The (Un)Necessary Use of Short-Tandem-Repeat (STR) Testing on Bone Marrow Samples in Patients After 1 Year Postallogeneic Hematopoietic Stem Cell Transplant

Abstract

Abstract Allogeneic hematopoietic stem cell transplant (HSCT) patients at our institution are monitored at multiple time points posttransplant for neoplastic disorders by short-tandem-repeat (STR) testing of unfractionated bone marrow (BM) and BM biopsy evaluation (morphology, flow cytometry, genetics) to assess for engraftment, rejection, relapse risk, graft-versus-host disease, and disease relapse. Since there is some overlap in the information provided by STR and BM biopsies, we asked if and when this combined testing may be redundant. To answer this question, we compared results of STR and flow cytometry (a surrogate for BM biopsy evaluation) at different time points posttransplant. We identified 1,199 cases, between June 2011 and November 2018, from 410 patients in which STR testing on unfractionated BM samples had concurrent flow cytometry analysis (within 7 days of each other). STR testing was considered positive if ≥1% of recipient DNA was identified in the sample. Flow cytometry was considered positive if an aberrant population was identified that correlated with the patient’s prior diagnosis. Analysis included concordance/discordance rates as a function of time from transplant and relapse (if relapse had occurred) and potential reasons for discordance. The overall concordance rate between STR and flow cytometry was 93%, with a majority of cases (1,063) being STR(–)/Flow(–). Of all discordant cases, 75 of those were STR(+)/Flow(–); however, the majority of these cases (70%) demonstrated <5% of recipient DNA. Five cases of graft failure/rejection were identified with an average recipient DNA of 93% at <1 year from transplant. The eight cases identified as STR(–)/Flow(+) represented relapsed/residual disease diagnosed on average 565 days after transplant (aberrant population ranged from 1% to 6% of total events by flow cytometry). Focusing on cases ≥1 year from transplant (54% of all cases), the concordance rate between STR and flow cytometry increased to 98%, with only nine cases identified as STR(+)/Flow(–). Five of the nine cases had an average recipient DNA of 2% and no clinical or pathologic evidence of relapse with at least 490 days posttransplant follow-up. The other four cases were at time points either after pathologic relapse had already been confirmed, at 3 months prior to cytogenetic relapse, or at least 2 years prior to STR(+)/Flow(+) relapse. Our main findings include the following: (1) STR testing is critical for identifying graft failure/rejection within 1-year posttransplant, (2) flow cytometry is superior to STR at detecting late relapses with low-level disease, and (3) STR testing on BM samples from patients ≥1 year post-HSCT may be unnecessary, as in our experience the BM biopsy evaluation is sufficient to determine disease relapse. Lastly, financial analysis suggests that elimination of STR testing 1-year post-HSCT would lead to a significant cost-avoidance to our health care system.