|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology Inflammatory Breast Carcinoma and Noninflammatory Locally Advanced Breast Carcinoma: Distinct Clinicopathologic Entities?From the National Cancer Institute/Division of Cancer Prevention and National Cancer Institute/Center to Reduce Cancer Health Disparities, Bethesda, MD. Address reprint requests to William F. Anderson, MD, NCI/Division of Cancer Prevention, EPN, Room 2144, 6130 Executive Blvd, Bethesda, MD 20892-7317; email: wanderso{at}mail.nih.gov.
Purpose: Inflammatory breast carcinoma (IBC) and noninflammatory locally advanced breast carcinoma (LABC) are both associated with poor prognosis; however, whether they are distinct clinicopathologic entities remains controversial. Materials and Methods: To determine whether IBC and LABC were different, we compared tumor characteristics, prognosis, and age-specific incidence rate patterns in the Surveillance, Epidemiology, and End-Results program. An age of 50 years served as a surrogate marker for menopause. Results: Younger age at diagnosis, poorer tumor grade, and negative estrogen receptors (ERs) were more predictive of IBC (n = 2,237) than of LABC (n = 7,985). Breast carcinoma survival was worse for patients with IBC than for those with LABC (log-rank test, P < .0001). Age-specific incidence rates for IBC increased until 50 years and then flattened, whereas rates for LABC increased for all ages. When rates for LABC were stratified by estrogen receptor–positive (ERP) and –negative (ERN) expression, rates for ERP and ERN diverged; that is, rates for ERP increased with advancing age, whereas rates for ERN flattened after 50 years. When rates for IBC were stratified by ER expression, rates for both ERP and ERN flattened after 50 years of age. Conclusion: IBC and LABC seemed to be distinct biologic entities, as indicated by different prognostic factor profiles and age-specific incidence rate patterns. Rates that increased before 50 years and then stabilized, possibly indicated that premenopausal exposures had a greater effect on maintaining rates for IBC than for LABC.
THE AMERICAN Joint Committee on Cancer (AJCC) T4d designation for inflammatory breast carcinoma (IBC) is based on Haagensen’s1,2 original clinicopathologic description of clinical inflammation with tumor embolization of the dermal lymphatic vessels of the breast. Typically, clinical changes arise quickly in IBC; therefore, the term IBC should not be applied to a neglected noninflammatory locally advanced breast carcinoma (LABC). However, despite well-established clinicopathologic features for IBC, several factors blur the distinction between IBC and LABC. First, IBC and LABC are both locally advanced breast carcinomas with poor prognosis. Second, IBC is an imprecise histopathologic term because pathologic embolization of the dermal lymphatic vessels may arise in any histopathologic breast carcinoma type.3 Third, IBC may have inflammation without lymphatic infiltration or lymphatic infiltration without inflammation.4 Fourth, the presence or absence of dermal lymphatic invasion has little effect on response to breast carcinoma treatment.5–8 Finally, classification schemes not predictive of therapeutic response suggest that IBC and LABC are part of an advanced breast carcinoma continuum rather than distinct clinicopathologic entities.9 To increase our understanding of IBC and LABC, we compared tumor characteristics, prognosis, and age-specific incidence rates in the United States using the large-scale population-based Surveillance, Epidemiology, and End-Results (SEER) database of the National Cancer Institute. Most other IBC studies have comprised relatively small case-series or cohort analyses. To date, the SEER database provides information about one of the largest groups of women with locally advanced breast carcinoma.
We used breast carcinoma records accrued during the years 1992 to 1999 from SEER’s 11 standard population-based cancer registries (November 2001 submission).10 This analysis was restricted to African-American and white women with invasive breast carcinomas because there were too few women in other racial-ethnic categories to perform the planned analysis. We used a combination of SEER’s Extent of Disease (EOD) codes and modified AJCC stage designation to separate IBC from LABC. SEER EOD code 70 was defined as "inflammatory carcinoma, including diffuse (beyond that directly overlying the tumor) dermal lymphatic permeation or infiltration,"11 which is consistent with the AJCC’s T4d designation.1 All other records contained within SEER’s AJCC stage IIIA or IIIB identified LABC. We stratified patient and tumor characteristics by IBC and LABC. We chose an age of 50 years as a surrogate indicator for menopause. SEER’s histopathologic grading conformed to the International Classification of Diseases for Oncology (ed 2; ICDO-2).12 We combined grades 1 (well differentiated) with 2 (moderately differentiated), and grades 3 (poorly differentiated) with 4 (undifferentiated) for low (or good) and high (or poor) tumor grades, respectively. Because no centralized laboratory was used to determine estrogen receptor (ER) expression, each SEER registry reported ER expression as positive (ERP), negative (ERN), missing, or unknown. Incidence rate data with SE were obtained with SEER stat 4.2.13 All rates were expressed per 100,000 person-years to the nearest 0.0001 value and age-adjusted by the direct method to the year 2000 United States standard population.14 Survival was estimated from SEER data using standard methods of observing persons with cancer until they died or were known to be alive at a particular time point.14 We used the Kaplan-Meier product-limit method to calculate breast carcinoma–specific survival15 and evaluated survival differences using the log-rank test.16 We performed Cox proportional hazards modeling to generate unadjusted and adjusted hazard ratios as estimates of the relative risks of breast carcinoma death for IBC compared with LABC.17
From SEER’s 174,994 female breast cancer records (Table 1  ), we identified 2,237 (1.3%) women with IBC and 7,985 (4.6%) women with LABC. Women with IBC were significantly younger than women with LABC (58.4 and 60.6 years, respectively; P < .0001). African-American women accounted for 9.1% of all breast carcinoma patients combined, 13% of IBC patients, and 14.2% of LABC patients.
As expected because the standard of care for IBC usually begins with systemic treatment and not surgery, there were large amounts of unknown data for tumor size (79.2%) and axillary lymph node status (34.6%). When tumor size was available, the mean tumor size was greater than 5.0 cm for both IBC and LABC. IBC was more likely to be axillary lymph node-negative, but this result is tenuous given the large amount of unknown lymph node data for IBC. Poor tumor grade and ERN were more predictive of IBC than of LABC. We observed a nearly two-fold increase in risk for IBC compared with LABC for poor grade (odds ratio, 1.91; 95% confidence interval, 1.69 to 2.16) and ERN (odds ratio, 1.99; 95% confidence interval, 1.78 to 2.22). Multivariate adjustment for age, tumor size, and lymph nodal status did not alter these univariate trends.
From 1992 to 1999, median duration of follow-up was 35 months. Kaplan-Meier product-limit method demonstrated worse survival for women with IBC than for those with LABC (log-rank test, P < .0001). Risk of death from IBC was nearly twice that from LABC after multivariate adjustment for race, age, tumor size, axillary lymph nodal status, grade, and ER expression (adjusted hazard ratio, 1.98; P < .0001). When we stratified breast carcinoma survival by ER expression (Fig 1
The age-specific rate curves for all patient cases combined and LABC increased rapidly until 50 years and then increased more slowly (Fig 2A and 2B). Rates for IBC increased until 50 years and then flattened (Fig 2B). When stratified by ER expression, ERP rates for all patient cases combined and LABC increased rapidly until 50 years and then increased more slowly, whereas ERN rates increased until 50 years and then flattened (Fig 3A and 3C). Conversely, both ERP and ERN rates for IBC flattened after 50 years (Fig 3B).
Our analysis of SEER data suggested that IBC and LABC represented distinct clinicopathologic entities with different prognostic factor profiles and age-specific incidence rate patterns. Whether these differences reflected unique biologic properties or vagaries in our classification scheme deserves additional clarification. We used SEER’s EOD 70 code to distinguish IBC from all other noninflammatory LABC, which are included within SEER’s AJCC stages IIIA or IIIB. The frequency distribution for IBC (1.3%) in our analysis was similar to the frequency distribution for IBC (1% to 6%) in other studies,18–22 indicating that the classification scheme worked with reasonable success.
Others have identified IBC patient cases in SEER using the ICDO-2 pathologic designation code 8530.23,24 ICDO-2 8530 is a conservative IBC definition because it only includes breast carcinoma patient cases with pathologic embolization of the dermal lymphatics and ignores clinical inflammation. Sixty-seven percent (1,488 of 2,237) of our IBC patient cases with SEER EOD 70 had matching ICDO-2 8350 codes. The remaining 33% (749 of 2,237) had ICDO-2 designations other than 8530. Rate patterns for these three groups (EOD 70, EOD 70 with ICDO-2 8530, and EOD 70 without ICDO-2 8530) were identical (Fig 4
Although similar incidence rate patterns suggested similar clinicopathologic entities (Fig 4  ), different rate patterns suggested that IBC and LABC differed (Figs 2 and 3). Better understanding of distinct prognostic factor profiles and incidence rate patterns for IBC and LABC may further clarify the differences between these two categories of locally advanced breast carcinoma. Although older age is generally associated with tumor progression and advancing AJCC stage, older women tend to have more favorable prognostic factor profiles than do younger women.25–27 In previous IBC studies, mean ages at diagnoses ranged from 45 to 57 years;5,20,24 these values are approximately a decade younger than those for breast carcinoma in the general breast cancer population. Consistent with younger age at diagnosis, IBC, compared with LABC, was associated with poorer tumor grade, ERN, and worse breast carcinoma survival. Kokal et al9 suggested that IBC was an ERN subtype or poor prognostic variant of LABC. However, multivariate modeling for ER and other covariates indicated that IBC was an independent predictor for poor prognosis, with nearly two-fold greater risk of breast carcinoma death. Incidence rate patterns for most epithelial tumors increase steadily for all ages and are unaffected by age-specific exposures.28 In contrast, female breast carcinoma rates increase sharply until 50 years of age, and then increase more slowly.29 When female breast carcinoma rates are stratified by ER expression, rates for ERP and ERN diverge. That is, rates for ERP increase until age 50 years and then increase more slowly, whereas rates for ERN flatten after age 50 years.30–32 Divergent ERP and ERN rates have been described for all cases combined, but not for LABC. Curiously, for IBC, we observed an ERN-like pattern for both ERP and ERN. To the best of our knowledge, concordantly flat ERP and ERN rates after age 50 years is a unique observation. Similar to our findings for IBC and LABC, studies with familial breast carcinoma aggregates and BRCA1–2 hereditary breast carcinoma (HBC) have observed two distinct age-specific incidence rate patterns. These studies include one study of mono- and dizygotic-twins and their relatives by Peto and Mack,33 one study from Easton et al34 in BRCA1-HBC, and another study by Struewing et al35 in Ashkenazi hereditary BRCA1- and BRCA2-HBC. These investigators observed age-specific rates that stabilized around age 50 years for familial aggregates and BRCA1-HBC, as do ERN and IBC. Conversely, rates for sporadic (nonfamilial) breast carcinoma and BRCA2-HBC increased continuously with aging, as do ERP and LABC.35,36 In sum, similar incidence rate patterns indicated similar etiologic processes, whereas divergent rate patterns indicated distinct etiologies. Indeed, we may be viewing the consequence of distinct age-specific risk factor profiles, which might be dependent on or independent of premenopausal estrogen.36,37 Rates that increased until 50 years and then flatten indicated an important etiologic link for premenopausal ovarian function for ERN, IBC, familial aggregates, and BRCA1-HBC. Conversely, rates that increase continuously indicated an important role for cumulative lifetime exposures with ERP, LABC, sporadic breast carcinoma, and BRCA2-HBC. It may seem counterintuitive that menopause has greater influence on maintaining population-based incidence rates for ERN than for ERP.30–32 However, the difference between tumor promotion or progression and tumor etiology may explain this quandary. ERN and ERP are terms well established as predictive and prognostic factors for tumor promotion or progression,38 but their relevance as risk factors for tumor initiation is less certain. Generally, ERN predicts estrogen-insensitive breast carcinomas with poor prognosis, whereas ERP predicts estrogen-sensitive tumors with good prognosis. However, treatment-related estrogen dependence may not apply to breast cancer initiation. Theoretically, estrogen-dependent etiologies could initiate an ERN progenitor with the subsequent capacity for estrogen-independent promotion or progression, whereas estrogen-independent carcinogenesis could initiate an ERP progenitor that is sensitive to estrogen-dependent promotion or progression.32 This observational analysis has several caveats. First, although age-specific rate patterns may indicate provocative conclusions,39 our speculations are untested. Future hypothesis-driven studies should examine these two fundamentally distinct patterns of risk, dependent on or independent of premenopausal exposures. Second, SEER had no systematic review to standardize for variations in histopathologic classification for IBC or LABC assigned by individual pathologists at centers that contributed to the SEER program. However, diagnostic misclassification would tend to blur histopathologic discrepancies rather than result in profound disturbances. Our results actually may underestimate the true difference between IBC and LABC. Similarly, SEER ER analyses were not carried out in one centralized laboratory, and thus were obtained from community laboratories using different assay methods with different threshold levels for ERP and ERN. However, the distinctive population-based rate patterns for ERP and ERN were observed in all 11 SEER registries31,32 and in the Danish Breast Cancer Cooperative Group project,30 which indicates the robustness of these observations. ER status was also unknown for a proportion of the patient cases (n = 40,698 of 174,994; 23.3%). However, the percentage of breast carcinoma patients with known ER expression has increased greatly in SEER since 1990. In 1992, two standard SEER registries were added to the original nine SEER sites. Therefore, we limited this study to breast cancer patients diagnosed from the 11 standard SEER registries for the years 1992 to 1999. Finally, SEER did not record menstrual status, so we used 50 years of age as a proxy for menopause. In the absence of menstrual history, 50 years of age seems to be a reasonable proxy for menopausal status.40 Notwithstanding potential weaknesses of the SEER program, the strength of our study is its large-scale population-based design. We were able to calculate prognostic factor profiles and age-specific rate patterns for relatively rare clinicopathologic entities such as IBC and LABC, which provided insights not obtainable from smaller case-review series. In conclusion, our results provided observational evidence that IBC and LABC are distinct clinicopathologic entities with different prognostic factor profiles and age-specific incidence rate patterns. Moreover, our findings suggest a role for premenopausal risks or exposures more for the development of IBC than for LABC, providing new etiologic clues that might also apply to familial breast carcinoma aggregates versus sporadic breast carcinoma and BRCA1- versus BRCA2-HBC. Although controversy will likely persist until there is an improved molecular definition for LABCs, these large-scale population-based observations indicate that IBC and LABC have distinct clinicopathologic characteristics with possibly different etiologic mechanisms. At this time, the impact of such information on the development of new prevention or treatment strategies is unknown for these relatively rare clinical conditions. However, understanding both the differences and the similarities of these two types of advanced breast carcinoma may elucidate fundamental carcinogenic mechanisms that are shared with more common types of breast carcinoma.
For his review, comments, and suggestions, the authors are indebted to Mark Sherman, MD, of the National Cancer Institute, Division of Cancer Epidemiology and Genetics, Bethesda, MD.
1. Breast, in Greene FL, Page DL, Fleming ID, et al (eds): AJCC Cancer Staging Handbook (ed 6). New York, NY, Springer, 2002, pp 255–281 2. Haagensen CD: Inflammatory carcinoma, in Diseases of the Breast (ed 2). Philadelphia, PA, W. B. Saunders, 1971, pp 576–584 3. Rosen PP, Oberman HA: Unusual clinical presentation of carcinoma, tumors of the mammary gland, in Atlas of Tumor Pathology. Washington, DC, Armed Forces Institute of Pathology, 1993, pp 259–274 4. Saltzstein SL: Clinically occult inflammatory carcinoma of the breast. Cancer 34:382–388, 1974[CrossRef][Medline] 5. Lucas FV, Perez-Mesa C: Inflammatory carcinoma of the breast. Cancer 41:1595–1605, 1978[CrossRef][Medline] 6. Levine PH, Steinhorn SC, Ries LG, et al: Inflammatory breast cancer: The experience of the surveillance, epidemiology, and end results (SEER) program. J Natl Cancer Inst 74:291–297, 1985[Medline] 7. Wilke D, Colwell B, Dewar R: Inflammatory breast carcinoma: Comparison of survival of those diagnosed clinically, pathologically, or with both features. Am Surg 64:428–431, 1998[Medline]
8. Donegan WL, Padrta B: Combined therapy for inflammatory breast cancer. Arch Surg 125:578–582, 1990 9. Kokal WA, Hill LR, Porudominsky D, et al: Inflammatory breast carcinoma: A distinct entity? J Surg Oncol 30:152–155, 1985[Medline] 10. Surveillance, Epidemiology, and End Results (SEER) Program (www.seer.cancer.gov) Public Use Data (1973–1999), National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch, released April 2002, based on the November 2001 submission. 11. SEER Extent of Disease: 1988 Codes and Coding Instructions, (3rd edition, 1988). http://www.seer.cancer.gov/tools/codingmanuals 12. Percy C, Holten VV, Muir C: International Classification of Diseases for Oncology (ed 2). Geneva, World Health Organization, 1990 13. Surveillance, Epidemiology, and End Results (SEER) Program (www.seer.cancer.gov) SEER*STAT Database: Incidence-SEER 11 Regs, Nov 2001 Sub (1973–1999), National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch, released April 2002, based on the November 2001 submission. 14. Edwards BK, Howe HL, Ries LAG, et al: Annual report to the nation on the status of cancer, 1973–1999, featuring implications of age and aging on U.S. cancer burden. Cancer 94:2766–2792, 2002 15. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457–481, 1958[CrossRef] 16. Peto R, Peto J: Asymptomatically efficient rank invariant test procedures. J R Stat Soc A 135:185–198, 1972[CrossRef] 17. Cox DR: Regression models and life-tables. J R Stat Soc B 34:187–220, 1972 18. Bonnier P, Piana L, Khouzami A, et al: Inflammatory carcinoma of the breast. Eur J Gynaecol Oncol 13:7–11, 1992[Medline] 19. Grace WR, Cooperman AM: Inflammatory breast cancer. Surg Clin North Am 65:151–160, 1985[Medline] 20. Lee BJ, Tannenbaum NE: Inflammatory carcinoma of the breast. Surg Gynecol Obstet 39:580–595, 1924 21. Taylor GW, Meltzer A: "Inflammatory carcinoma" of the breast. Am J Cancer 33:33–49, 1938 22. Barber KW, Dockerty MB: Inflammatory carcinoma of the breast. Surg Gynecol Obstet 112:406–410, 1961[Medline] 23. Berg JW, Hutter RV: Breast cancer. Cancer 75:257–269, 1995[CrossRef][Medline] 24. Chang S, Parker SL, Pham T, et al: Inflammatory breast carcinoma incidence and survival: The Surveillance, Epidemiology, and End Results program of the National Cancer Institute, 1975–1992. Cancer 82:2366–2372, 1998[CrossRef][Medline]
25. Diab SG, Elledge RM, Clark GM: Tumor characteristics and clinical outcome of elderly women with breast cancer. J Natl Cancer Inst 92:550–556, 2000 26. Albain KS, Allred DC, Clark GM: Breast cancer outcome and predictors of outcome: Are there age differentials? J Natl Cancer Inst Monogr 16:35–42, 1994 27. Clark GM: The biology of breast cancer in older women. J Gerontol 47:19–23, 1992 28. Armitage P, Doll R: The age distribution of cancer and a multi-stage theory of carcinogenesis. Br J Cancer 8:1–12, 1954[Medline]
29. Pike MC, Spicer DV, Dahmoush L, et al: Estrogens, progestogens, normal breast cell proliferation, and breast cancer risk. Epidemiol Rev 15:17–35, 1993 30. Yasui Y, Potter JD: The shape of age-incidence curves of female breast cancer by hormone- receptor status. Cancer Causes Control 10:431–437, 1999[CrossRef][Medline] 31. Tarone RE, Chu KC: The greater impact of menopause on ER- than ER+ breast cancer incidence: A possible explanation (United States). Cancer Causes Control 13:7–14, 2002[CrossRef][Medline] 32. Anderson WF, Chatterjee N, Ershler WB, et al: Estrogen receptor breast cancer phenotypes in the Surveillance, Epidemiology, and End Results database. Breast Cancer Res Treat 76:27–36, 2002[CrossRef][Medline] 33. Peto J, Mack TM: High constant incidence in twins and other relatives of women with breast cancer. Nat Genet 26:411–414, 2000[CrossRef][Medline] 34. Easton DF, Ford D, Bishop DT: Breast and ovarian cancer incidence in BRCA1-mutation carriers: Breast Cancer Linkage Consortium. Am J Hum Genet 56:265–271, 1995[Medline]
35. Struewing JP, Hartge P, Wacholder S, et al: The risk of cancer associated with specific mutations of BRCA1 and BRCA2 among Ashkenazi Jews. N Engl J Med 336:1401–1408, 1997 36. Narod SA: Modifiers of risk of hereditary breast and ovarian cancer. Nat Rev Cancer 2:113–123, 2002[CrossRef][Medline] 37. Anderson WF, Brawley OW, Chang S: Oophorectomy in carriers of BRCA mutations. N Engl J Med 347:1038, 2002 (letter) 38. Henderson IC, Patek AJ: The relationship between prognostic and predictive factors in the management of breast cancer. Breast Cancer Res Treat 52:261–288, 1998[CrossRef][Medline] 39. Peto J, Houlston RS: Genetics and the common cancers. Eur J Cancer 37: 88–96, 2001 (suppl 8)[CrossRef]
40. Morabia A, Flandre P: Misclassification bias related to definition of menopausal status in case-control studies of breast cancer. Int J Epidemiol 21:222–228, 1992 Submitted July 15, 2002; accepted March 12, 2003.
Related Correspondence
Related Reply
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
