|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 25, No 36 (December 20), 2007: pp. 5738-5741 © 2007 American Society of Clinical Oncology. DOI: 10.1200/JCO.2007.13.8271 Poor Survival for US Pacific Islander Cancer Patients: Evidence From the Surveillance, Epidemiology, and End Results Database: 1991 to 2004
From the School of Public Health; and the Nethersole School of Nursing, Chinese University of Hong Kong, Shatin, Hong Kong, China Address reprint requests to William Goggins, ScD, School of Public Health, Room 501, Chinese University of Hong Kong, Shatin, Hong Kong, China; e-mail: wgoggins{at}cuhk.edu.hk
Purpose: Although racial and ethnic differences in cancer survival in the United States have been studied extensively, little is known about cancer survival in US Pacific Islanders (PIs), a fast-growing and economically disadvantaged minority group. Methods: Using data from the US National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) registries, we compared cause-specific and all-cause survival for female breast, prostate, lung, colorectal, stomach and liver cancer for Native Hawaiians, Samoans, other PIs (including Tongans, Guamanians, and others), African Americans, and Native Americans with non-Hispanic whites using Cox proportional hazards models. Separate models were fitted adjusting for demographic factors only and demographic and disease severity variables. Results: Among all groups, Samoans were the most likely to present with advanced disease and had the worst cause-specific survival for all sites considered. Samoans had particularly poor results (adjusted for demographic variables only) for female breast (relative risk [RR] = 3.05; 95% CI, 2.31 to 4.02), colorectal (RR = 1.82; 95% CI, 1.37 to 2.41) and prostate (RR = 4.82; 95% CI, 3.38 to 6.88) cancers. Native Hawaiians and other PIs also had significantly worse cause-specific survival than did non-Hispanic whites for most sites, but generally had better survival than African Americans or Native Americans. Conclusion: Much of the survival disadvantage for PI groups appears to be a result of late diagnosis, and thus targeted interventions have much potential to reduce cancer mortality in this group. More research is needed to find explanations for the particularly poor cancer survival for Samoans in the United States.
Studies of ethnic variation in cancer survival rates in the United States have focused mainly on differences between African Americans and whites1 and have generally found that African Americans have a survival disadvantage for most sites relative to whites. Previously, most studies analyzed Asian and Pacific Islanders (APIs) as a single group, or excluded them from the analyses despite the fact that they are a diverse group in terms of genetics, socioeconomic status (SES), culture, and length of stay in the United States. Recently, some studies1-6 have examined larger API groups separately, with a few1-4 including separate analysis for Native Hawaiians, the largest PI group in the United States. To the authors' knowledge, no previous studies have looked at cancer survival for any other PI groups. Here, we present findings on cancer survival for three PI groups—Native Hawaiians, Samoans, and other PIs (including Tongans, Guamanians, Fiji Islanders, and others)—using population-based data from US cancer registries participating in the US National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) program.
We obtained data from all 17 US cancer registries participating in the SEER program. Analyses were restricted to cases diagnosed from 1991 to 2003 (with follow-up through 2004) because Samoans were not classified separately before 1991, and we wished to have sufficient follow-up time for meaningful analysis. The analysis included only histologically confirmed, malignant cases. Cox proportional hazards models were used to estimate adjusted relative risks (RRs) and their 95% CIs for all-cause and cause-specific mortality for cancers of the lung, female breast, prostate, colorectum, stomach, and liver. For the cause-specific mortality analyses, only deaths resulting from cancer of the site being analyzed were considered events, and patients were considered excluded at time of death if they died as a result of other causes. RRs were estimated for Native Hawaiians, Samoans, other PIs (Tongans, Guamanians, Fiji Islanders, others), African Americans, and Native Americans with non-Hispanic whites serving as the reference group. Results for African Americans and Native Americans, two other economically disadvantaged groups, were included for comparison purposes. The proportional hazards assumption was verified by visual examination of Kaplan-Meier survival plots. Two models were fitted for each type of mortality, one adjusting only for demographic factors including age (using linear and quadratic terms), sex, year of diagnosis and registry (through stratification). The other models additionally adjusted for disease characteristics including SEER modified stage (or SEER historic stage for stomach and liver cancers), tumor size (categorized into quartiles), grade (I, II, III, IV or none), histology, and estrogen and progesterone receptor status (for breast cancer). All disease variables were controlled for using stratification to avoid any problems resulting from nonproportional hazards. Analyses were restricted to malignant cases. All statistical tests were two-sided, and P values less than .05 were considered statistically significant.
Case numbers and stage distributions by ethnic group are shown in Table 1. For most sites, non-Hispanic whites had the most favorable stage distributions, whereas Samoans were the most likely to be diagnosed with advanced disease. Native Hawaiians tended to have less favorable stage distributions than did whites, but better than African Americans or Native Americans. Table 2 shows adjusted RRs, 95% CIs, and P values for the four types of models by site. Among all groups, Samoans had the worst cause-specific survival for all sites when adjusting for demographic variables only, and they had particularly poor results for prostate (RR = 4.82; 95% CI, 3.38 to 6.88), female breast (RR = 3.05; 95% CI, 2.31 to 4.02), and colorectal (RR = 1.82; 95% CI, 1.37 to 2.41) cancers. Adjustment for disease severity attenuated these results, but the Samoan survival disadvantage remained significant for prostate, breast, colorectal, and stomach cancers. In models adjusting for demographic variables, Native Hawaiians had modest but significantly worse cause-specific survival for all sites except liver cancer, but adjustment for disease severity reduced the RRs to nonsignificance for all sites except colorectal cancer. Patients in the "other PI" category had significantly worse cause-specific survival for prostate, breast, and lung cancers after adjustment for demographic variables. Further adjustment for disease severity attenuated the results for this group to nonsignificance for all sites. African Americans and Native Americans also had significantly worse cause-specific survival than did non-Hispanic whites for all sites, except for liver cancer for Native Americans. Their survival experience tended to be similar to or somewhat worse than the Native Hawaiian and other PI groups, but better than Samoans. Results for all-cause survival models were generally similar to those for cause-specific survival.
Our finding that US PIs have generally worse cancer survival than non-Hispanic Whites is not surprising since similar results have been found for other economically disadvantaged groups. However, the very poor cancer survival and stage at diagnosis distributions for Samoans were unexpected. The fact that Samoan cancer survival was worse both before and after adjustment for disease severity likely indicates both late diagnosis and poorer treatment after diagnosis. The lower SES of Samoans is almost certainly a contributing factor. Although data on SES indicators make it clear that Samoans and Native Hawaiians are economically and educationally disadvantaged relative to whites, Samoan SES status relative to African Americans and Native Americans is less clear. US Census data (2000)7 show that higher percentages of African Americans (23.5%) and Native Americans (24.5%) were below the poverty line than were Hawaiians (14.8%) and Samoans (18.3%); however, only 10.5% of Samoans 25 years of age or older had a Bachelor's degree or higher, compared with 26.1% of non-Hispanic whites, 15.2% of Hawaiians, 14.3% of African Americans, and 11.5% of Native Americans.7 In addition, although Samoan median household income in 2000 was higher than that of African Americans or Native Americans, their per-capita income is lower because of their higher mean household size.7 Cultural factors may also be important. A qualitative study on attitudes toward mammography and breast cancer in Samoan women found that Samoans tended to view sickness as unpreventable and cancer as an automatic death sentence.8 Thus, they may be less likely to be attentive to possible early warning signs of cancer. The same study also indicated that Samoan women would consider using Samoan traditional medicine to treat cancer,8 which could potentially lead to lower survival rates if they relied exclusively on traditional healers. In addition, because US Samoans are a relatively small minority group, with a population of 91,029 (133,281 including those of part-Samoan ancestry) in 2000,7 they may not have received targeted interventions to the same extent as other larger disadvantaged minority groups. Finally, genetic differences cannot be ruled out as a partial explanation for between group differences in disease severity and cancer survival.9 Our study does have some limitations. It is possible that Samoans with advanced disease came to SEER areas from American Samoa for treatment, thus biasing survival results. However, our analysis indicates that Samoan cancer survival for cases reported to the Hawaiian registry, which is geographically the closest to American Samoa and thus the most likely destination for American Samoans seeking treatment, was actually better than for the other registries for all sites (data not shown). The possibility of racial misclassification is always a concern for research on health disparities. A recent study10 found that SEER data on race, which is based on medical records, was generally of very high quality, except for the Native American classification. However, this study examined APIs as a single group, and thus did not examine the possibility of misclassification within this group (eg, misclassifying a Samoan as a Native Hawaiian). The "other PI" group used in our analysis is a racially diverse group, consisting of Polynesian groups such as Tongans, Micronesians (Guamanians, others), and Melanesians (Fiji Islanders, others). They are also diverse in terms of SES. Unfortunately, the sample sizes for these groups are presently too small to allow meaningful analysis of survival for individual groups. In the future, such analyses should be possible for some of these groups, particularly because SEER has recently initiated coverage of the entire state of California. Our analysis highlights the need for separate reporting of results for API2w3 subgroups in cancer epidemiology studies. Because our study has found that at least one API subgroup, Samoans, has worse cancer survival than African Americans, and other studies have found that some Asian subgroups have generally better survival than whites3,4,11 for some sites, this would indicate that differences in cancer survival within the API group is potentially greater than differences between African Americans and whites. Although further research is necessary to clarify the causes of their poorer survival rates, it does seem clear that US PIs, and particularly Samoans, would benefit from further interventions designed to enhance their cancer awareness, particularly because they have above-average incidence rates for some sites.12,13
The author(s) indicated no potential conflicts of interest.
Conception and design: William B. Goggins Data analysis and interpretation: William B. Goggins, Grace K.C. Wong Manuscript writing: William B. Goggins, Grace K.C. Wong Final approval of manuscript: William B. Goggins, Grace K.C. Wong
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. 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 2. Li CI, Malone KE, Daling JR: Differences in breast cancer stage, treatment, and survival by race and ethnicity. Arch Intern Med 163:49-56, 2003 3. Chien C, Morimoto LM, Tom J, et al: Differences in colorectal carcinoma stage and survival by race and ethnicity. Cancer 104:629-639, 2005[CrossRef][Medline] 4. Pagano IS, Morita SY, Dhakal S, et al: Time dependent ethnic convergence in colorectal survival in Hawaii. BMC Cancer 3:5, 2003[CrossRef][Medline] 5. Lin SS, Clarke CA, Prehn AW, et al: Survival differences among Asian subpopulations in the United States after prostate, colorectal, breast, and cervical carcinomas. Cancer 94:1175-1182, 2002[CrossRef][Medline] 6. Sun LM, Li CI, Huang EY, et al: Survival differences by race in nasopharyngeal carcinoma. Am J Epidemiol 165:271-278, 2007 7. US Census Bureau: Census 2000, Fact sheets for race, ethnic or ancestry group. http://factfinder.census.gov 8. Ishida DN, Toomata-Mayer TF, Braginsky NS: Beliefs and attitudes of Samoan women toward early detection of breast cancer and mammography utilization. Cancer 91:262-266, 2001[Medline] 9. Whitworth A: New research suggests access, genetic differences play role in high minority cancer death rate. J Natl Cancer Inst 98:669, 2006 10. Clegg LX, Reichman ME, Hankey BF, et al: Quality of race, Hispanic ethnicity, and immigrant status in population-based cancer registry data: Implications for health disparity studies. Cancer Causes Control 18:177-187, 2007[CrossRef][Medline] 11. Hashibe M, Gao T, Li G, et al: Comparison of bladder cancer survival among Japanese, Chinese, Filipino, Hawaiian and Caucasian populations in the United States. Asian Pac J Cancer Prev 4:267-273, 2003[Medline] 12. Mishra SI, Luce-Aoelua P, Wilkens LR, et al: Cancer among American-Samoans: Site-specific incidence in California and Hawaii. Int J Epidemiol 25:713-721, 1996 13. Pike MC, Kolonel LN, Henderson BE, et al: Breast cancer in a multiethnic cohort in Hawaii and Los Angeles: Risk factor-adjusted incidence in Japanese equals and in Hawaiians exceeds that in whites. Cancer Epidemiol Biomarkers Prev 11:795-800, 2002 Submitted August 1, 2007; accepted September 20, 2007.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2007 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
