The human vocal fold consists of three layers: the epithelium, the lamina propria, and the vocal muscle. The epithelium of the vocal fold is a stratified squamous cell epithelium [
Vocal fold trauma due to inflammation, injury, radiotherapy, or surgery can lead to the formation of scar tissue. The microstructure of the lamina propria changes drastically by losing the distinctive 3-layer structure. There is an increase of collagen formation to the full depth of the lamina propria of scarred vocal folds which seems to form thick, organized bundles. The elastic fibers are present in the three layers but are disorganized and although the hyaluronic acid has the same density compared to the normal vocal fold, its distribution is uniform in all three layers of the lamina propria and even more accentuated in the deep layers [
Numerous phonosurgical procedures have been attempted to restore the vibratory properties of the scarred vocal folds. This includes injection of collagen [
Recently, several studies have been performed in animal models regarding the effect of cell therapy on scarred vocal folds, human embryonic stem cells [
Plan of protocol.
The principle study design was used by several investigators. Our experimental models were three-month-old white male New Zealand rabbits with a body weight between 2.900 Kg and 3.950 Kg. Rabbits were chosen for their relatively small size and low cost as an experimental model but large enough vocal folds to work with. The vocal folds of rabbits also present a three-layered structure like in human beings. All the animal care and experimental procedures were conducted in accordance with the guide of animal experiments and the experimental protocol had been approved by the Scientific Committee of Experimental-Research Center ELPEN and by Veterinary Authority of East Attica Prefecture (pd 160/1991, EU Directive 609/1986).
We used 74 white New Zealand rabbits for the protocol. Sixteen of them were used as control group (normal/control group) to compare our results to normal, uninjured vocal folds. These rabbits were sacrificed by an overdose of phenobarbital sodium intravenously.
After premedication with glycopyrrolate (0.1 mg/Kg intramuscular) and diazepam (2 mg/Kg intramuscular), all animals were anesthetized with a solution of xylazine (Rompun, 35 mg/kg intramuscular) and ketamine (5 mg/kg intramuscular). The remaining 58 animals were put on a custom-made ramp and with the use of video monitor and a rigid endoscope (4 mm diameter, 0 degrees, MEGA) we proceeded with direct laryngoscopy and visualization of the vocal folds. If any anatomical anomalies were revealed, the animal was ruled out. The scarring procedure was performed with a Hartmann ear forceps, excising epithelium and the lamina propria of the anterior and middle portion of the vocal fold, creating a large defect (Figures
Normal vocal fold.
Creation of a trauma using cold instruments.
Bilateral trauma on vocal folds.
Scar tissue on bilateral vocal folds.
After 18 months 15 rabbits received anesthesia with 50 mg/Kg of ketamine and local anesthesia (lidocaine) in the right inguinal region. We isolated adipose tissue (max 0.2 grams). The incision was sutured in layers.
To isolate ADSC, we digested adipose tissue at 37°C for 40 min using 0, 2% collagenase type I, and 2% bovine serum albumin and neutralized the enzyme activity with Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum (FBS). Then, it was centrifuged at 430 ×g (1400) for 10 min without brakes and we resuspended pellet in 100
We chose the 25 LS MACS separator columns in order to get 108 cells as the maximum number of labeled cells and
The total volume of 5 mL ADSC (Figure
Hematoxylin-eosin. 40x vocal fold: adipose mesenchymal stem cells after isolation, before infusion in animal model.
A Cytomics FC500 by Beckman Coulter was used with CXP software for the Cytomics FC500 flow cytometry system version 2.2. We used 500
After three hours (time needed for the isolation of the adipose stem cell), we proceeded with the premedication and anesthesia of all rabbits, according to the previous protocol. All animals were put in a custom-made ramp and with the use of video monitor and a rigid endoscope (4 mm diameter, 0 degrees, MEGA) we proceeded with direct laryngoscopy and visualization of the vocal folds. If any anatomical anomalies were revealed, the animal was ruled out. In the 15 rabbits from which we isolated the adipose stem cells, we proceeded with the injection in both vocal folds of 0.1 mL of solution of stem cells, approximately 10,000 stem cells (Figure
Injection of ADSC.
After three months we proceeded with direct laryngoscopy to assess the morphological changes in the vocal folds (Figure
Vocal folds 3 months after the injection of ADSC.
Staining was made with hematoxylin and eosin and images analysis was made at 10x, 20x, and 40x magnifications (Axiolab ZEISS microscope). Lamina propria thickness was assessed by a hematoxylin and eosin stain at ×10 magnification and was assessed by measuring the distance from the basal lamina of the epithelium down to the thyroarytenoid muscle. Measurements were made at three spots of the lamina propria of each vocal fold, at the middle third of uninjured vocal folds, and at the site of the scar tissue in treated vocal folds. We measured the thickest part, the thinnest part, and an intermediate one and we calculated the median (Figure
Schematic overview of lamina propria thickness. Lines AB, CD, and EF represent three different points for the calculation of the median. Measurements were performed in mm. Lines GH and IJ represent the division of the lamina propria in three equal-in-thickness layers, SLP, ILP, and DLP.
For statistical analysis of the data we used SPSS, version 20.0. The thickness of lamina propria between the different groups was tested by a 1-way analysis of variance (ANOVA). For the structure of the elastin and the density of collagen, hyaluronan, and elastin, we proceeded to logarithmic transformations considering that the variables are not normally distributed (Kolmogorov-Smirnov control,
In the untreated group (control B), the surface of 100% vocal folds was irregular, with a scar and stiffness. In the group that received ADSC implantation, the surface of the vocal fold tends to be smoother, with fewer irregularities (Figure
In the normal population the mean value of the lamina propria is 0.1 mm (maximum: 0.35 mm, minimum: 0.03 mm, Figure
(a) Normal vocal fold. (b) Scarred vocal fold. (c) Injection of ADSC. (d) Injection of hyaluronan.
The collagen in the normal vocal folds is mainly distributed in the intermediate and the deep layers of the lamina propria (Figure
(a) Normal vocal fold. (b) Scarred vocal fold. (c) Injection of ADSC. (d) Injection of hyaluronan.
The elastin is located mostly in the intermediate layer of the lamina propria as seen by the median (Figure
(a) Normal vocal fold. (b) Scarred vocal fold. (c) Injection of ADSC. (d) Injection of hyaluronan.
In the normal vocal folds the hyaluronic acid is also mainly distributed in the intermediate and the deep layers of the lamina propria (Figure
(a) Normal vocal fold. (b) Scarred vocal fold. (c) Injection of ADSC. (d) Injection of hyaluronan.
Mean thickness (mm) of the lamina propria (LP), through the different treatment groups. Normal = normal/control group, control = scarred/control group, stem = stem cell group, and hyaluronan = hyaluronan group.
Density of collagen in the superficial lamina propria (SLP). Normal = normal/control group, control = scarred/control group, stem = stem cell group, and hyaluronan = hyaluronan group.
Density of collagen in the intermediate lamina propria (ILP). Normal = normal/control group, control = scarred/control group, stem = stem cell group, and hyaluronan = hyaluronan group.
Density of collagen in the deep lamina propria (DLP). Normal = normal/control group, control = scarred/control group, stem = stem cell group, and hyaluronan = hyaluronan group.
Density of elastin in the superficial lamina propria (SLP). Normal = normal/control group, control = scarred/control group, stem = stem cell group, and hyaluronan = hyaluronan group.
Density of elastin in the intermediate lamina propria (ILP). Normal = normal/control group, control = scarred/control group, stem = stem cell group, and hyaluronan = hyaluronan group.
Density of elastin in the deep lamina propria (DLP). Normal = normal/control group, control = scarred/control group, stem = stem cell group, and hyaluronan = hyaluronan group.
Level of disorganization of elastic fibers. Normal = normal/control group, control = scarred/control group, stem = stem cell group, and hyaluronan = hyaluronan group.
Density of hyaluronic acid in the superficial lamina propria (SLP). Normal = normal/control group, control = scarred/control group, stem = stem cell group, and hyaluronan = hyaluronan group.
Density of hyaluronic acid in the intermediate lamina propria (ILP). Normal = normal/control group, control = scarred/control group, stem = stem cell group, and hyaluronan = hyaluronan group.
Density of hyaluronic acid in the deep lamina propria (DLP). Normal = normal/control group, control = scarred/control group, stem = stem cell group, and hyaluronan = hyaluronan group.
Vocal folds have a highly specific structure, based on a specific distribution of the different cellular population and the different constituents of the extracellular matrix. When there is a trauma on the vocal fold, a high intensity inflammatory process takes place, in order to restore the structure and the function of the vocal fold. In case of a great trauma and as a consequence an intense inflammatory response, the healing process can lead to the formation of a scar tissue. This scar tissue mostly located in the lamina propria of the vocal fold adds great stiffness and interferes with the normal vibratory properties of the vocal fold by losing the three-layered structure of the lamina propria. Clinically patients present with dysphonia, hoarseness, loss of vocal range, and strain. Numerous treatment modalities and injection materials have been applied in order to treat the glottal insufficiency caused by scar tissue or vocal paralysis. None of them led to an optimal result.
According to Arnold [
Over the last years, much attention has been directed in the cell therapy research and regenerative medicine. Regenerative cell therapy involves the delivery of autologous or nonautologous mesenchymal stem cell, with the purpose of tissue and organ regeneration and reconstruction. Level of evidence was progressive within the years. Hanson et al. [
This study gives a great amount of information concerning the pathological changes following the implantation of adipose stem cell in a chronic vocal fold wound. In our study the density of collagen after the implantation of adipose stem cell is reduced significantly at the same levels as the normal, uninjured vocal folds. In studies where adipose stem cells have been used, Liang et al. [
When we want to compare the effects on quasi-mature scar regardless of the type of mesenchymal stem injected, we see that Svensson et al. [
In our study we failed to prove an increase of elastic fibers in the group of animals that were treated with adipose stem cell, although there is a clear tendency. The fact is that we demonstrate a significantly better structure and rearrangement of the fibers of elastin, when compared to the scarred/untreated group. We observed the same result in the vocal folds which were treated with hyaluronic acid. This result could be explained based on the hyaluronan anti-inflammatory profile as shown by the study of Hanson et al. [
There is no great amount of literature on the effects of cell therapy on the hyaluronan level in a scar tissue of the vocal folds. In our study, the levels of hyaluronic acid were significantly reduced in a chronic scar of vocal folds and returned to the same levels as in normal vocal folds. When we want to compare the studies in function of injection of ADSC, regardless of the maturity of the scar tissue, Hu et al. [
The major difference between the groups treated with mesenchymal stem cells and hyaluronic acid is the important dissolution of the excess of collagen fibers in the stem cell group through all the three layers of the lamina propria (
In our study we used the mesenchymal stem cells derived from a minimum amount of autologous fat while in all other studies the mesenchymal cells derived from bone marrow or the placenta, a nonautologous source. Few studies exist which use adipose-derived stem cells but the literature is growing [
The laryngoscopy was performed with the aid of mouth expander and a rigid endoscope of 0°/4 mm as opposed to the use of pediatric Pillings pediatric endoscope used in the majority of studies. This allows us to have a better exposure of the operative field.
In the past studies the implantation of mesenchymal stem cells had been some days (5–7) after the creation of the vocal cord injury [
In the present study we created three groups (untreated chronic scar group, adipose stem cells group, and hyaluronan group) and we compared the pathological findings (density of collagen, elastin, and hyaluronan and thickness of the lamina propria) between two treatment modalities and one control group. In previous studies [
Finally, on pathological sections differentiated mesenchymal cells from adipose tissue were identified. The identification was performed using 7-AAD.
Although many interesting points have been made in this research, it confirms that the amelioration in the microstructure of the vocal fold coexists with the amelioration of the vibratory properties also. Restoration of a chronic vocal wound with injection of ADSC has to improve the plasticity and rheology of the vocal folds.
Restoration of a chronic vocal wound still remains a great challenge. The data presented indicate that the transplantation of adipose-derived mesenchymal stem cells in a chronic scar can lead to an anatomical regeneration of the vocal fold by dissolution of the excessive collagen fibers forming the scar tissue and restoration of the normal structure of the elastic fibers. Further studies have to be conducted in order to assess the improvement of the viscoelastic properties.
The authors declare that there is no conflict of interests regarding the publication of this paper.
This study was funded by Scholarship-Grant by the Experimental-Research Center ELPEN Pharmaceuticals (ERCE).