CAR-T cells have been regarded as living drugs that undergo redistribution when they enter patients' bodies. Unlike the traditional therapies, the activation, diffusion, and expansion processes of CAR-T cells in patients' bodies influence the clinical efficacy. Therefore, dynamic monitoring of CAR-T cells in vivo is essential to explore their distribution and optimize the clinical responses. However, due to limitations of sampling in clinical studies, to date, no description of spatio-temporal distribution of CAR-T cells within patients' bodies has been reported.

This study included 43 patients with hematologic malignancies and received CAR-T cell therapy. The CAR transgene copies numbers in their peripheral blood (PB) samples and serous cavity effusion (SCE) after receiving CAR-T cell therapy were sequentially monitored. It was found that the CAR copies number in the SCE was consistent with that in PB, which reveals that good expansion of CAR-T cells in PB would be benefit for their distribution in SCE. However, the CAR copies number of PB was significantly higher than that of SCE in the early period (d0-7) after CAR-T infusion. Over time, the CAR copies numbers in SCE gradually equalized and exceeded those in PB. These results demonstrated delayed CAR-T expansion in SCE. We thus speculate that CAR-T cells diffuse from PB into the SCE by circulation and then undergo local expansion.

Next, we found that CAR copies number in cerebrospinal fluid (CSF) samples from patients with meningitis (n=7) was significantly higher than that in patients without meningitis (n=13). Moreover, patients with meningitis were more likely to undergo severe ICANS. However, there was no significant difference of CAR copies number between patients suffered ICANS or not. In addition, we found that the levels of IL-6 secretion in the CSF of patients with ICANS were significantly higher than those in time-paired PB samples, which reflecting local inflammation existed. Overall, meningitis appears to be considered as a stimulus for diffusion and local expansion of CAR-T into CSF.

To investigate the relationship between local tumor invasion and the accumulation as well as expansion of CAR-T cells in pleural effusion or ascites (PE/A), we next divided 13 patients with PE/A into an invasion group (n=6) and a non-invasion group (n=7). Intergroup analysis showed that the CAR transgene copies number in PE/A of patients in the invasion group was significantly higher than that of the non-invasion group. Overall, tumor invasion may promote local accumulation and expansion of CAR-T in PE/A.

According to the imagological examination, 6 of the 13 patients underwent increased PE/A after CAR-T therapy, and they experienced obvious clinical manifestations, such as chest tightness, cough, abdominal distension, nausea, and fatigue. In addition, the level of IL-6 concentration and CAR copies number in the PE/A of these 6 patients were significantly higher than those in time-paired PB samples. Thus, we proposed that these patients may have been experiencing local CRS at that time. Among them, 5 (83.3%) had local tumor invasion, and only 1 (16.7%) did not, which means that patients with tumor invasion were more likely to experience local CRS.

In this study, we demonstrated the spatio-temporal distribution characteristics of CAR-T cells in patients' bodies by examining the CAR copies number of SCE in patients with hematologic malignancies receiving CAR-T cell therapy and illustrated the relationship between CAR copies number and local inflammation, tumor invasion, and adverse events. Thereby, our results will help us accurately understand the diffusion, trafficking, and expansion of CAR-T cells, and provide a new perspective for clinical diagnosis and treatment.

Disclosures

No relevant conflicts of interest to declare.

This content is only available as a PDF.
Sign in via your Institution