Tyrosine kinase inhibitors (TKIs) have improved the outcome of chronic myeloid leukemia patients (CML) with faster reduction of the burden of disease. Robust and standardized definitions of molecular responses have been introduced and are now universally accepted (1). Early molecular monitoring has become particularly useful to identify patients who might be considered for a prompt change of treatment. First evidence correlated with the early reduction of the molecular burden with the cytogenetic response (2, 3) and the lack of disease progression (3, 4). The cut-off of nearly 10% achieved within the first 3 months was first identified by Marin et al. (5) as predictive of better overall survival (OS). The same level of response, namely the early molecular response or EMR (BCR::ABL1 ratio ≤ 10% at 3 months) was then replicated in other studies as always associated with better OS, event-free survival (EFS) and subsequent increased possibility to achieve in the long-term a deep molecular response (BCR::ABL1 ratio ≤ 0.0032% or MR4.5) (6–8). The significance of EMR was then validated also with second generation TKIs, dasatinib and nilotinib, when tested as frontline treatment in two sponsored clinical trials (9, 10).

Previous work across several cancer populations (11, 12) has shown that health-related quality of life (HRQoL) data provides independent prognostic information for survival and, some evidence in the CML setting, indicated that HRQoL data may also predict achievement of a major molecular response (13). Therefore, leveraging from an international trial (14), we explored the prognostic value of baseline HRQoL data for the achievement of a BCR::ABL1 transcript level of ≤10% after 3 months of TKI treatment, being a key milestone of patients management. A Secondary objective was to describe the baseline HRQoL profile by frequently used prognostic scores: Sokal, EUTOS and ELTS.

The SUSTRENIM trial (Clinical Trial number 02602314), was a prospective, interventional, randomized, two arms, study in newly diagnosed chronic phase CML (CP-CML) patients treated with a second generation TKI (Nilotinib, NIL) or with Imatinib (IM) followed by switching to NIL in absence of optimal response (14). Ethics committees of all participating sites approved this study protocol. For the purposes of this post-hoc analysis, the cohort consisted of CML patients who had a completed HRQoL assessment at study entry (i.e., before study treatment start) and were randomly assigned to receive first-line treatment with NIL or IM between November 2016 and January 2021.Treatment responses were expressed on the International Scale (IS %) and were evaluated according to ELN2020 recommendations every 3 months (15). Early molecular response (EMR) was defined by BCR::ABL1 transcript levels ≤10% on the International Scale at 3 months of TKI treatment (5). HRQoL was assessed using the cancer-generic EORTC QLQ-C30 (16) and the disease specific EORTC QLQ-CML24 questionnaires (17). The QLQ-C30 consists of five functioning scales: physical, role, emotional, cognitive, and social; three symptom scales: fatigue, nausea/vomiting, and pain; six single-item scales: dyspnea, insomnia, appetite loss, constipation, diarrhea, and financial impact; and the global health status/QoL scale. The EORTC QLQ-CML24 consists of six scales: symptom burden, impact on worry/mood, impact on daily life, satisfaction with care and information, body image problems and satisfaction with social life). The main characteristics of newly diagnosed CML patients were summarized overall, using frequencies, proportions, means and standard deviations (SD), medians and interquartile ranges (IQR), depending on the type of variable. Given the exploratory nature of the study, all HRQoL scales were considered for their potential prognostic value for EMR. Univariate and multivariate logistic regression analyses were conducted to assess the prognostic effect of selected baseline socio-demographic and clinical characteristics along with HRQoL scales on EMR at 3 months. The starting univariate model included: sex, ECOG performance status, Sokal risk score, treatment arm, and all the HRQoL scales as continuous variables.

Overall, 448 newly diagnosed CML patients were enrolled and randomized to NIL and IM, 436 (97%) had a HRQoL assessment completed. Of them, 222 (51%) and 214 (49%) patients were randomly assigned to the NIL and IM arms, respectively. EMR was achieved in 190/202 (94%) and 151/194 (78%) patients, respectively (p<0.001). The median age of patients was 55 years (interquartile range, IQR 46-64) with a male predominance (57%). The Sokal score stratification identified 179 (41%), 185 (43%), and 70 (16%) as low, intermediate and high risk, respectively. According to the ELTS score, 274 (63%), 121 (28%), and 39 (9%), were classified as low, intermediate, and high risk, respectively. According to the EUTOS score, 375 (91%) were classified as low risk and 35 (9%) were classified as high risk. Details are reported in Table 1 .

In the univariate logistic regression analysis, EMR was significantly associated with Sokal risk stratification, treatment arm and the following patient-reported HRQoL: physical functioning, role functioning, social functioning scales of the EORTC-QLQ-C30 and body image problems of the EORTC-QLQ-CML24. In the multivariate analysis, the following variables retained a significance: the Sokal score stratification for the high category (OR 0.13, 95% CI 0.05-0.33, p<0.001) and intermediate risk (OR 0.43, 95% CI 0.18-0.96, p=0.046), the treatment arm with nilotinib (OR 4.61, 95% CI 2.31-9.88, p<0.001), and the role functioning scale of the EORTC-QLQ-C30 (OR 1.34, 95% CI 1.10-1.63, p<0.001). This latter indicates a 34% increase in the odds of achieving an EMR with every 10-point increase (i.e., improving) in the baseline role functioning scale. Details are reported in Table 2 .

The three CML prognostic scoring systems were able to disentangle the baseline HRQoL profiles of patients are reported in the ( Supplementary Materials S1 - S3 ).

Patients classified as high-risk by the Sokal risk score reported a significantly worse HRQoL profile than those classified as intermediate- and low-risk, as described by either worse functional impairments (i.e. lower global health status/QoL, physical and role functioning scale scores) or higher symptom severity (i.e. higher pain and dyspnea scale scores) and lower satisfaction with social life (disease-specific domains).

We found that the role functioning scale of the EORTC QLQ-C30, which broadly reflects the ability of patients to perform daily activities, provides independent prognostic information for the achievement of EMR. This finding supports the value of a more patient-centric approach at diagnostic workup of CML as it may offer unique information to better identify those patients who are more likely to achieve an important treatment milestone.

Similar to our result, a recent longitudinal HRQoL analysis, including pooled data from the BFORE trial, showed that clinical improvement was associated with variable effect on different dimensions of HRQoL, and patients achieving deep response (i.e. MR⁵) reported the greatest improvement in emotional well-being and leukemia specific domains (18).

A possible explanation of our findings could be the mediating role of adherence to therapy. That is, those patients who have higher ability to perform daily activities at diagnosis may be facilitated in fitting treatment schedule into their daily life, thereby adopting a more effective medication taking behavior and ultimately achieving an EMR at 3 months. However, we note that in our cohort, achievement of MMR at 6 months was not significantly associated with any baseline HRQoL domain (data not shown). In any case, given the exploratory nature of the analysis, our findings of the prognostic value of the role functioning scale of the EORTC QLQ-C30, should be further corroborated in future prospective studies.

Of note, none of the EORTC QLQ-CML24 domains emerged in the multivariate model, possibly suggesting that the EORTC QLQ-C30 alone, well captures key important aspects of patients with CML. As expected, treatment with nilotinib was significantly associated with a higher probability of achieving EMR, and this is in line with previous evidence on the higher efficacy of second-generation TKIs in inducing faster molecular responses compared to imatinib (19). Similarly, patients with lower Sokal risk scores had higher EMR rates, consistent with its prognostic role in this disease (20).

Our descriptive analysis investigating HRQoL profile according to the commonly used prognostic scoring systems, suggests that Sokal-risk score can help to disentangle some important HRQoL parameters of newly diagnosed patients.

In conclusion, our results suggest that assessment of HRQoL at diagnostic workup may offer independent prognostic information associated with a greater likelihood to achieve EMR.

Presented in part at the 29th European Hematology Association Congress (EHA 2024 Hybrid Congress) held in Madrid, Spain, June 13 to 16, 2024.