DNA Sequencing of Bone Marrow Is Predictive of ALL Relapse Following Treatment With Tisagenlecleucel

DNA sequencing to detect minimal residual disease and eventual relapse was more accurate than flow cytometry for adult and pediatric patients with acute lymphoblastic leukemia who have been treated with tisagenlecleucel.

DNA sequencing-based detection of residual disease was found to be more predictive than flow cytometry and could accurately identify pediatric and young adults who would relapse with acute lymphoblastic leukemia following treatment with tisagenlecleucel (Kymriah), according to a study published in Blood Cancer Discovery.1 

Next-generation sequencing (NGS) with minimal residual disease (MRD) detection of more than 0 in bone marrow was reported to be highly associated with relapse. Patients who had B-cell recovery within the first year were associated with a hazard ratio (HR) for relapse of 4.5 (95% CI, 2.03-9.97; P <.001). At day 28 of the multivariate analysis, investigators reported that having bone marrow NGS-MRD of more than 0 (HR, 4.87; 95% CI, 2.18-10.8; P <.001) and B-cell recovery (HR, 3.33; 95% CI, 1.44-7.69; P = .005) were independently associated with relapse. At 3 months, the bone marrow NGS-MRD HR increased to 12 (95% CI, 2.87-50.0; P <.001) and the B-cell recovery was not found to be independently predictive (HR, 1.27; 95% CI, 0.33-4.79; P = .7).

“This is the first paper to show an approach that identifies markers of relapse that are very specific, allowing clinicians to add additional therapy prior to relapse that will prevent it,” lead author Michael Pulsipher, MD, a professor of pediatrics and the Division Chief of Pediatric Hematology and Oncology at Intermountain Primary Children’s Hospital and the Huntsman Cancer Institute, University of Utah, said in the press release.2

Investigators collected 1771 multiparamater flow cytometry (MFC) samples from 143 patients who were enrolled on the phase 2 ENSIGN (NCT02228096) and phase 2 ELIANA (NCT02435849) studies, both of which examined the safety and efficacy of tisagenlecleucel in patients with ALL. The median follow-up was 38.4 months. To analyze NGS-MRD, 474 samples were taken from 109 patients. The timepoints ranged from time of screening to 24 months and 79% of post infusion samples were drawn at 1, 3, and 6 months.

Investigators did not identify major differences between subgroups of patients with available NGS-MRD samples and those who did not. They did, however, observe a trend towards fewer patients in the non-NGS-MRD group receiving prior hemopoietic stem cell transplant (HSCT). Investigators noted that the patient population was high-risk, with 58% of patients relapsing after previous allogeneic HSCT and 69% having high bone marrow tumor burden.

Investigators observed that peripheral blood NGS-MRD was more sensitive in terms of detecting disease, with 9% more samples being detected through the method.Additionally, NGS-MRD yielded a sensitivity level of 10-6 than those detected via bone marrow MFC-MRD. Thirteen percent more samples were detected if any level of detection was allowed vs bone marrow MFC-MRD. Investigators reported that 10 patients had detectable NGS-MRD but negative bone marrow MFC-MRD, and 5 patients developed bone marrow or extramedullary relapse, 4 received HSCT or additional therapy, and 1 was lost to follow-up.

In the 287 post-infusion samples from 95 patients, MFC detected 18% of samples as being MRD-positive. At the sensitivity cutoff levels of 10-4, 10-5, 10-6, and 10-7, respectively, NGS detected 22%, 29%, 33%, and 41% of samples as MRD-positive. NGS detected 96 (88%) more samples than MFC (51) at sensitivity of 10-6.

Those with complete remission (CR) or CR incomplete hematologic recovery who did not achieve bone marrow MFC-negative remission at day 28 and/or month 3 following infusion did not do well, investigators stated. Of the 109 patients, on day 28 with a CR, 2 were bone marrow MRD–positive by MFC, with relapse occurring on day 80 and 85; the patients died without additional therapy. Five of 82 patients who achieved CR/CRi by 3 months were bone marrow MFC-MRD–positive, and 2 relapsed by day 92 and day 106. On day 93, 1 patient underwent HSCT, another received a different therapy on day 151, and 1 was lost to follow-up on day 86.


1. Pulsipher M, Han X, Maude S, et al. Next-generation sequencing of minimal residual disease for predicting relapse after tisagenlecleucel in children and young adults with acute lymphoblastic leukemia. Blood. Published online December 1, 2021. doi:10.1158/2643-3230.BCD-21-0095

2. Sequencing of bone marrow DNA may predict leukemia relapse after CAR-T therapy with sufficient time to intervene. News release. American Association for Cancer Research. December 1, 2021. Accessed December 1, 2021. https://bit.ly/3DbBMc