IMPACT OF THE MUTATIONAL PROFILE AT THE TIME OF DIAGNOSIS IN RESPONSE OUTCOMES IN PATIENTS WITH MYELODYSPLASTIC SYNDROMES AND CHRONIC MYELOMONOCYTIC LEUKEMIA TREATED WITH HYPOMETHYLATING AGENTS
(Abstract release date: 05/18/17)
EHA Library. Montalban-Bravo G. 06/24/17; 181776; S489
Guillermo Montalban-Bravo
Contributions
Contributions
Abstract
Abstract: S489
Type: Oral Presentation
Presentation during EHA22: On Saturday, June 24, 2017 from 16:45 - 17:00
Location: Room N105
Background
Hypomethylating agents (HMA) such as azacitidine and decitabine remain the standard of care for the treatment of myelodysplastic syndromes (MDS) however, loss of response to therapy is associated with poor outcomes. Multiple studies have tried do identify biomarkers of response but the impact of the mutational architecture present at the time of diagnosis in response outcomes is unclear.
Aims
To evaluate the impact of the mutational architecture present at the time of diagnosis in response outcomes is unclear.
Methods
We evaluated 222 previously untreated patients with MDS or CMML that received HMA therapy at The University of Texas MD Anderson Cancer Center. Next generation sequencing analyzing a panel of 28 genes was performed prior to therapy with HMA. VAF estimates were used to evaluate clonal and subclonal relationships within each individual sample with clonal heterogeneity being defined in cases with Pearson goodness-of-fit p-values <0.05. Generalized linear models were used to study association of response rates (ORR=overall and CR=complete) and risk factors. Response was defined following 2006 IWG criteria.
Results
A total of 143 patients (79%) had MDS and 43 (19%) had CMML, including 108 (49%) with lower-risk based on IPSS and 114 (51%) with higher-risk disease. Therapy consisted in azacitidine monotherapy in 60 (27%) patients, decitabine monotherapy in 57 (26%), guadecitabine in 46 (21) and combinations in 59 (27%). The ORR was 61% (135/222) with 80 (36%) patients achieving CR. A total of 161 (73%) patients had at least one detectable mutation. Median number of mutations was 1 (range 0-5). Frequencies of detected mutations are shown in Figure 1A. Among 70 (32%) patients evaluable for clonal heterogeneity testing, 38 (55%) where clonally heterogeneous and carried at least 1 subclone. Pairwise associations of mutations revealed distinct and significant co-mutation patterns (Figure 1B). Within these co-mutation associations, there were no clear hierarchical patterns of clonality in patients evaluable for clonal heterogeneity, as indicated in Figure 1B. By univariate analysis, presence of mutations in ASXL1 (OR 0.45, CI 0.22-0.93, p=0.03) and RUNX1 (0.44, CI 0.20-0.96, p=0.038) as well as that of TP53 mutations with VAF ≥0.31 (OR 0.21, CI 0.05-0.8, p=0.024) predicted for a lower likelihood of response. Analysis of functional pathways revealed that patients with mutations in chromatin (OR 0.43, CI 0.21-0.86, p=0.017) and signaling genes (OR 0.48, CI 0.23-1.00, p=0.049) had lower likelihood of achieving response. Additionally, patients with ASXL1 mutations (OR 0.24, CI 0.09-0.64, p=0.005), particularly in the absence of co-occurring TET2, as well as those with increased number of mutations, particularly if more than 3 (OR 0.21, CI 0.06-0.73, p=0.014), or signaling gene mutations (OR 0.32, CI 0.13-0.80, p=0.015), had a lower likelihood of achieving a CR. A longer time to response was observed in patients with DNMT3A mutations with VAF ≥0.35 (3.4 vs 1 months, OR 0.22, CI 0.06-0.76, p=0.017). Among patients who achieved CR, presence of 3 or more mutations (2.6 vs 1.3 months, OR 1.35, CI 1.00-1.83, p=0.049) and TP53 mutations with VAF ≥0.31 (0 vs 3.7 months, OR 2.03, CI 1.03-3.98, p=0.040) predicted for shorter CR duration. Presence of clonal heterogeneity, as well as the identified pairwise co-mutation patterns did not predict for any of the response outcomes.
Conclusion
The type, number and burden of mutations at the time of diagnosis may predict response to therapy with HMA in patients with MDS and CMML.
Session topic: 10. Myelodysplastic syndromes - Clinical
Keyword(s): Therapy, Myelodysplasia, mutation analysis
Abstract: S489
Type: Oral Presentation
Presentation during EHA22: On Saturday, June 24, 2017 from 16:45 - 17:00
Location: Room N105
Background
Hypomethylating agents (HMA) such as azacitidine and decitabine remain the standard of care for the treatment of myelodysplastic syndromes (MDS) however, loss of response to therapy is associated with poor outcomes. Multiple studies have tried do identify biomarkers of response but the impact of the mutational architecture present at the time of diagnosis in response outcomes is unclear.
Aims
To evaluate the impact of the mutational architecture present at the time of diagnosis in response outcomes is unclear.
Methods
We evaluated 222 previously untreated patients with MDS or CMML that received HMA therapy at The University of Texas MD Anderson Cancer Center. Next generation sequencing analyzing a panel of 28 genes was performed prior to therapy with HMA. VAF estimates were used to evaluate clonal and subclonal relationships within each individual sample with clonal heterogeneity being defined in cases with Pearson goodness-of-fit p-values <0.05. Generalized linear models were used to study association of response rates (ORR=overall and CR=complete) and risk factors. Response was defined following 2006 IWG criteria.
Results
A total of 143 patients (79%) had MDS and 43 (19%) had CMML, including 108 (49%) with lower-risk based on IPSS and 114 (51%) with higher-risk disease. Therapy consisted in azacitidine monotherapy in 60 (27%) patients, decitabine monotherapy in 57 (26%), guadecitabine in 46 (21) and combinations in 59 (27%). The ORR was 61% (135/222) with 80 (36%) patients achieving CR. A total of 161 (73%) patients had at least one detectable mutation. Median number of mutations was 1 (range 0-5). Frequencies of detected mutations are shown in Figure 1A. Among 70 (32%) patients evaluable for clonal heterogeneity testing, 38 (55%) where clonally heterogeneous and carried at least 1 subclone. Pairwise associations of mutations revealed distinct and significant co-mutation patterns (Figure 1B). Within these co-mutation associations, there were no clear hierarchical patterns of clonality in patients evaluable for clonal heterogeneity, as indicated in Figure 1B. By univariate analysis, presence of mutations in ASXL1 (OR 0.45, CI 0.22-0.93, p=0.03) and RUNX1 (0.44, CI 0.20-0.96, p=0.038) as well as that of TP53 mutations with VAF ≥0.31 (OR 0.21, CI 0.05-0.8, p=0.024) predicted for a lower likelihood of response. Analysis of functional pathways revealed that patients with mutations in chromatin (OR 0.43, CI 0.21-0.86, p=0.017) and signaling genes (OR 0.48, CI 0.23-1.00, p=0.049) had lower likelihood of achieving response. Additionally, patients with ASXL1 mutations (OR 0.24, CI 0.09-0.64, p=0.005), particularly in the absence of co-occurring TET2, as well as those with increased number of mutations, particularly if more than 3 (OR 0.21, CI 0.06-0.73, p=0.014), or signaling gene mutations (OR 0.32, CI 0.13-0.80, p=0.015), had a lower likelihood of achieving a CR. A longer time to response was observed in patients with DNMT3A mutations with VAF ≥0.35 (3.4 vs 1 months, OR 0.22, CI 0.06-0.76, p=0.017). Among patients who achieved CR, presence of 3 or more mutations (2.6 vs 1.3 months, OR 1.35, CI 1.00-1.83, p=0.049) and TP53 mutations with VAF ≥0.31 (0 vs 3.7 months, OR 2.03, CI 1.03-3.98, p=0.040) predicted for shorter CR duration. Presence of clonal heterogeneity, as well as the identified pairwise co-mutation patterns did not predict for any of the response outcomes.
Conclusion
The type, number and burden of mutations at the time of diagnosis may predict response to therapy with HMA in patients with MDS and CMML.
Session topic: 10. Myelodysplastic syndromes - Clinical
Keyword(s): Therapy, Myelodysplasia, mutation analysis
{{ help_message }}
{{filter}}