Dataset Information

Assessing Clinical Disease Recurrence Using Laboratory Data in Surgically Resected Patients From the TOPPIC Trial.

ABSTRACT:

Background

Machine learning methodologies play an important role in predicting progression of disease or responses to medical therapy. We previously derived and validated a machine learning algorithm to predict response to thiopurines in an inflammatory bowel disease population. We aimed to apply a modified algorithm to predict postsurgical treatment response using clinical trial data.

Methods

TOPPIC was a multicenter randomized double-blinded placebo-controlled trial of 240 patients, evaluating the effectiveness of 6-mercaptopurine in preventing or delaying postsurgical Crohn disease recurrence. We adapted a well-established machine learning algorithm to predict clinical recurrence postresection using age and multiple laboratory-specific covariates, and compared this to the thiopurine metabolite, 6-thioguanine.

Results

The random forest machine learning algorithm demonstrates a mean under the receiver operator curve (AuROC) of 0.62 [95% confidence interval (CI) 0.47, 0.78]. Similar results were evident when adding thiopurine metabolite (6-thioguanine) results. Alanine aminotransferase/mean corpuscular volume (ALT/MCV) and potassium × alkaline phosphatase (POT × ALK) predicted endoscopic and biologic recurrence, respectively, with AuROCs of 0.714 (95% CI 0.601, 0.827) and 0.730 (95% CI 0.618, 0.841).

Conclusions

A machine learning algorithm with laboratory data from within the first 3 months postsurgically does not discriminate clinical recurrence well. Alternative noninvasive measures should be considered and further evaluated.

SUBMITTER: Waljee AK

PROVIDER: S-EPMC9802488 | biostudies-literature | 2020 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Assessing Clinical Disease Recurrence Using Laboratory Data in Surgically Resected Patients From the TOPPIC Trial.

Waljee Akbar K AK Cohen-Mekelburg Shirley S Liu Yumu Y Liu Boang B Zhu Ji J Higgins Peter D R PDR

Crohn's & colitis 360 20201023 4

<h4>Background</h4>Machine learning methodologies play an important role in predicting progression of disease or responses to medical therapy. We previously derived and validated a machine learning algorithm to predict response to thiopurines in an inflammatory bowel disease population. We aimed to apply a modified algorithm to predict postsurgical treatment response using clinical trial data.<h4>Methods</h4>TOPPIC was a multicenter randomized double-blinded placebo-controlled trial of 240 patie ...[more]

PMID: 36777756

Similar Datasets

Project description:IntroductionAs recently evidenced by the ADAURA trial, most patients with stages IB to IIIA of resected EGFR-mutant lung adenocarcinoma benefit from osimertinib as adjuvant therapy. Nevertheless, predictive markers of response and recurrence are still an unmet need for more than 10% of these patients. Some circular RNAs (circRNAs) have been reported to play a role in tumor growth and proliferation. In this project, we studied circRNA expression levels in formalin-fixed, paraffin-embedded lung tumor samples to explore their biomarker potential and develop a machine learning (ML)-based signature that could predict the benefit of adjuvant EGFR tyrosine kinase inhibitors in patients with EGFR-mutant NSCLC.MethodsPatients with surgically resected EGFR mutant-positive, stages I to IIIB NSCLC were recruited from February 2007 to December 2015. Formalin-fixed, paraffin-embedded tumor samples were retrospectively collected from those patients with a follow-up period of more than or equal to 36 months (N = 76). Clinicopathologic features were annotated. Total RNA was purified and quantified prior nCounter processing with our circRNA custom panel. Data analysis and ML were performed taking into consideration circRNA expression levels and recurrence-free survival (RFS). RFS was defined from the day of surgery to the day when recurrence was detected radiologically or the death owing to any cause.ResultsOf the 76 patients with EGFR mutation-positive NSCLC included in the study, 34 relapsed within 3 years after resection. The median age of the relapsing cohort was 71.5 (range: 49-89) years. Most patients were nonsmokers (n = 21; 61.8%) and of female sex (n = 21; 61.8%). Most cases (n = 17; 50%) presented an exon 21 mutation, whereas 15 and four patients had an exon 19 and exon 18 mutation, respectively. Differential expression analysis revealed that circRUNX1, along with circFUT8 and circAASDH, was up-regulated in relapsing patients (p < 0.05 and >2 fold-change). A ML-based circRNA signature predictive of recurrence in patients with EGFR mutation-positive NSCLC, comprising circRUNX1, was developed. Our final model including selected 6-circRNA signature with random forest algorithm was able to classify relapsing patients with an accuracy of 83% and an area under the receiver operating characteristic curve of 0.91.RFS was significantly shorter not only for the subgroup of patients with high versus low circRUNX1 expression but also for the group classified as recurrent by the ML circRNA signature when compared with those classified as nonrecurrent.ConclusionsOur findings suggest that circRUNX1 and the presented ML-developed signature could be novel tools to predict the benefit of adjuvant EGFR tyrosine kinase inhibitors with regard to RFS in patients with EGFR-mutant NSCLC. The training and validation phases of our ML signature will be conducted including bigger independent cohorts.

Dataset Information

Assessing Clinical Disease Recurrence Using Laboratory Data in Surgically Resected Patients From the TOPPIC Trial.

Background

Methods

Results

Conclusions

Publications

Assessing Clinical Disease Recurrence Using Laboratory Data in Surgically Resected Patients From the TOPPIC Trial.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets