PEG-aspargase and DEP regimen combination therapy for refractory Epstein–Barr virus-associated hemophagocytic lymphohistiocytosis

HLH, an inflammatory cytokine storm caused by uncontrolled immune responses, is a rare life-threatening disease. HLH is divided into two major categories: primary and acquired. The latter is often caused by infections, malignant tumors, or autoimmune diseases. Infection-related HLH is mostly attributable to EBV infections, which account for approximately 70 % of infection-related cases [1]. EBV-HLH is common in Asian persons; however, it has also aroused attention in non-Asian countries in recent years. One study [10] has shown that adults with EBV-HLH have worse prognosis than children with this disease. However, there are no published large-scale clinical studies that adequately examine the prognoses of adults with HLH. In this study, 82.1 % of participants with EBV-HLH were adults. Thus, age may have some impact on the interpretation of the results of this study.

Immune system deficiencies are believed to be involved in the pathogenesis of EBV-HLH. It is well established that EBV and its antigens induce massive release of cytokines, thereby activating monocytes/macrophages and causing damage in multiple organs [11]. A series of Japanese-based studies have found that T cells and NK cells are targeted by EBV infection in patients with EBV-HLH [12, 13]. Fox et al. [13] reported abundant EBV genomes (12,900–1,816,550 genomes/10⁶ cells) in the circulating NK cells of such patients. Another study [14] also showed enhancement of proliferation of latent EBV-infected lymphocytes and that they contained new surface epitopes. Interestingly, etoposide (VP-16) inhibits EBV core epitope synthesis and has anti-EBV effects. One study [15] has shown that early administration of VP-16 and dexamethasone may reduce the mortality of EBV-HLH. Imashuku et al. used dexamethasone and VP-16 as the basic treatment regimen for 78 patients with EBV-HLH (73 of whom were ≤15 years old) and achieved a 43-month OS of 75.6 % [4]. Imashuku et al. [16] also reported enhancement of the long-term survival of patients with EBV-HLH treated with a regimen containing etoposide within 4 weeks of diagnosis; however, they found that treatment of patients with EBV-HLH with regimens containing cyclosporine made no significant difference to treatment efficacy or OS. Thus, the HLH-04 regimen is not significantly better than the HLH-94 regimen for treating patients with EBV-HLH. Although we found no significant difference in outcomes between 4-week HLH-04 and 4-week HLH-94 regimens in a previous study [2], patients with EBV-HLH who received these regimens had significantly better OS than patients who did not receive etoposide treatment. We therefore believe that etoposide-based immunochemotherapy improve the survival of patients with EBV-HLH. However, in a study by Henter et al. [3], >30 % of patients with EBV-HLH did not achieve response with the HLH-94 treatment regimen. To date, there is no established second-line treatment for patients with refractory HLH who do not respond to standard treatment regimens. Allo-HSCT-induced immune reconstitution in patients with EBV-HLH can result in them regaining the ability to remove EBV, thereby prolonging their survival and ultimately curing HLH. In one Japanese cohort, allo-HSCT resulted in an 85.7 % 10-year OS for patients with EBV-HLH [17]. However, allo-HSCT can cause an increase in transplant-related mortality in patients with poorly controlled EBV-HLH. In a previous study [2] in which we performed allo-HSCT in five patients with EBV-HLH, two of the five patients did not achieve response and died within approximately 3 months of diagnosis; one patient who did not achieve response after relapse died 7 months after diagnosis; and the remaining two patients, who received allo-HSCT while their disease was stable, survived. Thus, patients with EBV-HLH who do not achieve response before receiving allo-HSCT have a significantly reduced OS after transplantation, which is consistent with a previous report [5]. Hence, salvage treatment to stabilize the disease in patients with refractory EBV-HLH may be important in extending their survival after allo-HSCT.

Previous studies have demonstrated that anti-thymocyte globulin [18], tumor necrosis factor-alpha antagonists [19], anti-CD52 monoclonal antibody [9], anti-CD20 monoclonal antibody [20], and the DEP regimen [6] can be used as salvage therapy for refractory EBV-HLH. However, there is no established standard treatment for patients with EBV-HLH who fail to achieve response with the HLH-94 regimen. Beutel et al. [12] used rituximab (a CD20 monoclonal antibody) to label EBV-infected B lymphocytes and successfully treat B lymphocyte-related HLH. Imashuku et al. [21] postulated that EBV initially infects B cells and then continuously generates viral particles that further infect T or NK cells, suggesting that rituximab may be an effective treatment for EBV-infected T or NK cells. Based on the above theory, researchers have used rituximab to treat EBV-HLH; however, only retrospective studies and case reports are available. Chellapandian et al. [20] retrospectively analyzed the clinical information of 42 patients with EBV-HLH who received a rituximab monoclonal antibody treatment regimen. Rituximab-containing regimens appeared well tolerated and improved clinical status in 43 % of patients. However, no large follow-up studies of rituximab therapy in patients with EBV-HLH have been reported since 2013. In the study, one patient did not achieve response after a combination of rituximab monoclonal antibody and the HLH-94 regimen and further received the L-DEP regimen to achieve CR, which was followed by allo-HSCT. In our previous study, we used liposomal doxorubicin, etoposide, and high-dose methylprednisolone (i.e., the DEP regimen) to treat patients with EBV-HLH who did not achieve response after HLH-94; 72.7 % of those patients achieved PR or better treatment outcomes [6]. However, the duration of the consecutive treatment effect of the DEP regimen is relatively short and this regimen carries a significant risk of gastrointestinal bleeding. In this study, we modified the dosage and duration of methylprednisolone in the DEP regimen and combined it with PEG-aspargase (i.e., the L-DEP regimen) as salvage therapy for refractory EBV-HLH; this combination increased the overall response rate to 85.7 % and enabled more patients to receive allo-HSCT, including all those who achieved CR (100 %). Thirteen of the 28 patients with EBV-HLH received allo-HSCT. Only one patient received allo-HSCT from an HLA-identical donor and the remaining 12 patients underwent haploid allo-HSCT, which solved the limitation of donor problem due to “one-child” policy in China. Ten of these 13 patients survived (76.9 %) after allo-HSCT. Fifteen of the 28 study patients did not receive allo-HSCT for various reasons. Despite four (26.7 %) of the 15 patients are continuously receiving maintenance treatment at the time of publication, the remaining 11 patients do not survive.

L-DEP is a modified regimen in which PEG-aspargase is combined with the DEP regimen. PEG-aspargase, a new asparaginase preparation with a chemical coupling of modification in polyethylene glycol, not only retains the biological activity of L-asparaginase but also reduces its immunogenicity and its half-life from 20 h to 5.5 days, which may explain the improved treatment efficacy achieved by combining the DEP regimen with PEG-aspargase. Ando et al. [22] have reported that L-asparaginase induces in vitro apoptosis in NK lymphoma cells. In the present study, 15 of the 28 patients (53.6 %) demonstrated abnormal phenotypes of NK cells in the bone marrow as assessed by flow cytometry. In addition, these patients had significantly fewer EBV-DNA copies 2 weeks after L-DEP regimen than pretreatment (P = 0.048), EBV-DNA being undetectable after L-DEP regimen in four patients. We speculate that PEG-aspargase targets EBV-infected target cells, including T and NK cells, because these cells may not be able to synthesize L-asparagine themselves [22]. After entering cells, PEG-aspargase induces hydrolysis of L-asparagine, thus preventing target cells without L-asparagine (essential amino acids for protein syntheses) from synthesizing the corresponding proteins, ultimately inhibiting cellular proliferation and resulting in decline in EBV-DNA. An in vitro study by Jinta et al. [23] demonstrated that L-asparaginase dose-dependently reduces the number of EBV-positive T and NK cells; while not affecting the peripheral blood mononuclear cells of normal donors, suggesting that L-asparaginase inhibits the proliferation of EBV-positive T cells and NK tumor cells. A previous study [24] showed that high viral load of EBV-DNA may be a risk factor for poor outcomes. The patients less than 1 × 10³ copies/mL of EBVDNA showed a significantly higher clinical response and longer overall survival than those with high viral load of EBV-DNA. Another study [2] has shown that EBV-negativity is an independent risk factor for prognosis. Patients with EBV-negative conversion have a significantly better OS than patients who are continuously EBV-positive. During L-DEP treatment, the clinical status of patients with EBV-HLH improves, usually accompanied by reduction in EBV-DNA copies or conversion to EBV-DNA negativity. EBV replication is greater in patients with EBV-HLH with more serious disease manifestations. In this study, prognosis was not significantly associated with numbers of EBV-DNA copies before the L-DEP regimen (P > 0.05). However, reduction in EBV-DNA copies was associated with prognosis. After L-DEP treatment, reduction of ≥1 × 10²-fold in EBV-DNA copies was significantly associated with prolonged survival time (P = 0.018).

The DEP regimen is associated with increased risk of bleeding, especially from the gastrointestinal tract [6]. We therefore modified the DEP regimen by retaining high-dose corticosteroid therapy to suppress the inflammatory cytokine storm and rapidly control the disease. However, we rapidly decreased the dose of methylprednisolone and stopped it within 10 days, thus reducing the adverse effects of long-term and high-dose glucocorticoid therapy. In this study, no patient developed ≥grade 3 gastrointestinal bleeding. The major adverse effects of PEG-aspargase are liver dysfunction, decreased fibrinogen concentrations, and pancreatitis. In this study, no significant improvements in ALT or lactate dehydrogenase concentrations were found in the study patients in the fourth week after L-DEP treatment. In addition, total bilirubin concentrations 2 and 4 weeks after treatment did not differ significantly from pretreatment values (P > 0.05). However, other HLH indicators improved compared with pretreatment values, suggesting that impairment in liver function was caused by PEG-aspargase, especially in the five patients with higher bilirubin concentrations 2 and 4 weeks after treatment than pretreatment. However, this liver dysfunction was alleviated by liver protection therapy. In this study, fibrinogen concentrations were significantly lower 2 and 4 weeks after treatment than pretreatment in the study patients (P < 0.05), suggesting that this change was associated with coagulation disorders caused by PEG-aspargase. Supplementary human fibrinogen and fresh frozen plasma therapy prevented any serious bleeding in this cohort. Three patients had increased serum amylase concentrations after treatment, one of whom developed abdominal pain and radiographic changes. After somatostatin therapy, these patients’ amylase concentration reverted to normal values. In addition, we found that a low-fat diet after PEG-aspargase treatment appeared to reduce the incidence of adverse reactions.