Search

Donor site morbidity is an important consideration for follicular unit excision (FUE). We examined 103 male patients with adult androgenic alopecia. Patients were divided into three groups (Good, Fair, and Poor) based on visual assessment of the donor site. Hair density and hair diameter were measured using digital photography. A total of 72, 21 and 10 patients were classified into the Good, Fair and Poor appearance groups. The average hair density of each group was 127.8 ± 22.6 hair/cm

Background The scar alopecia after cranioplasty (SAC) may decrease the patient's quality of life. We have treated SAC using follicular unit extraction (FUE). The aim of this study was to discuss that efficacy of FUE and how much hair follicular unit (FU) should be transplanted intraoperatively for the treatment of SAC. Methods We treated 10 patients (4 men and 6 women) who had SAC using FUE. Results The average age, alopecia size, and intraoperative hair density on the graft area were 29.8 ± 12.1 years, 29.8 ± 44.5 cm2, and 34.6 ± 11.8 FU/cm2, respectively. One year postoperatively, the average hair survival rate on the graft area was 66.3 ± 6.1%. Hair appearance was rated as good in six, fair in three, and poor in one. Among patients whose 1-year postoperative hair density was ≥ 20 FU/cm2, five of six patients achieved good results. However, among patients whose 1-year postoperative hair density was Conclusions FU excision is useful for the treatment of scar alopecia after craniotomy. Our results suggest that the 1-year postoperative hair density should exceed 20 FU/cm2 to achieve good outcomes.

We performed two emergency microsurgical dura, skull, and scalp complex reconstructions. We used the rectus abdominis flap with free fascia lata and the anterolateral thigh flap with vascularized fascia lata. To achieve good postoperative result, reconstruction should be performed before meningitis or wound infection.

In living donor liver transplantation (LDLT) patients, the reconstruction of insufficiently long hepatic artery (HA) is difficult. However, no report has described the relationship between the length of HA and its reconstructive procedure. Herein, we aimed to identify the risk factors for the requirement of additional reconstructive procedures of direct anastomosis. Sixty-eight HA reconstructions in LDLT were conducted (37, men; 31, women). The causes of LDLT were hepatitis (

We performed nail fold reconstruction after digital mucous cyst (DMC) excision using an island-type lateral finger flap on seven patients (four males and three females). Our procedure is a simple and useful method to repair minor nail fold lesion defects after DMC excision.

BACKGROUND Microsurgical lower extremity reconstruction is challenging because of high incidence of vascular thrombosis compared to microsurgical head and neck reconstruction. The risk of vascular pedicle thrombosis increases, if patients have arterial sclerosis or intimal dissection at the recipient artery. We performed selective and continuous transarterial heparin infusion for postoperative anticoagulant therapy. METHODS Fifteen patients (10 men and 5 women; mean age of 55.1 years; range of 16–86 years) received lower leg reconstruction using free flap. Postoperatively, a catheter was inserted into the femoral artery during surgery. Heparin infusion was performed through the catheter as a postoperative therapy for patients who had a risk factor of vascular pedicle thrombosis. Until two days post-operation, heparin was started between 5,000 and 10,000 IU per day. In postoperative days 3 and 4, half of the initial dose of heparin was administered. In postoperative days 5 and 6, 25% of the initial dose of heparin was administered. RESULTS Recipient arteries were the posterior tibial (n=11), anterior tibial (n=2), lateral circumflex femoral (n=1), and medial sural (n=1) arteries. Thirteen of the 15 cases showed arterial sclerosis or intimal dissection at the recipient artery. There was no case of vascular thrombosis. Hematoma formation at flap recipient was observed in four cases. Their initial heparin dose was than 8.5±1.7 U/kg/h. CONCLUSION Continuous transarterial heparin infusion was an effective anticoagulant therapy for the patients who had received free tissue transfer to a lower extremity. The initial dose of heparin should not exceed 6.5 U/kg/h.

We have treated two patients who had an Achilles tendon and overlying tissue defect using an anterolateral thigh flap with fascia lata. Postoperatively, skeletal suspension of the affected leg and intra-arterial heparin infusion were performed for seven days. Six weeks postoperatively, the patients could walk again.

Summary. Surgical treatment of bronchobiliary fistula (BBF) is difficult. A 47-year-old woman presented with a cough with yellow yielding sputum due to BBF. The patient had the adhesion of the liver, diaphragm, and lung. We performed liver, diaphragm, and lung resections. Patient had a large defect of diaphragm. Diaphragm reconstruction was performed using a pedicled reverse latissimus dorsi muscle flap. No flap necrosis was observed. Seven months after surgery, the patient did not present yellow yielding sputum and the BBF was not observed in the computed tomography. This surgical procedure was useful for treating the diaphragm defect both safely and easily. We believe that the reverse pedicled latissimus dorsi muscle flap is a reliable alternative for large diaphragm reconstruction after severe BBF.

The reconstruction of large scalp and dural defects is difficult. Anterolateral thigh (ALT) flap is now widely used because of its reliable blood supply to the skin paddle. Additionally, ALT can be harvested with a large skin paddle and large, well-vascularized fascia. We have successfully treated eight scalp and dural composite defect cases (five male and three female) using ALT with vascularized fascia. The patients’ mean age was 59.1 ± 20.4 years ranging from 31 to 83 years. The mean dural defect size was 73 ± 21 cm2, ranging from 50 to 120 cm2. There were no postoperative infections, bleeding, cerebrospinal fluid leakage, or meningitis. Further discussion about the usefulness of vascularized fascia may be required and we believe that plastic surgeons, head and neck surgeons, and neurosurgeons should report on the results of dural reconstruction.

We treated nine patients with skin defect produced by digital mucous cyst (DMC) excision on the finger and toe using lateral finger flap (LFF). The postoperative scars were esthetically acceptable and no recurrence of mucous cysts was observed. Our LFF is a simple method to repair minor distal dorsal finger defects.

Background In this report, we described the use of hypoglossal-facial neurorrhaphy with end-to-side coaptation between the jump interpositional nerve graft and the hypoglossal nerve for facial reanimation and analyzed the relationship between the outcome of surgery and duration of preoperative paralysis. Methods We performed hypoglossal-facial neurorrhaphy with the jump interpositional nerve graft on nine men and 10 women with unilateral complete facial paralysis. The patients, with a mean age of 39.7 ± 18.1 years (range, 8–65 years) at the time of surgery, experienced preoperative paralysis ranging from 1 to 150 months (mean, 16.9 ± 34.9 months). The movement of the corners of the mouth was evaluated 12 months after surgery using a unique method based on the House–Brackmann grading scale. Results The mean follow-up was 5.6 ± 1.6 years (range, 3–9 years). The movement of the corners of mouth was classified as excellent in two cases, good in seven cases, fair in two cases, and poor in eight cases. Nine of the 11 cases with preoperative paralysis of 6 months or less had excellent or good results, whereas none of the eight cases with preoperative paralysis of 7 months or longer yielded excellent or good results, showing a significant difference (P = 0.01). Conclusions To achieve successful reanimation of the corners of the mouth, hypoglossal-facial neurorrhaphy with end-to-side coaptation between the jump interpositional nerve graft and the hypoglossal nerve should be performed within 6 months after the onset of facial nerve paralysis. © 2015 Wiley Periodicals, Inc. Microsurgery 36:460–466, 2016.

Highlights • Pancreaticoduodenectomy is a difficult and challenging operation, and a gastric tube is frequently used for reconstruction after subtotal esophagectomy for esophageal cancer. • The right gastro-omental artery is important for supplying blood flow to the gastric tube. • Surgeries for pancreatic head tumors become incredibly difficult in patients who have undergone esophageal reconstruction with a gastric tube. • We describe successful arterial reconstruction using the middle colic artery in a patient who had undergone esophageal reconstruction with a gastric tube and whose right gastro-omental artery had been resected., Introduction There are no reports on vessel reconstruction of right gastro-omental artery deficits due to pancreatic tumor resection. Here, we describe successful arterial reconstruction using the middle colic artery in a patient who had undergone esophageal reconstruction with a gastric tube and whose right gastro-omental artery had been resected. Presentation of case A 70-year-old man underwent subtotal esophagectomy and reconstructive surgery with a retrosternal gastric tube for esophageal cancer. A follow-up computed tomography (CT) scan revealed a tumor on the pancreatic head that was adjacent to the right gastro-omental artery. Pancreaticoduodenectomy (PD) was subsequently performed. The gastro-omental artery was resected along with the tumor, creating a 7-cm deficit. The anastomosis was performed between the right branch of the middle colic artery and the distal end of the right gastro-omental artery. No complications that involved blood flow to the reconstructed esophagus were postoperatively observed. Four months after surgery, the blood flow to the gastric tube was confirmed by a contrast CT scan. Discussion We reconstructed the right gastro-omental artery using the middle colic artery, and not a vein graft, as that would have required vessel anastomosis at two locations. The middle colic artery branches on the posterior surface of the pancreas, which is located close to the right gastro-omental artery. Conclusion The middle colic artery provides sufficient blood supply to the pulled-up gastric tube. PD can be performed even in patients who have undergone esophageal reconstruction.

Summary:. We performed a new procedure for reconstruction of donor site of a deltopectoral (DP) flap. A 58-year-old man presented with a wide subcutaneous abscess, which was caused by acute mandibular osteomyelitis due to dental caries. On admission, the patient received a neck incision for drainage. However, necrosis of the neck skin was observed after drainage. The patient had an 8 × 10 cm skin and soft-tissue defect, which we covered with a DP flap (15 × 7 cm). The DP flap donor site was reconstructed using a 16 × 8 cm pedicled thoracodorsal artery perforator (TDAP) flap. There was no flap necrosis, abscess formation, or scar contracture of the DP region. Debulking of the TDAP flap was not required. The pedicled TDAP flap is useful for the reconstruction of the donor site of DP flap. In this report, we describe our operative procedure.

Supplemental Digital Content is available in the text., Summary: We performed hepatic artery (HA) reconstruction on 24 patients between January 2010 and October 2016. Six of 24 patients used an Ikuta type A-II vascular clamp (A-II group). The mean age was 38.0 years (range, 1–61 years). There was no blood leakage at the anastomosed site in any of the patients. No patients required an additional vascular clamp, and none developed HA thrombosis. Eighteen of 24 patients used a conventional vascular clamp. The mean age was 36.1 years (range, 1–65 years; conventional group). Sixteen of 18 patients required an additional vascular clamp due to blood leakage from the HA. There was no significant difference between the 2 groups in mean age or diameter of the recipient HA. However, there was a significant difference in the proportion of patients who required an additional vascular clamp (n < 0.001). The Ikuta type A-II clamp is an effective vascular clamp for reconstruction of the HA in living donor liver transplantation.

Highlights • A graft hepatic artery we experienced too short to rotate a microclamp in living donor liver transplantation. • We performed two-step hepatic artery reconstruction. • In the first step, we cut the recipient right hepatic artery and used it as an arterial graft. • The graft hepatic artery was coapted to the distal stump of the arterial graft without a microclamp. • In the second step, the proximal stump of the arterial graft was coapted to the recipient right hepatic artery., Introduction We describe successful two-step hepatic artery reconstruction in a patient whose graft site hepatic artery was too short for the use of a microclamp in living donor liver transplantation. Presentation of case A 57-year-old woman was diagnosed as having hepatitis C and liver cirrhosis. Her 26-year-old son was the living liver donor. The living donor underwent right lobectomy. The dissected graft hepatic artery was too short for the use of a microclamp. The recipient right hepatic artery was cut and used as an arterial graft. The graft right hepatic artery was sutured to the right hepatic artery of the arterial graft and the graft posterior branch of the right hepatic artery was sutured to the middle hepatic artery of the arterial graft. After reconstruction of the portal vein and hepatic vein was completed, anastomosis was performed between the graft right hepatic artery and right hepatic artery. The patency of the vessels was checked using color Doppler ultrasonography for 1 week postoperatively. No postoperative complications involving blood flow of the hepatic artery were observed. Discussion In our case, the recipient hepatic artery was cut and used as an arterial graft. Although the number of anastomotic sites of the hepatic artery increased, we could perform hepatic artery reconstruction safely and easily. Conclusion Two-step hepatic artery reconstruction is a useful method in cases where the recipient hepatic artery does not have enough length.

Background To reanimate the mimetic muscles, crossface nerve graft (CFNG) is an effective surgical option. However, muscle atrophy after facial paralysis may influence the surgical result. We analyzed the relationship between surgical result and preoperative paralysis duration. Methods We performed CFNG on 15 patients. The sural nerve was transferred between the affected and nonaffected sides of the zygomatic branch. Eyelid function and eyelid lid were evaluated using the modified House–Brackmann scale. The effects of age, sex, cause of facial paralysis, graft nerve length, and preoperative paralysis duration were evaluated. Results The mean follow up period was 9.3 ± 3.3 (range 4–14) years. Eyelid closure was excellent in four patients, good in six, fair in one, and poor in four. Statistically, no significant difference was observed between those patients with excellent or good outcomes and fair or poor outcomes regarding age (40.9 ± 11.0 years vs. 22.6 ± 20.8; P = .067), sex (male/female = 2/8 vs. 3/2; P = .250), cause (tumor/trauma = 10/0 vs. 3/2; P = .095), and length of nerve graft (14.4 ± 0.8 cm vs. 13.8 ± 1.6 cm; P = .375). The average preoperative paralysis duration in the excellent/good patients was significantly shorter than that in the fair/poor patients (P = .005). All eight cases with preoperative paralysis of less than 6 months showed a marked excellent/good result. Two of the seven patients with preoperative paralysis was 6 months or longer marked fair/poor result. (P = .007). Conclusions To achieve successful results with CFNG, surgery should be performed within 6 months of the onset of paralysis.

Recently, oral antithrombotic treatment has been used to treat cerebral infarction, myocardial infarction, and vascular diseases, and these patients may have vascular damage. However, we know of no reports about the development of postoperative complications in patients who have free tissue transfers with preoperative antithrombotic prophylaxis. We retrospectively reviewed 150 patients who had had immediate reconstruction with free tissue transfer after excision of cancers in the head and neck between April 2001 and March 2006. We found nine patients who had been given preoperative antithrombotic prophylaxis: four of the nine had thrombotic complications (myocardial infarction (n = 1), cerebral infarction (n = 1), and necrosis of the flap caused by thrombus at an anastomotic site (n = 2)), whereas 4 of the remaining 141 patients had thrombotic complications, all of which were flap necrosis from thrombotic occlusion of an anastomotic site. Patients given antithrombotic prophylaxis may easily develop a thrombus, and our findings suggest that management of the coagulation system is necessary.

Lipoprostaglandin E1 (lipo-PGE1) has been found accumulating in injured vascular regions. This study examined the localization of lipo-PGE1 in the anastomotic region. The study was divided into three parts. First, we performed anastomosis of the rat femoral artery and vein (n = 17). Lipo-PGE1 labeled with 1,1'-dioctadecyl-1,3,3',3'-tetramethyl-indocarbocyanine was infused intravenously. Hematoxylin-Eosin staining and fluorescence microscopic findings showed that lipo-PGE1 markedly accumulated at the anastomotic site when compared to the contralateral non anastomotic region. Then, we measured laser Doppler flow (LDF) of a lower leg before and after infusion of lipo-PGE1 (n = 7) and saline (n = 7). Increase of blood flow was maintained 1 hour after the infusion of lipo-PGE1 (144% ± 25.0%) when compared to saline infusion. Finally, we performed immunohistochemical and electron microscopic examinations and found that Lipo-PGE1 was incorporated in vascular smooth muscle cells of the anastomotic region. These findings suggest selective accumulation of the lipo-PGE1 in the vascular anastomosis site and affect on the blood flow of repaired vessels. © 2011 Wiley-Liss, Inc. Microsurgery, 2011.

From October 1999 to July 2005, defects after total glossectomy with laryngectomy (TGL) for cancer of the tongue or middle pharynx were reconstructed with rectus abdominis myocutaneous (RAMC) flaps in nine patients. The skin flaps were designed larger in width than the original defect to create a funnel-shaped oropharynx and prevent stricture. Six patients had uneventful postoperative courses and began to eat perorally 8 to 15 days postoperatively. One patient suffered flap necrosis due to arterial thrombosis and two patients had leakage. Eventually, eight patients could eat soft foods or gruel, except one patient who had ingested food through a gastrostomy preoperatively. When TGL was performed without mandibulectomy, the blood supply for the remnant mucosa of the backside of the mandible is generally not good, for which the reconstruction with the overlapping deepithelialized flap and muscle is useful for prevention of the fistula formation. We considered that the RAMC flap is a good option for reconstruction after TGL.

Pharyngoesophageal reconstruction using free jejunal transfer is a reliable procedure, but the achievement of perfect functional results is still challenging. We present a devised method. Jejunoesophageal anastomosis is performed after 2 longitudinal incisions are made at the side corners of the esophageal stump. This maneuver not only enlarges the size of the esophageal stump but also provides a “Z-plasty-like” effect, which reduces the risk of delayed stricture formation. The pharyngojejunal anastomosis is performed in an end-to-end manner. Irrespective of the highest point of the pharyngeal defect, a longitudinal incision is made at the edge of the jejunal graft corresponding to the midpoint of the back wall of the pharyngeal stump. With this concept, considerable longitudinal tension is placed on the posterior side of the jejunal graft, whereas moderate tension is placed on the anterior side of the graft, which prevents the fistula formation that tends to occur in the anterior suture line. Twenty patients underwent the reconstruction using this operative procedure. In all cases, postoperative deglutition was satisfactory without jejunal redundancy or constriction. We believe that our method can be applied in most cases of pharyngoesophageal defects, providing simple and reasonable reconstruction using free jejunal transfer with stable results.

The Bernard method is a straightforward method for reconstructing lower lip defects after tumor resection. However, this method is difficult to apply when the defect is located on the unilateral side of the lower lip. This report describes the reconstruction of unilateral lower lip defects using a modified Bernard method, which is referred to as the hemi-Bernard method. Three patients (2 male and 1 female; mean defect, 55%) underwent reconstruction using the hemi-Bernard method after lower lip malignant tumor resection. No infection or flap necrosis occurred, and none of the 3 patients had difficulty with oral ingestion. Movement of the orbicularis oris muscle was retained in all patients. The hemi-Bernard method is straightforward and has several advantages, including extension of lower lip length. This method could be useful for reconstructing full-thickness defects located on the unilateral side of the lower lip.

Purpose The aims of the present study were to analyze the effectiveness of current perception threshold (CPT) testing to determine patients' minor paresthesia of the infraorbital region after open reduction and internal fixation (ORIF) for unilateral zygomaticomaxillary bone fracture (UZF) and to clarify which nerve fiber was related to the paresthesia. Materials and Methods We conducted a retrospective cohort study of patients who had undergone ORIF after UZF. We also performed neurosensory testing for healthy volunteers who served as the control group. The predictor variables were the period of measurement of Semmes-Weinstein monofilament (S-W) testing and CPT testing (preoperatively and 1 and 5 years postoperatively), measurement side, and disease status (UZF or control). The outcome variables were paresthesia status of the infraorbital nerve region and the results of S-W and CPT testing in both UZF and control groups. The differences in the S-W and CPT values between the affected and unaffected sides in the UZF group and between the UZF and control groups were analyzed by t test ( P Results The present study included 10 patients (6 males and 4 females), with an average age of 25.0 ± 12.7 years, and 21 controls (10 males and 11 females), with an average age of 24.3 ± 1.7 years. In the control group, the CPT and S-W test results did not show any significant differences between the left and right sides. All 10 patients had paresthesia at 1 and 5 years postoperatively. At 5 years postoperatively, the S-W values in all patients showed normalization. From the results of CPT testing, only the A-β fiber function showed significant improvement at 5 years postoperatively. Conclusion The CPT test was an effective sensory test for determining minor paresthesia that could not be detected using S-W testing. Paresthesia of the infraorbital nerve region was caused by the damaged A-δ and C fibers.

In this report, we described the use of hypoglossal-facial neurorrhaphy with end-to-side coaptation between the jump interpositional nerve graft and the hypoglossal nerve for facial reanimation and analyzed the relationship between the outcome of surgery and duration of preoperative paralysis.We performed hypoglossal-facial neurorrhaphy with the jump interpositional nerve graft on nine men and 10 women with unilateral complete facial paralysis. The patients, with a mean age of 39.7 ± 18.1 years (range, 8-65 years) at the time of surgery, experienced preoperative paralysis ranging from 1 to 150 months (mean, 16.9 ± 34.9 months). The movement of the corners of the mouth was evaluated 12 months after surgery using a unique method based on the House-Brackmann grading scale.The mean follow-up was 5.6 ± 1.6 years (range, 3-9 years). The movement of the corners of mouth was classified as excellent in two cases, good in seven cases, fair in two cases, and poor in eight cases. Nine of the 11 cases with preoperative paralysis of 6 months or less had excellent or good results, whereas none of the eight cases with preoperative paralysis of 7 months or longer yielded excellent or good results, showing a significant difference (P = 0.01).To achieve successful reanimation of the corners of the mouth, hypoglossal-facial neurorrhaphy with end-to-side coaptation between the jump interpositional nerve graft and the hypoglossal nerve should be performed within 6 months after the onset of facial nerve paralysis. © 2015 Wiley Periodicals, Inc. Microsurgery 36:460-466, 2016.

Reconstruction of the hepatic artery (HA) is challenging, because there are technical difficulties. Especially, it is difficult to repair the posterior wall. In 2006, we reported an experimental study of the posterior wall first continuous suturing combined with the interrupted suturing and we also confirmed the safety of this procedure. In this article, we report our clinical experiences using this procedure for the HA reconstruction in living-donor liver transplantation. First, we repaired the posterior wall of the HA with continuous suturing. Then, the anterior wall is repaired with the interrupted suturing using a nylon suture with double needle. Between 2006 and 2009, we performed 13 HA reconstructions using our procedure. In all patients, the HA reconstruction was completed easily and uneventfully without oozing from the posterior wall or postoperative HA thrombosis. Our procedure has the benefits of both continuous and interrupted suturing. We believe that it is useful for reconstruction of the HA in living-donor liver transplantation.

Lipoprostaglandin E1 (lipo-PGE1) has been found accumulating in injured vascular regions. This study examined the localization of lipo-PGE1 in the anastomotic region. The study was divided into three parts. First, we performed anastomosis of the rat femoral artery and vein (n = 17). Lipo-PGE1 labeled with 1,1'-dioctadecyl-1,3,3',3'-tetramethyl-indocarbocyanine was infused intravenously. Hematoxylin-Eosin staining and fluorescence microscopic findings showed that lipo-PGE1 markedly accumulated at the anastomotic site when compared to the contralateral non anastomotic region. Then, we measured laser Doppler flow (LDF) of a lower leg before and after infusion of lipo-PGE1 (n = 7) and saline (n = 7). Increase of blood flow was maintained 1 hour after the infusion of lipo-PGE1 (144% ± 25.0%) when compared to saline infusion. Finally, we performed immunohistochemical and electron microscopic examinations and found that Lipo-PGE1 was incorporated in vascular smooth muscle cells of the anastomotic region. These findings suggest selective accumulation of the lipo-PGE1 in the vascular anastomosis site and affect on the blood flow of repaired vessels.

In secondary reconstruction of the maxilla, skeletal reconstruction as well as soft tissue augmentation is required to obtain a good contour. We present a new strategy for combining a costal cartilage graft with a rib-latissimus dorsi flap. We used this method to treat a 39-year-old man who had previously undergone total maxillectomy. First, a vascularized rib, elevated together with a latissimus dorsi flap, was fixed between the middle of the maxilla and the edge of the zygomatic arch. The small defects that could not be reconstructed with the rib only were reconstructed with a costal cartilage graft. The patient did not develop any postoperative infection or flap necrosis. Thirteen months after the secondary reconstruction, he presented with a good contour of the cheek. Our method was effective for the reconstruction of a complex skeletal defect of the maxilla.

The reported success rates of free jejunal transfer are over 95%, but in cases of postoperative vascular thrombosis, the salvage of jejunal flap is often difficult because of poor ischemic tolerance of the jejunum. To reduce the incidence of jejunal necrosis due to vascular thrombosis to nearly zero, we employed one-segment double vascular pedicled free jejunal transfer. Different from conventional double pedicled free jejunal transfer (transfer of the two jejunal segments by anastomosing two pairs of jejunal root vessels), the arcade vessels are used as an additional feeder after the routine anastomosis of jejunal root vessels in our method. Between December 2004 and January 2006, 20 patients with laryngeal, pharyngeal, or cervical esophageal cancer underwent free jejunal transfer using this method. In all patients, the jejunal flap survived completely without any complication associated with vascular anastomosis or blood circulation of the flap. The disadvantage of this procedure is the approximately 1-hour prolonged operative time. Although we have experienced only 20 cases and not obtained statistically significant validity of this method compared with conventional one, we believe that the concept of our method is one of the help for safer pharyngoesophageal reconstruction, especially in patients with higher risk of vascular thrombosis.