Treatment of Newly Diagnosed Philadelphia Chromosome Positive Acute Lymphoblastic Leukemia Using Tyrosine Kinase Inhibitors in Combination with Chemotherapy: A Patient-Centered Benefit-Risk Assessment

BACKGROUND: Philadelphia Chromosome Positive Acute Lymphoblastic Leukemia (Ph+ALL), an aggressive disease with historically poor survival in adults, is treated with tyrosine kinase inhibitors (TKI) plus chemotherapy. Choosing a TKI involves considering a variety of factors, including the benefits and risks of each therapy. This study aims to understand patients’ preferences for these benefits and risks, and how this varies between patients.

METHODS: In a US-based study conducted between February and April 2020, adult patients (≥ 18 years) with self-reported Ph+ALL (newly diagnosed and relapsed/refractory) completed an online discrete choice experiment (DCE) comprising 12 experimental and 2 internal validity choice tasks. For each task, patients were asked to choose between 2 hypothetical treatments with different levels of benefits and risks. Attributes and levels in the DCE included overall survival (OS, 30–90 months in 5-month increments), duration of remission (DOR, 15–75 months in 5-month increments), risk of a major CV event (0%, 25%, 50%), and risk of myelosuppression (0%, 50%, 100%). A multinomial logit model was used to estimate patients’ preferences expressed as weights based on part-worth utilities with 95% CI . A benefit-risk assessment was subsequently conducted to estimate the proportion of patients who would be expected to prefer different TKI profiles. The hypothetical benefit-risk profiles included within this analysis are outlined in Table 1.

RESULTS: The DCE was completed by 201 patients. Mean age was 44.8 ± 12.9 years, 17% had ≥2 relapses, 58% were diagnosed >12 months previously, and 67% were in remission. The most common ongoing therapies were dasatinib (24%) and imatinib (18%). Internal validity was high: 88% passed the choice dominance test (where one treatment option was superior for all 4 attributes) and 98% never selected the same option. When selecting a preferred treatment, patients cared the most about a 1-month increase in OS (utility=0.032 [95% CI, 0.021–0.042]) and a 1-month increase in duration of remission (utility=0.017 [95% CI, 0.012–0.023]) and they cared less about reducing the risk of a major CV event by 1% (utility=0.011 [95% CI, 0.008–0.013]) and reducing the risk of myelosuppression by 1% (utility=0.009 [95% CI, 0.008–0.010]). Patients were willing to tolerate a 2.9% (95% CI, 1.78–4.02) increase in risk of a major CV event for 1 additional month of OS, and 1.59% (95% CI, 0.95%–2.24%) increase in risk of a major CV event for an additional 1 month in remission.

Figure 1 shows the proportion of patients who would be expected to prefer each of the TKIs, given only the choice between the 4 hypothetical TKIs outlined in Table 1. The results are presented by the overall sample and by subgroups of patients. When the analysis is run for the whole sample, 61.4% of patients are expected to prefer the benefit-risk profile of TKI A, choosing the longer OS and duration of remission, despite the higher CV and myelosuppression risk. However, treatment preference varied between patient subgroups, with greater proportions of patients expected to prefer the benefit-risk profile of TKI A if they had already experienced more than 3 prior lines of Ph+ALL treatment, were in remission, had not relapsed, or had fewer physical limitations. In contrast, other patients were less likely to tolerate higher risks for improvements in remission and survival, such as patients who were diagnosed less than 1 year ago and those who had just started treatment.

CONCLUSIONS: Most Ph+ALL patients are willing to accept an increased risk of CV events and myelosuppression in exchange for the improvements in survival and duration of remission generated by frontline treatment. However, risk tolerance varies between patients, with some subgroups of patients being more willing to tolerate increased risks. Consideration of patient specific preferences observed in this study can further assist clinicians in selection of appropriate treatment in discussion with their patients.