CT Perfusion Dynamics of Intracranial Tuberculomas

Tuberculomas constitute about 5-40% of all intracranial space occupying lesions [1]. Neuroimaging techniques like Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) play important roles in the diagnosis of central nervous system tuberculosis including tuberculoma. Routine imaging features of tuberculoma are non specific [2-4].

Advanced MR imaging techniques like diffusion weighted imaging, MR spectroscopy, Magnetization Transfer imaging and MR perfusion imaging have improved the sensitivity and specificity in the diagnosis of intracranial tuberculoma [5-8].

Earlier investigation using MR perfusion technique in tuberculoma revealed altered perfusion parameters. However, the lesions were presumed to be of tubercular aetiology in lieu of histological verification of the lesions [8]. Recent MR perfusion studies are promising in differentiating intracranial ring enhancing tuberculomas from ring enhancing metastases [9,10].

CT Perfusion (CTP) is an established method in the evaluation of hyper acute and acute ischemic strokes of brain. The results of CTP observations are comparable to those of MR perfusion and diffusion images [11,12]. The present study uses CT perfusion technique in histologically proven tuberculomata to understand the pathophysiology and its therapeutic implications.

After obtaining permission from the institute (NIMHANS, Bangalore), a prospective study was conducted in the year 2005 and 2006 in the Department of Neuroimaging and interventional radiology with following design:

Inclusion Criteria

1. The treatment naïve tuberculomas selected after CT guided stereo-tactic biopsy.

2. Patients on antitubercular treatment after CT guided stereo-tactic biopsy.

Exclusion Criteria

1. Paediatric patients were excluded from the study.

Informed consent was obtained in all patients

There were 19 patients in the study; nine patients were not on antitubercular treatment and ten were on antitubercular treatment at the time of study. In ten patients who were on antitubercular treatment at time of study six were on treatment for less than two months and four were more than six months. There were 13 females and six males in the age group of 17-56 y (Mean 32.93). All patients underwent plain and contrast enhanced CT scan of brain. In case of multiple tuberculomas, largest supratentorial lesion was selected for CTP evaluation.

CT Perfusion Imaging Technique and Data processing

Patients were advised fasting for 6 h prior to the CT perfusion study. An 18G cannula was introduced into the antecubital vein. Once the position of cannula was confirmed in situ, a non-enhanced CT of brain was performed to identify the axial sections demonstrating the lesion. Subsequently dual-slice dynamic CT bolus tracking perfusion studies using a multi-detector CT scanner was performed. The first slice was selected at the level of the maximum cross-sectional diameter of the mass lesion as seen on initial plain CT and the second slice was 10mm cranial to the first. The slice thickness was 10mm with a matrix size of 512 x 512. Fifty ml of 65% iodinated contrast agent was injected through an 18G cannula into the antecubital vein using a pressure injector at the rate of 6ml / second. Following a delay time of 4 seconds, a total of 40 images were acquired at a rate of one image / sec.

The image data was transferred to a workstation with Siemens MV 300 post processing software. Perfusion images were created with reference to superior sagittal sinus away from the bone. Irregular shape regions of interest (ROI) were drawn in the following regions: a) entire lesion, b) capsule of the lesion c) center of the lesion and d) within 1-2cm adjacent to the lesion. The CTP parameters from each of these ROIs were compared with perfusion parameters from ROIs in contra lateral normal-looking brain parenchyma and cerebral haemodynamic parameters such as relative cerebral blood flow (rCBF), relative cerebral blood volume (rCBV) and relative time to peak (rTTP) were obtained using deconvolution method. Colour maps also generated as shown in [Table/Fig-1a,b,2a,b].

Mean and standard deviations of these parameters were calculated in all patients. Paired Student t-test was performed to obtain statistical significance between two groups of patients. A p-value of less than 0.05 was considered statistically significant.

The CT imaging features of these tuberculomas are given in [Table/Fig-3].CT perfusion parameters are given in [Table/Fig-4,5]. [Table/Fig-6] shows mean and standard deviation of perfusion parameters of both the groups.

Contrast enhanced CT scan of brain showed single lesion in 14 patients and multiple lesions in five patients. The lesions in all the patients were isodense on plain CT and showed ring-enhancement on contrast studies. CTP parameters like rCBF and rCBV were significantly low in all the three ROIs in the group not on treatment compared to that of on treatment ; rCBF and rCBV of entire lesion (p=0.018 and p=0.005 respectively), capsule (p=0.045 and p=0.010 respectively) and center of the lesion (p=0.020 and p=0.009) respectively). Tuberculomas on antitubercular treatment for more than six months showed reduced rCBF and rCBV in entire lesion (p=0.01 & p=0.01 respectively), capsule (p=0.04 & p=0.03 respectively) and center (p=0.08 & p=0.05 respectively) compared to those on treatment for less than two months. Similarly tuberculomas on treatment for six months did not show significant difference in rCBF and rCBV compared to tuberculomas who were not on treatment. Compared to tuberculomas not on treatment, tuberculomas on treatment (for less than two months) showed statistically increased rCBF and rCBV in entire lesion (p=0.01 & p=0.04 respectively), capsule (p=0.03 & p=0.01 respectively) and center (p= 0.03 &=0.01). [Table/Fig-7,8] shows scatter and bar diagram of tuberculomas in entire lesion and capsule of the lesion.

CNS involvement in tuberculosis is common. The usual manifestations include acute and chronic meningitis, granulomata, encephalopathy and abscess. Tuberculoma is the focal granulomatous form, encountered in about 10%-40% of CNS tuberculosis [10]. Histologically, the mature tuberculomas are composed of necrotic caseous center surrounded by a capsule that contains fibroblasts, epitheloid cells, Langhans cells and lymphocytes formed by coalescence of microtuberculomas. In the formation of tuberculoma, initially there is predominant inflammatory reaction with collagen poor capsule and later collagen rich capsule forms with reduction in inflammatory reaction and caseated center due to coagulation necrosis [13].

Three layers are described in tuberculomas. The inner layer is central caseation necrosis with nuclear debris and adjacent inflammatory cells. Hyalinised collagenous tissue forms middle layer and outer layer is of inflammatory cells like lymphocytes and epitheloid [13].

Improved understanding of pathophysiology of a metabolically dynamic lesion such as a tuberculoma may help in better clinical management. Various imaging techniques such as PET, SPECT and MR perfusion have been used in an effort to understand and measure various physiological parameters in many neurological diseases [14-20].

Perfusion imaging is novel method to detect vasculature of tissue and can detect indirectly the metabolic activity of the tissue as perfusion of the tissue changes according to metabolic requirement of the tissue [21].

There are only a few pathophysiological studies in CNS tuberculosis using SPECT and MR perfusion imaging techniques [8-10,22].

However, PET is not available at all centers, uses isotopes and lacks resolution. MR is not cost effective and time-consuming. Considering these drawbacks, we performed CT perfusion in our patients.

In an earlier MR perfusion study of tuberculomas, Batra et al., observed increased perfusion similar to that of high grade tumours [8]. Jinkins et al., used CTP and reported changes in time density curve in intracranial tuberculoma with the treatment. They found early appearance of contrast in the lesion compared to normal brain parenchyma. This feature was attributed to the presence of inflammatory granulation tissue .The early wash-in phase was observed even during treatment which the authors explained by the presence of hypervascularity due to neovascularization and loss of auto regulation in the wall of lesion [23]. Recently MR perfusion studies in differentiating tuberculomas from metastasis have shown a cut-off rCBV value of ≥3.745 for differentiating ring-enhancing metastases from ring-enhancing tuberculomas [9]. Similar studies in lung lesions have shown significantly high peak height and increased relative flow (RF) in active tuberculoma as compared to inactive tuberculoma [24].

Infective lesions like abscess and toxoplasmosis have reduced perfusion compared to neoplastic lesions as shown by various other studies [25,26]. Thus it becomes clear from all these studies that infective lesions can be differentiated from neoplastic lesions based on perfusion studies.

In the present study we found that CT perfusion parameters were significantly low in treatment naive tuberculomas compared to those on treatment. In the treated group those tuberculomas which were on treatment for more than six months showed low perfusion than those who received treatment for less than two months. However, there was no statistically different perfusion in tuberculomas on treatment for more than six months as compared to treatment naive tubercuomas.

We observed reduced rCBF and rCBV in the capsule and the entire lesion compared to normal brain parenchyma in TWT group of patients. The reduction may be due to the presence of vasculitis in the tubercular granulomatous lesions. Dastur et al., in their important work have demonstrated total or partial vascular occlusion in the lesions. These changes were also noted in the reactive and surrounding edematous zones of tuberculomatous lesions. The core of tuberculomas contains caseation which showed no blood vessels or only ghost of blood vessels [27]. These pathological features explain findings of reduced rCBF and rCBV in our lesions.

On the other hand tuberculomas in TOT (<2 months) patients showed increased rCBF and rCBV parameters in all the three ROIs when compared to normal looking brain parenchyma. This increase may be due to loss of auto regulation in the new vessels developed in the lesion [23].

Jinkins et al., have described changes in micro and macro vasculature of various infective and neoplastic intracranial lesions. They observed that the universal response of cerebral tissue to any injury was coagulation necrosis following which there was reactive granulation repair resulting in reactive hypervascularity with loss of auto regulation [28]. In advanced cases there could be vascular shunting and early venous drainage. In 4-14 days, some of the hypervascularity can be accounted by a true neovascularization of the capsule. Damaged vessels and newly formed vessels in granulation tissue lack tight junction which led to leak of intravascular fluid and contrast into capsule leading to ring enhancement and oedema. These were demonstrated as elevated plateau of time density curve and late wash out phase [28]. These findings are also validated by the fact that there is increased neoangiogenesis with increased expression of vascular endothelial growth factor in the wall of tuberculomas leading to increased rCBV [29].

However, two patients (Cases 14 & 15) of TOT (<2 months) did not show increase in rCBF and rCBV parameters. This may be due to immunologically mediated secondary granulomatous vasculitis (with? vessel occlusion) [30].

The reasons for significant reduction of rCBF and rCBV parameters in patients of TOT (>6 months) compared to patients of TOT (<2 months) may include healing by fibrosis, reactive gliosis and calcification [27].

The tuberculomas showed a corresponding decrease in size in this interval period. The treatment has started the healing process resulting in reduced rCBF and rCBV in our patients. These were in accordance with the study by Jinkins et al., who found an initial hyper vascular phase which gradually returned to isovascular phase on treatment or a hypo vascular phase, attributed to atrophy or atrophy associated with calcification [23].

Similarly with treatment there is decreased neoangiogenesis and cellularity causing decrease in rCBV [29].

CT perfusion as a non-invasive technique uses an intravascular tracer to measure perfusion parameters whereas the traditional Xenon CT and PET measure perfusion by using diffusion tracers. While the two techniques may reflect different physiological mechanisms and the accuracy of flow values obtained by CT perfusion are not fully validated [31], studies have shown that quantitative results of perfusion CT are in agreement with xenon CT [32] and MR technique [33].

We did not use corrected perfusion values as disruption of BBB in enhancing lesions may not adversely affect the rCBF measurement [34]. CT perfusion involves radiation and the overall effective dose required for dynamic CT is 2.0-3.4mSv which is marginally higher than that required for routine head CT 1.5-2.5mSv [35].

The study group was small and the choice to study after six months of treatment was arbitrary. Jinkins et al., have suggested that the response of tuberculoma to treatment takes eight weeks [23]. Thus the choice of two months was based on this observation. A serial longitudinal study of tuberculomas with treatment till its resolution would perhaps help us to better understand the haemodynamics of tuberculomas and the changes thereof induced by treatment. Further, perfusion parameters would help in identification of tuberculomas which would either respond or resist treatment and also differentiate it from other mimicking lesions.

[Table/Fig-1a,b]:

rCBF (A)color map of tuberculoma without treatment in right thalamus showing low perfusion with corresponding contrast CT image(B)

[Table/Fig-2a,b]:

rCBF (A) colour map of tuberculoma on treatment in left thalamus showing high perfusion with corresponding contrast CT image (B)

[Table/Fig-7a,b]:

Scatter plot of rCBV (A) and rCBF (B)of entire lesion showing low perfusion in TWT and TOT

[Table/Fig-8a,b]:

Bar diagram of rCBF (A) and rCBV (B) at capsule of tuberculomas without treatment (TWT) and on treatment (TOT)

The findings in this study are 1) Intracranial tuberculomas not on treatment and those on treatment for around six months show low perfusion. 2) Tuberculomas on treatment for less than two months show high perfusion. While the temporal and morphological resolution of tuberculomas on treatment have traditionally been the markers of response to treatment, the findings of this study suggest that serial perfusion profiles of tuberculomas on treatment could possibly be seen as surrogate markers of response to treatment. Larger longitudinal studies could address the role of CT perfusion as a non-invasive technique to monitor response to treatment of intracranial tuberculomas.

[1]. GV Shah, Central nervous system tuberculosis: imaging manifestations Neuroimaging Clin N Am 2000 10(2):355-74.  [Google Scholar]

[2]. S Selvapandian, V Rajshekhar, MJ Chandy, J Idikula, Predictive Value of Computed Tomography- based diagnosis of intracranial Tuberculoms Neurosurgery 1994 35(5):845-50.  [Google Scholar]

[3]. AB Grayeli, A Redondo, J Salama, A Rey, Tuberculoma of Cavernous sinus: A case report Neurosurgery 1998 42(1):179-81.  [Google Scholar]

[4]. OH Del Brutto, A Mosquera, Brainstem tuberculoma mimicking glioma: the role of antituberculous drugs as a diagnostic tool Neurology 1999 52(1):210-11.  [Google Scholar]

[5]. OK Basoglu, R Savas, O Kitis, Conventional and diffusion-weighted MR imaging of intracranial tuberculomas: A case report Acta Radiol 2002 43(6):560-62.  [Google Scholar]

[6]. RK Gupta, DK Vatsal, N Husain, S Chawla, KN Prasad, R Roy, Differentiation of tuberculous from pyogenic brain abscesses with in vivo proton MR spectroscopy and magnetization transfer MR imaging AM J Neuroradiol 2001 22(8):1503-09.  [Google Scholar]

[7]. MH Pui, MN Ahmad, Magnetization transfer imaging diagnosis of intracranial tuberculomas. Neuroradiology 2002 44(3):210-15.  [Google Scholar]

[8]. A Batra, RP Tripathi, Perfusion magnetic resonance imaging in intracerebral parenchymal tuberculosis Preliminary findings. J Comput Assist Tomogr 2003 27(6):882-88.  [Google Scholar]

[9]. S Sankhe, A Baheti, A Ihare, S Mathur, P Dabhade, A Sarode, Perfusion magnetic resonance imaging characteristics of intracerebral tuberculomas and its role in differentiating tuberculomas from metastases Acta Radiol 2013 54(3):307-12.  [Google Scholar]

[10]. S Chatterjee, J Saini, C Kesavadas, HR Arvinda, M Jolappara, AK Gupta, Differentiation of tubercular infection and metastasis presenting as ring enhancing lesion by diffusion and perfusion magnetic resonance imaging Journal of Neuroradiology 2010 37:167-71.  [Google Scholar]

[11]. DG Na, JW Ryoo, KH Lee, CH Moon, CA Yi, EY Kim, Multiphasic perfusion computed tomography in hyperacute ischemic stroke: Comparison with diffusion and perfusion magnetic resonance imaging J Comput Assit Tomogr 2003 27(2):194-206.  [Google Scholar]

[12]. M Konig, R Banach-Planchamp, M Kraus, E Klotz, A Falk, W Gehlen, CT perfusion imaging in acute ischemic cerebral infarct: Comparison of cerebral perfusion maps and conventional CT findings Rofo 2000 172(3):219-26.  [Google Scholar]

[13]. TK Kim, KH Chang, CJ Kim, JM Goo, MC Kook, MH Han, Intracranial tuberculoma: Comparison of MR with pathologic findings Am J Neuroradiol 1995 16(9):1903-08.  [Google Scholar]

[14]. EE Kim, SK Chung, TP Haynie, CG Kim, BJ Cho, DA Podoloff, Differentiation of residual or recurrent tumors from post-treatment changes with F-18 FDG PET RadioGraphics 1992 12(2):269-79.  [Google Scholar]

[15]. PE Ricci, JP Karis, JE Heiserman, EK Fram, AN Bice, BP Drayer, Differentiating recurrent tumor from radiation necrosis: time for re-evaluation of positron emission tomography? Am J Neuroradiol 1998 19(3):407-13.  [Google Scholar]

[16]. VW Lee, V Antonacci, S Tilak, JD Fuller, TP Cooley, Intracranial Mass Lesions:Sequential Thallium and Gallium Scintigraphy in Patients with AIDS Radiology 1999 211(2):507-12.  [Google Scholar]

[17]. M Lorberboym, L Estok, J Machac, I Germano, M Sacher, R Feldman, Rapid differential diagnosis of cerebral toxoplasmosis and primary central nervous system lymphoma by thallium-201 SPECT J Nuc Med 1996 37(7):1150-54.  [Google Scholar]

[18]. S Cha, EA Knopp, G Johnson, SG Wetzel, AW Litt, D Zagzag, Intracranial Mass Lesions: Dynamic Contrast-enhanced Susceptibility-weighted Echo-planar Perfusion MR Imaging Radiology 2002 223(1):11-29.  [Google Scholar]

[19]. S Cha, S Pierce, EA Knopp, G Johnson, C Yang, A Ton, Dynamic contrastenhanced T2*-weighted MR imaging of tumefactive demyelinating lesions Am J Neuroradiol 2001 22(6):1109-16.  [Google Scholar]

[20]. SL Keir, JM Wardlaw, Systematic review of diffusion and perfusion imaging in acute ischemic stroke Stroke 2000 31(11):2723-31.  [Google Scholar]

[21]. WA Scott, Magnetic resonance imaging of brain and spine13rd Edition:215  [Google Scholar]

[22]. UK Misra, J Kalita, BK Das, Single photon emission computed tomography in tuberculous meningitis Post grad med J 2000 76(900):642-45.  [Google Scholar]

[23]. JR Jinkins, MZ Al-Kawi, R Bashir, Dynamic computed tomography of cerebral parenchymal tuberculomata Neuroradiology 1987 29(6):523-29.  [Google Scholar]

[24]. U Tateishi, M Kusumoto, Y Akiyama, F Kishi, M Nishimura, N Moriyama, Role of contrast-enhanced dynamic CT in the diagnosis of active tuberculoma Chest 2002 122(4):1280-84.  [Google Scholar]

[25]. TM Ernst, U Chang, MD Witt, HA Aronow, ME Cornfold, I Walot, Cerebral toxoplasmosis and lymphoma in AIDS:Perfusion MR imaging experience in 13 patients Radiology 1998 208(3):663-69.  [Google Scholar]

[26]. A Batra, RP Tipathi, SP Gorthi, Magnetic resonance evaluation of cerebral toxoplasmosis in patients with the acquired immunodeficiency syndrome Acta Radiol 2004 45(2):212-21.  [Google Scholar]

[27]. DK Dastur, UP Dave, Ultrastructural basis of the vasculopathy in and around brain tuberculomas. Possible significance of altered basement membrane Am J Pathol 1977 89(1):35-50.  [Google Scholar]

[28]. JR Jinkins, Dynamic CT of micro- and macroangiopathic states of the cerebrum Neuroradiology 1988 30(1):22-30.  [Google Scholar]

[29]. RK Gupta, M Haris, N Husain, KN Prasad, M Pauliah, C Srivastava, Relative cerebral blood volume is a measure of angiogenesis in brain tuberculoma and its therapeutic implications Proc Intl Soc Mag Reson Med 2006 14:826  [Google Scholar]

[30]. N Hejazi, W Hassler, Multiple intracranial tuberculomas with atypical response to tuberculostatic chemotherapy Acta Neurochir(Wien) 1997 139(3):194-202.  [Google Scholar]

[31]. HC Roberts, TP Roberts, WP Dillon, CT perfusion flow assessment: “up and coming” or “off and running”? Am J Neuroradiol 2001 22(6):1018-19.  [Google Scholar]

[32]. M Wintermark, JP Thiran, P Maeder, P Schnyder, R Meuli, Simultaneous measurement of regional cerebral blood flow by perfusion CT and stable xenon CT: A validation study Am J Neuroradiol 2001 22(5):905-14.  [Google Scholar]

[33]. M Wintermark, M Reichhart, O Cuisenaire, P Maeder, JP Thiran, JP Schnyder, Comparison of Admission Perfusion Computed Tomography and Qualitative Diffusion- and Perfusion-Weighted Magnetic Resonance Imaging in Acute Stroke Patients Stroke 2002 33:2025-31.  [Google Scholar]

[34]. GT Gobbel, CE Cann, HS Iwamoto, JR Fike, Measurement of regional cerebral blood flow in the dog using ultrafast Computed tomography experimental validation Stroke 1991 22(6):772-79.  [Google Scholar]

[35]. EG Hoeffner, I Case, R Jain, SK Gujar, GV Shah, JP Deveikeis, Cerebral perfusion CT- technique and clinical applications Radiology 2004 231(3):632-44.  [Google Scholar]