Epstein-Barr Virus Positive Pleural Effusion. Clinical Features, Cytomorphologic Characteristics, and Flow Cytometric Immunophenotyping

Transcription

1 AJCP / Original Article Epstein-Barr Virus Positive Pleural Effusion Clinical Features, Cytomorphologic Characteristics, and Flow Cytometric Immunophenotyping Hidehiro Takei, MD, and Dina Mody, MD From the Department of Pathology and Genomic Medicine, Houston Methodist Hospital, Houston, TX. Key Words: Epstein-Barr virus; Pleural effusion; Real-time quantitative PCR Am J Clin Pathol December 2014;142: ABSTRACT Objectives: The etiology of pleural effusions (PEs) varies, and a percentage of PEs remains unexplained despite an intensive workup. One previous study documented a high prevalence of Epstein-Barr virus (EBV) D in unselected and unexplained PEs. Our aim is to characterize the clinical and cytomorphologic features of EBV-associated PEs, which have not been described. Methods: A database search was performed for PEs with EBV-D identified in the fluid by real-time quantitative polymerase chain reaction. The corresponding fluid cytology and chemistry were reviewed, and the patients demographic data and clinical features were recorded. Results: A total of 20 cases of EBV-D positive PE were found. All patients had a history of lung transplantation. Most of the PE EBV loads were relatively low. Cytologically, polymorphous lymphocytosis was present in more than 70% of PEs. Scattered lymphocytic mitosis and apoptosis were seen in some cases. Mesothelial cells varied in number, and some cases showed reactive atypia. The lymphocytes were predominantly T cells with the CD4/CD8 ratio varying from 10:1 to 3:20. Conclusions: EBV infection/reactivation can account for certain proportions of idiopathic PEs. Polymorphous lymphocytosis is the most common cytologic feature, but atypical features (in both lymphocytes and mesothelial cells) can be seen. Pleural effusions (PEs) are frequently encountered specimens in cytopathology. Their etiology varies, and the most common cause in the United States includes congestive heart failure, pneumonia, and cancer. 1 The cause can be determined in most cases by clinical information, imaging techniques, and pleural fluid analysis, including cytomorphologic examination. Biochemical analysis of PE, as the first step for the evaluation, allows us to determine whether it is exudative or transudative, usually reflecting the physiologic mechanisms of its formation. Once identified as exudative, additional diagnostic tests are required for the determination of the cause. One of the most important tests of these is cytomorphologic examination, which is characterized by its very high positive predictive value and modest sensitivity for malignancies. 2-4 PE with WBC predominance may need additional tests to determine the cause (eg, microbiological culture). If lymphoma is suspected clinically and/or cytomorphologically, flow cytometry (FCM) is a method particularly useful for demonstrating the presence of a clonal B-cell population in the fluid. A certain percentage of PEs remains unexplained despite an intensive diagnostic workup, even after invasive procedures such as thoracoscopy or open pleural biopsy. 1 Intriguingly, Thijsen et al 5 recently reported a high prevalence of Epstein-Barr virus (EBV) D by polymerase chain reaction (PCR) in unselected PEs and an even higher prevalence in unexplained PEs in their prospective study in the Netherlands. EBV is a ubiquitous double-stranded D virus that belongs to the Herpesviridae family. EBV infection is highly prevalent worldwide, and as many as 95% of adults between 35 and 40 years of age have been infected in the United States. 6 Most individuals become infected during childhood, 788 Am J Clin Pathol 2014;142:

2 AJCP / Original Article adolescence, or young adulthood. EBV can infect and spread among humans by the oral route, usually followed by establishment of a lifelong latent infection in the human body as with other herpes viruses. Although epidemiologically infrequent, the most significant characteristic that is known to be associated with EBV latent infection is its oncogenic effects, leading to lymphomas (including posttransplant lymphoproliferative disorder [PTLD]), carcinomas, or sarcomas. Moreover, EBV infection has been shown in the PE to be associated with specific B-cell lymphomas such as pyothorax-associated lymphoma, primary effusion lymphoma, and primary PE PTLD. 7,8 EBV systemic reactivation can occur when the cellular immune response is compromised, for example, by human immunodeficiency virus infection or in patients who received immunosuppressive medications after organ transplantation. On the other hand, local reactivation of EBV takes place periodically in the oropharynx in EBVinfected healthy individuals, 5 and their saliva serves as the primary source of viral transmission. The aim of this study is to characterize the clinical and cytomorphologic features of EBV-associated PEs, which have not been previously described. In addition, results of FCM immunophenotyping performed for lymphocytic PEs are also shown. Materials and Methods A retrospective database search at the Houston Methodist Hospital (Houston, TX) was performed for PE cases, in which EBV-D was identified in the fluid by quantitative real-time PCR (qrt-pcr) in a 6-year period from January 1, 2007, to December 31, Briefly, total D from the PE samples was automatedly isolated using a MagNa Pure Compact instrument along with a MagNa Pure Compact Nucleic Acid isolation kit (Roche Applied Science, Indianapolis, IN). A qrt-pcr was performed for each sample using a LightCycler (Roche Applied Science). Specific primers and hybridization probes targeting the EBV latent membrane protein 2 (LMP-2) gene as well as a spiked control sequence were used. Quantitation (viral load) was performed using standard curves derived from the stock standards (EBV Template D; Roche Applied Science). To ensure target specificity, melt-temperature analysis was performed after completion of amplification on the LightCycler. In this assay, the limit of detection of EBV was 75 copies/ml, and the limit of reliable quantitation was 100 copies/ml. The corresponding fluid cytology (cytospin preparation, stained with Papanicolaou and Diff-Quik) and cell count and chemistry data of fluid were reviewed. The following cytologic parameters were evaluated: presence or absence of (1) lymphocytosis/lymphocyte predominance (lymphocytes >80% of total inflammatory cells) vs mixed inflammatory cells (eg, neutrophils, histiocytes), (2) large activated lymphocytes, (3) polymorphous lymphoid population, (4) mitosis or apoptosis in lymphocytes, and (5) mesothelial cells (semiquantitatively assessed as absent, 1+, 2+, and 3+). For those cases that were immunophenotyped by FCM, the results were reviewed. When bronchoalveolar lavage (BAL) and/or transbronchial biopsy (TBBx) specimens were taken on the same day as the thoracentesis for PE, those specimens were also reviewed. In addition, the patients demographic data and clinical features were recorded. This study was approved by the Institutional Review Board at Houston Methodist Hospital (ID: Pro ). Results A total of 20 cases (18 patients) of EBV-D positive PE were found; patient demographics, clinical information, and laboratory data/pathologic findings are listed in Table 1. The patients were 13 men and 5 women with a mean age of 64.6 years. All patients had a history of lung transplant for various underlying lung diseases. The duration between transplantation and thoracentesis for pathologic examination varied from approximately 1 month to more than 10 years (median, approximately 3 months). The most common presenting symptom was shortness of breath. Chemical analysis demonstrated that all PEs were exudates. In 20 cases of PEs, 14 were considered idiopathic, while six were most likely secondary to infection given the positive culture results (of the fluid or lung tissue) or a history of aspiration pneumonia. The EBV load varied from below the limit of quantitation (75-99 copies/ml) to 248,300 copies/ml (median, 530 copies/ ml). Two patients with an EBV load of 335 and 516 copies/ ml (cases 10 and 20, respectively) died within 1 week after thoracentesis. Serum EBV load was tested on the same day of thoracentesis in three cases, one (case 16) of which showed no detection. One patient (case 19) had a pleural biopsy, which was performed 2 days after thoracentesis. EBV-encoded small nuclear R in situ hybridization was performed (Ventana Medical Systems, Tucson, AZ) and was negative. The FCM study was performed in 13 cases. Ten cases were cytologically lymphocyte-rich PEs, and all fluids were shown to be T-cell predominant. No monoclonal B-cell population was noted in any PEs tested. The CD4 to CD8 ratio of T cells in PEs varied tremendously, from 10:1 to 3:20. BAL was performed in nine cases, all of which showed no cytological abnormality present. TBBx was performed in 10 cases, and two cases showed minimal acute cellular rejection (A1) and one was pneumonia with culture positive for Klebsiella pneumoniae. Am J Clin Pathol 2014;142:

3 Takei and Mody / EBV-Positive Pleural Effusion Table 1 Twenty Cases of EBV-D Positive Pleural Effusions: Patient Demographics, Clinical Information, and Laboratory Data/ Pathologic Findings Case No. Sex Age, y Flow Cytometry PE EBV Titer, Copies/mL History B-Cell Clonality, etc Cell Differentiation 1 F COPD; s/p single LT; diabetes mellitus No monoclonal population T cells (76%), B cells (2%), NK cells (1%), monocytes (5%) 2 M IPF; s/p double LT No monoclonal population T cells (75%), B cells (15%), NK cells (3.5%), plasma cells (<0.1%) 5:1 4 M COPD; s/p double LT No monoclonal population 5 M ,300 IPF; s/p double LT No monoclonal population 6 M 69 Pos (75-99) s/p double LT No monoclonal population 7 M 63 32,180 s/p double LT No monoclonal population 8 F 63 3,132 COPD due to AAT s/p double LT No monoclonal population 9 M COPD, s/p single LT No monoclonal population 10 F s/p BMT; single LT; kidney transplant; non-hodgkin ML 11 M COPD; s/p single LT No monoclonal population 12 M 62 3,765 COPD due to AAT; s/p double LT 13 F Scleroderma; s/p single LT T cells (61%), B cells (1%), plasma cells (<0.1%) T cells (25%), B cells (1.5%), NK cells (<1%), monocytes (42%), granulocytes (22%) T cells (6%), B cells (<1%), granulocytes (84%) T cells (87%), B cells (<1%), NK cells (<1%) T cells (70%), B cells (11%), NK cells (3%), plasma cells (<1%) T cells (68%), B cells (8%), NK cells (4%) T cells (86%), B cells (1%), NK (5%), plasma cells (<1%) 14 M 65 Pos (75-99) COPD, s/p double LT 15 M IPF; s/p double LT No monoclonal population or aberrant T-cell antigen 16 M IPF; s/p double LT No monoclonal population 17 F Pulmonary hypertension due to congenital heart disease; s/p double LT 18 M 60 41,520 IPF; s/p double LT No monoclonal population T cells (79%), B cells (15%), NK cells (<1%), plasma cells (<1%) T cells (66%), B cells (2%), NK cells (16%), monocytes (2%) T cells (76%), B cells (<1%), NK cells (1%), monocytes (4%), plasma cells (2%) 19 Pos (75-99) 20 M s/p single LT; skin squamous cell carcinoma s/p radical excision and irradiation No monoclonal population T cells (2%), monocytes (11%), granulocytes (74%) 10:1 CD4:CD8 Ratio 4:1 1:2 3:20 1:2 3:1 10:1 2:1 3:1 6:1 2:1 2:1 AAT, a1 antitrypsin deficiency; BAL, bronchial alveolar lavage; BMT, bone marrow transplant; CMV, cytomegalovirus; COPD, chronic obstructive pulmonary disease; EBV, Epstein-Barr virus; IPF, idiopathic pulmonary fibrosis; LT, lung transplant; ML, malignant lymphoma;, not applicable; NK, natural killer; PCR, polymerase chain reaction; PE, pleural effusion; Pos, positive; SOB, shortness of breath; s/p, status post; TBBx, transbronchial lung biopsy; VRE, vancomycin-resistant Enterococcus. 790 Am J Clin Pathol 2014;142:

4 AJCP / Original Article BAL TBBx Other Laboratory Results Clinical Information Negative Focal interstitial lymphocytic infiltrate Presented with chest pain and fever (1 month after LT) Presented with respiratory failure Serum EBV Titer, Copies/mL 110 Negative PE CMV PCR: Negative Post LT surgery continued to be complicated with acute renal failure, deep vein thrombosis, pseudomonas pneumonia; died 3 months after the test Negative Negative Presented with SOB Granulation tissue Presented with SOB Negative Negative Recurred PE Negative Peribronchial interstitial pneumonitis Pseudomonas pneumonia; died 2 days after test Admitted with SOB Negative Acute cellular rejection (A1) Presented with SOB and productive cough Negative Negative Presented with worsening of SOB with productive cough Negative Acute cellular rejection (A1) PE cultures: negative Presented with SOB and aspiration pneumonia Negative Klebsiella culture positive Presented with worsening SOB and chest pain Negative Negative PE culture: VRE (Enterococcus faecium) positive PE culture: Staphylococcus aureus coagulase recovered (in broth only) Presented with SOB Not detected Presented with SOB 630 PE culture: negative Presented with SOB and empyema 100 Presented with SOB and worsening dyspnea; died 5 days after the test Am J Clin Pathol 2014;142:

5 Takei and Mody / EBV-Positive Pleural Effusion Table 2 Twenty Cases of EBV-D Positive Pleural Effusions: Cytomorphologic Features Case No. Lymphocytosis Large Activated Lymphocytes Of 20 cases, 19 cases (17 patients) had corresponding fluid cytology specimens for review. Cytologic findings are listed in Table 2. One of these 19 cases was unsatisfactory for cytologic evaluation due to insufficient cellularity (case 6); however, granulocyte predominance was found by FCM. Another case had no cytology slides available for review (case 13). Cytologically, 13 PE specimens demonstrated lymphocytosis, characterized by a polymorphous lymphoid population Image 1, with varying numbers of large activated lymphocytes in three cases. One of these three cases was positive for vancomycin-resistant Enterococcus faecium in PE culture (case 17). Of those with lymphocytosis, rare lymphocytic mitosis was seen in three cases and scattered lymphocytic apoptosis was seen in three cases. Of these cases, one (case 11) showed abundant activated large lymphocytes with both mitotic and apoptotic figures, and malignant lymphoma was morphologically suspected as a differential diagnosis Image 2. The other three PEs exhibited a mixed inflammatory pattern with lymphocytes, neutrophils, and histiocytes, and two PEs showed abundant neutrophils (case 20 was consistent with a clinical diagnosis of empyema). There was no significant difference in EBV load between the fluids with lymphocytosis and those with mixed or neutrophilic inflammation (P =.3004, Mann-Whitney). All cases (except for one unsatisfactory case) contained various numbers of mesothelial cells, and three cases showed reactive atypia. In case 20 (with a history of aggressive skin cancer), mesothelial (reactive) atypia was so significant that a cytologic diagnosis of suspicious for metastatic malignancy was initially rendered for the pleural fluid Image 3. Mitosis of Lymphocytes Discussion Apoptosis of Lymphocytes Mesothelial Cells 1 Present Present Absent Present 1+ 2 Present No Absent Absent 2+ 3 Present No Absent Absent 1+ 4 Present No Absent Absent 2+ 5 Absent (mixed inflammation) No Absent Absent 1+ 6 Unsatisfactory 7 Present No Absent Absent 1+ 8 Present No Absent Absent 1+ 9 Present Present Absent Absent Absent (mixed inflammation) No Absent Absent Present (prominent) Present Present (rare) Present Present No Present (rare) Absent 3+ (reactive atypia) 13 Not available for review 14 Absent (mixed inflammation) No Present (rare) Absent Present No Absent Absent Present No Absent Present Present Present Absent Absent 1+ (reactive atypia) 18 Present No Absent Absent Absent (abundant neutrophils) No Absent Absent Absent (abundant neutrophils) No Absent Absent 3+ (marked reactive atypia), not applicable. Idiopathic PE is usually defined as any PE that fails to achieve a definitive diagnosis after usual clinical evaluation, including imaging techniques and cytomorphologic analysis and biochemical study of fluid. The prevalence of idiopathic etiologies in all PE cases varies depending on how strictly the definition is applied (ie, how intensively examined). Idiopathic PE cases excluding malignancy are reported to generally have favorable outcomes. 9 Although one of the most common causes of exudative PE is infectious pleurisy, the causative infectious agents may not be detected by conventional microbiological tests. Interestingly enough, it was recently reported for the first time by Thijsen et al 5 that a 40% and 59% prevalence of EBV-D was found in unselected and idiopathic PEs, respectively, using a very similar molecular technique as we did here. Given a high percentage of EBV positivity in pleural fluid among patients with unexplained PE, the EBV-D test should be considered in all apparently idiopathic PEs in which the etiology cannot be identified after a usual diagnostic workup. Identification of causative/possible etiology for PE is clinically important since repeated PE analysis and more invasive procedures to identify the etiology can be avoided, although no known effective treatment exists for EBV-associated PE. All cases with positive EBV-D in PE were following lung transplantation in our study. This is most likely due to selection bias since the EBV-D test is much more frequently performed in lung transplant patients compared with other patient populations in our institution. In most of our cases, the EBV-D levels of PE were relatively low, 792 Am J Clin Pathol 2014;142:

6 AJCP / Original Article Image 3 Neutrophil-predominant pleural fluid with scattered atypical mesothelial cells in a patient with a history of aggressive skin cancer (case 20) (Papanicolaou, 400). although three cases showed EBV-D levels more than 10,000 (copies/ml). In terms of viral copy number, our results are very similar to those documented by Thijsen et al5 (median, 530 in our series vs 454 copies/ml). They also documented that all patients with EBV-D levels more than 10,000 (copies/ml) died within 6 months, but our series did not show such an association. The previous study reported that two-thirds of patients had a positive EBV-PCR in PE with a negative serum EBVPCR (ie, both positive in one-third of patients),5 the pattern of which is suggestive of local reactivation of the virus in Image 2 Lymphocytosis with abundant large activated lymphocytes in Epstein-Barr virus positive pleural fluid (case 11) (Diff-Quik, 400). the pleural space. There have been no other reported cases of EBV reactivation in PEs or even EBV-related PEs with the exception of PEs associated with infectious mononucleosis (ie, EBV primary infection).10,11 In our 20 cases of PE, three were tested simultaneously for EBV-D in serum by PCR, and of these, one such example was found in a patient with culture-positive K pneumoniae (case 16). The precise triggers of EBV reactivation in vivo remain unknown. It is presumed that it occurs when latently infected B cells respond to unrelated infection, given that B-cell receptor stimulation triggers reactivation in B-cell lines.12 Five cases in our series demonstrated culture-positive bacterial infection in either PE or lung tissue, which may have been associated with EBV reactivation in these patients. Of these, case 16 may represent local EBV reactivation triggered by K pneumoniae. Cytologically, polymorphous lymphocytosis is the most common feature in our series. Lymphocytic mitosis and apoptosis can be found. Atypical features can be seen, and one of our cases demonstrated abundant activated large lymphocytes with both mitotic and apoptotic figures, prompting FCM immunophenotyping to exclude malignant lymphoma. Of note is that a mixed inflammatory pattern and even a neutrophilic pattern can be seen in EBV-associated PEs, although not so common. In other words, pleural empyema cannot exclude the presence of EBV-D in the PE. Mesothelial cells are shown to be present in all cases, although the number varies, and some cases have reactive atypia. The mesothelial reactive atypia can be associated with infection, chemotherapy, or surgical procedure (eg, thoracentesis, chest tube draining) in our cases. It is unknown whether pleural mesothelial cells are susceptible to EBV infection.5 Am J Clin Pathol 2014;142: Image 1 The most common feature of Epstein-Barr virus positive pleural fluid with a lymphocyte-predominant pattern (Diff-Quik, 400).

7 Takei and Mody / EBV-Positive Pleural Effusion In summary, EBV infection/reactivation can account for certain proportions of idiopathic PEs. Polymorphous lymphocytosis is the most common cytologic feature, but atypical features (in both lymphocytes and mesothelial cells) can be seen. FCM immunophenotyping is useful for cases with atypical lymphoid features. Although uncommon, neutrophilic PEs cannot exclude the presence of EBV-D. Address reprint requests to Dr Takei: Dept of Pathology and Genomic Medicine, Houston Methodist Hospital, 6565 Fannin St, Houston, TX 77030; takei327@aol.com. References 1. Light RW. Clinical practice: pleural effusion. N Engl J Med. 2002;346: Ong KC, Indumathi V, Poh WT, et al. The diagnostic yield of pleural fluid cytology in malignant pleural effusions. Singapore Med J. 2000;41: Brock MV, Hooker CM, Yung R, et al. Can we improve the cytologic examination of malignant pleural effusions using molecular analysis? Ann Thorac Surg. 2005;80: Bielsa S, Panades MJ, Egido R, et al. Accuracy of pleural fluid cytology in malignant effusions [in Spanish]. An Med Interna. 2008;25: Thijsen SFT, Luderer R, van Gorp JMH, et al. A possible role for Epstein-Barr virus in the pathogenesis of pleural effusion. Eur Respir J. 2005;26: Centers for Disease Control and Prevention. Epstein-Barr virus and infectious mononucleosis. ncidod/diseases/ebv.htm. Accessed January 4, Nakatsuka S, Yao M, Hoshida Y, et al. Pyothorax-associated lymphoma: a review of 106 cases. J Clin Oncol. 2002;20: Ohori NP, Whisnant RE, Nalesnik MA, et al. Primary pleural effusion posttransplant lymphoproliferative disorder: distinction from secondary involvement and effusion lymphoma. Diagn Cytopathol. 2001;25: Aleman C, Sanchez L, Alegre J, et al. Differentiating between malignant and idiopathic pleural effusions: the value of diagnostic procedures. QJM. 2007;100: Colebunders R, Pen J, Mathijs R. Pleural effusion and ascites in infectious mononucleosis. Acta Clin Belg. 1983;38: Cloney DL, Kugler JA, Donowitz LG, et al. Infectious mononucleosis with pleural effusion. South Med J. 188;81: Odumade OA, Hogquist KA, Balfour HH Jr. Progress and problems in understanding and managing primary Epstein- Barr virus infections. Clin Microbiol Rev. 2011;24: Am J Clin Pathol 2014;142: