Imaging actinic keratosis by high-definition optical coherencetomography. Histomorphologic correlation: a pilot study

Marc A. L. M. Boone1, Sarah Norrenberg1, Gregor B. E. Jemec2 and Veronique Del Marmol1

1Dermatology Department, Hopital Erasme, Universit�e Libre de Bruxelles, Brussels, Belgium; 2Dermatology Department, Roskilde Hospital,

University of Copenhagen, Copenhagen, Denmark

Correspondence: Marc A. L. M. Boone, 128 Kardinaal Sterckxlaan, Meise BE-1860, Belgium, Tel.: 0032 2 270 33 83, Fax: 0032 2 270 33 86,

e-mail: dr.boone@scarlet.be

Abstract: With the continued development of non-invasive

therapies for actinic keratosis such as PDT and immune

therapies, the non-invasive diagnosis and monitoring become

increasingly relevant. High-definition optical coherence

tomography is a high-resolution imaging tool, with micrometre

resolution in both transversal and axial directions, enable to

visualize individual cells up to a depth of around 570 lm filling

the imaging gap between conventional optical coherence

tomography and reflectance confocal microscopy. We sought to

determine the feasibility of detecting and grading of actinic

keratosis by this technique using criteria defined for reflectance

confocal microscopy compared to histology. In this pilot study,

skin lesions of 17 patients with a histologically proven actinic

keratosis were imaged by high-definition optical coherence

tomography just before excision and images analysed

qualitatively. The surrounding normal looking skin has been used

as control group. In lesional skin, dyskeratotic and atypical

keratinocytes could be noticed with this new technique. An

atypical honeycomb pattern in variable degree or a disarranged

epidermal pattern could be observed. A good correlation between

the dimension of atypia and/or disarrangement of the spinous–granular layer on en face images and the histopathological

grading could be demonstrated. Relevant cross-sectional imaging

criteria could be defined for the different histopathological

variants of actinic keratoses. The surrounding skin displayed

features of photodamage. Using features already suggested by

reflectance confocal microscopy, the study implies that high-

definition optical coherence tomography facilitates in vivo

diagnosis of actinic keratosis and allows the grading of different

actinic keratosis lesions for increased clinical utility.

Key words: actinic keratosis – high-definition optical coherence

tomography – histopathological grading – reflectance confocal microscopy

Accepted for publication 4 December 2012

IntroductionActinic keratosis (AK) causes thick, scaly or crusty erythematous

patches of the skin. Sites of predilection are areas of chronic,

high-dose or intermittent sun exposure (face, neck and forearm)

in persons with Fitzpatrick skin phototype I–III (1,2). The etio-

pathogenic concept of AK has been referred to as ‘actinic field

cancerization’. AK represents one stage in the continuum from

subclinical keratinocyte dysplasia to invasive squamous cell

carcinoma (SCC) (3,4). SCC requires extensive and intensive

therapy, while AK usually can be treated with simpler destructive

methods (5). Exact diagnosis is therefore important to segregate

premalignant and malignant lesions prior to treatment. The

exact progression rates are currently unknown (6).

The diagnosis of AK is actually based on clinical examination and

dermoscopy. Conventional dermoscopic criteria have been defined

(7,8). Histological evaluation is usually performed in clinically

indistinct cases and on suspicion of invasive squamous cell carcinoma.

Because of the high prevalence in the population, additional diagnos-

tic options are relevant to pursue, particularly if these are non-invasive as

non-surgical treatment modalities such as photodynamic therapy and

pharmacological immunomodulation are increasingly being used.

Reflectance confocal microscopy (RCM) is a non-invasive imag-

ing technique that allows the visualization of cellular and subcellu-

lar structures of the skin in vivo with near histological resolution

to a depth of approximately 200 lm. In contrast to histology,

which visualizes vertical sections of the skin, RCM obtains hori-

zontal (en face) optical sections in greyscale. RCM can be used as

a diagnostic and monitoring aid in actinic keratosis (4,9–12).RCM diagnostic features of AK have been defined (11).

We have recently introduced high-definition optical coherence

tomography (HD-OCT) (13,14). This is a technique based on the

principle of conventional OCT, with the ability to carry out optical

imaging up to 570 lm deep with a micrometre resolution in both

lateral and axial directions, giving it the potential to visualize indi-

vidual cells (13). HD-OCT is able to provide cross-sectional images

like the conventional OCT and en face images comparable to RCM

potentially giving the method considerable diagnostic utility (15).

The aim of this pilot study is to implement RCM features of AK

with en face HD-OCT images and to compare histological vertical

sections of these lesions with the corresponding cross-sectional and

en face HD-OCT images and on the basis of this suggest relevant

HD-OCT imaging criteria for the in vivo diagnosis and grading of

AK. The surrounding skin was used as control group.

MethodsPatientsSeventeen fair-skinned patients (Fitzpatrick types I–III) with a single

histologically proven AK lesion each located on the face, neck, arm

or trunk were recruited for this pilot study. Signed informed con-

sent was obtained. The group included nine women and eight men

with ages ranging from 44 to 81 years. Clinically, AK presented as

ª 2012 John Wiley & Sons A/SExperimental Dermatology, 2013, 22, 93–97 93

DOI: 10.1111/exd.12074

www.blackwellpublishing.com/EXDOriginal Article

erythematous to brown plaques with a scaly surface. In subclinical

cases, only a rough palpation was suggestive for diagnosis. All

patients gave informed consent to the study (Table S1).

High-definition optical coherence tomographyHigh-definition OCT is based on the principle of conventional

OCT, specifically the ability to carry out optical imaging deep within

highly scattering media such as skin, with micrometre resolution in

both transversal and axial directions, to visualize individual cells

(Skintell®; Agfa Healthcare, Mortsel, Belgium). Instead of a single

pin diode, it uses a two-dimensional, infrared-sensitive (1000–1700 nm) imaging array for light detection. This enables focus

tracking; the focal plane is continuously moved through the sample.

The movements of the focal plane and the reference mirror are syn-

chronized, and the refractive index of the sample is taken into

account. This results in a high lateral resolution of 3 lm at all

depths of the sample. A high axial resolution (3 lm in skin) is

achieved using a broadband thermal light source combined with a

special filter. This technology offers high resolution in all three

dimensions. Moreover, the system is capable of capturing a slice

image and an en face image in real time, as well as fast 3D acquisi-

tion. The spectral sensitivity makes it possible to work in the near

infrared range above 1000 nm. The field of view is 1.8 9 1.5 mm.

The tissue penetration depth goes up to 570 lm, and the total light

power at the tissue is <3.5 mW. The system works in direct contact

with the skin, using an optical matching gel (Skintell® optical gel;

Agfa Healthcare) comparable to ultrasound gel. The interference

signal detected by the 2D imaging sensor is digitized, and its enve-

lope of the interference signal is calculated. The result is transferred

to a computer and displayed using a grey scale or colour palette,

thereby generating an OCT image. Further technical details are

discussed elsewhere (13,14).

Study settingExcision under local anaesthesia was performed immediately after

HD-OCT imaging, and subsequent histopathological analysis

entered the standard histopathological procedure. The HD-OCT

images of histologically proven AK lesions were analysed accord-

ing the criteria for the diagnosis of AK and in RCM. These imag-

ing criteria of AK were implemented on the en face HD-OCT

images. These features are displayed in Table S2. Imaging criteria

of (surrounding) photodamaged skin were implemented on the

HD-OCT images (Table S3) (16).

Histopathological examination of the 17 biopsy specimens was

performed by two histopathologists. Haematoxylin and eosin

(H&E)-stained histological vertical sections of these 17 lesions

were correlated with the corresponding cross-sectional HD-OCT

images to identify other relevant HD-OCT imaging criteria for

both the in vivo diagnosis of AK and the determination of histo-

pathological variants of AK. The surrounding skin has been used

as control group (Tables S1 and S3).

Firstly, AKs were assessed according a three-tiered classification

scheme for keratinocytic intraepidermal neoplasia (KIN) as pro-

posed by Cockerell et al. (4,17). The histological features of KIN-I

on formalin-fixed, paraffin-embedded tissues include keratinocytic

atypia confined to the bottom third of the epidermis with the

basal and suprabasal cells showing some nuclear enlargement and

hyperchromasia. Nuclei maintain their round or oval shape

but show variation in size (Fig. 1). KIN-II is defined as atypia

involving the lower two-thirds of the epidermis. The majority of

clinically diagnosed AK would fall into this category. Abnormal

keratinocytes display more obvious nuclear enlargement, hyper-

chromasia and prominent nucleoli. This stage is further divided

into KIN-IIa (Fig. 2) where the process spares adnexal structures

and KIN-IIb (Fig. 3) where the atypical keratinocytes have

involved adnexa, presenting significant acanthosis or budding of

keratinocytes into the superficial papillary dermis or areas of acan-

tholysis. KIN-III (Fig. 4) represents carcinoma in situ with full-

thickness atypia involving the epidermis and adnexal structures.

Histological features of this three-tiered grading scale were corre-

lated with the corresponding cross-sectional and en face HD-OCT

images. If more than one grade was present in a lesion, the worst

grade has been taken into account (4,17).

Secondly, AKs were assessed according to different histopatho-

logical variants of actinic keratosis. Atrophic (Fig. 1), hypertrophic

(Fig. 2), lichenoid (Fig. 3) and bowenoid actinic keratosis (Fig. 4)

have been described (18–20). Histological vertical sections of these

variants of AK were correlated with the corresponding cross-

sectional HD-OCT images to identify relevant HD-OCT imaging

criteria. Furthermore, the correlation between histological/HD-OCT

variants and histopathological grading was evaluated (Table S4).

ResultsSeventeen patients were included in the study (8 men/9 women;

age range 44–81 years), contributing a total of 17 biopsy-proven

lesions including 2 AKs-KIN III, 9 AKs–KIN-II (2 9 IIa, 4 9 IIb

and 3 9 II unspecified) and 6 AKs-KIN-I. If more than one grade

was observed in a lesion, the worst grade has been taken into

account. KIN-IIa and KIN-IIb often cannot be determined by

Figure 1. Actinic keratosis (AK) graded keratinocytic intraepidermal neoplasia-I(face of patient #5). Cross-sectional (slice) and en face high-definition opticalcoherence tomography images (Z-values indicated: distance in micrometre fromskin surface to location). A substantive hyperkeratosis (red arrow and red brace)with pronounced parakeratosis (green arrow) is observed. Polygonal nucleated cellsare noticed in the disrupted stratum corneum (yellow circle). Atypical honeycombpattern (red circle) The basal cell layer and the spinous cell layer are atrophic(yellow brace). Vasodilation is observed (pink arrow). A histological vertical sectionof this lesion corresponding to the atrophic histological variant of AK is displayed.

94ª 2012 John Wiley & Sons A/S

Experimental Dermatology, 2013, 22, 93–97

Boone et al.

histology because of the shaving method. The anatomical distribu-

tion of the lesions was head (70%), neck (12%), trunk (12%) and

arm (6%; Table S1). The implemented RCM features of AK on

the en face HD-OCT images are summarized in Table S2.

All lesions displayed an adherent hyperkeratotic scale. Parakera-

tosis visualized as the presence of polygonal nucleated cells at the

stratum corneum was seen in all lesions. Single detached cells in

the superficial layer were also constant findings. In early lesions

(KIN-I), an atypical honeycomb pattern was confined to the lower

1/3 of the epidermis (Fig. 1). In fully developed lesions (KIN-II),

atypical keratinocytes provoking an atypical honeycomb pattern

involved mostly the lower two-third of the epidermis (Figs 2 and 3).

Severe architectural disarray of the spinous–granular layer in

which the honeycomb pattern was no longer visible, was only

observed in the two cases of AK graded KIN-III (Fig. 4). Round

nucleated cells sharply delineated with reflective outline were not

observed in early lesions. In the papillary dermis, the presence of

small reflective cells around dark oval, annular or circular spaces

corresponded to perivascular inflammation (Figure S1). In all

lesions, a variably perivascular inflammation was noticed. The

presence of a lace-like material in the papillary dermis correspond-

ing to solar elastosis was observed in only 13 of the 17 cases

(Figure S1). Adnexal involvement could be imaged by HD-OCT

and in six of nine cases correlated well with the histopathological

grading of AK. For the remaining three cases, KIN-IIa or KIN-IIb

could not be determined by histology because of shave biopsy

procedure. Table S3 summarizes the detailed analysis of the en

face HD-OCT images of all 17 patients and correlation to histo-

logical diagnosis and AK-KIN grading.

The histological vertical sections of different variants of AK

were correlated to the corresponding cross-sectional HD-OCT

images in order to identify relevant HD-OCT imaging criteria and

to correlate them with histopathological grading (Table S4). We

could demonstrate a good correlation between histopathology and

the cross-sectional HD-OCT imaging of these variants. Relevant

cross-sectional HD-OCT imaging criteria could be defined for the

different histopathological variants of AK (Table S4). Further-

more, we evaluated the correlation between histological/HD-OCT

variants and histopathological grading. The atrophic variant with

an atrophic stratum malpighi and typically a substantive overlying

hyperparakeratosis was only seen in early lesions graded KIN-I

(Fig. 1). The hypertrophic variant presented a pronounced

parakeratosis, acanthosis and psoriasiform hyperplasia on cross-

sectional HD-OCT images. This histological variant was only

observed in four of the nine AK lesions graded KIN-II (Fig. 2).

An interface inflammatory infiltrate was characteristic for the

lichenoid variant. This variant was present in 30% of AK lesions

graded KIN-II (Fig. 3). The bowenoid variant with at least lower

two-third thickness atypia but without hypertrophic or lichenoid

characteristics was observed in 30% of the AKs graded KIN-IIb.

The bowenoid variant with full-thickness atypia was only observed

in both AKs graded KIN-III (Fig. 4).

Figure 2. Actinic keratosis (AK) graded keratinocytic intraepidermal neoplasia-IIa(arm of patient #12). Cross-sectional (slice) and en face high-definition opticalcoherence tomography images (Z-values indicated). An atypical honeycomb patterninvolving mostly the lower two-third of the epidermis (red circle) is observed. Noadnexal involvement could be noticed. A substantive hyperkeratosis (red brace)containing polygonal nucleated cells (correlate of parakeratosis) is present (yellowcircle). Acanthosis (green brace) and psoriasiform hyperplasia are noticed. Roundnucleated bright cells are present in the spinous cell layer (brown arrow). Ahistological vertical section of this lesion corresponding to the hypertrophichistological variant of AK is displayed.

Figure 3. Actinic keratosis (AK) graded keratinocytic intraepidermal neoplasia-IIb(face of patient #6). Cross-sectional (slice) and en face high-definition opticalcoherence tomography images (Z-values indicated). In this lesion, the atypicalkeratinocytes have involved adnexal structures (hair follicle, green encircled). Thetrajectory of the same hairfollicle (green encircled) can be observed from the skinsurface over the different en face images up to the en face image with Z = 152.Dispersed polygonal nucleated cells (mild parakeratosis) are noticed in thedisrupted stratum corneum (yellow circle). The atypical honeycomb pattern involvesmostly the lower two-third of the epidermis (red circle). Because of the stronginterface inflammatory infiltrate, there is an almost total obliteration of thepapillary rings (violet circle). A histological vertical section of this lesioncorresponding to the lichenoid variant of AK is displayed.

ª 2012 John Wiley & Sons A/SExperimental Dermatology, 2013, 22, 93–97 95

Imaging actinic keratosis

On cross-sectional HD-OCT imaging, the surrounding photo-

damaged skin became less papillary. Features of solar elastosis were

present. Hyperkeratosis and parakeratosis were absent (Figure S2).

On en face HD-OCT imaging, the honeycomb pattern was slightly

irregular (not shown).

DiscussionAlthough no general consensus could be obtained, a reclassification

of AKs has been proposed classifying AKs as cutaneous squamous

cell carcinoma in situ (2,4,21), indicating that lesions may have the

potential to progress into invasive SCC. Lesions develop synchro-

nously as well as consecutively in areas of chronic sun exposure, in a

process commonly referred to as actinic field cancerization (3).

Treatment is evolving from classical destructive therapy such as sur-

gery, to non-invasive medical treatments like photodynamic therapy

or topical treatment options including imiquimod, 5-fluorouracil or

diclofenac in hyaluronic acid (22,23). It is suggested this offers a

better way of coping with the challenge of field cancerization; they

are, however, not a first-line therapy for invasive SCC.

The diagnosis of AK is based on clinical evaluation. Studies have

shown positive predictive values for clinical diagnosis between 81%

and 94% when compared to histopathological examination (24,25).

Nevertheless, biopsy and histopathology should always be performed

if invasive disease is suspected, but a large number of lesions exist in

which biopsy is neither feasible nor efficient (shaving), and where

other means of substantiating the diagnosis are therefore needed.

Classic dermoscopic features of AK have been described (7,8), but

the sensitivity and specificity of these features need to be evaluated.

Reflectance confocal microscopy has also been shown to be a

promising technology with clear features for diagnosis of AK,

which allows diagnosis at an early stage of the lesion. Horn et al.

have shown that the value of RCM in the discrimination of AK

from surrounding sun-damaged skin offered a high sensitivity and

specificity (9,11,26,27). AKs are histopathologically characterized

by a proliferation of atypical keratinocytes, starting at the basal

layer (18,20) indicating that deep penetration into the skin is a

relevant parameter when assessing the utility of a diagnostic

method. OCT may be a relevant method as it provides imaging of

the entire epidermis and often also of the whole dermis (28).

To stratify degrees of epidermal dysplasia, a three-tiered grading

scale has been proposed by Cockerell et al. for AKs that parallel that

used for evaluation of cervical dysplasia (4,17). The localized epider-

mal atypia in AKs reflects a partial disruption of the differentiation

programme, whereas a more complete disruption of differentiation

is associated with SCC in situ. While the KIN grading criteria evalu-

ate the macroscopic and microscopic features of AKs, identification

of genetic and molecular abnormalities associated with these lesions

has provided mechanistic insight into their pathogenesis (4).

One of the aims of this study was to investigate the HD-OCT

correlates of this histopathological three-tiered grading scale on

cross-sectional (Table S4) and en face (Table S3) images obtained

by HD-OCT. This new technology has recently been introduced

by our group as a possible non-invasive technique for morpholog-

ical investigation of tissue with cellular resolution filling the imag-

ing gap between reflectance confocal microscopy and conventional

optical coherence tomography (13).

In this study, we demonstrated that the RCM diagnostic features

of AK could be implemented on the en face HD-OCT images, that

is, comparing horizontal section with horizontal sections

(Table S3). These features could be divided into two groups. The

first group includes features having a good correlation with the his-

topathological AK-KIN grading. These features are atypical honey-

comb and/or disarranged epidermal pattern due to cellular

pleomorphism. These patterns do not appear to have clinical or

dermoscopic correlates and were unique to RCM assessment of AK

and SCC (7). In this study, we could observe the same epidermal

patterns beyond the limitations of RCM such as lack of cross-

sectional imaging. We could find a good correlation between the

dimension of the atypia and/or disarrangement of the spinous–granular layer observed in en face HD-OCT images and the histo-

pathological AK-KIN grading. In early or subclinical lesions (very

often only detected by a rough palpation), a mildly atypical honey-

comb pattern confined to the bottom third of the epidermis could

be noticed. This finding correlated well to KIN-I grading. HD-OCT

was, just like RCM, able to identify subclinical AK by visualization

of cellular and nuclear atypia within the lower spinous cell layers.

The clinical AKs included in this study have an atypical honeycomb

pattern involving the lower two-thirds of the epidermis. This was

the signature for AKs graded KIN-II. A full-thickness disarranged

epidermal pattern was only observed in AKs graded KIN-III. The

atypical and dyskeratotic keratinocytes noticed in histopathology of

AKs represent the round bright nucleated cells observed in RCM

(9,25). In our study, we observed these dyskeratotic and atypical

keratinocytes only in lesions graded KIN-II and KIN-III.

The second group includes features observed in variable degree

in all patients without any correlation with histopathological

(KIN) grading. These features are as follows: hyperkeratotic scale,

polygonal nucleated cells in the stratum corneum (as correlates of

Figure 4. Actinic keratosis (AK) lesions graded keratinocytic intraepidermalneoplasia-III (face of patient #11). Cross-sectional (slice) and en face high-definitionoptical coherence tomography images (Z-values indicated). A complete disarrangedepidermal pattern is observed. This lesion represents a carcinoma in situ with full-thickness atypia involving the epidermis and adnexal structures. Acantholysis isnoticed (white encircled). A histological vertical section of this lesion correspondingto the bowenoid histological variant of AK is displayed.

96ª 2012 John Wiley & Sons A/S

Experimental Dermatology, 2013, 22, 93–97

Boone et al.

parakeratotic cells) and stratum corneum disruption as well as

perivascular inflammation. All lesions demonstrated round blood

vessels in the superficial dermis which is a typical feature of AK in

RCM. Dermoscopic evaluations of the blood vessels have not been

performed. Live video recording of white blood cells moving in

the blood vessels, which is a very useful tool in RCM, is not

possible with HD-OCT (9).

According the histopathological three-tiered grading scale,

adnexal involvement is the discriminating factor between KIN-IIa

and KIN-IIb. Adnexal involvement observed in en face HD-OCT

images correlated well with histopathological AK-KIN grading in six

of nine cases. Due to shave biopsy procedure, it was not possible to

evaluate the remaining three KIN-II cases because of lack of dermal

tissue in the histological specimen. To the best of our knowledge,

this HD-OCT feature has not been described for RCM. HD-OCT

evaluation of adnexal involvement seems to be a promising non-

invasive tool in the clinical grading of AK (Figure S1).

Solar elastosis, represented by lace-like material adjacent to

thickened collagen bundles, could be observed in 13 of the 17

cases. Three of the cases without this lace-like material belonged

to AK graded KIN-I. This seems reasonable because the most

common risk factor for AK is chronic sun exposure. The more

severe the photodamage of the skin the higher the likelihood of

having a clinical AK. HD-OCT evaluation of solar elastosis can be

performed in a very fast way and is a helpful tool in the clinical

evaluation of the photodamaged skin (Figure S1).

Additionally, HD-OCT also provides cross-sectional imaging,

whereas RCM imaging is only possible in the en face mode poten-

tially giving the former method considerable diagnostic utility. We

could demonstrate a good correlation between histopathology and

the cross-sectional HD-OCT imaging of AK. Relevant cross-sec-

tional HD-OCT imaging criteria could be defined for the different

histopathological variants of actinic keratosis (Table S4). Our

study has several limitations including a small sample size and

inclusion of only biopsy-proven lesions. Therefore, it remains

speculative to suppose that the atrophic histopathological and

HD-OCT variant of AK is characteristic for subclinical lesions

graded KIN-I. Cutaneous squamous cell carcinoma typically mani-

fests as spectrum of progressively advanced malignancies, ranging

from a precursor actinic keratosis to squamous cell carcinoma in

situ, invasive and finally metastatic SCC (4). Skin atrophy is one

of the characteristic features of photodamaged skin. Furthermore,

AKs are predominantly found on sun-exposed surfaces. Therefore,

it seems plausible that the atrophic AK variant observed by both

methods is only noticed in early subclinical lesions graded AK-

KIN-I. Figure S2 presents an overview of cross-sectional images

from photodamaged skin via the different grades of AK to invasive

squamous cell carcinoma. We have used the surrounding photo-

damaged skin as a control group. This will make in further studies

calculations of sensitivity and specificity of each feature discussed

in this paper possible and hence the validation of our findings.

In conclusion, our findings indicate that the AK features of RCM

could be implemented on the en face HD-OCT images. Further-

more, a good correlation could be found between histopathological

AK variants and corresponding cross-sectional HD-OCT images.

On the basis of this relevant HD-OCT imaging criteria for the in

vivo diagnosis and grading of AK could be presented. Thereby HD-

OCT is also able to identify subclinical AK by visualization of cellu-

lar and nuclear atypia within the lower third of the spinous cell

layer. In this way, HD-OCT could also be used to survey sun-dam-

aged skin in the setting of field cancerization potentially increasing

diagnostic accuracy compared to clinical evaluation alone.

Author contributionM.A.L.M. Boone designed the research study, performed the study (acqui-

sition, analysis and interpretation of data) and wrote the paper. S. Norren-

berg performed the histopathological examination and critically revised the

paper. I. Neetens performed the histopathological examination. G.B.E. Je-

mec and V. Del Marmol critically revised the paper.

Conflict of interestsThe authors have declared no conflicting interests.

References1 Heaphy M R Jr, Ackerman A B. J Am Acad Der-

matol 2000: 43: 138–150.2 Ackerman A B. Arch Dermatol 2003: 139: 1216

–1217.3 Braakhuis B J, Tabor M P, Kummer J A et al.

Cancer Res 2003: 63: 1727–1730.4 Ratushny V, Gober M D, Hick R et al. J Clin

Invest 2012: 122: 464–472.5 Mogensen M, Thrane L, Jørgensen T M et al. J

Biophotonics 2009: 2: 442–451.6 Ulrich M, Gonzalez S, Lange-Asschenfeldt B

et al. J Eur Acad Derm 2011: 25: 276–284.7 Zalaudek I, Argenziano G, Leinweber B et al. Br

J Dermatol 2004: 150: 1112–1116.8 Zalaudek I, Giacomel J, Argenziano G et al. Br J

Dermatol 2006: 155: 951–956.9 Risphon A, Kim N, Scope A et al. Arch Dermatol

2009: 145: 766–772.10 Ulrich M, Krueger-Corcoran D, Roewert-Huber J

et al. Dermatology 2010: 220: 15–24.11 Richtig E, Ahlgrimm-Siess V, Koller S et al. J Eur

Acad Derm 2010: 24: 290–298.12 Wurm E M T, Claudia E S, Curchin M B B S

et al. J Am Acad Dermatol 2012: 66: 463–473.13 Boone M A L M, Jemec G B E, Del Marmol V.

Exp Dermatol 2012: 10: 740–74414 Boone M A L M, Norrenberg S, Jemec G B E

et al. Br J Dermatol 2012: 167: 855–864.

15 Mogensen M, Jemec G B. Dermatol Surg 2007:33: 1158–1174.

16 Longo C, Casari A, De Pace B et al. Skin ResTechnol 2012. June 7. doi: 10.1111/j.1600-0846.2012.00623.x.

17 Cockerell C J, Wharton J R. J Drugs Dermatol2005: 4: 462–467.

18 Goldberg L H, Joseph A K, Tschen J A. Int J Der-matol 1994: 33: 341–345.

19 Prieto V G, Casal M, McNutt N S. J Cutan Pathol1993: 20: 143–147.

20 Billano R A, Little W P. J Am Acad Dermatol1982: 7: 484–489.

21 R€owert-Huber J, Patel M J, Forschner T et al. BrJ Dermatol 2007: 156 (Suppl 3): 8–12.

22 Stockfleth E, Kerl H. Eur J Dermatol 2006: 16:599–606.

23 Ortonne J P, Gupta G, Ortonne N et al. Exp Der-matol 2010: 19: 641–647.

24 Venna S S, Lee D, Stadecker M J et al. ArchDermatol 2005: 41: 507–509.

25 Thompson S C, Jolley D, Marks R. N Engl J Med1993: 329: 1147–1151.

26 Horn M, Gerger A, Ahlgrimm-Siess V et al. Der-matol Surg 2008: 34: 620–625.

27 Koehler M J, Speicher M, Lange-Asschenfeldt Set al. Exp Dermatol 2011: 20: 589–594.

28 Mogensen M, Morsy H A, Thrane L et al. Der-matology 2008: 217: 14–20.

Supporting InformationAdditional Supporting Information may be found inthe online version of this article:Figure S1. Actinic keratosis graded keratinocytic in-

traepidermal neoplasia-IIb (face of patient #17).Figure S2. Actinic keratosis: from photodamaged

skin to invasive squamous cell carcinoma.Table S1. Patients characteristics and histological

diagnosis with grading and variants of actinic keratosis.Table S2. Reflectance confocal microscopy features

of actinic keratosis applied on en face images of high-definition optical coherence tomography.Table S3. Detailed analysis of implemented reflectance

confocal microscopy diagnostic features of actinic kera-tosis (AK) on en face high-definition optical coherencetomography images of all 17 patients and correlationto histopathological AK-keratinocytic intraepidermalneoplasia grading.Table S4. Histological vertical sections of the

observed variants of actinic keratosis were correlatedwith the corresponding cross-sectional high-definitionoptical coherence tomography images and correlationto histopathological AK-keratinocytic intraepidermalneoplasia grading in all 17 patients.

ª 2012 John Wiley & Sons A/SExperimental Dermatology, 2013, 22, 93–97 97

Imaging actinic keratosis