Prediction of clinical outcome in CLL based on recurrent gene mutations, CLL-IPI variables, and (para)clinical data

Genes selection using deep learning and explainable artificial intelligence for chronic lymphocytic leukemia predicting the need and time to therapy

Analyzing gene expression profiles (GEP) through artificial intelligence provides meaningful insight into cancer disease. This study introduces DeepSHAP Autoencoder Filter for Genes Selection (DSAF-GS), a novel deep learning and explainable artificial intelligence-based approach for feature selection in genomics-scale data. DSAF-GS exploits the autoencoder's reconstruction capabilities without changing the original feature space, enhancing the interpretation of the results. Explainable artificial intelligence is then used to select the informative genes for chronic lymphocytic leukemia prognosis of 217 cases from a GEP database comprising roughly 20,000 genes. The model for prognosis prediction achieved an accuracy of 86.4%, a sensitivity of 85.0%, and a specificity of 87.5%. According to the proposed approach, predictions were strongly influenced by CEACAM19 and PIGP, moderately influenced by MKL1 and GNE, and poorly influenced by other genes. The 10 most influential genes were selected for further analysis. Among them, FADD, FIBP, FIBP, GNE, IGF1R, MKL1, PIGP, and SLC39A6 were identified in the Reactome pathway database as involved in signal transduction, transcription, protein metabolism, immune system, cell cycle, and apoptosis. Moreover, according to the network model of the 3D protein-protein interaction (PPI) explored using the NetworkAnalyst tool, FADD, FIBP, IGF1R, QTRT1, GNE, SLC39A6, and MKL1 appear coupled into a complex network. Finally, all 10 selected genes showed a predictive power on time to first treatment (TTFT) in univariate analyses on a basic prognostic model including IGHV mutational status, del(11q) and del(17p), NOTCH1 mutations, β2-microglobulin, Rai stage, and B-lymphocytosis known to predict TTFT in CLL. However, only IGF1R [hazard ratio (HR) 1.41, 95% CI 1.08-1.84, P=0.013), COL28A1 (HR 0.32, 95% CI 0.10-0.97, P=0.045), and QTRT1 (HR 7.73, 95% CI 2.48-24.04, P<0.001) genes were significantly associated with TTFT in multivariable analyses when combined with the prognostic factors of the basic model, ultimately increasing the Harrell's c-index and the explained variation to 78.6% (versus 76.5% of the basic prognostic model) and 52.6% (versus 42.2% of the basic prognostic model), respectively. Also, the goodness of model fit was enhanced (χ2 = 20.1, P=0.002), indicating its improved performance above the basic prognostic model. In conclusion, DSAF-GS identified a group of significant genes for CLL prognosis, suggesting future directions for bio-molecular research.

Evaluation of the CLL-IPI and IPS-E Prognostic Indices in a Young Middle Eastern Population with Chronic Lymphocytic Leukemia: A Retrospective Analysis at the Kuwait Cancer Control Center

INTRODUCTION

The Chronic Lymphocytic Leukemia International Prognostic Index (CLL-IPI) is a powerful prognostic tool validated in multiple Western populations. However, its utility in the young Middle Eastern population is unknown.

METHODOLOGY

We conducted a retrospective analysis of 152 unselected patients with chronic lymphocytic leukemia (CLL) diagnosed between 2008 and 2022 at the Kuwait Cancer Control Center, which serves as the sole cancer center in Kuwait. The evaluation of the CLL-IPI was based on the assessment of event-free survival (EFS) across the entire cohort. Subsequently, we compared the CLL-IPI with the International Prognostic Score for Early-stage patients (IPS-E) in order to predict the time to first treatment specifically within the subgroup of patients diagnosed with early-stage disease.

RESULTS

The median age of the study cohort was 59.9 years (IQR, 53.1-68.8). The 5-year EFS rates for the low, intermediate, and high/very high-risk categories were approximately 82%, 34%, and 23%, respectively, p < 0.001 (C-statistic = 0.67). On multivariate analysis, advanced stage and unmated IGHV status were independent prognostic factors of EFS. In those with early-stage disease, cumulative 5-year treatment incidence rates for the low, intermediate, and high/very high-risk categories based on the CLL-IPI score were approximately 8%, 55%, and 55%, respectively, p = 0.001 (C-statistic = 0.70). However, based on the IPS-E score, the cumulative 5-year treatment incidence rates for the low, intermediate, and high-risk categories were approximately 0%, 10%, and 60%, respectively, p < 0.001 (C-statistic = 0.73).

CONCLUSIONS

The CLL-IPI and the IPS-E are valid stratification tool in our young Middle Eastern population.

Immune Dysfunction and Infection - Interaction between CLL and Treatment: A Reflection on Current Treatment Paradigms and Unmet Needs

BACKGROUND

Chronic lymphocytic leukemia (CLL) is a hematological malignancy characterized by immune dysfunction, which significantly contributes to increased morbidity and mortality due to infections.

SUMMARY

Advancement in therapeutic strategies based on combination chemoimmunotherapy and targeted treatment have increased life expectancy for patients affected by CLL. However, mortality and morbidity due to infection showed no improvement over the last decades. Although therapy options are highly efficient in targeting leukemic cells, several studies highlighted the interactions of different treatments with the tumor microenvironment immune components, significantly impacting their clinical efficacy and fostering increased risk of infections.

KEY MESSAGES

Given the profound immune dysfunction caused by CLL itself, treatment can thus represent a double-edged sword. Thus, it is essential to increase our understanding and awareness on how conventional therapies affect the disease-microenvironment-infection axis to ensure the best personalized strategy for each patient. This requires careful consideration of the advantages and disadvantages of efficient treatments, whether chemoimmunotherapy or targeted combinations, leading to risk of infectious complications. To this regard, our machine learning-based algorithm CLL Treatment-Infection Model, currently implemented into the local electronic health record system for Eastern Denmark, aims at early identification of patients at high risk of serious infections (PreVent-ACaLL; NCT03868722). We here review strategies for management of immune dysfunction and infections in CLL.

SOHO State of the Art Updates and Next Questions | Infections in Chronic Lymphocytic Leukemia Patients: Risks and Management

Although chronic lymphocytic leukemia (CLL) is a malignancy characterized by accumulation of tumor cells in the blood, bone marrow, lymph nodes and secondary lymphoid tissues, the hallmark of the disease and the major cause of death for patients with CLL is actually immune dysfunction and associated infections. Despite improvement in treatment based on combination chemoimmunotherapy and targeted treatment with BTK and BCL-2 inhibitors leading to longer overall survival for patients with CLL, the mortality due to infections have not improved over the last 4 decades. Thus, infections are now the main cause of death for patients with CLL, posing threats to the patient whether during the premalignant state of monoclonal B lymphocytosis (MBL), during the watch & wait phase for treatment naïve patients, or upon treatment in terms of chemoimmunotherapy or targeted treatment. To test whether the natural history of immune dysfunction and infections in CLL can be changed, we have developed the machine learning based algorithm CLL-TIM.org to identify these patients. The CLL-TIM algorithm is currently being used for selection of patients for the clinical trial PreVent-ACaLL (NCT03868722), testing whether short-term treatment with the BTK inhibitor acalabrutinib and the BCL-2 inhibitor venetoclax can improve immune function and decrease the risk of infections for this high-risk patient population. We here review the background for and management of infectious risks in CLL.

Genetic and Clinical Characteristics of Korean Chronic Lymphocytic Leukemia Patients with High Frequencies of MYD88 Mutations

Chronic lymphocytic leukemia (CLL) is the most common adult leukemia in Western countries. However, CLL is relatively rare in Asia; its genetic features are rarely studied. Here, we aimed to genetically characterize Korean CLL patients and to elucidate the genetic and clinical associations based on data obtained from 113 patients at a single Korean institute. We used next-generation sequencing to explore the multi-gene mutational data and immunoglobulin heavy chain variable gene clonality with somatic hypermutation (SHM). MYD88 (28.3%), including L265P (11.5%) and V217F (13.3%), was the most frequently mutated gene, followed by KMT2D (6.2%), NOTCH1 (5.3%), SF3B1 (5.3%), and TP53 (4.4%). MYD88-mutated CLL was characterized by SHM and atypical immunophenotype with fewer cytogenetic abnormalities. The 5-year time to treatment (TTT) of the overall cohort was 49.8% ± 8.2% (mean ± standard deviation) and the 5-year overall survival was 86.2% ± 5.8%. Patients with SHM, isolated del(13q), TP53-wild type, and NOTCH1-wild type showed better results than those without these conditions. In the subgroup analyses, patients with SHM and L265P presented shorter TTT than patients with SHM but not L265P. In contrast, V217F was associated with a higher SHM percentage and showed a favorable prognosis. Our study revealed the distinct characteristics of Korean CLL patients with high frequencies of MYD88 mutations and their clinical relevance.