After the respiratory consequences of COVID-19 and resulting ventilator shortages, acute kidney injury is emerging as the next healthcare and resource issue in this pandemic.
The prevalence of kidney failure in some ICUs is so high that centers are running out of dialysis machines and even dialysate solution, and they are also struggling to maintain access to renal replacement therapy (RRT) for noninfected patients.
Clinicians are devising creative solutions and workarounds but are also cautioning that acute kidney injury (AKI) is driving up patient deaths and may be a lingering sequela of the novel coronavirus.
“People are kind of shooting from the hip,” said Joel Topf, MD, assistant clinical professor of medicine, Oakland University William Beaumont School of Medicine in Detroit, and co-creator of the nephrology journal club, NephJC. “They’re doing the best they can, but a lot of what we would like to do is curbed by resource utilization.”
COVID-19 patients, he noted, are “tremendously catabolic” with hyperkalemia, hyperphosphatemia, and profound metabolic acidosis to a degree not seen in typical kidney failure patients.
“This ends up driving a lot of dialysis; the patient’s potassium is 7, so we absolutely need dialysis today, can’t wait, and we just don’t see this typically,” he told Medscape Medical News.
“This is acute kidney injury of a different stripe than we’re typically seeing in really sick people,” he commented.
Early COVID-19 reports rarely mentioned AKI or kidney failure. Other reports indicated that SARS-CoV-2 infection did not result in AKI or aggravate existing kidney disease among hospitalized patients in Wuhan, China.
“Many publications completely omit the complications of kidney failure, even in the ICU,” Matthew Sparks, MD, assistant professor of medicine, Duke University, Durham, North Carolina, and cofounder of the AJKD blog, told Medscape Medical News. “This is a very big problem. We’re starting to see more data come out, but it’s still a very slow trickle.”
Publications regarding ICU patients showed that the incidence of AKI varied but was more or less 5% and that 5% were treated with RRT, said Eric Hoste, MD, PhD, intensive care unit, Ghent University Hospital, Belgium, and chair of the AKI section of the European Society of Intensive Care Medicine (ESICM).
Still, this didn’t match with the experiences of many physicians.
“When we use real-world data from friends around the world, we found that moderate and severe AKI ― so AKI stage 2 and 3 ― occurs in 20% to 30%-plus of patients, and 25% to 30% of patients are treated with renal replacement therapy, with variation among centers,” Hoste said during an April 15 ESICM webinar.
Among the 11 centers that contributed data to the new analysis, the incidence of stage 2–3 AKI was 33% in Pittsburgh and 35% in London but varied from 8% in Montreal to 72% at Hoste’s own hospital. RRT use ranged from 0% to 37%, with some centers, especially in the United Kingdom, having 25% to 30% of patients on a filter, he noted.
The findings represent AKI as defined by KDIGO criteria. When only creatinine criteria were used, the incidence at Hoste’s hospital fell from 72% to 4.3%, highlighting one factor that can contribute to variance in the numbers.
Differences in baseline patient characteristics, varying degrees of rhabdomyolysis, and inflammation generated by mechanical ventilation and by acute respiratory distress syndrome itself may play a role, Hoste observed.
There’s also variation in practice, including when RRT is started and in the use of high positive end-expiratory pressure (PEEP), which was only recently shown to be unnecessary in the initial phase of the disease.
Very importantly, guidance on COVID-19, issued just 1 month ago, indicate that COVID-19 patients should be kept dry to protect the lungs from developing too much alveolar edema, Hoste said. But many patients arrive in the ICU dehydrated, probably from fever and gastrointestinal symptoms.
“When you have a dehydrated patient and you keep them dry and apply high levels of PEEP, that’s something the kidney doesn’t like,” he said. “We now use a more intelligent approach and look for signs of volume status. If a patient is dehydrated, especially in the first phase of the disease, we will look for optimization of volume status and not keep them as dry as possible.”
In hospitals in which dialysis machines are in short supply, workarounds include compressing dialysis sessions, possibly sharing machines, and using 12-hour continuous veno-venous hemodialysis (CVVHD) instead of 24-hour dialysis, but with faster blood flows and dialysate flows to improve efficiency, Chirag Parikh, MBBS, PhD, director of nephrology at Johns Hopkins University in Baltimore, Maryland, told Medscape Medical News.
“It’s not optimal, but I think a lot of centers are doing this already, based on what I have heard,” he said.
After fielding calls for dialysate from hard-hit New York hospitals and looking at their own stockpiles, Parikh and colleagues sought an alternative.
“Even if we gave 10% to 20% of our solution, it would last them for a few days and they’d be in the same situation,” he said. “So we decided that there’s a simple solution ― that we can create our own dialysate when the national supplies are short.”
The stopgap measure involves capturing the dialysate produced by inpatient dialysis units by attaching a 3D connector to the dialysis cartridge and using a total parenteral nutrition bag. Roughly 150 L of CVVHD dialysate can be produced over 5 hours. Parikh posted the method in a recent Twitter tutorial.
“We have submitted our protocol to the FDA, and they are giving us some guidance,” said Parikh, who noted that Columbia University, in New York City, and Cornell University, in Ithaca, New York, plan to implement the protocol this week.
Patients with COVID-19 are also very thrombogenic, leading to reports of premature clogging of dialysis filters and the formation of a so-called protein cake.
“This is resulting in the use of anticoagulation, which is another medication you have to use and another possibility of running out,” Sparks noted. “And a lot of these patients are on continuous dialysis.”
If continuous renal replacement therapy (CRRT) is not an option, acute peritoneal dialysis is an “absolutely acceptable solution,” according to Marlies Ostermann, MD, a consultant in critical care and nephrology at Guy’s and St. Thomas Hospital, London, United Kingdom. It provides continuous renal therapy and bypasses the need for a dialysis or CRRT machine as well as any anticoagulation.
That said, fluid removal is less precise with this approach, there’s a “clear risk of peritonitis, and for COVID-19 patients in particular, there’s a risk of acute leak, especially if patients need to be nursed in the prone position,” she said during the ESICM webinar.
Notably, the recent Surviving Sepsis and NICE COVID-19 guidelines contain no comments on RRT, and the very latest National Health Service guideline, published last week, notes shortened dialysis filtration but basically recommends standard renal replacement therapy, Ostermann observed.
“There is no universal plan for renal replacement therapy in COVID-19,” she said. “Circumstances change quickly; preparedness and flexibility are essential.”
Some people have suggested that plasmapheresis may be useful, especially for patients who experience a cytokine storm, noted Oakland University’s Topf.
“But if you’re already short of dialysis nurses and dialysis machines, you may not have the resources to do that kind of experimental therapy,” he observed.
It’s also proving extremely difficult to maintain metabolic control for these patients. Typically, such control is achieved with three dialysis sessions per week. Clinicians are adapting by using lower potassium baths with dialysis.
“The trend with dialysis has been to be less aggressive, and I think that is being reversed 180 degrees with this crisis, and we’re now saying: go ahead and dialyze them on a 1K or 0K bath,” Topf said. “I’m reading about people saying, ‘I’ve never used a 0K bath in my life, but I’m using 0K baths now, just to get their potassium under control.’ “
Specific to COVID-19
There is considerable debate as to how the novel coronavirus is damaging organs and cells, but emerging evidence suggests damage occurs in the kidneys, particularly the podocytes and proximal tubules.
To gain entry into host cells, the spike (S) protein on the outer membrane of the coronavirus binds to angiotensin-converting enzyme 2 (ACE2) and is activated by transmembrane serine proteases (TMPRSSs). Both ACE2 and TMPRSS are prevalent in bronchial endothelial and vascular tissue as well as in the kidneys. A recent analysis demonstrated high coexpression in podocytes and proximal tubule cells.
This distribution in the kidneys is very much concordant with autopsy data regarding 26 COVID-19 patients from Wuhan. That study identified coronavirus particles in proximal tubular epithelium and in the podocytes, ESICM’s Hoste said. Notably, only 9 of the 26 patients showed clinical signs of kidney injury.
Also prominent in COVID-19 patients are elevated levels of ferritin, which may be a marker of ferroptosis. Iron-mediated cell death may contribute to the enormous inflammation seen in these patients and could play a role in why they do worse.
“When ferritin levels are high, there’s active disease and patients get worse, and when [the levels] are going down, patients are most of the time improving,” he said.
A recent article also showed that among patients in whom the ACE1 D-allele is less prevalent, the incidence of COVID-19 cases per million population is higher. In addition, ACE2 and TMPRSS are not evenly distributed among patients of Western origin and Asian origin, Hoste said.
This “may explain why certain patient groups have a higher incidence of severe COVID-19 and a higher incidence of AKI in this disease,” he said.
However, like any ICU patients who have acute kidney injury, COVID-19 patients with AKI have a worse prognosis. In a prospective cohort of 701 patients from Wuhan, the hazard ratio for in-hospital death was 3.5 for stage 2 AKI and 4.8 for stage 3 AKI after controlling for age, severity of illness, and comorbidity.
Call for More Data
Although COVID-19 articles and preprints have been coming at a dizzying pace in recent weeks, all of those interviewed said better data are needed, including biopsy and autopsy information as well as data on race, biomarkers, and long-term outcomes.
“As far as the kidney is concerned, knowing whether patients are black or white is good, but actually knowing whether they’re APO1-positive will go a very long way,” Parikh said. “Because even among African Americans, people who are APO1-positive will get a much more severe course of disease and reduced recovery from kidney injuries.”
It can take the kidneys about 6 weeks to recover, and it will be equally vital in the coming weeks to obtain data on the number of survivors who have residual disease.
“I can tell you, the kidney may be one of the top in terms of long-term sequelae,” Parikh said.
Hoste reports academic research grants for biomarker research; participation in Astute Medical–sponsored research, receiving study materials from Bellco, speaker fees from Alexio, and travel grants from AM Pharma and Bioporto/Sopachem. Ostermann reports speaker honoraria from Fresenius, Mitsubishi Tanabe Pharma, Biomerieux, and Baxter; serving in an advisory role for Gilead, Gambro-Baxter, Fresenius, and Nxstage; and research funding from Fresenius Medical Care, LaJolla Pharma, and Baxter. Topf reports having an ownership stake in a few Davita-run dialysis clinics and a vascular access center and serving on advisory boards for AstraZeneca and Cara Therapeutics. Sparks reports serving on the board of directors for NephJC.org and receiving a grant from the Renal Research Institute.