|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 22, No 21 (November 1), 2004: pp. 4369-4375 © 2004 American Society of Clinical Oncology. DOI: 10.1200/JCO.2004.02.096 Effect of Pregnancy on Survival in Women With Cutaneous Malignant MelanomaFrom the Genetic Epidemiology Division, Cancer Research UK, St James's University Hospital, Leeds, UK; Division of Dermatology, Division of Pathology, University Hospital; Department of Biomedicine and Surgery, Faculty of Health Sciences, University of Linköping, Linköping; Department of Epidemiology, Institute of Environmental Medicine, Karolinska Institute and Department of Epidemiology, Stockholm Center for Public Health, Stockholm, Sweden Address reprint requests to Marko B. Lens, MD, PhD, Genetic Epidemiology Division, Cancer Research UK, St James's University Hospital, Beckett St, Leeds LS9 7TF, UK; e-mail: markolens{at}aol.com
PURPOSE: An adverse influence of pregnancy on the risk of death in women with cutaneous melanoma was suggested historically by anecdotal reports. Previous studies included small numbers of women observed for short periods. METHODS: Using data from the Swedish National and Regional Registries, we performed a retrospective cohort study of all Swedish women who were diagnosed with cutaneous melanoma during their reproductive period, from January 1, 1958, to December 31, 1999. The relationship between pregnancy status at the diagnosis of melanoma and overall survival was examined in multivariable proportional-hazards models.
RESULTS: The cohort comprised 185 women (3.3%) diagnosed with melanoma during pregnancy and 5,348 (96.7%) women of the same childbearing age diagnosed with melanoma while not pregnant. There was no statistically significant difference in overall survival between pregnant and nonpregnant groups (log-rank CONCLUSION: The survival of pregnant women with melanoma is not worse than the survival of nonpregnant women with melanoma. Pregnancy subsequent to the diagnosis of primary melanoma was not associated with an increased risk of death.
There has been a substantial increase in the incidence of cutaneous malignant melanoma in the last few decades.1 Melanoma in North America and Europe has a relatively flat age distribution curve, so that it is relatively common in young adults. The coexistence of melanoma and pregnancy or desire to conceive is therefore increasing, and melanoma may be the most frequently encountered malignancy during pregnancy. The estimated incidence of melanoma in pregnancy is from 2.8 to 5 cases per 100,000 pregnancies.2 For many years, there has been concern that the hormonal and immunological changes that occur during pregnancy may be important in the evolution or progression of melanoma. The concern about the effect of pregnancy began in 1951, when Pack and Scharnagel3 reviewed 32 cases of melanoma diagnosed during pregnancy, reporting extremely poor prognosis due to the rapid development of metastases. Data from subsequent epidemiological studies evaluating the influence of pregnancy on the prognosis in women with melanoma are conflicting.4 Several studies reported that women diagnosed with melanoma during pregnancy have a high risk of melanoma progression and increased mortality risk.5-7 Other studies failed to document an adverse effect of pregnancy on melanoma prognosis.8,9 However, the majority of these studies were case-series or uncontrolled studies without sufficient statistical power to detect an association. Five small controlled studies have examined the effect of pregnancy on outcome in women with melanoma,10-14 and these failed to demonstrate a difference in survival rates when comparing the women diagnosed with melanoma (American Joint Committee on Cancer stage I and II) during pregnancy with women of the same age diagnosed with melanoma while nonpregnant. Two studies9,13 reported a shorter disease-free interval in women diagnosed with melanoma during pregnancy when compared with the nonpregnant women of the same childbearing age. Currently, there is insufficient evidence regarding the effect of pregnancy on the clinical course of melanoma, and there remains anxiety in clinical practice about the advice that should be given to patients. Guidelines for the management of melanoma tend to advise that decisions about pregnancy should be based on social criteria (British Association of Dermatologists guidelines).15 The primary objective of this study was to examine whether overall survival is worse in women diagnosed with melanoma during pregnancy compared with women of the same childbearing age diagnosed with melanoma while nonpregnant. In the secondary analysis, we investigated the effect of pregnancy subsequent to diagnosis of the primary melanoma on overall survival.
Study Design We designed a population-based, retrospective cohort study of all Swedish women who were diagnosed with cutaneous melanoma during their reproductive age from January 1, 1958, to December 31, 1999. The study was approved by the Ethical Committee of Linköping, and by Ethical Committees of all six Swedish Regional Cancer Registries. Eligibility criteria for inclusion in the study required that subjects be: (1) female, (2) reside in Sweden when diagnosed with the primary melanoma, (3) diagnosed of first primary cutaneous melanoma during the reproductive age period (defined as age between 15 to 45 years), (4) diagnosis of first primary melanoma between January 1958 and December 1999. The cohort was established on the basis of data obtained from three Swedish National Registries and six Regional Registries. The Swedish National Cancer and Mortality Registry. The Registry contains information on all newly diagnosed cancers. The Registry contained 19,337 records of primary cutaneous melanoma in women who were diagnosed with primary melanoma during the period from January 1, 1958, to December 31, 1999. The data on death were available until December 31, 2001. Swedish National Medical Birth Registry. This Registry contains information on the number and dates of births for Swedish women from 1973 to present. In the Medical Birth Registry, only data on biological children are included. Among women diagnosed with cutaneous melanoma at any time of their life, there were 6,566 records of births. Multigeneration Registry, Statistics Sweden. The multigeneration registry contains information on dates of live and stillborn births for Swedish women. A total of 16,383 records were supplied with this registry. The first birth was recorded in January 1932, while the last one was recorded in February 1978. Information on the histology characteristics of primary cutaneous melanoma was obtained from six Regional Cancer Registries covering all geographical regions in Sweden: Stockholm-Gotland, Malmö-Lund, Uppsala-Örebro, Göteborg, Linköping, and Umeå. The unique identification number assigned to all residents in Sweden in 1947 or at birth thereafter permitted accurate linkage of information from different registries. Data sets were fused in Sweden, and all analyses were subsequently carried out on anonymous data sets.
Statistical Analysis The primary analysis was a comparison between survival in pregnant and nonpregnant groups. The primary exposure of interest for this cohort study was the pregnancy status at the time of the diagnosis of primary malignant melanoma. Only pregnancies terminated by a child's delivery were considered an exposure. Data on miscarriage or therapeutic abortion were not available. Pregnancies and deliveries before and/or subsequent to the diagnosis of melanoma were not considered an exposure in the primary analysis. Based on identified exposure status, individuals from the population-based databases were classified into two groups: (1) women diagnosed with their first primary melanoma during pregnancy (pregnant group), and (2) women diagnosed with primary melanoma while nonpregnant during their childbearing period (nonpregnant group). The Kaplan-Meier product-limit method was used to estimate curves for overall survival of pregnant and nonpregnant women. The log-rank test was used to test differences between the survival curves. The P values represented overall comparisons based on the survival curves and not on any particular time point.
The effect of pregnancy status at the time of diagnosis of primary melanoma was assessed by multivariable analysis using the Cox proportional hazards regression model, adjusting for the potential confounders in the 2,101 women for whom the data on the following prognostic factors were available: age (continuous measure, in years), Breslow tumor thickness (coded as: 1, In the secondary analysis, pregnancy status subsequent to the diagnosis of the primary melanoma was considered an exposure. The cohort was divided into two groups: women who had a pregnancy or pregnancies subsequent to the diagnosis of the primary melanoma, and women who did not have a pregnancy after the diagnosis of the melanoma. Thus, the secondary analysis was a comparison between survival in women who had a pregnancy subsequent to the diagnosis of the primary melanoma, and women who did not have a pregnancy subsequent to the diagnosis of the primary melanoma. In this analysis, the probability of a woman having a pregnancy will be dependent on whether or not she has survived. Therefore, a multivariable analysis was performed using pregnancy status after the diagnosis of melanoma as a time-dependent variable. All P values were based on two-sided tests. The data were analyzed using the SAS statistical package version 8.2 (SAS Institute Inc, Cary, NC).
Primary Analysis Patient characteristics. A total of 5,533 women diagnosed with their first primary cutaneous melanoma during their childbearing period were included in the study. The cohort comprised 185 women (3.3%) diagnosed with melanoma during pregnancy (pregnant group) and 5,348 women (96.7%) of the same childbearing age diagnosed with melanoma while not pregnant (nonpregnant group). Among women diagnosed with melanoma during pregnancy, the diagnosis of melanoma occurred during the first trimester in 58 patients (31.4%), during the second trimester in 52 patients (28.1%), and during the third trimester in 75 patients (40.5%). The length of follow-up ranged from 25 days to 41.9 years (median, 11.6 years) for women in the pregnant group and from 1 day to 42.9 years (median, 11.4 years) for women in the nonpregnant group. Baseline clinical and histological characteristics of the women in the pregnant and nonpregnant groups are presented in Table 1.
Tumor thickness was recorded for 2,262 women (40.9%)—137 women (74.1%) in the pregnant group and 2,125 women (39.7%) in the nonpregnant group. Although the mean thickness in the pregnancy-associated melanomas was higher (1.28 mm v 1.07 mm), analysis by t test did not reveal a statistically significant difference in the mean Breslow thickness between pregnant and nonpregnant groups (P = .068). Data on the Clark's level were available for 2,200 women (39.8%)—134 women in the pregnant group (72.4%) and in records of 2,066 women (38.6%) in the nonpregnant group. The majority of women had lesions with Clark's level less than IV (76.1% in the pregnant group and 82.1% in the nonpregnant group). Regional registries had the histological type of primary melanoma recorded for 1,957 women (35.4%)—92 women (49.7%) in the pregnant group and 1,865 women (34.9%) in the nonpregnant group. The most common type of melanoma was superficial spreading melanoma (74.7% in the pregnant group and 77.2% in the nonpregnant group). The primary site of the melanoma on the skin was reported for all women included in the cohort. Survival analysis. At the time of the last follow-up, a total of 991 deaths occurred among the women included in the cohort study (28 in pregnant and 963 in nonpregnant). The incidence mortality rate in pregnant and nonpregnant women was 1,315 per 100,000 person-years and 1,102 of 100,000 person-years of follow-up, respectively. A total of 4,544 women were alive at the end of follow-up—159 women (85.0%) in the pregnant group and 4,385 women (82.0%) in the nonpregnant group. The median follow-up of the surviving women was 12.9 years. Kaplan-Meier survival curves are presented in the Figure 1.
The log-rank test did not detect a statistically significant difference in overall survival in women with melanoma diagnosed during pregnancy when compared with women with melanoma diagnosed while nonpregnant (  21[r] = 0.84; P = .361). The following potential confounders were selected to be included in the Cox regression models: Breslow thickness, tumor site, Clark's level, and age. In the univariable analysis, all these factors were significant predictors of survival in women diagnosed with melanoma during their reproductive age. The multivariable Cox regression model included 2,101 women among whom there were 145 events (death). The population of 2,101 includes all women who had data on all variables included in the Cox model: Breslow thickness, tumor site, Clark's level, and age. After the adjustment for all included potential confounders, pregnancy status at the diagnosis of melanoma was not related to death (HR of death in women diagnosed with melanoma during pregnancy was: HR = 1.08; 95% CI, 0.60 to 1.93). Breslow thickness and tumor site, however, were highly significant independent prognostic factors of overall survival in these female melanoma patients. Women with higher Breslow tumor thickness category had significantly higher risk of death than those with lower Breslow category (HR = 2.16; 95% CI, 1.80 to 2.58). Also, women with axial tumors (head and neck, and trunk melanomas) had a poorer prognosis than those localized on the extremities (HR = 2.51; 95% CI, 1.78 to 3.56). Clark's level and age were not independent predictors of survival. Results of the multivariable Cox regression model are presented in Table 2. Testing of the assumption of proportionality verified that data follow the proportional hazards assumption.
Secondary Analysis In the analysis of the effect of pregnancy subsequent to the diagnosis of the primary melanoma, women were stratified into two groups: women who had pregnancies after the diagnosis of melanoma (966 women) and women without pregnancies subsequent to the diagnosis of melanoma (4,567 women). Baseline characteristics of these two groups of women are presented in Table 3.
The mean Breslow thickness in the group of women with pregnancy subsequent to diagnosis of melanoma was 0.93 mm (95% CI, 0.82 to 1.03), while in the group of women with no pregnancy subsequent to melanoma, the mean Breslow thickness was 1.11 mm (95% CI, 1.04 to 1.17). The difference in the mean Breslow thickness between two groups was statistically significant (t test, P = .003). The effect of pregnancy subsequent to the diagnosis of melanoma was tested in the multivariable Cox regression model. In the absence of data on locoregional relapse, which in clinical practice probably deters patients from contemplating pregnancy after melanoma, pregnancy status after diagnosis of melanoma was considered a time-dependent variable in the multivariable Cox model. After the adjustment for the Breslow thickness, tumor site, Clark's level, and age, pregnancy subsequent to the diagnosis of melanoma was not related to the survival (HR for death in women who had pregnancy after diagnosis of melanoma was: HR = 0.58; 95% CI, 0.32 to 1.05). Tumor thickness and tumor site were the most powerful predictors of survival. Results of the multivariable Cox regression model are presented in Table 4.
This study showed that the prognosis of pregnant women with melanoma is not worse than the prognosis of nonpregnant women with melanoma. The study therefore provides further evidence against pregnancy as a risk factor for overall survival. The multivariable Cox analysis of the subset of 2,101 women for whom the data on tumor thickness, Clark's level, melanoma site, and age were available, showed that the melanoma tumor thickness is the major single determinant of overall survival. Previously published studies14-16 have shown that pregnant women tend to present with thicker tumors. This study showed no overall statistically significant difference in tumor thickness between women who are diagnosed while pregnant versus women who are diagnosed while nonpregnant. However, the thickness of tumors presenting during the last trimester was greater. The reason is unclear, but it may be because malignant changes in pigmented lesions during pregnancy have been ignored, and the diagnosis has been delayed. It is often stated that nevi may darken and enlarge during pregnancy, though there are no data to support this view5,17,18; therefore, both pregnant women and their doctors might be less concerned about a changing mole than they would be otherwise. However, the suggestion of thicker tumors during the third trimester is of note. With a diagnosis of melanoma in the mother, there are no data justifying termination of pregnancy as a therapeutic intervention to improve survival.19 The observation in other studies, however, that pregnant women presented with thicker tumors, and our observation that the Swedish women who were diagnosed during the last trimester had thicker tumors, may argue that pregnant women should be advised against complacency about their moles and should seek advice earlier if one changes in appearance. Although pregnant women with melanoma should be treated similarly to the nonpregnant women, pregnancy status limits the treatment options. Although surgery is a definitive therapy for early-stage disease (stage I and II melanoma), and this does not differ between pregnant and nonpregnant women, the treatment of pregnant women with stage III and IV melanoma is less clear and more difficult.20 The use of chemotherapy in pregnancy is still an issue for discussion among oncologists,21 while there is yet insufficient evidence to warrant the use of adjuvant chemotherapy or biologic therapy during pregnancy.22 Also, radiologic examinations and the possible use of sentinel lymph node biopsies are considerations that modify management options for pregnant women. Therefore, early detection of thin melanomas should be the goal when treating women with pigmented lesions during childbearing age, as at any age. Although some clinical studies have suggested that women with a history of melanoma who become pregnant have no significant increased risk for developing a new melanoma or reactivating an existing melanoma,11,12,23 the question as to whether pregnancy activates micrometastatic disease is still a matter for debate. The results from this study shows that parity status after a diagnosis of primary melanoma is not an important prognostic factor in the multivariable Cox regression analysis: there was no statistically significant difference in the overall survival in women with no pregnancies subsequent to the diagnosis of primary melanoma, when compared with women who had pregnancies after the diagnosis and treatment of the primary cutaneous melanoma. Although pregnancy was not statistically significant, our analysis showed a modest protective effect of pregnancy (P = .07) on overall survival. It is possible indeed that pregnancy may have a protective effect after the diagnosis of melanoma, which would not be inconsistent with the well-recognized better survival in women than in men24,25; However, it is impossible, without data on locoregional recurrence, to exclude the probability that there is some bias associated with a reluctance to conceive after locoregional recurrence. There are insufficient data to fully understand the role of hormones in outcome for patients with melanoma. The sex-specific differences in prognosis of cutaneous malignant melanoma are interpreted by the hypothesis that estrogens may be responsible for the inhibition of melanoma growth.26,27 However, the classical estrogen-receptor pathway cannot explain the effect of estrogen on the biological behavior of melanoma, since the reports in the literature regarding the expression of estrogen receptors are conflicting.28-31 Several studies have suggested that the endogenous estrogen metabolite 2-methoxyoestradiol (2-ME) inhibits angiogenesis and suppresses tumor growth in many malignant tumors.32-34 The data suggest that 2-ME has antiproliferative and apoptotic activities mediated independently of estrogen receptors.35,36 Although various molecular targets have been proposed for 2-ME, the actual mechanism of its action is still undefined, and further research is necessary.37,38 There is some theoretical basis, therefore, for the suggestion that estrogen may be beneficial to melanoma patients rather than harmful. The finding of this study will help us in the future in advising women with melanoma regarding their family planning. Until now, the recommendation given to young female melanoma patients was to avoid pregnancies after the diagnosis of melanoma, particularly during the first 3 years after the diagnosis of the primary melanoma.39 This study provides no evidence to support deferral of pregnancy except that at the end of 3 years, patients have a better estimate of their risk of relapse. For patients with a good-prognosis thin tumor, in which the risk of relapse is small but changes comparatively little over time, there seems to be little justification at all for suggesting deferral. A limitation of our study is the lack of full histology data for all women included in the cohort. Thus, we were unable to include all histological prognostic variables in the multivariate model, which can control for several confounding variables simultaneously. Sensitivity analysis of overall survival comparing women with missing Breslow data with women with available Breslow data did not reveal that these data concerns may be considered a bias (P = .748). Furthermore, there were no data on locoregional recurrence. This issue is important since the presence of the locoregional and systemic metastases may confound the results from this study by reducing conception rates. We also had no information on therapeutic abortions associated with melanoma, which theoretically also may have confounded the data. However, in Sweden, therapeutic abortion was never thought justifiable in women diagnosed with primary melanoma; this therefore seems unlikely to be a significant confounder. Although potentially harmful therapies are probably less likely to be implemented during pregnancy, we do not think that this is an important bias due to the fact that as yet, there is no postsurgical therapy with significant improvement in survival in melanoma patients independently of the pregnancy status. This study shows no adverse effect of pregnancy on survival at the time of diagnosis with melanoma. There was, furthermore, no evidence that pregnancy subsequent to diagnosis has any effect on outcome. This issue should be addressed in a prospective cohort study, though the necessary size of such a study would be great.
The authors indicated no potential conflicts of interest.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
1. Greenlee RT, Hill-Harmon MB, Murray T, et al: Cancer statistics, 2001. CA Cancer J Clin 51:15-36, 2001 2. Dillman RO, Vandermolen LA, Barth NM, et al: Malignant melanoma and pregnancy. West J Med 164:156-161, 1996[Medline] 3. Pack GT, Scharnagel IM: Prognosis of malignant melanoma in pregnant women. Cancer 4:324-334, 1951[CrossRef][Medline] 4. Kjems E, Krag C: Melanoma and pregnancy. Acta Oncol 32:371-378, 1993[Medline] 5. Byrd BF, McGanty WJ: Effect of pregnancy on clinical course of melanoma. Southeastern Med J 47:196-200, 1954 6. Sutherland CM, Loutfi A, Mather FJ, Carter RD, Krementz ET: Effect of pregnancy upon malignant melanoma. Surg Gynecol Obstet 157:443-446, 1983[Medline] 7. Trapeznikov NN, Khasanov SR, Iavorskii VV: Melanoma of the skin and pregnancy. Vopr Onkol 33:40-46, 1987 8. Holly EA: Melanoma and pregnancy. Recent Results Cancer Res 102:118-126, 1986[Medline] 9. Houghton AM, Flannery J, Viola MV: Malignant melanoma of the skin occurring during pregnancy. Cancer 48:407-410, 1981[CrossRef][Medline] 10. Reintgen DS, McCarty KS Jr, Vollmer R, et al: Malignant melanoma and pregnancy. Cancer 55:1340-1344, 1985[CrossRef][Medline] 11. McManamny DS, Moss ALH, Pocock PV, et al: Melanoma and pregnancy: A long-term follow-up. Br J Obstet Gynaecol 96:1419-1423, 1989[Medline]
12. Wong JH, Sterns EE, Kopald KH, et al: Prognostic significance of pregnancy in stage I melanoma. Arch Surg 124:1227-1231, 1989 13. Slingluff CL Jr, Reintgen DS, Volmer RT, Seigler HF: Malignant melanoma arising during pregnancy: A study of 100 patients. Ann Surg 211:552-559, 1990[Medline] 14. MacKie RM, Bufalino R, Morabito A, et al: Lack of effect of pregnancy on outcome of melanoma: The World Health Organisation Melanoma Programme. Lancet 337:653-655, 1991[CrossRef][Medline] 15. Roberts DL, Anstey AV, Barlow RJ, et al: UK guidelines for the management of cutaneous melanoma. Br J Dermatol 146:7-17, 2002 16. Travers RL, Sober AJ, Berwick M, et al: Increased thickness of pregnancy-associated melanoma. Br J Dermatol 132:876-883, 1995[Medline] 17. Pennoyer JW, Grin CM, Driscoll MS, Dry SM, Walsh SJ, Gelineau JP, et al: Changes in size of melanocytic nevi during pregnancy. J Am Acad Dermatol 36:378-382, 1997 (3 Pt 1)[CrossRef][Medline] 18. Grin CM, Rojas AI, Grant-Kels JM: Does pregnancy alter melanocytic nevi? J Cutan Pathol 28:389-392, 2001[CrossRef][Medline]
19. Resnick R: Cancer during pregnancy. N Engl J Med 341:120-121, 1999 20. Johnston SRD: What should we tell patients about hormones after having melanoma? In: Newon Bishop JA, Gore M (eds): Melanoma: Critical Debates. Oxford, UK, Blackwell Science, 2002, pp 275-276 21. Dipaola RS, Goodin S, Ratzell M, et al: Chemotherapy for metastatic melanoma during pregnancy. Gynecol Oncol 66:526-530, 1997[CrossRef][Medline] 22. Squatrito RC, Harlow SP: Melanoma complicating pregnancy. Obstet Gynecol Clin North Am 25:407-416, 1998[CrossRef][Medline] 23. Osterlind A, Tucker MA, Stone BJ, et al: The Danish case-control study of cutaneous malignant melanoma, III; Hormonal and reproductive factors in women. Int J Cancer 42:821-824, 1988[Medline] 24. Ries LG, Pollack ES, Young JL Jr: Cancer patient survival: Surveillance, epidemiology and end results program, 1973-79. J Natl Cancer Inst 70:693-707, 1983 25. Thörn M, Adam HO, Ringborg U, et al: Long-term survival in malignant melanoma with special reference to age and sex as prognostic factors. J Natl Cancer Inst 79:969-974, 1987 26. Proctor JW, Auclair BG, Stokowski L: Endrocrine factors and the growth and spread of B16 melanoma. J Natl Cancer Inst 57:1197-1198, 1976
27. Feucht KA, Walker MJ, Das Gupta TK, et al: Effect of 17 beta-estradiol on the growth of estrogen receptor-positive human melanoma in vitro and in athymic mice. Cancer Res 48:7093-7101, 1988 28. Fisher RI, Neifeld JP, Lippman ME: Estrogen receptors in human malignant melanoma. Lancet 2:337-338, 1976[Medline] 29. Creagan ET, Ingle JN, Woods JE, et al: Estrogen receptors in patients with malignant melanoma. Cancer 46:1785-1786, 1980[CrossRef][Medline]
30. Flowers JL, Seigler HF, McCarty KS Sr, et al: Absence of estrogen receptor in human melanoma as evaluated by a monoclonal antiestrogen receptor antibody. Arch Dermatol 123:764-765, 1987 31. Lecavalier MA, From L, Gaid N: Absence of estrogen receptors in dysplastic nevi and malignant melanoma. J Am Acad Dermatol 23:242-246, 1990 (2 pt 1)[Medline] 32. Fotsis T, Zhang Y, Pepper MS, et al: The endogenous oestrogen metabolite 2-methoxyoestradiol inhibits angiogenesis and suppresses tumour growth. Nature 368:237-239, 1994[CrossRef][Medline] 33. Lin HL, Liu TY, Wu CW, et al: 2-methoxyestradiol-induced caspase-3 activation and apoptosis occurs through G(2)/M arrest dependent and independent pathways in gastric carcinoma cells. Cancer 92:500-509, 2001[CrossRef][Medline]
34. Chauhan D, Catley L, Hideshima T, et al: 2-methoxyestradiol overcomes drug resistance in multiple myeloma cells. Blood 100:2187-2194, 2002 35. Merriam GR, MacLusky NJ, Picard MK, et al: Comparative properties of the catechol estrogens, I: Methylation by catechol-O-methyltransferase and binding to cytosol estrogen receptors. Steroids 36:1-11, 1980[CrossRef][Medline]
36. LaVallee TM, Zhan XH, Herbstritt CJ, et al: 2-methoxyestradiol inhibits proliferation and induces apoptosis independently of estrogen receptors alpha and beta. Cancer Res 62:3691-3697, 2002 37. Pribluda VS, Gubish ER Jr, Lavallee TM, et al: 2-methoxyestradiol: An endogenous antiangiogenic and antiproliferative drug candidate. Cancer Metastasis Rev 19:173-179, 2000[CrossRef][Medline]
38. Yue TL, Wang X, Louden CS, et al: 2-Methoxyestradiol, an endogenous estrogen metabolite, induces apoptosis in endothelial cells and inhibits angiogenesis: Possible role for stress-activated protein kinase signaling pathway and Fas expression. Mol Pharmacol 51:951-962, 1997 39. Borden EC: Melanoma and pregnancy. Semin Oncol 27:654-656, 2000[Medline] Submitted February 12, 2004; accepted August 9, 2004.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2004 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
1.00 mm; 2, 1.01 to 2.00 mm; 3, 2.01 to 4.00 mm; 4, > 4.00 mm), site of melanoma (coded as: 0, limbs; 1, axial location), and the Clark's level of invasion (coded as: 0 if I-III and 1 if IV and V). These parameters were assessed individually in univariable hazards regression models with each independent variable in the model statement. The final multivariable Cox model was fitted, and the hazard ratio (HR) with 95% CI for all included variables was calculated. The proportional hazards assumption was tested analytically using time-varying covariates for each variable included in the final model. 
