Diffusion-weighted imaging (DWI) and 18F-Fluorodeoxy glucose-positron emission tomography/computed tomography (FDG-PET/CT) are commonly used imaging modalities for the evaluation of various oncologic processes, including musculoskeletal tumors [1, 2, 3, 4, 5, 6]. Many studies in the literature have reported that both imaging modalities were capable of differentiating benign from malignant tumors. FDG uptake, expressed as maximum standardized uptake value (SUVmax), increases in malignant tumor cells, and has been reported to correlate with aggressiveness [7, 8, 9, 10, 11]. On the other hand, apparent diffusion coefficient (ADC) derived from DWI has been also reported to have a role in predicting malignancy and aggressiveness of the lesion [12, 13, 14].
Previous studies in the literature reported significant correlation between histopathological findings and SUVmax of the tumors [9, 10, 15, 16]. FDG-PET/CT reflects the glycolytic activity in a tumor region, thus most cellular activities including mitosis. As DWI shows the diffusivity of water molecule in the examined tissue, restricted diffusion also reflects cellular abundance and decreased extracellular space. As both DWI and FDG-PET/CT are functional imaging techniques that are used to evaluate tumor characteristics, one may hypothesize that these two imaging modalities may show a significant correlation. There are several studies in the literature investigating the correlation of these two modalities in different organ malignancies, but few studies include musculoskeletal malignancies [7, 9, 17, 18]. The aim of this study was to evaluate whether there is a correlation between the SUVmax of FDG-PET/CT and ADC values of DWI in musculoskeletal tumors.
The institutional ethics committee approval was obtained for this retrospective study (OMUKAEK2020/16) and the requirement for informed consent was waived. The standards for the reporting of diagnostic accuracy studies (STARD) were used .
A retrospective search of our hospital database was performed to identify patients who underwent magnetic resonance imaging (MRI) including DWI for the evaluation of a musculoskeletal tumor between January 2017 and March 2020, and 132 patients were found. Among those, 98 were excluded as they had no FDG-PET/CT examinations available; 4 were excluded as they underwent chemo/radiotherapy within the interval period between DWI and FDG-PET/CT examinations; 3 patients were excluded as the DWI and FDG-PET/CT examination interval period was more than eight weeks. As a result, 27 patients (17 men and 10 women) with a mean age of 44.85 ± 24.07 (range: 2–82) met the enrollment criteria. Twenty-six patients had a single tumor whereas one patient had three tumors. Therefore, 29 musculoskeletal tumors were involved in the study (Figure 1). The mean interval period between DWI and FDG-PET/CT was 18 ± 12 days (range: 1–45).
MRI examinations of the patients were performed with a 3T system (Ingenia, Philips, Netherlands). Different coils and imaging protocols were used according to the body site involved. The standard protocol for a musculoskeletal malignancy in the authors’ center included a longitudinal fat-suppressed T2-weighted (T2W) TSE sequence, axial T1W and T2W TSE sequences, longitudinal and axial fat-suppressed contrast-enhanced T1W TSE sequences. The DWI sequence was obtained before the contrast-enhanced sequences with the following parameters: TR/TE, 1500/80 ms; matrix, 116 × 114; field of view, 350 × 350 mm; slice thickness, 5 mm; gap, 1 mm; sensitivity encoding factor, 2; b values, 0, 200, 400 and 800s/mm2. ADC maps were calculated from images with b values of 0 and 800 s/mm2.
FDG-PET/CT scans from the top of the head through the feet of the patients were performed using a hybrid PET/CT (GE Discovery IQ 3Ring; equipped with 16 slice Optima540 CT Model) device. Patients fasted for eight hours before the examination. The images were acquired 60 to 90 minutes after the injection of 18F-FDG ((0.14–0.20 mCi/kg) in patients with glucose level less than 200 mg/dL.
DWI interpretation was performed by two radiologists (reader 1, A.V.P.; reader 2, M.O.) independently. The reviewers were blinded to the histopathological diagnosis and other imaging results of the lesions. Each interpreter placed five ellipsoid regions of interest (ROI) boxes on the lesion by referring the solid and most enhancing portion on the contrast-enhanced T1W images. Cystic portions, calcifications and any artifacts were carefully avoided. The mean ADC (ADCmean) and minimum ADC (ADCmin) of each reader’s measurements were stored for final statistical analysis.
An experienced nuclear medicine physician (F.C.T.) evaluated the FDG-PET/CT images without knowing the other imaging findings and histopathological results. The lesion standardized uptake value (SUV) was calculated automatically (activity in volume of interest (VOI)/[injected dose*body weight]) using volume-of-interest segmentation of the lesion. The SUVmax was defined as the hottest voxel within the VOI (Figures 2 and 3).
Pathology results of the patients were obtained by reviewing the patients’ chart. Histopathological classification of the soft tissue tumors was according to the 2013 WHO classification. Immunochemical and molecular studies were used to confirm certain diagnosis in certain cases. Tissue specimens were obtained either by surgery (n = 8) or core needle biopsy (n = 17). For four lesions, histopathological diagnosis was not available but the lesions appeared during follow-up imaging so they were accepted as metastases. Core-needle biopsies were performed by an experienced musculoskeletal radiologist (M.B.S.) using a 14–18G automatic core-needle biopsy needle (22 mm excursion; Maxicore, Ankara, Turkey).
Statistical analysis was performed with the SPSS 15.0 (SPSS Inc.; Chicago, IL, USA). Shapiro-Wilk test was used to assess normal distribution of the continuous data. Data were presented as mean and standard deviation (SD) or median and range for continuous variables, and as frequencies for categorical variables. Student t-test was used to compare the ADCmean, ADCmin and SUVmax between soft tissue and bone tumors. Pearson correlation was used to evaluate the correlation between SUVmax and ADC values. The degree of correlation was classified as follows: 0 ≤ r < 2, none; 2 ≤ r < 4, weak; 4 ≤ r < 6, moderate; 6 ≤ r < 8, strong; and 8 ≤ r, very strong correlations. Interobserver agreement of the ADC measurements was assessed using intraclass correlation coefficient. A p value less than 0.05 was indicative of statistical significance.
Demographic data of the study population is shown in Table 1. A total of 29 lesions in 27 patients were included in the study. There were 18 (62.1%) soft tissue tumors and 11 (37.9%) bone tumors. Their histological subtypes are presented in Table 2. Thirteen lesions (44.8%) were located in the trunk, 13 lesions (44.8%) were located in the lower extremity and three lesions (10.4%) were located in the upper extremity. The mean maximum diameter of the lesions was 9.4 ± 6.2 cm (range: 1.7–25).
|VARIABLE||ALL PATIENTS||PATIENTS WITH STT||PATIENTS WITH BT||P VALUE*|
|Age (years)||44.85 ± 24.07||45.19 ± 26.16||44.36 ± 21.88||0.932|
|Male||19 (65.5%)||13 (68.4%)||6 (31.6%)|
|Female||10 (35.5%)||5 (50%)||5 (50%)|
|Maximum diameter (cm)||9.39 ± 6.21||9.78 ± 6.95||8.74 ± 5.01||0.668|
|ADCmean (x10–3mm2/s)||0.85 ± 0.28||0.92 ± 0.30||0.75 ± 0.22||0.105|
|ADCmin (x10–3mm2/s)||0.61 ± 0.27||0.67 ± 0.30||0.50 ± 0.19||0.094|
|SUVmax||12.93 ± 9.63||12.23 ± 8.87||14.07 ± 11.11||0.627|
|TYPES OF TUMORS||NUMBER OF TUMORS|
|Soft tissue tumors||18 (62.1%)|
|Undifferentiated spindle cell sarcoma||1|
|Undifferentiated pleomorphic sarcoma||1|
|Bone tumors||11 (37.9%)|
The mean SUVmax, ADCmean and ADCmin of the whole lesions were 12.93 ± 9.63, 0.85 ± 0.28 × 10–3mm2/s and 0.61 ± 0.27 × 10–3mm2/s, respectively. There was a weak positive correlation between maximum lesion diameter and SUVmax (r = 0.378, p = 0.043). ADCmean and ADCmin did not correlate with the maximum diameter of the lesions (r = –0.161, p = 0.403; r = –0.214, p = 0.265 respectively).
The relationship between SUVmax and ADC values were demonstrated in Table 3. Considering all lesions included in the study, a strong inverse correlation was found between SUVmax and ADCmean (r = –0.616, p < 0.001; Figure 4). ADCmin showed strong inverse correlation with SUVmax (r = –0.638, p < 0.001).
|SUVMAX VS. ADCMEAN||SUVMAX VS. ADCMIN|
|r||p value||r||p value|
ADCmean and ADCmin of the lesions demonstrated a strong inverse correlation with SUVmax both separately in soft tissue tumors (r = –0.651, p = 0.003; r = –0.683, p = 0.002) and in bone tumors (r = -0.623, p = 0.041; r = –0.656, p = 0.028).
ADCmean and ADCmin measurements of the readers demonstrated almost perfect interobserver agreement (ICC:0.926, CI:0.849–0.965, p < 0.001; ICC:0.883, CI:0.766–0.943, p < 0.001, respectively).
Lee et al.  evaluated 25 bone and 32 soft tissue tumors using MRI and FDG-PET/CT. They included both malignant and benign tumors and reported significant inverse correlation between ADC and SUV values. Our study also revealed significant inverse correlation between ADC and SUV, although all of our lesions were malignant. However, their study differed from ours as they did not perform a subgroup correlation analysis of ADC and SUV in bone and soft tissue tumors individually. We found significant correlation between SUVmax and ADC values in both soft tissue and bone tumors.
Sagiyama et al.  assessed the correlation between ADC and SUV of 35 musculoskeletal tumors using FDG-PET/MRI and reported significant correlation between these two entities. Rakheja et al.  evaluated the correlation of ADC and SUV in 52 non-osseous and 17 osseous tumors using FDG-PET/MRI. They reported significant correlation between ADC and SUV values in all osseous and non-osseous lesions; however, subgroup analysis of osseous lesions did not demonstrate correlation between ADC and SUV. Calcification and reparative sclerosis in bone tumors may correspond to low ADC values on DWI. Therefore, low ADC in bone tumors may reflect the calcification of the tissue instead of malignant cells, which might be the underlying reason that ADCmin and SUVmax did not correlate in bone tumors of their study. In our study, although the difference was not statistically significant, ADCmean and ADCmin of bone tumors were lower than those of the soft tissue tumors.
In another study evaluating 136 sarcomas by Rakheja et al. , SUV values were found to be correlated with mitotic cell count, presence of necrosis and histological grade of the tumors. Considering the significant heterogeneity and variety of sarcomas, authors suggested performing the biopsy of the lesions from the areas demonstrating maximum SUV uptake for accurate grading and treatment planning. Although we did not assess correlation between SUV, ADC and histological features, correlation between SUV and ADC confirmed that biopsy of musculoskeletal lesions should be performed from the areas with lowest ADC values.
This study has several limitations. First of all, it included only a limited number of tumors. More studies with larger populations are needed to validate our results. Second, this study did not evaluate the correlation between histological parameters and SUV and ADC, as most histological parameters were not available. Third, SUV and ADC measurements were performed by different physicians on different topologies. Therefore, we used blinded comparison of the SUV and ADC measurements. One may hypothesize that a non-blind comparison assessing how often SUVmax was on the same area with ADCnean and ADCmin may have resulted in more significant correlations. However, comparing these two modalities blindly, settings may be more valuable considering the daily practice. Future studies may assess whether SUVmax corresponds to the same area with ADCmin and vice versa.
In conclusion, SUVmax of FDG-PET/CT has a strong but inverse correlation with the ADCmean and ADCmin values obtained from DWI in musculoskeletal tumors.
The authors have no competing interests to declare.
Iima M, Le Bihan D. Clinical intravoxel incoherent motion and diffusion MR imaging: Past, present, and future. Radiology. 2016; 278(1): 13–32. DOI: https://doi.org/10.1148/radiol.2015150244
Koh DM, Collins DJ. Diffusion-weighted MRI in the body: Applications and challenges in oncology. AJR Am J Roentgenol. 2007; 188(6): 1622–35. DOI: https://doi.org/10.2214/AJR.06.1403
Lee SY, Jee WH, Jung JY, et al. Differentiation of malignant from benign soft tissue tumours: Use of additive qualitative and quantitative diffusion-weighted MR imaging to standard MR imaging at 3.0 T. Eur Radiol. 2016; 26(3): 743–54. DOI: https://doi.org/10.1007/s00330-015-3878-x
Sung JK, Jee WH, Jung JY, et al. Differentiation of acute osteoporotic and malignant compression fractures of the spine: Use of additive qualitative and quantitative axial diffusion-weighted MR imaging to conventional MR imaging at 3.0 T. Radiology. 2014; 271(2): 488–98. DOI: https://doi.org/10.1148/radiol.13130399
Nose H, Otsuka H, Otomi Y, et al. Correlations between F-18 FDG PET/CT and pathological findings in soft tissue lesions. J Med Invest. 2013; 60(3–4): 184–90. DOI: https://doi.org/10.2152/jmi.60.184
Shin DS, Shon OJ, Han DS, Choi JH, Chun KA, Cho IH. The clinical efficacy of (18)F-FDG-PET/CT in benign and malignant musculoskeletal tumors. Ann Nucl Med. 2008; 22(7): 603–9. DOI: https://doi.org/10.1007/s12149-008-0151-2
Kitajima K, Yamano T, Fukushima K, et al. Correlation of the SUVmax of FDG-PET and ADC values of diffusion-weighted MR imaging with pathologic prognostic factors in breast carcinoma. Eur J Radiol. 2016; 85(5): 943–9. DOI: https://doi.org/10.1016/j.ejrad.2016.02.015
Wang CL, MacDonald LR, Rogers JV, Aravkin A, Haseley DR, Beatty JD. Positron emission mammography: Correlation of estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2 status and 18F-FDG. AJR Am J Roentgenol. 2011; 197(2): W247–55. DOI: https://doi.org/10.2214/AJR.11.6478
Rakheja R, Makis W, Skamene S, et al. Correlating metabolic activity on 18F-FDG PET/CT with histopathologic characteristics of osseous and soft-tissue sarcomas: A retrospective review of 136 patients. AJR Am J Roentgenol. 2012; 198(6): 1409–16. DOI: https://doi.org/10.2214/AJR.11.7560
Folpe AL, Lyles RH, Sprouse JT, Conrad EU 3rd, Eary JF. (F-18) Fluorodeoxyglucose positron emission tomography as a predictor of pathologic grade and other prognostic variables in bone and soft tissue sarcoma. Clin Cancer Res. 2000; 6(4): 1279–87.
Bonarelli C, Teixeira PA, Hossu G, et al. Impact of ROI positioning and lesion morphology on apparent diffusion coefficient analysis for the differentiation between benign and malignant non fatty soft-tissue lesions. AJR Am J Roentgenol. 2015; 205(1): W106–13. DOI: https://doi.org/10.2214/AJR.14.13865
Einarsdóttir H, Karlsson M, Wejde J, Bauer HC. Diffusion-weighted MRI of soft tissue tumours. Eur Radiol. 2004; 14(6): 959–63. DOI: https://doi.org/10.1007/s00330-004-2237-0
Genovese E, Canì A, Rizzo S, Angeretti MG, Leonardi A, Fugazzola C. Comparison between MRI with spin-echo echo-planar diffusion-weighted sequence (DWI) and histology in the diagnosis of soft-tissue tumours. Radiol Med. 2011; 116(4): 644–56. DOI: https://doi.org/10.1007/s11547-011-0666-9
Hain SF, O’Doherty MJ, Bingham J, Chinyama C, Smith MA. Can FDG PET be used to successfully direct preoperative biopsy of soft tissue tumours? Nucl Med Commun. 2003; 24(11): 1139–43. DOI: https://doi.org/10.1097/00006231-200311000-00003
Hicks RJ, Toner GC, Choong PF. Clinical applications of molecular imaging in sarcoma evaluation. Cancer Imaging. 2005; 5(1): 66–72. DOI: https://doi.org/10.1102/1470-7330.2005.0008
Lee SY, Jee WH, Yoo IR, et al. Comparison of 3T diffusion-weighted MRI and 18F-FDG PET/CT in musculoskeletal tumours: Quantitative analysis of apparent diffusion coefficients and standardized uptake values. Br J Radiol. 2019; 92(1102): 20181051. DOI: https://doi.org/10.1259/bjr.20181051
Byun BH, Kong CB, Lim I, et al. Combination of 18F-FDG PET/CT and diffusion-weighted MR imaging as a predictor of histologic response to neoadjuvant chemotherapy: Preliminary results in osteosarcoma. J Nucl Med. 2013; 54(7): 1053–9. DOI: https://doi.org/10.2967/jnumed.112.115964
Bossuyt PM, Reitsma JB, Bruns DE, et al. STARD 2015: An updated list of essential items for reporting diagnostic accuracy studies. Radiology. 2015; 277(3): 826–32. DOI: https://doi.org/10.1148/radiol.2015151516
Sagiyama K, Watanabe Y, Kamei R, et al. Multiparametric voxel-based analyses of standardized uptake values and apparent diffusion coefficients of soft-tissue tumours with a positron emission tomography/magnetic resonance system: Preliminary results. Eur Radiol. 2017; 27(12): 5024–33. DOI: https://doi.org/10.1007/s00330-017-4912-y
Park S, Yoon JK, Chung NS, et al. Correlations between intravoxel incoherent motion diffusion-weighted MR imaging parameters and 18F-FDG PET/CT metabolic parameters in patients with vertebral bone metastases: Initial experience. Br J Radiol. 2018; 91(1086): 20170889. DOI: https://doi.org/10.1259/bjr.20170889
Rakheja R, Chandarana H, DeMello L, et al. Correlation between standardized uptake value and apparent diffusion coefficient of neoplastic lesions evaluated with whole-body simultaneous hybrid PET/MRI. AJR Am J Roentgenol. 2013; 201(5): 1115–9. DOI: https://doi.org/10.2214/AJR.13.11304