Deep learning-based predictive classification of functional subpopulations of hematopoietic stem cells and multipotent progenitors.

BACKGROUND: Hematopoietic stem cells (HSCs) and multipotent progenitors (MPPs) play a pivotal role in maintaining lifelong hematopoiesis. The distinction between stem cells and other progenitors, as well as the assessment of their functions, has long been a central focus in stem cell research. In recent years, deep learning has emerged as a powerful tool for cell image analysis and classification/prediction. METHODS: In this study, we explored the feasibility of employing deep learning techniques to differentiate murine HSCs and MPPs based solely on their morphology, as observed through light microscopy (DIC) images. RESULTS: After rigorous training and validation using extensive image datasets, we successfully developed a three-class classifier, referred to as the LSM model, capable of reliably distinguishing long-term HSCs, short-term HSCs, and MPPs. The LSM model extracts intrinsic morphological features unique to different cell types, irrespective of the methods used for cell identification and isolation, such as surface markers or intracellular GFP markers. Furthermore, employing the same deep learning framework, we created a two-class classifier that effectively discriminates between aged HSCs and young HSCs. This discovery is particularly significant as both cell types share identical surface markers yet serve distinct functions. This classifier holds the potential to offer a novel, rapid, and efficient means of assessing the functional states of HSCs, thus obviating the need for time-consuming transplantation experiments. CONCLUSION: Our study represents the pioneering use of deep learning to differentiate HSCs and MPPs under steady-state conditions. This novel and robust deep learning-based platform will provide a basis for the future development of a new generation stem cell identification and separation system. It may also provide new insight into the molecular mechanisms underlying stem cell self-renewal.

Deep learning-based predictive classification of functional subpopulations of hematopoietic stem cells and multipotent progenitors.

Background: Hematopoietic stem cells (HSCs) and multipotent progenitors (MPPs) play a pivotal role in maintaining lifelong hematopoiesis. The distinction between stem cells and other progenitors, as well as the assessment of their functions, has long been a central focus in stem cell research. In recent years, deep learning has emerged as a powerful tool for cell image analysis and classification/prediction. Methods: In this study, we explored the feasibility of employing deep learning techniques to differentiate murine HSCs and MPPs based solely on their morphology, as observed through light microscopy (DIC) images. Results: After rigorous training and validation using extensive image datasets, we successfully developed a three-class classifier, referred to as the LSM model, capable of reliably distinguishing long-term HSCs (LT-HSCs), short-term HSCs (ST-HSCs), and MPPs. The LSM model extracts intrinsic morphological features unique to different cell types, irrespective of the methods used for cell identification and isolation, such as surface markers or intracellular GFP markers. Furthermore, employing the same deep learning framework, we created a two-class classifier that effectively discriminates between aged HSCs and young HSCs. This discovery is particularly significant as both cell types share identical surface markers yet serve distinct functions. This classifier holds the potential to offer a novel, rapid, and efficient means of assessing the functional states of HSCs, thus obviating the need for time-consuming transplantation experiments. Conclusion: Our study represents the pioneering use of deep learning to differentiate HSCs and MPPs under steady-state conditions. With ongoing advancements in model algorithms and their integration into various imaging systems, deep learning stands poised to become an invaluable tool, significantly impacting stem cell research.

Changes in abdominal adipose tissue depots assessed by MRI correlate with hepatic histologic improvement in non-alcoholic steatohepatitis.

BACKGROUND & AIMS: Non-alcoholic steatohepatitis (NASH) is prevalent in adults with obesity and can progress to cirrhosis. In a secondary analysis of prospectively acquired data from the multicenter, randomized, placebo-controlled FLINT trial, we investigated the relationship between reduction in adipose tissue compartment volumes and hepatic histologic improvement. METHODS: Adult participants in the FLINT trial with paired liver biopsies and abdominal MRI exams at baseline and end-of-treatment (72 weeks) were included (n = 76). Adipose tissue compartment volumes were obtained using MRI. RESULTS: Treatment and placebo groups did not differ in baseline adipose tissue volumes, or in change in adipose tissue volumes longitudinally (p = 0.107 to 0.745). Deep subcutaneous adipose tissue (dSAT) and visceral adipose tissue volume reductions were associated with histologic improvement in NASH (i.e., NAS [non-alcoholic fatty liver disease activity score] reductions of ≥2 points, at least 1 point from lobular inflammation and hepatocellular ballooning, and no worsening of fibrosis) (p = 0.031, and 0.030, respectively). In a stepwise logistic regression procedure, which included demographics, treatment group, baseline histology, baseline and changes in adipose tissue volumes, MRI hepatic proton density fat fraction (PDFF), and serum aminotransferases as potential predictors, reductions in dSAT and PDFF were associated with histologic improvement in NASH (regression coefficient = -2.001 and -0.083, p = 0.044 and 0.033, respectively). CONCLUSIONS: In adults with NASH in the FLINT trial, those with greater longitudinal reductions in dSAT and potentially visceral adipose tissue volumes showed greater hepatic histologic improvements, independent of reductions in hepatic PDFF. CLINICAL TRIAL NUMBER: NCT01265498. IMPACT AND IMPLICATIONS: Although central obesity has been identified as a risk factor for obesity-related disorders including insulin resistance and cardiovascular disease, the role of central obesity in non-alcoholic steatohepatitis (NASH) warrants further clarification. Our results highlight that a reduction in central obesity, specifically deep subcutaneous adipose tissue and visceral adipose tissue, may be related to histologic improvement in NASH. The findings from this analysis should increase awareness of the importance of lifestyle intervention in NASH for clinical researchers and clinicians. Future studies and clinical practice may design interventions that assess the reduction of deep subcutaneous adipose tissue and visceral adipose tissue as outcome measures, rather than simply weight reduction.

Effect of 2-year caloric restriction on organ and tissue size in nonobese 21- to 50-year-old adults in a randomized clinical trial: the CALERIE study.

BACKGROUND: Sustained calorie restriction (CR) promises to extend the lifespan. The effect of CR on changes in body mass across tissues and organs is unclear. OBJECTIVES: We used whole-body MRI to evaluate the effect of 2 y of CR on changes in body composition. METHODS: In an ancillary study of the Comprehensive Assessment of Long-term Effects of Reducing Intake of Energy (CALERIE) trial, 43 healthy adults [25-50 y; BMI (kg/m2): 22-28] randomly assigned to 25% CR (n = 28) or ad libitum (AL) eating (n = 15) underwent whole-body MRI at baseline and month 24 to measure adipose tissue in subcutaneous, visceral, and intermuscular depots (SAT, VAT, and IMAT, respectively); skeletal muscle; and organs including brain, liver, spleen, and kidneys but not heart. RESULTS: The CR group lost more adipose tissue and lean tissue than controls (P < 0.05). In the CR group, at baseline, total tissue volume comprised 32.1%, 1.9%, and 1.0% of SAT, VAT, and IMAT, respectively. The loss of total tissue volume over 24 mo comprised 68.4%, 7.4%, and 2.2% of SAT, VAT, and IMAT, respectively, demonstrating preferential loss of fat vs. lean tissue. Although there is more muscle loss in CR than AL (P < 0.05), the loss of muscle over 24 mo in the CR group comprised only 17.2% of the loss of total tissue volume. Changes in organ volumes were not different between CR and AL. The degree of CR (% decrease in energy intake vs. baseline) significantly (P < 0.05) affected changes in VAT, IMAT, muscle, and liver volume (standardized regression coefficient ± standard error of estimates: 0.43 ± 0.15 L, 0.40 ± 0.19 L, 0.55 ± 0.17 L, and 0.45 ± 0.18 L, respectively). CONCLUSIONS: Twenty-four months of CR (intended, 25%; actual, 13.7%) in young individuals without obesity had effects on body composition, including a preferential loss of adipose tissue, especially VAT, over the loss of muscle and organ tissue. This trial was registered at www.clinicaltrials.gov as NCT02695511.

Genomic, Pathway Network, and Immunologic Features Distinguishing Squamous Carcinomas.

This integrated, multiplatform PanCancer Atlas study co-mapped and identified distinguishing molecular features of squamous cell carcinomas (SCCs) from five sites associated with smoking and/or human papillomavirus (HPV). SCCs harbor 3q, 5p, and other recurrent chromosomal copy-number alterations (CNAs), DNA mutations, and/or aberrant methylation of genes and microRNAs, which are correlated with the expression of multi-gene programs linked to squamous cell stemness, epithelial-to-mesenchymal differentiation, growth, genomic integrity, oxidative damage, death, and inflammation. Low-CNA SCCs tended to be HPV(+) and display hypermethylation with repression of TET1 demethylase and FANCF, previously linked to predisposition to SCC, or harbor mutations affecting CASP8, RAS-MAPK pathways, chromatin modifiers, and immunoregulatory molecules. We uncovered hypomethylation of the alternative promoter that drives expression of the ΔNp63 oncogene and embedded miR944. Co-expression of immune checkpoint, T-regulatory, and Myeloid suppressor cells signatures may explain reduced efficacy of immune therapy. These findings support possibilities for molecular classification and therapeutic approaches.

Multilevel Genomics-Based Taxonomy of Renal Cell Carcinoma.

On the basis of multidimensional and comprehensive molecular characterization (including DNA methalylation and copy number, RNA, and protein expression), we classified 894 renal cell carcinomas (RCCs) of various histologic types into nine major genomic subtypes. Site of origin within the nephron was one major determinant in the classification, reflecting differences among clear cell, chromophobe, and papillary RCC. Widespread molecular changes associated with TFE3 gene fusion or chromatin modifier genes were present within a specific subtype and spanned multiple subtypes. Differences in patient survival and in alteration of specific pathways (including hypoxia, metabolism, MAP kinase, NRF2-ARE, Hippo, immune checkpoint, and PI3K/AKT/mTOR) could further distinguish the subtypes. Immune checkpoint markers and molecular signatures of T cell infiltrates were both highest in the subtype associated with aggressive clear cell RCC. Differences between the genomic subtypes suggest that therapeutic strategies could be tailored to each RCC disease subset.

Extranodal hematopoietic neoplasms and mimics in the head and neck: an update.

The head and neck region is a common site for extranodal lymphomas, second only to the gastrointestinal tract; and 12% to 15% of all head and neck tumors are lymphomas. Non-Hodgkin lymphomas are most common, and Hodgkin lymphoma occurs rarely at extranodal sites in the head and neck. Most non-Hodgkin lymphomas of the head and neck region are of B-cell lineage, and the Waldeyer ring is the most common site. Head and neck lymphomas have distinctive epidemiological and clinicopathologic features, including an association with immunosuppression, infectious organisms, or autoimmune disorders; site-specific differences (eg, thyroid gland versus ocular adnexa) for common lymphomas, such as extranodal marginal zone B-cell lymphoma of mucosa-associated lymphoid tissue; and genetic differences that provide insights into etiology. Furthermore, the diagnosis of non-Hodgkin lymphomas at extranodal sites implies differences in prognosis and therapeutic implications with lymphomas at nodal sites. In this review, we discuss various types of non-Hodgkin lymphomas and Hodgkin lymphoma, focusing on unique aspects related to the head and neck region. We also discuss a number of newer entities that are clinically indolent as well as mimics of lymphoma that can occur in the head and neck region, including infectious mononucleosis, Kikuchi-Fujimoto disease, Kimura disease, Castleman disease, and immunoglobulin G4-related disease.

Biobanking in genomic medicine.

CONTEXT: Genomic medicine requires the identification of biomarkers and therapeutic targets, which in turn, requires high-quality biospecimens. Achieving high-quality biospecimens requires implementing standard operating procedures to control the variations of preanalytic variables in biobanking. Currently, most biobanks do not control the variations of preanalytic variables when collecting, processing, and storing their biospecimens. However, those variations have been shown to affect the quality of biospecimens and gene expression profiling. OBJECTIVE: To identify evidence-based preanalytic parameters that can be applied and those parameters that need further study. DATA SOURCES: We searched the Biospecimen Research and PubMed databases using defined key words. We retrieved and reviewed 212 articles obtained through those searches. We included 58 articles (27%) according to our inclusion and exclusion criteria for this review. CONCLUSION: -Preanalytic variables in biobanking can degrade the quality of biospecimens and alter gene expression profiling. Variables that require further study include the effect of surgical manipulation; the effect of warm ischemia; the allowable duration of delayed specimen processing; the optimal type, duration, and temperature of preservation and fixation; and the optimal storage duration of formalin-fixed, paraffin embedded specimens in a fit-for-purpose approach.

Squamous cell carcinoma of parotid gland associated with concurrent lymphoepithelial cysts and lymphoepithelial lesion: case report and proposed histogenesis.

Lymphoepithelial cyst and lymphoepithelial lesion have similar histologic features and an affinity for the parotid gland. Though considered as different entities, both conditions arise from heterotopic salivary epithelial rests or inclusions in intra- or peri-parotid lymph nodes. We present a case of squamous cell carcinoma of parotid gland associated with concurrent lymphoepithelial cyst and lymphoepithelial lesion in a patient who was not infected with human immunodeficiency virus. We propose that lymphoepithelial cyst and lymphoepithelial lesion have a similar histogenesis.

Mutational landscape of aggressive cutaneous squamous cell carcinoma.

PURPOSE: Aggressive cutaneous squamous cell carcinoma (cSCC) is often a disfiguring and lethal disease. Very little is currently known about the mutations that drive aggressive cSCC. EXPERIMENTAL DESIGN: Whole-exome sequencing was performed on 39 cases of aggressive cSCC to identify driver genes and novel therapeutic targets. Significantly, mutated genes were identified with MutSig or complementary methods developed to specifically identify candidate tumor suppressors based upon their inactivating mutation bias. RESULTS: Despite the very high-mutational background caused by UV exposure, 23 candidate drivers were identified, including the well-known cancer-associated genes TP53, CDKN2A, NOTCH1, AJUBA, HRAS, CASP8, FAT1, and KMT2C (MLL3). Three novel candidate tumor suppressors with putative links to cancer or differentiation, NOTCH2, PARD3, and RASA1, were also identified as possible drivers in cSCC. KMT2C mutations were associated with poor outcome and increased bone invasion. CONCLUSIONS: The mutational spectrum of cSCC is similar to that of head and neck squamous cell carcinoma and dominated by tumor-suppressor genes. These results improve the foundation for understanding this disease and should aid in identifying and treating aggressive cSCC.

Residual tumor thickness at the tumor-normal tissue interface predicts the recurrence-free survival in patients with liver metastasis of breast cancer.

Tumor response to neoadjuvant therapy is a significant predictive indicator of recurrence-free survival. We measured tumor response using residual tumor thickness at the tumor-normal tissue interface (TNI) and evaluated its association with outcome in patients with liver metastasis of breast cancer. We included 48 patients who underwent neoadjuvant therapy followed by partial liver resection at MD Anderson Cancer Center between 1997 and 2010. The hematoxylin-eosin-stained tumor sections were evaluated for both pathologic response and the residual tumor thickness at the TNI by 3 pathologists who were blinded to the clinical information, treatment regimen, and patient outcome. The residual tumor thickness at the TNI was correlated with recurrence-free survival using Kaplan-Meier method and log-rank test. Cox proportional hazard model was used to identify predictors of recurrence-free survival. All patients were women with a median age of 43 years. The median duration of follow-up was 52.1 months. Residual tumor thickness less than or equal to 3 mm at the TNI correlated with major pathologic response and was associated with longer recurrence-free survival in both univariate and multivariate analyses. Residual tumor thickness at the TNI predicts recurrence-free survival and provides an objective outcome end point in patients who underwent neoadjuvant therapy and liver resection of metastatic breast cancer. We suggest using both the pathologic response and the residual tumor thickness at the TNI to measure tumor response to therapy to improve accuracy.

Squamous cell carcinoma of the oral tongue in young non-smokers is genomically similar to tumors in older smokers.

PURPOSE: Epidemiologic studies have identified an increasing incidence of squamous cell carcinoma of the oral tongue (SCCOT) in younger patients. EXPERIMENTAL DESIGN: DNA isolated from tongue tumors of young (<45 years, non-smokers) and old (>45 years) patients at was subjected to whole-exome sequencing and copy-number analysis. These data were compared with data from similar patients in the TCGA (The Cancer Genome Atlas) project. RESULTS: In this study, we found that gene-specific mutation and copy-number alteration frequencies were similar between young and old patients with SCCOT in two independent cohorts. Likewise, the types of base changes observed in the young cohort were similar to those in the old cohort even though they differed in smoking history. TCGA data also demonstrate that the genomic effects of smoking are tumor site-specific, and we find that smoking has only a minor impact on the types of mutations observed in SCCOT. CONCLUSIONS: Overall, tumors from young patients with SCCOT appear genomically similar to those of older patients with SCCOT, and the cause for the increasing incidence of young SCCOT remains unknown. These data indicate that the functional impact of smoking on carcinogenesis in SCCOT is still poorly understood.

Immunohistochemical expression of hormone receptors in melanoma of pregnant women, nonpregnant women, and men.

The survival advantage of women over men with cutaneous melanoma and the reports of accelerated progression of melanoma during pregnancy have led to studies of the effect of hormones and hormone receptors on the development and progression of melanoma. However, the results are inconclusive. We therefore evaluated the expression of estrogen receptor α, estrogen receptor ß, and androgen receptor in melanomas of stage- and age-matched pregnant women, nonpregnant women, and men by immunohistochemical analysis of formalin-fixed, paraffin-embedded archival tissues. In addition, we also assessed the mitotic rate using the antiphosphohistone H3 antibody by immunohistochemistry. Our data showed a trend of more frequent expression of estrogen receptor ß in the melanomas of pregnant patients than in the melanomas of male patients, without a significant difference observed between pregnant and nonpregnant women. However, no association between the expression of estrogen receptor ß and survival was observed. The small cohort may have limited the statistical power of the study, and large-scale studies are needed to elucidate the potential role of estrogen receptor ß as a prognostic marker of melanoma.

Myeloid sarcoma of the head and neck region.

CONTEXT: Myeloid sarcoma of the head and neck region can pose diagnostic challenges because of the low frequency of myeloid sarcoma and the potential for tumors of almost any lineage to occur in the head and neck. OBJECTIVE: To study the clinicopathologic and immunohistochemical characteristics of myeloid sarcoma in the head and neck region and to review the differential diagnosis. DESIGN: We searched for cases of myeloid sarcoma involving the head and neck region for a 24-year period at our institution. The medical records and pathology slides were reviewed. Additional immunohistochemical stains were performed. RESULTS: We identified 17 patients, age 17 to 85 years. Most tumors involved the oral cavity. Myeloid sarcoma was the initial diagnosis in 9 patients (53%); the remaining 8 patients (47%) had a history of bone marrow disease. Immunohistochemical analysis using antibodies specific for lysozyme, CD43, and CD68 were highly sensitive for diagnosis but were not specific. By contrast, assessment for myeloperoxidase in this study was less sensitive but more specific. We also used antibodies specific for CD11c and CD33 in a subset of cases, and these reagents seem helpful as well. CONCLUSIONS: The clinical presentation of myeloid sarcoma involving the head and neck, particularly the mouth, is often nonspecific, and a high degree of suspicion for the possibility of myeloid sarcoma is needed. Immunohistochemistry is very helpful for establishing the diagnosis.

ALDH1 immunohistochemical expression and its significance in salivary adenoid cystic carcinoma.

BACKGROUND: Adenoid cystic carcinoma (ACC) cells with a high level of ALDH1, a known cancer stem cell (CSC) marker, had higher tumorigenic, invasive, and metastatic abilities. We examined the immunohistochemical expression of ALDH1 in ACC and its correlation with survival. METHODS: Archival paraffin blocks of ACC were analyzed. A tissue microarray was constructed and immunohistochemical expression of ALDH1 was analyzed using anti-ALDH1 monoclonal antibody. Correlations between ALDH1 expression and clinical and histological parameters were assessed by chi-square tests. Survival was assessed by the Kaplan-Meier method and log-rank test. RESULTS: Most of the tumors (63%) showed stromal staining only, 11% of the tumors showed both epithelial and stromal expression, and 26% of the tumors did not show either epithelial or stromal staining. Statistical analyses did not show any correlation between the pattern of ALDH1 expression and tumor histology, tumor size, or perineural invasion. There were no significant differences in survival among the 3 patterns of ALDH1 expression. CONCLUSION: Other factors, besides CSCs, may play important roles in tumorigenesis, cell differentiation, and tumor progression in these tumors.

Quercetin: a potential drug to reverse multidrug resistance.

This review centers on recent findings with respect to modulating cancer multidrug resistance (MDR) with the well-known flavonoid quercetin. After a short introduction of quercetin, major in vitro and in vivo findings are summarized showing that quercetin is a MDR modulator and thus a potential chemosensitizer. Finally, we contemplate future prospects of modulating MDR in the clinic.