Can quantitative parameters of spectral computed tomography predict lymphatic metastasis in lung cancer? a systematic...

This study evaluated the use of quantitative spectral computed tomography (CT) parameters to identify lymph node metastasis (LM) in lung cancer.

Literature about LM in lung cancer diagnosed using spectral CT up to September 2022 was retrieved from the PubMed, EMBASE, Cochrane Library, Web of Science, Chinese National Knowledge Infrastructure, and Wanfang databases. The literature was strictly screened according to the inclusion and exclusion criteria. Data were extracted, quality assessment was performed, and heterogeneity was evaluated. The pooled sensitivity (SEN), specificity (SPE), positive likelihood ratio (+LR), -LR, and diagnostic odds ratio (DOR) for normalized iodine concentration (NIC) and spectral attenuation curve (λHU) were calculated. The subject receiver operating characteristic (SROC) curves were used, and the area under the curve (AUC) was calculated.

Eleven studies, including 1,290 cases, without obvious publication bias were enrolled. In eight articles, the pooled AUC of NIC in the arterial phase (AP) was 0.84 (SEN = 0.85, SPE = 0.74, +LR = 3.3, -LR = 0.20, DOR = 16) while that of NIC in the venous phase (VP) was 0.82 (SEN = 0.78, SPE = 0.72). Additionally, the pooled AUC for λHU (AP) was 0.87 (SEN = 0.74, SPE = 0.84, +LR = 4.5, -LR = 0.31, DOR = 15) and that for λHU (VP) was 0.81 (SEN = 0.62, SPE = 0.81). Lymph node (LN) short-axis diameter was ranked last, with a pooled AUC of 0.81 (SEN = 0.69, SPE = 0.79).

Spectral CT is a suitable noninvasive and cost-effective method for determining LM in lung cancer. Additionally, NIC and λHU in the AP have good discrimination ability than short-axis diameter, providing a valuable basis and reference for preoperative evaluation.