Background: In the context of perioperative immunotherapy, it is crucial to personalize adjuvant treatment, identifying which patients should or should not receive post-surgical immunotherapy. Here we describe the changes in B cell peripheral blood immunophenotype during perioperative immunotherapy and its potential to predict disease progression in patients from NADIM I clinical trial (NCT03081689). Methods: Blood samples from 42 patients were obtained at pre-, post-neoadjuvant chemoimmunotherapy (pre-surgery) and at 6 and 12 months of adjuvant IO. Peripheral mononuclear cells were stained with CD19, CD20, CD38, CD24, CD27, CD25, CD10 antibody panel and analyzed by flow cytometry using FlowJo. Statistical analysis in paired samples was performed with Wilcoxon test and differences in parameters’ fold changes between disease status groups (Progression-free at 34.2 months) were determined by Mann-Whitney U tests. Variables cutoffs were stablished using ROC curve analysis. Progression-free survival (PFS) was evaluated using Kaplan-Meier curves and long-rank test. The median follow-up was 38.0 months. Results: Regarding the impact of NADIM complete treatment scheme in patient’s B cells, there was a reduction of memory B cells (CD19+CD20+CD38-CD24+CD27+; p < 0.001) and CD20 levels in total B cells (CD19+CD20+; p = 0.001) (n = 17). Importantly, this decrease occurs in all patients regardless their pathological response or disease progression. However, the magnitude of the memory B cell percentage and CD20 levels reduction was significantly more pronounced in patients whose disease progress thereafter (p = 0.023, p = 0.052 at 6 and 12 months for memory B cells; and p = 0.002, p = 0.009 at 6 and 12 months for CD20 expression). Patients with a relative drop of memory B cells greater than 90% at 6 or 12 months, showed a shorter PFS (p = 0.028, p = 0.007). Likewise, patients with a relative decrease of CD20 expression on B cells higher than 60% at 6 or 12 months had shorter PFS (p = 0.016, p < 0.001). The AUC ROCs to identify patients with disease progression, were 0.824 and 0.885 at 6 and 12 months for memory B cells; and 0.912 and 0.981 at 6 and 12 months for CD20 levels. Moreover, memory B cells percentage and CD20 levels decrease occur between surgery and the first 6 months of adjuvant therapy (n = 21, p < 0.001, p = 0.025). After that, these levels are maintained from 6 to 12 months of adjuvant treatment (n = 20, p = 0.290, p = 0.682), thus the overall reduction is also observed between surgery and 12 months samples (n = 17, p < 0.001, p = 0.001). Conclusions: Perioperative IO produces a significant reduction in memory B cells and CD20 levels. Importantly, stronger decreases might be helpful to predict progression since they are associated to shorter PFS, indicating that the maintenance of the memory B cell peripheral compartment is relevant for long-term survival of these patients.