Nivolumab in Lymphomatoid Granulomatosis (LYG) and Other Epstein-Barr Virus (EBV)-Positive Lymphoproliferative Disorders (LPDs) and Non-Hodgkin Lymphomas (NHLs) with Systemic and/or CNS Involvement: Preliminary Analysis of Efficacy and Safety from an Ongoing Phase II Study

Background: Programmed cell death protein-1 (PD-1) is a signaling molecule on the surface of T-cells that suppresses the cytotoxic effects of T-cells on tumor cells. We previously demonstrated increased PD-1 expression on CD8+ T-cells compared to healthy controls in pts with LYG, which could be reversed with immunotherapy using interferon alpha-2b (Melani et al. Lancet Haematol. 2023). Additionally, elevated expression of programmed death-ligand 1 (PD-L1) has been observed in other EBV-LPDs and NHLs as well as in chronic viral infections, such as EBV. Given this likely shared disease pathobiology, we hypothesized that treatment with the anti-PD-1 antibody, nivolumab, may reverse T-cell exhaustion and result in anti-tumor responses in pts with LYG and other EBV-LPDs and NHLs, including those with systemic and/or CNS involvement. Methods: Relapsed/refractory (R/R) EBV-LPD and NHL pts age ≥ 12y with adequate organ function were eligible. Untreated pts were eligible if EBV-LPD. Exclusions included prior use of PD-1/PD-L1/PD-L2/CD137/CTLA-4 antibodies, solid organ transplant, and HIV infection. Pts with a known immunodeficiency or autoimmune illness were eligible if not requiring steroids or immunosuppression within 14d of study. CNS involvement was permitted if no seizure activity within 4w of study. Nivolumab 480 mg IV was given q4w for up to 2y. Pts who achieved CR discontinued nivolumab after 1y of treatment. Responding patients with relapse or progression within 1y of discontinuing treatment were eligible for retreatment. Baseline evaluation included CT, PET, MRI brain, flow cytometry of peripheral blood and CSF, BM biopsy, and optional tumor biopsy. Restaging CT scans were performed after cycles 3, 6, 13, and 19, and end of treatment (EoT). PET was performed after cycles 1, 3, and EoT. Surveillance CT was performed q3m for 1y, q6m for yrs 2-5, and annually thereafter. Results: 11 pts were enrolled and treated; 6 (55%) with EBV-LPD and 5 (45%) with EBV-NHL (all DLBCL). EBV-LPD subtypes included 4 pts with G1-2 LYG and 1 pt each with EBV B-LPD of the CNS and CAEBV. Median age was 48y (range 20-67), and all pts had stage III/IV disease with extranodal involvement of the lung most common in 64% (7/11) and CNS disease in 18% (2/11). Median baseline absolute CD8 count (cells/mcL) was 86 (range 9-1237; normal: 178-853), CD4 count (cells/mcL) was 255 (range 99-1304; normal: 359-1565), and EBV VL in plasma (Log10 IU/mL) was undetected (range undetected-6.78; normal: <2.3). Median prior therapies were 1 (range 0-6) with 3 previously untreated LPD pts and 50% of previously tx pts were refractory to last tx. In 10 pts evaluable for response(1 came off tx prior to restaging), ORR and CR rate were both 60% (6/10). CR rate was 67% (4/6) in EBV-LPD, including 75% (3/4) with LYG, and 50% (2/4) in EBV-NHL (Fig 1A). One (50%) of 2 EBV-LPD pts with CNS involvement achieved CR after 3C of therapy and remains in ongoing response. Median TTR was 3.0m with 50% (3/6) responses ongoing from 3-56m after first response (Fig 1B). With a median follow up of 32m, 2y PFS and OS was 33% and 72% overall. By cohort, median PFS was 36m in EBV-LPD and 1m in EBV-NHL (p=0.003), with a 2y OS of 100% in EBV-LPD and 40% in EBV-NHL (p=0.04). Three EBV-NHL pts died, all of disease progression. One EBV-NHL pt stopped therapy after 2C due to radiographic disease progression in the kidney, but later developed CR without further therapy and remains in remission 56m after stopping therapy consistent with initial pseudoprogression. Adverse events (AEs) were as expected for nivolumab therapy with the most common AEs (% pts) including maculopapular rash (27%), diarrhea (18%), hypothyroidism (18%), and fatigue (18%). One pt developed an immune-mediated G2 myositis and G3 hepatitis following 3C of therapy which resolved after drug discontinuation and corticosteroids. Conclusion: Nivolumab is well tolerated in pts with LYG and other EBV-LPDs and NHLs without unexpected toxicities. Preliminary activity, including durable CRs, is noted across all pt cohorts, including 1 EBV-LPD pt with CNS involvement. With small numbers, statistically longer PFS and OS was observed with nivolumab in LYG and EBV-LPD compared to EBV-NHL, indicating that monotherapy may be sufficient for LYG and other EBV-LPDs, but combination therapy is likely needed for pts with EBV-NHL. Study enrollment continues to better assess the activity and durability of nivolumab in LYG and other EBV-LPDs and NHLs.