Date Published: February 13, 2018
Publisher: Public Library of Science
Author(s): Philip Rauber, Frank Lammert, Katharina Grotemeyer, Beate Appenrodt, Sheng-Nan Lu.
Thrombocytopenia occurs frequently in patients with cirrhosis. The immature platelet fraction (IPF%) is measured to differentiate the causes of thrombocytopenia. To date the relevance of thrombopoietin (TPO) in the context of cirrhosis is unknown. The aim of our study was to investigate the cause of thrombocytopenia in patients with liver cirrhosis by measuring IPF%, TPO and spleen size. In addition we examined the use of IPF% to evaluate the severity of cirrhosis and its complications.
Overall, we included 88 in-patients with cirrhosis in our study. The collected data comprises current health status, blood parameters, severity of cirrhosis evaluated by Child-Pugh score and MELD score, spleen diameter, ascites and esophageal varices. The IPF% was measured using an automatic hematology analyzer. TPO was measured with ELISA.
IPF% (p = 0.003) and spleen diameter (p = 0.001) were significantly higher in patients with thrombocytopenia. There was no significant difference in TPO between patients with and without thrombocytopenia. The mean values of IPF% varied significantly (p = 0.044) in Child-Pugh stages. IPF% was significantly (p = 0.005) elevated in patients with esophageal varices. Moreover, IPF% higher than 3.85% displayed sensitivity of 76.6% and specificity of 52.4% with an area under receiver operating curve characteristics of 0.669 for the presence of esophageal varices.
On closer examination of the three compartments known to have an influence on platelet count splenomegaly seems to be the major cause of thrombocytopenia in patients with cirrhosis according to current knowledge. Higher IPF% in patients with thrombocytopenia indicates peripheral consumption of platelets. The relation between spleen diameter and platelet count indicates the spleen to be the major place of platelets’ consumption. TPO did not differ between patients with and without thrombocytopenia. Furthermore, we cannot exclude an influence of impaired thrombopoietin synthesis on platelet counts. The association between IPF% and platelet count suggests that there is physiological regulation of platelets in patients with cirrhosis. In our study IPF% is associated with esophageal varices and the stage of cirrhosis. Further studies are needed to confirm these results.
Thrombocytopenia is a common complication in patients with liver cirrhosis and its pathogenesis is multifactorial. [1,2] Previous studies have reported the splenic sequestration of platelets to be a consequence of an enlarged spleen due to portal hypertension.  Furthermore recent studies propose that decreased TPO production in the liver and an impaired platelet production in the bone marrow are additional factors. [4,5] Increasing TPO levels and platelet recovery after orthotopic liver transplantation confirm these results as well as the availability of the TPO receptor agonist eltrombopag, which increases platelet counts in patients with liver cirrhosis caused by chronic hepatitis c virus infection. [6–9] In contrast, other studies have not been able to verify the associations between liver cirrhosis and decreased TPO production. [10–12] There are further factors that have an influence on platelet counts, such as anti platelet autoantibodies leading to shortened lifespan by removal via the reticuloendothelial system or chronic alcoholism, leading to the suppression of platelet production in the bone marrow. [13,14] Consequently the exact pathogenesis of thrombocytopenia in patients with liver cirrhosis is still unclear.
All data were obtained from participants that provided written informed consent. The following data were obtained: current health status, severity of cirrhosis evaluated by Child-Pugh score and Model for End-stage Liver Disease (MELD) score, spleen diameter, ascites and esophageal varices. The study was approved by the ethics committee of the Ärztekammer des Saarlandes (Ref. 271/11). In total, 88 in-patients with liver cirrhosis took part prospectively between March 2013 and February 2014 (30 females, 58 males; mean age 61±12 years). Exclusion criteria included: absence of liver cirrhosis, age < 18 years, acute bleeding, bone marrow disease or malignant disease (with the exception of hepatocellular carcinoma) with radio- or chemotherapy. Thrombocytopenia was defined as platelet count lower than 150.000/μl; subgroups were stratified into severe (lower than 50.000/μl), moderate (50.000–100.000/μl) and mild (100.000–150.000/μl). Etiology of cirrhosis, age and sex were recorded. Stage of liver cirrhosis, spleen diameter and ascites were assessed based on full history, clinical examination, abdominal ultrasound or magnetic resonance imaging. Esophageal varices were assessed by upper gastrointestinal endoscopy. Blood samples were collected from all patients for the following assays: complete blood count with platelet count and immature platelet fraction, thrombopoetin, albumin, bilirubin, international normalized ratio (INR), creatinin, cystatin C, c-reactive protein (crp), procalcitonin and n-terminal pro brain natriuretic peptide (nt-proBNP). Child-Pugh and MELD scores were calculated. The blood count assays were performed on a Sysmex XE-5000 hematology analyzer (SYSMEX CORPORATION, Kobe, Japan). The immature platelet fraction reference range is 1.1–6.1%.  Thrombopoietin measurement was performed with Quantikine Human TPO immunoassay (R&D Systems, Minneapolis, USA). The thrombopoietin reference range is 70-100ng/l.  Ascites was assessed by abdominal ultrasound and stratified into mild, moderate and large. Hepatorenal syndrome type I and II were defined according to criteria by the German S3 guideline by Gerbes et al. . Esophageal varices were defined according to criteria by Paquet . Thrombocytopenia is a common hematological abnormality that is either caused by increased sequestration and destruction of platelets or by decreased production. In patients with liver cirrhosis the pathogenesis is multifactorial but remains to be fully elucidated. Currently decreased platelet production in bone marrow and inadequate thrombopoietin production are purported to be the major causes. [5,35] In this context the immature platelet fraction (IPF%) might be a useful parameter for differential diagnosis.  The high IPF% in patients with thrombocytopenia in our study indicates peripheral consumption of platelets without bone marrow failure.  Therefore, on closer examination on whether the three compartments have an influence on platelet count, the significant relation between spleen diameter and platelet count indicates the spleen to be the place of consumption and destruction of platelets in our study, without taking into consideration the possible influence of the recently discovered Ashwell-Morell receptor.  These results are consistent with the studies by Zucker et al.  and Tana et al. . The inverse correlation between IPF% and platelet count in our study points to the physiological regulation of platelets in patients with cirrhosis.  These results are in contrast to a recent study that observed no correlation between platelet count and IPF% and proposed that this could indicate development of cirrhosis in patients with chronic liver disease.  Prior to the current study, the association between IPF% and stage of cirrhosis had not been investigated. One limitation of our study is the low number of patients with Child-Pugh stage C, and thus the physiological regulations might not function as they should with progressing cirrhosis. Indeed, serum TPO concentrations and platelet counts have been shown to increase post liver transplantation.  In patients with chronic viral hepatitis TPO levels are elevated and with progression from fibrosis to cirrhosis TPO synthesis in the liver decreases.  We observed no TPO differences in patients with and without thrombocytopenia, however, we cannot fully exclude an influence of impaired TPO synthesis on platelet counts. TPO serum levels are regulated by complex mechanisms. The TPO receptor c-MPL on platelets binds thrombopoietin  and the Ashwell-Morell receptor in the liver can affect platelet counts and TPO production via the JAK-STAT signaling pathway.  Our results are in line with those from Zucker et al.  and Temel et al. . They reported no difference in TPO serum levels between patients with and without thrombocytopenia and no difference according to Child-Pugh stages. TPO levels in controls without cirrhosis however, were higher and platelet counts correlate physiologically with TPO. [10,11] Furthermore IPF% can be elevated in some other entities other than cirrhosis. In addition to the falsely elevated measurement of IPF% values in case of macrothrombocytopenia , IPF% is thought to be elevated in acute coronary syndromes  and acute bacterial infections . In our study, patients with spontaneous bacterial peritonitis and all patients with elevated inflammatory parameters (elevated leucocyte count, c-reactive protein and procalcitonin) no significant difference in IPF% was observed when compared to patients without such conditions. There were no patients with acute coronary syndrome in our study. In conclusion, the present study demonstrates the complex associations between liver cirrhosis and its complications. In our study splenomegaly appears to play the major role in the development of thrombocytopenia in patients with cirrhosis. The regulation of thrombopoesis also appears to be physiological. Closer examination of several compartments did not reveal a single cause of thrombocytopenia. Further studies are needed to investigate the changes of the Ashwell-Morell receptor function and the formation of desialylated platelets in patients with liver cirrhosis. Furthermore, the correlation between the TPO serum levels and IPF% in patients with liver cirrhosis has not been described before. In further studies longitudinal measurements of IPF% would be useful for monitoring progression and development of esophageal varices. Source: http://doi.org/10.1371/journal.pone.0192271