This is the first large study to evaluate the correlates of CRP and SAA in a cohort of breast cancer survivors. In this cross-sectional analysis, several variables were associated with both CRP and SAA (measured ~31 months post diagnosis): age, BMI, study site, ethnicity, vitamin E supplementation and history of heart failure. CRP was associated with several additional factors: waist circumference, smoking status, tamoxifen use and treatment. SAA, but not CRP, was associated with history of myocardial infarction. All significant associations remained statistically significant or nearly significant in the context of a multivariate model, reducing the likelihood that these associations are due to confounding by other variables examined in this study. Excluding individuals with extreme CRP values did not change the interpretation of our results, with the exception of additional statistically significant associations between CRP and race/ethnicity/study site, SAA and race/ethnicity/study site, and SAA and smoking. These exclusions eliminated statistically significant associations between heart failure and both CRP and SAA. The variables examined explained more of the variation in CRP concentrations (R2=0.35) than in SAA concentrations (R2=0.18).
Many of the associations reported here are consistent with observations from previous studies of healthy individuals. Increasing age [40
], African American race (compared to non-Hispanic white) [40
], and smoking [41
] are known to be associated with increasing CRP concentrations. In healthy individuals, elevated levels of CRP and SAA are associated with body fatness [21
] and sedentary lifestyles [21
]. Weight loss and exercise training have been shown to reduce CRP levels in healthy individuals [27
], while the latter reduces CRP in breast cancer survivors [42
]. Intervention studies have shown that alpha-tocopherol, a form of vitamin E with purported anti-inflammatory properties [43
], reduces serum concentrations of CRP in healthy individuals [44
], type 2 diabetics [44
], and smokers with acute coronary syndromes [46
]. Intervention studies also suggest that low doses of tamoxifen decrease serum CRP concentrations in healthy women [47
] and in women with ER positive breast tumors [46
], consistent with the observed association in this study.
Several biological mechanisms have been suggested to explain the relationships between CRP, SAA, and their correlates. Associations between CRP and body fatness or weight loss are believed to be linked to adipose tissue. Adipose tissue secretes IL-6 [27
], an important trigger for CRP production, and its abundance is likely associated with CRP concentrations [49
]. It has also been suggested that the accumulation of macrophages in adipose tissue contributes to a heightened inflammatory state, as macrophages are an additional source of pro-inflammatory molecules [50
]. Obesity is a negative prognostic factor for breast cancer [51
] and is hypothesized to influence prognosis through effects on circulating concentrations of estrogens, insulin and insulin-like growth factors. Recently, obesity was observed be of heightened prognostic significance for ER positive cancers [52
], supporting the hypothesis that obesity effects prognosis through estrogens [53
]. The relationship between obesity, inflammation, and breast cancer survival has not yet been explored.
Exercise may reduce CRP, independent of changes in body fatness, through modification of cytokine production at non-adipose sites such as skeletal muscles and mononuclear cells. Reductions in CRP may also occur indirectly, through improved endothelial function, increased insulin sensitivity, or reduced body weight [54
]. Several studies suggest that physical activity is associated with a modest decrease in mortality for breast cancer patients [55
], although the evidence is not entirely consistent [57
]. Physical activity may influence breast cancer survival through the inflammation-related mechanisms above, through decreases in estrogen exposure, or increases in energy expenditure [60
The association between age and CRP is complex, and may be related to a wide variety of factors, including dysregulation of cytokine response due to a lifetime antigen exposure, decreases in production of sex hormones, and increases in cytokine-producing fat tissue [61
]. Age is a prognostic factor for breast cancer; decreased survival has been observed for women (≥75 years), whose age limits diagnostic tests and examinations, as well as treatment choices [62
]. Women diagnosed at a young age also have a poor prognosis [63
], but these cancers appear to be etiologically distinct from cancers occurring in older women [64
The well-established correlation between smoking and chronic inflammation [41
] is likely due to smoking-induced tissue damage, alterations in leukocyte concentrations, and/or increases in concentrations of pro-inflammatory cytokines [41
]. There is evidence that cigarette smoking is associated with an increased risk for total mortality [65
], but not breast cancer mortality [67
], although the evidence is not entirely consistent [68
]. Smoking could influence breast cancer survival through many mechanisms, including changes in local immune function, systemic anti-tumor defenses, and coagulation status, in addition to the direct effects of smoke constituents that promote the growth of metastases [69
Tamoxifen is a selective estrogen receptor modulator that competitively binds to the estrogen receptor, inhibiting the effects of estrogen. Adjuvant tamoxifen treatment decreases mortality in patients with ER positive tumors [70
]. Tamoxifen also decreases serum concentrations of CRP in a dose-dependent fashion in women with ER positive tumors [71
]. It has been hypothesized that tamoxifen-related decreases in CRP may be attributable to the anti-estrogenic effect of tamoxifen on adipocyte cytokine production [48
]. If this is the case, tamoxifen may improve survival by reducing systemic, chronic inflammation, in addition to its effects on estrogen signaling in tumors.
CRP and SAA show associations with the use of beta blockers, ACE inhibitors, and lipid lowering medications when unadjusted for age, BMI, ethnicity, and study site. These are likely due to confounding as overweight and older individuals are more likely to be prescribed these medications and have elevated CRP and SAA concentrations. After adjustment, these associations are no longer observed. Similarly, CRP is associated with both ER and PR status prior to adjustment, but not associated with ER and PR status after adjustment. The reductions in the magnitude of both of these associations are due primarily to adjustment for BMI. Interestingly, increased inflammation, as measured by the Glasgow prognostic score (based on CRP and albumin concentrations) has been shown previously to be associated with ER negative tumor status (borderline; p=0.06) in patients with metastatic breast cancer [33
]; however, this association was not adjusted for BMI.
This is the first large study of well-assessed correlates of both CRP and SAA in breast cancer survivors. This study was limited by the timing of the measurements of CRP and SAA, which were taken approximately 31 months after diagnosis. We could not assess the correlates of CRP and SAA in women who died prior to this measurement or did not return for a follow-up interview. As a result, our sample is representative of long-term breast cancer survivors (≥2 years survival). Also, in this study, it is difficult to disentangle the effects of study site and ethnicity, because they are strongly associated, and the results must be interpreted with that in mind.
Because inflammation status may be an important prognostic factor for breast cancer, it is important to understand its relationships with other demographic, lifestyle, and clinical factors of prognostic importance. As in healthy women, measures of body fatness emerged as the most important predictors of these inflammatory markers in this cohort of breast cancer survivors.