Time Stamps

  • 01:50 Pathophysiology of Steroids and Infection
  • 03:50 Steroids in PNA Studies
  • 08:56 CAPE COD Trial
  • 14:13 Discussion of CAPE COD Trial
  • 18:36 Takeaways/Discussion

Show Notes

Background

  • Pneumonia
    • Most common cause of hospitalization in the US (outside of people giving birth)
      • High mortality rates (that have not changed) despite advances in antibiotics and supportive care 
        • Mortality rate of 13% in 1996 and in 2017
    • Steroids may help treat an over-exuberant inflammatory response associated with infection 
      • Steroids come with side effects. To name a few:
        • Hyperglycemia 
        • GI bleeding 
        • Infections 
        • Neuropsychiatric

RCTs of Steroids in Pneumonia

  • 2005 RCT: IV hydrocortisone vs placebo 
    • DURATION: 7 days 
    • POPULATION: ICU patients with severe CAP
    • SAMPLE SIZE: 46 patients 
    • RESULT: Reduced 60-day mortality 
      • 38% in steroid group vs 0% (!) in placebo group
    • Caveats
      • Small sample size 
      • Multiple comparisons without statistical adjustment 
      • Somewhat unbelievable effect size (0% mortality in ICU patients?) 
  • 2010 RCT: PO or IV prednisolone vs placebo 
    • DURATION: 7 days
    • POPULATION: Hospitalized patients with CAP (included ICU patients)
    • SAMPLE SIZE: 213
    • RESULT: No change in 30-day mortality
      • Steroid group had: 
        • Faster decline in fever 
        • Decline in CRP 
        • But more late failure
  • 2011 RCT: IV dexamethasone vs placebo 
    • DURATION: 4 days
    • POPULATION: Hospitalized non-ICU patients with CAP 
    • SAMPLE SIZE: 304
    • RESULT: No difference in in-hospital mortality
      • Steroid group had:
        • Reduced length of stay 
        • Higher hyperglycemia
  • 2013 RCT: IV hydrocortisone vs placebo 
    • DURATION: 7 days
    • POPULATION: Hospitalized patients with CAP (included ICU patients)
    • SAMPLE SIZE: 80
    • RESULT: Reduced mortality (6.7% in steroid group vs 31.6% in placebo) 
      • Steroid group had:
        • Improved P:F ratio
        • Reduced duration of mechanical ventilation 
        • Reduced length of stay
  • 2015 RCT (STEP): PO prednisone vs placebo 
    • DURATION: 7 days
    • POPULATION: Hospitalized patients with CAP (included ICU patients)
    • SAMPLE SIZE: 785
    • RESULT: No difference in all-cause mortality, a pre-specified secondary endpoint
      • Steroid group had:
        • Reduced time to clinical stability 
        • Reduced length of stay
        • Increased hyperglycemia
  • 2022 RCT (ESCAPe): IV methylprednisolone vs placebo 
    • DURATION: 7 days with 20 day taper
    • POPULATION: Hospitalized patients with severe CAP (included ICU/intermediate level of care patients)
    • SAMPLE SIZE: 586
    • RESULT: No difference in 60 day mortality
      • No difference in shock, ARDS, length of stay, 1 year morbidity/mortality, or adverse events
    • Caveats
      • Did not meet its enrollment goal of 1,400 patients 
      • Done in VA population 
      • Raising concerns about generalizability 

Criticisms of Prior Trials 

  • Sample sizes were small
    • May have been underpowered to detect a small difference in mortality
  • Different inclusion and exclusion criterias in every trial
    • Patients in one trial may not have been as sick as in other trials
      • For example…trials with liberal inclusion criteria may have included “pneumonia mimics” 
        • Aspiration pneumonitis or heart failure
  • Steroid Protocols!
    • Every trial used different steroid protocols 
      • IV vs oral
      • Type of steroid 
      • Duration of treatment
      • With vs without taper

Ways to Address Criticisms:

  • Sample Sizes! 
    • Multiple meta-analyses were done to address the issue of the small sample sizes but also had mixed results
      • 2015 Meta-Analysis: 2005 hospitalized patients with CAP (included ICU)
        • RESULT: Non-significant reduction in mortality (RR 0.67, 95% CI 0.45-1.01) with more effect seen in severe CAP
          • Steroids were associated with: 
            • Lower need for mechanical ventilation 
            • Reduced length of stay
            • Reduced time to clinical stability
            • Reduced ARDS
            • Increased hyperglycemia
      • 2017 Meta-Analysis: 1954 hospitalized patients with pneumonia (included ICU patients and HCAP)
        • RESULTS: Steroids were associated with:
          • Reduction in mortality for severe pneumonia (RR 0.58, 95% CI 0.4-0.84) 
          • NO mortality reduction not non-severe pneumonia (RR 0.95)
          • Reduced time to clinical cure 
          • Reduced length of stay 
          • Less development of respiratory failure or shock 
          • Fewer pneumonia complications
          • Increased hyperglycemia 
      • 2018 Meta-Analysis: 1506 hospitalized patients with CAP (included ICU)
        • RESULTS: No difference in 30-day mortality 
          • Steroids were associated with:
            • Reduced time to clinical stability 
            • Reduced length of stay
            • Increased hyperglycemia
            • Increased CAP-related rehospitalization
      • 2023 Meta-Analysis: 4661 hospitalized patients with CAP (included ICU)
        • RESULTS: Steroids were associated with: 
          • Mortality reduction in severe CAP (RR 0.62, 95% CI 0.45-0.85) 
          • NO mortality reduction not in non-severe CAP (RR 1.08)
          • Reduced mechanical ventilation
          • Reduced ICU admission 
          • Reduced length of stay
          • Increased hyperglycemia
    • Based on these meta-analyses, the benefit of steroids appeared to be best in patients with severe CAP!
      • BUT, the data were so mixed that professional societies didn’t agree on recommendations!
        • European (ESICM/SCCM) guidelines on CAP:
          •  Recommended steroids 
        • American (ATS/IDSA) guidelines  
          • Recommended against steroids

CAPE COD Trial

  • QUESTION: Does IV hydrocortisone reduce mortality in patients hospitalized with severe CAP?
    • Recruited 800 ICU patients (largest trial to date) and used strict inclusion/exclusion criteria
      • Large sample size and strict criteria! (combats concerns of earlier trials)
      • Inclusion Criteria:
        • Patients with severe CAP, defined as: 
          • Mechanically ventilated with a PEEP >5
          • P:F ratio <300 on high-flow or non-rebreather
          • Highest pulmonary severity index (PSI) group (risk score for CAP)
            • NOTE: different than the definition of “severe” CAP used by IDSA
      • Exclusion criteria:
        • Baseline septic shock 
          • Steroids are already used in septic shock for different reasons
          • Influenza
            • Evidence for harm of steroids in patients with influenza
          • History suggestive of aspiration 
            • Excluding aspiration pneumonitis
          • Active tuberculosis, hepatitis, or fungal infection
                • Risk of worsening infection with steroid use
    • GROUPS
      • IV hydrocortisone for 4 or 7 days with taper of 8 or 14 days 
      • Placebo 
        • NOTE: Steroids were administered within 24 hours of meeting any severity criteria! 
          • Faster than in other trials 
            • For example, steroids could be given up to 96 hours after admission in the ESCAPe trial
    • RESULTS:
      • PRIMARY: 
        • Reduction in 28-day mortality in steroid group (6.2% vs 11.9%)
      • SECONDARY: 
        • Lower rates of 90 day mortality (9.3% vs 14.7%)
        • Reduced rates of new intubation or vasopressor use
      • SAFETY OUTCOMES:
        • Higher doses of insulin in the steroid group, but no differences in rates of GI bleed or hospital acquired infection
      • SUBGROUP ANALYSES (hypothesis-generating only):
        • Patients with a higher CRP (>15 mg/dL) had a greater mortality benefit!

Discussion of Trials 

  • Why was such a big mortality reduction seen in the steroid group
    • Patients in the steroid group had lower rates of shock 
      • 15% compared to 25% in the placebo group
        • Some people argued that because the control group was not able to be treated with open-label steroids, their higher mortality rate may have been driven by untreated shock!
  • The patient population in CAPE COD was particularly sick with severe pneumonia
    • Strict inclusion/exclusion criteria ensured only the sickest patients were included, which are  most likely to benefit
      • Remember, many therapies work in sick populations but not in less sick patients!
  • Some of the patients who benefited may have had concurrent ARDS
    • There is a known benefit of steroids in ARDS!
      • Many of the patients in CAPE COD also met criteria for ARDS so was the benefit from the steroids treating their ARDS and not CAP?
  • Hydrocortisone’s strong mineralocorticoid effect may be more effective than other steroids
    • The primary effect of steroids in pneumonia is thought to be from the immune modulation from the glucocorticoid effect, but hydrocortisone has strong mineralocorticoid activity
      • However, the mineralocorticoid benefit theory doesn’t make much sense physiologically!
        • ICU patients are already fluid positive – having them retain more salt and fluid should be detrimental… 
  • The patients in CAPE COD had a high percentage of strep pneumo infection
    • Strep pneumo was the most commonly identified pathogen (20%)
      • Remember, steroids are used in strep pneumo meningitis!
        • However, most patients had no identified pathogen and these actually did better in subgroup analysis than the patients with strep pneumo…

Takeaways from Trials

  • Steroids appear to benefit in patients who are:
    • Sick with severe pneumonia 
    • More inflamed (higher CRP)
  • Many questions remain about the ideal protocol for steroid delivery 
    • IV vs PO ?
    • Duration? 
    • Timing of initiation?
    • Best steroid to use?
  • Clinicians should be judicious and thoughtful about which patients to apply results from the steroid trials in CAP to!
  •  

Transcript

Dr. Greg Katz: Welcome to the third installment of Beyond Journal Club, a collaboration between Core IM and NEJM Group. 

Dr. Clem Lee: The goal of Beyond Journal Club is to take landmark clinical trials and put them into context, telling the story of how we got to where we are and what it means for how we take care of patients. 

Dr. Suellen Li: Today, we’re going to talk about the use of steroids in the treatment of community-acquired pneumonia. I’m Dr. Suellen Li, an editorial fellow at the New England Journal of Medicine and a hospitalist at MGH.

Dr. Greg Katz: I’m Dr. Greg Katz, a cardiologist at NYU. 

Dr. Clem Lee: I’m Dr. Clem Lee, a former fellow and current guest editor at the New England Journal of Medicine.

Dr. Suellen Li: Today, we’re talking about the CAPE COD trial, which was published in the New England Journal of Medicine in May of 2023. 

Dr. Clem Lee: The CAPE COD trial explored whether using glucocorticoids in the treatment of severe community-acquired pneumonia (or what we’ll call CAP from now on) would reduce mortality.

Dr. Greg Katz: Thinking about this trial, in talking to a bunch of my colleagues who take care of many patients with pneumonia, the theme that everyone kept coming back to is that interpreting a trial like this and applying it to our patients is one of the most exciting and wonderful things about being a doctor. 

Dr. Suellen Li: Right, there is so much conflicting evidence – the history of steroids in pneumonia is like a pendulum that swings back and forth between positive and negative trials. It’s ultimately up to you to decide whether the patient in front of you fits or doesn’t fit into T able 1.

Dr. Greg Katz: And putting trials like CAPE COD into the context of the body of research that came before is not only important because of what CAPE COD trial found but also because applying a trial like this in the face of conflicting evidence is part of the art of medicine and being a doctor.

Dr. Clem Lee: So today, we’re going to start off by laying the groundwork with the rationale for using glucocorticoids in infections, specifically in pneumonia. 

Dr. Greg Katz: Then, we’ll discuss the conflicting findings from past studies of steroids in pneumonia.

Dr. Suellen Li: And finally, we’ll take a deep dive into the CAPE COD study and talk about what we as clinicians can take away from the results. 

Pathophysiology of Steroids and Infection

Dr. Clem Lee: Before we talk about steroids, let’s first bring it back to basics and think about what happens when we get an infection and why steroids might even help. 

Dr. Suellen Li: So with an infection, there are two major problems: the first one is the direct effect of the invading organism and the second is a consequence of your body’s response to that stress, which is inflammation. And as we know, inflammation is a key component of our defense against infection, but an over-exuberant inflammatory response can cause harm. We can target the first problem, the invading organism with antibiotics but we don’t have the same kind of targeted therapies for the second problem, which is over-activation of the immune system. 

Dr. Greg Katz: Exactly. And just as with real estate, location matters. Where you get an infection, and the resultant inflammatory response is important. The location affects how big of an impact that inflammation has. 

Dr. Clem Lee: Right, to contrast that to another part of the body- if we think of inflammation in the GI tract with gastroenteritis, we lose fluids and electrolytes but we can just replace it. There’s not a finite amount of space. But with the lungs, there’s limited space. Sp if an inflammatory response spirals out of control, we quickly run out of room to ventilate. 

Dr. Suellen Li: Right! I love the comparison to real estate. So it sounds steroids would be the perfect drug to lower inflammation, right? I mean we use it in so many other conditions.

Dr. Greg Katz: It certainly makes sense that steroids, because they lower inflammation, may help curb the body’s immune response in the lungs, and improve outcomes if there’s an infection in that finite ventilatory space. But we know that steroids aren’t totally benign.

Dr. Clem Lee: Yeah there’s a whole bunch of adverse effects: hyperglycemia, GI bleeding, infections, neuropsychiatric side effects, the list goes on and on… I was surprised to learn from the recent CoreIM episode on steroids that even a short course of steroids can have side effects. 

Dr. Suellen Li: Yeah! I’ve had so many patients whose blood sugars go wildly out of control on steroids. And hospital-acquired delirium is already so common – I would hate to start steroids without a good reason. 

Dr. Clem Lee: Yeah, I totally agree – I think we would have to be sure the benefits outweigh the risks. So let’s take a closer look at the story of steroids.

Steroids in PNA Studies

Dr. Greg Katz: If you go back through the last several decades, there have been a ton studies investigating whether steroids reduce mortality in pneumonia. And spoiler alert – the results absolutely all over the place.

Dr. Suellen Li: You know guys, I read through so many of these studies and thought it would be good to pause and highlight some of the major positive and negative trials before we dive into CAPE COD. And show the story of how the pendulum has really swung in both directions. 

Dr. Clem Lee: Yeah. I think that’s a great idea. And as a heads up, these studies were so heterogenous. You’ll hear that different types of steroids were used, given in different ways, and measured using different outcomes.

Dr. Clem Lee: So, the first study that put steroids on the map was an RCT in 2005. It was a small study in ICU patients and randomized 46 patients with severe pneumonia to hydrocortisone infusion or placebo. The patients who got steroids had better vent mechanics, delayed septic shock, and less multi-organ dysfunction. 

Dr. Greg Katz: It’s a tiny sample size, but that study’s secondary endpoint was a shocker. 0% mortality in the hydrocortisone group versus 38% mortality with the placebo group after 60 days. 

Dr. Suellen Li: Yeah, Greg. I was pretty surprised too to see mortality rates of 0% and 38%. That’s so different from what I remember reading that the mortality rate for patients hospitalized with CAP is generally about 13%.

Dr. Greg Katz: Yeah! And also, a 0% mortality in ICU patients with pneumonia?? If it seems too good to be true, then it probably is. 

Dr. Clem Lee: Yeah, just like us, others were also skeptical about these results for a variety of reasons, including major concerns about how the trial was conducted, and the multiple comparisons without statistical adjustment. But the results inspired many other researchers to start studying steroids in pneumonia. 

Dr. Greg Katz: One of the other positive ones was a 2013 study out of Egypt that randomized 80 patients with CAP to IV hydrocortisone bolus plus infusion compared to placebo, which also found a treat benefit with steroids.

Dr. Clem Lee: Yeah and around the same timeline, there were also a bunch of negative RCTs on steroids, right?

Dr. Suellen Li: Right. There was a 2011 study in Lancet which took 304 patients with CAP and randomized them to to IV dexamethasone versus placebo, and found no mortality difference. 

Dr. Greg Katz: After that, the 2015 STEP Trial, which randomized 785 patients with CAP to prednisone versus placebo. The STEP investigators found a shortened time to clinical stability, but no difference at all in mortality rate. 

Dr. Suellen Li: And finally, another recent study was the ESCAPe trial, which randomized 586 patients with severe CAP to IV methylprednisolone versus placebo. This trial fell short of its target enrollment of over 1400 patients, but among those who were included, it found no significant difference in mortality or secondary outcomes. 

Dr. Clem Lee: We can zoom in on the ESCAPe trial for just a little bit. It is an important trial because it is one of the largest trials in this area and the most recent before CAPE COD. However, one of the criticisms of the trial was not only that it was underpowered, but that it was done in the VA system, which has mostly older and male patients, so it might be hard to generalize. 

Dr. Suellen Li: And it seems like a common problem with a lot of these studies was a small sample size. If you don’t enough patients, you’re not powered to detect a difference in mortality, especially if that difference is small. So wouldn’t a meta-analysis be the way to resolve this?

Dr. Greg Katz: Theoretically, I do get what you’re saying, Sullen. Combining data from multiple trials, you are often more likely to detect a difference if there really is one. But in this case the meta-analyses just adds to the pendulum swings. 

Dr. Clem Lee: Yep, and along that train of thought, Annals published a meta-analysis in 2015 concluding that “steroids may reduce mortality by approximately 3%.” Then, there were two more meta-analyses that came out in 2017 and 2018 looking at steroids in pneumonia. Long story short, one found a mortality benefit and the other didn’t. 

Dr. Suellen Li: Wow, and to add to the confusion, not only have the data been mixed, but even professional societies had conflicting guidance. When I compared the American guidelines with the European guidelines on CAP. The Europeans one recommended steroids and the American ones did not recommend steroids. 

Dr. Greg Katz: So when even the major professional societies don’t agree on what to do, it leaves us as clinicians in a space with a little bit of a lack of clarity.

Dr. Suellen Li: Yeah! And beyond that, the studies all had different steroid protocols. Some had IV infusion, some were oral, some started steroids within 24 hours, and others had delayed initiation. And they all used different drug doses.

Dr. Clem Lee: And not to bring up yet another issue, but all of these trials used different inclusion and exclusion criteria for CAP. So the patients in one study may not have been as sick as in another, which might influence the potential benefits and harms. 

Dr. Suellen Li: Right! In practice, we’ve all had patients that when they first present, it can be hard to tell if they have pneumonia or some other mimic. Studies that used looser inclusion criteria might have included some patients that didn’t actually have pneumonia, but had aspiration pneumonitis or heart failure instead. 

Dr. Greg Katz: Whenever I run into a situation like there,  where some trials are positive and some trials are neutral or negative, it makes me suspect that there is a particular subset of patients that benefits more from the treatment, and that some of the trials are just better at capturing that subgroup than others ones. 

Dr. Clem Lee: Yeah! In thinking about what you just said, Greg, maybe the signal for benefit in the studies we talked about was in patients who were sicker, or more inflamed. 

Dr. Greg Katz:  And that’s where the CAPE COD trial comes in! CAPE COD looked at a sicker population with more severe pneumonia and used hydrocortisone infusion to treat them.

CAPE COD Trial

Dr. Suellen Li: So to remind everyone, CAPE COD was the largest RCT to date on steroids in CAP. It took 800 ICU patients with severe CAP and randomized them to IV hydrocortisone vs placebo. 

Dr. Greg Katz: And really importantly to emphasize, the CAPE COD investigators were really strict in deciding who met criteria for severe CAP.  We’ve all certainly overcalled a pneumonia when we were actually just seeing heart failure. It was striking how precise they were in getting rid of common confounders like aspiration pneumonitis or volume overload when they were deciding who actually had pneumonia.

Dr. Suellen Li: They also excluded patients who had septic shock, significant baseline steroid use, or influenza. 

Dr. Clem Lee: Right, and it makes sense to exclude those patients from the study because we know from prior evidence that steroids cause harm in influenza.

Dr. Suellen Li: Exactly! And they excluded septic shock because the reason why steroids are used in refractory shock is a different mechanism. In septic shock, you have relative adrenal insufficiency so giving steroids helps with that.

Dr. Greg Katz: Exactly, the CAPE COD investigators wanted to answer the question – do steroids help in pneumonia? Not the question, do steroids help in shock? 

Dr. Clem Lee: Okay! So now that we know all the patients that they excluded, how did they define the severe CAP in the patients that were included – and specifically what did they mean by “severe”?

Dr. Greg Katz: They defined severe CAP as either patients who were mechanically ventilated with a PEEP over 5, patients who had a P:F ratio under 300 on high-flow or non-rebreather, or patients in the highest pulmonary severity index group, which is a validated risk score for CAP. 

Dr. Clem Lee: Yeah! It’s important to mention that these criteria are different from the IDSA criteria for severe CAP, which take into account other factors such as tachypnea, confusion, and uremia.

Dr. Suellen Li: And keep in mind, this is a fairly sick patient population. About half of the patients in CAPE COD were on mechanical ventilation and on average, these patients were admitted to the ICU within about 5 hours of hospital admission. So not your run-of-the-mill patient coming in for CAP. 

Dr. Clem Lee: Wow, Suellen! Did you say 5 hours? Cause sometimes I can’t even get admissions orders done in in that time. So, when you look at the methods a bit more, they screened 6,000 patients and only enrolled 800. So, why did so few patients make it past screening?

Dr. Suellen Li: Well, this study took a fairly rigorous approach to enrolling patients. We talked about their strict inclusion and exclusion criteria.

Dr. Greg Katz: Whenever a large number of patients don’t make it through the screening protocol, it sometimes makes me skeptical about the generalizability of a study. But when you look in detail at the way that the CAPE COD investigators applied the screening to figure out who actually had pneumonia, I me feel way more confident that these results truly are applicable to patients with CAP.

Dr. Clem Lee: What do we think about the method of steroid delivery that they chose? To remind ourselves, they gave hydrocortisone as an IV infusion for 4 or 7 days with a taper of 8 or 14 days based on improvement on day 4. That seems kind of arbitrary to me?

Dr. Greg Katz: Yeah! It’s totally arbitrary, but so is ever steroid protocol in ever single one of these trials. There is no standard way of doing steroid dosing in these cases and we talked about all of these studies, every single trial has it’s own unique steroid adjustment.

Dr. Clem Lee: And I think we can agree though that having a continuous infusion for up to 14 days is not what I would call the most practical approach.

Dr. Suellen Li: Totally! And another thing to point out is that the CAPE COD investigators administered steroids within 24 hours of onset of any of the severity criteria.

Dr. Clem Lee: That time to steroid initiation was much faster compared to some other studies. The ESCAPe trial, for example, allowed patients to be randomized to steroid treatment up to 96 hours after hospital admission.

Dr. Greg Katz: So, if we think that early administration of steroids matters, which I’m not actually sure that I do, the CAPE COD study would be better positioned to detect it. 

Dr. Clem Lee: Yeah, and one other thing. They also didn’t allow steroids in the control group, and I think that’s worth noting as we interpret the results.

Dr. Suellen Li: Alright, before we get too into the weeds about the methods, let’s talk about the results! The big headline was that there was significantly lower all-cause mortality at 28 days in the hydrocortisone group. 6.2% compared to 11.9% mortality in the placebo group. 

Dr. Clem Lee: They also found lower intubation rates in the steroid group (18% vs 29%) and lower rates of new vasopressor initiation (15% vs 25%), and both were pre-specified secondary endpoints.

Dr. Greg Katz: So, a mortality risk cut almost halved in the steroid group – that’s huge! And it’s perhaps a little bit surprising when you consider how many negative studies and how many neutral studies there are about steroids in pneumonia.

Dr. Suellen Li: And wait a minute, those mortality rates also sound lower than what you would expect for ICU patients with severe CAP. 

Dr. Clem Lee: Right! And the researchers themselves had actually anticipated a baseline mortality of 27% when they were doing their own power calculations. 

Dr. Suellen Li: I think even the researchers themselves were probably surprised by how big of a mortality benefit they found. And the size of that mortality difference made some people skeptical. 

Dr. Greg Katz: Yeah, that’s totally fair, but keep in mind that the investigators under enrolled their trial compared to what they were expecting to get. They initially planned on 1200 patients, but ultimately only recruited 800 because of the pandemic. And even though the study did not enroll that target sample size, and as a result the power of the study was smaller than planned, at the end of the day, the result was statistically significant. And that’s really the only thing that matters here.

Discussion of CAPE COD Trial

Dr. Clem Lee:…So what did people chalk the mortality benefit up to? Was it the hydrocortisone that they used, or was it that they started steroids early?

Dr. Suellen Li: Yeah. There is some thought it could be the type of steroid. And there was a new meta-analysis that came out in 2023 that included CAPE COD. It found that hydrocortisone was associated with more benefit compared to other steroids in patients with severe CAP.

Dr. Clem Lee: I guess I am confused about that because hydrocortisone does have more mineralocorticoid activity, but I’ve heard some ICU doctors have doubts about this being beneficial. Like why would it be good to have ICU patients retain more salt and fluid? They are sometimes already up 13 liters in their ICU stay.

Dr. Greg Katz: The cardiologist in me hears about patients being up to 13 liters of fluid and I can’t put in a lasix order fast enough.

Dr. Suellen Li: Honestly I’m not sure what the reasoning is for the signal with hydrocortisone. And the other thing people did point out, and there was even a letter to the editor written about this, was that there was a big difference in the rates of shock between the two groups. It was 15% in the hydrocortisone and 25% in the placebo group. And remember that steroids were not allowed in the placebo group.

Dr. Clem Lee: So maybe the intervention group did better not because the steroids were treating the pneumonia, but maybe because the steroids were treating the septic shock? I mean I guess, this could be the chicken or the egg though. Maybe the intervention group didn’t develop as much shock because they were on steroids. 

Dr. Suellen Li: Yeah! So it looks like there are some reasons to be skeptical here but it’s hard to explain away that 6% absolute risk reduction in mortality.

Dr. Clem Lee: Right, it seems like there’s a pretty convincing signal for benefit in the study population. What about the safety outcomes? 

Dr. Greg Katz: In terms of safety, the good news that the investigators did not see a major difference in the rate of GI bleeding or hospital-acquired infections. The bad news, is that the steroid group, sort of expectedly, did have higher blood glucoses and higher insulin doses. 

Dr. Suellen Li: Some side effects that CAPE COD did NOT report are the ones that unfortunately are most memorable to patients who stay in ICU. And these are PTSD, ICU myopathy, and delirium. So it would have been helpful to see those side effect data for CAPE COD.

Dr. Clem Lee: Yeah, I agree. Greg, you mentioned earlier about the idea that particular subsets of patients might benefit more from steroids than others. Did we see any signal of that in the study?

Dr. Greg Katz: The subgroup analyses in this study were really interesting but I’d need to remind everyone that subgroup analyses are hypothesis-generating, not hypothesis-testing.

Dr. Suellen Li: Right, remember that because of the pandemic, the trial didn’t enroll enough patients to meet its original sample size goals, so it was even more underpowered to look for subgroup effects. 

Dr. Clem Lee: I agree. With those caveats in mind, what were some of the interesting subgroup findings?

Dr. Greg Katz: Well, the most interesting subgroup to me is the patients who had higher CRP levels.

Dr. Suellen Li: Okay. Let’s unpack that,  patients with a CRP level greater than 15 mg/dL had a greater mortality benefit from steroids than those with a lower CRP. 

Dr. Clem Lee: Wait Suellen, there were patients in the ICU with CRP levels less than 15?? I find that to be very shock-ing! Heh! Get it?

Dr. Suellen Li: Clem, I think those jokes are the most shocking.

Dr. Greg Katz: It certainly makes intuitive sense that patients with more inflammation might have more benefit from steroids, which at their core are anti-inflammatory medication. 

Dr. Suellen Li: And then the last point worth discussing is that about 20% of patients in the study had strep pneumo. And we know steroids have a benefit in strep pneumo meningitis, so do they also help strep pneumonia? 

Dr. Greg Katz: Yeah. You can look at the pathogen breakdown in CAPE COD, where strep pneumo was the most commonly identified pathogen.

Dr. Clem Lee: Right, but then again, 45% of patients had no pathogen identified and those patients actually did better in the subgroup analysis. So maybe we shouldn’t be putting too much stock into which bug was identified. And in my experience, we frequently don’t identify a bug, and even when we do, it’s not often within the tight time frame for starting hydrocortisone, as was done in this trial. 

Dr. Suellen Li: Alright! So if let’s recap some of these discussion points (1) the placebo group didn’t get steroids and had more shock, so were the steroids just better at treating shock in the intervention group? (2) Patients with an elevated CRP showed more benefit. (3) There doesn’t seem to be any interaction between the identified pathogen and the effect of hydrocortisone.

Dr. Greg Katz: So these are all interesting theoretical ways to break down the study in more detail, but at the end of the day these are all just interesting points rather than solid takeaways that I’ll have from CAPE COD.

Takeaways/Discussion

Dr. Clem Lee: So now it boils down to this big question. Should the CAPE COD trial change your practice?

Dr. Suellen Li: Well, we talked to a bunch of people and it seems like there are two schools of thought. One is that, once patients get this level of sick with pneumonia, they also develop ARDS and septic shock so it gets hard to tease apart what we’re treating with the steroids. And we already use steroids for septic shock and ARDS. 

Dr. Greg Katz: Yeah, that school of thought is the school of therapeutic nihilism. It’s the argument that any patient who has pneumonia that’s severe enough to be included in this study also has some other indication for steroids. So, if all of these patients had ARDS, then the therapeutic nihilist says that this changes nothing, doesn’t teach us anything, and frankly, let’s just stop talking about this clinicial trial?

Dr. Suellen Li: The other school of thought is that maybe there’s the patient who has fallen through the cracks before. And this is the patient who has severe PNA but doesn’t COPD or ARDS or shock, but is still hypoxic and just really sick despite being on the right antibiotics. And that’s the kind of patient you should be reaching to treat with hydrocortisone, when you wouldn’t have before.

Dr. Clem Lee: Yeah. After some reflection, as a hospitalist, I think ascribe to the second school of thought. If I see a patient that is meeting the criteria for severe CAP how they defined it and could potentially end up in the ICU soon, I think I’m more likely to reach for hydrocortisone after reading this study than I would have before. 

Dr. Greg Katz: When I was reflecting on this trial and also talking to pulmonologists and ICU doctors, I was surprised by how many people brought up the RECOVERY Study, which looked at dexamethasone in severe COVID pneumonia, I was surprised because it’s not a CAP trial – and RECOVERY patients were generally sicker than those in CAPE COD. But some of these really smart pulmonologists and ICU doctors kept coming back to that trial as proof of concept that there’s probably a subset of patients who have pneumonia where the inflammation driving hypoxia is what is making them the most sick. And that subset of patients who benefits from treating the inflammation directly rather than just treating the underlying disease and providing high quality supportive care. While no single trial fully changes your practice unless you’re more gullible than you should be, different trials can push you in different directors for patients on the margin. And so for CAPE COD, its a trial that is going to push me to give steroids for that hypoxic patient that’s not doing well despite appropriate therapy.

Dr. Clem Lee: Yeah! I think there’s certainly a lot that has to be sorted out in future studies – the ideal method of delivery, whether or not a taper is needed, and which steroid is best. We have to keep these details in mind when we think about applying these results because we can’t necessarily extrapolate these findings to different treatment protocols.

Dr. Greg Katz: Exactly and before we get to the finish line, let’s recap the journey we’ve taken so far. 

Dr. Suellen Li: So first, we dove into the pathophysiology of how steroids may reduce inflammatory response in CAP.

Dr. Clem Lee: Then, we dug into the decades of previous studies that were, for the most part, small, under-power, and with mixed results and methodologies. There will be a one-page infographic of the key studies and, as always, details will be in our show notes.

Dr. Greg Katz: And then, finally, we went through the specific details of CAPE COD, and discussed how it’s findings may make us reconsider steroids on those patients with hypoxia and community-acquired pneumonia who we’re sort of on the fence about what the right next step is. 

Dr. Suellen Li: Right! I think the takeaway here is that these results are exciting but that we should be very judicious about who we apply them to. These were sick ICU patients with severe pneumonia – very few of the patients I see as a hospitalist would have even been eligible for the CAPE COD study. 

Dr. Clem Lee: Hey, Suellen! At least we’ve come a long way since the days of Hippocrates, when pneumonia was treated with hot baths and a blend of honey, milk, and vinegar.

Dr. Suellen Li: You know Clem, minus the vinegar, that doesn’t sound too bad to me!

Dr. Greg Katz: So there you have it! The third edition of Beyond Journal Club with NEJM Group 

Dr. Clem Lee: If you found this episode helpful, please share with your team and colleagues and give it a rating on Apple podcasts or whatever podcast app you use! It really does help people find us! 

Dr. Suellen Li: And if you have any feedback, please email us at hello@coreimpodcast.com.  Opinions expressed are our own and do not represent the opinions of any affiliated institutions.

References


Tags: , , , ,