Abstract
Abstract 71
Background:
The NCCN and the European LeukemiaNet guidelines for monitoring patients with chronic myeloid leukemia in chronic phase (CML-CP) provide recommendations for response assessment and treatment at 3, 6, 12, and 18 months based on evidence obtained in clinical trials. A clear limitation of such guidelines is their applicability at time-points different from those pre-specified. To overcome these limitations we have developed a novel statistical approach to CML prognostication.
Method:
In order to build our prognostic model, we used two cohorts of patients with CML-CP treated in the frontline DASISION phase III study (CA180-056) and the cohort of patients treated after imatinib failure in the dasatinib dose-optimization phase III study (CA180-034). Progression-free survival (PFS) was defined as any of the following: doubling of white cell count to >20×109/L in the absence of complete hematologic response (CHR); loss of CHR; increase in Ph+ BM metaphases to >35%; transformation to AP/BP; or death. A modified Cox proportional hazards model was used to build a prognostic nomogram.
Results:
A total of 1189 patients were used for this analysis: 519 from DASISION (259 dasatinib and 260 imatinib) and 670 from CA180-034. First, we devised a model to link a BCR-ABL1/ABL1 ratios (according to the International Standard) obtained at specific time points during the course of treatment with patientsÕ outcomes (PFS). For instance, at 18 months after front-line treatment, the future PFS probabilities are shown in Figure 1A. At 6 months after second-line treatment, the future PFS probabilities are shown in Figure 1B. Once the model was validated at specific time points, we next designed a nomogram to calculate patients' outcomes at any time point during the course of therapy by plotting ‘master PFS curves’ derived from the patient cohorts according to time. Figure 2A&B give the 90% quantile of the remaining PFS for patients at any time after front-line and second-line treatment, respectively. These may be used a guideline for considering other treatment options when patients' BCR-ABL1/ABL1 ratios exceed these values. Figure 2 shows that the remaining PFS times for either front- or second-line treated patients depend mostly on the current BCR-ABL/ABL ratio and less on the time at which the ratio is obtained, reflected by the fact that the curves showing future PFS probabilities are characterized by smooth slopes. Figure 2A shows that 10% of front-line treated patients whose BCR-ABL1/ABL1 ratios are 50% or higher will have remaining PFS times of less than 12 months. If BCR-ABL1/ABL1 ratios are 75% or higher, then 10% of them will have remaining PFS times of less than 6 months. Similarly, Figure 2B shows that for second-line treated patients whose BCR-ABL1/ABL1 ratios are 50% or higher, 10% of them will have remaining PFS time shorter than 6 months.
Conclusion:
We have designed a nomogram that predicts PFS for patients treated in the frontline and second line settings according to their BCR-ABL1/ABL1 ratios, independent from the time at which these ratios are obtained. A similar approach has been taken to predict failure-free and overall survival and will be presented at the meeting. This prognostic tool is readily available for clinical purposes and might greatly facilitate monitoring and prognostication in CML.
Disclosures:
No relevant conflicts of interest to declare.
Innovative Evaluation Techniques for Hanok Sustainability Based on Integrative Comfort Indices
