Molecular characteristics and metastatic mechanism of patients diagnosed with Krukenberg tumor from gastric cancer

Introduction

Gastric cancer (GC) poses a major global health challenge, ranking fifth in incidence worldwide, with metastasis and high recurrence rates contributing substantially to its poor prognosis (1). Common metastatic sites for GC include the lungs, liver and bones; notably, the ovaries represent a major site of metastasis in advanced-stage female patients (2,3). Ovarian metastases originating from gastrointestinal primaries, known as Krukenberg tumors (KTs), were first described by pathologist Friedrich Ernst Krukenberg.

KTs frequently exhibit bilateral ovarian involvement (4), which is more prevalent in tumors of GC origin (77%) compared to those arising from colorectal cancer (50%) (5). These metastases predominantly occur in young or middle-aged premenopausal women, and a significant proportion of patients are asymptomatic at diagnosis, often detected incidentally during routine examinations (6). The common pathological types of primary gastric tumors associated with KTs include poorly differentiated adenocarcinoma and signet ring cell carcinoma (3,7). GC with ovarian metastasis presents a complex clinical scenario with no established standard of care, and the prognosis remains poor, with a median survival of only 8 to 14 months, worse than that of ovarian metastases from other gastrointestinal tumors (8). Thus, it is urgent to elucidate the molecular characteristics and facilitate the development of therapies for this disease.

KTs typically present as cystic or cystic-solid ovarian masses with well-defined margins and lobulated contours (6), but these features are nonspecific. Although positron emission tomography/computed tomography (PET/CT) demonstrates greater sensitivity than non-contrast-enhanced CT in detecting ovarian lesions, it remains inadequate for reliably distinguishing ovarian metastases from primary ovarian neoplasms (9). In up to 38% of cases, ovarian metastases are identified before the primary tumor (10), highlighting the importance of accurately recognizing metastatic lesions to guide appropriate treatment strategies.

Studies investigating the mutation pattern of KTs are limited. Although several routes have been proposed to explain the formation of KTs, including lymphatic spread, peritoneal spread and hematogenous diffusion, the underlying molecular mechanism remains unclear (11,12). Previous studies have suggested that tumor pathogenesis and progression may be related to genetic heterogeneity or oncogenes such as TP53 (13). Understanding the concordance between primary tumors and their metastases is crucial for selecting optimal treatment strategies. Specifically, targeted therapies directed against variants present in the primary tumor but absent in the metastases will be ineffective. However, studies examining the overlap and divergence of mutational profiles between primary tumors and matched metastases at specific sites are still lacking. Therefore, elucidating the molecular mechanisms driving ovarian metastasis in GC is essential for improving diagnosis, prognosis, and therapeutic strategies.

In this study, we present a comprehensive molecular landscape of ovarian metastasis from GC, derived from whole-exome sequencing (WES) of paired primary gastric and metastatic ovarian samples. These findings illuminate the mutational characteristics of KTs and underscore the molecular heterogeneity between primary gastric and metastatic ovarian lesions, intending to contribute to the clinical management of this highly malignant tumor. We present this article in accordance with the REMARK reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1008/rc).

Methods

Patients’ selection

This study is a single-center retrospective study. A total of 11 patients diagnosed with KTs from GC treated in Beijing Friendship Hospital, Capital Medical University from 2013 to 2020 were collected. The inclusion criteria were: (I) the primary tumor was pathologically confirmed as GC by two board-certified pathologists; (II) the ovarian tumor was pathologically confirmed as ovarian metastasis from GC by two board-certified pathologists; (III) the patients did not receive preoperative chemotherapy, radiotherapy, or other anti-tumor treatment; (IV) the patients had no other primary tumors. Clinical and pathological data were collected for all included patients, including age at diagnosis, menstrual status, main presenting symptoms, preoperative serum tumor markers, primary tumor site, primary tumor pathology, tumor stage, number of metastatic lymph nodes, and information related to metastatic lesions. Patients were followed up after surgery through telephone calls monthly and outpatient reviews every 3 months. The cutoff date for follow-up was February 2021, and overall survival (OS) was recorded for each patient. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Beijing Friendship Hospital, Capital Medical University (No. 2020-P2-167-01) and informed consent was obtained from all the patients themselves or their families.

DNA extraction

All surgical specimens were independently reviewed by two board-certified pathologists blinded to clinical outcomes, and diagnostic consensus was achieved for all cases. Lymph node metastasis samples were excluded from downstream analyses because most archived specimens did not meet WES quality control criteria. Genomic DNA from formalin-fixed paraffin-embedded (FFPE) tumor samples and matched paracancerous tissues was isolated using QIAamp DNA FFPE Kit (Qiagen, Hilden, Germany) following the manufacturer’s protocol. The concentration and purity of DNA were quantified in Qubit 2.0 (Life Technologies, Carlsbad, CA, USA) and Agilent 2100 Bioanalyzer Instrument (Agilent Technologies, Santa Clara, CA, USA). DNA samples were fragmented for 200–250 bp peak size by sonication (Covaris M220, Waltham, MA, USA) before library construction.

Library preparation and sequencing

WES library preparation was performed using the SeqCap EZ MedExome Target Enrichment Kit (Roche Diagnostics, Basel, Switzerland) according to manufacturer’s instructions. The library preparations were sequenced on Illumina NovaSeq 6000 platform (Illumina, San Diego, CA, USA), and 150 bp paired-end reads were generated.

Data filtering and variant calling

Quality control statistics and pre-processing of raw sequence data were performed by fastp (version 0.19.1). Reads were aligned to the hg19 version of the human genome using Burrows-Wheeler Aligner software (BWA, version 0.7.12). Polymerase chain reaction (PCR) duplicates were marked using the MarkDuplicates tool in Picard. IndelRealigner and BaseRecalibrator on Genome Analysis Toolkit (GATK; version 3.8) were used to realign and recalibrate the BWA alignment results, respectively. Varscan was used to identify paired sample variant calling of single-nucleotide variations (SNVs) and insertions or deletions (INDELs) on the tumor and matched normal samples. All variants were annotated using Annovar. To ensure the quality of data, the following criteria were used to filter raw variant results: allele mutation frequency ≥10% for SNVs and INDELs; all reads were filtered by high mapping quality (≥30) and base quality (≥30).

Mutational signature

Both synonymous and non-synonymous somatic SNVs were analyzed to define mutational signatures including six categories of base substitutions in each included sample, including T>A, T>C, T>G, C>A, C>G and C>T. In view of the 5' and 3' flanking nucleotides of a specific mutant base, a total of 96 substitution types exist. We extracted the potential mutational signatures in each sample by using the 30 signatures documented by the Catalogue of Somatic Mutations in Cancer (COSMIC) as a reference (14).

Copy number variation (CNV)

CNV analysis was performed using CONTRA (version 2.1.0), which indicated CNV gain or loss for genes within the panel coverage. The software computes regions per reading and calculates the likely possibility based on dispersion measurements and coverage ratios. The hidden Markov model was then used to calculate a CNV classification. The ratios for each region given were 3.5 for copy number gain and 0.5 for copy number loss.

Tumor mutational burden (TMB) definition

The TMB was defined as the number of somatic, coding, and indel mutations and base substitutions per megabase of the genome examined. The cutoff value for the TMB in our cohorts was defined as the median TMB.

Phylogenetic tree construction

Multi-sample cell lineage trees were reconstructed by LICHeE (15) and the subclonal composition of each sample was inferred by using variant allele frequencies of somatic SNVs. The lineage tree of the somatic SNV clusters was built based on the constraint network (15,16).

Statistical analysis

Statistical analyses were performed using SPSS 25 and R 4.2. Continuous variables were compared using paired t-tests or Wilcoxon signed-rank tests for primary-metastatic pairs and Mann-Whitney U or Kruskal-Wallis tests for unpaired groups. Categorical variables were analyzed using Chi-squared or Fisher’s exact tests. OS was estimated by the Kaplan-Meier method and compared using log-rank tests. TMB was analyzed as a continuous variable and dichotomized at the cohort median for categorical analyses. Missing data were rare and handled by complete-case analysis. Two-sided P<0.05 was considered statistically significant.

Results

Baseline characteristics and treatment of patients

Table 1 summarizes the clinical and pathological features of 11 patients. The median age at GC diagnosis was 40 years (range, 26–70 years). In 7 cases (63.6%), the ovarian metastases were synchronously present at the time of GC diagnosis. In the other 4 cases, the ovarian metastases were diagnosed on median of 14.5 months later. Among them, 8 patients (72.3%) were premenopausal. The histopathological subtype distribution included poorly differentiated adenocarcinoma or signet ring cell carcinoma (81.81%, 9/11) and moderately differentiated adenocarcinoma (18.18%, 2/11). Bilateral ovarian metastases were found in 6 patients (54.5%), and 8 cases had at least 6 metastatic lymph nodes. Representative histopathological images of primary GCs and their matched ovarian metastases stained with hematoxylin and eosin (H&E) are presented in Figure 1.

Figure 1 Histopathological images (A,B) of primary gastric cancers and matched ovarian metastases (H&E). H&E, hematoxylin and eosin.

Nine patients underwent radical resection of the primary tumor and ovarian metastases, and two cases only underwent laparotomy and radical resection of ovarian metastases. Ten patients received postoperative chemotherapy. By the end of follow-up, the median OS time of these patients was 17 (range, 6–67) months (Figure 2).

Figure 2 The survival curve of patients grouped by treatment.

Mutational profile of primary gastric tumor and ovarian metastasis

We successfully conducted WES in 22 samples (11 primary gastric tumors and 11 metastatic ovarian tumors) from 11 patients with a median depth of 174× (101–371×). A total of 3,007 somatic mutations were identified in all primary gastric tumors (1,519 mutations) and metastatic ovarian tumors (1,488 mutations). TGFBR2 p.R553H, CDH1 p.I192N (patient 9) and RAD50 p.E431fs (patient 10) pathogenic germline variants were detected in 2 patients. TMB was generally low in this cohort, with the exception of patient 9, who showed elevated TMB in both primary and metastatic lesions.

We featured 1,488 SNVs, INDELs and 94 CNVs in KTs. The most commonly mutated cancer-related genes in KTs were TP53 (54.5%), followed by EIF1AX (45.5%), CCNE1 (36.4%) and MYC (36.4%, Figure 3A,3B). We also noted that 40.0% (2/5) of KTs with EIF1AX mutations were accompanied by TP53 alterations (Figure 3A). FATx, PTEN and POLE were rarely detected in metastatic lesions, while MYC and CCNE1 were more commonly found in KTs (all P>0.05, Figure 3C). In addition, KTs had significantly more CNVs than GCs (P=0.02, Figure 3D), and metastases were more prone to MYC and CCNE1 amplification (Figure 3B), implying large range of copy-number variation events may be a later genomic event during progression of ovarian metastasis in gastric carcinoma. Furthermore, receptor tyrosine kinase (RTK) gene amplifications were detected in ovarian metastases from five cases, including 1 case with EGFR-c-MET co-amplification and 4 cases with only one RTK genes amplification (ERBB2, n=2; EGFR, n=1; FGFR2, n=1, Figure 3B). Taken together, RTK genes amplification patterns in KTs were mainly mutually exclusive, which was consistent with previous reports (13).

Figure 3 The mutational landscape of primary GCs and paired KTs. Heatmap showing the top 30 genes with the highest mutation frequencies across 22 samples. Genes are sorted by mutation frequency. (A) Identified mutational landscape of primary lesions and matched ovarian metastases, including SNVs and INDELs. Top panel: the number of somatic mutations in each tumor. Left panel: the list of mutated genes. Middle panel: the matrix of mutations in a selection of frequently mutated genes. Columns represent samples. (B) The CNV events of the KTs and GCs. Top panel: the number of CNVs in each tumor. Left panel: the genes. Middle panel: the matrix of CNV in a selection of frequently mutated genes, including CNV gain (red) and CNV loss (blue). Columns represent samples. (C) Comparison of gene mutation frequencies between KTs and GCs, including SNVs, INDELs and CNVs. (D) Comparison of the number of CNVs between KTs and GCs. (E,F) The relative fraction of six SNV types of mutation spectra. CNV, copy number variation; GC, gastric cancer; INDEL, insertions or deletions; KT, Krukenberg tumor; SNV, single-nucleotide variant.

The mutational patterns of SNVs in ovarian metastasis were shown in Figure 3E. C>T mutations accounted for the highest proportion (59.7%) of the six basic transformation types. In contrast, the C>A transversion associated with environmental oncogenic effects was low. We further found that mutational signature 1 (associated with aging) and signature 6 (defective DNA mismatch repair-associated) were the most predominant in KTs, showing that the mutational signatures correlated with age at diagnosis and DNA mismatch repair deficiency (Figure 3F).

Mutational profile and clinical features in KTs

We further explored the relationship between clinical characteristics and somatic mutational profiles in patients with KTs. Stratified analysis revealed that TP53 mutations were inversely correlated with premenopausal status, whereas mutations in ARID1A and CCNE1 showed a positive association, with a higher prevalence in premenopausal patients (Figure 4A). In terms of metastatic patterns, ARID1A mutations were more commonly observed in patients with bilateral metastases, while TP53 mutations were enriched in those with unilateral involvement (Figure 4B). Moreover, metachronous metastases were characterized by a higher frequency of TP53, MYC, and RHOA mutations, whereas ERBB2 mutations were predominantly found in synchronous metastases (Figure 4C). Survival analysis indicated that TP53 and FGFR3 mutations correlated with inferior OS (<24 months), whereas PCLO mutations were associated with relatively prolonged survival (≥24 months) (Figure 4D). In addition, a higher CNV burden and TP53 mutation status showed a potential trend toward worse outcomes, although these associations did not reach statistical significance [CNV burden: P=0.14, hazard ratio (HR) =0.46, 95% confidence interval (CI): 0.12–1.72; TP53 mutation: P=0.16, HR =2.03, 95% CI: 0.60–6.88] (Figure 4E,4F).

Figure 4 Associations between somatic mutations and clinical characteristics in KTs of gastric origin. Volcano plots illustrate the relationships between gene mutations and clinical features, with the x-axis representing log₂ ORs and the y-axis representing −log₂ P values. The horizontal blue dashed line marks the significance threshold (P=0.05); genes above the line are considered statistically significant. (A) Mutation enrichment by menopausal status, with positive ORs indicating a higher prevalence in premenopausal patients. (B) Associations between gene mutations and metastatic laterality (unilateral vs. bilateral ovarian involvement). (C) Differences in mutation patterns between synchronous and metachronous metastases. (D) Gene mutations are associated with overall survival, with patients classified as having poor (<24 months) or good (≥24 months) outcomes. (E,F) Kaplan-Meier curves showing survival stratified by CNV burden (E) and TP53 mutation status (F); P values were calculated using the log-rank test, with no statistically significant differences observed. CNV, copy number variation; KT, Krukenberg tumor; MT, mutant type; OR, odds ratio; WT, wild type.

Consistency and discrepancy between matched GC and KT

For these genomes, TP53, EIF1AX, FBXW7 mutations and MYC-amplification were highly consistent between matched primary tumors and KTs (Figure 5A), indicating that they were predominantly key events that promote ovarian tumor metastasis. Only 22.1% (195/881) of KTs mutations were shared with primary GC in 10 patients, excluding patient no.11 with 636 co-mutations. To quantify intertumoral heterogeneity between KTs and GCs, we investigated “shared” (genetic variants present in both primary and metastatic sites) and “private” (variants present only in primary or metastatic sites) mutations based on previous studies (17). We observed that only a median of 6.3% (range, 1.2–83.3%) of mutations were shared, suggesting a high and variable degree of intertumoral heterogeneity between matched KTs and GC (Figure 5B). We inferred mutation-based evolutionary relationship of multiple biopsies from the same individuals with primary GC and metastatic KTs genomes available. In all cases, after sharing a period of clonality, the KTs and primary tumor seemed to have individual evolution patterns afterwards based on the phylogenetic tree (Figure 5C). This suggested that the parallel progression model of KTs may be predominant, as clones with KT-competent differentiated from matched primary tumor at a relatively early stage. Furthermore, most metastatic lesions (72.7%, 8/11) appeared to have more somatic mutations than the corresponding primary tumors (Figure 5A), suggesting that more mutations accumulated under evolutionary pressure during KTs formation.

Figure 5 Genetic heterogeneity between primary gastric cancer lesions and matched ovarian metastases. (A) Mutational landscape of primary tumors and corresponding ovarian metastases. (B) Proportion of shared (ubiquitous) and private (unique) mutations in each paired case. (C) Phylogenetic trees reconstructed for each patient based on somatic mutations. The fractions of shared and private mutations illustrate the evolutionary relationships between primary GC and KT. GC, gastric cancer; KT, Krukenberg tumor.

Discussion

Ovarian metastasis from GC remains a significant clinical challenge worldwide owing to its difficulty in early detection, poor prognosis, unclear metastatic mechanisms, and the absence of standardized and effective treatment protocols. To the best of our knowledge, this is the first study to systematically characterize both the clinical and mutational landscape of KT from GC and to elucidate the genetic mechanisms underlying their metastasis. Our findings revealed that TP53 and EIF1AX were the most commonly mutated genes in both KTs and GC, while CNVs exhibited marked differences. The genetic heterogeneity was observed between the primary tumors and the matched ovarian metastases. Phylogenetic analysis further demonstrated that KT-competent clones genetically diverged from their primary tumors at an early stage, supporting a parallel progression model of metastasis. These results highlight distinct molecular features and a unique evolutionary trajectory in the development of KTs from GC, providing new insights into their pathogenesis and potential therapeutic targets.

The median survival of patients with KTs has been reported to range from 7 to 14 months, consistent with the findings of our study (3,18). Various treatment strategies, including cytoreductive surgery (CRS), chemotherapy, and/or hyperthermic intraperitoneal chemotherapy (HIPEC), have been proposed (11). However, it remains unclear which approach significantly improves patient outcomes, and the absence of standardized, effective treatment guidelines may further contribute to the poor prognosis associated with KTs. Limited available studies have suggested that CRS, particularly complete (residual-free) CRS, is associated with improved survival, and that HIPEC may provide additional benefits when used alone or in combination with CRS (19). In our cohort, the three patients with the longest OS all underwent residual-free CRS followed by postoperative chemotherapy for both primary and metastatic lesions. By contrast, two patients who received CRS limited to metastatic sites with postoperative chemotherapy exhibited shorter survival times. However, these differences did not reach statistical significance, likely due to the small sample size.

Our genomic analysis reveals several distinctive molecular features of KTs, which have important clinical implications. Compared to a previously published GCs cohort from Memorial Sloan Kettering Cancer Center (MSKCC) (20), our cohort exhibited a higher frequency of mutations in TP53 (72.7% vs. 47.6%), EIF1AX (36.4% vs. 0%) and FAT1 (36.4% vs. 6.1%), along with a lower frequency of TTN and SYNE1 mutations in primary tumors. These findings suggest a unique mutational landscape that may be associated with ovarian metastasis. We identified recurrent mutations in TP53 and EIF1AX in KTs, with substantial concordance to those in matched GCs, indicating that these mutations may serve as potential driver events in KT pathogenesis. TP53 mutations have been reported to correlate with poor histological differentiation, vascular invasion, and unfavorable prognosis in GCs (21), particularly in the Asian population (22). Moreover, gene sequencing analyses of GCs and matched positive lymph nodes demonstrated that TP53 mutations may contribute to early lymphatic dissemination (23). In other malignancies, EIF1AX mutations have been shown to co-occur with RAS pathway alterations, promoting tumor progression. For instance, EIF1AX mutations frequently co-exist with RAS mutations in thyroid cancer (24), while in low-grade serous ovarian carcinoma, co-expression of mutant NRAS and EIF1AX enhances clonal survival and proliferation (25). A case report of Hürthle cell carcinoma further suggested a potential driver role of EIF1AX when combined with TP53 mutation (26). Retrospective studies in thyroid tumors reported that EIF1AX mutations are enriched in poorly differentiated thyroid carcinomas (PDACs), particularly when co-occurring with RAS or TP53 alterations, and that co-mutation with TP53 is associated with a higher risk of malignancy (27). Another analysis similarly noted that concurrent TP53 may increase aggressiveness in EIF1AX-mutated thyroid tumors, supporting a potential synergistic effect (28). In our cohort, 40.0% of patients with EIF1AX mutations also had concurrent TP53 mutations. Based on these observations, we hypothesize that TP53 and EIF1AX mutations may jointly drive ovarian metastasis and confer aggressive tumor behavior. However, this remains a hypothesis and requires functional validation.

Beyond somatic mutations, KTs demonstrated a higher CNV burden compared to primary GCs, reflecting elevated genomic instability. Ovarian cancer metastases commonly exhibit amplifications of MYC, CCNE1, ERBB2, and CDK6, as well as deletions of CDKN2A and SMAD4. A pattern of mutually exclusive RTK gene amplifications, most notably involving ERBB2, FGFR, and MET was observed, consistent with previous reports (13,20). Although statistical power is limited by cohort size, the recurrent nature of these amplifications implicates RTK-driven signaling pathways in mediating ovarian tropism and highlights their potential as actionable targets for metastasis-directed therapies. Comparative analysis of synchronous versus metachronous ovarian metastases revealed broadly comparable mutational landscapes; however, differential frequencies of MYC, RHOA and ERBB2 mutations suggest subtle genomic divergence that may contribute to the distinct clinical trajectories observed between these subtypes (29). Importantly, patients harboring TP53 mutations or exhibiting a high burden of CNVs showed a trend toward shorter OS; however, these associations did not reach statistical significance in our cohort, indicating that both may have potential prognostic relevance in KTs, but this observation should be interpreted cautiously and requires validation in larger studies. Furthermore, the detection of recurrent EIF1AX mutations raises the possibility of this gene serving as a metastasis-specific marker, meriting further investigation. Taken together, these data underscore the biological and clinical relevance of direct molecular profiling of metastatic lesions, advocating against sole reliance on the genomic features of primary tumors for therapeutic decision-making.

Several routes have been proposed to explain the formation of KTs, including peritoneal seeding, hematogenous dissemination, and lymphatic spread. In our cohort, only one patient exhibited serosal invasion of the primary tumor, whereas 72.3% of patients presented with intravascular tumor thrombi, suggesting that hematogenous dissemination may represent a predominant route for ovarian metastasis. Nevertheless, lymphatic involvement also appears to play an important contributory role. Previous studies have demonstrated that the risk of ovarian metastasis from GC positively correlates with the extent of lymph node involvement (30). Consistent with this observation, 88.9% (8/9) of patients in our cohort had metastases involving more than six lymph nodes. Together, these findings indicate that the development of KTs is likely driven by multiple metastatic routes rather than a single pathway, underscoring the biological complexity underlying the origin and evolution of distant metastases.

Intertumoral heterogeneity drives variability in tumor growth rate, invasion ability, drug response, and prognosis (31,32). Metastatic heterogeneity, reflecting differences in metastatic potential among subclones within the same tumor, arises from dynamic genetic alterations acquired during tumor evolution and may underlie the preferential colonization of distinct metastatic sites. Consistent with this concept, single-cell RNA sequencing studies of gastric adenocarcinoma have revealed marked cellular heterogeneity and distinct molecular subtypes during disease progression (33,34). In the present study, although partial genetic convergence was observed, substantial genetic heterogeneity persisted between primary GCs and ovarian metastases across patients. Phylogenetic analyses further demonstrated a predominance of a parallel progression pattern in KT development, indicating that ovarian metastases likely originate from early-disseminated subclones that evolve independently of the primary tumor. Notably, the genetic divergence between primary GCs and KTs exceeded that previously reported between primary tumors and lymph node metastases (23,35), suggesting that ovarian metastasis may follow an evolutionary trajectory distinct from lymphatic dissemination. In this context, lymph node metastases may predominantly reflect stepwise clonal selection and linear evolution, whereas ovarian metastases appear to arise through early systemic spread and parallel evolution. Importantly, these findings do not imply mutually exclusive metastatic models, but rather support a site-dependent evolutionary framework in GC metastasis. Different metastatic routes, such as lymphatic and hematogenous dissemination, may be governed by distinct biological constraints and selective pressures, resulting in heterogeneous evolutionary patterns. Interestingly, a previous study reported that patients exhibiting a parallel evolution pattern were more sensitive to paclitaxel-based chemotherapy and had a more favorable prognosis (36). Therefore, evolutionary classification may provide clinically relevant insights for prognostic stratification and therapeutic decision-making in GC patients with ovarian metastases.

Given the distinct genomic and evolutionary features of KTs, molecular profiling of ovarian metastases may provide clinically relevant information beyond that obtained from primary gastric tumors, but primarily in selected clinical contexts. Sequencing of ovarian lesions is most informative in patients with synchronous or early-onset ovarian metastasis, discordant responses between primary and ovarian tumors, or disease recurrence or progression after systemic therapy, when site-specific resistance mechanisms may emerge. Genomic features of the primary tumor, particularly alterations associated with aggressive behavior such as TP53 mutation, MYC amplification, or dysregulation of RTK pathways, may help identify patients at increased risk of ovarian dissemination who are more likely to benefit from metastatic profiling. In these settings, analysis of ovarian metastatic tissue may reveal prognostically or therapeutically relevant alterations that are not fully represented in the primary tumor and may inform subsequent systemic treatment selection, including following CRS. By contrast, in patients lacking high-risk molecular features and showing concordant disease behavior, additional sequencing of ovarian metastases is unlikely to meaningfully influence clinical management.

There are some limitations in this study. First, this was a single-center study with a small sample size due to the rarity of the disease. The identified genetic alterations need to be fully validated in larger multi-center cohorts. Moreover, the analysis was confined to primary tumors and ovarian metastases, without inclusion of lymph node metastases, potentially limiting the comprehensiveness of the metastatic landscape. Functional validation of key mutations, particularly TP53 and EIF1AX, was not performed; therefore, their pathogenic roles remain hypothetical. Although prior evidence from other tumor types supports their possible involvement, further mechanistic studies are essential to confirm their biological significance. Nevertheless, the study provides valuable insights with potential translational relevance.

Conclusions

In conclusion, we identified key molecular alterations potentially contributing to the ovarian metastasis of GC. The distinct mutational profiles observed between primary tumors and their matched ovarian metastases suggest a parallel progression model driven by early clonal divergence. These findings underscore the necessity of considering metastasis-specific genomic features when devising therapeutic strategies. Therefore, tailored treatments targeting KTs should be actively explored and optimized. Our study enhances the current understanding of the unique biology of KTs and provides a molecular basis for the development of more effective, site-specific therapeutic approaches. However, given the limited sample size of this study, these findings should be considered preliminary and require validation in larger, multi-center cohorts.

Acknowledgments

The authors wish to thank all the patients that participated in this study.

Footnote

Reporting Checklist: The authors have completed the REMARK reporting checklist. Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1008/rc

Data Sharing Statement: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1008/dss

Peer Review File: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1008/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1008/coif). M.L., X.M., H.W. and J.Q. are employees of Acornmed Biotechnology Co., Ltd. The other 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 Beijing Friendship Hospital, Capital Medical University (No. 2020-P2-167-01) and informed consent was obtained from all the patients themselves or their families.

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. Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024;74:229-63. [Crossref] [PubMed]
2. Huang X, Li Z, Weng Q. Clinicopathological features and prognostic significance of site-specific metastasis in gastric cancer: a population-based, propensity score-matched analysis. Discov Oncol 2025;16:1164. [Crossref] [PubMed]
3. Wang L, Fu T, Chen Y, et al. Mapping the landscape of ovarian metastases of gastric cancer: insights, trends, and emerging perspectives. World J Surg Oncol 2025;23:450. [Crossref] [PubMed]
4. Wu F, Zhao X, Mi B, et al. Clinical characteristics and prognostic analysis of Krukenberg tumor. Mol Clin Oncol 2015;3:1323-8. [Crossref] [PubMed]
5. Lobo J, Machado B, Vieira R, et al. The challenge of diagnosing a malignancy metastatic to the ovary: clinicopathological characteristics vary and morphology can be different from that of the corresponding primary tumor. Virchows Arch 2017;470:69-80. [Crossref] [PubMed]
6. Zulfiqar M, Koen J, Nougaret S, et al. Krukenberg Tumors: Update on Imaging and Clinical Features. AJR Am J Roentgenol 2020;215:1020-9. [Crossref] [PubMed]
7. Lin X, Han T, Zhuo M, et al. A retrospective study of clinicopathological characteristics and prognostic factors of Krukenberg tumor with gastric origin. J Gastrointest Oncol 2022;13:1022-34. [Crossref] [PubMed]
8. Ma F, Li Y, Li W, et al. Metastasectomy Improves the Survival of Gastric Cancer Patients with Krukenberg Tumors: A Retrospective Analysis of 182 patients. Cancer Manag Res 2019;11:10573-80. [Crossref] [PubMed]
9. Willmott F, Allouni KA, Rockall A. Radiological manifestations of metastasis to the ovary. J Clin Pathol 2012;65:585-90. [Crossref] [PubMed]
10. Crobach S, Ruano D, van Eijk R, et al. Somatic mutation profiles in primary colorectal cancers and matching ovarian metastases: Identification of driver and passenger mutations. J Pathol Clin Res 2016;2:166-74. [Crossref] [PubMed]
11. Wu SJ, Wu CY, Ye K. Strategies for the comprehensive treatment of gastric cancer ovarian metastasis. World J Clin Oncol 2025;16:106589. [Crossref] [PubMed]
12. Zhang M, Chen G, Lin X, et al. Comprehensive exome profiling identifies ARHGEF12 mutation as a driver in gastric cancer with ovarian metastasis. Theranostics 2025;15:8202-21. [Crossref] [PubMed]
13. Wang B, Tang Q, Xu L, et al. A comparative study of RTK gene status between primary tumors, lymph-node metastases, and Krukenberg tumors. Mod Pathol 2021;34:42-50. [Crossref] [PubMed]
14. Blokzijl F, Janssen R, van Boxtel R, et al. MutationalPatterns: comprehensive genome-wide analysis of mutational processes. Genome Med 2018;10:33. [Crossref] [PubMed]
15. Popic V, Salari R, Hajirasouliha I, et al. Fast and scalable inference of multi-sample cancer lineages. Genome Biol 2015;16:91. [Crossref] [PubMed]
16. Ricketts C, Popic V, Toosi H, et al. Using LICHeE and BAMSE for Reconstructing Cancer Phylogenetic Trees. Curr Protoc Bioinformatics 2018;62:e49. [Crossref] [PubMed]
17. Jamal-Hanjani M, Wilson GA, McGranahan N, et al. Tracking the Evolution of Non-Small-Cell Lung Cancer. N Engl J Med 2017;376:2109-21. [Crossref] [PubMed]
18. Lionetti R, DE, Luca M, Raffone A, et al. Clinics and pathology of Krukenberg tumor: a systematic review and meta-analysis. Minerva Obstet Gynecol 2022;74:356-63. [Crossref] [PubMed]
19. Lionetti R, De Luca M, Travaglino A, et al. Treatments and overall survival in patients with Krukenberg tumor. Arch Gynecol Obstet 2019;300:15-23. [Crossref] [PubMed]
20. Guo YA, Chang MM, Huang W, et al. Mutation hotspots at CTCF binding sites coupled to chromosomal instability in gastrointestinal cancers. Nat Commun 2018;9:1520. [Crossref] [PubMed]
21. Wang K, Yuen ST, Xu J, et al. Whole-genome sequencing and comprehensive molecular profiling identify new driver mutations in gastric cancer. Nat Genet 2014;46:573-82. [Crossref] [PubMed]
22. Wang J, Shao X, Liu Y, et al. Mutations of key driver genes in gastric cancer metastasis risk: a systematic review and meta-analysis. Expert Rev Mol Diagn 2021;21:963-72. [Crossref] [PubMed]
23. Lee HH, Kim SY, Jung ES, et al. Mutation heterogeneity between primary gastric cancers and their matched lymph node metastases. Gastric Cancer 2019;22:323-34. [Crossref] [PubMed]
24. Krishnamoorthy GP, Davidson NR, Leach SD, et al. EIF1AX and RAS Mutations Cooperate to Drive Thyroid Tumorigenesis through ATF4 and c-MYC. Cancer Discov 2019;9:264-81. [Crossref] [PubMed]
25. Etemadmoghadam D, Azar WJ, Lei Y, et al. EIF1AX and NRAS Mutations Co-occur and Cooperate in Low-Grade Serous Ovarian Carcinomas. Cancer Res 2017;77:4268-78. [Crossref] [PubMed]
26. Topf MC, Wang ZX, Furlong K, et al. EIF1AX Mutation in a Patient with Hürthle Cell Carcinoma. Endocr Pathol 2018;29:27-9. [Crossref] [PubMed]
27. Elsherbini N, Kim DH, Payne RJ, et al. EIF1AX mutation in thyroid tumors: a retrospective analysis of cytology, histopathology and co-mutation profiles. J Otolaryngol Head Neck Surg 2022;51:43. [Crossref] [PubMed]
28. Bandargal S, Chen T, Pusztaszeri MP, et al. Prognostic Indicators of EIF1AX-Mutated Thyroid Tumor Malignancy and Cancer Aggressiveness. Cancers (Basel) 2022;14:6097. [Crossref] [PubMed]
29. Rosa F, Marrelli D, Morgagni P, et al. Krukenberg Tumors of Gastric Origin: The Rationale of Surgical Resection and Perioperative Treatments in a Multicenter Western Experience. World J Surg 2016;40:921-8. [Crossref] [PubMed]
30. Feng Q, Pei W, Zheng ZX, et al. Clinicopathologic characteristics and prognostic factors of 63 gastric cancer patients with metachronous ovarian metastasis. Cancer Biol Med 2013;10:86-91. [PubMed]
31. Gullo I, Carneiro F, Oliveira C, et al. Heterogeneity in Gastric Cancer: From Pure Morphology to Molecular Classifications. Pathobiology 2018;85:50-63. [Crossref] [PubMed]
32. Craig AJ, von Felden J, Garcia-Lezana T, et al. Tumour evolution in hepatocellular carcinoma. Nat Rev Gastroenterol Hepatol 2020;17:139-52. [Crossref] [PubMed]
33. Jiang H, Yu D, Yang P, et al. Revealing the transcriptional heterogeneity of organ-specific metastasis in human gastric cancer using single-cell RNA Sequencing. Clin Transl Med 2022;12:e730. [Crossref] [PubMed]
34. Yasuda T, Wang YA. Gastric cancer immunosuppressive microenvironment heterogeneity: implications for therapy development. Trends Cancer 2024;10:627-42. [Crossref] [PubMed]
35. Sundar R, Liu DH, Hutchins GG, et al. Spatial profiling of gastric cancer patient-matched primary and locoregional metastases reveals principles of tumour dissemination. Gut 2021;70:1823-32. [Crossref] [PubMed]
36. Yu P, Hu C, Ding G, et al. Mutation characteristics and molecular evolution of ovarian metastasis from gastric cancer and potential biomarkers for paclitaxel treatment. Nat Commun 2024;15:3771. [Crossref] [PubMed]