Hyaluronic Acid/Alginate Hydrogel Containing Hepatocyte Growth Factor and Promotion of Vocal Fold Wound Healing

Abstract

Background: Hepatocyte growth factor (HGF) has been shown to facilitate vocal fold (VF) wound healing. This study was undertaken to determine whether the therapeutic efficacy of HGF could be enhanced by applying it in hyaluronic acid and alginate (HA/ALG) composite hydrogels into VFs after injury in a rabbit model.

Methods: HGF was loaded into HA/ALG composite hydrogel (HGF-HA/ALG) and its in vitro release profile was evaluated. In addition, HGF-HA/ALG was injected into the VFs of rabbits immediately after direct injury and HGF or PBS was injected in the same manner into control groups. Macroscopic features were observed by endoscopy at 3 months post-injury. Functional analyses including mucosal waves of VFs and viscoelastic properties were performed by kymography following high-speed digital imaging and rheometer. Histopathological and immunohistochemical evaluations were also conducted on VFs.

Results: HGF release from HGF-HA/ALG was sustained for up to 3 weeks. Rabbits treated with HGF-HA/ALG showed improved mucosal vibrations and VF viscoelastic properties as compared with the PBS and HGF controls. Histopathological staining revealed HGF-HA/ALG treated VFs showed less fibrosis than PBS and HGF controls, and immunohistochemical analysis demonstrated amounts of type I collagen and fibronectin were lower in HGF-HA/ALG treated animals than in PBS and HGF controls at 3 months post-injury.

Conclusion: HGF containing HA/ALG hydrogel enhanced healing in our rabbit model of VF injury.

Keywords: Alginate; Hepatocyte growth factor; Hyaluronic acid; Vocal fold; Wound healing.

Fig. 1

Experimental flow chart of in vivo study design

Fig. 2

Profile of hepatocyte growth factor (HGF) release from HA/ALG hydrogel. The amount of HGF released increased consistently over 21 days

Fig. 3

Endoscopic examination of VFs at 3 months post-injury. Endoscopic examination of VFs at 3 months post-injury revealed prominent fibrotic scars in the PBS group, but reduced scar formation in the HGF and HGF–HA/ALG groups (the PBS group; injured and PBS treated, the HGF group; injured and HGF treated, HGF–HA/ALG group; injured and HGF–HA/ALG treated)

Fig. 4

Micromorphological and immunohistochemical evaluations of ECM remodeling by HGF released from HGF–HA/ALG at 3 months post-injury. Scale bars represent 30 μm. Masson's trichrome (blue-stained collagen), collagen type I and fibronectin densities were quantified by summing numbers of pixels in positively stained areas using a software program. Results are presented as mean areas (%) ± SDs as determined by the Kruskal–Wallis test with Dunns’ post hoc test. *versus normal uninjured/non-treated controls, #versus injured/PBS treated animals, ^$versus injured/HGF treated animals, *p < 0.05, ***p < 0.001, ^###p < 0.001, ^$p < 0.05 and ^$$p < 0.01. (n = 5 rabbits per group) (the normal control group; non-injured and non-treated, the PBS group; injured and PBS treated, the HGF group; injured and HGF treated, HGF–HA/ALG group; injured and HGF–HA/ALG treated)

Fig. 5

Evaluation of mucosal wave oscillations after vocal fold injury. The amplitudes of mucosal waves (MWs) of VFs were evaluated by measuring the pixels of amplitudes of a mid-point of upper lips on high-speed digital images using image analysis software at 3 months post-injury. The amplitude ratios of MWs of treated VFs versus normal controls were calculated using Metamorph software. Normalized MW results are presented as mean ratios (right VFs to left VFs) ± SDs. Kruskal–Wallis test with Dunns’ post hoc test. *versus the PBS group, ^#versus the HGF group, **p < 0.01, ***p < 0.001 and ^###p < 0.001 (n = 5 rabbits per group) (PBS group; injured and PBS treated, HGF group; injured and HGF treated, HGF–HA/ALG group; injured and HGF–HA/ALG treated)

Fig. 6

Rheometric evaluations of viscoelasticity changes after vocal fold injury. Viscoelastic properties of regenerated VFs were measured at 3 months post-injury. Mean elastic moduli (G′) and viscous moduli (G″) are plotted on a log–log scale as a function of frequency. Rheological data were analyzed at all frequencies by two-way ANOVA with Bonferroni’s post hoc test (both p < 0.01) (n = 5 rabbits per group) (the normal control group; non-injured and non-treated, the PBS group; injured and PBS treated, the HGF group; injured and HGF treated, HGF–HA/ALG group; injured and HGF–HA/ALG treated)