#Episode 102- WATERFALL Trial - Fluid resuscitation in Acute Pancreatitis

[MUSIC] >> Podcasting from two beautiful countries in the world, India and Australia. This is E Crit Care Podcast. Podcast about evidence-based critical care medicine helping you to save lives. Here are thoughts on controversies and critical care. And now, here are your hosts, Dr. Swapnipawar from Sydney and Dr. Joe's Chaco from India. Welcome to E Crit Care Podcast. This is our episode number 102. And today I'm joined by my co-host, Dr. Joe's Chaco. So welcome, Dr. Chaco. >> Hi, Sapna, good evening. Things okay in Sydney. >> Yeah, I think they're pretty good. Overall, staffing issues are a bit better now. And then overall, activity seems to be settling down. How's things in India? >> It's been pretty cool, a little bit busy on occasions, but it's something we can handle. >> Yeah, now that's a good news. So as for today's episode, we decided to discuss recently published randomized controlled trial in New England Journal of Medicine around the use of fluid resuscitation in patients with acute pancreatitis. Now, this trial is quite interesting. It is called as waterfall trial. We do know that there is a controversy still exists around how much fluid to use in acute pancreatitis, which fluid to use in acute pancreatitis, especially between mild, moderate, and severe pancreatitis. Do these patients really need IV fluid? Traditional practice, which was out there for last couple of decades was whenever you get a patient with severe acute pancreatitis, they used to get central line, they used to get volus of IV fluid, and then large maintenance fluids. And always the traditional, especially the surgical teaching was keep them full, because there was always a concern around acute pancreatitis has this inflammatory response, which leads to third spacing or fluid losses. And that's why we need to be on the top with our fluid resuscitation strategies. As time has passed by, we did understand the fluid resuscitation paradigm quite better now. And I guess we become quite major in our practices around fluid resuscitation in patients, not only just with acute pancreatitis, but overall critically ill patients who come to the door of ICU. Having said that the ward practices around these illnesses have still remained somehow closer to the traditional practices. So on occasion, we do see some patients who go to the ward with acute pancreatitis, they are still on maintenance fluid of 125 to 250 mil per hour for first few 10 to 12 hours. And the traditional practice was like to finish one the drug backups line initially in two hours, then four hours, then six hours, and eight hours in that kind of way. So this trial is quite useful in that way, because this trial tried to assess the efficacy of aggressive fluid resuscitation in patients with acute pancreatitis. Now, what's the kind of evidence out there before this trial was performed? And there are only few small sites, randomized controlled trials out there, which are provided with very conflicting results. Now, there are two trials involving patients with severe pancreatitis showed that rapid fluid expansion was associated with decreased survival. And that has clearly helped us to change our practices in patients with acute severe acute pancreatitis in in the insecure unit. Now, there was also randomized controlled trial involving patients without baseline systemic inflammatory response syndrome or serves, who therefore had an initially low risk of moderate to severe disease, showed a quicker clinical improvement with rigorous hydration than with the moderate hydration. I guess the overall trial was quite low quality trial. And also, over the period of time, we have moved away from using the terminologies serves and we try not to use this to retain more now. There is also a systematic review, which was published. And also, it's quality is limited by heterogeneity and overall quality of the studies which are included within the systematic review. And it showed a lower incidence of adverse events and lower mortality with moderate hydration than with aggressive hydration. And because of this kind of couple of trials and also this systematic review, people kind of sway towards using more aggressive fluid resistion in patients with acute pancreatitis. So the waterfall trial, which basically is short form for the early way-based aggressive versus non-aggressive, gold-directed fluid resistion in early phase of active pancreatitis. And this is an open-level multi-central randomized controlled trial, which investigated the safety and efficacy of aggressive fluid resistion as compared with the moderate fluid resistion in diver sample of patients with acute pancreatitis with a range of severity of the disease. So Dr. Chako, can you please walk us through the methodology and the results of this trial, please? The waterfall trial, as you mentioned, was a multi-center open-level parallel group randomized controlled trial that enrolled patients from 18 centers in four countries. Most of the patients came from Spain. There were patients from Mexico, Italy, and from a center in India as well. Acute pancreatitis was diagnosed according to the revised Atlanta classification. And for inclusion in the study, patients were within 24 hours after the onset of pain. And eight hours or less since a confirmatory diagnosis of acute pancreatitis. They excluded patients who had moderate or severe disease, which included patients who met the criteria for moderately severe or severe disease at baseline, with organ failure, including shock, respiratory and renal failure. They excluded patients with heart failure, based on the Neochar Association functional class, two, three, or four, patients with uncontrolled hypertension, electric abnormalities for excluded patients who were deemed to have a life expectancy of less than a year excluded. And so were patients with chronic pancreatitis, chronic renal failure, and decommensated cirrhosis. So what all boils down to is they included patients with mild pancreatitis. So there were two arms in the study. The first arm was the aggressive resuscitation strategy in which they used a bolus of lactated ringer solution of 20 ml per kilogram over a period of two hours, which was followed by an infusion of three ml per kilogram per hour. So that's ringer selected solution. In an average 70 kilogram patient, this will come to 1.5 liters bolus, and you continue with 700 approximately 200 ml per hour. So one and a half liters bolus and 200 ml per hour roughly. That's what it comes to in the aggressive resuscitation group. And the other group was the moderate resuscitation group in whom they gave a bolus of 10 ml per kilogram over two hours if they were considered hypervalemic followed by a maintenance infusion of 1.5 ml per kilogram per hour. So that will come down to 700 ml bolus if you take a typical 70 kilogram patient, 700 bolus followed by 100 ml per hour, roughly half the aggressive strategy. Patients in both the groups had evaluation of their fluid status at three hours followed by biochemical and physical assessments at 12 hours, 24 hours, 48 and 72 hours. And once they made this assessment, they made adjustments to the rate of fluid administration and fluid administration was stopped if there was a suspicion of fluid overload. Oral feeds were allowed at 12 hours if the pain intensity was less than five on a scale called the pan promise scale, which is a relatively new scoring system which basis itself on a seven item scale based on symptoms that cause the most discomfort, pain and concern to the patients. So oral feeds allowed at 12 hours based on the pan promise score, which time fluid resuscitation was stopped if oral intake could be tolerated for more than eight hours. But the earliest cessation of IV fluid was 20 hours after in the moderate resuscitation strategy group and 48 hours in the aggressive resuscitation strategy group. So 20 hours 48 hours was a cutoff for cessation of IV fluids completely provided they could take adequate oral intake. How did they calculate the sample size? The calculation was based on an anticipated incidence of progression to moderately severe or severe acute pancreatitis of 35%. So as I mentioned before, these were patients who had mild disease. They assume that 35% among them would progress to moderate or severe disease. And they calculated the sample size of 744 based on this number of 35% 372 patients in each group for 80% power to detect the between group difference of 10 percentage points between the two arms for a two-sided significance or alpha level of 0.05. And they very rightly planned for two interim analysis after 248 and 496 patients. And at the first interim analysis, the study was terminated early due to safety concerns arising out of possible fluid overload with the aggressive strategy. So what did they find? They had 122 patients in the aggressive resuscitation strategy group and 127 patients in the moderate resuscitation strategy group. Their baseline characteristics were well matched except for a higher incidence of gallstone pancreatitis in the aggressive resuscitation arm. Fluid administered during the first 48 hours, obviously much higher with the aggressive strategy, 7.8 liters median at 48 hours in the aggressive group compared to 5.5 liters in the moderate strategy arm. So the volume of fluid administered in the first 48 hours was considerably more as you would expect because the rate of administration was almost twice as much. They diagnosed fluid overload based on fairly reliable criteria. They did not use dubious clinical criteria. They were very clear-cut criteria to diagnose fluid overload. And they went by the clinical symptoms signs as well as radiographic evidence. And they used these criteria to diagnose fluid overload, mild fluid overload if it was responsive to medical therapy or a decrease of hydration. And if the PF ratio never dropped below 300. So that was mild fluid overload. Moderate fluid overload was if it was responsive to medical therapy or decreased hydration, but the PF ratio was less than 300 at least once. Severe fluid overload was if the fluid overload led to mechanical ventilation or requirement for him or filtration. So the outcomes that they looked at, the primary outcome was, as I mentioned, the incidence of moderately severe or severe acute pancreatitis. So these are all patients with mild pancreatitis. Progression to moderately severe or severe acute pancreatitis occurred in 22.1% of patients with agressive strategy compared to 17.3 in the moderate strategy. Although this looks like very different with a fairly high number of patients who progressed to moderate to a severe disease in the agressive strategy group. It was not statistically significant. So it was a case with organ failures. Again, numerically much higher. Aggressive strategy 7.4% compared to 3.9% in the moderate strategy arm. Local complications persisting organ failure. These were all higher, but not statistically significant with the aggressive strategy. Respiratory failure, Necrotizing pancreatitis, need for ICU admission, hospital length of stay, and the pan promised code that I referred to earlier. These were all significantly worse in patients who had an aggressive strategy, but again, not statistically significant. But when they looked at fluid overload, the degree of fluid overload was much higher with the aggressive strategy. 20.5% versus 6.3% in the aggressive versus moderate strategy. This was statistically significant risk ratio of 2.85 and a confidence interval of 1.36 to 5.94 suggesting significance. Fluid overload with symptoms was again higher with aggressive resuscitation. 18 versus 7.9% fluid overload with signs. 26.2 versus 11% moderate to severe overload 4.9 versus 0.8%. So all these points towards a fairly high degree of fluid overload in patients who have go under an aggressive resuscitation strategy. Overall, if you look at their cohort from what I could gather, only one patient required mechanical ventilation. So obviously, these were all patients who had mild disease and some of them showed a significant number of them revealed signs of fluid overload with the aggressive strategy. So that's the bottom line, although the primary outcome of progression to moderately severe pancreatitis was not statistically significant. The organ failures were also worse off with the aggressive strategy, but not statistically significant. There were definite signs of fluid overload in the aggressive strategy. Thanks Dr. Chakur. So I guess this trial is, as I said, interesting. This trial has clearly given us a definitive answer that aggressive fluid resuscitation strategy is very harmful in patients with mild acute pancreatitis. Obviously, this trial has not given us any answer around what should we do with acute severe pancreatitis. Having said that, I guess this trial has clearly challenged the dogma and also probably for future, it will change the practices, especially in the water setting around the use of fluid in acute pancreatitis. This trial obviously comes with lots of limitations. Obviously, it's open-level randomized controlled trial. So the internal value itself questionable. This trial was stopped early due to futility and I guess that clearly makes this trial even more underpowered. Now, there are a lot of things with regards to the design of the trial. Like first and foremost, they compared the aggressive fluid resuscitation strategy versus moderate resuscitation strategy. Now, when you look at the moderate resuscitation strategy, the patients received on an average 5.5 liters fluid in force for 24 to 40 hours, which is still a lot of fluid. Now, if you think about how much fluid aggressive fluid resuscitation arm received, that's a lot. So, perhaps you might question whether they were too aggressive in their approach to view fluid resuscitation in aggressive treatment strategy arm. I perhaps that we definitely need to even question. It would have been much beneficial if the trial had a third arm, which is either a restrictive fluid arm or probably no fluid arm. If they would have compared these three different arms or aggressive fluid resuscitation practices versus no fluid resuscitation practices or minimal fluid resuscitation practices, then obviously we would have had even much better signal to guide our practices in future. That's the major problem. And also authors mandated to give the fluid in either moderate resuscitation volume or aggressive resuscitation volume for 48 hours. Now, often the current practice all over the world is if you really don't need fluid, you don't give fluid. And also, we de-escalate our fluid treatments very quickly nowadays. So, within this trial, there was this kind of protocol or dogma, which is quite surprising to me. Now, the trial was conducted in Mexico, Spain and one hospital in India, which probably limits its external validity and generalizability of the results. I guess having said that the whole hypothesis itself is questionable. And really, do we need to even think about doing another trial where we are going to compare giving a fluid in with a bucket to these patients. So, the answer to that is clearly no. And I guess this trial will definitely change practice in some parts of the world or some developing a world I see used where the these practices are still prevalent within the world settings. But obviously, this probably won't change any major practices in most of the developed countries where the fluid practices have already evolved over the period of time. So, that's the kind of my main takeaways or main analysis of this trial. What's your thoughts? It's interesting that you mentioned that in her possible third arm to the study with no resuscitation like the Feast trial perhaps they could have done that because if you look at their inclusion criteria, these were patients who had no shock, no renal failure, no respiratory failure. And you would ask the question, has to why would they require so much of fluid in excess of 45 liters in either arm? Is that really required? And predictably, they ended up with fluid overload. So, that is pretty much predictable. And I think over the years, we have realized not only in acute pancreatitis but in general in sick patients. If you try to resuscitate them with too much fluid, you're going to probably cause more harm than good. And that is, I think, quite evident from this study as well. So, at least in our practice, patients with essentially mild acute pancreatitis, they would eat and drink if that is possible. And we would give IV fluids only if they are not able to eat and drink adequately to sustain attention because of pain or vomiting. In that case, of course, it's understandable to use intravenous fluids to support them. But otherwise, we wouldn't particularly consider resuscitating them with the bolus and running a high dose infusion in patients with no evidence of organ failure. That's one of the problems. Secondly, these patients were, did not have much comorbidities at baseline. They were all experienced with comorbidities, but they were all included including heart failure, hypertension, electrolyte abnormalities and so on. So, essentially, very stable patients at baseline who had mild disease and the whole question is, why would they require resuscitation continued by a continued intravenous maintenance of up to 200 miles per hour? I agree, and I guess it's similar to Feast trial. This trial is quite useful to guide our practices in future. Now, obviously, with the Feast trial, there were a lot of ethical challenges to even do the trial. Similarly, I'm pretty sure this trial to get through ethics remedy would have been quite challenging because we are trying to give a fluid in abundance to patients who have not hypervolume. So, if you look at when the primary comparison between the two groups at the baseline, only 50% of the patients in each arm had some indication of hypervolumea. That means half of the patients did not have any evidence of hypervolumea and we still end up in giving so much of fluid in this study, which probably makes sense to start with. As you rightly mentioned, these patients, most of the times, they are in a worsening, they will probably commence their oral intake within 12 to 24 hours and then usually they will be just a kept monitored in a worsening, still they get better and then obviously they will discharge without any major interventions as the management is purely supportive. Now, nutrition support in these patients was interesting because as for the table, only seven patients in aggressive arm versus five patients in moderate precision arm received some nutritional support, which again, it makes me wonder that these patients might have looked probably deceptively well, but didn't get fed because they were on IV fluid. So, this probably trial does not reflect the practices that we follow here in Australia and New Zealand, but I get similar to Feast trial as Feast trial help develop being countries and especially African continent to change their practices of not giving Alvin boluses in patients with malaria. Similarly, I think this trial will help to stop these practices if they are still existing in some of the part of the world. In that regard, this trial is really good and has informed our future practice of fluid decision in patients with acute pantheries. The August mention in the introduction at one of the reasons why you might require aggressive fluid decision in pantherietitis even without shock is because of the possible lack of profusion to the pantherias and a higher incidence of pancreatic necrosis. So, that is the putative reason as to why you would require excess fluid, but this study fairly conclusively proves that a aggressive strategy of a substation in acute pantherietitis in my acute pantherietitis would certainly lead to harmful effects from fluid overload. It's a completely different matter when it comes to moderate or severe acute pantherietitis. There, of course, you would treat them in a much more aggressive fashion. Yeah, thanks Dr. Tako. So, as I said, probably it won't change my practices, but as I said, it will be definitely useful for some parts of the world where these patients are still managed by surgeons or less experienced colleagues in the water settings. So, again, this trial is good from those part of you. So, thanks to authors for completing a very good randomized controlled trial. Probably the design could have been a bit better, but at least, as I said, it will help to change the practice around the globe. Thanks, Dr. Tako. Thanks for your time. We'll be back in in four nights' time with another episode. Till then, goodbye and have a nice time. Thanks for no thanks, everyone for listening in. We will talk to you next time. Goodbye. Check us out on Facebook at Critical Care Education. Join us next time for another edition of E-CritCare Podcast.