Elsevier

Human Pathology

Volume 73, March 2018, Pages 7-15
Human Pathology

In this issue
Thymoma: a clinicopathological correlation of 1470 cases

https://doi.org/10.1016/j.humpath.2017.08.018Get rights and content

Highlights

  • A total of 1470 thymomas were assessed for classification and staging.

  • A simplified classification approach (Suster-Moran) correlates with clinical outcome.

  • A newly proposed staging system provides statistically significant correlation among the different stages.

Summary

We present 1470 surgical resections for thymoma identified in the pathology files of 14 institutions from 11 countries with the purpose of determining and correlating a simplified histological classification of thymoma and pathological staging with clinical outcome. The study population was composed of 720 men and 750 women between the ages of 12 and 86 years (average, 54.8 years). Clinically, 137 patients (17%) had a history of myasthenia gravis, 31 patients (3.8%) of other autoimmune disease, and 55 (6.8%) patients of another neoplastic process. Surgical resection was performed in all patients. Histologically, 1284 (87.13%) cases were thymomas (World Health Organization types A, B1, and B2, and mixed histologies), and 186 (12.7%) were atypical thymomas (World Health Organization type B3). Of the entire group, 630 (42.9%) were encapsulated thymomas, and 840 (57.9%) were invasive thymomas in different stages. Follow-up information was obtained in 1339 (91%) patients, who subsequently were analyzed by univariate and multivariate statistical analysis. Follow-up ranging from 1 to 384 months was obtained (mean, 69.2 months) showing tumor recurrence in 136 patients (10.1%), whereas 227 died: 64 (28.2%) due to tumor and 163 (71.8%) due to other causes. Statistical analysis shows that separation of these tumors into thymoma and atypical thymoma is statistically significant (P = .001), whereas tumor staging into categories of encapsulated, minimally invasive, and invasion into adjacent organs offers a meaningful clinical assessment with a P = .038. Our findings suggest that our simplified histological schema and pathological staging system are excellent predictors of clinical outcome.

Introduction

The classification and staging of thymomas have generated significant interest and controversy over the last 2 decades. Of note, some earlier publications have also shed important light on the pathology of the thymus. In 1955, in the first series of fascicles of the Armed Forces Institute of Pathology [1], under the designation of Atlas of Tumor Pathology, Section V—Fascicle 19, the author stated, “Many attempts have been made to classify thymomas on the basis of cell type.” In addition, the author stated, “… it is both difficult and hazardous to classify tumors of such cell derivatives,… there are so many variations within a given tumor … no attempt is made to give a special name to any particular variant.” Interestingly, only recently has such a statement been proven in a study of 630 thymomas evaluated for the impact on histologic heterogeneity [2]. In 1975, in the second series of fascicles from the Armed Forces Institute of Pathology [3], the authors stated, “… once the term thymoma is restricted to the tumor of epithelial thymic cells, with or without lymphocytic component, all further subdivisions are artificial.”

In 1961, Bernatz et al [4] put forward a classification system based on the proportion of lymphocytes and separated thymomas into lymphocyte rich, epithelial rich, and mixed thymomas in which the tumor shows approximately equal proportion of lymphocytes and epithelial cells. In a subsequent publication of 181 thymomas, Bernatz et al [5] clearly stated that all histological variants might become invasive neoplasms. In 1985, Marino and Müller-Hermelink [6] proposed a new histological classification so-called histogenetic classification, separating thymomas by “cell of origin” into cortical and medullary types. However, contrary to the system of Bernatz et al [4], which predicts clinical outcome based on tumor stage at the time of diagnosis, this so-called histogenetic classification bases its prediction of clinical outcome on histology—cortical or medullary type.

With the lack of an official classification for thymomas, the World Health Organization (WHO) organized a panel of experts who in 1999 produced their first official publication on the Histological Typing of Tumours of the Thymus [7]. However, this initial WHO classification was a compromised approach comparing the Bernatz et al [4] and the Marino and Müller-Hermelink [6] classification schemas. In the WHO proposal, a system of letters and numbers was put forth as a “facilitator” of those schemas. Two important statements, quoted below, emerged as a highlight of this initial WHO schema [7]:

“… the terminology chosen here is a non-committal one based on a combination of letters and numbers. It is not proposed as a new classification, but mainly to facilitate comparison among the many terms and classification schemes.”

“… the committee wishes to stress the importance of independently evaluating thymic epithelial tumours on the basis of their presence and degree of invasiveness and their cytoarchitectural features.”

Furthermore, the committee offered an additional statement: “It will be appreciated, of course, that the classification reflects the present knowledge and that modifications are almost certain to be needed as experience accumulates.” The last 2 publications from the WHO (2004 and 2015) [8], [9] have ignored the advice from the members of the 1999 committee and the abundant evidence on the subjectivity of this schema. In addition, the authors have also perpetuated a schema that was clearly not intended as a new classification.

The current study of 1470 resected thymomas represents the largest series of these tumors and clearly demonstrates that a simplified histological schema [10] and a more recently proposed pathological staging system [11] offer an excellent predictor for clinical outcome, which, in turn, will aid in determining who is best treated with surgery alone or who is best to be followed with additional treatment such as radiation and/or chemotherapy.

Section snippets

Materials

All of the cases correspond to surgical resections identified by the Thymic International Group in the files of their respective departments of pathology, following approved Institutional Review Board guidelines. The patient records were evaluated for additional significant clinical history specifically myasthenia gravis (MG), other autoimmune disorders, or associated neoplasia. In addition, all hematoxylin-eosin sections were reviewed. Clinical follow-up information was obtained in 1339

Clinical features

Tables 1 and 2 depict the most important clinical findings in this cohort. The cohort includes 1470 patients, 720 men and 750 women between the ages of 12 to 86 years (average, 54.8 years); 127 patients are younger than 35 years (8.6%). Clinical information regarding the presence of MG, other autoimmune disorders, or other neoplasia obtained in 807 patients shows that 137 patients (17%) had MG, 31 patients (3.8%) had another autoimmune disease, and 55 patients (6.8%) had another neoplastic

Discussion

The histological classification and the staging system of thymomas are subjects that have been debated for some time and continue to pose many challenges. Prior to the 1999 WHO schema of thymomas, there were numerous clinicopathological correlations that emphasized the importance of tumor invasion [12], [13], [14], [15], [16], [17], [18], including the incidence of recurrence in previously diagnosed “encapsulated tumors” [12], which was estimated to occur in less than 2% of all thymomas. In

Supplementary data

The following are the Supplementary data to this article.

Supplementary material.

References (37)

  • C.A. Moran et al.

    On the histologic heterogeneity of thymic epithelial neoplasms: impact of sampling in subtyping and classification of thymomas

    Am J Clin Pathol

    (2000)
  • J. Rosai et al.

    Tumors of the thymus

  • M. Marino et al.

    Thymoma and thymic carcinoma. Relation of thymoma epithelial cells to the cortical and medullary differentiation of thymus

    Virch Arch

    (1985)
  • J. Rosai

    Histological typing of tumours of the thymus. WHO international histological classification of tumours

    (1999)
  • S. Suster et al.

    Thymoma, atypical thymoma, and thymic carcinoma: a novel conceptual approach to the classification of thymic epithelial neoplasms

    Am J Clin Pathol

    (1999)
  • C.A. Moran et al.

    Thymoma II: a clinicopathologic correlation of 250 cases with a proposed staging system with emphasis on pathological assessment

    Am J Clin Pathol

    (2012)
  • Cited by (49)

    • Common thymomas: classification, histology, staging and prognosis

      2023, Diagnostic Histopathology
      Citation Excerpt :

      Previously, certain histological subtypes (type A and AB) were considered benign tumors without the potential to invade or metastasize.10 While it is true that certain histological types (type A, AB, B1 and B2) are associated with improved survival compared to type B3,36 all thymomas can behave in a malignant fashion and the most important prognostic parameter remains staging at the time of diagnosis.16,36 Interestingly, the presence of myasthenia gravis does not seem to have an effect on the prognosis, irrespective of thymoma type.12

    • Spindle cell thymoma (WHO type A) – Choriocarcinoma: An unusual association: A clinicopathological and immunohistochemical study of 3 cases

      2022, Pathology Research and Practice
      Citation Excerpt :

      Spindle cell thymomas (WHO type A) and choriocarcinomas are two distinct tumoral conditions, which in the scheme of thymomas and germ cell tumors, are the least common [1,2].

    • On the Histologic Classification of Thymoma

      2024, Advances in Anatomic Pathology
    View all citing articles on Scopus

    Disclosures: None of the authors has any conflict of interest regarding the elaboration of this manuscript. In addition, no specific funding was provided in the preparation of this manuscript.

    View full text