CD34+CD38− leukemic stem cell frequency to predict outcome in acute myeloid leukemia

Immunophenotypic Detection of Measurable Residual (Stem Cell) Disease Using LAIP Approach in Acute Myeloid Leukemia

“…Such LSCs may have different immunophenotypes (CD34+CD38+, CD34+CD38−, CD34−), however, CD34+CD38− LSCs have been shown to be most therapy resistant, both in vitro and in vivo, and leukemogenic (Bonnet & Dick, ; Costello et al., ; Ishikawa et al., ; van Rhenen et al., ). Moreover, in an earlier retrospective (Terwijn et al., ) and prospective study (Zeijlemaker et al., ), we showed the clinical relevance of these CD34+CD38− LSCs. Including the CD34+CD38− LSC frequency into the classical MRD frequency assessment strongly improves the prognostic impact of MRD frequency (Zeijlemaker et al., ).…”

“…Moreover, in an earlier retrospective (Terwijn et al., ) and prospective study (Zeijlemaker et al., ), we showed the clinical relevance of these CD34+CD38− LSCs. Including the CD34+CD38− LSC frequency into the classical MRD frequency assessment strongly improves the prognostic impact of MRD frequency (Zeijlemaker et al., ). In future AML studies, therapeutic strategies may thus be based on the combined MRD and LSC results.…”

“…F shows the results for all secondary gating used (FSC/SSC, CD34/SSC, and CD45/SSC), also showing that the CD45/SSC parameter does not contribute in this particular AML case. So, despite the initial poor separation between LSC and HSC (A), LSC and HSC can be fairly well distinguished using secondary gating parameters resulting in similar calculated LSC frequencies: 0.012% in both CD45RA and Combi-6 analyses which is below the previously defined cut-off of 0.03% (% of WBCs) for positivity at diagnosis (Zeijlemaker et al, 2019). Also, quite similar LSC/HSC ratios in the total CD34+CD38− compartment were found: 37:63 for CD45RA and 34:66 for Combi-6.…”

“…Although different targets can be used for molecular MRD detection, the NPM1 mutation offers the most suitable target because it is very specific, sensitive, and stable (Ivey et al., ). However, only part of the AML cases is characterized by the NPM1 mutation and because MFC enables the analysis of MRD in the vast majority (>90%) of AML cases for the best prediction of an AML relapse, at present both NPM1‐MRD and MFC‐MRD are necessary (Zeijlemaker et al., ). Newer technologies, including next generation sequencing (NGS), are promising but still in a preliminary stage, although they have already been suggested to be potentially highly sensitive (Jongen‐Lavrencic et al., ; Thol et al., ).…”

“…In rare cases both with no diagnosis of LAIP available and with shortage of follow-up material, only one, two, or three tubes can be measured at followup. Based on the frequencies of LAIPs, as used in the recent HO102 study (Zeijlemaker et al, 2019), the chances of finding a good LAIP (tube numbers shown in Table 1) are 59% for tube 1, 20% for tube 2, 11% for tube 3, and 10% for tube 4. So, with such a shortage of material and with no diagnosis information, the preferential choice would be tube 1, followed by tube 2, followed by tube 3 and tube 4. c. Importantly, if there is evidence for a CD34-negative AML (in most cases defined already at diagnosis), at least tube 4, with the antibody for the primitive marker CD133 present, should be used.…”