Sonographic Characteristics of Thymoma Compared With Mediastinal Lymphoma

Introduction

The thymus is an important organ that aids in the maturation of cells in the immune system. It begins as a paired organ originating from the third and fourth pharyngeal pouches.^1–3 The thymic buds migrate down the neck along the trachea and into the mediastinum toward the pericardium. In normal animals, the thymus reaches its largest size by the fourth or fifth month of life.^1,4 At that point, the thymus is a lobulated structure found in the ventral mediastinum within the cranioventral mediastinal reflection. It possesses both a cortex and medulla. The cortex is responsible for the maturation of T lymphocytes, whereas the medulla is made of epithelioid cells.^1,2,4,5 After the fifth month of life, the thymus begins to regress but never fully disappears. The functional portion of the thymus in an adult animal is mostly replaced by fat.^2–4

Thymomas develop from the thymic epithelium of the medulla but are generally highly infiltrated by mature lymphocytes.^6,7 Typically, thymomas are benign space-occupying masses in the ventral mediastinum that can occasionally invade regional structures, such as the caudal vena cava and aorta. Thymomas are the most frequently diagnosed type of thymic disesase.^3,4,6,8,9 Dogs are more likely to be diagnosed with thymomas than cats but in both species this disease more commonly affects older animals.^1,4,6,7

Thymomas are often not detected until very large in size due to nonspecific clinical signs.^10,11 Thymomas may be an incidental finding; however, presenting clinical signs include dyspnea, vomiting, cough, regurgitation, weight loss, salivation, lethargy, weakness, or regional edema.^{1,4,6,7,9,12–14} Those clinical signs often are attributed to the space-occupying lesion produced by the mass. A paraneoplastic syndrome causing megaesophagus has been associated with thymoma. Some patients show signs of regurgitation, concurrent myasthenia gravis, cytopenias, polymyositis, nonthymic cancer, systemic lupus, or other paraneoplastic syndromes.^1,4,6,7,9,13 Those syndromes are accredited to abnormal function of the T cells that normally mature in the thymus.^4,15 Those rogue T cells and their products have the ability to lead to the same signs seen in autoimmune diseases caused by other factors.

The available literature on the gross appearance of thymomas suggests that those neoplasms vary greatly in respect to size, weight, appearance, and color. However, most thymomas contain internal cysts filled with either serous or serosanguinous fluid.^{1,3,4,6–8,12} Most possess a capsule, but thymomas vary in the amount of extension through the capsule and invasion into adjacent tissues.^1,3,4,7,12

Histology of samples obtained from thymomas reflects their epithelial structure and immune function. It is suspected that the neoplastic cell is the thymic epithelial cell even though immune cells often predominate.¹ Thymomas can be categorized as lymphocyte-rich or epithelial-rich.^4,16 Thymomas can be cystic in nature, which has been identified on histopathology and correlates with gross pathology.³

Thoracic radiographs of patients with thymoma usually show a soft tissue opaque mass in the cranial mediastinum.⁴ Occasionally, opacities are identified in the craniodorsal aspect of the thorax.¹⁶ Tracheal compression or deviation in respect to the mass and pleural effusion may be present. Many cases show radiographic evidence of megaesophagus. Ventrally located peribronchial or alveolar patterns may be identified in the lung parenchyma due to aspiration pneumonia related to megaesophagus.^4,8,14,16

Sonographic evaluation of mediastinal masses is becoming more common due to its widespread availability and ability to characterize the internal appearance of the mass. Ultrasound is a safe and effective means of obtaining samples of the mass either by fine-needle aspiration or biopsy.^17,18 The sonographic appearance of thymomas has not been described thoroughly. There has been evidence that thymomas appear as a mixed echogenic mass with hypoechoic to anechoic(cystic) internal structures that can display through transmission artifact, with or without pleural effusion, but the statistical significance of those findings has not been determined to this point.^7,14,17,19

Lymphoma commonly manifests in the mediastinum and mediastinal lymph nodes. The sonographic appearance of mediastinal lymphoma has been characterized as homogenous and hypoechoic masses or lymph nodes.¹⁷ With lymphoma’s established sonographic appearance, this tumor can be used for comparison with thymoma to evaluate for any differences.

The purposes of this retrospective study were to characterize the sonographic appearance of thymomas and test the hypothesis that thymomas are more likely to be internally cystic compared with cranial mediastinal lymphoma.

Materials and Methods

A medical records search was performed at the Colorado State University Veterinary Medical Center from January 2005 to January 2012 for diagnosis of mediastinal lymphoma and thymoma. Inclusion criteria for the study were cases with definitive cytology or histology and accompanying ultrasound examinations. A total of 52 cases were found in the search; however, 2 cases were excluded from the study because they lacked reviewable ultrasonographic images. Forty-two dogs and 8 cats met the inclusion criteria. Static ultrasound images and reports were reviewed by a board-certified radiologist (A.M.). Masses were evaluated for echogenicity (hypoechoic or heterogeneous), cystic (yes or no), shape (ovoid or lobulated), margination (rounded or irregular), internal vascularity (avascular or vascular), and presence of effusion.

Results

There were 35 cases of thymomas and 15 cases of lymphoma, which met the inclusion criteria. Of the 35 thymomas; 30 were found in dogs while five were found in cats. Thirty-three (94.3%) showed heterogeneous echogenicity, 2 out of 35 (5.7%) were hypoechoic, 20 (57.1%) were distinctly cystic containing one or more cysts (Figure 1), and 15 (42.9%) showed no evidence of cysts. Thirty (85.7%) had a lobulated shape and five (14.2%) were ovoid. Thirty (85.7%) had irregular/indistinct margins and five (14.2%) had either rounded margins or margination could not be determined because it was outside the frame of the available images. Based on color Doppler interrogation, 15 thymomas were vascular and 5 were characterized as avascular. Fifteen thymomas had no Doppler images to evaluate and no mention of vascularity in the ultrasound report. Seven thymomas (20%) had evidence of plural effusion.

View Figure

• Download as Powerpoint
• Download Figure

Of the 15 mediastinal lymphomas, 12 were in dogs and 3 in cats. Eight of the lymphomas (53.3%) were of heterogeneous echogenicity (Figure 2A) and seven lymphomas (46.7%) were hypoechoic (Figure 2B). Twelve (80%) were solid and three (20%) cystic. Eleven (73.3%) were lobulated and four (26.7%) were ovoid in shape. In respect to margination of the lymphomatous lymph nodes, eight (53.3%) were irregular and four (26.7%) were smooth with three (20%) having no mention of margination. Nine lymphomas (60%) were vascular, one (6.7%) was avascular, and five (33.3%) had no mention of vascularity based on either Doppler images or the imaging report. Four (26.7%) had concurrent effusion and 11 (73.3%) had no sonographic evidence or mention of effusion in the report. Table 1 compared echogenicity and either presence or absence of cysts between thymoma and lymphoma. No significant differences were found between groups when comparing shape, margination, vascularity, or effusion.

View Figure

• Download as Powerpoint
• Download Figure

TABLE 1 Comparison of Echogenicity and Cavitation Between Thymoma and Lymphoma

View Fullscreen

Based on univariate logistic regression, echogenicity (P < .0028) and presence of cysts (P < .0219) were significant differentiators of thymoma from lymphoma, with thymoma being of heterogeneous echogenicity and containing one or more cysts. No other criteria were significant. With multivariate logistic regression analysis, only echogenicity (P = .0272) remained a significant discriminator between thymoma and lymphoma. The sensitivity and specificity of heterogeneous echogenicity as a predictor for thymoma was 94.3% and 46.7%, respectively. Echogenicity correctly identified thymoma with a PPV of 80.5, and lack of a mixed echogenicity had a NPV of 77.5.

Twenty-one thymomas also had histologic diagnosis. Of those 21 masses, 8 specifically described the cystic nature of the mass and 4 reported multifocal necrosis or “lakes of hemorrhage.” Seven masses were noted as being cystic by both ultrasound and histology. Many of the other reports made no mention of the gross appearance of the mass but only stated the cellular features of the mass. The number of thymomas that were stated to be cystic either by ultrasound or histology indicated that 25 out of 35 thymomas (71.43%) had cystic features. The presence of cysts in a mass has a PPV of 87.0% and the absence of cysts has a NPV of 44.4%.

Discussion

Based on the findings of this study, the heterogeneous echogenicity of thymomas may discriminate thymoma from mediastinal lymphoma. Presence of cysts was not statistically significant on multivariate analysis; however, 20 out of 35 thymomas had a cystic appearance on ultrasound, which correlates with gross and histopathological findings of the tumors in this study. These findings are also similar to previous studies in which thymomas were cystic on ultrasound examination.¹⁷ Additionally, the predominantly solid appearance of lymphomatous lymph nodes suggests that evaluating a mediastinal mass’ degree of cyst formation could be used to assess tumor type. Interestingly, half of the lymphomatous lymph nodes were heterogeneous and half were homogenous in echogenicity indicating that lymphoma doesn’t always have a hypoechoic echogenicity. As ultrasound is better able to detect the internal structure of a mass compared with radiography, that imaging modality could be more useful in diagnosing mediastinal masses. The use of static images in this study could have lowered the number of thymomas categorized as cystic versus solid because images were obtained at the sonographers’ discretion and not obtained in a controlled, prospective manner. Multiple thymomas classified as heterogeneous in echotexture could have contained small cysts that were not recognized due to the limited images reviewed. A large prospective study is needed to confirm the significance of cyst formation in identifying thymomas.

In humans, thymomas are often cystic in nature with variable presence of pleural effusion.²⁰ A study by Sakai et al. (1994) found that the majority of thymomas in people have either large cysts or a lacy appearance. Some of the masses evaluated in that study had more solid characteristics in regions of the mass.²⁰ In people, ultrasound is also used to obtain aspirates or biopsies that provide reliable definitive diagnoses.^21,22

The remaining parameters evaluated (including mass shape, margination, vascularity, and either presence or absence of effusion) were neither significant nor determined to differentiate between lymphoma and thymomas. Larger studies may be necessary to evaluate these parameters more fully.

Limitations of this study include its retrospective nature, small number of cases of thymoma and lymphoma, the use of still images, lack of protocol in obtaining sonographic examinations, and the variety of ultrasonographers performing the studies. A prospective study with a standardized protocol utilizing cine loops and standard sonographic views and protocols with histopathology is needed to further evaluate the sonographic appearance of thymomas.

Conclusion

The finding of a heterogeneous echogenic or cystic mediastinal mass increases the clinical suspicion of thymoma compared with lymphoma. The lack of cysts or heterogeneous echogenicity increases the suspicion of lymphoma compared with thymoma. Further studies are needed to fully evaluate the sonographic appearance of thymoma.