- Email: [email protected]
Influence of saline and glucose molecules to contrast properties of clinically used MRI contrast agents
Measurement 69 (2015) 109–114
Contents lists available at ScienceDirect
Measurement journal homepage: www.elsevier.com/locate/measurement
Influence of saline and glucose molecules to contrast properties of clinically used MRI contrast agents Oliver Strbak ⇑, Marta Masarova, Daniel Gogola, Pavol Szomolanyi, Ivan Frollo Institute of Measurement Science, SAS, Dubravska cesta 9, 84104 Bratislava, Slovakia
a r t i c l e
i n f o
Article history: Received 18 August 2014 Received in revised form 15 February 2015 Accepted 20 March 2015 Available online 26 March 2015 Keywords: Contrast agents MRI Contrast change Relaxation Saline Glucose
a b s t r a c t Contrast agents (CA) are usually used in clinical practice for contrast enhancement during Magnetic Resonance Imaging (MRI). They are iron oxides or gadolinium-based nanoparticles in a carrier fluid. We studied whether different concentration levels of the saline and glucose molecules are able to change the relaxation properties of both types of contrast agents during MRI. The reason is that they are essential biological molecules, and their modified concentration levels are accompanied with several pathological processes. We have found that the physiological concentration of saline and glucose molecules influences the CA contrast properties selectively on the basis of CA concentration (up to 30% for iron oxide CA, and 15% for gadolinium CA). Moreover, the altered concentration levels of saline and glucose change the signal intensity (contrast), for one selected pulse sequence and CA concentration, in range of 2–17%. Although, such contrast changes are on the visibility limit to the naked eye for our system (0.178 T), they can be clearly visible with high-field system, and can have influence to the next data analysis, e.g. relaxation times calculation. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Currently, Magnetic Resonance Imaging (MRI) is a technique routinely used in clinical practice. MRI provides an anatomical picture of tissue based on different intrinsic contrast. Contrast in MRI arises from the difference in signal intensity, which can be modified by intrinsic parameters (spin density, relaxation times T1 and T2) and by the pulse sequences parameters (repetition time TR, echo time TE). Modification of TR and TE results in T1, T2 or proton density weighted imaging. For contrast enhancement in clinical practice the paramagnetic contrast agents (CAs) are often used. Signal enhancement is caused by coupling of proton magnetic moments with larger magnetic moments of paramagnetic nanoparticles [1].
⇑ Corresponding author. Tel.: +421 2 59104546; fax: +421 2 54775943. E-mail address: [email protected] (O. Strbak). http://dx.doi.org/10.1016/j.measurement.2015.03.036 0263-2241/Ó 2015 Elsevier Ltd. All rights reserved.
CAs can be either iron oxide or gadolinium based nanoparticles. Generally, iron oxide CAs have a small effect on T1, but strongly affect T2. Due to this T2 shortening they produce hypointensive artefacts in T2-weighted images. On the other hand, gadolinium CA mainly affects T1 relaxation time constant and produce hyperintensive artefacts in T1 weighted images. Iron oxide CAs are especially used in imaging of liver pathology and include two clinically approved superparamagnetic (SPIO) agents: Endorem (ferumoxides with particle size of 120–180 nm) and Resovist (ferucarbotran with particle size of about 60 nm) [2]. They are absorbed by Kupffer cells which are present only in normal reticuloendothelial system (RES) but not in lesions. This results in different visibility of normal tissue and lesions. Gadolinium-based CAs are routinely used in all other clinical MRI applications, including tumour and vascular imaging [3]. In this study we have focused on MRI contrast properties of both types of contrast agents – SPIO and gadolinium
110
O. Strbak et al. / Measurement 69 (2015) 109–114
based CA, in the presence of saline and glucose molecules. As the SPIO contrast agent model we have chosen the Resovist (Bayer Schering Pharma AG), and as the gadolinium one, the MultiHance (ALTANA Pharma AG). We have chosen saline and glucose molecules because they are essential components of the human body, and their altered concentration levels are usually accompanied with serious pathological processes, which can be studied with MRI. The modified concentration levels of the glucose are usually accompanied with different pathological processes, including diabetes, cardiovascular disease of acute stroke risk, etc. [4,5]. The saline concentration imbalance can be caused by diseases of kidney, pituitary gland, hypothalamus or by hypernatremia in critically ill patients [6,7]. Their concentration variations are also present in other types of pathology or in life stress conditions, like diet [8]. The main goal is to find out if the altered concentration of saline and glucose molecules can have an influence to the clinical agents’ contrast properties, with using the low-field MRI techniques. We followed two main points: (i) if the influence is large enough to be visible to the naked eye, which would have a direct impact to the clinical practice; (ii) if there exists a ‘‘pattern of action’’ for the saline and glucose molecules in relation to contrast properties of CAs. 2. Material and methods 2.1. Contrast agents As an iron oxide CA model system we chose the Resovist (Bayer Schering Pharma AG), which consists of carboxydextran coated iron oxide nanoparticles. The gadolinium CA was represented by the MultiHance (ALTANA Pharma AG), which consists of gadobenate dimeglumine molecules. In the first series we investigated eight different concentration of CAs in various pools – distilled water, saline, glucose, and saline + glucose. Saline and glucose were in physiological concentration (saline 9 g/l, glucose 1 g/l). The concentration values of the CAs was chosen to copy the real possible concentrations in the blood stream, and simultaneously they had to fall between the experimentally determined limit values with a maximum contrast change for our system. For Resovist: 26.6, 52.7, 78.2, 103.3, 127.9, 152.1, 175.8, and 199 lg/ml. For MultiHance: 0.249, 0.494, 0.734, 0.969, 1.2, 1.426, 1.648, and 1.866 mg/ml. The aim was to find out how the physiological concentration of saline and glucose influence the contrast properties of CAs with different concentrations. It simulates the real changes of CA concentrations during in-vivo application.
Secondly, only one concentration of the CAs was chosen which was similar to recommended clinical in-vivo administration of CA: Resovist – 103.3 lg/ml of the iron oxide nanoparticles, MultiHance – 1.3 mg/ml of the gadolinium particles. Four different concentrations (around the physiological concentration) of the saline and glucose have been investigated: saline – 4.5, 9, 13.5, 18 g/l, and glucose – 0.5, 1, 1.5, 2 g/l. Here, the aim was to find out whether different concentration of saline and glucose has influence to the contrast properties of one selected concentration of CA. It simulates the use of CAs in pathological processes/diseases with altered concentrations of saline and glucose molecules [4–8]. 2.2. MRI experiments The MRI experiments were performed with the ESAOTE Opera (E-SCANTM XQ) 0.178 T system. Images were acquired using both Spin Echo (SE) and Gradient Echo (GE) pulse sequences in T1 and T2 weighted modes. In the first step we had to find the most suitable pulse sequence and its parameters for imaging the selected CAs. The selection of pulse sequence was based on three aspects: (i) artefacts formation, (ii) signal/noise ratio, and (iii) the biggest contrast change. In the last one we checked the shape of the curve of the contrast medium concentration gradient in distilled water, where we followed the difference between the reference (without particles) and the sample with the highest concentration of particles. Finally the pulse sequence with the highest change in signal intensity (and thus in contrast), without serious artefacts, and with good signal to noise ratio was chosen. On the basis of these three aspects, we selected the most suitable pulse sequence. We chose four sequences on the short-list: Spin Echo TR = 600 ms, TE = 26 ms, Spin Echo TR = 1800 ms, TE = 26 ms, Turbo Spin Echo TR = 3000 ms, TE = 80 ms, Turbo Spin Echo TR = 3000 ms, TE = 120 ms (Table 1). Subsequently, the most appropriate protocol for contrast imaging was determined. For Resovist we found that this was a T2-weighted Turbo Spin Echo (TSE) sequence with repetition time TR = 3000 ms and echo time TE = 120 ms. For MultiHance it was a T1-weighted SE sequence with repetition time TR = 600 ms and echo time TE = 26 ms. The GE sequences showed large artefacts in both types of contrast agents and selected concentrations. This was probably caused by the high sensitivity of the GE sequences to the field inhomogeneities, caused by the magnetic particles which are present in both types of contrast media. The naked-eye visibility limit for our system is
Table 1 Intensity changes for selected CA and pulse sequence from the short-list. The maximum change indicates the best suitability of pulse sequence. Pulse sequence
Resovist (difference in signal intensity for the reference sample – 0, and sample with the maximum CA concentration – 8)
MultiHance (difference in signal intensity for the reference sample – 0, and sample with the maximum CA concentration – 8)
SE T1 TR600 TE26 SE T1 TR1800 TE26 TSE T2 TR3000 TE80 TSE T2 TR3000 TE120
557.3 5.5 1526.1 1695.1
1068.5 604.6 238.3 634.4
The bold numbers defines the highest contrast change and thus the best suitability of a pulse sequence for imaging of selected contrast agent.
O. Strbak et al. / Measurement 69 (2015) 109–114
111
15%, when comparing contrast of the two points (areas). During study we focused on tracking of the ‘‘contrast change’’ value, which is the relative signal intensity of sample in comparison with reference (distilled water, and/or distilled water with CAs). For image data processing and basic analysis we used the following software tools: Marevisi (NRC – Institute for Biodiagnostic, Winnipeg, Canada), and Matlab R2011b (Mathworks Inc., Natic, USA). 3. Results and discussion In Figs. 1 and 2 MRI images of the selected contrast agents, together with the shape of the concentration curve, are shown in comparison with the signal intensity. Fig. 1a shows the change in the contrast of the Resovist contrast agent in comparison with the concentration increase. The appropriate concentration curve is in Fig. 1b, where we observe a decrease of signal with the concentration increase of iron oxide nanoparticles. The change is due to the influence of magnetite nanoparticles in Resovist when imaged with T2-weighted sequences. The same situation for MultiHance is in Fig. 2a and b. In this case we observed an increase in signal intensity when imaged with T1weighted sequences, caused by gadolinium particles. For both cases, we observe the expected shape of the curve of signal intensity in comparison with an increase of the contrast agent concentration: decrease for Resovist and increase for MultiHance.
Fig. 2. (a) MRI of the different concentration of MultiHance in distilled water, (b) MultiHance contrast change curve depending on gadolinium concentration.
In the first series, we investigated the effect of the physiological concentration of saline and glucose molecules on the contrast properties of selected contrast agents, which simulates the concentration changes of the contrast agent in the blood stream during in-vivo administration. For each contrast agent, we prepared four solutions with increasing concentration of magnetic particles: distilled water, saline, glucose, saline + glucose. The solutions were imaged with the selected pulse sequence.
Fig. 3 shows the intensity change of different concentrations of Resovist in comparison with reference (distilled water). We can see that a physiological concentration of saline and glucose molecules selectively influence the contrast properties of the CA depending on its concentration. These results were at variance with our expectations, where we assumed the same change in intensity value for all concentrations of contrast media. Saline molecules cause an extreme increase in contrast – up to 30% for samples with a higher concentration of contrast agent (above 130 lg/ml, recommended concentration for in-vivo administration is 97 lg/ml). The smallest change in contrast we observed for samples with the presence of glucose molecules (up to 20%). For a mixture of saline and glucose molecules the maximum change is about 25%. What is interesting is that we do not observe these maximum changes in the same sample, but always for a different concentration of CA: 7. for saline, 3. for glucose and 4. for a mixture of saline and glucose. In addition, for saline sample this change is positive, while for glucose
Fig. 1. (a) MRI of the different concentration of Resovist in distilled water, (b) Resovist contrast change curve depending on iron oxide concentration.
Fig. 3. Resovist contrast change curve in different pools in comparison with water.
3.1. CA vs. physiological concentration of MOI
112
O. Strbak et al. / Measurement 69 (2015) 109–114
Fig. 4. MultiHance contrast change curve in different pools in comparison with water.
Fig. 6. MultiHance contrast change values cleared from the MOI-water influence.
and a saline and glucose mixture the change is negative, relative to the reference. For application purposes are interesting the changes in the 4. sample, which contains the recommended concentration of contrast agent for invivo administration. However, we found that the mere presence of saline and glucose molecules in the absence of contrast agent, change the contrast: for saline 4.1%, for glucose 8.8%, and for a mixture of saline and glucose up to the 14.9% in comparison with reference. This is almost a naked-eye visibility limit for our system. Therefore, in Fig. 5 we show the impact of saline and glucose molecules deprived from their effect to the distilled water. In the figure we can see that only three columns reach the threshold of visibility to the naked eye, and only for saline samples. The CA concentrations of these samples are over the recommended in-vivo concentration, and therefore should not affect the situation in clinical practice. The contrast change in sample No. 4, which contains the recommended in-vivo concentration,
is for studied MOI, below the limit of naked-eye visibility. The largest change is for the mixture of saline and glucose ( 9%), then the glucose alone (3.4%), and finally for the saline sample (less than 1%). For the gadolinium-based contrast agent, these changes are even less visible. Here, we can also see the effect of saline and glucose molecules themselves on the contrast change of distilled water, although in different proportions in comparison with the first sample (compare first points in Figs. 3 and 4): saline 15% (before 4.1%), glucose 12.3% (before 8.8%), and the saline and glucose mixture only 1.4% (before 14.9%). From this point it seems like the ‘random’ effect of these molecules on the contrast change of distilled water. This effect is further reflected in samples with CAs. Fig. 6 shows the influence of the physiological concentration of MOI to the contrast properties of the contrast medium MultiHance, cleared from the effect of MOI. In this case, all contrast changes are below naked-eye visibility. It seems like the glucose decreased the effect of the saline, which is also the most visible in this case. For sample No. 5, which contains the recommended in-vivo concentration of MultiHance contrast agent, the impact on the contrast is as follows: glucose 8.9%, saline 11.2%, and saline with glucose only 0.5%. These changes do not affect the visibility to the naked eye for our system. 3.2. In-vivo concentration of CA in various MOI In the second case, we chose one concentration of the CA which is similar to the clinical in-vivo concentration, Table 2 Comparison of the relative contrast changes of the physiological concentration of MOI and CA acquired from both series.
Fig. 5. Resovist contrast change values cleared from the MOI-water influence.
CA with physiological concentration of MOI
Contrast change in 1st series (%)
Contrast change in 2nd series (%)
Resovist – saline Resovist – glucose MultiHance – saline MultiHance – glucose
0.6 3.4 11.2 8.8
4.1 2.1 5.0 12.4
113
O. Strbak et al. / Measurement 69 (2015) 109–114
Fig. 7. (a) Resovist and MultiHance contrast change values in comparison with reference for one CA concentration (Resovist – 103.3 lg/ml of the iron oxide nanoparticles, MultiHance – 1.3 mg/ml of the gadolinium particles). Samples: saline S1-4 (4.5, 9, 13.5, 18 g/l), glucose G1-4 (0.5, 1, 1.5, 2 g/l). (b) MRI of Resovist. (c) MRI of MultiHance.
and we followed the four different concentrations of MOI, around their physiological concentration. Such an arrangement simulated a concentration change of biological molecules in various pathological processes which may accompany an MRI examination. Saline concentration: S1 = 4.5 g/l, S2 = 9 g/l, S3 = 13.5 g/l, S4 = 18 g/l, glucose concentration: G1 = 0.5 g/l, G2 = 1 g/l, G3 = 1.5 g/l, G4 = 2 g/l. Changes visible to the naked eye occurred in only two samples: S1 16.6% for Resovist, and G1 16.1% for MultiHance (see Fig. 7). Changes also have the character of randomness and do not fit the concentration changes of MOI in this case. Almost the same change in contrast can be seen when samples S2 and G2 are compared from the second series, which include the physiological concentration of MOI, with the data from the first series (sample No. 4 for Resovist, and sample No. 5 for MultiHance), which also contain physiological concentrations of MOI (Table 2). From this point of view, it formerly excludes the random effect of influence, as discussed above. These results showed that the effect of selected MOI on the MRI contrast change exists and is not negligible. However, sometimes it rather appears at random (influence to distilled water), but in general the same ‘‘pattern’’ of influence can be seen (Tables 3 and 4). In case of Resovist it is a decrease of signal (Fig. 3) for concentrations up to 103.3 lg/ml (which is recommended in-vivo concentration), followed by a steep increase for all samples. In case of MultiHance we can see a decrease of signal (Fig. 4) for
Table 3 ‘‘Pattern of action’’ of the saline and glucose physiological concentration to the CA contrast properties. CA
Saline (9 g/l)
Glucose (1 g/l)
Saline (9 g/l) + Glucose (1 g/l)
Resovist
Decrease/ increase Mostly decrease
Decrease/ increase Mostly decrease
Decrease/ increase Mostly decrease
MultiHance
Table 4 ‘‘Pattern of action’’ of the increasing concentration of saline and glucose to the CA contrast properties. CA
Saline concentration increase (4.5–18 g/l)
Glucose concentration increase (0.5–2 g/l)
Resovist (103.3 lg/ml) MultiHance (1.3 mg/ml)
Full decrease
Mostly increase
Mostly decrease
Full decrease
whole selected concentration range. The data can be such transform to the pattern information as ‘‘decreasing/increasing’’ trend for Resovist (with the minimum around 100 lg/ml), and ‘‘decreasing’’ trend for MultiHance (Table 3). Such a selective influence of physiological concentration of molecules to the contrast properties of CAs depending on their concentration is the most surprising finding for us. In case of increasing concentration of saline we observed the ‘‘decreasing’’ trend for both CAs, while in case of glucose we found the ‘‘increasing’’ character for Resovist and ‘‘decreasing’’ trend for MultiHance (Table 4). The fact is that the changes are mostly below or at the limit of visibility to the naked eye for our system. However, without taking into consideration the question of randomness or pattern of action, the changes are large enough to be able to affect the calculation of parameters such as relaxation times. Moreover, these changes should be more visible with the high-field systems, which are more often used in clinical practice. Therefore, this MOI influence on the contrast of water molecules and CAs should not be forgotten and should be subjected to more precise study.
4. Conclusion The aim of the study was to determine the effect of biological MOI on the contrast properties of clinically used CAs for MRI. The study was undertaken due to the fact that
114
O. Strbak et al. / Measurement 69 (2015) 109–114
while there is routine use of CAs in clinical practice, the information on the influence of the in-vivo environment on the contrast properties is lacking or insufficient. Various physiological and especially pathological processes, accompanied by changes in concentration levels of biologically active molecules in the body, may affect with their properties the MRI signal and thereby distort the output data. Since saline and glucose molecules are well represented in the human body, we focused our attention on them. We followed their impact on the contrasting properties of clinically used contrast agents, both containing iron oxide nanoparticles as well as gadolinium particles. The planned outputs should be the information about the changes in MRI contrast, influenced by these molecules, which could have an impact on diagnostic conclusions in clinical practice. The images were acquired using both Spin Echo (SE) and Gradient Echo (GE) pulse sequences in T1- and T2weighted modes. For the purpose of data processing, the most suitable sequence was determined. For Resovist it was a T2-weighted Turbo Spin Echo sequence with a repetition time of TR = 3000 ms, and Echo Time TE = 120 ms. For MultiHance we used a T1-weighted Spin Echo sequence with a repetition time of TR = 600 ms, and an echo time of TE = 26 ms. The experiments were divided into two parts. At first, we were looking for the impact of physiological concentrations of selected MOI to the contrast properties of the studied contrast agents – Resovist and MultiHance in different concentrations. We were interested whether the physiological concentration of saline and glucose molecules has the same effect on the contrast medium, which simulates in-vivo conditions after the administration of the CA into the bloodstream. We found that contrast change for Resovist is up to 30. 6% compared with the reference. The largest changes in contrast were observed in the samples containing salt molecules. With the MultiHance CA the biggest changes of the contrast range are about 15% compared with the reference. It has been shown that the mere presence of saline and glucose molecules in distilled water change the contrast level by up to 15% in comparison with the reference for both contrast media. This is on the border for the nakedeye visibility changes for our system. The resulting change in the contrast deprived of the influence of MOI themselves to the distilled water for ranges up to 26.5% for Resovist in saline solution. For the MultiHance CA, the biggest intensity change is around 13%, as for the saline solution. In the second part, we investigated the effect of varying concentrations of the MOI on the contrast properties of the recommended in-vivo concentration of CA, which simulates the change in the concentration of biologically active
molecules in various pathological processes. It has been shown that the presence of MOI at various concentrations also affects the contrast of iron oxide-based Resovist, and the biggest change in the contrast is as high as 16.6% in the presence of saline molecules. For the MultiHance contrast agent, the biggest change reaches 16.1%. Although these changes are in most cases below or around the limit visibility to the naked eye for our system (15%), its height is not negligible and may affect the further analysis of data, e.g. calculation of the relaxation times. Thus, it may alter the final clinical conclusions. Moreover we found the non-random pattern of action of the saline and glucose molecules to the contrast properties of the selected CAs. In case of physiological concentration of saline and glucose we surprisingly found the selective influence to different concentrations of CA. Trends were as follows: ‘‘decreasing/increasing’’ for Resovist (with the minimum around 100 lg/ml), and ‘‘decreasing’’ for MultiHance. In case of increasing concentration of saline and glucose, the trend was ‘‘decreasing’’ for saline in both CAs, and ‘‘increasing’’ for Resovist and decreasing for MultiHance for glucose. Acknowledgements This work was supported within the project of the Slovak Research and Development Agency Nr. APVV0431-12, and Slovak Scientific Grant Agency VEGA 2/ 0013/14. References [1] P. Gillis, F. Moiny, R.A. Brooks, On T2 shortening by strongly magnetizes spheres: a partial refocusing model, Magn. Reson. Med. 47 (2002) 257–263, http://dx.doi.org/10.1002/mrm.10059. [2] Y.X.J. Wang, Superparamagnetic iron oxide based MRI contrast agents: current status of clinical application, Quant. Imaging. Med. Surg. 1 (2011) 35–40, http://dx.doi.org/10.3978/j.issn.2223-4292. 2011.08.03. [3] P. Caravan, J.J. Ellison, T.J. McMurry, R.B. Lauffer, Gadolinium(III) chelates as MRI contrast agents: structure, dynamics, and applications, Chem. Rev. 99 (1999) 2293–2352, http://dx.doi.org/ 10.1021/cr980440x. [4] S.S. Anand, G.R. Dagenais, V. Mohan, et al., Glucose levels are associated with cardiovascular disease and death in an international cohort of normal glycaemic and dysglycaemic men and women: the EpiDREAM cohort study, Eur. J. Prev. Cardiol. 19 (2012) 755–764, http://dx.doi.org/10.1177/1741826711409327. [5] K. Matz, K. Keresztes, C. Tatschl, et al., Disorders of glucose metabolism in acute stroke patients, Diabetes Care 29 (2006) 792– 797, http://dx.doi.org/10.2337/diacare.29.04.06.dc05-1818. [6] C.S. Lin, C.J. Cheng, K.C. Shih, S.H. Lin, Recurrent hyponatremia in a patient with chronic kidney disease, J. Nephrol. 19 (2006) 394–398. [7] G. Lindner, G.C. Funk, Hypernatremia in critically ill patients, J. Crit. Care 28 (2013) 216.e11, http://dx.doi.org/10.1016/j.jcrc.2012.05.001. [8] G. Riccardi, A.A. Rivellese, Effects of dietary fiber and carbohydrate on glucose and lipoprotein metabolism in diabetic patients, Diabetes Care 14 (1991) 1115–1125, http://dx.doi.org/10.2337/diacare.14. 12.1115.
Contents lists available at ScienceDirect
Measurement journal homepage: www.elsevier.com/locate/measurement
Influence of saline and glucose molecules to contrast properties of clinically used MRI contrast agents Oliver Strbak ⇑, Marta Masarova, Daniel Gogola, Pavol Szomolanyi, Ivan Frollo Institute of Measurement Science, SAS, Dubravska cesta 9, 84104 Bratislava, Slovakia
a r t i c l e
i n f o
Article history: Received 18 August 2014 Received in revised form 15 February 2015 Accepted 20 March 2015 Available online 26 March 2015 Keywords: Contrast agents MRI Contrast change Relaxation Saline Glucose
a b s t r a c t Contrast agents (CA) are usually used in clinical practice for contrast enhancement during Magnetic Resonance Imaging (MRI). They are iron oxides or gadolinium-based nanoparticles in a carrier fluid. We studied whether different concentration levels of the saline and glucose molecules are able to change the relaxation properties of both types of contrast agents during MRI. The reason is that they are essential biological molecules, and their modified concentration levels are accompanied with several pathological processes. We have found that the physiological concentration of saline and glucose molecules influences the CA contrast properties selectively on the basis of CA concentration (up to 30% for iron oxide CA, and 15% for gadolinium CA). Moreover, the altered concentration levels of saline and glucose change the signal intensity (contrast), for one selected pulse sequence and CA concentration, in range of 2–17%. Although, such contrast changes are on the visibility limit to the naked eye for our system (0.178 T), they can be clearly visible with high-field system, and can have influence to the next data analysis, e.g. relaxation times calculation. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Currently, Magnetic Resonance Imaging (MRI) is a technique routinely used in clinical practice. MRI provides an anatomical picture of tissue based on different intrinsic contrast. Contrast in MRI arises from the difference in signal intensity, which can be modified by intrinsic parameters (spin density, relaxation times T1 and T2) and by the pulse sequences parameters (repetition time TR, echo time TE). Modification of TR and TE results in T1, T2 or proton density weighted imaging. For contrast enhancement in clinical practice the paramagnetic contrast agents (CAs) are often used. Signal enhancement is caused by coupling of proton magnetic moments with larger magnetic moments of paramagnetic nanoparticles [1].
⇑ Corresponding author. Tel.: +421 2 59104546; fax: +421 2 54775943. E-mail address: [email protected] (O. Strbak). http://dx.doi.org/10.1016/j.measurement.2015.03.036 0263-2241/Ó 2015 Elsevier Ltd. All rights reserved.
CAs can be either iron oxide or gadolinium based nanoparticles. Generally, iron oxide CAs have a small effect on T1, but strongly affect T2. Due to this T2 shortening they produce hypointensive artefacts in T2-weighted images. On the other hand, gadolinium CA mainly affects T1 relaxation time constant and produce hyperintensive artefacts in T1 weighted images. Iron oxide CAs are especially used in imaging of liver pathology and include two clinically approved superparamagnetic (SPIO) agents: Endorem (ferumoxides with particle size of 120–180 nm) and Resovist (ferucarbotran with particle size of about 60 nm) [2]. They are absorbed by Kupffer cells which are present only in normal reticuloendothelial system (RES) but not in lesions. This results in different visibility of normal tissue and lesions. Gadolinium-based CAs are routinely used in all other clinical MRI applications, including tumour and vascular imaging [3]. In this study we have focused on MRI contrast properties of both types of contrast agents – SPIO and gadolinium
110
O. Strbak et al. / Measurement 69 (2015) 109–114
based CA, in the presence of saline and glucose molecules. As the SPIO contrast agent model we have chosen the Resovist (Bayer Schering Pharma AG), and as the gadolinium one, the MultiHance (ALTANA Pharma AG). We have chosen saline and glucose molecules because they are essential components of the human body, and their altered concentration levels are usually accompanied with serious pathological processes, which can be studied with MRI. The modified concentration levels of the glucose are usually accompanied with different pathological processes, including diabetes, cardiovascular disease of acute stroke risk, etc. [4,5]. The saline concentration imbalance can be caused by diseases of kidney, pituitary gland, hypothalamus or by hypernatremia in critically ill patients [6,7]. Their concentration variations are also present in other types of pathology or in life stress conditions, like diet [8]. The main goal is to find out if the altered concentration of saline and glucose molecules can have an influence to the clinical agents’ contrast properties, with using the low-field MRI techniques. We followed two main points: (i) if the influence is large enough to be visible to the naked eye, which would have a direct impact to the clinical practice; (ii) if there exists a ‘‘pattern of action’’ for the saline and glucose molecules in relation to contrast properties of CAs. 2. Material and methods 2.1. Contrast agents As an iron oxide CA model system we chose the Resovist (Bayer Schering Pharma AG), which consists of carboxydextran coated iron oxide nanoparticles. The gadolinium CA was represented by the MultiHance (ALTANA Pharma AG), which consists of gadobenate dimeglumine molecules. In the first series we investigated eight different concentration of CAs in various pools – distilled water, saline, glucose, and saline + glucose. Saline and glucose were in physiological concentration (saline 9 g/l, glucose 1 g/l). The concentration values of the CAs was chosen to copy the real possible concentrations in the blood stream, and simultaneously they had to fall between the experimentally determined limit values with a maximum contrast change for our system. For Resovist: 26.6, 52.7, 78.2, 103.3, 127.9, 152.1, 175.8, and 199 lg/ml. For MultiHance: 0.249, 0.494, 0.734, 0.969, 1.2, 1.426, 1.648, and 1.866 mg/ml. The aim was to find out how the physiological concentration of saline and glucose influence the contrast properties of CAs with different concentrations. It simulates the real changes of CA concentrations during in-vivo application.
Secondly, only one concentration of the CAs was chosen which was similar to recommended clinical in-vivo administration of CA: Resovist – 103.3 lg/ml of the iron oxide nanoparticles, MultiHance – 1.3 mg/ml of the gadolinium particles. Four different concentrations (around the physiological concentration) of the saline and glucose have been investigated: saline – 4.5, 9, 13.5, 18 g/l, and glucose – 0.5, 1, 1.5, 2 g/l. Here, the aim was to find out whether different concentration of saline and glucose has influence to the contrast properties of one selected concentration of CA. It simulates the use of CAs in pathological processes/diseases with altered concentrations of saline and glucose molecules [4–8]. 2.2. MRI experiments The MRI experiments were performed with the ESAOTE Opera (E-SCANTM XQ) 0.178 T system. Images were acquired using both Spin Echo (SE) and Gradient Echo (GE) pulse sequences in T1 and T2 weighted modes. In the first step we had to find the most suitable pulse sequence and its parameters for imaging the selected CAs. The selection of pulse sequence was based on three aspects: (i) artefacts formation, (ii) signal/noise ratio, and (iii) the biggest contrast change. In the last one we checked the shape of the curve of the contrast medium concentration gradient in distilled water, where we followed the difference between the reference (without particles) and the sample with the highest concentration of particles. Finally the pulse sequence with the highest change in signal intensity (and thus in contrast), without serious artefacts, and with good signal to noise ratio was chosen. On the basis of these three aspects, we selected the most suitable pulse sequence. We chose four sequences on the short-list: Spin Echo TR = 600 ms, TE = 26 ms, Spin Echo TR = 1800 ms, TE = 26 ms, Turbo Spin Echo TR = 3000 ms, TE = 80 ms, Turbo Spin Echo TR = 3000 ms, TE = 120 ms (Table 1). Subsequently, the most appropriate protocol for contrast imaging was determined. For Resovist we found that this was a T2-weighted Turbo Spin Echo (TSE) sequence with repetition time TR = 3000 ms and echo time TE = 120 ms. For MultiHance it was a T1-weighted SE sequence with repetition time TR = 600 ms and echo time TE = 26 ms. The GE sequences showed large artefacts in both types of contrast agents and selected concentrations. This was probably caused by the high sensitivity of the GE sequences to the field inhomogeneities, caused by the magnetic particles which are present in both types of contrast media. The naked-eye visibility limit for our system is
Table 1 Intensity changes for selected CA and pulse sequence from the short-list. The maximum change indicates the best suitability of pulse sequence. Pulse sequence
Resovist (difference in signal intensity for the reference sample – 0, and sample with the maximum CA concentration – 8)
MultiHance (difference in signal intensity for the reference sample – 0, and sample with the maximum CA concentration – 8)
SE T1 TR600 TE26 SE T1 TR1800 TE26 TSE T2 TR3000 TE80 TSE T2 TR3000 TE120
557.3 5.5 1526.1 1695.1
1068.5 604.6 238.3 634.4
The bold numbers defines the highest contrast change and thus the best suitability of a pulse sequence for imaging of selected contrast agent.
O. Strbak et al. / Measurement 69 (2015) 109–114
111
15%, when comparing contrast of the two points (areas). During study we focused on tracking of the ‘‘contrast change’’ value, which is the relative signal intensity of sample in comparison with reference (distilled water, and/or distilled water with CAs). For image data processing and basic analysis we used the following software tools: Marevisi (NRC – Institute for Biodiagnostic, Winnipeg, Canada), and Matlab R2011b (Mathworks Inc., Natic, USA). 3. Results and discussion In Figs. 1 and 2 MRI images of the selected contrast agents, together with the shape of the concentration curve, are shown in comparison with the signal intensity. Fig. 1a shows the change in the contrast of the Resovist contrast agent in comparison with the concentration increase. The appropriate concentration curve is in Fig. 1b, where we observe a decrease of signal with the concentration increase of iron oxide nanoparticles. The change is due to the influence of magnetite nanoparticles in Resovist when imaged with T2-weighted sequences. The same situation for MultiHance is in Fig. 2a and b. In this case we observed an increase in signal intensity when imaged with T1weighted sequences, caused by gadolinium particles. For both cases, we observe the expected shape of the curve of signal intensity in comparison with an increase of the contrast agent concentration: decrease for Resovist and increase for MultiHance.
Fig. 2. (a) MRI of the different concentration of MultiHance in distilled water, (b) MultiHance contrast change curve depending on gadolinium concentration.
In the first series, we investigated the effect of the physiological concentration of saline and glucose molecules on the contrast properties of selected contrast agents, which simulates the concentration changes of the contrast agent in the blood stream during in-vivo administration. For each contrast agent, we prepared four solutions with increasing concentration of magnetic particles: distilled water, saline, glucose, saline + glucose. The solutions were imaged with the selected pulse sequence.
Fig. 3 shows the intensity change of different concentrations of Resovist in comparison with reference (distilled water). We can see that a physiological concentration of saline and glucose molecules selectively influence the contrast properties of the CA depending on its concentration. These results were at variance with our expectations, where we assumed the same change in intensity value for all concentrations of contrast media. Saline molecules cause an extreme increase in contrast – up to 30% for samples with a higher concentration of contrast agent (above 130 lg/ml, recommended concentration for in-vivo administration is 97 lg/ml). The smallest change in contrast we observed for samples with the presence of glucose molecules (up to 20%). For a mixture of saline and glucose molecules the maximum change is about 25%. What is interesting is that we do not observe these maximum changes in the same sample, but always for a different concentration of CA: 7. for saline, 3. for glucose and 4. for a mixture of saline and glucose. In addition, for saline sample this change is positive, while for glucose
Fig. 1. (a) MRI of the different concentration of Resovist in distilled water, (b) Resovist contrast change curve depending on iron oxide concentration.
Fig. 3. Resovist contrast change curve in different pools in comparison with water.
3.1. CA vs. physiological concentration of MOI
112
O. Strbak et al. / Measurement 69 (2015) 109–114
Fig. 4. MultiHance contrast change curve in different pools in comparison with water.
Fig. 6. MultiHance contrast change values cleared from the MOI-water influence.
and a saline and glucose mixture the change is negative, relative to the reference. For application purposes are interesting the changes in the 4. sample, which contains the recommended concentration of contrast agent for invivo administration. However, we found that the mere presence of saline and glucose molecules in the absence of contrast agent, change the contrast: for saline 4.1%, for glucose 8.8%, and for a mixture of saline and glucose up to the 14.9% in comparison with reference. This is almost a naked-eye visibility limit for our system. Therefore, in Fig. 5 we show the impact of saline and glucose molecules deprived from their effect to the distilled water. In the figure we can see that only three columns reach the threshold of visibility to the naked eye, and only for saline samples. The CA concentrations of these samples are over the recommended in-vivo concentration, and therefore should not affect the situation in clinical practice. The contrast change in sample No. 4, which contains the recommended in-vivo concentration,
is for studied MOI, below the limit of naked-eye visibility. The largest change is for the mixture of saline and glucose ( 9%), then the glucose alone (3.4%), and finally for the saline sample (less than 1%). For the gadolinium-based contrast agent, these changes are even less visible. Here, we can also see the effect of saline and glucose molecules themselves on the contrast change of distilled water, although in different proportions in comparison with the first sample (compare first points in Figs. 3 and 4): saline 15% (before 4.1%), glucose 12.3% (before 8.8%), and the saline and glucose mixture only 1.4% (before 14.9%). From this point it seems like the ‘random’ effect of these molecules on the contrast change of distilled water. This effect is further reflected in samples with CAs. Fig. 6 shows the influence of the physiological concentration of MOI to the contrast properties of the contrast medium MultiHance, cleared from the effect of MOI. In this case, all contrast changes are below naked-eye visibility. It seems like the glucose decreased the effect of the saline, which is also the most visible in this case. For sample No. 5, which contains the recommended in-vivo concentration of MultiHance contrast agent, the impact on the contrast is as follows: glucose 8.9%, saline 11.2%, and saline with glucose only 0.5%. These changes do not affect the visibility to the naked eye for our system. 3.2. In-vivo concentration of CA in various MOI In the second case, we chose one concentration of the CA which is similar to the clinical in-vivo concentration, Table 2 Comparison of the relative contrast changes of the physiological concentration of MOI and CA acquired from both series.
Fig. 5. Resovist contrast change values cleared from the MOI-water influence.
CA with physiological concentration of MOI
Contrast change in 1st series (%)
Contrast change in 2nd series (%)
Resovist – saline Resovist – glucose MultiHance – saline MultiHance – glucose
0.6 3.4 11.2 8.8
4.1 2.1 5.0 12.4
113
O. Strbak et al. / Measurement 69 (2015) 109–114
Fig. 7. (a) Resovist and MultiHance contrast change values in comparison with reference for one CA concentration (Resovist – 103.3 lg/ml of the iron oxide nanoparticles, MultiHance – 1.3 mg/ml of the gadolinium particles). Samples: saline S1-4 (4.5, 9, 13.5, 18 g/l), glucose G1-4 (0.5, 1, 1.5, 2 g/l). (b) MRI of Resovist. (c) MRI of MultiHance.
and we followed the four different concentrations of MOI, around their physiological concentration. Such an arrangement simulated a concentration change of biological molecules in various pathological processes which may accompany an MRI examination. Saline concentration: S1 = 4.5 g/l, S2 = 9 g/l, S3 = 13.5 g/l, S4 = 18 g/l, glucose concentration: G1 = 0.5 g/l, G2 = 1 g/l, G3 = 1.5 g/l, G4 = 2 g/l. Changes visible to the naked eye occurred in only two samples: S1 16.6% for Resovist, and G1 16.1% for MultiHance (see Fig. 7). Changes also have the character of randomness and do not fit the concentration changes of MOI in this case. Almost the same change in contrast can be seen when samples S2 and G2 are compared from the second series, which include the physiological concentration of MOI, with the data from the first series (sample No. 4 for Resovist, and sample No. 5 for MultiHance), which also contain physiological concentrations of MOI (Table 2). From this point of view, it formerly excludes the random effect of influence, as discussed above. These results showed that the effect of selected MOI on the MRI contrast change exists and is not negligible. However, sometimes it rather appears at random (influence to distilled water), but in general the same ‘‘pattern’’ of influence can be seen (Tables 3 and 4). In case of Resovist it is a decrease of signal (Fig. 3) for concentrations up to 103.3 lg/ml (which is recommended in-vivo concentration), followed by a steep increase for all samples. In case of MultiHance we can see a decrease of signal (Fig. 4) for
Table 3 ‘‘Pattern of action’’ of the saline and glucose physiological concentration to the CA contrast properties. CA
Saline (9 g/l)
Glucose (1 g/l)
Saline (9 g/l) + Glucose (1 g/l)
Resovist
Decrease/ increase Mostly decrease
Decrease/ increase Mostly decrease
Decrease/ increase Mostly decrease
MultiHance
Table 4 ‘‘Pattern of action’’ of the increasing concentration of saline and glucose to the CA contrast properties. CA
Saline concentration increase (4.5–18 g/l)
Glucose concentration increase (0.5–2 g/l)
Resovist (103.3 lg/ml) MultiHance (1.3 mg/ml)
Full decrease
Mostly increase
Mostly decrease
Full decrease
whole selected concentration range. The data can be such transform to the pattern information as ‘‘decreasing/increasing’’ trend for Resovist (with the minimum around 100 lg/ml), and ‘‘decreasing’’ trend for MultiHance (Table 3). Such a selective influence of physiological concentration of molecules to the contrast properties of CAs depending on their concentration is the most surprising finding for us. In case of increasing concentration of saline we observed the ‘‘decreasing’’ trend for both CAs, while in case of glucose we found the ‘‘increasing’’ character for Resovist and ‘‘decreasing’’ trend for MultiHance (Table 4). The fact is that the changes are mostly below or at the limit of visibility to the naked eye for our system. However, without taking into consideration the question of randomness or pattern of action, the changes are large enough to be able to affect the calculation of parameters such as relaxation times. Moreover, these changes should be more visible with the high-field systems, which are more often used in clinical practice. Therefore, this MOI influence on the contrast of water molecules and CAs should not be forgotten and should be subjected to more precise study.
4. Conclusion The aim of the study was to determine the effect of biological MOI on the contrast properties of clinically used CAs for MRI. The study was undertaken due to the fact that
114
O. Strbak et al. / Measurement 69 (2015) 109–114
while there is routine use of CAs in clinical practice, the information on the influence of the in-vivo environment on the contrast properties is lacking or insufficient. Various physiological and especially pathological processes, accompanied by changes in concentration levels of biologically active molecules in the body, may affect with their properties the MRI signal and thereby distort the output data. Since saline and glucose molecules are well represented in the human body, we focused our attention on them. We followed their impact on the contrasting properties of clinically used contrast agents, both containing iron oxide nanoparticles as well as gadolinium particles. The planned outputs should be the information about the changes in MRI contrast, influenced by these molecules, which could have an impact on diagnostic conclusions in clinical practice. The images were acquired using both Spin Echo (SE) and Gradient Echo (GE) pulse sequences in T1- and T2weighted modes. For the purpose of data processing, the most suitable sequence was determined. For Resovist it was a T2-weighted Turbo Spin Echo sequence with a repetition time of TR = 3000 ms, and Echo Time TE = 120 ms. For MultiHance we used a T1-weighted Spin Echo sequence with a repetition time of TR = 600 ms, and an echo time of TE = 26 ms. The experiments were divided into two parts. At first, we were looking for the impact of physiological concentrations of selected MOI to the contrast properties of the studied contrast agents – Resovist and MultiHance in different concentrations. We were interested whether the physiological concentration of saline and glucose molecules has the same effect on the contrast medium, which simulates in-vivo conditions after the administration of the CA into the bloodstream. We found that contrast change for Resovist is up to 30. 6% compared with the reference. The largest changes in contrast were observed in the samples containing salt molecules. With the MultiHance CA the biggest changes of the contrast range are about 15% compared with the reference. It has been shown that the mere presence of saline and glucose molecules in distilled water change the contrast level by up to 15% in comparison with the reference for both contrast media. This is on the border for the nakedeye visibility changes for our system. The resulting change in the contrast deprived of the influence of MOI themselves to the distilled water for ranges up to 26.5% for Resovist in saline solution. For the MultiHance CA, the biggest intensity change is around 13%, as for the saline solution. In the second part, we investigated the effect of varying concentrations of the MOI on the contrast properties of the recommended in-vivo concentration of CA, which simulates the change in the concentration of biologically active
molecules in various pathological processes. It has been shown that the presence of MOI at various concentrations also affects the contrast of iron oxide-based Resovist, and the biggest change in the contrast is as high as 16.6% in the presence of saline molecules. For the MultiHance contrast agent, the biggest change reaches 16.1%. Although these changes are in most cases below or around the limit visibility to the naked eye for our system (15%), its height is not negligible and may affect the further analysis of data, e.g. calculation of the relaxation times. Thus, it may alter the final clinical conclusions. Moreover we found the non-random pattern of action of the saline and glucose molecules to the contrast properties of the selected CAs. In case of physiological concentration of saline and glucose we surprisingly found the selective influence to different concentrations of CA. Trends were as follows: ‘‘decreasing/increasing’’ for Resovist (with the minimum around 100 lg/ml), and ‘‘decreasing’’ for MultiHance. In case of increasing concentration of saline and glucose, the trend was ‘‘decreasing’’ for saline in both CAs, and ‘‘increasing’’ for Resovist and decreasing for MultiHance for glucose. Acknowledgements This work was supported within the project of the Slovak Research and Development Agency Nr. APVV0431-12, and Slovak Scientific Grant Agency VEGA 2/ 0013/14. References [1] P. Gillis, F. Moiny, R.A. Brooks, On T2 shortening by strongly magnetizes spheres: a partial refocusing model, Magn. Reson. Med. 47 (2002) 257–263, http://dx.doi.org/10.1002/mrm.10059. [2] Y.X.J. Wang, Superparamagnetic iron oxide based MRI contrast agents: current status of clinical application, Quant. Imaging. Med. Surg. 1 (2011) 35–40, http://dx.doi.org/10.3978/j.issn.2223-4292. 2011.08.03. [3] P. Caravan, J.J. Ellison, T.J. McMurry, R.B. Lauffer, Gadolinium(III) chelates as MRI contrast agents: structure, dynamics, and applications, Chem. Rev. 99 (1999) 2293–2352, http://dx.doi.org/ 10.1021/cr980440x. [4] S.S. Anand, G.R. Dagenais, V. Mohan, et al., Glucose levels are associated with cardiovascular disease and death in an international cohort of normal glycaemic and dysglycaemic men and women: the EpiDREAM cohort study, Eur. J. Prev. Cardiol. 19 (2012) 755–764, http://dx.doi.org/10.1177/1741826711409327. [5] K. Matz, K. Keresztes, C. Tatschl, et al., Disorders of glucose metabolism in acute stroke patients, Diabetes Care 29 (2006) 792– 797, http://dx.doi.org/10.2337/diacare.29.04.06.dc05-1818. [6] C.S. Lin, C.J. Cheng, K.C. Shih, S.H. Lin, Recurrent hyponatremia in a patient with chronic kidney disease, J. Nephrol. 19 (2006) 394–398. [7] G. Lindner, G.C. Funk, Hypernatremia in critically ill patients, J. Crit. Care 28 (2013) 216.e11, http://dx.doi.org/10.1016/j.jcrc.2012.05.001. [8] G. Riccardi, A.A. Rivellese, Effects of dietary fiber and carbohydrate on glucose and lipoprotein metabolism in diabetic patients, Diabetes Care 14 (1991) 1115–1125, http://dx.doi.org/10.2337/diacare.14. 12.1115.