Investigators report that a single leukemic cell unintentionally engineered into the chimeric antigen receptor (CAR) T-cell product can mask it from recognition and confer resistance to CAR T-cell therapy.
They described the case of a 20-year-old male who received the anti-CD19 CAR tisagenlecleucel (Kymriah) and relapsed at day 252 after the infusion.
The transduction of a leukemic cell during manufacture of the CAR T-cell product “is a rare event,” they wrote, and indicated that “this is the only case out of 369 patients reported worldwide at the time of publication.”
Lead author Marco Ruella, MD, of the University of Pennsylvania, and colleagues described the case in a Brief Communication published in Nature Medicine.
"In this case,” Dr. Ruella said, “we found that 100 percent of relapsed leukemic cells carried the CAR that we use to genetically modify T cells."
The patient had B-cell acute lymphoblastic leukemia (B-ALL) and relapsed three times after chemotherapy and a cord blood transplant before enrolling in the phase 1 trial of CTL019 (NCT 01626495).
The investigators reported that the infused CAR cells “displayed the typical pattern of in vivo engraftment and expansion.” At day 28 after the infusion, the patient was in complete remission.
But by day 252, he experienced a second expansion of CAR cells that did not correspond to the re-expansion of CAR+ T cells.
At day 261, the patient relapsed with more than 90% CD10+CD19- leukemic cells in the bone marrow and circulating blasts. The cells were CAR-transduced B-cell leukemia (CARB) cells.
The CARB cells continued to expand, and the patient died of progressive leukemia.
The investigators tracked the origin of the CARB cells by analyzing the relapsed CAR19+ cells using next-generation sequencing.
They hypothesized that the CAR19+ leukemia relapse occurred through lentiviral transduction during the manufacturing process, since they detected no replication-competent lentivirus when testing the patient’s peripheral blood at numerous time points after CTL019 infusion.
Further analysis confirmed the CARB cells were a byproduct made during CTL019 cell manufacturing.
To confirm that the leukemia relapse originated from a single clone, the investigators expanded in mice blast cells detected in the patient at month 9. Nine of 71 cells analyzed were positive for vector-host junctions. This confirmed that the relapsed cells originated from a single blast clone.
The investigators also excluded other possible reasons for the loss of CD19, including mutations, splicing variants, and structural alteration of the B-cell receptor complex.
They found that expression of the CAR in cis on B-ALL blasts masked the CAR target epitope.
The investigators concluded that their results “provide a direct confirmation of the cancer stem cell hypothesis in humans, given that clonal analysis indicated that the relapse and subsequent death of the patient were attributed to the progeny of a single leukemic blast cell with extensive replicative capacity, both in culture and in vivo.”
They called for improved manufacturing technologies that can eliminate contamination by residual tumor cells from engineered T cells.
Interestingly, this case developed not long after a case that showed essentially the opposite situation—a patient with chronic lymphocytic leukemia went into remission because of a single CAR T cell that reproduced and fought off the disease.