CLONAL ORIGIN OF THERAPY-RELATED MYELOID NEOPLASMS
(Abstract release date: 05/19/16)
EHA Library. Takahashi K. 06/11/16; 135202; S446
Dr. Koichi Takahashi
Contributions
Contributions
Abstract
Abstract: S446
Type: Oral Presentation
Presentation during EHA21: On Saturday, June 11, 2016 from 12:30 - 12:45
Location: Hall C11
Background
Therapy-related myeloid neoplasms (t-MNs) are secondary malignancies that develop in patients treated with chemotherapy and/or radiation therapy (C/RT). Previously, it was thought that genotoxic stress from C/RT induced driver mutations in hematopoietic stem cells leading to the development of t-MNs. However, a recent study suggested that pre-leukemic TP53 mutations could pre-exist in t-MN patients years before t-MN development. Further, hematological driver mutations have been detected in peripheral blood (PB) from apparently healthy individuals or patients with solid tumors, a phenomenon referred to as clonal hematopoiesis of indeterminate potential (CHIP).
Aims
To test the hypothesis that pre-leukemic driver mutations are detectable in t-MN patients at the time of primary cancer diagnosis and prior to C/RT exposures.
Methods
We identified 14 patients with t-MNs who were found to have paired samples of diagnostic bone marrow (BM) at the time of t-MN diagnosis and PB obtained after the diagnosis of their primary cancers but prior to C/RT. Targeted gene sequencing of the 280 leukemia-related genes was performed on the 14 t-MNs diagnostic BM to detect driver mutations. We then assessed presence of the same driver mutations in the patients’ matched PB samples taken at the time of primary cancer diagnosis. Because pre-leukemic driver mutations in the PB samples were expected to have very low variant allele frequency (VAF), we performed molecular barcoding deep sequencing of 32 genes that were reported as CHIP-associated mutation.
Results
Of the 14 t-MN patients, 5 (36%) had t-AML and 9 (64%) had t-MDS. The median age at primary cancer diagnosis and at t-MNs diagnosis was 62 years (range: 25-74) and 66 years (range: 28-77), respectively. The median latency from primary cancer to t-MNs was 3 years (range: 1-8). In the t-MN BM, 3 patients (21%) had normal karyotype and 7 (50%) had del 7q/-7, 4 (29%) had del 5q/-5 and 5 (36%) had complex karyotypes. Targeted gene sequencing of t-MN BM (median 289x): revealed 21 canonical hematological driver mutations in 14 t-MNs patients: mutations in TP53 (29%), DNMT3A (21%), TET2 (21%), RUNX1 (21%), IDH2 (14%), SRSF2 (7%), EZH2 (7%), FLT3 (7%), NRAS (7%), PTPN11 (7%) and GATA2 (7%).Molecular barcoding deep sequencing of the PB samples (median 3,000x) taken at the time of primary cancer diagnosis revealed that pre-leukemic driver mutations were detected in 10 out of 14 (71%) patients. Figure 1 shows the model of clonal evolution from pre-leukemic driver mutations to t-MN in 3 representative cases. For example, Case UID6982 had limited stage small cell lung cancer and received concurrent chemo-radiation therapy with carboplatin and etoposide. He developed t-AML 3 years after C/RT and was found to have an IDH2 p. R140Q (VAF 25%) and SRSF2 p.P95fs (46%) mutations in the diagnostic BM. His PB samples obtained before C/RT showed the same IDH2 p.R140Q and SRSF2 p.P95fs mutations with VAF of 16.4% and 9.8%, respectively.By genes, 75% of TP53, 67% of DNMT3A, 100% of TET2, 67% of RUNX1, 50% of IDH2, 100% of SRSF2, 100% of FLT3, 100% of NRAS, and 100% of PTPN11 mutations were detected as pre-leukemic driver mutations at time of primary cancer diagnosis. We did not detect EZH2 and GATA2 mutations as pre-leukemic mutations. The median VAF of the detected pre-leukemic driver mutations was 7.6% (range: 0.5-23%). There was no statistical difference between patients with and without detectable pre-leukemic mutations in terms of age at primary cancer diagnosis or latency time to t-MN development (P = 0.73 and 0.54, respectively).
Conclusion
In this study we have demonstrated evidence of detectable pre-leukemic driver mutations in multiple leukemia driver genes at time of diagnosis and before therapy of primary cancer in patients whose subsequent t-MNs also harbored identical driver mutations. These data suggest the potential to develop a risk stratification model based on presence of CHIP with canonical driver mutations at the time of primary cancer diagnosis. Determining predictive value of pre-leukemic driver mutations in t-MN development is currently undergoing.
Session topic: Myelodysplastic syndromes - Biology
Keyword(s): Clonal expansion, MDS/AML, Therapy-related AML
Type: Oral Presentation
Presentation during EHA21: On Saturday, June 11, 2016 from 12:30 - 12:45
Location: Hall C11
Background
Therapy-related myeloid neoplasms (t-MNs) are secondary malignancies that develop in patients treated with chemotherapy and/or radiation therapy (C/RT). Previously, it was thought that genotoxic stress from C/RT induced driver mutations in hematopoietic stem cells leading to the development of t-MNs. However, a recent study suggested that pre-leukemic TP53 mutations could pre-exist in t-MN patients years before t-MN development. Further, hematological driver mutations have been detected in peripheral blood (PB) from apparently healthy individuals or patients with solid tumors, a phenomenon referred to as clonal hematopoiesis of indeterminate potential (CHIP).
Aims
To test the hypothesis that pre-leukemic driver mutations are detectable in t-MN patients at the time of primary cancer diagnosis and prior to C/RT exposures.
Methods
We identified 14 patients with t-MNs who were found to have paired samples of diagnostic bone marrow (BM) at the time of t-MN diagnosis and PB obtained after the diagnosis of their primary cancers but prior to C/RT. Targeted gene sequencing of the 280 leukemia-related genes was performed on the 14 t-MNs diagnostic BM to detect driver mutations. We then assessed presence of the same driver mutations in the patients’ matched PB samples taken at the time of primary cancer diagnosis. Because pre-leukemic driver mutations in the PB samples were expected to have very low variant allele frequency (VAF), we performed molecular barcoding deep sequencing of 32 genes that were reported as CHIP-associated mutation.
Results
Of the 14 t-MN patients, 5 (36%) had t-AML and 9 (64%) had t-MDS. The median age at primary cancer diagnosis and at t-MNs diagnosis was 62 years (range: 25-74) and 66 years (range: 28-77), respectively. The median latency from primary cancer to t-MNs was 3 years (range: 1-8). In the t-MN BM, 3 patients (21%) had normal karyotype and 7 (50%) had del 7q/-7, 4 (29%) had del 5q/-5 and 5 (36%) had complex karyotypes. Targeted gene sequencing of t-MN BM (median 289x): revealed 21 canonical hematological driver mutations in 14 t-MNs patients: mutations in TP53 (29%), DNMT3A (21%), TET2 (21%), RUNX1 (21%), IDH2 (14%), SRSF2 (7%), EZH2 (7%), FLT3 (7%), NRAS (7%), PTPN11 (7%) and GATA2 (7%).Molecular barcoding deep sequencing of the PB samples (median 3,000x) taken at the time of primary cancer diagnosis revealed that pre-leukemic driver mutations were detected in 10 out of 14 (71%) patients. Figure 1 shows the model of clonal evolution from pre-leukemic driver mutations to t-MN in 3 representative cases. For example, Case UID6982 had limited stage small cell lung cancer and received concurrent chemo-radiation therapy with carboplatin and etoposide. He developed t-AML 3 years after C/RT and was found to have an IDH2 p. R140Q (VAF 25%) and SRSF2 p.P95fs (46%) mutations in the diagnostic BM. His PB samples obtained before C/RT showed the same IDH2 p.R140Q and SRSF2 p.P95fs mutations with VAF of 16.4% and 9.8%, respectively.By genes, 75% of TP53, 67% of DNMT3A, 100% of TET2, 67% of RUNX1, 50% of IDH2, 100% of SRSF2, 100% of FLT3, 100% of NRAS, and 100% of PTPN11 mutations were detected as pre-leukemic driver mutations at time of primary cancer diagnosis. We did not detect EZH2 and GATA2 mutations as pre-leukemic mutations. The median VAF of the detected pre-leukemic driver mutations was 7.6% (range: 0.5-23%). There was no statistical difference between patients with and without detectable pre-leukemic mutations in terms of age at primary cancer diagnosis or latency time to t-MN development (P = 0.73 and 0.54, respectively).
Conclusion
In this study we have demonstrated evidence of detectable pre-leukemic driver mutations in multiple leukemia driver genes at time of diagnosis and before therapy of primary cancer in patients whose subsequent t-MNs also harbored identical driver mutations. These data suggest the potential to develop a risk stratification model based on presence of CHIP with canonical driver mutations at the time of primary cancer diagnosis. Determining predictive value of pre-leukemic driver mutations in t-MN development is currently undergoing.
Session topic: Myelodysplastic syndromes - Biology
Keyword(s): Clonal expansion, MDS/AML, Therapy-related AML
Abstract: S446
Type: Oral Presentation
Presentation during EHA21: On Saturday, June 11, 2016 from 12:30 - 12:45
Location: Hall C11
Background
Therapy-related myeloid neoplasms (t-MNs) are secondary malignancies that develop in patients treated with chemotherapy and/or radiation therapy (C/RT). Previously, it was thought that genotoxic stress from C/RT induced driver mutations in hematopoietic stem cells leading to the development of t-MNs. However, a recent study suggested that pre-leukemic TP53 mutations could pre-exist in t-MN patients years before t-MN development. Further, hematological driver mutations have been detected in peripheral blood (PB) from apparently healthy individuals or patients with solid tumors, a phenomenon referred to as clonal hematopoiesis of indeterminate potential (CHIP).
Aims
To test the hypothesis that pre-leukemic driver mutations are detectable in t-MN patients at the time of primary cancer diagnosis and prior to C/RT exposures.
Methods
We identified 14 patients with t-MNs who were found to have paired samples of diagnostic bone marrow (BM) at the time of t-MN diagnosis and PB obtained after the diagnosis of their primary cancers but prior to C/RT. Targeted gene sequencing of the 280 leukemia-related genes was performed on the 14 t-MNs diagnostic BM to detect driver mutations. We then assessed presence of the same driver mutations in the patients’ matched PB samples taken at the time of primary cancer diagnosis. Because pre-leukemic driver mutations in the PB samples were expected to have very low variant allele frequency (VAF), we performed molecular barcoding deep sequencing of 32 genes that were reported as CHIP-associated mutation.
Results
Of the 14 t-MN patients, 5 (36%) had t-AML and 9 (64%) had t-MDS. The median age at primary cancer diagnosis and at t-MNs diagnosis was 62 years (range: 25-74) and 66 years (range: 28-77), respectively. The median latency from primary cancer to t-MNs was 3 years (range: 1-8). In the t-MN BM, 3 patients (21%) had normal karyotype and 7 (50%) had del 7q/-7, 4 (29%) had del 5q/-5 and 5 (36%) had complex karyotypes. Targeted gene sequencing of t-MN BM (median 289x): revealed 21 canonical hematological driver mutations in 14 t-MNs patients: mutations in TP53 (29%), DNMT3A (21%), TET2 (21%), RUNX1 (21%), IDH2 (14%), SRSF2 (7%), EZH2 (7%), FLT3 (7%), NRAS (7%), PTPN11 (7%) and GATA2 (7%).Molecular barcoding deep sequencing of the PB samples (median 3,000x) taken at the time of primary cancer diagnosis revealed that pre-leukemic driver mutations were detected in 10 out of 14 (71%) patients. Figure 1 shows the model of clonal evolution from pre-leukemic driver mutations to t-MN in 3 representative cases. For example, Case UID6982 had limited stage small cell lung cancer and received concurrent chemo-radiation therapy with carboplatin and etoposide. He developed t-AML 3 years after C/RT and was found to have an IDH2 p. R140Q (VAF 25%) and SRSF2 p.P95fs (46%) mutations in the diagnostic BM. His PB samples obtained before C/RT showed the same IDH2 p.R140Q and SRSF2 p.P95fs mutations with VAF of 16.4% and 9.8%, respectively.By genes, 75% of TP53, 67% of DNMT3A, 100% of TET2, 67% of RUNX1, 50% of IDH2, 100% of SRSF2, 100% of FLT3, 100% of NRAS, and 100% of PTPN11 mutations were detected as pre-leukemic driver mutations at time of primary cancer diagnosis. We did not detect EZH2 and GATA2 mutations as pre-leukemic mutations. The median VAF of the detected pre-leukemic driver mutations was 7.6% (range: 0.5-23%). There was no statistical difference between patients with and without detectable pre-leukemic mutations in terms of age at primary cancer diagnosis or latency time to t-MN development (P = 0.73 and 0.54, respectively).
Conclusion
In this study we have demonstrated evidence of detectable pre-leukemic driver mutations in multiple leukemia driver genes at time of diagnosis and before therapy of primary cancer in patients whose subsequent t-MNs also harbored identical driver mutations. These data suggest the potential to develop a risk stratification model based on presence of CHIP with canonical driver mutations at the time of primary cancer diagnosis. Determining predictive value of pre-leukemic driver mutations in t-MN development is currently undergoing.
Session topic: Myelodysplastic syndromes - Biology
Keyword(s): Clonal expansion, MDS/AML, Therapy-related AML
Type: Oral Presentation
Presentation during EHA21: On Saturday, June 11, 2016 from 12:30 - 12:45
Location: Hall C11
Background
Therapy-related myeloid neoplasms (t-MNs) are secondary malignancies that develop in patients treated with chemotherapy and/or radiation therapy (C/RT). Previously, it was thought that genotoxic stress from C/RT induced driver mutations in hematopoietic stem cells leading to the development of t-MNs. However, a recent study suggested that pre-leukemic TP53 mutations could pre-exist in t-MN patients years before t-MN development. Further, hematological driver mutations have been detected in peripheral blood (PB) from apparently healthy individuals or patients with solid tumors, a phenomenon referred to as clonal hematopoiesis of indeterminate potential (CHIP).
Aims
To test the hypothesis that pre-leukemic driver mutations are detectable in t-MN patients at the time of primary cancer diagnosis and prior to C/RT exposures.
Methods
We identified 14 patients with t-MNs who were found to have paired samples of diagnostic bone marrow (BM) at the time of t-MN diagnosis and PB obtained after the diagnosis of their primary cancers but prior to C/RT. Targeted gene sequencing of the 280 leukemia-related genes was performed on the 14 t-MNs diagnostic BM to detect driver mutations. We then assessed presence of the same driver mutations in the patients’ matched PB samples taken at the time of primary cancer diagnosis. Because pre-leukemic driver mutations in the PB samples were expected to have very low variant allele frequency (VAF), we performed molecular barcoding deep sequencing of 32 genes that were reported as CHIP-associated mutation.
Results
Of the 14 t-MN patients, 5 (36%) had t-AML and 9 (64%) had t-MDS. The median age at primary cancer diagnosis and at t-MNs diagnosis was 62 years (range: 25-74) and 66 years (range: 28-77), respectively. The median latency from primary cancer to t-MNs was 3 years (range: 1-8). In the t-MN BM, 3 patients (21%) had normal karyotype and 7 (50%) had del 7q/-7, 4 (29%) had del 5q/-5 and 5 (36%) had complex karyotypes. Targeted gene sequencing of t-MN BM (median 289x): revealed 21 canonical hematological driver mutations in 14 t-MNs patients: mutations in TP53 (29%), DNMT3A (21%), TET2 (21%), RUNX1 (21%), IDH2 (14%), SRSF2 (7%), EZH2 (7%), FLT3 (7%), NRAS (7%), PTPN11 (7%) and GATA2 (7%).Molecular barcoding deep sequencing of the PB samples (median 3,000x) taken at the time of primary cancer diagnosis revealed that pre-leukemic driver mutations were detected in 10 out of 14 (71%) patients. Figure 1 shows the model of clonal evolution from pre-leukemic driver mutations to t-MN in 3 representative cases. For example, Case UID6982 had limited stage small cell lung cancer and received concurrent chemo-radiation therapy with carboplatin and etoposide. He developed t-AML 3 years after C/RT and was found to have an IDH2 p. R140Q (VAF 25%) and SRSF2 p.P95fs (46%) mutations in the diagnostic BM. His PB samples obtained before C/RT showed the same IDH2 p.R140Q and SRSF2 p.P95fs mutations with VAF of 16.4% and 9.8%, respectively.By genes, 75% of TP53, 67% of DNMT3A, 100% of TET2, 67% of RUNX1, 50% of IDH2, 100% of SRSF2, 100% of FLT3, 100% of NRAS, and 100% of PTPN11 mutations were detected as pre-leukemic driver mutations at time of primary cancer diagnosis. We did not detect EZH2 and GATA2 mutations as pre-leukemic mutations. The median VAF of the detected pre-leukemic driver mutations was 7.6% (range: 0.5-23%). There was no statistical difference between patients with and without detectable pre-leukemic mutations in terms of age at primary cancer diagnosis or latency time to t-MN development (P = 0.73 and 0.54, respectively).
Conclusion
In this study we have demonstrated evidence of detectable pre-leukemic driver mutations in multiple leukemia driver genes at time of diagnosis and before therapy of primary cancer in patients whose subsequent t-MNs also harbored identical driver mutations. These data suggest the potential to develop a risk stratification model based on presence of CHIP with canonical driver mutations at the time of primary cancer diagnosis. Determining predictive value of pre-leukemic driver mutations in t-MN development is currently undergoing.
Session topic: Myelodysplastic syndromes - Biology
Keyword(s): Clonal expansion, MDS/AML, Therapy-related AML
{{ help_message }}
{{filter}}