BackgroundChimeric antigen receptor-modified (CAR) T-cell immunotherapy is a novel promising therapy for treatment of B-cell malignancy. Cytokine release syndrome (CRS) and infection are the most common adverse events during CAR T-cell therapy. Similar clinical presentation of concurrent CRS and infection makes it difficult to differentially diagnose and timely treat the condition.MethodsWe analyzed the features of infection events during the first 30 days after CAR T-cell infusion (CTI) in 109 patients from three clinical trials (ChiCTR-OPN-16008526, ChiCTR-OPC-16009113, ChiCTR-OPN-16009847). Based on the dynamic changes of interleukin (IL)-6 and ferritin, we proposed the “double peaks of IL-6” pattern as a feature of life-threatening infection during the first 30 days after CTI. Meanwhile, we screened candidate biomarkers from 70-biomarker panel to establish a prediction model for life-threatening infection.ResultsIn this study, 19 patients (17.4%) experienced a total of 19 infection events during the first 30 days after CAR T-cell infusion. Eleven patients (10.1%) had grade 4–5 infection, which were all bacterial infection and predominantly sepsis (N = 9). “Double peaks of IL-6” appeared in 9 out of 11 patients with life-threatening infection. The prediction model of three-cytokines (IL-8, IL-1β and interferon-γ) could predict life-threatening infection with high sensitivity (training: 100.0%; validation: 100.0%) and specificity (training: 97.6%; validation: 82.8%). On base of the aforementioned methods, we proposed a workflow for quick identification of life-threatening infection during CAR T-cell therapy.ConclusionsIn this study, we worked out two diagnostic methods for life-threatening infection during CAR T-cell therapy by analyzing inflammatory signatures, which contributed to reducing risks of infection-induced death.

Advanced central nervous system (CNS) lymphoma is an exclusion criterion for most chimeric antigen receptor (CAR) T-cell studies due to the associated levels of neurotoxicity. In this study, we described five patients with chemorefractory B-cell CNS lymphoma who received CAR19 and CAR22 T-cell “Cocktail” therapy and follow-up for 6–16 months. All patients experienced cytokine release syndrome (CRS). Two patients experienced CAR T-cell-related encephalopathy syndrome (CRES), which was controllable. The best response was observed in two patients, who successfully achieved complete remissions (CR), and the other three patients achieved partial remissions (PR). Four patients had progressive disease (PD) after remission. In addition, one CR patient and one PD patient accepted CAR T-cell infusion following hematopoietic stem cell transplantation therapy in the 3rd month and were in ongoing remission for 14 and 6 months of follow-up, respectively. The targeted antigens in two patients were still positive, and CAR T-cells were reboosted in the cerebrospinal fluid (CSF) after PD, but a small number of CD3-positive T-cells were observed to infiltrate into the tumor. Our study indicates the efficacy of CAR T-cell therapy for CNS lymphoma with an acceptable safety profile; however, the remission did not last long, perhaps due to the tumor immunosuppressive microenvironment (TME) of the CNS. CAR T-cell therapy should be combined with other treatments to help improve the TME of cerebral lymphoma.

Aberrant methylation of tumour suppressor genes is associated with the progression to a blast crisis in chronic myeloid leukaemia (CML). Methyl-CpG-binding domain protein 2 (MBD2) has been studied as a "reader" of DNA methylation in many cancers, but its role in CML is unclear. We constructed cell models of a homozygous deletion mutation of MBD2 using gene-editing technology in K562 cells and BV173 cells. Here, we demonstrated that the deletion of MBD2 inhibited cell proliferation capacity in vitro. MBD2 deletion also significantly inhibited K562 cell proliferation in a xenograft tumour model in vivo. Additionally, the JAK2/STAT3 signalling pathway, which is abnormally active in CML, was inhibited by MBD2 deletion, and MBD2 deletion could up-regulate the expression of SHP1. In conclusion, our findings suggest that MBD2 is a candidate therapeutic strategy for the CML blast phase.