Usefulness of an Artificial Neural Network in the Prediction of β-Lactam Allergy

Based on UHPLC-Q-TOF-MS and bioinformatics strategies, the potential allergens and mechanisms of allergic reactions caused by Danshen injection were explored

The aim is to investigate the potential allergens and mechanisms underlying allergic-like reactions induced by Danshen injection (DSI). Utilizing ultra-performance liquid chromatography quadrupole time-of-flight mass spectrometry (UHPLC-Q-TOF-MS), metabolomics, and bioinformatics, we identified the key allergens, targets, and metabolic pathways involved in DSI-induced allergic-like reactions, validating binding efficiency through molecular docking and molecular dynamics. A total of 45 compounds were identified within DSI, with 24 compounds exhibiting strong binding activity to the MrgprX2 activation site. DSI was found to cause changes in 89 endogenous metabolites, including arachidonic acid, prostaglandins, and leukotrienes, primarily affecting pathways such as phenylalanine metabolism and arachidonic acid metabolism. The key allergens identified were Cryptotanshinone, Miltipolone, Neocryptotanshinone, Salvianolic acid B, and Isosalvianolic acid C, which primarily trigger allergic-like reactions by regulating upstream signaling targets such as ALOX5, PTGS1, PPARD, and LTB4R. Validation confirmed the high binding affinity and stability between key allergens and targets. These findings indicate that the allergic components in DSI primarily induce allergic-like reactions by modulating the aforementioned signaling targets, activating the AA metabolic pathway, promoting mast cell degranulation, and releasing downstream endogenous inflammatory mediators, subsequently eliciting allergic-like reactions.

Predictive models and applicability of artificial intelligence-based approaches in drug allergy

PURPOSE OF REVIEW

Drug allergy is responsible for a huge burden on public healthcare systems, representing in some instances a threat for patient's life. Diagnosis is complex due to the heterogeneity of clinical phenotypes and mechanisms involved, the limitations of in vitro tests, and the associated risk to in vivo tests. Predictive models, including those using recent advances in artificial intelligence, may circumvent these drawbacks, leading to an appropriate classification of patients and improving their management in clinical settings.

RECENT FINDINGS

Scores and predictive models to assess drug allergy development, including patient risk stratification, are scarce and usually apply logistic regression analysis. Over recent years, different methods encompassed under the general umbrella of artificial intelligence, including machine and deep learning, and artificial neural networks, are emerging as powerful tools to provide reliable and optimal models for clinical diagnosis, prediction, and precision medicine in different types of drug allergy.

SUMMARY

This review provides general concepts and current evidence supporting the potential utility of predictive models and artificial intelligence branches in drug allergy diagnosis.

Predicting Penicillin Allergy: A United States Multicenter Retrospective Study

BACKGROUND

Using the reaction history in logistic regression and machine learning (ML) models to predict penicillin allergy has been reported based on non-US data.

OBJECTIVE

We developed ML positive penicillin allergy testing prediction models from multisite US data.

METHODS

Retrospective data from 4 US-based hospitals were grouped into 4 datasets: enriched training (1:3 case-control matched cohort), enriched testing, nonenriched internal testing, and nonenriched external testing. ML algorithms were used for model development. We determined area under the curve (AUC) and applied the Shapley Additive exPlanations (SHAP) framework to interpret risk drivers.

RESULTS

Of 4777 patients (mean age 60 [standard deviation: 17] years; 68% women, 91% White, and 86% non-Hispanic) evaluated for penicillin allergy labels, 513 (11%) had positive penicillin allergy testing. Model input variables were frequently missing: immediate or delayed onset (71%), signs or symptoms (13%), and treatment (31%). The gradient-boosted model was the strongest model with an AUC of 0.67 (95% confidence interval [CI]: 0.57-0.77), which improved to 0.87 (95% CI: 0.73-1) when only cases with complete data were used. Top SHAP drivers for positive testing were reactions within the last year and reactions requiring medical attention; female sex and reaction of hives/urticaria were also positive drivers.

CONCLUSIONS

An ML prediction model for positive penicillin allergy skin testing using US-based retrospective data did not achieve performance strong enough for acceptance and adoption. The optimal ML prediction model for positive penicillin allergy testing was driven by time since reaction, seek medical attention, female sex, and hives/urticaria.

How to Define and Manage Low-Risk Drug Allergy Labels

Risk stratification in drug allergy implies that specific risk categories (eg, low, moderate, and high) classify historical drug hypersensitivity reactions. These risk categories can be based on reaction phenotypic characteristics, the timing of the reaction and evaluation, the required reaction management, and individual characteristics. Although a multitude of frameworks have been described in the literature, particularly for penicillin allergy labels, there has yet to be a global consensus, and approaches continue to vary between allergy centers. Immune-mediated drug allergies can sometimes be confirmed using skin testing, but a negative drug challenge is required to demonstrate tolerance and remove the allergy from the electronic health record ("delabel" the allergy). Even for quintessential IgE-mediated drug allergy, penicillin allergy, recent data reveal that a direct oral challenge, without prior skin testing, is an appropriate diagnostic strategy in those who are considered low-risk. Drug allergy pathogenesis and clinical manifestations may vary depending on the culprit drug, and as such, the optimal approach should be based on risk stratification that considers individual patient and reaction characteristics, the likely hypersensitivity reaction phenotype, the drug class, and the patient's clinical needs. This article will describe low-risk drug allergy labels, focusing on β-lactam and sulfonamide antibiotics, nonsteroidal anti-inflammatory drugs, iodinated contrast media, and common chemotherapeutics. This review will also address practical management approaches using currently available risk stratification and clinical decision tools.

Delabeling of allergy to beta-lactam antibiotics in hospitalized patients: a prospective study evaluating cost savings

BACKGROUND

Patients with a penicillin allergy label are at risk of an associated increase in adverse antibiotic events and hospitalization costs.

AIM

We aimed to study the economic savings derived from the correct diagnosis and delabeling inpatients with suspected beta-lactam allergy, considering the acquisition cost of antimicrobials prescribed during a patient's hospital stay.

METHOD

We prospectively evaluated patients admitted to the University Hospital of Salamanca who had been labeled as allergic to beta-lactams and performed a delabeling study. Subsequently, cost differences between antibiotics administered before and after the allergy study and those derived from those patients who received alternative antibiotics during admission and those who switched to beta-lactams after the allergy study were calculated.

RESULTS

One hundred seventy-seven inpatients labeled as allergic to beta-lactams underwent a delabeling study; 34 (19.2%) were confirmed to have allergy to beta-lactams. Of the total number of patients, 136 (76.8%) received antibiotics during their hospitalization, involving a mean (SD) cost of €203.07 (318.42) and a median (IQR) cost of €88.97 (48.86-233.56). After delabeling in 85 (62.5%) patients, the antibiotic treatment was changed to beta-lactams. In this group of patients, the mean cost (SD) decreased from €188.91 (351.09) before the change to 91.31 (136.07) afterward, and the median cost (IQR) decreased from €72.92 (45.82-211.99) to €19.24 (11.66-168). The reduction was significant compared to the median cost of patients whose treatment was not changed to beta-lactams (p<0.001).

CONCLUSION

Delabeling hospitalized patients represents a cost-saving measure for treating patients labeled as allergic to beta-lactams.

Genetic association of beta-lactams-induced hypersensitivity reactions: A systematic review of genome-wide evidence and meta-analysis of candidate genes

Importance

Beta-lactams (BLs) are the most prescribed antibiotics, being the most frequent cause of drug allergy. However, the association between BL allergy and genetic variations is still unclear.

Objective

This systematic review and meta-analysis aimed to summarize the genetic effects of BL-induced hypersensitivity using existing evidence.

Methods

We searched PubMed, Medline, Scopus, EMBASE, CINAHL, and Cochrane Library from inception to September 15, 2022 with no language restriction. Genetic association studies investigating genetic variant/polymorphism and risk of drug-induced hypersensitivity reactions among individuals receiving BL-antibiotics were included. We excluded studies of acute interstitial nephritis, drug-induced liver injury, serum sickness, and isolated drug fever. Data were comprehensively synthesized and quality of study were assessed using STrengthening the Reporting of Genetic Association Studies (STREGA). The record screening, extraction and quality assessment were performed by two reviewers and discussions were made to resolve discrepancies. The effects of each variant were pooled and evaluated by modified Venice criteria.

Results

A total of 9276 records were identified, and 31 studies were eligible for inclusion. Twenty-seven were candidate-gene association studies (5416 cases and 5939 controls), while the others were next-generation sequencing (NGS) or genome-wide association studies (GWASs) (119 838 cases and 1 487 111 controls). Forty-nine polymorphisms were identified and most of them located in allergic reaction pathways. Meta-analyses of 15 candidate variants in a mixture of both immediate and non-immediate reactions revealed weak genetic effects of rs1801275 (8 studies; n = 1,560; odd ratio 0.73; 95%CI: 0.57-0.93) and rs20541 (4 studies; n = 1,482; odd ratio 1.34; 95%CI: 1.07-1.68) in IL4R and IL13, respectively. Results from GWASs and NGS identified, and confirmed associations in HLA regions including HLA-DRA, HLA-B, HLA-DQA, HLA-DRB1, and HLA-DRB3.

Conclusion

Our study summarized genetic evidence influencing BL-induced hypersensitivity and estimated effects of potential variants. We postulated that the genomic studies provide better insights to the mechanism of reactions and suggest potential effects of HLA Class II variants. However, results were inconsistent and unable to generalize in different settings. Further high-throughput studies with a well-defined function, epigenetic interaction, incorporated with clinical factors, would be beneficial for risk identification in BL-induced hypersensitivity.

Machine-learning-based adverse drug event prediction from observational health data: A review

Adverse drug events (ADEs) are responsible for a significant number of hospital admissions and fatalities. Machine learning models have been developed to assess the individual patient risk of having an ADE. In this article, we have reviewed studies addressing the prediction of ADEs in observational health data with machine learning. The field of individualised ADE prediction is rapidly emerging through the increasing availability of additional data modalities (e.g., genetic data, screening data, wearables data) and advanced deep learning models such as transformers. Consequently, personalised adverse drug event predictions are becoming more feasible and tangible.

Exploring the Intersection of Artificial Intelligence and Clinical Healthcare: A Multidisciplinary Review

Artificial intelligence (AI) plays a more and more important role in our everyday life due to the advantages that it brings when used, such as 24/7 availability, a very low percentage of errors, ability to provide real time insights, or performing a fast analysis. AI is increasingly being used in clinical medical and dental healthcare analyses, with valuable applications, which include disease diagnosis, risk assessment, treatment planning, and drug discovery. This paper presents a narrative literature review of AI use in healthcare from a multi-disciplinary perspective, specifically in the cardiology, allergology, endocrinology, and dental fields. The paper highlights data from recent research and development efforts in AI for healthcare, as well as challenges and limitations associated with AI implementation, such as data privacy and security considerations, along with ethical and legal concerns. The regulation of responsible design, development, and use of AI in healthcare is still in early stages due to the rapid evolution of the field. However, it is our duty to carefully consider the ethical implications of implementing AI and to respond appropriately. With the potential to reshape healthcare delivery and enhance patient outcomes, AI systems continue to reveal their capabilities.

Artificial Intelligence: Exploring the Future of Innovation in Allergy Immunology

PURPOSE OF REVIEW

Artificial intelligence (AI) has increasingly been used in healthcare. Given the capacity of AI to handle large data and complex relationships between variables, AI is well suited for applications in healthcare. Recently, AI has been applied to allergy research.

RECENT FINDINGS

In this article, we review how AI technologies have been utilized in basic science and clinical allergy research for asthma, atopic dermatitis, rhinology, adverse reactions to drugs and vaccines, food allergy, anaphylaxis, urticaria, and eosinophilic gastrointestinal disorders. We discuss barriers for AI adoption to improve the care of patients with atopic diseases. These studies demonstrate the utility of applying AI to the field of allergy to help investigators expand their understanding of disease pathogenesis, improve diagnostic accuracy, enable prediction for treatments and outcomes, and for drug discovery.

Changes in Sensitization Patterns in the Last 25 Years in 619 Patients with Confirmed Diagnoses of Immediate Hypersensitivity Reactions to Beta-Lactams

Beta-lactam (BL) drugs are the antibiotics most prescribed worldwide due to their broad spectrum of action. They are also the most frequently implied in hypersensitivity reactions with a known specific immunological mechanism. Since the commercialization of benzylpenicillin, allergic reactions have been described; over the years, other new BL drugs provided alternative treatments to penicillin, and amoxicillin is now the most prescribed BL in Europe. Diagnosis of BL allergy is mainly based on skin tests and drug provocation tests, defining different sensitization patterns or phenotypes. In this study, we evaluated 619 patients with a confirmed diagnosis of BL-immediate allergy during the last 25 years, using the same diagnostic procedures with minor adaptations to the successive guidelines. The initial eliciting drug was benzylpenicillin, which changed to amoxicillin with or without clavulanic acid and cephalosporins in recent years. In skin tests, we found a decrease in sensitivity to major and minor penicillin determinants and an increase in sensitivity to amoxicillin and others; this might reflect that the changes in prescription could have influenced the sensitization patterns, thus increasing the incidence of specific reactions to side-chain selective reactions.

Thermography based skin allergic reaction recognition by convolutional neural networks

In this work we present an automated approach to allergy recognition based on neural networks. Allergic reaction classification is an important task in modern medicine. Currently it is done by humans, which has obvious drawbacks, such as subjectivity in the process. We propose an automated method to classify prick allergic reactions using correlated visible-spectrum and thermal images of a patient’s forearm. We test our model on a real-life dataset of 100 patients (1584 separate allergen injections). Our solution yields good results—0.98 ROC AUC; 0.97 AP; 93.6% accuracy. Additionally, we present a method to segment separate allergen injection areas from the image of the patient’s forearm (multiple injections per forearm). The proposed approach can possibly reduce the time of an examination, while taking into consideration more information than possible by human staff.

Key use cases for artificial intelligence to reduce the frequency of adverse drug events: a scoping review

Adverse drug events (ADEs) represent one of the most prevalent types of health-care-related harm, and there is substantial room for improvement in the way that they are currently predicted and detected. We conducted a scoping review to identify key use cases in which artificial intelligence (AI) could be leveraged to reduce the frequency of ADEs. We focused on modern machine learning techniques and natural language processing. 78 articles were included in the scoping review. Studies were heterogeneous and applied various AI techniques covering a wide range of medications and ADEs. We identified several key use cases in which AI could contribute to reducing the frequency and consequences of ADEs, through prediction to prevent ADEs and early detection to mitigate the effects. Most studies (73 [94%] of 78) assessed technical algorithm performance, and few studies evaluated the use of AI in clinical settings. Most articles (58 [74%] of 78) were published within the past 5 years, highlighting an emerging area of study. Availability of new types of data, such as genetic information, and access to unstructured clinical notes might further advance the field.

Role of clinical history in beta-lactam hypersensitivity

PURPOSE OF REVIEW

Understand how the clinical history has been used to risk stratify patients reporting a beta-lactam allergy, both in clinical care pathways and predictive models.

RECENT FINDINGS

Drug allergy clinical care pathways have emerged as a safe and effective method of stratifying patients with a reported beta-lactam allergy into risk categories, with 'low-risk' patients able to proceed straight to direct challenges or test doses. These methods have streamlined antibiotic stewardship policies and penicillin allergy de-labeling. However, how to define 'low-risk' has been subject to much debate. New research has developed predictive models that utilize the clinical history to assess a patient's true risk of beta-lactam allergy.

SUMMARY

The clinical history has long been an essential part of drug allergy evaluation and has proven invaluable within the past decade in the development of drug allergy clinical pathways. Evidence-based predictive models that use the clinical history to assess a patient's true risk of beta-lactam allergy offer tremendous promise, but differ in crucial areas such as the populations they study, the predictor variables they use, and the ultimate accuracy they attain. These models highlight key aspects of the drug allergy history and pave the way for future large-scale research.

Urticaria: The 1-1-1 Criterion for Optimized Risk Atratification in β-Lactam Allergy Delabeling

BACKGROUND

A spurious label of β-lactam allergy compromises antibiotic stewardship. Delabeling protocols based on direct challenges (ie, not preceded by allergy tests) can be applied in low-risk patients.

OBJECTIVE

This study aims at determining the significance of the characteristics of urticaria in the risk stratification for delabeling.

METHODS

The characteristics of urticarial eruptions that had occurred during therapeutic courses with a β-lactam, namely the time interval between the exposure and onset, the dose (first or subsequent) after which urticaria appeared, and the duration of the eruption, were correlated to the results of a systematic allergy workup (skin tests, specific IgE measurements, and challenges). Data from 410 patients enrolled in 3 allergy centers (Rome and Troina, Italy, and Antwerp, Belgium) were analyzed. A multivariable logistic regression was performed, which included appearance within 1 hour after the first dose and regression within 1 day: a model that can be summarized as the "1-1-1" urticaria criterion.

RESULTS

An urticarial eruption that had appeared within 1 hour after the first dose and had regressed within 1 day was more frequently reported in the group with a positive allergy workup, with odds ratios of 17 (95% confidence interval [CI]: 9-31), 11 (95% CI: 6-20), and 48 (95% CI: 14-157), respectively (P < .005). The 1-1-1 criterion displayed a sensitivity and specificity of 85%, and a negative predictive value and a positive predictive value of 80% and 90%, respectively.

CONCLUSION

Patients with urticaria meeting the 1-1-1 criterion should be considered at high risk and referred for an allergy workup with skin testing and specific IgE measurement before challenging.