- Email: [email protected]
Selective neck dissection and the management of the node-positive neck
Selective
Neck Dissection and the Management of the Nodecpositive Neck
Sean J. Traynor,
MD, James I. Cohen, MD, PhD, Jason Gray, MD, Peter E. Andersen, Edwin C. Ever&, MD, fort/and, Oregon
OBJECTIVE: To assess the oncologic effectiveness of the selective neck dissection (SND) in patients with both clinically and pathologically proven regional metastases. METHODS: A 4-year retrospective medical chart review was conducted in an academic tertiary care referral center. Twenty-nine patients with a newly diagnosed upper aerodigestive tract squamous cell carcinoma, and both clinically and histologically proven cervical metastases who underwent 36 SND, had their records reviewed. Minimum follow-up was 2 years. RESULTS: Regional metastasis were staged Nl in 13 patients, N2A in 1, N2B in 8, and N2C in 7. Seventeen supraomohyoid and 19 lateral neck dissections were performed. Extracapsular spread of tumor was present in 11 patients. Postoperative radiation therapy was administered to 20 patients. Actuarial disease-specific survival at 4 years was 47% overall, 67% in Nl patients, and 41% in N2 patients. Only 1 failure in the treated neck occurred for a 4-year actuarial regional failure rate of 4%. The actuarial local failure and distant metastasis rate were 36% each. CONCLUSIONS: In carefully selected patients with clinically and histologically apparent regional metastases, the selective neck dissection can be an oncologically effective procedure. 01996 by Excerpta Medica, Inc. Am J Sot-g. 1996;172:654-657.
A
lthough cure remains the ultimate goal in the management of the patient with head and neck cancer, an increasing emphasis on improved postoperative function and cosmesis is evident. The evolution of the neck dissection is representative of this trend. First described by Crile in 1906, the radical neck dissection (RND) has served as the primary method of managing cervical metastases in patients with malignancies of the head and neck for the majority of this century.’ The goal of this operation is the complete en bloc removal of the cervical lymphatic components involved or at risk for metastatic disease, and includes the removal of the sternocleidomastoid muscle From the Department of Otolaryngology/Head and Neck Surgery, The Oregon Health Sciences University, Portland, Oregon. Requests for reprints should be addressed to James I. Cohen, MD, PhD, Department of Otolaryngology/Head and Neck Surgery, PV-01, The Oregon Health Sciences University, 3181 SW Sam Jackson Park Road, Portland, Oregon 97201. Presented at the Fourth International Conference on Head and Neck Cancer, Toronto, Ontario, Canada, July 28-August 1, 1996.
T 664
0 1996 by Excerpta All rights reserved.
Medica,
Inc.
MD,
(SCM), internal jugular vein (IJ), and the 1 lth cranial nerve. The identification of histopathologic evidence of factors known to increase the risk of regional recurrence, such as extracapsular spread (ECS) and multiple levels of involvement, may guide the use of radiation therapy.2-4 With time surgeons have challenged the necessity of such extensive measures to accomplish these goals and explored the feasibility of modifying the RND. The rationale for such modification stems from the significant cosmetic and functional deficits associated with the RND, and the realization that neck recurrence is still a significant problem despite its extensiveness. An improved understanding of lymph node drainage patterns,5,6 the identification of fascial compartments separating the lymphatic components of the neck from structures commonly sacrificed in the RND,7,X as well as an improved understanding of the benefits of adjuvant radiation therapy,2.4 have further fostered a movement away from the routine use of the RND. Selective neck dissection (SND), the selective removal of only those nodes at greatest risk for metastatic disease, represents the greatest divergence from the radical neck dissection routinely used today. Application of this technique in the clinically negative neck is accepted by many due to demonstrated rates of regional control and survival commensurate with the more extensive neck dissections.4*Ym” The use of the SND in the clinically positive neck, however, remains controversial at this writing. Extension of the indications of this operation to include the clinically positive neck seem logical, however, because, in the absence of factors that violate the fascial compartments of the neck or disrupt lymphatic flow such as massive adenopathy or gross extracapsular spread, the rationale behind this procedure remains viable. In addition, this approach can be extrapolated from the finding that in the elective selective neck dissection, histopathologic evidence of lymphatic spread has been identified in up to one-third of cases.‘? As a result, it seems the SND is, in fact, currently being used routinely in necks already involved with metastatic disease. Over the last decade at the Oregon Health Sciences University we have carefully expanded our use of the SND to include some patients exhibiting both clinical and histological evidence of metastatic neck disease. It is the intention of this paper to review our experience in this group of patients to help answer the question of whether the SND can provide the therapeutic measures and histopathologic information necessary to treat this patient group as effectively as the more comprehensive neck dissection.
MATERIALS
AND METHODS
The medical records of all patients who underwent neck dissection at the Oregon Health Sciences University be0002-961 O/96/$1 PII SOOO2-9610(96)00296-S
5.00
1 SELECTIVE
NECK
DISSECTION/TFtAYNOFl
ET AL
I
TABLE
I
TABLE Primaly
Primary
Site
Staging Staging
Nodal Patients
11 T2 T3 T4 Total
TABLE
Ill
Characteristics Nodal
1 4 16
Staging
Characteristics
Staging
Patients
Ni N2A N2B N2C N3 Total
a 29
11 1
a 9 0 29
II Primary Primary
Site
(Total)
Oral Cavity (9) Tongue Floor of Mouth Orophatynx (6) Base of Tongue Tonsil Larynx (a) Supraglottis Intrinsic Larynx Wpwhavnx Pyriform Total
(4) Sinus
Site
TABLE
Locations
IV Survival
Patients 4-year Nl N2
6 3 3 5 7 1 4 29
tween October 1, 1988 and October 1, 1992 were reviewed. Only those patients undergoing a SND with both clinically and pathologically proven regional metastasis were included. Patients with a malignancy other than squamous cell carcinoma, a history of a prior head and neck malignancy, previous treatment to the head and neck region with chemotherapy, radiation, or surgery, as well as those with follow-up of less than 2 years were excluded. Staging was done according to the criteria established by the American Joint Cancer Committee.” The type of SND performed depended on the site of the primary. For primaries of the oral cavity and oropharynx, a supraomohyoid neck dissection, involving the removal of nodal groups I-III, was performed. For primaries of the larynx or hypopharynx, a lateral neck dissection, involving the removal of nodal groups II-IV, was performed. In instances where the most inferior level of the dissection was clinically positive, the dissection was extended one level beyond this. Due to established patterns of lymphatic flow, extension of the dissection to include level V in instances of positive nodal disease at level IV was not done.6 Careful patient selection criteria were applied in all instances. To be eligible for SND the patient could not have had any previous treatment to the neck for this malignancy. In addition only those patients with mobile nodes, ie, without evidence of fixation to the surrounding structures such as the sternocleidomastoid muscle, were eligible. In general this restricted this approach to patients with nodes that were less than 3 cm in size, ie, Nl or N2B and N2C where no individual node was over 3 cm in size. Each neck was approached with the intention of performing a SND. If, however, at the time of operation, other structures, principally the internal jugular vein, were believed to be too close to the nodes to allow for safe and adequate removal, it was sacrificed. Similarly, in the instance of minor SCM THE
AMERICAN
actuarial
survival
rate
47% 67% 41%
involvement a cuff of muscle adjacent to the nodes was removed. All neck dissection specimens were reviewed by a staff pathologist. Any patients in whom evidence of extracapsular spread or multiple levels of involvement were identified received postsurgical radiation. Actuarial disease-specific survival and rates of local, regional, and distant recurrence were determined using the Kaplan-Meier method. Site of recurrence was considered to be at the most proximal site.
RESULTS Two hundred and seven patients underwent neck dissection at our institution over the 4 years included in this study. Of this group, 117 patients undergoing 137 neck dissections met the inclusion criteria of a newly diagnosed squamous cell carcinoma with a minimum follow-up of 2 years. Further limiting the group to include only those patients who underwent a SND in which both clinical and pathologic neck disease was evident limited the group to 36 patients. Elimination of 7 patients due to follow-up less than 2 years resulted in a treatment group of 29 patients undergoing 36 SNDs. Twenty-three males and 6 females were included with an average age of 59 years (range, 35-77 years). Median follow-up time was 3 years. Primary site characteristics are listed in Tables I and II, and the majority were staged T3 or greater. Histologic nodal staging is reported in Table III. The majority of these patients were staged N2, and no patients with N3 disease were included. Additionally, none of the patients had clinical evidence of nodal disease at level V. Twenty-two patients underwent unilateral SND: 13 supraomohyoid and 9 lateral neck dissections. Seven patients underwent bilateral selective neck dissections: 2 bilateral supraomohyoid and 5 bilateral lateral neck dissections. Based on the presence of extracapsular spread or multiple levels of nodal disease, 20 patients received postoperative radiation to the neck. Actuarial disease-specific survival at 4 years was 47% overall; 67% in the Nl patients and 41% in the N2 patients (Table IV). The 4-year actuarial regional failure rate was 4%, with neck recurrence in only 1 instance. This patient JOURNAL
OF
SURGERY”
VOLUME
172
DECEMBER
1996
655
( SELECTIVE
NECK
DISSECTION/TRAYNOR
ET AL 1 1
had a T3NZb well-differentiated squamous cell carcinoma and underwent composite resection and ipsilateral supraomohyoid neck dissection. Due to the presence of nodal disease at levels I and II and the presence of extracapsular spread she underwent postoperative radiation therapy. Recurrence in the dissected neck at level III, which was included in the original neck dissection, was noted at 5 months postoperatively. The 4-year actuarial local recurrence and distant failure rate were 36% each, with local recurrence occuring in 6 patients and distant metastases developing in another 8 (Table V).
DISCUSSION This investigation is unique in that it limited its focus to include only those patients with both clinical and histologic evidence of cervical metastases who were treated with a selective neck dissection. Traditionally, the surgical management of these patients has been the RND involving the en bloc removal of the cervical lymphatics along with the stemocleidomastoid muscle, the internal jugular vein, and the spinal accesory nerve.’ If the RND gauranteed regional control of disease and increased survival then, perhaps, the undeniable morbidity associated with its use would be acceptable. This, however, is not the case as demonstrated by the regional recurrence rates of 54.3% for node-positive patients and 71.3% with nodes positive at multiple levels reported by Strong in his review of the Memorial Hospital experience with the RND.4 Prompted by such imperfect regional control of disease, as well as a growing emphasis on improving posttreatment function and cosmesis, head and neck oncologic surgeons started to question the routine use of the RND and began to modify it. Modifications of the RND have not been made in a haphazard manner. Rather, they seem to be based on firm anatomic and oncologic principles not appreciated when the RND was first described nearly 100 years ago, An early impetus for modification was the recognition that the operation was, in fact, not a complete en bloc resection since other seemingly related structures such as the hypoglossal and vagus nerve and carotid artery were not removed.14 Additionally, gross and histological analysis of surgical and necropsy specimens demonstrated the cervical lymphatics are contained within distinct fascial compartments without intramuscular or intra-adventitial lymph nodes.’ Other anatomic studies have confirmed the lymphatic system of the neck lies within a system of aponeurotic compartments that, on the outer side, envelop muscles and vessels of the neck but, other than proximity, have no relationship to these structures.’ Bocca applied these principles in advocating the “functional” or complete neck dissection in which the lymphatics of all five nodal levels of the neck were removed by stripping the fascial cover from muscles and vessels and dissecting free major nerves in an effort to spare the structures commonly sacrificed during the RND. Regional recurrence rates of 2.4% for NO and 30% for Nl/N2 necks compared favorably to rates of 6.6% and 48% in similar patients in which members of the same treatment team performed RNDs.” A predictable pattern of cervical lymphatic flow based on the primary site is, by now, well described.5,6,‘6 In an early study conducted by Fisch employing lymphography, lym656
THE AMERICAN
JOURNAL
OF
SURGERY@
VOLUME
172
TABLE
V Recurrence 4 Year Local Regional Distant
Actuarial
Recurrence
Rates 36% 4% 36%
phatic flow from the upper posterior cervical (“junctional”) nodes into both the spinal and jugular nodal systems was demonstrated.5 Indirect flow from the spinal nodes to the jugular nodal system was evident, but filling of the spinal nodes from the jugular chain was never observed. This definite unidirectional flow was attributed to the presence of lymphatic valves, and its importance in predicting potential sites of cervical metastatic spread is apparent. The absence of retrograde flow from the jugular nodes into the spinal chain may explain why metastatic spread to the posterior triangle (region V) is uncommon. A review of the records of 1,277 patients undergoing RND identified positive nodes at region V in only 3% of instances, with a slight increase to only 5% when clinically positive nodes were evident.” Additional investigators have confirmed this low prevalence of posterior triangle involvement.‘* Studies conducted by Shah and associates have lended further credence to the existence of a predictable cervical drainage pattern, even in the presence of metastatic disease.‘j Specimens from more than 1,000 comprehensive neck dissections performed on individuals with squamous cell carcinoma of the head and neck were evaluated for the presence and site of metastatic disease. A predominance of certain regions was seen for each primary site. The vast majority of oral cavity lesion metastases occurred at levels I, II, and III, which comprise those nodal groups removed in the supraomohyoid neck dissection. Oropharyngeal, hypopharyngeal, and laryngeal lesions demonstrated a predilection for metastatic disease to levels II, III, and IV, which would be included in a lateral neck dissection. Metastatic nodal disease outside of the predominant regions was shown to be unusual in the absence of involvement of the expected regions. Thus, a compartmentalization of the cervical lymphatics along with a predictable pattern of lymphatic flow in which the posterior triangle is rarely involved have been demonstrated. These observations do not support the routine removal of all cervical regions and the sacrifice of the SCM, IJ, and spinal accessory nerve dictated by the radical or complete neck dissections, and suggest that the selective removal of only those nodes at greatest risk for metastatic disease might be sufficient to adequately treat and stage the at-risk neck. This type of neck dissection, termed the SND, represents the next logical step in the modification of the RND. Several investigators have demonstrated that the use of the SND in the clinically negative neck provides treatment results and staging information comparable to the more extensive lymphadenectomies. As a result, its use in these instances has become common at many institutions.‘-” The use of the SND in the clinically positive neck, however, has not yet been widely accepted. There is, however, evidence to support this approach in certain instances. DeDECEMBER
1996
) SELECTIVE
spite the finding of neck disease in the pathologic review of an elective neck dissection specimen control rates comparable to those seen with more extensive operations are still seen. Additionally, a review of the literature reveals the performance of a therapeutic SND in patients with both clinical and histologic evidence of metastatic disease is not a new concept. Several investigators have described its use in these instances, but a study confined soley to this group of patients seems to be lacking. Medina reported his experience in managing patients with upper aerodigestive tract squamous cell carcinoma using the supraomohyoid neck dissection. Forty-four patients had both clinical and pathologic evidence of metastatic disease, with the majority staged N 1. Regional recurrence rates of 10% for N 1 disease without extracapsular spread and 24% when multiple levels or ECS was found were described. Neck recurrence rates in the latter group decreased to 15% when postoperative radiation therapy was added, and no patients in the former group received radiation. ” Byers reported the same recurrence rates in a similar group of patients undergoing therapeutic supraomohyoid neck dissection, both with and without adjunctive radiation therapy, in which 70% had extracapsular spread or multiple levels of involvement.‘” Based on their favorable results these authors suggested the SND has a meaningful role in certain patients with Nl metastatic disease to the neck, and that the addition of adjunctive radiation therapy may improve regional control If ECS or multiple levels of involvement are identified. With this in mind it would seem reasonable to extend the indications of this operation to include not only Nl disease, but also N2B and N2C disease, where the individual nodes are nonfixed Nl nodes, because the rationale behind this procedure remains viable in the absence of gross extracapsular spread or nodal fixation. The results of our study seem to support such an approach. Importantly, it must be emphasized we do not advocate the use of the SND in all cases. Specifically, patients in which a SND is applied included only those without evidence of massive adenopathy, nodal fixation, and gross ECS or a history of prior neck surgery. More massive adenopathy with gross extracapsular spread may violate the fascial compartments of the neck and disrupt the predictable pattern of lymphatic flow. The presence of only 1 patient with N2A disease and the absence of any patients with N3 disease is demonstrative of our willingness to return to more traditional neck dissections when indicated to completely remove gross disease. Additionally, when a suspicious node is encountered at the lowest level of the dissection we routinely extend the operation to include the next echelon. This principle does not apply to level V because, due to the established pattern of lymphatic flow, this level is at limited risk from level IV disease.5x”,‘” In no instance is gross disease left behind, and the internal jugular vein, 11th cranial nerve, and sternocleidomastoid muscle must be removed if proximity to disease precludes safely sparing these structures.
THE AMERICAN
NECK
DISSECTION/TRAYNOR
ET AL 1
CONCLUSION Our experience with 29 patients who had both clinical and histopathological evidence of cervical metastases staged Nl to NZC, and underwent SND, was reviewed. Treatment outcomes including regional control rate and survival are comparable with those of more extensive neck dissections in similar patients. The SND is based on firm anatomlc and oncologic principles rhat have become apparent since the RND was first described nearly 100 years ago. The results of this study suggest the careful extension of its mdicatlons to include some patients with both clinical and histological neck disease is reasonable.
REFERENCES 1. Crile GW. Excision of cancer of the head and neck. JAMA. 1906;47:1780-1786. 2. O’Brien CJ, Stmth JW, Soong SJ, et al. Neck dissection with and without radiotherapy: prognostic factors, patterns of recurrence, and survival. Am J Surg. 1986;152:456-463. 3. Johnson JT, Myers EN, Bedetti CD, ct al. Cervical lymph node metastases: incidence and implications of cxtracapsular carcinoma. Arch Otolaryngol. 1985;111:534+537. 4. Strong EW. Preoperative radiation and radical neck dissection. Surg Clin North Am. 1969;19:271-276. 5. Fisch UP, Siegel ME. Cervical lymphatic system as viewed by lymphography. Ann Otol Rhinol Lqngoi 1964;73:869-882. 6. Shah JP. Patterns of cervical lymph node metastasis from squamous cell carcinomas of the upper aerodlgestive tract. Am 1 Surg. 1990;160:405-409. 7. Suarez 0. El Problema de las metastabia linfatlca y alejadas de1 cancer de laringe e hipofaringe. Reu Otorhinolanngol. 1963;23:83. 8. Bocca E, I’ignatoro 0. A conservation technique in radical neck dissection. Ann Otol Rhinol Luryngol. 1976;76:975-987. 9. Byers RM, Wolf PF, Ballantyne AJ. Rationale for elective modified neck dissection. Head Neck Surg. 19&X:160-167. 10. O’Brien CJ. A selective approach to neck dissection for mucosal squamous cell carcinoma. Aust N 2 J Surg 1994;64:236-241. 11. Medina JE, Byers RM. Supraomohy<,id neck dlssection: ratianale, indications and surgical technique. ffccid Neck. 1989:I 11-122. 12. Alvi A, Johnson JT. Extracapsular qwad in the clinically negative neck (NO): implications and outcomes. Otolaqngol Head and Neck Surg. 1996;114:65-70. 13. American Jomt Committee on Cancer. Manual for Staging of Cancer, 4th ed. Philedelphia: J.B. Lippincott CO., 1992:33-35. 14. DeSanto LW, Beahrs OH. Modified Jnd complete neck dissection in the treatment of squamous cell carcinoma of the head and neck. Surg Gynec und Obstet. 1988;167:259-269. 15. Bocca E, Pignataro 0, Oldini C, (:qxl C. Functional neck dissection: an evaluation and review of 843 cases. Laryngosc0@. 1984;94:942-945. 16. Lindberg R. Distribution of cervical lymph node metastases from squamous cell carcinoma of the upper respiratory and digestive tracts. Cancer. 1972;29:1446-1451. 17. Davidson BJ. Kulkarny V, Delacure MD, Shah JP. Posterior triangle metastases of squamous cell carcinom of the upper aerodigestive tract. Am] Surg. 1993;166:395-398. 18. Skolnik EM, Yee KF, Friedman M, Golden TA. The posterior triangle in radical neck surgery. Arch Otokqngol. 1976;102:1-4. 19. Byers, RM. Modified neck dissection: a study of 967 cases from 1970 to 1980. Am J Surg. 1985;150:414~421.
JOURNAL
OF
SURGERY”
VOLUME
172
DECEMBER
1996
657
Neck Dissection and the Management of the Nodecpositive Neck
Sean J. Traynor,
MD, James I. Cohen, MD, PhD, Jason Gray, MD, Peter E. Andersen, Edwin C. Ever&, MD, fort/and, Oregon
OBJECTIVE: To assess the oncologic effectiveness of the selective neck dissection (SND) in patients with both clinically and pathologically proven regional metastases. METHODS: A 4-year retrospective medical chart review was conducted in an academic tertiary care referral center. Twenty-nine patients with a newly diagnosed upper aerodigestive tract squamous cell carcinoma, and both clinically and histologically proven cervical metastases who underwent 36 SND, had their records reviewed. Minimum follow-up was 2 years. RESULTS: Regional metastasis were staged Nl in 13 patients, N2A in 1, N2B in 8, and N2C in 7. Seventeen supraomohyoid and 19 lateral neck dissections were performed. Extracapsular spread of tumor was present in 11 patients. Postoperative radiation therapy was administered to 20 patients. Actuarial disease-specific survival at 4 years was 47% overall, 67% in Nl patients, and 41% in N2 patients. Only 1 failure in the treated neck occurred for a 4-year actuarial regional failure rate of 4%. The actuarial local failure and distant metastasis rate were 36% each. CONCLUSIONS: In carefully selected patients with clinically and histologically apparent regional metastases, the selective neck dissection can be an oncologically effective procedure. 01996 by Excerpta Medica, Inc. Am J Sot-g. 1996;172:654-657.
A
lthough cure remains the ultimate goal in the management of the patient with head and neck cancer, an increasing emphasis on improved postoperative function and cosmesis is evident. The evolution of the neck dissection is representative of this trend. First described by Crile in 1906, the radical neck dissection (RND) has served as the primary method of managing cervical metastases in patients with malignancies of the head and neck for the majority of this century.’ The goal of this operation is the complete en bloc removal of the cervical lymphatic components involved or at risk for metastatic disease, and includes the removal of the sternocleidomastoid muscle From the Department of Otolaryngology/Head and Neck Surgery, The Oregon Health Sciences University, Portland, Oregon. Requests for reprints should be addressed to James I. Cohen, MD, PhD, Department of Otolaryngology/Head and Neck Surgery, PV-01, The Oregon Health Sciences University, 3181 SW Sam Jackson Park Road, Portland, Oregon 97201. Presented at the Fourth International Conference on Head and Neck Cancer, Toronto, Ontario, Canada, July 28-August 1, 1996.
T 664
0 1996 by Excerpta All rights reserved.
Medica,
Inc.
MD,
(SCM), internal jugular vein (IJ), and the 1 lth cranial nerve. The identification of histopathologic evidence of factors known to increase the risk of regional recurrence, such as extracapsular spread (ECS) and multiple levels of involvement, may guide the use of radiation therapy.2-4 With time surgeons have challenged the necessity of such extensive measures to accomplish these goals and explored the feasibility of modifying the RND. The rationale for such modification stems from the significant cosmetic and functional deficits associated with the RND, and the realization that neck recurrence is still a significant problem despite its extensiveness. An improved understanding of lymph node drainage patterns,5,6 the identification of fascial compartments separating the lymphatic components of the neck from structures commonly sacrificed in the RND,7,X as well as an improved understanding of the benefits of adjuvant radiation therapy,2.4 have further fostered a movement away from the routine use of the RND. Selective neck dissection (SND), the selective removal of only those nodes at greatest risk for metastatic disease, represents the greatest divergence from the radical neck dissection routinely used today. Application of this technique in the clinically negative neck is accepted by many due to demonstrated rates of regional control and survival commensurate with the more extensive neck dissections.4*Ym” The use of the SND in the clinically positive neck, however, remains controversial at this writing. Extension of the indications of this operation to include the clinically positive neck seem logical, however, because, in the absence of factors that violate the fascial compartments of the neck or disrupt lymphatic flow such as massive adenopathy or gross extracapsular spread, the rationale behind this procedure remains viable. In addition, this approach can be extrapolated from the finding that in the elective selective neck dissection, histopathologic evidence of lymphatic spread has been identified in up to one-third of cases.‘? As a result, it seems the SND is, in fact, currently being used routinely in necks already involved with metastatic disease. Over the last decade at the Oregon Health Sciences University we have carefully expanded our use of the SND to include some patients exhibiting both clinical and histological evidence of metastatic neck disease. It is the intention of this paper to review our experience in this group of patients to help answer the question of whether the SND can provide the therapeutic measures and histopathologic information necessary to treat this patient group as effectively as the more comprehensive neck dissection.
MATERIALS
AND METHODS
The medical records of all patients who underwent neck dissection at the Oregon Health Sciences University be0002-961 O/96/$1 PII SOOO2-9610(96)00296-S
5.00
1 SELECTIVE
NECK
DISSECTION/TFtAYNOFl
ET AL
I
TABLE
I
TABLE Primaly
Primary
Site
Staging Staging
Nodal Patients
11 T2 T3 T4 Total
TABLE
Ill
Characteristics Nodal
1 4 16
Staging
Characteristics
Staging
Patients
Ni N2A N2B N2C N3 Total
a 29
11 1
a 9 0 29
II Primary Primary
Site
(Total)
Oral Cavity (9) Tongue Floor of Mouth Orophatynx (6) Base of Tongue Tonsil Larynx (a) Supraglottis Intrinsic Larynx Wpwhavnx Pyriform Total
(4) Sinus
Site
TABLE
Locations
IV Survival
Patients 4-year Nl N2
6 3 3 5 7 1 4 29
tween October 1, 1988 and October 1, 1992 were reviewed. Only those patients undergoing a SND with both clinically and pathologically proven regional metastasis were included. Patients with a malignancy other than squamous cell carcinoma, a history of a prior head and neck malignancy, previous treatment to the head and neck region with chemotherapy, radiation, or surgery, as well as those with follow-up of less than 2 years were excluded. Staging was done according to the criteria established by the American Joint Cancer Committee.” The type of SND performed depended on the site of the primary. For primaries of the oral cavity and oropharynx, a supraomohyoid neck dissection, involving the removal of nodal groups I-III, was performed. For primaries of the larynx or hypopharynx, a lateral neck dissection, involving the removal of nodal groups II-IV, was performed. In instances where the most inferior level of the dissection was clinically positive, the dissection was extended one level beyond this. Due to established patterns of lymphatic flow, extension of the dissection to include level V in instances of positive nodal disease at level IV was not done.6 Careful patient selection criteria were applied in all instances. To be eligible for SND the patient could not have had any previous treatment to the neck for this malignancy. In addition only those patients with mobile nodes, ie, without evidence of fixation to the surrounding structures such as the sternocleidomastoid muscle, were eligible. In general this restricted this approach to patients with nodes that were less than 3 cm in size, ie, Nl or N2B and N2C where no individual node was over 3 cm in size. Each neck was approached with the intention of performing a SND. If, however, at the time of operation, other structures, principally the internal jugular vein, were believed to be too close to the nodes to allow for safe and adequate removal, it was sacrificed. Similarly, in the instance of minor SCM THE
AMERICAN
actuarial
survival
rate
47% 67% 41%
involvement a cuff of muscle adjacent to the nodes was removed. All neck dissection specimens were reviewed by a staff pathologist. Any patients in whom evidence of extracapsular spread or multiple levels of involvement were identified received postsurgical radiation. Actuarial disease-specific survival and rates of local, regional, and distant recurrence were determined using the Kaplan-Meier method. Site of recurrence was considered to be at the most proximal site.
RESULTS Two hundred and seven patients underwent neck dissection at our institution over the 4 years included in this study. Of this group, 117 patients undergoing 137 neck dissections met the inclusion criteria of a newly diagnosed squamous cell carcinoma with a minimum follow-up of 2 years. Further limiting the group to include only those patients who underwent a SND in which both clinical and pathologic neck disease was evident limited the group to 36 patients. Elimination of 7 patients due to follow-up less than 2 years resulted in a treatment group of 29 patients undergoing 36 SNDs. Twenty-three males and 6 females were included with an average age of 59 years (range, 35-77 years). Median follow-up time was 3 years. Primary site characteristics are listed in Tables I and II, and the majority were staged T3 or greater. Histologic nodal staging is reported in Table III. The majority of these patients were staged N2, and no patients with N3 disease were included. Additionally, none of the patients had clinical evidence of nodal disease at level V. Twenty-two patients underwent unilateral SND: 13 supraomohyoid and 9 lateral neck dissections. Seven patients underwent bilateral selective neck dissections: 2 bilateral supraomohyoid and 5 bilateral lateral neck dissections. Based on the presence of extracapsular spread or multiple levels of nodal disease, 20 patients received postoperative radiation to the neck. Actuarial disease-specific survival at 4 years was 47% overall; 67% in the Nl patients and 41% in the N2 patients (Table IV). The 4-year actuarial regional failure rate was 4%, with neck recurrence in only 1 instance. This patient JOURNAL
OF
SURGERY”
VOLUME
172
DECEMBER
1996
655
( SELECTIVE
NECK
DISSECTION/TRAYNOR
ET AL 1 1
had a T3NZb well-differentiated squamous cell carcinoma and underwent composite resection and ipsilateral supraomohyoid neck dissection. Due to the presence of nodal disease at levels I and II and the presence of extracapsular spread she underwent postoperative radiation therapy. Recurrence in the dissected neck at level III, which was included in the original neck dissection, was noted at 5 months postoperatively. The 4-year actuarial local recurrence and distant failure rate were 36% each, with local recurrence occuring in 6 patients and distant metastases developing in another 8 (Table V).
DISCUSSION This investigation is unique in that it limited its focus to include only those patients with both clinical and histologic evidence of cervical metastases who were treated with a selective neck dissection. Traditionally, the surgical management of these patients has been the RND involving the en bloc removal of the cervical lymphatics along with the stemocleidomastoid muscle, the internal jugular vein, and the spinal accesory nerve.’ If the RND gauranteed regional control of disease and increased survival then, perhaps, the undeniable morbidity associated with its use would be acceptable. This, however, is not the case as demonstrated by the regional recurrence rates of 54.3% for node-positive patients and 71.3% with nodes positive at multiple levels reported by Strong in his review of the Memorial Hospital experience with the RND.4 Prompted by such imperfect regional control of disease, as well as a growing emphasis on improving posttreatment function and cosmesis, head and neck oncologic surgeons started to question the routine use of the RND and began to modify it. Modifications of the RND have not been made in a haphazard manner. Rather, they seem to be based on firm anatomic and oncologic principles not appreciated when the RND was first described nearly 100 years ago, An early impetus for modification was the recognition that the operation was, in fact, not a complete en bloc resection since other seemingly related structures such as the hypoglossal and vagus nerve and carotid artery were not removed.14 Additionally, gross and histological analysis of surgical and necropsy specimens demonstrated the cervical lymphatics are contained within distinct fascial compartments without intramuscular or intra-adventitial lymph nodes.’ Other anatomic studies have confirmed the lymphatic system of the neck lies within a system of aponeurotic compartments that, on the outer side, envelop muscles and vessels of the neck but, other than proximity, have no relationship to these structures.’ Bocca applied these principles in advocating the “functional” or complete neck dissection in which the lymphatics of all five nodal levels of the neck were removed by stripping the fascial cover from muscles and vessels and dissecting free major nerves in an effort to spare the structures commonly sacrificed during the RND. Regional recurrence rates of 2.4% for NO and 30% for Nl/N2 necks compared favorably to rates of 6.6% and 48% in similar patients in which members of the same treatment team performed RNDs.” A predictable pattern of cervical lymphatic flow based on the primary site is, by now, well described.5,6,‘6 In an early study conducted by Fisch employing lymphography, lym656
THE AMERICAN
JOURNAL
OF
SURGERY@
VOLUME
172
TABLE
V Recurrence 4 Year Local Regional Distant
Actuarial
Recurrence
Rates 36% 4% 36%
phatic flow from the upper posterior cervical (“junctional”) nodes into both the spinal and jugular nodal systems was demonstrated.5 Indirect flow from the spinal nodes to the jugular nodal system was evident, but filling of the spinal nodes from the jugular chain was never observed. This definite unidirectional flow was attributed to the presence of lymphatic valves, and its importance in predicting potential sites of cervical metastatic spread is apparent. The absence of retrograde flow from the jugular nodes into the spinal chain may explain why metastatic spread to the posterior triangle (region V) is uncommon. A review of the records of 1,277 patients undergoing RND identified positive nodes at region V in only 3% of instances, with a slight increase to only 5% when clinically positive nodes were evident.” Additional investigators have confirmed this low prevalence of posterior triangle involvement.‘* Studies conducted by Shah and associates have lended further credence to the existence of a predictable cervical drainage pattern, even in the presence of metastatic disease.‘j Specimens from more than 1,000 comprehensive neck dissections performed on individuals with squamous cell carcinoma of the head and neck were evaluated for the presence and site of metastatic disease. A predominance of certain regions was seen for each primary site. The vast majority of oral cavity lesion metastases occurred at levels I, II, and III, which comprise those nodal groups removed in the supraomohyoid neck dissection. Oropharyngeal, hypopharyngeal, and laryngeal lesions demonstrated a predilection for metastatic disease to levels II, III, and IV, which would be included in a lateral neck dissection. Metastatic nodal disease outside of the predominant regions was shown to be unusual in the absence of involvement of the expected regions. Thus, a compartmentalization of the cervical lymphatics along with a predictable pattern of lymphatic flow in which the posterior triangle is rarely involved have been demonstrated. These observations do not support the routine removal of all cervical regions and the sacrifice of the SCM, IJ, and spinal accessory nerve dictated by the radical or complete neck dissections, and suggest that the selective removal of only those nodes at greatest risk for metastatic disease might be sufficient to adequately treat and stage the at-risk neck. This type of neck dissection, termed the SND, represents the next logical step in the modification of the RND. Several investigators have demonstrated that the use of the SND in the clinically negative neck provides treatment results and staging information comparable to the more extensive lymphadenectomies. As a result, its use in these instances has become common at many institutions.‘-” The use of the SND in the clinically positive neck, however, has not yet been widely accepted. There is, however, evidence to support this approach in certain instances. DeDECEMBER
1996
) SELECTIVE
spite the finding of neck disease in the pathologic review of an elective neck dissection specimen control rates comparable to those seen with more extensive operations are still seen. Additionally, a review of the literature reveals the performance of a therapeutic SND in patients with both clinical and histologic evidence of metastatic disease is not a new concept. Several investigators have described its use in these instances, but a study confined soley to this group of patients seems to be lacking. Medina reported his experience in managing patients with upper aerodigestive tract squamous cell carcinoma using the supraomohyoid neck dissection. Forty-four patients had both clinical and pathologic evidence of metastatic disease, with the majority staged N 1. Regional recurrence rates of 10% for N 1 disease without extracapsular spread and 24% when multiple levels or ECS was found were described. Neck recurrence rates in the latter group decreased to 15% when postoperative radiation therapy was added, and no patients in the former group received radiation. ” Byers reported the same recurrence rates in a similar group of patients undergoing therapeutic supraomohyoid neck dissection, both with and without adjunctive radiation therapy, in which 70% had extracapsular spread or multiple levels of involvement.‘” Based on their favorable results these authors suggested the SND has a meaningful role in certain patients with Nl metastatic disease to the neck, and that the addition of adjunctive radiation therapy may improve regional control If ECS or multiple levels of involvement are identified. With this in mind it would seem reasonable to extend the indications of this operation to include not only Nl disease, but also N2B and N2C disease, where the individual nodes are nonfixed Nl nodes, because the rationale behind this procedure remains viable in the absence of gross extracapsular spread or nodal fixation. The results of our study seem to support such an approach. Importantly, it must be emphasized we do not advocate the use of the SND in all cases. Specifically, patients in which a SND is applied included only those without evidence of massive adenopathy, nodal fixation, and gross ECS or a history of prior neck surgery. More massive adenopathy with gross extracapsular spread may violate the fascial compartments of the neck and disrupt the predictable pattern of lymphatic flow. The presence of only 1 patient with N2A disease and the absence of any patients with N3 disease is demonstrative of our willingness to return to more traditional neck dissections when indicated to completely remove gross disease. Additionally, when a suspicious node is encountered at the lowest level of the dissection we routinely extend the operation to include the next echelon. This principle does not apply to level V because, due to the established pattern of lymphatic flow, this level is at limited risk from level IV disease.5x”,‘” In no instance is gross disease left behind, and the internal jugular vein, 11th cranial nerve, and sternocleidomastoid muscle must be removed if proximity to disease precludes safely sparing these structures.
THE AMERICAN
NECK
DISSECTION/TRAYNOR
ET AL 1
CONCLUSION Our experience with 29 patients who had both clinical and histopathological evidence of cervical metastases staged Nl to NZC, and underwent SND, was reviewed. Treatment outcomes including regional control rate and survival are comparable with those of more extensive neck dissections in similar patients. The SND is based on firm anatomlc and oncologic principles rhat have become apparent since the RND was first described nearly 100 years ago. The results of this study suggest the careful extension of its mdicatlons to include some patients with both clinical and histological neck disease is reasonable.
REFERENCES 1. Crile GW. Excision of cancer of the head and neck. JAMA. 1906;47:1780-1786. 2. O’Brien CJ, Stmth JW, Soong SJ, et al. Neck dissection with and without radiotherapy: prognostic factors, patterns of recurrence, and survival. Am J Surg. 1986;152:456-463. 3. Johnson JT, Myers EN, Bedetti CD, ct al. Cervical lymph node metastases: incidence and implications of cxtracapsular carcinoma. Arch Otolaryngol. 1985;111:534+537. 4. Strong EW. Preoperative radiation and radical neck dissection. Surg Clin North Am. 1969;19:271-276. 5. Fisch UP, Siegel ME. Cervical lymphatic system as viewed by lymphography. Ann Otol Rhinol Lqngoi 1964;73:869-882. 6. Shah JP. Patterns of cervical lymph node metastasis from squamous cell carcinomas of the upper aerodlgestive tract. Am 1 Surg. 1990;160:405-409. 7. Suarez 0. El Problema de las metastabia linfatlca y alejadas de1 cancer de laringe e hipofaringe. Reu Otorhinolanngol. 1963;23:83. 8. Bocca E, I’ignatoro 0. A conservation technique in radical neck dissection. Ann Otol Rhinol Luryngol. 1976;76:975-987. 9. Byers RM, Wolf PF, Ballantyne AJ. Rationale for elective modified neck dissection. Head Neck Surg. 19&X:160-167. 10. O’Brien CJ. A selective approach to neck dissection for mucosal squamous cell carcinoma. Aust N 2 J Surg 1994;64:236-241. 11. Medina JE, Byers RM. Supraomohy<,id neck dlssection: ratianale, indications and surgical technique. ffccid Neck. 1989:I 11-122. 12. Alvi A, Johnson JT. Extracapsular qwad in the clinically negative neck (NO): implications and outcomes. Otolaqngol Head and Neck Surg. 1996;114:65-70. 13. American Jomt Committee on Cancer. Manual for Staging of Cancer, 4th ed. Philedelphia: J.B. Lippincott CO., 1992:33-35. 14. DeSanto LW, Beahrs OH. Modified Jnd complete neck dissection in the treatment of squamous cell carcinoma of the head and neck. Surg Gynec und Obstet. 1988;167:259-269. 15. Bocca E, Pignataro 0, Oldini C, (:qxl C. Functional neck dissection: an evaluation and review of 843 cases. Laryngosc0@. 1984;94:942-945. 16. Lindberg R. Distribution of cervical lymph node metastases from squamous cell carcinoma of the upper respiratory and digestive tracts. Cancer. 1972;29:1446-1451. 17. Davidson BJ. Kulkarny V, Delacure MD, Shah JP. Posterior triangle metastases of squamous cell carcinom of the upper aerodigestive tract. Am] Surg. 1993;166:395-398. 18. Skolnik EM, Yee KF, Friedman M, Golden TA. The posterior triangle in radical neck surgery. Arch Otokqngol. 1976;102:1-4. 19. Byers, RM. Modified neck dissection: a study of 967 cases from 1970 to 1980. Am J Surg. 1985;150:414~421.
JOURNAL
OF
SURGERY”
VOLUME
172
DECEMBER
1996
657