PS935 APPLICABILITY OF DROPLET DIGITAL PCR (DDPCR) FOR MINIMAL RESIDUAL DISEASE MONITORING (MRD) IN PHILADELPHIA‐POSITIVE ACUTE LYMPHOBLASTIC LEUKEMIA

Background: In Philadelphia‐positive acute lymphoblastic leukemia (Ph+ ALL), minimal residual disease (MRD) is the strongest independent prognostic factor. Q‐RT‐PCR represents the gold‐standard approach, but it is limited by the lack of standardization across laboratories. Moreover, a fraction of patients with very low MRD levels are classified as positive not‐quantifiable (PNQ), which represent a clinical gray zone. Droplet digital PCR (ddPCR) is an alternative approach which is based on an absolute quantification without the need of a standard curve, has the potential to overcome some limitations of RQ‐PCR and has shown to be advantageous in other lymphoproliferative disorders. In Ph+ ALL, few data have been reported and there are no guidelines for ddPCR MRD analysis and interpretation. Aims: 1) To develop a ddPCR assay, by establishing the reaction parameters (i.e. cDNA input and PCR conditions) for the quantification of BCR / ABL1 p190 and p210 transcripts; 2) to assess the limit of detection (LOD), specificity and reproducibility of the assay; 3) to compare ddPCR and Q‐RT‐PCR in samples from patients enrolled in the GIMEMA LAL2116 clinical trial. Methods: To perform ddPCR experiments we followed Biorad recommendations and used the same primers and probes of Q‐RT‐PCR (BIOMED1). After PCR, the plate was loaded into the QX200 Droplets Reader and data analyzed by the QuantaSoft analysis Software v1.7.4. Only replicates with a number of droplets ≥9000 were deemed acceptable. For results interpretation, we followed the FIL MRD NETWORK recommendations. Three volumes of cDNA (i.e. 1 μL, 2.5 μL and 5 μL) were tested to define the optimal input. To evaluate the specificity of the assay, we tested 4 pools of mononuclear cells from 4 healthy donors and not template controls. To assess the reproducibility, we tested 3 diagnostic Ph+ ALL samples and their 10 –1 , 10 –2 , 10 –3 dilutions in 2 independent experiments. To define the LOD, we tested serial dilutions of the lowest point (10 copies) of a plasmid standard curve corresponding to 1x10 –4 ; 5x10 –5 ; 1x10 –5 ; 5x10 –6 ; 1x10 –6 , in a logarithmic scale. Finally, we compared Q‐RT‐PCR and ddPCR in 10 samples at diagnosis and 35 collected during MRD monitoring. Results: The set‐up experiments revealed that: 1) 1 μL of undiluted cDNA for diagnostic samples and 5 μL for follow‐up samples are required; 2) the maximum sensitivity achieved by the method is 1x10 –5 and reproducible sensitivity is 5x10 –5 . We documented a high intra‐ and inter‐run reproducibility. The comparison of Q‐RT‐PCR and ddPCR on diagnostic samples highlighted a high concordance between the two methods (Pearson correlation coefficient = 0.72). The analysis of PNQ cases by Q‐RT‐PCR revealed that 55.5% (15/27) proved quantifiable by ddPCR, 25.9% (7/27) were confirmed as PNQ and 18.5% (5/27) were negative by ddPCR. The analysis of negative cases by Q‐RT‐PCR confirmed their negativity in 75% (6/8) also by ddPCR, while 25% (2/8) were classified as PNQ. Summary/Conclusion: ddPCR is as sensitive as Q‐RT‐PCR and can provide a more accurate quantification in PNQ cases by Q‐RT‐PCR, allowing to recover their quantification in 55% of cases. At present, clinical correlates analyses cannot be performed given the short follow‐up period of patients. Finally, the application of ddPCR in clinical protocols with well‐defined international guidelines is needed to define whether it can result in a refined management of Ph+ ALL outcome.