Download Diagnosis of Nonmelanoma Skin Cancer/Keratinocyte

Diagnosis of Nonmelanoma Skin Cancer/Keratinocyte
Carcinoma: A Review of Diagnostic Accuracy of Nonmelanoma
Skin Cancer Diagnostic Tests and Technologies
METTE MOGENSEN, MD,
AND
GREGOR B. E. JEMEC, MD, DMSC
BACKGROUND Nonmelanoma skin cancer (NMSC) is the most prevalent cancer in the light-skinned
population. Noninvasive treatment is increasingly used for NMSC patients with superficial lesions,
making the development of noninvasive diagnostic technologies highly relevant.
OBJECTIVE The scope of this review is to present data on the current state-of-the-art diagnostic methods for keratinocyte carcinoma: basal cell carcinoma, squamous cell carcinoma, and actinic keratosis.
METHODS AND MATERIALS MEDLINE, BIOSIS, and EMBASE searches on NMSC and physical and clinical examination, biopsy, molecular marker, ultrasonography, Doppler, optical coherence tomography,
dermoscopy, spectroscopy, fluorescence imaging, confocal microscopy, positron emission tomography,
computed tomography, magnetic resonance imaging, terahertz imaging, electrical impedance and sensitivity, specificity, and diagnostic accuracy.
RESULTS State-of-the-art diagnostic research has been limited in this field, but encouraging results
from the reviewed diagnostic trials have suggested a high diagnostic accuracy for many of the technologies. Most of the studies, however, were pilot or small studies and the results would need to be
validated in larger trials.
CONCLUSIONS Some of these new imaging technologies have the capability of providing new, threedimensional in vivo, in situ understanding of NMSC development over time. Some of the new technologies described here have the potential to make it from the bench to the clinic.
Mette Mogensen, MD, and Gregor B. E. Jemec, MD, DMSc have indicated no significant interest with commercial supporters.
N
onmelanoma skin cancer (NMSC) is the most
prevalent cancer in the light-skinned population, and the critical factor in assessment of patient
prognosis in skin cancer is early diagnosis.1 Serious
morbidity in NMSC is often the result of misdiagnosis or underestimation of the biologic potential
of the primary tumor.
All diagnostic tests can be accredited with precision
and accuracy. Precision, also known as reliability,
reproducibility, or repeatability, refers to agreement
of the test. If the diagnostic test is performed multiple times on the same subject with the same result,
the test is precise. Accuracy describes whether the
diagnostic test yields a correct or incorrect answer by
correlating the test result with the truth. Thus the
accuracy of a test tells us how efficient it is in arriving at the correct diagnosis. The accuracy of a
diagnostic test is usually reported in terms of its
sensitivity, specificity, and predictive values.
The truth, the correct diagnosis, cannot always unequivocally be reached, and for that reason, a ‘‘gold
standard’’ or reference standard is chosen to represent the truth. For skin cancer a biopsy specimen
obtained for histopathologic examination to diagnose skin cancer is considered the reference standard. Biopsy may be a time-consuming, expensive,
and sometimes a mutilating and painful experience
to the patient, especially because only approximately
Both authors are affiliated with the Department of Dermatology, University of Copenhagen, Roskilde Hospital, Roskilde,
Denmark
& 2007 by the American Society for Dermatologic Surgery, Inc. Published by Blackwell Publishing ISSN: 1076-0512 Dermatol Surg 2007;33:1158–1174 DOI: 10.1111/j.1524-4725.2007.33251.x
1158
MOGENSEN AND JEMEC
3% of presumed benign lesions are actually malignant.2 Furthermore, whereas treatment has hitherto
been almost synonymous with surgery, new noninvasive treatment strategies call for noninvasive diagnostics. To assist clinical diagnosis of NMSC and
malignant melanoma (MM), an extensive assortment of diagnostic technologies and tests have been
developed.3 The scope of this review is to present
available data on the diagnostic accuracy of the
different diagnostic methods. NMSC refers mainly
to basal cell carcinoma (BCC), squamous cell carcinoma (SCC), and premalignant actinic keratosis
(AK).
Accurate diagnostic test assessment involves four
phases4: Determining the normal range of values for
a diagnostic test through observational studies in
healthy people (1), diagnostic accuracy assessment
through case-control studies (2), assessment of clinical consequences of introducing a diagnostic test
through randomized trials (3), and determining the
effects of introducing a new diagnostic test into
clinical practice by surveillance in large cohort
studies (4).5
Materials
We have searched for clinical, human studies on the
diagnostic accuracy of the following NMSC diagnostic tests and techniques: clinical and physical
examination; biopsy techniques; histopathology and
molecular markers; high-frequency ultrasonography
(HFUS); Doppler sonography; dermoscopy,
dermatoscopy, epiluminescence microscopy, incident
light microscopy, and skin surface microscopy; optical coherence tomography (OCT); confocal microscopy (CM); near infrared and Raman
spectroscopy; fluorescence imaging/spectroscopy;
terahertz imaging; electrical impedance; positron
emission tomography (PET), computed tomography
(CT); and magnetic resonance imaging (MR, MRI).
Studies of NMSC diagnosis have been included in
this review according to the search protocol of the
review. Our interest has been to sample data on diagnostic accuracy in NMSC diagnostics. If sensitivity
and specificity values have not been calculated, we
have calculated values whenever the data were
available.
Methods
MEDLINE, BIOSIS, and EMBASE searches on
medical subject headings (MeSH) or similar terms:
skin, skin appendages, dermis, cutis, epidermis, integumentum, skin neoplasms, skin tumor and/or tumor, BCC, SCC, Bowen’s disease, AK and/or solar
keratosis, keratinocyte carcinoma and physical and
clinical examination, biopsy, molecular marker, ultrasonography, Doppler, OCT, dermoscopy,
dermatoscopy, epiluminescence microscopy and/or
incident light microscopy, spectroscopy, fluorescence
imaging, spectrophotometry, colorimetry, confocal
microscopy, PET, positron emission tomography, CT,
computed tomography, MR, MRI, magnetic resonance imaging, terahertz imaging, electrical impedance and diagnosis, sensitivity and specificity, and
diagnostic accuracy. Articles and abstracts have also
been retrieved from Web searches on Google Scholar,
Ixquick, Scirus, and reference lists of identified
studies. Only studies on diagnosis of NMSC have
been selected for this review. Only articles published
after 1990 have been included. Articles were restricted to those involving humans.
Results
Clinical and Physical Examination
Despite being the most accessible and used test,
precision and accuracy of the clinical examination
are often not known. The precision in clinical examination can be described as intra- and interobserver differences. Precision estimates must be
generated in a blinded fashion; this is feasible in interobserver observations, but as lesions change over
time intraobserver differences must be estimated
from evaluation of pictures of skin lesions assessed at
two different times.6,7 Physical examination for
NMSC can be performed by an expert, by a general
practitioner or by the patient. Results from the
33:10:OCTOBER 2007
1159
DIAGNOSTIC ACCURACY OF NONMELANOMA SKIN CANCER DIAGNOSTIC TESTS
selected studies2,6–14 are presented in Table 1. The
false-negative diagnostic rates as well as the falsepositive rates cannot be ignored as it was demonstrated in a study where 3% of lesions assessed as
benign proved malignant and 40% of suspected
malignancies were benign.2 In the reviewed studies
the overall sensitivity for clinical diagnosis of NMSC
is 56% to 90%, and specificity 75% to 90%, with
highest values for BCC diagnosis.
In studies where positive predictive value (PPV) is
mentioned, it must be taken into account that PPV is
prevalence-dependent; a study conducted in a high
prevalence setting would tend to increase PPV.
Patients themselves may also be involved in the diagnosis of cancer. When patients with nevi and/or
melanocytic lesions assessed new and changing
moles by performing a skin self-examination, the
accuracy was significantly higher with the aid of a
baseline digital photography of the skin.15 No similar studies have been performed with NMSC.
Biopsy Techniques
Shave and punch biopsy have been compared in a
study including 86 biopsy specimens and subsequent
total excision of the tumor.16 Punch biopsy was accurate in determining BCC in 81% of cases, and
shave biopsy correctly identified 76%. In a systematic review of exfoliative cytology in diagnosis
of BCC, a meta-analysis showed the pooled
sensitivity to be 97% [95% confidence interval
(CI), 94%–99%] and specificity 86% (95% CI,
80%–91%);17,18 however, cytology does not give
any information about subtype and none about
tumor borders.
Histopathology and Molecular Markers
Histopathology is a subjective assessment by a dermatopathologist, i.e., data acquisition guided by
classification standards and pattern recognition by
the human brain. Ideally histopathologic classification of NMSC should be able to identify subtypes
that correlate with clinical behavior and treatment
1160
D E R M AT O L O G I C S U R G E RY
requirements. In addition, the classification should
be easy to use and reproducible. Determining diagnostic accuracy of the reference standard itself is
hampered by the fact that there are several classification systems for NMSC.19,20 Difficulties confront
pathologists reporting BCCs: Many BCCs have more
than one growth pattern, and published data have
varied with regard to number necessary to designate
the presence of a specific subtype. Furthermore, the
accuracy of reporting different subtypes has not been
extensively investigated.19 In general, interobserver
differences cannot be neglected as a potential source
of error, with potential clinical consequences for the
patient, but also diagnostic researchwise, because all
other skin cancer diagnostic techniques must be
compared to histopathology, the reference standard.
The interobserver differences in this review ranged
from 1.2% to 7%. In Table 2 results from the reviewed studies are presented.21–26
A current research goal is to define skin cancer by its
phenotype in terms of molecular abnormalities as a
reference standard for NMSC diagnosis.19,20,27–30
To date a molecular marker with a high diagnostic
accuracy in NMSC diagnosis has not been identified.
HFUS and Doppler Sonography
The principle in HFUS is the emission of a pulsed
ultrasound (US) from a transducer and registration
of the intensity of the echo backscattered from the
tissue. The amplitude of the curve reproduces the
intensity and time delay of the returning US and is
called A-scan. A B-scan is created when the transducer is moved laterally creating a two-dimensional
image. Penetration depth and resolution of US is
inversely related to the frequency, which makes
HFUS suitable in dermatology. Axial resolution at
20 MHz is 50 mm and lateral resolution is 350 mm.
The sonographic characteristics of skin tumors have
been widely investigated during the past decades31–51
(Table 3). By means of HFUS, skin tumors generally
appear as a homogeneously echo-poor area in comparison to the surrounding echo-rich dermis. Because
all skin tumors appear echo-poor, HFUS alone is not
6
Plastic surgeons
Har-Shai et al., 200111
Whited et al., 1997110
835 lesions in 778
2582 NMSC lesions
excised from 1223
Dermatologists
493 NMSC patients
and family
practitioners
(FP)
Primary care phy- 190 NMSC
sicians (PCP)
patients
and dermatologists
Plastic surgeons
Morrison et al., 20019
Does history influence the
clinical diagnosis?
Question
53 patients with 143 Observer agreement
NMSC lesions
Choose between a
Observer agreement
diagnosis of either AK or SCC in
15 slides
141 patients with
Effect of additional data
BCC lesions
50 NMSC patients
Number of
participants
Diagnostic accuracy
Diagnostic accuracy
Diagnostic accuracy
Diagnostic accuracy
Dermatologists
complete a
questionnaire
before biopsy,
confidence
was plotted
level 1–3
Dermatologist di- 102 lesions in 70 re- Diagnostic accuracy
agnosis was
nal transplant pacompared to
tients
histopathology
Plastic surgeons
2058 lesions in 809
Diagnostic accuracy
NMSC patients
referred for tumor
excision
Dermatologists
blinded to the
patient history
Two dermatologists
77 pathologists
Group studied
Ek et al., 200512
Hallock and Lutz, 19982
Cooper and Wojnarowska, 200214
Schwartzberg et al.,
20058
Davis et al., 200513
Leffell et al., 199310
Whited et al., 1995
Authors, year
TABLE 1. Clinical Diagnosis
Sensitivity of the PCPs was 57% (95% CI, 44%–68%) and
specificity 88% (95% CI, 81%–93%).
FPs diagnosed 22% of skin cancer before biopsy,
dermatologists diagnosed 87% correctly.
Sensitivity 91% and a PPV of 71%.
Three-fourths of benign lesions were identified (sensitivity 93% and specificity 86%). Only 60% of malignant
lesions were identified (sensitivity 73% and specificity
90%). 3% of presumed benign lesions were
malignant.
BCC and SCC was diagnosed with a sensitivity of 89%
and 56% and a PPV of 65% (p = .001).
Sensitivity was 90.5% (SCC) and 66.6% (BCC). Specificity
was 75.3% (SCC) and 85.6% (BCC).
PPV was 80% when additional data was presented.
ICC was 0.96 for dermatopathologists and anatomic
pathologists and 0.65 for fellows.y
Kappa 0.78 in AK and 0.38 in SCC.
Kappa values 0.04 (blinded) and 0.76 (with history).
Result
MOGENSEN AND JEMEC
33:10:OCTOBER 2007
1161
1162
Oliveria et al., 200415
D E R M AT O L O G I C S U R G E RY
The kappa statistics is a measure of the agreement level beyond what might be expected by chance alone. The kappa values vary from 1 (perfect disagreement) to 1 1 (perfect
agreement).
y
Intraclass correlation coefficient (ICC) is an estimate of the interobserver reliability by two-way analysis of variance. Standard criteria for ICC are 0.40 = poor; 0.40–0.59 = fair;
0.60–0.74 = good; 0.75–1.00 = excellent correlation between observers.
AK, actinic keratosis; BCC, basal cell carcinoma; NMSC, nonmelanoma skin cancer; PPV, positive predictive value; SCC, squamous cell carcinoma.
50 patients with
dysplastic nevi
Diagnostic accuracy
GPs diagnosed SCC in 70, dermatologists in 24. 53 suspected SCCs were biopsied; 13 of these were confirmed SCC by pathologists.
The accuracy of self-examination was significantly higher (p = .001) with the aid of digital photography. Sensitivity without photo was 60% and specificity 96%
compared to sensitivity with photo was 72% and
specificity 98%.
Diagnostic accuracy
GPs, dermatologists, and pathologists
Patients
Westbrook et al.,
2006111
283 NMSC cases
Result
Group studied
Authors, year
TABLE 1. Continued
Number of
participants
Question
DIAGNOSTIC ACCURACY OF NONMELANOMA SKIN CANCER DIAGNOSTIC TESTS
suitable for differential diagnosis.52,53 Overestimation of tumor thickness appears to be a general
problem because fibrosis and inflammation generally
have the same echogenecity as NMSC. Skin tumor
vascularization studies using Doppler techniques are
presented in the Table 3 as well. Laser Doppler (LD)
devices send a monochromatic laser beam toward
the target and collect the reflected radiation. According to the Doppler effect, the change in wavelength of the reflected radiation is a function of the
targeted object’s relative velocity. The velocity of the
object can thus be calculated by measuring the
change in wavelength of the reflected laser light.
LD can be configured to act as flow meters.
In short HFUS in NMSC diagnosis is to some extent
capable of revealing the three-dimensional size,
margins, and relation to adjacent vessels of a suspicious skin lesion. Information on quality (such as
solid, cystic, or combined) and information about
the inner structure (homogeneous, inhomogeneous,
hypo- or hyperechoic, calcification, or necrosis) can
be obtained54 (Figure 1).
Dermoscopy, Dermatoscopy, Epiluminescence
Microscopy, Incident Light Microscopy, and
Skin Surface Microscopy
In dermoscopy the lesion is examined with a 10 to
100 magnification lens placed directly against skin
to which immersion oil has been applied to remove
scatter of light at the air–skin boundary. Dermoscopy has had the largest clinical impact regarding
new skin cancer diagnostic technologies in the diagnosis of pigmented skin lesions, but it has also
been used in NMSC. Dermoscopy of NMSC is still at
its infancy because consensus regarding diagnostic
criteria has not been established yet. Diagnostic accuracy regarding vessels characteristics in BCC and
Bowen’s disease, however, seems quite promising,
and sensitivity for BCC diagnosis ranges from 87%
to 96%, and specificity from 72% to 92% (Table 4).
Dermoscopy and spectroscopy devices, when linked
to a video microscope for computer analysis, have
MOGENSEN AND JEMEC
TABLE 2. Histopathology and Molecular Markers
Authors,
year
Group studied
Lind et al.,
199521
Two
pathologists
Olhoffer
et al.,
200222
Brochez
et al.,
200223
Dermatopathologists 336 cases
and pathologists
Trotter and
Bruecks,
200326
Renshaw
et al.,
2002112
Kamiya
et al.,
200324
Biesterfeld
and
Josef,
200225
Number of
cases
2694 slides
Question
Results
Interobserver
differences
Thirty-two major errors were
found, involving 1.2% of cases
reviewed. Errors were divided
into four types: (1) major: errors
in diagnosis that could directly
affect patient care; (2) diagnostic discrepancies: errors in diagnosis that should not affect
patient care; (3) minor: correct
diagnosis rendered, but report
correction required to add supportive information; (4) clerical:
typographical and grammatical
errors.
Discordance in 5.7% of cases. New
management in 18 of 19 severe
cases.
Overall sensitivity was 87% (range,
55%–100%) and specificity 94%
(range, 83%–100%).
Interobserver
differences
20 pathologists
48 slides
Interobserver
diagnosing
differences
pigmented skin
lesions and NMSC
Two pathologists
Blinded
Interobserver
Agreement was found in more
review of 592 histo- differences
than 93% of cases.
pathology slides
77 pathologists (both 15 SCC and
Interobserver
Mean ICC was 0.97 for all patholdermato- and
AK slides
differences
ogists. Agreement was above
anatomic)
70% for 11 of 15 slides.
Performing reverse 26 lymph nodes; 10 Molecular mark- This method showed a detection
transcriptase–polyhad histologically
er: micrometassensitivity of one tumor cell in
merase chain reac- proven metastasis,
tasis of SCC in
106 lymphocytes.
tion (RT-PCR) to
and 16 had none
lymph nodes
amplify keratin19
MIB-1 immuno49 keratoacanthomas Molecular mark- If specificity 85% is required senhistometry for the
and 48 SCCs
er: proliferation
sitivity decreases to 56%. MIB-1
differential diagnomarker MIB-1
is currently of limited value in
sis between KA and
SCC diagnosis.
SCC
Intraclass correlation coefficient (ICC) is an estimate of the interobserver reliability by two-way analysis of variance. Standard criteria for
ICC are 0.40 = poor; 0.40–0.59 = fair; 0.60–0.74 = good; 0.75–1.00 = excellent correlation between observers.
AK, actinic keratosis; NMSC, nonmelanoma skin cancer; SCC, squamous cell carcinoma.
been named ‘‘mole scanners.’’3 One pilot study on
NMSC diagnosis with computerized dermoscopy
combined with a two-dimensional in vivo reflectance
spectrophotometer included a few BCC and also MM
patients.55 Nodular BCC showed a characteristic
concentration of perilesional blood vessels combined
with slight central fibrosis. Pigmented BCC did not
show the erosion of dermal collagens specific for MM.
Dermoscopic features of BCC are arborizing vessels
defined as stem vessels with a large diameter,
branching irregularly into the finest terminal capillaries. Diameter of the vessels can be up to 0.2 mm,
branching irregularly into capillaries of 10 mm.56
Gray-brown lumps, often ovoid in shape, and amber-colored crusts are other dermoscopy features in
BCC (Figure 2). In Bowen’s disease, dermoscopic
33:10:OCTOBER 2007
1163
DIAGNOSTIC ACCURACY OF NONMELANOMA SKIN CANCER DIAGNOSTIC TESTS
TABLE 3. High-Frequency Ultrasonography
Authors, year
Number and pathology of
patients
Question asked
Results
181 patients with BCC before Thickness of the
BCC thickness predicts outcome 1 year
Moore and
and after photodynamic
lesions
after photodynamic therapy with dAllan, 200333
therapy treatment
aminolevulinic acid.
33 patients with SCC
HFUS diagnosis of
In SCC of the penis, HFUS provides diagLont et al.,
SCC of the penis
nostic value in staging the disease. PPV
2003113,114
compared to MRI
for corpus cavernosum infiltration,
67% (HFUS) and 75% (MRI).
16 BCC lesions and
Laser Doppler perfu- Tumor perfusion values higher than surStucker et al.,
27 MM
sion imaging
rounding skin. BCC perfusion values
1999115
similar in the tumor area as opposed to
MM, where perfusion was higher in the
tumor center.
81 clinically malignant
Laser Doppler perfu- Sensitivity 0.75 and specificity 0.79 if
Schroder
tumors
sion imaging
three to five vessels were visible in the
et al., 2001116
study with and
tumor, and sensitivity 0.58 and speciwithout contrast
ficity 0.88 if a parameter called ‘‘percentage vessel area’’ exceeded 5%.
19 benign, 32 BCC, and
Power Doppler
Specificity of 63% and sensitivity 88% in
Karaman
15 SCC
(Doppler indediagnosis based on vascular patterns
et al., 200132
pendent of angle)
alone in the lesions.
BCC, basal cell carcinoma; HFUS, high-frequency ultrasound; MM, malignant melanoma; MRI, magnetic resonance imaging; PPV, positive
predictive value; SCC, squamous cell carcinoma.
features are glomerular vessels (90%) and scaly surface (90%).57 Glomerular vessels are dotted vessels
often distributed in clusters mimicking the glomerular apparatus of the kidney. Compressing the blood
vessels renders them invisible. Magnification by 30
times or more must be applied to visualize vessels
down to 10 mm.58
OCT
OCT is a novel, noninvasive optical imaging technology. It can provide cross-sectional tomographic
images of tissue pathology in situ and in real time.
OCT is analog to B-mode US pulse-echo imaging
with an optical rather than acoustical reflectivity
being measured. In OCT, linear characteristics as
scattering, absorption, birefringence, and refractive
index are measured to produce images with micrometer resolution.59 When birefringence characteristics of the tissue are mapped in OCT images, it is
referred to as polarization sensitivity (PS) OCT. OCT
images can also be enhanced by Doppler function.
1164
D E R M AT O L O G I C S U R G E RY
OCT provides cross-sectional images of structures below the tissue surface in analogy to histopathology.60,61
NMSC and MM has been investigated with OCT with
quite promising results.62–67 Pilot studies have suggested that OCT can be used clinically in diagnosing
NMSC.62,63,65,67,68 In a study of 20 BCC lesions, there
was an excellent match of histologic features seen on
light microscopy with OCT images of superficial,
nodular, micronodular, and infiltrative BCCs.
The actual predictive value of OCT in BCC patients
could not be calculated from this study. In another
study of three patients with superficial BCC and
three patients with cutaneous MM, OCT images
were compared with histology. The size, allocation,
and form of BCC nests seemed to be similar to those
in histologic images.69 In a study of nine patients
with 12 BCC lesions, some lesions were distinct in
OCT images and BCC subtypes as superficial and
nodular could be identified in OCT images.70 Interestingly scar tissue could be distinguished from
BCC with PS-OCT. Therefore, PS-OCT may have
MOGENSEN AND JEMEC
Figure 1. HFUS and Doppler images of BCC lesion on the nose. (A) Photo of the BCC lesion. (B) 18-MHz real-time HFUS
image of BCC lesion in (A). (C) Doppler image of lesion in (A); vasculature is marked in red and blue. (D) Histopathology slide
HE stain from the lesion (A) showing BCC; original magnification, 20. Images by courtesy of Dr Ximena Wortsman, Chile.
additional advantages as BCC differ in content of
birefringent collagens from normal skin. In a study
of two patients with invasive BCC, PS-OCT was
used to discriminate tumors from normal skin. In
normal skin, a bright band of birefringence signal is
seen at middepth of the PS-OCT image, corresponding to the upper reticular dermis. In PS-OCT
images of invasive BCC, there was a dramatic alteration in the birefringence signal, it almost disappeared in the ulcerative, invasive BCC and showed a
haphazard distribution in the other invasive BCC
lesion (see also Figure 2). Furthermore, a gradual
transition from normal-appearing tissue to tumor
tissue could be detected by PS-OCT at the BCC
borders, indicating an ability of PS-OCT to delineate
tumor borders71 (Figure 3).
A study compared OCT and HFUS in diagnosis of
eyelid tumors. Examination of 38 patients (BCC
4/38, AK 1/38, and other benign and malignant
tumors) showed that OCT was superior in detecting
cystic lesions, but due to low penetration of the OCT
system in the skin, tumor margins could not be
determined.72
CM
Reflectance CM or confocal scanning laser microscopy has the highest resolution of all optical techniques used in NMSC diagnosis research.73 Current
CM systems have an axial resolution of 1 to 5 mm
and a lateral resolution of 0.5 to 1 mm. The penetration depth maximum is 300 mm. CM uses a point
source light to illuminate a small spot within a tissue.
The pinhole minimizes out-of-focus light reaching
the detector, and only confocal light is detected. For
in vivo imaging, a plastic cap filled with water must
be adapted to the skin. CM has proved a potentially
valuable diagnostic aid in BCC and SCC diagnosis,
both ex vivo and in vivo74,75 (Table 5). Characteristic CM features of NMSC are, in BCCs, abundant
blood vessels juxtaposed to BCC cells, sometimes in
33:10:OCTOBER 2007
1165
DIAGNOSTIC ACCURACY OF NONMELANOMA SKIN CANCER DIAGNOSTIC TESTS
TABLE 4. Dermoscopy
Authors, year
Number and pathology
of patients
Question asked
21 SCCs in situ
(Bowen’s disease)
Characteristic dermoscopy
features
Results
Glomerular vessels with a patchy
distribution and scaly surface
was found in 90%
of all lesions.
Characteristic
Arborizing vessels are detected in
Menzies, 2002118 71 pigmented BCC
lesions
dermoscopy
52%, sensitivity was 93%, and
features
specificity was 89%–92%.
Internet study regarding di- Sensitivity for BCC was 86.7% (95%
Zalaudek et al., 165 lesions (including 20
BCCs, otherwise mostly
agnostic accuracy, a three- CI, 76.9%–92.7%) and specificity
2006119
pigmented lesions, 150
point checklist (asymme71.9% (95% CI, 56.6%–83.3%).
dermatologists and other
try, atypical network, bluedoctors
white structures)
531 lesions (117 BCCs)
Diagnostic features: vascular Arborizing vessels were seen in
Argenziano
from 517 patients
structures
82% of BCCs, with a PPV of 94%
et al., 200456
(po.001) SCCs in situ: 13 of 16
lesions showed glomerular vessels, PPV 62%
Diagnostic features: vascular Sensitivity 96%, specificity 91%.
Kreusch, 200258 BCC patients, number
not retrieved
structures
7 BCC patients,
Diagnostic features: vascular Microvessel area fractions were inOtis et al.,
12 patients
structures with capillaroscreased 4.9-fold in BCC and 2.52004;120 Newell et al.,
copy
fold in AK compared to normal
2003121
skin.
111 skin samples
Blood vessels counted after Significant difference between the
Chin et al.,
(including
immunohistochemistry
groups and also between BCCs
2003122
20 SCCs, 50 BCCs)
and SCCs.
Zalaudek et al.,
2004,
200557,117
AK, actinic keratosis; BCC, basal cell carcinoma; PPV, positive predictive value; SCC, squamous cell carcinoma.
tightly packed nests, and rolling of leukocytes and
lymphocytes along the endothelial lining. BCC cells
appeared to be oval, elongated with a prominent,
monomorphic polarized nucleus. Tumor cells had a
high refractive index with dark-appearing nuclei,
and the cytoplasm appeared bright.76,77 In SCC,
features are irregular epithelial mass with a variable
proportion of normal and atypical keratinocytes,
along with areas of anaplasia; in AKs, CM revealed
hyperkeratosis, lower epidermal nuclear enlargement, and pleomorphism.74 Some features seen in
CM images (e.g., uniform polarization of BCC nuclei, margination and rolling of leukocytes) are
morphologic features not recognized in BCC histologypathology.76,77 In Mohs micrographic surgery
(MMS), CM may have a role in analyzing untreated
fresh biopsy specimens. Fluorescence fiber-optic CM
in vivo is a novel technique, where fluorophore distribution in the skin may illustrate morphologic
1166
D E R M AT O L O G I C S U R G E RY
changes in the epidermis. An application for fluorescence CM is the ability to image fluorescent
markers that target specific subcellular molecules
including proteins and therefore to monitor specific
pathologic and immune processes over time.78
Near Infrared, Diffuse Reflectance, and
Raman Spectroscopy
Light that penetrates the skin surface is variably
absorbed by different skin components termed
chromophores. The skin components, which subsequently emit radiation, are termed fluorophores.
Optical measurements of the skin can therefore be
based on the interactions of nonionizing electromagnetic radiation and the skin.73 The absorbed
energy may be dissipated as heat (tissue absorption),
reemitted as electromagnetic radiation of lower energy with a longer wavelength (fluorescence), or even
MOGENSEN AND JEMEC
Figure 2. Dermoscopy image from a BCC lesion (A) and the corresponding clinical photo (B). The dermoscopy image
visualizes the arborizing vessel in the upper right section of the lesion. Images courtesy of Prof. Kaare Weismann, MD,
DMSc, Horsholm Dermatology, Denmark.
reemitted as radiation of higher energy (Raman
scattering), the latter being the least probable event.
A spectroscope separates the returned light into individual wavelengths and assesses it. Raman spectroscopy provides molecular information of a sample
irradiated with laser light as a small fraction is
shifted in frequency (Raman effect). Several studies
have described a characteristic Raman spectra in
NMSC.79–88 A sensitivity of 97% and specificity of
98% were found in a study of 48 BCCs. In vivo
Raman spectroscopy is possible but the precision of
the spectra is low. Two ex vivo studies found distinct
Raman band differences between BCC and normal
skin. Ten samples of BCC suggested this from direct
observations of spectral differences, after reducing
endogenous autofluorescence by a confocal device,
and the two groups were significantly different from
each other.89 In an in vitro study of 15 BCCs, Raman
pseudo-color maps were compared to skin biopsies.
Pseudo-color maps assign areas with similar spectra
with the same color and are generated by multivariate statistical analysis and clustering analysis of
spectra. A prediction model could classify new tissue
samples from BCC lesions from their Raman spectra
with a sensitivity of 100% and specificity of 93%.82
Another study suggested that neural network analysis of near-infrared Fourier transform Raman
spectra may show some potential for ex vivo NMSC
diagnosis79 An in vivo study of 195 patients with a
variety of malignant and benign skin lesions (33
AKs, 32 BCCs) showed promising results for the
screening of skin lesions with near infrared spectroscopy. Spectra were compared to histopathology
in all lesions; univariate statistics showed significant
differences between spectra and healthy skin (normal
skin vs. AKs, BCCs, dysplastic nevi, lentigines, banal
nevi, and seborrheic keratosis) and also between the
spectra themselves. Significant differences, however,
are not always diagnostic differences. A pattern
recognition technique was applied and successfully
discriminated lesions with accuracy higher than
80%.81
Fluorescence Imaging
Specific autofluorescence emitted from malignant
tissue upon radiation with a laser, xenon light, or
halogen lamp has been used to distinguish normal
tissue from cancerous tissue in the head and neck
region.90 Fluorescence imaging is an attractive potential diagnostic technique for skin tumor demarcation. In a study of 21 patients with 80 BCCs, the
fluorescence intensity from BCCs was significantly
lower than surrounding, normal skin91 In a study of
33:10:OCTOBER 2007
1167
DIAGNOSTIC ACCURACY OF NONMELANOMA SKIN CANCER DIAGNOSTIC TESTS
Figure 3. OCT images of BCC. (A) Nodular BCC lesion on the back. (B) PS-OCT image of (A); notice the black band in the
upper part of the lesion, corresponding to lack of birefringence. The white birefringence band is clearly seen in (E). (C) OCT
image of the lesion in (A); notice the disruption of the layered skin structure. (D) OCT image of normal adjacent skin. (E) The
corresponding PS-OCT image. OCT Images from Department of Dermatology, Roskilde Hospital, recorded with an OCT
system developed by Risoe National Laboratory. Images provided by Dr Mette Mogensen.
18 patients with 25 NMSC lesions (20 BCCs and
5 SCCs), a fiber-optic–based fluorimeter collected
spectral data in vivo and microscopic fluorescence ex
vivo.92 The fluorescence of tryptophan moieties in
BCC was 2.9 7 1.9 SD, and for SCC 2 7 0.9 SD
times larger. A marked loss of fluorescence in the
middle of the tumor region was noticed in 78% of
the NMSC lesions due to a decrease in collagen and
elastin cross-links. Another study of 49 patients
(BCC, SCC, AK, and normal skin) compared diagnostic accuracy in laser-induced fluorescence spectroscopy for skin types I to III (Pathak) to determine
the skin colors effect on the results.84 Melanin absorbs fluorescence strongly. Typically normal skin
1168
D E R M AT O L O G I C S U R G E RY
exhibited stronger fluorescence emission than BCC
and SCC. The accuracy of classifying NMSC was
higher (93%) in Type I skin. It has also been suggested that due to the large variation in fluorescence
intensities developing, an algorithm for NMSC is
difficult.93 Bispectral fluorescence imaging combines
skin autofluorescence with d-aminolevulinic acid
(ALA) fluorescence. The agreement between bispectral fluorescence images and the histopathologic tumor boundary of ill-defined BCC in 12 patients with
an aggressive BCC undergoing MMS was examined.86 Only 5 patients had good correlation between histopathology and bispectral images of the
tumor, not significant (p = .057). Another study ap-
MOGENSEN AND JEMEC
TABLE 5. Confocal Microscopy
Authors, year
Gonzalez and
Tannous,
200276
Marra et al.,
2005123
Nori et al.,
2004124
Gerger et al.,
2005125
Number and
pathology of
patients
Question asked
Results
5 BCCs
In vivo CM compared
to histopathology
Characteristic CM features in all
specimens.
3 BCCs
CM compared to
histopathology
Diagnostic accuracy of
5 CM criteria for in
vivo BCC diagnosis
Diagnostic accuracy CM
in MMS untreated
fresh biopsies
Diagnostic accuracy in
vivo CM aluminum
chloride contrast
Ex vivo CM of Stage 1
MMS excisions
Above confirmed.
152 lesions
(83 BCCs and
benign)
20 BCCs
Tannous et al.,
2003126
5 BCCs
Chung et al.,
200475
92 BCCs,
23 SCCs
Sauermann
et al., 2002127
Aghassi et al.,
2000128
12 BCCs
6 AKs, 1 SCC
Diagnostic value of CM
vascular pattern
CM diagnostic features
Sensitivity 93.9% and specificity 78.3% was
found using three criteria and 95.7%
and 82.9%, respectively.
Sensitivity ranging from 44% to 100%
and a specificity of 100% for all
five criteria.
100% sensitivity in Stage 1 MMS and
80% sensitivity in Stage 2.
CM may be an alternative to frozen sections
in large nodular BCC. Difficulties in recognizing SCC in situ and poor image quality.
Vascular pattern of BCC in CM can be
used diagnostically.
CM able to distinguish pathologic features
of epidermal neoplasms: 100% show
nuclear enlargement and pleomorphism.
BCC, basal cell carcinoma; CM, confocal microscopy; MMS, Mohs micrographic surgery.
plied two different algorithms in data analysis of
bispectral fluorescence and showed promising results
in demarcation of skin lesions in 15 BCC patients.94
Also two different fluorescence systems showed a
clear demarcation of BCC in 2 patients with several
BCC lesions.95 In an in vivo study, 55 patients with
oral SCC were studied with respect to endogenous
fluorescence. The intensity of the fluorescence significantly corresponded with the pathologic tumor
and node categories of SCC (po.01).83,96
Terahertz Imaging
Terahertz pulsed imaging (TPI) is a novel, noninvasive, imaging modality. It uses pulses of electromagnetic radiation in the frequency range of 0.1 to
10 THz. Water has strong absorption over the entire
THz range, and water content in the skin is a source
of image contrast.97 TPI has potential use in NMSC
diagnosis. A significant difference between the re-
sponse of THz radiation in normal skin and BCC has
been reported.98,99 In a study of 18 BCCs both in
vivo (5 BCCs) and ex vivo TPI analysis was performed. In vivo regions of contrast were seen in all
THz images and correlated well with histology.100
Electrical Impedance
The impedance of the skin is an electrical entity that
can be described in complex numbers by resistance
and reactance. A pilot study had found statistical
difference in electrical impedance values was found
between BCC and normal skin.101 This was confirmed in a study of 34 BCC patients.102 Statistical
difference in electrical impedance values was found
between BCC and normal skin; however, diagnostic
accuracy could not be assessed. The electrical impedance system was elaborated further, and the
probe equipped with microinvasive electrodes to
bypass the barrier function of the high impedance of
stratum corneum. The lesions were 99 benign nevi,
33:10:OCTOBER 2007
1169
DIAGNOSTIC ACCURACY OF NONMELANOMA SKIN CANCER DIAGNOSTIC TESTS
28 BCCs, and 13 MMS.103 Sensitivity for separation
of BCC from benign nevi was 96% and specificity
86%, when using the noninvasive probe. The invasive probe had higher diagnostic accuracy only in
MM. The choice of electrode can be considered application-dependent. A study of 35 BCC patients
compares electrical impedance, transepidermal water loss (TEWL), and LD in diagnosing nodular and
superficial BCC.104 In accordance with other studies,
statistically significant differences between electrical
impedance and BCC was found (po.001), but no
differences between subtypes were found. In addition,
TEWL and LD values had similar p values in discriminating BCC from normal skin. Increased TEWL values
are ascribed to the decreased barrier function of the
skin due to the pathologic processes of BCC. The assumption is that the increased LD values are due to
increased angiogenesis and vasodilation in BCC.
depth, and the location of the tumor between MRI
and histopathology are in good agreement.107 A
retrospective study of 33 NMSC patients (20 BCCs,
12 SCCs, 1 mixed) estimated accuracy of MRI and
CT; the findings were compared to histopathology.108 Patients were seen for both primary assessment and follow-up. MR and CT localized the
lesions in 29 of 33 patients; of the 4 tumors not identified, three-fourths of patients had mean disease-free
survival at 33 months and the fourth patient developed
recurrence at 52 months. In a study of 35 patients of
whom 18 had perineural spread of BCC and SCC,
based on clinical and histopathologic investigation, CT
and MR109 showed that positive perineural spread inversely correlated with 5-year survival rate. Patients
who were imaging-positive had a 5-year survival of
50% and for imaging-negative patients it was 86%
(p = .049). In this study MRI seems to be informative in
estimating prognosis.
CT, PET, and MR
CT is based on the X-ray principal, whereas PET is
an imaging technique that detects positron release
from radioactive substances and provides cross-sectional physiologic information. PET imaging commonly uses 2-deoxy-2-18F-fluoro-D-glucose (FDG),
a positron-imaging agent, to measure the metabolic
rate of tissue noninvasively. Tumors can be metabolically more active than normal tissue, thus mobilization of the image tracer can be detected by PET
scanning. FDG-PET has been investigated in diagnosis of NMSC.105 Six patients with BCC larger than
1 cm were examined by PET scanning. BCC could
only be identified in 3 of 6 patients. Another study
compared FDG-PET NMSC diagnosis in patients
with head and neck tumors with physical examination, ultrasonography, and CT.106 In a group of
56 patients (43 SCCs), detecting the primary tumor
site with PET had a sensitivity of 95% (95% CI,
80%–98%) and a specificity of 100% (95% CI,
62%–100%). There was no statistical difference
between PET and CT diagnosis. MRI makes use of
the magnetic properties of hydrogen nucleus (the
protons). MRI has been applied to studies of SCC
and morphologic information about the shape, the
1170
D E R M AT O L O G I C S U R G E RY
Conclusion
A broad variety of diagnostic technologies are becoming available for noninvasive diagnosis of
NMSC. The reference standard, skin biopsy and
histopathologic assessment, is, however, not yet to be
replaced, because state-of-the-art diagnostic research
has only been performed with few of the mentioned
new diagnostic methods and technologies. Many of
the technologies seem to offer an adequate diagnostic accuracy, especially as a supplement to clinical
diagnosis. For all areas of new diagnostic tests and
technologies in NMSC the clinical role remains to be
established in larger, independent studies.
References
1. Rigel DS, Friedman R, Dzubow LM, et al. Cancer of the skin.
Philadelphia: Elsevier Saunders; 2004. p. 736S.
2. Hallock GG, Lutz DA. Prospective study of the accuracy of the
surgeon’s diagnosis in 2000 excised skin tumors. Plast Reconstr
Surg 1998;101:1255–61.
3. Marghoob AA, Swindle LD, Moricz CZ, et al. Instruments and
new technologies for the in vivo diagnosis of melanoma. J Am
Acad Dermatol 2003;49:777–97.
4. Gluud C, Gluud LL. Evidence based diagnostics. BMJ
2005;330:724–6.
MOGENSEN AND JEMEC
5. Knottnerus JA. The evidence base of clinical diagnosis. London:
BMJ Books; 2002.
6. Whited JD, Horner RD, Hall RP, et al. The influence of history
on interobserver agreement for diagnosing actinic keratoses and
malignant skin lesions. J Am Acad Dermatol 1995;33:603–7.
7. Whited JD, Hall RP. Diagnostic accuracy and precision in assessing dermatologic disease: problem or promise? Arch Dermatol 1997;133:1409–15.
8. Schwartzberg JB, Elgart GW, Romanelli P, et al. Accuracy and
predictors of basal cell carcinoma diagnosis. Dermatol Surg
2005;31:534–7.
9. Morrison A, O’Loughlin S, Powell FC. Suspected skin malignancy: a comparison of diagnoses of family practitioners and
dermatologists in 493 patients. Int J Dermatol 2001;40:104–7.
10. Leffell DJ, Chen YT, Berwick M, et al. Interobserver agreement
in a community skin cancer screening setting. J Am Acad
Dermatol 1993;28:1003–5.
11. Har-Shai Y, Hai N, Taran A, et al. Sensitivity and positive predictive values of presurgical clinical diagnosis of excised benign
and malignant skin tumors: a prospective study of 835 lesions in
778 patients. Plast Reconstr Surg 2001;108:1982–9.
12. Ek EW, Giorlando F, Su SY, et al. Clinical diagnosis of skin
tumours: how good are we? ANZ J Surg 2005;75:415–20.
13. Davis DA, Donahue JP, Bost JE, et al. The diagnostic concordance of actinic keratosis and squamous cell carcinoma. J Cutan
Pathol 2005;32:546–51.
14. Cooper SM, Wojnarowska F. The accuracy of clinical diagnosis
of suspected premalignant and malignant skin lesions in renal
transplant recipients. Clin Exp Dermatol 2002;27:436–8.
15. Oliveria SA, Chau D, Christos PJ, et al. Diagnostic accuracy of
patients in performing skin self-examination and the impact of
photography. Arch Dermatol 2004;140:57–62.
16. Russell EB, Carrington PR, Smoller BR. Basal cell carcinoma: a
comparison of shave biopsy versus punch biopsy techniques in
subtype diagnosis. J Am Acad Dermatol 1999;41:69–71.
17. Bakis S, Irwig L, Wood G, et al. Exfoliative cytology as a diagnostic test for basal cell carcinoma: a meta-analysis. Br J
Dermatol 2004;150:829–36.
18. Vega-Memije E, Martinez-de-Larios N, Waxtein LM, et al.
Cytodiagnosis of cutaneous basal and squamous cell carcinoma.
Int J Dermatol 2000;39:116–20.
19. Saldanha G, Fletcher A, Slater DN. Basal cell carcinoma: a dermatopathological and molecular biological update. Br J Dermatol 2003;148:195–202.
20. Walling HW, Fosko SW, Geraminejad PA, et al. Aggressive basal
cell carcinoma: presentation, pathogenesis, and management.
Cancer Metastasis Rev 2004;23:389–402.
24. Kamiya M, Ichiki Y, Kamiya H, et al. Detection of nonmelanoma
skin cancer micrometastases in lymph nodes by using reverse
transcriptase-polymerase chain reaction for keratin 19 mRNA.
Br J Dermatol 2003;149:998–1005.
25. Biesterfeld S, Josef J. Differential diagnosis of keratoacanthoma
and squamous cell carcinoma of the epidermis by MIB-1
immunohistometry. Anticancer Res 2002;22:3019–23.
26. Trotter MJ, Bruecks AK. Interpretation of skin biopsies by general pathologists: diagnostic discrepancy rate measured by
blinded review. Arch Pathol Lab Med 2003;127:1489–92.
27. Lacour JP. Carcinogenesis of basal cell carcinomas: genetics and
molecular mechanisms. Br J Dermatol 2002;146(Suppl 61):17–9.
28. Ortonne JP. From actinic keratosis to squamous cell carcinoma.
Br J Dermatol 2002;146(Suppl 61):20–3.
29. Park WS, Lee HK, Lee JY, et al. p53 mutations in solar keratoses.
Hum Pathol 1996;27:1180–4.
30. Wong CS, Strange RC, Lear JT. Basal cell carcinoma. BMJ
2003;327:794–8.
31. Jovanovic DL, Katic V, Jovanovic B. Value of preoperative determination of skin tumor thickness with 20-MHz ultrasound.
Arch Dermatol 2005;141:269–70.
32. Karaman GC, Karaman CZ, Sendur N, et al. Power Doppler
ultrasonography for the evaluation of skin tumors other than
malignant melanoma. Eur Radiol 2001;11:1111–6.
33. Moore JV, Allan E. Pulsed ultrasound measurements of depth
and regression of basal cell carcinomas after photodynamic
therapy. relationship to probability of 1-year local control.
Br J Dermatol 2003;149:1035–40.
34. Allan E, Pye DA, Levine EL, et al. Non-invasive pulsed
ultrasound quantification of the resolution of basal cell carcinomas after photodynamic therapy. Lasers Med 2002;17:
230–7.
35. Gupta AK, Turnbull DH, Foster FS, et al. High frequency
40-MHz ultrasound: a possible noninvasive method for the
assessment of the boundary of basal cell carcinomas. Dermatol
Surg 1996;22:131–6.
36. Harland CC, Bamber JC, Gusterson BA, et al. High frequency,
high resolution B-scan ultrasound in the assessment of skin tumours. Br J Dermatol 1993;128:525–32.
37. Gropper CA, Stiller MJ, Shupack JL, et al. Diagnostic highresolution ultrasound in dermatology. Int J Dermatol 1993;
32:243–50.
38. Jemec GB, Gniadecka M, Ulrich J. Ultrasound in dermatology.
Part I. High frequency ultrasound. Eur J Dermatol 2000;10:
492–7.
21. Lind AC, Bewtra C, Healy JC, et al. Prospective peer review in
surgical pathology. Am J Clin Pathol 1995;104:560–6.
39. Lassau N, Spatz A, Avril MF, et al. Value of high-frequency US
for preoperative assessment of skin tumors. Radiographics
1997;17:1559–65.
22. Olhoffer IH, Lazova R, Leffell DJ. Histopathologic misdiagnoses
and their clinical consequences. Arch Dermatol 2002;138:1381–3.
40. Marques J, Cueto L, Roldan F, et al. Ultrasound study of skin
tumors. Radiologia 2002;44:55–60.
23. Brochez L, Verhaeghe E, Grosshans E, et al. Inter-observer
variation in the histopathological diagnosis of clinically suspicious pigmented skin lesions. J Pathol 2002;196:459–66.
41. Vaillant L, Grognard C, Machet L, et al. High resolution ultrasound imaging to assess skin tumors prior to cryosurgery. Ann
Dermatol Venereol 1998;125:500–4.
33:10:OCTOBER 2007
1171
DIAGNOSTIC ACCURACY OF NONMELANOMA SKIN CANCER DIAGNOSTIC TESTS
42. Gross U, Suter L, Hundeiker M. 20-MHz sonography as a
planning aid in the therapy of skin tumours. Akt Dermatol
1993;19:32–5.
43. Costa P, Ghillani M, Papadia F, et al. Superficial ultrasound in
the assessment of skin tumours: indications and limitations. Riv
Ital Chir Plast 1992;24:263–9.
44. Nitsche N, Hoffmann K, Iro H. Ultrasound diagnosis of skin
tumours. HNO 1992;40:97–100.
45. Edwards C, Al-Aboosi MM, Marks R. The use of A-scan ultrasound in the assessment of small skin tumours. Br J Dermatol
1989;121:297–304.
46. Hoffmann K, Stucker M, el-Gammal S, et al. [Digital 20 MHz
sonography of basalioma in the B-scan]. Hautarzt 1990;41:
333–9.
47. el Gammal S, Auer T, Hoffmann K, et al. High-resolution ultrasound of the human epidermis. In: Non-invasive methods of
the skin. Boca Raton (FL): CRC Press; 1995.
48. el Gammal S, El Gammal C, Altmeyer PJ, et al. High resolution
sonography of the skin. In: Non-invasive methods of the skin.
2nd ed. Boca Raton (FL): CRC Press; 2006.
49. Serup J, Keiding J, Fullerton A, et al. High-frequency ultrasound
examination of the skin. In: Non-invasive methods of the skin.
2nd ed. Boca Raton (FL): CRC Press; 2006.
50. Wortsman XC, Holm EA, Wulf HC, et al. Real-time spatial
compound ultrasound imaging of skin. Skin Res Technol
2004;10:23–31.
51. Bessoud B, Lassau N, Koscielny S, et al. High-frequency sonography and color Doppler in the management of pigmented skin
lesions. Ultrasound Med Biol 2003;29:875–9.
52. Rallan D, Harland CC. Ultrasound in dermatologyFbasic principles and applications. Clin Exp Dermatol 2003;28:632–8.
53. Ruocco E, Argenziano G, Pellacani G, et al. Noninvasive imaging of skin tumors. Dermatol Surg 2004;30:301–10.
54. Schmid-Wendtner MH, Burgdorf W. Ultrasound scanning in
dermatology. Arch Dermatol 2005;141:217–24.
55. Bjerring P, Obitz ER, Cotton S. In vivo spectrophotometric
evaluation of skin tumours using a new chromophore imaging
system SIAscope. Melanoma Res 2001;11:s180.
56. Argenziano G, Zalaudek I, Corona R, et al. Vascular structures
in skin tumors: a dermoscopy study. Arch Dermatol
2004;140:1485–9.
57. Zalaudek I, Argenziano G, Leinweber B, et al. Dermoscopy of
Bowen’s disease. Br J Dermatol 2004;150:1112–6.
58. Kreusch JF. Vascular patterns in skin tumors. Clin Dermatol
2002;20:248–54.
59. Tycho A, Andersen P, Thrane L, et al. Optical coherence tomography in dermatology. In: Non-invasive methods and the
skin. 2nd ed. Boca Raton (FL): CRC Press; 2006.
60. Fujimoto JG, Pitris C, Boppart SA, et al. Optical coherence tomography: an emerging technology for biomedical imaging and
optical biopsy. Neoplasia 2000;2:9–25.
1172
D E R M AT O L O G I C S U R G E RY
61. Fujimoto JG. Optical coherence tomography for ultrahigh
resolution in vivo imaging. Nat Biotechnol 2003;21:
1361–7.
62. Gambichler T, Moussa G, Sand M, et al. Applications of optical
coherence tomography in dermatology. J Dermatol Sci
2005;40:85–94.
63. Pierce MC, Strasswimmer J, Park BH, et al. Advances in optical
coherence tomography imaging for dermatology.
2004;123:458–63.
64. Welzel J, Lankenau E, Birngruber R, et al. Optical coherence
tomography of the human skin. J Am Acad Dermatol
1997;37:958–63.
65. Welzel J. Optical coherence tomography in dermatology:
a review. Skin Res Technol 2001;7:1–9.
66. Steiner R, Kunzi RK, Scharffetter KK. Optical coherence tomography: clinical applications in dermatology. Med Laser
Application 2003;18:249–59.
67. Olmedo JM, Warschaw KE, Schmitt JM, et al. Optical
coherence tomography for the characterization of basal cell
carcinoma in vivo: a pilot study. J Am Acad Dermatol 2006;
55:408–12.
68. Gambichler T, Boms S, Stucker M, et al. Comparison of histometric data obtained by optical coherence tomography and
routine histology. J Biomed Opt 2005;10:44008.
69. Bechara FG, Gambichler T, Stucker M, et al. Histomorphologic
correlation with routine histology and optical coherence tomography. Skin Res Technol 2004;10:169–73.
70. Jensen LK, Thrane L, Andersen PE, et al. Optical coherence tomography in clinical examination of non-pigmented skin malignancies. Proc of SPIE-OSA Biomed Optics. SPIE
2003;5140:160–7.
71. Strasswimmer J, Pierce MC, Park BH, et al. Polarization-sensitive optical coherence tomography of invasive basal cell carcinoma. J Biomed Opt 2004;9:292–8.
72. Buchwald HJ, Muller A, Kampmeier J, et al. Optical coherence
tomography versus ultrasound biomicroscopy of conjunctival
and eyelid lesions. Klin Monatsblatter Augenheilkunde
2003;220:822–9.
73. Kollias N, Stamatas GN. Optical non-invasive approaches to
diagnosis of skin diseases. J Invest Dermatol Symp Proc
2002;7:64–75.
74. Selkin B, Rajadhyaksha M, Gonzalez S, et al. In vivo confocal
microscopy in dermatology. Dermatol Clin 2001;19:369–77.
75. Chung VQ, Dwyer PJ, Nehal KS, et al. Use of ex vivo confocal
scanning laser microscopy during Mohs surgery for nonmelanoma skin cancers. Dermatol Surg 2004;30:1470–8.
76. Gonzalez S, Tannous Z. Real-time, in vivo confocal reflectance
microscopy of basal cell carcinoma. J Am Acad Dermatol
2002;47:869–74.
77. Gonzalez S, Gilaberte-Calzada Y, Gonzalez-Rodriguez A, et al.
In vivo reflectance-mode confocal scanning laser microscopy in
dermatology. Adv Dermatol 2004;20:371–87.
78. Suihko C, Swindle LD, Thomas SG, et al. Fluorescence fibreoptic confocal microscopy of skin in vivo: microscope and fluorophores. Skin Res Technol 2005;11:254–67.
MOGENSEN AND JEMEC
79. Gniadecka M, Wulf HC, Nielsen OF, et al. Distinctive molecular
abnormalities in benign and malignant skin lesions: studies by
Raman spectroscopy. Photochem Photobiol 1997;66:418–23.
80. Gniadecka M, Philipsen PA, Sigurdsson S, et al. Melanoma diagnosis by Raman spectroscopy and neural networks: structure
alterations in proteins and lipids in intact cancer tissue. J Invest
Dermatol 2004;122:443–9.
81. McIntosh LM, Jackson M, Mantsch HH, et al. Infrared spectra
of basal cell carcinomas are distinct from non-tumor-bearing
skin components. J Invest Dermatol 1999;112:951–6.
82. Nijssen A, Bakker Schut TC, Heule F, et al. Discriminating basal
cell carcinoma from its surrounding tissue by Raman spectroscopy. J Invest Dermatol 2002;119:64–9.
83. Onizawa K, Yoshida H, Saginoya H. Chromatic analysis of autofluorescence emitted from squamous cell carcinomas arising in
the oral cavity: a preliminary study. Int J Oral Maxillofac Surg
2000;29:42–6.
84. Panjehpour M, Julius CE, Phan MN, et al. Laser-induced fluorescence spectroscopy for in vivo diagnosis of non-melanoma
skin cancers. Lasers Surg Med 2002;31:367–73.
85. Sigurdsson S, Philipsen PA, Hansen LK, et al. Detection of skin
cancer by classification of Raman spectra. IEEE Trans Biomed
Eng 2004;51:1784–93.
86. Stenquist B, Ericson MB, Strandeberg C, et al. Bispectral fluorescence imaging of aggressive basal cell carcinoma combined
with histopathological mapping: a preliminary study indicating a
possible adjunct to Mohs micrographic surgery. Br J Dermatol
2006;154:305–9.
tion for correlation with histopathologic extent of basal cell
carcinoma. J Biomed Opt 2005;10:034009.
95. Andersson-Engels S, Canti G, Cubeddu R, et al. Preliminary
evaluation of two fluorescence imaging methods for the detection and the delineation of basal cell carcinomas of the skin.
Lasers Surg Med 2000;26:76–82.
96. Onizawa K, Okamura N, Saginoya H, et al. Characterization of
autofluorescence in oral squamous cell carcinoma. Oral Oncol
2003;39:150–6.
97. Pickwell E, Cole BE, Fitzgerald AJ, et al. In vivo study of human
skin using pulsed terahertz radiation. Phys Med 2004;49:1595–607.
98. Woodward RM, Cole BE, Wallace VP, et al. Terahertz pulse
imaging in reflection geometry of human skin cancer and skin
tissue. Phys Med 2002;47:3853–63.
99. Woodward RM, Wallace VP, Pye RJ, et al. Terahertz pulse
imaging of ex vivo basal cell carcinoma. J Invest Dermatol
2003;120:72–8.
100. Wallace VP, Fitzgerald AJ, Shankar S, et al. Terahertz pulsed
imaging of basal cell carcinoma ex vivo and in vivo. Br J
Dermatol 2004;151:424–32.
101. Emtestam L, Nicander I, Stenstrom M, et al. Electrical impedance of nodular basal cell carcinoma: a pilot study. Dermatology
1998;197:313–6.
102. Aberg P, Nicander I, Holmgren U, et al. Assessment of skin lesions and skin cancer using simple electrical impedance indices.
Skin Res Technol 2003;9:257–61.
87. Svanberg K, Wang I, Colleen S, et al. Clinical multi-colour fluorescence imaging of malignant tumoursFinitial experience.
Acta Radiol 1998;39:2–9.
103. Aberg P, Geladi P, Nicander I, et al. Non-invasive and microinvasive electrical impedance spectra of skin cancerFa comparison between two techniques. Skin Res Technol
2005;11:281–6.
88. McIntosh LM, Summers R, Jackson M, et al. Towards non-invasive screening of skin lesions by near-infrared spectroscopy.
J Invest Dermatol 2001;116:175–81.
104. Kuzmina N, Talme T, Lapins J, et al. Non-invasive preoperative
assessment of basal cell carcinoma of nodular and superficial
types. Skin Res Technol 2005;11:196–200.
89. Choi J, Choo J, Chung H, et al. Direct observation of spectral
differences between normal and basal cell carcinoma (BCC)
tissues using confocal Raman microscopy. Biopolymers
2005;77:264–72.
105. Fosko SW, Hu W, Cook TF, et al. Positron emission tomography
for basal cell carcinoma of the head and neck. Arch Dermatol
2003;139:1141–6.
90. Inaguma M, Hashimoto K. Porphyrin-like fluorescence in oral
cancer: in vivo fluorescence spectral characterization of lesions
by use of a near-ultraviolet excited autofluorescence diagnosis
system and separation of fluorescent extracts by capillary
electrophoresis. Cancer 1999;86:2201–11.
91. Na R, Stender IM, Wulf HC. Can autofluorescence demarcate
basal cell carcinoma from normal skin? A comparison with
protoporphyrin IX fluorescence. Acta Derm Venereol
2001;81:246–9.
92. Brancaleon L, Durkin AJ, Tu JH, et al. In vivo fluorescence
spectroscopy of nonmelanoma skin cancer. Photochem Photobiol
2001;73:178–83.
93. Sterenborg NJ, Thomsen S, Jacques SL, et al. In vivo fluorescence
spectroscopy and imaging of human skin tumors. Dermatol Surg
1995;21:821–2.
94. Ericson MB, Uhre J, Strandeberg C, et al. Bispectral fluorescence
imaging combined with texture analysis and linear discrimina-
106. Sigg MB, Steinert H, Gratz K, et al. Staging of head and neck
tumors: [18F] fluorodeoxyglucose positron emission tomography compared with physical examination and conventional
imaging modalities. J Oral Maxillofac Surg 2003;61:1022–9.
107. Querleux B. Nuclear magnetic resonance (NMR) examination of
the skin in vivo. In: Non-invasive methods of the skin. 2nd ed.
Boca Raton (FL): CRC Press; 2006.
108. Lanka B, Turner M, Orton C, et al. Cross-sectional imaging in
non-melanoma skin cancer of the head and neck. Clin Radiol
2005;60:869–77.
109. Williams LS, Mancuso AA, Mendenhall WM. Perineural spread
of cutaneous squamous and basal cell carcinoma. CT and MR
detection and its impact on patient management and prognosis.
Int J Radiat Oncol Biol Phys 2001;49:1061–9.
110. Whited JD, Hall RP, Simel DL, et al. Primary care clinicians’
performance for detecting actinic keratoses and skin cancer. Arch
Intern Med 1997;157:985–90.
33:10:OCTOBER 2007
1173
DIAGNOSTIC ACCURACY OF NONMELANOMA SKIN CANCER DIAGNOSTIC TESTS
111. Westbrook RH, Goyal N, Gawkrodger DJ. Diagnostic accuracy
for skin cancer: comparison of general practitioner with dermatologist and dermatopathologist. J Dermatolog Treat
2006;17:57–8.
122. Chin CW, Foss AJ, Stevens A, et al. Differences in the vascular
patterns of basal and squamous cell skin carcinomas explain
their differences in clinical behaviour. J Pathol 2003;200:
308–13.
112. Renshaw AA, Pinnar NE, Jiroutek MR, et al. Blinded review as a
method for quality improvement in surgical pathology. Arch
Pathol Lab Med 2002;126:961–3.
123. Marra DE, Torres A, Schanbacher CF, et al. Detection of residual
basal cell carcinoma by in vivo confocal microscopy. Dermatol
Surg 2005;31:538–41.
113. Bertolotto M, Serafini G, Dogliotti L, et al. Primary and secondary malignancies of the penis: ultrasound features. Abdom
Imaging 2005;30:108–12.
124. Nori S, Rius-Diaz F, Cuevas J, et al. Sensitivity and specificity of
reflectance-mode confocal microscopy for in vivo diagnosis of
basal cell carcinoma: a multicenter study. J Am Acad Dermatol
2004;51:923–30.
114. Lont AP, Besnard AP, Gallee MP, et al. A comparison of physical
examination and imaging in determining the extent of primary
penile carcinoma. BJU Int 2003;91:493–5.
115. Stucker M, Horstmann I, Nuchel C, et al. Blood flow compared
in benign melanocytic naevi, malignant melanomas and basal
cell carcinomas. Clin Exp Dermatol 1999;24:107–11.
116. Schroder RJ, Maurer J, Zlowodski M, et al. Vascularization of
malignant and benign skin tumours measured by D-galactosebased signal-enhanced colour Doppler sonography. Acta Radiol
2001;42:294–301.
117. Zalaudek I. Dermoscopy subpatterns of nonpigmented skin tumors. Arch Dermatol 2005;141:532.
118. Menzies SW. Dermoscopy of pigmented basal cell carcinoma.
Clin Dermatol 2002;20:268–9.
119. Zalaudek I, Argenziano G, Soyer HP, et al. Three-point checklist
of dermoscopy: an open internet study. Br J Dermatol
2006;154:431–7.
125. Gerger A, Horn M, Koller S, et al. Confocal examination of
untreated fresh specimens from basal cell carcinoma: implications for microscopically guided surgery. Arch Dermatol
2005;141:1269–74.
126. Tannous Z, Torres A, Gonzalez S. In vivo real-time confocal
reflectance microscopy: a noninvasive guide for Mohs
micrographic surgery facilitated by aluminum chloride,
an excellent contrast enhancer. Dermatol Surg 2003;29:
839–46.
127. Sauermann K, Gambichler T, Wilmert M, et al. Investigation
of basal cell carcinoma [correction of carcionoma] by confocal
laser scanning microscopy in vivo. Skin Res Technol 2002;8:
141–7.
128. Aghassi D, Anderson RR, Gonzalez S. Confocal laser microscopic imaging of actinic keratoses in vivo: a preliminary report.
J Am Acad Dermatol 2000;43:42–8.
120. Otis LL, Piao D, Gibson CW, et al. Quantifying labial blood flow
using optical Doppler tomography. Oral Surg Oral Med Pathol
Oral Radiol Endod 2004;98:189–94.
121. Newell B, Bedlow AJ, Cliff S, et al. Comparison of the microvasculature of basal cell carcinoma and actinic keratosis using
intravital microscopy and immunohistochemistry. Br J Dermatol
2003;149:105–10.
1174
D E R M AT O L O G I C S U R G E RY
Address correspondence and reprint requests to: Mette
Mogensen, MD, Department of Dermatology, University
of Copenhagen, Roskilde Hospital, K^gevej 7-13, DK4000 Roskilde, Denmark, or e-mail: [email protected]