Abstract
Background: The objective of this study was to characterize tumor activity and mineralization status in newly-detected multiple myeloma (MM) bone lesions using 2-18F-fluoro-2-deoxy-D-glucose (18F-FDG)-PET/CT and 18F-sodium fluoride (18F-NaF)-PET/CT before and after antitumor treatment. Materials and Methods: In this retrospective study, seven patients with histologically-verified MM were included (four women, three men; median age=57 years, standard deviation=11.23 years). PET/CT was performed with 18F-FDG and with 18F-NaF, both at baseline and after treatment. All patients had positive scans. Volumes of interest (VOIs) were drawn over all 18F-FDG-PET/CT-positive bone lesions, as well as the corresponding regions in 18F-NaF-PET/CT. For characterization of bone lesions, semi-quantitative standard uptake value (SUV) parameters were measured. Results: 18F-FDG-PET/CT in the seven patients detected 39 metabolically active lesions that were correlated with the corresponding sites in 18F-fluoride-PET/CT. Overall, the lesions showed a response to therapy, with a significant decrease in SUVmax on PET/CT using 18F-FDG (p<0.001) and with 18F-NaF (p<0.001). In four patients with a second follow-up scan (at a median of 17 months after baseline scan), there was no significant change in lesion uptake. Conclusion: Based on our data, antitumor therapy in MM reduces not only tumor activity, but also the mineralization status of bone lesions. A second follow-up scan in a subset of the cohort yielded no change in mineralization status.
Multiple myeloma (MM) is a neoplastic plasma cell disorder that is characterized by clonal proliferation of malignant plasma cells in the bone marrow (1). Eighty percent of patients experience bone lesions during the course of their disease (1, 2), with focal osteolytic bone lesions being the radiographic hallmark of the disease (3). Changes in the bone microenvironment leads to increased recruitment and activation of osteoclasts and inhibition of osteoblasts (4, 5).
2-18F-Fluoro-2-deoxy-D-glucose (18F-FDG)-PET/CT is a sensitive functional imaging modality for many neoplasm types. The updated International Myeloma Working Group (IMWG) criteria consider focal skeletal lesions with increased 18F-FDG uptake and with osteolytic destruction in the CT component indicative of active myeloma (6). 18F FDG-PET/CT appears to be especially useful in the evaluation of the quality of treatment response in patients with MM (7-11).
18F-Sodium fluoride (18F-NaF) is a PET tracer used for skeletal imaging which reflects regional blood flow and bone remodeling and can accumulate in osteoblastic and osteolytic lesions (12-14). Although 18F-NaF is a well-known tracer, its use in PET/CT is considered an intriguing imaging method for the assessment of malignant bone diseases (15-18). Early studies have shown the possibility of using 18F NaF PET/CT as a valuable tool for diagnosis in MM (19-21). However, several studies have shown this method to have limited performance in the evaluation of myeloma bone disease (22-25). Data about the therapy response, as assessed by 18F-NaF PET/CT is, however, still very limited.
It is well known that the behavior of MM bone lesions changes under therapy (26). The hypothesis of the present study was that in response to therapy, MM bone lesions show increased bone turnover (mineralization) and reduced tumor activity; these changes can be detected with 18F-NaF-PET/CT and 18F-FDG-PET/CT. We expected to detect significant differences in the mineralization of MM bone lesions using 18F-NaF-PET/CT and in tumor activity using 18F-FDG-PET/CT for MM bone lesions before and during therapy.
The aim of this study was to characterize newly-detected MM bone lesions with regard to mineralization status and tumor activity using 18F-/18F-NaF-FDG-PET/CT and conduct a re-evaluation during therapy with these same diagnostic modalities to detect potential changes in the mineralization and tumor activity in MM bone lesions.
Materials and Methods
Patients. Seven patients (four women, three men; median (±standard deviation) age=57±11.23 years; age range=43-78 years) with newly-diagnosed MM, according to the International Myeloma Working Group (IMWG) criteria (6), were retrospectively recruited into the present study. All patients underwent baseline imaging (before anticancer therapy), including standard 18F-FDG-PET/CT with a diagnostic CT component as well as additional 18F-NaF-PET/CT with a low-dose CT component. PET/CT studies were performed within 1 week of each other (five patients underwent both studies on two consecutive days, the remaining 2 within 7 days).
Follow-up imaging, including 18F-NaF-PET/CT and 18F-FDGPET/CT, was performed after the introduction of anticancer therapy. At least one follow-up study was available for all patients within a median time-frame of 10 months (range=7-13 months). Additional follow-up scans were available for four patients within a median time-frame of 17 months after the baseline scan (range=12-22 months).
This study was approved by the local Ethics Committee (EK-Nr: 1845/2017). Informed consent was obtained from all individual participants included in the study.
Imaging techniques. Both PET tracers, 18F-NaF and 18F-FDG, were prepared on site using standardized protocols on a GE FASTlab synthesizer (GE Healthcare, Boston, MA, USA) with dedicated disposable cassettes on the day of PET/CT imaging. Full radiopharmaceutical quality control according to the monographs of the European Pharmacopoeia was performed prior to product release and patient application.
18F-NaF-PET/CT. PET/CT was performed using a modern hybrid scanner, a Siemens Biograph 64 True Point (Siemens Healthcare Sector, Erlangen, Germany). Helical CT slices and PET emission data were acquired from the skull to the foot in all patients. Image acquisition of emission data was started 30-60 min after i.v. injection of 185-370 MBq 18F-NaF. A low-dose CT scan was performed using the following acquisition parameters: 120 kV, 30-40 mAs; slice thickness 1.2 mm; increment 0.7 mm. To match CT slices and PET slices, CT was acquired with the patient in a shallow breathing position. Directly after CT imaging, the PET acquisition protocol was started. Acquisition time was 3 min per bed position (10-12 bed positions per patient). During imaging of the thorax, patients were instructed to maintain shallow breathing. PET images were reconstructed using a CT attenuation correction technique. The reconstruction was performed using the manufacturer's arithmetic reconstruction algorithms (“Iterative 2D uncor. recon” and/or “TrueD cor. recon”) in their respective default settings. CT images were converted into linear attenuation coefficients for the 511-keV energy radiation, as implemented in the system. CT and PET images were matched and fused into transaxial and coronal images of 5-mm thickness.
18F-FDG-PET/CT. Using the same hybrid scanner system, data were also acquired from the skull to the foot in all patients. All patients were in a fasting condition for at least 8 h prior to 18F-FDG injection. Serum glucose levels had to range from 80 to 160 mg/dl. Acquisition of emission data was started 60 min after i.v. injection of 300-400 MBq of 18F-FDG. A diagnostic CT without contrast medium was performed with the following parameters: 120 kV, 200-230 mAs; slice thickness 3 mm; increment 2 mm. The PET protocol was similar to that of 18F-NaF-PET/CT, with an acquisition time of 3-4 min per bed position (8-12 bed positions per patient). PET and CT images were also reconstructed as described above. CT and PET images were matched and fused into transaxial and coronal images of 5 mm thickness.
Image analysis. Two experienced nuclear medicine physicians analyzed 18F-FDG- and 18F-NaF-PET/CT based on a visual (qualitative) analysis and a semi-quantitative evaluation based on standard uptake value (SUV) calculations.
Qualitative assessment was based on increased focal uptake of 18F-FDG within the bone for which it was possible to exclude benign etiologies (trauma, inflammation, degenerative changes, arthritic disease, etc.) and these lesions were considered indicative of MM. 18F-FDG-PET findings were correlated with the CT component, and, in particular, clearly delineated osteolytic lesions were considered proof of myelomatous lesions according to IMWG guidelines (6). In addition, increased homogenous, diffuse bone marrow uptake of 18F-FDG on maximum intensity projections was considered indicative of myeloma. To minimize false-positive assessment, patient history was studied thoroughly. 18F-FDGPET/CT-positive lesions were correlated with those of 18F-NaF-PET/CT and served as a reference standard for this study, as there is no systemized standard for the evaluation of myelomatous lesions with this modality.
Semi-quantitative assessment was based on volumes of interest (VOIs) drawn semi-automatically using threshold-based delineation provided by the image analysis software (Hybrid 3D™ Viewer; Hermes Medical Solutions, Stockholm, Sweden), which was adjusted manually for optimal correlation with osteolytic bone destruction at the initial scan. Due to the very good response shown by the majority of patients after treatment, no focal 18F-FDG-uptake was detectable in some of the patients at follow-up. In these cases, cubic VOIs with a volume of 3 ml were placed either within osteolytic lesions detectable on CT or according to the anatomy of the initially metabolically active lesion.
Reference regions for the blood pool in the upper mediastinum and for bone metabolism using the fourth lumbar vertebra (L4), or iliac bone, if the previous region was affected, were drawn as cubic volumes of 10 ml and 3 ml, respectively. These VOIs served as reference tissue values for 18F-FDG-PET and 18F-NaF-PET lesions, respectively, to calculate the tumor-to-background ratio (TBR).
Patient-based therapy assessment was conducted according to the European Organization for Research and Treatment of Cancer (EORTC) 1999 criteria, which define four treatment response groups: Complete response (CR, with complete resolution of 18F-FDG uptake indistinguishable from the surroundings); partial response (PR); stable disease (SD); and progressive disease (PD) (27).
Clinical and histological data acquisition. Data about therapy, plasma cell infiltration status, and laboratory parameters were collected at the initial PET/CT scan and at each follow-up. Clinical therapy response was assessed according to the modified IMWG response criteria for MM (28).
Statistical analysis. Data were statistically evaluated using SPSS 24.0 (IBM, Armonk, NY, USA) software. The statistical evaluation was performed using descriptive statistics and a mixed model analysis of variance (ANOVA), as well as Pearson's correlation. Results were considered significant for p-values less than 0.05 (p<0.05).
Results
Seven patients (four women, three men; median age=57±11.23 years) newly diagnosed with MM were retrospectively recruited between March 2013 and October 2014 at the General Hospital of Vienna. All patients were followed-up for at least 30 months (median=45 months, range=21-53 months), with the exception of one patient who died 9 months after the first follow-up scan (Table I).
The initial stage, according to the Revised International Staging System for Multiple Myeloma (R-ISS) (29), was stage 1 for five patients and stage 2 and stage 3 for one patient. For one patient with R-ISS stage I, serological data to determine the stage was found only after the induction of anticancer therapy. Initial bone marrow biopsy in the iliac bone was positive for all patients, with one patient demonstrating a grade 1 plasma cell myeloma and the remaining six patients demonstrating a grade 2 plasma cell myeloma. The bone marrow showed a median infiltration of 60±23.7% (range=20-90%). Four patients were classified as having light-chain myelomas (two ĸ- and two λ-light-chain myelomas each) and the remaining as IgG myelomas (two ĸ- and one λ-IgG myelomas) (Table I).
After the initial PET/CT assessments, all patients underwent immunochemotherapy with a bortezomib-containing regimen. Four patients received additional local radiotherapy (57.1%) and three patients received autologous stem cell transplantation (ASCT) (42.9%).
After anticancer therapy, all patients showed either a very good PR or CR according to the IMWG guidelines (28), two and five patients, respectively. Bone marrow infiltration after therapy was reported to be either <5% or at 5%, with histological data missing for one patient (Table I).
The four patients who underwent a second follow-up scan all received different therapies after the first follow-up scan. One patient initially underwent ASCT. One patient received consolidating immunotherapy, while another received a second dose of immunochemotherapy. The last patient received no treatment. Three patients showed CR to the therapy. The fourth patient who underwent a second treatment-cycle of immunochemotherapy initially showed CR but presented with serological relapse before the second follow-up scan (Table I).
Baseline 18F-FDG-PET/CT revealed 34 lesions in six patients. One patient showed diffuse bone marrow uptake, with no clearly defined focal lesions. Five areas of interest were then determined based on Positron-Emission Tomography Response Criteria in Solid Tumors (PERCIST) criteria (30), which determines a set number of five lesions for therapy response assessment based on peak standard uptake values using lean body mass (SULpeak).
18F-NaF-PET/CT showed correlating focal uptake in 28 lesions in six patients. Other patterns of uptake were diffuse pattern, correlating the 18F-FDG-PET/CT in the patient with diffuse bone marrow uptake. Two lesions were 18F-NaF-negative, one within the marrow of the proximal femur of one patient, the other in the sacrum of another patient. Finally, elevated 18F-NaF in the bone adjacent to the FDG-avid focus was seen in four lesions of one patient.
The four patients with second follow-up PET/CTs initially had a total of 21 18F-FDG-positive lesions, including one patient with only one lesion and the aforementioned patient with diffuse bone marrow uptake.
After the initial treatment cycle, therapy response assessment by 18F-FDG-PET/CT according to EORTC guidelines (27) showed CR in one patient, PR in three patients, SD in two patients and PD in one patient (Figure 1). The patient with PD showed serological CR and displayed reduced 18F-FDG uptake but a new paravertebral lesion was discovered on the follow-up scan. Of the four patients with a second follow-up scan, three patients showed SD compared to the first follow-up having initially shown either PR or SD. The patient with PD initially showed an increase in 18F-FDG uptake in the known lesions compared to that at the first follow-up and additionally displayed newly diagnosed extramedullary disease in the vesical urinaria, thus showing PD once more. However, the SUV values for this patient did not reach the levels of the initial baseline scan.
In 18F-NaF-PET/CT treatment responses based on EORTC guidelines (27) showed partial response in all but one patient. Patient 7 showed PD with an actual increase in mineralization after treatment with radiotherapy and immuneochemotherapy. In the four patients with a second follow-up, the scans all showed SD compared to the first follow-up.
In addition to SUV values based on body weight, SUV based on body surface area [as recommended in the EORTC guidelines (27), as well as SUV based on lean body mass [SUL, as recommended in the PERCIST guidelines (30)], were measured. For 18F-FDG-PET/CT, SUVmax and SUVpeak were measured, as a reproducible SUVmean would not have been achievable in patients with CR after initial therapy. For 18F-NaF-PET/CT, SUVmax, SUVpeak, and SUVmean were measured, as many lesions still showed focal uptake after the initial anticancer therapy. Correlation of SUVmax with SUVpeak and SUVmean (in the case of 18F-NaF-PET/CT) yielded a high correlation of >0.9 in almost all cases except one. Similarly, SUV values based on body weight, body surface area, and lean body mass also showed a high correlation. For this reason, we decided to use only SUVmax based on body weight in our descriptive analysis, as it is the most commonly used parameter in the literature.
A mixed model ANOVA showed a significant decrease of SUVmax in the myeloma lesions before and after therapy on 18F-FDG-PET/CT (p<0.001). The average SUVmax at the initial scan was 7.19±3.15. After therapy, the mean SUVmax decreased to 3.96±2.78. 18F-NaF-PET/CT also showed a significant decrease in SUVmax (p<0.001). Mean SUVmax decreased from 41.71±26.58 to 21.95±18.44. Pearson's correlation showed a significant but weak positive correlation between relative changes in SUVmax between 18F-FDG-PET/CT and 18F-NaF-PET/CT, with a correlation coefficient of 0.363 (p=0.023). No significant changes in SUVmax were found between the first and second follow-up on either 18F-FDG- or 18F-NaF-PET/CT (p>0.99) (Figure 2).
When correcting for background activity using the upper mediastinum as the blood pool reference for 18F-FDG and, when available, L4 or the left iliac bone for 18F-NaF as normal bone uptake, there remained a significant decrease between baseline and the first follow-up (p<0.001 for both modalities). One patient with a diffuse pattern of uptake in the bone marrow had to be excluded from TBR calculations for 18F-NaF because it was not possible to establish healthy bone reference tissue in either the lumbar spine nor the hip. The TBR at the first and second follow-up continued to show no significant difference (p>0.99). The average TBR at baseline was significantly different from those at the first and second follow-ups but did not differ significantly between the follow-up studies in either 18F-FDG-PET/CT and 18FNaF-PET/CT as seen in Table II.
In 18F-FDG-PET/CT, the standard deviation at baseline for intra-patient SUVmax was lower than that for the combined lesion pool (1.20-1.96 compared to 3.15), suggesting a more homogenous lesion activity for each patient. 18F-NaF-PET/CT displayed a wider range of standard deviations from 3.91-37.23. Treatment response in both modalities tended toward a lower standard deviation but notable exceptions were present in both. While treatment response was displayed in all but one patient and 18F-FDG and 18F-NaF uptake decreased significantly, not all lesions displayed similar responses when looked at individually. In the first follow-up, all but one initially 18F-FDG-avid lesion showed reduced uptake to varying degrees. In 18F-NaF, six lesions displayed increased uptake including the lesion that had increased 18F-FDG uptake. Two lesions in patient 7 showed a marked increase in SUVmax, leading to this patient being evaluated as having PD. The four patients with a second follow-up study displayed a much more varied response for each individual lesion in that follow-up.
Discussion
MM is the second most prevalent hematological malignancy after non-Hodgkin's lymphoma (31) and many studies have shown 18F-FDG-PET/CT to be useful, especially in evaluation of treatment response, which led to its inclusion in IMWG guidelines (6). 18F-NaF-PET/CT, however, has only recently been studied as a potential diagnostic tool for the detection of myelomatous lesions, mostly with disappointing results (22-25, 32). There are only limited data about the changes in 18F-NaF uptake before and after therapy.
In our study, there was a significant reduction in 18F-NaF uptake after first-line therapy. Subsequent follow-up scans did not show any significant changes in either 18F-FDG-PET/CT or 18F-NaF-uptake. However, TBRs (using bone and the blood pool for 18F-NaF-PET/CT and 18F-FDG-PET/CT, respectively) continued to show an increased tracer uptake in 18F-NaF-PET/CT, indicating reduced but active mineralization processes after treatment. It is well known that the bone microenvironment is altered in the course of MM, leading to an environment that is osteoblast-inhibiting, osteoclastogenic, and in which osteoclast activity is promoted (4, 5, 33). However, the mechanisms that underlie the persistence of osteolytic lesions, even in responding patients, and the continued suppression of osteoclasts, are not well understood. In our study, we showed an effect of therapy on bone mineralization, but no significant changes between subsequent follow-ups even in patient 5 (one lesion) who received no further therapy after the initial regimen (SUVmax 9.54 and 10.52 at the first and second follow-ups, respectively).
Two studies have looked into 18F-NaF-PET/CT in MM before and after therapy. In 2017, Wang et al. released initial data describing 18F-NaF uptake as a surrogate marker for bone metabolism in patients undergoing treatment with Dickkopf-related protein 1 (DKK1)-inhibiting monoclonal antibodies. A small number of patients underwent static and dynamic 18F-NaF-PET/CT as well as dual-energy X-ray absorptiometry before and after six cycles of DKK1-inhibiting monoclonal antibody therapy. The mean SUV in the lumbar spine and left hip was measured in static PET/CT and tracer accumulation in the 10th thoracic vertebra in dynamic PET/CT. The results showed a stable or increasing uptake of 18F-NaF after six cycles of therapy and an increase in bone mineral density after therapy. No pathological lesions were measured (34). Comparisons with our study show similar average SUVmean values for the fourth lumbar vertebra at baseline scan but with a higher range [4.60 (range=0.59-8.53 compared to 5.0 (range=4.0-6.9)]. After anticancer therapy, SUVmean values dropped in our cohort while a rise was noted in the cohort treated with the DKK1-inhibiting monoclonal antibody [3.81 (range=1.59-7.27) compared to 5.6 (range=4.4-7.2)]. Another point of difference is the fact that our patient cohort was treated with first line therapy while disease in all the patients receiving the antibody was refractory or relapsed with at least one line of therapy (34). Interestingly, the four patients with a second follow-up scan and vastly different treatment regimens showed average SUVmean values and ranges similar to those of the baseline scan [4.68 (range=0.55-8.17)].
In 2017, Sachpekidis et al. evaluated 29 patients using static and dynamic 18F-FDG-PET/CT and 18F-NaF-PET/CT before and after high-dose chemotherapy and ASCT. 18F-FDG-positive lesions were used as a gold standard and were correlated to 18F NaF-positive lesions in the same way as in our study. Therapy assessment was made using EORTC guidelines (27) for both 18F-FDG-PET/CT and 18F-NaF-PET/CT. 18F-NaF-PET/CT revealed significantly fewer myelomatous lesions when compared to 18F-FDG-PET/CT, a result that was not reproducible in our study. Overall changes in therapy response were less pronounced than in our study, with the majority of patients showing SD and a number of patients showing PD that correlated with either PR or PD on 18F-FDG-PET/CT. In our study, using the same guidelines for 18F-NaF-PET/CT as for 18F-FDG-PET/CT, all of our patients except one showed PR. Patient 7 showed a marked increase in 18F-NaF uptake in two lesions while exhibiting reduced uptake in the other two lesions, as well as in the reference tissue. Patient 7 showed SD in the corresponding 18F-FDG-PET/CT. Semi-quantitative (SUVmax) and kinetic parameters showed a significant decrease after therapy, similarly to our study, but to a lesser extent (25). A marked difference in the two study cohorts is the larger number of patients in the study of Sachpekidis et al. and the uniform therapy regimen. All patients received high-dose chemotherapy with melphalan and ASCT before the second scan. In our study, only three patients received ASCT before the first scan while all received immunochemotherapy, including bortezomib, as the first-line treatment. Treatment response in 18F-FDG-PET/CT was 48.3% CR, 37.9% PR, and 13.8% PD in the study of Sachpekidis et al. and correspondingly 14.3%, 42.9%, 28.6%, and 14.3% in our study.
With regard to treatment response using 18F-FDG-PET/CT, patient 1 showed interesting results. This patient, with an initial ISS stage of 2 and IgG-myeloma, showed a serological CR after initial anticancer therapy. However, he showed progression on the follow-up scan, with a new 18F-FDG-avid paravertebral lesion (Figure 3). Two months after the follow-up scans, IgG paraproteins started to rise again and the decision to perform a second line of immunochemotherapy was made. Literature studies have shown the value of 18FFDG-PET/CT in treatment response evaluation with regard to overall survival and progression-free survival (35, 36), leading to its inclusion in the IMWG guidelines (6).
This study has certain limitations. The small number of patients in this cohort, the inhomogeneous treatment regimens, and the varying times of follow-up scans limit the possibilities of statistical analysis of certain therapy effects. As discussed above, longer follow-up with later scans might have revealed different uptake dynamics in 18F-NaF-PET/CT. The lack of standardized criteria with which to evaluate 18F-NaF uptake is a further limitation.
Conclusion
Baseline and post-therapy 18F-FDG-PET/CT and 18F-NaF-PET/CT in this study demonstrated the excellent therapy response-evaluation properties of 18F-FDG-PET/CT that have been shown repeatedly in the literature. 18F-NaF-PET/CT as a marker of bone mineralization was shown to be significantly decreased after first-line therapy in our study population, but no significant changes were detected between the first and second follow-up scans. Larger studies are needed to study 18F-NaF-PET/CT uptake dynamics over a longer period of time to show whether permanent changes in the bone microenvironment can be reliably shown with this tracer, which may lead to changes in treatment approaches for this disease.
Footnotes
Authors' Contributions
TN analyzed and interpreted the data and wrote the article. FP collected the data and aided in the image analysis. KE review the imaging data. MW executed the statistical analysis. VP, WW and MM provided the tracer and reviewed and corrected the parts of this article regarding chemical nomenclature and tracer production. AH and MH reviewed and corrected the article for intellectual content. GK serves as the Corresponding Author and primary investigator of this study. PP and HM reviewed and corrected this article for intellectual content in regards to accuracy of statements regarding haematology and pathology.
Compliance with Ethical Standards
This study received no funding.
Conflicts of Interest
The Authors declare no conflict of interest in regard to this study.
- Received February 27, 2019.
- Revision received March 21, 2019.
- Accepted March 25, 2019.
- Copyright© 2019, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved