3.1.1. Dip Coating Parameters Optimization

In the dipping process, the amount of drug coated onto a device can be optimized by balancing the viscosity of the drug solution, the incubation and drying times, and the number of dipping/drying iterations [54,55]. A comprehensive study was carried out to investigate the aforementioned effects on the amount of fluticasone loaded onto the string. Fluticasone (FTS)-loaded strings were designed to cover the entire human esophagus length (~25 cm) and release a clinically relevant dose of FTS (1–2 mg) in 24 h [54]. Initial optimization studies were accomplished with a commercially available biocompatible PCL string. The coating solution was comprised of FTS in a biocompatible polymer and organic solvent. The solvent used to load FTS onto the PCL string using a dip-coating process was selected based on high solubility of FTS in the solvent, its volatility, and ease of removal post drug loading. A number of solvents were screened, and acetone was selected as a dipping solvent based on its low boiling point and high saturation solubility of FTS (~20 mg/mL). To optimize the dip-coating process, poly(lactide-*co*-glycolide) (PLGA, 50:50) of various molecular weights (MW 11, 27, and 53 kDa) was used as a model to investigate the effect of dipping solution viscosity on percent PLGA loaded onto the string (Supplementary Figure S1D). Results from these experiments were used to extrapolate the amount of FTS that can be loaded on the strings using the drug-loaded dipping solution and minimize the use of an expensive drug for preliminary optimization studies.

A series of control experiments were performed to evaluate the effect of polymer molecular weight (MW) on the mass of polymer loaded onto strings under the same dipping conditions. The prototype PCL strings (3 mm OD, 20 mm L, *n* = 3), were pre-weighed (M<sup>i</sup> ) before the dip coating process. The dipping conditions involved incubating PCL strings in the dipping solution for 1 min followed by drying at RT for 24 h. After drying, the strings were weighed again (M<sup>f</sup> ) and mass of polymer loaded was calculated as the difference in mass (M<sup>f</sup> − M<sup>i</sup> ) (Supplementary Figure S1D). Results showed that the amounts of PLGA loaded onto the string increased with increasing the dipping solution viscosity. The viscos-

ity of the dipping solution significantly increased with increasing PLGA MW from 1.4 cP (PLGA MW 11 kDa) to 15.7 cP (PLGA MW 53 kDa). This increase in viscosity resulted in a 2.67-fold increase in PLGA loading onto the string (4.67 ± 0.09 vs. 1.75 ± 0.17 mg PLGA) with the highest viscosity solution (15.7 cP) compared to the lowest viscosity solution (1.4 cP). To investigate the effect of incubation time on PLGA loading, strings were incubated in PLGA solutions at RT for 1 min or 24 h (Supplementary Figure S1D). Results showed that there was no significant difference in PLGA loading with longer incubation time. The number of dipping iterations on PLGA loading was investigated over a 1 min incubation period. Strings were dipped once and incubated for the entire 1 min in PLGA solutions or dipped for 6 consecutive times (10 s each) over 1 min. Results showed that increasing the number of dipping iterations resulted in a significant increase in PLGA loading onto the strings for all three PLGA MWs. PLGA loading increased from 1.75 mg to 2.61 mg, 3.07 to 4.6 mg, and 4.67 to 7.27 mg for PLGA MW 11, 27, and 53 kDa, respectively (Supplementary Figure S1D). Finally, the effect of drying time between iterative dipping on PLGA loading was investigated. Results showed that increasing the drying time between iterative dips from 3 s to 30 s resulted in a significant increase in PLGA (MW 53 kDa) loading from 4.67 mg to 10.73 mg (Supplementary Figure S1E). Based on the collective data from the aforementioned optimization steps, the optimized drug loading conditions using the dip-coating process were set to a total of 1 min incubation time with 6 consecutive dipping steps (10 s each) and 30 s drying time between consecutive dipping steps. The loading solution used in subsequent studies with fluticasone (FTS) consisted of 1:5 *w/w* PLGA/acetone. The saturation solubility of FTS in the dipping solutions was ~10 mg/mL for solutions made with PLGA MW 11 and 27 kDa and ~6 mg/mL in the solution containing the highest MW PLGA (53 kDa) (Supplementary Figure S1E).

As the string would ultimately be for human use with an average esophageal length of 20–25 cm (the same length as in the porcine model), and for an overnight dwell time, the drug release was tested for a period of 24 h. Optimization studies were carried out with a 2 cm PCL string and extrapolated to a 25 cm string based on the homogenous and consistent drug loading data obtained from the optimization studies (Table 1). In vitro release studies showed that FTS had a slow burst release within 24 h (≤1%) corresponding to ~15, ~23, and ~44 µg for solutions containing PLGA MW 10, 27, and 53 kDa, respectively. These FTS concentrations were much lower than a targeted release of 1–2 mg FTS in 24 h (Table 1). The low FTS burst release within 24 h was attributed to a high affinity of FTS to PLGA given its hydrophobic nature (LogP = 2.78) resulting in a slow diffusion from the polymer layer to the release medium. To enhance FTS release from the strings, PLGA was removed from the loading solution and a saturated solution of FTS in acetone (20 mg/mL) was used. The drug loading process using the saturated FTS solution in acetone was further optimized to include a 4 min incubation time, with 24 consecutive dipping steps and 30 s drying steps between each consecutive dipping steps. These new loading parameters resulted in significantly higher FTS release within 24 h meeting the targeted 1 mg/day release. This optimized process was used to test FTS release both ex vivo and in vivo in a porcine model.

To design a clinically translatable fluticasone eluting string for human use and easy swallowing, a fabric (cotton) string was loaded with FTS using the optimized dip-coating process (4 min, 24 dips, and 30 s drying) and investigated for FTS release in vitro, ex vivo, and in vivo. In vitro release studies showed that FTS was released from the fabric string at the target rate with ~1.4 mg/day for a 25 cm string (Table 1).

**Table 1.** In vitro release kinetics of FTS from PCL strings and fabric strings (A) In vitro drug release profile of FTS loaded PCL string (2 cm L) and predicted profile for a 25 cm L string. (B) In vitro release profile of FTS loaded PCL string at varying dipping conditions. (C) In vitro release profile of FTS loaded fabric string. \* Dipping Time = 2 min, Number of dips = 12, Time between dips = 30 s. \*\* Dipping Time = 4 min, Number of dips = 24, Time between dips = 30 s.

