Association of Race and Breast Cancer Stage. DISCUSSION

Posted by James

We found that African-American women possess the same histological types of breast cancer, in similar locations in the breast, and have the same left/right breast distribution as Caucasian Americans. However, as with previous research in this area, our results found that African-American race is a predictor of advanced breast cancer stage at diagnosis. In addition, similar to Lannin et al., controlling for Medicaid insurance did not explain all of the racial variation in breast cancer stage. However, different from previous studies, we found no difference between the races for AJCC stage-4 breast cancer.

There are a few possible reasons for our findings. First, African Americans may have less access to care. Ideally, we would have liked to assess and control for a better indicator of socioeconomic status (SES) then just Medicaid insurance. Education level or annual income would be a preferred SES variable. When pre­vious researchers have investigated this topic and controlled for SES, they have still found a significant effect of race on stage at diagnosis. Second, breast cancer may act differently in African-American women. It may be harder to find on mammography and palpation or may grow at a faster pace than Caucasian Americans. Although there is no scientific data to suggest either of these possibilities, if true, these could cause the cancers of African-American women to be diagnosed at a more advanced stage. It is known that African Americans possess a lower incidence of breast cancer compared to Caucasians and, appropriately, their race is recognized as a predictor of a lower risk of developing the disease.

Our project was unique from previous research in that we investigated the variables of topography, morphology and laterality. We felt that one or more of these variables could explain, in addition to SES, some of the variation in stage demonstrated between the two races. However, since there was no difference for all three variables, this suggests that breast cancer is very similar for both races. Although several potential confounders were found to be different between the two races (age, receptor status, etc.), controlling for the variations had minimal impact on the outcomes of interest.

The study population should well represent the general population in Cincinnati. The TriHealth tumor registry is the largest in the county. The three hospitals serve urban, suburban and rural geographical populations. The percentage of African Americans in the study (11%) is identical to the overall county prevalence.

There are a few limitations of our project that need to be mentioned when considering our results. First, we have only regional data for our investigation. Although our results were consistent with previous investigations, there may be some geographical areas where this relationship of race and breast cancer stage is different. Second, our study possessed small populations in certain stages. In particular stages О, ЗА, ЗВ, ЗС, 2A and 2B all lacked power individually. The combined stages possessed better power; however, ideally our study would have had adequate numbers of patients to look at each individual stage. However, previous researchers with large, national data sets have also combined these stages for analysis. And third, our study would be stronger if we possessed a better control variable to assess SES than simply Medicaid insurance status. cialis super active online

Since these findings have been found across several populations, future research should focus on African Americans and why they get diagnosed later then Caucasian Americans. In particular, it would be important to see if access to care is the primary reason for the difference.