Rachel Giese and Richard Wong Abstract
Radical neck dissection was historically the standard of care for the management of cervical nodal metastases. A radical neck dissection includes the complete resection of all five levels of the cervical lymphatics, and additionally sacrifice of the spinal accessory nerve, sternocleidomastoid and the internal jugular vein. This procedure causes considerable morbidity, predominantly from shoulder dysfunction, and numerous studies have since demonstrated that the surgical preservation of non-in- vaded structures was oncologically sound. More recent modifications of neck dissection are now performed that selectively remove the lymph nodes in the anatomic levels that are at highest risk for harboring metastatic nodal disease while preserving key structures, and carry much less morbidity. The following chapter addresses the history, techniques, morbidity, post-operative considerations, and potential complications of a radical neck dissection.
Keywords: radical, neck, dissection
6.1 History
The initial concept of surgically removing cervical lymphatics en bloc represented a major conceptual advance in oncology. The first reports of removal of neck metastases came from independent efforts of four European surgeons: von Langenbeck, Billroth, von Volkmann, and Kocher.1 Later, Sir Henry Butlin demonstrated that surgically removing metastases from the neck for oral cavity cancer improved survival. In 1888, a Polish surgeon, Franciszek Jawdynski, reported on a procedure similar to a radical neck dissection (RND). Early surgeries resulted in serious complications and morbidity. In 1906, George Crile popularized the systematic en bloc RND2 that included the removal of all structures in the neck from the mandible to the clavicle, superficial to the deep muscles of the neck, except the carotid artery, brachial plexus and the vagal, lingual, and hypoglossal nerves. The specimen encompassed lymph nodes and fi- brofatty tissue spanning levels I to V and includes the spinal accessory nerve (SAN), sternocleidomastoid muscle (SCM), and internal jugular vein (IJV). Recognizing that loss of these structures caused patients significant functional and cosmetic morbidity, Suarez modified the procedure in 1963 to preserve nonlymphatic structures when oncologically feasible. Bocca3 and Martin et al4 further refined the surgical procedure terming it a “functional neck dissection.” The adapted procedure became widely adopted as it became clear that this modification was both oncologically safe and imparted less morbidity to the patient by preserving structures that were uninvolved by disease. Since then, many retrospective studies have repeatedly demonstrated the oncologic safety in preserving these structures. Knowledge of lymphatic drainage patterns specific to neck dissection has led to further evolution of this procedure into what is today referred to as the modified RND and selective neck dissection. Although Crile’s neck dissection was initially the “gold standard neck dissection,” with modifications in the subsequent four decades following the initial description, it is now referred to as the RND. This term is appropriate because it is a radical removal of sites of metastatic spread to the neck, sacrificing the SAN, SCM, and IJV. However, it is rarely required today, except in cases of extensive disease, and the evolution of the procedure to less aggressive resection has greatly improved the postoperative quality of life for patients.
6.2 Classification/Terminology
Surgeons at Memorial Hospital first classified the cervical lymph nodes into levels and later Byers et al described the levels most at risk for spread based on the primary tumor site in the head and neck.5 When it became clear not all cervical lymphatics needed to be resected that an RND encompassed, surgeons modified the neck dissection to involve the lymphatics at highest risk for disease spread. Furthermore, the SAN, IJV, and SCM were preserved unless there was disease involvement. Because there were so many variations between lymphatic levels and structures sacrificed, the American Head and Neck Society (AHNS) standardized the terminology for classification of neck dissection in 1991 and there have been multiple revisions thereafter.6 Most recently, the AHNS defined the RND as a cervical lymphadenectomy including en bloc removal levels I to V, the IJV, SCM, SAN, and the submandibular gland.7 According to the AHNS classification, any neck dissection that preserves the three structures should be termed a “modified RND” and preserved structures should be stated when naming the procedure, an adaptation of Medina’s terminology for structures preserved published in 1989. Ferlito et al proposed a further modification to this terminology suggesting that each level and structure resected should be clarified, with an RND termed ND (I-V, SCM, IJV, and cranial nerve [CN] XI).8
Due to the rarity of patients requiring an RND and sacrifice of the SAN, SCM, and IJV, the “modified” RND is now the standard of care for most patients with metastatic nodal disease where one or more of these structures can be preserved. RND imparts considerable comorbidity to the patients and should be avoided when oncologically possible. The term “modified RND” refers to preservation of nonlymphatic structures, whereas the term “selective neck dissection” refers to preservation of levels of lymphatics. The standardization of terminology is important to communicate to other practitioners which levels were dissected and which structures were resected for postoperative management. Stating that a patient had an RND accurately transmits the idea of extensive or invasive metastatic nodal disease in the neck.
6.3 Indications for Radical Neck Dissection
The goal of curative oncologic surgery should be the complete removal of all neoplastic tissue. Surgeons must balance the oncologic advantages of RND against the morbidity it brings to the patient. An RND is now performed only when necessary due to the extent and growth pattern of the disease. In most cases, this is due to disease invading the SCM, IJV, and SAN.
Uncommonly, an RND is performed due to direct spread of a primary tumor. Although neck dissections are not pathologically assessed for margins unless it is an en bloc resection of direct tumor spread, surgeons have applied the same reasoning that is used in tumor resection when resecting the lymphatics of the neck. That is, preferably, the disease will be surrounded by nondiseased (normal) tissue. In most dissections, this means the disease is within the capsule of the lymph node, but this is not typical of nodal disease in most cases requiring RND. However, in cases with extracapsular spread, the resection of adjacent structures such as SCM, IJV, and SAN may be required to get a “margin” around the disease, justifying the RND.
Some have suggested that RND is appropriate routinely for N3 disease of the upper neck, invasion of the SAN, and/or recurrent or persistent disease following definitive radiotherapy, chemora- diotherapy, or previous selective neck dissection.9 Presumably, the recommendation of RND for all N3 diseases implies that a 6-cm metastasis would encompass the SAN and involve the SCM and IJV due to the confined anatomy of the neck and restricted space for spread. Such a global suggestion also likely stems from concerns that the posttreatment scarring may obscure tissue planes, making the clean removal of disease from the SAN, IJV, and SCM challenging. However, the specific relationships between disease and each anatomic structure in the neck should be better evaluated on a case-by-case basis, and adaptations and consideration of risks of involvement of each structure must be applied to each individual patient. It is possible that not all three structures (SAN, IJV, and SCM) implicit in RND need to be resected in every patient who has had a previous neck dissection or radiation.
Regarding indications for RND, some patients may have such extensive disease that more extended resection is needed than a traditional RND. Surgery may necessitate resection of structures outside the scope of a typical RND, including the strap muscles, digastric, omohyoid, cervical rootlets, brachial plexus, hypoglossal nerve, facial nerve, phrenic nerve, carotid artery, the skull base, or deep muscles of the neck. Involvement of these structures, especially the carotid artery, portends a very poor prognosis and thus may call into question the goals of the operation.
It is important to note not all cervical nodal disease may be surgically resectable. If disease may not be cleared even with extended RND, the morbidity of the surgery may not be justified. For example, if the carotid artery is involved and cannot be safely sacrificed or when sacrifice of the carotid is not guaranteed to remove all disease in the neck, it may be wiser to consider nonsurgical options including systemic therapy, radiation therapy, or entry into a clinical trial. Ultimately, the surgeon needs to discuss the options with the patients and tailor individualized treatment to each patient’s particular clinical situation.
6.4 Preoperative Counseling and Evaluation
Documentation of arm and shoulder range of motion and strength should be made prior to surgery to assess SAN and brachial plexus function. Photos may aid in the preoperative counseling of the patient to set expectations for postoperative neck, shoulder, and arm function (Fig. 6.1). The patient should be counseled about potential postoperative morbidity and the need for rehabilitation prior to surgery. Expected contour changes to the neck, numbness, the expected incision line, and possible deficits of the particular CNs at risk should be reviewed. Other risks of surgery including bleeding, infection, chyle leak, carotid injury, stroke, nerve deficits, unresectable disease, and possible rapid recurrence despite surgery may be discussed. Postoperative morbidity intrinsic to the RND that is additional to the modified radial or selective neck dissection includes impaired shoulder and arm mobility, weakness, pain, cosmetic deformity, and lymphedema. In the rare case of bilateral IJV sacrifice, the patient must be counseled of the risk of potential facial and neck edema and decreased cerebral venous outflow. Counseling the patient preoperatively about the need for postoperative physical therapy may improve compliance with a postoperative rehabilitation program.
Patients should be counseled preoperatively about the possible morbidity of RND, but the decision to resect structures is sometimes made intraoperatively when assessment of resectability of the disease can be better assessed. As a general guideline, if a structure is not clearly involved by disease, and a clean resection can be achieved, then it should be structurally preserved.
One example may be a metastasis that is immobile in preoperative clinical evaluation because it is adherent to the SCM. Intraoperatively, after the skin and platysmal flaps are raised and superior and inferior SCM attachments are released, the tumor may become more mobile. Tumor fixation to the SCM may make it seem immobile, but there may be tissue planes between the IJV and SAN that facilitate its removal, and allow preservation of these structures. Conversely, some neck dissections may require removal of parapharyngeal and paraspinal lymphatics that are not included in the levels I to V of the RND. CT or MRI images with intravenous contrast should be carefully evaluated to assess relationships between nodal disease and normal structures (Fig. 6.2, Fig. 6.3, Fig. 6.4).
Another important feature to note on preoperative imaging tumor is the relationship of the nodal disease with the carotid artery. If tumor is more than three quarters abutting the circumference of the carotid artery, there is a higher possibility that the common carotid may need to be sacrificed. In these cases, the pros and cons ofcarotid resection should be considered, and justification for performing a neck dissection should be made. Often neck dissection may not be indicated due to the extent of disease, the higher risks of surgery, and the low likelihood of disease eradication. In such a situation, conducting a detailed discussion with the patient, the patient’s family, and the multidisciplinary medical team is essential to establish goals of care. If surgery is proposed, preoperative balloon occlusion angiography should be performed to assess the patency of the circle of Willis cerebral perfusion in the event of carotid resection. Vascular surgery should be on standby for possible carotid bypass and grafting if required. Regional or free tissue flap closure may be necessary to provide soft-tissue coverage of the exposed, dissected, or reconstructed carotid artery. Additional considerations of the carotid artery include thick atherosclerotic disease. In patients with atherosclerosis of the carotid, it is critical not to aggressively manipulate the carotid artery because it can cause embolus of atherosclerotic plaque material to the cerebral circulation.
6.5 Surgical Technique and Anatomic Considerations
A variety of skin incisions may be used to perform an RND, including those described by Martin4 (double Y), Crile2 (trifurcate), MacFee10 (double transverse) incision, Schobinger,11 and Babcock and Conley.12 A trifurcate incision offers the advantage of optimal exposure, easier posterior flap elevation, and dissection of the posterior triangle of the neck. The disadvantage is wound dehiscence at the union of the three flaps because the blood supply is divided between three flaps, especially in the irradiated neck.13 If the trifurcate incision is used, it is important to make sure the vertical point of the “T” incision is at least 2 cm posterior to the common carotid artery to reduce the risk of carotid exposure (Fig. 6.5).
After the skin incision, subplatysmal skin flaps are raised. If the tumor extends to the platysma, skin flap should be raised superficial to the platysma. Raising supraplatysmal a flap leads to a thinner skin flap that has less vascularity and offers less protection for underlying structures, particularly in the event that postoperative radiation is necessary. If the tumor is fixed to the dermis, resecting skin with the rest of the specimen is necessary to achieve clear margins (Fig. 6.6). In these scenarios, local or free tissue transfer may be needed to achieve appropriate skin closure and protection of underlying neck structures.
6.5.1 Internal Jugular Vein
Dissection of the IJV both above and below the level of the disease will achieve proximal and distal control. It may be advisable to use vessel loops to apply compression to the IJV for vascular control during the dissection (Fig. 6.7). Ligation of the IJV may be secured using a modified Halsted transfixation ligature (also known as a “stick tie”). With this approach, the vessel is clamped and a simple 2-0 silk ligature is placed. Next, before the clamp is released, a 2-0 silk with a needle is used to place a Halsted transfixation ligature (Fig. 6.8). The other end of the IJV can be ligated in a similar fashion.
6.5.2 Spinal Accessory Nerve
Resection of the SAN is arguably the most morbid effect of RND. The SAN contains both sensory and motor fibers. It connects to the cervical rootlets C2-C4 and some suggest it is better described as the SAN plexus.14 Three anatomic patterns of the SAN have been described.15 With the most common variation present in 66% of patients, the nerve courses into the SCM before dividing into branches to the trapezius and SCM. With the second most common variation, present in 22% of patients, the nerve divides before entering the SCM. This configuration facilitates preservation of the branch to the trapezius even when the branch to the SCM must be sacrificed. In every case, a meticulous attempt should be made to preserve the SAN.
Intraoperatively, the SAN may be reliably identified with retraction of the SCM laterally. The SAN descends from the skull base adjacent to the jugular vein and courses obliquely and posteriorly into the SCM. However, visualization of the SAN may not be feasible if there is bulky disease at level II. The SAN may also be identified posterior to the SCM near Erb’s point where the cutaneous nerves and cervical rootlets course. If the SAN is being traced and preserved, care must be taken to avoid confusing the branch to the trapezius with sensory nerves at Erb’s point (Fig. 6.9). Finally, the SAN may also be reliably identified at its entry deep to the trapezius muscle in the posterior triangle. Intraoperative nerve monitoring of trapezius function may have utility in facilitating the identification and preservation of the SAN, particularly in cases with high-volume disease in the neck. If the SAN requires resection, a nerve graft (e.g., sural nerve) may help decrease atrophy of the trapezius and preserve muscle tone.16.17 Nerve grafting is optimally performed at the time of SAN resection. Other experimental techniques have applied a bioabsorbable conduit to incorporate the two cut ends of the nerves.18
Fig.6.7 Vessel loops around internal jugular vein and carotid prior to dissection allow for vascular control.
Fig.6.8 Double suture ligation of the internal jugular vein using Halsted’s transfixation technique.
6.5.3 Other Structures
Although much attention is paid to the SCM, SAN, and IJV in discussions of the radical neck, it is important to acknowledge other structures that are at risk with neck dissection for aggressive or high-volume disease in the neck. Importantly, the vagus nerve, hypoglossal nerve, phrenic nerve, brachial plexus, cervical sympathetic chain, thoracic duct, deep neck musculature, and other structures may require resection depending on the extent of disease. Intraoperative nerve monitoring may have significant utility for monitoring motor nerves when disease is known to be in proximity to these structures, and aids in nerve identification, dissection, and prognostication of nerve function. Lymphatics should be carefully ligated with silk suture to avoid chylous fistula, and Valsalva maneuvers help identify lymphatic leaks.
6.5.4 Coverage of Carotid Artery
Regional flap closure or free tissue transfer may be necessary in situations with significant skin and soft-tissue resection, carotid exposure, and particularly in the setting of previously radiated tissues. Prior radiation leads to diminished vascular perfusion to the skin flaps. In addition, most patients undergoing a neck dissection will require postoperative radiation and well-perfused tissue coverage of the carotid artery will afford better protection against carotid rupture.
When the IJV is sacrificed in the RND, the transverse cervical and external jugular veins are often the only ipsilateral remaining veins to which a free tissue transfer donor vein can be anastomosed. Preservation of the external jugular vein is prudent for use in reconstructive vascular anastomosis. However, the external jugular vein is superficial and susceptible to compression, and oriented in a superoinferior direction susceptible to rotation and occlusion. The transverse cervical vein may be used, but is located very inferior in the neck and therefore a longer flap vascular pedicle may be required to reach this recipient vein.
If a portion of the SCM can be preserved, the remnant may sometimes be rotated over the carotid sheath and sutured to the strap muscles to achieve coverage. The segmental blood supply of the SCM is an important consideration when using it for a rotational flap. Another option is a local tissue transfer using the pectoralis, or deltopectoral flaps, which are good alternatives for patients in whom a free tissue transfer may not be an option due to patient comorbidity, poor recipient vessels, or inadequate donor tissue.
Fig. 6.9 (a,b) The branch of the spinal accessory nerve (SAN) to the trapezius runs should be identified, traced, and preserved if oncologically feasible. Care must be taken to distinguish the SAN from the cervical plexus.
6.6 Postoperative Complications and Management
The simultaneous resection of bilateral IJVs should be avoided, as this may lead to severe facial edema and rarely even cerebral edema. Angiography after bilateral IJV ligation has shown the sigmoid sinus and brain drain through the vertebral venous (Batson’s) plexus deep in the neck and empty into the subclavian vein at the cervicothoracic junction.19 Although Ahn et al reported that staging the procedures by doing one RND at a time does not decrease the risk of cerebral edema,20 anecdotal reports suggest that staged procedures are preferred and result in less severe facial edema, allowing for the interval development of vertebral venous plexus collaterals.
Sacrifice of the SAN, including the motor branch to the trapezius, motor branch to the SCM, and the sensory nerve plexus, is the most functionally devastating aspect of an RND. Selective neck dissection can also cause SAN palsy due to traction, demyelination, or devascularization, even when the nerve is not transected. The trapezius stabilizes the shoulder girdle, neck, back, and chest. It inserts on the occiput, the spinous process of the cervical and thoracic vertebrae, the acromion, and the clavicle. Accordingly, trapezius denervation affects shoulder abduction and can cause internal rotation (“winging”) of the scapula, and drooping of the shoulder. Trapezius denervation is not an isolated muscle weakness. Because the scapula is destabilized, muscles that insert on the scapula have decreased strength as well. These include the serratus anterior, deltoid, supraspinatus, teres, and infraspinatus muscles. Impairment of this complex of muscles further exacerbates the shoulder destabilization from the trapezius denervation. The resulting inability to abduct the shoulder against gravity is termed “shoulder syndrome,” first described by Ewing and Martin in 1952.21 Functionally, this is manifested as diminished ability to don and doff a jacket sleeve, to style one’s hair, or to lift objects.
All patients should be assessed for SAN palsy after any type of neck dissection and physicians should remember SAN palsy is not obvious in all patients immediately after RND surgery. While pain is almost always present, shoulder weakness may be a delayed finding.17 Pain may be due to cut afferent fibers in the SAN or due to traction on the shoulder from the weight of the unsupported arm.21 Often the patient has normal immediate postoperative shoulder function in the hospital because the trapezius has not yet atrophied and the shoulder girdle muscles are able to compensate for short term after injury. In addition, there is axoplasmic flow in the distal end of the nerve for weeks.22 Patients receiving adjuvant chemotherapy and radiation exhibit worse shoulder function than those treated with neck dissection alone, even in cases in which the SAN is spared.23
A reliable sign of SAN palsy is the scapular flip test24 in which the patient stands with arms by their sides and elbows bent 90 degrees. The patient attempts to externally rotate the shoulder, while the examiner resists at the distal forearm. If the patient has SAN palsy, the medial scapula will flip or lift from the thoracic wall. This maneuver distinguishes the SAN injury from long thoracic nerve injury. Interestingly, trapezius denervation may result in increased FDG activity on postoperative PET scan due to increased glucose utilization in denervated muscle.25
Occupational or physical therapy for SAN palsy after RND has been shown to improve patients’ postoperative function.22 Passive and active range of motion exercises are used to rehabilitate the shoulder. Therapy should be initiated within 8 weeks and most therapy lasts for up to 3 months.22.26 In addition to improving shoulder function, therapy can also help patients who may have resting pain due to muscle spasm of the “surviving muscles,” that is, muscle fibers of the trapezius with innervation from C2, C3, and C4 roots.27 In some patients, denervation is more painful and functionally disruptive than others. Sacrifice of the SAN can result in a range of impaired shoulder abduction, from as low as 35 degrees to as high as 80 degrees (with 180 being normal).27 These variations may be due to anatomic variations although the exact cause is not understood.14
SCM resection or denervation due to nerve resection does not cause as many functional complications as trapezius denervation. Cosmetically, however, loss of the soft tissue in a unilateral resection of the SCM can be very distressing to the patient. Furthermore, the brachial plexus and scalene muscles may be more evident just beneath the skin flap. There may be hollowing of the supraclavicular fossa as well. Accordingly, preoperative counseling of cosmetic as well as functional complications is critical.
Another complication of RND is chyle leak that results from inadvertent injury during dissection in level IV to the thoracic duct or its tributaries. Ideally a chyle leak will be identified intraoperatively and the duct and its branches will be ligated with suture or metal clips. Some surgeons have reported success with use of omohyoid rotational flaps and cyanoacrylate or fibrin glue to seal the area.28
Postoperative management is conservative at first with pressure dressings and a medium chain triglyceride and low fat diet. Somatostatin analogs (octreotide) or pancreatic lipase inhibitors (etilefrine) may be added. If high drain output fails to resolve, embolization or surgical ligation is necessary to resolve the leak. Although not a common complication, there are case reports of patients undergoing bilateral RND experiencing operative hypertension due to baroreflex dysfunction.29 This is attributed to dissection of the nerve to the carotid sinus during dissection along the carotid sheath (Fig. 6.10). The literature shows varying long-term sequelae: some patients have persistent hypertension, others become normotensive, and some have labile blood pressure. Risk factors include advanced age and preoperative radiation, and it usually happens in the immediate postoperative period when the vasodilatory effects of anesthesia have subsided. The best treatment for baroreflex failure is clonidine to stimulate the alpha-2 receptors centrally and decrease sympathetic tone.30 In summary, the RND can impart severe postoperative morbidity to the patient, which underscores the need for meticulous attention during surgery toward the preservation of noninvolved structures, as well as detailed preoperative patient counseling to appropriately set expectations.
6.7 Conclusion
Although the RND was the historical standard of cervical lym- phadenectomy, it has appropriately evolved such that it is only required currently in a select group of patients with aggressive nodal disease who require removal of the IJV, SAN, and SCM. Sacrifice of these structures, in particular the SAN, imparts functional and cosmetic morbidity to the patient. Careful preoperative image review is necessary to plan a neck dissection to achieve optimal clearance of disease, functional outcome, and appropriate coverage of the carotid artery. Patients should be counseled preoperatively about potential functional impairments and enrolled in rehabilitation programs postoperatively.
Fig. 6.10 Dissection along the carotid artery separates the internal jugular vein with the specimen.
References
[1] Rinaldo A, Ferlito A, Silver CE. Early history of neck dissection. Eur Arch Oto- rhinolaryngol. 2008; 265(12):1535-1538
[2] Crile G. Landmark article Dec 1, 1906: excision of cancer of the head and neck. With special reference to the plan of dissection based on one hundred and thirty-two operations. By George Crile. JAMA. 1987; 258(22):3286-3293
[3] Bocca E. Supraglottic laryngectomy and functional neck dissection. J Laryngol Otol. 1966; 80(8):831-838
[4] Martin H, Del Valle B, Ehrlich H, Cahan WG. Neck dissection. Cancer. 1951; 4 (3):441-499
[5] Byers RM, Wolf PF, Ballantyne AJ. Rationale for elective modified neck dissection. Head Neck Surg. 1988; 10(3):160-167
[6] Robbins KT, Medina JE, Wolfe GT, Levine PA, Sessions RB, Pruet CW. Standardizing neck dissection terminology. Official report of the Academy's Committee for Head and Neck Surgery and Oncology. Arch Otolaryngol Head Neck Surg. 1991; 117(6):601-605
[7] Robbins KT, Shaha AR, Medina JE, et al. Committee for Neck Dissection Classification, American Head and Neck Society. Consensus statement on the classification and terminology of neck dissection. Arch Otolaryngol Head Neck Surg. 2008; 134(5):536-538
[8] Ferlito A, Robbins KT, Shah JP, et al. Proposal for a rational classification of neck dissections. Head Neck. 2011; 33(3):445-450
[9] McCammon SD, Shah JP. Radical neck dissection. Oper Tech Otolaryngol. 2004; 15(3):152-159
[10] MacFee WF. Transverse incisions for neck dissections. Ann Surg. 1960; 151 (2):279-284
[11] Schobinger R. The use of a long anterior skin flap in radical neck resections. Ann Surg. 1957; 146(2):221-223
[12] Babcock WW, Jr, Conley J. Neck incision in block dissection. Experiences with the long anterior cervical flap incision. Arch Otolaryngol. 1966; 84(5):554-557
[13] Yii NW, Patel SG, Williamson P, Breach NM. Use of apron flap incision for neck dissection. Plast Reconstr Surg. 1999; 103(6):1655-1660
[14] Brown H, Burns S, Kaiser CW. The spinal accessory nerve plexus, the trapezius muscle, and shoulder stabilization after radical neck cancer surgery. Ann Surg. 1988; 208(5):654-661
[15] Lanisnik B, Zargi M, Rodi Z. Identification of three anatomical patterns of the spinal accessory nerve in the neck by neurophysiological mapping. Radiol Oncol. 2014; 48(4):387-392
[16] Goransson H, Leppànen OV, Vastamàki M. Patient outcome after surgical management of the spinal accessory nerve injury: a long-term follow-up study. SAGE Open Med. 2016; 4:2050312116645731
[17] Chandawarkar RY, Cervino AL, Pennington GA. Management of iatrogenic injury to the spinal accessory nerve. Plast Reconstr Surg. 2003; 111(2):611- 617, discussion 618-619
[18] Ducic I, Maloney CT, Jr, Dellon AL. Reconstruction of the spinal accessory nerve with autograft or neurotube? Two case reports. J Reconstr Microsurg. 2005; 21(1):29-33, discussion 34
[19] Ensari S, Kaptanoglu E, Tun K, et al. Venous outflow of the brain after bilateral complete jugular ligation. Turk Neurosurg. 2008; 18(1):56-60
[20] Ahn C, Sindelar WF. Bilateral radical neck dissection: report of results in 55 patients. J Surg Oncol. 1989; 40(4):252-255
[21] Ewing MR, Martin H. Disability following radical neck dissection; an assessment based on the postoperative evaluation of 100 patients. Cancer. 1952; 5 (5):873-883
[22] McGarvey AC, Chiarelli PE, Osmotherly PG, Hoffman GR. Physiotherapy for accessory nerve shoulder dysfunction following neck dissection surgery: a literature review. Head Neck. 2011; 33(2):274-280
[23] Gallagher KK, Sacco AG, Lee JS, et al. Association between multimodality neck treatment and work and leisure impairment: a disease-specific measure to assess both impairment and rehabilitation after neck dissection. JAMA Otolaryngol Head Neck Surg. 2015; 141(10):888-893
[24] Kelley MJ, Kane TE, Leggin BG. Spinal accessory nerve palsy: associated signs and symptoms. J Orthop Sports Phys Ther. 2008; 38(2):78-86
[25] Lee SH, Seo HG, Oh BM, Choi H, Cheon GJ, Lee SU. Increased (18)F-FDG uptake in the trapezius muscle in patients with spinal accessory neuropathy. J Neurol Sci. 2016; 362:127-130
[26] McGarvey AC, Hoffman GR, Osmotherly PG, Chiarelli PE. Maximizing shoulder function after accessory nerve injury and neck dissection surgery: a multicenter randomized controlled trial. Head Neck. 2015; 37 (7):1022-1031
[27] Chida S, Shimada Y, Matsunaga T, Sato M, Hatakeyama K, Mizoi K. Occupational therapy for accessory nerve palsy after radical neck dissection. Tohoku J Exp Med. 2002; 196(3):157-165
[28] Brennan PA, Blythe JN, Herd MK, Habib A, Anand R. The contemporary management of chyle leak following cervical thoracic duct damage. Br J Oral Maxillofac Surg. 2012; 50(3):197-201
[29] Prakash S, Rapsang A, Kumar SS, Bhatia PS, Gogia AR. Postoperative hypertension following radical neck dissection. J Anaesthesiol Clin Pharmacol. 2012; 28(1):121-123
[30] Wijeysundera DN, Naik JS, Beattie WS. Alpha-2 adrenergic agonists to prevent perioperative cardiovascular complications: a meta-analysis. Am J Med. 2003; 114(9):742-752