Eltrombopag with or without Tacrolimus for relapsed/refractory acquired aplastic anaemia: a prospective randomized trial – Blood Cancer Journal

Patient baseline characteristics

A diagram of screening and enrolment details was depicted in Supplementary Fig. 1. Ultimately, 114 patients were included in the analysis, with 96 (84.2%) classified as NSAA patients and 18 (15.8%) classified as SAA patients. Among them, 59 (51.8%) patients were refractory to previous IST treatment, and the others were relapsed patients. Seventy-six patients received EPAG+tacrolimus and 38 received EPAG only. No significant difference was found in the baseline characteristics of patients between the EPAG+tacrolimus and EPAG monotherapy groups, including the SAA/NSAA ratio, previous ATG/ATG+CsA treatment ratio, time to previous IST treatment, patient baseline data before EPAG like laboratory examinations, PNH clone presence proportion, and cytogenetic characteristics (Table 1).

Clinical efficacy

The ORR for patients treated with EPAG+tacrolimus and EPAG alone was 38.2% vs. 31.6% (P = 0.490) at the 3^rd month, 61.8% vs. 39.5% (P = 0.024) at the 6^th month, 64.5% vs. 47.1% (P = 0.097) at the 12^th month, and 60.5% vs. 34.2% (P = 0.008) at the last follow-up, with a median follow-up of 18 (6–24) months, comparable length of EPAG treatment (12 (6–24) months vs. 10 (6–24) months, P = 0.434), EPAG dose at the 3^rd month of follow-up (100 (25–150) mg/d vs. 100 (50–150) mg/d, P = 0.552), EPAG dose at the 6^th month of follow-up (75 (25–150) mg/d vs.100 (50–150) mg/d, P = 0.161), and the aggregated doses of EPAG exposure (26.5 (11.9–56.9)g vs. 27.0 (13.5–56.9)g, P = 0.559) for the EPAG+tacrolimus and EPAG monotherapy groups. No significant difference was found in the time to response (3 (1–12) months vs. 3 (1–8) months, P = 0.763) or the CRR at the 3^rd month/6^th month/12^th month/last follow-up between two groups (7.9%/13.2%/19.4%/15.8% vs. 5.3%/7.9%/11.8%/10.5%, P > 0.05 at each evaluated time point, Fig. 1).

Safety

AEs of various degrees attributed to either EPAG or tacrolimus were reported in 23 (30.3%) patients with EPAG+tacrolimus and 9 (23.4%) patients with EPAG only. The most common AEs in the EPAG+tacrolimus group included dyspepsia (11.8%), increased creatinine (3.9%), and pruritus (3.9%); the most common AEs in the EPAG monotherapy group included dyspepsia (10.6%), elevated ALT (5.3%) and pruritus (2.6%). All AEs were CTCAE grade 1-2 and were relieved after symptomatic treatment or EPAG dose adjustment. No patient discontinued EPAG or died due to AEs. There was no significant difference in the rate of different AEs between the EPAG+tacrolimus group and the EPAG monotherapy group (Table 2).

Survival and clonal evolution

Three patients (3.9%) died in the EPAG+tacrolimus group during follow-up, and the causes of death included pulmonary infection related to neutropenia for two patients and unknown reasons for one patient. None of the three patients responded, and they died at a median of 5 (1–6) months after discontinuing treatment. On the other hand, one NR patient (2.6%) in the EPAG monotherapy group died from unknown reasons at 3 months after discontinuing EPAG. No significant difference was found in the cumulative overall survival curve between the groups (P = 0.635, Fig. 2A).

A The OS curve was not significantly different between the groups (P = 0.635). B The RFS curve showed a trend of benefit in the EPAG+tacrolimus group. EPAG eltrombopag.

Malignant clonal evolution occurred in 3 (3.9%) patients in the EPAG+tacrolimus group and one (2.6%) patient in the EPAG monotherapy group (P = 1.000, Supplementary Table 1). For the EPAG+tacrolimus group, one NSAA patient had monosomy 7 at the 12^th month, one SAA patient had a low-risk evolution event (trisomy 8) at the 13^th month but did not progress to MDS, and one SAA patient progressed to MDS (MDS with low blasts, according to WHO 2022 classification) at the 13^th month of follow-up. One SAA patient in the EPAG monotherapy group had t [10, 11] (p12; q21) at the 24^th month of follow-up but did not progress to MDS. An increase in PNH clone size occurred in 3 (3.9%) other patients in the EPAG+tacrolimus group at the 5th or 6th month of follow-up: one patient’s PNH clone size increased from 5% to 33%, one from 0% to 4%, and one from 0% to 3%. For the EPAG monotherapy group, one patient’s PNH clone size increased from 0% to 3% at the 9^th month of follow-up. All 8 patients mentioned above responded before clonal evolution events occurred.

Predictors of overall response and relapse

Possible predictors of the ORR and relapse rate (RR) were analysed. For the EPAG+tacrolimus group, responders had a significantly higher baseline reticulocyte count (61.5 (11.5–230.5) × 10⁹/L vs. 21.6 (6.6–56.6) × 10⁹/L, P = 0.039) than nonresponders, but no other significant predictors were detected. Multivariate analysis showed that no factor was independently related to the ORR in the EPAG+tacrolimus group. For the EPAG monotherapy group, no significant predictor was found for the ORR (Table 3). No significant predictor was found for relapse in either the EPAG+tacrolimus or the EPAG monotherapy group (Supplementary Table 2).

Post hoc analysis of RFS and age subgroups

A post hoc analysis of RFS was conducted, and no difference was detected between the RFS curves of the two groups (P = 0.101, Fig. 2B). We also conducted a post hoc analysis to compare the ORR, CRR, RFS and clonal evolution in different age subgroups (<60 years old or ≥60 years old). A total of 78.9% (60/76) of patients receiving EPAG+tacrolimus and 65.8% (25/38) of patients receiving EPAG monotherapy were <60 years old (P = 0.128). For patients under 60 years old, the baseline characteristics before EPAG were comparable (Supplementary Table 3). The ORR at the 6th month/last follow-up was significantly higher in patients with EPAG+tacrolimus than in those with EPAG monotherapy (61.7% vs. 36.0%, P = 0.030 at the 6^th month and 61.7% vs. 32.0%, P = 0.013 at the last follow-up, respectively), and for the ORR at the 3^rd month and 12^th month, the difference between the groups was not significant (Fig. 3A). The CRRs were 8.3%/11.7%/18.4%/13.3% at the 3rd month/6th month/12th month/last follow-up in the EPAG+tacrolimus group and 4.0%/4.0%/8.7%/8.0% in the EPAG monotherapy group (P > 0.05 for the CRR at each evaluated time point between the groups, Fig. 3A). In this subgroup, a clear benefit in RFS was observed in the EPAG+tacrolimus group (P = 0.048, Fig. 4A). There was no significant difference in the death rate (1.7% vs. 0.0%, P = 1.000) or malignant clonal evolution rate (3.3% vs. 4.0%, P = 1.000) between the therapy groups in this age subgroup.

A The response rate of patients <60 years old revealed a significantly higher ORR at the 6^th month/last follow-up in patients treated with EPAG+tacrolimus than in those treated with EPAG alone. B The response rate of patients ≥60 years old was not significantly different between patients treated with either EPAG+tacrolimus or EPAG monotherapy at any evaluated time point. CR complete response, EPAG eltrombopag, PR partial response, ORR overall response. *P < 0.05.

A Cumulative RFS curves of patients <60 years old illustrated a significant benefit in the EPAG+tacrolimus group (P = 0.048). B Cumulative RFS curves of patients ≥60 years old illustrated no significant difference between the treatment groups. EPAG eltrombopag.

For patients ≥60 years old, the baseline characteristics were comparable except for the proportion of males (18.8% vs. 61.5%, P = 0.027, Supplementary Table 4). No significant difference was found in the ORR (25.0%/62.5%/61.5%/56.3% vs. 30.8%/46.2%/54.5%/38.5%, P > 0.05, Fig. 3B) or CRR (6.3%/18.8%/23.1%/25.0% vs. 7.7%/15.4%/18.2%/15.4%, P > 0.05, Fig. 3B) at the 3^rd/6^th/12^th/last follow-up between elderly patients with EPAG+tacrolimus and EPAG. No difference in cumulative RFS curves (P = 1.000, Fig. 4B), the death rate (12.5% vs. 7.7%, P = 1.000) or malignant clonal evolution rate (6.3% vs. 0.0%, P = 1.000) was found between the groups.

Post hoc analysis of the relapsed/refractory subgroups

We further conducted a post hoc subgroup analysis on relapsed and refractory patients. For the relapsed subgroup, there were 55 patients; 36 were in the EPAG+tacrolimus group, and 19 were in the EPAG monotherapy group. No significant differences were detected in the baseline characteristics between patients in the EPAG+tacrolimus and EPAG monotherapy groups except for in the baseline haemoglobin level (88 (44–159) g/L for the EPAG+tacrolimus group and 68 (30-114) g/L for the EPAG monotherapy group, P = 0.032, Supplementary Table 5). The ORR was significantly higher in the 6th month of follow-up and the last follow-up (58.3% vs. 26.3% in the 6^th month of follow-up, P = 0.024; 58.3% vs. 26.3% at the last follow-up, P = 0.024), while no significant difference was found in the ORR at the 3rd/12th month of follow-up (Supplementary Fig. 2A). The CRR was 11.1%/13.9%/10.7%/13.9% and 5.3%/5.3%/5.6%/5.3% (P > 0.05) for the EPAG+tacrolimus group and EPAG monotherapy group at the 3^rd month/6^th month/12^th month/last follow-up, respectively. No significant difference in the cumulative RFS curves was found between relapsed patients in the EPAG+tacrolimus and EPAG monotherapy groups (P = 0.422, Supplementary Fig. 3A). The death rates were 2.8% and 0% (P = 1.000) for relapsed patients in the EPAG+tacrolimus and EPAG monotherapy groups, respectively.

There were 59 refractory patients, with 40 patients in the EPAG+tacrolimus group and 19 in the EPAG monotherapy group. No significant difference was found in baseline characteristics (Supplementary Table 6) for refractory patients in the EPAG+tacrolimus and EPAG monotherapy groups. No significant difference was found in the ORR and CRR at the 3rd month/6th month/12th month/last follow-up (Supplementary Fig. 2B). No significant difference in the cumulative RFS curves was found between refractory patients in the EPAG+tacrolimus and EPAG monotherapy groups, yet a trend of superiority in the cumulative RFS of the EPAG+tacrolimus group was observed (P = 0.106, Supplementary Fig. 3B). The death rates were 5.0% and 5.3% (P = 1.000) for relapsed patients in the EPAG+tacrolimus and EPAG monotherapy groups, respectively.