2020 Volume 67 Issue 8 Pages 891-892
【Original Article】 EJ19-0478
【Letter to the Editor】 EJ20-0321
Browse “Advance online publication” versionDear Editor,
In response to the comments by Schönermarck, we would like to briefly discuss here the effects of glucocorticoids on level of cystatin C in serum based on our findings. As pointed out, serum cystatin C levels have been reported to be increased by pharmacological glucocorticoid administration in patients with a variety of disease conditions, including Graves’ ophthalmopathy [1] and heart failure [2]. To the best of our knowledge, our study [3] is the first to examine serum cystatin C in cases of endogenous glucocorticoid overexcretion. The findings showed that serum cystatin C levels were not significantly different between subjects with non-functioning adenoma (0.95 ± 0.01 mg/dL, mean ± standard error) and those with possible autonomous cortisol secretion (1.00 ± 0.02). In contrast, that level in the group with Cushing syndrome was 0.87 ± 0.04, significantly lower as compared to the non-functioning adenoma group. Thus, the effects of a chronic endogenous increase and pharmacological exogenous glucocorticoid dose may be markedly different. Another potential explanation for the opposite effects of these glucocorticoids on cystatin C level may be related to the original underlying disease in the patient. Indeed, glucocorticoid treatment has been shown to not increase but rather decrease serum cystatin C in asthmatic patients [4].
To further examine in detail the impact of endogenous glucocorticoid oversecretion on serum cystatin C, we performed multiple regression analysis to determine independent predictors of serum cystatin C levels using data from all subjects in the current report [3]. As expected, serum creatinine was the strongest predictor for serum cystatin C, with age, gender, and body mass index also independently and strongly associated (Table 1). With these 4 factors, nearly 60% of the cystatin C level results could be explained (R2 0.604, p < 0.01). Additionally, the presence of diabetes was also significantly and independently associated with higher cystatin C level. Notably, after adjustments for these significant confounding factors, serum cortisol level following dexamethasone administration (1 mg) was positively associated with serum cystatin C level. Addition of post-dexamethasone cortisol level in multiple regression analysis resulted in an approximately 4% increase in predictive value (R2 from 0.617 to 0.656, Table 1). These results suggest that the significantly lower cystatin C level observed in the Cushing group might be attributable to lower age and greater percentage of females.
| β | p value | |
|---|---|---|
| Age | 0.276 | <0.01 |
| Gender (male = 1, female = 0) | –0.259 | <0.01 |
| Body mass index | 0.148 | <0.01 |
| Current smoker (yes = 1, no = 0) | 0.070 | 0.07 |
| Cardiovascular disease (yes = 1, no = 0) | 0.019 | 0.63 |
| Hypertension (yes = 1, no = 0) | –0.039 | 0.34 |
| Dyslipidemia (yes = 1, no = 0) | 0.012 | 0.77 |
| Diabetes (yes = 1, no = 0) | 0.099 | 0.02 |
| Free T4 | 0.007 | 0.85 |
| Creatinine | 0.805 | <0.01 |
| ln cortisol after 1 mg DEX | 0.086 | 0.03 |
| R2 | 0.656 | <0.01 |
β, standard regression coefficient
Without a doubt, renal function in patients with hypercortisolism should be carefully interpreted, as eGFRcre with the use of serum creatinine can lead to an overestimation of true GFR. Conversely, as suggested by Schönermarck and shown by our additional analysis, eGFRcys using cystatin C might underestimate this parameter. However, as seen in the Table 1, the contribution of hypercortisolemia to serum cystatin C levels is not high. Precise determination of GFR based on inulin, DTPA, or iohexol clearance is not easily performed in clinical settings, thus estimation of GFR using cystatin C serum level may provide a relevant approach to check renal function.
