Choline PET/CT in Multiple Myeloma

The field of multiple myeloma (MM) imaging has evolved. The International Myeloma Working Group recently recommended performing 18F-fluorodeoxyglucose glucose (18FDG) positron emission tomography/computed tomography (PET/CT) with the aim of staging MM patients at baseline and evaluating response to therapy. Novel oncological radiotracers such as 11C-Choline and 18F-Fluorocholine, have been studied in comparison with 18FDG, mostly in MM patients presenting with refractory disease or suspected relapse. Choline-based tracers may overcome some limitations of 18FDG, which include a lack of sensitivity in depicting skull lesions and the fact that 10% of MM patients are FDG-negative. The majority of MM lesions display a higher uptake of Choline than FDG. Also, in many situations, Choline may offer better lesion visualization, with a higher tumor to background ratio; however, various patterns of Choline and FDG uptake have been observed in MM and some limitations, notably as regards liver lesions, should be recognized. Overall, Choline may provide additional detection of up to 75% more lesions. This article aims to provide a comprehensive review of the potential role of Choline in multiple myeloma, as compared to FDG, encompassing Choline physiopathology as well as data from clinical studies.

hematological and solid cancers. It reflects the level of glycolytic activity of cancer cells, which is increased in about 90% of newly diagnosed MM [19]. 18 FDG PET/CT offers a thorough evaluation of MM tumor burden at diagnosis and carries a powerful prognostic value at baseline and during therapy response assessment [20][21][22][23][24].
Bellow, we discuss the findings relating to the detection of focal bone lesions and extra-medullary disease by 18 FDG and Choline PET/CT.

Focal Bone Lesion Detection
The comparison of Choline PET/CT and 18 FDG PET/CT lesion detectability in MM has been performed in the past few years (Table 1). In a study conducted by Nanni and colleagues [25], 10 patients underwent both 11 C-Choline and 18 FDG PET/CT scans within one week. The settings were: evaluation after completion of initial therapy (n = 4), disease relapse (n = 4) and follow-up (n = 2). Six patients were positive on both examinations, with a total of 37 bone lesions detected on 11 C-Choline PET/CT vs. 22 lesions on 18 FDG PET/CT. Four patients had concordant negative findings. The last patient had only one lesion in the pelvis, which was 18 FDG-positive but 11 C-Choline negative. None of the patients had a positive 11 C-Choline PET combined with a negative 18 FDG PET. Choline intensity of uptake was globally superior to that of FDG, with a mean SUVmax of 5.0 vs. 3.8 respectively (p = 0.042); however, it is interesting to note that some lesions had a high 18 FDG uptake but a low Choline uptake. Moreover, a mix of uptake patterns of MM lesions could be observed within the same patient. In a second study by Cassou-Mounat et al. [26], 21 MM patients underwent FCH-PET/CT and FDG PET/CT within a median time of 7 days, for progression under treatment (n = 6) or suspected relapse (n = 15). Nineteen were biochemically confirmed. One patient was detected as positive for bone involvement by FCH-PET/CT only, while no patient was positive by FDG PET/CT only. After masked reading, the mean number of foci per patient was 4.6 for FDG vs. 8.1 for FCH (p < 0.001) with a total number of 69 lesions for FDG and 121 for FCH. Among the 69 FDG-positive lesions, 65 were also FCH-positive. The 56 foci that were FCH+/FDG-were predominantly located in the skull and the torso. The median SUVmax values for the FCH-detected lesions (3.8 and 5.7) were higher than for the FDG-detected lesions (3.0 and 4.5). Additionally, the tumor/background ratio of Choline was superior to that of FDG.
Overall, in relapsed/refractory MM, Choline PET/CT detects up to 75% more bone lesions than 18 FDG PET/CT [25,26]. Choline also offers better visualization of MM lesions than FDG, which can contribute to a higher reading comfort but also higher inter-observer agreement. The low sensitivity of FDG for skull lesions is explained by the high glucose uptake in adjacent brain tissues ( Figure 1). This limitation is overcome by Choline tracers (Figure 2). However, the high liver uptake of Choline may limit its sensitivity to depict lesions of surrounding bone areas such as the right costal grid.

5
In a second study by Cassou-Mounat et al. [26], 21 MM patients underwent FCH-PET/CT and FDG PET/CT within a median time of 7 days, for progression under treatment (n = 6) or suspected relapse (n = 15). Nineteen were biochemically confirmed. One patient was detected as positive for bone involvement by FCH-PET/CT only, while no patient was positive by FDG PET/CT only. After masked reading, the mean number of foci per patient was 4.6 for FDG vs. 8.1 for FCH (p < 0.001) with a total number of 69 lesions for FDG and 121 for FCH. Among the 69 FDG-positive lesions, 65 were also FCH-positive. The 56 foci that were FCH+/FDG-were predominantly located in the skull and the torso. The median SUVmax values for the FCH-detected lesions (3.8 and 5.7) were higher than for the FDG-detected lesions (3.0 and 4.5). Additionally, the tumor/background ratio of Choline was superior to that of FDG.
Overall, in relapsed/refractory MM, Choline PET/CT detects up to 75% more bone lesions than 18 FDG PET/CT [25,26]. Choline also offers better visualization of MM lesions than FDG, which can contribute to a higher reading comfort but also higher inter-observer agreement. The low sensitivity of FDG for skull lesions is explained by the high glucose uptake in adjacent brain tissues ( Figure 1). This limitation is overcome by Choline tracers (Figure 2). However, the high liver uptake of Choline may limit its sensitivity to depict lesions of surrounding bone areas such as the right costal grid.

Uptake Patterns of 18FDG and Choline-Based Tracers
A difference in uptake pattern between 18 FDG and Choline can be observed among patients but also within the same patient ( Figure 3) [25,26]. FDGhigh/Cholinelow uptake of MM lesions are relatively rare. One needs first to exclude false-positive findings of FDG, with for example recent fractures that can bring substantial inflammation leading to increased FDG uptake [26]. MM patients are prone to bone fractures, especially of the ribs and spine, as these patients are at higher risk of having osteoporosis [28]. FDGhigh/Cholinelow uptake can be seen in aggressive or "dedifferentiated" lesions [26]. This uptake pattern has already been documented in hepatocellular carcinoma and is a signature of aggressive disease [29]. The more frequently observed pattern of Cholinehigh/FDGlow uptake of MM lesions may have different causes. It may correspond to early lesions, as only a minority of these foci were accompanied by osteolytic changes on the corresponding CT images (35%) [26] (Figure 4). It Figure 2. A 60-year-old female with IgG lambda smoldering multiple myeloma. An osteolytic lesion of the skull was found on a follow-up CT. 18 F-Choline PET/CT was ordered to further characterize this lesion and search for additional bone lesions. 18 F-Choline PET/CT axial image of the skull shows a moderate uptake of an occipital osteolytic lesion (A). Additional focal uptake of 18 F-Choline was seen in the left femur, corresponding to a bone marrow lesion with no bone structural changes on CT (B).

Uptake Patterns of 18FDG and Choline-Based Tracers
A difference in uptake pattern between 18 FDG and Choline can be observed among patients but also within the same patient ( Figure 3) [25,26]. FDG high /Choline low uptake of MM lesions are relatively rare. One needs first to exclude false-positive findings of FDG, with for example recent fractures that can bring substantial inflammation leading to increased FDG uptake [26]. MM patients are prone to bone fractures, especially of the ribs and spine, as these patients are at higher risk of having osteoporosis [28]. FDG high /Choline low uptake can be seen in aggressive or "dedifferentiated" lesions [26]. This uptake pattern has already been documented in hepatocellular carcinoma and is a signature of aggressive disease [29]. The more frequently observed pattern of Choline high /FDG low uptake of MM lesions may have different causes. It may correspond to early lesions, as only a minority of these foci were accompanied by osteolytic changes on the corresponding CT images (35%) [26] ( Figure 4). It may also reflect a more indolent form of the disease [26]. It was recently pointed out that about 10% of newly diagnosed MM patients, with bone lesions on MRI, are FDG-negative, which is linked to a lower expression of hexokinase-2 [19]. Another study showed that newly diagnosed MM patients with false-negative findings on FDG PET/CT had longer PFS than FDG-positive patients [30]; however, other mechanisms also appear to be involved in the uptake as FDG-negative patients with relapsed/refractory disease were found to have normal expression of Hexokinase-2 [31]. Finally, the use of dexamethasone as a treatment option can decrease FDG-uptake by promoting MM cell apoptosis and by lowering the inflammatory environment of MM cells. Overall, these results corroborate the recent findings of multiple, spatially separated clones that coexist in the same patient, carrying different genotypic mutations [32]. These clones can be individualized even in patients with negative minimal residual disease. This highlights the complementary role of whole-body imaging, and the possible advantage of using more than one tracer in some situations that remain to be defined. Whether these different patterns are linked to patient prognosis should also be investigated in prospective studies.
Cancers 2020, 12, x FOR PEER REVIEW 7 of 13 7 may also reflect a more indolent form of the disease [26]. It was recently pointed out that about 10% of newly diagnosed MM patients, with bone lesions on MRI, are FDG-negative, which is linked to a lower expression of hexokinase-2 [19]. Another study showed that newly diagnosed MM patients with false-negative findings on FDG PET/CT had longer PFS than FDG-positive patients [30]; however, other mechanisms also appear to be involved in the uptake as FDG-negative patients with relapsed/refractory disease were found to have normal expression of Hexokinase-2 [31]. Finally, the use of dexamethasone as a treatment option can decrease FDG-uptake by promoting MM cell apoptosis and by lowering the inflammatory environment of MM cells. Overall, these results corroborate the recent findings of multiple, spatially separated clones that coexist in the same patient, carrying different genotypic mutations [32]. These clones can be individualized even in patients with negative minimal residual disease. This highlights the complementary role of whole-body imaging, and the possible advantage of using more than one tracer in some situations that remain to be defined. Whether these different patterns are linked to patient prognosis should also be investigated in prospective studies.   Although a good correlation can be observed between Choline uptake and plasma cell infiltration on BMB (R 2 = 0.52) [27], the performance of Choline for the diagnosis of diffuse infiltration pattern has not yet been addressed. Studies prospectively comparing MRI and Choline PET/CT in that regard are warranted.

MYELOCHOL: a Prospective Pilot Study
The MYELOCHOL study is a prospective monocentric study that is aimed at comparing FCH and 18 FDG PET/CT performance in newly diagnosed MM. Both of the previous studies performed have a common limitation [25,26]. Lesions were assessed without a pre-determined reference standard able to differentiate between true-positive and false-positive findings. MRI is the gold standard for bone marrow imaging [33], which makes it a good candidate for Choline PET/CT performance evaluation, as obtaining biopsy results for each lesion would not be feasible. Additionally, the performance Choline PET/CT is yet to be evaluated in a patient population that has not been exposed to MM drugs. The treatment of MM may indeed select clones that could behave differently from baseline [32].
The MYELOCHOL study plans to include 30 newly diagnosed MM patients (Haut-Lévêque Hospital, University of Bordeaux, Bordeaux, France). All patients will undergo Whole-Body MRI, FCH, and 18FDG PET/CT with a maximal interval of 7 days between each imaging modality. Whole-Body MRI will be the standard of reference. Each imaging study will be read blindly by two nuclear medicine physicians with expertise in MM. A second step will include a comparative reading of each PET lesion with MRI findings, to determine if the lesion is a true positive or a false positive. In case of a Choline-positive lesion not found on MRI, a multidisciplinary team, composed of nuclear medicine, radiology, and hematology physicians, will come to a consensus as to whether the lesion can be related to MM. Study enrollment started in September 2019 and is currently ongoing.

Extra-Medullary Disease
Multiple myeloma patients with extra-medullary disease (EMD) have a worse prognosis [34]. Extra-medullary lesions develop in any soft tissue and should be differentiated from soft tissue lesions expanding from adjacent bones ("breakout lesion") [35]. In line with what was found for bone lesions, Choline uptake within EMD can either be higher or lower than FDG-uptake [25,26]. The detectability of liver EMD lesions may be hampered due to the high physiological choline uptake by the liver. Moreover, benign hepatic lesions such as nodular hyperplasia may mimic MM EMD [6]. Generally, caution must be taken when diagnosing EMD with Choline or FDG, as inflammatory or neoplastic processes are known confounders for both tracers [10,36]. Although a good correlation can be observed between Choline uptake and plasma cell infiltration on BMB (R 2 = 0.52) [27], the performance of Choline for the diagnosis of diffuse infiltration pattern has not yet been addressed. Studies prospectively comparing MRI and Choline PET/CT in that regard are warranted.

MYELOCHOL: A Prospective Pilot Study
The MYELOCHOL study is a prospective monocentric study that is aimed at comparing FCH and 18 FDG PET/CT performance in newly diagnosed MM. Both of the previous studies performed have a common limitation [25,26]. Lesions were assessed without a pre-determined reference standard able to differentiate between true-positive and false-positive findings. MRI is the gold standard for bone marrow imaging [33], which makes it a good candidate for Choline PET/CT performance evaluation, as obtaining biopsy results for each lesion would not be feasible. Additionally, the performance Choline PET/CT is yet to be evaluated in a patient population that has not been exposed to MM drugs. The treatment of MM may indeed select clones that could behave differently from baseline [32].
The MYELOCHOL study plans to include 30 newly diagnosed MM patients (Haut-Lévêque Hospital, University of Bordeaux, Bordeaux, France). All patients will undergo Whole-Body MRI, FCH, and 18FDG PET/CT with a maximal interval of 7 days between each imaging modality. Whole-Body MRI will be the standard of reference. Each imaging study will be read blindly by two nuclear medicine physicians with expertise in MM. A second step will include a comparative reading of each PET lesion with MRI findings, to determine if the lesion is a true positive or a false positive. In case of a Choline-positive lesion not found on MRI, a multidisciplinary team, composed of nuclear medicine, radiology, and hematology physicians, will come to a consensus as to whether the lesion can be related to MM. Study enrollment started in September 2019 and is currently ongoing.

Extra-Medullary Disease
Multiple myeloma patients with extra-medullary disease (EMD) have a worse prognosis [34]. Extra-medullary lesions develop in any soft tissue and should be differentiated from soft tissue lesions expanding from adjacent bones ("breakout lesion") [35]. In line with what was found for bone lesions, Choline uptake within EMD can either be higher or lower than FDG-uptake [25,26]. The detectability of liver EMD lesions may be hampered due to the high physiological choline uptake by the liver. Moreover, benign hepatic lesions such as nodular hyperplasia may mimic MM EMD [6]. Generally, caution must be taken when diagnosing EMD with Choline or FDG, as inflammatory or neoplastic processes are known confounders for both tracers [10,36].

Therapy Response Assessment
The value of Choline PET/CT for response assessment in MM has not yet been explored. The IMPETUS criteria have been defined for harmonizing 18 FDG PET/CT readings with the use of the Deauville five-point scale [37]. This scale implies the use of liver uptake as a reference above which a lesion could be considered positive after treatment; however, Choline-specific criteria will have to be elaborated for response evaluation, mainly because of the markedly intense liver uptake of this tracer.
The main features of Choline PET/CT and FDG PET/CT are summarized in Table 2.

11. C-Choline PET/CT vs. 11 C-Methionine PET/CT and Other PET Tracers
Lapa and colleagues investigated the detectability rate of 11 C-Choline compared to 11 C-Methionine [27]. Nineteen pre-treated patients received both PET/CT scans within a median time of 10 days. No reference standard was used. Fifteen patients had positive findings, with only one patient with a methionine positive PET/CT but negative choline PET/CT. Choline and Methionine detected an equal number of lesions in 11/19 patients while Methionine detected more lesions in 8/19 patients. Seven patients had disseminated disease (>50 FL) that was identified by both tracers in four patients. In the remaining three patients with disseminated disease on Methionine PET/CT, Choline PET/CT only identified a total number of four lesions. The uptake of Methionine was superior to that of Choline with a median SUVmax of 9.3 (3.5-39) vs. 5.7 (3.5-14.6), respectively. A better tumor/background ratio was also observed with 11 C-Methionine (18.7 vs. 8.8, p = 0.0001). Although 11 C-Methionine appears to yield better results than 11 C-Choline, its availability is currently limited, considering the difficulties related to the short half-life of carbon-11.
Besides FDG, Choline, or Methionine, other radiotracers are of potential interest, with especially radiopharmaceuticals that can be used for disease imaging and treatment as well, constituting important theranostic applications [38], such as radiolabeled agents targeting the prostate-specific membrane antigen (PSMA), the chemokine receptor CXCR4 [39] or radiolabeled CD38-targeting antibodies [40,41]. A cluster of differentiation CD38 is highly expressed on myeloma cells. Daratumumab is the first fully human monoclonal antibody targeting CD38 for the treatment of MM. It improves the depth and duration of response in combination with other antimyeloma agents [42,43]. Daratumumab has been successfully labeled with 89 Zr and 64 Cu with promising immuno-PET imaging results in pre-clinical studies but also recently in small in-human studies [40,41]. CD38-targeted radioimmunotherapy has also been investigated preclinically [44,45]. The rate of patients with high CD38 expression will determine how many of them could beneficiate from anti-CD38 immuno-PET imaging and possibly CD38-targeted-immunoradiotherapy.

Conclusions
In multiple myeloma, PET/CT performed with Choline-based tracers may offer several advantages over 18 FDG PET/CT, showing better detectability of focal bone lesions; however, the diagnostic performance of Choline PET/CT should be further explored in larger prospective studies, which should be designed with a reference standard. These data are required before clinical routine implementation can take place. Studies focusing on the prognostic value of Choline PET/CT and its value as regards treatment response evaluation are also warranted.