Meta-Analysis
Copyright ©The Author(s) 2015. Published by Baishideng Publishing Group Inc. All rights reserved.
World J Gastroenterol. Jul 7, 2015; 21(25): 7895-7906
Published online Jul 7, 2015. doi: 10.3748/wjg.v21.i25.7895
Thrombocytopenia for prediction of hepatocellular carcinoma recurrence: Systematic review and meta-analysis
Qing Pang, Kai Qu, Jian-Bin Bi, Su-Shun Liu, Jing-Yao Zhang, Si-Dong Song, Ting Lin, Xin-Sen Xu, Yong Wan, Ming-Hui Tai, Hao-Chen Liu, Ya-Feng Dong, Chang Liu
Qing Pang, Kai Qu, Jian-Bin Bi, Su-Shun Liu, Jing-Yao Zhang, Si-Dong Song, Ting Lin, Xin-Sen Xu, Yong Wan, Ming-Hui Tai, Hao-Chen Liu, Chang Liu, Department of Hepatobiliary Surgery, the First Affiliated Hospital of Medical College, Xi’an Jiaotong University, Xi’an 710061, Shaanxi Province, China
Ya-Feng Dong, Department of Obstetrics and Gynecology, University of Kansas School of Medicine, Kansas City, KS 66045, United States
Author contributions: Pang Q and Qu K contributed equally to this work; Pang Q, Dong YF and Liu C designed the study; Pang Q and Zhang JY performed the literature search; Pang Q, Qu K, Song SD extracted the data; Lin T, Xu XS and Tai MH contributed to quality evaluation for each study and drew the forest plots; Pang Q, Qu K and Zhang JY wrote the paper; and Bi JB, Liu SS, Lin T, Xu XS, Wan Y and Liu HC revised the paper.
Supported by National Natural Science Foundation of China, No. 81272644 and No. 81072051.
Conflict-of-interest statement: No potential conflicts of interest relevant to this article were reported.
Data sharing statement: Technical appendix, statistical code, and dataset available from the corresponding author at email: liuchangdoctor@163.com. Participants gave informed consent for data sharing. No additional data are available.
Open-Access: This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Correspondence to: Chang Liu, Professor, MD, PhD, Department of Hepatobiliary Surgery, the First Affiliated Hospital of Medical College, Xi’an Jiaotong University, No. 277 West Road, Yanta district, Xi’an 710061, Shaanxi Province, China. liuchangdoctor@163.com
Telephone: +86-29-82653900 Fax: +86-29-82653905
Received: January 5, 2015
Peer-review started: January 6, 2015
First decision: February 13, 2015
Revised: March 1, 2015
Accepted: April 3, 2015
Article in press: April 3, 2015
Published online: July 7, 2015

Abstract

AIM: To investigate the association between thrombocytopenia and relapse after treatment for hepatocellular carcinoma (HCC).

METHODS: We searched the PubMed, EMBASE, and Web of Science databases to obtain eligible studies. The hazard ratios (HRs) values and 95% confidence intervals (CIs) were pooled by random effects model. Subsequently, we estimated the heterogeneity, performed a sensitivity analysis, determined the publication bias, and performed subgroup and meta-regression analyses. Study quality was assessed by using the Oxford Center for Evidence Based Medicine tool.

RESULTS: We identified 18 eligible studies by retrieval (published during 2000-2014). Out of the 4163 patients with HCC who were recruited, 2746 (66.0%) experienced recurrence. In general, our meta-analysis suggested that low platelet count (PLT) before therapy significantly increased the probability of postoperative recurrence (HR = 1.53, 95%CI: 1.29-1.81). PLT was also valuable in the prediction of intrahepatic distant recurrence (HR = 1.49, 95%CI: 1.25-1.77). Subgroup and meta-regression analyses identified various therapeutic modalities as the source of a high degree of heterogeneity. The pooled HR values showed no obvious change when a single study was removed, but otherwise, an opposite-effects model was used. In addition, no significant publication bias was detected.

CONCLUSION: Thrombocytopenia before treatment might be an inexpensive and useful predictor of postoperative recurrence in patients with HCC.

Key Words: Hepatocellular carcinoma, Blood platelets, Thrombocytopenia, Recurrence, Prognosis

Core tip: The probability of high postoperative recurrence is the greatest problem that affects the potential curative treatment of patients with hepatocellular carcinoma (HCC). No factors have been widely considered as useful predictors for postoperative recurrence in HCC. We analyzed 18 relevant studies to determine the significance of platelet count. We included 4163 patients with HCC and found that thrombocytopenia before therapy significantly increased the probability of postoperative recurrence as well as intrahepatic distant recurrence. We emphasize that thrombocytopenia is an inexpensive, helpful predictor of postoperative recurrence in patients with HCC.



INTRODUCTION

Hepatocellular carcinoma (HCC) is a common malignancy with an increasing incidence worldwide. Chronic infections with several types of hepatitis viruses, especially hepatitis B virus (HBV) and hepatitis C virus (HCV), are well-recognized risk factors for HCC[1]. Between these two, HBV infection is a major etiological factor, and thus, the majority of patients with HCC reside in developing countries where HBV is prevalent[2].

With the advancement in techniques, substantial progress has been made in the diagnosis and treatment of HCC over the past few decades. Current therapies primarily consist of liver resection, transplantation, radiofrequency ablation (RFA), and transcatheter arterial embolization (TACE). However, despite these treatments, the overall 5-year survival rate is 5%-6%, and therefore, the prognosis of HCC is unsatisfactory[3]. Many factors are related to the prognosis of HCC[4,5]. Among these factors, high probability of postoperative recurrence is the greatest issue that affects the potential curative treatments for HCC[5].

Platelet function in thrombosis, the inflammatory response, and liver regeneration via the release of several inflammatory mediators such as serotonin, platelet derived growth factor (PDGF), and transforming growth factor (TGF)-β[6,7]. Abnormalities in platelets, either quantitative or functional, would result in a series of pathological changes and would subsequently lead to some disorders. Platelet count (PLT) and/or several platelet-based noninvasive models have been validated as valuable indices for the detection of liver cirrhosis[8-10]; a disorder that is strongly associated with the incidence and outcome of HCC. Moreover, thrombocytopenia has been proposed as a useful tool to identify carcinogenesis in basic liver diseases[11,12], as well as to predict postoperative morbidity and mortality of patients with HCC[13]. Furthermore, our previous study and other reports have suggested that thrombocytopenia is related to poor survival of patients with liver cancer[14-16].

However, whether or not preoperative thrombocytopenia might increase the risk of relapse in patients with HCC is unclear and controversial. Some studies found that thrombocytopenia was significantly associated with HCC recurrence, whereas others failed to show a significant association between thrombocytopenia and HCC recurrence. However, we previously suggested that a higher preoperative PLT might lead to a higher postoperative recurrence rate[16]. Here, we further summarize all of the relative articles in order to clarify this issue.

MATERIALS AND METHODS
Search strategy and selection criteria

A systematic search in the PubMed, EMBASE, and ISI Web of Science databases for studies published until 31 August 2014 was performed by two independent investigators (Pang Q and Zhang JY). Our core search consisted of the terms (PLT or platelet or platelets or thrombopenia or thrombocytopenia or thrombocytosis) and (recurrence or relapse) and (hepatocellular carcinoma or HCC or liver cancer or hepatic carcinoma or hepatic cancer). In addition, we manually retrieved the reference lists of relevant reviews and included those studies as well.

We included studies that met the following predetermined criteria: (1) published as an original article; (2) HCC was determined by pathology/imaging; (3) studied the relationship between PLT and HCC recurrence and reported the hazard ratios (HRs) and 95% confidence intervals (CIs), or provided sufficient data to calculate them; (4) PLT was expressed as a binary variable (with a lower or a higher category as a reference); and (5) recruited no fewer than 20 patients. The exclusion criteria were: (1) studies of secondary liver cancer or other liver diseases; (2) HCC was assessed by serum markers; (3) studies of the effect of PLT on survival but not on recurrence; (4) studies that only provided the P value or other conditions that could not be used to calculate HR and 95%CI values; (5) PLT was expressed as a continuous variable; and (6) conference abstracts or letters. If two or more publications recruited identical populations, we only included the one with the largest number of cases or the one with the most adjusted HR values.

Data abstraction

Based on the above selection criteria, two researchers (Pang Q and Qu K) independently evaluated the retrieved studies. The κ statistic was calculated to assess the variability between the observers, and all discrepancies were resolved by discussion. For each included study, we abstracted the following data with a standardized data-collection protocol: first author, publication year, country where the study population lived, gender distribution, method of treatment, total cases, recurrent cases, duration of follow-up, cut-off value of PLT, and adjusted (prior use) or crude HR value and 95%CI. The quality of each study was appraised using the Oxford Center for Evidence-based Medicine appraisal approach[17]. We followed the MOOSE (Meta-analysis of Observational Studies in Epidemiology) guidelines[18] for the reporting of this meta-analysis. All data were verified by one author (Pang Q).

Statistical analysis

If a study considered a lower PLT as a reference, we converted its HR value to an estimated value with a higher category as a reference (which reflected the influence of a low PLT on recurrence). We assessed the heterogeneity among the studies with the Q value and the I2 statistic value (25%, 50%, and 75% correspond to the cut-off points for low, moderate, and high degrees of heterogeneity, respectively). Heterogeneity was considered to be statistically significant if P was < 0.10 for Q statistic, or I2 was > 50%; otherwise, no significant heterogeneity was observed. We calculated the pooled HR value for recurrence (or distant recurrence) by random-effects model. If high heterogeneity was found, we explored the potential source of heterogeneity by performing a subgroup analysis and meta-regression. Subsequently, we performed an influence analysis to evaluate whether any single study could markedly affect the result. Moreover, we compared the HR values that were pooled by a fixed-effect model and that were summed by a random-effects model. Eventually, publication bias was examined by funnel plots with Begg’s and Egger’s tests. The statistical methods of this study were reviewed by Dr. Kai Qu from the Department of Epidemiology, MD Anderson Cancer Center, University of Texas, and Prof. Ya-Feng Dong from the University of Kansas School of Medicine, United States.

We used STATA 12.0 software to analyze the data. A bilateral P value < 0.05 was indicated a significant difference.

RESULTS

The flow chart that details our literature search and selection process is shown in Figure 1. Of the total 964 citations, we finally included 18 studies. There was good agreement between the two observers on which studies to include (κ: 0.940). No additional studies were found within the references of reviews and included studies. All 18 studies were published in English.

Figure 1
Figure 1 Flow chart of the search strategy and study selection.
Characteristics of the included studies

The baseline values of the included studies in this meta-analysis are summarized in Table 1. Our meta-analysis consisted of 4163 patients, of which 2784 (66.9%) were men. During the mean or median follow-up period, which ranged from 12 to 151.2 mo, 2746 (66.0%) patients experienced recurrence. Sixteen studies adopted a cut-off value of 100 × 109/L for PLT while the remaining two[19,20] considered 150 × 109/L as the cut-off point. There was one individual inception cohort study with > 80% follow-up and this was assessed as level 1b using the Oxford Center for Evidence-based Medicine tool. Others were retrospective cohort studies and were identified as either level 2a/2b or 4. All studies were performed in Asia, and the effect of PLT on intrahepatic distant recurrence was estimated in six of them.

Table 1 Baseline characteristics of the studies included in the meta-analysis.
Ref.YearCountryCutoff(109/L)TherapyAdjusted HRRecurrence typeMenWomenRecurrenceFollow-up time (mo)LoE
Saito et al[32]2014Japan100PH + RFANoOverall38133939.9-46.64
Kaibori et al[19]2013Japan150PHYesOverall601414151.22a
Shiina et al[24]2012Japan100RFAYesDistant75141974038.22b
Nishikawa et al[25]2012Japan100RFAYesDistant114688636.42b
Kao et al[26]2012TaiWan100RFAYesOverall1629616328.52b
Miyatake et al[33]2012Japan100PH + PTAYesOverall26013527742-462b
Ikeda et al[20]2011Japan150RFANoOverall683955124
Amano et al[21]2011Japan100PHNoOverall1272410749.24
Ishikawa et al[31]2011Japan100TACEYesDistant47314636.61b
Hagihara et al[34]2011Japan100PH + PTANoOverall124588144.44
Nonaka et al[22]2010Japan100PHNoOverall422232124
Kang et al[23]2009Korea100PHNoOverall1254278384
Fuke et al[28]2008Japan100RFAYesDistant80379029.12b
Nouso et al[27]2008Japan100RFANoOverall + distant41520658934.74
Jeong et al[35]2007Japan100PH + medicalYesOverall651950322b
Tateishi et al[29]2006Japan100RFANoOverall276140277NR4
Yamanaka et al[30]2005Japan100RFAYesDistant17914182b
Ikeda et al[36]2000Japan100PH + RFANoOverall1378254
Pooled HR values for all of the studies

After calculation of the summed effects, we demonstrated that low PLT before treatment significantly increased the risk of HCC recurrence (HR = 1.53, 95%CI: 1.29-1.81). The forest plot is shown in Figure 2. We observed a moderate degree of heterogeneity between studies (I2 = 50.8%, P = 0.007).

Figure 2
Figure 2 Effect of platelet count on hepatocellular carcinoma recurrence.
Impact of thrombocytopenia on recurrence

Subsequently, we summed the 17 studies, which used a cutoff value of × 100 109/L for PLT and consisted of 4035 patients, in order to investigate the influence of thrombocytopenia on recurrence after treatment. Thrombocytopenia was found to be a useful tool for the prediction of recurrence in HCC (HR = 1.42, 95%CI: 1.27-1.60) (Table 2).

Table 2 Effects of platelet count in different studies using the two effects model.
Random-effects modelFixed-effect model
HR (95%CI)HR (95%CI)
All studies1.53 (1.29-1.81)1.41 (1.28-1.55)
Thrombocytopenia1.42 (1.27-1.60)1.39 (1.26-1.54)
Liver resection4.46 (1.57-12.65)3.48 (2.07-5.83)
RFA1.43 (1.24-1.65)1.39 (1.24-1.56)
Distant recurrence1.49 (1.25-1.77)1.45 (1.26-1.67)
HCV-HCC1.33 (1.12-1.60)1.33 (1.12-1.60)
Pooled HR value for patients who underwent liver resection, RFA

Different treatment modalities, which are given according to the tumor characteristics and stage, may influence the survival and risk of recurrence. Among these therapies, surgical resection is the primary curative treatment modality for HCC. Subsequently, we explored the impact of PLT on patients who received partial hepatectomy. Overall, 231 of the 470 (49.1%) patients who underwent resection experienced recurrence. By pooling the four eligible studies[19,21-23], we demonstrated a significant association between the preoperative PLT level and postoperative recurrence (HR = 4.46, 95%CI: 1.57-12.65), with a high degree of heterogeneity (Figure 3A).

Figure 3
Figure 3 Effect of platelet count on hepatocellular carcinoma recurrence in patients who underwent liver resection (A) and RFA (B), on distant recurrence (C), and on recurrence in patients with hepatitis C virus-related hepatocellular carcinoma (D).

Eight studies[20,24-30] recruited patients with HCC who underwent RFA. Of these included patients, 2014 of the 2897 (69.5%, about 1.42 times the number of patients who underwent resection) experienced recurrence during follow-up. Figure 3B indicates that PLT is still a useful indicator for the prediction of recurrence in patients who received RFA. There was no significant heterogeneity and the pooled HR value was 1.43 (95%CI: 1.24-1.65).

Effect of PLT on distant recurrence

Recurrence is divided into local and intrahepatic distant recurrence. Only one study[27] in our meta-analysis estimated the association between thrombocytopenia and local recurrence. In contrast, six studies[24,25,27,28,30,31], which consisted of 2193 patients (1541 had a relapse), reported the effect of PLT on distant recurrence. A cutoff point of 100 × 109/L for PLT was used in all six studies. This meta-analysis showed that thrombocytopenia was also a significant indictor for the prediction of distant recurrence (HR = 1.49, 95%CI: 1.25-1.77) (Figure 3C). Little heterogeneity (I2 = 15.8%, P = 0.312) was found among these studies.

Effect of PLT on recurrence of HCV-related HCC

Etiology is another major factor that is associated with the prognosis of HCC. Most of our included studies calculated the HR value without a distinction among the various etiologies. Seven studies[28,30,32-36] only recruited patients with HCV-related HCC. Similarly, thrombocytopenia also had significant potential in the prediction of relapse in these patients (HR = 1.33, 95%CI: 1.12-1.60), and no significant between-study heterogeneity was observed (I2 = 0%, P = 0.651) (Figure 3D).

Exploration of heterogeneity

To explore the source of heterogeneity, we performed subgroup and meta-regression analyses. We analyzed the covariates that might be responsible for the potential heterogeneity. We had at least two studies in each subgroup and included treatment method; follow-up time (≥ 3 or < 3 years); whether adjustment for confounding factors was performed; type of recurrence (overall or distant recurrence); and the number of recruited patients. The results of our exploration are shown in Table 3. We suggested that the treatment method might affect the pooled effect size (P < 0.05 in subgroup and meta-regression analysis). All of the subgroups showed a significant association between PLT and recurrence except the one subgroup whose studies recruited patients who received palliative therapies[31,35] (TACE or medical therapy, HR = 1.43, 95%CI: 0.96-2.11). For all five covariates, no significant between-group differences (all P > 0.05) were found when a multivariable meta-regression was performed.

Table 3 Subgroup and meta-regression analyses.
CovariatesSubgroupNo.HR (95% CI)Heterogeneity
Meta-regression
PaI2PbCrude PAdjusted P
TreatmentResection44.46 (1.57-12.65)0.01272.40.0050.0480.212
RFA81.43 (1.24-1.65)0.28318.6
Resection+RFA41.26 (1.03-1.55)0.5320
Others21.43 (0.96-2.11)0.9160
Follow-up≥ 3691.71 (1.25-2.35)0.00073.40.4540.5960.989
(mo)< 3691.35 (1.17-2.56)0.8590
ConfounderAdjusted91.70 (1.30-2.22)0.00464.10.1980.3450.863
Unadjusted91.37 (1.12-1.68)0.22524.6
Recurrence typeOverall131.51 (1.21-1.87)0.00557.60.3510.7200.999
Distant51.64 (1.25-2.14)0.24726.1
Recruited≥ 15091.49 (1.26-1.69)0.10339.70.6000.4430.944
No.< 15091.78 (1.10-2.88)0.00761.8
Sensitivity analysis and determination of publication bias

Furthermore, a sensitivity analysis was conducted to validate the certainty of our findings. When the opposite-effects model was used, the results did not significantly change (Table 2). Then, we performed an influence analysis and found that no single study affected the summary estimation (Figure 4).

Figure 4
Figure 4 Influence analysis of our included studies.

Eventually, we constructed a funnel plot to detect the existence of publication bias, and the figure reflects the basic symmetry (Figure 5). Indeed, no significant evidence of publication bias was found (P values were 0.11 and 0.36 in Begg’s test and Egger’s test, respectively).

Figure 5
Figure 5 Begg’s (A) and Egger’s (B) plots of our included studies.
DISCUSSION

A high risk of recurrence is considered one of the greatest concerns with regard to the treatment of patients with HCC[5]. The postoperative recurrence rate was reported to range from 50% to 100%[19]. To obtain a satisfactory prognosis in cases of HCC, it is crucial to determine the predisposing factors for recurrence and improve these factors before treatment. Platelets are associated with the prognosis of various solid tumors, including HCC[37]. However, the significance of platelets in the risk of recurrence of HCC remains unknown. This is the first time that an estimation by a quantitative summary of all relative studies has been performed. Our meta-analysis showed that patients with a low PLT before treatment had a significantly increased risk of postoperative relapse. In general, thrombocytopenia could increase the risk of recurrence by 53%. Specifically, thrombocytopenia was demonstrated to be a useful tool for the prediction of recurrence, no matter whether hepatic resection or RFA was performed for HCC.

Recurrence comprises both local and intrahepatic distant recurrence. The factors that influence the two types of recurrence are different[27,28]. It is essential to identify the risk factors separately for the two types in order to prevent recurrence. However, few studies have differentiated local from intrahepatic distant recurrence. Only one study exclusively reported the local recurrence, and six studies exclusively estimated the distant recurrence in our meta-analysis. By pooling the latter studies, we found that thrombocytopenia was also significantly associated with distant recurrence.

Our results are important in that they provide crucial guidance for the estimation of the prognosis of patients with HCC. No significant difference was found between our use of the random-effects and fixed-effect models. The pooled HR values were not markedly affected by a single study, which indicated the robustness of our results. By subgroup and meta-regression analyses, we demonstrated that the various treatment modalities might be a source of heterogeneity.

This study did have some limitations. First, gender, age, Child-Pugh grade, and several other parameters that are important prognostic factors for HCC were not adjusted in this meta-analysis and several included studies were level 4. Whether or not the significance of PLT is independent of these confounders remains unknown. However, after pooling the adjusted HR values (Table 3), we suggest that platelets might be independently associated with postoperative relapse. Second, pathological examination is universally viewed as the gold standard for the diagnosis of HCC. However, to obtain more data, we did not exclude the studies that assessed HCC by imaging. Third, thrombocytopenia is strongly associated with the progression of liver cirrhosis and is considered an independent risk factor for multicentric HCC[38]. In other words, thrombocytopenia is only a surrogate marker or a confounding factor of multicentric carcinogenesis associated with liver cirrhosis. It is therefore necessary to evaluate the relationship between thrombocytopenia and recurrence of HCC in the cases with a corresponding amount of liver cirrhosis. However, because all the retrieved studies failed to stratify the patients with HCC according to the presence or degree of liver cirrhosis, we could not explain whether thrombocytopenia is independent of liver cirrhosis. Nevertheless, based on our previous report, in patients with HCC without cirrhosis, but not in those with cirrhosis, PLT was significantly associated with recurrence[16]. In another study, a lower PLT was also associated with a significantly worse prognosis in patients with HCC and cirrhosis[4].

Due to portal hypertension, a decrease in thrombopoietin (TPO) production in the liver, and the capture of platelets by the liver[6], PLT tends to decrease in various liver diseases. Numerous studies have shown that thrombocytopenia is a major risk factor for the degree of cirrhosis, development of cirrhosis, as well as for carcinogenesis in patients with chronic hepatitis[8,39,40]. However, the exact mechanism of the effects of PLT in the prognosis of HCC is still unclear. Several clinical studies have demonstrated that thrombocytopenia could increase postoperative complications and morbidity, and could lead to the deterioration of liver function[41-43]. With the analysis of 202 patients with HCV-related HCC who underwent hepatectomy, Kubo et al[38] emphasized that PLT was the only independent predictor for multicentric HCC; in addition, PLT was significantly associated with the severity of active hepatitis and hepatic fibrosis (both P < 0.05). A previous study also indicated that PLT was a useful predictor of portal vein invasion[44]. Additionally, a decreased PLT level was found to be significantly associated with an elevated α-fetoprotein level[12,45,46]. Furthermore, patients with a lower preoperative PLT showed a significantly higher probability of cirrhosis, a higher level of bilirubin, a greater amount of bleeding, and a high level of indocyanine green retention at 15 min[23]. The above-mentioned factors were all related to recurrence in HCC[27,29], and thus, these findings may help hepatologists explore the exact mechanism. By both in vitro and in vivo studies, Nozaki et al[47] proved that TPO could promote liver regeneration and improve cirrhosis via an increase in platelets. PDGF, which is mainly stored in platelets, is associated with tumor progression and prognosis in HCC[48]. These findings support the relationships between PLT and HCC recurrence. In addition to its influence on recurrence, our previous research and several other studies also showed that thrombocytopenia could result in poor survival in patients with HCC[4,15,34]. However, data on the relationship between PLT and tumor stage or other crucial prognostic factors for HCC are lacking.

Additionally, thrombocytosis was found to be associated with the incidence of HCC as well[49] and could lead to an increase in the risk of death from HCC[37] and other cancers[50]. A positive correlation has been suggested between serum PLT and tumor size[49,51]. In in vitro studies, platelets are a stimulating factor for the growth and invasion of several HCC cell lines[52], which further indicates the adverse effects of excessive platelets.

Therefore, it is believed that a huge gap in the cut-off values of PLT (such as 400 × 109/L vs 100 × 109/L) may show opposite effects in the prognosis of HCC. Although thrombocytopenia and thrombocytosis are not contraindications for resection in patients with HCC[4], it is still recommended to normalize the serum platelet level by prophylactic platelet transfusions or by taking agents before treatment. On the contrary, although sorafenib, interferon, or other chemotherapeutic/anticancer drugs might be effective in the delay and/or management of postoperative recurrence, they could lead to many complications such as thrombocytopenia[32,33,53,54]. Therefore, for patients who are taking these drugs, a regular assessment of the platelet level in order to maintain the level within a normal range, is pivotal to obtain a favorable outcome.

In conclusion, our meta-analysis suggested that thrombocytopenia was a valuable, inexpensive predictor for recurrence in patients with HCC. Further experimental and clinical studies are needed to validate the results and to clarify the exact mechanism.

ACKNOWLEDGMENTS

We thank Dr. Wei Chen, Yan-Yan Zhou and Run-Chen Miao in the Department of Hepatobiliary Surgery, The First Affiliated Hospital of Medical College, Xi’an Jiaotong University for their support in this meta-analysis.

COMMENTS
Background

Hepatocellular carcinoma (HCC) is a common malignancy with an increasing incidence worldwide. No matter which treatments are given, the prognosis of HCC is poor. Many factors are related to the prognosis of HCC. Among these factors, a high postoperative recurrence rate is the greatest problem that influences potential curative treatments for HCC. It is crucial to seek several key risk factors that influence postoperative recurrence in HCC.

Research frontiers

Thrombocytopenia is significantly associated with poor survival in HCC. However, whether or not a preoperative low platelet count could increase the risk of recurrence in patients with HCC remains uncertain. The authors endeavored to validate this issue in a meta-analysis.

Innovations and breakthroughs

Several HCC prognostic models, which mainly focus on the tumor characteristics, have been proposed to assess postoperative survival. However, these models played a limited role in the prediction of recurrence. In contrast, the authors found that thrombocytopenia, a simple and inexpensive index, was significantly associated with a high probability of recurrence, including distant recurrence, in patients with HCC, no matter what treatments were given.

Applications

Through a validation of this novel prognostic predictive parameter, this study may represent a future strategy for cancer prediction in the follow-up of patients with HCC.

Terminology

Platelets are involved in thrombosis, inflammatory response, and liver regeneration via the release of several inflammatory mediators such as serotonin. The platelet count is a crucial reflection of platelet function, with a normal range of 100× 109 to 300 × 109/L. Thrombocytopenia occurs when the platelet count is < 100 × 109/L.

Peer-review

This is a prognostic meta-analysis and the manuscript is very well written.

Footnotes

P- Reviewer: Bashashati M, Sadeghi R S- Editor: Qi Y L- Editor: Kerr C E- Editor: Liu XM

References
1.  Forner A, Llovet JM, Bruix J. Hepatocellular carcinoma. Lancet. 2012;379:1245-1255.  [PubMed]  [DOI]
2.  Kew MC. Epidemiology of chronic hepatitis B virus infection, hepatocellular carcinoma, and hepatitis B virus-induced hepatocellular carcinoma. Pathol Biol (Paris). 2010;58:273-277.  [PubMed]  [DOI]
3.  Buonaguro L, Petrizzo A, Tagliamonte M, Tornesello ML, Buonaguro FM. Challenges in cancer vaccine development for hepatocellular carcinoma. J Hepatol. 2013;59:897-903.  [PubMed]  [DOI]
4.  Roayaie S, Obeidat K, Sposito C, Mariani L, Bhoori S, Pellegrinelli A, Labow D, Llovet JM, Schwartz M, Mazzaferro V. Resection of hepatocellular cancer ≤2 cm: results from two Western centers. Hepatology. 2013;57:1426-1435.  [PubMed]  [DOI]
5.  Bruix J, Gores GJ, Mazzaferro V. Hepatocellular carcinoma: clinical frontiers and perspectives. Gut. 2014;63:844-855.  [PubMed]  [DOI]
6.  Kondo R, Yano H, Nakashima O, Tanikawa K, Nomura Y, Kage M. Accumulation of platelets in the liver may be an important contributory factor to thrombocytopenia and liver fibrosis in chronic hepatitis C. J Gastroenterol. 2013;48:526-534.  [PubMed]  [DOI]
7.  Starlinger P, Assinger A, Haegele S, Wanek D, Zikeli S, Schauer D, Birner P, Fleischmann E, Gruenberger B, Brostjan C. Evidence for serotonin as a relevant inducer of liver regeneration after liver resection in humans. Hepatology. 2014;60:257-266.  [PubMed]  [DOI]
8.  Udell JA, Wang CS, Tinmouth J, FitzGerald JM, Ayas NT, Simel DL, Schulzer M, Mak E, Yoshida EM. Does this patient with liver disease have cirrhosis? JAMA. 2012;307:832-842.  [PubMed]  [DOI]
9.  Shaheen AA, Myers RP. Diagnostic accuracy of the aspartate aminotransferase-to-platelet ratio index for the prediction of hepatitis C-related fibrosis: a systematic review. Hepatology. 2007;46:912-921.  [PubMed]  [DOI]
10.  Kessoku T, Ogawa Y, Yoneda M, Imajo K, Sumida Y, Eguchi Y, Fujii H, Hyogo H, Ono M, Suzuki Y. Simple scoring system for predicting cirrhosis in nonalcoholic fatty liver disease. World J Gastroenterol. 2014;20:10108-10114.  [PubMed]  [DOI]
11.  Wang JH, Chang KC, Kee KM, Chen PF, Yen YH, Tseng PL, Kuo YH, Tsai MC, Hung CH, Chen CH. Hepatocellular carcinoma surveillance at 4- vs. 12-month intervals for patients with chronic viral hepatitis: a randomized study in community. Am J Gastroenterol. 2013;108:416-424.  [PubMed]  [DOI]
12.  Akuta N, Suzuki F, Seko Y, Kawamura Y, Sezaki H, Suzuki Y, Hosaka T, Kobayashi M, Hara T, Kobayashi M. Complicated relationships of amino acid substitution in hepatitis C virus core region and IL28B genotype influencing hepatocarcinogenesis. Hepatology. 2012;56:2134-2141.  [PubMed]  [DOI]
13.  Yang T, Zhang J, Lu JH, Yang GS, Wu MC, Yu WF. Risk factors influencing postoperative outcomes of major hepatic resection of hepatocellular carcinoma for patients with underlying liver diseases. World J Surg. 2011;35:2073-2082.  [PubMed]  [DOI]
14.  Pang Q, Zhang JY, Song SD, Qu K, Xu XS, Liu SS, Tai MH, Liu HC, Liu C. Thrombocytopenia as an inexpensive, valuable predictor for survival in patients with hepatocellular carcinoma. Scand J Gastroenterol. 2014;49:1507-1508.  [PubMed]  [DOI]
15.  Shen SL, Fu SJ, Chen B, Kuang M, Li SQ, Hua YP, Liang LJ, Guo P, Hao Y, Peng BG. Preoperative aspartate aminotransferase to platelet ratio is an independent prognostic factor for hepatitis B-induced hepatocellular carcinoma after hepatic resection. Ann Surg Oncol. 2014;21:3802-3809.  [PubMed]  [DOI]
16.  Pang Q, Zhang JY, Xu XS, Song SD, Qu K, Chen W, Zhou YY, Miao RC, Liu SS, Dong YF. Significance of platelet count and platelet-based models for hepatocellular carcinoma recurrence. World J Gastroenterol. 2015;21; 5607-5621.  [PubMed]  [DOI]
17.  Oxford Centre for Evidence-based Medicine. Critical appraisal. Accessed 2015 Feb 28.  Available from: http://www.cebm.net.  [PubMed]  [DOI]
18.  Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, Moher D, Becker BJ, Sipe TA, Thacker SB. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA. 2000;283:2008-2012.  [PubMed]  [DOI]
19.  Kaibori M, Kubo S, Nagano H, Hayashi M, Haji S, Nakai T, Ishizaki M, Matsui K, Uenishi T, Takemura S. Clinicopathological features of recurrence in patients after 10-year disease-free survival following curative hepatic resection of hepatocellular carcinoma. World J Surg. 2013;37:820-828.  [PubMed]  [DOI]
20.  Ikeda H, Tateishi R, Enooku K, Yoshida H, Nakagawa H, Masuzaki R, Kondo Y, Goto T, Shiina S, Kume Y. Prediction of hepatocellular carcinoma development by plasma ADAMTS13 in chronic hepatitis B and C. Cancer Epidemiol Biomarkers Prev. 2011;20:2204-2211.  [PubMed]  [DOI]
21.  Amano H, Tashiro H, Oshita A, Kobayashi T, Tanimoto Y, Kuroda S, Tazawa H, Itamoto T, Asahara T, Ohdan H. Significance of platelet count in the outcomes of hepatectomized patients with hepatocellular carcinoma exceeding the Milan criteria. J Gastrointest Surg. 2011;15:1173-1181.  [PubMed]  [DOI]
22.  Nonaka K, Fujioka H, Takii Y, Abiru S, Migita K, Ito M, Kanematsu T, Ishibashi H. mPGES-1 expression in non-cancerous liver tissue impacts on postoperative recurrence of HCC. World J Gastroenterol. 2010;16:4846-4853.  [PubMed]  [DOI]
23.  Kang CM, Choi GH, Kim DH, Choi SB, Kim KS, Choi JS, Lee WJ. Revisiting the role of nonanatomic resection of small (& lt; or = 4 cm) and single hepatocellular carcinoma in patients with well-preserved liver function. J Surg Res. 2010;160:81-89.  [PubMed]  [DOI]
24.  Shiina S, Tateishi R, Arano T, Uchino K, Enooku K, Nakagawa H, Asaoka Y, Sato T, Masuzaki R, Kondo Y. Radiofrequency ablation for hepatocellular carcinoma: 10-year outcome and prognostic factors. Am J Gastroenterol. 2012;107:569-577; quiz 578.  [PubMed]  [DOI]
25.  Nishikawa H, Osaki Y, Kita R, Kimura T, Inuzuka T, Takeda H, Nakajima J, Matsuda F, Sakamoto A, Henmi S. Transcatheter arterial infusion chemotherapy prior to radiofrequency thermal ablation for single hepatocellular carcinoma reduces the risk of intrahepatic distant recurrence. Int J Oncol. 2012;41:903-909.  [PubMed]  [DOI]
26.  Kao WY, Chiou YY, Hung HH, Su CW, Chou YH, Huo TI, Huang YH, Wu WC, Lin HC, Lee SD. Younger hepatocellular carcinoma patients have better prognosis after percutaneous radiofrequency ablation therapy. J Clin Gastroenterol. 2012;46:62-70.  [PubMed]  [DOI]
27.  Nouso K, Matsumoto E, Kobayashi Y, Nakamura S, Tanaka H, Osawa T, Ikeda H, Araki Y, Sakaguchi K, Shiratori Y. Risk factors for local and distant recurrence of hepatocellular carcinomas after local ablation therapies. J Gastroenterol Hepatol. 2008;23:453-458.  [PubMed]  [DOI]
28.  Fuke H, Sugimoto K, Shiraki K, Tanaka J, Beppu T, Yoneda K, Yamamoto N, Ito K, Takaki H, Nakatsuka A. Predictive factors for distant recurrence of HCV-related hepatocellular carcinoma after radiofrequency ablation combined with chemoembolization. Aliment Pharmacol Ther. 2008;27:1253-1260.  [PubMed]  [DOI]
29.  Tateishi R, Shiina S, Yoshida H, Teratani T, Obi S, Yamashiki N, Yoshida H, Akamatsu M, Kawabe T, Omata M. Prediction of recurrence of hepatocellular carcinoma after curative ablation using three tumor markers. Hepatology. 2006;44:1518-1527.  [PubMed]  [DOI]
30.  Yamanaka Y, Shiraki K, Miyashita K, Inoue T, Kawakita T, Yamaguchi Y, Saitou Y, Yamamoto N, Nakano T, Nakatsuka A. Risk factors for the recurrence of hepatocellular carcinoma after radiofrequency ablation of hepatocellular carcinoma in patients with hepatitis C. World J Gastroenterol. 2005;11:2174-2178.  [PubMed]  [DOI]
31.  Ishikawa T, Higuchi K, Kubota T, Seki K, Honma T, Yoshida T, Kamimura T. Prevention of intrahepatic distant recurrence by transcatheter arterial infusion chemotherapy with platinum agents for stage I/II hepatocellular carcinoma. Cancer. 2011;117:4018-4025.  [PubMed]  [DOI]
32.  Saito T, Chiba T, Suzuki E, Shinozaki M, Goto N, Kanogawa N, Motoyama T, Ogasawara S, Ooka Y, Tawada A. Effect of previous interferon-based therapy on recurrence after curative treatment of hepatitis C virus-related hepatocellular carcinoma. Int J Med Sci. 2014;11:707-712.  [PubMed]  [DOI]
33.  Miyatake H, Kobayashi Y, Iwasaki Y, Nakamura S, Ohnishi H, Kuwaki K, Toshimori J, Hagihara H, Nouso K, Yamamoto K. Effect of previous interferon treatment on outcome after curative treatment for hepatitis C virus-related hepatocellular carcinoma. Dig Dis Sci. 2012;57:1092-1101.  [PubMed]  [DOI]
34.  Hagihara H, Nouso K, Kobayashi Y, Iwasaki Y, Nakamura S, Kuwaki K, Toshimori J, Miyatake H, Ohnishi H, Shiraha H. Effect of pegylated interferon therapy on intrahepatic recurrence after curative treatment of hepatitis C virus-related hepatocellular carcinoma. Int J Clin Oncol. 2011;16:210-220.  [PubMed]  [DOI]
35.  Jeong SC, Aikata H, Katamura Y, Azakami T, Kawaoka T, Saneto H, Uka K, Mori N, Takaki S, Kodama H. Effects of a 24-week course of interferon-alpha therapy after curative treatment of hepatitis C virus-associated hepatocellular carcinoma. World J Gastroenterol. 2007;13:5343-5350.  [PubMed]  [DOI]
36.  Ikeda K, Arase Y, Saitoh S, Kobayashi M, Suzuki Y, Suzuki F, Tsubota A, Chayama K, Murashima N, Kumada H. Interferon beta prevents recurrence of hepatocellular carcinoma after complete resection or ablation of the primary tumor-A prospective randomized study of hepatitis C virus-related liver cancer. Hepatology. 2000;32:228-232.  [PubMed]  [DOI]
37.  Buergy D, Wenz F, Groden C, Brockmann MA. Tumor-platelet interaction in solid tumors. Int J Cancer. 2012;130:2747-2760.  [PubMed]  [DOI]
38.  Kubo S, Tanaka H, Shuto T, Takemura S, Yamamoto T, Uenishi T, Tanaka S, Ogawa M, Sakabe K, Yamazaki K. Correlation between low platelet count and multicentricity of hepatocellular carcinoma in patients with chronic hepatitis C. Hepatol Res. 2004;30:221-225.  [PubMed]  [DOI]
39.  Karasu Z, Tekin F, Ersoz G, Gunsar F, Batur Y, Ilter T, Akarca US. Liver fibrosis is associated with decreased peripheral platelet count in patients with chronic hepatitis B and C. Dig Dis Sci. 2007;52:1535-1539.  [PubMed]  [DOI]
40.  Carr BI, Guerra V. Hepatocellular carcinoma size: platelets, γ-glutamyl transpeptidase, and alkaline phosphatase. Oncology. 2013;85:153-159.  [PubMed]  [DOI]
41.  Taketomi A, Kitagawa D, Itoh S, Harimoto N, Yamashita Y, Gion T, Shirabe K, Shimada M, Maehara Y. Trends in morbidity and mortality after hepatic resection for hepatocellular carcinoma: an institute’s experience with 625 patients. J Am Coll Surg. 2007;204:580-587.  [PubMed]  [DOI]
42.  Maithel SK, Kneuertz PJ, Kooby DA, Scoggins CR, Weber SM, Martin RC, McMasters KM, Cho CS, Winslow ER, Wood WC. Importance of low preoperative platelet count in selecting patients for resection of hepatocellular carcinoma: a multi-institutional analysis. J Am Coll Surg. 2011;212:638-648; discussion 648-650.  [PubMed]  [DOI]
43.  Lee HS, Park SY, Kim SK, Kweon YO, Tak WY, Cho CM, Jeon SW, Jung MK, Park HG, Lee DW. Thrombocytopenia represents a risk for deterioration of liver function after radiofrequency ablation in patients with hepatocellular carcinoma. Clin Mol Hepatol. 2012;18:302-308.  [PubMed]  [DOI]
44.  Hagiwara S, Kudo M, Kawasaki T, Nagashima M, Minami Y, Chung H, Fukunaga T, Kitano M, Nakatani T. Prognostic factors for portal venous invasion in patients with hepatocellular carcinoma. J Gastroenterol. 2006;41:1214-1219.  [PubMed]  [DOI]
45.  Chen TM, Huang PT, Tsai MH, Lin LF, Liu CC, Ho KS, Siauw CP, Chao PL, Tung JN. Predictors of alpha-fetoprotein elevation in patients with chronic hepatitis C, but not hepatocellular carcinoma, and its normalization after pegylated interferon alfa 2a-ribavirin combination therapy. J Gastroenterol Hepatol. 2007;22:669-675.  [PubMed]  [DOI]
46.  Kobeisy MA, Morsy KH, Galal M, Sayed SK, Ashmawy MM, Mohammad FM. Clinical significance of elevated alpha-foetoprotein (AFP) in patients with chronic hepatitis C without hepatocellular carcinoma in upper EGYPT. Arab J Gastroenterol. 2012;13:49-53.  [PubMed]  [DOI]
47.  Nozaki R, Murata S, Nowatari T, Maruyama T, Ikeda N, Kawasaki T, Fukunaga K, Ohkohchi N. Effects of thrombopoietin on growth of hepatocellular carcinoma: Is thrombopoietin therapy for liver disease safe or not? Hepatol Res. 2013;43:610-620.  [PubMed]  [DOI]
48.  Stock P, Monga D, Tan X, Micsenyi A, Loizos N, Monga SP. Platelet-derived growth factor receptor-alpha: a novel therapeutic target in human hepatocellular cancer. Mol Cancer Ther. 2007;6:1932-1941.  [PubMed]  [DOI]
49.  Carr BI, Guerra V. Thrombocytosis and hepatocellular carcinoma. Dig Dis Sci. 2013;58:1790-1796.  [PubMed]  [DOI]
50.  Hu C, Chen R, Chen W, Pang W, Xue X, Zhu G, Shen X. Thrombocytosis is a significant indictor of hypercoagulability, prognosis and recurrence in gastric cancer. Exp Ther Med. 2014;8:125-132.  [PubMed]  [DOI]
51.  Sarpel U, Ayo D, Lobach I, Xu R, Newman E. Inverse relationship between cirrhosis and massive tumours in hepatocellular carcinoma. HPB (Oxford). 2012;14:741-745.  [PubMed]  [DOI]
52.  Carr BI, Cavallini A, D’Alessandro R, Refolo MG, Lippolis C, Mazzocca A, Messa C. Platelet extracts induce growth, migration and invasion in human hepatocellular carcinoma in vitro. BMC Cancer. 2014;14:43.  [PubMed]  [DOI]
53.  Waghray A, Balci B, El-Gazzaz G, Kim R, Pelley R, Narayanan Menon KV, Estfan B, Romero-Marrero C, Aucejo F. Safety and efficacy of sorafenib for the treatment of recurrent hepatocellular carcinoma after liver transplantation. Clin Transplant. 2013;27:555-561.  [PubMed]  [DOI]
54.  Sun HC, Tang ZY, Wang L, Qin LX, Ma ZC, Ye QH, Zhang BH, Qian YB, Wu ZQ, Fan J. Postoperative interferon alpha treatment postponed recurrence and improved overall survival in patients after curative resection of HBV-related hepatocellular carcinoma: a randomized clinical trial. J Cancer Res Clin Oncol. 2006;132:458-465.  [PubMed]  [DOI]