Hello, my name is Matthew Lissauer. I am currently the acute care surgery chief at Hartford HealthCare and Hartford Hospital. You may notice this is slightly different than the program, as between accepting the speaking invitation and creating this talk, I did change jobs. As part of the session on Lions, Tigers, and Bears, the World Wide Web of Resuscitation Adjuncts, I'm more than happy to give my talk, Vasopressors and Hemorrhagic Shock. Thank you to the Congress Planning Committee and our moderator, Gail Gessen, for the opportunity to speak. I have no disclosures related to the content of this presentation. When I was a resident and subsequently a trauma fellow, had I told my attending that I put a new trauma patient on pressors, I might not have graduated. The assumption was we either missed our resuscitation targets or missed an injury and the patient was still bleeding. It took quite a lot to convince them that the patient was indeed in some form of vasodilatory or other shock and that pressors were needed. Trauma patients on pressors were sort of treated like a never event, similar to the ones you see here. In fact, as our resuscitation has improved, the use of pressors has become less necessary in trauma. This is also in parallel with our more rapid ability to affect hemorrhage control. But moving through our resuscitation, when we were using a lot of crystalloid, those patients got pretty sick. We then realized that early blood products improved outcomes. Patients got a little bit better. We then learned that balanced 1 to 1 to 1 resuscitation helped even more. We now use thromboelastographic guided coagulopathic resuscitation, fresh or cold stored whole blood, and we're using more tranexemic acid. All of this helps to reduce the need for pressors and affect better resuscitation. And so even in the current era, our goal is volume restoration concurrent with hemorrhage control, and we save pressor use for those with hemodynamic instability refractory to blood or volume resuscitation. But the story gets a little more complicated as we shift towards using pressors only for patients refractory to resuscitation and begin to think about using pressors as part of an adjunct to our resuscitation. This is a graph looking at the mean arterial pressure of pigs after shedding blood and then returning that blood, and it doesn't make too much intuitive sense, so I'm going to walk you through it a little bit. While the x-axis is the amount of blood lost, it's not a linear timeline. On the lower curve, time moves left to right, but on the upper line, it moves right to left. So what these authors did was bleed 5 cc's per kilogram from a pig every 5 minutes. That lower curve is the pressure response to that bleeding. After 35 minutes, they began to give blood back, 5 cc's per kilogram at a time every 5 minutes. So for each time point, it represents the amount of blood loss, with the lower curve representing the relationship as blood was withdrawn, and the upper curve representing that same relationship for the same amount of blood loss after some blood was given back. You can see that after withdrawing blood, then replacing the same amount that was withdrawn, the animal's mean arterial pressure is higher, despite the volume theoretically being exactly the same. This demonstrates some of the physiologic compensatory mechanisms that happen early in hemorrhagic shock, and a bit of foreshadowing, they involve endogenous pressors. And we've known this for some time. We've known for a long time that hemogen resuscitation induced alterations in cytokine expression. In this paper, they looked at the development of acute lung injury. So we know it leads to adverse outcomes. Here's just a pictorial of those changes in IL-1, TNF-alpha, IF-gamma, TGF-beta. And since we know there's a whole cytokine and physiologic response to hemorrhagic shock, it's important to ask the question, what is the endogenous pressor response to this hemorrhage? Well, we know that vasopressin levels rise significantly after injury, as much as tenfold early after injury. Epinephrine levels also rise significantly early after injury, and that vasopressin and epinephrine, the amount of rise, is correlated to the severity of the injury. And here is some of that change quantified. You can see in the top left and bottom right squares that vasopressin and epinephrine levels rise pretty early, but that angiotensin and cortisol levels, the top right and bottom left boxes, have not risen yet. It is important to note that those changes in arginine vasopressin activity and plasma catecholamines are seen very early, even before there's a chance to see changes in cytokine expression. This group did an interesting experiment to look at the physiologic response to hypovolemia. They took healthy volunteers, put a neoprene suit over their lower abdomen, essentially put them in an iron lung, and generated negative pressure in the lower extremities. This caused blood to pool in the lower extremities, and the patients became hypovolemic. Well, as expected, they get a reduced cardiac filling, reduced stroke volume, reduced cardiac output, their blood pressure drops, and then they have baroreceptor stimulation, which leads to a physiologic response, including sympathetic activation, which leads to arginine vasopressin activity and an increase in plasma catechols, as well as a reduction in cardiac vagal tone. This compensatory response leads to peripheral vasoconstriction and elevated heart rate, better arterial blood pressure, and cerebral blood flow. So pressors early seem to be beneficial in the response to hypovolemia. As you move later on through the disease, however, you get metabolic disturbances and compensatory exhaustion. This is when patients will demonstrate peripheral vasodilatation, bradycardia. In these patients, they get presyncope. In the trauma bay, these are the patients who needed pressors. We've done their resuscitation, but they are beyond the point of compensatory exhaustion. Beyond baroreceptor changes in sympathetic and parasympathetic activity, we also see increases in nitric oxide synthesis. This production is significantly increased in decompensated hemorrhagic shock, and we know that catecholamine resistance can be ameliorated with nitric oxide inhibitors. We also see a deficiency in vasopressin in hemorrhagic shock. Now, wait a minute. I know what you're thinking. Didn't I just say a few slides ago that vasopressin actually increases in hemorrhagic shock? Well, the truth is a little bit nuanced. Vasopressin levels do rise early in injury, but they drop off relatively quickly, too. Because of this, fairly soon after the development of hemorrhagic shock, people tend to be in a vasopressin-deficient state. We know that low vasopressin levels are responsible for hypotension in hemorrhage in animal models as well, and we've known, again, for 40 years that just a small amount of hemorrhage and hypovolemia can lead to profound hypotension in brattle-borel rats, which cannot produce vasopressin. So since we know that the early response to hemorrhagic shock is quite complex, but includes an early rise in both vasopressin and catecholamines, followed by a rapid drop in vasopressin, can we utilize either of these as an adjunct to resuscitation to improve outcomes? Well, first, let's look at some animal models. In this porcine study, they subjected a pig to 40% blood loss. Then the pigs either got vasopressin, epinephrine, or a placebo. After this, they were resuscitated with fluids. In the vasopressin group, 7 of the swine survived. Unfortunately for the epinephrine and the placebo group, none of them lived. Here is some of the data from that study. Interestingly, in the top two graphs, you can see that both heart rate and mean arterial pressure were rescued by vasopressin. But interestingly, epinephrine did not have much of an effect, and the pigs still died. Here's another porcine study, where pigs were given either vasopressin and fluid resuscitation, placebo and fluid resuscitation, or just fluid resuscitation alone. In the vasopressin arm, 9 of the 9 pigs lived. But none in the placebo or fluid resuscitation groups lived. But the beneficial effects of arginine vasopressin are not limited to pigs. In this study utilizing ewes, the animals were first subjected to a hemorrhage, then given crystalloid, followed by a blood transfusion. And this was to mimic the pre-hospital transitioning into the hospital setting. After this, they were either given vasopressin, epinephrine, or remained control animals. Interestingly, the arginine vasopressin group had preserved cerebral blood flow. Carotid flow in this group was higher than either the epinephrine group or the control group. There was also noted to be improved regional brain oxygenation. This occurred in both the epinephrine and the arginine vasopressin groups. Going back to a porcine model, this group looked at survival in swine after cardiac arrest from hemorrhagic shock. They first induced hypotension via hemorrhagic shock. They then induced ventricular fibrillation and subject the pigs to 4 minutes of CPR. At that point, the pigs either received epinephrine, vasopressin, or saline, and were defibrillated. It would seem that survival was positively impacted by both vasopressin and epinephrine, as a majority of the pigs lived in these groups, while none of the placebo swine survived to return of spontaneous circulation. But the story gets a little more complicated. So 30 minutes post-resuscitation, renal blood flow was better in the vasopressin group compared to the epinephrine group. The epinephrine group did have a quicker rebound of blood pressure during resuscitation compared to vasopressin, but unfortunately this group developed a profound acidosis by 15 minutes post-resuscitation, and even though they survived immediately, all of those animals died before 60 minutes, thus leaving the only surviving animals to be the vasopressin group. So it seems that vasopressin, though not epinephrine, may be helpful as an adjunct to hemorrhagic shock in animals. Well what about human studies? So vasopressin use is associated with death in acute traumatic patients with shock. Maybe that's not what we thought was going to happen. Pays to look at the details of this study. It is important to note that this was a retrospective study looking at arginine vasopressin use as a second and or single agent. If you look in the fourth model, they adjusted their logistic regression for injury severity, age, initial lactate reading, and the presence of a severe head injury. And it certainly appears that when controlling for these confounders, arginine vasopressin use may be associated with increased mortality. While the data just presented suggests an association between arginine vasopressin and mortality, the truth in retrospective studies is always a little more complicated, as there certainly can be confounders we are just not aware of. Luckily we have some randomized prospective studies to look at to see if arginine vasopressin has a positive effect. This is one of the first ones. It was published in 2011. It was a randomized, double-blind, controlled trial which enrolled 70 hypotensive trauma patients. The experimental group added vasopressin to their resuscitative fluids as part of their trauma resuscitation. This was a 4 unit bolus followed by 2.4 units per hour for up to 5 hours. This bolus was held if the systolic blood pressure was over 160 mmHg. Inclusion criteria were pretty simple and included patients over the age of 18 presenting with traumatic injuries and a systolic blood pressure less than 90. Interestingly, this group found a significant decrease in utilization of crystalloids and blood and blood products in the arginine vasopressin group. This difference manifested early within the first 20-24 hours, but seemed to persist over the first 5 days. They did not notice any difference in mortality, multi-organ failure, or incidence of serious or moderate adverse events. This study does have some limitations. It was initially powered for 165 patients and was supposed to enroll in 3 separate centers. Only one of those centers was successful at recruiting and again, only 78 patients were enrolled. Additionally, 10-15 liters of fluid in the early resuscitation of trauma patients may not represent our modern resuscitation standards. Another recent randomized study looked at 100 trauma patients to see if low-dose supplementation of arginine vasopressin would reduce blood product transfusions in patients with trauma and hemorrhagic shock. This was a single-center, randomized, double-blind, placebo-controlled study enrolling adults who received 6 units or more of red blood cells within 12 hours of injury. You can see the exclusion criteria here, and as noted previously, they enrolled 100 patients in this study. The experimental group received 4-unit bolus of vasopressin followed by 0.04 units per hour. Titration was begun after hemorrhage control for an appropriate mean arterial pressure. If the patient was hypotensive despite the 0.04 unit vasopressin, they proceeded with standard resuscitation and pressors. The infusion could last up to 48 hours, but if the patient was hemodynamically stable after hemorrhage control, it was stopped. Their primary endpoint was 48-hour volume of blood product replacement. Secondary outcomes included volume of crystalloid, estimated blood loss, 48-hour fluid balance, and total pressor use. Additionally, they evaluated 30-day mortality, length of stay, and various complications. Similar to the first study, the intervention group here received fewer blood products, had significantly lower fluid balance, but they did have a higher rate of deep vein thrombosis. Interestingly, despite this higher rate of DVTs, the overall rate of all complications was not different between groups. There were no differences in other pressor use, open abdomen time, ventilator time, and there was a trend towards improvement in acute kidney injury and ICU length of stay, but these were not statistically significant. So it seems arginine vasopressin may lead to some improvements in trauma patients, at least in regards to blood product utilization and crystalloid utilization, when used as an adjunct to resuscitation. But are there other complications we need to be aware of, if we start using this drug liberally? Well, one, it can lead to diarrhea, two, it can lead to increase in pulmonary capillary leak and renal capillary leak in rats, but interestingly, a decreased capillary leak in the intestines. It can lead to DVTs, as noted in the prior human study. More importantly, we should probably consider not using it in the setting of brain injury as it might exacerbate SIADH. So given the information presented, should vasopressors in trauma really be a never event, at least in regards to vasopressin? Again, currently there are some non-controversial uses of pressors as it relates to trauma. For TBI, for instance, to maintain appropriate MAP, to maintain cerebral perfusion pressure, I think all of us would use a pressor. If someone is undergoing acute massive exsanguination and needs a bridge to their resuscitation and hemorrhage control to avoid arrest, it's probably appropriate to use a pressor. If someone has hypotension and shock despite appropriate blood product replacement due to the late vasodilatory effects of hemorrhage, then yes, they should also be on a pressor. But should we use it as an adjunct to our resuscitation early in all comers? Well, that question is still out, but the data is intriguing for vasopressin that perhaps we should utilize it in our trauma bay. Well, I've talked a significant amount about vasopressin use as an adjunct to our resuscitation, but what about other pressors? Now vasopressin has the most data behind it, likely because there is an early vasopressin deficiency in trauma patients. However, I'd be remiss if we didn't at least briefly discuss the use of other pressors early. While there is less data, there is some. Here is one study looking at the early use of norepinephrine on in-hospital mortality in hemorrhagic shock. The advantage of this study is it is a propensity score analysis, so hopefully eliminates a significant amount of confounders, although it is still a retrospective study. Interestingly, before matching, norepinephrine was associated with a worse odds of survival. But once the propensity score matching was done to eliminate confounders, survival appeared equal between groups. This suggests that maybe norepinephrine is necessary in those sick patients that need a bridge to resuscitation and hemorrhage control, or otherwise have that late vasodilatory effect, but it may not be an appropriate adjunct in the trauma bay. So should at least vasopressin be recommended as an adjunct to hemorrhagic shock resuscitation? Well, there is intriguing data that maybe it is helpful, but that data is still early and as of yet cannot be recommended as standard of care. I do look forward to seeing more studies. Once again, I would like to thank the Congress and Gail Gesson for the opportunity to present this talk. I look forward to seeing you all next year, hopefully in person, at the 52nd Critical Care Congress. Thank you.

vasopressors

hemorrhagic shock

resuscitation

blood product administration

trauma patients