Amphiregulin Appears to be a Promising Predictive Marker for Mortality in Acute Graft-Versus-Host Disease

Amphiregulin (AREG) appears to be a promising longitudinal monitoring biomarker for early mortality among patients with acute graft-vs-host disease (GVHD), according to a sample of 2 prospective clinical trials that was presented at the 2021 American Society of Hematology Annual Meeting; investigators noted that those with elevated AREG should be carefully monitored throughout serial assessments.

Specifically, in a multivariate analysis for survival, day 28 responses (risk ratio [RR], 9.14; P <.0001) and baseline AREG (>212 pg/mL; RR, 2.72; P <.05) were the only factors found to be independently associated with survival.

When this was also evaluated in a multivariate analysis out of a University of Minnesota study, day 28 responses (risk ratio [RR], 4.94; P <.0001) and baseline AREG (>212 pg/mL; RR, 4.17; P = .03) were also the only factors found to be independently associated with survival.

“Patients with life-threatening acute GVHD often have severe symptoms related to organ and tissue damage, elevated blood biomarkers of acute GVHD, or both. Monitoring can be challenging due to unpredictable variations in symptoms that may not be related to acute GVHD severity or pathogenesis, including changes in nutrition and infections,” lead study author Shernan Holtan, MD, of University of Minnesota, said in a presentation of the findings.

Researchers evaluated AREG, ST2, and Reg3a as monitoring biomarkers in a University of Minnesota study, which treated patients with uhCG/EGF, they then compared those findings with an independent cohort of patients with steroid-refractory acute GVHD receiving second-line ruxolitinib in the REACH-1 trial (NCT02953678).

Samples were collected at baseline, day 7, 14, 28, and 56. The University of Minnesota study (hCG/EGF) enrolled a total of 51 patients. Amphiregulin was measured using an enzyme-linked immunosorbent assay, and ST2 and REG3a were measured via a multiplex Luminex-based assay. In this study, there was significant positive correlations between assay platforms. The median age for first-line, high-risk patients (n = 26) in this study was 61 years (range, 22-72), there were 19 males (73%), and the number of patients with GVHD grade II, IIII, and IV was 2 (8%), 21 (81%), and 3 (12%), respectively.

Among the second-line cohort (n = 26), the mean age was 62 years (range, 2-69), there were 20 males (77%), and the number of patients with GVHD grade II, IIII, and IV was 9 (35%), 8 (31%), and 9 (35%). Severity was mostly grade 3 or 4 in both these cohorts, according to Holton.

REACH-1 enrolled 60 patients and all 3 biomarkers were assessed using microfluidic immunoassay. Among REACH-1 participants, the mean age was 52 years (range, 18-73), 31 were male (52%), and the number of patients with acute GVHD grade II, IIII, and IV was 22 (36.7%), 24 (40.0%), and 14 (23.3%), respectively.

In REACH-1, biomarker levels were assessed at baseline, and it was revealed that patients with day 28 progressive disease (PD) had higher baseline levels of AREG compared with patients with a day 28 complete response (CR; P <.01). Similarly, ST2 levels were also higher for patients with day 28 PD vs CR (P <.01) or very good partial response (VGPR)/partial response (PR; P <.05). Baseline REG3a did not adhere to the pattern; levels of this biomarker were not significantly different between patients with day 28 CR, VPGR/PR, or PD.

In the University of Minnesota study, patients with a 28-day CR experienced a 3-fold decrease in AREG levels between baseline and day 56 (mean, 98 vs 32 pg/mL, P = .006). However, among patients with a day 28 PR or no response (NR), AREG levels did not exhibit clinically meaningful change.

When this was evaluated in REACH-1, AREG levels decreased 2.8-fold from baseline to day 56 in patients with a day 28 CR (mean, 174.7 vs 63.6 pg/mL; P = .007). AREG levels also decreased 2.0-fold over time in patients with a day 28 PR (mean baseline 288 vs. 146.1 pg/mL at day 56; P = .017). AREG levels did not significantly change in those who had PD.

In the University of Minnesota study, patients with CR experienced a 1.4-fold decrease in ST2 over the course of 28 days (P = .02), compared with patients with PR or NR. In contrast, in the same study period, REG3a did not exhibit any significant change in any of the response categories. This was similar in REACH-1 (ST2, P = .01).

In regard to biomarker cutoffs, AREG was linked to rapidly fatal courses. Generally, the levels of AREG were higher in patients enrolled in REACH-1 compared with those in the University of Minnesota study. This may reflect a difference in assay, severity of illness of both, according to the study authors.

In the University of Minnesota study, biomarkers cutoffs were linked to a rapidly fatal course. With an AREFG cutoff of 212 pg/mL, the median survival was not reached vs 62 days (P = .006). With an ST2 cutoff of 292 ng/mL, the median survival was not reached vs 239 days (P <.001), and with a REG3a cutoff of 13.5 ng/mL, the median survival was not reached vs 416 days (P = .01).

In REACH-1, the AREG cutoff was found to be 336 pg/ml, and the median survival was not reached vs 74 days (P = .005). The ST2 cutoff was 188 ng/ml (P = .09), and the REG3a cutoff was 3.6 ng/ml (P = .03).

“Based upon our current samples, there is not a close correlation between these biomarkers,” Holton said in a post-presentation discussion. “That's a good thing. This is the new and additional way to do monitoring.”

Holton concluded that amphiregulin is a useful longitudinal monitoring biomarker with patients who have life threatening acute graft-vs-host disease.

“We have also shown that patients with a high baseline amphiregulin are at high risk of early death,” she said. “They should be monitored on a consistent platform during serial assessments during their therapy.”

Reference

Pratta M, El Jurdi N, Rashidi A, et al. Validation of amphiregulin as a monitoring during treatment of life-threatening acute graft-versus-host disease: a secondary analysis of 2 prospective clinical trials. Presented at: 2021 ASH Annual Meeting and Exposition; December 11-14, 2021; Atlanta, GA. Abstract 259.