B cell activating factor (BAFF) in the natural history of chronic hepatitis C virus liver disease and mixed cryoglobulinaemia

Authors


Correspondence: G. Lake-Bakaar, Liver Center, Beth Israel Deaconess Medical Center, Harvard Medical School, 110 Francis Street, LMOB 7A-055, Boston, MA 02215, USA.

E-mail: glakebak@bidmc.harvard.edu

Summary

B cell activating factor (BAFF) plays a crucial role in the process of development, maturation and activation of B lymphocytes. Chronic hepatitis C virus (HCV) infection is characterized by multiple B cell disorders. It is a major cause of type II mixed cryoglobulinaemia (MC). We measured serum BAFF levels in several clinical situations to elucidate the potential role of BAFF in chronic HCV infection. We used a commercially available solid phase enzyme-linked immunosorbent assay. We estimated serum BAFF in stored sera from uninfected controls (n = 8), patients with chronic hepatitis B virus infection HBV (n = 5) and chronic HCV infection with (n = 16) and without mixed cryoglobulinaemia (n = 14). In two patients with HCV and MC we correlated BAFF with HCV RNA after pegylated interferon (peg-I). We correlated serum BAFF levels at baseline and at 12 weeks with treatment response: sustained virological response SVR (n = 5), non-responders (n = 6) and relapsers (n = 2). Finally, we estimated BAFF levels after complete depletion of B cells with rituximab in patients with chronic HCV with MC (n = 3). Serum levels of BAFF were increased in chronic HCV with MC, but not in chronic HBV infection, suggesting an association between BAFF and cryoglobulinaemia. Peg-I increased BAFF levels in serum and this paralleled HCV RNA very closely. Serum BAFF levels at week 12 of therapy with peg-I and R were significantly higher in responders than non-responders. Finally, B cell depletion was associated with markedly increased levels of BAFF.

Introduction

The B cell activating factor (BAFF), also known as BLyS, a member of the tumour necrosis factor (TNF) superfamily, is central to the maturation and survival of peripheral blood B lymphocytes and is crucial to their development and activation. This process involves both the deletion of self-reactive B cells and the selective favouring of immunologically useful clones. In the absence of BAFF, B cell survival and maturation are defective. By contrast, elevated levels of BAFF subvert B cell self-tolerance and rescue weakly autoreactive B cells that would normally be deleted. Autoreactive B cells are more dependent on BAFF for survival than alloreactive B cells [1-3].

Over-expression of BAFF in mice and in humans is associated with autoimmune disorders such as Sjögren's syndrome (SS) and systemic lupus erythematosus (SLE) [4].

Chronic hepatitis C virus (HCV) infection is characterized by multiple B cell disorders. It is a major cause of type II mixed cryoglobulinaemia (MC), which is characterized by a monoclonal or oligoclonal proliferation of immunoglobulin (Ig)M-producing B cells. The IgM antibodies display rheumatoid factor (RF) activity and are autoreactive against IgG anti-HCV antibodies. The MC cryoprecipitates comprise monoclonal IgM-RF antibodies in complex with anti-HCV antibodies attached to hepatitis C virions and lipoproteins [5].

A common hypothesis for these HCV-related changes is chronic antigenic stimulation, which is thought to facilitate clonal B lymphocyte expansion [6], and can lead to non-Hodgkin's B cell lymphomas in 5–10% of patients [7]

It is unclear why cryoglobulins appear in just under half of all HCV patients, and why other chronic viral infections appear to be associated less frequently with lymphoproliferation [6]. It is possible that factors such as cytokines, growth factors and host genetic background may also link chronic HCV infection and lymphoproliferation [8, 9].

Effector signals such as BAFF that enhance survival of B lymphocytes may provide such a link. BAFF promoter polymorphisms may also be a contributor to MC pathogenesis. The frequency of the −871T allele is greater in HCV patients with than without MC [10, 11].

Therefore, in this preliminary study, we measured serum BAFF levels in clinical situations that could help to elucidate potential pathophysiological roles for BAFF in chronic HCV infection. The results remain to be confirmed in larger studies.

Methods

Human BAFF in serum was measured using a commercially available solid phase enzyme-linked immunosorbent assay enzyme-linked immunosorbent assay (ELISA) (Quantikine, R&D Systems, Minneapolis, MN, USA), according to the manufacturer's instructions. This assay was designed to eliminate interference from biological factors, including cryoglobulins. The assay provided sensitive and reproducible results within the range of values measured in our patients.

HCV RNA levels in serum were estimated using bDNA technology through a commercial laboratory (Bioreference Labs, Elmwood Park, NJ, USA).

Patients

BAFF and chronic HCV ± MC

Serum BAFF levels were measured in stored sera from a broad group of patients that included healthy controls (n = 8), patients with chronic HBV, chronic hepatitis B (CHB) infection (n = 5) and patients with chronic HCV infection with positive HCV antibody and HCV RNA (n = 30). The chronic HCV-infected patients were assigned to two groups: patients without evidence of MC (n = 14) and patients with clinical and or biochemical evidence of the mixed cryoglobulinaemia syndrome (n = 16) (Table 1). The diagnosis of MC syndrome was based on the presence of arthralgia, skin purpura and asthaenia, with or without HCV-induced systemic vasculitis, defined by the presence of symptoms such as purpura, vasculitic skin lesions and neuropathy with reduced levels of either C3 or C4. In several cases, cryoglobulinaemia was usually detected only after repeated testing. Serum samples from patients with significant co-morbidities, including autoimmune disorders or other disease states that could influence BAFF levels, were excluded. The majority of the patients were symptomatic at the time of presentation, when the serum samples were collected. Immunosuppressive treatment was avoided as much as possible in these patients with chronic HCV infection and none was receiving treatment at the time of their presentation.

Table 1. Serum B cell activating factor (BAFF) levels in controls, chronic hepatitis B (CHB), chronic hepatitis C virus (HCV) with and without mixed cryoglobulinaemia (MC).
Patient populationSerum BAFF (median and confidence interval)
Controls (n = 8)814 pg/ml (136·5–1466·5)
Chronic HBV (n = 5)377 pg/ml (207–512)
Chronic HCV naive without cryoglobulinaemia (n = 14)823·6 pg/ml (613–1017·2)
Chronic HCV with cryoglobulinaemia (n = 16)1694 pg/ml (1116–8461)
Chronic HCV all (n = 30)1066 pg/ml (801·5–1616·5)

The characteristics of our chronic HCV-infected patients with cryoglobulinaemia are shown in Table 2. The majority were females, aged more than 50 years with advanced hepatic fibrosis.

Table 2. Characteristics of patients with chronic hepatitis C virus (HCV) and cryoglobulinaemic syndrome with or without neuropathy or skin vasculitis.
Sex (female)9 (56%)
  1. GGT, gamma-glutamyl transpeptidase; LFT, liver function test.
Age > 50 years11 (74%)
Auto immune including diabetes4 (25%)
Renal insufficiency1 (6%)
Abnormal LFTs6 (38%)
Abnormal GGT8 (50%)
Rheum factor positive7 (45%)
Abnormal C3, C4 or CH5011 (71%)
Advanced fibrosis F2–F412 (73%)
HCV genotype 114 (87%)

The acute effect of interferon (IFN) on BAFF

Serum levels of BAFF were determined at short intervals after an initial dose of pegylated IFN-α2a in two patients with chronic HCV infection. The levels were correlated with serum HCV RNA concentrations measured in simultaneously obtained samples.

BAFF and response to chronic IFN and ribavirin therapy in chronic HCV patients

We estimated serum BAFF levels in patients with chronic HCV infection on standard treatment with pegylated IFN (180 μg by weekly subcutaneous injections) and ribavirin (weight-based and genotype-based, ranging between 800 and 1200 mg daily in divided doses). Levels were measured at baseline, after week 12 and after 24 or more weeks after discontinuation of treatment with IFN. The patients were divided into three groups, based on their response to treatment: sustained virological responders (SVR) (n = 5), non-responders (NR) (n = 6) and relapsers (n = 2).

BAFF and rituximab

Serum BAFF levels were determined in stored sera from three patients with chronic HCV infection and mixed cryoglobulinaemia, 8 or more weeks after depletion of B cells with rituximab. Complete depletion of CD20+ B cells in peripheral blood was demonstrated by fluorescence activated cell sorter (FACS) analysis.

Statistical analysis

Results are expressed as median (range) except where indicated. Continuous variables were compared using analysis of variance (anova) and a P-value of <0·05 was considered statistically significant. All statistical analyses were performed using NCSS statistical software (Kaysville, UT, USA).

Results

BAFF levels in patients with chronic HBV and HCV infection

Serum BAFF levels in controls, chronic HBV- and chronic HCV-infected patients with and without cryoglobulinaemia are shown in Fig. 1. BAFF levels did not differ significantly between controls and CHB patients. By contrast, chronic HCV-infected patients had significantly higher levels of BAFF. However, this difference was due largely to patients with cryoglobulinaemia. BAFF levels in chronic HCV patients without cryoglobulinaemia were not different from controls or CHB patients.

Figure 1.

Serum B cell activating factor (BAFF) levels in controls, chronic hepatitis B virus [(HBV)-infected], hepatitis C virus (HCV)-all (chronic HCV-infected patients with and without cryoglobulinaemia combined), HCV (chronic HCV without cryoglobulinaemia) and HCV-cryo (chronic HCV with cryoglobulinaemia).

BAFF levels after pegylated IFN stimulation

Serum levels of BAFF were determined at short intervals after an initial dose of pegylated IFN in two patients with chronic HCV. After an initial drop at 12 h, BAFF increased progressively up to threefold by day 7 (Fig. 2). HCV RNA levels measured concomitantly in the serum samples closely paralleled BAFF levels. Changes in B cell levels were not determined. The chronology of changes in BAFF and HCV RNA were similar in the two patients.

Figure 2.

B cell activating factor (BAFF) and hepatitis C virus (HCV) RNA levels after pegylated interferon stimulation. BAFF levels closely mirrored HCV RNA level after pegylated interferon.

BAFF and response to IFN and ribavirin treatment in chronic HCV

In SVR (n = 6), BAFF levels rose significantly from a baseline of 774 ± 242 [mean ± standard deviation (s.d.)] pg/ml to 1322 ± 403 (P = 0·02) by week 12 and dropped back to 596 ± 262 pg/ml after sustained virological response was achieved (Fig. 3). By contrast, in NR (n = 6), the rise in BAFF levels from a baseline of 1306 ± 991 pg/ml to 1673 ± 1665 by week 12 of treatment, and to 1728 ± 1762 post-therapy were not significant. The changes in BAFF levels during and after treatment in relapsers (n = 2) were also not significant.

Figure 3.

B cell activating factor (BAFF) and response to interferon (IFN) and ribavirin in chronic hepatitis C virus (HCV)-infected patients. BAFF levels at baseline, after week 12 and following treatment discontinuation in patients with sustained virological response (SVR), non-response and relapsers. The increase in BAFF at week 12 was significant only in patients with SVR.

Effect of B cell depletion with rituximab on serum BAFF levels

In three HCV-MC patients, the mean BAFF level (mean ± s.d.) 8 weeks after rituximab (9981 ± 1734 pg/ml) was significantly higher than in our control group of untreated HCV-MC patients, 2927 ± 3464 pg/ml (P = 0·0009). CD20+ B cell were depleted completely in peripheral blood at week 8. B cell depletion with rituximab in peripheral blood is associated with a more than threefold increase in levels of BAFF (Table 3).

Table 3. Effect of B cell depletion with rituximab on serum B cell activating factor (BAFF) levels.
Patient populationSerum BAFF mg/ml (median and confidence interval)
HCV with cryoglobulinaemia without rituximab (n = 16)1 694 pg/ml (1 116–8 461)
HCV with cryoglobulinaemia after rituximab (n = 3)9 796 pg/ml (8 348–11 801·5)

Discussion

The development and maturation of B lymphocyte cells require deletion of self-reactive B cells that may be potentially harmful. Elevated levels of BAFF subvert B cell self-tolerance by rescuing and increasing the survival rate of weakly autoreactive B cells that would be deleted normally. Over-expression in mice and in humans is associated with autoimmune disorders such as SS and SLE [12]. Type II mixed cryoglobulinaemia is an autoreactive disorder. It is characteristic of chronic HCV infection, but is uncommon in CHB. It is the result of a monoclonal or oligoclonal proliferation of B cells [13, 14], which produce IgM autoantibodies that are autoreactive with anti-HCV antibodies. Type II mixed cryoglobulins (MC-II) are complexes comprising HCV particles and lipoproteins with these antibodies [13].

Our study compared BAFF levels between patients with chronic HCV (with and without MC), uninfected controls and CHB patients. BAFF levels were elevated significantly in chronic HCV with MC, but not in chronic HCV without MC or chronic HBV patients. These data raise the possibility that elevated levels of BAFF in chronic HCV with MC maintain survival of autoreactive IgM antibody-producing B cells that would normally be deleted.

Others have reported similar elevations of serum levels of B lymphocyte stimulator (BLys/BAFF) in chronic HCV with MC [5, 15, 16]. The majority of these studies demonstrated higher serum BAFF levels in chronic HCV with MC patients compared to patients with chronic HBV infection. One study also demonstrated increased BAFF in chronic HCV in the absence of MC.

There also appeared to be a difference between type II and type III cryoglobulinaemia, with higher levels of BAFF in type II than in type III [5].

It is not clear why BAFF levels should be higher in chronic HCV than with CHB. Both are chronic viral infections with high levels of daily virion production. Clearance of these high levels of virus from serum presumably exerts significant pressure on B cells turnover. However, virus production in chronic HCV exceeds that in CHB by up to two orders of magnitude [17, 18]. The mutation rate of hepatitis B surface antigen is also much lower [19] compared to the highly mutable hepatitis C virus [20, 21]. It is likely that the impact of chronic HCV on B cells is both quantitatively and qualitatively greater than CHB. The higher numbers and diversity of B cells in chronic HCV infection might require higher levels of BAFF for maturation, activation and survival.

BAFF levels are also higher in patients with chronic HCV with cryoglobulinaemia than with chronic HCV without cryoglobulinaemia. The polymorphism −871 C/T has been described in the BAFF promotor. The presence of the mutated −871T genotype has been associated with higher BAFF mRNA levels in monocytes [22-24]. More recently, others have noted a higher prevalence of T allele homozygosity in a group of HCV patients with MC compared to HCV-infected patients without MC [10, 11]

BAFF is expressed locally in myeloid cells including macrophages and dendritic cells following activation by cytokines, such as IFNs and interleukin (IL)-10 [25-27]. IFN has been shown to stimulate BAFF secretion by myeloid cells in animals [28] and IFN-β stimulates BAFF in multiple sclerosis in man [29-31]. Treatment of chronic HCV with MC-related vasculitis with IFN has been associated with either the development or exacerbation of cutaneous vasculitis [32-34]. This might be explained by an effect of IFN on BAFF secretion with consequent effects on autoreactive B cells.

We investigated the acute effects of pegylated IFN-α on serum BAFF levels in two chronic HCV patients without MC. In both patients, there was an initial reduction in serum BAFF levels 12 h after pegylated IFN, followed by a progressive increase over the next 7 days. These data suggest that IFN-α increases BAFF levels in man and could account for the exacerbation in MC symptoms that have been described in patients treated with IFN. It might also begin to explain the superiority of combination treatment with rituximab, pegylated IFN and ribavirin over pegylated IFN and ribavirin alone in the treatment of chronic HCV-MC [35].

The demonstration that IFN stimulates BAFF secretion in patients with chronic HCV could provide clues on MC development in chronic HCV. Intercurrent infections in chronic HCV patients may lead to increased local cytokine expression and activation of macrophages and dendritic cells that express BAFF [25-27], thus providing a link between chronic HCV and chronic HCV with MC.

We noted a very striking correlation between changes in BAFF levels and in HCV RNA induced by IFN. There is no clear explanation for these findings, which need to be confirmed in a larger study. The effects of IFN on HCV may share a common link with its stimulation of BAFF. Alternatively, a direct effect of HCV on BAFF-secreting cells could account for similar findings. A recent study described higher BAFF levels in patients with acute hepatitis C that progressed to chronicity [36]. The higher hepatitis C viral loads that characterize persistent infection more probably reflect lower endogenous IFN expression. This argues for direct HCV stimulation of BAFF.

Current treatment of chronic HCV liver disease includes therapy with IFN and ribavirin. Approximately 50% of treated patients develop an SVR [37]. In three groups of patients who had been treated with IFN and ribavirin, we noted a significant rise in BAFF at week 12 of treatment in the group that achieved SVR, but not in the NR group or in the two patients who had responded initially, but relapsed when therapy was stopped. This is a new observation that has not been reported previously and deserves further study. We can only speculate about the inability of IFN and ribavirin to stimulate BAFF in NR. However, measurement of BAFF levels in patients after IFN therapy might prove useful as an early predictive marker of response.

BAFF is crucial for the survival and maturation of B cells, and may be associated with qualitative and quantitative control of peripheral blood B cell numbers. In an attempt to investigate the possibility that BAFF plays a role in the control of B cell numbers in peripheral blood, we measured serum BAFF levels after depletion of B cells with rituximab. Rituximab is a chimeric monoclonal antibody directed against the CD20 antigen expressed on mature B cells and is used to deplete B cells in a number of clinical situations, including chronic HCV-related cryoglobulinaemia. We measured BAFF levels in three patients with chronic HC with MC that were treated with rituximab. BAFF levels after B cell depletion were threefold higher than in our control population of patients with chronic HCV and MC who had not received rituximab.

Increased BAFF levels after rituximab have been described in patients with rheumatoid arthritis, SS, SLE and after renal transplantation rejection [38-40]. This study confirms similar findings in patients with chronic HCV with MC. The increased levels in serum could reflect release of BAFF bound to lysed B cell membrane. However, measurements were performed several weeks after B cell depletion, making this unlikely. The data suggest more probably that BAFF is involved in the feedback control of B cell numbers in peripheral blood.

In summary, our studies suggest that BAFF might play an important role in the natural history of chronic HCV infection. Serum levels of BAFF are increased in chronic HCV with MC, but not in CHB infection, suggesting a role for BAFF in the development of the autoreactive B cells that produce cryoglobulins. IFN increases BAFF levels in serum and could account for exacerbation in MC symptoms during IFN treatment. Changes in the level of BAFF correlate very closely with changes in HCV RNA. The significance of the close relationship between HCV RNA and BAFF is unclear. The pattern of changes in BAFF levels in serum during therapy with IFN and ribavirin differs significantly between responders and non-responders. This could be an important early marker of end-treatment response. Finally, significantly higher levels of BAFF in serum after B cell depletion suggest that it may play a central role in the control of B cell numbers in peripheral blood. Thus, BAFF may be an important factor in the natural history of chronic HCV infection.

Disclosure

No disclosures.

Ancillary