Chronic shoulder pain due to spinal accessory nerve palsies present opportunities for improved care integration

Abstract

Aim: Chronic shoulder pain due to iatrogenic spinal accessory nerve (SAN) injury continues to be under-recognized, resulting in delayed time-to-diagnosis and poorer outcomes. Solutions are needed to improve the management of this condition, which can be challenging as care needs to be coordinated across pain management, neurophysiology, rehabilitation and reconstructive surgery.

Cases: We present a series of six patients with shoulder pain refractory to conservative pain treatments to highlight how SAN injuries continued to be missed and treatment delayed, even at advanced care centers. The time to diagnosis of SAN palsy took an average of 21 months and treatment was inconsistent for all patients.

Discussion: None of the six cases had initial suspicion of SAN palsy and only one patient received targeted SAN injury care. SAN treatment should be started as early as possible so that patients can be referred for prompt surgical evaluation if they fail conservative management. Integrated care pathways may be a solution for formalizing multidisciplinary team involvement and improving SAN injury outcomes.

Conclusion: Systemic processes, such as integrated care pathways, are needed to optimize early recognition and targeted treatment of SAN injury and may be beneficial for other underdiagnosed and undertreated neuropathic pain conditions.

Plain Language Summary

Neck dissection surgeries, which are done to treat various head and neck cancers, can often lead to shoulder pain. This pain commonly involves damage to a nerve called the spinal accessory nerve (SAN). Even though this nerve injury has been well-described by experts, many patients continue to experience delays in getting a correct diagnosis and treatment for this pain.

This study describes six cases from a large cancer center where patients developed chronic shoulder pain after neck dissection. These patients waited an average of 21 months to reach a correct diagnosis of SAN injury. The pain and symptoms were often mistaken for other conditions. In one case, a patient received delayed treatment from a coordinated team of medical specialists, leading to improvements in her pain and shoulder function.

The delay in diagnosing SAN injuries is likely due to two main challenges: a lack of awareness among providers and a tendency for healthcare to focus on specific areas rather than a team-based approach. Early diagnosis is crucial to prevent and minimize long-term pain and muscle loss.

To improve patient care, this study suggests using an integrated care pathway approach and gives an example of how providers can adopt this approach. This method involves coordinating various medical specialists to ensure prompt diagnosis and treatment. By improving how SAN injuries are managed, patients can receive better care and avoid chronic pain and disability.

Plain language summary

Article highlights.

• Spinal accessory nerve injury is still commonly underdiagnosed, even at major medical centers, despite the condition being well-documented in the medical literature.

• Continued underdiagnosis leads to persistent chronic shoulder pain and disability, as well as missed opportunities to improve patient care.

• We discuss six recent cases and how a health systems approach can address the challenge of diagnosing and managing rare neuropathic conditions like spinal accessory nerve injury.

1. Introduction

There are over 22,000 neck dissection surgeries performed annually in the USA for a wide variety of head and neck neoplasms, with up to 100% of radical and modified radical neck dissections and 25% of selective neck dissections resulting in persistent unilateral shoulder pain [1,2]. There is well-documented literature that post-neck dissection unilateral shoulder pain and dysfunction is primarily due to spinal accessory nerve (SAN) injury, even when the nerve is surgically spared during modified and selective neck dissection [3–5]. This is because the SAN is prone to traction-related injury when its adjacent lymph nodes are dissected, which then results in trapezius motor neuropathy and pain [3,6]. However, even though the literature association is well-known, the clinical diagnosis of SAN related chronic shoulder pain continues to be significantly delayed, as seen in our multidisciplinary medical and surgical experiences at a comprehensive cancer care center. We highlight five recent consecutive cases of chronic unilateral shoulder pain after neck dissection that received either misdiagnoses or no diagnosis for over a year on average, describe one case that benefitted from coordinated multidisciplinary care, and propose an integrated care strategy to improve SAN palsy recognition and treatment. Despite much literature existing on SAN injury incidence and outcomes, the condition is still commonly misdiagnosed and mismanaged. We propose a health services approach to improving the diagnosis and management of SAN injury, an approach that has not yet been explored for rare pain conditions like SAN injury [7].

SAN palsy is often missed on clinical examinations [3,8,9]. Patients may display preserved strength on shoulder shrug due to compensation from the levator scapula and rhomboid muscles [3]. Sternocleidomastoid muscle testing is unreliable for SAN palsy as its strength is often intact due to partial cervical nerve root innervation [9]. The hallmark findings of trapezius atrophy and shoulder girdle depression may be missed if patients remain clothed for the physical examination [3]. Clinicians must have early suspicion when patients with unilateral shoulder pain after neck dissection present with greater loss of coronal shoulder abduction (often limited to active abduction of 90° or less) compared with forward flexion and with a positive scapular flip sign [3,10]. A positive scapular flip is lifting of the medial scapular border when patients perform resisted shoulder external rotation. It occurs due to the paralysis of the middle and lower trapezius muscles, resulting in unopposed deltoid and infraspinatus activation, and helps distinguish SAN palsy from the medial winging with forward flexion seen in long thoracic nerve palsies [11,12].

Early diagnosis of SAN palsy is crucial for proper recovery, so that targeted treatment plans can be started, including early consideration of surgical reconstruction. Current existing cancer survivorship care guidelines do not reflect the full extent of treatment options available to patients with neck dissection related shoulder pain, such as pain neurorehabilitation through neuromuscular electrical stimulation, nerve repair or nerve transfer, reconstructive muscle–tendon transfer surgery, and emerging peripheral nerve stimulation techniques [5,13–17]. The delayed diagnosis of SAN palsies and the limited utilization of available treatment options is likely a consequence of our modern healthcare system’s siloed healthcare approach, in which medical care is structured to focus on particular organ systems rather than on collaboration across disciplines [18,19]. The concept of siloed healthcare is not a new one, as healthcare policy experts have long acknowledged that healthcare silos play an enormous role in inefficient and ineffective healthcare [7,20]. Peripheral nerve pathologies, such as SAN injury, are particularly susceptible to being missed because of siloed healthcare, as their diagnosis and management need expertise across the fields of neurology/neurophysiology, pain management, rehabilitative medicine and peripheral nerve or reconstructive surgery. In some instances, providers may not even realize there are interventional pain or surgical reconstructive options for SAN palsy related shoulder pain and dysfunction.

2. Case series

First, we describe five cases of unilateral chronic shoulder pain that presented consecutively to our center’s clinical neurophysiology laboratory over 4 months and then a sixth case that was seen at our institution as a second opinion. While the electrodiagnostic localization process was straightforward, we use these cases to highlight how SAN palsies can be missed and that even after diagnosis of SAN palsies, additional treatment options are not consistently offered. This case series received research approval from our Institutional Review Board. The mean age of all patients was 57 years of age (range 44–69) at the time of electrodiagnostic testing, which was performed with surface nerve conduction studies with needle electromyography. Five were male patients with oropharyngeal squamous cell carcinoma and one was a female patient with papillary thyroid carcinoma. Patients waited a mean of 21 months (range 8–68 months) until they were referred to electrodiagnostic testing for persistent unilateral shoulder pain and formally diagnosed with SAN palsies. See Table 1 for a summary of cases.

Table 1.

Delayed confirmatory diagnosis of spinal accessory nerve palsies.

Case	Age	Sex	Cancer	Cancer treatment	Delay to diagnosis	EMG indication	SAN treatment
1	69	Male	Squamous cell carcinoma	Selective ND, RT, CT, IT	12 months	Clinical trial screen	Pain medications, physical therapy
2	55	Male	Squamous cell carcinoma	Selective ND, RT, CT	23 months	Shoulder pain	Pain medications
3	52	Male	Squamous cell carcinoma	Selective ND, RT, CT, IT	19 months	Shoulder impingement	Pain medications, physical therapy
4	56	Male	Squamous cell carcinoma	Selective ND, RT, CT	8 months	Shoulder weakness	Pain medications, injections, physical therapy
5	66	Male	Squamous cell carcinoma	Selective ND, RT, CT	68 months	Shoulder weakness/pain	Physical therapy
6	44	Female	Papillary thyroid carcinoma	Radical ND, RAI	8 months	Suspected axillary neuropathy	Pain medications, injections, physical therapy, reconstructive surgery

CT: Chemotherapy; EMG: Electromyography; IT: Immunotherapy; ND: Neck dissection; RAI: Radioactive iodine; RT: Radiation therapy; SAN: Spinal accessory nerve.

In Case 1, a 69-year-old male patient had a history of chemoradiation therapy to the bilateral neck before selective neck dissection for recurrent disease, which was followed by additional radiation therapy and immunotherapy. After radiation therapies, he developed mild submental lymphedema and progressive radiation-induced cervical fibrosis and dysphagia. He reported occasional shoulder pain before neck dissection when lifting heavy objects but noted new left shoulder muscle atrophy, weakness and pain after neck dissection. He was prescribed hydrocodone for shoulder concerns and referred to community physical therapy for shoulder pain. Therapy sessions were focused on lymphedema management and neck range of motion. His initial diagnosis was postsurgical lymphedema. One year after neck dissection, he was referred for electrodiagnostic testing as part of a clinical trial and was found to have severe ongoing left SAN palsy. A repeat electrodiagnostic examination showed continued chronic SAN palsy findings. The patient passed away shortly afterward.

In Case 2, a 55-year-old male patient had a history of radiation therapy, then developed local recurrence that was treated by induction chemotherapy, selective neck dissection and mandibulectomy, followed by postsurgical chemoradiation. He received inpatient physical therapy focused on mobility. Several months later, he was noted to have developed submental lymphedema, trismus, osteoradionecrosis and left shoulder pain requiring transdermal fentanyl and gabapentin after having failed multiple other treatments. His initial diagnosis was postsurgical shoulder pain. On examination, his left shoulder girdle was atrophic with scapular winging. Nearly 2 years after neck dissection, electrodiagnostic testing was completed showing severe chronic SAN palsy. He continued pharmaceutical management and later developed new findings of postradiation brachial plexopathy.

In Case 3, a 52-year-old male patient had a history of chemoimmunotherapy and chemoradiation therapy followed by selective neck dissection. Shortly after neck dissection, he developed unilateral shoulder weakness and discomfort and was given home stretching exercises. His initial diagnosis was post-surgical shoulder impingement. About 1.5 years after neck dissection, he had electrodiagnostic findings showing severe chronic SAN palsy with mild asymptomatic multilevel cervical radiculopathy. No SAN-specific treatment was pursued. A few years later, he developed bilateral cervical torticollis requiring chemodenervation and was then referred for physical therapy. A progressive resistance regimen for shoulder girdle strengthening aided in minimizing, but not resolving, pain relief and shoulder dysfunction.

In Case 4, a 56-year-old male patient had a history of chemoradiation followed by several years of disease-free survival until local recurrence requiring partial glossectomy and selective neck dissection, followed by radiation therapy. He noted shoulder weakness and pain after neck dissection and began outpatient physical therapy. He had minimal improvement after several months of compliant resistance-based physical therapy and a glenohumeral corticosteroid injection. His initial diagnosis was post-surgical shoulder girdle weakness, lymphedema and skin fibrosis. Eight months after neck dissection, electrodiagnostic testing was ordered and consistent with severe chronic SAN palsy. Disease progression was noted shortly afterward, resistant to multiple rounds of chemotherapy and radiation. The patient passed away while in hospice care.

In Case 5, a 66-year-old male patient had a history of chemoradiation therapy and no further evidence of disease. Several years after treatment, he presented with progressing unilateral shoulder pain and dysfunction with overhead activities. He had some improvement in symptoms with several months of physical therapy. His initial diagnosis was post-surgical shoulder girdle weakness and shoulder pain. More than 5 years after neck dissection, electrodiagnostic examination revealed moderate ongoing and chronic SAN palsy as well as moderate chronic cervical radiculopathy without active denervation. The patient reported no hand paresthesias. He stopped physical therapy after plateauing in progress.

In case 6, a 44-year-old female patient with a history of metastatic papillary thyroid carcinoma benefitted from coordinated multi-disciplinary care involving neurophysiology, pain management, rehabilitation, plastics and neurosurgery providers. This patient reported a gradual onset of excruciating right-sided neck spasms and shoulder weakness a few weeks after total thyroidectomy and radical neck dissection at an outside institution. She was initially diagnosed with suspected axillary neuropathy and underwent electrodiagnostic testing that confirmed SAN palsy about 8 months after neck dissection by an external provider. One month later, she underwent surgical exploration with neurolysis and had several weeks of pain relief and improved shoulder function. Unfortunately, her symptoms recurred, and she arrived at our institution with severe right shoulder pain and hypersensitivity, absent shoulder shrug and shoulder abduction to 30–40°. She underwent a diagnostic SAN block with 2% lidocaine which resulted in immediate but shortlived pain relief. Fifteen months after neck dissection, she then had a repeat neck exploration with nerve autograft taken from the lesser occipital nerve and was referred for extensive physical therapy and occupational therapy. Several months later, she no longer needed opioids, was able to perform shoulder shrug, and regained limited shoulder abduction to 80–90°.

3. Discussion

In all six cases, none were referred for electrodiagnostic testing for clinical suspicion of SAN palsy (Table 1). Time-to-diagnosis varied greatly from 8 months to over 5 years even though patients presented with shoulder symptoms shortly after neck dissection. Given that muscle atrophy occurs within days of muscle inactivity, a well-conducted physical examination of the shoulders within a few weeks post-neck dissection could reveal early findings of SAN palsy [3,21]. We suspect that the delayed diagnosis may be due to the lack of diagnostic awareness of SAN palsies as a cause for shoulder pain and/or the lack of knowledge of the components needed for a comprehensive shoulder physical examination in providers who initially evaluate these patients. Often, these providers may be mid-level practitioners unfamiliar with the full scope of head and neck cancer complications. Additionally, postsurgical video telemedicine follow-ups may be insufficient for conducting a thorough shoulder examination and miss the diagnosis of SAN palsies.

It is crucial to diagnose peripheral nerve injuries like SAN palsies early to minimize the development of chronic neuropathic pain, irreversible muscle atrophy, and secondary musculoskeletal injuries such as adhesive capsulitis and rotator cuff tendinopathy [22–24]. Surgical nerve repair outcomes have been noted to be better when performed within 8 months of SAN injury [25,26]. While surgical repair ultimately benefited Case 6, the patient may have achieved an even greater return of her shoulder abduction and functionality if she received earlier treatment. A systematic review has shown that the majority of surgical options for SAN palsies report improved pain and functional outcomes with high rates of fair-to-excellent patient satisfaction [15]. Benefits may be maintained long-term as well. In a 10-year outcomes study, the majority of SAN neurolysis cases reported minimal trapezius atrophy and pain [26]. Other nonsurgical therapies have also been reported for SAN palsies, such as transcutaneous neuromuscular stimulation, regional nerve blocks and peripheral nerve stimulation [14,16,17,27].

Similar to our findings, a leading tertiary care peripheral nerve center reviewed that their time to diagnosis and referral for SAN palsies was also over 1 year from injury onset [8]. Despite the common incidence of postoperative SAN palsies after neck dissection and the known risk of SAN palsies developing after radiation therapy, we continue to face barriers in directing patients to appropriate peripheral nerve providers or clinics for confirmatory electrodiagnostic testing, rehabilitation, pain management and surgical evaluation [28,29]. There is a need for more research on treatment options for SAN palsies in cancer patients, but this need is difficult to address if these patients are not being identified and referred for care promptly. This is the double bind that many rarer pain conditions are caught in, in which first-line providers do not recognize the condition for referral, and specialist providers see patients far too late in their diagnosis to research outcomes from early interventions.

Our ability to address siloed care and the double bind it creates for clinical practice and research will shape the future of high-quality pain care [18,30,31]. There are multiple emerging proposals for addressing these silos, from top-to-bottom organizational restructuring to the creation of regular multidisciplinary case conferences [18,30]. Both these two options can be time-intensive and require extensive stakeholder buy-in. We suggest, as providers in the fields of neurophysiology, cancer rehabilitation, interventional pain management and peripheral nerve surgery, that an integrated care pathways (ICPs) approach for SAN palsies is worth exploring as a way to address siloed care with minimal infrastructure change [32]. ICPs could be applied to other complex pain conditions as well.

ICPs are structured care plans that formalize multidisciplinary team involvement [33]. ICPs outline steps for providers to ensure that patients are receiving prompt diagnosis and treatment for conditions that require multiple teams. They have been shown to improve service quality and efficiency, with particular efficacy in expanding the scope of professional roles to meet evidence-based guidelines [33]. Our initial proposed plan is based on findings from our case series, clinical experiences and literature review on SAN care (Figure 1). This ICP forms the basis of further inter-departmental discussion and, ideally, formalization into a quality improvement or implementation pilot study. We share this framework here for those who may be similarly interested in developing ICPs to improve multidisciplinary service quality for pain conditions.

Figure 1.

Preliminary elements for a post-neck dissection shoulder integrated care pathway.

ADL: Activities of daily living; EMG: Electromyography; MSK: Musculoskeletal; ROM: Range of motion.

Specific to our practice environment, a pilot project for improving the care of SAN palsies post-neck dissection would be comprised of early postoperative recognition with champion care coordinators, timely trials of standard conservative and minimally invasive therapy and prompt surgical evaluation. The final ICP would be created through an iterative approach of refining how best to capture patients early and which local resources are available for conservative therapies. Pilot studies for this ICP and others like it should examine whether they are successful in improving patient care for under-recognized and under-treated conditions, whether they are effective tools for tracking clinical outcomes over time, and whether they can be sustained with minimal oversight.

4. Conclusion

Shoulder dysfunction from SAN injury is common after neck dissection yet even advanced care centers continue to struggle with recognizing and treating SAN palsies in a timely and appropriate manner. Patients will often go on to develop chronic shoulder pain and irreversible functional losses due to untreated SAN palsies. Achieving earlier diagnosis and better multidisciplinary peripheral nerve care would greatly aid the growing population of cancer survivors and requires a systematic quality-based approach to change current practice habits. We discuss ICPs as one such approach and how they can serve as a framework for other pain conditions that are also caught in the double bind of under-recognition and under-treatment.

Author contributions

AL Ye made substantial contributions in drafting of the final manuscript. All authors made written contributions to the final manuscript. AL Ye, S Tummala, RY North and M Zein provided critical review of the manuscript. M Zein provided final approval of the version to be published. All authors have agreed to be accountable for all aspects of work related to the accuracy and integrity of the manuscript. No writing assistance, including the use of artificial intelligence, was utilized in the production of this manuscript.

Financial disclosure

The authors have no financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Competing interests disclosure

The authors have no competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Ethical conduct of research

This work was conducted with the approval of the MD Anderson Cancer Center Institutional Review Board (IRB) under approval number PA17-0222, with an IRB-approved waiver of consent for human subject data review protocols such as case series.