BACKGROUND: In patients with metastatic colorectal cancer (mCRC) receiving highly active first-line combination treatments, early tumor shrinkage (ETS) and depth of response (DoR) are associated with survival, but their influence on outcomes during maintenance therapy is unknown. The Valentino study showed inferior PFS in 229 RAS wild-type mCRC patients randomized to panitumumab plus FOLFOX followed by maintenance with panitumumab vs. panitumumab + 5-FU/LV.PATIENTS AND METHODS: After blinded independent central review of ETS (≥20% reduction of the sum of target lesions) and DoR in patients enrolled in Valentino, the prognostic and predictive role of such parameters was investigated, along with their combination with PRESSING panel (uncommon genomic alterations associated with anti-EGFRs resistance beyond RAS and BRAF).RESULTS: One hundred and ninety-six patients were included (ETS in 132 [67.3%], median DoR: 44.1%). Both ETS and DoR ≥34% were associated with longer mPFS (p = 0.010 and p <0.001) and mOS (p = 0.006 and p <0.001). The PFS benefit of 5-FU/LV added to panitumumab maintenance, reported in the study, was independent from ETS and DoR status (interaction tests NS for both PFS and OS). However, outcomes were extremely poor in patients who received single-agent panitumumab and had no-ETS (mPFS and mOS: 7.7 and 18.7 months) or DoR <34% (mPFS and mOS: 6.5 and 18 months). Combining PRESSING panel ('molecular hyperselection') and response dynamics allowed to stratify both PFS (p <0.001 and p <0.001 for ETS and DoR, respectively) and OS (p <0.001 and p = 0.017 for ETS and DoR, respectively).CONCLUSIONS: ETS and DoR allow on-treatment anticipation of outcomes following an anti-EGFR-based strategy planning de-escalation, and poor radiological response may guide enrolment in crossover strategy trials. As in vivo markers of drug sensitivity, ETS and DoR may be integrated with several patient- and tumor-related factors to wisely drive decision-making on upfront treatment duration and intensity.