Effect of preoperative peripheral blood platelet volume index on prognosis in patients with invasive breast cancer

Breast cancer (BC) is a frequent malignant tumor in females occurring in the mammary epithelial tissue. It has the characteristics of incidence and rapid disease progression, which gravely threatens the lives and healthcare of women worldwide. According to the latest estimates from the WHO's International Agency for Research on Cancer [1], worldwide, there were 2.08 million new BC cases and 630,000 women died from BC in 2018. At present, early BC is mainly treated by surgery. Although such patients can obtain a good prognosis, the risk of metastasis and recurrence remains difficult to predict [2].

Traditional prognostic factors of BC, such as molecular typing and clinical staging, have poor accuracy, while new prognostic tools and genes for BC, such as Oncotype Dx and BC susceptibility gene BRCA, respectively, are expensive, which limits their clinical application. With increasing attention being paid to the interaction between the tumor inflammatory environment and tumor cells, the relationship between activated platelets and tumors has become a hot research field. The mutual interaction between activated platelets and cancer cells results in tumor growth, abnormal angiogenesis and metastasis [3,4]. Activated platelets can escape shear-induced injury by binding to tumor cells, thereby promoting tumor cell colonization [5]. In recent years, it has been reported that platelet count [6] and platelet-to-lymphocyte ratio (PLR) [7] have predictive value for the long-term prognosis of BC. In addition, larger platelets have been reported to store more particles and receptors and to have greater adhesivity compared with smaller platelets. Therefore, the activity of activated platelets is more accurately expressed by their volume than by count [8]. Therefore, platelet volume indices (PVI) are the best way to assess platelet activation. Platelet distribution width (PDW) and mean platelet volume (MPV) are markers of platelet activation and morphology [9] associated with various inflammatory states [10]. Recent studies have shown that in BC, esophageal cancer, hepatocellular carcinoma and non-small-cell lung cancer (NSCLC), the level of MPV-to-platelet count ratio (MPV/P) was related to poor prognosis [11–14]. PDW and PDW-to-platelet count ratio (PDW/P) are related to poor prognostic value in endometrial cancer, melanoma, gastric cancer as well as NSCLC [15–18], Meanwhile, Okuturlar et al. [19] indicated that there were no statistical differences in the univariate analysis of patients using MPV, PDW, neutrophil-to-lymphocyte ratio (NLR), PLR, MPV/P and other inflammatory cells between patients with metastatic BC and those without. However, Huang et al. [20] reported that PDW/P was not statistically meaningful in the long-term prognosis of BC in univariate analysis of overall survival, and MPV/P, PDW/P and MPV had no statistical significance in the multivariate analysis.

It can be concluded that the clinical significance of peripheral blood PVIs in BC remains obscure. To the best of our knowledge, PVIs in the Chinese population in the processing of information on how BC prognosis has changed remain restricted. Therefore, the intention of this study was to assess the preoperative influence of PVIs on disease-free survival (DFS) in invasive BC with the largest possible sample and longer follow-up time to ascertain their proportional meaning.

Materials & methods

Study population

The current review cohort study included all hospitalized BC patients first diagnosed at the People's Hospital of Guangxi Zhuang Autonomous Region (PHGZAR), China, between March 2013 and December 2016 (registration number: ChiCTR2200058542; http://www.chictr.org.cn/index.aspx).

After approval by the Ethics Committee of the PHGZAR, a historical cohort was constructed from March 2013 to December 2016, using the International Classification of Diseases code C50 and limiting all adult inpatients to individuals of Asian ethnic background. This study used the resource database of the PHGZAR, which is an advanced medical data management system in China. It connects and indexes the hospital's diagnostic and treatment records. The electronic data include demographic characteristics, medical diagnosis codes and services, surgical codes, medication prescriptions and death information. A series of 674 female participants were recruited for this study. A total of 22 patients with preoperative recurrence and metastasis, 58 patients with preoperative neoadjuvant therapy and incomplete pathological data, 66 patients with noninvasive cancer and no operative therapy in the authors' hospital, two male BC patients and two patients with no follow-up visits were excluded (Figure 1). A series of 524 invasive BC patients were enrolled in this research and were followed up until 30 June 2022. The median follow-up period was 6.76 years. Lymph node metastases, primary tumor size, distant metastases, histological grade, estrogen receptor (ER) status, progesterone receptor (PR) status and HER2 status were examined and recorded for each patient. The primary end point of this research was DFS, which was defined as the interval from the initial treatment date to initial observed disease recurrence.

Immunohistochemistry & HER2 analysis

The evaluation of ER and PR status was performed using immunohistochemistry (IHC). Samples were considered ER and PR positive if at least 1% immunostained tumor nuclei of the nucleus of an immune-stained tumor was detected (Pathology Department, PHGZAR, China). Detection of HER2 status positive by IHC or FISH and IHC score 3+ or HER2 score 2+ in tumor cells was considered positive when HER2 amplification was confirmed by FISH (Pathology Department, PHGZAR, Guangxi Province, China).

Clinical hematological examination

Fasting blood samples were collected via peripheral vein puncture before any treatment commenced. Platelets, MPV and PDW were measured with an automatic blood and fluid analyzer (Sysmex XN-9000, Kobe, Japan). PDW/P and MPV/P values were calculated as the PDW and MPV, respectively, divided by the platelet count. Receiver operating characteristic (ROC) curves were used to determine thresholds.

Statistical analysis

Median and interquartile range or frequency (percentage) were used for continuous variables and categorical variables are expressed in terms of frequency (percentage). Statistical comparisons of categorical variables by χ² test and Wilcoxon rank sum test were employed to detect differences between two groups of normally distributed continuous variables. The optimal threshold values for PDW, MPV, MPV/P and PDW/P were determined from ROC curves. The estimated hazard ratio (HR) using Cox analysis is the relative risk reported with a corresponding 95% CI. Kaplan-Meier curves were utilized to compare survival differences between the cohorts. All statistical analyses were completed and validated by SPSS 18.0 (Windows version. IBM Corp. NY, USA); p < 0.05 was assumed to be statistically meaningful.

Data collection & ethics compliance

Since this study was a retrospective study, it was authorized by the Ethics Review Committee of the PHGZAR. The medical data for all trial subjects were anonymized. The PHGZAR and Guangxi Academy of Medical Sciences announced the research plan and the option to freely reject participation. Patients were deemed to have agreed to the study unless they explicitly refused to participate.

Results

Subjects characteristics

Table 1 summarizes patient characteristics. The median age was 51.87 ± 11.59 years (range: 22–88 years). There were 290 patients (55.34%) >50 years of age and 234 patients (44.66%) <50 years of age. There were 215 patients (41.03%) with axillary lymph node metastasis, 288 patients (54.96%) without axillary lymph node metastasis and 21 patients (4.01%) without confirmed axillary lymph node metastasis. A total of 105 patients (20.04%) had luminal A and 172 patients (32.82%) had luminal B (HER2-), which accounted for the majority of molecular typing. There were 55 HER2+ (HR+) patients (10.50%); 76 HER2+ (HR-) patients (14.50%); 72 triple-negative patients (13.74%) and 44 clinically unconfirmed (8.40%).

According to the ROC curve analysis (Table 2), the optimal cutoff values for PDW, MPV, PDW/P and MPV/P were 10.05, 10.05, 0.06 and 0.53, respectively. Table 3 shows the associations between PVIs and clinicopathological variables. Increases in PDW/P and MPV/P were dramatically related to age (p < 0.05).

Survival

The median follow-up was 6.76 years, during which 81 patients (15.46%) experienced a recurrence. Clinicopathological factors and the values of PVIs for the long-term prognosis of invasive BC were further explored. For tumor size, patients with a size >20 mm had a worse DFS HR of 2.23 (95% CI: 1.36–3.66; p < 0.05). In multivariate analysis, tumor size was not statistically significant (Table 4).

The presence of axillary lymph node metastases was related to a worse long-term prognosis than the absence of axillary lymph node metastases, with an HR of 2.01 (95% CI: 1.29–3.15; p < 0.05). Axillary lymph node metastasis was not statistically significant in the multivariate analysis. However, clinicopathological factors such as age, histological grade and tumor immunological indicators were not statistically significant in relation to DFS in these patients (Table 4). For PVIs, using the optimum threshold as a cutoff point, higher PDW/P, PDW and MPV/P were significantly associated with poor DFS prognosis, with HRs of 2.42 (95% CI: 1.43–4.09; p < 0.05), 2.03 (95% CI: 1.05–3.93; p < 0.05) and 2.10 (95% CI: 1.24–3.54; p < 0.05). In multifactorial analysis, only PDW/P was statistically significant at an HR of 2.61 (95% CI: 1.14–5.99; p < 0.05; Table 4). To further demonstrate the prognostic long-term efficiency of PDW/P for invasive BC, survival analysis was performed using Kaplan-Meier curves showing that subjects with higher PDW/P had a worse long-term prognosis compared with subjects with low PDW/P (5-year survival rate: 74.17 vs 88.39%; p < 0.05; Figure 2).

Discussion

Platelets are small pieces of cytoplasm that are shed from the cytoplasm of mature bone marrow megakaryocytes by lysis. Their main functions are coagulation and hemostasis. Recent reports have confirmed that in addition to their regulatory roles in hemostasis, inflammatory responses and host immunity, activated platelets secrete a variety of growth factors. Among them are VEGF, PDGF and PAF, which enhance tumor cell invasiveness, epithelial–mesenchymal transition, and extravasation and promote angiogenesis and vascular remodeling [21–25]. The interaction between tumor cells in the blood and activated platelets can make cancer patients more prone to thrombotic events and further increase the degree of malignancy of tumor cells [26,27]. Platelets also have a vital function in modulating the metastatic efficiency and survival of various tumor cells in the blood [28,29]. They can adhere to circulating tumor cells, support their extravasation, induce their proliferation [30,31] and even activate and recruit inflammatory cells such as neutrophils or macrophages to induce tumor cells to escape natural killer cell lysis [32]. Kang et al. revealed that PDGF facilitates the invasiveness of BC cells through the NFkB signaling pathway [33]. The platelet-specific receptor glycoprotein VI of collagen and fibrin interacts with BC cells and uses ITAM signaling components in platelets to regulate platelet adhesion, aggregation, procoagulant activity and other functions, which is conducive to extravasation of BC cells [34].

PDW and MPV are morphological indicators of platelet volume size distribution and variability. Generally speaking, the two are inversely proportional [35] and are often used to distinguish between low-yield and high-destructive thrombocytopenia [36]. The quantity of research on the biological relationship between MPV and the long-term prognostic progression of various types of cancer is quite limited. However, recent clinical research by Riedl et al. [37] using data from 1544 participants with various types of cancer revealed that high MPV values were not only connected to a decrease in the risk of venous thromboembolism but also improved survival rates. Moreover, in patients with advanced NSCLC, low levels of MPV predict poor prognosis [14]. On the contrary, other research has shown that increased MPV is statistically significantly related to the poor prognosis of esophageal cancer, BC and hepatocellular cancer [11–13]. Thus, the value of MVP as a prognostic indicator for BC is presently unclear. These conclusions should be verified in larger prospective studies with larger samples and longer follow-ups. However, it is reasonable to estimate the prognosis of patients with malignant tumors by comparing MPV with platelet count as an indicator of platelet activation, since it seems to gather the advantages of two measures of platelet activation. In fact, inflammatory cell ratio indices such as PLR, monocyte-to-lymphocyte ratio or NLR are widely used, in part because the cutoff values for these indirect indices are not unique. Therefore, our cutoff values for MPV, PDW, PDW/P and MPV/P based on the ROC curves are reasonable, although they are not the dividing line between normal and abnormal. PDW is an index reflecting the average change in platelet volume. Because it does not accumulate in response to platelet bulking [38], it is often considered a more concrete marker of an activated platelet object. Compared with MPV, it is possible that PDW is a more accurate and reliable prognostic biomarker. In this study, MPV was not correlated with prognosis in BC. PDW, PDW/P and MPV/P were statistically significant for the prognosis of BC in univariate analysis. In multivariate analysis, only PDW/P had statistical significance for BC prognosis, which is similar to the recent report of Takeuchi et al. [39]. The current study also confirmed that PDW, PDW/P and MPV/P are relevant to the survival of Asian BC patients. Multicenter and large-sample clinical trials in other ethnic groups are needed to determine whether these findings are genetically race-specific.

Currently, the main mechanism linking PVIs to survival is unclear. Changes in PVIs may be accompanied by dysfunction of megakaryocytes and bone marrow cells. More recent research has revealed that cytokines secreted by tumor cells, for example, include granulocyte colony-stimulating factor and macrophage colony-stimulating factor, modulate megakaryocyte cell maturation, thrombocyte release and production as well as the size of thrombocytes [40]. In addition, local inflammatory infiltration plays a crucial function in the tumor microenvironment. Several inflammatory cytokines like IL-1, IL-8, IL-6, IL-18, IL-12, TNF and IFN-gamma are upregulated during the progression of tumor cell infiltration [41]. These cytokines promote abnormal maturation of bone marrow cells and megakaryocytes, which results in the increased generation and emission of premature platelets with different properties and sizes in the cardiovascular system. Subsequent interactions between activated platelets and cancer cells lead to angiogenesis, tumor development, tumor invasion and metastasis, further increasing the malignant degree of tumor cells. The convenience and low cost of measuring PVIs in clinical practice make it a potential biomarker for prognostic risk assessment and individualized treatment options in BC. However, before its clinical application, more quality basic and multicenter prospective studies are needed to better explain why higher PDW, PDW/P and MPV/P lead to adverse prognoses in invasive BC.

The strengths of this research include the large sample size, long follow-up time, good matching and multivariate analysis. However, there are some drawbacks that cannot be ignored. First, this study is single-center and cannot represent the entire Asian population. Second, the retrospective character of the studies could also lead to data selection and analysis bias. Therefore, high-quality basic studies and well-designed polycentric prospective studies are warranted to further explore the connection between the composite index of peripheral blood platelet volume and the prognosis of invasive BC, and to clarify the specific mechanism.

Conclusion

Based on the current findings, tumor size, PDW, MPV/P, PDW/P and axillary lymph node involvement are significant clinical factors that impact the prognosis of BC. Moreover, this analysis suggests that the PDW/P ratio has the potential as an independent predictor of prognostic risk in invasive BC. However, further research, including clinical and molecular/genomic validation, is necessary to determine if this data can explain and improve daily clinical treatment regimens.