Is hydrocortisone, vitamin C and thiamine therapy effective for sepsis?

Welcome back to Depth of Anesthesia. This is a podcast that critically explains I'm David Hao, and I'm an anesthesia resident at the Massachusetts General Hospital. This week's episode will be featured on Anesthesiology News, one of the most widely read resources in our specialty. If you haven't checked it out yet, take a look at our website, depth of anesthesia. com. We've collaborated with the Reed platform by QXMD to allow you to easily access all of the studies we discuss in each episode.

If you enjoy the work we do here at Depth of Anesthesia, consider joining our team. We're looking for people who are passionate about questioning dogma and sharing our findings with an ever-growing audience. We're specifically looking for episode writers and editors. Send me an email at depth of anesthesia at gmail. com if you want to learn more. My guest today is Dr. Charnin.

Dr. Charnin is an assistant professor of anesthesiology in the Division of Multi-Specialty Anesthesia and Critical Care Medicine at the Mayo Clinic. He completed both his critical care fellowship and anesthesia residency here at the Massachusetts General Hospital. Doctor Charnin, welcome to the show. Thanks, David. It's great to be on your show.

It's always great to connect with uh fellow MGHers. I know that we also have one more level of connection in that Dr. Daniel Saddawi-Konefka, who listeners may recognize from prior episodes, was one of your mentees. It's true, although it doesn't take long for that learner to surpass the master. Dr. Charnin, just for our listeners to get to know you a little bit more, how do you spend most of your clinical time at Mayo? I would say that I'm doing much more than half of my time doing critical care.

My home unit is the multi-specialty ICU that does vascular, thoracic, other surgical ICU work. But then I also rotate other uh ICUs throughout Mayo Rochester. In the OR, uh I'm still a hound for a great case. The multi-hallway has ET uh all kinds of general anesthesia cases, some urgent surgery, lap coley stuff, uh big urology cases. But then I also enjoy floating to other hallways as necessary to pick up a vascular case or another good case down the hall.

Sounds like they're keeping you busy. I really enjoy the practice. Uh it's been good for me. I know you presented very recently on this upcoming topic at the Year in Critical Care Session at the ASA Annual Meeting. How'd you get interested in exploring the subject?

Well, I've been thrilled to partner with uh Avery Tung and Mike Viero, both members of the Society for Critical Care Anesthesiologists If you're at all interested in critical care and anesthesia, I highly recommend a SOCA membership. SOCA has partnered with the ASA for several years now to do a critical care literature year and review. And the format changes from year to year. Sometimes we take a smaller number of articles and dive really deep. Sometimes we take a larger number of articles and skim the surface a little bit more This year was a shorter time frame because ASA was all online and so uh we really pared down the number of articles we were looking at.

But we try and bring forward articles that are relevant to essential critical care literature, but also relevant to practicing anesthesiologists who are trying to keep abreast of what's going on in the critical care news. So our case today is adapted from Dr. Charnin's audience response question at the session. A 68-year-old man with a recently diagnosed kidney stone is scheduled for a laser lithotripsy. A week before the procedure, however, he presents the ED with worsening flank pain, fevers, chills, and confusion. His lab work is notable for a lactic acidosis and he's booked for an urgent decompressing ureteral stent.

Urine microscopy shows gram-negative rods and he started on antibiotics. After stent placement, however, he's hypotensive and hypoxemic, requiring low dose norepinephrine and mechanical ventilation. As part of the sepsis care for the patient, the intensivist administers hydrocortisone, vitamin C, and thiamen. Here are two questions now for our listeners to think about. What's the level of your agreement with the claim that hydrocortisone, vitamin C, or thiamine are effective for treatment of sepsis?

And what is the level of evidence for what you believe? Dr. Charnin, you asked the audience whether they would give hydrocortisone, vitamin C, thiamine. all three or none of the above as part of the sepsis care of the patient. Do you remember what the results were? By far and away the majority said they wouldn't do any of those at the time. Only about 16% said that they would give the uh hydrochorazone, vitamin C, or ascorbate and thiam, sometimes referred to as hat therapy.

Yeah, and I'd say that's pretty consistent with my anecdotal experience. I've personally administered the therapy only about once or twice in residency so far. One of the seminal studies on this subject was published in JAMA in February of this year. What can you tell us about this trial Well I'd like to introduce you to the vitamins trial. It's uh definitely a randomized trial that deserves to be considered It is one of the top five JAMA articles in terms of looks on the website that does not relate to COVID.

So if you're not looking at COVID and you're looking at JAMA, you're probably looking at this article. And so I think it deserves a closer look So for an overview of the vitamins trial, it's a randomized, open label, multi-center trial that enrolled 216 septic patients who had at least two hours of a vasopressor requirement within the first 24 hours of a diagnosis of sepsis. It compared giving IV vitamin C, hydrocortisone, and thiamine in the intervention group to hydrocortisone alone in the control group. It did not demonstrate a more rapid resolution of shock in the intervention group. Dr. Charnin, what's the background behind this trial?

I seem to recall there was a particularly interesting study published prior to this as well. Yeah, the background is really interesting. Uh Dr. Paul Merrick and his colleagues uh in Virginia published in 2017. a before and after trial looking at hydrocortisone, vitamin C, and thiamine. And I think this is really where the moniker hat therapy was born. Their group came up with the idea of using this therapy for patients who required septic shock, and they all gathered together and decided to apply the same therapy in a similar way And therefore they could build a before and after trial, which was published and quite splashy.

The thing that is most notable from the trial is their figure one Where the predicted mortality and the observed mortality in the first 47 patients had a mortality around 42%, both in the predicted and the observed And then after they moved to HAT therapy, the treatment group had a similar predicted mortality of around 41%, but the observed mortality was only 8%. And it was such a dramatic reduction in mortality that many people joined the hat bandwagon immediately and many different trials have been spawned to look at whether or not this therapy is as beneficial in other hands as it has been in Merrick's groups. I remember that those results were the subject of many conversations on rounds in our ICUs.

I'm aware that severe vitamin C and thymine deficiency can cause diseases like scurvy and beriberi. uh and that in general relative deficiencies of these vitamins have been reported in septic patients. Is there any biological plausibility for, say, vitamin C? Vitamin C has several things that lend it biological plausibility in sepsis. It has always been well known to have antioxidant properties. It has certain anti-inflammatory properties And one of the things that is really attractive to the idea of vitamin C being helpful for patients in septis, particularly with septic shock. is that the body can't make norepinephrine without vitamin C as a cofactor to allow endogenous production of norepinephrine.

So if your vitamin C levels are low, it's hard to imagine that your body's going to do a good job of making norepinephrine. There are certain endothelial proteins that may have oxidative stress during the time of sepsis, and it's thought that vitamin C might help alleviate that. And there are endothelial receptors that help to regulate vascular tone that can be oxidized It's thought that the corticosteroids aren't able to do the job they want to do in tightening the vascular tone if the vitamin C can't help keep the corticosteroid receptors from being oxidized in sepsis. And what about thiamine? What's the biological plausibility for that?

Well thiamine is an important cofactor for three different enzymes, pyruvate dehydrogenase, alpha-ketoglutarate dehydrogenase, and transketalase. If you're like me, you're probably having an incomplete flashback to uh the metabolic section of your medical learning. So let me just say that thiamine pyrophosphate is the compound that crosses the mitochondrial membrane to allow pyruvate and the cytosol to undergo oxidative metabolism in the Krebs cycle. And if your pyruvate can't bind on to thiamine, it can't get in for oxidative metabolism. What happens then is that pyruvate that can't get into the mitochondria to be metabolized is converted to lactic acid.

There are other steps in the Krebs cycle where thiamine is also important. Further, transketolase requires thiamine to metabolize five carbon sugars And that is crucial for the production of NADPH. So a lot of the cellular machinery that provides oxidative metabolism doesn't work if your thiamine levels are low. And we know that in critically ill patients, thiamine levels can be very low. I was getting some nightmarish flashbacks there.

Tough times. Yeah. So it sounds like there's biological plausibility for both thiamen and vitamin C. Is there any downside to just giving these vitamins? One word of caution about vitamin C is that vitamin C is metabolized in the body to oxalate, and oxalate can cause kidney stones and sometimes full renal injury through oxalate crystals forming in the tubules.

Part of why vitamin C and thiamine are sometimes combined together is that thiamine helps to create another metabolic pathway in forming glyoxalate, which is less likely to form crystals. So before you start to give high dose vitamin C, please do think about the kidneys. Heading back to the vitamins trial itself, I know we're gonna get a little bit into the nitty-gritty here. How was the study designed I think it's a study that was well designed. It was randomized, and they did permuted block randomization in groups of two, four, and six, depending upon the size of the center And the overall ratio was a one-to-one ratio of intervention group to control group.

It was a multi-center trial. Ten different ICUs, eight ICUs in Australia, one in New Zealand and one in Brazil, lending it some international credibility. It was an open label trial, so all personnel were aware of the treatment allocation It was conducted in a short window of time. It was nice that it was conducted from May 2018 to June 2019. So we can assume that there wasn't a drift in the usual care of patients during that time.

So far so good. Was there anything notable about the study population that was enrolled? Well, the study population was actually sick. They were patients admitted to an ICU with a primary diagnosis of septic shock, and they were screened using sepsis 3 consensus criteria. The patients needed at least two points above their usual score on the SOFA score, a lactate level greater than two documented or suspected infection.

And so these weren't patients who might be sick. These were patients who actually needed to be in an ICU and were quite sick. And what were the interventions in each group? The intervention group received high-dose vitamin C, a total of 6 grams a day, given as 1.5 grams IV every six hours Thiamine, 400 mg a day, given as 200 mg IV every 12 hours, and hydrocortisone, 200 mg a day, given as 50 mg every 6 hours. The control group received hydrocortisone, a total of 200 mg a day, given as 50 mg every six hours.

In both groups, they got treatment until the septic shock resolved, the patient left the ICU, they developed a contraindication, the patient died, ten days of therapy had elapsed, or there was a serious adverse event. I want to point out what I think is an important aspect of this trial. The intervention group wasn't tested against no therapy. It was tested against a control group who actually got hydrocortisone. The control group definitely gets hydrocortisone.

So this isn't a trial of hat therapy versus nothing. This is a trial of hat therapy versus hydrocortisone What's nice is the hydrocortisone that was given is very similar to what is commonly given for stress dose steroids or rescue steroids for a patient who's ill I see. So it was at least a reasonable dose that you might see in clinical practice. Now I understand that the primary outcome was time alive and free of vasopressors on day seven after randomization. There were also a ton of secondary outcomes, right?

That's right. The authors set out to figure whether this therapy reduced the amount of vasopressors and the number of hours that the patients were on the vasopressors. And so that's why the primary outcome was time alive and free of vasopressors on day seven. And it's worth mentioning that if they came off of their vasopressors for a period of four hours and they continued to live, All the rest of the time was considered time alive and off vasopressors, even if subsequently they had a relapse and went on vasopressors. They also looked at a number of secondary outcomes, and there were two groups of secondary outcomes.

Feasibility outcomes, how easy is it to give these medications? How easy it is is it to screen patients, how many patients do you have to screen to enroll patients? And then there are a bunch of patient-centered outcomes, which included ICU mortality, hospital mortality, 28 90 day mortality, hospital length of stay, so on and so forth. All of these secondary outcomes uh seem very reasonable when you go through the list, and I found none of them particularly remarkable. And how did the authors go about analyzing the data in this trial?

Let's talk with the sample size calculation. I told you that Merrick's trial had a huge reduction in mortality, 38% down to 8% And so uh the trialists listening to your podcast will go, that's fantastic. We won't need a large N to figure out whether there's an effect. And what this group did was they started with a power analysis, only 50% of that scene in Merrick's trial. And they calculated that they would need a sample size of 126 patients to confirm or refute what Merrick found Then they pre-specified that they wanted to take another look at the power calculation and said that they would recalculate the power after the first 60 patients.

They did that and find a sample size calculation brought them to 180 patients. They enrolled 216 because there was the possibility for withdrawal of consent, so they wanted to make sure that they were powered for their primary outcome. The primary outcome was analyzed with a Wilcoxon rank sum test and presented using the Hodges-Layman estimator of the median and pair differences with 95% confidence intervals. The patient survival time was analyzed with the Cox proportional hazards regression and presented with a Kaplan-Meyer curve. Got it.

So after the initial 60 patients, the sample size calculation was performed again, and that's how the final enrollment got from the initial 126 to the final 180. That's right. And then they added a buffer zone on top of that. Before we get into the results of the study, was there anything particularly remarkable about the baseline participant characteristics? If you're used to talking about a trial in terms of its table one, I will say that the table one or the comparison between the two groups, in general, they were well matched.

They looked at many, many different variables, and so you'll find some differences. The intervention group had a lower Apache 3 score, but had a higher lactate and a higher white blood cell count There was a difference between the two groups in the terms of millrenown use. Millrenone use was used infrequently overall. But six people used it in the intervention group and only two people in the control group, which created a statistical difference. But I don't think that between the intervention and control group, there are large, meaningful differences in terms of the groups being imbalanced.

So it looks like a total of 786 patients were screened and 216 were randomized. Of the 216, five withdrew consent, so a total of 211 patients were analyzed, which is more than they were hoping for. What did the study authors actually find? In terms of the primary outcome, time alive and free of baser pressors, they found no difference. And it wasn't statistically close to trending in terms of a difference.

They found no difference. The intervention group had 122 hours. The control group had 124 hours. P-value of 0.83 I think we can call that no difference. In terms of the secondary outcomes, there was a statistical difference in one of them, which was change in SOFA score Where the intervention group had a minus two change in the sofa score, and the control group had a minus four difference change in the sofa score.

But I don't know that this is a clinically meaningful difference And because they looked at so many different secondary outcomes, we would expect to find something pop up as statistically significant. When you look at the amount of vasopressors that were used between the intervention and control group, there's a nice figure three that the authors present. which really show that there was not really a significant trend, even as they go out to day 10 for one of them being moving in a different direction than the other, they seem to be very well balanced between the two. In terms of survival, there was no difference in survival. When you look at figure two, the Kaplan-Meyer curve, it does look at first like the intervention and control group seem to separate But when you look closer, you can find that the confidence interval crosses one, suggesting that there is no statistical difference between the survival and the intervention and control group.

So from what you're saying, it sounds like the vitamins trial did not demonstrate an improvement in the time alive and off-vasopressors for hat therapy over hydrocortisone alone. and that the mortality benefit originally reported in 2017 for hat therapy was not supported. I totally agree. I understand you have one more trial that investigates vitamin C. This is the citrus ALI trial in JAM from last year. Yeah, just a quick word on the citrus ALI trial.

Merrick's trial has caused such a great interest in vitamin C and thiamine. I expect we will continue to see more trials coming down the line, so keep looking for more results. Citrus ALI looked at vitamin C in patients with acute lung injury and it chose its primary outcome as a change in SOFA score, and it had pre-specified secondary outcomes for a difference in C-reactor protein or thrombomodulin levels. Citrus ALI looked at 167 patients with ARDS, and they were randomized to 50 milligrams per kilogram of vitamin C every six hours versus placebo to see whether there was a difference. And what I will say is there was no difference in the primary outcome, which is a change in the SOFA scores, and there was no difference in the C-reactive protein or the thrombomodulin levels.

So the biological mechanism that we were assuming vitamin C might have a role, being a immune modulator, didn't appear to have any change in the C-reactive protein or thrombomodulin levels. There was a difference in mortality, but I think we will have to look farther for additional studies to see whether that is a meaningful difference, because the study wasn't powered to look for a difference in mortality So those were two really well-constructed trials and definitely both are thought-provoking. Dr. Charnin, would you be able to summarize some of the evidence that we discussed today? Let's recall that in 2017, Dr. Merrick and others introduced us to the concept of hat therapy, hatocortisone, ascorbic acid, and thiamine for the treatment of septic shock. and their results in a before and after trial were eye-catching.

Now we have a randomized open label trial, the vitamins trial. Done internationally through ten different ICUs, enrolling two hundred and seventeen patients, considering two hundred and eleven, and did not find a difference between hat therapy or hydrocortisone alone. Also, the citrus ALI trial looked at vitamin C what I would call high dose vitamin C for patients with acute lung injury, and didn't find a difference between primary outcome of a change in sulfascore or immune modulatory levels based on the vitamin C. So far in 2020, it hasn't been a great year for vitamins doing well in randomized trials. Here again are the two questions from the start of the show for our listeners to again think about.

What is the level of your agreement with the claim that hydrocortisone, vitamin C, or thiamine are effective for treatment of sepsis? And what is the level of evidence for what you believe in the future? Dr. Charnin, thank you so much for joining us on the show today. It was a real pleasure to chat with a former MGHer. David, so great to chat with you.

Thanks for inviting me. I really appreciate it. Thanks for joining us for this episode today. If you enjoy the show, leave us a five-star rating on Apple Podcasts. And if you really enjoy the show, leave us a review as well.

As always, stay hungry and keep asking questions. To this