Correlation between preoperative lymphoscintigraphy and metastatic nodal disease sites in 362 patients with cutaneous melanomas of the head and neck

OBJECTIVE: Lymphoscintigraphy for head and neck melanomas demonstrates a wide variation in lymphatic drainage pathways, and sentinel nodes (SNs) are reported in sites that are not clinically predicted (discordant). To assess the clinical relevance of these discordant node fields, the lymphoscintigrams of patients with head and neck melanomas were analyzed and correlated with the sites of metastatic nodal disease. METHODS: In 362 patients with head and neck melanomas who underwent lymphoscintigraphy, the locations of the SNs were compared with the locations of the primary tumors. The SNs were removed and examined in 136 patients and an elective or therapeutic regional lymph node dissection was performed in 40 patients. RESULTS: Lymphoscintigraphy identified a total of 918 SNs (mean 2.5 per patient). One or more SNs was located in a discordant site in 114 patients (31.5%). Lymph node metastases developed in 16 patients with nonoperated SNs, all underneath the tattoo spots on the skin used to mark the position of the SNs. In 14 patients SN biopsy revealed metastatic melanoma. After a negative SN biopsy procedure 11 patients developed regional lymph node metastases during follow-up. Elective and therapeutic neck dissections demonstrated 10 patients with nodal metastases, all located in predicted node fields. Of the 51 patients with involved lymph nodes, 7 had positive nodes in discordant sites (13.7%). CONCLUSIONS: Metastases from head and neck melanomas can occur in any SN demonstrated by lymphoscintigraphy. SNs in discordant as well as predicted node fields should be removed and examined to optimize the accuracy of staging.     

Redefinition of cutaneous lymphatic drainage with the use of lymphoscintigraphy for malignant melanoma

Lymphoscintigraphy was performed on 82 patients with melanoma registered at the University Melanoma Clinic. From these data, precise lymphatic drainage basins could be drawn for the head, neck, shoulder, and trunk. These drawings differed significantly from the classic anatomic studies, providing a functional look at the cutaneous lymphatic drainage. This new method correlates much better with clinical experiences and demonstrates much larger areas of ambiguous drainage than previously reported. Data from the lymphoscintigrams also emphasize the individuality of cutaneous lymphatic flow. The implications of these data in planning elective node dissections for intermediate thickness melanomas are obvious, since it is estimated that up to 59% of the dissections for trunk and head and neck primary melanomas may be misdirected if based on classic anatomic studies. The data indicate that all patients with head, neck, and shoulder lesions should undergo lymphoscintigraphy to define possible drainage basins at risk for metastatic disease. Similarly, truncal lesions require scintigrams except when they are within four well-defined areas with an extremely low probability of ambiguous drainage. Lesions in these areas show very reliable and predictable drainage to a single nodal group.     

Lymphoscintigraphic drainage patterns of the auricle in healthy subjects

BACKGROUND: In lymphoscintigraphies of the head and neck, multiple injections around a tumor result in variable drainage to multiple nodal basins. We undertook this study in healthy subjects to test whether single injections at specified points in the auricle display single predictable pathways and predict visualization of parotid sentinel lymph nodes (SLNs). METHODS: Twenty-five healthy subjects were classified according to their injection points in the auricle. Each was injected bilaterally with 99mTc nanocolloid. Parotid and extraparotid lymph nodes were topographically differentiated. The procedure was repeated 1 week later. RESULTS: Lymphoscintigraphy was reproducible. Each injection revealed a single SLN. Injection site predicted parotid SLN visualization. Two lymphatic territories with parotid or extraparotid drainage were identified. CONCLUSIONS: Lymphatic territories in the auricle coincide with the vascular territories and branchial origins. Our findings contradict the notion that lymphatic drainage of the head and neck is unpredictable and variably involves multiple nodal basins.     

Usefulness of preoperative lymphoscintigraphy for the identification of sentinel lymph nodes in melanoma.

BACKGROUND: The exact role of lymphoscintigraphy (LS) in the evaluation of sentinel lymph nodes (SLNs) in melanoma is controversial. METHODS: We reviewed our experience with preoperative LS for the determination of the lymph node drainage pattern of clinically node negative primary melanomas, with attention to the rate of ambiguous drainage and the effect of previous wide local excision (WLE). RESULTS: The scans of 87 patients who underwent LS at our institution for evaluation of their primary melanomas from 1995 to the present were reviewed. Fourteen of the primary tumor sites were in the head and neck region, 41 were truncal, and 32 were in the extremities. The average tumor thickness was 2.6 mm. Nine of 14 (64%) head/neck lesions and 12 of 41 (29%) truncal lesions displayed ambiguous drainage, as compared with only 2 of 32 (6%) extremity lesions (P <0.05). Forty-one of the 87 patients (47%) had undergone previous WLE of their primary lesion prior to their LS. The number of draining basins for the WLE and the non-WLE groups were not significantly different, and at least one SLN was found for all WLE cases. CONCLUSIONS: Preoperative LS is important for the treatment planning of SLN biopsy for head/neck and truncal melanomas, but adds little additional information for extremity lesions. Lymph node drainage scans and subsequent SLN biopsies are not contraindicated in the presence of a prior WLE.     

Utility of lymphoscintigraphy in directing surgical therapy for melanomas of the head, neck, and upper thorax

Lymphoscintigraphy with technetium99m antimony sulfur colloid or technetium99m human serum albumin helped direct the surgical management of 24 patients who had melanomas of the head, neck, and upper thorax. Eighteen (75%) patients had documented lymphatic flow to other than a single adjacent predictable lymph nodal group. Nineteen patients underwent lymphadenectomy. Availability of the scan altered surgical management in nine patients (47%) who required resection of nodes in addition to resection of adjacent nodes. The discovery of metastatic disease in one patient was clearly attributable to lymphoscintigraphy. This demonstrates the unpredictable lymphatic anatomy of this region and suggests that preoperative lymphoscintigraphy may be useful in directing the surgical management of cutaneous melanomas in which lymph node dissection is planned.     

Sentinel lymph-node biopsy after previous wide local excision for melanoma.

OBJECTIVE: To document experience with sentinel lymph-node biopsy in patients who have already undergone a wide local excision for melanoma because in many centres previous wide excision has been a contraindication for sentinel lymph-node biopsy. DESIGN: A prospective cohort study. SETTING: A tertiary care academic cancer centre. PATIENTS: One hundred patients who presented with cutaneous melanoma (depth >1 mm or Clark level IV) after having undergone wide local excision of the primary lesion that was not situated in the head or neck. The follow-up was 3 years. INTERVENTIONS: Sentinel lymph-node biopsy. Patients with truncal melanoma had preoperative lymphoscintigraphy to document the nodal basins at risk. Technetium-99m sulfur colloid (0.5-1 mCi in 0.5 mL) was injected intradermally around the scar, and the sentinel lymph node was excised with the aid of a hand-held gamma detector. OUTCOME MEASURES: Accuracy of the biopsy and false-negative rates in this setting. RESULTS: Of the 100 patients, 44 had truncal and 56 had extremity lesions. The average tumour depth was 3.47 mm and 3.07 mm respectively. Thirty-one patients had a sentinel lymph node positive for melanoma metastasis. Biopsies were positive for melanoma in 18 (41%) truncal lesions and 13 (23%) extremity lesions. There were 3 (9%) false-negative sentinel lymph-node biopsies as diagnosed by clinically evident nodal disease subsequently appearing in the nodal basin subjected to biopsy. Two occurred in patients after large rotation flap closures of truncal lesions. The third patient had a subungual melanoma of the great toe. No disease was found in the 2 nodes dissected. Two of the 3 false-negative biopsy results were obtained before serial sections and immunohistochemical staining were used to examine the sentinel lymph nodes. CONCLUSIONS: Sentinel lymph-node biopsies can successfully identify clinically occult nodal metastases in patients who have had previous wide local excision of a melanoma, but the false-negative rate in patients with rotation flap closures should be taken into consideration.     

Do nodal metastases from cutaneous melanoma of the head and neck follow a clinically predictable pattern?

BACKGROUND: Potential lymphatic drainage patterns from cutaneous melanomas of the head and neck are said to be variable and frequently unpredictable. The aim of this article is to correlate the anatomic distribution of pathologically involved lymph nodes with primary melanoma sites and to compare these findings with clinically predicted patterns of metastatic spread. METHODS: A prospectively documented series of 169 patients with pathologically proven metastatic melanoma was reviewed by analyzing the clinical, operative, and pathologic records. Clinically, it was predicted that melanomas of the anterior scalp, forehead, and face could metastasize to the parotid and neck levels I-III; the coronal scalp, ear, and neck to the parotid and levels I-V; the posterior scalp to occipital nodes and levels II-V; and the lower neck to levels III-V. Minimum follow up was 2 years. RESULTS: There were 141 therapeutic (97 comprehensive, 44 selective) and 28 elective lymphadenectomies (4 comprehensive dissections, 21 selective neck dissections, and 3 cases in which parotidectomy alone was performed). Overall, there were 112 parotidectomies, 44 of which were therapeutic and 68 elective. Pathologically positive nodes involved clinically predicted nodal groups in 156 of 169 cases (92.3%). The incidence of postauricular node involvement was only 1.5% (3 cases). No patient was initially seen with contralateral metastatic disease; however, 5 patients (2.9%) failed in the contralateral neck after therapeutic dissection. In 68% of patients, metastatic disease involved the nearest nodal group, and in 59% only a single node was involved. CONCLUSIONS: Cutaneous malignant melanomas of the head and neck metastasized to clinically predicted nodal groups in 92% of patients in this series. Postauricular and contralateral metastatic node involvement was uncommon.     

The role of cutaneous lymphoscintigraphy in determining regional lymph node drainage of truncal melanomas.

In the absence of clinically positive regional nodes, any value of prophylactic dissection in malignant melanomas depends on accurate preoperative determination of the pathway of lymphatic drainage. We report on the use of noninvasive radionuclide lymphoscintigraphy in the determination of regional patterns of lymph node drainage in patients with melanomas. Ten patients were studied; treatment was altered by test results in 2. Eleven node groups were excised in 7 patients. There have been no metastatic melanomas found in any nodal basins not detected by lymphoscintigraphy 23 to 42 months after operation.     

The use of lymphoscintigraphy and PET in the management of head and neck melanoma.

OBJECTIVES: Lymphoscintigraphy with sentinel node dissection and 18 fluoro-2-deoxyglucose positron emission tomography (PET) are being used independently in the management of many intermediate and thick melanomas of the head and neck. We report a series of patients with melanoma of the head and neck with Breslow depths greater than 1.0 mm and clinically negative regional nodes that were evaluated prospectively with PET and lymphoscintigraphy. STUDY DESIGN AND SETTING: Between July 1, 1998 and December 30, 2000 PET scans were obtained preoperatively on 18 patients undergoing resection of head and neck melanoma. Lymphoscintigraphy and sentinel node dissection was performed. Resection of the primary lesion was then carried out with adequate margins and the defects were reconstructed. RESULTS: Sentinel node(s) were found in 17/18 patients (94.4%); 5/18 (27.8%) of cases had metastases. PET detected nodal metastasis preoperatively in 3 patients (16.7%), one of which had a positive sentinel node dissection. CONCLUSION: PET and lymphoscintigraphy offer complimentary ways of evaluation for metastatic melanoma.     

Sentinel node biopsy for the detection of head and neck melanoma: a review

Worldwide incidence of malignant melanoma is on the rise. Early detection of this malignancy is key to survival, and in the case of more advanced malignancy, early and effective detection of micrometastatic disease is crucial for staging and therapy. Because melanoma spreads primarily via lymphatic drainage patterns, effective methods for tracing these pathways are of paramount importance. The authors summarize the efficacy of blue dye, gamma probe, and lymphoscintigraphy detection methods, both individually and combined; the "missed disease" (or false-negative) rate; and the clinical discordance between expected and actual location of metastatic disease in head and neck melanoma. A clinical meta-analysis of current studies in head and neck melanoma was used to evaluate clinical data. A success rate of 95% to 100% for detection of sentinel lymph nodes can be achieved when blue dye, gamma probe, and lymphoscintigraphy techniques are combined. This is associated with a false-negative rate of 7.7% to 10.4%. With respect to intermediate-depth melanomas of the head and neck, a significant discordance exists between expected and actual lymphatic drainage patterns. This problem is best addressed using a combination of lymphoscintigraphy, blue dye, and gamma probe localization, which yields a success rate of 95% to 100% for detection of sentinel lymph nodes and a low false-negative rate of 7.7% to 10.4%. In the instance of a failed study, one in which sentinel nodes are not detected by the aforementioned methods, elective node dissection is the treatment modality of choice.     

Sentinel nodes outside lymph node basins in patients with melanoma

BACKGROUND: Lymphoscintigraphy occasionally reveals hot spots outside lymph node basins in patients with melanoma. The aim of this study was to evaluate such abnormally located hot spots. METHODS: Sentinel node biopsy was studied prospectively in 379 patients with clinically localized cutaneous melanoma. One day after lymphoscintigraphy, sentinel node biopsy was performed guided by vital blue dye and a gamma ray detection probe. RESULTS: Persisting hot spots outside the regional node basins were seen in 25 patients (6.6 per cent). Several specific drainage patterns were discerned. In five patients, aberrant sentinel nodes were not explored. The hot spot represented a lymphangioma in two patients. Radioactive lymph nodes were identified in the remaining 18 patients (4.7 per cent). Four patients had metastasis in one of these aberrant lymph nodes. CONCLUSION: Sentinel nodes were found outside a lymph node basin in 5 per cent of patients. Particular drainage patterns exist. It is recommended to incorporate such sites in the late scintigraphy images and to pursue aberrant sentinel nodes, as they may be the only sites of metastasis.     

Interval sentinel lymph nodes in melanoma

HYPOTHESIS: For patients with melanoma, interval or in-transit sentinel lymph nodes (SLNs) have the same risk for nodal metastasis as SLN in traditional (ie, cervical, axillary, and inguinal) nodal basins. DESIGN: Prospective clinical trial. SETTING: Multicenter study. PATIENTS: Eligible patients were aged 18 to 70 years with melanomas of at least 1.0-mm Breslow thickness and nodes with clinically negative findings. INTERVENTION: Sentinel lymph node biopsy was guided by preoperative lymphoscintigraphy to identify all SLNs. MAIN OUTCOME MEASURES: We evaluated interval nodal sites, including epitrochlear, popliteal, and subcutaneous or intramuscular nodes outside of traditional basins, for the presence of metastases. RESULTS: The SLNs were identified in 2332 nodal basins from 2000 patients. In 62 patients (3.1%), interval SLNs were identified. We found SLN metastases in 442 (19.5%) of 2270 conventional nodal basins and 13 (21.0%) of 62 interval sites. In 11 (84.6%) of the 13 cases in which we found an interval node that was positive for metastatic disease, it was the only site of nodal metastasis. CONCLUSIONS: Although interval SLNs are identified infrequently, they contain metastatic disease at nearly the same frequency as SLNs in cervical, axillary, and inguinal nodal basins. Positive interval SLNs are likely to be the only site of nodal metastasis. Therefore, detailed preoperative lymphoscintigraphy and meticulous intraoperative search for interval nodes should be performed.     

Sentinel-Lymphknoten-Dissektion beim malignen Melanom

INTRODUCTION: In patients with cutaneous malignant melanoma, the sentinel lymph node (SLN) reflects the histopathological features of the lymphatic basin with high accuracy. MATERIAL AND METHODS: Three hundred eighty-one melanoma patients at the Hornheide clinic with an overall follow-up of 36 months (November 1998 to October 2001) underwent sentinel lymph node dissection (SLND). RESULTS: The SLNs were successfully found in 93% of truncal melanoma ( n=136), 97% of melanoma of the extremities ( n=184), and 86% of melanoma of the head and neck region ( n=61).Of truncal midline melanomas, 84% ( n=43) showed two or more regional basins, in contrast to 18% of nonmidline melanoma ( n=93). Histopathological analysis revealed occult nodal disease in 25% of all patients. Completion lymphadenectomy revealed residual nodal disease in 8% of all patients with low risk melanoma with a tumor thickness of 0-1.5 mm (two of 26 patients with positive SLN) and in 11% of all patients with high risk melanoma with tumor thickness above 1.5 mm (eight of 70 patients with positive SLN). Tumor relapse was noted in 5% of negative SLN patients and 14% of positive SLN patients. The results of the method were false negative in 2% with a sensitivity of 98%. CONCLUSION: Sentinel lymph node dissection is a reliable and accurate method of staging regional lymph nodes for all primary tumor sites. It can localize occult metastases in unexpected lymphatic basins and provides critical indications for completion lymphadenectomy. It represents an essential method of establishing stratification criteria for future adjuvant trials. Further long-term follow-up is needed to investigate its prognostic relevance to recurrence and overall survival.     

Lymphatic drainage patterns of head and neck cutaneous melanoma observed on lymphoscintigraphy and sentinel lymph node biopsy

BACKGROUND: The purpose of this study was to evaluate lymphatic drainage patterns of head and neck cutaneous melanoma observed on preoperative lymphoscintigraphy and sentinel lymph node biopsy (SLNB) and determine discordancy from clinically predicted lymphatic drainage patterns. METHODS: We conducted a retrospective chart review of 114 patients with head and neck cutaneous melanomas evaluated with preoperative lymphoscintigraphy and SLNB from January 2001 through July 2004. RESULTS: At least one sentinel lymph node (SLN) was identified in 97% of cases. On preoperative lymphoscintigraphy, an SLN was identified in an area not clinically predicted in 49 cases (43%). The most common sites of discordancy were in areas not typically dissected in standard neck dissections, such as the postauricular region, or in areas of more distant drainage than described previously, such as the inferior or posterior neck. Their percentages of discordant cases were 51%, 27%, and 22%, respectively. The sites of regional recurrence occurred in two cases not predicted on preoperative lymphoscintigraphy and in two cases of failed SLNB. CONCLUSIONS: On the basis of preoperative lymphoscintigraphy and the results of SLNB, head and neck cutaneous melanomas do have expected lymphatic drainage patterns despite perceived discordancy with previously clinically predicted drainage patterns that are based on standard neck dissection specimens. These "discordant" sites can still harbor melanoma, and all sites predicted on preoperative lymphoscintigraphy still need to be explored. The four cases of recurrences underscore the importance of close follow-up for all patients regardless of the SLNB result.     

Long-term prognosis and significance of the sentinel lymph node in head and neck melanoma

Objective To report the long-term significance of sentinel lymph node (SLN) biopsy on prognosis, determine false-negative SLN occurrences, and determine risk factors for death and recurrence in a large series of patients with head and neck melanoma. Study Design Case series with tumor registry review. Setting Academic tertiary care medical center. Subjects and Methods A database review was performed of all patients who underwent SLN biopsy for head and neck melanoma from 1994 to 2009. End points assessed were SLN status, recurrence, false-negative SLN results, and survival comparing SLN-positive and SLN-negative patients and different locations. Survival curves and multivariate analyses were performed. Results SLN biopsy was performed in 365 patients. SLNs were identified in 98.6% of patients with a mean of 3.7 nodes removed from 1.6 nodal basins per patient. Median follow-up was 8 years. The SLN was positive in 40 (11%) patients. SLN-positive patients had significantly thicker melanomas, higher recurrence (P < .0001), and a significant decrease in overall survival compared with SLN-negative patients (P < .002). Scalp melanoma patients had significantly thicker melanomas and an elevated risk of SLN positivity, recurrence, and death compared with other sites. Seventeen of 365 SLN-negative patients developed regional nodal disease for a false-omission rate of 5.2% and a negative predictive value of a negative SLN to be 94.8%. Risks for false negative-SLN occurrences included thick melanomas and scalp melanomas. Conclusion SLN biopsy is accurate in head and neck melanoma and provides significant prognostic data. Scalp melanoma patients present with thicker tumors with an increase in SLN positivity and false-negative SLN occurrences.     

Significance of Multiple Nodal Basin Drainage in Truncal Melanoma Patients Undergoing Sentinel Lymph Node Biopsy

Background: Although previous studies have demonstrated that truncal site is associated with an adverse prognosis, explanations for such risk are lacking. In addition, the number of nodal basins as well as the number of lymph nodes containing regional metastases are important prognostic factors in these patients. Because the lymphatic drainage pattern of truncal melanoma often includes more than one basin, we designed a study to evaluate (1) whether patients with multiple nodal basin drainage (MNBD) were at an increased risk of lymph node metastases identified by sentinel lymph node (SLN) biopsy, and (2) whether the histological status of an individual basin reliably predicted the status of the other draining basins in patients with MNBD.Methods: The records of 295 consecutive truncal melanoma patients who were managed primarily with intraoperative lymphatic mapping and SLN biopsy, between 1991 and 1997, were reviewed. All patients underwent preoperative lymphoscintigraphy, which established the number and location of draining nodal basins. Univariate and multivariate analyses of relevant clinicopathological factors were performed to assess which factors may predict the presence of a pathologically positive SLN.Results: At least one SLN was identified in 281 patients. MNBD was present in 86 (31%) patients, and a pathologically positive SLN was found in 56 (20%) patients. By multivariate analysis, the presence of MNBD (relative risk = 1.9; P = .03), tumor thickness (P = .007), and tumor ulceration (relative risk = 2.4; P = .01) were significant independent risk factors for the presence of at least one pathologically positive SLN. SLN pathology in one basin did not predict the histology of other basins in 19 (22%) of 86 patients with MNBD.Conclusions: MNBD is independently associated with an increased risk of nodal metastases in truncal melanoma patients. Because the histological status of an individual basin did not reliably predict the status of the other draining basins in patients with MNBD, it is important to adequately identify and completely assess all nodal basins at risk, as defined by lymphoscintigraphy, in truncal melanoma patients.Presented at the 52nd Annual Meeting of the Society of Surgical Oncology, Orlando, Florida, March 4–7, 1999.     

Lymphoscintigraphy of the neck

Lymph node imaging has been helpful in managing patients with lymphoma, melanoma, and breast cancer. To evaluate 38 patients with head and neck cancers, 99mTc minicolloid was injected adjacent to the tumor and into a similar area on the uninvolved side. Lymphoscintigraphy of the neck was performed at 3- and 5-hour intervals after injection and bilateral cervical lymphatic drainage was observed. Each patient then underwent a neck dissection. The pathologic node findings were then correlated with the neck scans. Results confirmed that cervical lymphatic drainage is unpredictable in 50% of the patients once the channels are involved with metastatic disease. Lymphoscintigraphy is not a reliable method of detecting early metastatic cervical carcinoma.     

Sentinel lymph-node biopsy for melanoma of the trunk and extremities: the McGill experience.

OBJECTIVE: To determine the effectiveness of sentinel lymph-node (SLN) biopsy for melanoma of the trunk and extremities. DESIGN: Case series review. SETTING: Royal Victoria Hospital, a Canadian university hospital. PATIENTS: Thirty-six patients (18 women and 18 men) seen between October 1996 and December 1998 with melanoma 1 mm or more in thickness with clinically negative lymph-node basins. Follow-up was 396 days. INTERVENTIONS: SLN biopsy. Technetium-99m filtered sulfur colloid (0.5 mCi) was injected intradermally around the melanoma or the excision scar 10 to 15 minutes before the surgical skin preparation. The identification of the SLN(s) was done with a hand-held gamma probe. Local anesthesia was used mostly for inguinal SLN biopsy whereas general anesthesia was usually required for axillary SLN biopsy. Preoperative lymphoscintigraphy was used only for trunk melanomas. OUTCOME MEASURES: Morbidity, successful identification of the sentinel node and locoregional recurrence. RESULTS: The mean age of patients at diagnosis was 53.4 years (range from 22-76 yr). The melanomas were distributed between the lower extremities (20 patients), upper extremities (8 patients) and trunk (8 patients). The mean Breslow thickness was 2.35 mm (range from 1-8 mm). Lymphoscintigraphy accurately localized the lymph-node drainage basin for trunk melanomas. In 1 patient the SLN could not be identified because the radiocolloid failed to migrate (failure rate 2.8%). The average number of SLNs removed was 1.97. Eight patients (22%) had sentinel nodes positive for malignant disease. The postoperative complication rate was 8.5%. Seven of 8 patients with positive SLNs underwent a complete node dissection (1 patient refused). Of the completion dissections only 2 patients had positive non-SLNs. All patients with positive nodes received interferon alpha-2b as adjuvant treatment. At follow-up, 34 patients are alive with no evidence of disease, 1 patient with a positive SLN is alive with distant metastatic disease and 1 patient with a negative SLN is dead of disseminated disease. CONCLUSION: SLN biopsy is a feasible technique with an acceptable failure rate and is thus a useful tool in the surgical management of melanoma.     

Preoperative assessment of cervical lymph node involvement in esophageal cancer

Lymphoscintigraphy (LS) and computed tomography (CT) were used preoperatively to assess nodal involvement in 23 patients with esophageal cancer. LS predicted nodal involvement with a high sensitivity and low specificity since false positives were frequent while in contrast, CT predicted it with high specificity and low sensitivity. The combination of LS and CT improved sensitivity by up to 87 per cent but not specificity. The positive image of nodes in LS was associated with malignant cell involvement and/or reactive changes in the nodes, especially sinus histiocytosis and germinal center hyperplasia. When cervical nodes were involved, bilateral images of lymph flows were often lost and by using the images of lymph flow as well as those of the nodes, the combination predicted cervical nodal involvement with a sensitivity of 83 per cent and a specificity of 86 per cent. Thus, the combination of LS and CT is beneficial for preoperatively estimating cervical nodal involvement in esophageal cancer.