Biopsy-proven mantle cell lymphoma in brain parenchyma

Abstract

Brain parenchymal involvement by mantle cell lymphoma is rare and confers a grim prognosis. More commonly, patients with central nervous system manifestations of mantle cell lymphoma have leptomeningeal involvement on radiographic studies with malignant cells found in the cerebrospinal fluid. Risk factors for central nervous system involvement include a high proliferation index, bone marrow involvement, and blastoid morphology. We present an unusual case of a biopsy-proven mantle cell lymphoma mass lesion in the brain parenchyma as the presentation of relapse 6 months after diagnosis.

Mantle cell lymphoma (MCL) accounts for approximately 5% to 10% of non-Hodgkin's lymphoma cases (1–4). Typically MCL presents in the fifth to sixth decades of life with generalized lymphadenopathy and extranodal involvement (bone marrow, gastrointestinal tract, blood) (5). The prognosis is generally poor, with a median survival of only 3 to 4 years (4). The prognosis is worse with the blastoid variant. MCL is characterized by strong expression of cyclin D-1, as well as CD5 and CD20 positivity; CD23 negativity helps distinguish it from chronic lymphocytic leukemia and small lymphocytic lymphoma. The t(11;14) translocation is detected in nearly every patient (6).

Central nervous system (CNS) involvement by MCL occurs in 4% to 26% of patients (1–3, 7, 8). It appears to be a late event in the course of the disease, as most reported cases occur during systemic relapse. The blastoid variant is associated with aggressive disease and is a predictor of CNS involvement. CNS disease is often identified via analysis of the cerebrospinal fluid (CSF), radiographic imaging studies, and clinical symptoms (1). Involvement of the brain or spinal cord is clinically important, as this warrants CNS-directed treatment. Usually there is leptomeningeal or CSF involvement without parenchymal lesions (9).

CASE REPORT

An 81-year-old Caucasian man presented to Baylor University Medical Center at Dallas in May 2009 with fatigue. His physical exam was remarkable only for a 4-cm nontender left supraclavicular lymph node. The patient's complete blood count showed 15,400 white blood cells, hemoglobin 14.6 g/dL, and platelets 41 K/uL. There were 38% blasts, 25% segs, 18% lymphocytes, 17% monocytes, 1% eosinophils, and 1% basophils. A computed tomography (CT) scan of the chest, abdomen, and pelvis showed diffuse adenopathy.

The patient underwent a supraclavicular lymph node core needle biopsy. Immunostains showed cyclin-D1 positivity and an 80% positive MIB-1/Ki-67 (Figure 1). Flow cytometry showed coexpression of CD5 and CD19, negative CD23, and kappa light chain restriction. A bone marrow biopsy showed 90% cellularity with 65% immature cells. The immature cells were medium to large with convoluted nuclei and prominent nucleoli. The immature cells stained positive for CD20 and negative for CD3. Cyclin D-1 was focally positive in the immature cells. Cytogenetic studies of the bone marrow determined that 31% of cells contained the t(11;14) translocation. Flow cytometry on the bone marrow specimen confirmed a monoclonal population of B lymphocytes that were CD5 positive, CD23 negative, and CD10 negative. The final diagnosis was MCL, pleomorphic variant (classified under blastoid morphology by the World Health Organization classification system) with bone marrow involvement. The patient's CSF contained a total cell count of 12, including 7 lymphocytes, 4 monocytes, and 1 neutrophil. There were no malignant cells.

Figure 1.

Lymph node biopsy. (a) 40×. The lymph node shows diffuse effacement of the node by a monotonous population of intermediate-sized lymphoid cells with fine, evenly distributed chromatin, small nucleoli, and very scant cytoplasm. Note several mitotic figures. (b) Cyclin D-1, 4×. Cyclin D-1 positivity supports the diagnosis of mantle cell lymphoma in this lymph node biopsy.

Chemotherapy was initiated with rituximab, cyclophosphamide, hydroxydaunorubicin, oncovin, and prednisone (R-CHOP), and the patient was given intrathecal methotrexate once for CNS chemoprophylaxis. After completing 6 cycles of R-CHOP, the patient was in complete remission with a normal positron emission tomography (PET)–CT scan and bone marrow evaluation.

Six months following diagnosis (in November 2009), the patient developed visual hallucinations. He did not display any other neurologic signs or symptoms. Magnetic resonance imaging of his brain showed a 2.4-cm posterior temporo-occipital lobe mass. CSF and bone marrow analysis done at this time were both negative for malignancy, as was a repeat PET-CT scan. The lesion was subsequently excised for diagnostic evaluation as well as to relieve the mass effect on the brain.

Intraoperatively, the tumor, which was adherent to the dura, was entirely resected. Further analysis of the excisional brain biopsy confirmed MCL (Figure 2). The morphology was similar to the lymph node biopsy done 6 months earlier: it was composed of sheets of discohesive cells with irregularly shaped pleomorphic nuclei and granular chromatin with increased nuclear to cytoplasmic ratios. Apoptotic and mitotic tumor cells were present. The tumor was associated with a granulation type connective tissue, but was also infiltrating the perivascular Virchow-Robin spaces in the adjacent brain tissue. Immuno-stains revealed positivity for CD20, CD79a, CD5, and cyclin D-1, and the MIB-1/Ki-67 index approached 100%. CD3 was negative. Flow cytometry was performed on a portion of the brain specimen, showing a monoclonal population of kappa-restricted lymphocytes with CD5 positivity and CD23 negativity. CSF was again negative for malignancy.

Figure 2.

Brain biopsy. (a) Hematoxylin and eosin, 40×. The malignant cells are surrounded by brain parenchyma. Note the similarity of tumor cells to those in the lymph node, as well as several mitotic figures. (b) Cyclin D-1, 4×. Cyclin D-1 positivity supports the diagnosis of mantle cell lymphoma. (c) MIB-1/Ki-61, 2×. The MIB-1 index approaches 100%, which denotes a very high mitotic rate.

DISCUSSION

This case illustrates a biopsy-proven MCL involving the brain parenchyma as the sole site of relapse. There was a 6-month time window from the initial diagnosis of MCL to CNS relapse, which is a shorter interval than the median in previous reports, in which relapse occurred at 12 months (1), 18 months (3), and 25 months (2). In a retrospective series of 62 MCL patients, four patients developed CNS relapse at a median of 12 months (range, 1–58 months), including two patients with confirmed blastic variant (1). All four of these patients had positive brain magnetic resonance imaging (one patient with blastic variant had multiple parenchymal lesions in the brain, while the other three patients had exclusively leptomeningeal involvement) (1). In a series by Ferrer et al (2007), which followed 82 MCL patients for 5 years, 10 developed CNS involvement (including 5 of 26 patients with blastic variant) with a median of 25 months from diagnosis (range, 5–130 months) (2).

It is of interest that this patient's relapse was isolated to the CNS. Feugier et al reported a frequency of isolated CNS disease at the time of relapse in approximately 1% of patients with diffuse large B-cell lymphoma (10). This entity has been rarely documented in MCL; even rarer is that this patient's relapse involved only the brain parenchyma. To our knowledge, isolated brain parenchymal relapse in MCL has been reported in only one other patient as described by Doolittle et al (9). Lumbar puncture for CSF analysis was negative for malignant cells at the time of diagnosis and on subsequent analysis during our patient's relapse. Our patient's tumor was fairly well circumscribed within the brain parenchyma and was associated with connective tissue, especially around blood vessels. Perhaps these characteristics did not allow for shedding into the CSF, which might explain the repeated negative CSF analysis. It appears that CSF analysis may be insufficient in determining cases of isolated brain parenchymal lesions. This is supported by Doolittle et al, who reported that only 11 of 67 (16%) patients with isolated brain parenchyma involvement at relapse of systemic non-Hodgkin's lymphoma had malignant cells in the CSF (9).

There are several reasons why biopsies of brain masses are not routinely done on patients with relapsed MCL. First, there is usually evidence of other sites of involvement during relapse, negating the necessity of a brain biopsy (9, 10). Additionally, CSF analysis and radiographic imaging may be sufficient to diagnose CNS involvement when it is present. In our review of the literature, we found four cases of mass lesions in the brain thought to be caused by MCL; however, none of these studies involved a tissue diagnosis (1, 7, 9). Since brain parenchymal involvement by non-Hodgkin's lymphoma is rare, Doolittle et al suggested that the clinician should strongly consider doing a stereotactic brain biopsy with microscopic examination (9).

Only a few studies report risk factors for CNS involvement in MCL patients. Ferrer et al followed 82 MCL patients from 1988 to 2003 in a single institutional study (2). Eleven patients (13%) developed CNS disease as determined by CSF analysis, radiology, and/or clinical suspicion. Risk factors for CNS involvement as determined in this study included blastic morphology, high MIB-1/Ki-67 immunostaining, high serum lactic acid dehydrogenase (LDH) levels, and a high-risk International Prognostic Index score. Our patient demonstrated all four of these characteristics. The MIB-1/Ki-67 was originally 80% on his lymph node biopsy and approached 100% on the brain biopsy. His LDH was 2133 U/L (reference range, 135–225 U/L), and he was considered to be at high risk based on the International Prognostic Index (age, stage IV, elevated LDH, and involvement of more than one extranodal site). Ferrer et al's study and other similar studies (1, 9, 10) categorize patients into elevated versus normal LDH, but additional studies that stratify LDH values at presentation are warranted.

CONCLUSION

CNS involvement is usually determined by other less invasive modalities. This case illustrates a biopsy-proven MCL involving the brain parenchyma.