Abstract
Background: Therapeutic resistance remains the primary challenge in adult acute myeloid leukemia (AML). Genetic profiling can refine the prediction of outcome across populations of AML patients and has enabled the development of integrated mutational/cytogenetic risk schemas that can separate patients with cytogenetically defined intermediate-risk AML into three subgroups with markedly different outcomes. Here, we investigated to what degree the prediction of therapeutic resistance and survival can be improved for individual patients by inclusion of data from genetic profiling.
Patients and Methods: We used data on adults aged 17-60 years with newly diagnosed AML who received treatment on a recent phase 3 trial from the Eastern Cooperative Study Group that investigated the value of escalated doses of daunorubicin during induction (E1900; NCT00049517). We used several criteria for the definition of therapeutic resistance: (a) failure to attain complete remission (CR) despite surviving at least 28 days from beginning induction therapy (“primary refractory”); (b) primary refractory or relapse-free survival (RFS) ²3 months; (c) primary refractory or RFS ²6 months; and (d) primary refractory or RFS ²12 months. We used logistic regression analyses to assess the relationship between individual covariates and measures of resistance and overall survival (OS): age, gender, white blood cell (WBC) count, platelet count, bone marrow blast percentage, disease type (primary vs. secondary), cytogenetic risk, and mutational status in the following genes: ASXL1, CEBPA, DNMT3A, FLT3, IDH1, IDH2, KIT, KRAS, MLL, NPM1, NRAS, PHF6, RUNX1, TET2, TP53, and WT1. The integrated mutational/cytogenetic risk schema was used as established by Patel et al. (NEJM 2012;366:1079-1089). We then used the area under the receiver operator characteristic curve (AUC) to quantify a multivariate modelÕs ability to predict therapeutic resistance; in this approach, an AUC of 1 indicates perfect prediction while an AUC of 0.5 indicates no prediction; AUC values of 0.6-0.7, 0.7-0.8, and 0.8-0.9 are commonly considered as poor, fair, and good, respectively.
Results: 298 patients surviving at least 28 days had data on all covariates and were included. 201 (67.4%) of these achieved CR and 97 (32.6%) were primary refractory; 103/297 patients (34.7%) with sufficient follow-up time were either primary refractory or had a RFS of ²3 months, 115/296 patients (38.9%) with sufficient follow-up time were either primary refractory or had a RFS of ²6 months, and 153/295 patients (51.9%) with sufficient follow-up time were primary refractory or had a RFS of ²12 months. The integrated mutational/cytogenetic risk schema was the most important individual predictor of resistance (AUCs ranging between 0.64 and 0.69 across the several definitions of resistance) and survival (AUC of 0.65), followed by cytogenetic risk and FLT3/NPM1 status (AUCs ranging between 0.59 and 0.64). Bootstrap-corrected multivariate models yielded AUCs of 0.76-0.79 for the prediction of primary refractoriness or primary refractoriness/RFS of 3 months or less, 6 months or less, or 12 months or less, respectively, and an AUC of 0.72 for the prediction of OS. Removal of information on FLT3/NPM1 status or mutational data from other profiled genes decreased the AUC to about the same degree each, yielding AUCs of 0.66-0.72 for simpler models including cytogenetic risk and other basic information (age, gender, performance status, white blood cells, platelet counts, marrow blast percentage; see table).
Conclusion: Genetic profiling increases the accuracy of multivariate models predicting therapeutic resistance or survival in adult AML. Nevertheless, even with inclusion of such data, our ability to predict these outcomes based on pre-treatment information remains relatively limited. This finding would argue for the integration of treatment response measures (e.g. minimal residual disease) to optimize prediction of resistance.
Table Parameter No CR No CR or RFS 3 months or less No CR or RFS 6 months or less No CR or RFS 12 months or less OS Basic model 0.60 0.60 0.63 0.63 0.59 Basic model + Cytogenetic risk 0.68 0.68 0.71 0.72 0.66 Basic model + Integrated mutational/ cytogenetic risk schema 0.68 0.68 0.71 0.74 0.68 Basic model + Cytogenetic risk + NPM1, FLT3/ITD 0.72 0.71 0.74 0.75 0.69 Basic model + Cytogenetic risk + NPM1, FLT3/ITD + Other mutations 0.76 0.76 0.78 0.79 0.72
Disclosures
Levine: Novartis: Consultancy, Grant support Other.
A Comparative Study of Digital Image Segmentation Algorithms for Acute Myeloid Leukemia M1 White Blood Cells Images
