Low hemoglobin as a novel prognostic biomarker in EBV-positive diffuse large B-cell lymphoma patients: a clinical feature analysis of 99 cases from Zhejiang, Southeast China.

Introduction

Epstein-Barr virus (EBV) is a gamma herpesvirus that has infected nearly 95% of the world population, making it one of the most ubiquitous human viruses [ 1 ]. First discovered in 1964, EBV is considered the first human tumor-causing virus [ 2 ]. Since its discovery, EBV has been implicated in a wide range of human tumors and is responsible for approximately 1.5% of all cancers worldwide [ 3 ]. EBV has the ability to infect human B lymphocytes and disrupt their normal proliferation, which can lead to the development of EBV-associated B cell tumors [ 4 ].

One particular form of EBV-associated malignancy is Epstein-Barr virus-positive diffuse large B-cell lymphoma (EBV+DLBCL). DLBCL is the most prevalent subtype of non-Hodgkin lymphoma, accounting for nearly 30% of all cases of non-Hodgkin lymphoma [ 5 ]. While EBV+ DLBCL represents only 5–15% of all DLBCL cases [ 6 ], this disease is associated with poor survival outcomes under standard chemoimmunotherapy regimens, a high frequency of extranodal involvement, and a predominance in elderly populations [ 6 , 7 ]. The relatively low incidence of EBV+ DLBCL could be partially attributed to the lack of clearly defined diagnostic criteria for EBV positivity in DLBCL, which complicates its identification and diagnosis. The most reliable method for diagnosing EBV+ DLBCL is the detection of EBV-encoded small RNAs (EBERs) within the nuclei of tumor cells, which serves as a definitive marker of EBV infection. However, the absence of a consensus cut-off value for EBER positivity has resulted in the use of divergent thresholds (from 5% to 80%) in the literature, consequently generating inconsistent findings [ 7 – 11 ]. According to the 4th and 5th edition of World Health Organization (WHO) lymphoma classifications, the diagnosis of EBV+ DLBCL, NOS, requires the demonstration of EBER positivity via in situ hybridization (ISH) in > 80% of tumor cells. Small numbers of EBER-positive cells may be present as bystander B cells in EBV-negative B-cell or T-cell lymphomas [ 12 , 13 ]. A meta-analysis has estimated that the pooled prevalence of EBV+ DLBCL among all DLBCL cases is approximately 7.9% [ 14 ]. Additionally, the rarity of EBV+ DLBCL and the small number of cases available for research also make it difficult to perform multivariate survival analyses and gain a comprehensive understanding of the disease. As a result, detailed clinicopathologic data and follow-up information regarding EBV+ DLBCL remain limited. International Prognostic Index (IPI) is a primary prognostic factor for risk stratification in DLBCL [ 15 ]. To complement IPI scoring system, another prognostic index called the Oyama score is established which incorporates age older than 70 years and the presence of B symptoms [ 16 ]. Although high IPI and Oyama scores have been shown to be associated with worse outcomes in EBV+ DLBCL patients [ 17 ], whether these two scores can accurately predict the prognosis in general EBV+ DLBCL patients remains to be validated. In addition, several studies have suggested that anemia as an adverse prognostic factor in DLBCL patients [ 18 , 19 ]. Anemia is a blood disorder defined as a low number of red bloods cells or reduction in hemoglobin (Hb) or hematocrit (Hct).

To address this gap, we retrospectively assessed 99 cases of EBV+ DLBCL at our institution, aiming to expand the knowledge of this disease through a thorough analysis of its clinical features and prognostic factors. We analyzed the clinical characteristics of EBV+DLBCL, with the aim of investigating the prognostic value of several simple laboratory parameters including hemoglobin levels.

Patients and methods

This study was conducted at The First Affiliated Hospital, College of Medicine, Zhejiang University and Ningbo No.2 Hospital, Zhejiang Province, Ningbo. From August 2016 and March 2023, a total of 99 patients with confirmed pathological diagnoses of Epstein-Barr virus-positive diffuse large B-cell lymphoma (EBV+DLBCL) were included. The enrolled criteria required the availability of both a confirmed pathological diagnosis and comprehensive medical records in this study. The patient enrollment period was from August 2016 to March 2023, with the final follow-up conducted on August 31, 2023. The clinical information used for this study was derived from laboratory reports and detailed clinical records. Survival data were anonymized and analyzed in accordance with ethical standards. The study was conducted in full compliance with the Declaration of Helsinki, and ethical approval was obtained from the institutional ethics committee of our hospital.

Pathology

A cutoff rate of 30% expression was applied to define a sample as positive for CD10, BCL6, and MUM1/IRF4, following the criteria established by the Hans algorithm [ 20 ]. To assess the presence of EBV, the Epstein-Barr-encoded RNA (EBER) was detected using a fluorescent-labeled peptide nucleic acid (PNA) probe (Dako), in conjunction with the Dako PNA in situ hybridization (ISH) detection kit. This method was performed on formalin-fixed paraffin-embedded tissue sections. A sample was considered positive for EBV when EBER was present in at least 80% of malignant cells, as determined by the sensitivity of the ISH technique.

Statistics

The clinical data collected for this study included patient age, sex, complete blood counts, Eastern Cooperative Oncology Group performance status (ECOG PS), presence of B symptoms, lactate dehydrogenase (LDH) levels, Ann Arbor stage, extranodal involvement, International Prognostic Index (IPI) score, therapy received, response to therapy, overall survival (OS) and progression-free survival (PFS). The Oyama score [ 16 ], which incorporates age greater than 70 years and presence of B symptoms, was calculated and used to stratify the cases into three distinct risk categories. Descriptive statistics were applied to summarize the clinical and pathological data. PFS was defined as the duration of survival without response, or relapse and progression from the time of disease diagnosis, while OS was defined as the time from diagnosis to death or last follow up. Univariate and multivariate Cox proportional hazards models were used to identify independent prognostic factors for PFS and OS. Variables that reached a significance level of p < 0.05 in the univariate analysis were considered for inclusion in the multivariate Cox model to identify independent prognostic factors. The results of the final multivariate models are presented as hazard ratios (HRs) with their corresponding 95% confidence intervals (CIs). Survival curves were compared using the log-rank test. All statistical analyses and graphical representations were conducted using the R software package.

Results

Characteristics of EBV+ DLBCL patients

The clinical profile of the 99 EBV+ DLBCL patients enrolled in the study is summarized in Table 1 . The median age of the cohort was 70 years, with male-to-female ratio of 1.9: 1(65/34). As an aggressive subtype of DLBCL, EBV+DLBCL patients exhibited a substantial proportion of extranodal involvement, with 75.8% of patients showing extranodal involvement. Notably, 54.4% (54/99) of patients had more than one site of extranodal involvement. The lymphoma cells were frequently observed in bone marrow, spleen, stomach, skin and lungs. These findings suggest that EBV+ DLBCL is an invasive disease, with a predisposition for extranodal involvement.

Immunophenotyping revealed that the non-GCB type of DLBCL, according to Hans classification, was predominant, accounting for 82.8% (82/99) of cases. MostEBV + DLBCL patients were diagnosed at an advanced stage, with 90.1% (90/99) presenting at Ann Arbor stage III or IV. Additionally, patients generally exhibited a frail performance status, with 76.7% (76/99) of patients having an ECOG PS of at least 2. The Oyama score [ 16 ], which is associated with poor prognosis, was calculated, and 12.1% (12/99) of patients presented with an Oyama score of 2, while 50.5% (50/99) had a score of 1. The up-to-date overall survival (OS) rate was 45.4%, and progression-free survival (PFS) was 30.3%.

Prognostic factors affecting OS and PFS

Table 2 summarized the factors influencing OS in EBV+ DLBCL patients. Figure 1 A depicts survival curves according to the therapy received. A significant difference in OS was observed based on age (age ≤ 60 vs. Age > 60, P = 0.012). Elevated LDH, ECOG PS of 2–4, high IPI score, B symptoms, low albumin and hemoglobin were identified as unfavorable prognostic factors. The IPI score of ≥ 4 ( P = 0.047) was associated with poor prognosis but failed to stratify patients into distinct survival risk categories in multivariate analysis (Table 3 ; Fig. 1 B). Figure 1 B depicts the overall survival curves according to IPI score in patients with EBV+ DLBCL. Although higher IPI scores corresponded to a trend toward worse survival, this association was not statistically significant in the multivariate analysis.

Univariate Cox analysis suggested that both IPI score and Oyama score performed well as risk stratifying tools, with IPI score ≥ 4 ( P = 0.047), Oyama score = 1 ( P = 0.034), and Oyama score = 2 ( P = 0.017) being associated with worse outcomes. However, multivariate analysis showed that neither IPI score nor Oyama score could stratify patients into subgroups with significant differences in OS (Table 3 ).

Univariate and multivariate Cox analyses for progression-free survival (PFS) also demonstrated results consistent with those for OS. The IPI score ≥ 4 ( P = 0.015) was associated with worse PFS in univariate analysis (Table 4 ). But both IPI score and Oyama score failed to stratify patients into distinct PFS groups in multivariate analysis. Regarding the IPI and Oyama scores, their lack of significance in multivariate analysis may be attributed to the limited sample size or potential incompatibility with the EBV+DLBCL subtype. We recommend validation in larger cohorts.

The impact of different therapy regimens on survival in EBV+ DLBCL patients

In the post-rituximab era, rituximab in combination with cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) chemotherapy remains the standard front-line treatment for DLBCL. Among patients with EBV+DLBCL, R-CHOP-type therapy is also the first-line treatment. The R-CHOP-type therapies included in this study were R-CHOP, R-CHOPE (R-CHOP+ etoposide), RDAEPOCH (dose-adjusted R-EPOCH). In contrast, other relatively milder therapies included R-GDP (Rituximab plus gemcitabine, dexamethasone and cis-platinum), R (Rituximab), COP (cyclophosphamide, vincristine, and prednisone), BR (Rituximab plus Bendamustine), and BSC (best supportive care). In this cohort, 71 patients (71.7%) were treated with R-CHOP-type therapy and 28 patients (28.3%) received less intensive regimens (Table 1 ).

Fig. 1 A depicts significantly superior survival outcomes for patients receiving R-CHOP-type chemotherapy compared to alternative regimens, a finding that was consistent in both univariate and multivariate analyses (OS: P<0.001; PFS: P=0.001; see Tables 4 , 5 ). Overall, R-CHOP-type therapy was associated with significantly better OS. However, EBV+DLBCL patients demonstrated more aggressive clinical features and lower OS compared to the general DLBCL population [ 21 ]. The probabilities of OS and PFS were estimated to be below 40% and 20%, respectively (Fig. 2 ). Among the 99 EBV+DLBCL patients, 37 patients (37.4%) achieved complete remission (CR), 15 patients (15.2%) achieved partial remission (PR), and 47 patients (47.5%) had stable disease (SD) or progressive disease (PD) (Table 1 ).

Prognostic effects of hemoglobin level on overall survival

By Cox regression analysis, we find the optimal hemoglobin level cutoff point as 90 g/L (Fig. 3 A) for different Hazard Ratio. Patients with hemoglobin level lower than 90 g/L were categorized into low Hb class. Regarding the prognostic value of hemoglobin level in patients with EBV+DLBCL, we draw a survival curve with R software (Fig. 3 B). As shown in Fig. 3 B, the Kaplan-Meier curve indicates that patients with Hb < 90 g/L have significantly poorer PFS compared to those with higher levels (log-rank p= 0.002). The OS rates were 51.3% (41/80) and 21.1% (4/19) for patients with high and low hemoglobin levels, respectively. It indicates that patients with high hemoglobin levels may have better survival compared to patients with low hemoglobin levels.

By multivariate analysis (Table 5 ), patients with low hemoglobin level were found to be associated with worse OS ( P = 0.001), and higher hemoglobin level demonstrated protective effect in patients with EBV+DLBCL (HR = 0.315). Low hemoglobin level (Hb <90 g/L) was considered as an independent prognostic factor for poor OS both in univariate and multivariate analysis (Table 5 ). Univariate and multivariate analysis for PFS in patients with EBV+DLBCL revealed similar results as OS (Table 6 ). Patients with low hemoglobin level was found to be associated with worse PFS ( P = 0.004) in multivariate analysis.

The presence of B symptoms was also considered as an adverse prognostic factor of OS (HR = 1.771, Table 5 ), but statistical significance was only found in univariate analysis. Univariate analysis revealed that IPI score ≥4 was associate with poor OS ( P = 0.047) (Table 5 ) and PFS ( P =0.015) (Tables 6 ). According to multivariate analysis, statistical significance ( P =0.035) was found only in IPI score ≥4 group with poor PFS (Table 6 ). Statistical significance of OS ( P = 0.113) was not found in multivariate analysis (Table 5 ).

Discussion

In this study we reported hemoglobin as an important prognostic factor in EBV+ DLBCL. This study presents a comprehensive analysis of 99 consecutively diagnosed EBV+DLBCL case series in China, contributing to the growing body of knowledge on this distinct subtype of diffuse large B-cell lymphoma. Conducted in the post-rituximab era, this retrospective analysis sought to characterize the clinical features, immunophenotypic profiles, and treatment outcomes of EBV+ DLBCL patients. Previous studies suggested that anemia with baseline hemoglobin levels of <10 g/dL as an independent prognostic factor in patients with DLBCL treated with R-CHOP standard immunochemotherapy [ 18 , 19 ]. Given that hemoglobin level was an acknowledged prognostic factor in many types of hematological malignancies, such as follicular lymphoma, Hodgkin lymphoma and multiple myeloma. It is reasonable that pretreatment anemia indicates high risk and advanced stage, poor treatment tolerance, and these might contribute to worse OS. Consistent with prior studies, EBV+DLBCL patients in this cohort exhibited high IPI scores and worse survival outcomes. These findings reinforce the notion that EBV+DLBCL is a more aggressive and clinically challenging disease compared to EBV-negative DLBCL, highlighting the need for tailored therapeutic strategies.

Despite advancements in lymphoma treatment, current prognostic tools are inadequate in effectively stratifying EBV+ DLBCL patients into distinct risk categories. This limitation underscores the necessity for further investigation into the unique characteristics of this disease, particularly across larger, multi-center cohorts from diverse geographical regions. Such studies would help refine prognostic models and contribute to more personalized treatment strategies for this patient population. Several studies have showed that EBV+ DLBCL patients is associated with poorer prognosis compared to its EBV-negative counterpart [ 22 – 24 ]. In our study, patients receiving R-CHOP-type chemotherapy demonstrated a survival benefit compared to those treated with less intensive regimens; however, the complete remission (CR) rate remained suboptimal at 37.4%. Additionally, only 15.2% (15/99) of patients achieved partial remission (PR), while nearly half (47.5%, 47/99) exhibited stable disease (SD) or progressive disease (PD) (Table 1 ). These results underscore the high relapse risk and limited efficacy of current therapeutic approaches in EBV+DLBCL. A separate predictive model suggested a long-term PFS probability of approximately 20%, highlighting the persistent relapse risk (Fig. 2 ), further highlighting the therapeutic challenges associated with this disease.

In our cohort, a significant proportion of patients (76.7%) had an ECOG PS greater than 1, and 54.5% (54/99) had a high International Prognostic Index (IPI) score (≥3) (Table 1 ), indicating a high-risk profile. The advanced age and poor PS of many EBV+ DLBCL patients contribute to their fragility, making them less likely to tolerate the intensity of standard chemotherapy regimens. This highlights the need for novel therapies specifically designed for patients who are refractory to stand treatments or cannot tolerate high-intensity regimens due to their old age or poor performance status.

This study has several limitations that should be considered. First, as a retrospective analysis conducted at two major centers in Zhejiang Province, the findings may be influenced by selection bias and might not be fully generalizable to the broader patient population within the province. Although efforts were made to ensure consistency in EBER detection—using in situ hybridization based on WHO criteria (defined as positivity in >80% of tumor cell nuclei)—the inherent variability in sample processing and interpretation across time and observers cannot be entirely excluded. Therefore, future prospective, multi-center studies with uniform testing protocols are warranted to validate these observations. Furthermore, experimental validation employing multi-omics approaches and representative model systems is crucial to functionally characterize these mechanisms.

In conclusion, despite these limitations, our study indicates that EBV+DLBCL constitutes a distinct clinicopathological subgroup within the DLBCL spectrum. This entity necessitates careful clinical evaluation and a personalized treatment approach. Anemia (Hb <90 g/L) is an independent prognostic factor associated with adverse outcomes in patients with EBV+ DLBCL. A hemoglobin level below 90 g/L may serve as a readily accessible and effective prognostic marker for risk stratification in this population. Given the aggressive clinical behavior and poor prognosis of EBV+DLBCL, it is critical to implement individualized therapeutic strategies that consider both the patient’s performance status and disease stage. Furthermore, the development of improved risk stratification tools and the exploration of novel treatment modalities are imperative to enhance survival outcomes in these patients.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1 (DOCX 38.7 KB)

Author contributions

Study concept: Jinghan Wang. Data collection, Initial draft of manuscript: Mengmeng Huang. Data analysis and creation of figures and tables: Mengmeng Huang and Jinghan Wang. Pathological data collection: Fang Yu. Data Integration: Kesang Li, Gangfeng Xiao. Jie Jin, Jinghan Wang. Supervision, article review and editing: Suying Qian, Jie Jin. Critical revision and approval of final version: all authors.

Funding

This work was supported by grants from Ningbo No.2 Hospital Key Discipline - Department of Hematologic and Oncology (2023-Y08);

Zhejiang Medical and Health Science and Technology Plan 2025(2025KY1407).

Data availability

No datasets were generated or analyzed during the current study.

Declarations

Ethics approval and consent to participate

This study was a retrospective analysis of de-identified data collected as part of routine clinical care. No personally identifiable information was used. Given its retrospective nature and the absence of direct patient contact or intervention, this study was exempted from institutional review board (IRB) approval and the requirement for informed consent. The study was approved by the Ethics Committee of Ningbo No.2 Hospital (Approval NO.: PJ-NBEY-KY-2025-027-01) and was conducted following the principles of the Helsinki Declaration in its current version.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Mengmeng Huang, Kesang Li and Fang Yu contributed equally.

Contributor Information

Mengmeng Huang, Email: mmh1213@126.com.

Jinghan Wang, Email: 1513084@zju.edu.cn.

Suying Qian, Email: qiansuyinghao@163.com.

References

1. Cohen JI et al (2011) Characterization and treatment of chronic active Epstein-Barr virus disease: a 28-year experience in the United States. Blood 117(22):5835–5849

2. Epstein MA, Achong BG, Barr YM (1964) Virus particles in cultured lymphoblasts from Burkitt’s lymphoma. Lancet 1(7335):702–703

3. Farrell PJ (2019) Epstein-barr virus and cancer. Annu Rev Pathol 14:29–53

4. Young LS, Yap LF, Murray PG (2016) Epstein-barr virus: more than 50 years old and still providing surprises. Nat Rev Cancer 16(12):789–802

5. Padala SA, Kallam A (2024) Diffuse Large B-Cell Lymphoma, in StatPearls. Treasure Island (FL)

6. Malpica L et al (2022) EBV-positive diffuse large B-cell lymphoma, not otherwise specified: 2022 update on diagnosis, risk-stratification, and management. Am J Hematol 97(7):951–965

7. Malpica L et al (2024) EBV-positive diffuse large B-cell lymphoma, not otherwise specified: 2024 update on the diagnosis, risk-stratification, and management. Am J Hematol

8. Gao X et al (2018) Clinical characteristics and prognostic significance of EBER positivity in diffuse large B-cell lymphoma: A meta-analysis. PLoS ONE 13(6):e0199398

9. Miyagi S et al (2020) Reappraisal of Primary Epstein-Barr Virus (EBV)-positive Diffuse Large B-Cell Lymphoma of the Gastrointestinal Tract: Comparative Analysis Among Immunosuppressed and Nonimmunosuppressed Stage I and II-IV Patients. Am J Surg Pathol 44(9):1173–1183

10. Ok CY et al (2014) Prevalence and clinical implications of epstein-barr virus infection in de novo diffuse large B-cell lymphoma in Western countries. Clin Cancer Res 20(9):2338–2349

11. Ziarkiewicz M et al (2016) Epstein-Barr Virus-Positive Diffuse Large B cell Lymphoma in the Experience of a Tertiary Medical Center in Poland. Arch Immunol Ther Exp (Warsz) 64(2):159–169

12. Swerdlow SH et al (2016) The 2016 revision of the World Health Organization classification of lymphoid neoplasms. Blood 127(20):2375–2390

13. Alaggio R et al (2022) The 5th edition of the World Health Organization Classification of Haematolymphoid Tumours: Lymphoid Neoplasms. Leukemia 36(7):1720–1748

14. Hwang J et al (2021) The Incidence of Epstein-Barr Virus-Positive Diffuse Large B-Cell Lymphoma: A Systematic Review and Meta-Analysis. Cancers (Basel) 13(8)

15. International Non-Hodgkin’s Lymphoma Prognostic Factors (1993) A predictive model for aggressive non-Hodgkin’s lymphoma. N Engl J Med 329(14):987–994

16. Oyama T et al (2007) Age-related EBV-associated B-cell lymphoproliferative disorders constitute a distinct clinicopathologic group: a study of 96 patients. Clin Cancer Res 13(17):5124–5132

17. Beltran BE et al (2011) EBV-positive diffuse large B-cell lymphoma of the elderly: a case series from Peru. Am J Hematol 86(8):663–667

18. Matsumoto K et al (2018) Anemia Associated with Worse Outcome in Diffuse Large B-Cell Lymphoma Patients: A Single-Center Retrospective Study. Turk J Haematol 35(3):181–184

19. Hong J et al (2014) Anemia as a useful biomarker in patients with diffuse large B-cell lymphoma treated with R-CHOP immunochemotherapy. Cancer Sci 105(12):1569–1575

20. Hans CP et al (2004) Confirmation of the molecular classification of diffuse large B-cell lymphoma by immunohistochemistry using a tissue microarray. Blood 103(1):275–282

21. Sehn LH, Salles G (2021) Diffuse large B-cell lymphoma. N Engl J Med 384(9):842–858

22. Park S et al (2007) The impact of Epstein-Barr virus status on clinical outcome in diffuse large B-cell lymphoma. Blood 110(3):972–978

23. Shimoyama Y et al (2009) Age-related EBV-associated B-cell lymphoproliferative disorders: diagnostic approach to a newly recognized clinicopathological entity. Pathol Int 59(12):835–843

24. Okamoto A et al (2015) Prognostic significance of Epstein-Barr virus DNA detection in pretreatment serum in diffuse large B-cell lymphoma. Cancer Sci 106(11):1576–1581

Associated Data

Supplementary Materials

Supplementary Material 1 (DOCX 38.7 KB)

Data Availability Statement

No datasets were generated or analyzed during the current study.