Outcome of Ethnic Minorities With Acute or Chronic Leukemia Treated With Hematopoietic Stem-Cell Transplantation in the United States
http://www.100md.com
《临床肿瘤学》
the University of Minnesota, Blood and Marrow Transplant Program, Minneapolis, MN
University of Nebraska Medical Center, Omaha, NE
Center for International Blood and Marrow Transplant Research, Medical College of Wisconsin, Milwaukee, WI
Columbia University, New York
University of Rochester Medical Center, Rochester
Our Lady of Mercy Comprehensive Cancer Center, Bronx, NY
Cleveland Clinic Foundation, Cleveland, OH
St Joseph Hospital, Orange, CA
South Florida Bone Marrow Stem Cell Transplant Institute, Boynton Beach, FL
Hospital Mexico, San Jose, Costa Rica
Hospital Privado de Cordoba, Cordoba, Argentina
Bristol Children’s Hospital, Bristol, United Kingdom
British Columbia Children’s Hospital, Vancouver, British Columbia
Princess Margaret Hospital, Toronto, Ontario, Canada
University of Hamburg-Eppendorf, Hamburg, Germany
ABSTRACT
PURPOSE: We previously reported a higher risk of mortality among Hispanics after allogeneic hematopoietic stem-cell transplantation (HSCT). However, it is not known how specific post-transplantation events (acute or chronic graft-versus-host disease [GVHD], treatment-related mortality [TRM], and relapse) may explain mortality differences. The purpose of this study was to examine the relationship between ethnicity and post-transplantation events and determine their net effect on survival.
PATIENTS AND METHODS: We identified 3,028 patients with acute myeloid leukemia, acute lymphoblastic leukemia, or chronic myeloid leukemia reported to the International Bone Marrow Transplant Registry between 1990 and 2000 who received an HLA-identical sibling HSCT after a myeloablative conditioning regimen in the United States. There were 2,418 white patients (80%) and 610 ethnic minority patients (20%), of whom 251 were black (8%), 122 were Asian (4%), and 237 were Hispanic (8%). Cox proportional hazards regression was used to compare outcomes between whites and ethnic minorities while adjusting for other significant clinical factors.
RESULTS: No statistically significant differences in the risk of acute or chronic GVHD, TRM, or relapse were found between whites and any ethnic minority group. However, Hispanics had higher risks of treatment failure (death or relapse; relative risk [RR] = 1.30; 95% CI, 1.08 to 1.54; P = .004) and overall mortality (RR = 1.23; 95% CI, 1.03 to 1.47; P = .02).
CONCLUSION: The higher risks of treatment failure and mortality among Hispanics may be the net result of modest but not statistically significant increases in both relapse and TRM and cannot be accounted for by any single transplantation-related complication. Further studies should examine the role of social, economic, and cultural factors.
INTRODUCTION
Several studies have examined inherent biologic differences between racial and ethnic subgroups of cancer patients to determine whether there are differences in tumor presentation, disease histology, stage at diagnosis, or response to therapy. For example, African American males with prostate cancer have higher rates of metastatic disease and more advanced-stage tumors at diagnosis.1-3 Similar findings have been reported for cancer of the breast,4-6 endometrium,7,8 colon,9 and oral cavity.7 However, no study to date has clearly identified any specific biologic factor to explain these differences in presentation. These factors have also been shown to have a negative impact on survival after a cancer diagnosis, which has been demonstrated by the facts that African American, American Indian, and Alaskan or Hawaiian native patients have higher risks of cancer death than whites10 and that, in the United States, African Americans have the highest cancer incidence and death rates overall.11
There are other factors that may influence differences in outcomes among racial and ethnic groups. Lower socioeconomic status (SES) has been found to contribute to higher mortality among racial and ethnic minorities with cancer. Similarly, cultural differences in attitudes towards seeking medical care and the potential lack of access to primary care (and subsequently delayed diagnosis) among ethnic minorities may contribute to discrepant outcome. However, even controlling for stage of disease, most studies suggest that SES alone cannot explain differences in outcome among racial and ethnic groups.12-15
Although a relatively large amount of literature exists on health disparities in ethnic minorities being treated for cancer, there are only limited published data that explore whether differences in outcomes exist for ethnic minorities when their treatment includes hematopoietic stem-cell transplantation (HSCT). We previously reported an analysis of trends in survival rates among ethnic groups after HSCT and examined three time periods (1985 to 1989, 1990 to 1994, and 1995 to 1999) to determine whether survival differences existed. This study found that, compared with whites, Hispanics had lower 1- and 3-year adjusted survival rates between 1995 and 1999, but there were no significant differences in the survival rates comparing whites with blacks or Asians in any of the time periods.16 However, that study did not investigate whether specific post-transplantation events, such as acute or chronic graft-versus-host disease (GVHD), treatment-related mortality (TRM), and relapse, differed among ethnic groups and contributed to mortality differences. Therefore, the purpose of this study was to examine the relationship between ethnicity and post-transplantation events and determine their net effect on survival.
PATIENTS AND METHODS
Data Sources
This observational study included patients who received HLA-identical sibling HSCT from 1990 to 2000 and were reported to the Center for International Blood and Marrow Transplant Research (CIBMTR). The CIBMTR is a voluntary working group of more than 450 transplantation centers worldwide that contribute detailed data on consecutive allogeneic HSCT to a Statistical Center at the Health Policy Institute of the Medical College of Wisconsin in Milwaukee or the National Marrow Donor Program Coordinating Center in Minneapolis. Approximately two thirds of all active transplantation centers worldwide report data to the registry. The registry database includes information on 40% to 45% of all patients who have received an allogeneic transplantation since 1970, with annual updates. Computerized checks for errors, review of submitted data by physicians, and on-site audits of participating centers are used to monitor the quality of the data.
The CIBMTR collects data at the following two levels: registration and research. Registration data include disease type, age, sex, pretransplantation performance status, disease stage and chemotherapy responsiveness, date of diagnosis, donor and graft type (bone marrow–and/or blood-derived stem cells), high-dose conditioning regimen, post-transplantation engraftment, disease recurrence and survival, development of a new malignancy, and cause of death. Requests for data on disease recurrence or death for registered patients are at 6-month intervals. All CIBMTR centers contribute registration data on all patients. Research data is collected on subsets of registered patients selected using a weighted randomization scheme, including comprehensive pre- and post-transplantation clinical information.
Patients
Patients with acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), or chronic myeloid leukemia (CML) who received an HLA-identical sibling HSCT between 1990 and 2000 in transplantation centers within the United States and had full research forms were included in this study. By limiting the analysis to HLA-identical sibling transplantations, we limited the possible influence of HLA-associated genetic factors on the outcomes of interest. Patients receiving reduced-intensity conditioning regimens were excluded. Only patients with full research data were included because we needed extensive information on the many clinical variables (see Statistical Analysis) to be adjusted for in the multivariate models. To provide assurance that the research patients were representative of all registered patients, basic patient-, disease-, and transplantation-related characteristics and relapse and survival rates between research and registered patients were compared; no differences were noted.
The patient’s ethnicity was abstracted from data submitted by the transplantation center to the CIBMTR. Patients were categorized into four ethnic cohorts: white, black, Hispanic, and Asian. Patients reported as being of European or Western Russian or Caucasian unspecified descent, according to the CIBMTR database, were classified as white. Patients reported as being of African American, Caribbean black, South or Central American black, or black unspecified descent were classified as black. Patients reported as being of Caribbean Hispanic, Mexican/Southwestern United States Hispanic, South or Central American Hispanic, or Hispanic unspecified descent were classified as Hispanic. Patients reported as being of Asian Indian, Filipino, Hawaiian/Polynesian, Japanese, Korean, Northern Chinese, Southeast Asian/Southern Chinese, or Asian unspecified descent were classified as Asian. Patients not classified into one of these four groups, such as Native Americans, were excluded because study numbers precluded further analysis.
Study End Points
The study outcomes evaluated were grades 2 to 4 acute GVHD and chronic GVHD in patients surviving more than 90 days with evidence of engraftment. Other clinical outcomes evaluated included overall survival, leukemia-free survival (LFS; survival without leukemia after transplantation), disease recurrence, and TRM. TRM is defined as death in the first 28 days after transplantation or death in remission. For LFS, patients were considered as having treatment failure at the time of relapse or death from any cause; patients who were alive were censored at the last follow-up evaluation.
Statistical Analysis
Patient-, disease-, and transplantation-related variables were compared among the four racial/ethnic groups using the 2 statistic for categoric variables and the Kruskal-Wallis test for continuous variables. Univariate probabilities of LFS and survival were calculated using the Kaplan-Meier method; the log-rank test was used for univariate comparisons. Probabilities of TRM and leukemia relapse were calculated using cumulative incidence curves to accommodate competing risks.
Multivariate analyses of acute and chronic GVHD, TRM, leukemia relapse, LFS, and overall survival were performed using Cox proportional hazards regression, with the racial/ethnic group forced in all models (main effect) and using whites as the reference group. Patient-, disease-, and transplantation-related covariates examined include (Table 1) age at transplantation, sex, Karnofsky performance score at transplantation, disease type, disease stage at transplantation, presence of comorbid medical risk factors (hypertension, diabetes mellitus, and smoking), type of graft, interval from diagnosis to transplantation, donor-recipient sex matching, donor-recipient cytomegalovirus serologic status, type of GVHD prophylaxis administered, preparative regimen administered, and year of transplantation. Disease type, disease stage, duration of first complete remission, and interval from diagnosis to transplantation were grouped together to form interactive covariates based on previous knowledge of their additive contributory effect on outcomes. The assumption of proportional hazards was tested using a time-dependent covariate. Models were stratified on covariates found to be nonproportional, except those of the main effect. Covariates significantly associated with the outcome were identified by forward stepwise selection at a significance level of P .05. First-order interactions between racial/ethnic group and all significant covariates were tested. No interactions were noted. However, interaction was noted between disease type and duration of first complete remission or interval from transplantation to diagnosis; thus, these covariates were entered in the model building in combinations. The overall covariate effects were tested using the Wald test. Because of multiple comparisons in the main effect, P .03 was considered statistically significant (Bonferoni correction). All P values were calculated as two-sided tests.
RESULTS
We identified 3,028 patients with ALL, AML, or CML reported to the CIBMTR who received an HLA-identical sibling HSCT after a myeloablative conditioning regimen in the United States between 1990 and 2000. There were 2,418 whites (80%) and 610 ethnic minorities including 251 blacks (8%), 122 Asians (4%), and 237 Hispanics (8%). Characteristics of the cohort by racial/ethnic group are listed in Table 2. In general, more Hispanics underwent transplantation at a younger age (< 1 to 20 years) than whites, blacks, or Asians, and more whites and blacks underwent transplantation at an older age (> 40 years). AML was the most common indication for HSCT in whites, CML was the most common indication for HSCT in blacks, and ALL was the most common indication for HSCT in Asians and Hispanics. Comorbid medical risk factors, including hypertension and diabetes, were more common in blacks, although whites were more likely to be smokers. Ethnic minorities were more likely to receive HSCT more than 1 year after diagnosis compared with whites. Ethnic minorities were also more likely to have both donor and recipient cytomegalovirus-positive status. Hispanics were more likely to have had their HSCT in the more recent treatment time period, with more than two thirds of the HSCTs performed in 1996 to 2000.
Acute and Chronic GVHD
Tables 3 and 4 show the multivariate analyses for acute and chronic GVHD. No statistically significant differences in acute and chronic GVHD were noted among the four racial/ethnic groups, as noted by the overall 3 df test. However, Asians tended to have a marginally reduced relative risk (RR) of having grades 2 to 4 acute GVHD (RR = 0.68; 95% CI, 0.47 to 0.98; P = .04) when compared with the white cohort.
TRM and Relapse
The risk of TRM (Table 5) was not increased when comparing blacks, Asians, or Hispanics with whites. There was also no statistically significant difference in the risk of relapse (Table 6) between whites and any of the racial/ethnic minority groups.
Treatment Failure and Mortality
The multivariate analyses of treatment failure (defined as the risk of death or relapse) and overall mortality are shown in Tables 7 and 8. Compared with the white cohort, Hispanics were found to have a significantly higher risk of treatment failure (RR = 1.30; 95% CI, 1.08 to 1.54; P = .004), which resulted in an inferior LFS, and a higher risk of overall mortality (RR = 1.23; 95% CI, 1.03 to 1.47; P = .02), which resulted in lower overall survival. Figures 1 and 2 show the probability of LFS and overall survival, respectively, according to racial/ethnic group.
DISCUSSION
In this study of patients undergoing HLA-identical sibling donor HSCT for acute or chronic leukemia in the United States, we have shown a surprising similarity of outcomes comparing ethnic minorities with whites for most major transplantation-related outcomes including acute and chronic GVHD, relapse, and TRM. However, for the Hispanic cohort, the overall risk of treatment failure (death or relapse) and mortality was increased. Our data suggest that this may be the net result of modest, although not statistically significant, increases in both relapse and TRM and cannot be accounted for by any single transplantation-related complication.
The trend towards a higher risk of relapse in the Hispanic group may be partially attributed to the fact that a higher percentage of Hispanic patients underwent HSCT for a diagnosis of ALL than the other groups. Previous studies have indicated that, depending on age at HSCT and stage of disease, survival after HSCT for ALL is inferior to that reported for AML or CML.17 Disease stage and age at time of HSCT also affect prognosis; however, there was no difference in the disease stage among the racial/ethnic groups, and as a whole, the Hispanic group was younger than the other groups. Similarly, the prevalence of the comorbid medical risk factors evaluated in our study was not increased in the Hispanic cohort. However, the differences in treatment failure and mortality between whites and Hispanics persisted after adjusting for the significant clinical variables in the multivariate models (including diagnosis), indicating that some specific ethnically associated biologic variable or other sociocultural factors that were not measured (eg, SES, compliance, cultural attitudes, and so on) may account for this difference. A large study on the outcome of childhood ALL from the Children’s Cancer Study Group found that, for children with high-risk leukemia, Hispanics (but not blacks or Asians) had significantly worse outcomes than whites.18 A similar finding was reported from a study in more than 5,000 children with ALL treated on Pediatric Oncology Group treatment protocols in which a significant excess risk of mortality was found for both black and Hispanic children.19 In both studies, clinically important prognostic variables were adjusted for, and in the Children’s Cancer Study Group study, adjustments were able to be made for SES, although in the Pediatric Oncology Group analysis, this could not be accounted for. However, these studies suggest that, at least for children with ALL treated with standard chemotherapy protocols, ethnicity (particularly for Hispanics) is an independent predictor of outcome. This question has not been as adequately addressed in adult patients with leukemia, except for one study that compared prognostic factors and outcomes between white and black patients with AML and found that black men (but not black women) had a worse overall survival and that there were differences with respect to important prognostic factors.20
It is also interesting to note that few transplantations for Hispanics were reported in the early time period and that 65% of transplantations for Hispanics occurred in the most recent time period (1996 to 2000). This may, in part, be explained by the fact that the number of individuals of Hispanic or Latino origin living in the United States increased by 58% over the 1990 to 2000 time period.21 However, if this were the only reason, one would expect similar increases in the number of transplantations in Asians, whose population increase was similar to that of Hispanics, and also in blacks, whose population increased by 15% to 20% compared with a 3% to 5% increase in the US white population over this time period. An alternative explanation is the apparent increase in offering HSCT as a feasible treatment for hematologic malignancies (possibly an improvement in access) among racial or ethnic minorities. Unfortunately, we are not able to analyze specific issues related to access to medical care. Despite this, with the increasing use of HSCT in the minority population, the inferior outcome in the Hispanic cohort is of potential concern.
Access to highly specialized therapies, such as HSCT, may certainly impact on the number of procedures that are performed, but this is difficult to analyze in treatments such as HSCT. Determining the population who received HSCT from the total pool of eligible patients is difficult to document. Patient choice, physician choice, or other reasons are vital determinants to the issue of access. One report examined this issue by looking at hospital discharge data for 2 years (between 1988 and 1991) from four states in patients admitted for leukemia or lymphoma. The study found that blacks with leukemia or lymphoma were less likely than whites to undergo HSCT. However, they found that patients with Medicaid, self-pay insurance, or health maintenance organization enrollees were less likely to undergo HSCT than patients with private insurance,22 and this was independent of race or ethnicity. This study also found that Hispanics with leukemia were not less likely to receive HSCT than whites, although Hispanics with lymphoma were less likely to receive HSCT than whites.
Our study is limited by the fact that we are not able to account for socioeconomic and other racial/ethnic disparities in health care delivery that may have an impact on the outcome of cancer therapies in ethnic minorities. The CIBMTR collects limited data on SES, but the quality and completeness of these data preclude their inclusion in our analysis. There is a large amount of literature describing the impact of SES on health disparities, although most of the data focuses on comparisons of blacks with whites, and few studies have examined this issue in other ethnic minority groups. It is also possible that disparities in health care delivery and access may have an impact on the assessment and treatment of late complications after HSCT in ethnic minorities when they return home to their community health care system. Lower SES is associated with lower overall health care use, the use of fewer expensive technologic procedures,23 including HSCT,22 and lower satisfaction in the health care system. There is an inverse relationship between SES (based on income and educational attainment) and mortality regardless of race, although ethnic minorities are over-represented in the lower socioeconomic classes.24 However, if these issues were the explanation for the findings in our study, we would have expected to see similar findings in the black and Asian cohorts.
Before this study, there was little information available on how specific post-transplantation events, such as acute or chronic GVHD, relapse, and transplantation-related complications, differ among ethnic groups and potentially contribute to mortality differences. We have shown that, in this cohort of patients undergoing related donor HSCT for acute or chronic leukemia, these factors generally do not contribute to significant differences in mortality. The exception of a higher mortality rate in Hispanics can likely be attributed to slightly higher, but not statistically significant, increases in relapse and TRM. Further prospective studies are necessary to analyze the cultural and socioeconomic or biologic factors that may contribute to these findings.
Appendix
Supported by Public Health Service Grant No. U24-CA76518 from the National Cancer Institute, the National Institute of Allergy and Infectious Diseases, and the National Heart, Lung and Blood Institute; Agency for Healthcare Research and Quality; and grants from Aetna; AIG Medical Excess; Allianz Life/Life Trac; American Red Cross; American Society of Clinical Oncology; Amgen, Inc; anonymous donation to the Medical College of Wisconsin; AnorMED, Inc; Aventis Pharmaceuticals; Baxter Healthcare Corp; Baxter Oncology; Berlex Laboratories, Inc; Biogen IDEC, Inc; Blue Cross and Blue Shield Association; The Lynde and Harry Bradley Foundation; BRT Laboratories, Inc; Cedarlane Laboratories Ltd; Celgene Corp; Cell Pathways; Cell Therapeutics, Inc; CelMed Biosciences; Centocor, Inc; Cubist Pharmaceuticals; Dynal Biotech ASA; Edwards Lifesciences RMI; Endo Pharmaceuticals, Inc; Enzon Pharmaceuticals, Inc; ESP Pharma; Excess, Inc; Fujisawa Healthcare, Inc; Gambro BCT, Inc; Genzyme; GlaxoSmithKline, Inc; Human Genome Sciences; ICN Pharmaceuticals, Inc; ILEX Oncology; Kirin Brewery Company; Ligand Pharmaceuticals, Inc; Eli Lilly and Company; Nada and Herbert P. Mahler Charities; Merck & Company; Millennium Pharmaceuticals; Miller Pharmacal Group; Milliman USA, Inc; Miltenyi Biotec; The Irving I. Moskowitz Foundation; National Leukemia Research Association; National Marrow Donor Program; NeoRx Corporation; Novartis Pharmaceuticals, Inc; Novo Nordisk Pharmaceuticals; Ortho Biotech, Inc; Osiris Therapeutics, Inc; PacifiCare Health Systems; Pall Medical; Pfizer US Pharmaceuticals; Pharmametrics; Pharmion Corp; Protein Design Labs; QOL Medical; Roche Laboratories; Schering AG; StemCyte, Inc; StemCell Technologies, Inc; Stemco Biomedical; StemSoft Software, Inc; SuperGen, Inc; Sysmex; THERAKOS, a Johnson & Johnson Co; University of Colorado Cord Blood Bank; Upside Endeavors; ViaCell, Inc; ViaCor Biotechnologies; WB Saunders Mosby Churchill; Wellpoint Health Network; and Zymogenetics, Inc.
Authors’ Disclosures of Potential Conflicts of Interest
Although all authors completed the disclosure declaration, the following author or immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed discription of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
NOTES
Authors’ disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Asbell SO, Vijayakumar S: Racial differences in prostate-specific antigen levels in patients with local-regional prostate cancer. Prostate 31:42-46, 1997
Iselin CE, Box JW, Vollmer RT, et al: Surgical control of clinically localized prostate carcinoma is equivalent in African-American and white males. Cancer 83:2353-2360, 1998
Kim JA, Kuban DA, el-Mahdi AM, et al: Carcinoma of the prostate: Race as a prognostic indicator in definitive radiation therapy. Radiology 194:545-549, 1995
Eley JW, Hill HA, Chen VW, et al: Racial differences in survival from breast cancer: Results of the National Cancer Institute Black/White Cancer Survival Study. JAMA 272:947-954, 1994
Ownby HE, Frederick J, Russo J, et al: Racial differences in breast cancer patients. J Natl Cancer Inst 75:55-60, 1985
Shavers VL, Harlan LC, Stevens JL: Racial/ethnic variation in clinical presentation, treatment, and survival among breast cancer patients under age 35. Cancer 97:134-147, 2003
Arbes SJ, Slade GD: Racial differences in stage at diagnosis of screenable oral cancers in North Carolina. J Public Health Dent 56:352-354, 1996
Sherman ME, Devesa SS: Analysis of racial differences in incidence, survival, and mortality for malignant tumors of the uterine corpus. Cancer 98:176-186, 2003
Mayberry RM, Coates RJ, Hill HA, et al: Determinants of black/white differences in colon cancer survival. J Natl Cancer Inst 87:1686-1693, 1995
Clegg LX, Li FP, Hankey BF, et al: Cancer survival among US whites and minorities: A SEER (Surveillance, Epidemiology, and End Results) Program population-based study. Arch Intern Med 162:1985-1993, 2002
National Cancer Institute: Cancer health disparities. http://www.nci.nih.gov/newscenter/healthdisparities
Franzini L, Williams AF, Franklin J, et al: Effects of race and socioeconomic status on survival of 1,332 black, Hispanic, and white women with breast cancer. Ann Surg Oncol 4:111-118, 1997
Lannin DR, Mathews HF, Mitchell J, et al: Influence of socioeconomic and cultural factors on racial differences in late-stage presentation of breast cancer. JAMA 279:1801-1807, 1998
Moul JW, Douglas TH, McCarthy WF, et al: Black race is an adverse prognostic factor for prostate cancer recurrence following radical prostatectomy in an equal access health care setting. J Urol 155:1667-1673, 1996
Bain RP, Greenberg RS, Whitaker JP: Racial differences in survival of women with breast cancer. J Chronic Dis 39:631-642, 1986
Serna DS, Lee SJ, Zhang MJ, et al: Trends in survival rates after allogeneic hematopoietic stem-cell transplantation for acute and chronic leukemia by ethnicity in the United States and Canada. J Clin Oncol 21:3754-3760, 2003
Loberiza FR Jr: Report on state of the art in blood and marrow transplantation. Volume 10, Issue 2. http://www.ibmtr.org/newsletter/newsletter.asp
Bhatia S, Sather HN, Heerema NA, et al: Racial and ethnic differences in survival of children with acute lymphoblastic leukemia. Blood 100:1957-1964, 2002
Pollock BH, DeBaun MR, Camitta BM, et al: Racial differences in the survival of childhood B-precursor acute lymphoblastic leukemia: A Pediatric Oncology Group study. J Clin Oncol 18:813-823, 2000
Sekeres MA, Peterson B, Dodge RK, et al: Differences in prognostic factors and outcomes in African-Americans and whites with acute myeloid leukemia. Blood 103:4036-4042, 2004
US Census Bureau: Population by race and Hispanic or Latino origin for the United States: 1990 and 2000. http://www.census.gov/population/www/cen2000/phc-t1.html
Mitchell JM, Meehan KR, Kong J, et al: Access to bone marrow transplantation for leukemia and lymphoma: The role of sociodemographic factors. J Clin Oncol 15:2644-2651, 1997
Escarce JJ, Epstein KR, Colby DC, et al: Racial differences in the elderly’s use of medical procedures and diagnostic tests. Am J Public Health 83:948-954, 1993
Pappas G, Queen S, Hadden W, et al: The increasing disparity in mortality between socioeconomic groups in the United States, 1960 and 1986. N Engl J Med 329:103-109, 1993(K. Scott Baker, Fausto R.)
University of Nebraska Medical Center, Omaha, NE
Center for International Blood and Marrow Transplant Research, Medical College of Wisconsin, Milwaukee, WI
Columbia University, New York
University of Rochester Medical Center, Rochester
Our Lady of Mercy Comprehensive Cancer Center, Bronx, NY
Cleveland Clinic Foundation, Cleveland, OH
St Joseph Hospital, Orange, CA
South Florida Bone Marrow Stem Cell Transplant Institute, Boynton Beach, FL
Hospital Mexico, San Jose, Costa Rica
Hospital Privado de Cordoba, Cordoba, Argentina
Bristol Children’s Hospital, Bristol, United Kingdom
British Columbia Children’s Hospital, Vancouver, British Columbia
Princess Margaret Hospital, Toronto, Ontario, Canada
University of Hamburg-Eppendorf, Hamburg, Germany
ABSTRACT
PURPOSE: We previously reported a higher risk of mortality among Hispanics after allogeneic hematopoietic stem-cell transplantation (HSCT). However, it is not known how specific post-transplantation events (acute or chronic graft-versus-host disease [GVHD], treatment-related mortality [TRM], and relapse) may explain mortality differences. The purpose of this study was to examine the relationship between ethnicity and post-transplantation events and determine their net effect on survival.
PATIENTS AND METHODS: We identified 3,028 patients with acute myeloid leukemia, acute lymphoblastic leukemia, or chronic myeloid leukemia reported to the International Bone Marrow Transplant Registry between 1990 and 2000 who received an HLA-identical sibling HSCT after a myeloablative conditioning regimen in the United States. There were 2,418 white patients (80%) and 610 ethnic minority patients (20%), of whom 251 were black (8%), 122 were Asian (4%), and 237 were Hispanic (8%). Cox proportional hazards regression was used to compare outcomes between whites and ethnic minorities while adjusting for other significant clinical factors.
RESULTS: No statistically significant differences in the risk of acute or chronic GVHD, TRM, or relapse were found between whites and any ethnic minority group. However, Hispanics had higher risks of treatment failure (death or relapse; relative risk [RR] = 1.30; 95% CI, 1.08 to 1.54; P = .004) and overall mortality (RR = 1.23; 95% CI, 1.03 to 1.47; P = .02).
CONCLUSION: The higher risks of treatment failure and mortality among Hispanics may be the net result of modest but not statistically significant increases in both relapse and TRM and cannot be accounted for by any single transplantation-related complication. Further studies should examine the role of social, economic, and cultural factors.
INTRODUCTION
Several studies have examined inherent biologic differences between racial and ethnic subgroups of cancer patients to determine whether there are differences in tumor presentation, disease histology, stage at diagnosis, or response to therapy. For example, African American males with prostate cancer have higher rates of metastatic disease and more advanced-stage tumors at diagnosis.1-3 Similar findings have been reported for cancer of the breast,4-6 endometrium,7,8 colon,9 and oral cavity.7 However, no study to date has clearly identified any specific biologic factor to explain these differences in presentation. These factors have also been shown to have a negative impact on survival after a cancer diagnosis, which has been demonstrated by the facts that African American, American Indian, and Alaskan or Hawaiian native patients have higher risks of cancer death than whites10 and that, in the United States, African Americans have the highest cancer incidence and death rates overall.11
There are other factors that may influence differences in outcomes among racial and ethnic groups. Lower socioeconomic status (SES) has been found to contribute to higher mortality among racial and ethnic minorities with cancer. Similarly, cultural differences in attitudes towards seeking medical care and the potential lack of access to primary care (and subsequently delayed diagnosis) among ethnic minorities may contribute to discrepant outcome. However, even controlling for stage of disease, most studies suggest that SES alone cannot explain differences in outcome among racial and ethnic groups.12-15
Although a relatively large amount of literature exists on health disparities in ethnic minorities being treated for cancer, there are only limited published data that explore whether differences in outcomes exist for ethnic minorities when their treatment includes hematopoietic stem-cell transplantation (HSCT). We previously reported an analysis of trends in survival rates among ethnic groups after HSCT and examined three time periods (1985 to 1989, 1990 to 1994, and 1995 to 1999) to determine whether survival differences existed. This study found that, compared with whites, Hispanics had lower 1- and 3-year adjusted survival rates between 1995 and 1999, but there were no significant differences in the survival rates comparing whites with blacks or Asians in any of the time periods.16 However, that study did not investigate whether specific post-transplantation events, such as acute or chronic graft-versus-host disease (GVHD), treatment-related mortality (TRM), and relapse, differed among ethnic groups and contributed to mortality differences. Therefore, the purpose of this study was to examine the relationship between ethnicity and post-transplantation events and determine their net effect on survival.
PATIENTS AND METHODS
Data Sources
This observational study included patients who received HLA-identical sibling HSCT from 1990 to 2000 and were reported to the Center for International Blood and Marrow Transplant Research (CIBMTR). The CIBMTR is a voluntary working group of more than 450 transplantation centers worldwide that contribute detailed data on consecutive allogeneic HSCT to a Statistical Center at the Health Policy Institute of the Medical College of Wisconsin in Milwaukee or the National Marrow Donor Program Coordinating Center in Minneapolis. Approximately two thirds of all active transplantation centers worldwide report data to the registry. The registry database includes information on 40% to 45% of all patients who have received an allogeneic transplantation since 1970, with annual updates. Computerized checks for errors, review of submitted data by physicians, and on-site audits of participating centers are used to monitor the quality of the data.
The CIBMTR collects data at the following two levels: registration and research. Registration data include disease type, age, sex, pretransplantation performance status, disease stage and chemotherapy responsiveness, date of diagnosis, donor and graft type (bone marrow–and/or blood-derived stem cells), high-dose conditioning regimen, post-transplantation engraftment, disease recurrence and survival, development of a new malignancy, and cause of death. Requests for data on disease recurrence or death for registered patients are at 6-month intervals. All CIBMTR centers contribute registration data on all patients. Research data is collected on subsets of registered patients selected using a weighted randomization scheme, including comprehensive pre- and post-transplantation clinical information.
Patients
Patients with acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), or chronic myeloid leukemia (CML) who received an HLA-identical sibling HSCT between 1990 and 2000 in transplantation centers within the United States and had full research forms were included in this study. By limiting the analysis to HLA-identical sibling transplantations, we limited the possible influence of HLA-associated genetic factors on the outcomes of interest. Patients receiving reduced-intensity conditioning regimens were excluded. Only patients with full research data were included because we needed extensive information on the many clinical variables (see Statistical Analysis) to be adjusted for in the multivariate models. To provide assurance that the research patients were representative of all registered patients, basic patient-, disease-, and transplantation-related characteristics and relapse and survival rates between research and registered patients were compared; no differences were noted.
The patient’s ethnicity was abstracted from data submitted by the transplantation center to the CIBMTR. Patients were categorized into four ethnic cohorts: white, black, Hispanic, and Asian. Patients reported as being of European or Western Russian or Caucasian unspecified descent, according to the CIBMTR database, were classified as white. Patients reported as being of African American, Caribbean black, South or Central American black, or black unspecified descent were classified as black. Patients reported as being of Caribbean Hispanic, Mexican/Southwestern United States Hispanic, South or Central American Hispanic, or Hispanic unspecified descent were classified as Hispanic. Patients reported as being of Asian Indian, Filipino, Hawaiian/Polynesian, Japanese, Korean, Northern Chinese, Southeast Asian/Southern Chinese, or Asian unspecified descent were classified as Asian. Patients not classified into one of these four groups, such as Native Americans, were excluded because study numbers precluded further analysis.
Study End Points
The study outcomes evaluated were grades 2 to 4 acute GVHD and chronic GVHD in patients surviving more than 90 days with evidence of engraftment. Other clinical outcomes evaluated included overall survival, leukemia-free survival (LFS; survival without leukemia after transplantation), disease recurrence, and TRM. TRM is defined as death in the first 28 days after transplantation or death in remission. For LFS, patients were considered as having treatment failure at the time of relapse or death from any cause; patients who were alive were censored at the last follow-up evaluation.
Statistical Analysis
Patient-, disease-, and transplantation-related variables were compared among the four racial/ethnic groups using the 2 statistic for categoric variables and the Kruskal-Wallis test for continuous variables. Univariate probabilities of LFS and survival were calculated using the Kaplan-Meier method; the log-rank test was used for univariate comparisons. Probabilities of TRM and leukemia relapse were calculated using cumulative incidence curves to accommodate competing risks.
Multivariate analyses of acute and chronic GVHD, TRM, leukemia relapse, LFS, and overall survival were performed using Cox proportional hazards regression, with the racial/ethnic group forced in all models (main effect) and using whites as the reference group. Patient-, disease-, and transplantation-related covariates examined include (Table 1) age at transplantation, sex, Karnofsky performance score at transplantation, disease type, disease stage at transplantation, presence of comorbid medical risk factors (hypertension, diabetes mellitus, and smoking), type of graft, interval from diagnosis to transplantation, donor-recipient sex matching, donor-recipient cytomegalovirus serologic status, type of GVHD prophylaxis administered, preparative regimen administered, and year of transplantation. Disease type, disease stage, duration of first complete remission, and interval from diagnosis to transplantation were grouped together to form interactive covariates based on previous knowledge of their additive contributory effect on outcomes. The assumption of proportional hazards was tested using a time-dependent covariate. Models were stratified on covariates found to be nonproportional, except those of the main effect. Covariates significantly associated with the outcome were identified by forward stepwise selection at a significance level of P .05. First-order interactions between racial/ethnic group and all significant covariates were tested. No interactions were noted. However, interaction was noted between disease type and duration of first complete remission or interval from transplantation to diagnosis; thus, these covariates were entered in the model building in combinations. The overall covariate effects were tested using the Wald test. Because of multiple comparisons in the main effect, P .03 was considered statistically significant (Bonferoni correction). All P values were calculated as two-sided tests.
RESULTS
We identified 3,028 patients with ALL, AML, or CML reported to the CIBMTR who received an HLA-identical sibling HSCT after a myeloablative conditioning regimen in the United States between 1990 and 2000. There were 2,418 whites (80%) and 610 ethnic minorities including 251 blacks (8%), 122 Asians (4%), and 237 Hispanics (8%). Characteristics of the cohort by racial/ethnic group are listed in Table 2. In general, more Hispanics underwent transplantation at a younger age (< 1 to 20 years) than whites, blacks, or Asians, and more whites and blacks underwent transplantation at an older age (> 40 years). AML was the most common indication for HSCT in whites, CML was the most common indication for HSCT in blacks, and ALL was the most common indication for HSCT in Asians and Hispanics. Comorbid medical risk factors, including hypertension and diabetes, were more common in blacks, although whites were more likely to be smokers. Ethnic minorities were more likely to receive HSCT more than 1 year after diagnosis compared with whites. Ethnic minorities were also more likely to have both donor and recipient cytomegalovirus-positive status. Hispanics were more likely to have had their HSCT in the more recent treatment time period, with more than two thirds of the HSCTs performed in 1996 to 2000.
Acute and Chronic GVHD
Tables 3 and 4 show the multivariate analyses for acute and chronic GVHD. No statistically significant differences in acute and chronic GVHD were noted among the four racial/ethnic groups, as noted by the overall 3 df test. However, Asians tended to have a marginally reduced relative risk (RR) of having grades 2 to 4 acute GVHD (RR = 0.68; 95% CI, 0.47 to 0.98; P = .04) when compared with the white cohort.
TRM and Relapse
The risk of TRM (Table 5) was not increased when comparing blacks, Asians, or Hispanics with whites. There was also no statistically significant difference in the risk of relapse (Table 6) between whites and any of the racial/ethnic minority groups.
Treatment Failure and Mortality
The multivariate analyses of treatment failure (defined as the risk of death or relapse) and overall mortality are shown in Tables 7 and 8. Compared with the white cohort, Hispanics were found to have a significantly higher risk of treatment failure (RR = 1.30; 95% CI, 1.08 to 1.54; P = .004), which resulted in an inferior LFS, and a higher risk of overall mortality (RR = 1.23; 95% CI, 1.03 to 1.47; P = .02), which resulted in lower overall survival. Figures 1 and 2 show the probability of LFS and overall survival, respectively, according to racial/ethnic group.
DISCUSSION
In this study of patients undergoing HLA-identical sibling donor HSCT for acute or chronic leukemia in the United States, we have shown a surprising similarity of outcomes comparing ethnic minorities with whites for most major transplantation-related outcomes including acute and chronic GVHD, relapse, and TRM. However, for the Hispanic cohort, the overall risk of treatment failure (death or relapse) and mortality was increased. Our data suggest that this may be the net result of modest, although not statistically significant, increases in both relapse and TRM and cannot be accounted for by any single transplantation-related complication.
The trend towards a higher risk of relapse in the Hispanic group may be partially attributed to the fact that a higher percentage of Hispanic patients underwent HSCT for a diagnosis of ALL than the other groups. Previous studies have indicated that, depending on age at HSCT and stage of disease, survival after HSCT for ALL is inferior to that reported for AML or CML.17 Disease stage and age at time of HSCT also affect prognosis; however, there was no difference in the disease stage among the racial/ethnic groups, and as a whole, the Hispanic group was younger than the other groups. Similarly, the prevalence of the comorbid medical risk factors evaluated in our study was not increased in the Hispanic cohort. However, the differences in treatment failure and mortality between whites and Hispanics persisted after adjusting for the significant clinical variables in the multivariate models (including diagnosis), indicating that some specific ethnically associated biologic variable or other sociocultural factors that were not measured (eg, SES, compliance, cultural attitudes, and so on) may account for this difference. A large study on the outcome of childhood ALL from the Children’s Cancer Study Group found that, for children with high-risk leukemia, Hispanics (but not blacks or Asians) had significantly worse outcomes than whites.18 A similar finding was reported from a study in more than 5,000 children with ALL treated on Pediatric Oncology Group treatment protocols in which a significant excess risk of mortality was found for both black and Hispanic children.19 In both studies, clinically important prognostic variables were adjusted for, and in the Children’s Cancer Study Group study, adjustments were able to be made for SES, although in the Pediatric Oncology Group analysis, this could not be accounted for. However, these studies suggest that, at least for children with ALL treated with standard chemotherapy protocols, ethnicity (particularly for Hispanics) is an independent predictor of outcome. This question has not been as adequately addressed in adult patients with leukemia, except for one study that compared prognostic factors and outcomes between white and black patients with AML and found that black men (but not black women) had a worse overall survival and that there were differences with respect to important prognostic factors.20
It is also interesting to note that few transplantations for Hispanics were reported in the early time period and that 65% of transplantations for Hispanics occurred in the most recent time period (1996 to 2000). This may, in part, be explained by the fact that the number of individuals of Hispanic or Latino origin living in the United States increased by 58% over the 1990 to 2000 time period.21 However, if this were the only reason, one would expect similar increases in the number of transplantations in Asians, whose population increase was similar to that of Hispanics, and also in blacks, whose population increased by 15% to 20% compared with a 3% to 5% increase in the US white population over this time period. An alternative explanation is the apparent increase in offering HSCT as a feasible treatment for hematologic malignancies (possibly an improvement in access) among racial or ethnic minorities. Unfortunately, we are not able to analyze specific issues related to access to medical care. Despite this, with the increasing use of HSCT in the minority population, the inferior outcome in the Hispanic cohort is of potential concern.
Access to highly specialized therapies, such as HSCT, may certainly impact on the number of procedures that are performed, but this is difficult to analyze in treatments such as HSCT. Determining the population who received HSCT from the total pool of eligible patients is difficult to document. Patient choice, physician choice, or other reasons are vital determinants to the issue of access. One report examined this issue by looking at hospital discharge data for 2 years (between 1988 and 1991) from four states in patients admitted for leukemia or lymphoma. The study found that blacks with leukemia or lymphoma were less likely than whites to undergo HSCT. However, they found that patients with Medicaid, self-pay insurance, or health maintenance organization enrollees were less likely to undergo HSCT than patients with private insurance,22 and this was independent of race or ethnicity. This study also found that Hispanics with leukemia were not less likely to receive HSCT than whites, although Hispanics with lymphoma were less likely to receive HSCT than whites.
Our study is limited by the fact that we are not able to account for socioeconomic and other racial/ethnic disparities in health care delivery that may have an impact on the outcome of cancer therapies in ethnic minorities. The CIBMTR collects limited data on SES, but the quality and completeness of these data preclude their inclusion in our analysis. There is a large amount of literature describing the impact of SES on health disparities, although most of the data focuses on comparisons of blacks with whites, and few studies have examined this issue in other ethnic minority groups. It is also possible that disparities in health care delivery and access may have an impact on the assessment and treatment of late complications after HSCT in ethnic minorities when they return home to their community health care system. Lower SES is associated with lower overall health care use, the use of fewer expensive technologic procedures,23 including HSCT,22 and lower satisfaction in the health care system. There is an inverse relationship between SES (based on income and educational attainment) and mortality regardless of race, although ethnic minorities are over-represented in the lower socioeconomic classes.24 However, if these issues were the explanation for the findings in our study, we would have expected to see similar findings in the black and Asian cohorts.
Before this study, there was little information available on how specific post-transplantation events, such as acute or chronic GVHD, relapse, and transplantation-related complications, differ among ethnic groups and potentially contribute to mortality differences. We have shown that, in this cohort of patients undergoing related donor HSCT for acute or chronic leukemia, these factors generally do not contribute to significant differences in mortality. The exception of a higher mortality rate in Hispanics can likely be attributed to slightly higher, but not statistically significant, increases in relapse and TRM. Further prospective studies are necessary to analyze the cultural and socioeconomic or biologic factors that may contribute to these findings.
Appendix
Supported by Public Health Service Grant No. U24-CA76518 from the National Cancer Institute, the National Institute of Allergy and Infectious Diseases, and the National Heart, Lung and Blood Institute; Agency for Healthcare Research and Quality; and grants from Aetna; AIG Medical Excess; Allianz Life/Life Trac; American Red Cross; American Society of Clinical Oncology; Amgen, Inc; anonymous donation to the Medical College of Wisconsin; AnorMED, Inc; Aventis Pharmaceuticals; Baxter Healthcare Corp; Baxter Oncology; Berlex Laboratories, Inc; Biogen IDEC, Inc; Blue Cross and Blue Shield Association; The Lynde and Harry Bradley Foundation; BRT Laboratories, Inc; Cedarlane Laboratories Ltd; Celgene Corp; Cell Pathways; Cell Therapeutics, Inc; CelMed Biosciences; Centocor, Inc; Cubist Pharmaceuticals; Dynal Biotech ASA; Edwards Lifesciences RMI; Endo Pharmaceuticals, Inc; Enzon Pharmaceuticals, Inc; ESP Pharma; Excess, Inc; Fujisawa Healthcare, Inc; Gambro BCT, Inc; Genzyme; GlaxoSmithKline, Inc; Human Genome Sciences; ICN Pharmaceuticals, Inc; ILEX Oncology; Kirin Brewery Company; Ligand Pharmaceuticals, Inc; Eli Lilly and Company; Nada and Herbert P. Mahler Charities; Merck & Company; Millennium Pharmaceuticals; Miller Pharmacal Group; Milliman USA, Inc; Miltenyi Biotec; The Irving I. Moskowitz Foundation; National Leukemia Research Association; National Marrow Donor Program; NeoRx Corporation; Novartis Pharmaceuticals, Inc; Novo Nordisk Pharmaceuticals; Ortho Biotech, Inc; Osiris Therapeutics, Inc; PacifiCare Health Systems; Pall Medical; Pfizer US Pharmaceuticals; Pharmametrics; Pharmion Corp; Protein Design Labs; QOL Medical; Roche Laboratories; Schering AG; StemCyte, Inc; StemCell Technologies, Inc; Stemco Biomedical; StemSoft Software, Inc; SuperGen, Inc; Sysmex; THERAKOS, a Johnson & Johnson Co; University of Colorado Cord Blood Bank; Upside Endeavors; ViaCell, Inc; ViaCor Biotechnologies; WB Saunders Mosby Churchill; Wellpoint Health Network; and Zymogenetics, Inc.
Authors’ Disclosures of Potential Conflicts of Interest
Although all authors completed the disclosure declaration, the following author or immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed discription of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
NOTES
Authors’ disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Asbell SO, Vijayakumar S: Racial differences in prostate-specific antigen levels in patients with local-regional prostate cancer. Prostate 31:42-46, 1997
Iselin CE, Box JW, Vollmer RT, et al: Surgical control of clinically localized prostate carcinoma is equivalent in African-American and white males. Cancer 83:2353-2360, 1998
Kim JA, Kuban DA, el-Mahdi AM, et al: Carcinoma of the prostate: Race as a prognostic indicator in definitive radiation therapy. Radiology 194:545-549, 1995
Eley JW, Hill HA, Chen VW, et al: Racial differences in survival from breast cancer: Results of the National Cancer Institute Black/White Cancer Survival Study. JAMA 272:947-954, 1994
Ownby HE, Frederick J, Russo J, et al: Racial differences in breast cancer patients. J Natl Cancer Inst 75:55-60, 1985
Shavers VL, Harlan LC, Stevens JL: Racial/ethnic variation in clinical presentation, treatment, and survival among breast cancer patients under age 35. Cancer 97:134-147, 2003
Arbes SJ, Slade GD: Racial differences in stage at diagnosis of screenable oral cancers in North Carolina. J Public Health Dent 56:352-354, 1996
Sherman ME, Devesa SS: Analysis of racial differences in incidence, survival, and mortality for malignant tumors of the uterine corpus. Cancer 98:176-186, 2003
Mayberry RM, Coates RJ, Hill HA, et al: Determinants of black/white differences in colon cancer survival. J Natl Cancer Inst 87:1686-1693, 1995
Clegg LX, Li FP, Hankey BF, et al: Cancer survival among US whites and minorities: A SEER (Surveillance, Epidemiology, and End Results) Program population-based study. Arch Intern Med 162:1985-1993, 2002
National Cancer Institute: Cancer health disparities. http://www.nci.nih.gov/newscenter/healthdisparities
Franzini L, Williams AF, Franklin J, et al: Effects of race and socioeconomic status on survival of 1,332 black, Hispanic, and white women with breast cancer. Ann Surg Oncol 4:111-118, 1997
Lannin DR, Mathews HF, Mitchell J, et al: Influence of socioeconomic and cultural factors on racial differences in late-stage presentation of breast cancer. JAMA 279:1801-1807, 1998
Moul JW, Douglas TH, McCarthy WF, et al: Black race is an adverse prognostic factor for prostate cancer recurrence following radical prostatectomy in an equal access health care setting. J Urol 155:1667-1673, 1996
Bain RP, Greenberg RS, Whitaker JP: Racial differences in survival of women with breast cancer. J Chronic Dis 39:631-642, 1986
Serna DS, Lee SJ, Zhang MJ, et al: Trends in survival rates after allogeneic hematopoietic stem-cell transplantation for acute and chronic leukemia by ethnicity in the United States and Canada. J Clin Oncol 21:3754-3760, 2003
Loberiza FR Jr: Report on state of the art in blood and marrow transplantation. Volume 10, Issue 2. http://www.ibmtr.org/newsletter/newsletter.asp
Bhatia S, Sather HN, Heerema NA, et al: Racial and ethnic differences in survival of children with acute lymphoblastic leukemia. Blood 100:1957-1964, 2002
Pollock BH, DeBaun MR, Camitta BM, et al: Racial differences in the survival of childhood B-precursor acute lymphoblastic leukemia: A Pediatric Oncology Group study. J Clin Oncol 18:813-823, 2000
Sekeres MA, Peterson B, Dodge RK, et al: Differences in prognostic factors and outcomes in African-Americans and whites with acute myeloid leukemia. Blood 103:4036-4042, 2004
US Census Bureau: Population by race and Hispanic or Latino origin for the United States: 1990 and 2000. http://www.census.gov/population/www/cen2000/phc-t1.html
Mitchell JM, Meehan KR, Kong J, et al: Access to bone marrow transplantation for leukemia and lymphoma: The role of sociodemographic factors. J Clin Oncol 15:2644-2651, 1997
Escarce JJ, Epstein KR, Colby DC, et al: Racial differences in the elderly’s use of medical procedures and diagnostic tests. Am J Public Health 83:948-954, 1993
Pappas G, Queen S, Hadden W, et al: The increasing disparity in mortality between socioeconomic groups in the United States, 1960 and 1986. N Engl J Med 329:103-109, 1993(K. Scott Baker, Fausto R.)