Comparative assessment of the role of the dynamic contrast-enhanced CT versus PET-CT in characterization of suspicious lung lesions: a pathologic correlated prospective study with ROC analysis

Abstract

Abstract Background The positron emission tomography–computed tomography (PET-CT) is the cornerstone in oncologic imaging because of its high sensitivity. However, its specificity is largely compromised by the false-positive uptake in benign infectious and inflammatory conditions. The dynamic contrast-enhanced computed tomography (DCE-CT) is a promising imaging technique that can eventually provide insight into the tumor vascularity and angiogenesis through quantification of tumor perfusion parameters like peak enhancement, net enhancement (wash-in), absolute loss of enhancement (wash-out), as well as the time–intensity curve morphology. Aim of the work : To compare the efficacy and accuracy of DCE-CT versus PET-CT in the characterization of suspicious lung lesions using clinico-laboratory correlation, pathological verification, and follow-up, as well as receiver-operating characteristics (ROC) statistical analysis. Results Forty patients with indeterminate pulmonary lesions were prospectively enrolled in this study during the period between September/2023 and January/2026. All patients underwent both DCE-CT and PET-CT examinations. Automatic calculations of relevant cutoff values were encountered; then, the ROC analyses and the tests of accuracy were applied. The study was conducted by four expert radiologists, single oncologist, and a single pulmonologist. Peak enhancement : At the initially used “40 HU” value, positive statistical significance was encountered ( p value = 0.003). The area under the ROC curve (AUC) was 0.74, 95% CI [0.57, 0.92], SE = 0.09, p = 0.006. At the estimated “50 HU” threshold, the performance was statistically better. Net enhancement/wash-in : At the initially used “25 HU” value, positive statistical significance was encountered ( p value = < 0.0001). The area under the ROC curve (AUC) was 0.83, 95% CI [0.68, 0.99], SE = 0.08, p < 0.001. At the estimated “23 HU” threshold, the performance was statistically better. Final curve : Type I curve was established in 95% of malignant lesions. On the other hand, type II and III curves were established in 78% and 11% of non-malignant lesions. Positive statistical significance was encountered ( p value = < 0.0001). The overall accuracy of the DCE-CT curve was higher than that of PET-CT SUV (92.5% compared to 80%), with much higher specificity; however, the sensitivity of the DCE-CT is lower. Conclusions The DCE-CT constitutes a reliable, cost-effective, and widely available alternative for the characterization of the suspicious pulmonary lesions. Quantitative perfusion parameters, particularly wash-in, wash-out, and curve-type analysis, can serve as valuable indicators of malignancy and improve diagnostic confidence in indeterminate cases. Integrating both modalities may yield a more holistic diagnostic model to enhance precision in lung cancer detection and management.