Fever and Increased Gastrointestinal Uptake on Positron Emission Tomography after Anti-Tumour Necrosis Factor Therapy: A Case Report of Whipple’s Disease Open Access

Whipple’s disease is a chronic, multisystem infectious disease caused by the actinomycete Tropheryma whipplei; “Tropheryma” means “Nourishment Barrier” in Greek, and “Whipplei” is attributed to George Hoyt Whipple, who described the first case of Whipple’s disease in 1907 [1]. Tropheryma Whipplei is usually found in soil; the exact transmission route and pathophysiology are poorly understood, but it is thought to involve multiple factors, including genetic, immunological, and bacterial virulence factors. Whipple’s disease is rare, with a reported prevalence of 1/1,000,000 affecting men more frequently; however, a recent population-based epidemiological study of Whipple’s disease in the USA showed a higher prevalence of 9.8/1,000,000 with equal prevalence in males and females [2].

Whipple’s disease has diverse symptomatology, including joint pain, weight loss, abdominal pain, diarrhoea, and skin hyperpigmentation. Other less common but more severe manifestations include neuropsychiatric, cardiac, respiratory, and eye involvement.

In this case report, we present a patient diagnosed with Whipple’s disease without typical gastrointestinal symptoms at the time of diagnosis. The non-rheumatological features of Whipple’s disease manifested only after treatment with an anti-tumour necrosis factor (TNF) agent for bilateral sacro-ileitis.

Written informed consent was obtained from the patient to publish this case report and any accompanying images. The CARE Checklist has been completed by the authors for this case report and is attached as online supplementary material (for all online suppl. material, see https://doi.org/10.1159/000538462).

A man in his 40s was admitted to our hospital with a history of fever, weight loss, ongoing arthralgia, and persistently high inflammatory markers. He did not have chronic medical issues or allergies and did not use any medications before this illness. He was married, never smoked, had an alcohol intake of 70–80 g weekly, and lived in an urban area. He had no family history of auto-immune, malignant, or other illnesses. He travelled to underdeveloped and tropical countries once or twice yearly for work; he had stayed in rural areas but never been unwell while overseas. He owned a dog for a long time with no other exposure to animals and never used intravenous or other illicit substances.

His symptoms started 4–5 years before the current admission, with intermittent back and migratory peripheral joint pain involving mainly the small joints of the hands and feet but also large joints, including knees and shoulders. The pain was described as a mild ache with inflammatory features, including severe pain at night and on waking up that improved with mobility; however, this was not associated with swelling, erythema, or systemic features, and he was generally feeling well. He used over-the-counter fish oil and turmeric for his arthralgia.

His arthralgia worsened 9 months before admission, with more severe lower back pain, specifically at the sacroiliac joints. He was reviewed by a rheumatologist who arranged blood tests which showed normocytic anaemia, neutrophilia, lymphopenia, elevated erythrocyte sedimentation rate, and C-reactive protein (CRP), with negative HLA-B27 test (Table 1). An MRI of the spine and sacrum showed bilateral sacro-ileitis and ankylosing spondylitis was diagnosed. He was initially treated with secukinumab for 4–5 months with no improvement in his clinical or biochemical markers. Due to the lack of response, further investigations were arranged, including an FDG-PET scan which showed a diffuse uptake throughout the stomach and mild focal uptake in the caecum and ascending colon, which was thought to be non-specific; there was an avid 21 × 16 mm enlarged para-aortic lymph node uptake and multiple mildly to moderately avid enlarged mesenteric nodes. Haematology review and opinion was that these lymph nodes were probably reactive. The patient used naproxen as needed to control his arthralgia, and the increased stomach uptake on FDG-PET was thought to be gastritis secondary to naproxen use.

Due to ongoing symptoms and elevated inflammatory markers, he was commenced on a second-line treatment for ankylosing spondylitis with adalimumab 2 months before admission. Paradoxically, he suffered worsening joint pain associated with new low-grade fever, night sweats, lethargy, anorexia, and worsening of his biochemical markers, necessitating admission for further investigations.

On admission, he reported a weight loss of 15 kg over 4 months, low-grade fever, and drenching night sweats for 2–4 weeks. He had recently developed a cough of 6 weeks duration, productive of a minimal amount of white-yellow sputum. He did not complain of abdominal pain, change in bowel habits, or other gastrointestinal symptoms.

On examination, he looked unwell and cachectic with generalised muscle wasting and a fever of 38°C; other vital observations were within normal limits, with no rashes or lymphadenopathy. He did not have a cardiac murmur or other stigmata of endocarditis. His abdominal examination was unremarkable, soft and non-tender to palpation, with no organomegaly and normal audible bowel sounds. He did not have joint swelling or erythema.

Blood test results showed worsening anaemia, ongoing neutrophilia, and further elevation of inflammatory markers. Extensive workup for the uncommon infectious causes of fever was arranged, which were within the normal or negative range (Table 2). The faecal calprotectin level was mildly elevated at 73 μg/g. Transthoracic echocardiography showed normal biventricular function, normal valvular structure, and function with no features of endocarditis. MRI of the spine and sacrum was repeated and showed active bilateral sacro-ileitis with chronic changes. Due to the undifferentiated nature of the illness, the FDG-PET scan was repeated, which showed similar results to the first FDG-PET scan with increased uptake in the stomach and caecum (Fig. 1). Endoscopic examinations of the upper and lower gastrointestinal tract were arranged; a paediatric colonoscope was used for upper examination and was inserted deeply into the jejunum; it showed severe nodular gastritis throughout the stomach, a medium-sized (2 cm) punched-out ulcer in the gastric body, the duodenal and jejunal mucosa appeared normal with no changes. Multiple biopsies were taken from the examined areas and sent for investigations, including Tropheryma Whipplei PCR; the colonoscopic examination was unremarkable, including the caecum and terminal ileum (Fig. 2).

Biopsies were reviewed and showed non-specific inflammatory changes throughout the gastrointestinal tract, with a mixed inflammatory cell infiltrate. Some villous blunting with foveolar metaplasia was noted in the small bowel mucosa. In many of the biopsies, more marked in the terminal ileum, there are scattered and focally clustered foamy histiocytes. These were demonstrated to contain granular periodic acid-Schiff (PAS) and diastase PAS-positive material (Fig. 3). PAS staining was not done on the gastric biopsies due to no inflammatory or other histological changes. The histological features were not deemed classic due to the lack of expansion of the lamina propria by foamy histiocytes and the absence of dilated lymphatic vessels. However, given the strong clinical suspicion, a diagnosis of Whipple’s disease remained probable. Tropheryma whipplei PCR was positive in the jejunum with a strong cycle threshold value of 23; this led to the conclusion that the results were compatible with the PCR diagnosis of Whipple’s disease. In retrospect, the granular material within foamy histiocytes is consistent with an atypical histological presentation of Whipple’s disease. Serum Tropheryma whipplei PCR was tested and returned positive; serum IgM and IgG against Tropheryma whipplei were not detected.

Despite no neurological features, proactive testing was conducted to investigate central nervous system (CNS) involvement. The cerebrospinal fluid (CSF) Tropheryma Whipplei PCR was positive with a cycle threshold value of 39. MRI of the Brain showed a small hyperintense FLAIR/T2 signal foci in the subcortical white matter of the frontal lobes bilaterally; this was thought to be non-specific, and there was no cerebral atrophy, mass lesions, or contrast enhancement suggestive of advanced neurological Whipple’s disease.

The patient was treated with intravenous ceftriaxone 2 g daily for 4 weeks (rather than 2 weeks due to CNS involvement) and continued to use naproxen for arthralgia as needed, with pantoprazole 20 mg daily due to gastritis. He was followed up closely and did not develop side effects of intravenous antibiotic treatment. He was monitored closely on the commencement of intravenous antibiotics with no evidence of immune reconstitution inflammatory syndrome.

He was reviewed 4 weeks later; his back pain was resolved with a significant improvement in his lethargy, anorexia, and he started regaining weight. His follow-up blood tests showed an albumin level of 42 g/L, ESR of 9 mm/h, haemoglobin of 127 g/L, and CRP level of 6 mg/L. Trimethoprim/sulfamethoxazole 160/800 mg twice daily for 1 year was commenced.

A further follow-up plan will involve monitoring for oral antibiotic side effects and symptoms of relapse during treatment, especially neurological relapse. MRI brain, CSF PCR, and endoscopic examinations with biopsies will be repeated at the end of 12 months of antibiotic treatment to confirm negative results and resolution of the infection before ceasing the antibiotic.

Whipple’s disease is rare, with high morbidity and mortality if not treated [3]. Diagnosis remains challenging, especially with atypical presentations, as in this case. It requires a high index of suspicion, and PCR testing leads to a more accurate diagnosis.

A Spanish clinical review of 91 patients reported that the earliest and most common pre-diagnosis symptoms were rheumatological symptoms (57% of the patients), mainly presenting as oligoarticular arthralgias rather than arthritis, with migratory polyarthralgia being the least common. The average time between joint symptoms and the diagnosis was 3–4 years. At the time of diagnosis, weight loss was the most prevalent symptom (80%), followed by gastrointestinal symptoms, including diarrhoea in 63% and abdominal pain in 24%. Skin manifestations are prevalent and include hyperpigmentation, purpura, and nodules. Joint symptoms are less prevalent in this phase of the illness (20%) compared to the pre-diagnosis [4].

The challenge in the presented case was the absence of typical gastrointestinal symptoms of Whipple’s disease. The worsening of symptoms with anti-TNF therapy raising the possibility of Whipple’s disease has been reported in the literature [5]. The globulin levels in our patient were consistently normal with mildly low IgM levels; this contrasts with usually elevated serum IgM levels in patients with Whipple’s disease [6].

CNS involvement in Whipple’s disease can be asymptomatic. Yet, it carries a worse prognosis and a higher risk of relapse even in patients who did not have neurological symptoms on initial diagnosis. Neurological manifestations can include non-specific symptoms like headache or specific symptoms related to the areas affected, which can involve any part of the brain or spinal cord. Investigations for neurological involvement include CSF analysis, which might show elevated protein level and pleocytosis; however, normal cytology and biochemical markers are reported in more than 50% of cases; hence, CSF Tropheryma Whipplei PCR and PAS staining are required. The neurological investigations also include an MRI of the whole spine and brain, which might show signs of atrophy, mass formation, enhancement, and white matter changes. Neurological involvement has significant implications on management; the initial intravenous antibiotic treatment should be extended to 4 weeks rather than 2 weeks. Oral antibiotic treatment might be extended for 2 years rather than 1 year, and all patients should have repeated MRI and CSF to confirm the resolution of abnormalities before ceasing the antibiotics [7, 8].

In conclusion, Whipple’s disease can present with no gastrointestinal manifestations. Arthritis worsening with anti-TNF treatment can be a clue to an infectious process (e.g., Whipple’s disease or other infectious diseases). Whipple’s disease should be considered a differential diagnosis in patients with increased uptake in the gastrointestinal tract on FDG-PET. Multiple tissue, fluid, and faeces testing for PCR should be sent. A paediatric colonoscope should be used to perform a push enteroscopy, or an antegrade balloon enteroscopy should be used, aiming for deeper insertion into the jejunum to take biopsies. Proactive testing for neurological involvement should be conducted even in asymptomatic patients. Treatment involves intravenous antibiotics (usually ceftriaxone 2 g daily) for 2 weeks; meropenem is an alternative for patients with penicillin allergy [9]. Followed by oral combined antibiotics (trimethoprim/sulfamethoxazole 160/800 mg twice a day) for 1 year, cefixime is an alternative for patients with sulphur allergy or renal impairment. The duration of antibiotics should be doubled if there is CNS involvement. All patients should have confirmation of infection resolution before ceasing antibiotics [10]. Patient education and vigilant monitoring for relapse should be conducted during and after stopping antibiotics.

The authors have no acknowledgements to make.

Ethical approval was obtained from the “Clinical Innovation and Audit Committee (CIAC)” at Macquarie Health on June 28, 2023, reference number: MQCIAC2023008. Written informed consent was obtained from the patient to publish the details of their medical care and any accompanying images.

The authors have no conflicts of interest to declare.

This case report did not receive funding.

A.A.: collecting the data, formatting, and writing the manuscript. B.J.: gastroenterology review, endoscopy images, and final approval. F.J.: rheumatology review and final approval. M.P.: infectious diseases review and final approval, editing the manuscript. T.L.: pathology review and providing histopathology photos. M.B.: cardiology review and final approval. R.G.: haematology review and final approval. All authors were directly involved in the clinical care of the patient.

The data that support the findings of this study are not publicly available due to containing information that could compromise the privacy of the patient but are available from the corresponding author (A.A.) upon reasonable request.