Safety Data for Baricitinib Use in Children With Severe SARS-CoV-2 Infection Free

The management of SARS-CoV-2 infection evolved rapidly with the development of novel drugs and the repurposing of existing agents. Remdesivir and corticosteroids became critical therapies for severe COVID-19.^1,2 Additional anti-inflammatory therapies, specifically cytokine inhibitors, including tocilizumab, anakinra, and Janus Kinase (JAK) inhibitors (which minimize exponentiated cytokine production),³ were investigated in adults with COVID-19 because steroids were inadequate in some cases.⁴ In November 2020, the US Food and Drug Administration issued an Emergency Use Authorization (EUA) for baricitinib,⁵ a JAK inhibitor,⁶ to be used in hospitalized patients with COVID-19 and a supplemental oxygen requirement. Drug trials investigating baricitinib use in adult patients established early evidence of efficacy and mortality reduction,^7–11 which accelerated clinical improvement compared with remdesivir alone.¹² Although baricitinib use in children with severe COVID-19 was available under this EUA, and there is an ongoing pediatric clinical drug trial (COV-BARRIER [NCT05074420]; a multicenter open-label dose confirmation, safety, and effectiveness study), there has been no published systematic experience of its use in children for this unique indication. The objective of this study is to describe the safety and outcomes following the use of baricitinib for children with severe COVID-19.

This retrospective case series included patients aged younger than 21 years who were treated with baricitinib for severe COVID-19 between January 2021 and October 2022 at 2 US children’s hospitals. The project was reviewed and approved by each institutional review board in accordance with institutional ethical standards and the Helsinki Declaration of 1975 for human experimentation. Investigation dates correspond with the time baricitinib was available and used locally to supplement recommended first-line therapies against SARS-CoV-2 (remdesivir and dexamethasone). Per institutional guidelines, baricitinib could be added for patients who met the definition of severe COVID-19, ie, those with evidence of lower respiratory tract infection requiring supplemental oxygen (or increase from home oxygenation) and satisfied any of the following 3 clinical scenarios: (1) lack of improvement on dexamethasone and remdesivir; (2) contraindication to steroids; or (3) on dexamethasone and remdesivir requiring critical care support, ie, intensive care unit (ICU) admission for any critical care need and not limited to only those who required mechanical ventilation. Patients received venous thromboembolism prophylaxis per institutional COVID-19 guidelines (Supplementary Table 2) unless there was an overt contraindication, ie, bleeding, allergy, etc. Treatment dose anticoagulation was used only when evidence of thrombosis was identified.

Electronic medical record data were collected retrospectively. Demographic data included institution, age, and sex assigned at birth. Patient characteristics included weight, comorbid conditions, dates and times of hospital (and ICU) admission/discharge, initiation and discontinuation of supplemental oxygen and other forms of respiratory support, complications, and discharge disposition. Results of laboratory studies including aspartate aminotransferase (AST), alanine aminotransferase (ALT), cholesterol, triglyceride, complete blood count, creatine kinase, microbiology studies (bacterial, fungal, viral including polymerase chain reaction (PCR) as needed clinically), and diagnostic imaging. Therapeutic data included all medications used to treat COVID-19—remdesivir, corticosteroids, and baricitinib (with start and stop dates), the reason for early discontinuation of baricitinib (if applicable), and medications used to treat secondary infections.

Patients who were immunocompromised¹⁴ were those with hematologic malignancy, bone marrow failure, or solid organ transplantation. Adverse drug events were defined according to the Common Terminology Criteria for Adverse Events (CTCAE) Version 5.¹⁵ Thrombocytosis was defined as a platelet count greater than 600 × 10³/μL. Per EUA recommendations (Table 1), dose interruption was considered when the absolute lymphocyte count was less than 200 cells/μL, the absolute neutrophil count was less than 500 cells/μL, or there was an increase in liver enzymes thought to be related to baricitinib. Discontinuation of baricitinib was considered “early” if it occurred before completion of the recommended 14-day course and was in relation to a possible adverse event or occurred before a patient met the criteria for hospital discharge but had improved from COVID-19 and the medical team deemed baricitinib no longer necessary.

Drug safety outcomes include drug duration, adverse events (evidence of blood cell lineage abnormalities, ie, thrombocytopenia, thrombocytosis, neutropenia, anemia), hepatorenal dysfunction, and dysregulated clotting (ie, deep vein thrombosis [DVT] confirmed by ultrasonography and pulmonary embolism [PE] confirmed by computed tomography angiogram), and incidence and reasons for discontinuation of baricitinib. The primary clinical outcomes were the type and duration of oxygen supplementation, length of hospital stay, and complications, including subsequent bacterial infections, fungal infections, and virus reactivations.

Descriptive analysis and statistics were performed to describe patient demographics, safety data, and patient outcomes. Time variables, such as time to initiation of therapies and durations of therapies or stay (in ICU or hospital), are reported as medians with ranges. Categorical variables such as transaminitis, cytopenia(s), presence of thrombocytosis, DVT, PE, and early baricitinib discontinuation were presented as frequencies (counts and proportions).

There were 37 patients, 32 at hospital #1 and 5 at hospital #2, who received baricitinib for severe COVID-19. All patients had a comorbidity (59% obesity, 5% malignancy) conferring risk for severe COVID-19 (Table 1). Baricitinib was initiated with a median of one day from hospital admission (78% treatment by 24 hours; 84% by 48 hours). The majority of patients received a 4 mg daily dose (n = 30, 81%), with the remainder receiving 2 mg because they were aged younger than 9 years (age range: 1.16–6 years; weight range: 8.8–32.5 kg). In addition to baricitinib, all patients received remdesivir, and 34 (92%) received corticosteroids (Table 2).

Baricitinib was discontinued early for clinical improvement (2) and adverse events (7; 6 elevated liver enzymes, 1 thrombocytosis). Only 1 of 6 patients with elevated liver enzymes met EUA discontinuation criteria as the enzymatic rise from baseline was indicative of drug-induced liver injury; this patient developed a grade 3 for AST and grade 4 for ALT during our investigation. The other 5 liver enzyme-related discontinuations were driven by provider caution. Of these 5 patients with early discontinuation due to provider caution, 1 met the criteria for a grade 2 event (ALT only), 2 met the criteria for a grade 1 event (AST and ALT), and 2 had concerns about rising liver function tests but were not high enough to meet CTCAE criteria. Notably, these 5 patients had elevated liver enzymes before administration of baricitinib, but levels did increase further after baricitinib initiation. There was no discontinuation of baricitinib for anemia or neutropenia. However, for the 11 of 37 patients who developed gradable anemia (CTCAE grade 1–3), 7 (64%) were anemic at presentation before the initiation of baricitinib. The single patient with grade 4 neutropenia had received induction chemotherapy for acute lymphoblastic leukemia in the preceding month and, thus, was not attributed to baricitinib. All adverse events resolved after baricitinib discontinuation (Table 3).

A total of 5 patients had concurrent infections during or after baricitinib use. These infections were compatible with the host’s underlying condition(s) and included the following: patient #1 developed Moraxella spp tracheitis, patient #2 first developed an Escherichia coli urinary tract infection and later an empyema (no causative organism testing was performed from the respiratory tract), patient #3 grew Staphylococcus aureus from a sputum culture and was treated for ventilator-associated pneumonia, patient #4 presented with bacterial pneumonia and oral herpes simplex virus (HSV) reactivation (note: this was the only patient with an infection secondary to COVID-19 who did not require mechanical ventilation), patient #5 presented with Pneumocystis jirovecii pneumonia and had a possible HSV reactivation (positive PCR from admission bronchoalveolar lavage specimen but demonstrated clinical improvement without targeted HSV therapy). All patients with concurrent or secondary infections were critically ill and required mechanical ventilation and/or were immunocompromised (Table 4). Supplementary Table 1 further details each unique patient, the timing of infection with respect to baricitinib administration, and the treatment(s) provided. Baricitinib was not discontinued because of infection for any patient.

There was one DVT without PE in an obese teenage boy. There were no deaths. The median duration of hospitalization was 7 days (range 2–108 days). The median duration of mechanical ventilation days was 4.5 days (range 1–86 days). Despite 76% of patients requiring ICU level care, only 2 of 37 patients (5%) required discharge with a new supplemental oxygen requirement, one of whom received continued care at a long-term care facility. One patient required a tracheostomy and was discharged without oxygen supplementation.

This drug use and safety data of baricitinib use for severe COVID-19 at 2 pediatric US institutions between January 2021 and October 2022 spanned multiple SARS-CoV-2 variants (Table 5). In this brief report, we show that baricitinib was well-tolerated in a high-risk population of children with severe COVID-19. Overall, incidence of discontinuation of drug and adverse events were lower than those reported in the adult COVID-19 literature. Only 19% of patients in this study discontinued baricitinib secondary to any degree of adverse event compared with a discontinuation incidence of 28% in a randomized control trial with baricitinib in critically ill adults. Additionally, the incidence of co- or secondary infection in our patients was acceptable and lower than what has been previously reported (14% in this study vs 70% by Ely et al⁷). Notably, those who presented with co-infection or secondary infections in this study were at high risk for infection because of critical illness with the need for mechanical ventilation or were immunocompromised. Venous thrombus events were rare. Only 1 patient in the pediatric and adult settings developed a DVT. There were no PE in children, but 2 occurred in adults.⁷ 

Baricitinib safety profile comparisons between children and adults are available from randomized placebo-controlled trials for autoimmune conditions such as juvenile idiopathic arthritis¹⁶ and atopic dermatitis.¹⁷ As with these studies, in our case series, patients received similar baricitinib dosages (2 and 4 mg), there were very low rates of thrombotic events, and no deaths were observed. Few adverse events prompted complete discontinuation of baricitinib in the juvenile idiopathic arthritis investigation (1% during the safety open-label lead-in period and 1% during the double-blind period) and in the atopic dermatitis study (0.6%). This is in contrast to our study, which does involve a smaller sample size but reports a slightly higher 5.4% adverse event discontinuation rate (n = 2 for those who met discontinuation criteria; one for thrombocytosis and one for elevated liver enzymes). Compared with our 14% concurrent infection rate, there was a higher incidence of secondary infections in juvenile idiopathic arthritis (25% during the safety open-label lead-in and 38% during the double-blind period) and atopic dermatitis (25.8% to 30.0% across dose groups).

The current study is limited by its retrospective nature, small sample size, and limited number of very young patients; however, it is the only report known in the literature describing the tolerability of baricitinib in multiple pediatric patients with acute COVID-19. In addition, we were unable to assess efficacy because of the lack of a comparator group; however, there were no deaths among treated patients, all of whom were high risk because of major comorbidities.

Baricitinib appears safe for both immunocompetent and immunocompromised children hospitalized for severe COVID-19. Most adverse events did not inhibit therapy administration, and there were no deaths following the initiation of baricitinib. Early baricitinib discontinuation for abnormal laboratory studies (6 of 7; 86%) was predominantly secondary to provider caution. The evaluation of baricitinib use for severe COVID-19 in children is limited because of the relatively small numbers that develop severe infection compared with adults. Additionally, of the children that reach eligible disease states, many are outside the current EUA. Further baricitinib safety and outcome data, particularly in those aged younger than 2 years, is needed to inform the management of severe COVID-19 in children.

We wish to acknowledge Kristen Ryan, RN who assisted with data collection.

ALT

alanine aminotransferase

AST

aspartate aminotransferase

CPK

creatine kinase

CTCAE

Common Terminology Criteria for Adverse Events

DVT

deep vein thrombosis

EUA

Emergency ssUse Authorization

ICU

intensive care unit

IRB

Institutional Review Board

JAK

Janus Kinase

LCH

Le Bonheur Children’s Hospital

PCR

polymerase chain reaction

PE

pulmonary embolism

SJCRH

St. Jude Children’s Research Hospital