|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 25, No 7 (March 1), 2007: pp. 869-875 © 2007 American Society of Clinical Oncology. DOI: 10.1200/JCO.2006.08.9755 Tumor-Infiltrating Lymphocytes Predict Sentinel Lymph Node Positivity in Patients With Cutaneous Melanoma
From the Departments of Surgery, Biostatistics, and Pathology, Memorial Sloan-Kettering Cancer Center, New York, NY Address reprint requests to Mary S. Brady, MD, Gastric and Mixed Tumor Service, Memorial Sloan-Kettering Cancer Center, 1275 York Ave, H1212, New York, NY 10021; e-mail: bradym{at}mskcc.org
Purpose Tumor-infiltrating lymphocytes (TILs) are considered a manifestation of the host immune response to tumor, but the influence of TILs on outcome remains controversial. Studies evaluating the prognostic significance of TILs were published before routine examination of draining lymph nodes by sentinel lymph node (SLN) biopsy, the most important predictor of survival in patients with melanoma. The prognostic implications of TILs were re-evaluated in a large group of patients undergoing SLN biopsy at our institution. Patients and Methods All patients who underwent SLN mapping for primary cutaneous melanoma between January 1996 and July 2005 were evaluated. Univariate and multivariate analyses were performed to assess factors that predict SLN positivity and survival. Factors analyzed included Breslow thickness, ulceration, anatomic site, sex, Clark level, age, mitotic rate, and the presence (brisk or nonbrisk) or absence of TIL. Results Eight hundred eighty-seven patients underwent SLN mapping, and a SLN was identified in 875 patients (98.8%). The SLN was positive for tumor in 156 patients (17.6%). Multivariate analysis revealed that only Breslow thickness (P < .0001), ulceration (P = .0004), male sex (P = .03), and absent TILs (P = .0003) were independently predictive of the presence of SLN metastases. In melanomas with a brisk TIL infiltrate, the probability of a positive SLN was 3.9% as compared with 26.2% for melanomas in which TILs were absent. TILs were not an independent predictive factor for survival. Conclusion The absence of TILs, together with increasing Breslow thickness, presence of ulceration and male sex, predicts SLN metastasis in patients undergoing SLN biopsy for primary cutaneous melanoma.
The immune system is thought to play a significant role in the control of tumor growth in melanoma patients. Tumor-infiltrating lymphocytes (TILs) have long been considered a manifestation of the host immune response to tumor, but the immunologic implications of TILs have not been clarified, and the prognostic significance of TILs remains controversial. A number of investigators have reported that the presence of TILs in the vertical growth phase (VGP) of primary cutaneous melanomas is associated with a better prognosis1-6; however, other reports suggest no such relationship.7-9 These studies antedate the current era of routine examination of the draining lymph nodes by sentinel lymph node (SLN) biopsy. The presence of metastases in the SLN is the most important independent predictor of recurrence and survival in patients with melanoma,10 and SLN status is an important component of the revised American Joint Committee on Cancer staging system10 for patients with melanoma. Tumor thickness and ulceration are factors that consistently predict SLN positivity.11-14 A recent report by Kruper et al15 demonstrated that tumor thickness, mitotic rate, and the absence of TILs were independent predictors of SLN metastases in a multivariate analysis of 327 patients with primary melanoma more than 1.0 mm in Breslow depth. The primary objective of this study was to determine whether the absence or presence of TILs in primary cutaneous melanoma can predict the histologic status of the SLN. The secondary objective was to determine whether the presence or absence of TILs had an impact on recurrence or survival.
Study Population Approval to conduct the study was obtained from the institutional review board of the Memorial Sloan-Kettering Cancer Center (New York, NY). A prospective melanoma database was used to identify all patients with histologically confirmed cutaneous melanoma, clinically negative lymph nodes, and no evidence of distant disease who presented for surgical management of primary melanoma between January 1996 and July 2005. Only patients who underwent lymphatic mapping and a SLN biopsy were included. Demographic factors including age at diagnosis, sex, and site of primary tumor were prospectively collected in all patients. All patients underwent wide local excision with margins of 1 cm for melanomas with a Breslow depth of 1 mm or less, 1 to 2 cm for more than 1 mm and less than 2 mm depth, and at least 2 cm for a depth of 2 mm or more. Patients with positive SLNs were offered a completion lymphadenectomy. Patients did not routinely receive adjuvant therapy, but were offered standard or investigational immunotherapy when appropriate. Patients were followed at regular intervals for evidence of recurrence using physical examination and appropriate radiologic studies. The type of recurrence was classified as local, regional in transit, regional nodal, or distant. The first and all subsequent recurrences were recorded.
SLN Biopsy Technique
Pathologic Features
Statistical Analysis
Clinical and Pathologic Characteristics A total of 887 patients were included in the analysis. The clinical characteristics of the patients grouped by TIL status (absent, nonbrisk, and brisk) are summarized in Table 1. The median age of patients was 57 years, and 60% were male. The most common site of melanoma was extremity (51%), followed by trunk (33%), and head and neck (16%). Adjuvant therapy was administered to 92 patients (10%). There were no significant differences in the clinical characteristics between patients who had TILs present (brisk and nonbrisk) in their primary melanoma and those in whom they were absent (not shown).
The pathologic characteristics of the primary melanoma according to TIL status are summarized in Table 1. The median Breslow depth was 1.8 mm, and most were Clark level IV (76%). Ulceration was present in 256 patients (29%) and was significantly correlated with tumor thickness (Fig 1). TILs were present in 692 patients (78%), whereas only 51 patients (5.7%) had "brisk" TILs. The TIL status of the lesion was also correlated with Breslow and Clark levels. The mean lesion thickness for patients with brisk TILs was 1.71 mm compared with 2.89 mm for patients with TILs absent (P = .001) and, among patients with thick lesions (> 4 mm), there was a significantly smaller proportion of lesions demonstrating brisk TIL infiltrate. Patients with absent TILs were significantly more likely to have a positive SLN compared with patients with nonbrisk or brisk TILs (P = .0001). Completion lymphadenectomy was also significantly more common in patients with absent TILs (P < .0001). There was no difference in other variables evaluated, including the mitotic rate or the presence of ulceration, among patients with brisk, nonbrisk, and absent TILs (Table 1).
SLN Status A total of 887 patients underwent SLN mapping, and an SLN was identified in 875 patients (98.6%). At least one SLN was positive in 156 patients (17.6%). The median number of SLNs collected was two (range, 0 to 20). A completion lymphadenectomy was performed in 131 patients with a positive SLN (82% of all SLN-positive patients). Univariate analysis demonstrated that the absence of TILs, increasing Breslow thickness, Clark level, mitotic rate, presence of ulceration, and male sex showed a significant relationship to SLN positivity (Table 2). By multivariate logistic regression analysis, male sex, Breslow thickness, presence of ulceration, and absence of TILs were independently associated with the finding of a positive SLN (Table 2). When brisk and nonbrisk TILs were analyzed separately, both levels were significant predictors of a negative SLN compared with absent TILs, by univariate and multivariate analysis. For this reason, we have decided to collapse the variable into positive versus negative TILs for the remaining analyses. Independent of primary tumor thickness and ulceration, the proportion of patients with a positive SLN was 3.9% with a brisk infiltrate, 16.1% with a nonbrisk infiltrate, and 26.2% with absent TIL.
As tumor thickness increased from less than 1 mm to more than 4 mm, the percentage of patients with a positive SLN increased in both the absent and nonbrisk TILs groups but not in the brisk-TILs group (Fig 2). In addition, Figure 2 demonstrates that among patients with thin melanomas (  1 mm), those with absent TILs had a significantly higher proportion of positive SLNs (18%) compared with those with nonbrisk or brisk TIL (2%; P = .002).
Recurrence and Survival The histologic status of the SLN was the most significant predictor of DFS (Fig 3B) and OS (not shown). Patients with a negative SLN had a 5-year DFS of 80.0%, compared with 32.8% in patients with a positive SLN (P < .0001). When SLN status was not considered, the 5-year DFS was marginally better for patients with TIL present (brisk, 80%; nonbrisk 71.4%) in their melanoma primary compared with those in whom TILs were absent (67.1%), but this was not statistically significant (P = .25; Fig 3A). When patients were stratified by SLN status, there was no survival advantage present with TILs (Fig 3B). This was true when brisk and nonbrisk TIL groups were analyzed together, or separately.
The OS was also similar between patients with TILs absent and present in the primary lesion (not shown), with a 5-year OS of 75% and 76%, respectively. The percentage of patients who eventually experienced recurrence was 26% in the group of patients with absent TILs and 18% in both the brisk and nonbrisk TILs group (Table 3). The pattern of recurrence was different among patients in the three TIL subgroups. Patients with absent TILs were more likely to develop an in-transit or nodal recurrence (15.4%), compared with patients with brisk (3.9%) or nonbrisk TILs (9.7%), with no difference in the percentage of patients developing a local or distant recurrence (Table 3). In patients with a positive SLN, a regional recurrence occurred in 28.8%, whereas only 6.7% of patients with a negative SLN experienced regional recurrence (P < .0001). When patients were stratified by SLN status, the TIL status of the primary was no longer associated with the prevalence of regional recurrence (P = .1), indicating that the regional failure was likely due to SLN metastasis and not TILs.
This study demonstrates an independent relationship between the TIL status of the primary melanoma and the histologic status of the SLN. TIL was evaluated in a multivariate analysis and was found to be an independent predictor of SLN status, as were male sex, Breslow thickness, and ulceration (Table 2). We have demonstrated that TILs predict SLN status even when other highly significant parameters are considered. Although a number of studies have described the presence of a lymphocytic infiltrate in primary melanoma and correlated that observation with clinical outcome,3,6,7,9 Clark et al1 were the first to clearly define and classify lymphocytes infiltrating primary melanoma in the VGP. They reported that TILs were an independent predictor of OS in 264 patients with a primary melanoma that had entered the VGP. The 8-year OS was 88% for patients with brisk TILs, 75% for nonbrisk TILs, and 59% for absent TILs, with an adjusted odds ratios of 11.3, 3.5, and 1, respectively. In this study the impact of TILs on survival appeared to be most pronounced in thicker lesions and those with a higher mitotic rate. The prognostic significance of TIL initially reported by Clark et al1 was supported by two subsequent studies.2,5 Clemente et al2 found that, in 285 patients with primary cutaneous melanoma in the VGP, the 5-year OS for patients with a brisk infiltrate was 77%, compared with 53% and 37% for patients in the nonbrisk and absent TIL groups. The impact of TILs were more pronounced in patients with lesions less than 6 mm in depth, suggesting that the prognostic significance was lost in very thick lesions. Tuthill et al5 found that a brisk TIL infiltrate predicted an improved survival (100% at 5 years) in 259 patients with primary cutaneous melanoma in the VGP. In contrast, Barnhill et al8 did not demonstrate a survival advantage resulting from TILs in the primary lesion. In this population-based study of 548 patients with localized cutaneous melanoma, the presence of TIL, as defined by Clark et al,1 did not predict an improved survival (5-year OS of 86% for TILs present v 91% for TILs absent). In contrast to prior studies, patients with melanoma in both the radial growth phase and VGP were included. In addition, only 25% of patients had lesions thicker than 1.7 mm. Lack of consensus regarding the prognostic significance of TIL in primary melanoma may be attributed to differences in the patient populations being investigated. Studies demonstrating a survival advantage associated with TILs tend to involve a high-risk population of patients. For example, in the study by Clemente et al,2 the proportion of patients with a lesion more than 2 mm in Breslow depth was 82%, and in the study by Tuthill et al,5 71% of patients had lesions thicker than 1.7 mm. By comparison, 74.4% of patients in the negative-results study by Barnhill et al8 has lesions thinner than 1.7 mm. Similarly, in the study by Clark et al,1 the impact of TILs on survival appeared to be most pronounced in lesions more than 1.7 mm in depth. We found no impact of TILs on survival, and in our study, only 44% of patients had lesions thicker than 2.0 mm. The recent availability of regional nodal staging with SLN biopsy18 allowed us to identify an association of TILs with SLN status that could explain the conflicting data on the relationship between TILs in the primary lesion and survival. It has been previously reported that the presence of TIL is more frequently seen in thin lesions.2,4,7,9,19 This observation was also true in the present study. Thick lesions (> 4 mm) were significantly less likely to have a brisk TIL infiltrate. The presence of brisk TILs may represent an immune response resulting in tumor regression or inhibition of growth.19,20 Alternatively, thicker lesions may have already metastasized to regional nodes or beyond, resulting in systemic immunosuppression inhibiting the presence of TILs at the primary site.7 Although brisk TILs are more commonly found in thinner lesions, we demonstrate that TILs are an independent predictor of SLN positivity, so the impact of TILs is not solely related to its association with Breslow depth. The proportion of patients who developed a recurrence was similar among the brisk, nonbrisk, and absent TILs groups but the pattern of recurrence was significantly different. There was an increased incidence of regional recurrences in patients with absent TILs, which probably reflects the increased incidence of SLN-positive disease in these patients. In the group of SLN-positive patients, the incidence of regional recurrence was much higher (28.8%) compared with the SLN-negative patients (6.7%). It is intriguing that, although patients with TILs present in their primary lesion were less likely to experience recurrence in a regional nodal basin, they still developed a similar number of distant recurrences. This may represent an alternative explanation for the apparent lack of a survival benefit associated with brisk TILs. Breslow thickness, Clark level, and ulceration are pathologic features of the primary most commonly used to indicate the need for SLN mapping as well as to estimate the risk of a SLN metastasis. In this study, we have identified TILs as an important variable to consider when estimating this risk. For example, in patients with a lesion less than 1 mm in thickness, the overall risk of a positive SLN is 5.5%, but it is 18.5% among patients with absent TILs and 0% among patients with brisk TILs in their primary melanoma. The addition of TIL status will improve the ability to accurately predict the rate of SLN positivity in each thickness category (Fig 2). In addition, the relationship between TILs and SLN status offers an opportunity for further studies into the nature of these infiltrating immune modulators. Although a number of studies have characterized the T-cell infiltrate in primary cutaneous melanoma, these have correlated immunophenotype with lesion thickness or signs of regression and not with survival or regional metastases. It has been demonstrated that T-cell infiltrates in regressing melanomas are predominantly CD4+ cells producing cytokines, such as interleukin-2 and interferon gamma, consistent with a Th1 subtype.21-25 By contrast, thicker melanomas are infiltrated with CD4+ cells of the Th2 subtype, producing immunosuppressive cytokines such as transforming growth factor beta and interleukin-10.21-25 In addition, studies of TILs in a variety of malignancies, including metastatic melanoma, suggest that, although effector cytotoxic CD8+ T cells may portend a better prognosis,26 infiltration with suppressor T cells (CD4+CD25+) may indicate the opposite.27-29 A study correlating the immunophenotype of the TILs and SLN status and survival in primary melanoma is currently being undertaken at our institution and may provide important clues toward understanding the relationship between TILs and prognosis. Our data demonstrate that the presence of TILs in primary cutaneous melanoma is independently associated with a lower risk for regional nodal metastasis. We found no evidence that TIL status determined outcome, however, probably because this end point is so powerfully influenced by SLN status. The progression and dissemination of tumor despite the presence of tumor-specific immune cells remains a major paradox in tumor immunology. A better understanding of the prognostic significance of TILs in patients with cutaneous melanoma may provide critical information for designing effective immunotherapy in the future.
The authors indicated no potential conflicts of interest.
Conception and design: Rebecca C. Taylor, Klaus J. Busam, Mary S. Brady Collection and assembly of data: Ami Patel, Mary S. Brady Data analysis and interpretation: Rebecca C.Taylor, Ami Patel, Katherine S. Panageas, Klaus J. Busam, Mary S. Brady Manuscript writing: Rebecca C.Taylor, Mary S. Brady Final approval of manuscript: Rebecca C.Taylor, Katherine S. Panageas, Klaus J. Busam, Mary S. Brady
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Clark WH Jr, Elder DE, Guerry DT, et al: Model predicting survival in stage I melanoma based on tumor progression. J Natl Cancer Inst 81:1893-1904, 1989 2. Clemente CG, Mihm MC Jr, Bufalino R, et al: Prognostic value of tumor infiltrating lymphocytes in the vertical growth phase of primary cutaneous melanoma. Cancer 77:1303-1310, 1996[CrossRef][Medline] 3. Sondergaard K, Schou G: Survival with primary cutaneous malignant melanoma, evaluated from 2012 cases: A multivariate regression analysis. Virchows Arch A Pathol Anat Histopathol 406:179-195, 1985[CrossRef][Medline] 4. Elder DE, Guerry DT, VanHorn M, et al: The role of lymph node dissection for clinical stage I malignant melanoma of intermediate thickness (1.51-3.99 mm). Cancer 56:413-418, 1985[CrossRef][Medline] 5. Tuthill RJ, Unger JM, Liu PY, et al: Risk assessment in localized primary cutaneous melanoma: A Southwest Oncology Group study evaluating nine factors and a test of the Clark logistic regression prediction model. Am J Clin Pathol 118:504-511, 2002 6. Johnson OK Jr, Emrich LJ, Karakousis CP, et al: Comparison of prognostic factors for survival and recurrence in malignant melanoma of the skin, clinical stage I. Cancer 55:1107-1117, 1985[CrossRef][Medline] 7. Larsen TE, Grude TH: A retrospective histological study of 669 cases of primary cutaneous malignant melanoma in clinical stage I: 3, The relation between the tumour-associated lymphocyte infiltration and age and sex, tumour cell type, pigmentation, cellular atypia, mitotic count, depth of invasion, ulceration, tumour type and prognosis. Acta Pathol Microbiol Scand 86A:523-530, 1978 8. Barnhill RL, Fine JA, Roush GC, et al: Predicting five-year outcome for patients with cutaneous melanoma in a population-based study. Cancer 78:427-432, 1996[CrossRef][Medline] 9. Thorn M, Ponten F, Bergstrom R, et al: Clinical and histopathologic predictors of survival in patients with malignant melanoma: A population-based study in Sweden. J Natl Cancer Inst 86:761-769, 1994 10. Balch CM, Soong SJ, Gershenwald JE, et al: Prognostic factors analysis of 17,600 melanoma patients: Validation of the American Joint Committee on Cancer melanoma staging system. J Clin Oncol 19:3622-3634, 2001 11. Mraz-Gernhard S, Sagebiel RW, Kashani-Sabet M, et al: Prediction of sentinel lymph node micrometastasis by histological features in primary cutaneous malignant melanoma. Arch Dermatol 134:983-987, 1998 12. Berk DR, Johnson DL, Uzieblo A, et al: Sentinel lymph node biopsy for cutaneous melanoma: The Stanford experience, 1997-2004. Arch Dermatol 141:1016-1022, 2005 13. Sondak VK, Taylor JM, Sabel MS, et al: Mitotic rate and younger age are predictors of sentinel lymph node positivity: Lessons learned from the generation of a probabilistic model. Ann Surg Oncol 11:247-258, 2004 14. Lowe JB, Hurst E, Moley JF, et al: Sentinel lymph node biopsy in patients with thin melanoma. Arch Dermatol 139:617-621, 2003 15. Kruper L, Botbyl B, Czerniecki B, et al: Sentinel lymph node status in stage II/III melanoma. J Clin Oncol 23:710S, 2005 (suppl; abstr 7501) 16. 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 233:250-258, 2001[CrossRef][Medline] 17. Busam KJ, Antonescu CR, Marghoob AA, et al: Histologic classification of tumor-infiltrating lymphocytes in primary cutaneous malignant melanoma: A study of interobserver agreement. Am J Clin Pathol 115:856-860, 2001 18. 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 17:976-983, 1999 19. Kopf AW, Welkovich B, Frankel RE, et al: Thickness of malignant melanoma: Global analysis of related factors. J Dermatol Surg Oncol 13:345-390, 401-420, 1987[Medline] 20. Elder DE, Gimotty PA, Guerry D: Cutaneous melanoma: Estimating survival and recurrence risk based on histopathologic features. Dermatol Ther 18:369-385, 2005[CrossRef][Medline] 21. Lowes MA, Bishop GA, Crotty K, et al: T helper 1 cytokine mRNA is increased in spontaneously regressing primary melanomas. J Invest Dermatol 108:914-919, 1997[CrossRef][Medline] 22. Wagner SN, Schultewolter T, Wagner C, et al: Immune response against human primary malignant melanoma: A distinct cytokine mRNA profile associated with spontaneous regression. Lab Invest 78:541-550, 1998[Medline] 23. Conrad CT, Ernst NR, Dummer W, et al: Differential expression of transforming growth factor beta 1 and interleukin 10 in progressing and regressing areas of primary melanoma. J Exp Clin Cancer Res 18:225-232, 1999[Medline] 24. Reed JA, McNutt NS, Prieto VG, et al: Expression of transforming growth factor-beta 2 in malignant melanoma correlates with the depth of tumor invasion: Implications for tumor progression. Am J Pathol 145:97-104, 1994[Abstract] 25. Halliday GM, Patel A, Hunt MJ, et al: Spontaneous regression of human melanoma/nonmelanoma skin cancer: Association with infiltrating CD4+ T cells. World J Surg 19:352-358, 1995[CrossRef][Medline] 26. Pages F, Berger A, Camus M, et al: Effector memory T cells, early metastasis, and survival in colorectal cancer. N Engl J Med 353:2654-2666, 2005 27. Viguier M, Lemaitre F, Verola O, et al: Foxp3 expressing CD4+CD25(high) regulatory T cells are overrepresented in human metastatic melanoma lymph nodes and inhibit the function of infiltrating T cells. J Immunol 173:1444-1453, 2004 28. Curiel TJ, Coukos G, Zou L, et al: Specific recruitment of regulatory T cells in ovarian carcinoma fosters immune privilege and predicts reduced survival. Nat Med 10:942-949, 2004[CrossRef][Medline] 29. Sasada T, Kimura M, Yoshida Y, et al: CD4+CD25+ regulatory T cells in patients with gastrointestinal malignancies: Possible involvement of regulatory T cells in disease progression. Cancer 98:1089-1099, 2003[CrossRef][Medline] Submitted September 2, 2006; accepted December 5, 2006.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2007 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 test for discrete variables. Multivariate logistic regression was performed to identify clinical and pathologic features predictive of SLN metastases. Disease-free survival (DFS) and overall survival (OS) were calculated from the date of SLN biopsy to the date of first recurrence, death, or last follow-up. The survival distributions were estimated using the Kaplan-Meier method. Univariate comparisons or survival distributions were made using the log-rank test and multivariate comparisons using the Cox proportional hazard model. P 
