Adenovirus and Kawasaki Disease: A Diagnostic Challenge

Kawasaki disease (KD), a pediatric medium vessel vasculitis, is the main cause of acquired heart disease in children in North America. Since it was first described by Dr Tomisaku Kawasaki in 1967 in Japan, its etiology remains unknown, its diagnosis relies mainly on clinical criteria without a single pathognomonic biomarker, and its cardiac complications continue to drive the need for timely clinical decision-making in the face of at times diagnostic uncertainty.¹

Adenovirus presents a diagnostic challenge to the clinician managing the febrile child. It shares several clinical features with KD in children 5 years of age or younger, the group most affected by KD. Yet it is also detected as a concomitant coinfection in children who meet the diagnostic criteria of KD.²

The study by McCrindle et al.³ in this issue provides the largest single-center comparison to date of patients with KD with and without concomitant adenovirus detection, and its findings compares the presentation and outcomes of both cohorts.

Insights From Clinical and Laboratory Data

The 2 groups did not differ in age of presentation, duration of fever, number of intravenous immunoglobulin (IVIG) doses or corticosteroid administration, and coronary outcomes. Children with adenovirus-positive KD were more likely to have respiratory (61% vs 20%) and gastrointestinal (47% vs 27%) symptoms, both hallmarks of adenovirus infection that are not typical of KD, whereas those with adenovirus-negative KD were more likely to present with rash (69% vs 88%), a finding that could either reflect the clinical profile of adenovirus disease or a KD-like phenotype.³

The laboratory findings show a lower white blood cell, neutrophil, eosinophil, and platelet count in the adenovirus-positive group. This differs from the classic KD profile of leukocytosis with neutrophilia and thrombocytosis.³ These findings could reflect the effect of a coinfection on the immune response in KD or simply that adenovirus-positive patients have a completely different inflammatory process. Elevated transaminases can be seen in both acute adenovirus infection⁴ and KD,⁵ yet the adenovirus-positive patients had lower alanine aminotransferase levels, which may reflect less hepatic involvement and a milder inflammatory response.

Should the Clinician Be Reassured by the Coronary Artery Outcomes?

The current study shows no coronary artery (CA) involvement among the 39 patients with adenovirus-positive KD, whereas the adenovirus-negative group had a considerable number of CA aneurysms. Twelve percent of patients with adenovirus-negative KD did not receive IVIG. Yet the 2 groups did not differ in the number of IVIG doses or steroid administration.³ This observation deserves careful interpretation within the broader body of evidence regarding concomitant infection detection and KD. Prior studies have investigated the role of viral or bacterial coinfection albeit not specifically adenovirus in patients with KD and have found no association between CA outcomes or IVIG resistance and coinfections.^6,7 Similarly, a meta-analysis looking at 18 observational studies found no increased risk of CA involvement in patients with KD with a concomitant viral infection.⁸ McCrindle et al.³ focus their findings specifically on adenovirus and suggest that adenovirus-positive patients may be at a lower risk for CA involvement. The jury is still out as to whether this reflects a KD-like disease, a potential misclassification of adenovirus disease cases as KD or simply a small sample size of adenovirus-positive patients.

A case series described 4 cases of KD with adenovirus coinfection in South Africa, all of whom developed severe disease including IVIG resistance, macrophage activation syndrome, and CA aneurysms.⁹ This contrasts with the current study and others, possibly reflecting an ascertainment bias with preferential reporting of severe cases. Yet it also may present a different end of the spectrum of adenovirus coinfection in KD, where the relationship between the virus and CA outcomes may be heterogeneous, influenced by the viral species, the viral load, genetics, and/or the timing of exposure relative to the onset of KD.

Deciphering Adenovirus Positivity in Kawasaki Disease: Mimicry, Coinfection, or Persistence?

The main challenge of this study acknowledged by the authors themselves is the inability to determine whether adenovirus positivity was indicative of acute coinfection, an incidental detection, or an adenovirus disease that mimicked KD enough to meet diagnostic criteria.³ This is not simply an academic exercise, yet it is a point that has major diagnostic and treatment implications. Studies have shown that quantitative polymerase chain reaction (PCR) and molecular typing may help resolve this ambiguity. Jaggi et al.² showed that patients with adenovirus-positive KD had significant lower viral loads than patients with adenovirus-positive infections mimicking KD. Species B predominated in the adenovirus infection group, whereas species C, the species that persists in tonsillar and adenoid tissue, was more common in the true KD group.² Similarly, Song et al.¹⁰ found that patients with acute adenovirus infections mimicking KD had fewer than 4 KD clinical features, a predominance of species B or E, and a higher viral load compared with KD infection with an adenovirus coinfection. Neither quantitative PCR nor molecular typing was available in the present study by McCrindle et al.³

The 2017 American Heart Association (AHA) KD statement recognizes the challenges of diagnosing KD in the setting of an infection and offers practical clinical guidance: “in a patient with fever, exudative pharyngitis, exudative conjunctivitis, and a positive adenovirus PCR, KD is extremely unlikely.”⁵ Yet when conjunctivitis and pharyngitis are nonexudative, adenovirus detection does not preclude a diagnosis of KD. The same principles were reiterated in the 2024 AHA update.¹ In the current study, the proportion of patients meeting AHA criteria for complete KD was identical (72%) between the 2 groups.³ This reinforces the principle that the detection of adenovirus alone is not sufficient to exclude KD.

Does Adenovirus Trigger KD?

The dramatic decline in KD incidence during periods of social distancing and masking with decreased exposure to respiratory and albeit other environmental agents, provided the strongest supporting evidence that infectious triggers may play a role in KD pathogenesis.¹¹ The emergence of the multisystem inflammatory syndrome in children as a postinfectious hyperinflammatory syndrome after SARS-CoV-2 further reinforced this hypothesis of a virally triggered KD-like inflammation.¹²

Several studies have examined the hypothesis of adenovirus triggering KD. Using a multivariate time series model in France, Valtuille et al.¹³ estimated that 24.4% of KD cases may be attributable to adenovirus: the highest proportions of pathogens examined in that study. A Taiwanese population–based study found that the incidence of KD was 5 times higher in those exposed to adenovirus compared with other infections.¹⁴ Finally, Marutani et al.¹⁵ found higher adenovirus positivity rates in KD compared with adenovirus non-KD patients. Patients with adenovirus-positive KD had higher anti-adenovirus antibody titers compared with non-KD patients, suggesting that a recent adenovirus infection may act as a trigger for KD.¹⁵ Yet a South Korean nationwide study found that while respiratory infections with varicella, rhinovirus, and the respiratory syncytial virus had a significant temporal correlation with KD, adenovirus did not.¹⁶

When examined within this ongoing debate, the findings take on a different significance. The mild clinical presentation along with the absence of CA involvement might suggest that adenovirus-triggered KD represents a milder phenotype or that the immune response in adenovirus is so robust that it triggers a KD-like inflammatory process without reaching the threshold required to result in CA involvement.

Implications for Clinical Care and Future Directions

The study by McCrindle et al.³ offers important contributions to the beside clinician. First, the detection of adenovirus should not be used as a criterion to withhold timely KD treatment in children meeting KD criteria. In the present study, all patients who tested positive for adenovirus and who met KD criteria received treatment with IVIG consistent with current best clinical practice. Second, the absence of CA involvement in the adenovirus-positive group should be interpreted with caution given the small sample size and the potential for a KD-like attenuated inflammatory response. This leads to a third and important consideration. When available, quantitative assessment of viral load and identification of adenovirus species may help differentiate an acute adenovirus infection from either KD with coinfection or KD with a recent exposure to adenovirus.

Further, large multicenter studies are needed to better characterize the relationship of adenovirus and KD focusing on the effect of adenovirus on KD severity, IVIG resistance, and CA outcomes. Artificial intelligence–guided and validated scoring systems using clinical data like the one developed by Fabi et al.,¹⁷ along with quantitative PCR, molecular typing, gene expression profiling, and cytokine panels, may help improve the diagnostic precision of adenovirus infection with a KD-like presentation from adenovirus-coinfected KD cases.

Funding Sources

The author has no funding sources to declare.

Disclosures

The author has no conflicts of interest to disclose.