96: Riddle Me This

Most people don't realize this, but I actually like to make riddles, like homemade riddles.

I mean, I could do a riddle now or at the end of the podcast, if you want, it's up to you.

On day two of fever, he presented to his primary care pediatrician and was prescribed a five day course of cefdinir for presumed left sided otitis media in the setting of a reported amoxicillin allergy with a rash. Despite the antibiotics, he continued to have fevers, so mom brought him to the ER. The resident asks our thoroughts on the matter if any additional workup we would like.

Dr. Kumar, I'd like to give you some more objective data in a bit, but before that, can you kind of describe your thought process when you receive a call like this about your differential diagnosis of a patient with prolonged fever and what sort of workup you may consider performing?

There's really a lot to parse through in terms of the calendar and timeline of fevers when you get a call like this to make sure that you're on the same page as the person who's consulting you and that you understand exactly what kind of category this falls in.

In this sort of story, this isn't the protracted fever without a source where most of our counseling comes from saying, you know, it's okay and as long as they're well, that we can do some rudimentary workup, but not get too far down the rabbit hole. In this case, it sounds like they did have a focus, and that focus might have been an acute otitis media. That's another one where diagnostics from a primary team can sometimes be challenging, and a lot of interrater reliability isn't quite there.

So if I have this story, I start to think, well, maybe they were on the wrong track, and that may not have been the focus since it didn't respond to antibiotic therapy with the caveat that they did use suboptimal antibiotic therapy, right? Cefdinir and pneumococcal coverage is always a concern. And so those are my sort of initial thoughts.

Really, what we're looking for is coming up soon, Jack, which is a little more objective data so we can start anchoring in on exactly where the source of this might be.

Additionally, he had decreased intake of both liquid and solids with reduced urine output. And he did have some loose stools that began when he started the cefdinir, and they've persisted while he's been on the cefdinir. At this point, we do go down to the ER to see the family and the patient and we get a little bit more deep into the history. He lives at home with his parents and three school age siblings. He does not attend daycare and the parents don't report any sick contacts.

His vaccinations are up to date and he's had no previous issues reported by the PCP. He had a normal birth history, full term, no NICU stay, no oxygen, no phototherapy for hyperbilirubinemia. Additionally, the family reports there's been no recent travel. They live in a nice, clean, reported suburban home just outside of the city of Chicago. They do have a dog at home, and there's no family history of immune deficiency or recurrent infections.

So, how does this information affect your differential, and what now additional diagnostic workup would you like the patient to receive at this point?

So whenever I hear rash and then eyes with some conjunctival changes, we've got potential mucous membrane changes when I hear about these new oral ulcers, you have to sort of wonder whether this is something non infectious, right?

Whether that's something more like a Stevens Johnson syndrome, a TEN, even serum sickness with oral ulcers obviously being a little bit odd there, but those are the kind of things that come up to my head, but we have to start by ruling out more common infectious pieces. You can always go with the old standby. You can say it's likely viral, which will always get you points in the ID world. And in this case, it might be true.

We have adenovirus and other adenovirus families that can present very much like this. It's very reasonable to test them. And in this story where they do have some persistence of symptoms and, you know, they're obviously in a healthcare setting or a more acute healthcare setting, it makes sense to send it, in my opinion, a respiratory viral panel, a respiratory pathogen panel, whatever your institution provides.

On that same note, as maybe some folks have been hinted to by the title of this podcast, I'm going to guess there is the concern that this is Kawasaki's disease based on the rash, some changes to the conjunctivae, and then of course these mucous membrane changes as well. So that has to be a part of your differential and workup. This is the child that's going to get their screening labs.

This is the child that's going to get a blood culture to ensure there's nothing we're missing there, probably respiratory viral testing and a formal infectious disease consult.

Many you already mentioned, you know, adenoviruses, enterovirus family, Epstein Barr virus, or other mononucleosis type illnesses. Perhaps a toxin producing bacterial infections such as Staphylococcus aureus or Streptococcus pyogenes, and perhaps also a drug related incident due to the cefdinir. However, also in the differential diagnosis was KD or incomplete Kawasaki disease, formerly called atypical Kawasaki disease.

And once again, I want to dive a little bit into KD before we get back to our case. Kawasaki disease is a vasculitis of the medium sized arteries. The etiology is unknown, but there are some fascinating science that's kind of been coming out, and I'll let Madan describe that in a bit that may suggest the true or an evolving true origin of it.

Epidemiologic and clinical features suggest an infectious origin, and, or perhaps an environmental cause or trigger in genetically susceptible individuals or perhaps within family lines or environments. The peak age of occurrence in the United States is 6 to 24 months.

50 percent of patients are younger than 2, and 80 percent are younger than 5 years old, so this is definitely a disease of the younger child for the most part, but there are cases that can be older than 8 years old, and very rarely they've occurred even in adults. There is a described definition of KD. It's the fever of five days at least in addition to the presence of at least four of the following five clinical criteria.

And this is kind of where the subjectivity of the clinical exam does come in to play. So you can have bilateral injection of the bulbar conjunctivate with limbic sparing and without exudate. I feel like that was something I got pimped on a lot as a resident, the limbic sparing part. Erythematous mouth and pharynx, strawberry tongue, and or red cracked lips, that's the second one.

Third is a polymorphous, generalized, erythematous rash, often with accentuation in the groin, which can be morbilliform, maculopapular, scarlatiniform, or erythema-multiforme like. The fourth is changes in the peripheral extremities consisting of erythema of the hands and soles and firm, sometimes painful, indurations of the hands and feet, often with periungual desquamation, usually beginning 10 to 14 days after fever onset.

So this can be a little bit more delayed and kind of a more transitionary finding. And then the fifth one is acute non suppurative, usually unilateral, anterior cervical lymphadenopathy with at least one node greater than or equal to 1. 5 centimeters in diameter. So that's the traditional KD, but I want to move on to incomplete or atypical KD. So this is our children with greater than or equal to five days of fever.

And then you might have two or three of the clinical criteria, which I think is what we often find more often, or, you can have infants with a fever for greater than, uh, are equal to seven days without explanation, which I'll have Dr. Kumar describe in a bit. And now you want to get some basic labs, so you want to get a CRP and ESR.

I'm going to describe kind of what the AAP Redbook uses for their numbers and their types of CRP and ESR, but if you have a CRP less than 3 mg/dL and an ESR less than 40 mm per hour, they recommend serial, clinical, and laboratory re evaluation if the fever persists, or you can consider an echocardiogram if peeling develops, interestingly enough. This can be done in the inpatient or outpatient setting, but I feel like more often than not we do admit the kids overnight for observation.

If the CRP is greater than 3 mg/dL, or the ESR is greater than or equal to 40 mm per hour, then you need three or more of laboratory findings. So, one is anemia for age. Two is platelet count greater than or equal to 450, 000 after the seventh day of fever. The third is albumin less than or equal to 3 g/dL.

If you have an elevated ALT level, if your white count is, white blood cell count is greater than or equal to 15, 000 per millimeter cubed, and then if your urine white blood cell count is greater than or equal to 10 white blood cells per high power field. Or, if you simply have a positive echocardiogram for coronary artery dilation. At that point, then you move on to treatment. So I just want to go back to our case really quick.

So the patient did end up getting an IV fluid bolus as he appeared dehydrated. A rapid group A streptococcal throat swab was performed in return negative. On our physical exam, the ID consultant physical exam, he was ill appearing, he was fussy and notable for sunken features. He did have the erythematous hands and feet, non purulent conjunctival injection. The oropharynx was erythematous and swollen with bloody and cracked lips and with areas of bleeding around the right posterior pharynx.

He was also noted to have small, shoddy cervical lymphadenopathy and was tachycardic. Interestingly enough though, there was no rash in the patients outside of the hands and feet, so no trunk, body, arms, or legs, just the hands and feet. We did get some blood tests, included an elevated white blood cell count of 18, 100, which is greater than the 15, 000 cutoff, with a neutrophilic predominance. He had an elevated platelet count of 550, 000. He did have a normal hemoglobin for his age.

He had an elevated ESR to 112 millimeters per hour and a very high CRP of 77 milligrams per deciliter, well above the 3. 0 cutoff. The patient's alkaline transferase, or ALT, was elevated to 263. And his remaining comprehensive metabolic panel was within normal limits. He also had a urinalysis performed which noted significant pyuria for 26 to 100 cells per high power field, and negative for nitrites or bacteria, and urine and blood cultures did remain no growth to date.

He did have a respiratory pathogen panel on nasal PCR swab, which was negative, and chest X ray was unremarkable. At this point, we did feel that he met three of the five compatible criteria, including the erythematous oropharynx features, non purulent conjunctivitis and changes in the peripheral extremities. He also had thrombocytosis, a low albumin, elevated ALT, the leukocytosis, and the pyuria, along with inflammatory markers, to meet a clinical diagnosis of incomplete KD.

Transthoracic echocardiogram was performed, which did show elevated Z scores of plus 2. 4 and plus 2. 1, respectively, and dilation of the left anterior descending and left medial coronary arteries, consistent with a diagnosis of Kawasaki disease. So, Dr. Kumar, can you give us thoughts on the clinical findings on the case, and I would also love if you wanted any more additional thoughts just on the diagnostic algorithms I described earlier.

So when we do that, sometimes trainees get caught in a memorization pattern and it makes it difficult to understand the mechanism of the disease, which really should be at the forefront in evaluating this disease. So if you go back to the basics and just remember that this is a medium to small vessel vasculitis, it's much better and easier to remember where that might manifest.

You might see redness in the eyes, but of course you shouldn't see purulence, not because you have to memorize it, but because this isn't a primary bacterial or virological purulent mechanism, it is a blood vessel inflammation. Similarly, with the rash, you can know that you're going to see a rash, and you can also know that if it is a vasculitis, that rash can be polymorphous, morbilliform, a lot of ways to say anything, any sort of red rash can fit with the parameters of KD.

The lymphadenopathy, it used to be a preeminent piece of the disease. We usually see it unilaterally again, of course, because it's not the same typical infectious lymphadenopathy process. And then the mucous membrane changes are other areas where hyperemia can be eminently visible. So those things can kind of help you just categorize why you're seeing what you're seeing and make it a little bit easier to remember.

The piece that's tough, and if we put, uh, caveats about where we're stepping away from the guideline verbally, this would be it. The piece that's tough is recognizing that not all of these parameters are made equal. If we're trying to diagnose someone with incomplete Kawasaki's disease, you're functionally trying to differentiate them from viral NOS, because you haven't established another clear diagnosis. If you have, it's easier to step away from Kawasaki's disease.

In a viral NOS, you're going to have a lot of overlapping lab parameters and clinical parameters, but there's some that kind of should jump out at you as being a typical for a viral process. So, for me, some of the things that help clinch the diagnosis, if we're sort of on the fence, are things like the sterile pyuria, right? We don't really see sterile pyuria as part of your typical adenovirus upper respiratory infection.

Something that's very hepatotropic, where we have elevated ALT, and then it's not necessarily part of the parameters, but it's something I always look for. Elevated GGT as well can also be indicative of something more in the KD pathway, or KD that's more likely to be resistant to initial therapy. There's some data to suggest that those things are genetically linked.

When you're dealing with Kawasaki's disease, you kind of have to make an internal judgment whether you're going to follow the criteria to the staunchest, most strict possible interpretation, or whether there may be children who look pathophysiologically like Kawasaki's disease, look miserable like most of these kids do, and warrant treatment based on meeting the incomplete criteria and fitting the mechanisms that you suspect. In this case, it's a little bit easier to make the decision to treat.

You've talked about the initial echocardiogram having elevated coronary artery scores. Oftentimes, variable cutoffs are used for the z score, and for trainees who are unfamiliar, the z score is essentially a measure of standard deviation. You're looking at the internal diameter of the artery itself and seeing how many standard deviations above the mean you are, respective to your body size or body surface area. And if you're more than 2, that's meaningful.

If you're more than 2.5, that's often used as even more meaningful of a cutoff point. This person was on the border for 2. 4. So besides making the diagnosis a little easier, it actually puts them into potentially a higher risk stratification. We can get into a little more when we talk about therapy, but in the initial diagnosis, risk stratification is important as well because it can inform your therapeutic decisions.

Unfortunately, because of just the way that the data is stratified and where this tends to be more prevalent, there's very clearly a underlying risk factor in the Asian population, which makes them more likely to get Kawasaki's disease.

When that happens, most of our data from Kawasaki's disease risk stratification comes from Asian countries, and it's unclear if our children in America follow the same risk profile or pattern, but in generality, infants less than six months and those with initial echocardiogram changes tend to be higher risk and require high risk stratification and high risk therapy.

Theoretically, if Kawasaki's disease was purely genetic and had no infectious trigger, the rates would be entirely consistent throughout the pandemic. And then, of course, the converse is true. And that's what we found is that when our general viral rates plummeted, when people were masking, when people were staying at home, similarly, our rates of Kawasaki's disease plummeted.

So epidemiologically, that presents a really compelling amount of data that even though there's clearly a genetic component to both risk for onset of disease and severity of disease. There is also an infectious trigger that we haven't identified. It sort of sets the chain off, which makes a lot of sense, in the pathophys of other similar disease processes.

The other big pieces, um, you know, of course, at this last national ID conference, we were presented data on partial sequencing of the causative virus as well. And that was very compelling. And it does look like within the next 5 years, we'll be able to have a little more substantive sequencing data of the causative virus. So at this point, personally, I think we've sort of clenched it, that this is a two fold process with a viral trigger and a secondary genetic predisposition.

Clinicians should consider KD in their differential diagnosis before the fifth day of course, if several of the features are present without an alternative explanation.

One of the issues that you've mentioned is that different things can have different temporalities, so certain things may present earlier in the course and some may present later in the course, and it's absolutely possible to have a concurrent viral upper respiratory infection in a patient with KD, particularly if it's during certain epidemiologic months of the year, like in the winter. The average duration of fever for untreated KD is 10 days. However, fever can last two weeks or longer.

After the fever resolves though, patient can remain anorexic and irritable with decreased energy for two weeks, and I feel like that's something we've encountered oftentimes in the outpatient setting after we see them in the hospital. The parents are really concerned because they're still not eating, they're still very tired, but it's something we kind of have to describe to them. It's, it's a natural phenomenon.

Also during this kind of recovery phase, the brawny desquamation of the fingers, toes, hands, and feet may occur. Transverse lines across the nails or Beau's lines sometimes are noted to occur even months later. The most serious complication, of course, is the coronary artery abnormalities. It occurs in about 20-25 percent of untreated children.

Certain increased risk factors for coronary artery abnormalities, there appears to be a biologic sex predisposition for males over females, if you're less than 12 months of age or greater than 8 years, so the very young or the very old. If your fever does last more than 10 days, if your white count is greater than 15, 000 with a high neutrophil predominance, If you're anemic, if you have a low albumin, if you have a low sodium, interestingly enough, and if you have high platelet count.

And then, additionally, if your fever persists greater than 36 hours, despite proper therapy, that also increases risk for coronary artery abnormalities. Aneurysms of the coronary arteries usually occur between 1 and 4 weeks after the onset of disease. Onset later than six weeks is extremely uncommon.

If the coronary artery aneurysm or ectasia is evident, as you mentioned before, a Z score greater than two, but really above two and a half, in any patient evaluated for fever, a presumptive diagnosis should be made. A normal early echocardiogram study is typical and does not exclude the diagnosis, but it might be useful in patients with suspected incomplete KD, perhaps that's the patient where you'd want to repeat the echo within 24 to 48 hours.

There was one study that showed that 80 percent of patients with KD who ultimately developed coronary artery disease had abnormalities in echocardiograms obtained during the first 10 days of illness. So, that is still a possibility before the 10 days.

Other exam findings, in many patients you might find urethritis, so pain with urination along with the sterile pyuria, a mild anterior uveitis, less likely you might have elevated serum, you know, transferase concentrations, arthralgias, or arthritis, perhaps CSF pleocytosis, and then even more rare would be hydrops of the gallbladder, a pericardial effusion, myocarditis, cranial nerve palsies. These are all kind of much less common things.

The current case fatality rate, fortunately, in the United States and Japan, where most of the studies are performed, is less than 0. 2 percent at this time. Primary cause of death is myocardial infarction resulting from coronary artery occlusion, attributable to thrombosis or progressive stenosis. The relative risk of mortality is highest within six weeks of of onset of acute symptoms, but that can occur many months to even years after the acute episode.

The prevalence of higher abnormalities is when you delay treatment beyond 10 days of illness, so that's kind of where we have our 10 day cutoff. The first line treatment is IVIG, 2 grams per kilogram, administered over 10 to 12 hours. It's important, particularly depending on the unit in the hospital, that they understand this prolonged infusion rate. A secondary cornerstone is aspirin.

There's the high dose aspirin, 80-100 mg per kg, or the middle to lower dose, 30-50 mg per kg per day, in 4 divided doses. In severe cases, you can consider steroids. If you have recurrence of fever after 36 hours of that first dose of IVIG, we recommend infliximab as one additional dose. So actually, Dr. Kumar, can you chat about the two different doses of aspirin I described, and actually why infliximab is used instead of a second dose of IVIG.

The aspirin piece has also been an area where we've traditionally had a paucity of data, and now, now finding more and more data to suggest that lower doses of aspirin do not have a higher risk of a lot of those complications, those coronary artery complications, that we were traditionally worried about when you have thrombocytosis and a artery abnormality or a aneurysm.

So from that high dose of aspirin, a lot of institutions now feel safe and comfortable switching to medium dose aspirin, and there's actually been even a push to reduce it even further and to start with the low dose aspirin and continue there. Although again, that hasn't made it into the general guidance yet. I suspect that that will be where we end up. As far as the infliximab, that's been an interesting piece.

So traditionally, if you refractory to a single dose of IVIG, we'd often give a 2nd dose of IVIG. And to be fair, many institutions still do that.

There was a multi center study that we were a part of that evaluated the respective risks for secondary coronary artery abnormalities, along with adverse events with each approach, and we found that there was a general overlap in terms of outcomes, and there was no stratification for outcomes with use of infliximab, but we had a much better safety profile compared to second dose of IVIG, particularly with things like autoimmune hemolytic anemia,

which we had a far higher incidence of with our second dose of IVIG. So we've made the choice to switch over to infliximab for our refractory patients.

If they do have abnormal coronary arteries, though, we oftentimes defer to our neighborhood friendly cardiologists, and they oftentimes will help monitor their patients for that and closely see them in the outpatient setting.

If you develop a giant coronary artery aneurysm or very large one with a luminal diameter of greater than or equal to eight millimeters, or perhaps larger in our infants with a z score of greater than or equal to 10, that usually requires the addition of anticoagulant therapy such as warfarin or low molecular weight heparin to prevent thrombosis.

Another interesting tidbit that I think is fair game for the ID board exam and it might have even been a practice question for my Pediatric board exam is the measles, mumps, rubella, and varicella containing vaccines should be deferred until 11 months after receipt of IVIG for treatment of KD because of the possible interference of the development of an adequate immune response.

Just to wrap up our case, the patient received a single dose of IVIG, 2 grams per kg, administered over 10 to 12 hours, in addition to initiating medium dose aspirin. The patient was noted to defervesce within 36 hours and did not require an additional dose of IVIG or infliximab, which is great. He did go home, but interestingly enough, he returned to the hospital 10 days later with low grade fever and upper respiratory infection symptoms, just a runny nose and a sore throat.

He did notice a desquamation of the hands and feet, so the ER asked us if this is actually the return of the Kawasaki disease. He was diagnosed with rhinovirus on nasal PCR swab, and his sibling was also sick and diagnosed too. We described to them that this was an expected finding on the 10-14 day range later, and patient was discharged to him with a close follow up and ended up doing well.

He had a follow up echocardiogram at 6 weeks of age, which showed complete resolution of coronary artery dilation, which was great. So this was a good thing. Good success story. Dr. Kumar, do you have any closing comments on KD or the

And unfortunately, Kawasaki's is not one of them. You hit the nail on the head here, Jack, when you talk about these concomitant viral positivities, right? It would be really nice if we could use those as a reason to say you don't have Kawasaki's disease. But we know kids who are in the right time and place are likely to have multiple viral positivities.

So since the causative trigger for this is viral, if you are Rhino Entero positive or some other viral positive, it may actually allude to the fact that you are more likely to have Kawasaki's disease because you're in daycare settings or around other children or other social risk factors that make you more likely to have these repeat viral positivities as this case highlights. The other piece is that the clinical phenotype can be very varied as well.

We have Kawasaki's disease that can be fairly mild, although those kids still tend to be fairly miserable and unhappy. But we also have Kawasaki's disease that presents with shock, that presents with macrophage activation syndrome, that presents quite fulminantly in children that end up in the ICU on pressers, and the root etiology is still Kawasaki, or what we sometimes sort of colloquially call "Kawa-shock-i". So it's nice to try to close the loop.

But this is one of those diagnoses that you shouldn't do that and shouldn't anchor and should still keep an open mind on, particularly since the outcome differences with treatment can be so substantial and preventing children from having long term coronary artery aneurysms is so very meaningful. So thank you for this case.

I, and we'll of course put some resources about some of these recent papers that you guys are talking about as far as trying to understand the causes of Kawasaki, and I also want to add, I'm very glad that you covered a lot of the specifics, which are, of course, board review, typical questions, but also highlighted the nuance and the thing that stands out to me the most from seeing these patients in residency and beyond was that irritability that isn't really captured in that.

But I remember, you know, talking through these cases with my clinical team and the attendings and learning a ton.

Most of the time I'm free, but occasionally I can be a pretty price to pay. You can find me almost anywhere from Dublin to New York or Paris to Mumbai.

Please reach out if you have any suggestions for future shows or want to be more involved with Febrile. Thanks for listening, stay safe, and I'll see you next time.