Microsurgical head and neck reconstruction: retrospective analysis of 30 free flaps

Introduction: Reconstructive head and neck surgery is a challenge for the plastic surgeon and requires a large technical arsenal, in which microsurgery plays an essential role. The objective to retrospectively analyze microsurgical head and neck reconstruction performed at the Pontifical Catholic University of Campinas over a period of 2 years. Methods: The medical charts of patients who underwent microsurgical reconstruction after head and neck surgery between June 2014 and October 2016 were reviewed to determine which flap was used, the length of the vascular pedicle, recipient vessels, microvascular anastomoses, surgical time, length of hospital stay, complications, and success rate. Results: Thirty microsurgical reconstructions were performed using 3 types of flaps: anterolateral thigh (n = 15), antebrachial radial (n = 8), and fibula (n = 7). Recipient vessels: facial artery (70%) and facial vein (50%), as 92.4% of microvascular anastomoses were end-to-end. The average operative time was 10.1 hours. The average length of hospital stay was 10.7 days. Two flaps were lost due to arterial thrombosis, resulting in a success rate of 93.3%. Conclusions: The microsurgical reconstructions were effective for the repair of complex defects as well as to partially restore the shape and function of the affected tissues. Complications were observed in less than 50% of the cases, although these presented with significant morbidity. The success rate was similar to that reported by major centers of reconstructive microsurgery. The learning curve is long, although it tends to improve with staff training and the acquisition of experience over time. ■ ABSTRACT


INTRODUCTION
Reconstruction performed on the head and neck is a great challenge to plastic surgeons, in view of the functional and 3-dimensional complexity associated with potential skin deficiency, as well as the inadequacy of mucosa, bone, and facial muscles.Fundamental functions such as breathing and swallowing as well as facial expression and speech may be compromised after the removal of a tumor in these areas.Therefore, evaluating possible surgical defects is an essential step to plan reconstruction 1 .
There are several options for the repair of head and neck defects, including primary closure, skin grafting, and local and pedicled flap 2 .Until the 1980s, pedicled flaps were considered the preferred option to cover surgical defects.However, they present several limitations, such as limited arc of rotation, necrosis of the distal portion, venous congestion due to pedicle compression, large volume of flap for the area to be covered, and significant injury to the donor area 3 .In addition, reconstruction of a bone defect with a pedicle flap is practically impossible 4 .
In the 1990s, advances in microvascular surgery techniques enabled free flap transfers to gain popularity.This was due to the high success rates and the ability to remove large amounts of tissue, intact or not, on a small axis of rotation in the region of microvascular anastomosis 5 .
Despite the numerous advantages of microsurgical reconstruction, the mastery of this technique demands much sacrifice and a long and difficult learning curve, since the consequences of failure are directly proportional to the complexity of the technique 6,7 .

OBJECTIVE
To retrospectively analyze microsurgical head and neck reconstructions performed at the Pontifical Catholic University of Campinas (PUC-Campinas), SP, over a period of 2 years.

METHODS
This retrospective study was conducted according to the ethical principles of the Declaration of Helsinki 8 .
The medical charts of patients who underwent microsurgical reconstruction after head and neck surgery at the Hospital e Maternidade Celso Pierro (HMCP) of PUC-Campinas between June 2014 and October 2016 were reviewed, to determine which flap was used, the length of the vascular pedicle, recipient vessels, microvascular anastomoses, surgical time, duration of hospital stay, complications, and success rates.

Surgical Technique
The author of the study was the surgeon responsible for most of the microsurgical reconstruction procedures.The dissection of the flap, usually performed with 2.5× magnification, was performed along with the removal of the tumor.This was followed by the preparation of recipient vessels, flap transfer and placement, and microvascular anastomoses.In cases of late reconstruction, a cervicotomy was performed along with the dissection and identification of recipient vessels, posterior dissection of the flap, transfer, and microvascular anastomosis.
Recipient vessels were chosen according to the similarity between their diameters in relation to those of the flap.In general, we chose between facial or upper thyroid arteries.However, the vein was usually chosen from among branches of the thyrolinguofacial trunk, external jugular vein, and occasionally the superficial temporal vein.Vascular stumps were prepared for anastomosis using a Zeiss S8 OPMI Vario ® surgical microscope.The position of the vessels was established in order to avoid twists, rotations, excessive traction, or compression caused by the flap itself.
Approximately 30 minutes before the sectioning of the pedicle flap, the patient received 2,000 IU of heparin sodium intravenously.At the onset of ischemia, the flap was washed with approximately 100 ml of warm saline solution with 5,000 IU of heparin sodium through cannulation of the flap artery using a 24-G catheter.
Whenever possible, an end-to-end anastomosis was performed, using 4 separate sutures of 10.0 nylon thread (Ethicon ® W2850).When the anastomosis was completed, we waited for 5 minutes to review possible leaks, and placed moist gauze under the anastomosis.If a leak was detected, we placed new sutures without clamping.
Then, we performed a patency test using 2 microsurgery forceps.If there was any change, we performed a new anastomosis; this was repeated as many times as the surgeon believed it was necessary.If the problem persisted, we changed the recipient vessel.Whenever possible, the mean arterial pressure was maintained above 70 mmHg without using vasopressors, which are considered deleterious for the microcirculation in the flap.
Although there is no consensus regarding prophylaxis for microvascular thrombosis during anastomosis 9 , all patients in this study received 20 mg of enoxaparin sodium (Clexane ® ) subcutaneously for the first 3 days, and 200 mg of oral acetylsalicylic acid (ASA) for 30 days.
The patients were kept in the intensive care unit (ICU) with the doctor on call present, to be able to monitor hemodynamic parameters in the early postoperative period.The flaps were evaluated approximately every 6 hours by means of clinical color parameters, venous return to digital pressure, and bleeding in response to a 25-G needle puncture.

RESULTS
The medical records of 28 patients who underwent microsurgical reconstruction in the Department of Head and Neck Surgery at HMCP of PUC-Campinas between June 2014 and October 2016 were analyzed.Two patients required 2 flaps, one as a consequence of tumor recurrence and the other because of arterial thrombosis of the first flap.This resulted in a total of 30 microsurgical reconstruction cases (Table 1).
Of 30 flaps dissected, 15 originated from the anterolateral thigh (ALT), 8 from the antebrachial radial (ABR) region, and 7 from the fibula.The mean length of the vascular pedicles was 15.3 cm from the ALT, 12.5 cm from the ABR, and 2.8 cm from the fibula.In this last flap, the excess proximal bone was discarded, which resulted in a longer vascular pedicle.Therefore, the final mean length of the vascular pedicle of the fibula flap was 8.4 cm.In all flaps, an arterial anastomosis was performed (n = 30), while in 6 of 8 ABR flaps we performed 2 venous anastomoses, for a total of 36 anastomosed veins.Recipient arteries comprised facial arteries in 80% (n = 21) of the cases, followed by upper thyroid arteries in 23.3% (n = 7), internal thoracic arteries in 3.3% (n = 1) and pharyngeal ascending arteries in 3.3% (n = 1).
The minimum duration of surgery was 5 hours and 45 minutes, while the maximum was hours and 30 minutes, with an average duration of hours and 6 minutes.The patients stayed in the hospital between 4 and 44 days, and the average hospital stay was 10.7 days.Of 28 operated patients, 27 spent an average of 3 days in the ICU.Twelve adverse events were observed: hematoma (n = 3), pneumonia (n = 3), salivary fistula (n = 2), arterial thrombosis (n = 2), partial flap necrosis (n = 1), and dehiscence of the donor area (n = 1).Two patients developed pneumonia and died.

DISCUSSION
Immediate or late sequelae following ablative head and neck surgery are always a challenge to the plastic surgeon, especially with the need to consider       functional rehabilitation.In this context, it is essential to add a free tissue transfer procedure to the therapeutic arsenal.Over the past 50 years, several advances in microsurgical techniques and a variety of potential flaps have been described 7,10 .The areas that most frequently require microsurgical reconstruction are located in the head and neck, as SCC is the most common malignancy in this area 11 .In this study, SCC was diagnosed in 58.6% of the cases.Smoking and alcoholism are the most common comorbidities, as also reported in other studies 12 .Although preoperative   radiotherapy is associated with an increased risk of adverse events 13 , two patients in this study who were previously irradiated did not have any complications.
Studies involving microsurgical reconstruction focus on the greater prevalence of soft tissue and bone defects.Fasciocutaneous and osteocutaneous flaps are most often used to treat these defects 5 .
In 1974, the first case of free-flap transfer with successful microsurgical vascular anastomosis was reported 14 .Since then, several flap types have been reported.These include ALT, ABR, and fibula flaps, due to their versatility and wide applicability 15,16 .
The ALT flap was described in 1984, and is based on a network of lateral femoral circumflex vessels 17 .Its first use in the head and neck dates back to 1993 18 .This is a versatile fasciocutaneous flap with a long pedicle that does not compromise the main vessel.It has a fatty thigh layer that can serve to treat important volumetric defects 16 .However, its cutaneous perforators significantly vary from an anatomical point of view.Since most of these are intramuscular, the dissection requires more work 19 .
The ABR flap was described in 1981 for cervicofacial reconstruction 20 .Its advantages rely on a thinner cutaneous component, flexibility, and a long and constant vascular pedicle based on the radial vessels.Venous drainage occurs through the superficial system, i.e., the cephalic vein, and the deep system, through the radial constituent veins.The latter is the dominant drainage system 21 .
The double drainage system explains why some ABR flaps included in the study were performed with two venous anastomoses.The donor area on the forearm should be assessed with the Allen test 22 .Disadvantages include sacrificing the radial artery, frequent need to graft the skin in order to cover the donor area, and an unsightly scar on the anterior surface of the forearm, besides the risk of complications that might occur in this region 23 .
The use of the fibula as a possible vascularized bone graft was described in 1975 24 .However, it was only in 1989 that Hidalgo reported its use for mandibular reconstruction 25 .Since the proximal and distal segments are preserved, its removal does not compromise gait.The fibula is a straight bone that can assume a mandibular shape for dental implants following osteotomy, once the bone has healed in the recipient bed.This is a chimeric bone flap that might include the gastrocnemius muscle for filling, and skin, to enable monitoring of perfusion in the oral mucosa.The disadvantages of this flap are: a short pedicle, laborious dissection, high morbidity of the donor area, and difficult molding to the skeleton of the lower face in order to maintain good oral occlusion 26 .
In most cases evaluated in this series, the defect involved soft tissues, thus justifying the use of 23 fasciocutaneous flaps.Whenever thicker skin and larger volume were required for filling, we chose the ALT flap.Conversely, when thinner and more flexible skin was required, with no volume, we chose the ABR flap, as in the case of nose, lip, and tongue reconstruction.However, for 7 cases in which the greatest defect involved the jaw, we chose to perform a microsurgical reconstruction with an osteocutaneous fibula flap, which was then fixed using a titanium plate and screws.
The vascular pedicles described in this study had an average length similar to that reported in the literature 21,22,27 .However, it was possible to increase the length of the ALT pedicle by dissecting and ligating other branches of the femoral circumflex system.The fibula, which has a short pedicle, was capable of vascular stretching, through distal branch ligation of the soleus muscle vein, and the unnecessary portion of the proximal fibular flap was discarded.Since Carrel's triangulation technique was first described 28 , adequate microsurgical technique has been essential for maintenance of visible patent vessels, either through a simple or continuous suture.It is known that poor coaptation of small vessels can lead to swirling blood flow, thus resulting in thrombosis 29 .End-to-side anastomosis offers a greater technical challenge; however, it is the option of choice in the absence of available recipient veins other than internal jugulars.Therefore, most cases described in this study underwent end-to-end microvascular anastomosis with simple and separate sutures.
The adequate choice of recipient vessels and atraumatic dissection is fundamental for the success of anastomotic patency and good tissue perfusion.The proximity to the location of the defect and the caliber and diameter compatibility between recipient vessels and the vascular pedicle, as well as the integrity of the intima layer, should be taken into account 30 .Whenever possible, in the cases described herein, these rules were respected.
In the first phase of the learning curve in microsurgical skills acquisition, the focus is on ensuring the vascular viability of the flap, which depends on the surgical technique performed.Long surgical times are common at this early stage of training, and these persist until there is synchrony between all members of the team 31 .In this case series, we observed that the duration of surgery tended to decrease with time and experience.
The support of the ICU for hemodynamic control and flap monitoring is of great value during the early postoperative period, i.e., when the clinical observation of flap perfusion is essential to diagnose possible changes 32 .By combining length of stay in the ICU with that in the ward, a prolonged hospital stay is the common denominator for microsurgical reconstructions.However, an average hospital stay of 10.7 days, as observed herein, is shorter than that reported by Al-Dam et al. 33 .
Prior radiotherapy and prolonged anesthesia, i.e., lasting more than 10 hours, are correlated with an increased risk of complications 34 ; moreover, age over 70 is correlated with increased severity of complications 35 .In this survey, the 2 patients that died were elderly (74 and 85 years old, respectively).During their prolonged hospitalization, they developed pneumonia, kidney failure, and cardiopulmonary dysfunction.However, their flaps were viable.
Partial necrosis of the ALT flap required a subsequent surgical revision, which required its re-utilization.The donor area of a different ALT developed dehiscence, and the wound was closed without dressings.The 3 cases of hematoma were clinically treated and resolved without surgery.
A low output fistula was observed in an ALT flap that was tubularized for the reconstruction of pharyngoesophageal transit.With clinical support, this improved spontaneously.A skin island was lost in an osteocutaneous fibula flap, and was caused by a venous thrombosis due to perforating veins.This finally evolved into a salivary fistula, which then resulted in a hemorrhagic shock.This required emergency surgery

Figures 2
Figures 2 to 12 illustrate some of the microsurgical reconstructions of this case series.

Figure 2 .
Figure 2. Donor area of the antebrachial radial flap.

Figure 7 .
Figure 7. Donor area of the anterolateral thigh flap.

Figure 10 .
Figure 10.Donor area of the fibula flap.Left leg and flap following osteotomy; excess bone discarded, with fixation on a precast titanium mandibular plate.

Figure 11 .
Figure 11.Fibula flap for left jaw reconstruction following removal of an ameloblastoma.