Physiologic MRI of a tumefactive multiple sclerosis lesion

White matter lesions caused by MS have variable appearance but rarely exceed 2 cm in diameter.^1,2 On MRI, acute MS plaques often show enhancement with gadolinium and may be associated with surrounding edema.³ However, when patients present with a solitary, large enhancing lesion with marked surrounding edema (>3 cm), possible diagnoses also include glioma or other neoplasms.¹ Because MRI may not differentiate these diseases, diagnosis of these tumefactive lesions often involves a brain biopsy.

Some of these diagnostic problems might be resolved by advanced physiologic MRI techniques. For example, magnetization transfer (MT) MRI improves the sensitivity of MRI for detecting MS lesions and may delineate the extent of demyelination.⁴ Other physiologic MRI techniques that might be useful for the evaluation of MS lesions are diffusion MRI,⁵ which evaluates tissue integrity, and perfusion MRI, which measures regional cerebral perfusion. Proton magnetic resonance spectroscopy (¹H MRS) also has been applied to the evaluation of MS lesions⁶; this technique allows for noninvasive measurement of certain biochemical changes associated with the lesions. We report the application of these physiologic MRI techniques in a young woman with MS who had a large contrast-enhancing lesion with marked surrounding edema. In this unusual case, physiologic MRI helped to exclude a neoplastic process.

Case report. A 20-year-old woman with a diagnosis of MS presented with a 1-week history of progressive weakness and coordination problems in her left hand, gait difficulty, and mildly slurred speech. She had been maintained on interferon beta-1b, 0.25 mg every other day, for the treatment of her MS. On examination her height was 62 inches, her weight was 147 pounds, and her vital signs were normal. Her cognitive function was intact, but she had difficulty writing with her left-dominant hand due to the left hemiparesis. Cranial nerves were intact except for nystagmus associated with leftward and upward gaze; she also had jerky pursuit movements bilaterally. Motor examination showed mild weakness primarily in the left upper extremity, involving the distal more than the proximal regions. Her sensory examination showed decreased light touch distal to her left elbow, but intact pinprick sensation, and decreased proprioception and vibratory sensation in her left fingers. She had mild truncal ataxia and dysmetria in the left upper extremity. She also had hyperactive deep tendon reflexes and the presence of Babinski sign bilaterally. Her gait was wide based and mildly ataxic. She scored 3.5 on the Expanded Disability Status Scale (EDSS).⁷

Figure 1 shows the results of the initial MR studies, performed on a 1.5-T GE scanner. The fluid-attenuated inversion recovery (FLAIR) images showed several small periventricular white matter lesions consistent with MS. However, one large lesion in the right frontoparietal white matter had unusual radiographic features for MS, such as large size (2.8 × 1.6 cm) with mass effect, marked surrounding edema (4.5 × 6.5 cm), and ring enhancement after gadolinium injection (see figure 1, A and B). These MRI findings raised the possibilities of a neoplastic process versus a tumefactive MS lesion.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Initial structural and physiologic MRIs and magnetic resonance spectroscopy obtained 1 month after onset of clinical symptoms. (A) Fluid-attenuated inversion recovery image, (B) postgadolinium T1-weighted image, (C) magnetization transfer ratio map, (D) regional cerebral blood flow map, and (E) apparent diffusion coefficient map of an axial slice through the center of the large enhancing lesion in the right frontoparietal region. Localized ¹H spectra were obtained from a voxel (3.7 cc) in the lesion (bottom right, lower spectrum) and in the contralateral normal-appearing white matter (upper spectrum). Spectra are scaled (repeat time/echo time = 3,000/30 msec). See text for details. MI = myoinositol; CHO = choline; CR = creatine; NA = N-acetyl; LAC/LIP = lactate and lipids.

¹H MRS of the mass lesion showed slightly elevated choline (CHO, +10%), decreased N-acetyl compounds (NA, -39%), decreased creatine (CR, -52%), decreased myoinositol (MI, -38%), and excess lactate and lipids (LAC/LIP) compared with the contralateral normal-appearing brain region. The MT ratio (MTR; see figure 1C) was markedly decreased in the enhancing lesion (MTR = 0.22) and moderately decreased in the surrounding edema (MTR = 0.30) compared with normal-appearing white matter (MTR = 0.45).

Regional cerebral blood flow (rCBF) was calculated from the signal changes of a dynamic gradient-echo planar imaging (EPI) scan (echo time [TE]/repeat time [TR] = 30/2,500, 39 time points) during a bolus injection of gadolinium contrast agent. The rCBF was decreased (-43%) in the central part of the enhancing lesion as well as in the surrounding edema (see figure 1D). With diffusion MRI, using spin-echo EPI (TE/TR = 109/5,000; b values 0 and 400 sec/mm²), the apparent diffusion coefficient (ADC) was increased in the enhancing lesion (+36%) and the surrounding edema (+67%; see figure 1E).

Because the physiologic MRI findings and the clinical history were suggestive of an acute MS plaque rather than a neoplasm, needle biopsy was deferred and a course of IV methylprednisolone was recommended instead. The patient initially declined the steroid treatment and waited for the exacerbation to subside. However, her left-arm weakness further progressed 1 week later, and she was hospitalized for a 3-day course of IV methylprednisolone (1 g/d). She showed moderate improvement in her symptoms and could return to school. Her left upper-extremity weakness and sensory deficits gradually normalized over the next 3 months; however, her EDSS score remained at 3.5 because she developed new-onset, right lower-extremity weakness.

On the repeat MRI, obtained 3 months after the initial scan, the patient showed several new, small enhancing MS lesions in the left periventricular and right parietal regions (figure 2B). However, the large initial lesion was substantially smaller and did not enhance with gadolinium. The mass effect with surrounding edema also had resolved at the time of the repeat scan. On the repeat ¹H MRS, the previously tumefactive MS lesion showed further elevated CHO (+60%), persistently decreased NA (-23%), excess LAC/LIP, and normalized CR (-3%). The previously decreased MI became elevated (+36%) (see figure 2). The MTR in the right frontal white matter normalized compared with the initial value. However, some areas, such as the center of the original large lesion, still showed markedly reduced MTR (see figure 2C, green areas). Finally, the rCBF and ADC values at the original lesion site also normalized in the repeat scan (see figure 2, D and E).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Structural and physiologic MRIs and magnetic resonance spectroscopy obtained 3 months after the initial MR studies. The resolving mass lesion is demonstrated in the (A) fluid-attenuated inversion recovery image, (B) postgadolinium T1-weighted image, (C) magnetization transfer ratio map, (D) regional cerebral blood flow map, and (E) apparent diffusion coefficient map of an axial slice through the center of the previously enhancing lesion in the right frontal lobe. Localized ¹H spectra were obtained from a voxel (4.0 cc) in the region of the previously enhancing lesion (bottom right, lower spectrum) and the contralateral normal-appearing white matter (upper spectrum; repeat time/echo time = 3,000/30 msec). See text for details. MI = myoinositol; CHO = choline; CR = creatine; NA = N-acetyl; LAC/LIP = lactate and lipids.

Discussion. The large ring-enhancing lesion in our patient represented a diagnostic dilemma between an acute tumefactive MS plaque and a neoplasm. Because conventional MRI cannot differentiate between these two pathologies, such lesions often require brain biopsies for diagnosis and treatment planning. Our data suggest that the differential diagnosis of tumefactive MS lesions from neoplasms may be improved by physiologic MRI, especially perfusion MRI; active areas of neoplasms typically have vascular proliferation and should show increased rCBF.⁸ In contrast, the rCBF in the enhancing lesion of our patient was reduced, suggesting decreased vasculature; therefore, perfusion MRI was consistent with a tumefactive MS lesion rather than a neoplastic process.

The reduced MTR in the enhancing lesion was consistent with a demyelinating MS lesion.⁴ However, because the MTR of neoplastic lesions is generally also decreased,⁹ the MTR might not differentiate the two lesion types. The MRS findings of decreased NA and CR most likely reflected axonal loss, whereas the elevated CHO and lipids probably resulted from demyelination and cell membrane injury, and the subsequently elevated MI represented increased glial content. Although these MRS profiles are characteristic of MS lesions,^4,6 gliomas may show similar biochemical changes.¹⁰

We present a patient with a history of relapsing-remitting MS who had new-onset symptoms due to a large tumefactive MS lesion. Conventional MRI couldn't determine whether the large enhancing lesion with mass effect was due to a neoplasm or an acute tumefactive MS lesion. Because physiologic MRI, especially perfusion MRI, did not support the presence of a neoplasm, the patient was treated for a presumptive MS exacerbation, and a needle biopsy, with its associated morbidity, was avoided.