Digital Health and Real-World Evidence in Onco-Hematological Patients

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.

Maintenance treatment is a pivotal part in the whole process management of multiple myeloma (MM), which further deepens response and improves survival. However, evidence of maintenance in non-transplant MM patients is inadequate in real-world practice. Here, we retrospectively analyzed the efficacy and survival of 375 non-transplant MM patients from 11 centers between 2010 and 2021 in north China. After a median of seven cycles of front-line regimens, there were 141, 79, and 155 patients receiving lenalidomide maintenance (L-MT), bortezomib maintenance (B-MT), or thalidomide maintenance (T-MT), respectively. Patients on L-MT and B-MT had significantly greater proportions of high-risk cytogenetic abnormalities (HRCAs) detected by fluorescence in situ hybridization (FISH), which was defined as 1q21 gain, 17p deletion, adverse immunoglobulin heavy chain (IgH) translocations. Although the progression-free survival (PFS) and overall survival (OS) were comparable among the three groups, L-MT and B-MT remedied the negative impact of HRCAs on survival (PFS of patients with HRCAs vs. patients without HRCAs: L-MT, 26.9 vs. 39.2 months, p=0.19; B-MT, 20.0 vs. 29.7 months, p=0.36; OS not reached in all groups). Patients with HRCAs in the T-MT group presented inferior clinical outcomes compared to standard-risk patients (PFS, 12.1 vs. 22.8 months, p=0.02, HR=1.8, 95% CI 1.0–3.4; OS, 54.9 months vs. NR, p<0.001, HR=3.2, 95% CI 1.5–7.0). Achieving complete response (CR) after induction therapy led to superior PFS compared to other degrees of response, regardless of maintenance medication. Furthermore, maintenance duration over 24 months correlated with favorable survival. Due to the large gap of transplant eligibility in China, optimizing maintenance therapy is important for non-transplant MM patients. In this real-world multi-centered study, our findings suggest that clinicians prefer to prescribe lenalidomide or bortezomib as maintenance therapy in high-risk settings, which are superior to thalidomide in non-transplant MM patients. Achievement of CR and maintenance duration over 2 years are positive factors that influence survival.