Is There a Role for β-Blockade in Septic Shock?

Abstract

Single-cell and multicell organisms have developed elaborate networks tominimize injury, repair damage, and fendoff invasion by other organisms1with the goal ofmaximizing the probability of surviving overwhelming stress. These networks in higher organismswere described byWalter Cannon2 as the acute stress response. The acute stress response includes centrallymediatedsympatheticneural andhumoral activation, increased vascular smooth muscle tone, catecholamine and cortisol release into the bloodstream, minimized pain perception, altered intercellular and intracellular signalingand intermediarymetabolism,andaproinflammatory,prothrombotic intravascular state. The clinical and physiologic manifestations of these responses include increased cardiac inotropy, peripheral vasoconstriction (although some vascular beds may be vasodilated), and tachycardia. Such manifestations, although intended to be helpful in the short term,may become detrimental if they persist. This detrimental reaction is well described for chronic conditions like low-output heart failure. In that state, the body responds by assuming that the low output is due to hypovolemia. Thus, the physiologic response is fluidretentionandincreasedvasomotor tone.3Hence, there is therapeutic benefit from afterload-reduction agents, β-adrenergic blockers, and diuretics.4 Long-term β-blockade of patientswith coronary artery disease improved survival in patients at risk of sudden death, but only if the heart ratewas controlled,5 presumably because that level of β-blockade has a measurable physiologic effect. The improved outcomes among thesepatients resulted fromamarkedreduction insudden death due to a slower progression of coronary artery disease and less tendency for malignant arrhythmias.6 Based on this logic, Morelli et al7 hypothesized that, because patients in septic shock had many of the manifestations of a hyperadrenergic response, pharmacologic reduction in β-adrenergic tone may be beneficial if patients were appropriately resuscitated. In this issue of JAMA, these investigators report the results of a single-center randomized clinical trial designed to measure the effects of the short-acting β-blocker, esmolol, in septic shock. The authors randomized 154 patients to receive esmolol (n = 77) or usual care (n = 77), and using sinoatrial adrenergic sensitivity as a barometer of adequate β-blockage,5 titrated esmolol to maintain the heart rate between 80/min and 94/min. Because patientswith sepsis have awide range of sympathetic activation and responsiveness, giving a fixed dose of a β-blocker would probably be less effective and potentially harmful if given to all patients. Furthermore, because adrenergic stress persists as long as the external stress (eg, infection or injury), treatment was continued for the entire intensive careunit stay. The choiceofusing a short-actingβ-blocker inpatients requiringvasopressor therapytomaintainorganperfusionpressurewas reasonablebecauseβ-blockadeallowsunrestricted α-adrenergic stimulation, minimizing the potential for worsening hypotension, while allowing its dosage to be accurately titrated up or down to effect. The intervention was associated with a number of cardiac effects. First, heart rate was lowered successfully to the target range without unwanted hypotension. Second, stroke work index and left ventricular stroke work improved, presumably because of improved diastolic filling with lower heart rate. Third, overall, because the heart became more efficient (better stroke work), cardiac index and systemic oxygen delivery only decreased minimally despite the relatively large decrease in heart rate. Beyond the primary cardiac effects, there were notable improvements in the norepinephrine and fluid requirements and in arterial lactate levels in the esmolol group compared with the control group. Perhaps most surprisingly, the 28-day mortality was considerably lower in the esmolol group than in the control group (49.4% vs 80.5%; P < .001). These data are consistent with selective blockage of β-adrenergic hyperactivity causing improved myocardial performance and decreased metabolic demand without compromising peripheral vascular function. Indeed, the PaO2 to FIO2 ratio also decreased, presumably due to decreased β-adrenergic pulmonary vasodilation causing better V/Q matching. Although these findings are potentially important, caution needs to be stressed before applying these results to all patients in septic shock. The reasons for this caution involve the limitations of this study and limitations in the current understanding of how β-blocker therapy can cause such effects. The study was a single-center open-label trial. Clearly, it would be difficult to mask heart rate titration because placebowould have little effect on heart rate other than to cause a large degree of fluid resuscitation. A largemulticenter clinical trial iswarrantedtoconfirmthesepreliminary findings.Second,more thanhalf of the septic shock candidates for this trial were excludedbecause theydidnothave tachycardia. It is unclear whether tachycardia is the right method to identify patients in septic shock who might benefit from β-blockers. β-Blockademaybebeneficial inpatientswith lowerheart rate, and the effects may be neither due to nor proportional to the degree of sinoatrial node blockade. Third, because outpatient use of β-blockers is common, it is unknown how such patients,whowere excluded from the trial,might have fared. Related article page 1683 Opinion