Droplet digital PCR represents a new sensitive method for detecting B-RAFV600E mutation at diagnosis and during the follow-up of patients affected by hairy cell leukemia

The BRAFV600E mutation has been found in up to 100% of patients affected by hairy cell leukemia (HCL). The aim of this study was the comparison of the real-time PCR (RQ-PCR) and the droplet digital PCR (DD-PCR) in the detecting the BRAFV600E mutation in patients affected by HCL. We assessed 20 cases with splenic marginal lymphoma and 28 HCL patients by a specific real-time PCR for the BRAFV600E mutation and a new PCR technique: the digital droplet PCR (PrimePCR™ DD-PCR™ Mutation Assay: JAK2 WT for p.V617F, Human- Biorad). At diagnosis, 44 cases were assessed: all patients showed the IgH clonality; when they were tested for the BRAFV600E mutation by QT-PCR, all marginal cases resulted wild-type, whereas the mutation was detected in 23 out 24 of the HCL patients; the wild-type patient was affected by a “variant” HCL. Same results were obtained by the DD-PCR. Other 4 cases were assessed as the first determination when in complete remission; all these patients resulted BRAF wild-type. Twenty-eight HCL cases have been monitored during the follow-up by QT-PCR (62 samples) and DD-PCR (118 samples): a significant reduction of “mutational burden” was observed after 2CdA, but even more significant it appeared after rituximab. Interestingly, patients concomitantly receiving rituximab and 2CdA showed a more rapid clearance of disease. At the last follow-up, 19/28 (68%) did not show anymore BRAF mutation; other 2 cases presented the mutation, still detectable but not quantifiable (<10 copies), and 3 patients showed a very low “mutational burden” (0.12). All these patients were in CR. The 4 remaining cases, positive after molecular assessment, were in partial response. Three patients relapsed: in all the BRAF mutation was still detectable at significant levels (more than 2) by real-time and DD-PCR. Sensitivity tests have been performed for both techniques, diluting a mutated DNA with a pool of wild-type DNAs, from 1x10 to 5x10-4. The sensitivity of the real-time PCR was 1x10-4, whereas that of the digital PCR was 5x10-5. When real-time and DD-PCR were applied to the same samples, only one case resulted wild-type by real-time PCR but mutated by DD-PCR, even if at low levels. In one third of cases values measured by DD-PCR were higher than those measured by real-time, whereas in 6% DD-PCR values were lower. In the remaining 65% of tests, both techniques gave the same values. In conclusion, the DD-PCR appeared as a promising molecular technique in HCL.