JRSM >
Volume 98, Number 4 >
Pp. 161-163
1 Department of Diabetes, Norfolk and Norwich University Hospital NHS Trust,
Norwich NR4 7UY, UK
2 Department of Microbiology, Norfolk and Norwich University Hospital NHS Trust,
Norwich NR4 7UY, UK
Correspondence to: Dr Ketan Dhatariya, Elsie Bertram Diabetes Centre, Norfolk and Norwich University Hospital NHS Trust, Colney Lane, Norwich NR4 7UY, UK E-mail: ketan.dhatariya{at}nnuh.nhs.uk
In a patient with unexplained fever and a potentially hypercoagulable state, a history of sore throat may be of crucial importance.
CASE HISTORY
A man aged 67 was admitted after six weeks of night sweats, rigors and
fatigue. He complained of nausea but there had been no vomiting, abdominal
pain or diarrhoea and currently he had no respiratory symptoms. The
patient's daughter later revealed that she and her father had experienced
mild sore throat and coryzal symptoms eight weeks earlier. The medical history
included tuberculosis at age 21, an ischaemic stroke, type 2 diabetes mellitus
diagnosed eighteen months ago and poorly controlled on diet (HbA1C
9.0% on admission), congestive heart failure with paroxysmal atrial
fibrillation, benign IgG (
) paraproteinaemia diagnosed twenty-two
months before admission, and stable chronic renal failure secondary to renal
vascular disease (creatinine 300 µmol/L). His drugs on admission were
amiodarone, gliclazide, furosemide, atorvastatin, spironolactone, and
warfarin. His international normalized ratio (INR) had been 2.92 two months
before admission and 3.48 two weeks before admission.
On examination he was apyrexial but sweaty, heart rate 95/min, blood pressure 140/80 mmHg; oxygen saturation was 96% on room air. He was noted to have very poor mouth and dental hygiene. Initial investigations showed haemoglobin 11.3 g/dL, white cell count 24.5 x 109/L (neutrophils 21.2), erythrocyte sedimentation rate 123 mm/h, C-reactive protein (CRP) 107 mg/L, INR 2.5, urea 18.9 mmol/L, creatinine 301 µmol/L, albumin 30 g/L, globulin 38 g/L, alkaline phosphatase 175 U/L, gamma-glutamyltransferase 180 U/L, glucose 14.3 mmol/L. On electrocardiography the only abnormality was long-standing first-degree atrioventricular block; the chest X-ray was reported as showing an enlarged heart with evidence of old tuberculosis. Blood and urine cultures taken over the next 4 days were negative. An echocardiogram showed mild impairment of left ventricular function but nothing else abnormal. While he was in hospital his temperature rose to 39.4°C, but no antibiotics were given because no source of infection was identified. The leukocyte count remained high at 24.1 x 109/L and his CRP rose to 130 mg/L. A haematology review (requested because of the benign paraproteinaemia) was unrewarding, and an autoantibody screen for collagen vascular disease was negative. There was no evidence of paroxysmal nocturnal haemoglobinuria, and levels of protein S and factor V Leiden were normal. Protein C levels were not determined because the patient was on warfarin. IgM and IgG anticardiolipin antibody levels were within normal limits. Antithrombin concentrations were normal. We did not investigate the possibility of a mutation in the prothrombin gene. CT of the abdomen and pelvis suggested the presence of a cholangiocarcinoma with tumour or thrombus within the portal vein. Subsequent MRI confirmed extensive portal venous thrombosis but showed no evidence of a cholangiocarcinoma.
During a rigor twelve days after admission another set of blood cultures was taken, and these grew anaerobic Gram-negative bacilli identified as Fusobacterium nucleatum. The bacterium was sensitive to clindamycin, metronidazole and cefotaxime. On intravenous clindamycin 600 mg three times daily the patient swiftly lost all his symptoms. He was discharged on long-term warfarin.
COMMENT
F. nucleatum has most often been associated with internal jugular venous thrombosis, initially described by Lemierre in 1936.1 To our knowledge, only five cases of portal vein thrombosis with F. nucleatum or F. necrophorum have been reported (Table 1).2–6
|
This case is unusual for two reasons. First, the portal vein thrombosis developed despite good anticoagulation. Second, the patient had diabetes. In two of the previous cases the patients had experienced pharyngitis in the weeks beforehand and the bacterium was presumed to have gained entry via the oropharynx;3,6 in the other three the gastrointestinal tract was judged the likely route.2,4,5 We assume that our patient's history of sore throat was relevant, though we disovered this only late in the course of his illness, when we quizzed his daughter closely
How might Fusobacterium spp. predispose to thrombosis? One suggested mechanism is thrombogenesis by the lipid A component of the bacterial lipopolysaccharide endotoxin; another is the binding of Fusobacterium to human plasminogen, activating local proteolysis and tissue damage.3,4 Our patient was well anticoagulated with warfarin and his clotting markers were normal; however, he did have two hypercoagulability factors—namely, diabetes7 and a paraproteinaemia.8 We conclude that, in a patient such as this with unexplained fever and abdominal symptoms, a history of upper respiratory tract infection should be sought and blood cultures should be taken to detect this fastidious anaerobic organism.
REFERENCES
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
