Computer Assisted Nodule Analysis and Risk Yield is Associated With Occult Lymph Node Status in Clinical Stage I-IIA Lung Adenocarcinoma Undergoing Resection

Abstract

Background

Current staging work-up does not capture all occult lymph node (OLN) disease. We sought to determine if Computer Assisted Nodule Analysis and Risk Yield (CANARY) analysis could help distinguish OLN status in early-stage lung adenocarcinoma.

Methods

Retrospective review of resected lung cancer patients from 2016 to 2021 was performed. Patients with surgically resected clinical stage I-IIA lung adenocarcinoma were included. Preoperative imaging was entered into the CANARY software, and each lesion was categorized into good, intermediate, and poor risk. OLN status was determined per pathology results. Pearson's Chi-square correlation, univariate and multivariate logistic regression models were used to assess OLN metastases as a function of CANARY risk profile, with statistical significance at $\alpha =0.05$.

Results

In total, the study cohort included 228 patients with median age of 70. By clinical stage, 195 (85.5%), 24 (10.5%), and 9 (3.9%) patients were determined to be in IA, IB, and IIA, respectively. 28 (12.3%) patients were found to have OLN metastases. Among them, 1 (3.6%), 3 (10.7%), and 24 (85.7%) patients had a good, intermediate, and poor CANARY risk profile, respectively. CANARY risk profile was significantly associated with OLN metastases ($\chi$2 = 9.9, P = .007). Relative to the good/intermediate group, patients with poor risk had a more-than 3-fold increase in likelihood of having OLN metastases (odd ratio [OR] = 3.3, 95% confidence interval [CI]:1.6-9.2, P = .007).

Conclusion

CANARY analysis was able to risk-stratify the likelihood of OLN metastases in early-stage lung adenocarcinoma. CANARY can provide an adjunctive non-invasive tool to aid in determining an appropriate individualized treatment plan.