Validation of Amphiregulin As a Monitoring Biomarker during Treatment of Life-Threatening Acute Gvhd: A Secondary Analysis of 2 Prospective Clinical Trials

Background: Patients with life-threatening acute graft-versus-host disease (aGVHD) often have severe symptoms related to organ/tissue damage, elevated blood biomarkers of aGVHD, or both. Monitoring such patients can be challenging due to unpredictable variations in symptoms that may not be related to aGVHD pathogenesis, such as dietary changes, medication side effects, or infections. In our previously published phase 1 study of urinary-derived human chorionic gonadotropin/epidermal growth factor (uhCG/EGF) for the treatment of life-threatening aGVHD (NCT02525029), we recognized amphiregulin (AREG) as a potential GVHD monitoring biomarker that is elevated at the start of treatment and decreases in patients who respond to treatment. Here we show the results of AREG as a monitoring biomarker in our larger uhCG/EGF treated cohort and validate them in an independent cohort of patients with steroid-refractory aGVHD receiving ruxolitinib as second-line therapy in the REACH1 study.

Methods: Longitudinal plasma samples were cryopreserved at study baseline and on study Days 7, 14, 28, and 56 from patients enrolled in both NCT02525029 (N=51) and REACH1 (N=60). In samples collected during NCT02525029, AREG was measured using enzyme-linked immunosorbent assay (R&D Systems, Minneapolis, MN), and ST2 and REG3a were measured using a multiplex Luminex-based array (R&D Systems). Samples from REACH1 were analyzed for concentrations of AREG, ST2, and REG3a using the ProteinSimple Ella platform (Bio-Techne, San Jose, CA). Comparisons of biomarker concentrations between response groups were performed using nonparametric one-way analysis of variance (Kruskal-Wallis test). Analyses of change of biomarkers from baseline to subsequent study days were performed using nonparametric matched pairs analysis (Wilcoxon signed rank test). Biomarker cutoffs relevant for overall survival at study baseline were identified using recursive partitioning on the NCT02525029 dataset and tested in the REACH1 dataset.

Results: In patients with a complete response (CR) at Day 28 to uhCG/EGF, AREG decreased 3-fold from baseline to Day 56 (mean, 98 vs 32 pg/mL, P=0.006, Figure 1A). AREG did not significantly change over time in patients with a partial response (PR) or no response (NR) to uhCG/EGF. Similarly, in patients with a Day 28 CR to ruxolitinib, AREG decreased 2.8-fold from baseline to Day 56 (mean, 174.7 vs 63.6 pg/mL, P=0.007, Figure 1B). AREG also decreased 2.0-fold over time in patients treated with ruxolitinib who had a PR to treatment (mean baseline 288.2 vs 146.1 pg/mL at Day 56, P=0.017) but no change in those with progressive disease (PD). ST2 declined with CR, but to a lesser extent, over time in patients with a CR in both studies (1.4-fold decrease in NCT02525029, P=0.04 and 2.2-fold decrease in REACH1, P=0.021). REG3a did not significantly change over time in responders in either clinical trial. Biomarker cutoffs associated with a rapidly fatal course were identified in NCT02525029 for AREG (>212 pg/mL associated with a median survival of 62 days, P=0.006, Figure 1C), ST2 (>292 ng/mL associated with a median survival of 239 days, P<0.001), and REG3a (>13.5 ng/mL associated with a median survival of 416 days, P=0.01). These cutoffs were not statistically significant in REACH1, although patients with a baseline AREG >212 pg/mL had a median survival of 165 days vs not reached (P=0.1, Figure 1D) in REACH1.

Conclusions: Using samples collected during 2 prospective clinical trials on 2 different measurement platforms, we conclude that AREG is a useful longitudinal monitoring biomarker for patients with life-threatening aGVHD. AREG levels were generally higher in REACH1 than NCT02525029, which could reflect differences in assay, severity of illness, or both. Patients with a high baseline AREG are at high risk of early death and should be monitored on a consistent platform during serial assessments.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal