P E T- C o m p u t e d Tom o g r a p h y

i n Vet e r i n a r y M e d i c i n e
Elissa K. Randall, DVM, MS

KEYWORDS

PET/CT
FDG
Veterinary
Staging
SUV

KEY POINTS

PET/computed tomography (CT) is used in veterinary medicine to diagnose and stage pa-
tients with cancer, using the glucose analogue 2-deoxy-2-18F-fluorodeoxyglucose (F-18
FDG).

FDG-PET/CT is good at detecting areas of increased glucose metabolism and can detect
primary tumors or metastatic lesions that are missed on routine examination and staging.

Hypermetabolic areas (areas of increased glucose metabolism) are not specific for
neoplastic tissue and may also represent physiologic variation or benign variation,
including inflammation.

Evaluation of the CT images may provide the information needed to determine the most
likely diagnosis for a visualized hypermetabolic area, but cytology or histopathology is
often needed for definitive diagnosis.

INTRODUCTION

Advanced imaging modalities, such as computed tomography (CT) and MRI, have been
in use for decades in veterinary medicine. They provide images with superior anatomic
information compared with radiographs and ultrasound. PET/CT is an advanced imag-
ing modality that is becoming more commonly used in veterinary medicine. Special im-
aging equipment and appropriate nuclear medicine facilities are required to perform the
studies, which are most commonly performed on patients with cancer.

WHAT IS PET/COMPUTED TOMOGRAPHY?

PET/CT was previously performed by acquiring a PET scan and CT scan on different
machines at different times; but more recently, new machines acquire images
on a dual PET/CT scanner as part of one imaging examination (Fig. 1). Combined

Disclosure statement: The author has nothing to disclose.

Department of Environmental and Radiological Health Sciences, Veterinary Teaching Hospital,
Colorado State University, ACC 135, 300 West Drake Road, Fort Collins, CO 80523-1620, USA
E-mail address: elissa.randall@colostate.edu

Vet Clin Small Anim 46 (2016) 515–533

http://dx.doi.org/10.1016/j.cvsm.2015.12.008 vetsmall.theclinics.com
0195-5616/16/$ – see front matter Ó 2016 Elsevier Inc. All rights reserved.
516 Randall

Fig. 1. A dual PET/CT scanner. The ring toward the front is the CT scanner. The ring in the
back is the PET scanner. The table moves patients into the appropriate position for each
phase of the PET/CT study. (Courtesy of Philips Healthcare. All rights reserved.)

PET/CT scanners provide automatic image fusion, whereas PET and CT studies ob-
tained independently require appropriate software for fusion.
PET is a form of nuclear medicine that uses radiopharmaceuticals that are positron
emitters. Positrons are positively charged particles, also called beta1 particles or b1.
These particles travel a short distance (1–2 mm) before colliding with a negatively
charged electron. When the two collide, 2 annihilation gamma photons are created
and travel 180 from each other (coincident photons). The special detectors in a
PET scanner detect and register these coincident photons, which arrive within a few
nanoseconds of each other. The annihilation photons have energy of approximately
511 keV, which is higher than the 140 keV energy of technetium 99m, the radiophar-
maceutical used for bone scans, glomerular filtration rate studies, and thyroid scans.1
CT is more familiar to the veterinary community. CT provides images based on x-ray
attenuation, similar to radiography. However, images are acquired in slices and super-
imposition of structures is eliminated. CT images also have better contrast resolution
than radiographs. The result is excellent anatomic depiction of patients. Multi-planar
or 3-dimensional (3D) reformatting of images can provide more complete assessment.2

WHAT RADIOPHARMACEUTICALS ARE USED IN PET/COMPUTED TOMOGRAPHY?

The most commonly used PET radiopharmaceutical is 2-deoxy-2-18F-fluorodeoxy-

glucose (F-18 FDG). FDG is a glucose analogue that is labeled to F-18, a radioactive
positron emitter. F-18 FDG is transported into the cell by glucose transporter proteins.
In the cell, it is phosphorylated, similar to glucose. Glucose is then transported into
mitochondria to be converted into energy or stored as glycogen. However, F-18
FDG undergoes no further reaction after this step and is effectively trapped in the
cell. This mechanism allows F-18 FDG to be used to map glucose metabolism in
the body.3,4 Tumor cells often use more glucose than normal cells, making F-18
FDG an excellent tool to detect neoplastic tissue. However, some normal tissues
use more glucose than other normal tissue and show high FDG uptake, so are hyper-
metabolic, or have higher FDG avidity, on PET images (Fig. 2). In addition, inflamma-
tory lesions often use more glucose and are hypermetabolic on PET images. Knowing
the appearance of variations of normal is key to image interpretation.
 PET-Computed Tomography 517

Fig. 2. Examples of organs that are normally hypermetabolic on FDG-PET/CT scans. All im-
ages are transverse fused PET/CT images. (A) The brain is diffusely hypermetabolic, with a
standardized uptake value (SUV) maximum (max) of between 7.7 and 9.3 in this patient.
(B) Salivary glands are generally hypermetabolic, though the degree varies with different
salivary glands. Both mandibular salivary glands are visualized as ovoid hypermetabolic
structures in this patient, with an SUV max of 5.7. (C) Cardiac muscle is variably hypermet-
abolic. This patient has several levels of metabolic activity in the heart, with the most hyper-
metabolic area of muscle having an SUV max of 7.7 to 9.3. (D) Intestines show variable
metabolic activity. The patient has more hypermetabolic segments of small intestine on
the left side of the image, with an SUV max of 5.0. Other, less hypermetabolic segments
of small intestine on the right side of the images have an SUV max of 2.5.

F-18 FDG is excreted by the kidneys and eliminated in the urine. The half-life of F-18
FDG is 110 minutes, so approximately 50% of the given dose is present in the urine
2 hours after administration.1 Because of this, patients typically have a urinary catheter
placed while anesthetized in order to prevent radioactive contamination due to urine
leakage.
Other PET radiopharmaceuticals used in clinical cases or clinical research in veteri-
nary medicine include F-18 sodium fluoride (NaF), used to image bone and in particular
osteoblastic activity; F-18 fluorothymidine (FLT), used to image DNA synthesis; and F-18
fluoromisonidazole and Cu-60,62,64 diacetyl-bis-N4-methylthiosemicarbazone (Cu-
ATSM), used to study tumor hypoxia.5–9 The focus of this article is F-18 FDG-PET/CT.

HOW IS A PET/COMPUTED TOMOGRAPHY STUDY ACQUIRED?

Protocols vary at different institutions. The protocol described is the one followed at
Colorado State University, where images are obtained on a dual PET/CT scanner.
Because a radiopharmaceutical is involved, having a clear understanding of the
518 Randall

nuclear medicine radiation safety protocol, in addition to the CT radiation safety pro-
tocol, is important for all those involved. Appropriate facilities for the intake and stor-
age of radiopharmaceuticals are necessary, as are appropriate facilities for housing
radioactive patients.
Patients are fasted overnight before the PET/CT scan. Serum glucose is measured
the morning of the scan. In humans, a moderate to severely elevated glucose (more
than 200 ng/dL with a normal range 70–110 ng/dL, according to one reference) usually
results in the scan being postponed until glucose levels are better regulated because
high serum glucose can compete with F-18 FDG for uptake into the cell and affect the
quality of the scan.1 The normal range for glucose at this institution is 75 to 130 mg/dL
for dogs and 69 to 136 mg/dL for cats. The author’s institution does not abide by a
strict cutoff for glucose values because this is an area of controversy, even in human
medicine. However, uncontrolled diabetic patients do not undergo FDG-PET/CT
scanning.
Patients are anesthetized according to the protocol determined by the anesthesia
section. Patients are then transported to the PET/CT suite and positioned in dorsal
or sternal recumbency. FDG is injected intravenously (0.14–0.17 mCi/kg), after which
there is a 1-hour uptake period. This period is to allow FDG to distribute throughout the
body and be transported into cells and trapped in those cells. Human patients are
asked to sit quietly and silently in a room during the uptake period to limit physiologic
uptake of FDG in skeletal muscle secondary to muscle use, which can make interpre-
tation more difficult.1 Because veterinary patients at this institution are anesthetized
during the 1-hour uptake period, they have limited skeletal uptake from motion during
this period. A preintravenous and postintravenous contrast whole-body CT scan is
performed during the 1-hour uptake period. After 1 hour, PET images are acquired
with patients in the same position. Whole-body PET images are obtained in 5 to 12
bed positions (2–3 minutes each) depending on the size of the patient. After the
PET images are acquired, patients are transported to a special nuclear medicine
ward (isolated from other hospital patients and personnel) and recovered there.
Because of the short half-life of FDG (110 minutes), patients are typically scanned in
the morning and released to go home the same evening, after their radiation level rea-
ches the institutional safety level. Because of urinary excretion, patients are catheter-
ized during the scan and eliminate in a specific nuclear medicine area after they have
recovered from anesthesia.
In veterinary medicine, whole-body PET/CT scans typically include the entire pa-
tient, from the tip of the nose to beyond the caudal soft tissues of the pelvis, including
all 4 limbs through the digits. Limbs may be positioned in extension or flexion. In com-
parison, most whole-body human PET/CT scans cover the base of the skull to the
midthigh.1

IMAGE PROCESSING

Non–attenuation-corrected (NAC) PET images are reconstructed using the line-of-

response row-action maximum likelihood algorithm method, which is a 3D reconstruc-
tion technique. CT attenuation-corrected (CTACs) images of PET are based on the
precontrast CT dataset. PET images are available as CTAC and NAC images. CTAC
images are the most commonly viewed of the PET images. The CT data provide infor-
mation about the degree of photon attenuation by different tissues or structures. This
information helps to provide proper scaling of the PET data by taking different tissue
densities into account when interpreting the degree of PET photon emission and
detection. However, when there are metallic implants or other very hyperattenuating
 PET-Computed Tomography 519

structures, these can be overinterpreted and depicted as hypermetabolic areas on

PET and potentially misinterpreted as malignant or inflammatory. Therefore, NAC im-
ages are also available. They will show normal FDG uptake in such areas and help the
interpreter realize when a hypermetabolic region is artifactual.10 Images are reviewed
as CT-only images, PET-only images, and fused PET/CT images. PET images are
fused with postcontrast CT images.

STANDARDIZED UPTAKE VALUE

Standardized uptake value (SUV) is a semiquantitative method of evaluating FDG up-

take. It indicates the degree of FDG uptake in a measured structure, and SUV is
directly proportional to metabolic activity. In simplified terms, if there were even distri-
bution of the radiopharmaceutical throughout the body, the SUV would be 1. There-
fore, a lesion with an SUV of 7 has 7 times the average uptake.3 SUV maximum
(max), SUV minimum, SUV average, and SUV total can be measured; but SUV max
is the most commonly used in lesion analysis.1 Examples of SUV max of normal struc-
tures in dogs are included in Table 1. SUV measurements have to be interpreted with
caution because factors, such as body weight, time between injection and imaging,
and serum glucose levels, affect SUV. In addition, comparison of SUV between insti-
tutions is difficult because of the different protocols, PET/CT equipment, and image
acquisition settings.8
Although SUV measurements can be helpful, there is literature that suggests that vi-
sual inspection of a lesion is as reliable as SUV measurements in determining malig-
nancy.11 In human medicine, threshold values are used for some lesions to determine
a benign from malignant nature. An example of this is solitary pulmonary nodules in
people. A nodule greater than 1 cm with an SUV max equal to or greater than 2.5 is
considered malignant.1 However, no specific thresholds or specific cutoffs have
been established in veterinary medicine. In addition, there is significant overlap in
SUVs of benign and malignant lesions. Large numbers of PET/CT cases of specific tu-
mor types would be necessary to establish SUV malignancy thresholds in veterinary
medicine, a difficult task in this field.

WHAT INFORMATION DOES A PET/COMPUTED TOMOGRAPHY STUDY PROVIDE?

The CT images provide excellent anatomic depiction of normal and abnormal struc-
tures. The PET images show areas of high glucose metabolism. CT and PET images

Table 1
SUV max of normal organs in dogs

Organ Typical SUV Max

Brain 4–7
Salivary glands 2–6
Liver 1–3
Spleen 1–3
Skeletal muscle 0.5–1.5
Bone 0.5–2.0
Spinal cord 0.5–2.0
Intestines 0.5–8.0
Heart (myocardium) 0.5–9.0
Lymph nodes 0.5–2.0
520 Randall

are compared, side by side and fused, to determine if areas of noticeably high meta-
bolic activity are normal or abnormal on CT. For example, the brain is normally mark-
edly hypermetabolic because of its high glucose metabolism. Cardiac muscle is
variably hypermetabolic because of the variation in myocardial metabolism relative
to fasting and individual variation.12,13 Salivary glands are hypermetabolic because
of excretion of FDG in the saliva.12 When a hypermetabolic area is noted, the CT
will be used to determine what specific structure is hypermetabolic. Then knowledge
of normal FDG uptake, and comparison with the contralateral structure if present, is
used to determine if the degree of uptake is normal or abnormal for that structure. If
a structure or area is deemed to be abnormally hypermetabolic, CT images provide
a better picture of what is occurring locally and determine if a mass, inflammation,
or other abnormalities are present.
FDG-PET/CT is used primarily to image patients with cancer. The ability to detect
lesions with increased glucose use is helpful in identifying primary neoplastic lesions,
determining the extent of known lesions, detecting metastatic lesions, and overall
staging of patients with cancer. PET/CT aids in the detection, staging, treatment plan-
ning, and response to treatment of human and veterinary patients with cancer. In
humans, the list of tumors covered for FDG-PET/CT by Medicare and Medicaid as
of June 11, 2013 included colorectal, esophageal, head and neck (not thyroid or cen-
tral nervous system) lymphoma, non–small cell lung, ovary, brain, cervix, small cell
lung, soft tissue sarcoma, pancreas, testes, thyroid, breast, melanoma, and myeloma;
a few of those have exceptions. Also listed as covered are all other solid tumors and all
other cancers not listed. One tumor type not covered for initial diagnosis and staging is
prostate cancer.14 This long list of covered tumors underscores the importance of
PET/CT in human oncology.
The normal distribution of FDG in cats and dogs has been described.4,12,13 Physiologic
variants, benign processes, and artifacts have also been described for FDG-PET/CT in
dogs and cats.15 The normal distribution of other radiopharmaceuticals has been inves-
tigated, including F-18 FLT in cats and F-18 NaF in dogs.9,16 These studies pave the way
for clinical investigation of disease processes in veterinary patients (Fig. 3).
PET/CT of the following tumor types have been reported on, or are currently being
studied, in veterinary medicine: lymphoma, feline oral squamous cell carcinoma, os-
teosarcoma (OSA), mast cell tumor (MCT), plasma cell tumor, primary lung tumor,
nasal tumors, mammary carcinoma, anal sac adenocarcinoma, pancreatic neuroen-
docrine carcinoma, soft tissue sarcoma, fibrosarcoma, hemangiopericytoma, squa-
mous cell carcinoma, histiocytic sarcoma, Sertoli cell tumor, and gastrointestinal
stromal tumor.5,6,17–24 Some tumors have been studied as part of a coordinated series
of cases of the same tumor type, and others have been cases that were part of a se-
lection of multiple tumor types.

INITIAL DIAGNOSIS AND STAGING OF PATIENTS WITH CANCER

PET/CT is most often used to stage patients with a suspected or known primary tu-
mor. It provides in-depth anatomic and metabolic information about the primary lesion
and can detect metastatic lesions or diagnose unknown second neoplasms. PET/CT
can also be used to look for the unknown primary lesion when metastasis is the first
detected lesion.
PET/CT provides a way to achieve whole-body staging in one imaging procedure.
The CT data are acquired in sufficiently small slices to provide excellent evaluation
of the thorax and abdomen as well as of all bony structures. Charges vary by institu-
tion; but at this institution, the charge for a PET/CT is actually slightly less expensive
 PET-Computed Tomography 521

Fig. 3. Examples of benign variants in dogs: regions that may be hypermetabolic because of
physiologic variation or inflammation, which may be subclinical. (A) A normal hypermeta-
bolic brain is visible. Ventral to the brain are 2 rounded, hypermetabolic structures, which
represent tonsils (arrow). Tonsils are frequently hypermetabolic, which may be because of
physiologic variation or secondary to oral or dental inflammation. (B) Ventral to the eyes
are mild to moderately hypermetabolic structures, which are normal zygomatic salivary
glands (small arrow). Also, lateral to the mandible bilaterally are focal hypermetabolic areas,
which may represent muscle uptake secondary to recent use or stretching during intubation,
uptake in vasculature or local soft tissues, or salivary pooling (large arrow). (C) Focal uptake
in the supraspinatus muscle (arrow), which may be due to altered weight bearing from
known lameness or other focal muscle injury or recent use. (D) The liver can show patchy,
mild to moderate hypermetabolic activity, as in this patient. This activity is seen is patients
with no known hepatic disease and is considered a normal variant. There is also focal hyper-
metabolic uptake in the gallbladder (arrow), which is also considered a normal variant. Clin-
ically silent hepatobiliary disease could be responsible for these changes.

than the combination of charges for a 3-view thoracic radiographic study 1 abdominal
ultrasound 1 single-region CT. This charge does not include the cost of anesthesia,
which is longer for PET/CT than CT alone. However, full-body staging is provided,
with a more detailed evaluation of the lungs (for metastasis or other lesions), abdomen,
head and neck, and spine/limbs.
The veterinary PET/CT literature up to this point has included a variety of tumor
types, often single case reports or small numbers of the same tumor type. The tumor
522 Randall

types with the greatest representation in veterinary FDG-PET or FDG-PET/CT litera-

ture include lymphoma, MCT, feline squamous cell carcinoma, and fibrosarcoma (5
cases or more of each).

LYMPHOMA AND MAST CELL TUMORS

Lymphoma and MCTs are neoplastic processes than are often disseminated to mul-
tiple organs and can involve liver, spleen, and lymph nodes. Patients with MCTs
commonly have multiple cutaneous nodules. PET/CT has been shown to be similar
to better than routine staging at detecting disease in lymph nodes. However, diffuse
disease of the liver and spleen has been variably detected with PET/CT, being some-
times hypermetabolic and sometimes normal on PET images.17,19 This data included a
study that involved FDG-PET only.19 Bone marrow infiltration has also been correctly
diagnosed by PET/CT.17 MCTs are variably hypermetabolic, and small or low-grade
lesions may be poorly visualized or not visualized on PET/CT. However, most MCTs
imaged at this institution have been visibly hypermetabolic, including low-grade
MCTs (Fig. 4).

OSTEOSARCOMA

The tumor type most commonly staged with FDG-PET/CT at this institution is OSA,
with approximately 60 cases having been imaged. Primary tumors are mildly to
severely hypermetabolic, with SUV max ranging from 1.8 to 24.9 in one report; but
an SUV max up to 34.9 has been seen.25 Metastatic disease has been detected in pa-
tients with OSA; but many patients show no other evidence of disease, making the de-
cision to pursue treatment more reasonable to owners. If metastasis is detected, those
sites may be included in the radiation treatment planning process, if radiation therapy
is pursued. Patients with a primary bone tumor often have areas of hypermetabolic ac-
tivity in muscles of the contralateral limb, or multiple limbs, attributed to altered weight
bearing (Fig. 5).

METASTASIS

When staging patients with cancer, primary lesions are usually hypermetabolic and
well visualized on PET/CT. Metastatic lesions are also typically hypermetabolic and
well visualized. One major benefit of PET/CT is in detecting metastatic lesions that
were not detected on physical examination or lesions that do not appear abnormal
on CT (Fig. 6). These lesions include lymph nodes that are normal on palpation and
CT but hypermetabolic on PET. It is important to remember that these hypermetabolic
lymph nodes may be reactive or metastatic, so cytology or histopathology should be
attempted when possible (Fig. 7).

QUESTIONABLE LESIONS

One difficulty of PET/CT is that it is possible to also see lesions that are mildly to moder-
ately hypermetabolic on PET but benign in appearance on CT. These lesions may be
difficult to access for aspirates or biopsies, making confirmation of the benign or malig-
nant nature difficult. Occasionally, repeat PET/CT can shed light on such lesions.

NON-NEOPLASTIC LESIONS

Many types of non-neoplastic lesions can be hypermetabolic and can range from
mildly to severely hypermetabolic. If the lesions are superficial or related to known
 PET-Computed Tomography 523

Fig. 4. A patient with 2 MCTs. (A) Dorsal and oblique lateral maximum intensity projection
(MIP) images. These images show the FDG uptake throughout the body and allow the inter-
preter to identify hot spots that need to be investigated further on CT, PET, and fused PET/CT
images. (B) CT image of a high-grade MCT on the right side of the nasal planum, which is
mildly contrast enhancing. (C) Fused PET/CT image of the high-grade MCT of the nasal pla-
num, which had an SUV max of 5.6. This mass is seen as a distinct hot spot on the nose on
both MIP images. (D) Fused PET/CT image of a mildly enlarged and mildly hypermetabolic
mandibular lymph node (arrow). This lymph node had an SUV max of 3.4, compared with
the contralateral mandibular lymph nodes, which had an SUV max of 1.6. The right lymph
node was considered possibly metastatic on imaging. It was interpreted as metastatic mast
cell disease on cytology but reactive on histopathology. (E) Fused PET/CT image of a low-
grade MCT on the abdominal body wall (arrow). This tumor had an SUV max of 1.8 and
could be visualized on MIP images when viewed in a straight lateral position. (F) Sagittal
fused PET/CT image of the left hind paw, which had multifocal areas of hypermetabolic ac-
tivity, with an SUV max of 5.0 to 6.3. These areas are also well visualized on the MIP images.
The paw was inspected, and multiple grass awns were removed.

procedures, it may be obvious to the interpreter why the lesion is hypermetabolic,

decreasing the chance of misinterpretation of malignancy. Examples include
hygromas, lick granulomas, feeding tube placement sites, and other superficial wounds
or sites of infection. Unfortunately, some lesions that are not ascribable to a benign pro-
cess or procedure, such as sclerotic, hypermetabolic lesions in vertebral bodies, are
difficult to access for sampling; the clinician will have to decide the best treatment
plan with the presumption of metastasis. Lesions that are logically diagnosed as met-
astatic disease, such as hypermetabolic lung nodules in a patient with a known primary
tumor, may be proven to be malignant based on cellular sampling or follow-up imaging.
However, some suspect metastatic lesions are proven to be benign on histopathology
or follow-up imaging (Fig. 8). These cases again underscore the importance of obtain-
ing histopathology of suspect metastatic lesions whenever possible.
524 Randall

Fig. 5. Patient with OSA. (A) Oblique lateral maximum intensity projection image that
shows the primary OSA in the left tibia (long arrow), a metastatic lesion to the left radius
(short arrow), and a metastatic lesion to the articular facet of the first lumbar vertebra
(star). The left popliteal lymph node was hypermetabolic and can be seen as a hot spot prox-
imal to the primary lesion in the tibia. (B) Transverse CT image in a bone window, showing
marked lysis and periosteal reaction of the primary tumor on the tibia. The image on the
right is a fused PET/CT image at the same location showing the hypermetabolic activity of
the tumor, which had an SUV max of 34.9. (C) Transverse CT image in a soft tissue window,
showing the contrast-enhancing soft tissue swelling surrounding the mass. The image on
the right is a fused PET/CT image at the same location showing the hypermetabolic activity
of the bone and soft tissue abnormalities. (D) Transverse CT image in a bone window,
showing lysis and expansion of the articular facet of the first lumbar vertebra and sclerosis
of the pedicle. The image on the right is a fused PET/CT image at the same location showing
the hypermetabolic activity in the articular facet, which had an SUV max of 6.2.

RADIATION TREATMENT PLANNING AND PET/COMPUTED TOMOGRAPHY AVID

TUMORS
Oral Squamous Cell Carcinoma
PET/CT is valuable for guiding radiation treatment planning and can also be used to
assess the response to radiation treatment. Cats with oral squamous cell carcinoma
 PET-Computed Tomography 525

Fig. 6. (A) Patient with a recently excised plasmacytoma of the mandible who returned for
staging after cutaneous nodules were noted. The PET/CT showed the cutaneous nodules
along the dorsum (seen on the maximum intensity projection), plus multiple bone lesions
that were new, surprising findings. The most hypermetabolic new bone lesions were in
both distal radii, both distal tibias, and the right scapula. (B) Fused PET/CT image of the
distal radii. At this level, the lesion in the left radius is visualized as a hypermetabolic
area in the medullary cavity that was contrast enhancing on CT (arrow). The SUV max
was 2.2. (C) Fused transverse PET/CT image showing a mildly hypermetabolic lesion in the
right scapula (arrow). The SUV max was 2.8. (D) Transverse CT image in a bone window,
showing mild, ill-defined lysis at the same location. Following the PET/CT, the cutaneous
soft tissue nodules were biopsied and the diagnosis was changed to multiple myeloma.

were studied; PET/CT provided more extensive margins for soft tissue tumors than CT
alone, whereas CT volume was greater than PET/CT volume in some cats, especially
those with bone involvement. PET/CT also detected metastatic lesions that were
equivocal on CT.23 The end result was a larger gross tumor volume for radiation ther-
apy treatment planning. Although it was not proven whether the hypermetabolic areas
were entirely neoplastic tissue or also included inflammatory tissue, the greater tumor
volume seen on PET/CT versus CT alone was used for radiation treatment to decrease
the chances of missing tumor cells (geographic miss), which may decrease local tu-
mor control.18

NASAL TUMORS

Similar conclusions were reached in a study of dogs with nasal tumors. PET/CT
showed different tumor margins than CT alone, with CT often showing a larger overall
526 Randall

Fig. 7. (A) Lateral maximum intensity projection of a cat with an oral squamous cell carci-
noma, seen as a large hypermetabolic area in the caudal mandibular region. The arrow is
pointing to a focal area of mild hypermetabolic activity in the ventral cervical region. This
area was further investigated on transverse images (B, C) and determined to be a deep cer-
vical lymph node. (B) Transverse fused image showing the mild hypermetabolic activity in
the area of the deep cervical lymph node. The SUV max was 3.6. (C) Transverse CT image
in a soft tissue window at the same level as image (B). The right deep cervical lymph
node (arrow) is mildly rounded and slightly larger than the left but is somewhat ill defined
and difficult to identify as abnormal. Cytology of the right deep cervical lymph node re-
vealed metastatic squamous cell carcinoma. (D) Fused transverse PET/CT image of a mildly
enlarged and hypermetabolic mandibular lymph node (arrow) in dog with nasal squamous
cell carcinoma. The SUV max was 4.1. Cytology revealed a reactive lymph node. (E) Fused
transverse PET/CT image of a mildly enlarged and hypermetabolic mandibular lymph
node (arrow) in dog with MCT. The SUV max was 3.5. Histopathology revealed a reactive
lymph node.

volume. However, PET/CT detected hypermetabolic tissue outside the primary tumor
that was normal on CT; in some dogs, the PET/CT volume was substantially greater
than the CT volume. The conclusion was that PET/CT offers the best chance of depict-
ing the maximal tumor volume, or potential areas of tumor, which would direct radia-
tion therapy better than CT alone26 (Fig. 9).

EVALUATING TREATMENT RESPONSE

PET/CT is valuable for evaluating the treatment response in human oncology patients
and can be used to determine if a patient is responding to a treatment protocol or if the
protocol needs to be altered. Evaluating the treatment response has also been inves-
tigated in veterinary medicine. FLT-PET/CT has been shown to depict the biological
tumor response to chemotherapy in dogs with lymphoma. The SUV max of lymphoid
tissues was significantly lower on the posttreatment scan compared with pretreatment
 PET-Computed Tomography 527

Fig. 8. (A) Fused transverse PET/CT image of a patient with OSA of the distal right femur.
This image shows a focal hypermetabolic area in the cortex of the left humerus (SUV max
5.0.) There were no abnormalities on CT images, and repeat PET/CT performed 2 weeks later
showed resolution of the hypermetabolic activity and a normal CT. The diagnosis was pre-
sumed focal trauma to the humerus before the first PET/CT. The repeat PET/CT helped
rule out metastasis to the humerus. (B) Fused PET/CT image using a lung window. The pa-
tient had a previously resected adrenal gland carcinoma and a current presumed meningi-
oma. There were multiple soft tissue nodules in the lungs that were mild to moderately
hypermetabolic. The nodule shown was 10 mm in diameter and had an SUV max of 3.7. His-
topathology of the nodules revealed benign granulomas. (C, D) Transverse fused PET/CT and
soft tissue window CT images of a dog with a maxillary hemangiosarcoma. There is a small,
hypermetabolic mass in the midventral mediastinum, adjacent to the heart. The SUV max
was 6.3. The CT images shows that the mass is a mix of soft tissue and fat attenuating tissue
and is heterogeneously contrast enhancing. Histopathology of the mass revealed focal gran-
ulomatous steatitis.

scans. This finding correlated with the patient clinical response. PET/CT also correctly
detected tumor recurrence before it was diagnosed clinically.5 FDG-PET/CT was used
to evaluate the response to toceranib phosphate (Palladia) in 6 dogs. There were
discordant findings between anatomic changes and metabolic changes on PET/CTs
performed at least 4 weeks after the initiation of Palladia treatment, and lack of histo-
pathology makes it difficult to determine whether the metabolic information from the
PET or the anatomic information from the CT better reflected the disease process in
the patients. However, the study supports the use of PET/CT to evaluate the treatment
528 Randall

Fig. 9. (A, B) Transverse CT and fused PET/CT images of a dog with nasal squamous cell car-
cinoma. The arrow is pointing to a portion of the large, mainly right-sided mass. The SUV
max was 13.5. There is also evidence of hyperattenuating soft tissue attenuating material
in the left nasal cavity at the same level. (C, D) Transverse CT and fused PET/CT images of
the midnasal cavity in the same dog, caudal to the large mass. PET/CT revealed hypermeta-
bolic activity in a small area of soft tissue attenuating material close to midline (arrows).
Based on the PET information, both the hypermetabolic area in the left nasal cavity and
the focal area in the midright nasal cavity were included in the radiation field in order to
prevent a geographic miss.

response or to restage patients and the need for histopathology of metabolic lesions
when possible (Fig. 10).

OTHER USES OF 2-DEOXY-2-18F-FLUORODEOXYGLUCOSE–PET/COMPUTED

TOMOGRAPHY

FDG-PET/CT can also be used to detect the source of fever of unknown origin
(FUO) in veterinary patients. It has a moderate success rate in humans (40%–
60%), but the degree of success in veterinary patients is still to be determined.27,28
One example of a successful study is a patient with FUO in which hypermetabolic
intrathoracic lymph nodes, pulmonary granulomas, and pulmonary interstitial dis-
ease were detected, guiding aspirates that led to a diagnosis of fungal disease
(Fig. 11).
FDG-PET/CT can also be used to find a source of lameness in patients that are diffi-
cult to diagnose. The study may detect muscular injury, joint abnormality, infection, or
neoplasia.
 PET-Computed Tomography 529

Fig. 10. Images from 2 PET/CT scans on a patient with a previous amputation for a tibial
OSA. (A) Transverse CT image in a soft tissue window showing a soft tissue mass in the
left side of the spinal canal, which is severely compressing the spinal cord. The spinal cord
(arrow) is displaced to the right. There is lysis and bone production associated with the
left pedicle of C7. (B) Fused PET/CT image at the same location as image (A) showing the hy-
permetabolic activity of the mass and abnormal bone, with an SUV max of 4.5. (C) Transverse
CT image in a bone window from a follow-up PET/CT performed 3 months after radiation
therapy for the presumed metastatic lesion of C7. The soft tissue component of the mass
is greatly reduced in size, and the compression of the cord has improved. Compression is
now mild and due to the smoothly mineralized portion of the mass that extends from
the pedicle. (D) Transverse fused PET/CT image at the same level as (C). The SUV max
decreased to 3.6. The patient’s clinical signs related to the C7 lesion (tetraparesis) also signif-
icantly improved after stereotactic radiation therapy.

PITFALLS OF 2-DEOXY-2-18F-FLUORODEOXYGLUCOSE–PET/COMPUTED
TOMOGRAPHY

As previously mentioned, there is overlap in the degree of hypermetabolic activity

between benign, inflammatory lesions and malignant lesions. Therefore, cytology or
histopathology should be performed when possible to confirm malignancy of
530 Randall

Fig. 11. (A, B) Dorsal and lateral maximum intensity projection (MIP) images of a Jack Rus-
sell terrier with a FUO. There are multiple areas of hypermetabolic activity in the thorax. (C,
D) Sagittal CT and fused PET/CT images. The study revealed hypermetabolic pulmonary nod-
ules (star), hypermetabolic, enlarged lymph nodes (small arrow pointing to trachea-
bronchial lymph nodes), and consolidated areas within multiple lung lobes (large arrow is
pointing to right caudal lung lobe). Disease was confined to the thoracic cavity and was
diagnosed as fungal disease (coccidioidomycosis). The hypermetabolic area in the cervical
region seen on MIP images is inflammation at the site of esophageal feeding tube
placement.

hypermetabolic lesions. False negatives are also possible. This result may happen in
tumors that are small, low grade, necrotic, have large amounts of mucin, or are simply
not as metabolically active.1
Aspirates of lesions that are detected before PET/CT should not be performed
within 1 to 2 days before the scan because the procedure can cause local inflamma-
tion that may be hypermetabolic on PET/CT images, which can confound interpreta-
tion. In addition, care must be taken when obtaining aspirates that are diagnosed on
PET/CT images. Ideally, aspirates should be performed on a separate day, when pa-
tient radioactivity is at background levels. Occasionally aspirates are obtained imme-
diately after PET/CT, while patients are still anesthetized, for lesions that would be
difficult to aspirate without anesthesia. In this case, only personnel with appropriate
radiation safety training should obtain the samples, as patients are still radioactive.
The samples also must be stored until they reach background levels of radioactivity
and patients reach institutional release levels.
PET/CT has lower spatial resolution compared with CT and MRI, related to the anni-
hilation process and technical factors of the PET/CT machine.8 PET technology can
detect lesions that are approximately 5 to 8 mm or greater. Therefore, lesions that
are in the range of 5 to 8 mm or less may be missed on visual or quantitative inspection
of images because of the lack of hypermetabolic activity.
 PET-Computed Tomography 531

There are many physiologic variants and artifacts that can complicate PET/CT
interpretation. Physiologic variants include variable uptake in intestines, variable
heterogeneity in liver and spleen, and muscle uptake. Artifacts include misregistra-
tion due to motion and vascular hypermetabolic activity upstream from the injection
site. Other areas that commonly have increased metabolic activity due to presumed
physiologic variation or inflammation include tonsils, gallbladder, nasal cavity,
and a region of linear activity lateral to the mandible.15 Experience reading canine
and feline PET/CTs allows the interpreter to recognize these areas of normal
variation.

SUMMARY

PET/CT is an imaging modality that is becoming increasingly available at veterinary

or human referral imaging sites. PET/CT scans typically include whole-body
imaging of patients with cancer and provide excellent anatomic and metabolic in-
formation for diagnostic and staging purposes. FDG-PET/CT has the ability to
detect metabolic changes before anatomic change is evident. Good candidates
for PET/CT are oncology patients who are healthy enough to undergo anesthesia
for 2 to 3 hours. It is important to remember that, although FDG-PET/CT is good
at detecting areas of increased metabolic activity, it is not specific for neoplasia,
and cytology or histopathology of hypermetabolic lesions should be pursued
when possible. In addition, strict cutoff values for SUV max have not been deter-
mined for specific malignancies in veterinary medicine; SUV max, although used
to guide the degree of suspicion for malignancy, cannot be used to definitively di-
agnose a malignant lesion.

REFERENCES

1. Workman R, Coleman RE. PET/CT essentials for clinical practice. In: Workman R,
Coleman RE, editors. New York: Springer; 2006. p. 1–22, 33–54.
2. Thrall D. Textbook of veterinary diagnostic imaging. 6th edition. St Louis (MO):
Elsevier; 2013.
3. Lynch TB. PET/CT in clinical practice. In: Lynch TB, editor. London: Springer-Verlag;
2007. p. 1–15.
4. LeBlanc AK, Jakoby B, Townsend DW, et al. Thoracic and abdominal organ up-
take of 2-deoxy-2-[18F] fluoro-D-glucose (18FDG) with positron emission tomog-
raphy in the normal dog. Vet Radiol Ultrasound 2008;49:182–8.
5. Lawrence J, Vanderhoek M, Barbee D, et al. Use of 30 -deoxy-30 -[18F] fluorothymi-
dine PET/CT for evaluating response to cytotoxic chemotherapy in dogs with non-
Hodgkin’s lymphoma. Vet Radiol Ultrasound 2009;50:660–8.
6. Ballegeer EA, Forrest LJ, Jeraj R, et al. PET/CT following intensity-modulated ra-
diation therapy for primary lung tumor in a dog. Vet Radiol Ultrasound 2006;47:
228–33.
7. Bruehlmeier M, Kaser-Hotz B, Achermann R, et al. Measurement of tumor hypoxia
in spontaneous canine sarcomas. Vet Radiol Ultrasound 2005;46:348–54.
8. Lawrence J, Rohren E, Provenzale J. PET/CT today and tomorrow in veterinary
cancer diagnosis and monitoring: fundamentals, early results and future per-
spectives. Vet Comp Oncol 2010;8:163–87.
9. Valdes-Martinez A, Kraft SL, Brundage CM, et al. Assessment of blood pool, soft
tissue, and skeletal uptake of sodium fluoride F 18 with positron emission
tomography-computed tomography in four clinically normal dogs. Am J Vet
Res 2012;73:1589–95.
532 Randall

10. Visvikis D, Costa DC, Croasdale I, et al. CT-based attenuation correction in the
calculation of semi-quantitative indices of [18F] FDG uptake in PET. Eur J Nucl
Med Mol Imaging 2003;30:344–53.
11. Boland GW, Blake MA, Holalkere NS, et al. PET/CT for the characterization of ad-
renal masses in patients with cancer: qualitative versus quantitative accuracy in
150 consecutive patients. AJR Am J Roentgenol 2009;192:956–62.
12. LEE MS, LEE AR, JUNG MA, et al. Characterization of physiologic 18F-FDG up-
take with PET-CT in dogs. Vet Radiol Ultrasound 2010;51:670–3.
13. LeBlanc AK, Wall JS, Morandi F, et al. Normal thoracic and abdominal distribution
of 2-deoxy-2-[18F] fluoro-d-glucose (18FDG) in adult cats. Vet Radiol Ultrasound
2009;50:436–41.
14. Jacques L, Jensen TS, Rollins J, et al. The Centers for Medicare and Medicaid
Services. Decision memo for positron emission tomography (FDG) for solid tu-
mors (CAG-00181R4). Baltimore (MD): United State Government; 2013.
15. Randall E, Loeber S, Kraft S. Physiologic variants, benign processes, and arti-
facts from 106 canine and feline FDG-PET/computed tomography scans. Vet Ra-
diol Ultrasound 2014;55:213–26.
16. Rowe JA, Morandi F, Wall JS, et al. Whole-body biodistribution of 30 -deoxy-30 -[18F]
fluorothymidine (18FLT) in healthy adult cats. Vet Radiol Ultrasound 2013;54:
299–306.
17. Ballegeer EA, Hollinger C, Kunst CM. Imaging diagnosis—multicentric lymphoma
of granular lymphocytes imaged with FDG PET/CT in a dog. Vet Radiol Ultra-
sound 2013;54:75–80.
18. Yoshikawa H, Randall EK, Kraft SL, et al. Comparison between 2-18F-fluoro-2-
deoxy-D-glucose positron emission tomography and contrast-enhanced
computed tomography for measuring gross tumor volume in cats with oral squa-
mous cell carcinoma. Vet Radiol Ultrasound 2013;54:307–13.
19. Leblanc AK, Jakoby BW, Townsend DW, et al. 18FDG-PET imaging in canine
lymphoma and cutaneous mast cell tumor. Vet Radiol Ultrasound 2009;50:
215–23.
20. LeBlanc AK, Miller AN, Galyon GD, et al. Preliminary evaluation of serial 18FDG-
PET/CT to assess response to toceranib phosphate therapy in canine cancer. Vet
Radiol Ultrasound 2012;53:348–57.
21. Hansen AE, McEvoy F, Engelholm SA, et al. FDG PET/CT imaging in canine can-
cer patients. Vet Radiol Ultrasound 2011;52:201–6.
22. Lee AR, Lee MS, Jung IS, et al. Imaging diagnosis—FDG-PET/CT of a canine
splenic plasma cell tumor. Vet Radiol Ultrasound 2010;51:145–7.
23. Randall EK, Kraft SL, Yoshikawa H, Larue SM. Evaluation of 18F-FDG PET/CT as a
diagnostic imaging and staging tool for feline oral squamous cell carcinoma. Vet
Comp Oncol 2013. [Epub ahead of print].
24. Seiler SM, Baumgartner C, Hirschberger J, et al. Comparative oncology: evalua-
tion of 2-deoxy-2-[18F] fluoro-D-glucose (FDG) positron emission tomography/
computed tomography (PET/CT) for the staging of dogs with malignant tumors.
PLoS One 2015;10:e0127800.
25. Mann K, Kraft SL, Hauke SM, et al. Quantitative analysis of 18F-fluorodeoxyglu-
cose standardized uptake values in pretreatment PET/computed tomography im-
aging of canine osteosarcomas. Proceedings, American College of Veterinary
Radiology Annual Scientific Conference. St Louis, MO, October 21–24, 2014.
(Available from Veterinary Radiology and Ultrasound).
26. Loeber SJ, Custis JT, Randall EK, et al. Incorporation of FDG-PET/CT into radia-
tion therapy planning to improve treatment of canine nasal tumors. Proceedings,
 PET-Computed Tomography 533

American College of Veterinary Radiology Annual Scientific Conference, St Louis,

MO, October 21–24, 2014. (Available from Veterinary Radiology and Ultrasound).
27. Kei PL, Kok TY, Padhy AK, et al. [18F] FDG PET/CT in patients with fever of un-
known origin: a local experience. Nucl Med Commun 2010;31:788–92.
28. Pelosi E, Skanjeti A, Penna D, et al. Role of integrated PET/CT with [(1)(8)F]-FDG
in the management of patients with fever of unknown origin: a single-centre expe-
rience. Radiol Med 2011;116:809–20.