Intended for healthcare professionals

Clinical Review State of the Art Review

Advances in regenerative medicine for otolaryngology/head and neck surgery

BMJ 2020; 369 doi: (Published 29 April 2020) Cite this as: BMJ 2020;369:m718
  1. Michael J McPhail, biomedical engineer1,
  2. Jeffrey R Janus, associate professor of otolaryngology - head and neck surgery2,
  3. David G Lott, professor of otolaryngology - head and neck surgery1 3
  1. 1Head and Neck Regenerative Medicine Laboratory, Mayo Clinic Arizona, Scottsdale, AZ, USA
  2. 2Department of Otolaryngology – Head and Neck Surgery, Mayo Clinic Florida, Jacksonville, FL, USA
  3. 3Department of Otolaryngology – Head and Neck Surgery, Mayo Clinic Arizona, Phoenix, AZ, USA
  1. Correspondence to: D G Lott lott.david{at}


Head and neck structures govern the vital functions of breathing and swallowing. Additionally, these structures facilitate our sense of self through vocal communication, hearing, facial animation, and physical appearance. Loss of these functions can lead to loss of life or greatly affect quality of life. Regenerative medicine is a rapidly developing field that aims to repair or replace damaged cells, tissues, and organs. Although the field is largely in its nascence, regenerative medicine holds promise for improving on conventional treatments for head and neck disorders or providing therapies where no current standard exists. This review presents milestones in the research of regenerative medicine in head and neck surgery.


Otorhinolaryngology—head and neck surgery (Oto-HNS) spans many different tissue types and functions, including hearing, balance, air filtration and humidification, smell, facial animation, deglutition, breathing, producing vocal sounds, and articulation during speech. Loss of these functions can result in high morbidity and, in some cases, mortality. Traditional strategies for replacement of tissues include grafts from other tissues, artificial materials, and transplants.1 Grafts can incur donor site morbidity, have limited availability of grafting material, and lack a purpose built form. Artificial materials can be hampered by immune response and risk of infection.2 Transplants require immunosuppressive drugs and also have limited availability. Furthermore, some sites within Oto-HNS have no possible functional replacement. Regenerative medicine aims to restore functions of the Oto-HNS sites through replacing or regenerating the relevant cells, tissues, and organs.3

In this review, we describe developments in regenerative medicine research for head and neck sites. We pay particular attention to the sites where regenerative therapies have been used in humans and note future research questions. Most trials described here are considered exploratory, with few randomized controlled trials or meta-analyses available in the literature. In an effort to represent the breadth of regenerative medicine work being done in Oto-HNS, we discuss pertinent basic science studies in areas in which human trials have not been conducted or to provide depth of understanding of work leading to a clinical trial. We expect this review to be relevant to clinicians, clinician scientists, academics, and regulatory specialists working in this rapidly developing field.

Sources and selection criteria

We identified articles through PubMed searches including peer reviewed articles published in English. We used the following search terms, and their combinations: “regenerative medicine” and “tissue engineering” with “ear”, “cochlea”, “nose”, “larynx”, “vocal fold”, “trachea”, “craniofacial”, and “head and neck”. These searches yielded more than 1000 publications varying from basic science research to human clinical trials. We also identified studies from reference lists of review articles. We prioritized randomized controlled trials the highest but also included exploratory trials and retrospective studies. Basic science or animal trials are briefly discussed where human trials have not yet been done or when informative to the development of a therapy with evidence in humans. Although often overlapping in goals and research methods with regenerative medicine, we excluded publications describing transplant medicine or conventional reconstructive surgery.

Regenerative medicine concepts

A widely used definition of regenerative medicine is “the process of replacing, engineering or regenerating human cells, tissues or organs to restore or establish normal function.”3 Early research in regenerative medicine hoped to engineer tissues and organs outside the body (ex vivo tissue engineering).4 However, growing mature tissues and organs outside the body has proven challenging and commercialization of these approaches has been limited. Regenerative medicine has expanded its methods to include in vivo regeneration of tissues. These in vivo regenerative approaches range from using the human body as a bioreactor to augmenting the body’s innate ability to regenerate and heal. Regenerative medicine approaches in Oto-HNS have varied widely, reflecting the variation in tissue type and function found across these sites.

Methods used can generally be grouped into cells, bioactive factors, and scaffolds, as illustrated in figure 1.567 Cells serve a central role in regenerative medicine. Transplanted cells can be used to regenerate new tissues, modulate immune response, and modify native cell behavior through paracrine signaling.8 Bioactive factors are used to modify cellular behaviors and include growth factors, cytokines, hormones, and other molecules.5 These factors can be used in vitro as tools to control cell behavior as well as in vivo as therapeutics to modify regenerative processes.910 Scaffolds provide a three dimensional structure for tissue regeneration. Scaffolds can be designed at the micro scale to control cell behaviors such as differentiation and migration.11 At the macro scale, scaffolds are designed to support mechanical loads and shaped to fit into defects and restore organ function.1213 Scaffolds can be used acellularly or in conjunction with bioactive factors and cells. A broad range of approaches have been used to generate scaffolds, including three dimensional printing,14 electrospinning,15 customizing hydrogels,16 and decellularizing tissue.1718

Fig 1
Fig 1

Concepts in tissue engineering and regenerative medicine.Cells, scaffolding, and bioactive factors are used in isolation or in complement to regenerative tissue. Reproduced with permission from Lott DG, Janus JR. Mayo Proc 2014;89:1722-335

Challenges to translation

Many challenges lie ahead for clinicians and scientists in Oto-HNS aiming to translate regenerative medicine therapies into practice. Although these therapies have been researched for decades, very few have been translated to clinical use.192021 A survey of 131 clinicians, scientists, and industry experts identified manufacturing as a key barrier to the adoption of cellular therapies, with efficacy, regulation, and cost effectiveness as other important barriers.22 Manufacturing of regenerative therapies can often be patient specific and/or use autologous tissue, thus being more costly and more difficult to scale up than traditional drugs. For this reason, regenerative therapies, such as autologous cell therapies, may need to show superior safety, efficacy, or both compared with standard of care.23 Otherwise, regenerative therapies may be best used for diseases with no current standard of care.

Developing and evaluating new regenerative therapies and obtaining regulatory approval is a lengthy and expensive process.20 For example, achieving clinical application of skin scaffolds—a notable scaffolding approach in tissue engineering—has taken decades.24 However, the regulatory approval process is critical for ensuring that safety and efficacy are properly evaluated and ethical controversies are avoided. As shown in the following discussions, the process moves from robust preclinical studies to first in human trials, prospective series, and ultimately randomized control trials, if possible.25

Clinicians may face ethical challenges unique to regenerative medicine therapies. Use of autologous or allogeneic cells and tissues needs careful consideration of informed consent and property rights.20 Clinicians may also face patients who have been exposed to overly optimistic claims in press releases,26 media reports,27 or patient testimonials.28

The science itself can present the most daunting challenges. As highlighted in the body of this review, regeneration and tissue engineering of connective tissues such as cartilage and bone have been shown to be feasible in preclinical and early clinical trials in Oto-HNS sites. However, regeneration of heterogeneous and complex tissues and organs, such as a vocal fold, remains a challenge to the field.24 Additionally, the presence and effect of cellular and bioactive factor therapies is often short lived, necessitating discovery of options to provide sustained clinical benefit.

Finally, bias toward publishing only research with positive results also serves as a challenge to the regenerative medicine community and may represent the field as overly positive, while not teaching valuable lessons from failure to other investigators. In the sections below, we present key research studies that have had a significant effect on the landscape of regenerative medicine in Oto-HNS.

Regenerative medicine advances in Oto-HNS

Maturity of regenerative medicine therapies and research varies substantially across Oto-HNS sites. We discuss progress in each site separately and cover the most investigated areas in the field.


Restoration of laryngeal anatomy and function is difficult given the complex nature of this biomechanical organ. Current reconstructive options are unable to restore appropriate tissue without scar formation or restore normal laryngeal function. Regeneration of new functional laryngeal tissue could theoretically help millions of patients, ranging from scarred vocal folds to laryngectomy.

Vocal fold microstructure restoration

The specialized layered microstructure of the vocal folds consists of squamous epithelium, lamina propria (superficial, middle, and deep layers), and muscle (fig 2). The superficial layer of the lamina propria (SLP) plays an integral role in sound production for voicing. Loss of pliability in this layer from scar formation can result from chronic vocal misuse, trauma, or surgery. Therapeutic materials used in practice to attempt to restore pliability (for example, lipoinjection, steroid injections) do not have viscoelastic properties similar to the natural lamina propria and do not promote normal tissue regeneration or repair. Regenerative scaffolds, cells, and bioactive factors are all being studied toward this end. Combining these different components seems to create a synergistic environment in which they support and increase one another’s effects.29303132

Fig 2
Fig 2

Schematics of vocal fold anatomy. A: superior view. B: coronal cross section. Reproduced with permission from Lott DG, Janus JR. Mayo Proc 2014;89:1722-335

Bioactive factors

Growth factors are peptide molecules that function to regulate cell proliferation and differentiation. Most vocal fold studies have focused on controlling the behavior of vocal fold fibroblasts. The bulk of the research has focused on basic fibroblast growth factor (bFGF) and hepatocyte growth factor (HGF).

The use of bFGF to improve vocal fold scarring has been thoroughly investigated.333435363738 The first injection into an atrophic human vocal fold was described in 2008.34 The injection improved aerodynamic and acoustic parameters one week after injection and lasted up to three months. As the initial results were promising, the same team conducted a non-randomized, single institution, human clinical trial investigating the use of bFGF for aged vocal folds, sulcus vocalis, and scarring. Ten patients (six men and four women; mean age 70.1 years) each had bFGF injected into each treated vocal fold, repeated up to seven times if necessary. After a one year follow-up, all patients showed improvement in vocal function, maximum phonation time, and acoustic/aerodynamic measures. No long term adverse effects were noted.38

HGF is a pleiotropic cytokine with a favorable profile for vocal fold healing. It is found in vocal folds after injury,39 has strong anti-fibrotic potency,40 increases hyaluronic acid and elastin synthesis, decreases collagen synthesis, induces cell growth and migration, and is highly angiogenic.41 In one study, a single injection of HGF into canine vocal folds was performed after mucosal stripping.42 At six months post-injury, HGF seemed to prevent excessive collagen deposition and tissue contraction, thereby reducing the effects of scarring on the vibratory properties of the vocal folds. Further studies have since confirmed these findings.4344 Building from this work, a phase I/II, uncontrolled, exploratory clinical trial in 18 patients was conducted using HGF injections for treatment of vocal fold scar or sulcus.45 Reported adverse events were hyperemia of the vocal fold in three patients. Improvements in voice measures were found, but further randomized controlled trials are needed.

Cell implantation

Two primary approaches are taken to the use of cells in laryngeal tissue regeneration. The first controls the processes within the native environment. The other uses cells to directly restore normal anatomy. Many cell types have been studied for vocal fold regeneration,46 but fibroblasts and stem cells seem to be the most promising to date.

Fibroblasts are responsible for maintenance of the vocal fold extracellular matrix (ECM) and have been shown to be functionally similar to mesenchymal stem cells (MSCs) with similar cell surface markers and differentiation potential.47 Given these findings, fibroblasts have been extensively studied as a regenerative therapy for vocal fold scarring and atrophy. In an initial canine study, buccal mucosa fibroblasts were injected into scarred vocal folds with three weekly injections. Vocal fold mucosal waves and acoustic parameters improved significantly.48 In 2011 the same team reported on a prospective, open label, single arm, pilot study at a single tertiary care center.49 Autologous fibroblasts from the buccal mucosa were injected into five people with vocal fold scarring. Three doses were injected into the superficial lamina propria layer of each scarred vocal fold at four week intervals, and patients were followed for 12 months. Reported adverse events were severe otalgia during injection, resulting in the early termination of treatment injections in two patients. Four of the five patients showed objective and subjective improvements in voice quality and in mucosal wave by the study endpoint. This work was followed by a randomized, double blind, placebo controlled, multi-institutional, phase II trial.50 Fifteen patients with either vocal fold scar or atrophy received fibroblasts derived from the postauricular skin, and six patients received saline injections. Each patient received three injections at four week intervals for each vocal fold. Primary outcomes were a mucosal wave grade from videostroboscopy, expert perceptual analysis scores, and Voice Handicap Index 30 score. The mucosal wave grade ranged from 1 (absent a mucosal wave) to 5 (normal mucosal wave). A significant improvement was found in mean mucosal wave grade in both atrophy and scar patients receiving fibroblast injection compared with their baseline. Expert perceptual analysis blindly rated the patient’s voice from 0 (normal) to 3 (severely dysphonic). Mean perceptual scores were significantly improved in the fibroblast treated scar group at 12 months (0.8) compared with the saline control (1.5). No significant difference was found in the Voice Handicap Index 30 scores. No severe adverse events were reported in this study.

Many types of stem cells have been evaluated for regeneration of vocal fold tissue, and all preclinical studies reviewed show improvement in healing to varying degrees.515253545556 Adipose derived stem cells (ASCs) are a particularly promising cell type for treatment of scarred vocal folds.57 In vitro trials have shown that ASCs secrete several growth factors that balance collagen and hyaluronic acid in the ECM.4052535455585960 Studies have shown that ASCs secrete HGF that attenuates collagen production and fibroblast proliferation in culture. ASCs can also produce elastic fibers, which typically does not occur in a scar environment.61 These stem cell therapies for vocal folds remain in the early stages of investigation and have not been used in human clinical trials.

Vocal fold scaffolds

Many different scaffolds have been investigated for three dimensional lamina propria replacement, including decellularized organ matrix, biologics and biologic polymers, and hydrogels.626364 Scaffolds for vocal fold regeneration can either be applied through injection or be attached during surgery. Scaffolds are designed to have biomechanical similarity to the native lamina propria, to deliver cells and bioactive factors, modulate the inflammatory response, and direct ECM remodeling.65

Early feasibility of tissue engineered composite structures to replace damaged vocal folds has been demonstrated with in vitro studies.326667 Three dimensional matrices have been used for culturing ASCs and human vocal fold fibroblasts and epithelial cells.3266 Functional testing has shown that these constructs can be designed to show similar mechanical properties and vibration to normal vocal folds.3267

Although these exciting discoveries are still relatively early in investigation, they represent substantial progress toward the clinical use of tissue engineered constructs for restoration of normal vocal fold structure and function. Once ready for clinical use, these types of constructs may be directly applied to a damaged vocal fold or used as a component of a tissue engineered larynx superstructure.

Laryngeal superstructure

Bioengineering the larynx superstructure adds layers of complexity in that the increased number of tissue types, larger surface area, and increased functional demand must all be considered. The laryngeal superstructure includes the three dimensional shape and structure of the larynx. Creation of a tissue engineered implant to replace part or all of the superstructure would consist of a composite scaffold that, at a minimum, includes analogs of the thyroid cartilage and vocal folds. Promising examples of bioengineered laryngeal structures have been reported. In 2003, a porcine derived xenogeneic ECM was used for reconstruction of the larynx in adult dogs and showed that regeneration can happen in a large laryngeal defect.68

Aortic allografts have been used in exploratory human trials and case studies to reconstruct hemilaryngeal defects (fig 3).6970 These reports showed normal epithelialization and acceptable laryngeal function. The drawbacks of prolonged time to epithelialization and lack of patient specific three dimensional architecture have necessitated investigation into other techniques.

Fig 3
Fig 3

A: hemilaryngectomy defect overlaid with cadaveric cryopreserved aortic allograft. Reproduced with permission from Cain RB, et al. Otolaryngol Head Neck Surg 2014;150:501-2.69 B: endoscopic image showing epithelialized aortic allograft. *Reconstructed side

Whole human larynges have been successfully decellularized to produce scaffolds.7172 A preclinical study found minimal immunostaining for major histocompatibility complex and only a few detectable chondrocytes.71 Both bFGF and vascular endothelial growth factor were maintained, which are important for angiogenesis and neovascularization of the graft. The mechanical response of the cartilaginous laryngeal structures was similar to native tissues and reflected their different compositions (elastic cartilage in the epiglottis and hyaline cartilage in cricoid and thyroid cartilages).

These advances represent an important step toward successful laryngeal superstructure bioengineering. With further refinement, they might provide an option for functional partial laryngectomy reconstruction or create a scaffold for total laryngeal regeneration.


Although simple in concept, reconstruction of the trachea can be extremely difficult, with life threatening implications. Tissue damage can be segmental or involve the entire trachea, and smaller defects are typically managed with tracheal resection and reanastomosis. However, no adequate reconstruction option exists for defects involving more than half of the trachea. The primary reasons for this are the need for mucociliary clearance and that the blood supply to the trachea depends on small perforating vessels with no primary large artery blood supply, thereby hindering transplant and free tissue transfer. Investigative treatment strategies include transplantation, reconstruction with autologous tissue, allograft reconstruction, and tissue engineering.73

As with laryngeal tissue engineering, regenerative reconstruction of the trachea involves a temporary or permanent scaffold, seeded with cells and/or bioactive factors. Significant advancements in this area have been made using decellularized tracheas, synthetic scaffolds, aortic allografts, and three dimensional printing.

Tracheal decellularization

Tracheal decellularization is the most investigated technique to date. Implantation of a decellularized tracheal allograft has been described in two children.747576 The first case, in 2010, was a 12 year old boy with long-segment congenital tracheal stenosis.74 Bone marrow mesenchymal stem cells (BMSCs) were seeded onto the scaffold with patches of autologous epithelium. Topical human recombinant erythropoietin and transforming growth factor β were applied to the graft to support angiogenesis and chondrogenesis. Intravenous human recombinant erythropoietin was continued postoperatively. Initially, multiple stenting and bronchoscopic procedures were needed to maintain the airway. However, at four years, the graft remained patent with evidence of a ciliated epithelial layer.75 The second case was a 15 year old girl. This surgery used a decellularized tracheal allograft seeded with mesenchymal stromal cells and autologous respiratory epithelial cells. Although the patient initially did well, she developed progressive narrowing at the implant site and on postoperative day 15 she had a prolonged respiratory arrest resulting in her death. No autopsy was performed, and the cause of death remains unknown.76 This case highlights a concern about the structural integrity of decellularized tracheas. Previous decellularization techniques required multiple cycles and prolonged exposure to the various decellularization chemicals, resulting in structural weakness and scaffolds that would collapse into the airway with respiration. Once a limitation, the time and cycles needed to decellularize tracheal tissue have decreased from 17 cycles to only four cycles of processing, resulting in improvement in the structural integrity of the graft.18 Further improvements to the decellularization technique may facilitate more widespread clinical applicability.

Synthetic scaffolds

Synthetic scaffolds to reconstruct the trachea were first reported in humans in 2005.7778 The scaffolding was made with a polypropylene mesh tube coated in a collagen sponge and clotted with autologous blood before implantation. An exploratory clinical study in four patients found good epithelialization of the implant in all patients without stenosis, with follow-up ranging from eight to 34 months.78 Despite promising results, this technique has not translated to widespread clinical use. These surgeries were performed for smaller partial tracheal defects, and the results do not necessarily translate to long-segment defects. Additionally, this implant was not compared with other conventional reconstruction techniques. If it is not shown to be significantly superior, the expertise and time needed to create this scaffold may not be justified.

Aortic allograft

Perhaps the most mature tracheal regeneration approach for long-segment defects is the bioengineered and stented aortic allograft illustrated in figure 4 (A and B). In animal studies with this allograft, de novo regeneration of cartilage within the graft (fig 4, C) and regeneration of ciliated epithelium on the lumen of the graft (fig 4, D) were seen.8081 The proposed explanation for cartilage regeneration is migration of BMSCs into the graft.82 The first four cases in a six patient exploratory human trial had problems including dehiscence, acute spinal cord ischemia, pneumonia, and graft necrosis due to fungal infection; however, the last two cases were uneventful in postoperative care.83 A refined approach for cryopreservation of the aortic allograft showed viable donor cells after thawing, which were capable of releasing relevant cytokines and growth factors.84 The refined aortic allograft has been used for reconstruction of trachea, bronchi, or carina defects in an uncontrolled single site feasibility trial in 13 patients. All five patients with trachea reconstructions were alive at the time of publication, all could breathe and speak without a tracheostomy, and three had their stents removed.79 Regenerated cartilage rings were visible during bronchoscopy (fig 4, E) and in computed tomography scans (fig 4, F).

Fig 4
Fig 4

Bioengineered and stented aortic allograft for trachea reconstruction. A and B: schematic of stented allograft concept. Reproduced with permission from Martinod E, et al. JAMA 2018;319:2212-22.79 C: excised and dissected sheep trachea after 24 months’ implantation with graft placed between forceps. D: histologic examination showing ciliated cells and goblet cells. Reproduced with permission from Martinod E, et al. Ann Thorac Surg 2003;75:1572-8.80 E and F: bronchoscopy and computed tomography images showing regenerated cartilage within aortic allograft in patients (black and yellow arrowheads). Reproduced with permission from Martinod E, et al. JAMA 2018;319:2212-2279

Three dimensional printing

Interest is growing in the use of three dimensional printed biomaterials that allow for the device to change with tissue growth. This property has been demonstrated with biodegradable external splints for tracheomalacia in children, first used in compassionate care at the University of Michigan.85 These external splints have since been implanted into multiple children, all with resolution of life threatening airway disease and continued growth of the primary airways.86 Other centers are now using this technology with similar outcomes.87 These splints have not yet been approved by the US Food and Drug Administration (FDA) but continue to be used under emergency FDA clearance for compassionate care while they are being investigated for regulatory approval.

Craniofacial bone

Craniofacial bones serve the purposes of articulation and deglutition and house the brain, globes, and sinuses. These bones have the added function of bearing loads in response to the forces of mastication and providing the scaffolding for the face, indirectly comprising one’s appearance and sense of self.

Bioactive factor based therapies

Much of the use of bioactive factor based therapies in the craniofacial region is centered on the introduction of these bioactive molecules into bony defect sites to encourage osteogenesis, angiogenesis, or a combination of both processes. The primary growth factors that are most commonly used in craniofacial regeneration are bone morphogenetic proteins (BMPs) and platelet derived growth factors (PDGFs). Although other factors are being explored, these two have had the most clinical application to date.88

BMP is a unique group of proteins in the transforming growth factor β (TGF-β) superfamily and directly influences regulation of bone growth, maintenance, and repair.89 BMP is crucial in regulating craniofacial development and plays a role in postnatal craniofacial morphology and lifelong maintenance of dental structures.90 BMPs act on distinct type II and type I serine/threonine kinase receptors, the downstream effects of which accelerate osteoblast activity.88 BMP-2, BMP-4, and BMP-7 all have established in vivo efficacy in mending critical sized bone defects, with recombinant human BMP-2 being approved by the FDA for interbody spinal fusion, open tibial fractures, sinus augmentation, and localized alveolar ridge augmentation after dental extraction.91 Many of the applications for BMP are facilitated by loading the material onto a scaffold, such as an absorbable collagen sponge or a more sophisticated polymer construct, to span a bony defect.

PDGF was initially identified in the 1970s as a serum growth factor for fibroblasts, smooth muscles cells, and glial cells.92 The role of PDGF in broad wound healing activities in both osseous and soft tissue has been extensively established.93 PDGF was used in 1989 for the treatment of teeth affected by periodontitis in beagle dogs. Histologic analysis of the bone from the treated group showed significant increases in new bone and cementum formation. Specimens also showed a nearly continuous layer of osteoblasts along the formed bone.94 Since then, PDGF has been used in bone matrix and a recombinant form (rh-PDGF) has been developed and approved by the FDA for use in diabetic foot ulcers.95 Much like BMP, this bioactive molecule is delivered with the help of a matrix or scaffold. rh-PDGF-BB has been applied to areas of bone loss between two teeth roots with freeze dried bone allograft, resulting in significant clinical improvement as measured by probing depth and histologic improvement as measured by new bone, cementum, and periodontal ligament coronal to the reference notch.96

Platelet rich plasma (PRP) is another interesting bioactive factor that is being investigated for use in craniofacial applications. PRP is obtained by sequestering and concentrating platelets by gradient density and centrifugation and is rich in many factors (PDGF, TGF-β1, and TGF-β2).91 PRP in cancellous cellular marrow grafts shows greater bone density in grafts in which PRP was added.97 Since then, several systems have been developed for both the acquisition and centrifugation of plasma to concentrate platelets. In the correct context, these systems negate the need for removal of blood products from the operating room and therefore allow the platelet concentrate to be treated more like a graft and less like a drug. Similar systems exist for the concentration of bone marrow, straddling the line of cell therapy and bioactive molecular therapy.98

Scaffold based therapies

Typical scaffolds for craniofacial reconstruction are composed of ceramics, synthetic and natural polymers, and composites.99 Among other considerations of scaffolds thought to be important, porosity, tensile strength, compressive strength, degradation, and bio-printability all play a role in the selection of a scaffold.100 In the realm of maxillofacial bone regeneration, most ceramics are calcium phosphate based, including calcium hydroxyapatite and tricalcium phosphate, whereas most polymers that have seen clinical use include poly(glycolic acid), poly(lactic acid), their copolymer poly(lactic-co-glycolic acid), and poly(E-caprolactone).88 Other details of scaffolds, such as piezoelectric and surface charge characteristics, have been shown to influence timed release of growth factors and bone growth. Neutral and negatively and positively charged scaffolds have been fabricated using oligo((polyethylene glycol) fumarate) hydrogels in the setting of BMP-2 microspheres.101

Human trials

Many examples exist for the application of cell therapies, bioactive agents, and scaffolds for the restoration of the craniofacial skeleton. Typically, a combination of these is used. In a culmination of many of the techniques described above, rh-PDGF-BB combined with particulate anorganic bovine bone mineral was used for maxillary sinus augmentation. This case series showed abundant numbers of osteoblasts in concert with significant osteoid in all sites, indicating ongoing osteogenesis.102 Additionally, horizontal bone augmentation, ridge preservation, and periodontal and peri-implant studies have all been done in humans with favorable outcomes.93103104

The use of regenerative technology as it applies to the mandible has been progressive. Several examples have shown success for smaller defects. PRP was used on cancellous cellular bone grafts in 1998,97 and in 2010 incubated bone marrow aspirate was applied to autogenous fibrin-rich and platelet-rich clot and membrane, β-tricalcium phosphate, and hydroxyapatite.105 As regards conquering the challenge of mandibular discontinuity in the setting of critical defects (those which are full thickness and greater than 3 cm), a handful of small clinical studies have shown variable success.106 Much of this stems from the fact that greater defect size requires adequate vascularization for appropriate cell seeding and growth. In a notable clinical case study, β-tricalcium phosphate granules were used as a scaffold along with recombinant human BMP-2 and autologous ASCs for mandibular reconstruction of a 10 cm defect in a single patient (fig 5).107 Bone regenerated in situ, without ex vivo engineering of the tissue, and dental implants were fitted to the regenerated mandible after 10 months.

Fig 5
Fig 5

Surgical planning and reconstruction of large anterior mandibular defect. A: three dimensional printed skull model used for surgical planning. B: reconstruction plate and titanium mesh pre-bent to fit three dimensional printed model. C: resected anterior mandible. D: anterior mandible defect reconstructed with pre-bent plate and mesh and filled with β-tricalcium phosphate scaffold seeded with adipose derived stem cells and recombinant human bone morphogenetic protein-2. Reproduced with permission from Sándor GK, et al. J Oral Maxillofac Surg 2013;71:938-50107

One larger study, which included 14 patients, used rhBMP-2 alone with a collagen carrier and without concomitant bone materials.108 Cases included patients with critical defects as a result of either neoplastic disease or osteonecrosis. Between 4 mg and 8 mg of rhBMP-2 was delivered to the surgical site on a collagen sponge and the defect stabilized with a titanium plate or titanium mesh. Patients were then followed for six to 18 months. All patients showed bony regrowth as early as three to four months, all patients regained continuity, two patients had mesh exposure, and two had dental implants placed at six and eight months. None of these patients had a malignancy, so they did not need postoperative radiation. Likewise, the mandibular periosteum, which is a rich source of mesenchymal stem cells, was preserved in all cases.

In a phase I/II, randomized, controlled feasibility trial ( NCT00755911), a mixed stem and progenitor cell population was transplanted for localized reconstruction of alveolar bone defects. Twelve patients received the cell transplant, and 12 received guided bone therapy as a control. No serious adverse events were reported. The cell transplant treatment accelerated bone regeneration and resulted in a statistically significant reduction in the need for secondary bone grafting.

Salivary gland

Hyposalivation, one of the characteristics of xerostomia or dry mouth, can occur in patients with Sjögren’s syndrome or other autoimmune disorders, in patients with a history of head and neck radiation, or as a result of polypharmacy. Saliva production comes primarily from three pairs of major glands (parotid, submandibular, and sublingual), as well as more than 1000 minor glands in the upper aerodigestive tract.109

Interestingly, beyond the mere secretion of saliva by stem cells derived from the salivary glands, the paracrine activity of the bioactive components secreted by ASCs and BMSCs has been explored as a reparative solution for salivary glands damaged by radiation. This bioactive milieu has paracrine qualities shown to improve the salivary microenvironment by way of epithelial repair, increased microvessel density, and reduced fibrosis.110111112 Additional efforts have explored the therapeutic potential of MSCs to direct differentiation into acinar-like cells both in vivo and in vitro.113

The first clinical trial ( NCT02513238) using MSCs for radiation induced xerostomia was completed in 2018.114115 This trial was single center, phase I/II, randomized, placebo controlled, and double blinded with a total of 30 patients. Autologous ASCs, obtained from liposuction, were isolated and expanded for 14 days. These ASCs or placebo (isotonic sodium chloride with 1% human albumin) were injected into the submandibular gland. The primary outcome was the unstimulated whole salivary (UWS) flow rate measured at one month and four months after injection and compared with a pre-injection baseline. The UWS flow rate increased at one month (33%) and four months (50%) in the ASC group compared with a pre-treatment baseline, whereas the placebo control group showed a 5% decrease at one month and a 0.5% increase in UWS at four months compared with baseline. No adverse events were reported.

Ear and nose cartilage

Cartilages of the ear and nose serve structural and aesthetic roles. Conventional reconstruction approaches for these sites rely on grafted cartilage or pre-made exogenous materials.116117 Regeneration efforts for these cartilages aim to produce a large cartilage volume that is mechanically similar to native cartilage.7 Decades of research have been dedicated to cartilage tissue engineering; here, we describe the early human applications for ear and nose reconstruction.118119

Cultured autologous chondrocytes were used as graft material for nasal reconstruction in a first in human exploratory trial in 2004.120 Chondrocytes were isolated from conchal cartilage, cultured in vitro, and injected into a subcutaneous pocket above the nasal bone in nine patients. No complications were reported, and no absorption was found in follow-up ranging from six months to two years. The same group then used similar autologous chondrocytes from the outer ear cartilage combined with autologous serum for injection into surgical defects of 32 patients with either a depressed cranial deformity or ear, nose, and chin deformities.121 In this exploratory trial, cells were cultured for roughly four weeks between harvest and injection. No complications were reported with the reconstruction, and follow-up patient satisfaction ranged from good to excellent.

A two stage process to amplify graft cartilage has been demonstrated in auricular reconstruction and nasal/chin reconstructions.122123 In the first study, four children (9-10 years old) with microtia were treated in an exploratory trial.122 The authors describe two of the patients as lacking acceptable costal cartilage for conventional reconstruction and all of the patients as wanting to avoid the donor site morbidity of the graft harvest. Instead, chondrocytes were isolated from remnant auricular cartilage (fig 6, A) and cultured in vitro for roughly four weeks. The cultured chondrocytes were then injected into a subcutaneous pocket of fascia in the lower abdomen of the patients, providing an “in vivo bioreactor.” A generated cartilage block was formed subcutaneously, harvested roughly six months later (fig 6, B), and sculpted into the auricular framework for reconstruction (fig 6, C and D). In follow-up ranging from two to five years, no appearance of chondrocyte reabsorption occurred and two of the patients had undergone a split thickness skin graft with ear elevation. No major complications were reported. This group used a similar approach for nasal and chin reconstruction in 18 patients.123 Magnetic resonance imaging was used to visualize blood vessels feeding the area surrounding the developing neocartilage within the patient’s abdomen. No serious complications and one case of partial absorption of the graft were reported.

Fig 6
Fig 6

Images showing generated cartilage graft used for auricular reconstruction. A: harvested cartilage from remnant auricular cartilage. B: cartilage block generated after six months in patient’s abdominal pocket. C: auricular framework sculpted from cartilage block. D: auricular framework in place and covered in skin flap. Reproduced with permission from Yanaga H, et al. Plast Reconstr Surg 2009;124:817-25123

A pre-designed scaffolding approach was used for auricular reconstruction in five patients with microtia in an exploratory study.124 The scaffold, shown in figure 7 (A) was made from biodegradable polymers with a patient specific shape based on the healthy contralateral ear, cultured with autologous chondrocytes (fig 7, B) and covered with a skin graft (fig 7, C). The longest follow-up, at 2.5 years, showed cartilage regeneration without extrusion or deformation of the implant and satisfactory aesthetic outcomes.

Fig 7
Fig 7

Tissue engineered and patient specific auricular cartilage framework. A: pre-shaped scaffolds before assembly and cell seeding. B: auricular framework after in vitro culture. C: auricular framework in place and covered with skin graft. Reproduced with permission from Zhou G, et al. EBioMedicine 2018;28:287-302124

Two exploratory trials have used autologous nasal septal chondrocytes for nasal alae reconstruction in humans.125126 In a first in human exploratory trial reported in 2014, chondrocytes were expanded and cultured with autologous serum on collagen membranes for four weeks. These engineered cartilage grafts were then used for nasal reconstruction in five patients. No adverse events were reported after 12 months. Interestingly, cartilage matrix was not apparent in histologic analysis of biopsies acquired from the reconstructed sites after six months.118

These early exploratory trials show the feasibility of some tissue engineering techniques for ear and nasal cartilage reconstruction. Approaches with autologous chondrocytes can amplify cartilage volume, and scaffolding provides a predetermined implant shape. These approaches remain experimental, and further trials are needed to compare their efficacy with that of conventional reconstructive surgeries.

Tympanic membrane

The tympanic membrane is a thin, three layered structure separating the external ear from the middle ear. Tympanic membrane perforations (TMP) are a common problem in otology, and most acute perforations heal spontaneously.127 However, large perforations may need surgery to close the defect.128 Many regenerative therapies have been evaluated for healing TMP in animal models and humans.127129130131 Given the relative maturity of this field, we discuss here the randomized controlled studies conducted in humans. Recent reviews provide more discussion of preclinical and exploratory clinical studies of tympanic membrane regeneration.127128131132133 A noteworthy challenge in the field is the lack of a standardized animal model for chronic TMP.127134135

Table 1 summarizes the randomized controlled trials of regenerative therapies for TMP. These studies targeted chronic and traumatic TMP and used a combination of scaffolds and growth factors, including epidermal growth factor, PDGF, and bFGF. The evaluated outcomes are typically the rate of TMP closure among patients and time to closure. Generally, healing outcomes with topical applications of growth factors and scaffolds for TMP show mixed results in these small trials, without a clear clinical benefit. A retrospective analysis of bFGF and an atelocollagen sponge treatment for TMP found that the location of the perforation, condition of the margin, and degree of tympanic membrane calcification were significant variables affecting TMP closure rates.144 Consideration of these variables in future clinical studies could improve understanding of the efficacy of treatment.

Table 1

Summary of randomized controlled trials for tympanic membrane regeneration

View this table:


Hearing and balance disorders arise from damage to or loss of the mechanosensitive cells and/or neuronal cells of the inner ear.145146 Hearing aids are used for moderate hearing loss and cochlear implantation for severe or complete loss of hearing. Basic science and preclinical research into regeneration of cochlear tissues is vast and is not covered in detail here.145147148 Strategies in cochlear regeneration have focused on transplantation of stem cells and delivery of growth factors.149

A prospective uncontrolled trial delivered insulin-like growth factor-1 on a gelatin hydrogel to the middle ear of 25 patients with glucocorticoid resistant sudden sensorineural hearing loss.150 No serious adverse events were reported, and improvement in hearing was found in roughly half of the patients.

Two studies have reported intravenous regenerative therapies for sensorineural hearing loss.151152 BMSCs were delivered intravenously in two patients with sensorineural hearing loss in an exploratory trial.151 No significant improvement in hearing was found in either patient, and no complications were reported. In a phase I clinical trial, autologous umbilical cord blood was delivered intravenously for acquired sensorineural hearing loss in 11 children between 6 months and 6 years of age.152 No adverse events were reported, and statistical significance was found in some hearing measures; however, further trials are needed to evaluate efficacy.

Restoration of hearing with hematopoietic stem cell transplant was evaluated in patients with mucopolysaccharidosis.153 Patients with mucopolysaccharidosis have a progressive build-up of glycosaminoglycans in their cells due to an enzyme deficiency, often resulting in some hearing loss. This retrospective chart analysis reviewed 30 patients with mucopolysaccharidosis and sensorineural hearing loss who were treated with hematopoietic stem cell transplant. Some measures showed improvements in hearing in all patients, but patients younger than 25 months showed statistically significant improvement.


Regenerative medicine in Oto-HNS is a rapidly developing field spanning from basic scientific research to clinical trials. The Oto-HNS field is unique in its need to restore vital functions such as breathing and swallowing to patients, as well as sense of self through voice, hearing, and physical appearance. The most substantial progress to date seems to be for indications for which no satisfactory conventional therapies exist, such as long-segment tracheal reconstruction. Major strides have been made in reconstruction of head and neck defects with regenerated cartilage and bone, reducing the need for grafting and providing an opportunity to improve aesthetic outcomes. Regenerating tissues that govern complex functions such as voice and hearing remains challenging owing to the inherent complexity of the tissues. This presents ample opportunity for meaningful future research that saves lives and improves quality of life.

Questions for future research

Basic science

  • Which animal model(s) should be used for evaluating regenerative therapies?

  • How can a patient’s regenerative capacity be characterized for personalized medicine?

  • How can complex tissues and organs be regenerated?


  • How do the safety and efficacy of regenerative medicine approaches compare with current standard of practice in randomized, multicenter, controlled trials?

  • How can biomanufacturing, sterilization, and quality protocols for complex regenerative constructs be standardized?

Clinical practice

  • How can patients’ expectations of regenerative therapies best be managed?

  • How can the isolation, processing, and manufacturing of regenerative medicine products be incorporated into practice?

  • How will reimbursement for novel therapies be handled?


  • Series explanation: State of the Art Reviews are commissioned on the basis of their relevance to academics and specialists in the US and internationally. For this reason they are written predominantly by US authors

  • Contributors: All authors defined the scope of the paper, did the literature search, and wrote and revised the manuscript. DGL is the guarantor.

  • Funding: All authors are supported by the Mayo Clinic Center for Regenerative Medicine.

  • Competing interests: We have read and understood the BMJ policy on declaration of interests and declare the following interests: none.

  • Provenance and peer review: Commissioned; externally peer reviewed.

  • Patient involvement: No patients were involved in developing this manuscript.


View Abstract