Laparoscopic pelvic lymph node dissection for prostate cancer: comparison of the extended and modified techniques

PURPOSE: We compared the results of extended (obturator, hypogastric, common and external iliac nodes) to modified (obturator and hypogastric nodes only) laparoscopic pelvic lymph node dissection in patients with clinically localized prostate cancer. MATERIALS AND METHODS: A total of 189 patients with stage T1 to T3 prostate cancer underwent modified (150) or extended (39) laparoscopic pelvic lymph node dissection for pelvic nodal assessment before definitive treatment. RESULTS: Twice as many lymph nodes were removed via extended than modified laparoscopic pelvic lymph node dissection (mean 17:8 versus 9.3). The overall positivity rate was 23 of 189 lymph nodes (12.2%), including 14 of 150 (7.3%) for modified and 9 of 39 (23.1%) for extended dissection (p = 0.02). Two patients (22%) who underwent extended dissection had positive lymph nodes in the external iliac area. Patients who presented with the high risk features of prostate specific antigen (PSA) greater than 20 ng./ml., Gleason score 7 or greater, or stage T2b disease or greater had a 26.5% (p = 0.0002), 22% (p = 0.0006) or 16.4% (p = 0.003) likelihood of positive lymph nodes, respectively. For extended versus modified laparoscopic pelvic lymph node dissection node positivity in high risk patients was 27% versus 18.8% (p = 0.4), 30 versus 26.4% (p = 0.8) and 25.4 versus 14.6% (p = 0.17) for Gleason score 7 or greater, PSA greater than 20 ng./ml. and disease stage T2b to T3a, respectively. Patients who underwent the extended procedure had a higher complication rate (35.9 versus 2%, p < 0.0001). No laparotomy was required. CONCLUSIONS: Despite yielding a 2-fold higher node count and higher node positivity rate, extended laparoscopic pelvic lymph node dissection offers no advantage over modified laparoscopic pelvic lymph node dissection for diagnosing positive lymph nodes when results are analyzed by prognostic factors. The extended procedure is associated with a much higher complication rate. In patients with the high risk features of PSA greater than 20 ng./ml., Gleason score 7 or greater and stage T2b to T3a disease modified laparoscopic pelvic lymph node dissection can be performed safely and effectively to help identify those who may benefit most from curative therapy.     

Inguinal lymphadenectomy in cancer of the penis: surgical techniques and indications

Lymph node invasion is one of the major prognostic factors of cancer of the penis. However, as it is difficult to evaluate clinically and by means of complementary investigations, inguinal or even ilioinguinal lymph node dissection is still indicated. As this surgery carries a certain morbidity (necrosis of skin edges, infection, lymphorrhoea and subsequent lymphoedema), the indications are presented according to the presence or absence of palpable inguinal lymph nodes and the stage of the primary tumour. Various surgical techniques are proposed: Superficial and deep inguinal lymph node dissection in the case of mobile and palpable inguinal nodes, simplified and superficial inguinal lymph node dissection in the absence of palpable inguinal nodes and in the case of invasive primary tumour.     

Treatment of the regional lymph nodes in squamous carcinoma of the penis

OBJECTIVES: We evaluated the management of the regional lymph nodes to determine the appropriate treatment for carcinoma of the penis. METHODS: The records of 36 patients with carcinoma of the penis were reviewed. Lymphadenectomy was performed in 18 patients, 17 were managed conservatively (watchful waiting) and 1 patient had a biopsy and received radiotherapy. RESULTS: Positive nodes were found in 2 of 2 pT4, 2 of 3 pT3, 8 of 13 pT2 and 2 of 12 pT1 patients submitted to lymphadenectomy. Concerning the histological grade, positive nodes were found in all of the 4 G3, 5 of 12 G2 and 3 of 20 G1 patients. The survival rate was 100% for the patients with negative lymph nodes (pNO = 6) or a single positive inguinal lymph node (pN1 = 5). A correlation was found between the T and the histological grade and the likelihood of lymph node invasion. CONCLUSIONS: The T and the histological grade of the primary lesion must be considered when deciding the approach in the management of the lymph nodes as unnecessary lymphadenectomy can be avoided and those at high risk of lymph node invasion can be treated radically and timely.     

A comparison between lymphangiography and pelvic node dissection in the staging of prostatic cancer

On 40 consecutive patients with prostatic cancer who had pedal lymphangiography during the initial evaluation and, subsequently, underwent pelvic node dissection or biopsy, a surprisingly high number had falsely positive (59 per cent) or negative (36 per cent) x-ray findings. Initially the tumors were considered clinically to be stage B in 24 cases, stage C in 13 and stage D in 3. After lymph node dissection only 17 tumors were still considered to be stage B and 7 were stage C, while 16 tumors were actually stage D. This surgical staging is important for the further management of the patient as well as the prognosis, Pedal lymphangiogrpahy alone is unreliable for accurate assessment of the regional lymph node status in clinically localized prostatic cancer.     

Surgical treatment of gastric adenocarcinoma: the role of lymph node dissection

In this retrospective study, 652 patients who had curative resections for gastric cancer from 1977 to 1991 were reviewed to evaluate improvements in gastric cancer surgery and the influence of the extent of lymphadenectomy on survival. The patients were grouped into three time periods: 1977 to 1981, 1982 to 1986 and 1987 to 1991. The percentage of patients with early gastric cancer increased from 17.7% during 1977 to 1981, to 24.3% during 1987 to 1991. The average number of dissected lymph nodes was 7.5 +/- 8.1 during 1977 to 1981 and 16.4 +/- 10.3 during 1987 to 1991, when more radical lymphadenectomy was adopted. Total gastrectomies increased from 10.9% to 25.9% in the same time periods while combined visceral resections increased from 26.7% to 38.1%. Operative mortality decreased from 5.0% to 1.7%. The overall 5-year survival rate increased from 34.8% to 59.4%. In subgroup analysis, significant improvement of the 5-year survival rate was noted in the following groups: patients with stage I, II and III tumors but not stage IV; both proximal and distally located tumors; tumors with or without lymph node metastases; T1 and T2 but not in T3 and T4 (cancer invasion beyond the serosa). The decreased surgical mortality in recent years suggests that curative resection with extensive lymph node dissection can now be safely performed. Radical gastrectomy with extended lymphadenectomy may be adopted in gastric cancer resection for better control of regional disease.     

Review of US medical school finances, 1995-1996

We discussed the rational extent of the lymph node dissection for carcinoma of the lower third of the esophagus of T2 or T3 stage with abdominal lymph node metastasis. Lymph node metastasis developed in 89.5% of patients. Cervical lymph node metastasis was seen in 35.8%. In the cases with positive abdominal lymph node, 40.9% of the patients had cervical node metastasis. The most frequent site of the positive node in the neck is the area along the right recurrent laryngeal nerve. On the stand point of removal of metastatic lymph node, neck dissection should be required. Three-field dissection yielded better survival rate than two-field dissection but statistical significance was not obtained. When the patients have cervical lymph node metastasis, they have greater possibility of developing blood borne metastasis. However, this observation does not deny the validity of the three-field dissection. Because this dissection may help reducing nodal spread and nodal recurrence. We have to wait for accumulation of the patients to analyze the definite extent of node dissection for T2 or T3 stage of carcinoma of the lower third of the esophagus with positive abdominal lymph node.     

An evaluation of lymphography in localized carcinoma of the prostate

Lymph node biopsies were positive in 20% (7/35) of stage T1 and T2 (stage B) tumors and 64% (21/33) of stage T3 (stage C) tumors in 69 previously untreated and unselected patients with apparently localized carcinoma of the prostate. One patient with a To (stage A) tumor had no evidence of lymph node metastasis. Prospective analysis demonstrated an overall lymphographic accuracy of 78%, sensitivity of 57% and specificity of 92%. The detection of lymph node metastases in the lymphogram is limited by the frequency of microscopic metastasis and the frequency of benign changes within pelvic lymph nodes in this older patient population. Diagnostic criteria for metastatic disease which gives a low incidence of false-positive interpretations should be maintained, since relaxing the criteria will not necessarily improve the detection rate of metastases and would decrease specificity.     

Lymph node involvement and tumor depth in rectal cancers: an analysis of 805 patients

BACKGROUND: Superficial rectal tumors are said to involve regional lymph nodes rarely. This presumption must be proven beyond any doubt if less radical surgery is to be offered for such patients. PATIENTS AND METHODS: Eight hundred five cases (467 males; median age, 64 (range, 19-97) years) of rectal cancer were reviewed. RESULTS: Lymph node positivity, number of lymph nodes involved, lymphatic vessel, and venous and perineural invasion were significantly increased with increasing depth of invasion of tumor through the bowel wall in univariate analysis. The percentage of lymph node involvement at each tumor depth was as follows: T1, 5.7 percent; T2, 19.6 percent; T3, 65.7 percent; T4, 78.8 percent. Overall lymph node involvement was 59 percent. For patients younger than 45 years of age, the percentage of lymph node involvement was 33.3, 30, 69.3, and 83.3 percent compared with 3.1, 8.4, 64.2, and 78.8 percent for patients aged 45 years or above for T1, T2, T3, and T4, respectively. CONCLUSION: Increased depths of tumor penetration beyond T1 and age less than 45 years have an excessive incidence of lymph node positivity. The finding of lymphatic vessel invasion on biopsy is highly indicative of lymph node metastasis.     

Systematic sextant biopsies improve preoperative prediction of pelvic lymph node metastases in patients with clinically localized prostatic carcinoma

PURPOSE: An algorithm including the results of systematic sextant biopsies was statistically developed and evaluated to predict the probability of pelvic lymph node metastases in patients with clinically localized carcinoma of the prostate. MATERIALS AND METHODS: Clinical stage, serum prostate specific antigen concentration, Gleason score, number of positive biopsies, number of biopsies containing any Gleason grade 4 or 5 cancer and number of biopsies predominated by Gleason grade 4 or 5 cancer were recorded in 345 patients undergoing pelvic lymph node dissection and correlated with the incidence of lymph node metastases. Multivariate logistic regression, and classification and regression trees analyses were performed. RESULTS: In univariate analysis all variables had a statistically significant influence on lymph node status. Logistic regression showed that the amount and distribution of undifferentiated Gleason grade 4 and 5 cancer in the biopsies were the best predictors of lymphatic spread followed by serum prostate specific antigen. Classification and regression trees analysis classified 79.9% of patients who had 3 or fewer biopsies with Gleason grade 4 or 5 cancer and no biopsies predominated by undifferentiated cancer as a low risk group. In this group positive lymph nodes occurred in only 2.2% (95% confidence interval 0.8 to 4.7%). CONCLUSIONS: Including the results of systematic sextant biopsies substantially enhances the predictive accuracy of algorithms that define the probability of lymph node metastases in prostatic cancer. Patients thus defined as having no lymphatic spread could potentially be spared pelvic lymph node dissection before definitive local treatment.     

Para-aortic lymph node metastasis in carcinoma of the distal bile duct

BACKGROUND/AIMS: Lymph node dissection plays an important role in radical surgery for pancreaticoduodenal carcinomas. The aim of this study was to identify the critical areas of lymph node dissection in carcinoma of the distal bile duct. METHODOLOGY: Between January 1995 and December 1996, 20 consecutive patients with distal bile duct cancer underwent pancreaticoduodenectomy with extended lymph node dissection (including the para-aortic nodes). Histopathologic findings were examined with special reference to lymph node metastasis. RESULTS: Histological evidence of lymph node metastasis was found in 11 patients (55%). The areas with frequent metastases were the posterior pancreaticoduodenal lymph nodes (35%), and the nodes around the hepatoduodenal ligament (35%) and around the common hepatic artery (30%). Para-aortic lymph node involvement was identified in 5 patients (25%). Most of these existed in the inter-aorticocaval space. Pancreatic parenchymal invasion was present in 10 patients. Half of the patients with pancreatic invasion had para-aortic nodal involvement. Para-aortic lymph node metastasis was significantly associated with pancreatic parenchymal invasion (p<0.05). CONCLUSIONS: In carcinoma of the distal bile duct with pancreatic parenchymal invasion, extended lymph node dissection (including para-aortic nodes) should be undertaken because of the relatively high incidence of metastasis.     

Outpatient pelvic lymph node dissection

PURPOSE: Staging pelvic lymph node dissection is an important part of the evaluation of most patients with prostatic carcinoma. While laparoscopic pelvic lymph node dissection provides an alternative to standard pelvic lymph node dissection, it has been associated with a significant learning curve, high major complication rate, frequent hospitalization and greater expense. We sought to modify the technique of pelvic lymph node dissection to allow its performance as an outpatient procedure. MATERIALS AND METHODS: Pelvic lymph node dissection was performed through bilateral 3 cm. incisions overlying the obturator fossa in 11 patients. RESULTS: Nodes sampled ranged from 5 to 26 per patient that differed insignificantly from those undergoing standard pelvic lymph node dissection. Four patients had lymph node metastases. Nine procedures were performed entirely on an outpatient basis. One complication (external vein injury) was repaired with a single suture after extending the incision to 6 cm. CONCLUSIONS: Outpatient pelvic lymph node dissection through small incisions provides an attractive alternative to standard or laparoscopic lymph node dissection. With further experience it may become the procedure of choice for staging prostatic carcinoma in patients with a high risk of pelvic lymph node metastases.     

Lymphatogenous spread of renal cell carcinoma: an autopsy study

PURPOSE: We investigated the occurrence and extent of metastatic spread, especially regarding lymph nodes, of renal cell carcinoma. MATERIALS AND METHODS: From 1958 to 1982, 554 cases of renal cell carcinoma were diagnosed at autopsy. Clinical data and autopsy findings were reevaluated, and the occurrence of lymph node metastases was analyzed by histological examination of retroperitoneal, mediastinal, supraclavicular, axillary and inguinal lymph nodes. RESULTS: Distant metastases were revealed in 119 cases (21.5%), including 31 (5.6%) with single metastases. In 88 cases (16%) renal cancer was the cause of death. Lymphatogenous dissemination was detected in 80 cases of which 75 had additional, mostly multifocal metastatic spread. Consequently lymph node metastases restricted to the paracaval and/or para-aortic lymph nodes were noted in only 5 cases (0.9%). CONCLUSIONS: Of the 554 cases of clinically unrecognized renal cell carcinoma almost all with lymphatic spread had additional distant metastases. Therefore, the therapeutic effect of extensive retroperitoneal lymph node dissection in association with radical nephrectomy seems to be low. However, more limited lymph node dissection may be useful, mainly as a staging procedure.     

Postoperative radiation for cervical cancer with pathologic risk factors

PURPOSE: To examine the efficacy of postoperative radiation therapy for early-stage cervical cancer with pathologic risk factors. METHODS AND MATERIALS: We reviewed the charts of 83 patients who received postoperative radiation therapy at our facility from March 1980 to November 1993 for early stage cervix cancer with positive surgical margins, positive pelvic or periaortic lymph nodes, lymphovascular space invasion, deep invasion, or for disease discovered incidently at simple hysterectomy. Twenty-eight patients received low dose rate (LDR) intracavitary radiation with or without external beam pelvic irradiation and 55 patients received external beam pelvic irradiation with high dose rate (HDR) intracavitary implants. Of these 83 patients, 66 were evaluable--20 LDR and 46 HDR patients. All patients received 45-50 Gy external beam irradiation and 20 Gy LDR equivalent intracavitary irradiation prescribed to 0.5 cm below the mucosa. Ninety percent of the LDR group and 92% of the HDR group completed treatment within < 56 days. Treatment-related toxicities were scored according to the GOG toxicity scale. Mean and median follow-up times were 101 months and 111 months (3-172 months) for the LDR group and 42 and 40 months (3-98 months) for the HDR group. RESULTS: The 5-year disease-free survival was 89% for the LDR group and 72% for the HDR group. Local control was observed in 90% (18 out of 20) of the LDR patients and 89% (41 out of 46) of the HDR patients for an overall local control rate of 89.5%. Two of 20 LDR patients (10%) experienced recurrence (two pelvic with distant metastasis). Nine of 46 HDR patients (22%) had recurrence of disease (three pelvic, four distant metastasis, and two pelvic with distant metastasis). In the HDR group, 6 out of 16 (38%) with positive lymph nodes died of disease whereas, 27 out of 30 (90%) of the patients with negative lymph nodes remain free of disease. Three of 20 (15%) LDR patients and 4 out of 46 (9%) HDR patients experienced Grade 2 or 3 late treatment- related complications. No patient in either group had Grade 4 or 5 complications. Pathologic risk factors were analyzed. Lymph node positivity and lymphovascular space invasion were found to be significant (p = 0.01 and p = 0.02). Positive margins, deep invasion, and age were not significant. CONCLUSION: Our results demonstrate the efficacy of postoperative irradiation for cervical cancer with pathologic risk factors. Overall, the local control rate was 89.5% The HDR results demonstrate that this method can be delivered safely and effectively.     

Radical vulvectomy with postoperative irradiation for vulvar cancer: therapeutic implications of a central block

PURPOSE/OBJECTIVE: To report the long-term results of vulvectomy, node dissection, and postoperative nodal irradiation using a midline vulvar block in patients with node positive vulvar cancer. METHODS AND MATERIALS: From 1971 through 1992, 27 patients with carcinoma of the vulva and histologically involved inguinal lymph nodes were treated postoperatively with radiation therapy after radical vulvectomy and bilateral lymphadenectomy (n = 25), radical vulvectomy and unilateral lymphadenectomy (n = 1), or hemivulvectomy and bilateral lymphadenectomy (n = 1). Federation Internationale de Gynecologic et d'Obstetrique stages were III (n = 14), IVA (n = 8), and IVB (n = 5) squamous cell carcinoma. Inguinal lymph nodes were involved with tumor in all patients (average number positive = 4, range 1-15). Postoperative irradiation was directed at the bilateral groin and pelvic nodes (n = 19), unilateral groin and pelvic nodes (n = 6), or unilateral groin only (n = 1). These 26 patients had the midline blocked. In addition, one patient received irradiation to the entire pelvis and perineum. Doses ranged from 10.8 to 50.7 Gy (median 45.5) with all patients except 1 receiving > or = 42.0 Gy. RESULTS: Actuarial 5-year overall survival and disease-free survival estimates were 40% and 35%, respectively. Recurrences developed in 63% (17/27) of the patients at a median of 9 months from surgery (range 3 months to 6 years) and 15 of these have died; two patients with recurrences are surviving at 24 and 96 months after further surgery and radiation therapy. Central recurrences (under the midline block) were present in 13 of these 17 patients (76%), either as central only (n = 8), central and regional (n = 4), or central and distant (n = 1). Additionally, three patients developed regional recurrences and one patient developed a concurrent regional and distant relapse. One patient developed a squamous cell cancer of the anus under the midline block 54 months after the initial vulvar cancer and an additional patient developed transitional cell carcinoma of the ureter (outside the radiation field) 12 months after diagnosis. Factors associated with a decreased relapse-free survival included increasing Federation Internationale de Gynecologic et d'Obstetrique stage (p = 0.01) and invasion of the tumor into the subcutaneous (SC) fat or deep soft tissue (p = 0.05). Chronic lower extremity edema developed in four patients, but there have been no other complications. CONCLUSIONS: Radical vulvectomy has often been considered sufficient central treatment for vulvar carcinoma, with postoperative irradiation directed only to the nodes. Although designed to protect the radiosensitive vulva, use of a midline block in this series resulted in a 48% (13/27) central recurrence rate, much higher than the 8.5% rate previously reported with this technique. Routine use of the midline block should be abandoned and, instead, postoperative irradiation volumes should be tailored to the individual patient.     

Frank I. Marcus

A series of 60 patients with squamous cell carcinoma of the larynx who underwent surgery of the primary tumor and elective bilateral neck dissection were reviewed to determine the importance of certain clinical and pathological features as risk factors for lymph node metastasis. Tumor location, extension to the vallecula and to the pyriform sinus, T stage, histological grade, palpable lymph nodes, laryngeal motility, and macroscopic aspects were studied. Logistic regression analysis demonstrated that T stage, tumor location, and palpable lymph nodes were the most important predictors of lymph node metastases.     

Pelvic lymphadenectomy can be omitted in selected patients with carcinoma of the prostate: development of a system of patient selection

OBJECTIVES: The prevalence of pelvic lymph node metastases in men with clinically localized prostate cancer has decreased dramatically over the past decade, possibly due to efforts at early detection. With a significantly lower incidence of pelvic node involvement, it may be possible to identify a segment of patients for whom pelvic lymph node dissection (PLND) may be omitted. This study was conducted to develop a method to select patients for whom PLND could be omitted. METHODS: We analyzed serum prostate-specific antigen (PSA), clinical stage, biopsy Gleason score, and final pathologic stage in 481 men with clinically localized prostate cancer. These variables were compared to the risk of positive pelvic lymph nodes. RESULTS: Logistic regression analysis determined that combining all three variables provided the best determination of final pathologic stage. A series of probability curves have been created to estimate the risk of positive lymph nodes in a given patient. Based on the distribution of patients in this study and using these probability functions, PLND could be avoided in up to 50% of patients with localized prostate cancer diagnosed by contemporary methods. CONCLUSIONS: In properly selected patients, pelvic lymphadenectomy can be omitted in the staging and treatment of localized prostate cancer.     

Results of radical surgical procedures after radiation for treatment of invasive carcinoma of the uterine cervix in a private practice

An analysis is made of the results of treatment of 96 women with carcinoma of the cervix, Stages IB and II, in a private practice. All 96 women were treated preoperatively with uterine intracavitary radium, followed 6 weeks later by Wertheim hysterectomy with pelvic lymphadenectomy. If malignant tumor was present in the lateral pelvic lymph nodes, external radiation was given postoperatively. The over-all survival rates were: Stage IB, 88% and 84% at 5 and 10 years; Stage II, 72% and 62% at 5 and 10 years. Regardless of the clinical stage, the highest survival rates were found in those patients who had no malignancy in the lateral pelvic lymph nodes and no residual cervical carcinoma. The lowest survival rates were found in those patients who had both residual cervical carcinoma and lymph node metastases.     

Paraaortic lymphadenectomy in gynecologic malignancies confined to the pelvis

Eighty-six patients with pelvic malignancies had paraaortic node dissection; 30% had positive paraaortic lymph nodes. At the time of laparotomy, 77 patients had malignancies apparently confined to the pelvis after extensive work-up and histologic examination of tissue biopsies outside the pelvis; 20 (26%) had metastatic cancer in the paraaortic lymph nodes. Thirteen of 49 patients with cancer of the cervix, 5 of 18 with uterine cancer, and 2 of 10 with ovarian cancer had positive paraaortic lymph nodes. The value of paraaortic dissection in patients with pelvic malignancies apparently confined to the pelvis is discussed.     

Lyme borreliosis in the southern United States: a review

OBJECTIVES: To evaluate the efficacy and morbidity of surgical treatment of carcinoma of the penis. This series of patients was derived from a retrospective multicentre study (1959-1989), initially concerning 506 patients, all treatments combined. METHODS: One hundred and sixty-eight patients treated by surgery or surgery followed by external radiotherapy between 1959 to 1989 were included. The mean follow-up is 14.4 years and the mean participation is 4.7 years. Total or partial amputation was performed as first-line treatment in 89 patients (53%) and as second-line treatment in 11 patients (7%). Thirty-two patients received external radiotherapy as a complement to surgery. Inguinal lymph node dissection was performed as first-line treatment in 68 patients (41%) and secondarily in 19 patients (11%). Postoperative inguinal radiotherapy was performed in 52 patients. Monofactorial statistical analysis of prognostic factors is proposed. RESULTS: The 5-year local control rate was 84%; it was independent of the stage of the tumour; the survival of the patients dying from any cause was 53%, the progression-free survival was 69% and the survival of patients dying from cancer of the penis was 75%. The vital prognosis is statistically significant related to the lymph node status. No significant relationship was observed between lymph node status and tumour stage. 61 complications involving the penis were observed in 40 patients (24%). Thirty-seven patients developed complications secondary to the lymph node dissection (24 cases of oedema of the lower limbs and 13 cases of inguinal sclerosis). CONCLUSION: Surgical treatment ensured a local control globally comparable to that obtained with brachytherapy, but unlike this technique, the local control is independent of the stage. Brachytherapy should not always be performed in favour of mutilation for advanced tumours because the iatrogenic effects of brachytherapy are increased in these cases. An approach to the treatment of lymph nodes is proposed based on the authors' experience and the data of the literature.     

Irradiation of clinically uninvolved cervical lymph nodes

Of 402 patients with cancers of the oral cavity, oropharynx, and supraglottic larynx treated at Stanford between 1957 and 1972, 164 had clinically uninvolved cervical lymph nodes prior to the initiation of radiation therapy. Lymph node metastases developed later in 38 per cent of patients with primary oral cavity carcinomas who were treated with interstitial radium implants alone. No late cervical lymph node involvement was found in those patients who received high dose external irradiation to at least the primary site and first echelon lymph nodes. Lymph node failures were ultimately noted in 20 of the 140 patients (14 per cent), who received partial or complete neck irradiation, but 18 of these occurred in patients with uncontrolled primary lesions, suggesting that re-seeding of cervical lymph nodes had taken place rather than failure of the initial irradiation to control subclinical metastases. Our present policy is to treat the primary lesion and adjacent lymph nodes with high dose megavoltage techniques, combined with interstitial irradiation if possible. Bilateral supplemental inferior neck radiation ports are added for patients with advanced primary neoplasms and for those with clinically involved cervical lymph nodes. All other patients undergoing radiation therapy for stage T1 primary lesions and clinically negative necks also receive ipsilateral low neck irradiation. In addition, cervical lymph nodes are electively irradiated when the primary lesion has been resected. When these policies are adopted, the incidence of cervical lymph node failures is extremely low in patients whose primary sites remain controlled, and morbidity from the cervical radiation fields is negligible.