**1. Introduction**

Norvir® (ritonavir) is an inhibitor of human immunodeficiency virus (HIV) protease and is indicated in combination with other antiretroviral (ARV) agents for the treatment of HIV-1 infection [1,2]. Due to the CYP3A inhibitory capabilities of ritonavir, it is also co-administered at lower doses as a pharmacokinetic (PK) enhancer to increase exposures of other HIV protease inhibitors (PIs) [3]. When used as a PK enhancer, ritonavir is most commonly administered at 100 to 200 mg once or twice daily [4]. As a PK enhancer, ritonavir has become a mainstay in the managemen<sup>t</sup> of both treatment-naïve and treatment-experienced patients and is typically no longer prescribed as a sole protease inhibitor in antiretroviral regimens today [5–7]. The PK enhancement often allows for a reduction of pill burden, dosing frequency, and food restrictions, while maintaining efficacy [6,8].

Ritonavir is practically insoluble in water (~5 μg/mL), however, this solubility can be enhanced to 1.2 mg/mL at approximately pH 1 (0.1 N hydrochloric acid solution) [9]. Due to the lack of aqueous solubility, ritonavir showed essentially no bioavailability in an animal model (dog) when administered as an unformulated solid in a capsule [9]. Attempts to enhance the oral bioavailability of ritonavir from a solid dosage form by incorporation of surfactants, acids, and other wetting agents failed to increase the bioavailability to greater than ~4%. Similar results were obtained by substituting ritonavir base with salt derivatives of ritonavir. However, bioavailability of 37% in dogs was achieved with a

solution formulation containing 5.0 mg ritonavir per mL in a solvent system consisting of 20:30:50 ethanol:propylene glycol:water [2]. Therefore, early formulation development of ritonavir was focused on oral solutions (OS) because of the poor oral bioavailability observed with solid oral dosage forms and the inability to find a solvent system that was compatible with hard or soft gelatin capsules.

OS development efforts concentrated on increasing the drug loading while maintaining oral bioavailability. It was concluded that: a) The relative bioavailability of ritonavir in liquid formulations is inversely proportional to the drug concentration and b) the presence of a surfactant, such as Cremophor EL, enhances the bioavailability of ritonavir in liquid formulations. OS also has a short shelf-life of six months at ambient storage to ensure ritonavir remains in solution (physical stability) while maintaining acceptable levels of degradation (chemical stability).

In addition to bioavailability and stability challenges, OS is known to possess multiple aversive sensory attributes: Basic taste, aromas, trigeminal irritation, and textures, collectively known as "flavor". These include bitterness and aromatic off-notes from the active pharmaceutical ingredient (API) and trigeminal irritation from the solvents. To address these aversive attributes, an identifying flavor system comprised of sweeteners and aromas (peppermint and caramel) was developed.

In a preliminary study, the sweetened/flavored formulation somewhat reduced aversive sensory attributes, but overall, the formulation remained relatively low in palatability. Five dose-administration approaches to further ameliorate the sensory effects of OS were subsequently evaluated, first by a trained adult sensory panel using the flavor profile method to identify the most promising one(s). The results are summarized in Table 1. Chasing liquid dose administration with foods was identified as the most promising approach. Six products were selected for confirmation testing with patients based on differences in form, flavor strength, and mastication characteristics: SnackWells® Fudge Brownie, Freshen-Up® Peppermint Gum, Toast Crackers with Peanut Butter, Oats'n Honey Crunch® Granola Bar, Riesen® Chocolate Chews, and Nutella® Spread. Using a 7-point intensity scale, 74 OS patients rated the strength of the "medicine flavor" remaining in the mouth two minutes after taking the OS (mean score 4.95) and chasing with these six foods. All of the food products tested reduced the intensity of the aversive flavor compared to the OS alone, with mean "medicine flavor" ranging from a score of 4.18 to 1.50. An ideal formulation would not require external vehicles, but when faced with significant formulation technical challenges, this chaser approach was effective [10].


**Table 1.** Results of approaches to ameliorate the flavor impact of Norvir® 80 mg/mL oral solution.

The development of an amorphous solid dispersion (ASD) formulation for ritonavir was initiated following the introduction of the Kaletra® (lopinavir/ritonavir) tablet in 2005. Experience gained with solid dispersion technology enabled the successful development of the Norvir® 100 mg tablet containing ritonavir ASD which achieved the desired bioavailability and acceptable ambient chemical and physical stability [11]. The Norvir® 100 mg tablet provides significant benefits to patients and physicians, primarily through non-refrigerated storage compared to the OS, offering more robust stability required for storage under global climatic conditions [12]. However, the need for a liquid formulation remains for pediatric patients and adults who are unable to swallow the tablets. To address

this gap, AbbVie developed a new powder formulation, Norvir® oral powder, 100 mg (NOP), to provide a more suitable formulation for the pediatric population and with the intent to replace the marketed OS. The objective of this study is to provide an overview of the development and palatability assessment of the age appropriate NOP.

### *Development Overview of NOP*

The NOP development program was initiated, primarily to mitigate or eliminate the risk of potential toxicities associated with ethanol and propylene glycol solvents in OS, which contains 43.2% (*v/v*) ethanol and 26.0% (*v/v*) propylene glycol [4,13]. Other excipients such as colorants, flavoring agents, and preservatives found in the OS were also removed from NOP. NOP also facilitates dose preparation and administration, aligned with current and future needs for pediatric patients.

NOP is manufactured by milling the ritonavir ASD extrudate intermediate and filling the resulting powder into sachets. For dose preparation, NOP is suspended in liquid vehicles or sprinkled on soft foods. The dispersion of NOP produces a supersaturated aqueous solution of ritonavir drug substance that maintains the bioavailability achieved with OS [14].

Key design targets for the development of the NOP are:


Early formulation development efforts evaluated both uncoated and coated powders. To mask the inherent bitter taste of ritonavir, a methacrylic acid–ethyl acrylate copolymer coating (enteric coating, insoluble in acidic media, and soluble above pH 5.5 to allow dissolution in the intestine) was applied to the uncoated powder. Exposure to various pH environments during dose preparation and administration was taken into consideration when assessing impact on the coated powder and potential for drug release and associated aversive attributes as evaluated in healthy adult volunteers. When exposed to various pHs, the uncoated powder provided a more homogeneous suspension compared to the coated powder. This is an important feature to ensure complete dose administration of the amorphous drug suspension for pediatric doses. The uncoated powder formulation achieves the majority of the key design targets, but still had opportunity to improve palatability for dose administration.

The coated and uncoated powders were evaluated for ritonavir pharmacokinetics and palatability as an aqueous suspension administered within approximately 5 min after suspending. In order to evaluate the potential impact of the dissolved aqueous soluble excipients on the ritonavir pharmacokinetics and palatability, uncoated powder was pre-dispersed and held for approximately 30 min prior to administration to allow dissolution of soluble excipients and suspension of the amorphous drug particles prior to administration. The pharmacokinetic results of these three ritonavir regimens (coated, uncoated, and uncoated pre-dispersed powders) showed comparable bioavailability. The bioavailability for all three regimens ranged from 80%–90% relative to OS. The palatability results from this study indicated that no taste-masking benefit was gained from coated powder formulation as compared to the uncoated powder and OS.

The initial milled ritonavir ASD extrudate intermediate had a rather broad particle size distribution with a significant fraction of fine particles (<100 μm) and a particle shape that is not ideal for polymeric coating. Coating for conventional tablets, round pellets, or mini-tablets can be effective in reducing or eliminating API taste, resulting in "taste-neutral" formulations [15–17]. However, coating for irregularly shaped particulates/granules often results in an imperfect barrier film coat, exposing API in the oral cavity where it can be perceived. The milling and coating processes were not optimized, and it is possible that the finer irregular particles may have had incomplete coating with additional ritonavir particles embedded in the outer polymeric coating layer (Figure 1) leading to potential premature dissolution or exposure to taste receptors during administration.

**Figure 1.** Ritonavir particle shape and small ritonavir particles embedded in eudragit coating. (Orange arrows point to the ritonavir particles).

To further investigate the pharmacokinetic behavior of a larger particle size, still meeting the recommended size for sprinkle products, a uniform 2-mm ritonavir extrudate particulate was coated with the methacrylic acid–ethyl acrylate copolymer (Figure 2) [18]. Although the coated particulate showed improvements in flavor as measured by a trained adult sensory panel, the pharmacokinetic results showed a reduced relative bioavailability of approximately 50% for the coated particulate relative to the OS. The delayed dissolution and drug release profile had a negative impact relative to the known narrow absorption window of ritonavir in the upper intestine. Given the challenges of achieving sufficient coating for taste-masking purposes, without negatively impacting bioavailability, development of NOP was focused on using uncoated powder.

**Figure 2.** Coated ritonavir extrudate 2-mm particulate.

### **2. Results and Discussion**

### *2.1. NOP Palatability Assessment, Part 1: Dose/Response Sensory Analysis*

All five strengths of NOP aqueous suspension were characterized by a strong and lingering bitter basic taste and secondary aromatic off-notes described as "burnt" (polyethylene, wax, and hair). As shown in Figure 3 bitterness of the five strengths spanned the upper half of the flavor profile supra-threshold intensity range, the most challenging from a taste-masking perspective. The bitterness of NOP is very strong at the 100mg clinically relevant dose and all strengths would be patient-perceptible (≥1 intensity) initially and for varying lengths of time in the aftertaste. In the absence of a taste-masking system, it would be necessary to reduce NOP strength to about 0.3 mg/10 mL in order for the bitterness to be imperceptible to patients (i.e., <1 intensity).

**Figure 3.** Bitterness dose/response of Norvir® oral powder (NOP).

The dose/response results for the burnt aromatics are shown in Figure 4. The four highest strengths of NOP (10, 25, 50, and 100 mg/10 mL) had moderate-to-strong intensity burnt aromatic off-notes. In the absence of an appropriate aroma masking system, it would be necessary to reduce NOP strength to about 1 mg/10 mL for the burnt aromatics to unrecognizable to patients.

**Figure 4.** Burnt aromatics dose/response of NOP.

### *2.2. NOP Palatability Assessment, Part 2: Evaluation with Foods*

As shown in Figure 5, and consistent with available information, the foods produced varied effects on bitterness reduction, ranging from <12 -unit (sweet potato chaser and chocolate milk mix-in) to a 112 -unit reduction in perception of bitterness being associated with food chasers high in fat and/or sugar content with strong flavor intensity (peanut butter, Nutella®). The 100-mg clinically relevant dose of NOP is shown for comparison.

**Figure 5.** Bitterness profile for NOP with food chasers.

The foods also had varied effects on the burnt aromatic off-notes of the NOP (Figure 6). All of the chasers produced greater reduction in the burnt aromatics than the chocolate milk mix-in (minimal reduction) with several at or below the threshold for perception (≤1). The poor performance of chocolate milk may have been due to its administration as a mix-in, which would have allowed for an extended time for hydration, potentially releasing volatile aromatics.

**Figure 6.** Burnt aromatics profile for NOP with food chasers.

During the bioavailability studies, palatability of NOP in various beverages (water, infant formula, and chocolate milk) as well as admixed into a soft food (applesauce or vanilla pudding) was compared to that of the OS through study participant questionnaires. Only the NOP mixed in chocolate milk showed a modest improvement in overall palatability compared to the OS. The bioavailability study also demonstrated that administration of NOP with infant formula, chocolate milk, applesauce, or pudding was bioequivalent to administration in water [19]. This observation suggests that NOP may be administered with a wide variety of vehicles (soft foods and liquids) with limited impact on bioavailability.

Based on these results, the foods were ranked in descending order of their overall ability to reduce the aversive sensory attributes of the drug product when administered as chasers. The composition and physical characteristics of foods can be important determinants of their ability to "mask" the aversive sensory attributes of drug products. Table 2 summarizes the fat, sugar content, and water activity of the model foods. Water activity is a measure of relative vapor pressure of water molecules in the headspace above a food versus vapor pressure above pure water from 0 to 1 (pure water). This type of food science-based categorization system will ensure that the most varied food types are tested with a minimum of overlap between samples. The best performing foods may then be screened for chemical compatibility, resulting in specific dosing recommendations that are both effective at reducing the aversive sensory attributes of the drug product and efficacious.

**Table 2.** Composition and physical characteristics of model foods selected as dosing vehicle or chaser.


### **3. Materials and Methods**
