BURN SURGERY AND RESEARCH
Wound Healing Complications With Intraoperative Brachytherapy for Head and Neck Cancer A Unique Form of Radiation Injury Erik J. Geiger, BA, Bryce A. Basques, BS, Christopher C. Chang, MD, Andre W. Alcon, BA, and Deepak Narayan, MD Background: Intraoperative brachytherapy (IOBT) to the tumor bed coupled with surgery has been shown to increase survival and to improve locoregional disease control after head and neck tumor extirpation. Flap reconstruction attempts to restore patient anatomy, while also covering the radioactive implants. The purpose of this study was to better characterize the wound healing complications experienced by patients undergoing reconstruction in the setting of IOBT after tumor ablation, as well as to identify risk factors predicting complications and the need for reoperation. Methods: A retrospective chart review of patients receiving IOBT for head and neck cancer at Yale-New Haven Hospital between 2005 and 2013 was conducted. Patient, tumor, treatment, and reconstructive details were recorded. The number and type of f lap complications, as well as instances in which patients had to be taken back to the operating room, were documented. Bivariate and multivariate logistic regressions were performed to identify risk factors associated with the occurrence of 1 or more f lap complications, as well as the need for reoperation. Results: Ninety-three patients aged 31 to 93 years (mean, 64 T 12 years) who underwent IOBT with f lap reconstruction were included in the study. Of these, 94% had a prior history of radiation (external beam or previous IOBT). Overall, 48 (51.6%) patients experienced at least 1 f lap complication, the most common of which was f lap dehiscence (32% of patients). Thirty-two patients (34% of the cohort) had to be taken back to the operating room at least once for f lap debridement or a revision procedure. On multivariate analysis, only the placement of mandibular hardware during f lap reconstruction was significantly associated with the risk of developing any type of f lap complication (odds ratio, 3.7; P = 0.009) or with subsequent return to the operating room (odds ratio, 3.9; P = 0.012). Conclusions: This study, the largest of its kind, demonstrated a very high complication rate for flaps used to cover brachytherapy implants in this patient cohort. However, many of the patient complications could be managed nonoperatively. Avoiding the use of mandibular hardware with IOBT suggests a method of reducing complications with reconstruction. Key Words: intraoperative brachytherapy, complications, reconstruction, hardware (Ann Plast Surg 2014;73: 378Y384)
W
ound healing typically progresses through an ordered sequence of cellular interactions in 3 stagesVan inf lammatory phase, a proliferative phase, and a maturation/remodeling phase. Radiation therapy disrupts each of these phases. It alters the levels of cytokines like transforming growth factor A, vascular endothelial growth factor, tumor necrosis factor >, and interferon F, all critical to the initial
Received February 18, 2014, and accepted for publication, after revision, April 28, 2014. From the Section of Plastic and Reconstructive Surgery, Yale University School of Medicine, New Haven, CT. Conflicts of interest and sources of funding: none declared. Reprints: Deepak Narayan, MD, Section of Plastic and Reconstructive Surgery, Yale University School of Medicine, 330 Cedar St, BB 3rd Floor, New Haven, CT 06520. E-mail: [email protected]. Copyright * 2014 by Lippincott Williams & Wilkins ISSN: 0148-7043/14/7304-0378 DOI: 10.1097/SAP.0000000000000277
378
www.annalsplasticsurgery.com
phases of wound healing. This inf lammatory milieu leads to endothelial cell apoptosis and microvascular thrombosis and obliteration. Radiation further impairs granulation tissue formation by altering the balance between matrix metalloproteinases and their tissue inhibitors. Then, fibroblastsVcells critical to the normal wound healing processVfunction aberrantly and deposit defective collagen in the radiated bed. All of these cellular processes lead to the familiar clinical elements of skin atrophy, ulceration, fistulization, and fibrosis characteristic of radiation injury.1,2 Osteoradionecrosis of the mandible is another feared complication of radiation therapy for head and neck cancer (HNC). Occurring secondary to radiation-induced microvascular damage, osteoradionecrosis can occur in more than 8% of HNC patients treated with radiotherapy and becomes a nidus for infection and chronic wounds.3,4 A mainstay of treatment for HNC is still external beam radiation therapy (EBRT), and multiple studies have confirmed its role in the treatment of primary disease.5Y7 Yet, despite multimodal therapy, there is a substantial risk for locoregional recurrence, which is the predominant pattern of treatment failure and accounts for most of disease mortality because therapeutic options are limited by prior treatments.8 Intraoperative brachytherapy (IOBT) is the surgical implantation of radioactive sources directly into the tumor bed. These implants are capable of supplying a continuous radiation dose within a limited volume, thereby minimizing injury to the overlying skin and surrounding neurovascular structures. Intraoperative brachytherapy to the tumor mass has proven effective for symptom palliation of HNC,9 and IOBT has been shown to increase actuarial survival and to improve locoregional disease control after head and neck tumor extirpation.7,10Y14 Despite its focal application, IOBT still presents considerable risk to neighboring tissues and can result in a unique form of radiation given the context of its use (Table 1). Moreover, the defect left after HNC resection is often complex (Fig. 1). Local, pedicle, or free flap reconstruction attempts to restore the form and function of the disfigured anatomy, while also covering the radioactive implants and displacing the skin and surgical wound away from areas of highest radiation. The variety of brachytherapy and reconstructive options available today allow head and neck oncology teams to offer multiple salvage options to recurrent HNC patients. However, the combination of these treatments potentially increases the risk of complications.15 The purpose of this studyVthe largest of its kindVwas to better characterize the wound healing complications experienced by TABLE 1. Factors Contributing to Unique Forms of Radiation Injury With IOBT for HNC Insult
Result
Prior radiation of wound bed Heavy oral bacterial contamination of implant site Comorbid conditions like smoking and cachexia
Annals of Plastic Surgery
&
Increased fibrosis and ulceration of aerodigestive tract Increased risk of infection and abscess formation Delayed healing or chronic wounds
Volume 73, Number 4, October 2014
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Plastic Surgery
& Volume 73, Number 4, October 2014
FIGURE 1. Complex defect after head and neck tumor extirpation in left mastoid region exhibits calvarium, exposed dura, and 125I seeds sewn in with vicryl suture (blue strands).
patients undergoing reconstruction in the setting of IOBT after head and neck tumor extirpation. Additionally, we sought to further differentiate minor complications from major ones, distinguished by their need to be taken back to the operating room, as well as to identify risk factors predicting complications and the need for reoperation.
MATERIALS AND METHODS A retrospective chart review of 329 patients receiving IOBT at Yale-New Haven Hospital between 2005 and 2013 was conducted
Wound Healing Complications With IOBT for HNC
with Yale University institutional review board approval (HIC #1310012868). Inclusion criteria included patients who had their first f lap reconstruction at the time of IOBT implant with a minimum of 1-month postoperative follow-up. Two hundred thirty-six patients were excluded (194 that did not require f lap reconstruction, 19 for nonsynchronous use of IOBT with their first f lap, and 23 who lacked sufficient follow-up or other information critical to the analysis), leaving 93 patients having 123 flap reconstructions in the study. Patient characteristics including age, sex, diagnosis, tumor stage, and flap type, as well as radiation, surgical, and medical histories were abstracted from the medical record. The number and types of complicationsV including flap dehiscence, infection, fistula, hematoma, and moreVwere recorded. Complications occurring amid cancer recurrence were excluded. Vascular complications were counted if they involved instances of carotid blowout or oropharyngeal bleeding requiring angiographic embolization. Infection and abscess documentation were mutually exclusive. Additionally, the number of times a patient was taken back to the operating room for f lap debridement or a revision procedure was documented. A ‘‘revision’’ was any operative procedure other than a f lap debridement. Patient treatment plans progressed according to Yale University Head and Neck Tumor Board recommendations. Intraoperatively, radioactive 125Iodine (125I) and/or 103Palladium (103Pd) permanent implants were placed in the surgical bed to provide maximum tumoricidal coverage. The mean activity for each 125I seed was 0.496 mCi and for each 103Pd seed was 1.24 mCi. The median total activity across all implants was 37.0 mCi. The half-lives and energies of 125I and 103 Pd are 60.2 days and 0.028 MeV and 17.0 days and 0.021 MeV, respectively. All patients underwent postoperative dosimetry computed tomographic scans of the IOBT site. Local, pedicle, or free f lap reconstruction and implant coverage was based on the discretion of the attending surgeons. Fifty-one pectoralis major f laps, 39 radial
FIGURE 2. Breakdown of f laps used for defect resurfacing and IOBT implant coverage in this study. * 2014 Lippincott Williams & Wilkins
www.annalsplasticsurgery.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
379
Annals of Plastic Surgery
Geiger et al
RESULTS
TABLE 2. Patient and Treatment Details Number
Percent
93
100.0
19 35 22 17 68 60 49 21 23 30 12 19 10
20.4 37.6 23.7 18.3 73.1 64.5 52.7 22.6 24.7 32.3 12.9 20.4 10.8
22 35 26 10
23.7 37.6 28.9 10.8
85 8
91.4 8.6
9 32 26 26 46 31 69 87
9.7 34.4 28.0 28.0 49.5 33.3 74.2 93.6
Overall Age, y G55 55Y64 65Y74 75+ Male sex History of smoking History of hypertension History of diabetes History of heart disease History of dyslipidemia History of vascular disease History of pulmonary disease History of CNS disease Tumor stage Cutaneous or T1 T2 T3 T4 Cancer diagnosis SCC Other* Flap type Local only Pedicle only Free only Mixed† IOBT total dose 9 37‡ Hardware present Prior surgery Prior radiation§
*Other diagnoses: adenoid cystic, parotid, and sarcoma. †Mixed f lap type refers to patients who had 91 f lap reconstruction. ‡Median total dose, 37 HCu. §Eighteen patients (19.4%) had prior IOBT, 69 patients (74.1%) had prior external beam radiation. CNS indicates central nervous system; SCC, squamous cell carcinoma.
forearm free flaps, 18 local f laps, and 4 fibula free flaps comprised the bulk of the reconstructions, with scapula free flaps, anterolateral thigh flaps, trapezius flaps, rhomboid flaps, latissimus flaps, and deep inferior epigastric perforator f laps all contributing to the remaining minority (Fig. 2). All f laps contacted the radioactive sources. Statistical analyses were conducted using STATA version 11.2 (StataCorp, LP, College Station, Tex). All tests were 2-tailed, and the statistical difference was established at a 2-sided > level of 0.05 (P G 0.05). Bivariate and multivariate logistic regressions were performed to identify risk factors for both the occurrence of 1 or more f lap complications, as well as the need for reoperation. The final multivariate model was constructed using a backward stepwise process that initially included all potential predictor variables and sequentially excluded variables with the highest P value until only those with P G 0.20 remained. Variables with 0.05 G P G 0.20 were left in the model to control for potential confounding but are not considered to be statistically significant. Variables with P G 0.05 were considered significant. 380
www.annalsplasticsurgery.com
& Volume 73, Number 4, October 2014
Patient demographics and other predictor variables are included in Table 2. Ninety-three patients (68 men) aged 31 to 93 years (mean, 64 T 12 years), with a median follow-up of 8 months (interquartile range, 4Y22 months) were included in the study. Sixty-five percent of patients had a history of smoking, and 91% were diagnosed with squamous cell carcinoma. Ninety-four percent had a prior history of radiation (EBRT or previous IOBT) to the site of reconstruction, and 74% of patients had undergone prior surgery. In 33% of patients, hardware (screws or plates) was used intraoperatively to reestablish mandibular continuity. Twenty-six patients received more than 1 f lap reconstruction. The number and percentage of patients with various complications is included in Table 3. Overall, 48 (51.6%) patients experienced at least 1 f lap complication, the most common of which was f lap dehiscence (32.3% of all patients), followed by fistula (19.4% of all patients), and f lap infection (16.1% of all patients). Overall, 41 patientsVor 44.1% of the cohortVhad to be taken back to the operating room at least once for a flap debridement or revision procedure. Table 4 includes the regression analyses for patients having any complication by demographics and comorbidities. On bivariate analysis, only the presence of mandibular hardware during soft tissue reconstruction in the setting of IOBT was significantly associated with the occurrence of any complication (OR, 3.4; P = 0.010). Similarly, on multivariate analysis, only having mandibular hardware was independently associated with subsequent f lap complication (OR, 3.7; P = 0.009). Table 5 includes the bivariate and multivariate analyses for factors predicting return to the operating room based on the variables in Table 1. As with flap complications, only the presence of mandibular hardware under the soft tissue reconstruction was significantly associated with subsequent return to the operating room for a f lap debridement or revision procedure on bivariate analysis (OR, 2.5; P = 0.047) and multivariate analysis (OR, 3.9; P = 0.012).
DISCUSSION Around 50,000 cases of HNC are diagnosed annually in the United States, most of which are locally advanced at presentation.8 The management of advanced HNC requires multimodal therapy, but options for locally recurrent disease are complicated by the EBRT used during primary therapy. Radiation is known to have a detrimental TABLE 3. The Number and Percentage of Patients Experiencing Flap Complications Any complication Dehiscence Infection Abscess Fistula Osteo Debridement Revision Hematoma Vascular Chronic wound Return to OR
No. Patients
Percent of All Patients
48 30 15 8 18 15 16 25 5 11 5 41
51.6 32.3 16.1 8.6 19.4 16.1 17.2 26.9 5.4 11.8 5.4 44.1
Revision pertains to any operative procedure other than f lap debridement; vascular is carotid blowout or bleeding requiring angiographic intervention. Infection and abscess are mutually exclusive. Osteo indicates osteonecrosis/myelitis.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Plastic Surgery
& Volume 73, Number 4, October 2014
Wound Healing Complications With IOBT for HNC
TABLE 4. Percentage of Patients With Any Complication Controlled for Demographics and Comorbidities Bivariate Analyses* Percent With Any Complication Overall Age, y G55 55Y64 65Y74 75+ Male sex History of smoking History of hypertension History of diabetes History of heart disease History of dyslipidemia History of vascular disease History of pulmonary disease History of CNS disease Tumor stage Cutaneous or T1 T2 T3 T4 Cancer diagnosis SCC Other IOBT total dose 9 37 Hardware present Prior surgery
OR
P
Multivariate Analysis*† OR
P
0.4
0.062
0.4
0.077
3.7 2.1
0.009 0.156
51.6 0.822 52.6 54.3 54.6 41.2 47.1 48.3 53.1 61.9 43.5 56.7 50.0 31.6 30.0
Ref. 1.1 1.1 0.63 0.5 0.7 1.1 1.7 0.6 1.3 0.9 0.4 0.4
45.5 54.3 50.0 60.0 0.5 50.6 62.5 56.5 71.0 56.5
Ref. 1.4 1.2 1.8 22 Ref. 1.6 1.5 3.4 2.2
0.151 0.394 0.768 0.286 0.370 0.502 0.905 0.056 0.161 0.864
0.349 0.010 0.112
*Bold values indicate statistical significance (P G 0.05). †The final multivariate model was constructed using a backward stepwise process that initially included all potential predictor variables and sequentially excluded variables with the highest P value until only those with P G 0.20 remained. Variables with 0.05 G P G 0.20 are left in the model to control for potential confounding, but are not considered to be statistically significant. CNS indicates central nervous system; SCC, squamous cell carcinoma.
impact on all phases of wound healing and can severely complicate attempts both of tumor ablation and of subsequent reconstruction. Because single-modality treatment is insufficient for salvage treatment of HNC, IOBT can be used to enhance the ability of today’s oncology teams to treat patients who otherwise carry a dismal prognosis. Although IOBT can improve overall survival and locoregional control of disease, it also complicates attempts at reconstruction of the surgical defect.7,10Y15 The purpose of this study was to familiarize plastic surgeons with the spectrum of wound healing complications that occur in reconstructions with IOBT after head and neck tumor extirpation. Additionally, we sought to differentiate minor complications from major ones, distinguished by their need to be taken back to the operating room, as well as to identify risk factors predicting complications and the need for reoperation. We found that 51.6% of our patient cohort experienced at least 1 complication, the most common of which was f lap dehiscence (32.3% of patients). Forty-one patients (44.1% of the cohort) had to eventually return to the operating room for a debridement or revision procedure. On multivariate analysis, only the presence of mandibular hardware in the setting of f lap reconstruction and IOBT was independently predictive both of developing any type of f lap complication (OR, 3.7; P = 0.009) and also of subsequent operative take-back (OR, 3.9; P = 0.012). The incidence of complications associated with IOBT reported in the literature varies widely up to 68%.16 Reconstruction of the * 2014 Lippincott Williams & Wilkins
brachytherapy sites with local, pedicle, or free f laps has also enjoyed variable results regarding the reduction of complications. After experiencing local soft tissue necrosis in 3 patients treated with Iridium192 (192Ir) afterloading catheters, Stafford and Dearnaley first described the use of myocutaneous f laps for defect resurfacing. They had no further instances of soft tissue complications in the remaining 5 patients of their treatment group.17 Cornes et al compared the complication rates in 13 patients treated with debulking surgery and brachytherapy with 26 patients receiving additional pedicle f lap (mostly deltopectoral and pectoral f lap) coverage. They described a reduction in severe radiation toxicity rates from 46% in the first group to 12% in the reconstructed group.13 Panchal et al18 had no complications associated with brachytherapy and free f lap reconstruction for HNC, but these patients underwent afterloading from 1 to 4 weeks postoperatively, which can permit increased f lap healing. Conversely, Park et al10 had a complication rate of 36% in 35 patients treated with 125I interstitial seeds, and this number increased to 56% when f lap reconstruction was required. Relatedly, Smith et al reviewed a series of 9 patients undergoing postsurgical implantation of 192Ir catheters. Their overall complication rate was 55% (5/9 patients), but all of the complications occurred in the 5 patients who had f lap reconstruction.19 This wide variability in the reported incidences of complications associated with brachytherapy may partly be due to the very small patient cohorts included in previous studies, the diversity of www.annalsplasticsurgery.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
381
Annals of Plastic Surgery
Geiger et al
& Volume 73, Number 4, October 2014
TABLE 5. Percentage of Patients Undergoing Reoperation, Controlled for Demographics and Comorbidities Bivariate Analyses* Percent With Return to OR Overall Age, y G55 55Y64 65Y74 75+ Male sex History of History of History of History of History of
OR
Multivariate Analysis*†
P
OR
P
44.1 0.405
smoking hypertension diabetes heart disease dyslipidemia
History of vascular disease History of pulmonary disease History of CNS disease Tumor stage Cutaneous or T1 T2 T3 T4 Cancer diagnosis SCC Other IOBT total dose 9 37 Hardware present Prior surgery
31.6 40.0 40.9 17.7 32.4 31.7 34.7 38.1 39.1 33.3
Ref. 1.4 1.5 0.5 0.7 0.7 1.0 1.23 1.3 0.9
25.0 26.3 30.0
0.6 0.6 0.8
22.7 40.0 42.3 20.0
Ref. 2.3 2.5 0.85
34.1 37.5 34.8 48.4 37.7
Ref. 1.2 1.03 2.5 1.8
0.199 Ref. 1.4 1.2 0.2
0.492 0.454 0.951 0.686 0.583 0.880 0.466 0.408 0.756 0.342
0.4
0.072
3.2
0.070
0.1 9
0.09
0.121 Ref. 2.9 5.2 1.2
0.847
0.940 0.047 0.264
3.9
0.012
*Bold values indicate statistical significance (P G 0.05). †The final multivariate model was constructed using a backward stepwise process that initially included all potential predictor variables and sequentially excluded variables with the highest P value until only those with P G 0.20 remained. Variables with 0.05 G P G 0.10 are left in the model to control for potential confounding, but are not considered to be statistically significant. CNS indicates central nervous system; SCC, squamous cell carcinoma.
their preceding treatment regimens, and the heterogeneity in brachytherapy applications. Furthermore, the potential impact of multiple patient details on f lap outcomes has been largely overlooked in prior series. The current study attempted to address many of those issues with a much larger sample size against a more uniform treatment background (more than 94% of our cohort had prior radiotherapy and almost 75% had prior surgery) using multivariate analysis to control for potentially confounding variables and to identify independent associations between preoperative variables and postoperative complications. Our study isolated only the presence of mandibular hardware as independently associated of subsequent complication and return to the operating room. Whenever radiation is delivered to tissue containing a foreign body or a foreign body is introduced into previously irradiated tissue, wound healing complications are a concern.20,21 Ryu et al retrospectively analyzed 47 patients with primary or recurrent HNC treated with composite resection, mandibular reconstruction with a titanium plate, and myocutaneous f lap coverage (mostly pectoralis f laps). Eleven patients received no radiation, 10 underwent radiation therapy greater than 10 months from surgery (the ‘‘remote’’ group), and 26 patients received radiation within 12 weeks of surgery (the ‘‘perioperative’’ group). The authors found that rates of both late complications and plate failures were highest in the perioperative group and speculated that the presence of a 382
www.annalsplasticsurgery.com
foreign body in injured tissue with a tenuous vascular supply may lead to a higher chance for infection and delayed wound healing.21 These issues related to mandibular reconstruction suggest the need to pursue treatment techniques that would abrogate the need of surgeons to perform mandibulectomies or mandible swing procedures for intraoral exposure. Performing a marginal mandibulectomy is one method of achieving satisfactory oncologic control while maintaining patients’ functionality and quality of life.22 Studies have also shown that transoral robotic surgery is capable of achieving equal or superior oncologic outcomes compared to open oral surgery in certain patients with either primary or recurrent HNC, while decreasing operative time, complications, and patient morbidity. Given its minimally invasive nature, patients treated via transoral robotic surgery are simply not at risk for many complicationsVincluding mandibular nonunion, free f lap vessel compromise, and need for hardware removalVthat can plague open procedures.23Y25 For the patient to truly benefit from robotic tumor ablation, plastic surgeons will have to reconstruct these defects using the same technology. Multiple studies have demonstrated the feasibility of transoral robotic oropharyngeal reconstruction, sparing the patient lip-split incisions and manibulotomies.26Y29 Certainly, robotic surgery will expand the ablative and reconstructive options available for patients in the future and will potentially obviate the need to expose patients to the risks associated with mandibular hardware. * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Plastic Surgery
& Volume 73, Number 4, October 2014
Interestingly, we found no association between patient variables and the outcomes of reconstruction. For example, no patient comorbidities (including smoking history, diabetes, vascular disease, and more) affected rates of flap complication. In support of our findings, Yoshimura et al30 also found that the incidence of complications due to brachytherapy was unrelated to patient comorbidity status. The high complication rates in this study may be due to prior radiotherapy. Yet, regarding the inf luence of prior radiotherapy on reconstruction, the literature is mixed. In a multivariate analysis of 81 free f lap reconstructions for HNC, Lee and Thiele31 revealed that only preoperative radiation was a significant predictor of free f lap complications. Other authors have also suggested prior radiation increases the risk of f lap complications.32,33 Conversely, Bengtson et al34 showed that prior radiotherapy did not inf luence f lap outcomes in their cohort of 354 patients undergoing free tissue transfer for HNC, and Aitasalo et al35 and Kadota et al36 also published success rates for head and neck free f laps in previously irradiated and operated necks that replicated the outcomes of f lap transfer in the native neck. There are weaknesses inherent to our study given its retrospective nature. As such, we were unable to comment on some details of the reconstructions. For example, we could not comment on whether flap compromise was specifically secondary to problems with the arterial or venous systems. Also, for f laps requiring operative revision, we did not comment on whether there was complete or partial f lap loss necessitating this intervention. Although an interesting and relevant detail, we wanted to describe overall how often patients may need to be taken back to the operating room after undergoing f lap reconstruction with IOBT. Once a patient requires additional operations, the excess cost and inherent risks are borne by our system and by the patient regardless if this was for partial or complete f lap loss. In the same way, we did not specify the specific indications for particular debridement procedures. Often times, the specific factor necessitating operative debridement was not specified in our data set; instead, patients usually presented with 1 or more combinations of issues like dehiscence, infection, and fistula, among others. Additionally, although this study is the largest of its size to date, a larger sample size and more statistical power may have enabled the identification of more associations between patient characteristics and flap outcomes. Finally, we did not divide mandibular hardware into its various components from lag screws to larger plates and can only comment on the presence of hardware in general as predictive of complications and reoperation. There may very well be a different degree of risk associated with the different methods of mandibular fixation.
Wound Healing Complications With IOBT for HNC
REFERENCES 1. Dormand EL, Banwell PE, Goodacre TE. Radiotherapy and wound healing. Int Wound J. 2005;2:112Y127. 2. Haubner F, Ohmann E, Pohl F, et al. Wound healing after radiation therapy: review of the literature. Radiat Oncol. 2012;7:162. 3. Cheriex KC, Nijhuis TH, Mureau MA. Osteoradionecrosis of the jaws: a review of conservative and surgical treatment options. J Reconstr Microsurg. 2013;29:69Y75. 4. Reuther T, Schuster T, Mende U, et al. Osteoradionecrosis of the jaws as a side effect of radiotherapy of head and neck tumour patientsVa report of a thirty year retrospective review. Int J Oral Maxillofac Surg. 2003;32:289Y295. 5. Al-Sarraf M, LeBlanc M, Giri PG, et al. Chemoradiotherapy versus radiotherapy in patients with advanced nasopharyngeal cancer: phase III randomized Intergroup study 0099. J Clin Oncol. 1998;16:1310Y1317. 6. Forastiere AA, Goepfert H, Maor M, et al. Concurrent chemotherapy and radiotherapy for organ preservation in advanced laryngeal cancer. N Engl J Med. 2003;349:2091Y2098. 7. Kupferman ME, Morrison WH, Santillan AA, et al. The role of interstitial brachytherapy with salvage surgery for the management of recurrent head and neck cancers. Cancer. 2007;109:2052Y2057. 8. Cabrera AR, Yoo DS, Brizel DM. Contemporary radiotherapy in head and neck cancer: balancing chance for cure with risk for complication. Surg Oncol Clin N Am. 2013;22:579Y598. 9. Son YH, Sasaki CT. Nonsurgical alternative therapy for bulky advanced head and neck tumors. Arch Otolaryngol Head Neck Surg. 1995;121:991Y993. 10. Park RI, Liberman FZ, Lee DJ, et al. Iodine-125 seed implantation as an adjunct to surgery in advanced recurrent squamous cell cancer of the head and neck. Laryngoscope. 1991;101:405Y410. 11. Martinez A, Goffinet DR, Fee W, et al. 125Iodine implants as an adjuvant to surgery and external beam radiotherapy in the management of locally advanced head and neck cancer. Cancer. 1983;51:973Y979. 12. Lee DJ, Liberman FZ, Park RI, et al. Intraoperative I-125 seed implantation for extensive recurrent head and neck carcinomas. Radiology. 1991;178:879Y882. 13. Cornes PG, Cox HJ, Rhys-Evans PR, et al. Salvage treatment for inoperable neck nodes in head and neck cancer using combined iridium-192 brachytherapy and surgical reconstruction. Br J Surg. 1996;83:1620Y1622. 14. Fee WE Jr, Goffinet DR, Paryani S, et al. Intraoperative iodine 125 implants. Their use in large tumors in the neck attached to the carotid artery. Arch Otolaryngol. 1983;109:727Y730. 15. Ross DA, Hundal JS, Son YH, et al. Microsurgical free flap reconstruction outcomes in head and neck cancer patients after surgical extirpation and intraoperative brachytherapy. Laryngoscope. 2004;114:1170Y1176. 16. Moscoso JF, Urken ML, Dalton J, et al. Simultaneous interstitial radiotherapy with regional or free-flap reconstruction, following salvage surgery of recurrent head and neck carcinoma. Analysis of complications. Arch Otolaryngol Head Neck Surg. 1994;120:965Y972. 17. Stafford N, Dearnaley D. Treatment of ‘inoperable’ neck nodes using surgical clearance and postoperative interstitial irradiation. Br J Surg. 1988;75:62Y64. 18. Panchal JI, Agrawal RK, McLean NR, et al. Early postoperative brachytherapy following free flap reconstruction. Br J Plast Surg. 1993;46:511Y515.
CONCLUSIONS This study demonstrated a very high complication rate for f laps used to cover brachytherapy implants in this patient cohort, and both patients and surgeons should be aware of these potential outcomes. Based upon review of our data, it seems that free f laps may pose a higher complication risk compared to pedicle f laps, although this subjective assessment demands further investigation and can be the topic of focused studies in the future. Notably, many of the patient complications could be managed nonoperatively. Thus, we do not believe that staged proceduresVmeaning delayed f lap reconstruction of IOBT implantsVare necessary. We do not feel that guaranteeing patients the risks inherent to 2 procedures is the correct protocol when one operation with conservative therapy may be sufficient. Avoiding the use of mandibular hardware with IOBT suggests a method of reducing complications with reconstruction. Despite these risks, the benefits of brachytherapy with f lap reconstruction should not be overlooked in the treatment of patients with advanced or recurrent HNC. * 2014 Lippincott Williams & Wilkins
19. Smith RV, Krevitt L, Yi SM, et al. Early wound complications in advanced head and neck cancer treated with surgery and Ir 192 brachytherapy. Laryngoscope. 2000;110:8Y12. 20. Ariyan S. Infections following surgery for head and neck cancer. Clin Plast Surg. 1979;6:523Y535. 21. Ryu JK, Stern RL, Robinson MG, et al. Mandibular reconstruction using a titanium plate: the impact of radiation therapy on plate preservation. Int J Radiat Oncol Biol Phys. 1995;32:627Y634. 22. Rao LP, Shukla M, Sharma V, et al. Mandibular conservation in oral cancer. Surg Oncol. 2012;21:109Y118. 23. White H, Ford S, Bush B, et al. Salvage surgery for recurrent cancers of the oropharynx: comparing TORS with standard open surgical approaches. JAMA Otolaryngol Head Neck Surg. 2013;139:773Y778. 24. Lee SY, Park YM, Byeon HK, et al. Comparison of oncologic and functional outcomes after transoral robotic lateral oropharyngectomy versus conventional surgery for T1 to T3 tonsillar cancer. Head Neck. 2013. doi: 10.1002/hed.23424. [Epub ahead of print]. 25. Sinclair CF, McColloch NL, Carroll WR, et al. Patient-perceived and objective functional outcomes following transoral robotic surgery for early oropharyngeal carcinoma. Arch Otolaryngol Head Neck Surg. 2011;137:1112Y1116.
www.annalsplasticsurgery.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
383
Annals of Plastic Surgery
Geiger et al
26. Song HG, Yun IS, Lee WJ, et al. Robot-assisted free flap in head and neck reconstruction. Arch Plast Surg. 2013;40:353Y358. 27. Longfield EA, Holsinger FC, Selber JC. Reconstruction after robotic head and neck surgery: when and why. J Reconstr Microsurg. 2012;28:445Y450. 28. Selber JC. Transoral robotic reconstruction of oropharyngeal defects: a case series. Plast Reconstr Surg. 2010;126:1978Y1987. 29. Selber JC, Robb G, Serletti JM, et al. Transoral robotic free flap reconstruction of oropharyngeal defects: a preclinical investigation. Plast Reconstr Surg. 2010;125:896Y900. 30. Yoshimura R, Shibuya H, Hayashi K, et al. Disease control using low-dose-rate brachytherapy is unaffected by comorbid severity in oral cancer patients. Br J Radiol. 2011;84:930Y938. 31. Lee S, Thiele C. Factors associated with free flap complications after head and neck reconstruction and the molecular basis of fibrotic tissue rearrangement in preirradiated soft tissue. J Oral Maxillofac Surg. 2010;68:2169Y2178.
384
www.annalsplasticsurgery.com
& Volume 73, Number 4, October 2014
32. Cordeiro PG, Hidalgo DA. Soft tissue coverage of mandibular reconstruction plates. Head Neck. 1994;16:112Y115. 33. Gilas T, Sako K, Razack MS, et al. Major head and neck reconstruction using the deltopectoral flap. A 20 year experience. Am J Surg. 1986;152: 430Y434. 34. Bengtson BP, Schusterman MA, Baldwin BJ, et al. Influence of prior radiotherapy on the development of postoperative complications and success of free tissue transfers in head and neck cancer reconstruction. Am J Surg. 1993;166:326Y330. 35. Aitasalo K, Relander M, Virolainen E. Microvascular free tissue transfers after preoperative irradiation in head and neck reconstructions. Acta Otolaryngol Suppl. 1997;529:247Y250. 36. Kadota H, Fukushima J, Yoshida S, et al. Microsurgical free flap transfer in previously irradiated and operated necks: feasibility and safety. Auris Nasus Larynx. 2012;39:496Y501.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Wound Healing Complications With Intraoperative Brachytherapy for Head and Neck Cancer A Unique Form of Radiation Injury Erik J. Geiger, BA, Bryce A. Basques, BS, Christopher C. Chang, MD, Andre W. Alcon, BA, and Deepak Narayan, MD Background: Intraoperative brachytherapy (IOBT) to the tumor bed coupled with surgery has been shown to increase survival and to improve locoregional disease control after head and neck tumor extirpation. Flap reconstruction attempts to restore patient anatomy, while also covering the radioactive implants. The purpose of this study was to better characterize the wound healing complications experienced by patients undergoing reconstruction in the setting of IOBT after tumor ablation, as well as to identify risk factors predicting complications and the need for reoperation. Methods: A retrospective chart review of patients receiving IOBT for head and neck cancer at Yale-New Haven Hospital between 2005 and 2013 was conducted. Patient, tumor, treatment, and reconstructive details were recorded. The number and type of f lap complications, as well as instances in which patients had to be taken back to the operating room, were documented. Bivariate and multivariate logistic regressions were performed to identify risk factors associated with the occurrence of 1 or more f lap complications, as well as the need for reoperation. Results: Ninety-three patients aged 31 to 93 years (mean, 64 T 12 years) who underwent IOBT with f lap reconstruction were included in the study. Of these, 94% had a prior history of radiation (external beam or previous IOBT). Overall, 48 (51.6%) patients experienced at least 1 f lap complication, the most common of which was f lap dehiscence (32% of patients). Thirty-two patients (34% of the cohort) had to be taken back to the operating room at least once for f lap debridement or a revision procedure. On multivariate analysis, only the placement of mandibular hardware during f lap reconstruction was significantly associated with the risk of developing any type of f lap complication (odds ratio, 3.7; P = 0.009) or with subsequent return to the operating room (odds ratio, 3.9; P = 0.012). Conclusions: This study, the largest of its kind, demonstrated a very high complication rate for flaps used to cover brachytherapy implants in this patient cohort. However, many of the patient complications could be managed nonoperatively. Avoiding the use of mandibular hardware with IOBT suggests a method of reducing complications with reconstruction. Key Words: intraoperative brachytherapy, complications, reconstruction, hardware (Ann Plast Surg 2014;73: 378Y384)
W
ound healing typically progresses through an ordered sequence of cellular interactions in 3 stagesVan inf lammatory phase, a proliferative phase, and a maturation/remodeling phase. Radiation therapy disrupts each of these phases. It alters the levels of cytokines like transforming growth factor A, vascular endothelial growth factor, tumor necrosis factor >, and interferon F, all critical to the initial
Received February 18, 2014, and accepted for publication, after revision, April 28, 2014. From the Section of Plastic and Reconstructive Surgery, Yale University School of Medicine, New Haven, CT. Conflicts of interest and sources of funding: none declared. Reprints: Deepak Narayan, MD, Section of Plastic and Reconstructive Surgery, Yale University School of Medicine, 330 Cedar St, BB 3rd Floor, New Haven, CT 06520. E-mail: [email protected]. Copyright * 2014 by Lippincott Williams & Wilkins ISSN: 0148-7043/14/7304-0378 DOI: 10.1097/SAP.0000000000000277
378
www.annalsplasticsurgery.com
phases of wound healing. This inf lammatory milieu leads to endothelial cell apoptosis and microvascular thrombosis and obliteration. Radiation further impairs granulation tissue formation by altering the balance between matrix metalloproteinases and their tissue inhibitors. Then, fibroblastsVcells critical to the normal wound healing processVfunction aberrantly and deposit defective collagen in the radiated bed. All of these cellular processes lead to the familiar clinical elements of skin atrophy, ulceration, fistulization, and fibrosis characteristic of radiation injury.1,2 Osteoradionecrosis of the mandible is another feared complication of radiation therapy for head and neck cancer (HNC). Occurring secondary to radiation-induced microvascular damage, osteoradionecrosis can occur in more than 8% of HNC patients treated with radiotherapy and becomes a nidus for infection and chronic wounds.3,4 A mainstay of treatment for HNC is still external beam radiation therapy (EBRT), and multiple studies have confirmed its role in the treatment of primary disease.5Y7 Yet, despite multimodal therapy, there is a substantial risk for locoregional recurrence, which is the predominant pattern of treatment failure and accounts for most of disease mortality because therapeutic options are limited by prior treatments.8 Intraoperative brachytherapy (IOBT) is the surgical implantation of radioactive sources directly into the tumor bed. These implants are capable of supplying a continuous radiation dose within a limited volume, thereby minimizing injury to the overlying skin and surrounding neurovascular structures. Intraoperative brachytherapy to the tumor mass has proven effective for symptom palliation of HNC,9 and IOBT has been shown to increase actuarial survival and to improve locoregional disease control after head and neck tumor extirpation.7,10Y14 Despite its focal application, IOBT still presents considerable risk to neighboring tissues and can result in a unique form of radiation given the context of its use (Table 1). Moreover, the defect left after HNC resection is often complex (Fig. 1). Local, pedicle, or free flap reconstruction attempts to restore the form and function of the disfigured anatomy, while also covering the radioactive implants and displacing the skin and surgical wound away from areas of highest radiation. The variety of brachytherapy and reconstructive options available today allow head and neck oncology teams to offer multiple salvage options to recurrent HNC patients. However, the combination of these treatments potentially increases the risk of complications.15 The purpose of this studyVthe largest of its kindVwas to better characterize the wound healing complications experienced by TABLE 1. Factors Contributing to Unique Forms of Radiation Injury With IOBT for HNC Insult
Result
Prior radiation of wound bed Heavy oral bacterial contamination of implant site Comorbid conditions like smoking and cachexia
Annals of Plastic Surgery
&
Increased fibrosis and ulceration of aerodigestive tract Increased risk of infection and abscess formation Delayed healing or chronic wounds
Volume 73, Number 4, October 2014
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Plastic Surgery
& Volume 73, Number 4, October 2014
FIGURE 1. Complex defect after head and neck tumor extirpation in left mastoid region exhibits calvarium, exposed dura, and 125I seeds sewn in with vicryl suture (blue strands).
patients undergoing reconstruction in the setting of IOBT after head and neck tumor extirpation. Additionally, we sought to further differentiate minor complications from major ones, distinguished by their need to be taken back to the operating room, as well as to identify risk factors predicting complications and the need for reoperation.
MATERIALS AND METHODS A retrospective chart review of 329 patients receiving IOBT at Yale-New Haven Hospital between 2005 and 2013 was conducted
Wound Healing Complications With IOBT for HNC
with Yale University institutional review board approval (HIC #1310012868). Inclusion criteria included patients who had their first f lap reconstruction at the time of IOBT implant with a minimum of 1-month postoperative follow-up. Two hundred thirty-six patients were excluded (194 that did not require f lap reconstruction, 19 for nonsynchronous use of IOBT with their first f lap, and 23 who lacked sufficient follow-up or other information critical to the analysis), leaving 93 patients having 123 flap reconstructions in the study. Patient characteristics including age, sex, diagnosis, tumor stage, and flap type, as well as radiation, surgical, and medical histories were abstracted from the medical record. The number and types of complicationsV including flap dehiscence, infection, fistula, hematoma, and moreVwere recorded. Complications occurring amid cancer recurrence were excluded. Vascular complications were counted if they involved instances of carotid blowout or oropharyngeal bleeding requiring angiographic embolization. Infection and abscess documentation were mutually exclusive. Additionally, the number of times a patient was taken back to the operating room for f lap debridement or a revision procedure was documented. A ‘‘revision’’ was any operative procedure other than a f lap debridement. Patient treatment plans progressed according to Yale University Head and Neck Tumor Board recommendations. Intraoperatively, radioactive 125Iodine (125I) and/or 103Palladium (103Pd) permanent implants were placed in the surgical bed to provide maximum tumoricidal coverage. The mean activity for each 125I seed was 0.496 mCi and for each 103Pd seed was 1.24 mCi. The median total activity across all implants was 37.0 mCi. The half-lives and energies of 125I and 103 Pd are 60.2 days and 0.028 MeV and 17.0 days and 0.021 MeV, respectively. All patients underwent postoperative dosimetry computed tomographic scans of the IOBT site. Local, pedicle, or free f lap reconstruction and implant coverage was based on the discretion of the attending surgeons. Fifty-one pectoralis major f laps, 39 radial
FIGURE 2. Breakdown of f laps used for defect resurfacing and IOBT implant coverage in this study. * 2014 Lippincott Williams & Wilkins
www.annalsplasticsurgery.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
379
Annals of Plastic Surgery
Geiger et al
RESULTS
TABLE 2. Patient and Treatment Details Number
Percent
93
100.0
19 35 22 17 68 60 49 21 23 30 12 19 10
20.4 37.6 23.7 18.3 73.1 64.5 52.7 22.6 24.7 32.3 12.9 20.4 10.8
22 35 26 10
23.7 37.6 28.9 10.8
85 8
91.4 8.6
9 32 26 26 46 31 69 87
9.7 34.4 28.0 28.0 49.5 33.3 74.2 93.6
Overall Age, y G55 55Y64 65Y74 75+ Male sex History of smoking History of hypertension History of diabetes History of heart disease History of dyslipidemia History of vascular disease History of pulmonary disease History of CNS disease Tumor stage Cutaneous or T1 T2 T3 T4 Cancer diagnosis SCC Other* Flap type Local only Pedicle only Free only Mixed† IOBT total dose 9 37‡ Hardware present Prior surgery Prior radiation§
*Other diagnoses: adenoid cystic, parotid, and sarcoma. †Mixed f lap type refers to patients who had 91 f lap reconstruction. ‡Median total dose, 37 HCu. §Eighteen patients (19.4%) had prior IOBT, 69 patients (74.1%) had prior external beam radiation. CNS indicates central nervous system; SCC, squamous cell carcinoma.
forearm free flaps, 18 local f laps, and 4 fibula free flaps comprised the bulk of the reconstructions, with scapula free flaps, anterolateral thigh flaps, trapezius flaps, rhomboid flaps, latissimus flaps, and deep inferior epigastric perforator f laps all contributing to the remaining minority (Fig. 2). All f laps contacted the radioactive sources. Statistical analyses were conducted using STATA version 11.2 (StataCorp, LP, College Station, Tex). All tests were 2-tailed, and the statistical difference was established at a 2-sided > level of 0.05 (P G 0.05). Bivariate and multivariate logistic regressions were performed to identify risk factors for both the occurrence of 1 or more f lap complications, as well as the need for reoperation. The final multivariate model was constructed using a backward stepwise process that initially included all potential predictor variables and sequentially excluded variables with the highest P value until only those with P G 0.20 remained. Variables with 0.05 G P G 0.20 were left in the model to control for potential confounding but are not considered to be statistically significant. Variables with P G 0.05 were considered significant. 380
www.annalsplasticsurgery.com
& Volume 73, Number 4, October 2014
Patient demographics and other predictor variables are included in Table 2. Ninety-three patients (68 men) aged 31 to 93 years (mean, 64 T 12 years), with a median follow-up of 8 months (interquartile range, 4Y22 months) were included in the study. Sixty-five percent of patients had a history of smoking, and 91% were diagnosed with squamous cell carcinoma. Ninety-four percent had a prior history of radiation (EBRT or previous IOBT) to the site of reconstruction, and 74% of patients had undergone prior surgery. In 33% of patients, hardware (screws or plates) was used intraoperatively to reestablish mandibular continuity. Twenty-six patients received more than 1 f lap reconstruction. The number and percentage of patients with various complications is included in Table 3. Overall, 48 (51.6%) patients experienced at least 1 f lap complication, the most common of which was f lap dehiscence (32.3% of all patients), followed by fistula (19.4% of all patients), and f lap infection (16.1% of all patients). Overall, 41 patientsVor 44.1% of the cohortVhad to be taken back to the operating room at least once for a flap debridement or revision procedure. Table 4 includes the regression analyses for patients having any complication by demographics and comorbidities. On bivariate analysis, only the presence of mandibular hardware during soft tissue reconstruction in the setting of IOBT was significantly associated with the occurrence of any complication (OR, 3.4; P = 0.010). Similarly, on multivariate analysis, only having mandibular hardware was independently associated with subsequent f lap complication (OR, 3.7; P = 0.009). Table 5 includes the bivariate and multivariate analyses for factors predicting return to the operating room based on the variables in Table 1. As with flap complications, only the presence of mandibular hardware under the soft tissue reconstruction was significantly associated with subsequent return to the operating room for a f lap debridement or revision procedure on bivariate analysis (OR, 2.5; P = 0.047) and multivariate analysis (OR, 3.9; P = 0.012).
DISCUSSION Around 50,000 cases of HNC are diagnosed annually in the United States, most of which are locally advanced at presentation.8 The management of advanced HNC requires multimodal therapy, but options for locally recurrent disease are complicated by the EBRT used during primary therapy. Radiation is known to have a detrimental TABLE 3. The Number and Percentage of Patients Experiencing Flap Complications Any complication Dehiscence Infection Abscess Fistula Osteo Debridement Revision Hematoma Vascular Chronic wound Return to OR
No. Patients
Percent of All Patients
48 30 15 8 18 15 16 25 5 11 5 41
51.6 32.3 16.1 8.6 19.4 16.1 17.2 26.9 5.4 11.8 5.4 44.1
Revision pertains to any operative procedure other than f lap debridement; vascular is carotid blowout or bleeding requiring angiographic intervention. Infection and abscess are mutually exclusive. Osteo indicates osteonecrosis/myelitis.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Plastic Surgery
& Volume 73, Number 4, October 2014
Wound Healing Complications With IOBT for HNC
TABLE 4. Percentage of Patients With Any Complication Controlled for Demographics and Comorbidities Bivariate Analyses* Percent With Any Complication Overall Age, y G55 55Y64 65Y74 75+ Male sex History of smoking History of hypertension History of diabetes History of heart disease History of dyslipidemia History of vascular disease History of pulmonary disease History of CNS disease Tumor stage Cutaneous or T1 T2 T3 T4 Cancer diagnosis SCC Other IOBT total dose 9 37 Hardware present Prior surgery
OR
P
Multivariate Analysis*† OR
P
0.4
0.062
0.4
0.077
3.7 2.1
0.009 0.156
51.6 0.822 52.6 54.3 54.6 41.2 47.1 48.3 53.1 61.9 43.5 56.7 50.0 31.6 30.0
Ref. 1.1 1.1 0.63 0.5 0.7 1.1 1.7 0.6 1.3 0.9 0.4 0.4
45.5 54.3 50.0 60.0 0.5 50.6 62.5 56.5 71.0 56.5
Ref. 1.4 1.2 1.8 22 Ref. 1.6 1.5 3.4 2.2
0.151 0.394 0.768 0.286 0.370 0.502 0.905 0.056 0.161 0.864
0.349 0.010 0.112
*Bold values indicate statistical significance (P G 0.05). †The final multivariate model was constructed using a backward stepwise process that initially included all potential predictor variables and sequentially excluded variables with the highest P value until only those with P G 0.20 remained. Variables with 0.05 G P G 0.20 are left in the model to control for potential confounding, but are not considered to be statistically significant. CNS indicates central nervous system; SCC, squamous cell carcinoma.
impact on all phases of wound healing and can severely complicate attempts both of tumor ablation and of subsequent reconstruction. Because single-modality treatment is insufficient for salvage treatment of HNC, IOBT can be used to enhance the ability of today’s oncology teams to treat patients who otherwise carry a dismal prognosis. Although IOBT can improve overall survival and locoregional control of disease, it also complicates attempts at reconstruction of the surgical defect.7,10Y15 The purpose of this study was to familiarize plastic surgeons with the spectrum of wound healing complications that occur in reconstructions with IOBT after head and neck tumor extirpation. Additionally, we sought to differentiate minor complications from major ones, distinguished by their need to be taken back to the operating room, as well as to identify risk factors predicting complications and the need for reoperation. We found that 51.6% of our patient cohort experienced at least 1 complication, the most common of which was f lap dehiscence (32.3% of patients). Forty-one patients (44.1% of the cohort) had to eventually return to the operating room for a debridement or revision procedure. On multivariate analysis, only the presence of mandibular hardware in the setting of f lap reconstruction and IOBT was independently predictive both of developing any type of f lap complication (OR, 3.7; P = 0.009) and also of subsequent operative take-back (OR, 3.9; P = 0.012). The incidence of complications associated with IOBT reported in the literature varies widely up to 68%.16 Reconstruction of the * 2014 Lippincott Williams & Wilkins
brachytherapy sites with local, pedicle, or free f laps has also enjoyed variable results regarding the reduction of complications. After experiencing local soft tissue necrosis in 3 patients treated with Iridium192 (192Ir) afterloading catheters, Stafford and Dearnaley first described the use of myocutaneous f laps for defect resurfacing. They had no further instances of soft tissue complications in the remaining 5 patients of their treatment group.17 Cornes et al compared the complication rates in 13 patients treated with debulking surgery and brachytherapy with 26 patients receiving additional pedicle f lap (mostly deltopectoral and pectoral f lap) coverage. They described a reduction in severe radiation toxicity rates from 46% in the first group to 12% in the reconstructed group.13 Panchal et al18 had no complications associated with brachytherapy and free f lap reconstruction for HNC, but these patients underwent afterloading from 1 to 4 weeks postoperatively, which can permit increased f lap healing. Conversely, Park et al10 had a complication rate of 36% in 35 patients treated with 125I interstitial seeds, and this number increased to 56% when f lap reconstruction was required. Relatedly, Smith et al reviewed a series of 9 patients undergoing postsurgical implantation of 192Ir catheters. Their overall complication rate was 55% (5/9 patients), but all of the complications occurred in the 5 patients who had f lap reconstruction.19 This wide variability in the reported incidences of complications associated with brachytherapy may partly be due to the very small patient cohorts included in previous studies, the diversity of www.annalsplasticsurgery.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
381
Annals of Plastic Surgery
Geiger et al
& Volume 73, Number 4, October 2014
TABLE 5. Percentage of Patients Undergoing Reoperation, Controlled for Demographics and Comorbidities Bivariate Analyses* Percent With Return to OR Overall Age, y G55 55Y64 65Y74 75+ Male sex History of History of History of History of History of
OR
Multivariate Analysis*†
P
OR
P
44.1 0.405
smoking hypertension diabetes heart disease dyslipidemia
History of vascular disease History of pulmonary disease History of CNS disease Tumor stage Cutaneous or T1 T2 T3 T4 Cancer diagnosis SCC Other IOBT total dose 9 37 Hardware present Prior surgery
31.6 40.0 40.9 17.7 32.4 31.7 34.7 38.1 39.1 33.3
Ref. 1.4 1.5 0.5 0.7 0.7 1.0 1.23 1.3 0.9
25.0 26.3 30.0
0.6 0.6 0.8
22.7 40.0 42.3 20.0
Ref. 2.3 2.5 0.85
34.1 37.5 34.8 48.4 37.7
Ref. 1.2 1.03 2.5 1.8
0.199 Ref. 1.4 1.2 0.2
0.492 0.454 0.951 0.686 0.583 0.880 0.466 0.408 0.756 0.342
0.4
0.072
3.2
0.070
0.1 9
0.09
0.121 Ref. 2.9 5.2 1.2
0.847
0.940 0.047 0.264
3.9
0.012
*Bold values indicate statistical significance (P G 0.05). †The final multivariate model was constructed using a backward stepwise process that initially included all potential predictor variables and sequentially excluded variables with the highest P value until only those with P G 0.20 remained. Variables with 0.05 G P G 0.10 are left in the model to control for potential confounding, but are not considered to be statistically significant. CNS indicates central nervous system; SCC, squamous cell carcinoma.
their preceding treatment regimens, and the heterogeneity in brachytherapy applications. Furthermore, the potential impact of multiple patient details on f lap outcomes has been largely overlooked in prior series. The current study attempted to address many of those issues with a much larger sample size against a more uniform treatment background (more than 94% of our cohort had prior radiotherapy and almost 75% had prior surgery) using multivariate analysis to control for potentially confounding variables and to identify independent associations between preoperative variables and postoperative complications. Our study isolated only the presence of mandibular hardware as independently associated of subsequent complication and return to the operating room. Whenever radiation is delivered to tissue containing a foreign body or a foreign body is introduced into previously irradiated tissue, wound healing complications are a concern.20,21 Ryu et al retrospectively analyzed 47 patients with primary or recurrent HNC treated with composite resection, mandibular reconstruction with a titanium plate, and myocutaneous f lap coverage (mostly pectoralis f laps). Eleven patients received no radiation, 10 underwent radiation therapy greater than 10 months from surgery (the ‘‘remote’’ group), and 26 patients received radiation within 12 weeks of surgery (the ‘‘perioperative’’ group). The authors found that rates of both late complications and plate failures were highest in the perioperative group and speculated that the presence of a 382
www.annalsplasticsurgery.com
foreign body in injured tissue with a tenuous vascular supply may lead to a higher chance for infection and delayed wound healing.21 These issues related to mandibular reconstruction suggest the need to pursue treatment techniques that would abrogate the need of surgeons to perform mandibulectomies or mandible swing procedures for intraoral exposure. Performing a marginal mandibulectomy is one method of achieving satisfactory oncologic control while maintaining patients’ functionality and quality of life.22 Studies have also shown that transoral robotic surgery is capable of achieving equal or superior oncologic outcomes compared to open oral surgery in certain patients with either primary or recurrent HNC, while decreasing operative time, complications, and patient morbidity. Given its minimally invasive nature, patients treated via transoral robotic surgery are simply not at risk for many complicationsVincluding mandibular nonunion, free f lap vessel compromise, and need for hardware removalVthat can plague open procedures.23Y25 For the patient to truly benefit from robotic tumor ablation, plastic surgeons will have to reconstruct these defects using the same technology. Multiple studies have demonstrated the feasibility of transoral robotic oropharyngeal reconstruction, sparing the patient lip-split incisions and manibulotomies.26Y29 Certainly, robotic surgery will expand the ablative and reconstructive options available for patients in the future and will potentially obviate the need to expose patients to the risks associated with mandibular hardware. * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Plastic Surgery
& Volume 73, Number 4, October 2014
Interestingly, we found no association between patient variables and the outcomes of reconstruction. For example, no patient comorbidities (including smoking history, diabetes, vascular disease, and more) affected rates of flap complication. In support of our findings, Yoshimura et al30 also found that the incidence of complications due to brachytherapy was unrelated to patient comorbidity status. The high complication rates in this study may be due to prior radiotherapy. Yet, regarding the inf luence of prior radiotherapy on reconstruction, the literature is mixed. In a multivariate analysis of 81 free f lap reconstructions for HNC, Lee and Thiele31 revealed that only preoperative radiation was a significant predictor of free f lap complications. Other authors have also suggested prior radiation increases the risk of f lap complications.32,33 Conversely, Bengtson et al34 showed that prior radiotherapy did not inf luence f lap outcomes in their cohort of 354 patients undergoing free tissue transfer for HNC, and Aitasalo et al35 and Kadota et al36 also published success rates for head and neck free f laps in previously irradiated and operated necks that replicated the outcomes of f lap transfer in the native neck. There are weaknesses inherent to our study given its retrospective nature. As such, we were unable to comment on some details of the reconstructions. For example, we could not comment on whether flap compromise was specifically secondary to problems with the arterial or venous systems. Also, for f laps requiring operative revision, we did not comment on whether there was complete or partial f lap loss necessitating this intervention. Although an interesting and relevant detail, we wanted to describe overall how often patients may need to be taken back to the operating room after undergoing f lap reconstruction with IOBT. Once a patient requires additional operations, the excess cost and inherent risks are borne by our system and by the patient regardless if this was for partial or complete f lap loss. In the same way, we did not specify the specific indications for particular debridement procedures. Often times, the specific factor necessitating operative debridement was not specified in our data set; instead, patients usually presented with 1 or more combinations of issues like dehiscence, infection, and fistula, among others. Additionally, although this study is the largest of its size to date, a larger sample size and more statistical power may have enabled the identification of more associations between patient characteristics and flap outcomes. Finally, we did not divide mandibular hardware into its various components from lag screws to larger plates and can only comment on the presence of hardware in general as predictive of complications and reoperation. There may very well be a different degree of risk associated with the different methods of mandibular fixation.
Wound Healing Complications With IOBT for HNC
REFERENCES 1. Dormand EL, Banwell PE, Goodacre TE. Radiotherapy and wound healing. Int Wound J. 2005;2:112Y127. 2. Haubner F, Ohmann E, Pohl F, et al. Wound healing after radiation therapy: review of the literature. Radiat Oncol. 2012;7:162. 3. Cheriex KC, Nijhuis TH, Mureau MA. Osteoradionecrosis of the jaws: a review of conservative and surgical treatment options. J Reconstr Microsurg. 2013;29:69Y75. 4. Reuther T, Schuster T, Mende U, et al. Osteoradionecrosis of the jaws as a side effect of radiotherapy of head and neck tumour patientsVa report of a thirty year retrospective review. Int J Oral Maxillofac Surg. 2003;32:289Y295. 5. Al-Sarraf M, LeBlanc M, Giri PG, et al. Chemoradiotherapy versus radiotherapy in patients with advanced nasopharyngeal cancer: phase III randomized Intergroup study 0099. J Clin Oncol. 1998;16:1310Y1317. 6. Forastiere AA, Goepfert H, Maor M, et al. Concurrent chemotherapy and radiotherapy for organ preservation in advanced laryngeal cancer. N Engl J Med. 2003;349:2091Y2098. 7. Kupferman ME, Morrison WH, Santillan AA, et al. The role of interstitial brachytherapy with salvage surgery for the management of recurrent head and neck cancers. Cancer. 2007;109:2052Y2057. 8. Cabrera AR, Yoo DS, Brizel DM. Contemporary radiotherapy in head and neck cancer: balancing chance for cure with risk for complication. Surg Oncol Clin N Am. 2013;22:579Y598. 9. Son YH, Sasaki CT. Nonsurgical alternative therapy for bulky advanced head and neck tumors. Arch Otolaryngol Head Neck Surg. 1995;121:991Y993. 10. Park RI, Liberman FZ, Lee DJ, et al. Iodine-125 seed implantation as an adjunct to surgery in advanced recurrent squamous cell cancer of the head and neck. Laryngoscope. 1991;101:405Y410. 11. Martinez A, Goffinet DR, Fee W, et al. 125Iodine implants as an adjuvant to surgery and external beam radiotherapy in the management of locally advanced head and neck cancer. Cancer. 1983;51:973Y979. 12. Lee DJ, Liberman FZ, Park RI, et al. Intraoperative I-125 seed implantation for extensive recurrent head and neck carcinomas. Radiology. 1991;178:879Y882. 13. Cornes PG, Cox HJ, Rhys-Evans PR, et al. Salvage treatment for inoperable neck nodes in head and neck cancer using combined iridium-192 brachytherapy and surgical reconstruction. Br J Surg. 1996;83:1620Y1622. 14. Fee WE Jr, Goffinet DR, Paryani S, et al. Intraoperative iodine 125 implants. Their use in large tumors in the neck attached to the carotid artery. Arch Otolaryngol. 1983;109:727Y730. 15. Ross DA, Hundal JS, Son YH, et al. Microsurgical free flap reconstruction outcomes in head and neck cancer patients after surgical extirpation and intraoperative brachytherapy. Laryngoscope. 2004;114:1170Y1176. 16. Moscoso JF, Urken ML, Dalton J, et al. Simultaneous interstitial radiotherapy with regional or free-flap reconstruction, following salvage surgery of recurrent head and neck carcinoma. Analysis of complications. Arch Otolaryngol Head Neck Surg. 1994;120:965Y972. 17. Stafford N, Dearnaley D. Treatment of ‘inoperable’ neck nodes using surgical clearance and postoperative interstitial irradiation. Br J Surg. 1988;75:62Y64. 18. Panchal JI, Agrawal RK, McLean NR, et al. Early postoperative brachytherapy following free flap reconstruction. Br J Plast Surg. 1993;46:511Y515.
CONCLUSIONS This study demonstrated a very high complication rate for f laps used to cover brachytherapy implants in this patient cohort, and both patients and surgeons should be aware of these potential outcomes. Based upon review of our data, it seems that free f laps may pose a higher complication risk compared to pedicle f laps, although this subjective assessment demands further investigation and can be the topic of focused studies in the future. Notably, many of the patient complications could be managed nonoperatively. Thus, we do not believe that staged proceduresVmeaning delayed f lap reconstruction of IOBT implantsVare necessary. We do not feel that guaranteeing patients the risks inherent to 2 procedures is the correct protocol when one operation with conservative therapy may be sufficient. Avoiding the use of mandibular hardware with IOBT suggests a method of reducing complications with reconstruction. Despite these risks, the benefits of brachytherapy with f lap reconstruction should not be overlooked in the treatment of patients with advanced or recurrent HNC. * 2014 Lippincott Williams & Wilkins
19. Smith RV, Krevitt L, Yi SM, et al. Early wound complications in advanced head and neck cancer treated with surgery and Ir 192 brachytherapy. Laryngoscope. 2000;110:8Y12. 20. Ariyan S. Infections following surgery for head and neck cancer. Clin Plast Surg. 1979;6:523Y535. 21. Ryu JK, Stern RL, Robinson MG, et al. Mandibular reconstruction using a titanium plate: the impact of radiation therapy on plate preservation. Int J Radiat Oncol Biol Phys. 1995;32:627Y634. 22. Rao LP, Shukla M, Sharma V, et al. Mandibular conservation in oral cancer. Surg Oncol. 2012;21:109Y118. 23. White H, Ford S, Bush B, et al. Salvage surgery for recurrent cancers of the oropharynx: comparing TORS with standard open surgical approaches. JAMA Otolaryngol Head Neck Surg. 2013;139:773Y778. 24. Lee SY, Park YM, Byeon HK, et al. Comparison of oncologic and functional outcomes after transoral robotic lateral oropharyngectomy versus conventional surgery for T1 to T3 tonsillar cancer. Head Neck. 2013. doi: 10.1002/hed.23424. [Epub ahead of print]. 25. Sinclair CF, McColloch NL, Carroll WR, et al. Patient-perceived and objective functional outcomes following transoral robotic surgery for early oropharyngeal carcinoma. Arch Otolaryngol Head Neck Surg. 2011;137:1112Y1116.
www.annalsplasticsurgery.com
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
383
Annals of Plastic Surgery
Geiger et al
26. Song HG, Yun IS, Lee WJ, et al. Robot-assisted free flap in head and neck reconstruction. Arch Plast Surg. 2013;40:353Y358. 27. Longfield EA, Holsinger FC, Selber JC. Reconstruction after robotic head and neck surgery: when and why. J Reconstr Microsurg. 2012;28:445Y450. 28. Selber JC. Transoral robotic reconstruction of oropharyngeal defects: a case series. Plast Reconstr Surg. 2010;126:1978Y1987. 29. Selber JC, Robb G, Serletti JM, et al. Transoral robotic free flap reconstruction of oropharyngeal defects: a preclinical investigation. Plast Reconstr Surg. 2010;125:896Y900. 30. Yoshimura R, Shibuya H, Hayashi K, et al. Disease control using low-dose-rate brachytherapy is unaffected by comorbid severity in oral cancer patients. Br J Radiol. 2011;84:930Y938. 31. Lee S, Thiele C. Factors associated with free flap complications after head and neck reconstruction and the molecular basis of fibrotic tissue rearrangement in preirradiated soft tissue. J Oral Maxillofac Surg. 2010;68:2169Y2178.
384
www.annalsplasticsurgery.com
& Volume 73, Number 4, October 2014
32. Cordeiro PG, Hidalgo DA. Soft tissue coverage of mandibular reconstruction plates. Head Neck. 1994;16:112Y115. 33. Gilas T, Sako K, Razack MS, et al. Major head and neck reconstruction using the deltopectoral flap. A 20 year experience. Am J Surg. 1986;152: 430Y434. 34. Bengtson BP, Schusterman MA, Baldwin BJ, et al. Influence of prior radiotherapy on the development of postoperative complications and success of free tissue transfers in head and neck cancer reconstruction. Am J Surg. 1993;166:326Y330. 35. Aitasalo K, Relander M, Virolainen E. Microvascular free tissue transfers after preoperative irradiation in head and neck reconstructions. Acta Otolaryngol Suppl. 1997;529:247Y250. 36. Kadota H, Fukushima J, Yoshida S, et al. Microsurgical free flap transfer in previously irradiated and operated necks: feasibility and safety. Auris Nasus Larynx. 2012;39:496Y501.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
