The technical backbone of the study was a targeted NGS workflow built on QIAGEN GeneReader panels, run on formalin-fixed, paraffin-embedded (FFPE) tissue as well as plasma samples. This is the material that pathology laboratories already produce every day, which makes the approach directly translatable to real-world practice rather than confined to specialized research cohorts. The team evaluated three classes of genomic alterations: single nucleotide variants (SNVs), which are single-letter changes in the DNA code; insertions and deletions (indels), which add or remove letters; and copy number variations (CNVs), which duplicate or delete larger stretches of the genome. Associations between these alterations and clinicopathological features such as sex, smoking status, and histological subtype were then analyzed statistically, providing a layered picture of how genomic profiles vary across patient groups.

The headline result concerns the two most notorious oncogenes in lung cancer. Alterations in EGFR, the epidermal growth factor receptor, were detected in 23.7 percent of patients, while KRAS mutations appeared in 22.5 percent. These two genes, which sit at the top of critical cell-signaling pathways controlling growth and division, together accounted for nearly half of the actionable genomic landscape in this cohort. Behind them, the team identified recurrent alterations in MET, STK11, PIK3CA, and FGFR1, each of which carries its own therapeutic implications. MET amplification and mutation can sensitize tumors to MET inhibitors, STK11 loss is increasingly recognized as a marker of immunotherapy resistance, PIK3CA alterations point to PI3K-pathway dependencies, and FGFR1 amplification has been explored as a target in squamous tumors.

One of the most clinically significant technical details in the study involves EGFR amplification, which was detected in 12.1 percent of patients and frequently co-occurred with EGFR mutations. This matters because EGFR copy number gain on top of an activating mutation may influence both tumor biology and treatment response, particularly in the context of tyrosine kinase inhibitors (TKIs), the targeted drugs that block EGFR signaling. The predominant EGFR mutations were exon 19 deletions and L858R substitutions, the two classic sensitizing alterations that predict strong responses to first- and later-generation EGFR TKIs. On the KRAS side, the G12C variant was the most common, and this is a particularly notable finding because KRAS G12C has recently become directly druggable with a new class of covalent inhibitors, transforming what was once considered an undruggable mutation into one of the hottest targets in thoracic oncology.

The study also revealed clear sex-based differences in the genomic landscape. EGFR alterations and ERBB2 (HER2) mutations were more common in female patients, whereas KRAS, FGFR1, and MET alterations were more frequent in male patients. These patterns align with well-established epidemiological trends, in which EGFR-mutant lung adenocarcinoma is more prevalent in women and in lighter or never-smokers, while KRAS-mutant tumors correlate more strongly with smoking exposure. But the fact that these differences hold up in a large Turkish cohort adds population-specific weight to the observation and reinforces the argument that molecular testing should be applied broadly rather than restricted by assumptions about who is likely to harbor which driver.

Perhaps the most scientifically intriguing aspect of the work is its characterization of co-mutation patterns, the combinations of alterations that appear together within the same tumor. EGFR and KRAS alterations demonstrated a largely mutually exclusive pattern, meaning tumors almost never carried activating mutations in both genes at once. This is consistent with the biology of these pathways, since constitutive activation of one driver typically removes the selective pressure for activating the other. In contrast, KRAS commonly co-occurred with STK11 and PIK3CA alterations, a combination that has been linked in prior studies to more aggressive disease biology and reduced benefit from immune checkpoint inhibitors when STK11 is lost. These co-alteration signatures are not merely descriptive; they carry direct implications for prognosis, immunotherapy response, and the design of rational combination therapies.

The team also observed distinct genomic profiles across histological subtypes, meaning that adenocarcinomas, squamous cell carcinomas, and NSCLC not otherwise specified each carried their own characteristic alteration patterns. This reinforces the argument that comprehensive molecular profiling should be performed on every NSCLC patient regardless of subtype, rather than reserving broad panels for adenocarcinoma alone. Squamous tumors, which have historically had fewer actionable targets, may still harbor alterations in FGFR1, PIK3CA, or MET that could open doors to targeted trials or therapies. The study’s real-world design, drawing on routine diagnostic samples rather than a curated clinical trial population, makes these subtype-specific patterns particularly relevant to everyday practice.

Technically, the study relied on DNA-based sequencing, which captures SNVs, indels, and CNVs but does not directly detect gene fusions such as ALK, ROS1, or NTRK rearrangements, which typically require RNA-based or fusion-specific assays. This is an important caveat when interpreting the results, since fusions represent another major class of actionable drivers in NSCLC. The authors’ focus on DNA-based panels reflects the reality of many molecular pathology laboratories, where DNA sequencing is often the first and most accessible NGS platform. The findings therefore highlight both the power and the boundaries of DNA-only profiling, and they suggest that optimal molecular assessment may require combining DNA-based panels with complementary assays for fusions and other rearrangements.

The broader significance of the study lies in its scale and its population. Most large genomic datasets in NSCLC come from North American, European, or East Asian cohorts, and population-specific factors can influence both the prevalence of certain mutations and the interpretation of co-mutation patterns. By generating one of the largest Turkish NSCLC genomic datasets to date, the team has filled a meaningful gap in the global molecular map of lung cancer. The work also demonstrates the value of retrospective real-world studies, which can capture the heterogeneity of patients seen in routine clinical practice, including those with comorbidities, unusual histologies, or atypical treatment histories who are often underrepresented in prospective trials.

For clinicians, the study reinforces the argument that comprehensive NGS profiling should be standard of care in advanced NSCLC, not an optional extra. The identification of recurrent co-alteration patterns involving EGFR, KRAS, STK11, MET, and PIK3CA provides a framework for thinking about tumor biology in a more integrated way, moving beyond single-gene testing toward a holistic view of the genomic landscape. As targeted therapies and immunotherapies continue to expand, the ability to read a tumor’s full genomic signature, including the alterations that travel together, will become increasingly central to selecting the right treatment for the right patient. This study, grounded in real-world samples from a large Turkish cohort, adds both technical depth and population-specific evidence to that evolving paradigm.

Subject of Research: Lung cancer genomics

Article Title: Genomic alterations and co-mutation patterns in advanced NSCLC: a real-world DNA-based NGS study in a large Turkish cohort

Article References: Sadioglu, A., Ekinci, O., & Akyurek, N. (2026). Genomic alterations and co-mutation patterns in advanced NSCLC: a real-world DNA-based NGS study in a large Turkish cohort. BMC Cancer. https://doi.org/10.1186/s12885-026-16949-5

Image Credits: AI Generated

DOI: 10.1186/s12885-026-16949-5

Keywords: NSCLC, next-generation sequencing, EGFR, KRAS, STK11, MET, PIK3CA, co-mutation, lung cancer, genomic profiling, Genomic, alterations

Lung cancer remains the most commonly diagnosed and deadliest malignant tumor worldwide, and while radical surgical resection is the treatment of choice for early-stage disease, postoperative recurrence continues to undermine long-term outcomes for a substantial fraction of patients. The clinical dilemma is familiar to every thoracic oncologist: two patients with seemingly identical tumor stage and histology can follow radically different trajectories after surgery, one remaining disease-free for a decade, the other relapsing within a year. Traditional staging, based on tumor size, nodal involvement and metastatic status, captures only part of this biological heterogeneity. The Shanghai team set out to determine whether the mutational landscape of resected tumors could close that gap, systematically delineating the gene mutation profile of a large real-world Chinese cohort and assessing the predictive value of critical mutations for postoperative disease-free survival (DFS).

The methodology was comprehensive. All 1,674 enrolled patients, treated between long follow-up windows and retrospectively identified from clinical records, had undergone NGS-based molecular profiling of their resected tumor specimens. The investigators collected detailed clinicopathological data, treatment information and survival outcomes, then constructed mutation profiles for the overall population and for clinically relevant subgroups. Statistical machinery included Kaplan-Meier survival curves to visualize differences in disease-free survival, univariate and multivariate Cox proportional-hazards regression to isolate the independent contribution of each genetic and clinical variable, and co-mutation analysis to test whether pairs of altered genes interact to shape prognosis more than either alteration alone.

The mutational census that emerged was broad: 144 relevant mutated genes were detected across the cohort, and the team mapped their frequencies across the population. Two alterations dominated the landscape. EGFR, the epidermal growth factor receptor gene long known to be unusually frequent in East Asian lung cancer populations, was mutated in 61.2 percent of patients, while TP53, the canonical guardian-of-the-genome tumor suppressor, was altered in 31.9 percent. The high EGFR prevalence itself underscores why population-specific genomic studies matter; mutation frequencies observed in predominantly Western cohorts do not simply translate to Chinese patients, and treatment guidelines built on foreign data risk misaligning with the biology of the patients they are meant to serve.

The prognostic signal, however, came less from EGFR than from a set of other genes. Univariate analysis revealed that patients harboring mutations in TP53, KRAS, MET, ROS1 or CDKN2A experienced significantly shorter disease-free survival than patients without those alterations. In contrast, and perhaps counterintuitively, ERBB2 mutations were associated with longer DFS, a finding that held up when the investigators pushed the data through multivariate modeling. That final model, which adjusted for confounders simultaneously, confirmed TP53 mutation, ERBB2 mutation, tumor and nodal stage (T/N classification) and pleural invasion as independent prognostic factors for postoperative disease-free survival. In other words, even after accounting for how advanced a tumor appeared anatomically, the mutational status of these genes carried genuine, standalone predictive information about whether a patient’s cancer would return.

The most consequential result concerned gene pairs. When the team performed co-mutation analysis, patients whose tumors carried both TP53 and MET mutations had the poorest prognosis of any molecular subgroup, with disease-free survival significantly worse than that of patients carrying either single mutation and dramatically worse than those wild-type for both genes. This synergistic worsening of outcome is biologically plausible: TP53 loss disables apoptotic safeguards and genomic integrity checks, while MET activation drives invasive growth and metastatic signaling pathways, and the combination may equip residual micrometastatic disease with a particularly aggressive phenotype. Critically, the TP53-MET co-mutation signal remained consistent when the analysis was repeated in the subgroup of patients whose disease did progress, reinforcing the robustness of the association rather than reflecting a statistical artifact of the full cohort.

For clinical practice, the implications are tangible. Postoperative risk stratification for lung cancer currently leans heavily on pathological stage, with adjuvant therapy decisions guided largely by nodal status and tumor size. This study provides evidence that NGS profiling of resected specimens can refine that stratification: a patient with a TP53-mutant tumor, or worse, a TP53 and MET co-mutant tumor, could be flagged for intensified surveillance imaging, consideration of adjuvant targeted therapy or chemotherapy trials, and earlier detection of recurrence through molecular monitoring. Conversely, an ERBB2-mutant patient with otherwise favorable anatomy may face a lower recurrence risk than staging alone would suggest, information that could spare unnecessary treatment toxicity. The authors explicitly frame the results as a reliable molecular basis for postoperative risk stratification and individualized adjuvant therapy in lung cancer.

The study’s scale and design give its conclusions unusual weight. Retrospective cohort studies of this kind are vulnerable to selection bias and to the vagaries of heterogeneous treatment patterns, but a sample of 1,674 consecutively sequenced patients with long-term follow-up is rare in the literature on resected lung cancer, and the concordance between univariate, multivariate and co-mutation analyses, together with validation in the progression subgroup, argues that the signals are genuine properties of the disease rather than artifacts. The work was approved by the Ethics Committee and Institutional Review Board of Shanghai Chest Hospital under reference number LS1808, conducted in line with the Declaration of Helsinki, and supported by the National Natural Science Foundation of China and the Chinese Society of Clinical Oncology. Corresponding authors Zhi-qiang Gao, Jun Lu and Bao-hui Han led the effort, with Feng Pan and Liang Zheng contributing equally as first authors.

What remains to be established is prospective validation. Retrospective association, however robust, must ultimately be tested in trials that assign adjuvant treatment based on mutational profile and measure whether outcomes improve; ongoing studies of adjuvant targeted therapy in EGFR-mutant disease suggest this paradigm is already moving into the clinic. The Shanghai data add momentum by identifying which patients, within the large EGFR-mutated population, are most and least likely to benefit from closer surveillance. As NGS testing becomes routine and cheaper for resected lung cancer specimens, studies like this one chart the path from genomic description to genomic prognosis, where every surgical specimen yields not just a diagnosis but a personalized forecast, and where the risk of recurrence, lung cancer’s most feared late event, can be anticipated and countered before the first recurrent cell ever shows up on a scan.

Subject of Research: Postoperative gene mutations as prognostic predictors of disease-free survival in Chinese patients with resected lung cancer

Article Title: Relationship between postoperative gene mutation and prognosis in chinese patients with lung cancer: a long-term follow-up retrospective cohort study

Article References: Pan, F., Zheng, L., Zhang, L.-L., Wang, X., Lu, A.-T., Zhou, C., Gao, Z.-Q., Lu, J., & Han, B.-H. (2026). Relationship between postoperative gene mutation and prognosis in chinese patients with lung cancer: a long-term follow-up retrospective cohort study. BMC Cancer. https://doi.org/10.1186/s12885-026-16960-w

Image Credits: AI Generated

DOI: 10.1186/s12885-026-16960-w

Keywords: lung cancer, gene mutation, TP53, EGFR, MET, disease-free survival, next-generation sequencing, postoperative recurrence, prognostic factors, co-mutation, retrospective cohort study, adjuvant therapy