Fever, Rash, and Cough in a 7-Year-Old Boy Free

A 7-year-old boy presented to the emergency department (ED) in December with fever, cough, congestion, abdominal pain, myalgias, and morbilliform rash. Thirteen days prior, he had developed intermittent fever, myalgias, a dry cough, and sore throat. Abdominal pain and rash started 8 days later, initially on his face and trunk, spreading caudally and to extremities, sparing palms and soles. Two days later, he saw his pediatrician, who obtained a rapid strep test and Antistreptolysin O titer, given concern for acute rheumatic fever. Amoxicillin was started for presumed group A Streptococcus infection. The rapid strep test was negative, and Antistreptolysin O titer returned negative the next day. Amoxicillin was discontinued; however, the rash persisted. The patient returned to the ED 3 days later. Several classmates had tested positive for influenza B. His unrelated, adopted brother had similar symptoms, without a rash. The patient is also adopted, with an unknown family history, and lives on a farm. He is up-to-date on childhood vaccinations, but did not receive the vaccine against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2).

On examination, he had temperature 37.1 C, heart rate 125, respiratory rate 30, blood pressure 105/64. He appeared fussy and coughed incessantly. His rash consisted of 1 mm monomorphic erythematous papules coalescing into sandpaper-like plaques over the face, chest, abdomen, groin, back, and buttocks. The plaques were confluent on his cheeks, with perioral sparing. These papules were scattered over proximal upper and lower extremities, and were pruritic. His tonsils were enlarged and erythematous bilaterally, but no other oral lesions or pharyngeal exudate were noted. There was no conjunctival injection, strawberry tongue, mucosal ulcerations in groin, involvement of palms or soles, or hand, feet, or face swelling. Lungs were clear. He had tender cervical and occipital lymphadenopathy.

Dr Meyer, what is your differential diagnosis? What further work-up would you pursue? When would you admit this child?

Evaluation of a child with fever and rash can be challenging. The differential diagnosis is broad, ranging from benign, self-limited to life-threatening diseases. Notable history included his febrile illness starting approximately 2 weeks before presentation, upper respiratory symptoms, and multiple sick contacts, all of which suggested an infectious or postinfectious illness. Many viruses are associated with exanthems: parvovirus, enteroviruses, human herpesvirus 6, rubella, measles, and varicella. Infectious mononucleosis, caused by Epstein-Barr virus (EBV) or cytomegalovirus (CMV), common causes of fever in childhood, can be associated with rash, especially following administration of ampicillin or amoxicillin. Mycoplasma pneumoniae infection can also cause atypical pneumonia with rash in this age group.

Given the sandpaper-like rash, scarlet fever was considered; however, streptococcal testing was negative. Acute rheumatic fever, another group A Streptococcus sequela, can cause fever and rash, but the rash consists of nonpruritic patches with central pallor (erythema marginatum), unlike our patient’s.

Tick-borne diseases, like Lyme disease and Rocky Mountain spotted fever, are infectious causes of fever and rash. This patient’s rash was not typical of these. He presented in winter, atypical for tick-borne disease.

Kawasaki disease (KD) can be considered in children with rash and fever. KD, a clinical diagnosis, requires fever (>38.5 C) for at least 5 days and 4 of 5 of the following: bulbar conjunctival injection, oral changes, peripheral extremity changes, rash, and cervical lymphadenopathy (>1.5 cm).¹  Multisystem inflammatory syndrome in children (MIS-C) should be considered. This patient had multisystem involvement and an apparent viral illness, which might have been coronavirus disease 2019 (COVID-19), a few weeks prior.

Here, the patient was nontoxic appearing and had features suggestive of viral illness. Viral exanthem was the leading diagnosis, care for which is primarily supportive. This patient was breathing comfortably without supplemental oxygen, drinking well, and was deemed clinically stable enough to discharge home by the ED physicians. Hospitalization was not indicated.

A respiratory pathogen panel and COVID-19 real time-polymerase chain reaction (RT-PCR) returned negative. Complete blood count revealed leukocytosis with neutrophilia and elevated atypical lymphocytes. Transaminases and renal function were within normal limits (Table 1). The patient was discharged from the hospital with instructions for supportive care. Antistreptolysin O titer, parvovirus B19 IgG and IgM, CMV IgG and IgM, EBV panel, and Mycoplasma pneumoniae IgG and IgM were sent and pending at discharge.

The next evening, he represented to the ED with worsening rash (Fig 1, rash day 6), continued fever, abdominal pain, and poor oral intake. The ED learned the patient traveled to a popular theme park in the Southeastern United States about 2 weeks before symptom onset. He denied history of insect bites. The ED also learned the patient began treatment with oxcarbamazepine for severe mood disorder 2 weeks before symptom onset. His pediatrician had discontinued oxcarbamazepine when he first developed the rash.

The patient was admitted to the Pediatric Hospitalist service. With history of a recent theme park visit, fever, and morbilliform rash, there was concern for measles. He was immediately moved to a negative pressure isolation room and hospital epidemiology was notified. Dermatology and Infectious Disease were consulted.

Vitals showed temperature 38.2 C, heart rate 130, blood pressure 91/47, respiratory rate 32, and oxygen saturation 95% on room air. Physical exam revealed monomorphic pink- to salmon-colored papules coalescing into sandpaper-like plaques almost confluent over the face with perioral sparing, trunk, groin, and buttocks, and scattered over upper and lower extremities (Fig 1, rash day 7). His tonsils were enlarged and erythematous bilaterally, with no other oral lesions or pharyngeal exudate. He had no Koplik spots, conjunctival injection, strawberry tongue, mucosal ulcerations in the groin, vesicles, bullae, urticaria, or involvement of palms or soles. His tender cervical and occipital lymphadenopathy persisted. He had hepatosplenomegaly. He continued to have mild leukocytosis to 10.46 K/uL with neutrophilia and elevated atypical lymphocytes but no eosinophilia. Transaminases, blood urea nitrogen, and creatinine remained within normal limits with mild hyponatremia to 135 mmol/L (Table 1). Erythrocyte sedimentation rate was normal; C-reactive protein was elevated. He tested negative for SARS-CoV-2 by qualitative real time-polymerase chain reaction (qRT-PCR), but positive for anti-SARS-CoV-2 spike protein IgG and negative for SARS-CoV-2 nucleocapsid protein IgG. Lactate dehydrogenase was mildly elevated. The ferritin, troponin, and creatinine kinase were within normal limits. D-dimer was high at 1006 ng/mL DDU, with fibrinogen within normal limits. Rubeola IgM and IgG, and bacterial blood cultures were drawn. A chest x-ray was negative for pneumonia and intrathoracic adenopathy.

Dr Zeichner, what is the significance of a possible measles diagnosis?

We must not only consider diseases that pose high risk to the patient, but also diseases that pose a public health risk. Fever, ill appearance, and morbilliform rash with an exposure opportunity triggered concerns that this patient might have measles, initiating a strong institutional response.

Measles (reviewed in Strebel 2019²  and Bester 2016³ ) is extraordinarily contagious, with an estimated R₀ (reproduction number, average number of secondary infections) of 12 or higher.⁴  The number of US measles cases remains low, but has increased recently because of inadequate vaccine coverage and case importation from endemic regions. Measles treatment is supportive, although vitamin A is helpful. Some experts suggest treating with ribavirin for severe disease.^5,6 

This patient received measles vaccination, but breakthrough cases in vaccinated persons occur.⁷  A notable outbreak occurred in 2015 in Disneyland.⁸  Among persons infected were immunocompromised children and babies too young to have received a vaccine. Large numbers of park attendees traveling from disparate points and mingling in close quarters raised substantial concerns about measles transmission in theme parks. The event triggered significant changes in California’s vaccination laws.⁹ 

The theme park that this patient visited hosts about 3 million visitors annually. Its policies changed several times during the pandemic; when the patient’s family visited in December 2021, a “play safe” policy was in effect under which fully vaccinated visitors were not required to wear masks and physical distancing was not required, although the park did not verify vaccination status. Reports from about the time of the patient’s visit indicate that few visitors wore masks or practiced social distancing. The COVID-19 pandemic has had secondary effects, with declining primary care pediatric visits and childhood vaccinations, including measles vaccination.¹⁰  These observations augmented concerns. When the diagnosis of measles was entertained, hospital epidemiology was alerted, the patient was placed in a negative pressure isolation room, and samples for measles testing were urgently obtained.

Dr Snavely, could you discuss the considerations behind measles laboratory testing, along with interpretation of results in vaccinated individuals like our patient?

Confirmation of measles disease requires either virus isolation, molecular detection of viral nucleic acid, or a positive serologic test: positive measles IgM, IgG seroconversion, or 4-fold rise in IgG titer between acute and convalescent sera. Specimens for measles diagnosis should be collected at presentation, along with the patient’s mumps, measles, and rubella vaccination status and date of rash appearance. For laboratory testing, sample collection timing relative to rash onset is important. False-negative laboratory results may occur when a specimen is collected too early or too late.^11,12 

Both molecular and serologic testing is performed when measles is suspected. Detecting measles RNA using real-time RT-PCR has greatest sensitivity during the first 3 days after rash onset, when measles IgM may not be observable.^11,12  Although RT-PCR sensitivity declines after 5 to 7 days,¹¹  RNA can be detected up to 2 weeks after rash appearance.¹³  Nasopharyngeal swabs are the preferred specimen type. Throat swabs are also acceptable; urine can be tested if necessary.^11,12 

In unvaccinated individuals, measles IgM antibodies are not detectable until 3 days after rash onset and are rarely present after 6 to 8 weeks. IgG production occurs a few days after the IgM response (7–10 days post rash) and persists for life.^11,12  Some vaccinated persons may not have detectable IgM responses when exposed to measles; molecular testing or testing of a second serum specimen to monitor for rising IgG titers is needed in such cases.^11,12  If a patient received the measles, mumps, rubella vaccine within 6 to 45 days of presenting with a measles compatible disease, viral genotyping and/or IgG avidity testing can be used to distinguish disease from vaccination.^11,12  Given the low prevalence of measles in the United States, false-positive or nonspecific IgM results occur. Specimens for testing should only be submitted from patients with clinically compatible disease. If a false-positive IgM result is suspected, a second serum sample collected 10 to 30 days after the first should be tested for measles IgM and IgG.¹² 

Measles is a notifiable disease. Laboratory-confirmed cases should be reported to public health authorities within 24 hours. Measles disease is unlikely in this patient, as measles RNA was not detected in oropharyngeal or nasopharyngeal swabs and the measles IgM result was negative.

Dr Zeichner, now that measles has been excluded, what are other entities on your differential diagnosis?

For this ill patient with a morbilliform rash, our infectious and periinfectious differential diagnosis also included MIS-C, Kawasaki disease, parvovirus infection, and Rickettsial diseases. Our noninfectious differential diagnosis further included drug reaction with eosinophilia and systemic symptoms (DRESS) syndrome, neoplastic disease such as acute lymphocytic leukemia, and autoimmune diseases. Here I discuss the infectious disease differential. Other consultants discuss other possibilities.

MIS-C, a potentially life-threatening hyperinflammatory syndrome, occurs at a low rate (∼164–224 per million cases of COVID-19) in children with recent COVID-19.^14–20  MIS-C was initially confused with KD and toxic shock syndrome, but the diseases have a number of differences. The true incidence of MIS-C is difficult to determine, because many pediatric COVID-19 cases are asymptomatic.¹⁹  The WHO case definition for MIS-C includes at least 2 of the following: rash, conjunctivitis or mucocutaneous inflammation; hypotension; cardiac disease; coagulopathy; or acute gastrointestinal problems.²¹  Neurologic and neuropsychiatric features have also been described.^22,23  Recommended management for MIS-C includes treatment with intravenous immunoglobulin (IVIG) and corticosteroids.²⁴ 

This patient had no prior history of vaccination against COVID-19, he had IgG against SARS-CoV-2 spike protein but not nucleocapsid protein, and his SARS-CoV-2 qRT-PCR test was negative. Serological assays for SARS-CoV-2 test for immune responses against 2 viral proteins: the nucleocapsid protein and the spike protein. All vaccines used in the United States result in the production of only the spike protein. Antibodies to the nucleocapsid only result from natural infection, and those antibodies can wane by 16 weeks. The receptor binding domain of the spike protein is the main target for neutralizing antibodies.^25,26  Testing positive for the anti-SARS-CoV-2 spike protein IgG suggests this patient had previously been infected with SARS-CoV-2, but the lack of detectable antinucleocapsid antibodies is consistent with a more distant infection. An exposure to SARS-COV-2 with clinical symptoms starting 13 days before, with negative SARS-CoV-2 qRT-PCR and positive serology, could be consistent with MIS-C. He could have been exposed to SARS-CoV-2 at the theme park.

Kawasaki disease (KD) (mucocutaneous lymph node syndrome) was considered, as discussed above. The clinical spectra of KD and MIS-C can overlap. KD often follows an episode of mild gastrointestinal and/or respiratory symptoms. Beyond the rash, this patient had oral mucosal changes, but less notable than those in KD (eg, no strawberry tongue or cracked, red lips), some cervical adenopathy, and no extremity changes. He did not have other laboratory findings seen with KD, for example thrombocytosis. It is unlikely he met formal fever criteria for KD. KD treatment, which prevents coronary artery disease, is IVIG.

Rickettsial diseases can present with fever, a rash, and systemic illness. This child had potential exposures to fields and forests. Rocky Mountain Spotted Fever (RMSF) (reviewed in Dantas-Torres 2007²⁷ ) is a common diagnosis among patients seen in our hospital, but the rash this patient had was not typical. It was morbilliform rather than petechial, and initially did not involve palms and soles. Laboratory features were not supportive: there was no hyponatremia and no thrombocytopenia.

The patient received 2 g/kg IVIG, divided over 2 days, given concerns for MIS-C or an atypical form of KD, after all serologies were obtained.

The next morning, he continued to be febrile and developed head and neck edema along with faint macular erythema on the palms and soles (Fig 1, rash day 8). Rheumatology was consulted and recommended treatment with corticosteroids in addition to IVIG, given concerns for MIS-C. Echocardiogram was normal. He was started empirically on doxycycline until RMSF IgG and IgM results were negative (<1:64). Later that day, he was noted to have eosinophilia and transaminitis (Fig 2), and Allergy was consulted, given concerns for DRESS syndrome, with oxcarbazepine as the potential culprit drug that was started 3 weeks prior.²⁸ 

Dr Zlotoff, what exam findings would allow you to distinguish between DRESS, measles, and RMSF?

Cutaneous findings in DRESS syndrome vary, but its most common presentation is erythematous and morbilliform. Morbilliform means “measles-like” in Latin; the rashes of measles and DRESS are similar.

Classically, a measles exanthem begins on the third or fourth day of fever with blanchable, erythematous macules and papules appearing first on the forehead, neck, hairline, and postauricular areas in a “paint-bucket” distribution, as if a bucket of paint was poured over the head. The rash spreads to the trunk and extremities over the next 2 to 3 days, reaching the feet by day 3. The lesions then fade slowly over a week to copper- or brown-colored macules.

Measles also has an enanthem: clusters of small, punctate, bluish-white papules on an erythematous base on buccal mucosa near the upper molars. These Koplik spots, pathognomonic for measles, often appear on a background of reddish granular mucosa, resulting in a distinctive “grain of salt on a red background” appearance.²⁹ 

Cervical and generalized lymphadenopathy are features of measles and DRESS. Although facial edema can occur in measles, it is more frequent and impressive in DRESS, with significant periorbital and midfacial edema. The morphology, distribution, and progression of the rashes is similar. There can be mucosal cheilitis, erosions, or erythema in DRESS.

Reports describe 60% of adult DRESS rashes as morbilliform³⁰ ; nearly 90% of pediatric DRESS cases have morbilliform morphology.³¹  Other cutaneous presentations of DRESS include vesicles, pustules, bullae, atypical targetoid plaques, epidermal necrolytic desquamation, and even purpura or erythroderma.³² 

RMSF begins with erythematous blanching macules on ankles, wrists, and in 65% of patients, palms or soles. The lesions then spread centripetally to the trunk. Although the eruption is classically considered petechial, only 60% of children with RMSF manifest petechiae as a later finding, after 5 days of illness. Reliance on petechiae to diagnose RMSF can delay treatment. RMSF should be considered in any sick patient from endemic areas with fever and acral rash in summer; doxycycline should be started immediately until tickborne illness is ruled out. Delayed initiation is associated with greater mortality.³³ 

The patient’s rash continued to spread, becoming more confluent on his extremities (Fig 1, rash day 9). Given the initiation of oxcarbazepine 3 weeks before onset of a morbilliform rash, significant head and neck edema, and new rising eosinophilia and transaminitis, he was diagnosed with DRESS. He began receiving intravenous methylprednisolone (1 mg/kg) for treatment of DRESS, and IVIG was discontinued. Of note, the eosinophilia did not appear earlier in the disease course (Fig 2).

Dr Borish, please describe important clinical and laboratory findings in DRESS.

DRESS presents a diagnostic challenge because of its variable clinical presentation and course. Symptoms typically begin 2 to 6 weeks after initiation of the medication,³²  so it is important to obtain an accurate and detailed medication history starting 2 to 3 months before presentation to cover that timeframe. Skin manifestations are nearly universal and generally are the first symptom recognized, but systemic symptoms, such as fever, lymphadenopathy, pruritus, and malaise present early and may predate the rash.³⁴ 

Internal organ involvement is present in most cases. The liver is most frequently involved, followed by lung, kidney, and thyroid.³⁵  Liver injury pattern can be variable with a cholestatic pattern most common, but hepatocellular patterns occur in younger patients.³⁶  Fulminant hepatitis leading to hepatic necrosis is a feared complication, primarily responsible for the estimated mortality rate of 10% in adults.³²  This complication is less commonly seen in children, who generally have better outcomes with a mortality rate of 1%.³⁵  Kidney manifestations are variable, ranging from proteinuria to acute interstitial nephritis to acute renal failure. Pulmonary involvement characteristically manifests with a dry cough and dyspnea. Imaging can show interstitial infiltrates.³²  This patient’s incessant cough, which initially raised concern for SARS-CoV-2, measles, or Mycoplasma pneumoniae, was likely because of pulmonary involvement of DRESS.

The laboratory abnormality timeline varies. Abnormalities include eosinophilia of greater than 1500 cells/µl and atypical lymphocytosis.³⁴  Eosinophilia can be transient or delayed for 1 to 2 weeks from symptom onset. The acute phase is associated with hypogammaglobulinemia, decreased natural killer and B lymphocytes, and increased T regulatory cells. Human herpesvirus-6 (HHV-6) reactivation has been implicated in DRESS pathogenesis and can usually be detected 2 to 3 weeks after symptom onset.³⁷  Other herpesviruses, EBV, HHV-7, and CMV, can reactivate in DRESS.³⁸  Although HHV-6 reactivation typically accompanies DRESS syndrome, it is not required to make the diagnosis. Some investigators advocate that testing for HHV-6 DNAemia be included in the criteria for DRESS diagnosis,³⁰  but others do not.³⁹  Since our treatment team felt secure in the clinical diagnosis and determination of HHV-6 DNA would not have altered management, HHV-6 testing was not pursued in this case.

The essential intervention is removal of the implicated medication. However, DRESS can worsen even after the inciting agent is discontinued, as happened in this patient. Continued worsening after a drug is stopped does not rule out that the drug caused DRESS.

Evidence-based treatment guidelines are wanting. For mild disease without internal organ involvement, symptomatic management with topical steroids is reasonable. With internal organ involvement, systemic steroids are employed, using a starting dose of 1 mg/kg per day of prednisone, with a prolonged 2 to 3 month taper or longer. Relapses are common if steroids are tapered rapidly. Other options include cyclosporine, mycophenolate, rituximab, plasmapheresis, or IVIG.^40,41  IVIG, when used, is typically added as a second agent in conjunction with systemic corticosteroids.⁴²  IVIG as monotherapy in DRESS, to our knowledge, has not been studied. Its use as monotherapy in this case for several days before initiation of systemic corticosteroids may explain why it was ineffective. Antiviral therapy’s role remains unclear. For viral reactivation with viral-induced end-organ damage, treatment with valganciclovir would be appropriate.⁴³ 

Over the next days, the patient transitioned from intravenous methylprednisolone to oral prednisone. By discharge day, his rash had substantially resolved, except for postinflammatory hyperpigmentation of legs (Fig 1, rash day 12). His laboratories were stable with absolute eosinophil count 2.28 K/uL, AST 64 U/L, and ALT 72 U/L.

Dr Makin, what considerations should be made regarding psychiatric medication use in this patient? Is HLA genotyping useful?

A few medications are responsible for most DRESS cases. Antiepileptics are the most commonly implicated group in children.³⁵  Patients who have experienced an episode of DRESS should avoid high-risk medications for DRESS for the rest of their life if possible: allopurinol, carbamazepine, oxcarbazepine, phenytoin, lamotrigine, phenobarbital, sulfasalazine, dapsone, cotrimoxazole, sulfadiazine, vancomycin, minocycline, nevirapine, rifampin, ethambutol, isoniazid, pyrazinamide, and mexiletine.^40,44  New medications should be avoided until resolution of DRESS.

Polymorphisms in HLA class I and II alleles are associated with higher risks of developing DRESS for specific medications. HLA allele A∗31:01 is associated with developing DRESS secondary to carbamazepine in white and Han Chinese populations. The utility of HLA alleles as a screening tool for DRESS has low predictive value.⁴⁰  Here, HLA screening has no use because the association with oxcarbazepine is clear. Testing would not change the recommendation to avoid other high-risk medications.

Our case presented a challenge. A previously healthy child presented with a morbilliform rash and infectious symptoms with many potentially confounding factors: living on a farm, exposure to sick contacts, recent visit to a theme park during a pandemic and reduced rates of childhood vaccinations, and use of an antiepileptic medication (Fig 3). The differential diagnosis for a morbilliform rash is broad. In this patient’s case, time revealed the diagnosis of DRESS when the patient developed eosinophilia and transaminitis several days into his hospital course. It is important to note the variable timeline of laboratory abnormalities in DRESS, including the late appearance of eosinophilia; be aware of all organ systems that can be affected, including liver, lung, kidney, and thyroid; and understand that the disease can continue to worsen even after discontinuation of the inciting drug.

We thank the patient and his family for allowing us to present this case; all of the clinical staff who helped care for the patient; and Dr Rachel Moon for her review and helpful suggestions for this manuscript.

CMV

cytomegalovirus

DRESS

drug reaction with eosinophilia and systemic symptoms

EBV

Epstein-Barr virus

ED

emergency department

HHV-6

human herpesvirus-6

IVIG

intravenous immunoglobulin

KD

Kawasaki disease

MIS-C

multisystem inflammatory syndrome in children

RMSF

Rocky Mountain spotted fever

SARS-CoV-2

severe acute respiratory syndrome coronavirus 2