A cost-effectiveness analysis of ruxolitinib versus best alternative therapy for patients with steroid-refractory chronic graft-versus-host disease aged > 12 years in Singapore.

BACKGROUND: Approximately 30-70% of patients who have undergone allogeneic (allo) hematopoietic stem cell transplantation (HSCT) eventually experience chronic graft-versus-host disease (cGVHD). Patients who develop steroid-refractory (SR)-cGVHD are the most severely impacted due to significant disease and financial burden. There remains an unmet need for safe, efficacious, and accessible treatments for these patients. The objective of this study was to determine the cost effectiveness of ruxolitinib for treatment of SR-cGvHD from the Singapore healthcare system perspective. METHODS: Based on data from the REACH3 randomized open-label trial, a semi-Markov model was developed to evaluate cost-effectiveness of ruxolitinib compared with investigators' choice of best alternative therapy (BAT) for treatment of patients > 12 years of age with SR-cGVHD in Singapore over a 40-year time horizon. The model only considered direct medical-care costs related to the treatment of SR-cGVHD and reported them in Singapore Dollars (SGD). Half-cycle correction was applied to all costs and outcomes, which were discounted at 3%. Probabilistic sensitivity analysis (PSA), one-way sensitivity analysis (OWSA), and scenario analysis were conducted to explore the drivers of uncertainty in the model. RESULTS: In the deterministic base case, more life years (LY; 10.28 vs. 9.42) and quality-adjusted life years (QALYs; 7.31 vs. 6.51) were gained with ruxolitinib than BAT at higher costs (SGD 303,214 vs. SGD 302,673) leading to an incremental cost-effectiveness ratio (ICER) of SGD 677/QALY. At a willingness-to-pay threshold of SGD 75,000/QALY gained, PSA found that ruxolitinib had a 78.52% probability of being cost-effective. Findings were sensitive to variations in non-responder utilities in the BAT arm and duration of BAT treatment in the OWSA, or comparison to either methotrexate (MTX) or mycophenolic acid as a single comparator in the scenario analysis. ICERs remained lower than SGD 75,000/QALY in all other tested variations and scenarios. CONCLUSION: Ruxolitinib is likely to be cost-effective from Singapore healthcare system's perspective for patients with SR-cGVHD, which is promising in the management of patients with unmet clinical needs.

Small Molecule Inhibition of the Autophagy Kinase ULK1 and Identification of ULK1 Substrates.

Many tumors become addicted to autophagy for survival, suggesting inhibition of autophagy as a potential broadly applicable cancer therapy. ULK1/Atg1 is the only serine/threonine kinase in the core autophagy pathway and thus represents an excellent drug target. Despite recent advances in the understanding of ULK1 activation by nutrient deprivation, how ULK1 promotes autophagy remains poorly understood. Here, we screened degenerate peptide libraries to deduce the optimal ULK1 substrate motif and discovered 15 phosphorylation sites in core autophagy proteins that were verified as in vivo ULK1 targets. We utilized these ULK1 substrates to perform a cell-based screen to identify and characterize a potent ULK1 small molecule inhibitor. The compound SBI-0206965 is a highly selective ULK1 kinase inhibitor in vitro and suppressed ULK1-mediated phosphorylation events in cells, regulating autophagy and cell survival. SBI-0206965 greatly synergized with mechanistic target of rapamycin (mTOR) inhibitors to kill tumor cells, providing a strong rationale for their combined use in the clinic.

MicroRNA dynamics in the stages of tumorigenesis correlate with hallmark capabilities of cancer.

While altered expression of microRNAs (miRs) in tumors has been well documented, it remains unclear how the miR transcriptome intersects neoplastic progression. By profiling the miR transcriptome we identified miR expression signatures associated with steps in tumorigenesis and the acquisition of hallmark capabilities in a prototypical mouse model of cancer. Metastases and a rare subset of primary tumors shared a distinct miR signature, implicating a discrete lineage for metastatic tumors. The miR-200 family is strongly down-regulated in metastases and met-like primary tumors, thereby relieving repression of the mesenchymal transcription factor Zeb1, which in turn suppresses E-cadherin. Treatment with a clinically approved angiogenesis inhibitor normalized angiogenic signature miRs in primary tumors, while altering expression of metastatic signature miRs similarly to liver metastases, suggesting their involvement in adaptive resistance to anti-angiogenic therapy via enhanced metastasis. Many of the miR changes associated with specific stages and hallmark capabilities in the mouse model are similarly altered in human tumors, including cognate pancreatic neuroendocrine tumors, implying a generality.

Genetic deletion of the desmosomal component desmoplakin promotes tumor microinvasion in a mouse model of pancreatic neuroendocrine carcinogenesis.

We used the RIP1-Tag2 (RT2) mouse model of islet cell carcinogenesis to profile the transcriptome of pancreatic neuroendocrine tumors (PNET) that were either non-invasive or highly invasive, seeking to identify pro- and anti-invasive molecules. Expression of multiple components of desmosomes, structures that help maintain cellular adhesion, was significantly reduced in invasive carcinomas. Genetic deletion of one of these desmosomal components, desmoplakin, resulted in increased local tumor invasion without affecting tumor growth parameters in RT2 PNETs. Expression of cadherin 1, a component of the adherens junction adhesion complex, was maintained in these tumors despite the genetic deletion of desmoplakin. Our results demonstrate that loss of desmoplakin expression and resultant disruption of desmosomal adhesion can promote increased local tumor invasion independent of adherens junction status.

Polymorphic genetic control of tumor invasion in a mouse model of pancreatic neuroendocrine carcinogenesis.

Cancer is a disease subject to both genetic and environmental influences. In this study, we used the RIP1-Tag2 (RT2) mouse model of islet cell carcinogenesis to identify a genetic locus that influences tumor progression to an invasive growth state. RT2 mice inbred into the C57BL/6 (B6) background develop both noninvasive pancreatic neuroendocrine tumors (PNET) and invasive carcinomas with varying degrees of aggressiveness. In contrast, RT2 mice inbred into the C3HeB/Fe (C3H) background are comparatively resistant to the development of invasive tumors, as are RT2 C3HB6(F1) hybrid mice. Using linkage analysis, we identified a 13-Mb locus on mouse chromosome 17 with significant linkage to the development of highly invasive PNETs. A gene residing in this locus, the anaplastic lymphoma kinase (Alk), was expressed at significantly lower levels in PNETs from invasion-resistant C3H mice compared with invasion-susceptible B6 mice, and pharmacological inhibition of Alk led to reduced tumor invasiveness in RT2 B6 mice. Collectively, our results demonstrate that tumor invasion is subject to polymorphic genetic control and identify Alk as a genetic modifier of invasive tumor growth.

Heterogeneity in the diagnosis and prognosis of ischemic stroke subtypes: 9-year follow-up of 22,000 cases in Chinese adults.

BACKGROUND: Reliable classification of ischemic stroke (IS) etiological subtypes is required in research and clinical practice, but the predictive properties of these subtypes in population studies with incomplete investigations are poorly understood. AIMS: To compare the prognosis of etiologically classified IS subtypes and use machine learning (ML) to classify incompletely investigated IS cases. METHODS: In a 9-year follow-up of a prospective study of 512,726 Chinese adults, 22,216 incident IS cases, confirmed by clinical adjudication of medical records, were assigned subtypes using a modified Causative Classification System for Ischemic Stroke (CCS) (large artery atherosclerosis (LAA), small artery occlusion (SAO), cardioaortic embolism (CE), or undetermined etiology) and classified by CCS as "evident," "probable," or "possible" IS cases. For incompletely investigated IS cases where CCS yielded an undetermined etiology, an ML model was developed to predict IS subtypes from baseline risk factors and screening for cardioaortic sources of embolism. The 5-year risks of subsequent stroke and all-cause mortality (measured using cumulative incidence functions and 1 minus Kaplan-Meier estimates, respectively) for the ML-predicted IS subtypes were compared with etiologically classified IS subtypes. RESULTS: Among 7443 IS subtypes with evident or probable etiology, 66% had SAO, 32% had LAA, and 2% had CE, but proportions of SAO-to-LAA cases varied by regions in China. CE had the highest rates of subsequent stroke and mortality (43.5% and 40.7%), followed by LAA (43.2% and 17.4%) and SAO (38.1% and 11.1%), respectively. ML provided classifications for cases with undetermined etiology and incomplete clinical data (24% of all IS cases; n = 5276), with area under the curves (AUC) of 0.99 (0.99-1.00) for CE, 0.67 (0.64-0.70) for LAA, and 0.70 (0.67-0.73) for SAO for unseen cases. ML-predicted IS subtypes yielded comparable subsequent stroke and all-cause mortality rates to the etiologically classified IS subtypes. CONCLUSION: This study highlighted substantial heterogeneity in prognosis of IS subtypes and utility of ML approaches for classification of IS cases with incomplete clinical investigations.

Development, validation and comparison of multivariable risk scores for prediction of total stroke and stroke types in Chinese adults: a prospective study of 0.5 million adults.

BACKGROUND AND PURPOSE: Low-income and middle-income countries have the greatest stroke burden, yet remain understudied. This study compared the utility of Framingham versus novel risk scores for prediction of total stroke and stroke types in Chinese adults. METHODS: China Kadoorie Biobank (CKB) is a prospective study of 512 726 adults, aged 30-79 years, recruited from 10 areas in China in 2004-2008. By 1 January 2018, 43 234 incident first stroke cases (36 310 ischaemic stroke (IS); 8865 haemorrhagic stroke (HS)) were recorded in 503 842 participants with no history of stroke at baseline. We compared the predictive utility of the Framingham Stroke Risk Profile (FSRP) with novel CKB stroke risk scores and included recalibration, refitting, stratifying by study area and addition of other risk factors. Discrimination was assessed using area under the receiver operating characteristic curve (AUC) and calibration was assessed using Greenwood-Nam-D'Agostino χ2 statistics. RESULTS: Incidence of total stroke varied fivefold by area in China. The FSRP had good discrimination for total stroke (AUC (95% CI); men: 0.78 (0.77 to 0.79), women: 0.77 (95% CI 0.76 to 0.78)), but poor calibration (χ2; men: 1,825, women: 3,053), substantially underestimating absolute risks. Recalibration reduced χ2 by >80%, but did not improve discrimination. Refitting the FSRP did not materially improve discrimination, but further improved calibration. Stratification by area improved discrimination (AUC; men: 0.82 (0.82 to 0.83); women: 0.82 (0.82 to 0.83)), but not calibration. Adding other risk factors yielded modest, but statistically significant, improvements in the AUCs. The findings for IS and HS were similar to those for total stroke. CONCLUSIONS: The FSRP reliably differentiated Chinese adults with incident stroke, but substantially underestimated the absolute risks of stroke. Novel local risk prediction equations that took account of differences in stroke incidence within China enhanced risk prediction of total stroke and major stroke pathological types.

Utility of single versus sequential measurements of risk factors for prediction of stroke in Chinese adults.

Absolute risks of stroke are typically estimated using measurements of cardiovascular disease risk factors recorded at a single visit. However, the comparative utility of single versus sequential risk factor measurements for stroke prediction is unclear. Risk factors were recorded on three separate visits on 13,753 individuals in the prospective China Kadoorie Biobank. All participants were stroke-free at baseline (2004-2008), first resurvey (2008), and second resurvey (2013-2014), and were followed-up for incident cases of first stroke in the 3 years following the second resurvey. To reflect the models currently used in clinical practice, sex-specific Cox models were developed to estimate 3-year risks of stroke using single measurements recorded at second resurvey and were retrospectively applied to risk factor data from previous visits. Temporal trends in the Cox-generated risk estimates from 2004 to 2014 were analyzed using linear mixed effects models. To assess the value of more flexible machine learning approaches and the incorporation of longitudinal data, we developed gradient boosted tree (GBT) models for 3-year prediction of stroke using both single measurements and sequential measurements of risk factor inputs. Overall, Cox-generated estimates for 3-year stroke risk increased by 0.3% per annum in men and 0.2% per annum in women, but varied substantially between individuals. The risk estimates at second resurvey were highly correlated with the annual increase of risk for each individual (men: r = 0.91, women: r = 0.89), and performance of the longitudinal GBT models was comparable with both Cox and GBT models that considered measurements from only a single visit (AUCs: 0.779-0.811 in men, 0.724-0.756 in women). These results provide support for current clinical guidelines, which recommend using risk factor measurements recorded at a single visit for stroke prediction.

Stroke risk prediction using machine learning: a prospective cohort study of 0.5 million Chinese adults.

OBJECTIVE: To compare Cox models, machine learning (ML), and ensemble models combining both approaches, for prediction of stroke risk in a prospective study of Chinese adults. MATERIALS AND METHODS: We evaluated models for stroke risk at varying intervals of follow-up (<9 years, 0-3 years, 3-6 years, 6-9 years) in 503 842 adults without prior history of stroke recruited from 10 areas in China in 2004-2008. Inputs included sociodemographic factors, diet, medical history, physical activity, and physical measurements. We compared discrimination and calibration of Cox regression, logistic regression, support vector machines, random survival forests, gradient boosted trees (GBT), and multilayer perceptrons, benchmarking performance against the 2017 Framingham Stroke Risk Profile. We then developed an ensemble approach to identify individuals at high risk of stroke (>10% predicted 9-yr stroke risk) by selectively applying either a GBT or Cox model based on individual-level characteristics. RESULTS: For 9-yr stroke risk prediction, GBT provided the best discrimination (AUROC: 0.833 in men, 0.836 in women) and calibration, with consistent results in each interval of follow-up. The ensemble approach yielded incrementally higher accuracy (men: 76%, women: 80%), specificity (men: 76%, women: 81%), and positive predictive value (men: 26%, women: 24%) compared to any of the single-model approaches. DISCUSSION AND CONCLUSION: Among several approaches, an ensemble model combining both GBT and Cox models achieved the best performance for identifying individuals at high risk of stroke in a contemporary study of Chinese adults. The results highlight the potential value of expanding the use of ML in clinical practice.

Immunotherapy combinations overcome resistance to bispecific T cell engager treatment in T cell-cold solid tumors.

Therapeutic approaches are needed to promote T cell-mediated destruction of poorly immunogenic, "cold" tumors typically associated with minimal response to immune checkpoint blockade (ICB) therapy. Bispecific T cell engager (BiTE) molecules induce redirected lysis of cancer cells by polyclonal T cells and have demonstrated promising clinical activity against solid tumors in some patients. However, little is understood about the key factors that govern clinical responses to these therapies. Using an immunocompetent mouse model expressing a humanized CD3ε chain (huCD3e mice) and BiTE molecules directed against mouse CD19, mouse CLDN18.2, or human EPCAM antigens, we investigated the pharmacokinetic and pharmacodynamic parameters and immune correlates associated with BiTE efficacy across multiple syngeneic solid-tumor models. These studies demonstrated that pretreatment tumor-associated T cell density is a critical determinant of response to BiTE therapy, identified CD8+ T cells as important targets and mediators of BiTE activity, and revealed an antagonistic role for CD4+ T cells in BiTE efficacy. We also identified therapeutic combinations, including ICB and 4-1BB agonism, that synergized with BiTE treatment in poorly T cell-infiltrated, immunotherapy-refractory tumors. In these models, BiTE efficacy was dependent on local expansion of tumor-associated CD8+ T cells, rather than their recruitment from circulation. Our findings highlight the relative contributions of baseline T cell infiltration, local T cell proliferation, and peripheral T cell trafficking for BiTE molecule-mediated efficacy, identify combination strategies capable of overcoming resistance to BiTE therapy, and have clinical relevance for the development of BiTE and other T cell engager therapies.

How Do Novice Hapticians Design? A Case Study in Creating Haptic Learning Environments.

Access to haptic technology is on the rise, in smartphones, virtual reality gear, and open-source education kits. However, engineers and interaction designers are often inexperienced in designing with haptics, and rarely have tools and guidelines for creating multisensory experiences. To examine the impact of this deficit, we supplied a haptic design kit, custom software, and technical support to nine teams (25 students) for an innovation challenge at a major haptics conference. Teams (predominantly undergraduate engineers with little haptics, interaction design, or education training) designed and built haptic environments to support learning of science topics. Qualitative analysis of surveys, interviews, team blogs, and expert assessments of teams' final demonstrations exposed three themes in these design efforts. 1) Novice teams tended to ignore many of ten design choices that experts navigate, such as explicitly choosing whether haptic and graphic feedback should reinforce versus complement one other. 2) Their design activities differed in timing and inclusion from the ten activities observed in expert process. 3) We identified three success strategies in how teams devised useful and engaging interactions and interpretable multimodal experiences, and communicated about their designs. We compare novice and expert design needs and highlight where future haptic design tools and theory need to support novice practice and training.

The CREB coactivator CRTC2 promotes oncogenesis in LKB1-mutant non-small cell lung cancer.

The LKB1 tumor suppressor is often mutationally inactivated in non-small cell lung cancer (NSCLC). LKB1 phosphorylates and activates members of the AMPK family of Ser/Thr kinases. Within this family, the salt-inducible kinases (SIKs) modulate gene expression in part via the inhibitory phosphorylation of the CRTCs, coactivators for CREB (cAMP response element-binding protein). The loss of LKB1 causes SIK inactivation and the induction of the CRTCs, leading to the up-regulation of CREB target genes. We identified CRTC2 as a critical factor in LKB1-deficient NSCLC. CRTC2 is unphosphorylated and therefore constitutively activated in LKB1-mutant NSCLC, where it promotes tumor growth, in part via the induction of the inhibitor of DNA binding 1 (ID1), a bona fide CREB target gene. As ID1 expression is up-regulated and confers poor prognosis in LKB1-deficient NSCLC, our results suggest that small molecules that inhibit CRTC2 and ID1 activity may provide therapeutic benefit to individuals with NSCLC.

Incomplete inhibition of the Rb tumor suppressor pathway in the context of inactivated p53 is sufficient for pancreatic islet tumorigenesis.

Here, we describe the surprising residual capability of the Rb pathway to negatively regulate proliferation and tumorigenesis in a SV40 large T antigen (Tag)-driven mouse model of pancreatic islet carcinogenesis. Heterogeneous Tag expression during all progression stages suggested that a threshold level of the T antigen oncoprotein might be deterministic for beta-cell hyperproliferation and led us to hypothesize that Tag might not be fully inhibiting the tumor suppressor activity of Rb. Moreover, genomic profiling of these tumors by array CGH pointed to regions of loss on chromosomes 6 and 14, where the Rb pathway inhibitor p27 and Rb itself, respectively, reside. Indeed, genetic ablation of the p27(Kip1) or Rb genes accentuated Tag-induced tumorigenesis, with loss of Rb in particular broadly enhancing multiple parameters of tumorigenesis including the frequency and growth rates of premalignant lesions, of nascent solid tumors, and of invasive carcinomas. The data indicate that attenuation rather than complete inactivation of Rb tumor suppressor gene function, in the context of p53 inhibition, is sufficient to initiate tumorigenesis in this model of islet cell cancer, with the demonstrable possibility that subsequent losses of Rb or its regulators can enhance malignant progression. The results may be relevant to human papillomavirus (HPV)-induced cervical neoplasias where E7 oncogene expression levels or activity (in the case of intermediate/low-risk HPV subtypes) incompletely inhibits Rb tumor suppressor functions, as well as to other neoplasias where initiating oncogenic or tumor suppressor events reduce but do not abrogate Rb function.