JBJS | Fluorine-18 Fluorodeoxyglucose-Positron Emission Tomography: A Highly Accurate Imaging Modality for the Diagnosis of Chronic Musculoskeletal Infections

The Journal of Bone & Joint Surgery, Volume 83, Issue 5

Articles | May 01, 2001

Fluorine-18 Fluorodeoxyglucose-Positron Emission Tomography: A Highly Accurate Imaging Modality for the Diagnosis of Chronic Musculoskeletal Infections

F. De Winter, MD; C. Van de Wiele, MD; D. Vogelaers, MD, PhD; K. De Smet, MD; R. Verdonk, MD, PhD; R. A. Dierckx, MD, PhD

View Disclosures and Other Information

Investigation performed at Ghent University Hospital, Ghent, Belgium
F. De Winter, MD C. Van de Wiele, MD D. Vogelaers, MD, PhD K. De Smet, MD R. Verdonk, MD, PhD R.A. Dierckx, MD, PhD Division of Nuclear Medicine (F. De W., C. Van de W., and R.A.D.), Division of Orthopedics (K. De S. and R.V.), and Section of Infectiology, Division of Internal Medicine (D.V.), Ghent University Hospital, De Pintelaan 185-9000 Ghent, Belgium. E-mail address for F. De W.: frederic.dewinter@rug.ac.be
No benefits in any form have been or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.

The Journal of Bone & Joint Surgery. 2001; 83:651-660

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

text A A A

Abstract

Background: The noninvasive diagnosis of chronic musculoskeletal infections remains a challenge. Recent studies have indicated that fluorine-18 fluorodeoxyglucose-positron emission tomography is a highly accurate imaging technique and is significantly more accurate than the combination of a bone scan and a white blood-cell scan for the diagnosis of chronic infection in the central skeleton (p < 0.05). However, patients who had had surgery within the previous two years were excluded from study. It was our aim to evaluate the technique in an unselected, clinically representative population.

Methods: Sixty patients with a suspected chronic musculoskeletal infection involving the central skeleton (thirty-three patients) or the peripheral skeleton (twenty-seven patients) were studied with fluorine-18 fluorodeoxyglucose-positron emission tomography. Thirty-five patients had had surgery within the previous two years. The fluorine-18 fluorodeoxyglucose-positron emission tomography studies were read in a blinded, independent manner by two experienced readers. The final diagnosis was based on histopathological studies or microbiological culture (eighteen patients) or on clinical findings after at least six months of follow-up (forty-two patients).

Results: On the final composite assessment, twenty-five patients had infection and thirty-five did not. All twenty-five infections were correctly identified by both readers. There were four false-positive findings; in two of these cases, surgery had been performed less than six months prior to the study. The sensitivity, specificity, and accuracy were 100%, 88%, and 93% for the whole group; 100%, 90%, and 94% for the subgroup of patients with a suspected infection of the central skeleton; and 100%, 86%, and 93% for the subgroup of patients with a suspected infection of the peripheral skeleton. Interobserver agreement was excellent (kappa = 0.97).

Conclusions: Fluorine-18 fluorodeoxyglucose-positron emission tomography is highly accurate as a single technique for the evaluation of chronic musculoskeletal infections. It is especially valuable in the evaluation of the central skeleton, where white blood-cell scans are less useful. Because of its simplicity and high degree of accuracy, it has the potential to become a standard technique for the diagnosis of chronic musculoskeletal infections. Further studies are needed to assess its ability to identify infections at the sites of total joint replacements and to distinguish infection from aseptic loosening of these prostheses.

Figures in this Article

Introduction

Introduction | Materials and Methods | Results | Discussion | References

Although many techniques have been proposed for the noninvasive assessment of musculoskeletal infections, clinicians are still confronted with an indeterminate diagnosis in many patients and action is often limited to a "wait-and-see" policy or empirical antibiotic treatment^1-3. Inflammatory parameters such as the C-reactive protein level, erythrocyte sedimentation rate, and white blood-cell count are helpful but lack both sensitivity (especially in the assessment of low-grade infections) and specificity^2,4-6. Computed tomography and magnetic resonance imaging provide excellent anatomical detail⁷. Magnetic resonance imaging is extremely helpful in the evaluation of patients who have not been managed operatively, but it is currently of limited value for discriminating between edema and active infection after surgery and in the presence of metallic implants^8,9. Technical improvements in magnetic resonance imaging have led to promising results in selected series of patients^10,11, but these improvements need to be confirmed in larger, clinically representative series.

The use of three-phase bone scintigraphy in combination with white blood-cell scanning is generally accepted as a method with clinically good accuracy (range, 79% to 100%) for the diagnosis of infection involving the peripheral skeleton^12-17. The accuracy of this combined strategy is decreased (1) in the presence of low-grade chronic infections (lower sensitivity)^9,12; (2) in the presence of periskeletal soft-tissue infection, because of the limited resolution of conventional nuclear imaging (lower sensitivity and specificity); (3) in the evaluation of the central skeleton, because of the presence of normal bone marrow and the possibility of so-called cold lesions (lower sensitivity and specificity)^{12,13,15,18-20}; and (4) after trauma or surgery, because of the presence of ectopic hematopoietic bone marrow (lower specificity)^21,22. A combination of colloid scintigraphy and white blood-cell scanning has been associated with substantially improved specificity for the diagnosis of infection at the sites of hip and knee prostheses^19,23,24, and a combination of bone and gallium scans has been proposed as a way to improve both sensitivity and specificity in the diagnosis of infection in the vertebral column¹⁶. However, the need for two techniques (a bone-marrow scan and a white blood-cell scan or a bone scan and a gallium scan) or even three techniques (a bone scan, a white blood-cell scan, and a bone-marrow scan) is not practical, adds to the cost and to the amount of radiation to which the patient is exposed, and is time-consuming. Therefore, an equally specific and sensitive all-in-one technique would be most welcome.

Fluorine-18 fluorodeoxyglucose-positron emission tomography has made it possible to evaluate cellular glucose metabolism^25-28. Specifically, this technique demonstrates the increased utilization of glucose by activated neutrophils and macrophages in inflammatory reactions^29-32. Sugawara et al.³³ showed, in a rat model, that the uptake of fluorine-18 fluorodeoxyglucose at the site of infection is higher than the uptake of other tracers and that the fluorine-18 fluorodeoxyglucose-uptake mechanism is related to the infiltration of inflammatory cells. Consequently, fluorine-18 fluorodeoxyglucose-positron emission tomography has been proposed as a promising technique for the evaluation of musculoskeletal infections^18,33-39.

The purpose of the present study was to evaluate the accuracy of fluorine-18 fluorodeoxyglucose-positron emission tomography in the diagnosis of chronic musculoskeletal infections in an unselected series of sixty consecutive patients.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1-A:: Coronal and sagittal fluorodeoxyglucose-positron emission tomography images showing intense uptake (arrows) in the anterolateral aspect of the middle third of the right femur after previous trauma. Staphylococcus aureus grew on culture of specimens taken from the site of increased uptake at the time of the operation.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1-B:: Corresponding anteroposterior and lateral radiographs.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2:: Case 1. Sagittal fluorodeoxyglucose-positron emission tomography scan through the vertebral column of a patient with microbiologically proven Staphylococcus aureus spondylodiscitis of the eleventh and twelfth thoracic vertebrae (arrow). The myocardium of the left ventricle is also shown for reference (arrowhead).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 3-A:: Figs. 3-A and 3-B Case 23. Coronal (Fig. 3-A) and sagittal (Fig. 3-B) fluorodeoxyglucose-positron emission tomography scans through the lumbar vertebrae of a thirteen-year-old girl in whom extreme scoliosis had been treated with osteosynthesis nine months earlier. She presented with low-back pain, an elevated erythrocyte sedimentation rate (109 mm at two hours), and an elevated level of C-reactive protein (3.18 mg/dL). The bone and white blood-cell scans were equivocal. The fluorodeoxyglucose-positron emission tomography scans show increased uptake (arrows) in the bodies of the third, fourth, and fifth lumbar vertebrae, where uptake is normally very low, as well as increased splenic uptake (arrowhead in Fig. 3-A). Staphylococcus aureus grew on culture of specimens obtained at the time of the operation.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 3-B:: Figs. 3-A and 3-B Case 23. Coronal (Fig. 3-A) and sagittal (Fig. 3-B) fluorodeoxyglucose-positron emission tomography scans through the lumbar vertebrae of a thirteen-year-old girl in whom extreme scoliosis had been treated with osteosynthesis nine months earlier. She presented with low-back pain, an elevated erythrocyte sedimentation rate (109 mm at two hours), and an elevated level of C-reactive protein (3.18 mg/dL). The bone and white blood-cell scans were equivocal. The fluorodeoxyglucose-positron emission tomography scans show increased uptake (arrows) in the bodies of the third, fourth, and fifth lumbar vertebrae, where uptake is normally very low, as well as increased splenic uptake (arrowhead in Fig. 3-A). Staphylococcus aureus grew on culture of specimens obtained at the time of the operation.

View Larger

TABLE I: Data on Patients with a Suspected Infection of the Central Skeleton

*THR = site of total hip replacement, and TSR = site of total shoulder replacement. †WBC = white blood-cell, MRI = magnetic resonance imaging, and CT = computerized tomography. ‡TP = true-positive finding, TN = true-negative finding, and FP = false-positive finding. #Bone and WBC scans were negative because of previous radiotherapy.

Case	Location of Infection*	Time Since Prev. Op.	Metallic Implants	Score (Reader 1/ Reader 2)	Imaging Technique(s)†	Type of Diagnosis	Final Diagnosis	Accuracy‡ (Reader 1/Reader 2)
?1	Spine	—	No	+/+	—	Microbiological	Spondylodiscitis	TP/TP
?2	Sacrum	—	No	+/+	—	Microbiological	Osteomyelitis#	TP/TP
?3	THR	12 mos	Yes	+/+	WBC, colloidscans; arthrography	Clinical	Infection	TP/TP
?4	TSR	3 yrs	Yes	—/—	Bone, WBCscans	Clinical	No infection	TN/TN
?5	Sacroiliac joint	—	No	—/—	Bone, WBC scans; MRI	Clinical	No infection	TN/TN
?6	Spine	4 yrs	Yes	—/—	Bone scan	Clinical	No infection	TN/TN
?7	Prox. femur	6 wks	No	+/+	—	Microbiological	No infection	FP/FP
?8	TSR	19 mos	Yes	—/—	—	Clinical	No infection	TN/TN
?9	Spine	—	No	—/—	Radiograph, MRI	Clinical	No infection, severe arthrosis deformans	TN/TN
10	Prox. femur	16 mos	Yes	+/+	—	Microbiological	No infection, screw insertion site in bone	FP/FP
11	TSR	14 mos	Yes	+/+	—	Microbiological	Infection	TP/TP
12	Sacrum	12 mos	No	+/+	—	Microbiological	Osteomyelitis	TP/TP
13	Prox. femur	4 yrs	No	—/—	Bone scan	Clinical	No infection	TN/TN
14	Prox. femur	8 mos	Yes	+/+	—	Microbiological	Osteomyelitis	TP/TP
15	Spine	10 mos	Yes	—/—	—	Microbiological	No infection	TN/TN
16	Clavicle	4 yrs	No	—/—	Bone, WBC scans	Clinical	No infection	TN/TN
17	THR	10 mos	Yes	+/+	Bone, WBC, colloid scans	Clinical	Infection	TP/TP
18	Coccyx	20 yrs	No	+/+	MRI	Clinical	Osteomyelitis	TP/TP
19	Prox. femur	11 mos	No	+/+	—	Microbiological	Osteomyelitis	TP/TP
20	THR	15 mos	Yes	+/+	WBC, colloid scans	Clinical	Infection	TP/TP
21	Spine	—	No	—/—	MRI	Clinical	Soft-tissue infection	TN/TN
22	THR	18 yrs	Yes	—/—	WBC, colloid scans	Clinical	No infection	TN/TN
23	Spine	9 mos	Yes	+/+	—	Microbiological	Osteomyelitis	TP/TP
24	THR	22 yrs	Yes	—/—	WBC scan	Clinical	No infection	TN/TN
25	THR	11 mos	Yes	—/—	Bone scan	Clinical	No infection	TN/TN
26	Spine	4 mos	No	—/—	Bone scan	Clinical (with clinical evolution)	Superficial infection	TN/TN
27	Spine	6 yrs	Yes	—/—	Bone scan; CT	Clinical	No infection	TN/TN
28	Spine	—	No	—/—	MRI	Clinical	No infection	TN/TN
29	Spine	—	No	—/—	MRI	Clinical	No infection	TN/TN
30	Spine	10 mos	Yes	—/—	—	Microbiological	No infection	TN/TN
31	THR	13 mos	Yes	+/+	—	Microbiological	Infection	TP/TP
32	Prox. femur	17 yrs	No	+/+	—	Microbiological	Osteomyelitis	TP/TP
33	Spine	4 mos	No	—/—	MRI	Clinical	No infection	TN/TN

Add to My JBJS

TABLE I: Data on Patients with a Suspected Infection of the Central Skeleton

Case	Location of Infection*	Time Since Prev. Op.	Metallic Implants	Score (Reader 1/ Reader 2)	Imaging Technique(s)†	Type of Diagnosis	Final Diagnosis	Accuracy‡ (Reader 1/Reader 2)
?1	Spine	—	No	+/+	—	Microbiological	Spondylodiscitis	TP/TP
?2	Sacrum	—	No	+/+	—	Microbiological	Osteomyelitis#	TP/TP
?3	THR	12 mos	Yes	+/+	WBC, colloidscans; arthrography	Clinical	Infection	TP/TP
?4	TSR	3 yrs	Yes	—/—	Bone, WBCscans	Clinical	No infection	TN/TN
?5	Sacroiliac joint	—	No	—/—	Bone, WBC scans; MRI	Clinical	No infection	TN/TN
?6	Spine	4 yrs	Yes	—/—	Bone scan	Clinical	No infection	TN/TN
?7	Prox. femur	6 wks	No	+/+	—	Microbiological	No infection	FP/FP
?8	TSR	19 mos	Yes	—/—	—	Clinical	No infection	TN/TN
?9	Spine	—	No	—/—	Radiograph, MRI	Clinical	No infection, severe arthrosis deformans	TN/TN
10	Prox. femur	16 mos	Yes	+/+	—	Microbiological	No infection, screw insertion site in bone	FP/FP
11	TSR	14 mos	Yes	+/+	—	Microbiological	Infection	TP/TP
12	Sacrum	12 mos	No	+/+	—	Microbiological	Osteomyelitis	TP/TP
13	Prox. femur	4 yrs	No	—/—	Bone scan	Clinical	No infection	TN/TN
14	Prox. femur	8 mos	Yes	+/+	—	Microbiological	Osteomyelitis	TP/TP
15	Spine	10 mos	Yes	—/—	—	Microbiological	No infection	TN/TN
16	Clavicle	4 yrs	No	—/—	Bone, WBC scans	Clinical	No infection	TN/TN
17	THR	10 mos	Yes	+/+	Bone, WBC, colloid scans	Clinical	Infection	TP/TP
18	Coccyx	20 yrs	No	+/+	MRI	Clinical	Osteomyelitis	TP/TP
19	Prox. femur	11 mos	No	+/+	—	Microbiological	Osteomyelitis	TP/TP
20	THR	15 mos	Yes	+/+	WBC, colloid scans	Clinical	Infection	TP/TP
21	Spine	—	No	—/—	MRI	Clinical	Soft-tissue infection	TN/TN
22	THR	18 yrs	Yes	—/—	WBC, colloid scans	Clinical	No infection	TN/TN
23	Spine	9 mos	Yes	+/+	—	Microbiological	Osteomyelitis	TP/TP
24	THR	22 yrs	Yes	—/—	WBC scan	Clinical	No infection	TN/TN
25	THR	11 mos	Yes	—/—	Bone scan	Clinical	No infection	TN/TN
26	Spine	4 mos	No	—/—	Bone scan	Clinical (with clinical evolution)	Superficial infection	TN/TN
27	Spine	6 yrs	Yes	—/—	Bone scan; CT	Clinical	No infection	TN/TN
28	Spine	—	No	—/—	MRI	Clinical	No infection	TN/TN
29	Spine	—	No	—/—	MRI	Clinical	No infection	TN/TN
30	Spine	10 mos	Yes	—/—	—	Microbiological	No infection	TN/TN
31	THR	13 mos	Yes	+/+	—	Microbiological	Infection	TP/TP
32	Prox. femur	17 yrs	No	+/+	—	Microbiological	Osteomyelitis	TP/TP
33	Spine	4 mos	No	—/—	MRI	Clinical	No infection	TN/TN

View Larger

TABLE II: Data on Patients with a Suspected Infection of the Peripheral Skeleton

*TKR = site of total knee replacement. †WBC = white blood-cell, and MRI = magnetic resonance imaging. ‡TN = true-negative finding, TP = true-positive finding, ind = indeterminate finding, and FP = false-positive finding.

Case	Location of Infection*	Time Since Prev. Op.	Metallic Implants	Score (Reader 1/ Reader 2)	Imaging Technique(s)†	Type of Diagnosis	Final Diagnosis	Accuracy‡ (Reader 1/Reader 2)
34	Distal femur	9 yrs	No	—/—	—	Microbiological	No infection	TN/TN
35	Elbow	7 mos	No	—/—	WBC scan	Clinical	No infection	TN/TN
36	Distal femur	6 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
37	Tibia	6 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
38	Tibia	13 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
39	Tibia	24 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
40	Tibia	30 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
41	Tibia	24 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
42	Tibia	28 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
43	TKR	25 mos	Yes	—/—	Bone, WBC scans	Clinical (with clinical evolution)	No infection	TN/TN
44	TKR	29 mos	Yes	—/—	Bone, WBC, colloid scans	Clinical	No infection	TN/TN
45	Metatarsal 1	16 mos	No	+/0	MRI	Clinical	Osteomyelitis	TP/ind
46	Calcaneus	7 mos	No	—/—	Bone, WBC scans	Clinical	Soft-tissue infection	TN/TN
47	Phalanx	18 mos	No	+/+	Bone scan, radiograph	Microbiological	Brodie abscess	TP/TP
48	Tibia	6 mos	Yes	—/—	WBC scan	Clinical	No infection	TN/TN
49	Knee	12 mos	No	—/—	—	Microbiological	Bursitis	TN/TN
50	Tibia	12 mos	No	—/—	WBC scan	Clinical	No infection	TN/TN
51	TKR	4 yrs	Yes	+/+	—	Microbiological	Aseptic loosening	FP/FP
52	TKR	5 yrs	Yes	—/—	Bone, WBC scans	Clinical	No infection	TN/TN
53	Tibia	4 mos	Yes	—/—	Bone, WBC scans	Clinical	Soft-tissue infection	TN/TN
54	Calcaneus	5 mos	No	—/—	WBC scan	Clinical	No infection	TN/TN
55	TKR	18 mos	Yes	+/+	WBC scan	Clinical	Infection	TP/TP
56	TKR	25 mos	Yes	+/+	WBC scan	Clinical	Infection	TP/TP
57	TKR	7 mos	Yes	+/+	WBC scan	Clinical	Infection	TP/TP
58	Ankle	—	No	—/—	Bone, WBC scans	Clinical	No infection	TN/TN
59	Tibia	5 mos	Yes	+/+	—	Microbiological	Osteomyelitis	TP/TP
60	Distal femur	4 mos	No	+/+	—	Microbiological	No infection	FP/FP

Add to My JBJS

TABLE II: Data on Patients with a Suspected Infection of the Peripheral Skeleton

Case	Location of Infection*	Time Since Prev. Op.	Metallic Implants	Score (Reader 1/ Reader 2)	Imaging Technique(s)†	Type of Diagnosis	Final Diagnosis	Accuracy‡ (Reader 1/Reader 2)
34	Distal femur	9 yrs	No	—/—	—	Microbiological	No infection	TN/TN
35	Elbow	7 mos	No	—/—	WBC scan	Clinical	No infection	TN/TN
36	Distal femur	6 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
37	Tibia	6 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
38	Tibia	13 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
39	Tibia	24 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
40	Tibia	30 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
41	Tibia	24 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
42	Tibia	28 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
43	TKR	25 mos	Yes	—/—	Bone, WBC scans	Clinical (with clinical evolution)	No infection	TN/TN
44	TKR	29 mos	Yes	—/—	Bone, WBC, colloid scans	Clinical	No infection	TN/TN
45	Metatarsal 1	16 mos	No	+/0	MRI	Clinical	Osteomyelitis	TP/ind
46	Calcaneus	7 mos	No	—/—	Bone, WBC scans	Clinical	Soft-tissue infection	TN/TN
47	Phalanx	18 mos	No	+/+	Bone scan, radiograph	Microbiological	Brodie abscess	TP/TP
48	Tibia	6 mos	Yes	—/—	WBC scan	Clinical	No infection	TN/TN
49	Knee	12 mos	No	—/—	—	Microbiological	Bursitis	TN/TN
50	Tibia	12 mos	No	—/—	WBC scan	Clinical	No infection	TN/TN
51	TKR	4 yrs	Yes	+/+	—	Microbiological	Aseptic loosening	FP/FP
52	TKR	5 yrs	Yes	—/—	Bone, WBC scans	Clinical	No infection	TN/TN
53	Tibia	4 mos	Yes	—/—	Bone, WBC scans	Clinical	Soft-tissue infection	TN/TN
54	Calcaneus	5 mos	No	—/—	WBC scan	Clinical	No infection	TN/TN
55	TKR	18 mos	Yes	+/+	WBC scan	Clinical	Infection	TP/TP
56	TKR	25 mos	Yes	+/+	WBC scan	Clinical	Infection	TP/TP
57	TKR	7 mos	Yes	+/+	WBC scan	Clinical	Infection	TP/TP
58	Ankle	—	No	—/—	Bone, WBC scans	Clinical	No infection	TN/TN
59	Tibia	5 mos	Yes	+/+	—	Microbiological	Osteomyelitis	TP/TP
60	Distal femur	4 mos	No	+/+	—	Microbiological	No infection	FP/FP

View Larger

TABLE III: Advantages and Disadvantages of Fluorine-18 Fluorodeoxyglucose-Positron Emission Tomography for the Diagnosis of Musculoskeletal Infections18,20,21,33,34,38,41,43

Advantages	Disadvantages
Early (1-hr) imaging	High cost
High target-to-background ratio	Possible lower sensitivity in diabetic patients
High resolution (±5 mm)	Patient must be sober for at least 4 hrs
High-count tomographic images	Technique is currently not widely available
Low bone and bone-marrow uptake	Differentiation between tumor and infection is not possible
Highly accurate in central skeleton
Not hindered by metal implants
No additional scans necessary, all-in-one technique
High interobserver agreement
Theoretically sensitive in low-grade infections
Use may be feasible in neutropenic patients

Add to My JBJS

TABLE III: Advantages and Disadvantages of Fluorine-18 Fluorodeoxyglucose-Positron Emission Tomography for the Diagnosis of Musculoskeletal Infections18,20,21,33,34,38,41,43

Advantages	Disadvantages
Early (1-hr) imaging	High cost
High target-to-background ratio	Possible lower sensitivity in diabetic patients
High resolution (±5 mm)	Patient must be sober for at least 4 hrs
High-count tomographic images	Technique is currently not widely available
Low bone and bone-marrow uptake	Differentiation between tumor and infection is not possible
Highly accurate in central skeleton
Not hindered by metal implants
No additional scans necessary, all-in-one technique
High interobserver agreement
Theoretically sensitive in low-grade infections
Use may be feasible in neutropenic patients

Materials and Methods

Introduction | Materials and Methods | Results | Discussion | References

Patient Selection

Sixty patients (thirty-two females and twenty-eight males [mean age, 48.2 years; range, thirteen to seventy-five years]) in whom chronic osteomyelitis, spondylodiscitis, or an infection at the site of a total joint prosthesis was suspected were prospectively studied from February to August 1999. To avoid selection bias, fluorine-18 fluorodeoxyglucose-positron emission tomography was performed in conjunction with other clinically requested investigations and the scans were read in a blinded fashion. Suspected chronic osteomyelitis was defined as a possible recurrence of formerly proven osteomyelitis or as the presence of symptoms of osteomyelitis for more than six weeks^18,34. Fifty-two patients had a history of surgery at the site of the suspected infection. Patients who had had recent musculoskeletal surgery were not excluded (median interval between last surgery and scan, fourteen months; range, 1.5 to 269 months). Eight patients with no history of surgery had a soft-tissue infection with clinical suspicion of concomitant osteomyelitis (four patients) or a history of needle insertion into a vertebra (four patients). No patient had hematogenous chronic osteomyelitis. Thirty-four patients had metallic implants in place at the time of scanning. Twenty-seven patients had a suspected infection of the peripheral skeleton, and thirty-three had a suspected infection of the central skeleton (the vertebrae, sacrum, coccyx, ribs, pelvis, proximal part of the femur, or proximal part of the humerus). Thirty-three patients were receiving antibiotics at the time of scanning.

Approval of the ethics committee was obtained, and the patients gave informed written consent prior to inclusion in the study. In eighteen patients the final diagnosis was based on the results of microbiological culture or histopathological examination, and in forty-two patients it was made after a clinical follow-up period of at least six months (median, nine months; range, six to twelve months). Clinical follow-up parameters⁴⁰ included (1) a decrease in pain and an improvement in local clinical inflammatory signs during antibiotic treatment, (2) improvement in laboratory inflammatory parameters, and (3) agreement among correlative imaging techniques (bone-scanning, white blood-cell scanning, magnetic resonance imaging, and, eventually, colloid scanning) (Tables I and II). The final diagnosis was made with use of all available clinical parameters by the referring orthopaedic surgeon, who was unaware of the results of fluorine-18 fluorodeoxyglucose-positron emission tomography.

Fluorine-18 Fluorodeoxyglucose-Positron Emission Tomography Imaging

All studies were performed with use of a positron emission tomography scanner (ECAT 951/31; Siemens/CTI, Knoxville, Tennessee) that provides thirty-one contiguous, 3.375-mm-thick slices with an axial field of view of 108 mm. Patients fasted for at least six hours before administration of the fluorine-18 fluorodeoxyglucose. The blood-glucose level was checked in all patients by means of capillary sampling (Glucometer Elite; Diagnostics Division, Bayer, Tarrytown, New York) in order to exclude false-negative results due to elevated serum-glucose concentrations. The serum-glucose level prior to the injection of fluorine-18 fluorodeoxyglucose did not exceed 6.9 mmol/L in any patient, and no patient was given insulin to lower the serum-glucose level. Fluorine-18 fluorodeoxyglucose was injected at a dose of 370 MBq. Positron emission tomography scanning was started sixty minutes after injection. Depending on the area of interest (one to six bed positions), the total acquisition time ranged from fifteen to sixty minutes. Attenuation correction was not performed.

Image reconstruction was performed with the ordered-subsets expectation-maximization algorithm with two iterations and eight subsets.

Image Interpretation

Fluorine-18 fluorodeoxyglucose-positron emission tomography images were evaluated independently by two certified nuclear-medicine physicians experienced in positron emission tomography imaging; these physicians were blinded to the final diagnosis and to the results of other studies and laboratory results. In order for the two readers to become accustomed to the normal variations in skeletal uptake, normal skeletal variations were agreed upon after a review of the whole-body scans of thirty patients who had been referred for the evaluation of malignant melanoma.

The results were graded on a three-point scale, with "—" indicating no increased uptake (no osteomyelitis), "0" indicating equivocal results (no clear spatial differentiation between increased uptake by soft tissue and bone could be made), and "+" indicating increased uptake inside the bone surface or adjacent to prosthetic material. Skeletal uptake was defined as being increased if it was higher than the muscle uptake at a comparable tissue depth and higher than the skeletal uptake on the contralateral side. When the contralateral side was not available for comparison, the uptake had to be deemed higher than that in healthy adjacent bone or marrow tissue. The uptake in what was judged to be the joint capsule was not considered positive for infection. The scores that were independently assigned by the two readers were compared with the final diagnosis. Sensitivity, specificity, and accuracy were calculated for each observer.

Statistical Analysis

Interobserver agreement concerning the absence or presence of osseous infection on the final composite assessment was determined with use of Cohen kappa statistics (a kappa value of 1.0 indicated perfect agreement; 0.81 to 0.99, very good agreement; 0.61 to 0.80, good agreement; 0.0 to 0.60, poor to moderate agreement; and 0.0, agreement that was no better than chance). Differences in proportion were assessed with use of a chi-square test.

Results

Introduction | Materials and Methods | Results | Discussion | References

The mean blood-glucose level of the sixty patients was 4.5 mmol/L (range, 3.4 to 6.9 mmol/L). Twenty-five patients (42%) were found to have osseous infection on the final composite assessment. Data on the individual patients are given in Tables I and II. Both readers correctly identified all infections on fluorine-18 fluorodeoxyglucose-positron emission tomography scans. When the "+" and "0" scores were classified as positive and the "—" scores were classified as negative, the sensitivity, specificity, and accuracy for the whole group were 100% (twenty-six of twenty-six), 88% (thirty of thirty-four), and 93% (fifty-six of sixty), respectively. The corresponding values for the subgroup of thirty-three patients who had a suspected infection of the central skeleton were 100% (thirteen of thirteen), 90% (eighteen of twenty), and 94% (thirty-one of thirty-three), and the values for the subgroup of twenty-seven patients who had a suspected infection of the peripheral skeleton were 100% (thirteen of thirteen), 86% (twelve of fourteen), and 93% (twenty-five of twenty-seven). The presence or absence of infection was accurately determined on the fluorine-18 fluorodeoxyglucose-positron emission tomography scans of all fifteen patients who had a suspected infection in the vertebral column (including the sacrum and the coccyx). Since the overall sensitivity was 100% in all populations, we concluded that sensitivity was not negatively influenced by antibiotic treatment or by the lack of attenuation correction.

Seventeen patients were evaluated for a suspected periprosthetic infection involving the shoulder (three patients), hip (seven patients), or knee (seven patients). Scanning revealed eight true-positive results, eight true-negative results, and one false-positive result. The one false-positive result was recorded for a female patient (Case 51) who had intense uptake of fluorine-18 fluorodeoxyglucose at the total knee prosthesis-bone interface and in the joint capsule. Aseptic loosening and a hyperplastic capsule were noted at the time of surgery. All cultures remained negative.

In the group of patients without prostheses, there were three additional false-positive findings, with both readers assigning the same score. One patient (Case 10) had increased uptake at the interface between bone and the locking screw of an intramedullary nail; one patient (Case 7), who had been operated on six weeks earlier for the removal of a hip prosthesis, had increased uptake but was found to be free of infection at the time of reoperation one week after scanning; and a third patient (Case 60) had slightly increased uptake in the distal aspect of the femur four months after surgery.

Two of the four false-positive findings were in a small subgroup of ten patients who had had an operation at the site within six months before the positron emission tomography study. In contrast, there were only two false-positive findings among the remaining fifty patients. This difference, however, was not significant (p = 0.25).

Fifty-nine scans were interpreted similarly by both readers (kappa = 0.97). The remaining scan was interpreted differently by the two readers, who disagreed as to the location of the infection (actual location, the first metatarsal joint).

Representative images are shown in Figures 1-A, 1-B, 2, 3-A, and 3-B.

Discussion

Introduction | Materials and Methods | Results | Discussion | References

Recently, three studies were performed to evaluate the usefulness of fluorine-18 fluorodeoxyglucose-positron emission tomography for the diagnosis of musculoskeletal infections^18,33,36.

Sugawara et al.³³ reported on a small series of eleven patients who had a known or suspected infection. On the basis of the final clinical diagnosis, fluorine-18 fluorodeoxyglucose-positron emission tomography was found to have correctly identified the presence or absence of infection in ten of the eleven patients. In one patient, who had diabetes and an increased serum-glucose level, the infectious focus was missed because of suboptimal image quality.

Guhlmann et al.¹⁸ reported on fifty-one patients in whom fluorodeoxyglucose-positron emission tomography and antigranulocyte antibody scintigraphy were used for the evaluation of suspected chronic osteomyelitis involving the peripheral skeleton (thirty-six patients) or the central skeleton (fifteen patients). Patients who had been operated on within the previous two years were excluded. The scans were evaluated in a blinded fashion by two readers. The two readers found that both techniques had an excellent degree of accuracy (97% and 95% for fluorodeoxyglucose-positron emission tomography and 86% and 92% for antigranulocyte antibody scintigraphy) when they were used to evaluate suspected lesions in the peripheral skeleton. However, the accuracy of fluorodeoxyglucose-positron emission tomography (93% and 100%) was found to be significantly higher (p < 0.05) than that of antigranulocyte antibody scintigraphy (73% and 80%) when the methods were used to evaluate suspected lesions in the central skeleton. In that prospective study, the presence or absence of infection was determined with surgical exploration in thirty-one patients and with clinical follow-up in twenty patients.

Kälicke et al.³⁶ reported on fifteen patients who had histologically confirmed chronic osteomyelitis (eight patients) or acute osteomyelitis (seven patients). Fluorodeoxyglucose-positron emission tomography yielded fifteen true-positive findings. However, the absence of negative findings in that series may raise questions concerning the selection criteria used.

In our early experience⁴¹, we found that fluorodeoxyglucose-positron emission tomography alone is more accurate than the combination of bone-scanning and white blood-cell scanning for the diagnosis of chronic musculoskeletal infections, especially those involving the central skeleton. In that study, two readers evaluated the scans of thirty-four patients who had a suspected infection involving the central skeleton (seventeen patients) or the peripheral skeleton (seventeen patients). The combination of bone-scanning and white blood-cell scanning was found to have an accuracy of 71% (reader 1) and 76% (reader 2) for the diagnosis of lesions in the central skeleton and an accuracy of 88% (both readers) for the diagnosis of lesions in the peripheral skeleton. The corresponding values for fluorodeoxyglucose-positron emission tomography were 94% (both readers) for lesions in the central skeleton and 94% (both readers) for those in the peripheral skeleton. The presence or absence of infection was determined with surgical exploration in nine patients and with clinical follow-up in twenty-five.

In concordance with the results reported by Guhlmann et al.¹⁸, the current study suggests that fluorodeoxyglucose-positron emission tomography is a highly accurate imaging technique for the evaluation of musculoskeletal infections. The advantages and disadvantages of fluorodeoxyglucose-positron emission tomography are summarized in Table III. Despite the absence of attenuation correction, the sensitivity in the present series was comparable with that reported by Guhlmann et al. However, the specificity in our series tended to be slightly lower. This finding was probably related to differences in patient selection: the prevalence of musculoskeletal infection in our series (twenty-five [42%] of sixty) was lower than that in the series of Guhlmann et al. (twenty-eight [55%] of fifty-one), and we did not exclude patients who had had an operation within the previous two years.

Both the study by Guhlmann et al.¹⁸ and the present study were performed to evaluate the usefulness of fluorodeoxyglucose-positron emission tomography in the diagnosis of chronic musculoskeletal infections. Except for the seven cases of acute osteomyelitis in the study by Kälicke et al.³⁶, data on the usefulness of fluorodeoxyglucose-positron emission tomography in the detection of acute musculoskeletal infections have been limited to those from experimental animal models. Sugawara et al.³³ recently showed that, two to four days after Escherichia coli inoculation, fluorodeoxyglucose uptake was highly elevated in the calf muscle of rats. The uptake of fluorodeoxyglucose increased as soon as activated macrophages and leukocytes accumulated³³. Larger series are needed to define the role of fluorodeoxyglucose-positron emission tomography in the evaluation of acute osteomyelitis. However, as accurate and less expensive techniques for the detection of acute osteomyelitis (such as magnetic resonance imaging and three-phase bone scintigraphy) are available, the added value of fluorodeoxyglucose-positron emission tomography is likely to be limited.

In concordance with the lower specificity that has been reported in association with the use of fluorodeoxyglucose-positron emission tomography for the evaluation of oncological patients in the early postoperative period⁴², it may be assumed that postoperative repair mechanisms can induce false-positive fluorodeoxyglucose uptake. Repair mechanisms after surgery are characterized by granulation tissue, which may be responsible for increased fluorodeoxyglucose uptake³¹, thus mimicking infection on positron emission tomography scanning. Since two of the four false-positive findings in our study occurred in the small subgroup of ten patients who had had surgery within the previous six months, larger series of patients who have had a recent operation may be needed to define the appropriate timing of fluorodeoxyglucose-positron emission tomography scanning.

As both malignant and infectious lesions are fluorodeoxyglucose avid, differentiation between a tumor and an infection is not possible¹⁸. Therefore, we do not use this technique to distinguish between infection and malignancy.

Although there were no patients with pseudarthroses or recent fractures in our series, four patients (Cases 35, 48, 50, and 53) had sustained a fracture relatively recently (within the previous four to twelve months). None of the patients had false-positive uptake at the fracture site. Kälicke et al.³⁶ recently demonstrated that fluorodeoxyglucose uptake at the sites of fractures and pseudarthroses is significantly lower than it is at the sites of infections, thereby facilitating differentiation.

Seventeen patients (Cases 3, 4, 8, 11, 17, 20, 22, 24, 25, 31, 43, 44, 51, 52, 55, 56, and 57) were evaluated for a suspected periprosthetic infection, and our preliminary results regarding the diagnosis of this type of infection are promising. There was, however, one false-positive finding in our study, in a patient with aseptic loosening of a total knee replacement. The cell-rich vascular areas of the interface tissue between implant and bone and the pseudocapsule around aseptically loose implants contain more activated macrophages and proliferating fibroblast-like cells than do the tissues around well-fixed implants^43-47. This factor may explain the increased uptake of fluorodeoxyglucose in our patient. Therefore, it remains questionable whether fluorodeoxyglucose-positron emission tomography is able to differentiate between aseptic and septic loosening with a high degree of accuracy. Larger series of patients with surgically confirmed infections are needed before the routine clinical use of this method can be recommended for the evaluation of the especially difficult area around a loose total joint prosthesis.

False-negative findings due to elevated blood-glucose concentrations (particularly those higher than 11 mmol/L) are a well-known problem associated with fluorodeoxyglucose-positron emission tomography scanning for tumors. However, recent findings suggest that hyperglycemia does not influence fluorodeoxyglucose uptake (and therefore does not influence sensitivity) when this method is used to study infectious processes⁴⁸. Until this relationship is clarified in larger series, we recommend stringent regulation of the serum-glucose concentration so that it does not exceed 11 mmol/L, as is the current practice in oncological fluorodeoxyglucose-positron emission tomography scanning. If necessary, rapid control of the blood-glucose level can be achieved with the subcutaneous administration of a glycemia-adjusted dose of rapid-acting insulin two hours prior to fluorodeoxyglucose injection.

Study Limitations

Although these results are promising and are in concordance with previously reported data^18,34,36,38, one must be aware that, because of ethical limitations (surgery was performed only when clinically indicated), histopathological and microbiological confirmation of the diagnosis was obtained for only eighteen of the sixty patients. In the other forty-two patients, the determination of whether an infection was indeed present was based on clinical findings after a minimum duration of follow-up of six months.

Thirty-three patients were receiving antibiotics at the time of the study, which might have negatively influenced the results of microbiological tests^2,49.

The results concerning the diagnosis of infections around musculoskeletal prostheses are promising, but the presence of one false-positive finding in this limited subgroup of seventeen patients has to be stressed. Larger series are needed to assess the accuracy of fluorine-18 fluorodeoxyglucose-positron emission tomography in this subpopulation, in whom accurate assessment is particularly difficult.

In conclusion, fluorine-18 fluorodeoxyglucose-positron emission tomography is a highly accurate technique for the diagnosis of chronic musculoskeletal infections in routine clinical practice. It has the advantage of combining an extremely high negative predictive value with a very good specificity without the need for additional imaging. If this technique becomes more available through a wider distribution of positron emission tomography and coincidence cameras and improved distribution of fluorodeoxyglucose, it has the potential to become a standard technique for the detection of musculoskeletal infections because it is simple and highly accurate. This is especially true with regard to chronic infections in the central skeleton, where fluorine-18 fluorodeoxyglucose-positron emission tomography can replace a combination of several other imaging techniques. Larger studies, however, are needed to assess its value for the diagnosis of infections in the first six months after surgery, for the detection of acute musculoskeletal infections, and especially for the evaluation of infections at the sites of total joint prostheses.

References

Introduction | Materials and Methods | Results | Discussion | References

Segreti J; Nelson JA; and Trenholme GM: Prolonged suppressive antibiotic therapy for infected orthopaedic prostheses. Clin Infect Dis,1998.27: 711-3, 27711 1998 [PubMed]

Spangehl MJ; Younger AS; Masri BA; and Duncan CP: Instructional Course Lecture, American Academy of Orthopaedic Surgeons. Diagnosis of infection following total hip arthroplasty. J Bone Joint Surg Am,1998.79: 1578-88, 791578 1998

Zimmerli W: Role of antibiotics in the treatment of infected joint prosthesis. Orthopade,1995.24: 308-13, German24308 1995 [PubMed]

Perry M: Erythrocyte sedimentation rate and C reactive protein in the assessment of suspected bone infection—are they reliable indices. J R Coll Surg Edinb,1996.41: 116-8, 41116 1996

Sanzen L, and Sundberg M: Periprosthetic low-grade hip infections. Erythrocyte sedimentation rate and C-reactive protein in 23 cases. Acta Orthop Scand,1997.68: 461-5, 68461 1997 [PubMed]

Shih LY; Wu JJ; and Yang DJ: Erythrocyte sedimentation rate and C-reactive protein values in patients with total hip arthroplasty. Clin Orthop,1987.225: 238-46, 225238 1987 [PubMed]

Ma LD; Frassica FJ; Bluemke DA; and Fishman EK: CT and MRI evaluation of musculoskeletal infection. Crit Rev Diagn Imaging,1997.38: 535-68, 38535 1997 [PubMed]

Crim JT, and Seeger LL: Imaging evaluation of osteomyelitis. Crit Rev Diagn Imaging,1994.35: 201-56, 35201 1994 [PubMed]

Seabold JE, and Nepola JV: Imaging techniques for evaluation of postoperative orthopedic infections. Q J Nucl Med,1999.43: 21-8, 4321 1999 [PubMed]

Morrison WB; Schweitzer ME; Bock GW; Mitchell DG; Hume EL; Pathria MN; and Resnick D: Diagnosis of osteomyelitis: utility of fat-suppressed contrast-enhanced MR imaging. Radiology,1993.189: 251-7, 189251 1993 [PubMed]

White LM; Kim JK; Mehta M; Merchant N; Schweitzer ME; Morrison WB; Hutchison CR; and Gross AE: Complications of total hip arthroplasty: MR imaging—initial experience. Radiology,2000.215: 254-62, 215254 2000 [PubMed]

Becker W: The contribution of nuclear medicine to the patient with infection. Eur J Nucl Med,1995.22: 1195-1211, 221195 1995 [PubMed]

Datz FL: Indium-111-labeled leukocytes for the detection of infection: current status. Semin Nucl Med,1994.24: 92-109, 2492 1994 [PubMed]

Kaim A; Maurer T; Ochsner P; Jundt G; Kirsch E; and Mueller-Brand J: Chronic complicated osteomyelitis of the appendicular skeleton: diagnosis with technetium-99m labelled monoclonal antigranulocyte antibody-immunoscintigraphy. Eur J Nucl Med,1997.24: 732-8, 24732 1997 [PubMed]

Krznaric E; Roo MD; Verbruggen A; Stuyck J; and Mortelmans L: Chronic osteomyelitis: diagnosis with technetium 99m-d, l-hexamethylpropylene amine oxime labelled leukocytes. Eur J Nucl Med,1996.23: 792-7, 23792 1996 [PubMed]

Palestro CJ, and Torres MA: Radionuclide imaging in orthopaedic infections. Semin Nucl Med,1997.27: 334-45, 27334 1997 [PubMed]

Peters AM: The use of nuclear medicine in infections. Br J Radiology,1998.71: 252-61, 71252 1998

Guhlmann A; Brecht-Krauss D; Suger G; Glatting G; Kotzerke J; Kinzl L; and Reske SN: Fluorine-18-FDG PET and technetium-99m antigranulocyte antibody scintigraphy in chronic osteomyelitis. J Nucl Med,1998.39: 2145-52, 392145 1998 [PubMed]

Palestro CJ; Kim CK; Swyer AJ; Capozzi JD; Solomon RW; and Goldsmith SJ: Total-hip arthroplasty: periprosthetic indium-111-labeled leukocyte activity and complementary technetium-99m-sulfur colloid imaging in suspected infection. J Nucl Med,1990.31: 1950-5, 311950 1990 [PubMed]

Palestro CJ; Kim CK; Swyer AJ; Vallabhajosula S; and Goldsmith SJ: Radionuclide diagnosis of vertebral osteomyelitis: indium-111-leukocyte and technetium-99m-methylene diphosphonate bone scintigraphy. J Nucl Med,1991.32: 1861-5, 321861 1991 [PubMed]

Kaim A; Ochsner P; Maurer T; Jundt G; and Mueller-Brand J: Ectopic hematopoietic bone marrow in the appendicular skeleton after trauma. J Nucl Med,1998.39: 1980-3, 391980 1998 [PubMed]

Papos M; Barat F; Narai G; Dillmann J; Lang J; and Csernay L: Tc-99m HMPAO leukocyte and Tc-99m nanocolloid scintigraphy in posttraumatic bone in fection. Clin Nucl Med,1998.23: 423-8, 23423 1998 [PubMed]

Palestro CJ; Roumanas P; Swyer AJ; Kim CK; and Goldsmith SJ: Diagnosis of musculoskeletal infection using combined In-111 labeled leukocyte and Tc-99m SC marrow imaging. Clin Nucl Med,1992.17: 269-73, 17269 1992 [PubMed]

Palestro CJ; Swyer AJ; Kim CK; and Goldsmith SJ: Infected knee prosthesis: diagnosis with In-111 leukocyte, Tc-99m sulfur colloid, and Tc-99m MDP imaging. Radiology,1991.179: 645-8, 179645 1991 [PubMed]

Delbeke D: Oncological applications of FDG PET imaging. J Nucl Med,1999.40: 1706-15, 401706 1999 [PubMed]

Maddahi J; Blitz A; Phelps M; and Laks H: The use of positron emission tomography imaging in the management of patients with ischemic cardiomyopathy. Adv Card Surg,1996.7: 163-88, 7163 1996 [PubMed]

Salanova V; Markland O; and Woth R: Longitudinal follow-up in 145 patients with medically refractory temporal lobe epilepsy treated surgically between 1984 and 1995. Epilepsia,1999.40: 1417-23, 401417 1999 [PubMed]

Won HJ; Chang KH; Cheon JE; Kim HD; Lee DS; Han MH; Kim IO; Lee SK; and Chung CK: Comparison of MR imaging with PET and ictal SPECT in 118 patients with intractable epilepsy. AJNR Am J Neuroradiol,1999.20: 593-9, 20593 1999 [PubMed]

Brudin LH; Valind SO; Rhodes CG; Pantin CF; Sweatman M; Jones T; and Hughes JM: Fluorine-18 deoxyglucose uptake in sarcoidosis measured with positron emission tomography. Eur J Nucl Med,1994.21: 297-305, 21297 1994 [PubMed]

Fantone JC, and Ward PA: Role of oxygen-derived free radicals and metabolites in leukocyte-dependent inflammatory reactions. Am J Pathol,1982.3: 395-418, 3395 1982

Kubota R; Yamada S; Kubota K; Ishiwata K; Tamahashi N; and Ido T: Intratumoral distribution of fluorine-18-fluorodeoxyglucose in vivo: high accumulation in macrophages and granulation tissues studied by microautoradiography. J Nucl Med,1992.36: 1972-80, 361972 1992

Yamada S; Kubota K; Kubota R; Ido T; and Tamahashi N: High accumulation of fluorine-18-fluorodeoxyglucose in turpentine-induced inflammatory tissue. J Nucl Med,1995.7: 1301-6, 71301 1995

Sugawara Y; Gutowski TD; Fisher SJ; Brown RS; and Wahl RL: Uptake of positron emission tomography tracers in experimental bacterial infections: a comparative biodistribution study of radiolabeled FDG, thymidine, L-methionine, 67Ga-citrate, and 125I-HSA. Eur J Nucl Med. ,1999.26: 333-41, 26333 1999 [PubMed]

Guhlmann A; Brecht-Krauss D; Suger G; Glatting G; Kotzerke J; Kinzl L; and Reske SN: Chronic osteomyelitis: detection with FDG PET and correlation with histopathologic findings. Radiology,1998.206: 749-54, 206749 1998 [PubMed]

Ichiya Y; Kuwabara Y; Sasaki M; Yoshida T; Akashi Y; Murayama S; Nakamura K; Fukumura T; and Masuda K: FDG-PET in infectious lesions: the detection and assessment of lesion activity. Ann Nucl Med,1996.10: 185-91, 10185 1996 [PubMed]

Kälicke T; Schmitz A; Risse JH; Arens S; Keller E; Hansis M; Schmitt O; Biersack HJ; and Grünwald F: Fluorine-18 fluorodeoxyglucose PET in infectious bone diseases: results of histologically confirmed cases. Eur J Nucl Med,2000.27: 524-8, 27524 2000 [PubMed]

Sasaki M; Ichiya Y; Kuwabara Y; Otsuka M; Tahara T; Fukumura T; Gunasekera R; and Masuda K: Ringlike uptake of [18F]FDG in brain abscess: a PET study. J Comput Assist Tomogr,1990.14: 486-7, 14486 1990 [PubMed]

Sugawara Y; Braun DK; Kison PV; Russo JE; Zasadny KR; and Wahl RL: Rapid detection of human infections with fluorine-18 fluorodeoxyglucose and positron emission tomography: preliminary results. Eur J Nucl Med,1998.25: 1238-43, 251238 1998 [PubMed]

Tahara T; Ichiya Y; Kuwabara Y; Otsuka M; Miyake Y; Gunasekera R; and Masuda K: High [18F]-fluorodeoxyglucose uptake in abdominal abscesses: a PET study. J Comput Assist Tomogr,1989.13: 829-31, 13829 1989 [PubMed]

Mader JT; Norden C; Nelson JD; and Calandra GB: Evaluation of anti-infective drugs for the treatment of osteomyelitis in adults. Infectious Diseases Society of America and the Food and Drug Administration. Clin Infect Dis,1992.15 Suppl 1: 155-61, 15 Suppl 1155 1992

De Winter F; Van de Wiele C; Vogelaers D; Verdonk R; De Smet K; De Clercq D; and Dierckx R: FDG-PET as a single technique is more accurate than the combination of bone scan and white blood cell scan in chronic orthopedic infections. J Nucl Med,2000.Suppl: 15, Suppl15 2000

Bakheet SM, and Powe J: Benign causes of 18-FDG uptake on whole body imaging. Semin Nucl Med,1998.28: 352-8, 28352 1998 [PubMed]

Chianelli M; Mather SJ; Martin-Comin J; and Signore A: Radiopharmaceuticals for the study of inflammatory processes: a review. Nucl Med Commun,1997.18: 437-55, 18437 1997 [PubMed]

Moreschini O; Fiorito S; Magrini L; Margheritini F; and Romanini L: Markers of connective tissue activation in aseptic hip prosthetic loosening. J Arthroplasty,1997.12: 695-703, 12695 1997 [PubMed]

Santavirta S; Xu JW; Hietanen J; Ceponis A; Sorsa T; Kontio R; and Konttinen YT: Activation of periprosthetic connective tissue in aseptic loosening of total hip replacements. Clin Orthop,1998.352: 16-24, 35216 1998 [PubMed]

Toumbis CA; Kronick JL; Wooley PH; and Nasser S: Total joint arthroplasty and the immune response. Semin Arthritis Rheum,1997.27: 44-7, 2744 1997 [PubMed]

Wooley PH; Petersen S; Song Z; and Nasser S: Cellular immune response to orthopaedic implant materials following cemented total joint replacement. J Orthop Res,1997.15: 874-80, 15874 1997 [PubMed]

Zhuang HM; Loman JC; Cortes-Blanco A; Martinez-Lazaro R; Lee JH; Pourdehnad M; Niu ZT; Atochina O; Rossman MD; and Alavi AA: Hyperglycemia does not adversely affect FDG uptake by inflammatory and infectious lesions in FDG PET imaging. J Nucl Med,2000.Suppl: 321, Suppl321 2000

Lonner JH; Desai P; Dicesare PE; Steiner G; and Zuckerman JD: The reliability of analysis of intraoperative frozen sections for identifying active infection during revision hip or knee arthroplasty. J Bone Joint Surg Am,1996.78: 1553-8, 781553 1996 [PubMed]

Topics

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1-B:: Corresponding anteroposterior and lateral radiographs.

View Large | Download Slide(.ppt)
Add to My JBJS

View Larger

TABLE I: Data on Patients with a Suspected Infection of the Central Skeleton

Case	Location of Infection*	Time Since Prev. Op.	Metallic Implants	Score (Reader 1/ Reader 2)	Imaging Technique(s)†	Type of Diagnosis	Final Diagnosis	Accuracy‡ (Reader 1/Reader 2)
?1	Spine	—	No	+/+	—	Microbiological	Spondylodiscitis	TP/TP
?2	Sacrum	—	No	+/+	—	Microbiological	Osteomyelitis#	TP/TP
?3	THR	12 mos	Yes	+/+	WBC, colloidscans; arthrography	Clinical	Infection	TP/TP
?4	TSR	3 yrs	Yes	—/—	Bone, WBCscans	Clinical	No infection	TN/TN
?5	Sacroiliac joint	—	No	—/—	Bone, WBC scans; MRI	Clinical	No infection	TN/TN
?6	Spine	4 yrs	Yes	—/—	Bone scan	Clinical	No infection	TN/TN
?7	Prox. femur	6 wks	No	+/+	—	Microbiological	No infection	FP/FP
?8	TSR	19 mos	Yes	—/—	—	Clinical	No infection	TN/TN
?9	Spine	—	No	—/—	Radiograph, MRI	Clinical	No infection, severe arthrosis deformans	TN/TN
10	Prox. femur	16 mos	Yes	+/+	—	Microbiological	No infection, screw insertion site in bone	FP/FP
11	TSR	14 mos	Yes	+/+	—	Microbiological	Infection	TP/TP
12	Sacrum	12 mos	No	+/+	—	Microbiological	Osteomyelitis	TP/TP
13	Prox. femur	4 yrs	No	—/—	Bone scan	Clinical	No infection	TN/TN
14	Prox. femur	8 mos	Yes	+/+	—	Microbiological	Osteomyelitis	TP/TP
15	Spine	10 mos	Yes	—/—	—	Microbiological	No infection	TN/TN
16	Clavicle	4 yrs	No	—/—	Bone, WBC scans	Clinical	No infection	TN/TN
17	THR	10 mos	Yes	+/+	Bone, WBC, colloid scans	Clinical	Infection	TP/TP
18	Coccyx	20 yrs	No	+/+	MRI	Clinical	Osteomyelitis	TP/TP
19	Prox. femur	11 mos	No	+/+	—	Microbiological	Osteomyelitis	TP/TP
20	THR	15 mos	Yes	+/+	WBC, colloid scans	Clinical	Infection	TP/TP
21	Spine	—	No	—/—	MRI	Clinical	Soft-tissue infection	TN/TN
22	THR	18 yrs	Yes	—/—	WBC, colloid scans	Clinical	No infection	TN/TN
23	Spine	9 mos	Yes	+/+	—	Microbiological	Osteomyelitis	TP/TP
24	THR	22 yrs	Yes	—/—	WBC scan	Clinical	No infection	TN/TN
25	THR	11 mos	Yes	—/—	Bone scan	Clinical	No infection	TN/TN
26	Spine	4 mos	No	—/—	Bone scan	Clinical (with clinical evolution)	Superficial infection	TN/TN
27	Spine	6 yrs	Yes	—/—	Bone scan; CT	Clinical	No infection	TN/TN
28	Spine	—	No	—/—	MRI	Clinical	No infection	TN/TN
29	Spine	—	No	—/—	MRI	Clinical	No infection	TN/TN
30	Spine	10 mos	Yes	—/—	—	Microbiological	No infection	TN/TN
31	THR	13 mos	Yes	+/+	—	Microbiological	Infection	TP/TP
32	Prox. femur	17 yrs	No	+/+	—	Microbiological	Osteomyelitis	TP/TP
33	Spine	4 mos	No	—/—	MRI	Clinical	No infection	TN/TN

Add to My JBJS

TABLE I: Data on Patients with a Suspected Infection of the Central Skeleton

Case	Location of Infection*	Time Since Prev. Op.	Metallic Implants	Score (Reader 1/ Reader 2)	Imaging Technique(s)†	Type of Diagnosis	Final Diagnosis	Accuracy‡ (Reader 1/Reader 2)
?1	Spine	—	No	+/+	—	Microbiological	Spondylodiscitis	TP/TP
?2	Sacrum	—	No	+/+	—	Microbiological	Osteomyelitis#	TP/TP
?3	THR	12 mos	Yes	+/+	WBC, colloidscans; arthrography	Clinical	Infection	TP/TP
?4	TSR	3 yrs	Yes	—/—	Bone, WBCscans	Clinical	No infection	TN/TN
?5	Sacroiliac joint	—	No	—/—	Bone, WBC scans; MRI	Clinical	No infection	TN/TN
?6	Spine	4 yrs	Yes	—/—	Bone scan	Clinical	No infection	TN/TN
?7	Prox. femur	6 wks	No	+/+	—	Microbiological	No infection	FP/FP
?8	TSR	19 mos	Yes	—/—	—	Clinical	No infection	TN/TN
?9	Spine	—	No	—/—	Radiograph, MRI	Clinical	No infection, severe arthrosis deformans	TN/TN
10	Prox. femur	16 mos	Yes	+/+	—	Microbiological	No infection, screw insertion site in bone	FP/FP
11	TSR	14 mos	Yes	+/+	—	Microbiological	Infection	TP/TP
12	Sacrum	12 mos	No	+/+	—	Microbiological	Osteomyelitis	TP/TP
13	Prox. femur	4 yrs	No	—/—	Bone scan	Clinical	No infection	TN/TN
14	Prox. femur	8 mos	Yes	+/+	—	Microbiological	Osteomyelitis	TP/TP
15	Spine	10 mos	Yes	—/—	—	Microbiological	No infection	TN/TN
16	Clavicle	4 yrs	No	—/—	Bone, WBC scans	Clinical	No infection	TN/TN
17	THR	10 mos	Yes	+/+	Bone, WBC, colloid scans	Clinical	Infection	TP/TP
18	Coccyx	20 yrs	No	+/+	MRI	Clinical	Osteomyelitis	TP/TP
19	Prox. femur	11 mos	No	+/+	—	Microbiological	Osteomyelitis	TP/TP
20	THR	15 mos	Yes	+/+	WBC, colloid scans	Clinical	Infection	TP/TP
21	Spine	—	No	—/—	MRI	Clinical	Soft-tissue infection	TN/TN
22	THR	18 yrs	Yes	—/—	WBC, colloid scans	Clinical	No infection	TN/TN
23	Spine	9 mos	Yes	+/+	—	Microbiological	Osteomyelitis	TP/TP
24	THR	22 yrs	Yes	—/—	WBC scan	Clinical	No infection	TN/TN
25	THR	11 mos	Yes	—/—	Bone scan	Clinical	No infection	TN/TN
26	Spine	4 mos	No	—/—	Bone scan	Clinical (with clinical evolution)	Superficial infection	TN/TN
27	Spine	6 yrs	Yes	—/—	Bone scan; CT	Clinical	No infection	TN/TN
28	Spine	—	No	—/—	MRI	Clinical	No infection	TN/TN
29	Spine	—	No	—/—	MRI	Clinical	No infection	TN/TN
30	Spine	10 mos	Yes	—/—	—	Microbiological	No infection	TN/TN
31	THR	13 mos	Yes	+/+	—	Microbiological	Infection	TP/TP
32	Prox. femur	17 yrs	No	+/+	—	Microbiological	Osteomyelitis	TP/TP
33	Spine	4 mos	No	—/—	MRI	Clinical	No infection	TN/TN

View Larger

TABLE II: Data on Patients with a Suspected Infection of the Peripheral Skeleton

Case	Location of Infection*	Time Since Prev. Op.	Metallic Implants	Score (Reader 1/ Reader 2)	Imaging Technique(s)†	Type of Diagnosis	Final Diagnosis	Accuracy‡ (Reader 1/Reader 2)
34	Distal femur	9 yrs	No	—/—	—	Microbiological	No infection	TN/TN
35	Elbow	7 mos	No	—/—	WBC scan	Clinical	No infection	TN/TN
36	Distal femur	6 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
37	Tibia	6 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
38	Tibia	13 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
39	Tibia	24 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
40	Tibia	30 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
41	Tibia	24 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
42	Tibia	28 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
43	TKR	25 mos	Yes	—/—	Bone, WBC scans	Clinical (with clinical evolution)	No infection	TN/TN
44	TKR	29 mos	Yes	—/—	Bone, WBC, colloid scans	Clinical	No infection	TN/TN
45	Metatarsal 1	16 mos	No	+/0	MRI	Clinical	Osteomyelitis	TP/ind
46	Calcaneus	7 mos	No	—/—	Bone, WBC scans	Clinical	Soft-tissue infection	TN/TN
47	Phalanx	18 mos	No	+/+	Bone scan, radiograph	Microbiological	Brodie abscess	TP/TP
48	Tibia	6 mos	Yes	—/—	WBC scan	Clinical	No infection	TN/TN
49	Knee	12 mos	No	—/—	—	Microbiological	Bursitis	TN/TN
50	Tibia	12 mos	No	—/—	WBC scan	Clinical	No infection	TN/TN
51	TKR	4 yrs	Yes	+/+	—	Microbiological	Aseptic loosening	FP/FP
52	TKR	5 yrs	Yes	—/—	Bone, WBC scans	Clinical	No infection	TN/TN
53	Tibia	4 mos	Yes	—/—	Bone, WBC scans	Clinical	Soft-tissue infection	TN/TN
54	Calcaneus	5 mos	No	—/—	WBC scan	Clinical	No infection	TN/TN
55	TKR	18 mos	Yes	+/+	WBC scan	Clinical	Infection	TP/TP
56	TKR	25 mos	Yes	+/+	WBC scan	Clinical	Infection	TP/TP
57	TKR	7 mos	Yes	+/+	WBC scan	Clinical	Infection	TP/TP
58	Ankle	—	No	—/—	Bone, WBC scans	Clinical	No infection	TN/TN
59	Tibia	5 mos	Yes	+/+	—	Microbiological	Osteomyelitis	TP/TP
60	Distal femur	4 mos	No	+/+	—	Microbiological	No infection	FP/FP

Add to My JBJS

TABLE II: Data on Patients with a Suspected Infection of the Peripheral Skeleton

Case	Location of Infection*	Time Since Prev. Op.	Metallic Implants	Score (Reader 1/ Reader 2)	Imaging Technique(s)†	Type of Diagnosis	Final Diagnosis	Accuracy‡ (Reader 1/Reader 2)
34	Distal femur	9 yrs	No	—/—	—	Microbiological	No infection	TN/TN
35	Elbow	7 mos	No	—/—	WBC scan	Clinical	No infection	TN/TN
36	Distal femur	6 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
37	Tibia	6 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
38	Tibia	13 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
39	Tibia	24 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
40	Tibia	30 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
41	Tibia	24 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
42	Tibia	28 mos	Yes	+/+	Bone, WBC scans	Clinical	Osteomyelitis	TP/TP
43	TKR	25 mos	Yes	—/—	Bone, WBC scans	Clinical (with clinical evolution)	No infection	TN/TN
44	TKR	29 mos	Yes	—/—	Bone, WBC, colloid scans	Clinical	No infection	TN/TN
45	Metatarsal 1	16 mos	No	+/0	MRI	Clinical	Osteomyelitis	TP/ind
46	Calcaneus	7 mos	No	—/—	Bone, WBC scans	Clinical	Soft-tissue infection	TN/TN
47	Phalanx	18 mos	No	+/+	Bone scan, radiograph	Microbiological	Brodie abscess	TP/TP
48	Tibia	6 mos	Yes	—/—	WBC scan	Clinical	No infection	TN/TN
49	Knee	12 mos	No	—/—	—	Microbiological	Bursitis	TN/TN
50	Tibia	12 mos	No	—/—	WBC scan	Clinical	No infection	TN/TN
51	TKR	4 yrs	Yes	+/+	—	Microbiological	Aseptic loosening	FP/FP
52	TKR	5 yrs	Yes	—/—	Bone, WBC scans	Clinical	No infection	TN/TN
53	Tibia	4 mos	Yes	—/—	Bone, WBC scans	Clinical	Soft-tissue infection	TN/TN
54	Calcaneus	5 mos	No	—/—	WBC scan	Clinical	No infection	TN/TN
55	TKR	18 mos	Yes	+/+	WBC scan	Clinical	Infection	TP/TP
56	TKR	25 mos	Yes	+/+	WBC scan	Clinical	Infection	TP/TP
57	TKR	7 mos	Yes	+/+	WBC scan	Clinical	Infection	TP/TP
58	Ankle	—	No	—/—	Bone, WBC scans	Clinical	No infection	TN/TN
59	Tibia	5 mos	Yes	+/+	—	Microbiological	Osteomyelitis	TP/TP
60	Distal femur	4 mos	No	+/+	—	Microbiological	No infection	FP/FP

View Larger

TABLE III: Advantages and Disadvantages of Fluorine-18 Fluorodeoxyglucose-Positron Emission Tomography for the Diagnosis of Musculoskeletal Infections18,20,21,33,34,38,41,43

Advantages	Disadvantages
Early (1-hr) imaging	High cost
High target-to-background ratio	Possible lower sensitivity in diabetic patients
High resolution (±5 mm)	Patient must be sober for at least 4 hrs
High-count tomographic images	Technique is currently not widely available
Low bone and bone-marrow uptake	Differentiation between tumor and infection is not possible
Highly accurate in central skeleton
Not hindered by metal implants
No additional scans necessary, all-in-one technique
High interobserver agreement
Theoretically sensitive in low-grade infections
Use may be feasible in neutropenic patients

Add to My JBJS

TABLE III: Advantages and Disadvantages of Fluorine-18 Fluorodeoxyglucose-Positron Emission Tomography for the Diagnosis of Musculoskeletal Infections18,20,21,33,34,38,41,43

Advantages	Disadvantages
Early (1-hr) imaging	High cost
High target-to-background ratio	Possible lower sensitivity in diabetic patients
High resolution (±5 mm)	Patient must be sober for at least 4 hrs
High-count tomographic images	Technique is currently not widely available
Low bone and bone-marrow uptake	Differentiation between tumor and infection is not possible
Highly accurate in central skeleton
Not hindered by metal implants
No additional scans necessary, all-in-one technique
High interobserver agreement
Theoretically sensitive in low-grade infections
Use may be feasible in neutropenic patients