Acquired MET alterations is one of the resistance mechanisms to advanced NSCLC patients treated with EGFR tyrosine kinase inhibitors (TKIs). Several clinical trials combined MET-TKI (such as capmatinib, tepotinib, savolitinib) with EGFR-TKI to overcome MET alterations resistance. We performed this meta-analysis to determine the efficacy and safety of MET-TKI plus EGFR-TKI combined therapy in NSCLC patients. Pubmed, Embase and the Cochrane Library were searched for relevant studies up to August 19, 2024. Data of objective response rate (ORR), disease control rate (DCR), progression-free survival (PFS), median duration of response (mDOR) and adverse events were extracted from the publications and analyzed. Six studies involving 562 patients were included in this meta-analysis. Our study showed a pooled ORR of 49.2% (95% confidence interval [CI] 0.402–0.582), a pooled DCR of 78.6% (95%CI 0.680–0.893), a mDOR of 6.85 months (95%CI 5.85–7.86), and a mPFS of 5.62 months (95%CI 4.74–6.50) in MET-TKI plus EGFR-TKI combination therapy for NSCLC patients with acquired MET-driven resistance after EGFR-TKI treatment. The pooled efficacy data suggested that combining MET-TKI with a third-generation EGFR-TKI was numerically superior to combining MET-TKI with a first-generation EGFR-TKI in patients who were T790M negative with MET-dependent resistance mechanism (ORR: 56.8% vs. 47.8%, p = 0.15; DCR: 81.6% vs. 75%, p = 0.57; mDOR: 9.08 vs. 7.00 months, p = 0.25; mPFS: 7.45 vs. 4.55 months, p = 0.05). The efficacy data of capmatinib plus EGFR-TKI, savolitinib plus EGFR-TKI and tepotinib plus EGFR-TKI (regardless of generation of EGFR-TKIs) was similar (ORR:47.7% vs. 50.7% vs. 48.8%, p = 0.96; DCR: 71.4% vs. 84.9% vs. 63.3%, p = 0.02; mDOR: NR vs. 8.4 vs. 8.01 months, p = 0.18; mPFS: 5.49 vs. 6.88 vs. 5.48 months, p = 0.56). Capmatinib subgroup seemed to demonstrate lower hepatotoxicity compared with savolitinib and tepotinib subgroups numerically (increased AST level: 12.8% vs. 18.8% vs. 17.4%, p = 0.66; increased ALT level: 14.2% vs. 17.6% vs. 20.1%, p = 0.91). And a lower occurrence rate of ≥ 3 grade TRAEs was observed in the capmatinib subgroup compared to the savolitinib or tepotinib subgroups (30.0% vs. 46.7% vs. 41.2%, p = 0.07). The findings from this meta-analysis suggest that the combination of MET-TKI and EGFR-TKI represents a promising therapeutic approach for NSCLC patients who have acquired MET alterations following EGFR-TKI treatment. Notably, the combination of MET-TKI and a third-generation EGFR-TKI demonstrated enhanced survival benefits compared to the combination with a first-generation EGFR-TKI. Furthermore, different MET-TKIs based combination therapy did not display significant differences in efficacy, while capmatinib based combination therapy showed better safety profile and lower hepatotoxicity.

Autofocus is necessary for high-throughput and real-time scanning in microscopic imaging. Traditional methods rely on complex hardware or iterative hill-climbing algorithms. Recent learning-based approaches have demonstrated remarkable efficacy in a one-shot setting, avoiding hardware modifications or iterative mechanical lens adjustments. However, in this paper, we highlight a significant challenge that the richness of image content can significantly affect autofocus performance. When the image content is sparse, previous autofocus methods, whether traditional climbing-hill or learning-based, tend to fail. To tackle this, we propose a content-importance-based solution, named SparseFocus, featuring a novel two-stage pipeline. The first stage measures the importance of regions within the image, while the second stage calculates the defocus distance from selected important regions. To validate our approach and benefit the research community, we collect a large-scale dataset comprising millions of labelled defocused images, encompassing both dense, sparse and extremely sparse scenarios. Experimental results show that SparseFocus surpasses existing methods, effectively handling all levels of content sparsity. Moreover, we integrate SparseFocus into our Whole Slide Imaging (WSI) system that performs well in real-world applications. The code and dataset will be made available upon the publication of this paper.