Intravascular large cell B lymphoma presenting as central nervous system pseudo-vasculitis: A rare diagnostic challenge

Abstract

Intravascular large B cell lymphoma (IVLBCL) is a very rare subtype of aggressive non-Hodgkin B cell lymphoma characterized by intravascular proliferation of clonal B lymphocytes, classically associated with pulmonary and cutaneous disease and, less frequently, with central nervous system (CNS) involvement. Brain imaging findings are usually non-specific, with evidence of multiple vascular occlusions and stroke as non-specific multifocal abnormalities. We present an exceptionally rare case of IVLBCL in a patient with unexplained inflammatory syndrome with B symptoms and rapidly progressive neurological impairment, with multifocal hemorrhagic and tumefactive brain lesions seen on MRI. We suggest that in this clinical setting, the presence of tumefactive and hemorrhagic lesions should raise suspicion for IVLBCL and lead to the decision to perform a biopsy, which, nonetheless, remains the diagnostic gold standard.

Introduction

Intravascular large B cell lymphoma (IVLBCL) is a very rare subtype of aggressive non-Hodgkin B cell lymphoma characterized by intravascular proliferation of clonal B lymphocytes. This malignant condition is a diagnostic challenge because of its non-specific and variable presentation, which earned it the name of oncologist’s “great imitator.” ¹ Prognosis is highly related to delay in diagnosis. CNS involvement has been historically described within the classical variant (formerly named “European”), associated with pulmonary and cutaneous disease. ² It has a poor survival rate compared to non-CNS IVBCL. About 60% of the cases are post-mortem diagnosis on autopsy pieces. ³

This report highlights the diagnostic challenges of IVLBCL and emphasize the potential role of neurological imaging in identifying atypical forms of IVLCBL.

Case presentation

We present a case of a 50-year-old Caucasian woman who was admitted for anemia (Hb 9.1 g/dl), weight loss (5 kg), low grade fever, and inflammatory syndrome. She had a history of hypermobile Ehlers-Danlos syndrome (type III) without relevant past medical history nor any chronic treatment. The symptoms were present for three months and they included fatigue, occasional fever peaks up to 38.5°C associated with night sweats. One month earlier she has been admitted for anemia work-up. Endoscopic investigations (gastroscopy and colonoscopy) and abdominal CT scan were both unable to identify an underlying cause. The patient was discharged after one unit of red blood cells transfusion and intraveinous iron supplementation (1000 mg), with temporarily clinical and biological improvement (Hb 10.4 g/dl).

She was readmitted 2 months later due to a recurrence of anemia, measured at 7.8 g/dL of hemoglobin without other blood count abnormalities (Table 1, column 2). She reported worsening of B symptoms. Laboratory findings revealed normocytic regenerative anemia associated with elevated LDH, inflammatory syndrome and hyperferritinemia (Table 1, column 2). Infectious (Brucella, Coxiella, Parvorvirus, Leishmania, Bartonella) and auto-immune serologies (FAN, ANCA, and anti-cardiolipines) were negative, as well as repeated prolonged blood cultures. Transthoracic echocardiography revealed no sign of endocarditis. PET-CT showed non-specific homogenous hypermetabolism of the spleen and bone marrow associated with diffuse hypermetabolism of both lung parenchyma without radiological abnormalities on CT scan. Bone marrow aspiration and random skin biopsy were not contributive.

Table 1.

Laboratory findings at admission of both hospitalizations with normal values.

Laboratory findings	First hospitalization	Second hospitalization (2 month later)	Normal value
Hemoglobin (g/dL)	9.1	7.8	13.0–18.0
White blood cells (103/µL)	6.58	6.47	4.00–10.00
Blood platelets (103/µL)	198	212	150–350
C-reactive protein ( mg/L)	17.5	99.7	<5
Ferritin (µg/L)	932	1831	13–150
Soluble transferrin receptor (mg/dl)	8.16	-	1.9–4.4
LDH (UI/L)	593	883	<2

At day 13 of hospitalization, she presented an acute episode of transient dysarthria. Non-enhanced cerebral CT scan showed subacute left parietal cortico-subcortical hypodensity, compatible with a subacute ischemic lesion (Figure 1(a)). EEG showed mild encephalopathy (occipital dominant rhythm 8.5 Hz, abundant theta-delta activities) but no epileptic abnormalities. Brain MRI demonstrated multiple tumefactive intraparenchymal lesions characterized by a hypersignal on T2-weighted images (T2-WI) and Fluid Attenuated Inversion Recovery (FLAIR) sequence, without enhancement after gadolinium injection, in frontal, occipital, and temporal lobe; as well as the left postcentral gyrus with corresponding areas of cortico-subcortical hemorrhagic lesions on diffusion-weighted images and T2*-weighted images (Figure 1(b)–(d)). An ill-defined T2-WI hyperintense lesion in the central pons was also observed (Figure 1(e)). A CT venography was performed to exclude cerebral venous thrombosis and turned out to be negative. Lumbar puncture revealed one nucleated element lymphocyte with normal CSF protein and glucose levels.

Figure 1.

CT scan shows subacute left parietal cortico-subcortical hypodensity, compatible with a subacute ischemic lesion (a). Brain MRI FLAIR images show multiple tumefactive intraparenchymal lesions characterized by a hypersignal (b). Diffusion-weighted images and T2*-weighted images (c and d) demonstrated the ischemic and hemorrhagic nature of the lesions (white arrows). FLAIR images show a central pontine hyperintense lesion (e).

Neurological state of the patient worsened some days after, with iterative episodes of transient dysarthria, right facial palsy and alteration of consciousness state with mutism. EEG showed mild encephalopathy described as generalized background slowing in the theta and delta frequency, but no epileptic abnormalities. Due to high clinical suspicion of seizure and recording compatible with post-ictal state, clobazam (5 mg three times a day with gradual taper) and levetiracetam (1 gr IV, then 1 g twice a day) were started. In parallel, a second brain MRI showed new foci of hyperintensity on FLAIR sequences and significative progression of the pre- and post-central tumefactive lesions as well as the cortical subcortical petechial foci on T2*-WI (Figure 2(a)). Enhanced black-blood sequence did not show vessel wall enhancement or vessels thickening suggestive of vasculitis (Figure 2(b)). However, post-contrast black-blood images showed subtle signs of blood-brain-barrier rupture, whether the 3D-T1 (SPGR) sequence did not show any enhancement; an unusual finding although possible, as increased sensitivity of black-blood sequences to blood–brain barrier rupture has already been described. ⁴

Figure 2.

An MRI study performed days after the first one shows new foci of hyperintensity on FLAIR sequences and significative progression of the pre- and post-central tumefactive lesions as well (a). 3D-T1 post-contrast images did not show any enhancement (b). Enhanced sequence did not show vessel wall enhancement or vessels thickening suggestive of vasculitis, however they showed subtle signs of blood-brain-barrier rupture (c).

At this point, we raised several hypotheses. The presence of multiple apparently ischemic, nonterritorial lesions as well as areas of vasogenic edema with modest signs of blood–brain barrier rupture visualized on black-blood images led from before to hypothesize lesions on a venous ischemic basis, which were excluded by contrast-enhanced CT. The other hypotheses formulated were PRES, unlikely due to the presence of more frankly ischemic areas and less compatible with clinical presentation; and vasculitis, unconfirmed on black-blood images. The presence of microhemorrhagic deposits at the lesions and areas of barrier rupture also led to the suspicion of amyloid-β related angiitis. Finally, we had a clinical suspicion of IVLBCL, however very rarely IVLBCL can manifest with cortical venous thrombosis or arterial dissection.^5–7

Due to uncertainty of diagnosis and major therapeutical implications with different treatment between hypothesis, especially in case of IVLBCL, cerebral biopsy of a right temporal posterior lesion was decided. Histologic evaluation revealed medium to large-sized atypical lymphoid cells tightly packed within small or intermediate-sized blood vessels mixed with red blood cells (Figure 3(a)). Under higher magnification, tumor cells exhibit scant cytoplasm, high nuclear/cytoplasmic ratio, irregular nuclear contours, and prominent nucleoli. Immunohistochemical staining revealed that neoplastic cells expressed pan B-cell markers (Figure 3(b)) and were positive for CD45, MUM-1, BCL-2, Bcl6, and c-Myc. Based on histopathological and immunohistochemistry results, the diagnosis of IVLBCL non-germinal center B-cell type was retained. Chemotherapy treatment was initiated with R-CHOP associated with high dose methotrexate, and the patient showed slow neurological recovery. MRI at 1-month and 2-month follow-up after biopsy showed a regression of the lymphomatous lesions (Figure 4). After six cycles the patient fully recovered clinically and showed a good tolerance to the chemotherapy.

Figure 3.

Histologic evaluation revealed medium to large-sized atypical lymphoid cells tightly packed within small or intermediate-sized blood vessels mixed with red blood cells (white asterisk)(a). Under higher magnification, tumor cells exhibit scant cytoplasm, high nuclear/cytoplasmic ratio, irregular nuclear contours, and prominent nucleoli. Immunohistochemical staining revealed that neoplastic cells expressed pan B-cell markers and were positive for CD45, MUM-1, BCL-2, Bcl6, and c-Myc (white asterisk)(b).

Figure 4.

FLAIR images of the MRI follow-up at 2 months after biopsy shows a regression of the lymphomatous lesions.

Discussion

We here reported a case of IVLBCL presenting with prevalent neurological involvement.

IVLBCL is a rare and aggressive form of lymphoma, representing sometimes a diagnostic challenge. In the presence of inflammatory syndrome of unknown origin, hyperferritinemia, high LDH level, associated with neurological symptoms and radiological abnormalities, despite negative skin biopsy and bone marrow aspiration, this diagnosis should be evoked in order to give prompt and tailored treatment. Brain biopsy was necessary to confirm the clinical suspicion due to insufficient radiological signs in favor of this entity and impossibility to disclose lymphoma by other means.

The abnormal findings on neuro-imaging of IVLBCL are due to the proliferation of clonal lymphocytes within small vessels generating venous congestion and demyelination and are reflected by five imaging patterns including (1) a hyperintensity lesion in the pons on T2WI, excluding the pontine tegmentum and ventrolateral region, (2) non-specific white matter lesions, especially in the periventricular lesions, (3) cortical subacute infarct-like lesions reflecting small vessel ischemia or demyelination, (4) meningeal enhancement, and (5) mass-like lesions suggestive of extensive vasogenic edema, which can disappear in follow-up MR studies.^8,9

The lesion pattern demonstrated in our case is quite rare and has only been reported in one sporadic case by Chen et al., ⁷ where the authors, on suspicion of lesions of a vascular nature, had performed an angiography prior to diagnostic biopsy, which demonstrated multiple dissections of medium-calibre intracerebral arterial vessels. The authors’ histopathological explanation for this phenomenon is transparietal infiltration of the cerebral arteries by malignant lymphomatous cells, resulting in destruction and thus dissection of the vessel wall. However, in our case, post-contrast black-blood images showed no parietal enhancement of intracranial vessels. Although Schaafsma et al. described two interesting cases of biopsy-proven intravascular lymphoma that mimicked the vessel wall MRI (vwMRI) appearance of vasculitis, with concentric arterial wall thickening and enhancement, ¹⁰ Havenon and McNally showed that the vwMRI findings of IVBCL are highly variable and that vessel wall enhancement and thickening was not consistent in biopsy proved IVBCL. ¹¹ An explanation to these results might be that the resolution of vwMRI protocols, 0.5–0.7 mm voxels, does not allow reliable assessment of the distal intracranial vessels, which introduces the possibility of a false negative study. Furthermore in vasculitis, the intramural lymphocytic invasion provides a pathophysiologic mechanism for concentric vessel wall enhancement, whereas in IVL the vessel wall itself is frequently spared. In our case the infiltration might, therefore, be predominantly perivascular and vascular damage may occur at the level of small-caliber vessels, not visible on black-blood images or on the CT venogram. Nonetheless, in the presence of a similar imaging picture and in the clinical suspicion of intravascular lymphoma, we believe that the indication for biopsy is a priority and completion with angiography may help to identify the underlying vascular lesion.

Abe et al. ⁸ demonstrated that more or less 90% of patients with IVLBCL present brain lesions on pretreatment MRI, with the hyperintensity in the pons as the most frequent sign (57.6%) and might have a potential diagnostic value. In this study, infarct lesions were associated with a poorer prognosis, which might be related to a more aggressive or advanced disease. After an adapted treatment, all abnormal findings improved or resolved in most of the patients on the 3 months follow-up MRI. Although centropontic hyperintensity may be also observed in pontine osmolytic demyelination and posterior reversible encephalopathy syndrome, clinical and medical history in a patient with IVLBL share little with those patients with the above-mentioned disease. Other differential diagnosis includes vasculitis, primary central system lymphoma, the last more often presenting as solitary lesions or inflammatory cerebral amyloid angiopathy notably characterized by subcortical white matter lesion and multiple cerebral microbleeds. ¹²

Conclusion

IVLBCL represents a diagnostic challenge and it is essential for a prompt treatment and a better prognosis. We suggest that in a patient with unexplained inflammatory syndrome with B symptoms and rapidly progressive neurological impairment, IVLBCL should be suspected and oriented brain imaging should be performed to exclude other neurological entities requiring other tailored treatment and to look for typical radiological signs of this pathology. Imaging findings such as diffuse petechial lesions on T2*-WI and ischemic and/or hemorrhagic with associated cortico-subcortical tumefactive lesion must raise suspicion for IVLBCL and potentially guide biopsy. However, biopsy stays the gold standard for diagnosis and to start chemotherapy treatment.

ORCID iDs

Denis Vandermeersch https://orcid.org/0000-0003-1277-2297

Valeria Onofrj https://orcid.org/0000-0003-1170-7282