Prediction model for postoperative delirium risk in elderly hypertensive patients: machine learning-based development and validation

BackgroundPostoperative delirium (POD) is a severe complication in elderly hypertensive patients, associated with poor long-term outcomes. Existing models often rely on intraoperative data, limiting preoperative risk stratification. This study aimed to develop a non-invasive machine learning model to predict POD and investigate its preoperative markers’ impact on three-year mortality.MethodsPreoperative variables were selected using LASSO regression from a cohort of 1,782 patients. Ten machine learning models were trained and validated (7:3 ratio). Model performance was evaluated via AUC-ROC and decision curve analysis (DCA). The optimal model was interpreted using SHAP values. Long-term prognosis within the POD cohort was assessed using Kaplan-Meier curves and multivariable Cox proportional hazards regression.ResultsThe POD incidence was 10.9%. The Gradient Boosting Machine (GBM) demonstrated optimal performance (AUC = 0.868, 95% CI: 0.819–0.917). SHAP analysis identified MMSE score as the most influential predictor, followed by HADS score, age, CFS, frailty, and PSQI score. Multivariable Cox analysis revealed that lower MMSE, alongside elevated HADS, CFS, frailty, and PSQI scores—but not chronological age—were independent predictors of increased three-year mortality in POD patients (all P < 0.05).ConclusionWe developed a robust machine learning tool for individualized POD prediction. Cognitive impairment, psychological distress, frailty, and poor sleep quality serve as critical dual-prognostic markers for both acute POD occurrence and long-term survival. These findings underscore the necessity of routine multidimensional preoperative assessment to facilitate personalized interventions for vulnerable hypertensive populations.

Clinical practice guidelines and quality standards for early intervention in psychosis: an AGREE II appraisal and systematic review of service components

BackgroundEarly intervention in psychosis (EIP) is a key component of youth mental health care, yet recommended models of care for individuals at clinical high risk for psychosis (CHR-P) and first-episode psychosis (FEP) remain heterogeneous across jurisdictions. No previous study has combined a formal AGREE II appraisal with a structured synthesis of core service components for EIP.MethodsWe conducted a systematic review and AGREE II-based appraisal of international clinical practice guidelines (CPGs) and quality standards (QSs). MEDLINE (via PubMed), the Cochrane Library (CENTRAL), Web of Science, guideline repositories, and grey literature sources were searched to March 2025 in accordance with PRISMA 2020 guidance. Eligible documents were CPGs or QSs published in English or Italian from 2005 onward containing recommendations on the organization, assessment, and treatment of CHR-P and/or FEP. Methodological quality was appraised with AGREE II (inter-rater agreement assessed using ICCs); recommendations were extracted, and core components required endorsement by at least one third of relevant documents with at least one strong or mandatory recommendation. Twenty-six documents (24 CPGs and 2 QSs) were included. Methodological quality was moderate overall, with substantial variability across AGREE II domains: scores were generally strongest for scope and purpose, stakeholder involvement, and clarity of presentation, and more variable for Rigour of development, applicability, and editorial independence.ResultsWe identified 32 core recommendations. For CHR-P, the most consistent components were specialized services or dedicated pathways, comprehensive multidisciplinary assessment, cognitive-behavioural therapy, family interventions, and a cautious approach to antipsychotic use. For FEP, the most consistently endorsed components were specialized multidisciplinary teams, assertive and continuous follow-up, family work, psychoeducation, structured pharmacological monitoring, and supported education and employment.ConclusionThese findings provide a guideline-derived framework for youth EIP service delivery and model development, while highlighting the need for more methodologically robust and implementation-oriented guidance across settings.Systematic review registrationhttps://osf.io/cek7u, identifier cek7u.

Reduced serum cortisol, IGF-1, GLP-1, and T3 levels in medication-free children and adolescents with obsessive–compulsive disorder: a case–control study

IntroductionObsessive–compulsive disorder (OCD) frequently emerges during childhood and adolescence and has been associated with alterations in multiple neurobiological systems. Neuroendocrine pathways involved in stress regulation, neurodevelopment, metabolic signaling, and thyroid function interact closely to influence brain circuits implicated in OCD. However, studies simultaneously examining several neuroendocrine parameters in pediatric OCD remain limited.MethodsThis case–control study included 52 medication-free children and adolescents with OCD and 42 healthy controls aged 8–17 years. Psychiatric diagnoses were established using the Schedule for Affective Disorders and Schizophrenia for School-Age Children—Present and Lifetime Version (K-SADS-PL) according to DSM-5 criteria. Obsessive–compulsive symptom severity was assessed using the Children’s Yale–Brown Obsessive Compulsive Scale (CY-BOCS). Anxiety and depressive symptoms were evaluated using the Screen for Child Anxiety Related Emotional Disorders (SCARED) and the Children’s Depression Inventory (CDI), respectively. Serum levels of adrenocorticotropic hormone (ACTH), cortisol, growth hormone (GH), Insulin-like growth factor-1 (IGF-1), Glucagon-like peptide-1 (GLP-1), relaxin-3 (RLN-3), triiodothyronine (T3), and thyroid-stimulating hormone (TSH) were measured using ELISA. Group differences were examined using covariance-based models controlling for age, sex, BMI percentile, anxiety (SCARED), and depressive symptoms (CDI).ResultsInitial analyses showed that children and adolescents with OCD had significantly lower serum levels of cortisol, IGF-1, GLP-1, and T3 compared with healthy controls. These differences remained significant after controlling for age, sex, BMI percentile, and the SCARED and CDI total score, with adjusted analyses showing lower levels of cortisol (p = 0.003), IGF-1 (p = 0.013), GLP-1 (p = 0.017), and T3 (p = 0.015) in the OCD group. Correlation analyses did not reveal significant associations between biochemical parameters and OCD symptom severity after correction for multiple comparisons.DiscussionThese findings suggest that pediatric OCD may be associated with alterations in several neuroendocrine pathways rather than isolated changes within a single hormonal system. Further longitudinal studies are needed to better clarify the biological basis and clinical significance of these findings.

Lilly, Ascidian Launch Up-to-$1.9B RNA Exon Editor Collaboration Targeting Inherited Kidney Diseases

Eli Lilly has expanded its genetic medicines pipeline and capabilities, agreeing to partner with Ascidian Therapeutics to develop RNA exon editors intended to treat inherited kidney diseases, through a collaboration that could generate more than $1.9 billion for the Boston biotech.

The companies have agreed to launch a global research collaboration focused on discovering and developing treatments for undisclosed monogenic kidney diseases, with the option to expand into additional targets.

At the heart of the collaboration are Ascidian’s RNA exon editors, which are designed to repair genetic instructions causing disease. A single RNA exon editor can address multiple mutations spanning multiple exons, resulting in the editing of multiple disease-causing exons simultaneously.

“What that allows us to do is to replace many exons at once, thousands of bases at a time, and use endogenous cellular machinery,” Michael Ehlers, MD, PhD, Ascidian’s president and CEO, told GEN. “We’re not editing letters in the genetic code. We’re rewriting whole chapters at the kilobase scale.

“It’s a very versatile technology to address a fairly wide-ranging set of genetic diseases, and potentially beyond as well,” he added.

Since only exons need to be replaced within the diseased protein, the exon editing payload is small enough to fit in an adeno-associated virus (AAV) or in other viral or nonviral delivery vehicles, including lipid nanoparticles.

Michael Ehlers, MD, PhD, president and CEO of Ascidian Therapeutics [Credit: Kevin Trimmer, courtesy Ascidian Therapeutics]

“The therapeutic is an AAV that expresses the designed RNA exon editor. The AAV infects the target cells. The episome of the AAV dwells in the nucleus and expresses this engineered RNA molecule, our exon editor, that then conducts the trans-splicing,” Ehlers explained. “We very intentionally, across programs and things, really emphasized being a differentiated cargo company in designing these RNAs, and tried where we can to use precedented clinical delivery.”

During the transcription of DNA into RNA, noncoding introns are usually removed, while exons that remain are spliced together to form messenger RNA (mRNA) that can be translated into protein. Mutations result in malformed proteins that cause disease.

Ascidian’s RNA exon editors are designed to bind to target pre-mRNA through what the company calls a highly specific binding domain. The editor molecules are delivered as a DNA construct and transcribed into mutation-free, exon-only RNA, designed with a highly specific binding domain.

Trans-splicing process

Through a binding process called pre-mRNA trans-splicing, exon editors replace disease-causing exons, leading to what Ascidian said is expression of wild-type mRNA and protein at proper levels, in the right cells at the right time.

The exon editors are designed to address large genes or genes with high mutational variance.

“There are other excellent targets which are dominant genes, where maybe you’re uncertain whether the disease-causing mutations are dominant because they’ve got a dominant toxic gain-of-function phenotype, or they’re dominant because it’s a haploinsufficiency. The nice thing about RNA exon editing is that it doesn’t really matter to us. We simply replace the sequence with wild type, so we don’t have to worry about whether you need allele-specific knockdown in those cases, or whether you are at risk of having a haploinsufficiency phenotype,” Ehlers explained.

The exon editing approach, Ascidian said, represents a sea change for RNA therapy, which has seen treatments incorporating antisense oligonucleotides, adenosine deaminase acting on RNA (ADAR)-mediated editing, and mRNA as applied in vaccines such as those for COVID-19, and in in vivo chimeric antigen receptor T-cell (CAR-T) therapies.

“RNA exon editing is really quite different, because you can target and dodge RNAs, and instead of editing one base at a time, you’re changing big swaths of sequence at a time, and unlike a lot of other forms of editing, whether DNA or RNA—not all, but for many—there’s no need to introduce exogenous enzymes or exogenous proteins to be able to conduct the editing,” Ehlers said. “That allows Ascidian to use RNA exon editing to go after a different set of diseases, a different set of disease genes, and potentially a larger patient population.”

In kidney disease, for example, more than 60 genetic diseases are known or suspected to affect the kidneys, with over 3.5 million Americans living with severe inherited kidney disease.

“I would not say all of them are treatable, but a reasonable portion of them are, and we’re prioritizing some that are clearly important, ones where RNA exon editing is particularly uniquely suited, we believe, to addressing the underlying genetic cause,” Ehlers said.

Undisclosed targets

Ascidian has granted Lilly exclusive, target-specific rights to Ascidian’s RNA exon editing technology for undisclosed kidney disease targets. The number of targets was also not disclosed. Ascidian has agreed to lead discovery and specified preclinical activities, with Lilly agreeing to oversee additional preclinical work, clinical development, manufacturing, and commercialization. Ascidian said it retains rights to pursue other kidney-focused targets, independently or with additional partners.

Ehlers said the Lilly collaboration resulted from conversations with the pharma giant held over several years that started with Ascidian discussing its research, and over time became more focused on the companies developing the research plan and programs that they agreed to work on.

“About a year and a half ago, I would say, we kind of got really quite a bit more specific about, well, what would be some interesting areas that might fit with Lilly’s strategic interests and capabilities, where Ascidian’s technology and know-how in RNA biology and splicing biology and using RNA exon editing could apply,” Ehlers recalled.

Genetic medicine expansion

Ascidian’s RNA research and technologies fit with Lilly’s commitment in recent years to expand in genetic medicines. Lilly launched the $700 million Institute for Genetic Medicine in 2021 in Boston’s Fort Point section as part of a strategy of advancing gene and gene editing therapies, and RNA- and other nucleic acid-based therapeutics. The Institute operates in Boston and New York City, where Lilly-owned neuroscience gene therapy developer Prevail Therapeutics is based. Prevail was acquired in 2021 for up to $1.04 billion.

A year later, Lilly expanded an RNA-focused collaboration with ProQR into a potential nearly $4 billion partnership to target disorders of the liver and nervous system by applying ProQR’s RNA-editing Axoimer™ platform. Lilly also snapped up another gene therapy developer, buying hearing loss-focused Akouos for up to $610 million, while last year Lilly purchased gene editing therapy developer Verve Therapeutics for up to $1.3 billion

This year, among its numerous acquisition and collaboration deals, Lilly announced plans in February to buy out circular RNA cell therapy developer Orna Therapeutics for up to $2.4 billion, targeting advancements in cell therapy, and last month acquired nonviral DNA delivery-focused drug developer Engage Biologics for up to $202 million cash.

In its latest collaboration with Ascidian, Lilly agreed to pay the biotech up to $1.9 billion, to consist of an undisclosed upfront payment and payments tied to achieving development and commercial milestones, as well as tiered royalties on global commercial sales.

Targeting Stargardt

Ascidian’s website discloses eight pipeline candidates, the most advanced of which is its sole clinical-phase candidate ACDN-01, a first-in-class RNA exon editing therapy designed to halt the progression of Stargardt disease or other ABCA4 retinopathies by targeting their genetic cause.

ACDN-01 contains a healthy copy of exons of ABCA4 RNA, designed to replace the sections of the ABCA4 RNA that contain mutations, thus creating healthy ABCA4 RNA in the retina. According to Ascidian, this is intended to produce normal ABCA4 protein that can then help clear the eye of toxic waste products. The FDA has granted ACDN-01 its Fast Track and Rare Pediatric Disease Designations.

“I anticipate this will be, by and large, the way that we will deliver exon editors. Although, in principle, it can be agnostic to the delivery method, it’s just whatever it takes to get that RNA exon editor into the cell, into the nucleus, effectively trans-splicing at the target.”

ACDN-01 is under study in the Phase I/II STELLAR trial (NCT06467344), an open-label, single ascending dose clinical study assessing the safety, tolerability, and preliminary efficacy of the treatment candidate when delivered subretinally in participants with ABCA4-related retinopathies. Last month, Ascidian said it had completed the adult dose escalation portion of the STELLAR trial and expanded the study to subjects over 12 years of age.

“We’re working as effectively as we can to get that up and going, so I’d say in the coming months,” Ehlers said when asked about the timing of the expansion.

Ascidian is also conducting an observational prescreening study called STARPATH (NCT06445322), which is designed to identify children ages 5+ and adults with Stargardt who may be eligible for future clinical trials evaluating ACDN-01.

ACDN-01 is among numerous candidates in clinical development to treat Stargardt. These candidates include genetic therapies such as:

  • Ocugen’s OCU410ST (AAV5-hRORA), a modifier gene therapy that uses an adeno-associated virus serotype 5 (AAV5) delivery platform to deliver the RORA (RAR-Related Orphan Receptor A) gene to the retina.
  • SpliceBio’s SB-007, a dual AAV gene therapy designed to restore expression of a functional, full-length ABCA4 protein in the retina through a protein splicing intein platform using two AAV serotype 8 (AAV8) vectors to overcome the size limitations of conventional AAVs, reconstituting biologically active ABCA4 through protein trans-splicing in target photoreceptor cells.
  • VeonGen Therapeutics’ VG801, a dual AAV gene therapy enabled by the company’s vgAAV capsid and vgRNA REVeRT large-gene delivery platform, which delivers the full-length functional ABCA4 gene.

Single vector approach

Ehlers said ACDN-01 would stand out from the other genetic therapies because of its single vector approach to delivery: “I’d say the others, because the nature of the ABCA4 gene, it’s too large just for gene replacement in a single vector. All the others have to use dual vector technologies to try to address that.”

“You can imagine a single vector approach being potentially simpler and more technically feasible,” he added. “But of course, what matters is clinical data, and all these programs will play out in the clinic.”

Also in Ascidian’s pipeline: A second retinal program in lead identification phase; four neuro and neuromuscular programs in phase from lead identification to lead optimization; and two lead identification-phase programs in undisclosed “other areas.” Ascidian is also partnering with Roche to discover and develop RNA exon editing therapeutics against undisclosed neurological targets, through an up-to-$1.842 billion collaboration ($42 million paid initially) launched in 2024.

Following two years in stealth mode, Ascidian was formally launched in 2022 by venture capital firm Apple Tree Partners (ATP), which incubated the company and funded it with $50 million in Series A financing. Ehlers led that incubation as the company’s founding CEO, then chaired the company’s board while Romesh Subramanian, PhD, served as Ascidian’s president and CEO. After Subramanian left the company, Ehlers returned to Ascidian’s helm in 2023, when it closed on $40 million in Series A extension funding from ATP, with the goal of financing the development of ACDN-01 and other pipeline programs.

Since then, Ascidian has grown its workforce to about 40 people, Ehlers said.

“For a clinical stage editing company, to be in the clinic with $90 million in equity financing and have 40 people plus or minus is no small feat, and I think it has been accomplished by having the focus that we’ve had, and having just one of the best scientific teams that I’ve ever had the fortune of working with,” Ehlers said.

“We might have little bits of growth here and there to be able to support this [Lilly] collaboration, to be able to expand the technology, but I’m not anticipating substantial growth,” he added. “We’ve built the company to be efficient and focused, and we’re going to maintain that going forward.”

Lilly’s hot streak

The Ascidian collaboration continues a hot streak of collaboration and acquisition deals for the pharma giant, which is flush with cash from sales of its blockbuster glucagon-like peptide 1 (GLP-1) receptor agonist drug tirzepatide, marketed for type 2 diabetes as Mounjaro® and for obesity as Zepbound®.

So far this year, Lilly has either acquired or is acquiring 10 biotechs, most recently three privately-held developers of vaccines for infectious diseases purchased for a combined up to $3.83 billion cash—Vaccine Company for up to $1.55 billion, Curevo for up to $1.5 billion, and LimmaTech Biologics for up to $780 million.

The Ascidian collaboration is Lilly’s third partnership with a biotech, announced just this week. The other two, totaling a combined amount of up to $4.304 billion-plus, were announced with Asian partners.

Seoul-based Hanmi Pharmaceutical said Lilly had agreed to license from it the rights to develop, manufacture, and commercialize sonefpeglutide (LAPSGLP-2 analog), a Phase II glucagon-like peptide 2 (GLP-2) receptor agonist, worldwide excluding South Korea. Hanmi is now studying sonefpeglutide in a global Phase II trial (NCT04775706) in short bowel syndrome (SBS).

Lilly agreed to pay Hanmi $75 million upfront, and up to an additional $1.185 billion in payments tied to achieving clinical development, regulatory approval, and commercialization milestones, plus royalties on sales following product launch.

Beijing-based Haisco Pharmaceutical Group announced a licensing and research collaboration with Lilly to develop treatments across multiple undisclosed therapeutic areas. Haisco agreed to oversee discovery and identification of up to five “innovative target” programs, while Lilly agreed to lead IND-enabling studies, clinical development, and commercialization. Lilly will obtain exclusive global rights for some programs, as well as exclusive rights worldwide, excluding China, Hong Kong, Macau, and Taiwan, for the other programs.

Lilly agreed to pay Haisco up to $87 million in upfront and near-term payments, up to $2.967 billion in unspecified milestone payments, and single-digit tiered royalties on future product sales.

The post Lilly, Ascidian Launch Up-to-$1.9B RNA Exon Editor Collaboration Targeting Inherited Kidney Diseases appeared first on GEN – Genetic Engineering and Biotechnology News.

Spatial Single‑Cell Platform Reveals Barriers to Antibody Delivery in Solid Tumors

Targeting solid tumors remains one of oncology’s most persistent challenges. Even when a therapeutic antibody is well‑designed, and its molecular target is clear, the drug often struggles to reach its destination inside the dense, heterogeneous architecture of human tumors. Understanding why these agents fail in patients has been a longstanding blind spot in cancer pharmacology.

A new study from Vanderbilt University Medical Center and Stanford University begins to close that gap. In work published in Nature Biotechnology, researchers developed a single-cell spatial pharmacology (SSP) platform, an experimental and analytical system that visualizes drug–tumor interactions directly in human solid tumors. The approach provides a high‑resolution view of drug delivery, target engagement, and the physical barriers that shape therapeutic response.

Eben Rosenthal, MD, the Barry and Amy Baker professor and chair of otolaryngology–head and neck surgery at Vanderbilt Health, is senior author of the paper, titled Single‑cell spatial pharmacobiology identifies conserved stromal barriers to therapeutic antibody delivery in human solid tumors.” Rosenthal and co‑author Guolan Lu, PhD, of Stanford University School of Medicine, developed SSP to quantify how antibody‑based therapies behave once they enter the tumor microenvironment.

Identifying the reason drugs fail in so many cancer patients is a high priority, and SSP can help,” Rosenthal said. “Current pharmacology tools and imaging methodologies do not provide the answers we need to understand which drugs fail due to poor delivery and which ones fail due to insufficient activity upon entering the tumor.

Using SSP, the team found pronounced spatial heterogeneity in both drug delivery and target engagement across head and neck, pancreatic, and other solid tumor types. The data point to a consistent culprit: the stromal architecture, known as the dense, noncancerous tissue surrounding tumors, which acts as a physical barrier that limits antibody penetration.

This approach allows us to examine how the drug distributes within the tumor, the cell types with which it interacts, how strongly it engages its molecular target, and how the architecture of the tumor microenvironment shapes its delivery and activity,” Rosenthal said.

The study included analysis of panitumumab‑IRDye800CW, an antibody used in Phase I trials and which is under investigation for fluorescence‑guided surgery. Rosenthal’s group has long been at the forefront of integrating fluorescence imaging into cancer research and surgical oncology.

“By directly measuring drug delivery at the site of targeted antibody therapy, SSP can distinguish tumor regions that are biologically unresponsive from those that are simply underexposed to the agent. We hope additional study in larger sample sizes of patients can help further validate the application of SSP to identify barriers to drug efficacy, Rosenthal added.

By exposing the physical and biological barriers that shape drug performance in human tissue, the platform offers a path toward designing tools that account for the true complexity of the tumor microenvironment.

The post Spatial Single‑Cell Platform Reveals Barriers to Antibody Delivery in Solid Tumors appeared first on GEN – Genetic Engineering and Biotechnology News.

STAT+: NIH cuts weakened network primed to respond to outbreaks like Ebola

In 2020, the National Institutes of Health funded a network of 10 centers intended to “expand knowledge on re-emerging and emerging infectious diseases.” But when dangerous Ebola and hantaviruses spilled over and caused outbreaks in recent months, those research centers have not been in a position to provide aid. In 2025, the centers’ grants were terminated by the Trump administration as part of cuts that targeted work related to Covid-19 and pandemic preparedness.

The network “has been deemed unsafe for Americans and not a good use of taxpayer funding,” the agency told its grantees in May 2025. Of the $82 million allocated to the program over the course of five years, $14.9 million remained unspent, according to data collected by Grant Witness

While the centers weren’t on the front lines of outbreak responses like the Centers for Disease Control and Prevention or USAID, some researchers involved in the network said the NIH cuts have weakened relationships with experts in other countries that they spent years fostering, with the hopes that it would streamline outbreak responses and the creation of diagnostics and treatments. 

Continue to STAT+ to read the full story…

GAIP ASD Research Study

Conditions: Autism Spectrum Disorder; Autism; ASD; Autism Spectrum Disorder (ASD)

Interventions: Biological: Umbilical Cord Blood-Derived Stem Cells and Exosomes; Drug: Glutathione

Sponsors: Greater Atlanta Integrative Pediatrics

Recruiting