C-TERMINAL PORTION OF PRO-ARGININE-VASOPRESSIN (CT-PRO-AVP) AS A PREDICTIVE BIOMARKER IN SEPSIS

The vasopressin axis is activated by hypotension (and increased osmolality, but hypotension takes precedence). Ironically, there is a deficiency of vasopressin in septic shock compared with cardiogenic shock. This has led to a therapeutic strategy to evaluate vasopressin versus norepinephrine in many small and one large multicenter randomized controlled trial of in septic shock (Vasopressin and Septic Shock Trial [VASST] (1)). Briefly, vasopressin did not change mortality, but may have decreased mortality in patients who had less severe shock (1). The vasopressin axis is also a viable candidate diagnostic strategy that could lead to a clinically viable biomarker kit. One diagnostic strategy that investigators have pursued is to test the hypothesis that biomarkers of the vasopressin axis predict response to and need for vasopressors—both vasopressin and adrenergic vasopressors—in septic shock. Vasopressin-related biomarkers have included vasopressin (1) (but that requires a radioimmune assay) and C-terminal portion of pro-arginine-vasopressin (2) (CT-pro-AVP; a surrogate biomarker of vasopressin level) also known as copeptin (3), a stable peptide that is cleaved from pre-pro-vasopressin and so is used as a surrogate biomarker of vasopressin level (Table 1). Semantics require clarification: diagnostic biomarkers diagnose a condition (e.g. troponin and acute myocardial infarction [AMI]), prognostic biomarkers determine the prognosis (e.g. protein C levels and organ dysfunction (4), and predictive biomarkers (also called a companion diagnostic; e.g. HER2neu and use of Herceptin [Genentech, San Francisco, Calif] in breast cancer) identify patients who have an improved response to a drug (or device). In this issue of Shock, Laribi et al. (5) aimed to stratify patients with septic shock for risk of vasopressor dependence and risk of death by measuring CT-pro-AVP within 24 h of onset of sepsis. This is a predictive biomarker in the sense that the test result on day 1 to 2 identifies which patients will respond quickly and be alive and free of vasopressors on day 7. Their rationale and the selection of CT-pro-AVP are appropriate because in essence, the study proposes a novel biomarker predictive of patients who are alive and free of vasopressors on day 7. Their single-center cohort (n = 113,102 with sepsis or shock and 11 nonseptic controls) was divided into two subgroups according to composite cardiovascular outcomes: first, patients who were either on vasopressors or dead at day 7 and second, patients who were alive and free of vasopressors at day 7. Plasma CT-pro-AVP level was not predictive of need for vasopressors; nonetheless, patients with shock had significantly higher plasma CT-pro-AVP levels than patients without shock and nonsurvivors had significantly higher CT-pro-AVP levels than survivors. The presentation is made more complete with a receiver operating characteristic curve with area under the curve (AUC) for comparisons of CT-pro-AVP versus need for vasopressors on day 7 or death by day 7. The AUC for CT-pro-AVP was 0.681 on day 1 to 2 and 0.736 on day 3 to 4. Then the authors compared the AUC for CT-pro-AVP to a popular sepsis biomarker (procalcitonin) and two widely used clinical scoring systems (Acute Physiology and Chronic Health Evaluation II [APACHE II] and Sequential Organ Failure Assessment [SOFA]). CT-pro-AVP AUC was better than the AUC of the procalcitonin (day 1–2: 0.608). Perhaps that is not surprising because procalcitonin is a better biomarker for antibiotic guidance in pneumonia, whereas a vasopressin axis predictive biomarker would be expected to be better for a composite cardiovascular outcome in septic shock. The AUC for CT-pro-AVP was not as high as baseline APACHE II (AUC = 0.747) and SOFA (AUC = 0.775), composite scores with many chronic health status (APACHE II) and acute physiology variables (APACHE II and SOFA). The authors conclude that plasma CT-pro-AVP level does not predict vasopressor need but confirms prior studies of prognostication for death, i.e. a prognostic biomarker. Strengths of the study include clear definition of outcomes – the clinical endpoint (alive and off vasopressors by day 7) is important and highly patient-centered, a novel hypothesis, and interesting results (that I interpret more positively than do the authors). I disagree that all the results regarding association of plasma CT-pro-AVP level with alive and free of vasopressors on day 7 are negative. The CT-pro-AVP (54 pmol/L) on day 3 to 4 in patients free of vasopressors on day 7 was significantly lower than in patients who were dead or on vasopressors on day 7 (CT-pro-AVP = 91 pmol/L). Furthermore, data of Laribi et al. (5) (Figure 1 and Table 1) clearly show that CT-pro-AVP discriminates early between no sepsis, severe sepsis, and septic shock and the differences are significant. Surely, the finding that there is discrimination on day 1 is important. That is not a negative finding.Table 1: Vasopressin-axis biomarkers in sepsis and other conditionsIn essence the authors propose CT-pro-AVP as a “predictive” biomarker for weaning successfully from vasopressor(s); however, I am not surprised with CT-pro-AVP AUCs of 0.681 to 0.736 for weaning successfully from vasopressor(s) on day 7 for several reasons related to pathophysiology of recovery of the vasopressin axis in septic shock. The ability of a rising CT-pro-AVP to predict successful weaning from vasopressors would be modulated in several ways – first, by recovery (or not) of the well-known downregulation of alpha-1 adrenergic receptors and AVP1a receptors (that drive vasopressin-induced vasoconstriction) in sepsis; second, by renal dysfunction because renal dysfunction decreases clearance of CT-pro-AVP (thus “confusing interpretation of CT-pro-AVP level as a sign of vasopressin axis recovery); and third, by the rate of recovery of sepsis-induced suppression of vasopressin synthesis and release centrally. So perhaps, it would be unrealistic to expect the increase of CT-pro-AVP to be the sole or even the major predictor of weaning successfully from vasopressor(s) on day 7. There are several limitations of this study. First, there were no a priori sample size and power calculations arguing that this was an exploratory study, but this is a gap. Second, this single-center small discovery cohort was not validated in a validation cohort (yet). Third, the comparison with other candidate biomarkers is limited, including only procalcitonin. Also, most patients were older (55 or older), quite different than a younger cohort with severe sepsis. More problematic, there was no multivariate analysis of a mix of clinical and laboratory biomarkers including CT-pro-AVP concentrations to test whether such multivariate predictions are better than clinical predictors alone. Vasopressin and CT-pro-AVP concentrations are prognostic and predictive in a wide range of other acute conditions that are modulated by the vasopressin axis including cardiac arrest (6), stroke (7), and diabetes insipidus after pituitary surgery (8) (Table 1). In summary, I find that this is an intriguing preliminary finding that requires further validation in larger, multicenter cohorts. CT-pro-AVP may be a useful predictive biomarker 1 day. Only subsequent studies will determine whether the authors’ “negative” or my “positive” interpretations are the closest to truth. I do still believe that biomarker(s) of the vasopressin axis could become predictive of need for and response to vasopressin and to other vasopressors.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.