Background: Independent early dynamic assessment (baseline [D0] and day 15 of first cycle [D15]) of both TKa and ctDNA was prognostic and predictive in pts with HR+, HER2− ABC treated with RIB+LET enrolled in the BioItaLEE trial (NCT03439046). Here we performed a combined analysis of these two biomarkers. Methods: 287 pts were enrolled in the study. Overall, early dynamics were assessable for both biomarkers in 241/287 pts (84.0%). Methods applied for ctDNA and TKa evaluation were previously reported. For ctDNA, samples were defined as wild type (WT) if no mutations were observed at D0 and D15, ctDNA positive (+) if with or negative (-) if without a primary target mutation at D15. Samples were TKa+ or TKa- if TKa levels were above or below the limit of detection at D15. According to ctDNA and TKa pts were classified as: WT/TKa-, WT/TKa+, ctDNA-/TKa-, ctDNA-/TKa+, ctDNA+/TKa- and ctDNA+/TKa+ and then divided into 3 main study groups (GRs) WT/TKa- (GR1, n = 126), WT/TKa+, ctDNA-/TKa-, ctDNA-/TKa+, ctDNA+/TKa- (GR2, n = 96) and ctDNA+/TKa+ (GR3, n = 19). The association between biomarkers and PFS (progression-free survival) was estimated using Kaplan-Meier analysis and multivariate Cox models with 95% confidence intervals (CIs) adjusted for clinical variables. Results: Median follow-up was 26.9 months. In multivariate Cox models both TKa dynamics and mutational tumor burden at D15 were independently predictive of PFS. Hazard ratios (HRs) were 0.37 (95% CI: 0.23-0.60; p < 0.0001) for WT vs ctDNA+ and 0.56 (95% CI: 0.32-1.00; p = 0.0506) for ctDNA- vs ctDNA+. For TKa, HR was 0.49 (95% CI: 0.30-0.80; p = 0.0040) in TKa- vs TKa+. Interestingly combining the two variables further improve prediction of outcome. HRs for TKa- vs TKa+ were 0.17 (95% CI: 0.09-0.32; p < 0.0001), 0.28 (95% CI: 0.13-0.59; p = 0.0009) and 0.44 (95% CI: 0.23-0.86; p = 0.0169) in WT, ctDNA- and ctDNA+ pts, respectively. Considering the 3 study GRs, median PFSs (95% CI) were not reached (27.89, NE), 19.58 (13.83, 23.39) and 6.65 (2.83, 12.16) months in GR1, GR2 and GR3, respectively, p < 0.001. At multivariate Cox models, HRs of GR1 and GR2 compared with GR3 were 0.17 (95% CI: 0.09-0.32; p < 0.0001) and 0.37 (95% CI: 0.20-0.67; p = 0.001) respectively. Conclusions: These findings suggest that combining the early dynamic assessment of both ctDNA and TKa may improve outcome prediction in pts treated with RIB+LET. Pts with ctDNA+/TKa+ are strongly enriched for non-responders. TKa and ctDNA capture different features of tumor biological activity and their combination warrants further evaluation in relation to other treatments, settings, and diseases. Clinical trial information: NCT03439046.