Serum Flt3 ligand is a biomarker of progenitor cell mass and prognosis in acute myeloid leukemia

Abstract

Fms-like tyrosine kinase 3 (Flt3) is expressed on progenitor cells and acute myeloid leukemia (AML) blasts. Fms-like tyrosine kinase 3 ligand (Flt3L) is detectable during homeostasis and increases in hypoplasia due to genetic defects or treatment with cytoreductive agents. Conversely, Flt3+ AML is associated with depletion of Flt3L to undetectable levels. After induction chemotherapy, Flt3L is restored in patients entering complete remission (CR) but remains depressed in those with refractory disease. Weekly sampling reveals marked differences in the kinetics of Flt3L response during the first 6 weeks of treatment, proportionate to the clearance of blasts and cellularity of the bone marrow. In the UK NCRI AML17 trial, Flt3L was measured at day 26 in a subgroup of 140 patients with Flt3 mutation randomized to the tyrosine kinase inhibitor lestaurtinib or placebo. In these patients, attainment of CR was associated with higher Flt3L at day 26 (Mann-Whitney UP < .0001). Day 26 Flt3L was also associated with survival; Flt3L ≤291 pg/mL was associated with inferior event-free survival (EFS), and Flt3L >1185 pg/mL was associated with higher overall survival (OS; P = .0119). The separation of EFS and OS curves increased when minimal residual disease (MRD) status was combined with Flt3L measurement, and Flt3L retained a near-significant association with survival after adjusting for MRD in a proportional hazards model. Serial measurement of Flt3L in patients who had received a hematopoietic stem cell transplant for AML illustrates the potential value of monitoring Flt3L to identify relapse. Measurement of Flt3L is a noninvasive test with the potential to inform clinical decisions in patients with AML.

Graphical abstract

Figure 1.

Association of Flt3L with progenitor cell mass and Flt3 expression in AML cohort A. (A) Serum Flt3L measured by ELISA in healthy controls (range, 48.3-173.8 pg/mL; shaded region), individuals with heterozygous germline GATA2 mutation, acquired aplastic anemia (AA), MDS, APL (open symbols), and AML. Asterisks in the MDS cohort indicate patients with excess blasts. Where Flt3L was reported as undetectable, data were set to the limit of detection at 7 pg/mL. Numbers in each cohort are shown in parentheses. (B) Level of Flt3 (CD135) expression indicated by MFI (minus isotype control) in control CD34⁺ BM progenitors, APL blasts, and AML blasts. The numbers in each cohort are indicated in parentheses. Note MFI is a log scale. (C) Inverse relationship between CD135 expression and Flt3L in 14 AML and 4 APL patients in whom both measurements were performed.

Figure 2.

Serial serum Flt3L measurements in AML patients. (A) Patients were sampled at diagnosis and following each of 2 cycles of intensive chemotherapy with daunorubicin and cytarabine (3+10) or fludarabine, cytarabine, and idarubicin (cohort A, Newcastle; shaded regions indicate normal range). Patients who entered continuing morphological CR after 2 courses maintained increasing Flt3L at remission assessments performed 28 to 42 days after the start of chemotherapy (range of detection, 7-1000 pg/mL; n = 15; broken lines). (B) Patients who did not achieve morphological CR after the first course of chemotherapy but subsequently entered remission after the second also demonstrated increasing Flt3L over both cycles (n = 5; filled circles and solid lines). (C) Patients who remained refractory after 2 courses of chemotherapy had undetectable Flt3L (n = 5; open circles and solid lines). The graph also includes a further 3 patients thought to be in morphological CR after the first course of chemotherapy who experienced relapse within 6 months.

Figure 3.

Kinetics of Flt3L response during induction. Sequential patients with weekly measurement of Flt3L during induction chemotherapy (n = 12; 4 in cohort A and 8 in cohort B). Dotted line indicates the timing of a single Flt3L measurement in patients treated on the AML17 trial. (A) Weekly Flt3L in 6 patients entering CR following induction chemotherapy (B) Weekly Flt3L in 4 patients with refractory disease or partial response following induction chemotherapy. (C) Weekly Flt3L in 2 patients rendered aplastic by induction chemotherapy

Figure 4.

Association between Flt3L and post–course 1 remission and survival. Serum Flt3L was measured on day 26 of induction in non-APL AML patients with Flt3 mutation randomized to lestaurtinib (days 11-24) on the UK NCRI AML17 trial. (A) Serum Flt3L at day 26 in patients who subsequently achieved or did not achieve CR. Dotted line indicates optimal cut point identified by the ROC curve: 116 pg/mL (Mann-Whitney U P value). (B) EFS of subgroups with serum Flt3L level ≥291 pg/mL (blue line) or <291 pg/mL (red line). Events include failure to achieve CR and relapse or death from any cause. Threshold of Flt3L for maximum discrimination identified by log-rank testing (log-rank P value). (C) OS of subgroups with a serum Flt3L level ≥1185 pg/mL (blue line) or <1185 pg/mL (red line). Threshold of Flt3L for maximum discrimination identified by log-rank testing (log-rank P value). (D) EFS of subgroups divided according to Flt3L level ≥291 pg/mL and MRD negative (green line, n = 17), ≥291 pg/mL and MRD positive (blue line, n = 15), Flt3L <291 pg/mL and MRD negative (black line, n = 23), or <291 pg/mL and MRD positive (red line, n = 46) (log-rank P value). (E) OS of subgroups divided according to Flt3L level ≥1185 pg/mL and MRD negative (green line, n = 10), ≥1185 pg/mL and MRD positive (blue line, n = 5), Flt3L <1185 pg/mL and MRD negative (black line, n = 30), or <1185 pg/mL and MRD positive (red line, n = 56) (log-rank P value).

Figure 5.

Serial measurement of Flt3L after hematopoietic stem cell transplantation. (A) Serial serum Flt3L measurements in patients with non-APL AML undergoing hematopoietic stem cell transplantation remaining in continuous remission (n = 8; open circles and broken lines; shaded region indicates normal range). Patients were sampled prior to conditioning (day −7), on the day of transplant (day 0), and on days and months posttransplant, as indicated. Patients were selected over the same period as panel B. (B) Serial serum Flt3L measurements in 7 patients with non-APL AML undergoing hematopoietic stem cell transplantation who relapsed (n = 7; filled circles and broken lines). Consecutive relapsed patients ≤2 years posttransplant are reported. (C) Interval between drop in Flt3L below the normal range and diagnosis of relapse. Summary of calendar-driven Flt3L measurement in 17 patients indicating timing of the point at which Flt3L first went below normal relative to the hematological diagnosis of relapse. Includes all patients in B with 10 additional patients in whom ≥1 Flt3L measurement was recording posttransplantation.