2025 Volume 12 Pages 255-260
Methotrexate-related lymphoproliferative disorder is a rare but serious complication of methotrexate treatment. Recurrent methotrexate-related lymphoproliferative disorder in the central nervous system is uncommon. An 86-year-old woman, who had been on methotrexate for 3 years to treat rheumatoid arthritis, was admitted with left hemiplegia and sensory disturbance. Brain magnetic resonance imaging revealed a right basal ganglia lesion that was hyperintense on diffusion-weighted imaging and showed patchy enhancement. Methotrexate-related lymphoproliferative disorder was suspected, and methotrexate was discontinued, which led to improvement in her symptoms and partial remission. Twenty-eight months after methotrexate cessation, the patient presented with reduced activity and left hemiparesis. Brain magnetic resonance imaging showed homogeneously enhancing lesions in the right frontal lobe, basal ganglia, and brainstem. A craniotomy for biopsy of the right frontal lobe lesion confirmed the diagnosis of methotrexate-related lymphoproliferative disorder. Whole-brain radiation (40 Gy in 20 fractions) was administered without chemotherapy due to her advanced age and poor performance status. Her symptoms improved, and she was transferred to a rehabilitation hospital. Methotrexate-related lymphoproliferative disorder of the central nervous system can recur spontaneously, even without resumption of methotrexate, making long-term follow-up necessary after methotrexate cessation.
Methotrexate-related lymphoproliferative disorder (MTX-LPD) is classified as "other iatrogenic immunodeficiency-associated LPD" in the fourth edition of the World Health Organization Classification of Tumors of Hematopoietic and Lymphoid Tissues.1) MTX-LPD constitutes a spectrum, ranging from polymorphic lymphoproliferation to diffuse large B-cell lymphoma, which occurs during or after MTX administration. Central nervous system (CNS) involvement is uncommon,1) and clinical understanding of CNS MTX-LPD, particularly recurrences, is limited. However, proton magnetic resonance spectroscopy and oligoclonal bands may aid in predicting benign-type lesions.2) We previously reported 3 cases of MTX-LPD in the CNS, all of which regressed after discontinuation of MTX.2) Here, we present a recurrence of one of these cases in a patient who did not resume MTX after remission of MTX-LPD.
An 86-year-old woman was admitted due to left hemiplegia and sensory disturbance. Her medical history included rheumatoid arthritis, for which she had been taking MTX (8 mg once weekly) for 3 years. Physical examination revealed weakness in the left upper and lower limbs (grade 3/5 on manual muscle testing), hypesthesia, and left lower facial weakness. Her body temperature was 37.7°C, and she was alert with normal cognition. Blood tests revealed: white blood count, 7,900/μL (normal range, 3,300-8,600/μL); lactate dehydrogenase, 211 U/L (normal range, 124-222 U/L); soluble interleukin-2 receptor, 881 U/L (normal range, 122-496 U/L); C-reactive protein, 2.26 mg/dL (normal range, 0.0-0.15 mg/dL); and Epstein-Barr virus (EBV) DNA, 3.36 logIU/mL. Oligoclonal bands were detected in the cerebrospinal fluid.
Brain computed tomography (CT) and magnetic resonance imaging (MRI) revealed a 10 mm-diameter lesion in the right basal ganglia (Fig. 1A-D). The lesion was low-density on CT, hyperintense on fluid-attenuated inversion recovery (FLAIR) imaging and diffusion-weighted imaging (DWI), with patchy enhancement after gadolinium administration. Proton magnetic resonance spectroscopy showed an elevated lipid peak and mildly elevated choline-to-N-acetyl aspartate ratio. Whole-body CT revealed no lymphadenopathy or other lesions suggestive of malignancy, aside from those in the brain.
Computed tomography (A), fluid-attenuated inversion recovery imaging (B), diffusion-weighted imaging (C), and gadolinium-enhanced T1-weighted imaging (D) at the time of admission. Computed tomography (E), fluid-attenuated inversion recovery imaging (F), diffusion-weighted imaging (G), and gadolinium-enhanced T1-weighted imaging (H) 16 months after discontinuation of methotrexate.
Based on her medical history and imaging findings, MTX-LPD was suspected, and MTX was discontinued. Her symptoms resolved within 4 weeks. Follow-up CT and MRI showed regression of the lesion over time, with partial remission achieved 16 months after MTX cessation (Fig. 1E-H). At this point, regular neurosurgical follow-up was discontinued. The otorhinolaryngology team, originally following her for a parapharyngeal space tumor suspected to be a schwannoma, continued to monitor her with regular MRIs, including brain imaging. These scans showed stable CNS disease. An immunosuppressive drug was not needed as her rheumatoid arthritis symptoms were well-controlled.
However, 12 months later, she presented with reduced activity and left hemiparesis, leading to hospitalization. Blood tests revealed: white blood count, 4,600 /μL; lactate dehydrogenase, 203 U/L; soluble interleukin-2 receptor, 763 U/L; C-reactive protein, 0.45 mg/dL; and EBV DNA, not detectable. Oligoclonal bands were not detected in the cerebrospinal fluid. Brain MRI detected multiple lesions in the brain stem, right basal ganglia, and right frontal lobe. All lesions were hyperintense on FLAIR and iso- to hyperintense on DWI, with relatively homogeneous enhancement (Fig. 2).
Magnetic resonance imaging of recurrent disease in the brain stem (A-C), basal ganglia (D-F), and right frontal lobe (G-I).
A right frontal craniotomy was performed to biopsy the right frontal lobe lesion for histological examination (Fig. 3A-F). Grossly, the lesion appeared as a grayish mass well-delineated from the surrounding normal brain. Hematoxylin and eosin staining of the specimen showed tightly arranged rosette-like lymphoid cells and necrotic areas. Immunohistochemical staining was negative for CD3, CD5, CD10, and EBER in situ hybridization but positive for CD20 and Ki-67. Recurrent MTX-LPD was diagnosed, and 6 mg of once-daily intravenous betamethasone was administered, followed by a taper to 0.5 mg daily over 2 weeks. This led to temporary symptom improvement. Repeat brain MRI revealed enlargement of the brainstem lesion and edema in the surrounding tissue (Fig. 4A-B). Given her advanced age and poor performance status, we chose not to administer chemotherapy. Whole-brain irradiation was initiated 14 days after the biopsy (40 Gy in 20 fractions), and betamethasone was increased to 2 mg once daily, where it was maintained. At the conclusion of radiotherapy, her symptoms improved.
(A) Intraoperative photograph of the right frontal lobe lesion. (B, C) Hematoxylin and eosin staining. Immunohistochemistry for CD20 (D), Ki-67 (E), and Epstein-Barr virus-encoded small RNA (F).
Imaging before initiating radiotherapy (A, B), 2 weeks after (C, D) and at the end of radiotherapy (E, F). Panels A, C, and E are fluid-attenuated inversion recovery images and B, D, and F are gadolinium-enhanced T1-weighted images.
Follow-up MRI showed the lesions had decreased in size, particularly the brainstem lesion (Fig. 4C-F). Due to her limited activities of daily living and her Karnofsky performance status score of 50, she was transferred to a rehabilitation hospital. Her activities remained stable for 2 months, at which point she developed a urinary tract infection and sepsis, causing a decline. An MRI taken 4 months after whole-brain irradiation showed no recurrent lesion or atrophic changes in the brain. She was still alive 14 months after irradiation but remained in the rehabilitation hospital.
Reported 2- to 3-year recurrence rates for systemic MTX-LPD range from 18% to 45%, with 5-year survival exceeding 70%.3-5) In contrast, a 2023 systematic review of 13 CNS MTX-LPD patients reported that only one experienced recurrence after an initial 1-month period of regression following MTX cessation.6) Proton magnetic resonance spectroscopy findings of an elevated lipid peak and slightly elevated choline-to-N-acetylaspartate ratio, in conjunction with the presence of oligoclonal bands in the cerebrospinal fluid, might reflect non-monoclonal lymphoproliferative histology and a benign nature.2) In our patient, the pathology changed from a benign phenotype to a malignant one at the time of recurrence. Although the data regarding recurrence of MTX-LPD in the CNS is limited, long-term follow-up is necessary even after stabilization in a state of partial remission, as shown by our patient.
EBV DNA was positive in our patient when CNS MTX-LPD was initially diagnosed, but negative at the time of recurrence. MTX can induce EBV-positive MTX-LPD by reactivating latent EBV infection, which activates nuclear signaling pathways and downregulates the anti-apoptotic protein BCL2.7) However, we do not assume that the conversion to EBV-negative MTX-LPD in our patient occurred solely due to MTX. It is possible that the patient's MTX-LPD changed to another phenotype at the time of recurrence, or it had polyclonal features originally. Kato et al.8) recently described a case of systemic MTX-LPD in which EBV testing was positive in the initial MTX-LPD lesion in the liver but negative in a recurrent lesion in the appendix. They concluded that the contradictory EBV testing results were due to a change in the tumor cell clone at the time of recurrence. Unfortunately, MTX-LPD that changes tumor phenotype at recurrence is not well understood. Further study is warranted, ideally with comprehensive molecular analysis data.
In our patient, MTX-LPD lesions appeared hyperintense on FLAIR imaging and iso- to hyperintense on DWI, with gadolinium enhancement. Although the presence of multiple lesions at the time of recurrence and her medical history highly suggested MTX-LPD,9) several radiological findings suggested otherwise. First, the enhancement was homogeneous and "nodular" in appearance. Most primary CNS lymphomas show homogeneous enhancement,10) while MTX-LPD lesions tend to exhibit ring-like enhancement.6,9) Second, 2 lesions were located adjacent to the cortex or the ventricular system. Although MTX-LPD usually occurs as multiple lesions in spatially distinct locations,9,11) lesions adjacent to the cerebrospinal fluid space generally suggest CNS lymphoma in immunocompetent patients.9) Alternatively, they can suggest leptomeningeal tumor dissemination. Moreover, we presumed that the fact that our patient had completely stopped taking MTX made recurrent MTX-LPD unlikely. Future studies describing and examining the radiological features of CNS MTX-LPD are warranted.
Although MTX discontinuation is considered the first-line treatment for MTX-LPD,6) treating intractable or recurrent MTX-LPD is challenging. In a literature review of 15 cases of MTX-LPD,12) regression after cessation of MTX was achieved in 10; chemotherapy was required in the other 5, with radiation therapy in one and rescue resection in one. In our patient, whole-brain radiotherapy was administered without chemotherapy due to her age and poor performance status. Previously published guidelines have indicated the recommended doses for primary CNS lymphoma when remission-induction therapy with high-dose MTX is contraindicated.13) We selected 40 Gy in 20 fractions, considering her advanced age and concerns about neurotoxicity. Palliative radiation therapy was not considered, as the estimated survival for CNS MTX-LPD is reasonably long with treatment. A report describes a primary CNS MTX-LPD case treated with 45 Gy of radiation therapy alone.14) However, to the best of our knowledge, this is the first report of using radiation therapy alone for recurrent CNS MTX-LPD after MTX cessation. Future studies are needed to establish the efficacy and safety of radiation therapy for CNS MTX-LPD.
ConclusionCNS MTX-LPD can spontaneously recur after remission following MTX cessation. Long-term follow-up is necessary for CNS MTX-LPD patients after stabilization to monitor for lesion relapse.
We thank Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.
Y. Okada and Y. Tomita designed the project and wrote the draft. Y. Tomita collected the data. H. Ueno, K. Matsuura, K. Makino, N. Kidani, and K. Muraoka aided in interpreting the imaging data. H. Nishimori, N. Hirotsune, and S. Nishino supervised the project. All authors reviewed and approved the manuscript.
Not applicable.
The authors obtained written informed consent from the patient.
The authors obtained written informed consent from the patient.
Not applicable.
Not applicable.
The authors have no conflicts of interest to declare.
