Comprehensive analysis of first-line, maintenance, and subsequent-line therapies for metastatic colorectal cancer: a systematic review, meta-analysis, and network meta-analysis

Background: Treatment sequencing in metastatic colorectal cancer (mCRC) remains controversial despite advances in systemic therapy. We conducted a systematic review and meta-analysis to evaluate the efficacy and safety of current strategies across treatment lines. Methods: A comprehensive search of major databases and conference proceedings was conducted from inception to August 2024. Eligible studies were phase II–III randomized controlled trials (RCTs) in mCRC patients, reporting overall survival (OS), progression-free survival (PFS), or grade ≥3 adverse events. Pairwise and network meta-analyses were performed using both Bayesian and frequentist models across first-line, maintenance, and subsequent-line settings, with subgroup analyses by RAS and BRAF mutation status. Results: We included 100 RCTs (41 215 patients) evaluating targeted therapy (Tar), chemotherapy (Chemo), Chemo + Tar, local plus systemic therapy (Local + ST), immunotherapy (IO), Chemo + Tar + IO, supportive care (SC), and Tar + IO. For first-line, Chemo + Tar improved PFS over Chemo (HR 0.83, 95% CI 0.75–0.93), but the OS benefit was not significant. Local + ST achieved the best PFS and OS outcomes, outperforming Chemo (HR 0.73; 95% CI, [0.52–0.99]) and Chemo + Tar (0.83 [0.72–0.93]). For maintenance therapy, Chemo provided the greatest PFS benefit, significantly better than SC (0.34 [0.13–0.74]). Chemo + Tar (0.81 [0.25–2.48]) and Tar (0.60 [0.19–1.71]) showed favorable but non-significant trends. De-escalation improved tolerability but compromised disease control. For subsequent-line, Chemo + Tar improved PFS over Chemo (0.70 [0.61–0.81]), while IO improved OS vs SC (HR 0.72, 95% CI 0.54–0.97). Tar + IO ranked highest for OS, suggesting potential synergy. Local + ST again showed the best PFS, outperforming SC (0.34 [0.18–0.63]) and Chemo (0.59 [0.42–0.83]). Subgroup analyses revealed distinct responses by molecular subtype, notably among RAS wild-type and BRAF mutant patients. Pairwise and network analyses yielded similar results. Conclusions: Local + ST benefits well-selected patients across lines. Chemo + Tar remains standard in first-line. Chemo-based maintenance offers the best PFS; de-escalation improves tolerance but reduces efficacy. In subsequent-line, Tar + IO shows promising OS, and Chemo + Tar remains effective.