Hardware complications in oromandibular defects: Comparing scapular and fibular based free flap reconstructions

Oral Oncology 71 (2017) 163–168

Contents lists available at ScienceDirect

Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology

Hardware complications in oromandibular defects: Comparing scapular and fibular based free flap reconstructions Gordon F.Z. Tsang a,b,⇑, Han Zhang a, Christopher Yao b, Mirko Kolarski b, Patrick J. Gullane a, Jonathan C. Irish a, Dale H. Brown a, Douglas B. Chepeha a, David P. Goldstein a, Ralph W. Gilbert a, John R. de Almeida a,⇑ a b

Department of Otolaryngology-Head & Neck Surgery/Surgical Oncology, Princess Margaret Cancer Centre, University Hospital Network, Toronto, Ontario, Canada Department of Otolaryngology-Head & Neck Surgery, Toronto, Ontario, Canada

a r t i c l e

i n f o

Article history: Received 9 April 2017 Received in revised form 3 June 2017 Accepted 19 June 2017

Keywords: Oromandibular reconstruction Free flap Squamous cell Carcinoma Hardware Complication Mandibular reconstruction

a b s t r a c t Background: Despite improvements in surgical technique and technology, hardware complications occur relatively frequently. This study analyzes hardware complications in patients undergoing oromandibular reconstruction using scapular (SFF) or fibular (FFF) free flaps. Methods: Retrospective data for 178 patients was obtained (1999–2014) at University Hospital Network (Toronto, Canada). Univariable and multivariable analyses were performed to identify risk factors for hardware complications. Results: Patients with FFF reconstruction (n = 129) had significantly more hardware complications than those with SFF (n = 49) (16% vs. 2%;p = 0.01). Surgical site infection (SSI) (OR = 7.05; p < 0.01), defect type (OR = 2.63; p < 0.01) and flap (OR = 0.12; p = 0.01) were significant predictors of hardware complications on univariable analysis. Flap type (OR = 0.12; p = 0.04) was an independent predictor of plate complication after adjusting for SSI. A subgroup analysis suggested a trend towards fewer hardware complications with SFF stratified by mandibular defect type. Conclusions: Scapular free flaps are associated with a lower rate of hardware-related complications in oromandibular reconstruction. Ó 2017 Elsevier Ltd. All rights reserved.

Introduction The advent of microvascular free tissue transfer has revolutionized oromandibular reconstruction. The combination of osseocutaneous or osseomyogeneous free flaps in addition to advances in instrumentation with locking screw technology and low profile plates has greatly improved the functional and cosmetic outcomes for patients with mandibular defects [1]. Despite these advances, hardware complications remains a significant challenge with roughly 15% of patients experiencing hardware related complications [2]. Plate exposure, plate fracture, plate infection, and loose screws are some of the more common hardware complications which in turn can lead to further operative procedures, prolonged ⇑ Corresponding authors at: Department of Otolaryngology-Head & Neck Surgery, University of Toronto, St. George Campus, 190 Elizabeth Street, Rm 3S-438, TGH RFE Building, Toronto, Ontario M5G 2C4, Canada (G.F.Z. Tsang). Princess Margaret Hospital, 610 University Avenue, 3-955, Toronto, Ontario M5G 2M9, Canada (J.R. de Almeida). E-mail addresses: [email protected] (G.F.Z. Tsang), john.dealmeida@ uhn.ca (J.R. de Almeida). http://dx.doi.org/10.1016/j.oraloncology.2017.06.020 1368-8375/Ó 2017 Elsevier Ltd. All rights reserved.

antibiotic use, and reduced quality of life (Fig. 1) [3]. Many factors such as previous radiation therapy, smoking status, diabetes, and types of hardware have been identified as factors that can contribute to hardware complications [3–6]. Although there are well established patient and disease related risk factors that are associated with hardware related complications, there is a paucity of data on donor site choice. Contemporary free tissue choices for oromandibular reconstruction often include fibula (FFF), iliac crest, osseocutaneous radial foream osseocutanous (OCRFFF), and the scapular system (SFF) [7]. Some studies comparing the FFF and OCRFFF have reported an increased risk of hardware complications with OCRFFF reconstructions [6,10,11]. No studies to date, however, have compared complications between the FFF and SFF. The FFF is commonly considered the standard in many centers for oromandibular reconstruction due to its length of bone, caliber of bone stock for dental rehabilitation, predictable anatomy, pedicle length, and ability to harvest simultaneous with the ablative procedure [8,9]. It is, however, limited in patients with large softtissue defects, advanced age, peripheral vascular disease, and with

164

G.F.Z. Tsang et al. / Oral Oncology 71 (2017) 163–168

and Gorham’s syndrome), (2) total flap loss including the osseous segment (3) incomplete records, (4) plate exposure secondary to tumor recurrence. In the comparison of patients undergoing SFF or FFF reconstruction, patients with external skin involvement in their primary resections were excluded for this part of the analysis. Total flap loss was excluded from data collection because these patients (n = 5) either required reconstruction with another flap or left hospital with hardware exposed secondary to the flap loss. Data collection

Fig. 1. Hardware complication (plate exposure) in patient who has undergone previous oromandibular reconstruction for squamous cell carcinoma of the oral cavity.

pre-existing ambulatory limitations. The osseous SFF can be harvested either based on the circumflex scapular vessels or based on the angular branch of the thoracodorsal artery, the latter of which provides longer pedicle length and are generally spared of microvascular disease. The SFF similarly provides good bone stock particularly when the crest of the scapula is utilized, a large volume of soft tissue particularly when chimeric flaps are utilized. This system of flaps, however, is limited by the inability to perform simultaneous harvest as well as the length of available bone. Overall, the soft tissue abundance combined with great versatility makes the SFF an excellent potential donor site for oromandibular reconstruction particularly to augment soft tissue defects that may predispose patients to hardware related complications such as plate exposure. This study aims to compare the risk of hardware complications between patients reconstructed with FFF and SFF for oromandibular defects.

Demographic, complications, hardware information, and clinicopathologic data was extracted from operative, clinical, and pathology notes as well as operating room instrumentation inventory datasets. Variables collected include: age adjusted Charlson Comorbidity Index (CCI), surgical site infection, presence of intraoral or external wound dehiscence, radiation exposure, diabetes, active smoking history, and hardware (plate profile height) details. Age adjusted-Charlson Comorbidity Index (CCI) scores, were calculated using relevant comorbidities [12]. Surgical site infection (SSI) were defined using the Center for Disease Control criteria and categorized based on perioperative clinical data [13,14]. Intraoral or external skin dehiscence were distinguished from SSI if no clinical evidence of infection was present. Radiation exposure was defined as any prior radiation therapy taking place before the hardware complication event. Active smoking history was defined as smoking up to four weeks prior to the surgical date (hardware insertion date). Oromandibular defects were classified using the Shaw Classification [15]. Where available, post-operative computed tomography scans were used to confirm defect sizes and classification. Outcomes The primary outcome for this study was the proportion of hardware complications. These included infection in the hardware site requiring antibiotics and with or without hardware removal, hardware exposure, pain or symptoms as associated with hardware requiring hardware removal, and device failure such as plate fracture [2,16].

Methods Data analysis This study was approved by the research ethics board at the University Health Network (04-0648-CE). Patients Patients were identified in an oral cavity registry of patients treated with osseous free flap reconstruction for oromandibular defects between 1999 and 2014 at the University Health Network in Toronto, Ontario, Canada. Inclusion criteria Patients who had: (1) composite oromandibular resection for pathology primarily originating from the oral cavity or bony mandible, (2) reconstruction with either a SFF or FFF, (3) had a minimum of at least one documented follow up visit after surgery, and (4) 18 years of age or older at the time of surgery.

Statistical analysis was performed using SPSS Statistics (v.24.0 Armonk, NY:IBM Corp.). An alpha level of 0.05 was set for statistical significance. Demographic data was summarized using descriptive statistics and groups were compared for baseline differences using the student t-test, Chi-Square test, and Mann Whitney U test depending on the type of variable and whether or not it was parametrically distributed. Univariable analysis of hardware complications was done with either Chi-Square test or with binomial logistic regression. Multivariable analysis was done with a multivariable binomial logistic regression. Two independent variables were selected for the multivariable binomial logistic regression based on the statistical rule of thumb recommending 10 events per variable included in the model. Subgroup analysis was performed based on patients stratified into defects and compared using Fisher’s exact test. Results

Exclusion criteria Demographics Patients with the following were excluded: (1) oromandibular resection not related to malignancy (e.g. osteoradionecrosis or osteomyelitis, craniofacial syndromes, osteomyelitis, pathological fracture unrelated to presence of a primary pathology, trauma,

A total of 178 patients from the database met inclusion criteria and were included in the analysis. 72.4% (n = 129) of patients received a FFF and 27.5% (n = 49) patients had a SFF. There were

165

G.F.Z. Tsang et al. / Oral Oncology 71 (2017) 163–168

more FFF performed earlier on for oromandibular reconstruction than SFF, though the correlation when comparing flap type or choice by time is relatively weak (R2 = 0.108). All reconstructive surgeons perform both fibula and scapular free flaps. Only one surgeon (RG) was clinically active throughout the study’s entire time frame. Patients with squamous cell carcinoma (77.0%) constituted the majority of the pathology seen followed by ameloblastoma (10.1%), salivary gland malignancies (3.9%), and soft tissue sarcomas (4.5%) (Table 1). Mean follow up time for all patients was

Table 1 Demographics of patients and tumor by free flap group. Variable

Fibula (n = 129)

Scapula (n = 49)

p Value

Age (yr) Sex, n (%) Male Female

57.7

60.1

0.373

87 (48.3) 42 (23.0)

30 (16.9) 19 (11.8)

0.435

Follow-up time (months) Mean Range

41.5 1.6–146.3

34.3 3.2–123.6

0.188

CCI Score Mean Median

3.8 4

4.6 4

0.027*

Active smoking, n (%) Radiation therapy, n (%) Diabetes, n (%) SSI, n (%)

40 89 12 19

23 (45.1) 29 (56.9) 4 (7.8) 5 (10.2)

0.118 0.090 0.812 0.430

Tumor pathology, n (%) Squamous cell carcinoma T4 T3 T2 T1 N/A N0 N1 N2a N2b N2c N3 N/A (includes NX) Ameloblastoma Sarcoma Malignant salivary tumor Others

103 (79.8) 78 (75.0) 3 (2.9) 16 (15.4) 6 (5.8) 1 (1.0) 60 (57.7) 12 (11.5) 0 19 (18.3) 9 (8.7) 1 (1) 3 (2.9) 13 (10.1) 5 (3.9) 4 (3.1) 4 (3.1)

34 (69.4) 4 (11.1) 4 (11.1) 6 (16.7) 22 (61.1) 0 19 (52.8) 2 (5.6) 0 12 (33.3) 2 (5.6) 0 1 (2.8) 5 (10.2) 3 (6.1) 3 (6.1) 3 (6.1)

Shaw defect classification, n (%) I II III IV Missing

35 58 30 4 2

19 26 4 0 0

Profile height (%) 1.0 mm 1.5/1.6 mm 2.0 mm 2.4/2.5 mm 2.8 mm Missing

7 (5.4) 100 (77.5) 3 (2.3) 2 (1.6) 4 (3.1) 13 (10.1)

1 (2.0) 39 (79.6) 3 (6.1) 0 (0) 1 (2.0) 5 (10.2)

Hardware complications Exposure Infection Pain/swelling

20 13 6 1

1 1 0 0

Management of complications Free flap Local flap Hardware removal Died before undertaking further treatment

5 1 7 2

0 1 0 0

(31.5) (70.1) (9.4) (14.7)

0.034*

38.8 ± 32.9 months (range = 1.6–147.9 months). Patients with FFF and SFF had a mean follow up time of 41.5 and 34.0 (p = 0.112), respectively. A statistically significant difference in CCI scores was found between FFF and SFF groups, average CCI 3.8 and 4.6, respectively (p = 0.027). A statistically significant difference was also found in T-stage between FFF and SFF groups (p = 0.034). There were a total of 21 (12.0%) hardware complications in the study population of patients. The most common type of complication was found to be hardware exposure (Table 1). The management of hardware complications is also detailed in Table 1. Seven patients underwent isolated removal of hardware, an additional five patients required an additional free flap (four had an RFFF, one with an anterolateral thigh free flap, one required another FFF), and two patients had a local advancement or rotation flap. Risk factors for hardware complications In univariable analysis SSI (OR = 7.05, p < 0.01), free flap type (HR = 0.12, p < 0.02), and defect type (OR = 2.63, p < 0.01) were significant predictors of plate complications (Table 2). No statistical difference was found amongst the groups for age, sex, radiation treatment, CCI, diabetes, smoking history, pathology, use of locking screws, and profile height (p > 0.05). A statistical significant higher number of hardware complications (p = 0.019) was evident in the FFF reconstruction group (16%) compared with the SFF group (2%). Multivariable analysis for hardware complication A multivariable analysis was then completed for free flap reconstruction type and SSI. A statistically significant difference between SFF and FFF was present in SSI (HR = 7.08, CI = 2.47–20.32, p < 0.001) and free flap type (HR = 0.12, CI = 0.015–0.92, p = 0.04) (Table 3). Subgroup analysis by defect type

0.482

0.670

0.794

Because the mandibular defect often determines flap choice, we performed a subgroup analysis stratifying patients by defect type (Table 4). The majority of patients had Class I and II defects. Longer mandible defects were preferentially reconstructed with FFF with 88.2% of Class III and 100% of Class IV (near-total or total mandibulectomy) defects receiving FFF. In class I defects, the absolute risk reduction (ARR) for a hardware complication was 8.6% (8.6% vs. 0%) for SFF (p = 0.287). In the class II defect type, the ARR was 6.3% (10.3% vs. 4%) for SFF (p = 0.324). The class III ARR is 26.7% (26.7% vs. 0%) for SFF (p = 0.261). Although no statistical significance was found in this subgroup analysis, a trend can be

Table 2 Univariable analysis for hardware complications. 0.458

Abbreviations: CCI – Age adjusted Charlson Comorbidity Index, SSI – Surgical Site Infection.

Variable

OR

CI

p-value

Age Flap type (SFF vs. FFF) Sex Radiation exposure CCI Diabetes Surgical site infection Shaw classification defect size Smoking history Plate profile height Pathology T-stage (SCCa only) N-stage (SCCa only) Use of locking screws

1.000 0.116 0.558 1.647 1.107 1.833 7.050 2.628 0.893 1.226 0.668 N/A N/A N/A

0.972–1.029 0.015–0.889 0.194–1.604 0.572–4.741 0.885–1.386 0.476–7.056 2.555–19.454 1.413–4.890 0.340–2.342 0.655–2.295 0.295–1.513 N/A N/A N/A

0.984 0.014* 0.274 0.351 0.373 0.372 <0.001* 0.002* 0.818 0.524 0.333 0.939 0.870 0.664

SFF = Scapular Free Flap, FFF = Fibular Free Flap, SCCa = Squamous Cell Carcinoma. * Statistically significant (p < 0.05).

166

G.F.Z. Tsang et al. / Oral Oncology 71 (2017) 163–168

Table 3 Selected multivariable analysis with binomial logistic regression.

*

Variable

OR

CI

p-value

Reconstruction (FFF vs SFF) SSI

0.115 7.080

0.015–0.915 2.467–20.322

0.041* <0.001*

Statistically significant (p < 0.05).

seen in the proportion of patients with hardware complications regardless of defect type (Fig. 2a and b). Discussion The fibular free flap has long been the workhorse at many centers across the world for osseocutaneous reconstruction. With excellent bone stock for osseointegrated implantation, abundant length for long-segment defects, and long vascular pedicle the FFF is well suited for many defects. At our institution, the scapular tip free flap based on the angular branch of the thoracodorsal artery is commonly utilized because of its long vascular pedicle, abundance of soft tissue, and relative sparing of atherosclerosis and intimal disease seen in fibular flaps [17]. In addition, patients are able to ambulate immediately post-operatively thus reducing their exposure to complications related to being immobile (e.g. pneumonia, deep vein thrombosis, etc.). In the present study, we compare hardware complications between patients undergoing free tissue transfer with the fibular or scapula system of free flaps. We demonstrated a lower rate of hardware complications in patients reconstructed with SFF compared to those reconstructed with FFF. The only other study comparing FFF and SFF in oromandibular reconstruction was published by Dowthwaite et al., who reported a hardware complication rate of 8.6% for FFF and 5.5% for SFF, although no statistical comparisons were performed [22]. The abundance of soft tissue available in the scapular system of flaps likely explains the difference in complication rate. We often harvest the bone segment with a separate soft tissue paddle in a chimeric design. Small mucosal defects may be reconstructed with serratus anterior or latissmus muscle (either based separately on the serratus artery branch of the thoracodorsal artery or using the serratus muscle attached to the scapula tip and folding this intraorally to reconstruct the mucosal defect). Larger mucosal defects or soft tissue defects on the buccal or labial surface of the mandible may be reconstructed with a chimeric latissimus myocutaneous flap or with a thoradorsal artery perforator fasciocutaneous skin paddle. These chimeric designs allow for versatility of reconstruction of the soft tissue envelope relative to the bony defect. Our experience with the fibular skin paddle is that it is

relatively immobile relative to the bone defect. In addition, the small cutaneous skin paddles provide inadequate soft tissue on the labial or buccal side of the mandible thus increasing the risk of plate exposure. One may consider de-epithelializing this skin paddle and folding the fascial component over the reconstruction plate to decrease this risk of exposure, although this is often still inadequate. Hardware contamination with bacterial seeding may be a possible mechanism for subsequent plate exposure. The strong association between surgical site infection and hardware complications in our cohort of patients corroborates this finding. In contrast to the reliable cutaneous/soft tissue paddle supplied through the scapular system, the sometimes tenuous blood supply of the fibular flap skin paddle based on cutaneous perforators may further increased the risk of intraoral wound breakdown and hardware contamination [23]. In the present study we excluded patients undergoing reconstruction of defects involving the external skin. When external skin is included as part of the tumor resection and a fibular flap is used for bony reconstruction, a second soft-tissue flap is often need for external coverage. In cases when a scapular flap is utilized, it is often used as a sole reconstructive option. Because of the complexity of the reconstructions and the use of multiple flaps for these types of reconstructions, we decided to exclude patients with external cutaneous defects. Factors independently associated with an increased risk of hardware complications on multivariable analysis included presence of a SSI and type of mandibular defect. The Shaw classification was chosen for the purposes of this study as it offers a simple but descriptive method of classifying defects which can be easily interpreted based on patient data from imaging or operating room records. The association between mandibular defect size and plate exposure has been previously observed, and our group has previously reported the association between SSI and plate exposure [21,24]. In contrast to other studies, smoking and a history of radiotherapy were not found to be associated with an increased risk of hardware complication [5,18,19,21]. Previous studies defined smoking as any history of smoking whereas we defined a smoking history as those who actively smoking within four weeks prior to surgery. Smoking cessation four weeks prior to surgery has been shown to normalize the pro-inflammatory state caused by smoking [20]. These differences in smoking definitions may account for the differences in the findings. We also failed to show an association between radiation therapy and hardware complication which has been corroborated by at least one other study [21]. Donor site was significantly correlated with hardware complication outcomes on both univariable analysis and multivariable analysis suggesting that SFF has a lower risk of hardware complication. Despite a relatively large cohort, the limited total number of

Table 4 Characteristics of patients stratified by shaw defect classification. Shaw defect classification (n)

Flap, n (%)

Hardware complication, n (%)

No hardware complication, n (%)

I Fibula Scapula

35 (64.8) 19 (35.2)

3 (8.6) 0

32 (91.4) 19 (100)

II Fibula Scapula

58 (69.0) 26 (31.0)

6 (10.3) 1 (3.8)

52 (89.7) 25 (96.2)

III Fibula Scapula

30 (88.2) 4 (11.8)

8 (26.7) 0

22 (73.3) 4 (100)

IV Fibula Scapula

4 (100) 0 (0)

2 (50) –

2 (50) –

NB: defect size missing for 2 patients.

p 0.287

0.324

0.261

–

G.F.Z. Tsang et al. / Oral Oncology 71 (2017) 163–168

167

Fig. 2. Breakdown of patients by defect size within each flap type group, and proportion of each defect class with hardware complications.

complications as a primary outcome limited the number of covariates utilized in the statistical analysis. We performed a post hoc subgroup analysis by defect size to address for this potential confounder. Bony defect size has been previously established as a risk factor for plate related complications [21]. In this analysis, a trend towards fewer hardware complications was found for SFF in all defect classifications, though none

demonstrated statistical significance. Likely this is due to limited sample size of the subgroups. Treatment of hardware complications carry a significant burden both to the individual patient as well as the health care system. These complications often require long-term antibiotic therapy and many require further surgery for hardware removal or soft tissue reconstruction following hardware exposure. Of these, a

168

G.F.Z. Tsang et al. / Oral Oncology 71 (2017) 163–168

second free flap or regional rotational flaps were often required. Furthermore, two patients were found to have facial nerve paralysis from hardware migration through the nerve. Based on the results from this study, with an absolute risk reduction of 14% using a SFF, a number needed to treat of seven patients was required to avoid one hardware complication. In the current financially constrained health care climate, the use of SFF for oromandibular reconstruction may result in significant cost savings. All reconstructive surgeons utilized both SFF and FFF in this cohort of patients. Only one reconstructive surgeon was present on staff throughout the entire duration of the study. Despite this, the incidence of hardware complications occurred fairly regularly throughout the cohort per year (data not shown), indicating that there was little surgeon to surgeon variation when considering some surgeons were only working during certain years that the database covers. The retrospective nature of this study is the main limitation to this study. Direct comparisons between different donor sites is fraught with challenges, particularly because different donor sites are often selected based on the defect and based on patient related factors. Furthermore, details regarding surgical techniques, hardware placement and use, and flap inset details are either affected by recall bias or are significantly limited with respect to obtaining more detailed information. We attempted to include all relevant variables in the analysis while restricting our analysis to factors with complete or near-complete data sets. In addition, because of the relatively large sample size, we were able to adjust for several variables within the analysis, although a multi-institutional review is ultimately needed to sufficiently power a more robust multivariable analysis and compare various reconstructive methods. Conclusions This study compares SFF and FFF used in oromandibular reconstruction with hardware complication as the primary outcome. Patients with defects that only involved the oral cavity were more likely to have a hardware complication if FFF was used in their reconstruction compared with a SFF. Ethics approval This study was approved by the University of Toronto’s Research Ethics Board. Submission This material has never been published and is not currently under evaluation in any other peer-reviewed publication. Conflict of interest The authors have no conflicts of interest to declare. Acknowledgements Colleen Simpson, Department of Otolaryngology UHN research coordinator, for her assistance in retrieving and preparing data for collection.

References [1] Mehta RP, Deschler DG. Mandibular reconstruction in 2004: an analysis of different techniques. Curr Opin Otolaryngol Head Neck Surg 2004;12(4). [2] Knott PD, Suh JD, Nabili V, Sercarz JA, Head C, Abemayor E, et al. Evaluation of hardware-related complications in vascularized bone grafts with locking mandibular reconstruction plate fixation. Arch Otolaryngol-Head Neck Surg 2007;133(12):1302. [3] Blackwell KE, Lacombe V. The bridging lateral mandibular reconstruction plate revisited. Arch Otolaryngol-Head Neck Surg 1999;125(9):988. [4] Fanzio PM, Chang K-P, Chen H-H, et al. Plate exposure after anterolateral thigh free-flap reconstruction in head and neck cancer patients with composite mandibular defects. Ann Surg Oncol 2015;22(9):3055–60. [5] van der Rijt EEM, Noorlag R, Koole R, Abbink JH, Rosenberg AJWP. Predictive factors for premature loss of Martin 2.7 mandibular reconstruction plates. Brit J Oral Maxillofacial Surg 53(2):121–5. [6] Day KE, Desmond R, Magnuson JS, Carroll WR, Rosenthal EL. Hardware removal after osseous free flap reconstruction. Otolaryngol Head Neck Surg 2013:0194599813512103. [7] Moubayed SP, L’Heureux-Lebeau B, Christopoulos A, et al. Osteocutaneous free flaps for mandibular reconstruction: systematic review of their frequency of use and a preliminary quality of life comparison. J Laryngol Otol 2014;128 (12):1034–43. [8] Wang KH, Inman JC, Hayden RE. Modern concepts in mandibular reconstruction in oral and oropharyngeal cancer. Curr Opin Otolaryngol Head Neck Surg 2011;19(2). [9] Okay D, Al Shetawi AH, Moubayed SP, Mourad M, Buchbinder D, Urken ML. Worldwide 10-year systematic review of treatment trends in fibula free flap for mandibular reconstruction. J Oral Maxillofac Surg. [10] Dean NR, Wax MK, Virgin FW, Magnuson JS, Carroll WR, Rosenthal EL. Free flap reconstruction of lateral mandibular defects indications and outcomes. Otolaryngol-Head Neck Sur; 2011:0194599811430897. [11] Militsakh On WAMN et al. Comparison of radial forearm with fibula and scapula osteocutaneous free flaps for oromandibular reconstruction. Arch Otolaryngol-Head & Neck Surg 2005;131(7):571. [12] Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity index. J Clin Epidemiol 47(11):1245–1251. [13] Grandis JR, Snyderman CH, Johnson JT, Yu VL, D’Amico F. Postoperative wound infection. A poor prognostic sign for patients with head and neck cancer. Cancer 1992;70(8):2166–70. [14] Horan TC, Gaynes RP, Martone WJ, Jarvis WR, Emori TG. CDC definitions of nosocomial surgical site infections, 1992: a modification of CDC definitions of surgical wound infections. Infect Control Hosp Epidemiol 1992;13(10):606–8. [15] Brown JS, Barry C, Ho M, Shaw R. A new classification for mandibular defects after oncological resection. Lancet Oncol 17(1):e23–e30. [16] Farwell DG, Kezirian EJ, Heydt JL, Yueh B, Futran ND. Efficacy of small reconstruction plates in vascularized bone graft mandibular reconstruction. Head Neck 2006;28(7):573–9. [17] Urken ML, Bridger AG, Zur KB, Genden EM. The scapular osteofasciocutaneous flap: a 12-year experience. Arch Otolaryngol-Head Neck Surg 2001;127 (7):862–9. [18] Furr AM, Schweinfurth JM, May WL. Factors associated with long-term complications after repair of mandibular fractures. Laryngoscope 2006;116 (3):427–30. [19] Kammerer PW, Klein MO, Moergel M, Gemmel M, Draenert GF. Local and systemic risk factors influencing the long-term success of angular stable alloplastic reconstruction plates of the mandible. J Cranio-Maxillofac Surg 2014(- 5):- e271. [20] Sørensen LT. Wound Healing and Infection in Surgery: The pathophysiological impact of smoking, smoking cessation, and nicotine replacement therapy: a systematic review. Annal Surg 2012;255(6). [21] Nicholson RE, Schuller DE, Forrest LA, Mountain RE, Ali T, Young D. Factors involved in long- and short-term mandibular plate exposure. Arch Otolaryngol-Head & Neck Surg 1997;123(2):217. [22] Dowthwaite SA, Theurer J, Belzile M, et al. Comparison of fibular and scapular osseous free flaps for oromandibular reconstruction: a patient-centered approach to flap selection. JAMA Otolaryngol-Head & Neck Surg 2013;139 (3):285–92. [23] Bak M, Jacobson AS, Buchbinder D, Urken ML. Contemporary reconstruction of the mandible. Oral Oncol 46(2):71–76. [24] Yao CM, Ziai H, Tsang G, Copeland A, Brown D, Irish JC, Gilbert RW, Goldstein DP, Gullane PJ, de Almeida JR. Surgical site infections following oral cavity cancer resection and reconstruction is a risk factor for plate exposure. Journal of Otolaryngology-Head & Neck Surgery 2017;46:30.