Pericardial Effusions Cytology: Experience at a Tertiary Care Center

Abstract

Introduction: Cytologic assessment is useful in determining the etiology of pericardial effusions. Recent pericardial effusion studies consist mainly of autopsy-based series and case reports of rare entities. In this study we reviewed pericardial fluid diagnoses at a tertiary care center, analyzed the frequencies of malignancies in pericardial fluid and assessed the rate of surgical correlation. Material(s) and Method(s): The electronic archive at Loma Linda University Medical Center was searched for pericardial effusion specimens and surgical follow-up over a 5 year period (2007-2011). Glass slide morphology, immunohistochemistry and diagnoses were reviewed. Result(s): A total of 59 pericardial cytology specimens from 48 different patients were retrieved, along with 26 correlating surgical specimens. The 16 malignant effusions consisted of clear cell renal cell carcinoma, chronic myelogenous leukemia, 5 breast carcinomas not otherwise specified, 6 poorly differentiated adenocarcinomas (4 of pulmonary origin, one signet ring gastric carcinoma, one signet ring cell carcinoma of unknown origin), a non-small cell poorly differentiated carcinoma of pulmonary origin, a non-small cell carcinoma with squamous features of pulmonary origin, and a poorly differentiated transitional cell carcinoma. Diagnosis of malignant pericardial effusions was mainly based on cytomorphology, supplemented by history if present of prior or known malignancy and immunohistochemistry. Pericardial cytology provided the initial cancer diagnosis in five specimens; patients with a previously identified cancer constituted the remaining 11/16 malignant specimens. Immunocytochemistry was utilized in 3/41 benign, 2/2 atypical, and 10/16 malignant cases. The primary neoplasm could not be definitively pinpointed in one poorly differentiated carcinomas despite additional tissue sampling and radiographic imaging. Atypical cells were reported in a patient with chemotherapy for squamous cell carcinoma with prior pericardial metastases and in a patient with metastatic mesothelioma on concurrent pericardial biopsy. The remainder of the pericardial effusions consisted of 21 reactive cases, 7 cases with chronic inflammation, and 13 cases with acute inflammation. Correlating histology was available for 26 specimens. The malignant cytology specimens had 7 correlating pericardial biopsies. Pericardial biopsies were received for both atypical cases, with histologic diagnoses of pericarditis with severely reactive changes and metastatic mesothelioma, respectively. No pericardial fluid involvement was seen in 3 cases of surgically confirmed cardiac tumor involvement (one case of metastatic melanoma and 2 cases of benign primary cardiac tumors - one rhabdomyoma and one atrial myxoma) and in five patients with localized malignancies (colon, prostate, breast, hepatocellular and follicular thyroid carcinoma). Conclusion(s): Pericardial effusions may comprise a small percentage of cytology specimens but have a high rate of malignancy, with adenocarcinomas of breast and lung being the most frequent in our patient population. In metastatic malignancies, cytology demonstrates a high correlation with histologic specimens. In conjunction with selected immunohistochemical stains and adequate patient history and imaging, cytomorphology is a valuable tool for pericardial effusion evaluation.