Imaging of Spondylodiscitis

Meera Raghavan, MD,* Elena Lazzeri, MD, PhD,† and Christopher J. Palestro, MD*

Spondylodiscitis is an infection of the vertebral body or disc and may also involve the epidural
space, posterior elements, and paraspinal soft tissues. It is a cause of morbidity and mortality,
and warrants early diagnosis and prompt treatment. Diagnosis can be difficult because of non-
specific signs and symptoms. Magnetic resonance imaging is sensitive and specific and is the
imaging modality of choice for spondylodiscitis. Gadolinium contrast can show the extent of soft
tissue and bone phlegmon and abscess. The test is less useful for evaluating treatment re-
sponse. When magnetic resonance imaging cannot be performed or is not diagnostic, radionuclide
imaging is a useful alternative. Although bone scintigraphy frequently is used as a screening test,
false-negative results can occur, especially in the elderly. This test is not useful for detecting soft
tissue infections that accompany or mimic spondylodiscitis. Gallium-67 citrate improves the speci-
ficity of the bone scan, can detect infection earlier than the bone scan, may be more sensitive,
especially in elderly patients, and identifies accompanying soft tissue infection. Performing SPECT
and SPECT/CT improves accuracy. The 2- to 3-day delay between radiopharmaceutical admin-
istration and the relatively poor image quality are significant disadvantages of gallium-67. Indium-
111 biotin, alone or in combination with streptavidin, accurately diagnoses spondylodiscitis;
unfortunately, this agent is not widely available. Currently, 18F-FDG imaging is the radionuclide
test of choice for spondylodiscitis. The procedure, which is completed in a single session, is sen-
sitive, has a high negative predictive value, and reliably differentiates degenerative from infectious
vertebral body end plate abnormalities. In comparative investigations, 18F-FDG has outper-
formed bone and gallium-67 imaging. Preliminary data suggest that 18F-FDG may be able to provide
an objective means to measure response to treatment. Gallium-68 citrate and 99mTc-radiolabeled
antimicrobial peptides have been investigated, but their role in spondylodiscitis has yet to be
established.
Semin Nucl Med 48:131–147 © 2017 Elsevier Inc. All rights reserved.

Spondylodiscitis etal infections.2 There is a 60% male predominance with mean

age of 66 years.3
S pondylodiscitis (SD), also known as spinal or vertebral
osteomyelitis or septic discitis, is an infection of the ver-
tebral body or disc. Infection may also involve the epidural
The majority of SD cases are pyogenic in origin, most com-
monly caused by Staphylococcus aureus (60%) followed by
Enterobacter species (30%). Less common, non-pyogenic agents
space, posterior elements, and paraspinal soft tissues.1 Pre-
include Myocobacterium tuberculosis, Brucella, fungi, and
existing conditions such as endocarditis, septic arthritis, urinary
parasites.4 SD can occur anywhere in the spinal column;
tract infections, and indwelling catheter infections can pre-
however, the most common location is the lumbar spine (60%)
dispose to SD. Other risk factors include advanced age, diabetes
followed by the thoracic (30%) and cervical spine (10%). In-
mellitus, coronary artery disease, spinal interventions, im-
volvement of a single spinal segment (which includes 2
munosuppression, and intravenous drug use.2 The rising
consecutive vertebral bodies and the intervening disc) occurs
incidence of SD is due in part to longer life expectancy, ma-
frequently (65%), followed by multilevel contiguous (20%)
lignancy, increased intravenous drug use, and spinal
and non-contiguous infection (10%).4,5
interventional procedures. SD accounts for about 1% of skel-
*Department of Radiology, Donald and Barbara Zucker School of Medicine Route of Infection
at Hofstra/Northwell, Hempstead, NY Hematogenous spread is the most common form of inocu-
†Regional Center of Nuclear Medicine, Pisa University Hospital, Pisa, Italy. lation, usually from a distant site of infection. This occurs
Address reprint requests to Meera Raghavan, MD, Department of Radiology,
Division of Nuclear Medicine and Molecular Imaging, Long Island Jewish
through the arterial blood supply or in a retrograde manner
Medical Center, 270-05 76th Ave, New Hyde Park, NY 11040. E-mail: through the venous system via the Batson lumbosacral plexus.
mraghavan@northwell.edu Other mechanisms include direct inoculation from surgery,

https://doi.org/10.1053/j.semnuclmed.2017.11.001 131
0001-2998/© 2017 Elsevier Inc. All rights reserved.
132 M. Raghavan et al.

Figure 1 Arterial anatomy of the lumbar vertebral body. Axial (A) and sagittal (B) diagrams demonstrate the paired seg-
mental arteries arising from the aorta. These arteries originate from the aorta, and course along the anterolateral surface
of the vertebral body. Segmental arteries give rise to paired metaphyseal arteries, which run parallel to the segmental
artery along the metaphyses. The segmental and metaphyseal arteries are connected by multiple periosteal arteries.
Metaphyseal and nutrient arteries (not shown) give rise to arteriolar anastomoses, which span multiple vertebral body
levels, and are most abundant at the end plates.

spinal procedures (such as lumbar puncture or discogra-
phy), or penetrating trauma. Contiguous spread is rare, but
can result from extension of an adjacent soft tissue infection
such as an infected aortic graft or retropharyngeal abscess.6
The vascular anatomy of the vertebral body is important
in understanding the typical spread of infection (Fig. 1). In
the lumbar spine, paired segmental arteries arise from the aorta,
and pass anterolaterally along the sides of the vertebral bodies.
Segmental arteries give rise to paired metaphyseal arteries,
which run parallel to the segmental artery along the me-
taphyses. The segmental and metaphyseal arteries are connected
by multiple periosteal arteries. The metaphyseal arteries give
rise to multiple branches; ascending branches anastomose with
the descending branches from the level above the end plates,
providing a rich anastomotic network in the end plates, similar
to that which is present in the metaphyses of long bones.4
In children, the network of arterial anastomoses extends into
the intervertebral disc; however, in adults, these end-
anastomoses regress and terminate at the end plates. In
children, therefore, SD typically begins in and usually is con-
fined to the disc, whereas in adults the infection usually begins
at the end plates. Infected microemboli lodge in the ante-
rior aspect of the end plate arterioles, producing infarction.
The infection then spreads in an anterior-to-posterior direction,7
sparing the relatively avascular intervertebral discs. The discs
usually become involved later in the course of infection because
of bacterial proteolytic enzymes. Approximately 95% of pyo-
genic SD cases involve the vertebral body, with only 5%
involving the posterior elements, because of the abundant
blood supply and presence of cellular bone marrow in the
vertebral body.8 Infection of the posterior elements in pyo-
genic SD is uncommon; it occurs more frequently in
tuberculous and fungal infections.
The venous blood supply is a less common, but equally
important mechanism for infection spread. Veins from the ver-
tebral bodies drain into the valveless Batson lumbosacral
plexus. Deep pelvic veins also drain into the Batson plexus,
allowing for retrograde spread of infection from abdominopelvic
organs, such as the urinary tract.9 Retrograde venous spread
is also postulated as the mechanism of tuberculous and fungal
SD.
Clinical Presentation
Back pain is the most common presenting symptom (86%).2
Fever can be present in up to 60% of patients; in a system-
atic review of 14 studies, however, fever was often absent at
presentation.2 The lack of fever in the setting of back pain
can mask the possibility of infection, delaying the diagnosis
from 11 to 59 days.2 Manifestations of neurologic deficits at
presentation, such as radiculopathy, limb weakness, or pa-
ralysis, can aid in the prompt and accurate diagnosis of SD,
but are present in only about 34% of cases.10 White blood
cell (WBC) count may be normal or elevated. Blood cul-
tures can be negative. Laboratory markers including erythrocyte
sedimentation rate and C-reactive protein (CRP) are sensi-
tive, but not specific. If elevated at the time of diagnosis, CRP
can be useful in monitoring response to treatment.
Because of its nonspecific clinical features, SD can be mis-
diagnosed as a degenerative process, leading to considerable
morbidity and mortality. Untreated infection can result in sig-
nificant spinal deformity (vertebral body collapse) and
neurologic deficits, hence the need for accurate and timely
diagnosis.
Radiologic Imaging
Radiography
Radiographs, which are almost always performed in the initial
workup of back pain, are not sensitive, especially early in the
course of the disease. Radiographs may be normal, or show
chronic degenerative changes, including loss of disc space
height and end plate definition (Fig. 2). A systemic review
Spondylodiscitis
Figure 2 Lateral radiograph of the lumbar spine in a patient with back
pain. There is mild disc space narrowing at L4-L5, and subtle ir-
regularity in the inferior end plate of L4 and the superior end plate
of L5 (arrows). Radiographic findings could easily be overlooked
or attributed to degenerative changes.
Figure 3 Thirty-year-old female with persistent back pain following multiple epidural steroid injections. Sagittal CT
images in bone (A) and soft tissue (B) windows show irregularity and fragmentation at L5-S1 with loss of the disc
space. Note the thickening of the anterior paraspinal soft tissues, consistent with phlegmon (B, white asterisk). Fungal
species was isolated on surgical specimen.
133
of 14 studies showed that radiographs performed for evalu-
ation of SD showed abnormalities in 89% of cases, whereas
another study revealed radiographic abnormalities in only
58%.2,8 Thickening of the paravertebral soft tissues may be
present, prompting additional cross-sectional imaging.
Computed Tomography
Because of its superior anatomic resolution, CT is more sen-
sitive than radiography. It is readily available, and can
demonstrate or confirm findings seen on radiographs. Loss
of disc height, end plate irregularity, and erosive changes are
often present (Fig. 3A). Loss of soft tissue planes and swell-
ing in the paravertebral soft tissues (Fig. 3B) can suggest
underlying abscess. Epidural extension of infection may also
be identified. CT is also helpful in confirming advanced de-
generative changes, or other bony causes of back pain such
as fracture or metastatic disease. Intravenous contrast sig-
nificantly aids in the evaluation of the soft tissues and in some
cases, can equal the diagnostic ability of magnetic reso-
nance imaging. CT is also useful in guiding percutaneous
biopsy and drainage. CT can be particularly useful in pa-
tients for whom MRI is contraindicated because of implanted
devices or if MRI is unavailable.
Magnetic Resonance Imaging
MRI is the imaging modality of choice for SD. MRI should
be performed when SD is suspected, as the nonspecific clini-
cal features can lead to delayed diagnosis and increased
morbidity and mortality. Advantages of MRI, in addition to
a lack of ionizing radiation, include multi-planar capability,
superior soft tissue contrast, and evaluation of the neural
134 M. Raghavan et al.

Figure 4 MRI images of patient in Figure 2. T1-weighted (A), T2-weighted with fat suppression (B), and post gadolinium T1 (C) images dem-
onstrate irregularity of the end plates of L4 and L5 and high T2 signal in the disc. Following gadolinium administration, there is diffuse
enhancement of the disc. Epidural abscess (red arrows) and intraosseous abscess (white asterisk) are identified on the postcontrast image
(C). Note the normal nuclear cleft, the linear low T2 linear signal in the non-infected L1-L2 disc (B, white arrow).

structures. MRI has a sensitivity of 97%, specificity of 93%,
and an accuracy of 94% for diagnosing SD.11-13 Standard
imaging protocols should include T1- and T2-weighted se-
quences, and gadolinium contrast should be administered.
T2 and post gadolinium T1-weighted sequences should also
be performed with fat suppression to increase the conspicuity
of findings.14,15 MRI allows for the evaluation of bone marrow
edema and disc space inflammation, as well as paraspinal and
epidural soft tissue involvement. The use of gadolinium allows
the differentiation of phlegmonous changes vs abscess
formation.
SD causes inflammatory exudate replacing normal marrow
with white cells as well as hyperemia.9 This results in hypo-
(low) or isointense T1 and hyperintense (high) or fluid T2
signal in the subchondral end plates and intervening disc
(Fig. 4).11,12,15 Signal changes are usually seen first in the an-
terior aspect of the vertebral body, involving a single spinal
segment. There is loss of end plate definition and dimin-
ished disc height. A positive nuclear cleft sign may be present.
The nuclear cleft is a T2 hypointense band in the nonin-
fected disc, which is effaced in the presence of infection
(Fig. 4B).12 However, the nuclear cleft sign is not specific for
SD as it may also be effaced in degenerative disc disease
(Fig. 4B).
Following gadolinium contrast administration, enhance-
ment of the subchondral bone or vertebral body can be seen,
and the affected disc enhances diffusely (Fig. 6C).16 If present,
abscess within the disc (Fig. 6D) or bone (Fig. 4B) can be
identified. The posterior elements of the spine are generally
not involved in pyogenic SD.
MRI can be limited in the evaluation of patients with spinal
hardware because of magnetic susceptibility artifact. The
artifact is amplified at higher field strengths because of
the ferromagnetic hardware. Titanium implants are
nonferromagnetic and produce fewer artifacts.17 The use of
specific imaging sequences and coils can be used to miti-
gate artifact; however, even with these techniques, evaluation
may still be limited.18
Epidural abscess extension (Figs. 4C and 6B) can cause com-
pression of the spinal cord resulting in a surgical emergency.
Soft tissue phlegmons reflect inflammation and hyperemia and
are not drainable.4 They are poorly defined areas of high T2
signal and diffuse enhancement, in contrast to abscesses, which
are T2 hyperintense, rim-enhancing, drainable collections
(Fig. 6C-E).
Modic changes refer to a spectrum of end plate signal
changes seen with degenerative disc disease. End plate marrow
signal changes are postulated to occur as a result of axial
loading and stress, producing various degrees of fibrovascu-
lar tissue and marrow replacement.19 Modic type I changes
are acute to subacute in nature, histologically reflecting dis-
ruption and fissuring of the end plates and formation of
fibrovascular tissue.19 Modic type I changes characterized by
low T1 and hyperintense T2 signals caused by hypervascularity
and marrow edema mimic the signal changes seen in SD.20
However, the key factor in distinguishing Modic type I changes
from SD is the T2 signal intensity in the intervening disc. In
SD, the disc shows high T2 signal, whereas in Modic type I
changes, the disc is usually low signal (Fig. 5). Severely de-
generated discs, however, may also show T2 hyperintensity.
Spondylodiscitis
Figure 5 Modic type 1 changes. T1-weighted (A) and T2-weighted fat-suppressed (B) MRI images. There is low T1
and high T2 signal at the L1-L2 end plates (arrows). Although the disc space is narrowed, it does not demonstrate a
high T2 signal. The vertebral end plates are smooth, not irregular. These findings support degenerative change, not
infection.
There are imaging features that may help to differentiate Modic
type 1 changes from infection. The “claw” sign refers to well-
defined linear areas of high signal in adjacent vertebral bodies
seen on diffusion-weighted images, which are thought to rep-
resent a physiologic response to axial loading and stress, and
are suggestive of Modic changes.21 Other imaging findings that
suggest Modic changes include lack of change over time,
absence of disc or end plate enhancement, and the presence
of vacuum changes in the disk. The “psoas sign,” hyperin-
tense T2 signal in the psoas muscle(s) on noncontrast MRI
examinations, identifies SD with 92% sensitivity and 92%
specificity.16 Distinguishing tuberculous SD from pyogenic SD
is important to institute appropriate treatment, thereby re-
ducing morbidity and mortality. Although most cases of SD
are bacterial, tuberculous spondylitis is a frequent cause of
SD in endemic areas, and is increasing in prevalence because
of the resurgence of tuberculosis during the past decade, es-
pecially in immunocompromised patients.
Although there is considerable overlap between pyogenic
and tuberculous SD, several imaging features can suggest a
diagnosis of tuberculous SD. Tuberculous SD more com-
monly affects the thoracic spine, usually with involvement
of the posterior elements.7,22 A lack of proteolytic enzymes
is commonly seen with Mycobacterium tuberculosis; this leads
to late or lack of involvement of the disc and greater degree
of soft tissue involvement.7 Well-defined, thin-walled ab-
scesses can be seen in the paraspinal and psoas muscles
(Fig. 6C-E). Infection usually spreads along the anterior lon-
gitudinal ligament in a craniocaudad direction, which can lead
to noncontiguous spinal segment involvement and skip lesions
(Fig. 6C).
135
Follow-up imaging is not routinely indicated, as there is
poor correlation between clinical response and improve-
ment on MRI. Symptoms and clinical parameters such as pain,
fever, and CRP levels are used to guide management. Kow-
alski et al23 demonstrated no association between follow-up
MRI findings and clinical status. Soft tissue and bony changes
on MRI can persist or even appear worse despite clinical im-
provement. MRI imaging follow-up is warranted in patients
with lack of clinical improvement 4-6 weeks after treat-
ment. Because of the more indolent nature of tuberculous SD
and higher incidence of spinal collapse and deformity, CT or
MRI can be useful in long-term follow-up.
Radionuclide Imaging
Single Photon Emitting Agents
99m
Tc-diphosphonates
Bone scintigraphy is performed with technetium-99m labeled
diphosphonates. These radiopharmaceuticals bind to the hy-
droxyapatite crystalline matrix of bone, and their uptake
depends on blood flow and rate of new bone formation. When
osteomyelitis is suspected, bone scintigraphy is often per-
formed as a triple or 3-phase bone scan. The first, or perfusion,
phase consists of a dynamic imaging sequence, followed im-
mediately by the second, or blood pool or tissue phase, in
which static images of the region(s) of interest are obtained.
The third, or skeletal, phase consists of images of the region(s)
of interest, acquired 2-4 hours after radiopharmaceutical in-
jection. Focal hyperperfusion, focal hyperemia, and focally
increased bony uptake are the classic presentation of
136 M. Raghavan et al.

Figure 6 Tuberculous spondylodiscitis. Sagittal T1 (A), T2 fat-suppressed (B), postcontrast (C), and axial post contrast (D, E) MRI. There is
decreased T1 and increased T2 signal at the L2-L3 level with high T2 signal in the disc. There is diffuse enhancement of the vertebral bodies
and disc (6C). Note the epidural abscess narrowing the spinal canal (6B, black arrow). There is non-contiguous spinal level involvement,
with abscess tracking superiorly along the anterior longitudinal ligament (6C, white arrow). Thin-walled, peripherally enhancing bilateral
psoas abscesses (6E, F, black arrows) and bone abscess (6D, black arrowheads) also are present.

osteomyelitis on 3-phase bone scintigraphy. The test is both
sensitive and specific for diagnosing osteomyelitis in bones
not affected by underlying conditions. Abnormalities on bone
scintigraphy reflect the rate of new bone formation in general,
and in the setting of preexisting conditions such as degen-
erative joint disease, fracture, and orthopedic hardware, because
of decreased specificity, the test is less useful.24
Adatepe et al25 reviewed the results of planar bone scin-
tigraphy performed on 12 patients with hematogenous
pyogenic vertebral osteomyelitis and reported that the test was
positive in all 12. Abnormal uptake was diffuse and in-
volved 1 vertebral body in 11 patients and 2 vertebrae in the
12th patient. The abnormal uptake extended into the sur-
rounding soft tissues in 2 patients with a paravertebral abscess.
Lisbona et al26 retrospectively reviewed the results of bone
scintigraphy in 21 cases of infectious spondylitis, including
6 cases of tuberculous spondylitis. The test was positive in
16 of 17 sites of nontuberculous infection. Bone scintigra-
phy was definitely abnormal in 6 of 9 sites of tuberculous
spondylitis and subtly abnormal in the other 3. None of the
8 paraspinal infections (6 non-tuberculous, 2 tuberculous)
were detected.
Gratz et al27 prospectively studied 30 patients, all of whom
had vertebral osteomyelitis at the time of imaging. They re-
ported that planar bone scintigraphy was 86% sensitive for
diagnosing spinal osteomyelitis. Performing SPECT in-
creased the sensitivity to 92%. Love et al28 retrospectively
reviewed bone scintigraphy performed on 22 patients with
Spondylodiscitis
Figure 7 L3-L4 spondylodiscitis. There is hyperperfusion, hyperemia, and increased bony uptake in the lower lumbar
spine on this 3-phase bone scan, a pattern that is specific, but not sensitive for spondylodiscitis.
suspected SD. Planar imaging was 73% sensitive, 31% spe-
cific, and 50% accurate for infection. Sensitivity increased to
82%, whereas specificity decreased to 23% with SPECT. Ac-
curacy was unchanged at 50%. When the SPECT images were
interpreted alone, the test was 73% sensitive, 69% specific,
and 71% accurate. The presence of abnormal uptake in 2 con-
tiguous vertebrae on SPECT was the single most accurate
criterion (71%) for detecting spinal osteomyelitis in this in-
vestigation. None of the 11 sites of extraosseous infection were
identified on planar images or on SPECT.
Performing the test as a 3-phase bone scan does not improve
accuracy. Gratz et al27 found that 3-phase bone imaging was
positive in patients with severe infection, but not in indi-
viduals with mild or moderate infection. Love et al28 reported
that although specificity improved with the 3-phase tech-
nique, it did so at the expense of sensitivity, which decreased
from 92% to 36% (Fig. 7).
Gallium-67
Several factors contribute to gallium-67 (67Ga) uptake in in-
fection. About 90% of circulating 67Ga is transferrin bound
in the plasma. Increased blood flow and vascular mem-
brane permeability result in increased delivery and
accumulation at infectious foci. 67Ga binds to lactoferrin, which
is present in high concentrations in sites of infection. Direct
bacterial uptake, complexing with siderophores, and leuko-
cyte transport also may contribute to 67Ga uptake in infection.
Imaging generally is performed 18-72 hours after injection.24
Hadjipavlou et al29 reviewed the results of 67Ga imaging per-
formed on 41 patients with suspected SD. The test was positive
in all 39 patients with infection and negative in the 2 pa-
tients without infection. Lisbona et al26 reported that 67Ga
scintigraphy was positive in all 17 sites of spinal infection as
well as in 6 paraspinal abscesses; bone scintigraphy was posi-
tive in 16 of the 17 sites of spinal infection and negative in
the paraspinal infections. Gratz et al,30 in a prospective in-
vestigation, reported sensitivity, specificity, and accuracy of
73%, 61%, and 80%, respectively, for diagnosing SD with 67Ga
planar plus SPECT imaging. 67Ga was less sensitive, but more
specific, than bone scintigraphy. The accuracy of the 2 tests
was the same: 80%.
Love et al28 retrospectively reviewed the results of planar
and SPECT 67Ga imaging performed on 22 patients with sus-
pected SD. 67 Ga images were interpreted alone and in
combination with bone scintigraphy. The results of com-
137
bined bone/67Ga SPECT and 67Ga SPECT were identical (91%
sensitivity, 92% specificity, and 92% accuracy) and were more
accurate than planar (50%) and SPECT (71%) bone, planar
67
Ga (79%), and combined planar bone/67Ga (75%) imaging.
Nine of 11 (82%) sites of extraosseous infection were iden-
tified on 67Ga imaging; none were seen on bone scintigraphy.
The authors concluded that 67Ga SPECT was sufficient for di-
agnosing SD, and bone scintigraphy was not necessary.
More recent investigations have focused the role of 67Ga
SPECT/CT for diagnosing SD31-34 (Figs. 8 and 9). Domin-
guez et al32 investigated 9 patients with SD who underwent
67
Ga SPECT/CT as well as planar and SPECT bone imaging.
67
Ga SPECT/CT was abnormal in all 9 patients and detected
adjacent soft tissue infection in 3 patients (Fig. 10). Bone scin-
tigraphy was abnormal in 8 of the 9 patients and detected
none of the soft tissue infections. Fuster et al33 prospec-
tively investigated 34 patients with combined planar bone
scintigraphy and 67Ga SPECT/CT. The sensitivity, specificity,
and accuracy of the combined test were 78%, 81%, and 79%,
respectively. The positive predictive value was 82%, and the
negative predictive value was 76%. 67Ga SPECT/CT de-
tected 10 of 12 cases of soft tissue involvement. Tamm et al34
retrospectively reviewed 34 patients who had undergone bone
SPECT/CT and 67Ga SPECT/CT for suspected SD. They re-
ported that combined bone/67Ga SPECT/CT detected 17 of
18 cases of SD and was negative in all 16 patients without
infection.
111
In-Biotin
Biotin, also known as vitamin B7 and vitamin H, plays a key
role in glucose metabolism including regulation of hepatic
glucose uptake, gluconeogenesis, lipogenesis, insulin recep-
tor transcription, and pancreatic β-cell function. Biotin also
is used as a growth factor by certain bacteria.35,36 In initial
studies, the 2-step streptavidin/111In-biotin imaging accu-
rately diagnosed various forms of osteomyelitis, including SD.
The predominant uptake mechanism at the site of infection
is related to nonspecific accumulation of streptavidin because
of locally increased vascular permeability. 111In-biotin accu-
mulates at the infected focus because of its high affinity for
streptavidin.37,38 Lazzeri et al36 prospectively studied 110 pa-
tients with 111In-biotin alone, without streptavidin pretreatment.
Seventy-one patients with suspected hematogenous spread
of infection and 39 patients with suspected postoperative in-
fection were included. Planar and SPECT imaging of the region
138 M. Raghavan et al.

Figure 8 Spondylodiscitis T12-L1. There are destructive vertebral body end plate changes (arrow) on the CT component of the examination
(left), with increased gallium-67 uptake in the same region. The infection does not extend to the hardware in the lower lumbar spine.

of interest were performed 2 hours after radiopharmaceuti-
cal administration. The overall sensitivity, specificity, and
accuracy of 111In-biotin were 93%, 90%, and 92%, respec-
tively. There were no significant differences in any of these
parameters between the 2 groups of patients. In a subse-
quent investigation, Lazzeri et al39 incorporated SPECT/CT
into their 111In-biotin imaging protocol. They found that al-
though there were no significant differences in sensitivity and
specificity between SPECT and SPECT/CT, SPECT/CT impacts
patient management by precisely localizing the infection and
facilitating selection of the most appropriate treatment (Fig. 11).
Advantages of 111In-biotin include same-day imaging, little
or no bone marrow uptake, and less patient radiation than
labeled leukocyte and 67Ga studies. Antibiotic therapy does
not appear to affect test sensitivity.40 Unfortunately, this agent
is available in only a few centers worldwide.
Labeled leukocytes
In vitro leukocyte (WBC) labeling usually is performed with
111
In oxyquinolone (In) or 99mTc-exametazime (Tc). Uptake
depends on intact chemotaxis, number and types of cells
labeled, and cellular response in a particular condition. A cir-
culating WBC count of at least 2000/µL is needed for
satisfactory image quality. The majority of WBCs labeled usually
are neutrophils, and the test is most sensitive for detecting
neutrophil-mediated infections.24
In contrast to other locations in the skeleton, labeled leu-
kocyte imaging is not useful for diagnosing SD. Although
increased uptake is virtually diagnostic, half of all cases present
as areas of decreased, or absent activity.41 The explanation for
this phenomenon is not known. It has been suggested that
the photopenic presentation is caused by occlusion of the mi-
crocirculation of the involved bone. Infection-induced death
of reticuloendothelial cells that normally accumulate labeled
leukocytes also may be involved. Marrow uptake in the normal
spine may be greater than in infected bone, masking the ab-
normality or causing a relative photopenia.42 Duration of
symptoms appears to affect the likelihood of a photopenic
presentation. Patients who are symptomatic for less than 1
month more frequently demonstrate increased uptake, whereas

Figure 9 Degenerative changes L5-S1. There are vertebral body end plate irregularities (arrow) on the CT component (left) of the examina-
tion. There is normal gallium-67 uptake in this region, however, making infection unlikely.
Spondylodiscitis 139

Figure 10 Spondylodiscitis and infected hardware with psoas abscesses. There is abnormal gallium-67 uptake in L4-
L5, which extends into both psoas muscles (arrows). Gallium-67 is superior to bone scintigraphy for detecting soft
tissue infections that often accompany spinal infections.

photopenia is more common in individuals who are symp-
tomatic for more than 1 month. Regardless of the explanation,
photopenia is associated with several noninfectious condi-
tions including previously treated SD, tumor, infarction,
compression fracture, and Paget disease24,41,42 (Fig. 12). The
photopenic presentation of SD on labeled leukocyte imaging
is not unique to the in vitro labeled leukocyte procedure. It
also has been observed with in vivo labeled leukocyte imaging
agents.43,44
Radiolabeled Antimicrobial Peptides
Antimicrobial peptides are an important component of the
natural defenses of most living organisms. They are small, cat-
ionic, and amphipathic (hydrophilic and hydrophobic), and
attack pathogens by multiple mechanisms. They exhibit broad-
spectrum activity against Gram-positive and Gram-negative
bacteria, yeasts, fungi, and enveloped viruses. They also are
involved in apoptosis, wound healing, and immune modu-
lation. Their expression may be constant or induced on contact
with microbes.45,46
Radiolabeled synthetic fragments of ubiquicidin, a murine
antimicrobial peptide present in the cytosolic fraction of mac-
rophages, have been investigated as infection-specific imaging
agents.47 Dillmann-Arroyo et al48 studied 99mTc-ubiquicidin-
29-41 in 27 patients with suspected SD. Twelve patients had
metallic hardware and 14 patients had previous spinal surgery.
Twenty (74%) patients had a final diagnosis of SD. Imaging
was performed approximately 2 hours after radiopharma-
ceutical administration. They reported that the test was 100%
sensitive and 87.5% specific for SD.
Positron Emitting Agents
Positron emitting radiopharmaceuticals have several advan-
tages over single photon emitting radiopharmaceuticals. PET
is intrinsically a high-resolution tomographic technique, which
facilitates the precise localization of radiopharmaceutical uptake,
especially when performed as PET/CT. Semiquantitative analy-
sis is more feasible with PET agents than with single photon
emitting agents, and this potentially could be used to differ-
entiate infectious from noninfectious conditions and to monitor
treatment response.
Fluorine-18-fluorodeoxyglucose
18
F-FDG is a structural analog of 2-deoxyglucose. Cellular
uptake of 18F-FDG is governed by several mechanisms: passive
diffusion, active transport via a Na1-dependent glucose trans-
porter (GLUT), and, most importantly, via GLUT-1 through
GLUT-13 transporters. After entering the cell, 18F-FDG is phos-
phorylated to 18F-2'-FDG-6 phosphate by the hexokinase
enzyme. In contrast to glucose-6-phosphate, 18F-2'-FDG-6

Figure 11 Surgical wound infection and osteomyelitis. There is abnormal 111In-biotin accumulation extending posteriorly from L3 into the
surgical wound (arrows), in a patient who had undergone removal of the intervertebral disc 3 months previously. Cultures grew Staphylo-
coccus aureus.
140
Figure 12 (A) T6 spondylodiscitis. There is an area of decreased labeled
leukocyte accumulation (arrow) in the mid-thoracic spine. (B) T10
compression fracture. There is an area of decreased labeled leuko-
cyte accumulation (arrow) in the lower thoracic spine, similar to
A. Although more than half of all cases of spondylodiscitis present
as decreased activity on labeled leukocyte images, this pattern is not
specific for infection.
phosphate is not a substrate for the enzymes of the glyco-
lytic pathway or the pentose-phosphate shunt. It is not
metabolized and does not diffuse back into the extracellular
space and is trapped intracellularly.49
Several factors influence 18F-FDG uptake in infection.
In activated inflammatory cells, such as neutrophils,
lymphocytes, monocytes, and macrophages, there are both
an increased number and an increased expression of glucose
transporters, as well as an increased affinity of these trans-
porters for deoxyglucose.50 Kubota et al51 demonstrated that
Figure 13 L3-L4 spondylodiscitis. There is intense uptake of 18F-FDG in the intervertebral L3-L4 disc space and body of L3, with extension
into the right ileo-psoas muscle. Staphylococcus aureus was the causative organism.
M. Raghavan et al.
uptake in tumor was due, at least in part, to newly formed
granulation tissue and activated macrophages associated with
tumor necrosis and growth. They observed that 18F-FDG
uptake by non-neoplastic cells was even higher than its uptake
in viable tumor cells. Histopathologic and autoradiographic
analysis of a turpentine inflammation model in rats showed
that uptake of 18F-FDG corresponded to areas of inflamma-
tory cell infiltration: neutrophils in the acute phase and
macrophages in the chronic phase.52 Kaim et al53 reported
similar results in a rodent abscess model. 18F-FDG uptake was
greatest in areas of inflammatory cell infiltrates, predomi-
nantly in neutrophils in the acute phase and in macrophages
in the chronic phase of the infection.
18
F-FDG is an attractive agent for imaging musculoskel-
etal infection for several reasons. It is a small molecule that
rapidly enters poorly perfused regions, and the procedure is
completed within hours rather than days. Normal bone marrow
uptake of this radiopharmaceutical is low, which may facili-
tate the differentiation of inflammatory cellular infiltrates from
marrow. 18FDG uptake in spinal infection is higher than normal
bone marrow uptake (Figs. 13 and 14). Degenerative bone
changes generally show only faintly increased uptake (Fig. 15).
18
F-FDG uptake in traumatized bone usually normalizes within
3-4 months.50
The role of 18F-FDG in the diagnosis of SD has been ex-
tensively investigated.43,54-60 Guhlmann et al55 reported that
18
F-FDG-PET was positive in all 3 cases of infection and nega-
tive in the 1 patient without infection. In another investigation,
Spondylodiscitis 141

Figure 14 Multilevel spondylodiscitis. There is intense 18F-FDG uptake in the T8-T9, T11-T12, and L3-L4 vertebrae. Staphylococcus aureus
was the causative organism.

Guhlmann et al43 reported that 18F-FDG-PET was signifi-
cantly more accurate than the radiolabeled antigranulocyte
antibody, 99mTc-besilesomab, for diagnosing SD. Schiesser et al56
reported that 18F-FDG-PET was true negative for infection in
6 patients with spinal hardware. Kalicke et al57 reported that
18
F-FDG-PET correctly identified all 7 patients with spinal
osteomyelitis. In another investigation, the test was positive
in 12 patients with spinal infection and negative in 3 of the
4 patients without infection.58 Meller et al59 prospectively in-
vestigated 8 patients suspected of having SD and reported that
18
F-FDG-PET was 100% accurate. Hungenbach et al60 per-
formed 18F-FDG-PET examinations on 42 patients including
13 who had previously undergone surgery. They evaluated
both intensity and patterns of 18F-FDG uptake. Using visual

Figure 15 Degenerative spine changes. There is mildly increased 18F-FDG uptake in the vertebral end plates of L5-S1. 18F-FDG uptake in de-
generative changes, as this case illustrates, usually is mild and less intense than uptake in infection.
142
analysis, they found that the test was 86% sensitive and 95%
specific for SD. The 3 false-negative results were attributed
to low-grade tuberculous infection and to reduced uptake in
poorly controlled diabetes. The 1 false-positive result was
because of a recent vertebral fracture. Hartman et al61 re-
ported that 18F-FDG-PET/CT was positive in all 7 patients with
spinal osteomyelitis, and negative in 2 patients without spinal
osteomyelitis. They found that the improved anatomic lo-
calization provided by PET/CT facilitated the differentiation
of soft tissue from bone involvement, which was helpful for
therapy planning.
Tuberculous SD, or Pott disease, can be difficult to diag-
nose because of nonspecific symptoms and similarities with
non-tuberculous spinal infections. Appropriate treatment
depends on accurate differentiation of tuberculous from
nontuberculous infections. The potential of 18F-FDG PET/
CT for differentiating tuberculous from pyogenic SD has been
studied by a few investigators, with variable results. Bassetti
et al62 performed 18F-FDG PET/CT on 30 patients with SD,
including 10 with tuberculous infection. Imaging was per-
formed about 1 hour after radiopharmaceutical administration.
They found that the maximum standardized uptake value (SUV
max) was significantly higher for tuberculous than for pyo-
genic infections. Kim et al63 performed dual time point
18
F-FDG-PET/CT studies on patients with SD, including 11
with pyogenic and 11 with tuberculous infection. Imaging
was performed approximately 1 and 2 hours after radiophar-
maceutical administration. Like Bassetti et al,62 they observed
that the SUV max, at 1 hour, was significantly higher for tu-
berculous than for nontuberculous infection. At 2 hours,
however, the differences were not significant and they con-
cluded that delayed imaging was not useful for differentiating
between the 2 entities. Gunes et al64 investigated 18F-FDG PET/
CT in 32 patients with SD, including 8 with tuberculous
infection. Nineteen of the patients underwent dual time point
imaging. Images were acquired at about 90 minutes and 2
and one-half hours after radiopharmaceutical administra-
tion. These investigators found no significant differences in
Figure 16 Uninfected spinal hardware. Focally increased 18F-FDG uptake is confined to the spinal hardware, probably related to reactive in-
flammation. The lack of uptake in the adjacent bone and surrounding soft tissues makes infection very unlikely.
M. Raghavan et al.
SUV max between pyogenic and tuberculous spondylitis on
early or delayed imaging, and concluded that it is not pos-
sible to differentiate between these 2 types of infection with
18
F-FDG PET/CT. It should be noted that the time intervals
between radiopharmaceutical administration and early and
delayed imaging were longer in this investigation than in those
of Bassetti et al62 and Kim et al,63 which could explain the
differences in the results obtained.
The diagnosis of SD in the patient with spinal hardware
is a diagnostic challenge. The presence of a foreign body can
incite an intense immune response, and increased 18F-FDG
uptake around uninfected spinal implants, in the absence of
infection, has been reported (Fig. 16). Using 18F-FDG PET,
de Winter et al65 reported the specificity of the test was con-
siderably lower in patients with, than in patients without,
implants (65% vs 92%). In subsequent investigations in which
PET/CT rather than PET was performed, the results have been
better. As part of a larger investigation, Hartman et al61 re-
ported that 18F-FDG PET/CT had a 100% accuracy in patients
with metallic implants. Inanami et al66 studied 14 patients
with spinal hardware, including 8 with suspected spinal
implant infection and 6 controls. Imaging was performed 5-33
weeks after surgery. Six of the 14 patients eventually were di-
agnosed with infection. These investigators reported that using
visual analysis, all 6 infected devices were positive and all 8
infected devices were negative (100% accuracy). Both the mean
and the range of SUV max values were significantly higher
in the infected than in the uninfected groups (9.0 and 5.5-
14.7 vs 3.3 and 2.0-4.3, respectively; p = 0.003).
Bagrosky et al67 retrospectively reviewed 18F-FDG PET/CT
studies performed on 20 patients, 6-31 years old, with pos-
terior spinal fusion rods. Six of the patients ultimately were
diagnosed with hardware infection. In all 6 patients there was
intense, confluent increased 18F-FDG uptake in soft tissue and
bone immediately adjacent to the hardware at multiple con-
tiguous levels. In 3 of the 6 patients, there was 18F-FDG uptake
at the bone-hardware interface of one or more interpedicular
screws. Nine patients had noninfectious hardware complica-
Spondylodiscitis
tions, primarily hardware loosening or bone remodeling, which
was characterized by focal 18F-FDG uptake adjacent to 1 or
2 hooks, screws, or anchors, usually at the upper or lower
aspects of the spinal hardware. The intense contiguous areas
of uptake in bone and soft tissue surrounding infected hard-
ware was not present in these patients. The patients in this
series underwent whole-body imaging, and alternative sources
of infection were identified in 7: pneumonia (n = 5), pyo-
nephrosis (n = 1) and superficial wound infection (n = 1).
18
F-FDG has outperformed single photon emitting
radiopharmaceuticals in comparative studies. Guhlmann et al43
reported that 18F-FDG PET was significantly more accurate
than 99mTc-besilesomab for diagnosing spinal osteomyelitis.
Gratz et al30 observed that 18F-FDG PET was superior to 67Ga
for detecting paraspinal soft tissue infection and was supe-
rior to bone scintigraphy for differentiating infection from
advanced arthritic changes. Fuster et al33 studied 34 pa-
tients, including 18 with SD. The sensitivity, specificity, and
accuracy of 18F-FDG PET/CT were 89%, 88%, and 88%, re-
spectively. The sensitivity, specificity, and accuracy of combined
planar bone/67Ga SPECT/CT imaging were 78%, 81%, and
79%, respectively.18F FDG-PET/CT identified soft tissue in-
fection in 12 patients; 67Ga SPECT/CT identified soft tissue
involvement in 10 patients.
18
F-FDG also has been compared with MRI. Gratz et al30
reported that18 FDG-PET was superior to MRI for detecting
low-grade spondylitis or discitis. Ohtori et al68 studied 18 pa-
tients with Modic type 1 changes on MRI, including 11 with
spinal osteomyelitis and found a 100% concordance between
18
F-FDG PET results and final diagnoses. Stumpe et al69 com-
pared 18F-FDG PET and MRI in 30 patients (38 sites) with
vertebral end plate abnormalities. 18F-FDG PET was true posi-
tive in all 5 infected sites and true negative in all 33 uninfected
sites (100% accuracy). The sensitivity and specificity of MRI
were 50% and 96%, respectively. Seifen et al70 analyzed 38
consecutive cases of suspected SD in patients with inconclu-
sive results on conventional imaging modalities. The sensitivity,
specificity, and accuracy of FDG-PET/CT were 81.8%, 100%,
89.5%, respectively; positive and negative predictive values
were 100% and 80%, respectively. Sensitivity, specificity, and
accuracy of MRI performed on 27 patients were 75%, 71.4%,
and 74.1%, respectively; the positive and negative predic-
tive values were 88.2% and 50%, respectively. Fuster et al71
compared 18F-FDG-PET/CT and MRI in 26 patients with clini-
cal symptoms of spinal infection. Sensitivity, specificity, and
positive and negative predictive values for 18F-FDG-PET/CT
were 83%, 88%, 94%, and 70%; for MRI they were 94%, 38%,
77%, and 75%, respectively. Accuracy of 84% for 18F-FDG
PET/CT and 81% for MRI was similar. MRI, however, was
superior for detecting soft tissue involvement.
Skanjeti et al72 reported that 18F-FDG PET had a sensitiv-
ity, specificity, and accuracy of 92.9%, 50%, and 75%,
respectively for SD. MRI had a sensitivity, specificity, and ac-
curacy of 100%, 50%, and 81.3%, respectively. These
investigators found that 18F-FDG-PET is a useful adjunct to
MRI and can be used when MRI is inconclusive or cannot
be performed. Ioannou et al73 studied 10 patients with brucellar
SD. 18F-FDG PET/CT identified all foci of infection seen on
143
MRI, identified additional spinal lesions in 3 patients, as well
as new paravertebral soft tissue involvement and epidural
masses. The additional information provided by 18F-FDG PET/
CT helped determine the duration of treatment. Nakahara
et al,74 in a study of 9 patients, reported that 18F-FDG PET/
CT was superior to MRI for localizing sites of active infection
and providing a guide for minimally invasive surgery.
18
F-FDG appears to be useful for monitoring response to
treatment in patients with SD (Figs. 17 and 18). Nanni et al75
evaluated the potential of 18F-FDG PET/CT for early evalu-
ation of response to antibiotic therapy in 34 patients with
hematogenous SD. They compared the SUV max of the base-
line study (SUV1) with the SUV max of the second study
(SUV2), which was performed 2-4 weeks after the start of
treatment, as well as the change in SUV max (delta-SUV max)
between the 2 studies. In responders, SUV2 was signifi-
cantly lower than SUV1, with sensitivity and specificity of 83%
and 46%, respectively. Using delta-SUV max, sensitivity and
specificity both were 82%. Kim et al76 found that the change
in SUV max between the initial and the follow-up studies was
the best predictor of residual infection. Ioannou et al73 re-
ported that successful treatment of brucellar SD was associated
with a significant decrease in SUV max values.
Riccio et al77 reported that evaluating uptake patterns was
more useful than quantification for differentiating patients with
active infection from those without active infection or from
those who had had a successful response to therapy. In their
investigation, patients with a poor response to treatment for
hematogenous SD had persistent 18F-FDG uptake in bone and
soft tissue, whereas uptake confined to the margins of a de-
stroyed disc after treatment was not indicative of infection.
Skanjeti et al72 reported that, in 21 patients with 24 sites of
disease, 18F-FDG-PET/CT was more accurate than MRI for as-
sessing treatment response (90% vs 61.5%).
Dauchy et al78 prospectively investigated 30 patients, in-
cluding 19 with spinal hardware, with microbiologically
confirmed SD, who were suspected of having relapsed in-
fection. All 30 patients underwent 18F-FDG PET/CT and 27
also underwent MRI 1 month after completion of antibiotic
treatment. Patients were followed for a minimum of 1 year.
Seven of the 30 patients were diagnosed with relapsed SD,
which was based on microbiological samples. Studies were
interpreted using a combination of visual and semiquantitative
analysis. Sensitivity, specificity, positive predictive value, and
negative predictive value were 85.7%, 82.6%, 60.0%, and
95.0%, respectively, for 18F-FDG PET/CT, and 66.6%, 61.9%,
33.3%, and 86.6%, respectively, for MRI. Although the results
of 18F-FDG PET/CT were better than those of MRI, the dif-
ferences were not significant.
Gallium-68
The positron emitting gallium isotope, gallium-68 (68Ga) in
the form of 68 Ga-citrate, offers certain advantages over
67
Ga-citrate for diagnosing infection. The physical half-life of
68
Ga is 68 minutes, which is much shorter than the 78-
hour half-life of 67Ga, and allows injection of larger quantities
of radioactivity. Imaging is performed within a few hours after
radiopharmaceutical administration, in contrast to the typical
144 M. Raghavan et al.

Figure 17 (A) L5-S1 spondylodiscitis. There is intense 18F-FDG uptake in L5-S1 on the baseline study. (B) There is near-complete resolution
of the abnormal 18F-FDG uptake on the follow-up PET/CT scan performed after 3 months of antibiotic treatment, consistent with resolving
infection.

48-72 hour interval between administration of 67Ga and
imaging. Because 68Ga is a positron emitter, image quality is
vastly superior to what can be obtained with 67Ga.
There are few data on the role of 68Ga-citrate imaging in
SD. Nanni et al79 reported on 31 patients with suspected SD
who underwent 40 68Ga PET/CT scans. There were 23 cases
of infection. The authors reported a sensitivity of 100%, a
specificity of 76%, and an accuracy of 90%. False-positive
results were associated with tumors. Unfortunately, the uptake
mechanisms of 68Ga in inflammation and infection are the same
nonspecific mechanisms that are responsible for uptake of 67Ga.
Although 68Ga offers the advantages of superior image quality
and same-day imaging, the other disadvantages associated with
67
Ga remain.
In an attempt to develop a more specific infection imaging
agent, some investigators have complexed 68Ga with various
agents. Based on the results reported with 111In-biotin in SD,
efforts have been made to complex biocytin with 68Ga. In an
initial investigation, 68Ga-DOTA-biocytin was very stable over
3 hours and may have potential for diagnosing infection.80
At the present time, MRI is the imaging test of choice for
diagnosing SD.18F-FDG is the radionuclide imaging test of
choice for this entity and is a useful compliment to MRI. In
addition to its value for diagnosing SD, 18F-FDG has shown
Spondylodiscitis 145

Figure 18 (A) Multifocal spondylodiscitis. There is intense 18F-FDG uptake involving T11-L1 and L4-L5 on the baseline study. (B) There is
complete resolution of the abnormal 18F-FDG uptake on the follow-up PET/CT scan performed after 3 months of antibiotic treatment, con-
sistent with resolved infection. Mildly increased 18F-FDG uptake at the L5-S1 level, also present on baseline study (6a), is related to degenerative
changes.
146
potential for monitoring response to treatment. When 18F-FDG
cannot be performed, 67 Ga-SPECT/CT is an acceptable
alternative.
References
1. Hopkinson N, Stevenson J, Benjamin S: A case ascertainment study of
septic discitis: Clinical, microbiological and radiological features. QJM
94:465-470, 2001
2. Mylona E, Samarkos M, Kakalou E, et al: Pyogenic vertebral osteomyelitis:
A systematic review of clinical characteristics. Semin Arthritis Rheum
39:10-17, 2009
3. Boody BS, Jenkins TJ, Maslak J, et al: Vertebral osteomyelitis and spinal
epidural abscess: An evidence-based review. J Spinal Disord Tech
28:E316-E327, 2015
4. Prodi E, Grassi R, Iacobellis F, et al: Imaging in spondylodiskitis. Magn
Reson Imaging Clin N Am 24:581-600, 2016
5. Hong SH, Choi JY, Lee JW, et al: MR imaging assessment of the spine:
Infection or an imitation? Radiographics 29:599-612, 2009
6. Yeom JA, Lee IS, Suh HB, et al: Magnetic resonance imaging findings
of early spondylodiscitis: Interpretive challenges and atypical findings.
Korean J Radiol 17:565-580, 2016
7. de Souza CG, Gasparetto EL, Marchiori E, et al: Pyogenic and tuberculous
discitis: Magnetic resonance imaging findings for differential diagnosis.
Radiol Bras 46:173-177, 2013
8. Butler JS, Shelly MJ, Timlin M, et al: Nontuberculous pyogenic spinal
infection in adults: A 12-year experience from a tertiary referral center.
Spine 31:2695-2700, 2006
9. Sundaram VK, Doshi A: Infections of the spine: A review of clinical and
imaging findings. Appl Radiol 10-20, 2016
10. Zimmerli W: Clinical practice. Vertebral osteomyelitis. NEJM 362:1022-
1029, 2010
11. Dagirmanjian A, Schils J, McHenry M, et al: MR imaging of vertebral
osteomyelitis revisited. AJR Am J Roentgenol 167:1539-1543, 1996
12. Ledermann HP, Schweitzer ME, Morrison WB, et al: MR imaging findings
in spinal infections: Rules or myths? Radiology 228:506-514, 2003
13. Berbari EF, Kanj SS, Kowalski TJ, et al: 2015 Infectious Diseases Society
of America (IDSA) clinical practice guidelines for the diagnosis and
treatment of native vertebral osteomyelitis in adults. Clin Infect Dis
61:e26-e46, 2015
14. Longo M, Granata F, Ricciardi K, et al: Contrast-enhanced MR imaging
with fat suppression in adult-onset septic spondylodiscitis. Eur Radiol
13:626-637, 2003
15. Diehn FE: Imaging of spine infection. Radiol Clin North Am 50:777-798,
2012
16. Ledbetter LN, Salzman KL, Shah LM: Imaging psoas sign in lumbar spinal
infections: Evaluation of diagnostic accuracy and comparison with
established imaging characteristics. AJNR Am J Neuroradiol 37:736-741,
2016
17. Mazzie JP, Brooks MK, Gnerre J: Imaging and management of postop-
erative spine infection. Neuroimaging Clin N Am 24:365-374, 2014
18. Hancock C, Quencer R, Falcon S: Challenges and pitfalls in postoperative
spine imaging. Appl Radiol 37:23-34, 2008
19. Modic MT, Steinberg PM, Ross JS, et al: Degenerative disk disease:
Assessment of changes in vertebral body marrow with MR imaging.
Radiology 166(1 Pt 1):193-199, 1988
20. Zhang YH, Zhao CQ, Jiang LS, et al: Modic changes: A systematic review
of the literature. Eur Spine J 17:1289-1299, 2008
21. Patel KB, Poplawski MM, Pawha PS, et al: Diffusion-weighted MRI “claw
sign” improves differentiation of infectious from degenerative modic type
1 signal changes of the spine. AJNR Am J Neuroradiol 35:1647-1652,
2014
22. Jung NY, Jee WH, Ha KY, et al: Discrimination of tuberculous spondylitis
from pyogenic spondylitis on MRI. AJR Am J Roentgenol 182:1405-1410,
2004
23. Kowalski TJ, Layton KF, Berbari EF, et al: Follow-up MR imaging in
patients with pyogenic spine infections: Lack of correlation with clinical
features. AJNR Am J Neuroradiol 28:693-699, 2007
M. Raghavan et al.
24. Palestro CJ: Radionuclide imaging of musculoskeletal infection: A review.
J Nucl Med 57:1406-1412, 2016
25. Adatepe MH, Powell OM, Isaacs GH, et al: Hematogenous pyogenic
vertebral osteomyelitis: Diagnostic value of radionuclide bone imaging.
J Nucl Med 27:1680-1685, 1986
26. Lisbona R, Derbekyan V, Novales-Diaz J, et al: Gallium-67 scintigraphy
in tuberculous and nontuberculous infectious spondylitis. J Nucl Med
34:853-859, 1993
27. Gratz S, Dörner J, Oestmann JW, et al: 67Ga-citrate and 99Tc-MDP for
estimating the severity of vertebral osteomyelitis. Nucl Med Commun
21:111-120, 2000
28. Love C, Patel M, Lonner BS, et al: Diagnosing spinal osteomyelitis: A
comparison of bone and Ga-67 scintigraphy and magnetic resonance
imaging. Clin Nucl Med 25:963-977, 2000
29. Hadjipavlou AG, Cesani-Vazquez F, Villaneuva-Meyer J, et al: The
effectiveness of gallium citrate Ga 67 radionuclide imaging in vertebral
osteomyelitis revisited. Am J Orthop (Belle Mead NJ) 27:179-183, 1998
30. Gratz S, Dorner J, Fischer U, et al: F-18-FDG hybrid PET in patients
with suspected spondylitis. Eur J Nucl Med Mol Imaging 29:516-524,
2002
31. Lievano P, De la Cueva L, Navarro P, et al: 67Ga SPECT/low-dose CT.
A case report of spondylodiscitis and Schmorl’s node. Rev Esp Med Nucl
28:288-290, 2009
32. Domínguez ML, Lorente R, Rayo JI, et al: SPECT-CT with (67)Ga-citrate
in the management of spondylodiscitis. Rev Esp Med Nucl Imagen Mol
31:34-39, 2012
33. Fuster D, Solà O, Soriano A, et al: A prospective study comparing
whole-body FDG PET/CT to combined planar bone scan with 67Ga
SPECT/CT in the diagnosis of spondylodiskitis. Clin Nucl Med 37:827-
832, 2012
34. Tamm AS, Abele JT: Bone and gallium single-photon emission computed
tomography-computed tomography is equivalent to magnetic resonance
imaging in the diagnosis of infectious spondylodiscitis: A retrospective
study. Can Assoc Radiol J 68:41-46, 2017
35. Kennedy DO: B Vitamins and the brain: mechanisms, dose and
efficacy—a review. Nutrients 8:68, 2016
36. Lazzeri E, Erba P, Perri M, et al: Scintigraphic imaging of vertebral
osteomyelitis with 111in-biotin. Spine 33:E198-E204, 2008
37. Lazzeri E, Manca M, Molea N, et al: Clinical validation of the avidin/
indium-111 biotin approach for imaging infection/inflammation in
orthopaedic patients. Eur J Nucl Med 26:606-614, 1999
38. Lazzeri E, Pauwels EK, Erba PA, et al: Clinical feasibility of two-step
streptavidin/111In-biotin scintigraphy in patients with suspected
vertebral osteomyelitis. Eur J Nucl Med Mol Imaging 31:1505-1511,
2004
39. Lazzeri E, Erba P, Perri M, et al: Clinical impact of SPECT/CT with In-111
biotin on the management of patients with suspected spine infection.
Clin Nucl Med 35:12-17, 2010
40. Palestro CJ, Glaudemans AWJM, Dierckx RAJO: Multiagent imaging of
inflammation and infection with radionuclides. Clin Transl Imaging
1:385-396, 2013
41. Palestro CJ, Kim CK, Swyer AJ, et al: Radionuclide diagnosis of
vertebral osteomyelitis: Indium-Ill leukocyte and technetium-
99mmethylene diphosphonate bone scintigraphy. J Nucl Med 32:1861-
1865, 1991
42. Palestro CJ, Love C: Radionuclide imaging of musculoskeletal
infection: Conventional agents. Semin Musculoskelet Radiol 11:335-352,
2007
43. Guhlmann A, Brecht-Krauss D, Suger G, et al: Fluorine-18-FDG PET
and technetium-99m antigranulocyte antibody scintigraphy in chronic
osteomyelitis. J Nucl Med 39:2145-2152, 1998
44. Gratz S, Braun HG, Behr TM, et al: Photopenia in chronic vertebral
osteomyelitis with technetium-99 m antigranulocyte antibody (BW
250/183). J Nucl Med 38:211-216, 1997
45. Wang G: Human antimicrobial peptides and proteins. Pharmaceuticals
7:545-594, 2014
46. Hiemstra PS, van den Barselaar MT, Roest M, et al: Ubiquicidin, a novel
murine microbicidal protein present in the cytosolic fraction of
macrophages. J Leukoc Biol 66:423-428, 1999
Spondylodiscitis
47. Lupetti A, Pauwels EK, Nibbering PH, et al: 99mTc-antimicrobial
peptides: Promising candidates for infection imaging. Q J Nucl Med
47:238-445, 2003
48. Dillmann-Arroyo C, Cantú-Leal R, Campa-Núñez H, et al: Utilidad de
la gammagrafía con ubiquicidina 29-41 en el diagnóstico de osteomielitis
piógena de columna vertebral. Acta Ortop Mex 25:27-31, 2011
49. Meller J, Sahlmann CO, Scheel AK: 18F-FDG PET and PET/CT in fever
of unknown origin. J Nucl Med 48:35-45, 2007
50. Palestro CJ: FDG-PET in musculoskeletal infections. Sem Nucl Med
43:367-376, 2013
51. Kubota R, Yamada S, Kubota K, et al: Intratumoral distribution of
fluorine-18-fluorodeoxyglucose in vivo: High accumulation in
macrophages and granulation tissues studied by microautoradiography.
J Nucl Med 33:1972-1980, 1992
52. Yamada S, Kubota K, Kubota R, et al: High accumulation of fluorine-
18-fluorodeoxyglucose in turpentine-induced inflammatory tissue. J Nucl
Med 36:1301-1306, 1995
53. Kaim AH, Weber B, Kurrer MO, et al: Autoradiographic quantification
of 18F-FDG uptake in experimental soft-tissue abscesses in rats.
Radiology 223:446-451, 2002
54. Zhuang H, Duarte PS, Pourdehand M, et al: Exclusion of chronic
osteomyelitis with F-18 fluorodeoxyglucose positron emission
tomographic imaging. Clin Nucl Med 25:281-284, 2000
55. Guhlmann A, Brecht-Krauss D, Suger G, et al: Chronic osteomyelitis:
Detection with FDG PET and correlation with histopathologic findings.
Radiology 206:749-754, 1998
56. Schiesser M, Stumpe KD, Trentz O, et al: Detection of metallic implant-
associated infections with FDG PET in patients with trauma: Correlation
with microbiologic results. Radiology 226:391-398, 2003
57. Kalicke T, Schmitz A, Risse JH, et al: Fluorine-18 fluorodeoxyglucose
PET in infectious bone diseases: Results of histologically confirmed cases.
Eur J Nucl Med Mol Imaging 27:524-528, 2000
58. Schmitz A, Risse JH, Grunwald F, et al: Fluorine-18 fluorodeoxyglucose
positron emission tomography findings in spondylodiscitis: Preliminary
results. Eur Spine J 10:534-539, 2001
59. Meller J, Koster G, Liersch T, et al: Chronic bacterial osteomyelitis:
Prospective comparison of F-18-FDG imaging with a dual-head
coincidence camera and In-111-labelled autologous leucocyte
scintigraphy. Eur J Nucl Med Mol Imaging 29:53-60, 2002
60. Hungenbach S, Delank KS, Dietlein M, et al: 18F-fluorodeoxyglucose
uptake pattern in patients with suspected spondylodiscitis. Nucl Med
Commun 34:1068-1074, 2013
61. Hartmann A, Eid K, Dora C, et al: Diagnostic value of 18F-FDG PET/CT
in trauma patients with suspected chronic osteomyelitis. Eur J Nucl Med
Mol Imaging 34:704-714, 2007
62. Bassetti M, Merelli M, Di Gregorio F, et al: Higher fluorine-18
fluorodeoxyglucose positron emission tomography (FDG-PET) uptake
in tuberculous compared to bacterial spondylodiscitis. Skeletal Radiol
46:777-783, 2017
63. Kim SJ, Lee JS, Suh KT, et al: Differentiation of tuberculous and pyogenic
spondylitis using double phase F-18 FDG PET. Open Med Imaging J
2:1-6, 2008
64. Gunes BY, Onsel C, Sonmezoglu K, et al: Diagnostic value of F-18 FDG
PET/CT in patients with spondylodiscitis: Is dual time point imaging
time worthy? Diagn Microbiol Infect Dis 381-385, 2016
147
65. de Winter F, Gemmel F, Van de Wiele C, et al: 18-fluorine
fluorodeoxyglucose positron emission tomography for the diagnosis
of infection in the postoperative spine. Spine 28:1314-1319,
2003
66. Inanami H, Oshima Y, Iwahori T, et al: Role of 18 F-Fluoro- D
-deoxyglucose PET/CT in diagnosing surgical site infection after spine
surgery with instrumentation. Spine 40:109-113, 2015
67. Bagrosky BM, Hayes KL, Koo PJ, et al: 18F-FDG PET/CT evaluation of
children and young adults with suspected spinal fusion hardware
infection. Pediatr Radiol 43:991-1000, 2013
68. Ohtori S, Suzuki M, Koshi T, et al: 18F-fluorodeoxyglucose-PET for
patients with suspected spondylitis showing Modic change. Spine
35:E1599-E1603, 2010
69. Stumpe KD, Zanetti M, Weishaupt D, et al: FDG positron emission
tomography for differentiation of degenerative and infectious endplate
abnormalities in the lumbar spine detected on MR imaging. AJR Am J
Roentgenol 179:1151-1157, 2002
70. Seifen T, Rettenbacher L, Thaler C, et al: Prolonged back pain attributed
to suspected spondylodiscitis. The value of 18F-FDG PET/CT imaging
in the diagnostic work-up of patients. Nuklearmedizin 51:194-200,
2012
71. Fuster D, Tomás X, Mayoral M, et al: Prospective comparison of
whole-body 18F-FDG PET/CT and MRI of the spine in the diagnosis
of haematogenous spondylodiscitis. Eur J Nucl Med Mol Imaging
42:264-271, 2015
72. Skanjeti A, Penna D, Douroukas A, et al: PET in the clinical work-up
of patients with spondylodiscitis: A new tool for the clinician? Q J Nucl
Med Mol Imaging 56:569-576, 2012
73. Ioannou S, Chatziioannou S, Pneumaticos SG, et al: Fluorine-18
fluoro-2-deoxy-D-glucose positron emission tomography/computed
tomography scan contributes to the diagnosis and management of
brucellar spondylodiskitis. BMC Infect Dis 13:73, 2013
74. Nakahara M, Ito M, Hattori N, et al: 18F-FDG-PET/CT better localizes
active spinal infection than MRI for successful minimally invasive surgery.
Acta Radiol 56:829-836, 2015
75. Nanni C, Boriani L, Salvadori C, et al: FDG PET/CT is useful for the
interim evaluation of response to therapy in patients affected by
haematogenous spondylodiscitis. Eur J Nucl Med Mol Imaging 39:1538-
1544, 2012
76. Kim SJ, Kim IJ, Suh KT, et al: Prediction of residual disease of spine
infection using F-18 FDG PET/CT. Spine 34:2424-2430, 2009
77. Riccio SA, Chu AK, Rabin HR, et al: Fluorodeoxyglucose positron
emission tomography/computed tomography interpretation criteria for
assessment of antibiotic treatment response in pyogenic spine infection.
Can Assoc Radiol J 66:145-152, 2015
78. Dauchy FA, Dutertre A, Lawson-Ayay S, et al: Interest of
[18F]fluorodeoxyglucose positron emission tomography/computed
tomography for the diagnosis of relapse in patients with spinal infection:
A prospective study. Clin Microbiol Infect 22:438-443, 2016
79. Nanni C, Errani C, Boriani L, et al: 68Ga-Citrate PET/CT for evaluating
patients with infections of the bone: Preliminary results. J Nucl Med
51:1932-1936, 2010
80. Asti M, Iori M, Erba P, et al: Radiosynthesis of 68Ga-labelled DOTA-
biocytin (68Ga-r-BHD) and assessment of its pharmaceutical quality for
clinical use. Nucl Med Commun 33:1179-1187, 2012