Journal of Viral Hepatitis, 2014, 21, 416–423
doi:10.1111/jvh.12158
Ribavirin suppresses erythroid differentiation and proliferation in chronic hepatitis C patients L. Ronzoni,1 A. Aghemo,2 M. G. Rumi,3 G. Prati,2 A. Colancecco,1 L. Porretti,4 S. Monico,2 M. Colombo2 and M. D. Cappellini1 1Department of Clinical Sciences and Community, Fondazione IRCCS Ca Granda Ospedale Maggiore Policlinico, University of Milan, Milan, Italy; 2A.M. and A. Migliavacca Center for Liver Disease, First Division of Gastroenterology, Fondazione IRCCS Ca Granda Ospedale Maggiore Policlinico, University of Milan, Milan, Italy; 3Division of Hepatology, Ospedale San Giuseppe, University of Milan, Milan, Italy; and 4Cytometry Laboratory, Department of Regenerative Medicine, Fondazione IRCCS Ca Granda, Ospedale Maggiore Policlinico, University of Milan, Milan, Italy Received March 2013; accepted for publication June 2013
SUMMARY. Combination therapy with pegylated interferon (pegIFN) plus ribavirin (RBV) is the standard of care for chronic hepatitis C. One of the major treatment-related side effects is anaemia, attributed to RBV-induced haemolysis. However, haemolysis biomarkers are not present in all patients supporting the existence of other pathogenetic mechanisms. We studied the role of RBV in inducing haemolysis and its effects on erythropoiesis. In 18 hepatitis C virus (HCV) genotype 2 patients treated with pegIFN-alpha-2a (180 mcg/week) plus RBV (800 mg/day) for 24 weeks and in 10 hepatitis B virus (HBV) patients treated with pegIFNalpha-2a (180 mcg/week) for 48 weeks, haemolysis was assessed by serum LDH, haptoglobin and reticulocyte count. Erythropoiesis was evaluated both ex vivo, analysing the clonogenic activity of patients’ erythroid progenitors, as well as in vitro adding pegIFN and RBV to liquid cultures obtained
INTRODUCTION The guanosine analogue ribavirin (RBV) is a key component in the treatment of patients with chronic hepatitis C Abbreviations: ALT, alanine aminotransferase; BFU-E, burst-forming unit – erythroid; CFU-GEMM, colony-forming unit – mixed; DAA, directly acting antiviral; FBS, foetal bovine serum; GPA, glycophorin A; Hb, haemoglobin; HBV, hepatitis B virus; HCV, hepatitis C virus; IFN, interferon; IL-3, interleukin-3; IMPDH, inosine monophosphate dehydrogenase; ITPase, inosine triphosphatase; LDH, lactate dehydrogenase; MNC, mononuclear cells; NS5B, nonstructural 5B protein; PBS, phosphate-buffered saline; pegIFN, pegylated interferon; Rbv, ribavirin; SCF, stem cell factor; SNPs, single nucleotide polymorphisms; SVR, sustained virological response. Correspondence: Maria Domenica Cappellini, Department of Clinical Sciences and Community, Fondazione IRCCS Ca Granda Ospedale Maggiore Policlinico, University of Milan, Via F. Sforza 35, 20122 Milan, Italy. E-mail: [email protected]
from CD34+ cells of healthy volunteers. The majority of patients developed anaemia; the week 4 mean haemoglobin decrease was greater in HCV than in HBV patients (1.7 vs 0.47 g/dL, P = 0.01). Only three HCV patients (17%) and no HBV patients showed signs of haemolysis. The 15 nonhaemolytic HCV patients and all HBV patients showed a delay in erythroid differentiation, with a reduction in colony number and a relative increase in undifferentiated colony percentage. Haemolytic HCV patients had an increase in colony number at week 4 of therapy. In vitro, erythroid cell proliferation and differentiation were inhibited by both pegIFN and RBV. Both pegIFN and RBV have an inhibitory effect on erythroid proliferation and differentiation. Keywords: anaemia, erythropoiesis, haemolysis, hepatitis C virus, pegylated interferon, ribavirin.
virus (HCV) infection. RBV is part of the standard of care regimen for patients with HCV genotype 2, 3 and 4 where it is coupled with peglyated interferon (pegIFN) as well as for patients with HCV genotype 1 where the standard of care regimen consists of pPegIFN/RBV and a directly acting antiviral (DAA) [1,2]. Although the exact mechanisms through which RBV exerts its anti-HCV properties are still somewhat elusive, as RBV shows direct antiviral properties due to its analogue function, as well as inducing the activation of interferon-stimulated genes that de facto potentiate the antiviral state determined by pegIFN, still most attempts to design RBV-free regimens have failed due to unsatisfactory sustained virological response (SVR) rates [3–5]. Unfortunately, RBV is known to determine a plethora of side effects that include rash, xerostomia and most importantly anaemia [6]. Anaemia is observed in up to 30% of patients receiving pegIFN plus RBV combination therapy, with a haemoglobin (Hb) decline that peaks in the first 12 weeks of therapy and that often leads to symptoms, © 2013 John Wiley & Sons Ltd
Ribavirin effects on erythropoiesis impaired quality of life and unnecessary RBV dose reductions [7]. RBV has been shown to cause haemolytic anaemia following its concentration in erythrocytes where it enhances pro-oxidant pathways [8]. This is supported by the recent demonstration that single nucleotide polymorphisms (SNPs) on chromosome 20 in the inosine triphosphatase (ITPase) coding region confer protection against RBV-induced haemolytic anaemia [9,10]. However, the wide heterogeneity in Hb decrease usually observed once RBV is given supports the existence of other pathogenic mechanisms. In theory, RBV could contribute to the wellknown pegIFN-induced myelosuppressive effect, by eliciting an inhibitory effect on cellular proliferation, through the inhibition of the inosine monophosphate dehydrogenase enzyme (IMPDH), which catalyses the oxidation of inosine 5 monophosphate to xanthosine 5 monophosphate [11,12]. Furthermore, RBV may activate the gene p53, which has known antiproliferative activity [13]. With this as a background, we decided to study the impact of RBV on erythropoiesis function by designing an in vivo and in vitro study in patients with chronic hepatitis C receiving pegIFN and RBV, using patients with chronic hepatitis B (HBV) who received pegIFN monotherapy, as controls.
MATERIAL AND METHODS Study design This prospective, open label, controlled study was divided into three parts: in vivo, ex vivo and in vitro. In the in vivo part, haematological measurements were assessed at baseline and during treatment in HCV patients receiving RBV plus pegIFN-alfa-2a combination therapy and HBV patients receiving pegIFN-alfa-2a monotherapy. In the ex vivo part, the clonogenic activity of peripheral blood erythroid progenitors from the HCV and HBV patients was assessed at baseline and at week 4 of antiviral therapy. In the in vitro part, we evaluated the effect of RBV and pegIFN-alfa-2a on cellular proliferation (cell growth and viability) and differentiation (morphology, glycophorin A and gene expression) of erythroid progenitors derived from peripheral blood of healthy volunteers. Written informed consent has been obtained from each patient and healthy volunteer. The study has been performed according to the World Medical Association Declaration of Helsinki and has been approved by the institutional ethics committee.
In vivo study Included were previously untreated patients 18–70 years of age, with higher than normal alanine aminotransferase (ALT) activity, serum HCV RNA positivity, HCV genotype 2 or serum HBV DNA higher than >2000 IU/mL in the presence of anti-HBe positivity. Excluded were patients with © 2013 John Wiley & Sons Ltd
417
haemoglobin ≤12 g/dL for women and ≤13 g/dL for men; alternated white blood cell counts ≤2.5 9 103/mm3; neutrophil count ≤1.5 9 103/mm3; platelet count ≤75 9 103/mm3 and serum creatinine level >1.5 times the upper limit of normal; other liver disease; HIV co-infection; autoimmune diseases; and general contraindications to IFN or RBV. Patients with HCV-2a received RBV 800 mg/day combined with pegIFN-alfa-2a (Pegasys, Roche, Basel Switzerland) 180 mcg/week for 24 weeks, while patients with HBV chronic hepatitis received monotherapy with pegIFNalfa-2a at the same dosage for 48 weeks. Haematological parameters (red blood cells count), iron metabolism (serum iron, transferrin and ferritin), markers of haemolysis (serum lactate dehydrogenase [LDH], serum haptoglobin and reticulocyte count) and serum B12 vitamin and folic acid were assessed at baseline and during treatment weeks 2, 4, 12 and 24 in both HCV and HBV chronic hepatitis patients. Anaemia was defined as grade 2 if haemoglobin levels were 8.5 g/dL.
Ex vivo study (Colony assay) Peripheral blood from HCV and HBV patients was obtained, after informed consent, before starting therapy and after 4 weeks of treatment. Mononuclear cells (MNCs) were isolated by separation on Lymphoprep density gradient (Nycomed Pharma, Oslo, Norway), and single-cell suspensions (2 9 105/mL) were plated in triplicate in methylcellulose (MethoCult H4230; StemCell Technologies, Vancouver, BC, USA) supplemented with cytokines supporting erythroid differentiation: 3 U/mL human recombinant Epo (JanssenCilag, Milan, Italy), 50 ng/mL human recombinant stem cell factor and 10 ng/mL human recombinant interleukin-3 (PeproTech, London, UK). After 14 days of culture in erythroid stimulating conditions, individual progenitor cells give rise to undifferentiated mixed colonies (CFU-GEMM) or to mature erythroid colonies (BFU-E) depending on their maturation status. Colonies were counted after 14 days of incubation at 37 °C in 5% CO2 atmosphere. Colonies were morphologically classified as burst-forming unit – erythroid (BFU-E) or colony-forming unit – mixed (CFU-GEMM) using a Leica microscope (49/0.10). Images were acquired by a Nikon digital camera DS-Fi1 (Nikon Instruments S.p.A., Florence, Italy). Colonies were harvested from methylcellulose for further molecular analysis. After three washings in phosphate-buffered saline (PBS; GIBCO, Grand Island, NY, USA), cells were dissolved in denaturing solution (4M guanidine isothiocyanate, 25 mM sodium citrate pH 7, 0.5% lauryl sarcosyl, 0.1 M b-mercaptoethanol (all from Fluka, St. Louis, MO, USA)) and stored at 80 °C.
In vitro study Peripheral blood from healthy volunteers was collected into sterile heparinized tubes. Light-density mononuclear cells
418
L. Ronzoni et al.
obtained by centrifugation on Lymphoprep (Nycomed Pharma, Oslo, Norway) density gradient were enriched for CD34+ cells by positive selection using CD34+ microbeads (Miltenyi Biotech, Auburn, CA, USA) according to the manufacturers’ instructions. CD34+ cells were cultured at a density of 105 cells/mL in alpha-minimal essential medium (a-MEM; GIBCO, Grand Island, NY, USA) supplemented with 30% foetal bovine serum (FBS; GIBCO, Grand Island, NY, USA), as previously described [14]. To induce cell proliferation and erythroid differentiation, cells were cultured with 20 ng/mL recombinant human (rH) stem cell factor (SCF, PeproTech, London, UK), 10 ng/mL rH interleukin-3 (IL-3, PeproTech, London, UK) and 3 U/mL rH erythropoietin (rHuepo, Janssen-Cilag, Milan, Italy). Cells were incubated at 37 °C with an atmosphere of 5% CO2 for 14 days; after 7 days of culture, the medium was changed to ensure good cell feeding. RBVv (ScheringPlough, Milan, Italy) and pegIFNa-2a (Pegasys, Roche, Basel Switzerland) were added at day 0 of culture at different concentration obtained by dilution with phosphate-buffered saline (PBS; GIBCO, Grand Island, NY, USA) (RBV: 0.1, 0.5 and 1 mM; PegIFN: 100U and 1000U). RBV was read every 2 days due to its short half-life. The RBVv dose added to cell cultures was based on previous in vitro studies [8]. Cell sample were collected on days 7 and 14 for morphology, flow cytometer and molecular analysis. Cells number and viability were determined by trypan blue exclusion.
Statistical analysis Comparisons between groups were made using the Mann– Whitney U-test or Student’s t-test for continuous variables. Ex vivo experiments were run in triplicate, and in vitro experiments on cell lines were repeated at least three times. Twotailed Student’s t-test was applied, and a probability value of P < 0.05 was considered statistically significant. Data handling and analysis were performed with StataViewâ package (SAS Institute Inc., Cary, NC, USA).
RESULTS In- vivo results A minority of hepatitis C virus patients show signs of haemolysis A total of 18 patients with chronic hepatitis C treated with pegIFN-alpha-2a (180 mcg/week) plus RBV (800 mg/day) for 24 weeks and 10 patients with chronic hepatitis B treated with pegIFN-alpha-2a 180 mcg/week monotherapy for 48 weeks were consecutively evaluated. The baseline characteristics of these patients are shown in Table S1. Haematological parameters were similar in the two groups. During the first 4 weeks of treatment, one HCV patient had to reduce RBV to 600 mg/day for anaemia, whereas in HBV
patients, no treatment modifications occurred. The mean haemoglobin decline was greater in HCV than in HBV patients (1.7 vs 0.4 g/dL at week 4, P = 0.01) (Figure S1). One (6%) patient in the HCV group and none in the HBV group developed grade 2 anaemia at week 4. Three (17%) HCV patients showed an increase in LDH values, reticulocyte count and/or a decrease in haptoglobin values suggesting a haemolytic pattern of anaemia. The remaining 15 (83%) HCV patients and all HBV patients did not show any serum marker of haemolysis, as LDH, reticulocyte count and/or haptoglobin values were unchanged. In the three HCV haemolytic patients, the mean Hb decline at week 4 was higher than in the 15 nonhaemolytic HCV patients and HBV patients (3.4 g/dL vs 1.4 g/dL vs 0.4 g/dL, P = 0.04 and 0.003, respectively) (Figure S2). The case of grade 2 anaemia was in a patient with peripheral signs of haemolysis.
Ex vivo results Erythroid colony number supports two different mechanisms of anaemia in hepatitis C virus patients: haemolytic and nonhaemolytic Total colony number was analysed at baseline (W0) and after 4 weeks of treatment (W4) in all patients. By comparing week 4 and baseline colony number, the three HCV patients with signs of haemolysis showed increasing colony values (P < 0.05) that are consistent with increased bone marrow activity in response to peripheral anaemia. On the other hand, the 15 HCV patients without signs of peripheral haemolysis and all HBV patients showed either a significant decrease or relative stability in colony number, a result consistent with an impaired response of the bone marrow to compensate for peripheral anaemia. (Table 1) Erythroid colony morphology and gene expression analysis show delayed erythroid differentiation in the majority of hepatitis C virus patients The morphological analysis of colonies aimed to distinguish between mature erythroid BFU-E and undifferentiated CFUGEMM colonies. The relative percentage of BFU-E and CFUGEMM, calculated as total colony number, is reported in Fig. 1. All the patients at baseline (W0) and at week 4 (W4) showed a prevalence of erythroid mature BFU-E colonies. However, in nonhaemolytic HCV patients and in all HBV patients, a significant increase in CFU-GEMM percentage was observed after 4 weeks of treatment (+14% in nonhaemolytic HCV; +40% in HBV; P < 0.05), while in the three HCV haemolytic patients, the CFU-GEMM percentage did not significantly change between W0 and W4 (10.2% at W0 and 13.3% at W4) (Fig. 1). To estimate colony differentiation stage, the erythroid c-globin and GATA2 gene expression was analysed and compared at W0 and W4 of treatment. These genes are physiologically more expressed at an early stage of differentiation (CFU-GEMM). In all HCV patients, the © 2013 John Wiley & Sons Ltd
Ribavirin effects on erythropoiesis
419
Table 1 Colony assay in semisolid cultures: colonies per 105 seeded cells at baseline (W0) and at week 4 of treatment (W4) HCV pts
W0
Haemolytic 1 6.6 2 1.6 3 12.9 Nonhaemolytic 4 2.2 5 4.0 6 2.6 7 4.5 8 3.0 9 8.5 10 13.0 11 6.5 12 43.2 13 8.0 14 14.0 15 14.8 16 24.7 17 42.7 18 43.6 HBV pts
W0
1 2 3 4 5 6 7 8 9 10
4.3 29.6 3.3 11.8 16.5 10.8 15.8 19 20.1 72.5
W4
W4-W0
1.1 0.8 3.5
61.1 19.1 28.8 7.3 29.5 8.2
54.5* 27.2* 16.6*
6.0 6.8 2.5 4.0 1.6 6.6 10.7 4.0 39.4 1.2 4.0 1.1 4.4 12.9 9.7
3.8 2.8 0.1 0.5 1.4 1.9 2.3 2.5 3.8 6.8 10.0 13.7 20.3 29.8 33.9
0.0 0.0 0.0 0.0 0.0 5.6 2.2 2.6 9.1 4.4 0.0 4.6 7.9 3.8 7.9
2.7 0.8 1.1 1.4 1.4 2.6 3.3 0.0 4.6 1.1 1.2 0.2 1.8 1.6 2.3
W4
2.4 7.1 0.8 3.1 0.9 1.8 1.2 7.0 4.7 13.9
12.6 28.8 0.8 5.5 5.5 0.8 3.9 4.1 0.2 5.3
W4-W0
6.4 5.9 0.1 1.8 2.3 0.9 1.6 1.2 0.1 1.0
8.3 0.8 2.5 6.3 11.0 10.0 11.9 14.9 19.9 67.2
HBV, hepatitis B virus; HCV, hepatitis C virus. For each patients, three independent experiments were set up; values are expressed as means SD. *P < 0.05 by Student’s t-test. expression of these two genes increased at week 4 of treatment. The same was true in HBV patients, thus confirming an increase in undifferentiated colonies in both treatment groups (c-globin: 2.2-fold increase in haemolytic HCV, 3 in nonhaemolytic HCV and 2.7 in HBV; GATA2: 1.2-fold increase in haemolytic HCV, 1.4 in nonhaemolytic HCV and 1.6 in HBV).
In vitro results The ex vivo demonstrations of a delay in erythroid progenitor differentiation and maturation prompted us to evaluate © 2013 John Wiley & Sons Ltd
Fig. 1 Colony morphology. Percentage of different types of colonies (BFU-E and CFU-GEMM) derived from peripheral blood mononuclear cells of hepatitis C virus (HCV) (haemolytic and nonhaemolytic) and hepatitis B virus (HBV) patients at baseline and week 4 of treatment. Values are means of each group patients. *P < 0.05.
the effect of pegIFN and RBV in an in vitro culture system. Liquid cultures of CD34+ human stem cells isolated from peripheral blood of healthy volunteers and stimulated towards erythroid differentiation were set up. Different concentrations of pegIFN (100U and 1000U) and RBV (0.1, 0.5 and 1 mM), which were chosen on the basis of previous in vitro studies [8,13], were added at day 0 on isolated CD34+ cells. Cell viability and growth, erythroid differentiation and haemolysis were analysed under drug influence in vitro. Ribavirin inhibits erythroid cell proliferation Pegylated interferon 100U and 1000U had no significant effects on cell proliferation (Fig. 2a) and viability (Fig. 2b). RBV significantly inhibited cell proliferation (Fig. 2a). 1 mM RBV reduced cell growth by 49-fold compared with untreated cells (P < 0.004), with a significant reduction in cell viability also (P = 0.01) (Fig. 2b). Lower concentrations of RBV (0.1 and 0.5 mM) decreased cell growth by 7-fold (P < 0.05) without significantly affecting cell viability. Higher RBV doses (1 mM) were excluded from further analysis due to toxic effect. Pegylated interferon and ribavirin delay erythroid differentiation at different maturation steps The effect of pegIFN and BV on erythroid differentiation was evaluated by analysing glycophorin A (GPA), cell morphology and primitive erythroid-specific c-globin and GATA2 gene expression at day 14 of culture. Figure 3a,b shows the cytofluorimetric expression of GPA, a specific marker of erythroid maturation. In physiological conditions, after 2 weeks of culture, more than 70% of cells were GPA+. Concentrations of 100U and 1000U of PegIFN
420
L. Ronzoni et al.
Fig. 2 Growth kinetics and viability of erythroid progenitor cells treated with ribavirin (RBV) or pegylated interferon (pegIFN). CD34 + cells derived from peripheral blood of healthy volunteers were grown in the presence of RBV (0.1, 0.5 and 1 mM) or pegIFN (100 and 1000U). (a) Cell numbers of three representative experiments determined at regular time intervals are shown as means SD. (b) Viability of control and treated cells of three independent experiments are reported as means SD. *P < 0.005; ‡P < 0.05.
did not affect GPA expression, which instead was significantly reduced by RBV 0.1 mM and 0.5 mM (26% and 27%, respectively, P < 0.05). Figure 3c shows the results of the morphological studies of erythroid differentiation. Physiological erythroid differentiation includes maturation from proerythroblasts to the mature orthochromatic erythroblasts, with intermediate steps including basophilic and polychromatophilic erythroblasts. In physiological conditions, at day 14 of culture, the majority of cells are orthochromatic erythroblasts (52%), including 20% polychromatophilic and basophilic erythroblasts and almost no proerythroblasts (only 0.5%). Both pegIFN and RBV led to delayed erythroid differentiation, increasing the percentage of undifferentiated cells and consequently decreasing the percentage of more mature erythroblasts. Concentrations of 100U and 1000U of pegIFN increased proerythroblast percentage from 0.5% of untreated cells to 11.5% and 10.5% (P < 0.05), respectively, and reduced orthochromatic erythroblasts from 52% to 37% and 41%, respectively. Concentrations of 0.1 and 0.5 mM of Rbv significantly increased the percentage of polychromatophilic erythroblasts from 21% in untreated cells to 41% and 57%, respectively (P < 0.03), while decreasing the percentage of orthochromatic erythroblasts from 52% to
Fig. 3 Ribavirin and pegylated interferon (pegIFN) effect on erythroid differentiation. (a and b) Glycophorin A expression of control and treated cells at day 14 of culture analysed by flow cytometry. Means SD of three independent experiments are presented. *P < 0.05. Histograms represent the MFI relative to the cell number. (c) Aliquots of cells at day 14 of culture were subjected to cytocentrifugation and staining with benzidine and histological dyes and analysed for the proportion of proerythroblasts, basophilic, polychromatophilic and orthochromatic erythroblast, erythrocytes and nonerythroid cells. Means of three independent experiments are reported. (d) GATA2 and c-globin gene expression profile of control and treated cells at day 14 of culture. Results are expressed as means SD of three independent experiments.
44% and 23%, respectively. Figure 3d shows the expression of the primitive erythroid-specific c-globin and GATA2 genes at day 14 of culture. Physiologically, these genes are © 2013 John Wiley & Sons Ltd
Ribavirin effects on erythropoiesis more expressed in undifferentiated cells (proerythroblasts), while their expression declines during maturation. PegIFN increased both c-globin and GATA2 gene expression (1.6and 1.4-fold increase, respectively, with 100U pegIFN; 1.8and 1.5-fold increase, respectively, with 1000U pegIFN) as a consequence of increased numbers of undifferentiated proerythroblasts. RBV increased GATA2 gene expression (4.3-fold increase) without affecting c-globin gene expression, suggesting that RBV affects mainly polychromatophilic erythroblasts not expressing c-globin genes. These data are consistent with an inhibitory effect of both pegIFN and RBV on erythroid differentiation, however, at different time points of the physiological maturation process.
DISCUSSION The present study challenges the paradigm that RBVinduced anaemia is a mere consequence of extravascular haemolysis. Our in vivo and in vitro findings, in fact, directly correlate anaemia with RBV-induced suppression of erythroid differentiation and proliferation. In the in vivo part of our study, only 17% of patients exposed to pegIFN and RBV actually showed biochemical evidence of haemolysis, which was confirmed by an increase in colony numbers after 4 weeks of treatment compared with baseline values in the ex vivo experiments. Consistently with these findings, the remaining HCV patients (83%) did not have serum markers of peripheral haemolysis, while also showing in the ex vivo analysis a reduction in colony numbers with a relative increase in undifferentiated colonies (CFU-GEMM), a pattern consistent with suppression of erythroid proliferation and differentiation. The suppressive effect of pegIFN/RBV on erythroid differentiation was also confirmed by the increased expression of c-globin and GATA2 genes that are typical of the early stages of differentiation, which was observed after 4 weeks of anti-HCV therapy. Interestingly, the prominent role of RBV in this process is clarified by the in vitro demonstration of a significant inhibition of cell proliferation, as well as a delay in erythroid differentiation, once RBV was added to CD34+ human stem cells isolated from the peripheral blood of healthy volunteers who were stimulated towards erythroid differentiation. Concentrations of 0.1 and 0.5 mM of RBV significantly increased the percentage of polychromatophilic erythroblasts from 21% in untreated cells to 41% and 57%, respectively, while they decreased the percentage of orthochromatic erythroblasts from 52% to 44% and 23%, respectively. While our study confirms the well-known pegIFN myelosuppressive effect, as concentrations of 100U and 1000U of pegIFN increased proerythroblast percentage from 0.5% in untreated cells to 11.5% and 10.5%, respectively, and reduced orthochromatic erythroblasts from 52% to 37% and 41%, respectively, it also highlights that pegIFN and RBV act at different points of the erythroid differentiation process. PegIFN increased both c-globin and GATA2 gene © 2013 John Wiley & Sons Ltd
421
expression as a consequence of increased numbers of undifferentiated proerythroblasts, while RBV increased GATA2 gene expression without affecting c-globin gene expression, suggesting that RBV affects more mature forms of erythroid precursors such as polychromatophilic erythroblasts that do not express c-globin genes. Although the myelosuppressive effect of IFN is not a novelty, our study is the first to show RBV to directly suppress erythroid differentiation and proliferation. Given the multiple functions of RBV, several mechanisms through which RBV alone determines erythroid suppression may be accounted for. RBV is known to inhibit the IMPDH, which catalyses the oxidation of inosine 5 monophosphate to xanthosine 5 monophosphate, leading to a concomitant reduction in nicotinamide adenine dinucleotide to thio-nicotinamide adenine dinucleotide. This reaction is regulating the rate-limiting step of guanine nucleotide biosynthesis, whose inhibition apparently explains the negative effect of RBV on cell proliferation [12,15]. The latter can also be inhibited by the tumour suppression activity of the p53 gene whose antiproliferative activity can be primed by RBV [13]. Owing to the fact that IMPDH activity is downregulated by p53 [16], the p53-RBV interactions can partly account for IMPDH inhibition by RBV. Finally, RBV may act synergistically with IFN in upregulating IFN-stimulated genes causing enhanced expression of cell genes involved in IFN signalling pathways [11,13]. From a clinical standpoint, our observations may have practical implications also with respect to the therapeutic algorithms of HCV protease inhibitors in combination with pegIFN and RBV, which bear a higher risk of anaemia compared with dual pegIFN and RBv therapy [17–20]. To minimize the risk of anaemia, the concomitant use of drugs with known myelosuppressive activity should be avoided; moreover, management of treatment-related anaemia through blood transfusions should be generally discouraged as transfused red blood cells will only transiently increase Hb values while potentially further inhibiting bone marrow function. Based on our findings, stepwise RBV down dosing and eventual concomitant pegIFN down dosing should be the standard of care management of treatment-related anaemia. Indeed, given that the two drugs act on different points of erythroid differentiation, a synergistic interaction may potentiate the observed haemoglobin drop in most HCV patients. This was the case in a Phase II study where RBV was combined with an NS5B polymerase inhibitor with or without pegIFN. The mean haemoglobin decline was far less pronounced in the pegIFN-free regimen as compared to the pegIFN/RBV/NS5B polymerase inhibitor arm, demonstrating the key role of pegIFN in treatmentrelated anaemia [21]. While our study shows that in the majority of patients anaemia is likely to be a direct consequence of RBV- and pegIFN-induced suppression of erythroid differentiation and proliferation, our findings do not completely negate RBV-
422
L. Ronzoni et al.
induced haemolysis. Indeed, a subgroup of HCV patients had peripheral signs of haemolysis that was confirmed by an increase in erythroid progenitor colony number. Moreover, from a clinical standpoint, these patients were the most difficult to manage as they were those with a more rapid and pronounced haemoglobin decline at week 4 of pegIFN/RBV therapy. Still, the in vitro section of our study could not replicate previous studies on RBV-induced haemolysis. One of the possible explanations is that in previous in vitro studies [8], RBV was dosed at 1 mM, which in the present study and in another concomitant study [13], led to cell death as assessed by reduced cell viability. We do acknowledge that our study has the limitation that RBV monotherapy was studied only in vitro; however, in clinical practice, RBV monotherapy is currently not feasible because RBV is active against HCV in combination with IFN or DAA only. Another limitation is the small sample size of the study which prevented the identification of pretreatment predictors of haemolysis or severe bone marrow suppression during pegIFN plus RBV therapy. Indeed, the three patients with the haemolytic pattern did not differ in terms of age, kidney function or ITPA genotype (data not shown) from the nonhaemolytic patients. Likewise with previous studies, we could not correlate the level of colony-forming erythroid progenitor cells with the development and severity of anaemia or with the haemolytic or myelosuppressive pattern of anaemia at the individual level [22]. Lastly, we did not assess the ITPA genotype in healthy volunteers, as the role of ITPA in treatment-
induced anaemia had not yet been discovered at the time of the study design. These caveats notwithstanding, our study is the first to demonstrate a direct myelosuppressive effect of RBV on red blood cells proliferation and differentiation, shedding further lights on the multiple mechanisms of action of this drug that since its discovery remains a fundamental backbone of anti-HCV therapy.
STATEMENT OF INTERESTS Alessio Aghemo has served as a speaker for Roche and Janssen and has received research funding from Roche and Gilead Sciences. Maria Grazia Rumi has served as an advisory board member for Tibotec, Roche and Janssen Cilag. Massimo Colombo has served as a speaker for Tibotec, Roche, Novartis, Bayer, BMS, Gilead Science, Vertex and as an advisory board member for Merck, Roche, Novartis, Bayer, BMS, Gilead Science, Tibotec, Vertex, Achillion; he has received research funding Merck, Roche, BMS, Gilead Science. Maria Domenica Cappellini has served as an advisory board member for Novartis, Genzyme, Shire. Other authors have no financial disclosure to declare. This study was funded in part by a grant from the Universit a degli Studi di Milano, Milan, Italy (FIRST 2009) to MC; a PRIN grant from ‘Ministero Italiano dell’Universit ae della Ricerca’ (PRIN 2008) to MDC; and grants from Fondazione IRCCS C a Granda, Ospedale Maggiore Policlinico, Milan, Italy (RC 2009 and RC 2010) to MDC.
REFERENCES 1 Ghany MG, Nelson DR, Strader DB, Thomas DL, Seeff LB. An update on treatment of genotype 1 chronic hepatitis C virus infection: 2011 practice guideline by the American Association for the Study of Liver Diseases. Hepatology 2011; 54: 1433–1444. 2 European Association for the Study of the Liver. EASL clinical practice guidelines: management of hepatitis C virus infection. J Hepatol 2011; 55 (2): 245–264. 3 Feld JJ, Hoofnagle JH. Mechanism of action of interferon and ribavirin in treatment of hepatitis C. Nature 2005; 436(7053): 967–972. 4 Feld JJ, Lutchman GA, Heller T et al. Ribavirin improves early responses to peginterferon through improved interferon signaling. Gastroenterology 2010; 139(1): 154–162. 5 Hezode C, Forestier N, Dusheiko G et al. Telaprevir and peginterferon
with or without ribavirin for chronic HCV infection. N Engl J Med 2009; 360(18): 1839–1850. 6 Aghemo A, Rumi MG, Monico S et al. Ribavirin impairs salivary gland function during combination treatment with pegylated interferon alfa-2a in hepatitis C patients. Hepat Mon 2011; 11(11): 918– 924. 7 Sulkowski MS, Cooper C, Hunyady B et al. Management of adverse effects of Peg-IFN and ribavirin therapy for hepatitis C. Nat Rev Gastroenterol Hepatol 2011; 8(4): 212– 223. 8 De Franceschi L, Fattovich G, Turrini F et al. Hemolytic anemia induced by Ribavirin therapy in patients with chronic hepatitis C virus infection: role of membrane oxidative damage. Hepatology 2000; 31: 997–1004.
9 Fellay J, Thompson AJ, Ge D et al. ITPA gene variants protect against anaemia in patients treated for chronic hepatitis C. Nature 2010; 464(7287): 405–408. 10 Thompson AJ, Fellay J, Patel K et al. Variants in the ITPA gene protect against ribavirin-induced hemolytic anemia and decrease the need for ribavirin dose reduction. Gastroenterology 2010; 139(4): 1181–1189. 11 Feld JJ, Nanda S, Huang Y et al. Hepatic gene expression during treatment with peginterferon and ribavirin: identifying molecular pathways for treatment response. Hepatology 2007; 46(5): 1548– 1563. 12 Jayaram HN, Cooney DA, Gruseh M, Krupitza G. Consequences of IMP dehydrogenase inhibition and its relationship to cancer and apop-
© 2013 John Wiley & Sons Ltd
Ribavirin effects on erythropoiesis
13
14
15
16
tosis. Curr Med Chem 1999; 6: 561– 574. Su WC, Liu WL, Cheng CW et al. Ribavirin enhances interferon signaling via stimulation of mTOR and p53 activities. FEBS Lett 2009; 583 (17): 2793–2798. Ronzoni L, Bonara P, Rusconi D, Frugoni C, Libani I, Cappellini MD. Erythroid differentiation and maturation from peripheral CD34+ cells in liquid culture: cellular and molecular characterization. Blood Cells Mol Dis 2008; 40(2): 148–155. Wang W, Hedstrom L. Kinetic mechanism of human inosine 5-monophosphate dehydrogenase type II: random addition of substrates and ordered release of products. Biochemistry 1997; 36: 8479–8483. Liu Y, Bohn SA, Sherley JL. Inosine5-monophosphate dehydrogenase is
423
a rate determining factor for p53dependent growth regulation. Mol Biol Cell 1998; 9: 15–28. 17 Kwo PY, Lawitz EJ, McCone J et al. Efficacy of boceprevir, an NS3 protease inhibitor, in combination with peginterferon alfa-2b and ribavirin in treatment-naive patients with genotype 1 hepatitis C infection (SPRINT-1): an open-label, randomised, multicentre phase 2 trial. Lancet 2010; 376: 705–716. 18 Jacobson IM, McHutchison JG, Dusheiko G et al. Telaprevir for previously untreated chronic hepatitis C virus infection. NEJM 2011; 364: 2405–2416. 19 Poordad F, McCone J Jr, Bacon BR et al. Boceprevir for untreated chronic HCV genotype 1 infection. N Engl J Med 2011; 364: 1195– 1206.
20 Bacon BR, Gordon SC, Lawitz E et al. Boceprevir for previously treated chronic HCV genotype 1 infection. N Engl J Med 2011; 364: 1207–1217. 21 Gane EJ, Stedman CA, Hyland RH et al. Once daily PSI-7977 plus Rbv: pegylated interferon-alfa not required for complete rapid viral response in treatment-naive patients with HCV GT2 or GT3 [Abstract]. Hepatology 2011; 54(4): 76A. 22 Brunati C, Cappellini MD, De Feo T et al. Uremic inhibitors of erythropoiesis: a study during treatment with recombinant human erythropoietin. Am J Nephrol 1992; 12(1– 2): 9–13.
Figure S1. Quantitative hemoglobin reduction from the baseline in HCV and HBV patients.
Figure S2. Quantitative hemoglobin reduction from the baseline in hemolytic and non-hemolytic HCV patients and in HBV patients.
SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article: Table S1. Baseline characteristics of patients with chronic hepatitis C (HCV) and B (HBV).
© 2013 John Wiley & Sons Ltd
doi:10.1111/jvh.12158
Ribavirin suppresses erythroid differentiation and proliferation in chronic hepatitis C patients L. Ronzoni,1 A. Aghemo,2 M. G. Rumi,3 G. Prati,2 A. Colancecco,1 L. Porretti,4 S. Monico,2 M. Colombo2 and M. D. Cappellini1 1Department of Clinical Sciences and Community, Fondazione IRCCS Ca Granda Ospedale Maggiore Policlinico, University of Milan, Milan, Italy; 2A.M. and A. Migliavacca Center for Liver Disease, First Division of Gastroenterology, Fondazione IRCCS Ca Granda Ospedale Maggiore Policlinico, University of Milan, Milan, Italy; 3Division of Hepatology, Ospedale San Giuseppe, University of Milan, Milan, Italy; and 4Cytometry Laboratory, Department of Regenerative Medicine, Fondazione IRCCS Ca Granda, Ospedale Maggiore Policlinico, University of Milan, Milan, Italy Received March 2013; accepted for publication June 2013
SUMMARY. Combination therapy with pegylated interferon (pegIFN) plus ribavirin (RBV) is the standard of care for chronic hepatitis C. One of the major treatment-related side effects is anaemia, attributed to RBV-induced haemolysis. However, haemolysis biomarkers are not present in all patients supporting the existence of other pathogenetic mechanisms. We studied the role of RBV in inducing haemolysis and its effects on erythropoiesis. In 18 hepatitis C virus (HCV) genotype 2 patients treated with pegIFN-alpha-2a (180 mcg/week) plus RBV (800 mg/day) for 24 weeks and in 10 hepatitis B virus (HBV) patients treated with pegIFNalpha-2a (180 mcg/week) for 48 weeks, haemolysis was assessed by serum LDH, haptoglobin and reticulocyte count. Erythropoiesis was evaluated both ex vivo, analysing the clonogenic activity of patients’ erythroid progenitors, as well as in vitro adding pegIFN and RBV to liquid cultures obtained
INTRODUCTION The guanosine analogue ribavirin (RBV) is a key component in the treatment of patients with chronic hepatitis C Abbreviations: ALT, alanine aminotransferase; BFU-E, burst-forming unit – erythroid; CFU-GEMM, colony-forming unit – mixed; DAA, directly acting antiviral; FBS, foetal bovine serum; GPA, glycophorin A; Hb, haemoglobin; HBV, hepatitis B virus; HCV, hepatitis C virus; IFN, interferon; IL-3, interleukin-3; IMPDH, inosine monophosphate dehydrogenase; ITPase, inosine triphosphatase; LDH, lactate dehydrogenase; MNC, mononuclear cells; NS5B, nonstructural 5B protein; PBS, phosphate-buffered saline; pegIFN, pegylated interferon; Rbv, ribavirin; SCF, stem cell factor; SNPs, single nucleotide polymorphisms; SVR, sustained virological response. Correspondence: Maria Domenica Cappellini, Department of Clinical Sciences and Community, Fondazione IRCCS Ca Granda Ospedale Maggiore Policlinico, University of Milan, Via F. Sforza 35, 20122 Milan, Italy. E-mail: [email protected]
from CD34+ cells of healthy volunteers. The majority of patients developed anaemia; the week 4 mean haemoglobin decrease was greater in HCV than in HBV patients (1.7 vs 0.47 g/dL, P = 0.01). Only three HCV patients (17%) and no HBV patients showed signs of haemolysis. The 15 nonhaemolytic HCV patients and all HBV patients showed a delay in erythroid differentiation, with a reduction in colony number and a relative increase in undifferentiated colony percentage. Haemolytic HCV patients had an increase in colony number at week 4 of therapy. In vitro, erythroid cell proliferation and differentiation were inhibited by both pegIFN and RBV. Both pegIFN and RBV have an inhibitory effect on erythroid proliferation and differentiation. Keywords: anaemia, erythropoiesis, haemolysis, hepatitis C virus, pegylated interferon, ribavirin.
virus (HCV) infection. RBV is part of the standard of care regimen for patients with HCV genotype 2, 3 and 4 where it is coupled with peglyated interferon (pegIFN) as well as for patients with HCV genotype 1 where the standard of care regimen consists of pPegIFN/RBV and a directly acting antiviral (DAA) [1,2]. Although the exact mechanisms through which RBV exerts its anti-HCV properties are still somewhat elusive, as RBV shows direct antiviral properties due to its analogue function, as well as inducing the activation of interferon-stimulated genes that de facto potentiate the antiviral state determined by pegIFN, still most attempts to design RBV-free regimens have failed due to unsatisfactory sustained virological response (SVR) rates [3–5]. Unfortunately, RBV is known to determine a plethora of side effects that include rash, xerostomia and most importantly anaemia [6]. Anaemia is observed in up to 30% of patients receiving pegIFN plus RBV combination therapy, with a haemoglobin (Hb) decline that peaks in the first 12 weeks of therapy and that often leads to symptoms, © 2013 John Wiley & Sons Ltd
Ribavirin effects on erythropoiesis impaired quality of life and unnecessary RBV dose reductions [7]. RBV has been shown to cause haemolytic anaemia following its concentration in erythrocytes where it enhances pro-oxidant pathways [8]. This is supported by the recent demonstration that single nucleotide polymorphisms (SNPs) on chromosome 20 in the inosine triphosphatase (ITPase) coding region confer protection against RBV-induced haemolytic anaemia [9,10]. However, the wide heterogeneity in Hb decrease usually observed once RBV is given supports the existence of other pathogenic mechanisms. In theory, RBV could contribute to the wellknown pegIFN-induced myelosuppressive effect, by eliciting an inhibitory effect on cellular proliferation, through the inhibition of the inosine monophosphate dehydrogenase enzyme (IMPDH), which catalyses the oxidation of inosine 5 monophosphate to xanthosine 5 monophosphate [11,12]. Furthermore, RBV may activate the gene p53, which has known antiproliferative activity [13]. With this as a background, we decided to study the impact of RBV on erythropoiesis function by designing an in vivo and in vitro study in patients with chronic hepatitis C receiving pegIFN and RBV, using patients with chronic hepatitis B (HBV) who received pegIFN monotherapy, as controls.
MATERIAL AND METHODS Study design This prospective, open label, controlled study was divided into three parts: in vivo, ex vivo and in vitro. In the in vivo part, haematological measurements were assessed at baseline and during treatment in HCV patients receiving RBV plus pegIFN-alfa-2a combination therapy and HBV patients receiving pegIFN-alfa-2a monotherapy. In the ex vivo part, the clonogenic activity of peripheral blood erythroid progenitors from the HCV and HBV patients was assessed at baseline and at week 4 of antiviral therapy. In the in vitro part, we evaluated the effect of RBV and pegIFN-alfa-2a on cellular proliferation (cell growth and viability) and differentiation (morphology, glycophorin A and gene expression) of erythroid progenitors derived from peripheral blood of healthy volunteers. Written informed consent has been obtained from each patient and healthy volunteer. The study has been performed according to the World Medical Association Declaration of Helsinki and has been approved by the institutional ethics committee.
In vivo study Included were previously untreated patients 18–70 years of age, with higher than normal alanine aminotransferase (ALT) activity, serum HCV RNA positivity, HCV genotype 2 or serum HBV DNA higher than >2000 IU/mL in the presence of anti-HBe positivity. Excluded were patients with © 2013 John Wiley & Sons Ltd
417
haemoglobin ≤12 g/dL for women and ≤13 g/dL for men; alternated white blood cell counts ≤2.5 9 103/mm3; neutrophil count ≤1.5 9 103/mm3; platelet count ≤75 9 103/mm3 and serum creatinine level >1.5 times the upper limit of normal; other liver disease; HIV co-infection; autoimmune diseases; and general contraindications to IFN or RBV. Patients with HCV-2a received RBV 800 mg/day combined with pegIFN-alfa-2a (Pegasys, Roche, Basel Switzerland) 180 mcg/week for 24 weeks, while patients with HBV chronic hepatitis received monotherapy with pegIFNalfa-2a at the same dosage for 48 weeks. Haematological parameters (red blood cells count), iron metabolism (serum iron, transferrin and ferritin), markers of haemolysis (serum lactate dehydrogenase [LDH], serum haptoglobin and reticulocyte count) and serum B12 vitamin and folic acid were assessed at baseline and during treatment weeks 2, 4, 12 and 24 in both HCV and HBV chronic hepatitis patients. Anaemia was defined as grade 2 if haemoglobin levels were 8.5 g/dL.
Ex vivo study (Colony assay) Peripheral blood from HCV and HBV patients was obtained, after informed consent, before starting therapy and after 4 weeks of treatment. Mononuclear cells (MNCs) were isolated by separation on Lymphoprep density gradient (Nycomed Pharma, Oslo, Norway), and single-cell suspensions (2 9 105/mL) were plated in triplicate in methylcellulose (MethoCult H4230; StemCell Technologies, Vancouver, BC, USA) supplemented with cytokines supporting erythroid differentiation: 3 U/mL human recombinant Epo (JanssenCilag, Milan, Italy), 50 ng/mL human recombinant stem cell factor and 10 ng/mL human recombinant interleukin-3 (PeproTech, London, UK). After 14 days of culture in erythroid stimulating conditions, individual progenitor cells give rise to undifferentiated mixed colonies (CFU-GEMM) or to mature erythroid colonies (BFU-E) depending on their maturation status. Colonies were counted after 14 days of incubation at 37 °C in 5% CO2 atmosphere. Colonies were morphologically classified as burst-forming unit – erythroid (BFU-E) or colony-forming unit – mixed (CFU-GEMM) using a Leica microscope (49/0.10). Images were acquired by a Nikon digital camera DS-Fi1 (Nikon Instruments S.p.A., Florence, Italy). Colonies were harvested from methylcellulose for further molecular analysis. After three washings in phosphate-buffered saline (PBS; GIBCO, Grand Island, NY, USA), cells were dissolved in denaturing solution (4M guanidine isothiocyanate, 25 mM sodium citrate pH 7, 0.5% lauryl sarcosyl, 0.1 M b-mercaptoethanol (all from Fluka, St. Louis, MO, USA)) and stored at 80 °C.
In vitro study Peripheral blood from healthy volunteers was collected into sterile heparinized tubes. Light-density mononuclear cells
418
L. Ronzoni et al.
obtained by centrifugation on Lymphoprep (Nycomed Pharma, Oslo, Norway) density gradient were enriched for CD34+ cells by positive selection using CD34+ microbeads (Miltenyi Biotech, Auburn, CA, USA) according to the manufacturers’ instructions. CD34+ cells were cultured at a density of 105 cells/mL in alpha-minimal essential medium (a-MEM; GIBCO, Grand Island, NY, USA) supplemented with 30% foetal bovine serum (FBS; GIBCO, Grand Island, NY, USA), as previously described [14]. To induce cell proliferation and erythroid differentiation, cells were cultured with 20 ng/mL recombinant human (rH) stem cell factor (SCF, PeproTech, London, UK), 10 ng/mL rH interleukin-3 (IL-3, PeproTech, London, UK) and 3 U/mL rH erythropoietin (rHuepo, Janssen-Cilag, Milan, Italy). Cells were incubated at 37 °C with an atmosphere of 5% CO2 for 14 days; after 7 days of culture, the medium was changed to ensure good cell feeding. RBVv (ScheringPlough, Milan, Italy) and pegIFNa-2a (Pegasys, Roche, Basel Switzerland) were added at day 0 of culture at different concentration obtained by dilution with phosphate-buffered saline (PBS; GIBCO, Grand Island, NY, USA) (RBV: 0.1, 0.5 and 1 mM; PegIFN: 100U and 1000U). RBV was read every 2 days due to its short half-life. The RBVv dose added to cell cultures was based on previous in vitro studies [8]. Cell sample were collected on days 7 and 14 for morphology, flow cytometer and molecular analysis. Cells number and viability were determined by trypan blue exclusion.
Statistical analysis Comparisons between groups were made using the Mann– Whitney U-test or Student’s t-test for continuous variables. Ex vivo experiments were run in triplicate, and in vitro experiments on cell lines were repeated at least three times. Twotailed Student’s t-test was applied, and a probability value of P < 0.05 was considered statistically significant. Data handling and analysis were performed with StataViewâ package (SAS Institute Inc., Cary, NC, USA).
RESULTS In- vivo results A minority of hepatitis C virus patients show signs of haemolysis A total of 18 patients with chronic hepatitis C treated with pegIFN-alpha-2a (180 mcg/week) plus RBV (800 mg/day) for 24 weeks and 10 patients with chronic hepatitis B treated with pegIFN-alpha-2a 180 mcg/week monotherapy for 48 weeks were consecutively evaluated. The baseline characteristics of these patients are shown in Table S1. Haematological parameters were similar in the two groups. During the first 4 weeks of treatment, one HCV patient had to reduce RBV to 600 mg/day for anaemia, whereas in HBV
patients, no treatment modifications occurred. The mean haemoglobin decline was greater in HCV than in HBV patients (1.7 vs 0.4 g/dL at week 4, P = 0.01) (Figure S1). One (6%) patient in the HCV group and none in the HBV group developed grade 2 anaemia at week 4. Three (17%) HCV patients showed an increase in LDH values, reticulocyte count and/or a decrease in haptoglobin values suggesting a haemolytic pattern of anaemia. The remaining 15 (83%) HCV patients and all HBV patients did not show any serum marker of haemolysis, as LDH, reticulocyte count and/or haptoglobin values were unchanged. In the three HCV haemolytic patients, the mean Hb decline at week 4 was higher than in the 15 nonhaemolytic HCV patients and HBV patients (3.4 g/dL vs 1.4 g/dL vs 0.4 g/dL, P = 0.04 and 0.003, respectively) (Figure S2). The case of grade 2 anaemia was in a patient with peripheral signs of haemolysis.
Ex vivo results Erythroid colony number supports two different mechanisms of anaemia in hepatitis C virus patients: haemolytic and nonhaemolytic Total colony number was analysed at baseline (W0) and after 4 weeks of treatment (W4) in all patients. By comparing week 4 and baseline colony number, the three HCV patients with signs of haemolysis showed increasing colony values (P < 0.05) that are consistent with increased bone marrow activity in response to peripheral anaemia. On the other hand, the 15 HCV patients without signs of peripheral haemolysis and all HBV patients showed either a significant decrease or relative stability in colony number, a result consistent with an impaired response of the bone marrow to compensate for peripheral anaemia. (Table 1) Erythroid colony morphology and gene expression analysis show delayed erythroid differentiation in the majority of hepatitis C virus patients The morphological analysis of colonies aimed to distinguish between mature erythroid BFU-E and undifferentiated CFUGEMM colonies. The relative percentage of BFU-E and CFUGEMM, calculated as total colony number, is reported in Fig. 1. All the patients at baseline (W0) and at week 4 (W4) showed a prevalence of erythroid mature BFU-E colonies. However, in nonhaemolytic HCV patients and in all HBV patients, a significant increase in CFU-GEMM percentage was observed after 4 weeks of treatment (+14% in nonhaemolytic HCV; +40% in HBV; P < 0.05), while in the three HCV haemolytic patients, the CFU-GEMM percentage did not significantly change between W0 and W4 (10.2% at W0 and 13.3% at W4) (Fig. 1). To estimate colony differentiation stage, the erythroid c-globin and GATA2 gene expression was analysed and compared at W0 and W4 of treatment. These genes are physiologically more expressed at an early stage of differentiation (CFU-GEMM). In all HCV patients, the © 2013 John Wiley & Sons Ltd
Ribavirin effects on erythropoiesis
419
Table 1 Colony assay in semisolid cultures: colonies per 105 seeded cells at baseline (W0) and at week 4 of treatment (W4) HCV pts
W0
Haemolytic 1 6.6 2 1.6 3 12.9 Nonhaemolytic 4 2.2 5 4.0 6 2.6 7 4.5 8 3.0 9 8.5 10 13.0 11 6.5 12 43.2 13 8.0 14 14.0 15 14.8 16 24.7 17 42.7 18 43.6 HBV pts
W0
1 2 3 4 5 6 7 8 9 10
4.3 29.6 3.3 11.8 16.5 10.8 15.8 19 20.1 72.5
W4
W4-W0
1.1 0.8 3.5
61.1 19.1 28.8 7.3 29.5 8.2
54.5* 27.2* 16.6*
6.0 6.8 2.5 4.0 1.6 6.6 10.7 4.0 39.4 1.2 4.0 1.1 4.4 12.9 9.7
3.8 2.8 0.1 0.5 1.4 1.9 2.3 2.5 3.8 6.8 10.0 13.7 20.3 29.8 33.9
0.0 0.0 0.0 0.0 0.0 5.6 2.2 2.6 9.1 4.4 0.0 4.6 7.9 3.8 7.9
2.7 0.8 1.1 1.4 1.4 2.6 3.3 0.0 4.6 1.1 1.2 0.2 1.8 1.6 2.3
W4
2.4 7.1 0.8 3.1 0.9 1.8 1.2 7.0 4.7 13.9
12.6 28.8 0.8 5.5 5.5 0.8 3.9 4.1 0.2 5.3
W4-W0
6.4 5.9 0.1 1.8 2.3 0.9 1.6 1.2 0.1 1.0
8.3 0.8 2.5 6.3 11.0 10.0 11.9 14.9 19.9 67.2
HBV, hepatitis B virus; HCV, hepatitis C virus. For each patients, three independent experiments were set up; values are expressed as means SD. *P < 0.05 by Student’s t-test. expression of these two genes increased at week 4 of treatment. The same was true in HBV patients, thus confirming an increase in undifferentiated colonies in both treatment groups (c-globin: 2.2-fold increase in haemolytic HCV, 3 in nonhaemolytic HCV and 2.7 in HBV; GATA2: 1.2-fold increase in haemolytic HCV, 1.4 in nonhaemolytic HCV and 1.6 in HBV).
In vitro results The ex vivo demonstrations of a delay in erythroid progenitor differentiation and maturation prompted us to evaluate © 2013 John Wiley & Sons Ltd
Fig. 1 Colony morphology. Percentage of different types of colonies (BFU-E and CFU-GEMM) derived from peripheral blood mononuclear cells of hepatitis C virus (HCV) (haemolytic and nonhaemolytic) and hepatitis B virus (HBV) patients at baseline and week 4 of treatment. Values are means of each group patients. *P < 0.05.
the effect of pegIFN and RBV in an in vitro culture system. Liquid cultures of CD34+ human stem cells isolated from peripheral blood of healthy volunteers and stimulated towards erythroid differentiation were set up. Different concentrations of pegIFN (100U and 1000U) and RBV (0.1, 0.5 and 1 mM), which were chosen on the basis of previous in vitro studies [8,13], were added at day 0 on isolated CD34+ cells. Cell viability and growth, erythroid differentiation and haemolysis were analysed under drug influence in vitro. Ribavirin inhibits erythroid cell proliferation Pegylated interferon 100U and 1000U had no significant effects on cell proliferation (Fig. 2a) and viability (Fig. 2b). RBV significantly inhibited cell proliferation (Fig. 2a). 1 mM RBV reduced cell growth by 49-fold compared with untreated cells (P < 0.004), with a significant reduction in cell viability also (P = 0.01) (Fig. 2b). Lower concentrations of RBV (0.1 and 0.5 mM) decreased cell growth by 7-fold (P < 0.05) without significantly affecting cell viability. Higher RBV doses (1 mM) were excluded from further analysis due to toxic effect. Pegylated interferon and ribavirin delay erythroid differentiation at different maturation steps The effect of pegIFN and BV on erythroid differentiation was evaluated by analysing glycophorin A (GPA), cell morphology and primitive erythroid-specific c-globin and GATA2 gene expression at day 14 of culture. Figure 3a,b shows the cytofluorimetric expression of GPA, a specific marker of erythroid maturation. In physiological conditions, after 2 weeks of culture, more than 70% of cells were GPA+. Concentrations of 100U and 1000U of PegIFN
420
L. Ronzoni et al.
Fig. 2 Growth kinetics and viability of erythroid progenitor cells treated with ribavirin (RBV) or pegylated interferon (pegIFN). CD34 + cells derived from peripheral blood of healthy volunteers were grown in the presence of RBV (0.1, 0.5 and 1 mM) or pegIFN (100 and 1000U). (a) Cell numbers of three representative experiments determined at regular time intervals are shown as means SD. (b) Viability of control and treated cells of three independent experiments are reported as means SD. *P < 0.005; ‡P < 0.05.
did not affect GPA expression, which instead was significantly reduced by RBV 0.1 mM and 0.5 mM (26% and 27%, respectively, P < 0.05). Figure 3c shows the results of the morphological studies of erythroid differentiation. Physiological erythroid differentiation includes maturation from proerythroblasts to the mature orthochromatic erythroblasts, with intermediate steps including basophilic and polychromatophilic erythroblasts. In physiological conditions, at day 14 of culture, the majority of cells are orthochromatic erythroblasts (52%), including 20% polychromatophilic and basophilic erythroblasts and almost no proerythroblasts (only 0.5%). Both pegIFN and RBV led to delayed erythroid differentiation, increasing the percentage of undifferentiated cells and consequently decreasing the percentage of more mature erythroblasts. Concentrations of 100U and 1000U of pegIFN increased proerythroblast percentage from 0.5% of untreated cells to 11.5% and 10.5% (P < 0.05), respectively, and reduced orthochromatic erythroblasts from 52% to 37% and 41%, respectively. Concentrations of 0.1 and 0.5 mM of Rbv significantly increased the percentage of polychromatophilic erythroblasts from 21% in untreated cells to 41% and 57%, respectively (P < 0.03), while decreasing the percentage of orthochromatic erythroblasts from 52% to
Fig. 3 Ribavirin and pegylated interferon (pegIFN) effect on erythroid differentiation. (a and b) Glycophorin A expression of control and treated cells at day 14 of culture analysed by flow cytometry. Means SD of three independent experiments are presented. *P < 0.05. Histograms represent the MFI relative to the cell number. (c) Aliquots of cells at day 14 of culture were subjected to cytocentrifugation and staining with benzidine and histological dyes and analysed for the proportion of proerythroblasts, basophilic, polychromatophilic and orthochromatic erythroblast, erythrocytes and nonerythroid cells. Means of three independent experiments are reported. (d) GATA2 and c-globin gene expression profile of control and treated cells at day 14 of culture. Results are expressed as means SD of three independent experiments.
44% and 23%, respectively. Figure 3d shows the expression of the primitive erythroid-specific c-globin and GATA2 genes at day 14 of culture. Physiologically, these genes are © 2013 John Wiley & Sons Ltd
Ribavirin effects on erythropoiesis more expressed in undifferentiated cells (proerythroblasts), while their expression declines during maturation. PegIFN increased both c-globin and GATA2 gene expression (1.6and 1.4-fold increase, respectively, with 100U pegIFN; 1.8and 1.5-fold increase, respectively, with 1000U pegIFN) as a consequence of increased numbers of undifferentiated proerythroblasts. RBV increased GATA2 gene expression (4.3-fold increase) without affecting c-globin gene expression, suggesting that RBV affects mainly polychromatophilic erythroblasts not expressing c-globin genes. These data are consistent with an inhibitory effect of both pegIFN and RBV on erythroid differentiation, however, at different time points of the physiological maturation process.
DISCUSSION The present study challenges the paradigm that RBVinduced anaemia is a mere consequence of extravascular haemolysis. Our in vivo and in vitro findings, in fact, directly correlate anaemia with RBV-induced suppression of erythroid differentiation and proliferation. In the in vivo part of our study, only 17% of patients exposed to pegIFN and RBV actually showed biochemical evidence of haemolysis, which was confirmed by an increase in colony numbers after 4 weeks of treatment compared with baseline values in the ex vivo experiments. Consistently with these findings, the remaining HCV patients (83%) did not have serum markers of peripheral haemolysis, while also showing in the ex vivo analysis a reduction in colony numbers with a relative increase in undifferentiated colonies (CFU-GEMM), a pattern consistent with suppression of erythroid proliferation and differentiation. The suppressive effect of pegIFN/RBV on erythroid differentiation was also confirmed by the increased expression of c-globin and GATA2 genes that are typical of the early stages of differentiation, which was observed after 4 weeks of anti-HCV therapy. Interestingly, the prominent role of RBV in this process is clarified by the in vitro demonstration of a significant inhibition of cell proliferation, as well as a delay in erythroid differentiation, once RBV was added to CD34+ human stem cells isolated from the peripheral blood of healthy volunteers who were stimulated towards erythroid differentiation. Concentrations of 0.1 and 0.5 mM of RBV significantly increased the percentage of polychromatophilic erythroblasts from 21% in untreated cells to 41% and 57%, respectively, while they decreased the percentage of orthochromatic erythroblasts from 52% to 44% and 23%, respectively. While our study confirms the well-known pegIFN myelosuppressive effect, as concentrations of 100U and 1000U of pegIFN increased proerythroblast percentage from 0.5% in untreated cells to 11.5% and 10.5%, respectively, and reduced orthochromatic erythroblasts from 52% to 37% and 41%, respectively, it also highlights that pegIFN and RBV act at different points of the erythroid differentiation process. PegIFN increased both c-globin and GATA2 gene © 2013 John Wiley & Sons Ltd
421
expression as a consequence of increased numbers of undifferentiated proerythroblasts, while RBV increased GATA2 gene expression without affecting c-globin gene expression, suggesting that RBV affects more mature forms of erythroid precursors such as polychromatophilic erythroblasts that do not express c-globin genes. Although the myelosuppressive effect of IFN is not a novelty, our study is the first to show RBV to directly suppress erythroid differentiation and proliferation. Given the multiple functions of RBV, several mechanisms through which RBV alone determines erythroid suppression may be accounted for. RBV is known to inhibit the IMPDH, which catalyses the oxidation of inosine 5 monophosphate to xanthosine 5 monophosphate, leading to a concomitant reduction in nicotinamide adenine dinucleotide to thio-nicotinamide adenine dinucleotide. This reaction is regulating the rate-limiting step of guanine nucleotide biosynthesis, whose inhibition apparently explains the negative effect of RBV on cell proliferation [12,15]. The latter can also be inhibited by the tumour suppression activity of the p53 gene whose antiproliferative activity can be primed by RBV [13]. Owing to the fact that IMPDH activity is downregulated by p53 [16], the p53-RBV interactions can partly account for IMPDH inhibition by RBV. Finally, RBV may act synergistically with IFN in upregulating IFN-stimulated genes causing enhanced expression of cell genes involved in IFN signalling pathways [11,13]. From a clinical standpoint, our observations may have practical implications also with respect to the therapeutic algorithms of HCV protease inhibitors in combination with pegIFN and RBV, which bear a higher risk of anaemia compared with dual pegIFN and RBv therapy [17–20]. To minimize the risk of anaemia, the concomitant use of drugs with known myelosuppressive activity should be avoided; moreover, management of treatment-related anaemia through blood transfusions should be generally discouraged as transfused red blood cells will only transiently increase Hb values while potentially further inhibiting bone marrow function. Based on our findings, stepwise RBV down dosing and eventual concomitant pegIFN down dosing should be the standard of care management of treatment-related anaemia. Indeed, given that the two drugs act on different points of erythroid differentiation, a synergistic interaction may potentiate the observed haemoglobin drop in most HCV patients. This was the case in a Phase II study where RBV was combined with an NS5B polymerase inhibitor with or without pegIFN. The mean haemoglobin decline was far less pronounced in the pegIFN-free regimen as compared to the pegIFN/RBV/NS5B polymerase inhibitor arm, demonstrating the key role of pegIFN in treatmentrelated anaemia [21]. While our study shows that in the majority of patients anaemia is likely to be a direct consequence of RBV- and pegIFN-induced suppression of erythroid differentiation and proliferation, our findings do not completely negate RBV-
422
L. Ronzoni et al.
induced haemolysis. Indeed, a subgroup of HCV patients had peripheral signs of haemolysis that was confirmed by an increase in erythroid progenitor colony number. Moreover, from a clinical standpoint, these patients were the most difficult to manage as they were those with a more rapid and pronounced haemoglobin decline at week 4 of pegIFN/RBV therapy. Still, the in vitro section of our study could not replicate previous studies on RBV-induced haemolysis. One of the possible explanations is that in previous in vitro studies [8], RBV was dosed at 1 mM, which in the present study and in another concomitant study [13], led to cell death as assessed by reduced cell viability. We do acknowledge that our study has the limitation that RBV monotherapy was studied only in vitro; however, in clinical practice, RBV monotherapy is currently not feasible because RBV is active against HCV in combination with IFN or DAA only. Another limitation is the small sample size of the study which prevented the identification of pretreatment predictors of haemolysis or severe bone marrow suppression during pegIFN plus RBV therapy. Indeed, the three patients with the haemolytic pattern did not differ in terms of age, kidney function or ITPA genotype (data not shown) from the nonhaemolytic patients. Likewise with previous studies, we could not correlate the level of colony-forming erythroid progenitor cells with the development and severity of anaemia or with the haemolytic or myelosuppressive pattern of anaemia at the individual level [22]. Lastly, we did not assess the ITPA genotype in healthy volunteers, as the role of ITPA in treatment-
induced anaemia had not yet been discovered at the time of the study design. These caveats notwithstanding, our study is the first to demonstrate a direct myelosuppressive effect of RBV on red blood cells proliferation and differentiation, shedding further lights on the multiple mechanisms of action of this drug that since its discovery remains a fundamental backbone of anti-HCV therapy.
STATEMENT OF INTERESTS Alessio Aghemo has served as a speaker for Roche and Janssen and has received research funding from Roche and Gilead Sciences. Maria Grazia Rumi has served as an advisory board member for Tibotec, Roche and Janssen Cilag. Massimo Colombo has served as a speaker for Tibotec, Roche, Novartis, Bayer, BMS, Gilead Science, Vertex and as an advisory board member for Merck, Roche, Novartis, Bayer, BMS, Gilead Science, Tibotec, Vertex, Achillion; he has received research funding Merck, Roche, BMS, Gilead Science. Maria Domenica Cappellini has served as an advisory board member for Novartis, Genzyme, Shire. Other authors have no financial disclosure to declare. This study was funded in part by a grant from the Universit a degli Studi di Milano, Milan, Italy (FIRST 2009) to MC; a PRIN grant from ‘Ministero Italiano dell’Universit ae della Ricerca’ (PRIN 2008) to MDC; and grants from Fondazione IRCCS C a Granda, Ospedale Maggiore Policlinico, Milan, Italy (RC 2009 and RC 2010) to MDC.
REFERENCES 1 Ghany MG, Nelson DR, Strader DB, Thomas DL, Seeff LB. An update on treatment of genotype 1 chronic hepatitis C virus infection: 2011 practice guideline by the American Association for the Study of Liver Diseases. Hepatology 2011; 54: 1433–1444. 2 European Association for the Study of the Liver. EASL clinical practice guidelines: management of hepatitis C virus infection. J Hepatol 2011; 55 (2): 245–264. 3 Feld JJ, Hoofnagle JH. Mechanism of action of interferon and ribavirin in treatment of hepatitis C. Nature 2005; 436(7053): 967–972. 4 Feld JJ, Lutchman GA, Heller T et al. Ribavirin improves early responses to peginterferon through improved interferon signaling. Gastroenterology 2010; 139(1): 154–162. 5 Hezode C, Forestier N, Dusheiko G et al. Telaprevir and peginterferon
with or without ribavirin for chronic HCV infection. N Engl J Med 2009; 360(18): 1839–1850. 6 Aghemo A, Rumi MG, Monico S et al. Ribavirin impairs salivary gland function during combination treatment with pegylated interferon alfa-2a in hepatitis C patients. Hepat Mon 2011; 11(11): 918– 924. 7 Sulkowski MS, Cooper C, Hunyady B et al. Management of adverse effects of Peg-IFN and ribavirin therapy for hepatitis C. Nat Rev Gastroenterol Hepatol 2011; 8(4): 212– 223. 8 De Franceschi L, Fattovich G, Turrini F et al. Hemolytic anemia induced by Ribavirin therapy in patients with chronic hepatitis C virus infection: role of membrane oxidative damage. Hepatology 2000; 31: 997–1004.
9 Fellay J, Thompson AJ, Ge D et al. ITPA gene variants protect against anaemia in patients treated for chronic hepatitis C. Nature 2010; 464(7287): 405–408. 10 Thompson AJ, Fellay J, Patel K et al. Variants in the ITPA gene protect against ribavirin-induced hemolytic anemia and decrease the need for ribavirin dose reduction. Gastroenterology 2010; 139(4): 1181–1189. 11 Feld JJ, Nanda S, Huang Y et al. Hepatic gene expression during treatment with peginterferon and ribavirin: identifying molecular pathways for treatment response. Hepatology 2007; 46(5): 1548– 1563. 12 Jayaram HN, Cooney DA, Gruseh M, Krupitza G. Consequences of IMP dehydrogenase inhibition and its relationship to cancer and apop-
© 2013 John Wiley & Sons Ltd
Ribavirin effects on erythropoiesis
13
14
15
16
tosis. Curr Med Chem 1999; 6: 561– 574. Su WC, Liu WL, Cheng CW et al. Ribavirin enhances interferon signaling via stimulation of mTOR and p53 activities. FEBS Lett 2009; 583 (17): 2793–2798. Ronzoni L, Bonara P, Rusconi D, Frugoni C, Libani I, Cappellini MD. Erythroid differentiation and maturation from peripheral CD34+ cells in liquid culture: cellular and molecular characterization. Blood Cells Mol Dis 2008; 40(2): 148–155. Wang W, Hedstrom L. Kinetic mechanism of human inosine 5-monophosphate dehydrogenase type II: random addition of substrates and ordered release of products. Biochemistry 1997; 36: 8479–8483. Liu Y, Bohn SA, Sherley JL. Inosine5-monophosphate dehydrogenase is
423
a rate determining factor for p53dependent growth regulation. Mol Biol Cell 1998; 9: 15–28. 17 Kwo PY, Lawitz EJ, McCone J et al. Efficacy of boceprevir, an NS3 protease inhibitor, in combination with peginterferon alfa-2b and ribavirin in treatment-naive patients with genotype 1 hepatitis C infection (SPRINT-1): an open-label, randomised, multicentre phase 2 trial. Lancet 2010; 376: 705–716. 18 Jacobson IM, McHutchison JG, Dusheiko G et al. Telaprevir for previously untreated chronic hepatitis C virus infection. NEJM 2011; 364: 2405–2416. 19 Poordad F, McCone J Jr, Bacon BR et al. Boceprevir for untreated chronic HCV genotype 1 infection. N Engl J Med 2011; 364: 1195– 1206.
20 Bacon BR, Gordon SC, Lawitz E et al. Boceprevir for previously treated chronic HCV genotype 1 infection. N Engl J Med 2011; 364: 1207–1217. 21 Gane EJ, Stedman CA, Hyland RH et al. Once daily PSI-7977 plus Rbv: pegylated interferon-alfa not required for complete rapid viral response in treatment-naive patients with HCV GT2 or GT3 [Abstract]. Hepatology 2011; 54(4): 76A. 22 Brunati C, Cappellini MD, De Feo T et al. Uremic inhibitors of erythropoiesis: a study during treatment with recombinant human erythropoietin. Am J Nephrol 1992; 12(1– 2): 9–13.
Figure S1. Quantitative hemoglobin reduction from the baseline in HCV and HBV patients.
Figure S2. Quantitative hemoglobin reduction from the baseline in hemolytic and non-hemolytic HCV patients and in HBV patients.
SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article: Table S1. Baseline characteristics of patients with chronic hepatitis C (HCV) and B (HBV).
© 2013 John Wiley & Sons Ltd
