IVIG-Resistant Kawasaki Disease with a New Small Aneurysm: A Trial That Didn’t Quite Test This Patient
The randomized comparison of second-line options answered its primary question cleanly — about fever. His coronary arteries are the endpoint the same trial looked at and could not separate.
Marcus J. is a 3-year-old boy who was admitted three days ago with five days of fever, bilateral non-purulent conjunctivitis, a diffuse polymorphous rash, cracked erythematous lips, and swollen hands and feet — a presentation that met full clinical criteria for Kawasaki disease and prompted immediate treatment with IVIG at 2 g/kg along with high-dose aspirin. His fever, which had briefly dipped during the infusion, returned to 38.9°C thirty-eight hours after the infusion completed, meeting the standard definition of IVIG resistance rather than a simple slow initial response. A baseline echocardiogram obtained on admission, before any coronary change would typically be expected this early in the illness, was normal; a repeat echocardiogram obtained this morning, now eight days into the illness, showed a coronary artery Z-score of 2.6 in the left anterior descending artery. That number is not merely “abnormal”: the 2017 AHA classification puts dilation alone at a Z-score of 2 to under 2.5 and calls anything from 2.5 up to 5 a small aneurysm, so at 2.6 he has crossed out of dilation and into the aneurysm range — by three days ago he had none at all.
His inflammatory markers remain markedly elevated — CRP still above 150 mg/L and a platelet count that has begun trending upward, the pattern that typically follows the initial thrombocytopenic phase of the illness as it evolves. A repeat albumin, drawn alongside this morning's echocardiogram, has fallen further to 2.6 g/dL, consistent with ongoing vascular inflammation rather than early recovery. The KIDCARE trial (Burns and colleagues, 2021) settled its primary question: fever resolution at 24 hours without recurrence, 77 percent on infliximab against 51 percent on a second IVIG infusion. On its own secondary analyses it found no difference between the arms in inflammatory markers or coronary artery outcome. That is the awkward shape of the evidence for Marcus: it favors infliximab decisively on fever and separates the two drugs not at all on the endpoint his echocardiogram just made the point of this admission. The RAISE trial (Kobayashi and colleagues, 2012) speaks to coronary outcome directly, but tested corticosteroids added to first-line IVIG in patients scored high-risk at the outset, not second-line therapy in a child who has already failed an infusion.
Whichever second-line agent gets chosen tonight, the team keeps returning to the same underlying question none of the available trial evidence answers cleanly for a patient exactly like him: which approach most directly protects coronary arteries that have already started to change.
A trial result that doesn’t quite match his echo
His coronary arteries have crossed into the small-aneurysm range, which changes what I'm weighing here beyond just fever resolution. KIDCARE is not ambiguous on its primary endpoint — 77 percent versus 51 percent for fever resolution without recurrence, favoring infliximab — and infliximab infuses in about two hours against eight to twelve for a second 2 g/kg IVIG dose. In a child whose coronaries are already involved, I'd rather not add that volume load and that many more hours of ongoing inflammation when the better-performing option is also the faster one.
I want to be careful about what KIDCARE actually showed, because I think it is being read as more helpful than it is. Its primary outcome was fever, and on fever it is clear — I'm not disputing the 77 against 51. But the same trial looked specifically at inflammatory markers and at coronary artery outcome and found no difference between the arms on either. So the endpoint it settles is the one that is no longer what worries me about this child, and the endpoint I care about is the one it explicitly failed to separate.
I'm not saying infliximab is wrong for him — on this evidence it is probably right. I'm saying we shouldn't let a decisive fever result stand in for a coronary result the trial went looking for and did not find. If we choose infliximab tonight and his Z-score keeps climbing, I don't want anyone surprised, because nothing in KIDCARE promised us otherwise.
Then you agree more than it sounds like, and the agreement points somewhere neither of you has gone. If KIDCARE settles fever and leaves coronaries open, the coronary question needs its own drug. The RAISE trial — Kobayashi and colleagues, 2012, randomized, not a cohort — found that adding prednisolone to IVIG cut coronary artery abnormality rates substantially in patients scored high-risk at presentation. That is a finding about coronary protection directly, not fever resolution standing in as a proxy for it.
That trial tested steroids as initial therapy in high-risk patients, not this second-line, IVIG-resistant scenario, so I won't overstate the match. But given that his coronary arteries, not just his fever, are the actual concern driving tonight's urgency, I'd want corticosteroids added to whichever second-line agent you two choose, rather than treated as a separate third option to pick instead of theirs.
Infliximab was given with prednisolone added, and a repeat echocardiogram was scheduled at 48 hours to track the coronary Z-score directly rather than infer coronary protection from fever resolution or inflammatory markers alone.
Not agreed, and explicitly left open: the Pediatric Rheumatologist maintains that if the coronary Z-score improves at 48 hours, that outcome would not actually distinguish whether infliximab, the added corticosteroid, or their combination deserves credit — and, given that KIDCARE found no coronary difference between the agents at all, would want the improvement attributed to the steroid until something shows otherwise. The Pediatric Cardiologist and Clinical Pharmacologist agreed the attribution is genuinely unresolved, but both held that for tonight's decision, protecting his coronary arteries took priority over knowing afterward which part of the regimen deserved the credit.