MRI versus CT: Which Modality Is Better in Gross Tumor Volume Delineation for External Beam Radiotherapy in Cancer Patients? A Systematic Review and Meta-Analysis

Abstract

Background: are a variety of imaging modalities for contouring the GTV of EBRT that strive for more accurate delineation and better concordance between observers. Nowadays, there is a great controversy about which modality (MRI or CT) is more accurate for GTV delineation. We conducted a systematic review and meta-analysis to investigate whether MRI or CT is better in GTV delineation for EBRT in cancer patients. Methods: We identified MRI, CT and GTV studies by searching PubMed, Embase and the Cochrane Library databases until April 18, 2019. We included original articles reporting GTV delineation using CT and MRI. The outcomes of interest included GTV, interobserver agreement (IOA) and intermodality agreement (IMA) among different modalities (MRI only, MRI+CT fusion, CT). The accuracy of GTV delineation of these imaging modalities was compared with pathology (path). The MD or SMD and their corresponding 95% CIs were selected as the effect size. Subgroup and sensitivity analyses were performed. Findings: Thirty-two studies encompassing 672 participants (677 lesions) were included. We did not identify any significant difference between GTVMRI only / GTVMRI+CT fusion and GTVCT (SMD: -0.02, 95% CI: -0.20~0.17, P=0.84; SMD:0.53, 95% CI: -0.22~1.28, P=0.17 respectively). GTVCT was larger than GTVpath (SMD:0.92, 0.26~1.58, P=0.006), whereas no difference was observed when MRI only was compared with path (SMD:0.71, -0.36~1.78, P=0.19). No significant difference was observed between the IOA of MRI only / MRI+CT fusion and CT (MD: 0.00, -0.09~0.08, P=0.92; MD: 0.03, -0.01~0.08, P=0.14, respectively). Subgroup analysis showed that GTVMRI+CT fusion was significantly larger for the delineation of nasopharyngeal cancer (NPC) (SMD: 6.83, 5.89~7.78, P<0.001). MRI only had the highest tumor coverage by comparing with path (Mean:0.88, 0.82~0.94), followed by CT (Mean:0.85, 0.79~0.91). Only the IMA of MRI only versus MRI+CT fusion showed good concordance (Mean:0.84, 0.79~0.89) among the three pairs composed of three images (MRI only, MRI+CT fusion, and CT); all others showed moderate concordance. Interpretation: CT, MRI only and MRI+CT fusion for GTV delineation are all clinically applicable for preoperative adjuvant radiotherapy. Prospective studies emphasizing the collection of patient-reported data and functional follow-ups are strongly recommended. Funding Statement: The authors state: There was no funding source for this study. Declaration of Interests: The authors declare: None. Ethics Approval Statement: The authors declare: Not applicable. The authors conducted this systematic review and meta-analysis in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. The study protocol was registered with PROSPERO (Number CRD4209127649).