Clinical application of 18F-FDG PET/CT parameters in predicting tumor spread through air spaces in patients with lung adenocarcinomas at T1-2 stage / 中华核医学与分子影像杂志

ABSTRACT

Objective:To evaluate the clinical value of 18F-FDG PET/CT findings in patients with T1-2 lung adenocarcinoma spread through air spaces (STAS). Methods:From June 2018 to June 2020, a total of 80 patients (36 males, 44 females; age 19-84 (59.9±11.8) years) with surgically and pathologically confirmed T1-2 lung adenocarcinomas in Jiangmen Central Hospital were enrolled retrospectively. All patients underwent 18F-FDG PET/CT examination preoperatively and were divided into STAS positive and negative groups according to the histopathological diagnosis. Independent-sample t test, Mann-Whitney U test, χ2 test and Fisher exact test were used to analyze differences of gender, age, tumor biomarker, SUV max, SUV mean, features showed on high resolution CT (HRCT; including diameter, lesion location, morphology, density, lobulated sharp, spiculated sign, vacuole sign, air bronchgram sign, pleural traction and para-emphysema), and pathologic findings (micropapillary pattern, lymphvascular inversion, pleural inversion and lymph node metastasis) between the two groups, and then multivariate logistic regression was performed. The ROC curve was employed to evaluate the predictive value of parameters for STAS of T1-2 lung adenocarcinomas. Results:Among the 80 patients with T1-2 lung adenocarcinomas, 12 (15.0%) were STAS positive and 68 (85.0%) were STAS negative. Significant differences were shown in SUV max, SUV mean, micropapillary pattern, lymphvascular inversion and lymph node metastasis between the two groups ( z values -2.60, -2.17; χ2 values 29.56, 9.28, 17.40, P<0.001 or P<0.05). SUV max (odds ratio ( OR) 1.348 (95% CI 1.071-1.695), P=0.011), micropapillary pattern ( OR=47.444 (95% CI 4.592-490.214), P=0.001) and lymph node metastasis ( OR=8.201 (95% CI 1.129-59.576), P=0.038) were independent risk factors for STAS positive in multivariation logistic regression analysis. The optimum cut-off value for SUV max was 3.85 in the ROC analysis with the AUC of 0.737 (95% CI 0.614-0.859), the sensitivity of 11/12, the specificity of 55.9%(38/68) and the accuracy of 61.2%(49/80). The AUC of the SUV max combined with micropapillary pattern and lymph node metastasis was 0.945 (95% CI 0.892-0.999) with the sensitivity of 11/12, the specificity of 88.2%(60/68) and the accuracy of 88.7%(71/80). Conclusions:The PET/CT characteristics may be useful in differentiating STAS status among patients with T1-2 lung adenocarcinoma. SUV max >3.85, pathological papillary pattern and lymph node metastasis are independent risk factors to predict STAS.