Multicenter Retrospective Analysis on the Clinical Benefit of Post-Transplant Tyrosine Kinase Inhibitor (TKI) Maintenance Therapy in Ph-Positive Acute Lymphoblastic Leukemia Suggests Improved GvHD-Free, Relapse Free Survival and Non-Relapse Mortality, but Not the Risk of Relapse

Introduction: Post-transplant maintenance therapy with tyrosine kinase inhibitor (TKI) is very common practice in the management of Philadelphia positive acute lymphoblastic leukemia (Ph+ALL) following allogeneic hematopoietic stem cell transplantation (HCT) with an expectation to reduce the risk of relapse. However, it is still not clearly demonstrated whether TKI brings clinical benefit to the patients (pts). Conflicting retrospective results were reported while data from a randomized controlled trial is lacking. The present study evaluated the benefit of TKI in 245 pts from 7 international HCT centers with respect to for overall (OS) and relapse-free survival (RFS), non-relapse mortality (NRM), cumulative incidence of relapse (CIR), GvHD-free, relapse-free survival (GRFS).

Patients and Method: We have retrospectively reviewed 245 pts with Ph+ ALL who underwent their first HCT in complete remission (CR) from 2000 till 2024. The OS, RFS and GRFS were calculated using the Kaplan-Meier method and analyzed by a log-rank test, while the CIR, NRM and chronic GvHD (cGvHD) incidence were calculated considering competing events and analyzed using the Fine-Gray model. Time-dependent analysis using Cox or Fine-gray model was conducted using time to TKI starts as a time-dependent covariate, and compared the outcomes between the TKI vs no-TKI groups using Mantel-Byar test which will avoid immortal bias.

Results: Out of 245 pts, 126 (51.4%) received TKI. ABL1 kinase-domain mutation (KDM) was detected in 18 pts prior to HCT: T315I (n=8) and others (n=10), of whom 12 pts received TKI. The median time to start TKI after HCT was 3 months (range: 0.5-115 months) with imatinib (n=33), dasatinib (n=53), ponatinib (n=31), or other TKIs (n=7). The median duration of TKI was 23 months (0-163 months).

Between the TKI and no-TKI groups, the characteristics were very similar in terms of age (p=0.17), sex (p=0.78), BCR-ABL qPCR level prior to HCT (p=1.0), and the presence of ABL kinase mutation (p=0.18). However, a significant difference was observed with respect to the remission status prior to HCT, with a higher proportion of TKI group received HCT in second CR or beyond (26.6% vs 12.4%, p=0.012).

With a median follow-up duration of 42 months among survivors, 101 (41.2%) deaths and 52 (21%) relapses were observed out of 245 pts, resulting in 63.4% of OS, 55.2% of RFS, 61.3% of GRFS, 15.9% of CIR, 24.8% of NRM and 33.0% of cGvHD rates at 3 years.

In comparison to non-TKI group, the TKI group showed less frequent occurrence of deaths (27.0% vs 56.3%, p<0.001), NRM (8.7% vs 47.1%, p<0.001), cGvHD (22.4% vs 38.7%, p=0.008), but a more frequent occurrence of relapse (27.8% vs 14.3%, p=0.012). However, because this analysis did not take account of TKI start day as a time-dependent covariate, we have conducted time-dependent analysis subsequently. Of interest, the time-dependent analysis showed trends of favouring TKI use for OS (HR, 0.8, P=0.33), RFS (HR, 0.73, P=0.14), and the significantly favourable impact of TKI use toward GRFS (HR, 0.74, P<0.0001), NRM (HR, 0.37, P= 0.003), and cGVHD (HR, 0.58, P= 0.04). However, it did not affect CIR (HR, 1.674, P= 0.07).

Multivariable analysis did not demonstrate the TKI as an independent factor for OS, RFS or CIR. Of surprise, TKI use was confirmed as an independently favourable prognostic factor for GRFS (HR 0.565, p=0.012) and for NRM (HR 0.338, p=0.001).

Out of 126 pts who received TKI, TKI was discontinued in 51 (41%) pts, of whom 22 pts completed the planned treatment, while 29 pts discontinued due to either relapse (n=12), TKI-related toxicity (n=14), and death (n=3). In the TKI group, 17 pts relapsed of whom 11 pts relapsed during TKI therapy and 6 pts relapsed after TKI stopped due to toxicity. The incidence of relapse after TKI starts was 22.8%, 33.4% and 33.4% at 3, 4 and 5 years, showing a plateau after 4 years.

The current study strongly suggested a significant risk of relapse in the patients having ABL1 KDM, both T315I mutation and others. While only 11 (9.6%) out of 227 pts without ABL1 KDM relapsed, 6 (50%) out of 18 pts carrying ABL1 KDM had relapsed (p=0.001).

Conclusion: The present results again can not confirm TKI's benefit with respect to relapse risk while it strongly suggests TKI’s benefit by improving GRFS and reducing the risk of NRM. The patients with ABL1 KDM prior to HCT carry the highest risk of relapse even with TKI, implying additional intervention would be required on top of TKI-based maintenance.