Background/Purpose: A Glucocorticoid Toxicity Index app (GTI 2.0) is now used as a clinical trial outcome measure in adults, but glucocorticoid (GC) toxicity issues in children differ from those relevant to adults in a number of respects. We sought to design a pediatric GTI (pGTI) for use as an outcome measure in clinical trials and other settings to assess the impact of treatment interventions on GC toxicity.

Methods: Sixteen experts from 7 subspecialties participated. Six were from the U.S. and 10 from Europe, Canada, or New Zealand. Group consensus methods and multi-criteria decision analysis (MCDA) were employed. The pGTI consists of a Composite Index and a Specific List. Toxicities included in the Composite GTI occur commonly and may vary with GC exposure. The Composite GTI reflects GC toxicity with the potential to change over time: to worsen if GC doses increase, or to improve if successful GC sparing is achieved. The Composite pGTI was evaluated by application to paper cases derived from real-life cases. Weights for individual items of the Composite pGTI were assigned through MCDA, facilitated by 1000Minds software. Similar MCDA approaches were used in the development of the adult GTI and ACR/EULAR Classification Criteria for several rheumatic diseases (RA, SLE, SSc). In contrast, the Specific List describes clinical events that are less common but often severe and typically not reversible by GC reduction. Items of the Specific List are not weighted but rather recorded as a cumulative index of glucocorticoid-related damage.

Results: The complexity of longitudinal GC toxicity assessment in children lends itself well to app technology, which offers significant advantages in ease of calculation. The impact of normally-anticipated growth on several pGTI domains poses particular challenges addressed effectively by app technology. The pGTI Composite Index consists of 10 toxicity domains: body mass index (BMI), growth, glucose tolerance, lipid metabolism, systolic blood pressure, bone mineral density, GC-induced myopathy, skin toxicity, neuropsychiatric impact, and infections. Weighting of items of the Composite GTI permits the calculation of cumulative and aggregate scores. The pGTI weights improvement in GC toxicity within a given domain as highly as worsening: improvements yield negative toxicity scores; worsening, positive scores. As an example, an increase in BMI of >2 but <5 BMI units is associated with an increase in the cumulative GTI of +21 points. In contrast, decline in BMI to a corresponding degree is associated with a reduction in that domain of -21 points. All pGTI scores are calculated based on a change in toxicity between two visits rather than an absolute score at one point in time. The Specific List is designed to capture GC toxicity not included in the Composite GTI. The Specific List includes six additional domains addressing features of GC toxicity such as pubertal delay/sex hormone access interruption, ocular toxicity (cataracts, central serous retinopathy), and bone health (osteonecrosis).

Conclusion: We describe the development and initial evaluation of a comprehensive, weighted index for the assessment of GC toxicity. A pGTI app will be available in the autumn of 2018.

« Back to 2018 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/development-of-a-pediatric-glucocorticoid-toxicity-index/