Imaging characteristics of pathologically proven thymic hyperplasia: identifying features that can differentiate true from lymphoid hyperplasia

Abstract

Objective: The purpose of this article is to investigate the imaging characteristics of pathologically proven thymic hyperplasia and to identify features that can differentiate true hyperplasia from lymphoid hyperplasia.

Materials and methods: Thirty-one patients (nine men and 22 women; age range, 20-68 years) with pathologically confirmed thymic hyperplasia (18 true and 13 lymphoid) who underwent preoperative CT (n=27), PET/CT (n=5), or MRI (n=6) were studied. The length and thickness of each thymic lobe and the transverse and anterior-posterior diameters and attenuation of the thymus were measured on CT. Thymic morphologic features and heterogeneity on CT and chemical shift on MRI were evaluated. Maximum standardized uptake values were measured on PET. Imaging features between true and lymphoid hyperplasia were compared.

Results: No significant differences were observed between true and lymphoid hyperplasia in terms of thymic length, thickness, diameters, morphologic features, and other qualitative features (p>0.16). The length, thickness, and diameters of thymic hyperplasia were significantly larger than the mean values of normal glands in the corresponding age group (p<0.001). CT attenuation of lymphoid hyperplasia was significantly higher than that of true hyperplasia among 15 patients with contrast-enhanced CT (median, 47.9 vs 31.4 HU; Wilcoxon p=0.03). The receiver operating characteristic analysis yielded greater than 41.2 HU as the optimal threshold for differentiating lymphoid hyperplasia from true hyperplasia, with 83% sensitivity and 89% specificity. A decrease of signal intensity on opposed-phase images was present in all four cases with in- and opposed-phase imaging. The mean maximum standardized uptake value was 2.66.

Conclusion: CT attenuation of the thymus was significantly higher in lymphoid hyperplasia than in true hyperplasia, with an optimal threshold of greater than 41.2 HU in this cohort of patients with pathologically confirmed thymic hyperplasia.

Fig. 1

Technique for thymic size measurements including length, thickness, and transverse and anteroposterior diameters.

Fig. 2

Thymic attenuation on CT in true hyperplasia and lymphoid hyperplasia. A, Box-and-whisker plots represent CT attenuation of thymus in true and lymphoid hyperplasia. Box outlines represent values from lower to upper quartile (25th to 75th percentile), and horizontal line within each box represents median CT attenuation of each group. Vertical lines extend from minimum to maximum value. CT attenuation of lymphoid hyperplasia was significantly higher than that of true hyperplasia (p = 0.03). B, Receiver operating characteristic curve analysis identified > 41.2 HU as optimal threshold for differentiating lymphoid hyperplasia from true hyperplasia, with 83% sensitivity and 89% specificity (Youden index score, 0.72).

Fig. 3

31-year-old woman with lymphoid thymic hyperplasia and Graves disease. Contrast-enhanced CT image shows pyramidal-shaped thymus gland (arrow) with CT attenuation of 67.7 HU.

Fig. 4

51-year-old woman with true thymic hyperplasia. Contrast-enhanced CT image shows thymus gland (arrow) with attenuation of −5.0 HU. Patient had myasthenia gravis and history of steroid use for asthma.

Fig. 5

31-year-old man with true hyperplasia. A and B, Transaxial chemical-shift in-phase (TR/TE, 120/2.5) MRI (A) and opposed-phase (TR/TE, 3.1/1.1) gradient-echo T1-weighted MRI (B) show decreased signal intensity (arrows) on opposed-phase (B) compared with in-phase (A) imaging.

Fig. 6

52-year-old man with true hyperplasia. A–C, Axial CT (A), ¹⁸FDG PET (B), and fused FDG PET/CT (C) images show thymus gland (arrows) with maximum standardized uptake value of 2.62.