Stability testing of gadoteridol and gadobenate dimeglumine formulations under exposure to high-intensity focused ultrasound

Abstract

Objective:

Contrast-enhanced MRI could be useful to guide high-intensity focused ultrasound treatment (HIFU), but the effects of HIFU on gadolinium-based agents is not known. Here, we tested in vitro the stability of gadoteridol and gadobenate dimeglumine, two widely used MR contrast agents, after exposure to HIFU at power levels typically applied in the clinical practice.

Methods:

0.5 M (gadoteridol and gadobenate dimeglumine) and diluted formulations (1:10 gadoteridol in saline) were exposed to different HIFU sequences. Unexposed and exposed solutions were characterized by high-performance liquid chromatography in terms of concentration of gadolinium complex, free gadolinium and free ligand.

Results:

Gadoteridol formulation after treatment showed concentrations of the complex not significantly different from control. Free Gd and/or free ligand concentrations in the order of 0.002/0.004% w/w, were observed occasionally without significant correlation with intensity and duration of exposure to HIFU. Gadobenate dimeglumine formulation after treatment showed complex assay content values, by-products (0.24–0.26%) and free BOPTA levels (0.07%) comparable to control sample within the experimental error.

Conclusion:

In the range of conditions explored, HIFU exposure did not induce significant dissociations of gadoteridol and gadobenate dimeglumine, nor a detectable increase in the concentration of free species.

Advances in knowledge:

Our study strengthens the hypothesis that gadolinium-based contrast agents are stable during HIFU treatment for body applications (e.g. thermal ablation of uterine fibroids).

Introduction

High-intensity focused ultrasound (HIFU) is a non-invasive technique introduced in the last decades into clinical practice to perform thermal ablation in a deep-seated target volume. ^1,2 Current HIFU treatments are performed under either ultrasound-HIFU or MR-HIFU imaging guidance, to precisely guide the ultrasound beam. Currently, thermal ablation of tissues is well established for the treatment of uterine fibroids and for pain alleviation in patients suffering from bone metastases. ¹ Moreover, several new applications are emerging (breast cancer, ³ prostate cancer, neurological disorders, abdominal applications). ²

For uterine fibroids, contrast-enhanced MRI enables precise evaluation of the efficacy of HIFU treatment, since non-perfused volume (NPV) is indicative of successful ablation. ^4–6 Monitoring NPV during treatment would help physicians identify remaining viable fibroid tissue more accurately, thus preventing unnecessary treatment of necrotic tissue and potentially drastically reducing treatment time. However, despite potentially important advantages, there are concerns over the use of Gd-based contrast agents (GBCAs) just before or during HIFU ablation, ⁷ including dissociation of Gd-chelates under thermal and mechanical stress and potential entrapment of the complexes in the ablated region. A preclinical study by Hijnen et al ⁷ investigated the use of Gd-DTPA before HIFU in rats, with respect to dissociation, trapping, and long-term deposition of gadolinium in the body. Those authors did not report evidence of variation in pharmacokinetic, retention or in vivo relaxivity of Gd-DTPA in HIFU-exposed vs unexposed animals.

However, to the best of our knowledge, a dedicated in vitro study to directly measure GBCAs dissociation under high-intensity sonication, up to rather extreme conditions, has never been performed. The aim of this study is to systematically assess the effects of HIFU at different power levels and ON-OFF cycles on the stability of gadoteridol (ProHance^™) and gadobenate dimeglumine (MultiHance^™), two widely used MR GBCAs, and to measure the potential release of free species under controlled experimental conditions.

Methods and materials

Two GBCAs were analysed: gadoteridol for injection (ProHance; Bracco Imaging SpA, Milan, Italy; batch no. VD2002) and gadobenate dimeglumine for injection (MultiHance; Bracco Imaging SpA, Milan, Italy; batch no. Z19803). Original formulations with a concentration of 500 mM were used as received; gadoteridol was also diluted 1:10 in saline to a concentration of 50 mM.

A home-made HIFU system was developed, as follows. A very thin dialysis tube (20–30 microns thick membrane) of 6.4 mm inner diameter (Spectra/Por 6, Repligen, Waltham, MA), transparent to ultrasound, was filled with about 1 ml of solution. A focused 0.5 MHz transducer (H107-41, Sonic Concepts, Bothell, WA) with diameter of 64 mm and a radius of curvature of 63 mm was used to expose GBCA samples to high-intensity sonication. The transducer was driven by a 0.5 MHz sinusoid tone burst generated by a waveform generator (3,3600A, Keysight, Santa Rosa, CA) and amplified by a 400W RF amplifier (1040L, E&I, Rochester, NY) with a 470Ὠ load input. The HIFU transmit sequences (n seconds ultrasound ON, m seconds ultrasound OFF) were programmed through the Burst mode of the Keysight generator by tuning the number of cycles and the repetition period. The power delivered by the transducer was defined as the electrical power dissipated during the ON time of the sequence and was calculated from the complex impedance and the voltage amplitude applied to the transducer. A dedicated setup was built in a water tank to align accurately the transducer focus on the cell containing the GBCA sample.

Two (gadoteridol a and d) and four (gadobenate a,b,c, and d) HIFU sequences were selected for exposing GBCA samples: (a) 5 s ON and 5 s OFF repeated five times, (b) 5 s ON and 5 s OFF repeated 10 times, (c) 10 s ON and 5 s OFF repeated five times, (d) 20 s ON and 5 s OFF repeated five times. Four HIFU power levels were set, namely 100 W, 200 W, 300 W and 400 W. The highest power levels (300 W and 400 W) were tested only with sequence (a) (i.e. 5 s ON and 5 s OFF repeated five times), to avoid destruction of the cell or the transducer due to strong cavitation and powerful reverberation in the tank. Power levels are comparable to those applied in clinical practice for body application (especially during thermal ablation of uterine fibroids), while ON/OFF timing are more stressful than in typical HIFU treatment. In the case of gadoteridol, three samples were subjected to each HIFU protocol and samples were analysed separately. In the case of gadobenate, four samples were subjected to each condition and pooled to obtain a single sample of volume about 4.5 ml, the minimum volume needed for accurate analytical characterisation with our assay. Identical samples not subjected to any HIFU exposure were also analysed, as control.

Analytical methods

Analytical methods strictly derive from those routinely used for product release as to guarantee the most reliable performances for gadoteridol and gadobenate dimeglumine assay, free species and by-products testing.

Gadoteridol content was performed by high-performance liquid chromatography (HPLC)/FLD (excitation 274 nm, emission 311 nm). Column: Supelcosil LC-18-DB (250 × 4.6 mm, 5 µm) set at 30°C. Mobile phase: phosphate buffer (0.025 M, pH 5) and acetonitrile (98:2 %v/v). Flow rate: 1 mL/min in. Elution: in isocratic mode. Sample solution concentration: 0.562 mg ml⁻¹ in terms of gadoteridol content in phosphate buffer (0.025 M, pH 5). Injection volume: 20 µl.

Free gadolinium content of gadoteridol samples was performed by HPLC/FLD (excitation 274 nm, emission 311 nm) equipped with autosampler with sample holder thermostated at 5°C. Column: Supelcosil LC-18-DB column (250 × 4.6 mm, 5 µm) set at 20°C. Mobile phase phosphate buffer/EDTA 0.025 M and acetonitrile (98:2% v/v). Flow:1 ml/min. Elution: in isocratic mode. Sample solution concentration: 28 mg ml⁻¹ in terms of gadoteridol content in phosphate buffer/EDTA 0.025M. Injection volume: 20 µl.

Free ligand content of gadoteridol samples was performed with an HPLC/DAD (280 nm) equipped with autosampler with sample holder thermostated at 5°C. Column: Hamilton, PRP-X100 (15 cm x 4.1 mm, 10 µm) set at 30°C. Mobile phase: TRIS/EDTA aqueous buffer with acetonitrile and THF (98.8:1:0.2% v/v). Preparation of TRIS/EDTA buffer: 0.3 g of TRIS and 1.86 g of EDTA in 2 l of water, pH 7 with 5 M NaOH, filter with 0.45 µm filters and cool to 5°C. Flow: 2 ml/min. Elution: in isocratic mode. The method determines by UV analysis the free ligand complex with Cu2+ after derivatization with copper acetate. Sample preparation: 1 ml of the solution in a 20 ml flask with 9 ml of TRIS/EDTA buffer cooled to 5°C. Shake and add 10 ml of copper acetate. Copper acetate solution: approximately 2 g of copper acetate and TRIS in 100 ml with water, pH 7 with glacial acetic acid, refrigerate at 5°C. Injection volume: 50 µl.

Gadobenic acid content was performed using an HPLC/DAD (210 nm). Column: LiChrospher 100 RP-8 (250 × 4 mm, 5 µm) set at 50°C. Mobile phase: aqueous solution of n-octylamine (7.7 mM) and acetonitrile (73:27% v/v), pH 6. Flow: 1 ml/min. Elution: in isocratic mode. Sample concentration: 2 mg ml⁻¹ in terms of gadobenate dimeglumine in water. Injection volume: 10 µl.

By-product content of gadobenate dimeglumine samples was performed using an HPLC/DAD (210 nm). Column: Superspher 100 RP-18 LiChroCART (250 × 4 mm, 4 µm) set at 45°C. Mobile phase: aqueous solution of Na₂HPO₄ (60 mM)/EDTA (0.04 mM)/tetrahexylammonium hydrogensulfate (3.0 mM), and acetonitrile (68:32% v/v). Flow: 1 ml/min. Elution: in isocratic mode. Sample concentration: 52.90 mg ml⁻¹ in terms of gadobenate dimeglumine in water. Injection volume: 10 µl.

Free BOPTA and free gadolinium of gadobenate dimeglumine samples were determined complexometrically by titration with a solution of gadolinium chloride and disodium edetate (Titriplex^® III), respectively, at pH 5.8. Indicator: xylenol orange (0.3 mg ml⁻¹). Instrument: Dosimat plus 876. Sample solution preparation: 2 ml of the solution into a beaker with 50 ml of pH 5.8 buffer and 1 ml of 0.3 mg ml⁻¹ orange xylenol. If the colour of the sample solution is yellow, BOPTA is present as a free species in the sample and titration is performed with a 0.001 M gadolinium chloride solution to a reddish-yellow end point. If the sample solution color is reddish-yellow to purple, gadolinium is present as a free species in the sample and titration is performed with a 0.001 M disodium edetate solution (Titriplex^® III) to a yellow end point.

Results

Gadoteridol and gadobenate samples were exposed to various HIFU protocols with different powers and exposure times. In the case of gadoteridol, commercial and diluted solutions (n = 3 for each condition) were exposed to protocol a (at 100 W and 400 W) and protocol d (at 100 W and 200 W) and analysed to assess degradation of the Gd chelate. Single-sample and averaged results are shown in Table 1. Gadoteridol percentage mean values are never statistically lower (Student’s t test) than the control sample, showing no significant degradation detectable with this assay. Free Gd and/or free ligand in the order of 0.002/0.004% are occasionally observed without any correlation with respect to the intensity of HIFU exposure, consistent with the previous result.

Table 1.

Gadoteridol content expressed in mg/ml, averaged values over n = 3 samples for each HIFU condition and over n = 2 experiments for the untreated sample, free gadolinium and free ligand content expressed as weight to weight percentage for different HIFU exposure time and power (n.q. stays for not quantifiable, n.d. not detectable)

Sample	HIFU protocol (sequence-power)	Gadoteridol content (mg/ml) – single values	Gadoteridol content (mg/ml) – averaged values over group	Free Gd (% w/w)	Free ligand (% w/w)
Gadoteridol	None	275	274 ± 1	n.q.	0.001
Gadoteridol	None	274		n.q.	0.001
Gadoteridol	a-400W	265	275 ± 10	n.q.	0.003
Gadoteridol	a-400W	285		0.002	0.002
Gadoteridol	a-400W	273		n.q.	0.002
Gadoteridol 1:10	a-400W	285	289 ± 4	n.q.	n.q.
Gadoteridol 1:10	a-400W	292		n.q.	n.q.
Gadoteridol 1:10	a-400W	292		n.q.	n.q.
Gadoteridol	d-200W	282	278 ± 7	n.q.	0.002
Gadoteridol	d-200W	281		n.q.	0.003
Gadoteridol	d-200W	270		n.q.	0.004
Gadoteridol 1:10	d-200W	258	277 ± 16	n.q.	n.q.
Gadoteridol 1:10	d-200W	289		0.002	n.q.
Gadoteridol 1:10	d-200W	284		0.002	n.q.
Gadoteridol	a-100W	268	269 ± 10	0.002	n.q.
Gadoteridol	a-100W	271		n.d.	n.q.
Gadoteridol	a-100W	252		n.d.	n.q.
Gadoteridol 1:10	a-100W	258	274 ± 9	n.d.	n.q.
Gadoteridol 1:10	a-100W	274		n.d.	n.q.
Gadoteridol 1:10	a-100W	274		n.d.	n.q.
Gadoteridol	d-100W	263	273 ± 6	n.d.	n.q.
Gadoteridol	d-100W	270		n.d.	n.q.
Gadoteridol	d-100W	276		n.d.	n.q.
Gadoteridol 1:10	d-100W	269	271 ± 3	n.d.	n.q.
Gadoteridol 1:10	d-100W	274		n.d.	n.q.
Gadoteridol 1:10	d-100W	269		n.d.	n.q.

Data on gadobenate samples are reported in Table 2. Gadobenate content values are comparable to control in all cases, within the experimental error. By-products (0.24–0.26%) and free BOPTA (0.07%) are unvaried with respect to control sample, independently on the intensity or the duration of HIFU exposure.

Table 2.

Gadobenate assay expressed in mg/ml and as percentage vs the control sample, by-products, free gadolinium and free ligand content expressed as percentage for different HIFU exposure time and power

Sample	HIFU protocol (sequence-power)	Gadobenate assay (mg/ml)	Assay % vs Control	By-products (%)	Free Gd or free BOPTA (%)
Gadobenate	None	334	-	0.26	0.07 BOPTA
Gadobenate	a-100W	327	98	0.25	0.07 BOPTA
Gadobenate	a-200W	331	99	0.26	0.07 BOPTA
Gadobenate	a-300W	323	97	0.25	0.07 BOPTA
Gadobenate	a-400W	324	97	0.25	0.07 BOPTA
Gadobenate	b-100W	326	98	0.25	0.07 BOPTA
Gadobenate	b-200W	329	99	0.26	0.07 BOPTA
Gadobenate	c-100W	329	99	0.26	0.07 BOPTA
Gadobenate	c-200W	331	99	0.26	0.07 BOPTA
Gadobenate	d-100W	322	97	0.25	0.07 BOPTA
Gadobenate	d-200W	327	98	0.24	0.07 BOPTA

Altogether, we did not find any indication that either gadoteridol or gadobenate might decompose and release free gadolinium or other by-products under the thermal and mechanical stress induced by the HIFU protocols used in this study.

Discussion

Collection of experimental evidence to investigate the behaviour of GBCA under HIFU exposure is fundamental to address concerns in the medical community over the use GBCA just before or during HIFU ablation, despite their undisputed effectiveness in monitoring the effects of treatment. In this study, the first assessment of the stability of the gadoteridol and gadobenate dimeglumine under HIFU treatment is presented, showing no evidence of degradation of the complex or release of free species. These results are well in agreement with observations of the preclinical study performed by Hijnen et al. ⁷ Those authors suggested that Gd3+ remains chelated by DTPA during MR-HIFU ablation treatment of muscle and tumour mass of rats at an acoustic power of 30 W on the basis of two indirect evidences: (i) no significant difference in the amount of Gd present in the bone between HIFU treated and untreated animals 14 days after Gd-DTPA injection and (ii) no variation of Gd-DTPA relaxivity in the ablated region after HIFU treatment.

A few clinical studies reporting administration of GBCA prior to MR-HIFU treatment are also described in literature. First cases were reported by Furusawa et al ^3,5 using MR-contrast agent for the treatment planning of 30 patients with breast cancer. Similarly, at UMC Utrecht, MR-contrast agent was used for the treatment planning of 10 patients with breast cancer. ⁶ No complications or adverse events related to the interaction of gadolinium with ultrasound were reported in these clinical studies.

This communication is focused on assessing the stability of gadoteridol and gadobenate dimeglumine formulations under a wide range of HIFU exposures in terms of complex assay and residual amount of free species and by-products. More stressful exposure conditions, in terms of ON/OFF times of irradiation, than those applied in clinical applications were tested, determining a range of exposures that can be considered safe. The effects of HIFU irradiation were tested at various concentrations, up to 0.5M, the highest attainable.

Some limitations of the study should be pointed out. All experiments were performed in aqueous solutions, and this study may be extended to phantoms that better mimic the properties of human tissues. In addition, despite HIFU protocol parameters (frequency, power and duration) were set to values similar or more stressful than to those applied in clinical treatment, ⁸ an experimental setup was used for the experiment, rather than a clinical device. However, we note that the experimental rig used here enabled accurate determination of the experimental conditions, as well as the exploration of conditions beyond those used in the clinic.

Conclusion

Our study strengthens the hypothesis that GBCAs are stable during HIFU treatment for body applications (e.g. thermal ablation of uterine fibroids). Specifically, experimental evidence demonstrate that the mechanical and thermal stress induced by a highly focused ultrasound beam operating at 0.5 MHz up to a power of 400 W and up to five cycles of 20/5 s of ON/OFF sonication does not induce degradation of the complex nor appreciable release of free species in either gadoteridol and gadobenate dimeglumine.