Abstract
PET with 18F-FDG is the standard modality in nuclear medicine for imaging multiple myeloma (MM). However, viable MM as detected by MRI or PET with other metabolic tracers, including 11C-methionine, may be missed—for example, because of low hexokinase 2 (HK2) expression of tumor cells. The aim of this study was to further investigate potential reasons for PET false negativity. Methods: A cohort of 15 mainly pretreated patients with relapsed or refractory biopsy-proven, serologically active MM who underwent both 18F-FDG and 11C-methionine PET/CT was retrospectively analyzed. Results: In 9 of the 15 patients, 18F-FDG PET was negative in the presence of viable disease. In the remaining 6 patients, both 18F-FDG and 11C-methionine PET/CT revealed the same number of MM lesions. At immunohistochemistry, 18F-FDG–negative myeloma did not exhibit significant differences in HK2 or glucose-6-phosphatase expression from 18F-FDG–positive disease (P = 0.57 and P = 0.44, respectively). Conclusion: Beyond HK2 expression, 18F-FDG negativity in (mainly pretreated) MM patients seems to be associated with additional causes not yet known.
PET with 18F-FDG as the standard nuclear medicine imaging modality is increasingly used in the diagnosis, prognostication, and management of multiple myeloma (MM) (1–7). Although sensitivity is generally high, particularly in extramedullary disease (8), subsets of viable malignant plasma cells might not be 18F-FDG–avid and therefore might be missed by standard PET imaging (4,9). Recently, so-called false-negative 18F-FDG PET results have been reported in 11% of patients with viable disease detectable on diffusion-weighted MRI and were linked to low hexokinase 2 (HK2) gene expression (10).
Potentially more sensitive approaches using radiolabeled tracers targeting metabolic pathways other than glycolysis or membrane receptors expressed by MM cells have been investigated, including 11C-/18F-choline (11,12) and 11C-acetate (13) as markers of cell membrane lipid turnover and lipid metabolism.
In addition, we and others have reported on 18F-FDG–negative, viable myeloma detectable with PET using 11C-methionine, a radiolabeled amino acid (14–19). The aim of this study was to further investigate the underlying biology of metabolically active, so-called 18F-FDG false-negative MM.
MATERIALS AND METHODS
We analyzed all patients with biopsy-proven, serologically active MM who underwent both 18F-FDG and 11C-methionine PET/CT as part of a previously published prospective study (16) approved by the local ethics committee (University of Würzburg). Of the entire cohort, patients with exclusively 11C-methionine–positive (and 18F-FDG–negative) disease were identified and compared with those who presented with identical results for both imaging examinations. All subjects gave written informed consent to sequential 18F-FDG and 11C-methionine PET/CT imaging in accordance with the Declaration of Helsinki. 11C-methionine was administered under the conditions of the pharmaceutical law (German Medicinal Products Act, AMG §13 2b) according to German law and the responsible regulatory body (Regierung von Oberfranken).
In total, 9 patients (4 female; mean age, 59 ± 10 y) with exclusively 11C-methionine–positive disease were identified and compared with 6 (control) subjects (1 female; mean age, 61 ± 8 y) in whom both PET tracers revealed an identical number of focal lesions, as well as tumor burden at identical sites. Thirteen of the 15 patients presented with relapsed or refractory, progressive disease, and the remaining two (patient 3 from the 18F-FDG–negative cohort and patient C2 from the 18F-FDG–positive cohort) presented with newly diagnosed, treatment-naïve disease. Imaging was performed before an intended change in therapy (in relapsed or refractory disease) or before treatment initiation (in the newly diagnosed cases). All patients had undergone recent (within 1 wk before PET imaging) random bone marrow biopsy of the iliac crest for histopathologic work-up. The patients’ characteristics are detailed in Table 1.
Patients’ Characteristics
PET/CT was performed after injection of 302 ± 30 MBq of 18F-FDG or 658 ± 143 MBq of 11C-methionine, and the images were qualitatively and semiquantitatively analyzed as previously described (16).
The imaging results were compared with HK2 and glucose-6-phosphatase (G6Pase) expression of the myeloma cells as assessed by standard immunohistochemistry testing on trephine biopsies. The following antibodies were used: rabbit anti-HK2 (HPA028587 [Sigma-Aldrich]; dilution, 1:50 with Dako Advanced) and rabbit anti-G6Pase (ab83690 [Abcam]; dilution, 1:50 with Dako Advanced). For staining for HK2, healthy myocardium served as a positive control. For G6Pase, hepatocytes served as a reference. For both enzymes, vascular endothelium and mesenchymal stromal cells were used as negative controls. The stained sections were analyzed semiquantitatively by light microscopy according to the immunoreactive score (IRS) of Remmele and Stegner (20). The percentage of HK2- and G6Pase-positive cells was scored as follows: 0 (no positive cells), 1 (<10% positive cells), 2 (10%–50% positive cells), 3 (>50%–80% positive cells), 4 (>80% positive cells). Additionally, the intensity of staining was graded: 0 (no color reaction), 1 (mild reaction), 2 (moderate reaction), 3 (intense reaction). Multiplication of both scores for a given sample yielded the IRS classification: 0–1 (negative), 2–3 (mild), 4–8 (moderate), 9–12 (strongly positive). Statistical analysis was performed using the Student t test (GraphPad Prism, version 5.0).
An analysis of glucose transporter 1 and L-type amino acid transporter 1 (CD98) expression as the major routes of transport of 18F-FDG and 11C-methionine into the myeloma cell had already been performed (17).
RESULTS
Because of the selection criteria, 11C-methionine PET/CT was positive in all 15 patients whereas 18F-FDG did not reveal any active lesions in 9 of the 15 patients (Fig. 1). On a lesion basis, 11C-methionine detected more than 50 focal lesions in 5 of these 9 18F-FDG PET–negative patients, 20 lesions in a single patient, and less than 20 focal lesions in the remaining 3 patients. Patient 3 presented with 11C-methionine–positive, 18F-FDG–negative extramedullary disease. In the 6 positive controls, 18F-FDG PET/CT revealed the same number of lesions as 11C-methionine PET/CT (all patients had >50 focal lesions on both scans).
Example of 18F-FDG–positive disease (A, patient C1) in comparison to 18F-FDG–negative viable myeloma (B, patient 4). Shown are maximum-intensity projection (MIP) of 18F-FDG and 11C-methionine (inset) PET/CT, as well as immunohistochemistry results for HK2 and G6Pase from iliac crest biopsies (PET/CT). Both subjects presented with serologically progressive myeloma. Despite pronounced differences in imaging, both patients display almost equally high HK2 expression (IRS, 12 vs. 12) and G6Pase expression (IRS, 6 vs. 8).
Analysis of bone marrow aspirates confirmed monoclonal plasma cells in all 15 cases and demonstrated intense expression of glucose transporter 1 (as the major route of transport of 18F-FDG into the myeloma cell) in all myeloma samples, with little variation between cases (19). In analogy to previously published findings of low HK2 expression in false-negative 18F-FDG PET studies (10), immunohistochemical analysis of HK2 was performed and also revealed intense enzyme expression by both 18F-FDG PET–negative and 18F-FDG PET–positive samples (median IRS, 9.7 ± 2.4 vs. 10.3 ± 1.9; P = 0.57) (Fig. 2; Table 2). To further elucidate the phenomenon of 18F-FDG–negative, viable MM, we focused on G6Pase, an enzyme that hydrolyzes glucose-6-phosphate to free glucose and a phosphate group. Overexpression of G6Pase has been demonstrated as a reason for 18F-FDG negativity in hepatocellular carcinoma, in which low-grade tumors tended to demonstrate higher G6Pase levels than high-grade carcinoma (21). However, irrespective of 18F-FDG PET positivity or negativity, G6Pase was highly expressed on myeloma cells in all bone marrow samples, without a significant differences between the 2 cohorts (median IRS, 8.3 ± 2.4 vs. 7.3 ± 2.4; P = 0.44) (Fig. 2; Table 2).
Box plot analysis of HK2 and G6Pase levels in 18F-FDG–negative vs. 18F-FDG–positive viable MM (unpaired Student t test, 2-tailed).
Individual Immunohistochemical Results for HK2 and G6Pase
DISCUSSION
To our knowledge, this study was the first to investigate protein HK2 and G6Pase expression as an underlying cause for 18F-FDG PET negativity in MM. In contrast to Rasche et al. (10), who recently reported on low HK2 gene expression in so-called PET false-negative MM, we did not detect lower HK2 protein levels in 18F-FDG–negative, 11C-methionine–positive patients than in either 18F-FDG– or 11C-methionine–positive controls. Additionally, no significant differences in G6Pase expression between the 2 cohorts could be identified. Thus, additional—yet-unidentified—factors in 18F-FDG negativity must be present. At the moment, although we cannot provide exact explanations for the apparent existence of different 18F-FDG–negative myeloma cell clones, our observation further underscores the distinct heterogeneity of MM (22–24). The presence of disease that is viable as detected by whole-body diffusion-weighted MRI (10,25) but not positive with 18F-FDG or non–18F-FDG radiotracers, including markers of amino acid and lipid metabolism (e.g., 11C-choline, 11C-acetate) (11,13) or of cell membrane receptor expression such as C-X-C motif chemokine receptor 4 (26,27), raises the question of distinct biologic subcohorts with distinct behavior and, potentially, distinct susceptibility to therapeutic regimens.
Noteworthy, the present study enrolled mainly patients with relapsed or refractory disease, whereas the previous study (10) included subjects with newly diagnosed MM. Thus, it is conceivable that both studies investigated patients with biologically different disease, contributing to the differences observed in both studies. Further research on the biologic and prognostic differences in different morphologic and metabolic disease patterns is required to elucidate the significance of 18F-FDG–negative disease and to aid in further understanding the complexity of myeloma biology. Given the pronounced inter- and intraindividual tumor heterogeneity, it can be assumed that a combination of different imaging approaches might be needed to comprehensively depict MM in a specific patient.
This study has some limitations. First, only a small number of patients could be enrolled, thus limiting statistical power. Second, expression of only 2 proteins was assessed, and no definite explanations for the presence of viable, 18F-FDG–negative MM can be provided. Potentially, bone marrow involvement tended to be lower in 18F-FDG–negative patients than in 18F-FDG–positive patients. However, random bone marrow biopsy is prone to sampling bias, rendering falsely low percentages for malignant plasma cell infiltration. Of note, a patchy pattern of intramedullary involvement was observed in most patients (as assessed by 11C-methionine PET/CT). In addition, serum parameters documented viable myeloma in all subjects. Third, no comparison of PET imaging results to diffusion-weighted MRI was performed. Last, further differences (beyond the inclusion of different patient cohorts) between our study and the study by Rasche et al. (10) are to be appreciated: whereas immunohistochemistry on trephine biopsies to investigate differences in HK2 and G6Pase expression was investigated by our group, Rasche et al. used gene expression profiling of CD138 purified plasma cells in a considerably larger patient cohort (n = 227).
CONCLUSION
This pilot study reports on the presence of 18F-FDG PET–negative viable myeloma with relatively high expression of HK2 (and G6Pase). Further research to elucidate the underlying mechanisms and prognostic implications of HK2-positive, 18F-FDG–negative MM is highly warranted.
DISCLOSURE
No potential conflict of interest relevant to this article was reported.
Footnotes
Published online Nov. 2, 2018.
- © 2019 by the Society of Nuclear Medicine and Molecular Imaging.
REFERENCES
- Received for publication July 11, 2018.
- Accepted for publication October 10, 2018.
Jump to section
Related Articles
Cited By...
- No citing articles found.