Gastric cancer (GC) ranks as the fifth most common cancer worldwide, and is the third leading cause of cancer-related deaths. It is an aggressive disease, often diagnosed at advanced stages, with a 5-year survival rate of less than 30% ^[1][2][1,2].

In terms of classification, GC is a heterogeneous disease with multiple clinical, histological, and molecular variables influencing disease presentation and patient prognosis ^[3][4][3,4]. Geographical differences have been observed between Asian and Western countries, with GC being more prevalent in Asian regions. In fact, some high-incidence countries have implemented screening strategies that have improved early detection and patient outcomes ^[5][6][7][5,6,7]. Furthermore, geographic variations related to clinical, histological, prognostic, surgical, and treatment response factors have been noted ^[8][9][10][8,9,10].

Clinically, GC can be divided into proximal and distal types, each with distinct epidemiological characteristics. Proximal GC is associated with obesity, gastroesophageal reflux, and Barrett’s esophagus, while the more prevalent distal type is linked to Helicobacter pylori infections, the male gender, smoking, and dietary habits ^[1][11][12][1,11,12].

From a macroscopic perspective, GC can be classified using the Paris classification for superficial lesions, the Borrmann classification for advanced GC (stage pT2 or higher), or the Japanese Society of Endoscopy classification, encompassing both early and advanced GC ^[13][14][15][13,14,15].

With respect to histological features, notable classifications include the Laurén and the World Health Organization (WHO) systems ^[16][17][16,17]. Laurén’s classification, established in 1965 as a histoclinical classification, categorizes GC into intestinal and diffuse types. Intestinal GC forms tubules and may include papillary or solid structures, occurs in older patients, and is associated with H. pylori infection and environmental factors. It develops through the carcinogenic process of chronic gastritis—intestinal metaplasia—dysplasia [18]. In contrast, diffuse GC is composed of loosely cohesive cells, potentially displaying signet ring morphology, appears in younger patients, and is induced by active inflammation or genetic factors. Previous studies have shown that this classification correlates with patient prognosis, treatment response, and the molecular characteristics of GC [8]. On the other hand, the WHO classification is more complex, morphology-based, and identifies the following four main types of GC: tubular, papillary, poorly cohesive, and mucinous [19]. This classification has shown a lower correlation with non-histological factors [20]. It should be noted that both classifications establish a "mixed" subtype.

Regarding GC treatment, surgery remains the only curative option for GC [21]. Endoscopic techniques can be employed in early stages, while more advanced stages, prevalent in Western countries, require a total or subtotal gastrectomy with lymphadenectomy [22]. For non-surgical patients, chemotherapy is the main therapeutic approach [23]. Approved targeted drugs include antiangiogenics (anti-VEGFR-2) and anti-HER2 agents ^[24][25][24,25]. Additionally, the approval of pembrolizumab for solid tumors with high microsatellite instability (MSI-H) or mismatch repair deficiency (dMMR) included GC cases ^[26][27][26,27]. The indication for immunotherapy also depends on PD-L1 expression or the tumor mutational burden (TMB) [28]. Therefore, the only established and broadly available biomarkers for GC treatment are HER2 amplification, MSI-H, and PD-L1 expression [29]. The therapeutic arsenal for GC is limited when compared to other tumor types, and current therapies have not significantly improved patient prognosis ^[30][31][30,31].

In terms of molecular characteristics, technological advancements in recent years have allowed the identification of multiple molecular alterations in various types of tumors [32]. Among these, alterations with prognostic or therapeutic value have significantly impacted clinical practice in tumors such as lung or breast cancer, enabling personalized treatment, improving patient outcomes, and reducing the side effects associated with conventional treatment ^[33][34][33,34].

In GC, multiple studies have analyzed its genetic, epigenetic, transcriptomic, proteomic, or metabolomic profiles, revealing numerous molecular changes and dysregulated pathways, some of which carry prognostic and/or therapeutic significance ^{[35][36][37][38][39][40][41]}[35,36,37,38,39,40,41]. The synthesis of this information has given rise to several molecular classifications, with notable examples being those published by The Cancer Genome Atlas (TCGA) and the Asian Cancer Research Group (ACRG) ^[42][43][42,43]. Despite these efforts, the practical impact of these classifications on clinical practice remains limited, primarily due to the complexity of their implementation. Beyond these pivotal studies, various authors have proposed alternative molecular classifications of GC that require external validation in other cohorts and the identification of surrogate markers for their application. Consequently, there is an urgent need to reach a consensus on molecular categories, establish easily detectable subgroups, and identify optimal surrogate markers for each molecular subtype.

As previously mentioned, recent technological advancements have propelled cancer research into the molecular era. Comprehensive genetic, transcriptomic, and proteomic analyses are now possible, resulting in vast databases of molecular changes, including mutations, copy number variants, epigenetic alterations, gene expression profiles, or disrupted pathways across various tumor types. This wealth of information has enabled the identification of molecular alterations with prognostic and therapeutic significance. Prognostic alterations allow for personalized patient management, improving the cost-effectiveness of treatment. Meanwhile, predictive molecular alterations have transformed cancer treatment from a generic approach to an individualized approach. Targeted drugs have enhanced patient prognosis, often with fewer side effects and better tolerance than conventional chemotherapy [44].

Among the technological advances that have impacted the molecular characterization of cancer in the last decade, microarrays and second- or next-generation sequencing platforms (NGS) stand out. Microarrays allow the detection of molecular alterations at the DNA, RNA, or protein level ^[45][46][47][45,46,47]. They are primarily applied in research studies, although some commercial microarray-based platforms are used in clinical routine, mainly in breast cancer ^[48][49][48,49]. NGS techniques, which have been implemented in clinical practices in institutions worldwide, are typically employed for DNA sequencing. They facilitate the simultaneous analysis of multiple samples, either at the whole genome or whole exome level, or through the utilization of targeted panels containing dozens of genes of interest. These techniques have been refined, automated, and modified to allow for the analysis of RNA or epigenetic alterations. In GC, NGS and microarrays have played a pivotal role in elucidating the landscape of molecular alterations ^{[50][51][52][53]}[50,51,52,53]. However, in the daily practice of GC, these techniques do not present significant applications because the necessary biomarkers are currently analyzed using immunohistochemistry (IHC) and in situ hybridization methods. NGS could be useful for determining the TMB, or as a complementary technique for assessing MSI status ^{[54][55][56][57]}[54,55,56,57].

In the early 2010s, third-generation sequencing techniques emerged, enabling sequencing at a single-molecule level. Despite their potential, these techniques have not been integrated into clinical practice, and their utilization in research studies remains limited. Advantages over second-generation sequencing methods include the fact that they do not require sample pre-amplification and can read longer fragments of DNA, but the error rate is generally higher (10–15%) ^{[58][59][60][61]}[58,59,60,61].

Lastly, another interesting molecular approach that has garnered attention in recent years is single-cell sequencing (SCS), which, using second- or third-generation methodologies, enables the analysis of DNA, RNA, or methylome at the single-cell level ^{[62][63][64][65]}[62,63,64,65]. Its main advantage lies in its ability to scrutinize the molecular profile of cell subclones, thereby offering significant potential for evaluating tumor heterogeneity, refining the personalization of patient management, and enhancing the monitoring of treatment response and resistance detection ^[66][67][66,67]. Furthermore, SCS requires a small sample size, thus making it suitable for analyzing circulating tumor cells in liquid biopsy specimens [68]. However, these techniques have not yet been implemented in clinical routine and require technical refinement, standardization, and cost reduction to have a practical impact ^[69][70][69,70]. In GC, research in this area is in its early stages, but promising results have been obtained ^[71][72][73][71,72,73].

Multiple molecular alterations and dysregulated pathways have been identified in GC. Notably, mutations in the TP53 and CDH1 genes are prominent ^[74][75][74,75]. TP53 mutation is the most common in GC, occurring in over 50% of cases, and is often associated with chromosomal instability and an increased expression of cell-cycle progression genes ^[76][77][76,77]. While the TP53 mutation has been correlated with a worse prognosis in other tumors, its significance in GC remains unclear ^{[75][78][79][80]}[75,78,79,80]. This uncertainty may stem from the specific impact of different mutations on the function of the p53 protein, concomitant molecular alterations, or treatment effects ^[75][81][75,81]. Additionally, most studies have focused on the p53 protein rather than the gene, with some exceptions ^{[75][77][82][83]}[75,77,82,83]. As for CDH1, it encodes for E-cadherin, a transmembrane glycoprotein responsible for maintaining cell–cell adhesion [84]. Germline mutations in CDH1 are associated with hereditary diffuse GC syndrome, which increases the risk of diffuse GC and lobular breast carcinoma [85]. In sporadic GC cases, mutations and the abnormal methylation of CDH1 are predominantly found in diffuse GC ^[86][87][86,87]. Other key mutations in GC include those within the ARID1A, PIK3CA, or BRCA2 genes ^[88][89][90][88,89,90].

Regarding copy number alterations, the amplification of genes involved in tyrosine kinase receptor pathways, such as FGFR2, HER2, EGFR, or MET, stand out ^{[91][92][93][94]}[91,92,93,94]. Among these genes, HER2 amplification has significant clinical implications, serving as an indication for treatment with trastuzumab in advanced HER2-positive GC patients ^[95][96][95,96]. HER2 amplification occurs in 10–15% of patients with advanced GC, with a higher prevalence in intestinal-type GC and a lower prevalence in diffuse GC ^[97][98][99][97,98,99]. The most frequent copy number variation is the amplification of FGFR2, which is observed in 15% of patients and associated with high-grade tumors and a poorer prognosis [93].

Finally, the main dysregulated pathways in GC include those related to genome integrity, cell adhesion, chromatin remodeling, cell motility and cytoskeletal structure, Wnt signaling, and tyrosine kinase receptors [74].

As for the management of GC, surgery remains the only curative option, and most resectable tumors are treated with total or subtotal gastrectomy associated with D2 lymphadenectomy. Early stage tumors meeting certain criteria may undergo endoscopic procedures, such as endoscopic mucosal resection or endoscopic submucosal dissection [100]. The assessment of tumor depth, size, grade, and the presence of ulceration is crucial to determine the suitability of these techniques [101].

Surgery for GC typically forms part of a multimodal treatment, with the two following options, depending on the context: surgery followed by adjuvant chemotherapy or perioperative therapy. Regarding the surgical procedure, according to the European Society for Medical Oncology (ESMO) guidelines, T1 tumors can be treated with partial gastrectomy and D1 lymphadenectomy, while for IB-III disease, total or subtotal gastrectomy with D2 lymphadenectomy is recommended [101]. Perioperative chemotherapy has become the standard of care, supported by findings from clinical trials which have been conducted since the 2000s, demonstrating a survival benefit for patients undergoing this approach ^{[102][103][104]}[102,103,104]. ESMO guidelines advocate for the pre- and post-operative administration of FLOT regimen (5-FU, leucovorin, oxaliplatin, and docetaxel) in patients who can tolerate it [101]. The choice of the chemotherapy regimen may vary depending on the guideline, and the role of radiotherapy as an adjunct is still under investigation ^{[105][106][107]}[105,106,107].

The main innovations in the surgical treatment of GC include the use of laparoscopy, which has been shown to be non-inferior to open surgery in both Asian and Western countries, and robot-assisted gastrectomy ^{[108][109][110][111]}[108,109,110,111].

Regarding non-surgical cases, it is worth noting that, despite the detection of numerous molecular alterations and the development of multiple molecular classifications in GC, the clinical application of this information lags behind other cancers. For instance, breast cancer has successfully integrated molecular classification into daily practice, surpassing the practical impact of traditional histological features. In lung cancer, multiple targetable alterations have been identified, leading to recommendations for testing as many as nine molecular biomarkers and PD-L1 expression in all adenocarcinomas and in squamous cell carcinomas that meet certain criteria [112]. As a final example, in endometrial cancer, molecular and histopathological features have been integrated to develop a new FIGO staging system with prognostic and therapeutic value, which has been in effect since 2023 [113].

Contrastingly, in unresectable GC, the main therapeutic approach continues to be conventional chemotherapy, typically involving a platinum-fluoropyrimidine doublet [114]. Nonetheless, many patients develop resistance to this treatment, which often leads to adverse effects ^{[115][116][117]}[115,116,117].

The administration of targeted therapy has the potential to enhance the specificity and efficacy of oncological treatment, while mitigating adverse effects [118]. As drawbacks, the effectiveness of these therapies is heavily reliant on the molecular profile of the tumor at a given time, and they are not entirely devoid of toxicity ^[44][119][44,119]. According to the latest National Comprehensive Cancer Network guidelines, the main targeted therapies approved for advanced GC include anti-HER2 agents (trastuzumab and fam-trastuzumab deruxtecan-nxki), anti-VEGFR-2 agents (ramucirumab), and immunotherapy (nivolumab, pembrolizumab and dostarlimab-gxly) [120]. Anti-HER2 therapy is indicated in HER2-amplified GC, and immunotherapy may be indicated in cases with MSI-H, PD-L1 overexpression, or a high TMB [120]. However, the latest ESMO guidelines only include PD-L1 expression and MSI-H as indications for immunotherapy [101]. Lastly, tumors with NTRK1, NTRK2, or NTRK3 gene fusions may be treated with entrectinib and larotrectinib, although such cases are exceptionally rare in GC, with only one case published so far [121].

Early GC has demonstrated favorable outcomes for decades, with survival rates of over 90% with surgical treatment ^[122][123][122,123]. However, in Western countries, the lack of widespread screening techniques coupled with mild and nonspecific symptoms has lead to over 80% of patients being diagnosed at advanced stages [124]. Despite advancements in molecular biology and personalized therapy, the prognosis for advanced GC has seen limited improvement [125]. Even in resectable cases, the recurrence rates range from 14–80%, often exceeding 40% within the first years following surgery ^[126][127][126,127]. The addition of neoadjuvant therapy in surgical cases has slightly improved patient prognosis, but studies report recurrence rates exceeding 30% ^{[128][129][130]}[128,129,130]. Unresectable cases present a dismal prognosis, with median overall survival rates ranging from 11 to 14 months, and 5-year survival rates of less than 30% ^{[131][132][133][134]}[131,132,133,134]. Notably, patients eligible for targeted therapy or immunotherapy exhibit significantly higher survival rates overall ^[135][136][135,136]. However, numerous authors highlight the need to refine patient selection for these treatments, enhance drug efficacy, identify new therapeutic targets, and overcome treatment resistance ^{[137][138][139][140]}[137,138,139,140].

The modest impact of the aforementioned advancements on the prognosis and management of advanced GC and the scarcity of therapeutic targets could be due to the heterogeneity that characterizes this tumor, both phenotypically and molecularly ^[3][4][141][3,4,141]. This heterogeneity is also evident at the intratumoral and tumor microenvironmental levels, as demonstrated by recent single-cell studies ^[142][143][142,143]. Additionally, molecular heterogeneity exists among primary tumors, lymph node metastases, and distant metastatic sites ^{[144][145][146]}[144,145,146]. Understanding spatial and temporal heterogeneity, both phenotypically and molecularly, at primary and metastatic sites holds promise for improving prognosis and treatment outcomes for GC patients.

New potential treatment strategies for GC encompass perioperative targeted therapy or immunotherapy, personalized treatment guided by molecular tumor characterization, the utilization of trastuzumab conjugates, and the development of new anti-HER2 agents. Additionally, ongoing studies are investigating novel therapeutic approaches, such as Claudin 18.2 targeted therapy or FGFR, MET, and EGFR inhibitors ^[147][148][147,148].

In summary, these circumstances highlight the need for enhancing patient stratification in both clinical trials and practice. Additionally, identifying new biomarkers and improving the currently available drugs is crucial to expand the range and effectiveness of targeted therapies for GC and translate the progress seen in other tumors to GC.