William A. Pace

William A. Pace

Auto-generated profile, based on publicly available information

William A. Pace

University of California, San Francisco 3 Department of Urology San Francisco, California

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
6 October 2021 in Molecular Metabolism
Verified
Journal Article
Open access
KBKarin J. Bosma
Karin J. Bosma
SASpencer R. Andrei
Spencer R. Andrei
LKLiora S Katz
Liora S Katz
ASAshley A. Smith
Ashley A. Smith
JDJennifer C. Dunn
Jennifer C. Dunn
MGMaureen A Gannon
Maureen A Gannon
Type 2 diabetes is characterized by hyperglycemia and inflammation. Prostaglandin E2, which signals through four G protein-coupled receptors (EP1-4), is a mediator of inflammation and is upregulated in diabetes. We have shown previously that EP3 receptor blockade promotes β-cell proliferation and survival in isolated mouse and human islets ex vivo. Here, we analyzed whether systemic EP3 blockade could enhance β-cell mass and identity in the setting of type 2 diabetes using mice with a spontaneous mutation in the leptin receptor (Leprdb). Four- or six-week-old, db/+, and db/db male mice were treated with an EP3 antagonist daily for two weeks. Pancreata were analyzed for α-cell and β-cell proliferation and β-cell mass. Islets were isolated for transcriptomic analysis. Selected gene expression changes were validated by immunolabeling of the pancreatic tissue sections. EP3 blockade increased β-cell mass in db/db mice through enhanced β-cell proliferation. Importantly, there were no effects on α-cell proliferation. EP3 blockade reversed the changes in islet gene expression associated with the db/db phenotype and restored the islet architecture. Expression of the GLP-1 receptor was slightly increased by EP3 antagonist treatment in db/db mice. In addition, the transcription factor nuclear factor E2-related factor 2 (Nrf2) and downstream targets were increased in islets from db/db mice in response to treatment with an EP3 antagonist. The markers of oxidative stress were decreased. The current study suggests that EP3 blockade promotes β-cell mass expansion in db/db mice. The beneficial effects of EP3 blockade may be mediated through Nrf2, which has recently emerged as a key mediator in the protection against cellular oxidative damage.

22 September 2023 in Clinical Radiology
Verified
Journal Article
Closed access
WPW.A. Pace
W.A. Pace
KSK. Smith
K. Smith
AGA. Gamboa
A. Gamboa
SZS. Zamirpour
S. Zamirpour
LGL. Ge
L. Ge
ETE. Tseng
E. Tseng

2 February 2024 in Journal of Vascular Surgery
Verified
Journal Article
Open access
ZMZachary A. Matthay
Zachary A. Matthay
WPWilliam A. Pace
William A. Pace
ESEric J. Smith
Eric J. Smith
RGRichard D. Gutierrez
Richard D. Gutierrez
WGWarren J. Gasper
Warren J. Gasper
JIJames C. Iannuzzi
James C. Iannuzzi
Inadequate vein quality or prior harvest precludes use of autologous single segment greater saphenous vein (ssGSV) in many patients with chronic limb-threatening ischemia (CLTI). Predictors of patient outcome after infrainguinal bypass with alternative (non-ssGSV) conduits are not well-understood. We explored whether limb presentation, bypass target, and conduit type were associated with amputation-free survival (AFS) after infrainguinal bypass using alternative conduits. A single-center retrospective study (2013-2020) was conducted of 139 infrainguinal bypasses performed for CLTI with cryopreserved ssGSV (cryovein) (n = 71), polytetrafluoroethylene (PTFE) (n = 23), or arm/spliced vein grafts (n = 45). Characteristics, Wound, Ischemia, and foot Infection (WIfI) stage, and outcomes were recorded. Multivariable Cox proportional hazards and classification and regression tree analysis modeled predictors of AFS. Within 139 cases, the mean age was 71 years, 59% of patients were male, and 51% of cases were nonelective. More patients undergoing bypass with cryovein were WIfI stage 4 (41%) compared with PTFE (13%) or arm/spliced vein (27%) (P = .04). Across groups, AFS at 2 years was 78% for spliced/arm, 79% for PTFE, and 53% for cryovein (adjusted hazard ratio for cryovein, 2.5; P = .02). Among cases using cryovein, classification and regression tree analysis showed that WIfI stage 3 or 4, age >70 years, and prior failed bypass were predictive of the lowest AFS at 2 years of 36% vs AFS of 58% to 76% among subgroups with less than two of these factors. Although secondary patency at 2 years was worse in the cryovein group (26% vs 68% and 89% in arm/spliced and PTFE groups; P < .01), in patients with tissue loss there was no statistically significant difference in wound healing in the cryovein group (72%) compared with other bypass types (72% vs 87%, respectively; P = .12). In patients with CLTI lacking suitable ssGSV, bypass with autogenous arm/spliced vein or PTFE has superior AFS compared with cryovein, although data were limited for PTFE conduits for distal targets. Despite poor patency with cryovein, wound healing is achieved in a majority of cases, although it should be used with caution in older patients with high WIfI stage and prior failed bypass, given the low rates of AFS.

4 July 2023 in Journal of Experimental Orthopaedics
Verified
Journal Article
Open access
PDF file
MVMichael S. Valerio
Michael S. Valerio
WPWilliam A. Pace
William A. Pace
CDConnor P. Dolan
Connor P. Dolan
JEJorge B. Edwards
Jorge B. Edwards
NJNaveena B. Janakiram
Naveena B. Janakiram
SGStephen M. Goldman
Stephen M. Goldman
Purpose: This study aimed to develop and characterize a closed intra‐articular fracture (IAF) mediated post‐traumatic osteoarthritis (PTOA) model in rats to serve as a testbed for putative disease modifying interventions.Methods: Male rats were subject to a 0 Joule (J), 1 J, 3 J, or 5 J blunt‐force impact to the lateral aspect of the knee and allowed to heal for 14 and 56 days. Micro‐CT was performed at time of injury and at the specified endpoints to assess bone morphometry and bone mineral density measurements. Cytokines and osteochondral degradation markers were assayed from serum and synovial fluid via immunoassays. Histopathological analyses were performed on decalcified tissues and assessed for evidence of osteochondral degradation.Results: High‐energy (5 J) blunt impacts consistently induced IAF to the proximal tibia, distal femur, or both while lower energy (1 J and 3 J) impacts did not. CCL2 was found to be elevated in the synovial fluid of rats with IAF at both 14‐ and 56‐days post‐injury while COMP and NTX‐1 were upregulated chronically relative to sham controls. Histological analysis showed increased immune cell infiltration, increased osteoclasts and osteochondral degradation with IAF relative to sham.Conclusion: Based on results from the current study, our data indicates that a 5 J blunt‐forced impact adequately and consistently induces hallmark osteoarthritic changes to the articular surface and subchondral bone at 56 days after IAF. Marked development of PTOA pathobiology suggest this model will provide a robust testbed for screening putative disease modifying interventions that might be translated to the clinic for militarily relevant, high‐energy joint injuries.

6 January 2026 in JMIR Bioinformatics and Biotechnology
Verified
Journal Article
Open access
MCMarvin N Carlisle
Marvin N Carlisle
WPWilliam A Pace
William A Pace
ALAndrew W Liu
Andrew W Liu
RKRobert Krumm
Robert Krumm
JCJanet E Cowan
Janet E Cowan
AOAnobel Y Odisho
Anobel Y Odisho
Background: The manual abstraction of unstructured clinical data is often necessary for granular clinical outcomes research but is time consuming and can be of variable quality. Large language models (LLMs) show promise in medical data extraction yet integrating them into research workflows remains challenging and poorly described. Objective: This study aimed to develop and integrate an LLM-based system for automated data extraction from unstructured electronic health record (EHR) text reports within an established clinical outcomes database. Methods: We implemented a generative artificial intelligence pipeline (UODBLLM) utilizing a flexible language model interface that supports various LLM implementations, including Health Insurance Portability and Accountability Act-compliant cloud services and local open-source models. We used extensible markup language (XML)-structured prompts and integrated using an open database connectivity interface to generate structured data from clinical documentation in the EHR. We evaluated the UODBLLM’s performance on the completion rate, processing time, and extraction capabilities across multiple clinical data elements, including quantitative measurements, categorical assessments, and anatomical descriptions, using sample magnetic resonance imaging (MRI) reports as test cases. System reliability was tested across multiple batches to assess scalability and consistency. Results: Piloted against MRI reports, UODBLLM processed 1800 clinical documents with a 100% completion rate and an average processing time of 8.90 seconds per report. The token utilization averaged 2692 tokens per report, with an input-to-output ratio of approximately 13:2, resulting in a processing cost of US $0.009 per report. UODBLLM had consistent performance across 18 batches of 100 reports each and completed all processing in 4.45 hours. From each report, UODBLLM extracted 16 structured clinical elements, including prostate volume, prostate-specific antigen values, Prostate Imaging Reporting and Data System scores, clinical staging, and anatomical assessments. All extracted data were automatically validated against predefined schemas and stored in standardized JSON format. Conclusions: We demonstrated the successful integration of an LLM-based extraction system within an existing clinical outcomes database, achieving rapid, comprehensive data extraction at minimal cost. UODBLLM provides a scalable, efficient solution for automating clinical data extraction while maintaining protected health information security. This approach could significantly accelerate research timelines and expand feasible clinical studies, particularly for large-scale database projects.

1 January 2026 in JCO Clinical Cancer Informatics
Verified
Journal Article
Closed access
AOAnobel Y. Odisho
Anobel Y. Odisho
ALAndrew W. Liu
Andrew W. Liu
WPWilliam A. Pace
William A. Pace
MCMarvin N. Carlisle
Marvin N. Carlisle
RKRobert Krumm
Robert Krumm
MCMatthew R. Cooperberg
Matthew R. Cooperberg
PURPOSE: Radiology reports are stored as plain text in most electronic health records, rendering the data computationally inaccessible. Large language models are powerful tools for analyzing unstructured text but relatively untested in urologic oncology. We aimed to develop a pipeline to extract data from plain text prostate magnetic resonance imaging (MRI) reports using GPT4.0 and compare the accuracy to manually abstracted data. METHODS: We developed a data pipeline using a secure, enterprise-wide deployment of OpenAI's GPT-4.0 to automatically extract data elements from prostate MRI report text when presented with prostate MRI reports. Identical prompts and reports were sent multiple times to determine response variability. We extracted 15 data elements per report and compared accuracy to a manually abstracted gold standard. RESULTS: Across 424 prostate MRI reports, GPT-4.0 response accuracy was consistently above 95%. Individual field accuracies were 98.3% (96.3%-99.3%) for prostate-specific antigen density, 97.4% (95.4%-98.7%) for extracapsular extension, and 98.1% (96.3%-99.2%) for TNM stage, and had a median of 98.1% (96.3%-99.2%), a mean of 97.2% (95.2%-98.3%), and a range of 99.8% (98.7%-100.0%) for number of suspicious lesions to 87.7% (84.2%-90.7%) for identification of lesion location in the base of the prostate. Response variability over five repeated runs ranged from 0.14% to 3.61%, differed based on the data element extracted ( P < .001), and was inversely correlated with accuracy ( P < .001). In disagreements between manual and GPT-4.0 extracted data, GPT-4.0 responses were more often deemed correct by an additional reviewer. CONCLUSION: GPT-4.0 had high accuracy with low variability in extracting data points from prostate cancer MRI reports with low upfront programming requirements. This represents an effective tool to expedite medical data extraction for clinical and research use cases.