|Year : 2020 | Volume
| Issue : 1 | Page : 326-331
Role of magnetic resonance imaging in diagnosis of pediatric posterior fossa tumors
Zeinab A.E. Ali1, Rehab M Habib1, Shaimaa A.E. Fotoh2
1 Department of Radiodiagnosis, Faculty of Medicine, Menoufia University, Menoufia, Egypt
2 Department of Radiodiagnosis, Tala General Hospital, Menoufia Governorate, Menoufia, Egypt
|Date of Submission||12-Aug-2018|
|Date of Decision||01-Oct-2018|
|Date of Acceptance||09-Oct-2018|
|Date of Web Publication||25-Mar-2020|
Shaimaa A.E. Fotoh
Berket El Sabae City, Menoufia Governorate
Source of Support: None, Conflict of Interest: None
The aim was to study the role of MRI in diagnosis of pediatric posterior fossa tumors.
Brain tumors represent the most common solid neoplasms in children and the second most common pediatric malignancy following leukemia.
Patients and methods
A retrospective study was conducted on 25 pediatric patients with known brain posterior fossa masses, who ranged in age from 1 to 15 years and had suspected posterior fossa tumor by computed tomography. All participants attended the MR unit of Radiology Department, Menoufia University, Menoufia Governorate, during the period from October 2016 to November 2017. Complete history taking, clinical examination, and MR data were compared with histopathology obtained from all patients.
The mean age of the cases was 7.2 ± 3.9 years. The most common tumors were medulloblastoma (MB) and pilocytic astrocytoma. MB showed restricted diffusion with low apparent diffusion coefficient value, and pilocytic astrocytoma showed free diffusion with higher apparent diffusion coefficient value. High-grade tumors such as MB and brain stem glioma showed elevated choline and reduced N-acetyl aspartate in magnetic resonance spectroscopy curve. Low-grade tumors such as low-grade glioma showed mild rise of choline and mild reduction of creatine in magnetic resonance spectroscopy curve.
Conventional MRI provided limited information regarding tumor type and grade, falling short as a definitive diagnostic examination. Advanced brain MRI techniques provided incremental diagnostic value over conventional MRI. However, no single advanced technique was perfect, but different techniques typically complement one another.
Keywords: brain tumors, diagnosis, magnetic resonance imaging, posterior fossa tumors
|How to cite this article:|
Ali ZA, Habib RM, Fotoh SA. Role of magnetic resonance imaging in diagnosis of pediatric posterior fossa tumors. Menoufia Med J 2020;33:326-31
|How to cite this URL:|
Ali ZA, Habib RM, Fotoh SA. Role of magnetic resonance imaging in diagnosis of pediatric posterior fossa tumors. Menoufia Med J [serial online] 2020 [cited 2020 Aug 15];33:326-31. Available from: http://www.mmj.eg.net/text.asp?2020/33/1/326/281281
| Introduction|| |
Brain tumors represent the most common solid neoplasms in children and the second most common pediatric malignancy following leukemia. Approximately 65% of all brain pediatric tumors arise in the posterior fossa. The commonest tumors are pilocytic astrocytoma (commonest), medulloblastoma (MB), ependymoma, brain stem glioma, atypical teratoid rhabdoid tumor, hemangioblastoma (uncommon except with Von Hippel-Lindau syndrome (VHL)), and teratoma (in infant). Children with posterior fossa brain tumors present with signs and symptoms of increase intracranial pressure, gait disturbances, and/or cranial nerve deficits, depending on type, size, and location of the tumor. Earlier diagnosis of brain tumors in children and young adults improves long-term outcome. Conventional MRI is an essential tool for diagnosis and evaluation of location, quality, and extent of posterior fossa tumors, and also for the follow-up, but it offers limited information regarding grade and type. Diffusion-weighted imaging (DWI) and apparent diffusion coefficient (ADC) map increase both sensitivity and specificity of MRI and provide information about cellularity of the tumor and have an important role in preoperative differentiation of different tumor types. ADC value is a simple and available technique for the evaluation of pediatric cerebellar neoplasm and accurately differentiates between the most common tumor types. ADC value is irreversibly proportionate to the tumor cellularity (primary brain tumor with higher cellularity or higher grade have lower ADC value). Proton magnetic resonance spectroscopy (MRS) is a noninvasive in-vivo technique that provides additional metabolic diagnostic indices and has been used widely in the imaging of brain tumors, as it can help differentiation of common pediatric brain tumors. Therefore, correlation of conventional and advanced MRI techniques may improve the specific diagnosis of brain tumors in posterior fossa in children.
| Patients and Methods|| |
A retrospective study was conducted on 25 pediatric patients with known brain posterior fossa masses, who ranged in age from 1 to 15 years and had suspected posterior fossa tumor by computed tomography. All participants attended the MR unit of Radiology Department, Menoufia University, Menoufia Governorate, during the period from October 2016 to November 2017.
Approval of the study protocol was obtained from the Ethical Scientific Committee of Menoufia University, and an informed consent was taken from all patients or from the members of their families before their enrollment in the study.
The inclusion criteria were as follows: pediatric age group (1 day to 18 years), patients with clinical findings suggestive of brain neoplasm, and patients with radiological findings (computed tomography) suggestive of posterior fossa neoplasm.
Patients older than 18 years and refusal of parental consent were the exclusion criteria.
For all pediatric patients, the following procedures were performed: careful history taking regarding headache, neck pain, drowsiness, nausea, vomiting, uncoordinated walk, and vision problems and full clinical examination. Patient preparations, stressing on any metallic (non-titanium) body implants, such as cardiac pacemakers, brain aneurysm clips, cochlear implants, and vascular stents, are considered an absolute contraindication to the performance of the procedure. Detachable metallic implants like teeth prosthesis are considered as proportional contraindication and should be removed before entrance to the magnetic area. Sedatives given is chloral hydrate 50 mg/kg by mouth.
Conventional MRI was performed on a 1.5-tesla superconducting MR imager (Toshiba 1.5-tesla MRI system Vantage Titan; Toshiba, Tokyo, Japan) with head coil using the following protocol: axial T1-weighted noncontrast images, with imaging parameters of 405/12 repetition time/echo time (TR/TE), a slice thickness of 4 mm, a slice gap of 1 mm, a flip angle of 90/180, and a matrix of 192 × 352; axial and coronal T2-weighted images, with imaging parameters of 4300/100 (TR/TE), a slice thickness of 4 mm, an interslice gap of 1 mm, a flip angle of 90/160, and a matrix of 129 × 448; axial fluid-attenuated inversion recovery, with imaging parameters of 7200/100 (TR/TE), a slice thickness of 4 mm, an interslice gap of 1 mm, a flip angle of 90/160, and a matrix of 304 × 304; and postcontrast-enhanced axial, coronal, and sagittal T1-weighted images. The lesions detected on contrast-enhanced MRI were evaluated regarding the following: number of lesions, signal behavior, heterogeneousity (presence of intratumoral hemorrhagic or necrotic components), peritumoral edema, mass effect, enhancement degree, and pattern.
DWIs were acquired by using a single-shot T2-weighted echo-planar spin-echo sequence with imaging parameters of 6800/120 (TR/TE), a slice thickness of 4 mm an interslice gap of 1 mm, a matrix of 128 × 128, a flip angle of 90/180, and echo-planar imaging factor of 120. The diffusion gradient encoding was done in three (x, y, and z) orthogonal directions. The DWIs were acquired with b values of 0, 0, 1000 s/mm, 0, 1000, 0 and 1000, 0, 0 s/mm; isotope 1000; and calculated ADC maps. The ADC maps were calculated automatically by MRI software and included in the sequences. Measurements of ADC were made in regions of interest (enhancing solid portion of the lesion after it was identified on postcontrast axial, coronal, and sagittal T1 WIs and in the normal brain tissue). The ADC values were expressed in 10−3 mm2/s. All DW imaging data were transferred to a computer workstation for determination of the signal intensity and ADC.
Magnetic resonance spectroscopy
MRS was obtained by the single-voxel technique with a volume of interest of 4.5 cm3 (20 × 15 × 15 mm3), and the spectra were acquired by using a spin-echo sequence (point-resolved spectroscopy) with short and long TEs of 30 and 135 ms, respectively. This volume of interest was placed within the solid portion of the tumorous lesion, avoiding necrotic areas, calcifications, cysts, or hemorrhagic foci. Sequences were performed with water saturation.
Histopathology was obtained from all patients and compared with our MR data.
Data were collected, tabulated, and statistically analyzed using an IBM personal computer with statistical package for the social sciences (SPSS; SPSS Inc., Chicago, Illinois, USA) version 20 and Epi Info 2000 programs (Centers for Disease Control and Prevention (CDC), Atlanta, Georgia, US), where the following statistics were applied: descriptive statistics, in which quantitative data were presented in the form of mean, SD, and range, and qualitative data were presented in the form of numbers and percentages.
| Results|| |
The pool of our study was 25 patients (nine were females and 16 were males) with posterior fossa brain tumors. Conventional MRI and MR diffusion studies were performed in an attempt to diagnose preoperative tumor. Tumors were all located in the posterior fossa. Histopathological types of the tumors studied included the following: pilocytic astrocytoma, [Figure 1] MB classic type, [Figure 2] ependymoma, brain stem glioma, low-grade glioma, and ganglioglioma. Patients' age ranged from 1 to 15 years. The mean age of the cases was 7.2 ± 3.9 years. The numbers and percentage of frequency of each tumor in our study group are shown in [Table 1]. The most common tumors were MB and pilocytic astrocytoma [Table 1]. Moreover, MB shows restricted diffusion, whereas pilocytic astrocytoma shows free diffusion [Table 2].
|Table 1: Distribution of the studied patients regarding sex, age, mean age for each disease, and percentage of frequency of each tumor|
Click here to view
|Figure 1: A 5-year-old female child with cerebellar pilocytic astrocytoma. (a) Axial contrast-enhanced T1-weight image and (b) apparent diffusion coefficient map image.|
Click here to view
|Figure 2: A 3-year-old female child with medulloblastoma. (a) Axial T2-weighted image. (b) Single-voxel magnetic resonance spectroscopy shows exceedingly high choline/creatine ratio and reduced N-acetyl aspartate strongly, suggesting high-grade tumor.|
Click here to view
Nine cases had calculated ADC values. Higher values were shown with low-grade glioma and pilocytic astrocytoma (low-grade tumors) and lower values with MB (high-grade tumor). Free diffusion (low-grade tumor) had high ADC value, and restricted diffusion (high-grade tumor) had low ADC value. ADC value for pilocytic astrocytoma was higher than that of MB and ependymoma and that of ependymoma was higher than that of MB. MRS was done for three cases; high-grade tumors such as MB and brain stem glioma showed elevated choline and reduced N-acetyl aspartate (NAA) in MRS curve. Low-grade tumors such as low-grade glioma show mild rise of choline and mild reduction of creatine in MRS curve [Table 3]. Based on our study inclusion of diffusion imaging, ADC value and MRS in imaging protocol of pediatric posterior fossa tumor imaging had a significant value, with sensitivity frequency percentage of 60% [Table 4].
|Table 3: Calculated apparent diffusion coefficient values for some cases and its relation to diffusion and pathology result as well as magnetic resonance spectroscopy curve result in relation to tumor grade and pathology result|
Click here to view
|Table 4: The number and sensitivity frequency percentage of positive and negative cases|
Click here to view
| Discussion|| |
This study included 25 patients with posterior fossa tumors. Age ranged between 1 and 15 years. The mean age of the cases was 7.2 years. Rumboldt et al. included 31 patients with neuroglial tumors, and the mean age of the patients was 10 years (range: 6 months to 16 years). The study group in this work contained mostly common posterior fossa tumors in children: 10 (40%) cases of MB, four (20%) cases of juvenile pilocytic astrocytoma (JPA), and one (4%) case of ependymoma. It also included other posterior fossa tumors such as brain stem glioma (16%), ganglioglioma (8%), and low-grade glioma (12%). The study by Zitouni et al. contained 42 pediatric patients, where 14 lesions were JPA, 10 were ependymoma, and 18 were MB.
In our study, the results showed that the most common pediatric posterior fossa tumors are MB and JPA. Rumboldt et al. in their study showed that the most common pediatric posterior fossa tumors were JPA and MB. Poretti et al. reported that MB was the most frequent malignant brain tumor of childhood. Cerebellar astrocytomas were the most frequent posterior fossa tumors in children.
In our study, MB was hyperintense in DW image (restricted diffusion) and pilocytic astrocytoma was hypointense in DW image (free diffusion). ADC values were different between MB and low-grade astrocytoma. It was lower for MB than pilocytic astrocytoma. Moreover, ADC value for pilocytic astrocytoma was higher than that of MB and ependymoma and that of ependymoma was higher than that of MB. Lower ADC values were seen with restricted diffusion and higher values coincided with free diffusion. So, DWI with ADC value can differentiate and predict most common pediatric posterior fossa tumors, regarding type and grade, as high-grade gliomas were hyperintense on DW images with decreased ADC values. These results coincided with that of Poretti et al.. Diffusion was restricted (high intensity on DWI, low ADC values) in densely packed cerebral tumors with high cellularity, small extracellular space, and high nuclear-to-cytoplasmic ratio. High-grade tumors were typically characterized by higher cellularity and larger nuclear area than low-grade tumors. Restricted diffusion has been demonstrated in high-grade brain tumors with high cellularity. Moreover, Koral et al. showed that inclusion of diffusion MRI data in the imaging protocol significantly improved the accuracy of preoperative diagnosis of common pediatric cerebellar tumors. The improvement in results increased correct diagnoses of Pas and MBs. JPA consists of low-density cells and Rosenthal fibrils; the extracellular space is wider, and the cellularity is less than in other posterior fossa tumors. Thus, they cause less restricted diffusion compared with extracellular compartments. MBs were composed of high-density cells having high nucleus/cytoplasm ratio, thus having less extracellular matrix. As a result, increased diffusion restriction was observed. On the contrary, ependymoma has intermediate cellularity. Moreover, in the same study, low ADC value indicated restricted diffusion, whereas a high ADC value indicated increased diffusion. Zitouni et al. found that preoperative ADC values could differentiate the main histological subtypes of pediatric posterior fossa tumors with high sensitivity and specificity in a study composed of 42 pediatric patients (mean age: 7.76 ± 4.58 years) with intra-axial tumors in the infratentorial region. ADC ratio thresholds of less than or equal to 1.18, 1.18–1.7, and greater than or equal to 1.7 allowed preoperative diagnosis of MB, ependymoma, and JPA with high specificity and sensitivity. Rumboldt et al. have reported that ADC values and ratios were simple and readily available techniques for evaluation of pediatric cerebellar neoplasms that may accurately differentiate the two most common tumors: JPA and MB. Proposed cut-off values of greater than 1.4 × 10−3 mm2/s for JPA and less than 0.9 × 10−3 mm2/s for MB seem to reliably provide the diagnosis. Ependymomas are also significantly different from other tumor types and, in most of cases, show ADC values of 1.00–1.30 × 10−3 mm2/s. Our results disagree with Jaremko et al., who in their study showed that contrary to some recent reports, there was true overlap between diffusion characteristics of JPA, ependymoma, and MB. Moreover, ADC overlap could not be accounted solely for technical difficulties in ADC measurement (small, hemorrhagic, or calcific tumors) but included variation in tumor pathology (desmoplastic MB, anaplastic ependymoma, and restricted diffusion in a JPA tumor nodule).
In our result also MB and brain stem glioma showed elevated choline-reduced NAA, which was a characteristic of high-grade tumor. Low-grade glioma showed mild rise of choline and mild reduction of creatine which was a characteristic that lead us to expect that we have a low-grade tumor. Zitouni et al. in his study shows that MR spectral patterns with elevated choline and lipid levels in the absence of NAA were histologically verified to represent regions of active tumor with extensive areas of necrosis. Such MR spectral patterns contribute to high-grade tumor. Astrakas et al. reported that high-grade tumors, which are highly acellularized and have a high proliferative rate, show increased Cho levels in comparison with normal brain tissue. In a pediatric population, high Cho levels seemed to correlate with tumor progression and a faster growth. Vicente et al. reported that there were significant differences in metabolite concentrations of pilocytic astrocytomas, ependymoma, and MB. Cho was higher in MB and ependymoma compared with pilocytic astrocytoma, which is in agreement with Cho as an indicator of cell proliferation and tumor malignancy. Low concentrations of Cr were seen in pilocytic astrocytoma. Based on our study, inclusion of diffusion imaging, ADC value, and MRS in imaging protocol of pediatric posterior fossa tumor imaging had a significant value, with sensitivity frequency percentage of 60%. Koob et al. reported that the optimal brain imaging protocol to best determine tumor grade was obtained by combining DWI and PWI, with a predictive diagnostic accuracy of 73.24%, whereas tumor types were better evaluated with the combination of DWI, PWI and MRS, although with a lower performance (55.76%). Therefore, DWI combined with either PWI or MRS can be performed interchangeably because the diagnostic efficacy of these models is quite similar for tumor grading and typing. On the contrary, Qiuhong et al. reported that in some common locations (e.g., the brain stem) of pediatric brain tumors, tumor grading based on MRS findings has been difficult because of excessive magnetic susceptibility perturbations and low spatial resolution. Moreover, Porto et al. have a relatively different opinion. They reported that general morphological aspect of a pediatric brain tumor has a very high diagnostic reliability. Adding information of signal intensities on T2 and DWI further increased the diagnostic accuracy of conventional MRI. Multimodal approaches with MR perfusion, ADC maps, and MRS also provided important information but are time consuming, are sophisticated approaches, and can be difficult to perform, especially in an urgent clinical situation in children with acute intracranial pressure signs.
Finally, DWI with ADC was highly useful in the diagnosis of pediatric posterior fossa tumors. High tumor grade, for example, MB, showed restricted diffusion and low ADC value. Pilocytic astrocytoma showed free diffusion and higher ADC value than that of MB. ADC value could differentiate MB from pilocytic astrocytoma or low-grade glioma. Moreover, MRS data had an important role in tumor grading, as high-grade tumors show elevated choline and reduced NAA ratios. Advanced brain MRI techniques provided important diagnostic role over conventional MRI. No single advanced technique is perfect, but different techniques typically complement one another.
| Conclusion|| |
Conventional MRI provided limited information regarding tumor type and grade, falling short as a definitive diagnostic examination. Advanced brain MRI techniques provide incremental diagnostic value over conventional MRI. However, no single advanced technique is perfect, but different techniques typically complement one another.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
William TO, Brien SR. Imaging of primary posterior fossa brain tumours in children. J Am Osteopath Coll Radiol 2013; 3
Poretti A, Meoded A, Huisman TA. Neuroimaging of pediatric posterior fossa tumours including review of the literature. J Magn Reson Imaging 2012; 35
Wilne S, Dineen RA, Dommott RM. Identifying brain tumours in children and young adults. BMJ 2013; 9
Mohammed FF, Ismail AA, Hassan DI. The role of ADC value in the differentiation between most common pediatric posterior fossa tumours. Egypt J Radiol Nucl Med 2013; 44
Schneider JF, Confort-Gouny S, Viola A. Multiparametric differentiation of posterior fossa tumours in children using DWI and short echo time IH-MR Spectroscopy. J Magn Reson Imaging 2007; 5
Soares DP, Law M. Magnetic resonance spectroscopy of the brain: review of metabolites and clinical application. Clin Radiol 2009; 64
Rumboldt Z, Camacho DLA, Lake D, Welsh CT, Castillo M. Apparent diffusion coefficients for differentiation of cerebellar tumors in children. AJNR Am J Neuroradiol 2006; 27
Zitouni S, Koc G, Doganay S. Apparent diffusion coefficient in differentiation of pediatric posterior fossa tumors. Jpn J Radiol 2017; 35
Koral K, Zhang S, Gargan L, Moore W, Garvey B, Fiesta M, et al
. Diffusion MRI improves the accuracy of preoperative diagnosis of common pediatric cerebellar tumors. AJNR Am J Neuroradiol 2013; 34
Jaremko JL, Jans LBO, Coleman LT. Value and limitations of diffusion-weighted imaging in grading and diagnosis of pediatric posterior fossa tumors. AJNR Am J Neuroradiol 2010; 31
Astrakas L, Ye S, Zarifi M, Makedon F, Tzika AA. The clinical perspective of large scale projects: a case study of multiparametric MR imaging of pediatric brain tumors. Oncol Rep 2006; 15
Vicente J, Fuster-Garcia E, Tortajada S. Accurate classification of childhood brain tumours by in vivo
1H MRS. Eur J Cancer 2013; 49
Koob M, Girard N, Ghattas B, Fellah S, Confort-Gouny S, Figarella-Branger D, et al
. The diagnostic accuracy of multiparametric MRI to determine pediatric brain tumor grades and types. J Neorooncol 2016; 127
Qiuhong H, Ray Z, Xu G, Shkarin P. Magnetic resonance spectroscopic imaging of tumor metabolic markers for cancer diagnosis, metabolic phenotyping, and characterization of tumor microenvironment. Dis Markers 2004; 19
Porto L, Jurcoane A, Schwabe D. Conventional magnetic resonance imaging in the differentiation between high and low-grade brain tumours in pediatric patients. Eur J Pediatr Neurol 2014; 18
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4]