Clinicopathologic Characteristics and Outcomes of Transformed Diffuse Large B-Cell Lymphoma in Hepatitis C Virus-Infected Patients

Abstract

Hepatitis C virus (HCV) causes a chronic but curable infection associated with the development of marginal zone lymphoma (MZL) and diffuse large B-cell lymphoma (DLBCL). Preliminary data have shown frequent transformation of indolent lymphoma to DLBCL in HCV-infected patients. To compare their clinicopathologic characteristics and oncologic outcomes, we reviewed the medical records and pathology reports of HCV-infected and uninfected patients with DLBCL that transformed from indolent lymphoma seen at The University of Texas MD Anderson Cancer Center (June 2004–May 2015). To identify predictors of DLBCL relapse, patients with relapse after first-line chemotherapy were compared with those without it using univariate and logistic regression analyses. Compared with the uninfected patients (n=63), HCV-infected patients (n=21) were younger (median age =54 years [interquartile range= 49–62 years] vs 62 years [53–66 years]; p= 0.01) and more often had advanced DLBCL (Ann Arbor stage 3–4; 95% vs 76%; p= 0.05). Immunophenotypically, more HCV-infected than uninfected patients had CD10-negative B cells (76% vs 43%; p= 0.008), CD5-positive B cells (39% vs 7%; p= 0.004), and activated B-cell phenotypes (57% v 31%; p= 0.07). Comparison of the patients who had relapse after first-line chemotherapy (n= 42) and those who did not (n= 40) revealed that having CD5-positive B cells was the only factor associated with DLBCL relapse in multivariate analysis (odds ratio= 10.7; p= 0.02). HCV-infected patients with transformed DLBCL have unique clinicopathologic characteristics that make their lymphoma difficult to treat, potentially leading to unfavorable outcome. The impact of HCV eradication should be explored in such patients.

INTRODUCTION

Hepatitis C virus (HCV) is an oncogenic virus that was responsible for about 7.8% of all human cancers in 2012.^{1, 2} HCV infection is curable because of its very high rate of response to antivirals.³ Apart from HCV-attributable hepatocellular carcinoma, the strongest association between HCV and cancer is the epidemiologic link between the virus and subtypes of B-cell non-Hodgkin lymphoma (NHL),^4–7 particularly marginal zone lymphoma (MZL), diffuse large B-cell lymphoma (DLBCL), and lymphoplasmacytic lymphoma.^{1, 4, 8} Researchers have proposed multiple mechanisms of HCV-induced lymphomagenesis, mainly chronic viral stimulation of lymphocytes and HCV replication in B cells mediated by HCV envelope proteins interacting with lymphocyte-specific glycoproteins.^{5, 9, 10} Furthermore, regression of indolent lymphomas induced by only antiviral therapy clinically confirms the epidemiologic association of subtypes of NHL with HCV, described since 1994.^11–13 In fact, antiviral therapy for HCV infection is currently the first step in the management of HCV-associated splenic MZL when chemotherapy is not immediately indicated.¹⁴

DLBCL is an aggressive subtype of NHL that can be de novo or arise from indolent lymphoma.¹⁵ Transformed DLBCL is mostly described in patients with follicular lymphoma,^{15, 16} and less frequently in those with chronic lymphocytic leukemia,¹⁷ lymphoplasmacytic lymphoma,¹⁸ and MZL.¹⁹ Histologic transformation is an important challenge in the management of DLBCL, as such patients have tumors that are refractory to chemotherapy and unfavorable oncologic outcomes.^{16, 20}

A small study including 26 HCV-infected patients reported a higher frequency of transformed DLBCL in HCV-infected patients than in their uninfected counterparts.²¹ Consequently, because of the strong association between HCV infection and B-cell NHL, we conducted the present study to determine if transformed DLBCL in HCV-infected patients arises from specific indolent lymphomas and has different clinical or pathologic characteristics compared to HCV-uninfected patients. Additionally, we explored the characteristics associated with DLBCL relapse to identify a possible role for HCV infection in inducing resistance of DLBCL to chemotherapy.

PATIENTS AND METHODS

The medical records and pathology reports for patients with chronic HCV infection and DLBCL seen at The University of Texas MD Anderson Cancer Center from June 2004 to May 2015 were reviewed retrospectively. Only patients with proven HCV infections and evidence of indolent NHL that transformed to DLBCL were included. Patients with DLBCL that transformed from Hodgkin lymphoma and those with Richter’s transformation (transformation from chronic lymphocytic leukemia/small lymphocytic lymphoma to DLBCL) were excluded. HCV-infected patients were compared with randomly selected uninfected patients with transformed DLBCL diagnosed in the same study period at a ratio of 1:3. This study was approved by the MD Anderson Institutional Review Board.

Definitions

Transformed DLBCL was defined as either that in a patient with a previous history of indolent lymphoma (MZL, follicular lymphoma, or lymphoplasmacytic lymphoma) before DLBCL diagnosis or with concomitant biopsies demonstrating simultaneous indolent lymphoma and DLBCL, as previously described.^{21, 22} Pathology reports were reviewed by a pathologist. Available immunohistochemistry, flow cytometry, and fluorescence in situ hybridization data were collected. The panel of markers in flow cytometry immunophenotype analysis included CD5, CD10, CD19, CD20, CD22, CD23, CD30, CD38, CD43, CD45, CD79b, FMC7, and the surface immunoglobulin kappa and lambda light chains. The panel of markers detected using immunohistochemistry included CD10, CD20, CD30, CD43, CD45, PAX5, BCL2, BCL6, MUM-1, and cyclin D1. Fluorescence in situ hybridization was performed to determine the MYC gene rearrangement using break-apart probes. DLBCL was further classified as having the germinal center B-cell or activated B-cell phenotype according to the Hans criteria.

Proven HCV infection was defined by either detectable serum HCV RNA (chronic HCV infection) or undetectable HCV RNA with a prior history of antiviral treatment. Chronic hepatitis B virus (HBV) infection was defined by the presence of positive hepatitis B surface antigen, hepatitis B DNA, and hepatitis B core antibodies. Past HBV infection was defined by negative hepatitis B surface antigen and positive hepatitis B core antibodies regardless of the hepatitis B surface antibody result.

As described previously,²³ patients with liver, stomach, pancreatic, and/or splenic involvement of DLBCL were combined in the upper gastrointestinal (GI) involvement group. The revised 2007 Cheson criteria were used to define the response of DLBCL to chemotherapy, distinguishing complete response, partial response, stable disease, and progressive disease.²⁴ The overall survival (OS) duration was measured from the start date for chemotherapy to the date of last follow-up examination or death²⁵. The progression-free survival (PFS) duration was measured from the start date for chemotherapy to the date of DLBCL relapse or progression.²⁵

Statistical Analysis

To determine the effect of HCV infection on the clinicopathologic presentation and oncologic outcomes of DLBCL, a retrospective cohort analysis of patients with transformed DLBCL was conducted to compare those with and without HCV infection (Figure 1). A chi-square or the Fisher exact test was used to compare categorical variables, and the Wilcoxon-Mann-Whitney test was performed to compare continuous variables. The 2-year OS and PFS rates in HCV-infected and uninfected patients were compared, with the differences evaluated using a log-rank test. To identify predictors of DLBCL relapse, a case-control analysis was conducted to compare patients who had DLBCL relapse or progressive disease after first-line chemotherapy with those without relapses irrespective of HCV status (Figure 1). Logistic regression analysis was used to identify independent predictors of transformed DLBCL relapse. All statistical tests were two-sided and conducted using the SPSS software program (version 23.0; IBM Corporation, Armonk, NY). P values less than or equal to 0.05 were considered statistically significant.

Figure 1.

Flow diagram of HCV-infected patients with transformed DLBCL analyzed in this study.

RESULTS

Cohort Study

We identified 25 patients with detectable HCV antibodies and evidence of transformed DLBCL. We excluded four of these patients from the study: two because of undetectable HCV viral loads without a history of HCV infection treatment and two because they had not undergone HCV RNA testing. We analyzed the remaining 21 patients and compared them with 63 HCV-uninfected patients with transformed DLBCL (Figure 1).

The majority of the HCV-infected patients were white (n = 17 [81%]) and male (n = 13 [62%]), and the median age was 54 years (interquartile range [IQR] = 49–62 years). Three HCV-infected patients (14%) had a prior history of indolent lymphoma, and 18 (86%) had biopsies demonstrating simultaneous indolent lymphoma and DLBCL. The indolent lymphomas in the 21 biopsies were follicular lymphoma (n = 6), nodal MZL (n = 3), splenic MZL (n = 2), mucosa-associated lymphoid tissue (n = 2), and low-grade lymphoma not otherwise specified (n = 8) due to a small percentage of bone marrow infiltration limiting further specification or due to a heterogeneous low grade component. None of the five HCV-infected patients tested with fluorescence in situ hybridization had MYC rearrangement, and 3 of the 11 (27%) patients tested had identified cytogenetic aberrations. Of 15 patients with available data, 12 (80%) were infected with HCV genotype 1. None of the 21 patients were co-infected with human immunodeficiency virus, whereas one (5%) and seven (33%) had chronic and past HBV infections, respectively. Only, 6 patients (29%) had cirrhosis at the time of DLBCL diagnosis.

Comparison of Clinicopathologic Characteristics and Outcomes in HCV-Infected and Uninfected Patients

Compared with the uninfected patients, the HCV-infected patients in our study were younger (median age = 54 years [IQR = 49–62 years] v 62 years [IQR = 53–66 years]; p = 0.01) and more often had advanced (Ann Arbor stage 3–4) disease (95% v 76%; p = 0.05), upper GI involvement (48% v 25%; p = 0.05), and bone marrow involvement (66% v 41%; p = 0.07) (Table 1). Immunohistochemical data demonstrated that more HCV-infected patients than uninfected patients had CD5-positive B cells (39% v 7%; p = 0.004), CD10-negative B cells (76% v 43%; p = 0.008), and activated B-cell phenotypes (57% v 31%; p = 0.07) (Table 2). HCV-infected patients with and without cirrhosis had worse 2-year OS rates than did uninfected patients (67% v 80% v 92%; p = 0.02) (Figure 2). We did not find a difference in the 2-year PFS rate between the HCV-infected and uninfected patients.

Table 1.

Comparison of clinical characteristics of HCV-infected and uninfected patients with transformed DLBCL

	n (%)
Characteristic	HCV-infected patients (n = 21)	Uninfected patients (n = 63)	p
Median age at DLBCL diagnosis, years (IQR)	54 (49–62)	62 (53–66)	0.01
Sex
Male	13 (62)	37 (59)	0.79
Female	8 (38)	26 (41)
Race
White	17 (81)	48 (76)	0.76
Black	3 (14)	3 (5)	0.16
Hispanic	0	9 (14)	0.10
Other	1 (5)	3 (5)	>0.99
Ann Arbor stage
1–2	1 (5)	15 (24)	0.05
3–4	20 (95)	48 (76)
Extranodal involvement	18 (86)	44 (70)	0.15
Upper GI involvement	10 (48)	16 (25)	0.05
Bone marrow involvement	14 (66)	26 (41)	0.07
B symptoms	9 (43)	16 (25)	0.12
ECOG performance status
0–1	13/18 (72)	50/59 (85)	0.29
2–4	5/18 (28)	9/59 (15)
IPI score
Low	4/16 (25)	20/61 (33)	>0.99
Intermediate	8/16 (50)	30/61 (49)	>0.99
High	4/16 (25)	11/61 (18)	0.50
Cirrhosis	6 (29)	0	<0.01
HIV infection	0	0	--
Chronic HBV infection	1 (5)	2 (3)	>0.99
Past HBV infection	7 (33)	1 (2)	<0.01

HCV indicates hepatitis C virus; DLBCL, diffuse large B-cell lymphoma; IQR, interquartile range; GI, gastrointestinal; ECOG, eastern cooperative oncology group; IPI, international prognostic index; HIV, human immunodeficiency virus; HBV, hepatitis B virus.

Table 2.

Comparison of pathologic characteristics of HCV-infected and uninfected patients with transformed DLBCL

	n (%)
Characteristic	HCV-infected patients (n = 21)	Uninfected patients (n = 63)	p
Site of biopsy
Lymph node	13 (62)	34 (54)	0.52
Extranodal	8 (38)	29 (46)
Immunophenotype
CD5+	7/18 (39)	4/54 (7)	0.004
CD10+	5/21 (24)	35/61 (57)	0.008
CD5−/CD10−	9/20 (45)	19/56 (34)	0.37
CD5−/CD10−/CD23−	6/18 (33)	7/45 (16)	0.16
CD19+	12/12 (100)	22/22 (100)	>0.99
CD20+	21/21 (100)	61/62 (98)	>0.99
CD22+	6/6 (100)	8/8 (100)	>0.99
CD23+	2/12 (17)	9/28 (32)	0.45
CD30+	0/8	6/22 (27)	0.15
CD38+	4/5 (80)	6/8 (75)	>0.99
CD43+	2/8 (25)	7/21 (33)	>0.99
CD45+	6/6 (100)	13/13 (100)	>0.99
CD79a+	6/6 (100)	8/8 (100)	>0.99
FMC+	5/6 (83)	7/9 (78)	>0.99
PAX5+	6/6 (100)	19/19 (100)	>0.99
BCL2+	13/16 (81)	43/49 (88)	0.67
BCL6+	12/15 (80)	43/50 (86)	0.68
MUM-1+	9/12 (75)	18/28 (64)	0.71
Cyclin D1+ *	0/12	3/32 (9)	0.55
Kappa light chain+	11/15 (73)	18/31 (58)	0.31
Lambda light chain+	4/15 (27)	12/31 (39)	0.42
c-MYC rearrangement	0/4	3/13 (23)	0.54
Activated B-cell phenotype	8/14 (57)	16/51 (31)	0.07

HCV indicates hepatitis C virus; DLBCL, diffuse large B-cell lymphoma.

^*Cyclin D1 was tested in 8 out of our 11 patients with CD5-positive B-cells. It was negative in 7 patients and positive in one patient in which further testing was done, including CCND1 (negative) and IGH-BCL2 (positive) gene rearrangements, ruling out mantle cell lymphoma. Cyclin D1 was not tested in the other 3 patients, because CD5-positivity was focal in one patient, and because the low grade component of the other two biopsies were identified as follicular lymphoma and marginal zone lymphoma respectively, making the diagnosis of mantle cell lymphoma less likely.

Figure 2.

Comparison of the 2-year OS comparing A) HCV-infected and uninfected patients B) HCV-infected DLBCL patients with cirrhosis, HCV-infected patients without cirrhosis, and uninfected control patients.

Case-Control Study

To identify predictors of DLBCL relapse, we compared patients with relapsed DLBCL or progressive disease after first-line chemotherapy (n = 42) with those with complete or partial responses and no relapses (n = 40). We found no differences in patient characteristics, lymphoma stage, International Prognostic Index scores, type of chemotherapy received, cirrhosis, or rate of infection with human immunodeficiency virus, HBV, or HCV (Table 3). The presence of CD5-positive B cells was the only factor associated with DLBCL relapse in univariate analysis (24% in patients with relapsed or primary refractory DLBCL vs 6% in those without DLBCL relapses; p = 0.03) and logistic regression analysis after adjustment for potential confounders, including type of first-line chemotherapy used, HCV infection status, and DLBCL stage (Table 4).

Table 3.

Comparison of clinical characteristics of patients with primary refractory or relapsed DLBCL and patients without relapse after first-line chemotherapy

	n (%)
Characteristic	Relapsed or primary refractory DLBCL (n = 42)	No relapse (n = 40)	p
Median age, years (IQR)	58 (52–65)	60 (51–65)	0.71
Sex			0.39
Male	27 (64)	22 (55)
Female	15 (36)	18 (45)
Race			>0.99
White	32 (80)	31 (78)
Nonwhite	10 (20)	9 (22)
Indolent lymphoma diagnosed prior to DLBCL diagnosis	4 (10)	3 (8)	>0.99
Ann Arbor stage			0.22
1–2	6 (14)	10 (25)
3–4	36 (86)	30 (75)
Extranodal involvement	32 (76)	28 (70)	0.52
Upper GI involvement	15 (36)	10 (25)	0.29
Bone marrow involvement	22 (52)	16 (40)	0.26
B symptoms	12 (29)	13 (33)	0.69
ECOG performance status			0.17
0–1	27/36 (75)	34/39 (87)
2–4	9/36 (25)	5/39 (13)
IPI score			0.25
Low	11/36 (31)	13/39 (33)
Intermediate	15/36 (42)	21/39 (53)
High	10/36 (27)	5/39 (13)
Cirrhosis	2 (5)	2 (5)	>0.99
HCV infection	12 (29)	7 (18)	0.23
HIV infection	0	0
Chronic HBV infection	0	3 (8)	0.11
Past HBV infection	5 (12)	3 (8)	0.71
Immunohistochemistry
CD5+	9/37 (24)	2/34 (6)	0.03
CD10+	20/42 (48)	20/38 (53)	0.65
Kappa light chain+	13/23 (57)	14/21 (67)	0.49
Lambda light chain+	8/23 (35)	8/21 (38)	0.82
BCL2+	27/31 (87)	27/32 (84)	>0.99
BCL6+	24/28 (86)	31/36 (86)	>0.99
R-CHOP	32 (76)	29 (73)	0.70

HCV indicates hepatitis C virus; DLBCL, diffuse large B-cell lymphoma; GI, gastrointestinal; ECOG, eastern cooperative oncology group; IPI, international prognostic index; HIV, human immunodeficiency virus; HBV, hepatitis B virus; R-CHOP, rituximab-cyclophosphamide, doxorubicin, vincristine, and prednisone.

Table 4.

Multivariate logistic regression analysis comparing patients with primary refractory or relapsed DLBCL (n = 42) with patients without relapse (n = 40) after first line-chemotherapy

Variable	OR	95% CI	p
CD5-positive vs CD5-negative B cells	10.7	1.4–79.0	0.02
HCV-infected vs uninfected	1.7	0.6–16.0	0.43
R-CHOP vs other first-line chemotherapy†	0.6	0.3–1.3	0.28
Ann Arbor stage 1–2 vs 3–4	0.5	0.1–2.0	0.34
ECOG performance status 0–1 vs 2–4	1.7	0.4–7.0	0.43

OR indicates Odds ratio; C.I; confidence interval; HCV, hepatitis C; ECOG, eastern cooperative oncology group; R-CHOP, rituximab-cyclophosphamide, doxorubicin, vincristine, and prednisone.

^†Rituximab-etoposide, prednisone, Oncovin and hydroxydaunorubicin hycrochloride or rituximab-cyclophosphamide, vincristine, doxorubicin and dexamethasone.

DISCUSSION

To the best of our knowledge, this is the first study to demonstrate that HCV-infected patients with DLBCL transformed from indolent lymphoma have clinicopathologic characteristics different from those of uninfected patients. HCV-infected patients appear to have more advanced stages at clinical presentation of DLBCL, with transformation occurring at a younger age than in uninfected patients. In addition, we found higher frequencies of CD5-positive and CD10-negative B cells in HCV-infected patients, suggesting more frequent transformation of DLBCL from specific indolent lymphomas such as MZL.

In a large series of 325 patients with follicular lymphoma studied over 25 years, transformation to DLBCL was more common in older patients (age >60 years) than in younger ones and in patients with higher International Prognostic Index scores than in those with lower scores.²⁶ Unfortunately, the HCV status in that study went unreported. Our results demonstrated transformation to DLBCL at an earlier age in HCV-infected patients than in uninfected ones and that transformation was associated with advanced Ann Arbor stage, extranodal involvement (particularly in the upper GI tract), and a B-cell phenotype compatible with MZL. Sustained B-cell activation and inhibition of B-cell apoptosis by HCV core proteins may explain the earlier transformation to DLBCL in infected patients.²⁷ In addition, authors have described extranodal involvement of HCV-associated DLBCL in the liver, spleen, and stomach,^21–23 which may explain the advanced Ann Arbor stages seen in our HCV-infected patients. MZL cells express pan-B-cell markers, are CD10-negative, and are usually CD5-negative, with the exceptions being subsets of CD5-positive mucosa-associated lymphoid tissue, nodal MZL, and splenic MZL cells.^28–31 Also, CD5-positive mucosa-associated lymphoid tissue tends to disseminate and involve the bone marrow frequently.^{30, 31} Therefore, the significantly higher CD10-negative and CD5-positive B-cell numbers and more frequent bone marrow involvement in our HCV-infected patients with transformed DLBCL than in our uninfected patients suggest transformation from MZL.

CD5-positive B cells are rare in adults but predominant in fetuses.³² Researchers have identified expansion of this subset of cells in cases of autoimmune diseases.^33–35 Patients with de novo CD5-positive DLBCL have clinical features of more aggressive disease than de novo CD5-negative DLBCL,³⁶ and authors reported CD5 positivity of B cells to be an independent predictor of poor prognosis for DLBCL.³⁷ Also, CD5-positive B cells are proven to be resistant to apoptosis by acquiring an interleukin-10 autocrine loop.³⁸ Furthermore, CD5 protein seems to be important for HCV entry into human lymphocytes;³⁹ CD5 positivity therefore may be linked with enhance viral entry into B cells, leading to a high viral burden and increased oncogenic potential. The findings of our study confirm the high frequency and aggressive nature of CD5-positive DLBCL cells in HCV-infected patients. It may also explain, at least in part, the high relapse rate of CD5-positive DLBCL seen in this study and the HCV-driven resistance of DLBCL to chemotherapy we observed in our previous work.²³

Authors have reported CD5-positive B-cell expansion in peripheral blood^{40, 41} and hepatic B-cell in HCV-infected patients.⁴² Of note, a previous study of 15 HCV-infected patients treated with interferon and ribavirin, the previous standard-of-care therapy for HCV infection, demonstrated disappearance of CD5-positive B cells from the peripheral blood in all 9 patients who achieved sustained virologic responses, regarded as virologic cure of HCV.⁴¹ We believe that further research of the interplay between chronic HCV infection and CD5-positive B cells in patients with DLBCL is needed, as the elimination of these apoptosis-resistant cells may be crucial to improve oncologic outcomes. Antiviral therapy with new direct-acting agents is very effective, even in HCV-infected cancer patients, with sustained virologic response rates higher than 90%.⁴³ This therapy may eliminate HCV infection, downregulate the expansion of CD5-positive B cells, and most importantly improve oncologic outcomes as already reported for selected patients with indolent NHL.^{14, 44}

Our study has some limitations. First, due to its retrospective nature, some clinical and pathologic characteristics were missing. Second, we conducted this study in a single tertiary cancer center to which a selective patient population is referred. Therefore, our results may not be generalizable to the general population of NHL patients. Third, the majority of our HCV-infected and uninfected patients had indolent lymphoma diagnosed concomitantly with DLBCL. Despite being suggestive of transformation, the presence of composite biopsies does not confirm a transformation without proven clonal relationship by molecular techniques. However, data in the literature is still scant and DLBCL in this group of patients is still considered a transformation from indolent lymphoma.^21,22 Fourth, the inclusion of patients with indolent lymphoma preceding the diagnosis of DLBCL could have created a selection bias, as these patients received chemotherapy before the development of DLBCL, potentially affecting the pathologic characteristics and response of DLBCL to chemotherapy. Nevertheless, the number of such patients was small (7 patients in total) with a similar distribution between groups (4 cases vs 3 controls respectively). Fifth, because of the small number of patients, we could not draw definite conclusions regarding the difference in 2-year OS rate between the HCV-infected and uninfected patients in univariate analysis. Thus, we analyzed the 2-year OS rate in three groups of patients, differentiating cirrhotic HCV-infected patients, to control for this important prognostic factor. Sixth, past HBV infections were more frequent in HCV-infected patients with transformed DLBCL than in HCV-uninfected patients. However, we found no difference in the rate of chronic HBV infection between HCV-infected and HCV-uninfected patients, making the effect of chronic HCV infection more important than a past HBV infection. HBV is also carcinogenic associated with the development of hepatocellular carcinoma but the lymphomagenic potential of HBV remains unclear.⁴⁵ Finally, physicians may have treated DLBCL in HCV-infected patients differently from that in uninfected patients owing to abnormal hepatic function, explaining the difference in DLBCL response to chemotherapy in these two groups. We attempted to control for these confounders by forcing HCV infection status and type of first-line chemotherapy received into the multivariate model.

In conclusion, HCV-infected patients with transformed DLBCL have unique clinicopathologic characteristics that make this cancer difficult to treat, potentially leading to unfavorable outcome. The increased numbers of CD5-positive and CD10-negative B cells in these patients suggest transformation from a specific indolent lymphoma, such as MZL, to DLBCL. In addition, the increased numbers of CD5-positive B cells may explain the previously described high incidence of DLBCL relapse in HCV-infected patients. Our study highlights the importance of HCV testing in all lymphoma patients and sheds light on the mechanism by which HCV eradication improves oncologic outcomes in patients with DLBCL as we recently reported.^{23, 46} Prospective studies of direct-acting agent-based therapy for HCV eradication are needed to confirm this benefit, particularly in CD5-positive DLBCL patients.

Brief description.

This work represents to our knowledge, the first series to show that hepatitis C virus (HCV)-infected patients with transformed diffuse large B-cell lymphoma (DLBCL) have unique clinicopathologic characteristics that make their lymphoma difficult to treat. Compared to uninfected patients, HCV-infected patients have higher frequency of CD5-positive B-cells increasing the risk of DLBCL relapse.

Abbreviations

DLBCL

diffuse large B-cell lymphoma

GI

gastrointestinal

HBV

hepatitis B virus

HCV

hepatitis C virus

IQR

interquartile range

MZL

marginal zone lymphoma

NHL

non-Hodgkin lymphoma

OS

overall survival

PFS

progression-free survival