(T1530-09-61) Effect of Collagen Source on the In Vitro Performance of Collagen Implant

Purpose: Bupivacaine HCl collagen implant is a matrix type drug delivery system that is implanted in the tissue to manage post-surgical pain. The implant matrix consists of purified type I collagen as the only excipient. Type I collagen may be obtained from different sources. The sources of collagen may affect the morphology of the implant matrix which in turn may affect the quality, drug release and hence the in vitro performance of the implant. The purpose of this work was to investigate the effects of various sources of collagen on the quality and performance of bupivacaine collagen implant.

Methods: Purified type I collagens were obtained from different sources (A, B and C). The structural property of collagen was characterized by measuring the hydroxyproline to nitrogen ratio (Hyp/N) using colorimetric method. Fourier transform infrared spectroscopy (FTIR) was used to evaluate the structural composition of the raw collagens. Several placebo collagen dispersions were prepared in acidified water (pH 4.5) at 38°C ± 2°C with high shear homogenizer by varying the speed (1000-2000 rpm) and time (15 to 30 mins). The morphology of fibrillar structure of collagen dispersions were investigated by morphologically directed Raman spectroscopy (MDRS). Bupivacaine HCl was dissolved in acidified water (pH 4.5) at 38°C ± 2°C. which was subsequently mixed with the collagen dispersion using optimized homogenization speed and time. The resulting dispersion was filtered and filled in polyethylene glycol terephthalate containers for lyophilization. The effect of lyophilization process parameters, specifically primary drying temperature, and primary drying time on the quality of the resulting implants was investigated. The implants were characterized for porosity by scanning electron microscopy (SEM), tensile strength by texture analyzer, % drug content by HPLC and in vitro release test (IVRT) by USP apparatus II. The moisture content of implants was determined by thermogravimetric analysis. The solid state of API (amorphous vs crystalline state) in the implants was investigated using X-ray diffraction and differential scanning calorimetry.

Results: Hyp/N ratio for raw collagens from A, B and C sources were comparable (1.18, 1.19 and 1.18, respectively) and comparable for collagen from all sources. The FTIR spectra of collagens from three sources exhibited similar peak patterns with characteristic wavenumbers at 1500-1650 cm^-1(polypeptide backbone C-N stretching), 1700-1800 cm^-1 (polypeptide backbone C-O stretching), 2900-3000 cm^-1 (N-H in amide stretching) and 3200-3500 cm^-1 (N-H in amide stretching) cm^-1 (Figure 1). The homogenization conditions did not impact the collagen fibrillar size in dispersions prepared using collagen from three sources. However, based on MDRS images, the dispersion containing collagen from Source A exhibited a loose clustered micro fibrillar structure (Figure 2A) whereas collagen dispersions from Sources B and C appeared as large interconnected tangled fibrillar structures (Figure 2B and 2C). The images in Figure 3 show that implants manufactured using collagen from Sources A and B (F1 and F2) have smooth surfaces and were non-sticky in nature. In contrast, the implant manufactured using collagen from Source C (F3) exhibited a rough and cracked surface. SEMs of all implants showed a randomly interconnected fibrous and porous structure. The moisture content of implants containing collagen from Sources A, B and C were 5.02%, 13.52%, and 11.84%, respectively. The implant from Source A contained significant low amount of moisture. All implants (F1, F2 and F3) did not show a distinct melting point compared to crystalline bupivacaine (111.83°C) indicating amorphous nature of bupivacaine in the implants.

Conclusion: In conclusion, the structure of collagen from various sources were similar but the morphology of the resulting implants was affected by the source of collagen. These morphological differences may be critical for in vitro performance of the implants.

Acknowledgements: This poster reflects the views of the authors and should not be construed to represent FDA’s views or policies.
This project is funded by FDA/CDER/OGD

Figure 1. FTIR spectra of raw collagen from different sources (A, B and C)

Figure 2. MDRS images of different dispersions: (a) loose clustered micro fibrillar structure of dispersion from collagen A , (b, c) large interconnected tangled fibrillar structure of dispersions from collagen B and C

Figure 3. Photographs of implants following lyophilization. F1 and F2: smooth top surfaces of implants containing collagen A and B; F3: rough and cracked surface of implant containing collagen C