High tumor mutational burden assessed through next-generation sequencing predicts favorable survival in microsatellite stable metastatic colon cancer patients Authors: Annabella Di Mauro, Mariachiara Santorsola, Giovanni Savarese, Roberto Sirica, Monica Ianniello, Alessia Maria Cossu, Anna Ceccarelli, Francesco Sabbatino, Marco Bocchetti, Anna Chiara Carratù, Francesca Pentimalli, Gerardo Ferrara, Guglielmo Nasti, Michele Caraglia, Alessandro Ottaiano Published: 2024 ABSTRACT AbstractBackgroundMicrosatellite instability (MSI) is a well-established predictive biomarker for immune checkpoint inhibitor (ICI) response in metastatic colon cancer. Both high MSI and tumor mutational burden (TMB) are markers of genomic instability. However, the prognostic and predictive value of TMB in patients with microsatellite stable (MSS) tumors remains unclear.MethodsWe evaluated the prognostic significance of TMB levels in MSS metastatic colon cancer patients undergoing standard treatments. Tumor responses were assessed using RECIST v1.1 criteria. Comprehensive clinical and molecular profiling was conducted, including next-generation sequencing (NGS) for TMB evaluation with the TruSight Oncology® kit. Overall survival (OS) was the primary endpoint. Multivariate Cox regression analysis was utilized to assess the relationship among potential prognostic factors.ResultsAmong 102 MSS metastatic colon cancer patients, high TMB (> 10 mut/mb) was associated with a significantly longer median OS compared to low TMB (70.0 vs 45.0 months, respectively; HR: 0.45; 95% CIs 0.21 to 0.96; P = 0.0396). Multivariate analysis, adjusting for age, gender, number of metastatic sites, response to first-line chemotherapy, RAS mutational status, and liver involvement, identified TMB as an independent prognostic factor, along with response to first-line chemotherapy.ConclusionsOur results highlight the prognostic significance of TMB in MSS metastatic colon cancer patients, suggesting its potential role in patient stratification and treatment decision-making.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12967-024-05927-9.Keywords: Microsatellite instability, Tumor mutational burden, Colon cancer, Prognosis, NGS Background Colorectal cancer (CRC) is one of the most common malignancies worldwide, with over 1 million new cases and 600,000 deaths annually. Despite significant advances in early detection and treatment, about 40% of patients present with or will eventually develop metastatic disease [1]. In recent years, the treatment of metastatic colorectal cancer (mCRC) has been enriched with new biologic drugs, including small molecules and monoclonal antibodies, which can be administered as monotherapy or in combination with chemotherapy [2]. However, despite the improvement in median survival to over 24 months through a continuum-of-care approach using all available treatments, mCRC remains a significant cause of morbidity and mortality, with a five-year survival rate of only 10–15% [1, 2]. One of the latest advancements in treatment is immunotherapy. The Food and Drug Administration (FDA) has recently approved pembrolizumab, an anti-PD-1 immune checkpoint inhibitor (ICI), for treating microsatellite instability (MSI) or high tumor mutational burden (TMB) solid tumors, including mCRC. Pembrolizumab works by restoring the immune system ability to attack tumor cells [3, 4]. MSI tumors are characterized by numerous errors in microsatellite DNA sequences due to a deficient mismatch repair system (dMMR), leading to mutation accumulation [5]. Microsatellite status can be characterized using different techniques (from immunohistochemistry to next-generation sequencing, NGS), all of which show good concordance [6]. Only about 5% of mCRC patients exhibit MSI status, limiting the patient population that can benefit from ICIs. Another biomarker that directly measures the mutational propensity of neoplastic cells is the TMB. TMB is typically defined as the total number of somatic non-synonymous mutations per coding region of a tumor genome (mutations per megabase, mut/mb) [7]. Similar to MSI, the focus on TMB as a biomarker in solid tumors is primarily due to its potential to predict response to immunotherapy. A higher TMB likely indicates a greater presence of neoantigens capable of eliciting an antitumor immune response. In melanoma and non-small cell lung cancer (NSCLC), the predictive role of TMB is well-established, showing that high TMB correlates with excellent responses to ICIs [8, 9]. Prognostically, the results are mixed. In lung cancer, high TMB (> 8 mut/Mb vs ≤ 8 mut/Mb) is associated with better survival, especially in stage I-III patients who underwent lung resection and standard post-operative treatments. High TMB is an independent variable in multivariate analysis adjusted for treatment type, age, sex, performance status, histology, and T/N stage [10]. Assessing the prognostic role of TMB in melanoma outside the context of ICIs is challenging, as ICIs are the standard treatment. However, a significant study evaluated the mortality risk of patients with high versus low TMB across various solid tumors, both before and after the introduction of ICIs treatment. In NSCLC and melanoma, high TMB shifted from a negative to a positive prognostic factor with the introduction of ICIs. In contrast, high TMB consistently favored prognosis in colorectal, endometrial, and bladder cancers, independently from ICI treatment [11]. However, in CRC, the results are contradictory. A study analyzing two public datasets (MSKCC, Memorial Sloan Kettering Cancer Center, and TCGA, The Cancer Genome Atlas) did not find a significant difference in survival between patients with high vs low TMB, except in those with KRAS mutations, where high TMB was associated with better prognosis. However, these patients were highly heterogeneous, especially regarding staging, which included all stages from I to IV [12]. Another study examining TMB in mCRC patients enrolled in the FIRE-3 trial found no prognostic role for TMB (high > 8 Mut/Mb vs low ≤ 8 Mut/Mb) [13]. Another study, which categorized TMB as low, intermediate, or high (< 7, 7–16, or ≥ 17 Mut/Mb, respectively), suggested a trend indicating that high TMB is associated with improved prognosis in metastatic colorectal cancer [14]. The present study was undertaken to measure the prognostic effect of TMB in a well-characterized cohort of MSS metastatic colon cancer patients outside the context of ICIs treatment. Methods Study design and primary objective This single-center, observational study aimed to explore the prognostic power of TMB in metastatic colon cancer (mCC) patients with MSS tumors and who received standard treatments. Patients’ selection and clinical management Between January 2016 and January 2024, patients were enrolled and managed according to European Society of Medical Oncology (ESMO) guidelines [15]. To control prognostic variables, exclusions were made for patients with Performance Status (PS) Eastern Cooperative Oncology Group (ECOG) ≥ 2, cachexia risk > 1 [16], peritoneal carcinomatosis, and life expectancy < 3 months (as assessed by the oncologists involved in the study). Inclusion criteria included age > 18 years, MSS tumors confirmed by immunohistochemical analysis of MMR proteins [17] and verified through NGS. Only patients with colon cancer were included to ensure homogeneous and interpretable results, recognizing the distinct clinical, biological, and molecular characteristics of colon versus rectal cancers. Patients were monitored regularly with total body Computed Tomography (tbCT) scans and/or Magnetic Resonance Imaging (MRI) every six months, and treatment responses were evaluated according to Response Evaluation Criteria In Solid Tumors (RECIST v1.1) [18]. The disease control (DC) rate was defined as the proportion of patients achieving complete or partial responses or stable disease, while progressive disease indicated a lack of disease control. As the study was non-therapeutic, formal ethics approval was not required per institutional guidelines. The study adhered to the Declaration of Helsinki principles, and all participants provided written informed consent before treatment or genetic testing. Tumor specimens and sequencing Primary colon cancer samples were collected as formalin-fixed and paraffin-embedded (FFPE) tissue specimens, and tumor cell microdissection was performed under morphological guidance. DNA was extracted using the MGF03-Genomic DNA FFPE One-Step Kit following the manufacturer’s instructions (MagCore Diatech). The DNA quality was evaluated in triplicate with the FFPE QC Kit, following the manufacturer’s guidelines (Illumina, San Diego, USA). Libraries were created using the TruSight Oncology® 500 kit, which targets 523 cancer-related genes (full list in Supplementary File 1). This assay detects a range of genomic alterations, including small nucleotide variants (SNVs), indels, splice variants, copy number variants, fusions, and immunotherapy biomarkers like TMB and microsatellite status. Sequencing was performed on an Illumina NovaSeq 6000 platform (San Diego, USA). TMB quantification was performed as described by Chalmers et al. [19]. This process included all coding somatic base substitutions and insertions/deletions (indels) within the targeted regions, encompassing synonymous mutations. Variant calling and TMB calculation algorithms were kept independently to ensure accurate coding variant counts (see Manufacturer Instructions at https://emea.support.illumina.com/). The targeted coding genomic region spanned 1.9 Mb. An accurate exome-based predictive model for MSI classification was used, employing a statistical classifier derived from somatic mutation profiles to differentiate MSI from MSS tumors. This classifier was trained on 999 exome-sequenced TCGA tumor samples with known MSI status (based on mononucleotide markers) and demonstrated a positive predictive value of 98.9% and a negative predictive value of 98.8% in an independent test set of 427 samples (see Manufacturer Instructions at https://emea.support.illumina.com/). Bioinformatics analysis and data presentation The bioinformatics pipeline of Illumina TruSight Oncology 500 was used for the analysis of sequencing data. A median of 117 million reads was generated per sample, with coverage in the target region exceeding the manufacturer's suggested threshold of 150X. Sequence data were aligned to the human reference genome GRCh37 using the Burrows–Wheeler Aligner with default parameters [20]. Population- and cancer-specific variants were cross-referenced with several databases, including GENCODE, dbNSFP, ICGC-PCAWG, COSMIC, 1000Genomes, ClinVar, CancerMine, OncoScore, CIViC, and CBMDB, to evaluate clinical significance. Variants were filtered using unmatched normal datasets and excluded if the global minor allele frequency was < 1%. Prioritization of variants followed a four-tiered structure (Tier 1–4), in line with AMP/ACMG joint consensus recommendations [21]. Variants with strong clinical significance in cancer were identified based on evidence levels from databases such as CIViC and Cancer Biomarkers. The prognostic impact of clinic-pathological variables on overall survival (OS) was investigated. OS was calculated from metastatic disease diagnosis until death from colon cancer (cancer-specific survival). Progression-free survival (PFS) was not included due to heterogeneous treatments and radiologic evaluations, making vital status a more reliable outcome. Data were obtained from an electronic database (analyzed in April 2024) containing clinical and pathological information of mCC patients treated at the unit of Innovative Therapies for Abdominal Metastases of the Istituto Nazionale Tumori di Napoli “G. Pascale”. Covariates were dichotomized for univariate and multivariate analyses: age (≥ 70 vs < 70 years), gender (male vs female), metastatic involvement (more than one site vs one site), response to first-line therapy (disease control vs no disease control), RAS mutational status variants (wild-type vs mutated), liver involvement (yes vs no), and TMB high (high ≥ 10 mut/mb vs low < 10 mut/mb). Regarding the definition and dichotomization of TMB, given the FDA approval of pembrolizumab as an agnostic treatment for all metastatic solid tumors with ≥ 10 mut/Mb, and considering the effectiveness of this cut-off in real-world populations as a predictive biomarker for immunotherapy [22, 23], a consensus among the authors led to the adoption of this 10 mut/Mb cut-off. OS was assessed using the Kaplan–Meier method, with statistical significance determined by the two-tailed log-rank test for univariate analysis. Multivariate analysis, based on the Cox proportional-hazards regression model, examined prognostic interactions between OS and covariates, with Hazard Ratios (HRs) reported alongside 95% confidence intervals (CIs). Covariates were carefully selected based on clinical relevance. The “enter” method was used for the Cox regression model, incorporating all selected covariates simultaneously to assess their independent prognostic significance. No stepwise selection method was applied; all variables were retained in the model to ensure consistency and minimize potential biases associated with variable exclusion. The t-test was used to compare the mean values of continuous variables (ages) between high TMB and low TMB groups. Associations between clinicopathological variables and classification groups were studied using contingency tables and the chi-square test. A p-value less than the conventional level of 0.05 was considered significant. Statistical analyses were conducted using Excel software and MedCalc® version 20.112. Results Clinicopathological characteristics of patients and tumors The flowchart describing the selection of the cohort is shown in Fig. 1, and the clinical and pathological characteristics of the study population are summarized in Table 1. Flowchart describing the inclusion and exclusion process of screened and analyzed patients, with reasons for exclusion listed on the right Table 1Clinico-pathological characteristics of analysed patientsVariableNo. (%)TMBP < 10 ≥ 10Age Median (range)64 (31–82)63 (31–82)68 (48–81)0.0389Gender Female36 (35.3)288 Male66 (64.7)47190.4748Side Left70 (68.6)5119 Right32 (31.4)2480.8208Grading G17 (6.9)43 G2/395 (93.1)71240.3109pT 1/213 (12.7)94 352 (50.9)3814 415 (14.7)1140.9591 Unknown22 (21.6)175pN 029 (28.4)209 134 (33.3)268 217 (16.7)1250.7859 Unknown22 (21.6)175No. of metastatic sites 178 (76.5)5622 ≥ 224 (23.5)1950.4763Liver involvement Yes72 (70.6)5319 No30 (29.4)2280.9770 Clinico-pathological characteristics of analysed patients The median age of the entire cohort was 64 years, ranging from 31 to 82 years. When stratified by TMB, the median age was 63 years (range 31–82) in the TMB < 10 mut/mb group and 68 years (range 48–81) in the TMB ≥ 10 mutations/mb group. The difference in median age between the two TMB groups was statistically significant (P = 0.0389). Among the 102 patients, 36 (35.3%) were female and 66 (64.7%) were male. The primary tumor was predominantly located on the left side (68.6%), and most tumors were graded as G2/3 (93.1%). At initial diagnosis, 13 patients (12.7%) had pT1/2 stage tumors, 52 patients (50.9%) had pT3 stage tumors, 15 patients (14.7%) had pT4 stage tumors, and the pT stage was unknown for 22 patients (21.6%) since these patients did not undergo to resection of the primary tumor. The majority of patients had pN1 pathological stage tumors (33.3%). Seventy-eight patients (76.5%) developed metastasis at one organ, while 24 patients (23.5%) had more than two metastatic sites at the time of diagnosis. Liver involvement was observed in 70.6% of cases. The distribution of gender, tumor side, pT stage, pN stage, number of metastatic sites, and liver as a metastatic site did not differ significantly between the two TMB groups (P > 0.05). Associations between TMB, gene mutations, and treatment responses With an exploratory and descriptive aim, we conducted analyses of TMB levels within genetic profiles of patients (BRAF p.V600E, PIK3CA, RAS, and TP53) (Table 2) and their response to treatments (Table 3). Table 2Associations between tumor mutational burden (TMB) and mutations in specific key driver genesGene statusTotal no. (%)TMBPLowHighBRAF p.V600E Wild-type96 (94.1)7323 Mutated6 (5.9)240.0221PIK3CA Wild-type86 (84.3)6422 Mutated16 (15.7)1150.6386RAS Wild-type58 (56.9)3919 Mutated44 (43.1)3680.1001TP53 Wild-type39 (38.2)2910 Mutated63 (61.8)46170.8818 Associations between tumor mutational burden (TMB) and mutations in specific key driver genes Table 3Associations between tumor mutational burden (TMB) and Response to First-Line ChemotherapyVariableTotal no. (%)TMBPLowHighResponse to first-line CT DC80 (78.4)5921 No DC22 (21.6)1660.9237Response to first-line anti-EGFR CT DC49 (87.5)3118 No DC7 (12.5)610.2449Response to first-line anti-VEGF* CT DC31 (67.4)283 No DC15 (32.6)1050.0497CT: chemotherapy; DC: Disease control; EGFR: Epidermal growth factor receptor; VEGF: Vascular endothelial growth factor*bevacizumab Associations between tumor mutational burden (TMB) and Response to First-Line Chemotherapy CT: chemotherapy; DC: Disease control; EGFR: Epidermal growth factor receptor; VEGF: Vascular endothelial growth factor *bevacizumab The analysis revealed a significant association only between BRAF p.V600E mutation status and TMB levels (P = 0.0221), indicating that patients with this mutation are more likely to have higher TMB. No significant correlations were found between TMB levels and mutations in PIK3CA (P = 0.6386), RAS (P = 0.1001), and TP53 (P = 0.8818). The correlation between TMB levels and the propensity for mutations may be linked to variations in responses to first-line chemotherapies. Upon further examination, we found that, although there were no significant associations between TMB levels and response rates in the overall cohort or among patients treated with anti-EGFR therapies, our analysis suggests a potential effect of TMB levels on the efficacy of treatment regimens containing bevacizumab. Specifically, patients with low TMB showed a slightly better response compared to those with high TMB (P = 0.0497). Prognostic significance of TMB The primary objective of this study was to investigate the prognostic significance of TMB in patients with MSS metastatic colon cancer (mCC). All participants had an ECOG performance status of 0–1, a life expectancy of more than three months (as assessed by the oncologists involved in the study), and a cachexia risk score below 1. After a median follow-up of 46 months, there were 53 deaths attributed to colon cancer. Kaplan–Meier survival curves for overall survival (OS) revealed a distinct separation based on TMB levels (low vs. high), with the difference becoming particularly evident after one year of follow-up (Fig. 2). Kaplan–Meier survival curves for patients with high tumor mutational burden (TMB) versus low TMB show a significantly divergent course (P = 0.0372, log-rank test), favoring patients with high TMB. Hazard ratios and confidence intervals are reported in Table 4 The median OS (mOS) was 70.0 months for patients with high TMB compared to 45.0 months for those with low TMB (P = 0.0372, log-rank test; HR: 0.47; CI 0.22–0.99). In multivariate analysis, after adjusting for age (≥ 70 years vs < 70 years), gender (male vs female), extent of metastatic disease (multiple sites vs single site), response to first-line chemotherapy (disease control vs no disease control), RAS mutational status (mutated vs wild-type), and liver involvement (involved ns not involved), TMB levels remained an independent prognostic factor (P = 0.0396), alongside the response to first-line chemotherapy (P = 0.0134). HRs with CIs for each covariate are presented in Table 4. Table 4Uni- and multi-variate analyses of the prognostic impact of TMB with selected covariatesCo-variateDicothomizationMedian survivals (months)No. of events/patientsP at univariateHR95% CIsP at multivariateAge≥70y vs < 70y47.0 vs 87.039/73 vs 14/290.49290.940.48 to 1.840.8689GenderM vs F61.0 vs 42.036/66 vs 17/360.51991.140.61 to 2.100.6704No. of metastatic sites> 1 vs 1 site32.0 vs 61.016/24 vs 37/780.00380.610.31 to 1.190.1514Response to first-line CTDC vs no DC29.0 vs 63.036/80 vs 17/220.00020.440.22 to 0.840.0134RAS mutational statusMutated vs wild-type34.0 vs 62.026/44 vs 27/580.09831.290.73 to 2.270.3798Liver involvementYes vs No45.0 vs 70.041/72 vs 12/300.13031.350.66 to 2.750.4022TMB high vs lowYes vs No70.0 vs 45.09/27 vs 44/750.03720.450.21 to 0.960.0396CI: Confidence Interval; CT: chemotherapy; DC: Disease Control; F: Female; HR: Hazard Ratio; M: Male; y: years; TMB: Tumor mutational burden Uni- and multi-variate analyses of the prognostic impact of TMB with selected covariates CI: Confidence Interval; CT: chemotherapy; DC: Disease Control; F: Female; HR: Hazard Ratio; M: Male; y: years; TMB: Tumor mutational burden Discussion The present study aimed to explore the prognostic value of TMB in a well-characterized cohort of metastatic colon cancer patients with microsatellite stable tumors who received standard treatments. None of the enrolled patients received immune checkpoint inhibitor (ICI)-based therapy, as per European pharmacy regulatory guidelines. Our findings suggest that high TMB (> 10 mut/Mb), observed in 26.5% of the patients, is associated with improved overall survival (OS). Specifically, patients with high TMB had a median OS of 70 months, compared to 45 months for those with low TMB (P = 0.0372). Importantly, TMB remained an independent prognostic factor in multivariate analysis, highlighting its potential significance as a biomarker in colon cancer. Previous studies by Valero et al. [11] and Wang et al. [12] identified high TMB as a positive prognostic factor in colon cancer. In contrast, the analysis of patients in the FIRE-3 trial found no prognostic role for TMB [13], while another study demonstrated only a trend suggesting high TMB as a potential positive prognostic factor [14]. Our study contributes to clarify these discrepancies by focusing on a single-center, consecutive, and homogeneous cohort of MSS mCC patients characterized using the same NGS test. Factors such as variability in sequencing platforms, panel design (e.g., the size of the targeted genomic region used to define TMB), patient characteristics, analyzed tissue types (colon vs. rectum, primary vs. metastatic), tumor location (left vs. right-sided primary), and differing thresholds for classifying low versus high TMB may all contribute to the inconsistent and non-uniform results observed across studies. It is interesting to note that the first association emerging from our analysis is with age. The median age tends to be higher in patients with high TMB. This phenomenon has been extensively described elsewhere [24] and partially explains the improved response to ICIs in some tumor types with advanced age. However, in younger patients, hereditary conditions such as Lynch syndrome, which are more prevalent in this population, are often associated with microsatellite instability (MSI) and elevated TMB [25, 26]. In contrast, in sporadic cases involving older patients, aging cells exhibit epigenetic and genetic changes associated with increased replication errors, telomere shortening, and reduced DNA repair efficiency [27, 28]. These factors collectively contribute to the accumulation of somatic mutations and an age-related increase in mutational burden. Additionally, although not a primary objective of our study, an exploratory analysis suggested that high TMB levels might influence the efficacy of first-line anti-VEGF therapy (i.e. bevacizumab). Specifically, patients with low TMB demonstrated a better response compared to those with high TMB. Given the limited number of cases and the heterogeneity of chemotherapy regimens combined with bevacizumab, these results should be interpreted with caution. However, this finding warrants further discussion. In fact, it is known that anti-VEGF drugs and immune checkpoint inhibitors (ICIs) demonstrate synergistic clinical efficacy in terms of response and survival improvements in hepatocellular carcinoma, renal cell carcinoma, and soft tissue sarcomas [29–31]. The mechanistic and molecular bases of this synergy remain unclear. One possible hypothesis is that high TMB may lead to the activation of alternative angiogenic pathways, thereby influencing the response to these therapies. Recent studies have shown that PD-L1 overexpressing non-small cell lung cancers (NSCLCs) secrete angiogenic factors such as angiopoietin-1, platelet-derived growth factor-beta, and fibroblast growth factor-alpha, promoting neo-angiogenesis independent of the canonical VEGF-dependent pathway [32]. It is conceivable that tumors with higher TMB may evade VEGF inhibition by upregulating immunological checkpoints and secreting pro-angiogenic factors, thereby enhancing their survival. These hypotheses are scientifically intriguing and warrant further investigation to elucidate the underlying mechanisms. Moreover, our findings are consistent with previous reports linking the BRAF p.V600E variant with TMB levels in larger cohorts [33]. BRAF p.V600E mutations are more common in microsatellite instability (MSI) tumors, and similar increased mutational propensity and genomic instability observed in high TMB tumors may account for a higher incidence of BRAF mutations [33–35]. However, the molecular basis of these associations remains unclear beyond this hypothesis. Additionally, while BRAF mutations are generally associated with a poor prognosis in colon cancer [36], our study found that high TMB was linked to better survival outcomes. This apparent paradox suggests a complex interplay between TMB and specific genetic alterations, which influences tumor biology and patient prognosis. The impact of different treatments on clinical outcomes should also be considered in this context. The mechanism by which high TMB, independently of MMR protein deficiency, is associated with a good prognosis in MSS patients could be linked to the greater accumulation of somatic mutations, resulting in the generation of neoantigens [37]. A similar phenomenon is believed to be responsible for the better prognosis of metastatic patients with MSI [38]. These neoantigens arise from non-synonymous alterations in the coding regions of the tumor genome and are recognized as foreign by the immune system. Consequently, high TMB tumors possess a broader repertoire of neoantigens, increasing the likelihood of eliciting an immune response. The presence of these neoantigens can lead to tumor cell recognition and subsequent destruction by cytotoxic T lymphocytes (CTLs) and other immune effectors [39] conferring a better prognosis. This subgroup, which represents a larger proportion than the mere 5% of MSI patients, likely has additional prognostic impact and warrants specific biological and immunological characterization. As previously explained, TMB is considered a marker of tumor neoantigen load. High TMB increases the likelihood of generating neoantigens, which can trigger an immune response. However, the immunoediting process plays a critical role in determining whether these neoantigens are successfully targeted or if the tumor evolves mechanisms to evade immune detection [40, 41]. Immunoediting reflects the dynamic interplay between the immune system and tumor evolution. Initially, immune surveillance suppresses tumorigenesis by recognizing and eliminating abnormal cells. However, as tumors accumulate mutations (especially those driving immune evasion) they enter a state of equilibrium, where immune responses are unable to fully eradicate the tumor. Eventually, tumor cells that adapt to immune pressure escape detection, leading to cancer progression. Thus, TMB is intrinsically linked to the immunoediting process. Without intending to delve into exhaustive technical details, it is important to note that the quantification of immunoediting is an emerging biomarker [42, 43] that, in the near future, could integrate and enhance the prognostic and predictive significance of TMB. Some limitations deserve to be highlighted. Firstly, the retrospective nature of the study, although partially mitigated by the single-center design and the homogeneous NGS technique, may introduce potential selection biases related to clinical characteristics such as disease severity, access to care, and treatment eligibility, all of which can influence outcomes. The single-center design may not fully capture the diversity of clinical and demographic characteristics seen in broader, multi-institutional cohorts, which could limit the generalizability of our findings to other clinical settings or geographic regions. However, it is important to emphasize that the single-center design also offers certain strengths, such as the homogeneity of treatments and the completeness of follow-up data, which enhance the robustness of our study. Secondly, although we adopted the FDA-approved cut-off of 10 mut/mb, there is inherent arbitrariness in this threshold, and different studies may use varying cut-offs. Finally, rather than a limitation, a distinctive feature of this cohort is its higher median survival of 50 months compared to the approximately 30 months typically reported for metastatic colon cancer. This is attributable to the patient selection criteria: patients with excellent performance status (85% with ECOG PS 0 and the remainder with PS 1), an estimated life expectancy of at least 3 months, absence of peritoneal disease, and a cachexia risk < 1. These criteria identify patients with uniformly excellent overall clinical conditions. This methodological approach is intended to highlight the prognostic impact of biological factors that may operate independently of the multiple factors related to overall clinical status. Conclusions Our study reinforces the prognostic value of high tumor mutational burden (assessed through next generation sequencing) in microsatellite stable metastatic colon cancer patients. This finding adds to the growing evidence supporting tumor mutational burden as a crucial biomarker in oncology. Supplementary Information Supplementary Material 1 Acknowledgements We thank Daniela Capobianco for technical editing and writing assistance. We would like to express our gratitude to the TRIAL scientific association (CF: 92088670622) for their invaluable and unwavering collaboration on this work. Abbreviations Colorectal cancer Metastatic colorectal cancer Food and Drug Administration Programmed cell death 1 Immune checkpoint inhibitor Microsatellite instability-high Tumor mutational burden-high Deficient mismatch repair Next-generation sequencing Non-small cell lung cancer Memorial Sloan Kettering Cancer Center The Cancer Genome Atlas Kirsten RAt Sarcoma viral oncogene homolog European Society of Medical Oncology Performance status Eastern Cooperative Oncology Group Microsatellite stable Total body Computed Tomography Magnetic Resonance Imaging Response Evaluation Criteria In Solid Tumors Disease control Formalin-fixed and paraffin-embedded Small nucleotide variants Association for Molecular Pathology American College of Medical Genetics Overall survival Progression-free survival Hazard ratios Confidence interval V-raf murine sarcoma viral oncogene homolog B Phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha Tumor Protein p53 Epidermal Growth Factor Receptor Programmed Death-Ligand 1 Vascular Endothelial Growth Factor Median overall survival Chemotherapy Female Male Cytotoxic T lymphocytes Author contributions Conceptualization, A.D.M., M.S., A.O., M.C., G.N.; methodology, A.D.M., M.S., A.O., G.S., M. C; software, A.O., A.C.C., R.S.; validation, R.S., G.S., F.S., A.C., A.O.; formal analysis, R.S., G.S., A.O.; investigations, A.D.M., M.S., R.S., U.P., R.P., A.M.C., M.B., G.S., G.F., M.C.; resources, A.O.; data curation, A.D.M., M.S.; writing-original draft preparation, A.D.M., M.S., A.O.; writing-review and editing, G.N., M.C.; supervision, M.C., A.O. All authors have read and agreed to the published version of the manuscript. Funding This work was supported by grants from the Italian Government, Ministry of Health (Ricerca Corrente L4/8) and by the “Ministero delle Imprese e del Made in Italy” project number F/310034/01-03/X56, “Epi-MET—Funzionalizzazione delle aberrazioni (epi)genomiche nei tumori metastatici.” We also extend our deepest gratitude to the late Mrs. Antonietta Nacca, a private citizen, for her generous and unconditional financial support. Availability of data and materials The genetic data supporting the findings of this study are available upon reasonable request, which can be sent to giovanni.savarese@centroames.it. Declarations Ethics approval and consent to participate The Ethical Committee of the University of Campania “Luigi Vanvitelli” approved this study with the numbers 790 and 68 for project i-Cure. All patients provided written informed consent before undergoing treatment administration and molecular assessments. Consent for publication Not applicable. Competing interests Roberto Sirica, Giovanni Savarese, and Monica Ianniello are employed by AMES, Centro Polidiagnostico Strumentale srl, 80013 Naples, Italy. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Footnotes The original version of this article was revised: Ethics approval and consent to participate declaration was published incorrectly. It is now as follows: Ethics approval and consent to participate In accordance with local and national laws, ethics approval and consent to participate were exempted due to the retrospective/observational and noninterventional nature of the work. However, all patients involved in this study provided written informed consent before undergoing treatment administration and molecular assessments. It should be as follows: Ethics approval and consent to participate The Ethical Committee of the University of Campania “Luigi Vanvitelli” approved this study with the numbers 790 and 68 for project i-Cure. All patients provided written informed consent before undergoing treatment administration and molecular assessments. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Annabella Di Mauro and Mariachiara Santorsola have contributed equally to this work. Michele Caraglia and Alessandro Ottaiano are co-last authors. Change history 1/8/2025 A Correction to this paper has been published: 10.1186/s12967-024-05971-5 Contributor Information Michele Caraglia, Email: michele.caraglia@unicampania.it. Alessandro Ottaiano, Email: a.ottaiano@istitutotumori.na.it. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement The genetic data supporting the findings of this study are available upon reasonable request, which can be sent to giovanni.savarese@centroames.it.