Abstract
Background
There is scarcity of data on differences in survival in acute myeloid leukemia (AML) patients by ethnicity. We utilized data from the Surveillance Epidemiology and End Result (SEER) database to investigate the ethnic disparities of survival in general U.S. population.
Methods
The SEER-18 Registry was used to identify adult (>=18) patients with AML as the only or the first primary cancer diagnosed from 1992 to 2010. We only included cases which were microscopically confirmed and actively followed. Cases that were alive without survival time, those resulted in death certificate/autopsy, and those with ethnicity recorded as unknown were excluded from this study. A total of 29,477 patients (54.5% males) were identified. For the subsequent analyses, various cohorts were formed. Age group cohorts included: 18-44 (5394; 18.3%), 45-54 (3751; 12.7%), 55-64 (4913; 16.7%), 65-74 (6513; 22.1%) and 75+ (8906; 30.2%). The total study period was divided into four groups, 1992-1995 (3409; 11.6%), 1996-2000 (5816; 19.7%), 2001-2005 (9984; 33.9%) and 2006-2010 (10268; 34.8%) for the survival analyses over time. Ethnic stratification used included White (21338; 72.4%), African American (AA: 2322; 7.9%), Asians/Pacific Islanders (A/PI: 2389; 8.1%), Native American/Alaskan Natives (NA/AN: 137; 0.5%) and Hispanics (3291; 11.2%). NA/AN categories were excluded from the final analysis due to their small numbers. Kaplan Meier (KM) curve and log rank test were used to evaluate association between patient characteristics and survival in overall population, OS, and AML-specific survival(AMLSS). Cox proportional hazards model was used for the analysis of association between patient characteristics and survival. Statistical analyses were carried out using SPSS version 16.0.0
Results
Median age at diagnosis for the patient population was 66 years. Median follow-up period was 6.17 years for the whole population.
Median OS for whole population was 6 months with highest survival among Hispanics and lowest among Whites (10 months versus 5 months, p <0.001). The AMLSS was highest for Hispanics and lowest for Whites (24 months versus 12 months, p <0.001). Median OS and AMLSS deteriorated significantly with advancing age (p<0.001). The median OS and AMLSS were the same for males and females (p>0.05), in overall population. OS for females were better than males among AA and Hispanic patients (p value <0.001). AMLS survival was better for A/PI females compared to males (median AMLSS 22 months vs. 17 months, p =0.015). When comparing survival among year of diagnosis cohorts, OS as well as AMLSS were comparable among 1992-1995 and 1996-2000 cohorts, (p>0.05); however, there was a gradual improvement in the more recent time period cohorts.
Results of the proportional hazard models indicated that when compared to Whites, the OS was best for A/PI and worst for AA patients (HR= 0.933, p= 0.006, and HR = 1.139, p <0.001 ,respectively). The OS was higher for females, younger patients, and for patients diagnosed during recent time period cohorts. Similarly, AMLS survival among Hispanics and AA was comparable to whites and, best for Asians/PI (HR 0.911, p =0.003).
Conclusions
This study demonstrated significant differences in survival rates among AML patients belonging to various ethnic groups with highest OS and AMLSS among A/PI AML patients.
Disclosures:
No relevant conflicts of interest to declare.
Nature and Extent of International Diversification: Empirical Evidence from India
