 This three-year-old has increasing head size, lethargy, and vomiting. Let's have a look at the imaging. The axial diffusion-restricted image demonstrates high intensity in a mass consistent with some diffusion restriction. We'll look at the ADC map in a moment. The axial T2 shows a heterogeneous mass. You'll have to decide if it's intra or extraaxial over the course of these images, and whether it's intra or extraventricular. Here are some foci of high signal within. An axial flare image, an axial T1, a coronal T2, an axial ADC map or parametric map, an axial contrast-enhanced T1, and a coronal inversion recovery, contrast-enhanced T1. Let's tackle our questions. What is the most likely diagnosis? Is it A, a typical teratoid-rabdoid tumor, or ATRT? B, glioblastoma multiforme. C, polycytic astrocytoma. D, ependymoma. E, astroblastoma. You can pause if you wish, and move on to question number two. Which is not true regarding ATRT? A, infant. B, median survival, six years. C, bubbly or cystic lesion. D, world health organization grade four. E, lesion is infertentorial 50% of the time. Question number three. Which is not true regarding ATRT? A, bleeds. B, presents with dissemination in 20%. C, rarely calcifies. D, aggressive spect or spectroscopy MRI pattern. E, diffusion restricts. Let's go back to the imaging. We'll start out with the first images presented. The diffusion restricted image, which shows high signal, diffusion, restriction, an indicator of an aggressive or nasty lesion. The axial T2, which shows little evidence of cytoplasm, or it's cytoplasmic poor lesion, since it's not very bright, but there are cystic foci within. The lesion is parenchymal or intraaxial, one of the first things you should decide. We'll look at the angle of the lesion relative to the brain parenchymal, which is acute. In an extraaxial lesion, the interface is obtuse. And the white matter will typically not wrap all the way around the lesion if it's extraaxial, but it will wrap around the lesion, along with a little bit of gray matter when it's intraaxial. Our mass, as we look at it serially, is extraventricular, although it is severely compressing the ventricular system, severely displacing the temporal horn and the uncus medially, leading to impending, uncle herniation, a potential catastrophe. On the axial T1 weighted image, look at the acute angle between the lesion and the white and gray matter, an indicator of a parenchymal intraaxial mass. There is midline shift, another indicator of a severe, serious, life-threatening lesion. The coronal T2, once again, demonstrates an intraaxial position, further supported by the absence of a membrane, rim or pseudo membrane around the lesion between it and the parenchyma. The mass seems to blend somewhat imperceptibly, especially superiorly, with the brain parenchyma. More cystic foci not to be confused with the compressed, surrounded and trapped temporal horn. The axial ADC parametric map shows predominantly decreased intensity, another sign of diffusion, restriction, a negative prognostic sign. The axial T1 contrast image and the coronal inversion recovery contrast image further support the diagnosis of an intraaxial parenchymal mass. It'd be highly unlikely to have an extra axial mass of this size in a young individual who's in infancy. Now let's go back to our questions. What is the most likely diagnosis? The answer is a typical, teratoid, rhabdoid tumor. These are lesions that present in two, three and one-year-olds. The age is appropriate, the size is appropriate, the aggressiveness is appropriate, but not appropriate for GBM, glioblastoma multiforme, which occurs in adults, is uncommonly seen in children unless it de-differentiates from pontine astrocytoma. This lesion is not in the pons. Pillocytic astrocytoma. These are classically seen in young individuals in the posterior fossa. They are typically cystic with an enhancing nodule. They're not this aggressive. They don't diffusion restrict. It's not a good choice. Appendimoma, when it's parenchymal, when it's supertentorial, it's an adult. It's not a two-year-old, a three-year-old, not a good choice. Astroblastoma. This occurs in young individuals, but not in infants. Therefore, not a good choice. Question number two, which is not true regarding ATRT? The answer is median survival, six years. Unfortunately, the median survival is far shorter than this. It's a poor prognosis lesion. It is a World Health Organization grade four lesion. It's aggressive. It is bubbly. It is cystic. It occurs in infants. The lesion is infratentorial 50% of the time, even though this one isn't. It's supertentorial at least 40% of the time. Which is not true regarding ATRT? The answer is rarely calcifies. That is an incorrect statement. It calcifies not uncommonly. The rest of the choices are true. It does bleed. It does disseminate not uncommonly, at least 20% of the time. It does have an aggressive spectroscopy MRI pattern, as we'll see in a moment, and it does diffusion restrict because it's aggressive as all get-out. The ATRT, or atypical, teratoid, rabdoid tumor, is rare. It's aggressive. It's large. It's seen in an infant under age three. It resembles the primitive neuroectodermal tumor or meduloblastoma, especially the parenchymal type. The supertentorial type of ATRT is large. It occurs above the tentorium about 40% of the time. It is cystic. It can be hemispheric or supracellar. It occurs infratentorially 50% of the time, usually off midline. It may involve the CP angle. It may involve the brain stem, in some ways simulating glioblastoma multiforme, but glioblastoma multiforme doesn't occur in two-year-olds, one-year-olds, three-year-olds. It may occur in the cerebellum, mimicking the meduloblastoma, pinet tumor, which is the primary differential diagnosis. The age is helpful. An infant with a meduloblastoma-like lesion below age three, you've got to think ATRT. The average size is about two centimeters, but they can and often are huge. They're heterogeneous. They bleed. They calcify, not uncommonly. They enhance like our lesion moderately, heterogeneously, and they are, although not shown, hyper-dense on non-contrast CT, except for the cystic foci. Obstructive hydrocephalus is common. Did you notice in our case that the entire ventricular system, especially the ventricles on the left side, were dilated? Therefore, this mass is not only producing herniation on the right, it's obstructing the entire ventricular system. On MRI, the lesions are solid with cystic foci. They're bloody. There is transapendymal migration. There is peri-ventricular infiltration. They seed. They involve the leptomaninges, the diffusion restrict, and on spec-MRI, the choline level is elevated, the lactate lipid peak is elevated, and the NAA, a cell neuron marker, is low or absent. The differential diagnosis of ATRT includes appendemoma. One-third of appendemomas are parenchymal above the tentorium, but in the fourth ventricle or infratentorial, they are plastic. They squeeze into small spaces like the foramen of Lusca. They do bleed. They do make cysts. They do calcify. They do enhance heterogeneously, but ATRTs are not soft. They're expansile, and they occur in a younger age group under age three. Teratomas. These may have some high signal inside, but usually not blood. It's fat. They may be pineal, pericellar. They calcify. They can occasionally bleed, but they don't seed if they are benign teratomas, and they do not. Diffusion restrict. Hemangioblastoma. This would not be a consideration in a two or three- or one-year-old unless there were von Hippel-Lindau, and even then you don't see them that young. The average age of a hemangioblastoma at presentation is 30 years of age. So the patient is a little bit old, about 27 years too old. The patient also doesn't have von Hippel-Lindau. If you see a hemangioblastoma, always check the globes for a prior retinal bleed. Retinal bleed plus posterior fossa-enhancing mass equals von Hippel-Lindau. Glyosarcoma. These are exophitic from the brainstem. They may rise from pontine gliomas like glioblastoma multiformi. Chorideplexus papilloma. These are intraventricular. Our mass, in this case, is not intraventricular. Therefore, chorideplexus papilloma is not a consideration. The enhancement is more intense. It's more homogeneous, and the mass looks like a giant piece of cauliflower. This mass is a little bit smoother even though it infiltrates the brain parenchyma. Pillocidic astrocytoma. A large cystic mass is more frequent. A single, loculated cyst with an enhancing nodule. The age is a little different. 5 to 15 as opposed to less than 3 years of age. Another uncommon lesion that you might have considered is the pylomixoidastrocytoma. These are typically found in infants, like our mass, but they are chiasmatic hypothalamic lesions. They are solid like this lesion. They enhance like this lesion, but the location in the hypothalamus and chiasm is quite different. The primitive neuroectodermal tumor or the medulloblastoma, this is the one that is the major differential diagnosis. But if you are shown a mass that looks like a medullo and somebody 3 years of age or younger, you probably won't get this as a choice, but you should err towards or at least think of ATRT when you're practicing clinically. Medullos are a little older. They are less likely to have cystic foci. They also seed. The pathology of ATRT includes monosome 22 abnormalities especially, monosome 22 and deletion of band 22q11. The raddoid cells are similar to malignant raddoid tumor in the kidney. So in summary, a young infant with a large head with nausea and vomiting 3 years of age or younger with obstructive hydrocephalus with a large mass with moderate enhancement of cystic foci with seeding, with bleeding, with cysts, think ATRT. Let's move on.