Article Text

PDF

Fever of unknown origin with seronegative spondyloarthropathy: an atypical manifestation of Whipple's disease
  1. C Várvölgyi1,
  2. T Bubán1,
  3. S Szakáll2,
  4. Z Hargitai2,
  5. L Galuska3,
  6. C Jeney4,
  7. G Kakuk1,
  8. J Gaál1
  1. 1Ist Department of Internal Medicine, University of Debrecen, Medical and Health Science Centre, Debrecen, Hungary
  2. 2Institute of Pathology, University of Debrecen
  3. 3Department of Nuclear Medicine, University of Debrecen
  4. 4H-MED Laboratory, Central Military Hospital, Budapest, Hungary
  1. Correspondence to:
    Dr J Gaál, Ist Department of Internal Medicine, University of Debrecen, Medical and Health Science Centre, Debrecen, 4012 Hungary, POB 19;
    gaalj{at}ibel.dote.hu

Statistics from Altmetric.com

Many authors emphasise the diagnostic difficulties and point out the multifaceted nature of Whipple's disease.1,2 Joint symptoms are present in 90% of all cases and may precede other disease manifestations by decades.3 We report here a case with fever of unknown origin accompanied by seronegative spondyloarthropathy with no typical gastrointestinal symptoms and initially negative upper panendoscopy. To confirm the diagnosis, the bacterial 16S ribosomal RNA sequence of Tropheryma whippelii was determined by polymerase chain reaction (PCR).

CASE REPORT

A 58 year old white man had a 12 year history of intermittent arthralgias and seronegative polyarthritis. In 1993, monolateral stage II sacroiliitis was disclosed with no definite cause. Low dose methylprednisolone treatment was started, but there was no clinical improvement. In 1998 the patient became febrile, lost 10 kg of weight but had no gastrointestinal symptoms. He underwent an extensive examination, including radiological examinations of the chest and paranasal sinuses, abdominal sonography, echocardiography, abdominal computed tomography, upper panendoscopy, bone marrow biopsies, whole body gallium-67 citrate scan, multiple blood, stool, and urine cultures, as well as serological investigations for known viruses and autoantibodies. Results of all these tests were normal or negative except for mild splenomegaly, transitory otitis, and temporary antinuclear antibody positivity. Fever of unknown origin was diagnosed. The patient often took antibiotics.

He was referred to our department in October 1999. Physical examination showed limitation in the lumbar spine, bilateral swollen and tender wrists, right sided proximal interphalangeal synovitis of the hand, minimal synovial fluid in the right knee, and bilateral tenderness of Achilles tendons. Radiographic examination showed bilateral stage II sacroiliitis. Non-differentiated seronegative spondyloarthropathy was diagnosed, and meloxicam (15 mg/day) and methotrexate (7.5 mg/week) were started.

In January 2000 the patient's fever (39°C) reappeared, but no malignant, infective, or autoimmune cause of the disease was shown. The differential diagnosis included rheumatoid arthritis, rheumatic fever, connective tissue disorders, reactive arthritis, adult Still's disease, sarcoidosis, arthritis with haematological and solid malignancies, familial Mediterranean fever, arthropathy associated with HIV infection, histiocytosis, some fungal infections, other seronegative arthropathies, including inflammatory bowel disease with spondyloarthropathy. Laboratory examinations showed raised C reactive protein level (160 mg/l), increased erythrocyte sedimentation rate (82 mm/1st h), and mild iron deficiency anaemia (haemoglobin, 116 g/l). All other laboratory parameters, including serum albumin, calcium, bilirubin, and HLA-B27 antigen, were normal or negative. Colonoscopy did not show mucosal changes characteristic of inflammatory bowel disease. Upper gastrointestinal series with small bowel roentgenograms showed abnormal enteral peristalsis. A whole body 67Ga citrate scan was performed again and showed increased accumulation in the small intestinal regions. Whipple's disease was suspected and repeated upper panendoscopy disclosed small whitish plaques in the duodenum. In the duodenal biopsy specimens, numerous periodic acid-Schiff (PAS) positive macrophages were seen (fig 1). PCR was unequivocally positive for bacterial 16S rRNA from the formalin fixed and deparaffinated samples.

Figure 1

Biopsy specimen of the duodenal mucosa. The stroma is infiltrated with macrophages having a coarse granular intracytoplasmic PAS positivity. Bar represents 100 μm. Periodic acid-Schiff (PAS) stain.

Meloxicam and methotrexate were continued with sulfamethoxazole-trimethoprim. From that time on the patient was free of fever and the arthralgia diminished markedly. Two months later the whitish plaques disappeared from the duodenum, but histological examination still detected macrophages with PAS positive inclusion bodies. PCR was intensely positive for bacterial 16S rRNA from the native samples.

The diagnosis of Whipple's disease is usually straightforward in cases with characteristic clinical manifestations. However, difficulties may arise when clinical features are atypical or the disease is oligosymptomatic. Despite the fact that dyspeptic symptoms were lacking, we repeated the upper panendoscopy in light of the patient's iron deficiency anaemia, weight loss, and questionable gallium scan report. Histological examination of duodenal mucosa showed PAS positive inclusions in macrophages in the lamina propria. In one series, endoscopic mucosal appearance was characteristic of Whipple's disease only in 72% of cases.4 In our patient, endoscopy was performed three times, but no report suggestive of Whipple's disease was recorded on the initial checkup.

DISCUSSION

The diagnosis of Whipple's disease generally relies on the histological detection of sickle form particle-containing cells stained red by the PAS reaction. However, the mucosal histological features are not always pathognomonic. Because some other PAS positive cells (for example, macrophage infiltrates in Mycobacterium avium or Rhodococcus equi infection of patients with advanced HIV infection) may mimic sickle form particle-containing cells of Whipple's disease, confirmation of the diagnosis is recommended.3 Electron microscopy is generally accepted as an accurate diagnostic tool for detecting T whippelii, but it is expensive, time consuming, and needs special handing of biopsy samples.5 Recently, the organism has been characterised by amplification of its unique 1321 base sequence of the 16S rRNA (rRNA, small subunit) gene from tissue infected with T whippelii.6 It offers a suitable alternative for diagnosing Whipple's disease in cases with atypical clinical features or when the histological diagnosis requires confirmation.3,5 PCR provides an important aid to histological examination of extraintestinal manifestations unaccompanied by gastrointestinal symptoms.2 PCR results did correlate with therapeutic results, but correlation between histology after treatment and clinical outcome has not been established.7

In conclusion, Whipple's disease should also be considered in clinical syndromes of fever of unknown origin and seronegative spondyloarthropathy, even in those patients who do not develop gastrointestinal symptoms. PCR is a useful diagnostic tool and provides definite diagnosis of this rare but treatable disease

REFERENCES

View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.