SCCM Resource Library-The Practical Magic of Steroids in Sepsis

Video Transcription

Welcome to our session entitled The Practical Magic of Steroids and Sepsis. It's my honor to act in the role of moderator for this session. My name is Paul Zamita.  I'm the Director of Clinical Pharmacy Services at Brigham Women's Hospital in Boston, Massachusetts. I have no disclosures. I want to introduce the faculty for today's talk.  First is going to be Dr. Kimberly Clark, speaking on the laboratory tests, are they necessary or wasteful?  Dr. Clark is a clinical pharmacy specialist at Prisma Health Upstate in Greensville. The next speaker will be Dr. Peitz, speaking on the steroids and sepsis shock.  Gregory is a clinical pharmacy coordinator in adult intensive care medicine at Nebraska Medical Center.  And finally, Dr. Steven Pastorez is going to be speaking on when and how should steroids be discontinued.  Dr. Pastorez is the Program Director for Critical Care Medicine at Memorial Sloan Kettering Cancer Center in New York, New York. Thank you and enjoy the session. Good afternoon.  Our topic for this session is the practical magic of steroids and sepsis. I do need to briefly mention that I do not have anything to disclose at this time.  Let's start with a very basic review of the hypothalamic pituitary adrenal or HPA axis, since this is very central to our discussion today.  The hypothalamus is located in the brain and releases corticotropin-releasing hormone in response to a variety of stimuli.  CRH then acts on the anterior pituitary gland to release adrenocorticotropic hormone. And then ACTH works on the adrenal gland to release cortisol and a variety of other hormones.  Cortisol works negatively on this axis, providing negative feedback to decrease further secretion.  During times of stress, the APH axis is activated and we see increases in CRH, ACTH, as well as cortisol.  And the negative feedback of cortisol is reduced, which allows higher than normal concentrations of cortisol to be produced.  Cortisol usually increases in amounts roughly proportional to the severity of illness and cortisol plays a very vital role in mediating a lot of the physiologic and pathophysiologic  responses to stressful events, including an enhanced cardiovascular response to vasopressors and inhibition of the synthesis of inflammatory markers.  Now sepsis is a stressor that can lead to the activation of the HPA axis. This results in the release of pro-inflammatory mediators, in particular cytokines, and cortisol  can act to attenuate these inflammatory molecules. Under normal circumstances, the body's natural increase in cortisol should manage the pro-inflammatory effects of sepsis.  However, in some cases, the levels are inadequately elevated for the degree of stress in the body.  When the patient's own cortisol levels are not sufficient for the severity of the patient's  illness, the patient is said to have critical illness-related corticosteroid insufficiency, or SIRSI.  And SIRSI is the result of multiple derangements in the HPA axis, so we see a decrease in the production of CRH, ACTH, and cortisol.  In addition, we have recent studies that show that the metabolism of cortisol is also altered.  You see reduced metabolism due to a decrease in the cortisol metabolizing enzymes in both the kidney and the liver.  And then finally, there is a reduction in the glucocorticoid receptor density and activity leading to cortisol resistance in the tissues.  These abnormalities can coexist and change dynamically over time. The incidence of SIRSI is said to be as high as 60% in patients with septic shock, and  in studies, it's associated with worse outcomes, including an increase in mortality. This inadequacy of natural cortisol suggests a role of exogenous corticosteroids in sepsis  management. So, from the 1970s to about 2017, we had over 20 randomized placebo-controlled trials that  looked at either 28-day mortality or in-hospital mortality as an outcome. These studies started in the 1970s and 80s, and these studies used short courses of high-dose  steroids, but these studies actually found that patients with steroids had worse outcomes.  So, in the 1990s, there was a series of several small studies, usually about 40 patients in each, that were published looking at lower doses for longer courses.  And then we have the first landmark study that was published in 2002 by NAHN. Since then, we've had three other landmark studies, the Corticus, Adrenal, and Approach studies.  We also have multiple guidelines that make recommendations regarding the use of steroids. However, despite this wealth of literature and guidelines, there's still a great deal  of conflict and controversy over the practical or day-to-day use of steroids in sepsis.  So, during this session, we're going to look at some of these specific controversies, starting with what is the role of laboratory testing in determining the need for steroids?  Now, there are a variety of tests that can be used to assess the need for steroids, but today we're going to focus on the ACTH stimulation test and the plasma cortisol level.  So, the ACH stimulation test traditionally consists of measuring cortisol levels at baseline and then at 30 to 60 minute intervals after the administration of cosentropin, 250  micrograms IV, and cosentropin is just a synthetic form of ACTH. And the current definition of CRC is a change in baseline of cortisol at 60 minutes of less  than 9 micrograms per deciliter. Now, the rationale behind the ACH stimulation test is that if the adrenal gland responds  well to an exogenous stimulus, we would expect it to respond in a similar fashion to an endogenous stress factor.  Now, while this rationale certainly seems logical, there are several limitations to this test. First, the test was originally described in non-stress subjects, and its applicability  and interpretation in the setting of critical illness is uncertain. There is also no consistent relationship between cortisol response to cosentropin and illness severity.  So while the currently accepted cutoff for defining CRC is less than 9 micrograms per deciliter, this number really has not been fully investigated.  In addition, there is question regarding what is the appropriate dose of cosentropin. As I mentioned, the traditional test includes a 250 microgram dose.  However, this dose tests the maximal secretory capacity of the adrenal cortex and can result in superphysiologic levels of ACTH.  Therefore, some people believe that this may override adrenal resistance to ACTH, and therefore you may miss some mild secondary adrenal insufficiency cases.  To avoid this, some people prefer to use a smaller 1 microgram dose. However, this dose has its own limitations.  While it does produce a more natural physiologic concentration and patients do respond differently,  we don't really know what an appropriate response to this dose looks like because it has been minimally studied in the critically ill population.  In addition, there is not a commercially available formulation for this dose, and so a dilution is required which creates a logistical challenge.  Here I've included a study that highlights one of the inaccuracies of the ACH stimulation test.  So, this study looks at 20 adult ICU patients that were diagnosed with sepsis, and in the study, they measured cortisol levels every hour for 24 hours, and then at that 24-hour  period, they performed an ACTH stimulation test.  They then calculated what was the maximal spontaneous hourly rise in ACTH and compared it to what the 60-minute rise was after the ACTH stimulation test.  And so you can see the results here in this graph at the bottom, where we have the plasma cortisol on the y-axis, and the individual patient results are on the x-axis.  And we have the spontaneous change of the patient highlighted in green versus the ACTH stimulated change in blue.  And what we see from these 20 patients is that there is really no consistent relationship  between the maximal spontaneous hourly rise in plasma cortisol and the corticotropin-induced rise.  So, these results really don't support that initial thought that the ACTH stimulation test can be used as a surrogate for the body's natural response.  The inaccuracies in this test certainly could explain some of the different results we see in the literature.  So, both the ANON 2002 study, as well as the CORTICA study, performed corticotropin or ACTH stimulation tests.  In the ANON study, they found a decrease in 28-day mortality among non-responders between those that received steroids and those that received placebo.  And so, those authors concluded that the test was useful. On the other hand, the CORTICA study did not see a difference in 28-day mortality among  non-responders and concluded that the test was not useful. When we look at our two most recent studies, first, we have the ADRENAL study.  The ADRENAL study did not even perform ACTH stimulation tests, citing that there was poor literature for the test and a lack of current recommendations for utilizing it.  And so, they did not include it as part of the study protocol. The APPROACHUS study did plan to include the ACTH stimulation test as part of their protocol.  However, there was a widespread drug shortage of cosentropin that prevented them from performing the test in 461 patients.  For those that they did perform the test in, there was no difference in 90-day mortality among the non-responders who received steroids versus placebo.  Now, another option for testing is the plasma cortisol level. And currently, SIRSI is defined as a random plasma cortisol level less than 10 micrograms per deciliter.  This test has its own limitations. While the diurnal pattern of cortisol secretion is lost in critical illness, there are still significant fluctuations over a 24-hour period.  And so, depending on when you check a patient's plasma cortisol level, they may or may not meet the criteria for the SIRSI definition.  Another complicating factor is that 90% of circulating cortisol is bound to plasma proteins, such as the corticosteroid binding protein and albumin.  And concentrations of these proteins decline in critical illness, resulting in a higher free cortisol for the same amount of total cortisol.  So if you check a total cortisol, you're not differentiating between that protein bound and free cortisol, which is that active form.  So you might ask, why not always check a free cortisol? Well, the ability to check a free cortisol is going to be limited, because this measurement  requires a mechanical separation of the protein bound and unbound forms, and is technically difficult to perform and cumbersome.  And so, a lot of facilities do not offer this test in-house. In addition, temperature control is pretty critical for this test, and even slight variations  in temperature can cause marked influence on the level. The final limitation of these cortisol levels that we need to discuss is high variability among the various assays.  So here's a study that highlights this variability. It was decided ahead of time to do a sub-study of the corticus population, and they wanted  to compare cortisol measurements performed at the local hospital laboratories with those from a central reference laboratory.  And the results of the study showed that 23% of patients were misclassified as non-responders, depending on which laboratory result you used.  So here in the table, you can see that 5% of patients were classified as non-responders by the central lab that the local lab said were responders, and there were 18% of patients  that the local lab said were non-responders that the central lab said were responders. And so based on this study, it's unclear whether the criteria for SIRSI can be readily  transferred to all cortisol assays and whether this test can really be used to determine the role of steroids.  Now both the 2016 and 2016 surviving sepsis guidelines provide recommendations about the use of these laboratory tests. The 2020 guidelines make very specific recommendations.  In the graded recommendations, they state not to use the ACTH simulation test to identify  patients with septic shock who should receive hydrocortisone, and in the commentary section, they discuss the random cortisol levels and mention that they have not demonstrated to  be useful. Now the 2016 guidelines do not make any graded recommendations about laboratory tests, but  if you read in the commentary section, it reiterates the same recommendations as the 2012 guidelines.  Now in addition to the surviving sepsis guidelines, the Society of Critical Care Medicine and  the European Society of Intensive Care Medicine provided an update to the diagnosis and management of SIRSI guidelines back in 2017.  And this guideline makes no recommendation about whether you should use the ACTH simulation test or a random plasma cortisol test for the diagnosis of SIRSI.  And again, the rationale section of these guidelines explains this a little bit more. It states that, in elaboration, that after two rounds of voting, the task force members  were not able to reach a consensus, and a consensus was defined as greater than 80% agreement.  And so they were not able to reach a consensus on whether the ACTH simulation test was superior to the random cortisol for the routine diagnosis of SIRSI.  And ultimately, the task force felt that it is unlikely that a single test can reliably diagnose SIRSI.  So to summarize, in terms of laboratory testing, I think that we can all agree that the currently  available laboratory tests have some significant limitations, which could lead to the erroneous  conclusion about adrenal function in the patient and the initiation of an inappropriate therapy.  So therefore, at this time, our laboratory tests should not be routinely used to assess adrenal function or to guide steroid therapy in critical illness.  So if you cannot use a laboratory test, then you're left with some very determining patient-specific  factors that could be helpful, which leads us to our next practical use question. Hi. My name is Greg Pites.  I'm very thankful to be here today to speak on this topic, and also I'd like to give a big thanks to Dr. Clark and Dr. Pastore for being solid bookends to this presentation today.  I have the distinct pleasure to follow up with Dr. Clark's presentation with the hope  that I can add some information on how steroids should be considered for use in a septic shock population.  I have no financial disclosures to report today or related to any of the material being presented. Before we begin, however, I wanted to comment on the title of this presentation,  The Practical Magic of Steroids and Septic Shock. To me, this seems like a strange and almost oxymoronic title for this material.  I know we've all witnessed those events where steroids were absolutely the key at turning a patient around and other scenarios where there may be more ambiguity in their effect  on outcomes. So perhaps the best way to summarize these variable magic effects then can be reflected by the following quote, there is no magic in magic, it's all in the details.  With that said, I am about to dive into more specific detail that may help us aid in our decision regarding the use of these agents.  As we get started, let's take a broader 10,000 foot view, perhaps reviewing the latest guidelines  on the critical illness related corticosteroid insufficiency article published in 2017. We can see from the recommendation that there's support given to corticosteroids for septic  shock patients who are failing initial therapies including fluids and requiring vasopressors.  When we make these decisions on the applications of steroids in a critically ill patient though, it's probably prudent to consider things on a more granular level.  Do all patients benefit from steroids? What added benefit or benefits are we aiming for? Are there meaningful outcomes outside of mortality?  Also, when should we begin contemplating initiating corticosteroid therapy? And lastly, are there some corticosteroid regimens that are better than others?  If we reflect back on the slide from Dr. Clark's previous presentation, we're fortunate to  have accumulated a number of studies and guidelines over the past 20 plus years that can aid in  our application of corticosteroid management and perhaps answer some of these questions. When we're asked the question regarding the holy grail of outcomes when it comes to studies,  the first thought that often comes to mind is how have corticosteroids potentially influenced mortality?  In the past 20 years, we've had two really well-designed trials that have perhaps shown this demonstration in mortality benefit.  As we're all aware, Anand's initial corticosteroid therapy paper in 2002 was one of the first  to demonstrate the benefit of corticosteroids in a relatively adrenal insufficient septic shock patient population.  A follow-up 15 to 16 years later in the APPROACHES trial, we again saw a mortality benefit in a septic shock population, this time at 90 days from an index event and as you can see  in a much larger patient population. While both of these studies have resulted in mortality benefits, we've also witnessed  similar trials that have failed to yield similar survival outcomes. The CORTICUS trial published in 2008 did not demonstrate similar findings to Anand's 2002  publication in that relative adrenal insufficient septic shock patient population. But we know that there may have been downsides to this study, particularly having a lower  number of patients than needed to be adequately powered. Likewise, the recent adrenal trial from 2018, which enrolled almost 4,000 patients and instituted  a continuous infusion of hydrocortisone, did not show any difference in 90-day mortality, unlike the APPROACHES trial, which was also published in that same year.  So due to the discrepancies in these reported outcomes, this begs the question as to whether there are substantial differences between the trials in terms of their enrollment.  This slide takes an aim to potentially present baseline demographics of the patients in these four trials.  When examining baseline and clinical characteristics of the patients in these four large trials,  there appears that they are quite similar in the patient populations that are studied. As you look across the various studies, similar in age, gender, most of these patients are  mechanically ventilated and we're seeing a larger proportion of patients perhaps in the APPROACHES trial that were admitted with a primary indication of pulmonary infection.  So although we have no major differences between some of these baseline demographics, if we look closer at the severity of illness, we begin to see a difference amongst the four  trials. In an on-end process trial, not only enrolled patients with a higher predicted in-hospital  mortality, as evident here on the left-hand side, but they also demonstrated higher observed mortality in both the placebo and the treatment groups.  If you compare them to both the corticus and the adrenal trials. So given that these two trials demonstrated steroids affect the mortality, we can perhaps  rationalize that patients with greater acuity may benefit more from the use of corticosteroids.  This proposed rationale for increased response to greater acuity should be approached with caution, however.  The subgroup analysis from the adrenal study did not support this argument when it compared  outcomes for those subjects who were presented with more severe septic shock, as identified either via their catecholamine dose, as indicated here above, or their APACHE-2 score.  Therefore, I think we should take that piece of information with a grain of salt that steroids are only impactful in a dramatically ill patient population.  So while we may have more compelling evidence then to initiate corticosteroids in higher acuity illness patients, what are we to do for those patients who present to us with  all sorts of a range of acuity? Should corticosteroids be abandoned or not used in patients who present with less severe septic shock, or are there other benefits worth considering?  We attempted to make decisions primarily based on mortality, but I do not think that we should ignore other tangible outcomes that these four trials have illustrated.  As we review this table, you'll see that although the corticose and adrenal trials did not demonstrate  survival outcomes, we can see their benefits in many other ways from shortened duration  of the vasopressors, resolution of shock, and in the case of the adrenaline trial, improvement  in both the mechanical ventilation and ICU length of stay, all which are extremely valuable outcomes for patients that we care for in a critically ill environment.  It's important to distinguish, however, that broad-reaching benefits does not apply to all the application of steroids in other septic shock states.  So because of the theoretical mechanisms of using steroids in septic shock, a hyper study was developed to assess whether or not we could blunt the progression of shock with  the use of hydrocortisone administration. So this study, which included 353 patients who were not on vasopressors at the time of  enrollment, was unable to find a difference in outcomes in terms of their mortality and length of stay or other pertinent findings with the use of hydrocortisone, and therefore  should likely not be a major driver in the use of corticosteroids early in therapy before the presence of shock or before the presence of vasoactive medications.  When should we administer steroids then? If we look closer at the four major trials, we can see some considerable differences in  both when patients were allowed to be enrolled in the study as well as the initial vasopressor requirements seen in the baseline demographics.  If you look at each study, the one thing that is noticeable is the magnitude of vasoactive support that were included in each trial.  While we are not necessarily comparing apples to apples between these four trials, you may identify that the norepinephrine doses for patients in the corticosteroid trial and the  adrenal trial may be less than the other two studies when looking at a weight-based dosing fashion of norepinephrine.  It should be noted that the CERCI guidelines suggest that corticosteroids would be warranted if patients are on vasoactive support of an equivalent of norepinephrine of greater  than 0.1 microgram per kilogram per minute. Although this dose is not, as you can reflect upon among these four trials, is not necessarily  substantiated and perhaps would need further evaluation. The only study that perhaps gives some extra granularity on the matter was again the recently  published adrenal trial from 2018 in which the subgroup analysis demonstrated a possible benefit for corticosteroids if patients were randomized within the 6-12 hour time period.  This should be also taken very cautiously and this should only serve as hypothesis generating. It does not reflect upon best practice for patients at this time.  What steroid regimen should be used? Is flugelcortisone the magic that we're looking for?  If we review the two trials that demonstrated a mortality benefit, we notice that both trials  were consistent with their dual corticosteroid regimen using both hydrocortisone 50 milligrams every 6 hours as well as flugelcortisone 50 micrograms daily for 7 days.  The rationale for this combination therapy was due to the potential for concomitant primary adrenal insufficiency which could be exacerbated during septic shock.  So although despite the investigator's rationale for using flugelcortisone, it prompts an interesting  question as to whether or not flugelcortisone should be preferentially used in this patient population.  For more on this or to dive a little bit deeper into the question, let's focus briefly on the pharmacologic profile of flugelcortisone.  At a very peripheral view, flugelcortisone has a distinct ability to stimulate mineral corticoid receptors much more than its corticosteroid counterparts.  I believe there is another session during this Congress that will discuss these caveats of corticosteroid pharmacology in greater detail, so I will plan on foregoing any further  conversation on the matter, but however, with flugelcortisone's degree of affinity for the mineral corticoid receptor, it is compelling that this may play a role why these  trials demonstrate a mortality benefit. Let's take a little bit closer look at some of the literature on the matter.  In 2002, there was a multi-center randomized trial from Anand and colleagues who sought to explore the possibility of flugelcortisone's benefit in this patient population.  There were a total of 509 total patients with septic shock who were randomized into four interventional groups.  While the primary goal of this study was to test the efficacy of intensive insulin therapy  for those patients with septic shock on steroids, there was also a secondary objective to assess whether or not there was a benefit for flugelcortisone when added to hydrocortisone.  Patients were randomized to these four groups as seen here. The bottom line for the secondary outcomes were that there were no differences in hospital  mortality between those patients who received hydrocortisone alone versus those who received hydrocortisone plus flugelcortisone.  Furthermore, in addition to the paucity of substantial trial validating the role of flugelcortisone,  there's some other compelling data, uncompelling data, that argues against the reliability of flugelcortisone in the septic shock population.  This study from 2016 was a small pharmacokinetic study that looked at 21 patients directly  from the CRYSTL trial, and the results potentially cast a shadow on the use of flugelcortisone in the population.  This PK analysis found that about one-third of the patients had undetectable flugelcortisone  plasma concentrations after administration, which definitely calls into question the availability of enteroflugelcortisone in critically ill patient populations.  Also of note was the fact that flugelcortisone was shown to have a fairly short half-life with just a bit over an hour, which, again, puts into question its potential to play a  continuous role for hemodynamic effect in this tenuous population. So then, what corticosteroid regimens should be utilized?  If we rely strictly on the four most robust clinical trials investigating this matter,  it would seem most logical to initiate hydrocortisone therapy with a dose of 200 milligrams a day, either a divided or continuous dose strategy.  While that may be one hydrocortisone regimen, there have definitely been others that have been studied as well.  To assess the potential effects of other corticosteroid regimens, Zhang and colleagues recently performed a meta-analysis using Bayesian modeling.  They included a total of 35 randomized controlled trials that compromised almost 9,000 patients with the intent of analyzing various steroid regimens on 28- to 31-day mortality.  Among the comparisons they made, they evaluated how corticosteroid dosing affected outcomes.  What they found was that when assessing various total daily dose of corticosteroids, they  discovered that those patients who received 200 to 400 milligrams of hydrocortisone equivalent  a day were the only dosing group that had a statistically significant reduction in short-term mortality.  These results were obviously subject to various biases in trial heterogeneity, but perhaps  due to confirmed CERCI guidelines that suggest using hydrocortisone doses of 400 milligrams a day for the treatment, they do suggest that we could use doses up to 400 milligrams a  day for these particular patients. In summary, I believe that we have identified substantial evidence to prove that there are  meaningful benefits for patients to receive corticosteroids who are in septic shock. After reviewing primary studies, inclusion criteria, and mission demographics, we really  should only use these drugs in those patients who are exhibiting shock. And lastly is I believe that doses between 200 milligrams to 400 milligrams of a hydrocortisone  equivalent a day should be used. Although the addition of fluidicortisone may be added to the care, I believe that further  robust studies may be helpful to further define its role in application of these patients. With that, I'd like to again thank SCCM for this opportunity, and I personally welcome  any further discussion on this topic. Thank you. Hi, this is Steve Pastoris, Program Director, Critical Care Medicine, and Vice Chair of  Education in the Department of Anesthesiology and Critical Care Medicine at Memorial Sloan Kettering Cancer Center in New York City.  I'm delighted to participate in this session on the practical magic of steroids in sepsis. My specific topic for the next 15 minutes or so is on when and how should steroids be  discontinued. This is my disclosure slide. I have none related to the content of this presentation. We just heard from Greg the role of steroids in septic shock.  Steroids in patients with septic shock have been used for decades. The use of steroids in septic shock has evolved certainly since the late 1990s following the  demonstration in small trials that when compared to placebo, low-dose hydrocortisone in doses of 200 to 400 milligrams per day resulted in faster shock reversal and withdrawal of  vasopressors. This then led to larger major trials, the ANAIN-FRENCH trial in 2002 and the APPROACH  trial in 2018, which also showed faster shock reversal and mortality benefit with the administration  of both hydrocortisone and flutocortisone, as well as other randomized controlled trials, including the CORTICUS trial in 2008, the HYPRESS trial, and the ADRENAL trial in 2018,  which also showed faster shock reversal but without a mortality benefit. It's important to recognize that steroids are not benign drugs.  Steroids have been associated with many side effects, including hyperglycemia, hypernatremia,  neuromuscular weakness, superinfections, upper gastrointestinal hemorrhage, arrhythmias, and psychoses.  In the latest systematic review and meta-analysis from the Cochrane in late 2019, looking at  thousands of patients that were enrolled in several randomized controlled trials and observational  studies using steroids for septic shock, they concluded that the use of steroids for septic shock was associated with an increased risk of muscle weakness, an increased risk of  hypernatremia, as well as hyperglycemia. But they also found little or no difference in GI bleeding, stroke, or cardiac events,  little to no difference in neuropsychiatric events with low certainty of evidence, and most importantly, corticosteroids probably do not increase the risk of superinfection  in the doses that were being used in the trials for septic shock patients. It is important to understand the pharmacologic principles that guide glucocorticoid therapy  in patients with septic shock, as well as in patients with sepsis-induced ARDS. The duration of glucocorticoid administration and the tapering of corticosteroids are the  main determinants of treatment efficacy. So the question is, when should steroids be discontinued in patients with septic shock?  Well, there has been no comparative study between a fixed duration and a clinically guided regimen.  There have been at least three studies that look at randomized controlled trials with fixed duration of corticosteroids.  The Ballard study, published in 1998 in Critical Care Medicine, used a five-day fixed dose of hydrocortisone, 100 milligrams intravenously, Q8 hours.  The ANAIN trial in 2002 used the combination of hydrocortisone and flutocortisone for seven days.  The CORTICUS trial, published in 2008, used hydrocortisone for at least five days and then tapered it over the next six days for a total duration of 11 days of treatment.  When we look at the most recent steroid trials in septic shock, the ADRENAL trial and the APPROACH trial, which were both published in 2018, as we heard in the previous talk,  they gave corticosteroids for seven days without tapering. So what about comparing three days versus seven days of corticosteroid therapy in patients with septic shock?  So in a study by WHO published in Respirology in 2011, this was a prospective randomized pilot study of 130 patients with septic shock and relative adrenal insufficiency.  These were patients that did not have a response to the cortisol stimulation test of at least nine or delta-cortisol that was less than nine.  In those patients, the administration of hydrocortisone, 50 milligrams IV, Q6 hours was compared for three days versus seven days.  Patients in both groups had similar baseline variables and when they looked at 28-day mortality, there was no significant difference, 34% versus 37% in those that got seven days.  The median time to withdrawal of vasopressor was slightly less in the three-day group, five days, versus 6.4 days in the patients that got steroids for seven days, but this  did not turn out to be statistically significant. In the guidelines that we published in 2017 on the diagnosis and management of critical  illness-related corticosteroid insufficiency, we suggested that in patients with septic shock that are not responsive to fluid and moderate to high-dose vasopressor therapy,  we suggested the use of intravenous hydrocortisone of less than 400 milligrams a day for at least  three days at full dose before initiating a taper in those patients that were no longer in shock.  Question then is when should steroids be discontinued in patients with septic shock? Well the consensus is that when the patient is no longer requiring vasopressor therapy  to maintain blood pressure, and typically this would be at least 12 to 24 hours, unless the patient has an underlying endocrine or corticosteroid use history, in which case  corticosteroid therapy has to continue perhaps in lower doses for maintenance therapy in these types of patients.  The next question then becomes how should steroids be discontinued in patients with septic shock? Again here, there has been no comparative study between steroid tapering and abrupt  discontinuation of steroids. So one might wonder, well, is there potential harm with the abrupt withdrawal of corticosteroids? And the answer appears to be yes.  In a classic paper by Didier K published in the Blue Journal in 2003, they analyzed 40 patients in which hydrocortisone was given in one group at 100 milligrams loading dose  for over 30 minutes, followed by 10 milligrams per hour until day 3, and then isotonic saline was administered from day 4 to 6.  In the other group that got corticosteroids, they first received placebo, followed by hydrocortisone.  The norepinephrine infusion was tapered individually with ongoing hemodynamic stabilization, keeping the mean arterial pressure greater than 70 millimeters of mercury.  The patients continued to receive treatment after day 6, and in case of ongoing hemodynamic instability, the hydrocortisone was given according to physician discretion.  When they analyzed the levels of cortisol and the levels of inflammatory cytokines in  these patients, what they showed was hydrocortisone infusion induced a significant hypercortisolemia, as you can see in the top left of this slide.  This was completely reversed after steroids were discontinued within 24 hours. On the right panel, you can see the graphs of IL-6 and IL-8.  The steroid group in dark circles or black circles showed a dramatic reduction in IL-6 levels and in IL-8 levels, which after day 3, when the hydrocortisone was switched to  placebo, the inflammatory cytokines IL-6 and IL-8 both went up, suggesting that in patients  who are abruptly withdrawn from corticosteroids, there is a potential of rebound inflammation with increase in pro-inflammatory cytokines.  We've known for years, even before that study by DDRK, that rebound inflammation can occur with rapid steroid taper.  We've seen that in the patients that were getting adjunctive therapy with corticosteroids for the treatment of pneumocystis carinii pneumonia or PJP pneumonia.  We saw that in the study of the steroid use in patients with persistent ARDS from the ARDS network in the 1990s and early 2000, and we saw that also abrupt withdrawal of  corticosteroids in patients with severe community-acquired pneumonia was also associated with rebound inflammation and worsening of the patient's clinical condition.  A more recent study looking at the timing, method, and discontinuation of hydrocortisone administration in patients with septic shock published in 2017 of a prospective cohort  study of 59 patients showed that there was no benefit with tapering strategy and in fact they found a higher incidence of hyperglycemia and hypokalemia.  So in summary, low-dose corticosteroids, when used for septic shock, can be certainly associated  with side effects, most commonly hyperglycemia, hypernatremia, and neuromuscular weakness.  Steroids can be discontinued when the patient in septic shock is no longer requiring vasopressor therapy.  If used for three days, corticosteroids should be gradually weaned as rebound inflammation  can occur if the steroids are abruptly discontinued, and steroid use and taper should always be guided by clinical response and assessment of side effects.  Thank you very much for your attention.

Video Summary

In this session, the speakers discuss the practical use of steroids in sepsis. They start by discussing the role of laboratory tests in determining the need for steroids, focusing on the ACTH stimulation test and the plasma cortisol level. They highlight the limitations of these tests, including uncertainties in their applicability in critical illness and lack of consistent relationship with illness severity. The speakers conclude that at this time, laboratory tests should not be routinely used to guide steroid therapy in critical illness. <br /><br />Next, they discuss the use of steroids in septic shock. They review the evidence from several landmark studies and guidelines, emphasizing the mortality benefit and other positive outcomes associated with the use of steroids in septic shock. They recommend considering the acuity of illness when deciding to initiate corticosteroid therapy. They also discuss the duration of therapy, highlighting the lack of consensus on the optimal duration of treatment. <br /><br />Finally, they discuss the discontinuation of steroids in septic shock. They stress the importance of tapering corticosteroids to prevent rebound inflammation and highlight the potential harm of abrupt withdrawal. They recommend discontinuing steroids when the patient no longer requires vasopressor therapy and caution against rapid tapering.

Asset Subtitle

Sepsis, Pharmacology, 2021

Asset Caption

The Society of Critical Care Medicine's Critical Care Congress features internationally renowned faculty and content sessions highlighting the most up-to-date, evidence-based developments in critical care medicine. This is a presentation from the 2021 Critical Care Congress held virtually from January 31-February 12, 2021.

Learning Objectives

Assess available laboratory tests for the diagnosis of critical illness-related corticosteroid insufficiency
Determine the role of fludrocortisone as an additive agent
Discuss criteria for when, and strategies for how, to discontinue steroids

Laboratory Tests: Necessary or Wasteful?
Kimberly B. Clark

Steroids in Septic Shock
Gregory J. Peitz

When and How Should Steroids Be Discontinued?
Stephen M. Pastores

Meta Tag

Content Type Presentation

Knowledge Area Sepsis

Knowledge Area Pharmacology

Knowledge Level Foundational

Knowledge Level Intermediate

Knowledge Level Advanced

Membership Level Select

Tag Shock

Tag Inflammation

Tag Pharmacokinetics Pharmacodynamics

Year 2021

Keywords

steroids

sepsis

laboratory tests

critical illness

septic shock

mortality benefit

corticosteroid therapy

discontinuation of steroids

	Society of Critical Care Medicine 500 Midway Drive Mount Prospect, IL 60056 USA Phone: +1 847 827-6888 Fax: +1 847 439-7226 Email: support@sccm.org	Contact Us About SCCM Newsroom Advertising & Sponsorship DONATE	MySCCM LearnICU Patients & Families Surviving Sepsis Campaign Critical Care Societies Collaborative	GET OUR NEWSLETTER
© Society of Critical Care Medicine. All rights reserved. \| Privacy Statement \| Terms & Conditions The Society of Critical Care Medicine, SCCM, and Critical Care Congress are registered trademarks of the Society of Critical Care Medicine.