Is It Time to Reconsider the Concept of "Salvage Therapy" in Refractory Shock?

Distributive shock is the most common form of circulatory failure and is characterized by a systemic arterial tone insufficient to adequately deliver oxygenated blood and clear metabolic waste products (1). This syndrome may lead to an impairment in effective oxygen extraction at the level of the tissues, inadequate cellular utilization of oxygen, rapid conversion to anaerobic respiration, and progressive multiple organ failure. As such, distributive shock must be considered a medical emergency that warrants prompt diagnosis and reestablishment of perfusion. In patients requiring restoration of systemic arterial tone despite adequate fluid resuscitation, the Surviving Sepsis Campaign recommends vasopressor support with norepinephrine as the first-line choice (2). For patients that fail to achieve desired hemodynamic targets, recommendations regarding timing and selection of secondary agents are vague and have been contested widely in the literature, leaving clinicians with varying opinions (3). In this issue of Critical Care Medicine, Smith et al (4) report a retrospective observational analysis of persons with refractory distributive shock who received angiotensin II in a so-called “salvage” fashion, which they compared with a matched angiotensin II-unexposed group. Importantly, the mean baseline vasopressor dose in norepinephrine equivalents (NEE) was 0.62 mcg/kg/min, which was the only criterion used to identify the matched group, leaving multiple differences between the treatment groups, such that angiotensin II recipients appeared to be sicker than the control group with higher Sequential Organ Failure Assessment scores (12 vs 10; p < 0.001), higher lactate concentrations (8.5 vs 5.9 mmol/L; p < 0.001), lower mean arterial pressures (69 vs 73 mm Hg; p = 0.004), and greater frequency of various comorbid conditions, suggesting they may be, at baseline, more likely to experience a suboptimal outcome than the control group. Additionally, those patients in the angiotensin II group had already progressed to more organ failures than the control group at the time of angiotensin II receipt with nearly twice as many receiving kidney replacement therapy (33% vs 17%; p < 0.001), and more were mechanically ventilated (92% vs 71%; p < 0.001) compared with unexposed controls. Despite these imbalances, the authors found similar 30-day (60% vs 56%; p = 0.292) and 90-day (65% vs 63%; p = 0.440) mortality rates between angiotensin II recipients and the control group, respectively. In their multivariable model, receipt of angiotensin II did not alter odds of 30-day mortality (odds ratio [OR], 0.95; 95% CI, 0.61–1.48), nor the odds of new kidney replacement therapy or mechanical ventilation, or the propensity to experience thrombotic complications. The report by Smith et al (4) suggests that administering angiotensin II in patients with distributive shock that is refractory to high-dose vasopressors may not change outcomes, but leaves critically unanswered questions including: 1) is refractory shock the optimal setting to apply alternative vasopressors and 2) should we reconsider the concept of “salvage therapy” in shock? Refractory shock appears to be a distinct subset of distributive shock that presents with additional molecular mechanisms that lead to impaired responsiveness to vasopressors and promote complete cardiovascular collapse, such as dysregulated mitochondrial respiration, membrane hyperpolarization, vascular hyperreactivity, and compromised microcirculatory flow (5). Although the diagnostic criteria, risk factors, and frequency of refractory shock are not universally accepted, most clinicians would consider the requirement of high doses of vasopressor agents to be an important part of the diagnosis (6). Indeed, shock refractory to high doses of vasopressors is associated with extraordinarily high mortality rates, ranging from 60% to greater than 90% (7). These high mortality rates are consistent with those observed by Smith et al (4) in their sample of patients with a mean baseline NEE vasopressor dose of approximately 0.62 mcg/kg/min. This poses additional questions, such as, in the hostile chemical environment of refractory shock requiring extremely high doses of vasopressors, has the opportunity to meaningfully intervene passed, such that the concept of applying a “salvage therapy” could be considered inherently flawed? Are the opportunities for reversing organ hypoxia and restoring vascular integrity the same in refractory shock as they are in shock that has not progressed to this extent? Is it plausible for discrete interventions to drastically alter the course of refractory shock? As the vasopressor requirements rise in refractory shock, it seems apparent that the likelihood of a beneficial outcome to secondary and tertiary vasopressors declines. There is some suggestion that delayed addition of vasopressin to catecholamine vasopressors is associated with worse outcomes: the odds of death rise by over 20% (OR, 1.21; 95% CI, 1.09–1.34) for every 10 mcg/min increase in vasopressor dose at the time vasopressin is added to catecholamine vasopressors (8). Furthermore, patients randomized to vasopressin when the norepinephrine dose was low (< 15 mcg/min), had a lower risk of death (risk ratio, 0.78; 95% CI, 0.61–0.99) than those who received continued escalation of norepinephrine, an effect not seen at higher vasopressor doses (9). Similarly, when the vasopressor dose at the time of randomization to angiotensin II was low (≤ 0.25 mcg/kg/min), the hazard of death was nearly half that of continued escalation of other vasopressors (hazard ratio, 0.51; 95% CI, 0.27–0.95) (10). Importantly, a similar relationship was not seen when the baseline vasopressor dose was high (> 0.25 mcg/kg/min). Taken altogether, these data suggest that contrary to the environment in which Smith et al (4) assessed angiotensin II, potential benefits of the addition of alternative vasopressors, including angiotensin II, to catecholamine vasopressors such as norepinephrine, may be realized in settings where progression to refractory shock has not yet occurred. The stepwise approach to vasopressor management in distributive shock consists of application of a first-line vasopressor, with subsequent dose up-titration in an effort to achieve hemodynamic and perfusion goals (Fig. 1). While many will respond to this first-line vasopressor, those not achieving goals—at some specified timepoint and vasopressor dose—a secondary vasopressor agent is added and adjunct treatments (e.g., corticosteroids) are considered (2). One of the most important limitations of such a strategy is the time lost in providing satisfactory perfusion when allowing each vasopressor an opportunity to restore hemodynamics. Therefore, once the time has come to add a secondary, tertiary, quaternary agent, there has been greater exposure to hypotension, subsatisfactory perfusion, tissue hypoxia, and organ failure. Indeed, as the time spent under mean arterial pressure less than 65 mm Hg increases, the odds of acute kidney injury (OR, 1.07; 95% CI, 1.047–1.095), myocardial injury (OR, 1.05; 95% CI, 1.004–1.087), and death (OR, 1.11; 95% CI, 1.078–1.151) increase (11). Similarly, as the vasopressor load increases for every 10 mcg/min during early resuscitation, the odds of death increase dramatically (OR, 1.33; 95% CI, 1.16–1.53) (12). Progressive decline in vasopressor responsiveness, a characteristic of refractory shock, must also be considered a flaw of the stepwise vasopressor approach, as for every 1 mmol/L increase in lactate concentrations, the odds of a favorable hemodynamic response to the addition of vasopressin (OR, 0.93; 95% CI, 0.89–0.97) (13) and angiotensin II (OR, 0.89; 95% CI, 0.83–0.95) (14) in vasopressor-refractory shock decline.Figure 1.: Simplified conceptual framework of a stepwise approach and an early multimodal approach to vasopressor application in the resuscitation of distributive shock. Created with BioRender.Taken altogether, it may be reasonable to rationalize an early multimodal vasopressor strategy (Fig. 1), where vasopressors bearing differing mechanisms, and adjunctive treatments, are instituted in concert from the onset. The goals of deploying such a strategy are to maximize the number of treatment responders and minimize the time spent under subsatisfactory perfusion targets. Some key factors we must understand to operationalize such a strategy include how to enrich responders to specific vasopressor agents with biologic endotypes, better predict who will otherwise progress to refractory shock, and ultimately shift to perfusion-focused resuscitation. As we continue to investigate the optimal settings and approaches to vasopressor treatment in shock, the findings of Smith et al (4) should remind us that we must critically reconsider the potentially flawed concept of “salvage therapy” and seek every opportunity possible to prevent progression to refractory shock.