Citations

8549 Background: First-line treatment of early stage non-small cell lung cancer (NSCLC) is surgical resection, but with a 5-year survival of only 54%, rates of future relapse are high. Identifying patients at high risk of relapse can help guide adjuvant treatment decisions. Deep convolutional neural network (DCNN) AI models trained on tumor histology have shown incredible flexibility as potential biomarkers, both in lung cancer and more generally across many malignancies. While DCNN models can obtain extremely accurate results when used for routine purposes such as diagnosis, subtyping, and grading of malignancies, most models trained for prognostication or treatment response prediction do not reach performance sufficient for clinical application. Uncertainty quantification (UQ) – a family of techniques that give DCNN models the ability to report confidence alongside predictions – is an underexplored avenue in cancer AI that may help further improve performance and clinical application of models designed to provide clinicians with estimations of risk. Methods: To explore the potential use of UQ in clinically-oriented DCNN models, we trained models on a single-institution, retrospective digital tumor histology slide cohort from patients with Stage I-III NSCLC who underwent surgical resection to predict risk of future relapse. Estimation of uncertainty was performed using dropout as a Bayesian approximation, and uncertainty thresholds were calculated from training sets using a novel method to identify and remove low-confidence predictions. For comparison, a separate multivariate logistic regression was trained in cross-validation using known clinical risk factors. Results: We trained DCNN models on slides from 198 patients (40 relapsed, 158 without relapse). In this cohort, 130 patients had stage I disease (65.6%), 42 had stage II (21.2%), and 26 had stage III (13.2%). The average age was 69 years, 85% were current or previous smokers, and 38 received guideline-concordant adjuvant chemotherapy. Without UQ, a DCNN model predicted risk of future relapse with an average area under receiver operator curve (AUROC) of 0.74 across three-fold cross-validation. Using UQ estimation, 63% of slides were reported with high confidence. Relapse prediction was significantly improved in the high-confidence cohort, with an average AUROC of 0.83 in cross-validation. With a specificity of 70%, this corresponds to an average sensitivity of 86.8% across the three cross-folds (79.1%, 85.5%, and 95.7%). In comparison, a clinical-only multivariate regression model predicted relapse with a cross-validated AUROC of 0.67. Conclusions: This method of uncertainty quantification appears to be a powerful tool to predict lung cancer recurrence risk from digital histopathology while simultaneously providing clinicians with a measurement of algorithm trustworthiness.

4146 Background: Targeting focal adhesion kinase (FAK) renders checkpoint immunotherapy effective in pancreatic ductal adenocarcinoma (PDAC) mouse models. Defactinib is a highly potent oral FAK inhibitor shown to have a tolerable safety profile. We evaluated the safety and recommended phase 2 dose (RP2D) of defactinib in combination with pembrolizumab and gemcitabine for PDAC patients. Methods: We conducted a multicenter, open-label, phase I study with dose escalation and expansion phases. In 3x3 dose escalation, patients with refractory solid tumors were treated at five escalating dose levels of defactinib and gemcitabine to identify a RP2D. In expansion phase, patients with metastatic PDAC who progressed on frontline treatment (refractory cohort) or had treatment response or stable disease (SD) on standard gemcitabine/nab-paclitaxel (maintenance cohort) were treated at RP2D. Pre- and post-treatment tumor biopsies were performed to evaluate changes in tumor immunity. Results: The triple drug combination was well-tolerated with no dose-limiting toxicities. Among 17 treated patients with refractory PDAC, the disease control rate (DCR) was 58.8% with one partial response (PR) and nine SDs and the median progression-free survival (PFS) and overall survival (OS) were 4.2 months and 9.1 months, respectively. Among the evaluable patients in the maintenance cohort, DCR was 63.6% with one PR and six SD. Three patients with SD came off study due to treatment- or disease-related complications. The median PFS and OS were 5.0 months and 8.3 months, respectively. Conclusions: The combination of defactinib, pembrolizumab, and gemcitabine was well-tolerated, had promising preliminary efficacy, and showed increased infiltrative T lymphocytes in post-treatment tumor biopsies. Incorporation of a more potent chemotherapy backbone should be considered to achieve better clinical response in future trial design. Clinical trial information: NCT02546531.

e15567 Background: Radiographic surveillance of colorectal cancer (CRC) after therapy is costly and insensitive. We have reported high cross-sectional performance of methylated DNA markers (MDMs) for detection of primary CRC and metastatic recurrence. This study evaluated MDMs to detect recurrence and measure response to anti-cancer therapy in a longitudinal cohort. Methods: This was a nested case-control study drawn from a prospective cohort of adult patients with CRC who completed definitive treatment and followed forward from the start of surveillance by cross-sectional imaging (CT or MRI) and carcinoembryonic antigen (CEA), collected every 3-6 months under routine care. Blood was collected in LBgard® tubes (Biomatrica, San Diego CA). Cases had recurrent CRC (defined by RECIST v1.1 or radiographic findings that justify re-initiation of anti-cancer therapy as determined by treating clinicians). Controls remained NED (no radiographic recurrence and off therapy for ≥4 surveillance visits). From 6mL of plasma, MDMs ( CNNM1, ANKRD13B, FER1L4, ZNF568, CHST2, ZNF671, VAV3, QKI, GRIN2D, DTX1, PDGFD, SFMBT2, JAM3) were measured by target enrichment long-probe quantitative-amplified signal assays, normalized to B3GALT6, in blinded fashion. Previously published random forest models of MDMs +/- CEA, developed on independent patients with intact primary tumors, were used to develop a summary score. Results: With a median follow-up of 276 days (IQR 187-343) and 3 visits per patient (54 visits total), 18 patients had recurrence after start of surveillance. Another 18 patients remained NED after median 555 days (IRQ 469-662) and 4 surveillance visits per patient (76 visits total). Using a 90th %-ile cut-off summary score from 60 NED patients in a prior cross-sectional measurement, the MDM panel was positive in 15 of 18 patients who recurred and only 2 of the 18 NED patients (Table). Of the 76 patient visits for NED patients, 67 (88%) were below the threshold. The MDM panel score crossed the positivity threshold by a median of 127 days (IRQ 0-233) preceding clinical or radiographic documentation of recurrence. A random forest model with CEA was more sensitive but less specific (Table). In patients who resumed therapy, the MDM panel scores correlated closely with subsequent RECIST scores (not shown). Conclusions: Plasma MDMs rise in anticipation of recurrent CRC prior to radiographic detection and remain low in patients without radiographic disease over serial measurements, warranting further development of this tumor-agnostic liquid biopsy approach as a cost-effective tool to assist cancer surveillance.[Table: see text]

Abstract Glioblastoma (GBM) is a rapidly fatal malignancy typically treated with radiation and temozolomide (TMZ), an alkylating chemotherapeutic. These cytotoxic therapies cause oxidative stress and DNA damage, yielding a senescent-like state of replicative arrest in surviving tumor cells. Unfortunately, recurrence is inevitable and may be driven by surviving tumor cells eventually escaping senescence. A growing number of so-called “senolytic” drugs have been recently identified that are defined by their ability to selectively eliminate senescent cells. A growing inventory of senolytic drugs is under consideration for several diseases associated with aging, inflammation, DNA damage, as well as cancer. Ablation of senescent tumor cells after radiation and chemotherapy could help mitigate recurrence by decreasing the burden of residual tumor cells at risk of recurrence. This strategy has not been previously explored for GBM. We evaluated a panel of 10 previously described senolytic drugs to determine whether any could exhibit selective activity against human GBM persisting after exposure to radiation or TMZ. Three of the 10 drugs have known activity against BCL-XL and preferentially induced apoptosis in radiated or TMZ-treated glioma. This senolytic activity was observed in 12 of 12 human GBM cell lines. Efficacy could not be replicated with BCL-2 inhibition or senolytic agents acting against other putative senolytic targets. Knockdown of BCL-XL decreased survival of radiated GBM cells, whereas knockdown of BCL-2 or BCL-W yielded no senolytic effect. Implications: These findings imply that molecularly heterogeneous GBM lines share selective senescence-induced BCL-XL dependency increase the significance and translational relevance of the senolytic therapy for latent glioma.

Background: Scattered tubular-like cells (STCs) are dedifferentiated renal tubular cells endowed with progenitor-like characteristics to repair injured parenchymal cells. STCs may be damaged and rendered ineffective by renal artery stenosis (RAS), but the underlying processes remain unclear. We hypothesized that RAS alters the epigenetic landscape on DNA and the ensuing gene transcriptional profile of swine STCs. Methods: CD24+/CD133+ STCs were isolated from pig kidneys after 10 weeks of RAS or sham (n = 3 each) and their whole 5-methylcytosine (5mC) and 5-hydroxymethylcytosine (5hmC) profiles were examined by 5mC and 5hmC immunoprecipitation sequencing (MeDIP-/hMeDIP-seq, respectively). A subsequent integrated (MeDIP/hMeDIP-seq/mRNA-seq) analysis was performed by comparing all online available gene sets using Gene Set Enrichment Analysis. Apoptosis, proteolysis, and mitochondrial structure and function were subsequently evaluated in vitro. Results: Differential expression (DE) analysis revealed 239 genes with higher and 236 with lower 5mC levels and 275 genes with higher and 315 with lower 5hmC levels in RAS-STCs compared to Normal-STCs (fold change ≥1.4 or ≤0.7, p ≤ 0.05). Integrated MeDIP-/hMeDIP-seq/mRNA-seq analysis identified several overlapping (DE-5mC/mRNA and DE-5hmC/mRNA levels) genes primarily implicated in apoptosis, proteolysis, and mitochondrial functions. Furthermore, RAS-STCs exhibited decreased apoptosis, mitochondrial matrix density, and ATP production, and increased intracellular amino acid concentration and ubiquitin expression. Conclusions: Renal ischemia induces epigenetic changes in apoptosis-, proteolysis-, and mitochondria-related genes, which correlate with alterations in the transcriptomic profile and corresponding function of swine STCs. These observations may contribute to developing novel targeted interventions to preserve the reparative potency of STCs in renal disease.

Abstract BackgroundMany studies have evaluated bundled interventions to improve hand hygiene compliance (HHC). However, the compliance rates remain low. We aimed to use Kotter's change model (KCM) to improve HHC and conduct a complete process evaluation for HHC among medical staff in the intensive care unit (ICU). MethodsKCM was administrated in the ICU of Zhongnan Hospital of Wuhan University from March 2018 to August 2021, and 41-month longitudinal monitoring of HHC was carried out. The primary outcome was the absolute change in monthly HHC. The secondary outcomes included the characteristics of HHC in different phases, the different change trends of HHC for different HH opportunities and occupations, the quarterly incidence of the central line-associated bloodstream infection (CLABSI), and the catheter-associated urinary tract infection (CAUTI).Results. 20,222 HH actions and 24,195 HH opportunities were included in this study. The overall HHC was 83.58% (95%CI, 83.11%-84.04%). After implementing the KCM project, the HHC increased from 35.71% (95%CI, 22.99%-50.83%) to 87.75% (95%CI, 85.53%-89.67%), with an increased rate of 145.73%. The fastest-growing rate of HHC was after contact patients, which increased from 35.29% to 89.8%. Quarterly trends of the CLABSI rate (0‰-3.53‰) and CAUTI rate (0.96‰-4.26‰) remained stable at a low level under the background of the number of cases and the treatment complexities of the patients increasing year by year.ConclusionsUsing KCM can systematically change the cognition of HH in healthcare providers and create an atmosphere to promote HH to improve and maintain the stability of the HHC of healthcare providers in ICU.

Abstract Purpose: This phase Ib/2 trial investigated pembrolizumab-containing trimodality therapy in patients with gastroesophageal junction (GEJ) adenocarcinoma. Patients and Methods: Patients with GEJ adenocarcinoma (cT1–3NanyM0) received neoadjuvant pembrolizumab-containing chemoradiation (CROSS regimen) followed by surgical resection and adjuvant pembrolizumab. The primary endpoints were tolerability in the first 16 patients and pathologic complete response [pCR (ypT0N0)]. Secondary endpoints included progression-free survival (PFS) and overall survival (OS). An independent propensity-score-matched cohort (treated with CROSS without immunotherapy) was used for comparison. Exploratory analyses included immune biomarkers in the tumor microenvironment (TME) and plasma. Results: We enrolled 31 eligible patients, of whom 29 received all expected doses of neoadjuvant pembrolizumab and 28 underwent R0 resection. Safety endpoints were met. The primary efficacy endpoint was not met [7/31 (22.6%) achieved pCR]. Patients with high [i.e., combined positive score (CPS) ≥ 10] baseline expression of programmed death (PD)-L1 in the TME had a significantly higher pCR rate than those with low expression [50.0% (4/8) vs. 13.6% (3/22); P = 0.046]. Patients with high PD-L1 expression also experienced longer PFS and OS than propensity-score-matched patients. Among trial patients with PD-L1 CPS < 10, unprespecified analysis explored whether extracellular vesicles (EV) could identify further responders: an elevated plasma level of PD-L1–expressing EVs was significantly associated with higher pCR. Conclusions: Adding pembrolizumab to trimodality therapy showed acceptable tolerability but did not meet the pre-specified pCR endpoint. Exploratory analyses suggested that high PD-L1 expression in the TME and/or on EVs may identify patients most likely to achieve tumor response.

AbstractIdentifying biomarkers associated with disease progression and drug resistance are important for personalized care. We investigated the expression of 121 curated genes, related to immunomodulatory drugs (IMiDs) and proteasome inhibitors (PIs) responsiveness. We analyzed 28 human multiple myeloma (MM) cell lines with known drug sensitivities and 130 primary MM patient samples collected at different disease stages, including newly diagnosed (ND), on therapy (OT), and relapsed and refractory (RR, collected within 12 months before the patients’ death) timepoints. Our findings led to the identification of a subset of genes linked to clinical drug resistance, poor survival, and disease progression following combination treatment containing IMIDs and/or PIs. Finally, we built a seven‐gene model (MM‐IMiD and PI sensitivity‐7 genes [IP‐7]) using digital gene expression profiling data that significantly separates ND patients from IMiD‐ and PI‐refractory RR patients. Using this model, we retrospectively analyzed RNA sequcencing (RNAseq) data from the Mulltiple Myeloma Research Foundation (MMRF) CoMMpass (n = 578) and Mayo Clinic MM patient registry (n = 487) to divide patients into probabilities of responder and nonresponder, which subsequently correlated with overall survival, disease stage, and number of prior treatments. Our findings suggest that this model may be useful in predicting acquired resistance to treatments containing IMiDs and/or PIs.

Abstract Clinical question In adults with low density lipoprotein (LDL) cholesterol levels >1.8 mmol/L (>70 mg/dL) who are already taking the maximum dose of statins or are intolerant to statins, should another lipid-lowering drug be added, either a proprotein convertase subtilisin/kexin 9 (PCSK9) inhibitor or ezetimibe, to reduce the risk of major cardiovascular events? If so, which drug is preferred? Having decided to use one, should we add the other lipid-lowering drug? Current practice Most guidelines emphasise LDL cholesterol targets in their recommendations for prescribing PCSK9 inhibitors and/or ezetimibe in adults at high risk of experiencing a major adverse cardiovascular event. However, to achieve these goals in very high risk patients with statins alone is almost impossible, so physicians are increasingly considering other lipid-lowering drugs solely for achieving LDL cholesterol treatment goals rather than for achieving important absolute cardiovascular risk reduction. Most guidelines do not systematically assess the cardiovascular benefits of adding PCSK9 inhibitors and/or ezetimibe for all risk groups across primary and secondary prevention, nor do they report, in accordance with explicit judgments of assumed patients’ values and preferences, absolute benefits and harms and potential treatment burdens. Recommendations The guideline panel provided mostly weak recommendations, which means we rely on shared decision making when applying these recommendations. For adults already using statins, the panel suggests adding a second lipid-lowering drug in people at very high and high cardiovascular risk but recommends against adding it in people at low cardiovascular risk. For adults who are intolerant to statins, the panel recommends using a lipid-lowering drug in people at very high and high cardiovascular risk but against adding it in those at low cardiovascular risk. When choosing to add another lipid-lowering drug, the panel suggests ezetimibe in preference to PCSK9 inhibitors. The panel suggests further adding a PCSK9 inhibitor to ezetimibe for adults already taking statins at very high risk and those at very high and high risk who are intolerant to statins. How this guideline was created An international panel including patients, clinicians, and methodologists produced these recommendations following standards for trustworthy guidelines and using the GRADE approach. The panel identified four risk groups of patients (low, moderate, high, and very high cardiovascular risk) and primarily applied an individual patient perspective in moving from evidence to recommendations, though societal issues were a secondary consideration. The panel considered the balance of benefits and harms and burdens of starting a PCSK9 inhibitor and/or ezetimibe, making assumptions of adults’ average values and preferences. Interactive evidence summaries and decision aids accompany multi-layered recommendations, developed in an online authoring and publication platform ( www.magicapp.org ) that also allows re-use and adaptation. The evidence A linked systematic review and network meta-analysis (14 trials including 83 660 participants) of benefits found that PCSK9 inhibitors or ezetimibe probably reduce myocardial infarctions and stroke in patients with very high and high cardiovascular risk, with no impact on mortality (moderate to high certainty evidence), but not in those with moderate and low cardiovascular risk. PCSK9 inhibitors may have similar effects to ezetimibe on reducing non-fatal myocardial infarction or stroke (low certainty evidence). These relative benefits were consistent, but their absolute magnitude varied based on cardiovascular risk in individual patients (for example, for 1000 people treated with PCSK9 inhibitors in addition to statins over five years, benefits ranged from 2 fewer strokes in the lowest risk to 21 fewer in the highest risk). Two systematic reviews on harms found no important adverse events for these drugs (moderate to high certainty evidence). PCSK9 inhibitors require injections that sometimes result in injection site reactions (best estimate 15 more per 1000 in a 5 year timeframe), representing a burden and harm that may matter to patients. The MATCH-IT decision support tool allows you to interact with the evidence and your patients across the alternative options: https://magicevidence.org/match-it/220504dist-lipid-lowering-drugs/ . Understanding the recommendations The stratification into four cardiovascular risk groups means that, to use the recommendations, physicians need to identify their patient’s risk first. We therefore suggest, specific to various geographical regions, using some reliable risk calculators that estimate patients’ cardiovascular risk based on a mix of known risk factors. The largely weak recommendations concerning the addition of ezetimibe or PCSK9 inhibitors reflect what the panel considered to be a close balance between small reductions in stroke and myocardial infarctions weighed against the burdens and limited harms. Because of the anticipated large variability of patients’ values and preferences, well informed choices warrant shared decision making. Interactive evidence summaries and decision aids linked to the recommendations can facilitate such shared decisions. The strong recommendations against adding another drug in people at low cardiovascular risk reflect what the panel considered to be a burden without important benefits. The strong recommendation for adding either ezetimibe or PCSK9 inhibitors in people at high and very high cardiovascular risk reflect a clear benefit. The panel recognised the key uncertainty in the evidence concerning patient values and preferences, namely that what most people consider important reductions in cardiovascular risks, weighed against burdens and harms, remains unclear. Finally, availability and costs will influence decisions when healthcare systems, clinicians, or people consider adding ezetimibe or PCSK9 inhibitors.

Background It has been reported that metabolic switch from fatty acid oxidation to glycolysis is associated with cardiomyocyte cell cycle induction. However, it is unknown if increased glycolysis and ATP production are indispensable for cardiomyocyte cell cycle induction. Here, we aim to address this question. Methods Cell cycle was induced by overexpression of CCND2 gene or treatment of FGF1 (100 ng/mL) and CHIR99021 (5 µM) on human induced pluripotent stem cells‐derived cardiomyocytes (hiPSC‐CMs). Contraction inhibition was induced by treatment of verapamil (0.5 μM). Cell cycle was evaluated by immunostaining using antibodies against cell cycle markers. Real‐time ATP production and mitochondrial oxidative phosphorylation were determined by Seahorse assay. Metabolic profile was assessed by liquid chromatography–mass spectrometry. Results Treatment of FGF1 and CHIR99021 induces cell cycle of hiPSC‐CMs, reduced carbohydrate metabolism and real‐time ATP production from both mitochondria and cytoplasm. Data from RNA Sequencing showed this treatment increased expression of CCND2 and CDK4 and decreased expression of sarcomeric contractile related genes. Overexpression of CCND2 in hiPSC‐CMs also induces cell cycle, increased aerobic glycolysis (Warburg effect) and no change of real‐time ATP production. Inhibition of contraction via treatment of verapamil resulted in reduced overall metabolism and real‐time ATP production. Conclusion Our data suggest that cardiomyocyte cell cycle may be induced without increased glycolysis or real‐time ATP production.