Prognostic prediction of new risk scores (AML-DRG and AML-HCT-CR) in acute myeloid leukemia patients with allogeneic hematopoietic stem cell transplantation

Patient characteristics

Among the 172 AML patients analyzed, 48.3% (n = 83) were women. The median age was 31.5 years (range 14 to 62 years), two patients were over 60 years old (1.2%). The median follow-up was 44 months (range 1 to 94 months). According to the AML-DRG model, 109, 49, and 14 patients were in the low, intermediate, and high risk group, respectively. According to the AML-HCT-CR model, 108, 30, 20 and 14 patients were in the low, intermediate, high and very high risk group, respectively. The basic clinical data of the patients are shown in Table 1.

Table 1 Characteristics of the patients.

GVHD

Among the 172 patients, grade II to IV acute GVHD developed in 39 patients (22.7%) and chronic GVHD in 38 patients (22.1%). For the entire cohort, the cumulative incidence of grade II to IV acute GVHD at day 100 was 22.7% (95% CI: 20.0–25.4), the cumulative incidence of chronic GVHD of all degrees at 2 years was 21.7% (95% CI: 19.0 –24.4). The 100-day cumulative incidence of grade II to IV acute GVHD for the low-, intermediate-, and high-risk AML-DRG groups was 22.0% (95% CI: 18.3 to 25.7), 18, 4% (95% CI 11.8 to 25.0) and 7.1% (95% CI 0 to 21.1), the differences are not statistically significant ( p= 0.605). The incidence of chronic GVHD of all grades at 2 years was 23.2% (95% CI: 19.4–27.0), 22.4% (95% CI: 15.5–29 .3) and 7.1% (95% CI: 0–22.1), respectively (p= 0.430) (Table 2).

Table 2 AHCT results according to the AML-DRG and AML-HCT-CR models.

For the AML-HCT-CR model, the 100-day cumulative incidence of grade II to IV acute GVHD for the low, intermediate, high, and very high risk groups was 21.3% (95% CI: 17, 5–25.1). 13.3% (95% CI 4.3–22.3), 30.0% (95% CI 15.8–44.2), and 7.1% (95% CI 0–21.1), respectively ( p= 0.582). The incidence of chronic GVHD of all grades at 2 years was 23.4% (95% CI 19.6–27.2), 30.3% (95% CI 19.8–40.8), 10 .0% (95% CI 0–21.7) and 7.1% (95% CI 19.6–27.2). CI 0–22.7), respectively ( p= 0.287) (Table 2).

Relapse and NRM

For the entire cohort, the cumulative incidences of relapse and NRM at 3 years were 21.1% (95% CI: 18.5–23.7) and 24.8% (95% CI: 22, 1–27.5), respectively. We found that relapse and NRM occurred in 37 (21.5%) and 44 (25.6%) of 172 patients. The reasons for relapse were hematological in 30 patients (17.4%), extramedullary in 4 patients (2.3%), and hematological plus extramedullary in 5 patients (2.9%).

The 3-year cumulative incidence of relapse for the low-, intermediate-, and high-risk AML-DRG groups was 12.2% (95% CI: 9.5 to 16.3), 32.7% (95% CI: 95%: 25.4 to 40.0) and 50.0% (95% CI: 95 to 16.3). % CI 30.2–69.8), respectively (p<0.001) (Fig. 1B), with the corresponding 3-year MRL was 18.8% (95% CI 15.1–22.5), 32.9% (95% CI 25.6–40.2) and 42.9% (95% CI 22.9–62.9), the differences are not statistically ( p= 0.072) (Figure 1C). The cumulative incidence of relapse at 3 years for the low, intermediate, high, and very high risk AML-HCT-CR groups was 12.3% (95% CI: 8.9–15.7), 26 .7% (95% CI: 16.4–37.0), 45.0% (95% CI 30.1–59.9), and 42.9% (95% CI 23.4–62.4 ), respectively (p<0.001) (Fig. 2B), with the 3-year NRM was 19.0% (95% CI 15.3–22.7), 20.4% (95% CI 10.5–30.3) , 45.0% (95% CI 29.9–60.1) and 50.0% (95% CI 30.0–70.0), respectively (p= 0.008) (Figure 2C).

Figure 1
Figure 1

Comparison of overall survival (A), cumulative incidence of relapse (B.) and the cumulative incidence of NRM (C) by the AML-DRG model in patients with low, intermediate and high risk groups.

Figure 2
Figure 2

Comparison of overall survival (A), cumulative incidence of relapse (B.) and the cumulative incidence of NRM (C) by the AML-HCT-CR model in patients with low, intermediate and high risk groups.

OS and PFS

The 3-year OS and PFS for the entire cohort were 56.9% (95% CI 49.9–64.9) and 53.1% (95% CI 45.9–61.3). Patients in the low-, intermediate-, and high-risk AML-DRG groups had a median OS of 33.0 (1 ~ 94), 16.0 (1 ~ 73), and 4.5 (1 ~ 69) months, respectively ( p < 0.001), with 3-year OS of 69.6% (95% CI 61.0–79.6), 38.6% (95% CI 27.1–55.0), and 7.1 % (95% CI 1.1–47.2), respectively (p < 0.001). PFS at 3 years was 67.3% (95% CI 58.8–77.2), 34.5% (95% CI 23.4–50.8), and 7.1% (95% CI 1 .1–47.2), respectively (p < 0.001) (Fig. 1A). The median OS for the low, intermediate, high, and very high risk AML-HCT-CR groups was 33.0 (1 ~ 94), 28.9 (2 ~ 73), 6.5 (1 ~ 43) and 4.0 (1 ~ 69) months, respectively (p < 0.001), with corresponding 3-year OS of 69.4% (95% CI 60.7–79.4), 59.6% (95% CI % 44.3–80.2), 10.0% (95% CI 2.7–37.2) and 7.1% (95% CI 1.1–47.2), respectively (p < 0.001) (Fig. 2A). The 3-year PFS was 67.1% (95% CI 58.5–77.0), 52.9% (95% CI 37.6–74.4), 10.0% (95% CI 37.6–74.4), 95% 2.7–37.2) and 7.1% (95% CI 1.1–47.2), respectively (p < 0.001) (Table 2).

In univariate analysis for OS, patients in the intermediate- and high-risk AML-DRG groups had a significantly increased risk of death with a hazard ratio (HR) of 2.62 (95% CI: 1.60– 4.31; p<0.001) and 7.08 (95% CI 3.68-13.63; p< 0.001), respectively compared to the low-risk group. In addition, the risk of death was higher in the high-risk group compared to the intermediate-risk group (HR 2.63, 95% CI 1.35–5.10; p= 0.004), confirming the ability of the AML-DRG model to predict post-transplant survival. For the AML-HCT-CR groups, patients with high (HR 5.44, 95% CI 3.03–9.78; p<0.001) and very high risk groups (HR 8.35, 95% CI 4.32-16.14; p< 0.001) had a significantly higher risk of death than the low-risk group, while there was no difference between low and intermediate risk (HR 1.39, 95% CI 0.72–2.69; p= 0.330) and between the high and very high groups (HR 1.53, 95% CI 0.74-3.17; p= 0.251). Similar results were found in a univariate analysis for PFS as summarized in Table 3.

Table 3 Univariate analysis evaluating the impact of the HCT-CR model on OS and PFS.

Multivariate analysis

Multivariate analysis confirmed that the HCT-CR and AML-HCT-CR models could be used to predict OS for patients in different risk groups after adjusting for other variables, such as age, gender, type of conditioning, modality of transplant and source of stem cells. See Table 4 for more information.

Table 4 Multivariate analysis for OS.

Comparison of prognostic stratification

The AML-DRG, AML-HCT-CR, DRI, ELN2017 and HCT-CI/Age genetic risk model C-scores were 0.69 (95% CI 0.61 to 0.78), 0.71 ( 95% CI: 0.63 to 0.79), 0.61 (95% CI 0.52–0.69), 0.52 (95% CI 0.43–0.60), and 0.59 ( 95% CI 0.50–0.67), respectively. Compared with the DRI and HCT-CI/Age models, the AML-DRG and AML-HCT-CR models had significantly better discriminative ability in predicting OS with the C index. AML-DRG and AML-HCT-CR may be better than that of the ELN2017, DRI and HCT-CI/Age genetic risk models.

Leave a Comment