Improved Outcome for AML Relapse after Allogeneic Transplant with High Intensity Chemotherapy Followed By 2nd Allogeneic Stem Cell Transplant or Donor Lymphocyte Infusion; A Retrospective Analysis

Abstract

<h3>Background</h3> Acute myeloid leukemia (AML) relapse after allogeneic stem cell transplant (allo SCT) remains a major therapeutic challenge. While patients with longer remission after initial allo SCT are recommended to receive either 2^nd allo SCT or donor lymphocyte infusion (2^nd allo SCT/DLI), survival probability for patients relapsing within 6 months of SCT are dismal. We evaluated the outcomes of AML relapse after allo SCT to assess the impact of 2^nd allo SCT/DLI on long term survival. <h3>Methods and Results</h3> One hundred and seventy-two patients with AML underwent allo SCT at Penn State Cancer Institute from Jan 2014 to Aug 2019. Sixty-nine patients relapsed after allo SCT (median age, 60 years; range, 10-75). Of these, 4 patients underwent 2nd allo SCT and 26 received DLI. Overall survival (OS) was defined as days from the relapse and evaluated with Kaplan-Meier survival curves and log rank tests. One-year OS in all cases was 22.4% (95% CI: 13.3-32.9%). Patients with ECOG performance status (PS) 0-2 at relapse (n=60) showed a better 1-year OS compared to those with ECOG PS 3-4 (n=9) (24 vs 11.1%, p<0.001). Blast counts at relapse (≥20 vs <20% in the bone marrow or peripheral blood) did not affect patient outcome (1-year OS: 20 vs 28.6%, p=0.47). Median OS for patients who received no treatment (n=8), chemotherapy only (n=31), or 2^nd allo SCT/DLI (n=30) were 20, 68 and 186 days respectively (p<0.01). Relapsed patients receiving conventional re-induction chemotherapy were categorized as the high intensity (H) group, while those receiving treatments such as hypomethylating agents, protein kinase inhibitors, or venetoclax were categorized as the low intensity (L) group (Table 1). H group showed a better 1-year OS compared to L group (37 vs 11.9%, p<0.01). High intensity chemotherapy + 2^nd allo SCT/DLI (H+D) showed a better 1-year OS compared to the other 3 groups (H-D, L+D, and L-D; 52.9 in H+D vs 11.6% in all others, p<0.001, Fig 1). Among patients aged ≥ 50 years, the same tendency was seen (1-year OS: 50 vs 10.8%, p<0.01). Even for patients who relapsed < 6 months post allo SCT, there was the same trend (1-year OS: 45.5 vs 7.5%, P<0.01, Table 2). Factors including age and donor source at the 1^st allo SCT, time to relapse, blast counts and PS at relapse, and treatment types after relapse were used for multivariate analysis with Cox proportional hazard regression and only treatment types after allo SCT were identified as independent risk factors for improved OS (L±D: Hazard ratio 2.7, 95% CI: 1.2-6.5, p<0.05; H-D: Hazard ratio 3.4, 95% CI: 1.3-8.5, p<0.05; L-D: Hazard ratio 4.1, 95% CI: 1.9-8.9, p<0.001, compared to H+D). The number of patients treated with venetoclax was too small to allow meaningful analysis. <h3>Conclusions</h3> Relapsed AML after allo SCT treated with high intensity chemotherapy followed by 2^nd allo SCT/DLI showed a better OS. Similar benefit was shown even for patients who relapsed within 6 months of allo SCT.