AB021. Histopathologic characteristics of resected thymic cyst walls, with imaging correlation

BACKGROUND: Thymic cysts can change in volume, CT attenuation, and MRI signal over time, raising possibility that spontaneous hemorrhage and resorption could contribute to these changes. A previous study showed some of their CT features to prompt misinterpretation as thymomas. We studied the pathological features of thymic cysts and correlated them with preoperative imaging findings. METHODS: The MGH Pathology archives were searched to identify thymic cyst resections between April 2000 and May 2020. Exclusion criteria included non-prevascular mediastinal localization, location within an enhancing mass (imaging), and/or presence of solid-cystic mass identified on gross examination. Cases without available imaging were also excluded. Various macroscopic/microscopic pathological parameters were evaluated and correlated with cyst imaging characteristics seen on CT or MRI performed closest to the time of surgery. RESULTS: Upon application of exclusion criteria, we identified 18 thymic cysts from the initial 84 mediastinal cystic specimens. The median age at resection was 60.5 with a range of 45–77 years. Most cysts were unilocular (11/18; 61%), with a gross maximum diameter ranging from 1.5 to 11.2 cm (mean ± SD: 4.2±2.7 cm). Microscopic review showed that most cysts had mixed patterns (8/16; 50%) of cuboidal, flat, squamous and/or pseudostratified epithelia. Common findings included hemosiderin deposition in the cyst wall (5/18; 28%), calcifications (6/18; 33%), chronic inflammation (6/18; 33%), hyalinosis (11/18; 61%), and significant fibrosis (12/18; 67%). Other findings included cyst walls with denuded epithelium (6/18; 33%); adjacent cholesterol clefts (3/18; 17%); and granulation tissue (2/18; 11%). The adjacent thymic tissue was involuted in 16/18 (89%) cysts, and showed fat necrosis in 11/18 (61%), microcystic Hassall’s corpuscles change in 4/18 (22%), lymphoid follicular hyperplasia in 3/18 (17%), and true thymic hyperplasia in 1/18 (6%). A total of 17/18 cysts were imaged by CT, and 4/18 imaged by MRI, including 3 imaged by CT. On CT, 6/17 (%) cysts demonstrated wall calcification, 11/17 (%) had attenuation values ≥20 HU, and mean/median wall thickness on CT [solely perceivable and measurable in 5/17 (%)] was 3 mm/3 mm ±1.5 (1 SD). Of the 4 cysts imaged by MRI, three were T1-isointense and the other T1-hypointense to muscle. CONCLUSIONS: Thymic cysts encompass a series of morphologically diverse lesions, which often show histological features suggestive of microbleeding, inflammation, and fibrosis, which may explain their variable CT and MRI appearance.