British Journal of Dermatology (1976J 94, 319.
Lichen scrofulosorum A REPORT OF FOUR CASES N.P.SMITH, T.J.RYAN,* K.V.SANDERSON AND I.SARKANYt Skin Department, St George's Hospital, Hyde Park Comer, London SWIX 7EZ, •The Slade Hospital, Headington, Oxford OX3 7JH, and jThe Royal Free Hospital, Pond Street, London NW3 2QG Accepted for publication 2 July
SUMMARY
Four cases of lichen scrofulosorum associated with lymph node tuberculosis are reported, and the clinical and histologica! appearances of the disorder are described.
Lichen scrofulosorum was first recognized by Hebra and his description is a model of precision and clarity. In a footnote to the English translation (1868) the editors commented that'. . . English writers seem generally unable to identify the lichen scrofulosorum with any disease observed by themselves . . .'. To later generations of dermatologists it became a readily recognized but uncommon entity which was associated with tuberculosis of lymph nodes or bone, although the infection was not always obvious at the time of development of the rash (Miescher, 1955). The eruption charaaeristically occurs in children and yoting adults. All Hebra's cases were males under 25 years old. Volk (1931) suggested that women may be more commonly affected in the older age group. Other forms of cutaneous tuberculosis might at times be present in the same patient. Phlyctenular conjunctivitis was present in 50% ofpatients in one series (Rauschkolb, 1934). Lichen-scrofulosorum-like lesions have also been seen following B.C.G. vaccination (Dostrovsky & Sagher, 1963). There have been few case reports imder this title in the last two decades. The situation has been somewhat confused by the description of'lichenoid tuberculid', an apparently heterogeneous disorder (Ockuly & Montgomery, 1950). Some of the original cases described by Ockuly & Montgomery may have been cutaneous sarcoidosis. A rather extreme position is taken by Lever (1967) who concluded from the clinical and histological data that lichen scrofulosorum is identical with the lichenoid type of sarcoidosis. We are reporting the following four cases to rectify this erroneous suggestion and to show how well the condition responds to antituberculous treatment.
CASE REPORTS
Case I
A 27-year-old nurse, born in Vietnam but living for 5 years in Britain, was seen with a symptomless rash on the trunk which had been gradually spreading for i year. Her past health had been good apart 319
320
N.P.Smith et al.
from an episode of headache and vomiting 4 months previously, when her submandibular and cervical lymph nodes were swollen. Some swelling of the cervical nodes had probably been present for 6 years. The patient's maternal grandmother had been treated for pulmonary tuberculosis 4 years previously. Examination showed an eruption consisting of follicular skin-coloured papules, 1-2 mm in diameter, on the lower chest and abdomen with a tendency to grouping. The left submandibular and upper cervical lymph nodes were enlarged. Investigations. Chest X-ray showed an old calcified tuberculous focus in the right mid-zone, but no evidence of an active lesion. ESR was 40 mm/h (Westergren). The Mantoux reaction (i: io,cxx>) was strongly positive, producing induration and swelling of the whole forearm within 48 h. Twelve days after the Mantoux test a crop of lichenoid papules identical to the original eruption developed over the test-site area. Biopsy of one of the enlarged cervical nodes showed tuberculous histology; fluorescent microscopy and culture were positive for Myobacterium tuberculosis. Skin biopsy showed some parakeratosis in the region of the hair follicles, with follicular plugging. There was an upper dermal granulomatous infiltrate of epithelioid cells closely surrounding the follicular ducts. Some Langhans giant cells were present in the infiltrate and there were lymphocytes at the periphery. No acid-fast bacilli were seen in the sections examined and none could be cultured from the skin. The patient was treated with isoniazid 300 mg/day, ethambutol 1000 mg/day and pyridoxine 10 mg/day. After 7 weeks the dose of ethambutol was reduced to 600 mg/day. After 4 months there was marked reduction in the size of the cervical glands and the skin eruption had cleared completely.
FIGURE I. Low power view ot lesion trom case 2, showing a granulomatous infiltrate of epithelioid cells and lymphocytes surrounding a hair follicle (H & E, x 83).
Lichen scrofulosorum
321
mmm FIGURE 2. Case 2, high power view, showing a perifollicular non-caseating tuberculoid granuloma ( H & E , X2o8).
Case 2
An iS-year-old Tanzanian male student had had a rash affecting the trunk and upper arms for 2 months. For 6 weeks there had been a rapidly progressive enlargement of the lymph nodes on both sides of his neck, and he had had a low grade fever. There was no relevant past medical history and no family history of tuberculosis. A buff-coloured lichenoid rash was patchily distributed over the trunk and upper limbs. There was some follicular hyperkeratosis and the individual lesions were up to 3 mm in diameter. Bilateral cervical lymphadenopathy was present. The nodes were soft, fluctuant and fixed to deeper tissues. Investigations. Chest X-ray showed enlargement of the left hilar and right paratracheal nodes. ESR was 46 mm/h. Mantoux reaction ( i : 10,000) was strongly positive with 12 mm of indtn-ation and blister formation. Biopsy of cervical and nasopharyngeal lymph nodes showed typical tuberculous histology. Tubercle bacilli were identified both on direct smear and culture of the pus from a cervical node. Histology of the skin lesions (Figs i & 2) showed non-caseating tuberculoid granulomata both in the upper dermis and surrounding hair follicles. No acid-fast bacilli were seen. The patient was treated with streptomycin 0 75 g/day, sodium aminosalicylate 12 g/day and isoniazid 300 mg/day. After 7 weeks the skin eruption was much less obvious, chest X-ray revealed reduction in size of the hilar lymph nodes and the ESR had fallen to 10 mm/h. After 11 weeks the skin was
322
N.P.Smith et al.
clear and there was no cervical lymphadenopathy although on chest X-ray there was still some enlargement of the left hilar nodes. Case 3
A 66-year-old English woman developed a rash 6 years after receiving treatment for lymph node tuberculosis. When she was first seen she had had symptoms of dysphagia, a persistent cough and night sweats. Enlarged supraclavicular lymph nodes were found on examination. Biopsy of the nodes showed the presence of tuberculoid granulomata, and Mycobacterium tuberculosis was cultured from the biopsy material. M. tuberculosis was also cultured from the sputum on one occasion but the chest X-ray then showed only some paralysis of the left diaphragm with no evidence of active pulmonary tuberculosis. The patient was treated with calcium benzamidosalicylate and isoniazid for i year, and streptomycin which was stopped after 5 weeks because of tinnitus. The lymphadenopathy disappeared, and the patient was free of symptoms for 6 years. She then developed discomfort in the right eye and was treated for an anterior uveitis. A few months later she developed a symptomless rash over the back, chest and abdomen. Though there was some spontaneous regression of the rash during the following 12 months, after about 18 months the eruption recurred in a more severe form.
FIGURE 3. Case 3, distribution of the lesions over the back.
Examination showed a pinkish-brown grouped micropapular rash over the trunk and back (Fig. 3), individual lesions being 2-4 mm in diameter. There was no evidence of lymphadenopathy and examination of the eyes showed the appearances of a quiescent right uveitis with posterior synechiae formation. Investigations. Chest X-ray showed slight residual weakness of the left diaphragm. ESR was 26 mm/h. Mantoux reaction (i; 10,000) was positive, producing 12 mm of induration. A Kveim test was
Lichen scrofulosorum
323
negative. Two skin biopsies were taken during the course of the illness, both showing essentially similar features. The epidermis was slightly thinned and there was some oedema of the papillary dermis. Non-caseating epithelioid cell granulomata were present in the upper dermis adjacent to hair foUicles in some areas. A few giant cells were seen and there was a patchy peripheral lymphocytic infiltrate. The patient was treated with rifampicin 450 mg/day and isoniazid 300 mg/day. After 3 months of treatment the skin eruption had cleared completely. Case 4
A 53-year-old Englishman was seen in the Skin Department of the Royal Free Hospital in 1969 with a 6-week history of a widespread eruption on the trunk and limbs. Nine years previously he had been discovered to be suffering from tuberculosis of a cervical node on the left side of the neck and was treated with streptomycin and sodium aminosalicylate/isoniazid (Pasinah, Wander) for 2 years. At that time, the diagnosis of tuberculosis was confirmed by the finding of acid-fast bacilli in the pus, and guinea-pig inoculation was also positive. In 1964 the lump in the neck reappeared and a further i year's treatment with sodium aminosalicylate was given. He was well until 1968 when the lymphadenopathy recurred and he was treated with streptomycin for 3 months and Pasinah for 7 months. In January 1969, while on antituberculous treatment, the lesion in his neck became swollen and tender and surgical drainage of a collar-stud abscess was carried out. In May 1969 a diffuse lichenoid eruption appeared on his trunk and limbs. He was off all antituberculous therapy at that time. The patient's mother had a history of diabetes and tuberculosis and a brother aged 37 had died of tuberculosis. On examination, he was found to have many discrete, in some parts almost confluent, lichenoid papules (Fig. 4) resembling those of lichen planus, but less shiny and less violaceous in colour, over the chest, abdomen, back, buttocks and limbs. There was a string of swellings in the left posterior triangle of the neck and he had a left hydrocele.
FIGURE 4. Case 4, detail of the lesions on the back.
Investigations. Chest X-ray showed pleural thickening and bilateral apical scarring. There was also a density in the left upper zone which was thought to be due to old tuberculosis. A Mantoux test (1:100,000) was positive. Histology of the skin from the left scapular region showed several noncaseating granulomata in the dermis. No acid-fast bacilli were seen in the section, and none were cultured from the cervical lymph node.
324
N.P.Smith et al.
A diagnosis of lichen scrofulosorum was made and in view of continued discharge from his neck, a course of sodium aminosalicylate and isoniazid (Inapasade, Smith & Nephew Pharmaceuticals) was started. Within 2 weeks the rash was much less obvious, and i month later it had disappeared altogether.
DISCUSSION
The four cases described correspond well to Hebra's original description of lichen scrofulosorum. All the patients had lymph node tuberculosis, the skin lesions are comparable in clinical appearance and distribution and in all cases the eruptions appear to have improved or resolved completely within months of commencing antituberculous therapy. The cutaneous lesions of lichen scrofulosorum are sufficiently characteristic for a diagnosis to be suspected on clinical grounds alone. Typically the eruption consists of symptomless papules, usually follicular in distribution and 0-5-3 "^ni ^" diameter. Although usually skin-coloured, they may vary from yellow to reddish-brown. The papules occur in groups sometimes with a patchy, annular arrangement and are most characteristically seen on the abdomen, chest and back. There may be some scaling over the lesions and micropustule formation has been described. Skin lesions somewhat resembling those of lichen scrofulosorum may occasionally be seen in other conditions. The papules of lichen nitidus tend to be shiny and commonly occur on the limbs and genital area as well as the trunk. Keratosis spinulosa is a condition which may be indistinguishable from lichen scrofulosorum clinically although the cutaneous lesions tend to be less inflammatory with more obvious follicular hyperkeratosis and rarely show the histological appearances of lichen scrofulosorum. Limb and face involvement is common in the papular and lichenoid types of cutaneous sarcoidosis and the Mantoux reaction should help to distinguish the two conditions. In lichen scrofulosorum the Mantotix reaction is always positive and usually strongly so. Fimgal infections of the skin may produce patchy annular eruptions but small follicular papules are seldom seen. Secondary syphilis and lichenoid drug eruptions may occasionally produce a picture superficially similar to that of lichen scrofulosorum. The typical histological appearances of lichen scrofulosorum (Montgomery, 1937) consist of superficial dermal granulomata, usually surrounding hair follicles and sweat ducts, but occasionally independent of either. The infiltrate is composed of epithelioid cells with some Langhans giant cell formation and a narrow margin of lymphocytes at the periphery. The adjacent capillaries may show a lymphocytic perivascular infiltrate and there may be non-specific secondary changes in the epidermis above the lesions (Johnson, 1972). There is no caseation necrosis and changes are not seen in the subcutaneous tissue. Normally tubercle bacilli are not seen in the pathological specimens, nor can they be cultured from biopsy material. Some of the earlier writers, however, have claimed to identify bacilli in the skin lesions (Volk, i93O- The distribution of the granulomata around the appendages and the fact that the infiltrate often occupies several dermal papillae are helpful in differentiating lichen scrofulosorum from lichen nitidus. The cutaneous lesions of secondary syphilis are essentially perivascular and plasma cells are frequently seen in the infiltrate. In cutaneous sarcoidosis the granulomata tend to be more discrete with less surrounding lymphocytic infiltrate, but distinction of sarcoidosis from lichen scrofulosorum may be impossible on histological grounds alone. The course of the untreated disease tends to be slow, the lesions persisting for months or years. Spontaneous involution without scarring usually occurs eventually. In all our cases there has been an excellent response to antituberculous treatment, although not all writers have had the same experience (Schuhmachers, 1967). There is little doubt that tuberculous infection is an aetiological factor in the development of the cutaneous lesions of lichen scrofulosorum. The exact nature of the relationship between systemic
Lichen scrofulosorum
325
tuberculosis and the skin eruption is not completely clear. At one time the cutaneous lesions were considered to be reactions to 'toxins' of tubercle bacilli. It is now thought more likely that they result from haematogenous dissemination of bacilli in an individual strongly sensitive to Mycobacterium tuberculosis, though the primary focus of dissemination may not always be apparent. Lichen scrofulosorum appears to be a definite clinical and pathological entity, distinct from other forms of cutaneous tuberculosis and unrelated to sarcoidosis. Its prevalence in any community probably depends upon the prevalence of tuberculosis of the lymph nodes and it is felt that the condition is perhaps not so uncommon as the lack of recently doctunented cases would suggest. ACKNOWLEDGMENTS
We would like to thank Dr T.J.McEIwain of the Royal Marsden Hospital and Professor J.G.Scadding of the Brompton Hospital for permission to publish details of case 2. Case I was presented by Dr T.J.Delaney at a meeting of the St John's Hospital Dermatological Society on 6 June 1974. Case 3 was presented by Dr J.P.Ellis at a meeting of the Dermatological Section of the Royal Society of Medicine on 21 February 1974, and was subsequently listed in the Proceedings of the Royal Society of Medicine. REFERENCES DOSTROVSKY, A. & SAGHER, F . (t963) Dcrmatological complications of B.C.G. vaccination. British Journal of Dermatology, 75, 181. HEBRA, F . (t868) Lichen scrofulosorum. In: Diseases of the Skin (Translated and ed. by C.H.Fagge and P.H.PyeSmith), Vol. 2, p. 52. The New Sydenham Society, London. JOHNSON, W.C. (1972) Lichen scrofulosorum. In: Dermal Pathology (Ed. by J.H.Grahatn, W.C.Johnson and E.B.Helwig), p. 394. Harper Sc Row, Hagerstown, Maryland. LEVER, W.F. (t967) In: Histopathology of the Skin, 4th. edn, p. 301. Pitman Medical, London. MIESCHER, G . (1955) Dber katamnestischc Untersuchungen bei Fallen mit Tuberkulid. Dermatologica, n o , 23. MONTGOMERY, H . (1937) Histopathology of various types of cutaneous tuberculosis. Archives of Dermatology and Syphilology, 35, 698. OCKULY, O.E. & MONTGOMERY, H . (1950) Lichenoid tuberculid. Journal of Investigative Dermatology, 14, 415. RAUSCHKOLB, J.E. (1934) Tuberculosis of the skin: a clinical report. Archives of Dermatology and Syphilology, 29, 398. SCHUHMACHERS, R. (1967) 2 Fiille eines lichenoides Tuberkulids (= Lichen scrofulosorum). Hautarst, 18, 81. VOLK, R. (i93t) Lichen scrofulosorum. In: Handbuch der Haia-und Geschlechtskrankheiten (Ed. by J. Jadassohn) Vol. 10, p. 331, Springer, Berlin.
Lichen scrofulosorum A REPORT OF FOUR CASES N.P.SMITH, T.J.RYAN,* K.V.SANDERSON AND I.SARKANYt Skin Department, St George's Hospital, Hyde Park Comer, London SWIX 7EZ, •The Slade Hospital, Headington, Oxford OX3 7JH, and jThe Royal Free Hospital, Pond Street, London NW3 2QG Accepted for publication 2 July
SUMMARY
Four cases of lichen scrofulosorum associated with lymph node tuberculosis are reported, and the clinical and histologica! appearances of the disorder are described.
Lichen scrofulosorum was first recognized by Hebra and his description is a model of precision and clarity. In a footnote to the English translation (1868) the editors commented that'. . . English writers seem generally unable to identify the lichen scrofulosorum with any disease observed by themselves . . .'. To later generations of dermatologists it became a readily recognized but uncommon entity which was associated with tuberculosis of lymph nodes or bone, although the infection was not always obvious at the time of development of the rash (Miescher, 1955). The eruption charaaeristically occurs in children and yoting adults. All Hebra's cases were males under 25 years old. Volk (1931) suggested that women may be more commonly affected in the older age group. Other forms of cutaneous tuberculosis might at times be present in the same patient. Phlyctenular conjunctivitis was present in 50% ofpatients in one series (Rauschkolb, 1934). Lichen-scrofulosorum-like lesions have also been seen following B.C.G. vaccination (Dostrovsky & Sagher, 1963). There have been few case reports imder this title in the last two decades. The situation has been somewhat confused by the description of'lichenoid tuberculid', an apparently heterogeneous disorder (Ockuly & Montgomery, 1950). Some of the original cases described by Ockuly & Montgomery may have been cutaneous sarcoidosis. A rather extreme position is taken by Lever (1967) who concluded from the clinical and histological data that lichen scrofulosorum is identical with the lichenoid type of sarcoidosis. We are reporting the following four cases to rectify this erroneous suggestion and to show how well the condition responds to antituberculous treatment.
CASE REPORTS
Case I
A 27-year-old nurse, born in Vietnam but living for 5 years in Britain, was seen with a symptomless rash on the trunk which had been gradually spreading for i year. Her past health had been good apart 319
320
N.P.Smith et al.
from an episode of headache and vomiting 4 months previously, when her submandibular and cervical lymph nodes were swollen. Some swelling of the cervical nodes had probably been present for 6 years. The patient's maternal grandmother had been treated for pulmonary tuberculosis 4 years previously. Examination showed an eruption consisting of follicular skin-coloured papules, 1-2 mm in diameter, on the lower chest and abdomen with a tendency to grouping. The left submandibular and upper cervical lymph nodes were enlarged. Investigations. Chest X-ray showed an old calcified tuberculous focus in the right mid-zone, but no evidence of an active lesion. ESR was 40 mm/h (Westergren). The Mantoux reaction (i: io,cxx>) was strongly positive, producing induration and swelling of the whole forearm within 48 h. Twelve days after the Mantoux test a crop of lichenoid papules identical to the original eruption developed over the test-site area. Biopsy of one of the enlarged cervical nodes showed tuberculous histology; fluorescent microscopy and culture were positive for Myobacterium tuberculosis. Skin biopsy showed some parakeratosis in the region of the hair follicles, with follicular plugging. There was an upper dermal granulomatous infiltrate of epithelioid cells closely surrounding the follicular ducts. Some Langhans giant cells were present in the infiltrate and there were lymphocytes at the periphery. No acid-fast bacilli were seen in the sections examined and none could be cultured from the skin. The patient was treated with isoniazid 300 mg/day, ethambutol 1000 mg/day and pyridoxine 10 mg/day. After 7 weeks the dose of ethambutol was reduced to 600 mg/day. After 4 months there was marked reduction in the size of the cervical glands and the skin eruption had cleared completely.
FIGURE I. Low power view ot lesion trom case 2, showing a granulomatous infiltrate of epithelioid cells and lymphocytes surrounding a hair follicle (H & E, x 83).
Lichen scrofulosorum
321
mmm FIGURE 2. Case 2, high power view, showing a perifollicular non-caseating tuberculoid granuloma ( H & E , X2o8).
Case 2
An iS-year-old Tanzanian male student had had a rash affecting the trunk and upper arms for 2 months. For 6 weeks there had been a rapidly progressive enlargement of the lymph nodes on both sides of his neck, and he had had a low grade fever. There was no relevant past medical history and no family history of tuberculosis. A buff-coloured lichenoid rash was patchily distributed over the trunk and upper limbs. There was some follicular hyperkeratosis and the individual lesions were up to 3 mm in diameter. Bilateral cervical lymphadenopathy was present. The nodes were soft, fluctuant and fixed to deeper tissues. Investigations. Chest X-ray showed enlargement of the left hilar and right paratracheal nodes. ESR was 46 mm/h. Mantoux reaction ( i : 10,000) was strongly positive with 12 mm of indtn-ation and blister formation. Biopsy of cervical and nasopharyngeal lymph nodes showed typical tuberculous histology. Tubercle bacilli were identified both on direct smear and culture of the pus from a cervical node. Histology of the skin lesions (Figs i & 2) showed non-caseating tuberculoid granulomata both in the upper dermis and surrounding hair follicles. No acid-fast bacilli were seen. The patient was treated with streptomycin 0 75 g/day, sodium aminosalicylate 12 g/day and isoniazid 300 mg/day. After 7 weeks the skin eruption was much less obvious, chest X-ray revealed reduction in size of the hilar lymph nodes and the ESR had fallen to 10 mm/h. After 11 weeks the skin was
322
N.P.Smith et al.
clear and there was no cervical lymphadenopathy although on chest X-ray there was still some enlargement of the left hilar nodes. Case 3
A 66-year-old English woman developed a rash 6 years after receiving treatment for lymph node tuberculosis. When she was first seen she had had symptoms of dysphagia, a persistent cough and night sweats. Enlarged supraclavicular lymph nodes were found on examination. Biopsy of the nodes showed the presence of tuberculoid granulomata, and Mycobacterium tuberculosis was cultured from the biopsy material. M. tuberculosis was also cultured from the sputum on one occasion but the chest X-ray then showed only some paralysis of the left diaphragm with no evidence of active pulmonary tuberculosis. The patient was treated with calcium benzamidosalicylate and isoniazid for i year, and streptomycin which was stopped after 5 weeks because of tinnitus. The lymphadenopathy disappeared, and the patient was free of symptoms for 6 years. She then developed discomfort in the right eye and was treated for an anterior uveitis. A few months later she developed a symptomless rash over the back, chest and abdomen. Though there was some spontaneous regression of the rash during the following 12 months, after about 18 months the eruption recurred in a more severe form.
FIGURE 3. Case 3, distribution of the lesions over the back.
Examination showed a pinkish-brown grouped micropapular rash over the trunk and back (Fig. 3), individual lesions being 2-4 mm in diameter. There was no evidence of lymphadenopathy and examination of the eyes showed the appearances of a quiescent right uveitis with posterior synechiae formation. Investigations. Chest X-ray showed slight residual weakness of the left diaphragm. ESR was 26 mm/h. Mantoux reaction (i; 10,000) was positive, producing 12 mm of induration. A Kveim test was
Lichen scrofulosorum
323
negative. Two skin biopsies were taken during the course of the illness, both showing essentially similar features. The epidermis was slightly thinned and there was some oedema of the papillary dermis. Non-caseating epithelioid cell granulomata were present in the upper dermis adjacent to hair foUicles in some areas. A few giant cells were seen and there was a patchy peripheral lymphocytic infiltrate. The patient was treated with rifampicin 450 mg/day and isoniazid 300 mg/day. After 3 months of treatment the skin eruption had cleared completely. Case 4
A 53-year-old Englishman was seen in the Skin Department of the Royal Free Hospital in 1969 with a 6-week history of a widespread eruption on the trunk and limbs. Nine years previously he had been discovered to be suffering from tuberculosis of a cervical node on the left side of the neck and was treated with streptomycin and sodium aminosalicylate/isoniazid (Pasinah, Wander) for 2 years. At that time, the diagnosis of tuberculosis was confirmed by the finding of acid-fast bacilli in the pus, and guinea-pig inoculation was also positive. In 1964 the lump in the neck reappeared and a further i year's treatment with sodium aminosalicylate was given. He was well until 1968 when the lymphadenopathy recurred and he was treated with streptomycin for 3 months and Pasinah for 7 months. In January 1969, while on antituberculous treatment, the lesion in his neck became swollen and tender and surgical drainage of a collar-stud abscess was carried out. In May 1969 a diffuse lichenoid eruption appeared on his trunk and limbs. He was off all antituberculous therapy at that time. The patient's mother had a history of diabetes and tuberculosis and a brother aged 37 had died of tuberculosis. On examination, he was found to have many discrete, in some parts almost confluent, lichenoid papules (Fig. 4) resembling those of lichen planus, but less shiny and less violaceous in colour, over the chest, abdomen, back, buttocks and limbs. There was a string of swellings in the left posterior triangle of the neck and he had a left hydrocele.
FIGURE 4. Case 4, detail of the lesions on the back.
Investigations. Chest X-ray showed pleural thickening and bilateral apical scarring. There was also a density in the left upper zone which was thought to be due to old tuberculosis. A Mantoux test (1:100,000) was positive. Histology of the skin from the left scapular region showed several noncaseating granulomata in the dermis. No acid-fast bacilli were seen in the section, and none were cultured from the cervical lymph node.
324
N.P.Smith et al.
A diagnosis of lichen scrofulosorum was made and in view of continued discharge from his neck, a course of sodium aminosalicylate and isoniazid (Inapasade, Smith & Nephew Pharmaceuticals) was started. Within 2 weeks the rash was much less obvious, and i month later it had disappeared altogether.
DISCUSSION
The four cases described correspond well to Hebra's original description of lichen scrofulosorum. All the patients had lymph node tuberculosis, the skin lesions are comparable in clinical appearance and distribution and in all cases the eruptions appear to have improved or resolved completely within months of commencing antituberculous therapy. The cutaneous lesions of lichen scrofulosorum are sufficiently characteristic for a diagnosis to be suspected on clinical grounds alone. Typically the eruption consists of symptomless papules, usually follicular in distribution and 0-5-3 "^ni ^" diameter. Although usually skin-coloured, they may vary from yellow to reddish-brown. The papules occur in groups sometimes with a patchy, annular arrangement and are most characteristically seen on the abdomen, chest and back. There may be some scaling over the lesions and micropustule formation has been described. Skin lesions somewhat resembling those of lichen scrofulosorum may occasionally be seen in other conditions. The papules of lichen nitidus tend to be shiny and commonly occur on the limbs and genital area as well as the trunk. Keratosis spinulosa is a condition which may be indistinguishable from lichen scrofulosorum clinically although the cutaneous lesions tend to be less inflammatory with more obvious follicular hyperkeratosis and rarely show the histological appearances of lichen scrofulosorum. Limb and face involvement is common in the papular and lichenoid types of cutaneous sarcoidosis and the Mantoux reaction should help to distinguish the two conditions. In lichen scrofulosorum the Mantotix reaction is always positive and usually strongly so. Fimgal infections of the skin may produce patchy annular eruptions but small follicular papules are seldom seen. Secondary syphilis and lichenoid drug eruptions may occasionally produce a picture superficially similar to that of lichen scrofulosorum. The typical histological appearances of lichen scrofulosorum (Montgomery, 1937) consist of superficial dermal granulomata, usually surrounding hair follicles and sweat ducts, but occasionally independent of either. The infiltrate is composed of epithelioid cells with some Langhans giant cell formation and a narrow margin of lymphocytes at the periphery. The adjacent capillaries may show a lymphocytic perivascular infiltrate and there may be non-specific secondary changes in the epidermis above the lesions (Johnson, 1972). There is no caseation necrosis and changes are not seen in the subcutaneous tissue. Normally tubercle bacilli are not seen in the pathological specimens, nor can they be cultured from biopsy material. Some of the earlier writers, however, have claimed to identify bacilli in the skin lesions (Volk, i93O- The distribution of the granulomata around the appendages and the fact that the infiltrate often occupies several dermal papillae are helpful in differentiating lichen scrofulosorum from lichen nitidus. The cutaneous lesions of secondary syphilis are essentially perivascular and plasma cells are frequently seen in the infiltrate. In cutaneous sarcoidosis the granulomata tend to be more discrete with less surrounding lymphocytic infiltrate, but distinction of sarcoidosis from lichen scrofulosorum may be impossible on histological grounds alone. The course of the untreated disease tends to be slow, the lesions persisting for months or years. Spontaneous involution without scarring usually occurs eventually. In all our cases there has been an excellent response to antituberculous treatment, although not all writers have had the same experience (Schuhmachers, 1967). There is little doubt that tuberculous infection is an aetiological factor in the development of the cutaneous lesions of lichen scrofulosorum. The exact nature of the relationship between systemic
Lichen scrofulosorum
325
tuberculosis and the skin eruption is not completely clear. At one time the cutaneous lesions were considered to be reactions to 'toxins' of tubercle bacilli. It is now thought more likely that they result from haematogenous dissemination of bacilli in an individual strongly sensitive to Mycobacterium tuberculosis, though the primary focus of dissemination may not always be apparent. Lichen scrofulosorum appears to be a definite clinical and pathological entity, distinct from other forms of cutaneous tuberculosis and unrelated to sarcoidosis. Its prevalence in any community probably depends upon the prevalence of tuberculosis of the lymph nodes and it is felt that the condition is perhaps not so uncommon as the lack of recently doctunented cases would suggest. ACKNOWLEDGMENTS
We would like to thank Dr T.J.McEIwain of the Royal Marsden Hospital and Professor J.G.Scadding of the Brompton Hospital for permission to publish details of case 2. Case I was presented by Dr T.J.Delaney at a meeting of the St John's Hospital Dermatological Society on 6 June 1974. Case 3 was presented by Dr J.P.Ellis at a meeting of the Dermatological Section of the Royal Society of Medicine on 21 February 1974, and was subsequently listed in the Proceedings of the Royal Society of Medicine. REFERENCES DOSTROVSKY, A. & SAGHER, F . (t963) Dcrmatological complications of B.C.G. vaccination. British Journal of Dermatology, 75, 181. HEBRA, F . (t868) Lichen scrofulosorum. In: Diseases of the Skin (Translated and ed. by C.H.Fagge and P.H.PyeSmith), Vol. 2, p. 52. The New Sydenham Society, London. JOHNSON, W.C. (1972) Lichen scrofulosorum. In: Dermal Pathology (Ed. by J.H.Grahatn, W.C.Johnson and E.B.Helwig), p. 394. Harper Sc Row, Hagerstown, Maryland. LEVER, W.F. (t967) In: Histopathology of the Skin, 4th. edn, p. 301. Pitman Medical, London. MIESCHER, G . (1955) Dber katamnestischc Untersuchungen bei Fallen mit Tuberkulid. Dermatologica, n o , 23. MONTGOMERY, H . (1937) Histopathology of various types of cutaneous tuberculosis. Archives of Dermatology and Syphilology, 35, 698. OCKULY, O.E. & MONTGOMERY, H . (1950) Lichenoid tuberculid. Journal of Investigative Dermatology, 14, 415. RAUSCHKOLB, J.E. (1934) Tuberculosis of the skin: a clinical report. Archives of Dermatology and Syphilology, 29, 398. SCHUHMACHERS, R. (1967) 2 Fiille eines lichenoides Tuberkulids (= Lichen scrofulosorum). Hautarst, 18, 81. VOLK, R. (i93t) Lichen scrofulosorum. In: Handbuch der Haia-und Geschlechtskrankheiten (Ed. by J. Jadassohn) Vol. 10, p. 331, Springer, Berlin.
