Jourtiat of Cutaneous Pattiotogy 1977: 4: 51-67

Traumatic Alopecia in Trichotillomania: a Pathogenic Interpretation of Histologic Lesions in the Pilosebaceous Unit^ .1. M. LACHAPELLE and G. E. PIERARD LInit and Laboratory tor Occupational Dermatoses, University of Louvain, and Department of Dermatology. University of Liege, Belgium

Ten patients with tratimatic alopeeia (trichotillomania) were being investigated histologieally and. in five of them, hairs trom the affected scalp area were plucked out tor direct microscopic examination. Some histologie features appear to be specific markers for traumatic alopeeia: empty hair duets, plucked out hair bulbs, elefts in hair matrix, eatagen involution of empty outer root sheaths. Mieseher's trichomalaeia in the deep dermis and torn-ol f .sebaeeous glands. Other signs are unspecilic, such as presence of catagen and anagen VI hairs, infundibular plugging, melanin in keratin pitigs and in the dermal papilla. The relative frequcney of the different histopathologic features was evaluated. When little clinical information is available, a diapno.sis of traumatic alopecia ean be supported by skin biopsy. The histolopie picture of trichotillomania is always incomplete, depending upon factors sueh as intensity of pulling or/and time of biopsy alter plucking. (Received for publication January 10, 1977}

The clinical diagnosis of trichotil'.omania of the scalp may sometimes be uncertain. In two well-substantiated clinicopathologic studies, Mehregan (1970) and Muller & Winkelman (1972) etnphasized the interest of skin biopsy in questionable cases. Indeed, they pointed out some specific histologic markers for trichotillomania and our pteliminary results have supported their conelusions (Lachapelle et al. 1975). The purpose of this study was to determine the sequence of events and the relative frequency of specific and non-specific histologic lesions encountered in trichotillomania in order to build up a pathogenic interpretation.

Material and Methods

Ten patients, presenting symptoms

of trichotillomania of the scalp were biops.ed

under local anesthesia. Special attention was paid to deeply removing the scalp tissue in order to include hair bulbs in the hypodertnis. Skin biopsies were fixed in Bouin's fluid, paraffin embedded and cut to 5 ^(m. Three hundred serial sections were obtained from each specimen and most of them were stained with hematoxylin and eosin. Special stainings included periodic acid-Schiff for basement membrane zone, Fontana's silver method for melanin. Perls' method for iron, Chevreniont's tnethod and DDD stain for -SH (sulphydryl) groups, and orcein stain for elastic fibres. Unstained sections were exaniined under polarized light, In five patients, hairs from the affected area of the scalp were plucked out for direct microscopic examination at low magnifica^.^^^ ^^^^

g^^^^ ^j ^j ,^57^ -^^ ^Uriyi\,i and

^^^^^^^ ^^^ ^ ^^.^ ,^y^^ ^^ ^^^^ palladium

•Presented in part at the 55tli annual tneeting of the British Association of Dermatologists, Cambridge (Laehapelle et al. 1975).



Table 1 Relative frequency of the different histopathologic signs in Traumatic Alopecia

Histopathologic signs Specific markers (SM) Empty hair canals Plucked out hair bulbs Clefts in hair matrix Catagen involution of empty outer root sheaths Mieseher's triehomalaeia in the deep dermis Torn off sebaceous glands SM/P Unspecific markers (UM) Nortnal anagen VI hairs Catagen hairs Infundibular plugging Melanin in keratin plugs Melanin in the papilla UM/P



Patients (P) 5 6 7


8 9 1 0

frequency %

100 20 50 70 40 30 3 4 2

6 2 3

2 3


100 70 +














were examined by scanning electron microscopy (Stereoscan, Cambridge).

100 70 80



surrounding connective tissue sheaths. Varying degrees of dissociation between cells of the hair matrix formed clefts (Fig. 4) in 50% of the biopsied lesions. Results In 70% of the specimens, when hair had The main histopathologic features and their been plucked out, an accumulation of basarelative frequency are summarized in Table 1. loid and isomorphic cells, probably derived The various lesions are described below from the oviter root sheath, formed a strand according to the site and type of alteration. evoluting as in a catagen phase (Fig. 5). The glassy membrane was greatly thickened and 1. Hair fotticlc: In all specimens, significant convoluted. lesions affected the different structures of Trichomalaeia, as described by Miescher the hair follicle. They were present at differ- (1942) was ob.setved in 40% of the cases in ent microanatomical sites but they did not our survey. The hair bulbs, structurally modiall occur in all patients, as indicated in Table fied and sometimes completely distorted, 1. were located in the deep part of the skin. A majority of hair ducts were empty The shafts were short, wavy and sotnetimes (Figs. 1 and 2). Some other hair follicles coiled up (Fig. 6). Unstained sections and were normal, in anagen VI, early anagen or those stained by Chevremont's method for catagen stage. Telogen hairs were rare. The -SH groups were examined under polarized occurrence of normally growing hairs light. The keratogenous zone of the hair between empty hair ducts in non-inflamma- shaft appeared progressively birefringent in tory dermis was a constant feature in our trichomalacic hairs as observed in normal survey (Table 1). Partially or completely hairs. plucked-out hair bulbs (Fig. 3) were obvious Release of melanin in the dermal papilla in only 20% of the cases. Exudative serosily and in the penbulbar connective ti.ssue filled the space between bulbar cells and the sheaths was ascertained by Fontana's silver





a E



Fig. 2. Hair in an early anagen stage in the vicinity of an empty hair duct (II & E stain).

method (Fig. 7) in 80% of the cases. Melanin granules were present in the papilla with different lesions; trichomalacic hairs, hairs with bulbar clefts, epithelial strand in a catagenlike involution. Iron pigment, searched for as

a hypothetical consequence of traumatic hemorrhage, could not be identified in any case. Someempty hair ducts were partially filled with keratin debris, usually clumped in plugs


Fw. .?. Hair Inilh half-way phicked out (H & F .•it.iin). in the infundibulum and occasionally in the isthmus. In 70% of the cases, melanin casts were also present in the empty hair ducts, which sometimes assumed a corkscrew shape in the more highly keratinized upper portion of the follicle (Fig. 8). In serial sections

melanin pigment was seen to migrate with the cells of the sheaths and to pour into the emply hair canals (Fig. 9). 2. Sebaceous glands and ducts: In 70% of the cases, the .structure of seba-




(^. 4. Clefts formed from a dissociation of cells within the hair matrix (11 & E stain).

ceous glands was unmodified. Empty spaces within sebaceous glands (Fig. 10) were observed in other specimens and corresponded to torn-off sebaceous cells. Indeed, parts of sebaceous glands were found lying within the horny layer of the epidermis (Fig. 11).

Sometimes, basal cells of partly empty sebaceous glands proliferated more than in normal skin, forming a double layer of flattened basal cells with occasional mitoses. Sebaceous ducts remained unaffected.



Fig. 5. Accumulation of hasaloid and i.somorphic cells (prohably derived from Ihe outer root sheath) forming a strand cvohiiing as in eatagen The glassy membrane is greatly thickened (H & E stain).



Fig. 6. Mieseher's trichomalaeia. The hair bulb, located in the hypodermis is completely distorded and the shaft is coiled up (H & E stain).

3. Epidermis: The epidermis was normal or slightly hyperkeratotic. In one case we observed typical eczematous changes of the epidermis, i.e. spongiosis, spongiotic vesiculation, exoserosis, and exocytosis of lymphoid cells.

A. Dermis: An almost complete absence of inflammatory infiltrate was a prominent feature in most sections (Figs. 1 and 2). A very scarce infiltrate of lymphoid cells was present in a few cases throughout the dermis. A dense peri-


r I 4




Fig. 7. Release of melanin pigment in (lie papilla and in Ihe surrounding connective tissue sheaths from an epithelial strand in a catagen-like involution (1 ontana's silver method).




Fig. 8. Melanin casts in two empty hair ducts assuming a corkscrew shape (Fontana's silver method).




Fig. 9. Melanin pigment migrating with the cells of the epithelial sheaths pouring into the hair duel (Fontana's silver method).







Fig. 12. Trichoptilosis (feathery appearance) observed by scanning electron microscopy.

vascular infiltrate of mononuclear cells was only observed in the specimens where epidermal eezematous changes were present. Alteration of the connective tissue architecture was never noted.

believed to be related to mechanical trauma and not to a metabolic process involving the hair bulb (Pierard 1975).

5. Sweat glands and ducts: There was no pathological change in the sweat glands nor in the sweat ducts of the different lesions.

The histopathologic pattern of trichotillomania is complex and different histologie features may be present in the same specimen, combining the presence of both damaged and normal hair follicles. As previously emphasized by Muller & Winkelmann (I 972), it is important to study many sections, ideally strictly serial sections, "in order to ensure finding hair follicle reactions that are clear and representative". Indeed, in some cases, the characteristic changes are only present in a few sections. Plucked-out hair bulbs, presence of clefts in hair bulbs, and toni-off sebaceous glands, have been observed by examining hundreds of strictly serial sections. Some of the histologie lesions described may be considered as specific markers for trichotillomania, whereas others are non-

6. Direct examination of plucked-out hairs in trichograms: Direct examination of plucked-out hairs from the affected area was performed in five cases, with the technique used for trichrograms derived from Van Scott et al. (1957). Telogen hairs were reduced in number ( 2 5%), while catagen (1 -3%) and anagen hairs (84-90%) were in the normal range and dystrophic hairs were present in an increased proportion (7 - 1 2%). Some alterations of the hair shafts, like loss of cuticular cells and trichoptilosis, were present (Fig. 12) and




specific. In Table 1, six histologie features MuUer & Winkelmann (1972) considered aie classified as specific, and five as non- that the presence of catagen hairs was the specific. most significant finding in trichotillomania. Follicular plugging is a trivial histopatho- The presence of catagen hairs, and the logic symptotn in many skin diseases of the frequent catagen involution of empty outer scalp and therefore cannot be taken into root sheaths were encountered in our survey. consideration as an argument for hair pulling It seems likely that hairs that have been and/or plucking. Nevertheless, it must be pulled out can be precipitated into a catagen added that follicular plugging is particularly phase. It is also probable that, when hairs prominent in tricholillomania, dilating the have been entirely plucked-out, remnants of infundibulum and occasionally extending hair sheaths undergo a basaloid and isomordown to the isthmus. The presence of phic accumulation which may evoke the melanin casts in keratin plugs filling the folli- morphological appearance of a catagen stage. cular infundibulum has been mentioned in Such a secondary proliferative reaction also previous reports (Mehregan 1970, Muller & occurs al the level of the basal cells of the Winkelmann 1972, Lachapelle et al. 1975). sebaceous glands. It has already been menIt is evident from our serial sections that tioned by Strauss & Kligman (1958). these melanin deposits in the empty hair Trichomalacia has been considered as a canals are produced and released by pro- specitic histologie lesion for trichotillomania liferating outer root sheaths passively and (Miescher 1942, Miescher & Schmuziger migrate upwards with the cells. 1957, Haensch & Blaich I960). We agree Similarly, release of melanin in the dermal that trichomalacia encountered in trichotillopapilla from the matrix cannot be considered mania, i.e. a complete distortion of a fully as specific for trichotillomania, as it usually developed terminal hair in an anagen stage is occurs in the hairs blocked in anagen IV specific for pulling hairs. It probably results stage during alopecia areata. Evidence for on a new hair cycle initiated in the deeper traumatized connective tissue sheaths and parts of the skin, from the remnants of the vessels was absent in our survey but has been matrix of a plucked out anagen VI hair. pointed out by Mehregan (1970). None of these hair follicles undergoes a cataSome other features need to be considered gen phase as pointed out by Mehiegan (1970) as specific for hair pulling and/or plucking. and contrary to Miesclier's observation, we Indeed, they are never encountered in other did not find a loss of birefringence in these skin diseases of the scalp and it is not con- tortuous shafts. The newly-formed .soft epiceivable that they can be spontaneous in thelial strand grows in the partially collapsed origin. The presence of empty hair canals in and tortuous hair duct and even if keratinizaserial sections is an obvious histopathological tion occurs, tiair is kept convoluted like a trifeature of hair plucking, as mentioned pre- chomalacic unkeratinized hair. It must be viously by Mehregan (1970) and by Muller emphasized that some other types of tricho& Winkelmann (1972). The absence or the malacia exist, as tor example in some pilary scarcity of an inflammatory response and agencsias and hypotrichosis (personal unconversely the presence of early and late published data) where small hairs are wavy, anagen stages in the affected area are two and eventually coiled up. Pinkus (1965) also features indicating the absence of an under- described a peculiar type of trichomalacia lying disease. The initiation, pursute and during a transient alopecia in B D mice. maintenance of new hair cycles are not im- which is structurally close to that obseived paired and the standard scheme of the hair in pilary agenesia of man. cycle (Morctti et al. 1976) controlled by A pathogenic interpretation of these t h e connective tissue function during the events should be related to the scheme of anagen IV stage (Pierard & de la Brassinne the hair cycle modifications depending on the more or less extended avulsion of the 1974, 1975a, 1975b) is unmodified.



hair shaft. Incomplete pulling may result in the development of clefts in the hair bulb. If the structure and function of the matrix are sufficiently altered, the standard response of any anagen VI hair follicle to insult leads to a catagen involution. A release of melanin from the epithelium into the papilla concomitantly occurs (Ebling & Rook 1972). If the trauma is less severe, clefts separating the hair over the matrix could be followed by the formation of a dystrophic and atrophic hair issued from the hair bulb lying in the deep part of the skin, at the level reached in a normal anagen VI stage. Complete plucking of hair may be accompanied by torn-off clumps of sebaceous glands. This is followed by an isomorphic proliferative reaction of the outer root sheath and of the basal layer of the sebaceous gland. The deeper part of the epithelial strand evolves as in a normal catagen phase while the ostial and infundibular cells form a tortuous and extensive keratinized plug. New anagen cycles can then be initiated. In conclusion, gross observation of plucked hairs is of limited value in supporting a diagnosis of trichotillomania. A decrease in the proportion of telogen hairs with a concomitant increase in dystrophic anagen VI hairs may also be encountered in other conditions like intoxications by heavy metals (Pierard et al. 1977). When little clinical information is available, diagnosis of traumatic alopecia should be supported by a skin biopsy. The histologie pattern of trichotillomania is incomplete, depending upon factors such as intensity of pulling and/or time of biopsy after plueking. Acknowledgement

We are greatly indebted to Miss F. D. M. de Cook for her skillful laboratory work and to Mrs N. Noelen and Miss I. Strivay for secretarial help. References

Ebling, F. J. & Rook, A. (1972) Hair. In Textbook

of Dermatology, ed. Rook, A., Wilkinson, D s & Ebling, F. .1. G., pp. 1559-1641. OxfordBlackwell. Haensch, R. & Blaich, W. (1960) Trichomalacie und Trichotillomanie. Archiv fiir klinisehe und experimentelle Dermatologie 210, 447-452. Lachapelle, J. M., Picrard, G. E., & Fierens, F (1975) Trichotillomania: a pathogenic interpretation of histologie changes in the pilosebaceous unit. British Journal of Dertnatology 9 j supplement 11, 21. Mehregan, A. II. (1970) Trichotillomania. A clinicopathologic study. Arehives of Dermatology 1O2 129-133. Miescher, G. (1942) Trichomalacie. Arehiv fa^ Dermatologie und Syphiligrafie (Berlin) fgjj 117. Miescher, G. & Schmuziger, P. (1957) Irichomalacic und Trichotillomanie. Dertnatologica 114, 199-203. Morctti, G., Rampini, E. & Rcbora, A. (1976) The hair cycle rc-cvaluatod. Ititertiational Journal of Dermatology 15, 277-285. Muller, S. A. & Winkelmann, R. K. (1972) Tricliotillomania. A clinicopathologic study of 24 caaes. Archives of Dermatology 10.5. 535-540 Pierard, G. E. (1975) Dystrophies acquiscs du cheveu. Structure et interpretation pathogiiniquc. Archives Franeaises de Dermatologie et de Syphiligraphie 102, \ 27-143. Pierard, G. E. & de la Brassinne, M. (1974) Connective tissue cells and hair growth in rat skin plucked human scalp and alopecia areata. Proeeedings of the fourth meeting of the European Society for Derntatologieal Research, Atitsterdam. Abstract 20. Liege: Presses Universitaires Pierard, G. E. & de la Brassinne, M. (1975a) Modulation of dermal cell activity during hair growth in the rat. Journal of Cutaneous Pathology 2 35-41. Picrard, G. E. & dc la Brassinne, M. (1975b) Cellular activity in the dermis surrounding the hair bulb in alopecia areata. .Journal of Cutaneous Pathology 2, 240-245. Picrard, G. E., Hermanns, J. F., Lctot. B., Cryns, IC & Lapicre, Ch. M. (1977) Cellular dynamics iii toxic alopecia. Proceedings of the XV International Congress of Dertnatology, Mexico. Pinkus, H. (1965) Transient alopecia in weanling BD mice (Trichomalacia). In Biology of the skin and hair growth, ed. Lyne, A. G.& Shorl,B. F. pp. 141-153, Sydney, Angus and Robertson. Strauss, J. S. & Kligman, A. M. (1958) Pathologic patterns of the sebaceous gland. Journal of Investigative Dermatology 30, 51-61. Van Scott, E. J., Reinertson, R. P. & Stainmuller, R. (1957) The growing hair roots of the human scalp hair and morphologic changes therein following amethopterin. Journal of Investigative Dermatology 29, 197-204.

HISTOLOGIC LESIONS IN TRAUMATIC ALOPECIA Address: Dt-. J. M. Lachapelle Utni(> de Dermatologie Professiotjnelle Departetnetit de Medeeinc du Travail-UCL 3033 'Jnivcrsiti Catholique de Louvain Clos Chapelle-aux-Champs, 30 B-1200 Brussels

Dr. G. K. Pierard Service dc Dermatologie Hopital de Baviere Vttiversitc de Liege B-4020 Liege Belgium


Traumatic alopecia in trichotillomania: a pathogenic interpretation of histologic lesions in the pilosebaceous unit.

Jourtiat of Cutaneous Pattiotogy 1977: 4: 51-67 Traumatic Alopecia in Trichotillomania: a Pathogenic Interpretation of Histologic Lesions in the Pilo...
8MB Sizes 0 Downloads 0 Views