Multiple Myeloma (MM) is a plasma-cell malignancy in which a single clone of plasma cells proliferates and produces a monoclonal protein (MC). Myeloma clonal plasma cells accumulate in the bone marrow (BM), resulting in diffuse skeletal involvement, hypercalcemia, anemia and extramedullary localizations ^[1][2][3][4][1,2,3,4]. In addition, MC and free-light chains are nephrotoxic and may result in renal failure. Many advances have been made in MM treatment even in aggressive forms of disease ^{[5][6][7][8][9]}[5,6,7,8,9]. Early diagnosis of organ damage has recently been improved. Symptomatic MM is described by the typical CRAB features (hypercalcemia, renal failure, anemia, bone disease) where bone disease is now referred to as >1 osteolytic lesion (≥5 mm) on skeletal radiography, computed tomography (CT) or positron emission tomography with computed tomography (PET/CT) ^[10][11][10,11] as one of the myeloma defining events (MDE) defined by the International Myeloma Working Group (IMWG) which can identify active disease and the need to treat patients [12]. In addition, creatinine 2 mg/dL could be a MDE although renal involvement can be seen at the MGRS stage (monoclonal gammopathy of renal significance (MGRS)) [13]. Bone disease is indeed a mainstay of active MM; hence, it is identified in 70% of patients and invariably indicates a need to start therapy [14]. The established imaging modalities to manage MM patients are CT, Magnetic Resonance Imaging (MRI) and PET/CT, whereas there is consensual agreement on their replacement of the old X-ray skeletal survey because of its significantly lower sensitivity for bone lesions’ detection. Recently, MRI and PET/CT have acquired a major role with respect to low dose CT. These imaging modalities can functionally quantify bone disease before bone damage, and lessons can be obtained from the treatment of lymphomas [15]. MM is characterized by the focal and patchy distribution of plasma cells in the BM, the focal lesions (FLs) identified with MRI and PET/CT are focal accumulations of plasma cells and they are different from the lytic lesions detected with low dose CT, where bone destruction has already occurred [16]. FLs are defined as focal bone uptake on two consecutive PET slices without evident changes on CT for PET/CT and a low T1-weighted signal and a high T2w-STIR signal on MRI [17].

2. PET/CT: Methods and Role in Response to Therapy in MM

¹⁸F-Fluorodeoxyglucose (FDG)-PET/CT is considered the best functional imaging technique in defining the metabolic activity of bone lesions, plasmacytomas and extra-medullary disease (EMD) caused by MM. ¹⁸F-FDG is currently the most used tracer for MM and it is a radiolabeled glucose analogue in which the C-2 hydroxyl group is replaced by the positron-emitting fluorine-18 atom (¹⁸F) ^[18][19]. FDG is taken up by high-glucose-using cells through the glucose transporters GLUT1 and GLUT3 and it is phosphorylated by a hexokinase to ¹⁸F-FDG-6P, that is stored in cells as it cannot be further metabolized. After intravenous administration, ¹⁸F-FDG reaches an equilibrium state in 60 min and it starts to decay in 80–90 min, allowing an accurate assessment of glucose metabolic activity. MM plasma cells usually overexpress hexokinase-2, displaying a higher glycolytic activity than surrounding cells [17]. PET positivity is defined as the presence of areas of ¹⁸F-FDG uptake higher than the liver uptake, that is taken as a background with a Deauville score of 4. Consequently, PET positive lesions have a Deauville Score of 4 or 5 ^[19][20]. For a better localization of metabolically active FLs, PET is usually combined with low dose CT (120 kV, 80 mA). Mostly the field of view (FOV) includes the region spanning from the skull to the femur, encompassing the upper limbs, and there are only a few centers using whole body (WB) PET/CT ^[20][21]. ¹⁸F-FDG-PET/CT is used to clarify a dubious diagnosis between MM and smoldering multiple myeloma (SMM), when low-dose WB-CT and WB-MRI are inconclusive but there is a strong suspicion of active MM, because of the high cost of this imaging technique. ¹⁸F-FDG-PET/CT is also used for the diagnosis of suspected extramedullary solitary plasmacytoma [6] and in the prognostic assessment of newly diagnosed MM (NDMM): more than three FLs ^[21][22], EMD and a maximum standardized uptake value (SUVmax) greater than 4.2 are related to a poor outcome ^[20][21]. These three factors are independently associated with progression-free survival (PFS). High FDG-avidity (SUV > 4.2) and the presence of EMD are also correlated with shorter overall survival (OS) ^[22][23]. The main implication of ¹⁸F-FDG-PET/CT is in the assessment of treatment response. After the first line therapy, patients undergo a repeat bone marrow (BM) aspiration, for the evaluation of MRD, and ¹⁸F-FDG-PET/CT is the only imaging technique able to distinguish between active and inactive FLs, with high prognostic value: ¹⁸F-FDG uptake in bone lesions has higher sensitivity in detecting tumor residual disease than immunofixation electrophoresis ^[23][24] and appears to be closely related with PFS and OS. In the standardization of the ¹⁸F-FDG-PET/CT Deauville criteria for MM, Zamagni et al. defined the complete metabolic response as the ¹⁸F-FDG uptake minus the liver activity in BM sites and FLs previously involved (including extramedullary and para-medullary disease (DS score 1–3)) and partial metabolic response as the decrease in number and/or activity of BM/FLs present at baseline, but persistence of lesion(s) with an uptake greater than the liver activity (DS score 4 or 5). The metabolic disease is defined as stable if there are no significant changes in BM/FLs and progressive when new FLs appear compared with baseline ^[24][25].

2.1. New Tracers

The sensitivity of ¹⁸F-FDG-PET/CT in bone lesions’ detection at diagnosis ranges from 59% to 100%, whereas the specificity ranges from 75% to 82% ^[20][21]. False positive can be obtained in the case of other malignancies, inflammation, infections, fractures, post-surgical areas (including BM biopsy), recent chemotherapy infusion and use of growth factors, therefore PET/CT should be performed at least one month after the use of these agents ^[24][25]. False negative can be obtained in the case of administration of high-dose steroids, hyperglycemia ^[20][21] and in cases of a lack of hexokinase that can occur in 10–15% of MM patients. To overcome this limitation, new tracers have been evaluated for patients with MM, specific for plasma cells or proliferating cells or conjugated with anti-CD38 Daratumumab (immune PET tracers). Lipid tracers, such as choline and acetate, were first investigated. Choline is an indicator of the synthesis of plasma membrane, and it can be labelled with ¹¹C and ¹⁸F, becoming more sensitive than ¹⁸F-FDG ^[25][26][26,27]. Indeed, acetate is converted in acetyl-CoA in Krebs’ cycle, producing energy for cells and it is also considered more sensitive than ¹⁸F-FDG ^[27][28][28,29]. Amino acid tracers are used by plasma cells in the synthesis of new proteins, such as immunoglobulins, and they can be more specific biomarkers of active MM. Among them, ¹¹C-methionine is more sensitive than ¹⁸F-FDG and ¹¹C-choline ^[29][30], even if it has a short half-life (20 min) and an on-site cyclotron is necessary to produce it ^[30][31]. Stokke et al. demonstrated that ¹⁸F-fluciclovine, a leucine analogous, is also a promising tracer for MM, reaching a similar uptake pattern and similar sensitivity, with a half-life of 110 min ^[31][32]. ¹⁸F-fluoro-ethyl-tyrosine could be used as a tracer, but it is less sensitive than ¹¹C-metionine and ¹⁸F-FDG, having minimal transportation into plasma cells ^[32][33]. Nucleoside tracers are related to the rate of DNA synthesis, reflecting highly proliferating cells. ¹¹C-thiothymidine sensitivity is similar to the ¹¹C-methionine one and it is better than ¹⁸F-FDG sensitivity, especially during the early stage of the disease ^[33][34]. ¹⁸F-fluorothymidine was investigated as a new tracer, but it is not considered a useful biomarker for MM ^[34][35]. ¹⁸F-fluoride reflects the early phase of bone calcification, but results concerning the diagnostic assessment in MM are divergent ^[35][36]. ⁸⁹Zr-bevacizumab, a radiolabeled antibody directed to the VEGF receptor, is uniformly expressed on plasma cells and it could be useful for the detection of MM ^[36][37]. In the end, there are also tracers targeting molecules expressed on cells surface such as CD38, expressed on plasma cells, and CXCR4, a chemokine expressed on hematopoietic stem cells. The activation of CXCR4/stromal-derived factor 1 axis correlates with bone activation, playing an important role in MM. Pentixafor radiolabeled with gallium-68 is used to target CXCR4. Its sensitivity compared to ¹⁸F-FDG is not clear, but a positive ⁶⁸Ga-pentixafor has a negative prognostic significance, with a poorer OS ^[37][38]. CD38 is highly expressed on all MM cells and anti-CD38 monoclonal antibody (Daratumumab) can be labelled with Zirconium-89 or Copper-64 to target them, but data are not yet available from recent clinical trials.

2.2. PET/CT and MRD in MM

Due to the patchy dissemination of the disease, MRD evaluation in the BM could sometimes lead to false negative results in the presence of minimal disease after therapy ^[38][39]. In the absence of recognized new tools for blood MRD evaluation, PET is a valid technique able to identify focal active metabolic lesions that can be reservoirs for MM relapse. MRD negative is a deeper level of response than a complete response, that requires the absence of phenotypically abnormal clonal plasma cells from BM aspirates, detected by next-generation flow cytometry (NGF) or next-generation sequencing (NGS). However, MM is a heterogenous disease with patchy infiltration and EMD is not uncommon. The association between BM analysis and ¹⁸F-FDG-PET/CT leads to a more accurate assessment of response after a treatment line with the highest prognostic value, well related to PFS and OS. In fact, although MRD-negativity is associated with improved outcomes, in MRD-negative patients relapse still occurs, potentially due to the presence of focal bone disease that could be detected with ¹⁸F-FDG-PET/CT. The IFM 2009 trial showed patients who were double negative (with no residual disease assessed by NGF and negative PET/CT) achieving better PFS than patients who were not double negative ^[37][38]. Alonso et al. selected 103 NDMM patients who received their first-line therapy, underwent BM MRD assessment with NGF and ¹⁸F-FDG-PET/CT evaluation at diagnosis and one month after the end of the treatment. It was observed that patients MRD-/PET- had the best 4-years OS (94.2%) and PFS (92 months), patients MRD+/PET− had a not significant difference in 4-years OS (100%) but a shorter PFS (45 months) and PET+ patients had the worst 4-years OS (73.8%) and PFS (28 months) ^[39][40].

2.3. Clinical Studies

Numerous studies evaluated the accuracy of ¹⁸F-FDG-PET/CT in the therapy response assessment, even if they are not exactly concordant at the time of evaluation and in the interpretation of results. The main disadvantages of the ¹⁸F-FDG-PET/CT, beyond the high cost and the limited availability, are the lack of standardization criteria and of interobserver reproducibility in the interpretation of results. More than 30 clinical studies on the prognostic role of PET/CT in the evaluation of treatment response were found.

3. MRI: Methods and Role in Response to Therapy in MM

3.1. MRI and MRD in MM

NGF with functional imaging has been used to define responses in MM ^[40][66], though the role of MRI to complement MRD needs confirmation. The use of MRI for the assessment of MRD is not clear because there are no data for a comparison with FDG-PET/CT in evaluating MRD after therapy. Moving forward, it is crucial to standardize guidelines about the use of imaging techniques and the time point to be assessed. It is necessary to establish the role of MRI to the definition of MRD with new comparative studies.

3.2. Clinical Studies

Although IMWG guidelines indicate the use of DWI-MRI to evaluate plasma cell disorders, the use of MRI to assess response to treatment is a matter of debate and it is investigated in several clinical studies, in small series of patients. In recent years, some studies have researched the role of DWI in treatment response assessment in patients with MM. They have proven that DWI is more sensitive than the conventional MRI and it gives a quantitative and non-invasive evaluation of BM after therapy, distinguishing responders from non-responders. Some of these studies also hypothesize that WB-DWI could be equivalent or superior to PET/CT assessment of the response and the MRD. To shed light on the role of WB-MRI, a multidisciplinary, international, and expert panel of radiologists, medical physicists and hematologists developed the Myeloma Response Assessment and Diagnosis System (MY-RADS) imaging recommendations ^[41][67].

4. Comparison of PET/CT and Functional MRI in Response Evaluation in MM

Evidence shows that both MRI and PET/CT have significant potential in evaluating treatment response as separate tools. Specifically, data from PET/CT studies point out its prognostic and predictive value whereas MRI emerges as particularly sensitive technique to detect myeloma FLs and background marrow infiltration, with the addition of DWI bringing further improvement in sensitivity and potential for differentiation between active and treated disease in cases of disease relapse. Several studies have also investigated which modality performs better in the setting of treatment-response evaluation. A recent comparative meta-analysis of Yokoyama et al. has shown the major impact of PET/CT rather than MRI in response to therapy (sensitivity 80% vs. 20%, specificity 58% vs. 83%) ^[42][80]. Other studies reach the same conclusions even with different data (same sensitivity 75%, specificity 86% vs. 43%) ^[43][81]. The prognostic value of PET/CT and WB-MRI has also been investigated for response evaluation at different time points of treatment, finding only PET/CT significantly able to predict response to autologous stem cell transplantation (ASCT) in the post-induction phase ^[31][33][44][32,34,71] and OS and PFS in the post-transplant evaluation ^[45][82]. The glucose metabolism exploited by FDG-PET/CT rapidly follows the dynamic change of FLs during treatment, whereas the persistence of non-viable lesions on MRI images may explain its reduced specificity and the lack of prognostic value ^[46][47][83,84]. Sometimes active lesions may be misinterpreted as scar tissue, affecting sensitivity ^[42][80]. Implementation with DWI technique increases the specificity of MRI sections ^[43][81], though it seems to not gain advantage over PET/CT in predictive value. On the other hand, another recent meta-analysis of data from 12 comparative studies assessing the accuracy of WB-MRI and FDG-PET/CT, identify the first as the most sensitive technique (90% vs. 66%) for determining response through earlier detection of post-treatment recurrence, though the finding was not significant. Similar evidence is reported in a retrospective study comparing PET/CT and MRI in different treatment phases ^[46][83]. These results could also be biased from the delayed healing of pre-treatment myeloma lesions which can still be detected from MRI. The bias has been addressed by the meta-analysis investigator Rama et al. by considering five studies comparing PET/CT with DWI-WB-MRI. The use of DWI though displayed the same specificity of WB-MRI (57% vs. 56%), not overcoming the limitation.

Comparison of PET/CT and Functional MRI for MRD in MM

Concerning comparison of PET/CT and functional MRI for MRD assessment, clinical studies are few. According to IMWG, PET/CT represents the best tool for MRD evaluation as it has been shown to predict survival after therapy. However, most clinical trials using PET/CT as MRD tool are still ongoing and preliminary data do not show significant concordance between imaging and BM assessment ^[48][85]. A recent study from Rashe et al. evaluating the combination of functional imaging and flow cytometry for MRD evaluation on a cohort of patients undergoing first or successive lines of therapy, proposed DWI-MRI as complementary imaging tool to PET/CT, as the PFS of patients with FLs detected with DWI-MRI or PET/CT were not statistically significant (3.4 years vs. 3 years). Furthermore, the study identified that the combination of both approaches yielded the highest rate of FLs detection in patients achieving CR ^[49][86].