bjd dermatopathology british journal of dermatology
TRANSCRIPT
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.
� 2012 The Authors
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
� 2012 The Authors
BJD � 2012 British Association of Dermatologists 2012
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.
� 2012 The Authors
BJD � 2012 British Association of Dermatologists 2012
4 Actinic keratosis in high-definition optical coherence tomography, T. Maier et al.
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.
� 2012 The Authors
BJD � 2012 British Association of Dermatologists 2012
Actinic keratosis in high-definition optical coherence tomography, T. Maier et al. 5
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).
� 2012 The Authors
BJD � 2012 British Association of Dermatologists 2012
6 Actinic keratosis in high-definition optical coherence tomography, T. Maier et al.
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).
� 2012 The Authors
BJD � 2012 British Association of Dermatologists 2012
Actinic keratosis in high-definition optical coherence tomography, T. Maier et al. 7
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).
� 2012 The Authors
BJD � 2012 British Association of Dermatologists 2012
8 Actinic keratosis in high-definition optical coherence tomography, T. Maier et al.
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].
� 2012 The Authors
BJD � 2012 British Association of Dermatologists 2012
Actinic keratosis in high-definition optical coherence tomography, T. Maier et al. 9