C L I N I C A L A N D LA B O R A T O R Y I N V E S T I G A T I O N S DOI 10.1111/j.1365-2133.2006.07426.

Dermoscopy of facial nonpigmented actinic keratosis

I. Zalaudek, J. Giacomel,* G. Argenziano, R. Hofmann-Wellenhof, T. Micantonio, A. Di Stefani, M. Oliviero,§
H. Rabinovitz,§ H.P. Soyer and K. Peris
Department of Dermatology, Medical University of Graz, Auenbruggerplatz 8, 8036 Graz, Austria
*Mends St Medical Centre, South Perth, Western Australia
Department of Dermatology, Second University of Naples, Naples, Italy
Department of Dermatology, University of L’Aquila, L’Aquila, Italy
§Skin and Cancer Associates and Department of Dermatology, School of Medicine, University of Miami, Miami, FL, U.S.A.

Summary

Correspondence Background The accuracy of clinical diagnosis of nonpigmented, facial actinic kera-
Iris Zalaudek. tosis (AK) is often suboptimal, even for experienced clinicians.
E-mail: iris.zalaudek@meduni-graz.at Objectives To investigate the dermoscopic features of nonpigmented AK located on
the head/neck that may assist the clinical diagnosis.
Accepted for publication
17 April 2006 Methods Forty-one nonpigmented AKs on facial sites were examined by dermosco-
py for any consistent underlying features. Lesions were gathered from skin cancer
Key words centres in Australia, Austria, Italy and the U.S.A. All cases were diagnosed histo-
actinic keratosis, dermoscopy, diagnosis, skin pathologically.
cancer Results Four essential dermoscopic features were observed in facial AK: (i) ery-
thema, revealing a marked pink-to-red ‘pseudonetwork’ surrounding the hair fol-
Conflicts of interest
None declared. licles (95%); (ii) white-to-yellow surface scale (85%); (iii) fine, linear-wavy
vessels surrounding the hair follicles (81%); and (vi) hair follicle openings filled
with yellowish keratotic plugs (66%) and/or surrounded by a white halo
(100%). These features combined, in 95% of cases, to produce a peculiar ‘straw-
berry’ appearance.
Conclusions A dermoscopic model of ‘strawberry’ pattern is presented, which may
prove helpful in the in vivo diagnosis of nonpigmented, facial AK. A limitation of
this study is the lack of testing of the specificity of the described dermoscopic
criteria in differentiating nonpigmented AKs from other nonpigmented skin
lesions at this site.

Actinic keratoses (AKs), or solar keratoses, are neoplasms
within the continuum of squamous cell carcinoma (SCC).
They are considered by some to be premalignant and by oth-
ers to be incipient SCCs. Although traditionally diagnosed by
clinical appearance alone, AK may be difficult to differentiate
from other skin lesions such as Bowen’s disease (BD), plaque-
type or superficial basal cell carcinoma (sBCC) and initial
seborrhoeic keratosis (SK).1,2
Dermoscopy is an in vivo technique that has gained popular-
ity in recent years as an aid to the clinical diagnosis of many
pigmented skin lesions (PSL) and nonpigmented skin lesions
(non-PSL). The improvement in diagnostic accuracy of PSL
has been well established,3,4 but recent work has also suggest-
ed an improved diagnosis of lesions that are typically nonpig-
mented, using this technique.2,5–8 This is because dermoscopy
allows the uncovering of key structures in non-PSL, including
vascular patterns, that are usually not visible to the naked eye.
While dermoscopic criteria have been formulated to
improve diagnostic accuracy for a variety of skin lesions, the
dermoscopic characteristics of nonpigmented facial AK have
yet to be described in detail. Consequently, there is currently
no dermoscopic model for the diagnosis of such AKs.
Materials and methods
Clinical and dermoscopic features were studied in 41 cases of
nonpigmented AK randomly selected from skin cancer centres
in Perth (Australia), Graz (Austria), L’Aquila and Naples
(Italy), and Miami (FL, U.S.A.) between February and October
2005. Clinical data were recorded for each patient, namely:
(i) age, (ii) sex, (iii) location on the head/neck, (iv) clinical
diagnosis and (v) histopathological diagnosis.
Before therapeutic interventions according to established
protocols, clinical and dermoscopic images of each lesion

 2006 The Authors

Journal Compilation  2006 British Association of Dermatologists • British Journal of Dermatology 2006 155, pp951–956 951
952 Dermoscopy of facial nonpigmented AK, I. Zalaudek et al.

were obtained using a Heine Delta 20 hand-held dermato-

scope (Heine Optotechnik, Herrsching, Germany) coupled to
a Nikon Coolpix 4500 digital camera (Nikon Corporation,
Tokyo, Japan), or a Dermlite Foto lens without a glass plate
(3Gen, LLC, Dana Point, CA, U.S.A.). All instruments operate
generally at 10-fold magnification.
In all cases punch or shave biopsy was undertaken for histo-
pathological confirmation, with specimens being routinely
stained by haematoxylin and eosin.
All lesions were examined by two of us (J.G., I.Z.) for the pres-
ence of the following dermoscopic features: (i) surface scales,
(ii) vascular structures and (iii) other dermoscopic features.
Table 1 Patient demographics and dermoscopic features seen in nonpigmented actinic keratosis (AK) on facial sites
Results
General results and patient demographics
We examined 41 nonpigmented AKs in 32 patients, with a
remarkable male predominance: 24 (75%) men and eight
(25%) women. Patient age ranged from 48 to 91 years (mean
69).
Most facial AKs were located on the forehead (34%), fol-
lowed by the temple (24%), cheek (17%), scalp (12%),
nose (10%) and eyebrow (2%). The diameter of the lesions
varied from 4 to 13Æ5 mm (median 6). All AKs revealed

Age Clinical Follicular plugs Erythema/red Linear, wavy Coiled Dotted

Patient Sex (years) diagnosis Scales with whitish halo pseudonetwork vessels vessels vessels
1 M 52 AKa + + + ) ) )
2 M 48 AK,a sBCC ) + + + ) )
3 M 69 AK,a BD + ) + + + )
4 M 73 AK + + + + ) )
5 M 84 AK + + + + ) )
6 F 78 AK,a BD + + + + ) )
7 M 77 sBCC + + + + ) )
8 M 91 AK, sBCC + ) + + ) )
9 F 54 AK + + + + ) )
10 M 70 AK,a BD + + + + ) )
AK, BD + + + + ) )
11 M 66 AKa + + + + ) +
12 F 79 AKa + + + + ) )
13 M 91 AKa + + + + + )
14 M 52 AK,a BD + + + + ) )
15 M 75 AK,a BD + + + ) + )
AK,a BD + + + + ) )
AK,a BD + + + ) + )
16 M 88 AKa ) ) + + ) )
17 M 74 AK, BD ) ) + + ) )
AK, BD + + + ) + )
AK + + + ) + +
AKa + + + + ) )
18 M 59 AK ) ) + + ) )
19 M 62 AK,a BD + + + + ) )
20 M 71 AK, sBCC + + + + ) )
21 M 74 AK + + + + ) +
22 M 60 AK,a BD + + + + + )
AK,a BD + ) + + ) )
23 F 55 AKa + + + + ) )
24 M 63 AK + ) ) + ) )
25 F 74 AK, sBCC + ) + + ) )
26 F 70 AK + + + + ) )
27 M 76 AK, BD + ) + + ) )
28 F 63 AK,a BD + + + + ) )
29 F 64 AK ) ) + ) + )
30 M 57 AK + ) + + ) )
31 M 72 AK + + + ) ) )
32 M 84 AK + ) + + + )
AK ) ) ) ) + )
AK + ) + + ) )
Total: 41 lesions 35 (85%) 27 (66%) 39 (95%) 33 (81%) 10 (24%) 3 (7%)
a
Histopathological diagnosis of hyperkeratotic AK. BD, Bowen’s disease; sBCC, superficial basal cell carcinoma.

 2006 The Authors

Journal Compilation  2006 British Association of Dermatologists • British Journal of Dermatology 2006 155, pp951–956

clinical features of a macule, papule or plaque surmounted
by an adherent surface scale, with some degree of erythema
ranging from pale pink to dark red. In all but one case, the
clinical differential diagnosis included AK, but in approxi-
mately half of the lesions (46%) BD and/or sBCC were addi-
tionally considered (Table 1). In all cases the diagnosis of
AK was confirmed by histopathology and, in 19 lesions, his-
topathological examination revealed a hyperkeratotic type of
AK.
Specific results: dermoscopic features of nonpigmented
actinic keratoses on facial sites
The dermoscopic features seen in our series of 41 facial AKs
are shown in Table 1. The most striking pattern in 39 of
41 (95%) facial AKs was a ‘background’ erythema that was
Fig 1. Dermoscopy of an actinic keratosis located (A) on the cheek of
a 60-year-old man (patient 22) and (B) on the forehead of a 75-year-
old man (patient 15). The former lesion displays a stereotypical
‘strawberry’ appearance, with white-to-yellow follicular keratotic
plugs (circle) surrounded by a whitish halo, and background
erythema/red pseudonetwork. In addition, wavy vessels are seen
encircling the whitish halo of the follicles. By contrast, the latter
lesion reveals, despite the ‘strawberry’ pattern, coiled vessels (circle)
surrounding the keratin-filled hair follicle openings. Scale
graduations ¼ 1 mm; original magnification · 10.
 2006 The Authors
Journal Compilation  2006 British Association of Dermatologists • British Journal of Dermatology 2006 155, pp951–956
Dermoscopy of facial nonpigmented AK, I. Zalaudek et al. 953
accentuated dermoscopically by a peculiar reddish ‘pseudonet-
work’. Dermoscopically, the latter consisted of unfocused,
large-calibre telangiectatic vessels located between the hair fol-
licles. This erythema/red pseudonetwork contrasted with
prominent hair follicle openings, which were surrounded by a
white halo in all lesions.
In 66% of AKs, particularly when located on the nose and/
or of the hyperkeratotic type, we found yellowish keratotic
plugs within the hair follicles, resulting in a whitish-yellowish
‘targetoid-like’ appearance (Fig. 1). As this overall, composite
appearance of reddish pseudonetwork and hair follicles was
reminiscent of the surface of a strawberry (Fig. 1), we have
coined the phrase ‘strawberry’ pattern to describe it. In fact,
95% of our facial AKs exhibited this ‘strawberry’ appearance.
Apart from the erythema/red pseudonetwork structures, we
found fine, focused linear ‘wavy’ vessels surrounding the hair
follicles in more than 80% of facial AKs (Fig. 1A). Further
vascular structures, but much less frequent, were small,
Fig 2. (A) Clinically, the lesion is seen as a reddish scaly plaque
located on the forehead of a 72-year-old man (patient 31). (B)
Dermoscopically, this actinic keratosis displays a ‘strawberry’ pattern
in the absence of other specific vascular structures. The strawberry
appearance is typified by a reddish pseudonetwork intermingled with
hair follicle openings, which are surrounded by whitish halos
(original magnification · 10).
954 Dermoscopy of facial nonpigmented AK, I. Zalaudek et al.
coiled vessels (Fig. 1B) and dotted vessels surrounding the
hair follicles, seen in 10 and three lesions, respectively. The
coiled vessels were the only type present in five of 10 lesions,
and in four lesions they were associated with linear wavy ves-
sels. Only two lesions lacked any specific vascular type, but
they none the less displayed the typical ‘strawberry’ appear-
ance of erythema/red pseudonetwork intermingled with whi-
tish or targetoid hair follicles (Fig. 2). Surface scales,
dermoscopically characterized by structureless whitish areas,
were present in 35 facial AKs (85%).
Discussion
AKs are common dysplastic epidermal lesions that usually pre-
sent as nonpigmented, rough, scaly lesions on sun-exposed
areas of the skin. These latter sites include particularly the dor-
sal hands, forearms, and head and neck. The major risk factors
for the development of AK are a combination of chronic,
cumulative exposure to ultraviolet radiation and fair skin
A
Fig 3. (A) Clinical image of a flat, histopathologically proven basal
cell carcinoma, located on the forehead of a 68-year-old man.
Clinically, a reddish plaque can be seen, while dermoscopy of the
lesion (B) exhibits arborizing vessels, within a red to white
structureless background. Superficial ulceration is also apparent, and
there is an absence of follicular keratotic plugs and/or openings
(original magnification · 10).
Journal Compilation  2006 British Association of Dermatologists • British Journal of Dermatology 2006 155, pp951–956
phenotype (particularly skin phototypes I and II). AKs are sig-
nificant in that they have the potential to progress, if left
untreated, to full-thickness in situ or invasive SCC.9 Indeed, AK
and SCC are considered to represent extreme ends of the same
disease spectrum,9 evolving from intraepidermal containment
to dermal invasion. Although the risk of progression of an
individual AK to invasive SCC is unclear, estimates vary
between 0Æ1% and 10%.10 Moreover, patients with multiple
AKs (i.e. more than 10) may have a 14% cumulative probabil-
ity of developing SCC, either within the AK or de novo, within
5 years.11 Therefore, patients with AK should be treated using
various modalities, and periodic full cutaneous examinations
are recommended.
A problem facing the clinician is that AKs cannot always be
distinguished from other non-PSL on clinical grounds alone.2
These latter lesions include particularly plaque-like BCC and
BD (Figs 3 and 4). Given that skin cancer can cause progres-
sive local tissue destruction or may metastasize, early diagnosis
is crucial to reduce morbidity and mortality rates. On the
A
Fig 4. (A) Clinical image of histopathologically proven Bowen’s
disease, clinically presenting as a reddish scaly plaque located on the
forehead of a 62-year-old man. (B) Dermoscopy of the lesion reveals
glomerular vessels arranged in clusters (circle), and surface scales
characterized by a whitish structureless coloration (original
magnification · 10).
 2006 The Authors

A
B encircle the hair follicles as single and uniform units, which is
Fig 5. (A) Clinical image of telangiectasias seen on the cheek of a 65-
year-old man with chronic sun-damaged, atrophic skin. (B)
Dermoscopically, the vessels of the dermal vascular plexus are
regularly branched and tree like, appearing slightly unfocused with a
pinkish hue. A blurred, mesh-like pseudonetwork is not seen (original
magnification · 10).
other hand, the overtreatment of false positive AKs (i.e. AKs
misdiagnosed as skin cancer) should be avoided.
Dermoscopy is primarily designated as an adjunct to the
clinical diagnosis of PSL, but has also been discovered in
recent years to assist the clinical examination of many non-
PSL. It achieves this, in part, by allowing the visualization of
specific vascular patterns which are mostly clinically invisible.
Specific morphological types of vessels have been described in
the great majority of non-PSL included in the differential diag-
nosis of nonpigmented AK.1,2,7,8 Yet, to date, the dermoscopic
patterns of facial AK have not been formally investigated.
In our pilot study we observed, in the vast majority (95%)
of facial AKs, repetitive dermoscopic features of a background
erythema that, surrounding follicles with whitish halos
(100%) or a whitish-yellowish ‘targetoid’ appearance (66%),
resembled a blurred reddish pseudonetwork structure. Because
this composite appearance, observed in 95% of our AKs, was
reminiscent of the surface of a strawberry, we suggest
employing the simple term ‘strawberry’ pattern to describe it.
 2006 The Authors
Journal Compilation  2006 British Association of Dermatologists • British Journal of Dermatology 2006 155, pp951–956
Dermoscopy of facial nonpigmented AK, I. Zalaudek et al. 955
To our knowledge, these particular features have not been
previously described in other non-PSL at this site. The reddish
pseudonetwork of AK differs from the telangiectasias fre-
quently seen in elderly patients with chronic sun-damaged or
atrophic skin and/or in patients with telangiectatic rosacea. In
these latter conditions, the vascular plexus is typically less
blurred and is composed of large-stem vessels that communi-
cate with each other (Fig. 5).6
Furthermore, we found additional specific vascular patterns
of linear, wavy vessels, mainly encircling the whitish halo
around the hair follicles, in more than 80% of lesions. These
peculiar linear, wavy vessels of facial AK clearly differ in mor-
phology from the arborizing, short fine telangiectatic, and
regular hairpin vessels that are characteristic of BCC, sBCC and
SK, respectively.2,6,8,12 Furthermore, wavy vessels typically
in contrast to the irregularly sized and distributed linear-irre-
gular vessels that can be seen in amelanotic/hypomelanotic
melanoma, areas of regression in melanoma, or invasive
SCC.5–8
In addition, the small coiled and dotted vessels seen in our
facial AKs should be distinguished from the glomerular vessels
that are typically seen in BD.1,8 While coiled vessels are typic-
ally located around the hair follicle openings, the latter are
usually distributed in clusters and are larger in size (being
‘globular’), their name deriving from their fanciful resem-
blance to the renal glomerulus (Fig. 4).
A limitation of our study is that we did not assess the spe-
cificity of the described dermoscopic criteria, in differentiating
AK from other non-PSL. This is especially the case for the
coiled vessels, which might be difficult to differentiate from
the glomerular vessels in BD, due to their convoluted mor-
phology. It was noteworthy that only five of 15 lesions with a
clinical differential diagnosis of BD revealed coiled vessels by
dermoscopy, although the additional presence of a reddish
pseudonetwork in all five lesions allowed the correct diagnosis
of AK to be made. However, this number is too small to draw
definitive conclusions regarding the validity of this combined
pattern. We therefore suggest performing biopsy and histo-
pathological examination of those lesions for which the clin-
ical and/or dermoscopic criteria are uncertain.
In conclusion, our descriptive, pilot study suggests that
dermoscopy may be considered a promising tool for improv-
ing the diagnostic accuracy of nonpigmented AKs on facial
sites, by revealing a typical ‘strawberry’ pattern and a scaly
surface.
References
1 Zalaudek I, Argenziano G, Leinweber B et al. Dermoscopy of
Bowen’s disease. Br J Dermatol 2004; 150:1112–16.
2 Giacomel J, Zalaudek I. Dermoscopy of superficial basal cell carci-
noma. Dermatol Surg 2005; 31:1710–13.
3 Argenziano G, Soyer HP, Chimenti S et al. Dermoscopy of pigment-
ed skin lesions: results of a consensus meeting via the internet.
J Am Acad Dermatol 2003; 48:679–93.
956 Dermoscopy of facial nonpigmented AK, I. Zalaudek et al.
4 Argenziano G, Soyer HP. Dermoscopy of pigmented skin
lesions—a valuable tool for early diagnosis of melanoma. Lancet On-
col 2001; 2:443–9.
5 Zalaudek I, Argenziano G, Kerl H et al. Amelanotic/hypomelanotic
melanoma—is dermatoscopy useful for diagnosis? J Dtsch Dermatol
Ges 2003; 1:369–73.
6 Kreusch JF. Vascular patterns in skin tumors. Clin Dermatol 2002;
20:248–54.
7 Argenziano G, Zalaudek I, Corona R et al. Vascular structures in
skin tumors: a dermoscopy study. Arch Dermatol 2004; 140:1485–9.
8 Zalaudek I. Dermoscopy subpatterns of nonpigmented skin tumors.
Arch Dermatol 2005; 141:532.
Journal Compilation  2006 British Association of Dermatologists • British Journal of Dermatology 2006 155, pp951–956
9 Cockerell CJ. Histopathology of incipient intraepidermal squamous
cell carcinoma (‘actinic keratosis’). J Am Acad Dermatol 2000;
42:S11–17.
10 Sober AJ, Haluska FG. Atlas of Clinical Oncology: Skin Cancer. Hamilton,
ON: B.C. Decker Inc., 2001.
11 Salasche SJ. Epidemiology of actinic keratoses and squamous cell
carcinoma. J Am Acad Dermatol 2000; 42:S4–7.
12 Menzies SW, Westerhoff K, Rabinovitz H et al. Surface microscopy
of pigmented basal cell carcinoma. Arch Dermatol 2000; 136:1012–
16.
 2006 The Authors