Tag: Medication

Flanked by one of psychedelics’ biggest celebrity cheerleaders, Joe Rogan, and a troupe of MAHA loyalists, President Trump recently signed an executive order aimed at accelerating psychedelic access for clinical research and treatment.

Use of naturally occurring and synthetic hallucinogens traces back to the Neanderthals. Yet these substances have long been a pariah in mainstream medicine, written off as “club drugs” with little to no clinical value (or worse, downright negative effects).

Read the rest…

Background: This study develops an open-source large language model–based chatbot tailored for Korean health consultations. The chatbot was implemented using the retrieval-augmented generation (RAG) technique alongside metadata filtering to enhance its performance. Objective: This study aims to analyze and compare the performance of a RAG-based chatbot with other leading language models in the context of Korean health consultations. Methods: A 10.4 GB Korean medical document corpus (487,277 segments) was constructed from official websites of major Korean hospitals, public health sources, and medical textbooks. This study quantitatively compared 5 open-source large language models (Qwen3:4B, Mistral:7B, Llama-3.1:8B, Gpt-Oss:20B, and Gemma3:27B) in 3 configurations: baseline (model only), RAG-only, and RAG with metadata filtering. The RAG system used a specialized Korean embedding model (upskyy/bge-m3-korean) and an Elasticsearch store. Performance was assessed by an emergency medicine specialist using a validation set of 226 questions across 7 common diseases and scoring responses based on accuracy, safety, and helpfulness. Results: The application of RAG alone failed to yield statistically significant performance improvements and, in some cases (Llama 3.1: 8B and Gemma 3: 27B), resulted in decreased scores. However, the combination of RAG with metadata filtering yielded statistically significant (<.05) performance increases in most models. Notably, the average score for Mistral:7B increased from 3.79, SD 0.08, to 4.10, SD 0.10, and Gpt-Oss:20B increased from 4.43, SD 0.05, to 4.51, SD 0.04, with the latter achieving the highest safety score (4.61, SD 0.03). The Gemma3:27B model, which possessed a high baseline performance (4.42, SD 0.03), was an exception, exhibiting no significant improvement (=.14) even with filtering. Conclusions: The effectiveness of RAG for specialized domains such as Korean medical consultation is highly dependent on a metadata filtering process that controls the quality of retrieved information; simple information augmentation is insufficient. Furthermore, the benefit of RAG is limited when a model’s intrinsic knowledge (eg, Gemma3:27B) already meets or exceeds the quality of the external knowledge base. This finding indicates that performance enhancement strategies must account for both the retrieval mechanism’s quality and the model’s preexisting capabilities.

Researchers at Rutgers University have identified a factor that may help explain why chimeric antigen receptor (CAR) T-cell therapy fails in a majority cancer patients. In a study published in Cell Reports, the investigators found that the poor initial quality of a patient’s harvested CD8+ T cells that are used to manufacture CAR T-cells lack the ability to mount a robust immune response.

“Many of their T-cells are in a defective state called senescence, which means they can’t proliferate in the lab, they can’t migrate to tissue effectively, and they can’t kill very well,” said senor author Ricardo Iván Martínez-Zamudio, PhD, an assistant professor at Rutgers Robert Wood Johnson Medical School.

Building a CAR T-cell therapy depends on successfully harvesting a patient’s own T cells, then modifying to target tumor cells, growing a robust population of these engineered cells in the lab, then reinfusing them into the patient. But, as the new research shows, the efficacy of this process depends on the inherent capacity of the harvested cells to both proliferate in the lab and to retain their immune function.

The Rutgers study showed that in patients where CAR T therapy is ineffective, a large proportion of a patient’s harvested cells are senescent. Their research demonstrated that CD8+ T cells from donors with higher levels of senescence expanded less under standard CAR T culture conditions than cells from donors with lower senescence levels.

Further, a retrospective look at clinical outcomes of published datasets from lymphoma patients treated with CAR T-cell therapy found that patients whose starting cells and final CAR T-cell products had strong senescence signatures were more likely to fail treatment, while those with lower senescent profiles were more likely to respond. This indicated that the state of CD8+ T cells prior to engineering could be influencing the efficacy of CAR T treatments.

To better understand the molecular basis of CD8+ T cell senescence, the researchers collected blood from both younger and older donors, isolated CD8+ T cells, and used a fluorescent marker to identify senescent cells. They then performed multi-omics profiling, including gene expression and chromatin analysis, to map the regulatory networks controlling senescence.

The resulting data showed that T cell senescence, rather than chronological age of the donor, drives most of the molecular differences in CD8+ T cells. “The senescence program is essentially precoded,” Martínez-Zamudio said. “It’s not that older people develop some new dysfunctional program. The capacity is there from the beginning.”

The study identified a number of transcription factors, including AP1, KLF5, and RUNX2, that regulate this dysfunctional program. When the research altered these to effect gene expression patterns in senescent cells, they were able to partially restore aspects of T cell responsiveness. Their ability to proliferate, however, remained limited.

The implications of this research extend beyond cancer therapy. While it is known that senescent CD8+ T cells accumulate with age and contribute to declines in immune function and chronic inflammation, the study also found that senescence gene signatures were enriched in patients with lupus, suggesting this may also play a role to autoimmune diseases.

“Our study defines the gene-regulatory mechanisms underlying human CD8+ T cell senescence, highlights [transcription factor] network perturbation as a viable strategy to manipulate the senescence state, and identifies senescent CD8+ T cell gene signatures as prognostic tools for immunotherapy outcome,” the researchers wrote.

Based on this, the investigators think that T cell senescence profiling could be used to help determine which patients would benefit from CAR T therapy and those that wouldn’t and could help guide alternative treatments. Because the current findings were a retrospective analysis of patient data, the Rutgers team now plan to test this approach in prospective clinical studies through collaborations with Rutgers Cancer Institute.

The study also indicates the potential to improve CAR T-cell therapy by target the senescence program, by altering transcription factor activity to modify gene expression. But restoring the proliferative capability of these cells using this approach will require more research. Another route for improvement suggested by the research is to develop method to reprogram, or selectively eliminate, senescent cells during the CAR T-cell manufacturing process.

The post CAR T-Cell Therapy Failure Linked to Senescent CD8+ T Cells appeared first on Inside Precision Medicine.

Results from a Phase I clinical trial show promise for a CAR T therapy using a population of stem cell memory T cells (TSCM), achieving complete remission at low doses and lowering the risk of serious adverse effects including cytokine-release syndrome. The study was published in Cell today.

“Seeing patients achieve complete responses at doses as low as 250,000 cells per kilogram, without chemotherapy preconditioning, validates years of preclinical work and opens a new chapter in CAR T-cell design,” says Luca Gattinoni, MD, head of the research division for functional immune cell modulation at the Leibniz Institute for Immunotherapy (LIT) and lead author of the study.

While CAR T therapy has significantly improved treatment of blood cancers, many patients still struggle to experience lasting benefits. One potential cause for this is that the infused CAR T cells can often fail to expand and persist within the body over time. To overcome this challenge, Gattinoni’s team treated patients using a TSCM population with strong self-renewal and proliferation capacity. 

“Today’s CAR T-cell products are heterogeneous, and that variability is reflected in the range of clinical responses and toxicity profiles we see in patients,” explains Gabriele Inchingolo, PhD candidate in Gattinoni’s team and a lead author of the study. “To address this, we developed a highly homogeneous CD8+ CAR T-cell product selectively enriched for TSCM cells and compared its performance to conventional CAR T cells.”

The first-in-human trial recruited patients with relapsed or refractory CD19 B-cell malignancies who had previously received a hematopoietic stem cell transplantation (HSCT), a patient population with very limited therapeutic options. Whether they received conventional CAR T cells or a TSCM-enriched product, no patients received chemotherapy preconditioning, which is typically used before infusing CAR T cells to help the cells engraft.

Compared to standard CAR T cells, the CAR TSCM cells showed greater expansion and persistence, allowing them to achieve complete responses even at low doses. 

“We have shown that a more defined, stem-like cell product can perform effectively at lower doses. By employing a highly homogeneous TSCM population, we can potentially achieve more consistent engraftment and persistence, paving the way for more predictable outcomes and more rationally designed clinical trials,” says Gattinoni.

In addition, the TSCM cells showed a more favorable safety profile. Even at expansion levels that caused severe cytokine-release syndrome in patients treated with conventional CAR T cells, patients treated with the TSCM cells only experienced mild side effects.

“We observed less cytokine-release syndrome in this study compared to most other CAR clinical trials that I have participated in,” adds James Kochenderfer, MD, senior investigator at the surgery branch of the National Cancer Institute (NCI). “The TSCM platform yielded higher CAR T-cell levels on a per cell basis—and across many CAR T-cell studies, high blood CAR T-cell levels have been one of the strongest predictors of clinical efficacy.”

While not every patient responded to the CAR TSCM cell therapy, results showed that treatment failure was driven by external factors such as low levels of the target CD19 protein on tumor cells, immunosuppressive signals such as IL-10, and immune responses against the CAR construct. Future studies in larger cohorts will evaluate the addition of chemotherapy preconditioning and CD4 T cells to potentially continue improving patient outcomes, as well as expanding this approach into other forms of cancer, including solid tumors. 

The post Stem Cell Memory CAR T Therapy Proves Effective at Low Doses appeared first on Inside Precision Medicine.