CME JOURNAL OF MAGNETIC RESONANCE IMAGING 41:935–940 (2015)
Original Research
Evaluation of Femoral Perfusion in a Rabbit Model of Steroid-Induced Osteonecrosis by Dynamic Contrast-Enhanced MRI with a High Magnetic Field MRI System Shigeki Hayashi, MD,1 Mikihiro Fujioka, MD, PhD,1* Kazuya Ikoma, MD, PhD,1 Masazumi Saito, MD, PhD,1 Keiichiro Ueshima, MD, PhD,1 Masashi Ishida, MD, PhD,1 Masaaki Kuribayashi, MD, PhD,1 Akira Ikegami, MD,1 Osam Mazda, MD, PhD,2 and Toshikazu Kubo, MD, PhD1 Key Words: steroid-induced osteonecrosis; evaluation of femoral perfusion; dynamic contrast-enhanced MRI; high magnetic field MRI system J. Magn. Reson. Imaging 2015;41:935–940. C 2014 Wiley Periodicals, Inc. V
Purpose: To evaluate perfusion during the early phase after steroid administration in vivo using dynamic contrast-enhanced magnetic resonance imaging (DCEMRI) with a high magnetic field MRI system. The main pathogenesis of steroid-induced osteonecrosis is considered to be ischemia. Materials and Methods: A single dose of methylprednisolone (MPSL) was injected into nine rabbits. DCE-MRI was performed for these rabbits before MPSL administration and 1, 5, 10, and 14 days after administration. Time–signal intensity curves were created for each femur based on the signal intensity to evaluate perfusion. Enhancement ratio (ER), initial slope (IS), and area under the curve (AUC) were calculated and the value before MPSL administration and the minimal value after administration were compared statistically. Results: ER, IS, and AUC values after MPSL administration significantly decreased (P < 0.05, P < 0.01, and P < 0.01, respectively). All of them decreased by the 5th day in 56% of the femora and by the 14th day in 83%, and some femora even showed a decrease from the 1st day. Conclusion: In this study, decreased perfusion in the femora after steroid administration was proven. Additionally, we could show that it occurred from the early days after steroid administration.
1 Department of Orthopaedics, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kyoto, Japan. 2 Department of Immunology, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kyoto, Japan. Contract grant sponsor: Research on Rare and Intractable Diseases, Health and Labour Sciences Research; Contract grant number: H24nanchi-ippan-001; Contract grant sponsor: Hip Joint Foundation of Japan. *Address reprint requests to: M.F., Department of Orthopaedics, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kawaramachi-Hirokoji, Kamigyo-ku, Kyoto, 602-8566, Japan. E-mail: [email protected] Received December 31, 2013; Accepted March 12, 2014. DOI 10.1002/jmri.24632 View this article online at wileyonlinelibrary.com. C 2014 Wiley Periodicals, Inc. V
ISCHEMIA AFTER STEROID ADMINISTRATION is thought to be the main pathogenesis for steroidinduced osteonecrosis (ON) (1). We previously reported that in patients who underwent kidney transplantation, a band image in the femoral head shown by T1weighted images was observed on magnetic resonance images as early as 6 weeks after steroid administration (2). An experimental study in beagle dogs reported that band images appeared 4 weeks after the occurrence of ischemia in the femoral head on magnetic resonance images (3). Based on these findings, in steroid-induced ON, ON in the femur is thought to occur within 2 weeks after steroid administration. Recently, in a rabbit model of steroid-induced ON, it was reported that the expression of vascular endothelial growth factor (VEGF) (4) and oxidative DNA damage (5) occurred early on the 3rd to 5th days after steroid administration. These results suggest that steroids exert significant effects on the body soon after steroid administration. Magnetic resonance imaging (MRI), especially dynamic contrast-enhanced MRI (DCE-MRI) is often used in the evaluation of perfusion (6–13). In this study, we sought to especially evaluate the early temporal changes in perfusion in vivo after steroid administration. MATERIALS AND METHODS Animals Male Japanese white rabbits older than 28 weeks, weighing 3.0–4.0 kg (Oriental BioService, Kyoto,
935
936
Hayashi et al.
Japan), were kept in separate cages. The animals were housed at the Animal Center of Kyoto Prefectural University of Medicine, fed nutritionally adequate food daily, and had free access to clean drinking water. This study fully followed the guidelines of the Kyoto Prefectural University of Medicine Animal Care and Use Committee. Treatment Steroid-induced ON was produced by an intramuscular injection of methylprednisolone (MPSL, 20 mg/kg body weight) in nine rabbits (13). This experimental model induces ON in the femoral bones of 60–70% of treated rabbits (13,14). MRI Study MR images of the proximal half of bilateral femora were obtained before MPSL administration and 1, 5, 10, and 14 days after administration using a 7.04 T MRI system (MRI Systems, Agilent Technologies, Palo Alto, CA). Before MRI scanning, each rabbit was anesthetized by inhalation of isoflurane at 2.5 L/min. A plastic venous cannula was placed in an auditory vein. Each rabbit was placed in an imaging coil in the supine position. A total of 40 DCE-MR images were taken consecutively at the same interval; 10 before the contrast agent administration (“phase a”) and 30 images after the contrast agent administration. Further, the period after the contrast agent administration was divided into two phases for analysis; the phase just after the contrast agent administration (“phase b,” 15 images) and the phase when contrast agent distribution is considered to reach steady state (“phase c,” 15 images). The total acquisition time was 136 seconds. DCE-MR images (relaxation time / echo time [TR/ TE] ¼ 18/3 msec, flip angle ¼ 20 ) were obtained in the coronal plane with spoiled gradient echo sequencing. Imaging parameters were as follows: slice thickness, 2 mm; imaging matrix, 128 128; and square field of view, 100 mm. The acquisition time was 2.3 seconds per image and the interval between acquisitions was 1.1 seconds. As the contrast agent, a bolus of 0.2 mmol per kg of body weight of gadopentetate dimeglumine (Gd-DTPA, Magnevist, Schering, Berlin, Germany) was injected into an auditory vein followed by a 5-mL saline flush.
Figure 1. A representative image of the serial images of DCE-MRI. The ROI was set at the common site of osteonecrosis development in bilateral femora. Blue and red areas show ROIs for the common site of development of osteonecrosis in the right and left femora, respectively. SI in the ROI was measured.
The measurement was performed twice by the author and one of the co-authors (A.I.) under blind condition, and the results were compared and evaluated statistically to verify that the inter- and intraexaminer variability was not significant. The mean SI of 10 measurements before injecting the contrast agent (pre-SI) was calculated. A total of 40 images were taken before and after the contrast agent injection, and their SIs were individually divided by pre-SI to obtain RIs at each timepoint. RI(t) for each femur was plotted against time to create a time– signal intensity curve (Fig. 2). Each of the time–signal intensity curves created were divided into three phases: “phase a,” which included 10 timepoints before the contrast agent injection; “phase b,” which included 15 timepoints just after the injection and showed increasing RI; and “phase c,” which included 15 timepoints when RI reached a steady state after the injection. The following four items were evaluated. Change in the Time–Signal Intensity Curve Over Time Changes in the shape of the time–signal intensity curve were evaluated from before steroid administration to 1, 5, 10, and 14 days after administration. Determination of the Enhancement Ratio (ER)
Analysis of DCE-MRI The region of interest (ROI) was set between the proximal metaphysis and proximal diaphysis where is the common site of intra-femoral ON in rabbits (Fig. 1). The variations of arterial input function and the vascular structure of each subject can affect the results, which makes it essential to normalize the perfusion parameters in the ROI to those of a reference region. In light of this, the ROI was set so that as much bonemarrow space as possible between the proximal metaphysis and proximal diaphysis was included. The signal intensity (SI) in the ROI was measured.
SI reaches steady state 30 seconds after a contrast agent injection. The ER was obtained by dividing the mean RI(t) of 15 timepoints in “phase c” when steady state was achieved by the mean RI(t) of 10 timepoints in “phase a” before the contrast agent injection (pre-SI). Determination of the Initial Slope (IS) RI level started to rise statistically at t ¼ 40.8 (second image after contrast agent injection) compared with the average RI level during the “phase a” (pre-SI) (paired t-test, P < 0.01). While, the level at t ¼ 57.8 (seventh image after the injection) was significantly
Femoral Perfusion in Steroid-Induced ON
937
(M.I., M.S.). Both of them are orthopedists with more than 10 years of experience who are skilled in tissue evaluation regarding bone necrosis. Statistical Analysis
Figure 2. A representative example of a time–signal intensity curve. Blue and red lines signify the time–signal intensity curves for the right and left femora. RI was enhanced after bolus injection of a contrast agent at the time shown by the arrow. Forty sequential RIs were divided into three phases: (a) 10 measurements before contrast agent injection (in the yellow frame); (b) 15 measurements during the period when SI was enhanced just after the contrast agent injection (in the red frame); and (c) 15 measurements at the stage when RI reached steady state after the contrast agent injection (in the green frame).
higher than the level on the previous image (paired ttest, P < 0.01) but not significantly different from that on the next image (paired t-test, P ¼ 0.18). Thus, we considered that the contrast effect appeared sufficiently at t ¼ 57.8. The initial slope of RI was obtained by dividing the increment at t ¼ 57.8 from pre-SI with t ¼ 20.4, the time between when the first and seventh images were taken. The equation used was: Initial slope ¼ (RI(57.8)-1)/20.4. Calculation of the Area Under the Curve (AUC) The AUC was calculated by adding RI(t) of 15 timepoints in “phase b,” directly after the injection of the contrast agent, using the following equation: AUC ¼ P85 RIðT Þ. T ¼34 Histological Study Immediately after the MRI was obtained 14 days after steroid administration, the rabbits were sacrificed with an excessive dose of intravenously injected sodium pentobarbital. The bilateral femora were immediately harvested and divided at the middle to obtain the proximal halves. They were fixed in 10% buffered formalin for 1 week and then decalcified with 10% EDTA for 5 weeks. After decalcification, the femora were embedded in paraffin. Four-micrometer-thick coronal sections were prepared and stained with hematoxylin and eosin. ON was defined as the presence of empty lacunae or obvious pyknotic nuclei of osteocytes within the bone trabeculae, accompanied by necrosis of surrounding bone marrow hematopoietic cells, or the presence of diffuse necrosis of bone marrow hematopoietic cells and fat cells (16). ON and all other histologic measures were assessed blindly by two independent authors
In this study we evaluated perfusion, for which interindividual differences often occur (12,17). Therefore, we also evaluated interindividual differences using the Wilcoxon signed-rank test in statistical analysis. For the evaluation of blood flow in the ROI in the femur, each value of the ER, IS, and AUC was statistically compared between the value before steroid administration and the minimal value after steroid administration. SPSS Statistics (v. 21.0, IBM, Somers, NY) was used as the statistical analysis software. We considered P < 0.05 statistically significant. RESULTS DCE-MRI of the Femur Sequential Changes in the Shape of the Time–Signal Intensity Curves Sequential changes in the shape of the time–signal intensity curves of two rabbits that represent typical time course changes are shown in Figs. 3 and 4. In the first rabbit, RI levels after the contrast agent injection were lower on the 5th day after steroid administration compared with those before steroid administration. RI levels improved to the same levels as before steroid administration on the 10th and 14th days after steroid administration (Fig. 3A–E). In the second rabbit, RI levels were higher on the 1st day after steroid administration compared with those before administration in both femora, but especially in the left femur. RI levels decreased to the same levels as before steroid administration on the 5th day after administration, and became lower than the level before steroid administration on the 10th day. RI levels improved on the 14th day (Fig. 4A–E). As mentioned above, some rabbits showed changes in the time–signal intensity curves even from the 1st day after steroid administration, and all of the rabbits showed changes over time. Changes in ER of the Femora The ER increased in three out of 18 femora and decreased in all of the other 15 femora after steroid administration. Statistically, the ER changed after steroid administration compared with before administration, and the direction of change was toward decrease, which was significant (P < 0.05; Fig. 5A). The day when the ER of each femora showed the minimal value was the 1st day after steroid administration in five femora (28%), the 5th day in five femora (27%), the 10th day in three femora (17%), and the 14th day in two femora (11%). Change in the IS of the Femora The IS increased in two out of 18 femora and decreased in all of the other 16 femora after steroid
938
Hayashi et al.
Figure 3. A representative example of changes in the time–signal intensity curve over time. The time–signal intensity curve shows the following. A: before steroid administration; B: the 1st day after steroid administration; C: the 5th day after steroid administration; D: the 10th day after steroid administration; and E: the 14th day after steroid administration. The increase in RI after the contrast agent injection was slightly suppressed on the 1st day after steroid administration and was further suppressed on the 5th day compared with the curve before steroid administration. However, on the 10th and 14th days after steroid administration, RI after the contrast agent injection was increased to as high as that before steroid administration.
administration. Statistically, the IS changed after steroid administration compared with before administration, and the direction of change was toward decrease, which was significant (P < 0.01; Fig. 5B). The day when the IS of each femora showed the minimal value was the 1st day after steroid administration in four femora (22%), the 5th day in six femora (33%), the 10th day in two femora (11%), and the 14th day in four femora (22%). Change in the AUC of the Femora The AUC increased in three out of 18 femora, and decreased in all of the other 15 femora after steroid administration. Statistically, the AUC changed after steroid administration compared with before administration, and the direction of change was toward decrease, which was significant (P < 0.01; Fig. 5C). The day when the AUC of each femora showed the
minimal value was the 1st day after steroid administration in five femora (27%), the 5th day in five femora (27%), the 10th day in two femora (11%), and the 14th day in three femora (17%).
Comparison Between Histological Results and DCE-MRI Results One out of 18 femora in one of nine rabbits showed pyknotic nuclei of osteocytes within the bone trabeculae. This was accompanied by necrosis of surrounding bone marrow hematopoietic cells and fat cells, which was determined to be ON. The ER, IS, and AUC of the one femur that developed ON were decreased on the 5th day after steroid administration compared with those before steroid administration. The ER and IS showed the third lowest
Femoral Perfusion in Steroid-Induced ON
939
Figure 4. Changes in the time–signal intensity curve over time in another rabbit. The time signal intensity curve shows the following. A: before steroid administration; B: the 1st day after steroid administration; C: the 5th day after steroid administration; D: the 10th day after steroid administration; and E: the 14th day after steroid administration. On the 1st day after steroid administration, both of the femora, but especially the left femur, showed higher RIs after the injection of contrast agent compared with those before steroid administration. On the 5th day, RI of the left femur, which had been high, decreased to the level observed before the steroid administration. On the 10th day, RI after the contrast agent injection clearly decreased compared with that before steroid administration. On the 14th day, RI after the contrast agent injection was higher than that before steroid administration.
value among the 18 femora, and its AUC showed the second lowest value among the 18 femora. DISCUSSION DCE-MRI is often used for the evaluation of blood flow with MRI (12). Parameters used for this evaluation include the ER (6), IS (7), AUC (8), relative SI (9), time to peak enhancement (10), and washout slope (11). Because of the effect of interrater differences and motion artifacts, one parameter alone cannot precisely explain the phenomena. Therefore, evaluation of blood flow should be comprehensively made based on multiple parameters (12). In the current study, the IS and AUC were used to evaluate the change in SI from immediately after the contrast agent injection, and the
ER was used to evaluate the change in SI after it reached a plateau after the contrast agent injection (12,13). The major etiological factor for ON is considered to be ischemia, which occurs after steroid administration (1). However, the presence of ischemia in ON in vivo has not been proven. In our study, all of the three parameters (IS, ER, and AUC) were significantly decreased at the common development site of ON of the femoral head. Therefore, judging these three parameters comprehensively, we considered that blood flow was actually decreasing after steroid administration. Based on clinical cases, the timing when ischemia occurs after steroid administration had been considered to be 2 weeks after administration (2,3). However, recent studies have shown that VEGF expression levels increase on the 3rd day after steroid administration (4)
940
Hayashi et al.
uated indirectly through the confirmation of increased VEGF levels, which increase after the occurrence of ischemia, by sacrificing animals. This study is important from two perspectives. First, perfusion was able to be evaluated in vivo over time, and second, we showed that blood flow starts to decrease in the early phase after steroid administration. The limitation of this study is that the number of femurs that developed ON was small, 1 out of 18 femora, which was not sufficient to discuss any possible relationship between ON development in the tissue and the perfusion status observed in this study. In conclusion, we assessed perfusion in the femora during early days after steroid administration and could confirm that perfusion actually decreases soon after steroid administration. REFERENCES
Figure 5. Changes in the ER (graph A), IS (graph B), and AUC (graph C). “before steroid administration” are values before the steroid administration and “after steroid administration” are minimal values after the steroid administration. ER, IS, and AUC values after MPSL administration significantly decreased (P < 0.05, P < 0.01, and P < 0.01, respectively).
and that oxidative damage of DNA occurs by the 5th day after steroid administration (5). Therefore, we considered that the change in perfusion might occur by the 5th day after steroid administration, and focused on this time window for evaluation. We found that blood flow decreased from the 1st day after steroid administration in some femora. We showed that blood flow actually decreases at the time when VEGF levels increase and oxidative damage of DNA occurs. In our study we observed decreased blood flow in the bone marrow of femora after steroid administration in vivo. We also showed that blood flow decreased from the early phase, by observing decreased blood flow in some femora from the 1st day after steroid administration. Without MRI, perfusion cannot be evaluated over time in vivo. Ischemia can only be eval-
1. Kerachian MA, Harvey EJ, Cournoyer D, Chow TY, S eguin C. Avascular necrosis of the femoral head: vascular hypotheses. Endothelium 2006;13:237–244. 2. Kubo T, Yamazoe S, Sugano N, et al. Initial MRI findings of nontraumatic osteonecrosis of the femoral head in renal allograft recipients. Magn Reson Imaging 1997;15:1017–1023. 3. Nakamura T, Matsumoto T, Nishino M, Tomita K, Kadoya M. Early magnetic resonance imaging and histologic findings in a model of femoral head necrosis. Clin Orthop Relat Res 1997; 68–72. 4. Kabata T, Matsumoto T, Yagishita S, Wakayama T, Iseki S, Tomita K. Vascular endothelial growth factor in rabbits during development of corticosteroid-induced osteonecrosis: a controlled experiment. J Rheumatol 2008;35:2383–2390. 5. Ichiseki T, Ueda Y, Katsuda S, Kitamura K, Kaneuji A, Matsumoto T. Oxidative stress by glutathione depletion induces osteonecrosis in rats. Rheumatology (Oxford) 2006;45:287–290. 6. Mussurakis S, Gibbs P, Horsman A. Primary breast abnormalities: selective pixel sampling on dynamic gadolinium-enhanced MR images. Radiology 1998;206:465–473. 7. Karahaliou A, Vassiou K, Arikidis NS, Skiadopoulos S, Kanavou T, Costaridou L. Assessing heterogeneity of lesion enhancement kinetics in dynamic contrast-enhanced MRI for breast cancer diagnosis. Br J Radiol 2010;83:296–309. 8. Evelhoch JL. Key factors in the acquisition of contrast kinetic data for oncology. J Magn Reson Imaging 1999;10:254–259. 9. Johansen R, Jensen LR, Rydland J, et al. Predicting survival and early clinical response to primary chemotherapy for patients with locally advanced breast cancer using DCE-MRI. J Magn Reson Imaging 2009;29:1300–1307. 10. Liney GP, Gibbs P, Hayes C, Leach MO, Turnbull LW. Dynamic contrast-enhanced MRI in the differentiation of breast tumors: user-defined versus semi-automated region-of-interest analysis. J Magn Reson Imaging 1999;10:945–949. 11. Jia G, O’Dell C, Heverhagen JT, et al. Colorectal liver metastases: contrast agent diffusion coefficient for quantification of contrast enhancement heterogeneity at MR imaging. Radiology 2008;248:901–909. 12. Yang X, Knopp MV. Quantifying tumor vascular heterogeneity with dynamic contrast-enhanced magnetic resonance imaging: a review. J Biomed Biotechnol 2011;2011:732848. 13. Kiessling F, Lichy M, Grobholz R, et al. Simple models improve the discrimination of prostate cancers from the peripheral gland by T1-weighted dynamic MRI. Eur Radiol 2004;14:1793–1801. 14. Yamamoto T, Irisa T, Sugioka Y, Sueishi K. Effects of pulse methylprednisolone on bone and marrow tissues: corticosteroid-induced osteonecrosis in rabbits. Arthritis Rheum 1997;40:2055–2064. 15. Kuribayashi M, Fujioka M, Takahashi KA, et al. Vitamin E prevents steroid-induced osteonecrosis in rabbits. Acta Orthop 2010;81:154–160. 16. Yamamoto T, Hirano K, Tsutsui H, Sugioka Y, Sueishi K. Corticosteroid enhances the experimental induction of osteonecrosis in rabbits with Shwartzman reaction. Clin Orthop Relat Res 1995; 235–243. 17. Byun J, Lai D, Luo S, Risser J, Tung B, Hardy RJ. A hybrid method in combining treatment effects from matched and unmatched studies. Stat Med 2013;32:4924–4937.
Original Research
Evaluation of Femoral Perfusion in a Rabbit Model of Steroid-Induced Osteonecrosis by Dynamic Contrast-Enhanced MRI with a High Magnetic Field MRI System Shigeki Hayashi, MD,1 Mikihiro Fujioka, MD, PhD,1* Kazuya Ikoma, MD, PhD,1 Masazumi Saito, MD, PhD,1 Keiichiro Ueshima, MD, PhD,1 Masashi Ishida, MD, PhD,1 Masaaki Kuribayashi, MD, PhD,1 Akira Ikegami, MD,1 Osam Mazda, MD, PhD,2 and Toshikazu Kubo, MD, PhD1 Key Words: steroid-induced osteonecrosis; evaluation of femoral perfusion; dynamic contrast-enhanced MRI; high magnetic field MRI system J. Magn. Reson. Imaging 2015;41:935–940. C 2014 Wiley Periodicals, Inc. V
Purpose: To evaluate perfusion during the early phase after steroid administration in vivo using dynamic contrast-enhanced magnetic resonance imaging (DCEMRI) with a high magnetic field MRI system. The main pathogenesis of steroid-induced osteonecrosis is considered to be ischemia. Materials and Methods: A single dose of methylprednisolone (MPSL) was injected into nine rabbits. DCE-MRI was performed for these rabbits before MPSL administration and 1, 5, 10, and 14 days after administration. Time–signal intensity curves were created for each femur based on the signal intensity to evaluate perfusion. Enhancement ratio (ER), initial slope (IS), and area under the curve (AUC) were calculated and the value before MPSL administration and the minimal value after administration were compared statistically. Results: ER, IS, and AUC values after MPSL administration significantly decreased (P < 0.05, P < 0.01, and P < 0.01, respectively). All of them decreased by the 5th day in 56% of the femora and by the 14th day in 83%, and some femora even showed a decrease from the 1st day. Conclusion: In this study, decreased perfusion in the femora after steroid administration was proven. Additionally, we could show that it occurred from the early days after steroid administration.
1 Department of Orthopaedics, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kyoto, Japan. 2 Department of Immunology, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kyoto, Japan. Contract grant sponsor: Research on Rare and Intractable Diseases, Health and Labour Sciences Research; Contract grant number: H24nanchi-ippan-001; Contract grant sponsor: Hip Joint Foundation of Japan. *Address reprint requests to: M.F., Department of Orthopaedics, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kawaramachi-Hirokoji, Kamigyo-ku, Kyoto, 602-8566, Japan. E-mail: [email protected] Received December 31, 2013; Accepted March 12, 2014. DOI 10.1002/jmri.24632 View this article online at wileyonlinelibrary.com. C 2014 Wiley Periodicals, Inc. V
ISCHEMIA AFTER STEROID ADMINISTRATION is thought to be the main pathogenesis for steroidinduced osteonecrosis (ON) (1). We previously reported that in patients who underwent kidney transplantation, a band image in the femoral head shown by T1weighted images was observed on magnetic resonance images as early as 6 weeks after steroid administration (2). An experimental study in beagle dogs reported that band images appeared 4 weeks after the occurrence of ischemia in the femoral head on magnetic resonance images (3). Based on these findings, in steroid-induced ON, ON in the femur is thought to occur within 2 weeks after steroid administration. Recently, in a rabbit model of steroid-induced ON, it was reported that the expression of vascular endothelial growth factor (VEGF) (4) and oxidative DNA damage (5) occurred early on the 3rd to 5th days after steroid administration. These results suggest that steroids exert significant effects on the body soon after steroid administration. Magnetic resonance imaging (MRI), especially dynamic contrast-enhanced MRI (DCE-MRI) is often used in the evaluation of perfusion (6–13). In this study, we sought to especially evaluate the early temporal changes in perfusion in vivo after steroid administration. MATERIALS AND METHODS Animals Male Japanese white rabbits older than 28 weeks, weighing 3.0–4.0 kg (Oriental BioService, Kyoto,
935
936
Hayashi et al.
Japan), were kept in separate cages. The animals were housed at the Animal Center of Kyoto Prefectural University of Medicine, fed nutritionally adequate food daily, and had free access to clean drinking water. This study fully followed the guidelines of the Kyoto Prefectural University of Medicine Animal Care and Use Committee. Treatment Steroid-induced ON was produced by an intramuscular injection of methylprednisolone (MPSL, 20 mg/kg body weight) in nine rabbits (13). This experimental model induces ON in the femoral bones of 60–70% of treated rabbits (13,14). MRI Study MR images of the proximal half of bilateral femora were obtained before MPSL administration and 1, 5, 10, and 14 days after administration using a 7.04 T MRI system (MRI Systems, Agilent Technologies, Palo Alto, CA). Before MRI scanning, each rabbit was anesthetized by inhalation of isoflurane at 2.5 L/min. A plastic venous cannula was placed in an auditory vein. Each rabbit was placed in an imaging coil in the supine position. A total of 40 DCE-MR images were taken consecutively at the same interval; 10 before the contrast agent administration (“phase a”) and 30 images after the contrast agent administration. Further, the period after the contrast agent administration was divided into two phases for analysis; the phase just after the contrast agent administration (“phase b,” 15 images) and the phase when contrast agent distribution is considered to reach steady state (“phase c,” 15 images). The total acquisition time was 136 seconds. DCE-MR images (relaxation time / echo time [TR/ TE] ¼ 18/3 msec, flip angle ¼ 20 ) were obtained in the coronal plane with spoiled gradient echo sequencing. Imaging parameters were as follows: slice thickness, 2 mm; imaging matrix, 128 128; and square field of view, 100 mm. The acquisition time was 2.3 seconds per image and the interval between acquisitions was 1.1 seconds. As the contrast agent, a bolus of 0.2 mmol per kg of body weight of gadopentetate dimeglumine (Gd-DTPA, Magnevist, Schering, Berlin, Germany) was injected into an auditory vein followed by a 5-mL saline flush.
Figure 1. A representative image of the serial images of DCE-MRI. The ROI was set at the common site of osteonecrosis development in bilateral femora. Blue and red areas show ROIs for the common site of development of osteonecrosis in the right and left femora, respectively. SI in the ROI was measured.
The measurement was performed twice by the author and one of the co-authors (A.I.) under blind condition, and the results were compared and evaluated statistically to verify that the inter- and intraexaminer variability was not significant. The mean SI of 10 measurements before injecting the contrast agent (pre-SI) was calculated. A total of 40 images were taken before and after the contrast agent injection, and their SIs were individually divided by pre-SI to obtain RIs at each timepoint. RI(t) for each femur was plotted against time to create a time– signal intensity curve (Fig. 2). Each of the time–signal intensity curves created were divided into three phases: “phase a,” which included 10 timepoints before the contrast agent injection; “phase b,” which included 15 timepoints just after the injection and showed increasing RI; and “phase c,” which included 15 timepoints when RI reached a steady state after the injection. The following four items were evaluated. Change in the Time–Signal Intensity Curve Over Time Changes in the shape of the time–signal intensity curve were evaluated from before steroid administration to 1, 5, 10, and 14 days after administration. Determination of the Enhancement Ratio (ER)
Analysis of DCE-MRI The region of interest (ROI) was set between the proximal metaphysis and proximal diaphysis where is the common site of intra-femoral ON in rabbits (Fig. 1). The variations of arterial input function and the vascular structure of each subject can affect the results, which makes it essential to normalize the perfusion parameters in the ROI to those of a reference region. In light of this, the ROI was set so that as much bonemarrow space as possible between the proximal metaphysis and proximal diaphysis was included. The signal intensity (SI) in the ROI was measured.
SI reaches steady state 30 seconds after a contrast agent injection. The ER was obtained by dividing the mean RI(t) of 15 timepoints in “phase c” when steady state was achieved by the mean RI(t) of 10 timepoints in “phase a” before the contrast agent injection (pre-SI). Determination of the Initial Slope (IS) RI level started to rise statistically at t ¼ 40.8 (second image after contrast agent injection) compared with the average RI level during the “phase a” (pre-SI) (paired t-test, P < 0.01). While, the level at t ¼ 57.8 (seventh image after the injection) was significantly
Femoral Perfusion in Steroid-Induced ON
937
(M.I., M.S.). Both of them are orthopedists with more than 10 years of experience who are skilled in tissue evaluation regarding bone necrosis. Statistical Analysis
Figure 2. A representative example of a time–signal intensity curve. Blue and red lines signify the time–signal intensity curves for the right and left femora. RI was enhanced after bolus injection of a contrast agent at the time shown by the arrow. Forty sequential RIs were divided into three phases: (a) 10 measurements before contrast agent injection (in the yellow frame); (b) 15 measurements during the period when SI was enhanced just after the contrast agent injection (in the red frame); and (c) 15 measurements at the stage when RI reached steady state after the contrast agent injection (in the green frame).
higher than the level on the previous image (paired ttest, P < 0.01) but not significantly different from that on the next image (paired t-test, P ¼ 0.18). Thus, we considered that the contrast effect appeared sufficiently at t ¼ 57.8. The initial slope of RI was obtained by dividing the increment at t ¼ 57.8 from pre-SI with t ¼ 20.4, the time between when the first and seventh images were taken. The equation used was: Initial slope ¼ (RI(57.8)-1)/20.4. Calculation of the Area Under the Curve (AUC) The AUC was calculated by adding RI(t) of 15 timepoints in “phase b,” directly after the injection of the contrast agent, using the following equation: AUC ¼ P85 RIðT Þ. T ¼34 Histological Study Immediately after the MRI was obtained 14 days after steroid administration, the rabbits were sacrificed with an excessive dose of intravenously injected sodium pentobarbital. The bilateral femora were immediately harvested and divided at the middle to obtain the proximal halves. They were fixed in 10% buffered formalin for 1 week and then decalcified with 10% EDTA for 5 weeks. After decalcification, the femora were embedded in paraffin. Four-micrometer-thick coronal sections were prepared and stained with hematoxylin and eosin. ON was defined as the presence of empty lacunae or obvious pyknotic nuclei of osteocytes within the bone trabeculae, accompanied by necrosis of surrounding bone marrow hematopoietic cells, or the presence of diffuse necrosis of bone marrow hematopoietic cells and fat cells (16). ON and all other histologic measures were assessed blindly by two independent authors
In this study we evaluated perfusion, for which interindividual differences often occur (12,17). Therefore, we also evaluated interindividual differences using the Wilcoxon signed-rank test in statistical analysis. For the evaluation of blood flow in the ROI in the femur, each value of the ER, IS, and AUC was statistically compared between the value before steroid administration and the minimal value after steroid administration. SPSS Statistics (v. 21.0, IBM, Somers, NY) was used as the statistical analysis software. We considered P < 0.05 statistically significant. RESULTS DCE-MRI of the Femur Sequential Changes in the Shape of the Time–Signal Intensity Curves Sequential changes in the shape of the time–signal intensity curves of two rabbits that represent typical time course changes are shown in Figs. 3 and 4. In the first rabbit, RI levels after the contrast agent injection were lower on the 5th day after steroid administration compared with those before steroid administration. RI levels improved to the same levels as before steroid administration on the 10th and 14th days after steroid administration (Fig. 3A–E). In the second rabbit, RI levels were higher on the 1st day after steroid administration compared with those before administration in both femora, but especially in the left femur. RI levels decreased to the same levels as before steroid administration on the 5th day after administration, and became lower than the level before steroid administration on the 10th day. RI levels improved on the 14th day (Fig. 4A–E). As mentioned above, some rabbits showed changes in the time–signal intensity curves even from the 1st day after steroid administration, and all of the rabbits showed changes over time. Changes in ER of the Femora The ER increased in three out of 18 femora and decreased in all of the other 15 femora after steroid administration. Statistically, the ER changed after steroid administration compared with before administration, and the direction of change was toward decrease, which was significant (P < 0.05; Fig. 5A). The day when the ER of each femora showed the minimal value was the 1st day after steroid administration in five femora (28%), the 5th day in five femora (27%), the 10th day in three femora (17%), and the 14th day in two femora (11%). Change in the IS of the Femora The IS increased in two out of 18 femora and decreased in all of the other 16 femora after steroid
938
Hayashi et al.
Figure 3. A representative example of changes in the time–signal intensity curve over time. The time–signal intensity curve shows the following. A: before steroid administration; B: the 1st day after steroid administration; C: the 5th day after steroid administration; D: the 10th day after steroid administration; and E: the 14th day after steroid administration. The increase in RI after the contrast agent injection was slightly suppressed on the 1st day after steroid administration and was further suppressed on the 5th day compared with the curve before steroid administration. However, on the 10th and 14th days after steroid administration, RI after the contrast agent injection was increased to as high as that before steroid administration.
administration. Statistically, the IS changed after steroid administration compared with before administration, and the direction of change was toward decrease, which was significant (P < 0.01; Fig. 5B). The day when the IS of each femora showed the minimal value was the 1st day after steroid administration in four femora (22%), the 5th day in six femora (33%), the 10th day in two femora (11%), and the 14th day in four femora (22%). Change in the AUC of the Femora The AUC increased in three out of 18 femora, and decreased in all of the other 15 femora after steroid administration. Statistically, the AUC changed after steroid administration compared with before administration, and the direction of change was toward decrease, which was significant (P < 0.01; Fig. 5C). The day when the AUC of each femora showed the
minimal value was the 1st day after steroid administration in five femora (27%), the 5th day in five femora (27%), the 10th day in two femora (11%), and the 14th day in three femora (17%).
Comparison Between Histological Results and DCE-MRI Results One out of 18 femora in one of nine rabbits showed pyknotic nuclei of osteocytes within the bone trabeculae. This was accompanied by necrosis of surrounding bone marrow hematopoietic cells and fat cells, which was determined to be ON. The ER, IS, and AUC of the one femur that developed ON were decreased on the 5th day after steroid administration compared with those before steroid administration. The ER and IS showed the third lowest
Femoral Perfusion in Steroid-Induced ON
939
Figure 4. Changes in the time–signal intensity curve over time in another rabbit. The time signal intensity curve shows the following. A: before steroid administration; B: the 1st day after steroid administration; C: the 5th day after steroid administration; D: the 10th day after steroid administration; and E: the 14th day after steroid administration. On the 1st day after steroid administration, both of the femora, but especially the left femur, showed higher RIs after the injection of contrast agent compared with those before steroid administration. On the 5th day, RI of the left femur, which had been high, decreased to the level observed before the steroid administration. On the 10th day, RI after the contrast agent injection clearly decreased compared with that before steroid administration. On the 14th day, RI after the contrast agent injection was higher than that before steroid administration.
value among the 18 femora, and its AUC showed the second lowest value among the 18 femora. DISCUSSION DCE-MRI is often used for the evaluation of blood flow with MRI (12). Parameters used for this evaluation include the ER (6), IS (7), AUC (8), relative SI (9), time to peak enhancement (10), and washout slope (11). Because of the effect of interrater differences and motion artifacts, one parameter alone cannot precisely explain the phenomena. Therefore, evaluation of blood flow should be comprehensively made based on multiple parameters (12). In the current study, the IS and AUC were used to evaluate the change in SI from immediately after the contrast agent injection, and the
ER was used to evaluate the change in SI after it reached a plateau after the contrast agent injection (12,13). The major etiological factor for ON is considered to be ischemia, which occurs after steroid administration (1). However, the presence of ischemia in ON in vivo has not been proven. In our study, all of the three parameters (IS, ER, and AUC) were significantly decreased at the common development site of ON of the femoral head. Therefore, judging these three parameters comprehensively, we considered that blood flow was actually decreasing after steroid administration. Based on clinical cases, the timing when ischemia occurs after steroid administration had been considered to be 2 weeks after administration (2,3). However, recent studies have shown that VEGF expression levels increase on the 3rd day after steroid administration (4)
940
Hayashi et al.
uated indirectly through the confirmation of increased VEGF levels, which increase after the occurrence of ischemia, by sacrificing animals. This study is important from two perspectives. First, perfusion was able to be evaluated in vivo over time, and second, we showed that blood flow starts to decrease in the early phase after steroid administration. The limitation of this study is that the number of femurs that developed ON was small, 1 out of 18 femora, which was not sufficient to discuss any possible relationship between ON development in the tissue and the perfusion status observed in this study. In conclusion, we assessed perfusion in the femora during early days after steroid administration and could confirm that perfusion actually decreases soon after steroid administration. REFERENCES
Figure 5. Changes in the ER (graph A), IS (graph B), and AUC (graph C). “before steroid administration” are values before the steroid administration and “after steroid administration” are minimal values after the steroid administration. ER, IS, and AUC values after MPSL administration significantly decreased (P < 0.05, P < 0.01, and P < 0.01, respectively).
and that oxidative damage of DNA occurs by the 5th day after steroid administration (5). Therefore, we considered that the change in perfusion might occur by the 5th day after steroid administration, and focused on this time window for evaluation. We found that blood flow decreased from the 1st day after steroid administration in some femora. We showed that blood flow actually decreases at the time when VEGF levels increase and oxidative damage of DNA occurs. In our study we observed decreased blood flow in the bone marrow of femora after steroid administration in vivo. We also showed that blood flow decreased from the early phase, by observing decreased blood flow in some femora from the 1st day after steroid administration. Without MRI, perfusion cannot be evaluated over time in vivo. Ischemia can only be eval-
1. Kerachian MA, Harvey EJ, Cournoyer D, Chow TY, S eguin C. Avascular necrosis of the femoral head: vascular hypotheses. Endothelium 2006;13:237–244. 2. Kubo T, Yamazoe S, Sugano N, et al. Initial MRI findings of nontraumatic osteonecrosis of the femoral head in renal allograft recipients. Magn Reson Imaging 1997;15:1017–1023. 3. Nakamura T, Matsumoto T, Nishino M, Tomita K, Kadoya M. Early magnetic resonance imaging and histologic findings in a model of femoral head necrosis. Clin Orthop Relat Res 1997; 68–72. 4. Kabata T, Matsumoto T, Yagishita S, Wakayama T, Iseki S, Tomita K. Vascular endothelial growth factor in rabbits during development of corticosteroid-induced osteonecrosis: a controlled experiment. J Rheumatol 2008;35:2383–2390. 5. Ichiseki T, Ueda Y, Katsuda S, Kitamura K, Kaneuji A, Matsumoto T. Oxidative stress by glutathione depletion induces osteonecrosis in rats. Rheumatology (Oxford) 2006;45:287–290. 6. Mussurakis S, Gibbs P, Horsman A. Primary breast abnormalities: selective pixel sampling on dynamic gadolinium-enhanced MR images. Radiology 1998;206:465–473. 7. Karahaliou A, Vassiou K, Arikidis NS, Skiadopoulos S, Kanavou T, Costaridou L. Assessing heterogeneity of lesion enhancement kinetics in dynamic contrast-enhanced MRI for breast cancer diagnosis. Br J Radiol 2010;83:296–309. 8. Evelhoch JL. Key factors in the acquisition of contrast kinetic data for oncology. J Magn Reson Imaging 1999;10:254–259. 9. Johansen R, Jensen LR, Rydland J, et al. Predicting survival and early clinical response to primary chemotherapy for patients with locally advanced breast cancer using DCE-MRI. J Magn Reson Imaging 2009;29:1300–1307. 10. Liney GP, Gibbs P, Hayes C, Leach MO, Turnbull LW. Dynamic contrast-enhanced MRI in the differentiation of breast tumors: user-defined versus semi-automated region-of-interest analysis. J Magn Reson Imaging 1999;10:945–949. 11. Jia G, O’Dell C, Heverhagen JT, et al. Colorectal liver metastases: contrast agent diffusion coefficient for quantification of contrast enhancement heterogeneity at MR imaging. Radiology 2008;248:901–909. 12. Yang X, Knopp MV. Quantifying tumor vascular heterogeneity with dynamic contrast-enhanced magnetic resonance imaging: a review. J Biomed Biotechnol 2011;2011:732848. 13. Kiessling F, Lichy M, Grobholz R, et al. Simple models improve the discrimination of prostate cancers from the peripheral gland by T1-weighted dynamic MRI. Eur Radiol 2004;14:1793–1801. 14. Yamamoto T, Irisa T, Sugioka Y, Sueishi K. Effects of pulse methylprednisolone on bone and marrow tissues: corticosteroid-induced osteonecrosis in rabbits. Arthritis Rheum 1997;40:2055–2064. 15. Kuribayashi M, Fujioka M, Takahashi KA, et al. Vitamin E prevents steroid-induced osteonecrosis in rabbits. Acta Orthop 2010;81:154–160. 16. Yamamoto T, Hirano K, Tsutsui H, Sugioka Y, Sueishi K. Corticosteroid enhances the experimental induction of osteonecrosis in rabbits with Shwartzman reaction. Clin Orthop Relat Res 1995; 235–243. 17. Byun J, Lai D, Luo S, Risser J, Tung B, Hardy RJ. A hybrid method in combining treatment effects from matched and unmatched studies. Stat Med 2013;32:4924–4937.
