BPX-501 T-cells interfere with minimal residual disease evaluation of B-cell acute lymphoblastic leukemia

Cellular therapies are increasingly used to treat hematologic malignancies requiring providers to recognize and learn to treat newly prominent complications of therapy such as cytokine release syndrome as well as adapting their practice to incorporate the special requirements for working with cellular therapies. Here we present an unusual complication of BPX-501 cells affecting the clinical workflow and emphasizing the need for good communication between providers and the laboratory.

The patient is a 24-year-old female with Philadelphia chromosome positive precursor B-cell acute lymphoblastic leukemia (pre B-ALL) in first complete remission who received an alpha beta T-cell depleted, CD34+ selected haploidentical peripheral blood hematopoietic stem cell transplant (HSCT) after myeloablative conditioning with fludarabine, cyclophosphamide, anti-thymocyte globulin, rituximab, and total body irradiation. The patient received a planned infusion of BPX-501 T-cells on day 13 as a part of a clinical trial (ClinicalTrials.gov identifier: NCT0174423). BPX-501 T-cells are allogeneic donor T-cells which have been modified with an inducible apoptosis safety switch based upon the fusion of human caspase 9 to human FK506-binding protein (FKBP12). Upon binding of remiducid (AP1903), a bioinert agent with high affinity to FKBP12, caspase 9 dimerizes and is activated, inducing apoptosis. The development of acute graft versus host disease can be abrogated in recipients of haploidentical transplants who have received infusions of BPX-501 T cells upon exposure to remiducid.¹

Neutrophil and platelet engraftment occurred 11 days after transplant. The patient was discharged to her home on day 22, and received follow-up in clinic twice weekly. Her post-transplant course was complicated by relative eosinophilia and transplant related microangiopathy which was resolved by discontinuing tacrolimus. The patient developed mild grade 1 acute graft versus host disease of the skin (stage 1) which was treated with topical corticosteroids. Bone marrow biopsies at 30, 60, and 180 days following HSCT did not reveal any evidence of residual leukemia by morphology, cytogenetics, or standard flow cytometry.

As part of a validation study for the detection of minimal residual disease (MRD) in B ALL, a sample was sent to an outside institution for multiparameter flow cytometry at 6-months post-HSCT. ² This test demonstrated a small population of abnormal cells positive for CD45 (bright) and CD19 but lacking CD10, CD20, CD38, CD34, and the myeloid antigens CD13 and CD33 (Figure 1). The pretreatment phenotype and details of therapy received were not available to the laboratory at the time of analysis. The population identified occupied a region of multiparameter space in which no normal B cell would fall; because of this, the test was interpreted as positive for MRD, although this would be an unusual phenotype for B ALL.

Figure 1. Multiparameter flow cytometric analysis showing atypical CD19 positive cells.

Six color flow cytometry analysis was performed using the two-tube combination CD20-FITC / CD10-PE / CD38-PerCPCy5.5 / CD19-PECy7 / CD58-APC / CD45-APCH7 and CD9-FITC / CD13+CD33-PE / CD34-PerCPCy5.5 / CD19-PECy7 / CD10-APC / CD45-APCH7. Only selected plots from the two tubes are shown in the figure. CD19+ cells were gated (A) and the gate refined on a display of CD45 vs CD19 (not shown). Normal B cell precursors (pink) were identified on a CD45 vs CD10 plot (B). An unusual CD19+CD10-negative B cell population (blue) was identified (C) and also found to lack CD20 (D) and CD34 (F and G) while expressing CD58 (E) and lacking CD38 (E and H). This CD19+CD45+ CD58+CD10-CD20-CD34-CD38− population did not correspond to any known normal B cell immunophenotypes and was interpreted as B cell acute lymphoblastic leukemia minimal residual disease.

These findings were a cause of consternation given the clinical implications of MRD pre-B ALL to the patient. Clinically this was a highly unanticipated result: diagnostic testing performed at the treating institution including bone marrow morphology, cytogenetic analysis, donor chimerism studies, and flow cytometry based MRD testing for pre B-ALL did not indicate any residual pre B-ALL. Upon reconsidering the case, we recalled that the manufacturing process for BPX-501 T-cells uses a CD19 marker to identify the donor T cells that have been successfully transduced with the transgene encoding the caspase9 / FK506 binding protein fusion suicide gene switch. We thus hypothesized that the abnormal CD19+ cells represented the BPX-501 T-cells which had been infused on day 13 and had progressively expanded in the intervening 167 days. Review of the patient’s blood immunophenotyping studies demonstrated a small unusual population of CD19+CD3+ cells (Figure 2a), which confirmed our hypothesis that the abnormal population detected earlier was consistent with the BPX-501 T-cells. Further review of the sample received by the reference laboratory also demonstrated a small CD19+CD3+ population (Figure 2b). Notably, communication of the patient’s therapeutic history, specifically having received BPX-501 T-cells, with the reference laboratory would have affected the interpretation of the unusual CD19+CD3+ population that was observed.

Figure 2. BPX-501 T-cell population co-expressing CD19 and CD3.

A. Immunophenotyping from the treating center. There is a dual CD19+CD3+ population (red dots). B. Immunophenotyping from the reference laboratory. The same sample illustrated in Figure 1 stained with CD19 and CD3. There is a small population of BXP-501 T cells (arrow) similar to those shown in Figure 2a.

Achieving MRD negativity predicts for improved clinical outcomes in many hematologic malignancies such as ALL and multiple myeloma.^2–5 Thus, MRD monitoring has become a standard clinical practice over the last decade. The use of cellular therapies, however, can affect MRD assessment by multiparameter flow cytometry. CD19 chimeric antigen receptor T cell therapies are highly effective against B-cell acute lymphoblastic leukemia. Of the few patients who relapse, as many as 65% will have CD19 negative disease which will not be detected by CD19 based MRD gating strategies.^{6, 7} In response, a new ALL MRD assessment panel has been developed to overcome this diagnostic challenge which gates on CD66b negative and CD24 or CD22 positive cells instead of CD19 positive cells.⁸

Less frequently, false positive MRD results can occur, as in this case, when a cellular therapy expresses an antigen in common with a hematologic malignancy. In this patient, the clonal population of BPX-501 cells expressed CD19 which was misinterpreted as a relapse of pre-B ALL. Theoretically, the presence of BPX-501 cells in the context of any other CD19 expressing malignancy could also be misinterpreted as relapse. Other potential false positives could occur in cellular therapies which utilize the newly described RQR8 antigen which combines epitopes from human CD34 and CD20.⁹ Falsely positive MRD assays could occur in the context of any CD20 expressing lymphoma or CD34 expressing acute myeloid or lymphoid leukemia.

This is the first case to report the misinterpretation of BPX-501 T-cells as MRD which could have resulted in the incorrect conclusion that a current therapy for acute lymphoblastic leukemia had failed to maintain remission. Subsequent changes to the patient’s therapeutic plan based upon this result, such as the addition of chemo- targeted, or immuno- therapies, would have greatly impacted the patient’s life. This case is informative for two reasons: 1) knowledge of the nature of the cellular therapy being used is key, especially any unusual features used in the manufacturing process such as anomalous immunophenotypes used to identify the cells, and 2) good communication with the flow cytometry lab is essential so that it can be aware of and appropriately evaluate any anomalous findings. As technological advancement continues to grow rapidly, it is important to recognize that good communication and shared knowledge are still essential to good medicine.