A Machine Learning Approach to Predict Remission in Patients With Psoriatic Arthritis on Treatment With Secukinumab

Venerito, Vincenzo; Lopalco, Giuseppe; Abbruzzese, Anna; Colella, Sergio; Morrone, Maria Concetta; Tangaro, Sabina; Iannone, Florenzo

doi:10.3389/fimmu.2022.917939

Cited by 9 publications

(5 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is a group of metabolic abnormalities that include hypertension, obesity, dyslipidaemia and insulin resistance ( 16 ). MetS is also thought to be a comorbid condition of many immune diseases, such as psoriatic arthritis (PsA) and OA ( 17 , 18 ). No doubt about it, OA and MetS are both leading public health problems with increasing rates of illness and disability.…”

Section: Discussionmentioning

confidence: 99%

Identification of S100A8 as a common diagnostic biomarkers and exploring potential pathogenesis for osteoarthritis and metabolic syndrome

Liu

Huang

2023

Front. Immunol.

View full text Add to dashboard Cite

BackgroundOsteoarthritis (OA) is the most frequent musculoskeletal disease and the major contributor to disability worldwide. Metabolic syndrome (MetS) has been recognized as being associated with the pathogenesis of osteoarthritis. However, the exact mechanisms and links between the two are not clear.MethodsWe downloaded clinical information data and gene expression profiles for OA and MetS from the database of Gene Expression Omnibus (GEO), and immune related gene (IRG) from the database of Immunology Database and Analysis Portal (IMMPORT). After screening OA-DEG and MetS-DEG, we identified the common immune hub gene by screening the overlapping genes between OA-DEG, MetS-DEG and IRG. Then we conducted single-gene analysis of S100A8, assessed the correlation of S100A8 with immune cell infiltration, and verified the diagnostic value of S100A8 in OA and MetS database respectively.Results323 OA-DEGs,101 MetS-DEGs and an immune-related hub gene, S100A8, were identified. In single gene analysis of S100A8 in OA samples, GSEA suggested that immune-related biological processes were more significantly enriched. The results of immune cell infiltration analysis showed that the enrichment fraction of M2 macrophages was significantly higher in the high S100A8-expressing group, and the level of S100A8 expression was positively correlated with M2 macrophage infiltration. The results of the dataset validation showed that S100A8 expression levels were significantly upregulated in the OA group and performed well in the diagnosis of OA. In single gene analysis of S100A8 in MetS samples, immune cell infiltration analysis showed that monocyte infiltration was higher in the S100A8 high expression samples and that there was a positive correlation between the two. Dataset validation showed that S100A8 is of high value for the diagnosis of MetS. In the validation of the dataset for the four metabolism-related diseases (obesity, diabetes, hypertension and hyperlipidaemia), S100A8 was expressed at higher levels in the disease group and also had a higher diagnostic value for the four metabolism-related diseases.ConclusionS100A8 is a common hub gene and diagnostic biomarker for OA and MetS, and the immune regulation involved in S100A8 may play a central role in the pathogenesis of OA and MetS.

show abstract

Section: Discussionmentioning

confidence: 99%

Identification of S100A8 as a common diagnostic biomarkers and exploring potential pathogenesis for osteoarthritis and metabolic syndrome

Liu

Huang

2023

Front. Immunol.

View full text Add to dashboard Cite

show abstract

“…For classi cation with small training samples and high dimensionality, feature selection plays an essential role in avoiding over tting and improving classi cation performance. One commonly used feature selection method for small samples problems is the wrapper feature selection using the recursive feature elimination (RFE) algorithm (27,28). RFE needs an algorithm to be embedded.…”

Section: Methodsmentioning

confidence: 99%

“…RFE can generate different subsets of features based on various criteria. The subgroup generated in each step will be used to build a model and train the learning algorithm iteratively (27,29). This is achieved by tting the given ML algorithm used in the RFE core, ranking features by importance, discarding the least important features, and re-tting the model (Supplementary Material).…”

Section: Methodsmentioning

confidence: 99%

Machine learning applied to MRI evaluation for the detection of lymph node metastasis in patients with locally advanced cervical cancer treated with neoadjuvant chemotherapy

Arezzo

Cormio

Mongelli

et al. 2022

Arch Gynecol Obstet

Self Cite

View full text Add to dashboard Cite

Purpose: Concurrent cisplatin-based chemotherapy and radiotherapy (CCRT) plus brachytherapy is the standard treatment for locally advanced cervical cancer (LACC). Platinum-based neoadjuvant chemotherapy (NACT) followed by radical hysterectomy is an alternative for patients with stage IB2-IIB disease. Therefore the correct pre-treatment staging is essential to the proper management of this disease. Pelvic magnetic resonance imaging (MRI) is the gold standard examination but studies about MRI accuracy in the detection of lymph node metastasis (LNM) in LACC patients show con icting data.Machine learning (ML) is emerging as a promising tool for unraveling complex non-linear relationships between patient attributes that cannot be solved by traditional statistical methods. Here we investigated whether ML might improve the accuracy of MRI in the detection of LNM in LACC patients.Methods: We analyzed retrospectively LACC patients who underwent NACT and radical hysterectomy from 2015 to 2020. Demographic, clinical and MRI characteristics before and after NACT were collected, as well as information about post-surgery histopathology. Random features elimination wrapper was used to determine an attribute core set. A ML algorithm, namely Extreme Gradient Boosting (XGBoost) was trained and validated with 10-fold cross-validation. The performances of the algorithm were assessed.Results: Our analysis included n.92 patients. FIGO stage was IB2 in n.4/92 (4.3%), IB3 in n.42/92 (45%), IIA1 in n.1/92 (1.1%), IIA2 in n.16/92 (17.4%) and IIB in n.29/92 (31.5%). Despite detected neither at pretreatment and post-treatment MRI in any patients, LNM occurred in n.16/92 (17%) patients. The attribute core set used to train ML algorithms included grading, histotypes, age, parity, largest diameter of lesion at either pre and post-treatment MRI, presence/absence of fornix in ltration at pre-treatment MRI and FIGO stage. XGBoost showed a good performance (accuracy 89%, precision 83%, recall 78%, AUROC 0.79). Conclusions:We developed an accurate model to predict LNM in LACC patients in NACT, based on a ML algorithm requiring few easy-to-collect attributes.Authors contributions: Francesca Arezzo and Vera Loizzi performed the study concepyion. Gerardo Cazzato and Claudio Lombardi contributed to the study design. Material preparation and data collection were performed by Vincenzo Venerito, Ambrogio Cazzola and Michele Mongelli.

show abstract

“…Machine learning (ML) techniques can investigate patterns from large clinical datasets and identify distinct clusters of patients with potential therapeutic or prognostic significance, leading to a better understanding of the disease and evolution towards precision medicine (5). ML can deal with complex, non-linear relationships between patient attributes which are difficult to model with statistical methodologies and hence, support the development of models predicting responses to treatments and disease progression (6). The potential of ML to identify distinct clusters of patients with potential therapeutic or prognostic significance has been demonstrated in a pooled analysis of four Phase III trials in patients with PsA (FUTURE 2-5) (5).…”

Section: Introductionmentioning

confidence: 99%

Patient clusters identified by machine learning from a pooled analysis of the clinical development programme of secukinumab in psoriatic arthritis, ankylosing spondylitis and psoriatic arthritis with axial manifestations

Baraliakos,

Pournara,

Coates

et al. 2023

Clinical and Experimental Rheumatology

View full text Add to dashboard Cite

ObjectiveTo identify patient clusters based on baseline demographics and clinical indicators. MethodsPooled baseline demographics and clinical data of secukinumab-treated patients from ten Phase III studies in psoriatic arthritis (PsA;, ankylosing spondylitis (AS; MEASURE 1-4), were analysed by machine learning (ML) algorithms. The longitudinal responses of secukinumab 300 mg versus 150 mg were investigated across the clusters and three clinical indicators of tender joints, swollen joints and enthesitis. Results patients were grouped into eight distinct clusters based on patient demographics and baseline clinical characteristics.Patients with PsA and axial manifestations (MAXIMISE) were overrepresented in clusters 6-8. Patients in cluster 6 (mean age 48 years; 46% male) were overweight with pronounced psoriasis, higher articular burden in knees, shoulders, elbows and wrists. Patients in cluster 7 (mean age 47 years; 53% male) were less overweight with lower polyarticular joint counts and tenderness of the joints of the feet, wrists and hands. Patients in cluster 8 were predominantly with AS (mean age 43 years; 64% male) with a mean body mass index of 27.3 kg/m 2 , oligoarthritis and high prevalence of spinal pain. Patients with PsA (FUTURE) were overrepresented in clusters 1-5. Longitudinal analysis showed improvements with secukinumab 300 mg versus 150 mg in clusters 6 and 8 for tender joint counts, and cluster 7 for swollen joint counts. Conclusion PsA clusters obtained by ML in pooled dataset indicate phenotypical heterogeneity of patients with PsA andaxial manifestations and overlapping features across the spondyloarthritis spectrum.

show abstract

A Machine Learning Approach to Predict Remission in Patients With Psoriatic Arthritis on Treatment With Secukinumab

Cited by 9 publications

References 22 publications

Identification of S100A8 as a common diagnostic biomarkers and exploring potential pathogenesis for osteoarthritis and metabolic syndrome

Identification of S100A8 as a common diagnostic biomarkers and exploring potential pathogenesis for osteoarthritis and metabolic syndrome

Machine learning applied to MRI evaluation for the detection of lymph node metastasis in patients with locally advanced cervical cancer treated with neoadjuvant chemotherapy

Patient clusters identified by machine learning from a pooled analysis of the clinical development programme of secukinumab in psoriatic arthritis, ankylosing spondylitis and psoriatic arthritis with axial manifestations

Contact Info

Product

Resources

About