Section : Seborrhoeic Keratosis (seborrheic keratosis)(seborrheic warts)(age spots)

Etiology (For Professionals)

Etiology (For Professionals)

The origin and cause of SK is to the best of our knowledge unknown and what follows below is a list of proposed hypotheses.

Origin of SK

Hair follicle epithelium

Whatever, the triggers and mechanisms of pathogenesis of seborrheic keratosis, they seem to develop from the follicular epithelium.This hypothesis come from the fact that SKs are absent from the palms and soles, where hair follicles are absent. (Lever histopathology of the skin)

IFKs (Inverted follicular keratoses) represent the endophytic pattern of growth to the opposite of classical SKs. IFK develops from the follicular infundibulum and simulates a proliferative pattern by penetrating the epidermis {Degos, 1968; Mehregan, 1964}.

Histological staining of SK also seem to support the fact that SKs come from the follicular epithelium, by showing that antibodies to constituents of the follicular apparatus are found in SK keratinocytes. Different antibodies are known to react with different constituents of the follicular apparatus (HKN-2,4, 5, 6, 7, BKN-1).  The monoclonal antibodies HKN-2 and HKN-4 react with all constituents of the follicular apparatus but also with the epidermis {Ito, 1986; Ito, 1986}. HKN-6 and HKN-7 react with the inner root sheath and HKN-5 also reacts with the inner layer of the outer root sheath. BKN-1 reacts with the basal cells of the interfollicular epidermis and infundibulum and with the cells below the follicular isthmus (lower follicular cells) in the normal human skin {Shimizu, 1987}(figure available in the article).

Shimizu et al (1989) took 9 biopsies of SK, 3 were histologically irritated and 6 non-irritated. All the lesions were stained with the monoclonal antibodies in the figure above. Only RGE53 and CEA, which  are not follicular markers, were absent.

In non irritated specimens, BKN-1 was found in the basal cells facing the dermis; it was sometimes present suprabasally. HKN-2 showed an opposite pattern of staining, with presence principally in the suprabasal to the granular layer.

In irritated specimens, BKN-1 and HKN-2 staining showed the same distribution with more positive cells of the former being present suprabasally. The weakness of this study, is that HKN-2 can also be found in the normal epidermis.

Solar Lentigo

Solar or senile lentigo is a skin lesion considered by some as a plane SK. The SK surrounding skin often shows signs of photoaging: multiple solar lentigo are present. Often, there are patients who develop SK lesions of the reticulate type on macular spots {Mehregan, 1975}. (See figure in “clinical features”)

Solar lentigo have as histological characteristics, a slight acanthosis with preservation of the granular layer, and a slight hyperkeratosis {Cawley, 1950}. A slight increase in the number of melanocytes has been reported {Hodgson, 1963}.

Solar lentigo and SK have a common etiological cause, UV radiation (see UV light). Moreover, both lesions show an enhanced expression of P16. If we adhere to the retentional model (see “accumulation by senescence” in Suggested mechanisms of growth) into account, solar lentigo could represent a beginning of accumulation of keratinocytes continued by SK.

Causal factors

Even though the cause of SK is unknown some “risk” factors seem to be implicated.

UV exposure (see pigmentation, retentional)

Exposure to sunlight has already been suggested as a cause of SK in a study in the United States{Engel, 1988}.In addition Yeatman{Yeatman, 1997}compared his study group in Australia with a study group in the UK {Memon, 1995}and found that the population in Australia was affected earlier. He supported Engel’s hypothesis that sunlight had a role in the development of SK. Yeatman also found an increased prevalence of SKs on sun-exposed sites.

UVB light provokes the expression of p16 {Pavey, 1999}. Pavey et al. applied suberythematous doses of UVB on normal skin. 16 hours after, a peak of p16 was present p16 was shown to stay present 24 hours after exposition but was not found in the dermis. p16 subsequently decreased 72 hours after exposition. UVC, which is more prevalent in Australia due to the diminished ozone layer (authors’ opinion) also induces p16 expression {Wang, 1996} in normal skin.

In SK, p16 is expressed even without UV exposure {Nakamura, 2003}(see retentional model in mechanisms of growth).

HPV infection

Verruca Vulgaris and SK have many histological characteristics in common and the HPV viral origin of the former has been postulated to explain the latter. Human papillomavirus (HPV) has been detected in genital and non genital lesions of SK mostly by polymerase chain reaction (PCR).

In genital skin, HPV has been detected by PCR in 72% (18 of 25 specimens) cases {Bai, 2003}. Unfortunately, the lesions found in this location resemble condyloma accuminata and Li et al (1994) {Li, 1994} suggests that genital SKs are merely condyloma accuminata (due to HPV). This could explain why HPV would be absent in non genital skin.

Non genital skin reveals HPV present by PCR ranging from 19,65% to 83,7% of specimens. The most recent work was done by Gushi et al (2003) {Gushi, 2003} who revealed by in situ hybridation 28,8% (30 of 104 specimens) cases positive with HPV DNA in epidermal keratinocytes. Even a greater amount of  positive specimens was obtained by PCR. 83,7% ( 87/103) specimens were positive for HPV 18, 77,9%(81/103) positive for HPV 6, 70,2% (73/103) positive for HPV 18 and 6 together. Control normal skin did not reveal HPV by in situ hybridation, even though seven of them were positive on PCR, thus suggesting together with the work of Gushi et al. a lack of sensitivity and specificity of PCR for HPV detection. In another study by Lee and al. (2003), oncogenic HPV (6, 11, 31, 33) were never found in a study of 40 patients.

One study goes further postulating that the HPV found in SK is the same of that found in Epidermodysplasia verruciformis (Detection of epidermodysplasia verruciformis-associated human papilloma virus DNA in non genital seborrheic keratosis). HPV was found in 42 of 55 specimens and genotypes included HPV 20, 23, 5, 17 et RTR. These subtypes were found together in 5 specimens.

Tsamboas and al (1995) {Tsambaos, 1995} showed 19,65%  (34   of 173 specimens) HPV positive PCRs, whereas all of the control group, consisting also of 173 specimens was negative.

To summarize, non oncogenic HPV are  present in some SKs in contrast to normal skin, but this does not give any information of a link of causality between its existence and  the development of SKs; HPV in one study has been found in 79% on the top of 43 SK specimens but only in 8% of strip biopsies (superficial layers of the skin){Forslund, 2004}. This result suggests that HPV doesn’t really penetrate the surface of SK. This suggests HPV would be a contaminant of the skin. Indeed no data tells us whether HPV appears because of the development of SK, or if SK is a consequence of HPV.

As a conclusion HPV is postulated as a cause of SK but no study with an absolute causal link has been proven or published.

Deficiency in presenilin

This is the only study trying to link non eruptive SK with an internal disease, in this case Alzheimer’s disease. Tournoy and al (2004) {Tournoy, 2004} studied lesions clinically and histologically equivalent to SK in mice and found a deficiency in presenilin.

Presenilins 1 and 2 are components of a gamma secretase enzyme which releases the amyloid beta components of the APP protein in Alzeimer’s disease. Over one hundred mutations of presenilins 1 and 2 have been documented, which can account for the familial form of Alzheimer’s disease, because these mutations enhance the generation of amyloid beta peptides number. Thanks to these generated peptides, cleavage of signalling proteins such as notch is done. On the contrary, cleavage of the notch protein is not done in “knockout” mice, thus not activating its signalling pathway {De Strooper, 1998; Shen, 1997; Hartmann, 1999; Wong, 1997}.

The notch pathway is implicated in the fate of the cell and the regeneration of tissues {Pear, 2003; Ohishi, 2003; Mori, 2003; Duvic, 2002; Conboy, 2002}. It also has a role in vascular differentiation {Nijjar, 2002; Lawson, 2001; Uyttendaele, 2000}. In active notch (non cleaved), the cell is maintained in a proliferative state {Hitoshi, 2002} or hyperproliferative {Jhappan, 1992} as suggested by an absence of presenilin. But on the contrary in keratinocytes, notch induces a differentiation and acts as a tumour suppressor, suggesting that a cleavage of notch and its impaired signalling pathway could induce dedifferentiation and proliferation. Notch is also implicated in cell fate decisions between B and T-lymphocyte lineages {Allman, 2002; De Smedt, 2002} and between different subsets of T lymphocytes such as CD4+ and CD8+ cells.

In the studies of Tournoy and al, PS1 was measured to be partially deficient in mice by western blotting after culture and cellular extraction, compared with normal fibroblasts as a control (ref). Functionally, this was measured by a reduction of the gamma secretase activity which resulted in the accumulation of carboxyterminal APP fragments, which were detected in the brain and in keratinocytes by Western blotting.

The (PS1) knockout mice remain phenotypically normal until about 6 months of age, when the majority develop dermatitis and keratitis, as well as vasculitis and glomerulonephritis.

The knockout (in PS1) mice also developped  after 6 months in 75% of cases a benign epidermal hyperplasia analogous to SK (hyperplastic epithelium, intraepidermal horn cysts, papillomatosis, no architectural or cytological atypia). None of these mice had presented such lesions when under three months old.

In addition, the mice developed severe autoimmune disease with hypergammaglobulinemia and immune complex deposits in the subcutis and in the kidney suggestive of leucocytoclasic vasculitis and positve anti-nuclear factors suggestive of systemic lupus erythematosis. In addition there was an increase in the CD4/CD8 rapport.

The problem with this study is that it links two very prevalent diseases in old age. As with Basal Cell carcinoma, Alzheimer’s disease and SK could well be associated with no causal factors. The second problem is one of definition: is SK in mice equivalent as in humans (same morphologic features, but different kinetics of growth). This interesting study raises the question in general as an associated internal disease associated with SK.

Vitamin A

Elevated dehydroretinol levels, a form of vitamin A is comptable with hyperkeratosis which is present in some SKs. Vitamin A can be detected as retinol or 3-4 dehydroretinol in the normal human skin {Vahlquist, 1980}. 8 specimens of non inflammed acanthotic seborhheic keratosis were compared with normal control skin and no significant difference was found {Rollman, 1981}.  This study done over twenty years ago with no subsequent attempts to determine vitamin A levels should be tried again with HPLC, or unconclusive results should be communicated. An elevation of cutaneousVitamin A, that is  contrary to that described by Rollman levels could offer interesting therapeutical methods; tazarotene has already been attempted. (see treatment)

This advice is for informational purposes  only and does not replace therapeutic judgement done by a skin doctor.


Dr Christophe HSU – dermatologist. Geneva, Switzerland

Category : cause - Modifie le 10.6.2012Category : causes - Modifie le 10.6.2012Category : étiologie - Modifie le 10.6.2012Category : etiology - Modifie le 10.6.2012Category : Kératose séborrhéique - Modifie le 10.6.2012Category : Seborrheic keratosis - Modifie le 10.6.2012Category : Seborrhoeic keratosis - Modifie le 10.6.2012Category : verrue séborrhéique - Modifie le 10.6.2012