Case report

In December 1993 a 61 year old woman presented with a two month history of asymptomatic “warty” lesions on her vulva but also involving the axillae, submammary areas and, to a lesser extent, the upper limbs. The lesions were variable in size, appearing as smooth yellow papules some of which were confluent, and one vulval lesion was becoming large and uncomfortable. The patient's full blood count, erythrocyte sedimentation rate, C-reactive protein, urea, electrolytes, liver function tests, thyroid function, lipid profile, glucose tolerance test, serum immunoglobulins, and serum and urine electrophoresis were all normal. Biopsy specimens from the vulva and skin showed different stages of the time course of this entity with vacuolated histiocytic cells interspersed with mixed inflammatory infiltrate and giant cells (fig 1A). There was a predominance of fibrotic changes with predominantly spindle shaped histiocytes in the larger nodular lesions. Immunocytochemical tests were negative for markers of histiocytosis-X (S-100 protein, LN3 and peanut agglutination) and positive for factor XIIIa, HAM 56, and KP1. There were no cells containing Birbeck granules at electron microscopy and the cells showed an irregular scalloped border with a centrally placed nucleus. In view of the clinicopathological presentation and immunocytochemical markers the diagnosis of xanthoma disseminatum was confirmed. She was commenced on topical clobetasol propionate 0.05%; over the following few months she developed lesions elsewhere on the trunk and buttocks with little response to the topical steroids.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

(A) Section of skin showing mixed cellular infiltrate composed of pale vacuolated histiocytic cells with some multinucleated forms. Lymphocytes and other inflammatory cells are present in the background (stain: H & E; high power magnification). (B) Section of trachea (post mortem) showing nodular masses of pale tissue (arrow) consisting almost entirely of histiocytes with a few giant cells. The nodules are interspersed between the cartilage plates (arrow head) (stain: H & E; medium power magnification). (C) Immunostaining of section of trachea using KP1 showing positive staining of macrophages (high power magnification).

In February 1995 she developed dyspnoea, wheeze, and a weak voice. Examination at that time revealed a fixed wheeze in the left lung and slight hoarseness of the voice. Spirometric tests revealed obstructive airflow with a forced expiratory volume in one second (FEV₁) of 1.0 litre (40% predicted) and forced vital capacity (FVC) of 2.5 litres (85% predicted). The chest radiograph was unremarkable. Bronchoscopic examination showed bilateral vocal cord oedema and a nodule at the anterior commissure. Throughout the trachea and main airways there were numerous flat pale white lesions and biopsy specimens showed a mucosal infiltrate of histiocytic cells with a foamy cytoplasm. There was no evidence of granuloma formation and Ziell-Nielsen stain failed to reveal acid fast bacilli. Blood investigations including serum angiotensin converting enzyme (ACE) were normal. She was commenced on oral prednisolone 10 mg daily and after six months of monitored treatment there had been no subjective or objective evidence of improvement in the chest. The prednisolone was reduced to 5 mg daily without deterioration in symptoms.

In 1996 she was reviewed and still complained of a weak voice and dyspnoea. During the interval she had continued to develop further cutaneous lesions consistent with XD. Chest examination was normal. Pulmonary function tests were unchanged and confirmed airflow obstruction: FEV₁ 1.0 litre (40% predicted) and FVC 2.6 litres (87% predicted) with normal gas transfer. Flow volume loops excluded fixed upper airway obstruction. Computed tomographic (CT) scanning of the thorax showed significant thickening of the trachea and large airways, and minor bronchial wall thickening throughout the lungs. There was no evidence of interstitial lung disease. Blood investigations were again normal. She was given a trial of inhaled beclomethasone which had no beneficial effect and was discontinued after two months.

The patient was reassessed in 1997 because of worsening dyspnoea. Chest examination was again unremarkable and oxygen saturation was 95%. FEV₁ had fallen slightly to 0.8 litres. A repeat bronchoscopic examination showed small submucosal nodules in the trachea and main airways, but biopsy specimens were non-diagnostic.

Forty eight hours later she became increasingly unwell and suffered a respiratory arrest at home. Resuscitation was unsuccessful and necroscopic examination confirmed numerous submucosal nodules in the epiglottis, trachea, and all the bronchi (fig 1B). Nodules were also present on the visceral pleura, the epicardium, and the lower third of the oesophagus. Histologically all of these tissues showed infiltration of foamy macrophages and histiocytic proliferation with lipid storing cells and immunocytochemical examination confirmed that they were also non-X histiocytic cells (fig 1C). There was no cytological atypia or evidence of neoplastic transformation in any of the sites; these have been compared with histological examination of the skin and confirm the appearances of XD. The cause of death was stated as respiratory failure and bronchospasm secondary to XD.

Discussion

Xanthoma disseminatum is one of several heterogeneous conditions known collectively as the primary histiocytic dermatoses1 ,2 or cutaneous syndromes of non-X histiocytoses,3 so called to differentiate them from the more recognised condition histiocytosis-X (HX). The condition is caused by a proliferation of non-X histiocytic cells, probably from a macrophage/monocyte origin,4 and was probably first described in 1867 by von Grafe5 and subsequently as a distinct entity by Montgomery and Osterberg in 1938.6 Only about 100 cases of XD have been reported in the world literature; the condition appears to affect predominantly male children and young adults but has been described in both sexes and in all age groups.7

As XD and HX share clinical and pathological features, difficulties may arise in diagnosis but immunocytochemical and ultrastructural differences have recently been recognised.4 Electron microscopy shows Birbeck (Langerhans') granules in HX which are absent in XD, while more phagosomes are seen in XD and other non-X histiocytoses compared with HX.4 ,9

Immunocytochemistry has shown differences between the proliferating Langerhans' cells in HX which label positively for S-100, LN3, and peanut agglutination in contrast to cells of macrophage/monocyte origin in XD (and other non-X histiocytoses) which are strongly positive for factor XIIIa, KiM6, and KP1 markers.4 ,9

Typically XD affects the flexural areas and skin lesions may regress spontaneously after several years.1 ,8 ,9 More commonly a persistent mucocutaneous form occurs and treatment with steroids, immunosuppressive agents, and radiotherapy is usually unsuccessful1 ,2 ,8-10 despite the successful use of agents such as etoposide, the vinca alkaloids, and methylprednisolone in Langerhans' cell histiocytosis.11 Selective cutaneous lesions in XD may be ablated by dermabrasion or electrodesiccation.1 Diabetes insipidus is a common systemic manifestation of the disease but may be mild and transient2 and involvement of mucous membranes may be observed in 30–50% of cases, typically affecting most frequently (in order) the larynx, pharynx, mouth, trachea, epiglottis, and tongue.2 ,8 ,9

When the pharynx and upper airways are involved symptoms of dyspnoea and dysphagia may occur and endoscopic examination may reveal plaques covering the mucosal surfaces. These can cause life threatening obstruction and asphyxiation requiring tracheostomy and has been reported in a total of eight patients.7 ,9 ,12-14 This is the most recognised cause of morbidity in XD, but tracheostomy may successfully relieve obstruction to the larynx with patients continuing in good health at subsequent follow up.

In our patient necroscopy showed no evidence that upper airway obstruction was responsible for asphyxiation but that death was caused by bronchospasm secondary to extensive involvement of the medium and small sized bronchi leading to respiratory failure. This complication has not been described since 1925 when Turner et al 14 reported the death of a 22 year old girl with XD who also died from respiratory failure secondary to airway infiltration including the bronchi. Although infiltration of the bronchi has been reported at necropsy in one other case of XD,7 it was not the cause of death as seen in our patient.

Xanthoma disseminatum is a rare mucocutaneous xanthomatosis considered to be benign and self-limiting and unresponsive to various modes of treatment. Involvement of the larynx may, however, lead to life threatening complications necessitating tracheostomy with a satisfactory long term prognosis but, when the lower respiratory tract is affected as in this case, the prognosis is poor leading to respiratory failure and death.