Professional Documents
Culture Documents
A Single Typical Trichoscopic Feature Is Predictive of Tinea Capitis: A Prospective Multicentre Study
A Single Typical Trichoscopic Feature Is Predictive of Tinea Capitis: A Prospective Multicentre Study
Summary
Correspondence Background Specific trichoscopic signs of tinea capitis (TC) were first described in
Florie Dhaille. 2008. The accuracy of this diagnostic tool has not been evaluated.
E-mail: floriedhaille@gmail.com. Objectives To assess the diagnostic accuracy of trichoscopy.
Methods A prospective, multicentre study was done between March 2015 and
Accepted for publication
March 2017 at the dermatology departments of four French university medical
4 March 2019
centres. Patients with a presumed diagnosis of TC were included. Trichoscopy
Funding sources was considered to be positive if at least one specific trichoscopic sign was
Grant from PHYTO laboratory. observed. Trichoscopy results were compared with the gold standard for diagno-
sis of TC (mycological culture).
Conflicts of interest Results One hundred patients were included. Culture was positive for 53
None declared.
patients and negative for 47. The sensitivity of trichoscopy was 94% [95%
DOI 10.1111/bjd.17866 confidence interval (CI) 88–100], specificity was 83% (95% CI 72–94), posi-
tive predictive value was 92% and negative predictive value was 86%. Comma
hairs, corkscrew hairs, zigzag hairs, Morse-code-like hairs and whitish sheath
were significantly more frequent in patients with a positive mycological cul-
ture (P < 0001). Comma hairs were more frequent in patients with Trichophy-
ton TC (P = 0026), and zigzag hairs were more frequent in patients with
Microsporum TC (P < 0001). Morse-code-like hair was not observed in any
patients with Trichophyton TC and therefore appears to be highly specific for
Microsporum TC.
Conclusions The presence of a single trichoscopic finding is predictive of TC. Tri-
choscopy is a useful, rapid, painless, highly sensitive tool for the diagnosis of TC
– even for dermoscopists with little experience of trichoscopy. It enhances physi-
cians’ ability to make treatment decisions.
Tinea capitis (TC) is a common, contagious, dermatophyte (African, Asian or Caucasian), and any international travel in the
infection of the scalp that occurs predominantly in children.1 previous 6 months. A Wood lamp examination was performed.
The diagnosis of TC is based on a clinical assessment and a The trichoscopy assessment was performed by dermatology res-
Wood lamp examination.2 However, this diagnosis must be idents who had no previous training in trichoscopy or the iden-
confirmed by a mycological culture before systemic treatment tification of trichoscopic signs of TC, and had only been shown
is initiated. The culture may take up to 6 weeks – delaying the trichoscopic aspects of TC described in the literature: photos
treatment initiation and increasing the risk of contagion.3 or drawings of comma hairs, corkscrew hairs, zigzag hairs,
Specific trichoscopic signs of noninflammatory TC were first Morse-code-like hairs and whitish sheath. Trichoscopy was per-
described in 2008, mostly with video-dermoscopy.4–12 Tri- formed at a magnification of 910 with a handheld, nonpolar-
choscopy (dermoscopy of hair and scalp) with a pocket der- ized light dermoscope (Deltaâ 20; HEINE Optotechnik,
moscope may be a useful, rapid, painless tool for diagnosing Herrsching, Germany), in the absence of immersion liquid. The
TC. However, the accuracy of trichoscopy for diagnosing TC dermoscope was cleaned after each assessment with a quater-
has not previously been evaluated. We therefore conducted a nary ammonium surface disinfectant. Photographs of the tricho-
prospective, nonrandomized, multicentre, noninferiority study scopic features were taken directly through the dermoscope
to compare the accuracy of trichoscopy with that of the gold- with a smartphone, and embedded in an electronic case report
standard method (mycological culture) for the diagnosis of form (eCRF). All the trichoscopic features highlighted during
TC. the examination were reported on the eCRF.
A dermatologist with expertise in hair disorders (referred to
hereafter as the expert) and training in the identification of
Patients and methods
TC-specific signs reviewed the photographs and reported the
findings on the eCRF. The expert was blinded to the culture
Design overview
results and the dermatologist’s assessment of the scalp. The tri-
The study was performed between March 2015 and March choscopy assessment was considered to be positive by the resi-
2017 at the dermatology departments of four university medi- dent and the expert if at least one TC-specific sign described
cal centres in France. The protocol was approved by the local in the literature was present on at least two hairs.
institutional review board (CPP Nord Ouest II, Amiens, The sample for mycological culture (skin scales and plucked
France) and registered at ClinicalTrials.gov (NCT 02550496). hairs) was collected by the dermatologist or a mycologist after
Patients were included after they had provided verbal and the trichoscopy assessment. The latter assessment sometimes
written consent. For patients < 18 years of age, written, guided the practitioner’s choice of areas to be sampled. Myco-
informed consent was obtained from one of the parents or logical examination (direct microscopic observation with 40%
legal guardians. potassium hydroxide, and culture in mycobiotic agar incu-
bated at 24°C) was performed in all patients. TC was con-
firmed if the mycological culture was positive.
Setting and participants
The study population consisted of consecutive ambulatory
Outcomes and measurements
patients treated at or referred to one of the four investigating
centres for suspected TC. Included patients had one or more The primary outcome measures were the sensitivity, speci-
signs of TC: dandruff, scales with pruritus or scaling alopecia. ficity, positive predictive value (PPV) and negative predictive
There was no age restriction on inclusion. Patients having value (NPV) of the trichoscopy procedure (on the basis of the
taken systemic antifungal medications in the 2 weeks prior to resident’s data and those of the expert). The trichoscopy
observation were excluded, as were patients with inflamma- results were compared with the gold standard for the diagno-
tory TC (e.g. patients with kerions, in whom a trichoscopy sis of TC (mycological culture).
assessment tends only to reveal old crusts, suppuration or The secondary outcome measures included the correlation
pustules). between the trichoscopic aspects of TC with the type of hair
or the dermatophyte species (based on the expert analysis),
the degree of concordance between the resident and the
Study procedures
expert, the sensitivity and specificity of the direct mycological
During a consultation with a dermatology resident, standard examination, and (for Microsporum TC) the sensitivity and speci-
clinical data were recorded: the patient’s age and hair type ficity of the Wood lamp assessment.
Cases analysed
n = 100
(a) (b)
(c) (d)
Fig 2. Trichoscopic images of tinea capitis (TC) acquired with a handheld dermoscope (magnification 910). (a) TC caused by Microsporum canis:
zigzag hair (grey arrows) and whitish sheath (red arrows). (b) TC caused by Trichophyton tonsurans: corkscrew hair (circles). (c) TC caused by M.
canis: Morse-code-like hair (black arrows) and whitish sheath (red arrows). (d) TC caused by T. tonsurans: comma hair (squares).
Data in parentheses are the 95% confidence interval. PPV, positive predictive value; NPV, negative predictive value.
and lack of equipment in four cases). The technique’s sensitiv- hairs might represent a specific trichoscopic pattern for TC.
ity and specificity (calculated after excluding cases of Trichophy- Their finding might have been due to a lack of power in a
ton TC) were, respectively, 708% and 903%. study with a sample size half that of the present work.
Of the 53 patients with a PC, 26 were also positive in the The 2008 report by Slowinska et al. on two patients was the
direct examination. The sensitivity of direct examination was first to state that comma hair was a distinctive marker for TC.4
491% (95% CI 361–621) and the specificity was 979% Since then, several studies have described other specific signs:
(95% CI 889–996%). corkscrew hairs,8,9,17 zigzag hairs,11,18 horizontal white bands
(referred to as ‘Morse-code-like hairs’),12 and weak-looking
whitish or translucent hairs. These white hairs may result from
Discussion
fungal invasion of the whole hair shaft.18–20 Short broken
The primary objective of this prospective study was to evalu- hairs, black dots or dystrophic hairs are frequently described
ate the accuracy of trichoscopy for diagnosing TC. We com- but are not specific for TC.21
pared the results of the trichoscopy examination with those of In the literature, some researchers have sought to establish a
the gold-standard technique (i.e. mycological culture); the correlation between specific signs on the one hand and the
sensitivity and specificity were, respectively, 94% and 83% for type of hair or the type of parasite on the other hand. How-
the expert, and 89% and 77% for the residents who included ever, the sample sizes were too small to enable statistically sig-
the patients. The correlation coefficient for concordance was nificant conclusions to be drawn. Indeed, comma hairs have
high between the resident and the expert, even when calcu- been mostly described in cases of ectothrix infection.4,7,20,21
lated for each individual sign; this value indicates that tri- Corkscrew hairs were described for the first time by Hughes
choscopy is reproducible. We found that the presence of a et al. in 2011 in six children, although the researchers could
single trichoscopic finding (comma, corkscrew, zigzag or not link this sign to the type of hair (African) or the incrimi-
Morse-code-like hair or whitish sheath) was predictive of TC. nated dermatophyte (T. soudanense).7 Corkscrew hairs seem to
The accuracy of trichoscopy assessment had not previously be short comma-shaped hairs with a more exaggerated coiled
been evaluated. appearance, resulting from cracking and bending of the hair
The clinical and demographic characteristics of our study shaft filled with the hyphae.4 Indeed, Lu et al. described a
population were similar to those reported in the literature; direct microscopic examination with a 10% potassium hydrox-
our population mainly comprised children < 13 years of age ide smear of corkscrew hair, which revealed that curved hairs
(with a median age of 624 years for the patients in the PC were filled with extremely high numbers of endothrix spores
group). In the literature, 80% of cases of TC are said to occur (TC caused by T. violaceum).22 Zigzag hairs have been described
in children < 10 years of age, with around half of these in in cases of ectothrix infection.18,23 Lastly, Morse-code-like hair
children < 5 years of age.13,14 In the PC group, we did not is mostly frequently described as a dermoscopic finding in the
observe the female predominance mentioned in some litera- setting of an ectothrix infection.12,18,23
ture studies; on the contrary, 62% of the cases occurred in In the present study, comma hair was more frequently
males.13,15 As recently described in the literature, we noted observed in the subset of patients with African-type hair than
the predominance of anthropophilic dermatophytes (51%, in the subset with Caucasian-type hair. Comma hair and cork-
mainly T. tonsurans) and zoophilic dermatophytes (45%, mainly screw hair were more frequently observed in the ‘endothrix
M. canis).13 We found only one similar study in the literature: TC’ subgroup. Zigzag hair and Morse-code-like hair were
a prospective study of 50 children with a possible diagnosis of more frequently observed in the subset of patients with Cau-
TC.16 However, the researchers used a video-dermoscope casian-type hair, and in the ‘ectothrix TC’ subgroup. Morse-
(magnification 940–9400), which is not widely available in code-like hair was not observed in any patients with Trichophy-
routine clinical practice. The mycological culture was positive ton TC, and so this feature appears to be highly specific for
for 38 of the 50 patients. The researchers found very few Microsporum TC. In cases of infection by an endothrix-type para-
specific signs, and concluded that the association of perifollic- site, the hair breaks close to the scalp (comma and corkscrew
ular scaling with any type of dystrophic hair or with broken hair). In contrast, the hair breaks further from the scalp in
cases of infection by an ectothrix-type parasite, and thus may 5 Sandoval AB, Ortiz JA, Rodriguez JM et al. [Dermoscopic pattern in
be have a zigzag shape or a Morse-code-like aspect. Hence, tinea capitis]. Rev Iberoam Micol 2010; 27:151–2 (in Spanish).
the results of the trichoscopy assessment may give the physi- 6 Rudnicka L, Olszewska M, Rakowska A, Slowinska M. Trichoscopy
update 2011. J Dermatol Case Rep 2011; 5:82–8.
cian an idea of the type of dermatophyte present. The present
7 Hughes R, Chiaverini C, Bahadoran P, Lacour JP. Corkscrew hair:
study is the first to have highlighted these factors, which may a new dermoscopic sign for diagnosis of tinea capitis in black
have value in routine dermatological practice. Nevertheless, children. Arch Dermatol 2011; 147:355–6.
mycological assessments and cultures remain essential for con- 8 Vazquez-Lopez F, Palacios-Garcia L, Argenziano G. Dermoscopic
firming the diagnosis and identifying the disease-causing corkscrew hairs dissolve after successful therapy of Trichophyton vio-
pathogen. laceum tinea capitis: a case report. Australas J Dermatol 2012; 53:118–
A Wood lamp examination was not performed on 31% of 19.
9 Pinheiro AM, Lobato LA, Varella TC. Dermoscopy findings in tinea
the patients, mainly due to a lack of time. For the tested
capitis: case report and literature review. An Bras Dermatol 2012;
patients, the relatively low sensitivity and specificity (708% 87:313–14.
and 903%, respectively, after the exclusion of cases of Tri- 10 Mapelli ET, Gualandri L, Cerri A, Menni S. Comma hairs in tinea
chophyton TC) highlighted the limitations of this examination. capitis: a useful dermatoscopic sign for diagnosis of tinea capitis.
The sensitivity of direct examination was 491% (95% CI Pediatr Dermatol 2012; 29:223–4.
361–621) and the specificity was 979% (95% CI 889– 11 El-Taweel AE, El-Esawy F, Abdel-Salam O. Different trichoscopic
996). The value of trichoscopy in diagnosing TC is further features of tinea capitis and alopecia areata in pediatric patients.
Dermatol Res Pract 2014; 2014:848763.
highlighted by the low sensitivity of direct examination and
12 Wang HH, Lin YT. Bar code-like hair: dermoscopic marker of
the long turnaround time required for a PC. tinea capitis and tinea of the eyebrow. J Am Acad Dermatol 2015; 72
One limitation of our study relates to the representativeness (1 Suppl.):S41–2.
of the study population; all study participants had consulted a 13 Cremer G, Bousseloua N, Roudot-Thoraval F et al. [Tinea capitis in
dermatology department at a university hospital. Furthermore, Creteil. Trends over ten years]. Ann Dermatol Venereol 1998;
a second mycological examination was not performed if the 125:171–3.
initial examination was negative. The power of the statistical 14 Reichert-Penetrat S, Contet-Audonneau N, Barbaud A et al. Epi-
demiology of dermatophytoses in children living in northeast
analysis in the trichoscopic subgroups was limited by the
France: a 5-year study. Pediatr Dermatol 2002; 19:103–5.
small sample sizes. Nevertheless, it is important to note that 15 Elewski BE. Tinea capitis: a current perspective. J Am Acad Dermatol
the present series is the largest yet analysed. 2000; 42:1–20.
In conclusion, trichoscopy is a very useful tool for the diag- 16 Brasileiro A, Campos S, Cabete J et al. Trichoscopy as an additional
nosis of TC; the technique is rapid, painless and highly sensi- tool for the differential diagnosis of tinea capitis: a prospective
tive – even for dermoscopists inexperienced in trichoscopy. clinical study. Br J Dermatol 2016; 175:208–9.
Our results demonstrated that the presence of a single feature 17 Haliasos EC, Kerner M, Jaimes-Lopez N et al. Dermoscopy for the
pediatric dermatologist part I: dermoscopy of pediatric infectious
(comma hair, corkscrew hair, zigzag hair, Morse-code-like
and inflammatory skin lesions and hair disorders. Pediatr Dermatol
hair or whitish sheath) is predictive of TC. Morse-code-like 2013; 30:163–71.
hair appears to be highly specific for Microsporum TC. The use 18 Lacarrubba F, Verzi AE, Micali G. Newly described features result-
of trichoscopy greatly enhances the physician’s ability to make ing from high-magnification dermoscopy of tinea capitis. JAMA
treatment decisions. In the event of doubt, a mycological cul- Dermatol 2015; 151:308–10.
ture is still required. 19 Gomez-Moyano E, Crespo Erchiga V, Martinez Pilar L et al. Using
dermoscopy to detect tinea of vellus hair. Br J Dermatol 2016;
174:636–8.
References 20 Isa Isa R, Yanez Amaya B, Isa Pimentel M et al. Dermoscopy in
tinea capitis: a prospective study on 43 patients. Med Cutan Iber Lat
1 Vena GA, Chieco P, Posa F et al. Epidemiology of dermatophytoses:
Am 2014; 42:18–22.
retrospective analysis from 2005 to 2010 and comparison with
21 Ekiz O, Sen BB, Rifaioglu EN, Balta I. Trichoscopy in paediatric
previous data from 1975. New Microbiol 2012; 35:207–13.
patients with tinea capitis: a useful method to differentiate from
2 Kefalidou S, Odia S, Gruseck E et al. Wood’s light in Microsporum
alopecia areata. J Eur Acad Dermatol Venereol 2014; 28:1255–8.
canis positive patients. Mycoses 1997; 40:461–3.
22 Lu M, Ran Y, Dai Y et al. An ultrastructural study on corkscrew
3 Fuller LC, Barton RC, Mohd Mustapa MF et al. British Association
hairs and cigarette-ash-shaped hairs observed by dermoscopy of
of Dermatologists’ guidelines for the management of tinea capitis
tinea capitis. Scanning 2016; 38:128–32.
2014. Br J Dermatol 2014; 171:454–63.
23 Yu-Ting Lin Y-CL. The dermoscopic comma, zigzag, and bar
4 Slowinska M, Rudnicka L, Schwartz RA et al. Comma hairs: a der-
code-like hairs: markers of fungal infection of the hair follicles.
matoscopic marker for tinea capitis: a rapid diagnostic method. J
Dermatol Sinica 2014; 32:160–3.
Am Acad Dermatol 2008; 59(5 Suppl.):S77–9.