A review of advanced head and neck osteoradionecrosis

Osteoradionecrosis (ORN) of the head and neck can be a devastating complication following radiation therapy. ORN is associated with pain, chronic infection, and non-healing wounds. Radiation fibrosis, chronic infection, fistula formation, and necrotic tissues can make treatment challenging. The following review article is a narrative on the management of advanced head and ORN.


INTRODUCTION
The majority of patients with advanced head and neck cancer will receive radiation as part of their treatment [1] . For a subset of patients, radiation treatment has devasting side effects on the surrounding noncancerous tissues. One of the most crippling outcomes is the development of osteoradionecrosis (ORN) in the head and neck region. These cases can range from mildly symptomatic areas of exposed bone to pathological fractures, often leaving patients with disabling symptoms. The current understanding of ORN pathophysiology suggests a progression of hypovascular-hypoxic-hypocellular tissue. Changes in the metabolic homeostasis following radiation eventually progress to a state of hypoxia and hypovascularity. This ultimately leads to tissue breakdown and a non-healing wound [2] .
The mean duration from completion of radiation therapy to the development of head and neck ORN is estimated at 22-47 months [3] . Dental extractions are commonly found to be a precipitating factor, with some studies noting a recent dental extraction in 50%-60% of cases [4,5] . The incidence of head and neck ORN throughout the literature ranges from 3%-15% [3,6] . While ORN can occur in multiple head and neck subsites, the mandible is the most common location [7] . The treatment of ORN is dependent on the severity of symptoms. The following narrative focuses on patients with advanced head and neck ORN with an emphasis on refractory cases requiring a free flap reconstruction.

CLINICAL PRESENTATION
Patients with advanced ORN commonly present with pain and exposed bone [ Figure 1]. Patients with malignancy can also present with pain; therefore, an underlying malignancy should be ruled out prior to the start of ORN treatment. As ORN progresses, the bone may experience loss of density and strength, resulting in a pathologic fracture and/or orocutaneous fistula.
On panorex, early stages of ORN may present with findings of sclerotic bone or a poorly defined radiolucent lesion. Panorex may also show findings of cortication loss following dental extraction [8] .
Common computed tomography (CT) findings of ORN include cortical defect or lucency, disorganized bony architecture, intraosseous air, and ultimately pathologic fracture [9] . When considering surgical intervention, CT imaging with 1mm cuts of the maxillofacial skeleton is recommended [10] .

NON-SURGICAL MANAGEMENT
ORN treatment typically begins conservatively, with free flap reconstruction being reserved for refractory and advanced cases [11] . Conservative therapy for the early-stage disease includes optimizing oral hygiene, eliminating dental disease, and the use of systemic antibiotics [12] . While often non-curative, these conservative interventions can provide symptomatic relief and slow progression. Antibiotics are commonly administered for acute infections or in the setting of chronically draining orocutaneous fistulas [ Table 1].
Hyperbaric oxygen (HBO) therapy was introduced as a possible treatment for ORN in 1983 by Marx [13] . HBO therapy works by increasing local tissue oxygen concentrations, thereby promoting tissue epithelialization and bone regeneration. HBO therapy is sometimes used prophylactically in patients requiring dental extractions after radiation to potentially prevent ORN development. However, a randomized controlled trial of previously radiated patients requiring dental extraction found HBO therapy before dental trauma did not prevent ORN complications [14] . Currently, there are limited data demonstrating efficacy for the use of HBO therapy in the treatment of ORN [ Table 2] [15][16][17] .
PENTOCLO is an antioxidant therapy that consists of pentoxifylline, tocopherol, and clodronate. Previous publications suggest improved wound healing when patients with ORN were administered PENTOCLO [18,19] . Pentoxifylline is thought to improve microcirculation, while tocopherol (vitamin E) acts as an antioxidant. Clodranate is a first-generation bisphosphonate that reduces osteoclast activity and stimulates osteoblasts. A retrospective study found that patients who received pentoxifylline and tocopherol after radiation had a lower incidence of ORN [20] . Dissard et al. [21] also found that administration of the PENTOCLO regimen daily had a low side effect profile with high rates of symptom improvement when given antibiotics and steroids.

SURGICAL MANAGEMENT
ORN has significant quality of life (QOL) implications, including pain, infection, draining fistulas, and/or pathologic fractures. The decision to proceed with more invasive treatments is typically dictated by the Table 1. Summary of literature evaluating the efficacy of antibiotic therapy in radiated patients receiving dental extraction or implantation

Ref. Design n Treatment Outcomes Conclusion
Sandhu et al. [55] (2020) Al-Bazie et al. [56] (2016) RCT 89 Ten days of Amoxicillin 500 mg q8h and chlorhexidine before extractions No reported cases of ORN at a mean of 63 months Perioperative Abx with antibiotic mouthwashes are effective in preventing ORN after extractions Sultan et al. [15] (

HBO for dental extraction/implantation is unnecessary
Bennett et al. [58] ( patient's symptomology and the subsequent impact on QOL. Conservative surgical measures can include debridement, sequestrectomy, local tissue rearrangement, and marginal mandibulectomy. If conservative medical and surgical management fails to provide resolution or symptom relief, more invasive surgical interventions can be considered. Definitive surgery typically involves resection of the involved bone and soft tissues. The resulting defect often requires reconstruction with vascularized non-radiated tissue, typically in the form of a free flap [22][23][24][25][26] .

FREE FLAP DONOR SITES
When selecting a donor site for ORN reconstruction, one must consider the amount of soft tissue required and the length of bone needed for continuity.

Fibula
The fibula is the donor site used most commonly for free flap reconstruction of head and neck ORN. Advantages of the fibula free flap include a long segment of bone stock (22-25 cm), adequate pedicle length for tension-free anastomosis in the neck, and low donor site morbidity [27] . The fibula is harvested as a long segment of vascularized bone which allows for multiple osteotomies if needed to restore the natural contouring of the mandible or midface. The fibula has a high volume of cortical bone to bear the forces of mastication. Additionally, the bicortical bone of the fibula allows for dental implantation. The associated fasciocutaneous paddle has reliable perforators and can be harvested with the bone if coverage is needed for a mucosal or soft tissue defect. However, patients with severe peripheral vascular disease or those lacking three-vessel (anterior tibial, posterior tibial, peroneal) arterial runoff of the lower extremity may not be candidates for fibula reconstruction.

Scapula
In cases where the fibula is contraindicated or a large component of soft tissue is required, the scapula free flap is an ideal donor site. The lateral border of the scapula can reliably provide 10 cm of bone. Compared to the fibula, the scapula may have a thinner bone. The subscapular system often has a shorter pedicle length. However, the scapula tip is supplied by the angular artery, which originates from the thoracodorsal system. It has been reported that the angular artery is able to supply up to 10 cm of the lateral scapula border [28] . As a result, incorporation of the angular artery allows for increased pedicle length. The subscapular system can be extremely versatile, providing skin, muscle, and bone in a multitude of combinations. The latissimus can be incorporated and utilized for intraoral or external plate coverage. An additional chimeric component of rib and serratus muscle can be harvested with the serratus branch of the thoracodorsal vessels. The versatility of soft tissue combinations and bulk makes the scapular flap ideal for complex soft tissue reconstructions requiring a large volume of soft tissue and a shorter linear segment of bone [29] .
Dental implantation is also possible with scapula bone, and success rates are similar to the fibula [28] . In addition, the scapula may be advantageous in elderly patients who have significant peripheral vascular disease. Additionally, patients tend to ambulate sooner, possibly reducing postoperative complications [30] . The main disadvantages of the scapula are the shorter pedicle length, longer operative time, and intraoperative patient positioning.

Osteocutaneous radial forearm
The osteocutaneous radial forearm free flap (OCRFFF) is another less commonly used option to restore bony continuity. Similar to the scapula, the OCRFFF provides 10-12 cm of bone. However, the entire diameter of the radius bone cannot be harvested, limiting it to a single cortex following the harvest of 40%-60% of the radius circumference. Subsequently, there is insufficient bone stock to support dental implantation and necessitates prophylactic plating the radius to prevent fracture [31,32] . In addition, the fasciocutaneous tissue volume available is often less than the soft tissue available with the fibula.

Fasciocutaneous and myocutaneous only donor sites
In patients who do not require a vascularized segment of bone, a radial forearm or anterolateral thigh flap may be utilized. These flaps are applicable in ORN of the anterior skull base or lateral temporal bone for soft tissue coverage [33] . The radial forearm free flap is a fasciocutaneous flap that is thin and pliable, with reliable pedicle length and high rates of flap survival. However, this flap contains limited subcutaneous adipose in the majority of patients and lacks a muscle component making it inadequate for defects with large volume loss. Conversely, the anterolateral thigh free flap (ALT) may be composed of a combination of skin, adipose, fascia, or muscle. Adipose allows for improved volume retention as it is less subject to atrophy and contracture over time. However, depending on the thickness of the ALT subcutaneous adipose, it may provide more bulk than desired.

OPERATIVE CONSIDERATIONS
When reconstructing with an osseous donor site, consideration of bone segment length is important. For the periosteum to provide sufficient blood study, it is recommended that each bone segment measure at least 2 cm in length. The ability to create multiple osteotomies can improve facial contouring. However, it also reduces the length of the pedicle. This can create additional challenges in vessel depleted, radiated necks where contralateral vessels or vessels in the base of the neck (transverse cervical, dorsal scapular, or internal mammary) may necessitate a longer pedicle length.

VIRTUAL SURGICAL PLANNING
Virtual surgical planning (VSP) can be utilized to aid in flap design preoperatively. VSP allows for the creation of patient-specific cutting guides and plates, which may increase free flap accuracy and reduce operative times [34,35] . The use of VSP is thought to improve bone-to-bone contact, reducing rates of malunion or nonunion and the subsequent sequalae [36,37] . The VSP software also provides information regarding the thickness of bone to aid in selecting sites for screw placement, ensuring adequate bone thickness, and reducing screw mobility and subsequent hardware extrusion [36][37][38] . VSP can also be used to mirror the healthy bone allowing for more symmetric reconstruction that more accurately reflects the anatomical positioning prior to developing ORN. Additionally, VSP can aid in planning the ideal orientation of the pedicle [39] .

TECHNICAL CONSIDERATIONS
Flap geometry and vessel selection can be quite challenging in patients with ORN. Tissue damage and fibrosis from prior radiation therapy and chronic infection can complicate dissection of recipient vasculature and impact the vessel wall integrity and caliber. A previous publication found that over half of free flap reconstructions for ORN required the use of the contralateral neck vessels [17] . Similarly, we have found improved free flap survival rates when using contralateral neck vessels [40] .
Reconstruction bar thickness varies among surgeons, ranging from 1 to 3 mm and averages 2.1 mm. An example of a 2.0 mm reconstruction bar is shown in Figure 2A. Recent data suggests that reconstruction bar thickness may correlate with overlying soft tissue loss and hardware exposure rates, with a higher incidence of complications as reconstruction bar thickness increases [40] . Another variation in surgical design includes orientation and placement of the soft tissue component. Orientation of the fasciocutaneous paddle, in particular when a fibula donor site is used, can either be medial to the reconstructed bone segment or lateral to the reconstructed segment. When the soft tissue is oriented lateral to the hardware, it may provide an additional layer of soft tissue coverage over the reconstruction hardware. An example of a fasciocutaneous paddle that was brought medially into the floor of the mouth is depicted in Figure 2B. Interestingly, the medial orientation of the fasciocutaneous paddle was found to result in a lower incidence of postoperative bone exposure and required fewer postoperative tissue debridement and local flaps procedures [40] . The authors hypothesize that the medial orientation for the fasciocutaneous paddle reduces tension on the perforating vessels and provides protection from vascular insufficiency to the soft tissues [40] .
Dental implantation can significantly improve a patient's quality of life. Dental implantation success in ORN reconstruction has been cited around 95% [41] . The literature cites similar complications rates with primary vs. secondary implantation in the setting of ORN. Secondary implantation has been found to correlate with higher fixed costs [41] , while primary implantation leads to a faster return of oral intake [41] . Implant success rates were found to be similar for fibula vs. scapula free flaps [42] .

QUALITY OF LIFE
Pain is thought to be a large contributor to poor quality of life for patients with advanced ORN [23,43] . Exposed nerve endings are subjected to infection and inflammation within a non-healing ORN wound. Free flaps are advantageous in their ability to provide coverage of exposed nerves and improve the blood flow to the area. One prospective study found a consistent reduction in pain-related domains following surgical resection and immediate free flap reconstruction of ORN, leading the authors to conclude that free flap reconstruction of advanced ORN improves QOL [43] .
In addition to pain, other QOL variables of concern for patients included speech, chewing, swallowing, and appearance [24,[44][45][46] . The University of Washington Quality of Life (UW-QOL) survey is commonly utilized for the assessment of QOL following head and neck reconstruction with a free flap [23,24,43,[46][47][48] . It was previously found that UW-QOL domain scores following free flap reconstruction for advanced ORN were higher for those patients who received dental implants, did not have a history of prior head and neck surgery, and did not develop a cancer recurrence [22] . In contrast, Sweeny et al. [22] found UW-QOL domain scores were not impacted by ORN recurrence, anatomic subsite of the ORN, or donor tissue used for the reconstruction. A summary of QOL outcomes can be found in Table 3.

Ref. Design n Treatment Outcomes Conclusion
Sweeny et al. [22] (2021) Retrospective 137 UW-QOL survey in patients following free flap reconstruction for ORN 45% reported no pain, 28% no swallowing abnormalities, 93% no speech difficulty Data suggests a good return of function and QOL following surgery Lofstrand et al. [26] (2018) Retrospective 41 SF-36, EORTC QLQ-C30, and QLQ-H&N35 questionnaires in cancer vs. ORN patients ORN group had lower scores in swallowing and social eating compared to cancer, but general QOL did not differ from the reference population Cancer and ORN patients have similar QOL following reconstruction with the exception of swallowing/social eating Jacobson et al. [45] (2013) Retrospective 30 PSS, SHI, QLQ-H&N35, and EAT-10 surveys in ORN after reconstruction 89% had abnormal EAT-10 and SHI scores following reconstruction, indicating abnormal speech and swallow Many patients remain unhappy with speech and swallowing outcomes following reconstruction Wang et al. [24] ( An important component of quality of life includes nutritional status. At 3 months following free flap reconstruction for advanced ORN, the rates of feeding tube dependence ranged from 13%-16% [22,44] . When compared to their preoperative nutritional status, 47% of patients were tolerating a regular diet at 5 years following free flap reconstruction, and 31% had improvement in their diet status following free flap reconstructive surgery [22] . This data suggest that for a subset of patients, free flap reconstruction can lead to an improvement in diet function.

SURGICAL COMPLICATIONS
Complication rates following free flap reconstruction of head and neck ORN are cited between 30%-60% [16,26,44,46,49,50] . While most institutions cite free flap survival rates following head and neck reconstruction as 95% or greater, free flap survival rates following ORN reconstruction are cited at 89-93% [26,46,50,51] . Additionally, it was found that patients undergoing free flap reconstruction for ORN had a higher incidence of late complications compared to patients undergoing free flap reconstruction for malignancy [3] . A retrospective study of 277 patients found that 24% of patients developed a postoperative fistula, 16% developed exposed bone, and 20% developed plate extrusion following free flap reconstruction for ORN [40] . These complications are attributed to poor tissue quality (radiation fibrosis, decreased perfusion) and a chronic inflammatory reaction in response to persistent infection and saliva exposure. Although uncommon, these complications can result in increased patient morbidity and healthcare costs [50,52,53] .

ORN RECURRENCE
ORN recurrence following free flap reconstruction is cited at 10%-14%, with a median time to onset of 11 months [17,22] . Poor wound healing and failure of osseointegration postoperatively were found to correlate with higher rates of ORN recurrence [22] . While the donor site selected for free flap reconstruction did not impact the recurrence of ORN [54] . To avoid ORN recurrence at the surgical site, every attempt should be made to resect necrotic non-viable bone. The periosteum of the remaining bone should be inspected to ensure it is viable. The periosteum is a dense fibrous membrane with a rich vascular supply that envelopes the outer cortex of the bone and provides the superficial cortex with nutrients. The cortex should also be evaluated for viability. The cortex of healthy bone is dense, typically white in color, and will have visible bright red bleeding from trans-cortical capillaries. In addition, the bone marrow should be evaluated for viability. Healthy bone marrow is able to maintain its trabecular structure and will bleed bright red. Any bone which does not appear viable should be removed until viable bone is confirmed.

CONCLUSIONS
ORN is one of the most crippling complications following radiation for head and neck malignancies. The development of ORN has a significant impact on quality of life, leaving patients with disabling pain and chronic wounds. Free flap reconstruction is an integral part of the definitive management of advanced ORN. Although often effective at treating ORN, there remains a risk for postoperative wound complications and functional decline. Informed discussions with patients regarding expectations and anticipated outcomes and careful preoperative planning are essential in treating these complex cases.