Presence of RB1 or Absence of LRP1B Mutation Predicts Poor Overall Survival in Patients with Gastric Neuroendocrine Carcinoma and Mixed Adenoneuroendocrine Carcinoma
Article information
Abstract
Purpose
Neuroendocrine carcinomas (NECs) of the stomach are extremely rare, but fatal. However, our understanding of the genetic alterations in gastric NECs is limited. We aimed to evaluate genomic and clinicopathological characteristics of gastric NECs and mixed adenoneuroendocrine carcinomas (MANECs).
Materials and Methods
Fourteen gastric NECs, three gastric MANECs, and 1,381 gastric adenocarcinomas were retrieved from the departmental next-generation sequencing database between 2017 and 2022. Clinicopathological parameters and next-generation sequencing test results were retrospectively collected and reviewed.
Results
Gastric NECs and MANECs frequently harbored alterations of TP53, RB1, SMARCA4, RICTOR, APC, TOP1, SLX4, EGFR, BRCA2, and TERT. In contrast, gastric adenocarcinomas exhibited alterations of TP53, CDH1, LRP1B, ARID1A, ERBB2, GNAS, CCNE1, NOTCH, and MYC. Mutations of AKT3, RB1, and SLX4; amplification of BRCA2 and RICTOR; and deletion of ADAMTS18, DDX11, KLRC3, KRAS, MAX, NFKBIA, NUDT7, and RB1 were significantly more frequent in gastric NECs and MANECs than in gastric adenocarcinomas. The presence of LRP1B mutation was significantly associated with longer overall survival (OS), whereas RB1 mutation and advanced TNM stage were associated with shorter OS.
Conclusion
We identified frequently mutated genes and potential predictors of survival in patients with gastric NECs and MANECs.
Introduction
The 2019 World Health Organization classification categorized neuroendocrine neoplasms (NENs) of the stomach into three distinct groups—neuroendocrine tumors, neuroendocrine carcinomas (NECs), and mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs)—based on histologic features, mitotic count, and the Ki-67 proliferation index [1,2]. Gastirc NECs, which account for 6%-16% of all gastric NENs, have a worse prognosis than typical adenocarcinomas because of their tendency to be diagnosed at advanced clinical stages and their early metastasis to the liver or lymph nodes [3]. Gastric NECs are characterized by the presence of disorganized trabecular structures or sheets of poorly differentiated epithelial cells that display neuroendocrine markers, and are further classified into small and large cell subtypes. Small cell NECs are characterized by cancer cells with scant cytoplasm, hyperchromatic nuclei, and no nucleoli. In contrast, large cell NECs are characterized by abundant eosinophilic cytoplasm, large vesicular nuclei, and prominent nucleoli. Mixed adenoneuroendocrine carcinomas (MANECs) are a type of MiNEN composed of both NEC and associated adenocarcinoma, with at least 30% of either component [1].
NECs that originate from various anatomical locations exhibit distinct genetic alterations. For example, gastroenteropancreatic NECs exhibit a higher frequency of KRAS and BRAF mutations, but TP53 and RB transcriptional corepressor 1 (RB1) mutations are less common than in pulmonary NECs [4]. Similarly, gastrointestinal NECs have shown higher frequency of APC and CTNNB1 mutations but a lower frequency of RB1 mutation than small cell lung cancers [5]. However, genomic analyses of gastric NECs remain limited, largely because of the rarity of this disease.
Next-generation sequencing (NGS) has recently been developed to enable clinicians and researchers to analyze a broad range of genomic and transcriptomic alterations. Since solid cancer NGS was implemented in clinical practice in the Republic of Korea in 2016, its use has rapidly expanded and a large amount of genetic data has been produced [6]. Therefore, in this study, we aimed to collect NGS data on gastric NECs and MANECs from the hospital database, analyze genetic alterations and clinicopathological data, and evaluate the prognostic significance of genetic alterations.
Materials and Methods
1. Patient collection
Between March 2017 and June 2022, 15,931 NGS tests for solid tumors were performed at Asan Medical Center. From the NGS database, we retrieved 14 NECs, 3 MANECs, and 1,381 gastric adenocarcinomas. An expert pathologist collected and reviewed clinical parameters and pathological reports, including age, sex, survival status, survival time, histological diagnosis, and biomarker status, from the database. Hematoxylin and eosin (H&E)–stained slides were reviewed to verify the diagnosis and identify whether the NEC components were of the small or large cell type (Fig. 1). In addition, we tested 23 cases of non-gastric extrapulmonary NECs: seven from the colorectum, three from the ampulla of Vater, three from the bile duct, two from the esophagus, two from the gallbladder, two from the liver, two from the prostate, and two from the uterus.
Representative microscopic images of gastric neuroendocrine carcinomas (NECs) and mixed adenoneuroendocrine carcinomas (MANECs). (A) Small cell gastric NEC exhibits compactly arranged tumor cells with hyperchromatic nuclei, little cytoplasm, and nuclear molding (H&E, ×600). The tumor cells are weakly positive for synaptophysin by immunohistochemistry and show Ki-67 labeling index of > 90%. (B) Large cell gastric NEC consists of tumor cells showing abundant cytoplasm and prominent nucleoli. Mitotic figures are frequently seen (arrows) (H&E, ×600). The tumor cells are diffusely positive for chromogranin A by immunohistochemistry and show Ki-67 labeling index of approximately 50%. (C) Gastric MANEC composed of two different components, poorly differentiated adenocarcinoma and small cell NEC. Poorly differentiated adenocarcinoma component exhibits poorly formed glandular structures and intracytoplasmic and extracellular mucin production. In contrast, small cell NEC component shows no glandular differentiation but solid arrangement of tumor cells with hyperchromatic nuclei, little cytoplasm, nuclear molding, and frequent mitotic figures (H&E, ×600). The NEC component shows diffuse insulinoma-associated protein 1 (INSM-1) immunolabelling, which supports neuroendocrine differentiation.
2. DNA extraction, library preparation, and targeted NGS
DNA was extracted as described previously [7]. Briefly, expert pathologists reviewed H&E-stained slides of tumor tissues in each test and confirmed whether the amount of tissue and tumor cellularity was sufficient for NGS. Samples were considered adequate when the size of the tissue was bigger than 0.5×0.5 cm and tumor cellularity exceeded 20%. For MANECs, both the adenocarcinoma and NEC components were included in each experiment. After deparaffinization, genomic DNA was extracted from formalin-fixed, paraffin-embedded tumor tissues using the NEXprep FFPE Tissue kit (#NexK-9000, Genes Laboratories). Extracted genomic DNA was fragmented by sonication (Covaris Inc.) and selected by size using Agencourt AMPure XP beads (Beckman Coulter).
Library construction was performed using the OncoPanel AMC v4.5 panel as described previously [8]. Briefly, OncoPanel AMC v4.5 is a DNA-based hybrid capture targeted gene panel of 1.2 Mbp targeting 343 genes. Detailed information about this panel is provided in S1 Table. Pooled libraries were loaded onto a NextSeq 550Dx Sequencing System (Illumina) for paired-end sequencing.
3. Bioinformatics analysis
Bioinformatics analysis was performed as described previously [9]. Briefly, sequenced reads were aligned to the human reference genome (GRCh37;hg19) using the Burrows-Wheeler Aligner and processed using the Genome Analysis Toolkit pipeline (v4.2.6.1).
Small-scale mutations (SSMs) were called using VarDict (v1.6) [10]. The following variants were filtered out: (1) total read depth < 30, mutated read counts < 3, and variant allele frequency < 3%, and (2) minor allele frequency > 1% in the common dbSNP build 141, Exome Aggregation Consortium release 0.3.1 (http://exac.broadinstitute.org), and Korean Reference Genome database (http://152.99.75.168/KRGDB). After benign/likely benign variants were excluded, pathogenic/likely pathogenic variants and variants with unknown significance were included.
Copy number variation analysis was performed using a CNV kit [11]. The estimated copy numbers were calculated using fold changes, and tumor purity was estimated from H&E-stained slides. Genes with an estimated copy number ≥ 5 were classified as amplifications, whereas those with an estimated copy number ≤ 0 were classified as copy number losses. Gene fusion was detected using BreaKmer v0.0.2 [12]. Split reads with read counts of < 2 were filtered out. All SSMs and fusions were manually reviewed by expert molecular pathologists using Integrative Genomics Viewer [13] to exclude false-positive variants.
Tumor mutation burden (TMB) was obtained from nine NEC or MANEC and 431 adenocarcinoma cases. All detected single nucleotide variant, insertion, and deletion were used to calculate TMB [14].
4. LRP1B immunohistochemistry
LDL receptor related protein 1B (LRP1B) immunohistochemical staining was performed to evaluate expression levels. Among 17 cases of NEC or MANEC, 14 had tissue available for immunohistochemical staining. The same formalin-fixed paraffin-embedded blocks used for next-generation sequencing were selected. Immunohistochemical staining was performed using a Benchmark XT autoimmunostainer (Ventana Medical Systems) with an OptView DAB IHC detection kit (Ventana Medical Systems), following the manufacturer’s instructions. Whole tissue sections were stained with anti-LRP1B antibody (1:200 dilution, PA5-64396, ThermoFisher Scientific/Invitrogen). Cytoplasmic staining in cancer cells was considered as positive. Evaluation was conducted using a semi-quantitative H-score method in the entire NEC area [H score=(% of cells stained at intensity score 1×1)+(% of cells stained at intensity score 2×2)+(% of cells stained at intensity score 3×3)] [15]. For MANECs, NEC areas were selected for evaluation.
5. Statistical analysis
The Wilcoxon rank-sum test was used to compare continuous variables between the two groups. The chi-square test or Fisher’s exact test were used, as appropriate, to compare categorical data between the two groups. Survival analyses were performed using the log-rank test and Kaplan-Meier plots. Multiple comparisons were corrected for false discovery rate using the Benjamini-Hochberg method. A p-value or Q-value less than 0.05 was considered statistically significant. All statistical analyses were performed using R ver. 4.2.3 (R Core Team). Genetic alterations were visualized using ComplexHeatmap [16] and G3viz [17] R packages.
Results
1. Clinicopathologic characteristics of patients
The clinicopathological characteristics of the patients included in this study are summarized in Table 1. Among the 17 patients with gastric NECs or MANECs, 16 were available for histology review. Of those, 12 (75.0%) had small-cell NEC components and four (25.0%) had large-cell NEC components. Patients with gastric NECs had either small cells (64%, 9/14) or large cells (29%, 4/14), whereas all three patients with gastric MANECs had small cell NEC components. The non-neuroendocrine component observed in the three MANECs was tubular adenocarcinoma (100%). One gastric NEC case (6%) did not have histological slides available for review.
Clinicopathological characteristics of patients with gastric NEC or MANEC, or adenocarcinoma included in this study
Four patients (23.5%) with initial stage III disease underwent curative surgery and adjuvant chemotherapy, whereas the other 13 patients (76.5%) were initially at stage IV and received palliative chemotherapy. There were no cases of microsatellite instability–high (MSI-H) or defective mismatch repair (dMMR) among the gastric NEC or MANEC cases, whereas 61 (4.4%) gastric adenocarcinomas showed MSI-H or dMMR. None of the gastric NEC or MANEC tumors were positive for the Epstein-Barr virus (EBV) by in situ hybridization, whereas 44 (3.2%) gastric adenocarcinomas were EBV-positive. No human epidermal growth factor receptor 2 (HER2)–positive tumors were found among gastric NECs or MANECs, whereas 158 (11.4%) gastric adenocarcinoma showed HER2 positivity. However, these were not significantly different (p > 0.99, p > 0.99, and p=0.243, respectively).
Among the 17 patients with gastric NECs and MANECs, 15 died, with a median overall survival (OS) of 6.7 months. We categorized the 17 patients into two groups based on survival duration: a shorter survival group (n=11), in which OS after diagnosis was less than 1 year, and a longer survival group (n=6), in which OS after diagnosis was 1 year or more. Two patients who were currently alive survived longer than 2 years and were therefore included in the longer survival group.
2. Frequent genetic alterations of gastric NECs and MANECs, gastric adenocarcinomas and extrapulmonary non-gastric NECs
Fig. 2 summarizes the frequent genetic alterations observed in gastric NECs and MANECs. The frequently mutated genes of gastric NECs and MANECs were TP53 (88%, 15/17), RB1 (59%, 10/17), SMARCA4 (47%, 8/17), RICTOR (41%, 7/17), APC (35%, 6/17), TOP1 (35%, 6/17), SLX4 (29%, 5/17), EGFR (29%, 5/17), BRCA2 (29%, 5/17), TERT (29%, 5/17), DDR2 (24%, 4/17), AURKA (24%, 4/17), and SRC (24%, 4/17). Frequent alterations in TP53 (87%, 20/23), RB1 (57%, 13/23), TERT (26%, 6/23), and APC (22%, 5/23) were similarly observed in extrapulmonary non-gastric NECs (S2A Fig.). In contrast, common genetic alterations in gastric adenocarcinomas included TP53 (61.2%), CDH1 (24.9%), LRP1B (24.5%), ARID1A (22.7%), ERBB2 (20.8%), GNAS (19.3%), CCNE1 (16.9%), NOTCH1 (15.1%), MYC (15.1%), APC (14.2%), BRCA2 (14.2%), CDK12 (12.7%), NOTCH4 (13.6%), ATM (12.1%), NOTCH3 (12.1%), ROS1 (12.2%), PIK3CA (12.0%), SMARCA4 (11.1%), CDKN2A (11.5%), and ERBB3 (10.7%) (S2B Fig.). ATM, FANCA, LTK, MSH2, BRD3, DDR1, MITF, GNAQ, SDHA, and ZNF141 mutations were found in both MANECs (33%, 1/3, each) and adenocarcinomas (0.65%-11.66%), but not in NECs. The results are presented in S3 Table.
OncoPrint plot for gastric neuroendocrine carcinomas (NECs) and mixed adenoneuroendocrine carcinomas (MANECs). TP53, RB1, and SMARCA4 are the three most frequently altered genes. CNA, copy number alteration; OS, overall survival.
We further investigated mutations in gastric NECs and MANECs to identify recurrent amino acid changes (S4 Fig.). We identified two mutations in V143—V143M and V143E—and two mutations in A159—A159L and A159P—within TP53. However, no other genes exhibited recurrent amino acid changes, likely due to the small sample size.
Next, we performed a more comprehensive comparative investigation of gastric NEC and MANEC and stomach adenocarcinomas. The results are presented in Fig. 3 and S5 Table. The median number of SSMs in each case was 8 (interquartile range, 7 to 11) in gastric NEC and MANEC and 8 (interquartile range, 5 to 11) in gastric adenocarcinoma, and was not significantly different between the two groups (p=0.698, q=0.835). The median number of copy number alterations in each case, including amplifications and losses, were slightly higher in gastric NEC and MANEC (median, 4; range, 2 to 12) than in gastric adenocarcinoma (median, 2; range, 0 to 6), although the statistical significance was marginal (p=0.069, q=0.144). AKT3 (p=0.002), APC (p=0.046), DDR2 (p=0.023), RB1 (p < 0.001), SLX4 (p=0.004), SMARCA4 (p=0.031), and TP53 (p=0.023) showed a higher occurrence of SSMs in gastric NEC and MANEC than in gastric adenocarcinomas. In contrast, ARID1A (p=0.032) and CDH1 (p=0.018) were more prevalent in gastric adenocarcinomas. Of these, AKT3, RB1, and SLX4 were statistically significant after multiple corrections (q=0.036, q=0.001, and q=0.045, respectively). The prevalence of amplifications in AURKA (p=0.032), BRCA2 (p=0.001), EGFR (p=0.043), FOXO1 (p=0.013), RICTOR (p=0.001), SMO (p=0.017), TERT (p=0.023), and TOP1 (p=0.025) was higher in gastric NEC and MANEC than in gastric adenocarcinoma, with BRCA2 and RICTOR being statistically significant after multiple correction (q=0.032 and q=0.029, respectively). The deletions of ADAMTS18, DDX11, KLRC3, KRAS, MAX, NFKBIA, NUDT7, and RB1 were more prevalent in gastric NEC and MANEC, with statistical significance after multiple corrections (q=0.037, q=0.032, q=0.044, q=0.036, q=0.037, q=0.029, q=0.037, and q=0.001, respectively). Neither group exhibited detectable canonical or druggable fusion.
Comparison of genetic alterations detected by next-generation sequencing between gastric neuroendocrine carcinoma (NEC) or mixed adenoneuroendocrine carcinoma (MANEC) and gastric adenocarcinoma. (A) Total number of small-scale mutations in each case. (B) Total number of copy number alterations in each case, including amplification and loss. (C) Frequency of small-scale mutations. (D) Frequency of gene amplifications. (E) Frequency of homozygous deletions. *p < 0.05, **q < 0.05.
The median TMB value was 12.5 mutations/Mb in NEC/MANEC (range, 7.8 to 20.3) and 14.1 mutations/Mb in adenocarcinoma (range, 1.6 to 128.1), with no statistically significant difference (p=0.567).
3. Comparison of genetic alterations of gastric NECs and MANECs between survival groups and NEC component
A comparison of the clinical parameters and genetic alterations of gastric NECs and MANECs between the two survival groups is shown in Fig. 4 and S6 Table. The longer survival group had a significantly lower TNM stage (p=0.006) and higher occurrence of LRP1B mutation (p=0.029), AURKA amplification (p=0.099), and BRCA2 amplification (p=0.099) than the shorter survival group, although they were not statistically significant after multiple corrections (q=0.674, q=1, q=1, and q=1, respectively), probably due to the limited sample size of both groups. There were no significant differences in age, sex, histologic diagnosis, cell components of NEC, EBV status, number of SSMs and copy number alterations, or other genetic alterations between the two survival groups.
Comparison of genetic alterations detected by next-generation sequencing of gastric neuroendocrine carcinoma or mixed adenoneuroendocrine carcinoma between survival groups. (A) Total number of small-scale mutations per case. (B) Total number of copy number alterations per case, including amplification and loss. (C) Frequency of small-scale mutations. (D) Frequency of gene amplifications. (E) The longer overall survival group seemed to harbor more frequent homozygous deletions of multiple genes than the shorter overall survival group, but these were not statistically significant. *p < 0.10, **p < 0.05.
A comparison of the clinical parameters and genetic alterations of gastric NECs and MANECs between small cell carcinoma and large cell carcinoma components is shown in S7 Fig. and S8 Table. Only the large cell carcinoma components showed EGFR and NOTCH3 mutations (2/4 for both), while small cell carcinoma components did not show any (0/12 for both; p=0.050 for both). However, these were not significantly different after multiple corrections (q=1 for both). Other clinicopathological parameters and genetic alterations were not significantly different between the small and large cell carcinoma groups.
4. Survival analysis of gastric NEC and MANEC in relation to genetic alterations
Survival analysis of gastric NEC and MAMEC was performed using Kaplan-Meier curves and the log-rank test. We investigated various factors, including TNM stage, the total number of genes with amplification or deletion, LRP1B mutation, RB1 mutation, BRCA2 amplification, and AURKA amplification. These factors demonstrated the previously mentioned differences between the two survival groups. As shown in Fig. 5, the presence of LRP1B mutation was associated with longer OS (p=0.026), whereas the presence of RB1 mutation and advanced TNM stage were associated with shorter OS (p=0.040 and p=0.003). The total number of genes with amplification or deletion, BRCA2 amplification, and AURKA amplification did not significantly affect the OS time (p=0.753, p=0.455, and p=0.098, respectively).
Kaplan-Meier curve of overall survival in patients with gastric neuroendocrine carcinoma (NEC) and mixed adenoneuroendocrine carcinoma (MANEC). Each is in relation to TNM stage and genetic alterations (A), including copy number alteration (CNA) (B), LRP1B mutation (C), RB1 mutation (D), BRCA2 amplification (E), AURKA amplification (F), and LRP1B or RB1 mutation (G).
When we divided the cases into three groups (RB1-mutated, LRP1B-mutated, and neither RB1-nor LRP1B-mutated), the RB1-mutated group exhibited the worst OS, followed by the no-mutation group and LRP1B-mutated group (p=0.030). None of the cases harbored both RB1 and LRP1B mutations. Among the three patients in the LRP1B-mutated group, two (66.7%) were in stage III and one (33.3%) was in stage IV, while all six patients (100%) in the RB1-mutated group were in stage IV and six of the eight patients (75.0%) in no-mutation group were in stage IV (Table 2). Furthermore, two (66.7%) patients in the LRP1B-mutated group were diagnosed with MANEC, whereas only one of the eight patients (12.5%) in the no-mutation group was diagnosed with MANEC and all six patients in the RB1-mutated group was diagnosed with NEC.
Clinicopathological characteristics of patients with gastric NEC or MANEC based on LRP1B and RB1 mutation status
5. Functional validation of LRP1B mutation and in gastric NEC and MANEC
We identified three cases harboring LRP1B mutations: one with a truncating mutation (p.Y3865*), another with multiple mutations (p.E158D, p.I321V and c.4170-1_4170delGA) and one with a missense mutation (p.G314S). Among these, p.E158D and p.G314S were predicted to be probably damaging by PolyPhen in silico prediction (scores, 0.523 and 0.961, respectively).
For further functional validation, immunohistochemical staining using anti-LRP1B antibody was performed on 14 cases of NEC or MANEC. Of these, two had LRP1B missense mutations while 12 did not. The median H-score of LRP1B in LRP1B-mutated tumors was 25 (range, 0 to 50), while in LRP1B-wild type tumors, it was 75 (range, 10 to 200). Decreased or total loss of LRP1B expression was noted in cases with mutant LRP1B (S9 Fig.). Statistical comparison was not possible due to the small sample size.
Discussion
The key findings of this study were the detection of frequent genetic alterations and their association with patient survival in gastric NEC and MANEC.
First, we found that the most frequent genetic alterations in gastric NECs and MANECs were TP53, RB1, SMARCA4, RICTOR, APC, TOP1, SLX4, EGFR, BRCA2, and TERT. Notably, truncating mutations or biallelic losses of TP53, RB1, SMARCA4, and APC were frequently observed. Compared to those in gastric adenocarcinoma, RB1 mutation, SLX4 mutation, AKT3 mutation, BRCA2 amplification, and RICTOR amplification were more frequent in gastric NEC and MANEC. Our results are consistent with previous studies (Table 3) [4,5,18-25]. TP53 mutation, APC mutation, and RB1 deletion or mutation were common in gastric NEC or MANEC. Of these, the TP53 mutation was the most prevalent genetic mutation (88%), which was consistent with findings from other studies (62%-100%). In addition, we identified several frequently altered genes, such as SMARCA4, RICTOR, TOP1, SLX4, BRCA2, and TERT, in gastric NEC and MANEC. Owing to the infrequency of gastric NEC and MANEC, most published research has encompassed a limited number of patients, usually approximately 10, with the largest study including 87 patients. We consider our study to be significant because it confirms earlier discoveries and reveals novel genetic modifications. However, further extensive research is required.
Review of the literature on genetic alterations in gastric neuroendocrine carcinoma or mixed adenoneuroendocrine carcinoma
Our findings indicate that the presence of LRP1B mutation may serve as an indicator of extended survival. We further validated the functional significance of LRP1B mutations using immunohistochemical staining and found decreased or total loss of LRP1B expression in LRP1B-mutated cancers, which suggests a loss of function. However, the presence of RB1 mutation is an unfavorable prognostic marker in patients with gastric NEC and MANEC. This observation is consistent with the findings of previous studies. Wu et al. [5] conducted a study of 87 patients diagnosed with gastric NEC, which is the most extensive study to date. They observed that biallelic RB1 inactivation was strongly linked to shorter OS time. However, our observations differ from those of Venizelos et al.’s study [4]. They did not find any noteworthy correlation between the examined mutations and OS. Wang et al. [19] reported a strong correlation between chromosomal instability and poor OS in patients with stomach NEC.
LRP1B encodes the low-density lipoprotein (LDL) receptor-related protein 1B, which belongs to the LDL receptor family. The LDL receptor family has diverse functions, including transporting lipids between cells, facilitating cell communication, and functioning as scavenger receptors [26]. LRP1B is normally expressed in various human tissues such as the central nervous system, thyroid gland, lungs, gastrointestinal tract, hepatobiliary system, soft tissue, and gynecological organs. It also plays roles in receptor-mediated endocytosis and signal transduction [27]. LRP1B is often inactivated in cancers and is a potential tumor suppressor. The presence of LRP1B mutation has been linked to favorable survival outcomes in lung adenocarcinoma [28], small cell lung cancer [29], biliary tract cancer [30], and colorectal cancer [31]. However, it is associated with poor prognosis in gastric cancer [32] and hepatocellular carcinoma [33,34], indicating different prognostic significance across tumor types. In one previous study, LRP1B mutation was an independent risk factor for shorter disease-free survival in stage II-III gastric cancer and associated with less infiltration of CD4-positive T cells and macrophages [32]. Moreover, LRP1B mutation has been documented as a prognostic indicator of a favorable immune checkpoint inhibitor response in various cancer types [35]. Therefore, we propose that the presence of LRP1B mutation in gastric NEC and MANEC is a potentially favorable survival indicator. However, further investigation and validation are needed, particularly to explore its possible association with a lower TNM stage and higher prevalence in MANEC.
We used the total number of amplified or deleted genes as an indicator of chromosomal instability in our survival analysis because of its straightforwardness and accessibility. However, its influence on OS rate was negligible. Hence, additional validation is required to establish a correlation between chromosomal instability and survival in patients with gastric NEC and MANEC.
The strengths and limitations of this study are as follows. Despite the rarity of gastric NECs and MANECs, we enrolled a sizable number of patients and used targeted NGS to examine various genes. Our findings support the previously reported prognostic significance of the RB1 mutation and raise the possibility that the absence of LRP1B mutation may be a new potential prognostic marker. However, the number of included patients was insufficient for in-depth statistical analysis. To overcome this limitation, we divided the patients into longer survival and shorter survival groups using an arbitrary criteria of 1-year survival. We performed comparative analyses and identified genetic differences between the two groups. But the sample size was still insufficient for multiple corrections or multivariable survival analysis. Therefore, further validation is required, which can be achieved by expanding the patient cohort using multivariate survival analysis and analyzing protein expression profiles. In addition, we did not perform multiregional sequencing for MANEC to compare the adenocarcinoma and NEC components. Several previous studies on gastric MANEC have revealed that the adenocarcinoma and NEC components within each MANEC share some pathogenic mutations, most frequently TP53, and that NEC components are associated with more aggressive genomic properties than their adenocarcinoma counterparts [20,25]. However, the application of these findings to patient management still needs to be investigated.
We comprehensively analyzed the genetic alterations and clinicopathological parameters in 17 gastric NECs and MANECs using targeted NGS. As the disease is rare, we included only a small number of cases. However, we found frequently altered genes, and the presence of RB1 and the absence of LRP1B mutations could be potential predictors of patient survival.
Electronic Supplementary Material
Supplementary materials are available at Cancer Research and Treatment website (https://www.e-crt.org).
Notes
Ethical Statement
All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1964 and later versions. This study was approved with a waiver of informed consent by the Institutional Review Board of Asan Medical Center (2022-0998).
Author Contributions
Conceived and designed the analysis: Hong SM.
Collected the data: Song IH, Park YS.
Contributed data or analysis tools: Kim DH.
Performed the analysis: Song IH, Ahn B, Park YS, Kim DH.
Wrote the paper: Song IH, Kim DH, Hong SM.
Conflict of Interest
Conflict of interest relevant to this article was not reported.
Acknowledgements
We would like to thank Editage (www.editage.co.kr) for English language editing.