I Io

Letters to the Editor

function tests 1-4 and P U O 3-5 has been described recently. We suggest, as Bailey et al.fl and Pearl and Strauchen ~ have done, that K D must be considered a potential cause of P U O . Our patient's illness resolved following treatment with naproxen which is as Kapadia et al. 4 described, but why this drug is effective is not understood. T h e aetiology of the disease remains uncertain but an association with yersinia, 4,6 toxoplasma, 7 herpes virus 6, ~ EBV, 8 and more recently h u m a n immunodeficiency virus 9 and parvovirus B 191 infections has been suggested. Like Rivano et al., s we found an association with the Epstein-Barr virus.

Department of Infectious Diseases, University Hospital and School of Medicine, Coimbra, Portugal

Eduardo M . Rabaddo* Joaquim F. Oliveira Saraiva da Cunha Rui Cdrte-Real Antdnio A. Melifo-Silvestre

* Address correspondence to: Dr E.M. Rabadfio, Clinica de Doenqas Infecciosas, Hospital da Universidade, 3049 Coimbra Codex, Portugal.

References I. Meyer O, Ribard P, Belmatoug Net al. Trois cas de lymphad~nite de Kikuchi au cours du lupus eryth6mateux syst6mique. R61e du parvovirus BI9. Ann Mdd Intern 1991 ; I42(4): 259-264. 2. Ohta A, Matsumoto Y, Ohta, T et al. Still's disease associated with necrotizing lymphadenitis (Kikuchi's disease): report of 3 cases. J Rheumatol 1988; 15: 981-983. 3. Bailey EM, Klein NC, Cunha AB. Kikuchi's disease with liver dysfunction presenting as fever of unknown origin (letter). Lancet 1989; ii: 986. 4- Kapadia V, Robinson BA, Helen HB. Kikuchi's disease presenting as fever of unknown origin (letter). Lancet 1989; ii: 1519-I52O. 5. Pearl D, Strauchen JA. Kikuchi's disease as a cause of fever of unknown origin (letter). N EnglJ Med I989; 3zo(17): II47-II48. 6. Feller AC, Lennert K, Stein H et al. Immunohistology and etiology of histiocytic necrotizing lymphadenitis: report of three instructive cases. Histopathology I983; 7: 825-839. 7. Kikuchi M, Yoshizumi T, Nakamura H. Necrotizing lymphadenitis: possible acute toxoplasmic infection. Virchows Arch (A) 1977; 376: 247-253. 8. Rivano MT, Falini B, Stein H et al. Histiocytic necrotizing lymphadenitis without granulocytic infiltration (Kikuchi's lymphadenitis). Morphological and immunohistochemical study of eight cases. Histopathology I987; I I : 1oi3-Io27. 9. Pileri SA, Sabattini E, Costigliola P e t al. Kikuchi's lymphadenitis and HIV infection. A I D S I99I ; 5(4): 459-46I.

Gemella morbillorum

e n d o c a r d i t i s in a n i n t r a v e n o u s drug abuser

Accepted for publication I9 December ~991 Sir,

Gemella species rarely cause infective endocarditis. We wish to report a case of endocarditis caused by Gemella morbillorum. There appear to be only two previous reports of endocarditis caused by G. morbillorum, formerly Streptococcus morbillorum in the U.K. 1 Streptococcus morbillorum was recently transferred to the genus Gemella as a result of D N A homology studiesfl A i9-year-old male intravenous drug abuser presented in October 199o with a 2 weeks' history of malaise, night sweats and intermittent fever. He had started injecting heroin at the age of I4 years. Since then, he had been on various drug rehabilitation schemes, all unsuccessful. In the months before his admission to hospital, he had been

Letters to the Editor

III

injecting Temgesic ® (buprenorphine) into his arm veins. H e was a low-risk hepatitis B carrier. His H I V status was negative at that time. Clinical examination on admission revealed a swinging fever and pallor. H e had neither a cardiac m u r m u r nor enlargement of the liver or spleen. L a b o r a t o r y investigations revealed a haemoglobin concentration of 13 g/dl, white blood cell count of Io x IO9/1 and C-reactive protein (CRP) of 94 rag/1. On the p r e s u m p t i v e diagnosis of infective endocarditis, treatment began with intravenous flucloxacillin 2 g, 6 hourly, benzylpenicillin r 8 g, 6 hourly and gentamicin 80 rag, 8 hourly. T h r e e sets of blood cultures, within 24 h of admission, yielded an organism of variable m o r p h o l o g y and indeterminate Gram-staining. It was oxidase and catalase negative, sensitive to penicillin, ampicillin and vancomycin but resistant to the aminoglycosides on disc sensitivity testing. Antibiotic therapy was therefore changed to intravenous benzylpenicillin I.z g, 4 hourly and gentamicin 80 mg, 8 hourly. T h e organism was identified by means of ' A P I 20 S t r e p ' as G. morbillorum. T h i s finding was confirmed by the Streptococcus Reference Unit, Central Public Health Laboratory, London. T h e m i n i m u m inhibitory concentration ( M I C ) and m i n i m u m bactericidal concentration (MBC) for penicillin of this isolate were ~< 0"03 mg/1. T h e M I C and M B C for gentamicin were both 4'0 mg/1, i.e. a low degree of resistance. Over the ensuing days, a m u r m u r of tricuspid incompetence could be auscultated. An echocardiogram showed vegetations on the tricuspid valve. After 4 weeks of intravenous antibiotic therapy, the patient improved clinically while the C R P gradually fell to within the normal range ( < Io mg/1). An echocardiogram repeated after therapy showed a reduction in the size of the tricuspid vegetations. Gemella morbillorum is part of the commensal flora of the healthy oropharynx. On reviewing the literature on S. morbillorum endocarditis, we found that most cases have been reported in the U.S.A. : one case by Coto and Berkfl two cases (both from failed dental prophylaxis) by D u r a c k 4 and eight of a total of 364 cases of streptococcal endocarditis by Facklam. 5 O f the only two cases reported in the U . K . , 1 one was associated with previous anal surgery (anal stretching and dilatation followed by left lateral subcutaneous sphincterotomy post-sigmoidoscopy) and one with dental instrumentation (drainage of an abscess and tooth extraction). F u r t h e r m o r e , of the only four reported cases of endocarditis caused by G. haemolysansfl which is closely related to G. morbillorum, 2 three were associated with dental procedures. In our case, there was neither an acute dental problem nor a history of recent dental work. O n questioning the patient and by observation it was evident that the patient o f t e n ' licked his fingers' before rubbing them over the venepuncture site where he had injected drugs. T h i s practice may therefore have been the cause of infection. I n Glasgow Royal Infirmary, over the past 6 years, there have been 30 cases of endocarditis a m o n g intravenous drug abusers. I n c o m m o n with other centres, Staphylococcus aureus has accounted for most of the cases. We believe that ours is the third case of G. morbillorum endocarditis reported in the U . K . and the first in an intravenous drug abuser. (We thank Professor D. H. Lawson for permission to report this case.)

Department of Medicine, Royal Infirmary, Glasgow G4 oSF Department of Microbiology, Royal Infirmary, Glasgow G4 oSF, Scotland U.K. * A u t h o r to w h o m c o r r e s p o n d e n c e s h o u l d b e s e n t .

Ewan Bell

A. Christine McCartney*

I IS

Letters to the Editor References

I. Maxwell S. Endocarditis due to Streptococcus morbillorum. J Infect I989; x8: 67-72. 2. Kilpper-B~ilz R, Schleifer KH. Transfer of Streptococcus morbillorum to the genus Gemella as Gemmella morbillorum comb. nov. Int J System Bact I988; 38: 442-443. 3. Coto H, Berk SL. Endocarditis caused by Streptococcus morbillorum. Am J Sci r984; 287: 54-58. 4. Durack DT, Kaplan EL, Bisno AL. Apparent failures of endocarditis prophylaxis. Analysis of 52 cases submitted to a national registry. J A M A 1983; 25o: 23 I8-2322. 5. Facklam RR. Physiologial differentiation of viridans streptococci, ff Clin Microbiol I977; 19: x84-2or. 6. Buu-Hoi A, Sapoetra A, Branger, Acar JF. Antimicrobial susceptibility of Gemmella haemolysans isolated from patients with subacute endocarditis. Fur J Clin Microbiol I982; I : IO2-IO6.

Capnocytophaga canimorus infection and acute renal failure Accepted for publication 27 November I 9 9 I Sir, T h e bacterium Capnocytophaga canimorus, a m o u t h commensal of cats and dogs, is an unusual cause of septicaemia in h u m a n beings. H u m a n infection usually follows animal bites or licking of broken skin?. ~ We wish to report a case complicated by acute renal failure and to review the literature with regard to renal involvement. A previously well 56-year-old man, without any recent animal bite, presented with a 2 days' history of malaise, fever and dyspnoea. H e rapidly developed multiple organ failure requiring ventilatory and circulatory support. Bacilli were seen within neutrophils in a conventionally Wright-stained peripheral blood smear made on the patient's admission to hospital. Gram-staining revealed intracellular G r a m - n e g a t i v e rods. T r e a t m e n t with piperacillin and gentamicin began. T h e organism was later identified as C. canimorus, sensitive to penicillins, resistant to gentamicin. On transfer to our unit for haemodialysis, the patient's A P A C H E - I I score was 30. 3 S e r u m creatinine was 560/~mol/l and the W B C count was I8"6 x iog/l (85 % neutrophils). T a r g e t cells and Howell-Jolly bodies were later noted in blood films. Abdominal ultrasound examination revealed an abnormal spleen with an irregular, possibly infarcted, lower pole. T h e extremities became gangrenous, requiring below-knee amputations and disarticulation of the left-hand. Ultrasound scanning of the splenic bed revealed a collection of cystic and solid material with little residual splenic tissue. T h e patient was discharged from hospital 6 weeks after admission with normal renal function. Capnocytophaga canimorus is increasingly recognised as a cause of catastrophic h u m a n infection. Morphology, cultural techniques and characteristics are well described? It is invariably resistant to gentamicin but sensitive to a wide range of other antibiotics. 5 Clinical features 5-7 range f r o m m i n o r local infection to overwhelming septicaemia. T h e course of our case was dominated by established septic shock and consequent multiple organ failure. Such severe shock is unusual in the i m m u n o c o m p e t e n t but C. canimorus septicaemia characteristically presents in this fashion. Pre-existing i m m u n o i n c o m p e t e n c e is often associated with infection?' 7-9 Our patient's history did not suggest i m m u n o i n c o m p e t e n c e ; he had an appropriate leucocytosis and normal concentrations of immunoglobulins. His spleen was noted to be abnormal on admission and the haematological changes (target cells and Howell-Jolly bodies) seen later were compatible with functional asplenism due to multi-infarct ' auto-splenectomy.'