Hematology
Contributions
Type: Publication Only
Background
patients (pts) affected by acute myeloid leukemia (AML) resistant to a first line treatment are a relatively small group with a very poor prognosis. There are few data about clinical and molecular features predictive of primary induction failure (PIF) at baseline. Ravandi observed in a series of 1597 AML patients that higher age, unfavourable cytogenetic, high white blood cells (WBC) count at presentation were associated with PIF while they failed to identify any association with molecular abnormalities such as FLT3 mutations.
Aims
aim of our work was to evaluate clinical and molecular features of PIF AML patients at baseline in order to identify any possible clinical characteristics predictive of PIF.
Methods
Clinical and molecular data at baseline of 385 pts affected by AML fit for standard chemotherapy from 2000 to 2013 were collected in a prospective manner in our institution. The distribution of patients’ characteristics were summarized using percentiles, for continuous variables, and percentages and frequencies for categorical variables. In order to test difference in distribution of patients’ characteristics according to PIF we performed the Wilcoxon rank sum test for continuous variables and the χ2 test for categorical variables. We performed a logistic regression analysis to identify risk factors associated of failure to achieve complete remission after induction treatment..
Results
Results: 104 (27%) out of 450 pts resulted refractory to the first line treatment. Clinical characteristics at baseline were as follows: median age 58 (± 19) years, 207 out of 385 pts were male. 25,8% of patients who achieved complete remission showed organomegalies versus 28,6% of PIF pts. High WBC count at presentation was evident in 22% of PIF versus 23,9% of controls. Median platelets count at presentation was similar in the two subgroups: 60 in controls versus 48 in PIF pts. Cytogenetic risk was as follows: 9% of controls and 2,3% of PIF pts were low risk, 23,8% of controls and 39,1% of PIF patients were high risk, the remaining pts were intermediate risk
The majority of the pts (64%) were treated by standard dose anthracycline based induction regimen, while 26% were treated by fludarabine based regimen and 10% were treated by high dose chemotherapy.
At univariate analysis age, unfavourable cytogenetic risk and fludarabine based regimen were predictive factors of induction failure (OR 1,02, 2.05 and 1,77 respectively with 1,01-1,04, 1,22-3,45 and 1,09-2,87 95% CI) while NPM mutated status, high dose ARA-C regimen and favourable cytogenetic risk were associated with complete remission (OR 0,33, 0,4 and 0,23 with 0,1-1,07, 0,17-0,99 and 0,05-1,02 95% CI). We failed to identify a statistically significant association with WBC, platelets count and FLT3 status and PIF status.
Summary
In our experience the only predictive factors of PIF in adults affected by AML are higher age and unfavourable cytogenetic at baseline. Flt3 status is confirmed not to be a molecular marker of treatment failure while NPM mutation seems to have a positive predictive value but the data need to be confirmed in a larger series of cases. Even if our study is only observational, our data showed that pts treated by conditioning regimen based on high dose ARA-C showed a better outcome than pts who underwent fludarabine based induction chemotherapy. Further studies are needed in order to identify more clinical and biological factors predictive of primary treatment failure at baseline.
Keyword(s): AML
Session topic: Publication Only
Type: Publication Only
Background
patients (pts) affected by acute myeloid leukemia (AML) resistant to a first line treatment are a relatively small group with a very poor prognosis. There are few data about clinical and molecular features predictive of primary induction failure (PIF) at baseline. Ravandi observed in a series of 1597 AML patients that higher age, unfavourable cytogenetic, high white blood cells (WBC) count at presentation were associated with PIF while they failed to identify any association with molecular abnormalities such as FLT3 mutations.
Aims
aim of our work was to evaluate clinical and molecular features of PIF AML patients at baseline in order to identify any possible clinical characteristics predictive of PIF.
Methods
Clinical and molecular data at baseline of 385 pts affected by AML fit for standard chemotherapy from 2000 to 2013 were collected in a prospective manner in our institution. The distribution of patients’ characteristics were summarized using percentiles, for continuous variables, and percentages and frequencies for categorical variables. In order to test difference in distribution of patients’ characteristics according to PIF we performed the Wilcoxon rank sum test for continuous variables and the χ2 test for categorical variables. We performed a logistic regression analysis to identify risk factors associated of failure to achieve complete remission after induction treatment..
Results
Results: 104 (27%) out of 450 pts resulted refractory to the first line treatment. Clinical characteristics at baseline were as follows: median age 58 (± 19) years, 207 out of 385 pts were male. 25,8% of patients who achieved complete remission showed organomegalies versus 28,6% of PIF pts. High WBC count at presentation was evident in 22% of PIF versus 23,9% of controls. Median platelets count at presentation was similar in the two subgroups: 60 in controls versus 48 in PIF pts. Cytogenetic risk was as follows: 9% of controls and 2,3% of PIF pts were low risk, 23,8% of controls and 39,1% of PIF patients were high risk, the remaining pts were intermediate risk
The majority of the pts (64%) were treated by standard dose anthracycline based induction regimen, while 26% were treated by fludarabine based regimen and 10% were treated by high dose chemotherapy.
At univariate analysis age, unfavourable cytogenetic risk and fludarabine based regimen were predictive factors of induction failure (OR 1,02, 2.05 and 1,77 respectively with 1,01-1,04, 1,22-3,45 and 1,09-2,87 95% CI) while NPM mutated status, high dose ARA-C regimen and favourable cytogenetic risk were associated with complete remission (OR 0,33, 0,4 and 0,23 with 0,1-1,07, 0,17-0,99 and 0,05-1,02 95% CI). We failed to identify a statistically significant association with WBC, platelets count and FLT3 status and PIF status.
Summary
In our experience the only predictive factors of PIF in adults affected by AML are higher age and unfavourable cytogenetic at baseline. Flt3 status is confirmed not to be a molecular marker of treatment failure while NPM mutation seems to have a positive predictive value but the data need to be confirmed in a larger series of cases. Even if our study is only observational, our data showed that pts treated by conditioning regimen based on high dose ARA-C showed a better outcome than pts who underwent fludarabine based induction chemotherapy. Further studies are needed in order to identify more clinical and biological factors predictive of primary treatment failure at baseline.
Keyword(s): AML
Session topic: Publication Only