Erythrocytosis after liver transplantation: The experience of a university hospital


Address reprint requests to Gianpaolo Parrilli, M.D., Gastroenterology Unit, Department of Medicine and Surgery, University of Salerno, Salerno, Italy. E-mail:


The prevalence and causes of erythrocytosis after liver transplantation have never been studied, even though this condition is known to predispose patients to thrombosis leading to graft failure or death. Erythrocytosis after orthotopic liver transplantation (OLT) can be defined as an increase in the red cell mass >125% in patients without a pre-OLT history of this condition. The study population was composed of 96 patients: 33 had undergone transplantation for a hepatitis B virus (HBV) infection (18 had a hepatitis D virus coinfection), 43 had undergone transplantation for a hepatitis C virus infection, 9 had undergone transplantation for alcohol abuse, and 11 had undergone transplantation for other causes [autoimmune liver disease (6), Wilson's syndrome (1), or cryptogenetic liver cirrhosis (4)]. Idiopathic erythrocytosis was reported in 11 male patients with a history of HBV infection. Patients with the diagnosis of erythrocytosis underwent phlebotomy every 3 weeks until the hematocrit level reached 45%, and this was repeated if the level exceeded 49%, so no patient presented with cardiovascular accidents during the follow-up. In conclusion, a history of HBV infection, male sex, and hepatitis B immune globulin therapy are all possible cofactors for an increased risk of erythrocytosis in OLT patients. Liver Transpl 19:420–424, 2013. © 2013 AASLD.






hepatitis B virus




hepatitis C virus


orthotopic liver transplantation.

Red blood cells constitute the main fraction of circulating cells in the blood, and they influence blood viscosity. The hematocrit (Hct) and the hemoglobin (Hb) concentration are related parameters. Erythrocytosis is defined as an Hct increase >51% in males (Hb level >18.5 g/dL) and as an Hct increase >48% in females (Hb level >16.5 g/dL in females).[1] To establish whether there is a definite increase in red blood cells, the red cell mass is measured; it is considered to be increased if it is 125% greater than that expected for the sex and body mass.[1, 2]

As shown in Table 1, erythrocytosis can be divided into a primary form, in which there is an intrinsic defect in the erythropoietic component of the bone marrow, and a secondary form, in which the mechanism is not intrinsic to the bone marrow.[1]

Table 1. Causes of Erythrocytosis
  1. NOTE: This table was adapted with permission from International Journal of Laboratory Hematology.[1] Copyright 2008, John Wiley & Sons, Inc.

✓ Polycythemia vera
✓ Genetical mutation of oxygen handling enzymes and hemoglobin
Erythropoietin-mediated :
✓ Chronic diseases: pulmonary, cardiopulmonary shunt, central nervous system diseases
✓ Hypoxia: CO2 poisoning, smoke, sleep apnea, altitude, renal diseases including post-renal transplant erythrocytosis
Pathologic Erythropoietin production
✓ Benign and malign tumors of brain, liver, parathyroid, kidney, and others
✓ Erythropoietin administration
✓ Androgen administration

The main cause of primary erythrocytosis is polycythemia vera, in which not only red cells but also white cells and platelets are overproduced by bone marrow. The erythropoietin (EPO) level in this condition is usually below the normal range (4–24 U/L).[3]

Secondary erythrocytosis can be congenital or acquired. Congenital erythrocytosis is usually diagnosed in young people with a family history (genetic defects of the EPO signaling pathway can be found, and this condition is associated with low EPO levels, whereas genetic defects of the oxygen-sensing pathway lead to elevated EPO levels). Acquired erythrocytosis is caused by central or local renal hypoxic processes, excess EPO production due to a pathological event, or drug administration.[1]

There is also an idiopathic form of erythrocytosis for which no cause has been identified. Reactive thrombocytosis after liver transplantation has been described in the literature,[4] but there are no reports on reactive erythrocytosis; instead, the latter has been described in patients with hepatocellular carcinoma,[5] Budd-Chiari syndrome,[6] and liver hemangioma.[7]

The purpose of our study was to evaluate the prevalence of erythrocytosis in patients with orthotopic liver transplantation (OLT).


The study population consisted of 96 adult patients who had undergone OLT at least 1 year previously; they were consecutively recruited from September 2009 to March 2010 at the Liver Transplantation Center of Frederick II University (Naples, Italy). All subjects gave their informed consent for the study, which was carried out in agreement with the Declaration of Helsinki and was approved by the local ethics committee. Data were collected on age, sex, year, etiology of liver transplantation, and laboratory indices. The exclusion criteria were any symptoms or signs of cirrhosis, a history of pre-transplant erythrocytosis, kidney or respiratory failure, and a diagnosis of a malignant or nonmalignant tumor. All patients with increased Hct values underwent a focused examination at the hematology department to establish the form: primary, secondary or idiopathic.

Thirty-three of the 96 patients who were enrolled had undergone transplantation for a hepatitis B virus (HBV) infection (18 had a hepatitis D virus coinfection), 43 had undergone transplantation for a hepatitis C virus (HCV) infection, 9 had undergone transplantation for alcohol abuse, and 11 had undergone transplantation for other causes [autoimmune liver disease (6), Wilson's syndrome (1), or cryptogenetic liver cirrhosis (4)]. Demographic and diagnostic data are shown in Table 2.

Table 2. Demographics and Diagnostic Parameters of OLT Patients (n=96)
  1. a

    The data are shown as medians and ranges.

Sex [n (%)] 
Males67 (69.8)
Females29 (30.2)
Age at time of study (years)a58 (28-;72)
Time from OLT (years)a7.5 (0-;21)
Etiology [n (%)] 
HCV43 (44.79)
HBV33 (34.38)
Alcohol9 (9.37)
Other11 (11.46)

Statistical Analysis

The Barnard's exact test was used to compare males/females ratio and the number of subjects with Hepatitis D coinfection; the Fligner-Policello test was used to compare median ages and Kaplan-Meier method was used to estimate the median time of erythrocytosis manifestation.


An elevated red cell mass (>125%) was reported in 13 patients. One of these patients had a history of HCV infection and was excluded because he had a diagnosis of recurrent hepatocellular carcinoma 5 years after OLT and was also a carrier of the genetic mutation p.V617F in the Janus kinase 2 gene; another patient with a history of alcohol abuse was excluded because he lived at an unusually high altitude. All 11 patients with erythrocytosis had a history of HBV infection (8 had a hepatitis D virus coinfection), and all were male, were nonsmokers, were hepatitis B surface antigen–negative during therapy with lamivudine and hepatitis B immune globulins, and had good renal function and high EPO levels. Their family histories were negative for erythrocytosis. The Janus kinase 2 gene was the wild type; 4 patients were on cyclosporine therapy, 3 were on tacrolimus, 2 were on sirolimus, and 2 were on mycophenolate mofetil and tacrolimus. In addition, Hb electrophoresis, pulse oximetry, carboxyhemoglobin levels, chest X-rays, and abdominal ultrasound examinations (including the spleen volume) had all given negative results. Erythrocytosis appeared in all 11 patients from the third to twelfth month after liver transplantation. Because they showed no evidence of the causes known for primary or secondary erythrocytosis, a diagnosis of idiopathic erythrocytosis was made for all 11 patients. Table 3 shows differences between HBV-infected patients with erythrocytosis and HBV-infected patients without erythrocytosis.

Table 3. Differences Between HBV-Infected Patients With Erythrocytosis and HBV-Infected Patients Without Erythrocytosis
 Patients Without Erythrocytosis (n=22)Patients With Erythrocytosis (n=11)P Value
  1. a

    The data is shown as median and range.

Male/female (n/n)14/811/00.03
Age at OLT (years)a48.5 (34-;60)43 (32-;58)0.10
Hepatitis D virus coinfection (n)1080.17
Treated with hepatitis B immune globulinAllAllNot applicable
Treated with lamivudineAllAllNot applicable

Follow-Up Evaluation

Blood cell counts were run for all patients monthly for 10 months. Table 4 shows the characteristics of the HBV-infected patients developing erythrocytosis after OLT.

Table 4. Characteristics of HBV-Infected Patients Developing Erythrocytosis After OLT
Patient NumberSexAge at OLT (Years)Hepatitis D Virus CoinfectionOLT DateErythrocytosis DateWhite Blood Cells (×103 cells/μL)aRed Blood Cells (×106 cells/μL)bHb (g/dL)cHct (%)dPlatelets (×103 cells/μL)e
  1. a

    Normal range = 4.8−10.8 × 103 cells/μL.

  2. b

    Normal range = 4.2−5.6 × 106 cells/μL.

  3. c

    Normal range = 12−16 g/dL.

  4. d

    Normal range = 37%−54%.

  5. e

    Normal range = 130−400 × 103 cells/μL.


The Kaplan-Meier curve in Fig. 1 shows that the median time to the development of erythrocytosis was 18 months. Figure 2 shows the Hct fluctuations in the 11 patients developing erythrocytosis. Patients with the diagnosis of erythrocytosis underwent phlebotomy every 3 weeks until the Hct level reached 45%, and this was repeated if the level exceeded 49%, so no patient presented with cardiovascular accidents during the follow-up.

Figure 1.

Kaplan-Meier estimates of survival (hazard ratio=0.0468).

Figure 2.

Hct fluctuations in the 11 patients developing erythrocytosis.


In 1984, Chan et al.[8] described a patient on maintenance hemodialysis who developed a spontaneous increase in erythropoiesis associated with an episode of viral B hepatitis; the level returned to normal with the resolution of the hepatitis. In 2001, Ifudu and Fowler[9] reported that patients on maintenance hemodialysis with an HBV infection required lower doses of EPO than patients without an HBV infection. Even if HBV and HCV are mainly hepatotrophic viruses, they replicate in many extrahepatic tissues. The bone marrow has been termed a sanctuary for HBV, and it displays tropism for immature hematopoietic cells.[10, 11] HCV also shows tropism for bone marrow cells but with a different spectrum of findings.[12]

An HCV infection has no evident directly suppressive effect on bone marrow, whereas an HBV infection frequently results in the suppression of hematopoiesis.[13] In this study, 11 of the 33 patients in the HBV group developed erythrocytosis; this could depend on the genotype of the virus, but all patients were HBV-negative, which however only means that the virus is not circulating in the bloodstream. Because of the intrinsic invasiveness of bone marrow biopsy, we were not allowed to use it as a method of unveiling HBV virus in our patients. Moreover, erythrocytosis developed only after OLT and not when HBV was active, so we can suggest a possible relationship with hepatitis B immune globulin therapy, which was followed by all patients in the HBV group after OLT. The fact that all patients with erythrocytosis were male raises the possibility of a role of androgens. The literature presents evidence that testosterone therapy induces erythrocytosis. Recently, Bachman et al.[14] reported that the suppression of the iron regulatory peptide hepcidin in patients receiving testosterone therapy induced increased Hb and Hct levels more in older men than younger men. Moreover, as is well known, the prevalence of HBV in Italy is 1.5%; males constitute 76% of infected subjects, and 6% of all HBV-infected patients need liver transplantation.[15] Therefore, the joint probability of observing an HBV-infected male who needs OLT is 0.0684%. Similarly, the prevalence of erythrocytosis is 1:200,000, and males constitute 58% of the affected subjects16; the joint probability of observing a male with erythrocytosis is 0.00029%. If these pathologies are independent of each another, the probability of observing an HBV-infected male developing erythrocytosis after OLT is the product of both probabilities and thus 1:500,000,000. In our cohort of liver transplant subjects, 11 of the 33 HBV-infected males developed erythrocytosis: according to a binomial distribution, the probability of observing this is approximately 8·× 10−90. This is strongest evidence that these pathologies are linked.

In conclusion, our findings show that males with an HBV infection who undergo liver transplantation have a higher risk than the general population of developing erythrocytosis, but longitudinal studies are required to confirm our observation. It is generally accepted that a high Hct can increase the risk of thrombosis17; 2 of the patients with erythrocytosis had a history of cardiovascular stroke after OLT, and another patient had a history of hepatic artery thrombosis 8 years after OLT. Therefore, we believe that the type of patient described here deserves more frequent blood cell count screening and/or evaluations for anticoagulant therapy.


The authors thank Mary Spears for her linguistic revisions of the manuscript.