A retrospective analysis of EBV‐DNA status with the prognosis of lymphoma

Abstract Epstein–Barr virus (EBV) infection is proved to be associated with clinicopathology of lymphoma. However, little is known about the relationship between EBV‐DNA status after treatment and prognosis. In this study, real‐time polymerase chain reaction (PCR) was used for quantitative detection of EBV‐DNA load in peripheral blood of all 26,527 patients with lymphoma, and the clinical characteristics and prognosis of 202 patients were retrospectively analysed, including 100 patients with positive EBV‐DNA and 102 randomly selected patients with negative EBV‐DNA. We found that the average rate of EBV‐DNA positivity in lymphomas was 0.376%, and EBV‐DNA‐positive patients presented higher risk with elevated lactate dehydrogenase (LDH) and β2‐MG level, B symptoms, secondary hemophagocytic syndrome and lower objective response rate compared to EBV‐DNA‐negative patients. Multivariate analysis revealed EBV‐DNA‐positive patients had inferior progression‐free survival (PFS) and overall survival (OS) and EBV‐DNA level before treatment was related to PFS but not OS of T/NK cell lymphoma. In T/NK cell lymphoma, EBV‐DNA converting negative after treatment was correlated with better PFS but not OS, and second‐line therapy could induce more EBV‐DNA‐negative conversion compared to CHOP‐based therapy. In all, EBV‐DNA positivity before treatment can be a biomarker representing the tumour burden and an independent prognostic factor. EBV‐DNA‐negative conversion after treatment is a good prognostic factor for T/NK cell lymphomas.

immunodeficient individuals, which are true virally driven lymphomas, such as PTLD and HIV-associated immunoblastic lymphoma, and those occurring in immunocompetent individuals. The latter group includes Burkitt lymphoma, Hodgkin's lymphoma, diffuse large B-cell lymphoma, extranodal NK/T cell lymphoma nasal type, and so on 5 and EBV infection is a cofactor rather than the driving influence.
Epstein-Barr virus spreads through oral cavity, proliferates in the throat, and then lurks in B lymphocytes, usually presenting as a latent infection without clinical symptoms. Given the long-term latent infection, EBV antibody detection is insufficient to diagnose EBV-associated lymphomas. 6 Epstein Barr encoded RNA (EBER) and EBV-DNA are used to define reactivation of EBV. EBER in situ hybridization (FISH) assay is considered to be the gold standard for the detection and diagnosis of EBV active infection. 7 However, biopsies are usually not performed when tumour biopsy tissue is difficult to obtain or when the patient is refractory or relapsed lymphoma. 8 Detection and quantification of EBV nucleic acids in peripheral blood by polymerase chain reaction (PCR) has also been widely used in the diagnosis and monitoring of EBV-associated lymphomas. 9 However, whether EBV-DNA conversion to negative after treatment and qualitative results of EBV-DNA level before treatment could affect prognosis are unclear. To elucidate these questions, we analysed the clinical characteristics and prognosis of immunocompetent patients with EBV associated lymphoma.
HLH is a rapidly progressive disease with a high fatality rate, which can occur secondary to lymphoma or severe pathogen infection, and its main clinical manifestations are persistent fever, hepatosplenomegaly, and pancytopenia. Previous studies showed that EBV-associated HLH was the most prevalent subtype. 10 In order to study the relationship between EBV active infection and the development of HLH in lymphoma patients, we also analysed the clinical characteristics and prognosis of EBV-associated lymphoma patients with HLH.

| Patients
The positivity rates of EBV-DNA in peripheral blood of 26,527 patients who were definitively diagnosed as lymphoma between May

| Treatment
The first-line chemotherapy regimen was based on cyclophosphamide, doxorubicin, vincristine and prednisone (CHOP). The

| Clinical evaluation index
The response evaluation was divided into complete response (CR), partial response (PR), stable disease (SD), and progressive disease    5 × 10 2 copy/ml was defined to be the critical value. Higher than 5 × 10 2 copy/ml was considered to be EBV-DNA positive. EBV-DNA load in peripheral blood was measured before initial treatment and after four cycles of chemotherapy.

| Statistical analysis
Univariate analysis was performed using the Kaplan-Meier model, multivariate analysis was performed using the Cox regression model, and the differences were assessed using the log-rank test. The qualitative data were compared using the χ 2 test. p < 0.05 was considered to be statistically significant different. All statistical analyses were performed using SPSS 26.0 software. Taken together, T/NK cell lymphomas were more commonly associated with EBV-DNA active infection ( Table 1).

| Clinical characteristics between EBV-DNApositive and -negative patients
The patients' characteristics are listed in Table 2. The total number

| Different EBV-DNA status after treatment
EBV-DNA status in patients' peripheral blood was re-detected after four cycles of treatment. The negative conversion rates of different subtypes of lymphomas were shown in Figure 3, in which indolent B-cell lymphoma is higher than other subtypes. Kaplan-Meier analysis revealed that EBV-DNA converting negative after treatment was correlated with improved PFS of T/NK cell lymphoma (Table 3, p = 0.001) but had no effect on OS (

| Prognostic analysis
The univariate analysis showed that EBV-DNA, age, response status,  Multivariate analysis found that age, EBV-DNA, response status, and HLH were significantly independent prognostic factors for both poor PFS and OS (p < 0.05, Table 4).

| EBV active infection was related to secondary HLH
Among the 26,527 recruited lymphoma patients, a total of thirteen patients were accompanied with HLH. EBV-DNA was positive in all HLH patients, indicating that EBV active infection was related to the occurrence of HLH (p < 0.001, Table 2). Univariate and multivariate analyses indicated that HLH was an independent factor affecting the prognosis (Table 4). In addition, we found that the HLH patients had higher levels of LDH (p = 0.045), β2 MG (p < 0.001), and EBV-DNA (p < 0.001) compared to those without HLH. The average value of EBV-DNA in patients with HLH (36 × 10 5 copy/ml) was significantly higher than that in EBV-positive patients without HLH (6 × 10 5 copy/ ml) (p < 0.001).

| DISCUSS ION
Epstein-Barr virus, a member of the human herpesvirus family, is widespread in world's population 11 and is carried as a latent asymptomatic infection in individuals. 12 Persistent EBV active infection is considered a high-risk factor for nasopharyngeal carcinoma and malignant lymphomas, including Hodgkin and non-Hodgkin lymphomas. 5 B cells are known to be the primary lymphoid target of EBV infection. EBV-associated NK and T cell lymphoproliferative diseases are more common in Asia. 13 A study in China indicated that EBV-DNA was more frequently detected in T/NK cell lymphomas than in B cell lymphomas. 14 A similar result was observed in our study, and EBV-DNA positivity rate was higher in T/NK cell lymphomas than that in B-cell lymphomas.
Some studies revealed that EBV-DNA positivity before treatment could reflect the tumour burden. [15][16][17][18][19] We found that EBV- Our multivariate analysis also revealed that EBV-DNA positivity was an independent prognostic factor. However, no reports on EBV-DNA level to the prognosis of lymphoma patients were published as we known. We found that EBV-DNA level before treatment could affect PFS of patients with T/NK cell lymphoma, but had no effect HLH is a rapidly progressive, highly fatal disease that may occur either as a result of the lymphoma disease itself or pathogen infection during immunosuppression. 22 Secondary HLH is often associated with a variety of underlying diseases, such as infection, tumour, and rheumatic diseases. Among the infectious factors, EBV infection is the most important one. We found that HLH patients in the study were all EBV-DNA positive, indicating that EBV active infection increased the risk of HLH in lymphoma patients. Our study also showed that the prognosis of HLH patients was significantly worse, which may be related to the high copy number of EBV-DNA in lymphoma patients.
In our study, we found that EBV active infection rate in T/NK cell lymphoma was significantly higher than that in B cell lymphoma,

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data sets used and/or analysed during the current study are available from the corresponding author on reasonable request.