Latest Kawasaki Series Offers Clues to COVID-19 in Kids as Debate Persists

Investigators in Bergamo province, one of the regions deeply affected by COVID-19 in Italy, say that the number of children who have tested positive for the virus and presented with severe “Kawasaki-like” symptoms amounts to a 30-fold increase over the number of Kawasaki patients seen in prior years.

“Although this complication remains very rare, our study provides further evidence on how the virus may be affecting children,” lead author Lucio Verdoni, MD (Hospital Papa Giovanni XXIII, Bergamo, Italy), said in a press statement. “Parents should follow local medical advice and seek medical attention immediately if their child is unwell. Most children will make a complete recovery if they receive appropriate hospital care.”

Speaking with TCTMD, senior author Lorenzo D'Antiga, MD (Hospital Papa Giovanni XXIII), said that he and his colleagues have continued to track Kawasaki presentations and have now seen 10 additional patients, reaching a total of 20 who meet the criteria for this typically rare diagnosis; half also had cardiac involvement. But it’s very difficult to put these numbers in perspective because the actual prevalence of COVID-19 in the population is unknown, he pointed out.

“As long as we don’t have screening of the entire population, we can’t work out the rate of the complications, including the rate of Kawasaki disease,” D'Antiga said. “However, considering the population of our province, and considering what we gathered from previous studies, we calculated that one child out of 1,000 infected with COVID-19 develops Kawasaki disease.” That figure is based in part on the estimated 10% infection rate for the region, he noted. “This should be reassuring for families. It is not something that a child [automatically] develops if the child gets COVID-19.”

Earlier in the pandemic, children and young adults looked to be largely spared the higher infection rates and greater disease severity seen in older adults. Then public health officials in New York and the United Kingdom began warning physicians about an uptick in cases of “pediatric multisystem inflammatory syndrome” temporally associated with COVID-19 and bearing a number of Kawasaki-disease features. In the Lancet last week, UK investigators described an “unprecedented cluster” of eight children treated over a 10-day period with “hyperinflammatory shock” and characteristics “similar to atypical Kawasaki disease.” Verdoni et al’s Bergamo series appeared in the Lancet yesterday and together the papers add heft to what has become a mounting concern in pandemic hot spots—debate continues over whether this condition should be described as “Kawasaki-like” disease, toxic shock, or something altogether different.

The Bergamo Data

Verdoni and colleagues compared all patients presenting with Kawasaki symptoms between February 18 and April 20, 2020, at a single hospital to those who presented over the 5 years prior to the pandemic. In all, the hospital saw 19 patients between January 1, 2015, and February 17, 2020, and 10 patients between February 18 and April 20, with all of them presenting in the latter four weeks. That’s an incidence of 10 cases per month at the height of the COVID-19 period, as compared to 0.3 cases per month in earlier years. All of the children admitted during the COVID-19 period were tested for active infection (nasal/throat swab plus RT-PCR assay) or previous infection (serology testing), and all were managed as per Kawasaki-disease recommendations, namely intravenous immunoglobulin (IVIG), steroids, and aspirin.

This should be reassuring for families. It is not something that a child [automatically] develops if the child gets COVID-19. Lorenzo D'Antiga

Children diagnosed during the study period were older than patients diagnosed before the pandemic (mean 7.5 years vs 3.0 years) and were more likely to develop macrophage activation syndrome (MAS; 50% vs none), Kawasaki Disease Shock Syndrome (KDSS) characterized by a 20% decrease in systolic blood pressure (50% vs none), and were more likely to have cardiac involvement (six out of 10 patients vs 2 out of 19). All of these difference reached statistical significance.

Cardiac complications included aneurysm, reduced ejection fraction, mitral valve regurgitation, and/or pericardial effusion. None of the typical Kawasaki patients treated during the earlier period had elevated troponin or proBNP levels, whereas these were seen in 55% and 100%, respectively, of the young patients in the COVID-19 period; their mean proBNP levels were 1,255 ng/L.

In terms of COVID-19 positivity, two of the 10 patients tested positive after RT-PCR testing, whereas on subsequent testing, eight of 10 patients were IgG positive and three were also IgM positive. “This finding and the positivity of IgG antibodies suggest a late onset of the disease compared with the primary infection, due to the host immune response,” the authors write. “This might be the reason why, in the past, no active viral infection could be shown in this disease. All these results and considerations support the hypothesis that the immune response to SARS-CoV-2 is responsible for a Kawasaki-like disease in susceptible patients.”

In the case of the lone patient who tested negative on both tests and whose serology results could not have been influenced by treatment, it may be that his case was a true case of Kawasaki unrelated to the ongoing COVID-19 outbreak, the authors say.

What’s in a Name?

Commenting on the study for TCTMD, Moshe Arditi, MD (Cedars-Sinai Medical Center, Los Angeles, CA), took issue with the increased use of “Kawasaki-like” to describe this emerging phenomenon, arguing that the presentation described by Verdoni and colleagues is much more in keeping with classic toxic shock symptoms, not Kawasaki disease.

Conjunctivitis, swollen lymph nodes, red lips, red tongue, and a rash may all be symptoms of Kawasaki, but they also are symptoms of scarlet fever and measles, Arditi noted. Moreover, the dramatic blood pressure drops and severe myocardial dysfunction seen here, along with multiorgan involvement, kidney failure, abdominal pain, and diarrhea, denote a distinct, more serious phenomenon.

“None of these things are seen with Kawasaki disease, not even the rare part of shock syndrome in Kawasaki disease, which affects just 5% of patients,” Arditi said. “This is a lot more than that.” In fact, he continued, this new entity in children—which he pointed out is affecting older children and teens, not the infants and toddlers typically diagnosed with Kawasaki—“is actually not too dissimilar to the later phases of adult COVID-19 infection that we are seeing.” In many severe adult cases, he reminded TCTMD, patients appear to recover then a week later develop cytokine storm, systemic inflammation, multiorgan injury, vomiting, and gastrointestinal involvement.

“At this early stage of our experience, the available data support a strong argument that this is a new syndrome, it shouldn’t have Kawasaki or ‘Kawasaki-like’ in the title,” Arditi urged. “And if you want to compare it with anything you can say it has some Kawasaki features, but it’s much more a toxic shock syndrome picture.”

At this early stage of our experience, the available data support a strong argument that this is a new syndrome, it shouldn’t have Kawasaki or ‘Kawasaki-like’ in the title. Moshe Arditi

D'Antiga, however, disputed this, noting that Kawasaki is a clinical diagnosis and all of the children in their published series met the criteria for diagnosis in terms of having 5 days of fever plus four of the five hallmark conditions. “So this is Kawasaki disease, but it is more severe than we usually see and the severity is mainly the cardiovascular involvement,” D'Antiga argued. “These patients tend to have myocarditis and gastrointestinal symptoms, which we have not seen in the past, but then they have all the criteria of Kawasaki as well: high fever, rash, glossitis, conjunctivitis, lymph node enlargement: they all fulfilled these criteria.”

Part of the pushback from the United States and the United Kingdom, he added, may relate to the fact that centers there are only seeing the most-severe, hospitalized children—the broader spectrum of late presentation Kawasaki-like symptoms may be underdiagnosed in the community, or may yet emerge.

Critically, when all of these pediatric patients in the Bergamo series were treated according to Kawasaki guidelines, including the use of steroids for patients who met criteria (eight out of 10 patients), they recovered.

“All responded very quickly,” D’Antiga stressed. “If you don’t recognize the disease and you don’t it treat it properly, you can have complications,” he said, likening it to stamping out a fire before it blazes out of control. “These patients will go on and on with the cytokine storm, they will have cardiac impairment, they will need to go on a ventilator, and they may die. But if you treat them as Kawasaki disease, they will respond to treatment.” All of the original 10 patients have gone home. Three of the subsequent 10 patients who presented after April 20 remain on the wards, including one that was previously in the intensive care unit, but all he said are “absolutely well” and will recover.

Where D'Antiga agreed with Arditi was on the belief that the underlying pathology—the immune system response to the virus—is similar in children and adults. “For some reason in adults the main presentation is the respiratory syndrome and in the kids it is Kawasaki disease. But it is still the immune response of the host causing most of the injury,” D’Antiga said.

In an accompanying editorial, Russell M Viner, MBBS, PhD (UCL Great Ormond Street Institute of Child Health, London, England), and Elizabeth Whittaker, PhD, MB BCh BAO (Imperial College Healthcare NHS Trust London, England), write that gaining a better understanding of the immune system’s role may provide “vital information” not only on the immune response to the SARS-CoV-2 virus, but also on “possible correlates of immune protection that might have relevance both for adults and children.” As knowledge grows in this arena, it may conversely shed new light on Kawasaki disease itself, they note, but for the more pressing task of combating this pandemic, there are other important lessons to be learned.

“If this is an antibody-mediated phenomenon, there might be implications for vaccine studies, and this might also explain why some children become very ill with COVID-19, while the majority are unaffected or asymptomatic,” the editorialists suggest.

That has, of course, been the million-dollar question for adult patients as well an active area of research. Whether the same genetic, physiologic, or pathologic factors figure into disease severity for both children and adults needs much more study, D’Antiga told TCTMD. While acquired cardiovascular disease, hypertension, diabetes, and obesity have all emerged as predisposing factors in adults, these conditions are typically absent in infants, although cardiometabolic diseases are increasingly diagnosed in children and teens—the age of the children presenting with Kawasaki-like disease during this pandemic, a point D’Antiga made to TCTMD. Body mass index (BMI) is the only one of the cardiometabolic risk factors reported in the Italian series, and D’Antiga highlighted the fact that the children presenting during the COVID-19 period had higher BMIs than children seen in the control period ( mean 19.11 kg/m² vs 15.93 kg/m²; P = 0016.

“But I don’t know if that is just an effect of time, because our children are also becoming more [overweight] over time, or whether there is a predisposition to disease in heavier and overweight children,” he added.

The very small numbers of children affected makes it difficult to draw any firm conclusions, he continued. In the British series, authors speculated that African origin might be a predisposing feature for Kawasaki-like disease in children; in the Italian series, only one child was of African descent.

In the press release, study coauthor Annalisa Gervasoni, MD (Hospital Papa Giovanni XXIII) emphasized something that will be welcome news for parents, namely, that “only a very small proportion of children infected with SARS-CoV-2 develop symptoms of Kawasaki disease.” That said, she continued, “it is important to understand the consequences of the virus in children, particularly as countries around the world grapple with plans to start relaxing social distancing policies.”

Indeed, the authors end their paper with a grim prediction, concluding: “A similar outbreak of Kawasaki-like disease is expected in countries affected by the SARS-CoV-2 pandemic.”