The latest medical research on Gastroenterology

Diagnosis code classification is a common method for cohort identification in cirrhosis research, but it is often inaccurate and augmented by labor-intensive chart review. Natural language processing (NLP) using large language models (LLMs) is a potentially more accurate method. To assess LLMs' potential for cirrhosis cohort identification, we compared code-based versus LLM-based classification with chart review as a "gold standard."

We extracted and conducted a limited chart review of 3,788 discharge summaries of cirrhosis admissions. We engineered zero-shot prompts using Generative Pre-trained Transformer (GPT)-4 to determine whether cirrhosis and its complications were active hospitalization problems. We calculated positive predictive values (PPVs) of LLM-based classification versus limited chart review, and PPVs of code-based versus LLM-based classification as a "silver standard" in all 3,788 summaries.

Versus gold standard chart review, code-based classification achieved PPVs of 82.2% for identifying cirrhosis, 41.7% hepatic encephalopathy, 72.8% ascites, 59.8% gastrointestinal bleeding, and 48.8% spontaneous bacterial peritonitis. Compared to chart review, GPT-4 achieved 87.8-98.8% accuracies for identifying cirrhosis and its complications. Using LLM as a silver standard, code-based classification achieved PPVs of 79.8% for identifying cirrhosis, 53.9% hepatic encephalopathy, 55.3% ascites, 67.6% gastrointestinal bleeding, and 65.5% spontaneous bacterial peritonitis.

LLM-based classification was highly accurate versus manual chart review in identifying cirrhosis and its complications - this allowed us to assess the performance of code-based classification at scale using LLMs as a silver standard. These results suggest LLMs could augment or replace code-based cohort classification and raise questions regarding the necessity of chart review.

Neuroendocrine neoplasms of gastrointestinal tract (GIT) and pancreas are heterogenous tumors. World Health Organization (WHO) 2019 classification introduced Grade (G)3 neuroendocrine tumor (NET) distinct from neuroendocrine carcinoma (NEC), based on molecular differences and to triage the patients for appropriate therapy. This distinction largely relies on morphology, which can be challenging at times. Genomic profiling has revealed TP53 and RB1 mutations in NECs, while death domain-associated protein 6 (DAXX) and alpha-thalassemia/mental retardation X-linked (ATRX), in G3NET. Their role as biological markers in differentiating these entities and their significance as prognostic markers are not yet established. This study aims at analyzing the diagnostic and prognostic role of p53 and ATRX in neuroendocrine neoplasms of GIT and pancreas.

A single-centre, eight-year retrospective study of neuroendocrine neoplasm of GIT and pancreas comprised G2NET, G3NET and NEC. Tumor slides were stained by immunohistochemistry for p53 and ATRX. Strong nuclear staining of > 50% of tumor cells for p53 was considered mutated. Nuclear staining of ATRX in < 5% of tumor cells was considered ATRX loss. Expression of p53 and ATRX was analyzed and correlated with tumor grades and patient survival.

Fifty-five patients with gastro-entero-pancreatic neuroendocrine neoplasm were studied, comprising G2NET (58%), G3NET (16%) and NEC (26%). Median age of diagnosis was 59 years with male predominance. The pancreas was the most common site followed by the small bowel. NEC showed lower survival compared to G3 and G2NET. Mutated p53 immunohistochemical expression was more frequent among NEC than G3NET. Patients with mutated p53 had significantly lower survival irrespective of the grade (p = 0.001). There was no association of ATRX loss with grade or survival.

G3NETs are genetically different from NECs. Use of immunohistochemistry for p53 in addition to histomorphology may facilitate accurate categorization of NEC and G3NET. Mutated p53 may also be used as an independent prognostic marker in neuroendocrine tumors of GIT and pancreas.