Figure 1. Recurrence-free survival estimates (Kaplan-Meier) of patients with squamous cell carcinoma of the upper digestive tract.
Figure 2. Overall survival estimates (Kaplan-Meier) of patients with squamous cell carcinoma of the upper digestive tract.
Figure 3. Recurrence-free survival estimates (Kaplan-Meier) of patients with squamous cell carcinoma of the larynx.
Figure 4. Overall survival estimates (Kaplan-Meier) of patients with squamous cell carcinoma of the larynx.
Canis M, Plüquett S, Ihler F, Matthias C, Kron M, Steiner W. Impact of Elective Neck Dissection vs Observation on Regional Recurrence and Survival in cN0-Staged Patients With Squamous Cell Carcinomas of the Upper Aerodigestive Tract. Arch Otolaryngol Head Neck Surg. 2012;138(7):650-655. doi:10.1001/archoto.2012.1026
Author Affiliations: Department of Otorhinolaryngology–Head and Neck Surgery, University of Göttingen, Göttingen, Germany (Drs Canis, Ihler, Matthias, and Steiner); Department for Shoulder Surgery, Arthroscopic Surgery, and Sports Traumatology, Roland Clinic Bremen, Bremen, Germany (Dr Plüquett); and Department of Biometry and Medical Documentation, University of Ulm, Ulm, Germany (Dr Kron).
Objective To evaluate the impact of elective selective neck dissection vs observation on regional recurrence and survival in cN0-staged patients after transoral microscopic laser surgery of squamous cell carcinomas of the upper aerodigestive tract.
Design Retrospective unicenter study performed from 1986 through 2003.
Setting Tertiary referral center.
Patients A total of 202 of 425 patients (48%) with previously untreated squamous cell carcinoma of the upper aerodigestive tract (T1-T4) underwent transoral laser surgery and selective neck dissection with curative intent. A total of 223 patients (52%) received tumor resection only. Stage distribution was as follows: stage I, 50 cases (12%); stage II, 216 cases (51%); stage III, 120 cases (28%); and stage IVa, 39 cases (9%).
Main Outcome Measures Overall survival, recurrence-free survival.
Results In the neck dissection group, 15% of occult metastases were found. In pN0 necks we found 4 late metastases (4%), and in pN+ necks, 1 recurrent neck metastasis (7%) (pT1-pT2) (n = 109). In patients with pT3-pT4 tumors (n = 93), 5 late neck metastases (5%) were observed. The wait-and-see group comprised 7 late metastases (4%) in patients with pT1-pT2 tumors (n = 171) and 9 late metastases (17%) in patients with pT3-pT4 tumors (n = 52). In early-stage T1-T2 tumors, the 5-year recurrence-free survival rate was 95% in the neck dissection group and 96% in the wait-and-see group. Advanced stages T3 to T4 presented a 5-year recurrence-free survival rate of 90% in the neck dissection group and 75% in the wait-and-see group. In early-stage T1-T2 tumors, the 5-year overall survival rate was 83% in the neck dissection group and 72% in the wait-and-see group. Advanced T3-T4 categories showed 5-year overall survival rates of 58% in the neck dissection group and 60% in the wait-and-see group.
Conclusions A wait-and-see approach is justified in patients with early-stage disease. Elective selective neck dissection should be considered in patients with advanced cancer of the upper aerodigestive tract. Because of selection effects, the possible conclusions of this study are limited. There is a definite need for prospective, randomized studies.
Squamous cell carcinomas (SCCs) of the upper aerodigestive tract have a relatively high potential of spreading to cervical lymph nodes, and adequate diagnosis and treatment of the clinically metastases-free neck (cN0) is still challenging for head and neck surgeons, radiologists, and radiotherapists. Depending on risk factors, such as site and size of the primary tumor, pT category, depth of tumor infiltration, grading, and biological parameters, there is a relevant rate of occult metastases even if the neck is proven to be cN0.1 Rates in the literature vary among different tumor categories and range from 8% to 45%.2- 4 However, the most informative prognostic marker, besides the size of the primary tumor, is the presence of metastases at the time point of first diagnosis, which predicts an average reduction of overall survival by about 50%.5- 7
To improve neck control, elective surgery of the neck has been used to treat tumors with a high risk (>20%) of occult metastases.8 The main advantage of surgery is to provide histologic information that is useful for subsequent treatment. If the neck dissection is performed, small, clinically occult lymph node metastases can be detected and verified histologically.1 This is particularly important if there is extranodal spread of tumor cells, which definitively influences the decision for postoperative therapy. Since 1979, the regionally limited functional neck dissection (RFND) of levels II and III in laryngeal carcinoma has been performed in our hospital. According to the recommendation of the American Academy of Otolaryngology–Head and Neck Surgery, RFND corresponds with selective neck dissection (SND).9 Since the early 1980s, SND has been routinely performed in patients presenting with tumors of the upper aerodigestive tract and N0-2 neck disease. Selective neck dissection was mainly conducted after transoral laser microsurgery of the primary tumor as a delayed procedure. The results of more than 500 cases treated with SND at the Department of Otorhinolaryngology–Head and Neck Surgery, at the University of Göttingen (Göttingen, Germany), from 1986 to 1997 were published by Ambrosch et al.10
However, when the risk of occult disease is low, adverse effects and complications of anesthesia and the potential risks of surgery must be considered, and a “wait-and-see” strategy may be justified. In the present study we therefore aimed to evaluate the impact of selective neck dissection vs a wait-and-see strategy on regional recurrence and survival in cN0-staged patients after transoral laser microsurgery for SCCs of the upper aerodigestive tract.
This retrospective study comprised 425 patients who presented with SCC of the upper aerodigestive tract and who were treated (from1986 through 2003) by transoral laser microsurgery with curative intend. All patients were previously untreated and were clinically staged as having N0 neck disease. Primary tumors were located in the oral cavity (in 80 patients), the oropharynx (in 41 patients), the hypopharynx (in 25 patients), the supraglottic larynx (in 95 patients), and the glottic larynx (in 184 patients). Exclusion criteria for this study were non-SCC tumors, clinical N1 to N3 neck disease, impossibility of organ-preserving surgery, primary or adjuvant radio(chemo)therapy, previous primary tumor, simultaneously second primary tumor, simultaneous distant metastasis, recurrent disease of tumors treated elsewhere primarily, and palliative treatment. Further exclusions were patients who experienced locoregional recurrences during follow-up, because the failure in the neck may be assumed to be result of failure to control the primary tumor with subsequent reseeding in the neck.
Preoperatively, all patients were assed for lymph node metastases by palpation of the neck, ultrasonography, and/or computed tomography (CT) or magnetic resonance imaging (MRI). Nodes smaller than 1 cm were staged as cN0. Presupposing that the neck remained negative for nodal disease after imaging, a decision for further treatment (neck dissection vs a wait-and-see strategy) was made regarding the primary tumor–and patient-related factors. The criteria for the decision were site (larynx vs hypopharynx), size (T1 vs T4), depth of infiltration (≤3 mm vs >3 mm), and grading (G1 vs G3) of primary lesion. If the patient presented with advanced primary disease (T3-T4), the tumor infiltration depth was more than 3 mm, and the primary lesion was undifferentiated (G3), a unilateral or bilateral selective neck dissection was performed in a 2-stage manner (1-2 weeks after tumor resection). Patients underwent bilateral neck dissection if the primary tumor was of an advanced stage with midline localization. If the primary lesion was an early tumor (T1-T2), tumor infiltration depth was 3 mm or less, grading was differentiated (G1-G2), or if the patient refused the recommendation for elective selective neck dissection, a wait-and-see approach was used. Follow-up of all patients consisted of ultrasonography and ear, nose, and throat examination every 6 to 8 weeks during the first 2 years of follow-up and then every 6 months thereafter.
Preoperative examination consisted of magnifying laryngoscopy on the awake patient and ultrasonography of the neck for lymph node evaluation and staging; CT or MRI of the neck was undertaken unless the patient presented with satisfactory imaging performed at the referring hospital. Further standard preoperative investigations included radiographic examination of the chest and ultrasonography of the abdomen.
At the commencement of the planned surgery under general anesthesia, a panendoscopy was performed to exclude any second primary tumor in the aerodigestive tract and upper digestive tract prior to tumor resection. The transoral laser microsurgery was undertaken with the carbon dioxide laser in continuous superpulse mode. Resections were performed using the step-by-step technique described by Steiner,11 cutting through the tumor and thus allowing the surgeon to inspect the surrounding under microscopic magnification. With this technique, the differentiation between tumor and healthy tissue is feasible and allows the surgeon to follow the tumor while preserving as much healthy tissue as possible. Frozen sections were necessary in all cases of larger tumors. If definite histologic resection margins showed positive or unclear results, further resections were conducted until R0 resection was verified.
All survival rates were calculated by using the Kaplan-Meier method.12 Survival curves were compared between groups with the log-rank test. Postoperative follow-up data were available for all patients. The overall survival time was defined as the interval between the date of surgery and the date of the last consultation (censored) or date of death (event). For the purposes of the statistical analysis, recurrence-free survival, intercurrent deaths, and deaths due to secondary primary tumors, as well as patients alive without recurrences, were regarded as censored observations. Events included recurrences, distant metastases, and deaths due to disease.
The total of 425 patients who met the inclusion criteria for the present study consisted of 71 women (17%) and 354 men (83%) with a mean (SD) age of 61 (16) years (range, 20-91 years). Follow-up ranged from 4 to 263 months (median, 64 months). All 425 patients were classified according to the current Union Internacional Contra la Cancrum/American Joint Committee on Cancer classification formulated in 2002. Stage distribution was as follows: stage I, 50 cases (12%); stage II, 216 cases (51%); stage III, 120 cases (28%); and stage IVa, 39 cases (9%). Postoperative T categories according to tumor site are given in Table 1.
A total of 223 patients (52%) were exclusively treated by transoral laser surgery of the primary site and with the wait-and-see policy afterward. Neck dissection was performed in 202 patients (48%), in 127 patients (30%) unilaterally and in 75 (18%) bilaterally. In all patients, a level II/III selective neck dissection was performed. In addition, level I and/or IV was completed in 43 of 227 neck sides.
In the neck dissection group (all patients) (n = 280), histopathologically positive lymph nodes were found in 30 patients and 30 neck dissections (15%). Fifteen occult metastases (14%) were found in the pT1-pT2 group (n = 109) and 15 (16%) in the pT3-pT4 group (n = 93). Table 2 gives primary, late, and recurrent neck metastases in both the wait-and-see group and the neck dissection group. Within the neck dissection (pN0/pN+) group, 4 late (4%) and 1 recurrent (7%) neck metastases were found in patients with tumors staged as pT1 to pT2, and 5 late neck metastases (5%) were observed in those with tumors staged as pT3 to pT4. Neck dissection failure therefore is calculated as 5%. The wait-and-see group (n = 145) comprised 7 late neck metastases (4%) in patients with tumors staged as pT1 to pT2 and 9 late neck metastases (17%) in those with tumors staged as pT3 to pT4. All patients in the neck dissection group (n = 10) and the wait-and-see group (n = 16) who showed late or recurrent neck metastases received salvage surgery. Within the neck dissection and the wait-and-see group no patient received adjuvant radio(chemo)therapy.
The subgroup of patients with SCC of the upper digestive tract (oral cavity, oropharynx, and hypopharynx) (n = 146) primarily showed occult metastases in 21 of 102 patients who had undergone neck dissection (21%). Late neck metastases within the pN0 group occurred in 3 of 49 patients with T1 to T2 tumors (6%) and 2 of 32 patients with T3 to T4 tumors (6%). Recurrent neck metastases in the pN+ group were found in 1 of 13 patients with T1 to T2 tumors (8%) and in none of the 8 patients with T3 to T4 tumors. Within the wait-and-see group, 3 of 35 patients in the T1-T2 group (9%) and 3 of 9 patients in the T3-T4 group (33%) showed late neck metastases.
The subgroup of patients with laryngeal cancer (glottis and supraglottis) (n = 279) showed occult metastases in 9 of the 100 who underwent neck dissection (9%). Late neck metastases within the pN0 group occurred in 1 of 45 patients with T1 to T2 tumors (2%) and 3 of 46 patients with T3 to T4 tumors (7%). No recurrent neck metastases in the pN+ group were found in 2 patients with T1-T2 tumors (8%) and in 7 patients with T3-T4 tumors. Within the wait-and-see group, 4 of 136 patients in the T1-T2 group (3%) and 6 of 43 (14%) in the T3-T4 group showed late neck metastases.
The 5-year recurrence-free survival rates of 146 patients with SCC of the upper digestive tract (oral cavity, oropharynx, and hypopharynx) (n = 146) were 92% in the neck dissection group (n = 102) and 83% in the wait-and-see group (n = 44). There was no significant difference between both groups (P = .08) In early T1-T2 tumors, 5-year recurrence-free survival rates were 93% in the neck dissection group (n = 62) and 89% in the wait-and-see group (n = 35). Advanced T3-T4 tumor categories presented 5-year, recurrence-free survival rates of 89% in the neck dissection group (n = 40) and 57% in the wait-and-see group (n = 9) (Figure 1). Comparison between the groups showed no significant differences (T1-T2 group, P = .44; T3-T4 group, P = .07). However, among those with T3-T4 tumors, a tendency toward a better recurrence-free survival within the neck dissection group was seen. The 5-year overall survival rates were 68% in the neck dissection group and 61% in the wait-and-see group, which are not significantly different (P = .69). Among those with early T1-T2 tumors, the 5-year overall survival rates were 78% in the neck dissection group (62 patients) and 63% in the wait-and-see group (35 patients). Advanced T3-T4 categories showed 5-year overall survival rates of 52% in the neck dissection group (n = 40) and 47% in the wait-and-see group (n = 9) (Figure 2). Comparison between the groups revealed no significant differences (P = .69).
Five-year recurrence-free survival rates of patients with laryngeal carcinoma (supraglottic and glottis) (n = 96) were 93% in the neck dissection group (n = 100) and 94% in the wait-and-see group (n = 179). There was no significant difference between the groups (P = .98). In those with early T1-T2 tumors, the 5-year recurrence-free survival rates were 97% in the neck dissection group (n = 47) and 97% in the wait-and-see group (n = 197). Advanced stages T3 to T4 presented 5-year recurrence-free survival rates of 90% in the neck dissection group (n = 53) and 79% in the wait-and-see group (n = 43) (Figure 3). No significant differences were observable between the groups (T1-T2 tumors, P = .27; T3-T4 tumors, P = .21). The 5-year overall survival rates were 77% in the neck dissection group and 71% in the wait-and-see group, which were not significantly different (P = .59). In early T1-T2 tumors, the 5-year overall survival rates were 91% (n = 47) in the neck dissection group and 74% in the wait-and-see group (n = 136). The advanced T3-T4 categories showed 5-year overall survival rates of 63% in the neck dissection group (n = 53) and 59% in the wait-and-see group (n = 43) (Figure 4). Comparison between the groups revealed no statistical differences (T1-T2 tumors, P = .30; p3-p4 tumors, P = .59).
The presence of cervical lymph node metastasis decreases overall survival in head and neck SCC by approximately 50%.13,14Identification of patients with a high risk of occult metastases therefore presents a primary concern in treatment of SCC. However, the rate of occult lymph node metastases varies from 8% to 45%2- 4 according to tumor site and tumor extent. Previous investigations have found increased survival in patients undergoing elective neck dissection compared with the wait-and-see strategy when having a risk of occult lymph node metastases larger than 20%.8
Despite these findings, elective surgery of occult neck disease is still under discussion. Incidence rates of occult lymph node metastases detected by neck dissection were 21% for patients with SCC of the upper digestive tract (n = 102) and 9% for patients with laryngeal cancer (n = 100). These results are in line with those of previous investigations, which showed incidence rates of occult metastases in SCC of the upper digestive tract of 25% and 23%.4,15 In patients with laryngeal tumors, the rate of occult metastases has been found to be lower.16,17 In this regard, neck dissection provides valuable staging information in the cN0 neck and allows the earliest treatment of occult nodal disease. However, in pN0 necks, the rate of late neck metastases showed no significant difference between the neck dissection group and the wait-and-see group. This fact may be explained by a successful clinical preselection of patients with a high risk of occult metastases into the neck dissection group. Notwithstanding, compared with the neck dissection group (with rates of 6% and 7%, respectively), there was a tendency toward late neck metastases in advanced tumors within the wait-and-see group (with rates of 33% and 14%, respectively).
Previous investigations have confirmed selective neck dissection as an effective procedure for staging the clinically negative neck by having similar regional recurrence rates compared with modified radical and radical neck dissection (RND).18- 20 Up to the late 1960s, RND remained the unquestioned standard surgical approach. Based on anatomic observations, a more conservative technique has been proposed.17,21,22 This operation, with lymph node removal of levels I to V and preservation of 1 or more nonlymphatic structures, was defined as modified radical neck dissection (MRND) by the American Academy of Otolaryngology–Head and Neck Surgery in 1991.9 Based on the observation that in specimens of RNDs and MRNDs, most occult metastases occurred in levels II and III, one of us (W.S.) introduced the SND for laryngeal carcinomas in our clinic. In SND, only the high-risk lymph node groups are removed, and the nonlymphatic structures are preserved. The first results achieved with SND in treatment of laryngeal carcinomas were published in 1984.23 In the following years, indications for SND were widened to other head and neck sites and more advanced neck disease. In 2001, a study10 on 503 cases of selective neck dissection in patients with N0-N2 neck disease was conducted and showed comparable results of SND (with or without adjuvant radiotherapy) and MRND or RND in regard to regional control. The authors concluded that SND is a contribution to the concept of less invasive surgery and offers functional and esthetic advantages without oncologic compromise. The rates of neck failure published range from 3% to 10%24,25 in the previously dissected levels, which is in line with 5% neck failure rate in our series. Therefore, selective neck dissection combines a low comorbidity without comprising regional control and survival.
The benefits of elective selective neck dissection in patients with early-stage tumors and N0 neck are still not clear because results of existing studies have been divergent. Most investigations have failed to show statistically significant differences in recurrence-free and overall survival in neck dissection and wait-and-see groups.26 In our series, we also did not observe any significant difference between both groups in early-stage SCC of the upper aerodigestive tract and the larynx. However, there have been few retrospective studies, mostly in patients with SCC of the oral cavity, which showed a significant survival benefit in favor of elective neck dissection.27- 29 In contrast, a prospective study by Kligerman et al30 revealed a statistically significant disease-free survival benefit of elective neck dissection in early-stage T1-T2 SCC of the oral cavity.
Because there were no differences in mortality and survival in our groups with early-stage SCC, both observation and elective neck dissection should be offered to the patients with a clear explanation of the advantages and disadvantages. However, one should consider the risk of 15% occult metastases, which might be psychologically stressful. Of those patients who cannot be followed up every 4 to 6 weeks, the recommendation of elective selective neck dissection might be preferable.
In patients presenting with advanced T3-T4 SCC of the upper digestive tract, there was a tendency toward improvement of recurrence-free survival in the neck dissection group (89% vs 57%). However, neck dissection did not significantly improve overall survival rates (52% vs 47%). These data are in line with those of a study conducted by Poeschl et al,31 who found a tendency toward neck dissection in T4 oral cavity SCC. In patients with laryngeal carcinoma, this trend was not as obvious as in those with cancer of the upper digestive tract. Notwithstanding, in patients whose tumors are staged as T3 to T4, there also was a tendency toward a better recurrence-free survival (90% vs 79%). We therefore support the recommendation of selective neck surgery in patients with advanced SCC of the upper aerodigestive tract. However, our data may be biased because the treatment decision is influenced by many factors, such as psychological factors, patient's wishes for security, and the degree of medical information. Further effects (eg, selection of patients with a high risk of occult metastases into the neck dissection group) may also influence the therapeutic outcome and limit possible conclusions. For a more conclusive answer, prospective randomized studies with a large number of patients need to be conducted.
In conclusion, from our retrospective results, a wait-and-see policy is justified in cooperative patients with early-stage disease. Elective selective neck dissection should be considered in patients with advanced cancer of the upper digestive tract and the larynx. Because of the disadvantages of a retrospective study, this question warrants further research.
Correspondence: Martin Canis, MD, Department of Otorhinolaryngology–Head and Neck Surgery, University of Göttingen, Robert-Koch-Strasse 40, 37075 Göttingen, Germany (email@example.com).
Submitted for Publication: February 23, 2012; final revision received April 4; accepted April 21, 2012.
Author Contributions: All authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Steiner. Acquisition of data: Canis, Plüquett, Matthias, and Steiner. Analysis and interpretation of data: Canis, Plüquett, Ihler, Matthias, Kron, and Steiner. Drafting of the manuscript: Canis, Ihler, and Matthias. Critical revision of the manuscript for important intellectual content: Canis, Plüquett, Ihler, Matthias, Kron, and Steiner. Statistical analysis: Canis, Plüquett, Ihler, and Kron. Obtained funding: Plüquett. Administrative, technical, and material support: Canis, Ihler, and Matthias. Study supervision: Canis, Matthias, and Steiner.
Financial Disclosure: None reported.
Previous Presentation: The results from this study were presented at the Combined Otolaryngology Spring Meetings; April 28-29, 2010; Las Vegas, Nevada.