Nuclear Grading System Predicts Recurrence in Stage I Lung Adenocarcinoma Patients.
Kyuichi Kadota, Kei Suzuki, Valerie W Rusch, Andre L Moreira, Prasad S Adusumilli, William D Travis. Memorial Sloan-Kettering Cancer Center, New York, NY; Memorial Sloan-Kettering, Cancer Center, New York, NY
Background: There is no established histologic grading system for lung adenocarcinoma. Architectural and cytologic grading systems are widely accepted for prostate (Gleason) and breast (nuclear) cancers, respectively. We sought to investigate whether a nuclear grading system can predict recurrence among stage I lung adenocarcinoma.
Design: H&E slides of 506 stage I lung adenocarcinoma patients from a single institution (1995 to 2005) were evaluated for the following nuclear features: nuclear diameter, nuclear atypia, nuclear/cytoplasmic ratio, chromatin pattern, presence of intranuclear inclusion, prominence of nucleoli, mitotic count per 10 high power fields (HPF), and presence of atypical mitoses. We also correlated the nuclear grade with histologic subtyping according to the new IASLC/ATS/ERS classification of lung adenocarcinoma based on predominant morphology: lepidic, acinar, papillary, micropapillary and solid. Log-rank tests and Cox proportional hazard models were used to analyze the association between histological variables and disease-free survival (DFS).
Results: Larger nuclear diameter (P=0.024), nuclear atypia (P=0.020), higher mitotic count (P<0.001), and atypical mitosis (P<0.001) were associated with shorter DFS. Based on these results, we developed a scoring system based on the sum of two factors: nuclear diameter (1: small, 2: intermediate, 3: large) and mitotic count (1: 0-1/10HPF, 2: 2-4/10HPF, 3: ≥5/10HPF). We further established a 3-tier grading system based on this scoring.
|Nuclear grade||N||Recurrence (%)|
|Grade I (score 2-3)||252||88.6|
|Grade II (score 4-5)||177||77.7|
|Grade III (score 6)||77||71.6|