bjd dermatopathology british journal of dermatology

DERMATOPATHOLOGY

BJD

British Journal of Dermatology

Actinic keratosis in the en-face and slice imaging modeof high-definition optical coherence tomography andcomparison with histologyT. Maier,1 M. Braun-Falco,1 R.P. Laubender,2 T. Ruzicka1 and C. Berking1

1Department of Dermatology and Allergology, Ludwig-Maximilian University of Munich, Frauenlobstr. 9–11, D-80337 Munich, Germany2Department of Biometry and Epidemiology, Institute of Medical Informatics, Ludwig-Maximilian University of Munich, Campus Grosshadern, Munich, Germany

CorrespondenceTanja Maier.

E-mail: [email protected]

Accepted for publication31 July 2012

Funding sourcesThis work was supported by the Curd-Bohnewand-

Fonds of the University of Munich (to T.M.), by

the Matthias Lackas Foundation and the Dr

Helmut Legerlotz Foundation (to C.B.).

Conflicts of interestThe HD-OCT Skintell� device used in this study

was provided by Agfa HealthCare GmbH.

Dr Maier Served as lectures for Agfa Healthcare

GmbH.

DedicationThis work is dedicated to Prof. Hans-Christian

Korting who always gave essential support and has

recently passed away.

DOI 10.1111/j.1365-2133.2012.11202.x

Summary

Background Optical coherence tomography (OCT) allows real-time, in vivo examina-tion of nonmelanoma skin cancer. An innovative high-definition (HD)-OCT witha horizontal (en-face) and vertical (slice) imaging mode offers additional informa-tion in the diagnosis of actinic keratosis (AK) and may potentially replaceinvasive diagnostic biopsies.Objectives To define the characteristic morphological features of AK by using HD-OCT in the two imaging modes compared with histopathology as gold standard.Methods In total, 20 AKs were examined by HD-OCT in the en-face and slice imag-ing modes and characteristic features were described and evaluated in comparisonwith the histopathological findings. Furthermore, the HD-OCT images of a sub-group of AKs were compared with those of the clinically normal adjacent skin.Results The preoperative in vivo diagnostics showed the following features in theen-face imaging mode of HD-OCT: disruption of stratum corneum, architectural dis-array, cellular ⁄nuclear polymorphism in the stratum granulosum ⁄stratum spino-sum, and bright irregular bundles in the superficial dermis. In the vertical sliceimaging mode the following characteristics were found: irregular entrance signal,destruction of layering, white streaks and dots, and grey areas. In contrast, the clini-cally healthy adjacent skin showed mainly a regular epidermal ‘honeycomb’ patternin the en-face mode and distinct layering of the skin in the slice mode.Conclusions HD-OCT with both the en-face and slice imaging modes offers additionalinformation in the diagnosis of AK compared with conventional OCT and mightenhance the possibility of the noninvasive diagnosis of AK prior to treatmentprocedures and possibly in the monitoring of noninvasive treatment strategies.

The incidence of nonmelanoma skin cancer including actinic

keratosis (AK) is rising worldwide. As AKs are precursor lesions

of potentially metastasizing squamous cell carcinoma (SCC) the

increasing prevalence of AK is alarming. It is assumed that about

10% (6–16%) of patients with AK will develop over decades an

SCC.1,2 Epidemiological data show a high prevalence of AK in

fair skin types and in countries with high exposure to ultraviolet

radiation. The highest incidence of AK has been documented in

Australia with 55% in men and 37% in women.3 According to

studies in the U.K. the prevalence of AK was 15% in men and

6% in women, and in the population over the age of 70 years it

increased to 34% in males and 18% in females.4 In the U.S.A.

the incidence rate of AK is between 11% and 26%.2 Addition-

ally, the growing subgroup of immunosuppressed patients

worldwide is known to develop a high rate of frequently aggres-

sive AK,5 who will need regular monitoring and treatment.

AKs are mainly diagnosed clinically although histological

confirmation is recommended to exclude an invasive SCC. As

noninvasive treatments of AK such as topical immune response

modifiers, topical chemotherapy, chemical peeling or photo-

dynamic therapy are becoming more and more popular, the

invasive biopsy for the diagnosis and treatment follow-up is

often unfavourable.

To date there is no commonly accepted classification of AK

which accounts for the grade of invasiveness of the lesion.

According to the Guidelines for Dermatology in Europe AK

can be divided clinically into different subtypes (keratotic,

atrophic, cornu cutaneum, verrucous, pigmented, lichenoid)

� 2012 The Authors

BJD � 2012 British Association of Dermatologists 2012 1

but without a distinct parameter for the prediction of malig-

nant transformation.6 In 2006 a review article tried to identify

clinical features for the recognition of aggressive AK.7 As risk

factors for malignancy the following parameters were pro-

posed: induration ⁄ inflammation, diameter > 1 cm, rapid

enlargement, bleeding, erythema and ulceration. In certain

cases, the use of dermoscopy assists in the identification of

pigmented AK. The strawberry sign has been described as a

characteristic pattern of AK, meaning a red pseudonetwork of

vessels around prominent hair follicles with a white halo in

dermoscopy.8,9 As AKs are the most common premalignant le-

sions worldwide a reliable diagnostic evaluation would be

preferable as is established in pigmented lesions (ABCD rule).

In this context the new imaging techniques such as reflec-

tance confocal microscopy (RCM) and optical coherence

tomography (OCT) offer a noninvasive option for diagnostics

and therapeutic monitoring. RCM has already been studied for

the evaluation of melanoma10,11 and nonmelanoma skin can-

cer such as AK,8 and also OCT has been performed in differ-

ent types of epithelial skin cancer including AK.12–16

Nevertheless, the diagnosis of AK with RCM is time-consum-

ing especially when screening multiple lesions. With respect

to conventional OCT the visualization of AK is still insufficient

because of the relatively low lateral and axial resolution of the

available OCT devices.

In this study, an innovative high-definition (HD)-OCT

device with an improved resolution of about 3 lm in all

dimensions and the possibility of vertical (slice-mode) and

horizontal (en-face mode) imaging similar to the horizontal

RCM scanning mode was used to evaluate noninvasively the

clinical diagnosis of AK in correlation with the histological

diagnosis.

Subjects and methods

Participants

We investigated 20 lesions clinically suspicious for AK prior

to the surgical intervention. The study was performed at the

Department of Dermatology, Ludwig-Maximilian University of

Munich after obtaining written informed consent from each

patient. The study population consisted of nine men and four

women with one or more lesions and an age range of 50–

82 years (Table 1). In a subgroup of five lesions (lesions 10,

16–19) the clinically healthy adjacent skin was additionally

imaged by HD-OCT. After obtaining OCT images of the

suspected lesions a punch or shave biopsy was performed

preferentially from the centre of the lesions followed by the

histopathological examination. There were no biopsies taken

from the adjacent normal skin. The study was approved by

the local ethics committee of the Ludwig-Maximilian

University of Munich and followed the principles expressed in

the Declaration of Helsinki.

High-definition optical coherence tomography

A commercially available full-field HD-OCT system (Skintell�;

Agfa HealthCare, Mortsel, Belgium) was used for OCT

Table 1 Clinical patient data and histopathological characteristics of the 20 actinic keratoses (AKs) included in the high-definition optical

coherence tomography study

Lesion

Patient age

(years) ⁄ sex Site

Histopathological

type of AK Parakeratosis

Destruction of

epidermal structure

Pleomorphic

keratinocytes

Inflammatory

infiltrate

1 78 ⁄M Forehead Bowenoid + + + +2 78 ⁄M Eyebrow right Atrophic + + + +

3 71 ⁄M Cheek right Unclassified + + + +4 71 ⁄M Temple left Unclassified + + + +

5 71 ⁄M Cheek left Atrophic + + + +6 78 ⁄M Cheek left Hypertrophic + + + +

7 54 ⁄F Cheek right Unclassified + + + +8 54 ⁄F Infraorbital right Initial ) + + )9 74 ⁄M Forehead Acantholytic + + + +

10 50 ⁄F Eyebrow right Atrophic + + + +

11 74 ⁄M Head left Atrophic + + + +12 78 ⁄M Head Unclassified + + + +

13 82 ⁄F Cheek left Atrophic + + + +14 74 ⁄M Head right Unclassified + + + )15 74 ⁄M Occipital right Unclassified + + + )16 65 ⁄F Cheek right Atrophic + + + +

17 65 ⁄F Temple left Initial ) ) + )18 55 ⁄M Nose Hypertrophic + + + +

19 55 ⁄M Cheek left Unclassified + + + +20 75 ⁄M Forehead left Unclassified + + + +

+, present; ), absent.


BJD � 2012 British Association of Dermatologists 2012

2 Actinic keratosis in high-definition optical coherence tomography, T. Maier et al.

imaging. It works on the principle of a ‘time domain’ OCT

system including dynamic focus tracking which is a synchro-

nized motion of the imaging lens system and the reference

optical system. In the Skintell� (Agfa HealthCare) system a

halogen lamp with a Gaussian filter and an ultrahigh band-

width centred at 1300 nm (infrared radiation) serves as the

light source which allows a high-depth resolution and no lat-

eral cross-talk (speckling). Two different modes are available

in this system: real-time b-scan and en-face which allows a fast

capture of a three-dimensional tomogram. According to the

producer’s instructions, Skintell� allows a resolution of 3 lm

in three dimensions. The penetration depth in skin is about

750 lm and the field of view in the en-face mode is

1Æ8 · 1Æ5 mm.

As coupling medium an optical gel (Skintell� optical gel;

Agfa HealthCare) is used and the probe is applied directly on

to the skin.

The clinically suspicious lesions were systematically evalu-

ated by HD-OCT in the slice and en-face mode and images

were recorded. The images were evaluated starting from the

stratum corneum (SC) through the epidermis and upper der-

mis as far as it was displayed in the images. The HD-OCT

probe was preferentially placed in the centre of the lesion,

where the consecutive biopsy was taken later on. The imprint

of the HD-OCT probe was visible right after the imaging and

was marked with a pen for the consecutive biopsy. In a sub-

group of patients adjacent clinically normal skin was addition-

ally imaged by HD-OCT for comparison with the lesional

skin. The lesions were examined by an experienced investiga-

tor (T.M.) prior to excision. Characteristic features were evalu-

ated in relation to the established histological criteria of

AK (Table 1).

Histological evaluation

Biopsies were taken of all 20 lesions clinically suspicious for

AK and tissue sections were prepared for conventional hist-

ology. As there was no exact labelling of the HD-OCT position

possible within the lesion the evaluation and consecutive bi-

opsy were always done in the centre of each lesion. The histo-

logical evaluation was performed by a board-certified

dermatopathologist (M.B.-F.).

Results

En-face mode of high-definition optical coherence

tomography in actinic keratosis

The evaluation of the documented images of the AKs revealed

certain features which presented repeatedly in the examined

lesions (see Figs 1–4).

For the new en-face (horizontal) imaging mode in HD-OCT,

we defined the following four criteria: disruption of SC, archi-

tectural disarray, cellular ⁄nuclear polymorphism in the stratum

granulosum ⁄stratum spinosum (SG ⁄SS), and bright irregular

bundles in the superficial dermis (Table 2). Disruption of SC

was observed as irregular ruptures in the otherwise bright SC.

Architectural disarray in the epidermis was defined as altera-

tions in the usually regular honeycomb pattern with broad-

ened intercellular demarcations and increased brightness.

Cellular ⁄nuclear polymorphism was characterized by focal

changes in shape and size of the keratinocytes and nuclei

(bright polygonal formations with central dark ⁄oval structure

consistent with the nucleus). In the superficial dermis bright

irregular bundles were detectable when an increase of bright-

ness and clusters of the otherwise regular network occurred.

In the examined AKs we found disruption of the SC in 18 of

20 (90%) lesions, including three (15%) lesions with only

moderate signs of disruption. Architectural disarray and cellu-

lar ⁄nuclear polymorphism were present in 20 of 20 (100%)

examined AKs, while these changes were only moderate in six

(30%) of these lesions. Bright irregular bundles were detected

in 16 of 20 (80%) AKs with seven (35%) lesions showing

only a moderate grade of these changes.

Slice mode of high-definition optical coherence

tomography in actinic keratosis

In the classical vertical OCT imaging mode (slice mode), AKs

showed a destruction of the typical layering of the epidermis

and dermis as described earlier elsewhere.17 In reference to

the criteria of Jørgensen et al.17 we chose the following charac-

teristics for AK in the slice mode of HD-OCT: destruction of

layering, white streaks and dots, and grey areas (Table 3).

Additionally, we defined the feature of an irregular entrance

signal meaning alterations in refractivity and thickness of the

topmost band appearing in the HD-OCT image (Figs 2, 3).

This irregular entrance signal was found in all of the 20

(100%) examined AKs. A destruction of layering was also

present in all of the 20 (100%) lesions. White streaks and dots

were located in 19 of 20 (95%) AKs. Lesion 16 did not show

streaks and dots and was histologically defined as atrophic AK.

Grey areas were observed in 18 of 20 AKs (90%).

High-definition optical coherence tomography in normal

skin

The clinically healthy adjacent skin showed the following HD-

OCT features: SC without interruption, regular epidermal hon-

eycomb pattern in the SG ⁄SS, reticulated meshwork in the

upper dermis, oval dark structures with bright border (hair

follicles) in the en-face mode, and a regular bright entrance

signal of the SC, and distinct layering of the epidermis ⁄dermis

in the slice mode. In the en-face mode of HD-OCT the clinically

normal skin at the level of the SG ⁄SS showed a regular pattern

of small dark cells with bright cell borders without significant

changes in shape and size (Fig. 1). These features were pres-

ent in all of the five examined samples of clinically healthy

skin.

The transition of SG ⁄SS can be recognized by a slight

decrease in size of the keratinocytes. The conversion of the

dermoepidermal junction into the superficial dermis is not



Actinic keratosis in high-definition optical coherence tomography, T. Maier et al. 3

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

Fig 1. (a–d) High-definition optical

coherence tomography (HD-OCT) of normal

skin of the forehead in the slice (upper part)

and en-face (lower part) mode. In the slice

mode the regular layering of the epidermis

and dermis is displayed, disrupted centrally by

a hair follicle (HF). (a) Bright stratum

corneum (SC) of normal skin is interrupted

by dark skin folds and parts of regular stratum

granulosum (SG) (arrow) in the en-face mode.

(b) SG of normal skin presents as regular

round dark structures surrounded by a bright

web consistent with keratinocytes (arrow). (c)

In the stratum spinosum (SS) the cells appear

smaller and form the typical honeycomb

pattern (arrow) in the en-face mode. (d) In the

en-face mode the transition of stratum basale

(SB) ⁄dermoepidermal junction (DEJ) to the

upper dermis (UD) is depicted as bright

bundles with oval-shaped dark structures.

(e–h) HD-OCT of an atrophic actinic keratosis

(lesion 10) displays in the slice mode (upper

part) the destruction of the epidermal ⁄dermal

layering consistent with parakeratosis and the

destruction of the epidermal structure. (f, g)

The regular web and honeycomb pattern of

the SG ⁄SS as displayed in the normal adjacent

skin (a–d) is disrupted and larger dark round

structures surrounded by irregular bright

borders are found coherent with pleomorphic

keratinocytes. (h) The regular composition of

the DEJ is disordered.




sharp because of the flattened rete ridges of the facial skin. In

other regions of the body with more prominent dermal papil-

lae the papillary dermis was more clearly depicted in HD-OCT

(data not shown). The superficial dermis is characterized by a

regular network of bright bundles most probably consistent

with the reticular collagen bundles.

Histological evaluation and comparison with high-

definition optical coherence tomography

The histological evaluation (Table 1) revealed eight unclassi-

fied, six atrophic, two hypertrophic, two initial, one acantho-

lytic and one bowenoid AK. Typical histological features of AK

Table 2 Evaluation criteria of actinic keratoses in the en-face imaging mode of high-definition optical coherence tomography

Lesion Disruption of SC Architectural

disarray in SG ⁄SSCellular ⁄nuclear

polymorphism in SG ⁄SSBright irregular

bundles in SD

1 + + + +2 + + + +

3 + + + +4 + + + +

5 + + ⁄) + ⁄) + ⁄)6 ) + ⁄) + ⁄) +

7 ) + ⁄) + ⁄) )8 + + + +

9 + + + )10 + + + )11 + + + )12 + + + +

13 + + + +14 + + + +

15 + ⁄) + ⁄) + ⁄) + ⁄)16 + ⁄) + ⁄) + ⁄) + ⁄)17 + + + + ⁄)18 + + + + ⁄)19 + + + + ⁄)20 + ⁄) + ⁄) + ⁄) + ⁄)

SC, stratum corneum; SG ⁄SS, stratum granulosum ⁄ stratum spinosum; SD, superficial dermis; +, present; ), absent; + ⁄), moderately present.

Table 3 Evaluation criteria of the actinic keratoses in the slice imaging mode of high-definition optical coherence tomography

Lesion Irregular entrance signal Destruction of layering White streaks White dots Grey areas

1 + + + + +2 + + + + +

3 + + + + +4 + + + + + ⁄)5 + + ⁄) + ⁄) + ⁄) + ⁄)6 + ⁄) + ⁄) + ⁄) + ⁄) +

7 + ⁄) + ⁄) + + +8 + + + ⁄) + ⁄) +

9 + + + + +10 + + + + +

11 + + + + +12 + + + ⁄) + ⁄) +

13 + + + + + ⁄)14 + + + + +

15 + + ⁄) + ⁄) + ⁄) +16 + ⁄) + ⁄) ) ) )17 + + + + +18 + + + + +

19 + + + + +20 + ⁄) + ⁄) + ⁄) + ⁄) )

+, present; ), absent; + ⁄), moderately present.




were documented. Parakeratosis was found in 18 ⁄20 (90%),

destruction of the epidermal structure in 19 ⁄20 (95%), pleo-

morphic keratinocytes in 20 ⁄20 (100%) and inflammatory

infiltrate in 16 ⁄20 (80%) histologically examined AKs.

The histological changes of the SC such as hyper- and para-

keratosis and the ‘pink and blue’ sign correlated with structural

changes in HD-OCT, which we defined as irregular entrance

signal and disruption of the SC. The presence of nuclei in the

corneocytes as a sign of parakeratosis could not be detected dis-

tinctly in HD-OCT. The histologically described destruction of

the epidermal structure correlated with the destruction of layer-

ing in the slice mode and the architectural disarray in the en-face

mode of HD-OCT. The presence of pleomorphic keratinocytes

correlated with cellular ⁄nuclear polymorphism in the en-face

HD-OCT. The observed bright bundles in the superficial dermis

presumably correlated with solar elastosis. The histologically

described inflammatory infiltrate could not be definitely seen

using HD-OCT, although clusters of small bright roundish

structures were sometimes found in the upper layers.

We correlated selected HD-OCT features with the corre-

sponding histological findings as presented in Table 4. The

following pairs were compared: parakeratosis in histology and

disruption of SC in the en-face mode, destruction of epidermal

structure in histology and architectural disarray in SG ⁄SS in

the en-face mode, pleomorphic keratinocytes and cellu-

lar ⁄nuclear polymorphism in the en-face mode, parakeratosis

and irregular entrance signal in the slice mode, destruction of

epidermal structure and destruction of layering in the slice

mode. The sensitivity, specificity and P-value were evaluated

using McNemar’s v2 test with continuity correction. Due to

the small sample size (n = 20) and the finding that the criteria

are all positive in the histology (20 of 20) a specificity of 0%

occurs with a very broad 95% confidence interval (0–84%)

that covers the whole range of the domain for a proportion.

The highest sensitivity was found in the correlating pair

parakeratosis in histology and disruption of SC in the en-face

mode (88%), and second highest sensitivity was 80% in the

correlated pair pleomorphic keratinocytes in histology and cel-

lular ⁄nuclear polymorphism in the en-face mode. Comparing the

histological characteristics of a single lesion with the findings

in the en-face HD-OCT images, individual histological features

such as focal acantholysis and dyskeratosis could be identified.

As depicted in Figure 3 we found in the SG single dark round

structures with a bright irregular centre using the en-face mode

of HD-OCT. These were larger than the regularly composed

surrounding keratinocytes and thus possibly consistent with

dyskeratotic cells as described by conventional histology.

The histologically commonly found acantholysis was most

probably presented by a disruption of the regular honeycomb

pattern with blurred cell borders of keratinocytes and widened

dark spaces between the single cells (Fig. 4). Altogether, the

HD-OCT images correlated well with the features found by

conventional histology.

Discussion

Because of the rising incidence of AK worldwide, the effective

screening, treatment and monitoring of AK is becoming

increasingly important. Therefore the purpose of this preli-

minary study was to test the improved HD-OCT device in a

(a)

(b)

(c)

(d)

(e)

Fig 2. (a) High-definition optical coherence

tomography (HD-OCT) of actinic keratosis

(AK) lesion 9 with irregular corneal layer

displayed as inhomogeneous bright entrance

signal (arrow) and destruction of the

epidermal layering in the slice mode

comparable with the histological parakeratosis

(arrow, c). (b) In the en-face mode of HD-

OCT at the basal layer acantholysis is depicted

as dark fields within the brighter layers (white

circle, micrometer bar 200 lm) consistent

with the histopathological (haematoxylin and

eosin) image of acantholysis (c, white circle,

micrometer bar 50 lm). In contrast to the AK

the adjacent clinically normal skin at the same

level shows a intact layering of the epidermis

and dermis in the slice mode (d) and a

regular web pattern of the stratum

granulosum (SG) in the en-face mode (e).




series of AK to define and evaluate characteristic morphologi-

cal features in comparison with the histological examination

as the gold standard.

The development of HD-OCT allows a higher resolution of

morphological skin structures compared with the conventional

OCT devices. It enables the visualization of skin changes not

only in the typical vertical (slice) imaging mode but also in

the horizontal (en-face) mode and has recently been shown to

be a valuable tool in the diagnostic imaging of basal cell carci-

noma (BCC).18

The main findings of our study were: first, morphological

alterations of AK could be visualized in the en-face mode of

HD-OCT in correlation with histological findings. Second,

lesional skin of AK showed different morphological features in

the HD-OCT than nonlesional adjacent skin. Third, the en-face

mode of HD-OCT allowed the description of cellular altera-

tions present in AK including pleomorphism of keratinocytes,

dyskeratosis and acantholysis.

In earlier studies using conventional OCT, AKs were defined

by the destruction of the epidermal ⁄dermal layering and the

presence of white streaks, dots and grey areas corresponding

to hyperkeratosis.17 Furthermore, thinning of epidermis and

disruption of the entrance signal were described in the OCT

imaging of AK. Barton et al.19 found a dark band in the SC

characteristic for AK with a sensitivity of 79% and a specificity

of 100% in conventional OCT. However, in an observer-

blinded study by Mogensen et al.12 the sensitivity of recogniz-

ing AK and BCC in regular OCT was only 46%. They found

that the OCT diagnosis was less reliable than the clinical diag-

nosis concerning BCC and AK, but the pathological lesions

could be distinguished from healthy skin. These findings are

comparable with our results of alterations in AK present in the

slice mode of HD-OCT such as irregular entrance signal,

destruction of layering, white streaks and dots, and grey areas.

We found that there were distinct features present in the en-

face mode of HD-OCT imaging of AK. Herein, we were able to

visualize cellular and nuclear polymorphism, which have not

been described in conventional OCT so far. We found the

en-face mode of HD-OCT superior to the slice mode in the

imaging of individual histological alterations such as cellular

polymorphism, dyskeratosis and acantholysis.

In this context, our findings of morphological changes of

AK in the en-face mode of HD-OCT could be better compared

with features of AK described in RCM images than in conven-

tional OCT. RCM features of AK were first described by Aghassi

et al.20 They found enlarged, pleomorphic nuclei varying in

size, shape and orientation contrasting with small, uniform

nuclei from normal skin. Later on, Ulrich et al.8 refined evalu-

ation parameters such as superficial disruption, architectural

disarray, and cellular polymorphism at the granular and spi-

nous layer. Additionally, RCM analysis was found helpful in

the noninvasive monitoring of treatment response.21,22 In

RCM, keratinocytes are described as roundish structures with

dark nuclei and bright cytoplasm forming a regular web,

which is known as a honeycomb pattern.23 This description is

comparable with the image of SG ⁄SS in the en-face mode of

HD-OCT, although the differences are not as prominent as in

RCM, where the granular structure of the cells in the granular

layer can be clearly visualized. Nevertheless, in consistency

with RCM features for AK, we recognized alterations of kerati-

nocytes in AK in the en-face mode of HD-OCT with enlarge-

ment of nuclei and varying size and shape of keratinocytes.

(a)

(b)

(c)

Fig 3. (a) Actinic keratosis lesion 7 presents in the slice mode of

high-definition optical coherence tomography (HD-OCT) with an

irregular epidermal band alternating in width and brightness (white

arrow), thus consistent with the histological picture of alternating

ortho- and hyperkeratosis (‘pink and blue’ sign) (black long arrow, c).

(b) In the en-face mode the regular honeycomb pattern of the stratum

granulosum ⁄ stratum spinosum (SG ⁄SS) shows alterations with single

dark round structures with bright centres (white arrows, micrometer

bar 200 lm) possibly related to dyskeratosis of single cells as

depicted in the histology (c, black arrows, micrometer bar 20 lm).

(c) The cellular polymorphism and spongiosis of the keratinocytes as

depicted in the histology (dark circle) could be correlated to

variations in shape, size and brightness of cellular structures in the

en-face mode of HD-OCT (b, white circle).




Even though an exact mapping of the investigated area was

not possible in the HD-OCT, we observed several large dark

roundish bodies with a bright irregular centre in the lesions

where histologically dyskeratotic cells were described. In the

literature, the description of dyskeratotic cells in RCM is some-

how inconsistent. In the latest RCM atlas large round bright

cells with a dark nucleus are described suggestive of dyskera-

totic keratinocytes in SCC,24 whereas Terhorst et al.25 found

large cells with a bright centre and a dark halo corresponding

possibly to apoptotic cells in wound healing. In our opinion,

dyskeratotic and ⁄or apoptotic cells with a condensed nucleus

are more likely to be displayed in HD-OCT and in RCM as

dark cells with a bright centre. We presume that atypical,

pleomorphic cells as found in neoplasia could presumably

present as large bright cells with large darker centres.

Another feature characteristic of AK in histology is acanthol-

ysis, which could be referred to as dark blurry spaces in HD-

OCT in our study. This is consistent with the latest atlas of

RCM describing microvesicle formation in spongiotic derma-

titis as dark hollow spaces. In RCM inflammatory cells are

depicted as scattered small bright spots.24 In the images of en-

face mode of HD-OCT we found inconsistently clusters of

small bright roundish structures in the upper layers possibly

correlating with inflammatory cells.

In comparison with conventional OCT devices, HD-OCT has

the advantage of the additional en-face mode and a high resolu-

tion of about 3 lm in all dimensions which allows supple-

mentary information on cellular changes of the investigated

lesions. On the other hand, conventional OCT has a higher

penetration depth of about 1–2 mm compared with approxi-

mately 750 lm using HD-OCT. In contrast, RCM offers the

highest resolution of about 1Æ5 lm compared with conven-

tional and HD-OCT and allows the visualization of the cellular

and subcellular morphology. However, RCM imaging is lim-

ited to horizontal sections with a penetration depth of about

250 lm and the examination time is about 15 min depending

on the investigator’s experience. In comparison, HD-OCT

imaging takes only about 5–7 min. Another advantage of

RCM is the possibility of a topographical mapping of the

investigated lesion with an integrated dermoscopic picture,

which is to date not available with the HD-OCT device.

Table 4 High-definition optical coherence tomography (HD-OCT) criteria in the en-face and slice mode compared with the matching criteria inroutine histology as the gold standard with individual sensitivity and P-value. Specificity was 0 in all cases because of the small sample size

Histology ParakeratosisDestruction ofepidermal structure

Pleomorphickeratinocytes Sensitivity (%) P-value

HD-OCT criteria

Disruption of SC (en-face) 17 ⁄20 – – 88 1Architectural disarray in SG ⁄SS (en-face) – 19 ⁄20 – 68 0Æ136

Cellular ⁄nuclear polymorphism in SG ⁄SS (en-face) – – 20 ⁄20 80 0Æ133Irregular entrance signal (slice) 18 ⁄20 – – 77 0Æ683

Destruction of layering (slice) – 19 ⁄20 – 68 0Æ13

SC, stratum corneum; SG ⁄SS, stratum granulosum ⁄ stratum spinosum.

(a)

(b)

(c)

Fig 4. (a) High-definition optical coherence tomography imaging of

actinic keratosis lesion 4 shows an irregular entrance band with bright

and dark areas and focally broadened epidermal band in the slice

mode consistent with hyperparakeratosis and acanthosis in histology.

(b) In the en-face mode at the layer of the stratum granulosum ⁄ stratum

spinosum (SG ⁄SS) the regular web structure shows focally broadened

bright interspaces and blurred dark cell borders (white arrows,

micrometer bar 200 lm) possibly consistent with spongiotic

disaggregation as displayed in the histology (c, dark arrows,

micrometer bar 50 lm).




Another promising skin imaging technique with high potential

is multiphoton tomography (MPT), which is based on a tita-

nium : sapphire laser system and provides horizontal optical

sections up to 200 lm tissue depth with an axial and trans-

verse resolution of < 2 lm and < 0Æ5 lm, respectively. It has

shown promising results in the imaging of various dermatoses

especially in combination with OCT or RCM.26,27 Neverthe-

less, MPT currently has a relatively slow imaging process with

a narrow field of view of about 0Æ3 mm3, which makes it at

this time point inconvenient as an imaging tool for daily clini-

cal practice.

Altogether, HD-OCT with the integrated slice and en-face

mode offers an additional noninvasive investigation tool in the

diagnostics of AK by combining certain advantages of both

conventional OCT and RCM. For the determination of sensitiv-

ity and specificity levels of HD-OCT in the diagnosis of AK

additional, observer-blinded studies are required.

What’s already known about this topic?

• Conventional optical coherence tomography (OCT) has

been studied in the diagnosis of actinic keratosis (AK),

without the discrimination of reliable criteria because of

an insufficient resolution.

What does this study add?

• Using the new en-face imaging mode of an innovative high-

definition OCT morphological features of AK are now

described in correlation with the histological findings.

References

1 Glogau RG. The risk of progression to invasive disease. J Am AcadDermatol 2000; 42:23–4.

2 Salasche SJ. Epidemiology of actinic keratoses and squamous cellcarcinoma. J Am Acad Dermatol 2000; 42:4–7.

3 Frost C, Williams G, Green A. High incidence and regression ratesof solar keratoses in a Queensland community. J Invest Dermatol

2000; 115:273–7.4 Memon AA, Tomenson JA, Bothwell J et al. Prevalence of solar

damage and actinic keratosis in a Merseyside population. Br J Der-matol 2000; 142:1154–9.

5 Stockfleth E, Ulrich C, Meyer T et al. Epithelial malignancies inorgan transplant patients: clinical presentation and new methods of

treatment. Recent Results Cancer Res 2002; 160:251–8.6 Sterry W, Griffiths C, Korting HC, eds. EDF Guidelines for Dermatology

in Europe, Vol. 1. Berlin: ABW Wissenschaftsverlag Berlin, 2009.7 Quaedvlieg PJ, Tirsi E, Thissen MR et al. Actinic keratosis: how to

differentiate the good from the bad ones? Eur J Dermatol 2006;16:335–9.

8 Ulrich M, Maltusch A, Rius-Diaz F et al. Clinical applicability of invivo reflectance confocal microscopy for the diagnosis of actinic

keratoses. Dermatol Surg 2008; 34:610–19.9 Zalaudek I, Giacomel J, Argenziano G et al. Dermoscopy of facial

nonpigmented actinic keratosis. Br J Dermatol 2006; 155:951–6.

10 Pellacani G, Cesinaro AM, Seidenari S. In vivo assessment of melano-cytic nests in nevi and melanomas by reflectance confocal micros-

copy. Mod Pathol 2005; 18:469–74.11 Maier T, Sattler EC, Braun-Falco M et al. Reflectance confocal

microscopy in the diagnosis of partially and completely amelanoticmelanoma: report on seven cases. J Eur Acad Dermatol Venereol 2012.

doi: 10.1111 ⁄ j.1468-3083.2012.04465.x. [Epub ahead of print].12 Mogensen M, Jørgensen TM, Nurnberg BM et al. Assessment of

optical coherence tomography imaging in the diagnosis of non-melanoma skin cancer and benign lesions versus normal skin:

observer-blinded evaluation by dermatologists and pathologists.

Dermatol Surg 2009; 35:965–72.13 Gambichler T, Orlikov A, Vasa R et al. In vivo optical coherence

tomography of basal cell carcinoma. J Dermatol Sci 2007; 45:167–73.

14 Welzel J. Optical coherence tomography in dermatology: a review.Skin Res Technol 2001; 7:1–9.

15 Korde VR, Bonnema GT, Xu W et al. Using optical coherencetomography to evaluate skin sun damage and precancer. Lasers Surg

Med 2007; 39:687–95.16 Gambichler T, Jaedicke V, Terras S. Optical coherence tomography

in dermatology: technical and clinical aspects. Arch Dermatol Res2011; 303:457–73.

17 Jørgensen TM, Tycho A, Mogensen M et al. Machine-learning clas-sification of non-melanoma skin cancers from image features

obtained by optical coherence tomography. Skin Res Technol 2008;14:364–9.

18 Maier T, Braun-Falco M, Hinz T et al. Morphology of basal cell car-cinoma in high definition optical coherence tomography: en-face

and slice imaging mode, and comparison with histology. J Eur AcadDermatol Venereol 2012. doi: 10.1111 ⁄ j.1468-3083.2012.04551.x.

[Epub ahead of print].19 Barton JK, Gossage KW, Xu W et al. Investigating sun-damaged

skin and actinic keratosis with optical coherence tomography: apilot study. Technol Cancer Res Treat 2003; 2:525–35.

20 Aghassi D, Anderson RR, Gonzalez S. Confocal laser microscopicimaging of actinic keratoses in vivo: a preliminary report. J Am Acad

Dermatol 2000; 43:42–8.21 Ulrich M, Krueger-Corcoran D, Roewert-Huber J et al. Reflectance

confocal microscopy for noninvasive monitoring of therapy anddetection of subclinical actinic keratoses. Dermatology 2010; 220:15–

24.22 Ulrich C, Johannsen A, Roewert-Huber J et al. Results of a random-

ized, placebo-controlled safety and efficacy study of topical diclofe-

nac 3% gel in organ transplant patients with multiple actinickeratoses. Eur J Dermatol 2010; 20:482–8.

23 Gonzalez S, Gill M, Halpern AC, eds. Reflectance Confocal Microscopy ofCutaneous Tumors: An Atlas with Clinical, Dermoscopic and Histological Correla-

tions. London: Informa Healthcare, 2008.24 Hofmann-Wellenhof R, Pellacani G, Malvehy J, Soyer HP, eds.

Reflectance Confocal Microscopy for Skin Diseases, Vol. 1. Berlin: Springer-Verlag, 2012.

25 Terhorst D, Maltusch A, Stockfleth E et al. Reflectance confocalmicroscopy for the evaluation of acute epidermal wound healing.

Wound Repair Regen 2011; 19:671–9.26 Konig K, Speicher M, Buckle R et al. Clinical optical coherence

tomography combined with multiphoton tomography of patientswith skin diseases. J Biophotonics 2009; 2:389–97.

27 Alex A, Weingast J, Weinigel M et al. Three-dimensional multipho-ton ⁄optical coherence tomography for diagnostic applications in

dermatology. J Biophotonics 2012. doi: 10.1002 ⁄ jbio.201200085.[Epub ahead of print].




bjd dermatopathology british journal of dermatology

Documents