The optimal treatment regimen for breast cancer patients with gBRCA mutations remains controversial given the availability of numerous options, such as platinum-based agents, polymerase inhibitors (PARPis), and other agents. We included phase II or III RCTs and estimated the HR with 95% CI for OS, PFS, and DFS, in addition to the OR with 95% CI for ORR and pCR. We determined the treatment arm rankings by P-scores. Furthermore, we carried out a subgroup analysis in TNBC and HR-positive patients. We conducted this network meta-analysis using R 4.2.0 and a random-effects model. A total of 22 RCTs were eligible, involving 4253 patients. In the pairwise comparisons, PARPi + Platinum + Chemo was better than PARPi + Chemo for OS (in whole study group and in both subgroups) as well as PFS. The ranking tests demonstrated that PARPi + Platinum + Chemo ranked first in PFS, DFS, and ORR. Platinum + Chemo showed higher OS than PARPi + Chemo. The ranking tests for PFS, DFS, and pCR indicated that, except for the best treatment (PARPi + Platinum + Chemo) containing PARPi, the second and third treatments were platinum monotherapy or platinum-based chemotherapy. In conclusion, PARPi + Platinum + Chemo might be the best regime for gBRCA-mutated BC. Platinum drugs showed more favorable efficacy than PARPi in both combination and monotherapy.

This study aims to evaluate deep learning (DL) performance in differentiating low- and high-grade glioma. Search online database for studies continuously published from 1st January 2015 until 16th August 2022. The random-effects model was used for synthesis, based on pooled sensitivity (SE), specificity (SP), and area under the curve (AUC). Heterogeneity was estimated using the Higgins inconsistency index (I²). 33 were ultimately included in the meta-analysis. The overall pooled SE and SP were 94% and 93%, with an AUC of 0.98. There was great heterogeneity in this field. Our evidence-based study shows DL achieves high accuracy in glioma grading. Subgroup analysis reveals several limitations in this field: 1) Diagnostic trials require standard method for data merging for AI; 2) small sample size; 3) poor-quality image preprocessing; 4) not standard algorithm development; 5) not standard data report; 6) different definition of HGG and LGG; and 7) poor extrapolation.