Key Points
-
•
Joint model suggests that zanubrutinib resulted in a lower patient-reported symptom burden than ibrutinib in CLL.
-
•
Recurrent symptomatic deterioration may act as a predictor for an increased risk of PFS events and shortened DoR.
Visual Abstract

Abstract
The ALPINE trial was a randomized, open-label study comparing the efficacy and safety of zanubrutinib vs ibrutinib monotherapy in patients with relapsed/refractory (R/R) chronic lymphocytic leukemia (CLL)/small lymphocytic lymphoma (SLL). Using patient-reported outcome (PRO) data from the ALPINE trial, this study aimed to investigate the associations between disease-specific patient-reported symptoms and the clinical outcomes of progression-free survival (PFS) and duration of response (DoR). After adjusting for baseline stratification factors, zanubrutinib was associated with lower PRO-based recurrent symptomatic deterioration than ibrutinib. Zanubrutinib also demonstrated significantly lower risk of disease progression and better DoR across all disease-specific patient-reported symptoms. Compared with ibrutinib, zanubrutinib demonstrated a 28% to 33% reduction in the risk of a PFS event after adjusting for global health status, physical functioning, role functioning, fatigue, pain, nausea and vomiting, diarrhea, and constipation (all P < .02). Similarly, zanubrutinib demonstrated a 31% to 51% reduced risk of shortened DoR after adjusting for the patient-reported scores (all P < .02). These are crucial aspects in clinical decision-making because they highlight zanubrutinib’s association with improved clinical outcomes and patient-reported symptomatology. These findings reinforce the enhanced efficacy profile of zanubrutinib in patients with R/R CLL/SLL by demonstrating its lower patient-reported symptom deterioration in addition to increased PFS and better DoR. This trial was registered at www.clinicaltrials.gov as NCT03734016.
Introduction
Chronic lymphocytic leukemia (CLL), including small lymphocytic lymphoma (SLL), is the most common form of leukemia in Western countries, with an incidence of 4.2 per 100 000 person-years.1, 2, 3, 4 Despite advancements in treatment, CLL/SLL remains a chronic, progressive disease, often leading to substantial symptom burden and eventual disease progression that affect both health-related quality of life and survival outcomes.1 Patients frequently experience symptoms such as fatigue, pain, nausea, vomiting, and diarrhea, with symptom severity usually increasing as the disease progresses.3,4
Primary analyses of the phase 3 ALPINE trial (BGB-3111-305; ClinicalTrials.gov identifier: NCT03734016)5 demonstrated that, at a median follow-up of 29.6 months, zanubrutinib was superior to ibrutinib in investigator-assessed progression-free survival (PFS; hazard ratio [HR], 0.65; 95% confidence interval [CI], 0.49-0.86; P = .002) and investigator-assessed overall response rate (83.5% vs 74.2%). Furthermore, analyses based on ALPINE’s extended follow-up period (median, 42.5 months) reported that the PFS benefit with zanubrutinib vs ibrutinib was sustained (HR, 0.68; 95% CI, 0.54-0.84).6 Notably, PFS benefit was also sustained in high-risk patients (ie, those with del(17p)/TP53 mutation status), as evidenced by a 49% reduction in the risk of disease progression or death (HR, 0.51; 95% CI, 0.33-0.78) and a substantial increase in patients progression-free at 3 years with zanubrutinib compared with ibrutinib (59% vs 39%).6 In addition, compared with ibrutinib, treatment with zanubrutinib was associated with a lower incidence of serious adverse events (AEs) that led to treatment discontinuation (15.4% vs 22.2%) and a lower incidence of cardiac disorders (21.3% vs 29.6%).5 ALPINE included patient-reported outcomes (PROs) as a secondary objective, with patients in the zanubrutinib arm demonstrating clinically meaningful improvements in the global health status/quality of life (GHS/QoL), physical and role functioning, and pain and fatigue symptoms of the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30).7 Because PRO questionnaires measure the patient’s experience of symptoms during treatment, PROs are direct reports by patients regarding their health (eg, pain, fatigue), whereas AEs are physician-reported, broad, untoward medical occurrences that may or may not be caused by the treatment.
Previous research has examined correlations between patient-reported symptoms and survival outcomes in various cancer populations; however, most studies have focused on baseline PRO scores or changes at fixed time points, overlooking the transient nature of PROs (ie, patients may experience repeated cycles of deterioration and improvement throughout a randomized clinical trial).8, 9, 10, 11, 12 In addition, except for Yan et al,13 most evidence is derived from solid tumor populations, and there are limited data available for hematologic malignancies. Notably, these previous studies did not directly assess the association between PRO-based recurrent symptomatic deterioration events and their influence on key clinical outcomes (PFS and duration of response [DoR]) in patients with CLL/SLL.
These analyses aimed to examine the relationship between patient-reported symptoms and clinical end points (PFS and DoR) in patients with relapsed/refractory (R/R) CLL/SLL within the ALPINE trial. We hypothesized that this framework would (1) further differentiate zanubrutinib from ibrutinib using clinical-event outcomes and sustained PRO symptom management and (2) help identify PRO end points associated with clinical events that may be explored prospectively as complementary markers to optimize patient-centered treatment strategies in future CLL/SLL trials.
Methods
Study design and patient population
The ALPINE trial (ClinicalTrials.gov identifier: NCT03734016) was a global, randomized, open-label, phase 3 study comparing the efficacy and safety of zanubrutinib and ibrutinib monotherapy in adult patients with R/R CLL/SLL.6 Eligible patients were aged ≥18 years, had a confirmed diagnosis of CLL or SLL that met the International Workshop on Chronic Lymphocytic Leukemia criteria, were R/R to ≥1 previous systemic therapy, and had an Eastern Cooperative Oncology Group performance status of ≤2. Exclusion criteria included previous treatment with a Bruton tyrosine kinase inhibitor and known prolymphocytic leukemia or possible Richter transformation. Patients who met the inclusion criteria were randomized 1:1 to receive oral zanubrutinib 160 mg twice daily or oral ibrutinib 420 mg once daily until disease progression or withdrawal. Randomization was stratified by age (<65 vs ≥65 years), geographic region (China vs non-China), refractory status (yes vs no), and del(17p)/TP53 mutation status (present vs absent). The study was conducted in accordance with the Good Clinical Practice guidelines, the principles of the Declaration of Helsinki, and local laws and regulations.14 A list of ethics committees for the study sites can be found in a previous ALPINE publication.6 All patients provided a written informed consent before participating.
Study end points
Investigator-defined PFS and DoR were analyzed as the clinical-event measures. PROs were assessed using the EORTC QLQ-C3015 and the EuroQol-5-Dimension-5-Level questionnaire16 at baseline, cycle 1, and then every third cycle until the end of treatment. Each cycle constituted 28 days. For these analyses, data up to cycle 43 were used, representing ∼3.5 years of treatment. Fatigue, pain, and physical functioning were considered disease related, whereas the remaining symptoms were considered treatment related. Although pain is not usually a symptom associated with CLL/SLL, it is classified as such in these analyses because it is a common side effect of AEs. The PRO questionnaires were completed by patients at the clinical site before any other study procedure.
Statistical analyses
All randomized patients in the intent-to-treat population who completed the baseline and ≥1 postbaseline PRO questionnaire were included in the analyses. PRO end points comprised the GHS/QoL, physical and role functioning, fatigue, pain, nausea and vomiting, diarrhea, and constipation symptom domains of the EORTC QLQ-C30, as well as the EuroQol-5-Dimension-5-Level questionnaire’s visual analog scale (EQ-VAS). A clinically meaningful recurrent deterioration event was defined as a 10-point change from baseline in a direction that indicated worsening of symptoms (ie, negative direction for the GHS/QoL, physical and role functioning, and EQ-VAS and positive direction for the symptom scales).17 Each unique deterioration event (a deterioration separated by ≥1 nondeterioration change score) was counted as a recurrent event. For any given analysis, a patient may have had 0 to many recurrent deterioration events.
Model framework
The framework included a linear mixed model (LMM) to assess adjusted changes from baseline treatment effects (eg, PRO change from baseline–zanubrutinib treatment effect). This parameter indicates how much PRO change scores differ between treatment arms. The treatment-by-time interaction, represented by the PRO change from baseline–zanubrutinib treatment × time interaction parameter, quantifies the difference in treatment effects on adjusted PRO change scores over time.
A recurrent-event Cox proportional hazards model was used to assess how treatment affects the recurrence of symptomatic deterioration and whether longitudinal PRO symptom trajectories are predictive of repeated deterioration over time. The recurrent deterioration event–zanubrutinib treatment effect illustrates the first type of effect. In contrast, the recurrent deterioration event–longitudinal effect illustrates the second type. The recurrent-event submodel operationalizes PRO worsening as discrete, clinically meaningful deterioration episodes defined by the 10-point threshold, capturing the frequency and timing of symptom worsening events.
A clinical-event Cox proportional hazards model was fit to evaluate the impact of treatment, longitudinal PRO trajectories, and recurrent symptomatic deterioration on the clinical end points of investigator-assessed PFS and DoR. In particular, the clinical event–zanubrutinib treatment effect parameter measures the direct effect of treatment on the risk of experiencing a clinical event. The clinical event–longitudinal effect parameter evaluates whether changes in PRO scores over time are predictive of clinical events. Finally, the clinical event–recurrent deterioration event (frailty) parameter quantifies the association between the recurrent symptomatic deterioration process and the clinical event through shared, patient-level risk. The relationships between the longitudinal and time-to-event outcomes are represented by association parameters, which are essential in joint models because they quantify the relationship between the longitudinal and time-to-event processes. The association parameter captures the effect of the longitudinal outcome and the hazard or risk of the events of interest. In this analysis, the frailty term adjusts for the patient-specific risk of repeated events. The frailty association parameter quantifies the linkage between the recurrent deterioration process and the clinical-event process (eg, PFS/DoR), such that patients with a higher underlying tendency for recurrent deterioration also tend to have a higher hazard of the clinical event.18 The proportional hazards assumptions for each model were assessed by examining the Schoenfeld residuals. A graphical illustration of this joint model framework is presented in Figure 1.
Figure 1.

Graphical illustration of the joint model framework.
Per the study design, each model included the stratification factors of age, geographic region, refractory status, del(17p)/TP53 mutation status, cancer type (CLL vs SLL), and baseline PRO measurement (in the LMM). All analyses were conducted in R statistical software (version 4.4.1) using the JMBayes2 (0.5-0), nlme (3.1-165), and survival (3.7-0) packages.
Results
Baseline characteristics
At the data cutoff date (28 February 2024), the intent-to-treat population consisted of 652 patients (327 randomized to zanubrutinib and 325 to ibrutinib). Patient demographics and baseline disease characteristics, which have been described previously7 and are presented in Table 1, were generally well balanced between the treatment arms. The average completion rate for PROs was >80% throughout the trial.
Table 1.
Baseline demographics and clinical characteristics
| Characteristic | PFS |
DoR |
||
|---|---|---|---|---|
| Zanubrutinib (n = 308) | Ibrutinib (n = 298) | Zanubrutinib (n = 275) | Ibrutinib (n = 249) | |
| Age, median (range), y | 67 (35-90) | 67.5 (35-89) | 67 (35-89) | 67 (35-86) |
| Sex, n (%) | ||||
| Male | 201 (65.3) | 212 (71.1) | 178 (64.7) | 175 (70.3) |
| Female | 107 (34.7) | 86 (28.9) | 97 (35.3) | 74 (29.7) |
| Race/ethnicity, n (%) | ||||
| White | 247 (80.2) | 242 (81.2) | 220 (80.0) | 203 (81.5) |
| Asian | 45 (14.6) | 42 (14.1) | 40 (14.5) | 34 (13.7) |
| Black or African American | 4 (1.3) | 1 (0.3) | 4 (1.5) | 1 (0.4) |
| Native Hawaiian or other Pacific Islander | 3 (1.0) | 0 | 3 (1.1) | 0 |
| Not reported | 6 (1.9) | 11 (3.7) | 6 (2.2) | 9 (3.6) |
| Other | 2 (0.6) | 2 (0.7) | 1 (0.4) | 2 (0.8) |
| Unknown | 1 (0.3) | 0 | 1 (0.4) | 0 |
| ECOG performance status, n (%) | ||||
| 2 | 6 (1.9) | 9 (3.0) | 6 (2.2) | 7 (2.8) |
| 1 | 181 (58.8) | 176 (59.1) | 161 (58.5) | 154 (61.8) |
| 0 | 121 (39.3) | 113 (37.9) | 108 (39.3) | 88 (35.3) |
| Geographic region, n (%) | ||||
| Asia | 47 (15.3) | 43 (14.4) | 42 (15.3) | 35 (14.1) |
| Australia/New Zealand | 27 (8.8) | 28 (9.4) | 22 (8.0) | 23 (9.2) |
| Europe | 185 (60.1) | 177 (59.4) | 166 (60.4) | 154 (61.8) |
| North America | 49 (15.9) | 50 (16.8) | 45 (16.4) | 37 (14.9) |
| Del(17p) status, n (%) | ||||
| Absent | 239 (77.6) | 233 (78.2) | 214 (77.8) | 197 (79.1) |
| Present | 69 (22.4) | 65 (21.8) | 61 (22.2) | 52 (20.9) |
| Refractory status, n (%) | ||||
| No | 224 (72.7) | 215 (72.1) | 203 (73.8) | 182 (73.1) |
| Yes | 84 (27.3) | 83 (27.9) | 72 (26.2) | 67 (26.9) |
ECOG, Eastern Cooperative Oncology Group.
Depending on the PRO domain, 44 to 46 patients were excluded from the current PFS modeling analyses and 126 to 128 were excluded from the current DoR modeling analyses due to incomplete or missing PRO data. A summary of the numbers of patients in each domain, including all randomized patients who completed the baseline questionnaire and had available data on the clinical and recurrent deterioration events, is presented in supplemental Table 1. The number of recurrent deterioration events from cycles 1 to 43, including unscheduled visits, ranged from 0 to 5. The number of recurrent diarrhea symptom deterioration events by clinical event and treatment arm is presented in Table 2.
Table 2.
Observed number of recurrent diarrhea symptom deterioration events
| No. of recurrent events | n (%) | Cumulative n (%) | n (%) | Cumulative n (%) | |
|---|---|---|---|---|---|
| PFS | Zanubrutinib (n = 308) | Ibrutinib (n = 298) | |||
| 0 | 213 (69.2) | 213 (69.2) | 193 (64.8) | 193 (64.8) | |
| 1 | 60 (19.5) | 273 (88.6) | 60 (20.1) | 253 (84.9) | |
| 2 | 20 (6.5) | 293 (95.1) | 22 (7.4) | 275 (92.3) | |
| 3 | 7 (2.3) | 300 (97.4) | 16 (5.4) | 291 (97.7) | |
| 4 | 6 (1.9) | 306 (99.4) | 6 (2.0) | 297 (99.7) | |
| 5 | 2 (0.6) | 308 (100) | 1 (0.3) | 298 (100) | |
| DoR | Zanubrutinib (n = 275) | Ibrutinib (n = 249) | |||
| 0 | 193 (70.2) | 193 (70.2) | 164 (65.9) | 164 (65.9) | |
| 1 | 57 (20.7) | 250 (90.9) | 47 (18.9) | 211 (84.7) | |
| 2 | 13 (4.7) | 263 (95.6) | 21 (8.4) | 232 (93.2) | |
| 3 | 7 (2.5) | 270 (98.2) | 11 (4.4) | 243 (97.6) | |
| 4 | 4 (1.5) | 274 (99.6) | 6 (2.4) | 249 (100) | |
| 5 | 1 (0.4) | 275 (100) | 0 | 249 (100) | |
All models achieved acceptable convergence based on visual inspection of the Markov chain trace and density plots, as well as evaluation of the statistic. Further details of model convergence are presented in the supplemental Methods.
LMMs
PFS
In the LMM with PFS as the clinical event, a statistically significant zanubrutinib treatment effect on change from baseline was observed for diarrhea (−2.088; 95% CI, −3.812 to −0.352; P = .0179; Table 3). On average, patients in the zanubrutinib arm experienced a 2.1-point lower level of deterioration in diarrhea symptoms than patients in the ibrutinib arm over the course of the study. The corresponding statistic for this effect was ∼1.00, indicating acceptable convergence. For the remaining EORTC QLQ-C30 domains (GHS/QoL, physical and role functioning, fatigue, pain, nausea and vomiting, and constipation) and the EQ-VAS, no significant changes from baseline or treatment-by-time effects were observed.
Table 3.
Linear mixed and survival models for PFS clinical events adjusted for treatment efficacy, PRO ΔBL, and recurrent symptomatic deterioration events
| PRO domain | Parameter | β (95% CI) | P value | ∗ | HR (95% CI) |
|---|---|---|---|---|---|
| EORTC QLQ-C30 GHS/QoL | PRO ΔBL–zanubrutinib treatment effect† | 0.558 (−1.608 to 2.674) | .6000 | 1.001 | NA |
| PRO ΔBL–zanubrutinib treatment† × time‡ interaction | 0.028 (−0.038 to 0.094) | .3997 | 1.005 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect† | NA | .6841 | 1.002 | 0.927 (0.651-1.327) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.002 | 0.860 (0.846-0.873) | |
| Clinical event–zanubrutinib treatment effect† | NA | .0153 | 1.000 | 0.717 (0.543-0.939) | |
| Clinical event–longitudinal effect | NA | .2337 | 1.003 | 0.995 (0.987-1.003) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .9007 | 1.029 | 0.757 (0.012-40.124)§ | |
| EORTC QLQ-C30 physical functioning | PRO ΔBL–zanubrutinib treatment effect | 1.004 (−0.836 to 2.814) | .2785 | 1.000 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | −0.010 (−0.063 to 0.043) | .7064 | 1.001 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .9457 | 1.005 | 1.016 (0.642-1.596) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.034 | 0.851 (0.832-0.867) | |
| Clinical event–zanubrutinib treatment effect | NA | .0072 | 1.003 | 0.708 (0.549-0.913) | |
| Clinical event–longitudinal effect | NA | .6903 | 1.001 | 1.002 (0.993-1.011) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .2373 | 1.007 | 1.264 (0.824-1.933)§ | |
| EORTC QLQ-C30 role functioning | PRO ΔBL–zanubrutinib treatment effect | −0.306 (−2.644 to 2.045) | .8031 | 1.000 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | 0.042 (−0.031 to 0.113) | .2713 | 1.001 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .5495 | 1.002 | 1.120 (0.781-1.627) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.035 | 0.908 (0.897-0.918) | |
| Clinical event–zanubrutinib treatment effect | NA | .0125 | 1.009 | 0.709 (0.537-0.930) | |
| Clinical event–longitudinal effect | NA | .6456 | 1.000 | 0.999 (0.992-1.005) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .0263 | 1.190 | 1.988 (1.074-7.009)§ | |
| EORTC QLQ-C30 fatigue | PRO ΔBL–zanubrutinib treatment effect | −0.212 (−2.607 to 2.194) | .8619 | 1.002 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | −0.009 (−0.082 to 0.062) | .8137 | 1.001 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .9421 | 1.002 | 0.987 (0.695-1.399) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.009 | 1.116 (1.102-1.132) | |
| Clinical event–zanubrutinib treatment effect | NA | .0132 | 1.009 | 0.719 (0.552-0.934) | |
| Clinical event–longitudinal effect | NA | .1557 | 1.000 | 1.005 (0.998-1.011) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .4593 | 1.021 | 1.704 (0.289-25.381)§ | |
| EORTC QLQ-C30 pain | PRO ΔBL–zanubrutinib treatment effect | −0.904 (−3.308 to 1.485) | .4671 | 1.001 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | 0.035 (−0.033 to 0.103) | .3299 | 1.003 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .4468 | 1.004 | 0.873 (0.616-1.235) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.012 | 1.110 (1.097-1.124) | |
| Clinical event–zanubrutinib treatment effect | NA | .0088 | 1.001 | 0.713 (0.554-0.917) | |
| Clinical event–longitudinal effect | NA | .6957 | 1.001 | 1.001 (0.995-1.008) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .7979 | 1.013 | 0.925 (0.499-1.490)§ | |
| EORTC QLQ-C30 nausea and vomiting | PRO ΔBL–zanubrutinib treatment effect | −0.509 (−1.234 to 0.221) | .1697 | 1.001 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | 0.016 (−0.010 to 0.042) | .2521 | 1.006 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .6340 | 1.006 | 0.902 (0.590-1.383) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.006 | 1.288 (1.241-1.337) | |
| Clinical event–zanubrutinib treatment effect | NA | .0112 | 1.033 | 0.696 (0.505-0.919) | |
| Clinical event–longitudinal effect | NA | .3136 | 1.001 | 1.010 (0.990-1.031) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .5867 | 1.103 | 3.082 (0.037-202.201)§ | |
| EORTC QLQ-C30 diarrhea | PRO ΔBL–zanubrutinib treatment effect | −2.088 (−3.812 to −0.352) | .0179 | 1.004 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | 0.010 (−0.046 to 0.066) | .7308 | 1.001 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .9708 | 1.003 | 0.989 (0.632-1.525) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.047 | 1.182 (1.158-1.207) | |
| Clinical event–zanubrutinib treatment effect | NA | .0104 | 1.064 | 0.693 (0.485-0.922) | |
| Clinical event–longitudinal effect | NA | .8197 | 1.006 | 1.001 (0.992-1.010) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .5020 | 1.200 | 3.922 (0.050-179.322)§ | |
| EORTC QLQ-C30 constipation | PRO ΔBL–zanubrutinib treatment effect | 1.335 (−0.638 to 3.321) | .1820 | 1.003 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | −0.012 (−0.075 to 0.053) | .7303 | 1.008 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .6717 | 1.002 | 1.104 (0.703-1.728) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.034 | 1.144 (1.123-1.168) | |
| Clinical event–zanubrutinib treatment effect | NA | .0104 | 1.000 | 0.709 (0.544-0.921) | |
| Clinical event–longitudinal effect | NA | .5593 | 1.002 | 0.998 (0.990-1.006) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .8993 | 1.134 | 1.060 (0.143-6.461)§ | |
| EQ-VAS | PRO ΔBL–zanubrutinib treatment effect | 1.761 (−0.084 to 3.548) | .0607 | 1.000 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | −0.024 (−0.087 to 0.038) | .4536 | 1.006 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .1936 | 1.001 | 0.781 (0.544-1.125) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.029 | 0.861 (0.847-0.874) | |
| Clinical event–zanubrutinib treatment effect | NA | .0125 | 1.004 | 0.666 (0.477-0.916) | |
| Clinical event–longitudinal effect | NA | .5056 | 1.001 | 1.003 (0.994-1.012) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .0000 | 1.058 | 4.522 (1.773-27.519)§ |
Each model was adjusted for the following stratification factors: age (<65 vs ≥65 years), geographic region (China vs non-China), refractory status (yes vs no), del(17p)/TP53 mutation status (present vs absent), cancer type (CLL vs SLL), and baseline PRO measurement (in the LMMs).
ΔBL, change from baseline; NA, not applicable.
An statistic with a value of 1.0 indicated acceptable convergence.
Treatment effect was coded as zanubrutinib vs ibrutinib, with the former as the effect group.
Time was defined as the number of months from baseline.
Association parameter, not HR.
DoR
In the LMM with DoR as the clinical event, the findings were similar to those for PFS. In particular, a statistically significant change from baseline treatment effect was observed for diarrhea (−2.290; 95% CI, −4.223 to −0.375; P = .0189; Table 4). On average, patients in the zanubrutinib arm experienced a 2.3-point lower level of deterioration in diarrhea symptoms over the course of the study than the ibrutinib arm. No significant changes from baseline effects or treatment-by-time interactions were observed for any of the remaining PRO domains.
Table 4.
Linear mixed and survival models for DoR clinical events adjusted for treatment efficacy, PRO ΔBL, and recurrent symptomatic deterioration events
| PRO domain | Parameter | β (95% CI) | P value | ∗ | HR (95% CI) |
|---|---|---|---|---|---|
| EORTC QLQ-C30 GHS/QoL | PRO ΔBL–zanubrutinib treatment effect† | 0.755 (−1.489 to 3.002) | .5080 | 1.004 | NA |
| PRO ΔBL–zanubrutinib treatment† × time‡ interaction | 0.030 (−0.047 to 0.108) | .4557 | 1.004 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect† | NA | .6164 | 1.002 | 0.899 (0.592-1.352) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.003 | 0.868 (0.852-0.883) | |
| Clinical event–zanubrutinib treatment effect† | NA | .0157 | 1.019 | 0.683 (0.484-0.924) | |
| Clinical event–longitudinal effect | NA | .5988 | 1.000 | 1.002 (0.993-1.011) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .8777 | 1.116 | 0.625 (0.007-30.820)§ | |
| EORTC QLQ-C30 physical functioning | PRO ΔBL–zanubrutinib treatment effect | 1.303 (−0.603 to 3.239) | .1913 | 1.005 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | −0.018 (−0.083 to 0.049) | .5916 | 1.017 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .8043 | 1.000 | 0.937 (0.561-1.529) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.083 | 0.839 (0.818-0.859) | |
| Clinical event–zanubrutinib treatment effect | NA | .0067 | 1.002 | 0.648 (0.463-0.890) | |
| Clinical event–longitudinal effect | NA | .0148 | 1.002 | 1.014 (1.003-1.025) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .0188 | 1.034 | 2.269 (1.123-12.960)§ | |
| EORTC QLQ-C30 role functioning | PRO ΔBL–zanubrutinib treatment effect | −0.378 (−2.763 to 2.016) | .7591 | 1.003 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | 0.032 (−0.052 to 0.116) | .4533 | 1.007 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .2852 | 1.000 | 1.243 (0.823-1.831) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.086 | 0.910 (0.897-0.922) | |
| Clinical event–zanubrutinib treatment effect | NA | .0111 | 1.004 | 0.670 (0.482-0.920) | |
| Clinical event–longitudinal effect | NA | .1835 | 1.003 | 1.005 (0.997-1.012) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .1044 | 1.005 | 3.168 (0.772-54.945)§ | |
| EORTC QLQ-C30 fatigue | PRO ΔBL–zanubrutinib treatment effect | −0.054 (−2.427 to 2.346) | .9627 | 1.007 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | −0.016 (−0.095 to 0.064) | .6765 | 1.008 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .9177 | 1.001 | 0.976 (0.658-1.418) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.012 | 1.117 (1.103-1.133) | |
| Clinical event–zanubrutinib treatment effect | NA | .0145 | 1.001 | 0.642 (0.405-0.919) | |
| Clinical event–longitudinal effect | NA | .8981 | 1.006 | 0.999 (0.991-1.008) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .0953 | 1.040 | 10.221 (0.464-371.870)§ | |
| EORTC QLQ-C30 pain | PRO ΔBL–zanubrutinib treatment effect | −0.907 (−3.364 to 1.570) | .4685 | 1.003 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | 0.023 (−0.058 to 0.101) | .5728 | 1.011 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .2113 | 1.001 | 0.801 (0.558-1.125) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.200 | 1.111 (1.097-1.126) | |
| Clinical event–zanubrutinib treatment effect | NA | .0176 | 1.001 | 0.688 (0.495-0.932) | |
| Clinical event–longitudinal effect | NA | .3115 | 1.002 | 0.996 (0.988-1.004) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .3425 | 1.070 | 2.898 (0.143-86.441)§ | |
| EORTC QLQ-C30 nausea and vomiting | PRO ΔBL–zanubrutinib treatment effect | −0.666 (−1.535 to 0.220) | .1363 | 1.004 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | 0.016 (−0.017 to 0.049) | .3260 | 1.007 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .6367 | 1.000 | 0.886 (0.534-1.443) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.021 | 1.328 (1.273-1.386) | |
| Clinical event–zanubrutinib treatment effect | NA | .0144 | 1.046 | 0.648 (0.397-0.922) | |
| Clinical event–longitudinal effect | NA | .3627 | 1.007 | 1.011 (0.985-1.036) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .4524 | 1.020 | 7.498 (0.038-520.352)§ | |
| EORTC QLQ-C30 diarrhea | PRO ΔBL–zanubrutinib treatment effect | −2.290 (−4.223 to −0.375) | .0189 | 1.002 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | 0.013 (−0.062 to 0.087) | .7139 | 1.006 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .8721 | 1.001 | 0.954 (0.559-1.576) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.077 | 1.199 (1.171-1.230) | |
| Clinical event–zanubrutinib treatment effect | NA | .0183 | 1.028 | 0.630 (0.379-0.928) | |
| Clinical event–longitudinal effect | NA | .6771 | 1.011 | 1.003 (0.990-1.016) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .1721 | 1.127 | 14.825 (0.117-596.595)§ | |
| EORTC QLQ-C30 constipation | PRO ΔBL–zanubrutinib treatment effect | 1.273 (−0.932 to 3.442) | .2611 | 1.005 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | −0.033 (−0.113 to 0.046) | .4009 | 1.008 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .7029 | 1.004 | 1.098 (0.665-1.770) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.043 | 1.142 (1.121-1.164) | |
| Clinical event–zanubrutinib treatment effect | NA | .0119 | 1.005 | 0.670 (0.459-0.920) | |
| Clinical event–longitudinal effect | NA | .3176 | 1.000 | 0.995 (0.984-1.005) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .7967 | 1.149 | 1.757 (0.020-185.583)§ | |
| EQ-VAS | PRO ΔBL–zanubrutinib treatment effect | 1.699 (−0.325 to 3.708) | .0991 | 1.005 | NA |
| PRO ΔBL–zanubrutinib treatment × time interaction | −0.006 (−0.073 to 0.062) | .8728 | 1.004 | NA | |
| Recurrent deterioration event–zanubrutinib treatment effect | NA | .3512 | 1.003 | 0.823 (0.547-1.230) | |
| Recurrent deterioration event–longitudinal effect | NA | .0000 | 1.006 | 0.862 (0.848-0.876) | |
| Clinical event–zanubrutinib treatment effect | NA | .0073 | 1.055 | 0.489 (0.217-0.846) | |
| Clinical event–longitudinal effect | NA | .0215 | 1.013 | 1.016 (1.002-1.036) | |
| Clinical event–recurrent deterioration event (frailty) | NA | .0000 | 1.032 | 47.968 (4.102-1551.194)§ |
Each model was adjusted for the following stratification factors: age (<65 vs ≥65 years), geographic region (China vs non-China), refractory status (yes vs no), del(17p)/TP53 mutation status (present vs absent), cancer type (CLL vs SLL), and baseline PRO measurement (in the LMMs).
Abbreviations are explained in Table 3.
An statistic with a value of 1.0 indicated acceptable convergence.
Treatment effect was coded as zanubrutinib vs ibrutinib, with the former as the effect group.
Time was defined as the number of months from baseline.
Association parameter, not HR.
Time-to-event models
PFS
In the recurrent-event models for symptomatic deterioration, no statistically significant differences were observed between treatment arms after adjusting for the PFS and longitudinal models, as well as stratification factors, across all PRO domains. As expected, irrespective of treatment arm, increasing changes from baseline scores in all symptom scales (fatigue, pain, nausea and vomiting, diarrhea, and constipation) were associated with an increased risk of recurrent symptomatic deterioration events (HRs, 1.11-1.29, reflecting an 11%-29% increased risk for each 10-point increase in PRO score), whereas increasing changes from baseline scores in GHS/QoL and both functioning scales, as well as the EQ-VAS, were linked to a reduced risk (protective factor) of recurrent symptomatic deterioration events (HRs, 0.91-0.85, reflecting a 9%-15% reduced risk for each 10-point increase in PRO score). This is illustrated by the recurrent deterioration event–longitudinal effect parameter in Table 3.
In the PFS clinical-event models, after adjusting for recurrent symptomatic deterioration, change from baseline in corresponding symptoms, and stratification factors, zanubrutinib was associated with a statistically significant reduction in the risk of investigator-assessed PFS events vs ibrutinib (HRs, 0.72-0.67, reflecting a 28%-33% reduced risk of disease progression with zanubrutinib treatment). For example, after adjusting for change from baseline in diarrhea symptom scores and corresponding symptomatic deterioration events, zanubrutinib provided a 31% reduction in the risk of a PFS event (HR, 0.693; 95% CI, 0.485-0.922; P = .0104). This is depicted by the clinical event–zanubrutinib treatment effect parameter in Table 3.
Importantly, the frailty prediction of PFS was statistically significant for the role functioning and EQ-VAS domains, irrespective of treatment arm (see clinical event–recurrent deterioration event [frailty] parameter in Table 3). The frailty prediction exhibited acceptable convergence for the role functioning domain and for the EQ-VAS domain ( = 1.190 and 1.058, respectively); we have found that the frailty parameter is the hardest parameter for which to achieve acceptable, stable convergence.
DoR
In the recurrent-event models for symptomatic deterioration, after adjusting for DoR, change from baseline in corresponding symptoms, and stratification factors, the results were consistent with the evidence for PFS (no statistically significant difference between treatment arms in the risk of recurrent deterioration events for any PRO domain). Similar to the PFS model, worsening changes from baseline scores in all of the symptom scales were associated with an increased risk of recurrent symptomatic deterioration events (HRs, 1.11-1.33; reflecting an 11%-33% increased risk for each 10-point increase in PRO score), whereas improved changes from baseline scores in GHS/QoL, both functioning scales, and the EQ-VAS were linked to a reduced risk of recurrent symptomatic deterioration events (HRs, 0.91-0.84; reflecting a 9%-16% reduced risk for each 10-point increase in PRO score; Table 4).
In the DoR clinical-event models, after adjusting for recurrent symptomatic deterioration, change from baseline in corresponding symptoms, and stratification factors, zanubrutinib was associated with a statistically significant reduction in the risk of shortened DoR compared with ibrutinib. In particular, HRs for zanubrutinib compared with ibrutinib ranged from 0.69 to 0.49, reflecting a 31% to 51% reduced risk of shortened DoR (Table 4).
The recurrent symptomatic deterioration frailty prediction of DoR was statistically significant for the physical functioning and EQ-VAS domains, irrespective of treatment (Table 4). The frailty prediction exhibited acceptable convergence for both domains ( = 1.034 and 1.032, respectively).
Kaplan-Meier plots
Kaplan-Meier plots were used to show time to disease progression and DoR in the PRO-evaluable population. These analyses, using log-rank tests, showed statistically significant improvements in both PFS and DoR with zanubrutinib treatment across all PRO domains. The patients included in the EORTC QLQ-C30 fatigue domain analyses for PFS and DoR, respectively, are represented in Figures 2 and 3 (also shown in supplemental Table 2).
Figure 2.

Kaplan-Meier plot for time to disease progression. The dashed lines represent the median time to event for each treatment arm.
Figure 3.

Kaplan-Meier plot for time to DoR. The dashed lines represent the median time to event for each treatment arm.
Discussion
These analyses demonstrate that patient-reported symptoms are important for predicting clinical outcomes. Coupled with this, it is also important to consider functional scales, especially physical functioning, that are associated with disease- and treatment-related symptoms. The analyses showed that, regardless of the treatment, recurrent deterioration in PRO disease-related and treatment-related symptoms was significantly associated with worsening of PFS and shortened DoR. However, treatment with zanubrutinib was not associated with deterioration in patient-reported symptoms, and zanubrutinib-treated patients had significantly better PFS and longer DoR than their ibrutinib-treated counterparts. These findings both support the prognostic value of PROs in assessing treatment efficacy and also reinforce the clinical benefits of zanubrutinib beyond traditional survival metrics. To clarify, worsening patient-reported symptoms do not indicate disease progression. Rather, this research suggests that these patient-reported symptoms are predictive and antecedents of progression and loss of response (PFS and DoR events).
In the Cox proportional hazards model component, the association between recurrent deterioration events and longitudinal parameters was significant for all patient-reported symptoms. Worsening patient-reported symptoms may reflect underlying disease activity and were associated with a higher hazard of progression, suggesting that symptom trajectories can serve as an early warning signal for impending disease progression. Although traditional PRO analytic methods provide essential insights into PROs by assessing treatment effects over time and providing group-level differences, these analyses capture the dynamic relationship between patient-reported symptom trajectories and disease progression. These findings highlight the relevance of PROs in assessing disease progression. In particular, the observed associations between recurrent symptomatic deterioration and clinical events suggest worsening in patient physical functioning and may serve as an early indicator of disease progression. From a practical standpoint, monitoring these patient-reported symptoms could enable earlier intervention, potentially guiding treatment adjustments before disease progression becomes clinically apparent.
The strengths of this study include the use of high-quality data from the ALPINE trial to examine the application of a multicomponent model in this specific patient population. Previously published PRO data from this trial have demonstrated robust results in patients on zanubrutinib, thereby enhancing the reliability of this post hoc analysis.7 The current model approach accounts for the association between the longitudinal and time-to-event data, leading to more efficient estimation of the treatment effect on both time-to-event and longitudinal outcomes. This approach also provides a comprehensive adjustment for missing data, particularly in later treatment cycles when patient dropout owing to progression or death can bias traditional analyses.
Although this study provides support for modeling PRO data using an innovative modeling approach, it has some limitations. Patients might terminate participation or be withdrawn by an investigator due to AEs, protocol violations, or other factors that are captured in patient-level data. Therefore, adding relevant information as either competing risks or data-truncation markers would improve the adjustment for informative data loss resulting from such study departures. Finally, assessing the severity and frequency of treatment AEs was not an objective of the current analysis. AEs should be included in future analyses, especially those directly attributable to treatment.
After predicting PFS and DoR from the risk of recurrent symptomatic deterioration events using survival analyses adjusted for baseline stratification factors and change from baseline in corresponding symptoms, zanubrutinib remained more effective than ibrutinib with regard to disease progression in patients with R/R CLL/SLL from the ALPINE trial. Recurrent, patient-reported symptomatic deteriorations in role functioning and physical functioning were identified as leading predictors of PFS events and shortened DoR, respectively. These post hoc analyses provide a comprehensive framework for modeling PRO data and identify previously unreported associations between the risk of recurrent symptomatic deterioration and the risk of disease-progression events. Regulatory bodies are placing increasing emphasis on patient-reported symptomatology, and these results provide a strong foundation for integrating PRO data into future drug evaluations. The findings support differentiation of zanubrutinib from ibrutinib by highlighting its lower patient-reported symptom burden alongside its efficacy advantages and set a precedent for how joint model can be leveraged in future clinical trials to highlight the full spectrum of a drug’s benefits.
Conflict-of-interest disclosure: T.V. reports receiving all support for this study from BeOne Medicines Ltd. G.B. is an employee of BeOne Medicines Ltd and holds stock in the company. M.S. reports receiving all support for this study and research funding from BeOne Medicines Ltd; received research funding from Mustang Bio, Genentech, AbbVie, AstraZeneca, Genmab, MorphoSys/Incyte, and Vertex; consulting fees from AbbVie, Genentech, AstraZeneca, Genmab, Janssen, BeOne Medicines Ltd, Bristol Myers Squibb, MorphoSys/Incyte, Kite Pharma, Eli Lilly, Fate Therapeutics, Nurix, Merck, Pfizer, and Pierre Fabre; holds stock in Koi Biotherapeutics; and has an immediate family member employed by Bristol Myers Squibb. C.S.T. has received honoraria from BeiGene, Janssen, AbbVie, AstraZeneca, and Gilead; and research funding from BeiGene, Janssen, and AbbVie. L.Q. has served in a consulting or advisory role for BeOne Medicines, Sanofi, Eli Lilly, and Johnson & Johnson; and has participated in speakers’ bureaus for BeOne Medicines, Pfizer, Johnson & Johnson, Sanofi, and Roche. T.S. and F.B.B. are employees of BeOne Medicines and hold stock in the company. R.K. is a BeOne employee and holds stock. J.R.B. has received research funding paid to her institution by BeOne Medicines Ltd, Gilead, iOnctura, Loxo/Eli Lilly, MEI Pharma, Nagoon Therapeutics, and TG Therapeutics; has received royalties from UpToDate; has served as a consultant for AbbVie, Acerta/AstraZeneca, Alloplex Biotherapeutics, BeiGene Medicines, Bristol Myers Squibb, EcoR1 Capital, Galapagos NV, Genentech/Roche, Grifols Worldwide Operations, InnoCare Pharma Inc, Kite Pharma, Loxo/Eli Lilly, Magnet Biomedicine, Merck, Numab Therapeutics, Nurix Therapeutics, Pfizer, and Pharmacyclics; and served on the data safety monitoring board for Grifols Therapeutics. N.L. declares no competing financial interests.
Acknowledgments
Medical writing support, under the direction of the authors, was provided by Jason Allaire of Generativity Solutions Group and was funded by BeOne Medicines USA, Inc. Editorial support was provided by Allison Michaelis of Twist Medical and was funded by BeOne Medicines USA, Inc.
This study was funded by BeOne Medicines USA, Inc.
Authorship
Contribution: T.V., G.B., F.B.B., and R.K. were responsible for developing the concept and design for patient-reported outcome joint model, extracting and analyzing the data, interpreting the results, writing the manuscript, and providing critical feedback on the manuscript; C.S.T., L.Q., T.S., and J.R.B. were responsible for developing the ALPINE trial protocol and collecting data for the ALPINE trial, interpreting results, writing the manuscript, and providing feedback; M.S. and N.L. were responsible for collecting data for the ALPINE trial, interpreting results, writing the manuscript, and providing feedback; T.S. was responsible for developing the ALPINE trial protocol and conducting the ALPINE trial, interpreting results, writing the manuscript, and providing feedback; and all authors were involved in reviewing/revising the manuscript, provided approval for submission, and were accountable for all aspects of the work.
Footnotes
On request, and subject to certain criteria, conditions, and exceptions, BeOne Medicines, Ltd. will provide access to individual deidentified participant data from applicable BeOne Medicines-sponsored studies. BeOne Medicines shares data only when permitted by applicable data privacy and security laws and regulations, shares when it is feasible to do so without compromising the privacy of the study participants, and other considerations. Data requests may be submitted to ClinicalTrials@beonemed.com.
The full-text version of this article contains a data supplement.
Supplementary Material
References
- 1.Eichhorst B, Robak T, Montserrat E, et al. Chronic lymphocytic leukaemia: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2021;32(1):23–33. doi: 10.1016/j.annonc.2020.09.019. [DOI] [PubMed] [Google Scholar]
- 2.Waweru C, Kaur S, Sharma S, Mishra N. Health-related quality of life and economic burden of chronic lymphocytic leukemia in the era of novel targeted agents. Curr Med Res Opin. 2020;36(9):1481–1495. doi: 10.1080/03007995.2020.1784120. [DOI] [PubMed] [Google Scholar]
- 3.Holtzer-Goor KM, Schaafsma MR, Joosten P, et al. Quality of life of patients with chronic lymphocytic leukaemia in the Netherlands: results of a longitudinal multicentre study. Qual Life Res. 2015;24(12):2895–2906. doi: 10.1007/s11136-015-1039-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Shanafelt TD, Bowen D, Venkat C, et al. Quality of life in chronic lymphocytic leukemia: an international survey of 1482 patients. Br J Haematol. 2007;139(2):255–264. doi: 10.1111/j.1365-2141.2007.06791.x. [DOI] [PubMed] [Google Scholar]
- 5.Brown JR, Eichhorst B, Hillmen P, et al. Zanubrutinib or ibrutinib in relapsed or refractory chronic lymphocytic leukemia. N Engl J Med. 2023;388(4):319–332. doi: 10.1056/NEJMoa2211582. [DOI] [PubMed] [Google Scholar]
- 6.Brown JR, Eichhorst B, Lamanna N, et al. Sustained benefit of zanubrutinib vs ibrutinib in patients with R/R CLL/SLL: final comparative analysis of ALPINE. Blood. 2024;144(26):2706–2717. doi: 10.1182/blood.2024024667. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Tam CS, Lamanna N, O'Brien SM, et al. Health-related quality of life outcomes associated with zanubrutinib versus ibrutinib monotherapy in patients with relapsed/refractory chronic lymphocytic leukemia and small lymphocytic lymphoma: results from the ALPINE Trial. Curr Med Res Opin. 2023;39(11):1497–1503. doi: 10.1080/03007995.2023.2262378. [DOI] [PubMed] [Google Scholar]
- 8.Friis RB, Hjøllund NH, Pappot H, Taarnhøj GA, Vestergaard JM, Skuladottir H. Patient-reported outcome measures used in routine care predict for survival at disease progression in patients with advanced lung cancer. Clin Lung Cancer. 2021;22(2):e169–e179. doi: 10.1016/j.cllc.2020.09.014. [DOI] [PubMed] [Google Scholar]
- 9.Efficace F, Collins GS, Cottone F, et al. Patient-reported outcomes as independent prognostic factors for survival in oncology: systematic review and meta-analysis. Value Health. 2021;24(2):250–267. doi: 10.1016/j.jval.2020.10.017. [DOI] [PubMed] [Google Scholar]
- 10.Fang FM, Liu YT, Tang Y, Wang CJ, Ko SF. Quality of life as a survival predictor for patients with advanced head and neck carcinoma treated with radiotherapy. Cancer. 2004;100(2):425–432. doi: 10.1002/cncr.20010. [DOI] [PubMed] [Google Scholar]
- 11.Earlam S, Glover C, Fordy C, Burke D, Allen-Mersh TG. Relation between tumor size, quality of life, and survival in patients with colorectal liver metastases. J Clin Oncol. 1996;14(1):171–175. doi: 10.1200/JCO.1996.14.1.171. [DOI] [PubMed] [Google Scholar]
- 12.Gotay CC, Kawamoto CT, Bottomley A, Efficace F. The prognostic significance of patient-reported outcomes in cancer clinical trials. J Clin Oncol. 2008;26(8):1355–1363. doi: 10.1200/JCO.2007.13.3439. [DOI] [PubMed] [Google Scholar]
- 13.Yan JT, Bel C, Trask PC, Lo E. Are changes in patient-reported outcomes prognostic for diffuse large B-cell lymphoma survival? Results from the GOYA trial. J Patient Rep Outcomes. 2024;8(1):31. doi: 10.1186/s41687-024-00708-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) adopts consolidated guideline on good clinical practice in the conduct of clinical trials on medicinal products for human use. Int Dig Health Legis. 1997;48(2):231–234. [PubMed] [Google Scholar]
- 15.Aaronson NK, Ahmedzai S, Bergman B, et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst. 1993;85(5):365–376. doi: 10.1093/jnci/85.5.365. [DOI] [PubMed] [Google Scholar]
- 16.Herdman M, Gudex C, Lloyd A, et al. Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L) Qual Life Res. 2011;20(10):1727–1736. doi: 10.1007/s11136-011-9903-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Osoba D, Rodrigues G, Myles J, Zee B, Pater J. Interpreting the significance of changes in health-related quality-of-life scores. J Clin Oncol. 1998;16(1):139–144. doi: 10.1200/JCO.1998.16.1.139. [DOI] [PubMed] [Google Scholar]
- 18.Balan TA, Putter H. A tutorial on frailty models. Stat Methods Med Res. 2020;29(11):3424–3454. doi: 10.1177/0962280220921889. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
