DYNAMICS OF SOMATIC MUTATIONS COMMONLY DETECTED IN OTHER MYELOID NEOPLASMS BEYOND THE BCR-ABL GENE REARRANGEMENT IN RESPONSE TO TYROSINE KINASE INHIBITOR THERAPY IN CHRONIC MYELOID LEUKEMIA
(Abstract release date: 05/19/16)
EHA Library. Kim T. 06/11/16; 135249; S493
Mr. Taehyung Kim
Contributions
Contributions
Abstract
Abstract: S493
Type: Oral Presentation
Presentation during EHA21: On Saturday, June 11, 2016 from 16:30 - 16:45
Location: Hall C11
Background
Chronic myeloid leukemia (CML) is mainly characterized by a translocation event t(9;22)(q34;q11.2) which fuses the ABL1 gene and the BCR gene, yielding a bcr-abl onco-protein. Despite remarkable improvement in the treatment of CML via tyrosine kinase inhibitors (TKIs), a portion of patients show resistance to these treatments and have a much higher risk of progression to accelerated phase (AP) or blastic phase (BP) as a result. The survival of patients who progress to BP is comparable to the pre-TKI era, even with other salvage treatment modalities such as allogeneic stem cell transplantation. Using high-throughput sequencing technologies, extensive investigations in other myeloid malignancies have discovered commonly mutated genes and the biological pathways they effect. Other than ABL1 kinase domain (KD) mutations, very little is known about the origins and dynamics of relevant somatic variants, as well as their associations with BCR-ABL transcript changes and their clinical implications on TKI response in CML.
Aims
We aimed to determine whether patterns of mutation acquisition, persistence, and clearance can provide insight into treatment outcomes in CML.
Methods
100 CML patients were included for deep sequencing targeting 92 genes that are recurrently mutated in other myeloid neoplasms. Treatment failure was estimated to be 10.2% at 3 years while PFS at 5 years was 91.0%. The OS at 5 years was 92.6%. Based on ELN criteria, 74 patients were determined to have optimal response, 18 failed but remained in CP, and 8 had progressed to accelerated or blast phases. For each patient, samples taken at the initial diagnosis (prior to TKI therapy) and after TKI treatment, as well as T-cell samples, were sequenced in this study.
Results
In total, 64 variants from 32 genes in 37 patients were detected in 300 serial samples from 100 CML patients. ASXL1, ABL1, and TET2 were the 3 genes most frequently mutated (n=9, n=6, and n=6, respectively). Unsupervised hierarchical clustering of the 51 non-silent mutations across all serial samples revealed 5 distinct pattern of mutation dynamics throughout the course of CML (Figure 1). A majority of cases with mutations only had mutations from one pattern. Pattern 1 mutations arise at diagnosis and persist at follow-up. Despite this, all patients with Pattern 1 mutations were TKI-responsive. Since these mutations were not cleared in spite of significant reduction of BCR-ABL transcript level at the time of follow-up, they are likely to be indicative of abnormal and clonal Ph-negative hematopoiesis that existed prior to Ph-positive clones. Pattern 2 mutations are acquired during TKI treatment. This pattern included a high portion of ABL1 KD mutations (7/13 mutations in 6 patients). Perhaps unsurprisingly, all patients with Pattern 2 mutations failed TKI therapy. Pattern 3 mutations rise at the time of diagnosis and vanish or substantially decrease following TKI therapy. Patients with these mutations showed mixed patterns of clinical outcomes. Interestingly, Pattern 3 mutations were frequently within genes associated with chromatin modification and DNA methylation, which are epigenetic regulation pathways (11/17, 64.7%). Patients with Pattern 4 and 5 mutations showed evidence of preleukemic mutations in CML. Too few cases had these mutations to assess their association with treatment outcomes.
Conclusion
Overall, this study demonstrates that patterns of mutation acquisition, persistence, and clearance vary but have a number of interesting correlations with clinical outcomes. Our data show that mutation burden often persists despite successful TKI response in CML, particularly in mutations that are likely in persistent Ph-negative clones, while mutation clearance is not associated with particular outcomes. Patients that acquired new mutations during treatment all failed TKI therapy. We found evidence of preleukemic mutations in some CML patients. These patterns show that CML mutation dynamics following TKI therapy are markedly distinct from other hematologic malignancies. Figure 1. BCR-ABL level indicated on a log10 percentage scale for both subplots. Each circle in the left-hand plot represents a variant. Filled circles indicate TKI treatment failure. The right-hand plot shows patients without extra mutations. Each dot indicates an individual patient. The boxes on the right indicate the number of patients within each decade on TKI response coloured by pattern.
Session topic: Chronic myeloid leukemia - Biology
Keyword(s): Chronic myeloid leukemia, Drug resistance, Genomics, Progression
Type: Oral Presentation
Presentation during EHA21: On Saturday, June 11, 2016 from 16:30 - 16:45
Location: Hall C11
Background
Chronic myeloid leukemia (CML) is mainly characterized by a translocation event t(9;22)(q34;q11.2) which fuses the ABL1 gene and the BCR gene, yielding a bcr-abl onco-protein. Despite remarkable improvement in the treatment of CML via tyrosine kinase inhibitors (TKIs), a portion of patients show resistance to these treatments and have a much higher risk of progression to accelerated phase (AP) or blastic phase (BP) as a result. The survival of patients who progress to BP is comparable to the pre-TKI era, even with other salvage treatment modalities such as allogeneic stem cell transplantation. Using high-throughput sequencing technologies, extensive investigations in other myeloid malignancies have discovered commonly mutated genes and the biological pathways they effect. Other than ABL1 kinase domain (KD) mutations, very little is known about the origins and dynamics of relevant somatic variants, as well as their associations with BCR-ABL transcript changes and their clinical implications on TKI response in CML.
Aims
We aimed to determine whether patterns of mutation acquisition, persistence, and clearance can provide insight into treatment outcomes in CML.
Methods
100 CML patients were included for deep sequencing targeting 92 genes that are recurrently mutated in other myeloid neoplasms. Treatment failure was estimated to be 10.2% at 3 years while PFS at 5 years was 91.0%. The OS at 5 years was 92.6%. Based on ELN criteria, 74 patients were determined to have optimal response, 18 failed but remained in CP, and 8 had progressed to accelerated or blast phases. For each patient, samples taken at the initial diagnosis (prior to TKI therapy) and after TKI treatment, as well as T-cell samples, were sequenced in this study.
Results
In total, 64 variants from 32 genes in 37 patients were detected in 300 serial samples from 100 CML patients. ASXL1, ABL1, and TET2 were the 3 genes most frequently mutated (n=9, n=6, and n=6, respectively). Unsupervised hierarchical clustering of the 51 non-silent mutations across all serial samples revealed 5 distinct pattern of mutation dynamics throughout the course of CML (Figure 1). A majority of cases with mutations only had mutations from one pattern. Pattern 1 mutations arise at diagnosis and persist at follow-up. Despite this, all patients with Pattern 1 mutations were TKI-responsive. Since these mutations were not cleared in spite of significant reduction of BCR-ABL transcript level at the time of follow-up, they are likely to be indicative of abnormal and clonal Ph-negative hematopoiesis that existed prior to Ph-positive clones. Pattern 2 mutations are acquired during TKI treatment. This pattern included a high portion of ABL1 KD mutations (7/13 mutations in 6 patients). Perhaps unsurprisingly, all patients with Pattern 2 mutations failed TKI therapy. Pattern 3 mutations rise at the time of diagnosis and vanish or substantially decrease following TKI therapy. Patients with these mutations showed mixed patterns of clinical outcomes. Interestingly, Pattern 3 mutations were frequently within genes associated with chromatin modification and DNA methylation, which are epigenetic regulation pathways (11/17, 64.7%). Patients with Pattern 4 and 5 mutations showed evidence of preleukemic mutations in CML. Too few cases had these mutations to assess their association with treatment outcomes.
Conclusion
Overall, this study demonstrates that patterns of mutation acquisition, persistence, and clearance vary but have a number of interesting correlations with clinical outcomes. Our data show that mutation burden often persists despite successful TKI response in CML, particularly in mutations that are likely in persistent Ph-negative clones, while mutation clearance is not associated with particular outcomes. Patients that acquired new mutations during treatment all failed TKI therapy. We found evidence of preleukemic mutations in some CML patients. These patterns show that CML mutation dynamics following TKI therapy are markedly distinct from other hematologic malignancies. Figure 1. BCR-ABL level indicated on a log10 percentage scale for both subplots. Each circle in the left-hand plot represents a variant. Filled circles indicate TKI treatment failure. The right-hand plot shows patients without extra mutations. Each dot indicates an individual patient. The boxes on the right indicate the number of patients within each decade on TKI response coloured by pattern.
Session topic: Chronic myeloid leukemia - Biology
Keyword(s): Chronic myeloid leukemia, Drug resistance, Genomics, Progression
Abstract: S493
Type: Oral Presentation
Presentation during EHA21: On Saturday, June 11, 2016 from 16:30 - 16:45
Location: Hall C11
Background
Chronic myeloid leukemia (CML) is mainly characterized by a translocation event t(9;22)(q34;q11.2) which fuses the ABL1 gene and the BCR gene, yielding a bcr-abl onco-protein. Despite remarkable improvement in the treatment of CML via tyrosine kinase inhibitors (TKIs), a portion of patients show resistance to these treatments and have a much higher risk of progression to accelerated phase (AP) or blastic phase (BP) as a result. The survival of patients who progress to BP is comparable to the pre-TKI era, even with other salvage treatment modalities such as allogeneic stem cell transplantation. Using high-throughput sequencing technologies, extensive investigations in other myeloid malignancies have discovered commonly mutated genes and the biological pathways they effect. Other than ABL1 kinase domain (KD) mutations, very little is known about the origins and dynamics of relevant somatic variants, as well as their associations with BCR-ABL transcript changes and their clinical implications on TKI response in CML.
Aims
We aimed to determine whether patterns of mutation acquisition, persistence, and clearance can provide insight into treatment outcomes in CML.
Methods
100 CML patients were included for deep sequencing targeting 92 genes that are recurrently mutated in other myeloid neoplasms. Treatment failure was estimated to be 10.2% at 3 years while PFS at 5 years was 91.0%. The OS at 5 years was 92.6%. Based on ELN criteria, 74 patients were determined to have optimal response, 18 failed but remained in CP, and 8 had progressed to accelerated or blast phases. For each patient, samples taken at the initial diagnosis (prior to TKI therapy) and after TKI treatment, as well as T-cell samples, were sequenced in this study.
Results
In total, 64 variants from 32 genes in 37 patients were detected in 300 serial samples from 100 CML patients. ASXL1, ABL1, and TET2 were the 3 genes most frequently mutated (n=9, n=6, and n=6, respectively). Unsupervised hierarchical clustering of the 51 non-silent mutations across all serial samples revealed 5 distinct pattern of mutation dynamics throughout the course of CML (Figure 1). A majority of cases with mutations only had mutations from one pattern. Pattern 1 mutations arise at diagnosis and persist at follow-up. Despite this, all patients with Pattern 1 mutations were TKI-responsive. Since these mutations were not cleared in spite of significant reduction of BCR-ABL transcript level at the time of follow-up, they are likely to be indicative of abnormal and clonal Ph-negative hematopoiesis that existed prior to Ph-positive clones. Pattern 2 mutations are acquired during TKI treatment. This pattern included a high portion of ABL1 KD mutations (7/13 mutations in 6 patients). Perhaps unsurprisingly, all patients with Pattern 2 mutations failed TKI therapy. Pattern 3 mutations rise at the time of diagnosis and vanish or substantially decrease following TKI therapy. Patients with these mutations showed mixed patterns of clinical outcomes. Interestingly, Pattern 3 mutations were frequently within genes associated with chromatin modification and DNA methylation, which are epigenetic regulation pathways (11/17, 64.7%). Patients with Pattern 4 and 5 mutations showed evidence of preleukemic mutations in CML. Too few cases had these mutations to assess their association with treatment outcomes.
Conclusion
Overall, this study demonstrates that patterns of mutation acquisition, persistence, and clearance vary but have a number of interesting correlations with clinical outcomes. Our data show that mutation burden often persists despite successful TKI response in CML, particularly in mutations that are likely in persistent Ph-negative clones, while mutation clearance is not associated with particular outcomes. Patients that acquired new mutations during treatment all failed TKI therapy. We found evidence of preleukemic mutations in some CML patients. These patterns show that CML mutation dynamics following TKI therapy are markedly distinct from other hematologic malignancies. Figure 1. BCR-ABL level indicated on a log10 percentage scale for both subplots. Each circle in the left-hand plot represents a variant. Filled circles indicate TKI treatment failure. The right-hand plot shows patients without extra mutations. Each dot indicates an individual patient. The boxes on the right indicate the number of patients within each decade on TKI response coloured by pattern.
Session topic: Chronic myeloid leukemia - Biology
Keyword(s): Chronic myeloid leukemia, Drug resistance, Genomics, Progression
Type: Oral Presentation
Presentation during EHA21: On Saturday, June 11, 2016 from 16:30 - 16:45
Location: Hall C11
Background
Chronic myeloid leukemia (CML) is mainly characterized by a translocation event t(9;22)(q34;q11.2) which fuses the ABL1 gene and the BCR gene, yielding a bcr-abl onco-protein. Despite remarkable improvement in the treatment of CML via tyrosine kinase inhibitors (TKIs), a portion of patients show resistance to these treatments and have a much higher risk of progression to accelerated phase (AP) or blastic phase (BP) as a result. The survival of patients who progress to BP is comparable to the pre-TKI era, even with other salvage treatment modalities such as allogeneic stem cell transplantation. Using high-throughput sequencing technologies, extensive investigations in other myeloid malignancies have discovered commonly mutated genes and the biological pathways they effect. Other than ABL1 kinase domain (KD) mutations, very little is known about the origins and dynamics of relevant somatic variants, as well as their associations with BCR-ABL transcript changes and their clinical implications on TKI response in CML.
Aims
We aimed to determine whether patterns of mutation acquisition, persistence, and clearance can provide insight into treatment outcomes in CML.
Methods
100 CML patients were included for deep sequencing targeting 92 genes that are recurrently mutated in other myeloid neoplasms. Treatment failure was estimated to be 10.2% at 3 years while PFS at 5 years was 91.0%. The OS at 5 years was 92.6%. Based on ELN criteria, 74 patients were determined to have optimal response, 18 failed but remained in CP, and 8 had progressed to accelerated or blast phases. For each patient, samples taken at the initial diagnosis (prior to TKI therapy) and after TKI treatment, as well as T-cell samples, were sequenced in this study.
Results
In total, 64 variants from 32 genes in 37 patients were detected in 300 serial samples from 100 CML patients. ASXL1, ABL1, and TET2 were the 3 genes most frequently mutated (n=9, n=6, and n=6, respectively). Unsupervised hierarchical clustering of the 51 non-silent mutations across all serial samples revealed 5 distinct pattern of mutation dynamics throughout the course of CML (Figure 1). A majority of cases with mutations only had mutations from one pattern. Pattern 1 mutations arise at diagnosis and persist at follow-up. Despite this, all patients with Pattern 1 mutations were TKI-responsive. Since these mutations were not cleared in spite of significant reduction of BCR-ABL transcript level at the time of follow-up, they are likely to be indicative of abnormal and clonal Ph-negative hematopoiesis that existed prior to Ph-positive clones. Pattern 2 mutations are acquired during TKI treatment. This pattern included a high portion of ABL1 KD mutations (7/13 mutations in 6 patients). Perhaps unsurprisingly, all patients with Pattern 2 mutations failed TKI therapy. Pattern 3 mutations rise at the time of diagnosis and vanish or substantially decrease following TKI therapy. Patients with these mutations showed mixed patterns of clinical outcomes. Interestingly, Pattern 3 mutations were frequently within genes associated with chromatin modification and DNA methylation, which are epigenetic regulation pathways (11/17, 64.7%). Patients with Pattern 4 and 5 mutations showed evidence of preleukemic mutations in CML. Too few cases had these mutations to assess their association with treatment outcomes.
Conclusion
Overall, this study demonstrates that patterns of mutation acquisition, persistence, and clearance vary but have a number of interesting correlations with clinical outcomes. Our data show that mutation burden often persists despite successful TKI response in CML, particularly in mutations that are likely in persistent Ph-negative clones, while mutation clearance is not associated with particular outcomes. Patients that acquired new mutations during treatment all failed TKI therapy. We found evidence of preleukemic mutations in some CML patients. These patterns show that CML mutation dynamics following TKI therapy are markedly distinct from other hematologic malignancies. Figure 1. BCR-ABL level indicated on a log10 percentage scale for both subplots. Each circle in the left-hand plot represents a variant. Filled circles indicate TKI treatment failure. The right-hand plot shows patients without extra mutations. Each dot indicates an individual patient. The boxes on the right indicate the number of patients within each decade on TKI response coloured by pattern.
Session topic: Chronic myeloid leukemia - Biology
Keyword(s): Chronic myeloid leukemia, Drug resistance, Genomics, Progression
{{ help_message }}
{{filter}}