| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
SUPPLEMENT |
From the National Cancer Institute, Naples, Italy.
Correspondence: Address correspondence and reprint requests to: Corrado Caracò, MD, via Orazio 31, 80122 Naples, Italy; Fax: 39-0815903239; E-mail: corrado.caraco{at}fastwebnet.it
ABSTRACT
Management of patients with cutaneous melanoma in the absence of lymph node metastases is still controversial. The experience at the National Cancer Institute in Naples was analyzed to evaluate 3-year disease-free survival and overall survival for all patients who underwent sentinel lymph node biopsy (SLB) with Breslow thickness greater than 4 mm. Data from 359 sentinel biopsies performed in the past 5 years were reviewed to determine the effect of the treatment on disease-free survival and overall survival after stratifying patients for node status, tumor ulceration, and Breslow thickness. Statistical analysis showed a better 3-year survival for sentinel node–negative patients than for sentinel node–positive cases (88.4% and 72.9%, respectively; P < .05). Tumor ulceration retained its prognostic significance despite lymph node status, indicating a higher risk for development of distant metastases. Survival curves associated with thicker melanomas did not show significant differences between negative- and positive-SLB patients. SLB provides accurate staging of nodal status in melanoma patients who have no clinical evidence of metastases. Longer follow-up and final results from ongoing trials are necessary to definitively clarify the role of this procedure.
Key Words: Cutaneous melanoma • Lymph node metastases • Lymphoscintigraphy • Sentinel biopsy
Management of cutaneous melanoma changed in the past decade because of the introduction of the sentinel lymph node concept. Sentinel lymph node biopsy (SLB) can accurately identify patients with nodal micrometastases who are eligible for complete lymph node dissection. Data from 359 SLB procedures were analyzed to determine 3-year disease-free survival (DFS) and overall survival (OS), particularly for patients with thicker melanomas (greater than 4 mm), who have a worse prognosis for the early appearance of distant metastasis.1,2
MATERIALS AND METHODS
At the National Cancer Institute of Naples, 359 SLBs were performed in 331 patients with cutaneous melanoma from 1996 to 2002. All patients with a primary melanoma thicker than 1 mm or of Clark level IV or V, without clinical evidence of nodal metastases, underwent SLB and were retrospectively analyzed to evaluate DFS and OS. All patients had undergone excision of the primary melanoma with a margin less than 1 cm within 3 months before SLB. Clinical and pathological characteristics of these patients are listed in Table 1.
|
Lymphoscintigraphy was performed 2 to 4 hours before surgery to identify the drainage basin. Human serum albumin labeled with 99Tc was injected intradermally around the tumor. Coronal, sagittal, and oblique images were obtained, and the sentinel node "hot spot" was marked on the skin. Cases that showed drainage to more than two basins or no drainage were excluded from this study.
About 20 minutes before the surgical procedure, 1 mL of Patent Blue V dye (Jacopo Monico, Mestre, Italy) was injected intradermally around the primary scar. Just before the incision, a gamma probe was used to verify correct position of the cutaneous marker and to guide the cutaneous incision directly over the hot lymph node. The hand-held gamma probe then was used to guide the identification of sentinel nodes, by correlating radioactivity in vivo, ex vivo, and in operative fields; this correlation ensured that all possible sentinel nodes were removed.
Serial sections of each sentinel node were analyzed by hematoxylin-eosin (H&E) and immunohistochemical (IHC) staining. Patients with tumor-positive sentinel nodes underwent complete lymph node dissection of the involved basin. Those with groin localization underwent superficial and deep dissection when nodal metastases were identified by H&E staining or superficial dissection if nodal metastases were identified only by IHC staining. Negative-SLB patients had no adjuvant therapy; follow-up visits were scheduled every 3 months for the first 2 years and every 6 months thereafter. Stage III patients were chosen, after radical dissection, to participate in adjuvant trials or to undergo observation only, according to their decision.
Statistical Analysis
All statistical analyses were performed with SPSS software (SPSS, Chicago, IL). Frequency distributions for the basin of drainage, sentinel node identification, and sentinel/nonsentinel metastasis were computed. The 
2) test was used in univariate analysis to determine significant differences between the two groups. Actuarial survival curves were calculated by the Kaplan-Meier method, and differences between these parameters were tested for significance with the log-rank test. All factors found to be significant in univariate analysis were included in a multivariate analysis in which a Cox proportional hazards model was used to determine independent prognostic factors for survival.
RESULTS
The overall rate of sentinel node identification was 98%. The rate of sentinel node identification was higher for the groin region (100%) and lower for the axillary region, probably because of the fatty tissue in the axillary area, which might interfere with the procedure. In 28 cases, lymphoscintigraphy revealed drainage to two basins. These basins were bilateral in 14 patients (9 in the axilla, 4 in the groin, and 1 in the neck), ipsilateral in 8 cases, and ambiguous in 6 cases. In 10 patients, lymphoscintigraphy identified an unusual drainage pattern to popliteal, scapular triangle, epigastric, or occipital regions; in these cases the rate of sentinel node identification was 80%.
Lymph node micrometastases were found in 68 cases (15.8%), and 11 patients (3.2%) had micrometastases identified only by IHC staining. The nonsentinel node identification rate was 8%, with two nodes identified in 21 patients and three or more lymph nodes identified in 10 patients. Complete dissection of basins with sentinel nodes revealed no other tumor-positive nodes in 77.1% and 81.8% after H&E and IHC staining, respectively. All patients with Breslow thickness less than 2 mm had no other micrometastatic nodes, whereas those with greater Breslow values had two or more positive nonsentinel nodes.
There were no major perioperative complications. Three patients (0.9%) had cutaneous erythema after the injection of Patent Blue V dye, and 25 (7.5%) experienced a seroma in the dissected basin.
The median follow-up was 34.2 months. The false-negative rate, calculated as the rate of lymph node recurrence in the SLB basin, was 3.9% (13 cases). The rate of lymph node recurrence in a different basin was 3.3% (11 cases). Among 68 cases of radical dissection for positive sentinel nodes, only two recurred locally, whereas three showed nodal metastases in a different basin. By univariate analysis, Breslow thickness (P < .001), sentinel node tumor status (P < .002), and ulceration (P < .003) were statistically significant independent prognostic factors for survival; age was not significant. The DFS rates for patients with negative and positive sentinel nodes were 39.3% and 33.5%, respectively, without statistical significance. The 3-year OS rates for patients with negative and positive sentinel nodes were 88.4% and 72.9%, respectively (P < .05) (Fig. 1). When stratified by the presence or absence of tumor ulceration, OS was 92.3% for patients with a tumor-negative sentinel node and a nonulcerated primary, compared with only 54.8% when both adverse factors were present (P < .005) (Fig. 2). Three-year OS curves for cases with Breslow thickness greater than 4 mm showed no differences between those with tumor-negative versus tumor-positive sentinel nodes, indicating the influence of other factors on disease outcome (Fig. 3).
|
|
|
Management of cutaneous melanoma changed in the past decade, after the introduction of the sentinel concept by Morton et al.3 in 1992. This technique permits staging at diagnosis for the presence or the absence of lymph node metastases, avoiding the morbidity of useless nodal dissections in about 80% of cases. Many investigators have recently reported their experience with SLB and its feasibility as a staging procedure; patients with tumor-negative sentinel nodes had better survival than those with micrometastatic nodal involvement.1,4–7 The Multicenter Selective Lymphadenectomy Trial (MSLT) will provide final results on DFS and OS, shaping the real role of this revolutionary technique.8
Our study’s results indicate that there are different outcomes for micrometastatic (at IHC) and metastatic (at H&E) nodal disease in terms of overall survival (88.4% and 72.9%, respectively). In addition, our findings suggest that patients with primaries thicker than 4 mm represent a heterogeneous group whose prognosis depends not only on sentinel node status (no differences in 3-year OS for patients with tumor-negative vs. tumor-positive sentinel nodes) but also on factors such as ulceration. In a study of patients with thick primary melanomas, Kim et al.2 reported thickness, ulceration, and sentinel node status as predictors of survival, stressing the important role of sentinel node status. Clary et al.9 and White et al.10 stressed the prognostic value of early appearance of distant metastases in patients with primary melanomas thicker than 4 mm. Essner et al.11 recently reported that tumor thickness, ulceration, and sentinel node status are not predictive of survival in 4-mm melanoma. Similar results were reported by Ferrone et al.12; these authors critically analyzed the outcome of patients with melanomas of higher Breslow thickness, assessing nodal status alone as a predictor of survival. They proposed that thicker tumors have a biological complexity, and prognosis seems to be governed by multiple factors such as tumor thickness, ulceration, age, sex, and site of primary. Massi et al.13 reported a multifactorial analysis with Breslow thickness, type of invasive front, ulceration, and even mitotic rate as independent predictors of OS. The serum TA90 immune complex (IC) assay performed at John Wayne Cancer Institute (JWCI) may add further important information to predict survival associated with thicker melanomas.14 Data from JWCI showed that a simple serum assay discriminated cases at low and high risk of recurrence; TA90-IC status was the strongest independent prognostic factor for both DFS and OS.
In our experience, tumor ulceration adds important information to understand the outcome of cutaneous melanoma and may be more important than sentinel node status. Patients with tumor-negative sentinel nodes and ulceration have a worse prognosis than patients with tumor-positive sentinel nodes, with a 3-year OS of 79.8% and 95.6%, respectively (P < .005). This means that tumor ulceration retains its prognostic significance despite lymph node status, with a higher risk of development of distant metastases.
Sentinel node biopsy provides a correct classification regarding lymphatic involvement in patients with cutaneous melanoma, but probably it is only one aspect of the biological complexity of melanoma when tumor thickness is greater than 4 mm.15,16 Other factors and prognostic models, like those proposed by Ferrone’s group, might identify the different subgroups with differences in survival and clarify the role of SLB in cases with Breslow thickness greater than 4 mm.
FOOTNOTES
Sentinel lymph node biopsies (SLBs) performed in the past 5 years were reviewed to determine the effect of the treatment on disease-free survival and overall survival. Statistical analysis showed a better 3-year survival for sentinel node–negative versus node-positive patients. SLB provides accurate staging of nodal status in melanoma patients who have no clinical evidence of metastases.
Received for publication December 4, 2003. Accepted for publication December 12, 2003.
REFERENCES
This article has been cited by other articles:
 |
|
 |
|
  J. N. Cormier, Y. Xing, M. Ding, J. E. Lee, P. F. Mansfield, J. E. Gershenwald, M. I. Ross, and X. L. Du Population-Based Assessment of Surgical Treatment Trends for Patients With Melanoma in the Era of Sentinel Lymph Node Biopsy J. Clin. Oncol., September 1, 2005; 23(25): 6054 - 6062. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
