Clinical significance of dual-energy dual-layer CT parameters in differentiating small-sized gastrointestinal stromal tumors from leiomyomas

Jpn J Radiol. 2023 Dec;41(12):1389-1396. doi: 10.1007/s11604-023-01473-4. Epub 2023 Jul 19.

Abstract

Purpose: Small gastrointestinal stromal tumors (GISTs) can generally have nonspecific CT findings similar to those with benign submucosal tumors of the stomach. The purpose of this study was to explore the potential dual-layer dual-energy CT (dlDECT) parameters to differentiate small-sized (≤ 4 cm) GISTs from leiomyomas of the stomach.

Materials and methods: This retrospective study included 26 SMTs ≤ 4 cm in diameter with pathological confirmation of either GIST (n = 17) or leiomyoma (n = 9) from May 2018 to January 2022. All patients received contrast-enhanced CT. The normalized iodine concentration (NIC) and spectral slope (λHU) were compared between GIST and leiomyoma. Receiver-operating characteristic (ROC) curves were plotted and the areas under the curve (AUCs) were calculated to estimate the diagnostic performance of these markers for differentiating GISTs from leiomyomas.

Results: NIC was significantly higher in GIST than in leiomyoma in the portal (P = 0.0019) and delayed phases (P = 0.0011). λHU was significantly higher in GIST than in leiomyoma in the portal (P = 0.0006) and delayed phases (P = 0.0009). AUC of the ROC curves using NIC to differentiate between GIST and leiomyoma were 0.875 and 0.895 in the portal and delayed phase; using λHU, they were 0.918 and 0.902 in the portal and delayed phase.

Conclusion: dlDECT parameters including NIC and λHU show promise as indicators for differentiating small-sized GISTs from leiomyomas.

Keywords: Dual-energy CT; Gastrointestinal stromal tumors; Iodine density; Leiomyomas.

MeSH terms

  • Clinical Relevance
  • Diagnosis, Differential
  • Gastrointestinal Stromal Tumors* / diagnostic imaging
  • Humans
  • Leiomyoma* / diagnostic imaging
  • Leiomyoma* / pathology
  • Retrospective Studies
  • Tomography, X-Ray Computed