Association of serum copeptin and urinary uromodulin with kidney function, blood pressure and albuminuria at 6 weeks post-partum in pre-eclampsia

Abstract

Background: Preeclampsia (PE) is associated with subsequent higher risk of cardiovascular and kidney disease. Serum copeptin, as a proxy for vasopressin, and urinary uromodulin, were associated with PE physiopathology and kidney functional mass respectively. We describe concentrations of these proteins in the post-partum period and characterize their association with persistent hypertension (HTN) or albuminuria.

Methods: Patients with PE and healthy controls with uncomplicated pregnancy were prospectively included at two teaching hospitals in Switzerland. Clinical parameters along with serum copeptin and urinary uromodulin were measured at 6 weeks post-partum. PE patients were further characterized based on presence of HTN (defined as either systolic BP (SBP) ≥140 mmHg or diastolic (BP) ≥90 mmHg) or albuminuria [defined as urinary albumin to creatinine ratio (ACR) ≥3 mg/mmol].

Results: We included 226 patients with 35 controls, 120 (62.8%) PE with persistent HTN/albuminuria and 71 (37.1%) PE without persistent HTN/albuminuria. Median serum copeptin concentration was 4.27 (2.9-6.2) pmol/L without differences between study groups (p > 0.05). Higher copeptin levels were associated with higher SBP in controls (p = 0.039), but not in PE (p > 0.05). Median urinary uromodulin concentration was 17.5 (7.8-28.7) mg/g with lower levels in PE patients as compared to healthy controls (p < 0.001), but comparable levels between PE patients with or without HTN/albuminuria (p > 0.05). Higher uromodulin levels were associated with lower albuminuria in PE as well as control patients (p = 0.040).

Conclusion: Serum copeptin levels at 6 weeks post-partum are similar between PE patients and healthy controls and cannot distinguish between PE with or without residual kidney damage. This would argue against a significant pathophysiological role of the vasopressin pathway in mediating organ damage in the post-partum period. On the opposite, post-partum urinary uromodulin levels are markedly lower in PE patients as compared to healthy controls, potentially reflecting an increased susceptibility to vascular and kidney damage that could associate with adverse long-term cardiovascular and kidney outcomes.

Keywords: albuminuria; copeptin; hypertension; organ damage; post-partum; preeclampsia; proteinuria; uromodulin.

Figure 1

Study flowchart. PE, preeclampsia.

Figure 2

Serum copeptin concentration (pmol/L, log-transformed) at 6 weeks post-partum according to PE status and presence or absence of persistent HTN or albuminuria. Data are presented as median, 25th and 75th percentiles as well as upper and lower adjacent values. Comparison between control participants (N = 35) and PE patients as a whole (N = 191). T-test p value 0.247. Comparison between PE patients with (N = 120) and without (N = 71) HTN/albuminuria. T-test p value 0.614. PE, preeclampsia; HTN, hypertension.

Figure 3

Urinary uromodulin concentration (mg/g, log-transformed) at 6 weeks post-partum according to PE status and presence or absence of persistent HTN or albuminuria. Data are presented as median, 25th and 75th percentiles as well as upper and lower adjacent values. Comparison between control participants (N = 35) and PE patients as a whole (N = 191). T-test p value <0.001. Comparison between PE patients with (N = 120) and without (N = 71) HTN/albuminuria. T-test p value 0.109. PE, preeclampsia; HTN, hypertension.

Figure 4

Association of serum copeptin concentration (pmol/L, log-transformed) at 6 weeks post-partum with outcomes according to PE status. (A) Association with SBP. Interaction between serum copeptin and PE status is significant with p value of 0.017 (contrast of marginal linear predictions) confirming a modification effect. Consequently, sub-group associations for control participants (N = 32) and PE patients (N = 139) are presented separately with associated p values of 0.039 and 0.254 respectively (average marginal effects of multivariate linear regression). (B) Association with eGFR. Interaction between serum copeptin and PE status is significant with p value of 0.045 (contrast of marginal linear predictions) confirming a modification effect. Consequently, sub-group associations for control participants (N = 32) and PE patients (N = 186) are presented separately with associated p values of 0.123 and 0.131 respectively (average marginal effects of multivariate linear regression). PE, preeclampsia; SBP, systolic blood pressure; eGFR; estimated glomerular filtration rate.

Figure 5

Association of urinary uromodulin concentration (mg/g, log-transformed) at 6 weeks post-partum with outcomes. (A) Association with eGFR (ml/min/1.73 m²). N = 218. P value of 0.020 (multivariate linear regression). (B) Association with ACR (mg/mmol creatinine, log-transformed). N = 218. P value of 0.040 (multivariate linear regression). PE, preeclampsia; eGFR, estimated glomerular filtration rate; ACR, urinary albumin to creatinine ratio.