Predictive value of lymphocyte-to-monocyte ratio in the preoperative setting for progression of patients with breast cancer

Breast cancer is the most common type of cancer in women. Although neoadjuvant chemotherapy (NAC) increases the options for breast-conserving surgery and reduces the risk of postoperative recurrence in patients with resectable breast cancer [1‐4], recurrence and metastasis still remain major problems, especially in patients with advanced-stage disease [5]. The tumor node metastasis (TNM) staging system and its associated phenotypes are important and useful tools for predicting prognosis [6, 7]. In breast cancer, molecular subtypes also affect the prognosis. Patients with triple-negative breast cancer (TNBC) or human epidermal growth factor receptor-2 (HER2)-enriched breast cancer have high recurrence rates and poor prognosis [8]. However, it has been reported that not only the tumor characteristics but also the host inflammatory response are important for cancer progression [9].

Inflammation affects cancer progression, and a chronic systemic inflammatory response is involved in poor outcome in breast cancer patients [10]. Furthermore, the host immune system has been found to influence the clinical response to chemotherapy, and should thus be taken into account even during conventional chemotherapy treatment [11]. Systemic inflammatory markers such as neutrophil to lymphocyte ratio (NLR) and platelet lymphocyte ratio (PLR) have been reported as prognostic factors in various cancers [12, 13]. Furthermore, the lymphocyte-to-monocyte ratio (LMR), which also reflects the degree of systemic inflammation, has recently been reported to correlate with survival in various types of malignancies, such as diffuse large B cell lymphoma, colon cancer, esophageal carcinoma, lung cancer [14‐17]. In this single-center, retrospective study, we aimed to evaluate LMR and NLR as a possible marker for predicting the outcome of NAC in a consecutive series of patients with breast cancer treated with a standardized protocol.

Clinical responses (pCR + partial response) were observed in 209 patients (87.5%, 209/239). NAC-related pCR was observed in 91 patients (38.1%, 91/239). The pCR rates were 48.2% (40/83) and 32.7% (51/156) in patients with TNBC and non-TNBC, respectively (Table 1). Among all cases, DFS was significantly better in the pCR group compared with the non-pCR group (p = 0.040) (Additional file 1: Figure S1a), while OS tended to be better in the pCR group (p = 0.058) (Additional file 1: Figure S1b).

LMR was determined in every sample and ranged from 1.8–15.2 (mean, 6.2; median, 5.9; standard deviation, 2.3). The LMR cut-off value for DFS was 6.00 (area under the curve (AUC): 0.57; sensitivity: 61.3%; specificity: 59.2%) (Additional file 2: Figure S2a). We therefore classified patients into low-LMR (n = 120, 50.2%) and high-LMR (n = 119, 49.8%) groups according to a cut-off value of 6.0 (Fig. 1). LMR was significantly correlated with age (p = 0.004), menopausal status (p = 0.008), and tumor size (p = 0.036). There was no significant correlation between LMR and any other tested clinicopathological parameter, including pCR (Table 2). In continuous variables, older age (p < 0.001), menopause (p < 0.001), and higher Ki-67 (p = 0.016) were significantly associated with higher LMR (Additional file 3: Figure S3). NLR was also determined in every sample, and ranged from 0.5–10.6 (mean, 2.3; median, 2.0; standard deviation, 1.3). The NLR cut-off value for DFS was 1.63 (AUC: 0.56; sensitivity: 75.6%; specificity: 36.7%) (Additional file 2: Figure S2b). We therefore classified patients into low-NLR (n = 74) and high-NLR (n = 165) groups according to a cut-off value of 1.63 (Fig. 1). NLR was significantly correlated with age (p < 0.001) and menopausal status (p < 0.001), but there was no significant correlation between NLR and any other tested clinicopathological parameter, including pCR. In continuous variables, older age (p = 0.011) and menopause (p = 0.008) were significantly associated with lower NLR (Additional file 4: Figure S4).

DFS was significantly worse in the low- compared with the high-LMR group (p = 0.005) (Fig. 2a), while OS tended to be worse in the low-LMR group (p = 0.059) (Fig. 2b). Among 83 TNBC patients, DFS was significantly longer (p = 0.006) in the high- compared with the low-LMR group (Fig. 2c), but OS was not significantly different (p = 0.191) (Fig. 2d). Among 156 non-TNBC patients, there was no significant difference between the low- and high-LMR groups in terms of DFS (p = 0.170) (Fig. 2e) or OS (p = 0.176) (Fig. 2f). There were no significant differences between the low- and high-NLR groups in terms of DFS or OS for all breast cancers (p = 0.143 and p = 0.359, respectively) (Fig. 3a,b), TNBC (p = 0.150 and p = 0.416, respectively) (Fig. 3c,d), and non-TNBC patients (p = 0.376 and p = 0.191, respectively) (Fig. 3e,f).

The correlations between DFS and the various clinicopathological factors in 239 all breast cancers are shown in Table 3. According to the results of univariate analysis, DFS exhibited significant relationships with lymph node status (p = 0.020), complete response to chemotherapy (p = 0.034), and LMR (p = 0.005). In contrast, NLR was not a prognostic factor for DFS (p = 0.116). Multivariate analysis indicated that lymph node status (HR = 2.826, 95 %CI: 1.281–7.474, p = 0.008), and LMR (HR = 2.245, 95 %CI: 1.237–4.233, p = 0.008) were independent prognostic factors for survival (Table 3). Additionally, in 83 TNBC patients, multivariate analysis revealed that pathological response (HR = 2.921, 95 % CI: 1.015–10.470, p = 0.047) and LMR (HR = 4.675, 95 % CI: 1.500–20.445, p = 0.006) were significantly correlated with RFS (Additional file 5: Table S1).

Inflammation and cancer show a strong association, and pretreatment levels of peripheral inflammatory cells, including neutrophils, lymphocytes and monocytes are reported as prognositic factors in various cancers [27‐29]. Lymphocytes play an important role in host tumor immunity (for example, in cytotoxic cell death and inhibition of tumor cell proliferation and migration) [30‐33]. Decreased lymphocyte numbers are therefore considered to be responsible for an insufficient immunologic reaction to the tumor, thus promoting tumor progression and metastasis [15]. Monocytes are known to infiltrate tumors and differentiate into tumor-associated macrophages, which are involved in tumor proliferation, invasion, metastasis, neovascularization, and recurrence [34, 35]. Increased levels of monocytes thus reflect a high tumor burden in patients with cancer. In such a mechanism, LMR is believed to reflect the host immune status and the degree of tumor progression. Given that both a low lymphocyte count and high monocyte count reflect insufficient anti-tumor immunity and an elevated tumor burden, a low LMR is therefore associated with a poorer prognosis.

In this study, we evaluated the predictive value of pre-NAC LMR and NLR for progression and chemosensitivity in breast cancer, and compared the predictive values of these systemic inflammatory markers, and showed that LMR was significantly associated with DFS in all breast cancer patients who received NAC. This result suggests that LMR may be one of the criteria for deciding whether to perform adjuvant chemotherapies with active regimens from the beginning.

There are many publications concerning the prognostic value of LMR or NLR in breast cancer [36‐44]. The appropriate cut-off values for LMR and NLR were set by ROC curve in most studies, and were not unified. Recently, Ethier et al. conducted meta-analysis on the relationship between NLR and prognosis of breast cancer, and reported that the median cut-off value for NLR was 3.0 [45]. However, to our knowledge, no meta-analysis on the relationship between LMR and prognosis of breast cancer has been conducted. In addition, some studies evaluated the predictive value of LMR or NLR for progression and chemosensitivity in breast cancer patients treated with NAC, however, there are few reports to investigate the prognostic value of both LMR and NLR in breast cancer patients following NAC. Marín et al. analysed 150 breast cancer patients treated with NAC and reported that patients with high LMR and low NLR were associated with good DFS [44]. On the other hand, our study reported that only LMR was significantly correlated with prognosis. There are two possible reasons for this inconsistent result. First, the appropriate cut-off values for LMR and NLR differed in each study. Second, the standardized protocol of NAC differed in each study. In Marín’s study, NAC regimen was based on anthracyclines and taxanes ± trastuzumab, and pCR rate was 17.6%. In our study, NAC regimen was based on FEC100 followed by weekly paclitaxel ± trastuzumab, and pCR rate was higher than Marín’s study, 38.1%. 5-fluorouracil is shown to increase the sensitivity of cancer cells to killing by cytotoxic T cells (CTLs). In addition, a DNA alkylating agent cyclophosphamide can lead to improved T cell effector functions and stimulate an immunogenic death. Moreover, CTLs are effective against paclitaxel-treated tumor cells and induce tumor cell apoptosis [46, 47]. In other words, our regimen plays a role in enhancing the immune response.

Moreover, the present study is the first to evaluate the prognostic value of both LMR and NLR in different molecular subtypes of breast cancer following NAC. Weijuan et al. reported that decreased LMR was significantly associated with a poor prognosis for TNBC subtype in a study of 1570 patients [38]. The present study found no relationship between LMR and intrinsic subtype, but did reveal significant differences in DFS among all 239 patients, and among 83 TNBC patients stratified by LMR. TNBC is known to exhibit poor clinical outcomes compared with non-TNBC, and a recent study suggested that TNBC may be more strongly influenced by systemic inflammatory function [41].

Recently, it is becoming evident that the surrounding cancer microenvironment greatly influences cancer cells and plays a role in the development of characteristic cancer features [48]. Lymphocytes play not only as systemic inflammatory markers but also as tumor-infiltrating lymphocytes (TILs) that evaluate tumor immune responses [49‐51], and monocytes also relate to tumor-associated macrophages. Matsumoto et al. reported that high levels of tumor-infiltrating CD8+ T-cells may reflect an improved prognosis in terms of chemotherapy sensitivity in TNBC, and that tumor-associated macrophages were associated with a relatively poor outcome in patients with TNBC [52]. We also examined the prognostic significance of TILs before NAC in same breast cancer patients [53]. As in present study, DFS was significantly longer in the high TILs group than in the low TILs group among all breast cancer and TNBC patients. These results suggested that local immune tumor microenvironment and systemic inflammation had relationship. From now on, by further evaluating tumor-infiltrating lymphocytes or tumor-associated macrophages, along with systemic inflammatory markers such as LMR or NLR, more accurate identification of patient-specific immune mechanisms and prediction of prognosis may be possible.

As a potential limitation, this study is a single-center retrospective study, and the sample size is small, and the numbers of TNBC patients are thus even smaller. Further prospective multicenter studies are therefore needed to identify the strengths and weaknesses of our findings.

The findings of this study indicate that pretreatment LMR is a useful prognostic marker in patients with breast cancer.