Indeterminate solitary pulmonary nodule: Surgical diagnosis and models for predicting malignancy

Objectives:To analyze the histology of patients with indeterminate solitary pulmonary nodule(ISPN) after surgery.To describe clinical(C) and radiological (Rx) characteristics.To estimate a model for predicting cancer in our series and apply the Mayo Clinic model(MCM). Methods:Retrospective study of patients with ISPN seen on CT referred for surgery with a diagnosis(Jan98-Aug13).C and Rx variables collected from patient files.Independent predictors for malignant NPS were estimated developing a predictive model with a multivariate logistic regression model.Probability of cancer was calculated with the MCM along with the areas under the curve(AUC) of both. Results:74cases:82.4%malignant (60.8%adenocarcinoma(Ade))17.6% benign (9.5% hamartoma (H)). ![Figure][1] Significant differences were found between the two groups for nodule edges and pack-years;both variables were identified as independent risk factors for malignant SPN.Malignant probability formula=1/1+e–f(x) f(x)=-2.318+3.694x jagged edge(JE) +0.04x pack-years(PY) . Sensitivity(S)=96.7%;specificity(Sp)=69.2%;AUC=0.904(95%CI0.823-0.984).When the MCM was applied to our series, both the S and Sp obtained were lower(p=0.31). Conclusions:82.4% of the ISPN were malignant mostly Ade;half of the benign ISPN were H. JE and accumulated PY are the only 2 characteristics with significant differences between malignant and benign lesions.This estimated model may enable us to diagnose ISPN. [1]: pending:yes