Real-world survival outcomes in first-line ibrutinib-treated patients with high-risk CLL/SLL Open Access

Chronic lymphocytic leukemia (CLL)/small lymphocytic lymphoma (SLL) is the most common type of blood cancer among older adults in the United States, with a median age at diagnosis of 70 years.^1,2 In 2021, >200 000 people were estimated to be living with CLL/SLL in the United States, with >20 000 new cases estimated in 2024.^2,3 Although the 5-year relative overall survival (OS) of CLL/SLL is 88.5%, there is a large variability in survival across patient subgroups, ranging from a few years from diagnosis to normal life expectancy, owing in part to cytogenetic features.^1,3 Approximately 80% of patients with CLL/SLL have a karyotypic abnormality and, particularly among those with del(17p), del(11q), or unmutated immunoglobulin variable heavy chain (IGHV), there is a higher risk of poor response to therapy and relapse. Furthermore, patients with ≥3 aberrations (ie, complex karyotype) have significantly shorter OS owing to higher genomic instability, highly aggressive disease subsets, and reduced response to treatment, leading to poorer clinical outcomes in patients with CLL/SLL.^4-7 Aberrations in TP53 are also of interest in the context of high-risk CLL/SLL disease.⁸ The TP53 gene is located on the short arm of chromosome 17(p), and del(17p) typically results in the loss of the TP53 gene, leading to increased genomic instability through the loss of the tumor-suppressor protein p53.⁸ 

With the approval of novel targeted therapies such as Bruton tyrosine kinase inhibitors (BTKis), which have replaced chemoimmunotherapy as the standard of care, the CLL/SLL treatment landscape has changed dramatically over the last decade.^9,10 Ibrutinib was the first BTKi to receive approval from the US Food and Drug Administration for the treatment of CLL/SLL in 2014. It was subsequently approved in 2016 for first-line (1L) use followed by acalabrutinib in 2019 and most recently, zanubrutinib in 2023.^11-13 

OS is the gold standard outcome in cancer research¹⁴ because it effectively assesses both efficacy and safety, whereas progression-free survival (PFS) focuses mainly on tumor burden and has been shown to be a suboptimal predictor of OS.^15,16 Previous clinical trials have established ibrutinib as a standard of care for patients with CLL/SLL in the 1L setting, demonstrating OS benefits over up to 10 years of follow-up, with an increased depth of response over time.^17-22 In a pooled post hoc analysis of clinical trial data collected over a 12-year period, OS was comparable for patients with CLL/SLL who were treated with 1L ibrutinib (82%) and the age-matched general population (80%).²³ Real-world studies have also confirmed the benefit of 1L ibrutinib treatment; specifically, real-world OS (rwOS) rates were reported to be 89% at 30 months after 1L ibrutinib treatment initiation.²⁰ In addition, 1L ibrutinib has been shown to be an effective treatment option for patients with high-risk cytogenetics, such as those with del(17p), del(11q), and unmutated IGHV, improving long-term clinical outcomes in both clinical trials and real-world studies.^19,24,25 However, to our knowledge, no studies to date have formally compared rwOS between patients with and without high-risk CLL/SLL treated with 1L ibrutinib in real-world clinical practice.

The objective of this study was to describe the characteristics of patients with and without high-risk CLL/SLL (ie, del(17p), del(11q), or unmutated IGHV) treated with single-agent ibrutinib in 1L therapy and compare rwOS between both cohorts. Considering the evolving definition of high-risk CLL/SLL,²⁶ a sensitivity analysis with an alternative definition for high-risk CLL/SLL (ie, excluding del(11q) from the high-risk definition) was conducted. In addition, because CLL/SLL is primarily prevalent in older populations who have an increased comorbidity burden,²⁷ mortality risk, and susceptibility to adverse events,²⁸ a sensitivity analysis was conducted in a subgroup of patients with Medicare coverage (ie, aged ≥65 years), which can inform the management of older patients with CLL/SLL in a US and non-US setting.

Deidentified data were extracted from the nationwide Flatiron Health database from 1 January 2011 to 31 January 2023. This electronic health record (EHR)–derived database is a longitudinal database comprising deidentified patient-level structured and unstructured data originating from ∼280 cancer clinics (∼800 sites of care) during the study period and were curated using technology-enabled abstraction.^29,30 The registry includes a comprehensive view of patient clinical characteristics, including laboratory test results and vital sign data, oncologist-defined rule-based lines of therapy (LOTs), cytogenetic and molecular testing results, and Eastern Cooperative Oncology Group performance status (ECOG PS) scores. The data are deidentified and subject to obligations to prevent reidentification and protect patient confidentiality. Institutional review board approval of the study protocol was obtained before study conduct and included a waiver of informed consent.

To be included in the registry, patients met the following criteria: CLL/SLL diagnosis confirmed at any time; ≥2 visits at a clinic contributing data to Flatiron Health on, or after, 1 January 2011; and ≥1 order of an antineoplastic agent on, or after, 1 January 2011.

Adult patients with or without high-risk CLL/SLL were included in this retrospective cohort study if they had initiated 1L ibrutinib on, or after, 4 March 2016 (ie, date of ibrutinib US Food and Drug Administration approval for 1L CLL/SLL treatment) with confirmation of single-agent ibrutinib use within 28 days of 1L initiation (ie, LOT identification window). The index date was defined as the date of 1L ibrutinib initiation plus 28 days. The baseline period was defined as the 12 months before 1L ibrutinib initiation (Figure 1). Exclusion criteria included patients who started 1L therapy >14 days before their CLL/SLL diagnosis date; were involved in clinical trials on, or before, 1L treatment; were diagnosed with other cancers in the 12 months before 1L treatment; or who had no information on high-risk status. Those who received an add-on treatment within 28 days or died within 28 days of 1L treatment initiation were also excluded. Patients were considered to have high-risk CLL/SLL if they had del(17p), del(11q), or unmutated IGHV genomic alterations. TP53 alteration status, another potentially important prognostic indicator⁸ was not available in the data source and was therefore not included in the definition. Those with an absence of all 3 abnormalities were considered to have non–high-risk CLL/SLL. The follow-up period spanned from the index date until the last date of clinical activity, death, or the end of data availability, whichever occurred first.

Any systemic treatment (ie, an order for or administration of an antineoplastic agent recorded in the EHR) was considered a LOT, excluding corticosteroids, radiotherapy, surgery, and megestrol. Patients were considered to have received 1L ibrutinib therapy if they had a record of ibrutinib initiation after CLL/SLL diagnosis. Patients were considered to have received single-agent ibrutinib if they were not prescribed any other antineoplastic agent within 28 days of initiating ibrutinib (ie, during the LOT identification window).

Baseline characteristics were assessed during the 12-month period before 1L ibrutinib initiation, except for cytogenetic risk stratification, which could be assessed at any time before. rwOS, defined as the time between the index date and all-cause mortality, was the primary outcome measure. Mortality data in the registry were obtained from various sources, including EHR, external commercial data sources that mine data from obituaries, funeral homes, and the US National Social Security Index. Mortality data were validated by Flatiron Health with high sensitivity, specificity, and accuracy.³¹ Patients without mortality information were censored at the date of their last confirmed activity.

Baseline characteristics were described using proportions for categorical variables and means and standard deviations (SDs) or medians and interquartile ranges (IQRs) for continuous variables.

rwOS was analyzed as a time-to-event outcome using Kaplan-Meier survival curves and Cox proportional hazards models, with the non–high-risk cohort as the reference cohort for all analyses. Results were reported using hazard ratios (HR) and corresponding 95% confidence intervals (CIs). Inverse probability of treatment weighting (IPTW) was used to account for potential imbalances between cohorts for the following baseline covariates: sex, age, race, index year, region, disease type, stage, ECOG PS scores, time from diagnosis to index treatment initiation, Quan–Charlson Comorbidity Index (CCI), practice type, del(13q), and trisomy 12. Sensitivity analyses with an alternative high-risk CLL/SLL definition based on the presence of either del(17p) or unmutated IGHV only (ie, excluding del(11q) from the definition), and a subgroup analysis for Medicare beneficiaries (based on insurance information provided by Flatiron Health) were conducted using IPTW to account for potential imbalances, as described earlier.

This article is based on previously collected data from a commercially available database and does not contain any studies of human participants or animals performed by any of the authors. Because data were deidentified and comply with the Health Insurance Portability and Accountability Act, no review by an institutional review board was required per Title 45 of the Code of Federal Regulations, Part 46.101(b)(4).

A total of 16 387 patients had a diagnosis of CLL/SLL in Flatiron Health during the study period (Figure 2). Among 9850 adults initiating 1L therapy and meeting other study eligibility criteria, 2531 were treated with 1L single-agent ibrutinib on, or after, 4 March 2016. Of these patients, 1242 had known cytogenetic risk, with 969 considered high risk and 273 non–high risk.

Before applying IPTW, the mean age of patients with high- and non–high-risk CLL/SLL was 70.0 (SD, 10.2) and 70.8 (SD, 9.0) years (Table 1). Most patients were White (White patients in the high-risk cohort, 72.0%; White patients in the non–high-risk cohort, 74.7%), male (high-risk, 63.8%; non–high-risk, 56.4%), and treated in a community practice setting (high-risk, 91.6%; non–high-risk, 93.8%). Among those with high- and non–high-risk disease, the mean Quan-CCI scores were 0.2 (SD, 0.6) and 0.1 (SD, 0.4). The proportion of patients with an ECOG PS score of 0 to 1 was 68.3% among those with high-risk CLL/SLL and 72.9% among those with non–high-risk CLL/SLL. Rai stage 0 was observed in 26.8% of patients with high-risk CLL/SLL and 33.7% in those with non–high-risk CLL/SLL; 15.0% had Rai stage I in both cohorts. Rai stage II was observed in 7.0% of patients with high-risk CLL/SLL and 4.4% of patients with non–high-risk CLL/SLL. Rai stages III and IV were observed in similar proportions across the cohorts (stage III [high-risk cohort, 7.4%; non–high-risk cohort, 6.2%]; and stage IV [high-risk cohort, 7.7%; non–high-risk cohort, 6.2%]). The mean time from first observed diagnosis to 1L ibrutinib initiation was 35.0 (SD, 48.0) and 64.4 (SD, 73.2) months for patients with high- and non–high-risk CLL/SLL, respectively. Statistically significant differences were found for sex and time from diagnosis to 1L ibrutinib initiation (P = .03 and P < .0001, respectively).

Medicare beneficiaries accounted for 37.5% and 39.6% of the high- and non–high-risk cohorts, respectively. Among Medicare beneficiaries, the mean age was 74.4 (SD, 6.8) and 74.2 (SD, 6.4) years in the high- and non–high-risk cohorts, respectively. Most patients included in the Medicare subgroup were White (high-risk cohort, 73.3%; non–high-risk cohort, 80.6%), male (high-risk cohort, 62.0%; non–high-risk cohort, 56.5%), and treated in a community practice setting (high-risk cohort, 91.7%; non–high-risk cohort, 96.3%). Among Medicare beneficiaries with high- and non–high-risk disease, the mean Quan-CCI score was 0.2 (SD, 0.6) and 0.1 (SD, 0.3), respectively. The proportion of patients with an ECOG PS score of 0 to 1 was 68.0% among those with high-risk CLL/SLL and 75.0% among those with non–high-risk disease. Rai stage 0 was observed among 29.2% of patients with high-risk CLL/SLL and 30.6% among those with non–high-risk CLL/SLL; 14.0% and 17.6%, respectively, had Rai stage I disease. Rai stage II was observed in 5.8% of patients with high-risk CLL/SLL and 1.9% of patients with non–high-risk CLL/SLL. Rai stages III and IV were observed in slightly lower proportions in the high-risk cohort relative to the non–high-risk cohort (stage III [high-risk cohort, 7.7%; non–high-risk cohort, 9.3%]; and stage IV [high-risk cohort, 6.3%; non–high-risk cohort, 9.3%]). The mean time from first observed diagnosis to 1L ibrutinib initiation was 38.0 (SD, 45.3) months among patients with high-risk CLL/SLL and 68.3 (SD, 72.7) months among those with non–high-risk CLL/SLL. No statistically significant differences were found between patients with high- and non–high-risk CLL/SLL in the Medicare subgroup.

Among patients with high-risk disease, 319 (32.9%) had del(17p), 356 (36.7%) had del(11q), and 569 (58.7%) had unmutated IGHV (Table 2). Approximately half of the patients (47.0%) with high-risk disease had del(13q) compared with 63.0% among those with non–high-risk CLL/SLL (P < .0001). The prevalence of trisomy 12 was comparable between patients with high- and non–high-risk CLL/SLL at 23.5% and 24.2%, respectively (P = .82; Table 2). In the Medicare subgroup, of patients with high-risk CLL/SLL, 132 (36.4%) had del(17p), 133 (36.6%) had del(11q), and 203 (55.9%) had unmutated IGHV. Approximately half (44.4%) of patients with high-risk disease had del(13q) compared with 59.3% of those with non–high-risk disease (P = .007); the prevalence of trisomy 12 was comparable between patients with high- and non–high-risk CLL/SLL at 25.1% and 26.9%, respectively.

The median follow-up (from index date to the last date of clinical activity, death, or the end of data availability) was 32 months (IQR, 17-48) in the high-risk cohort and 31 months (IQR, 19-44) in the non–high-risk cohort. The median time on ibrutinib treatment was 18.9 and 14.3 months in the high- and non–high-risk cohorts, respectively. After applying IPTW, median rwOS was not reached in either cohort (Figure 3). Kaplan-Meier curves were similar for both cohorts, and the Cox proportional hazard model found no significant difference in rwOS between patients with high- and non–high-risk CLL/SLL (HR, 1.09; 95% CI, 0.79-1.51; P = .60). In the sensitivity analysis using an alternative high-risk definition (presence of del(17p) or unmutated IGHV only), results were similar, with a HR of 1.19 (95% CI, 0.86-1.64; P = .30) and median rwOS not reached in either cohort (Figure 4). Within the Medicare subgroup, the HR was 0.98 (95% CI, 0.63-1.53; P = .94), and median rwOS was also not reached in either cohort (Figure 4).

An understanding of real-world treatment outcomes is essential when choosing optimal 1L therapy for patients with CLL/SLL. Using rwOS as an outcome reflecting both efficacy and safety, this study addresses the need for real-world treatment outcomes data in patients with cytogenetic high- and non–high-risk CLL/SLL, using a large community health care data set. Results show that patients with high-risk CLL/SLL treated with 1L single-agent ibrutinib had similar rwOS compared with patients with non–high-risk disease. Similar findings were observed in sensitivity analyses using an alternative definition for high-risk CLL/SLL based on the presence of del(17p) or unmutated IGHV only, and in a subgroup of Medicare beneficiaries, further bolstering the generalizability of study findings.

These data demonstrating the rwOS benefit of 1L ibrutinib treatment in patients with CLL/SLL align with previously reported findings from clinical trials and real-world studies.^{17-20,25,32-35} Specifically, long-term data from the RESONATE-2 clinical trial, with 10 years of follow-up, show sustained OS benefit for single-agent ibrutinib in 1L therapy for patients with CLL/SLL.²² In retrospective analyses of patients with CLL/SLL receiving 1L ibrutinib in real-world practices, the treatment was found to be associated with better clinical outcomes than alternative real-world 1L regimens used for the treatment of CLL/SLL.^25,35 Importantly, a deep and durable response to ibrutinib and sustained PFS is also observed among patients with high-risk CLL/SLL.^{18,24,25,36-38} By integrating results from phase 3 clinical trials with up to 6.5 years of follow-up, Burger et al showed that ibrutinib treatment in the 1L was associated with a 5-year OS of 84% among patients with high-risk CLL.²⁴ Similarly, real-world studies have shown that 1L single-agent ibrutinib treatment was associated with a longer time to next treatment than chemotherapy, as well as a rwOS benefit among patients with high-risk disease (which were no different relative to those with non–high-risk features), consistent with the results of this study.^37,38 

This study adds to the literature by providing a dedicated comparison of rwOS among patients with high- and non–high-risk CLL/SLL receiving 1L single-agent ibrutinib in clinical practice, using a large community health care data set, including a sensitivity analysis that considered an alternative definition for high-risk CLL/SLL. The definition for high-risk CLL/SLL continues to evolve, and understanding the therapeutic implications of high-risk molecular features warrants ongoing investigation. Although cytogenetic abnormality del(11q) is 1 of the most frequently occurring chromosomal deletions in CLL/SLL, detected in ∼20% of patients at diagnosis, its importance as a prognostic factor remains unclear.³⁹ Although previous studies have shown that patients with del(11q) had a worse prognosis even on chemotherapy,^32,39 recent evidence has found limited prognostic value for del(11q) on rwOS among patients with CLL/SLL.⁶ Indeed, using pooled data from 3 phase 3 trials, a study by Kipps et al showed that the presence of del(11q) in patients with CLL/SLL did not affect outcomes with single-agent ibrutinib.⁴⁰ Moreover, Kipps et al reported that patients with del(11q) had a significantly better OS response to single-agent ibrutinib treatment than those without del(11q).⁴⁰ In the recent phase 2 CAPTIVATE trial, which evaluated a 1L ibrutinib-based therapy for patients with high-risk CLL/SLL,³⁶ del(11q) was not included in the definition for high-risk CLL/SLL, which guided the alternative definition considered in the current study’s sensitivity analysis. The results of the sensitivity analysis confirm that the rwOS benefit of single-agent ibrutinib treatment in the 1L for patients with CLL/SLL was unchanged when patients with del(11q) were excluded from the high-risk cohort.

The second sensitivity analysis focused on the subgroup of Medicare beneficiaries. Considering that the median age at CLL/SLL diagnosis is 70 years, and with the increasing incidence of CLL/SLL in the United States together with the continually aging population, older patients represent a population of particular interest for research, owing to their increased mortality risk and susceptibility to adverse events.^41-43 Clinical trial data previously demonstrated the efficacy of ibrutinib single-agent as 1L treatment for patients with CLL/SLL who were aged ≥65 years.^18,21 The findings of this study in the Medicare subgroup confirm the value of single-agent 1L ibrutinib treatment for older patients (age of ≥65 years) with CLL/SLL in clinical practice, regardless of their risk status.

Although BTKis share similarities, including cardiovascular adverse events as a class-specific BTKi effect, there is little comparative evidence to support a better clinical response among high-risk patient groups treated with a second-generation BTKi, such as acalabrutinib or zanubrutinib, given that ibrutinib is the only treatment with sufficient long-term data to show an OS benefit.^9,22,44-46 For ibrutinib specifically, the OS for patients with CLL/SLL treated with 1L ibrutinib is comparable with that of an age-matched cohort from the general population based on a pooled analysis of 3 phase 3 clinical studies (RESONATE-2, ECOG1912, and iLLUMINATE).²³ Further research is warranted to understand real-world treatment outcomes among patients with high- and non–high-risk CLL/SLL treated with other BTKis.

The results of this study should be interpreted considering some limitations. Because Flatiron Health collects information from real-world clinical oncology practices, the data are subject to inherent miscoding and errors. As such, data on some important summary prognostic factors (eg, comorbidities, ECOG PS scores, and Rai stage) may have been missing for some patients in the database. As this is expected to equally affect all patient records, the impact on study findings and interpretation is limited. Missing data in baseline variables were reported as a subcategory in the analyses, and patients with missing data for del(17p), del(11q), or IGHV were excluded from analyses. Exploratory analyses imputing missing values in baseline variables were conducted, with the results of the study remaining unchanged. In addition, although the objective was to study patients treated with single-agent ibrutinib in the 1L setting, it is possible that some patients with add-ons after the first 28 days after treatment initiation may have been included in the analysis. Information on TP53 alteration status, which may be another important prognostic indicator,⁸ was not available in Flatiron Health records and was therefore not considered in the definition for high-risk CLL/SLL in this study. In addition, information on PFS or response was not available in the Flatiron Health data used for this study; however, the reliability of rwOS data from Flatiron Health has been established previously.³¹ Considering the limitations of PFS data in the real world,^15,16 rwOS as captured in Flatiron Health data at the patient level provides an opportunity for unbiased comparisons between therapies or patient groups across diverse clinical practice settings. Furthermore, information about treatments received by patients outside of the specific cancer care sites included in Flatiron Health data may have not been captured in the structured EHR data. In addition, information about patients’ history before the adoption of EHR by clinical oncology practices may not have been available, particularly in the structured data. Finally, although IPTW was used to account for potential baseline confounding, residual confounding due to unmeasured characteristics may remain.

This real-world study of patients with CLL/SLL, selected from a large community health care database and treated with 1L ibrutinib, demonstrated that patients with high-risk cytogenetic features had similar rwOS compared with patients with non–high-risk CLL/SLL when receiving this well-established treatment option. Similar results were found in sensitivity analyses using an alternative high-risk definition and in a Medicare subgroup. These results support the use of single-agent ibrutinib as 1L treatment for patients with CLL/SLL, regardless of cytogenetic risk.

Editorial and medical writing support were provided by Cindi A. Hoover of ApotheCom and Loraine Georgy of Analysis Group, Inc, and funded by Janssen Biotech, Inc, administered by Janssen Scientific Affairs, LLC.

This study was sponsored by Janssen Scientific Affairs, LLC, a Johnson & Johnson company.

Contribution: J.N.A. and S.M.O. contributed to study conceptualization and design, interpretation of results, and drafting and revision of the manuscript; and T.R., Z.D., J.H., A.B., and Z.P.Q. contributed to study conception and design, acquisition of data, data analysis, interpretation of results, and drafting and revision of the manuscript; all authors met the International Committee of Medical Journal Editors criteria for authorship for this article, took responsibility for the integrity of the work as a whole, and approved the final version for publication.

Conflict-of-interest disclosure: J.N.A. reports a consulting or advisory role with AbbVie, Adaptive Biotechnologies, ADC Therapeutics, AstraZeneca, BeiGene, Epizyme, Genentech, Janssen, Eli Lilly and Company, Merck, NeoGenomics, Pharmacyclics, and TG Therapeutics; reports research funding from BeiGene, Celgene, Genentech, Janssen, and TG Therapeutics; and reports a speakers bureau role for AbbVie, BeiGene, Janssen, and Pharmacyclics. T.R., Z.D., J.H., A.B., and Z.P.Q. are employees of Janssen Scientific Affairs, LLC, and stockholders of Johnson & Johnson. S.M.O. reports a consulting or advisory role with AbbVie, Alexion, Amgen, Aptose Biosciences Inc, Astellas, AstraZeneca, Autolus, Bristol Myers Squibb, Celgene, DynaMed, Eli Lilly and Company, Gilead, GlaxoSmithKline, Janssen Oncology, Johnson & Johnson, Juno Therapeutics, MEI Pharma Inc, Merck, NOVA Research, Pfizer, Pharmacyclics, TG Therapeutics, Vaniam Group LLC, Verastem Oncology, and Vida Ventures; and reports research funding from Acerta, Alliance, BeiGene, Caribou Biosciences Inc, Gilead, Kite, Loxo Oncology Inc, Mustang, Nurix Therapeutics Inc, Pfizer, Pharmacyclics, Regeneron, and TG Therapeutics.

Correspondence: Zaina P. Qureshi, Janssen Scientific Affairs, LLC, a Johnson & Johnson company, 800 Ridgeview Dr, Horsham, PA 19044; email: zquresh3@its.jnj.com.