Biomarkers and risk factors in flares of aGvHD: MAGIC analysis

Patients who undergo allogeneic hematopoietic cell transplant (allo-HCT) are at risk of developing acute graft-versus-host disease (aGvHD).¹ The current standard first-line treatment for aGvHD is systemic steroids.¹ However, aGvHD symptoms often flare in patients upon tapering or discontinuation of steroids.¹ The lack of data available for incidence, clinical presentations, and outcomes of aGvHD flares, and the absence of a standardized definition for GvHD flares makes it challenging to assess patients at risk of these occurrences.¹

Focusing on the use of serum biomarkers, below, we summarize data evaluating flares of aGVHD in the Mount Sinai Acute GvHD International Consortium (MAGIC) analysis, published by Akahoshi et al.¹ in Blood Advances.

Study design¹

•  A retrospective study of adult and pediatric patients who received allo-HCT at 23 international allo-HCT centers in North America, Europe, and Asia, between 2014 and 2021.
• Patients were eligible for this study if they achieved complete response (CR) or very good partial response (VGPR) to systemic steroid treatment, within 4 weeks without primary disease relapse.
• The definition of aGvHD flare was defined according to expert consensus in the Mount Sinai Acute GvHD International Consortium.
• Serum samples were collected at the initial achievement of CR/VGPR.
• The MAGIC algorithm probability was calculated as a value between 0.001 and 0.999.
• Validated thresholds for Ann Arbor (AA) scores were used (AA1 < 0.14; 0.14 ≤ AA2 < 0.29; AA3 ≥ 0.29).

Key findings¹

• In total, 968 patients were included in this study.
• The cumulative incidence of flares after CR/VGPR at 6 months was 21.6% (95% confidence interval [CI], 19.0–24.2%).
• Patients with Grade 3–4 GvHD, at the time of flare, were more likely to experience non-relapse mortality than those with Grade 1–2 GvHD (47% vs 12%; p<0.001).
• After measuring the MAGIC algorithm probability score in all patients at CR/VGPR, the cumulative incidence of eventual flares increased with each AA score increase (Figure 1.).

AA, Ann Arbor; CR, complete response; MAP, Mount Sinai Acute GvHD International Consortium algorithm predictor; VGPR, very good partial response.
*Data from Akahoshi, et al.¹

• Patients whose steroids were tapered rapidly were nearly twice as likely to experience a flare compared with patients who were tapered slowly (30% vs 17%; p < 0.001).
• In patients who developed Grade 3–4 GvHD, the severity of flare and proportion of patients with lower gastrointestinal involvement (Stage 2–4) increased with each increase of AA score (Table 1).

Table 1. Severity and LGI involvement measured by MAPs at CR/VGPR*

AA, Ann Arbor; CR, complete response; GI, gastrointestinal; LGI, lower GI; MAP, Mount Sinai Acute GVHD International Consortium algorithm predictor; VGPR, very good partial response. *Data from Akahoshi, et al.1
	AA1	AA2	AA3
Severity, %	5.4	11.4	20.0
Lower GI involvement, %	4.9	9.9	18.0

Key learnings1

The use of MAGIC biomarkers at CR/VGPR can help to predict flares of GvHD. The MAGIC biomarkers have the potential to guide future treatment strategies, including steroid tapering, in patients with GvHD.