Preoperative angiography algorithm.
Alam DS, Khariwala SS. Technical Considerations in Patients Requiring a Second Microvascular Free Flap in the Head and Neck. Arch Otolaryngol Head Neck Surg. 2009;135(3):268-273. doi:10.1001/archoto.2008.554
Copyright 2009 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2009
To identify the difficulties associated with performing a second free flap reconstruction in the head and neck as well as the techniques used to achieve successful reconstruction.
A retrospective review of a prospectively maintained patient database.
The study population comprised 33 patients who required a second free flap in the head and neck.
Main Outcome Measures
Several variables were analyzed in this cohort. Patient-related factors included the following: the availability of recipient vessels, the need to access the contralateral side of the neck, history of neck surgery, history of radiation therapy, and type of free flap used. Complications associated with the reconstruction were also reported.
In this cohort of 33 patients, 34 free flap reconstructions were performed. All were successful, and there were no flap-related complications. Owing to the paucity of good recipient vessels in many cases, the contralateral side of the neck was commonly used, but no patients required vein interposition grafts.
Second free flap reconstructions in the head and neck can be safely accomplished. We have found that the availability of recipient vessels is the most important consideration in these patients. The dissection of the contralateral side of the neck is often required, but the use of flaps with long pedicles obviates the need for vein interposition grafts. Compared with the success rate in this study, the relevant literature suggests that flap survival rates are lower when interposition grafts are used. Modifications in flap-harvesting techniques and inset geometry can also be used to facilitate insets in complicated surgical fields that have been reoperated on.
Many patients with head and neck squamous cell carcinoma present with advanced disease. As a result, surgical resection often results in large defects requiring complex reconstructive techniques. Reconstruction using vascularized free flaps is generally the procedure of choice following tumor ablation in the head and neck. Unfortunately, free flaps are generally required in patients with large primary tumors who are also at higher risk for local recurrence. Patients who develop second tumors or tumor recurrence may in some situations require salvage surgical treatment, often requiring a second free flap.
Another population of patients seen in tertiary referral centers who require a second free flap are patients whose primary reconstruction had failed. Our practice at the Cleveland Clinic, Cleveland, Ohio, performs a considerable number of salvage procedures for failed reconstructions that were done at outside institutions. Four percent of the free flap reconstructions performed by the senior author (D.S.A.) fall within this category.
Free tissue reconstruction in these patients is highly complex owing to myriad factors related to the surgical wound bed. Difficulties can occur relating to issues with prior surgical scarring, the availability of recipient vessels, complications due to infection, open wounds, draining fistulae, and poor local tissue vascularity. This study aimed to clearly identify the difficulties associated with performing a second free flap reconstruction in the head and neck as well as the techniques used to achieve successful reconstruction. In doing so, we hope to identify strategies for success and potential obstacles that may be overcome with careful preoperative and intraoperative planning.
The patients undergoing second free flap reconstructions in this series were most easily divided into the following 2 groups, which we believe are 2 distinct populations: (1) patients with recurrent disease requiring a second ablation and reconstruction and (2) patients with an initial failed free flap necessitating another reconstruction. We sought to identify differences between these 2 groups with regard to the technical aspects of performing a second free flap reconstruction as well as postoperative function and survival.
This study was performed as a retrospective review of a prospectively maintained patient database. All patients undergoing microvascular reconstruction by the senior author (D.S.A.) are cataloged in this database for a series of preoperative, intraoperative, and postoperative variables. All of the microvascular reconstructions performed from 2001 through 2007 are included in this patient database. For the purpose of this study, all patients who required a second free flap in the head and neck region were included (Table 1 and Table 2). This includes patients whose initial free flap reconstruction performed at an outside institution had failed as well as patients with new tumors or recurrent tumors who had a successful initial reconstruction. Several variables were analyzed in this cohort, including patient characteristics such as prior therapy (surgery, chemotherapy, and/or radiation treatment), age, sex, and presenting diagnosis. Technical operative data including the availability of recipient vessels, the need to access the contralateral side of the neck, and anastomosis sites are also logged in the database. Finally, postoperative measures of flap survival, acute surgical and medical complications, and patient survival are also monitored and recorded. An analysis of these data and resulting findings are presented herein for this cohort of patients.
The patient demographics and variables analyzed are presented in Table 1 and Table 2. In this cohort of 33 patients, a total of 34 free flap reconstructions were performed. The extensive nature of the ablative defect required using 2 simultaneous flaps at the second surgery in 1 patient. Two subgroups of patients were defined based on the clinical nature of presentation. The first were patients requiring a second free flap for recurrent tumor (n = 19). These patients all had their initial reconstructive surgery at our institution. The second group in the cohort presented with failed free flaps from outside institutions (n = 14).
The preoperative diagnoses of the patients were as follows: 25 patients were initially treated for carcinoma and 6 patients for osteoradionecrosis, 1 patient had an ameloblastoma of the mandible, and 1 patient was treated for a gunshot wound. All of the patients in the cancer recurrence group and 9 of the 14 in the failed flap group had prior full-course radiation therapy. Many of the patients also had concurrent or adjuvant chemotherapy as a part of a combined therapy protocol (24 of 33 patients).
The initial free flaps used were from the forearm (n = 13), fibula (n = 16), rectus abdominis muscle (n = 2), iliac crest (n = 1), and jejunum (n = 1) (Table 1). All 34 second free flaps had successful microvascular transfers and were viable. The second free flaps were from the forearm (n = 11), fibula (n = 20), latissimus dorsi muscle (n = 1) and rectus abdominis muscle (n = 2) (Table 2). The second free flaps used in the patients with cancer recurrence following initial extirpation and free flap reconstruction were from the forearm (n = 8), fibula (n = 10), and latissimus dorsi muscle (n = 1), and the flaps in the patients with failure of their initial reconstruction were from the forearm (n = 2), fibula (n = 10), latissimus dorsi muscle (n = 1), and rectus abdominis muscle (n = 2).
There were no perioperative mortalities. The only major perioperative surgical complication was a hematoma in 1 patient requiring evacuation in the operating room. Major medical complications included 3 cases of postoperative pneumonia, which were likely caused by aspiration, and 2 cases of postoperative new-onset atrial fibrillation, which were managed medically with spontaneous resolution. One patient had a pulmonary embolism documented on spiral computed tomographic examination and was treated with long-term anticoagulation without further sequelae.
Of the 33 patients in this study, 14 with a failed initial free flap from an outside institution presented both immediately and in a delayed fashion after their failure occurred and were treated at our institution. Because all of the patients with failed free flaps were transferred from outside institutions, accurate preoperative records of vessel availability in this subset of patients was often not available. These failures included frank free flap failure as well as initial reconstructions that had not maintained functional viability (ie, a segmental mandibular defect reconstructed using a reconstruction plate and soft-tissue free flap with subsequent plate exposure and wound breakdown). In this group, 12 patients (86%) presented with open wounds or draining fistulae. The most common failed flap in patients who were transferred to our care was the fibula flap. Owing to the preoperative complications due to infection seen in these patients in the ipsilateral side of the neck, we used the vessels from the contralateral side in 11 of the 14 patients (79%).
In our study, 19 patients were treated for recurrent carcinoma. Notably, only 2 of the 19 patients (11%) in this group presented with an open wound or fistula site. All had been previously irradiated. The availability of vessels from the ipsilateral side of the neck in this group of patients was higher, but given that a number of these patients had undergone very aggressive neck dissection procedures on the affected side, many still lacked suitable vessels on the inset side. Of the 19 patients, 10 (53%) required surgical exploration of the contralateral side of the neck for recipient vessels. Because vessels from the ipsilateral side were used more often in this group, the internal jugular vein was more commonly used to drain the flap because the facial vein was frequently not available. This was usually a consequence of the ipsilateral facial vein being removed during the prior neck dissection.
While our data suggest that multiple free flaps can be performed safely and effectively, the long-term prognosis of these patients is poor. Of the 19 patients in this cancer recurrence group, 13 were alive at the time of analysis, with a mean follow-up of 14 months, and the other 6 died from their malignant disease. At the time of analysis, 1 of the 13 living patients had a recurrent tumor and was in palliative care.
Aggressive management of stage 4 malignant neoplasms in the head and neck often requires microvascular free flap reconstruction. Many of these patients are predisposed to tumor recurrence resulting in the need for a second free flap. While such reconstructions have become routine in major medical centers, the data to support the technique have been limited to a handful of small patient series based on retrospective medical record reviews focused on flap survival and outcomes. This study supports prior data suggesting that multiple microsurgical reconstructions are safe and effective.1- 4 The unique aspect of this report compared with its predecessors is the prospective data collection and focus on intraoperative variables. These are logged on a case-by-case basis.
All of the patients in the present study had a successful transfer of a second free flap to the head and neck. This is similar to the survival rates in other comparable cohort series (90%-95%). Complication rates were also similar to those reported in other series and to those seen in primary flap operations in our practice. Major complications were limited to 1 patient with a hematoma requiring a re-exploration. The medical complication rate in the cohort was similar to that reported in the literature (18%) but higher than that seen when compared with the population with primary reconstructions at our institution (10%). This may be because of the limited size of this cohort and therefore cannot be proven statistically significant at this time. If the trend is, however, an accurate finding, it may reflect the general poor health of patients with recurrent disease or those recovering from the strains of a prior failed flap. An analysis of a larger series should help to better answer this question.
From a technical standpoint, the patients in this study are best divided into those requiring a second reconstruction for recurrent carcinoma and those requiring reconstruction following failure of the first flap. Our findings show that these are 2 distinct populations.
The common primary factor in determining operative complexity in our series was the availability of recipient vessels. The 2 subsets of patients in this study did, however, have different recipient vessel profiles. In patients with failed flaps initially treated at outside institutions, we found the operative reports often underestimated the number of vessels ligated during the first procedure. The active infections and open wounds in these patients further reduced the viability of available vessels. This makes preoperative planning for the second free flap difficult. Most of our patients did not have healthy vessels on the ipsilateral side of the neck, and we opted to use vessels from the contralateral side. The transverse cervical vessels were available in most of these cases, but a better vascular size match made the contralateral facial system preferable. The other potential advantage of using the contralateral side, which may correlate with our good success rate in this series, is the ability to move the microvascular anastomosis well away from the infected operative field.
In contrast to the acute flap failures, the patients being treated for recurrent carcinoma often had ipsilateral vessels available for use. The lack of infectious complications in this group also provided for generally healthier vessels. We did, however, note a relative absence of venous drainage sources on the ipsilateral side, presumably due to surgical resection of the external jugular and common facial veins during the prior neck dissection. These patients more commonly required end-to-side anastomosis into the internal jugular vein compared with those patients with a failed first free flap.
The wide range of vessels used as recipients for the second flap led us to pursue preoperative angiography. Any patient with a suspected paucity of recipient vessels underwent this study (ie, patients with multiple neck operations or infected wounds) to better identify potential sites for vascular anastomosis. In our practice, this test also served as a useful planning tool to help estimate required pedicle length and inset geometry prior to flap harvest. Of the 33 patients, 23 underwent preoperative angiography. All of the patients with a prior failed flap underwent a preoperative vascular workup. This was primarily owing to 2 factors. The first was the increased prevalence of open wounds, active inflammation, and scarring in this population. The second was the limited availability and occasional inaccuracies found in the operative records from outside institutions. The preoperative decision-making paradigm is shown in the flow diagram presented in the Figure.
Unfortunately, a paucity of recipient vessels was a frequent finding. Considering that many patients treated for head and neck cancer undergo radiation therapy in combination with surgery, there were patients who were vessel depleted in both sides of the neck at the time of the second free tissue transfer. In such patients, the transverse cervical artery and vein were a useful alternative option. In addition, in these cases, vessels outside the neck may be accessed to supply the free flap. Two useful options for vascular supply in this situation are the transposition of the thoracodorsal artery and vein or the use of the internal mammary artery and vein.5,6 In the case of thoracodorsal transposition, an incision is made in the posterior axillary fold, and thoracodorsal vessels are identified and traced to the subscapular artery. The vessels are then transposed to the neck by passing them above the pectoralis minor and clavicle for free flap anastomosis.5 Exposure and transposition of the internal mammary vessels requires removal of a portion of the medial third rib via a subperichondrial dissection. A small window is then made in the posterior perichondrium to expose the vessels, which are then dissected, ligated, and transposed superiorly.6
Another approach to manage the inherent vascular limitations in the neck that has undergone multiple surgical procedures is to consider the appropriate flap harvest and design. In contrast to primary cases in which pedicle length is not a common factor, it was often a limiting factor in our second free flap procedures. Efforts were made to extend pedicle length when feasible. For example, radial forearm flaps are harvested to the point of the brachial vein just proximal to the antecubital fossa. Fortunately, radial forearm and scapular system flaps can be harvested with pedicle lengths ranging from 10 to 15 cm. Our fibula flaps were based on distal cutaneous perforators to maximally preserve proximal arterial length. Inset geometry can also play a role in extending the pedicle length. In cases in which facial vessels from the contralateral side were used, we oriented our fibula flaps with the proximal pedicle placed anteriorly to allow it to traverse easily to the other side of the neck. Through careful planning, we were able to avoid vein interposition grafts in all of the cases included in this study.
Interposition grafts have been often seen as a useful option in this patient population because of problems with cervical vessel availability. This subset of patients has usually undergone a modified or radical neck dissection resulting in the depletion of donor vessels, such as with the patients in this series. Although reported rates of flap survival using vein interposition grafts range from 75% to 95%, these grafts are an additional potential source of thrombosis in patients with complicated conditions such as those described in this study.7 We believe that our avoidance of vein grafts in this patient series contributed to our flap survival rate.
In summary, based on the results of this study, we believe that the first step in any case requiring a second free tissue transfer in the head and neck should be the identification of recipient vessels. The establishment of recipient vessel location is the foundation for planning the geometry and inset of the remainder of the flap. Given the variable nature of available vessels, early identification is, in our opinion, paramount. This portion of the procedure generally relies on considerable preoperative planning and potentially computed tomographic angiography or standard angiography to help define available vessels. Once this step has been successfully completed, special maneuvers such as those previously described can help ensure the success of the free flap.
A final but equally important consideration is the survival and quality-of-life outcomes associated with a second free tissue transfer. Before undertaking such a procedure, the resulting functional loss should be carefully considered and discussed with the patient. Whereas a first large surgical ablation may result in tolerable long-term morbidity, a second ablation often removes much of a patient's remaining native tissue in a given anatomic subsite. Owing to the widely divergent nature of preoperative function in these patients as well as the significant variability of the ablative resections and flaps used, it was impossible for us to quantify the true additional morbidity of the second flap. It is clear that these patients experience a significant loss in their quality of life. For all of these reasons, we recommend that both patients and practitioners carefully consider the potential complications and survival prior to a second free tissue reconstruction in the head and neck. Unfortunately in our series, the prognosis of patients with a cancer recurrence requiring a second free flap was poor. At a mean follow-up of only 13 months, only 21 patients (63%) remained disease free and alive. Extrapolating to a 5-year survival point, the survival rates would most certainly be worse. While the success rate and ability to perform a second microvascular reconstruction is supported by this series, the consideration of outcomes and patient quality of life must be considered in the preoperative discussion with the patient.
The performance of a second free flap in the head and neck is complicated but can be successfully completed with a combination of planning and flexibility. Several technical points are illustrated in this study. The availability of recipient vessels is the most important consideration in these patients. The use of vessels from the contralateral side of the neck and even extracervical sources may be considered if vessels from the ipsilateral side are unavailable. This in turn mandates attention to long pedicle harvests and flap inset modifications to allow a tension-free microvascular anastomosis and avoidance of vein interposition grafts. The avoidance of this additional series of anastomosis may have contributed to the improved flap survival rate in this series of patients. Finally, the potential complications and benefits associated with a second free tissue transfer in the head and neck must be carefully considered before undertaking such a procedure. This is particularly important in patients with salvage procedures for local recurrence who often have a very poor prognosis.
Correspondence: Daniel S. Alam, MD, Cleveland Clinic, 9500 Euclid Ave, Ste A-71, Cleveland, OH 44195 (email@example.com).
Submitted for Publication: November 4, 2007; final revision received February 22, 2008; accepted February 25, 2008.
Author Contributions: Dr Alam had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Alam and Khariwala. Acquisition of data: Alam. Analysis and interpretation of data: Alam and Khariwala. Drafting of the manuscript: Alam and Khariwala. Critical revision of the manuscript for important intellectual content: Alam and Khariwala. Administrative, technical, and material support: Alam and Khariwala. Study supervision: Alam.
Financial Disclosure: None reported.