[Background] Methotrexate - induced lymphoproliferative disorders (MTX-LPDs) belong to the group of other iatrogenic immunodeficiency-associated LPDs, as defined in the revised 4th edition of the World Health Organization (WHO) classification. The phenomenon of LPD regression after MTX withdrawal is a specific event that develops in patients with autoimmune diseases (ADs), such as rheumatoid arthritis (RA). Although patients with regressive MTX-LPDs have better overall survival (OS), relapse/regrowth (R/R) events after regression are one of the factors of poor prognosis. Recent studies including ours have shown that lymphocytes play an important role in the regressive LPD phenomenon in MTX-LPDs (Rheumatology (Oxford). 2017;56:940-946, Front Immunol. 2018;4;9:621) ; however, little is known of the details regarding differences among the LPD subtypes and the influence of lymphocytes on R/R events. In this study, we analyzed patients with regressive Hodgkin lymphoma (HL) and diffuse large B cell lymphoma (DLBCL), especially focusing on the influence of the absolute lymphocyte count (ALC) on R/R events and clinical outcomes.

[Methods] Data were collected from 25 patients with ADs who developed LPD and had regressive LPD after MTX withdrawal at our institutions. All diagnoses were confirmed based on immunohistochemistry analysis of paraffin-embedded samples. All pathological samples were classified and diagnosed according to the WHO classification. The ALCs were determined at the time of MTX withdrawal due to LPD development (0 M), 1 month after MTX withdrawal (1 M), 6 months after MTX withdrawal (6 M), and at the time of R/R. The OS was calculated between 0 M and the time of the last observation. Statistical analyses were performed using EZR software.

[Results] Of 25 patients with regressive MTX-LPDs, the median age at 0 M was 67 years (range: 44ー84). Ten men and 15 women were included. The basal ADs were RA (N = 23), psoriasis vulgaris (N = 1), and systemic lupus erythematosus (N = 1). The median duration of MTX treatment was 5.5 years. Seventeen cases of DLBCL and 8 cases of HL were included. The median ALCs at 0 M and 1 M in all patients were 627/µL and 1364/µL, respectively. The median ALC ratio (ALC at 1 M divided by ALC at 0 M) was 2.3 (0.6-8), and it indicated over 1 in 24 patients, confirming ALC recovery at 1 M. Regarding the ALCs at 0 M and 1 M, significant differences between DLBCL and HL were not detected (p=0.215 and p=0.77, respectively).

 Of all patients, the median duration from the time of MTX duration to R/R was 12 months (range, 7-41). Thirteen patients remained in the remission state without R/R (Regression group), whereas 12 patients experienced an R/R event (R/R group). In the latter group, the median ALCs at 0 M, 1 M, 6 M, and R/R were 582/µL, 1194/µL, 1158/µL, and 565/µL, respectively. The ALC significantly increased at 1 M from 0 M (p=0.006) and decreased at R/R from 1 M (p=0.05), suggesting that the ALC recovery diminished when the LPD underwent R/R. On the other hand, among 13 patients in the regressive group, the median ALCs at 1 M, 6 M, and at the time of the last observation were over 1500/µL (1732/µL, 1782/µL, and 1574/µL, respectively), although that at 0 M was less than 1000/µL (991/µL). To investigate the influence of ALC on the OS, we analyzed the clinical definition of the ALC cut-off of 1000/µL after the ALC recovery observed at 1 M. The result indicated that patients with ALC greater than or equal to 1000/µL after 6 M had significantly better OS compared to those with ALC less than 1000/µL after 6 M (5-year OS, 100% vs. 43.8%, p=0.00048), considering that the ALC is one of the strong prognostic factors in the clinical outcomes of MTX-LPDs. It is of note that the R/R rate in patients with HL was higher than that in those with DLBCL (100% and 23.4%, respectively), and all 8 patients with HL had an ALC at R/R of less than 1000/µL.

[Summary] In this study, we demonstrated that ALC recovery was detected at 1 M, which continued during the clinical course in patients who maintained a remission state, although it decreased to less than 1000/µL after ALC recovery in patients with R/R LPDs. In addition, the ALC of 1000/µL was a critical level affecting the clinical outcomes. Regarding LPD subtypes, the reason for higher R/R rates observed in HL compared to those in DLBCL was thought to be dependent on the rates of lower ALCs.

[Conclusion] Our data suggest that the ALC is one of the predictive factors for R/R and OS in patients with regressive MTX-LPDs.

Disclosures

Tokuhira:Mitsubishi Tanabe Pharma Corporation: Speakers Bureau; AYUMI Pharmaceutical Corporation: Speakers Bureau; Bristol-Myers Squibb: Speakers Bureau; Chugai: Speakers Bureau. Tamaru:Nichirei Bioscience INC.: Research Funding; Takeda Pharmaceutical Company Limited: Speakers Bureau. Kizaki:Nippon Shinyaku,: Research Funding, Speakers Bureau; Celgene: Research Funding, Speakers Bureau; Bristol-Myers Squibb: Research Funding, Speakers Bureau; Novartis: Speakers Bureau.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution