These authors contributed equally to this work as co-first authors.
Retrospective study.
This study aims to semiquantitatively evaluate the standardized uptake value (SUV) of 99mTc-methylene diphosphonate (MDP) radionuclide tracer in the normal vertebrae of breast cancer patients using an integrated single-photon emission computed tomography (SPECT)/computed tomography (CT) scanner.
Molecular imaging techniques using gamma cameras and stand-alone SPECT have traditionally been utilized to evaluate metastatic bone diseases. However, these methods lack quantitative analysis capabilities, impeding accurate uptake characterization.
A total of 30 randomly selected female breast cancer patients were enrolled in this study. The SUV mean (SUVmean) and SUV maximum (SUVmax) values for 286 normal vertebrae at the thoracic and lumbar levels were calculated based on the patients’ body weight (BW), body surface area (BSA), and lean body mass (LBM). Additionally, 106 degenerative joint disease (DJD) lesions of the spine were also characterized, and both their BW SUVmean and SUVmax values were obtained. A receiver operating characteristic (ROC) curve analysis was then performed to determine the cutoff value of SUV for differentiating DJD from normal vertebrae.
The mean±standard deviations for the SUVmean and SUVmax in the normal vertebrae displayed a relatively wide variability: 3.92±0.27 and 6.51±0.72 for BW, 1.05±0.07 and 1.75±0.17 for BSA, and 2.70±0.19 and 4.50±0.44 for LBM, respectively. Generally, the SUVmean had a lower coefficient of variation than the SUVmax. For DJD, the mean±standard deviation for the BW SUVmean and SUVmax was 5.26±3.24 and 7.50±4.34, respectively. Based on the ROC curve, no optimal cutoff value was found to differentiate DJD from normal vertebrae.
In this study, the SUV of 99mTc-MDP was successfully determined using SPECT/CT. This research provides an approach that could potentially aid in the clinical quantification of radionuclide uptake in normal vertebrae for the management of breast cancer patients.
The skeleton has been identified as the most frequent site for tumors to metastasize to after the treatment of primary breast cancer [
In addition to initial investigation methods such as blood cell count tests, bone scans (i.e., bone scintigraphy) can also be conducted for accurate early detection of bone metastases. A bone scan is a type of nuclear medicine-based imaging that provides an entire skeletal visualization based on the activity of a radionuclide tracer within a short amount of time. Two of the most widely used bone scan instruments are gamma cameras and single-photon emission computed tomography (SPECT), both of which are used with 99mTc-methylene diphosphonate (MDP) tracer [
In recent years, the introduction of SUVs into the SPECT/CT method has offered a more accessible tool for diagnostic purposes in comparison with the expensive and less available positron emission tomography (PET) method. In general, an SUV is a semiquantitative biomarker that assesses the tissue concentration of a radionuclide tracer measured by the scanner and divides it by the activity injected divided by body size [
Furthermore, few researchers have studied the measurement of SUV in SPECT/CT bone scans with 99mTc tracers thus far [
Patient data from a single center were retrospectively analyzed. Ethical approval, including a written informed consent waiver, was given by the institution’s Human Research Ethics Committee (USM/JEPeM/18110694) following the standards of the Declaration of Helsinki. A total of 30 female patients (mean age, 52.6±9.7 years; age range, 30–70 years) positively diagnosed with breast cancer who had undergone a 99mTc-MDP bone scan via SPECT/CT from January 2018 to December 2018 were randomly selected based on the following inclusion criteria: (1) a SPECT/CT procedure for the thoracic spine (T1–T12) and lumbar spine levels (L1–L5) was conducted 3 hours after intravenous 99mTc-MDP administration; (2) data available on the patient’s weight and height; (3) and data available on the injection activity and times of measurement and injection.
The bone scans were performed using a SPECT/CT system (Discovery NM/CT 670 Pro; GE Healthcare, Chicago, IL, USA) equipped with a low-energy high-resolution (LEHR) collimator 3 hours after an intravenous injection of 736.3±22.2 MBq (19.9±0.6 mCi) 99mTc-MDP, corresponding to 12.21±1.85 MBq (0.33±0.05 mCi/kg). A planar scan over the anterior and posterior regions of the whole body was first performed using the LEHR collimator at 256×256 pixels (
The following set of exclusion criteria was first outlined to determine the normal vertebrae (i.e., healthy bone): compression fractures, diffuse bone metastasis, ankylosing spondylitis, and metabolic bone disease. In addition, DJD of the spine was also characterized by correlating the tracer uptake with its CT morphology, which encompassed osteophytes, end plates, facet joints, and the area around the joints [
Information regarding camera sensitivity, patient data, activity injected, administration, and scan time was acquired and loaded into the Q.Metrix software before the SUV assessment. The calculated camera sensitivity was determined as 176.18 counts per minute per microcurie (cnt/min/μCi). The SUVs were expressed as the SUV mean (SUVmean) and SUV maximum (SUVmax). The SUVmean was the average multiple pixels within the VOI sampled, while the SUVmax was defined as the highest pixel value within the VOI with the highest activity concentration:
Note that the SUVs listed above accounted for body size or body weight (BW) measurements only. To reduce the dependence of the SUV on BW, the values for lean body mass (LBM) and body surface area (BSA) were also examined and calculated by replacing the BW parameter in both the SUVmax and SUVmean [
All statistical analyses were performed using the IBM SPSS ver. 24.0 software (IBM Corp., Armonk, NY, USA). The degrees of dispersion of the SUVmean and SUVmax for the non-rejected vertebrae (or the noise signal amount [
Of the 30 patients included in this study, 18 (treatment group; 60%) were identified to have undergone either a single cancer treatment or combination of treatments (chemotherapy, radiotherapy, and hormonal therapy), while the remaining 12 (no-treatment group; 40%) had not received any cancer treatment prior to their bone scan procedure. From the SPECT/CT scans, 286 normal vertebrae from T1 to T12 and L1 to L5 were obtained. No vertebral levels were excluded in the statistical analyses, as the sampling number was deemed adequate (n≥10). This criterion was accomplished in the same manner as the statistical considerations by Kaneta et al. [
The box and whisker plots of the SUVmean and SUVmax for BW, BSA, and LBM at each vertebral level are shown in
The mean±SD of the BW SUVmean and SUVmax in the no-treatment group (120 vertebrae) was 3.23±1.61 and 5.37±2.81, respectively. Comparatively, an independent
A total of 106 DJD lesions of the spine were identified. The mean±SD of the BW SUVmean and SUVmax was 5.26±3.24 and 7.50±4.34, respectively. An ROC curve analysis was performed to determine the diagnostic accuracy of both SUVs in discerning DJD from normal vertebrae (
Metastases of the bone are known to not show any discernible uptake patterns during scanning [
Most SUVs are measured based on BW, since this is the most popular method. In this study, the BW skeletal SUVmean and SUVmax were relatively low at 3.92±0.27 and 6.51±0.72, respectively. These values were regarded as commensurable to previously reported BW SUVs for normal vertebrae, e.g., 4.4±0.5, 4.6±1.7, and 5.9±1.5 for SUVmean, and 7.1±0.4 and 7.6±2.4 for SUVmax [
In this study, based on the Spearman correlation test, it was demonstrated that all SUVs showed a weak and no significant correlation with age (
A significant negative correlation between the BW SUVmean with height was also found in this study. Bone density composition, in general, is higher in taller subjects compared with smaller subjects. This increase of bone density has been suggested to be a result of the rise in physical burden due to the high center of gravity [
Breast cancer therapy can affect normal bone homeostasis by altering the osteoclast and osteoblast functions [
Degenerative changes of the spine, especially in the elderly, are common and often mistaken for bone metastases during bone scans. Therefore, SUV acquisition could potentially be a useful parameter to differentiate bone lesions and normal vertebrae. In a recently published paper by Mohd Rohani et al. [
The implementation of quantitative SPECT/CT is deemed to be more challenging compared with PET imaging due to the former’s many technical limitations. Regarding bone SPECT/CT scans, quantitative uncertainties may appear because of tracer-drug interactions (e.g., with iron supplements), metal-induced artifacts from prostheses affecting the computed SUV, variations in hormone levels such as estrogen, and unknown rates of uptake and clearance from the blood [
In sum, SUVs can be used as a parameter to quantify tracer uptake in normal vertebrae SPECT/CT studies in breast cancer patients. In this research, the SUVmean and SUVmax values calculated based on the patients’ BW, BSA, and LBM for normal vertebrae showed a wide variability, though they were comparable to other published results. More effort needs to be put into the clinical interpretation of SUVs to understand the quantitative nature of tracer uptake before it can be used routinely in the clinical setting.
No potential conflict of interest relevant to this article was reported.
Conception and design: NMN, MFMR; data acquisition: YBU, SNBMY, MFMR, NT; analysis of data and drafting of manuscript: MFMR, SNBMY, NT; revision: NMN, WMNMZ; and supervision: NMN, WMNWZ
This study was funded by the short-term research grant of Universiti Sains Malaysia (304/PPSP/6315121). The study was approved by the Human Research Ethics Committee USM (HREC), Universiti Sains Malaysia (reference: USM/JEPeM/18110694).
(A–D) Normal bone planar scintigraphy. R, right; L, left.
(A–K) Coronal, sagittal, and trans-axial images of patients’ single-photon emission CT/CT-fused datasets. The different colors depict the different selected volumes of interest. CT, computed tomography; R, right; L, left; A, anterior; P, posterior.
Box and whisker plot of the SUVmean (A) and SUVmax (B) for BW showing a quantitative distribution of standard statistics. SUV, standardized uptake value; BW, body weight; SD, standard deviatio; CoV, coefficient of variation.
Box and whisker plot of SUVmean (A) and SUVmax (B) for BSA showing a quantitative distribution of standard statistics. SUV, standardized uptake value; BSA, body surface area; SD, standard deviation; CoV, coefficient of variation.
Box and whisker plot of the SUVmean (A) and SUVmax (B) for LBM showing a quantitative distribution of standard statistics. SUV, standardized uptake value; LBM, lean body mass; SD, standard deviatio; CoV, coefficient of variation.
Receiver operating characteristic curves to differentiate between degenerative joint disease and normal vertebrae using the SUVmean (black line) and SUVmax (red line). SUV, standardized uptake value; AUC, area under the curve; CI, confidence interval.
CoV for normal vertebrae in breast cancer patients
Vertebral level | No. | Body weight |
Body surface area |
Lean body mass |
|||
---|---|---|---|---|---|---|---|
Suvmean | Suvmax | Suvmean | Suvmax | Suvmean | Suvmax | ||
T1 | 13 | 0.354 | 0.391 | 0.358 | 0.388 | 0.349 | 0.385 |
T2 | 13 | 0.376 | 0.402 | 0.360 | 0.404 | 0.356 | 0.402 |
T3 | 16 | 0.460 | 0.488 | 0.461 | 0.501 | 0.457 | 0.501 |
T4 | 14 | 0.495 | 0.541 | 0.468 | 0.539 | 0.468 | 0.546 |
T5 | 15 | 0.595 | 0.599 | 0.571 | 0.583 | 0.571 | 0.591 |
T6 | 18 | 0.479 | 0.509 | 0.462 | 0.495 | 0.462 | 0.501 |
T7 | 20 | 0.477 | 0.472 | 0.459 | 0.451 | 0.456 | 0.453 |
T8 | 19 | 0.452 | 0.518 | 0.436 | 0.521 | 0.437 | 0.523 |
T9 | 18 | 0.476 | 0.535 | 0.462 | 0.529 | 0.462 | 0.534 |
T10 | 18 | 0.443 | 0.422 | 0.434 | 0.422 | 0.434 | 0.424 |
T11 | 20 | 0.352 | 0.339 | 0.358 | 0.326 | 0.361 | 0.329 |
T12 | 23 | 0.354 | 0.340 | 0.359 | 0.351 | 0.357 | 0.348 |
L1 | 26 | 0.339 | 0.351 | 0.337 | 0.365 | 0.334 | 0.360 |
L2 | 18 | 0.306 | 0.304 | 0.359 | 0.299 | 0.307 | 0.299 |
L3 | 14 | 0.341 | 0.341 | 0.333 | 0.321 | 0.335 | 0.324 |
L4 | 11 | 0.247 | 0.339 | 0.236 | 0.336 | 0.242 | 0.346 |
L5 | 10 | 0.255 | 0.408 | 0.221 | 0.203 | 0.221 | 0.205 |
Mean CoV | 0.400 | 0.430 | 0.390 | 0.414 | 0.389 | 0.416 |
CoV, coefficient of variation; SUV, standardized uptake value.
Correlation coefficients between SUVs and age, weight, height
Variable | Body weight |
Body surface area |
Lean body mass |
|||
---|---|---|---|---|---|---|
Suvmean | Suvmax | Suvmean | Suvmax | Suvmean | Suvmax | |
Age | 0.239 | 0.101 | 0.207 | 0.118 | 0.221 | 0.124 |
Weight | 0.158 | 0.457 |
-0.160 | 0.136 | -0.166 | -0.008 |
Height | -0.382 |
-0.192 | -0.315 | -0.238 | -0.299 | -0.073 |
SUV, standardized uptake value.