A Systematic Review of MRI, Scintigraphy, FDG-PET and PET/CT for Diagnosis of Multiple Myeloma Related Bone Disease - Which is Best?

Aim: The purpose of the current study was to conduct a systematic review of the published literature to evaluate the diagnostic accuracy of FDG-PET, PTE/CT, MRI and scintigraphy for multiple myeloma related bone disease. Methods: Through a search of PubMed, EMBASE, and the Cochrane Library, two reviewers independently assessed the methodological quality of each study. We estimated pooled sensitivity, specificity, positive and negative likelihood ratios (PLR and NLR), and two sample Z-tests were conducted to evaluate for differences in sensitivity, specificity, area under the curve (AUC), and the Q* index between any two diagnostic modalities. Results: A total of 17 studies were reviewed. The MRI had a pooled sensitivity of 0.88, specificity of 0.68, AUC of 0.897, and Q*index of 0.828, whereas for MIBI, the corresponding values were 0.98, 0.90, 0.991, and 0.962, respectively, and for bone scan, they were 066, 0.83, 0.805, and 0.740, respectively. The corresponding values of MIBI were 0.98, 0.90, 0.991, and 0.962, respectively. For PET and PET/CT, the values were 0.91, 0.69, 0.927 and 0.861, respectively. Statistically significant differences were not found in the sensitivity, specificity, AUC, and Q* index between MRI, scintigraphy, FDG-PET and PET/CT. Conclusions: On the condition that X ray is taken as a reference in our study, we suggested that FDG-PET, PTE/CT, MRI and scintigraphy are all associated with high detection rate of bone disease in patients with MM. Thus, in clinical practice, it is recommended that we could choose these tests according to the condition of the patient.

Introduction articles with the same keywords mentioned above in Chinese.

Data Selection
Two investigators independently checked all of the retrieved articles to determine whether they satisfied the following selection criteria: (a) CT, MRI, FDG-PET, FDG-PET/CT and scintigraphy was used to identify and characterize suspected multiple myeloma patient without treated. If the study includ patients before and after treatment, should have valid results of pro-treatment patients; (b) Studies assessed the diagnostic accuracy of bone disease can use WBXR or/and CT as a reference standard; (c) Sufficient data were presented to calculate the true-positive (TP), false-positive (FP), falsenegative (FN), and true-negative (TN) values for the imaging techniques; (d) Patients with a secondary malignoma were excluded from the study. (e) Concerning to the quality of study design, only the article in which the number of the answer 'yes' for the 14 questions in QUADAS quality assessment tool was larger than 8 was included. if the number of the answer 'No' or 'unclear' was larger than 5, the article was excluded. Theoretically, the reference standard would be a local biopsy to confirm clonal plasmacytosis. The Durie and Salmon staging system introduced in 1975 relied on the radiographic skeletal survey as the sole imaging criterion. Given that it is both practically and ethically impossible to perform biopsies of all detected lesions, the use of this method as a reference standard in clinical studies was not achievable. The International Myeloma Working Group (IMWG) reported a consensus statement on the role of imaging techniques in multiple myeloma1 in which whole body X-ray (WBXR) was considered the gold standard in initial staging, and if available WBXR might be replaced by CT (Barlogie et al., 2004). So we used X-rayor/and CT and clinical follow-up as alternative reference standards.

Data extraction
Data were obtained for author, year of publication, country, patient characteristics, reference standard and diagnostic performance of imaging modalities. Data were extracted independently by 2 investigators. To resolve disagreement between reviewers, a third reviewer assessed all discrepant items, and the majority opinion was used for analysis.

Study design characteristics
The QUADAS quality assessment tool was used to extract relevant study design characteristics of each study. This tool and the definitions of the characteristics are fully described by Penny Whiting. It is the first systematically developed evidence based quality assessment tool to be used in systematic reviews of diagnostic accuracy studies. Two investigators independently assessed whether each item of QUADAS was fulfilled (yes, no or unclear).

Data analysis
Data were separately analyzed for MRI, scintigraphy, PET and PET/CT. We calculated pooled sensitivity, specificity and diagnostic odds ratio (DOR) for each modality Data were separately analyzed for PET, CT, and MRI, we calculated pooled sensitivity, specificity and diagnostic odds ratio (DOR) for each modality, and we also calculated summary receiver operating characteristic curves (SROC) and the Q* index.

Study selection
After the computerized search was performed and reference lists were extensively cross-checked, we extracted 2150 abstracts for analysis; however, 1919 articles were excluded on the basis of their titles, abstracts or full text. We screened the full text of 134 articles, 23 studies were excluded because the aim of the articles was not to reveal the diagnostic value of MRI, FDG-PET, PET/ CT or scintigraphy for identification and characterization of multiple myeloma related bone disease; 36 studies were excluded because the study compared patients before and after treatment, but the valid results for protreatment patients could not be extracted; 31 studies were excluded owing to insufficient information construct or calculate true-positive, false-positive, true-negative, and/ or false-negative results. 24 studies were excluded for the identification of bone disease could not using WBXR or CT as a reference standard.Finally, 17 articles fulfilled all inclusion criteria and were selected for data extraction and data analysis. The selection process and reasons for exclusion of the articles are summarized in Figure 1.

Study characteristics
There were total 575 patients in the selected studies and the age ranged from 20 to 93 years. 14 studies use one imaging techniques compared with the reference test when four studies use two imaging techniques compared with the reference. Twelve studies were performed prospectively. Six studies included patients with MM staged according to the diagnostic criteria of Durie-Salmon or Durie -Salmon Plus, one described patients with SPC only, and the remaining seven studies included patients with MM, SPC and/or MGUS. The number of patients included per study varied from 9 to 119. Scintigraphy was the index test in 8 (include four MIBI and four bone scan), MRI in 7, FDG-PET in 2 and FDG-PET-CT in 3 studies. Three papers included more than one index test. The characteristics of the included studies are presented in Table 1.

Methodological Quality Assessment
Methodological quality was assessed by 14 items

Discussion
Multiple myeloma (MM) is not a rare disease, and is a research focus in China (Chen et al., 2012;Liu et al., 2012;Lin et al., 2013;Zhang et al., 2014). It is reported that patients with MM could have symptoms, e.g., fractures, bone pain, and elevated blood calcium etc, caused by bone abnormalities (Lecouvet et al., 1999;Umeda et al., 2002;Terpos et al., 2003;Kitano et al., 2005;Harousseau et al., 2004). The consequence is a decreasing of quality of life. Guideline based diagnosis and treatment includes imaging examinations of head, chest, extremities, vertebra, and the pelvis. Regarding imaging techniques for bone lesions, X ray, MRI, scintigraphy, FDG-PET and PET/CT are usually considered. In clinical practice, the value of these imaging techniques could be different when different patients with MM are encountered, and the characteristics of each imaging techniques are compared. The sensitivity of these imaging techniques in detecting myeloma related bone lesions is different. X ray is reported to have a low sensitivity in diagnosing MM related bone disease, when >50% bone mineral content has been lost. It was demonstrated that MRI is associated with a very good sensitivity in particular at the early stage of the disease. The main limitation of MRI is caused by partial field, usually, only spine and pelvis could be viewed. Another point should be mentioned regarding MRI is its poor performance when metallic prosthesis or claustrophobia is presented.
Scintigraphy was also employed to diagnose MM related bone disease. However, it is considered that both 99mTc-diphosphonate and MIBI were associated with low sensitivity due to low tracer uptake by MM related bone disease and to high physiological uptake by liver that could mislead vertebral or right rib disease.
Within these years, PET/CT was frequently performed in the diagnosed of MM related bone disease. 18F-FDG PET is associated with higher sensitivity due to its characteristic that the whole body could be imaged and both medullary and extra-medullary lesions could be detected.
However, the sensitivity and specificity of PET/CT are depended on the uptake of tracer, and the hot spot could be equivocal when disease is in early stage or is disseminated.
Our results showed that MRI had a pooled sensitivity of 0.88, specificity of 0.68, AUC of 0.897, and Q*index of 0.828; whereas for MIBI, the corresponding values were 0.98, 0.90, 0.991, and 0.962; and for bone scan, the values were 066, 0.83, 0.805, and 0.740; and for PET and PET/CT, the value were 0.91, 0.69, 0.927 and 0.861, respectively. Statistically significant differences were not found in the sensitivity, specificity, AUC, and Q* index between MRI, scintigraphy, FDG-PET and PET/CT. Several confounding factors should be mentioned before we draw the conclusion. First, we are not sure whether the diagnostic consensus of PET/CT is reached in each institute. The technological skill of staff members in each institute could be greatly different, and the diagnostic standard adopted by each institute could not be standardized. Second, there could be difference between ethnic groups in our selected studies, Oriental and Caucasian were both included into current study. We hypothesize that the sensitivity and specificity of imaging test could be different between ethnic groups. And third, imaging techniques and quality for one device in different study group could be greatly differed. For CT and MRI machines, we are not sure these machines are produced during the same period of time. And we suppose machine manufactured in recent years could have higher sensitivity and specificity than those come into use before 20 years. Fourth, the stage of patients in each study group was not standardized. We know patients with more advanced disease should be diagnosed with MM related bone lesions more frequently than those with early disease. In conclusions: on the condition that X ray is taken as a reference in our study, we suggested that FDG-PET, PTE/CT, MRI and scintigraphy are all associated with high detection rate of bone disease in patients with MM. Thus, in clinical practice, it is recommended that we could choose these tests according to the condition of patient.