Characterization of real‐world treatment practices and outcomes among patients with chronic lymphocytic leukemia treated in a Finnish tertiary center

Abstract Objectives We conducted this retrospective study to characterize the change in chronic lymphocytic leukemia (CLL) treatment patterns between 2005 and 2019, to understand the treatment sequencing across the course of the disease, and to investigate how targeted agents and prognostic testing were implemented into the patient care. Methods This study included adult patients with CLL treated at the Hospital District of Southwest Finland during the study period. Data were collected from the Turku University Hospital data lake. Results In total, 122 and 60 patients received first‐ and second‐line treatments for CLL, respectively. The shift from conventional chemoimmunotherapy to targeted treatments in recent years (2014–2019) was observed. The median overall survival times were not reached in patients treated with targeted agents compared to conventional standard treatments in first‐ and second‐line settings and improved toward the end of the study period. Prognostic testing increased during the study follow‐up and patients with unmutated immunoglobulin heavy‐chain variable showed significantly poorer overall survival and time‐to‐next‐treatment outcomes than patients with mutated immunoglobulin heavy‐chain variable. Conclusions This real‐world study implicated added value of targeted chemo‐free therapies as reported in randomized clinical trials, and highlighted the necessity of prognostic testing in order to improve treatment selection and patient outcomes.


INTRODUCTION
Chronic lymphocytic leukemia (CLL) is the most common adult leukemia in western countries. In Finland, 2900 individuals live with CLL currently and the age-adjusted incidence was 5.09 per 100,000 people in 2019 [1]. CLL is a malignancy of B-cells characterized by CD5positive lymphocytes accumulating in the blood and lymphoid organs [2,3]. Various chromosomal alterations have proven to be prognostically important and can guide treatment selection [2,4].
The most common prognostic markers in CLL are chromosomal alterations such as del(13q), del(11q), del(17p), trisomy 12, and TP53 mutation, as well as the mutation status of immunoglobulin heavychain variable (IGHV) genes. Traditionally, high-risk CLL is defined as having del(11q), unmutated IGHV gene, and/or TP53 aberration (TP53 mutation or del[17p]) [19]. Especially the patients harboring TP53 aberration are relatively refractory to conventional CIT [4,20]. As a consequence, CIT is currently the most beneficial in the first-line treatment of young fit patients with mutated IGHV and without a TP53 aberration [9,21]. As the effectiveness of targeted therapies is not substantially dependent on conventional prognostic variables (e.g., disease state, age, and general fitness), the characterization of the disease's molecular biology has become increasingly important. According to the International Workshop on Chronic Lymphocytic Leukemia (iwCLL) and European Society for Medical Oncology guidelines, assessment of cytogenetic abnormalities and testing for TP53 mutations and IGHV status should be applied in routine clinical practice to guide the treatment selection alongside the patient's clinical stage and symptoms [4,21]. Over time, Finnish treatment guidelines have aligned with international recommendations [4,8]. In addition, Finnish treatment practices reflect the national reimbursement decisions. In recent years, the reimbursement of ibrutinib, idelalisib, and venetoclax in relapsed CLL and in CLL with TP53 aberration has translated into growing numbers of patients receiving targeted therapy.
Amid the changing tides of CLL therapy, up-to-date data on routine clinical care in Finland was lacking, while these data from other Nordic countries are available [22][23][24]. Therefore, we conducted this retro-

Outcome measures
Treatment regimens were categorized as follows: fludarabine- Patients were additionally categorized based on IGHV mutational status (mutated and unmutated) and del(17p) and/or TP53 mutation (with TP53 aberration and without TP53 aberration), or unknown if the status was not tested. Patients with 17p or 11q deletion, TP53 mutation, or unmutated IGHV were categorized as high-risk and patients without as low-risk.

Statistical analyses
The descriptive findings for continuous variables were reported as medians along with lower and upper quartiles (Q1-Q3), except for age for which range (min-max) was used instead of quartiles, and categorical variables were reported as observed proportions and frequencies.
Treatment regimens at first-, second-, and third-line were stratified by the treatment initiation period and/or treatment regimen. The followup times, including all patients, not just event-free, were calculated for each subgroup from the beginning of each treatment line to the end of the follow-up (Table S1).
The OS and TTNT were assessed from the initiation of first-or second-line treatments and explored using the Kaplan-Meier estimator. The median OS (mOS) and median TTNT (mTTNT) with 95% confidence intervals were reported if reached. These values indicate the time point by which half of the patients in the group are still alive or have initiated the next treatment, respectively. The differences in survival distributions were compared using the log-rank test. OS-and TTNT-specific hazard ratios (HRs) with 95% confidence intervals (CIs) were estimated using multivariate weighted Cox proportional hazard regression models adjusted for patient characteristics (presented in Table 1) at the start of the first treatment line and the treatment regimen given within the treatment line (historical SOC therapy vs. targeted therapy groups) [25]. Missing data were described and used without imputation in the analyses. The statistical significance (P) threshold was set at 0.05. All analyses were conducted with R software version 3.6.3 [26].

Ethical considerations
The study approval was obtained from the Hospital District of Southwest Finland (T288/2019) and was performed in accordance with the declaration of Helsinki and in compliance with applicable national laws.
According to Finnish legislation, informed consent is not required for studies based on patient records.

Patient characteristics
Altogether, 122 patients initiated first-line treatment during the study period (    (Figure 1).

Treatment pattern changes over time
An overview of treatment sequencing throughout the study period (2005-2019) is shown in Figure 2 and Table S2.

First-line treatment outcomes
In   Figure 4B). Factors associated with OS and TTNT  Table 3.

Second-line treatment outcomes
In  Table S3).
The mOS was 27.6 months (95% CI, 24.2-66.9) for patients whose second-line treatment was initiated during the early period and 78.5 months (95% CI, 36.9-NR) for patients in the late period ( Figure 4C) Figure 4D).

Treatment outcomes based on high-risk disease features
In this study cohort, mOS and mTTNT were shorter in patients with unmutated IGHV in both first-and second-line treatment without treatment stratification compared to patients with mutated IGHV (Table S4) (Table S3).
In high-risk patients (17p or 11q deletion, TP53 mutation, or unmutated IGHV) and in patients with TP53 aberration (TP53 mutation or 17p deletion), the mOS in first-and second-line treatment was shorter compared to low-risk patients or patients without TP53 aberration  (Table S4). Of note, a high proportion of deaths was observed among patients with unknown prognostic test results (Table S4), which might indicate a significant number of patients in fact having high-risk disease features in this group.

DISCUSSION
CLL treatment has changed over the last decades reflecting the advances in the treatment armamentarium. At first, the addition of CD20-antibody to conventional chemotherapy improved patient outcomes, while more extensive benefits have been seen recently with novel targeted treatments becoming available [27,28]. In this study, we set out to characterize CLL treatment practices, outcomes, and the role of prognostic testing in real-life clinical practice from a well-defined geographical region in Southwest Finland between 2005 and 2019.
In general, the patients in the present study had been treated in accordance with the Finnish and European treatment guidelines prevailing during the study period [8,29]. Chlorambucil-based therapy was widely used in the first-line during the study period, which was in line with similar reports from Sweden [22]. On the other hand, the first-line use of FC/FCR was lower than reported in another Finnish study [30]. to late period, which corresponds to other reports from Nordic countries [23,30]. The overall use of FC/FCR declined during the study but was seemingly still at a high percentage in second-line treatment during the late period. However, the number of patients in the different treatment groups was small, and hence no conclusions can be drawn.
During the late period, targeted treatments were the most used regimens in the second and third lines.
The superior efficacy of targeted treatments over conventional CITs has been demonstrated in several randomized clinical trials in both untreated and relapsed diseases [9,10,12,13,[15][16][17][18]. Of note, only ibrutinib has been compared to most recommended CIT alternatives across all patient segments in previously untreated and relapsed CLL [10,12,13,17,18], and significant OS benefit has been demonstrated over FCR and chlorambucil monotherapy in the first-line treatment [12,18]. It has also been shown, that by using CIT the long-term survival is achieved only for a subset of patients, and over 50% of patients are relapsing, progressing, or deceased within five years of receiving CIT [31,32]. In this study, patients receiving any targeted treatment were grouped together and thus, no conclusions concerning individual treatment regimens can be made. The mOS was not reached for patients treated with targeted therapies in either the first or second line, and a greater proportion of patients was alive at the end of followup compared to patients on other treatments. The mOS was also longer among patients whose treatment was initiated during the late period compared to the early period, in agreement with previously published data from Finland [30]. In Swedish retrospective studies, no significant improvement in the OS was observed between 2003 and 2013 [22,23].
The mTTNT in this study was shorter in the targeted therapy group and in patients who initiated treatment during the late period, which may indicate that during recent years it has become easier to change the treatment in case of intolerance or suboptimal response as several targeted treatment options are now available. A number of treatment switches were related to the use of idelalisib, which has been associated with severe treatment-emergent adverse events [33].
Additionally, the targeted treatments were presumably distributed to patients with difficult-to-treat or rapidly progressing diseases. Due to the low number of events and patients in the targeted therapy group, statistical comparisons for OS and TTNT could not be performed.
Unmutated IGHV is known to be a negative predictor of survival [24,34]. Clinical trials show that progression-free survival benefit is achieved in patients with unmutated IGHV using targeted treatments in the first-line compared to conventional therapy [10,12,15,17,18].
IGHV testing was recommended by the iwCLL in 2018 and introduced shortly after that to the Finnish treatment guidelines in which the testing was recommended at the physician's discretion during the study period [4,29].  [31,35]. Interestingly, first-line mOS was shortest for the subset of patients whose IGHV status was unknown, suggesting that these patients may not have received optimal treatment. In order to provide patients treatment, they will benefit the most from, the IGHV mutation status should be investigated prior to the treatment initiation, as advised by the iwCLL and European Society for Medical Oncology guidelines [4,21].
In addition to IGHV mutation status, molecular cytogenetics for del(13q), del(11q), del(17p), and trisomy 12, as well as TP53 mutation, should always be examined at the baseline [4]. Testing is paramount as targeted therapies confer consistent long-term disease control and progression-free survival benefit in patients with high-risk CLL, where benefit with traditional chemotherapy or CIT is limited [4,[10][11][12]15,17,18]. In this study, the testing for cytogenetic lesions and TP53 mutation status increased in general during the study followup. However, these markers, especially TP53 that was not routinely tested until 2018, were unknown for a large subset of patients. Recent real-world data from the informCLL registry from the USA show that low rates of prognostic testing may prevent patients from getting the most appropriate therapy, and even when prognostic testing was conducted, a large subset of high-risk patients (with TP53 aberration) was treated with CIT that did not align with treatment recommendations [36]. These observations highlight the necessity to increase prognostic testing and the utilization of that knowledge to guide optimal treatment decisions.
This study provides novel information on CLL treatment practices in Finland and is the only observational study conducted in Finland in which the follow-up period extends to the era of targeted therapies. Additionally, the current study is the most comprehensive Finnish This was a single-center study, which may also affect the generalizability of the results, along with the retrospective nature of the investigation. This reflects the need for national quality registries with systemic and structured data entry to provide a coherent and costeffective way to monitor the quality and effectiveness of care and patient safety in the real-world clinical setting.

CONCLUSION
In conclusion, targeted therapies, ibrutinib, idelalisib, and venetoclax, implicated benefits for patients with CLL over conventional chemoor CIT regimens in both first-and second treatment lines, and the clinical outcomes mirror those reported in randomized clinical trials.
The favorable OS times achieved by targeted therapies in both first and second lines suggest that the targeted agents should be used as early as possible, in order to improve treatment outcomes. The results also highlight the importance of routine assessment of del(17p), TP53 mutation and IGHV mutational status as these markers serve as a treatment guiding factor and affect the treatment outcomes.

ACKNOWLEDGMENTS
Juha Mehtälä from MedEngine and Samu Kurki from Auria Biobank contributed to statistical analyses. Harlan Barker from MedEngine Oy performed the language review. We would like to thank Carola Ekstedt and Eeva Kronqvist for administrative support, Amy Leval for her excellence and dedication in driving the development of the study protocol, and Christina Wennerström for analytic insights and support. We would like to acknowledge the patients, the doctors, and the nurses whose lives and efforts made this evaluation possible. This study was funded by Janssen-Cilag Oy, Finland. Janssen identifier number PCSONC001602.

NOVELTY STATEMENT
What is the NEW aspect of your work? We here present the first retrospective registry study of real-world Finnish CLL treatment patterns describing the outcomes of targeted therapy over more historical SOC treatments.
What is the CENTRAL finding of your work? Implementation of targeted therapies suggests improved treatment outcomes among the Finnish patients with CLL in first-and second-line settings.
What is (or could be) the SPECIFIC clinical relevance of your work?
Our findings suggest that previously reported benefits of targeted therapies can also be demonstrated in real-life settings mandating their use in earlier lines of treatment.

CONFLICT OF INTEREST
Juha Ranti served as a consultant for Janssen, AbbVie, and