Outcomes of the Osteocutaneous Radial Forearm Free Flap for Mandibular Reconstruction

Importance Limited donor and recipient site complications support the osteocutaneous radial forearm free flap (OCRFFF) for mandibular reconstruction as a useful option for single-stage mandibular reconstruction.

Objective To examine and report long-term outcomes and complications at the donor and recipient sites for patients undergoing the OCRFFF for mandibular reconstruction.

Design Retrospective review.

Setting Academic, tertiary care medical center.

Patients The study population comprised 167 consecutive patients who underwent single-staged mandibular reconstruction with an OCRFFF.

Mean Outcome Measures Rates of complications at the donor and recipient sites.

Results The mean patient age was 61 years (range, 20-93 years). Men compromised 68% of the population. Follow-up interval ranged from 2 to 99 months (mean, 25.9 months). The median length of bone harvested was 7 cm (range, 2.5-12.0 cm). Prophylactic plating was completed for each of the radii at the time of harvest. Donor site complications included radial fracture (1 patient [0.5%]), tendon exposure (47 patients [28%]), and donor hand weakness or numbness (13 patients [9%]). Recipient site complications included mandible hardware exposure (29 patients [17%]), mandible nonunion or malunion (4 patients [2%]), and mandible bone or hardware fracture (4 patients [2%]). Using regression analysis, we found that patients were 1.3 times more likely to have plate exposure for every increase of 1 cm of bone harvest length; this was statistically significant (P = .04).

Conclusions and Relevance This is the largest single study reporting outcomes and complications for patients undergoing OCRFFF for mandibular reconstruction. Prophylactic plating of the donor radius has nearly eliminated the risk of pathologic radial bone fractures. Limited long-term donor and recipient site complications support the use of this flap for single-stage mandibular reconstruction.

Mandibular reconstruction second to malignancy, osteoradionecrosis, or trauma remains a challenge for head and neck surgeons. Restoration of function including speech and swallow, while limiting long-term morbidity, are primary objectives. Restoring 3-dimensional anatomical relationships and mandibular continuity are crucial in achieving these goals. Traditionally, the fibula free flap has been the mainstay for mandibular reconstruction, with its increased length and quality of bone stock.¹ One of its counterparts, the osteocutaneous radial forearm free flap (OCRFFF), was first described in 1974.² Initially, this flap showed great promise for oromandibular reconstruction. Specific advantages are well known and include ease of patient positioning during the harvest; a long vascular pedicle; predictable vascular anatomy; reliable skin paddle; and a thin, pliable, soft-tissue component. This thin and pliable soft-tissue component allows for superior intraoral surface contouring when reconstructing the labiomandibular sulcus, resulting in improved tongue mobility and dental rehabilitation. Limitations include the length of bone stock available with the OCRFFF compared with the fibula flap as well as the historic reports of increased donor site morbidity, specifically pathologic radial fractures.

In the late 1990s, this flap fell out of popular use following multiple published series reporting increased rates of pathologic radius fractures occurring at the donor site.^3-6 In an attempt to address this increased rate of donor site fractures, prophylactic plating of the donor radius with a 3.5-mm dynamic compression plate was introduced by orthopedic surgeons and otolaryngologists at the University of Kansas Medical Center.⁷ With the use of prophylactic plating of the donor radius, previous studies have shown a reduction in incidence of donor site morbidity and radius fracture.^8-13 This study was designed to further assess the long-term donor and recipient site complications of patients undergoing OCRFFF for mandibular reconstruction.

A retrospective medical chart review was undertaken for patients undergoing OCRFFF between January 1, 2000, and December 31, 2010, at the University of Kansas Hospital. Institutional review board approval was obtained. The patient list was generated by searching Current Procedural Terminology (CPT) codes 20969 (osteocutaneous free flap microvascular reconstruction) and 29125 (static short arm splint). We excluded patients with less than 2 months of documented clinical follow-up. The final study number was 167 patients. Data points collected for each patient included demographic data and procedural specifics. Outcome measures included rates of complications at the donor and recipient sites. Data variables collected for each patient are summarized as follows:

• Demographics/preoperative: Age, sex, tobacco abuse, alcohol abuse, preoperative radiography, and primary etiology

• Complications: Recipient site and donor site

• Procedural specifics: Skin paddle area, radius harvest length, mandibular defect location, mandible plate, radial plate, tourniquet time, ischemia time, and operative room time.

Data were analyzed using χ², Fisher exact test, and regression analyses to assess continuous and categorical data.

Our institution's technique for radius harvest involves harvesting 50% of the radius width, with keel-shaped osteotomies on either side. The length of radial bone harvested is determined by the size of the mandibular defect, although limited by the insertion points of the brachioradialis and pronator teres tendons. We have found that 12 cm is most consistently the maximal length of radial bone available for harvest. We plate the donor radius with an 8- to 14-hole low-contact dynamic compression plate. The plate is positioned over the dorsal side of the radius and bent to the contour of the bone. Two bicortical screws are placed distally, 3 bicortical screws are placed proximally. Initially, monocortical screws were placed into the remaining bony defect. After additional subclinical radial fractures were noted using this technique, it was believed that these screws were potentially weakening the osteotomized segment, and this technique was abandoned. Currently, no screws are placed in the remaining bony defect. The overlying forearm muscles are repositioned over the plate, and a split-thickness skin graft harvested from the thigh is placed over the distal forearm defect. Postoperatively, the patient is placed in a rigid ulnar gutter splint for 5 days.

A total of 167 patients underwent OCRFFF for single-stage mandibular reconstruction. The mean patient age was 61 years, ranging from 20 to 93 years. Men compromised a majority of the patient population (68%). A notable number (61%) of patients had a history of smoking, and 47% had a history of radiation therapy. Follow-up data ranged from 2 to 99 months, with a mean of 25.9 months. The most common indication for surgery at the recipient site was squamous cell carcinoma (77%). Other causes included osteoradionecrosis (11%), trauma (3%), and odontogenic tumors (3%). A majority of the primary cancer sites were in the oral cavity, with tumors originating along the alveolar ridge being the most common (29%). The median length of bone harvested was 7 cm (range, 2.5-12.0 cm). Prophylactic plating was completed for each of the radii at the time of harvest with a 3.5-mm low-contact dynamic compression plate. The most common plate used for mandibular reconstruction was a 2.0-mm locking-reconstruction plate (47% of patients); other plates used included a 2.4-mm locking reconstruction plate (LRP) (27%), a 2.5-mm LRP (7%), and a 2.8-mm LRP (19%). The harvested radial bone underwent between 0 to 2 osteotomies to obtain appropriate mandibular contour for reconstruction.

Our donor site complications are summarized in Table 1. The most common donor site complication was tendon exposure, occurring in 28% of patients. All of these patients were treated with local wound care including wet-to-dry dressing changes. Only 1 patient required operative intervention for tendon excision after developing an infection. There was only 1 radial fracture (0.5%). The fracture occurred at the distal one-third of the radius after 2 bouts of methicillin-resistant Staphylococcus aureus cellulitis and osteomyelitis. Ultimately, the radial plate was removed, and 10 months later, the patient was treated with open reduction and internal fixation and iliac crest bone grafting. He has no resulting functional deficits.

Complications occurring at the recipient site are summarized in Table 2. The most common was mandible hardware exposure occurring in 17% of patients. The patients with mandibular hardware exposure were not without significant risk factors for poor wound healing. Of these patients, 83% had either preoperative or postoperative radiation therapy, and 79% had a history of smoking. Of the 29 patients who had hardware exposure, 19 ultimately underwent surgical intervention, most commonly with partial or complete removal of the mandibular plate. The remaining 10 patients were treated conservatively with local wound care and occasionally hyperbaric oxygen therapy. Using a regression analysis on our data, we also did not find age, sex, chemotherapy, plate size, or history of radiation therapy to be predictive of hardware exposure. We found that patients with a larger amount of bone harvested were more likely to have eventual plate exposure. The odds ratio was 1.32, suggesting that patients are 1.3 times more likely to have plate exposure for every 1-cm increase of bone harvest length; this was statistically significant (P = .04). History of smoking did not reach statistical significance (P = .06). Only 2% of patients had either mandibular malunion or nonunion. Likewise, only 2% of patients had either mandibular hardware or bone fractures. Our flap failure rate was 1%, occurring in only 2 patients, making our overall flap success rate 99%.

After first being introduced in 1974, the OCRFFF showed great promise for oromandibular reconstruction.² In the late 1990s, this flap fell out of popular favor after multiple published series reported frequent pathologic donor radius fractures. Incidence of this complication was reported in 25% of published reports and ranged from 0% to 67% of patients after undergoing OCRFFF.^3-6 In an attempt to reduce the risk of donor radius fracture, prophylactic compression plate fixation of the radius was introduced. This technique was initially evaluated using a biomechanical study assessing the strength of the plated cadaveric bone after ostectomy and compared with unplated matched cadaveric radii. The plate was shown to significantly increase the strength of donor radii after ostectomy.⁷ Reports following the introduction of prophylactic plating showed substantial reduction in incidence of pathologic radial bone fracture, particularly after plating screws were no longer being placed into the remaining bony defect.⁸ Prophylactic plating is now routinely performed at our institution after harvest for OCRFFF.

This study retrospectively reviewed our institution's outcomes over the past 10 years for 167 patients undergoing OCRFFF for mandibular reconstruction, to assess for both donor and recipient site morbidity. We limited our study to patients who had at least 2 months of documented clinical follow-up. We found limited long-term complications occurring at both the donor and recipient site, with only 1 radial fracture (0.5%). We are aware of several other articles assessing the outcomes and complications for patients undergoing an OCRFFF.^9-13 Our complication rates, and those previously reported in the literature, are summarized in Table 3. A previous publication by Militsakh et al¹⁰ reports a retrospective review of patients from the University of Kansas Medical Center undergoing OCRFFF. Our 2 studies overlap slightly (2000-2003); 50 patients were included in the present study and in the previously published article.¹⁰ These patients' were rereviewed using our study's criteria described in the “Methods” section, with the goal of extending follow-up time and adding additional data not previously collected. Complications among these 50 patients are specifically outlined in Table 3.

Our highest recipient site complication was mandibular hardware exposure, which occurred in 17% of patients. This was similar to hardware exposure rates recently published by Sinclair et al⁹ and Militsakh et al,¹⁰ who reported plate exposure in 15.6% and 16%, respectively. Although these rates are similar among recent articles assessing outcomes of the OCRFFF, they are still higher than the generally quoted rates for fibula free flap.^1,14 Speculations as to the causes for increased plate exposure have been made, including different sizes of mandibular plates. Militsakh et al¹⁰ showed a lower complication rate when a 2.0-mm LRP was used, which was statistically significant; however, this trend was not seen in the study by Sinclair et al⁹ or in our data. This patient population certainly has increased risk factors for poor wound healing, with a majority of patients having history of smoking and/or radiation therapy. Sinclair et al⁹ found that radiation, chemotherapy, diabetes mellitus, age older than 60 years, size of plate, and sex were not predictive of plate extrusion.⁹ Using a regression analysis on our data, we also did not find age, sex, chemotherapy, plate size, or history of radiation therapy to predict higher rates of hardware exposure. We did, however, find that patients with a larger amount of bone harvested were more likely to have plate exposure, which was statistically significant (P = .04). Ultimately, our statistics were limited by the smaller number of patients with hardware exposure; larger studies are needed to further delineate the actual predictive nature of these variables.

Our technique for closure around the plate involves a single-layer intraoral closure of flap epithelium to the mucosa with Vicryl suture (Ethicon Inc), an external multilayer closure with Vicryl suture reapproximating the platysma and deep dermal layers, and skin closure with running Prolene suture (Ethicon Inc). Comparisons with other institutions' plating and closure techniques may be useful to see if certain methods were found to be helpful in preventing mandibular plate exposure.

Although this study did not directly compare the OCRFFF with other osteocutaneous free flaps, several other reports have been published that not only illustrate comparable complication rates but also highlight specific advantages that are unique to this flap.^9-13 Militsakh et al¹⁰ found long-term donor and recipient site morbidity and postoperative function of the OCRFFF were comparable to those of other commonly used osteocutaneous free flaps. Advantages were decreased operative time and intensive care unit stay, which reached statistical significance.¹⁰ Virgin et al¹³ found that OCRFFF had comparable functional outcomes at recipient sites between fibula free flap and OCRFFF. This study reported that patients with fibula free flap had comparable rates of flap failure and malunion and actually had a greater number of donor site complications compared with patients who underwent OCRFFF.¹³

Despite the known advantages, and seemingly comparable complication rates, there is still some reluctance to use the OCRFFF for single-staged mandibular reconstruction. One criticism of this flap is that its bone stock is not as structurally sturdy as the fibula, potentially unable to withstand the forces of mastication with resulting malunion. Several studies with large population numbers have shown limited incidence of malunion or nonunion (3.2%⁹ and 2%¹³), as we have in this study (2%), which are comparable to rates for fibula free flap. An additional concern with regard to the strength of the bone is its potential inability to accommodate osseointegrated implants. Although we did not look specifically at this in our study, Virgin et al¹³ actually had more patients receive osseointegrated implants after OCRFFF; of note, all of these patients had augmentation with iliac crest bone grafting. In addition, there is a concern for donor site morbidity associated with hand and wrist immobility after harvest of the donor radius. Two studies have used the Disability of the Arm, Shoulder, and Hand (DASH)¹⁵ questionnaire to obtain subjective data from patients regarding their presumed disability after this procedure. Deleyiannis et al,¹⁶ in 2008, showed objective reductions in wrist, forearm, and thumb range of motion are frequent after harvest of an OCRFFF and that decreased wrist range of motion had the greatest impact on patient self-report of disability.¹⁶ Sinclair et al¹³ recently compared DASH scores of those undergoing OCRFFF with those with fasciocutaneous radial forearm free flap. They found that mean DASH scores were not statistically different between these 2 groups and concluded that mild wrist weakness and stiffness are common after OCRFFF but do not impede ability to perform activities of daily living.⁹

In conclusion, this is the largest single study reporting outcomes and complications for patients undergoing OCRFFF for mandibular reconstruction. We have shown that harvest of 50% of the donor radius width with up to 12 cm of length can safely be performed without increased morbidity. Prophylactic plating of the donor radius has nearly eliminated the risk of pathologic radial bone fractures. This study shows limited long-term donor and recipient site complications, supporting this flap as a useful option for single-stage mandibular reconstruction.

Correspondence: Jill M. Arganbright, MD, Department of Otolaryngology–Head and Neck Surgery, University of Kansas Medical Center, 3901 Rainbow Blvd, MS 3010, Kansas City, KS 66160 (jarganbright@kumc.edu).

Submitted for Publication: August 21, 2012; final revision received October 10, 2012; accepted November 5, 2012.

Author Contributions: Drs Arganbright and Shnayder had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Arganbright, Tsue, Girod, Sykes, and Shnayder. Acquisition of data: Arganbright, Militsakh, and Markey. Analysis and interpretation of data: Arganbright, Girod, Sykes, Markey,and Shnayder. Drafting of the manuscript: Arganbright and Markey. Critical revision of the manuscript for important intellectual content: Arganbright, Tsue, Girod, Militsakh, Sykes, and Shnayder. Statistical analysis: Arganbright and Sykes. Administrative, technical, and material support: Girod, Militsakh, Sykes, Markey, and Shnayder. Study supervision: Tsue, Girod, Militsakh, and Shnayder.

Previous Presentation: This study was presented orally at the American Head and Neck Society Eighth International Conference on Head and Neck Cancer; July 22, 2012; Toronto, Ontario, Canada.