6606 Background: Mutations in isocitrate dehydrogenase (IDH1 and IDH2) occur in 10-20% of AML patients and result in production of 2-hydroxyglutarate (2-HG). We hypothesize that serial 2-HG quantification may be used to monitor response and predict relapse in patients with AML. Methods: We are conducting a prospective study at MGH and DFCI to (1) assess changes in serum and urine 2-HG levels during treatment in patients with newly diagnosed AML, (2) compare 2-HG levels in serum, urine, and bone marrow and, (3) serially compare 2-HG levels with IDH1/2mutant allele burden. In those with elevated serum 2-HG (≥1000ng/ml), 2-HG is monitored serially or at relapse. Results: To date, 20 patients have been enrolled, 5 (25%) of whom had elevated baseline serum 2-HG. All 5 were found to have IDH1/2 mutations. The serum 2-HG for these patients was significantly higher than for those who were IDH-wt (median 1933 vs 87ng/mL; p<0.001). Urine 2-HG was also higher in IDH-mutant patients (median 30500 vs 4230ng/mL; p<0.021), as was bone marrow 2-HG (median 9870 vs 309ng/mL; p<0.005). Serum 2-HG levels strongly correlated with that in urine (R2 0.987). Serum 2-HG decreased in all IDH mutant patients on therapy, but more rapidly in those receiving induction chemotherapy (Table), all of whom achieved remission, than those receiving hypomethylatingagents. Conclusions: Patients with IDH-mutant AML have markedly elevated serum and urine 2-HG. 2-HG levels decreased with therapy and there is a strong correlation between serum and urine 2-HG. Data on the sensitivity of serial 2-HG monitoring as well as comparison with mutant IDH1/2allele burden will be presented. This data suggests that serial 2-HG quantification may be a valuable non-invasive biomarker in this genetically defined subset of AML. [Table: see text]