In this retrospective study, we highlighted that pretreatment high NLR (≥ 2.53) and lymphopenia (ALC< 1.5G/L) were prognostic markers associated with poorer PFS in women treated with first-line CdK4/6i and ET for metastatic or locally advanced HR+/HER2- breast cancer.
No consensus has been reached to define a cut-off or threshold value for each ratio (NLR, PLR, LMR). We first determined these cut-offs with ROC curves. In our study, NLR cut-off was similar to those found in the literature mostly ranging between 2 and 5. A meta-analysis conducted in BC reported a median NLR cut-off value of 2.5 in 10 out of 15 studies .
Our results are consistent with previous studies. Four recent meta-analyses corroborate our findings showing that NLR is an independent prognostic factor for PFS and OS in patients with BC at different stages [13, 22, 23], especially for luminal A subtype . Wariss et al.  reported an association between high NLR and worse OS in 2,374 eBC and mBC patients, but only for patients with luminal subtypes. In another study concerning mTNBC, NLR> 2.5 at diagnosis was a useful predictor of poor OS, regardless of the subsequent treatment . In HR+/HER2- eBC, high NLR (>2.25) after neoadjuvant chemotherapy was correlated with poorer disease free survival (DFS) and OS, especially in patients with non-pathologic complete response (pCR) . Furthermore, NLR has been evaluated as a prognostic factor for other endpoints besides survival in different tumor types. For example, it was shown in esophageal cancer that patients with high NLR (≥ 2.2) had lower pCR rates (and a delayed response to chemotherapy) compared to those with a NLR< 2.2 (21% vs. 56%, p=0.001) . To date, our study is the first to report the prognostic impact of NLR on survival and response rates to first line treatment for HR+/HER2- mBC.
Low NLR appears to be a protective factor for progression-free survival rate with a 40% risk reduction of progression or death (HR=0.59 [CI95%: 0.32-1.08]), independently of ECOG-PS status in multivariable analysis. However, this result was no significant probably due to the lack of power resulting from a limited number of events as well as PFS results were not mature enough to support these. Similarly, we could not validate ECOG-PS as a significant independent predictive factor of disease progression, although this has been acknowledged in the literature.
Lymphopenia and NLR are two complementary prognostic factors. Increased systemic inflammation markers have been reported in lymphopenic patients, with an inverse increase in the percentage of peripheral neutrophils, due to the expression of pro-inflammatory cytokines such as IL-6 and IL-7, CD4+CD8+ double-positive (DP) thymocytes and age-related thymic function .
Lymphopenia is multifactorial and can be associated with patient characteristics (as age, ECOG-PS)  or tumor burden. In our study, the median age was similar between the two groups of pretreatments NLR, but they differed on ECOG-PS: patients in the high NLR group had a worse ECOG-PS, that was associated with a lower PFS. Otherwise, cell death secondary to inflammation and reduced thymic function have been suggested as possible mechanisms of peripheral lymphopenia observed in metastatic patients .
Lymphocytes, whether in peripheral blood or as tumor-infiltrating lymphocytes, play a major role in controlling disease progression. For example, in patients receiving everolimus for mBC, Schettini et al. reported that responders had a higher number of circulating CD8+ and CD4+ T cells, and increased pretreatment infiltration of these cells in the tumor microenvironment than non-responders . Concerning patients treated with palbociclib for HR+/HER2- mBC, we previously showed that those with pretherapeutic ALC< 1.5 G/L had significantly shorter PFS time (6 vs. 10 months, p=0.004), shorter OS time (20 vs. 33 months, p=0.018) and more disease progression at first imaging evaluation .
It is necessary to thoroughly understand the impact of the immune system on tumor control. On the one hand, neutrophils, B lymphocytes and some CD4+ T cells may stimulate cancer growth. On the other hand, cytotoxic CD8+ T cells are crucial components of tumor-specific cellular adaptive immunity as Thelper 1 (TH1), TH17, CD4+ T cells and Natural Killer cells are in the tumor microenvironment are. They inhibit tumor growth by producing interferon gamma, subsequently leading to angiostasis, cell cycle inhibition, apoptosis and tumor phagocytosis by macrophages . A retrospective study of 1,902 patients with eBC showed that a high total and peripheral CD8+ T cell count was associated with significantly longer breast cancer-specific survival (BCSS) . More specifically, in patients with ER-positive tumor, the total number of infiltrating CD8+ T cells was not significantly associated with patient outcome, whereas peripheral CD8+ count was associated with longer BCSS . Furthermore, Coffelt et al. demonstrated that elevated neutrophil counts induced by BC tumor cells suppressed CD8+ T cells and promoted metastasis through immunosuppression .
Furthermore, CdK4/6i have been reported to increase tumor immunogenicity by overcoming two principal mechanisms of tumor immune evasion. They limit the proliferation of Regulatory T cells (Tregs) leading to reduced immunosuppression and enhance antitumor immunity by increasing T cell activation, promoting T cell tumor infiltration, and expanding the functional capacity of tumor cells to present antigens [34, 35].
Other biomarkers evaluated in our study, such as LMR and PLR, had no impact on PFS or OS. Nevertheless, after 12 months of treatment, patients with high PLR (≥ 174) had a non-significant shorter median PFS than patients with low PLR. PLR has been described as a reliable prognostic marker in many cancers including BC . One study identified low pretreatment PLR value to be an independent predictive factor of pCR and DFS in patients with luminal B-like (HER2-) BC, suggesting that patients with a low platelet count and high lymphocyte count exhibit increased antitumor activity . In luminal BC, PLR> 185.5 had a strong prognostic value for DFS in patients with lymph node metastases . Platelets promote angiogenesis and invasiveness of tumor cells by enhanced metalloproteinase-9 (MMP-9) secretion, degradation of the extracellular matrix, release of adhesion molecules and growth factors such as VEGF-A or PDGF [39, 40]. Also, IL-6 implicated in the systemic inflammatory response and recruitment of neutrophils, can stimulate differentiation and multiplication of megakaryocytes . Concerning LMR, it was associated with poor OS in a meta-analysis of patients with solid tumors . In the BC neoadjuvant chemotherapy setting, a recent study confirmed this result in a multivariable analysis and showed that patients with low LMR had shorter DFS .
None of the parameters we studied had an impact on OS possibly due to limited follow-up with an OS not mature. Of note, the median PFS of 27 months in this study was similar to that obtained in the registration trials of CdK4/6 inhibitors .
This study had some limitations. Due to the retrospective nature of the study, we were unable to collect the values of certain inflammation parameters (albumin, C-reactive protein, vitamin D, and LDH) or data concerning concomitant treatments (e.g., corticosteroid therapy) and radiotherapy, which may influence neutrophil and lymphocyte counts [44, 45]. The sample size is limited and results must be interpreted with caution. Moreover, PLR and LMR were not significant on the primary outcome possibly due to the small sample size and a lack of power but also because the cut-off determined was not sufficiently discriminating. Is that it prompted us to design a prospective study, in order to improve and to confirm our results more powerfully.
Immune status is emerging as an essential biomarker of the tumor biology and microenvironment with an impact on patient outcome. Other biomarkers, such as tumor infiltrating lymphocytes (TILS) and circulating tumor cells (CTC), are still being evaluated in clinical research as prognostic factors but are not easily obtained in routine clinical practice .
Our study shows that NLR and ALC, indicators of the immune status, are interesting biomarkers in routine clinical practice because they are both available, easy-to-use, reliable and inexpensive prognostic factors of PFS in mBC patients treated with CdK4/6i in the first-line setting. These results could eventually help select patients more likely to benefit from CdK4/6i. These findings need to be confirmed in a larger prospective study.