Diagnostic Value of Common Inflammatory Markers on Fever of Unknown Origin

Transcription

1 Jpn. J. Infect. Dis., 69, , 2016 Original Article Diagnostic Value of Common Inflammatory Markers on Fever of Unknown Origin Cui-Ping Liu 1,Zhi-YongLiu 2, Jun-Ping Liu 1, Yi Kang 1, Chong-Shan Mao 1, and Jia Shang 1 * 1 Department of Infectious Diseases, Henan Provincial People's Hospital, People's Hospital of Zhengzhou University, Henan; and 2 Department of Bone and Soft Tumor, Affiliated Cancer Hospital of Zhengzhou University, Henan Cancer Hospital, Henan, China SUMMARY: This study wasdesigned toevaluate the diagnostic value of common inflammatory markers with regard to fever of unknown origin (FUO). We investigated 383 patients who were hospitalized with FUO at the Henan Province People's hospital between January 2009 and June Of all the cases, infectious diseases accounted for 33.9z, neoplasms for 21.1z, collagen vascular diseases for 25.1z, miscellaneous diseases for 4.7z, and no diagnosis for 15.1z. Patients in the neoplasm group were older than those in the infectious disease, collagen vascular disease, and miscellaneous disease groups (p = 0.006, p < , and p = 0.001, respectively). The duration of fever before admission of patients in the neoplasm and collagen vascular disease group was longer than that of patients in the infectious disease group (p = and p = 0.007, respectively). The diagnostic time after admission of patients from the neoplasm and collagen vascular disease groups was longer than that for patients from the infectious disease group (both p < ). Serum ferritin levels of patients in the infectious disease group were lower than those of patients in the neoplasm and collagen vascular disease groups (p = and p = 0.032, respectively), while serum procalcitonin (PCT) levels in the infectious disease group was higher than that in the neoplasm and collagen vascular disease groups (p = and p = 0.007, respectively). Therefore, FUO remains a clinical problem in China and serum ferritin and PCT may be useful in discriminating infectious from non-infectious causes (neoplasms and collagen vascular diseases) in patients with FUO. INTRODUCTION Fever of unknown origin (FUO) was first defined by Petersdorf and Beeson as a body temperature higher than 38.39C, which last morethan 3 weeks, with a negative diagnostic evaluation after 1 week of investigation in the hospital (1). However, with a rise in the proportion of HIV-infected patients in addition to those with neutropenia and increasing opportunity to investigate FUO in an outpatient clinic, the definition was revised later by Durack and Street. They divided FUO into classic FUO, nosocomial FUO, neutropenic FUO, and HIV-associated FUO (2). FUO is a challenging clinical problem worldwide with a variety of etiologies and clinical manifestations. As diagnostic methods and tools are developed, most patients can reach a definite diagnosis. However, it is reported that 7z to 51z of FUO cases remain undiagnosed (3 13). There are more than 200 causes for FUO (14). Infectious diseases, neoplasms, and collagen vascular diseases are the main causes. In the past decade in China, infections were the most prevalent cause and collagen vascular diseases ranked second, followed by malignancy (5,6,15). However, Received September 20, Accepted December 21, J-STAGE Advance Publication January 8, DOI: /yoken.JJID *Corresponding author: Mailing address: Department of Infectious Diseases, Henan Province People's hospital, People's Hospital of Zhengzhou University, No. 7, Weiwu Road, Zhengzhou , Henan, China. Tel: , Fax: , shangjia666@ 126.com These two authors contributed equally to this study. despite the improvement of the diagnostic methods and tools, only a few studies on FUO have been published over the last decade. Furthermore, it remains unknown whether changes with respect to the markers for FUO have occurred in China. Obviously, fever is the common symptom of FUO, resulting from inflammation caused by biological factors, necrotic tissues, and allergy. Inflammatory markers, including white blood cell (WBC) count, C-reactive protein (CRP), erythrocyte sedimentation rates (ESR), serum ferritin, and serum procalcitonin (PCT), can increase to different levels when inflammation happens. Some of these non-specific laboratory tests may help excluding possible causes of FUO and narrowing down the range of diseases. However, a paucity of studies investigated changes in the levels of these inflammatory markers and whether some differences exist in the different etiologies of FUO. Therefore, we conducted a retrospective study to analyze the clinical and epidemiological data from patients with FUO in China in the latest years and to evaluate the diagnostic value of common inflammatory markers on FUO, which may help the diagnosis of FUO. MATERIALS AND METHODS Study population: This study was conducted at the Henan Province People's hospital in China and was approved by the Ethics Committee of the hospital. A total of 1,654 patients were hospitalized with fever as the chief complaint between January 2009 and June The patients were included in this study if they met the criteria of classical FUO defined by Petersdorf and Beeson: (i) body temperature higher than 38.39C, (ii) fever 378

2 Value of Inflammatory Markers on Fever of Unknown Origin lasting more than 3 weeks, (iii) with a negative diagnosis after 1 week of investigation in the hospital. Exclusion criteria were as follows: (i) Patients with neutropenia (neutrophil count < /L), (ii) patients who developed fever after being admitted to the hospital, (iii) patients with HIV. As a result, 383 (23.2z) patients with fever met the criteria and they were included in this study. Because the present study was merely observational and did not demand deviations from standard care, no informed consent was obtained. Data collection: We collected the data from the inpatient services at the Department of Infectious Diseases using the electronic medical record system and hospital paper charts. The epidemiological and clinical data recorded included sex, age, medical history, findings in physical examination, major causes, axillary temperature, duration of fever before admission, diagnostic methods, diagnostic time after admission, and duration of hospitalization. Besides, we collected all auxiliary tests carried out during the course of disease diagnosis, including complete blood count, peripheral blood smear, ESR, CRP, serum ferritin, PCT, biochemical tests (urea, creatinine, alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, bilirubins, albumin, and globulin), urinalysis, tuberculin skin test, 3 sets of blood cultures, culture for urine, chest X-ray, computed tomography (CT), magnetic resonance imaging (MRI), PET/CT with 18F- FDG, and invasive procedures such as biopsies of the bone marrow, liver, and enlarged lymph nodes. ForserologicaltestsofpatientswithFUO,WBC count,crp,esr,serumferritin,andpctmeasured for the first time after admission were selected for the analysis in this study. WBC count was measured by XS- 800i Automatic Hematology Analyzer (SYSMEX Co., Hyogo, Japan). CRP was tested using an Automatic Biochemistry analyzer (Olympus AU2700, Olympus Corporation, Tokyo, Japan). ESR was determined by the Westergren method. Serum ferritin levels were measured by ARCHITECT T2000 SR equipment (ARCHITECT ferritin reagent kit, Abbott Ireland Diagnostics Division, Dublin, Ireland). Serum PCT was measured by immunochemiluminometric assays (LUMI test PCT, Brahms Diagnostica, Berlin, Germany). Patients confirmed as FUO were classified according to their final diagnoses into 5 groups, including infectious diseases, neoplasms, collagen vascular diseases, miscellaneous diseases, and no diagnosis. Statistical analysis: The data analyses were performed using the Statistical Package for Social Science Version 22.0 (SPSS Inc., Chicago, IL, USA). The normality of data distribution was assessed by Kolmogorov-Smirnov tests. The continuous variables following normal distribution were expressed as mean ± standard deviation and variables not following normal distribution were expressed as median and inter-quartile range. If the continuous variables followed normal distribution and homoscedasticity, one-way ANOVA was used to compare more than 2 groups and LSD-test was used to compare 2 groups. However, if the continuous variables did not follow a normal distribution or homoscedasticity, comparisons of multiple samples were analyzed by the Kruskal-Wallis test. When the results of the Kruskal- Wallis test were statistically significant, groups were reanalyzed using the Mann-Whitney test. Data were analyzed by sensitivity, specificity, positive andnegative likelihood ratios derived from the receiver operating characteristic (ROC) curve, and area under the ROC curve (AUC). All tests were 2-tailed, with the risk level set at 5z and statistical significance defined as p value of <0.05. RESULTS Table 1 shows the distribution of causes underlying FUO.Ofthe383cases,325 (84.9z) individuals reached a final diagnosis and,for 58 cases (15.1z), the diagnosis remained unknown. Infectious diseases, neoplasms, and collagen vascular diseases were the main etiologies of FUO. Infectious diseases were the most common and accounted for 33.9z of FUO cases, neoplasms for 21.1z, collagen vascular diseases for 25.1z, andmiscellaneous diseases for 4.7z. The most prevalent FUO Table 1. Causes of fever unknown origin Causes of fever Proportion (z) Infectious diseases 130 (33.9) Bacterial infection 101 (26.4) Tuberculosis 50 (13.1) Endocarditis 13 (3.4) Septicaemia 16 (4.2) Typhoid fever 6 (1.6) Brucellosis 5 (1.3) Urinary tract infection 5 (1.3) Intraabdominal abscess 4 (1.0) Actinomycosis 2 (0.5) Viral infection 16 (4.2) Parasitic infection 10 (2.6) Fungal infection 3 (0.8) Neoplasms 81 (21.1) Hodgkin lymphoma 10 (2.6) Non-Hodgkin lymphoma 45 (11.7) Leukaemia 11 (2.9) Hepatoma 2 (0.5) myeloma 1 (0.3) Pulmonary cacinoma 2 (0.5) Colon carcinoma 1 (0.3) Pancreatic carcinoma 1 (0.3) Malignant histiocytosis 6 (1.6) Malignancy of unknown primary 2 (0.5) Collagen vascular diseases 96 (25.1) Adult-onset Still disease 65 (17.0) Systemic lupus erythematosus 8 (2.1) Sjogren syndrome 5 (1.3) Polyarteritis nodosa 2 (0.5) Rheumatic arthritis 5 (1.3) Mixed connective tissue disease 3 (0.8) Undifferentiated connective tissue disease 3 (0.8) Vasculitis 2 (0.5) Polymyositis 3 (0.8) Miscellaneous diseases 18 (4.7) Subacute thyroiditis 3 (0.8) Necrotizing lymphadenitis 7 (1.8) Hemophagocytic syndrome 5 (1.3) Drug fever 3 (0.8) No diagnosis 58 (15.1) 379

3 Table 2. Patients' demographic characteristics of disease groups in diagnosed FUO (means ± SD) Infectious disease (n = 130) Neoplasm (n = 81) Collagen vascular disease (n = 96) Miscellaneous disease (n = 18) P-value Age (yrs) 42.8 ± ± ± ± 16.7 < Duration of fever before admission (days) 28.0 ± ± ± ± Duration of hospitalization (days) 24.9 ± ± ± ± Diagnostic time after admission (days) 11.3 ± ± ± ± 9.3 < FUO, fever unknown origin; SD, standard deviation. Table 3. Comparison of inflammatory makers' levels in disease groups (median [IQR] or [means ± SD]) Inflammatory marker Infectious disease (n = 130) Neoplasm (n = 81) Collagen vascular disease (n = 96) P-value White blood cell count ( 1,000/mm 3 ) 11.7 ( ) 11.0 ( ) 10.8 ( ) 0.76 C-reactive protein (mg/l) 82.4 ± ± ± Ferritin (ng/ml) ( ) 1,913.2 ( ,676.5) ( ,742.6) Erythrocyte sedimentation rate (mm/h) 68.2 ± ± ± Serum procalcitonin (ng/ml) 3.64 ± ± ± IQR, inter-quartile range; SD, standard deviation. causes were determined as non-hodgkin lymphoma, tuberculosis, and adult Still's disease in this order, accounting for 11.7z, 13.1z, and 17.0z of the cases, respectively. Among the 383 cases with FUO, the temperature fluctuated from a low temperature of 38.39C toahigh temperature of 41.59C. Irregular fever accounted for 60.0z, remittent fever for 26.4z, recurrent fever for 12.3z, and undulant fever for 1.3z of the cases. Ninety seven percent of patients presented with other symptoms such as chills, weakness, night sweats, pharyngalgia, cough, expectoration, myalgia, and arthralgia. Among the 325 (84.9z) patients with a final diagnosis, 108 (28.2z) patients were finally diagnosed mainly by serology or pathogen testing, 52 (13.6z)patientsbyimageology, 64 (16.7z) patients by pathology of bonemarrow, lymph node, liver, or skin biopsy, 75 (19.6z) cases by the effect of diagnostic treatment, 26 (6.8z) by medical history or the course of disease progression. Twelve patients were diagnosed by more than 2 bonemarrow or lymph node biopsies. Age, duration of fever before admission, and diagnostic time after admission significantly differed among the infectious disease, neoplasm, collagen vascular disease, and miscellaneous disease groups (p < , p = 0.006, and p < , respectively), while duration of hospitalization did not ( p = 0.054). Patients in the neoplasm group were older than those in the other 3 groups, including infectious disease, collagen vascular disease, and miscellaneous disease groups ( p = 0.006, p < , and p = 0.001, respectively). The duration of fever before admission of the patients in the neoplasm and collagen vascular disease groups was longer than that of the patients in the infectious disease group (p = and p = 0.007, respectively). The diagnostic time after admission of patients from the neoplasm and collagen vascular disease groups was longer than that of the patients in the infectious disease group (both p < ) (Table 2). No significant difference was observed in terms of WBC, CRP, and ESR between the infectious disease, neoplasm, and collagen vascular disease groups (p = 0.76, p = 0.32, and p = 0.26, respectively). However, serum ferritin and PCT levels were statistically different among the 3 groups (p = and p = 0.011, respectively). Serum ferritin levels of the infectious disease group were lower than those of the neoplasm and collagen vascular disease groups (p = and p = 0.032), while no significant difference was observed between the neoplasm and collagen vascular disease groups (p = 0.24). In contrast, serum PCT levels of the infectious disease group were higher than those of the neoplasm and collagen vascular disease groups ( p = and p = 0.007, respectively) (Table 3). Besides, PCT levels of patients with bacterial infections were higher than those of patients with non-bacterial infections (4.56 ± 0.58 vs. 0.43± 0.15 ng/ml, p = 0.007). Fig. 1 shows the ROC curve of inflammatory markers for discriminating infectious from non-infectious causes (neoplasms and collagen vascular diseases) in patients with FUO. The AUC of serum ferritin for noninfectious diseases was 0.78 (95z CI, ), with a sensitivity of 62z, a specificity of 83z, positive likelihood ratio of 3.65, and negative likelihood ratio of The AUC of PCT for infectious diseases was 0.92 (95z CI, ), with a sensitivity of 81z, a specificity of 88z, a positive likelihood ratio of 6.75, and a negative likelihood ratio of DISCUSSION FUO is a challenging clinical problem with a long course and much complex etiologies. Although the diagnostic methods and tools developed rapidly, the diagnostic time after admission is still very long and the cost is significantly high. More importantly, 7z to 51z of cases with FUO remained undiagnosed (15.1z reported inthisstudy).therefore,itisurgenttodeterminethe distribution of etiologies and their clinical data dynamically from patients with FUO and explore some detection indices for the differential diagnosis of FUO. In this study, the etiologies of FUO were identified as 380

4 Value of Inflammatory Markers on Fever of Unknown Origin Fig. 1. Receiver operating characteristic (ROC) curve of inflammatory markers for discriminating infectious from non-infectious causes in patients with fever of unknown origin. (A) ROC curve of serum ferritin for noninfectious diseases (neoplasms and collagen vascular diseases). (B) ROC curve of procalcitonin for infectious diseases. follows: infectious diseases, 33.9z; neoplasms, 21.1z; collagen vascular diseases, 25.1z; miscellaneous diseases, 4.7z; and unidentified, 15.1z. Infectionwasthe most common cause and collagen vascular diseases ranked second followed by malignancy. These 3 causes accounted for about 80z of all etiologies. The result was consistent with that of previous studies in China (5,6,15). About 84.9z of cases with FUO reached a definite diagnosis. The rate of definite diagnosis determined herein is higher than that previously reported (5,6), which can be attributed to the progress of diagnostic methods and tools such as the use of PET/CT with 18F-FDG (16 19). However, 15.1z of cases remained undiagnosed in this study. FUO is likely to remain a clinical problem and more advanced diagnostic techniques need to be explored. The present study showed that patients from the neoplasm group were the oldest among all cases with FUO. Therefore, clinicians should remain on high alert for neoplasm in the elderly with FUO. The duration of fever before admission and the diagnostic time after admission in the neoplasm and collagen vascular disease groups was longer than that of the infectious disease group. Adult Still's disease is the principal cause in the collagen vascular disease group, accounting for 67.7z (65/96) in this group. It usually lacks specific manifestations and its diagnosis is always based on clinical features after excluding other diseases (20). Therefore, its diagnostic time after admission is rather long. In this study, there were 55 cases with lymphoma, including non-hodgkin lymphoma and Hodgkin lymphoma, accounting for 67.9z (55/81) of the patients in the neoplasm group. Lymphoma is frequently lacking specific clinical presentations in the early stage. Although medical imaging technologies such as ultrasound, CT, and MRI greatly progressed, the diagnosis of lymphoma remained difficult. In some cases, biopsies from the bone marrow or enlarged lymph node have to be repeatedtwiceormoreforthediagnosisoflymphoma.asa result, the diagnostic time after admission is longer. Serum ferritin was discovered by Laufberger in 1937 and was first measured in serum by a radioimmunoassay (21). It is not only used as an index of the extent of iron storage in the body, but also as an acute phase reactant in response to inflammation (22). If serum ferritin is elevated, infectious diseases, neoplastic diseases, liver diseases, and hemophagocytic syndrome should be considered (23 25). Our study demonstrated that serum ferritin levels of patients in the infectious disease group were lower than those of patients in the neoplasm and collagen vascular disease groups (p = and p = 0.032, respectively). This result is in agreement with previous studies (11,26,27). Cunha et al. (26) asserted that serum ferritin levels should be used combined with medical history, physical examination, and other serological tests to exclude infections because the likelihood of infectious disease is low when early serum ferritin levels are elevated. Efstathiou et al. (11) reported that ferritin concentration <500 ng/ml, eosinopenia <40/mm 3, and CRP >6 mg/dl were independently associated with infectious diseases. Another study by Kim et al. (27) demonstrated that the median serum ferritin level in infectious diseases was lower than that in the hematologic disease group and noninfectious inflammatory diseases. The optimal cutoff value of serum ferritin levels to predict FUO caused by a noninfectious disease (hematologic diseases and noninfectious inflammatory diseases) was established as 561 ng/ml. In addition, Cunha et al. (26) recognized that serum ferritin is an acute phase reactant and ferritin levels may be elevated as part of the acute phase. However, the process of FUO is prolonged for more than 3 weeks according to the definition. Therefore, the authors considered that highly elevated serum ferritin levels should not be ascribed to an acute phase reaction occurring early in the first few days of an acute process, but rather are reflective of the disorder causing FUO in the FUO context. Hazard et al. (28) showed that patients with 381

5 hematologic malignancies presented with a reduced cellular immunity level and increased serum ferritin levels. They concluded that serum ferritin is associated with immunity. In addition, elevated serum ferritin is related to chronic inflammation such as hematologic malignancy or autoimmune diseases. When the chronic inflammation occurs, the body produces hepcidin in the liver to prevent the utilization of serum iron by pathogens or tumor cells by suppressing intestinal absorption and sequestration of iron in the macrophages, causing an increase in serum ferritin (27,29,30). Therefore, serum ferritin may be a useful index for the discrimination between infectious diseases and noninfectious diseases (neoplasms and collagen vascular diseases) in the treatment of patients with FUO. Our data also showed that serum PCT levels of patients in the infectious disease group were higher than that of patients in the neoplasm and collagen vascular disease groups ( p = 0.016and p = 0.007, respectively). The finding was not in line with the previous study by Naito et al. (31), probably due to the limited number of subjects in the previous study. PCT levels were measured in only 45 patients with FUO. PCT is the prohormone of calcitonin secreted by C cells in the thyroid in response to hypercalcemia (32). Its secretion is stimulated by various cytokines in systemic inflammatory conditions, especially, in bacterial infections, while the PCT production is decreased in viral infections, likely resulting from increased interferon gamma production (33). In addition, elevated PCT levels are not observed in noninfectious inflammatory conditions. The levels of PCT in bacterial infection conditions could be detectable within 3 4 hours and peak within 6 24 hours, which is earlier than both CRP and ESR. Simon et al. (34) recognized that serum PCT was particularly good for differentiating bacterial infections from viral infections and differentiating bacterial infections from other noninfectious causes of systemic inflammation. In our study, PCT levels of patients from the infectious disease group was higher than those of patients from other noninfectious disease groups resulting from the fact that the majority of the infectious diseases, about 76z (99/130), were bacterial infections. This result was in agreement with previous findings. Thus, PCT is also a valuable marker in differentiating infectious causes from non-infectious causes in patients with FUO. In addition, no significant differences were detected in the levels of WBC, CRP, and ESR among the infectious disease, neoplasm, and collagen vascular disease groups (p = 0.76, p = 0.32, and p = 0.26, respectively) in our study. These findings are in line with previous studies (27,31). Although these 3 indices would be elevated in inflammatory conditions, they are nonspecific for the differential diagnosis of FUO. Our study has some limitations. The results were obtained at a single hospital and there is a probability of selective bias. The results should be confirmed by multicenter studies due to the possibility of region-specific differences. Besides, we included only inflammatory markers measured for the first time after admission. They may fluctuate with time, which may influence the results of this study. In conclusion, with the progress in the development of diagnostic methods and tools, more and more patients with FUO are eventually diagnosed, but, for some, the origin of FUO remains unknown. FUO remains a clinical problem in China. Among the causes of FUO, it takes longer to diagnose neoplasms and collagen vascular diseases. Elevated serum ferritin levels in patients with FUO are associated with a low probability of an infectious disease and a high probability of a neoplasm and collagen vascular disease. In contrast, increased serum PCT level is associated with a higher probability of an infectious disease rather than a neoplasm or collagen vascular disease. Therefore, serum ferritin and serum PCT may be useful in discriminating infectious from non-infectious causes (including neoplasmsand collagen vascular diseases) in patients with FUO. Acknowledgments We gratefully thank for Hui-Bin Ning for collecting the data and Yan-Hong Kang for comments on the manuscript. This work was supported by the National Key Specialty Construction of Clinical Projects of China (2012). Conflict of interest None to declare. REFERENCES 1. Petersdorf RG, Beeson PB. Fever of unexplained origin: report on 100 cases. Medicine (Baltimore). 1961;40: Durack DT, Street AC. Fever of unknown origin reexamined and redefined. Curr Clin Top Infect Dis. 1991;11: 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 (Baltimore). 2007;86: Tabak F, Mert A, Celik AD, et al. Fever of unknown origin in Turkey. Infection. 2003;31: Liu KS, Sheng WH, Chen YC, et al. Fever of unknown origin: a retrospective study of 78 adult patients in Taiwan. J Microbiol Immunol Infect. 2003;36: Zhiyong Z, Bingjun L, Xiaoju L, et al. Fever of unknown origin: a report from China of 208 cases. Int J Clin Pract. 2003;57: 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: Ergonul O, Willke A, Azap A, et al. Revised definition of `fever of unknown origin': limitations and opportunities. J Infect. 2005; 50: ErtenN,SakaB,OzturkG,etal.Feverofunknownorigin:a report of 57 cases. Int J Clin Pract. 2005;59: Kucukardali Y, Oncul O, Cavuslu S, et al. The spectrum of diseases causing fever of unknown origin in Turkey: a multicenter study. Int J Infect Dis. 2008;12: 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: Naito T, Mizooka M, Mitsumoto F, et al. Diagnostic workup for fever of unknown origin: a multicenter collaborative retrospective study. BMJ Open. 2013;3:e Yamanouchi M, Uehara Y, Yokokawa H, et al. Analysis of 256 cases of classic fever of unknown origin. Intern Med. 2014;53: Knockaert DC, Vanderschueren S, Blockmans D. Fever of unknown origin in adults: 40 years on. J Intern Med. 2003;253: 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: Roy SG, Karunanithi S, Dhull VS, et al. 18 F-FDG PET/CT aids the diagnosis of adult onset Still's disease in a patient with fever of unknown origin. Rev Esp Med Nucl Imagen Mol. 2014;33: Krem MM, Pan L, Blinder MA. 18 F-FDG-PET-facilitated diagnosis of lymphoma presenting with fever of unknown origin and cold agglutination. Leuk Lymphoma. 2007;48: Gafter-Gvili A, Raibman S, Grossman A, et al. [18F] FDG- 382

6 Value of Inflammatory Markers on Fever of Unknown Origin PET/CT for the diagnosis of patients with fever of unknown origin. QJM. 2015;108: Buch-Olsen KM, Andersen RV, Hess S, et al. 18F-FDG-PET/CT in fever of unknown origin: clinical value. Nucl Med Commun. 2014;35: Larkin JG, Sturrock RD. Adult Still's disease. A new consideration in pyrexia of unknown origin. Scott Med J. 1983;28: Addison GM, Beamish MR, Hales CN, et al. An immunoradiometric assay for ferritin in the serum of normal subjects and patients with iron deficiency and iron overload. J Clin Pathol. 1972;25: Tran TN, Eubanks SK, Schaffer KJ, et al. Secretion of ferritin by rat hepatoma cells and its regulation by inflammatory cytokines and iron. Blood. 1997;90: Koduri PR, Carandang G, DeMarais P, et al. Hyperferritinemia in reactive hemophagocytic syndrome report of four adult cases. Am J Hematol. 1995;49: Lee MH, Means RT Jr. Extremely elevated serum ferritin levels in a university hospital: associated diseases and clinical significance. Am J Med. 1995;98: Di Bisceglie AM, Axiotis CA, Hoofnagle JH, et al. Measurements of iron status in patients with chronic hepatitis. Gastroenterology. 1992;102: Cunha BA. Diagnostic significance of non-specific laboratory abnormalities in infectious diseases. In:Gorbach SL, Bartlett JG, Blacklow NR, editors. Infectious Diseases. 3rd ed. Philadelphia: Lippincott Williams & Wilkins; p Kim SE, Kim UJ, Jang MO, et al. Diagnostic use of serum ferritin levels to differentiate infectious and noninfectious diseases in patients with fever of unknown origin. Dis Markers. 2013;34: Hazard JT, Drysdale JW. Ferritinaemia in cancer. Nature. 1977; 265: Zandman-Goddard G, Shoenfeld Y. Ferritin in autoimmune diseases. Autoimmun Rev. 2007;6: Ganz T, Nemeth E. Iron sequestration and anemia of inflammation. Semin Hematol. 2009;46: Naito T, Torikai K, Mizooka M, et al. Relationships between causes of fever of unknown origin and inflammatory markers: a multicenter collaborative retrospective study. Intern Med. 2015; 54: Whicher J, Bienvenu J, Monneret G. Procalcitonin as an acute phase marker. Ann Clin Biochem. 2001;38: Gilbert DN. Procalcitonin as a biomarker in respiratory tract infection. Clin Infect Dis. 2011;52 Suppl 4:S Simon L, Gauvin F, Amre DK, et al. Serum procalcitonin and C- reactive protein levels as markers of bacterial infection: a systematic review and meta-analysis. Clin Infect Dis. 2004;39: