Early measurement of ST2 and REG3α has a prognostic impact in haploidentical transplantation with post-transplant cyclophosphamide

Abstract

Research on predictive biomarkers in haploidentical SCT (haplo-HSCT) with post-transplant cyclophosphamide (PTCy) remains limited, particularly regarding the critical first week post-infusion. We prospectively evaluated serum ST2, REG3α, and the MAGIC algorithm pre-conditioning and at days 0, +7, +14, and +21 in 81 adult patients undergoing haplo-HSCT with PTCy.

ST2 and REG3α surged significantly from pre-conditioning to day 0 (p=0.031 and p<0.001). At day 0, elevated ST2 independently predicted inferior OS (p=0.016) and EFS (p=0.002), while REG3α predicted inferior GRFS (p=0.041) in multivariate analysis. Post-infusion, ST2 independently portended higher NRM at days +7 and +14 (p=0.009 and p=0.011), inferior OS across days +7 (p= 0.001), +14 (p= 0.005), and +21 (p= 0.019), EFS at days +7 (p< 0.001), +14 (p= 0.002), and +21 (p= 0.019) and grade II–IV aGVHD at days +14 (p=0.013) and +21 (p<0.001). REG3α independently predicted NRM at days +14 (p=0.028) and +21 (p=0.023), lower OS at days +7 (p = 0.045), +14 (p = 0.007), and +21 (p=0.021) and inferior GRFS at days +14 (p=0.007) and +21 (p=0.006). Longitudinal ST2 trajectory models significantly discriminated NRM (p=0.006), OS (p=0.031), EFS (p=0.046), and aGVHD (p=0.014). Multivariable models confirmed the MAGIC algorithm as an independent predictor of EFS at days 0 (p=0.013) and +14 (p=0.043) and GRFS at day +14 (p=0.025).

This study confirms the relevance of ST2 and REG3α in haplo-SCT and demonstrates the prognostic impact of pre-infusion and day +7 levels across multiple endpoints, establishing the MAGIC algorithm’s utility in haplo-SCT with PTCy.