Identifying a predictive relationship between maximal flow rate and viscosity for subcutaneous administration of macromolecules with recombinant human hyaluronidase PH20 in a miniature pig model

Abstract

Subcutaneous (SC) infusion of large volumes at rapid flow rates has historically been limited by the glycosaminoglycan hyaluronan (HA), which forms a barrier to bulk fluid flow in the SC space. Recombinant human hyaluronidase PH20 (rHuPH20) depolymerizes HA, temporarily eliminating this barrier to rapid SC delivery of large volume co-administered therapeutics. Using a miniature pig model, in-line pressure and applied force to the delivery hardware were measured when subcutaneously infusing a representative macromolecule (human polyclonal immunoglobulin [Ig]), at varying concentrations and viscosities (20–200 mg/mL), co-formulated with and without rHuPH20 (2000 U/mL and 5000 U/mL). Maximal flow rate (Q_max) was calculated as the flow rate producing a statistically significant difference in mean applied force between injections administered with or without rHuPH20. There was a significant reduction in mean applied force required for SC delivery of 100 mg/mL Ig solution with 5000 U/mL rHuPH20 versus Ig solution alone. Similar significant reductions in mean applied force were observed for most Ig solution concentrations, ranging from 25–200 mg/mL when administered with or without 2000 U/mL rHuPH20. Q_max was inversely proportional to Ig solution viscosity and Q_max for solutions co-formulated with 5000 U/mL rHuPH20 was approximately double that of 2000 U/mL rHuPH20 solutions. Mathematical simulation of a hypothetical 800 mg Ig dose co-formulated with rHuPH20 showed that delivery times <30 s could be achieved across a broad range of concentrations. Addition of rHuPH20 can help overcome volume and time constraints associated with SC administration across a range of concentrations in a dose-dependent manner.

Keywords:

• Large volume
• maximal flow rate
• recombinant human hyaluronidase PH20
• subcutaneous
• viscosity

Introduction

Subcutaneous (SC) administration is becoming a more common delivery route for biotherapeutics, such as monoclonal antibodies and immunoglobulin (Ig) G. An increasing number of therapeutics that were previously administered intravenously have been approved for delivery via SC injection in the past decade (Locke et al., 2019).

SC delivery of therapeutic proteins is an attractive alternative to intravenous (IV) administration, due to reduced infusion-related reactions, risk of infection, administration time, and healthcare costs (Dychter et al., 2012; Burcombe et al., 2013; Wynne et al., 2013; De Cock et al., 2016a; De Cock et al., 2016b; Bittner et al., 2018). SC delivery also offers the possibility of self-administration at home, rather than in a hospital setting, for treatments that are compatible with self-administration, such as biotherapeutics for the treatment of rheumatoid arthritis, primary immunodeficiency, or multiple sclerosis (Dychter et al., 2012; Bittner et al., 2018). As a result, both healthcare professionals and patients often prefer SC administration over IV administration (Pivot et al., 2014; Rummel et al., 2017; Bittner et al., 2018). The main reasons given by patients for this preference were the time saved and reduced pain of SC administration compared with IV administration (Pivot et al., 2014; Rummel et al., 2017; Bittner et al., 2018).

Hyaluronan (HA), a naturally occurring glycosaminoglycan, forms a highly hydrated viscous gel in the extracellular matrix throughout the body (Frost 2007; Dicker et al., 2014; Cowman et al., 2015). Hyaluronan resists bulk fluid flow through the SC space of the skin and acts as a barrier, limiting large-volume SC drug delivery and dispersion (Laurent 1972; Frost 2007; Buhren et al., 2016; DuFort et al., 2016). For this reason, SC administration has historically been limited to volumes of <3.5 mL, or it requires multiple injections to deliver the appropriate dose and/or very slow infusion flow rates. Therefore, this route of administration is not suitable for agents requiring large-volume delivery (Locke et al., 2019; Dychter et al., 2012; Frost 2007; McDonald et al., 2010; Hunter 2008; Amgen Inc., 2021).

One strategy designed to address volumetric constraints is to increase the concentration of the drug; however, this can lead to formulation issues (such as aggregation) and often results in higher viscosities, which can limit flow rates and require longer delivery times (Garidel et al., 2017; Wright and Jones 2017; Bittner et al., 2018). An alternative approach is to increase the SC dispersion and absorption by co-formulating therapeutic agents for SC delivery with recombinant human hyaluronidase PH20 (rHuPH20), a highly purified, recombinant human form of the naturally occurring human hyaluronidase PH20 enzyme (Locke et al., 2019).

rHuPH20 facilitates SC delivery of co-administered therapeutic agents by locally reducing HA in the SC space for up to 24 h after administration, thus temporarily removing the barrier to large-volume administration (Bookbinder et al., 2006; Frost 2007; Kang et al., 2012; Locke et al., 2019). As a result, rHuPH20 can facilitate SC administration of co-administered therapeutics at both large volumes and rapid flow rates (Locke et al., 2019; Knowles et al., 2021). rHuPH20 has been approved by the US Food and Drug Administration (FDA) as an adjuvant to facilitate SC administration of hydration fluids and for increasing the dispersion and absorption of other injected drugs (Halozyme, Inc., 2016). In addition, several monoclonal antibodies have received approval for co-formulation with rHuPH20 (Genentech Inc., 2020, 2022; U.S. Food and Drug Administration, 2019, 2020a), and rHuPH20 has been approved for sequential injection with human Ig (U.S. Food and Drug Administration, 2014).

The development of devices for rapid large-volume SC injections led us to consider the role and importance of understanding the applied forces required to inject biotherapeutics of varying viscosities. We set out to evaluate administration parameters during SC injection of human polyclonal Ig solutions at varying concentrations and viscosities with or without the addition of two different concentrations of rHuPH20, using a Yucatan miniature pig model.

The skin of Yucatan miniature pigs is anatomically and physiologically similar to human skin, including its general morphology, epidermal thickness, cellular composition, immunological reactivity, permeability, and metabolic properties (Mahl et al., 2006), making it a suitable nonclinical model for assessing SC administration of biotherapeutics. In addition, previous pre-clinical studies have demonstrated that infusion of human Ig facilitated by rHuPH20 was effective in reducing HA in the SC space in a Yucatan miniature pig model (Kang et al., 2012). A previous study reported similarities between Yucatan miniature pigs and humans in terms of injection-site swelling and back leakage when test solutions were administered in the SC space (Shi et al., 2021). It has also been demonstrated that rHuPH20 can facilitate SC infusion of a polyclonal human gamma Ig solution in a Yucatan miniature pig model by significantly reducing in-line pressure, resulting in increased fluid dispersion while minimizing infusion-related local swelling and induration (Kang et al., 2012).

Here, we present results from studies in Yucatan miniature pigs, which demonstrate the effect of rHuPH20 on in-line pressure and the required applied force during SC administration of Ig solutions, and we identify a predictable relationship between the viscosity of the drug solution and maximal flow rate.

Ethics approval statement

All animal experiments were performed in the Halozyme Therapeutics, Inc. facility and conducted employing sound scientific practices and in accordance with the written study plans (adhering to the Animal Research: Reporting of In Vivo Experiments [ARRIVE] Guidelines) and Halozyme’s standard operating procedures. This study was approved by the Institutional Animal Care and Use Committee (IACUC; Halozyme IACUC P16C).

The use of live animals was essential for this study because the functionally active HA that forms the highly hydrated viscous gel in the SC space can only be mimicked in live animals. To study SC fluid/drug dispersion, the Yucatan miniature pig was selected because it most closely resembles the anatomy of human skin, including underlying components such as the dermis, hypodermis, adipose, and muscle tissue.

A total of six Yucatan miniature pigs were obtained from Premier Biosource (formerly S&S Farms; Ramona, CA), a USDA-regulated facility near the vivarium in which the animals were housed for the duration of the study. All six miniature pigs were utilized in the study and each received two injections per experiment – one test and one control – such that each animal served as its own control. Animals were housed in steel pens with automatic water provided ad libitum and were fed twice daily (AM and PM), except on the PM prior to the study start, due to potential complications with anesthesia. Animals were group housed to promote social interactions and environmental enrichment toys were provided at all times. The housing environment was maintained at a temperature of 17–27 °C and a relative humidity of 40–70%, with a 12-h light/dark cycle. Animals were acclimated to the vivarium and the staff members for ≥7 days prior to study start, enabling the animals to become familiar with their handler, which facilitated the treatment and anesthetic processes while minimizing stress for the animals.

On the day of the study, familiar handlers anesthetized the animals with isoflurane gas. Animals were occasionally immobilized in a sling to facilitate safe treatment, but were kept in the sling for no longer than 15 min. Animals were humanely euthanized by an intravenous injection of a sodium barbiturate cocktail only after the animal had been anesthetized by isoflurane gas.

Any deviations were recorded and assessed for impact in the final study reports. Data verification of raw source data files and statistical analysis was completed to ensure that the study reports accurately reflected the raw data and aligned with all statements, tables, and appendices. The reports and data were reviewed for proper content, documentation errors, and inclusion in the data files.

Methods

Animals

Yucatan miniature pigs (Sus scrofa domestica) were chosen as a model for SC delivery of test solutions with different concentrations and viscosities. All animals were more than 3 months of age and weighed approximately 10–27 kg. Animals were housed in steel pens with automatic water provided ad libitum and were fed twice daily (AM and PM), except on the PM prior to the study start. Animal body weights were recorded from the day of delivery to 1 day post-completion of the study to assess animal health. The room environment was set to maintain a temperature of ∼17–27 °C and a relative humidity of 40–70%, with a 12-h light/dark time cycle. All animals were acclimated to the facility for at least 48 h prior to the study start.

All animal experiments were conducted in full compliance with local, national, ethical, and regulatory principles and local licensing regulations under approved Institutional Animal Care and Use Committee protocols following the United States Department of Agriculture guidelines and regulations for research.

Test solutions

Human polyclonal Ig was chosen for the test solution as a non-specific, representative macromolecule and was investigated at concentrations encompassing the clinical concentrations of most approved monoclonal antibody therapies (Garidel et al., 2017). Ig solutions at varying concentrations and viscosities were prepared by reconstitution of a lyophilized polyclonal Ig product (BioMed Supply, Carlsbad, CA) in a saline-based buffer solution comprising 10 mM histidine and 56 mM sodium chloride at pH 6.5.

For each concentration of Ig (0 mg/mL, 20 mg/mL, 50 mg/mL, 100 mg/mL, 150 mg/mL, and 200 mg/mL), test solutions were prepared either with the Ig plus buffer solution alone or with the Ig solution co-mixed with 5000 U/mL rHuPH20. In addition, test solutions of Ig formulated at 0 mg/mL, 25 mg/mL, 100 mg/mL, 150 mg/mL, and 200 mg/mL were prepared either with or without 2000 U/mL rHuPH20. For both 0 mg/mL solutions, low-viscosity lactated Ringer’s (LR) solution (Baxter, Deerfield, IL) was used in place of Ig solution. All test solutions assessed are shown in Table 1. Test solutions were all equilibrated to ambient room temperature (20 ± 1.0 °C) prior to use.

Table 1. Formulations of test solutions administered subcutaneously to miniature pigs.

Ig concentration (mg/mL)	rHuPH20 concentration (U/mL)
0^†	5000
	0
20	5000
	0
50	5000
	0
100	5000
	0
150	5000
	0
200	5000
	0
0^†	2000
	0
25	2000
	0
100	2000
	0
150	2000
	0
200	2000
	0

^†0 mg/mL Ig solutions comprised of LR solution mixed with rHuPH20.

Ig: human immunoglobulin; LR: lactated Ringer’s; rHuPH20: recombinant human hyaluronidase PH20.

Infusion hardware and injection force measurement

A high-pressure syringe pump (KD Scientific, Holliston, MA) was used for all bench testing and SC infusions. Disposable 20 mL plastic syringes with a Luer-Lok™ tip (BD, Franklin Lakes, NJ) were connected to a Deltran-1 disposable in-line pressure transducer (Utah Medical Products, Midvale, UT) attached to a Surflo^® 23-, 25-, or 27-gauge infusion set (Terumo Medical Corporation, Somerset, NJ). The 23- and 25-gauge infusion sets utilized a ¾” (approximately 19 mm) thin-walled needle and 12” (approximately 305 mm) tubing, while the 27-gauge infusion set utilized a ½” (approximately 13 mm) thin-walled needle and 8” (approximately 203 mm) tubing. Only the 23-gauge needle was used for infusions in miniature pigs, while 23-, 25-, and 27-gauge needles were used for baseline measurements of applied force and in-line pressure when delivering solutions of different viscosities into air.

The instrumental set-up for the injection force and in-line pressure measurements is shown in Figure 1(A,B). To measure applied force, a high‑pressure infusion pump was used in combination with an RSB5 sub-miniature load cell force sensor placed in an adapter (0–100 N, accuracy ± 0.5%; Loadstar Sensors, Fremont, CA). The load cell was connected to a DI-100U interface (Loadstar Sensors, Fremont, CA) and force exerted on the syringe plunger was recorded by the load cell at a sampling rate of 1 Hz. For in-line pressure measurements, a disposable Deltran-1 in-line pressure transducer (Utah Medical Products Inc., Midvale, UT) was connected to a PowerLab 4/30 data acquisition system (AD Instruments, Colorado Springs, CO).

Figure 1. Instrumental set-up and injection sites. A high-pressure infusion pump (A) was used in combination with an RSB5 sub-miniature load cell force sensor (red box) (B) to record force exerted on the syringe plunger at a sampling rate of 1 Hz. Ig or LR solutions were administered on contralateral sides in Yucatan miniature pigs (C), either with (i) or without (ii) the addition of rHuPH20. Ig: human immunoglobulin; LR: lactated Ringer’s; rHuPH20: recombinant human hyaluronidase PH20.

The injection equipment was equilibrated to ambient room temperature (20 °C ± 1.0 °C) prior to use. In addition, the dynamic viscosity of each Ig solution was measured using a cone and plate rheometer (AR-1000 or Discovery HR-3 Rheometer; TA Instruments, New Castle, DE) at a temperature of 20 °C and a shear rate of 100 s⁻¹.

Study design: SC delivery in the miniature pig model

SC delivery of Ig or LR solutions with or without the addition of rHuPH20 was assessed in miniature pigs. On the day of study, animals were anesthetized with isoflurane gas and placed in dorsal recumbence on a heated surgical table. Animals were maintained under isoflurane gas for the entire duration of the procedure. Injection sites were located on the left and right regions of the lower abdomen of the animal (two sites/animal), approximately 3 cm lateral to the midline and approximately 6 cm cranially from the center of the inguinal fold (Figure 1(C)). The injection sites were marked with a permanent marker and then photographed using a standard digital camera (Canon PowerShot S120, Canon Inc., Tokyo, Japan). Injections on the left and right sides were performed sequentially for each animal.

Ig solutions of various concentrations and viscosities, and LR solution, with or without the addition of rHuPH20 at either 2000 or 5000 U/mL were delivered to the abdominal SC space at the indicated sites. The Ig and LR solutions were infused at flow rates ranging from 3–16 mL/min for solutions delivered with and without 2000 U/mL rHuPH20 (n = 6 miniature pigs/group at each concentration and flow rate). Flow rates ranged from 3–36 mL/min for solutions delivered with and without 5000 U/mL rHuPH20 (n ≤ 6 miniature pigs/group at each concentration and flow rate). Each procedure consisted of 1 infusion containing the Ig or LR solution alone, while the second infusion, on the contralateral side in the same animal, contained the Ig or LR solution plus rHuPH20 (2000 U/mL or 5000 U/mL). Due to the molecular weight of Ig exceeding that of macromolecules that can be directly absorbed, the solution entered systemic circulation via the lymphatic system after SC administration (Supersaxo et al., 1990). Applied injection force and in-line pressure measurements were obtained during the delivery of 10 mL of fluid for each infusion. Following the completion of injections, post-infusion local swelling height was measured using digital calipers and photographs of the injection area were captured.

In addition to measurements in miniature pigs, measurements into air were performed to establish the baseline force and in-line pressure required to deliver different types of fluid (distilled water [dH₂O], 100 mg/mL polyclonal Ig solution, or 120 mg/mL monoclonal antibody solution) through the hardware without any tissue back-pressure. The relationship between applied force and flow rate was assessed by delivering dH₂O and various concentrations of polyclonal Ig solution (100 mg/mL, 120 mg/mL, and 200 mg/mL) into air through a 23-gauge needle. In addition, the relationship between applied force and needle gauge was assessed by measuring the applied force while delivering a 100 mg/mL polyclonal Ig solution through different needle gauges (23-, 25-, and 27- gauge needles) into air.

Data analysis

LabChart 8 (AD instruments, Colorado Springs, CO) and SensorVue software (Loadstar Sensors, Fremont, CA) were used to analyze in-line pressure and applied force measurements, respectively. Statistical comparisons of in-line pressure and applied force between injections with and without rHuPH20 for different concentrations of Ig solution were performed using a paired t-test and considered significant if p ≤ 0.05.

Mean applied forces measured during infusions of Ig solutions at concentrations of 20–200 mg/mL and of LR solution (0 mg/mL Ig solution), either with or without the addition of 2000 U/mL rHuPH20 (n = 6/group) or 5000 U/mL rHuPH20 (n ≤ 6/group) were used to determine maximum flow rate (Q_max). Q_max was defined as the fastest flow rate that produced a statistically significant difference in mean applied force between an injection administered with versus without rHuPH20 for each concentration of Ig solution. SC delivery times were modeled for a given dose based on predicted Q_max values and volumes from 5000 U/mL rHuPH20 data using GraphPad Prism, v.9.0. (GraphPad Software LLC, San Diego, CA).

Results

Relationships between in-line pressure, applied force, and needle gauge

The relationship between in-line pressure and applied force was established by measuring both parameters while delivering different types of fluid (dH₂O, 100 mg/mL polyclonal Ig solution, or 120 mg/mL monoclonal antibody solution) through the infusion hardware into air at flow rates of 3–24 mL/min (dH₂O), 3–18 mL/min (100 mg/mL polyclonal Ig solution), and 3–12 mL/min (120 mg/mL monoclonal antibody solution) using a 23-gauge needle. These parameters represent the innate system pressure and the baseline minimum applied force required to deliver the water, polyclonal Ig, and monoclonal antibody solution due to the hardware configuration. A high degree of correlation was observed (R²=0.99; Figure 2(A)) between applied force and in-line pressure, suggesting that applied force may serve as a potential endpoint for monitoring SC infusions.

Figure 2. Baseline measurements of in-line pressure and applied force. (A) Correlation between in-line pressure and applied force when delivering dH₂O (flow rate 3–24 mL/min), 100 mg/mL polyclonal Ig solution (flow rate 3–18 mL/min), or 120 mg/mL monoclonal antibody solution (flow rate 3–12 mL/min) into air using a 23-gauge needle. (B) Applied force for delivering polyclonal Ig solutions of 100, 120, and 200 mg/mL, or dH₂O using a 23-gauge needle at various flow rates. dH₂O: distilled water; Ig: immunoglobulin; SD: standard deviation.

The relationship between applied force and needle gauge was examined using 100 mg/mL polyclonal Ig solutions (Supplementary Figure 1). For all flow rates, there was a strong inverse correlation between the inner needle diameter and applied force. Therefore, needles with larger inner diameters (lower needle gauge) require less applied force during injection and allow for substantively faster flow rates (e.g. the increase in inner diameter from a 27- to 23-gauge needle allows for 10 mL/min flow rate to be delivered with approximately one sixth the amount of applied force, a reduction from 63.9 N to 12.1 N, respectively).

In addition, applied force was higher for more concentrated polyclonal Ig solutions, as observed with a 23-gauge needle at a range of concentrations (100 mg/mL, 120 mg/mL, and 200 mg/mL polyclonal Ig solutions, and dH₂O) and flow rates (1–24 mL/min) (Figure 2(B)). With a 23-gauge needle, the highest applied force was 39.7 N, measured when a 200 mg/mL polyclonal Ig solution was administered at a flow rate of 9 mL/min, and the lowest applied force was 2.4 N, measured when dH₂O was administered at a flow rate of 3 mL/min.

Effect of rHuPH20 on in-line pressure and applied force required for SC delivery in the miniature pig model

We tested whether the addition of 5000 U/mL rHuPH20 to an Ig solution could reduce the in-line pressure and applied force required for SC infusion in miniature pigs. A 100 mg/mL Ig solution was infused subcutaneously at flow rates of 5, 12, and 15 mL/min either with or without the addition of 5000 U/mL rHuPH20 (n ≤ 6/group; Figure 3(A–E)).

Figure 3. In-line pressure and applied force during SC delivery of 100 mg/mL Ig solution with or without the addition of 5000 U/mL rHuPH20 in the miniature pig model. (A) Mean in-line pressure at a flow rate of 5 mL/min. Mean applied force at flow rates of (B) 5 mL/min, (C) 12 mL/min, and (D) 15 mL/min. Steady-state mean applied force during delivery of 100 mg/mL Ig solution administered with or without rHuPH20, and baseline infusions at various flow rates (E). Solid lines represent the mean applied force for each test solution and dotted lines represent the SEM. Ig: immunoglobulin; rHuPH20: recombinant human hyaluronidase PH20; SC: subcutaneous; SEM: standard error of the mean.

At a flow rate of 5 mL/min, there was a significant decrease in both the mean in-line pressure (p < 0.05; Figure 3(A)) and mean applied force (p < 0.05; Figure 3(B)) required for SC delivery of the Ig solution with 5000 U/mL rHuPH20 versus the Ig solution without rHuPH20. A significant decrease in mean applied force was also observed for flow rates of 12 mL/min (p < 0.0001; Figure 3(C)) and 15 mL/min (p < 0.0001; Figure 3(D)). A similar decrease in mean applied force was observed for all other concentrations of Ig solution (20 mg/mL, 50 mg/mL, 150 mg/mL, and 200 mg/mL) and LR solution with the addition of 5000 U/mL rHuPH20 (data not shown). Steady-state mean applied forces during delivery of 100 mg/mL Ig solution administered with or without rHuPH20 at flow rates of 5, 12, and 15 mL/min are shown in Figure 3(E).

As in experiments with 5000 U/mL rHuPH20, the addition of 2000 U/mL rHuPH20 to Ig solutions at concentrations ranging from 25–200 mg/mL (n = 6/group) also resulted in reductions in applied forces in comparison with Ig solutions without rHuPH20 (Figure 4). Percentage reductions in applied force in the presence of 2000 U/mL rHuPH20 versus test solutions without rHuPH20 ranged from 11.5% to 47.3% for all solutions (p < 0.05 for all solutions with the exception of the 100 mg/mL Ig solution). A significant reduction (p < 0.05) in applied force was also observed with the addition of 2000 U/mL rHuPH20 to LR solution, in comparison with LR solution alone.

Figure 4. Applied force at flow rates of 3–16 mL/min during SC delivery of LR solution or 25–200 mg/mL Ig solutions with or without 2000 U/mL rHuPH20 in the miniature pig model. Mean applied force when administering (A) LR solution at a 16 mL/min flow rate, (B) Ig 25 mg/mL at a 12 mL/min flow rate, (C) Ig 100 mg/mL at a 5 mL/min flow rate, (D) Ig 150 mg/mL at a 4 mL/min flow rate, and (E) Ig 200 mg/mL at a 3 mL/min flow rate. Solid lines represent the mean applied force for each test solution and dotted lines represent the SEM. Baseline represents mean applied force obtained when delivering LR solution and Ig solutions into air. p-values represent statistical significance of difference in mean applied force between solutions administered with and without rHuPH20. ^†Percentage reduction in applied force in the presence of rHuPH20. Ig: immunoglobulin; LR: lactated Ringer’s; NS: not significant; rHuPH20: recombinant human hyaluronidase PH20; SC: subcutaneous; SEM: standard error of the mean.

Qualitative comparison of abdominal injection sites in miniature pigs showed that the addition of rHuPH20 to Ig solutions resulted in increased dispersion at the injection site (Figure 5). Quantitative assessment of local swelling at the abdominal injection sites post-infusion demonstrated that injections facilitated by the addition of 2000 U/mL rHuPH20 had significantly lower mean swelling heights in comparison with injections administered without rHuPH20 for all concentrations of Ig solution (Figure 6).

Figure 5. Injection site pre- and post-10 mL total infusion volume with or without the addition of 2000 U/mL rHuPH20 in miniature pigs. Ig: immunoglobulin; rHuPH20: recombinant human hyaluronidase PH20.

Figure 6. Mean post-infusion swelling height in miniature pigs following administration of 10 mL total infusion volume of Ig solutions with and without 2000 U/mL rHuPH20. ^†0 mg/mL Ig solutions comprised LR solution mixed with or without rHuPH20. Statistical significance of difference between swelling heights with and without rHuPH20: ***p < 0.001, **p < 0.01, *p < 0.05. Ig: immunoglobulin; LR: lactated Ringer’s; rHuPH20: recombinant human hyaluronidase PH20; SEM: standard error of the mean.

Modeling maximal flow rate (Q_max) across a range of Ig solution concentrations

Q_max was determined for a range of concentrations of Ig solutions (20–200 mg/mL), and for LR solution (0 mg/mL Ig solution) with 2000 U/mL and 5000 U/mL rHuPH20 (Table 2). Q_max is the flow rate that produced a statistically significant difference in mean applied force between an injection administered with rHuPH20 versus an injection administered without the addition of rHuPH20 for each concentration of Ig solution. With both concentrations of rHuPH20 (2000 and 5000 U/mL), Q_max was inversely proportional to viscosity. Q_max for the 5000 U/mL rHuPH20 concentration was approximately double that of the 2000 U/mL rHuPH20 concentration (Figure 7). This mathematical relationship between viscosity and Q_max allowed potential flow rates to be predicted for Ig solutions of various viscosities.

Figure 7. Maximal flow rate versus viscosity of Ig solutions administered subcutaneously with 2000 or 5000 U/mL rHuPH20. cP: centipoise; Ig: immunoglobulin; LR: lactated Ringer’s; Q_max: maximum flow/delivery rate; rHuPH20: recombinant human hyaluronidase PH20.

Table 2. Viscosity and maximal flow rate during SC administration of various concentrations of Ig solution with 2000 or 5000 U/mL rHuPH20, using a 23-gauge needle in a miniature pig model.

rHuPH20 concentration (U/mL)	Ig concentration (mg/mL)	Viscosity (cP)	Qmax
2000	0^†	1.05	16
	25	1.27	12
	100	2.92	6
	150	6.76	4
	200	19.63	2
5000	0^†	1.09	36
	20	1.35	27
	50	1.90	21
	100	3.69	15
	150	9.51	7.5
	200	19.09	4

^†0 mg/mL Ig solutions comprised LR solution mixed with rHuPH20.

cP: centipoise; Ig: human immunoglobulin; LR: lactated Ringer’s; Q_max: maximal flow/delivery rate; rHuPH20: recombinant human hyaluronidase PH20; SC: subcutaneous.

Using the in silico model developed with 5000 U/mL rHuPH20 data, a simulation of a hypothetical dose of 800 mg Ig with rHuPH20 showed that the shortest delivery times were achieved across a broad range of concentrations, when extremes in both viscosity/concentration and volume were avoided. Notably, the shortest delivery times (<30 s) did not require the most concentrated drugs but were achieved across a broad range of concentrations 62–140 mg/mL (Figure 8).

Figure 8. Q_max modeling to estimate delivery times and optimal range of concentrations using a modeled dose of 800 mg Ig with 5000 U/mL rHuPH20.

Ig: immunoglobulin; Q_max: maximum flow/delivery rate; rHuPH20: recombinant human hyaluronidase PH20.

Discussion

We used a miniature pig model to evaluate the mean in-line pressures and applied forces required for SC delivery of Ig solutions at concentrations of 20–200 mg/mL with or without the addition of 2000 or 5000 U/mL rHuPH20. Baseline measurements, established by delivering monoclonal antibody solutions, polyclonal Ig solutions, or dH₂O into air, demonstrated a high degree of correlation between in-line pressure and applied force, suggesting that applied force may serve as an endpoint for monitoring high-speed, high-volume SC infusions. In addition, we have demonstrated that an increase in needle diameter may result in significant decreases in applied force, allowing for higher-speed, larger-volume SC injections at increased needle diameters.

Ig solutions co-mixed with different concentrations of rHuPH20 showed an rHuPH20 dose-dependent reduction in applied injection forces during SC administration, allowing for a higher Q_max to be achieved with increased concentrations of rHuPH20. Increased dispersion and a significantly reduced mean swelling height were observed at the injection site with the addition of rHuPH20 in comparison with injections without rHuPH20. These results suggest that rHuPH20 can improve the performance of delivery devices by degrading HA in the extracellular matrix of the SC space, reducing the SC tissue constraints to bulk fluid flow that otherwise impede the device hardware, thereby allowing for higher delivery rates required for large-volume SC administration (Bookbinder et al., 2006; Kang et al., 2012). Similar results were observed in a recent study of miniature pigs, in which the addition of rHuPH20 to test solutions administered subcutaneously resulted in shorter injection times and reduced swelling at the injection site with some auto-injector devices (Shi et al., 2021).

Historically, rapid SC administration has been limited to volumes of <3.5 mL owing, in part, to constraints imposed by HA resistance to bulk fluid flow in the extracellular matrix (Frost 2007; Hunter 2008; McDonald et al., 2010; Dychter et al., 2012; Locke et al., 2019; Amgen Inc. 2021). One method to increase SC dispersion and absorption is to use rHuPH20 to locally and transiently degrade HA in the SC space (Bookbinder et al., 2006; Frost 2007; Kang et al., 2012; Locke et al., 2019). This study suggests that reducing SC tissue resistance by co-formulating therapeutics with rHuPH20 could help to minimize constraints on SC delivery and may allow for the future design of medical devices, such as auto-injectors, patch-pumps, and off-body injectors, that permit faster SC injections of macromolecules at greater volumes (Bittner et al., 2018; Shi et al., 2021).

The development of such large-volume high-flow rate devices could be particularly beneficial in an oncology setting, where treatment with monoclonal antibodies typically requires administration of large volumes. The range of Ig concentrations assessed in this study, from 20–200 mg/mL, encompasses the clinical concentrations of most approved monoclonal antibody therapies (Garidel et al., 2017), suggesting that the advantages of co-formulation could be applied to a broad range of therapeutics. Several monoclonal antibodies, including trastuzumab, rituximab, daratumumab, and a combination of pertuzumab and trastuzumab, have already received approval for co-formulation with rHuPH20 (Genentech Inc., 2020, 2021; U.S. Food and Drug Administration, 2019, 2020a) and have been used to treat more than 400,000 patients worldwide (Bittner et al., 2022).

Further, the recent COVID-19 pandemic has highlighted the need for administration options outside of a hospital or clinic, and has resulted in increased interest in such treatments (Hanna et al., 2021). Limiting exposure to COVID-19 and other infections is of particular importance in patients with cancer, inflammatory, or autoimmune diseases. Treatment of these patients often requires large-volume administration, and the therapies themselves may weaken the immune response, increasing the risk of poor outcomes following COVID-19 infection (Hanna et al., 2021).

Home administration delivered by a healthcare professional can reduce the possibility of patient exposure to COVID-19, and has been shown to be effective and preferred by patients over hospital or clinic-based treatment (Polinski et al., 2017; Hanna et al., 2021). Home administration can result in improved patient quality of life, reduced healthcare costs, and increased scheduling flexibility, allowing patients to receive treatment at more convenient times and resulting in a positive impact on adherence (Wasserman 2014; Polinski et al., 2017; Finnigan et al., 2018). Self-administration of SC-formulated therapies is also associated with many of these benefits, permitting increased flexibility in time and place of injection administration, and increased patient autonomy in managing their treatment, as well as reduced cost and burden to caregivers (van den Bemt et al., 2019).

Historical limits on large-volume, rapid delivery of SC formulations have presented a barrier to more widespread home- and self-administration, but the development of rapid-delivery, large-volume SC devices using rHuPH20 may facilitate the shift of an increasing number of high-dose biologic therapeutics from hospital-based IV administration to SC delivery in a home setting. Rapid large-volume SC delivery facilitated by rHuPH20 already permits self-administration, or home-administration by a healthcare professional, for FDA-approved therapeutics for the treatment of patients with primary immunodeficiency disease and HER2-positive early- and metastatic breast cancer (Genentech Inc., 2020; U.S. Food and Drug Administration, 2020b; U.S. Food and Drug Administration, 2014).

In addition to demonstrating the feasibility of using rHuPH20 to reduce constraints in SC delivery, mathematical modeling showed that favorable SC delivery times (<30 s) were achieved with a broad range of Ig concentrations, when the balance between drug concentration and delivery volume was optimized using rHuPH20. The shortest delivery times did not require the most concentrated drugs, allowing for flexibility in identifying the optimal concentration/viscosity of each therapeutic when developing drugs co-formulated with rHuPH20 for SC delivery.

Limitations of this investigation include a few anatomical and physiological differences between human and miniature pig skin. The SC fat layer in miniature pigs differs in thickness from humans, although this may vary between strains of miniature pigs and between individuals; human skin has a pH of approximately 5, in comparison with pH 6–7 in miniature pigs; and there are slight differences in vascularization (Mahl et al., 2006). In the 5000 U/mL rHuPH20 group, multiple infusion cycles were administered to individual animals; it is possible that this could have stretched the skin and affected the measured parameters in an undetermined manner in this group. Nevertheless, the skin of miniature pigs is considered to be generally translatable to humans and a suitable clinical model (Mahl et al., 2006).

In conclusion, the addition of rHuPH20 to different concentrations of Ig formulation resulted in a significant reduction in mean applied forces required for SC delivery versus the formulations without rHuPH20 in Yucatan miniature pigs. The shortest SC delivery times were achieved across a broad range of solution concentrations in an in silico simulation of a hypothetical dose of Ig solution co-formulated with rHuPH20. By co-formulating solutions of macromolecules with rHuPH20, high-volume SC infusions can be delivered rapidly across a broad range of concentrations. These results demonstrate that co-formulating rHuPH20 with macromolecules can help overcome the volume constraints associated with SC delivery across a range of concentrations commonly used for biologic therapeutics. This provides potentially valuable information for the future design of devices for administering rapid large-volume SC injections with rHuPH20, especially those intended for the home setting.

Author contribution

RJC was responsible for the conception, design, and execution of the study, as well as for interpreting the data, drafting the paper, revising it critically for intellectual content, and providing final approval of the final draft for publication.

RC was responsible for the conception, design, and execution of the study, as well as for interpreting the data, drafting the paper, revising it critically for intellectual content, and providing final approval of the final draft for publication.

DWK was responsible for the conception, design, and execution of the study, as well as for interpreting the data, drafting the paper, revising it critically for intellectual content, and providing final approval of the final draft for publication.

All authors agree to be accountable for all aspects of the work.

Acknowledgments

The authors would like to thank all investigators involved in this study, and Drs Anna-Maria Hays Putnam, Michael J LaBarre, and Stephen Knowles for reviewing the manuscript. Medical writing support, including assisting authors with the development of the outline and initial draft, and incorporation of comments, was provided by Rachel O’Meara, PhD, and editorial support was provided by Sarah Christopher, PhD, of Paragon, Knutsford, UK, supported by Halozyme Therapeutics, Inc. Halozyme follows all current policies established by the International Committee of Medical Journal Editors and Good Publication Practice guidelines (https://www.acpjournals.org/doi/10.7326/M22-1460). The sponsor was involved in the study design and collection, analysis, and interpretation of data, as well as data verification of information provided in the manuscript. However, ultimate responsibility for opinions, conclusions, and data interpretation lies with the authors.

Disclosure statement

RJC and DWK are employees of Halozyme Therapeutics, Inc., and hold shares in the company. RC is a former employee of Halozyme Therapeutics, Inc.

Data sharing statement

Halozyme Therapeutics, Inc. follows policies established by the International Committee of Medical Journal Editors and Good Publication Practice guidelines (link). The studies were conducted by Halozyme Therapeutics, Inc. and the data are held by the company. The datasets supporting the results reported in this article may be provided to researchers upon reasonable request. Such requests can be made by contacting Halozyme Therapeutics, Inc.: 12390 El Camino Real, San Diego, CA 92130, USA; Phone: [phone]; Email: [email protected].

Additional information

Funding

Development of this manuscript was supported by Halozyme Therapeutics, Inc.