Nancy C. and Craig M. Berge Endowed Chair for the Director of the UA Cancer Center Professor of Urology and Cellular and Molecular Medicine
8:40 AM 9:50 AM Team Science Breakouts
· Immunology · Cancer Engineering · Precision Health · Environmental Impact · IITs and Phase 1
· Community Engagement
9:50 AM 10:10 AM Bridging Care and Discovery: Building Data Systems for Skin Cancer Innovative Research
Clara Curiel-Lewandrowski, MD
Co-Director, UA Skin Cancer Institute Professor, Medicine, College of Medicine (Dermatology) Chief, Division of Dermatology, Department of Medicine Program Director, Dermatology Residency Program
10:10 AM 10:30 AM Immunomodulation Tumor Microenvironment in the Gastric Cancer Patient of Arizona
Junaid Arshad, MD, MS
Assistant Professor & Clinical Scholar College of Medicine
Poster Speaker
10:30 AM 10:50 AM
10:50 AM 11:00 AM
HPV16 productivity vs tumorigenicity in the oropharynx
Robert Jackson
College of Medicine – Tucson, Department of Immunobiology
CTOP POSTER SPEAKER
Machine Learning for Breast Tumor Classification: Insights from Nuclear Morphology
Célia Sahli
College of Pharmacy, Pharmacology and Toxicology
PI: Kenry, PhD
11:00 AM 11:15 AM BREAK
11:15 AM 11:35 AM
Development of a Liquid Biopsy Assay to Detect Cancer in Unexplained Pleural Effusion
Mark Nelson, PhD
Investigator, Center for Toxicology Member of the Graduate Faculty Professor, Cancer Biology - GIDP Professor, Pathology
11:35 AM 11:55 AM Evolving Our Thinking on T Cell Activation and Biomimetic Engineering of Immunotherapies
Mike Kuhns, PhD
Department of Immunobiology
Senior Scientific Advisor, Center for Advanced Molecular and Immunological Therapies
11:55 AM 12:05 PM CBP POSTER SPEAKER
Hypoxia-induced Centrosome Elimination as a Driver of Chromosomal Instability in Prostate Cancer
John Ryniawec
College of Medicine, Cellular and Molecular Medicine PI: Anne Cress, PhD and Gregory Rogers, PhD
12:05 PM 12:20 PM Poster Lightning Round hosted by Jacob Schwartz, PhD
1:50 PM 2:10 PM Architects of Disease: Fusions and Phase Separation in Sarcoma
Jacob C Schwartz, PhD
Associate Director, Cancer Research Training and Education Coordination, UA Cancer Center
Associate Research Professor, Pharmacology
2:10 PM 2:30 PM
Southwest Data Resources and Tools to Advance Research Across the Cancer Continuum
Gloria Coronado, PhD
Associate Director, Population Sciences, UA Cancer Center Department of Epidemiology and Biostatistics, College of Public Health
2:30 PM 2:50 PM Affordable, Non-invasive Microscopy Methods for Aiding Cancer Diagnosis and Treatment
Dongkyun Kang, PhD
Associate Professor, Optical Sciences and Biomedical Engineering
2:50 PM 3:10 PM Water and Metals in the Southwest US: Implications for Human Health
Joseph Hoover, PhD
Assistant Professor, Department of Environmental Science
3:10 PM 3:20 PM CPCP POSTER SPEAKER
Identifying Cardiorespiratory Symptoms in Medical Notes with Open Source Large Language Models
Yunbing Bai
College of Medicine, Arizona Telemedicine Program
PI: Joseph Finkelstein, MD, PhD, FAMIA, FACMI
3:20 PM 3:30 PM Navigating the Shared Resources for Project Efficiency
Gillian D. Paine-Murrieta
Lab Manager III, EMSR Laboratory University of Arizona Cancer Center
3:30 PM 4:00 PM The future of Health in the Age of AI
David Ebert, PhD
Chief AI And Data Science Officer
Associate Vice President of Research and Partnerships
Computer Science Engineering Endowed Innovation Chair, Department of Electrical and Computer Engineering, University of Arizona
4:00 PM 4:05 PM Closing Remarks
4:05 PM 5:00 PM Reception (HSIB Patio South)
UACC Annual Scientific Retreat – Member Input Survey
Your input will guide Cancer Center program development, strategic planning, and CCSG renewal priorities across research, training, community impact, operations, and philanthropy.
QR code contains link to Google Form: or TAP HERE
Welcome and Introduction
8:30 - 8:40am
Dan Theodorescu, MD, PhD
Nancy C. and Craig M. Berge Endowed Chair for the Director of the UA Cancer Center Professor of Urology and Cellular and Molecular Medicine
Welcome to the 2025 Scientific Retreat.
We extend our gratitude to all member researchers, students, community, and industry partners for joining us.
As we look to the future, the U of A Comprehensive Cancer Center is preparing for CCSG renewal, seeking input from program leaders, and aiming to strengthen and build new partnerships. We are collaborating with Banner Health on clinical trials and are excited to be part of the Office of Research and Partnerships, which aligns with the university’s convergent science initiatives.
This year’s retreat themes emphasize these strategic research goals, encouraging us to explore AI-driven healthcare innovation and other cuttingedge approaches.
Although our keynote speaker, NCI interim director Douglas Lowy, cannot attend due to the government shutdown, this has opened new opportunities for our retreat. We invite you to participate in the Team Science Breakout Sessions from 8:40 to 9:50 a.m., where you can explore topics such as immunology, precision health, and community engagement, and potentially discover new collaborative avenues.
Team Science Breakouts
8:40-9:50am
• Immunology
• Precision Health
• IITs and Phase 1
• Cancer Engineering
• Environmental Impact
• Community Engagement
Clinical and Translational Oncology
9:50-10:10am
Clara N Curiel-Lewandrowski, MD
Co-Director, UA Skin Cancer Institute Professor, Medicine, College of Medicine (Dermatology) Chief, Division of Dermatology, Department of Medicine Program Director, Dermatology Residency Program
Bridging Care and Discovery: Building Data Systems for Skin Cancer Innovative Research
10:10-10:30am
Junaid Arshad, MD, MS
Assistant Professor & Clinical Scholar College of Medicine
Immunomodulation Tumor
Microenvironment in the Gastric Cancer Patient of Arizona
10:30 - 10:50am
PI: Koenraad Van Doorslaer, PhD
10:50 - 11:00am
PI: Kenry, PhD
Poster Presenter
Robert Jackson
College of Medicine – Tucson, Department of Immunobiology
HPV16 productivity vs tumorigenicity in the oropharynx
Poster Presenter
Célia Sahli
College of Pharmacy, Pharmacology and Toxicology
Machine Learning for Breast Tumor Classification: Insights from Nuclear Morphology
Break
11:00-11:15am
Cancer Biology
11:15-11:35am
Mark Nelson, PhD
Investigator, Center for Toxicology
Member of the Graduate Faculty Professor, Cancer Biology - GIDP Professor, Pathology
Development of a Liquid Biopsy Assay to Detect Cancer in Unexplained Pleural Effusion
11:15-11:35am
Mike Kuhns, PhD
Department of Immunobiology
Senior Scientific Advisor, Center for Advanced Molecular and Immunological Therapies
Evolving Our Thinking on T Cell Activation and Biomimetic Engineering of Immunotherapies
11:55am - 12:05pm
PI: Anne Cress, PhD and Gregory Rogers, PhD
John Ryniawec
College of Medicine, Cellular and Molecular Medicine
Hypoxia-induced Centrosome Elimination as a Driver of Chromosomal Instability in Prostate Cancer
Poster Lightning Round
hosted by Jacob Schwartz, PhD
12:05pm - 12:20pm
Helen Zukoski-Bartlett
College of Engineering
Biomedical Engineering
PI: Dr. Swarna Ganesh
Brandon Eich
College of Science Psychology
PI: Stephen Adamo
Cancer Diagnostics: Printing Viable Microfluidic Chips Using Volumetric 3D Printing Techniques
Comparing the Subsequent Search Miss Effect Between Mammography and Tomosynthesis
Lunch & Poster Session
12:20pm - 12:50pm
Lunch: Room 306
Poster Session: Forum Floor
1:50 - 2:10pm
Jacob Schwartz, PhD
Associate Director, Cancer Research Training and Education Coordination, UA Cancer Center
Associate Research Professor, Pharmacology
Architects of Disease: Fusions and Phase Separation in Sarcoma
Cancer Prevention
2:10 - 2:30pm
Gloria Coronado, PhD
Associate Director, Population Sciences, UA Cancer Center Department of Epidemiology and Biostatistics, College of Public Health
Southwest Data Resources and Tools to Advance Research Across the Cancer Continuum
2:30 - 2:50pm
Dongkyun Kang, PhD
Associate Professor, Optical Sciences and Biomedical Engineering
2:30 - 2:50pm
Joseph Hoover, PhD
Assistant Professor, Department of Environmental Science
Water and Metals in the Southwest US: Implications for Human Health
11:55am - 12:05pm
PI: Joseph Finkelstein, MD, PhD, FAMIA, FACMI
Poster Presenter
Yunbing Bai
College of Medicine, Arizona Telemedicine Program
Hypoxia-induced Centrosome Elimination as a Driver of Chromosomal Instability in Prostate Cancer
Shared Resources
3:20 - 3:30pm
Gillian D. Paine-Murrieta
Lab Manager III, EMSR Laboratory University of Arizona Cancer Center
Navigating the Shared Resources for Project Efficiency
Special Guest Speaker
3:30- 4:00pm
David Ebert, PhD
Chief AI And Data Science Officer
Associate Vice President of Research and Partnerships
Computer Science Engineering Endowed Innovation Chair, Department of Electrical and Computer Engineering, University of Arizona
The future of Health in the Age of AI
Closing
Remarks
Dan Theodorescu, MD, PhD
Nancy C. and Craig M. Berge Endowed Chair for the Director of the UA Cancer Center
Professor of Urology and Cellular and Molecular Medicine
Reception
4:00 - 5:00pm
HSIB Patio
Ajibola Adelakun - Poster: 23
Pharmacy Pharmaceutical Sciences
Cancer Biology Program
PI(s):Dr.
Jacob C. Schwartz
The Interplay of G-quadruplexes and R-loops in Ewing Sarcoma Cell Survival
The Interplay of G-quadruplexes and R-loops in Ewing Sarcoma Cell Survival
Click to edit Master title style
Ajibola D. Adelakun1,2,4, Megan Zhu1, Haining Zhu3, Jacob C. Schwartz1,4
1 Department of Pharmacology, University of Arizona College of Medicine, Tucson AZ ,
2 Department of Pharmaceutical Sciences, University of Arizona College of Pharmacy, Tucson, AZ,
3 Department of Pharmacology & Toxicology, University of Arizona College of Pharmacy, Tucson, Az,
4 University of Arizona Cancer Center, University of Arizona, Tucson AZ
Introduction and Background
Maintaining genomic stability is fundamental for cellular health and disease prevention, particularly in the context of cancer During transcription and replication, cells encounter various structural challenges within DNA and RNA Notable among these are G-quadruplexes, fourstranded structures formed by guanine-rich sequences, and R-loops, which are RNA-DNA hybrids These non -canonical structures pose significant obstacles to replication and transcription, frequently leading to genomic instability by creating conflicts between these vital cellular processes The resolution of these complex nucleic acid structures is paramount to preserving genomic integrity To explore the implications of these structures in drug discovery, researchers employ modified Ewing sarcoma cell line that incorporates an auxin-inducible degron system, enabling the rapid and reversible depletion of the EWSR1-FLI1 fusion protein, a key driver in Ewing sarcoma Drug screening experiments are designed to assess the effects of various compounds on these cells, both in the presence and absence of EWSR1-FLI1 depletion This strategic approach holds the potential to uncover novel therapeutic strategies for Ewing sarcoma and other diseases linked to genomic instability
Methods
Discussion and Conclusion
Our research illuminates the complex interactions between Gquadruplexes (G4s), R-loops, and specific helicases in maintaining genomic stability within cancer cells We identified four compounds out of 30 that demonstrated significant effects on A673 cells, with enhanced efficacy in cells lacking the EWS-FLI1 fusion protein The differential response based on EWS-FLI1 presence indicates a possible avenue for targeted Ewing sarcoma therapies Our findings emphasize the importance of further investigating the specificity and effectiveness of G4-targeting compounds Additionally, they highlight the potential for exploring combination therapies and applications in various genetic disorders This study lays the groundwork for future research aimed at leveraging these molecular interactions to develop novel therapeutic strategies in cancer and beyond Future Research
Investigate potential synergistic effects between G4-targeting compounds and drugs targeting other cellular processes affected by G4s and R-loops
Funding
This research is funded by the Bi-National Science Foundation (BSF)2021273 to RT and JCS
Figure 2: Schematic depicting AID based degron approach for depletion of endogenous
Figure 3: Drug Screening
Figure 4: Viability test
Figure 1: Formation and resolution of G4s and
Figure 5: Dose response: Ewing cells vs Mesenchyma cells
Figure 6: Dose response: Ewing cells vs Non-Ewing cells
Figure 7: Dose response: Ewing cells vs Human embryonic kidney cells
Alia Starman - Poster: 24
Engineering Biomedical Engineering
Cancer Biology Program
PI(s):Dr.
Alexander McGhee
Granular Environment for Lattice Self-assembly (GELS)
Analea Poppen - Poster: 25
University of Arizona Cellular and Molecular Medicine
Cancer Biology Program
PI(s):Dr. Gregory Rogers
Building a Stable Core: How Cep135 Maintains Centriole Architecture to Prevent Genomic Instability
Anastasia Amoiroglou - Poster: 26
College of Medicine Cellular and Molecular Medicine
Cancer Biology Program
PI(s):Gregory C. Rogers, PhD
Plk4 phosphorylates the Distal Tip Complex to regulate centriole growth during mitosis.
Christina Arnoldy - Poster: 27
College of Medicine Immunobiology
Cancer Biology Program
PI(s):Koenraad Van Doorslaer
RNA Modification of HPV Transcripts Changes Upon Host Keratinocyte Differentiation
Christina Lyons - Poster: 28
University of Arizona Basic Medical Sciences
Cancer Biology Program
PI(s):Dr. Shalini Sharma, PhD
Colin Nelson - Poster: 29
University of Arizona University of Arizona Cancer Center
Cancer Biology Program
PI(s):Dr. Anne Cress
Prostate Cancer Biophysical Phenotype Controls Immunosensitivity Against Gamma Delta T Cells
Diksha Manhas - Poster: 30
College of Pharmacy Department of Pharmacology and Toxicology
Cancer Biology Program
PI(s):Dr. Xinxin Ding
Cytochrome P450-Mediated Naphthalene Bioactivation and DNA Damage in the Brain: Implications for Brain Tumorigenesis
Cytochrome P450-Mediated Naphthalene Bioactivation and DNA Damage in Brain: Implications for Brain Tumorigenesis
➢ Reactive NA metabolites are detected in the brain, including specific adducts of 1,2-NQ with DNA following inhalation or IP NA exposure, and NA-GSH and
1,2-NQ and/or 1,2-NQ-GSH
the brains of liver-Cpr-null mice ➢ Further studies are needed to examine the impact of brain CYP induction on the bioactivation of NA and its potential role in brain tumorigenesis
RESULTS
REFERENCES
Diksha Manhas1, Xiangmeng Wu1, Sarrah Hannon1, Weiguo Han1, Michelle Hollon1, Qing-Yu Zhang1, Hans-Joachim Lehmler2, Laura S. Van Winkle3, Xinxin Ding1
Franziska Kuehner - Poster: 31
Medicine IMB
Cancer Biology Program
PI(s):Koenraad Van Doorslaer
Pokemon goes viral: ZBTB7A in HPV associated cancer
Heng Wu - Poster: 32
College of Science Department of Mathematics; Applied Mathematics Graduate Interdisciplinary Degree Program
PI(s):Heng Wu, Elizabeth S. Borden, Karen T. Hastings, Ryan N. Gutenkunst Cancer Biology Program
What Single-Tumor Sequencing Reveals About Immune Recognition and Killing via Neoantigen Depletion Curves
Noorie Kaur Sandhu - Poster: 33
University of Arizona Cancer Center
Cancer Biology Program
PI(s):Dr.
Anne Cress
Protease Activation of ?6?1 Integrin Primes Cancer Invasion Membrane Responses and Signaling
Joel Parker - Poster: 34
Mel and Enid Zuckerman College of Public Health Epidemiology and Biostatistics
Cancer Biology Program
PI(s):Bonnie LaFleur
Deconvolution methods to guide opportunities and timing to intervene with topical immunoprevention strategies in cSCC
Deconvolution
1.
2.
John Ryniawec - Poster: 35
College of Medicine Cellular and Molecular Medicine
Cancer Biology Program
PI(s):Anne Cress and Gregory Rogers
Hypoxia-induced centrosome elimination as a driver of chromosomal instability in prostate cancer
Natascha Schippel - Poster: 36
College of Medicine - Phoenix Basic Medical Sciences
Cancer Biology Program
PI(s):Dr. Shalini Sharma
Stage-Specific Multiomic Analysis of Human Erythroid Differentiation
Stage-Specific Multiomic Analysis of Human Erythroid Differentiation
Significance:
Understanding the
that regulate Epodependent erythroid differentiation is critical for identifying potential new therapeutic targets that may stimulate erythropoiesis to restore normal RBC production.
Background: Erythropoiesis, the development of erythrocytes from hematopoietic stem cells (HSCs), occurs through four phases: erythroid progenitor development, early erythropoiesis (EE), terminal erythroid differentiation (TED), and maturation. According to the classical model, cells progress from HSCs through EE comprised of erythroid-committed progenitors burst-forming unit-erythroid (BFU-E) and colony-forming unit-erythroid (CFU-E) which then undergo TED and maturation to form functional RBCs. Recent studies have resolved heterogeneity within EE, but investigations into the Epo-dependence of these populations are lacking.
Methods: To dissect the mechanisms underlying these transitions, multiparametric flow cytometry and scRNA-seq were applied to human bone marrow (BM)-derived CD34⁺ hematopoietic stem and progenitor cells cultured ex vivo in the presence or absence of Epo Untargeted LC-MS/MS–based lipidomics was used to assess Epodependent changes in lipid metabolite profiles, while intracellular flow cytometry validated expression of key enzymes involved in glycerophospholipid (GPL) metabolism.
Results: Based on expression of CD34, CD71, and CD105, we developed a strategy to immuno-phenotypically resolve five EE populations: early BFU-E (CD34 CD71loCD105lo), late BFU-E (CD34+CD71hiCD105lo), early CFU-E (CD34–CD71loCD105lo), mid CFU-E (CD34–CD71hi CD105lo), and late CFU-E (CD34–CD71hiCD105hi). These populations were prospectively detected in BM and, when isolated and recultured, continued to differentiate along the erythroid trajectory. Functional assessment of cultures ±Epo revealed that Epo is required for the transition from mid to late CFU-E, i.e. acquisition of the CD71hiCD105hi phenotype. To probe the molecular mechanisms underlying this Epo-dependent switch, we performed scRNA-seq, which 1. confirmed a differentiation arrest preceding this transition in –Epo cultures and 2. identified transcriptional programs associated with Epo signaling. In addition to canonical STAT5, PI3K, and MAPK pathways, transient upregulation of genes involved in lipid and cholesterol metabolism was observed at the Epo-dependent transition. Untargeted lipidomics demonstrated corresponding changes in specific GPL species, consistent with differential expression of GPL metabolism genes, including MBOAT2 LPCAT3 CHPT1, and PEMT Epo-mediated changes in these enzymes were validated at the protein level by intracellular flow cytometry. Discussion: This work establishes a mechanistic framework for studying Epo-dependent human erythroid differentiation, integrating immunophenotypic, transcriptomic, and lipidomic analyses to define five EE subpopulations and their molecular transitions. Through this
Terminal Difluoro-methyl alkynes, with no reactivity towards GSH, as an isosteric replacement in covalent inhibitors of WRN-helicase
1
Paulina Anguiano Garcia - Poster: 38
College of Pharmacy Pharmacology and Toxicology
Cancer Biology Program
PI(s):Xiong Rui
A Covalent DNA-Encoded Library Platform to Accelerate Drug Discovery in Cancer Research.
Paulina
Rafael Sainz - Poster: 39
PI(s):Dr. Anne E. Cress Cancer Biology Program
Niche-Specific Cancer Cell Specialization at Perineural, Perivascular, and Capsular Invasive Fronts: A Spatial Transcriptomics Analysis Cancer Center Cancer Center
Robert Jackson - Poster: 40
College of Medicine - Tucson Department of Immunobiology
Cancer Biology Program
PI(s):Koenraad Van Doorslaer
HPV16 productivity vs tumorigenicity in the oropharynx
Saptarshi Mallick - Poster: 41
College of Medicine Dept of CMM
Cancer Biology Program
PI(s):Anne Cress/Kelvin Pond
Single Cell Spatial Transcriptomics unveils novel cell populations in Prostate Cancer Evolution to Aggressive Disease during Muscle Invasion
Sydney Verdugo - Poster: 42
College of Medicine Immunobiology
PI(s):Justin Wilson Cancer Biology Program
Toll-like Receptor Adaptor WDFY1 provides protection against intestinal inflammation
Toll-Like Receptor Adaptor WDFY1 provides protection against intestinal inflammation
Sydney A.N, Verdugo1, Chloe A. Sairs1, Christina R. Arnoldy1, Dakota M. Reinartz1 and Justin E. Wilson, Ph. D.1 Department of Immunobiology1, Levy Cancer Center, University of Arizona
Figure 1. WDFY1
Figure 3. WDFY1 protects against AOM/DSS induced colitisassociated cancer (CAC)
Figure 4. WDFY1 expression by non-immune cells protects against CAC
University of Arizona College of Medicine - Tucson Medicine
Cancer Biology Program
PI(s):Jennifer Stern, PhD
Therapeutic Potential of Targeting Glucagon Receptor Signaling in Hepatocellular Carcinoma
Uloma Beauty Elvis-Offiah - Poster: 44
Medicine Cancer Biology
Cancer Biology Program
PI(s):Dr.
Juanita L. Merchant
Extracellular signaling blocks Menin suppressor function and nuclear localization through phosphorylation.
1
signaling reverses gastrin suppression by MENIN by inducing its nuclear export
Uloma
Figure 5 EGFR-Induced MENIN Nuclear Translocation and Degradation (A-B) Immunofluorescent images and quantification of FLAGMENIN staining (red) co-stained with DAPI (blue) and Btubulin (green) (C-F) Western blot analysis and quantification of nuclear and cytoplasmic protein extracts AGS
Western blot analysis of kinase pathway activation in AGS cells under the same treatment conditions
V Phosphorylation
Employ
Aim
B. Elvis-Offiah2, Sulaiman Sheriff1, Juanita L Merchant1,2
Emily Kaelin - Poster: 11
College of Medicine - Phoenix Dermatology
Cancer Prevention & Control Program
PI(s):Karen Hastings
Structural features of the neopeptide-MHC complex are associated with immunogenicity in human cancer
Meghan B Skiba - Poster: 19
Nursing Nursing & Health Sciences
Cancer Prevention & Control Program
PI(s):Meghan Skiba
and
Text Message Intervention for Cancer Survivors and Caregivers
n
Yunbing Bai - Poster: 21
College of Medicine Arizona Telemedicine Program
Cancer Prevention & Control Program
PI(s):Joseph Finkelstein
Identifying Cardiorespiratory Symptoms in Medical Notes with OpenSource Large Language Models
Identifying Cardiorespiratory Symptoms in Medical Notes with Open-Source Large Language Models
Yunbing Bai, MA, Joseph Finkelstein, MD, PhD, Arizona Telemedicine Program, the University of Arizona, Tucson, AZ
Introduction
Understanding signs and symptoms (S&S) is vital for the early detection and diagnosis of cancer. These clinical indicators often provide the first evidence of disease progression but are typically embedded within unstructured clinical notes, limiting their use in cancer surveillance and predictive modeling. In particular, cardiorespiratory S&S such as changes in breathing patterns, cough, or chest discomfort can serve as early indicators of lung cancer and related malignancies. Large language models (LLMs) like ChatGPT show potential for extracting S&S but raise data privacy concerns. This study investigates the feasibility of using the locally operable Meta-Llama model to extract cardiorespiratory S&S from clinical notes, enabling secure and scalable applications in cancer informatics.
Result
Performance improved with progressive prompt refinement.
• Instruction-only: High recall but low precision
• +ICD-10 definitions: Marked improvement
• +Assumption-free constraints: Balanced accuracy
• Multi-Agent LLM + Post-Cleaning: Best overall performance Strong precision–recall trade-off and effective false-positive control.
Method
Data Source: 96 clinical notes from MTSamples containing cardiorespiratory-related terms.
Model: Llama 3.3-70B, deployed locally via Ollama for extended-context processing.
Prompt Strategies Tested:
• Instruction-only: Basic task description.
• +ICD-10 definitions: Added code meanings and lists.
• +Assumption-free constraints: Extract explicitly stated S&S only. Multi-Agent LLM (EA + RA): Separate Extraction Agent and Refinement Agent for stable, structured output.
EA extracted explicit S&S; RA mapped outputs to ICD-10 using provided definitions. Post-processing removed false positives (mentioned red-flag words).
Evaluation Metrics: Precision, recall, and F1-score compared to expert annotations.
Extraction Agent A sign is an observation by a medical professional obtained from examination, test results, or a questionnaire that indicates a patient may have a disease. Symptom is a physical or mental experience or observation reported by a patient that may indicate a disease.
TASK: Based on the provided definitions of symptom and sign, extract the patient’s current presented symptoms and signs from the clinical note.
REQUIREMENT:ONLY include explicitly mentioned symptoms and signs. Exclude the denied or negative symptoms and signs. Exclude the signs that are examination results with numerical data.
CLINICAL NOTE: [note here]
Instruction Only
TASK: Extract the patient’s current presented symptoms and signs involving the circulatory and respiratory systems from the clinical note.
REQUIREMENT: Exclude the denied or negative symptoms and signs. List the symptoms and signs, and the ICD-10-CM code
CLINICAL NOTE: [note here] Only include the symptoms and signs and their ICD-10CM code that involve the circulatory and respiratory systems.
ICD-10 Code-Based Symptom Extraction Instruction Only prompt + Only include the symptoms and signs in the following ICD-10-CM list: [List of ICD-10-CM codes with definitions, e.g., R00 Abnormalities of heartbeat]
Assumption-Free Constraints
ICD-10 Code-Based Symptom Extraction prompt + ONLY INCLUDE explicitly mentioned symptoms and signs, do not make assumption
Refinement Agent Prompt
TASK: From the description of symptoms and signs of a patient in the INPUT, pick any symptoms and signs involving the circulatory and respiratory systems.
REQUIREMENT: Only output the original symptoms and signs from the INPUT that can match with ICD-10-CM code in the following list:[List of ICD-10-CM codes with definitions, e.g.,
Discussion & Conclusion
Conclusion • Llama 3.3-70B effectively extracted and coded cardiorespiratory S&S from clinical text. Structured, multi-agent prompting enhanced accuracy and stability.
Multi-agent design mitigated reasoning and formatting overload.
Limitations: • Small dataset (96 notes).
• Focused on a single physiological domain.
Future Work: Expand dataset and clinical domains. Refine prompt orchestration for broader biomedical use.
Reference
1. Hamilton W, Peters TJ, Round A, Sharp D. What are the clinical features of lung cancer before the diagnosis is made? A population-based case-control study. Thorax. 2005;60(12):1059–1065. doi:10.1136/thx.2005.045880.
2. Rivera MP, Mehta AC, Wahidi MM. Establishing the diagnosis of lung cancer: ACCP evidence-based clinical practice guidelines (3rd ed.). Chest. 2013;143(5 Suppl):e142S–e165S. doi:10.1378/chest.12-2353.
3. Touvron H, Lavril T, et al. LLaMA: open and efficient foundation language models. ArXiv 2023;abs/2302.13971.
4. Bai Y, Cui W, Finkelstein J. Performance of OpenSource Large Language Models to Extract Symptoms from Clinical Notes. Stud Health Technol Inform. 2025 Aug 7;329:663-667. doi: 10.3233/SHTI250923. PMID: 40775941.
Zenen Salazar - Poster: 22
College of Science Psychology
Cancer Prevention & Control Program
PI(s):Heidi Hamann, Ph.D.
VIRTUAL
TEACHING & LEARNING
(VTL): Adapting an effective commercial tobacco dependence treatment intervention and prevention education program for virtual platforms
Benjamin Rembetski - Poster: 7
University of Arizona College of Medicine, Tucson Dept of Surgery
Cancer Prevention & Control Program
PI(s):Madhav Chopra
The Banner University Medical Center-Tucson Lung Cancer Screening Program: Four-Year Experience with a Low-Dose CT Lung Cancer Screening Program: Growth, Compliance, and Early Detection
INTRODUCTION:
CONCLUSION:
The implementation of LDCT lung cancer screening was associated with substantial growth of patients screened. The overall adherence was comparable to other decentralized lung cancer screening programs with a rate of 37.3%. This is significantly lower than the 70% adherence rate reported in centralized programs.3 Higher risk patients with index LDCT with Lung-RADS 3,4a,4b,4x had a lower follow up rate; however, these patients were discussed at Lung Tumor Board and referred to our Multidisciplinary Lung Nodule Clinic. Our system’s rate of lung cancer detection from screening is similar when compared to reported averages . These findings support the current model of lung cancer screening as a feasible model for improving access and adherence to lung cancer surveillance. Furthermore, these data support the development of a centralized lung cancer screening program to improve patient adherence and early detection.
OBJECTIVE:
STUDY HYPOTHESIS:
will be the same as other decentralized programs. 2. The percentage of screened nodules that are malignant will be congruent with the national averages as reported the NLST1
METHODS:
RESULTS:
From
Nodule
2024, 927 patients were evaluated and a total of 1186 LDCT scans were completed. Participant characteristics for a representative sample are shown in Table 1. Of the total LDCTs completed, 864 were index LDCT scans to be used as the patient’s baseline and 322 were subsequent follow-up scans for continued surveillance, representing a compliance rate of 37.3%. The number of baseline scans increased as shown in (Table 1). The rate of follow up was lower in the high-risk patients when categorized by Lung Rads (Figure 2). Of the total patients reviewed, 184 patients were discussed at the UACC multidisciplinary tumor board (19.8%), with 16 patients diagnosed with lung cancer (1.7%) (Table 3).
REFERENCES:
1. National Lung Screening Trial Research Team; Aberle DR, Adams AM, Berg CD, Black WC, Clapp JD, Fagerstrom RM, Gareen IF, Gatsonis C, Marcus PM, Sicks JD. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011 Aug 4;365(5):395-409.
2. Kim RY, Rendle KA, Mitra N, Neslund-Dudas C, Greenlee RT, Honda SA, Schapira MM, Simoff MJ, Jeon J, Meza R, Ritzwoller DP, Vachani A. Adherence to Annual Lung Cancer Screening and Rates of Cancer Diagnosis. JAMA Netw Open. 2025 Mar 3;8(3):e250942.
3. Smith HB, Ward R, Frazier C, Angotti J, Tanner NT. Guideline-Recommended Lung Cancer Screening Adherence Is Superior With a Centralized Approach. Chest. 2022 Mar;161(3):818-825.
4. PDQ® Screening and Prevention Editorial Board. PDQ Lung Cancer Screening. Bethesda, MD: National Cancer Institute. Updated <04/17/2025>. Available at: https://www.cancer.gov/types/lung/hp/lung-screening-pdq. Accessed <10/19/2025>. [PMID: 26389268]
5. de Koning HJ, van der Aalst CM, de Jong PA, Scholten ET, Nackaerts K, Heuvelmans MA, Lammers JJ, Weenink C, Yousaf-Khan U, Horeweg N, van ‘t Westeinde S, Prokop M, Mali WP, Mohamed Hoesein FAA, van Ooijen PMA, Aerts JGJV, den Bakker MA, Thunnissen E, Verschakelen J, Vliegenthart R, Walter JE, Ten Haaf K, Groen HJM, Oudkerk M. Reduced Lung-Cancer Mortality with Volume CT Screening in a Randomized Trial. N Engl J Med. 2020 Feb 6;382(6):503-513. doi: 10.1056/NEJMoa1911793. Epub 2020 Jan 29.
Divya Maikhuri - Poster: 10
College of Nursing Nursing
Cancer Prevention & Control Program
PI(s):Thaddeus Wesley Warren Pace
A Dyadic Investigation of Psychological Distress and Cortisol Rhythms in Breast Cancer Survivorship
A Dyadic Investigation of Psychological Distress and Diurnal Cortisol Rhythm in Breast Cancer Survivors and Their Supportive Partners
Background
• Breast cancer survivors and their supportive partners often experience psychological distress (i.e., depression and anxiety).
These challenges are accompanied by disruptions in stress -related physiology, especially diurnal cortisol rhythm (DCR), the typical decline in cortisol from morning to evening. 2
• Emotional well-being is shaped not only by individual factors but also by relationship dynamics: partners’ distress and coping are interdependent. 1
• Although distress has been linked to altered cortisol, few studies have incorporated social dynamics, and even fewer have examined these processes within survivor–partner dyads.
This study examines associations between psychological distress, social connectedness, and DCR metrics within survivor –partner dyads to better understand the biopsychosocial pathways influencing distress in survivorship.
Methods
Primary breast cancer treatments (chemotherapy, radiation, surgery) completed ≥3 months and <5 years ago; partners must cohabitate with survivors.
Dyads completed PROMIS Depression 8a, PROMIS Anxiety 8a, Social Connectedness Scale, Relationship Assessment Scale, and Mutual Psychological Development Questionnaire (mutuality).
Methods, continued
• Dyads (Tabel 1) collected saliva samples at home immediately upon waking, 30 minutes after waking, and at bedtime over two consecutive days. Salivary cortisol concentrations were determined using a commercially available enzyme immunoassay kit (Salimetrics, State College, PA). Cortisol awakening response (CAR) and diurnal slope were computed from averaged values across the two collection days.
• Multiple regressions examined associations between psychological/social variables and cortisol indices, controlling for BMI and participant role.
• Analyses were repeated in subsamples with mild or higher depression or anxiety (PROMIS T -Score ≥ 55).
Results
, continued
Conclusions
•
Results
Figure 4: Associations between social connectedness and depression (left) and anxiety (right), after controlling for group roles in dyad
Table 1: Sample characteristics.
Figure 1: Biopsychosocial model that guides the current research
Figure
(left) and partners (right).
Kseniia Korchagina - Poster: 12
MEZ College of Public Health CB2
Cancer Prevention & Control Program
PI(s):Dr. Shravan Aras
Feruz Oripov1, Kseniia Korchagina1, Marjorie A. Nelson2, Emmanuel Katnasis3,4, Terry A. Badger2,4, Laura B. Oswald5, Rina S. Fox4,6, Shravan Aras1
Liz Olivarez - Poster: 14
College of Nursing Symptoms Health
community Engagement (SHINE) Lab
PI(s):Dr. Terry Badger Cancer Prevention & Control Program
Mariella Rodriguez - Poster: 16
Cancer Prevention & Control Program
PI(s):Adriana Maldonado
BACKGROUND
Prevalence of metabolic dysfunctionassociated
(MASLD) among Mexican-origin adults with overweight and obesity is nearly 50% in Southern Arizona.
• Primary care providers are less likely than hepatologists to be familiar with the American Association of Clinical Endocrinology (AACE) Clinical Practice Guidelines for the Diagnosis and Management of MASLD.
While ample research attributes disparities on MASLD prevalence to individual-level practices, limited work has focused on the role of physician’s consulting behaviors on the diagnosis and management of MASLD.
PURPOSE & OBJECTIVES
• Characterize healthcare providers’ awareness and implementation of the AACE clinical practice guidelines for the diagnosis and management of MASLD. As well as gain a comprehensive understanding of physicians' experiences diagnosing and managing MASLD.
Conduct semi-structured interviews with board-
El Rio Health providers (MD, DO, NP,
•
•
METHODS
Between May 2024 and February 2025, 10 interviews were completed.
Participant Characteristics El Rio Providers (n=10)
Lack
RESULTS
Limitations of Guidelines
“Society guidelines focus on a single medical problem. Particularly in the FQHC setting, I'm managing eight or more active clinical problems in a 20-minute period. I can't have all my focus be on discovering whether somebody has fatty liver disease since I have to manage all the other cancer screenings, chronic, and acute medical problems during that same period.”
• “We go back to the "Oh, we know they have it, but there's nothing we can do.” But like, "Oh, just eat better and exercise more." Well, everybody knows they should do that, but it's not that easy, right? That doesn't translate to real lasting change.”
Openness to Guideline Implementation
“I'm pretty open. Since I started knowing about this, I use them almost every day in my clinical practice trying to see patients who are at risk for NAFLD.”
56.25% of patients screened for MASLD
Diagnosis & Management of MASLD
“I would start by assessing risk factors. There may be something going on if there is elevation in liver enzymes, which would then initiate an investigation of the cause. If no other factors suggest a secondary cause other than MASLD, depending on the severity of the liver enzyme elevation and the likelihood of progression, I would consider imaging or a referral.”
• It is a conversation about lifestyle change, weight loss, and reducing the other influencing factors, particularly usually elevated sugar, elevated insulin.”
Barriers in Diagnosing & Managing of Patients with MASLD
“You need multiple visits a year to have enough real estate to put control of the diabetes and address the liver disease and screening. The frequency of visits and more time in visits to actually address this stuff, that's it.
“I just think of how expensive it is to eat right and the knowledge base and the cultural heritage of food. The challenge of dietary change that is so manyfold and availability and access to therapies that can really make the difference.“
DISCUSSION
While gaps in awareness of the AACE clinical guidelines exist, providers are open to implementing them into their clinical practice to improve diagnosis and management of MASLD.
Improving guidelines readability and useability for practice may be necessary to increase their implementation across a range of healthcare providers.
Reassessing length of visits and directly addressing socioeconomic disparities may further support the diagnosing and management of MASLD within the community.
FUTURE DIRECTION
• At study completion, it is expected to have a better understanding of barriers and facilitators to physicians’ adherence to AACE clinical guidelines which will help develop a collaborative care model to improve liver cancer prevention among populations at high risk for the disease.
Mariella R. Rodriguez, BS1, Adriana Maldonado, PhD,MA1, Victoria Rueda2, Jessica Austin, PhD,MPH3, Melissa Flores, PhD 2, Joy Mockbee, MD,MPH4, Douglas Spegman, MD,MSPH,FACP4, and David O. Garcia, PhD,FACSM1
Brandon Eich - Poster: 8
Science Psychology
Cancer Prevention & Control Program
PI(s):Stephen Adamo
Comparing the Subsequent Search Miss Effect Between Mammography and Tomosynthesis
Comparing the Subsequent Search Miss Effect Between Mammography and Tomosynthesis
Brandon Eich, Lyndon Rakusen, & Stephen Adamo Departments of Psychology and Radiology & Imaging Sciences
Subsequent search miss (SSM; also known as Satisfaction of Search or SOS) effect: a 2nd lesion is more likely to be missed after finding a 1st lesion²
Experienced observers (i.e., radiologists) miss fewer second targets when they are similar²
Digital breast tomosynthesis (DBT) improves breast cancer detection compared to digital mammography (DM) due to searching in depth⁵
- “Expert” searchers (novices searching for a T among Ls) did not exhibit an SSM effect within a 3D segmented search¹ 3 types of misses occur during search⁴:
- Scanning: never fixate the lesion
- Recognition: fixate lesion for less than 1s
- Decision: fixate lesion for more than 1s
What is the extent to which searching with DBT reduces the SSM effect?
Does expertise play a role in search performance across imaging modality?
Can expertise explain the different types of errors made during search?
Mammographers and undergraduate students performed an untimed visual search task with VICTRE generated DM or DBT images³
DMs and DBTs were created with identical architecture using the VICTRE pipeline (i.e., the same “breast” and lesion locations were used to produce both image types).
Analysis: 2 (expertise) x 2 (1st or 2nd lesion) Mixed ANOVA per image type (DM & DBT), with post-hoc comparisons between groups
Behavior:
Novices: SSM occurred in DMs and DBTs
Experts: SSM occurred in DMs butnot in DBTs
Overall detection was better with DBTs, suggesting this imaging modality improves breast cancer detection
Takeaway: DBT and expertise mitigate SSMs
Eye-tracking:
Scanning errors: more common in DBT
Recognition errors: most common error when participants fixated a lesion location
Decision errors: minimal. If searchers fixate a lesion for more than 1s they are likely to identify it
Takeaway: Eye-movements vary by expertise, but general patterns suggest failures in perception
•
Devin Saunders - Poster: 9
Kylie Wilson - Poster: 13
University of Arizona Cancer Center
Cancer Prevention & Control Program
PI(s):Chris Lim
Promoting daily physical activity in Arizona elementary schools, 2021ñ2024: Progress and opportunities
Title: Promoting daily physical activity in Arizona elementary schools, 2021-2024: Progress and opportunities
Background:
• Regular physical activity (PA) supports overall health & reduces cancer risk into adulthood1
• Schools are key settings for fostering youth & community PA
Tracking PA-related policies & practices helps identify gaps and opportunities for improvement
Purpose:
• Examine trends in Arizona (AZ) public elementary school wellness, physical education (PE) recess, and out-ofschool PA policies and practices from 2021–2024
Methods:
Design:
• Repeated cross-sectional online survey of AZ public elementary school PE teachers and administrators
• Stable wellness & recess policies, but low PE frequency & fewer certified PE teachers
• Most schools offer 2+ daily recesses, consistent with ARS 15-118
• Out-of-school
Lyndon Rakusen - Poster: 15
University of Arizona Psychology
Cancer Prevention & Control Program
PI(s):Stephen Adamo
Explainable AI Heatmaps and Mammography Screening
Marjorie Nelson - Poster: 17
Nursing Division of Nursing and Health Sciences
Cancer Prevention & Control Program
PI(s):Dr. Rina Fox
Comparing 6-sulfatoxymelatonin collected from liquid and dried urine samples: A proof-of-concept study
Mary Hadeed - Poster: 18
Public Health and Nursing Public Health (HPS) and Nursing
Cancer Prevention & Control Program
PI(s):Terry Badger
Understanding Factors Influencing Engagement in a Symptom Management Intervention After Cancer Treatment: The Role of Self-efficacy
Understanding factors influencing engagement in a symptom management intervention after cancer treatment: The role of self-efficacy
Background and Purpose
Cancer survivorship begins the moment a person is diagnosed and continues throughout their life.1 By the year 2040, the number of people living with a cancer diagnosis will rise to 26 million.2 More people are surviving cancer, yet the physical, emotional, social, and psychological effects continue long after treatment.3
Symptom management as supportive care4:
• Includes physical, emotional, behavioral, and lifestyle strategies
• Reduces symptom severity
• Improves quality of life
1. Who engages in symptom management and why?
Gap in understanding the personal, behavioral, and environmental factors that may facilitate or impeded a survivor from engaging in symptom management strategies after cancer treatment.
2. Is self-efficacy related to engagement?
• Confidence to manage one’s symptoms can reduce symptom burden and increase quality of life.
Methods
Aim: Determine the role of self-efficacy and other predictors on engagement in symptom self-management interventions.
Study Design & Sample:
• Secondary analysis of SMART trial data (N=375) 5
• 45% Hispanic; 84% female; mean age 57.6
• Post-chemo survivors in high-need of symptom management randomized to one of two evidence-based interventions:
Symptom Management and Survivorship Handbook (SMSH)
A health professional called survivors weekly for 12 weeks, assessed symptoms, and referred them to specific chapters and strategies.
Results and Implications
Telephone Interpersonal Counseling (TIPC)
A Master’s prepared social worker called survivors weekly for 12 weeks to address interpersonal communication, relationships, and social support.
Weeks 1-4 = First Randomization (all high-need survivors)
• Weeks 5-12= Nonresponders re-randomized
Measures & Analysis:
Primary Outcome: Engagement measured by:
1. Weekly session completion 2. Weekly use of SMSH strategies
Analysis: Generalized Linear Mixed Models (GLMM) adjusted for SCT-based covariates
Interaction Tested: Self-efficacy x Study group
• Early engagement may be driven by self-efficacy; symptom burden reduced ongoing participation.
• Intervention design should address physical and psychological barriers in tandem with early additional support.
Acknowledgement: This study was supported by the National Institutes of Health (R01 CA225615), MPI Badger T, Sikorskii A. Data collection by the Behavioral Intervention and Measurement Shared Resource at the University of Arizona Cancer Center (P30 CA023074). University of Arizona Institutional Review Board Protocol #1711069340
Mary Hadeed, MPA; Terry A Badger, PhD, RN, PMHNP-BC; Chris Segrin, PhD; Alla Sikorskii, PhD
Social Cognitive Theory (SCT)
Thomas Brower - Poster: 20
College of Medicine - Tucson N/A - Medical Student
Cancer Prevention & Control Program
PI(s):Rina Fox, PhD, MPH
Text Message Health Behavior Intervention and Patient-Reported Outcomes among Cancer Survivors and Caregivers
Thomas Brower1, Meghan B. Skiba2, Terry A. Badger2, Marjorie A. Nelson2, Chris Segrin3, Alejandro Recio-Boiles1, Rina S. Fox4
Evelyn Alexander - Poster: 2
University of Arizona- Tucson COMT Surgery
Clinical and Translational Oncology Program
PI(s):Stephanie Worrell
Beyond Pathologic Response: Survival Advantage with Perioperative Chemoimmunotherapy in Resectable Esophageal and GEJ Cancer
Real-World Outcomes of Chemoimmunotherapy vs Chemoradiation ± Immunotherapy in Resectable Esophageal and GEJ Adenocarcinoma
While chemoradiation ± IO offered superior locoregional control, perioperative chemoimmunotherapy demonstrated a survival advantage, reflecting a trade-off between local and systemic efficacy.
• Therapeutic timing in esophageal adenocarcinoma is evolving
• Immunotherapy is redefining treatment strategies
• Real-world data is needed to evaluate long-term surgical and survival outcomes
To evaluate the clinical and survival outcomes of perioperative chemoimmunotherapy and neoadjuvant chemoradiation ± adjuvant IO
• Chemo + IO was associated with improved overall survival compared to nCRT ± IO, despite lower pCR and higher ypN+ rates
• nCRT ± IO continues to provide superior locoregional control
• Together, these findings highlight the trade-off between systemic survival advantage and locoregional disease control
Evelyn Alexander MD, Ahmed Elkamel MBBS, Shamele Battan-Wraith MD, Mazin Abdalgadir MBBS, Chiu Hsu PhD, Jonathan Rice MD PhD, Praveen Sridhar MD, Stephanie G. Worrell MD
Celia Sahli - Poster: 1
College of Pharmacy Pharmacology and Toxicology
Clinical and Translational Oncology Program
PI(s):Kenry
Machine Learning for Breast Tumor Classification: Insights from Nuclear Morphology
Helen Zukoski-Bartlett, Connar Smith - Poster: 3
College of Engineering Biomedical Engineering
Clinical and Translational Oncology Program
PI(s):Dr. Swarna Ganesh
Cancer Diagnostics: Printing Viable Microfluidic Chips Using Volumetric 3D Printing Techniques
Cancer Diagnostics: Printing Viable Microfluidic Chips Using Volumetric Printing Techniques Helen Bartlett1, Conner Simth1, and Swarna
•
Department
BIO5
University
Yanhao Jiang - Poster: 6
College of Pharmacy Pharmaceutical Science
Clinical and Translational Oncology Program
PI(s):Jianqin Lu
Enhanced Delivery of Camptothecin to Colorectal Carcinoma Using a Tumor-Penetrating Peptide Targeting p32
Enhanced Delivery of Camptothecin to Colorectal Carcinoma Using a Tumor-Penetrating Peptide Targeting p32
1, Skaggs Pharmaceutical Sciences Center, Department of
Introduction
LinTT1/Camptothesome markedly enhances the intracellular delivery
xenograft CRC mouse model
Camptothecin nanovesicles (Camptothesome) improves tumor drug delivery over free CPT but still has limited tumor distribution (~5%) Although the enhanced permeability and retention (EPR) effect aids nanotherapeutics' accumulation at tumor peripheries, effective intracellular internalization and deep tissue penetration remain difficult due to the dense extracellular matrix and interstitial fluid pressure Therefore, strategies to enhance tumor delivery are crucial to maximize Camptothesome’s therapeutic potential in cancer treatment
In this study, We successfully functionalized Camptothesome nanovesicles with the CendR motif using thiol-maleimide chemistry, linking an N-terminal cysteine on LinTT1 to DSPE-PEG2K-Maleimide, which anchors into the lipid bilayer Optimizing peptide ratios revealed that a 4 3% molar ratio maximized stability and cellular uptake in HCT116 and CT26 CRC cells, while showing no uptake difference in normal Beas-2b cells, demonstrating cancer selectivity LinTT1/Camptothesome exhibited enhanced tumor delivery via Golgi-dependent transcytosis, lower offtarget effects, and improved anti-CRC efficacy over Camptothesome and Onivyde in a CRC tumor model Elevated cleaved caspase-3 and γH2AX levels confirmed the increased anticancer activity, highlighting the strategy's therapeutic promise
Methods and Results
The schematic of self-assembly and endocytosis/transcytosis process of LinTT1/Camptothesome
LinTT1/Camptothesome-induced transcytosis is Golgi-dependent
LinTT1/Camptothesome beats Camptothesome on therapeutic efficacy in a human xenograft CRC carcinoma mouse model
LinTT1-decorated Camptothesome enhances tumor delivery, enables deeper tumor penetration, and reduces the non-specific tissue distribution in a
Jesse Ritter - Poster: 4
University of Arizona College of Medicine Internal Medicine
Clinical and Translational Oncology Program
PI(s):Muhammad Husnain
Clinical and Immunologic Correlates of Out-of-Specification Ciltacabtagene Autoleucel in Multiple Myeloma: A Univerisity of Arizona Analysis
Clinical and Immunologic Correlates of Out -of-Specification Ciltacabtagene Autoleucel in Multiple Myeloma: A Univerisity of Arizona Analysis
Jesse Ritter1, Danielle Bailey 1, Numaan Mahmood 1, Laura McPheeters2, Krishna Moturi2, Abhijeet Kumar2, Akshay Amaraneni2 , Ravitharan Krishnadasan 2, Sharad Khurana2, Muhammad Husnain2 University of Arizona College of Medicine – Tucson1, University of Arizona Cancer Center2
• BCMA-directed CAR T-cell therapy improves outcomes in relapsed/refractory multiple myeloma.
Manufacturing variability can produce out-of-specification (OOS) products (e.g., low viability, vector persistence) that may still be infused under institutional oversight.
• Real-world data on the safety, efficacy, and immune recovery of OOS products remain limited.
Objective: Evaluate clinical outcomes and lymphocyte kinetics in patients receiving OOS vs in-specification (IS) cilta-cel.
METHODS
• Retrospective single-center study of adults with relapsed/refractory myeloma undergoing leukapheresis for cilta-cel (Sept 2023–Sept 2025). Twenty-five patients collected; 22 infused (17 IS, 5 OOS). OOS defined as deviation from commercial release (e.g., viability < 80% or vector persistence) infused under institutional oversight.
• Endpoints:
o Primary: Non-inferiority of PFS (OOS vs IS)
o Secondary: OS, Day-90 composite toxicity (≥ grade 3 CRS, ICANS, infection, cytopenia, or ICU admission), and ALC/CD3⁺ T-cell kinetics.
Analyses: Kaplan–Meier and Cox models for survival; χ²/Wilcoxon for group comparisons; mixed-effects models for log₁₀-CD3⁺ trajectories; Pearson correlation for ALC–CD3⁺; landmark analyses for day +14/+30 thresholds.
RESULTS
Cohort: 25 leukapheresis, 22 infused (17 IS, 5 OOS); median follow-up 8.5 mo among survivors.
• Manufacturing Interval: Median leukapheresis-to-infusion 50 vs 76 days (p = 0.0025), indicating delayed release for OOS products. Baseline: OOS recipients had more prior therapy (7 vs 4 lines, p = 0.045), higher talquetamab exposure (62 % vs 6 %, p = 0.01), and lower ALC/CD3⁺ counts at collection (p = 0.086, 0.027).
• Efficacy: Median PFS not reached; 6-mo PFS 63 % (OOS) vs 60 % (IS); HR 0.96, p = 0.96. Median OS 16.9 vs 14.6 mo; 12-mo OS 65 % vs 80 %; HR 1.00, p = 0.99. CR: 6/10 IS vs 2/5 OOS; MRD-negativity 6/6 IS vs 2/2 OOS.
• Toxicity: Composite day-90 grade ≥ 3 events in 60 % OOS vs 40 % IS (RR 1.50, p = 0.61); no grade ≥ 3 CRS or ICANS. Immune Recovery: At day +30, OOS ALC 368 vs 744 cells/µL (p = 0.06); CD3⁺ 130 vs 658 cells/µL (p = 0.03). Mixed-effects model β = –0.42 ± 0.20 (≈2.6-fold lower CD3⁺).
• Correlations / Prognosis: ALC–CD3⁺ ρ = 0.88 (p < 0.001). Day +14 ALC ≥ 1000 cells/µL predicted inferior PFS and OS (HR ≈ 14, p = 0.04).
CONCLUSIONS
• OOS cilta-cel products showed non-inferior PFS and comparable safety to inspec (IS) infusions.
OOS manufacturing was enriched in heavily pretreated, talquetamab-exposed patients, consistent with T-cell exhaustion as a mechanistic driver.
• Low ALC and CD3⁺ counts at leukapheresis were associated with OOS manufacturing, suggesting impaired T-cell reserve rather than production failure. Despite delayed immune recovery, outcomes remained similar.
• These findings support selective, clinically supervised use of OOS CAR T products when re-manufacturing is impractical and warrant validation in larger, multi-center cohorts.
REFERENCES
N, Mooyaart JE, Hoogenboom JD, Daskalakis M, Tudesq J -J, Ram R, et al. CAR-T cell manufacturing failures and out-of-specification products in the real-world setting: A survey from the EBMT cellular therapy and immunobiology working party. Bone Marrow Transplantation. 2025;60(8):1184 -6.
Worel
Wenpan Li - Poster: 5
College of Pharmacy Dept. of Pharmacology/Toxicology
Clinical and Translational Oncology Program
PI(s):Jianqin Lu
A sphingolipid-derived paclitaxel nanovesicle enhances efficacy of combination therapies in triple-negative breast cancer and pancreatic cancer
