Cardiac fibrosis as a predictor for sudden cardiac death after transcatheter aortic valve implantation.

BACKGROUND: Cardiac fibrosis plays a major pathophysiological role in any form of chronic heart disease, and high levels are associated with poor outcome. Diffuse and focal cardiac fibrosis are different subtypes, which have different pathomechanisms and prognostic implications. The total fibrosis burden in endomyocardial biopsy tissue was recently proved to play an independent prognostic role in aortic stenosis patients after transcatheter aortic valve implantation (TAVI). AIMS: Here, for the first time, we aim to assess the specific impact of different fibrosis subtypes on sudden cardiac death (SCD) as a primary reason for cardiovascular mortality after TAVI. METHODS: The fibrosis pattern was assessed histologically in the left ventricular biopsies obtained during TAVI interventions in 161 patients, who received a structured follow-up thereafter. RESULTS: Receiver operating characteristic analyses, performed 6, 12, 24 and 48 months after TAVI, showed diffuse, but not focal, fibrosis as a significant predictor for SCD at all timepoints, with the highest area under the curve at the first time point and a decrease in its SCD predictivity over time. In both multivariate Cox proportional hazards and Fine-Gray competing risk models, including both fibrosis subtypes, as well as age, sex and ejection fraction, high diffuse fibrosis remained statistically significant. Accordingly, it represents an independent SCD predictor, most importantly for the occurrence of early events. CONCLUSIONS: The burden of diffuse cardiac fibrosis plays an important and independent prognostic role regarding SCD early after TAVI. Therefore, the histological evaluation of fibrosis topography has value as a prognostic tool for TAVI patients and may help to tailor individualised approaches to optimise their postinterventional management.

Radiomics and machine learning for renal tumor subtype assessment using multiphase computed tomography in a multicenter setting.

OBJECTIVES: To distinguish histological subtypes of renal tumors using radiomic features and machine learning (ML) based on multiphase computed tomography (CT). MATERIAL AND METHODS: Patients who underwent surgical treatment for renal tumors at two tertiary centers from 2012 to 2022 were included retrospectively. Preoperative arterial (corticomedullary) and venous (nephrogenic) phase CT scans from these centers, as well as from external imaging facilities, were manually segmented, and standardized radiomic features were extracted. Following preprocessing and addressing the class imbalance, a ML algorithm based on extreme gradient boosting trees (XGB) was employed to predict renal tumor subtypes using 10-fold cross-validation. The evaluation was conducted using the multiclass area under the receiver operating characteristic curve (AUC). Algorithms were trained on data from one center and independently tested on data from the other center. RESULTS: The training cohort comprised n = 297 patients (64.3% clear cell renal cell cancer [RCC], 13.5% papillary renal cell carcinoma (pRCC), 7.4% chromophobe RCC, 9.4% oncocytomas, and 5.4% angiomyolipomas (AML)), and the testing cohort n = 121 patients (56.2%/16.5%/3.3%/21.5%/2.5%). The XGB algorithm demonstrated a diagnostic performance of AUC = 0.81/0.64/0.8 for venous/arterial/combined contrast phase CT in the training cohort, and AUC = 0.75/0.67/0.75 in the independent testing cohort. In pairwise comparisons, the lowest diagnostic accuracy was evident for the identification of oncocytomas (AUC = 0.57-0.69), and the highest for the identification of AMLs (AUC = 0.9-0.94) CONCLUSION: Radiomic feature analyses can distinguish renal tumor subtypes on routinely acquired CTs, with oncocytomas being the hardest subtype to identify. CLINICAL RELEVANCE STATEMENT: Radiomic feature analyses yield robust results for renal tumor assessment on routine CTs. Although radiologists routinely rely on arterial phase CT for renal tumor assessment and operative planning, radiomic features derived from arterial phase did not improve the accuracy of renal tumor subtype identification in our cohort.

Stable feature selection utilizing Graph Convolutional Neural Network and Layer-wise Relevance Propagation for biomarker discovery in breast cancer.

High-throughput technologies are becoming increasingly important in discovering prognostic biomarkers and in identifying novel drug targets. With Mammaprint, Oncotype DX, and many other prognostic molecular signatures breast cancer is one of the paradigmatic examples of the utility of high-throughput data to deliver prognostic biomarkers, that can be represented in a form of a rather short gene list. Such gene lists can be obtained as a set of features (genes) that are important for the decisions of a Machine Learning (ML) method applied to high-dimensional gene expression data. Several studies have identified predictive gene lists for patient prognosis in breast cancer, but these lists are unstable and have only a few genes in common. Instability of feature selection impedes biological interpretability: genes that are relevant for cancer pathology should be members of any predictive gene list obtained for the same clinical type of patients. Stability and interpretability of selected features can be improved by including information on molecular networks in ML methods. Graph Convolutional Neural Network (GCNN) is a contemporary deep learning approach applicable to gene expression data structured by a prior knowledge molecular network. Layer-wise Relevance Propagation (LRP) and SHapley Additive exPlanations (SHAP) are methods to explain individual decisions of deep learning models. We used both GCNN+LRP and GCNN+SHAP techniques to construct feature sets by aggregating individual explanations. We suggest a methodology to systematically and quantitatively analyze the stability, the impact on the classification performance, and the interpretability of the selected feature sets. We used this methodology to compare GCNN+LRP to GCNN+SHAP and to more classical ML-based feature selection approaches. Utilizing a large breast cancer gene expression dataset we show that, while feature selection with SHAP is useful in applications where selected features have to be impactful for classification performance, among all studied methods GCNN+LRP delivers the most stable (reproducible) and interpretable gene lists.

Short- and Long-Term Outcomes of Patients with Postoperative Arrhythmia after Liver Surgery.

BACKGROUND: New-onset postoperative arrhythmia (PA) has previously been described as a pivotal risk factor for postoperative morbidity and mortality after visceral surgery. However, there is a lack of data concerning liver surgery. The incidence and impact of new-onset postoperative arrhythmia after liver surgery was, therefore, analyzed in a monocentric study. METHODS: In total, n = 460 patients (221 female, 239 male) who underwent liver surgery between January 2012 and April 2020 without any prior arrhythmia in their medical history were included in this retrospective analysis. Clinical monitoring started with the induction of anesthesia and was terminated with discharge from the intensive care unit (ICU) or intermediate care unit (IMC). Follow-up included documentation of complications during the hospital stay, as well as long-term survival analysis. RESULTS: Postoperative arrhythmia after liver surgery was observed in 25 patients, corresponding to an incidence of 5.4%. The occurrence of arrhythmia was significantly associated with intraoperative complications (p < 0.05), liver fibrosis/cirrhosis (p < 0.05), bile fistula/bile leakage/bilioma (p < 0.05), and organ failure (p < 0.01). Survival analysis showed a significantly poorer overall survival of patients who developed postoperative arrhythmia after liver surgery (p < 0.001). CONCLUSIONS: New-onset postoperative arrhythmia after liver surgery has an incidence of only 5.4% but is significantly associated with higher postoperative morbidity and poorer overall survival.

Postoperative facial appearance of patients with extensive oral squamous cell carcinoma can be adequately preserved with in­house virtually planned mandibular reconstruction.

The present study aimed to assess the concordance of preoperative and postoperative hard and soft tissues in patients with advanced oral squamous cell carcinoma (OSCC) following virtual surgical planning (VSP) mandibular reconstruction. In the present study, a cohort of 32 patients with OSCC underwent in-house VSP, followed by guided mandibular reconstruction utilizing vascularized free tissue grafts sourced from the fibula or scapula. A morphometric analysis was conducted comparing preoperative and postoperative three-dimensional virtual models to evaluate discrepancies and identify potential risk factors associated with poor reconstruction outcomes. The outcome variables were the differences in root mean square (RMS) and mean surface distance (MSD) resulting from the application of an iterative closest point algorithm to the virtual data. The validity of soft tissue comparison data is limited due to its susceptibility to various confounding variables. The present study conducted a comprehensive re-evaluation of these variables. High tumor stage, positive N status and the use of adjuvant therapy contributed to more noticeable differences in preoperative and postoperative facial soft tissue appearance. The accuracy of postoperative bone reconstruction results was higher in patients who underwent neomandibular formation using a fibular graft compared with those who received a scapular graft. Preoperative and postoperative soft tissue analyses were conducted for comparison. The MSD showed a deviation of 3.2 mm (± 2.0 mm SD; range 1.3-9.5 mm), whereas the RMS was 5.3 (± 2.9 SD; range 2.1-14). In conclusion, in-house VSP and guided mandibular reconstructions can yield clinically accurate results, preserving patient appearance and offering the advantage of rapid feasibility.

Combined Biomarker System Predicts Prognosis in Patients with Metastatic Oral Squamous Cell Carcinoma.

BACKGROUND: Metastatic oral squamous cell carcinoma (OSCC) is associated with poor patient prognosis. Metastasis is a complex process involving various proteins, tumor cell alterations, including changes attributable to the epithelial-to-mesenchymal transition (EMT) process, and interactions with the tumor microenvironment (TME). In this study, we investigate a combined protein marker system consisting of connexin 43 (Cx43), EMMPRIN (CD147), E-cadherin, and vimentin, with a focus on their roles in the invasive metastatic progression of OSCC and their potential utility in predicting prognosis. METHODS: We conducted an immunohistochemical analysis to assess the protein expression profiles of Cx43, EMMPRIN, E-cadherin, and vimentin using tissue samples obtained from 24 OSCC patients. The metastatic process was mapped through different regions of interest (ROIs), including adjacent healthy oral mucosa (OM), center of primary OSCC, invasive front (IF), and local cervical lymph node metastases (LNM). The primary clinical endpoints were disease-free survival (DFS) and overall survival (OS). RESULTS: Substantial changes in the expression profiles of the different marker proteins were observed among the different ROIs, with all p-values < 0.05, signifying statistical significance. Multivariable Cox regression analysis results showed a significant effect of increased EMMPRIN expression toward the IF on DFS (p = 0.019) and OS (p = 0.023). Furthermore, the combined predictive analysis showed a significant predictive value of the marker system for DFS (p = 0.0017) and OS (p = 0.00044). CONCLUSIONS: The combined marker system exhibited a significant ability to predict patient prognosis. An increase in EMMPRIN expression toward the IF showed the strongest effect and could be an interesting new antimetastatic therapy approach.

Checkpoint blockade and BRAF/MEK therapy in the therapeutic setting improved the overall survival after sentinel node biopsy: A retrospective study comparing patients with primary care between 1998-2009 and 2010-2017.

Immunotherapies using checkpoint blockade and BRAF/MEK therapies have improved overall survival (OS) in patients with unresectable melanoma metastases. In this retrospective study, we aimed to demonstrate the resulting increase in melanoma-specific survival (MSS) and OS after the excision of primary melanomas (≥1 mm thick) and sentinel lymph node (SN) biopsy (SNB). Using Kaplan-Meier estimates and Cox models, we compared two consecutive cohorts. Patients in cohort 1 (N = 518) underwent SNB between 1998 and 2009, and patients in cohort 2 (N = 460) between 2010 and 2017, when checkpoint blockade and BRAF/(MEK) inhibition became available for the treatment of unresectable relapses. The median follow-up times were 120 and 73 months, respectively. While recurrence-free and distant metastasis-free survival rates remained very similar, MSS and OS increased in favor of cohort 2. The estimated 5-year OS rate of SN-positive patients increased by 14.3% (78.5% vs 64.2%, logrank test: P = .005). The MSS benefit was significant even with low SN tumor burden (metastasis diameter < 1 mm). On multivariate analyses, the risk-reduction in favor of cohort 2 was significant in the total population and in the SN-negative and SN-positive subgroups. In SN-positive patients, besides the availability of modern therapies, SN metastasis diameter and ulceration were independent factors of MSS and OS. Treatment of unresectable melanoma recurrences with modern drug therapies results in significantly higher survival rates in a population with SNB. The survival benefit measured from primary melanoma affects both the SN-positive and SN-negative subpopulations.

Incidence and impact of new-onset postoperative arrhythmia after surgery of the lower gastrointestinal tract.

Postoperative arrhythmias (PAs) are common events and have been widely investigated in cardiothoracic surgery. Within visceral surgery, a recent study revealed a significant occurrence of PA in esophageal resections. In contrast, PA in lower gastrointestinal surgery is rarely investigated and has been rudimentary described in the medical literature. The present work is a retrospective cohort study of 1171 patients who underwent surgery of lower gastrointestinal tract between 2012 and 2018. All included patients were treated and monitored in the intensive care unit (ICU) or intermediate care unit (IMC) after surgery. Follow-up, performed between January and May 2021, was obtained for the patients with PA investigating the possible persistence of PA and complications such as permanent arrhythmia or thromboembolic events after discharge. In total, n = 1171 patients (559 female, 612 male) without any history of prior arrhythmia were analyzed. Overall, PA occurred in n = 56 (4.8%) patients after surgery of the lower GI. The highest incidence of PA was seen in patients undergoing bowel surgery after mesenteric ischaemia (26.92%), followed by cytoreductive surgery (CRS) combined with hyperthermic intraperitoneal chemotherapy (HIPEC; 16.67%). PA was significantly associated with higher age (72 years (IQR 63-78 years) vs. 64 years (IQR 55-73.5 years), p < 0.001) and longer length of stay in the ICU (median 15 days (IQR 5-25 days) vs. median 2 days (IQR 1-5 days), p < 0.001). PA was independently associated with organ failure (OR = 4.62, 95% CI 2.11-10.11, p < 0.001) and higher in-house mortality (OR = 3.37, 95% CI 1.23-9.28, p < 0.001). In median, PA occurred 66.5 h after surgery. In follow-up, 31% of all the patients with PA showed development of permanent arrhythmia. The incidence of PA after lower GI surgery is comparatively low. Its occurrence, however, seems to have severe implications since it is significantly associated with higher rates of organ failure and in-house mortality. Also, compared to the general population, the development of permanent arrhythmia is significantly higher in patients who developed new-onset PA.

Cytogenetic peripheral blood monitoring in azacitidine treated patients with high-risk MDS/sAML: A monocentric real-world experience.

In this single center retrospective analysis 76 patients with high-risk (HR) myelodysplastic syndrome (MDS) treated with azacitidine (AZA) were reviewed for response, especially cytogenetic response (cyR) using repeated chromosome banding analyses (CBA) of bone marrow (bm) metaphases and frequent sequential Fluorescence-in-situ Hybridization (FISH) analyses of immunomagnetically enriched CD34 + circulating peripheral blood cells (CD34 +pb-FISH). In total, 526 CD34 +pb-FISH analyses and 236 CBA were examined. Median observation time was 8.45 months, median number of AZA cycles applied was 8, median overall survival (OS) was 14.9 months, 42.1 % of patients responded to therapy according to IWG criteria: 5 complete response (CR), 0 partial response (PR), 12 bmCR, 15 stable disease with hematologic improvement (HI). HI was reached in 36.8 % of patients, 31.5 % became transfusion-independent. By CBA or CD34 +pb-FISH 20.4 % and 31.6 % of patients showed cyR, respectively. HI rate was significantly higher in cytogenetic responders than in non-responders, but there was no impact on OS or leukemia-free-survival. Cytogenetic responders showed significantly better OS than non-responders. Patients with ≥ 6 AZA cycles had significantly better OS than patients with < 6 cycles applied. Karyotype evolution (KE) as a manifestation of cytogenetic progression was diagnosed in 29.5 % and 17.1 % of patients by CBA and CD34 +pb-FISH, respectively. KE was associated with significantly poorer OS and leukemia-free-survival.

Biomarker profiles in heart failure with preserved vs. reduced ejection fraction: results from the DIAST-CHF study.

AIMS: Chronic heart failure (HF) is a common disease and one of the leading causes of death worldwide. Heart failure with preserved ejection fraction (HFpEF) and with reduced ejection fraction (HFrEF) are different diseases with distinct as well as comparable pathophysiologies and diverse responses to therapeutic agents. We aimed to identify possible pathobiochemical signalling pathways and biomarkers in HFpEF and HFrEF by using a broad proteomic approach. METHODS AND RESULTS: A total of 180 biomarkers in the plasma of a representative subgroup (71 years old) of HFpEF (70% female) with a left ventricular ejection fraction (LVEF) ≥ 50% and HFrEF (18% female) with an LVEF ≤ 40% patients (n = 127) from the Prevalence and Clinical Course of Diastolic Dysfunction and Diastolic Heart Failure (DIAST-CHF) trial were examined and compared with a healthy control group (n = 40; 48% female). We were able to identify 35 proteins that were expressed significantly different in both HF groups compared with the control group. We determine 29 unique proteins expressed in HFpEF and 33 unique proteins in HFrEF. Significantly up-regulated trefoil factor 3 (TFF3) and down-regulated contactin-1 could be identified as previously unknown biomarkers for HF. However, TFF3 is also a predictive factor for the occurrence of a cardiovascular event in HFpEF patients. In HFpEF, serine protease 27 was found at reduced levels for the first time, which could offer a new therapeutic target. Additionally, network analyses showed a special role of platelet-derived growth factor subunit A, Dickkopf-related protein 1, and tumour necrosis factor receptor superfamily member 6 in HFpEF patients, whereas perlecan and junctional adhesion molecule A stood out in the HFrEF group. Overall, signalling pathways of metabolic processes, cellular stress, and iron metabolism seemed to be important for HFrEF, whereas for HFpEF, oxygen stress, haemostasis, cell renewal, cell migration, and cell proliferation are in the foreground. CONCLUSIONS: The identified proteins and signalling pathways offer new therapeutic and diagnostic approaches for patients with chronic HF.

Influence of Shear Stress, Inflammation and BRD4 Inhibition on Human Endothelial Cells: A Holistic Proteomic Approach.

Atherosclerosis is an important risk factor in the development of cardiovascular diseases. In addition to increased plasma lipid concentrations, irregular/oscillatory shear stress and inflammatory processes trigger atherosclerosis. Inhibitors of the transcription modulatory bromo- and extra-terminal domain (BET) protein family (BETi) could offer a possible therapeutic approach due to their epigenetic mechanism and anti-inflammatory properties. In this study, the influence of laminar shear stress, inflammation and BETi treatment on human endothelial cells was investigated using global protein expression profiling by ion mobility separation-enhanced data independent acquisition mass spectrometry (IMS-DIA-MS). For this purpose, primary human umbilical cord derived vascular endothelial cells were treated with TNFα to mimic inflammation and exposed to laminar shear stress in the presence or absence of the BRD4 inhibitor JQ1. IMS-DIA-MS detected over 4037 proteins expressed in endothelial cells. Inflammation, shear stress and BETi led to pronounced changes in protein expression patterns with JQ1 having the greatest effect. To our knowledge, this is the first proteomics study on primary endothelial cells, which provides an extensive database for the effects of shear stress, inflammation and BETi on the endothelial proteome.

Direct proteomic and high-resolution microscopy biopsy analysis identifies distinct ventricular fates in severe aortic stenosis.

The incidence of aortic valve stenosis (AS), the most common reason for aortic valve replacement (AVR), increases with population ageing. While untreated AS is associated with high mortality, different hemodynamic subtypes range from normal left-ventricular function to severe heart failure. However, the molecular nature underlying four different AS subclasses, suggesting vastly different myocardial fates, is unknown. Here, we used direct proteomic analysis of small left-ventricular biopsies to identify unique protein expression profiles and subtype-specific AS mechanisms. Left-ventricular endomyocardial biopsies were harvested from patients during transcatheter AVR, and inclusion criteria were based on echocardiographic diagnosis of severe AS and guideline-defined AS-subtype classification: 1) normal ejection fraction (EF)/high-gradient; 2) low EF/high-gradient; 3) low EF/low-gradient; and 4) paradoxical low-flow/low-gradient AS. Samples from non-failing donor hearts served as control. We analyzed 25 individual left-ventricular biopsies by data-independent acquisition mass spectrometry (DIA-MS), and 26 biopsies by histomorphology and cardiomyocytes by STimulated Emission Depletion (STED) superresolution microscopy. Notably, DIA-MS reliably detected 2273 proteins throughout each individual left-ventricular biopsy, of which 160 proteins showed significant abundance changes between AS-subtype and non-failing samples including the cardiac ryanodine receptor (RyR2). Hierarchical clustering segregated unique proteotypes that identified three hemodynamic AS-subtypes. Additionally, distinct proteotypes were linked with AS-subtype specific differences in cardiomyocyte hypertrophy. Furthermore, superresolution microscopy of immunolabeled biopsy sections showed subcellular RyR2-cluster fragmentation and disruption of the functionally important association with transverse tubules, which occurred specifically in patients with systolic dysfunction and may hence contribute to depressed left-ventricular function in AS.

Incidence, Associated Risk Factors, and Outcomes of Postoperative Arrhythmia After Upper Gastrointestinal Surgery.

Importance: New-onset postoperative arrhythmia, which most often presents as postoperative atrial fibrillation (AF), is a frequent complication in patients undergoing visceral surgery of the upper gastrointestinal tract. Its relevance for patients' outcomes is unknown. Objective: To assess the incidence of arrhythmia after upper gastrointestinal surgery, its risk factors, and its short- and long-term implications for patient outcomes. Design, Setting, and Participants: This cohort study included 1210 patients who underwent surgery of the upper gastrointestinal tract (esophagus, stomach, or pancreas) at the University Medical Center Göttingen in Germany between January 2012 and December 2018. Follow-up was performed between February and May 2020. Patients were excluded if they had a preexisting cardiac arrhythmia or pacemaker. Main Outcomes and Measures: The incidence of atrial fibrillation (AF) was recorded in most cases of postoperative arrhythmia; therefore, the analysis focused on postoperative AF. A multivariable logistic regression model was used to assess associations between surgical complications and postoperative AF occurrence, with odds ratios and 95% CIs reported. Results: A total of 1210 patients (median [IQR] age, 62 [19-90] years; 704 [58.2%] men) were enrolled in this study. Postoperative arrhythmia was recorded in 100 patients (8.3%). Among the different procedures, esophagectomy was associated with the highest incidence of postoperative AF (45.5% in complex esophageal resections and 17.1% in elective thoracoabdominal esophagectomies). The incidence of postoperative AF was associated with prolonged length of stay in the intensive care unit (23.4 days for patients with postoperative AF vs 5.9 days for those without; P < .001). Four factors were associated with the occurrence of postoperative AF: patients' age (OR, 1.06; 95% CI, 1.03-1.08; P < .001), intraoperative surgical complications (OR, 2.47; 95% CI, 1.29-4.74; P = .006), infections (OR, 2.23; 95% CI, 1.31-3.80; P = .003), and organ failure (OR, 4.01; 95% CI, 2.31-6.99; P < .001). In the multivariable analysis, postoperative AF (OR, 7.08; 95% CI, 2.75-18.23; P < .001) and sepsis (OR, 10.98; 95% CI, 3.91-30.81; P < .001) were associated with in-hospital mortality. At a median 19-month follow-up, 20 of 74 patients (27.0%) with postoperative AF developed recurring episodes of arrhythmia after discharge. Conclusions and Relevance: This cohort study found that the postoperative AF was associated with an increased length of stay in the intensive care unit and in-hospital mortality in patients after upper gastrointestinal tract surgery. In addition, postoperative AF was associated with development of permanent or paroxysmal arrhythmia after discharge.

The Influence of Conventional Treatment on Symptoms and Complaints in Patients With Chronic Postsurgical Hypoparathyroidism.

Quality of life (QoL) is impaired in patients with chronic hypoparathyroidism (HypoPT). With a recently developed specific patient questionnaire, the 28-item Hypoparathyroid Patient Questionnaire (HPQ 28), we were able to demonstrate an effect of laboratory parameters on symptoms and complaints identified by scales and items of the HPQ 28. Here, we evaluated the effect of conventional treatment modalities on QoL using this specific questionnaire. In this cross-sectional study, we included 49 HypoPT (41 female and 8 male) patients. Laboratory values of total serum calcium, magnesium, phosphate, calcium-phosphate product (CPP), and 24-hour urine for calcium and phosphate were analyzed. Patients completed the HPQ 28 questionnaire during the corresponding visit. Mean age was 57.3 ± 10.5 years and duration of disease 12.6 ± 9.8 years. Most patients (86%, n = 42) were treated with the active vitamin D analogs calcitriol, alfacalcidol, or dihydrotachysterol (DHT). The use of calcium and magnesium supplements influenced scales on HPQ 28 in a dose-dependent manner. We detected a dose-dependent increase on the HPQ 28 scales "depression and anxiety" and "pain and cramps," and the item "numbness and tingling" related to calcitriol. This effect was independent of gender, age, underlying disease, kind of surgery, serum 25-hydroxyvitamin D3, calcium, or phosphate values. This study presents the first data on specific symptoms of HypoPT patients dependent on different treatment modalities. Our data suggest that in part the reduced QoL in these patients might be caused by conventional treatment. © 2022 The Authors. JBMR Plus published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research.

The Nasolacrimal Drainage System in 143 Children with the Microphthalmos-Anophthalmos Complex. / Tränenwegsbefunde von 143 Kindern mit Mikrophthalmus-Anophthalmus-Komplex.

BACKGROUND: Report of clinical findings relating to the lacrimal system in congenital clinical anophthalmos and severe blind microphthalmos (MAC-complex patients). METHODS: A retrospective study of the notes of 207 consecutive patients treated surgically at least once with highly hydrophilic self-inflating expanders for MAC between 1998 and 2021. The lacrimal drainage system was always probed and irrigated under general anaesthesia before any other procedure was started. RESULTS: 64 patients were excluded due to possible misdiagnosis because of previous lid or orbit surgery elsewhere or due to missing data. The analysis therefore included 67 girls and 76 boys aged between 1 and 126 months (median age: 5 months). 72 patients presented with unilateral and 42 with bilateral anophthalmos, and 24 had unilateral and 5 bilateral microphthalmos; consequently, 286 orbits (of which, 190 with probable pathology) were available for assessment. In unilateral cases the lacrimal system on the normal side was never affected. On the anophthalmic or microphthalmic side the lacrimal system was normal in 68 orbits only (35.8%). The most frequent finding was canalicular stenosis (91 orbits; 48%). Common canaliculus stenosis was observed in 12 orbits (6.3%) and nasolacrimal duct obstruction in 9 orbits (4.7%). There were four cases of punctal aplasia, but no other anomalies. In unilateral MAC pathologic findings during lacrimal probing were found to be associated with anatomical malformation of the contralateral fellow eye. Only in unilateral anophthalmos there was a significant association with cleft lip and palate, which was not found in the three other groups. CONCLUSIONS: In congenital clinical anophthalmos the lacrimal system is affected in up to 66.5% of cases, mostly due to canalicular stenosis. Even if there is no clear evidence of an embryological connection, this association is certainly not a random finding.

Gender and risk-taking behaviors influence the clinical presentation of oral squamous cell carcinoma.

OBJECTIVE: The common risk factors for oral squamous cell carcinoma (OSCC) are smoking and alcohol abuse. A small percentage of patients, mostly women, are demonstrating oral cancer without the common risk behavior. This study investigates how gender and different patterns of lifestyle factors influence the clinical presentation of OSCC. PATIENTS AND METHODS: From this retrospective study, demographical and tumor-specific data and lifestyle factors were analyzed. Statistical analyses were performed using the χ2 test or Fisher's exact test for categorical analysis and the t test, ANOVA test, or Kruskal-Wallis test for continuous variables. The influence of the respective lifestyle factors together with their interactions with the gender on tumor characteristics has been tested using logistic and ordinal cumulative link regression models. RESULTS: Among a total of 308 patients, men represented the majority of smokers (87.2%) and the female cohort were largely non-smokers and non-drinkers (64.9%). For age, tumor site and N-stage it looks like that differences of men and women are driven by the different risk behavior. But if the lifestyle factors are taken into account, we observe contrary effects between men and women for T-, N-, and UICC-stage. For different cancer locations we saw opposite effects with gender and risk profile. These effects are not dose-dependent explainable for gender. CONCLUSION: Some but not all differences in the development of OSCC for men and women are explainable by the respective difference in lifestyle behavior. Some further investigations are necessary to find explanations for the obvious differences between men and women in developing OSCC.

The sentinel node invasion level (SNIL) as a prognostic parameter in melanoma.

Sentinel lymph node (SN) tumor burden is becoming increasingly important and is likely to be included in future N classifications in melanoma. Our aim was to investigate the prognostic significance of melanoma infiltration of various anatomically defined lymph node substructures. This retrospective cohort study included 1250 consecutive patients with SN biopsy. The pathology protocol required description of metastatic infiltration of each of the following lymph node substructures: intracapsular lymph vessels, subcapsular and transverse sinuses, cortex, paracortex, medulla, and capsule. Within the SN with the highest tumor burden, the SN invasion level (SNIL) was defined as follows: SNIL 1 = melanoma cells confined to intracapsular lymph vessels, subcapsular or transverse sinuses; SNIL 2 = melanoma infiltrating the cortex or paracortex; SNIL 3 = melanoma infiltrating the medulla or capsule. We classified 338 SN-positive patients according to the non-metric SNIL. Using Kaplan-Meier estimates and Cox models, recurrence-free survival (RFS), melanoma-specific survival (MSS) and nodal basin recurrence rates were analyzed. The median follow-up time was 75 months. The SNIL divided the SN-positive population into three groups with significantly different RFS, MSS, and nodal basin recurrence probabilities. The MSS of patients with SNIL 1 was virtually identical to that of SN-negative patients, whereas outgrowth of the metastasis from the parenchyma into the fibrous capsule or the medulla of the lymph node indicated a very poor prognosis. Thus, the SNIL may help to better assess the benefit-risk ratio of adjuvant therapies in patients with different SN metastasis patterns.

Explaining decisions of graph convolutional neural networks: patient-specific molecular subnetworks responsible for metastasis prediction in breast cancer.

BACKGROUND: Contemporary deep learning approaches show cutting-edge performance in a variety of complex prediction tasks. Nonetheless, the application of deep learning in healthcare remains limited since deep learning methods are often considered as non-interpretable black-box models. However, the machine learning community made recent elaborations on interpretability methods explaining data point-specific decisions of deep learning techniques. We believe that such explanations can assist the need in personalized precision medicine decisions via explaining patient-specific predictions. METHODS: Layer-wise Relevance Propagation (LRP) is a technique to explain decisions of deep learning methods. It is widely used to interpret Convolutional Neural Networks (CNNs) applied on image data. Recently, CNNs started to extend towards non-Euclidean domains like graphs. Molecular networks are commonly represented as graphs detailing interactions between molecules. Gene expression data can be assigned to the vertices of these graphs. In other words, gene expression data can be structured by utilizing molecular network information as prior knowledge. Graph-CNNs can be applied to structured gene expression data, for example, to predict metastatic events in breast cancer. Therefore, there is a need for explanations showing which part of a molecular network is relevant for predicting an event, e.g., distant metastasis in cancer, for each individual patient. RESULTS: We extended the procedure of LRP to make it available for Graph-CNN and tested its applicability on a large breast cancer dataset. We present Graph Layer-wise Relevance Propagation (GLRP) as a new method to explain the decisions made by Graph-CNNs. We demonstrate a sanity check of the developed GLRP on a hand-written digits dataset and then apply the method on gene expression data. We show that GLRP provides patient-specific molecular subnetworks that largely agree with clinical knowledge and identify common as well as novel, and potentially druggable, drivers of tumor progression. CONCLUSIONS: The developed method could be potentially highly useful on interpreting classification results in the context of different omics data and prior knowledge molecular networks on the individual patient level, as for example in precision medicine approaches or a molecular tumor board.

NOTCH Activation via gp130/STAT3 Signaling Confers Resistance to Chemoradiotherapy.

Resistance of tumor cells to chemoradiotherapy represents a fundamental problem in clinical oncology. The underlying mechanisms are actively debated. Here we show that blocking inflammatory cytokine receptor signaling via STAT3 re-sensitized treatment-refractory cancer cells and abolished tumor growth in a xenograft mouse model when applied together with chemoradiotherapy. STAT3 executed treatment resistance by triggering the expression of RBPJ, the key transcriptional regulator of the NOTCH pathway. The mandatory RBPJ interaction partner, NOTCH intracellular domain, was provided by tumor cell-intrinsic expression of NOTCH ligands that caused tonic NOTCH proteolysis. In fact, NOTCH inhibition phenocopied the effect of blocking STAT3 signaling. Moreover, genetic profiling of rectal cancer patients revealed the importance of the STAT3/NOTCH axis as NOTCH expression correlated with clinical outcome. Our data uncovered an unprecedented signal alliance between inflammation and cellular development that orchestrated resistance to chemoradiotherapy. Clinically, our findings allow for biomarker-driven patient stratification and offer novel treatment options.

The detection of isochromosome i(12p) in malignant germ cell tumours and tumours with somatic malignant transformation by the use of quantitative real-time polymerase chain reaction.

AIMS: Malignant germ cell tumours (GCTs) of the testis are rare neoplasms, but the most common solid malignancies in young men. World Health Organization guidelines divide GCTs into five types, for which numerous immunohistochemical markers allow exact histological subtyping in the majority of cases. In contrast, a germ cell origin is often hard to prove in metastatic GCTs that have developed so-called somatic malignant transformation. A high percentage, up to 89%, of GCTs are characterised by the appearance of isochromosome 12p [i(12p)]. Fluorescence in-situ hybridisation has been the most common diagnostic method for the detection of i(12p) so far, but has the disadvantages of being time-consuming, demanding, and not being a stand-alone method. The aim of the present study was to establish a quantitative real-time polymerase chain reaction assay as an independent method for detecting i(12p) and regional amplifications of the short arm of chromosome 12 by using DNA extracted from formalin-fixed paraffin-embedded tissue. METHODS AND RESULTS: A cut-off value to distinguish between the presence and absence of i(12p) was established in a control set consisting of 36 tumour-free samples. In a training set of 149 GCT samples, i(12p) was detectable in 133 tumours (89%), but not in 16 tumours (11%). In a test set containing 27 primary and metastatic GCTs, all 16 tumours with metastatic spread and/or somatic malignant transformation were successfully identified by the detection of i(12p). CONCLUSION: In summary, the qPCR assay presented here can help to identify, further characterise and assign a large proportion of histologically inconclusive malignancies to a GCT origin.