10.1245/s10434-007-9357-0
Annals of Surgical Oncology 14:1934-1942 (2007)
© 2007 Society of Surgical Oncology
Patterns of First-Recurrence and Post-recurrence Survival in Patients with Primary Cutaneous Melanoma After Sentinel Lymph Node Biopsy
Kimberly M. Dalal, MD,
Ami Patel, BA,
Mary S. Brady, MD,
David P. Jaques, MD and
Daniel G. Coit, MD
Department of Surgery, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, New York 10021, USA
Correspondence: Address correspondence and reprint requests to: Daniel G. Coit, MD; E-mail: coitd{at}mskcc.org
 |
ABSTRACT
|
|---|
Background: Sentinel lymph node biopsy (SLNB) has become well accepted in management of patients with primary cutaneous melanoma. An understanding of the pattern of recurrence after SLNB is helpful in coordinating a rational plan of follow-up in these patients. We sought to determine the site and timing of initial recurrence and post-recurrence survival after SLNB.
Methods: Stage I/II melanoma patients who underwent SLNB during 1991–2004 were identified from a prospective single-institution database. Site and date of first recurrence after SLNB were recorded. Patterns of recurrence after SLNB and post-recurrence survival were analyzed.
Results: One thousand and forty-six patients underwent SLNB. The sentinel lymph node (SLN) was positive in 164 patients (16%). Median follow-up was 36 months for survivors. Median and 3-year relapse-free survival for SLN-positive patients were 41 months and 56%, and for SLN-negative patients were not reached and 87%, respectively (P < .0001). Of the SLN-positive patients, 47% experienced recurrence, compared with 14% SLN-negative patients. The pattern of recurrence stratified by SLN status was similar between the two groups (P = NS). After recurrence, the site of recurrence was the only significant prognostic factor influencing survival (P < .0001).
Conclusions: Although SLN-positive patients experience recurrence far earlier and more frequently than SLN-negative patients, the pattern of recurrence is similar. After recurrence, its site is the primary determinant of survival.
Key Words: Melanoma • Recurrence • Sentinel lymph node biopsy • Post-recurrence survival
|
INTRODUCTION
|
|---|
In 2006, an estimated 62,190 patients were diagnosed with melanoma in the United States, and 7910 of them died of the disease.1 The presence of meta-static melanoma in the regional lymph node basin has been shown to be the most powerful prognostic indicator for patients with American Joint Committee on Cancer (AJCC) stage I or II disease.2–5 Since the seminal work of Morton et al.,6 SLN mapping and sentinel lymph node biopsy (SLNB) in patients with primary cutaneous melanoma have been shown to be extremely accurate in staging nodal basins at risk for regional metastases. On this basis, SLNB has become widely accepted in the initial management of patients with melanoma 
1 mm in depth or Clark level IV.7
Several reports have addressed the pattern of recurrence in patients undergoing SLNB.4,8,9 Clary et al.3 reported that the most common site of initial recurrence was locoregional in these patients, regardless of the status of the sentinel lymph node (SLN).9,10 Recurrence occurred earlier, however, in the SLN-positive patients.3
In this study, we determined the site and timing of initial relapse after SLNB and postrelapse survival by using a large, prospective single-institution database of patients with cutaneous melanoma.
|
MATERIALS AND METHODS
|
|---|
Patients
Data from all patients with melanoma seen at Memorial Sloan-Kettering Cancer Center are entered into a prospective, single-institution melanoma database. Study patients had histologically confirmed melanoma with clinically negative lymph nodes and no evidence of distant metastatic disease at the time of the SLNB. SLNB was offered to patients with primary melanoma 1.0 mm Breslow thickness, or Clark level IV–V of any thickness, or to selected patients who were thought to be at high risk of occult nodal metastases (those primaries tumors with ulceration, major regression, or both).11 Patients with multiple primary tumors who underwent SLNB in multiple basins or those with clinical stage III disease were excluded from the current study.
Patient data, including clinical characteristics, pathologic findings, and follow-up, were recorded. Postoperative follow-up included physical examination every 3 to 4 months for the first year, every 3 to 6 months for the second year, and every 6 to 12 months thereafter. Chest radiographs were performed every 6 to 12 months during the first 2 to 3 years. Serum lactate dehydrogenase levels and a complete blood count were also obtained annually during the first 2 to 3 years of follow-up. Further investigations, including computed tomography and positron emission tomography scanning, were performed on a selective basis as clinically indicated.
Data for analysis were extracted from the database and computerized patient records. The data examined included patient demographics, tumor and SLN pathologic characteristics, site and date of first recurrence after SLNB, and status at last follow-up. This study was conducted after institutional review board approval.
SLNB Technique
Visual blue dye and radiolocalization were used to help ensure identification and removal of all SLNs. A dose of 1–2 mCi of technetium-99m sulfur colloid was injected intradermally at the tumor site 2 to 4 hours before the operation. After induction of anesthesia and immediately before wide excision of the primary tumor, .5–1.0 mL of isosulfan blue dye (Lymphazurin 1%; Hirsch Industries, Richmond, VA) was injected intradermally at the site of the primary melanoma. All basins identified by lymphoscintigraphy were explored through small incisions guided by the handheld gamma probe. All blue nodes and all nodes with 10% of the highest ex vivo nodal counts were removed.12 Each SLN removed was submitted separately for analysis. All patients underwent wide local excision of the primary melanoma or reexcision of the previous biopsy site. Of the patients with metastases in the SLN on review of the permanent sections, 90% (147 of 163) underwent a completion lymph node dissection (CLND) of the affected nodal basin within 2 to 3 weeks of SLNB.
Histological Analysis
Permanent sections of SLNB specimens were analyzed with hematoxylin and eosin staining and light microscopy. Nodes found to be negative for metastases by this method were analyzed by step sectioning and staining with immunohistochemical markers (S-100, HMB-45).
Recurrence Analysis
Data were analyzed for patterns of first site recurrence. For patients with multisite recurrence, the first site was coded as that associated with the worst prognosis; this was defined such that systemic (including distant nodal) is worse than nodal, which is worse than in-transit disease, which is worse than local (within 2 cm of the primary site). Nodal recurrence was defined as a recurrence in the lymph node basin regardless of whether a CLND had been performed previously. Multisite recurrence was defined as any recurrence detected within 4 weeks of each other (e.g., liver metastases detected within 4 weeks after a local recurrence).
Statistical Analysis
Clinicopathologic correlates were analyzed by the 
2 test for categorical variables. Relapse-free survival (RFS) was defined as the time from SLNB to date of first relapse or last follow-up. Means testing of continuous variables was performed with the Student t-test. RFS analyses were performed considering pathologic variables (primary tumor site, Breslow thickness, ulceration, Clark level, and SLN status) and patient variables (age, sex). Patients whose Clark level or ulceration status was unknown were not included in univariate analysis. RFS was determined by Kaplan-Meier methodology, and comparisons were made by the log rank test for categorical variables. Tumor thickness and age were treated as continuous variables for univariate analyses by a Cox proportional hazard model. Multivariate analyses were performed by a Cox proportional hazard model.
Post-recurrence survival (PRS) was defined as the time from initial recurrence after SLNB to date of death or last follow-up.
|
RESULTS
|
|---|
Patient Demographics
From October 1991 to October 2004, 1046 patients with stage I or II primary cutaneous melanoma underwent successful SLNB at Memorial Sloan-Kettering Cancer Center. The clinicopathologic characteristics are shown in Table 1
. The median age of patients was 56 years, with a range of 6 to 97 years. Men comprised 59% of the group. The most common primary site was the extremity (n = 531, 51%), followed by the trunk (n = 353, 34%) and head and neck (n = 162, 15%). The median Breslow thickness was 1.7 mm (range, .20–24 mm). A Clark level IV or V primary lesion was seen in 761 (73%) and 89 (9%) patients, respectively. Ulceration was present in 296 (28%) of the primary lesions; the ulceration status was unknown in 142 patients (14%). Approximately half of the patients presented with 2002 AJCC stage I (IA, n = 68, 7%; IB, n = 493, 47%) disease. The SLN was positive in 164 patients (16%) (Table 2). The median age of SLN-positive and SLN-negative patients was 59 (17–90) and 57 (6–97), respectively (P = NS).
Follow-up and RFS
The median follow-up from the time of SLNB for 1046 patients was 36 months (range, 0–151 months). The median time to recurrence for those who experienced recurrence of disease was 19 months. At last follow-up, 895 patients were alive. The RFS distribution for the entire cohort is illustrated in Fig. 1. Three- and 5-year RFS was 81% and 72%, respectively.
Incidence of Recurrence
At a median follow-up of 36 months, 203 patients (19%) developed recurrent disease. There were 230 recurrences in these 203 patients. Most patients (87%, n = 177) experienced recurrence at one site, 25 patients (12%) had two sites of recurrence, and one patient had three sites of recurrence. Recurrence was far more common in SLN-positive patients. Seventy-seven (47.2%) of 163 patients with a positive SLN experienced recurrence of disease, compared with 126 (14.4%) of 883 patients with a negative SLN (P < .001). The median time to recurrence was shorter in the 77 SLN-positive patients who experienced recurrence (13 months) compared with the 126 SLN-negative patients who experienced recurrence (24 months; P < .001).
Patterns of Recurrence
Of the 203 patients who developed recurrent mel-anoma, the most common site of initial recurrence was distant (n = 103 patients, 51%), followed by intransit disease (n = 50, 25%), regional nodal failure (n = 40, 20%), and local recurrence (n = 10, 4.9%). Although distant failure was most slightly common in both groups, in-transit recurrences were more common in SLN-positive compared with SLN-negative patients (30% vs. 21%, respectively, P = NS). In addition, regional nodal recurrences were slightly more common in SLN-negative patients compared with SLN-positive patients (22% vs. 16%, respectively), but this was again not statistically significant. The pattern of recurrence in all patients, and stratified by SLN status, is shown in Table 3. Outcomes are presented both as a percentage of all recurrences in the entire cohort and within SLN-positive and SLN-negative groups as well as a percentage of patients who experienced recurrence of disease within the SLN-positive and SLN-negative groups.
Although recurrence was seen more often in SLN-positive patients, the overall pattern was similar to that seen in SLN-negative patients (Fig. 2). Approximately half of the initial recurrences in both the SLN-positive and SLN-negative groups were systemic. One-third of the initial recurrences in both groups were of local or in-transit disease.
View larger version (17K):
[in this window]
[in a new window]
|
FIG. 2. Pattern of first-site recurrence defined by worst site. Approximately half of all recurrences in both the sentinel lymph node (SLN)-positive and -negative groups were systemic. One-third of the recurrences in both groups comprised local or in-transit disease. Among recurrences within each SLN status group, the overall pattern of recurrence was similar (P = NS).
|
|
There was no statistically significant difference in the site of first recurrence based on primary Breslow depth. Although recurrence was seen more often in patients with thick melanoma (P < .001), the pattern of recurrence was similar to that seen in patients with thinner lesions (P = NS) (Table 4). Moreover, although there was an insignificant trend toward increased in-transit and nodal first-site recurrence in patients with Clark level IV–V melanoma compared with those with Clark level II or III primaries, this was not statistically significant (P = NS) (Table 5).
Although recurrence was seen more often in patients with ulcerated tumors (P < .001), the pattern of recurrence was similar to that seen in patients with nonulcerated tumors (P = NS) (Table 6). SLN-positive patients with ulcerated melanomas >2 mm in thickness had a slightly higher rate of nodal failure (12.7%) compared with SLN-negative patients (8.7%) with the same lesions (P = NS).
Table 7 demonstrates the stratification of first site of recurrence by stage and SLN status. A greater percentage of stage II patients experienced recurrent disease compared with stage I patients, regardless of positive or negative SLN status. The disease of SLN-positive patients was more likely to recur than that of SLN-negative patients. Figure 3 demonstrates the Kaplan-Meier curves for stage I and stage II patients; 5-year RFS rates are 87% and 56%, respectively (P < .0001).
Post-recurrence Survival
The median survival after recurrence was 16 months in the 203 patients who developed recurrent disease (Fig. 4A). The median survival was least favorable after systemic recurrence (9 months), followed by nodal recurrence (20 months), intransit recurrence (37 months), and local recurrence (not reached) (Fig. 4B; Table 8).
View larger version (28K):
[in this window]
[in a new window]
|
FIG. 4. Post-recurrence survival (PRS) after sentinel lymph node (SLN) biopsy. PRS for all patients (A), patients stratified by site of worst recurrence (B), or patients stratified by SLN status (C).
|
|
When stratified by SLN status, median PRS for SLN-positive and SLN-negative patients was not markedly different (13 months and 20 months, respectively; Fig. 4C
). The 3-year PRS rates were 26% and 34% for SLN-positive and SLN-negative patients, respectively.
By univariate analysis, factors associated with a worse PRS were site of recurrence (P < .0001) and primary site (P = .03; data not shown). On multivariate analysis, the site of recurrence was the only significant predictor of PRS (P < .0001).
|
DISCUSSION
|
|---|
We have demonstrated that although SLN-positive patients experience recurrent disease far more frequently and earlier than SLN-negative patients, the pattern of first recurrence is similar to that seen in SLN-negative patients. At a median follow-up of 36 months, 47% of SLN-positive patients experienced recurrence at a median interval of 13 months, whereas 14.3% of SLN-negative patients experienced recurrence at a median interval of 24 months. This is consistent with many series of similar length of follow-up, in which the recurrence rate of SLN-negative patients is 9% to 14% (Tables 9 and 10).3,4,7–9,13–18 Although SLN-positive patients were significantly more likely to experience recurrence at in-transit (P < .001), nodal (P < .05), and distant (P < .001) sites than patients with a negative SLN, the overall pattern of recurrence within each SLN status group was similar. This particular finding of a similar distribution of sites of recurrence within SLN-positive and SLN-negative patients has not been demonstrated previously.
The incidence of nodal recurrence in SLN-negative patients was 3.2% of the entire cohort, which is consistent with previous reports3,4,13 (Table 9). Nodal recurrence was seen in 22% of SLN-negative patients who experienced recurrence of disease. This is also consistent with previous studies, in which the regional nodal basin occurred in 6.3%13 to 64%8 of SLN-negative patients who developed recurrent disease. Fincher et al.13 reported a low nodal recurrence rate of 6.3%, which may be attributed to a median thickness of <1.5 mm, which is less than our cohort, at a mean follow-up time of 50 months. Conversely, Gadd et al.8 demonstrated a 64% nodal recurrence rate, detected by immunoperoxidase stains with S-100, HMB-45, NK1C3, and MART-1; this relatively high rate may be due to a small number of patients who experienced recurrence and the relatively low incidence of locoregional and systemic recurrence in the cohort compared with other reports. The overall low rate of nodal recurrence in the literature for SLN-negative patients (Table 9) is most likely related to performance of the procedure by a small, consistent group of surgeons and nuclear medicine personnel. More importantly, however, the pathologic evaluation of the SLN now routinely includes both serial sectioning and immunohistochemical evaluation with S-100 and HMB-45 in the nodes that were found to be negative for disease by routine hematoxylin and eosin staining. This method supports the ability of SLNB to accurately stage the regional lymph node basin.
In our cohort, 28% of the recurrences in the SLN-negative patients were locoregional, defined as local or in transit, consistent with previous reports, in which recurrence ranges from 18%8 to 48%.19 Jansen et al.16 reported that nearly half of the recurrences were locoregional (Table 9). This emphasizes the importance of the physical examination, both by the patient and by the physician, in follow-up of patients with melanoma.
Half of all recurrences in the SLN-negative group were systemic; this has recently been shown by Zogakis et al.15 Our findings differ from the pattern described by Gershenwald et al.4 in patients with negative nodes, where, at a median follow-up of 35 months, local, in-transit, nodal, and systemic recurrences accounted for 13%, 26%, 32%, and 29% of the first-site recurrences, respectively. Our cohort may be skewed away from local and toward systemic recurrences because 85% of our patients presented with Clark level IV–V, compared with 58% of patients in his study.
When we review SLN-positive patients and patterns of recurrence, 16% of our patients who experienced recurrence did so in the mapped nodal basin (Table 10). Although this finding is consistent with a prior report from our institution,3 this is a higher proportion than has been reported elsewhere. One possible explanation for the high rate of nodal failure is that only 90% of patients with disease-positive SLN biopsy underwent CLND. Among SLN-positive patients, the nodal recurrence rate in those who did and did not undergo CLND was 5% (8 of 147) and 25% (4 of 16), respectively. Interestingly, two studies report no nodal recurrences in SLN-positive patients at a median follow-up of 32 months16 and 38 months13 (Table 10); however, these series had the highest locoregional recurrence rates, and nodal and locoregional recurrences were categorized separately. One explanation may be attributed to the skew of the cohort away from nodal to systemic16 or locoregional areas,13 respectively. Another explanation may be the differences in the definition of what comprises a nodal versus a locoregional recurrence.15–17 One-third of the recurrences in our SLN-positive group were local or in transit, similar to other series and within the reported range from 16%18 to 60%.13 Finally, most initial recurrences in our series of SLN-positive patients were at distant sites, which again is similar to most series.13,17
Interestingly, although a greater percentage of SLN-positive patients experience recurrence, a higher absolute number of recurrences arise within the much larger SLN-negative cohort. This underscores two points. First, SLN status is not a perfect predictor of recurrence. Second, patients with a negative SLN require follow-up because most recurrences will occur in this group.
Survival after recurrence is poor, which is a clear reflection of aggressive biology. The 3-year survival after systemic recurrence of 16% is consistent with one previous study that reported a rate of 17%;15 however, our 3-year PRS after nodal recurrence of 29% was lower than 64% seen in another cohort.15 The only independent predictor of PRS is the worst site of initial recurrence. Balch et al.19 demonstrated that local recurrence was associated with a highly statistically significant difference in survival compared with patients who never experienced a local recurrence; however, when he compared patients who had local recurrence as their first sign of relapse (n = 13) and patients who developed local recurrence later in their course of metastatic disease (n = 11), there were no statistically significant differences in PRS. To our knowledge, our report is the first to formally analyze predictors of PRS after SLNB in the literature.
Delineating the pattern of first recurrence with regard to site and timing as well as PRS in patients undergoing SLNB raises additional questions. Identifying predictors of recurrence as well as elucidating how recurrences are detected will be helpful in formulating a rational plan for routine follow-up and surveillance of patients after SLNB.
In conclusion, although patients with positive SLNs experience recurrence of disease earlier and more frequently than their node-negative counterparts, most recurrences occur in SLN-negative patients after long-term follow-up. The pattern of recurrence is similar. Once disease recurs, the site of initial relapse is the primary determinant of PRS.
Received for publication July 29, 2006.
Accepted for publication November 9, 2006.
|
REFERENCES
|
|---|
- Jemal A, Murray T, Ward E, et al. Cancer Statistics, 2006. CA Cancer J Clin 2006; 56:106–30.[Abstract/Free Full Text]
- Stadelmann W, Rapaport D, Soong S, et al. (1998) Prognostic clinical and pathologic features. In: Balch CHA, Sober A, Soong S (edsCutaneous Melanoma Quality Medical, St. Louis, MO, pp 11–35.
- Clary BM, Brady MS, Lewis JJ, Coit DG. Sentinel lymph node biopsy in the management of patients with primary cutaneous melanoma: review of a large single-institutional experience with an emphasis on recurrence. Ann Surg 2001; 233:250–8.[CrossRef][Medline]
- Gershenwald JE, Colome MI, Lee JE, et al. Patterns of recurrence following a negative sentinel lymph node biopsy in 243 patients with stage I or II melanoma. J Clin Oncol 1998; 16:2253–60.[Abstract]
- Gadd MA, Coit DG. Recurrence patterns and outcome in 1019 patients undergoing axillary or inguinal lymphadenectomy for melanoma. Arch Surg 1992; 127:1412–6.[Abstract/Free Full Text]
- Morton DL, Wen DR, Wong JH, et al. Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 1992; 127:392–9.[Abstract/Free Full Text]
- Gershenwald JE, Thompson W, Mansfield PF, et al. Multi-institutional melanoma lymphatic mapping experience: the prognostic value of sentinel lymph node status in 612 stage I or II melanoma patients. J Clin Oncol 1999; 17:976–83.[Abstract/Free Full Text]
- Gadd MA, Cosimi AB, Yu J, et al. Outcome of patients with melanoma and histologically negative sentinel lymph nodes. Arch Surg 1999; 134:381–7.[Abstract/Free Full Text]
- Essner R, Conforti A, Kelley MC, et al. Efficacy of lymphatic mapping, sentinel lymphadenectomy, and selective complete lymph node dissection as a therapeutic procedure for early-stage melanoma. Ann Surg Oncol 1999; 6:442–9.[Abstract]
- McCarthy W, Shaw H, Thompson J, et al. Time and frequency of recurrence of cutaneous stage I malignant melanoma with guidelines for follow-up study. Surg Gynecol Obstet 1988; 166:497–502.[Medline]
- Wagner JD, Gordon MS, Chuang TY, et al. Predicting sentinel and residual lymph node basin disease after sentinel lymph node biopsy for melanoma. Cancer 2000; 89:453–62.[CrossRef][Medline]
- McMasters KM, Reintgen DS, Ross MI, et al. Sentinel lymph node biopsy for melanoma: how many radioactive nodes should be removed?. Ann Surg Oncol 2001; 8:192–7.[Abstract/Free Full Text]
- Fincher TR, McCarty TM, Fisher TL, et al. Patterns of recurrence after sentinel lymph node biopsy for cutaneous melanoma. Am J Surg 2003; 186:675–81.[CrossRef][Medline]
- Chao C, Wong SL, Ross MI, et al. Patterns of early recurrence after sentinel lymph node biopsy for melanoma. Am J Surg 2002; 184:520–5.[CrossRef][Medline]
- Zogakis TG, Essner R, Wang H, et al. Melanoma recurrence patterns after negative sentinel lymphadenectomy. Arch Surg 2005; 140:865–72.[Abstract/Free Full Text]
- Jansen L, Nieweg OE, Peterse Jl, et al. Reliability of sentinel lymph node biopsy for staging melanoma. Br J Surg 2000; 87:484–9.[CrossRef][Medline]
- Statius Muller MG, van Leeuwen PA, van Diest PJ, et al. Pattern and incidence of first site recurrences following sentinel node procedure in melanoma patients. World J Surg 2002; 26:1405–11.[CrossRef][Medline]
- Wagner JD, Ranieri J, Evdokimow DZ, et al. Patterns of initial recurrence and prognosis after sentinel lymph node biopsy and selective lymphadenectomy for melanoma. Plast Reconstr Surg 2003; 112:486–97.[CrossRef][Medline]
- Balch CM, Soong S, Smith T, et al. Long-term results of a prospective surgical trial comparing 2 cm vs. 4 cm excision margins for 740 patients with 1–4 mm melanomas. Ann Surg Oncol 2001; 8:101–108.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
K. M. Dalal, Q. Zhou, K. S. Panageas, M. S. Brady, D. P. Jaques, and D. G. Coit
Methods of Detection of First Recurrence in Patients with Stage I/II Primary Cutaneous Melanoma After Sentinel Lymph Node Biopsy
Ann. Surg. Oncol.,
August 1, 2008;
15(8):
2206 - 2214.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
