Utilizing clinical features, genetic mutations, and a 14-gene molecular assay for optimizing the management of multiple primary lung cancer

Introduction

Lung cancer remains one of the leading causes of cancer-related morbidity and mortality worldwide and imposes a substantial global health burden (1,2). Multiple primary lung cancer (MPLC), characterized as the presence of two or more primary lung cancer lesions within the same patient, has emerged as a challenging clinical issue in recent years and is widely recognized for its diagnostic and therapeutic complexity (3,4). The differentiation of MPLC from intrapulmonary metastasis (IM) has traditionally been achieved through the application of criteria based on anatomical location and pathological features, such as those proposed by the American College of Chest Physicians (ACCP) in 2013 (5). However, these criteria have notable limitations, and the accurate distinction between MPLC and IM remains difficult in many cases (6-8).

It has been demonstrated that lesions in MPLC harbor different mutations and exhibit varying drug sensitivities. However, current guidelines lack standardized protocols or models for incorporating molecular techniques into clinical decision-making (6,8-12). The pronounced interpatient heterogeneity of MPLC further complicates the establishment of uniform treatment strategies (13-16). In clinical practice, multidisciplinary teams (MDTs) often tailor treatment plans by integrating lesion-specific characteristics, prioritizing tumors with radiographically solid components, invasive pathological features, and other clinical or genomic risk factors (17-22). Despite these efforts, the correlation between the clinicopathological features and molecular genetic characteristics of MPLC has not been sufficiently determined, potentially limiting a deeper understanding of this disease. With the rapid advancement of molecular testing technologies and predictive models, there is a pressing need to integrate multidimensional molecular data to refine diagnosis and optimize individualized treatment strategies for MPLC.

The 14-gene RNA-level assay is a validated prognostic model that stratifies risk based on the RNA expression levels of selected genes. This risk stratification correlates strongly with 5-year mortality in patients with early-stage nonsquamous non-small cell lung cancer (NSCLC) and has been validated in large-scale international cohorts (23,24). In addition, the assay has undergone analytical validation on formalin-fixed, paraffin-embedded (FFPE)-derived RNA, demonstrating high precision and reproducibility of the quantitative polymerase chain reaction (qPCR) measurements and the derived risk score across repeat testing, supporting its robustness for clinical application (25). This assay can also accurately identify patients who may benefit from adjuvant chemotherapy or EGFR mutation-targeted therapy independently of tumor-node-metastasis (TNM) staging and is even capable of identifying high-risk patients within stage I (26-28). More recently, the randomized phase 3 AIM-HIGH trial further showed that adjuvant platinum-based chemotherapy significantly improved disease-free survival (DFS) versus observation among stage IA–IIA nonsquamous NSCLC patients classified as molecular high risk by this 14-gene expression profile, providing trial-level evidence that the assay can identify a subgroup with clinically actionable residual risk (29). The 11 cancer-related genes included in this RNA testing protocol (BAG1, BRCA1, CDC6, CDK2AP1, ERBB3, FUT3, IL11, LCK, RND3, SH3BGR, and WNT3A) are closely associated with a number of tumor biological processes, including cell proliferation (i.e., CDC6, ERBB3, and WNT3A) (30-33), apoptosis and antiapoptosis (i.e., BAG1 and CDK2AP1) (34,35), invasion and metastasis (i.e., FUT3, IL11, and RND3) (36-38), and immune regulation (i.e., LCK and IL11) (39,40). These genes are key factors in cancer recurrence and chemotherapy sensitivity. Our previous study that included a cohort of 102 patients demonstrated that a risk score based on the RNA expression levels of 14 genes is associated with higher computed tomography (CT) values, tumor mutational burden (TMB), and mutation rate of certain genes, including TP53, KRAS, and LRP1B, and is capable of independently predicting the inherent recurrence risk of lesions (41). This suggests that this method has the potential to identify nodules within MPLC that are at higher risk for recurrence or metastasis from another independent dimension, which can help to identify potential IM and inform adjuvant treatment strategies focusing on high-risk nodules in MPLC.

Therefore, this study was designed to further elucidate the clinical and genetic characteristics of MPLC and to propose an RNA-based approach to optimize its clinical management from a novel perspective. To this end, we prospectively collected clinical data from 69 pulmonary nodules in a cohort of 30 MPLC patients. Mutation profiles of multiple lesions were further characterized in 27 patients using next-generation sequencing (NGS). In addition, a 14-gene RNA-level molecular assay was applied to assess risk scores and stratify all nodules. We present this article in accordance with the TRIPOD reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0338/rc).

Methods

Sample collection

A total of 30 patients with MPLC who underwent surgery at the Department of Thoracic Surgery, Peking Union Medical College Hospital, were enrolled in this study. All patients included in this study were diagnosed with synchronous MPLC according to the 2013 ACCP criteria (5) or formally assessed by MDT meetings in alignment with the latest expert consensus guidelines for distinguishing MPLC from IM (42,43). The inclusion criteria for the MPLC cohort are shown in Figure 1. Briefly, tumors presenting as adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), or lepidic predominant adenocarcinoma (LPA) were considered primary lesions (42); otherwise, pathological evaluation was performed. When two lesions exhibited different histological features, such as predominant architectural patterns, they were classified as MPLC (43,44). According to the ACCP criteria, lesions with similar pathological features but located in different lung lobes and without N2 or N3 involvement or systemic metastases were identified as MPLC (5). Finally, regarding molecular information, lesions with distinct driver mutations were more likely to be classified as MPLC (42). Demographic information of the enrolled patients, including age, gender, smoking history, family history of cancer, TNM stage, and number of nodules, was collected. In addition, tumor-specific characteristics such as radiological features and pathological types were recorded for each lesion. The criterion of a family history of cancer was considered to be whether the patient had a first-degree relative with a history of malignant tumors (45). Detailed radiographical characteristics such as the maximum diameter, average CT value, maximum CT value, average enhanced CT value, maximum enhanced CT value, and maximum standardized uptake value (SUVmax) of the nodules were also recorded in our analysis. In addition, we collected the mutation information of 63 nodules that underwent targeted sequencing of 520 genes to investigate the genetic features of MPLC. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Peking Union Medical College Hospital (No. K4171), and informed consent was taken from all the patients.

Figure 1 The inclusion criteria for the MPLC cohort. AIS, adenocarcinoma in situ; LPA, lepidic predominant adenocarcinoma; MDT, multidisciplinary team; MIA, minimally invasive adenocarcinoma; MPLC, multiple primary lung cancer; N, node.

DNA isolation and targeted sequencing

The sequencing of a commercial 520-gene panel (OncoScreen Plus) was performed at Burning Rock Biotech (Guangzhou, China). Briefly, the QIAamp DNA FFPE tissue kit (Qiagen, Hilden, Germany) was used to extract DNA. DNA purification with the Agencourt AMPure XP Kit (Danaher Corporation, Washington D.C., USA), targeted capture with OncoScreen Plus (Burning Rock Biotech, Guangzhou, China), and quality control with Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA) were then conducted. The NextSeq 500 (Illumina, Inc., San Diego, CA, USA) was used for sequencing with paired-end reads and an average sequencing depth of 1,000×. Data were aligned to the reference human genome (hg19) via the Burrows-Wheeler Aligner version 0.7.10 (46). TMB was calculated as the ratio of the nonsynonymous mutation count, with the exclusion of hotspot mutations, copy number variations (CNVs), structural variations, and germline single nucleotide polymorphisms, to the total coding region size of the panel.

For tissue-derived samples, single nucleotide variants required a minimum of eight supporting reads, whereas insertion and deletion events were retained with at least two and five supporting reads, respectively. Variants with a population allele frequency exceeding 0.1% in public germline databases, including ExAC, 1000 Genomes, dbSNP, and ESP6500SI-V2, were considered common polymorphisms and excluded from downstream analyses. Functional annotation of the remaining variants was performed using ANNOVAR (release February 1, 2016) together with SnpEff v3.6. Structural variants were identified using Factera v1.4.3. CNVs were inferred from sequencing coverage across target regions after correction for guanine-cytosine (GC)-content and probe-related capture bias. To minimize inter-sample variability, coverage data were normalized using the average sequencing depth of all captured regions. CNV status was determined by calculating the ratio of sequencing depth in tumor samples to the mean coverage of a sufficient number of CNV-negative control samples (n>50) for each captured interval. Genomic regions exhibiting significant quantitative and statistical deviations from controls were classified as CNVs. Thresholds for CNV calling were defined as 1.5 for copy loss and 2.64 for copy gain. Microsatellite instability status was evaluated using a previously validated algorithm based on sequencing read-count distribution patterns.

Risk stratification by the 14-gene molecular assay

Procedures related to the 14-gene molecular assay were conducted with commercially available kits and in accordance with the protocol described in published literature (23). Total RNA was extracted from FFPE samples using the RNeasy FFPE Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. RNA yield (ng) and purity (A260/280 and A260/230 ratios) were assessed using a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). The isolated RNA was then reverse-transcribed into cDNA, followed by qPCR using TaqMan assays, as previously described (23). The relative expression level of 11 cancer-related target genes (BAG1, BRCA1, CDC6, CDK2AP1, ERBB3, FUT3, IL11, LCK, RND3, SH3BGR, and WNT3A) and 3 reference genes (ESD, TBP, and YAP1) was appraised with the comparative cycle threshold method. The risk score and stratification of each lesion were then generated according to a validated and published prognostic model (23). The assay measured RNA expression levels of 11 cancer-related genes and 3 reference genes in tumor tissue and calculated a continuous risk score using a patented algorithm (Chinese invention patent ZL201410053021.X). Briefly, the relative gene expression values were input into a predefined L2-penalized Cox proportional hazards model [least absolute shrinkage and selection operator (LASSO)-Cox] to generate the risk score, which was then used for stratification according to predefined cutoffs established in prior study cohorts. Three risk groups were defined a priori: low risk (risk score ≤33rd percentile), intermediate risk (33rd–67th percentile), and high risk (risk score >67th percentile) (23). In the present MPLC study, these validated thresholds were directly applied for risk classification.

Statistical analyses

R version 4.3.2 software (The R Foundation for Statistical Computing, Vienna, Austria) was used to conduct statistical analyses. The Fisher exact test was used to determine the association of categorical variables, such as gender, smoking history, family history, and radiological features. Meanwhile, the Shapiro-Wilk test was used to check for the normal distribution of continuous variables, such as age, maximum tumor diameter, SUVmax, and various CT values, and the var.test() function was used to examine the homogeneity of variances. The Mann-Whitney U test was then applied to analyze nonnormally distributed or heteroscedastic two-group data, while the t-test was used for normally distributed and homoscedastic two-group data. The R package ggplot2 was employed to generate the plots.

Results

Clinical features of patients and nodules

A total of 69 tumors from 30 patients were prospectively included in our analysis (Table S1). The median age of the population was 59.5 years, with a range from 27 to 81 years. The majority of patients were female (86.7%) and had no smoking history (90.0%). Notably, a considerable proportion (46.7%) of patients had a family history of cancer, with 7 (23.3%) patients having a first-degree relative diagnosed with lung cancer. Regarding the characteristics of the nodules, 43 (62.3%) pure ground-glass opacities (pGGOs), 15 (21.7%) mixed ground-glass opacities (mGGOs), and 11 (15.9%) solid nodules (SNs) were included in the analysis. Postoperative pathological analyses revealed that 15 (21.7%) nodules were diagnosed as AIS, 25 (36.2%) nodules as MIA, 28 (40.6%) nodules as IAC, and 1 (1.4%) nodule as carcinoid.

To identify which nodules should be prioritized for surgical intervention, we compared the clinical characteristics of IAC and non-IAC nodules. A higher maximum diameter, average CT value, maximum CT value, average enhanced CT value, maximum enhanced CT value, SUVmax, and solid component on imaging were indicative of lesions with greater invasiveness (Figure S1). This observation remained largely unchanged when nodules were classified into AIS and non-AIS (Figure S2).

Genetic sequencing results of MPLC

The complete DNA mutation results of multiple lesions in 27 patients were obtained via NGS (Figure 2A, Table S2). The EGFR mutation was the most prevalent mutation, present in 44% of all detected lesions. A total of 17 (63.0%) patients had at least one lesion with the EGFR mutation; of these, 7 patients had only one lesion with the EGFR mutation, while 10 patients had two or multiple lesions with EGFR mutations. Among all nodules, the next most common mutation was RBM10 (19%), followed by ERBB2 (14%), BRAF (10%), KRAS (10%), KMT2C (8%), LRP1B (8%), and TP53 (8%). It is worth noting that the majority of ERBB2 mutations were ERBB2 exon 20 insertion (ERBB2 20ins; 88.9%), and among the BRAF-mutated nodules, only one nodule (16.7%) exhibited the BRAF V600E mutation.

Figure 2 Mutation data of MPLCs in 27 patients. (A) Spectrum of the key genomic mutations in MPLC. (B) Relationship between mutation frequencies of the key genes and pathological invasiveness or radiological solidity. (C) Correlation between TMB and pathological invasive components. *, P<0.05; ***, P<0.001. AIS, adenocarcinoma in situ; CN, copy number; IAC, invasive adenocarcinoma; mGGO, mixed ground-glass opacity; MIA, minimally invasive adenocarcinoma; MPLC, multiple primary lung cancer; pGGO, pure ground-glass opacity; SN, solid nodule; SUVmax, maximum standardized uptake value; TMB, tumor mutation burden.

We subsequently sought to determine how the frequency of these mutations varied according to the progression of the nodules. Combining the pathological features of each lesion, we found that AIS exhibited a higher frequency of ERBB2 20ins. As the pathologically invasive component increased, the frequency of ERBB2 20ins decreased, while EGFR mutations increased and became the dominant mutation. mGGOs exhibited the highest frequency of EGFR mutations. Moreover, TP53 mutations were observed only in IAC and SNs in this cohort (Figure 2B). In addition, the aggressiveness of the nodules increased with the TMB, although this association was not significant (P=0.09) (Figure 2C).

Risk of nodules according to the 14-gene molecular assay

Risk scores and stratifications for all 69 nodules from 30 patients were evaluated with the 14-gene molecular assay. Among these nodules, 50 (72.5%) were classified as low risk, 11 (15.9%) as medium risk, and 8 (11.6%) as high risk. Moreover, 17 (56.7%) patients had all their lesions classified as low risk, while 13 (43.3%) had at least one lesion evaluated as medium or high risk.

In the same patient, nodules harboring comutation sites did not necessarily share the same risk stratification, while those with the same risk stratification could exhibit entirely different mutation profiles (Figure S3A,S3B). The comparison of clinical information between patients with exclusively low-risk nodules and those with medium- or high-risk nodules revealed that older patients or males were more likely to have medium- or high-risk nodules (Figure S3C).

Correspondingly, nodules in male patients and smokers were more likely to be evaluated as medium- or high-risk nodules. Nodules with higher tumor diameter and average and maximum CT values also tended to have a higher risk. However, such an association was not observed for enhanced CT values and SUVmax, suggesting that the 14-gene molecular assay is associated with certain clinical features while also demonstrating prognostic capability independent of several conventional clinical imaging parameters (Figure 3A). Additionally, medium- or high-risk lesions had a significantly higher frequency of TP53 mutations, and they tended to bear a higher TMB (Figure 3B,3C), indicating an association between mutational features and the 14-gene molecular assay. However, EGFR mutations did not appear to be associated with the long-term risk of lesions (Figure 3B).

Figure 3 Risk stratification and its correlation with clinical factors and mutations. (A) Relationship between risk stratification and clinical factors. (B) Relationship between risk stratification and key gene mutations. (C) Relationship between risk stratification and TMB. (D) Distribution of high, medium, and low risk across lesions. *, P<0.05; **, P<0.01. AIS, adenocarcinoma in situ; CT, computed tomography; IAC, invasive adenocarcinoma; mGGO, mixed ground -lass opacity; MIA, minimally invasive adenocarcinoma; pGGO, pure ground-glass opacity; SN, solid nodule; SUVmax, maximum standardized uptake value; TMB, tumor mutation burden.

We assessed whether these risk stratifications accurately reflected the invasiveness of the nodules and found that nearly all (93.3%) AIS were classified as low-risk lesions. Surprisingly, 67.9% of IACs were also determined to be low risk, while only 10.7% were diagnosed as high risk (Figure 3D). IACs with a higher average or maximum CT value, radiological solid component, and micropapillary or solid pathological component tended to have a higher long-term risk (Figure S4).

To assess the ability of the 14-gene molecular assay to identify lesions requiring adjuvant treatment among multiple nodules in a patient, we examined the risk stratification of each nodule in each patient. Of note, among the 4 nodules that should be considered for adjuvant therapy according to the National Comprehensive Cancer Network (NCCN) guideline, 3 were identified as medium- or high-risk lesions by the 14-gene molecular assay. Furthermore, the risk scores of nodules deemed suitable for adjuvant therapy after resection were the highest among the multiple lesions in the respective patients. This suggests that the 14-gene molecular assay can assist in identifying nodules within MPLC that may require adjuvant therapy.

Utilizing molecular information to optimize MPLC management

An exemplary case involving both MPLC and IM further demonstrated the potential of the 14-gene molecular assay to evaluate nodules requiring adjuvant treatment in MPLC. A patient with IAC in the right upper lobe and MIA in the right lower lobe was unexpectedly found to have additional pulmonary nodules during the surgery (Figure 4A). The IAC in the right upper lobe and one of the nodules exhibited similar pathological subtypes and overlapping mutation profiles, while the MIA in the right lower lobe displayed distinct mutations (Figure 4B). Gene copy number plots of these lesions further indicated that the IAC in the right upper lobe and the unexpectedly discovered nodule shared similar alterations (Figure 4C). Additionally, the risk score for the IAC in the right upper lobe was 56.053 (high risk), while that for the MIA in the right lower lobe was 18.039 (low risk) (Figure 4B). Based on these findings, the MDT considered that the unexpectedly discovered nodules were metastatic lesions originating from the IAC in the right upper lobe. We observed that other larger solid tumors in this cohort, which were assessed as low or medium risk, did not exhibit metastasis. In this case, the 14-gene molecular assay demonstrated its potential as a novel approach for identifying nodules that may require adjuvant treatment in patients with MPLC and can help determine the likelihood of IM by assessing nodule risk.

Figure 4 A patient with synchronous MPLC and IM. (A) Schematic diagram and CT images of nodules in this patient. (B) Pathological characteristics, genetic features, risk score, and stratification of these nodules. (C) Gene copy number plots of these nodules. The black arrows indicate the nodules identified during surgery in the right lower lobe. CT, computed tomography; Est. CN, estimated copy number; IAC, invasive adenocarcinoma; IM, intrapulmonary metastasis; MIA, minimally invasive adenocarcinoma; MPLC, multiple primary lung cancer.

Another case involving a patient with IACs in both the right upper lobe and left lower lobe confirmed that the 14-gene molecular assay can assist in the arrangement of adjuvant treatment plans (Figure S5A). The nodule in the left lower lobe with the EGFR exon 19 deletion was assessed as medium risk, while the nodule in the right upper lobe with the RET rearrangement was assessed as low risk. Ultimately, according to the combination of clinical risk factors and the 14-gene molecular assay results, osimertinib was selected as the drug for adjuvant treatment, which has yielded more than 4 years of DFS thus far (Figure S5B).

To date, the majority of patients have been followed up for 3 years. In the real-world setting, patients with medium- or high-risk nodules had a higher rate of receiving adjuvant therapy (4/13, 30.8%) compared to those with exclusively low-risk nodules (1/17, 5.9%) (Figure 5A). Thus far, no patients with exclusively low-risk nodules have died, experienced recurrence, or have been loss to follow-up. Meanwhile, among patients with medium- or high-risk nodules, one (P5) died due to pleural effusion and complicated pulmonary infection. Tumor recurrence was observed in two other patients (P9 and P16). Notably, tumor recurrence occurred only in patients with medium- or high-risk nodules who did not receive adjuvant therapy postoperatively. Patients with all nodules classified as low-risk showed significantly improved DFS (Figure 5B). These findings highlight the significant prognostic value of the 14-gene molecular assay in patients with MPLC.

Figure 5 Follow-up data of the cohort. (A) Swimlane plot of follow-up data for patients in the MPLC cohort. (B) Kaplan-Meier survival curve of the cohort. *, P<0.05. L, patients with exclusively low-risk nodules; M/H, patients with medium- or high-risk nodules; MPLC, multiple primary lung cancer.

Our study identified preoperative clinical indicators of nodules with higher invasiveness in MPLC and demonstrated the value of the 14-gene molecular assay as an independent tool for postoperatively diagnosing high-risk nodules. Based on these results and clinical experience, we constructed a theoretical decision tree in order to contribute to the refinement and standardization of the clinical management process for MPLC (Figure 6). In simple terms, female patients and those with a family history of cancer may warrant heightened attention during MPLC screening. Preoperatively, nodules with larger diameters, higher densities, elevated SUVmax values, and solid components should receive attention, and surgical intervention should be preferred for these higher-risk cases. In addition to pathological features, oncogenic driver mutations, and traditional clinical risk factors, we recommend incorporating risk stratification implemented via the 14-gene molecular assay to identify tumors with a higher risk of recurrence and inform adjuvant therapy for patients with MPLC.

Figure 6 Theoretical decision tree for MPLC. CT, computed tomography; IM, intrapulmonary metastasis; MDT, multidisciplinary team; MPLC, multiple primary lung cancer; NCCN, National Comprehensive Cancer Network; SBRT, stereotactic body radiation therapy; SUVmax, maximum standardized uptake value; TNM, tumor-node-metastasis.

Discussion

With the increasing prevalence of CT scans, MPLC are being discovered at a higher rate (8,47,48). However, the clinical management of MPLC is currently not well established. In this study, we prospectively collected cases with multiple lung cancer lesions to clarify the clinical and genetic features of MPLC and to assess the ability of a novel 14-gene molecular assay to identify nodules that warrant heightened attention or adjuvant treatment in MPLC management. Our findings indicate that different lesions within a single synchronous MPLC patient may carry distinct long-term risk profiles, highlighting the need for lesion-specific assessment.

In analyzing the demographic characteristics of these patients with MPLC and the clinical features of their nodules, we found that the bulk of nodules in Chinese patients with MPLC manifest as pGGOs or mGGOs and are adenocarcinomas. Moreover, the patient population had a higher proportion of females, consistent with the conclusions of previous studies (49-51). A significant number of patients with MPLC had a family history of tumors, especially lung cancer, suggesting a potential hereditary tendency (44). Therefore, females with a family history of tumors should be a target group in MPLC screening.

The surgery remains a critical part of the treatment strategy for patients with MPLC and involves identifying those nodules for which intervention is most necessary, especially in cases where radical excision of all lesions is challenging (48). In current MPLC guidelines and expert consensus, surgical resection remains the preferred treatment modality for MPLC (51-53). Postoperative adjuvant therapy is generally recommended for patients with MPLCs presenting high-risk features, including positive surgical margins, TNM stage II disease, stage IB tumors with poor differentiation, vascular invasion, visceral pleural invasion, or intraparenchymal dissemination, following curative resection (51). However, when the benefit of adjuvant therapy and the optimal treatment strategy remain uncertain, management should follow the principles established for solitary primary lung cancer (51). In clinical practice, the selection of the “highest-risk” lesion still largely depends on radiologic features, histopathological parameters, and clinician experience. A standardized framework for lesion-level postoperative risk stratification has not yet been established. Our data suggest that integrating an imaging- and pathology-independent 14-gene expression-based risk signal may facilitate preoperative lesion prioritization and further refine postoperative therapeutic escalation strategies. To preoperatively determine which nodules are more likely to be IACs and thus be prioritized in surgical planning, we analyzed the clinical features of nodules diagnosed as IAC in patients with MPLC. We found that imaging characteristics, such as maximum diameter, average CT value, maximum CT value, average enhanced CT value, SUVmax, and solid component, could reflect the invasiveness of MPLC nodules. According to the risk stratification conducted via the 14-gene molecular assay, nodules in males and smokers may require more aggressive management. These results provide evidence for the preoperative identification of nodules requiring priority resection.

With regard to the genetic mutation profiles of MPLC, EGFR, RBM10, ERBB2, BRAF, KRAS, KMT2C, LRP1B, and TP53 exhibited the highest mutation frequencies, consistent with similar studies (54,55). Moreover, AIS exhibited a high frequency of ERBB2 20ins, and the frequency of EGFR mutations increased successively with invasiveness, from AIS to IAC. This suggests that using EGFR tyrosine kinase inhibitors to treat inoperable MPLC may be effective, as it may help control multiple invasive lesions. Moreover, TMB also increased with the proportion invasive components, consistent with our accumulated experience in solitary lung cancers (56).

Due to the limitations of traditional radiology and histology in distinguishing between MPLC and IM, growing number of NGS-based discrimination algorithms have been developed (51,57-59). However, it is important to bear in mind that there is no standardized procedure for distinction based on sequencing data, with many studies still heavily relying on empirical and personalized analyses (59,60). Additionally, the reliability of sequencing-based discrimination might be heavily influenced by the scale of mutation-test panels (42,60,61). In current MPLC care, the lack of standardized sequencing-based procedures for distinguishing MPLC from IM can directly worsen treatment decisions and prognosis (42,61). If IM is misclassified as MPLC, patients may receive lesion-limited local therapy without timely systemic escalation, increasing the risk of early recurrence. Conversely, if true MPLC is misclassified as IM, patients may be inappropriately upstaged, denied potentially curative resections, or exposed to unnecessary systemic toxicity. Uncertainty in clonal interpretation also complicates “dominant lesion” selection, potentially prioritizing the wrong nodule for adjuvant therapy, leaving the most aggressive lesion insufficiently treated. Inconsistent molecular interpretation further impairs lesion-level surveillance intensity, creating inter-institutional variability in MDT decisions and patient counseling. Novel approaches such as DNA methylation and metabolomics are being developed to assist in the diagnosis of MPLC (62,63), and the application of machine learning and deep learning in diagnosing MPLC is intensifying (42). However, these methods remain at the preliminary (12). Panel breadth can influence clonality inference, as smaller panels may provide limited informative value in certain patients (61). Accordingly, a gene-expression-based risk assay is complementary in nature; it does not assess clonality through direct mutational profiling, but rather estimates lesion-specific long-term recurrence risk that is not directly captured by imaging, pathology, or clonality-oriented sequencing approaches, thereby providing additional, indirect evidence that may support differential diagnostic interpretation from a distinct biological dimension.

The postoperative identification of nodules with long-term risk in MPLC and selecting treatment plans based on risk stratification of lesions with various mutations are also critical components of MPLC management. In this study, we assessed the ability of 14-gene molecular assay, which had demonstrated good performance in survival prediction among patients with solitary nonsquamous NSCLC, to identify nodules with a higher long-term higher risk and requiring adjuvant treatment in MPLC. Currently, no standardized guideline specifically addresses adjuvant therapy in MPLC. Treatment decisions are largely empirical and rely on a multidisciplinary integration of imaging characteristics, histopathological findings, and molecular alterations at the lesion level (51,53). We found that risk stratification based on the 14-gene molecular assay was not significantly associated with imaging features, suggesting that it can serve as an independent dimension beyond traditional clinical features in determining which lesions in MPLC require adjuvant treatment. The 14-gene expression profile was developed in resected nonsquamous NSCLC to further refine recurrence risk beyond routine clinicopathologic factors and guide adjuvant treatment decisions. In the AIM-HIGH trial, patients with assay-defined molecular high-risk stage IA–IIA disease derived a DFS benefit from platinum-based adjuvant chemotherapy, supporting the clinical utility of this risk stratification approach (29). When applied to MPLC, lesion-level risk scoring may assist in prioritizing therapeutic escalation for the highest-risk lesions while avoiding overtreatment of indolent nodules. However, it should be used as a complement rather than a substitute for imaging, pathology, and, when necessary, genomic testing. Importantly, this assay does not provide definitive discrimination between MPLC and IM. In an exemplary case, one nodule in MPLC, which developed metastasis, was assessed as high risk, while the other with no metastasis was assessed as low risk, confirming the reliability of this risk-assessment assay. This further suggests another potential application, namely, determining the likelihood of IM being present among multiple lesions according to their risk stratification. Multiple nodules in patients with MPLC may harbor different driver gene mutations, necessitating distinct types of adjuvant therapy (64). Another case in our study demonstrated that applying the 14-gene molecular assay was reliable for selecting adjuvant therapy drugs in patients with nodules harboring different driver mutations. We believe that this assay can help assess the long-term risk of each nodule and provide evidence-based management for MPLC as a supplement to clinical and genetic risk factors. It is important to note that, as there have been no prior studies applying the 14-gene molecular assay in MPLC, we cannot ascertain whether this assay better reflects the recurrence risk of lesions compared to traditional high-risk factors. Therefore, based on our study, we recommend considering the 14-gene molecular assay as an additional dimension in clinical scenarios where treatment options are difficult to determine; larger prospective cohorts are needed to demonstrate the superiority of this assay over other clinical high-risk factors in MPLC management.

Our study involved several limitations that should be acknowledged. First, the sample size was insufficient to support the construction of decision trees. This is because the proportion of patients with multiple clinically significant lung cancer lesions was relatively low. Larger prospective studies should be conducted to validate the clinical utility of the 14-gene molecular assay in the management of MPLC. Second, the 14-gene molecular assay has only been proven effective in distinguishing patients requiring adjuvant chemotherapy or EGFR-targeted therapy, and further investigation is needed to determine its ability to identify postoperative patients who may benefit from immunotherapy. Third, due to the limited sample size and the small number of outcome events, we were unable to integrate clinical variables, genomic data, and the 14-gene expression-based risk stratification into a unified joint prediction model. Future studies with larger cohorts and longer follow-up will be conducted to develop and validate a robust integrated predictive model.

Conclusions

MPLCs exhibit unique clinical characteristics and mutation profiles. Risk stratification with the 14-gene molecular assay can facilitate the clinical management of MPLC.

Acknowledgments

We would like to thank the patients and their families for their consent and support in participating in this study.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0338/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0338/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0338/prf

Funding: This study was supported by the CAMS Innovation Fund for Medical Sciences (No. 2023-I2M-C&T-B-0192), the National High Level Hospital Clinical Research Funding (No. 2022-PUMCH-B-011), the National Key Research and Development Program of China (No. 2024YFB4708805), the Beijing Natural Science Foundation (No. L258029), and the Wu Jieping Medical Foundation (No. 320.6750.2025-20-24).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0338/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Peking Union Medical College Hospital (No. K4171), and informed consent was taken from all the patients.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
2. Leiter A, Veluswamy RR, Wisnivesky JP. The global burden of lung cancer: current status and future trends. Nat Rev Clin Oncol 2023;20:624-39. [Crossref] [PubMed]
3. Guo W, Zhou B, Bie F, et al. Single-cell RNA sequencing analysis reveals transcriptional heterogeneity of multiple primary lung cancer. Clin Transl Med 2023;13:e1453. [Crossref] [PubMed]
4. Yang Y, Lin H, Shi L, et al. The characteristics and prognoses of 72 postoperative synchronous multiple primary lung cancer patients. World J Surg Oncol 2025;23:355. [Crossref] [PubMed]
5. Kozower BD, Larner JM, Detterbeck FC, et al. Special treatment issues in non-small cell lung cancer: Diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest 2013;143:e369S-99S.
6. Yang R, Li P, Wang D, et al. Genetic and immune characteristics of multiple primary lung cancers and lung metastases. Thorac Cancer 2021;12:2544-50. [Crossref] [PubMed]
7. Tian S, Li F, Pu J, et al. Differential Diagnostic Value of Histology in MPLC and IPM: A Systematic Review and Meta-Analysis. Front Oncol 2022;12:871827. [Crossref] [PubMed]
8. Liufu Y, Su R, Wen Y, et al. A CT-based deep learning approach to differentiate multiple primary lung cancers, metastases, and benign nodules. BMC Cancer 2026;26:177. [Crossref] [PubMed]
9. Murphy SJ, Harris FR, Kosari F, et al. Using Genomics to Differentiate Multiple Primaries From Metastatic Lung Cancer. J Thorac Oncol 2019;14:1567-82. [Crossref] [PubMed]
10. Xu L, Chen J, Zeng Y, et al. Differential diagnosis of multiple primary lung cancers and intra-lung metastasis of lung cancer by multiple gene detection. Chin Med J (Engl) 2022;135:86-8. [Crossref] [PubMed]
11. Bai G, Li Y, Ji Y, et al. Treatments and whole exon sequencing of a case with multiple primary lung cancer. J Cardiothorac Surg 2023;18:58. [Crossref] [PubMed]
12. Wang Z, Yuan X, Jiang G, et al. Towards the molecular era of discriminating multiple lung cancers. EBioMedicine 2023;90:104508. [Crossref] [PubMed]
13. Xue X, Liu Y, Pan L, et al. Diagnosis of multiple primary lung cancer: a systematic review. J Int Med Res 2013;41:1779-87. [Crossref] [PubMed]
14. Detterbeck FC, Franklin WA, Nicholson AG, et al. The IASLC Lung Cancer Staging Project: Background Data and Proposed Criteria to Distinguish Separate Primary Lung Cancers from Metastatic Foci in Patients with Two Lung Tumors in the Forthcoming Eighth Edition of the TNM Classification for Lung Cancer. J Thorac Oncol 2016;11:651-65.
15. Shi J, Liang L, Gao Y. Mapping the clinical trial landscape of multiple primary lung cancer in the era of precision oncology: persistent exclusion and design limitations hamper evidence-based treatment development. Front Oncol 2025;15:1642062. [Crossref] [PubMed]
16. Ren ZR, Wu L, Lu C, et al. Multiple Primary Lung Cancer With Driver Gene Mutations: Is Targeted Therapy Always the Optimal Choice?-A Case Report. Thorac Cancer 2026;17:e70215. [Crossref] [PubMed]
17. Gregoire J. Guiding Principles in the Management of Synchronous and Metachronous Primary Non-Small Cell Lung Cancer. Thorac Surg Clin 2021;31:237-54. [Crossref] [PubMed]
18. Handa Y, Tsutani Y, Okada M. Transition of Treatment for Ground Glass Opacity-Dominant Non-Small Cell Lung Cancer. Front Oncol 2021;11:655651. [Crossref] [PubMed]
19. Mimae T, Tsutani Y, Miyata Y, et al. Solid Tumor Size of 2 cm Divides Outcomes of Patients With Mixed Ground Glass Opacity Lung Tumors. Ann Thorac Surg 2020;109:1530-6. [Crossref] [PubMed]
20. Travis WD, Asamura H, Bankier AA, et al. The IASLC Lung Cancer Staging Project: Proposals for Coding T Categories for Subsolid Nodules and Assessment of Tumor Size in Part-Solid Tumors in the Forthcoming Eighth Edition of the TNM Classification of Lung Cancer. J Thorac Oncol 2016;11:1204-23.
21. Travis WD, Brambilla E, Noguchi M, et al. International association for the study of lung cancer/american thoracic society/european respiratory society international multidisciplinary classification of lung adenocarcinoma. J Thorac Oncol 2011;6:244-85. [Crossref] [PubMed]
22. Sullivan I, Salazar J, Arqueros C, et al. KRAS genetic variant as a prognostic factor for recurrence in resectable non-small cell lung cancer. Clin Transl Oncol 2017;19:884-90. [Crossref] [PubMed]
23. Kratz JR, He J, Van Den Eeden SK, et al. A practical molecular assay to predict survival in resected non-squamous, non-small-cell lung cancer: development and international validation studies. Lancet 2012;379:823-32. [Crossref] [PubMed]
24. Kratz JR, Van den Eeden SK, He J, et al. A prognostic assay to identify patients at high risk of mortality despite small, node-negative lung tumors. JAMA 2012;308:1629-31. [Crossref] [PubMed]
25. Kratz JR, Tham PT, Mulvihill MS, et al. Analytical validation of a practical molecular assay prognostic of survival in nonsquamous non-small cell lung cancer. Diagn Mol Pathol 2013;22:65-9. [Crossref] [PubMed]
26. Woodard GA, Wang SX, Kratz JR, et al. Adjuvant Chemotherapy Guided by Molecular Profiling and Improved Outcomes in Early Stage, Non-Small-Cell Lung Cancer. Clin Lung Cancer 2018;19:58-64. [Crossref] [PubMed]
27. Jiang Y, Lin Y, Fu W, et al. The impact of adjuvant EGFR-TKIs and 14-gene molecular assay on stage I non-small cell lung cancer with sensitive EGFR mutations. EClinicalMedicine 2023;64:102205. [Crossref] [PubMed]
28. Jiang Y, Lin Y, Fu W, et al. Clinicopathological factors vs. molecular model for predicting adjuvant EGFR-TKI benefit in stage I EGFR-mutant non-small cell lung cancer. Transl Lung Cancer Res 2025;14:1531-42.
29. Spigel DR, Westeel V, Anderson IC, et al. Adjuvant chemotherapy for stage IA-IIA non-squamous, non-small-cell lung cancer identified as molecular high-risk by a 14-gene expression profile (AIM-HIGH): an international, randomised, phase 3 trial. Lancet Respir Med 2025;13:887-96. [Crossref] [PubMed]
30. Lim N, Townsend PA. Cdc6 as a novel target in cancer: Oncogenic potential, senescence and subcellular localisation. Int J Cancer 2020;147:1528-34. [Crossref] [PubMed]
31. Foreman PK, Gore M, Kobel PA, et al. ErbB3 inhibitory surrobodies inhibit tumor cell proliferation in vitro and in vivo. Mol Cancer Ther 2012;11:1411-20. [Crossref] [PubMed]
32. Lu C, He Y, Duan J, et al. Expression of Wnt3a in hepatocellular carcinoma and its effects on cell cycle and metastasis. Int J Oncol 2017;51:1135-45. [Crossref] [PubMed]
33. Luo Y, Li Y, Liu S, et al. Single-cell profiling of ERBB family receptors identifies ERBB3 as a key regulator in head and neck squamous cell carcinoma progression. Discov Oncol 2025;16:1843. [Crossref] [PubMed]
34. Sharp A, Crabb SJ, Townsend PA, et al. BAG-1 in carcinogenesis. Expert Rev Mol Med 2004;6:1-15. [Crossref] [PubMed]
35. Zolochevska O, Figueiredo ML. Expression of cell cycle regulator cdk2ap1 suppresses tumor cell phenotype by non-cell-autonomous mechanisms. Oral Oncol 2009;45:e106-12. [Crossref] [PubMed]
36. He C, Li A, Lai Q, et al. The DDX39B/FUT3/TGFβR-I axis promotes tumor metastasis and EMT in colorectal cancer. Cell Death Dis 2021;12:74. [Crossref] [PubMed]
37. Chen Z, Chen X, Xie R, et al. DANCR Promotes Metastasis and Proliferation in Bladder Cancer Cells by Enhancing IL-11-STAT3 Signaling and CCND1 Expression. Mol Ther 2019;27:326-41. [Crossref] [PubMed]
38. Paysan L, Piquet L, Saltel F, et al. Rnd3 in Cancer: A Review of the Evidence for Tumor Promoter or Suppressor. Mol Cancer Res 2016;14:1033-44. [Crossref] [PubMed]
39. Bommhardt U, Schraven B, Simeoni L. Beyond TCR Signaling: Emerging Functions of Lck in Cancer and Immunotherapy. Int J Mol Sci 2019;20:3500. [Crossref] [PubMed]
40. Tang M, Xu M, Wang J, et al. Brain Metastasis from EGFR-Mutated Non-Small Cell Lung Cancer: Secretion of IL11 from Astrocytes Up-Regulates PDL1 and Promotes Immune Escape. Adv Sci (Weinh) 2024;11:e2306348. [Crossref] [PubMed]
41. Huang Z, Zhao M, Li B, et al. Correlations between 14-gene RNA-level assay and clinical and molecular features in resectable non-squamous non-small cell lung cancer: a cross-sectional study. Transl Lung Cancer Res 2024;13:3202-13. [Crossref] [PubMed]
42. Chou TY, Dacic S, Wistuba I, et al. Differentiating Separate Primary Lung Adenocarcinomas From Intrapulmonary Metastases With Emphasis on Pathological and Molecular Considerations: Recommendations From the International Association for the Study of Lung Cancer Pathology Committee. J Thorac Oncol 2025;20:311-30. [Crossref] [PubMed]
43. Yang CY, Yeh YC, Wang LC, et al. Genomic Profiling With Large-Scale Next-Generation Sequencing Panels Distinguishes Separate Primary Lung Adenocarcinomas From Intrapulmonary Metastases. Mod Pathol 2023;36:100047. [Crossref] [PubMed]
44. Yang P, Liu S, Chen Y, et al. Decoding the enigma of multiple primary lung cancers: from mechanism to bedside-a narrative review. Transl Lung Cancer Res 2025;14:5181-97. [Crossref] [PubMed]
45. Wang CL, Hsu KH, Chang YH, et al. Low-Dose Computed Tomography Screening in Relatives With a Family History of Lung Cancer. J Thorac Oncol 2023;18:1492-503. [Crossref] [PubMed]
46. Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 2009;25:1754-60. [Crossref] [PubMed]
47. Tie H, Luo J, Shi R, et al. Characteristics and prognosis of synchronous multiple primary lung cancer after surgical treatment: A systematic review and meta-analysis of current evidence. Cancer Med 2021;10:507-20. [Crossref] [PubMed]
48. Watanabe T, Tanahashi M, Suzuki E, et al. Surgical treatment for synchronous multiple primary lung cancer: Is it possible to achieve both curability and preservation of the pulmonary function? Thorac Cancer 2021;12:2996-3004. [Crossref] [PubMed]
49. Bhaskarla A, Tang PC, Mashtare T, et al. Analysis of second primary lung cancers in the SEER database. J Surg Res 2010;162:1-6. [Crossref] [PubMed]
50. Barclay ME, Lyratzopoulos G, Walter FM, et al. Incidence of second and higher order smoking-related primary cancers following lung cancer: a population-based cohort study. Thorax 2019;74:466-72. [Crossref] [PubMed]
51. Chen K, Liu A, Wang C, et al. Multidisciplinary expert consensus on diagnosis and treatment of multiple lung cancers. Med 2025;6:100643. [Crossref] [PubMed]
52. Detterbeck FC, Jones DR, Kernstine KH, et al. Lung cancer. Special treatment issues. Chest 2003;123:244s-58s.
53. Dong H, Tian Y, Xin S, et al. Diagnosis and management of multiple primary lung cancer. Front Oncol 2024;14:1392969. [Crossref] [PubMed]
54. Yang Z, Zhou B, Guo W, et al. Genomic characteristics and immune landscape of super multiple primary lung cancer. EBioMedicine 2024;101:105019. [Crossref] [PubMed]
55. Liang Z, Zeng G, Wan W, et al. The Unique Genetic Mutation Characteristics Based on Large Panel Next-Generation Sequencing (NGS) Detection in Multiple Primary Lung Cancers (MPLC) Patients. Discov Med 2023;35:131-43. [Crossref] [PubMed]
56. Lin Y, Li D, Hui H, et al. Genomic landscape and tumor mutational features of resected preinvasive to invasive lung adenocarcinoma. Front Oncol 2024;14:1389618. [Crossref] [PubMed]
57. Chang JC, Alex D, Bott M, et al. Comprehensive Next-Generation Sequencing Unambiguously Distinguishes Separate Primary Lung Carcinomas From Intrapulmonary Metastases: Comparison with Standard Histopathologic Approach. Clin Cancer Res 2019;25:7113-25. [Crossref] [PubMed]
58. Vincenten JPL, van Essen HF, Lissenberg-Witte BI, et al. Clonality analysis of pulmonary tumors by genome-wide copy number profiling. PLoS One 2019;14:e0223827. [Crossref] [PubMed]
59. Wang Z, Yuan X, Sun K, et al. Optimizing the NGS-based discrimination of multiple lung cancers from the perspective of evolution. NPJ Precis Oncol 2025;9:14. [Crossref] [PubMed]
60. Wang Z, Zhang Q, Wang C, et al. Multiple primary lung cancer: Updates and perspectives. Int J Cancer 2024;155:785-99. [Crossref] [PubMed]
61. Ni CH, Wang MT, Lin PH, et al. Different prognosis of multiple lung cancer identified by 116-gene panel by next-generation sequencing based on an Asian population. Transl Lung Cancer Res 2025;14:1699-714. [Crossref] [PubMed]
62. Sano A, Kage H, Sugimoto K, et al. A second-generation profiling system for quantitative methylation analysis of multiple gene promoters: application to lung cancer. Oncogene 2007;26:6518-25. [Crossref] [PubMed]
63. Liu Z, Wang L, Gao S, et al. Plasma metabolomics study in screening and differential diagnosis of multiple primary lung cancer. Int J Surg 2023;109:297-312. [Crossref] [PubMed]
64. Cheng H, Guo Z, Zhang X, et al. Lack of evolutionary convergence in multiple primary lung cancer suggests insufficient specificity of personalized therapy. J Genet Genomics 2023;50:330-40. [Crossref] [PubMed]

(English Language Editor: J. Gray)