Table 1.
Summary of 20 reviewed articles on radiogenomics in clear cell renal cell carcinoma. Nature of feature extraction is indicated by “Radiologist” if features are scored by one or more radiologists. Elsewise, software derived features are indicated by “Computational”. Number of selected features indicated in parenthesis. TAT (total adipose tissue), VAT (visceral adipose tissue), AUC (area under the curve), OR (odds ratio), HR (hazard ratio), CSS (cancer specific survival), OS (overall survival), PFS (progression free survival).
| Author | Title | Year of Publication | Patient # | Feature Extraction (Number) | ±Machine Learning | Image Phase Used | Genes Studied | Outcome |
|---|---|---|---|---|---|---|---|---|
| Karlo et al. [9] | Radiogenomics of Clear Cell Renal Cell Carcinoma: Associations between CT Imaging Features and Mutations | 2014 | 233 | Radiologist (10) | − | CT | BAP1 VHL KD5MC |
BAP1 and KD5MC: renal vein invasion (OR 3.50 and 3.89) VHL: ill-defined margin (OR 0.49), nodular enhancement (OR 2.33), intratumoral vasculature (OR 0.51) |
| Shinagare et al. [10] | Radiogenomics of clear cell renal cell carcinoma: Preliminary findings of the cancer genome atlas–renal cell carcinoma (TCGA–RCC) imaging research group | 2015 | 103 | Radiologist (6) | − | Contrast-enhanced CT | BAP1 MUC-4 |
BAP1: Ill-defined margin and calcification MUC4: Exophytic growth |
| Greco et al. [11] | Relationship between visceral adipose tissue and genetic mutations (VHL and KDM5C) in clear cell renal cell carcinoma | 2021 | 97 | Computational (3) | − | CT | KDM5C vs. VHL | KDM5C higher TAT and VAT area than VHL |
| Feng et al. [12] | Identifying BAP1 Mutations in Clear-Cell Renal Cell Carcinoma by CT Radiomics: Preliminary Findings | 2020 | 54 | Computational (58) | + (Random Forest) | CT | BAP1 | AUC 0.77 |
| Kocak et al. [13] | Machine learning-based unenhanced CT texture analysis for predicting BAP1 mutation status of clear cell renal cell carcinomas | 2020 | 65 | Computational (6) | + (Random Forest) | CT | BAP1 | AUC 0.897 |
| Ghosh et al. [14] | Imaging-genomic pipeline for identifying gene mutations using three-dimensional intra-tumor heterogeneity features | 2015 | 78 | Computational (1636) | + (Random Forest) | CT nephrographic phase | BAP1 | AUC 0.71 |
| Kocak et al. [15] | Radiogenomics in Clear Cell Renal Cell Carcinoma: Machine Learning-Based High-Dimensional Quantitative CT Texture Analysis in Predicting PBRM1 Mutation Status | 2019 | 45 | Computational (10) | + (Random Forest) | CT | PBRM1 | AUC 0.987 |
| Chen et al. [16] | Reliable gene mutation prediction in clear cell renal cell carcinoma through multi-classifier multi-objective radiogenomics model | 2018 | 57 | Computational (43) | + (6 classifier composite) | CT | VHL PBRM1 BAP1 |
AUC 0.88 0.86 0.93 Mutation status prediction |
| Marigliano et al. [17] | Radiogenomics in clear cell renal cell carcinoma: correlations between advanced CT imaging (texture analysis) and microRNAs expression | 2019 | 20 | Computational (6) | − | CT | miR-21-5p | R2 = 0.25 between entropy and change in miR-21-5p expression between tumor and surrounding parenchyma |
| Cen et al. [18] | Renal cell carcinoma: predicting RUNX3 methylation level and its consequences on survival with CT features | 2019 | 106 | Radiologist (9) | − | CT | RUNX3 methylation | High methylation: left side (OR 2.70), ill-defined margin (OR 2.69), intratumoral vascularity (OR 3.29)—AUC of 0.73 |
| Yu et al. [19] | Renal Cell Carcinoma: Predicting DNA Methylation Subtyping and Its Consequences on Overall Survival With Computed Tomography Imaging Characteristics | 2020 | 212 | Radiologist (12) | − | CT | Tumor methylation (M1-M3 subtype) | M1: >7 cm (OR 2.45), necrosis (OR 4.76) M2: necrosis (OR 0.047), enhancement (OR 0.083) M3: Long axis > median (OR 0.30), necrosis (OR 3.26) |
| Jamshidi et al. [20] | The radiogenomic risk score: construction of a prognostic quantitative, noninvasive image-based molecular assay for renal cell carcinoma | 2015 | 70 | Radiologist (4) | − | Contrast CT | SPC gene signature | RRS correlation with SPC (R = 0.45), HR 3.32 for CSS after surgery |
| Jamshidi et al. [4] | The radiogenomic risk score stratifies outcomes in a renal cell cancer phase 2 clinical trial | 2016 | 41 | Radiologist (4) | − | Contrast CT | SPC gene signature | PFS: 6 mo (high RRS) vs. >25 mo (low RRS)—After bevacizumab tx OS: 25 mo (high RRS) vs. >37 months (low RRS) |
| Bowen et al. [21] | Radiogenomics of clear cell renal cell carcinoma: associations between mRNA-based subtyping and CT imaging features | 2019 | 177 | Computational (8) | − | CT | mRNA subtyping (m1-m4) | M1: OR 2.1—well-defined margin M3: OR 0.42 (well-defined margin), OR 2.12 (renal vein involvement) |
| Yin et al. [22] | Integrative radiomics expression predicts molecular subtypes of primary clear cell renal cell carcinoma | 2018 | 8 | Computational (4) | + (Fisher’s linear discriminant analysis) | PET and MRI | Molecular subtype of ccRCC (ccA vs. ccB) | Accuracy of classification—86.96% |
| Lee et al. [23] | Integrative radiogenomics approach for risk assessment of post-operative metastasis in pathological T1 renal cell carcinoma: a pilot retrospective cohort study | 2020 | 58 | Computational (4) | + (Random Forest) | Contrast CT | Multiple gene-mediated pathways | AUC 0.955—Metastasis |
| Zhao et al. [24] | Validation of CT radiomics for prediction of distant metastasis after surgical resection in patients with clear cell renal cell carcinoma: exploring the underlying signaling pathways | 2021 | 547 | Computational (9) | + (Logistic regression) | CT | 19 gene pathway signatures | AUC 0.84—Metastasis |
| Lin et al. [25] | Radiomic profiling of clear cell renal cell carcinoma reveals subtypes with distinct prognoses and molecular pathways | 2021 | 160 | Computational (122) | + (Consensus clustering) | Unenhanced CT | VHL, MUC16, FBN2, and FLG Cell cycle related pathways |
C1: Lower OS and PFS than C2 and C3 C1: Lower VHL expression C3: Higher FBN2 expression |
| Huang et al. [26] | Exploration of an integrated prognostic model of radiogenomics features with underlying gene expression patterns in clear cell renal cell carcinoma | 2021 | 205 | Computational (4) | + (LASSO/SVM for feature selection, random forest for classification) | Contrast CT | Gene modules | AUC 0.837, 0.806 and 0.751—1-, 3-, and 5-year OS (combined radiogenomic model) |
| Zeng et al. [27] | Integrative radiogenomics analysis for predicting molecular features and survival in clear cell renal cell carcinoma | 2021 | 207 | Computational (4) | + (Random Forest) | Contrast CT | VHL, BAP1, PBRM1, SETD2, molecular subtypes (m1–m4) | AUC 0.846—5-year OS (Combined radiogenomic model) |
The # refers to number (as in number of patients).