Clinical Significance of Expression of Immunoadjuvant Molecules (LAG-3, TIM-3, OX-40) in Neoadjuvant Chemotherapy for Breast Cancer

Background/Aim: Various immunosuppressive factors that inhibit the immune response to cancer are present in cancer cells and the cancer microenvironment. Co-inhibitory and co-stimulatory receptors are dynamically expressed on T-cells as immunoadjuvant molecules that regulate the state of T-cell activity. In this report we focus on immunoadjuvant molecules such as LAG-3, TIM-3, and OX-40, for which there have been few published reports. We investigated the expression of LAG-3, TIM-3 and OX-40 in tumor-infiltrating lymphocytes (TILs), and clinically verified the significance of that expression in relation to neoadjuvant thermotherapy (NAC). Patients and Methods: A total of 177 patients with resectable early-stage breast cancer were treated with NAC. Estrogen receptor (ER), progesterone receptor (PgR), human epidermal growth factor receptor 2 (HER2), Ki67, LAG-3, TIM-3 and OX-40 status were assessed by immunohistochemistry. Results: The group with low-LAG-3 expression was significantly smaller than the group with high expression in triple-negative breast cancer (TNBC) (p=0.038) and HER2-enriched breast cancer (HER2BC) (p=0.021), while the total number of pathological complete response (pCR) patients was greater (p<0.001). In TNBC and HER2BC, the pCR rate was significantly higher in the low-LAG-3 expression group than in the high-LAG-3 expression group (p<0.001 and p=0.02, respectively). Moreover, on multivariate analysis low-LAG-3 expression status was an independent predictor of favorable prognosis (TNBC: p=0.014, HR=8.124; HER2BC: p=0.048, HR=10.400). Conclusion: Our findings suggest that LAG-3 may become a biomarker in highly malignant breast cancers such as TNBC and HER2BC that can predict the therapeutic efficacy of NAC.


Conclusion: Our findings suggest that LAG-3 may become a biomarker in highly malignant breast cancers such as TNBC and HER2BC that can predict the therapeutic efficacy of NAC.
Various immunosuppressive factors that inhibit the immune response to cancer are present in cancer cells and the cancer microenvironment (1,2). Co-inhibitory (suppression) and costimulatory receptors (activation) are dynamically expressed on T-cells as immunoadjuvant molecules that regulate the state of T-cell activity, and the inhibitory molecules act as immunological checkpoints. There are a number of immunological checkpoints in the progression of the immune response, and the co-inhibitory functions of receptors such as cytotoxic T-lymphocyte-associated antigen 4 (CTLA-4) and programmed cell death 1 (PD-1: CD279) are considered particularly important checkpoints for regulating the autoimmune response (3,4). Antitumor effects of immunological checkpoint inhibition therapy utilizing an anti-CTLA-4 antibody and an anti-PD-1 antibody as antitumor Tcell effectors have been discovered in many types of cancers, and these findings have dramatically changed the positioning of cancer immunotherapy in clinical settings (5)(6)(7)(8). In breast cancer (BC) as well, it has been reported that the anti-PD-1 antibody pembrolizumab has antitumor activity toward triplenegative breast cancer (TNBC), and that the levels of expression of PD-1 and its ligand, programmed cell deathligand-1 (PDL-1: CD274), correlate with prognosis (9)(10)(11)(12).These findings could affect individualized therapy for breast cancer. We have previously suggested that the expression of PD-1 and PD-L1 may serve as biomarkers for predicting therapeutic efficacy in neoadjuvant chemotherapy (NAC) for breast cancer (13). In this report we focus on immunoadjuvant molecules such as lymphocyte activation gene-3 (LAG-3: CD223), T-cell immunoglobulin and mucin containing protein-3 (TIM-3), and orexin-40 (OX-40: CD134), for which there have been few published reports. LAG-3 works as a molecule that suppresses the proliferation and activation of T-cells, and it is also expressed by tumorinfiltrating lymphocytes (TILs), T-cells that have been impoverished by chronic infection (exhausted T cells), and by regulatory T cells (Tregs) (14)(15)(16). TIM-3 not only functions as an immunological checkpoint regulator that suppresses Tcell function synergistically with PD-1, but it also suppresses the natural immune response of myeloid-lineage cells and is responsible for important functions in cancer cell support and survival (17)(18)(19). OX-40 is an immuno-co-stimulatory molecule that is expressed by activated T-cells, is expressed by Tregs, and supports memory T-cells (20)(21)(22).
The importance of regulating and improving the immune microenvironment in cancer has been recognized because the immune microenvironment in cancer tissues affects not only the efficacy of immunotherapy, but also the efficacy and prognosis of conventional chemotherapy and other modes of anticancer therapy (23,24). Therefore, monitoring the host's immune response to cancer in the microenvironment is believed to play a key role in predicting therapeutic efficacy and prognosis. TILs, which are considered indicators of immune response monitoring, have been reported as prognostic factors and predictors of therapeutic efficacy (25)(26)(27). In this study, we investigated the expression of LAG-3, TIM-3, and OX-40 in TILs, and clinically verified the significance of that expression in relation to NAC. A total of 177 patients with resectable early-stage  breast cancer diagnosed as stage IIA (T1, N1, M0 or T2, N0, M0),  IIB (T2, N1, M0 or T3, N0, M0) or IIIA (T1-2, N2, M0 or T3, N1-2,  M0) were treated with NAC between 2007 and 2013. Tumor stage and T and N factors were stratified based on the TNM Classification of Malignant Tumours, UICC Seventh Edition (28). Breast cancer was confirmed histologically by core needle biopsy and staged by systemic imaging studies using computed tomography (CT), ultrasonography (US) and bone scintigraphy. Breast cancer was classified into subtypes according to the immunohistochemical expression of estrogen receptor (ER), progesterone receptor (PgR), human epidermal growth factor receptor 2 (HER2) and Ki67. Based on their immunohistochemical expression, the tumors are categorized into the immuno-phenotypes of luminal A (ER+ and/or PgR+, HER2-, Ki67-low), luminal B (ER+ and/or PgR+, HER2+) (ER+ and/or PgR+, HER2-, Ki67-high), HER2-enriched breast cancer (HER2BC) (ER-, PgR-, and HER2+) and TNBC (negative for ER, PgR and HER2). In this study, luminal A and luminal B were assumed to be hormone receptor-positive breast cancers (HRBC).

Patient background.
All patients received a standardized protocol of NAC consisting of four courses of FEC100 (500 mg/m 2 fluorouracil, 100 mg/m 2 epirubicin and 500 mg/m2 cyclophosphamide) every 3 weeks, followed by 12 courses of paclitaxel (80 mg/m 2 ), administered weekly (29,30). Forty-five patients were diagnosed with HER2positive breast cancer and trastuzumab was administered on a weekly (2 mg/kg) or tri-weekly (6 mg/kg) basis, during paclitaxel treatment (31). All patients underwent chemotherapy as outpatients. Therapeutic anti-tumour effects were assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) criteria (32). Pathological complete response (pCR) was defined as the complete disappearance of the invasive compartment of the lesion with or without intraductal components, including the lymph nodes. Patients underwent mastectomy or breast-conserving surgery after NAC. All patients who underwent breast-conserving surgery were administered postoperative radiotherapy to the remnant breast. We operated for seven cases of progressive disease (PD) during the NAC enforcement when progression was confirmed. Overall survival (OS) was defined as the period from the initiation of NAC to the time of death from any cause. Disease-free interval (DFS) was defined as the interval in years from the date of the primary surgery to the first local recurrence, distant recurrences, or death from any cause. All patients were followed up by physical examination every 3 months, US every 6 months and CT and bone scintigraphy annually. The median follow-up period for the assessment of OS was 3.4 years (range=0.6-6.0 years) and 3.1 years for DFS (range=0.1-6.0 years). The design of this study is a retrospective chart review study. Written informed consent was obtained from all subjects. This research conformed to the provisions of the Declaration of Helsinki in 1995. All patients were informed of the investigational nature of this study and provided their written, informed consent. The study protocol was approved by the Ethics Committee of Osaka City University (#926).
Immunohistochemistry. All patients underwent a core needle biopsy prior to NAC, and they had undergone a curative operation involving a mastectomy or conservative surgery with axillary lymph node dissection after NAC at Osaka City University. Immunohistochemical studies were performed as previously described on core needle biopsy specimens (33). Tumor specimens were fixed in 10% formaldehyde solution, embedded in paraffin, and 4-μm-thick sections were mounted onto glass slides. Slides were deparaffinized in xylene and heated for 20 min (105˚C, 0.4 kg/m 2 ) in an autoclave in Target Retrieval Solution (Dako, Carpinteria, CA, USA). Specimens were then incubated with 3% hydrogen peroxide in methanol for 15 min to block endogenous peroxidase activity, followed by incubation in 10% normal goat or rabbit serum to block non-specific reactions.
Immunohistochemical scoring. Immunohistochemical scoring was performed by two pathologists specialized in mammary gland pathology, using the blind method to confirm the objectivity and reproducibility of diagnosis. The cut-off value for ER and PgR positivity was set at ≥1% in accordance with previous studies (34). HER2 expression was scored according to the accepted grading system (0, no reactivity or membranous reactivity in less than 10% of cells; 1+, faint/barely perceptible membranous reactivity in ≥10% of cells or reactivity in only part of the cell membrane; 2+, weak to moderate complete or basolateral membranous reactivity in ≥10% of tumor cells; or 3+, strong complete or basolateral membranous reactivity in ≥10% of tumor cells). HER2 expression was considered positive if the immunostaining score was 3+, or in cases where the score was 2+ and included gene amplification determined using fluorescent in situ hybridization (FISH). For FISH analyses, each copy of the HER2 gene and its centromere 17 (CEP17) reference were counted. The interpretation followed the criteria of the ASCO/CAP guidelines for HER2 IHC classification for breast cancer; it considered positive if the HER2/CEP17 ratio was higher than 2.0 (35,36). A Ki67-labeling index with ≥14% of tumor cells with nuclear staining was considered positive (37).
Histopathological analysis of the percentage of TILs was evaluated on a single full-face hematoxylin and eosin (HE)-stained tumor section using criteria described by Salgado et al. (38). TILs were defined as the infiltrating lymphocytes within tumor stroma and are expressed by the proportion of the field investigated (39,40). To evaluate LAG-3, TIM-3, and OX-40 expression, the number of infiltrating T-lymphocytes stained with anti-LAG-3, anti-TIM-3 and anti-OX-40 antibody in areas surrounding cancer cells was measured under 400-times magnified microscopy in each of 5 fields of view (FOVs) selected in darkly stained areas ( Figure 1A-C). Using previous reports as a reference, the median value of the average number in the 5 FOVs was determined, and that number was set as a cut-off value (9,16,38).
Statistical analysis. Statistical analysis was performed using the SPSS ® version 19.0 statistical software package (IBM, Armonk, NY, USA). Categorical data are reported with numbers and percentages, and continuous data as a median and range. The association between LAG-3, TIM-3, OX-40, and clinicopathological variables, and the significance of different prognostic markers were analyzed using the chi-square test (or Fisher's exact test when necessary). Association with survival was analyzed using the Kaplan-Meier plot and the log-rank test. The Cox proportional hazards model was used to compute univariable and multivariable hazards ratios (HR) for the study parameters with a 95% confidence interval (CI), and was used in a backward stepwise method for variable selection in multivariate analyses. In all of the tests, a pvalue of less than 0.05 was considered statistically significant. Cutoff values for different biomarkers included in this study were chosen prior to statistical analysis.  Expression of immunoadjuvant molecules in all breast cancer patients. Among the total number of breast cancer patients (n=177), 47 patients (26.6%; range=0 to 46%; mean=8%; median=7%; standard deviation=5%) tested positive for LAG-3 expression, 31 patients (17.5%; range=0 to 37%; mean=6%; median=6%; standard deviation=4%) tested positive for TIM-3 expression, and 32 patients (18.1%; range=0 to 38%; mean=7%; median=8%; standard deviation=4%) tested positive for OX-40 expression. When the clinicopathological features were examined, the group with low-LAG-3 expression was significantly smaller than the group with high expression in TNBC (p=0.038) and HER2BC (p=0.021), but significantly larger in HRBC (p<0.001), while the total number of pCR patients was greater (p<0.001). In the TIM-3 low-expression group, tumor diameter was significantly smaller (p=0.007) and the Ki-67 level was significantly higher (p=0.043) than in the high-expression group, and no correlation was found between intrinsic subtype and pCR rate. OX-40 expression also showed ANTICANCER RESEARCH 42: 125-136 (2022) 128 no correlation between intrinsic subtype and pCR rate (Table  I). In the prognostic analysis, we found a significant DFS extension in the LAG-3 low-expression group over the highexpression group (p=0.001, log-rank) (Figure 2A). We also found a significant OS extension in the LAG-3 low-expression group over the high expression group (p<0.001) ( Figure 3A). There was no significant difference in DFS and OS between TIM-3 and OX-40 expression level groups ( Figure 4A-D). In the univariate analysis, low-LAG-3 expression significantly contributed to extension of the disease-free survival interval (p=0.002, HR=3.239). Low-LAG-3 expression was also an independent factor for good prognosis (p=0.006, HR=3.047) in the multivariate analysis (Table II).

LAG-3 expression in triple-negative breast cancer.
Among the 61 TNBC patients, 22 (36.1%) were in the high-LAG-3 expression group, and 39 (63.9%) were in the low-LAG-3 expression group. The pCR rate was significantly higher (p<0.001) in the low-LAG-3 expression group than in the high-LAG-3 expression group (Table III). Outcome analysis showed significantly longer DFS (p<0.001, log-rank) ( Figure  2B) and OS (p=0.003, log-rank) ( Figure 3B) in the low-LAG-3 expression group than that in the high-LAG-3 expression group. Low-LAG-3 expression group in patients with TNBC demonstrated significantly lower rate for the recurrence by univariate (p=0.003, HR=10.107) and multivariate analysis (p=0.014, HR=8.124) (Table II).

expression. We found a significant disease-free survival (DFS) extension in the LAG-3 low-expression group over the high-expression group in 177 breast cancer patients (p=0.001, log-rank) (A). In TNBC and HER2BC, outcome analysis showed significantly longer DFS in the low-LAG-3 expression group than that in the high-LAG-3 group (p<0.001 and p=0.013 respectively, using the log-rank test) (B, C). DFS did not differ significantly between patients with low vs high LAG-3 expression (p=0.652, log-rank) (D).
tumor diameter was significantly smaller in the LAG-3 lowexpression group than in the high-expression group (p=0.028), and the number of pCR patients was greater (p=0.020) (Table III). Outcome analysis showed significantly longer DFS (p=0.013, log-rank) ( Figure 2C) and OS (p=0.003, log-rank) ( Figure 3C) in the low-LAG-3 expression group than that in the high-LAG-3 expression group. On univariate analysis for recurrence, low-LAG-3 expression status was a good prognostic factor (p=0.042, HR=10.969). Multivariate analysis also showed that low-LAG-3 expression status was an independent good prognostic factor (p=0.048, HR=10.400) (Table II).

LAG-3 expression in hormone receptor-positive breast cancer.
Among the 80 HRBC patients, 10 (12.5%) were in the high-LAG-3 expression group and 70 (87.5%) were in the low-LAG-3 expression group. No correlation was found between LAG-3 expression and any clinicopathological factor in HRBC patients (Table III) or between LAG-3 expression and outcomes (Table II, Figure 2D and Figure 3D).

Discussion
Recently, cancer cells has become clear that the peripheral environment (tumor microenvironment) greatly affects cancer cells and contributes to the formation of characteristics particular to cancer (41). The immune microenvironment in tumor tissues affects not only the efficacy of immunotherapy, but also the efficacy and prognosis of chemotherapy and other modes of anticancer therapy (23,24). Tumor antigens are present on tumor cells, and they induce an antitumor immune response in the host. Moreover, tumors activate mechanisms to suppress antitumor immunity, particularly in the microenvironment of tumor tissue, and they utilize those mechanisms for selfproliferation (1). Furthermore, there are diverse activation and suppression signals for antitumor immunity (3). In the group of immunoadjuvant molecules that regulate T-cell activation, inhibitory molecules such as CTLA-4, PD-1, PD-L1/2, LAG-3 and TIM-3 function as immunologic checkpoints (15,19). Furthermore, immunotherapy utilizing a blocking antibody to inhibit the signals of the immunological checkpoints has shown promising therapeutic efficacy in a clinical setting (5,7,8). Meanwhile, activators include OX-40, and it has been reported that administration of an anti-OX-40 agonistic antibody may enhance the antitumor immune response (42). In cancers such as malignant melanoma, renal cell carcinoma and breast cancer, a correlation has been suggested between PD-1 and PD-L1 expression levels and both the malignancy of cancer and extent of a poor prognosis (9,11,23). However, there have been few reports that have examined the level of expression of immunoadjuvant molecules such as LAG-3, TIM-3 and OX-40 clinically, and therefore monitoring of the tumor immune microenvironment has been conducted in NAC breast cancer patients using TILs. In this study, the LAG low-expression group had a significantly higher pCR rate than the high LAG expression group in NAC breast cancer patients, and low-LAG expression contributed to an extension of the disease-free survival interval. LAG-3 is a molecule with a structure similar to CD4, and it appears on the cell surface when T-cells are activated (14,16,43). The signal transduction pathway of LAG-3 is still unclear, but it is believed to function as a molecule that not only suppresses the proliferation and activation of T-cells, but that also plays a key role as an immune checkpoint similar to PD-1 and CTLA-4 (15). Basic research has demonstrated that the antitumor immune response in mice is enhanced by inhibiting the LAG-3 signal using an anti-LAG-3 antibody while concurrently administering an anti-PD-1 antibody (44). Furthermore, an anti-LAG-3 antibody was adapted for human use, and it was discovered that the use of this antibody in combination with paclitaxel in phase I and II clinical studies of breast cancer raised the response rate from 25% to 50% compared with groups treated with anticancer monotherapy (45). Paclitaxel is considered to improve the immune escape in the host by suppressing Tregs, but expression of LAG-3 has also been found in Tregs. In other words, an anti-LAG-3 antibody in combination with paclitaxel may effectively act to relieve immunosuppression by suppressing Tregs. In the NAC regimen used in our study that treats paclitaxel as a key drug, it appears that these mechanisms enable immune response monitoring via LAG-3 expression.

ANTICANCER RESEARCH
On the other hand, TIM-3 has galectin-9 as a ligand and suppresses the activation of effector T-cells mediated by ligand-receptor interaction (17)(18)(19). Clinical studies have reported that the level of expression of TIM-3 in renal cell carcinoma and head and neck cancer exhibits a negative correlation to prognosis (46,47). In our study, we found no correlation between TIM-3 expression and the therapeutic efficacy of NAC. TIM-3 is a molecule that contributes to the suppression of T-cell function synergistically with PD-1, and simply activating antitumor immunity by lowering TIM-3 expression alone may not improve the tumor immune microenvironment. Moreover, OX-40 is a member of the TNF receptor superfamily, and is expressed by activated Tcells, NK cells and Tregs (20)(21)(22). It has been reported that administration of an anti-OX-40 agonistic antibody enabled rejection of fully established tumors in a mouse model (42). In our study, however, we found no correlation between OX-40 expression and therapeutic efficacy of NAC. Under the assumption that no potent antitumor effect will be obtained with anti-OX-40 agonistic antibody monotherapy, a search is now underway for a combination therapy with another drug (48). In other words, in the case of OX-40 as well, this finding indicates that a clear improvement of the tumor immune microenvironment cannot be obtained clinically through modulation of OX-40 expression alone. In addition, no correlations between expression of LAG-3, TIM-3 and OX-40 were found in this study.
In previous reports, breast cancer was not considered as a cancer that develops as the result of an immune disorder (49). Recently, however, breast cancer has come to be viewed as an immunogenic tumor, and the highly malignant subtypes TNBC and HER2BC have a high level of immune activity (50,51). We believe that in these highly malignant breast cancers the tumor immune microenvironment can be ANTICANCER RESEARCH 42: 125-136 (2022) 132 reliably monitored by TILs because of the high immune activity. In our study, when breast cancer was stratified by intrinsic subtype and TILs were evaluated, the groups with low-LAG expression in highly malignant breast cancers had a significantly higher pCR rate, but no significant difference was found for HRBC.

Conclusion
In TNBC and HER2BC, low LAG3 expression was shown to contribute to pCR and to have a favorable prognosis. Our findings suggest that LAG-3 may become a biomarker in highly malignant breast cancers such as TNBC and HER2BC that can predict the therapeutic efficacy of NAC.

Conflicts of Interest
None of the Authors has any conflicts of interest to disclose regarding this study.