Hepatitis C and non-Hodgkin lymphoma: The clinical perspective

Abstract

Hepatitis C virus (HCV) is a commonly transmitted infection that has both hepatic and extrahepatic repercussions. These range from the inflammatory to the oncologic with an undisputed link to hepatitis, liver cirrhosis, and hepatocellular carcinoma. Its role in the development of B cell non-Hodgkin lymphoma (B-NHL) is becoming better understood, leading to opportunities for research, therapy, and even prevention. Research in the field has progressed significantly over the last decade, with the number of patients diagnosed with HCV and B-NHL rising incrementally. It is therefore becoming crucial to fully understand the pathobiologic link of HCV in B cell lymphomagenesis and its optimal management in the oncologic setting. (HEPATOLOGY 2012)

Over 180 million people are infected with hepatitis C virus (HCV), accounting for 3% of the global population.1 HCV is well-recognized as a cause of hepatic disease and hepatocellular carcinoma, while its hematologic manifestations (mixed cryoglobulinemia [MC] and B cell non-Hodgkin lymphoma [B-NHL]) are less appreciated.2 Clarifying the oncogenic role of HCV has been hampered by difficulties in viral culture, the heterogeneity of B-NHL, and the differences between viral genotypes, prevalence, and affected populations. This is further complicated by the host environment and the availability of effective therapy. Nevertheless, a molecular and clinical link between HCV and B-NHL has been made placing an emphasis on cause and effect rather than simple association.

Abbreviations

BCR, B cell receptor; B-NHL, B cell non-Hodgkin lymphoma; DLBCL, diffuse large B cell lymphoma; HCV, hepatitis C virus; IFN, interferon; MALT, mucosa-associated lymphoid tissue lymphoma; MC, mixed cryoglobulinemia; MZL, marginal zone lymphoma; OR, odds ratio; OS, overall survival; RIBA, recombinant immunoblot assay; WM, Waldenström's macroglobulinemia.

Mechanisms of Lymphoproliferation

Sung et al.2 established a flourishing HCV-infected B-NHL line whose virions could infect hepatic cells, peripheral blood mononuclear cells, and B cells in vitro. Indeed, viral replication within “lymphoid reservoirs” is thought to be responsible for HCV persistence after apparently successful therapy: these reservoirs potentially function as a viral “storage facility” and allow for positive selection of different viral subtypes which may influence the natural history of infection.3, 4 Current evidence suggests there may be a predilection for B cells5 and that different viral strains may be more lymphotropic than others.6

The biologic rationale for a causal relationship between HCV and B-NHL is based upon epidemiologic data and clinical observation. One study showed evidence of oligoclonal and monoclonal B cell expansion in 100% of HCV-infected patients with a mixture of pathologies, including MC, Waldenström's macroglobulinemia (WM), and B-NHL.7 However, the mechanism by which HCV infection leads to B-NHL is still unclear.

The proposed mechanisms include direct infection of hematopoietic cells, antigen drive, and the “hit and run” hypothesis. Despite evidence that HCV can infect and replicate within lymphoid cells, there are few data demonstrating that HCV can induce malignant lymphoproliferation. For example, active viral replication in the peripheral blood and bone marrow of chronically infected patients with MC or B-NHL could not be demonstrated.8 Also, the identification of HCV-infected peripheral blood mononuclear cells was not a significant predictor of NHL risk when compared with serologic evidence of infection.9 However, a recent study found that peripheral blood B cells from chronic HCV patients were infected and also had enhanced gene expression associated with B cell NHL development when compared with healthy controls.10

Similar to the association of Helicobacter pylori and gastric mucosa-associated lymphoid tissue lymphoma (MALT) lymphoma, the concept of chronic antigen stimulation leading to a monoclonal malignant proliferation can also be applied to HCV. The HCV E2 viral protein binds to CD81 expressed on B lymphocytes11 and CD81 facilitates viral entry.12 Moreover, costimulation of the CD19/CD21/CD81 complex with the B cell receptor (BCR) can lower the B cell activation threshold and induce proliferation.13 CD81 is also up-regulated in HCV-infected and MC patients and increases with viral load.14 Therefore, B cells with anti-HCV surface immunoglobulins receive a strong proliferation signal through binding of the HCV-specific BCR and viral binding to CD81.15 Furthermore, experimental sequencing of clonal immunoglobulin variable regions from both MC and HCV-associated NHL patients shows restricted expression of V_H and V_L genes (V_H1-69 and VκA27) and evidence of somatic hypermutation, suggesting exposure and response to a common antigen.16 Such sequence analysis has allowed identification of premalignant oligoclonal cell populations in MC patients years before lymphoma development.17 Whether HCV is this common antigen has been demonstrated by research from Stanford School Medical Center. The group showed that both normal B cells and HCV-associated B-NHL preferentially expressed the V_H1-69 gene in response to E218 and that the BCRs from an HCV-associated B-NHL bound E2.19 This provides compelling evidence for the role of HCV and mechanism of antigen drive in B-NHL. This concept is already accepted in gastric MALT and Helicobacter pylori.20 However, despite differences in antigenic origin, the outcomes are similar: chronic B cell proliferation and malignant lymphomagenesis.

The jump from lymphoproliferation to malignancy may require a second “hit and run” transforming event such as the antiapoptotic Bcl-2 rearrangement. The translocation t(14;18) is significantly associated with chronic HCV infection,20 particularly in MC.21 Moreover, research has identified B cell clonal expansion with this translocation in MC22 and HCV-positive patients with MALT lymphoma.23 However, whether HCV is directly mutagenic or responsible for a clonal B cell population that becomes vulnerable to transforming mutations remains unclear.

Epidemiology of HCV and B-NHL

Epidemiologic studies have demonstrated a causal relationship between HCV and B-NHL (Table 1). However, the odds ratios are moderate (2-3 on average) in comparison to HCV and hepatocellular carcinoma. One meta-analysis reviewed data from 23 studies (4,049 NHL patients and 1,813,480 controls) and found a strong association (odds ratio [OR] 5.70).24 It should be noted that studies reporting a significant association have originated from countries with a high HCV prevalence, such as Italy,25 Egypt,26 and Japan,27 as opposed to low in Northern Europe, North America, and the United Kingdom.28 These findings echo the north-south divide in European HCV prevalence, with recent figures of 0.1%-1%, 0.2%-1.2%, and 2%-5%-3%-5% quoted for Northern, Central, and Southern Europe, respectively. Such differences are thought to reflect viral transmission trends, with intravenous drug use (IVDU) transmission among the younger population predominating in the North and iatrogenic transmission amongst the older population, followed by IVDU transmission in the young, in the South.29 One study of female patients in Connecticut found an increased risk of B-NHL despite the low prevalence, with an OR of 2.0.30 Therefore, despite the comparatively weak ORs, the accumulation of evidence has prompted a shift from association to causation. Recent epidemiologic evidence also suggests that genotype 2 may be more prevalent and carcinogenic in lymphoma.31

Table 1. Studies Demonstrating the Epidemiologic Link Between HCV and Indolent NHL

Study	Geographic Region	No. of Patients (Cases/Controls)	HCV RNA Prevalence (Cases/Controls)	Association of HCV with B-NHL by Subtype
Abbreviations: CI, confidence interval; FL, follicular lymphoma; LPL, lymphoplasmacytic lymphoma; MZL, marginal zone lymphoma; OR, odds ratio. * ORs provided for active HCV infection (HCV-RNA) and NHL major subtypes. ORs for past HCV infection (anti-HCV antibodies) and NHL major subtypes as follows: MZL, 3.4 (95% CI, 1.4-8.5; P = 0.0076); FL, 1.5 (95% CI, 0.6-3.8; P = 0.3928). † Prevalence among all NHL subtypes. ‡ Included small lymphocytic lymphoma/chronic lymphocytic leukemia, LPL, FL, MZL, and MALT. § Crude OR provided (not adjusted for possible confounding variables, including sex, age, level of education, and place of birth). Adjusted OR, 2.3 (95% CI, 1.3-44). || OR is provided for patients in whom HCV infection was defined as a positive enzyme-linked immunosorbent assay (ELISA) test and a positive or intermediate recombinant immunoblot assay (RIBA) for anti-HCV antibodies. No OR is provided for when HCV infection was more conservatively defined as a positive ELISA test and a positive RIBA or was restricted to individuals who tested positive for both anti-HCV antibodies and HCV RNA.
Goldman et al.26*	Egypt	486/786	26%-48%/27%†	MZL: OR, 4.4 (95% CI, 1.8-10.6)
				FL: OR, 3.3 (95% CI, 1.3-8.0)
Talamini et al.63	Italy	225/504	19.6%/8.9%	Low-grade B cell: OR, 3.21 (95% CI, 1.63-6.33)‡
Mele et al.34	Italy	400/396	17.5%/5.6%	Indolent: OR, 3.2 (95% CI, 1.8-5.8)§
Morton et al.30	United States	464/534	2%/1%	FL: OR, 4.1 (95% CI, 0.8-19.4)
de Sanjose et al.33	Europe, North America, Australia	4,784/6,269	3.6%/2.7%	MZL: OR, 2.47 (95% CI, 1.44-4.23)
				LPL: OR, 2.57 (95% CI, 1.14-5.79)
Schöllkopf et al.64	Denmark, Sweden	2,819/1,856	2%/1%	LPL: OR, 5.2 (95% CI, 1.0-26.4)||

Manifestations and Subtypes of B-NHL Associated with HCV

The most common associations with HCV are marginal zone lymphoma (MZL) and lymphoplasmacytic, WM, and diffuse large B cell lymphoma (DLBCL),32-35 with MZL being the most common.36-38 Transformed DLBCL is also seen.

The International Lymphoma Epidemiology Consortium reported an association of MZL (OR 2.47), lymphoplasmacytic lymphoma (OR 2.57), and DLBCL (OR 2.24) with HCV.33 Interestingly, a large population-based study in the United States found an increased risk of Burkitt lymphoma (OR 5.21) and follicular lymphoma (OR 1.88) in comparison to DLBCL (OR 1.52) and MZL (OR 2.20).32 One of the largest case-control studies to date found a higher OR (3.5 versus 2.3) for aggressive versus indolent lymphomas, respectively, and suggested that previous data may have been influenced by the relatively poorer prognosis with aggressive lymphomas.34 Patients with HCV-related DLBCL may have more aggressive clinical features at presentation in comparison to HCV-negative patients.39, 40 Other studies have also reported rare lymphoma sites with HCV infection, most notably primary splenic or hepatic41 and ocular adnexal B-NHL.42 A recent study of 12 HCV-positive patients identified a new subcutaneous “lipoma-like” primary extranodal MZL, characterized by subcutaneous nodules containing a lymphoid infiltrate. Functional IGH rearrangements were identified in nine and somatic mutations in 82%, indicating the cells were likely derived from germinal-center experienced cells.43

Management of HCV Infection in B-NHL

The management of B-NHL according to grade is outlined by guidelines from the National Clinical Cancer Network (2011), European Society of Medical Oncology (2008), and the International Working Group Guidelines for Lymphoma (2007). In turn, guidelines for the management of HCV infection are available from the World Health Organization, the National Institutes of Health, and the European Association for the Study of the Liver.

Evidence for a link between antigen drive, HCV, and lymphoma derives from research showing lymphoma regression with antiviral therapy (Table 2). The work of Hermine et al.44 demonstrated complete remissions in patients with HCV infection and splenic lymphoma with villous lymphocytes after interferon (IFN)-α treatment. All patients were treated with IFN-α and all patients with HCV and complete viral clearance had durable lymphoma remissions of more than 2 years, with no response in the six HCV-negative patients. These results have been corroborated by other groups, including those in high-prevalence countries such as Italy45, 46 and Japan,47 where lymphoma regression was positively correlated with viral load reduction.

Table 2. Case Series Showing Indolent Lymphoma Regression in HCV-Infected Patients Treated with Antiviral Therapy

Study	Lymphoma Histology	No. of HCV-Positive Patients Receiving Antiviral Therapy	Type of Antiviral Therapy Administered	Lymphoma Response After Antiviral Therapy	Duration of Lymphoma Response to Antiviral Therapy	Lymphoma Response Correlated to Virologic Response?
Abbreviations: B-NHL, B cell non-Hodgkin lymphoma; CR, complete remission; FL, follicular lymphoma; LPL, lymphoplasmacytic lymphoma; NOS, not otherwise specified; NR, no response; POD, progression of disease; PR, partial remission; RBV, ribavirin; SD, stable disease; SL, splenic lymphoma. * Cause was hepatic failure and anasarca after a long latency period of chronic HCV infection had been observed. † No P value available. ‡ No response information available for one patient. § Patients are divided into group A (n = 8) and group B (n = 10). Group A received IFN + RBV; group B received pegylated IFN + RBV. || Six patients discontinued antiviral treatment due to absence of virologic response or toxicity.
Moccia et al.65	Splenic MZL	3	IFN (n = 3)	CR (n = 2); death (n = 1)*	One CR still in remission 4.5 years after diagnosis	Not available
Hermine et al.44	SL with villous lymphocytes	9	IFN (n = 9)	CR (n = 8); PR (n = 1)	At median follow-up of 27 months (range, 15-40 months): CR, n = 7; PR, n = 1; relapse, n = 1 (re-treated to CR)	Yes†
Vallisa et al.46	FL; lymphoplasma; cytoid lymphoma; MZL	13	IFN + RBV (n = 13)	CR (n = 7); PR (n = 2); SD (n = 2); POD (n = 1)‡; unable to assess (n = 1)	Overall response of 14.1 ± 9.7 months (range, 2-24 months; median time of follow-up, 14 months)	Yes (P = 0.005)
Kelaidi et al.66	MZL	8	IFN (n = 1); IFN + RBV (n = 7)	CR (n = 5); POD (n = 1); NR (n = 2)	Unavailable	Correlation noted in some cases
Pitini et al.67	Splenic MZL	2	IFN (n = 2); IFN + RBV (n = 0)	CR (n = 2)	9 months	Yes; both patients achieved a complete virologic response at the end of treatment
Saadoun et al.68	SL with villous lymphocytes	18	IFN (n = 8); IFN + RBV (n = 10)	CR (n = 16); PR (n = 2)	All 14 patients who achieved complete virologic response had durable CRs at a mean follow-up of 62.3 months	Yes; of the 16 CRs, 14 patients achieved a complete virologic response, and two patients achieved a partial virologic response
Mazzaro et al.69	FL; SL with villous lymphocytes; LPL	18§	IFN + RBV (n = 8); pegylated IFN + RBV (n = 10)	CR (n = 9); PR (n = 4); NR (n = 5)	Unavailable	Yes
	Unavailable	Yes; all nine CR patients achieved complete virologic response at the end of treatment
Arcaini et al.45	Splenic, nodal MZL; MALT; FL; LPL; Mantle cell lymphoma; small lymphocytic lymphoma; low-grade B- NHL NOS	94	IFN (n = 36 [26 with RBV]); pegylated IFN (n = 57 [53 with RBV])||	Antiviral treatment as first-line treatment (n = 76 [CR, n = 36; PR, n = 23; SD, n = 14]); antiviral treatment as second-line treatment (n =18 [CR, n = 5; PR, n = 9])	Antiviral treatment as first-line treatment: median duration of response, 23 months (median follow-up, 3.3 years); antiviral treatment as second-line treatment: median duration of response, 26 months (median follow-up, 4.3 years)	Yes; hematologic response (CR + PR) (in patients treated with antiviral treatment as first-line treatment) was associated with achievement of sustained virologic response (P < 0.001)

For aggressive lymphoma or symptomatic indolent lymphoma, HCV eradication alone is not an option. These patients require systemic therapy with rituximab-based regimens as first treatment. From a practical vantage point, antiviral therapy to eradicate HCV is a logical recommendation after successful lymphoma therapy. Whether HCV eradication after chemoimmunotherapy may impact future survival outcome remains uncertain. Recent data suggest this may be possible, with improved disease-free survival (DFS) and clinical outcome.48 Further investigation is needed to clarify this important question. It will also be interesting to observe the impact of newer antiviral therapies (such as telaprevir and boceprevir) on the future prevalence and outcomes of B-NHL.49-51

Although lymphoma therapy is administered with the intent of curative or prolonged remission, concerns have been raised regarding its intensity and side effect profile in HCV-positive patients. Paradoxically, although the addition of rituximab to chemotherapy heralded a new treatment era,52 hepatotoxicity and viral recurrence risk are important considerations. Intriguingly, the incidence and severity of hepatotoxicity appears to have increased with time: while early reports found little evidence of liver dysfunction in HCV-positive patients treated with chemotherapy,53 recent studies have shown the opposite,39, 40, 54, 55 with the percentage of patients affected differing widely.39, 55 However, authors have cautioned that this disparity is probably a reflection of treatment differences, meaning the cause is difficult to isolate.

Besson et al.39 stated toxicity could not be attributed to pretreatment liver abnormalities or to a specific drug whereas Ennishi et al. found hepatotoxicity was more likely to occur if pre-treatment aminotransferase levels were high.40 Besson et al. also found the degree of toxicity increased with each chemotherapy cycle and as treatment progressed.39, 40 However, the authors also proposed other mechanisms, including direct cytotoxicity as a result of accelerated HCV replication after chemotherapy, hepatitis induced by immune reactivation after treatment, and increased drug toxicity from suboptimal drug metabolism.39 It is recommended that HCV-RNA levels are monitored carefully throughout treatment, because they can increase significantly.40 Whether this is potentiated by the use of agents such as rituximab is unclear, because earlier studies were completed before its routine use. Marignani et al.56 reported that hepatitis “flares” occurred in three of nine HCV-positive patients treated with rituximab, with none seen in the HCV-negative group (n = 95). Promisingly, all three of the patients were in remission with no further occurrences of hepatitis at 12 months follow-up.

Although reactivation risk is well recognized with HBV, and prophylactic antiviral therapy to suppress HBV-DNA now routinely recommended, such HCV prophylaxis during chemotherapy has not been found to be beneficial.57 Nevertheless, hepatotoxicity during lymphoma therapy appears less of a problem with HCV than HBV reactivation. Clearly, this is an area in which further study is required to prevent unnecessary dose reductions, regimen modifications, or chemotherapy cessation while balancing hepatotoxicity risk that might be severe or even fatal.39, 40

Management of hepatotoxicity is also key, because stopping or delaying treatment may translate into poorer overall survival (OS). Arcaini et al.55 reported a shorter median progression-free survival (2 years versus 3.7 years) for patients who experienced hepatotoxicity. Arcaini et al.55 and Besson et al.39 reported a decreased OS (56% versus 80%) for HCV-positive patients versus negative patients at 2 years follow-up, attributed in part to the impact of hepatotoxicity, which led to the death of three patients. However, this drop in survival was not seen by Ennishi et al.,40 who found that HCV infection was not an adverse prognostic factor despite significant levels of hepatotoxicity (27% of patients), with a 3-year progression-free survival of 69% versus 77% (P = 0.22) and OS of 75% versus 84% (P = 0.07). Significant immunosuppression may also change the tempo of HCV natural history and accelerate complications such as cirrhosis, as seen in HCV-infected allogeneic bone marrow transplant recipients.58 Based on the current literature, it is therefore advisable that patients with HCV and lymphoma are monitored for hepatotoxicity and that appropriate specialist referrals are made for the management of infection throughout lymphoma treatment, with hepatologists and oncologists working closely together to optimize outcome.

Prognostication and Prevention of HCV-Related Lymphoma

With the growing realization that HCV may be responsible for a higher lymphoma prevalence, research efforts have moved toward prognostication and prevention. A new HCV prognostic score was recently presented that attempts to differentiate between three risk categories (low, intermediate, and high) and ascribe a numerical score according to specific factors.59 The group from Italy studied HCV-positive B-NHL, and their multivariate analysis on 1,043 patients identified three factors associated with poorer OS: Eastern Cooperative Oncology Group score ≥2, HCV-RNA >106 IU/mL and serum albumin <3.5 g/dL.

There is also evidence to show that antiviral therapy may prevent lymphoma development in HCV-positive patients. Zuckerman et al.60 demonstrated the correlation of antiviral treatment response with elimination of either t(14;18) or immunoglobulin H rearrangements in HCV patients. Not all groups have confirmed these findings, however.44, 46 It is not yet known, therefore, whether t(14;18) is a neoplastic marker in HCV-positive patients, as some have experienced concurrent “molecular” and clinical remissions in response to antiviral therapy, whereas others have only achieved clinical remissions. This is compounded by uncertainty regarding the significance of t(14;18) as a neoplastic biomarker, because studies have found low-level circulating t(14;18) positive cells in healthy individuals, albeit at different frequencies, which may be age-related.61

There have also been promising retrospective data from Japan comparing 501 HCV-infected patients who had never received antiviral (IFN) therapy with 2,708 patients who had. The group reported an annual incidence of lymphoma of 0.23% overall, but the results between the groups were strikingly different. In the non-IFN group, the cumulative rates of lymphoma were reported as 0.6% at 5 years, 2.3% at 10 years, and 2.6% at 15 years, whereas a flat rate of 0% was seen in the IFN-treated group who achieved a sustained viral response.62

Conclusions

We have summarized the current literature that supports a link between HCV and B-NHL and have reviewed management strategies for HCV-associated lymphomas. Despite research advances, knowledge gaps remain regarding the in vivo mechanisms that link viral infection to malignant lymphoproliferation and the optimal management of a clinically disparate set of lymphomas. Prospective clinical trials are required to prove whether antiviral therapy alone can induce effective, durable remissions in indolent lymphomas, and as consolidation, contribute to curative outcome in aggressive lymphomas. Finally, a proven link between HCV and B-NHL may create a new therapeutic dimension in public health by providing the opportunity to successfully prevent associated B cell lymphoproliferation and lymphomas.

Summary points:

• The most common B-NHL subtypes associated with HCV infection include MZL, WM, lymphoplasmacytic lymphoma, and DLBCL.

• Antiviral therapy may have a significant role in the treatment and prevention of some HCV-associated B-NHL disorders.

• Primary treatment of HCV infection may be an alternative to standard lymphoma therapy in some HCV-associated indolent lymphomas.

• Systemic therapy of B-NHL in HCV-positive patients requires close monitoring of hepatic function and viral activity.

• Posttreatment consolidation with HCV antiviral eradication should be studied/considered in all eligible patients with HCV-associated B-NHL.

• Collaboration between hepatologists and medical oncologists is essential to optimize outcome in HCV-associated lymphomas.