Customize your JAMA Network experience by selecting one or more topics from the list below.
Head C, Sercarz JA, Abemayor E, Calcaterra TC, Rawnsley JD, Blackwell KE. Microvascular Reconstruction After Previous Neck Dissection. Arch Otolaryngol Head Neck Surg. 2002;128(3):328–331. doi:10.1001/archotol.128.3.328
Microvascular reconstruction of defects in the head and neck is more challenging in patients who have undergone a previous neck dissection, owing to prior resection of potential cervical recipient blood vessels used for free flap perfusion.
To evaluate the reliability and safety of free flap reconstruction in patients with previous neck dissection.
Patients and Methods
Sixty free flaps were performed in 59 patients with a medical history of neck dissection for head and neck cancer. This included patients undergoing salvage surgery for recurrent cancer as well as patients undergoing secondary reconstruction of cancer surgery–related defects. Flap selection included 25 radial forearm flaps, 20 fibula flaps, 7 rectus abdominis flaps, 7 subscapular system flaps, and 1 iliac crest flap.
Recipient vessels were used in the field of previous neck dissection in approximately half the patients with previous selective neck dissection, while contralateral recipient vessels were always used in patients with a history of modified radical or radical neck dissection. Vein grafts were not necessary in any cases. One arterial anastomosis that was created under excessive tension required urgent reoperation and revision, but there were no cases of free flap failure.
Free flap reconstruction of the head and neck is highly successful in patients with a history of neck dissection, despite a relative paucity of potential cervical recipient blood vessels. Heavy reliance on free flaps with long vascular pedicles obviated the need to perform vein grafts in the present series, probably contributing to the absence of free flap failure. Previous neck dissection should not be considered a contraindication to microvascular reconstruction of the head and neck.
MICROVASCULAR free flaps have proven to be both reliable and functionally effective for reconstruction of major head and neck defects. The most recent clinical series of head and neck free flap reconstructions following ablative cancer surgery have reported flap survival rates in the 98% to 99% range.1,2
Free flap reconstruction is more challenging in patients who have undergone previous neck dissection ipsilateral to the site of defect reconstruction, as previous neck dissection reduces the availability of potential recipient cervical blood vessels for free flap perfusion. The lack of potentially suitable cervical recipient blood vessels can increase the complexity of achieving successful free flap perfusion and thereby may increase the risk of free flap thrombosis and failure. To our knowledge, no previous series have focused on the impact of previous neck dissection on microvascular head and neck reconstruction. In this series, we document the reliability and safety of free flap transfer following previous neck dissection; recommendations for optimizing free flap survival are also described.
Fifty-nine patients with a medical history of neck dissection for treatment of cancer underwent a total of 60 microvascular free flaps (1 patient received 2 simultaneous free flaps) for reconstruction of defects in the head and neck region. Medical records were reviewed to determine patient age, sex, cancer histologic characteristics, classification of previous neck dissection, indication and timing of free flap reconstruction, defect classification and laterality, and cervical recipient vessel selection. For patients with large recurrent cancers that crossed the midline, defect laterality was classified according to the initial site of origin of the tumor or the location of the epicenter of the defect. Previous neck dissections were classified as radical neck dissections, modified radical neck dissections (sparing the spinal accessory nerve), or selective neck dissections (including supraomohyoid, lateral, posterolateral, or anterior type).3 The internal jugular vein was preserved in all cases of previous selective neck dissection.
There were 41 men and 18 women with a medical history of neck dissection for head and neck cancer who underwent microvascular flap reconstruction. Age at the time of therapy ranged from 32 to 85 years. All defects arose as a result of treatment of head and neck cancer, with the specific pathologies consisting of squamous cell carcinoma (55 cases), epimyoepithelial carcinoma (1 case), adenoid cystic carcinoma (1 case), and metastatic renal cell carcinoma (1 case). Forty-seven reconstructions were carried out in conjunction with resection of recurrent cancers or secondary primary cancers, while 10 cases entailed secondary reconstruction of cancer therapy–related defects among patients in remission. Two additional reconstructions were done in conjunction with segmental mandibulectomy for treatment of advanced osteoradionecrosis of the mandible. The wounds undergoing reconstruction were classified as oral-oropharyngeal (48 cases), pharyngoesophageal (8 cases), or skull base–midface defects (3 cases).
All patients had a medical history of previous neck dissection. In 50 (85%) of 59 cases, a previous neck dissection had been performed ipsilateral to the site of the defect. Of the 9 patients (15%) who had previous neck dissections performed contralateral to the site of the defect, 2 had functional neck dissections and 7 had selective neck dissections. Overall, there were 29 cases of previous unilateral radical or modified radical neck dissection and 24 cases of previous unilateral selective neck dissection. In addition, 3 patients had a history of ipsilateral radical neck dissection combined with contralateral selective neck dissection. Three additional patients had a history of bilateral selective neck dissection, resulting in a total of 33 previous selective neck dissections among the 59 cases in this series. Forty-two cases (71%) had preoperative radiation therapy.
Flap selection included 25 radial forearm flaps, 20 fibula flaps, 7 rectus abdominis flaps, 7 subscapular system flaps, and 1 iliac crest flap. Recipient vessel selection is summarized in Table 1. The number of recipient arteries and veins used for free flap perfusion exceeds the number of reconstructions performed because dual venous drainage was used in select cases of radial forearm flap and fibula flap reconstruction, and there was 1 case of simultaneous transfer of 2 free flaps. Overall, cervical recipient vessels located on the side of the neck that was contralateral to the defect site were used in 36 (61%) of 59 cases of microvascular reconstruction. The contralateral facial artery was the most common recipient artery, while the most common recipient vein was divided relatively evenly between the contralateral internal and external jugular veins (Table 1). Cervical recipient vessels for free flap perfusion were located in the field of previous selective neck dissection much more frequently than in the field of previous radical or modified radical neck dissection. Recipient blood vessels for free flap perfusion were located in the field of 17 (52%) of 33 previous selective neck dissections, compared with no instances where cervical recipient vessels were in the field of previous radical or modified radical neck dissection (0 of 29 cases). It was not necessary to use vein grafts to lengthen the vascular pedicles that supplied the free flaps in any cases.
One case required urgent reexploration 3 days postoperatively for disruption of a microarterial anastomosis that was created under excessive tension. In this case, the patient had previously undergone a right radical neck dissection that was ipsilateral to the defect site. A through-and-through defect of the tongue, mandible, and skin of the chin and neck was reconstructed using simultaneous transfer of a fibula free flap and a radial forearm free flap. Both flaps were perfused using recipient vessels in the contralateral left neck. The microarterial anastomosis created between the peroneal artery of the fibula flap to the left lingual artery ruptured 3 days postoperatively when the patient turned his head position to the extreme right. This anastomosis was urgently revised and the fibula flap was successfully salvaged. There were no cases of free flap failure, resulting in a success rate of 100%. Preoperative radiotherapy had no impact on free flap viability.
Over the past decade, the use of microvascular free flaps has greatly enhanced the armamentarium of methods available to achieve surgical reconstruction of defects in the head and neck. Microvascular free flaps allow single-stage reconstruction at the time of surgical resection. Although the first microscope-assisted transfer of a free flap was reported in 1973,4 prior to the 1990s there was limited enthusiasm in the United States to apply free flaps for reconstruction of head and neck defects.5 This reluctance arose from several perceived shortcomings of microvascular tissue transfer. Such concerns included questions regarding the reliability of a technique that was dependent on small vessel vascular anastomoses for a successful outcome and the potential for an adverse impact on the costs and complications of therapy.
As surgeons became more experienced with microvascular free flap techniques, the reliability of free flaps has improved steadily. An early survey revealed that the rate of successful free flap transfer was 89% during the first decade of clinical experience with microvascular surgery6 By the mid-l990s, several large series reported successful head and neck reconstruction using free flaps in 91% to 95% of cases.7-11 In l999, Blackwell1 described a success rate of 99% in 119 cases of microvascular head and neck reconstruction, while Singh et al2 reported success in 98% of 200 cases. Improved free flap reliability has been due to improved microvascular techniques, increased reliance on free flaps with long vascular pedicles that contain large-caliber blood vessels, and greater experience by individual surgeons.
The most common cause of free flap failure is thrombosis of the vascular pedicle in the region of the microvascular anastomosis.12 Microvascular reconstruction in the head and neck is more challenging in patients who have undergone previous neck dissection, owing to prior resection of potential recipient blood vessels. It is conceivable that a paucity of potential cervical recipient blood vessels might result in increased risk of free flap failure in patients with a history of previous neck dissection.
The present series details 59 patients who underwent free flap reconstruction following previous neck dissection, achieving a success rate of 100%. About 15% of the neck dissections were contralateral to the defect site. Prior radiation therapy does not appear to have a negative impact on flap survival since 71% of our patients had preoperative radiation therapy. The study therefore does not identify previous neck dissection as a risk factor associated with an increased rate of free flap failure. The data indicate that free flaps need not be avoided when there is a history of previous neck dissection. Although the flaps proved reliable, 61% of patients required use of cervical recipient blood vessels in the contralateral neck, reflecting an increased complexity of reconstruction. Use of contralateral cervical recipient blood vessels is rarely necessary in the absence of previous neck surgery.
The extent of the previous neck dissection had an impact on the need to rely on contralateral vessels. In approximately half the cases of previous selective neck dissection, cervical recipient blood vessels were successfully located in the field of the previous neck dissection. By contrast, all patients with a history of modified radical or radical neck dissection required use of cervical recipient blood vessels in the opposite side of the neck. This is likely a reflection of the unavailability of suitable cervical recipient veins in the field of previous modified radical or radical neck dissection. The facial and superior thyroid arteries are commonly ligated during cases of selective, modified radical, and radical neck dissection. However, the external carotid artery and medial branches of the external carotid artery such as the lingual artery are commonly preserved during neck dissection and may be available to use as cervical recipient blood vessels during subsequent cases of microvascular flap reconstruction. This may explain why the external carotid artery and lingual artery were used frequently within the ipsilateral neck (52% of ipsilateral recipient arteries) compared with the contralateral neck (11% of contralateral arteries) in the present series. However, the external and internal jugular veins are routinely sacrificed during a modified radical or radical neck dissection. This makes isolation of a recipient vein within the field of previous modified radical or radical neck dissection more difficult compared with after previous selective neck dissection, where the external or internal jugular veins are more commonly preserved.
Another approach when there is not a suitable recipient vein relies upon cephalic vein transposition.13 In cases of previous modified radical or radical neck dissection, the cephalic vein can be transposed from the ipsilateral arm to the neck to serve as a recipient vein for free flap perfusion. Advantages of this technique include the fact that only one microvenous anastomosis is required, and the high-flow, low-pressure cephalic-subclavian system may be resistant to stasis and thrombosis. The primary disadvantage of this technique arises from the increased potential of kinking or extrinsic compression of the venous pedicle within its long subcutaneous course, particularly where the cephalic vein crosses over the clavicle.
In cases where ipsilateral cervical recipient blood vessels are unavailable, vein grafts can be used to lengthen vascular pedicle to reach remote recipient vessels. In the present series, the need to use vein grafts was eliminated by careful preoperative planning and heavy reliance on free flaps that contain long vascular pedicles. Two previous large series of microvascular head and neck reconstruction have correlated the use of vein grafts with an increased risk of free flap failure, so they are best avoided whenever feasible.10,11 In the present series, radial forearm flaps, fibula flaps, rectus abdominis flaps, and subscapular system flaps accounted for the 98% of the donor sites selected. All of these flaps contain long vascular pedicles that usually can reach cervical recipient blood vessels in the contralateral neck without requiring the use of vein grafts.
Based on our experience with the patients in this series, we propose the following algorithm for microvascular flap reconstruction in patients with a history of neck dissection as related to factors that affect recipient vessel selection. The status of potential recipient veins within the field of previous neck dissection is usually the most critical factor in determining the selection of cervical recipient vessels for flap perfusion. Some length of the external carotid artery is usually preserved during most neck dissections, and an end-to-side arterial anastomosis to the external carotid artery can usually be performed even when all external carotid branches have been previously ligated.
In patients with a history of radical neck dissection or modified radical neck dissection, plans should be made for use of recipient vessels in the unoperated-on neck, as it has been our experience that recipient veins are seldom available owing to prior resection of the internal and external jugular venous systems in patients with previous radical neck dissection or modified radical neck dissection. In cases where the unoperated-on neck is contralateral to the defect site, selection of a free flap that offers a long vascular pedicle usually obviates the need to perform vein grafting, although all patients with previous neck dissection are routinely informed of and consented for the possibility to perform vein grafting if necessary.
In patients with a history of a selective neck dissection that is ipsilateral to the defect site, careful review of the previous operative report to determine the status of the internal jugular vein is recommended. In cases where the internal jugular vein was preserved during previous neck dissection, we have found that recipient vessels are usually available within the field of the prior neck dissection in approximately half the cases. In the other half of cases, it is necessary to use recipient vessels within the contralateral neck, owing to difficulty in isolating and preparing suitable recipient veins within the field of previous selective neck dissection, due to periadventitial scarring or perioperative thrombosis of an internal jugular vein that had been preserved during a prior selective neck dissection. Even in these cases, selection of flaps that offer long vascular pedicles and planning for possible vein grafting is desirable, as use of contralateral recipient vessels will be necessary in approximately 50% of cases.
Free flap reconstruction of the head and neck is highly successful in patients with a history of neck dissection, despite a relative paucity of potential cervical recipient blood vessels. Recipient vessels can be identified in the field of previous selective neck dissection in approximately half of such cases, while recipient vessels are rarely available in the field of a previous radical neck dissection. In most cases of microvascular reconstruction after previous neck dissection, it is necessary to use recipient vessels in the neck that is contralateral to the side of the defect. Heavy reliance on free flaps with long vascular pedicles eliminated the need to perform vein grafts in the present series, probably contributing to the absence of free flap failure. Previous neck dissection should not be considered a contraindication to microvascular reconstruction of the head and neck.
Accepted for publication December 6, 2001.
This study was presented at the annual meeting of the American Head and Neck Society, Palm Desert, Calif, May 14, 2001.
Corresponding author and reprints: Keith E. Blackwell, MD, Department of Surgery, Box 951624, UCLA School of Medicine, Los Angeles, CA 90095-1624 (e-mail: firstname.lastname@example.org).