ASH 2022 | Using serum proteomics & the HT10 score to distinguish between infections & CRS after CAR-T therapy

One of the clinical problems that we are facing when it comes to CAR-T cell therapy is fever in these patients. A significant proportion, almost all patients develop fever early on after their CAR-T cell transfusion. It’s always the question, is this fever that we are seeing associated to the cytokine release syndrome that is CAR-T cell driven or is this associated with some kind of infection? Because CAR-T cell patients are usually heavily pretreated, particularly in the relapsed/refractory setting, they also do have a significant risk of infection. Treatment for infection apparently is completely different to the treatment of cytokine release syndrome. For the one treatment you use antibiotics, and for the other you use things like tocilizumab. Looking at a cohort of a total of 60 patients, we tried to understand if we can do something like a prediction for a patient if a fever event in these patients is associated with CRS or if it’s associated with infection.

We checked for different routine serum markers, inflammatory markers, namely CRP, interleukin-6 and procalcitonin, PCT. Looked if one of these factors might allow a prediction if this is an infectious event that we’re looking at or if this is a CRS related event. We found that PCT, so procalcitonin, is the marker that is best suited to differentiate between CRS and infection. When looking at the total cohort of patients, we additionally looked on ways to improve the accuracy of that prediction. What we did there was that we used the baseline CAR-HEMATOTOX score. This is a score that has been developed by our group, by Kai Rejeski, and has been published.

This is a score that incorporates information about the baseline hematopoietic reserve of the patient, so ANC levels and thrombocytes, and also incorporates information about the baseline inflammatory state of the patient, namely ferratin, and calculates from these different information, these values that we get for the patient prior to lymphodepletion, a score. The score ranges between zero and seven and then gives us a risk stratifier for hematologic toxicities after CAR-T cell therapy. When looking at the question about infection versus CRS, we combined this baseline HEMATOTOX score together with elevated levels of the PCT, so above a threshold of 1.5. When we combine this information, we can improve the prognostication for CRS versus infection and we think that this might be a suitable marker to be used in clinics to evaluate that risk and make decisions on treatment of fever events accordingly.