Abstract

Case report
A 68-year-old man was transferred to our unit from a local hospital for further investigation of pyrexia of unknown origin (PUO). He had consulted his GP intermittently over the previous two years with non-specific aches and pains, and mild fatigue. Two weeks prior to admission he noticed low back pain followed by high fevers, chills, nausea and diarrhoea. Past medical history included a nephrectomy for renal cell carcinoma and ischaemic heart disease. He was born in Kenya, but had been resident in the UK for 30 years, and was married with three children. He had never smoked and drank minimal alcohol. Two months prior to admission he had travelled to Kenya for a holiday, and had not taken anti-malarial prophylaxis.
On presentation to our unit the patient was haemodynamically stable, with a temperature of 36.8°C. There was no lymphadenopathy or organomegaly, and there were no skin changes. Examination of the cardiovascular, respiratory, and neurological systems was unremarkable. Initial blood parameters demonstrated a mild normocytic anaemia (Hb 10.4 g/dL, MCV 85.8 fl), moderate renal impairment (urea 13.6 mmol/L, creatinine 234 mmol/L), significantly raised inflammatory markers (ESR 80 mm/hr, CRP 179 mg/L), and abnormal liver function (albumin 26 g/L, PT 14.5 s, ALT 47 U/L, ALP 715 U/L, GGT 300 U/L). Three thick and thin blood films were negative for malarial parasites as were multiple cultures of blood, stool and urine. A CT scan of the whole body was normal. Viral hepatitis, HIV and autoimmune serology were negative.
Over subsequent days low-grade fevers were noted, and the patient complained of mild intermittent headaches. Renal function improved with increased oral fluid intake, but liver function continued to deteriorate (albumin 16 g/L, PT 15.3 s, ALT 338 U/L, ALP 1071 U/L, GGT 496 U/L). An ultrasound scan of the liver and biliary system was normal. A percutaneous liver biopsy revealed granulomatous arteritis with prominent giant cell involvement (Figure 1). No acid-fast bacilli were identified. Prompted by this unexpected finding, a temporal artery biopsy was undertaken which confirmed patchy granulomatous inflammation with giant cell involvement. A final diagnosis of giant cell arteritis (GCA) was, therefore, made and treatment with high-dose oral prednisolone was initiated. His symptoms improved rapidly, and he was discharged shortly afterwards on a tapering prednisolone regimen. Liver function has since normalized.
Discussion
PUO is not uncommon and most often ends up being attributed to infective, malignant or
inflammatory conditions. GCA is a systemic vasculitis that almost exclusively affects
adults older than 50 years and classically involves medium- to large-sized extra-cranial
branches of the carotid artery. Liver biopsy showing giant cells and swollen arterial endothelial cells
Mild biochemical abnormalities of liver function are common in GCA and the closely-related condition polymyalgia rheumatica (PMR), which may cause diagnostic confusion and lead to a delay in diagnosis. Most often this comprises an elevation in the serum level of ALP, but also occasionally ALT or AST. 2 Radiological abnormalities of the liver have been demonstrated in some patients with GCA/PMR, with abnormal liver isotope scans and reduced hepatic arterial blood flow reported in a small case series. 3 The histological basis of these abnormalities remains incompletely defined, and liver biopsy in unselected GCA/PMR cases may be unrewarding. Histological findings encountered under light microscopy have included normal liver, non-specific steatosis, cholestasis, hepatocellular necrosis, portal tract inflammation, and perisinusoidal lipocyte hyperplasia. 4 Electron microscopy has demonstrated ultrastructural damage to bile canaliculi in a solitary case report. 5 Overt hepatic giant cell arteritis 6, 7 or granulomatous inflammation of the liver 8 in patients with clinical and/or temporal artery biopsy-proven GCA/PMR is exceedingly rare.
GCA/PMR is an important, potentially sight- or life-threatening, but eminently treatable condition. The patient we have described had few of the classic features to suggest the diagnosis on clinical grounds. We suggest that abnormalities of liver biochemistry in the setting of a PUO should alert the clinician to the possibility of GCA even in the absence of cranial features, and consideration should be given to temporal artery biopsy. Liver biopsy may also be valuable in selected cases in defining the mechanism of hepatic involvement and directly influencing patient management. 9
An emerging investigational technique that may prove to be useful in grey PUO cases such as ours is that of fluorodeoxyglucose (FDG)-positron emission tomography (PET). FDG accumulates in malignant tissues as well as inflammatory cells and hence has the potential to confirm or refute several of the major differential diagnoses in PUO, including large vessel vasculitis. 10 Hybrid imaging combining FDG-PET with either computed tomography or magnetic resonance imaging is also the subject of active research, but it remains to be seen whether, and how, such techniques will fit into current diagnostic algorithms.
Footnotes
DECLARATIONS
Footnotes
Acknowledgements
Helene Mellor – for providing the histology photograph
