Successful combined venetoclax to PD‐1 blockade and ruxolitinib for refractory Epstein–Barr virus‐associated haemophagocytic lymphohistiocytosis

Haemophagocytic lymphohistiocytosis (HLH) is a rare and life-threatening syndrome characterized by immune dysregulation and excessive inflammation. Epstein–Barr virus-associated HLH (EBV-HLH) is a type of secondary HLH which progresses rapidly and lacks effective and safe treatment options.1 The immune checkpoint inhibitor (ICI), programmed cell death protein 1 blockade (PD-1 blockade), has demonstrated effects in treating EBV-HLH.2, 3 However, only 30%–50% of EBV-HLH patients respond well to PD-1 blockade.3 For those who did not response, there is an increasing interest in exploring collaborative regimens. BCL-2 inhibitor venetoclax has been applied in haematological malignancies, attributing to its pro-apoptotic effect. Recent research suggests that combining BCL-2 inhibitors to ICIs proved enhanced anti-tumour effects.4 To illustrate the potential assistance of BCL-2 inhibitors, we present two cases of EBV-HLH patients who failed in responding to PD-1 blockade, surprisingly achieved remission in both of HLH and EBV infection after combined venetoclax. A 53-year-old male presented with a recurrent fever lasting over 2 weeks. After evaluation, EBV-HLH was diagnosed (details in Table S1). EBV-infected lymphocytes were confirmed in B and NK cells. Noteworthy, flow cytometry analysis of bone marrow cells revealed that 2% of NK cells exhibited abnormal immunophenotype. He initially received etoposide and dexamethasone, along with rituximab (375 mg/m2, weekly) and ruxolitinib (10 mg, twice daily). The fever subsided after 2 weeks, but cytopenias and EBV infection persisted. Three weeks later, he began experiencing a low-grade fever (37.3–37.6°C), anti-PD-1 antibody (Sintilimab) was introduced but laboratory parameters kept deteriorating. Considering that HLH was at relapse, therapy strategy switched to Venetoclax (100 mg daily) combined with weekly PD-1 blockade, along with ruxolitinib (5 mg, twice daily). Remarkably, EBV-DNA became undetectable within just 3 days, accompanied by the cessation of fever. Within 1 week, neutrophil and platelet counts reached normal levels. Hepatic dysfunction and coagulopathy resolved, and the serum ferritin level declined significantly. Upon discharge from the hospital on day 36, complete remission was achieved. EBV-DNA in peripheral blood remained undetectable. Ruxolitinib was tapered and eventually stopped after 1 month. As a maintenance therapy, PD-1 blockade was administrated every 3 weeks. He has been followed up for over 6 months and remains in good health (Figure 1). A 28-year-old female presented with recurrent fever lasting over 3 months. After evaluation, EBV-HLH was diagnosed (details in Table S1). She initially received etoposide and dexamethasone, along with ruxolitinib (10 mg, twice daily). During therapy, HLH remained in stable, but EBV infection persisted. Eight weeks later, a recurrence of fever (maximum 38.9°C) prompted her transfer to our hospital. EBV infection remained positive (3.4 × 104 copies/mL) and analysis of EBV-infected lymphocytes indicated the presence of virus in B and NK cells. Flow cytometry analysis of bone marrow cells showed that 81.7% of NK cells held an abnormal immunophenotype (CD56+, CD7−, CD57−, CD2+, CD38+). Ruxolitinib was increased to 15 mg twice daily for better control of HLH, and a 200 mg dose of PD-1 blockade was administered for EBV infection. Five days later, as all indicators deteriorated continually, relapse of HLH was confirmed. Salvage treatment with DEP regimen was conducted but proved ineffective in 10 days. Venetoclax (100 mg, daily) was initiated along with continuous PD-1 blockade and ruxolitinib. Surprisingly, the fever resolved within 48 h. One week later, cytopenia recovered. Serum ferritin and cytokines decreased significantly. Two weeks later, EBV-DNA decreased to 7.3 × 103 copies/mL. She was discharged from the hospital on day 36. PD-1 blockade (200 mg, every 3 weeks); venetoclax and ruxolitinib were administrated as maintenance therapy. One month later, EBV-DNA climbed back to 1.05 × 105 copies/mL but without any evidence of HLH relapse. Two months after discharge, EBV-DNA became undetectable. She has been followed up for over 18 months and remains continuously free from EBV infection and HLH (Figure 2). The expression of PD-1 and PD-L1 on T cells (CD3+) and NK cells (CD56+) was tested by flow cytometry. Before treatment, the expression of PD-L1 on NK cells was elevated (T cells: PD1+ 82.9%, PDL1+ 1.6%; NK cells: PD1+ 0.1%, PDL1+ 66.7%), which gradually decreased along with treatment and dropped to 0% after 18 months (Figures S1 and S2). Persistent EBV infection often contribute to the relapse of HLH in EBV-HLH patients. In these two cases, the application of mono-PD-1 blockade failed in eliminating EBV, resulting in the relapse of HLH. In these settings, the combined venetoclax have surprisingly received remarkable results. In these two cases, localization of EBV-infected on NK cells and B cells in peripheral blood and identified abnormal phenotypic NK cells in bone marrow cells indicated specific EBV infection of NK cells. In case 2, high PD-L1 expression on NK cells was detected in the peripheral blood (66.7%), but PD-L1 expression on T cells was normal (1.6%), which also confirmed the EBV infection of NK cells. These EBV-infected NK cells may contribute to the resistance to mono-PD-1 blockade at first place. EBV-encoded latent membrane protein 1 (LMP-1) can inhibits apoptosis, promotes Bcl-2 expression and cell proliferation.5, 6 Additionally, inhibition of BCL-XL induces apoptosis in EBV–associated T/NK-cell lymphoma cell line.7 The combination of venetoclax may eliminate EBV by directly inducing apoptosis in EBV-infected NK cells. In case 1, EBV continued to be negative after rapid clearance, but in case 2, there was a resurgence of EBV-DNA in later period. The different pattern of EBV clearance between two cases suggests the efficacy of BCL-2 inhibitors in EBV-HLH operates differently in long and short term respectively. BCL-2 has been identified playing an important role in immunomodulatory, resulting in an increase in the proportion of CD8+ and CD4+ T effector cells.4, 5 Overall, in the short term, the pro-apoptotic effect of BCL-2 inhibitor promotes the apoptosis of EBV-infected cells, leading to rapid control of HLH and EBV. However, this effect may not last. In the long term, the immunomodulatory effect of BCL-2 inhibitors assists PD-1 blockade in restoring the function of EB virus-specific T cells, ensuring continuous control of EBV and HLH. This is the first report on the combined use of BCL-2 inhibitors to support PD-1 blockade in EBV-HLH patients, leading to the elimination of EBV infection and remission of HLH. This study provides a potential resolution to address the variability in the effectiveness of PD-1 blockade in EBV-HLH. In situations where PD-1 blockade proves ineffective and the application of intensive chemotherapy or HSCT entails substantial risks, co-administration with BCL-2 inhibitors may present therapeutic benefit. XH designed and supervised this study; SC and DW coordinated the study; YS and XY conducted the data analysis and wrote the manuscript. Other authors did the research. None. This work was supported by the National Natural Science Foundation of China (82020108003), National Center for Translational Research Open Project (2021ZKMB04) and China Postdoctoral Science Foundation (2023M732540). All authors declare no competing interests. The study was approved by the Ethics Committee at The First Affiliated Hospital of Soochow University. We have obtained consent to participate from the participants. We have obtained consent to publish from the participants. All relevant raw data will be freely available to any researcher wishing to use them for non-commercial purposes, without breaching participant confidentiality. For publication-related data, please contact the corresponding author. Data S1. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.