Is ChatGPT a trusted source of information for total hip and knee arthroplasty patients?

Aims: While internet search engines have been the primary information source for patients' questions, artificial intelligence large language models like ChatGPT are trending towards becoming the new primary source. The purpose of this study was to determine if ChatGPT can answer patient questions about total hip (THA) and knee arthroplasty (TKA) with consistent accuracy, comprehensiveness, and easy readability. Methods: We posed the 20 most Google-searched questions about THA and TKA, plus ten additional postoperative questions, to ChatGPT. Each question was asked twice to evaluate for consistency in quality. Following each response, we responded with, "Please explain so it is easier to understand," to evaluate ChatGPT's ability to reduce response reading grade level, measured as Flesch-Kincaid Grade Level (FKGL). Five resident physicians rated the 120 responses on 1 to 5 accuracy and comprehensiveness scales. Additionally, they answered a "yes" or "no" question regarding acceptability. Mean scores were calculated for each question, and responses were deemed acceptable if ≥ four raters answered "yes." Results: The mean accuracy and comprehensiveness scores were 4.26 (95% confidence interval (CI) 4.19 to 4.33) and 3.79 (95% CI 3.69 to 3.89), respectively. Out of all the responses, 59.2% (71/120; 95% CI 50.0% to 67.7%) were acceptable. ChatGPT was consistent when asked the same question twice, giving no significant difference in accuracy (t = 0.821; p = 0.415), comprehensiveness (t = 1.387; p = 0.171), acceptability (χ2 = 1.832; p = 0.176), and FKGL (t = 0.264; p = 0.793). There was a significantly lower FKGL (t = 2.204; p = 0.029) for easier responses (11.14; 95% CI 10.57 to 11.71) than original responses (12.15; 95% CI 11.45 to 12.85). Conclusion: ChatGPT answered THA and TKA patient questions with accuracy comparable to previous reports of websites, with adequate comprehensiveness, but with limited acceptability as the sole information source. ChatGPT has potential for answering patient questions about THA and TKA, but needs improvement.

Enhanced BRAF engagement by NRAS mutants capable of promoting melanoma initiation.

A distinct profile of NRAS mutants is observed in each tumor type. It is unclear whether these profiles are determined by mutagenic events or functional differences between NRAS oncoproteins. Here, we establish functional hallmarks of NRAS mutants enriched in human melanoma. We generate eight conditional, knock-in mouse models and show that rare melanoma mutants (NRAS G12D, G13D, G13R, Q61H, and Q61P) are poor drivers of spontaneous melanoma formation, whereas common melanoma mutants (NRAS Q61R, Q61K, or Q61L) induce rapid tumor onset with high penetrance. Molecular dynamics simulations, combined with cell-based protein-protein interaction studies, reveal that melanomagenic NRAS mutants form intramolecular contacts that enhance BRAF binding affinity, BRAF-CRAF heterodimer formation, and MAPK > ERK signaling. Along with the allelic series of conditional mouse models we describe, these results establish a mechanistic basis for the enrichment of specific NRAS mutants in human melanoma.

Impact of age-related T cell dynamics on the identification of biomarkers predictive of immunotherapy discontinuation: a prospective cohort study.

BACKGROUND: The impact of biologic aging on immune checkpoint inhibitor (ICI) toxicity and efficacy is underexplored in metastatic melanoma (MM). In peripheral blood T-lymphocytes (PBTLs), biologic aging is characterized by changes in T-cell composition and cellular senescence. Whether indicators of PBTL biologic aging vary in MM patients or can be used to predict premature ICI discontinuation (pID) is unknown. METHODS: We prospectively collected PBTLs from 117 cancer-free controls and 46 MM patients scheduled to begin pembrolizumab or nivolumab monotherapy. 74 mRNAs indicative of T-cell subsets, activation, co-stimuation/inhibition and cellular senescence were measured by Nanostring. Relationships between each mRNA and chronologic age were assessed in patients and controls. Candidate biomarkers were identified by calculating the hazard ratio (HR) for pID in patients divided into low and high groups based on log-transformed mRNA levels or the magnitude by which each mRNA measurement deviated from the control trend (Δage). Area under the curve (AUC) analyses explored the ability of each biomarker to discriminate between patients with and without pID at 6 months and 1 year. RESULTS: Fifteen mRNAs correlated with chronologic age in controls, including markers of T-cell subsets, differentiation, cytokine production and co-stimulation/inhibition. None of these mRNAs remained correlated with age in patients. Median follow-up was 94.8 (1.6-195.7) weeks and 35 of 46 patients discontinued therapy (23 progression, 7 toxicity, 5 comorbidity/patient preference). Elevated pre-therapy CD8A (HR 2.2[1.1-4.9]), CD45RB (HR 2.9[1.4-5.8]) and TNFRSF14 (HR 2.2[1.1-4.5]) levels predicted pID independent of Δage-correction. CD3ε, CD27 and FOXO1 predicted pID only after Δage-correction (HR 2.5[1.3-5.1]; 3.7[1.8-7.8]; 2.1[1.1-4.3]). AUC analysis identified Δage-CD3ε and -CD27 as candidate predictors of pID (AUC=0.73; 0.75). CONCLUSIONS: Correlations between transcriptional markers of PBTL composition and chronologic age are disrupted in MM. Correcting for normal, age-related trends in biomarker expression unveils new biomarker candidates predictive of ICI outcomes.