Magnetic Resonance Imaging Appearance of Giant Intracerebral Tuberculoma: A Retrospective Analysis

. Background : Giant intracerebral tuberculomas are rare lesions but should be considered in the differential diagnosis of intracranial space-occupying lesion in an endemic region. Objective: The purpose of this study is to analyze the clinical data and magnetic resonance imaging (MRI) findings of giant intracerebral tuberculomas to improve the diagnostic precision. Material and Methods : The clinical and MRI findings of 22 patients of giant intracerebral tuberculoma were analyzed retrospectively. For the statistical analysis independent sample Student t-test was used. Results: For 22 patients included in this sample the giant intracerebral tuberculoma was of size more than 2.5cm. The majority of the giant tuberculomas (19 patients (86.4%))was located in the supratentorial area.T2-weighted hypointense core of giant tuberculoma was observed in 12 patients (54.5%) and T1 hyperintensities were observed in peripheral (wall) of the giant tuberculoma in 14 patients (63.6%). The mean ADC value of the peripheral (wall) of the giant tuberculoma was 1.034± 0.466[SD] x 10 -3 mm 2 /s and the core was 0.994± 0.455[SD] x 10 -3 mm 2 /s with a statistically significant difference (p-value <0.0005) in between. MR spectroscopy showed raised lipid peak at 0.9 to 1.33 ppm in 10 patients (45.5%),raised lipid-lactate peak in 12 patients (54.5%),raised Choline/Cr ratio more than 1.2 in 14 patients(63.6%) and Choline/Cr ratio less than 1.2 in 5 patients (22.7%). Associated involvement of lung was observed in the 6patients (27.3%), cervical lymph node in 1 patient (4.5%) and spine in 1patient (4.5%). Conclusions: MRI plays a vital role in distinguishing giant intracerebral tuberculomas from other intracranial space-occupying lesions, thereby allows the early institution of anti-tubercular treatment (ATT), decreased patient morbidity, mortality, and prevents unnecessary neurosurgical excision.


Introduction
Tubercular infection of the central nervous system (CNS) is a major public health problem in developing countries like India, with a recent increase in the incidence among immunocompromised patients [1].CNS tuberculosis is the most dangerous manifestation of systemic TB leading to high mortality and morbidity because of its possible serious complications and sequelae [2]. CNS involvement occurs in approximately 5-10% of all cases of tuberculosis [3], and 20% of cases of CNS tuberculosis (TB) are related to acquired immunodeficiency syndrome(AIDS).
Different forms of CNS tuberculosis can be in the form of meningitis, cerebritis, intracranial tuberculomas, miliary tuberculosis, tubercular abscess, calvarial and spinal involvement [4].Of these forms, tubercular meningitis (TBM) followed by intracranial tuberculomas are the most common forms [5]. A tuberculoma consists pathologically of a capsule of collagenous tissue surrounding a central caseation zone [6]. In endemic regions, tuberculomas account for 10% to 30% of intracranial masses [7].
Magnetic resonance imaging (MRI) is superior to computed tomography (CT) for imaging characterization of giant intracerebral tuberculomas. But, there is a substantial overlap in the imaging features with other lesions, such as metastases, gliomas, lymphoma, neurocysticercosis, and fungal granulomas [6,8]. Early differentiation of giant intracranial tuberculoma from other lesions is clinically important because giant intracerebral tuberculomas are potentially curable with prompt antitubercular treatment (ATT) or surgery leading to a better prognosis [9].Therefore newer, advanced MR imaging techniques like MR spectroscopy, MR perfusion, diffusion tensor imaging (DTI), and magnetization transfer (MT) imaging are needed in addition to the conventional MRI sequences to improve the diagnostic precision, thereby precluding the requirement for tissue biopsy [10,11].
To date, only a few case reports and case series of such giant tuberculomas have been reported in the literature. But this giant intracerebral tuberculoma is often misdiagnosed as a brain tumor and taken for surgical evacuation.
The purpose of this study is to analyze the clinical data and magnetic resonance imaging (MRI) findings of such giant intracerebral tuberculomas to improve the diagnostic precision.

Methods and Materials
After approval from the institutional ethics review committee, a hospital-based retrospective study was conducted in a tertiary care hospital of Northeast India from May 2019 to April 2021 with a retrospective review of the clinical and MR imaging data of consecutive 22 patients of giant intracerebral tuberculomas. Informed consent was obtained from patients/guardians before undergoing an MRI scan.

MRI protocols
All patients were subjected to an MRI scan using the Philips Ingenia1.5 Tesla machine (Netherlands). All patients were subjected to an MRI scan of the brain in a supine position and various MRI sequences were obtained and parameters shown in Table 1.
Conventional MRI sequences protocol includes Axial T1WI, T2WI, fluid-attenuated inversion recovery (FLAIR), diffusion-weighted imaging (DWI) and susceptibility-weighted imaging (SWI) sequences followed by sagittal T1WI and coronal T2WI sequences. Post-gadolinium T1WI sequences were obtained in all three planes. Multi-voxel MR spectroscopy was done with TE 144 and TE 35.

Satellite lesion
Relatively smaller nodular or irregular plaque-like lesions near to the edges of a larger giant tubercular lesion were categorized as satellite lesions.

ADC calculation
Apparent diffusion coefficient (ADC) value measurement of the giant intracerebral tuberculoma were done in b value=1000 s/mm2 DWI images. ADC calculation was obtained both in the peripheral (wall) and central portion(core) of the giant intracerebral tuberculoma. The ADC values were measured independently by the two radiologists and the results were obtained from the mean ADC values. Calculation of the mean ADC values was done by placing either round or elliptical ROIs (region of interest). We measured the mean ADC value in the operating system console using 3 uniform sizes of ROIs (with areas maximum 50 mm 2 and minimum 10 mm 2 ) placed in the ADC map image of b=1000sec/mm2, and the mean ADC value was calculated for statistical analysis.
The conventional MRI findings and MR spectroscopy findings were considered as the reference standards in this study for the diagnosis of tuberculoma. Cerebrospinal fluid (CSF) analysis was performed in 19 patients, in 3 patients it was avoided due to fear of intracranial herniation.
Following the provisional MRI imaging diagnosis of tuberculoma, the final diagnosis was confirmed on histopathogical examination in the resected specimens of 4 patients. In the rest of the 18 patients, giant tuberculoma was diagnosed based on a favorable response to the ATT, CSF analysis findings and on associated extra-cranial involvement of tuberculosis. All patients received the standard doses of ATT for a variable interval ranging from 12 months up to 18 months. They were followed up regularly by clinical and serial CT or MRI examinations for clinical and neurological improvement.

Statistical analysis
All statistical analysis was performed using Statistical Package for Social Science (SPSS, version 16). An independent sample student t-test was used to compare the mean ADC values between the cen-

Results
In this study sample of  Abbreviations: ADC=apparent diffusion coefficient, Hypo=hypointense, Hyper=hyperintense, Iso= isointense, NAD=no abnormality detected     CSF analysis showed pleocytosis in 19 patients, lymphocytic predominance in 17 patients (89.5%) and neutrophilic predominance in 2 patients (10.5%). CSF protein was elevated in 16 patients (84.2%) and glucose was low in 15 patients (78.9%). Four patients were treated with surgical decompression followed by medical treatment with ATT and the rest of the 18 patients managed with only medical treatment. Histopathogical examination in the resected specimens of 4 patients showed epitheloid cells admixed with lymphocytes representing tubercular granuloma. In the rest of the 18 patients, giant tuberculoma was diagnosed based on a favorable response to the ATT, CSF analysis findings and on associated extra-cranial involvement of tuberculosis.

Discussion
CNS tuberculosis is associated with a high mortality rate and morbidity due to grave neurological complications and sequelae [5]. Tuberculous meningitis and intracranial tuberculomas are the main forms of CNS tuberculosis [5]. In endemic regions like India, intracranial tuberculomas presenting as space-occupying lesions are not uncommon, where it still accounts for 10% to 30% of all intracranial masses [7,12]. Intracranial tuberculomas arise when small tubercles conglomerate and enlarge in the brain parenchyma [5]. Rupture of these intracranial tuberculomas into the subarachnoid space may sometimes result in tuberculous meningitis [13,14]. Intracranial tuberculomas char-acteristically show granulomatous reaction forming noncaseating granuloma, that later develops caseating necrosis in the central area, which is at first solid, but subsequently liquefy [12].Mostly, typical tuberculomas are oval or rounded in shape, measuring around 2-12 mm in size [15]. Rarely, multiple intracranial tuberculomas may merge and increase in size resulting in giant tuberculoma, presenting clinically with features of mass effect and raised intracranial pressure [6]' Intracranial tuberculomas can develop at any age [12]. However, previous studies in the literature [5] reported the mean age of occurrence of intracranial tuberculomas between 24 and 42 years of age, which was comparable to our series, with a mean age27.4 ± 1.36 years.
Intracranial tuberculomas may present with various clinical symptoms attributable to their size and location. In the previous studies of giant intracranial tuberculomas, the headache was the most common presentation, followed by seizure, features of raised intracranial pressure, and focal neurological deficits, limb weakness [6]. In our study also, the headache was the most common presentation followed by fever and vomiting.
The previous studies on giant tuberculomas reported extracranial tubercular involvement in approximately 21% of cases[ Table 4]. In our series concomitant extracranial tubercular involvement in the form of pulmonary TB was seen in 6 patients (27.3%), cervical lymphadenitis in 1 patient (4.5%) and Pott's spine in another 1 patient (4.5%).
The supratentorial location is most often observed in adults, whereas the infratentorial location is most often observed in children [6,12,16]. In our series also, the majority of the giant tuberculomas were in the supratentorial area, detected in 19 patients (86.4%).
MRI plays a vital role in the diagnosis of intracranial tuberculomas due to its innate specificity and sensitivity in its early detection as compared to CT imaging [16].In MRI the appearance of intracranial tuberculoma depends on the stage of maturation. Noncaseating intracranial tuberculomas appear hypointense on T1WI, hyperintense on T2WI with homogeneous nodular enhancement on the post-contrast study. Tuberculomas with solid central caseation appear as hypo-to isointense on both T1 and T2 weighted images with iso-to hyperintense rim on T2WI and peripheral rim-like post-contrast enhancement. On the other hand, intracranial tuberculomas with central liquefaction appear hypointense on T1WI and hyperintense on T2WI with peripheral hypointense rim which shows irregular rim-like peripheral enhancement on the post-contrast study [6,9,12,14,17,18]. Central region of T2 hypointensity is a valuable finding in tuberculoma as it is not found in several other space-occupying lesions [19].In our series, 12 patients (54.5%)of giant tuberculoma demonstrated a central region of T2 hypointensity.
Intracranial tuberculomas may show different patterns of enhancement on post-contrast study ranging from open rings, complete rings, lobular, or irregular patterns [6]. In our series, the most common patterns of enhancement of giant tuberculoma on post-contrast study were irregular nodular and shaggy rim enhancementsin 13 patients (59.1%)[ Table 3].
The features of conventional MRI are sometimes nonspecific and may overlap with other tumoral and infective lesions leading to diagnostic dilemmas [1,6,8,20]. So additional advanced MR imaging techniques like magnetic resonance spectroscopy (MRS), MR perfusion, diffusion tensor imaging (DTI), and magnetization transfer (MT) imaging are needed to improve diagnostic accuracy.
On magnetisation transfer (MT) T1WI, the cellular content of the noncaseating intracranial tuberculomas appear brighter, which is specific and helps in distinguishing them from other intracranial space-occupying lesions like lymphoma, metastases, and other infective lesions. However, on MT T1WI, these lesions surrounded by a hyperintense rim favors tuberculoma [11,12].
Solid caseating intracranial tuberculomas do not demonstrate diffusion restriction, which helps in distinguishing them from other lesions like lymphoma and medulloblastoma. On the other hand, intracranial tuberculomas with central liquefaction show diffusion restriction on DWI [13]. In our series, irregular nodular to peripheral rim-like diffusion restriction was seen in 9 patients (40.9%), concentric target-appearing diffusion restriction in 6 patients (27.3%) and smooth rim-like diffusion restriction in 2 patients (9.1%) [Table2].
Magnetic resonance spectroscopy (MRS) demonstrates typically elevated lipid peak which helps in distinguishing tuberculomas from other malignant and infective conditions [11,21]. Intracranial tuberculomas can also demonstrate a decrease in N-acetyl aspartate (NAA) and creatinine, with a raised choline/creatinine ratio of >1 [6,22]. NCC can demonstrate elevated lactate, with increased proteins like succinate, glutamate, alanine, glycine, and a decrease in creatinine and NAA [23]. Pyogenic brain abscess demonstrates raised levels of amino-acid peaks. Glioma or metastatic lesions reveal an elevated choline peak significantly with or without lipid/lactate peaks depending on tumor grade or necrosis [6,23].
In our series, the presumptive diagnosis of intracranial giant tuberculomas was based on clinical features and characteristic MR imaging findings. Previous studies [6,20,24] have reported a favorable outcome in most cases of giant tuberculomas following medical treatment alone with ATT for 6 to 24 months [ Table 4]. The present study also showed a favorable outcome in 18 cases of giant tuberculomas following only medical treatment with ATT. Therefore, the empirical use of ATT can be tried after a presumptive diagnosis of giant tuberculoma in endemic regions. However, sometimes giant tuberculoma may need urgent surgical intervention in view of increasing mass effect and neurological deficit.
Positive responsiveness of a giant intracerebral tuberculoma to the anti-tubercular therapy (ATT) can be assessed with reduction of thickness and signal intensity of peripheral T1WI hyperintensity, squeezing and reduction of T2WI hypointense central core of the tuberculoma, increased peripheral wall blooming of micro-haemorrhages on SWI images as compared to the pre-treatment MRI appearances of tuberculoma.
The common differential for rim-enhancing SOL in the cerebellum would be metastasis and pyogenic brain abscess, where dynamic contrast-enhanced MR perfusion and MR spectroscopy play important role in their differentiation. The mean rCBV ratio of peripheral wall of tuberculoma to normallyappearing white matter is around one while mean rCBV ratio is more than 5 in metastasis [28]. MR spectroscopy also plays important role in differentiating neoplasm and pyogenic abscess from tuberculomas. MR spectroscopy reveals larger lipid peak in tuberculoma. Presence of lactate is observed in caseating tuberculomas, however lipid lactate peak is not specific for tuberculoma, and it can be raised in glioma and metastasis [29].

Limitations of the study
Because of small sample in our study, a larger sample size is needed to confirm these MRI findings in the diagnosis of the giant intracerebral tuberculomas, and to differentiate them from other mimics.

Conclusion
Giant tuberculomas are rare lesions but should be considered in the differential diagnosis of intracranial space-occupying lesions in an endemic regions like India. Early diagnosis of giant tuberculomas is clinically important, as they are treatable with ATT, leading to a better prognosis. MR imaging plays a vital role in distinguishing giant tuberculomas from other intracranial space-occupying lesions, thereby precluding the requirement for tissue biopsy. However, sometimes diagnosis can be challenging due to atypical imaging features leading to diagnostic dilemma. An accurate MRI diagnosis allows the early institution of ATT, decreased patient morbidity, mortality, and prevents unnecessary neurosurgical excision.