Esophageal cancer (EC), one of the most lethal malignancies worldwide, represents the seventh predominant cause of global mortality related to cancer (1). Though there have been notable advancements in therapeutic approaches and critical care in recent decades, the EC population demonstrates one of the least favorable five-year survival outcomes in contrast to other cancers (2). The GLOBOCAN 2020 global cancer statistics showed approximately 604,000 new and 544,000 dead cases attributable to EC, ranking seventh and sixth in incidence and mortality, among all malignancies (3). Notably, nearly half of all new cases and deaths occurred in East Asia (4). Its treatment presents unique challenges owing to the anatomical proximity of the esophagus to the airway and main blood vessels, no serosal layer, and its rich surrounding lymphatic network. Furthermore, patients are commonly diagnosed at a late stage and frequently malnourished, further complicating treatment (5). We accurately determined the stage of esophageal cancer according to the international esophageal cancer diagnostic guidelines, which is important for us to choose the appropriate treatment, so that we can provide individualized treatment pathways for patients with different stages of esophageal cancer (6).In the early stage, surgical resection is still the mainstay and most effective approach. However, due to the insidious onset and aggressive nature of EC, many cases are detected at advanced stages, thereby missing the best opportunity for curative surgery (7). With a better understanding of the disease, various treatment strategies have been developed for the locally advanced EC population, like neoadjuvant chemoradiotherapy (NCRT), neoadjuvant chemotherapy (NCT), as well as immunotherapy with surgery (8). Evidence suggests that NCRT significantly improves overall survival (OS) in locally advanced EC, with clinical benefits observed across different histological subtypes (9); Similarly, several studies have reported a survival advantage with NCT in this patient population (10). In terms of recurrence-free survival (RFS), Zhang et al. demonstrated that neoadjuvant treatment - whether NCRT or NCT - contributes positively to improved outcomes (11); Furthermore, Lewis et al. found that neoadjuvant therapy increases the pathological complete response (pCR) rate compared to surgery alone (12). Hence, early and accurate prognostic assessment is critical for guiding individualized treatment decisions in clinical practice.

Tumor progression is influenced by both tumor-specific factors and the host immune response (13, 14). Therefore, systemic inflammatory markers are prospective prognostic indicators. A low neutrophil-to-lymphocyte ratio (NLR) influences favorable outcomes in various malignancies, including non-small cell lung cancer, gastric cancer, colorectal cancer, and hepatocellular carcinoma (15, 16). Systemic inflammation can be assessed via alterations in peripheral blood cell counts of lymphocytes, monocytes, neutrophils, and platelets. Composite hematological indices, like the NLR and platelet-to-lymphocyte ratio (PLR), have been proposed as accessible prognostication biomarkers. In 2021, Li et al.’s study, which included 127 patients, NLR was a dependable indicator of prognosis for EC sufferers undergoing neoadjuvant therapy (17). Conversely, a contemporaneous study by Anand et al. reported limited predictive value of NLR in this context (18). These conflicting findings underscore the need for further investigation into the prognosis utility of NLR among the EC population after neoadjuvant therapy.

To date, a substantial number of both retrospective and prospective investigations have examined the predictive value of NLR in this context, yet a comprehensive meta-analysis to consolidate the available research findings remains lacking. Therefore, our study unveiled the prognosis relevance of NLR among EC individuals taking neoadjuvant therapy via a systematic review and meta-analysis. GRADE methodology and subgroup analyses were employed to assess the evidence quality and identify possible sources of heterogeneity. Ultimately, this study seeks to present evidence-based recommendations for the clinical adoption of NLR as a readily available hematological biomarker to facilitate prognostic stratification and guide therapeutic decision-making, ameliorating outcomes for the EC population.

Inflammation, a key feature of the tumor microenvironment, is crucial in cancer initiation, cell proliferation, angiogenesis, as well as metastasis during tumor progression (34, 35). Several studies have demonstrated that inflammatory markers hold prognostic value in cancer treatment and outcome prediction (36). Among them, NLR has been a proven prognostic indicator. In non-small cell lung, colorectal, gastric, ovarian cancers, hepatocellular and renal cell carcinomas (37–42), NLR is related to patient prognosis. Zhang CL et al. identified that NLR notably predicts survival and treatment response among the ovarian cancer population (43). Similarly, Mi HL et al.’s retrospective research on NLR in lung cancer proved that higher NLR can independently be adopted for forecasting reduced RFS (44). These findings suggest that NLR, a non-invasive, easily accessible, and cost-effective hematologic marker, may provide valuable prognostic information across various types of cancer.

This meta-analysis, which included 2,220 patients, demonstrated that NLR evidently correlates with OS, RFS, and pCR among the EC population. These findings have been validated by other studies. For instance, a 2021 study reported that among EC patients who received NCRT, those who achieved pCR exhibited evidently lower NLR levels (HR = 1.218, 95% CI = 1.050–1.414, P = 0.009) (45). Another meta-analysis published in 2024 unveiling NLR’s prognostic relevance in EC also proved that elevated NLR is markedly linked to worse OS (HR = 1.47, 95% CI = 1.29–1.67, P < 0.00001) (46). These consistent results reinforce our conclusion that NLR is a qualified and reliable prognostic indicator. As a cost-effective and easily obtainable hematological marker, NLR can facilitate the rapid construction of prognostic models and serve as a feasible component in the development of comprehensive prognostic prediction tools. Our subgroup analysis showed that NCRT offers greater prognostic benefits than NCT among the EC population. Furthermore, the subgroup analyses for RFS across various parameters demonstrated robust statistical power. Prognostic models constructed through this approach can assist clinicians in accurately assessing patient outcomes and adjusting treatment strategies in a timely manner, thereby potentially improving prognosis in high-risk individuals.

Our study elucidated the link of NLR to OS and RFS in EC sufferers and included subgroup analyses by sample size, follow-up duration, age, and treatment regimen. For OS, NLR showed a significant association among patients with a follow-up length of ≤36 months, while no marked link was observed in those with a follow-up lasting over 36 months. This may be due to increased loss to follow-up over longer durations, which can bias statistical analyses in retrospective studies and reduce the significance of subgroup findings with extended follow-up. Among patients aged <60, higher NLR was linked to shorter OS, whereas no significant association was found in patients aged >60. Rui et al. (47) demonstrated that the mortality risk of EC increases with age. Aging is accompanied by a gradual decline in immune function and cumulative exposure to risk factors, both of which contribute to higher mortality. This could account for the lack of statistical significance in NLR’s prediction value for OS among patients aged above 60. Notably, in the NCRT cohort, the prognostic effect of NLR (HR = 2.84, 95% CI = 1.58–5.12, P = 0.0005) was stronger than the NCT receivers(HR = 1.58, 95% CI = 1.15–2.18, P = 0.004). Similarly, for RFS, the link of NLR to prognosis was stronger in the NCRT group (HR = 3.25, 95% CI = 1.68–6.28, P = 0.0005) than in the NCT subgroup (HR = 2.07, 95% CI = 1.37–3.28, P = 0.002). Multiple studies have shown that perioperative NCRT, compared to NCT alone, offers superior benefits in terms of prognosis and postoperative complications among the EC population (48, 49). In our RFS subgroup analyses, all predefined subgroups demonstrated a significant link between NLR and patient outcomes. This consistent cross-subgroup finding further supports the reliability of NLR as a prognosis biomarker in EC, suggesting that its predictive value is largely independent of these clinical variables. Of note, our subgroup analysis also demonstrated that the subgroup with an NLR cut-off value > 3 exhibited better predictive performance for RFS compared to the subgroup with an NLR cut-off value ≤ 3. The superior predictive ability of an NLR cut-off > 3 may be attributed to the fact that a higher threshold more effectively identifies patients with severe systemic inflammation, which is closely associated with tumor recurrence.

This study explored the link of NLR to the prognosis of EC and conducted a comprehensive sensitivity analysis to ensure result robustness and reliability. All sensitivity analyses were performed utilizing a random-effects model. The stability of our pooled estimates was examined utilizing a leave-one-out approach. For NLR and RFS (I² = 47%, P = 0.13), heterogeneity was relatively low, and the effect sizes remained the same within the original range upon sequential exclusion of individual studies, indicating good stability in the RFS-related findings. In contrast, for NLR and pCR (I² = 62%, P = 0.07) and NLR and OS (I² = 88%, P < 0.00001), significant heterogeneity persisted even after stepwise study ostracization. This possibly arises from variations in study populations from various countries and regions, variability in study periods, and the relatively small sample sizes, which introduce both temporal and geographic heterogeneity. Taken together, the sensitivity analysis suggests that, despite differences in patient characteristics and study design, the prognostic value of NLR remains robust. Probable publication bias was examined utilizing Egger’s test. No publication bias existed in the analyses of OS, RFS, and pCR in relation to NLR. Furthermore, GRADE assessments were executed and evidence was rated as “very low” for each outcome. This was mainly due to the observational nature of eligible studies, which are inherently prone to methodological limitations. Therefore, Additional high-quality prospective research is necessitated for verifying and strengthening the evidence base for the prognostic utility of NLR in this setting. Ultimately, the goal is to provide clinicians with better tools for making informed decisions on the post-neoadjuvant treatment of EC sufferers. From the perspective of outcome stability, our findings further confirm that elevated NLR is significantly associated with worse OS and RFS, while its association with lower pCR rates is less stable due to the influence of individual studies; this is consistent with the sensitivity analysis result that excluding the McLaren P. J (2017). study altered the statistical significance of the pCR-related association.

Neutrophils exhibit an intricate role in the tumor microenvironment. Once recruited by tumor cells into the tumor milieu, neutrophils become tumor-associated neutrophils (TANs): N1 and N2 phenotypes. TANs support tumor growth primarily by releasing proliferative factors, suppressing T-cell activity, and promoting tumor angiogenesis (50). In contrast, lymphocytes are central to antitumor immunity. CD8⁺ T cells are cytotoxic against tumor cells via the secretion of IFN-γ and TNF-α, while CD4⁺ T cells activate antigen-presenting cells and enhance the cytotoxic effects of both CD8⁺ T cells and NK cells, thereby inhibiting tumor growth and metastasis (51, 52). Lymphocyte count decrease is often connected with poor prognosis among the EC population (53). A risen NLR usually reflects a relative elevation in neutrophils and/or a drop in lymphocytes. Neutrophils suppress the activity of lymphocytes and NK cells, diminishing antitumor immune responses, while also contributing to systemic inflammation through the release of pro-inflammatory mediators. Conversely, lymphopenia weakens the ability of the immune system to target and eliminate tumor cells, thereby facilitating tumor invasion and metastasis (54). Lymphocytes play crucial roles in restraining tumor cell proliferation, and metastasis, and promoting cytotoxicity and apoptosis, whereas neutrophils facilitate tumor progression by producing growth factors, chemokines, and proteases (55). When elevated neutrophil counts and reduced lymphocyte counts co-occur, the predictive value for tumor progression may be further enhanced. Thus, NLR is regarded as a potential prognostic biomarker for EC sufferers following neoadjuvant therapy. Despite presenting valuable insights, this meta-analysis has limitations. Most incorporated studies were retrospective, with only two being prospective. The retrospective design may introduce uncontrolled confounding factors that could bias the conclusions. Moreover, most were conducted in Asia and Europe, suggesting potential geographic limitations. Therefore, broad generalizations to non-studied populations should be approached with caution. Further research is required to validate NLR’s prognostic relevance among the EC population undergoing neoadjuvant therapy in other geographic settings. Additionally, variability in the cutoff values and timing of NLR measurement may introduce heterogeneity and confounding effects that are difficult to control. Differences in treatment regimens across studies may also contribute to the inconsistency in results. Therefore, subsequent researchers should focus on standardizing when NLR is measured, selecting consistent treatment approaches, and determining clear, measurable NLR thresholds.

In conclusion, following neoadjuvant therapy, elevated NLR is markedly linked to poor prognosis among the EC population, including reduced OS and RFS and lower pCR. Therefore, NLR is a low-cost, non-invasive, and easily accessible biomarker for forecasting the prognosis of EC sufferers. It can aid clinicians in timely developing optimal treatment plans based on the potential risk of poor outcomes. Notably, this hematological biomarker may be valuable for forecasting the prognosis of EC in underdeveloped or resource-limited regions, especially in clinical settings where genomic monitoring is not feasible. However, due to some inevitable limitations and confounding factors in the included studies, it is recommended to carry out a multi-center prospective study to standardize the NLR measurement time, such as baseline and cutoff values before treatment, in order to confirm the conclusions of this study.