Best Practice & Research Clinical Rheumatology xxx (2016) 1e13

Contents lists available at ScienceDirect

Best Practice & Research Clinical

Rheumatology
journal homepage: www.elsevierhealth.com/berh

Rheumatologic diseases as the cause of fever of

unknown origin
C.M. Mulders-Manders*, A. Simon, C.P. Bleeker-Rovers
Department of Internal Medicine, Radboud University Medical Center, PO Box 9101, 6500 HB Nijmegen,
The Netherlands

a b s t r a c t
Keywords:
Fever of unknown origin In 30% of patients with fever or inflammation of unknown origin
Inflammation of unknown origin (FUO/IUO), the cause is eventually found to be a rheumatologic
Autoimmune disease disease such as autoimmune or granulomatous disease or vasculitis.
Rheumatology Most of these patients suffer from an uncommon presentation of a
FDGePET common disease, instead of an uncommon disease. We demon-
strate the diagnostic challenge with several cases. The workup of
FUO is based on the identification of potential diagnostic clues
(PDCs). In the absence of PDCs, a standardized diagnostic protocol
should be followed, including early FDGePET/CT. Other imaging
techniques or invasive diagnostic techniques should be reserved for
those in whom PDCs are present.
© 2016 Elsevier Ltd. All rights reserved.

Introduction

Although approximately 30% of patients with long-term unexplained fever (fever of unknown
origin (FUO), see Table 1 for definition) are eventually diagnosed with rheumatologic diseases such as
autoimmune disease, vasculitis, or granulomatous disease [1], these kinds of diseases are often not the
first considered in the differential diagnosis by physicians.
Other patients present with long-term unexplained inflammation (inflammation of unknown
origin (IUO), see Table 2 for definition) in the absence of fever. Underlying causes, workup, and
prognosis are the same for FUO and IUO [2].
It should be kept in mind that most patients with FUO suffer from an uncommon presentation of a
common disease instead of an uncommon disease. This makes FUO a diagnostic challenge. We present

* Corresponding author. Tel.: þ31 24 36 81 819; fax: þ31 24 36 16 519.

E-mail address: Karin.Mulders-Manders@radboudumc.nl (C.M. Mulders-Manders).

http://dx.doi.org/10.1016/j.berh.2016.10.005
1521-6942/© 2016 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Mulders-Manders CM, et al., Rheumatologic diseases as the cause of
fever of unknown origin, Best Practice & Research Clinical Rheumatology (2016), http://dx.doi.org/
10.1016/j.berh.2016.10.005
2 C.M. Mulders-Manders et al. / Best Practice & Research Clinical Rheumatology xxx (2016) 1e13

Table 1
Definition of fever of unknown origin (FUO).

1. Fever 38.3  C (101  F) on 3 occasions

2. Duration of illness 3 weeks
3. No diagnosis despite extensive evaluation including (but not limited to):
- Extensive medical history and physical examination.
- Laboratory: erythrocyte sedimentation rate (ESR) or C-reactive protein (CRP), hemoglobin, platelet count, leukocyte count
including differentiation, sodium, potassium, calcium, creatinine, total serum protein in protein electrophoresis, alkaline
phosphatase (AF), aspartate aminotransferase (ASAT), alanine aminotransferase (ALAT), lactate dehydrogenase (LDH),
creatine kinase (CK), antinuclear antibodies (ANAs), rheumatoid factor (RF), microscopic urinalysis, ferritin.
- Microbiology: blood cultures (minimal 3), urine culture, tuberculin skin test, or interferon-gamma release assay.
- Imaging: chest X-ray, abdominal ultrasound, or chest and abdominal CT scans.
4. No known immunocompromised state:
- Neutropenia (leukocyte count <1.0  109/L and/or granulocyte count <0.5  109/L during at least 1 week within the 3
months before the start of the fever.
- Known human immunodeficiency virus (HIV) infection.
- Known hypogammaglobulinemia (IgG <50% of normal value).
- Use of 10 mg prednisone or equivalent dose of steroids during at least 2 weeks in the 3 months before the start of the fever.

Table 2
Definition of inflammation of unknown origin (IUO).

1. Elevated erythrocyte sedimentation rate (ESR), C-reactive protein (CRP) or serum amyloid A (SAA) on  3 separate
occasions
2. Duration of illness 3 weeks
3. No diagnosis despite:
-Extensive medical history taking and physical examination.
-Laboratory: hemoglobin, platelet count, leukocyte count including differentiation, sodium, potassium, calcium, creatinine,
total serum protein in protein electrophoresis, alkaline phosphatase (AF), aspartate aminotransferase (ASAT), alanine
aminotransferase (ALAT), lactate dehydrogenase (LDH), creatine kinase (CK), antinuclear antibodies (ANAs), rheumatoid
factor (RF), microscopic urinalysis, ferritin.
-Microbiology: blood cultures (minimal 3), urine culture, tuberculin skin test, or interferon-gamma release assay.
-Imaging: chest X-ray, abdominal ultrasound, or chest and abdominal CT scans.
4. No known immunocompromised state:
-Neutropenia (leukocyte count <1.0  109/L and/or granulocyte count <0.5  109/L during at least 1 week within the 3
months before the start of the fever.
-Known human immunodeficiency virus (HIV) infection.
-Known hypogammaglobulinemia (IgG <50% of normal value).
-Use of 10 mg prednisone or equivalent dose of steroids during at least 2 weeks in the 3 months before the start of the fever.

a case series of four patients with FUO because of a rheumatologic disease who were treated at our
university expertise FUO center in the past years.

Case 1. Polymyalgia rheumatica

A 70-year-old woman was referred to our outpatient department after three previous admissions in
a Dutch community hospital because of episodes of fever that had been present for 5 months and were
accompanied by rigors, night sweats, 17 kg weight loss, fatigue, chest and abdominal pain, myalgia,
arthralgia, muscle weakness, and arthritis. Laboratory evaluation showed microcytic anemia, elevated
C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR), and elevated liver enzymes. Anti-
nuclear antibodies (ANAs) were positive at multiple occasions, with repeatedly negative extractable
nuclear antibodies (ENAs) and anti-double stranded DNA (anti-dsDNA) antibodies. Antineutrophil
cytoplasmic antibodies (ANCAs) were negative. Extensive microbiological investigation, including 19
blood cultures and serology for different pathogens, was negative. PET/CT scan had shown high FDG-
uptake in the right hip compatible with large-joint arthritis. A culture of the synovial fluid of the hip
had remained sterile. Bone marrow biopsy had shown anemia of chronic disease. She had been treated
with multiple courses of broad-spectrum antibiotics, which had resulted in normalization of body
temperature. However, fever and inflammation reappeared quickly after each course.

Please cite this article in press as: Mulders-Manders CM, et al., Rheumatologic diseases as the cause of
fever of unknown origin, Best Practice & Research Clinical Rheumatology (2016), http://dx.doi.org/
10.1016/j.berh.2016.10.005
 C.M. Mulders-Manders et al. / Best Practice & Research Clinical Rheumatology xxx (2016) 1e13 3

At physical examination, we found a maximum body temperature of 39.2  C, oral aphthous ulcers;
limb-girdle pattern muscle weakness; and frank arthritis of the knees, shoulders, hip, and elbow.
Microbiological evaluation also did not show any signs of infection. A second PET/CT was performed,
showing arthritis of both shoulders and hips with bursitis and enthesopathy, and symmetrically
elevated FDG uptake in the muscles around both knees. Electromyography (EMG) showed myopathy
without myositis or polyneuropathy. Because of these findings, a muscle biopsy was taken from the
femoral muscle. This showed abundant presence of target fibers, which is compatible with polymyalgia
rheumatica. The patient was started on prednisone 30 mg/day. After 3 days, there was a remarkable
improvement in her clinical condition. CRP decreased from 117 to 19 mg/L in 1 week.

Case 2. Behcet's disease

A 28-year-old man of Bosnian origin was referred because of a 13-year history of episodes of fever.
He had previously been evaluated by 10 different medical specialists (internists, rheumatologists, and
ophthalmologists) in seven different community and university hospitals in both Serbia and The
Netherlands. Long febrile episodes of 2 weeks to 3 months duration had occurred in 1997, 1998, 1999,
2007, and 2009. In between these protracted attacks, he had suffered from less severe and shorter
attacks several times a year. Fever could be accompanied by loss of appetite, weight loss, night sweats,
abdominal pain, headache, arthralgia, aphthous ulcers, and pancytopenia, but all these were not
present during all the attacks. From 2005 onwards, he had multiple episodes of anterior uveitis, which
was unrelated to the fever. He also described Raynaud's phenomena of both hands and feet. During the
extensive previous evaluations, infections, autoimmune diseases, porphyria, thalassemia, familial
Mediterranean fever (FMF), hyperimmunoglobulin D and periodic fever syndrome (HIDS), and TNF-
receptor-associated periodic syndrome (TRAPS) had all been ruled out. Trials with multiple drugs,
including nonsteroidal anti-inflammatory drugs (NSAIDs), colchicine, prednisone, and anti-TNF (eta-
nercept) had been ineffective.
Short-term treatment with the interleukin-1 receptor antagonist anakinra during an episode of
fever had been given twice and had resulted in rapid reduction of inflammation both times.
At the time of our evaluation, his anti-inflammatory medication consisted of naproxen 500 mg once
daily. He had no fever, but oral aphthous ulcers and hepatomegaly were present on physical exami-
nation. A month later, he developed a recognizable fever attack accompanied by malaise, arthralgia,
arthritis of the fingers, pyrosis, abdominal pain, and anterior uveitis. PET/CT during the fever attack
showed increased FDG-uptake in a single axillary lymph node, several small mediastinal lymph nodes,
and a single para-aortic lymph node. Furthermore, there was increased uptake in the ligaments around
the right knee. HLA-B51 was positive. We diagnosed this patient with Behcet's disease. Criteria for
diagnosis were the presence of oral aphthous ulcers, anterior uveitis, and HLA-B51 positivity. He was
started on continuous treatment with anakinra 100 mg daily. Despite this treatment, he developed a
new fever attack with arthralgia, myalgia, and erythema nodosum, and we started colchicine 0.5 mg
twice daily. Anakinra and NSAIDs were continued because previous trials with colchicine monotherapy
had been unsuccessful. This combination effectively eliminated fever and inflammation, but arthralgia
remained.

Case 3. Systemic lupus erythematosus (SLE)

A 75-year-old man was referred after he had been admitted in two other hospitals because of fever;
night sweats; diarrhea; hypoalbuminemia (serum albumin 19 g/L); and edema, which had been pre-
sent for 4 months. His medical history consisted of a basal cell carcinoma located at the shoulder, deep
venous thrombosis, and cataract. In the past, he had visited Africa and South America multiple times.
He had been evaluated because of these problems by multiple internists, a dermatologist, otolaryn-
gologist, dental surgeon, cardiologist, and a neurologist. Besides other imaging, two PET/CT scans had
been performed, which showed high FDG-uptake in the left maxillary sinus. Laboratory investigation
showed polyclonal light chain production, low complement levels, and type 3 cryoglobulinemia. ANAs,
ENAs, and anti-dsDNA had all tested negative. Infections, including multiple types of tropical in-
fections, malignancy, nephrotic syndrome, protein losing enteropathy, and amyloidosis had been

Please cite this article in press as: Mulders-Manders CM, et al., Rheumatologic diseases as the cause of
fever of unknown origin, Best Practice & Research Clinical Rheumatology (2016), http://dx.doi.org/
10.1016/j.berh.2016.10.005
4 C.M. Mulders-Manders et al. / Best Practice & Research Clinical Rheumatology xxx (2016) 1e13

excluded. He had been treated for Strongyloides infection in the absence of microbiological proof, but
this had not led to resolution of the fever. Under suspicion of a dental infection, all upper teeth had
been extracted, again without any effect on the inflammation.
At the time of our first evaluation, the fever and night sweats had subsided, but fatigue, weight loss,
and diarrhea remained. He suffered from dryness of the mouth and decreased sensation of the feet. On
physical examination, we only found small erythematous lesions on his back, which had previously
been biopsied. CRP was normal, but ESR was 126 mm/h. There was a mild leukopenia with neutro- and
lymphopenia.
ANA was repeated and was positive with a homogeneous pattern this time, while ENA remained
negative. Complement C3 and C4 were low. We consulted a neurologist, who diagnosed symmetric
sensorimotor axonal polyneuropathy of the lower legs. PET/CT again showed high FDG-uptake in the
left maxillary sinus with increased thickness of the mucosa, several slightly active hilar lymph nodes,
elevated uptake in the sigmoid, and hardly increased uptake in a subpleural lesion in the left upper
lobe. Because of the persistently increased uptake in the sinus, an otolaryngologist was consulted, who
diagnosed minimal sinusitis, which was considered not to be the cause of the fever.
We diagnosed this patient as having SLE based on four positive Systemic Lupus International
Collaborating Clinics (SLICC) criteria: polyneuropathy, leukopenia, positive ANA, and low complement
levels. He was referred back to the referring hospital for further treatment, where he was given
hydroxychloroquine.

Case 4. Eosinophilic granulomatosis with polyangiitis

A 45-year-old male roadworker was referred because of fever, accompanied by weight loss,
arthralgia, arthritis, headache, dyspnea, and cough. His symptoms had started after surgical drainage of
an abscess of the infrapatellar bursa, a complication of bursitis attributed to his job. Because of these
conditions, he had been evaluated for 5 months in two different hospitals by multiple specialists:
rheumatologists, a dermatologist, and a pulmonologist. Laboratory evaluation showed elevated levels
of ESR and CRP, and eosinophilia (eosinophils maximal 5.3  109/L). At the start of the complaints, a
chest X-ray had been made, showing bilateral hilar lymphadenopathy. Over the following months,
multiple chest CTs had been performed, all showing mediastinal, hilar, and axillar lymphadenopathy.
These lymph nodes were proven to be FDGePET positive. Bronchoalveolar lavage had shown many
leukocytes with elevated eosinophils (8%). A lymph node had been biopsied by endobronchial ultra-
sound, but pathological examination revealed only reactive changes. Other lymph nodes had been
removed by mediastinoscopy, which again showed only reactive changes. Because of suspected
sarcoidosis, a cardiac magnetic resonance imaging (MRI) had been performed, which did not show any
signs of cardiac sarcoidosis. Extensive microbiological evaluation including cultures of blood, lymph
nodes, bronchoalveolar lavage fluid and synovial fluid, many serological tests and PCR tests for
tuberculosis on lymph nodes and bronchoalveolar fluid had all remained negative. Bone marrow bi-
opsy had shown remarkable eosinophilia of 18%. A kidney biopsy revealed no abnormalities.
In the absence of a diagnosis, the patient had been treated with a 3-month course of high-dose
steroids under suspicion of sarcoidosis. This treatment was very effective: besides remission of fever
and other symptoms, the lymphadenopathy also decreased on repeated chest CT scans. After cessation
of treatment, symptoms, biochemical inflammation, and lymphadenopathy reappeared.
Two months after cessation of the steroids, we examined this patient for the first time. At that time,
there were no complaints, and physical examination was normal, as were ESR, CRP, and eosinophil
count. A month later, his arthralgia increased and according to the patient's request, we treated him
with 20 mg prednisone again. This again effectively reduced fever and arthralgia. Prednisone was
quickly tapered and 2 months later, while on prednisone 2.5 mg daily, he again developed fever;
arthralgia; and arthritis of ankles, wrist, and elbows, accompanied by eosinophilia (eosinophils
1.37  109 per liter). Because of this, we stopped the steroids and admitted the patient to our internal
medicine ward. A new bone marrow biopsy was performed, which showed hypercellular bone marrow
with marked eosinophilia and activated T cells. A new FDGePET/CT showed multiple FDG-positive
mediastinal, hilar, and axillary lymph nodes; increased uptake in both upper pulmonary lobes and
around the elbows, wrists, hips, and ankles; and in the left foot. Chest CT showed new miliary

Please cite this article in press as: Mulders-Manders CM, et al., Rheumatologic diseases as the cause of
fever of unknown origin, Best Practice & Research Clinical Rheumatology (2016), http://dx.doi.org/
10.1016/j.berh.2016.10.005
 C.M. Mulders-Manders et al. / Best Practice & Research Clinical Rheumatology xxx (2016) 1e13 5

abnormalities in both lungs. His symptoms spontaneously remitted and the patient was discharged
after 7 days. However, his condition deteriorated after discharge and he developed severe dyspnea.
Because of this and the findings on the latest chest CT, a video-assisted thoracoscopy (VATS) with
wedge excision of the left lung was performed. Pathologic evaluation showed granulomatous
inflammation with central collagenous necrosis. On the basis of this finding, in combination with the
marked eosinophilia, we diagnosed this patient with eosinophilic granulomatosis with polyangiitis. He
was started on prednisone 60 mg again, which again was immediately effective and could be tapered
slowly.

The workup of FUO

The diagnostic process for FUO starts with detailed medical history taking. This history should
include information on medical history, medication use (both prescription and nonprescription drugs;
this is important as almost all drugs can cause drug fever), travel history, animal contacts, occupation,
exposure to, for instance, toxins or recreational drugs; and a sexual history. This is followed by a
physical examination with specific attention to mucous membranes; eyes; temporal arteries; cervical,
supraclavicular, axillar and inguinal lymph nodes; liver; spleen; and skin. Any abnormalities identified
at this stage are called potential diagnostic clues (PDCs) [3,4]. It is rare for patients not to show any PDCs
at this stage, but misleading PDCs (i.e. PDCs not leading to a diagnosis) are very common [3,4]. All anti-
inflammatory drugs should be stopped at this stage, as they may mask PDCs.
The choice of further diagnostic tests at this stage is guided by the PDCs present and by local disease
prevalence. When no PDCs are identified or all PDCs have proven to be misleading after further
investigation, a standardized diagnostic protocol should be followed (Fig. 1). Fundoscopy and mea-
surement of serum cryoglobulin should be performed early, as localizing symptoms may often be
absent in vasculitis or cryoglobulinemia. When these tests are negative, the following step is whole
body FGDePET scan, including diagnostic CT scan if available. FDGePET/CT has a very high diagnostic
yield in FUO [5,6]. When this does not lead to a diagnosis, it should be followed by temporal artery
biopsy in elderly patients, because due to high FDG uptake in the brain, the specificity of FDGePET for
temporal arteritis is not high enough to exclude this diagnosis based on negative FDGePET alone. Chest
and abdominal CT scans may identify FDGePET-negative lesions, and thus should be performed at this
stage if not previously performed.
When no diagnosis is made despite all these investigations, a period of watchful waiting starts.
History and physical examination should be repeated periodically to identify any new PDCs. In stable
patients, NSAIDs or paracetamol can be used to lower body temperature. Empirical treatment with
antibiotics, corticosteroids, or other anti-inflammatory drugs should be reserved only for those pa-
tients who show clinical deterioration during this follow-up period, as they may mask new symptoms
indicating a diagnosis. The rapid effects of corticosteroids in some rheumatologic diseases are illus-
trated by cases 1 and 4 described above.
There are no PDCs that specifically indicate rheumatologic diseases. However, these diseases may be
more frequent in elderly patients (65 years of age) [7e9] and in patients with arthritis, skin rash, or
lymphadenopathy on physical examination [10,11].

FUO in rheumatologic diseases

The causes of FUO are divided into four main groups: infection, malignancy, noninfectious in-
flammatory disease (NIID), and miscellaneous causes. Rheumatologic diseases will be grouped under
the category NIID, which includes autoimmune, autoinflammatory and granulomatous diseases, and
vasculitides. The distribution of causes of FUO depends on the geographical region of the cohort, with
Western studies reporting less infections and higher numbers of undiagnosed cases than non-Western
studies. However, the percentage of NIID in the most recent Western studies [2,4,7,11e19] is roughly
equal to that of the largest non-Western studies in the past 10 years [8e10,20e28], with approximately
30% of all patients with FUO being diagnosed with NIID. This does not reflect the true percentage of
rheumatologic diseases in these studies for several reasons. First, besides true rheumatologic diseases,
other diseases such as autoimmune hepatitis or thyroiditis may also be classified as NIID. Second, not

Please cite this article in press as: Mulders-Manders CM, et al., Rheumatologic diseases as the cause of
fever of unknown origin, Best Practice & Research Clinical Rheumatology (2016), http://dx.doi.org/
10.1016/j.berh.2016.10.005
6 C.M. Mulders-Manders et al. / Best Practice & Research Clinical Rheumatology xxx (2016) 1e13

Fever ≥38.3oC (101 oF) and illness ≥3 weeks and

no known immunocompromised state

History and physical examinaon

Stop anbioc treatment and corcosteroids

Obligatory invesgaons:
ESR or CRP, hemoglobin, platelet count, leukocyte count and differenaon, electrolytes,
creanine, total protein, protein electrophoresis, alkaline phosphatase, ASAT, ALAT, LDH,
creane kinase, annuclear anbodies, rheumatoid factor, urinalysis, 3 blood cultures, urine
culture, chest X-ray, abdominal ultrasonography and tuberculin skin test or interferon
gamma release assay

Exclude manipulaon with thermometer

Stop or replace medicaon to exclude drug fever

PDCs present PDCs absent or misleading

Guided diagnosc tests Cryoglobulin and fundoscopy

FDG-PET/CT
Diagnosis No diagnosis (or labeled leykocyte scingraphy or galliumscingraphy)

Scingraphy abnormal Scingraphy normal

Confirmaon of abnormality Repeat history and physical examinaon

(e.g. biopsy, culture) PDC driven invasive tesng

Diagnosis No diagnosis Diagnosis No diagnosis

Chest and abdominal CT

Temporal artery biopsy (≥55 years)

Diagnosis No diagnosis

Stable condion: Deterioraon:

follow up for new PDCs Further diagnosc tests
Consider NSAID Consider therapeuc trial

Fig. 1. Flowchart of the diagnostic workup of fever of unknown origin. PDC ¼ potential diagnostic clue.

all NIIDs with (frequent) musculoskeletal disorders are consistently reported in the NIID category.
Instead, they may be placed in the miscellaneous category in some studies. This has occurred for
instance with FMF [2,10,18], Schnitzler syndrome [2], sarcoidosis [7,16], (pseudo)gout [10,12], temporal
arteritis [24], Behcet's disease [26], and reactive arthritis [22]. Lastly, it may be impossible to diagnose
the exact underlying rheumatologic disease in FUO patients, as many of these diseases can only be
diagnosed on the basis of a standard set of criteria without a golden standard test. This includes, among

Please cite this article in press as: Mulders-Manders CM, et al., Rheumatologic diseases as the cause of
fever of unknown origin, Best Practice & Research Clinical Rheumatology (2016), http://dx.doi.org/
10.1016/j.berh.2016.10.005
 C.M. Mulders-Manders et al. / Best Practice & Research Clinical Rheumatology xxx (2016) 1e13 7

others, unclassified vasculitides; autoimmune diseases; and autoinflammatory syndromes without a

known genetic defect. Patients with signs and symptoms fitting these diseases, but not fulfilling all
required criteria will eventually be categorized in the no diagnosis category. It is not possible to esti-
mate the additional number of patients with a rheumatologic disease in this category, as there are no
FUO studies that report suspected diagnoses in the no diagnosis category. The number of additional
patients with a rheumatologic diseases in this category may be substantial, as up to 53% of patients in
FUO cohorts remain undiagnosed despite extensive investigation (Table 3).
The most frequently diagnosed rheumatologic diseases in patients with FUO from Western pop-
ulations are adult-onset Still's disease (AOSD), large-vessel vasculitis (including temporal arteritis), SLE,
and sarcoidosis, which are respectively diagnosed in 5%, 5%, 3%, and 2% of all cases. In non-Western
populations, AOSD is the most frequent rheumatologic disease in FUO, making up 5% of all FUO
cases and being followed by unclassified vasculitis, polymyalgia rheumatica, and SLE, which are all
diagnosed in 2% of all cases. Figs. 2e5 show the distribution of rheumatologic diseases diagnosed in
patients with FUO in the most recent and largest studies. Sarcoidosis is not included in these tables, but
is diagnosed in 1e3% of patients in both Western and non-Western cohorts.

How to diagnose rheumatologic diseases in patients with FUO

Role of autoantibodies in FUO

As it is clear from Tables 1 and 2, ANAs and rheumatoid factor (RF) should be measured in all pa-
tients with long-term fever or inflammation. ANCAs should be measured in patients with PDCs
pointing toward vasculitis. The presence of autoantibodies in patients with FUO varies between
studies. Table 4 shows the presence of autoantibodies in patients with FUO.
Although quite common and more frequently present in patients ultimately diagnosed with NIIDs
[11], the presence of autoantibodies is helpful in a diagnosis in 2e12% of patients [13,25]. Positive ANAs

Table 3
Patients with no diagnosis in the largest recent non-Western and the most recent non-Western FUO cohorts 100 patients.

Study Patients Undiagnosed (%) Study Patients Undiagnosed (%)

Non-Western Western
Hu 2008 142 20 (14%) De Kleijn 1997 167 50 (30%)
China The Netherlands
Kucukardali 2008 154 24 (16%) Vanderschueren 2003 290 154 (53%)
Turkey Belgium
Bandyopadhyay 2011 164 20 (12%) Zenone 2006 144 37 (26%)
India France
Mete 2012 100 20 (20%) Bleeker-Rovers 2007 73 37 (51%)
Turkey The Netherlands
Ma 2012 397 39 (10%) Mansueto 2008 91 29 (32%)
China Italy
Ryuko 2013 174 33 (19%) Vanderschueren 2009 114 46 (40%)
Japan Belgium
Mahmood 2013 205 24 (12%) Efstathiou 2010 200 39 (20%)
Pakistan Greece
Naito 2013 121 28 (23%) Pedersen 2012 52 21 (40%)
Japan Denmark
Alavi 2013 106 16 (15%) Robine 2014 103 52 (51%)
Iran France
Yamanouchi 2014 256 74 (29%) Vanderschueren 2014 436 168 (39%)
Japan Belgium
Yu 2014 107 23 (22%) Popovska 2016 74 3 (4%)
China Serbia
Montasser 2015 374 29 (8%)
Egypt

Total 2052 350 (17%) 1756 636 (36%)

Please cite this article in press as: Mulders-Manders CM, et al., Rheumatologic diseases as the cause of
fever of unknown origin, Best Practice & Research Clinical Rheumatology (2016), http://dx.doi.org/
10.1016/j.berh.2016.10.005
8 C.M. Mulders-Manders et al. / Best Practice & Research Clinical Rheumatology xxx (2016) 1e13

Fig. 2. Prevalence of autoimmune diseases in Western and non-Western FUO cohorts 100 patients and published 2006 and later.
SLE ¼ systemic lupus erythematosus; RA ¼ rheumatoid arthritis; MCTD ¼ mixed connective tissue disease; UDC ¼ undifferentiated
connective tissue disease.

are helpful in 3e20% of all diagnoses, and most of these patients will be diagnosed with SLE [3,4,10].
ANCAs are found to be helpful in the diagnosis of microscopic polyangiitis or polyarteritis nodosa in
2e3% of all FUO cases [3,4,18]. RF may contribute to a diagnosis in 0.5e2% of patients [3,10].
Autoantibodies may also be false negative in FUO. A single study reported that in 4 of 39 FUO pa-
tients with an autoimmune disease, autoantibodies were not present [25]. As is illustrated in case 3
described above, repeated measurement of ANAs, ENAs, and anti-dsDNA may be helpful [29], especially
when other PDCs for autoimmune disease are present.

Role of imaging in diagnosing rheumatologic diseases in patients with FUO

No single imaging technique that can diagnose or rule out all rheumatologic diseases in FUO has
been identified thus far. As outlined in the section on the investigation of FUO, the choice of additional
imaging techniques should be guided by the presence of PDCs. Except for FDGePET and chest and
abdominal CT, there is no place for random imaging in the workup of FUO.
The first-line imaging techniques in the workup of FUO, chest X-ray and abdominal ultrasound, will
by definition not directly lead to the final diagnosis, but they may contribute to the diagnosis of
rheumatologic diseases in 8% of patients with NIIDs and FUO [25].
Abnormalities on abdominal ultrasound may be present for instance in sarcoidosis or SLE [3,25].
CT scans may be helpful in the diagnosis of rheumatologic diseases in these patients. Chest CT scans
can lead to the diagnosis in 28% and abdominal CT scans in 15% [25]. Chest CT scans may for instance
show lung fibrosis in patients with autoimmune diseases [3] or interstitial abnormalities in patients

Please cite this article in press as: Mulders-Manders CM, et al., Rheumatologic diseases as the cause of
fever of unknown origin, Best Practice & Research Clinical Rheumatology (2016), http://dx.doi.org/
10.1016/j.berh.2016.10.005
 C.M. Mulders-Manders et al. / Best Practice & Research Clinical Rheumatology xxx (2016) 1e13 9

Fig. 3. Prevalence of vasculitis in Western and non-Western FUO cohorts 100 patients and published 2006 and later.
RA ¼ rheumatoid arthritis; RS3PE ¼ Remitting seronegative symmetrical synovitis with pitting edema.

with pulmonary vasculitis, as was seen in case 4. Abdominal CT was reported to contribute to the
diagnosis of rheumatoid arthritis, showing vasculitis/pericarditis [3], which can be extra-articular
symptoms of this disease. Cerebral CT may show abnormalities in patients with vasculitis leading to
cerebral infarction [10].
Transthoracic or transesophageal echocardiography will mostly be performed to rule out endo-
carditis in patients with FUO, but it may also contribute to the diagnosis of NIID, for instance when
pericarditis is present [3]. However, all patients in whom transthoracic echocardiography contributed
to the diagnosis, showed other PDCs for the underlying disease in this study [3]. Therefore, echocar-
diography should only be performed when PDCs for endocarditis, pericardial effusion, or pericarditis
are present.

Role of FGDePET/CT in diagnosing rheumatologic diseases in patients with FUO

FDGePET/CT has a high-diagnostic yield in FUO, because it can show any inflammation irrespective
of the cause. Any inflammation due to rheumatologic disease may lead to positive FDGePET/CT.
Sensitivity of FDGePET/CT in the diagnosis of NIID is 65%, which is lower than for the diagnosis of
malignancy or infection [30]. Arthritis may be shown as increased uptake in any joint and may be
present in many rheumatologic diseases. PET-positive foci in the lungs may be found in patients with
vasculitis [17,18] or sarcoidosis [18]. In patients with large-vessel vasculitis, FDGePET/CT scan with
increased large-vessel uptake may immediately reveal the diagnosis. Early initiation of corticosteroids
may confound FDGePET results. This was reported by Robine et al., who found an abnormal abdominal
CT scan in a patient with giant cell arteritis with aortitis but a negative PET during corticosteroid

Please cite this article in press as: Mulders-Manders CM, et al., Rheumatologic diseases as the cause of
fever of unknown origin, Best Practice & Research Clinical Rheumatology (2016), http://dx.doi.org/
10.1016/j.berh.2016.10.005
10 C.M. Mulders-Manders et al. / Best Practice & Research Clinical Rheumatology xxx (2016) 1e13

Fig. 4. Prevalence of autoinflammatory diseases in Western and non-Western FUO cohorts 100 patients and published 2006 and
later. AOSD ¼ adult-onset Still's disease; FMF ¼ familial Mediterranean fever; TRAPS ¼ TNF-receptor-associated periodic syndrome;
HIDS ¼ hyperimmunoglobulin D and periodic fever syndrome.

therapy [18]. Because of this effect, FDGePET/CT should not be performed during therapy with
corticosteroids.
FDGePET may also be false negative in NIID. Bleeker-Rovers et al. reported 2 patients with limited
pleural effusions that were eventually diagnosed with SLE, but had negative PET scans [4].

Role of invasive diagnostic techniques in diagnosing rheumatologic diseases in patients with FUO

At any stage of the workup, PDCs may necessitate invasive diagnostic techniques to confirm or
exclude a diagnosis. As a general rule, there is no indication for the use of invasive techniques in the
absence of PDCs except for temporal artery biopsy in elderly patients. As described earlier, this should
be performed at an early stage in patients >55 years in the absence of PDCs and in all patients with
PDCs for temporal arteritis. Biopsies in younger patients in the absence of PDCs for temporal arteritis
are not helpful [18]. Bleeker-Rovers reported temporal artery biopsy being performed in 10 patients
with PDCs for vasculitis, two patients with vasculitis on FGDePET, and two patients without any ev-
idence for vasculitis; the only patient with a positive biopsy was one of the two patients in whom
FDGePET already showed vasculitis [4].
Bone marrow biopsy may be helpful in some rheumatologic diseases presenting as FUO, for instance
granulomatous diseases or AOSD, which may be accompanied by hemophagocytic syndrome [14].
There is no indication for bone marrow biopsy when no PDCs for bone marrow abnormalities (anemia,
leukopenia, and thrombocytopenia) are present.

Please cite this article in press as: Mulders-Manders CM, et al., Rheumatologic diseases as the cause of
fever of unknown origin, Best Practice & Research Clinical Rheumatology (2016), http://dx.doi.org/
10.1016/j.berh.2016.10.005
 C.M. Mulders-Manders et al. / Best Practice & Research Clinical Rheumatology xxx (2016) 1e13 11

Fig. 5. Prevalence of other rheumatologic diseases in Western and non-Western FUO cohorts 100 patients and published 2006 and
later.

Table 4
Frequencies of positive autoimmune antibodies in patients with FUO (%).

Study De Kleijn 1997 Bleeker 2007 Kucukardali 2008 Efstathiou 2010

Netherlands Netherlands Turkey Greece

ANAs 14 16 22 34
ENAs 1 11
Anti-dsDNA 2 2
RF 2 4 38
ANCAs 6 8 11

ANAs ¼ antinuclear antibodies; ENAs ¼ extractable nuclear antibodies; anti-dsDNA ¼ anti-double strand DNA antibodies;
RF ¼ rheumatoid factor; ANCAs ¼ antineutrophilic cytoplasmic antibodies.

When skin abnormalities are present, a skin biopsy can contribute to the diagnosis. It was reported
to be helpful in the diagnosis of vasculitis [3,25,29], HenocheSchonlein purpura, SLE, AOSD [4], der-
matomyositis [10], and Behcet's disease [29].
Renal biopsies can be helpful in the presence of an abnormal urinary sediment for the diagnosis of
granulomatosis with polyangiitis, SLE, vasculitis, and cryoglobulinemia [3,25].
When arthritis is present, joint aspiration may prove gout or pseudogout [3].
Liver biopsy may be performed when PDCs for liver disease, such as elevated liver enzymes, are
present. De Kleijn and Kucukardali both diagnosed two patients with granulomatous hepatitis on the
basis of a biopsy [3,10]. Even in the presence of PDCs for liver disease, liver biopsy may be noncon-
tributory in 90% of patients [3,10].

Please cite this article in press as: Mulders-Manders CM, et al., Rheumatologic diseases as the cause of
fever of unknown origin, Best Practice & Research Clinical Rheumatology (2016), http://dx.doi.org/
10.1016/j.berh.2016.10.005
12 C.M. Mulders-Manders et al. / Best Practice & Research Clinical Rheumatology xxx (2016) 1e13

Other invasive diagnostic techniques that contributed to the diagnosis of rheumatologic diseases in
the workup for FUO were pulmonary wedge excision ([4] and case 4 presented here) and laparotomy or
laparoscopy [3,4,31], aiding the diagnosis of SLE and granulomatous diseases. A history of laparoscopy
or laparotomy for suspected appendicitis in which a noninflamed appendix was found in a patient with
clinical signs of peritonitis, should raise the suspicion of an autoinflammatory disease, such as FMF.

Conclusion

Approximately 30% of patients with FUO/IUO will eventually be diagnosed with a rheumatologic
disease. The true number of rheumatologic diseases in these patients may be even higher, as many of
these diseases can only be diagnosed based on criteria; patients that do not fulfill all criteria will be
categorized in the no diagnosis category in the literature. The presentation of rheumatologic diseases in
this group of patients is mostly noncharacteristic, leading to a diagnostic challenge. PDCs should guide
investigations and a standardized diagnostic protocol including early stage FDGePET/CT should be
followed in cases in which PDCs are absent or misleading. Immunologic serology such as ANA, ENA,
anti-dsDNA, and ANCA will often be false positive, but may also be false negative and repeated mea-
surements may be helpful. There is no indication for random imaging or invasive diagnostic tech-
niques, but these may be helpful in the diagnosis of rheumatologic diseases when PDCs are present.
Treatment with antiinflammatory drugs should be postponed as much as possible in patients in whom
no diagnosis has been made, as it may mask important PDCs.

Conflict of interest statement

None of the authors report a conflict of interest.

References

[1] Mulders-Manders C, Simon A, Bleeker-Rovers C. Fever of unknown origin. Clin Med 2015;15(3):280e4.
*[2] Vanderschueren S, Del Biondo E, Ruttens D, et al. Inflammation of unknown origin versus fever of unknown origin: two of
a kind. Eur J Intern Med 2009;20(4):415e8.
*[3] de Kleijn EM, van Lier HJ, van der Meer JW. Fever of unknown origin (FUO). II. Diagnostic procedures in a prospective
multicenter study of 167 patients. The Netherlands FUO Study Group. Medicine 1997;76(6):401e14.
*[4] Bleeker-Rovers CP, Vos FJ, de Kleijn EM, et al. A prospective multicenter study on fever of unknown origin: the yield of a
structured diagnostic protocol. Medicine 2007;86(1):26e38.
*[5] Bleeker-Rovers CP, Vos FJ, Mudde AH, et al. A prospective multi-centre study of the value of FDG-PET as part of a
structured diagnostic protocol in patients with fever of unknown origin. Eur J Nucl Med Mol Imaging 2007;34(5):
694e703.
*[6] Kouijzer IJ, Bleeker-Rovers CP, Oyen WJ. FDG-PET in fever of unknown origin. Seminars Nucl Med 2013;43(5):333e9.
[7] Popovska-Jovicic B, Canovic P, Gajovic O, et al. Fever of unknown origin: most frequent causes in adults patients. Voj-
nosanit Pregl 2016;73(1):21e5.
[8] Yu KK, Chen SS, Ling QX, et al. Fever of unknown origin: report of 107 cases in a university hospital. Int J Clin Exp Med
2014;7(12):5862e6.
[9] Yamanouchi M, Uehara Y, Yokokawa H, et al. Analysis of 256 cases of classic fever of unknown origin. Intern Med 2014;
53(21):2471e5.
[10] Kucukardali Y, Oncul O, Cavuslu S, et al. The spectrum of diseases causing fever of unknown origin in Turkey: a multi-
center study. Int J Infect Dis IJID Off Publ Int Soc Infect Dis 2008;12(1):71e9.
[11] Efstathiou SP, Pefanis AV, Tsiakou AG, et al. Fever of unknown origin: discrimination between infectious and non-
infectious causes. Eur J Intern Med 2010;21(2):137e43.
*[12] de Kleijn EM, Vandenbroucke JP, van der Meer JW. Fever of unknown origin (FUO). I A. prospective multicenter study of
167 patients with FUO, using fixed epidemiologic entry criteria. The Netherlands FUO Study Group. Medicine 1997;76(6):
392e400.
[13] Vanderschueren S, Knockaert D, Adriaenssens T, et al. From prolonged febrile illness to fever of unknown origin: the
challenge continues. Archives Intern Med 2003;163(9):1033e41.
*[14] Hot A, Jaisson I, Girard C, et al. Yield of bone marrow examination in diagnosing the source of fever of unknown origin.
Archives Intern Med 2009;169(21):2018e23.
[15] Zenone T. Fever of unknown origin in adults: evaluation of 144 cases in a non-university hospital. Scand J Infect Dis 2006;
38(8):632e8.
[16] Mansueto P, Di Lorenzo G, Rizzo M, et al. Fever of unknown origin in a Mediterranean survey from a division of internal
medicine: report of 91 cases during a 12-year-period (1991-2 002). Intern Emerg Med 2008;3(3):219e25.
[17] Pedersen TI, Roed C, Knudsen LS, et al. Fever of unknown origin: a retrospective study of 52 cases with evaluation of the
diagnostic utility of FDG-PET/CT. Scand J Infect Dis 2012;44(1):18e23.

Please cite this article in press as: Mulders-Manders CM, et al., Rheumatologic diseases as the cause of
fever of unknown origin, Best Practice & Research Clinical Rheumatology (2016), http://dx.doi.org/
10.1016/j.berh.2016.10.005
 C.M. Mulders-Manders et al. / Best Practice & Research Clinical Rheumatology xxx (2016) 1e13 13

[18] Robine A, Hot A, Maucort-Boulch D, et al. Fever of unknown origin in the 2000s: evaluation of 103 cases over eleven
years. Presse Med 2014;43(9). e233-40.
*[19] Vanderschueren S, Eyckmans T, De Munter P, et al. Mortality in patients presenting with fever of unknown origin. Acta
Clin Belg 2014;69(1):12e6.
[20] Hu Y, Lu H, Zhang Y, et al. Fever of unknown origin: revisit of 142 cases in a tertiary Chinese hospital. Biosci Trends 2008;
2(1):44e6.
[21] Alavi SM, Nadimi M, Zamani GA. Changing pattern of infectious etiology of fever of unknown origin (FUO) in adult
patients in Ahvaz, Iran. Casp J Intern Med 2013;4(3):722e6.
[22] Bandyopadhyay D, Bandyopadhyay R, Paul R, et al. Etiological study of Fever of unknown origin in patients admitted to
medicine ward of a teaching hospital of eastern India. J Glob Infect Dis 2011;3(4):329e33.
[23] Ma J, Shi X, Zhao S, et al. Fever of unknown origin in an older chinese population. J Am Geriatrics Soc 2012;60(1):169e70.
[24] Mahmood K, Akhtar T, Naeem M, et al. Fever of unknown origin at a teritiary care teaching hospital in Pakistan. Southeast
Asian J Trop Med Public Health 2013;44(3):503e11.
[25] Mete B, Vanli E, Yemisen M, et al. The role of invasive and non-invasive procedures in diagnosing fever of unknown
origin. Int J Med Sci 2012;9(8):682e9.
[26] Montasser MF, Abdelkader NA, Montasser IF, et al. Changing the face of fever of unknown origin in Egypt: a single
hospital study. Braz J Infect Dis Off Publ Braz Soc Infect Dis 2015;19(3):334e5.
[27] Ryuko H, Otsuka F. A comprehensive analysis of 174 febrile patients admitted to Okayama University Hospital. Acta Med
Okayama 2013;67(4):227e37.
*[28] Naito T, Mizooka M, Mitsumoto F, et al. Diagnostic workup for fever of unknown origin: a multicenter collaborative
retrospective study. BMJ open 2013;3(12):e003971.
[29] Saltoglu N, Tasova Y, Midikli D, et al. Fever of unknown origin in Turkey: evaluation of 87 cases during a nine-year-period
of study. J Infect 2004;48(1):81e5.
*[30] Kubota K, Nakamoto Y, Tamaki N, et al. FDG-PET for the diagnosis of fever of unknown origin: a Japanese multi-center
study. Ann Nucl Med 2011;25(5):355e64.
[31] Tabak F, Mert A, Celik AD, et al. Fever of unknown origin in Turkey. Infection 2003;31(6):417e20.

Please cite this article in press as: Mulders-Manders CM, et al., Rheumatologic diseases as the cause of
fever of unknown origin, Best Practice & Research Clinical Rheumatology (2016), http://dx.doi.org/
10.1016/j.berh.2016.10.005