Poster Session II: Acute myeloid leukemia - Biology - translational research
The screening of isocitrate dehydrogenase 1 and 2 (IDH1 and IDH2) mutations plays an important role in categorization of acute myeloid leukemia (AML) and can drive ab initio the treatment strategy due to the introduction of IDH inhibitors, ivodesinib and enadisenib, in the clinical practice. IDH2 mutations are found in about 20% of AML patients; they are single-nucleotide variants involving the arginine hotspots R140 or R172. Prognostic assessment of IDH2 mutations is still controversial, but the introduction of enadisenib leads the possibility to use IDH2 as a marker of minimal residual disease (MRD), nevertheless the stability and suitability of this marker is not still confirmed (Petrova et al., 2018). Sanger sequencing is the gold-standard method for detection, but it is not quantitative and it has limited sensitivity (15%). Various allele-specific qPCR methods have been developed, based on the amplification refractory mutation system PCR techniques.
We set a new multiplex ddPCR method to detect IDH2 mutations, then used at diagnosis and as tool for MRD detection.
A “drop-off” ddPCR FAM/HEX Assay (Biorad®) method was used as basis for setting the new technique for IDH2 mutations detection (ddIDH2 G&G). “Drop-off” requires a single pair of probes to detect and quantify different mutations in a single reaction: the FAM-labeled probe binds a reference sequence distant from the target but still within the same amplicon, while the HEX probe binds the wild-type sequence in the target site. In a 2D-plot wild-type samples present signals from both FAM and HEX probes, while the mutated ones display only the FAM signal. This method reached a sensitivity of 2x10-3. As qPCR assay, the qBiomarker Somatic Mutation PCR Assay kit was employed (Qiagen®).
At diagnosis, we identified 11/60 (18.3%) IDH2-mutated patients, 9 carrying the R140Q and 2 had R172K mutation. The median mutational burden was 13.7% (range 0.4%>44%); it did not correlate with blasts percentage or histotype. Interestingly, Sanger sequencinq was not able to identify mutations in 5 out of the 11 cases already positive by qPCR, probably because of their low mutational burden (0.4-12%). In our series of patients, the IDH2 mutational status did not significantly impact on survival, neither on the achievement of the clinical response after treatment. We also performed ddPCR in follow-up samples after induction and consolidation therapy in 4 IDH2 mutated patients (2 CRMRD-, 1 CRMRD+ and 1 resistant). In the CRMRD- patients the IDH2 allele burden was reduced and became negative, such as NPM1 and WT1 markers and they remained in remission; in the resistant one the IDH2 status correlated with WT1 and FLT3 persistence; in the CRMRD+ patient the IDH2 mutation was 0% while the MRD positivity was still detected by NPM1 and WT1 test (30% and 100 copies respectively). This last patient relapsed, confirming NPM1 as a good predicting marker.
We demonstrated that the ddIDH2 G&G method could represent a valid tool for detecting and quantifying IDH2 mutations, even in the MRD setting. Our results about MRD are very preliminary, but suggest that the application of ddPCR to IDH2 mutations is today a hot matter of debate. Finally, it is relevant that the sensitivity of our method is really promising, being higher than that of Sanger and comparable to that of qPCR.