PATTERN OF CNS RELAPSE IN ACUTE LYMPHOBLASTIC LEUKEMIA BCR-ABL POSITIVE, THE IMPORTANCE OF CHARACTERIZATION OF ABL1 MUTATIONS IN CEREBROSPINAL FLUID.
(Abstract release date: 05/19/16)
EHA Library. Martinez-Lopez J. 06/09/16; 132414; E865
Joaquin Martinez-Lopez
Contributions
Contributions
Abstract
Abstract: E865
Type: Eposter Presentation
Background
he incidence of central nervous system (CNS) relapse among patients with BCR/ABL-positive acute lymphoblastic leukemia (ALL) is 8-17%. Although tyrosine-kinase inhibitors (TKI), especially imatinib, are included in the first-line treatment in these patients, their concentration in CNS is too low to effectively prevent CNS relapse, making CNS prophylaxis mandatory.
Aims
To study the frequency, predictors and evolution of BCR-ABL positive ALL patients with CNS relapse in two consecutive clinical trials of the PETHEMA group using imatinib and chemotherapy. As secondary objective we proposed the introduction of a new method for the study of variants of uncertain significance (VUS) in kinase domain of the BCR-ABL from cDNA of cerebrospinal fluid (CSF) blasts, in order to adapt the TKI used in relapse according to the clonal evolution from bone marrow (BM) to CSF cells.
Methods
It has been reviewed data of CNS relapse of patients included in two consecutive clinical PETHEMA trials (PETHEMA-LAL-PH-2008 and PETHEMA-LAL-OPH-2007), for treatment of BCR-ABL- Ph+ ALL patients. In two patients with combined BM and CSF we analyzed the BCR-ABL mutations at diagnosis and at relapse in both sites. Kinase domain of cDNA was amplified in two steps by nested PCR in the samples of BM at diagnosis and BM or CSF at relapse, covering the whole kinase domain from residues Gly227 to Gly514. Further enzymatic fragmentation of the domain yielded ~200 bp of small fragments. Ion Torrent ultra deep-sequencing of the resulting fragments allowed us to evaluate the variants of the chimeric protein.
Results
A total of 138 ALL BCR/ABL positive patients were analyzed and 128 reached complete remission, 30 of them have relapsed (13 [43%] involving CNS [isolated or combined], 16 [53%] BM and 1 in unknown site). The overall survival probability at 2 years was 69% for the 'CNS Relapse' group (IC95%: 44-94%; p=0.067) and 44% for the 'Not CNS Relapse' group (IC95%: 20-68%; p=0.067, Figure 1a). In a multivariable analysis any clinical variable was associated with CNS relapse probability. In two of the CNS and BM relapsed patients we have performed massive sequencing experiments of ABL1 Kinase domain mutational status. In the first patient ultra-deep NGS of BM samples at diagnosis and at relapse did not show VUS or pathogenic variants, but the CSF study at relapse confirmed the variant c.1159 T>A, p.L387M. The same study in the second patient also found the same pathogenic variant only in CSF blast cells (Figure 1b), whereas, the VUS c.733 A>G, p.K245E was found only in the BM sample at diagnosis.
Conclusion
In BCR-ABL ALL patients treated with imatinib and chemotherapy CNS relapse was a significant feature, despite CNS prophylaxis, In our series we did not find any clinical variable predicting CNS relapse . We have found the pathogenic variant p.L387M in CSF blasts of two patients with combined CNS and BM recurrence, this variant not being found in BM samples at diagnosis or at relapse. These mutations were sensitive to other TKIs with better penetration to CSF. Based on these results, a mutational study of the kinase domain of the BCR-ABL in blast cells obtained from CSF, should also be integrated in the mutation study of these patients in order to select the TKI according to the clonal evolution from BM to CSF cells.
Session topic: E-poster
Keyword(s): CNS, Mutation analysis, Relapsed acute lymphoblastic leukemia
Type: Eposter Presentation
Background
he incidence of central nervous system (CNS) relapse among patients with BCR/ABL-positive acute lymphoblastic leukemia (ALL) is 8-17%. Although tyrosine-kinase inhibitors (TKI), especially imatinib, are included in the first-line treatment in these patients, their concentration in CNS is too low to effectively prevent CNS relapse, making CNS prophylaxis mandatory.
Aims
To study the frequency, predictors and evolution of BCR-ABL positive ALL patients with CNS relapse in two consecutive clinical trials of the PETHEMA group using imatinib and chemotherapy. As secondary objective we proposed the introduction of a new method for the study of variants of uncertain significance (VUS) in kinase domain of the BCR-ABL from cDNA of cerebrospinal fluid (CSF) blasts, in order to adapt the TKI used in relapse according to the clonal evolution from bone marrow (BM) to CSF cells.
Methods
It has been reviewed data of CNS relapse of patients included in two consecutive clinical PETHEMA trials (PETHEMA-LAL-PH-2008 and PETHEMA-LAL-OPH-2007), for treatment of BCR-ABL- Ph+ ALL patients. In two patients with combined BM and CSF we analyzed the BCR-ABL mutations at diagnosis and at relapse in both sites. Kinase domain of cDNA was amplified in two steps by nested PCR in the samples of BM at diagnosis and BM or CSF at relapse, covering the whole kinase domain from residues Gly227 to Gly514. Further enzymatic fragmentation of the domain yielded ~200 bp of small fragments. Ion Torrent ultra deep-sequencing of the resulting fragments allowed us to evaluate the variants of the chimeric protein.
Results
A total of 138 ALL BCR/ABL positive patients were analyzed and 128 reached complete remission, 30 of them have relapsed (13 [43%] involving CNS [isolated or combined], 16 [53%] BM and 1 in unknown site). The overall survival probability at 2 years was 69% for the 'CNS Relapse' group (IC95%: 44-94%; p=0.067) and 44% for the 'Not CNS Relapse' group (IC95%: 20-68%; p=0.067, Figure 1a). In a multivariable analysis any clinical variable was associated with CNS relapse probability. In two of the CNS and BM relapsed patients we have performed massive sequencing experiments of ABL1 Kinase domain mutational status. In the first patient ultra-deep NGS of BM samples at diagnosis and at relapse did not show VUS or pathogenic variants, but the CSF study at relapse confirmed the variant c.1159 T>A, p.L387M. The same study in the second patient also found the same pathogenic variant only in CSF blast cells (Figure 1b), whereas, the VUS c.733 A>G, p.K245E was found only in the BM sample at diagnosis.
Conclusion
In BCR-ABL ALL patients treated with imatinib and chemotherapy CNS relapse was a significant feature, despite CNS prophylaxis, In our series we did not find any clinical variable predicting CNS relapse . We have found the pathogenic variant p.L387M in CSF blasts of two patients with combined CNS and BM recurrence, this variant not being found in BM samples at diagnosis or at relapse. These mutations were sensitive to other TKIs with better penetration to CSF. Based on these results, a mutational study of the kinase domain of the BCR-ABL in blast cells obtained from CSF, should also be integrated in the mutation study of these patients in order to select the TKI according to the clonal evolution from BM to CSF cells.
Session topic: E-poster
Keyword(s): CNS, Mutation analysis, Relapsed acute lymphoblastic leukemia
Abstract: E865
Type: Eposter Presentation
Background
he incidence of central nervous system (CNS) relapse among patients with BCR/ABL-positive acute lymphoblastic leukemia (ALL) is 8-17%. Although tyrosine-kinase inhibitors (TKI), especially imatinib, are included in the first-line treatment in these patients, their concentration in CNS is too low to effectively prevent CNS relapse, making CNS prophylaxis mandatory.
Aims
To study the frequency, predictors and evolution of BCR-ABL positive ALL patients with CNS relapse in two consecutive clinical trials of the PETHEMA group using imatinib and chemotherapy. As secondary objective we proposed the introduction of a new method for the study of variants of uncertain significance (VUS) in kinase domain of the BCR-ABL from cDNA of cerebrospinal fluid (CSF) blasts, in order to adapt the TKI used in relapse according to the clonal evolution from bone marrow (BM) to CSF cells.
Methods
It has been reviewed data of CNS relapse of patients included in two consecutive clinical PETHEMA trials (PETHEMA-LAL-PH-2008 and PETHEMA-LAL-OPH-2007), for treatment of BCR-ABL- Ph+ ALL patients. In two patients with combined BM and CSF we analyzed the BCR-ABL mutations at diagnosis and at relapse in both sites. Kinase domain of cDNA was amplified in two steps by nested PCR in the samples of BM at diagnosis and BM or CSF at relapse, covering the whole kinase domain from residues Gly227 to Gly514. Further enzymatic fragmentation of the domain yielded ~200 bp of small fragments. Ion Torrent ultra deep-sequencing of the resulting fragments allowed us to evaluate the variants of the chimeric protein.
Results
A total of 138 ALL BCR/ABL positive patients were analyzed and 128 reached complete remission, 30 of them have relapsed (13 [43%] involving CNS [isolated or combined], 16 [53%] BM and 1 in unknown site). The overall survival probability at 2 years was 69% for the 'CNS Relapse' group (IC95%: 44-94%; p=0.067) and 44% for the 'Not CNS Relapse' group (IC95%: 20-68%; p=0.067, Figure 1a). In a multivariable analysis any clinical variable was associated with CNS relapse probability. In two of the CNS and BM relapsed patients we have performed massive sequencing experiments of ABL1 Kinase domain mutational status. In the first patient ultra-deep NGS of BM samples at diagnosis and at relapse did not show VUS or pathogenic variants, but the CSF study at relapse confirmed the variant c.1159 T>A, p.L387M. The same study in the second patient also found the same pathogenic variant only in CSF blast cells (Figure 1b), whereas, the VUS c.733 A>G, p.K245E was found only in the BM sample at diagnosis.
Conclusion
In BCR-ABL ALL patients treated with imatinib and chemotherapy CNS relapse was a significant feature, despite CNS prophylaxis, In our series we did not find any clinical variable predicting CNS relapse . We have found the pathogenic variant p.L387M in CSF blasts of two patients with combined CNS and BM recurrence, this variant not being found in BM samples at diagnosis or at relapse. These mutations were sensitive to other TKIs with better penetration to CSF. Based on these results, a mutational study of the kinase domain of the BCR-ABL in blast cells obtained from CSF, should also be integrated in the mutation study of these patients in order to select the TKI according to the clonal evolution from BM to CSF cells.
Session topic: E-poster
Keyword(s): CNS, Mutation analysis, Relapsed acute lymphoblastic leukemia
Type: Eposter Presentation
Background
he incidence of central nervous system (CNS) relapse among patients with BCR/ABL-positive acute lymphoblastic leukemia (ALL) is 8-17%. Although tyrosine-kinase inhibitors (TKI), especially imatinib, are included in the first-line treatment in these patients, their concentration in CNS is too low to effectively prevent CNS relapse, making CNS prophylaxis mandatory.
Aims
To study the frequency, predictors and evolution of BCR-ABL positive ALL patients with CNS relapse in two consecutive clinical trials of the PETHEMA group using imatinib and chemotherapy. As secondary objective we proposed the introduction of a new method for the study of variants of uncertain significance (VUS) in kinase domain of the BCR-ABL from cDNA of cerebrospinal fluid (CSF) blasts, in order to adapt the TKI used in relapse according to the clonal evolution from bone marrow (BM) to CSF cells.
Methods
It has been reviewed data of CNS relapse of patients included in two consecutive clinical PETHEMA trials (PETHEMA-LAL-PH-2008 and PETHEMA-LAL-OPH-2007), for treatment of BCR-ABL- Ph+ ALL patients. In two patients with combined BM and CSF we analyzed the BCR-ABL mutations at diagnosis and at relapse in both sites. Kinase domain of cDNA was amplified in two steps by nested PCR in the samples of BM at diagnosis and BM or CSF at relapse, covering the whole kinase domain from residues Gly227 to Gly514. Further enzymatic fragmentation of the domain yielded ~200 bp of small fragments. Ion Torrent ultra deep-sequencing of the resulting fragments allowed us to evaluate the variants of the chimeric protein.
Results
A total of 138 ALL BCR/ABL positive patients were analyzed and 128 reached complete remission, 30 of them have relapsed (13 [43%] involving CNS [isolated or combined], 16 [53%] BM and 1 in unknown site). The overall survival probability at 2 years was 69% for the 'CNS Relapse' group (IC95%: 44-94%; p=0.067) and 44% for the 'Not CNS Relapse' group (IC95%: 20-68%; p=0.067, Figure 1a). In a multivariable analysis any clinical variable was associated with CNS relapse probability. In two of the CNS and BM relapsed patients we have performed massive sequencing experiments of ABL1 Kinase domain mutational status. In the first patient ultra-deep NGS of BM samples at diagnosis and at relapse did not show VUS or pathogenic variants, but the CSF study at relapse confirmed the variant c.1159 T>A, p.L387M. The same study in the second patient also found the same pathogenic variant only in CSF blast cells (Figure 1b), whereas, the VUS c.733 A>G, p.K245E was found only in the BM sample at diagnosis.
Conclusion
In BCR-ABL ALL patients treated with imatinib and chemotherapy CNS relapse was a significant feature, despite CNS prophylaxis, In our series we did not find any clinical variable predicting CNS relapse . We have found the pathogenic variant p.L387M in CSF blasts of two patients with combined CNS and BM recurrence, this variant not being found in BM samples at diagnosis or at relapse. These mutations were sensitive to other TKIs with better penetration to CSF. Based on these results, a mutational study of the kinase domain of the BCR-ABL in blast cells obtained from CSF, should also be integrated in the mutation study of these patients in order to select the TKI according to the clonal evolution from BM to CSF cells.
Session topic: E-poster
Keyword(s): CNS, Mutation analysis, Relapsed acute lymphoblastic leukemia
{{ help_message }}
{{filter}}