Suiting up for COVID-19 autopsies, sharing findings

As for the autopsy findings themselves, some of them have surprised the University Medical Center team, and some haven’t. “One of the things we’ve found very characteristic of the COVID patients that was not really reported up until that point was a lot of pulmonary hemorrhage,” Dr. Vander Heide says. “It’s a very consistent finding in all the patients.”

On the other hand, reports from China and Seattle of mysterious heart damage in patients who have died of COVID-19 have not been borne out at University Medical Center. “So far, we have not seen any cardiac deaths,” says Dr. Vander Heide. “The causes of death have all been lung related. However, we’re in the process of doing microscopic analysis and have found cardiac findings that while not a cause of death, we believe are specific to COVID-19 virus infection.” He calls “interesting and surprising” the findings they have submitted to a cardiology journal. “Some of the autopsied COVID-19 patients had some kidney disease during their clinical course and we are finding glomerular lesions that we also believe were produced by SARS-CoV-2,” he says.

Another of the team’s autopsy findings has led to a change in protocol for the entire hospital. “When we started doing these autopsies, we noticed, of course, that they have the typical lung features everybody talks about—diffuse alveolar damage,” Dr. Vander Heide says. “We knew this was happening. And we had some [autopsied] patients who were actually negative on the nasal swab RT-PCR test [for SARS-CoV-2]. So we contacted the chief medical officer of the hospital and indicated we were concerned that the testing being done at that early time point may not have been completely accurate. We told them that if someone looks like they’re COVID-positive in terms of their clinical presentation, they should treat them as COVID-positive.”

Of the testing itself, Dr. Love says, “It’s been a real struggle to obtain the [SARS-CoV-2 virus] testing that we need.” At first, University Medical Center relied on Louisiana State Public Health Laboratory. “Of course they performed magnificently, but soon they were just overcome,” he says. “We finally brought it all in-house.”

Dr. Love and his team had instituted the purchase of a Roche Cobas 6800 system before the pandemic began, as a replacement for a previous system. However, “The week that we were supposed to have the unit arrive on site to be installed, the installation was interrupted,” he says. “It’s a terrific technology; everyone wants it, and the company is doing what it can with a limited resource.”

While Dr. Love worked to get the hospital’s Cobas unit installed at a later date, he worked out a stop-gap solution with Roche and an old friend at Tulane University School of Medicine, which had its own Cobas 6800 but was not testing for SARS-CoV-2 because the reagents were not easily available. After getting Roche to agree to provide reagents, Dr. Love reached out to the interim pathology chair at Tulane—with whom he’d attended college 50 years ago—and proposed a deal: Tulane would allow University Medical Center patient specimens to be tested on its Cobas 6800, while the medical center would provide the necessary Roche reagents so that Tulane could do its own testing, “and we cooperated in validating their unit.” (Roche would not release reagents to Tulane for SARS-CoV-2 virus testing because of the shortage, Dr. Love notes.) That arrangement “bridged us” until the medical center’s Cobas 6800 arrived, at which time Roche expedited installation, he says. The medical center now has not one but three platforms on which to run SARS-CoV-2 virus testing: Cobas 6800, Abbott ID Now, and Cepheid GeneXpert.

“We did not have the Abbott unit” prior to the pandemic, Dr. Love reports. “We had used a Cepheid four-bay model for Mycobacterium tuberculosis identification/rifampin sensitivity and influenza rapid testing in the past. We are obtaining a Cepheid 16-bay model. The Cepheid takes 45 minutes processing time per SARS-CoV-2 specimen and is manually loaded. The Abbott instrument takes five minutes to produce a positive result and 14 minutes to produce a negative result and is also manually loaded.”

He’s not in favor of maintaining multiple systems to perform the same test. “But we are desperate to bring SARS-CoV-2 testing in-house,” he says. “Now the challenge is convincing clinicians that all SARS-CoV-2 testing is not equal.” The sensitivity and specificity of the Roche Cobas 6800 that processes specimens over 3.5 hours is likely better than a unit that produces a result in 15 minutes, he says. “We work with clinicians to help them understand that false-negatives can occur, the main problems being seen with poorly collected swabs containing lower amounts of virus.”

Dr. Love and his colleagues are pleased to be able to perform autopsies of COVID-19 patients, particularly as the medical community worldwide continues to parse the particulars of the virus that has led to this shocking and uncertain time. “I believe autopsies have been undervalued,” he says. “And I hope as we proceed forward, analyzing our cases, that we will be able to contribute to the understanding of this disease.”

Anne Ford is a writer in Evanston, Ill.