Decreased platelet distribution width predicts a worse prognosis in patients undergoing surgical resection for hepatocellular carcinoma

Abstract

BACKGROUND:

Hepatocellular carcinoma (HCC) is the sixth most common malignant tumor and second most common cause of tumor-related deaths worldwide. Activated platelets play a prominent role in tumor. Platelet distribution width (PDW) indicates platelets activation and is altered in malignancies. The aim of this study was to explore the prognostic value of PDW for overall survival (OS) in HCC patients.

METHODS:

We retrospectively reviewed 273 HCC patients at a single institution from 2010 to 2014. The relationship between PDW and clinicopathological characteristics was analyzed. Kaplan-Meier curves and multivariate Cox regression analyses were used to evaluate the relationship of PDW with OS.

RESULTS:

Low PDW levels were observed in 127 (46.5%) out of 273 patients. A significant correlation was found between PDW and liver cirrhosis. Median follow-up was 36 months, survival curves revealed that the patients with increased PDW had significantly shorter survival time than those with normal PDW ( $p=$ 0.001). Cox regression analysis demonstrated that PDW was an independent prognostic factor for overall survival (hazard ratio, 2.464; 95% confidence interval [CI], 1.402–4.330, $p=$ 0.001).

CONCLUSION:

PDW is significantly associated with OS in HCC. This result suggests activated platelet may affect clinical outcome and warrant continued investigation.

Keywords

Hepatocellular carcinoma platelet distribution width prognosis

1. Introduction

Hepatocellular carcinoma (HCC) is the sixth most common malignant tumor and second most common cause of tumor-related deaths worldwide [1, 2]. Although surgical resection and liver transplantation are still the best options for HCC patients, the overall 5-year survival rate remains low due to high postoperative recurrence rate and metastasis [3, 4]. Thus, there is an urgent need of discovering new biomarkers aiding in prognosis and developing curative strategies against HCC.

There is abundant evidence that platelets are associated with tumor development and tumor progression. Pretreatment thrombocytosis is a poor prognostic factor as well and reflects aggressive tumor behavior in malignant mesothelioma, cervical, colon, and non-small cell lung cancer patients [5, 6]. There are accumulated data on the substantial influence of blood platelets in the process of blood vessel formation during malignancy [7]. Moreover, platelets contribute to early and late stages of angiogenesis through platelet-derived microparticles, microRNA, lipids, and variety of surface receptors [7].

Platelet distribution width (PDW), a platelet marker, indicates variation in platelet size [8]. Elevated PDW levels were associated with poor prognosis in a variety of tumors, such as gastric cancer, non-small cell lung cancer, laryngeal cancer, and melanoma [9, 10, 11, 12]. However, the clinical implication of PDW in HCC has not been well defined.

The current study aims to test whether pretreatment PDW could predict survival in HCC patients.

2. Methods

2.1 Study population

The present study retrospectively reviewed the clinical data of 273 patients histologically diagnosed with HCC from Harbin Medical University Cancer Hospital between January 2010 and December 2014 (Harbin, China). All patients underwent surgical resection. None of the patients received any adjuvant therapy before resection. We excluded patients with hematological disorders, hypertension, diabetes mellitus, and medical treatment with anticoagulant therapy, statins, and acetylic salicylic acid. Overall survival (OS) was defined as the interval from the date of diagnosis to date of death or last follow-up. Follow-up evaluations were performed every 3 months. Follow-up was completed on December 31, 2017. The follow-up duration was 36 months.

The clinicopathohistological data including age, sex, HbsAg, liver cirrhosis, tumor size, tumor number, tumor differentiation, regional lymph node metastasis, Child-pugh’s grade, and the therapy administered were all collected through medical records. Enhanced computed tomography or magnetic resonance imaging was performed to rule out distant metastases. In case of multiple tumors with different size, largest diameter was used. Cirrhosis was diagnosed based on the data from history or on a combination of endoscopic or radiological evidence, or of portal hypertension. All patients were confirmed to have liver cirrhosis by pathologic assessment at the time of surgery. The status of lymph nodes larger than 4.0 mm was reported pathologically. The removed lymph nodes were pathologically confirmed to be classified into a metastatic group or non-metastatic group. The information on blood tests was obtained from the test report from the Department of Clinical Laboratory at the Harbin Medical University Cancer Hospital. The normal range for $\alpha$ -fetoprotein (AFP) in our hospital was 0–7 ng/ml. Platelet indices were determined with an automated hematological analyzer (Sysmex XE-2100, Kobe, Japan). The normal range for PDW in our hospital was 11–17%. Cigarette smoking was defined as having smoked at least 100 cigarettes during one’s lifetime. Alcohol drinking was defined as the consumption of at least 30 g of alcohol per week for 1 year or more.

The present study is a retrospective study and was reviewed and approved by the Institutional Ethics Review Board of Harbin Medical University Cancer Hospital. All patients provided written informed consent.

2.2 Statistical analysis

Data are expressed as the mean $\pm$ standard deviation or percentages. The significance of association between PDW and the clinicopathological characteristics were assessed using the $\chi^{2}$ test. The cut-off value of PDW was determined with receiver operating characteristic (ROC) curve with the best accuracy (the greatest sensitivity and specificity) using MedCalc version 15.0. Kaplan-Meier curve with log-rank test and Cox proportional hazards regression were used to determine the prognostic roles of the biomarkers. Hazards ratio (HR) and 95% confidential interval (CI) were used to measure the strength between them. All variables with a statistically significant univariate association ( $p<$ 0.10) were included in the multivariate analysis. All statistical tests were performed in SPSS version 22.0 software (SPSS Inc., Chicago, IL, USA). A $p$ value $<$ 0.05 was considered statistically significant.

Figure 1.

The ROC curve for PDW. The optimized cut-off value for PDW using standard ROC curve analysis.

Table 1

Baseline characteristics of HCC patients according to PDW levels

Variables	Total	PDW $\leqslant$ 17.1	PDW $>$ 17.1	$P$ value
	$n$ (%)	$n$ (%)	$n$ (%)
Age (years)				0.185
$\leqslant$ 55	162 (59.3)	92 (63.0)	70 (55.1)
$>$ 55	111 (40.7)	54 (37.0)	57 (44.9)
Gender				0.885
Male	132 (48.4)	70 (47.9)	62 (48.8)
Female	141 (51.6)	76 (52.1)	65 (51.2)
AFP				0.278
$>$ 20 ng/ml	156 (57.1)	79 (54.1)	77 (60.6)
$\leqslant$ 20 ng/ml	117 (42.9)	67 (45.9)	50 (39.4)
HbsAg				0.238
Negative	58 (21.2)	35 (24.0)	23 (18.1)
Positive	215 (78.8)	111 (76.0)	104 (81.9)
Liver cirrhosis				0.001 ${}^{\ast}$
With	172 (63.0)	79 (54.1)	93 (73.2)
Without	101 (37.0)	67 (45.9)	34 (26.8)
Tumor Size				0.989
$\geqslant$ 5 cm	142 (52.0)	76 (52.1)	66 (52.0)
$<$ 5 cm	131 (48.0)	70 (47.9)	61 (48.0)
Tumor number				0.130
Single	235 (86.1)	130 (89.0)	105 (82.7)
Multiple	38 (13.9)	16 (11.0)	22 (17.3)
Tumor differentiation				0.809
Poor	75 (27.5)	41 (28.1)	34 (26.8)
Moderate/well	198 (72.5)	105 (71.9)	93 (73.2)
Regional lymph node metastasis				0.855
Absent	255 (93.4)	136 (93.2)	119 (93.7)
Present	18 (6.6)	10 (6.8)	8 (6.3)
Child-pugh’s grade				0.181
A	266 (97.4)	144 (98.6)	122 (96.1)
B	7 (2.6)	2 (1.4)	5 (3.9)

HCC, hepatocellular carcinoma; AFP, $\alpha$ -fetoprotein; PDW, platelet distribution width. ${}^{\ast}p<$ 0.01.

Table 2

Baseline characteristics of HCC patients according to PDW levels

Variables	PDW $\leqslant$ 17.1	PDW $>$ 17.1	$P$ value
Age (years)	51.8 (8.7)	53.2 (9.8)	0.198
BMI (kg/m ${}^{2}$ )	24.1 (2.9)	23.8 (3.0)	0.303
Cigarette smoking ( $n$ , %)	53 (36.3)	47 (37.0)	0.904
Alcohol drinking ( $n$ , %)	26 (17.8)	23 (18.1)	0.948
TBIL ( $\mu$ mol/L)	12.4 (9.2–16.3)	12.6 (9.7–18.9)	0.178
AST (U/L)	23 (16–38)	26 (18–45)	0.181
ALT (U/L)	22 (17–32)	22 (18–35)	0.248
GGT (U/L)	52 (36–88)	66 (38–106)	0.086
Albumin (g/L)	41.0 (3.8)	41.9 (5.4)	0.117
FPG (mmol/L)	5.00 (4.60–5.30)	5.00 (4.61–5.30)	0.600
WBC ( $\times$ 10 ${}^{9}$ /L)	5.67 (1.88)	4.95 (1.75)	0.001 ${}^{\ast}$
NLR	2.04 (1.49)	2.22 (1.29)	0.297
RDW (%)	13.6 (2.4)	14.1 (1.2)	0.030 ${}^{\ast\ast}$
Hemoglobin (g/L)	136.0 (17.5)	139.1 (18.3)	0.157
Platelet count ( $\times$ 10 ${}^{9}$ /L)	177.2 (76.8)	132.1 (63.1)	$<$ 0.001 ${}^{\ast}$
MPV (fL)	9.7 (1.5)	10.2 (2.0)	0.025 ${}^{\ast\ast}$

Data are expressed as means (SD) or median (IQR). BMI, body mass index; AST, aspartate aminotransferase; ALT, alanine aminotransferase; GGT, gamma-glutamyl transpeptidase; TBIL, total bilirubin; FPG, fasting plasma glucose; WBC, white blood cell; RDW, red cell distribution width; NLR, neutrophil-to-lymphocyte ratio; MPV, mean platelet volume. ${}^{\ast}p<$ 0.01; ${}^{\ast\ast}p<$ 0.05.

Table 3

The univariate and multivariate analysis of overall survival in HCC patients

Variables	Univariate analysis		Multivariate analysis
	HR (95% CI)	$P$ -value	HR (95% CI)	$P$ -value
Age (years) ( $>$ 55 versus $\leqslant$ 55)	1.313 (0.791–2.181)	0.292
Gender (male versus female)	0.840 (0.505–1.396)	0.500
BMI (kg/m ${}^{2}$ )	1.025 (0.943–1.114)	0.563
Cigarette smoking (yes versus no)	1.068 (0.635–1.797)	0.805
Alcohol drinking (yes versus no)	0.787 (0.388–1.599)	0.508
HbsAg (positive versus negative)	0.681 (0.385–1.208)	0.189
Liver cirrhosis (with versus without)	0.727 (0.436–1.211)	0.221
TBIL ( $\mu$ mol/L)	1.360 (0.823–2.248)	0.231
AST (U/L) (log-value)	1.033 (0.715–1.493)	0.861
ALT (U/L) (log-value)	0.704 (0.417–1.188)	0.188
GGT (U/L) (log-value)	1.940 (1.456–2.584)	$<$ 0.001 ${}^{\ast}$	1.373 (0.997–1.891)	0.052
Albumin (g/L)	1.008 (0.954–1.065)	0.782
FPG (mmol/L) (log-value)	0.560 (0.061–5.172)	0.609
AFP (ng/ml) ( $>$ 20 versus $\leqslant$ 20)	1.592 (0.931–2.724)	0.089	1.194 (0.660–2.161)	0.557
Hemoglobin (g/dl)	1.003 (0.989–1.018)	0.645
WBC ( $\times$ 10 ${}^{9}$ /L)	1.180 (1.044–1.335)	0.008 ${}^{\ast}$	1.125 (0.960–1.319)	0.146
RDW (%)	1.044 (0.954–1.141)	0.351
NLR	1.312 (1.174–1.467)	$<$ 0.001 ${}^{\ast}$	1.063 (0.900–1.256)	0.473
Platelet count ( $\times$ 10 ${}^{9}$ /L)	1.001 (0.998–1.005)	0.360
Tumor size (cm) ( $\geqslant$ 5.0 versus $<$ 5.0)	2.833 (1.598–5.023)	$<$ 0.001 ${}^{\ast}$	1.921 (1.057–3.491)	0.032 ${}^{\ast\ast}$
Tumor number (multiple versus single)	0.822 (0.374–1.807)	0.625
Tumor differentiation (moderate/well versus poor)	2.897 (1.745–4.810)	$<$ 0.001 ${}^{\ast}$	2.526 (1.450–4.399)	0.001 ${}^{\ast}$
Regional lymph node metastasis (present versus absent)	3.893 (1.970–7.694)	$<$ 0.001 ${}^{\ast}$	2.595 (1.263–5.332)	0.009 ${}^{\ast}$
TNM stage (III versus I $+$ II)	2.849 (1.675–4.846)	$<$ 0.001 ${}^{\ast}$	1.818 (0.901–3.665)	0.095
Child-pugh’s grade ( $B$ versus $A$ )	5.335 (2.124–13.400)	$<$ 0.001 ${}^{\ast}$	2.406 (0.829–6.982)	0.106
MPV (fL)	0.891 (0.757–1.047)	0.162
PDW (%) ( $>$ 17.1 versus $\leqslant$ 17.1)	2.371 (1.395–4.031)	0.001 ${}^{\ast}$	2.464 (1.402–4.330)	0.001 ${}^{\ast}$

Variables that showed a $p$ -value $a<$ 0.10 in univariate analysis were included in a multivariate Cox proportional hazards regression model. CI, confidence interval. Abbreviations: see to Tables 1 and 2. ${}^{\ast}p<$ 0.01; ${}^{\ast\ast}p<$ 0.05.

3. Results

Among the 273 patients, the mean age was 52.5 $\pm$ 9.2 years (range 24–74). One hundred and thirty-two (48.4) were male and 141 (51.6) were female.

According to the ROC curve analysis, we identified that the optimal cut-off value of 17.1 for PDW had a 65.0% sensitivity and 58.7% specificity for survival rate ( $\text{AUC}=$ 0.623, 95% CI: 0.548–0.699, $p=$ 0.001) (Fig. 1). Then, patients were divided into 2 groups: patients with $\text{PDW}\leqslant$ 17.1% and patients with $\text{PDW}>$ 17.1%. There were 146 (53.5%) patients with $\text{PDW}\leqslant$ 17.1% and 127 (46.5%) patients with $\text{PDW}>$ 17.1%.

The relationships between PDW and clinical characteristics were summarized in Tables 1 and 2. Our study found that PDW was significantly associated with platelet count, WBC, and liver cirrhosis. However, there were no significant differences between the groups in terms of age, gender, smoking status, drinking status, BMI, hemoglobin, AFP, HbsAg, tumor size, tumor number, tumor differentiation, and regional lymph node metastasis.

With a follow up of 36 months, 60 (22.0%) patients had death events. Patients with $\text{PDW}>$ 17.1% exhibited significantly shorter overall survival compared to those with $\text{PDW}\leqslant$ 17.1% (69.3% vs. 85.6%, $p=$ 0.001). The Kaplan-Meier OS curves of the normal versus elevated PDW showed a significant separation (Fig. 2).

Figure 2.

The correlation between PDW levels and overall survival in patients with HCC. The cumulative incidence of overall survival is significantly higher in patients with low PDW levels.

Cox univariate and multivariate regression analyses were performed for the factors that could impact survival. GGT, WBC, PDW, neutrophil-to-lymphocyte ratio, tumor size, regional lymph node metastasis, tumor differentiation, and Child-pugh’s grade were determined to be significant in univariate analyses. AFP showed weak associations. Other parameters were not found to be in correlation with OS (Table 3). Next, all the factors with a $P$ value less than 0.10 in univariate analysis were included in multivariate analysis (Table 3). Multivariate analysis determined that PDW was an independent variable associated with survival. Patients with $\text{PDW}>$ 17.1% had a hazard ratio (HR) of 2.464 [95% confidence interval (CI): 1.402–4.330, $p=$ 0.001] for OS.

4. Discussion

Here is the first study to investigate the prognostic value of preoperative PDW in HCC. We revealed that increased PDW was associated with liver cirrhosis. In addition, elevated PDW was an independent indicator of overall survival in HCC.

Despite new advances in treatment, the 5-year survival of HCC patients after resection ranges between 50% and 70% with high recurrence rate and metastasis [3, 4]. Therefore, it is of great importance in identifying novel prognostic factors. More and more evidence has emerged showing a pro-metastatic role of platelets in cancer [13].

Activated platelets inhibit hepatocellular carcinoma cell differentiation and promote tumor progression via platelet-tumor cell binding [14]. Platelet granule contents promote the proliferation of HCC cells by suppressing the expression of Kruppel like factor 6 [15]. Clinical studies found that platelet-derived endothelial cell growth factor is an angiogenic factor and increased in HCC patients [16]. Moreover, a recent report confirmed that platelet-derived growth factor receptor alpha overexpression is a prognostic marker independent of liver fibrosis status in HCC [17]. Furthermore, antiplatelet therapy delays immune-mediated hepatocarcinogenesis and improves survival in a mouse model of chronic hepatitis B [18].

In accordance with the studies above, our report indirectly confirmed the results using a simple platelet marker. Furthermore, our study can form the basis for further mechanistic studies and ultimately help clinicians with better individualization of their therapeutic strategy.

The biologic mechanisms underlying the correlation between PDW and prognosis in HCC remains poorly understood. Altered PDW maybe related to bone marrow cells (including megakaryocytes) dys-function. PDW is a measure of platelet heterogeneity caused by heterogeneous demarcation of megakarocytes [19]. Recent reports found that several cytokines, such as interleukin-6 (IL-6), granulocytes colony stimulating factor (G-CSF) and macrophage colony stimulating factor (M-CSF), regulate megakaryocytic maturation, platelet production and platelet size [20]. IL-6 promotes tumor angiogenesis, metastasis and metabolism [21]. In addition, G-CSF and M-CSF secreted by tumor cells stimulate megakaryopoiesis and subsequent thrombopoiesis in cancer [22].

This study has several limitations. First, this study had a retrospective design and included a relatively small number of patients. Second, the mechanisms underlying the involvement of PDW in HCC should be clarified. Third, this study lacked of internal/external validation. Finally, the patients were composed of Chinese. Multicenter prospective studies with larger samples are needed to verify our results.

In conclusion, increased PDW may predict the clinical outcomes of HCC patients. Further studies are warranted.

Footnotes

Acknowledgments

This work was supported by grants from the Harbin Medical University Cancer Hospital (JJZD2017-05) and from Nn10 program of Harbin Medical University Cancer Hospital.

References

Ferlay

Soerjomataram

Dikshit

Eser

Mathers

Rebelo

Parkin

D.M.

Forman

and Bray

, Cancer incidence and mortality worldwide: Sources, methods and major patterns in GLOBOCAN 2012, Int J Cancer 136 (2015), 359–386.

Jemal

Bray

Center

M.M.

Ferlay

Ward

and Forman

, Global cancer statistics, CA Cancer J Clin 61 (2011), 69–90.

Bruix

and Llovet

J.M.

, Major achievements in hepatocellular carcinoma, Lancet 373 (2009), 614–616.

Shimozawa

and Hanazaki

, Longterm prognosis after hepatic resection for small hepatocellular carcinoma, J Am Coll Surg 8 (2004), 356–365.

Gao

Zhang

and Wang

, Prognostic value of combination of preoperative platelet count and mean platelet volume in patients with resectable non-small cell lung cancer, Oncotarget (2017), 15632–15641.

Wang

Y.H.

Deng

S.J.

Yang

Y.D.

Yao

Zhao

J.M.

Min

G.T.

Wang

T.F.

Zhao

P.Y.

Wang

H.P.

and Chen

, The pretreatment thrombocytosis may predict prognosis of patients with colorectal cancer: A systematic review and meta-analysis, Biomark Med (2017), 195–210.

Wojtukiewicz

M.Z.

Sierko

Hempel

Tucker

S.C.

and Honn

K.V.

, Platelets and cancer angiogenesis nexus, Cancer Metastasis Rev (2017), 249–262.

Kaito

Otsubo

Usui

Yoshida

Tanno

Kurihara

Matsumoto

Hirata

Domitsu

and Kobayashi

, Platelet size deviation width, platelet large cell ratio, and mean platelet volume have sufficient sensitivity and specificity in the diagnosis of immune thrombocytopenia, Br J Haematol 8 (2005), 698–702.

Cui

M.M.

Liu

Yun

Z.Y.

Niu

Zhang

Gao

Liu

and Wang

R.T.

, Platelet distribution width correlates with prognosis of non-small cell lung cancer, Sci Rep 3456 (2017).

10.

Diao

Huang

Niu

Liu

Z.P.

Wang

R.T.

and Yu

K.J.

, Increased platelet distribution width predicts poor prognosis in melanoma patients, Sci Rep 2970 (2017).

11.

Zhang

Liu

Y.S.

Wang

Liu

Z.P.

Wang

R.T.

and Yu

K.J.

, Higher platelet distribution width predicts poor prognosis in laryngeal cancer, Oncotarget (2017), 48138–48144.

12.

Zhang

Cui

M.M.

L.L.

Liu

Y.S.

Liu

Z.P.

Liu

Wang

R.T.

and Yu

K.J.

, Platelet distribution width correlates with prognosis of gastric cancer, Oncotarget (2017), 20213–20219.

13.

Schlesinger

, Role of platelets and platelet receptors in cancer metastasis, J Hematol Oncol 125 (2018).

14.

Zhang

Guo

Liu

Y.J.

Cheng

Zhou

Zhao

and Liu

, Activated platelets inhibit hepatocellular carcinoma cell differentiation and promote tumor progression via platelet-tumor cell binding, Oncotarget (2016), 60609–60622.

15.

A.D.

Xie

Song

Y.Y.

Liu

Liang

M.L.

X.W.

Yao

G.Q.

Zhang

B.X.

Gao

C.J.

Xiang

J.Z.

and Ming

Z.Y.

, Platelet releasates promote the proliferation of hepatocellular carcinoma cells by suppressing the expression of KLF6, Sci Rep 3989 (2017).

16.

Jin-no

Tanimizu

Hyodo

Nishikawa

Hosokawa

Endo

Doi

Mandai

and Ishitsuka

, Circulating platelet-derived endothelial cell growth factor increases in hepatocellular carcinoma patients, Cancer (1998), 1260–1267.

17.

J.H.

Kim

J.M.

Kim

J.K.

Choi

S.J.

Lee

K.S.

Lee

J.W.

Chang

H.Y.

and Lee

J.I.

, Platelet-derived growth factor receptor α in hepatocellular carcinoma is a prognostic marker independent of underlying liver cirrhosis, Oncotarget (2017), 39534–39546.

18.

Sitia

Aiolfi

L.P.

Mainetti

Fiocchi

Mingozzi

Esposito

Ruggeri

Z.M.

Chisari

F.V.

Iannacone

and Guidotti

L.G.

, Antiplatelet therapy prevents hepatocellular carcinoma and improves survival in a mouse model of chronic hepatitis B, Proc Natl Acad Sci USA 9 (2012), 2165–2172.

19.

Paulus

J.M.

, Recent advances in the story of megakaryocyte physiology, Pathol Biol (Paris), (1981), 133–135.

20.

Kaushansky

, Growth factors and hematopoietic cell fate, A new feature: controversies in hematology, Blood (1998), 345–344.

21.

Kumari

Dwarakanath

B.S.

Das

and Bhatt

A.N.

, Role of interleukin-6 in cancer progression and therapeutic resistance, Tumour Biol (2016), 11553–11572.

22.

Kowanetz

Lee

et al., Granulocyte-colony stimulating factor promotes lung metastasis through mobilization of ly6G+Ly6C+ granulocytes, Proc Natl Acad Sci USA 7 (2010), 21248–21255.