AML, ALL, and CML classification and diagnosis based on bone marrow cell morphology combined with convolutional neural network: A STARD compliant diagnosis research.

Leukemia diagnosis based on bone marrow cell morphology primarily relies on the manual microscopy of bone marrow smears. However, this method is greatly affected by subjective factors and tends to lead to misdiagnosis. This study proposes using bone marrow cell microscopy images and employs convolutional neural network (CNN) combined with transfer learning to establish an objective, rapid, and accurate method for classification and diagnosis of LKA (AML, ALL, and CML). We collected cell microscopy images of 104 bone marrow smears (including 18 healthy subjects, 53 AML patients, 23 ALL patients, and 18 CML patients). The perfect reflection algorithm and a self-adaptive filter algorithm were first used for preprocessing of bone marrow cell images collected from experiments. Subsequently, 3 CNN frameworks (Inception-V3, ResNet50, and DenseNet121) were used to construct classification models for the raw dataset and preprocessed dataset. Transfer learning was used to improve the prediction accuracy of the model. Results showed that the DenseNet121 model based on the preprocessed dataset provided the best classification results, with a prediction accuracy of 74.8%. The prediction accuracy of the DenseNet121 model that was obtained by transfer learning optimization was 95.3%, which was increased by 20.5%. In this model, the prediction accuracies of the normal groups, AML, ALL, and CML were 90%, 99%, 97%, and 95%, respectively. The results showed that the leukemic cell morphology classification and diagnosis based on CNN combined with transfer learning is feasible. Compared with conventional manual microscopy, this method is more rapid, accurate, and objective.

BCR-ABL Gene Transcript Types of Patients with Chronic Myelogenous Leukemia in Yogyakarta, Indonesia.

The aim of this study was analyzing the BCR-ABL transcript types of patients with chronic myeloid leukemia (CML) in Dr Sardjito General Hospital, Yogyakarta, Indonesia. This study is very relevant because the data concerning BCR-ABL gene transcript types is very limited in Indonesia. Furthermore, it is important for patient's management, particularly in defining the tyrosine kinase inhibitors (TKIs) therapy and monitoring after therapy. The introduction of TKIs has become a major advance in the management of patients with CML, especially in the chronic phase (CML-CP), in which most patients are diagnosed. METHODS: One hundred eighty five (185) of 370 recruited patients were included in this study (2010-2014). RNA samples were isolated from mononuclear cells of peripheral blood of the subjects taken at primary diagnosis. Detection of BCR-ABL gene transcript types was done using multiplex reverse transcriptase PCR (multiplex RT-PCR) and/or nested PCR following the cDNA synthesis. When the first PCR set failed to amplify the BCR-ABL gene, RT-conventional PCR and/or nested PCR would be applied. The proportion of each transcript type was calculated among the BCR-ABL positive CML patients. RESULTS: Approximately 99% (183/185) of CML patients are BCR-ABL positive, with the most common type is major b3a2 (136/183; 74.3%), followed by major b2a2 (41/183; 22.4%). Two samples (1.1%) showed co-expression of b3a2 and b2a2; 1 sample showed co-expression of b3a2 and fragment at 500bp; and 3 samples showed uncommon fragments. CONCLUSION: Ninety nine percent (99%) of CML patients in Yogyakarta, Indonesia are BCR-ABL positive, with 74.3% have b3a2 transcript, 22.4% have b2a2 trascript, 1.1% have co-expression of b3a2 and b2a2 transcript, and the rest (2.2%) have uncommon bands that still need to be confirmed.

Concordance among hematopathologists in classifying blasts plus promonocytes: A bone marrow pathology group study.

Enumeration of blasts and promonocytes is essential for World Health Organization (WHO) classification of myelomonocytic neoplasms. The accuracy of distinguishing blasts, promonocytes and monocytes, including normal vs abnormal monocytes, remains controversial. The objective of this analysis is to assess concordances between experienced hematopathologists in classifying cells as blasts, promonocytes, and monocytes according to WHO criteria. Each of 11 hematopathologists assessed glass slides from 20 patients [12 with chronic myelomonocytic leukemia (CMML) and 8 with acute myeloid leukemia (AML)] including blood and BM aspirate smears, and limited nonspecific esterase (NSE) stains. All cases were blindly reviewed. Fleiss' extension of Cohen's kappa for multiple raters was used on these variables, separately for peripheral blood (PB) and bone marrow (BM). Spearman's rank correlation was used to assess correlations between each pair of hematopathologists for each measurement. For the classification based on the sum of blasts and promonocytes in the BM, Fleiss' kappa was estimated as 0.744. For PB, categorizing patients according to the sum of blasts and promonocytes, Fleiss' kappa was estimated as 0.949. Distinction of abnormal monocytes from normal monocytes in PB did not achieve a good concordance and showed strong evidence of differences between hematopathologists (P < .0001). The hematopathologists achieved a good concordance rate of 74% in CMML vs AML classification and a high k rate, confirming that criteria for defining the blasts equivalents (blasts plus promonocytes) could be applied consistently. Identification of monocyte subtypes (abnormal vs normal) was not concordant. Our results support the practice of combining blasts/promonocytes into a single category.

Towards rapid prediction of drug-resistant cancer cell phenotypes: single cell mass spectrometry combined with machine learning.

Combined single cell mass spectrometry and machine learning methods is demonstrated for the first time to achieve rapid and reliable prediction of the phenotype of unknown single cells based on their metabolomic profiles, with experimental validation. This approach can be potentially applied towards prediction of drug-resistant phenotypes prior to chemotherapy.

Validation of the 2017 revision of the WHO chronic myelomonocytic leukemia categories.

The 2017 revision of the World Health Organization (WHO) classification includes substantial changes to the subclassification of chronic myelomonocytic leukemia (CMML): (1) a 3-tiered blast-based scheme including a novel "CMML-0" category replacing a 2-tiered system in place since 2001 and (2) 2 CMML subtypes, myelodysplastic (MDS-CMML) and myeloproliferative (MP-CMML), based on a white blood cell count cutoff of 13 × 109/L. The clinical utility of this subclassification scheme, particularly the expansion of blast-based subgroups, has not been validated. In this study, a large single-institution CMML patient cohort (n = 629) was used to assess the prognostic impact of the newly proposed categories. Patients were risk stratified according to the CMML-specific Prognostic Scoring System (CPSS) and the MD Anderson Prognostic Scoring System. MP-CMML patients had significantly shorter overall survival (OS; P < .0001; hazard ratio: 0.53, 95% confidence interval: 0.42-0.65) and median duration to acute myeloid leukemia (AML) transformation (P < .0001; 15.2 vs 22.0 months) compared with MDS-CMML patients. The CMML-0 group included 36.4% patients with higher risk CPSS categories and 11.2% of patients with high-risk cytogenetics. Among treatment-naïve patients (n = 499), there was a marginal difference in OS between the CMML-0 and CMML-12017 subgroups (P = .0552). The WHO 2017 blast-based categories were not associated with AML-free survival. Incorporation of the WHO 2017 blast-based subgroups in a modified CPSS scheme had a neutral effect and did not improve its prognostic strength. Our data support the inclusion of MP-CMML and MDS-CMML subtypes in the WHO 2017 revision. Although of some utility in MP-CMML, the 3-tiered blast-based system is not well supported in this study.

The impact of BCR-ABL1 transcript type on tyrosine kinase inhibitor responses and outcomes in patients with chronic myeloid leukemia.

Although the majority of patients with chronic myeloid leukemia do well with treatment with tyrosine kinase inhibitors (TKIs), some patients still have inferior outcomes. There are many factors that might play a part, including the different BCR-ABL1 transcript types at baseline. The current study was performed to determine the possible impact of different transcripts on the treatment responses and outcomes of patients with chronic myeloid leukemia who are receiving TKI therapy. The authors performed a systematic literature search by using the terms "b2a2/b3a2," "e13a2/e14a2," or "transcript type." e14a2 was the more common transcript type. The majority of the studies demonstrated no significant difference regarding age, sex, leukocyte counts, and hemoglobin levels between patients with the e13a2 and e14a2 transcripts. However, in approximately one-half of the studies, the e14a2 transcript was associated with higher platelet counts. Almost no studies demonstrated a significant association between disease risk scores and transcript types. In the majority of studies, having the e14a2 transcript was associated with earlier, deeper, and higher molecular response rates. Although better event-free survival was observed in patients with the e14a2 transcript in some of the studies, the majority demonstrated that transcript type did not have an impact on progression-free and overall survival. Treatment-free remission currently is a topic of much interest, and to the authors' knowledge there are limited data with conflicting results regarding the possible effects of transcript types on the outcomes of patients after discontinuation of TKIs. Because having the e14a2 transcript appears to be related to a favorable outcome, choosing second-generation TKIs for frontline therapy might be a convenient approach in patients with chronic myeloid leukemia with the e13a2 transcript. The authors believe this finding warrants further investigation.

Differentiation status of primary chronic myeloid leukemia cells affects sensitivity to BCR-ABL1 inhibitors.

Tyrosine kinase inhibitors (TKI) are the mainstay treatment of BCR-ABL1-positive leukemia and virtually all patients with chronic myeloid leukemia in chronic phase (CP CML) respond to TKI therapy. However, there is limited information on the cellular mechanisms of response and particularly on the effect of cell differentiation state to TKI sensitivity in vivo and ex vivo/in vitro. We used multiple, independent high-throughput drug sensitivity and resistance testing platforms that collectively evaluated 295 oncology compounds to characterize ex vivo drug response profiles of primary cells freshly collected from newly-diagnosed patients with BCR-ABL1-positive leukemia (n = 40) and healthy controls (n = 12). In contrast to the highly TKI-sensitive cells from blast phase CML and Philadelphia chromosome-positive acute lymphoblastic leukemia, primary CP CML cells were insensitive to TKI therapy ex vivo. Despite maintaining potent BCR-ABL1 inhibitory activity, ex vivo viability of cells was unaffected by TKIs. These findings were validated in two independent patient cohorts and analysis platforms. All CP CML patients under study responded to TKI therapy in vivo. When CP CML cells were sorted based on CD34 expression, the CD34-positive progenitor cells showed good sensitivity to TKIs, whereas the more mature CD34-negative cells were markedly less sensitive. Thus in CP CML, TKIs predominantly target the progenitor cell population while the differentiated leukemic cells (mostly cells from granulocytic series) are insensitive to BCR-ABL1 inhibition. These findings have implications for drug discovery in CP CML and indicate a fundamental biological difference between CP CML and advanced forms of BCR-ABL1-positive leukemia.

Combined Population Dynamics and Entropy Modelling Supports Patient Stratification in Chronic Myeloid Leukemia.

Modelling the parameters of multistep carcinogenesis is key for a better understanding of cancer progression, biomarker identification and the design of individualized therapies. Using chronic myeloid leukemia (CML) as a paradigm for hierarchical disease evolution we show that combined population dynamic modelling and CML patient biopsy genomic analysis enables patient stratification at unprecedented resolution. Linking CD34(+) similarity as a disease progression marker to patient-derived gene expression entropy separated established CML progression stages and uncovered additional heterogeneity within disease stages. Importantly, our patient data informed model enables quantitative approximation of individual patients' disease history within chronic phase (CP) and significantly separates "early" from "late" CP. Our findings provide a novel rationale for personalized and genome-informed disease progression risk assessment that is independent and complementary to conventional measures of CML disease burden and prognosis.

Multiplex RT-PCR Assay for Detection of Common Fusion Transcripts in Acute Lymphoblastic Leukemia and Chronic Myeloid Leukemia Cases.

BACKGROUND: Acute lymphoblastic leukemia (ALL) is a heterogeneous disease which requires a risk-stratified approach for appropriate treatment. Specific chromosomal translocations within leukemic blasts are important prognostic factors that allow identification of relevant subgroups. In this study, we developed a multiplex RT-PCR assay for detection of the 4 most frequent translocations in ALL (BCR-ABL, TEL-AML1, MLL-AF4, and E2A- PBX1). MATERIALS AND METHODS: A total of 214 diagnosed ALL samples from both adult and pediatric ALL and 14 cases of CML patients (154 bone marrow and 74 peripheral blood samples) were assessed for specific chromosomal translocations by cytogenetic and multiplex RT-PCR assays. RESULTS: The results showed that 46 cases of ALL and CML (20.2%) contained the fusion transcripts. Within the positive ALL patients, the most prevalent cryptic translocation observed was mBCR-ABL (p190) at 8.41%. In addition, other genetic rearrangements detected by the multiplex PCR were 4.21% TEL-AML1 and 2.34% E2A-PBX1, whereas MLL-AF4 exhibited negative results in all tested samples. Moreover, MBCR-ABL was detected in all 14 CML samples. In 16 samples of normal karyotype ALL (n=9), ALL with no cytogentic result (n=4) and CML with no Philadelphia chromosome (n=3), fusion transcripts were detected. CONCLUSIONS: Multiplex RT-PCR provides a rapid, simple and highly sensitive method to detect fusion transcripts for prognostic and risk stratification of ALL and CML patients.

Refined medullary blast and white blood cell count based classification of chronic myelomonocytic leukemias.

Since 2001, chronic myelomonocytic leukemia (CMML) is classified by the WHO as myeloproliferative/myelodysplastic neoplasm. Herein we tried to better describe CMML patients with regard to hematological characteristics and prognosis using data of the Duesseldorf registry. We created 6 CMML subgroups, by dividing dysplastic and proliferative CMML at the cut-off of white blood cell count of 13,000/µL and splitting these two groups into 3 subgroups: CMML 0 with <5% blasts (n=101), CMML I with 5-9% blasts (n=204) and CMML II with 10-19% blasts (n=81). For comparison we included patients with RCMD, RAEB I and II. The newly created CMML 0 group had better prognosis than CMML I and II, median survival times were 31 months (ms), 19ms and 13ms, respectively (p<0.001). Median survival times between the corresponding dysplastic and proliferative subgroups 0 and 1 differed significantly: CMML 0 dysplastic 48ms and CMML 0 proliferative 17ms (p=0.03), CMML I dysplastic 29ms and CMML I proliferative 15ms (p=0.008), CMML II dysplastic 17ms and CMML II proliferative 10ms (p=0.09). Outcome of CMML patients worsens with increasing medullary blasts and when presenting as proliferative type. Therefore it is justified to separate CMML with <5% medullary blasts.

[A clinical and laboratory study of chronic myeloid leukemia with atypical BCR-ABL fusion gene subtypes].

OBJECTIVE: To explore the clinical and laboratory features of chronic myeloid leukemia (CML) with atypical e14a3 and e19a2 BCR-ABL fusion gene subtypes. METHODS: We retrospectively analyzed a cohort of CML patients with Ph chromosome positive confirmed by cytogenetic and FISH but classical e13a3(b2a2), e14a2(b3a2)and e1a2 fusion transcripts negative identified by conventional real-time quantification RT-PCR (RQ-PCR). Further RQ-PCR was done with the forward primer and reverse primer designed to detect rare atypical BCR-ABL fusion genes including e14a3 and e19a2 transcripts. Direct sequencing analysis was performed on the PCR products and mutations in the BCR-ABL kinase domain were detected. The clinical data of patients were retrospectively analyzed. RESULTS: Six CML patients were found to carry t(9;22) abnormality and BCR-ABL rearrangement confirmed by FISH but classical BCR-ABL fusion genes negative detected by RQ-PCR. Further RQ-PCR and sequencing analysis confirmed the fusion of BCR exon 14 and ABL exon 3 in five CML patients (case 1-5) and the fusion of BCR exon 19 and ABL exon 2 in one CML patient (case 6). E255K and I293T IM-resistant mutations were detected in case 1 and 2, respectively. Among five cases with e14a3 transcripts, four were CML-CP, one CML-AP. Four patients were male and one was female. The median age was 48 years. The patient (case 6) with e19a2 transcripts was 40-year-old female with a diagnosis of CML-CP and PLT count was more than 1 000×109/L. Imatinib (IM) therapy was administer in case 1, 2, 3, 4 and hematopoietic stem cell transplantation (HSCT) was undergone in case 5 after hydroxyurea (Hu) or interferon failure. Case 1 who had E255K IM resistant mutation, responded poorly to IM but obtained a complete cytogenetic remission (CCyR) after a substitution of dasatinib for IM. Case 2 and 3 achieved CCyR 6 months later after IM treatment and had been maintained well with IM despite I293T mutation in case 2. Case 4 attained CCyR 3 months later after IM treatment but relapsed and died soon. Case 5 was still in CCyR after HSCT. Case 6 with e19a2 transcripts got complete hematologic response after Hu treatment and CCyR was achieved soon after IM therapy. CONCLUSION: Incidence of CML with atypical transcripts is extremely low. They could benefit from tyrosine kinase inhibitors or HSCT. Rare and atypical BCR- ABL fusion gene subtypes could be missed by conventional RQ-PCR.

Chronic myelomonocytic leukemia: myelodysplastic or myeloproliferative?

Chronic myelomonocytic leukemia (CMML) is a clonal disease of the hematopoietic stem cell that provokes a stable increase in peripheral blood monocyte count. The World Health Organisation classification appropriately underlines that the disease combines dysplastic and proliferative features. The percentage of blast cells in the blood and bone marrow distinguishes CMML-1 from CMML-2. The disease is usually diagnosed after the age of 50, with a strong male predominance. Inconstant and non-specific cytogenetic aberrations have a negative prognostic impact. Recurrent gene mutations affect mainly the TET2, SRSF2, and ASXL1 genes. Median survival is 3 years, with patients dying from progression to AML (20-30%) or from cytopenias. ASXL1 is the only gene whose mutation predicts outcome and can be included within a prognostic score. Allogeneic stem cell transplantation is possibly curative but rarely feasible. Hydroxyurea, which is the conventional cytoreductive agent, is used in myeloproliferative forms, and demethylating agents could be efficient in the most aggressive forms of the disease.

The 2008 World Health Organization classification system for myeloproliferative neoplasms: order out of chaos.

The first formal classification of chronic myeloid neoplasms is credited to William Dameshek, who in 1951 described the concept of "myeloproliferative disorders (MPD)" by grouping together chronic myelogenous leukemia, polycythemia vera (PV), essential thrombocythemia (ET), and primary myelofibrosis (PMF). The 2001 World Health Organization (WHO) classification of myeloid malignancies included these MPDs under the broader category of chronic myeloproliferative diseases (CMPD), which also included chronic neutrophilic leukemia, chronic eosinophilic leukemia/hypereosinophilic syndrome (CEL/HES), and "CMPD, unclassifiable." The revised 2008 WHO classification system featured the following changes: 1) the term "CMPD" was replaced by "myeloproliferative neoplasm (MPN)," 2) mast cell disease was formally included under the category of MPN, and 3) the subcategory of CEL/HES was reorganized into "CEL not otherwise specified (CEL-NOS)" and "myeloid and lymphoid neoplasms with eosinophilia and abnormalities of PDGFRA, PDGFRB, and FGFR1"; CEL-NOS remained a subcategory of "MPN," whereas the latter neoplasms were now assigned a new category of their own. Furthermore, diagnostic criteria for PV, ET, and PMF were revised by incorporating recently described molecular markers (eg, JAK2 and MPL mutations) as well as underscoring the role of histology in differentiating reactive from clonal myeloproliferations. As a result, red cell mass measurement is no longer necessary for the diagnosis of PV, and ET can now be diagnosed at a lower platelet count threshold. The revised WHO document continues to promote the recognition of histologic categories as a necessary first step toward the genetic characterization of myeloid malignancies.

The myelodysplastic/myeloproliferative neoplasms: myeloproliferative diseases with dysplastic features.

The 2001 World Health Organization (WHO)-sponsored classification of hematopoietic tumors has, for the first time, clearly defined a group of rare myeloid neoplasms termed myelodysplastic/myeloproliferative diseases (MDS/MPDs). This group includes three main entities, chronic myelomonocytic leukemia, atypical chronic myeloid leukemia and juvenile myelomonocytic leukemia, and also several less well defined, 'unclassifiable' disorders with MDS/MPN-like features. In the upcoming fourth edition of the WHO fascicle, due out later this year, the term 'MPD' is replaced by 'myeloproliferative neoplasm (MPN)'. Accordingly, the term MDS/MPD is being replaced by 'MDS/MPN' that will be used in this review. Although much progress has been made in understanding the molecular pathogenesis of myeloid neoplasms, most of the diseases included in the group of MDS/MPN still remain 'clinicopathologically assigned'. In other words, they can only be accurately categorized by a careful multiparametric approach that is based on the integration of bone marrow and peripheral blood morphology with other laboratory and clinical findings. The current 'spotlight' review provides practical guidelines, which should allow for a reproducible classification of these uncommon neoplasms when encountered in clinical practice.

Leucemia Mielóide Crônica: história natural e classificação / Chronic Myeloide Leukemia: natural history and classification

A Leucemia Mielóide Crônica (LMC), cuja incidência é de um a dois casos para cada 100 mil habitantes por ano, corresponde de 15 por cento a 20 por cento das leucemias. É uma doença mieloproliferativa crônica clonal, caracterizada por leucocitose com desvio à esquerda, esplenomegalia e pela presença do cromossomo Philadelphia (Ph), que resulta da translocação recíproca e equilibrada entre os braços longos dos cromossomos 9q34 e 22q11, gerando a proteína híbrida BCR-ABL, com atividade aumentada de tirosino quinase. A proteína BCR-ABL está presente em todos os pacientes com LMC, e sua hiperatividade desencadeia liberação de efetores da proliferação celular e inibidores da apoptose, sendo sua atividade responsável pela oncogênese inicial da LMC. A doença evolui em três fases: crônica, acelerada e aguda. Na fase crônica (FC) ocorre proliferação clonal maciça das células granulocíticas, mantendo estas a capacidade de diferenciação. Posteriormente, num período de tempo variável, o clone leucêmico perde a capacidade de diferenciação e a doença passa a ser de difícil controle (fase acelerada - FA) e progride para uma leucemia aguda (crise blástica - CB). Nesse artigo discutimos a história natural e as definições das fases da doença, de acordo com os critérios mais utilizados.

Chronic myeloid leukemia (CML) is estimated at approximately 1 to 2 cases per 100,000 individuals and accounts for approximately 15 percent to 20 percent of all patients with leukemia. CML is a clonal disease characterized by balanced translocation between chromosomes 9 and 22 (Philadelphia chromosome). The resulting BCR-ABL gene has abnormal tyrosine kinase activity which stimulates cell growth and is responsible for the transformed phenotype of CML cells. The disease is characterized by a triphasic course that includes a chronic phase (CP), an accelerated phase (AP) and an acute or blastic phase (BP). Unless the disease is controlled or eliminated, patients progress to AP and BF in variable periods of time. Several staging classification systems are used for CML all of which were designed in the pre-imatinib era. In this article we discuss the natural history of CML and phase definitions according to the most useful criteria.