
Lhermitte-Duclos disease, or dysplastic cerebellar gangliocytoma, is a slow-growing cerebellar mass that expands the folia and produces a striated tiger stripe pattern on T2. That pattern was described in only 58.7 percent of published cases, so its absence does not exclude it. The 2025 PTEN hamartoma tumour syndrome guideline calls the lesion pathognomonic of Cowden syndrome, particularly in adults, which makes genetic referral the key recommendation [1][2].
- Key facts for practice and radiology board exams
- What is Lhermitte-Duclos disease?
- MRI features of Lhermitte-Duclos disease
- Is the tiger stripe pattern pathognomonic?
- Does Lhermitte-Duclos disease enhance?
- CT, PET and perfusion findings
- Differential diagnosis
- Lhermitte-Duclos disease and Cowden syndrome
- What to report and what to recommend
- Quiz
- Frequently asked questions
- References
Key facts for practice and radiology board exams
- Usually infratentorial and unilateral. Of the 208 published cases with imaging described: right hemisphere 39 percent, left 36 percent, bilateral 5.7 percent, vermis alone 1.9 percent [1]. A site outside the cerebellum is atypical [2].
- The tiger stripe pattern is not sensitive: described in 122 of those 208 cases, 58.7 percent [1]. Four in ten do not show it, so its absence does not count against the diagnosis.
- Bright on DWI without true restriction. ADC is not reduced and is often mildly raised, 967.8 against 770.4 x 10-6 mm2/s for normal cerebellum [4]: T2 shine-through [3].
- Enhancement is variable, not absent: 14.9 percent of published cases, 64.3 percent in one dedicated imaging series [1][4]. Describe it; it does not exclude.
- Cowden syndrome in 32.8 percent of 302 published cases, confirmed PTEN variant in 19.9 percent. The association was not recorded in 53 percent, so both are floors [1].
- Pathognomonic of Cowden syndrome, particularly in adults, per the 2025 PTEN hamartoma tumour syndrome guideline [2]. The lesion alone justifies genetic referral.
- Demographics: mean age at presentation 33.6 years, 56.6 percent female. Headache in 57.6 percent, ataxia in 36.1 percent [1].
- Outcome: surgery 64.2 percent, recurrence 8.6 percent, mortality 4.3 percent, postoperative cerebellar mutism about 1 percent [1]. Recurrence can follow a long disease-free interval.
- Synonyms, the lesion never settled as neoplasm or hamartoma: dysplastic cerebellar gangliocytoma, dysplastic gangliocytoma of the cerebellum, granular cell hypertrophy of the cerebellum, cerebellar hamartoma, diffuse ganglioneuroma.
What is Lhermitte-Duclos disease?
Lhermitte-Duclos disease is a disorganised overgrowth of one cerebellar hemisphere in which the internal granular layer is replaced by large dysplastic ganglion cells, the Purkinje cells degenerate and the molecular layer becomes hypermyelinated. The folia thicken but keep their laminar arrangement, and it is that preserved layering, a hypermyelinated outer and a cell-rich inner band, that produces the alternating T2 signal. Germline inactivation of PTEN on chromosome 10q23 disinhibits the PI3K-AKT pathway, which accounts for both the overgrowth and the link to Cowden syndrome [8].

On grading, the 2021 WHO classification lists dysplastic cerebellar gangliocytoma among the glioneuronal and neuronal tumours, states grades in Arabic rather than Roman numerals, and grades within tumour type [5]. The published summary assigns it no grade, so the familiar “WHO grade I” shorthand is at best older phrasing.
MRI features of Lhermitte-Duclos disease
Contrast-enhanced MRI is the diagnostic study. The combination to look for is an expansile cerebellar mass with thickened striated folia whose ADC is not reduced.
| Sequence | Finding | Why it matters |
|---|---|---|
| T2 and FLAIR | Alternating hyperintense inner and hypointense outer bands in enlarged folia: the tiger stripe, tigroid or corduroy pattern | The diagnostic feature, but present in only 58.7 percent of published cases [1] |
| T1 | Expansile iso- to hypointense mass crossed by linear hypointense bands | The folial architecture is distorted but preserved, not destroyed |
| DWI and ADC | Bright on DWI, ADC not reduced and often mildly raised above normal cerebellum [4] | T2 shine-through, not hypercellularity [3]. True restriction is atypical and warrants neurosurgical review [2] |
| Post-contrast T1 | None, or thin linear enhancement of the veins running between the folia | Reported in 14.9 percent of published cases and 64.3 percent of one imaging series [1][4] |
| SWI and T2* | Prominent veins coursing between the enlarged folia | Thought to be the source of the linear vascular enhancement [3] |
| MR spectroscopy | Reduced metabolites with a prominent lactate peak; choline usually reduced [3] | A raised choline argues for a high-grade tumour instead |

Is the tiger stripe pattern pathognomonic?
It is specific but not sensitive, and the two are routinely confused. Of 302 published cases, 208 described the imaging and the tiger stripe appearance was present in 122 of those, 58.7 percent [1]. A striated cerebellar mass in an adult is close to diagnostic, but a cerebellar mass without stripes is no argument against Lhermitte-Duclos disease. A striated appearance is also reported in acute cerebellar infarction, which has been misdiagnosed as it outright [7].
Does Lhermitte-Duclos disease enhance?
Sometimes, and the textbook line that it never enhances is wrong. Pooled published cases report enhancement in 14.9 percent, usually mild, but a dedicated imaging series found it in 64.3 percent, with small enhancing veins in half [1][4]. The 2025 guideline nonetheless counts enhancement among the atypical features that should prompt neuro-oncology or neurosurgical assessment [2]. So describe it explicitly rather than omitting it, and let the team weigh it alongside growth rate and diffusion.
CT, PET and perfusion findings
On CT the lesion is hypodense or isodense and rarely calcifies, with no necrosis or haemorrhage. What CT contributes is the surgical question: fourth ventricular and cisternal effacement, and obstructive hydrocephalus. Two advanced findings mislead if read at face value. Perfusion shows raised relative cerebral blood volume, flow and mean transit time, yet it is not specific enough to separate the lesion from a cerebellar tumour [3]. FDG PET shows it to be hypermetabolic despite its benign behaviour, more so on delayed imaging [6]. Neither is evidence of malignancy. Fibre tractography serves only surgical planning.
Differential diagnosis
| Differential | Overlap with Lhermitte-Duclos disease | Discriminator |
|---|---|---|
| Acute cerebellar infarction | Non-enhancing, bright on DWI, and a striated appearance has caused direct misdiagnosis [7] | Acute onset, confined to a vascular territory, ADC genuinely reduced |
| Medulloblastoma | Lateral cerebellar hemispheric mass in adults | Enhances solidly and restricts diffusion with a genuinely low ADC |
| Ganglioglioma | Posterior fossa glioneuronal tumour | Enhances solidly, often cystic with a mural nodule, and does not follow the folia |
| Cerebellar cortical dysplasia | Disorganised cerebellar cortex, non-enhancing | Neither expansile nor progressive, so mass effect and hydrocephalus are rare |
Lhermitte-Duclos disease and Cowden syndrome
This is the part of the report most often left out, and the part that changes a patient’s life. Cowden syndrome sits within PTEN hamartoma tumour syndrome, an autosomal dominant predisposition to breast, thyroid and endometrial cancer, with moderate risks of renal and colorectal cancer and melanoma [10]. The 2025 guideline calls Lhermitte-Duclos disease pathognomonic of Cowden syndrome, particularly in adults [2]; across the published series Cowden syndrome was present in 32.8 percent of cases [1]. A confirmed diagnosis commits the patient to lifelong multi-organ surveillance, and the two current guidelines do not agree on all of it.
| Organ | International PHTS consensus, 2025 [9] | ERN GENTURIS, 2026 [10] |
|---|---|---|
| Breast | MRI annually from 25, mammography annually from 40; risk-reducing mastectomy may be offered from 25 | Same ages, following the BRCA1 pattern |
| Thyroid | Ultrasound every 2 to 5 years from 12; consider annual TSH from 18 | Ultrasound yearly from 10 to 18, then every 2 to 3 years after two normal studies |
| Colorectum | Baseline at 35 to 40; every 1 to 3 years only for 5 or more adenomas, a polyp of 1 cm or more, or high-grade dysplasia | Baseline at 35 to 40; repeat at 5 to 10 years if only hamartomas or no polyps |
| Kidney | Ultrasound every 2 years from 35 to 40 | A single baseline study at 35 to 40, MRI preferred, then none unless that study requires it |
| Endometrium | None outside a clinical trial; prompt assessment of abnormal bleeding | Transvaginal ultrasound at least annually from 35 to 40 |
| Skin | Not specified | Baseline dermatology examination at 30 |
Note what is absent from both: the 2025 guideline recommends against screening neuroimaging of the brain or neuraxis without neurological symptoms or signs, so a confirmed PTEN variant is not by itself a reason to scan [2].
What to report and what to recommend
- State laterality, the involved folia and the degree of fourth ventricular effacement, and say explicitly whether there is hydrocephalus. That decides urgency.
- Name any atypical feature: a site outside the cerebellum, contrast enhancement, diffusion restriction, or rapid growth on serial imaging. Any of these should prompt neuro-oncology or neurosurgical assessment for biopsy or resection [2].
- Recommend genetic referral for PTEN testing in an adult. This is the recommendation most often missing, and it starts the surveillance above.
- For a typical asymptomatic lesion without CSF obstruction, contrast-enhanced MRI every 3 to 12 months is reasonable; after resection, imaging and clinical review every 6 to 12 months [2].
Quiz
Lhermitte-Duclos disease is closely associated with which syndrome?
- Lynch syndrome
- Peutz-Jeghers syndrome
- Familial adenomatous polyposis
- Cowden syndrome
Frequently asked questions
References
- Alanazi AI, Alanezi T, Aljofan ZF, Alarabi A, Elwatidy S. Lhermitte-Duclos disease: a systematic review. Surg Neurol Int. 2023;14:351.
- Dhawan A, Liu D, Baitamouni S, Anthony K, Srivastava S, Hardan AY, et al. Neurodevelopmental and neurologic manifestations of PTEN hamartoma tumor syndrome: management recommendations. Neurol Genet. 2025;11(5):e200299.
- Thomas B, Krishnamoorthy T, Radhakrishnan VV, Kesavadas C. Advanced MR imaging in Lhermitte-Duclos disease: moving closer to pathology and pathophysiology. Neuroradiology. 2007;49(9):733-738.
- Dhamija R, Wood CP, Porter AB, Hu LS, Weindling SM, Hoxworth JM. Updated imaging features of dysplastic cerebellar gangliocytoma. J Comput Assist Tomogr. 2019;43(2):277-281.
- Louis DN, Perry A, Wesseling P, Brat DJ, Cree IA, Figarella-Branger D, et al. The 2021 WHO classification of tumors of the central nervous system: a summary. Neuro Oncol. 2021;23(8):1231-1251.
- Nakagawa T, Maeda M, Kato M, Terada N, Shimizu S, Morooka Y, et al. A case of Lhermitte-Duclos disease presenting high FDG uptake on FDG-PET/CT. J Neurooncol. 2007;84(2):185-188.
- Kumar GK, Rachatte M, Pahari S, Sharma A, Sharma A. Atypical presentation of acute cerebellar infarct as Lhermitte-Duclos disease. Surg Neurol Int. 2025;16:340.
- Zhou XP, Marsh DJ, Morrison CD, Chaudhury AR, Maxwell M, Reifenberger G, et al. Germline inactivation of PTEN and dysregulation of the phosphoinositol-3-kinase/Akt pathway cause human Lhermitte-Duclos disease in adults. Am J Hum Genet. 2003;73(5):1191-1198.
- Dhawan A, Baitamouni S, Liu D, Yehia L, Anthony K, McCarther A, et al. Cancer and overgrowth manifestations of PTEN hamartoma tumor syndrome: management recommendations from the International PHTS Consensus Guidelines Working Group. Clin Cancer Res. 2025;31(9):1754-1765.
- Hoogerbrugge N, Blatnik A, Lautrup CK, Huneburg R, Turchetti D, van Altena A, et al. ERN GENTURIS cancer surveillance guideline for individuals with PTEN hamartoma tumour syndrome (PHTS). Eur J Hum Genet. 2026 Jul 16. Online ahead of print.
Case co-authored by TeamGyan Member Dr Bhargavi Sovani.
Illustration by Dr. Disha Lokhandwala.
