Introduction
Esophageal cancer (EC) is one of the most common gastrointestinal cancers and a leading cause of cancer-related mortality [1]. Although the incidence of esophageal adenocarcinoma has been rising, particularly in developed countries, esophageal squamous cell carcinoma (ESCC) remains the predominant histological subtype worldwide, especially in Asia [2,3].
Although the tumor–node–metastasis (TNM) classification is the primary tool for assessing prognosis and guiding clinical treatment in EC including ESCC, it does not fully address the need for individualized treatment strategies [4]. Therefore, convenient and reliable biomarkers for prognostic evaluation alongside TNM classification are desirable to support personalized treatment planning. Recent studies have suggested a strong link between the host inflammatory response and cancer progression [5-7]. Among inflammatory markers, the neutrophil-to-lymphocyte ratio (NLR)—calculated by dividing the neutrophil count by the lymphocyte count—is one of the most well-established indicators, with reported associations with tumor progression and prognosis in various solid cancers, including EC [8-12]. However, existing studies on NLR in EC including ESCC have exclusively focused on patients treated with photon radiotherapy [13-15].
Proton beam therapy (PBT), a form of radiotherapy, delivers radiation via a characteristic depth-dose curve, with the highest dose deposited at the Bragg peak and a rapid dose fall-off beyond that point [16]. This unique dose distribution considerably reduces radiation exposure to surrounding normal tissues compared with photon radiotherapy, potentially lowering treatment-related toxicity [17]. In particular, cardiac irradiation is associated with cardiac complications and cardiac-related death in patients with EC including ESCC [18,19]. Because PBT can reduce cardiac irradiation dose, volume, and complications compared to photon radiotherapy [20], the relationship between NLR and treatment outcomes in patients with ESCC receiving photon radiotherapy may not necessarily apply to those treated with PBT. Therefore, this study aimed to evaluate the association between pre-treatment NLR and treatment outcomes in patients with ESCC undergoing PBT.
Materials and Methods
1. Patients
Consecutive patients with ESCC who underwent PBT with concurrent chemotherapy at our institution between January 2015 and January 2020 were included in this study. Clinical staging was determined according to the 8th edition of the TNM classification established by the Union for International Cancer Control (UICC) [4]. The study was approved by the institutional ethics committee, and the requirement for informed consent was waived (IRB No. 2017-440).
2. Treatment
A total radiation dose of 60–66 Gy (relative biological effectiveness [RBE]) (median, 60 Gy [RBE]) was prescribed to the planning target volume and delivered in 30–33 fractions using passive scattering PBT. The number of irradiation fields of PBT was 2 ports in principle; however, the number of irradiation fields was increased to achieve dose constraints for critical organs at risk in some cases. Treatment planning was based on computed tomography (CT), 18F-fluorodeoxyglucose positron emission tomography/CT, and endoscopic findings.
Gross tumor volume included the primary tumor (GTVp) and metastatic lymph nodes (GTVn) identified on CT and/or endoscopy. Clinical target volume (CTV) for the primary tumor was defined as the GTVp plus a 2-cm margin in the craniocaudal direction and a 0 to 0.5-cm margin in the other directions, whereas for the metastatic lymph nodes it was defined as GTVn plus a 0 to 0.5-cm margin. Prophylactic regional lymph nodes were also added as needed for CTV. Planning target volume was defined as the overall CTV plus a 0.5-cm margin in all directions to account for respiratory motion.
Concurrent chemotherapy was administered to all patients. The cisplatin plus fluorouracil regimen was the primary regimen, with two dosing variations: (1) cisplatin 75 mg/m2 intravenously + fluorouracil 1,000 mg/m2 intravenously (q28, one cycle) and (2) cisplatin 70 mg/m2 intravenously + fluorouracil 700 mg/m2 intravenously (q28, one cycle). Other regimens included oxaliplatin, leucovorin, and 5-fluorouracil, nedaplatin plus fluorouracil, or fluorouracil alone. Induction chemotherapy—administered before PBT—was used for patients who initially planned to undergo surgery but later declined. Induction chemotherapy regimens included (1) cisplatin plus fluorouracil regimen, (2) docetaxel, cisplatin, and 5-fluorouracil regimen, and (3) paclitaxel, cisplatin, and 5-fluorouracil regimen. In addition, consolidation chemotherapy—administered after PBT—was used for some patients who were in clinical stages Ⅱ–Ⅳ and had received no induction chemotherapy. Consolidation chemotherapy regimen extended the concurrent chemotherapy regimen.
3. Blood examination
Laboratory tests were performed within 0–14 days before initiation of PBT (median, 1 day). The NLR was calculated as the ratio of total neutrophils to total lymphocytes and was used as a biomarker of systemic inflammation.
4. Statistical analysis
All statistical analyses were conducted using JMP software version 14.3.0 (SAS Institute, Cary, NC, USA). The Kaplan-Meier method was used to estimate overall survival (OS) and progression-free survival (PFS) rates. OS was defined as the time from the start of PBT to death from any cause; in contrast, PFS was defined as the time from the initiation of PBT to death from any cause or disease progression. Disease progression was assessed by radiologists and gastroenterologists using the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 [21] and/or histopathological confirmation of local tumor progression. Patients lost to follow-up were censored at their last clinical visit. To determine prognostic factors, the Cox proportional hazards model was used to calculate hazard ratios (HRs), 95% confidence intervals (CIs), and p-values. Although no established NLR cutoff values exist for patients with EC undergoing PBT with concurrent chemotherapy, in a systematic review, the median NLR cutoff value for EC prognostic markers was “3.00” [12]. In this study, receiver operating characteristic curve analysis was performed to determine the optimal cutoff for predicting OS. Additionally, Fisher’s exact test was used to analyse the likelihood of a complete response based on RECIST 1.1 and the occurrence of acute and late adverse events of grade ≥3 according to the Common Terminology Criteria for Adverse Events version 5.0. Statistical significance was set at p < 0.05.
Results
A total of 133 patients with ESCC underwent PBT with concurrent chemotherapy at our institution between January 2015 and January 2020. Of these, 17 patients were excluded: 14 who did not receive concurrent chemotherapy and three who lacked laboratory data within 2 weeks before the start of PBT. Consecutive 116 patients who underwent definitive PBT with concurrent chemotherapy were retrospectively evaluated. Patient characteristics are summarized in Table 1. Consolidation chemotherapy was administered for 44 patients; however, this factor could not be included in the analyses because of the immortal time bias.
The area under the receiver operating characteristic curve for NLR in predicting OS was 0.65, with a sensitivity of 71.4% and a specificity of 62.1%. The optimal NLR cutoff was identified as 3.02. Since this NLR 3.02 is essentially equal to previously reported NLR cutoff values of 3.0, NLR of 3.0 was set as the cutoff for this study. Patients were stratified into two groups: those with an NLR ≥3.0 (high NLR group) or those with an NLR <3.0 (low NLR group).
The median follow-up time for OS was 51.2 months (range, 3.0 to 114.0), and the median survival follow-up time was 55.6 months (range, 3.0 to 114.0). At the time of analysis, 21 patients had died. The 1-year and 3-year OS rates were 89.5% and 82.4%, respectively (Fig. 1A). In univariate analysis, an Eastern Cooperative Oncology Group Performance Status (ECOG PS) ≥1 (HR, 3.76; 95% CI, 1.58 to 8.97; p = 0.003), clinical stage Ⅳ (HR, 3.11; 95% CI, 1.28 to 7.56; p = 0.012), tumor length ≥6 cm (HR, 3.60; 95% CI, 1.32 to 9.85; p = 0.013), and high NLR (HR, 4.76; 95% CI, 1.84 to 12.29; p = 0.001) were significantly associated with poorer OS (Table 2, Fig. 1B, 1C). Additionally, in ESCC patients who did not receive induction chemotherapy (n = 93), high NLR (HR, 3.98; 95% CI, 1.51 to 10.52; p = 0.005) was also significantly associated with poorer OS (Fig. 1D). In multivariate analysis, an ECOG PS ≥1 (HR, 2.52; 95% CI, 1.00 to 6.34; p = 0.049), clinical stage Ⅳ (HR, 3.37; 95% CI, 1.30 to 8.77; p = 0.013), and high NLR (HR, 3.89; 95% CI, 1.36 to 11.19; p = 0.012) remained significant independent predictors for poor OS (Table 2). Among patients with both clinical stage IV disease and a high NLR, the 1-year OS rate was 24.6% (HR, 17.54; 95% CI, 2.08 to 147.49; p = 0.008) (Supplementary Fig. S1A–S1C).
A complete clinical complete response (cCR) was observed in 75 patients (64.7%); in contrast, 41 patients (35.3%) had residual disease (non-cCR). The cCR rate was significantly associated with the clinical stage (Ⅳ vs. Ⅰ–Ⅲ, p = 0.005), T classification (T4 vs. T1–3, p < 0.001), tumor length (<6 cm vs. ≥6.0, p = 0.011), and NLR (high vs. low NLR, p < 0.001) (Supplementary Table S1). At the time of analysis, disease progression was observed in 48 (41.4%). The most common sites of initial progression, including overlapping distribution, were the primary tumor (n = 26), distant metastases (n = 16), lymph node (n = 15), and peritoneal dissemination (n = 1).
The median follow-up time for PFS was 22.8 months (range, 0.7 to 103.4). The 1- and 3-year PFS rates were 71.4% and 59.4%, respectively (Fig. 2A). In univariate analysis, clinical stage Ⅳ (HR, 2.36; 95% CI, 1.27 to 4.37; p = 0.006), T4 classification (HR, 2.37; 95% CI, 1.33 to 4.22; p = 0.003), and high NLR (HR, 2.02; 95% CI, 1.13 to 3.59; p = 0.017) were significantly associated with poorer PFS (Table 3, Fig. 2B, 2C). Additionally, in ESCC patients who did not receive induction chemotherapy (n = 93), high NLR (HR, 2.16; 95% CI, 1.19 to 3.91; p = 0.010) was also significantly associated with poorer PFS (Fig. 2D). In multivariate analysis, clinical stage Ⅳ (HR, 2.86; 95% CI, 1.50 to 5.44; p = 0.003) and high NLR (HR, 2.40; 95% CI, 1.32 to 4.39; p = 0.004) remained independent unfavourable prognostic factors for PFS (Table 3). Among patients with both clinical stage IV disease and a high NLR, high NLR had a significant impact on poorer PFS (HR, 10.07; 95% CI, 2.10 to 48.28; p = 0.004) (Supplementary Fig. S2A–S2C).
Grade ≥3 acute adverse events occurred in 13 patients (11.2%), with the most common being esophagitis (n = 12, 10.3%). Grade 3 pneumonitis and anorexia were observed in one patient (0.9%) and two patients (1.7%), respectively. No cases of acute dermatitis of grade ≥3 were reported. Grade 3 late adverse events occurred in 14 patients (12.0%), with perforation being the most frequently observed (n = 6, 5.2%). Other grade ≥3 late adverse events included pneumonitis (n = 3, 2.6%), cardiac complication (n = 3, 2.6%), esophageal stenosis (n = 2, 1.7%), gastric bleeding (n = 2, 1.7%), and pleural effusion (n = 2, 1.7%).
Discussion and Conclusion
This study examined the prognostic impact of NLR on survival outcomes in patients with ESCC who underwent PBT with concurrent chemotherapy. In addition to the clinical stage, NLR was identified as a significant prognostic factor for OS, PFS, and cCR rates. Patients with clinical stage IV and a high NLR demonstrated particularly poor OS and PFS.
A systematic review and meta-analysis previously established NLR as a prognostic marker in patients with EC including ESCC treated with surgery or chemoradiotherapy using photon irradiation [12]. However, most studies have focused on preoperative NLR, with some investigating NLR before chemoradiotherapy with photon irradiation. This study newly identified NLR before chemoradiotherapy with proton beam irradiation as a prognostic marker. Chronic inflammation within the tumor microenvironment is known to promote malignant progression [6]. Based on the results of our study and many previous studies, as a marker of systemic inflammation, an elevated NLR has been associated with poor prognosis across different treatment modalities in ESCC. That is, our findings substantiated that pretreatment NLR is a prognostic, rather than a predictive factor for the prognosis of ESCC. Moreover, the clinical stage—strongly correlated with prognosis in this study—is the most widely recognized classification system, as defined by the UICC [4]. Given the extremely poor prognosis of patients with clinical stage IV ESCC with a high NLR, definitive treatment, including PBT, may not be appropriate for this subgroup. Therefore, for patients with clinical stage Ⅳ ESCC with a high NLR, the development of new treatment methods is strongly desirable.
Nemoto et al. [22] reported outcomes of chemoradiotherapy with photon irradiation in Japanese patients. Although their study, based on the UICC 7th edition, had similar clinical stage characteristics to the present study, the reported cCR rate in patients with ESCC treated with 60 Gy of photon irradiation was approximately 45% [23]. In contrast, the cCR rate in patients treated with PBT in this study was approximately 60%, suggesting that PBT may offer a comparable or superior local therapeutic effect. Furthermore, the cCR rate was correlated with both the clinical stage and NLR. These findings suggest that pretreatment NLR, similar to TNM classification, may serve as a valuable predictive marker of treatment response in patients with ESCC undergoing PBT.
Patients with EC including ESCC receiving radiotherapy may develop acute and late adverse events, including pneumonitis and cardiac complications [18,24]. The risk of these events is closely linked to the radiation dose delivered to the lungs or heart [25-27]. Recent studies have suggested that PBT may reduce radiation exposure to these organs compared with intensity-modulated radiation therapy using photon irradiation [28]. In this study, grade ≥3 adverse events were adequately low, supporting the potential benefit of PBT in reducing severe toxicity. Furthermore, although cardiac complications could influence survival outcomes [19], the grade ≥3 cardiac complication rate was only 2.6% in our study. This low cardiac complication of PBT may have a positive implication for survival outcome of patients with ESCC. When a relatively long survival time is anticipated for patients with ESCC, PBT should be used more aggressively than photon radiotherapy.
This study has some limitations, due to its retrospective design. First, the distribution of clinical stages was unbalanced, with relatively few patients in stage II, necessitating the combination of stages I–III for statistical analysis. Second, survival outcomes may have been influenced by variations in local salvage treatments, chemotherapy regimens, and definitive PBT protocols; especially, the use of consolidation chemotherapy had potential risks of selection bias and immortal time bias. Third, induction chemotherapy may influence baseline NLR values. The laboratory data before induction chemotherapy could not be collected for almost all ESCC patients who underwent induction chemotherapy because most of them were underwent induction chemotherapy at other institutions. However, we believe that this NLR before PBT treatment is sufficiently useful as a factor that radiation oncologists can evaluate prognosis at the time of consulting for PBT. Fourth, this study included patients with ESCC who were followed for only 3 months. This short follow-up time in some patients with ESCC was because they were transferred to other hospitals when their acute side effects improved. However, to reduce the influence of this selection bias, we decided to include all 116 consecutive patients with ESCC in our analyses. Fifth, adverse event assessments were based on medical records, which may not have fully captured all toxicities, potentially underestimating the true incidence of treatment-related complications. Despite these limitations, this study is the first to highlight the prognostic significance of NLR in patients with ESCC treated with PBT.
In conclusion, NLR was a significant prognostic factor for OS, PFS, and cCR in patients with ESCC undergoing PBT and concurrent chemotherapy. Among these patients, those with clinical stage IV disease and a high NLR had particularly poor OS and PFS.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
