Pre-treatment Elevated Platelet Count Associates with HER2 Overexpression and Prognosis in Patients with Breast Cancer

Purpose: To research the association between pre-treatment elevated platelet count and clinicopathologic characteristics in breast cancer (BC), as well as explore the relationship between pre-treatment elevated platelet count and HER2 status and prognosis of BC patients. Materials and Methods: A retrospective cohort of BC patients who were newly diagnosed or treated by surgery only and had pathological detection results and platelet values in the Department of Oncology, the First Affiliated Hospital of Liaoning Medical College were enrolled from 1/1/2008 until 31/12/2009, and followed up until 31/12/2014. Age, thrombocyte parameters before chemotherapy and/or radiotherapy, immunohistochemical (IHM) indexes, and regional lymph node (LN) involvement and progression-free survival (PFS) were recorded. Results: A total of 447 eligible subjects were included in this research. As we analyzed, for HER2, positive and negative, the incidence rates of elevated platelet count were 25.8% and 14.7% ( P <0.05). In the Cox proportional hazards model both variables were independent risk factors for BC (for HER2, OR, 0.592, 95% confidence interval, CI, 0.355 to 0.985, P =0.044;f or PLT, OR, 0.998, 95% CI, 0.996 to 1.000, P =0.042). For ER, PR, Ki67 and LN involvement, the differences were not statistically significant ( P >0.05). Conclusions: In this research, pre-treatment elevated level of platelet count demostrated a significantrelationship with HER2 amplification/overexpression, and both variables significantly influenced the prognosis of BC. However, elevated platelet count did not exhibit any association with ER, PR, Ki67 and LN involvement. performed a multivariate, retrospective cohort analysis in an attempt to assess the correlations between elevated platelet count with HER2 and BC prognosis.


Introduction
As a adverse prognostic factor,elevated level of platelet count has certain incidence in patients of a variety of solid cancers (Monreal et al., 1998;Buergy et al., 2012;Stravodimou et al., 2013;Fang-Xuan Li et al., 2014;Yan Li et al., 2014;Ying Chen et al., 2015), whereas the reasons were not exactly clear (Taucher et al., 2003;Buergy et al., 2012). But evidence of the prognostic role of elevated platelet count is inadequate. Previous studies (Gerdes et al., 1983;Colleoni et al., 2006;Gown, 2008;Davies et al., 2011;Pathmanathan et al., 2014;Ramirez et al., 2014) also reported that pathological indexes including HER2, ER, PR, Ki67 and LN involvement play a key role in the progress of diagnosis, treatment and prognosis of BC. And in the process of clinical practice, patients had different expression status of HER2 observed a concomitant of different incidence rates of elevated platelet count, and had different survival time, but the mechanisms remain to be further studied. We therefore

Pre-treatment Elevated Platelet Count Associates with HER2 Overexpression and Prognosis in Patients with Breast Cancer
Mei-Ling Gu 1 , Cai-Jun Yuan 1 *, Xiao-Mei Liu 1 * , Yi-Chao Zhou 2 , Shu-Huan Di 3 , Fei-Fei Sun 4 , Quan-Ying Qu 5 performed a multivariate, retrospective cohort analysis in an attempt to assess the correlations between elevated platelet count with HER2 and BC prognosis.

Study subjects
The study was approved by The First Affiliated Hospital of Liaoning Medical University ethical committee, and all patients signed a written informed consent.We conducted this multivariate, retrospective cohort analysis by filtrating 447 eligible subjects from 1372 BC patients from 1/1/2008 until 31/12/2009,and following up 6 years until 31/12/2014.These 447 cases were all newly diagnosed,and have detection results of IHM indexes and LN involvement,with the tissue samples were obtained by surgery or puncture,and had platelet parameters record detected before adjuvant therapy.Then large amounts of data statistics was taken to collect information such as: age,platelet counts, HER2, ER, PR, Ki67, LN involvement and PFS.

Classification of subgroups
On account of the normal reference ranges of platelet count was 100-300*10 9 /L, the standard of the elevated was defined as a platelet count exceeding 350*10 9 /L in this research,for purpose of minimizing the false positive rate.And in order to ensure the accuracy of the HER2 amplification/overexpression detection results,we selected those the IHM results were "3+"or those the FISH detection results were amplified or positive to be the HER2-positive group.In spite of the cut-off values to distinguish Ki67-high from Ki67-low in BC were uncertain,on the base of the highest Youden's index (0.346) (Bewick et al., 2004), <10 and ≥10 were chosen to divide the Ki67 subgroups.The outcome factor was the untreated elevated platelet count (≥350*10 9 /L) while the study factors were HER2 (positive or negative), ER (positive or negative), PR(positive or negative), LN (positive or negative), Ki67 (<10 or ≥10; or continuous).

Follow-up visits
All patients were followed up according to the National Comprehensive Cancer Network (NCCN) guidelines recommendations since the first visit to our department to 12/31/2014.

Statistical analysis
Chi-square tests was performed to assess the relationship between elevated platelet count and subgroups of every variable, significance inferred at P<0.05. And Cox proportional hazards regression analysis was used to estimate or confirm hazard ratios of the related variables, choosing 95 percent as confidence intervals. Kaplan-Meier curves were presented, and log-rank tests for the comparison of survival curves. Furthermore, to minimize the risk of identifying a chance retrospective statistical correlation, all computations throughout the total cohort were repeated.

Results
The analysis of demographic characteristics of the cohort A total of 447 untreated patients with BC during the past 72 months were analyzed. These patients has a mean (SD) age of 52.7(10.0),range from 20 to 84 years old. Among the total cohort (Table 1), 99.6% patients were female, 86.4% patients were treated with surgery, and the rest 13.6% were diagnosed by puncture. All the 447 patients were treated by radiotherapy and/or chemotherapy. Until the end of follow-up, 62.4% patients were alive without progressive disease, 25.3% patients were alive with recurrence or metastasis and 11.4% patients were dead,3.7% patients lost to follow-up.The positive rates of HER2 expression and LN involvement were 16.8% (62/369) and 55.3% (213/385), respectively.

Correlations between pre-treatment elevated platelet count and clinicopathologic characteristics
The analysis results of platelet count in terms of five variables were shown in Table 2. In the HER2-positive group, the incidence rates of elevated platelet count was higher than that in the HER2-negative group (25.8% vs 14.7%, p=0.031, Figure 1). However, the associations between elevated platelet count and ER, PE, Ki67 and LN did not reach statistical significance (P=0.691,P=0.197,P=0.467,P=0.361,respectively,.

Influence of platelet count and HER2 status on prognosis of BC
As we analyzed, both the two variables were identified as the independent risk factors for poor prognosis of   Pedersen et al.,1996;Nakano et al.,1998;Eltabbakh et al., 1999;O'Byrn et al., 1999;Ikedaet al., 2002). The positive finding of the study was that the elevated platelet count and the HER2 status were statistically significant correlated, and both these two factors were significantly related to the prognosis of BC.Elevated platelet count remarkably influences the OS and PFS of BC patients, by affecting the metastatic potential of tumor cells. At the level of molecular biology, interaction between high platelet count and malignancy was a complicated network of signaling pathways via cytokines like vascular endothelial growth factor (VEGF), plateletderived growth factor (PDGF), transforming growth factor β (TGF-β), interleukin 6(IL-6), and others that have been implicated in phases of tumor growth and progression. Among these biochemical markers, VEGF played a critical role to connect elevated platelet value and HER2 together (Verheul et al., 1997;Laughner et al., 2001). As expected,our research demonstrated the confidence intervals, CI:0.996 to 1.000, P=0.042),which was in accord with previous research (Taucher et al., 2003;Burstein, 2005;Pathmanathan et al., 2014). Figure 6, 7 showed the survival functions.

Discussion
From this retrospective research,the central insight is that the elevated platelet count had a statistical significant positive correlation with HER2 status.As the identification of the correlation,the author considered it has important implications to further study.
The percentage of elevated platelet value in the total cohort appeared to be 15.21%, comparable to 10-63% in other solid cancers (Olesen et al., 1988;Gastl et al.,1993; significant relevant between elevated platelet value and HER2 upregulation.But question remains unresolved as to how exactly the elevated platelet count is correlated with HER2 amplification/overexpression in BC patients. One explanation is that the HER2-positive BC triggers a upregulation of VEGF,therefore lead to the elevated level of platelet count. Other possible perspectives are HER2positive BC promotes a high level of platelet, platelets therefore release large doses of VEGF in the plasma. In this research, due to the elimination of patients whose IHM results of HER2 were"2+", the positive percentage of HER2 was 16.80%,which was not comparable to the 20-30% in previous study (Foy et al.,2012).However,the lower HER2 positive percentage minimized the false positive rate of elevated platelet count in the HER2positive subgroup. HER2 was reported overexpressed in gastric cancer, and especially in BC. The upregulation of HER2 in BC promotes elevated risk of recurrence, poor prognosis,and even do harm to the efficacy of hormonal therapy (Puhalla et al.,2013),so the identification of its relevance to high platelet values in our research may help to provide synergism to the diagnose and adjuvant therapy targeting biochemical markers of BC.
Advantages of this research were a multivariate cohort analysis and a relatively big sample size.But some limitations should be mentioned. Firstly, our research was a hospital-based study in China,the subjects may not well represent the general population.Secondly,the results should be cautioned because of the differences of region and race.Thirdly,we need to further study the mechanisms of the correlation between elevated platelet count and HER2 amplification/overexpression,and large sample size multi-center prospective clinical trials were needed to confirm the effect of antiplatelet treatment in HER2-positive BC therapy.
In conclusion, the positive finding of the study was that pre-treatment elevated platelet count and the HER2 status were statistically significant associated, and both these two factors were significantly related to the prognosis of BC. Therefore, antiplatelet treatment maybe useful to improve therapeutic efficacy and prognosis of HER2positive BC patients.