Deep Learning-Enabled Diagnosis of Liver Adenocarcinoma.

BACKGROUND & AIMS: Diagnosis of adenocarcinoma in the liver is a frequent scenario in routine pathology and has a critical impact on clinical decision making. However, rendering a correct diagnosis can be challenging, and often requires the integration of clinical, radiologic, and immunohistochemical information. We present a deep learning model (HEPNET) to distinguish intrahepatic cholangiocarcinoma from colorectal liver metastasis, as the most frequent primary and secondary forms of liver adenocarcinoma, with clinical grade accuracy using H&E-stained whole-slide images. METHODS: HEPNET was trained on 714,589 image tiles from 456 patients who were randomly selected in a stratified manner from a pool of 571 patients who underwent surgical resection or biopsy at Heidelberg University Hospital. Model performance was evaluated on a hold-out internal test set comprising 115 patients and externally validated on 159 patients recruited at Mainz University Hospital. RESULTS: On the hold-out internal test set, HEPNET achieved an area under the receiver operating characteristic curve of 0.994 (95% CI, 0.989-1.000) and an accuracy of 96.522% (95% CI, 94.521%-98.694%) at the patient level. Validation on the external test set yielded an area under the receiver operating characteristic curve of 0.997 (95% CI, 0.995-1.000), corresponding to an accuracy of 98.113% (95% CI, 96.907%-100.000%). HEPNET surpassed the performance of 6 pathology experts with different levels of experience in a reader study of 50 patients (P = .0005), boosted the performance of resident pathologists to the level of senior pathologists, and reduced potential downstream analyses. CONCLUSIONS: We provided a ready-to-use tool with clinical grade performance that may facilitate routine pathology by rendering a definitive diagnosis and guiding ancillary testing. The incorporation of HEPNET into pathology laboratories may optimize the diagnostic workflow, complemented by test-related labor and cost savings.

Upper Tract Urinary Cancer Recurrence after Radical Cystectomy: Risk Assessment of Intraoperative Frozen Section.

INTRODUCTION: Upper tract urinary cancer recurrence (UTUCR) after radical cystectomy (RC) is outcome-limiting. Surgical recommendations on intraoperative performance of frozen section analysis (FSA) and management of positive ureteral margin (PUM) are lacking. METHODS: 634 RC cases were identified (2010-2018). In patients with PUM, sequential ureteral resections up to a negative margin were performed. We investigated the accuracy of FSA, significance of PUM, and identified risk factors (RFs) to stratify patients for UTUCR. RESULTS: FSA was performed in 355 patients, including a total of 693 ureters. FSA sensitivity was 0.93 and specificity 0.99. PUM conversion was possible in 52 (91.2%) cases. UTUCR occurred in 17 (4.8%) patients. Identified UTUCR RFs are non-muscle invasive bladder carcinoma (NMIBC) (OR 3.8, 95% confidence intervals [CI] 1.4-10.2, p = 0.008), multifocal bladder cancer in cystectomy specimen (OR 4.7, CI 1.1-20.8, p = 0.042), and recurrent NMIBC (OR 4.1, CI 1.5-10.9, p = 0.006). Risk-group stratification showed a six-fold increase in UTUCR risk (low-to high-risk). CONCLUSION: FSA is a sensitive and specific method to identify PUM. UTUCR occurs significantly more often in patients with recurrent, multifocal NMIBC at the time of RC. Patients can be risk stratified for UTUCR. In case of NMIBC-PUM, surgeons can safely opt for a kidney preserving strategy.

Singleton-Merten syndrome: A rare cause of femoral head necrosis.

Singleton-Merten syndrome (SMS) is a type I interferonopathy. In this report, we disclose the first-to the best of our knowledge-direct association of SMS with femoral head necrosis (FHN). The following case report presents the condition of a 38-year-old male suffering from SMS with FHN, characterized by acute symptoms and rapid disease progression. As per the recommendations of the Association Research Circulation Osseous (ARCO) and the S3-guidelines, we successfully treated the FHN with core decompression. Our histological results correlate with the changes described in medical literature in patients with SMS and MDA5-knockout in vivo experiments such as osteopenia, widened medullary cavity, and thin cortical bone. Moreover, the conducted immunohistochemistry shows strong CD56 positivity of the osteoblasts and osteocytes, as well as significant CD68 and CD163 positivity of the middle-sized osteoclasts. Collectively, these findings suggest an underlying syndrome in the FHN. A six-month post-operative follow-up revealed complete recovery with the absence of the initial symptoms and ability to resume normal daily activities. Taken together, our findings suggest that SMS is an additional cause of FHN in young adults. Early detection and adequate treatment using well-established joint-preserving techniques demonstrate a favorable improvement of the patient's clinical condition.

RNA Expression of DNA Damage Response Genes in Muscle-Invasive Bladder Cancer: Influence on Outcome and Response to Adjuvant Cisplatin-Based Chemotherapy.

BACKGROUND: Perioperative cisplatin-based chemotherapy (CBC) can improve the outcome of patients with muscle-invasive bladder cancer (MIBC), but it is still to be defined which patients benefit. Mutations in DNA damage response genes (DDRG) can predict the response to CBC. The value of DDRG expression as a marker of CBC treatment effect remains unclear. MATERIAL AND METHODS: RNA expression of the nine key DDRG (BCL2, BRCA1, BRCA2, ERCC2, ERCC6, FOXM1, RAD50, RAD51, and RAD52) was assessed by qRT-PCR in a cohort of 61 MICB patients (median age 66 y, 48 males, 13 females) who underwent radical cystectomy in a tertiary care center. The results were validated in the The Cancer Genome Atlas (TCGA) cohort of MIBC (n = 383). Gene expression was correlated with disease-free survival (DFS) and overall survival (OS). Subgroup analyses were performed in patients who received adjuvant cisplatin-based chemotherapy (ACBC) (Mannheim n = 20 and TCGA n = 75). RESULTS: Low expression of RAD52 was associated with low DFS in both the Mannheim and the TCGA cohorts (Mannheim: p = 0.039; TCGA: p = 0.017). This was especially apparent in subgroups treated with ACBC (Mannheim: p = 0.0059; TCGA: p = 0.012). Several other genes showed an influence on DFS in the Mannheim cohort (BRCA2, ERCC2, FOXM1) where low expression was associated with poor DFS (p < 0.05 for all). This finding was not fully supported by the data in the TCGA cohort, where high expression of FOXM1 and BRCA2 correlated with poor DFS. CONCLUSION: Low expression of RAD52 correlated with decreased DFS in the Mannheim and the TCGA cohort. This effect was especially pronounced in the subset of patients who received ACBC, making it a promising indicator for response to ACBC on the level of gene expression.

[Solitary locoregional metastasis of an undifferentiated pleomorphic sarcoma in the M. quadratus femoris]. / Solitäre lokoregionäre Metastase eines undifferenzierten pleomorphen Sarkoms im M. quadratus femoris.

The undifferentiated pleomorphic sarcoma (UPS) is a part of the soft tissue sarcoma group and represents almost 10% of all soft tissue sarcomas. The case of a 49-year-old patient is presented who was diagnosed with a primary UPS in the left gluteus maximus muscle, which was treated with compartmental resection with adjuvant radiotherapy (60 Gy). During tumor follow-up (3 years later) a locoregional metastasis at an unusual location in the quadratus femoris muscle was detected, which was treated by in toto resection with intraoperative radiotherapy (10 Gy). The intra and postoperative outcome was without complications and without neurological deficits. In the last follow-up, 6 months postoperatively, the patient was free of tumors and symptoms.

Phosphodiesterase SMPDL3B Gene Expression as Independent Outcome Prediction Marker in Localized Prostate Cancer.

Current outcome prediction markers for localized prostate cancer (PCa) are insufficient. The impact of the lipid-modifying Sphingomyelin Phosphodiesterase Acid Like 3B (SMPDL3B) in PCa is unknown. Two cohorts of patients with PCa who underwent radical prostatectomy (n = 40, n = 56) and benign prostate hyperplasia (BPH) controls (n = 8, n = 11) were profiled for SMPDL3B expression with qRT-PCR. Publicly available PCa cohorts (Memorial Sloane Kettering Cancer Centre (MSKCC; n = 131, n = 29 controls) and The Cancer Genome Atlas (TCGA; n = 497, n = 53 controls)) served for validation. SMPDL3B's impact on proliferation and migration was analyzed in PC3 cells by siRNA knockdown. In both cohorts, a Gleason score and T stage independent significant overexpression of SMPDL3B was seen in PCa compared to BPH (p < 0.001 each). A lower expression of SMPDL3B was associated with a shorter overall survival (OS) (p = 0.005) in long term follow-up. A SMPDL3B overexpression in PCa tissue was confirmed in the validation cohorts (p < 0.001 each). In the TCGA patients with low SMPDL3B expression, biochemical recurrence-free survival (p = 0.011) and progression-free interval (p < 0.001) were shorter. Knockdown of SMPDL3B impaired PC3 cell migration but not proliferation (p = 0.0081). In summary, SMPLD3B is highly overexpressed in PCa tissue, is inversely associated with localized PCa prognosis, and impairs PCa cell migration.

Predicting survival from colorectal cancer histology slides using deep learning: A retrospective multicenter study.

BACKGROUND: For virtually every patient with colorectal cancer (CRC), hematoxylin-eosin (HE)-stained tissue slides are available. These images contain quantitative information, which is not routinely used to objectively extract prognostic biomarkers. In the present study, we investigated whether deep convolutional neural networks (CNNs) can extract prognosticators directly from these widely available images. METHODS AND FINDINGS: We hand-delineated single-tissue regions in 86 CRC tissue slides, yielding more than 100,000 HE image patches, and used these to train a CNN by transfer learning, reaching a nine-class accuracy of >94% in an independent data set of 7,180 images from 25 CRC patients. With this tool, we performed automated tissue decomposition of representative multitissue HE images from 862 HE slides in 500 stage I-IV CRC patients in the The Cancer Genome Atlas (TCGA) cohort, a large international multicenter collection of CRC tissue. Based on the output neuron activations in the CNN, we calculated a "deep stroma score," which was an independent prognostic factor for overall survival (OS) in a multivariable Cox proportional hazard model (hazard ratio [HR] with 95% confidence interval [CI]: 1.99 [1.27-3.12], p = 0.0028), while in the same cohort, manual quantification of stromal areas and a gene expression signature of cancer-associated fibroblasts (CAFs) were only prognostic in specific tumor stages. We validated these findings in an independent cohort of 409 stage I-IV CRC patients from the "Darmkrebs: Chancen der Verhütung durch Screening" (DACHS) study who were recruited between 2003 and 2007 in multiple institutions in Germany. Again, the score was an independent prognostic factor for OS (HR 1.63 [1.14-2.33], p = 0.008), CRC-specific OS (HR 2.29 [1.5-3.48], p = 0.0004), and relapse-free survival (RFS; HR 1.92 [1.34-2.76], p = 0.0004). A prospective validation is required before this biomarker can be implemented in clinical workflows. CONCLUSIONS: In our retrospective study, we show that a CNN can assess the human tumor microenvironment and predict prognosis directly from histopathological images.

The EEF1A2 gene expression as risk predictor in localized prostate cancer.

BACKGROUND: Besides clinical stage and Gleason score, risk-stratification of prostate cancer in the pretherapeutic setting mainly relies on the serum PSA level. Yet, this is associated with many uncertainties. With regard to therapy decision-making, additional markers are needed to allow an exact risk prediction. Eukaryotic translation elongation factor 1 alpha 2 (EEF1A2) was previously suggested as driver of tumor progression and potential biomarker. In the present study its functional and prognostic relevance in prostate cancer was investigated. METHODS: EEF1A2 expression was analyzed in two cohorts of patients (n = 40 and n = 59) with localized PCa. Additionally data from two large expression dataset (MSKCC, Cell, 2010 with n = 131 localized, n = 19 metastatic PCa and TCGA provisional data, n = 499) of PCa patients were reanalyzed. The expression of EEF1A2 was correlated with histopathology features and biochemical recurrence (BCR). To evaluate the influence of EEF1A2 on proliferation and migration of metastatic PC3 cells, siRNA interference was used. Statistical significance was tested with t-test, Mann-Whitney-test, Pearson correlation and log-rank test. RESULTS: qRT-PCR revealed EEF1A2 to be significantly overexpressed in PCa tissue, with an increase according to tumor stage in one cohort (p = 0.0443). In silico analyses in the MSKCC cohort confirmed the overexpression of EEF1A2 in localized PCa with high Gleason score (p = 0.0142) and in metastatic lesions (p = 0.0038). Patients with EEF1A2 overexpression had a significantly shorter BCR-free survival (p = 0.0028). EEF1A2 expression was not correlated with serum PSA levels. Similar results were seen in the TCGA cohort, where EEF1A2 overexpression only occurred in tumors with Gleason 7 or higher. Patients with elevated EEF1A2 expression had a significantly shorter BCR-free survival (p = 0.043). EEF1A2 knockdown significantly impaired the migration, but not the proliferation of metastatic PC3 cells. CONCLUSION: The overexpression of EEF1A2 is a frequent event in localized PCa and is associated with histopathology features and a shorter biochemical recurrence-free survival. Due to its independence from serum PSA levels, EEF1A2 could serve as valuable biomarker in risk-stratification of localized PCa.

Loss of syntaxin 3 causes variant microvillus inclusion disease.

Microvillus inclusion disease (MVID) is a disorder of intestinal epithelial differentiation characterized by life-threatening intractable diarrhea. MVID can be diagnosed based on loss of microvilli, microvillus inclusions, and accumulation of subapical vesicles. Most patients with MVID have mutations in myosin Vb that cause defects in recycling of apical vesicles. Whole-exome sequencing of DNA from patients with variant MVID showed homozygous truncating mutations in syntaxin 3 (STX3). STX3 is an apical receptor involved in membrane fusion of apical vesicles in enterocytes. Patient-derived organoid cultures and overexpression of truncated STX3 in Caco-2 cells recapitulated most characteristics of variant MVID. We conclude that loss of STX3 function causes variant MVID.

Point Cloud Registration for Measuring Shape Dependence of Soft Tissue Deformation by Digital Twins in Head and Neck Surgery.

Introduction: A 2½ D point cloud registration method was developed to generate digital twins of different tissue shapes and resection cavities by applying a machine learning (ML) approach. This demonstrates the feasibility of quantifying soft tissue shifts. Methods: An ML model was trained using simulated surface scan data obtained from tumor resections in a pig head cadaver model. It hereby uses 438 2½ D scans of the tissue surface. Tissue shift was induced by a temperature change from 7.91 ± 4.1°C to 36.37 ± 1.28°C. Results: Digital twins were generated from various branched and compact resection cavities (RCs) and cut tissues (CT). A temperature increase induced a tissue shift with a significant volume increase of 6 mL and 2 mL in branched and compact RCs, respectively (p = 0.0443; 0.0157). The volumes of branched and compact CT were decreased by 3 and 4 mL (p < 0.001). In the warm state, RC and CT no longer fit together because of the significant tissue deformation. Although not significant, the compact RC showed a greater tissue deformation of 1 µL than the branched RC with 0.5 µL induced by the temperature change (p = 0.7874). The branched and compact CT forms responded almost equally to changes in temperature (p = 0.1461). Conclusions: The simulation experiment of induced soft tissue deformation using digital twins based on 2½ D point cloud models proved that our method helps to quantify shape-dependent tissue shifts.

Spatial Omics Imaging of Fresh-Frozen Tissue and Routine FFPE Histopathology of a Single Cancer Needle Core Biopsy: A Freezing Device and Multimodal Workflow.

The complex molecular alterations that underlie cancer pathophysiology are studied in depth with omics methods using bulk tissue extracts. For spatially resolved tissue diagnostics using needle biopsy cores, however, histopathological analysis using stained FFPE tissue and the immunohistochemistry (IHC) of a few marker proteins is currently the main clinical focus. Today, spatial omics imaging using MSI or IRI is an emerging diagnostic technology for the identification and classification of various cancer types. However, to conserve tissue-specific metabolomic states, fast, reliable, and precise methods for the preparation of fresh-frozen (FF) tissue sections are crucial. Such methods are often incompatible with clinical practice, since spatial metabolomics and the routine histopathology of needle biopsies currently require two biopsies for FF and FFPE sampling, respectively. Therefore, we developed a device and corresponding laboratory and computational workflows for the multimodal spatial omics analysis of fresh-frozen, longitudinally sectioned needle biopsies to accompany standard FFPE histopathology of the same biopsy core. As a proof-of-concept, we analyzed surgical human liver cancer specimens using IRI and MSI with precise co-registration and, following FFPE processing, by sequential clinical pathology analysis of the same biopsy core. This workflow allowed for a spatial comparison between different spectral profiles and alterations in tissue histology, as well as a direct comparison for histological diagnosis without the need for an extra biopsy.

Modulation of PD­L1 expression by standard therapy in head and neck cancer cell lines and exosomes.

Although checkpoint inhibitors (CPI) have recently extended the treatment options and improved clinical response of advanced stage head and neck squamous cell carcinoma (HNSCC), treatment success remains unpredictable. Programmed cell death ligand­1 (PD­L1) is a key player in immunotherapy. Tumor cells, and exosomes derived therefrom, are carriers of PD­L1 and efficiently suppress immune responses. The aim of the present study was to analyze the influence of established therapies on PD­L1 expression of HNSCC cell lines and their exosomes. The HNSCC cell lines, UM­SCC­11B, UM­SCC­14C and UM­SCC­22C were treated with fractionated radiotherapy (RT; 5x2 Gy), cisplatin (CT) and cetuximab (Cetux) as monotherapy, or combined therapy, chemoradiotherapy (CRT; RT and CT) or radioimmunotherapy (RT and Cetux). The expression of PD­L1 and phosphorylated (p)ERK1/2 as a mediator of radioresistance were assessed using western blotting, immunohistochemistry and an ex vivo vital tissue culture model. Additionally, exosomes were isolated from concentrated supernatants of the (un­)treated HNSCC cell lines by size exclusion chromatography. Exosomal protein expression levels of PD­L1 were detected using western blotting and semi­quantitative levels were calculated. The functional impact of exosomes from the (un­)treated HNSCC cell lines on the proliferation (MTS assay) and apoptosis (Caspase 3/7 assay) of the untreated HNSCC cell lines were measured and compared. The HNSCC cell lines UM­SCC­11B and UM­SCC­22B showed strong expression of pERK1/2 and PD­L1, respectively. RT upregulated the PD­L1 expression in UM­SCC­11B and UM­SCC­14C and in exosomes from all three cell lines. CT alone induced PD­L1 expression in all cell lines. CRT induced the expression of PD­L1 in all HNSCC cell lines and exosomes from UM­SCC­14C and UM­SCC­22B. The data indicated a potential co­regulation of PD­L1 and activated ERK1/2, most evident in UM­SCC­14C. Exosomes from irradiated UM­SCC­14C cells protected the unirradiated cells from apoptosis by Caspase 3/7 downregulation. The present study suggested a tumor cell­mediated regulation of PD­L1 upon platinum­based CRT in HNSCC and in exosomes. A co­regulation of PD­L1 and MAPK signaling response was hypothesized.

Inhibition of lysyl oxidases synergizes with 5-azacytidine to restore erythropoiesis in myelodysplastic and myeloid malignancies.

Limited response rates and frequent relapses during standard of care with hypomethylating agents in myelodysplastic neoplasms (MN) require urgent improvement of this treatment indication. Here, by combining 5-azacytidine (5-AZA) with the pan-lysyl oxidase inhibitor PXS-5505, we demonstrate superior restoration of erythroid differentiation in hematopoietic stem and progenitor cells (HSPCs) of MN patients in 20/31 cases (65%) versus 9/31 cases (29%) treated with 5-AZA alone. This effect requires direct contact of HSPCs with bone marrow stroma components and is dependent on integrin signaling. We further confirm these results in vivo using a bone marrow niche-dependent MN xenograft model in female NSG mice, in which we additionally demonstrate an enforced reduction of dominant clones as well as significant attenuation of disease expansion and normalization of spleen sizes. Overall, these results lay out a strong pre-clinical rationale for efficacy of combination treatment of 5-AZA with PXS-5505 especially for anemic MN.

Natural Language Processing in Diagnostic Texts from Nephropathology.

INTRODUCTION: This study investigates whether it is possible to predict a final diagnosis based on a written nephropathological description-as a surrogate for image analysis-using various NLP methods. METHODS: For this work, 1107 unlabelled nephropathological reports were included. (i) First, after separating each report into its microscopic description and diagnosis section, the diagnosis sections were clustered unsupervised to less than 20 diagnostic groups using different clustering techniques. (ii) Second, different text classification methods were used to predict the diagnostic group based on the microscopic description section. RESULTS: The best clustering results (i) could be achieved with HDBSCAN, using BoW-based feature extraction methods. Based on keywords, these clusters can be mapped to certain diagnostic groups. A transformer encoder-based approach as well as an SVM worked best regarding diagnosis prediction based on the histomorphological description (ii). Certain diagnosis groups reached F1-scores of up to 0.892 while others achieved weak classification metrics. CONCLUSION: While textual morphological description alone enables retrieving the correct diagnosis for some entities, it does not work sufficiently for other entities. This is in accordance with a previous image analysis study on glomerular change patterns, where some diagnoses are associated with one pattern, but for others, there exists a complex pattern combination.

Assessment of glomerular morphological patterns by deep learning algorithms.

BACKGROUND: Compilation of different morphological lesion signatures is characteristic of renal pathology. Previous studies have documented the potential value of artificial intelligence (AI) in recognizing relatively clear-cut glomerular structures and patterns, such as segmental or global sclerosis or mesangial hypercellularity. This study aimed to test the capacity of deep learning algorithms to recognize complex glomerular structural changes that reflect common diagnostic dilemmas in nephropathology. METHODS: For this purpose, we defined nine classes of glomerular morphological patterns and trained twelve convolutional neuronal network (CNN) models on these. The two-step training process was done on a first dataset defined by an expert nephropathologist (12,253 images) and a second consensus dataset (11,142 images) defined by three experts in the field. RESULTS: The efficacy of CNN training was evaluated using another set with 180 consensus images, showing convincingly good classification results (kappa-values 0.838-0.938). Furthermore, we elucidated the image areas decisive for CNN-based decision making by class activation maps. Finally, we demonstrated that the algorithm could decipher glomerular disease patterns coinciding in a single glomerulus (e.g. necrosis along with mesangial and endocapillary hypercellularity). CONCLUSIONS: In summary, our model, focusing on glomerular lesions detectable by conventional microscopy, is the first sui generis to deploy deep learning as a reliable and promising tool in recognition of even discrete and/or overlapping morphological changes. Our results provide a stimulus for ongoing projects that integrate further input levels next to morphology (such as immunohistochemistry, electron microscopy, and clinical information) to develop a novel tool applicable for routine diagnostic nephropathology.

Management of Hepatic Sarcoidosis.

BACKGROUND AND AIMS: Liver involvement in sarcoidosis may occur in up to 60% of all patients. As many patients experience only minor symptoms, a high number of undiagnosed cases must be assumed. In order to successfully identify patients with hepatic sarcoidosis, a throughout characterization of these patients and their course of disease is necessary. METHODS: We collected 40 patients from four German centers to evaluate current treatment standards and course of disease. All of our patients underwent liver biopsy with histologically proven granulomatous hepatitis. RESULTS: Detailed characterization of our patients showed an overall benign course of disease. Treatment was very diverse with glucocorticoids for 1 year in 55% (22/40), 5-10 years in 18% (7/40), and permanently in 18% (7/40). Other treatments included disease-modifying anti-rheumatic drugs (DMARDs), the conventional non-biological type in 53% of all patients (of these 81% received azathioprine, 46% metotrexate, 10% hydroxychloroquine, 10% mycophenolate mofetil and 10% cyclophosphamide and biologicals in 8%. Despite these very diverse treatments, patients generally showed slow progression of the disease. Two patients died. None of our patients received a liver transplantation. CONCLUSIONS: Patients received diverse treatments and generally showed slow progression of the disease. Based on our experience, we proposed a diagnostic work up and surveillance strategy as a basis for future, prospective register studies.

Preclinical evaluation of eltrombopag in a PDX model of myelodysplastic syndromes.

Preclinical research of myelodysplastic syndromes (MDSs) is hampered by a lack of feasible disease models. Previously, we have established a robust patient-derived xenograft (PDX) model for MDS. Here we demonstrate for the first time that this model is applicable as a preclinical platform to address pending clinical questions by interrogating the efficacy and safety of the thrombopoietin receptor agonist eltrombopag. Our preclinical study included n = 49 xenografts generated from n = 9 MDS patient samples. Substance efficacy was evidenced by FACS-based human platelet quantification and clonal bone marrow evolution was reconstructed by serial whole-exome sequencing of the PDX samples. In contrast to clinical trials in humans, this experimental setup allowed vehicle- and replicate-controlled analyses on a patient-individual level deciphering substance-specific effects from natural disease progression. We found that eltrombopag effectively stimulated thrombopoiesis in MDS PDX without adversely affecting the patients' clonal composition. In conclusion, our MDS PDX model is a useful tool for testing new therapeutic concepts in MDS preceding clinical trials.

Humanized three-dimensional scaffold xenotransplantation models for myelodysplastic syndromes.

Patient-derived xenograft (PDX) models have emerged as versatile preclinical platforms for investigation of functional pathomechanisms in myelodysplastic syndromes (MDS) and other myeloid neoplasms. However, despite increasingly improved methodology, engraftment efficiencies frequently remain low. Humanized three-dimensional scaffold models (ossicle xenotransplantation models) in immunocompromised mice have recently been found to enable improved engraftment rates of healthy and malignant human hematopoiesis. We therefore interrogated the feasibility of using four different three-dimensional ossicle-based PDX models for application with primary MDS samples. In a fully standardized comparison, we evaluated scaffold materials such as Gelfoam, extracellular matrix (ECM), and human or xenogenous bone substance in comparison to intrafemoral (IF) co-injection of bone marrow (BM)-derived mesenchymal stromal cells (MSCs) and CD34+ hematopoietic stem and progenitor cells (HSPCs). Our study included13 primary MDS patient samples transplanted in parallel according to these five different conditions. Engraftment of MDS samples was assessed by flow cytometry, immunohistological staining, and molecular validation. We determined that three-dimensional ossicle-based methods achieved higher relative rates of engraftment and enabled long-term retrievability of patient-derived MSCs from implanted ossicles. In summary, HSPCs and MSCs derived from MDS BM, which did not significantly engraft in NSG mice after intrafemoral injection, were able to colonize humanized scaffold models. Therefore, these models are promising new xenotransplantation techniques for addressing preclinical and functional questions of the interaction between hematopoiesis and the BM niche in MDS.