Background: Checkpoint inhibitors (CPI) targeting the PD1/PD-L1 axis improved outcomes for stage IV non-small cell lung cancer (NSCLC). Poor response to PD-L1 antibody has been linked to increased cancer-associated fibroblasts (CAF), which interact with cytotoxic T cells (CTLs) similarly to regulatory T cells (Tregs). CAFs, through cytokine production, impair antitumor immunity by affecting CTL function (TGF beta) and inhibiting CTL recruitment into tumors, suggesting that CAFs could be a therapeutic target in CPI-resistant NSCLC. Pirfenidone (P) exhibits anti-fibrotic effects by blocking fibroblast differentiation into CAFs and suppressing TGF beta production and associated signaling pathways/collagens. Atezolizumab (A) targets PD-L1, inhibiting its interaction with PD-1 and B7-1 (CD80) receptors on T cells. We initiated a phase I/II trial to test the combination of AP in recurrent NSCLC after CPI progression. The phase I study aims to determine the maximum tolerated dose (MTD) of AP and assess its safety and tolerability. The phase II study will evaluate its efficacy. The Exploratory Objectives are to measure circulating levels of TGF beta and investigate the expression of CAF-related proteins. Methods: phase I will enroll 3 patients using P at 801 mg PO TID and A at 1200mg IV every 3 weeks. If there is ≤ 1 Dose Limiting Toxicity (DLT), the study will proceed to phase II. If there are ≥2 DLTs, P will be reduced to 534 mg TID. If there is ≤ 1 DLT at this dose, we will proceed to phase II. If there are ≥2 DLTs, then the study will be terminated. Phase II will enroll 16 patients to assess efficacy. The main inclusion criteria are patients with recurrent NSCLC without actionable mutation, after progression with first-line therapy CPI with or without platinum-based doublet chemotherapy, no more than 2 lines of therapy, measurable disease, ECOG 0-2, and adequate organ function. Results: The study enrolled a total of 5 patients (4 males, 1 female), median age of 63.4; 5 were white; 3 patients did not complete the required 21 days of therapy for evaluation (1 non-compliance, 1 disease progression,1 unrelated Urinary Tract Infection); 2 patients were evaluable for toxicity, without DLT; 1 exhibited a partial response to cutaneous metastatic disease, with overall stable disease, 1 had disease progression. Adverse events (AE) in more than 5% of patients included blood disorders, cardiac, gastrointestinal, administration site, metabolic, musculoskeletal, psychiatric, respiratory, and skin disorders. Serious AEs were noted in 2 patients: one experienced gastrointestinal bowel perforation due to disease progression resulting in death, and the other had a respiratory disorder. Conclusions: The combination of AP did not produce unexpected side effects or reach the dose-limiting toxicity threshold. After the phase I portion is completed, the study will proceed to phase II expansion to enroll patients with recurrent NSCLC. Clinical trial information: NCT04467723.