News

The Glucocorticoid Toxicity Index: Measuring Change in Trials and Practice

For decades, nearly 70 distinct toxicities have been attributed to long-term glucocorticoid use. Yet, until the Glucocorticoid Toxicity Index (GTI) arrived, clinicians had no standardized, validated instrument to capture the burden on the human body across a patient's treatment course. The gap mattered: a clinical worry without a number attached to it cannot be managed, priced, or written into a label claim.


What the GTI measures and how

The GTI is a clinician-reported outcome assessment (ClinRO) developed to provide a weighted, quantitative score of steroid-toxicity between two points in time. That temporal design is deliberate. Earlier attempts to document glucocorticoid harm focused on treating toxicity as a fixed state rather than a trajectory. The GTI was built to recognize and register whether a patient's burden worsened or improved from one clinic visit to the next, or from a trial's baseline to its endpoint. The instrument was assembled by 17 international physician investigators across 11 subspecialties, from rheumatology and nephrology to dermatology, ophthalmology, and psychiatry. They used multi-criteria decision analysis via the 1000Minds software platform (Dunedin, New Zealand) to weight each domain as minor, moderate, or major and to weight it by consensus.[1]


The output is two comprehensive scores. The cumulative worsening score (CWS) captures all toxicities that occur during the assessment period. The aggregate improvement score (AIS) captures all toxicities that occur MINUS those that resolve. Together they give clinicians and trial investigators a composite view of a patient's steroid-toxicity profile shifts over time.


John H. Stone, MD MPH, Professor of Medicine at Harvard Medical School, and the Edward A. Fox Chair in Medicine at the Massachusetts General Hospital, who led the team, described the central purpose in commentary accompanying a 2022 paper published in Seminars in Arthritis and Rheumatism, which provides the dual scoring approach and minimum clinically important difference (MCID). The MCID confirms which changes in steroid-toxicity are not only statistically significant but clinically important.[2]


That MCID matters. A numerical change in a GTI score is not sufficient on its own; the MCID indicates whether that change crosses the threshold a clinician would recognize as meaningful. The estimate was derived from 510 patient assessments by 34 clinicians, providing a distribution-based anchor for interpreting trial results.[2]


Validation and change-sensitivity

The GTI's validation spanned several years and involved a real-world cohort of 101 patients with severe asthma beginning treatment with mepolizumab, an anti-interleukin-5 receptor antagonist.[3] The GTI demonstrated strong correlation with two asthma-specific patient-reported outcome measures: asthma-related quality of life and the St. George's Respiratory Questionnaire.


By 2026, the GTI had been deployed as a pre-specified endpoint in phase 3 trials across several diseases. In REPLENISH, secukinumab in polymyalgia rheumatica, the GTI recorded a clinically meaningful reduction in steroid-toxicity across every health domain while the drug doubled sustained remission versus placebo.[4] In INDIGO, obexelimab in IgG4-related disease, the GTI showed toxicity reductions across the majority of domains alongside a 56% cut in disease flares.[9] In GCAptAIN, secukinumab in giant cell arteritis, the trial missed its primary remission endpoint, yet the GTI still showed significantly lower steroid-toxicity in the higher-dose arm, with a clear dose-response.[10]


Quantifying the risk-benefit ratio of steroids in myasthenia gravis extends this validation work into a real-world claims setting. Using data from the US-based Optum electronic health record database, researchers from Harvard Medical School, Argenx, ZS Associates, and Steritas applied the GTI-MD to 377 steroid-initiating myasthenia gravis patients matched against 305 steroid-naive controls. The CWS was 22.6 in the steroid-initiated group against 18.7 in the steroid-naive group (p < 0.05). Among patients who received multiple steroid courses rather than one, the CWS rose further to 25.7 versus 19.2 for single-course patients (p < 0.05), with AIS values of 9.6 and -0.3, respectively.[5] The findings confirmed the GTI-MD's ability to detect dose-dependent worsening from electronic health record data at scale.


By 2024, more than 45 studies had licensed the GTI, including 12 phase 3 clinical trials, across a range of inflammatory conditions. Paul Brunetta, MD, Chief Medical Officer, Hinge Bio, and a member of the Steritas Advisory Board, stated the practical conclusion plainly:


"Anywhere that steroids are being used in clinical studies, especially randomized controlled studies, there is a place for the GTI."


Integrating measurement into routine practice and trials

The full GTI is essential for clinical trial settings, but less practical in time-pressured clinical care settings. The solution is the GTI-MD (Metabolic Domains), an abridged version developed by John Stone and colleagues and described in a study published in The Lancet Rheumatology.[7] The analysis identified four metabolic domains and separated active treatment and control arms: body mass index, glucose tolerance, blood pressure, and lipid metabolism.[7] These four domains are routinely captured in a clinic visit, which means the GTI-MD can run in the background of an existing workflow rather than requiring additional clinical time.


An independent signal of the GTI-MD's domain selection comes from a New England Journal of Medicine analysis of 1,518,028 participants by the Global Cardiovascular Risk Consortium, which found that 57.2% of cardiovascular disease cases in women and 52.6% in men are attributable to five modifiable factors: body mass index, systolic blood pressure, non-high-density lipoprotein cholesterol, diabetes, and smoking.[8] Four of those five map directly to the GTI-MD domains, and the risk factors for steroid-toxicity and cardiovascular disease overlap article examines what that convergence implies for both measurement and prevention.


Frank Buttgereit, MD, among the researchers contributing to steroid-toxicity work in GCA and PMR, has documented why dose reduction without a direct toxicity measure is incomplete, a position detailed in the steroid-sparing is not enough discussion of why GCA trials need toxicity data.


The GTI Family of COAs now supports trials driving the next wave of steroid-sparing innovation across multiple therapeutic areas. Martha N. Stone, CEO of Steritas, an IQVIA business, described the moment after EULAR 2023:


"Just a year ago, the CEO of one of our clients wondered whether the world was really ready to taper steroids, and now just one year later, that same CEO told me he thought we were at a tipping point."


The instrument behind the argument

The entire case for steroid-toxicity measurement rests on an instrument that is validated, change-sensitive, and deployable across settings. Without the GTI and its derivatives, the argument stays theoretical. With them, it becomes a trial endpoint, a clinical workflow module, and a health economics variable.


For clinicians managing steroid-treated patients, the GTI-MD offers a practical solution: metabolic data already being collected and calculated in the background can alert the physician when emerging toxicity crosses a threshold requiring attention. For drug developers building the steroid-sparing case into a target product profile, the GTI provides the structured endpoint to enable and defend a steroid-toxicity reduction label claim. For health economics and outcomes research teams working on measuring what matters: steroid-toxicity and patient outcomes, the instrument is now the starting point, not the aspiration.


References

  1. Miloslavsky EM, Naden RP, Bijlsma JWJ et al. Development of the Glucocorticoid Toxicity Index (GTI) using multicriteria decision analysis. Ann. Rheum. Dis. (2017). 76:543-546. DOI: 10.1136/annrheumdis-2016-210002

  2. John H. Stone, P. Jane McDowell, David R.W. Jayne, Peter A. Merkel, Joanna Robson, Naomi J. Patel, Yuqing Zhang, Huibin Yue, Pirow Bekker, Liam G. Heaney, The glucocorticoid toxicity index: Measuring change in glucocorticoid toxicity over time, Seminars in Arthritis and Rheumatism, Volume 55, 2022, 152010, ISSN 0049-0172, https://doi.org/10.1016/j.semarthrit.2022.152010

  3. McDowell PJ, Stone JH, Zhang Y, Honeyford K, Dunn L, Logan J et al. Quantification of glucocorticoid-associated morbidity in severe asthma using the Glucocorticoid Toxicity Index. The Journal of Allergy and Clinical Immunology: In practice. 9: 365–372. 2021 https://doi.org/10.1016/j.jaip.2020.08.032

  4. Stone JH, Buttgereit F, Saraux A, et al. Phase 3 Trial of Secukinumab in Polymyalgia Rheumatica. New England Journal of Medicine. 2026; Published online June 3, 2026. https://doi.org/10.1056/NEJMoa2602567

  5. Phillips G, Stone JH, Qi CZ, Stone M, Gelinas D, Chamberas A, Amirthaganesan D, Kulkarni R and Whangbo A. Adaptation of the Glucocorticoid Toxicity Index-Metabolic Domains to Electronic Health Records to Evaluate Steroid Toxicity in Adults with Myasthenia Gravis in the United States. Value in Health, Volume 27, Issue 6, S1 (June 2024)

  6. Paul Brogan, Ray Naden, Stacy P. Ardoin, Jennifer C. Cooper, Fabrizio De Benedetti, Jean-Francois Dicaire, Despina Eleftheriou, Brian Feldman, Jon Goldin, Seth E. Karol, Fiona Price-Kuehne, David Skuse, Constantine A. Stratakis, Nicholas Webb, John H. Stone, The Pediatric Glucocorticoid Toxicity Index, Seminars in Arthritis and Rheumatism, 14 July, 2022 152068, https://doi.org/10.1016/j.semarthrit.2022.152068

  7. Patel NJ, Jayne, DRW, Merkel PA, Bekker P, Zhang Y, McDowel PJl, Johal J, Heaney LG, Murrel D, Stone MN, Yue H, Stone JH, The Glucocorticoid Toxicity Index-Metabolic Domains, an abridged version of the Glucocorticoid Toxicity Index: post-hoc analysis of data from the ADVOCATE trial, The Lancet Rheumatology, Volume 5, Issue 7, 2023, e413-e421, https://doi.org/10.1016/S2665-9913(23)00131-5

  8. The Global Cardiovascular Risk Consortium. Global effect of modifiable risk factors on cardiovascular disease and mortality. N Engl J Med (2023) 389:1273-1285.

  9. Della-Torre E, Baker MC, Zhang W, et al; INDIGO Trial Investigators. Obexelimab for the Treatment of IgG4-Related Disease. N Engl J Med. 2026.https://doi.org/10.1056/NEJMoa2601337

  10. Stone JH, Venhoff N, Buttgereit F, et al. Secukinumab for Giant Cell Arteritis. NEJM Evidence. Published online June 3, 2026. https://doi.org/10.1056/EVIDoa2600112

Categories: