In a sub-analysis of the Belgian STEMI registry, including 20,3% of Belgian STEMI patients undergoing pPCI for STEMI, we found that renal dysfunction at the time of hospital admission (eGFR < 60 mL/min per 1.73 m2 or CKD class 3 or higher), assessed by the CKD-EPI formula, was a common finding and that more women (42.3%) than men (25.3%) suffered from this condition. As expected, a CKD class 3 or higher on admission was associated with in-hospital mortality, and this independently of the TIMI risk score. Although there was a trend towards higher mortality for women with renal dysfunction compared to men with this condition, we could not demonstrate a gender difference in the impact of renal dysfunction on in-hospital mortality.
Despite the fact that male STEMI patients had a higher serum creatinine concentration at admission, the prevalence of renal dysfunction defined by an eGFR <60 mL/min per 1.73 m2 and assessed with the CKD-EPI formula was almost doubled in women as compared to men, even after correction for observed differences in age and risk profile between men and women. The higher serum creatinine concentrations in men can easily be explained by the larger muscle mass, resulting in greater creatinine generation, and a higher serum creatinine concentration for a given GFR. This illustrates that serum creatinine values should not be used to evaluate gender differences in the impact of renal function on outcome. It is not completely clear why the incidence of renal dysfunction is higher in women, we cannot exclude that there are unknown confounders that may explain this difference.
Many authors have demonstrated that women with STEMI have a higher risk of in-hospital mortality and some, but not all, could explain this higher mortality based on age and the presence of more comorbidities, especially hypertension and diabetes [1–5]. Lawesson et al. recently demonstrated in a small single center study, including 274 STEMI patients undergoing pPCI, that female gender was a strong and independent predictor of renal dysfunction and that renal dysfunction had a possibly higher impact on 1-year mortality in women (p for interaction = 0.08) . We confirmed that renal dysfunction was more prevalent in pPCI treated women but we did not find a gender-impact on in-hospital mortality. Given the fact that the prevalence of renal dysfunction was independently related to female gender and that there was no difference in mortality between women and men with preserved renal function we speculate that renal dysfunction could be an important reason why women with STEMI die more than men.
The presented data do not reveal why renal dysfunction was associated with worse outcome. There are numerous data that demonstrate that chronic kidney disease serves as an important modifier for outcome. Renal dysfunction may serve as a surrogate marker for general health and for unknown risk factors that may explain the worse outcome. Also renal dysfunction may be associated with complications, of which development of (contrast induced) acute kidney injury (AKI) and bleeding are the most likely to occur . By implementing routine calculation of eGFR, based on the CKD-EPI formula, high risk patients can be identified, and strategies for prevention of contrast induced AKI (bicarbonate administration, optimization of hemodynamic status and discontinuation of nephrotoxic drugs)  and other complications (e.g. bleeding) can be initiated. More over, evidence based therapies such as β-blockers, ACE-inhibitors and statins should not be withheld in this subgroup .
Today it is not clear whether the prognostic performance of the currently most used simple and bedside risk scores could improve by adding eGFR to the model. Kidney function, represented by various cut-offs for serum creatinine but not by eGFR, is already incorporated in the GRACE risk score . Kidney function is not incorporated at all in the TIMI risk score for STEMI and this might be one of the reasons why the GRACE risk score performed better in STEMI patients in a recent meta-analysis of 15 derivation studies and 17 validation studies . Given our finding that serum creatinine underestimates renal dysfunction, especially in women, future risk stratification should preferably use the eGFR based kidney function estimates such as the CKD-EPI equation.
Strenghts and limitations
This study is unique as it represents the first dataset that links gender, renal dysfunction, assessed by the CKD-EPI equation, and outcomes in a subgroup of PCI-treated STEMI patients included in the Belgian STEMI registry. As 71.2% of this Belgian STEMI cohort presented with an eGFR > 60 mL/min/1.73 m2, the use of the CKD-EPI equation was appropriate.
The study has some limitations. First, we only studied patients who underwent pPCI in a subgroup of 8 tertiary care centers that participated in the Belgian STEMI registry; since it has been demonstrated that patients with renal failure have less access to invasive therapy, a selection bias is not excluded . Second, it is uncertain whether renal dysfunction at the time of hospital admission represents a steady state condition of chronic kidney disease or whether there was also a component of AKI. Third, we only collected data on in-hospital mortality rates; hence, it remains unclear whether our results can be extrapolated to long-term outcomes. Finally, it is not clear how missing values could have influenced our results, however we found no differences in proportions of women and baseline characteristics, except for a higher incidence of CPR in the group with missing values. This and the higher in-hospital mortality in this group suggest that these patients were more severely ill and probably also had renal dysfunction at the time of hospital admission, which would have reinforced our findings.