Regular health screening plays a crucial role in the early detection of common chronic diseases and prevention of their progression. An AI system capable of recapitulating early disease detection, staging and incidence prediction would help to improve healthcare access and delivery, particularly in resource-poor or remote settings. Using a total of 115,344 retinal fundus photographs from 57,672 patients (with data split into mutually exclusive training, internal testing, and external validation sets), we first developed AI models capable of identifying chronic kidney disease (CKD) and type 2 diabetes mellitus (T2DM) based on fundus images. The AI system was shown to be capable of predicting the clinical indicators of CKD and T2DM (including eGFR and blood glucose levels), which indicates its potential for extracting quantitative clinical metrics embedded subtly within retinal fundus images. We further developed an AI system to predict the risk of disease progression using baseline images of 10,269 patients for whom longitudinal clinical data were available for up to 6 years, which demonstrated potential utility in optimizing health screening intervals and clinical management. The generalizability of the AI system in identifying and predicting the progression of CKD and T2DM was evaluated using population-based external validation cohorts. Moreover, a prospective pilot study with 3,081 patients was also conducted to demonstrate the broader applicability of the AI system at the 'point-of-care' using fundus images captured with smartphones. The results provide proof-of-concept for a reliable and non-invasive AI-based clinical screening tool based on fundus photographs for the early detection and incidence prediction of two common systemic diseases.

Elevated inflammatory markers are associated with poor outcomes in various types of cancers; however, their clinical significance in multiple myeloma (MM) have seldom been explored. This study investigated the prognostic relevance of neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and monocyte-to-lymphocyte ratio (MLR) in MM. Totally 559 MM patients were included in this study. NLR, PLR and MLR were calculated from whole blood counts prior to therapy. Kaplan-Meier curves and multivariate Cox proportional models were used for the evaluation of the survival. It has shown that newly diagnosed MM patients were characterized by high NLR and MLR. Elevated NLR and MLR and decreased PLR were associated with unfavorable clinicobiological features. Applying cut-offs of 4 (NLR), 100 (PLR) and 0.3 (MLR), elevated NLR, MLR and decreased PLR showed a negative impact on outcome. Importantly, elevated NLR and decreased PLR were independent prognostic factors for progression-free survival. Thus, elevated NLR and MLR, and decreased PLR predict poor clinical outcome in MM patients and may serve as the cost-effective and readily available prognostic biomarkers.

The common features shared by primary plasma cell leukemia (pPCL) and multiple myeloma (MM) with circulating plasma cells (CPCs) are peripheral blood invasion and expansion of plasma cells independent of the protective bone marrow (BM) microenvironment niche. However, few studies have addressed the relationship between pPCL and MM with CPCs. Here, we quantitated the number CPCs by conventional morphology in 767 patients with newly diagnosed MM; their clinic features were compared with those of 33 pPCL cases. When the presence of CPCs was defined as more than 2 % plasma cells per 100 nucleated cells on Wright-Giemsa stained peripheral blood smears, the incidence of MM with CPCs was 14.1 % in newly diagnosed MM. Patients with CPCs shared many clinical features with pPCL, especially clinical parameters related to tumor burden. However, no commonalities were found in immunophenotyping and cytogenetics. The prognosis of pPCL was poor, with a median progression free survival (PFS) of 12 months and an overall survival (OS) of 15 months. MM patients with CPCs had a clearly inferior PFS and OS as compared with the control cohort. Most interestingly, although the CPCs were not high enough to meet the diagnostic criteria for pPCL, the survival of MM patients with CPCs was comparable with that of pPCL, with a median PFS of 17 months and an OS of 25 months.

Purpose: Accumulating evidence indicates that intratumor heterogeneity is prevalent in multiple myeloma and that a collection of multiple, genetically distinct subclones are present within the myeloma cell population. It is not clear whether the size of clonal myeloma populations harboring unique cytogenetic abnormalities carry any additional prognostic value.Experimental Design: We analyzed the prognostic impact of cytogenetic aberrations by fluorescence in situ hybridization at different cutoff values in a cohort of 333 patients with newly diagnosed myeloma and 92 patients with relapsed myeloma.Results: We found that nearly all IgH-related arrangements were observed in a large majority of the purified plasma cells; however, 13q deletion, 17p deletion, and 1q21 amplification appeared in different percentages within the malignant plasma cell population. Based on the size of subclones carrying these cytogenetic aberrations, the patients were divided into four groups: 0%-10%, 10.5%-20%, 20.5%-50%, and >50%. Receiver-operating characteristics analysis was applied to determine the optimal cutoff value with the greatest differential survival and showed that the most powerful clone sizes were 10% for 13q deletion, 50% for 17p deletion, and 20% for 1q21 gains, which provided the best possible cutoffs for predicting poor outcomes.Conclusions: Our study indicated that the impact of clone size on prognostic value varies between specific genetic abnormalities. Prognostic value was observed for even a subgroup of plasma cells harboring the cytogenetic aberration of 13q deletion and 1q21 gains; however, 17p deletion displayed the most powerful cutoff for predicting survival only if the predominant clones harbored the abnormality.