using cell-free DNA in the diagnosis of malignant pleural effusion. J Oncol (2023) 1(2023):5888844. 10.1155/2023/5888844

  • CrossRef
  • Google Scholar

• 59.

  ZhangJHuangHYuFBianYWangRLiuHet alA comprehensive diagnostic scheme of morphological combined molecular methylation under bronchoscopy. Front Oncol (2023) 13:1133675. 10.3389/fonc.2023.1133675

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 60.

  ZhangCYuWWangLZhaoMGuoQLvSet alDNA methylation analysis of the SHOX2 and RASSF1A panel in bronchoalveolar lavage fluid for lung cancer diagnosis. J Cancer (2017) 8(17):3585–91. 10.7150/jca.21368

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 61.

  XieBDongWHeFPengFZhangHWangWet alThe combination of SHOX2 and RASSF1A DNA methylation had a diagnostic value in pulmonary nodules and early lung cancer. Oncology (2024) 102(9):759–774. 10.1159/000534275

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 62.

  ShiJChenXZhangLFangXLiuYZhuXet alPerformance evaluation of SHOX2 and RASSF1A methylation for the aid in diagnosis of lung cancer based on the analysis of FFPE specimen. Front Oncol (2020) 10:565780. 10.3389/fonc.2020.565780

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 63.

  RenMWangCShengDShiYJinMXuS. Methylation analysis of SHOX2 and RASSF1A in bronchoalveolar lavage fluid for early lung cancer diagnosis. Ann Diagn Pathol (2017) 27:57–61. 10.1016/j.anndiagpath.2017.01.007

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 64.

  LuHLinD. Diagnostic value of exfoliated tumor cells combined with DNA methylation in bronchoalveolar lavage fluid for lung cancer. Medicine (2023) 102(36):e34955. 10.1097/md.0000000000034955

  • CrossRef
  • Google Scholar

• 65.

  LiuJBianTSheBLiuLSunHZhangQet alEvaluating the comprehensive diagnosis efficiency of lung cancer, including measurement of SHOX2 and RASSF1A gene methylation. Bmc Cancer (2024) 24(1):282. 10.1186/s12885-024-12022-1

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 66.

  LiangCLiuNZhangQDengMMaJLuJet alA detection panel of novel methylated DNA markers for malignant pleural effusion. Front Oncol (2022) 12:967079. 10.3389/fonc.2022.967079

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 67.

  JinYLuRLiuFJiangGWangRZhengM. DNA methylation analysis in plasma for early diagnosis in lung adenocarcinoma. Medicine (2024) 103(28):e38867. 10.1097/MD.0000000000038867

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 68.

  JiX-YLiHChenHHLinJ. Diagnostic performance of RASSF1A and SHOX2 methylation combined with EGFR mutations for differentiation between small pulmonary nodules. J Cancer Res Clin Oncol (2023) 149(11):8557–71. 10.1007/s00432-023-04745-8

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 69.

  GaoHYangJHeLWangWLiuYHuYet alThe diagnostic potential of SHOX2 and RASSF1A DNA methylation in early lung adenocarcinoma. Front Oncol (2022) 12:849024. 10.3389/fonc.2022.849024

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 70.

  ChenSHuangKZouLChenLHuP. Diagnostic value of SHOX2, RASSF1A gene methylation combined with CEA level detection in malignant pleural effusion. Bmc Pulm Med (2023) 23(1):160. 10.1186/s12890-023-02462-z

  • CrossRef
  • Google Scholar

• 71.

  FengHShaoWDuLQingXZhangZLiangCet alDetection of SHOX2 DNA methylation by methylation-specific PCR in non-small cell lung cancer. Translational Cancer Res (2020) 9(10):6070–7. 10.21037/tcr-20-887

  • CrossRef
  • Google Scholar

• 72.

  DietrichDKneipCRajiOLiloglouTSeegebarthASchlegelTet alPerformance evaluation of the DNA methylation biomarker SHOX2 for the aid in diagnosis of lung cancer based on the analysis of bronchial aspirates. Int J Oncol (2012) 40(3):825–32. 10.3892/ijo.2011.1264

  • Pubmed Abstract
  • CrossRef
  • Google Scholar