10 Salvage Neck Dissection: Indications, Workup, and Technical Considerations

Salvage neck dissections are characterized by a technically more challenging surgery, due to the presence of scarring and fibrosis from previous treatments ( Fig. 10.1). In general, the potential acute and long-term morbidity of a salvage neck dissection is higher than in primary procedures. It is also well documented that after radiation, neck dissection has an increased rate of postsurgical complications, in addition to increased subdermal fibrosis, neck and shoulder stiffness, pain, and dysphagia.¹ Additionally, disease control with salvage neck dissection is poorer than with primary neck dissection, as the disease tends to be biologically more aggressive and may be resistant to adjuvant treatment. Furthermore, fewer adjuvant treatment options may be available due to toxicity from previous therapies.² Finally, it is critically important to place salvage neck surgery in the overall context of the patient, and the overall disease burden. Patients with local disease recurrence in addition to regional disease recurrence often have poorer prognosis and are faced with more significant morbidity and radical changes to their quality of life in the setting of salvage surgery. In these patients, the prognosis and morbidity tend to be highly correlated with the site of recurrence. Generally, among patients presenting with synchronic local and regional recurrence, those with laryngeal or oral cavity primaries tend to fare better than those with oropharyngeal disease.³ Given these inherent challenges, it is crucial for clinicians to be familiar with the workup, decision-making process, and technical considerations associated with nodal disease that may necessitate a salvage neck dissection.

Fig. 10.1 Salvage modified radical neck dissection after chemoradiation therapy for squamous cell carcinoma of the oropharynx, with sacrifice of the internal jugular vein, and preservation of the sternocleidomastoid muscle and spinal accessory nerve. Vagus and hypoglossal nerves are preserved and visible. Dense fibrotic tissue in the neck encasing the cervical rootlets demonstrates the extent of treatment-related soft-tissue fibrosis and scarring.

There are several clinical scenarios in which a salvage neck dissection may be considered: (1) partial nodal remission following (chemo) radiation therapy; (2) complete nodal remission in a post (chemo) radiotherapy neck as part of a planned neck dissection; (3) disease recurrence in the neck following previous surgery (with or without radiation); (4) local disease recurrence in a previously treated N0 neck. These four potential indications for salvage surgery will be sequentially reviewed in subsequent subsections.

There is an inverse relationship between the likelihood of complete response to treatment in relation to the size and overall burden of nodal disease,⁴ with N1 nodal disease showing the best complete response to (chemo) radiotherapy, with relatively less complete response among patients with N2 and N3 disease.⁵ For patients with less than complete response in neck lymph nodes, salvage neck dissection should be performed assuming the disease is surgically resectable.^6,7 For patients who have had complete remission of disease in response to therapy, those with original N1 disease do not require a salvage neck dissection.^8,9 It is among the N2 and N3 group patients wherein controversy has existed over whether a salvage neck dissection should be performed, even with complete response in the neck to (chemo) radiotherapy. The evidence for and against planned neck dissection for patients with N2/N3 disease will be explored in the following three subsections.

10.2.2 Evidence Supporting Planned Neck Dissection for N2/N3 Disease Following (Chemo) radiotherapy

Several studies lend support to planned neck dissection following irradiation only for head and neck squamous cell carcinoma (SCC) patients with N2/N3 disease. Mendenhall et al looked at 161 patients (oral cavity, oropharynx, nasopharynx, hypopharynx, supraglottic larynx, unknown head and neck primary) and found that after radiotherapy, initial control of neck disease dropped from 92% for N1 disease to 65% for N2a disease.¹⁰ Dubray et al found similar poor results for N2/N3 necks after radiotherapy. In a series of 1,251 consecutive patients with node-positive oropharyngeal or pharyngolaryngeal SCC, the overall 3-year actuarial neck failure rates were 33% for N2 (n = 103) and 45% for N3 disease (n = 699).¹¹

Studies incorporating both, radiotherapy arms and concurrent CRT arms, show similar results. In Lavertu et al’s⁹ retrospective review of a 100-patient, phase III randomized clinical trial comparing definitive concurrent CRT with radiotherapy alone, 35 of 53 patients with N2/N3 disease had a neck dissection. None of the patients in this cohort who had a complete response in the neck, and who also underwent a posttreatment neck dissection, experienced regional recurrence. In contrast, 25% of the patients who had a clinically complete response but did not undergo posttreatment neck dissection had regional neck recurrence.

From a histopathologic perspective, several studies have shown that a significant proportion of patients with N2/N3 disease continue to have pathologic disease, despite achieving a clinically complete response post-CRT.^12,13 McHam et al,¹² for example, looked at a series of 109 patients with head and neck SCC (excluding nasopharynx, paranasal sinus, and salivary gland tumors) who had N2 and N3 disease. All patients underwent CRT and a subsequent planned neck dissection at least 6 weeks after treatment. The authors found that 28% of N2 disease patients, and 42% of N3 disease patients had pathologic evidence of disease in the neck dissection specimens.

Additionally, it is also argued that salvage surgery in an unplanned fashion, when disease recurs, is often not successful, with local control rates ranging from 22 to 35%.¹⁴ This is due to difficulty in achieving microscopically clear margins in a treated/fibrotic neck, coupled with limited adjuvant treatment options for microscopic disease. Finally, despite increased technical challenge, the immediate reported postoperative complication rates of planned neck dissection remain relatively low. Lavertu et al looked at a series of 100 patients treated in a phase 3 trial comparing radiotherapy alone with concurrent CRT for stage III and IV head and neck SCC. Twenty-nine planned neck dissections were performed for persistent neck disease or initial-stage N2 or greater. They found a major complication rate of 7% after neck dissection for radiation only, and 0% after neck dissection for chemoradiation.¹⁵

10.2.3 Evidence Refuting Planned Neck Dissection for N2/N3 Disease Following (Chemo) radiotherapy

Other studies have shown relatively low rates of neck failure after treatment with radiotherapy and CRT. Greven et al retrospectively reviewed 103 patients with stage III/IV, node-positive SCC (larynx, oropharynx, oral cavity, and hypopharynx) who had been treated with definitive radiotherapy or chemoradiation at a single institution with median follow-up of 42 months. CT scans were performed at a median of 4 weeks after treatment. The authors found that patients who had a radiographic complete response on posttreatment CT, and who underwent a neck dissection, had a nodal control rate of 94%, compared with 97% among those without neck dissection. On the other hand, patients with partial radiographic response who were treated with neck dissection had a nodal control rate of 94% compared with 73% among those without neck dissection. The authors concluded that patients who had a complete radiographic response on posttreatment imaging 4 to 6 weeks after radiation did not need a neck dissection.¹⁶

A series of 62 node-positive patients with oropharyngeal cancer from MD Anderson Cancer Center were treated with concomitant boost radiation and observed following complete response. The isolated neck failure rate was less than 5%, suggesting that observation is a reasonable approach in patients with nodal complete remission.¹⁷ Subsequently, the same group looked at 880 patients with T1–T4, N1–N3 M0 SCC of the oropharynx, larynx, or hypopharynx who received CRT or radiotherapy alone. Nodal complete response occurred in 377 (43%) patients, of whom 365 patients did not undergo nodal dissection. The 5-year actuarial regional control rate of patients with complete response was 92%. Two hundred sixty-eight (53%) of the remaining 503 patients without complete response underwent salvage neck dissections, and the 5-year actuarial regional control rate for patients without a complete response was 84%. Those who had a neck dissection fared better, with 5-year actuarial regional control rates of 90% for those operated, versus 76% for those not operated (p < 0.001). Following initial (chemo) radiotherapy, this large study thus supports observation for patients with complete nodal response, but salvage surgery for patients with partial nodal response.¹⁸

10.2.4 Planned Neck Dissection for N2/N3 Disease: Current Evidence-Based Recommendations

Drawing conclusions from studies that support and refute the decision for planned salvage neck dissection for N2/N3 disease can be challenging, due to significant heterogeneous and confounding factors inherent in individual studies. Many factors influence neck disease remission following CRT, including location of the primary tumor (e.g., the high response rate observed in nasopharyngeal and human papillomavirus–associated oropharyngeal cancer) and nodal disease burden.

In particular, for posttreatment neck dissections that demonstrate positive pathologic disease, the timing of the neck dissection may be crucial in interpreting results. Even among patients with complete radiologic response to therapy, as many as 30 to 40% of neck dissection pathologic specimens may contain cancer.^19,20 However, neck dissections in such studies were performed relatively early posttreatment (4–6 weeks). Surgical pathology from an early (4–6 week) posttreatment neck dissection may produce falsely positive results related to tumor cells that are in the process of dying, but yet appear histopathologically positive in the initial posttreatment period. The use of proliferation markers such as Ki-67 may be useful in differentiating active proliferating tumor cells versus nonviable tumor cells.²¹ When long-term clinical outcomes are analyzed, rather than analysis of early neck dissection pathology results, patients who obtain a complete clinical and radiological response to chemoradiation have a low (< 5%) risk of isolated neck recurrence.²²

It has been difficult to prove significant benefit of salvage planned neck surgery for post (chemo) radiotherapy patients with complete radiographic response in the neck, regardless of the initial N staging.^9,23,24 It is now generally agreed that it is reasonable to observe an N2/N3 neck after (chemo) radiotherapy if there is no evidence of disease on cross-sectional imaging at 6 weeks after treatment. A negative PET scan at 12 weeks (to be discussed in subsequent sections) makes this approach even more attractive. A routine neck observation approach following CRT is especially favored if patients are reliable to undergo appropriate follow-up imaging at regular intervals. Hence, patient factors and relevant resources of the health care system may also be taken into account. Surveillance post-CRT, rather than planned neck dissection, is now widely accepted as a recommended practice.^25,26 However, in select cases, a planned neck dissection may be appropriate in the setting of adverse risk factors, which predispose to poorer ultimate regional control. For example, one may consider a planned salvage neck dissection in the context of large nodal disease burden (especially N3 disease) in non-nasopharyngeal, nonoropharyngeal SCC.

10.2.5 Disease Recurrence in the Neck Following Previous Surgery (with or without Radiation)

Salvage neck dissection may also be performed after a patient has been treated with previous neck dissection (with or without adjuvant postoperative therapy after the neck dissection). Disease recurrence may result from missed lymph nodes in dissected neck levels, extracapsular nodal disease extension that recurs in the soft tissues of the neck, or recurrence in lymph nodes in areas not previously dissected. Aggressive tumor histology has been found to be associated with recurrent neck disease.²⁷ In general, patients with recurrent nodal disease following a neck dissection should be offered a salvage neck dissection provided the disease is grossly resectable.

In a retrospective review of 699 radical neck dissections, Jones et al found that 119 patients who had undergone a radical neck dissection developed a nodal recurrence; of these, 69 were considered candidates for salvage surgery. Factors that increased the risk of neck recurrence were neck node (N) status and absence of adjuvant radiotherapy. The 5-year survival for salvage neck dissection was 31%, with younger patients and low T and N classification having improved survival.²⁸

10.2.6 Local Disease Recurrence in an N0 Neck Previously Treated with (Chemo) radiation Therapy

Salvage neck dissection may be considered in the context of a patient with local disease recurrence, and an N0 neck after previous (chemo) radiotherapy. Salvage neck dissections in this scenario typically yield relatively low rates of neck nodal involvement (< 10%).²⁹ Radiation causes hyalinization and fibrosis of lymphatic structures, resulting in eventual narrowing of these structures,³⁰ which may explain the relative lower rates of subsequent nodal involvement when the primary tumor recurs.

Additionally, neck dissection performed in this setting seems to increase late adverse effects of radiation therapy (RT). In an analysis of three separate Radiation Therapy Oncology Group (RTOG) trials, severe late toxicity after CRT was quite common at 43%, and neck dissection after CRT was an independent risk factor for severe late toxicity with an odds ratio of 2.4.³¹

Dagan et al performed a retrospective review of patients treated with elective nodal radiation for T1–T4 N0 M0 SCC of the oropharynx, hypopharynx, or larynx who later developed an isolated local recurrence and remained N0. Fifty-seven patients were salvaged, 40 with neck dissection and 17 with neck observation. Four of 46 (9%) heminecks were found to have occult metastases in the surgical specimen. The 5-year local–regional control rate was 75% for all patients. Neck dissection resulted in poorer outcomes compared with observation. In the dissected group, the 5-year local control, regional control, cause-specific survival, and overall survival rates were 71, 87, 60, and 45%, respectively, compared to 82, 94, 92, and 56%, respectively, for the observation group. Toxicity was more likely with neck dissection. The authors concluded that in the setting of previous elective nodal irradiation, routine elective neck dissection should not be included during salvage surgery for locally recurrent head and neck SCC. The risk of occult neck disease is low, outcomes do not improve, and the likelihood of adverse effects increases.²⁹ For the case of elective salvage neck dissection associated with larynx cancer, the rate of occult nodal metastases has been shown to be about 5%, and there has not been demonstrated a regional control benefit to elective salvage neck dissection in this setting.³²

In summary, elective salvage neck dissection in the locally recurrent primary with an already treated neck is not typically recommended. However, it should be recognized that many patients with local recurrence for head and neck SCC require free flap reconstruction of the primary site. Since in these cases the neck must be explored for flap recipient blood vessels, it is reasonable to resect any nodal basin (s) in the area of the respective exploration.

10.3 Workup

10.3.1 Confirmation of Disease

Some authors make an arbitrary distinction between residual and recurrent diseases, wherein residual disease is disease that remains after treatment (i.e., no disease-free interval) and recurrent disease is disease that becomes apparent after at least 6 months of complete regression. Regional, residual, or recurrent disease can be detected on clinical or radiologic followup. Depending on the philosophy and resources of the treating institution, follow-up imaging may be performed at various intervals in the first 5 years after treatment, and may take the form of ultrasound, CT, MRI, or PET scan.

CT scan with contrast is a quick and efficient means of obtaining a radiologic evaluation of the neck nodal status. The performance of CT scan in the post-CRT setting has been reported as 57 to 85% sensitivity, 24 to 79% specificity, 22 to 59% positive predictive value, and 73 to 94% negative predictive value.^20,33,34 In some institutions, MRI is the preferred modality of imaging for surveillance. MRI, with the use of diffusion weighted imaging (DWI) protocols based on tissue cellularity, may aid in discriminating between metastatic and benign neck adenopathy, with metastatic adenopathy demonstrating significantly higher apparent diffusion coefficient.³⁵

Additionally, some have suggested that MRI may be better in distinguishing posttreatment changes versus recurrence in the primary tumor bed, based on T2 signal characteristics.³⁶ Hence, the neck is followed with the same imaging modality as the primary in this setting. However, other studies suggest that the superiority of MRI over CT scan in the follow-up of the primary bed is exclusively for nasopharyngeal cancers, rather than laryngeal and pharyngeal neoplasms.³⁷ A large meta-analysis by de Bondt et al comparing CT and MRI in the detection of lymph node metastases pretreatment in head and neck cancer showed that MRI and CT scan had similar accuracy.³⁸ In many centers, CT is employed as a relatively cheap, quick, and readily available means of imaging surveillance.

While ultrasound-guided FNA biopsy is not first line in the routine surveillance of neck disease, it can be useful to confirm the presence of disease suspected on imaging on CT/MRI before committing to salvage surgery. In the de Bondt et al meta-analysis, the authors concluded that ultrasound-guided FNA biopsy was the most accurate imaging modality to detect cervical lymph node metastases. Ultrasound-guided FNA biopsy showed the highest diagnostic odds ratio (DOR) of 260, compared to ultrasound without FNA (DOR = 40), CT (DOR = 14), and MRI (DOR = 7).³⁸ Nonetheless, it must be stated that ultrasound-guided FNA biopsy is only useful with a positive result. In situations where cytological proof cannot be obtained from an ultrasound-guided FNA biopsy, intraoperative frozen section can be performed to confirm presence of disease before proceeding with surgical resection. In most cases, however, intraoperative frozen section is not feasible due to potential oncologic compromise in obtaining the samples (vs. favored en bloc specimen removal). As such, preoperative counseling and disclosure regarding the possibility of a pathologically negative specimen is a recommended practice.

PET-CT scan has a very high negative predictive value in the posttreatment neck. If the PET-CT is negative, it is very unlike that there is viable tumor present in the lymph nodes, even if there are persistent lymphadenopathies on other imaging modalities (CT, MRI, ultrasound). Its negative predictive value of 93 to 95% for the primary site, and of 94 to 100% for the neck, makes this imaging modality an attractive choice for the initial posttreatment follow-up regimen.³⁹ PET scans have the additional advantage of simultaneous assessment of distant disease; although these tests are resource intensive, patients with negative PET scan at 12 weeks posttreatment have a very low risk of subsequent failure.^40,41

A large, multicentric randomized phase 3 trial involving 37 head and neck cancer centers in the United Kingdom recruited 564 patients to study the utility of PET in the evaluation of planned neck dissection post-CRT. The patients were randomized to planned neck dissection before or after CRT (control), or CRT followed by fludeoxyglucose PET-CT 10 to 12 weeks post-CRT with neck dissection only if PET-CT showed incomplete or equivocal response of nodal disease (intervention). There were 54 neck dissections performed in the surveillance arm, with 22 surgical complications, and 221 neck dissections in the neck dissection arm, with 85 complications. Quality-of-life scores were slightly better in the surveillance arm. The authors concluded that PET-CT-guided active surveillance after CRT showed similar survival outcomes compared to planned neck dissection but resulted in considerably fewer neck dissections, fewer complications, and lower cost, supporting the routine use of PET imaging in the posttreatment setting.⁴² Although PET is useful in surveillance of the primary and neck after CRT, the resources of the treating institution and country must be taken into account when considering this as an standard practice. A resource-balanced approach that is employed in many centers in the United States is a single PET scan at 12 weeks posttreatment to rule out residual active tumor in nodal remnants, with further follow-up imaging (if necessary) limited to more cost-efficient modalities such as CT.

10.3.2 Extent of Disease and Operative Planning

The timing of posttreatment imaging is critically important. False-positive results commonly occur if imaging is performed too early, as nodal remnants and hypermetabolic activity often continue to resolve over time.⁴³ However, delaying posttreatment imaging too long may result in loss of opportunity to intervene in the optimal surgical window following (chemo) radiation therapy. This window represents a short time frame, when fibrosis is less marked and surgery is technically easier to perform, or between resolution of the acute CRT injury and the onset of chronic CRT injury. Another obvious risk in delaying too long is the risk for disease progression. It is, therefore, recommended that CT or MRI be performed approximately 8 to12 weeks after completion of CRT.²²

For PET and PET-CT, the timing posttreatment is equally important. As treatment-related inflammation abates, false-positive results on posttreatment PET scan decrease proportionally (increased specificity and negative predictive as time from treatment completion increases).⁴⁰ The current recommendation for optimal timing of PET is approximately 12 weeks posttreatment.^22,40

10.4 Technical Considerations

10.4.1 Which Neck Levels Should Be Included in Salvage Dissection?

The appropriate neck level to address in salvage dissection depends on the primary site, extent of original nodal disease, and current nodal status post (chemo) radiotherapy. Historically, neck dissection performed for presumed residual disease after (chemo) radiotherapy was performed as a comprehensive radical neck dissection to encompass all levels of disease ( Fig. 10.2a). However, in recent decades, modified radical and selective neck dissections have been demonstrated to offer equivalent disease control and decreased morbidity, unless there is invasion of the sternocleidomastoid muscle, internal jugular vein, and/or spinal accessory nerve.

Neck dissection after chemoradiation for advanced head and neck cancer is not without risks, especially related to wound healing. Davidson et al showed that a preoperative radiotherapy dose greater than 70 Gy was associated with complications in 58% of the patients, versus 29% if the dose was less than 70 Gy (p = 0.09). This trend was reflected primarily in wound complications, and reached significance for skin flap necrosis.⁴⁴ Others have shown initial increased incidence of spinal accessory nerve dysfunction, and a negative impact on quality of life, when all five neck levels were dissected.⁴⁵ In general, more extensive neck dissections with associated removal of fibrofatty tissue result in greater long-term fibrosis with subsequent deleterious effects on swallowing and neck mobility.⁴⁶

Selective neck dissection is generally preferred to comprehensive neck dissection in the salvage setting. A selective neck dissection is one that involves the removal of at least three contiguous neck levels ( Fig. 10.3), whereas a superselective neck dissection is a neck dissection where only two contiguous levels are addressed ( Fig. 10.4). The concept of selective neck dissection is based on predictable patterns of lymphatic drainage, and lymphatic tumor spread in association with upper aerodigestive tract cancers.^47,48 However, this philosophy is based on the understanding of lymphatic drainage and lymphatic tumor spread in the untreated neck. In the situation of previous CRT, drainage patterns may potentially change. Nonetheless, several studies have shown that a selective or even superselective neck dissection is generally the preferred salvage option to both limit morbidity and ensure appropriate oncologic disease control.^49,50

A study examining 28 postradiotherapy neck dissections for SCC found that only 1 of 28 neck dissections (23 of which were comprehensive) revealed cancer outside nodal levels II, III, and IV. Primary sites in this study were oropharynx (n = 19), nasopharynx (n = 3), and larynx/hypopharynx (n = 3). The one case with nodal disease outside levels II to IV was a patient with a 10-cm neck mass.⁵¹ Another study by Stenson et al¹³ suggests that level V need not be addressed unless there is gross disease in this level prior to RT.

Robbins et al examined a series of 106 neck dissections performed after concurrent intra-arterial chemoradiation. The extent of neck dissection was determined by extent of disease, as elucidated by posttreatment imaging. Three subgroups of patients were delineated by the extent of neck dissection: radical or modified radical neck dissection, selective neck dissection, and superselective neck dissection. With a median follow-up of 58 months, regional failure occurred in 11 (5%) of 240 patients: 2/12 (17%) in the modified radical neck dissection group, 3/65 (5%) in the selective neck dissection group, 0/7 in the superselective neck dissection group, and 6/156 (4%) in the no neck dissection group. The rates of overall survival and distant metastases were not significantly different among the three neck dissection subsets. Notwithstanding the selection bias among treatment groups, the authors concluded that selective and superselective neck dissections are viable therapeutic alternatives for patients with residual disease confined to one neck level after intra-arterial chemoradiation, and possibly other chemoradiation protocols.⁵²

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Extracapsular Spread: Biology, Imaging, Surgical Management, Pathological Diagnosis, and Therapeutic Implications
2. Clinical Assessment of Neck Lymphadenopathies
3. Sentinel Node Dissection
4. Robotic Neck Dissection: The Retroauricular Approach

Stay updated, free articles. Join our Telegram channel

Join