Raltegravir-associated Drug-Reaction With Eosinophilia and Systemic Symptoms Syndrome in a Pediatric Patient Without Characteristic Human Leukocyte Antigen B*57:01 or B*53:01 alleles

Abstract

We present the first published case of raltegravir-associated drug-reaction with eosinophilia and systemic symptoms (DRESS) syndrome in a child without characteristic human leukocyte antigen haplotypes HLA-B*57:01 or HLA-B*53:01. A 4-year-old African American female with perinatally acquired human immunodeficiency virus infection was hospitalized for DRESS after starting a raltegravir-based antiretroviral regimen.

Drug-reaction with eosinophilia and systemic symptoms (DRESS) syndrome is a rare but potentially life-threatening adverse effect of certain medications, including antiretrovirals. Unlike other drug reactions, DRESS can result in a mortality rate of up to 10% if unrecognized [1]. Within individuals infected with human immunodeficiency virus (HIV), the association between human leukocyte antigens (HLAs) and DRESS syndrome secondary to drugs such as cotrimoxazole and anti-tuberculosis agents have been well studied [2]. More recently, there has been increasing evidence that shows an association between HLA haplotypes and DRESS secondary to antiretrovirals including tenofovir, efavirenz, nevirapine, raltegravir, and abacavir [3]. HLA-C*04:01 and nevirapine-associated DRESS has been reported in Malawian individuals [4]. The association between abacavir-induced DRESS and the HLA-B*57:01 allele is also well established [5]. More recently, a case series was published that reported an association between raltegravir-associated DRESS and HLA-B*53:01 [6], but there is limited literature in pediatric patients, particularly those without characteristic HLA haplotypes.

Here, we report a young pediatric patient with perinatally acquired HIV infection who developed raltegravir-associated DRESS. However, HLA haplotyping did not demonstrate either B*57:01 or B*53:01 alleles.

CASE REVIEW

A 4-year-old African American female with a past medical history of prematurity and perinatally acquired HIV presented to the hospital with fever and a new onset rash of 2 days duration. Her symptoms began approximately 2 weeks after initiation of an antiretroviral therapy (ART) regimen that included raltegravir. Of note, her HIV viral load 1 week prior to initiation of raltegravir was 138 copies/mL (Table 1). The rash began as nonpruritic, erythematous papules on the patient’s face, later spreading to the chest and upper extremities. Soon after her rash developed, she became febrile to 41.3ºC on home tympanic thermometry, prompting further evaluation. Upon initial history, systemic symptoms were notable for decreased appetite and abdominal pain; pertinent negatives included cough, coryza, sore throat, vomiting, and changes in bladder or bowel habitus. Her immunizations were complete for age with the exception of the second measles-mumps-rubella and annual influenza vaccines. Exposure history was unremarkable for introduction of new medications, except for change in ART as detailed below; introduction of new foods; changes in diet; recent animal exposure; changes in hygiene products; exposures to ill contacts, including persons with known measles infection; and recent travel. She had negative Streptococcus pyogenes rapid antigen and culture testing on a pharyngeal swab sample prior to hospital admission.

Table 1.

Review of Laboratory Investigations

A. Hematology and chemistries
Laboratory Results	2 Months Prior to Admission			Day 1	Day 2	Day 3	Day 4	Day 5	Day 6
White blood cells (× 109/L)	4.1			3.4	3.9	...	2.5	3.0	4.1
Segmented neutrophil (%)	34			34	38	...	33	16	14
Band (%)	3			3	7	...	3	...	18
Basophil (%)	...			...	...	...	...	...	1
Eosinophil (%)	3			12	9	...	19	19	8
Lymphocyte (%)	54			48	42	...	40	56	61
Monocyte (%)	6			3	4	...	5	9	4
Variant lymphocyte (%)	...			...	...	...	...	...	4
Hemoglobin (g/dL)	11.2			10.7	9.6	...	8.8	8.9	9.4
Platelets (× 109/L)	197			122	143	...	160	136	152
Aspartate aminotransferase (IU/mL)	36			442	...	182	123	153	179
Alanine aminotransferase (IU/mL)	30			677	...	439	302	267	263
Creatinine (mg/dL)	0.1			0.3	...	0.2	0.2	0.1	0.3
Erythrocyte sedimentation rate (mm/h)	...			...	...	...	...	92	...
C-reactive protein (mg/dL)	...			...	7.41	...	...	...	...
B. Infectious studies
Adenovirus PCR, plasma	Not detected
Blood culture	No growth
Cytomegalovirus IgM and IgG	Not detected
EBV viral capsid antigen IgM and IgG	Not detected
EBV nuclear antigen	Not detected
EBV early antigen IgG	Not detected
Enterovirus PCR, plasma	Not detected
Extended respiratory viral panel (nasopharyngeal swab)	None detected
Hepatitis A IgM	Not detected
Hepatitis B core IgM	Not detected
Hepatitis B surface antigen	Not detected
Hepatitis C total antibody	Not detected
Parvovirus B19 IgM and IgG	Nonreactive
Rapid plasma reagin test	Nonreactive
Toxoplasma IgM and IgG	Not detected
C. HLA typing (peripheral blood)
HLA B 5701 (Luminex method)	Not detected
HLA B locus typing (Luminex method)	B*15:XX (inferred to align with B*15:17, similar to B62) detected; B*57:02 detected
D. CD4 count and viral load (peripheral blood)
	Prior to Hospitalization			Post-Hospital Discharge
	6 Months	4 Months	1–2 Months	2 Weeks	1 Month	2 Months	4 Months a	5 Months
Antiretroviral therapy use	Yes	Yes	Yes	No	No	No	Noa	Yes
CD4 count (cells/mm3)	...	852	727	...	...	640	459	...
Viral load, copies/mL (log)	Not detected	127 (2.1)	138 (2.1)	105 000 (5.0)	62 300 (4.8)	...	215 000 (5.3)	65 (1.8)

Abbreviations: EBV, Epstein-Barr virus; HLA, human leukocyte antigen; Ig, immunoglobulin; PCR, polymerase chain reaction; ..., data not available.

^aAntiretroviral therapy was restarted soon after human immunodeficiency virus type 1 RNA was obtained.

Her ART history was notable for zidovudine prophylaxis from birth until 6 weeks of age. Once the diagnosis of HIV infection was established, she began an ART regimen of abacavir, lamivudine, and combination lopinavir/ritonavir. Despite this regimen, she had a documented history of chronic, low-level HIV viremia suspected to be secondary to poor medication adherence. We could not pursue ART resistance testing as her HIV viral loads were below the threshold for the HIV-1 GenoSure PRIME panel. She had a period of no ART about 1 month prior to the current hospital presentation. The maladherence was attributed to poor tolerance of the liquid-formulated lopinavir/ritonavir component. Therefore, she restarted abacavir and lamivudine, but lopinavir/ritonavir was replaced with raltegravir in an attempt to increase medication adherence. Two weeks later, she developed the aforementioned facial rash.

Upon admission, she remained with a temperature of 39.3ºC and associated tachycardia. Her physical exam was notable for a nontoxic appearance despite a diffuse (>50% skin involvement) nonpruritic, morbilliform rash with confluence over the face and trunk. There were discrete papules on the bilateral upper and lower extremities with palmar involvement but initially sparing her soles. Her physical examination was also notable for soft, nontender, mobile bilateral anterior and posterior cervical lymphadenopathy and hepatomegaly spanning 3 cm below the right costal margin. The remainder of her exam was within normal limits including absence of an audible cardiac murmur.

She initially received a 10-mL/kg normal saline bolus and antipyretics with interval improvement in vital signs. Admission laboratory results were notable for transaminitis, neutropenia, eosinophilia (absolute count, 408 × 10⁶/L [408 cells/μL]), and thrombocytopenia (Table 1, part A). She underwent a broad infectious evaluation (Table 1, part B) that did not yield an etiology; human herpesvirus 6 testing was not performed. Subsequently, DRESS syndrome was more strongly considered. Therefore, ART was discontinued and HLA B typing was performed. Within 24 hours of ART cessation, she developed conjunctivitis, palpable splenomegaly, and a rash that involved the soles of her feet. She remained intermittently febrile with associated tachycardia and mild tachypnea for 5 days, which also prompted evaluation for Kawasaki disease. Inflammatory marker evaluation revealed an elevated erythrocyte sedimentation rate and C-reactive protein (Table 1, part A), but the remainder of her evaluation did not fulfill diagnostic criteria for typical Kawasaki disease. She did not receive aspirin or intravenous immunoglobulin therapy, nor did she receive corticosteroids or antimicrobials at any point during her hospital admission.

By the fourth day of hospitalization, she had defervescence, marked improvement of the rash, and complete resolution of conjunctivitis. However, she continued to have palpable hepatosplenomegaly and ongoing neutropenia with a nadir of 0.48 × 10⁹/L (480 cells/μL). The remainder of the laboratory abnormalities, including thrombocytopenia, eosinophilia, and transaminitis, markedly improved, and she returned to baseline activity and appetite with an otherwise reassuring clinical disposition prior to discharge on hospital day 5. ART was held through discharge and outpatient follow-up. ART was resumed 4 months post-discharge after transaminases had normalized. During this time, the patient’s HIV RNA steadily increased (maximum 215 000 copies/mL [log 5.3]; Table 1, part D). She did not have intercurrent illnesses or recurrence of her symptoms. She resumed ART with tablet formulations of lamivudine, abacavir, and lopinavir/ritonavir. She tolerated this regimen well without recurrence of DRESS syndrome and improved medication adherence.

DISCUSSION

DRESS syndrome is a severe form of drug hypersensitivity that occurs about 2–6 weeks after initiation of the offending drug. In some pediatric cases, however, there is a shorter latency period between drug exposure and onset of DRESS [7]. Three independent groups have proposed different scoring systems to diagnose DRESS: Bocquet et al; J-SCAR; and RegiSCAR [1]. Using the RegiSCAR scoring system [1, 8], the patient discussed in this report met criteria for “probable DRESS syndrome” with fever, lymphadenopathy, eosinophilia, atypical lymphocytes, widespread (>50% body surface area) morbilliform rash, transaminitis, and prolonged time to symptom resolution after discontinuation of the drug. According to the adverse drug reaction probability scale, raltegravir was the probable causative agent with a score of 6 [9]. The patient’s robust infectious and noninfectious evaluation was unremarkable, and she did not develop recurrent symptoms upon restarting abacavir or lamivudine. Moreover, her acute presentation was temporally associated with a novel exposure to raltegravir, all suggestive of raltegravir-associated DRESS as the most likely diagnosis.

Some HLA haplotypes confer a genetic predisposition to DRESS when the patient is exposed to the offending drug [1]. Within the antiretroviral drug classes, this association is best appreciated with abacavir, a nucleoside reverse transcriptase inhibitor. There is an increased risk of abacavir-associated DRESS in white individuals and persons who carry the HLA-B*57:01, HLA-DR7, or HLA-DQ3 allele [5]. Studies have not yet been performed to definitively link HLA haplotypes and integrase strand transfer inhibitor–associated DRESS. However, there have been reports of dolutegravir-associated DRESS or hypersensitivity reaction [10, 11]. Otherwise, an exhaustive review of the literature reveals 6 case reports to date associating DRESS with the initiation of raltegravir [4, 10]. All except 1 of these patients were adults. The majority of the patients developed fever, eosinophilia, cutaneous involvement, and transaminitis as part of the DRESS syndrome. Onset of symptoms ranged from 4 to 8 weeks after initiation of therapy, all required hospitalization, and the majority of patients required initiation of corticosteroids along with cessation of ART. Subsequent time to normalization of laboratory abnormalities, including transaminitis, ranged from 1 to 8 weeks. The only previously reported pediatric case did not undergo extensive HLA genotyping. Although the patient was negative for HLA-B*57:01, it is unknown whether she was a carrier of the other alleles that have previously been associated with raltegravir-induced DRESS [12]. Interestingly, 4 of these 5 individuals were of African ethnicity and 1 was presumably Hispanic (since she recently moved from Mexico). Thomas et al also recognized a linkage between raltegravir-associated DRESS and ethnicity. They collaborated with the respective clinicians and performed HLA genotyping on 3 of these patients and another patient with raltegravir-associated DRESS encountered at their own facility. HLA typing revealed that all 4 patients carried HLA-B*53; 3 of these patients underwent high-resolution haplotyping that helped subclassify the allele as HLA-B*53:01. The majority (75%) of the patients also had HLA-C*04 and HLA-DQB1*04 in common [6]. Using structural features of the HLA-B*53:01 molecule and the raltegravir compound, the researchers illustrated the interaction between both molecules [6]. This suggests a potential antigen-like presentation of raltegravir by the HLA molecule that would mediate the T cell–dependent hypersensitivity reaction that results in DRESS [6]. Reportedly, 20% of individuals with African ethnicity carry HLA-B*53:01 compared with 6% of individuals with Hispanic ethnicity and 0.5% in those with European ethnicity [6]. Our patient underwent HLA typing of the B locus alone but did not carry the B*53:01 allele (Table 1, part C). Although the typing was performed using a Luminex medium resolution HLA test (Luminex Corporation, Texas), it would still be able to identify B*53. She was not evaluated for carriage of either HLA-C*04 or HLA-DQB1*04. These alleles are an important target of future studies, especially with increasing use of integrase inhibitors as part of ART.

It is crucial to consider DRESS on the differential diagnosis for any person who presents with fever, rash, lymphadenopathy, eosinophilia, and transaminitis while taking integrase strand transfer inhibitors. If unrecognized, DRESS can result in a mortality rate of up to 10% [1]. Further studies are needed to determine the association between raltegravir-associated DRESS syndrome and HLA haplotypes, especially B*53, HLA-C*04, and HLA-DQB1*04, particularly in people of African ethnicity.

Notes

Acknowledgments. This work is supported by the following National Institutes of Health awards: B. A. B., Eunice Kennedy Shriver National Institute of Child Health and Human Development (K12-HD000850); C. K. C. and A. B. T., Duke University Center for AIDS Research (5P30 AI064518); J. T. K., Ruth L. Kirschstein National Research Service Award Institutional Research Training (grant T32-AI007062); and Children’s Miracle Health Network (4531787) (S. T.).

Potential conflicts of interest. All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.