Lhermitte-Duclos Disease or Neoplasm?
A 5-month-old girl was brought to her local emergency department because she displayed increasing fussiness and back arching for 3 days. She vomited once and was febrile on the day of presentation. Meningitis was suspected. A lumbar puncture was performed. It revealed a cerebrospinal fluid (CSF) protein level of 120 mg/dL and a CSF white blood cell count of 10/µL, with 50% polymorphonuclear cells. Antibiotics were empirically administered, and the infant was transferred to a tertiary care facility for further treatment.
Figure. MRI reveals a posterior fossa mass in the 5-month-oldpatient.
HISTORY
A 5-month-old girl was brought to her local emergency department because she displayed increasing fussiness and back arching for 3 days. She vomited once and was febrile on the day of presentation. Meningitis was suspected. A lumbar puncture was performed. It revealed a cerebrospinal fluid (CSF) protein level of 120 mg/dL and a CSF white blood cell count of 10/µL, with 50% polymorphonuclear cells. Antibiotics were empirically administered, and the infant was transferred to a tertiary care facility for further treatment.
PHYSICAL EXAMINATION
On arrival at the tertiary care facility, the child had normal vital signs and was alert, mildly irritable, and consolable. Findings from her physical examination were significant for a large head, a bulging fontanelle, and a downward gaze preference.
LABORATORY AND RADIOLOGICAL STUDIES
The CSF cultures were pending from the referring facility, and radiological studies were then obtained. A CT scan of the brain revealed obstructive hydrocephalus and a posterior fossa mass. MRI (Figure) was obtained to better define the abnormalities; the findings revealed an enlarged right cerebellar hemisphere with no focal dominant mass but with diffuse nodular enlargement of cerebellar folia and widening of the white matter in a striated pattern with obstructive hydrocephalus and associated tonsillar herniation.
DIAGNOSIS: Medulloblastoma
Pediatric neurosurgery was consulted for placementof a ventriculoperitoneal shunt, and the patientwas subsequently discharged to home withthe agreed-on diagnosis of Lhermitte-Duclos disease(LDD) based on the distinctive radiologicalfindings. No further treatment was administered,but close monitoring by the pediatric neurologist,neurosurgical team, and primary care pediatricianwas instituted.
One month later, the child was rehospitalized becausesymptoms recurred. On repeated MRI, themass was noted to have increased significantly insize, and the decision was made to excise it. Tissueanalysis revealed a tumor consistent with medulloblastoma.Despite aggressive treatment with 7 cyclesof chemotherapy, the patient succumbed todisease at approximately 18 months after initialpresentation.
LDD VERSUS MEDULLOBLASTOMA
LDD, or dysplastic cerebellar gangliocytoma, is a rare benign brain lesion described in approximately 200 cases in the medical literature,1 with the majority occurring in adult patients presenting with occipital head pain and other typical signs of a mass in the posterior fossa. In contrast, medulloblastoma is the most common, poorly differentiated CNS malignancy, accounting for 20% of all pediatric brain tumors.2 Both LDD and medulloblastoma arise in the posterior fossa, but the distinctive appearance of LDD on MRI typically allows for a straightforward diagnosis based on the MRI findings alone.3-5 This scenario of a 5-month-old girl in whom medulloblastoma was initially misdiagnosed as LDD based on her MRI findings is similar to one other case reported by Chen and colleagues6 concerning an 18-month-old patient with medulloblastoma whose MRI findings mimicked LDD.
PITFALLS OF DIAGNOSING RADIOLOGICALLY
Accurate and prompt determination of the cause of a CNS mass in a child is crucial to administering appropriate therapy in a timely manner. Histological examination of any tumor is the definitive method of determining the exact type of lesion present, but obtaining a tissue specimen is fraught with serious risks. Although establishing a diagnosis based on radiological findings alone is attractive and noninvasive, it also carries the risk of misdiagnosis.
The presence of a lesion rare for a patient's age should have raised doubts in the image-based diagnosis. LDD is a rare entity. Our case highlights the importance of questioning its diagnosis in any child. This case illustrates the pitfall of diagnosis based on morphological imaging alone, since our patient's tumor displayed the features (Figure) of enlarged cerebellar folia, enlargement of the white matter between adjacent cerebellar folia, and absence of a well-defined focal dominant mass that are so typically present with LDD.
The Table highlights the imaging features typically associated with LDD, medulloblastoma, and medulloblastoma with extensive nodularity (MBEN) in comparison with the features present in the case. Note that the features of the case, LDD, and the MBEN variant of medulloblastoma are highly concordant. This concordance of imaging findings contributed to the misdiagnosis. Although the case did have the finding of hyperintense signal on diffusion-weighted imaging, this was not considered highly in determining the diagnosis, since there is great variability and inconsistency in the presence of this finding in neoplastic brain lesions.
On MRI, LDD appears as a non-enhancing striated mass in the cerebellar hemisphere, hyperintense on T2-weighted images, and hypointense on the corresponding T1-weighted images. The folia are thickened because of an excess of dysplastic cortical neurons that results in a loss of Purkinje cells and thinning of the medullary white matter.4,5 This finding is unique to LDD because most other cerebellar masses destroy the folial pattern and produce enhancement.
In contrast to LDD, the typical medulloblastoma in a young child is a midline, spherical mass that fills the fourth ventricle, arising from the superior medullary velum of the fourth ventricle. It appears to be a discrete mass that displaces, rather than invades, adjacent normal cerebellar tissue. Typically, medulloblastoma will be hyperdense to normal brain tissue on a noncontrasted CT scan, and it may calcify. Medulloblastoma enhances intensely on either CT or MRI. T2-weighted images of typical medulloblastoma are near isointense to gray matter.
MBEN, which affects children younger than 3 years, is a rare variant of medulloblastoma, which has alternating layers, or zones, of differing tumor histology.7 These alternating layers mimic the striated pattern seen in LDD.
Differentiating medulloblastoma from LDD is essential in order to counsel families about further therapy and prognosis. The long-term prognosis of medulloblastoma can be poor and is affected by age (poorer in the patient younger than 3 years), extent of postoperative residual disease (poorer if residual disease is greater than 1.5 cm2), and the presence of metastatic disease.8 In contrast, the few reported cases of pediatric LDD describe a fairly benign course.9
Since the management of LDD and medulloblastoma differ markedly, differentiating these 2 entities is paramount. The optimum management of either disease is yet to be decided, but the diagnosis must be definitively determined before a treatment pathway can even be established in a particular patient.
In general, the treatment of both tumors includes surgical resection. Although LDD is "benign," it is managed by surgical decompression or tumor excision if the patient is symptomatic.3 If surgery is not needed, the patient is monitored closely for neurological sequelae. Children with this entity have survived by using a wait and see approach.9
Whereas surgery or close observation alone will be effective treatments for LDD,3 management of medulloblastoma also includes combined modalities (radiotherapy and chemotherapy) for improved survival rates.10 Pediatric survivors of medulloblastoma are at particular risk for serious adverse effects secondary to their treatments.
Our case of misdiagnosed medulloblastoma reiterates the pitfalls associated with applying radiological findings alone to diagnosis of a tumor. Included is the serious psychological distress that misdiagnosis causes patients' parents and older patients.
Each case of posterior tumor fossa must be evaluated individually, but the clinician should give strong consideration to obtaining a histological diagnosis in all cases despite the great risks that accompany tissue sampling, even if MRI findings are very highly suggestive of the nonmalignant entity of LDD.
References:
References
1.
Robinson S, Cohen AR. Cowden disease and Lhermitte-Duclos disease: an update. Case report and review of the literature.
Neurosurg Focus.
2006;20:E6.
2.
Frosch MP, Anthony DC, De Girolami U. The central nervous system. In: Kumar V, Abbas AK, Fausto N, eds.
Robbins and Cotran's Pathologic Basis of Disease.
7th ed. Philadelphia: Saunders; 2005:1401-1420.
3.
Carlson JJ, Milburn JM, Barre GM. Lhermitte-Duclos disease: case report.
J Neuroimaging.
2006;16:157-162.
4.
Kulkantrakorn K, Awwad EE, Levy B, et al. MRI in Lhermitte-Duclos disease.
Neurology.
1997;48:725-731.
5.
Meltzer CC, Smirniotopoulos JG, Jones RV. The striated cerebellum: an MR imaging sign in Lhermitte-Duclos disease (dysplastic gangliocytoma).
Radiology.
1995;194:699-703.
6.
Chen KS, Hung PC, Wang HS, et al. Medulloblastoma or cerebellar dysplastic gangliocytoma (Lhermitte-Duclos disease)?
Pediatr Neurol.
2002;27: 404-406.
7.
Suresh TN, Santosh V, Yasha TC, et al. Medulloblastoma with extensive nodularity: a variant occurring in the very young-clinicopathological and immunohistochemical study of four cases.
Childs Nerv Syst.
2004;20:55-60.
8.
Gilbertson RJ. Medulloblastoma: signaling a change in treatment.
Lancet Oncol.
2004;5:209-218.
9.
Capone Mori A, Hoeltzenbein M, Poetsch M, et al. Lhermitte-Duclos disease in 3 children: a clinical long-term observation.
Neuropediatrics.
2003;34: 30-35.
10.
Walker DA, Wilne S. Treatment of medulloblastoma in young children.
Lancet Oncol.
2005;6:541-542.
