|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2002 American Society for Clinical Oncology Tumor Vascularity in the Prognostic Assessment of Primary Cutaneous MelanomaByFrom the Melanoma Center, Cutaneous Oncology Program, Cancer Center, University of California at San Francisco, San Francisco, CA. Address reprint requests to Richard W. Sagebiel, MD, University of California at San Francisco Cancer Center, 1600 Divisadero St, Fourth Floor, San Francisco, CA 94115; email: moledoc{at}itsa.ucsf.edu
PURPOSE: The vascular supply of the primary tumor is recognized to play an important role in the progression of a number of solid tumors. However, the role of tumor vascularity in the prognostic assessment of melanoma remains unclear. The purpose of this study was to determine the prognostic impact of patterns of vascularity on the outcome associated with cutaneous melanoma. PATIENTS AND METHODS: Tumor vascularity was documented prospectively using routine histopathologic analysis of 417 primary cutaneous melanomas from the University of California at San Francisco Melanoma Center database. Four patterns of tumor vascularity were recorded: absent, sparse, moderate, and prominent. RESULTS: Increasing tumor vascularity significantly increased the risk of relapse and death associated with melanoma, corresponding to reduced relapse-free and overall survival. By multivariate analysis, tumor vascularity was the most important determinant of overall survival, surpassing tumor thickness. Increasing tumor vascularity was associated with increased incidence of ulceration in the primary tumor. CONCLUSION: Tumor vascularity is an important prognostic factor in melanoma, rivaling tumor thickness. Increasing tumor vascularity is highly correlated with ulceration, possibly helping to explain the biologic basis of this known prognostic factor.
THE PROGNOSTIC assessment of primary cutaneous melanoma allows the identification of homogeneous subgroups with defined risks of relapse and death in order to facilitate treatment decisions and to identify stratification factors for clinical trials. Although tumor thickness has consistently emerged as the dominant factor affecting melanoma prognosis, numerous additional histologic factors have been examined for their ability to improve on the information gained by knowledge of the tumor thickness alone (see review in Kashani-Sabet et al1 and Montone and Elder2). Recently, ulceration has been incorporated into the American Joint Committee on Cancer staging classification for cutaneous melanoma.3 This is based on analyses of large databases that confirmed the reduced survival associated with ulcerated melanomas within a given range of tumor thickness.4 Ulceration is commonly defined as disruption of epidermal continuity and thus can be easily recorded in melanoma pathology reports. However, the biologic basis of this prognostic factor is poorly understood. One possible factor driving the development of ulceration is vascularity of the primary tumor. Tumor vascularity may be altered by increased vascular supply to existing vessels or development of new blood vessels. This latter concept, termed angiogenesis, has emerged as a powerful mechanism for tumor progression in a number of solid tumors.5 However, the role of tumor vascularity in the prognostic assessment of primary cutaneous melanoma remains controversial. In this study, we examined the significance of tumor vascularity (recorded histopathologically) in the outcome associated with melanoma.
Data regarding tumor vascularity was collected using a database of 417 patients from the University of California at San Francisco Melanoma Center archives, with at least 2 years of follow-up or first relapse. The cohort included 225 men (54%) and 192 women (46%), with an age range of 17 to 88 years (median, 47 years). The distribution of anatomic location of the primary tumor was as follows: trunk, 39.1%; extremity, 35.5%; and head and neck, 25.4%. The mean tumor thickness of the cohort was 2.48 mm. Tumor vascularity was recorded histopathologically, without the aid of immunoperoxidase stains, and evaluated at the base of the vertical growth phase of the tumor, in the immediately adjacent dermis and within the tumor mass. Four degrees of tumor vascularity were observed and recorded prospectively by a single pathologist (R.W.S.): absent, sparse, moderate, or prominent. "Absent" referred to no apparent difference from the normal vascular plexus in or about the tumor (Fig 1A). "Sparse" was recorded when a few additional small vessels were noted, usually located at the base of the thickest part of the tumor (vertical growth phase) (Fig 1B). "Moderate" reflected a pattern of both small and dilated vessels seen at intermediate powers of magnification (Fig 1C). Such vessels may have lumens up to three to four times normal and/or increased numbers of smaller vessels throughout the vertical growth phase both at the base and within all levels of the tumor. "Prominent" referred to widely dilated large vessels in easily recognized increased numbers within and at the base of the vertical growth phase, usually apparent at low magnification (Fig 1D). Other histopathologic attributes recorded included tumor thickness, mitotic index, ulceration, and microsatellites. Clinical parameters recorded included age, sex, and location of the primary tumor. Metastasis was defined to include local dermal or subcutaneous recurrence, satellitosis, regional nodal, or distant metastasis.
Statistical analysis of the cross-tabulated frequency data was performed (1) via the Fisher’s exact test, in the case of all 2 x 2 tables, and (2) via the Pearson  2 test, if either the number of rows in the table or the number of columns or both exceeded two. Statistical analysis of all differences between means (eg, mean Breslow tumor thickness and mean elapsed time intervals until relapse or death) was performed (1) via the two-sample (independent) t test, in the case of all two-sample comparisons, and (2) via a one-way analysis of variance (fixed effects model), if more than two means were compared. Where indicated, survival curves were generated using the Kaplan-Meier method,6 and differences between Kaplan-Meier curves were assessed using both log-rank and generalized Wilcoxon tests. Multivariate comparisons of the relative contributions of various high-risk prognostic factors were assessed via Cox regression analysis, assuming proportional hazards. To remain conservative in our statistical inferences, all tests of statistical significance reported are two sided.
In this study, the significance of tumor vascularity of primary cutaneous melanoma was examined histopathologically in a database of 417 patients. The four degrees of vascularity (defined in Patients and Methods and shown in Fig 1) were present in the following frequency: absent, documented in 154 patients (36.9%); sparse, in 68 patients (16.3%); moderate, in 103 patients (24.7%); and prominent, in 92 patients (22.1%). Initially, the possible relationship between tumor thickness and vascularity was examined. The mean tumor thickness increased significantly with increasing tumor vascularity, from 1.55 mm in absent cases to 4.01 mm in prominent cases (Table 1; P < .00005).
Several analyses were performed to evaluate the impact of increasing tumor vascularity on the outcome associated with melanoma. Statistical analysis of Kaplan-Meier survival curves suggested that three analyses were meaningful: (1) a two-tiered comparison of absent vascularity versus any degree of vascularity (ie, combining sparse, moderate, and prominent), (2) a three-tiered comparison of absent vascularity versus sparse and moderate (combined) versus prominent, and (3) a comparison of absent versus prominent vascularity. This was feasible given that the overall survival of the sparse group was lower than that of the absent group (P < .00005 by log-rank test), that the overall survival of the prominent group was lower than the sparse/moderate group (P = .001 by log-rank test), and that there was no significant difference in overall survival between the sparse and moderate groups (P = .120 by log-rank test). Finally, there was a significant difference in survival between the absent and prominent groups (P < .02 by log-rank test). The relationship between tumor vascularity and melanoma prognosis was analyzed by examining several measures of melanoma outcome. As shown in Table 2, 33.1% of patients in the absent group experienced relapse of their melanoma, compared with 50.6% of patients in the sparse/moderate group and 68.1% of patients in the prominent group (P < .00005). A total of 12.3% of patients died in the absent group, compared with 25.2% of patients in the sparse/moderate group and 42.4% of patients in the prominent group (P < .00005). There were similar increases in the rates of regional nodal and distant metastasis with increasing vascularity (Table 2). When the three groups with increased vascularity were combined and compared with the absent group with respect to all four outcome parameters, the differences remained statistically significant at the P < .00005 level (data not shown).
Next, the impact of tumor vascularity on survival was evaluated. The median overall survival of patients with absent vascularity was 65 months, compared with 35 months for patients with any degree of vascularity (P < .02). The median relapse-free survival of patients with absent vascularity was 20 months, compared with 10 months for patients with any degree of vascularity (P = .0034). Kaplan-Meier survival plots showed significant differences in overall survival (P < .00005) and relapse-free survival (P < .00005) between patients with absent vascularity and those with any degree of vascularity (Fig 2A and 2B, respectively).
The independent significance of tumor vascularity on outcome was assessed using Cox regression of several well-recognized histologic factors, including tumor thickness, ulceration, mitotic index, and microsatellitosis. When the impact of these factors was examined on the overall survival of melanoma patients in this data set, surprisingly, tumor vascularity emerged as the most powerful prognostic factor, surpassing tumor thickness (Table 3). In this multivariate analysis, tumor thickness and mitotic index retained significance, whereas the other factors provided no additional prognostic value.
Finally, we hypothesized that tumor vascularity may be associated with the presence of ulceration. Strikingly, the prevalence of ulceration increased with increasing vascularity, from 9.7% in cases with absent vascularity up to 48.9% of cases with prominent vascularity (P < .00005; Table 4).
Our results indicate the prognostic significance of vascularity of the primary tumor in the outcome associated with cutaneous melanoma. An increasing degree of vascularity was associated with increased rates of relapse, including regional nodal and distant metastasis. Moreover, increasing tumor vascularity was correlated with reduced relapse-free and overall survival associated with melanoma. By Cox regression analysis, the degree of vascularity was the single most powerful predictor of overall survival. Finally, increasing tumor vascularity was positively correlated with ulceration, providing one possible explanation for the biologic basis of this known prognostic factor. Although the significance of angiogenesis as a prognostic factor for other solid tumors (namely, breast cancer5) is well established, its importance in melanoma has been controversial because of the presence of conflicting reports in the literature. Analysis of tumor vascularity has been confounded by the different definitions, techniques, and immunohistochemical stains used to record vascularity, as well as the divergent measures of clinical outcome used, usually in small case series. The early work of Srivastava et al7,8 indicated that melanomas are vascular and suggested a possible correlation with relapse. In one study, 20 melanomas of intermediate (0.76 to 4.0 mm) thickness were analyzed using Ulex europaeus I agglutinin staining, 10 with metastasis and 10 without. The percentage vascular area at the base of the tumor was more than two-fold higher in the recurrence group than in the relapse-free group.7 Subsequently, in a study of 71 melanomas, using Doppler ultrasonography and Ulex staining, these authors demonstrated that the neovascular bed develops as tumor thickness approaches 0.8 mm.8 Moreover, there were higher peak and mean systolic frequencies over melanomas with either regional or systemic relapse. Followfield and Cook9 analyzed the vascular volume of 64 melanomas using a similar staining method and found a correlation with tumor thickness. Interestingly, increased tumor vascularity was noted in in situ melanomas, suggesting that angiogenesis may precede tumor invasion and therefore may not be predictive of metastatic spread. More recently, Barnhill et al10 have performed studies examining the role of angiogenesis in melanoma progression. In one study, these authors examined the microvessel density of 89 nevi and melanomas using U europaeus I agglutinin staining. There was increased vascularity in melanomas when compared with nevi. However, no significant differences in microvessel density could be shown between radial and vertical growth phase melanomas, between thin and thick melanomas, and between primary and metastatic melanomas. These findings suggested that the angiogenic phenotype was induced early in the tumor progression of melanoma. In another study, Barnhill and Levy11 showed that the mean microvessel counts of thin regressing melanomas were higher than those of melanomas without regression, especially those of radial growth phase melanomas. Finally, Busam et al12 performed a matched-pair analysis of 60 cases of metastasizing and nonmetastasizing melanomas matched for tumor thickness, age, sex, and anatomic site. Three immunohistochemical stains were evaluated, and U europaeus I agglutinin emerged as the most suitable for analysis. These authors found no significant differences in the microvessel density between the two groups. Interestingly, the authors examined four separate patterns of vascular microarchitecture (distinct from those described in this study), and were unable to find a pattern that was significantly associated with increased relapse. Given the importance of patterns of vascularity in uveal melanoma,13 the authors speculated that heretofore unidentified patterns of vascularity may yield prognostic information. Finally, Straume et al14 examined the microvessel density of 88 nodular melanomas using factor VIII–related antigen. Increased microvessel density was associated with reduced overall survival in univariate but not multivariate analysis. Interestingly, the authors found a correlation between vessel density and ulceration in this subgroup of melanomas. Significant differences exist in the approach taken here when compared with these prior reports. In general, many of these studies have attempted to quantify tumor vascularity by analyzing the microvessel density as determined by an immunohistochemical stain. It is possible that such an approach, pioneered in studies of angiogenesis for breast cancer,5 is not an accurate surrogate for vascularity in melanoma. Moreover, matched-pair analyses may be undermined by the exclusion of other known or unknown prognostic factors (such as ulceration or vascular invasion) that may be unbalanced between the two groups, yielding a false-negative result. In contrast, our study used four histopathologic patterns of vascularity that seemed to describe separate morphologic subgroups, recorded prospectively in a large cohort. This morphologic classification was proposed because, to date, the optimal stain to quantitate tumor vascularity for melanoma has yet to be described. The prognostic impact of tumor vascularity was rigorously analyzed using several measures of the outcome associated with cutaneous melanoma, including regional and distant metastases and relapse-free and overall survival. The independent significance of the morphologic patterns of vascularity on overall survival was demonstrated using Cox regression analysis. Of particular interest was the relationship found between increasing vascularity and ulceration. Although the prognostic importance of ulceration is well known,4,15 resulting in its incorporation into the revised American Joint Committee on Cancer staging classification,3 factors in the primary tumor that drive the development of ulceration are poorly understood. Ulceration is known to be more prevalent in thicker tumors4 and has been correlated with the mitotic index.16,17 Several hypotheses have been advanced in attempts to account for ulceration, including, paradoxically, decreased blood supply, external trauma, and destruction of the epidermis by the expanding tumor mass.16,17 To our knowledge, our study is the first to demonstrate a link between graded morphologic patterns of increasing tumor vascularity and the development of ulceration in melanoma. Although it may be challenging to reproduce the four patterns of vascularity described here, our group is initiating a study examining the interobserver and intraobserver variability of these patterns. In this regard, recent studies using a novel stain for lymphatic endothelium to propose an important role for lymphangiogenesis in tumor progression18 may be of interest to further characterize the vascular patterns described here. Until the optimal immunohistochemical stain for determining vascularity in melanoma is determined, our studies suggest that morphologic criteria for tumor vascularity have the greatest prognostic impact in the evaluation of melanoma patients. In practical terms, however, it may be most important for pathologists to recognize the absent and prominent groups that seem to carry the greatest prognostic significance. Moreover, the strong relationship with ulceration suggests that ulceration may represent a surrogate marker for the vascularity of the primary melanoma. In conclusion, our results indicate that histopathologically defined patterns of vascularity are highly correlated with ulceration, metastasis, and death in malignant melanoma. Further studies to determine the molecular basis of angiogenesis in melanoma are warranted.
M.K.-S. was supported by the Leaders Society Clinical Career Development Award of the Dermatology Foundation, Evanston, IL. C.M.M.F. was supported by a fellowship sponsored by the Conselho Nacional de Desenvolvimento Cientifico e Tecnologico, Brasilia, Brazil. We thank Celia Hamilton for preparation of the manuscript.
Presented in part at the Thirty-Seventh Annual Meeting of the American Society of Clinical Oncology, San Francisco, CA, May 12-15, 2001.
1. Kashani-Sabet M, Leong SPL, Sagebiel RW: Prognostic factors in malignant melanoma. Surg Oncol Clin North Am 6: 599-623, 1997[Medline] 2. Montone KT, Elder DE: Factors predictive of metastasis and outcome, in Miller SJ, Maloney ME (eds): Cutaneous Oncology: Pathophysiology, Diagnosis and Management. Malden, MA, Blackwell Science, Inc, 1998, pp 278-291
3.
Balch CM, Buzaid A, Atkins M, et al: Final version of the American Joint Committee on Cancer staging system for cutaneous melanoma. J Clin Oncol 19: 3635-3648, 2001
4.
Buzaid AC, Ross MI, Balch CM, et al: Critical analysis of the current American Joint Committee on Cancer staging system for cutaneous melanoma and proposal of a new staging system. J Clin Oncol 15: 1039-1051, 1997 5. Weidner N, Semple JP, Welch WR, et al: Tumor angiogenesis and metastasis: Correlation in invasive breast carcinoma. N Engl J Med 324: 1-8, 1991[Abstract] 6. Kaplan EL, Meier P: Nonparametric estimations from incomplete observations. J Am Stat Assoc 53: 457-481, 1958[CrossRef] 7. Srivastava A, Laidler P, Davies RP, et al: The prognostic significance of tumor vascularity in intermediate-thickness (0.76-4.0 mm thick) skin melanoma: A quantitative histologic study. Am J Pathol 133: 419-423, 1988[Abstract] 8. Srivastava A, Hughes LE, Woodcock JP, et al: Vascularity in cutaneous melanoma detected by Doppler sonography and histology: Correlation with tumor behaviour. Br J Cancer 59: 89-91, 1989[Medline] 9. Followfied ME, Cook MG: The vascularity of primary cutaneous melanoma. J Pathol 164: 241-244, 1991[CrossRef][Medline] 10. Barnhill RL, Fandrey K, Levy MA, et al: Angiogenesis and tumor progression of melanoma: Quantification of vascularity in melanocytic nevi and cutaneous malignant melanoma. Lab Invest 67: 331-337, 1992[Medline] 11. Barnhill RL, Levy MA: Regressing thin cutaneous malignant melanomas (< or = 1.0 mm) are associated with angiogenesis. Am J Pathol 143: 99-104, 1993[Abstract] 12. Busam KJ, Berwick M, Blessing K, et al: Tumor vascularity is not a prognostic factor for skin melanoma. Am J Pathol 147: 1049-1056, 1995[Abstract] 13. Folberg R, Pe’er J, Gruman LM, et al: The morphologic characteristics of tumor blood vessels as a marker of tumor progression in primary human uveal melanoma: A matched case-control study. Hum Pathol 23: 1298-1305, 1992[CrossRef][Medline] 14. Straume O, Salveson HB, Akslen LA: Angiogenesis is prognostically important in vertical growth phase melanomas. Int J Oncol 15: 595-599, 1999[Medline] 15. Balch CM, Wilkerson JA, Murad TM, et al: The prognostic significance of ulceration of cutaneous melanoma. Cancer 45: 3012-3017, 1980[CrossRef][Medline] 16. Mascaro JM, Castro J, Castel T, et al: Why do melanomas ulcerate? J Cutan Pathol 11: 269-273, 1984[CrossRef][Medline] 17. Day CLJ, Mihm MCJ, Lew RA, et al: Prognostic factors for patients with clinical stage I melanoma of intermediate thickness (1.51 - 3.39 mm): A conceptual model for tumor growth and metastasis. Ann Surg 195: 35-43, 1982[Medline] 18. Skobe M, Hawighorst T, Jackson DG, et al: Induction of tumor lymphangiogenesis by VEGF-C promotes breast cancer metastasis. Nat Med 7: 151-152, 2001[CrossRef][Medline] Submitted June 15, 2001; accepted December 27, 2001.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2002 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
