There are several biomarkers that can predict less durable responses and lower rates of complete remissions with CAR T-cell therapy. However, importantly, we do not have any biomarkers that can identify patients who would not stand to benefit from this treatment, assuming that they are eligible to receive it. So, biomarkers do not generally dissuade me from considering CAR T-cell therapy, and this treatment should be considered for any individual who is eligible, as it is the only option that can induce a cure in a substantial proportion of patients with DLBCL who are relapsing after second-line therapy.

High tumor burden, which is defined in many ways, can indicate a lower likelihood of achieving a durable or complete remission with CAR T-cell therapy. Those with bulkier disease, high lactate dehydrogenase levels, and the highest tumor burdens appear to have less durable remission compared with those with less bulky disease. These findings may reflect exhaustion of the CAR T cells after they encounter abundant tumor antigen, so we will likely need to find ways to offset T-cell exhaustion and/or to induce cytoreduction.

The prior use of bridging therapy has also been associated with lower rates of complete and durable remission in patients with DLBCL. Now, there is certainly overlap between high tumor burden and the use of bridging therapy (ie, patients who require bridging therapy are more likely to have a high tumor burden and/or more rapidly progressive symptomatic disease). Patients also often have disease that is difficult to cytoreduce with conventional therapies. All CAR T-cell products were US Food and Drug Administration approved based on the receipt of a median of 3 prior lines of therapy, and we have been using them in patients with chemotherapy-resistant disease; it is very difficult to reduce the tumor burden with chemotherapy in such individuals. It is also possible that the bridging therapy itself might contribute to inferior outcomes, potentially by inducing toxicity that makes patients less able to tolerate the subsequent lymphodepleting chemotherapy and CAR T cells.

Other biomarkers that are associated with lower rates of durable remission in patients with DLBCL include high baseline levels of C-reactive protein, ferritin, and interleukin 6. These markers might correlate with high tumor burden or with a more inflamed microenvironment that may be more prone to inducing CAR T-cell exhaustion. More recently, high levels of monocytes and myeloid-derived suppressor cells and deletion or mutations of CD58 have been linked to poor outcomes. Monocytes can have an immunosuppressive effect and thus may contribute to lower CAR T-cell activity. CD58 is on the surface of the lymphoma, and it is the receptor for CD2, one of the primary binding domains of the T cell. So, if you delete or mutate the binding domain of T cell, which contributes to one of the major activation signals, you can decrease activity. While these biomarkers are still exploratory, they are potential avenues that could be investigated as therapeutic targets to optimize the depth and durability of CAR T-cell responses.