The LEukemia Artificial Intelligence Program (LEAP) in chronic myeloid leukemia in chronic phase: A model to improve patient outcomes.

Extreme gradient boosting methods outperform conventional machine-learning models. Here, we have developed the LEukemia Artificial intelligence Program (LEAP) with the extreme gradient boosting decision tree method for the optimal treatment recommendation of tyrosine kinase inhibitors (TKIs) in patients with chronic myeloid leukemia in chronic phase (CML-CP). A cohort of CML-CP patients was randomly divided into training/validation (N = 504) and test cohorts (N = 126). The training/validation cohort was used for 3-fold cross validation to develop the LEAP CML-CP model using 101 variables at diagnosis. The test cohort was then applied to the LEAP CML-CP model and an optimum TKI treatment was suggested for each patient. The area under the curve in the test cohort was 0.81899.Backward multivariate analysis identified age at diagnosis, the degree of comorbidities, and TKI recommended therapy by the LEAP CML-CP model as independent prognostic factors for overall survival. The bootstrapping method internally validated the association of the LEAP CML-CP recommendation with overall survival as an independent prognostic for overall survival. Selecting treatment according to the LEAP CML-CP personalized recommendations, in this model, is associated with better survival probability compared to treatment with a LEAP CML-CP non-recommended therapy. This approach may pave a way of new era of personalized treatment recommendations for patients with cancer.

Molecular responses in patients with chronic myelogenous leukemia in chronic phase treated with imatinib mesylate.

PURPOSE: To determine the clinical significance of molecular response and relapse among patients with chronic myelogenous leukemia (CML) treated with imatinib. EXPERIMENTAL DESIGN: We analyzed the results of quantitative PCR in 280 patients with CML in chronic phase who achieved complete cytogenetic remission with imatinib (117 after IFN-alpha failure and 163 previously untreated). Median follow-up was 31 months (range, 3-52 months). RESULTS: Median BCR-ABL/ABL ratio before the start of therapy was 39.44 (range, 0.252-170.53). A major molecular response (BCR-ABL/ABL ratio <0.05%) was achieved in 174 (62%), and transcripts became undetectable (complete molecular response) in 95 (34%). By multivariate analysis, only treatment with high-dose imatinib (P = 0.02) was associated with achievement of a major molecular response. Nine of 166 (5%) patients who achieved a major molecular response lost their cytogenetic remission, compared with 25 of 68 (37%) among those who did not achieve this response (P < 0.0001). Patients achieving a major molecular response 12 months after the start of therapy had significantly better complete cytogenetic remission duration than others. A >1-log reduction in transcript levels after 3 months of therapy predicted for an improved probability of achieving a major molecular response at 24 months. Increasing levels of BCR-ABL transcripts predicted for a loss of cytogenetic remission only among patients who did not achieve a major molecular response. CONCLUSIONS: Achieving a major molecular response, particularly within the first year of therapy, is predictive of a durable cytogenetic remission and may be the future goal of therapy in CML.

Staging of chronic myeloid leukemia in the imatinib era: an evaluation of the World Health Organization proposal.

BACKGROUND: Several staging classification systems, all of which were designed in the preimatinib era, are used for chronic myeloid leukemia (CML). The World Health Organization (WHO) recently proposed a new classification system that has not been validated clinically. The authors investigated the significance of the WHO classification system and compared it with the classification systems used to date in imatinib trials ("standard definition") to determine its impact in establishing the outcome of patients after therapy with imatinib. METHODS: In total, 809 patients who received imatinib for CML were classified into chronic phase (CP), accelerated phase (AP), and blast phase (BP) based on standard definitions and then were reclassified according to the new WHO classification system. Their outcomes with imatinib therapy were compared, and the value of individual components of these classification systems was determined. RESULTS: With the WHO classification, 78 patients (10%) were reclassified: 45 patients (6%) were reclassified from CP to AP, 14 patients (2%) were reclassified from AP to CP, and 19 patients (2%) were reclassified from AP to BP. The rates of complete cytogenetic response for patients in CP, AP, and BP according to the standard definition were 72%, 45%, and 8%, respectively. After these patients were reclassified according to WHO criteria, the response rates were 77% (P = 0.07), 39% (P = 0.28), and 11% (P = 0.61), respectively. The 3-year survival rates were 91%, 65%, and 10%, respectively, according to the standard classification and 95% (P = 0.05), 63% (P = 0.76), and 16% (P = 0.18), respectively, according to the WHO classification. Patients who had a blast percentage of 20-29%, which is considered CML-BP according to the WHO classification, had a significantly better response rate (21% vs. 8%; P = 0.11) and 3-year survival rate (42% vs. 10%; P = 0.0001) compared with patients who had blasts > or = 30%. CONCLUSIONS: Different classification systems had an impact on the outcome of patients, and some prognostic features had different prognostic implications in the imatinib era. The authors believe that a new, uniform staging system for CML is warranted, and they propose such a system.

Granulocyte-colony-stimulating factor (filgrastim) may overcome imatinib-induced neutropenia in patients with chronic-phase chronic myelogenous leukemia.

BACKGROUND: Imatinib mesylate administration has become standard treatment for patients with chronic myelogenous leukemia (CML). Although the safety profile of imatinib is favorable, Grade > or = 3 neutropenia (according to the National Cancer Institute Common Toxicity Criteria) occurs in 35-45% of patients with CML in chronic phase who receive standard-dose imatinib. Myelosuppression results in treatment interruptions, which may compromise responses to imatinib. The authors investigated the ability of granulocyte-colony-stimulating factor (filgrastim) to reverse imatinib-associated neutropenia, thereby allowing for more continuous imatinib administration. METHODS: Thirteen patients with chronic-phase CML and Grade > or = 3, imatinib-induced neutropenia were treated with filgrastim. Treatment with filgrastim was initiated after a median of 22 months from the start of imatinib. Eleven patients received filgrastim 5 microg/kg 1-3 times weekly, and 2 patients received filgrastim 5 microg/kg daily; doses were titrated to maintain an absolute neutrophil count (ANC) > or = 10(9)/L. RESULTS: Seven of 11 patients (64%) who began treatment with an ANC < 1.5 x 10(9)/L had responses (i.e., their ANC improved to > or = 2 x 10(9)/L within 21 days); the other 4 patients experienced slower recovery but were able to continue receiving imatinib uninterrupted. Before filgrastim administration was initiated, patients did not receive imatinib (due to neutropenia-related treatment interruptions) for an average of 21% of the total time since the start of imatinib. This figure decreased to 6% after the start of filgrastim treatment (P = 0.0008). Before filgrastim treatment was initiated, only one patient had achieved a major (partial) cytogenetic response. After the start of filgrastim treatment, five patients had major cytogenetic responses (including two complete responses). CONCLUSIONS: The authors concluded that filgrastim may overcome imatinib-associated neutropenia and allow improved delivery of imatinib. Some patients may experience improvements in their responses to therapy as a result.

Efficacy of imatinib mesylate in the treatment of idiopathic hypereosinophilic syndrome.

Idiopathic hypereosinophilic syndrome (HES) is a myeloproliferative disorder characterized by persistent eosinophilia and organ involvement. Different treatments have been investigated in HES with modest success. It has been suggested that imatinib is active in HES. We treated 9 patients with HES with 100 mg imatinib daily. Doses for patients without response after 4 weeks were increased to 400 mg daily. Prior therapy had failed for 7 patients. Five patients responded: 4 achieved sustained complete remission lasting a median of 12+ weeks (range, 9+ to 36+ weeks), and 1 had a transient response. One patient died in complete remission. Responses occurred within 4 weeks of therapy; only 1 responder required an increase in dose to 400 mg daily. Three of 4 nonresponders failed to respond to an increase in dose. Toxicity was minimal. We conclude that imatinib therapy is effective for HES.

Results of imatinib mesylate therapy in chronic myelogenous leukaemia with variant Philadelphia chromosome.

Five to 10 per cent of patients with Philadelphia chromosome (Ph)-positive chronic myelogenous leukaemia (CML) have variant translocations involving chromosomes other than 9 and 22. We investigated the characteristics and outcome of patients with variant translocations treated with imatinib. Among 721 patients, 44 (6%) had variant translocations, involving one (n = 39) or two (n = 4) additional chromosomes. Nineteen patients (44%) were in chronic (12 previously untreated), 24 (55%) in accelerated and one (2%) in blastic phase. A major cytogenetic response was achieved in 14 (74%) patients treated in chronic phase and in 14 (58%) treated in accelerated phase. Six of 13 (46%) evaluable patients had deletion of derivative chromosome 9, and there was a trend for a lower response rate in these patients. We compared the 43 patients in chronic or accelerated phase to 678 patients with classic Ph treated with imatinib. The only significant difference in clinical characteristics was a higher frequency of accelerated phase among those with variant translocations (56%) compared with those with classic translocations (38%). No differences in outcome were evident. In a multivariate analysis, variant Ph translocations had no impact in response rate, overall survival or duration of response. We conclude that patients with variant Ph translocations have a similar prognosis to those with classic Ph translocations when treated with imatinib.