1 Department of Diabetes and Metabolism, Royal London Hospital, London
2 Department of Histopathology, St Bartholomew's Hospital, London, UK
Correspondence to: Dr T A Chowdhury, Consultant Physician, Department of
Diabetes and Metabolism, 5th Floor Alexandra Wing, Royal London Hospital,
Whitechapel, London E1 1BB, UK
E-mail:
TahseenChowdhury{at}bartsandthelondon.nhs.uk
In a person coming from the Indian subcontinent to the UK the combination of fever, night sweats and lymphadenopathy is most likely to be due to tuberculosis; but a less common diagnosis must be borne in mind.
CASE HISTORY
A Bangladeshi woman age 24, who had arrived in the UK three months previously, described episodes of fever and night sweats without rigors for the past three weeks; she had also developed non-tender cervical lymphadenopathy and an itchy rash over the trunk and arms. There was no story of cough, dyspnoea, haemoptysis, dysuria or gastrointestinal symptoms and she had not lost weight. A year earlier she had been investigated in Bangladesh for similar symptoms, when a fine-needle aspirate of a cervical node had shown chronic non-specific lymphadenitis; at that time a Heaf test had been negative and the chest X-ray normal, though the erythrocyte sedimentation rate (ESR) had been raised at 20 mm/h. She had been started on antituberculosis chemotherapy, but this had been changed after a few weeks to doxycyline for a presumed diagnosis of rickettsial fever.
On the present admission her temperature was 38°C. Firm lymphadenopathy was apparent in the anterior cervical chain of the neck and a single pea-sized submental lymph node could be felt on the right. There were no cardiac murmurs or stigmata of infective endocarditis, the lungs were clear and nothing abnormal was found on abdominal examination. The rash on her trunk and arms was pale pink, itchy and scaly. Haemoglobin was 12.3 g/L; mean cell volume 87.5 fL; white cell count low at 2.7x109/L (normal range 4-11); neutrophil count low at 1.2x109/L (2.5-7.58). Urea and electrolytes, liver function tests and calcium were all normal. ESR was 41 mm/h and C-reactive protein 10 mg/L (normal <5). Chest radiograph was normal, urinalysis clear and pregnancy test negative. Blood and urine cultures were negative, as was a full autoantibody screen, including antinuclear antibody and anti-dsDNA antibody.
Over the next 72 hours her low-grade pyrexia settled without antibiotic therapy. An excision biopsy specimen from the submental node showed zones of necrosis containing apoptotic cellular debris surrounded by pale staining histiocytes and scattered lymphocytes (Figure 1). There was no evidence of granuloma formation. The features were consistent with histiocytic necrotizing lymphadenitis, or Kikuchi's disease. At review, her pyrexia was less troublesome but the lymphadenopathy and rash had persisted. She was reassured that the condition was likely to settle spontaneously.
|
COMMENT
Kikuchi's disease was first described in Japan in 1972.1 Typically it presents with painless unilateral cervical lymphadenopathy, although adenopathy elsewhere, including mediastinal and retroperitoneal sites,2 is reported. 20% of cases show generalized lymphadenopathy, and splenomegaly is an occasional feature.3 Fever and neutropenia are common. Cutaneous manifestations of Kikuchi's disease are well recognized, and in a review of ten patients these included indurated erythematous macules, papules, plaques and nodules on the face, arms and trunk.4 Seemingly the patients with cutaneous lesions are those most likely to have high fever and liver dysfunction with raised transaminases. Less common clinical features include arthralgia, epididymitis, pancytopenia and bone marrow involvement. Most patients recover in time, although many have recurrences. No specific treatment is suggested apart from symptomatic therapy; severe cutaneous manifestations have been treated with oral prednisolone.4 A single fatal outcome, from myocardial infarction, has been reported.5
In the affected lymph nodes, involvement is almost confined to T-cell regions, with patchy paracortical necrosis consisting of eosinophilic fibrinoid material and nuclear debris.6 Foamy histiocytes and reactive T cells infiltrate the tissue but polymorphonuclear leucocytes are absent and plasma cells infrequent. The proliferation of reactive foamy histiocytes and abnormal nodal architecture make for possible confusion with malignant lymphoma or Hodgkin's disease; infective causes of lymph node hyperplasia such as tuberculosis can be readily distinguished.
The pathogenesis of Kikuchi's disease is not clear. This uncommon condition seems to affect women more than men, with a male to female ratio of around 4:1, and a mean age of 30 years at onset.6 Most of the reported cases are in persons of south Asian or oriental descent, though some of these were born in Europe or the US, suggesting a racial or genetic predisposition. An infective agent has been suggested, with toxoplasma, Yersinia enterocolitica and herpesvirus 6 all proposed.7 Apoptosis may be important: cytotoxic T lymphocytes are considered to be apoptotic effectors, as well as target cells, while histiocytes could enhance apoptosis.8 One suggestion is that Kikuchi's disease is a variant of systemic lupus erythematosusbased on cases in which histologically diagnosed cases have progressed to SLE.9 An association with autoimmune Hashimoto's thyroiditis has also been reported.9 Nevertheless, in the largest case series SLE developed in only 2 of 108 patients, so the association may be a chance finding.6
REFERENCES
This article has been cited by other articles:
![]() |
B. Rhodes and A. S M Jawad Kikuchi's disease J R Soc Med, October 1, 2004; 97(10): 507 - 507. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||