Torque teno virus in pediatric CAR-T recipients

Abstract BACKGROUND Chimeric antigen receptor T-cell (CART) therapy has improved outcomes in relapsed/refractory (R/R) patients with B-acute lymphoblastic leukemia (B-ALL) and diffuse large B-cell lymphoma (DLBCL), although immune-mediated toxicities remain a concern. Torque Teno Virus (TTV), a highly prevalent and ubiquitous DNA virus, is the main component of human virome. A recent study in adults suggests that plasma TTV viral load (TTV-VL) in CART patients may correlate with immune effector cell-associated neurotoxicity syndrome (ICANS) severity and progression-free survival (Benzaquén, 2024). AIMS To analyze plasma TTV-VL kinetics in pediatric CART recipients, its association with complications, outcome, lymphocytes and CART subsets, and its predictive value. METHODS Retrospective single center study including patients <18 years treated with CART CD19 or CD19/22 (2019-2023) for R/R B-cell malignancies. Blood samples were collected pre-lymphodepletion (pre-LD), at CART infusion (0d), at +7, +14, +21, +28, +35, +42, +60 and + 90 days (d), and at the onset of immunotoxicities, which were graded according to ASTCT consensus (Lee, 2019). Immune cell typing was performed by flow cytometry. CART was monitored via HIV qPCR and flow cytometry. TTV in plasma was measured with TTV R-GENE® (Biomérieux, France). RESULTS Twenty-seven patients were included, treated with CD19 (70.4%) or CD19/22 (29.6%) CART. Most were male (59.3%), mean age 12.5 years (SD 5.6), predominantly Caucasian (55.6%) and Latin American (37%); 18.5% Down Syndrome. Underlying disease was 81.5% B-ALL, 18.5% DLBCL. Cytokine release syndrome (CRS) occurred in 85.2% patients (median 3 days, [1,6.5]), ICANS in 29.6% (mean 8.13 days, SD 2.6), and hemophagocytic lymphohistiocytosis (HLH) in 37% (mean 8 days, SD 2.6). Tocilizumab was administered in 40.7%, siltuximab in 29.6%, steroids in 48.1%, and anakinra in 33.3% cases. At +30d, 22.2% had pancytopenia, 25.9% had had infections, and 51.9% had been admitted to pediatric intensive care unit (PICU). B-cell recovery without relapse occurred in 29.6% (mean 92 days, SD 61.4), 40.7% patients relapsed (mean 139 days, SD 105), and 11.1% had progressive disease (PD). Twenty patients (74.1%) received treatment after CART, 48.1% including hematopoietic stem cell transplantation (HSCT). Survival was 55.6% (median follow up 752 days, [134.5,1421]). Main causes of death were PD (64.3%) and HSCT-related complications (14.3%). TTV was analyzed in 173 samples. Pre-LD, it was detectable in 81.5%, with no differences regarding patient characteristics. It was significantly lower at +7 (p=0.0435) and peaked at +60d. Receiver Operating Characteristics (ROC) curve analysis at Pre-LD revealed a TTV-VL threshold of 7.52 log copies/ml predicted CRS (specificity [Sp] 50%, sensitivity [Se] 95%), 2.02 grade 3-4 CRS (Sp 100%, Se 37.5%), 3.17 tocilizumab necessity (Sp 92.3%, Se 66.7%), 5.94 steroids (Sp 40%, Se 91.7%), 4.91 PICU admission (Sp 70%, Se 75%), and 6.14 log copies/ml B-cell recovery (Sp 92.9%, Se 37.5%). Each increase in pre-LD TTV-VL diminished the probability of receiving tocilizumab by 45.1% (p=0.0496). At 0d, TTV-VL was inversely correlated with CRS grade (p=0.0165). ROC curve analysis showed that 6.51 log copies/ml predicted CRS (Sp 66.7%, Se 100%) and 2.93 grades 3-4 (Sp 91.7%, Se 100%); 3.29 ICANS (Sp 76.9%, SE 100%), HLH (Sp 90%, Se 75%), infections at +30d (Sp 80%, Se 50%), PICU admission (Sp 90%, Se 75%), and steroid treatment (Sp 90%, Se 75%). Treatment with tocilizumab was predicted by 3.78 (Sp 100%, Se 100%) and siltuximab by 1.64 log copies/ml (Sp 100%, Se 50%). At +7d, a threshold of TTV-VL of 4.39 log copies/ml classified patients suffering infection (Sp 75%, Se 80%) and 5.62 those admitted to PICU (Sp 30%, Se 100%). TTV-VL could predict mortality at different points: 4.99 log copies/ml at +28d (Sp 80%, Se 75%), 4.2 at +35d (Sp 100%, Se 75%), 3.77 at +42d (Sp 100%, Se 75%), and 6.08 at +60d (Sp 100%, Se 66.7%). TTV-VL was directly correlated with absolute NK cells count at +7 (p=0.0371), double negative lymphocytes at +21 (p=0.0356), and regulatory T lymphocytes at +42d (p=0.0386). The correlation resulted inverse with total T lymphocytes at +21d (p=0.0064). TTV-VL at 0d inversely correlated with HIV qPCR at +28d (p=0.0231). We found no associations with CART flow subsets. CONCLUSIONS In pediatric CART recipients, TTV-VL does not correlate with CART monitoring methods, but could predict toxicity and outcomes.