RESUMO
For most patients with childhood myelodysplastic syndrome (cMDS), allogeneic hematopoietic stem cell transplantation (allo-HSCT) remains the only curative option. In the case of increased blasts (cMDS-IB), the benefit of pretransplant cytoreductive therapy remains controversial. In this multicenter retrospective study, the outcomes of all French children who underwent allo-HSCT for cMDS reported in the SFGM-TC registry between 2000 and 2020 were analyzed (n = 84). The median age at transplantation was 10.2 years. HSCT was performed from matched sibling donors (MSD) in 29% of the cases, matched unrelated donors (MUD) in 44%, haploidentical in 6%, and cord blood in 21%. Myeloablative conditioning was used in 91% of cases. Forty-eight percent of patients presented with cMDS-IB at diagnosis (median BM blasts: 8%). Among them, 50% received pretransplant cytoreductive therapy. Five-year overall survival (OS), cumulative incidence of nonrelapse mortality (NRM), and relapse were 67%, 26%, and 12%, respectively. Six-month cumulative incidence of grade II-IV acute graft-versus-host disease was 46%. Considering the whole cohort, age under 12, busulfan/cyclophosphamide/melphalan conditioning or MUD were associated with poorer 5-year OS. In the cMDS-IB subgroup, pretransplant cytoreductive therapy was associated with a better OS in univariate analysis. This seems to be mainly due to a decreased NRM since no impact on the incidence of relapse was observed. Overall, those data may argue in favor of cytoreduction for cMDS-IB. They need to be confirmed on a larger scale and prospectively.
RESUMO
BACKGROUND: Acute leukaemias are life-threatening haematological cancers characterised by the infiltration of transformed immature haematopoietic cells in the blood and bone marrow. Prompt and accurate diagnosis of the three main acute leukaemia subtypes (ie acute lymphocytic leukaemia [ALL], acute myeloid leukaemia [AML], and acute promyelocytic leukaemia [APL]) is of utmost importance to guide initial treatment and prevent early mortality but requires cytological expertise that is not always available. We aimed to benchmark different machine-learning strategies using a custom variable selection algorithm to propose an extreme gradient boosting model to predict leukaemia subtypes on the basis of routine laboratory parameters. METHODS: This multicentre model development and validation study was conducted with data from six independent French university hospital databases. Patients aged 18 years or older diagnosed with AML, APL, or ALL in any one of these six hospital databases between March 1, 2012, and Dec 31, 2021, were recruited. 22 routine parameters were collected at the time of initial disease evaluation; variables with more than 25% of missing values in two datasets were not used for model training, leading to the final inclusion of 19 parameters. The performances of the final model were evaluated on internal testing and external validation sets with area under the receiver operating characteristic curves (AUCs), and clinically relevant cutoffs were chosen to guide clinical decision making. The final tool, Artificial Intelligence Prediction of Acute Leukemia (AI-PAL), was developed from this model. FINDINGS: 1410 patients diagnosed with AML, APL, or ALL were included. Data quality control showed few missing values for each cohort, with the exception of uric acid and lactate dehydrogenase for the cohort from Hôpital Cochin. 679 patients from Hôpital Lyon Sud and Centre Hospitalier Universitaire de Clermont-Ferrand were split into the training (n=477) and internal testing (n=202) sets. 731 patients from the four other cohorts were used for external validation. Overall AUCs across all validation cohorts were 0·97 (95% CI 0·95-0·99) for APL, 0·90 (0·83-0·97) for ALL, and 0·89 (0·82-0·95) for AML. Cutoffs were then established on the overall cohort of 1410 patients to guide clinical decisions. Confident cutoffs showed two (0·14%) wrong predictions for ALL, four (0·28%) wrong predictions for APL, and three (0·21%) wrong predictions for AML. Use of the overall cutoff greatly reduced the number of missing predictions; diagnosis was proposed for 1375 (97·5%) of 1410 patients for each category, with only a slight increase in wrong predictions. The final model evaluation across both the internal testing and external validation sets showed accuracy of 99·5% for ALL diagnosis, 98·8% for AML diagnosis, and 99·7% for APL diagnosis in the confident model and accuracy of 87·9% for ALL diagnosis, 86·3% for AML diagnosis, and 96·1% for APL diagnosis in the overall model. INTERPRETATION: AI-PAL allowed for accurate diagnosis of the three main acute leukaemia subtypes. Based on ten simple laboratory parameters, its broad availability could help guide initial therapies in a context where cytological expertise is lacking, such as in low-income countries. FUNDING: None.
Assuntos
Leucemia Mieloide Aguda , Aprendizado de Máquina , Humanos , França , Leucemia Mieloide Aguda/diagnóstico , Feminino , Masculino , Pessoa de Meia-Idade , Adulto , Idoso , Leucemia-Linfoma Linfoblástico de Células Precursoras/diagnóstico , Leucemia Promielocítica Aguda/diagnóstico , AlgoritmosRESUMO
Background: Thanks to an improved therapeutic regimen in childhood B-cell precursor acute lymphoblastic leukemia (BCP-ALL), 5 year-overall survival now exceeds 90%. Unfortunately, the 25% of children who relapse have an initial poor prognosis, potentially driven by pre-existing or emerging molecular anomalies. The latter are initially and essentially identified by cytogenetics. However, some subtle alterations are not visible through karyotyping. Methods: Single nucleotide polymorphisms (SNP) array is an alternative way of chromosomal analysis allowing for a more in-depth evaluation of chromosomal modifications such as the assessment of copy number alterations (CNA) and loss of heterozygosity (LOH). This method was applied here in retrospective diagnosis/relapse paired samples from seven children with BCP-ALL and in a prospective cohort of 38 newly diagnosed childhood cases. Results: In the matched study, compared to the initial karyotype, SNP array analysis reclassified two patients as poor prognosis cases. Modulation during relapse was seen for 4 CNA and 0.9 LOH. In the prospective study, SNP reclassified the 10 patients with intermediate karyotype as 7 good prognosis and 3 poor prognosis. Ultimately, in all the children tested, SNP array allowed to identify additional anomalies compared to conventional karyotype, refine its prognostic value and identify some druggable anomalies that could be used for precision medicine. Overall, the anomalies detected could be segregated in four groups respectively involved in B-cell development, cell proliferation, transcription and molecular pathways. Conclusion: SNP therefore appears to be a method of choice in the integrated diagnosis of BCP ALL, especially for patients initially classified as intermediate prognosis. This complementary method of both cytogenetics and high throughput sequencing allows to obtain further classified information and can be useful in case of failure of these techniques.