Faculty Bibliography

PURPOSE/OBJECTIVE:To develop and evaluate a hybrid large language model (LLM)- and rule-based algorithm for tracking pancreatic cystic lesions (PCLs) across serial radiology reports and apply it to a single-center cohort to characterize cyst growth using first-to-last and consecutive comparisons. METHODS:This retrospective study included all abdominal imaging reports interpreted by abdominal radiologists between 2005 and 2024. A validated LLM identified reports describing PCLs. A hybrid algorithm combining LLM-extracted cyst features (size, location) with rule-based logic matched dominant cysts across serial reports. Two blinded radiologists independently annotated cyst matches in a subset of reports. Agreement between algorithm and radiologists was assessed using percent agreement and Cohen's κ. Discrepancies were adjudicated by a third radiologist for error analysis. Absolute size-based growth was assessed according to American College of Radiology (ACR) criteria, whereas cyst growth rate as a worrisome feature was assessed per Kyoto criterion, using both first-to-last and consecutive studies. Predictors of meeting the Kyoto cyst growth rate criterion were analyzed using multivariable Cox regression. RESULTS:In 58 patients with 418 longitudinal reports, agreement with the adjudicated reference standard was 96%. Cohen's κ was 0.81 (95% CI: 0.66-0.91). The overall cohort included 8,176 patients (61.8% female; mean age 67.6 ± 11.7 years) with 8,389 cysts across 28,489 reports over 40.6 ± 28.5 months (annual growth rate = 0.35 mm/year). ACR absolute growth criteria classified a higher proportion of cysts than the Kyoto cyst growth rate criterion using first-to-last examinations (16.3% vs. 7.4%, p < 0.001), whereas Kyoto cyst growth rate criterion identified a higher proportion using consecutive examinations (21.4% vs. 14%, p < 0.001). Larger baseline cyst size predicted meeting the Kyoto cyst growth rate criterion (HR 1.9-4.0; highest HR 4.0, 95% CI 3.08-5.21; p < 0.001). CONCLUSION/CONCLUSIONS:A hybrid LLM-rule-based algorithm enables reliable automated longitudinal tracking of PCLs across serial radiology reports. When applied at scale, cyst growth classification varied according to the guideline framework and interval comparison method used, and baseline cyst size was the strongest predictor of meeting Kyoto growth rate threshold.

Many bacteria produce polysaccharide-based capsules that protect them from environmental insults and play a role in virulence, host invasion, and other functions. Understanding how the polysaccharide components are synthesized could provide new means to combat bacterial infections. We have previously characterized two pairs of homologous enzymes involved in the biosynthesis of capsular sugar precursors GDP-6-deoxy-D-altro-heptose and GDP-6-OMe-L-gluco-heptose in Campylobacter jejuni. However, the substrate specificity and mechanism of action of these enzymes-C3 and/or C5 epimerases DdahB and MlghB and C4 reductases DdahC and MlghC-are unknown. Here, we demonstrate that these enzymes are highly specific for heptose substrates, using mannose substrates inefficiently with the exception of MlghB. We show that DdahB and MlghB feature a jellyroll fold typical of cupins, which possess a range of activities including epimerizations, GDP occupying a similar position as in cupins. DdahC and MlghC contain a Rossman fold, a catalytic triad, and a small C-terminal domain typical of short-chain dehydratase reductase enzymes. Integrating structural information with site-directed mutagenesis allowed us to identify features unique to each enzyme and provide mechanistic insight. In the epimerases, mutagenesis of H67, D173, N121, Y134, and Y132 suggested the presence of alternative catalytic residues. We showed that the reductases could reduce GDP-4-keto-6-deoxy-mannulose without prior epimerization although DdahC preferred the pre-epimerized substrate and identified T110 and H180 as important for substrate specificity and catalytic efficacy. This information can be exploited to identify inhibitors for therapeutic applications or to tailor these enzymes to synthesize novel sugars useful as glycobiology tools.