Colorectal cancer (CRC) is the second leading cause of cancer-related mortality in men and the third in women within the European Union (1). At the time of diagnosis, 15 to 25% of patients already have metastatic disease, and an additional 20% develop metastases during follow-up. Despite advancements in treatment over the past decade, metastatic CRC (mCRC) remains incurable, with a median overall survival (OS) exceeding 30 months (2). mCRC is a heterogeneous disease, and the European Society for Medical Oncology (ESMO) guidelines emphasize the importance of tailoring therapy based on patient and tumor characteristics (3, 4). First and second-line treatments are based on chemotherapy regimens incorporating three cytotoxic agents: 5-fluorouracil, irinotecan, and oxaliplatin. These are used either as monotherapy or in combination with targeted therapy added depending on RAS/BRAF status (5–8). Treatment selection depends on the patient’s general condition, treatment goals, surgical resectability, and microsatellite instability (MSI) and RAS/BRAF molecular testing results (9). Inflammation plays a crucial role in the development of various diseases, including cancer, highlighting the need to explore markers that can assess systemic inflammation. Chronic inflammation, characterized by immune activity and the presence of cytokines, growth factors, and chemokines, can promote tumor initiation and support tumor growth, survival, and angiogenesis (10). Recently, there has been increased recognition of the tumor microenvironment’s significance, with the inflammatory response being considered a key factor in tumor progression and metastasis (11). Furthermore, systemic inflammation has been identified as a negative prognostic indicator in many malignancies (12). The systemic inflammatory response can be evaluated using blood cell counts and ratios, including the neutrophil-to-lymphocyte ratio (NLR) and the systemic immune-inflammation index (SII), as well as markers such as C-reactive protein (CRP) and serum albumin. NLR and SII are simple, low-cost biomarkers derived from routine complete blood counts (CBCs) and have been found to have prognostic value, correlating with poorer survival outcomes in patients with various tumor types, including CRC (13, 14). Elevated NLR is not specific to cancer and can be seen in various conditions, including infections, chronic inflammatory diseases, and cardiovascular events (15). However, there is currently no consensus on specific threshold values for NLR, and its prognostic significance varies across different stages and histological types of solid tumors (13, 15), as well as concerning clinical data. In stage II–III CRC a high NLR has been associated with poorer outcomes, likely reflecting greater tumor proliferative activity. In contrast, its predictive value in stage IV disease remains uncertain (16). On the other hand, the prognostic role of the SII in malignant disease has received less attention in research, especially in stage IV (14). In CRC patients who underwent surgery, SII was identified as an independent predictive factor for OS, demonstrating superior diagnostic performance compared to NLR (17, 18). Limited research has focused on the clinical application of NLR and SII specifically in mCRC, particularly in the context of first-line treatments. Consequently, there is a lack of data regarding the use of NLR and SII for patient selection and monitoring in modern treatment protocols, which include targeted therapies (19–21).

Cancer-associated inflammation plays a significant role in disease progression and is linked to poor outcomes in patients undergoing systemic therapy for metastatic disease (24). This process involves various inflammatory cells that produce inflammatory mediators. Neutrophils can support tumor growth, angiogenesis, and metastasis, and may also contribute to treatment resistance. They are functionally plastic and can adopt either pro-tumor (N2) or anti-tumor (N1) phenotypes, depending on the tumor microenvironment. Platelets similarly promote angiogenesis and help tumor cells evade immune surveillance. In contrast, lymphocytes are crucial for anti-tumor immunity and the clearance of tumor cells (25). While many inflammatory markers have been studied for their prognostic significance, there is still insufficient data on their predictive value in patients receiving first-line treatment for mCRC. NLR was the first immune-related marker extensively researched in various tumors, showing in CRC to be predictive of survival (26). Also, it was reported as an independent prognostic factor in patients receiving first-line chemotherapy for mCRC (27). SII is a promising biomarker that incorporates neutrophils, platelets, and lymphocytes into a single score, where high values may indicate a pro-tumor inflammatory state with a compromised immune response (16).

In this study, we conducted serial assessments of NLR and SII to investigate their prognostic significance for survival outcomes and treatment response among 232 patients with mCRC who received first-line chemotherapy, with targeted therapy added in most cases (85.4%). Specifically, we analyzed both baseline (pretreatment) and 3-month post-treatment values of NLR and SII to determine their prognostic value. Our results indicate that post-treatment NLR and especially SII are stronger prognostic factors of survival, particularly PFS on first-line chemotherapy. Patients with high NLR and SII after treatment had significantly shorter PFS compared to those with lower values. Multivariate analysis confirmed that pretreatment NLR and SII are independent prognostic factors for OS, but not for PFS. Regarding post-treatment assessment, NLR appears to be an independent prognostic factor for PFS, but not for OS. Post-treatment SII is independently associated with improved OS and PFS. This suggests that post-treatment values of NLR and even more of SII better reflect treatment response and survival outcomes than baseline values alone. Therefore, serial assessment of NLR and SII early during treatment could provide valuable prognostic information and support more timely and informed therapeutic decisions.

Similar findings have been reported by Liu et al. with NLR on a smaller cohort of patients (n = 127) receiving FOLFOX chemotherapy for mCRC, showing that elevated baseline NLR was associated with poor PFS, but not OS. However, the change from high to low post-treatment NLR values was associated with treatment benefit and improved survival. Patients included in this study did not receive any targeted therapy (28). In a study conducted by Kim et al., the pattern of change in NLR was evaluated, demonstrating that a reduction in NLR served as an independent predictor of chemotherapy response and improved survival outcomes. Conversely, a persistently elevated NLR, or a shift from a low to a high NLR, was associated with poorer survival. This study focused on the chemotherapy regimen and did not explore the prognostic role of targeted therapy (19). Similarly, another study involving 71 patients investigated the dynamic changes in inflammatory indices (NLR, platelet-to-lymphocyte ratio, PLR, lymphocyte-to-monocyte ratio, LMR) during first-line chemotherapy and their prognostic significance in patients with mCRC. This study found that among the evaluated indices, only NLR was significantly correlated with PFS and OS, while no association was observed between NLR dynamics and response to treatment. The prognostic value of SII has not been investigated in this research (29). In study by Kocak et al., prognostic value for dynamic changes in NLR or PLR was not observed, while improvements in the prognostic nutritional index (PNI) were associated with longer survival. The PNI, calculated from serum albumin and lymphocyte count, reflects the patient’s immunonutritional status and has repeatedly been shown to correlate with outcomes in mCRC, with higher values indicating better treatment tolerance and more favorable tumor biology (30). Although our study did not specifically analyze dynamic changes, our findings align with these observations by demonstrating that post-treatment inflammatory status has greater prognostic significance than pretreatment levels alone, underscoring the need for longitudinal assessment.

While our results were independent of RAS/BRAF mutational status, conflicting data from Miyamoto et al. suggest that systemic inflammation biomarkers are predictive in the KRAS wild-type population but not in the KRAS-mutated population. However, this study was unable to conclude whether the selection of targeted therapy was effective (31). On the other hand, two retrospective studies concluded that pre-treatment inflammatory indices can predict the efficacy of cetuximab therapy, with elevated NLR being associated with shorter PFS and OS. Other indices, including the SII, did not demonstrate a statistically significant correlation with survival outcomes. Both studies lack data on post-treatment values and focus on a highly selected mCRC population – RAS wild-type patients treated with cetuximab (32, 33). Conversely, Passardi et al. identified both SII and, more prominently, NLR as prognostic and predictive biomarkers of survival in patients treated with chemotherapy plus bevacizumab in two separate studies. Notably, none of the aforementioned studies assessed the prognostic value of post-treatment inflammatory indices in relation to survival (34, 35). Anti-VEGF therapy can reduce tumor-driven angiogenesis and indirectly attenuate pro-inflammatory signaling, while anti-EGFR therapy may alter tumor cell proliferation and downstream cytokine production (36). However, despite these potential differences, no significant variation in PFS or OS was observed between patients receiving anti-VEGF therapy or anti-EGFR therapy in our cohort. Likewise, the distribution of NLR and SII at baseline and during follow-up did not differ meaningfully between these groups. In our study, SII appeared to be a more robust prognostic factor than NLR. An improvement in PFS was observed in patients receiving targeted therapy, with no benefit in OS. However, this could be influenced by other factors not evaluated in this study, as patients who did not receive targeted therapy may have had contraindications for it. Low pre-treatment NLR and SII were associated with a better response to treatment.

A recent meta-analysis has demonstrated that SII is linked to poorer outcomes in patients with CRC. However, in stage IV patients, the association was not statistically significant. Nonetheless, none of the included studies assessed the temporal patterns of inflammatory biomarkers or the potential impact of treatment on outcomes. The authors suggest that longitudinal assessment in future research could improve the utility of inflammatory indices (20). Conversely, Tan et al. conducted a meta−analysis showing that pre−treatment SII was more strongly associated with both PFS and OS than post−treatment values, suggesting that the prognostic impact of SII may be attenuated following therapy. The studies included in the analysis were heterogeneous, spanning stages I-IV, and only 6 out of the 27 studies investigated chemotherapy combined with targeted therapy as the treatment modality (37). In contrast, our findings highlight the importance of serial assessment of NLR and SII in mCRC, supporting the need for stage-specific studies to better define the temporal evolution of these inflammatory markers during systemic treatment.

We also examined other factors that may have an impact on survival and the prognosis of mCRC patients. In our cohort, improvement in OS and PFS was also influenced by the number of metastatic sites and the use of local ablative treatment. Local ablative treatments, including metastasectomy, MWA, and SBRT, may improve outcomes in selected patients with mCRC by reducing overall tumor burden, delaying systemic progression, and extending the interval to subsequent systemic therapy. These interventions are generally reserved for patients with oligometastatic disease, favorable performance status, and favorable tumor biology, which are associated with improved survival. Beyond their cytoreductive effects, local therapies may influence systemic inflammation and alter NLR and SII (38, 39). While this study was not designed to determine the independent effects of local therapy, the potential interaction between local interventions and systemic inflammatory markers should be considered when interpreting the prognostic value of longitudinal NLR and SII trends. Conversely, liver and/or peritoneal involvement did not show an association with OS, although patients with liver involvement had shorter PFS. Liver metastases represent the most prevalent site in mCRC, followed by metastases to the lungs and the peritoneum (40). They are known to negatively impact survival, with the number of lesions affecting the outcome (41). Peritoneal metastases represent adverse prognostic and predictive factors, as patients with this pattern of spread derive limited benefit from systemic chemotherapy (42). Also, patients with left-sided tumors experienced prolonged OS. Resection of the primary tumor had a significant impact on OS only when analyzed with post-treatment NLR and SII, but not in pretreatment assessment. The impact of primary tumor resection on survival in patients with mCRC remains controversial, likely due to the heterogeneity of the populations, although patients with left-sided tumors do experience prolonged survival (43). Moreover, the systemic inflammatory response induced by primary tumor resection may enhance immunosuppression and potentially accelerate metastatic progression (44). Most patients in our study were in good general condition, with an ECOG PS of 0. A poor ECOG PS has been identified as a risk factor for worse survival and increased toxicity, and it is included in treatment guidelines for mCRC (4). However, most studies group ECOG PS 0 and 1 patients as having a good general condition, although patients with ECOG PS 0 show significantly better treatment outcomes, both in terms of survival and a reduction in serious adverse events (45). In our analysis, ECOG PS was associated with improved OS in the univariate analysis, although this association did not remain significant in the multivariate model.

Finally, there is considerable variability in the cutoff values for the NLR across studies, with thresholds ranging from 2 to 5. These thresholds have been established using various methods, such as receiver operating characteristic (ROC) curves, median or mean values from study populations, or reference to previously published data (46). In contrast, cutoff values for the SII are even more limited and heterogeneous, ranging from 340 to over 6000 (17, 37). Although ROC curves provide an objective method for defining cut-offs, their implementation in routine clinical practice remains challenging. Despite consideration of ROC-based cut-offs, the cohort heterogeneity, variability in PFS/OS endpoints, and the absence of a universally validated reference cut-off for NLR/SII in metastatic disease led to limited clinical interpretability. Because our objective was to evaluate prognostic trends rather than define clinically adoptable thresholds, we selected the median as a robust, distribution-based, and non-model-dependent method, which is frequently applied in similar retrospective biomarker studies. Additionally, NLR and SII values may be influenced by factors such as age, race, and comorbidities, further complicating the determination of standardized cutoff values (24). Despite these limitations, NLR and SII possess a prognostic role, and combining them with other validated biomarkers (RAS and BRAF molecular testing, and carcinoembryonic antigen (CEA) monitoring) and clinical parameters (such as ECOG PS, radiological assessment) in routine clinical practice for individual prognosis and treatment selection could enhance long-term prognostic assessments (19, 26). Furthermore, integrating inflammatory biomarkers derived from CBC with multi-omics approaches, such as interleukin-27 receptor alpha (IL27RA) expression, may enhance the assessment of the inflamed tumor microenvironment and its prognostic relevance (47).

Unlike the mentioned studies, which focused predominantly on pre-treatment inflammatory indices, this study emphasizes the prognostic significance of serially assessing NLR and SII at 3 months post-treatment to better reflect treatment response and survival outcomes. Baseline inflammatory markers reflect a composite of tumor-related inflammation, comorbid conditions, and pre-treatment physiological state, and therefore may not accurately represent the biology driving subsequent disease behavior (25). In contrast, post-treatment NLR and SII capture the host response after the first cycles of systemic therapy, integrating both tumor sensitivity to treatment and the degree of treatment-induced or residual systemic inflammation. Patients in whom inflammatory markers decrease after treatment likely experience effective cytoreduction and reduced tumor-driven myelopoiesis, whereas persistently elevated values may indicate ongoing tumor activity, inadequate response, or a sustained pro-tumor inflammatory milieu (19). This dynamic information appears to provide more accurate prognostic insight than baseline measurements alone, which is consistent with the superior predictive performance of 3-month post-treatment NLR and SII in our cohort. By conducting serial assessments of NLR and SII early in the treatment process, we can gain valuable prognostic information that may lead to more timely and informed therapeutic decisions. This aspect has not been extensively explored in previous research.

This study has several limitations that should be acknowledged. The retrospective, single-center design introduces potential selection bias and limits external validity, as patient management and data collection may not reflect broader practice patterns. Heterogeneity in patient characteristics and therapeutic regimens may also confound outcomes. Furthermore, the absence of standardized cut-off values for these markers reduces comparability across studies and may impair generalizability.

A better understanding of the role of systemic inflammation in the prognosis of mCRC and response to treatment could improve patient selection for a more personalized therapeutic approach. Inflammatory biomarkers derived from CBC are readily available, robust, and cost-effective prognostic factors for outcomes in patients with mCRC. Consequently, there is a need for further investigation into the role of NLR and SII in identifying high-risk patient groups and guiding the treatment of mCRC. To validate the prognostic value of NLR and SII and their dynamics, future prospective trials should focus on specific mCRC patient subpopulations based on disease biology and treatment modalities. Additionally, it should be investigated whether a serial assessment of NLR and SII has prognostic value in later lines of treatment.