Evolution of epidermoid cyst into dermoid cyst: Embryological explanation and radiological-pathological correlation

Abstract

Intracranial dermoid and epidermoid cysts are usually considered to be two different entities in the radiological and surgical literature. Epidermoid cysts are classically off midline in location, isointense to cerebrospinal fluid on T1 and T2-weighted images and have restricted diffusion, whereas dermoid cysts are classically midline in location, have T1-hyperintense regions due to the presence of fat and show facilitated diffusion. We report a case of radiological epidermoid cyst in baseline imaging, which evolved into a radiological dermoid cyst over time, and explain this unique occurrence with a review of the embryology and histopathogenesis of these cysts.

Introduction

Intracranial dermoid and epidermoid cysts are usually considered to be two different entities in the radiological and surgical literature.¹Both of these cysts are epidermal inclusion cysts, lined by stratified squamous epithelium, similar to skin. Dermoid cysts, in addition, have skin appendages including sweat and sebaceous glands in the cyst wall, which secrete oily fluid producing the characteristic hyperintense signal in T1-weighted imaging (T1WI). Dermoid and epidermoid cysts are formed during the early stages of development, between 3 and 5 weeks of gestation, secondary to either failure of separation between the surface ectoderm and neural tube or due to sequestration of surface ectoderm during fusion.²Dermoid cysts arise at an earlier age of development and are usually midline in location, while intradural epidermoid cysts are off midline with cerebellopontine angle and middle cranial fossa being the commonest locations.

The radiological evolution of epidermoid cyst into dermoid cyst has not been reported earlier in the literature. We report a case of a radiological epidermoid cyst evolving into a dermoid cyst over time, and review the case in the special context of pathological findings and embryological origin to explain this unique occurrence. This case report emphasises the embryological origin of epidermal inclusion cysts, their histopathology and hence explains the imaging findings of these cysts and how they can change over time.

Case report

A 24-year-old man presented to us with refractory seizures for the past 12 years. A baseline non-contrast computed tomography (CT) scan performed 12 years earlier showed a well-defined extra-axial hypodense lesion with peripheral calcifications in the right choroid fissure and medial temporal region (Figure 1(a)). The density of the lesion was similar to cerebrospinal fluid (CSF); however, Hounsfield unit (HU) measurements were not available. Baseline magnetic resonance imaging (MRI) performed at the same time revealed the lesion to be hypointense on T1WI and hyperintense on T2-weighted imaging (T2WI) with incomplete suppression in fluid-attenuated inversion recovery (FLAIR) images (Figure 1 (b–d)). In T1WI, no definite hyperintense areas were present within the lesion. Diffusion-weighted imaging (DWI) showed a hyperintense signal throughout the lesion (Figure 1(e)). These imaging features were suggestive of an epidermoid cyst, and provisional diagnosis of the same was made. MRI performed three years from the baseline scan revealed few focal hyperintense areas on T1WI in the anterior and medial aspect of the lesion (Figure 1(f)). MRI performed at five and eight years from the baseline scan showed a progressive increase in hyperintensities on T1WI within the lesion (Figure 1(g,h)). Hyperintense signals on T1WI were also noted in the subarachnoid spaces suggesting rupture of the lesion. The current MRI, performed 12 years from the baseline (Figure 1(i)), showed the lesion to be predominantly of T1-hyperintense signal with a small cystic component along the posterior and lateral aspect of the lesion. The T1-hyperintense signal was completely suppressed in spectral fat-saturated images (Figure 1(j)). The cystic area was hyperintense on DWI, whereas most of the lesion was showing facilitated diffusion (Figure 1(k)). In steady state free precision (SSFP) imaging (Figure 1(l)), the T1-hypointense component of the lesion showed sheet-like orientation, suggestive of keratin flakes, whereas multiple small lobules with peripheral chemical shift artifacts were seen in the ‘fatty’ T1-hyperintense component of the lesion. Minimal peripheral enhancement was seen in the T1WI after gadolinium-based contrast administration. These imaging features were suggestive of a dermoid cyst.

Figure 1.

(a–e) Baseline computed tomography (CT) and magnetic resonance imaging (MRI) showed features of typical epidermoid cyst. A well-defined extra-axial lesion was seen in the middle cranial fossa in the right side, which is hypodense on CT similar to cerebrospinal fluid with a peripheral speck of calcification (white arrow in (a)), hyperintense on T2-weighted imaging (T2WI) (black arrow in (b)), almost completely suppressed on fluid-attenuated inversion recovery images (white arrow in (c)) and hypointense on T1-weighted imaging (T1WI) (white arrow in (d)). Diffusion-weighted imaging performed at baseline showed complete diffusion restriction (white arrow in (e)). (f–i) T1-weighted magnetic resonance images performed over a period of 12 years showed a progressive increase in hyperintense signals on T1WI, initially seen in the anteromedial aspect of the lesion (white arrows in (f–i)), progressing over time. (j) Spectral fat saturated T1WI shows suppression of hyperintense signal in the lesion, confirming the presence of fat (white arrow in (j)). (k) In the current MRI (12 years from baseline), restricted diffusion is seen only in the small part of the posterior and lateral aspect of the lesion (white arrowhead in (k)), while most of the lesion showed facilitation of diffusion (white arrow in (k)). (l) Steady state free precision images show layers of ‘sheet-like’ orientation in the cystic component of the lesion (white arrowhead in (l)), corresponding to epidermoid cyst-like component of the lesion. Fatty component of the lesion shows small globular areas with chemical shift artifacts (white arrow in (l)).

The lesion was excised by way of a temporal craniotomy, one week after the current MRI (12 years from the time of baseline imaging). An extra-axial cyst with keratin flakes, hair and fat was seen in the choroid fissure, confirming the diagnosis of dermoid cyst. The postoperative period was uneventful.

Histopathological examination (Figure 2) showed skin appendages including hair follicle and sebaceous glands, along with keratin flakes. The wall of the cyst showed inflammatory infiltrate with haemorrhages, changes secondary to rupture. These findings were consistent with a ruptured dermoid cyst.

Figure 2.

(a) Histopathological examination of the lesion showed typical keratin flakes (black arrow in (a)), in the posterior aspect of the lesion. This corresponds to the diffusion restricting ‘epidermoid’-like area of the lesion in magnetic resonance imaging. Note the similarity of the orientation of keratin flakes in histopathological examination (HPE) and steady state free precision magnetic resonance image (white arrowhead in Figure 1 (l)). (b) In the anterior aspect of the lesion, sebaceous glands (black arrows in (b)) and hair follicle (arrow head in (b)) are seen in the cyst wall, characteristic of a dermoid cyst.

Discussion

Dermoid and epidermoid cysts are epidermal inclusion cysts, lined by stratified squamous epithelium, similar to skin. The superficial layer contains only layers of dead cells or the ‘keratin’, which is shed off into the cavity, being replaced by the cells from the deeper layers. With time, multiple sheets of keratin in concentric lamellar arrangement accumulate within the cavity. Sheets of keratin allow free diffusion of water only along the plane of the sheet, but not perpendicular to it, hence producing the characteristic hyperintense signal in DWI.³Unlike the white matter, where diffusion of free water is facilitated in only one direction (linear anisotropy), in epidermoid cysts diffusion is facilitated in two directions along the plane of keratin sheets (planar anisotropy).^4,5Recent magnetic resonance techniques, diffusion tensor imaging and tensor metrics imaging can be used to identify the planar anisotropy in the epidermoid cysts. Dermoid cysts are similar to epidermoid cysts with additional skin appendages in the cyst wall. These skin appendages include hair follicles, sweat and sebaceous glands, which shed hair and secrete sweat and sebum into the cavity. These secretions along with their breakdown products form an oily mixture, resulting in the characteristic contents of a dermoid cyst including fat and hair.⁶Continuous active secretion of fluid into the cavity increases intracystic pressure over time leading to cyst rupture. On the contrary, there is no ‘active secretion’ within the epidermoid cyst due to the absence of skin appendages explaining the low incidence of rupture in epidermoid cysts.⁷

The pathological distinction of epidermoid and dermoid cysts is based only on the presence of skin appendages in the cyst wall. When an epidermal inclusion cyst has the majority of its wall lining as only stratified squamous epithelium, with only small ‘microscopic’ areas of skin appendages, it is still a pathological dermoid cyst. In the initial imaging, these cysts might have the appearance of epidermoid cysts as the majority of the contents would be just desquamated keratin. With time, as the glands secrete more secretions, it might produce the classic appearance of a dermoid cyst, as seen in our case. The presence of secretions both dilutes the contents and disturbs the directional orientation of keratin, resulting in the loss of the hyperintense DWI signal. This is evident in our patient in whom initial imaging (hypointense on T1WI, hyperintense on T2WI with diffusion restriction) was suggestive of a typical epidermoid cyst, which showed a progressive increase in fatty components over time and a change in diffusion characteristics (Figure 3). In the literature, a similar case with no hyperintense signals on T1WI in baseline imaging, and subsequent imaging showing a hyperintense signal on T1WI has been reported. However, the patient was not subjected to fat-saturated techniques and was lost on follow-up before surgery.⁸The same theory can be used to explain the reason behind the few cases of dermoid cysts reported in the literature, which showed imaging findings similar to epidermoid cysts, as in the fourth and fifth case described by Orakcioglu et al.⁹and the case described by Fanous et al.¹⁰These dermoid cyst cases reported might only have small microscopic areas of skin appendages, hence having the imaging appearance of epidermoid cyst. There is also a possibility that a few of the cases reported as ‘ruptured epidermoid cyst’ or ‘epidermoid cyst with fat’, as in the first representative case described by Chowdhury et al.¹¹may actually be dermoid cysts with minute quantities of skin appendages missed in histopathological examination due to inadequate sampling of the cyst wall.

Figure 3.

Schematic diagram of radiological evolution of an epidermoid cyst into a dermoid cyst. (a) To start with, these cysts have a small component of skin appendages in the wall (blue structures). These components are less compared to the rest of the squamous epithelium, thereby the contents of the cyst are predominantly keratin flakes (asterisk in (a)), producing the characteristic appearance of an epidermoid cyst in the imaging. Note that pathologically they are dermoid cysts, as they have skin appendages in their wall. (b) and (c) Over time, as the skin appendages secrete more fatty components (arrows in (b) and (c)), they produce the characteristic T1-weighted hyperintense appearance. They also distort the keratin orientation within the cyst, resulting in diffusion facilitation.

Typical or ‘black’ epidermoid cysts have characteristic CT and MRI appearance. They are hypodense on CT, hypointense on T1WI and hyperintense on T2WI, showing an almost similar signal to CSF. They also show incomplete suppression in FLAIR images and diffusion restriction, thus differentiating them from arachnoid cysts.¹²The so-called ‘white’ epidermoids, which are hyperintense on T1WI, have variable appearance. The reason for their appearance is theorised to be increased protein, haemorrhagic or lipid content.^13,14White epidermoid cysts secondary to increased protein or haemorrhagic content are usually hyperdense on CT and hypointense on T2WI. They are also less likely to show diffusion restriction, secondary to more liquid contents or disturbance in the directional orientation of keratin, or both. The T1-hyperintense signal does not get suppressed in fat-saturated sequences. Most of the white epidermoid cysts reported in the literature belong to this category.^15,16Out of the 15 cases of hyperdense epidermoids reported by Li et al.,¹⁷six patients had MRI. All of them showed a hypointense signal on T2WI and five of them showed an iso to hyperintense signal on T1WI.

Although the contents of the epidermoid cyst include cholesterol and triglycerides, they are never high enough to produce a hyperintense signal in T1WI.¹⁸To the best of our knowledge, none of the white epidermoids reported in the literature, which were hypodense on CT and not hypointense on T2WI, were subjected to spectral fat saturation techniques to prove the fat content. Chemical analysis of the contents of the epidermoid cyst showed increased protein content in these cysts probably explaining the signal intensity on T1WI. Similar to the change in signal intensity of sinonasal secretions with varying protein concentrations, as described by Som et al.,¹⁹epidermoid cysts can also have a varied magnetic resonance appearance based on the protein concentration. Epidermoid cysts can be hypointense on T1WI and hyperintense on T2WI (typical epidermoid cyst), intermediate signal on T1WI and hyperintense on T2WI, hyperintense on T1WI and T2WI, intermediate signal on T1WI and hypointense on T2WI and hypointense on both T1WI and T2WI, with progressively increasing protein concentration. Lipids in epidermoid cysts are caused by the saponification of dead cells, and are not due to active secretion. Triglycerides are seen in ‘white’ epidermoids which might return a T1-hyperintense signal. They are absent in ‘black’ epidermoids, which show more cholesterol content.²⁰

Conclusion

Radiological epidermoid cysts can turn out to be pathological dermoid cysts, but not vice versa. The presence of ‘radiological’ fat in the lesion would mostly turn out to be a pathological dermoid cyst, provided the entire cyst wall is adequately sampled and examined. Although there is radiological evolution of an epidermoid cyst into a dermoid cyst, pathologically there is no real transformation. As both dermoid and epidermoid cysts have the same pathogenesis, these cysts should be better termed as dermoid/epidermoid as in the pathological literature, and not as separate entities.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Conflict of interest

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.