- Email: [email protected]
Sentinel lymph node biopsy in thick malignant melanoma: A 10-year single unit experience
Journal of Plastic, Reconstructive & Aesthetic Surgery (2012) 65, 1396e1402
Sentinel lymph node biopsy in thick malignant melanoma: A 10-year single unit experience* Neil G. Fairbairn*, Georgios Orfaniotis, Mark Butterworth Department of Plastic Surgery, St John’s Hospital, Livingston, West Lothian, EH54 6PP, Scotland, UK Received 18 November 2011; accepted 12 April 2012
KEYWORDS Sentinel lymph node biopsy; Thick melanoma; Survival
Summary Between the years 2000e2010, 195 patients were diagnosed with 4 mm Breslow thickness malignant melanoma in our unit. Median follow-up was 36.8 months. 49% of patients were male and 51% were female. Median age was 74 years. The commonest melanoma type was nodular (55%). The commonest tumour location was on the extremity (45%). 64% of tumours were ulcerated. Median mitotic rate was 9. Median Breslow thickness was 7 mm 66 patients underwent sentinel lymph node biopsy. 44 (67%) patients had negative results and the remaining 22 (33%) patients were positive for metastatic melanoma. There was no statistically significant correlation between any of the patient or tumour variables (age, sex, melanoma type, melanoma site, Clark level, Breslow thickness, mitotic rate, ulceration) and sentinel lymph node status. Patients with Breslow thickness melanoma of <6 mm had a significantly better 5-year disease free and overall survival compared with those patients with >6 mm Breslow thickness melanoma (63.5% vs. 32.9%; P Z 0.004 and 73.9% vs. 54.7%; P Z 0.02 respectively). Recurrence rate was 50% in those with positive sentinel lymph node biopsy compared to 23% in those with negative results. Distant recurrence was the commonest in both groups. 5-year disease free survival was 64.1% in the SLNB eve group and 35.4% in the SLNB þve group (P Z 0.01). There was no significant difference in overall survival between the SLNB eve and SLNB þve groups (70.3% vs. 63.7% respectively; P Z 0.66). We conclude that sentinel lymph node biopsy in our unit has provided no survival benefit in those with thick melanoma over the past 10 years but is an important predictor of recurrence free survival. Breslow thickness remains an important predictor of disease free and overall survival in thick melanoma. ª 2012 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved.
*
Article presented at BAPRAS Winter Scientific Meeting 2011. * Tel.: þ44 7764514281; fax: þ44 1506523034. E-mail address: [email protected] (N.G. Fairbairn).
1748-6815/$ - see front matter ª 2012 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.bjps.2012.04.019
SLNB in thick melanoma
1397
Introduction
Results
Sentinel lymph node biopsy (SLNB) is the most sensitive and accurate investigative modality for establishing regional node status in patients with melanoma.1 The status of the regional node has established itself as the most powerful predictor of recurrence and overall survival in patients with intermediate thickness melanoma.2 The role of SLNB in patients with thick melanoma is more uncertain. Opponents argue that the procedure is futile in this cohort of patients due to the high incidence of distant metastases at presentation and the overall poor survival of patients.3 This has been supported by studies showing no benefit in overall survival according to the regional node status.4 However, recent evidence suggests that patients with thick melanoma are a heterogenous group with regards to tumour behaviour and prognosis. Those with negative SLNB may represent a subset of thick melanoma patients with significantly better prognosis compared to those with regional node metastases.5e8 Controversy around the subject of SLNB in thick melanoma continues and the aim of our retrospective single centre study is to report how SLNB and other selected variables have affected outcome in our unit over the last 10 years.
Patient demographics
Methods Study design and patients Approval for the study was gained from the local research and ethics committee and the Scottish Melanoma Group. We conducted a retrospective analysis of the regional melanoma database provided by the Scottish Melanoma Group. All patients presenting between 2000 and 2010 with thick (4 mm Breslow) malignant melanoma were included in the study. Information including patient date of birth, age at diagnosis, date of diagnosis, melanoma type, melanoma site, Breslow thickness, ulceration status, Clark level, lymphovascular invasion, regression, mitotic rate, date and result of SLNB, date and result of regional node dissection, recurrence date, recurrence site, follow-up status and date of death were recorded. Histology results were obtained from our local electronic patient management system. All information was compiled into an Excel database for analysis.
Statistics All statistical analysis was performed using GraphPad Prism version 5. We assessed whether any of the measured patient or tumour variables were predictive of sentinel lymph node status. We also assessed whether these variables and particularly sentinel lymph node status had any significant effect on disease free and overall survival. Categorical variables such as melanoma type, melanoma site, Clark level, ulceration status, and mitotic rate were analysed using the chi-squared test. Continuous variables such as age and Breslow thickness were analysed using the Student t-test. Disease free (DFS) and overall survival (OS) were analysed using the Kaplan Meier method followed by the log rank test. Results with P < 0.05 were judged to be statistically significant.
Between 2000 and 2010, there were 195 patients with thick (4 mm) melanoma. There were 100 females and 95 males. Median age was 74 (range 18e98). Details according to sentinel lymph node status are shown in Table 1. Neither age nor sex was predictive of sentinel lymph node status, DFS or OS. Median follow-up was 36.8 months (range 1.4e121.9 months).
Tumour characteristics From the 195 patients, 55% of melanomas were nodular, 17% were superficial spreading (SSMM), 11% were lentigo maligna melanoma (LMM), 2% were amelanotic, 6% were spindle cell/desmoplastic and 4% were spitzoid. In 5% of cases, the results were unavailable. In 45% of patients, the melanoma was located on an extremity, 22% were located on the trunk and 28% were diagnosed on the head and neck. In the remaining 5% of patients, this information was unavailable. In 5% of cases, the tumour was reported as Clark level III, 54% were Clark level IV and 30% were Clark level V. In 12% of cases, this information was absent from the histology report. Ulceration was present at the time of presentation in 64% of patients and was absent in 23%. In 8% of cases ulceration was judged to be incipient. In 5% of cases, this information was not available. Median mitotic rate was 9 (range <1e100). Median Breslow thickness was 7 mm (range 4.1e50 mm). Tumour variable sub-groups were amalgamated for statistical analysis. Tumour characteristics specific to SLNB status are shown in Table 1. None of the patient or tumour characteristics (sex, age, melanoma type, melanoma site, Clark level, ulceration status, mitotic rate or Breslow thickness) reached statistical significance with the Chi squared test or paired student t-test with regards to predicting sentinel lymph node status (Table 1). Breslow thickness and SLNB status were the only variables on univariate analysis to have significant correlation with DFS (Table 2). Breslow thickness was the only variable to retain significance for OS (Table 3). 5-year DFS in patients with 4e6 mm Breslow thickness melanoma was 63.5%. For patients with >6 mm Breslow thickness melanoma, 5-year DFS was 32.9% (P Z 0.004; see Figure 1). OS at 5 years was 73.9% and 54.7% for those patients with 4e6 mm and >6 mm Breslow thickness melanoma respectively (P Z 0.02; see Figure 2).
Sentinel lymph node biopsy From the 195 patients with thick melanoma, 66 (34%) patients had sentinel lymph node biopsy. 44 (67%) of these patients had negative results and in the remaining 22 (33%) patients, the result was positive for metastatic melanoma. Recurrence was more common in those with positive sentinel lymph node biopsy (11 of 22 patients/50%) compared to those with negative SLNB (10 of 44 patients/ 23%). From those patients in the SLNB eve group who suffered a recurrence, 30% suffered local or in transit recurrence, 20% had regional recurrence and 50% had
1398 Table 1
N.G. Fairbairn et al. Patient and tumour characteristics as predictors of sentinel lymph node status. SLNB eve (n Z 44)
SLNB þve (n Z 22)
No SLNB (n Z 129)
Total (n Z 195)
Sex (no./%) Male Female
19 (43) 25 (57)
14 (64) 8 (36)
62 (48) 67 (52)
95 (49) 100 (51)
Age (years) Mean Median
59.1 17.1 61
54.1 18.4 60
75.7 12.7 78
69.5 16.9 74
Melanoma type (no./%) Nodular Superficial spreading Other Unknown
24 (55) 10 (23) 8 (18) 2 (5)
15 (68) 2 (9) 5 (23) 0 (0)
69 (53) 22 (17) 30 (23) 8 (6)
108 (55) 34 (17) 43 (22) 10 (5)
Melanoma site (no./%) Extremity Trunk Head and neck Unknown
25 (57) 13 (30) 5 (11) 1 (2)
11 (50) 7 (32) 3 (14) 1 (5)
52 (40) 23 (18) 46 (36) 8 (6)
88 43 54 10
Clark level (no./%) III IV V Unknown
3 (7) 28 (64) 11 (25) 2 (5)
1 (5) 13 (59) 7 (32) 1 (5)
5 (4) 64 (50) 40 (31) 20 (16)
9 (5) 105 (54) 58 (30) 23 (12)
Ulceration (no./%) Present Absent Incipient Unknown
24 (54) 14 (32) 1 (2) 5 (11)
13 (59) 6 (27) 2 (9) 1 (5)
88 (68) 24 (19) 13 (10) 4 (3)
125 (64) 44 (23) 16 (8) 10 (5)
Mitotic rate (no./%) 0e10 11e20 >20 Unknown Median
26 (59) 5 (11) 4 (9) 9 (20) 7
13 (59) 3 (14) 2 (9) 4 (18) 9
38 34 11 46 10
77 42 17 59 9
Breslow thickness Mean SD Median
6.9 3.54 5.75
6.9 2.84 5.5
9.7 6.33 7.9
Characteristic
P value (predictive of SN status) 0.19
0.31
0.75
0.88 (45) (22) (28) (5) 0.62
0.78
0.69 (29) (26) (9) (36)
(39) (22) (9) (30) 0.42
distant recurrence. From those patients in the SLNB þve group who suffered a recurrence, 36% had local or in transit recurrence, 9% had regional recurrence and 55% had distant recurrence (Table 4). There was no significant relationship between SLNB status and site of recurrence (P Z 0.13). 5-year disease free survival was 64.1% in the SLNB eve group and 35.4% in the SLNB þve group (log rank p Z 0.01; see Figure 3). Overall 5-year survival was 70.3% in SLNB eve patients compared to 63.7% in SLNB þve patients (log rank p Z 0.66; see Figure 4). 129 patients did not undergo SLNB. None of these patients went on to have elective node dissection. 45 patients were judged to be unsuitable surgical candidates for SLNB and potential completion lymphadenectomy either due to age or the existence of medical comorbidities. Despite being offered the procedure, 25 patients made an informed decision not to undergo SLNB. In
8.7 5.6 7
31 patients, the reasons were not clearly documented although there was no evidence of any medical or disease specific factors that would have rendered them unfavourable surgical candidates. In 28 patients, SLNB was not performed due to the presence of regional node metastases. In 12 of these patients regional disease was detected at the time of presentation or wide local excision. In the remaining 16 patients, regional disease was detected between 12 and 14 months after initial presentation. All of these patients proceeded to regional node clearance. 6 of these patients originally elected not to undergo SLNB.
Discussion Sentinel lymph node biopsy is recommended by the American Joint Committee on Cancer (AJCC) Melanoma Staging
SLNB in thick melanoma Table 2
1399
Factors associated with disease free survival.
Variable
No. (%) Patients
Age (n Z 155) 0e50 >50
20 (13) 135 (87)
Sex (n Z 151) Male Female
75 (50) 76 (50)
SLNB (n Z 66) Positive Negative
22 (33) 44 (67)
Breslow thickness (n Z 150) 4e6 mm >6 mm Clark level (n Z 150) IeIII IVeV Mitotic rate (n Z 130) 0e5 >5 Ulceration (n Z 147) Present Absent Melanoma site (n Z 150) Extremity Non-extremity Histology (n Z 151) Nodular Other
Table 3
Factors associated with overall survival.
Hazard ratio (95% CI)
P value
Variable
0.64 (0.35e1.19)
0.16
Age (n Z 155) 0e50 >50
20 (13) 135 (87)
Sex (n Z 151) Male Female
75 (50) 76 (50)
SLNB (n Z 66) Positive Negative
22 (33) 44 (67)
1.15 (0.72e1.82)
3.03 (1.24e7.34)
0.51 (0.32e0.81)
0.56
0.01
0.004
61 (41) 89 (59) 0.44 (0.25e1.42)
0.19
8 (5) 142 (95) 0.79 (0.40e1.60)
0.47
50 (38) 80 (62) 1.34 (0.81e2.19)
0.25
102 (69) 45 (31) 1.28 (0.82e1.94)
0.52
85 (57) 65 (43) 1.22 (0.73e2.03)
0.59
95 (63) 56 (37)
Committee as a staging procedure in patients where this information will be useful for planning subsequent management. The current guidelines advise SLNB in those with T2, T3 and T4 melanoma with clinically negative regional nodes and in those with 1b melanoma and the coexistence of mitotic rate >1 or ulceration.14 The information gained from SLNB guides decisions on regional node clearance, adjuvant therapy, entry into clinical trials, and forms the basis of informed discussions between patient and surgeon regarding likely prognosis. Establishing reliable predictive factors of sentinel lymph node status has helped shape the current guidelines. Younger patient age, high mitotic rate, the presence of ulceration, lymphovascular invasion, regression, satellitosis and tumour location on the trunk or lower extremity have all been linked with increasing rates of nodal micrometastases.9e11 Breslow thickness remains the most important of these factors.12,13
Breslow thickness (n Z 150) 4e6 mm >6 mm Clark level (n Z 150) IeIII IVeV Mitotic rate (n Z 130) 0e5 >5 Ulceration (n Z 147) Present Absent Melanoma site (n Z 150) Extremity Non-extremity Histology (n Z 151) Nodular Other
No. (%) Patients
Hazard ratio (95% CI)
P value
0.61 (0.26e1.41)
0.25
0.98 (0.53e1.80)
0.94
1.07 (0.40e2.89)
0.66
0.49 (0.27e0.91)
0.02
0.51 (0.19e1.70)
0.31
0.66 (0.38e1.54)
0.39
1.66 (0.88e3.14)
0.12
1.33 (0.71e2.34)
0.55
1.64 (0.85e3.20)
0.28
61 (41) 89 (59)
8 (5) 142 (95)
50 (38) 80 (62)
102 (69) 45 (31)
85 (57) 65 (43)
95 (63) 56 (37)
We were unable to show any significant correlation between the measured patient or tumour variables and the status of the sentinel lymph node. Many conflicting results exist in the literature on the predictive value of these variables. It is likely that variations in data collection, histological techniques and reporting are responsible. A notable problem in our study was that a significant proportion of patients entered into the database at the beginning of the 10-year period had no record of variables such as mitotic rate in the histology report. Other potential variables such as lymphovascular invasion, regression and satellitosis were largely absent. The lack of numbers led to us excluding these variables from our final analysis. A recent multi-institutional study has shown that the presence of micrometastases, number of tumour containing lymph nodes, mitotic rate, tumour thickness, patient age, ulceration and anatomic site of primary tumour were
1400
N.G. Fairbairn et al. Table 4 Recurrence, survival and association between recurrence site and SLN status.
Figure 1 5-year DFS according to Breslow thickness. 5-year DFS for 4e6 mm Breslow thickness melanoma was 63.5%. 5-year DFS for >6 mm Breslow thickness melanoma was 32.9% (P Z 0.004).
all significant predictors of patient survival.15 The status of the sentinel node however is widely regarded as the most significant predictor of disease free survival and overall survival.2,7,8,16e18 The only prospective, randomized controlled trial to date involved those patients with intermediate thickness melanoma.19 This showed no significant difference in overall survival between those undergoing wide local excision, sentinel lymph biopsy and immediate lymphadenectomy (in those with positive nodes) and those patients undergoing wide local excision, nodal observation and lymphadenectomy (in those developing clinically evident regional disease). However, within the biopsy group, patients with positive sentinel lymph nodes were found to have a significantly reduced 5-year survival. In those patients with positive sentinel lymph nodes, 5-year survival was also significantly reduced in those with delayed lymphadenectomy compared to immediate clearance. In addition, the mean number of tumourinvolved nodes was significantly greater in the observation group, suggesting disease progression in the group not undergoing sentinel lymph node biopsy. Although these
Figure 2 OS according to Breslow thickness. The 5-year OS in patients with 4e6 mm Breslow thickness melanoma was 73.9%. The 5-year OS for patients with >6 mm Breslow thickness melanoma was 54.7% (P Z 0.02).
SLNB eve (n Z 44)
SLNB þve (n Z 22)
P value
Total Recurrence
10 (23%)
11 (50%)
0.13
Local and in-transit Regional Distant
3 (30%) 2 (20%) 5 (50%)
4 (36%) 1 (9%) 6 (55%)
5-year DFS 5-year OS
61.2% 70.3%
35.4% 63.7%
0.01 0.66
results are not representative of a population with thick melanoma, this represents the first prospective randomised controlled trial indicating a survival advantage for those undergoing SLNB. Our results show that Breslow thickness and sentinel lymph biopsy are the only variables significantly predicting DFS. OS was predicted by Breslow thickness alone. The lack of significance between sentinel lymph node status and overall survival is in keeping with the results of Essner although these results are not consistent with many other recent studies showing a significant relationship.20 It is possible that our small sample size may be responsible. Only 66 (34%) patients with thick melanoma in our unit underwent SLNB over the 10-year period. Decisions were very much based on an individual basis and as with any intervention, reflected fitness for surgery, stage of disease and ultimately personal choice. 19% (25/129) of patients elected not to have a SLNB and, although it was not possible to establish the exact information provided, it is likely that the surgeon responsible at that time emphasised the lack of overall survival benefit associated with the procedure. It is only relatively recently that evidence has emerged suggesting a possible survival benefit from SLNB in thick melanoma. It will be of great interest to follow how this evidence influences the number of SLNB performed for thick melanoma in our unit in the future. Recurrence was more common in the SLNB þve group compared with the SLNB eve group. Table 4 shows that the most common site of recurrence in both the SLNB þve and
Figure 3 5-year DFS according to SLNB status. The 5-year DFS in the SLNB eve group was 64.1%. The disease free survival in the SLNB þve group was 35.4% (log rank p Z 0.01).
SLNB in thick melanoma
1401
Acknowledgements We would like to thank the Scottish Melanoma Group for their help and assistance during this project.
References
Figure 4 5-year OS according to SLNB status. The 5-year OS in the SLNB eve group was 70.3%. The 5-year OS in the SLNB þve group was 63.7% (log rank p Z 0.66).
SLNB eve groups was distant metastases. This is consistent with other reports in the literature and with the notion that increasing Breslow thickness predicts a greater incidence of distant metastatic spread.21 Patients with positive sentinel lymph node biopsy and who went on to have regional node clearance had a lower incidence of regional node recurrence compared with those patients in the SLNB eve group who did not have regional node clearance (9% vs. 20% respectively). The relationship between site of recurrence and sentinel lymph node status was not statistically significant. Similar findings have been reported in previous studies and may reflect the fact that a cleared regional drainage basin is less likely to develop a recurrence. Quoting percentages with such a small sample size may be a little misleading but the results remain consistent with previous studies. Despite having a negative SLNB, 2 (20%) patients developed regional metastases at the site of biopsy. This equates to a false negative rate of 5% (2/44). False negative sentinel lymph node biopsy may be explained by nuclear medicine, surgical or pathological error and our results are comparable to others.22
Conclusion Patients with thick melanoma in our unit over the last 10-years with negative sentinel lymph node biopsy can expect a significantly improved 5-year recurrence free survival. No significant improvement in overall survival has been demonstrated although this is not consistent with many other larger studies and may be related to our small sample size. Breslow thickness has been shown to predict improved 5-year disease free and overall survival in this cohort of patients. Despite our inability to show overall survival benefit, sentinel lymph node biopsy provides useful prognostic information and helps direct subsequent management in patients with thick melanoma.
Conflict of interest/funding None.
1. Morton DL, Cochrane AJ, Thompson JF, et al. Sentinel lymph node biopsyfor early stage melanoma: accuracy and morbidity in MSLT-1, an international multi-centre trial. Ann Surg 2005; 242:302e11. 2. Gershenwald JE, Thomson W, Mansfield PF, et al. Multi-institutional melanoma lymphatic mapping experience: the prognostic value of sentinel node status in 612 stage I or II melanoma patients. J Clin Oncol 1999;17(3):976e83. 3. Thompson JF, Shaw HM. The prognosis of patients with thick primary melanoma: is regional lymph node status relevant, and does removing positive regional nodes influence outcome? Ann Surg Oncol 2002;9(8):719e22. 4. Gutzmer R, Al Ghazal M, Geerlings H, et al. Sentinel lymph node biopsy in melanoma delays recurrence but does not change melanoma related survival: a retrospective analysis of 673 patients. Br J Dermatol 2005;153:1137e41. 5. Gershenwald JE, Mansfield PF, Lee JE, et al. Role for lymphatic mapping and seninel lymph node biopsy in patients with thick (4 mm) primary melanoma. Ann Surg Oncol 2000;7(2):160e5. 6. Jacobs IA, Chang CK, Salti GI. Role of sentinel lymph node biopsy in patients with thick (>4 mm) primary melanoma. Am Surg 2004;70(1):59e62. 7. Carlson GW, Murray DR, Hestley A, et al. Sentinel lymph node mapping for thick (4 mm) melanoma: should we be doing it? Ann Surg Oncol 2003;10(4):408e15. 8. Gajdos C, Griffith KA, Wong SL, et al. Is there a benefit to sentinel lymph node biopsy in patients with T4 melanoma? Cancer 2009;115(24):5752e60. 9. Paek SC, Griffith KA, Johnson TM, et al. The impact of factors beyond Breslow depth on predicting sentinel lymph node positivity in melanoma. Cancer 2007;109:100e8. 10. Kruper LL, Spitz FR, Czerniecki BJ, et al. Predicting sentinel node status in AJCC stage 1/11 primary cutaneous melanoma. Cancer 2006;107:2436e45. 11. Nguyen CL, McClay EF, Cole DJ, et al. Melanoma thickness and histology predict sentinel lymph node status. Am J Surg 2001; 181:8e11. 12. Clary BM, Brady MS, Lewis JJ, et al. 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(2):250e8. 13. Thomson JF, Shaw HM. Should tumour mitotic rate and patient age, as well as tumour thickness, be used to elect melanoma patients for sentinel lymph node biopsy? Ann Surg Oncol 2004; 11(3):233e5. 14. Balch CM, Gershenwald JE, Soong SJ, et al. Final version of 2009 AJCC melanoma staging and classification. J Clin Oncol 2009;27(36):6199e206. 15. Balch CM, Gershenwald JE, Soong SJ, et al. Multivariate analysis of prognosptic factors among 2313 patients with stage II melanoma: comparison of nodal micrometastases versus macrometastases. J Clin Oncol 2010;28(14):2452e9. 16. Rondelli F, Vedovati M, Becattini C, et al. Prognostic role sentinel lymph node biopsy patient thick melanoma a metaanalysis. J Eur Acad Dermatol Venereol 2011 [Epub ahead of print]. 17. Meguerditchian AN, Asubonteng K, Young C, et al. Thick primary melanoma has a heterogenous tumour biology: an institutional series. World J Surg Oncol 2011 Apr 14;9:40.
1402 18. Scoggins CR, Bowen AL, Martin RC, et al. Prognostic information from sentinel lymph node biopsy in patients with thick melanoma. 2nd. Arch Surg 2010 Jul;145(7):622e7. 19. Morton DL, Thompson JF, Cochran AJ, et al. Sentinel-node biopsy or nodal observation in melanoma. N Eng J Med 2006 Sep 28;355(13):1307e17. 20. Essner R, Chung MH, Bleicher R, et al. Prognostic implications of thick (4 mm) melanoma in the era of intraoperative lymphatic mapping and sentinel lymphadenectomy. Ann Surg Oncol 2002;9(8):754e61.
N.G. Fairbairn et al. 21. Essner R, Conforti A, Kelly 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(5): 442e9. 22. Nowecki ZI, Rutkowski P, Nasierowska-Guttmejer A, et al. Survival analysis and clinicopathological factors associated with false-negative sentinel lymph node biopsy findings in patients with cutaneous melanoma. Ann Surg Oncol 2006 Dec; 13(12):1655e63.
Sentinel lymph node biopsy in thick malignant melanoma: A 10-year single unit experience* Neil G. Fairbairn*, Georgios Orfaniotis, Mark Butterworth Department of Plastic Surgery, St John’s Hospital, Livingston, West Lothian, EH54 6PP, Scotland, UK Received 18 November 2011; accepted 12 April 2012
KEYWORDS Sentinel lymph node biopsy; Thick melanoma; Survival
Summary Between the years 2000e2010, 195 patients were diagnosed with 4 mm Breslow thickness malignant melanoma in our unit. Median follow-up was 36.8 months. 49% of patients were male and 51% were female. Median age was 74 years. The commonest melanoma type was nodular (55%). The commonest tumour location was on the extremity (45%). 64% of tumours were ulcerated. Median mitotic rate was 9. Median Breslow thickness was 7 mm 66 patients underwent sentinel lymph node biopsy. 44 (67%) patients had negative results and the remaining 22 (33%) patients were positive for metastatic melanoma. There was no statistically significant correlation between any of the patient or tumour variables (age, sex, melanoma type, melanoma site, Clark level, Breslow thickness, mitotic rate, ulceration) and sentinel lymph node status. Patients with Breslow thickness melanoma of <6 mm had a significantly better 5-year disease free and overall survival compared with those patients with >6 mm Breslow thickness melanoma (63.5% vs. 32.9%; P Z 0.004 and 73.9% vs. 54.7%; P Z 0.02 respectively). Recurrence rate was 50% in those with positive sentinel lymph node biopsy compared to 23% in those with negative results. Distant recurrence was the commonest in both groups. 5-year disease free survival was 64.1% in the SLNB eve group and 35.4% in the SLNB þve group (P Z 0.01). There was no significant difference in overall survival between the SLNB eve and SLNB þve groups (70.3% vs. 63.7% respectively; P Z 0.66). We conclude that sentinel lymph node biopsy in our unit has provided no survival benefit in those with thick melanoma over the past 10 years but is an important predictor of recurrence free survival. Breslow thickness remains an important predictor of disease free and overall survival in thick melanoma. ª 2012 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved.
*
Article presented at BAPRAS Winter Scientific Meeting 2011. * Tel.: þ44 7764514281; fax: þ44 1506523034. E-mail address: [email protected] (N.G. Fairbairn).
1748-6815/$ - see front matter ª 2012 British Association of Plastic, Reconstructive and Aesthetic Surgeons. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.bjps.2012.04.019
SLNB in thick melanoma
1397
Introduction
Results
Sentinel lymph node biopsy (SLNB) is the most sensitive and accurate investigative modality for establishing regional node status in patients with melanoma.1 The status of the regional node has established itself as the most powerful predictor of recurrence and overall survival in patients with intermediate thickness melanoma.2 The role of SLNB in patients with thick melanoma is more uncertain. Opponents argue that the procedure is futile in this cohort of patients due to the high incidence of distant metastases at presentation and the overall poor survival of patients.3 This has been supported by studies showing no benefit in overall survival according to the regional node status.4 However, recent evidence suggests that patients with thick melanoma are a heterogenous group with regards to tumour behaviour and prognosis. Those with negative SLNB may represent a subset of thick melanoma patients with significantly better prognosis compared to those with regional node metastases.5e8 Controversy around the subject of SLNB in thick melanoma continues and the aim of our retrospective single centre study is to report how SLNB and other selected variables have affected outcome in our unit over the last 10 years.
Patient demographics
Methods Study design and patients Approval for the study was gained from the local research and ethics committee and the Scottish Melanoma Group. We conducted a retrospective analysis of the regional melanoma database provided by the Scottish Melanoma Group. All patients presenting between 2000 and 2010 with thick (4 mm Breslow) malignant melanoma were included in the study. Information including patient date of birth, age at diagnosis, date of diagnosis, melanoma type, melanoma site, Breslow thickness, ulceration status, Clark level, lymphovascular invasion, regression, mitotic rate, date and result of SLNB, date and result of regional node dissection, recurrence date, recurrence site, follow-up status and date of death were recorded. Histology results were obtained from our local electronic patient management system. All information was compiled into an Excel database for analysis.
Statistics All statistical analysis was performed using GraphPad Prism version 5. We assessed whether any of the measured patient or tumour variables were predictive of sentinel lymph node status. We also assessed whether these variables and particularly sentinel lymph node status had any significant effect on disease free and overall survival. Categorical variables such as melanoma type, melanoma site, Clark level, ulceration status, and mitotic rate were analysed using the chi-squared test. Continuous variables such as age and Breslow thickness were analysed using the Student t-test. Disease free (DFS) and overall survival (OS) were analysed using the Kaplan Meier method followed by the log rank test. Results with P < 0.05 were judged to be statistically significant.
Between 2000 and 2010, there were 195 patients with thick (4 mm) melanoma. There were 100 females and 95 males. Median age was 74 (range 18e98). Details according to sentinel lymph node status are shown in Table 1. Neither age nor sex was predictive of sentinel lymph node status, DFS or OS. Median follow-up was 36.8 months (range 1.4e121.9 months).
Tumour characteristics From the 195 patients, 55% of melanomas were nodular, 17% were superficial spreading (SSMM), 11% were lentigo maligna melanoma (LMM), 2% were amelanotic, 6% were spindle cell/desmoplastic and 4% were spitzoid. In 5% of cases, the results were unavailable. In 45% of patients, the melanoma was located on an extremity, 22% were located on the trunk and 28% were diagnosed on the head and neck. In the remaining 5% of patients, this information was unavailable. In 5% of cases, the tumour was reported as Clark level III, 54% were Clark level IV and 30% were Clark level V. In 12% of cases, this information was absent from the histology report. Ulceration was present at the time of presentation in 64% of patients and was absent in 23%. In 8% of cases ulceration was judged to be incipient. In 5% of cases, this information was not available. Median mitotic rate was 9 (range <1e100). Median Breslow thickness was 7 mm (range 4.1e50 mm). Tumour variable sub-groups were amalgamated for statistical analysis. Tumour characteristics specific to SLNB status are shown in Table 1. None of the patient or tumour characteristics (sex, age, melanoma type, melanoma site, Clark level, ulceration status, mitotic rate or Breslow thickness) reached statistical significance with the Chi squared test or paired student t-test with regards to predicting sentinel lymph node status (Table 1). Breslow thickness and SLNB status were the only variables on univariate analysis to have significant correlation with DFS (Table 2). Breslow thickness was the only variable to retain significance for OS (Table 3). 5-year DFS in patients with 4e6 mm Breslow thickness melanoma was 63.5%. For patients with >6 mm Breslow thickness melanoma, 5-year DFS was 32.9% (P Z 0.004; see Figure 1). OS at 5 years was 73.9% and 54.7% for those patients with 4e6 mm and >6 mm Breslow thickness melanoma respectively (P Z 0.02; see Figure 2).
Sentinel lymph node biopsy From the 195 patients with thick melanoma, 66 (34%) patients had sentinel lymph node biopsy. 44 (67%) of these patients had negative results and in the remaining 22 (33%) patients, the result was positive for metastatic melanoma. Recurrence was more common in those with positive sentinel lymph node biopsy (11 of 22 patients/50%) compared to those with negative SLNB (10 of 44 patients/ 23%). From those patients in the SLNB eve group who suffered a recurrence, 30% suffered local or in transit recurrence, 20% had regional recurrence and 50% had
1398 Table 1
N.G. Fairbairn et al. Patient and tumour characteristics as predictors of sentinel lymph node status. SLNB eve (n Z 44)
SLNB þve (n Z 22)
No SLNB (n Z 129)
Total (n Z 195)
Sex (no./%) Male Female
19 (43) 25 (57)
14 (64) 8 (36)
62 (48) 67 (52)
95 (49) 100 (51)
Age (years) Mean Median
59.1 17.1 61
54.1 18.4 60
75.7 12.7 78
69.5 16.9 74
Melanoma type (no./%) Nodular Superficial spreading Other Unknown
24 (55) 10 (23) 8 (18) 2 (5)
15 (68) 2 (9) 5 (23) 0 (0)
69 (53) 22 (17) 30 (23) 8 (6)
108 (55) 34 (17) 43 (22) 10 (5)
Melanoma site (no./%) Extremity Trunk Head and neck Unknown
25 (57) 13 (30) 5 (11) 1 (2)
11 (50) 7 (32) 3 (14) 1 (5)
52 (40) 23 (18) 46 (36) 8 (6)
88 43 54 10
Clark level (no./%) III IV V Unknown
3 (7) 28 (64) 11 (25) 2 (5)
1 (5) 13 (59) 7 (32) 1 (5)
5 (4) 64 (50) 40 (31) 20 (16)
9 (5) 105 (54) 58 (30) 23 (12)
Ulceration (no./%) Present Absent Incipient Unknown
24 (54) 14 (32) 1 (2) 5 (11)
13 (59) 6 (27) 2 (9) 1 (5)
88 (68) 24 (19) 13 (10) 4 (3)
125 (64) 44 (23) 16 (8) 10 (5)
Mitotic rate (no./%) 0e10 11e20 >20 Unknown Median
26 (59) 5 (11) 4 (9) 9 (20) 7
13 (59) 3 (14) 2 (9) 4 (18) 9
38 34 11 46 10
77 42 17 59 9
Breslow thickness Mean SD Median
6.9 3.54 5.75
6.9 2.84 5.5
9.7 6.33 7.9
Characteristic
P value (predictive of SN status) 0.19
0.31
0.75
0.88 (45) (22) (28) (5) 0.62
0.78
0.69 (29) (26) (9) (36)
(39) (22) (9) (30) 0.42
distant recurrence. From those patients in the SLNB þve group who suffered a recurrence, 36% had local or in transit recurrence, 9% had regional recurrence and 55% had distant recurrence (Table 4). There was no significant relationship between SLNB status and site of recurrence (P Z 0.13). 5-year disease free survival was 64.1% in the SLNB eve group and 35.4% in the SLNB þve group (log rank p Z 0.01; see Figure 3). Overall 5-year survival was 70.3% in SLNB eve patients compared to 63.7% in SLNB þve patients (log rank p Z 0.66; see Figure 4). 129 patients did not undergo SLNB. None of these patients went on to have elective node dissection. 45 patients were judged to be unsuitable surgical candidates for SLNB and potential completion lymphadenectomy either due to age or the existence of medical comorbidities. Despite being offered the procedure, 25 patients made an informed decision not to undergo SLNB. In
8.7 5.6 7
31 patients, the reasons were not clearly documented although there was no evidence of any medical or disease specific factors that would have rendered them unfavourable surgical candidates. In 28 patients, SLNB was not performed due to the presence of regional node metastases. In 12 of these patients regional disease was detected at the time of presentation or wide local excision. In the remaining 16 patients, regional disease was detected between 12 and 14 months after initial presentation. All of these patients proceeded to regional node clearance. 6 of these patients originally elected not to undergo SLNB.
Discussion Sentinel lymph node biopsy is recommended by the American Joint Committee on Cancer (AJCC) Melanoma Staging
SLNB in thick melanoma Table 2
1399
Factors associated with disease free survival.
Variable
No. (%) Patients
Age (n Z 155) 0e50 >50
20 (13) 135 (87)
Sex (n Z 151) Male Female
75 (50) 76 (50)
SLNB (n Z 66) Positive Negative
22 (33) 44 (67)
Breslow thickness (n Z 150) 4e6 mm >6 mm Clark level (n Z 150) IeIII IVeV Mitotic rate (n Z 130) 0e5 >5 Ulceration (n Z 147) Present Absent Melanoma site (n Z 150) Extremity Non-extremity Histology (n Z 151) Nodular Other
Table 3
Factors associated with overall survival.
Hazard ratio (95% CI)
P value
Variable
0.64 (0.35e1.19)
0.16
Age (n Z 155) 0e50 >50
20 (13) 135 (87)
Sex (n Z 151) Male Female
75 (50) 76 (50)
SLNB (n Z 66) Positive Negative
22 (33) 44 (67)
1.15 (0.72e1.82)
3.03 (1.24e7.34)
0.51 (0.32e0.81)
0.56
0.01
0.004
61 (41) 89 (59) 0.44 (0.25e1.42)
0.19
8 (5) 142 (95) 0.79 (0.40e1.60)
0.47
50 (38) 80 (62) 1.34 (0.81e2.19)
0.25
102 (69) 45 (31) 1.28 (0.82e1.94)
0.52
85 (57) 65 (43) 1.22 (0.73e2.03)
0.59
95 (63) 56 (37)
Committee as a staging procedure in patients where this information will be useful for planning subsequent management. The current guidelines advise SLNB in those with T2, T3 and T4 melanoma with clinically negative regional nodes and in those with 1b melanoma and the coexistence of mitotic rate >1 or ulceration.14 The information gained from SLNB guides decisions on regional node clearance, adjuvant therapy, entry into clinical trials, and forms the basis of informed discussions between patient and surgeon regarding likely prognosis. Establishing reliable predictive factors of sentinel lymph node status has helped shape the current guidelines. Younger patient age, high mitotic rate, the presence of ulceration, lymphovascular invasion, regression, satellitosis and tumour location on the trunk or lower extremity have all been linked with increasing rates of nodal micrometastases.9e11 Breslow thickness remains the most important of these factors.12,13
Breslow thickness (n Z 150) 4e6 mm >6 mm Clark level (n Z 150) IeIII IVeV Mitotic rate (n Z 130) 0e5 >5 Ulceration (n Z 147) Present Absent Melanoma site (n Z 150) Extremity Non-extremity Histology (n Z 151) Nodular Other
No. (%) Patients
Hazard ratio (95% CI)
P value
0.61 (0.26e1.41)
0.25
0.98 (0.53e1.80)
0.94
1.07 (0.40e2.89)
0.66
0.49 (0.27e0.91)
0.02
0.51 (0.19e1.70)
0.31
0.66 (0.38e1.54)
0.39
1.66 (0.88e3.14)
0.12
1.33 (0.71e2.34)
0.55
1.64 (0.85e3.20)
0.28
61 (41) 89 (59)
8 (5) 142 (95)
50 (38) 80 (62)
102 (69) 45 (31)
85 (57) 65 (43)
95 (63) 56 (37)
We were unable to show any significant correlation between the measured patient or tumour variables and the status of the sentinel lymph node. Many conflicting results exist in the literature on the predictive value of these variables. It is likely that variations in data collection, histological techniques and reporting are responsible. A notable problem in our study was that a significant proportion of patients entered into the database at the beginning of the 10-year period had no record of variables such as mitotic rate in the histology report. Other potential variables such as lymphovascular invasion, regression and satellitosis were largely absent. The lack of numbers led to us excluding these variables from our final analysis. A recent multi-institutional study has shown that the presence of micrometastases, number of tumour containing lymph nodes, mitotic rate, tumour thickness, patient age, ulceration and anatomic site of primary tumour were
1400
N.G. Fairbairn et al. Table 4 Recurrence, survival and association between recurrence site and SLN status.
Figure 1 5-year DFS according to Breslow thickness. 5-year DFS for 4e6 mm Breslow thickness melanoma was 63.5%. 5-year DFS for >6 mm Breslow thickness melanoma was 32.9% (P Z 0.004).
all significant predictors of patient survival.15 The status of the sentinel node however is widely regarded as the most significant predictor of disease free survival and overall survival.2,7,8,16e18 The only prospective, randomized controlled trial to date involved those patients with intermediate thickness melanoma.19 This showed no significant difference in overall survival between those undergoing wide local excision, sentinel lymph biopsy and immediate lymphadenectomy (in those with positive nodes) and those patients undergoing wide local excision, nodal observation and lymphadenectomy (in those developing clinically evident regional disease). However, within the biopsy group, patients with positive sentinel lymph nodes were found to have a significantly reduced 5-year survival. In those patients with positive sentinel lymph nodes, 5-year survival was also significantly reduced in those with delayed lymphadenectomy compared to immediate clearance. In addition, the mean number of tumourinvolved nodes was significantly greater in the observation group, suggesting disease progression in the group not undergoing sentinel lymph node biopsy. Although these
Figure 2 OS according to Breslow thickness. The 5-year OS in patients with 4e6 mm Breslow thickness melanoma was 73.9%. The 5-year OS for patients with >6 mm Breslow thickness melanoma was 54.7% (P Z 0.02).
SLNB eve (n Z 44)
SLNB þve (n Z 22)
P value
Total Recurrence
10 (23%)
11 (50%)
0.13
Local and in-transit Regional Distant
3 (30%) 2 (20%) 5 (50%)
4 (36%) 1 (9%) 6 (55%)
5-year DFS 5-year OS
61.2% 70.3%
35.4% 63.7%
0.01 0.66
results are not representative of a population with thick melanoma, this represents the first prospective randomised controlled trial indicating a survival advantage for those undergoing SLNB. Our results show that Breslow thickness and sentinel lymph biopsy are the only variables significantly predicting DFS. OS was predicted by Breslow thickness alone. The lack of significance between sentinel lymph node status and overall survival is in keeping with the results of Essner although these results are not consistent with many other recent studies showing a significant relationship.20 It is possible that our small sample size may be responsible. Only 66 (34%) patients with thick melanoma in our unit underwent SLNB over the 10-year period. Decisions were very much based on an individual basis and as with any intervention, reflected fitness for surgery, stage of disease and ultimately personal choice. 19% (25/129) of patients elected not to have a SLNB and, although it was not possible to establish the exact information provided, it is likely that the surgeon responsible at that time emphasised the lack of overall survival benefit associated with the procedure. It is only relatively recently that evidence has emerged suggesting a possible survival benefit from SLNB in thick melanoma. It will be of great interest to follow how this evidence influences the number of SLNB performed for thick melanoma in our unit in the future. Recurrence was more common in the SLNB þve group compared with the SLNB eve group. Table 4 shows that the most common site of recurrence in both the SLNB þve and
Figure 3 5-year DFS according to SLNB status. The 5-year DFS in the SLNB eve group was 64.1%. The disease free survival in the SLNB þve group was 35.4% (log rank p Z 0.01).
SLNB in thick melanoma
1401
Acknowledgements We would like to thank the Scottish Melanoma Group for their help and assistance during this project.
References
Figure 4 5-year OS according to SLNB status. The 5-year OS in the SLNB eve group was 70.3%. The 5-year OS in the SLNB þve group was 63.7% (log rank p Z 0.66).
SLNB eve groups was distant metastases. This is consistent with other reports in the literature and with the notion that increasing Breslow thickness predicts a greater incidence of distant metastatic spread.21 Patients with positive sentinel lymph node biopsy and who went on to have regional node clearance had a lower incidence of regional node recurrence compared with those patients in the SLNB eve group who did not have regional node clearance (9% vs. 20% respectively). The relationship between site of recurrence and sentinel lymph node status was not statistically significant. Similar findings have been reported in previous studies and may reflect the fact that a cleared regional drainage basin is less likely to develop a recurrence. Quoting percentages with such a small sample size may be a little misleading but the results remain consistent with previous studies. Despite having a negative SLNB, 2 (20%) patients developed regional metastases at the site of biopsy. This equates to a false negative rate of 5% (2/44). False negative sentinel lymph node biopsy may be explained by nuclear medicine, surgical or pathological error and our results are comparable to others.22
Conclusion Patients with thick melanoma in our unit over the last 10-years with negative sentinel lymph node biopsy can expect a significantly improved 5-year recurrence free survival. No significant improvement in overall survival has been demonstrated although this is not consistent with many other larger studies and may be related to our small sample size. Breslow thickness has been shown to predict improved 5-year disease free and overall survival in this cohort of patients. Despite our inability to show overall survival benefit, sentinel lymph node biopsy provides useful prognostic information and helps direct subsequent management in patients with thick melanoma.
Conflict of interest/funding None.
1. Morton DL, Cochrane AJ, Thompson JF, et al. Sentinel lymph node biopsyfor early stage melanoma: accuracy and morbidity in MSLT-1, an international multi-centre trial. Ann Surg 2005; 242:302e11. 2. Gershenwald JE, Thomson W, Mansfield PF, et al. Multi-institutional melanoma lymphatic mapping experience: the prognostic value of sentinel node status in 612 stage I or II melanoma patients. J Clin Oncol 1999;17(3):976e83. 3. Thompson JF, Shaw HM. The prognosis of patients with thick primary melanoma: is regional lymph node status relevant, and does removing positive regional nodes influence outcome? Ann Surg Oncol 2002;9(8):719e22. 4. Gutzmer R, Al Ghazal M, Geerlings H, et al. Sentinel lymph node biopsy in melanoma delays recurrence but does not change melanoma related survival: a retrospective analysis of 673 patients. Br J Dermatol 2005;153:1137e41. 5. Gershenwald JE, Mansfield PF, Lee JE, et al. Role for lymphatic mapping and seninel lymph node biopsy in patients with thick (4 mm) primary melanoma. Ann Surg Oncol 2000;7(2):160e5. 6. Jacobs IA, Chang CK, Salti GI. Role of sentinel lymph node biopsy in patients with thick (>4 mm) primary melanoma. Am Surg 2004;70(1):59e62. 7. Carlson GW, Murray DR, Hestley A, et al. Sentinel lymph node mapping for thick (4 mm) melanoma: should we be doing it? Ann Surg Oncol 2003;10(4):408e15. 8. Gajdos C, Griffith KA, Wong SL, et al. Is there a benefit to sentinel lymph node biopsy in patients with T4 melanoma? Cancer 2009;115(24):5752e60. 9. Paek SC, Griffith KA, Johnson TM, et al. The impact of factors beyond Breslow depth on predicting sentinel lymph node positivity in melanoma. Cancer 2007;109:100e8. 10. Kruper LL, Spitz FR, Czerniecki BJ, et al. Predicting sentinel node status in AJCC stage 1/11 primary cutaneous melanoma. Cancer 2006;107:2436e45. 11. Nguyen CL, McClay EF, Cole DJ, et al. Melanoma thickness and histology predict sentinel lymph node status. Am J Surg 2001; 181:8e11. 12. Clary BM, Brady MS, Lewis JJ, et al. 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(2):250e8. 13. Thomson JF, Shaw HM. Should tumour mitotic rate and patient age, as well as tumour thickness, be used to elect melanoma patients for sentinel lymph node biopsy? Ann Surg Oncol 2004; 11(3):233e5. 14. Balch CM, Gershenwald JE, Soong SJ, et al. Final version of 2009 AJCC melanoma staging and classification. J Clin Oncol 2009;27(36):6199e206. 15. Balch CM, Gershenwald JE, Soong SJ, et al. Multivariate analysis of prognosptic factors among 2313 patients with stage II melanoma: comparison of nodal micrometastases versus macrometastases. J Clin Oncol 2010;28(14):2452e9. 16. Rondelli F, Vedovati M, Becattini C, et al. Prognostic role sentinel lymph node biopsy patient thick melanoma a metaanalysis. J Eur Acad Dermatol Venereol 2011 [Epub ahead of print]. 17. Meguerditchian AN, Asubonteng K, Young C, et al. Thick primary melanoma has a heterogenous tumour biology: an institutional series. World J Surg Oncol 2011 Apr 14;9:40.
1402 18. Scoggins CR, Bowen AL, Martin RC, et al. Prognostic information from sentinel lymph node biopsy in patients with thick melanoma. 2nd. Arch Surg 2010 Jul;145(7):622e7. 19. Morton DL, Thompson JF, Cochran AJ, et al. Sentinel-node biopsy or nodal observation in melanoma. N Eng J Med 2006 Sep 28;355(13):1307e17. 20. Essner R, Chung MH, Bleicher R, et al. Prognostic implications of thick (4 mm) melanoma in the era of intraoperative lymphatic mapping and sentinel lymphadenectomy. Ann Surg Oncol 2002;9(8):754e61.
N.G. Fairbairn et al. 21. Essner R, Conforti A, Kelly 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(5): 442e9. 22. Nowecki ZI, Rutkowski P, Nasierowska-Guttmejer A, et al. Survival analysis and clinicopathological factors associated with false-negative sentinel lymph node biopsy findings in patients with cutaneous melanoma. Ann Surg Oncol 2006 Dec; 13(12):1655e63.