MétaCan
Menu
Back to cohort

Computational extraction and analysis of de-identified medical records to characterize hyperammonemia in patients with fibrolamellar carcinoma (FLC).

2021· article· en· W3167119435 on OpenAlexaff
Travis Zack, Samantha Maisel, Allison F. O’Neill, Michael P. LaQuaglia, Michael Herman, Jennifer J. Knox, Amin Yaqubie, Alan P. Venook, Robert J. Mayer, John D. Gordan, Ghassan K. Abou‐Alfa

Bibliographic record

VenueJournal of Clinical Oncology · 2021
Typearticle
Languageen
FieldMedicine
TopicRadiomics and Machine Learning in Medical Imaging
Canadian institutionsPrincess Margaret Cancer CentreUniversity Health NetworkUniversity of Toronto
Fundersnot available
KeywordsHyperammonemiaMedicineInternal medicineContext (archaeology)EverolimusIncidence (geometry)OncologyCancerGastroenterology

Abstract

fetched live from OpenAlex

e16169 Background: Rare cancers including FLC make up 25% of adult tumors, but are difficult to study due to low incidence and incomplete case identification. FLC occurs in adolescents and young adults without liver dysfunction. Hyperammonemia has been frequently reported in FLC patients, but is poorly understood. Methods: Data from three clinical trials allowed us to establish the incidence of hyperammonemia (serum ammonia value > 75 µmol/L) in FLC. In these studies, FLC patients received everolimus, estrogen deprivation therapy (EDT) with leuprolide + letrozole or everolimus + EDT (Oncologist. 2020 25(11):925-e1603), ENMD-2076 (Oncologist. 2020 25(12):e1837-e1845), or neratinib (J Clin Oncol 39, 2021 (suppl 3; abstr 310); ammonia was tested prospectively in the latter two studies. To assess impacts of cancer therapy or liver dysfunction, we studied hyperammonemia in FLC and non-FLC patients at UCSF in parallel. Using Natural Language Processing (NLP) of pathology reports and oncology notes from > 2300 liver cancer patients from the last 12 years of UCSF records, we identified a cohort of patients with FLC, contrasting their laboratory data to all UCSF patients with ammonia testing for the last 10 years. We used leiden clustering and umap dimensionality reduction to contrast FLC and other patients to assess the clinical context of hyperammonemia. Results: Data from the 3 trials showed hyperammonemia in 10 of 32 (31.3%) FLC patients during study participation, independent of the therapy received. These patients exhibited hyperammonemia with varying levels, and at different points in their treatment. NLP identified 37 patients with FLC ( < 0.1% of liver cancer patients), with 33% showing hyperammonemia. Across all UCSF patients, we found 24,000 independent visits where ammonia was tested, with > 2400 demonstrating hyperammonemia. Using leidan clustering on all encounters with ammonia > 75 µmol/L, we found distinct subsets of hyperammonemia corresponding to known metabolic and physiologic processes (e.g., fulminant liver failure, tumor lysis syndrome, etc). FLC patients clustered separately from hepatocellular carcinoma patients with hyperammonemic encephalopathy due to cirrhosis. Conclusions: NLP of large EMRs is a valuable tool to study FLC, a rare cancer. Herein, we have defined hyperammonemia as a frequent event in FLC, not directly linked to hepatic dysfunction or individual therapies. Further investigation may determine whether hyperammonemia is related to FLC tumor biology.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.002
metaresearch head score (Gemma)0.009
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Observational · Consensus signal: Observational
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.010
Threshold uncertainty score0.021

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0020.009
Meta-epidemiology (narrow)0.0010.000
Meta-epidemiology (broad)0.0010.002
Bibliometrics0.0100.004
Science and technology studies0.0010.000
Scholarly communication0.0010.001
Open science0.0010.001
Research integrity0.0010.001
Insufficient payload (model declined to judge)0.0020.001

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.034
GPT teacher head0.408
Teacher spread0.374 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designObservational
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations0
Published2021
Admission routes1
Has abstractyes

Explore more

Same venueJournal of Clinical OncologySame topicRadiomics and Machine Learning in Medical ImagingFrench-language works237,207