Abstract
Recent studies have investigated the pathogenesis of the class of conditions known as “other iatrogenic immunodeficiency-associated lymphoproliferative diseases” (OIIA-LPDs), particularly in patients with rheumatoid arthritis (RA). Methotrexate (MTX) is a potent cause of LPDs, and withdrawal of MTX can result in spontaneous regression of LPD, which suggests that this drug plays an important role in the tumorigenesis of LPDs. In addition, an impaired immunity against Epstein-Barr virus (EBV) has been obserbed in RA patients. A number of reports describe LPD regression in patients with OIIA-LPDs-RA, but its precise etiology and pathogenesis remain unclear. Furthermore, the phenomenon of relapse/regrowth of LPDs after initial regression has not been well documented. This study retrospectively analyzed the clinicopathological features of OIIA-LPDs-RA patients to determine the influence of EBV infection on regression/relapse of the disease.
Data were collected from 35 patients with RA who developed LPD and who were treated at our institute between 1998 and 2013. All patients had received treatment with MTX. The diagnosis of RA was made according to the American College of Rheumatology criteria. Based on immunohistochemistry performed on paraffin-embedded tissue sections, diagnoses were as follows: diffuse large B cell lymphoma (DLBCL; n=14), Hodgkin lymphoma (HL; n=7), follicular lymphoma (FL; n=4), mucosal-associated lymphoid tissue (MALT; n=3), Hodgkin-like lymphoma (HL-like; n=3), T-cell lymphoma (n=3), polymorphic LPD (P-LPD; n=2) according to the 4th WHO classification. Regarding EBV infection, 16 patients (44%) were positive. Patients with FL, MALT, and T-cell lymphoma were negative for EBV, except for one patient with T-cell lymphoma. In contrast, EBV infection positivity was prevalent in patients with DLBCL, HL, HL-like and P-LPD (46%, 100%, 100%, and 50%, respectively). Although HL indicated a specific phenotype, such as positivity for CD15 and CD30 (83%, and 100%, respectively), and rarely expressed CD20, OCT2 or BOB1 (0%, 14%, 14%, respectively), the phenotypes of HL-like and P-LPD were supposedly intermediate between DLBCL and HL. The phenotypes of FL, MALT, and T-cell lymphoma were the same as those of de novo cases. LPD regression was observed in 23 (66%) of 35 patients, which is more common than that seen in previous reports. Although LPD regression was not documented in patients with T-cell lymphoma, it did occur in all patients with HL, HL-like and P-LPD. In addition, the incidence of regression among patients with DLBCL, FL and MALT was 46%, 75% and 33%, respectively. The relationship between EBV infection and LPD regression among patients with HL, DLBCL, HL-like and P-LPD was statistically significant (p=0.048, Fisher's exact test). Of 23 patients with regression, 13 patients (56%) subsequently showed relapse/regrowth, and the incidence of this phenomenon was relatively high in patients with HL, HL-like and P-LPD (100%, 67%, and 50%, respectively), whereas a lower incidence was seen in patients with DLBCL, FL, and MALT (7%, 33%, and 0%, respectively).
LPD regression was relatively common (66%) in patients with OIIA-LPDs-RA, particularly in patients with B cell phenotypes. There was a significant relationship between LPDs and EBV infection in patients with HL, DLBCL and HL-like, suggesting that underlying EBV infection might influence the immunosuppressant effect of MTX against EBV in those phenotypes. Further, LPD relapse/regrowth was common in patients with HL, HL-like and P-LPD and was unlikely in patients with DLBCL. Further studies would be of benefit to investigate the underlying molecular mechanism of regression/relapse of LPD after withdrawal of MTX.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.
