Does Early Vasopressin in Septic Shock Improve Outcomes? An Important Piece to This Emerging Puzzle Has Arrived

Abstract

In this month’s Annals of Pharmacotherapy, the largest observational study assessing the early versus later use of vasopressin has been published. When this new study is combined with the other available observational studies, there are 2 important outcomes to focus on. When all the observational studies are pooled together, no reduction in new onset arrhythmias is seen (odds ratio [OR] = 0.91, 95% confidence interval [CI] = 0.41-1.95) with early versus late vasopressin use while the reduction in renal replacement therapy just missed statistical significance (OR = 0.56, 95% CI = 0.32-1.00). Early vasopressin likely does not reduce new onset arrhythmias versus later use but might reduce the need for renal replacement therapy.

Background

As vasopressin stimulates V1a receptors to induce vasoconstriction, it can be used as an alternative when alpha-1 adrenoceptor stimulators like norepinephrine fail or as an adjunctive strategy to enhance the level of vasconstriction. ¹ However, the extent to which early use of vasopressin can enhance outcomes versus waiting until patients fail norepinephrine therapy is not as clear. Conducting clinical trials during critical illness or emergency situations is very difficult, so an amalgam of clinical trials and observational studies are commonly used when informing clinical practice. ²

In this month’s Annals of Pharmacotherapy, Betthauser and colleagues ³ present the results of an article titled “Effect of early administration of vasopressin on new onset arrhythmia development in patients with septic shock: a retrospective, observational cohort study.” While this is not the first observational study to assess the impact of early versus later initiation of vasopressin on clinical outcomes in patients with septic shock, it is by far the largest and provides important insights to critical care clinicians. ⁴ In previous observational studies, the patient populations ranged from 71 to 130 patients while this new study has 436 patients. ⁴ This is a similar sample size (n = 409) to the only published clinical trial by Gordon and colleagues.^4,5

Placing the Findings in Context

In 2021, a systematic review conducted by Huang and colleagues assessed the clinical trial and observational studies that evaluated early versus later vasopressin therapy in septic shock.^4-9 While their search strategy was strong and they found the available literature, their meta-analysis strategy had 2 main weaknesses. First, they statistically pooled the single clinical trial together with the observational studies which is not recommended given the inherent methodological differences.^10,11 They also chose a fixed effects model for their meta-analysis which is not recommended given the underlying methodological and clinical heterogeneity or moderate to high statistical heterogeneity.^10,11 As such, in this section, I used the study-level data from the observational studies included in the Huang and colleagues ⁴ systematic review and meta-analyze them using a more conservative random-effects model including the data from this new study by Betthauser and colleagues ³ to show how this study impacts clinical understanding of early versus later vasopressin therapy.

The primary endpoint in the study by Betthauser and colleagues was new onset arrhythmias. While Reardon and colleagues found a significant reduction in new onset arrhythmias (odds ratio [OR] = 0.33, 95% confidence interval [CI] = 0.13-0.88), Hammond and colleagues in 2019 (OR = 0.73, 95% CI = 0.16-3.47) and Gordon and colleagues (OR = 0.39, 95% CI = 0.08-2.04) had the same direction of effect but insignificant findings.^4-7 However, Hammond and colleagues ⁸ in 2018 (OR = 2.17, 95% CI = 0.50-9.35) and this new study from Betthauser and colleagues ³ (OR = 1.34, 95% CI = 0.67-2.69) found an opposite direction of effect with early versus later administration of vasopressin with insignificant differences between the groups. ⁴ In the previous systematic review by Huang and colleagues, ⁴ the investigators meta-analyzed the observational trials with the clinical trial by Gordon and colleagues and found a strong trend toward a reduction in the odds of developing new onset arrhythmias with early versus later vasopressin therapy (OR = 0.59, 95% CI = 0.31-1.10). I conducted a random-effects meta-analysis confined to just the previous observational studies as well as the new results from Betthauser and colleagues, and found no reduction in new onset arrhythmias (OR = 0.91, 95% CI = 0.41-1.95) with moderate heterogeneity (I ² = 51.9%). As such, this new study by Betthauser and colleagues greatly dampens enthusiasm that early vasopressin therapy will reduce new onset arrhythmias. The lack of difference between the initial use of norepinephrine or vasopressin on new onset arrhythmias may be because neither drug alters chronotropy or coronary blood flow to an appreciable extent. ¹

A secondary endpoint to focus on from this new observational study was the impact of early versus later use of vasopressin on the need for renal replacement therapy. ³ Daley and colleagues ⁹ (OR = 0.21, 95% CI = 0.21-0.88) found significant reductions in the need for renal replacement therapy with early vasopressin while Hammond and colleagues ⁸ in 2018 (OR = 1.28, 95% CI = 0.48-3.42) did not. ⁴ Gordon and colleagues ⁵ also found a significant reduction in need for renal replacement therapy (OR = 0.62, 95% CI = 0.41-0.95). ⁴ In the previous fixed effects meta-analysis, Huang and colleagues ⁴ found a significant reduction in the need for renal replacement therapy (OR = 0.63, 95% CI = 0.44-0.88) but with the exclusion of the clinical trial by Gordon and colleagues, I calculated no significant reduction (OR = 0.63, 95% CI = 0.35-1.12) using a fixed effects model. In this new study, Betthauser and colleagues ³ found a significant (OR = 0.45, 95% CI = 0.27-0.77) reduction in the need for renal replacement therapy with early versus late vasopressin therapy. When meta-analyzing the 2 previous observational studies together with those of Betthauser and colleagues using a more appropriate random-effects model, I found a strong trend toward a reduction in the odds of needing renal replacement therapy (OR = 0.56, 95% CI = 0.32-1.00). While a fixed effects model would have returned a significant reduction in the odds for renal replacement therapy, the single site nature of these trials and some of the nuances in the methods make this test less appropriate, especially because there was moderate statistical heterogeneity (I ² = 47.5%).^10,11 As such, early vasopressin therapy is a promising modality to reduce the need for renal replacement therapy in patients with septic shock, although further investigation is needed.

Betthauser and colleagues focused on the significant 1-day reduction in intensive care unit (ICU) length of stay with early versus later vasopressin therapy in their study (P = 0.02). However, the other observational studies assessing this endpoint were mixed with Daley and colleagues plus Hammond and colleagues in 2019 trending toward reductions with early vasopressin therapy while Reardon and colleagues plus Hammond and colleagues in 2018 were trending toward increases in ICU length of stay. ⁴ ,^6-9 In addition, the clinical trial by Gordon and colleagues found nonsignificant increases in ICU length of stay (mean difference = 0.70, 95% CI = −0.69 to 2.29 days) in their trial with early vasopressin use. ⁵ Future studies could better clarify the impact of early versus late vasopressin therapy on this endpoint.

Conclusions

When viewed considering the body of literature, this new observational study helps to clarify the potential benefits of early versus later vasopressin therapy in patients with septic shock. The results of this new study cast doubt on the ability of early vasopressin use to reduce new onset arrhythmias versus later use. While this study found reductions in ICU length of stay, the literature is mixed, and these benefits may not materialize with further research. This new study does strengthen the possibility that early vasopressin therapy will reduce the need for renal replacement therapy, but more research is still needed. As such, early vasopressin therapy is a viable alternative to later vasopressin use but whether it truly offers clinical advantages remains unknown.