Is radiomic MRI a feasible alternative to OncotypeDX® recurrence score testing? A systematic review and meta-analysis

Clinical applications of radiomics and deep learning in breast and lung cancer: A narrative literature review on current evidence and future perspectives

Radiomics, analysing quantitative features from medical imaging, has rapidly become an emerging field in translational oncology. Radiomics has been investigated in several neoplastic malignancies as it might allow for a non-invasive tumour characterization and for the identification of predictive and prognostic biomarkers. Over the last few years, evidence has been accumulating regarding potential clinical applications of machine learning in many crucial moments of cancer patients' history. However, the incorporation of radiomics in clinical decision-making process is still limited by low data reproducibility and study variability. Moreover, the need for prospective validations and standardizations is emerging. In this narrative review, we summarize current evidence regarding radiomic applications in high-incidence cancers (breast and lung) for screening, diagnosis, staging, treatment choice, response, and clinical outcome evaluation. We also discuss pro and cons of the radiomic approach, suggesting possible solutions to critical issues which might invalidate radiomics studies and propose future perspectives.

The use of radiomic analysis of magnetic resonance imaging findings in predicting features of early osteoarthritis of the knee-a systematic review and meta-analysis

The primary aim of this study was to systematically review current literature evaluating the use of radiomics in establishing the role of magnetic resonance imaging (MRI) findings in native knees in predicting features of osteoarthritis (OA). A systematic review was performed with respect to PRISMA guidelines in search of studies reporting radiomic analysis of magnetic resonance imaging (MRI) to analyse patients with native knee OA. Sensitivity and specificity of radiomic analyses were included for meta-analysis. Following our initial literature search of 1271 studies, only 5 studies met our inclusion criteria. This included 1730 patients (71.5% females) with a mean age of 55.4 ± 15.6 years (range 24-66). The mean RQS of included studies was 16.6 (11-21). Meta-analysis demonstrated the pooled sensitivity and specificity for MRI in predicting features of OA in patients with native knees were 0.74 (95% CI 0.71, 0.78) and 0.85 (95% CI 0.83, 0.87), respectively. The results of this systematic review suggest that the high sensitivities and specificity of MRI-based radiomics may represent potential biomarker in the early identification and classification of native knee OA. Such analysis may inform surgeons to facilitate earlier non-operative management of knee OA in the select pre-symptomatic patients, prior to clinical or radiological evidence of degenerative change.

An overview of meta-analyses on radiomics: more evidence is needed to support clinical translation

OBJECTIVE

To conduct an overview of meta-analyses of radiomics studies assessing their study quality and evidence level.

METHODS

A systematical search was updated via peer-reviewed electronic databases, preprint servers, and systematic review protocol registers until 15 November 2022. Systematic reviews with meta-analysis of primary radiomics studies were included. Their reporting transparency, methodological quality, and risk of bias were assessed by PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) 2020 checklist, AMSTAR-2 (A MeaSurement Tool to Assess systematic Reviews, version 2) tool, and ROBIS (Risk Of Bias In Systematic reviews) tool, respectively. The evidence level supporting the radiomics for clinical use was rated.

RESULTS

We identified 44 systematic reviews with meta-analyses on radiomics research. The mean ± standard deviation of PRISMA adherence rate was 65 ± 9%. The AMSTAR-2 tool rated 5 and 39 systematic reviews as low and critically low confidence, respectively. The ROBIS assessment resulted low, unclear and high risk in 5, 11, and 28 systematic reviews, respectively. We reperformed 53 meta-analyses in 38 included systematic reviews. There were 3, 7, and 43 meta-analyses rated as convincing, highly suggestive, and weak levels of evidence, respectively. The convincing level of evidence was rated in (1) T2-FLAIR radiomics for IDH-mutant vs IDH-wide type differentiation in low-grade glioma, (2) CT radiomics for COVID-19 vs other viral pneumonia differentiation, and (3) MRI radiomics for high-grade glioma vs brain metastasis differentiation.

CONCLUSIONS

The systematic reviews on radiomics were with suboptimal quality. A limited number of radiomics approaches were supported by convincing level of evidence.

CLINICAL RELEVANCE STATEMENT

The evidence supporting the clinical application of radiomics are insufficient, calling for researches translating radiomics from an academic tool to a practicable adjunct towards clinical deployment.

MRI Radiomics and Machine Learning for the Prediction of Oncotype Dx Recurrence Score in Invasive Breast Cancer

AIM

To non-invasively predict Oncotype DX recurrence scores (ODXRS) in patients with ER+ HER2- invasive breast cancer (IBC) using dynamic contrast-enhanced (DCE) MRI-derived radiomics features extracted from primary tumor lesions and a ML algorithm.

MATERIALS AND METHODS

Pre-operative DCE-MRI of patients with IBC, no history of neoadjuvant therapy prior to MRI, and for which the ODXRS was available, were retrospectively selected from a public dataset. ODXRS was obtained on histological tumor samples and considered as positive if greater than 16 and 26 in patients aged under and over 50 years, respectively. Tumor lesions were manually annotated by three independent operators on DCE-MRI images through 3D ROIs positioning. Radiomic features were therefore extracted and selected using multistep feature selection process. A logistic regression ML classifier was then employed for the prediction of ODXRS.

RESULTS

248 patients were included, of which 87 with positive ODXRS. 166 (66%) patients were grouped in the training set, while 82 (33%) in the test set. A total of 1288 features was extracted. Of these, 1244 were excluded as 771, 82 and 391 were excluded as not stable (n = 771), not variant (n = 82), and highly intercorrelated (n = 391), respectively. After the use of recursive feature elimination with logistic regression estimator and polynomial transformation, 92 features were finally selected. In the training set, the logistic regression classifier obtained an overall mean accuracy of 60%. In the test set, the accuracy of the ML classifier was 63%, with a sensitivity of 80%, specificity of 43%, and AUC of 66%.

CONCLUSIONS

Radiomics and ML applied to pre-operative DCE-MRI in patients with IBC showed promises for the non-invasive prediction of ODXRS, aiding in selecting patients who will benefit from NAC.

A Novel Surrogate Nomogram Capable of Predicting OncotypeDX Recurrence Score©

Background: OncotypeDX Recurrence Score© (RS) is a commercially available 21-gene expression assay which estimates prognosis and guides chemoendocrine prescription in early-stage estrogen-receptor positive, human epidermal growth factor receptor-2-negative (ER+/HER2−) breast cancer. Limitations of RS testing include the cost and turnaround time of several weeks. Aim: Our aim is to develop a user-friendly surrogate nomogram capable of predicting RS. Methods: Multivariable linear regression analyses were performed to determine predictors of RS and RS > 25. Receiver operating characteristic analysis produced an area under the curve (AUC) for each model, with training and test sets were composed of 70.3% (n = 315) and 29.7% (n = 133). A dynamic, user-friendly nomogram was built to predict RS using R (version 4.0.3). Results: 448 consecutive patients who underwent RS testing were included (median age: 58 years). Using multivariable regression analyses, postmenopausal status (β-Coefficient: 0.25, 95% confidence intervals (CIs): 0.03−0.48, p = 0.028), grade 3 disease (β-Coefficient: 0.28, 95% CIs: 0.03−0.52, p = 0.026), and estrogen receptor (ER) score (β-Coefficient: −0.14, 95% CIs: −0.22−−0.06, p = 0.001) all independently predicted RS, with AUC of 0.719. Using multivariable regression analyses, grade 3 disease (odds ratio (OR): 5.67, 95% CIs: 1.32−40.00, p = 0.037), decreased ER score (OR: 1.33, 95% CIs: 1.02−1.66, p = 0.050) and decreased progesterone receptor score (OR: 1.16, 95% CIs: 1.06−1.25, p = 0.002) all independently predicted RS > 25, with AUC of 0.740 for the static and dynamic online nomogram model. Conclusions: This study designed and validated an online user-friendly nomogram from routinely available clinicopathological parameters capable of predicting outcomes of the 21-gene RS expression assay.

Value of the 21-gene expression assay in predicting locoregional recurrence rates in estrogen receptor-positive breast cancer: a systematic review and network meta-analysis

Purpose

The Oncotype DX© 21-gene Recurrence Score (RS) estimates the risk of distant disease recurrence in early-stage estrogen receptor-positive, human epidermal growth factor receptor-2-negative (ER+/HER2− ) breast cancer. Using RS to estimate risk of locoregional recurrence (LRR) is less conclusive. We aimed to perform network meta-analysis (NMA) evaluating the RS in estimating LRR in ER+/HER2− breast cancer.

Methods

A NMA was performed according to PRISMA-NMA guidelines. Analysis was performed using R packages and Shiny.

Results

16 studies with 21,037 patients were included (mean age: 55.1 years (range: 22–96)). The mean RS was 17.1 and mean follow-up was 66.4 months. Using traditional RS cut-offs, 49.7% of patients had RS < 18 (3944/7935), 33.8% had RS 18–30 (2680/7935), and 16.5% had RS > 30 (1311/7935). Patients with RS 18–30 (risk ratio (RR): 1.76, 95% confidence interval (CI): 1.32–2.37) and RS > 30 (RR: 3.45, 95% CI: 2.63–4.53) were significantly more likely to experience LRR than those with RS < 18. Using TAILORx cut-offs, 16.2% of patients had RS < 11 (1974/12,208), 65.8% had RS 11–25 (8036/12,208), and 18.0% with RS > 30 (2198/12,208). LRR rates were similar for patients with RS 11–25 (RR: 1.120, 95% CI: 0.520–2.410); however, those with RS > 25 had an increased risk of LRR (RR: 2.490, 95% CI: 0.680–9.390) compared to those with RS < 11. There was a stepwise increase in LRR rates when applying traditional and TAILORx cut-offs (both P < 0.050).

Conclusion

RS testing accurately estimates LRR risk for patients being treated for early-stage ER+/HER2− breast cancer. Future prospective, randomized studies may validate the predictive value of RS in estimating LRR.