Clinicians have struggled for years to bring a successful immunotherapy to the clinic that has a significant and lasting impact on patients with cancer. Genetically engineered autologous T cell based therapeutics have cemented immunotherapy as a cornerstone modality in the fight against tumors. Although many successes are being realized with T cell receptor (TCR) engineering such as with NY-ESO specific T cells in HLA-A2 recipients with advanced synovial cell sarcoma, chimeric antigen receptors (CAR) will revolutionize how we treat leukemia and lymphoma.
The power of CAR T cells is multifaceted: 1) CARs free T cells from MHC restriction thereby allowing universal application, 2) CARs maintain antibody specificity, and 3) CARs provide both signals required for full T cell activation simultaneously (Figure 1). So, through a single interaction with its cognate extracellular target, CAR T cells rapidly kill the target cell, secrete inflammatory cytokines which recruit other immune components, and replicate many fold with each of the progeny CAR T cells able to carry out all of these functions. Indeed, this replicative capacity is the principal reason CAR therapy has been so successful against leukemia and lymphoma.
Several groups have demonstrated >70% complete response rates in early clinical trials of autologous CD19 CAR T cells in children and young adults with highly refractory, multiply relapsed pre-B acute lymphoblastic leukemia (ALL), and responses for adults with B-lineage lymphomas are not far behind. Clinical responses are always accompanied by CAR T cell expansion (Figure 2) the degree of which appears to be dependent on tumor burden, at least in leukemia. This is currently the only known surrogate for response, so our ability to detect and track their expansion in a peripheral blood sample is of high importance. Although PCR tests are available for all CARs produced, such testing is typically done retrospectively and does not indicate whether the CAR is expressed on the cell surface, an important piece of data as several CARs have been shown to be downregulated upon T cell activation. Flow cytometry, therefore, represents the best modality for real-time monitoring of expansion and response.
Since CARs are generally composed of single chain variable fragments (scFv) to endow antigenic specificity with little else on the extracellular surface (Figure 1), detecting the CAR is not a straightforward task. Three modalities exist: 1) develop an anti- idiotype antibody to the scFv 2) use Fc-conjugated soluble antigen such as Fc-CD22 that interacts with CD22 CAR T cells followed by a secondary detection antibody (anti-Fc) and 3) use biotinylated Protein L followed by a strepavidin-conjugated florophore.
Anti-idiotype antibodies generate the best sensitivities and are simple to use in multi-parameter flow assays, but making them is expensive and takes time. Fc-conjugated antigens and biotinylated Protein L are cheaper and quicker to make but are less sensitive and more time consuming to use. Finally, Protein L will not bind all scFv's but importantly will bind to most flow antibodies. Therefore, Protein L must be incubated with cells in a separate step before the cells are labeled with other reagents.
These challenges are not insurmountable. Labs at several institutions are routinely assaying blood, bone marrow, and CSF for CAR T cells and the articles accompanying this issue demonstrate several of these approaches. Establishing assays and validating them separately for each CAR (or each scFv) is required and time consuming, but once done a wealth of information can be provided to the clinicians.
Daniel W. "Trey" Lee, MD
National Cancer Institute
Bethesda, MD, USA