Physical and pharmacokinetic characterization of Soluvec™, a novel, solvent-free aqueous ivermectin formulation

Ivermectin (IVM) is the best studied and most widely used of the avermectin family of macrocyclic lactones [1–3]. IVM acts as an endectocide in livestock and companion animals, targeting both endoparasitic helminths and ectoparasitic arthropods. In humans, IVM has revolutionized the management of river blindness and lymphatic filariasis. At low nanomolar concentrations, IVM disrupts neural and neuromuscular transmission in parasitic nematodes by constitutively activating glutamate-gated chloride (GluCl) channels [3].

While GluCl genes are absent from the vertebrate genome [2], IVM and other avermectins can also act on various ion channels and other molecular targets in invertebrate and vertebrate cells, albeit with lower affinity [4–6]. As such, avermectins are potentially promising agents for treating viral [7] and mycobacterial [8] infections as well as cancers [9].

The low aqueous solubility of IVM (∼0.004 mg/ml) [10] greatly limits its bioavailability following oral dosing or injection. Various solubilization approaches have been proposed, particularly for parenteral formulations [10–13]. Thus, while avermectins dissolve readily in organic solvents, these excipients may not be well tolerated, and their use in subcutaneous (SC) or intramuscular (IM) injection may lead to local precipitation of the drug within the injection depot, with unpredictable consequences for absorption kinetics. Therefore, an ideal formulation for parenteral delivery would have high bioavailability and could be administered at high concentration without organic solvents [12,14].

Quicksol™ is a two-step approach to the solubilization of macrocyclic lactones [15,16], in which a poorly water-soluble compound is first complexed with a cyclodextrin [17] and then dissolved in the presence of a polysorbate nonionic surfactant [18]. Here, we describe the use of the Quicksol approach to prepare an aqueous formulation of IVM (Soluvec™). We report the physical characterization of Soluvec and provide an initial pharmacokinetic (PK) analysis of this novel IVM formulation in beagles.

Materials & methods

Preparation of solubilized ivermectin

Following the Quicksol solubilization strategy [15,16], hydroxypropyl-β-cyclodextrin (HP-β-CD; Trappsol^®, CTC, Inc., FL, USA) and IVM USP (Letco Med, AL, USA) (85:15 w/w) were mixed in 2:1 methanol:water at 50°C. The mixture was lyophilized, and the dry residue collected and ground to a powder. Gas chromatography-flame ionization detection established that residual methanol was below International Council for harmonization limits for solvents in pharmaceuticals. Incubation of the IVM/HP-β-CD lyophilizate at 50°C in 9% aqueous polysorbate 80, followed by filtration (0.2 μm), yielded a clear, sterile solution.

Analytical chemistry

For differential scanning calorimetry (DSC) of IVM USP and the IVM/HP-β-CD lyophilizate, 5 mg samples were equilibrated to 20.0°C and ramped at 10°C/min to 350°C in a TA Instruments Q20 DSC. Fourier transform infrared spectroscopy (FTIR) was performed using a Perkin Elmer Spectrum 65 FT-IR equipped with an attenuated total reflectance accessory, scanning over the range of 4000–600 cm^-1.

IVM concentration in the IVM/HP-β-CD lyophilizate was quantified using high performance liquid chromatography (HPLC; Agilent 1100 with an Xbridge C8 column), using 100% methanol as the diluent and acetonitrile–methanol–water (42.5:42.5:15) as the mobile phase. IVM was quantified by absorption at 248 nm [19].

Dynamic light scattering (DLS) of the IVM/HP-β-CD/polysorbate 80 solution (Soluvec) was analyzed using a Zetasizer (Malvern Instruments, Ltd, Malvern, UK).

Pharmacokinetics

Eighteen 2-year-old healthy male beagle dogs (body weights 10.1 ± 0.5 kg) were allocated 6:6:6 to oral dosing (Avemac^® [Opsonin Pharma, Ltd], one 3 mg tablet per dog) or parenteral (IM or SC) dosing with Soluvec (0.3 mg/kg). IM injections were to the caudal thigh; SC injections were to the dorsal cervical skin. Dogs were fed ad libitum 2 h before dosing. Venous blood was collected at 15 and 30 min and 1, 2, 4, 8, 12, 24, 36, 48, 72, 96, 144, 216, 288, 360 and 480 h post-dose.

Plasma IVM was determined by liquid chromatography/tandem mass spectroscopy. IVM standards and study samples (50 μl each) were supplemented with abamectin (serving as an internal standard for peak area-ratios) and extracted with water and acetonitrile. Following centrifugation at 20,000 × g for 10 min at 4°C, the organic phase was collected, dried under nitrogen at 50°C and reconstituted with 50% methanol. Chromatography was performed using a Phenomenex Kinetex column (150 × 4.6 mm, 2.6 μm) with a gradient of aqueous 10 mM ammonium bicarbonate and 0.1% formic acid in methanol, flow rate 1.0 ml/min over 10 min. IVM and abamectin were detected using a TSQ Vantage triple quadrupole mass spectrometer coupled with an Accela 1250 UHPLC pump and Accela Open autosampler (ThermoFisher Scientific).

Single-dose PK parameters were evaluated by Phoenix^® WinNonlin^® version 8.3 software (Certara Inc.).

Results

Preparation of solubilized ivermectin

Following incubation of IVM with HP-β-CD, the mixture was lyophilized and ground to produce a white, flowable powder, which was analyzed by DSC (Figure 1) and FTIR (Figure 2). Compared with the starting IVM, which showed a single sharp melting point at the expected temperature (158°C), the IVM/HP-β-CD lyophilizate had a second melting point at 241°C.

Comparison of the FTIR findings showed the expected spectrum for pure IVM, whereas, in the IVM/HP-β-CD lyophilizate, two absorption peaks in the carbonyl stretch region of the spectrum (1676 and 1730 cm^-1) were present at reduced intensity, relative to pure IVM, as has been reported previously [14]. Following solubilization of the lyophilizate in 9% polysorbate 80, the sterile-filtered aqueous solution (Soluvec) contained 10 mg/ml IVM. The solution was clear and viscous, with pH 7.13. DLS showed a nearly monodisperse (polydispersity index 0.137) solution with mean particle size of 28.0 nm.

To investigate the physical basis of this reduction in absorption intensity at 1676 and 1730 cm^-1, the IVM/HP-β-CD complex was prepared again and solubilized in water (rather than 9% polysorbate 90, as described above). Following filtration to eliminate any insoluble, non-complexed IVM, this preparation was lyophilized a second time, for FTIR analysis. In contrast to starting IVM/HP-β-CD complex, the water-resolubilized, filtered IVM/HP-β-CD complex showed no absorption at 1676 cm^-1 or 1730 cm^-1 (Supplementary Figure 1).

Pharmacokinetic analysis in beagle dogs

Table 1 shows PK parameters derived for dogs treated with Soluvec 0.3 mg/kg parenterally (SC or IM), compared with those receiving an oral dose (3 mg/dog) of a commercially available IVM tablet.

With all dosing approaches, IVM absorption was rapid in most animals (Figure 3), with median T_Max ranging from 1.25 to 4 h across dosing groups. Overall IVM exposure was greater with parenteral Soluvec (mean area under the curve (AUC) 6237 ng*h/ml and 6053 ng*h/ml for SC and IM dosing, respectively) versus oral IVM (895 ng*h/ml), a difference of nearly sevenfold. Peak exposure was likewise greater for parenteral dosing, albeit by a smaller factor. In addition, duration of exposure to therapeutic levels of IVM was consistently greater with Soluvec treatment. None of the six animals receiving oral IVM had plasma IVM ≥50 ng/ml beyond the 24-h collection time point, whereas all 12 of the animals receiving parenteral Soluvec (100%) had plasma IVM in this range for at least 48 h after injection.

Inter-animal PK variability was greatest in SC-treated dogs. This variability was primarily due to two animals in the SC Soluvec group, which showed substantially extended exposure (T_Max at 24 and 72 h), relative to all other animals in any treatment group (T_Max ranging from 0.5 to 4 h).

Three of the dogs receiving IM Soluvec experienced mild hives or local inflammation within the first hour post-injection. In all cases, these episodes resolved within 1 h without treatment. No other tolerability issues were evident in oral IVM- or Soluvec-treated animals.

Discussion

Despite a variety of innovative preparations described in the literature [10–14,20,21], there remains a substantial need for improved parenteral dosing of avermectins. In veterinary medicine, formulations with good bioavailability are needed to overcome parasite resistance in livestock and companion animals [22,23]. In the clinical setting, oral dosing of IVM is the rule, and no injectable formulations are currently available. However, given the unfavorable PK properties of IVM tablets, as well as practical limitations on patients’ daily pill burden, treaters have explored SC administration of IVM in some high-need patients.

IVM is an effective treatment for Strongyloides stercoralis infection, among other helminthic diseases [24–26]. Individuals with advanced strongyloidiasis may experience a significant decline in serum albumin. Because this protein is the major carrier protein for IVM in the circulation [27], hypoalbuminemia leads to rapid IVM turnover and restricted drug distribution volume, further complicating the maintenance of therapeutic IVM levels [28]. SC IVM has shown some success in increasing IVM exposure and achieving parasite clearance in severely infected patients [24,25,28–30].

The Quicksol approach was developed to produce solvent-free, aqueous preparations of macrocyclic lactones. As shown here, this technique yields a high-concentration, solvent-free, aqueous preparation of IVM, Soluvec, providing proof of principle that avermectins and related compounds may be formulated similarly through a scalable, Good Manufacturing Practice-compliant process. Ongoing stability studies show the formulation remains stable at both 5 and 25°C storage conditions for at least 9 months. Other than phosphate buffer, the sole excipients in Soluvec are HP-β-CD and polysorbate 80 [17,18], which are widely used in food and pharmaceutical preparations, with well-defined safety and tolerability profiles and rapid excretion kinetics. Both are US FDA approved inert materials, used in intramuscular and other parenteral or oral formulations [31].

FTIR spectroscopy and melting point analysis confirmed that IVM formed a complex with HP-β-CD [14]. After solubilization in aqueous 9% polysorbate 80, the IVM concentration was 10 mg/ml, thus achieving a ∼2500-fold increase in solubility, relative to free IVM in a detergent-free aqueous solution. Soluvec is also more than twice as concentrated as another aqueous preparation in which an ionic detergent was used to solubilize IVM [10]. This relatively high-concentration stock could facilitate repeat-dosing of livestock. The sterile aqueous solution could also be adapted for nasogastric, topical or intraocular dosing.

As previously observed [14], complex formation with HP-β-CD likely constrains the molecular motion of IVM, suppressing carbonyl stretch vibrations that create FTIR absorption peaks at 1676 and 1730 cm^-1. In water-solubilized, filtered IVM/ HP-β-CD complexes, we observed the complete loss of these two absorption peaks, whereas, in the crude lyophilizate, these peaks were present at reduced intensity relative to pure IVM. Hence, it appears that IVM in Soluvec is present as a mixture of two physical forms: free IVM (solubilized by the presence of the nonionic detergent polysorbate 80) and IVM/HP-β-CD complex. The complex forms a homogeneous population (PDI < 0.15) of 28 nm particles. The small size and relative uniformity of these complexes are expected to enhance and improve the reliability of drug delivery from the complexed form of IVM in Soluvec [32,33].

As expected [34–36], parenteral dosing of Soluvec in beagles was associated with substantially improved PK properties, relative to oral dosing with an IVM tablet. Thus, overall exposure was nearly seven-times higher, with increased peak exposure. Without exception, all animals receiving parenteral Soluvec maintained plasma IVM levels above 50 ng/ml for an extended duration, relative to those ingesting an IVM tablet.

In previous dosing studies, PK parameters are reported to vary among different commercial IVM preparations, reflecting sensitivity to excipients and manufacturing processes [36]. Eraslan et al. [37] carried out a direct comparison of seven IVM preparations in dogs, all dosed SC at 0.2 mg/kg. Both mean C_Max (minimal plasma concentration, C_Max) and mean AUC values varied significantly over approximately a twofold range (C_Max 29.5 ng/ml to 49.7 ng/ml; AUC 3074 ng*h/ml to 5980 ng*h/ml). Because of different dosing (0.3 mg/kg) and substantial PK variability, specifically among the five dogs treated SC with Soluvec, the values reported here cannot be compared directly with those of Eraslan et al. [37], but they are of a similar magnitude.

Whereas IVM is acutely toxic in collie dogs and some others that commonly carry a mutation in the drug transport protein MDR1, beagles can be safely treated with single doses up to 80 mg/kg [38]. The local inflammation observed in some IM-injected dogs in this study was expected, given the presence of polysorbate 80 in the preparation, since this excipient has been reported to cause pseudo-allergic responses in beagles [39]. As reported elsewhere, this adverse event was brief and self-limiting. While it is intriguing that no such effect was seen in SC-injected dogs, further studies will be needed to assess the tolerability of Soluvec when used in other animals, particularly for repeat dosing.

Although the mean PK parameters for SC and IM dosing were generally similar, SC dosing of Soluvec was associated with substantially greater scatter in the PK data, relative to IM dosing. Indeed, two of the SC-injected animals continued to show high plasma IVM levels beyond the 72-h collection time point. Inter-individual variation in absorption kinetics has been reported previously in animals receiving SC IVM. Thus, the optimal mode of parenteral dosing of Soluvec remains to be established. PK properties of IVM vary across species and by physiological state, as well as with the formulation and mode of delivery [12,36,40]. Further PK and safety studies of Soluvec in livestock are ongoing.

Limitations of the current study include the absence of a parenteral dosing arm using a commercially available liquid formulation of IVM. In addition, the small number of animals, coupled with the substantial inter-animal variability in the SC arm, limits the precision of the PK findings.

Conclusion

Quicksol solubilization was used to generate Soluvec, a high-concentration, sterile, aqueous preparation of IVM. Using this formulation, it may be possible to increase systemic exposure of IVM or related macrocyclic lactones in the treatment of parasitic infections. The Quicksol approach may also be useful for other avermectins used to treat various infections, infestations and cancers, and it may be generalized to formulate other therapeutic molecules exhibiting poor solubility and low bioavailability in their free form.

Despite the extraordinary success of IVM as a broad-spectrum endectocide, the poor solubility of the avermectins remains a barrier to achieving and maintaining therapeutic drug levels in humans and livestock species. Over the coming decade, this challenge will continue and may worsen with growing drug resistance. For this reason, many research groups have explored new IVM formulations that might be produced at high aqueous concentration. Such preparations would simplify parenteral dosing and bring higher circulating IVM levels within reach. Over the same time frame, parenteral administration in human patients, which has already been explored as an off-label option, will likely be approved for some indications, including filarial and helminthic diseases and, more speculatively, in novel anti-cancer applications.