*3.2. Methods*

Samples were evaluated using the flavor profile method of sensory analysis, an internationally recognized and approved open-source method [20]. Flavor profile is used to identify, characterize, and quantify the sensory attributes of the samples. Flavor profile measures the perceived strength or intensity of the attributes, the order in which they appear; and a description of all flavors—basic tastes, feeling factors, and aromatics—remaining at specified time intervals in the aftertaste. Per the methodology, 4–6 panelists evaluated each sample and collectively arrived at a consensus judgement of the attribute intensity using chemical reference standards to establish the intensity scale. Sensory characteristics above a slight intensity on the flavor profile scale (>1) are clearly perceptible to consumers/patients; this intensity is known as the recognition threshold. Therefore, in order not to be perceived, negative attributes (e.g., bitterness or irritation) should be below this threshold. The flavor profile terms are as shown in Figure 7.

### **Figure 7.** Flavor profile definitions.

A two-part study was conducted with healthy adult sensory panelists (subjects). Part 1 was a dose/response sensory analysis of NOP to determine the maximum concentration that is patient perceptible. Part 2 was an assessment of the sensory performance of NOP when administered with model foods. The study was conducted in accordance with good clinical practice.

Part I: Dose/Response Sensory Analysis of NOP

The NOP was prepared by milling ritonavir extrudate intermediate material. To characterize NOP across a range of concentrations, five strengths (1, 10, 25, 50, and 100 mg/10 mL) were evaluated, the highest being clinically relevant.

A 10-mL aliquot of sample was dispensed into individual 1-ounce plastic cups using a graduated oral syringe and distributed to each panelist. Starting at the same time, each panelist poured the sample into their mouth, swished the contents around the oral cavity for 10 s, and expectorated. The panelists independently evaluated and recorded the sensory characteristics at nine discrete time points (0, 1, 3, 5, 10, 15, 20, 25, and 30 min) using the flavor profile method. Each sample was evaluated twice to generate the final flavor profile. Multiple sessions were required to complete the evaluations to limit exposure to the drug active and minimize sensory fatigue.

Part 2: Evaluation of NOP with Model Foods

There are several approaches that should be considered when improving the "palatability" of drug products with extremely aversive sensory attributes, e.g., bitterness. For NOP, taste-masking was not achievable via a formulation strategy, so two potential alternative approaches were considered, including dosing with food (the term "food" also includes beverages) and cleansing the palate with a food immediately following dose administration (using a food "chaser") [21]. The principle of food "chasers", consumption of a food immediately following dose administration, was explored based on historical experience evaluating various approaches with common food products to reduce the bitterness intensity with OS.

The selection of food products is based on multiple factors including patient age, ease of preparation, and availability. Choices for infants and toddlers often include applesauce, yogurt, and formula [22]. However, limited consideration is given to the composition and chemical properties of the foods, e.g., fat, sugar moisture content, pH, water activity, flavor impact (intensity and duration). Seven products were selected in part based on their compositional diversity from a food science perspective as well as availability and fit with the target demographics and geography (e.g., pediatrics in Africa). These products were peanut butter, hazelnut spread, black currant concentrate, golden syrup, chocolate syrup, sweet potato puree, and chocolate milk. Nutrient data was determined from the commercial nutrition facts labels and water activity was measured on an AquaLab Model 4TEV water activity meter.

The commercial food products were evaluated as dosing vehicles for NOP ("mix-in") or following administration of the NOP ("chaser") to help cleanse the palate. The sensory panelists first evaluated each of the model food products alone (i.e., without NOP) to develop a flavor profile of the native product.

For dosing with model liquid food (mix-in), the panelists were provided a sachet containing NOP and a 1-oz sample cup containing 10 mL of the model liquid food. Starting at the same time, panelists emptied the sachet in to the model food and mixed. After mixing, panelists took the sample into the mouth and agitated for 10 s and expectorated the liquid. The panelists independently evaluated and recorded the sensory characteristics at nine discrete time points (0, 1, 3, 5, 10, 15, 20, 25, and 30 min) using the flavor profile method. The process was repeated for a second evaluation of each sample to generate the final flavor profile for the sample. Multiple sessions were required to complete the evaluations to limit exposure to the drug active and minimize sensory fatigue.

For evaluating the effects of "chasers", the panelists were provided a sachet containing NOP, a 1-oz sample cup containing 10 mL water, and a disposable spoon containing 5 g of the tested food chaser. Starting at the same time, panelists emptied the sachet into the water and mixed, forming a suspension. After mixing, starting at the same time, the panelists sipped the NOP suspension into the oral cavity, agitated in the mouth for 10 s, and expectorated the liquid. Immediately following expectoration, panelists took the food chaser from the spoon into the oral cavity, agitated in the mouth for 10 s, and swallowed the food chaser. The panelists independently evaluated and recorded the sensory characteristics at the same nine time intervals using the flavor profile method. As before, the process was repeated for a second evaluation of each sample to generate the final flavor profile for the sample.
