Systemic and Tumor Level Iron Regulation in Men With Colorectal Cancer: A Case Control Study

Cenk K Pusatcioglu; Elizabeta Nemeth; Giamila Fantuzzi; Xavier Llor; Sally Freels; Lisa Tussing-Humphreys; Robert J Cabay; Rose Linzmeier; Damond Ng; Julia Clark; Carol Braunschweig


Nutr Metab. 2014;11(21) 

In This Article


Individuals with CRC frequently have anemia or low iron status which may, at least in part, be due to an inflammatory response termed the anemia of chronic disease (ACD).[18,20,21,28,29] The ACD is thought to manifest through increased hepatic production of the iron regulatory protein hepcidin, which can down-regulate the iron exporter FPN.[10,13,30] Reduction of FPN limits iron flow into circulation from both the diet and storage sites and promotes tissue iron sequestration. Beyond impaired iron status and red blood cell production, the ACD may be relevant for persons with CRC given that hepcidin-induced reductions in dietary iron absorption in the proximal gut may increase colonic iron exposure. Colonic tumors may benefit from this alteration since iron is a substrate for cell proliferation, cancer progression and metastasis.[1,2,5,6] The primary aims of this case–control study were to simultaneously examine systemic and tumor level iron status and regulation in men with CRC compared to controls and determine if systemic or tumor level hepcidin expression was associated with tumor iron accumulation.

We found cases had significantly more iron-restricted erythropoiesis based on Hb and sTfR compared to controls. We could not differentiate whether this mild anemia was due to inflammation or also to frank iron deficiency. Often, gastrointestinal bleeding can contribute to iron deficiency in CRC; however, we excluded persons with known gastrointestinal bleeding, suggesting other factors contributed to the iron restriction observed in cases.[29,31] Typically when iron deficiency is present, serum hepcidin concentrations are undetectable.[8] Indeed, cases had lower serum hepcidin compared to controls; however, this difference was not significant and was within the normal range. This suggests hepcidin levels were inappropriately elevated given the patients' iron-restricted erythropoiesis. Hepcidin expression is simultaneously regulated by inflammation, iron stores and erythropoiesis.[7,10] Further, hepcidin levels are ultimately determined by the relative strength of these opposing stimuli.[12,32] In contrast to controls where serum hepcidin was correlated with Hb, serum hepcidin within cases was not significantly correlated with Hb suggesting that multiple signals may be influencing its production. Significantly greater systemic and colonic mucosal inflammation was observed in cases compared to controls which may have provided the necessary stimuli to promote increased hepatic hepcidin production, despite elevated sTfR and low iron status. This phenotype, in which serum hepcidin is simultaneously regulated by low iron status and inflammation, has been previously reported in morbid obesity.[26,33,34] Similar to our cases, obese participants in these studies presented with low iron status, elevated systemic inflammation and serum hepcidin concentrations within the normal range.[26] Like obesity, the iron profile observed in our cases is consistent with a mixed anemia for which hallmarks of the ACD and iron deficiency coexist.[9] This phenotype allows for the mobilization of iron from body stores but impairs iron repletion as a result of reduced dietary iron absorption.[35] Therefore, over time, inflammation-induced chronically elevated hepcidin would precipitate cellular iron depletion as body iron losses exceed dietary absorption effort.[27,36] Importantly, this type of mixed anemia has the potential to increase gastrointestinal iron exposure in persons with CRC.

We assessed colonic mucosal iron transport and regulation and found no statistical difference in expression of DMT-1 or FPN between cases and controls. Directionally our DMT-1 and FPN expression were similar to those of Brookes et al..[1] They reported increased colonocyte expression of iron influx (DMT-1) and decreased iron efflux (FPN) proteins in CRC compared to controls. This suggests that CRC tissue is associated with greater colonic adenocarcinoma iron uptake and sequestration compared to healthy mucosa. We observed very low hepcidin expression in the colonic tumors and healthy mucosa. This is not unexpected given that hepcidin is primarily produced in hepatocytes. In our previous analysis, abdominal subcutaneous and visceral adipose tissue also expressed very low mRNA expression compared to the liver.[26] Very little is known regarding the role of extra-hepatic hepcidin in systemic and tissue-level iron regulation. Our findings indicate that colonic hepcidin has a minimal role in tumor iron regulation. Ward et al. reported elevated tumor hepcidin expression within 34% of cases compared to non-involved healthy mucosa; however, other factors that regulate hepcidin expression including inflammation, iron status and dietary iron intake were not examined, limiting interpretations of their findings.[15] Given that few studies have examined the role of extra-hepatic hepcidin in regulating systemic or tissue-level iron metabolism, further investigations are warranted to understand mechanisms related to iron metabolism in colonic tumors.[15,37–39]

We observed significantly more cases (30%) had detectable iron accumulation compared to controls (5%). Perls' staining has relatively low sensitivity. We speculate iron accumulation would have been much more prevalent in cases had we used a more sensitive assessment method. Brookes et al. used DAB-enhanced Perls' methodology and detected visible iron accumulation in all (n = 20) of the CRC cases but not controls.[1] To further characterize the cases with iron accumulation, we examined systemic hepcidin and iron status parameters. The iron (+) group had elevated serum hepcidin compared to cases without iron accumulation. However, the iron (+) group was also more iron sufficient and had higher Hb. Given the small sample of cases with iron accumulation, it is hard to make any definitive conclusions. However, our data coupled with the report by Brookes et al. suggests that a subset of persons with CRC have increased expression of systemic hepcidin, which may precipitate greater intestinal iron exposure, and promote tumor iron retention.[1] In a murine model of CRC, animals had increased tumor growth when fed a high iron diet compared to a low iron diet.[5] Also, in persons with ulcerative colitis (UC), who have co-existing systemic and colonic inflammation and hepcidin-mediated dietary iron malabsorption, luminal iron exposure is associated with greater colonic inflammation and mucosal proliferation.[40–42] Therefore, studies examining the effect of dietary iron restriction in persons with CRC should be explored.

The link with hepcidin and excess tumor iron accumulation provides further evidence for its role in cancer. Only one other study has comprehensively explored the hepcidin-FPN axis and its relationship in cancer progression. Pinnix et al. using human breast cancer tissue found that low FPN and high hepcidin expression may enable rapid cell proliferation.[14] Unfortunately, the authors did not measure systemic hepcidin concentrations and future exploration should focus on its role in tumor iron retention in other epithelial cancers.

Our study had strengths and limitations. A significant strength was that we simultaneously measured hepcidin systemically and in colonic mucosa while accounting for several factors that can contribute to hepcidin regulation, including iron status, inflammation and dietary iron intake. Additionally, we included a metabolically healthy control group. However, this study had limitations, including a relatively small sample size. Therefore, when we stratified by presence of iron accumulation in the CRC cases, this may have reduced the statistical power to detect differences. Tissue analysis with adenocarcinoma in cases to healthy mucosa in a control group was not a direct comparison. Future studies should also include non-involved mucosa in cases to accurately describe our analysis among all groups. Due to limited tissue allocated for this study, we were unable to accurately measure protein expression of the colonic iron exporter via FPN. Ferroportin is post-translationally modified by hepcidin, which may have precluded us from observing a significant correlation between colonic FPN and hepcidin. Although we suggest dietary iron absorption may be impaired in persons with CRC, we did not examine dietary iron absorption in this study. Finally, our design included only males, thus generalizability to females may be limited.