Improved survival in patients with head and neck cancer in the 1990s

It is generally felt amongst the medical profession and the lay public that cancer is being treated more successfully than in the past. This is certainly true for childhood malignancies and leukaemia but evidence that significantly improved survival is occurring in the common solid tumours is lacking. Since 1963 the University of Liverpool Department of Otolaryngology/Head and Neck Surgery has collected data on all patients with head and neck tumours presenting to the department. The present study investigates patients with histologically proven squamous cell carcinoma of the four main sites: larynx, hypopharynx, oral cavity and orophyarnx. From 1963 until the end of 1989, 2738 patients were seen by the department and from 1990 a further 717 patients have been seen. Since 1990 patients have tended to be in better general physical condition but, on the other hand, have tended to have more advanced disease at the primary site. The department has latterly tended to see fewer laryngeal cancers and more cancers of the oropharynx. Significantly fewer patients have presented with neck node metastases. Multiple logistic regression suggests that the most significant difference between the two groups is the great reduction in neck node recurrence rates in the group of patients seen since 1990 (P = 0.0001). The recurrence of tumours at the primary site since 1990 has been 35% compared with 41% before 1990, and recurrence in the neck nodes since 1990 has been 12%, compared with 15% before 1990. These differences are significant (P = 0.0141 and P = 0.0494, respectively). When studying survival in the 1960s, 1970s and 1980s, the 5-year cure rate was 50%, whereas since 1990 the figure has risen to 60% tumour-specific 5-year survival—a significant difference. A similar effect was noted in observed survival. This improvement in cure rate occurred for all four main sites. The results were confirmed by Cox’s proportional hazards model where year of treatment was highly significantly associated with improved survival (P = 0.0001). It has been demonstrated that locoregional recurrence has improved since 1990 and this is reflected in improved survival figures. Although there are differences in the parameters of tumours referred before 1990 and since 1990, multivariate analysis suggests that the improvement in neck node recurrence rates may be responsible for this improved survival rate. Multivariate analysis for survival also suggests that the improvement in cure rates is independent of compounding variables and dependent on the year of presentation of the tumour. This improved survival may be related to factors, such as the administration of radical postoperative radiotherapy.     

The effects of age on survival and other parameters in squamous cell carcinoma of the oral cavity,pharynx and larynx

Many papers have been written on the effect of age on survival from cancer and a number of these papers have concentrated on cancer of the head and neck. The literature is fairly evenly split between those studies that claim that the young patient has a better chance of survival and those that suggest the older patient has a better chance of survival. The present study investigates 2647 patients with histologically proven squamous cell carcinoma of the oral cavity, oropharynx, larynx and hypopharynx. The tumour-specific 5-year survival of patients with head and neck cancer from the third decade through to the seventh decade at presentation was 54%, whereas this figure dropped to 44% for the eighth, ninth and tenth decades. This difference was statistically significant (P = 0.0001). When the patients in the third to seventh decades of presentation were compared with those from the eighth to tenth decades, it was found that older patients tended to have significantly more advanced disease at the primary site and fewer neck node metastases when compared with younger patients at presentation. These differences were confirmed by multiple logistic regression. Multivariate analysis of survival confirmed that advanced age was associated with poor survival (P = 0.0001). Whilst patients with head and neck cancer in their eighth, ninth and tenth decades fared worse than younger patients, their mean tumour specific survival at 5 years was in the region of 44%, which makes treatment worthwhile, certainly in selected cases.     

Metastatic squamous cell carcinoma in cervical lymph nodes from an unknown primary tumour: prognostic factors

The aim of this retrospective review is the study of the prognostic factors related to cervical metastases of squamous cell carcinoma from an unknown primary tumour. Sixty-seven patients were selected and surgery and postoperative radiotherapy was the treatment used. Nineteen tumours were subsequently found (27%). The 5-year actuarial survival rate of all patients was 22%. Survival rates were significantly related to lymph node stages and to the histological degree of differentiation. Nevertheless, actuarial survival rates were not related to the appearance of the primary tumour (P = 0.07). In our series, the single most important prognostic factor was the neck stage. The value close to statistical significance observed when the primary tumour subsequently appeared (P = 0.07), suggests that this could worsen the prognosis.     

Soft tissue cervical metastases of squamous carcinoma of the head and neck

Some 497 of 3085 patients with squamous cell carcinoma of the head and neck treated between 1963 and 1990 had a later radical neck dissection at some time after initial treatment. The histological slides were all reviewed, firstly to confirm the presence of squamous cell carcinoma within the neck, and secondly to ascertain whether the metastasis was to soft tissue, to a lymph node or to both. The presence of extracapsular rupture in lymph node deposits was also assessed. Of the 497 patients, 138 had soft tissue deposits only, and 359 had nodal deposits only. Of the patients with nodal deposits 165 had extracapsular rupture and 194 did not. The 5-year survival of the 138 patients with soft tissue metastases was 27% compared with 33% for patients with extracapsular rupture and 50% for patients with no extracapsular rupture. Weighted logistic regression showed that soft tissue deposits were significantly more common in patients in poor general condition, plus poorly differentiated squamous cell carcinoma plus T₄ tumours (P < 0.005), and in patients with poorly differentiated squamous cell carcinoma plus T₄ tumours (P < 0.025). Cox's multivariate analysis with backward elimination showed that gender, histological differentiation, site of primary tumour and age of patient had no statistically significant effect on survival. The number of nodes (P < 0.0001), the presence of extracapsular rupture (P < 0.0001) and the presence of soft tissue free metastases (P < 0.001) were all highly significant. The N-status at recurrence also reached statistical significance (P < 0.0001).     

Adenoid cystic carcinoma of the head and neck

Adenoid cystic carcinoma has a long natural history but frequently proves fatal. The present study describes 108 patients with an adenoid cystic carcinoma of the head and neck seen over a 30-year period. Analysis of the data utilized both univariate and multivariate methods. Forty per cent of patients had tumours arising from the oral cavity and half of these were in the hard palate; 29% occurred in the major salivary glands; 41% of tumours were locally advanced at presentation and 11% had lymph node metastases at this time. The histological pattern was solid in 25%, cribriform in 40% and tubular in 20%. In addition, 15% of patients had a polymorphous low-grade adenocarcinoma and these were analysed separately. Primary site recurrence was more common in the presence of locally advanced tumours at presentation (T_3–4) (P = 0.0093). Only six patients had surgery with adjuvant radiotherapy. Six patients had no curative treatment, 21 had primary radiotherapy, 39 had local excision and 42 radical excision. The actuarial primary site recurrence rate was 100% at 30 years. The neck node recurrence rate was 23% at 15 years. Tumour specific survival was 40% at 20 years. Solid histology had a worse prognosis than other histological types (P = 0.0429) but those patients with polymorphous low-grade adenocarcinomas fared very well. Patients with tumours of the hard palate fared better than those patients with tumours at other sites (P = 0.0301). Early disease at the primary site (T_1–2) was a good prognostic sign (P = 0.0013). Patients with neck node metastases at presentation tended to do badly (P = 0.009).     

Significado pronóstico de la extensión extranodal en pacientes con carcinomas escamosos de cabeza y cuello cN0 con metástasis ganglionares ocultas

Introduction and objectives

Extranodal extension in nodal metastases is an independent adverse prognostic factor in head and neck squamous cell carcinoma patients. However, few studies specifically address the subgroup of patients with no clinical evidence of nodal disease.

Surgical treatment for pulmonary metastases of squamous cell carcinoma of the head and neck

Purpose: As locoregional control of head and neck cancer has improved, distant metastases have become increasingly common problems.Patients and Methods: To determine the role of surgical treatment, we reviewed 32 patients with squamous cell carcinoma (SCC) of the head and neck who underwent thoracotomy for pulmonary metastases.Results: The overall 5-year survival rate was 32%. The 5-year survival rate of the patients with SCC of the oral cavity was significantly poorer than that of the patients with other primary site (15.4% v 45.2%; P = .01). In the patients with single nodule, extent of the tumor was a significant prognostic factor (P = .007). Mediastinal lymph node involvement (P = .004) and pleural invasion (P = .04) also correlated with survival.Conclusion: TNM classification of the primary tumor did not have an impact on survival in this study. Further studies of a large series should be performed to determine the indications and modalities of the surgical treatment for pulmonary metastases of the SCC of head and neck.     

Non-squamous malignancy in lymph nodes: the occult primary

The present study presents 105 patients seen at a head and neck specialist clinic with a neck gland which subsequently proved to be a non-squamous malignancy. Of the 105 patients, 50 patients were eventually found to have a tumour in the head and neck region, 30 to have a distant primary and in 25 no primary site was ever found. The majority of patients were diagnosed in the clinic after careful examination and most of the remainder were diagnosed during endoscopy/biopsy. Chest radiography was the most useful investigation for diagnosing primary tumours of the lung. The 5-year-survival for the whole group of 105 patients was 28% (95% CI 17–39). The 5-year-survival for the head and neck primary tumour group was 44% (95% CI 25–60). The median survival of patients with a distant primary tumour was only a 6 months, there was one 5-year-survivor. The median survival for those in whom the primary was never discovered was 18 months. However, a reasonable proportion of these patients survived, five being alive at 5 years. The difference between survival for the three groups was statistically significant (P < 0.001). The most common histological type was undifferentiated/anaplastic tumours (37 out of 105) and this was followed by adenocarcinoma (33 out of 105). There was a significant difference in the survival between these two groups (χ²= 2.02, d.f. = 1, P= NS). Multi-variate analysis suggested that survival was better in the older age group and was affected by histology (P= 0.0093, P= 0.0332 respectively). The present study suggests that the treatment of patients in whom the primary site is eventually found to be in the head and neck region is rewarding with the same survival as a similar group of patients with squamous cell carcinoma. Sixty of the group of 105 patients had excision biopsies of the neck node and this did not affect survival.     

The effect of occult nodal metastases on survival and regional control in patients with head and neck squamous cell carcinoma

Objectives: To determine factors associated with disease‐free survival (DFS) and regional control in clinically node‐negative head and neck squamous cell cancer (HNSCC) patients with occult metastasis. Study Design: Non‐randomized retrospective analysis. Materials and Methods: Patients who underwent elective neck dissection (END) from 1985 to 2002 were analyzed. Results: A total of 337 patients underwent END. The majority of patients (67%) had advanced stage disease (T3/T4). Occult metastases were present (pN+) in 168 patients (50%), with extracapsular spread (ECS) present in 72 patients (43%). Five‐year DFS for patients with histologically node negative necks was 62% versus 36% for pN+ patients (P < .0001). Postoperative radiation (XRT) did not significantly influence DFS for pN+ patients with less than three nodes involved, but had a significant association with DFS with three or more nodes involved (P < .0001). XRT showed a trend toward improved regional control rates in patients with less than three positive nodes (86% vs. 78%; P = .7579) and patients with three or more positive nodes (62% vs. 50%; P = .0014). When ECS was present, XRT did not affect DFS in patients with less than three nodes (36%), but had a significant effect on DFS in patients with three or more nodes (20% vs. 0%; P = .0075). Regional control rates were not improved with XRT in ECS‐positive patients with less than three nodes (62% vs. 75%) or with three or more nodes involved (43% vs. 50%; P = .0678). Conclusions: There is a high incidence of occult metastases in clinically node‐negative patients which adversely affects survival, regardless of the use of adjuvant XRT. Postoperative XRT did not significantly affect regional control or survival rates in patients with <3 positive nodes. When ECS was present, survival was poor regardless of the number of nodes. These data emphasize the prognostic and therapeutic role of END and highlight the need for the development of novel therapeutic regimens to improve disease control and survival in HNSCC patients with nodal metastases.     

Control of neck nodes in squamous cell carcinoma of the head and neck by radiotherapy: prognostic factors

313 patients with cervical metastases from a squamous carcinoma of the head and neck treated with radiotherapy, were studied by means of a multivariant analysis in order to determine the prognostic factors for cure. These were: lymph node response to irradiation (P= 0.0000), size of node (P= 0.0000), radiotherapy dose (P= 0.0037), condition of the primary (controlled vs non-controlled) (P= 0.0015), recurrent cervical metastases post-surgery (P= 0.0286).     

The impact of prelaryngeal node metastases on early glottic cancer treatment results

The aim of the study was to examine the impact of positive prelaryngeal node on the outcome of early glottic cancer and to compare the rate of local and regional recurrences and overall survival rates in patients with positive and negative Delphian node (DN). In the years 1989–2008, a consecutive cohort of 212 patients with T1b and T2 glottic cancer with anterior commisure involvement were treated by means of supracricoid partial reconstructive laryngectomies. No adjuvant radiotherapy was administered. Out of 212 patients, in 75 suspected prelaryngeal tissue was found, harvested and separately sent for histological investigation (16-thyroid, 11-fat, 14-connective tissue, 34-lymph nodes). In 137 remaining cases, there were only muscles and fascia without even a small amount of tissue to be taken. In 16 cases out of the whole group, metastases were found. 33 patients with positive ultrasound findings on the lateral neck underwent selective neck dissection; in 4 cases metastases were confirmed. Local and regional recurrence developed in 37 out of 212 patients (17.5%). There was significant correlation between local relapse and prelaryngeal node metastases; out of 20 cases with local recurrence, 13 had positive DN (P < 0.005). There was also significant correlation between nodal relapses and DN metastases; out of 22 cases with nodal relapse, 12 had positive DN and 10 were DN negative (P < 0.005). The organ preservation rates for DN positive and DN negative patients were 62.5 and 93.88%, respectively. There was noted a significant difference in the mean survival between the groups with positive and negative DN (P = 0.004; 38.7 vs. 49.3 months, respectively). In conclusion, positive DN seems to be a strong isolated factor influencing prognosis in patients with early glottic cancer. DN metastases are responsible for the increased rates of local and nodal relapses, decreased chances of organ preservation and poor overall survival rates.     

Survivin expression is significantly higher in pN+ oral and oropharyngeal primary squamous cell carcinomas than in pN0 carcinomas

Conclusions. The preliminary results reported here suggest that survivin expression in primary oral and oropharyngeal squamous cell carcinomas (SCCs) may identify patients at risk of disease disseminating to neck lymph nodes. If these results are confirmed in larger series of patients it may imply that elective neck dissection should be considered in clinically N0 patients with oral and oropharyngeal SCCs who show high expression of survivin. Objective. To investigate the expression of survivin, a member of the inhibitor of apoptosis proteins family, in patients with primary oral and oropharyngeal SCCs with and without neck lymph node metastases. Material and methods. We considered 13 consecutive cases of oral and oropharyngeal SCCs with lymph node metastases (pN?+?) and 13 cases of pN0 oral and oropharyngeal SCCs. The survivin reactivity of primary SCCs and lymph node metastases was evaluated immunohistochemically. A lesion was considered positive if >9.5% of the tumour cells showed diffuse strong staining. Results. Sporadic groups of normal basal and parabasal epithelial cells showed weak survivin staining. In SCCs, a nuclear reaction predominated. Eight primary pN+ SCCs were survivin-positive (mean expression 34.7%), compared to 5 primary pN0 SCCs (mean expression 12.3%; p=0.017). Statistical analysis disclosed significantly higher survivin expression in primary oral and oropharyngeal SCCs that developed distant non-lymphatic metastases (p=0.012).     

Diagnostic work-up and outcome of cervical metastases from an unknown primary

Conclusions. An intensive diagnostic work-up including ¹⁸F-fluorodeoxyglucose positron emission tomography (FDG-PET) detects many unknown primary tumours, leads to a low emergence rate of primary tumours, and selects carcinoma of unknown primary with much more favourable results after neck dissection and postoperative radiotherapy. Objective. To investigate the optimal diagnostic approach and best treatment modality for rare head and neck cancer of unknown primary. Patients and methods. In a retrospective study, 69 patients admitted from 1987 to 2002 with cervical lymph node metastases without apparent primary were reviewed. Test characteristics of all diagnostic procedures were calculated. Disease-free and overall survival rates were calculated. Major prognostic factors were analysed univariately. Results. At the primary site FDG-PET showed the best sensitivity with 69% and the highest negative predictive value with 87%. Computed tomography and magnetic resonance imaging had a better specificity with 87% and 95%, respectively. The primary tumour was detected in 23 cases (33%). Frequent primary tumour origin was the palatine tonsil (n=8, 35%), base of the tongue (n=6, 26%) and lung (n=4, 17%). All patients with unknown primary were treated by neck dissection. Adjuvant radiotherapy was performed in 26 patients (57%), concurrent radiochemotherapy was performed in 12 patients (26%). The primary emergence rate was 7%. The 5-year overall survival rate was inferior in patients with detected primary in comparison with patients with unknown primary (22% versus 52%). Significant prognostic factors in case of unknown primary were M stage, smoking, alcohol consumption and tonsillectomy. Radiotherapy but not chemotherapy with carboplatin influenced the overall survival.     

Prognostic value of histopathological parameters in 51 supraglottic squamous cell carcinomas

The prognosis for supraglottic squamous cell carcinomas is difficult to establish. A study of 19 histopathological parameters (14 from the primary tumor and 5 from the lymph node metastases) was undertaken in 51 supraglottic epidermoid carcinomas. All patients were subjected to supraglottic horizontal laryngectomy with bilateral functional neck dissection and postoperative radiotherapy. A discriminant analysis was performed to relate the different histopathological variables to 5-year survival. Capsule rupture and the number of metastatic lymph nodes were the only two parameters significantly related to survival (P <.01). Histopathology (in particular, lymph node metastases) proved more effective in predicting patient evolution than TNM staging.     

Treatment of oropharyngeal carcinoma by irradiation or by surgery

Of previously untreated patients with squamous cell carcinoma of the oropharynx, 145 are reviewed in this study. All were treated in the Department of Head and Neck Surgery at the University of Liverpool from 1990 to 1997. Seventy-seven patients were treated with irradiation, 28 patients by surgery and 40 patients were deemed not suitable for any curative treatment. Univariate analysis showed no difference in the two groups treated by curative modalities but multivariate analysis did suggest that the surgical group tended to have larger neck node metastases. The 5-year tumour specific actuarial survival for all patients was 53%, 65% for the radiotherapy group and 51% for the surgery group. The difference was not statistically significant (χ²₁ = 1.5070). The modality of treatment had no affect on either the development of a primary or neck node recurrence or the survival after such a recurrence. Where neck node disease was present it was treated as appropriate. As is generally standard practice, lymph nodes over 2 cm were treated with radical neck dissection whether the patient was having irradiation therapy or surgery. If the patient was having irradiation therapy, the neck dissection was carried out before and irradiation after operation, both on the primary and on the neck, if appropriate. It is concluded that irradiation therapy in properly selected cases in combined head and neck clinics is a safe and effective treatment for squamous cell carcinoma of the oropharynx. Neck node disease should be treated appropriately, but there is no support for the old adage that whatever form of treatment is being used for the neck node should also be used for the primary site.     

Distribution and impact of lymph node metastases in oropharyngeal cancer

Conclusions. In the treatment of oropharyngeal cancers, possible metastases to retropharyngeal lymph nodes (RPLNs) should be taken into account, especially in tumors arising in the lateral wall and/or posterior wall. Patients with multiple positive neck nodes must have intensified adjuvant therapy, especially when they have extracapsular spread (ECS). Objective. To develop optimal treatment strategies for oropharyngeal cancers, we retrospectively analyzed the lymph node metastases of oropharyngeal squamous cell carcinoma. Patients and methods. Between 1988 and 2003, 77 patients with previously untreated oropharyngeal squamous cell carcinoma underwent neck dissections. Results. Among the patients with tumor arising in the lateral wall or posterior wall, retropharyngeal nodes were involved in 29% (11/38), while RPLN metastasis was not observed in patients with tumors arising in the superior wall or anterior wall. The survival rate of patients with two or fewer positive lymph nodes was significantly better than that of patients with three or more positive lymph nodes (p<0.05). The survival rate of the patients who had ECS was significantly worse than that of the patients who had lymph node metastases but not ECS (p<0.05). There was no significant difference between the survival rates of the patients with and without RPLN metastases.     

Occurrence of lymph node metastasis in early-stage parotid gland cancer

Lymph node metastasis is one of the most important factors in therapy and prognosis for patients with parotid gland cancer. Nevertheless, the extent of the primary tumor resection and the necessity of a neck dissection still is a common issue. Since little is known about lymph node metastasis in early-stage parotid gland cancer, the purpose of the present study was to evaluate the occurrence of lymph node metastases in T1 and T2 carcinomas and its impact on local control and survival. We retrospectively analyzed 70 patients with early-stage (T1 and T2) primary parotid gland cancer. All patients were treated with parotidectomy and an ipsilateral neck dissection from 1987 to 2009. Clinicopathological and survival parameters were calculated. The median follow-up time was 51.7 months. A positive pathological lymph node stage (pN+) was found in 21.4% of patients with a significant correlation to the clinical lymph node stage (cN) (p = 0.061). There were no differences in the clinical and histopathological data between pN− and pN+ patients. In 73.3% of pN+ patients, the metastases were located intraparotideal. The incidence of occult metastases (pN+/cN−) was 17.2%. Of all patients with occult metastases, 30.0% had extraparotideal lymphatic spread. A positive lymph node stage significantly indicated a poorer 5-year overall as well as 5-year disease-free survival rate compared to pN− patients (p = 0.048; p = 0.011). We propose total parotidectomy in combination with at least a level II–III selective neck dissection in any case of early-stage parotid gland cancer.     

Implications of tumour in resection margins following surgical treatment of squamous cell carcinoma of the head and neck

Presence of tumour at the resection margin following primary surgical treatment for squamous cell carcinoma of the head and neck is thought to adversely affect prognosis. To confirm this we performed a review of 478 patients treated by primary surgery for squamous cell carcinoma of the head and neck and sub-divided them into those exhibiting postive margins and those with negative margins following resection. Uni-variate and multi-variate statistical methods were used to analyse survival figures and a variety of parameters associated with the presence of positive resection margins. We found 5-year survival was decreased if resection margins were found to be positive (P < 0.025). The presence of positive resection margins was also significantly associated with time to tumour recurrence (P < 0.001) and survival with nodal recurrence (P < 0.001). Other factors which were significantly associated with survival using Cox's multi-variate analysis were site of tumour (P < 0.005), nodal extracapsular rupture (P < 0.05) and pathological T-stage (P < 0.05). Uni-variate analysis revealed no significant associations betweent the presence of positive margins and the patient's age, sex, tumour site, degree of tumour differentiation, and nodal status, though using multiple logistic regression, the general condition of the patient (P < 0.01) and the tumour site P < 0.05) were significantly related. The results support the concept that every effort should be made to obtain negative resection margins when undertaking primary ablative surgery for squamous cell carcinoma of the head and neck.     

Risk of nodal metastases from malignant salivary gland tumors related to tumor size and grade of malignancy

Indications for elective treatment of the neck are not well defined in salivary gland tumors. We retrospectively reviewed 153 cases of malignant salivary gland tumors treated from 1965 to 1985 at the Hospital de Oncología, Mexico City. There were 106 parotid cancers, 26 in the submandibular gland and 21 in minor salivary glands. Median follow-up was 48 months. In T1-2 tumors there was a 12% incidence of nodal metastases as compared with 27% in T3-4 cancers (P = 0.01). Thirty-six elective neck dissections were performed. Patients with high-grade tumors had an increased risk (50%) of occult node metastases, while no cases were found in low-grade carcinomas (P < 0.05). The risk of neck recurrence was higher in N+ (23.5%) than in NO patients (3.2%). The 5-year actuarial survival was significantly better in lowgrade tumors (78%), T1-2 tumors (85%) and negative nodes (63%) than in high-grade neoplasms (49%), T3-4 tumors (35%) and positive nodes (36%) (P = 0.001, P = 0.001 and P = 0.04, respectively).Presented at the combined meeting of the Society of Head and Neck Surgeons and the European Organization for Research and Treatment of Cancer, Paris, France, 25–28 May 1994     

THE LEVEL OF CERVICAL LYMPH NODE METASTASES: THEIR PROGNOSTIC RELEVANCE AND RELATIONSHIP WITH HEAD AND NECK SQUAMOUS CARCINOMA PRIMARY SITES

It would seem logical that patients with nodal metastases low in the neck would fare less well than patients with disease high in the neck. The penultimate UICC classification suggested that neck node level was important although there was no mention of this in the most recent classification. In addition, patients with carcinomas at the various sites would be expected to have different patterns of nodal involvement. Of 3419 patients with head and neck squamous carcinoma on the Liverpool University Head and Neck Unit database, 947 had neck node metastases. The neck node levels were coded as (I)sub-mandibular, (II) above the thyroid notch, (III) below the thyroid notch and (IV) supra-clavicular/posterior triangle nodes. Levels II and III contained the deep jugular chain. The relationship between node level and site and sub-site and survival were analysed with particular emphasis on multivariate methods. The 5-year survival for the whole group was 51% and survival fell with decreasing node level (I-IV) being 37% for sub-mandibular nodes, 32% for deep cervical nodes and 25% for lower deep cervical nodes. The 18-month survival for supra-clavicular and posterior triangle nodes was 21%. The difference in survival was significant (x²₃= 24.42, P < 0.001). Multivariate analysis confirmed that as the level of the nodes fell from the sub-mandibular refion to the supra-clavicular region the prognosis worsened (estimate = -0.3378, P = 0.0003). Level II (upper deep cervical) nodes were the most commonly involved with regards to all primary sites and formed 69% of all neck node metastases. Over three quarters of laryngeal oropharyngeal and hypopharyngeal metastases went ot this level whereas only 47% of oral cancers did. Most of the remainder of these latter lesions metastasized to level I (42%). These findings were confirmed by multiple logistic regression. When studying survival for lymph node level with regard to site all sites had a reducing prognosis with decrasing node level except for larynx. Multiple linear regression showed an association between decreasing node level and increasing N-stage (P = 0.001) with increasing T-stage (P = 0.0014) and as the site moved from the mouth to the larynx (P = 0.0047). The present data support the view that neck node level is important as regards prognosis for most sites in the head and neck. The data confirm the clinical view that deep cervical nodes are most frequently affected by head and neck cancer with level IV nodes being unusual and clinically tending to herald a non head and neck tumour and that level III nodes are relatively uncommon. This is surprising as one would expect at least a proportion of laryngeal carcinomas and quite a high proportion of lhypopharyngeal carcinomas to metastasize to this region.