Association of High‐Grade Components < 20% on DFS in Stage I Lung Adenocarcinoma: A Real‐World Study of Risk Analysis and Predictive Nomogram Development

OBJECTIVE: To assess whether high-grade components occupying < 20% of the tumor have prognostic value and whether an Adjusted Grade nomogram improves risk stratification in Stage I invasive lung adenocarcinoma (LUAD). METHODS: We retrospectively analyzed 1539 patients with pathologically confirmed Stage I invasive LUAD undergoing R0 resection; 1321 with follow-up, including 967 with preoperative CT. Tumors were classified by IASLC Grade and reclassified according to predominant pattern and high-grade components; Adjusted Grade 3 was subdivided into MHGC ≤ 5%, DHGC > 5% to < 20%, and SHGC ≥ 20%. Cox models identified DFS predictors. Nomograms were internally validated with 1000 bootstrap resamples and assessed for discrimination, calibration, and net benefit. RESULTS: During a median 38-month follow-up, 124 DFS events occurred. Adjusted Grade 3 independently predicted recurrence (HR 3.07, 95% CI 1.54-6.12), while MHGC and DHGC were associated with poorer DFS (HR 3.95, 95% CI 1.53-10.15; HR 2.55, 95% CI 1.12-5.80). In the CT subgroup, consolidation-to-tumor ratio (CTR) independently predicted DFS (HR 3.08, 95% CI 1.46-6.51). Optimism-corrected C-index was 0.740 for the Adjusted Grade-based nomogram, compared with 0.727 for the IASLC-based nomogram and 0.612 for TNM staging. After adding CTR, they were 0.763 and 0.745 for the Adjusted Grade- and IASLC-based nomogram. Decision-curve analysis showed similar net benefit profiles for the two nomograms. CONCLUSION: High-grade components below 20% were associated with poorer DFS. The Adjusted Grade-based nomogram improved internal discrimination versus IASLC Grade and TNM staging but showed no clear clinical-benefit advantage over IASLC Grade. External validation is needed to establish optimal thresholds and clinical utility.