Optimal Depth of the Treatment Response Before Allogeneic Hematopoietic Transplantation for Chronic Myeloid Leukemia in Chronic Phase

Abstract

The outcomes of chronic myeloid leukemia (CML) have dramatically improved since the introduction of tyrosine kinase inhibitors (TKIs) [1, 2]. The development of TKIs has improved the life expectancy in CML patients, which is approaching that in the general population. However, resistance and intolerance to TKIs can still occur, and allogeneic hematopoietic cell transplantation (allo-HCT) is still a therapeutic option for CML. According to the European LeukemiaNet 2020 recommendations, CML patients in the first chronic phase (CP1) who are resistant or intolerant to multiple TKIs, those who progress to the accelerated phase (AP) or blast phase (BP) during treatment, and those who initially present with BP are recommended to receive allo-HCT [3]. Regarding disease status, it is well known that CML patients in chronic phase (CML-CP) have favorable outcomes compared to those in the AP or BP [4]. However, it is not yet clear whether a deeper treatment response in CML-CP results in better outcomes after allo-HCT.

In this study, we aimed to clarify the optimal depth of the treatment response at allo-HCT for CML-CP. Based on the Japanese nationwide registry database, we retrospectively analyzed transplant outcomes in CML-CP who received their first allo-HCT and who had received TKI therapy before allo-HCT. Details of additional methods are provided in the Supporting Information. Between 2002 and 2022, 1413 patients who were aged 16 years or older received their first allo-HCT, and 689 patients met the inclusion criteria. Disease status at diagnosis was CP or AP in 350 and BP in 339 patients (de novo BP group). Among the patients whose disease status at diagnosis was CP or AP, disease status at allo-HCT was CP1 in 207 (insufficient response group) and second CP (CP2) or later in 143 (disease progression group; Figure S1).

Regarding the depth of the treatment response, the number of patients with major molecular response (MMR) at allo-HCT has increased in recent years, whereas the number of patients with complete cytogenetic response (CCyR) or complete hematologic response (CHR) has decreased (p < 0.001, Table 1). Accordingly, second- and third-generation TKIs before allo-HCT (p < 0.001) and post allo-HCT TKI maintenance (p < 0.001) were more commonly administered in patients with MMR compared to the rates in those with CCyR or CHR. The relationship between the highest generation of TKIs before allo-HCT and transplantation years was shown in Figure S2. The details of TKIs before and after allo-HCT were provided in Table S1.

In the entire cohort, disease-free survival (DFS) at 3 years was higher in patients with MMR (76.3%, 95% confidence interval (CI): 70.6–80.9) than in those with CCyR (65.1%, 95% CI: 56.6–72.3) or CHR (64.8%, 95% CI: 58.3–70.5, p = 0.017). Overall survival (OS) at 3 years was also higher in patients with MMR (82.6%, 95% CI: 77.4–86.7) than in those with CCyR (70.3%, 95% CI: 61.9–77.1) or CHR (69.3%, 95% CI: 62.9–74.8, p = 0.007). On the other hand, the cumulative incidence of relapse at 3 years was not significantly different between patients with MMR (9.9%, 95% CI: 6.7–13.8), CCyR (13.7%, 95% CI: 8.7–19.9), and CHR (14.5%, 95% CI: 10.4–19.2, p = 0.15). Multivariable analyzes confirmed that MMR before allo-HCT was significantly associated with superior DFS (HR 0.64, 95% CI: 0.46–0.88, p = 0.0058), OS (HR 0.59, 95% CI: 0.41–0.83, p = 0.003), and a lower risk of relapse (HR 0.57, 95% CI: 0.34–0.97, p = 0.039). The cumulative incidence of non-relapse mortality (NRM) was comparable among MMR, CCyR, and CHR in both univariate and multivariable analyzes (Figure S3).

Next, we performed subgroup analyzes according to the change in disease status from the initial diagnosis to allo-HCT. Clinical characteristics of the subgroups are provided in Tables S2 and S3. Regarding OS, we found a potential interaction between the depth of the treatment response and the subgroups (p for interaction = 0.11; Figure S4).

In the insufficient response group, DFS at 3 years was not significantly different between MMR (71.6%, 95% CI: 52.5–84.1), CCyR (83.5%, 95% CI: 68.4–91.8), and CHR (72.5%, 95% CI: 63.8–79.4, p = 0.32). OS at 3 years was also comparable between MMR (83.8%, 95% CI: 65.4–92.9), CCyR (83.5%, 95% CI: 68.4–91.8), and CHR (75.2%, 95% CI: 66.6–81.9, p = 0.75). The cumulative incidence of relapse at 3 years was higher in MMR (18.9%, 95% CI: 7.5–34.2) than in CCyR (2.4%, 95% CI: 0.2–11.0) and CHR (7.0%, 95% CI: 3.4–12.3, p = 0.028), whereas this difference was not significant in the multivariable analysis (HR 2.18, 95% CI: 0.79–5.99, p = 0.13; Figure 1A–E).

In the disease progression group, DFS at 3 years seemed to be higher in MMR (66.8%, 95% CI: 51.9–77.9) than in CCyR (57.2%, 95% CI: 40.4–70.9) and CHR (49.2%, 95% CI: 33.8–63.0, p = 0.055). OS at 3 years also seemed to be higher in MMR (76.7%, 95% CI: 62.6–86.1) than in CCyR (62.0%, 95% CI: 44.9–75.1) and CHR (59.4%, 95% CI: 43.3–72.3, p = 0.17). The cumulative incidence of relapse at 3 years was lower in MMR (11.9%, 95% CI: 4.8–22.7) than in CCyR (17.3%, 95% CI: 7.5–30.5) and CHR (27.6%, 95% CI: 15.7–41.0, p = 0.048). Multivariable analyzes confirmed that MMR was significantly associated with superior DFS (HR 0.43, 95% CI: 0.22–0.87, p = 0.018) and a lower risk of relapse (HR 0.26, 95% CI: 0.07–0.99, p = 0.048), whereas OS was not significantly different between the depth of the treatment response at allo-HCT (HR 0.50, 95% CI: 0.24–1.07, p = 0.075; Figure 1F–J).

In the de novo BP group, DFS at 3 years was higher in MMR (79.7%, 95% CI: 73.1–84.8) than in CCyR (57.6%, 95% CI: 44.2–68.9) and CHR (60.6%, 95% CI: 47.7–71.2, p < 0.001). OS at 3 years was also higher in MMR (83.9%, 95% CI: 77.6–88.6) than in CCyR (67.0%, 95% CI: 53.5–77.3) and CHR (65.0%, 95% CI: 52.1–75.2, p < 0.001). Moreover, the cumulative incidence of relapse at 3 years was lower in MMR (7.8%, 95% CI: 4.5–12.1) than in CCyR (19.5%, 95% CI: 10.7–30.3) and CHR (19.7%, 95% CI: 11.1–30.2, p = 0.004). Multivariable analyzes also confirmed that MMR was significantly associated with superior DFS (HR 0.44, 95% CI: 0.26–0.72, p = 0.0011) and OS (HR 0.38, 95% CI: 0.22–0.65, p < 0.001), and the risk of relapse (HR 0.30, 95% CI: 0.12–0.72, p = 0.007; Figure 1K–O). The cumulative incidence of NRM was comparable among MMR, CCyR, and CHR across all subgroups.

As a result, in the insufficient response group, patients might be recommended to proceed to allo-HCT without delay rather than aiming for a deeper response. However, it should be noted that most of them only used imatinib before allo-HCT. Moreover, patients who achieve CCyR or a better response in CP1 no longer receive allo-HCT under the current guidelines [3, 5]. Further investigation is warranted to determine whether the strategy of immediately proceeding to allo-HCT should still be recommended, given the availability of multiple TKIs. On the other hand, in the disease progression group, a deeper treatment response was associated with better transplant outcomes, and the same trend was observed more clearly in the de novo BP group. Because patients in the de novo BP group were assumed to have developed in CP but were not diagnosed until the disease progressed to BP, patients in these groups seemed to be similar from a pathological perspective. Thus, both groups might exhibit a similar trend and might be recommended to achieve MMR before allo-HCT.

Although patients who achieved MMR showed superior outcomes, especially in the disease progression and de novo BP groups, it has been unclear whether achieving a deeper treatment response in itself contributed to better transplant outcomes. It is possible that patients with sensitivity to TKIs or chemotherapy were more likely to achieve a deeper treatment response. Moreover, in the current study, post allo-HCT TKI maintenance was more frequently performed in patients with MMR than in those with CCyR or CHR. This may have contributed to a longer DFS and a lower incidence of relapse because we did not treat prophylactic or preemptive intervention with TKIs as relapse. Indeed, post allo-HCT TKI maintenance is still under debate [6]. Based on the results of the current study, patients who do not achieve MMR might be recommended to receive post allo-HCT TKI maintenance because they are considered to be at a high risk of relapse, especially in the disease progression and de novo BP groups.

In conclusion, transplant outcomes in CML-CP were superior in patients with MMR. Patients with disease progression or who initially present BP might be recommended to achieve MMR before allo-HCT, whereas those with an insufficient response might need to proceed to allo-HCT without delay. Further prospective studies are warranted to clarify the optimal depth of the treatment response for CML-CP.

This study was approved by the Institutional Review Board of Jichi Medical University Saitama Medical Center in accordance with the Helsinki Declaration.

The authors declare no conflicts of interest.