Cholesterol granuloma of the sphenoid wing with long-term growth dynamics: illustrative case

Abstract

BACKGROUND

Cholesterol granulomas typically arise in the petrous apex of the temporal bone, and involvement of the sphenoid wing has not been reported. This case is notable not only for its rare sphenoid wing origin but also for the ability to document its growth dynamics over long-term imaging follow-up.

OBSERVATIONS

A 35-year-old man was found to have a mass along the lateral aspect of the left sphenoid wing during evaluation after minor head trauma. Long-term radiological follow-up demonstrated gradual enlargement of the lesion, prompting resection. Histopathological examination revealed cholesterol clefts surrounded by granulomatous tissue, consistent with a diagnosis of cholesterol granuloma originating from the sphenoid wing.

LESSONS

Although cholesterol granulomas most often occur in pneumatized regions of the skull base, particularly the petrous apex, this lesion developed in an area without bone aeration. Together with the documented long-term growth dynamics, this case suggests that cholesterol granulomas can develop independently of bone pneumatization and highlights the importance of considering alternative pathogenic mechanisms, supported by unique long-term radiological documentation of lesion growth.

https://thejns.org/doi/10.3171/CASE26724

Cholesterol granulomas most commonly originate in the petrous apex of the temporal bone, with additional occurrences reported in paranasal sinuses such as the frontal and sphenoid sinuses.¹ These lesions typically arise adjacent to pneumatized bone, supporting the prevailing hypothesis that aeration contributes to their pathogenesis.² We report a rare case of a cholesterol granuloma arising from the sphenoid wing, in which long-term radiological follow-up enabled detailed observation of its natural history. To our knowledge, this is the first report documenting long-term growth dynamics of a cholesterol granuloma arising from the sphenoid wing. This case not only expands the anatomical spectrum of cholesterol granulomas but also provides insight into their natural history and potential pathogenesis in nonpneumatized bone. Because the sphenoid wing lacks pneumatization, the present case raises the possibility of a pathogenic mechanism distinct from those previously proposed.

Illustrative Case

A 35-year-old man with a history of chronic sinusitis presented to an outside hospital in year X after sustaining minor head trauma. Head CT revealed a hyperdense lesion along the lateral aspect of the left sphenoid wing with associated bone destruction (Fig. 1A). MRI demonstrated a cystic lesion with T1 hyperintensity, T2 hypointensity, and diffusion-weighted imaging hypointensity (Fig. 1B–D). A dermoid cyst was suspected, although CT hyperdensity did not match. The asymptomatic lesion was therefore followed conservatively.

FIG. 1.

Radiological images. A–C: Axial T1-weighted (A), T2-weighted (B), and diffusion-weighted (C) MR images obtained at the initial presentation, showing a mass centered on the left sphenoid wing (arrow). D: Serial axial head CT images obtained over a 10-year period (year X, X+3, X+6, and X+9), demonstrating gradual enlargement of the lesion centered on the left sphenoid wing, with progressive extension into the temporal region and increasing bone destruction over time.

During annual imaging follow-up, the mass showed progressive enlargement, and the patient was referred to our institution in year X+9 (Fig. 1E). On examination, he exhibited a firm, elastic mass with bulging of the left temple, although no neurological deficits were present. CT suggested proptosis of the left eye, but it was not shown on examination. MRI revealed a well-circumscribed, nonenhancing lesion measuring up to 50 mm in maximal diameter, which was hyperintense on T1-weighted sequences and heterogeneously hyperintense on a T2-weighted sequence (Fig. 2A–C). CT demonstrated bone destruction extending from the sphenoid wing to the lateral orbital wall (Fig. 2D). In comparison with the initial imaging, the lesion exhibited interval enlargement from 28 mm to 50 mm, along with changes in both CT attenuation and MRI signal intensity. The tumor volume initially decreased at 1 year and then increased again thereafter. After 5 years, the growth rate accelerated and the volume increased linearly. The timing of the increase in growth rate coincided with the change in CT density from hyperdense to hypodense. Differential diagnoses included lipoma, osteoma, fibrous dysplasia, and hemangioma. Because of its continued enlargement, resection was recommended.

FIG. 2.

Preoperative MRI and CT findings. A: Axial T1-weighted MR image demonstrating a well-circumscribed hyperintense mass along the left sphenoid wing. B: Axial contrast-enhanced T1-weighted MR image showing no significant enhancement of the lesion. C: Axial T2-weighted MR image revealing heterogeneous hyperintensity within the mass. D: Bone-window CT scan demonstrating focal bone destruction along the lateral orbital wall (arrow) and the temporal bone (arrowhead).

A left frontotemporal craniotomy was performed (Fig. 3A). After skin incision and elevation of the temporalis muscle, a focal bony defect was identified in the temporal bone, exposing the tumor capsule (Fig. 3B). On entering the lesion, a mixture of muddy fluid and yellow, glossy material was encountered (Fig. 3C). After confirming the tumor capsule, the lesion was carefully dissected from the intact dura mater and completely removed. Intraoperatively, no continuity was observed between the lesion and the paranasal sinuses or mastoid air cells, and the dura mater remained intact, supporting an origin from the sphenoid wing rather than secondary extension. The presence of bone destruction at the sphenoid ridge also suggests an origin from the sphenoid ridge. The bony defect was subsequently reconstructed using a mesh-style titanium plate, as the defect was relatively small and adequately covered without the need for a patient-specific implant (Fig. 3D). The postoperative course was uneventful, and the patient was discharged home shortly thereafter. MRI performed 2 years after surgery showed no evidence of recurrence.

FIG. 3.

Intraoperative photographs. A: Intraoperative photograph obtained after left frontotemporal craniotomy, demonstrating exposure of the temporal region and an underlying bone defect (arrowheads). The inset shows a 3D reconstruction of the entire skull, providing an overview of the cranial anatomy and the bone-cutting line. B: Gross appearance of the resected tumor capsule showing a focal defect corresponding to the site of bone erosion (arrow). C: Intraoperative view of the opened lesion revealing glossy, muddy cystic contents (asterisk) and adjacent dura mater forming part of the cyst wall (double asterisks). D: Intraoperative photograph obtained after reconstruction of the bony defect using a titanium mesh plate.

Histopathological examination demonstrated abundant cholesterol clefts within granulomatous tissue (Fig. 4A), confirming the diagnosis of a cholesterol granuloma arising from the sphenoid wing. The cholesterol clefts were surrounded by foreign body giant cells (Fig. 4B), indicating a granulomatous inflammatory reaction. Focal intralesional hemorrhage with abundant erythrocytes was also observed (Fig. 4C).

FIG. 4.

Histopathological findings of the lesion (hematoxylin and eosin staining). A: Low-power view showing cholesterol clefts containing granulation tissue surrounded by a thick fibrous capsule (asterisk). B: High-power view showing numerous cholesterol clefts (arrowheads) surrounded by macrophages and foreign body–type multinucleated giant cells (arrow) in a fibrous stroma with hemosiderin deposition. C: Cholesterol clefts (arrowhead) within the hematoma, surrounded by foreign body–type multinucleated giant cells (arrow). Scale bars: 200 μm (A), 50 μm (B and C).

Informed Consent

The necessary informed consent was obtained in this study.

Discussion

Observations

Cholesterol granulomas are cystic lesions characterized by cholesterol crystals within a fibrous capsule.³ They most commonly arise in the middle ear and, intracranially, have been reported predominantly in association with pneumatized bone.⁴ Several cases of cholesterol granulomas involving the orbital portion of the sphenoid bone have been reported. In contrast, the present lesion originated from the sphenoid ridge, separate from the orbit, and to our knowledge no such case has previously been described.

Two major hypotheses have been proposed to explain the pathogenesis of cholesterol granulomas.⁴ The first, termed the obstruction-vacuum hypothesis, suggests that occlusion of a pneumatized space, such as the mastoid air cells or paranasal sinuses, leads to the development of negative pressure that subsequently causes hemorrhage. Breakdown of extravasated erythrocytes results in cholesterol crystal formation, which elicits a chronic inflammatory response and ultimately granuloma formation.⁵ The second, known as the exposed marrow hypothesis, proposes that exposure of bone marrow to a pneumatized cavity causes vascular disruption and hemorrhage, followed by the same cascade of cholesterol crystal deposition and granulomatous reaction.⁶ Importantly, both hypotheses presuppose the presence of a pneumatized cavity, which was absent at the lesion site in the present case.

In the present case, the lesion was located away from both the mastoid air cells and the paranasal sinuses, and the site lacked pneumatized bone. Moreover, the initial MRI demonstrated T1 hyperintensity and T2 hypointensity, findings consistent with subacute hemorrhage. These observations suggest that the pathogenesis was likely unrelated to pneumatization. Although a definitive causal relationship cannot be established, trauma-induced hemorrhage or inflammatory extension from chronic sinusitis may have contributed to the initial hemorrhagic event suggested by the early MRI findings (Fig. 5). The initial CT high density and the presence of intralesional hemorrhage on histopathological examination further support the involvement of hemorrhage.

FIG. 5.

Schematic illustration of the tentative pathogenesis of cholesterol granuloma. A: Trauma or inflammation induces hemorrhage from the exposed bone marrow. B: Hemorrhage leads to the formation of a hemorrhagic cavity, and subsequent breakdown of erythrocytes results in the formation of cholesterol clefts. C: A foreign body inflammatory reaction to the accumulated cholesterol clefts leads to the development of a cholesterol granuloma. D: Repeated hemorrhage promotes further enlargement of the cholesterol granuloma. E: Progressive expansion of the lesion results in thinning of the surrounding bone with focal bone loss.

We speculate that an initial hemorrhagic event, potentially related to trauma or chronic inflammation, led to cholesterol crystal formation and a subsequent granulomatous reaction. During follow-up, the lesion changed from hyperdense to hypodense on CT, coinciding with acceleration of lesion growth. These radiological changes may reflect evolution from an acute hemorrhagic stage to chronic inflammatory cystic lesion. Repeated hemorrhage and the resulting inflammatory response may have contributed to the progressive enlargement observed over the 10-year follow-up period.

Recent advances in patient-specific implants have enabled one-stage cranial reconstruction for large or anatomically complex skull defects.^7–9 However, because the bony defect in the present case was relatively small and localized, reconstruction with a titanium mesh was considered sufficient to provide satisfactory structural support and cosmetic reconstruction.

Lessons

Although cholesterol granuloma of the sphenoid wing is extremely rare, it should be included in the differential diagnosis of sphenoid wing lesions with hemorrhagic imaging features. Long-term follow-up may demonstrate progressive enlargement, and complete resection can provide definitive diagnosis and durable local control.

Acknowledgments

Support for this study was provided by the Japan Agency for Medical Research and Development (AMED) (grant no. JP256f0137001) to M.K.

Disclosures

The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

Author Contributions

Conception and design: Kinoshita, Hirano. Acquisition of data: Kinoshita, Hirano, Tamai. Analysis and interpretation of data: Kinoshita, Hirano. Drafting the article: Hirano. Critically revising the article: Kinoshita, Hirano, Tamai, Nakada. Reviewed submitted version of manuscript: Kinoshita, Tamai, Tanaka, Nakada. Approved the final version of the manuscript on behalf of all authors: Kinoshita. Administrative/technical/material support: Nakada. Study supervision: Misaki. Pathological analysis: Ikeda.

Correspondence

Masashi Kinoshita: Faculty of Medicine, Institute of Medical, Pharmaceutical and Health Sciences, Kanazawa University, Kanazawa, Ishikawa, Japan. m-kinoshita@med.kanazawa-u.ac.jp.