Pre‐treatment systemic immune‐inflammation index is a useful prognostic indicator in patients with breast cancer undergoing neoadjuvant chemotherapy

Abstract The systemic immune‐inflammation index (SII = N × P/L) based on neutrophil (N), platelet (P) and lymphocyte (L) counts is used to predict the survival of patients with malignant tumours and can fully reflect the balance between host inflammatory and immune status. This study is conducted to explore the potential prognostic significance of SII in patients with breast cancer undergoing neoadjuvant chemotherapy (NACT). A total of 262 patients with breast cancer received NACT were enrolled in this study. According to the receiver operating characteristic curve, the optimal cut‐off value of SII was divided into two groups: low SII group (<602 × 109/L) and high SII group (≥602 × 109/L). The associations between breast cancer and clinicopathological variables by SII were determined by chi‐squared test or Fisher's exact test. The Kaplan‐Meier plots and log‐rank test were used to determine clinical outcomes of disease‐free survival (DFS) and overall survival (OS). The prognostic value of SII was analysed by univariate and multivariate Cox proportional hazards regression models. The toxicity of NACT was accessed by National Cancer Institute Common Toxicity Criteria (NCICTC). According to univariate and multivariate Cox regression survival analyses, the results showed that the value of SII had prognostic significance for DFS and OS. The patients with low SII value had longer DFS and OS than those with high SII value (31.11 vs 40.76 months, HR: 1.075, 95% CI: 0.718‐1.610, P = .006; 44.47 vs 53.68 months, HR: 1.051, 95% CI: 0.707‐1.564, P = .005, respectively). The incidence of DFS and OS in breast cancer patients with low SII value was higher than that in those patients with high SII value in 3‐, 5‐ and 10‐year rates. The common toxicities after NACT were haematological and gastrointestinal reaction, and there were no differences by SII for the assessment of side effects of neoadjuvant chemotherapy. Meanwhile, the results also proved that breast cancer patients with low SII value and high Miller and Payne grade (MPG) survived longer than those breast cancer with high SII value and low MPG grade. In patients without lymph vessel invasion, these breast cancer patients with low SII value had better prognosis and lower recurrence rates than those with high SII value. Pre‐treatment SII with the advantage of reproducible, convenient and non‐invasive was a useful prognostic indicator for breast cancer patients undergoing neoadjuvant chemotherapy and is a promising biomarker for breast cancer on treatment strategy decisions.


| INTRODUC TI ON
Breast cancer is one of the most frequent female malignancies and is the second cause of morbidity and mortality in women all over the world. 1 In recent years, the incidence of breast cancer is increasing year after year, and the number of breast cancer survivors sustains growth. 2 Moreover, 1.67 million new breast cancer cases and 522 000 deaths of breast cancer are reported worldwide per year, and the age of onset tends to be younger. 3 In China, breast carcinoma is the leading major cause of cancer-related mortality for females and has the highest incidence rate. 4 The incidence of breast carcinoma is increasing rapidly in coastal developed cities, and the post-menopausal breast cancer patients in China will reach 100/100 000 in the future. 5 Because of the progress of early diagnosis and the improvement of treatment strategy, lots of patients have been successfully treated, and the average 5-year survival rate is around 90%. 6 However, approximately 20%-25% patients are diagnosed with locally advanced breast cancer, and these patients account for early relapse and deaths. 7,8 A large number of studies have proved that surgery combined with adjuvant chemotherapy and radiotherapy can effectively improve the survival rate of patients according to the progress of early detection and treatment. [9][10][11] Neoadjuvant chemotherapy (NACT) is widely used to the treatment for breast cancer and has been testified to be beneficial for breast cancer patients. The application of neoadjuvant chemotherapy in breast cancer has attracted more and more attention as a result of increasing breast-conserving surgery rate in patients who might have required a mastectomy at initial diagnosis and decreasing tumour stage to achieve surgical operation chance and not increasing post-operative recurrence risk. 12,13 The main purpose of NACT is to establish a therapeutic strategy based on proven therapeutic efficacy and reduce the size of tumours and improve breast-conserving rates. 14,15 Although a number of NACT regimens have been used to treat the breast cancer, there is no internationally generally accepted NACT regimen for advanced breast carcinoma. 16,17 Therefore, it is necessary to seek some sensitive and effective indicators of breast carcinoma in order to improve the long-term survival and provide better treatment measures.
There are some factors influencing the prognosis of the breast carcinoma, for instance, pathological stage, histologic type, molecular subtypes, tumour size, lymph vessel invasion, menopause age and so forth. 18  In recent years, inflammation has been suggested as a critical and potentially intervenable mechanism in the pathogenesis of patients with carcinoma. Tumour associated inflammation is a fundamental part of the tumour microenvironment and could be responsible for treatment response. The changes in inflammatory cells may affect tumour carcinogenesis, development and metastasis, for example neoplastic cell proliferation, invasion migration and metastasis. 20 The detection of peripheral blood can reflect the inflammatory state of tumours for the diagnosis and treatment of tumours. A large number of studies have clarified that the elevated inflammatory biomarkers, for instance, white blood cell (W), neutrophil (N), lymphocyte (L), monocyte (M), platelet counts (P), C-reactive protein (CRP), as well as neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), lymphocyte-to-monocyte ratio (LMR) and systemic inflammation response index (SIRI), are detected in the systemic circulation and widely considered as prognostic factors for many malignant tumours. [21][22][23][24] An integrated and novel indicator that named systemic immune-inflammation index (SII = N × P/L), which is based on neutrophil (N), platelet (P) and lymphocyte (L) counts, is demonstrated to be related to clinical outcomes and predict the survival time of patients with cancer. 25,26 This integrated indicator may fully reflect the balance between host inflammatory and immune status compared with NLR, LMR/MLR and PLR and other conventional haematological parameters. However, the SII has been studied rarely in breast cancer patients received NACT. As far as we are concerned, the relationship between SII and breast cancer has not breast cancer patients with low SII value had better prognosis and lower recurrence rates than those with high SII value. Pre-treatment SII with the advantage of reproducible, convenient and non-invasive was a useful prognostic indicator for breast cancer patients undergoing neoadjuvant chemotherapy and is a promising biomarker for breast cancer on treatment strategy decisions.

K E Y W O R D S
breast cancer, neoadjuvant chemotherapy, prognosis, survival, systemic immune-inflammation index (SII) been building-up. Hence, the purpose of this study is to investigate the prognostic significance of SII in patients with breast cancer undergoing NACT.

| Study population
The retrospective analysis was enrolled 262 patients with advanced breast carcinoma undergoing NACT at Cancer Hospital Chinese Academy of Medical Sciences from January 1999 to December 2014. All patients were confirmed in accordance with the histology by core needle biopsy, and all cases were received NACT treatment.
The details of treatment for all patients were extracted from the patients' medical record. This study was approved by the ethics com- The exclusion criteria for the patients were as follows: (a) patients had accepted chemotherapy, radiotherapy, endocrine therapy and targeted therapy before beginning therapy; (b) patients with serious complications, such as infection, pneumonia; (c) patients with chronic inflammatory diseases or autoimmune disease, such as liver cirrhosis, systemic lupus erythematosus (SLE); (d) patients with distant metastasis; (e) patients were treated with blood product transfusion within one month before treatment.

| Chemotherapy regimens
We used anthracyclines-based and/ or taxanes-based neoadjuvant chemotherapy regimens, and one cycle of these regimens was re- On the first day, Anthracyclines (Zhejiang Hisun Pharmaceutical Co., LTD) at 90 mg/m 2 was given by intravenous injection in 100 mL of 0.9% Normalsaline, and Cyclophosphamide (Baxter Oncology GmbH, Halle, Germany) at 600 mg/m 2 was given by intravenous injection in 100 mL of 0.9% Normalsaline. One cycle of this regimen was repeated every 3 weeks.
On the first day, Anthracyclines at 90 mg/m 2 was given by intravenous injection in 100 mL of 0.9% Normalsaline, and Cyclophosphamide at 600 mg/m 2 was given by intravenous injection in 100 mL of 0.9% Normalsaline. From the first day to the second day, 5-Fluorouracil (Tianjin Jinyao Pharmaceutical Co., LTD) 500 mg/m 2 was given by intravenous injection in 500 mL of 5% glucose above 46 h. One cycle of this regimen was repeated every 3 weeks.
On the first day, Cyclophosphamide at 600 mg/m 2 was given by intravenous injection in 100 mL of 0.9% Normalsaline, and Taxol (Jiangsu Hengrui Medicine Co., Ltd) at 175 mg/m 2 was given by intravenous injection in 500 mL of 0.9% Normalsaline above 3 hours. One cycle of this regimen was repeated every 3 weeks.
On the first day, Anthracyclines at 90 mg/m 2 was given by intravenous injection in 100 mL of 0.9% Normalsaline, and Cyclophosphamide at 600 mg/m 2 was given by intravenous injection in 100 mL of 0.9% Normalsaline, Taxol at 175 mg/m 2 was given by intravenous injection in 500 mL of 0.9% Normalsaline above 3 hours.
One cycle of this regimen was repeated every 3 weeks.
On the first day, Anthracyclines at 90 mg/m 2 was given by intravenous injection in 100 mL of 0.9% Normalsaline, and Taxol at 175 mg/m 2 was given by intravenous injection in 500 mL of 0.9% Normalsaline above 3 hours. One cycle of this regimen was repeated every 3 weeks.
On the first day, Taxol at 175 mg/m 2 was given by intravenous injection in 500 mL of 0.9% Normalsaline above 3 hours, and Platinum compounds (Bristol-Myers Squibb biopharmaceutical company, Srl.) at AUC 4-6 was given by intravenous injection in 500 mL of 0.9% Normalsaline or 500 mL of 5% glucose. One cycle of this regimen was repeated every 3 weeks.

| Classification criteria and response evaluation
The tumour size was determined by the maximum diameter of the sample. TNM stage system was performed by the eighth edition of the American Joint Committee on Cancer (AJCC) and the Union for International Cancer Control (UICC) TNM stage classification. 27,28 Pathology type was classified as ductal carcinoma, lobular carcinoma and others. Molecular subtypes of breast cancer were divided into Luminal A, Luminal B HER2-positive, Luminal B HER2-negative, HER2-enriched and Triple negative. 29 The histologic response was estimated on the basis of Miller and Payne grade (MPG), and the MPG was categorized into five grades in line with the number of tumour cells in excision/ mastectomy specimens compared with the pretreatment core biopsy as follows 30 :(a) grade 1, no reduction in total number of cancer cells; (b) grade 2, minor (<30%) loss of total number of cancer cells; (c) grade 3, estimated 30%-90% reduction in total number of cancer cells; (d) grade 4, over (>90%) loss of total number of cancer cells; 5) grade 5, no infiltrating cancer (IC), or ductal carcinoma in situ. Histologic tumour grade was accessed by the Elston-Ellis modification of Scarff-Bloom-Richardson grading system and based on three factors, 31 such as: (a) the amount of gland formation; (b) the nuclear features; and (c) the mitotic activity. Each feature score was 1-3 points and then added the score to get the final total score of 3-9 points and was categorized into three grades by the following way 32 : (a) grade I, the total score of tumours was 3-5 points; (b) grade II, the total score of tumours was 6-7 points; (c) grade III, the total score of tumours was 8-9 points. The response rate was defined in the light of the Response Evaluation Criteria In Solid Tumours (RECIST) guidelines and was stratified into four groups: complete response (CR), partial response (PR), stable disease (SD) and progression of disease (PD). 33 The sum of CR and PR were the clinical objective response rate, the sum of SD and PD were the non-clinical response rate, and the sum of CR, PR and SD were the clinical benefit rate. Lymph vessel invasion and neural invasion were diagnosed by haematoxylin and eosin (HE) stain. The toxicity of NACT was accessed on the basis of the National Cancer Institute Common Toxicity Criteria (NCICTC). 34

| Peripheral venous blood parameters
Peripheral venous blood samples were taken within 1 week after breast cancer diagnosis and before neoadjuvant chemotherapy. The sterile EDTA tube was used to collect the blood samples. The XE-2100 haematology analyser (Sysmex, Kobe, Japan) was used to analyse the haematological parameters.

| Follow-up
All enrolled patients had post-operative follow-up in inpatients or outpatients every 3 months for the 1st to 2nd year, every 6 months for the 3rd to 5th year after surgery, then annually every year and until death. 35 Follow-up assessments included laboratory tests (routine blood test,  and GraphPad prism software (version 5.0). The optimal cut-off value was accessed by the receiver operating characteristic curve (ROC) analyses. The clinicopathologic categorical variables were performed as frequencies and percentages (%), and the associations between breast carcinoma and clinicopathological variables were evaluated by chi-squared test or Fisher's exact test. The Kaplan-Meier product limit estimator method was to determine the survival time of DFS and OS, and the association between breast carcinoma and survival was analysed by the Kaplan-Meier plots and log-rank test. The Cox proportional hazards regression model was used to examine the independent prognostic factors. P < .05 was considered to be statistical significance.

| Demographic and clinicopathologic characteristics of all breast cancer patients
Two hundred and sixty two patients with breast carcinoma were included in this study. The clinicopathological characteristics of patients with breast cancer were shown in Table 1. The ROC analysis was intended for evaluating the optimal cut-off value of the SII (602 × 10 9 /L).

| Correlation between SII and side effects of chemotherapy
In our study, the side effects of neoadjuvant chemotherapy for two cycles were evaluated and analysed by NCICTC. We also found that were the haematological and gastrointestinal reaction were the common toxicities after NACT.

| Univariate and multivariate Cox regression survival analyses
The

| Survival rate and prognostic significance of SII
The 20.8% (22/106), 6.6% (7/106), respectively. The results also indicated that the incidence of DFS and OS in patients with low SII was higher than that in patients with high SII in 3-, 5-and 10-year rates.
Although there was a trend towards significance for the association between low SII and high SII at 3-, 5-and 10-year rates, there were no significant differences between the two groups. (Table 7, Figure 2A-D).

| Association of molecular subtypes by postoperative pathology IHC and SII in patients with breast carcinoma
The results showed that molecular subtypes by post-operative pathology IHC were the important prognostic factor (Table 6). In order to go deep into evaluating the prognostic value of SII, the SII was analysed by the molecular subtypes. And the SII with different mo-

| Correlation between Miller and Payne grade (MPG) and SII in breast cancer patients
According to univariate and multivariate analyses, the MPG was the significant prognostic factor (Table 3). We also analysed the SII by MPG. We defined the patients with MPG grade 1 and 2 as

MPG-A group, the patients with MPG grade 3 as MPG-B group
and patients with MPG grade 4 and 5 as MPG-C group, respectively. By using the log-rank test, the mean DFS and OS time for patients with low SII were longer than in those with high SII in MPG-A group (χ 2 = 2.657, P = .103 and χ 2 = 2.953, P = .086, respectively; Figure 4A  .005 1 (reference) .005 1 (reference) .001 1 (reference) .002 Luminal

| Association of ER status and SII in patients with breast carcinoma
The results indicated that ER status in both core needle biopsy and post-operative pathology IHC was the significant prognostic factor.
Hence, for the sake of further to study the prognostic efficiency of SII, the SII was analysed by ER status. In core needle biopsy, the mean DFS and OS time in patients with low SII by the log-rank test were longer than in those patients with high SII in ER negative (χ 2 = 5.401, P = .020 and χ 2 = 5.476, P = .019, respectively; Figure

| Association of lymph vessel invasion and SII in breast cancer patients
The results also proved that lymph vessel invasion was the significant prognostic factor by univariate and multivariate analyses ( Table 6). The results showed that the mean DFS and OS time in patients with low SII by the log-rank test were longer than in

| D ISCUSS I ON
Breast carcinoma is the most common diagnostic cancer and with more than 2.1 million newly diagnosed female breast cancer cases in 2018, accounting for almost one in four cancer cases among women. 36 The surgery, radiotherapy and endocrine therapy are used to the treatment of early breast cancer; however, the neoadjuvant treatment which can turn inoperable tumours into operable tumours or reduce tumour stage for more frequent conservative breast surgery. 37 [54][55][56][57] Numerous studies showed that inflammatory markers and immune-based prognostic indexes, such as NLR, d-NLR, MLR/LMR and PLR, were considered to have prognostic value for breast cancer. [58][59][60][61][62][63][64][65][66] TA B L E 7 3-, 5-and 10-year DFS and OS rates of patients with breast cancer Parameters Cases (n) However, SII based on three inflammatory cells, including neutrophil, platelet and lymphocyte and can fully reflect the balance between host immune and inflammatory status. SII has been investigated in some malignancies, such as gastric cancer, prostate cancer, lung cancer and pancreatic cancer, and SII has been proved to be an independent predictor of malignant tumours. [67][68][69][70] In our study, we confirmed that the SII was the significant prognostic factor by univariate and multivariate analyses and could predict survival in breast cancer patients receiving NACT.

3-year (%) 5-year (%) 10-year (%) 3-year (%) 5-year (%) 10-year (%)
In our study, the clinicopathologic and demographic characteristics of the 262 patients with breast cancer were enrolled and analysed. The optimal cut-off value of the SII was 602 × 10 9 /L by ROC analysis, and this value was used to the data analysis. The results indicated that the low SII was significantly correlated with menopause and US-LNM. The blood parameters were determined by its median value, and the low SII was significantly correlated with ALB, are needed to support our findings, and the SII may be used in combination with other biomarkers to help predict clinical outcome in patients with breast cancer and is adopted in routine practice.

| CON CLUS IONS
SII is the significant prognostic factor for patients with breast cancer and can effectively predict the survival in patients with breast cancer receiving NACT. It is very important to consider the high incidence of breast cancer and the unbalanced distribution of medical condition in China and that reproducible, conveniently and non-invasive biomarkers should be applied to the prevention and treatment of breast cancer. It is very critical and meaningful to understand of haematological parameters and look for new targets for subjective treatment.

ACK N OWLED G EM ENTS
This work was supported by CAPES, FAPESC, CNPQ and UNESC.

CO N FLI C T O F I NTE R E S T
The authors declare no competing financial interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The material supporting the conclusion of this article has been included within the article.