Hot Seat #28 Denouement: 19 day old with pustules

By Rosemary Thomas, Children’s National
With Jennifer Anders, Johns Hopkins, on the Hot Seat
Case edited by Emily Willner, Children’s National

The Case
This is a case about a well-appearing, almost 3 week old with a history of treated E. Coli sepsis now presenting with pustules on her hand. The picture is below. For a complete case presentation, click here to go to the original post.

Here’s how you answered our questions (n=17)
In this well appearing neonate with the described skin findings, what work-up would you do initially?

The audience was split between the two extremes. About 1/3 wanted to test the lesions only while an almost equal number would do the full workup with LP.

The gram stain reveals no bacteria, rare WBC/RBC, moderate epithelial cells. You have contacted the PMD and can assure close follow-up.

What disposition would you choose for this patient?

Most people would admit for IV antibiotics and acyclovir, likely the same group of people who wanted to do the full sepsis work up in the previous question. However, there was variation among the remaining people on whether to send the patient home or admit for observation.
 
Clinical Pearls
Thanks to Dave Mathison, Shilpa Patel, Rasha Sawaya, Todd Chang, Paul Mullan, and Joelle Simpson for adding your clinical opinions and Jennifer Anders for sitting on the Hot Seat this week.

The discussion focused on what this might be and how aggressive we should be on the workup of this patient. There was no consensus although it seemed that most people would at least get blood and a blood culture. Reflective of the respondents to the above questions, there was practice variation on whether or not to perform a lumbar puncture based on age and risk and whether to admit this patient. Others debated whether this patient should get empiric antibiotics. The discussion is a useful elaboration on a challenging question without clear answers. Click here to read more on their clinical reasoning and what kinds of challenges readers were tackling.

On the Clinical Approach

“My general approach in the ED is: shotgun diagnosis, what else could it be, does the patient have risk factors to make me worry more.” –Dave Mathison, Children’s National

“This is one of those times I wish we had a microscope as well and could do a gram stain of the lesion or at least get a quick path read from a swab.” –Joelle Simpson, Children’s National

“My takeaway (aha!) learning point was: "Common things are common." When you see something that is unusual, go back to "most common" diagnoses and consider those first. If you saw on the tracking board "3 week old with new rash," what would be #1 in your mind for all-comers?” –Paul Mullan, Children’s National

On What to Consider

“The candidiasis could be secondary to the antibiotics, but what if the candidiasis, this rash, the E coli sepsis were all from an immunodeficiency which would also put this baby at high risk for HSV which could start cutaneously and quickly go systemic.” –Rasha Sawaya, Children’s National

“The lesions are described as having an erythematous base, are coalescing and some appear to be drying. Also there are two rashes in different stages (palm and thighs), all concerning for HSV. Classically isolated SEM (skin, eyes, mucous membranes) HSV presents at 10-14 days, so this child is a little old but still high risk.” –Shilpa Patel, Children’s National

“The formation looks herpetic to me…Clear vesicles with some that join up; haven't seen many other lesions do those clustering in linear fashion except perhaps poison ivy or local reaction.” –Todd Chang, CHLA

“Herpes? This is always a tricky one because if you suspect it in this age group, it really buys the whole shebang workup. But I would argue that these are really pustular and not vesicular so I wouldn't necessarily go down that route.” –Dave Mathison, Children’s National

“I would be cautious in only considering HSV in vesicular neonatal rashes - HSV can appear pustular.” -Shilpa Patel, Children’s National

On What to Do

“I hesitate to do an LP but would be worried if a 19d F already had a viral infection, a bacterial infection, and a fungal infection not responsive or even getting better on nystatin, and would at minimum admit for IV acyclovir; the floor can LP and do abx if needed.” -Todd Chang, CHLA

“I would check the newborn screen for galactosemia to be safe. Recent antibiotic use is a setup for candida infections, but a candida infection in the setting of recent illness also makes me concerned about an underlying immune disease. and systemic candidiasis can be fatal.” –Dave Mathison, Children’s National

“I usually refer to the Healthy Children website from the AAP to provide parents with their synopsis of typical rashes in a healthy newborn. It is very informative and can be found at: http://www.healthychildren.org/English/ages-stages/baby/bathing-skin-care/Pages/Your-Newborns-Skin-Birthmarks-and-Rashes.aspx. I have made it into a routine handout for patients” –Joelle Simpson, Children’s National

From the Hot Seat
By Jennifer Anders
Johns Hopkins

With any baby less than 28 days, presenting with either fever or signs of a focal infection, I am going to be pretty concerned about systemic infection. I am especially concerned given this baby’s history of perinatal sepsis. So despite the reassuring lack of fever, given the pustular and erythematous appearance of the hand I would still proceed with a full septis workup including HSV evaluation. Since risk stratification by lab results or clinical appearance is poor in this age group, I would also admit with empiric treatment for both bacteremia and HSV.

So that covers high risk, but I don’t think HSV is the highest likelihood diagnosis. I’ve organized my thinking around three main areas to narrow the differential. This is a very young infant, the rash has a distinct distribution on the hand, and the baby has prior medical history and concurrent infection with thrush.

Age. HSV must be considered in a newborn with a papulovesicular rash. This rash is more confluent than I expect from neonatal HSV, and more pustular than vesicular. The tight clustering and pustules look more like herpetic whitlow, which I have never seen at this early age. Could an adult shedding oral HSV been kissing the baby’s hand? I would ask more about caregivers’ cold sore history.

Could this be a neonatal pustulosis syndrome (transient neonatal pustular melanosis or acropustulosis of infancy)? TNPM should be present at birth, resolving over the first week of life, so that doesn’t fit for this baby. Acropustulosis of infancy is often on the palms and presents in the first weeks of life, so the timing is good. In both those cases, the lesions should be filled with neutrophils, which was negative for this patient.

Erythema toxicum typically shows up earlier in the newborn period than this (first week of life), should not cluster like this, not show up on the palms and the photos don’t look right for e tox. Those lesions should have smaller rounder papules or vesicles on pink base.

Distribution of rash. The description of location in the hand is intriguing. Since there was no picture of the scattered papules on the thighs, I disregarded them, which could leave me vulnerable to missing something!

I like the possibility of acropustulosis given the location and appearance. The description of vesicles/pustules in the webspaces and sides of fingers makes me curious about scabies – although that classic distribution in bigger people is not usually the distribution in infants. But location on the palms is common in babies.

I have seen several newborns with impetigo/staph with pustules that look just like this. Could this baby be sucking her fingers and creating skin break down? Was this hand used for her IV in the NICU and became secondarily infected? Like the pustulosis syndromes, I would expect more neutrophils in the smear of the lesion if this was staph infection.

Past history/associated signs and symptoms. With the thrush and diaper rash described, could this be some sort of candidal skin infection? Candida on a dry skin surface with thick epidermis (aka the palm) is unusual, but baby might be sucking her fingers.

Other questions that I pondered but don’t have good answers for include: Could this be soft tissue seeding with E. coli related to the prior sepsis? I don’t think so. Could this be something picked up in the NICU? I don’t even know what to suspect beyond staph. My blood culture and culture of lesions should cover both those possibilities. Could this be allergic or post traumatic response to the IV access or tape used in the previous hospitalization? If so, there is nothing to do but supportive care and if baby has a tape allergy I will probably see new evidence after this episode of care.

So in conclusion, I am going to do a full sepsis work up and admit on antibiotics and acyclovir. I will have a dermatologist consult while the baby is an inpatient.

My ranked differential diagnosis for the rash itself is:
1. Staph infection
2. Acropustulosis of infancy
3. HSV
4. Allergic reaction (to tape used for IV?)
5. Candida infection

Denouement
By Rosemary Thomas
Children’s National Medical Center
 
A bacterial culture and gram stain was sent from the lesion. In consultation with ID and dermatology, initially outpatient management with PO antibiotics was considered in this well-appearing patient with an isolated skin finding. Given the patient’s young age and history of NICU stay, admission for observation was also considered. However, when the Gram stain of lesion was negative for WBC and bacteria, a full ROS was performed.

The working differential based on expert consensus: HSV, staph (impetigo), coxsackie, contact dermatitis (from IV tape given the band-like distribution), vesicular/pustular stage of incontinentia pigmenti, yeast

Labs were notable for CBC with WBC 10.8 with lymphocytic predominance and mild eosinophilia (n 19, m7, L 67, e 6). Urine and CSF studies showed no WBCs.

HSV cultures were sent from lesions, mouth, nose, and rectum and later returned negative. Blood, urine, and CSF cultures were also negative.

The bacterial culture of the lesion grew 2+ yeast, and a KOH prep was sent showing fungal elements. A fungal culture of the lesion grew 2+ candida albicans.

The baby was discharged home 2 days later after cultures came back, with Nystatin PO and topical for rash.

The Hot Seat is a recurrent online case series aimed at facilitating asynchronous sharing of PEM knowledge and experience. Faculty are placed on The Hot Seat and publish their opinions without knowing the case or its outcomes. Cases are based on real cases and written by PEM fellows at Children's National in DC, Johns Hopkins in MD, and INOVA Fairfax in VA to highlight teaching points. Emily Willner is the faculty advisor for the Hot Seat.