Summary
Background
The Brazilian Group of CLL (BGCLL) has proposed a more restrictive approach for treatment initiation compared to the International Workshop on Chronic Lymphocytic Leukemia (IWCLL) guidelines. Here, we report on the safety and efficacy of this strategy.
Methods
We performed a retrospective analysis of patients with CLL registered in the Brazilian CLL Registry between January 2009 and July 2023 who met the minimum data availability criteria for analysis. The BGCLL criteria employ stricter thresholds for cytopenias, defining them as hemoglobin levels below 9.5 g/dL or platelet counts below 50,000/mm3, as opposed to the IWCLL criteria of below 10 g/dL and below 100,000/mm3, respectively. Furthermore, the BGCLL criteria do not consider progressive lymphocytosis or disease-related symptoms to be criteria for treatment initiation when cytopenias or symptomatic masses are absent. Survival outcomes were estimated using the Kaplan–Meier method and compared with log-rank tests. Cox proportional hazards models were used for multivariable analysis, with results expressed as hazard ratios and 95% confidence intervals. A P-value <0.05 was considered statistically significant.
Findings
A total of 2511 patients were enrolled from 41 centers. Of these, 1404 patients (56%) met the IWCLL criteria for treatment initiation (liberal criteria), whereas only 788 patients (31%) met the BGCLL criteria (restrictive criteria). Patients with liberal criteria had a better overall survival than those with restrictive criteria (85% vs. 68%, respectively), suggesting that restrictive criteria were more predictive of prognosis than liberal criteria. In addition, patients treated for cytopenias had significantly worse overall survival (69%) compared to those treated for any other indication (97%, P < 0.0001). Patients with disease-related symptoms, progressive lymphocytosis, and extranodal involvement as isolated indications for treatment had similar overall survival to those with no indication for treatment.
Interpretation
Our results demonstrate that a more restrictive approach to treatment initiation in CLL, as proposed by the BGCLL, better identifies patients with higher disease burden and poorer outcomes, while sparing others unnecessary therapy.
Funding
Brazilian Registry of CLL–Brazilian Association of Hematology and Hemotherapy (ABHH)/Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES).
Keywords: Chronic lymphocytic leukemia, IWCLL guidelines, Treatment indications, Restrictive approaches, Watchful waiting, Overtreatment
Research in context.
Evidence before this study
According to the 2018 International Workshop on Chronic Lymphocytic Leukemia (IWCLL) guidelines, most patients with chronic lymphocytic leukemia (CLL) are managed with watchful waiting, as treatment is indicated only for symptomatic disease. Some patients have mild cytopenias or isolated lymphocytosis, which does not require therapy; in these cases, observation is recommended, though data on adherence and outcomes are limited. The Brazilian Group of CLL (BGCLL) has proposed a more conservative approach, but its safety and efficacy remain untested.
We searched PubMed and Google Scholar from January 2018 to June 30, 2024, using the terms “chronic lymphocytic leukemia” and “criteria for treatment initiation.” No language restrictions were applied.
Few studies have evaluated the prognostic impact of individual treatment indications specifically. A retrospective analysis by Mozas et al. found that specific IWCLL treatment triggers, such as lymphadenopathy, splenomegaly, and bone marrow failure, are linked to distinct biological disease features and clinical outcomes. These outcomes include shorter survival for patients with cytopenias due to marrow infiltration. Flowers et al. (2018) analyzed the Connect® CLL cohort and showed that patients with early-stage (Rai 0/1) CLL often begin treatment for nonspecific symptoms or disease progression. They also identified clinical predictors of treatment response. These findings underscore the variability in clinical decision-making processes and the necessity of evaluating the outcomes of more conservative management strategies.
Added value of this study
Using data from over 2500 patients in the Brazilian CLL Registry, we compared the prognostic value and safety of the BGCLL criteria to the standard IWCLL criteria. We found that patients who met only the IWCLL criteria—but not the more restrictive BGCLL criteria—had an overall survival rate similar to that of untreated patients. In contrast, those who met the BGCLL criteria had significantly worse outcomes. These results imply that the BGCLL strategy more accurately identifies patients with high-risk disease while safely deferring treatment for others.
Implications of all the available evidence
Current IWCLL guidelines may lead to overtreatment of patients with stable or low-risk features, such as progressive lymphocytosis or mild symptoms. Our study provides real-world evidence that a more restrictive, risk-adapted approach to treatment initiation is safe and informative in terms of prognosis. By limiting therapy to patients with high-burden disease, particularly those with significant cytopenias, the BGCLL criteria can avoid unnecessary exposure to treatment-related toxicity without compromising outcomes. Furthermore, adopting a more restrictive approach in low-income countries could optimize resource allocation by reducing unnecessary treatments and focusing efforts on patients with more aggressive disease. This would ultimately improve patient outcomes while minimizing healthcare costs.
Introduction
Chronic lymphocytic leukemia (CLL) is a heterogeneous disease with a wide range of clinical presentations that, over time, may or may not require treatment. While a minority of patients require immediate therapeutic intervention, most patients are managed with a watchful waiting approach for years or may never require treatment. While targeted therapies have changed the outlook for CLL management over the past decade, the criteria for initiating treatment remain unchanged since the National Cancer Institute-sponsored Working Group and the International Workshop on Chronic Lymphocytic Leukemia (IWCLL) guidelines, last updated in 2018, which are widely used in both clinical practice and research settings.1
The decision to start treatment is based on the presence of symptomatic disease. However, some patients present with one or more IWCLL criteria during the course of their disease without an actual clinical need for intervention. These include a platelet count of less than 100,000/μL or mild anemia that remains stable over time, or isolated progressive lymphocytosis with no other clinical indication.2,3 Appropriately, the IWCLL criteria state that patients with mild cytopenia may be observed without treatment. However, data on compliance with this recommendation and its impact are lacking.
CLL predominantly affects the elderly, who often have comorbidities that influence therapeutic decisions and outcomes. It is estimated that approximately 45% of CLL patients have significant comorbidities that negatively impact both treatment response and overall survival. These comorbidities often require dose adjustments and/or treatment interruptions and lead to cumulative toxicity.4,5
In contrast to the IWCLL criteria, the Brazilian Group of CLL (BGCLL) has proposed a more conservative approach to treatment initiation. This conservative approach was initially derived from clinical practice in Brazilian public hospitals. Treatment delays revealed that many patients were asymptomatic, prompting a reassessment of the initiation criteria. This approach avoids strict criteria for cytopenias and does not consider progressive lymphocytosis or disease-related symptoms as criteria for treatment initiation. However, the safety and efficacy of this approach compared to the strict IWCLL criteria has yet to be systematically evaluated.
In an era of expanding therapeutic options and the increasing use of real-world evidence to inform medical practice, it is of critical importance to understand how different thresholds for treatment initiation impact patient outcomes. This is particularly relevant in settings with limited resources, where overtreatment can pose unnecessary risks and undertreatment can delay effective care. While prior studies have explored individual treatment triggers, this study offers a comprehensive real-world comparison of outcomes between patients managed with restrictive vs. liberal interpretations of IWCLL criteria. Our findings aim to support more individualized, context-aware treatment decisions in CLL.
Methods
Study design, population, and data source
We conducted a retrospective analysis of data from the Brazilian Registry of Chronic Lymphocytic Leukemia (BRCLL) to identify treatment initiation patterns in the first line setting in Brazil. The BRCLL is an observational database in which clinical interventions and follow-up visits are performed at the treating physician's discretion. Clinical data were collected from patients' routine medical records — electronic or paper, depending on the center — and manually entered into the Research Electronic Data Capture (REDCap) platform by trained personnel to ensure standardized data capture across sites. Affiliated with the Brazilian Association of Hematology and Hemotherapy (ABHH), the BRCLL includes approximately 4350 registered patients from 57 centers. This analysis includes patients enrolled in the BRCLL from January 2009 to July 2023 from 41 participating centers, which are listed in the Supplementary Material. Only patients who had not yet received treatment and did not meet criteria for treatment initiation at the time of registry inclusion were considered, regardless of the time since diagnosis.
Patients were eligible for inclusion in this analysis if they had a confirmed diagnosis of CLL according to the IWCLL criteria, were 18 years of age or older, and had at least three months of clinical follow-up after diagnosis. A total of 3491 patients met these criteria. However, information was insufficient to determine the exact indication of treatment initiation for 945 patients (27%). Additionally, 35 patients (1%) had no data on their final follow-up or death. Consequently, the final analysis included 2511 patients.
The BGCLL criteria provide a more restrictive interpretation of the commonly used IWCLL criteria. They set lower thresholds for cytopenias (hemoglobin <9.5 g/dL and/or platelets <50,000/mm3) and exclude progressive lymphocytosis and disease-related symptoms as indications for treatment. And in accordance with the IWCLL, BGCLL treatment criteria also include massive or symptomatic lymphadenopathy, massive or symptomatic splenomegaly, autoimmune complications unresponsive to steroids, and symptomatic or functional extranodal involvement. Therefore, for this analysis we separated patients in two groups.
-
•
Liberal criteria group: Patients who meet the broader, internationally established criteria outlined in the 2018 IWCLL guidelines. These criteria include progressive marrow failure (e.g., hemoglobin <10 g/dL or platelets <100,000/mm3); progressive or symptomatic lymphadenopathy or splenomegaly; progressive lymphocytosis; autoimmune cytopenias unresponsive to corticosteroids; disease-related symptoms (e.g., B symptoms); and symptomatic extranodal involvement.
-
•
Restrictive criteria group: Patients who met the more conservative criteria proposed by the Brazilian Group of CLL (BGCLL). These criteria adopt stricter thresholds for cytopenias (hemoglobin <9.5 g/dL and/or platelets <50,000/mm3) and exclude progressive lymphocytosis and disease-related symptoms as standalone indications for treatment. The BGCLL criteria include massive or symptomatic lymphadenopathy or splenomegaly, autoimmune complications unresponsive to steroids, and symptomatic or functional extranodal involvement, which is consistent with the IWCLL criteria.
Supplementary Figure S1 shows a Flowchart of patient selection and stratification by treatment indication criteria.
It is important to note that the BGCLL criteria adopted a more stringent threshold for anemia, defining cytopenia as hemoglobin levels below 9.5 g/dL. This is in contrast to the IWCLL cut-off of 10 g/dL. This decision was based on expert consensus and a retrospective analysis of our cohort, which included over 100 patients with counts within this range. Many of these patients had stable hemoglobin levels and no clinical evidence of progression. The lower threshold better reflects cases with a declining hemoglobin trend and marrow involvement, helping to identify patients who truly need treatment.
To compare the IWCLL and BGCLL treatment criteria, all patients were retrospectively classified by the central study team based on individual-level clinical data available in the registry. This allocation was not performed by local investigators or determined by treatment site. Instead, each patient was independently assessed to determine whether they met the treatment initiation criteria defined by the IWCLL 2018 guidelines (“liberal” group) or by the BGCLL proposal (“restrictive” group), according to their clinical status at the time treatment was initiated. This per-patient classification ensured consistent application of both sets of criteria across all participating centers.
The primary outcomes were treatment-free survival (TFS) and overall survival (OS), analyzed by indication group and treatment status, to evaluate the clinical safety and efficacy of the BGCLL's restrictive approach compared to standard IWCLL criteria.
Statistical analysis
Continuous variables were expressed as medians and ranges, and categorical variables were expressed as percentages. Overall survival was calculated from the date a patient met at least one IWCLL criterion until the patient died or reached the last follow-up. Treatment-free survival was defined as the period from meeting an IWCLL criterion until initiating first-line treatment, experiencing death, or reaching the last follow-up. Survival estimates were calculated using the Kaplan–Meier method and compared using log-rank tests.
Adjusted analyses were conducted using Cox proportional hazards models, and the results are presented as hazard ratios (HRs) and 95% confidence intervals (CIs). Variables with a P-value of less than 0.05 in the univariate analysis were included in the multivariable models. The main covariates included in the models were age (categorized as <65 vs. ≥65 years), Binet stage (A vs. B/C), hemoglobin (categorized as <10 vs. ≥10 g/dL), and beta-2 microglobulin (categorized as normal vs. elevated). A two-sided P-value of less than 0.05 was considered statistically significant. All analyses were performed using SPSS version 20 (SPSS Inc., Chicago, IL).
Ethical considerations
The registry was approved by the involved Institutional Ethics Committees in accordance with the Declaration of Helsinki and is registered in Plataforma Brasil under CAAE number 09984919.0.1001.5505. No personal identifiable information was collected in the database used for the analysis and informed consent were not required. Data was compiled by trained professionals and stored in a secure database.
Role of funding source
The funders had no role in study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication. Three authors (FMM, VP and FP) are data managers of the BRCLL.
Results
Baseline characteristics
Among the 2511 patients included, the median follow-up time was 58 months (interquartile range [IQR] 3–340). Most patients were male (n = 1,410, 56%) with a median age of 65 years (IQR 58–73). At the time of diagnosis, 1409 patients (59%) were classified as Binet stage A, 489 (20%) as stage B, and 492 (21%) as stage C. The median hemoglobin level at diagnosis was 13.2 g/dL (IQR 11.7–14.5), the median leukocyte count was 25,000/μL (IQR 16,705–54,350), the median lymphocyte count was 19,724/μL (IQR 9890–42,780), and the median platelet count was 181,000/μL (IQR 137,000–231,250). Beta 2-microglobulin levels were available for 1361 patients (54%), of whom 671 (49%) had elevated levels. IGHV mutation status was available in only 423 patients (17%), of whom 248 patients (59%) were unmutated. And the presence of del (17p) and/or TP53 mutation was available in 749 patients (30%), of whom 97 (13%) were positive. Table 1 shows the clinical and laboratory characteristics of the study population.
Table 1.
Baseline clinical and laboratory characteristics of patients with chronic lymphocytic leukemia (CLL) at diagnosis.
| Variable | Value |
|---|---|
| N | 2511 |
| Male Sex—n (%) | 1410 (56%) |
| Age—median (interquartile range) | 65 years (58–73) |
| Binet | |
| A—n (%) | A: 1409 (59%) |
| B—n (%) | B: 489 (20%) |
| C—n (%) | C: 492 (21%) |
| Hb, g/dL—median (interquartile range) | 13.2 g/dL (11.7–14.5) |
| Lymphocytes,/mm3—median (interquartile range) | 19,724/mm3 (9890–42,780) |
| Platelets,/mm3—median (interquartile range) | 181,000/mm3 (137,000–231,250) |
| Elevated B2 microglobulin—n (%) [available in 1361–54%] | 671 (49%) [available in 1361 patients] |
| Unmutated IgHV—n (%) | 248 (59%) [available in 423 patients] |
| FISH for del17p positive and/or TP53 mutation present—n (%) | 97 (13%) [available in 749 patients] |
Values are expressed as median (interquartile range) or number (%), as appropriate. Percentages of available data are shown for each variable.
Distribution by treatment indication criteria
The majority of patients (n = 1,404, 56%) met at least one IWCLL criterion for treatment indication during the follow-up period, after a median of 35 days (range 0–149 months) from the diagnosis of CLL, while only 788 patients (31%) met the restrictive criteria. Median time between diagnosis and meeting a IWCLL criteria was 0 months in the restrictive criteria group and 36 months in the liberal criteria group (P < 0.0001). Some patients had multiple indications for treatment. After a comprehensive retrospective review of each case, the primary indication for each patient was determined based on the severity ratings assigned by the local treatment team. The primary indications for treatment initiation in the group of patients with restrictive criteria (N = 788) included cytopenias in 413 patients (52%), symptomatic adenopathy in 316 patients (40%), autoimmune complications in 22 patients (3%), extranodal involvement in 16 patients (3%), symptomatic splenomegaly in 7 patients (1.5%), and disease-related symptoms in 11 patients (1.5%). It is important to note that although constitutional symptoms were not considered a treatment criterion under the restrictive BGCLL approach when present alone, they may have been recorded as the main indication for treatment in patients who presented with multiple concurrent indications, if symptoms were deemed the most clinically relevant factor at the time of therapy initiation. The primary indications for treatment initiation in the group with liberal criteria but not restrictive criteria (N = 616) were: cytopenias in 358 patients (58%), symptomatic adenopathy in 14 patients (2%), and disease-related symptoms in 101 patients (17%). The remaining patients presented with various symptoms, including progressive lymphocytosis (14%), symptomatic splenomegaly (8%), extranodal involvement (1%), and steroid-refractory autoimmune complications (0.2%). Data comparing the primary treatment indications between the group with liberal criteria and the group with restrictive criteria are presented in Table 2. A significant discrepancy was observed in the frequency of most indications between the two groups. Cytopenias, disease-related symptoms, progressive lymphocytosis, and symptomatic splenomegaly were more frequent in the group with liberal criteria, whereas symptomatic adenopathy and autoimmune complications were more frequent in the group with restrictive criteria. Notably, there were only a few cases (n = 23) in which extranodal involvement was the sole indication.
Table 2.
Comparison of primary treatment indications between patients meeting liberal vs. restrictive IWCLL criteria.
| Indication | Liberal criteria | Restrictive criteria | P |
|---|---|---|---|
| N (%) | 616 | 788 | |
| Cytopenias | 358 (58) | 413 (52) | 0.03 |
| Symptomatic adenopathy | 14 (2) | 316 (40) | <0.0001 |
| Disease-related symptoms | 101 (16.8) | 11 (1.5) | <0.0001 |
| Progressive lymphocytosis | 86 (14) | 0 | <0.0001 |
| Symptomatic splenomegaly | 49 (8) | 7 (1.5) | <0.001 |
| Extranodal involvement | 7 (1) | 16 (2) | 0.21 |
| Autoimmune complications | 1 (0.2) | 22 (3) | <0.0001 |
P-values refer to comparisons between groups using the chi-square test.
When comparing the clinical characteristics of patients between the two groups, patients with liberal criteria alone were younger, had less advanced disease, lower creatinine and Beta 2-microglobulin levels, were less likely to be IGHV unmutated, and were less likely to have del (17p) or TP53 mutations (Table 3).
Table 3.
Comparison of clinical and prognostic features between patients meeting liberal versus restrictive IWCLL criteria.
| Liberal criteria | Restrictive criteria | P | |
|---|---|---|---|
| N | 616 | 788 | |
| Age (median) | 64 | 66 | 0.019 |
| Binet B or C n (%) | 320 (54%) | 537 (73%) | <0.0001 |
| Hemoglobin (median) | 13 | 11.8 | <0.0001 |
| Lymphocytes (median) | 25,900 | 24,900 | 0.88 |
| Platelets (median) | 157,000 | 169,000 | 0.20 |
| Elevated Creatinine n (%) | 32 (10%) | 88 (20%) | <0.001 |
| Elevated B2-Microglobulin n (%) | 190 (54%) | 258 (68%) | <0.001 |
| Unmutated IGHV n (%) | 65 (55%) | 131 (69%) | 0.015 |
| Del17p or TP53 Mutation n (%) | 21 (9%) | 60 (18%) | 0.0020 |
P-values derived from Mann–Whitney U test for continuous variables and chi-square test for categorical variables.
Outcomes by indication group
Median treatment-free survival at 5 years was significantly longer in patients with liberal criteria (21 months) than in those with restrictive criteria (7 months, P < 0.0001). Treatment-free survival at 5 years was significantly lower in patients whose indication was lymphadenopathy (9%, P = 0.0030), and significantly higher in those with disease-related symptoms (34%, P < 0.0001), but similar across all other different primary treatment indications: 16% for cytopenias, 18% for splenomegaly, 19% in those with progressive lymphocytosis and autoimmune complications, 23% for extranodal disease.
Overall survival from diagnosis to last recorded follow-up or death for all 2511 patients was 82% at 5 years. Overall survival at 5 years was significantly higher in patients with no treatment indication (90%), in those whose primary indication was disease-related symptoms (87%), progressive lymphocytosis (90%), extranodal disease (89%), compared to those with treatment indications for cytopenias (69%), lymphadenopathies or splenomegaly (82%), or autoimmune complications (77%). A significantly lower Overall survival at 5 years (69%) was observed in patients with a treatment indication for cytopenias compared to other indications (84%, P < 0.0001, Fig. 1).
Fig. 1.
Overall survival of 2511 patients with chronic lymphocytic leukemia according to treatment indication by the IWCLL criteria.
Among the 1404 patients with a treatment indication, the Overall survival from the time the patient met a treatment indication until the last visit or death was 69% at 5 years. Overall survival at 5 years was significantly worse in patients with restrictive criteria (60%) compared to those with liberal criteria (79%, P < 0.0001, Fig. 2). Table 4 shows the overall survival according to all other potentially significant risk factors and the differences between the restrictive and liberal groups. After adjusting for age, Binet staging, and elevated beta-2 microglobulin, multivariate analysis revealed that having a restrictive treatment indication remained an independent risk factor for worse overall survival (hazard ratio [HR]: 1.728, 95% confidence interval [CI]: 1.30–2.30, P < 0.0001) (Table 5). Variables with excessive missing data, such as creatinine, IGHV status, and del (17p)/TP53, were excluded from the final multivariate model. The proportional hazards assumption was tested using log-minus-log survival plots and Schoenfeld residuals; no significant violations were observed.
Fig. 2.
Overall survival of 1404 patients with chronic lymphocytic leukemia with treatment indication according to the liberal criteria (yellow line) or according to the restrictive criteria (blue line).
Table 4.
Univariate analysis of overall survival (OS) according to risk factors at diagnosis in patients with IWCLL treatment indication.
| Risk factors | N | OS (%) | P |
|---|---|---|---|
| All | 1404 | 69 | |
| Age | |||
| <65 years | 743 | 76 | <0.0001 |
| ≥65 years | 661 | 59 | |
| Binet | |||
| A | 470 | 72 | 0.0070 |
| B or C | 1327 | 65 | |
| Hemoglobin | |||
| ≥10 g/d | 906 | 73 | <0.0001 |
| <10 g/dL | 288 | 49 | |
| Creatinine (missing data: 637–45%) | |||
| Normal | 647 | 69 | <0.0001 |
| Elevated | 120 | 50 | |
| B2-Microglobulin | |||
| Normal | 281 | 80 | <0.0001 |
| Elevated | 448 | 64 | |
| IGHV status (missing data: 1084–77%) | |||
| Mutated | 114 | 67 | 0.76 |
| Unmutated | 206 | 67 | |
| Del (17p) and/or TP53 mutation (missing data: 1084–77%) | |||
| Absent | 484 | 76 | 0.0020 |
| Present | 81 | 56 |
P-values calculated using the log-rank test.
Table 5.
Multivariate Cox proportional hazards model for overall survival (OS) among patients with IWCLL treatment indication.
| Risk factors | Hazard ratio | 95% confidence interval | P |
|---|---|---|---|
| Liberal criteria | 1.728 | 1.301–2.295 | <0.0001 |
| Binet B or C | 1.051 | 0.778–1.419 | 0.75 |
| Age ≥65 years | 1.990 | 1.526–2.594 | <0.0001 |
| Hemoglobin <10 g/dL | 1.183 | 0.854–1.639 | 0.31 |
| Elevated B2-Microglobulin | 1.498 | 1.127–1.992 | 0.0050 |
Hazard ratios (HRs) are adjusted for all covariates shown. Reference categories: liberal criteria, Binet A, age <65 years, hemoglobin ≥10 g/dL, normal B2-microglobulin.
Treatment distribution and impact
Of the 1404 patients who met the IWCLL criteria for CLL, 1149 (82%) received treatment. An additional 80 patients received treatment despite not meeting the criteria. Of the 1229 patients who received frontline therapy, the most common treatment regimens were fludarabine-cyclophosphamide (FC, n = 475) and chlorambucil-based regimens (n = 471). These were followed by CHOP/CVP-like regimens (n = 127). Bruton tyrosine kinase (BTK) inhibitors (n = 73), venetoclax-based regimens (n = 38), bendamustine-rituximab (n = 25), and other therapies (n = 20), including monoclonal antibodies, corticosteroids, and other agents. Patients without an IWCLL treatment indication more frequently received lymphoma-directed therapies (e.g., CHOP/CVP) than those with an indication (P < 0.0001). No significant differences were observed between the liberal and restrictive IWCLL groups regarding the use of CLL-directed therapies or novel agents. There was no statistically significant difference in overall survival between patients treated with chemoimmunotherapy and those treated with targeted agents. However, the small number of patients receiving targeted therapies limits the reliability of this comparison.
The median time to initiation of first-line treatment was 3.3, 0.6, and 0.2 years for patients with Binet stage A, B, and C, respectively.
Of the 788 patients eligible for treatment based on restrictive criteria, 670 (85%) received treatment. We then compared overall survival at 5 years between untreated and treated patients and found a significantly inferior overall survival at 5 years in the untreated group (44% vs. 72% in the treated group, P < 0.0001). Treatment was therefore clearly beneficial for these patients.
And of the 616 patients eligible for treatment based on liberal criteria alone, 479 (78%) received treatment. We also compared overall survival at 5 years from the time the patient met an indication for treatment to the last visit or death between untreated and treated patients and found the opposite result: significantly worse overall survival at 5 years in the treated group (77% vs. 96% in the untreated group, P < 0.0001, Fig. 3), showing that treatment may have been harmful to these patients. However, there were 438 (92%) patients treated with chemoimmunotherapy and only 37 (8%) treated with targeted therapies. Among patients who received targeted therapies there was also a trend toward a lower overall survival between untreated and treated patients at 5 years in the treated group (96% vs. 75%, P = 0.06). Table 6 shows the overall survival according to all other potentially significant risk factors and differences between the restrictive and liberal groups.
Fig. 3.
Overall survival of 616 patients with chronic lymphocytic leukemia with treatment indication according to the liberal criteria who received a first line treatment (blue line) or did not receive treatment (yellow line).
Table 6.
Univariate analysis of overall survival (OS) in patients who met only liberal IWCLL criteria.
| Risk factors | N | OS (%) | P |
|---|---|---|---|
| All | 1404 | 69 | |
| Having received treatment | |||
| No | 137 | 95 | <0.0001 |
| Yes | 479 | 83 | |
| Age | |||
| <65 years | 298 | 88 | <0.0001 |
| ≥65 years | 318 | 70 | |
| Binet | |||
| A | 268 | 80 | 0.53 |
| B or C | 320 | 78 | |
| Hemoglobin | |||
| ≥10 g/dL | 469 | 80 | <0.0001 |
| <10 g/dL | 42 | 60 | |
| B2-Microglobulin | |||
| Normal | 159 | 84 | 0.0030 |
| Elevated | 190 | 75 |
P-values calculated using the log-rank test.
On multivariate analysis, among those with liberal criteria, the fact of having been treated remained an isolated risk factor for worse overall survival (HR 4.83, 95% CI 1.75–13.32, P = 0.0020) after adjustment for age, Binet staging and elevated beta-2-microglobulin (Table 7). Among treated patients, 65 patients (42%) died of infection, including 15 cases of COVID-19. In addition, 17 patients died of disease progression, 16 patients died of cardiovascular causes, including ventricular arrhythmias, and four patients died of secondary neoplasms.
Table 7.
Multivariate Cox proportional hazards model for overall survival (OS) among patients who met only liberal IWCLL criteria.
| Risk factors | Hazard ratio | 95% confidence interval | P |
|---|---|---|---|
| Having received treatment | 4.480 | 1.631–12.302 | 0.0040 |
| Age ≥65 years | 2.155 | 1.393–3.333 | <0.0001 |
| Hemoglobin <10 g/dL | 2.183 | 1.145–3.969 | 0.017 |
| Elevated B2-Microglobulin | 1.405 | 0.919–2.149 | 0.12 |
Hazard ratios (HRs) are adjusted for all covariates shown. Reference categories: no treatment, age <65 years, hemoglobin ≥10 g/dL, normal B2-microglobulin.
A total of 80 patients received treatment despite a lack of indication according to the IWCLL criteria. The most common reasons for initiating treatment were cytopenias above the IWCLL recommended cut-offs, small or asymptomatic lymphadenopathies seen only on imaging studies, mild constitutional symptoms related to other causes, and isolated absolute lymphocytosis without meeting criteria for progressive lymphocytosis. Most of these patients received chemotherapy in combination with immunotherapy. The observed five-year survival rate for this cohort was 84%. The most common cause of death was infection (63%), including four cases of coronavirus disease 2019 (COVID-19). In addition, two patients died of disease progression, one of a cardiac event and one of a secondary neoplasm.
Discussion
The present study evaluated the safety and efficacy of a more conservative approach to treatment indication in CLL. This study highlights the value of disease registry data in identifying variations in clinical practice. The BRCLL registry collects data from multiple centers, reflecting heterogeneity in healthcare access, disease management, and socioeconomic background. Our findings underline the need for individualized CLL management, even with advances in therapies and risk stratification tools, and raise questions about prognostic variation within IWCLL treatment indications.
Compared to the liberal criteria group, patients meeting restrictive criteria were older, more frequently presented with Binet stage B or C, and had higher creatinine levels, suggesting greater disease burden and comorbidities. This group also had a higher prevalence of adverse prognostic factors, including elevated Beta 2-microglobulin, del (17p) or TP53 mutation, and unmutated IgHV, which correlated with poorer survival. These findings suggest that the BGCLL criteria may be more predictive of mortality risk and disease severity. However, treatment did not appear to significantly improve survival in this high-risk population, highlighting the limitations of currently available therapies.
The availability of prognostic markers and treatment options varies significantly across socioeconomic contexts. For example, disparities exist between public and private healthcare in Brazil.6 This limitation primarily stems from restricted access to specialized diagnostic tools, such as TP53 and IGHV mutation testing, rather than a lack of clinical awareness of their prognostic and therapeutic value. Although international guidelines recommend stratification using Beta 2-microglobulin, FISH, TP53 mutation, and IGHV mutation status, these data are often unavailable. This reflects the systemic challenges reported in real-world studies worldwide.7, 8, 9 In this context, the BGCLL's more conservative treatment approach may be particularly relevant. It offers a pragmatic framework to guide clinical decisions when comprehensive biological profiling is not feasible.
Our analysis showed that patients meeting liberal treatment criteria were often treated for constitutional symptoms and progressive lymphocytosis, the latter of which is not a treatment criterion for BGCLL. In contrast, patients with cytopenias had significantly worse overall survival. These findings suggest that some treatment indications, such as disease-related symptoms, progressive lymphocytosis, and extranodal disease, may not strongly predict survival and should be evaluated on a case-by-case basis. Clinical observation without immediate intervention may be appropriate in these cases, as they did not appear to significantly affect overall survival.
Similar survival was observed between patients with subjective symptoms and those with no treatment criteria, supporting the idea that symptoms often attributed to CLL may be related to other causes. Weight loss or fatigue may be due to other conditions, leading to unnecessary CLL treatment and delay in alternative diagnoses. Delaying treatment allows for a full evaluation of symptoms or laboratory abnormalities, which may lead to better outcomes.
A study of 530 patients treated in Spain between 1978 and 2014 found that the most common indications for treatment were symptomatic adenopathy, cytopenias, and progressive lymphocytosis.10 Consistent with our findings, cytopenias were associated with worse overall survival, while adenopathy was associated with longer overall survival compared to other grouped criteria.10
Studies evaluating early intervention with chemotherapy or targeted agents have not shown a significant overall survival benefit, but have shown increased treatment-related toxicity.11, 12, 13, 14 Risks of early treatment include immediate and delayed adverse effects such as myelosuppression, immunosuppression, cardiac events, and secondary malignancies, potentially reducing quality of life.15,16 Early intervention would only be justified if it effectively reduces CLL-related complications, a hypothesis that remains to be proven. A critical assessment of the risks of overtreatment is necessary, considering late side effects and potential resistance to future treatments. Our results suggest that prolonged observation until restrictive criteria are met is a viable strategy to minimize unnecessary treatment-related risks.
Despite enrolling patients until 2023, our study found that 92% of treated CLL patients in Brazil received chemoimmunotherapy, while only 8% received targeted therapies. This is consistent with previous registries such as Connect CLL17 (until 2014) and the Ontario Cancer Registry18 (until 2017), which reported a predominant use of chemoimmunotherapy. In contrast, the informCLL Registry19 (2015–2019) showed a greater uptake of targeted therapies, with 45% of first line and 49% of relapsed/refractory patients receiving ibrutinib. The minimal integration of targeted therapies in our cohort highlights delays in availability and accessibility in Brazil, compared to regions with greater access. As a result, the benefits observed in informCLL are not fully realized in our population, as evidenced by significantly inferior overall survival in patients treated according to liberal criteria. These findings highlight the urgent need to improve access to novel therapies in Brazil to keep pace with global advances and improve patient outcomes.
The disparity in survival between the treated and untreated groups among those with liberal criteria highlights the need to prevent deaths in treated patients, especially in light of treatment-related toxicity. Many deaths were due to infectious complications, a significant concern with chemoimmunotherapies but also with targeted therapies such as continuous BTKi, which also increase the risk of cardiovascular and bleeding events.
Our findings support the potential utility of refining treatment indications to focus on clinical criteria that may better identify patients with high-risk disease and meaningful morbidity. This approach could reduce unnecessary treatment, costs, and toxicity while minimizing interference with disease biology and the selection of resistant clones and could prevent avoidable deaths. The financial impact of CLL is significant for patients and healthcare systems, especially in developing countries. It is critical to optimize resource use by identifying patients who will truly benefit from treatment initiation. Resources currently spent on unnecessary treatment could be redirected to patients with clear indications for therapy.
It is important to note that the restrictive approach proposed by the BGCLL improves the precision of treatment decisions. In our cohort, patients who met the restrictive criteria had shorter overall survival, reflecting that these patients generally presented with more advanced or aggressive disease at the time of diagnosis rather than a detrimental effect of the restrictive approach itself. By excluding borderline or low-risk cases, the restrictive criteria help avoid premature treatment of indolent disease while focusing therapeutic efforts on those who are more likely to benefit. In this sense, “clinical benefit” refers to improved patient selection, risk-adapted care, and more efficient use of healthcare resources, not a survival advantage per se.
As a non-randomized retrospective observational study, limitations include lack of control over study group classification, non-random assignment, and data collection by different professionals. To reduce bias, regular data review was performed and staff were trained in accurate data entry. Patients with significant missing or questionable data were excluded. Another limitation of our study is the absence of ethnicity data. In Brazil, a country marked by extensive racial admixture, the classification of individuals into distinct ethnic groups is particularly complex and often unreliable. Ethnicity is typically self-reported and influenced by social, cultural, and regional factors, which can introduce significant misclassification bias. For this reason, we opted not to include ethnicity in our analysis. A prospective, multicenter, international, randomized trial is needed to validate these findings. Such a study should compare the IWCLL criteria with a more restrictive approach, such as the BGCLL criteria. Initiating treatment based on symptomatic, persistent, and progressive cytopenias rather than fixed thresholds may be both safe and beneficial.
Our study spans a lengthy inclusion period, during which the management of CLL has evolved considerably. Advances in diagnostic techniques, prognostic marker assessment, and the advent of novel targeted therapies have significantly changed treatment paradigms. This temporal heterogeneity likely influenced treatment choices and patient outcomes observed in our cohort. While this reflects real-world practice across different eras, it also represents a limitation when interpreting longitudinal data. Nonetheless, our findings highlight the relevance and applicability of the restrictive BGCLL criteria in diverse clinical contexts, including both historical and contemporary treatment landscapes.
In summary, the BGCLL's more restrictive approach to treatment initiation does not confer a direct survival advantage but offers clinical benefit by improving prognostic discrimination. It more accurately identifies patients with higher disease burden and poorer outcomes, while sparing others from unnecessary treatment and its associated toxicities. This selective strategy supports a more individualized approach to CLL management in real-world settings. These results highlight the potential of restrictive criteria to optimize survival outcomes and minimize treatment-related toxicities, supporting a more tailored approach to CLL management.
Contributors
Conception and design: Fernanda Marques, Verena Pfister, Carlos Sérgio Chiattone and Celso Arrais- Rodrigues. Collection and assembly of data: Fernanda de Morais Marques, Verena Pfister, Flavia Parra, Leila Perobelli, Valeria Buccheri, Rodrigo Santucci, Sergio Costa Fortier, Vinicius Campos de Molla, Maura Rosane Valerio Ikoma, Glaciano Ribeiro, Nelson Hamerschlak, Abel da Costa Neto, Laura Maria Fogliatto, Vera P. Figueiredo, Talita Silveira, Marcelo Pitombeira de Lacerda, Matheus Vescovi Gonçalves, Carlos Sérgio Chiattone, Celso Arrais-Rodrigues. Data analysis and interpretation: Fernanda Marques, Vinicius Molla, Matheus Vescovi Gonçalves, Jennifer R. Brown and Celso Arrais-Rodrigues. Manuscript writing: Fernanda Marques, Vinicius Molla, Matheus Vescovi Gonçalves, Jennifer R. Brown and Celso Arrais-Rodrigues. Final approval of manuscript: All authors. Accountable for all aspects of the work: All authors.
Data sharing statement
The data that support the findings of this study are available from the corresponding author upon reasonable request. Due to privacy and ethical concerns, data are not publicly available. Access to the data will be granted in accordance with institutional and ethical guidelines.
Declaration of interests
The authors declare no relevant conflict of interests. A preliminary version of this work was presented as an abstract and published in Hematology, Transfusion and Cell Therapy (DOI: https://doi.org/10.1016/j.htct.2024.09.551).
Acknowledgements
The authors express their profound gratitude to all the physicians, nurses, and data managers who contribute to the Brazilian Registry of CLL. They would also like to extend their appreciation to the Brazilian Association of Hematology and Hemotherapy (ABHH) team for their unwavering support and funding, with particular recognition to Aline Pimenta and Renata Fava.
Fernanda Marques was supported by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES, code process no. 001).
Footnotes
Disclaimer: This summary is available in Portuguese in the Supplementary Material.
Supplementary data related to this article can be found at https://doi.org/10.1016/j.lana.2025.101234.
Appendix A. Supplementary data
References
- 1.Hallek M., Cheson B.D., Catovsky D., et al. iwCLL guidelines for diagnosis, indications for treatment, response assessment, and supportive management of CLL. Blood. 2018;131(25):2745–2760. doi: 10.1182/blood-2017-09-806398. [DOI] [PubMed] [Google Scholar]
- 2.Hallek M., Al-Sawaf O. Chronic lymphocytic leukemia: 2022 update on diagnostic and therapeutic procedures. Am J Hematol. 2022;96(12):1679–1705. doi: 10.1002/ajh.26367. [DOI] [PubMed] [Google Scholar]
- 3.Eichhorst B., Ghia P., Niemann C.U., et al. ESMO Clinical Practice Guideline interim update on new targeted therapies in the first line and at relapse of chronic lymphocytic leukaemia. Ann Oncol. 2024;35(9):762–768. doi: 10.1016/j.annonc.2024.06.016. [DOI] [PubMed] [Google Scholar]
- 4.Rozovski U., Hazan-Halevy I., Keating M.J., Estrov Z. Personalized medicine in CLL: current status and future perspectives. Cancer Lett. 2014;352(1):4–14. doi: 10.1016/j.canlet.2013.07.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Steingrímsson V., Lund S.H., Dickman P.W., et al. Survival, causes of death, and the prognostic role of comorbidities in chronic lymphocytic leukemia in the pre-ibrutinib era: a population-based study. Eur J Haematol. 2022;108(2):145–153. doi: 10.1111/ejh.13720. [DOI] [PubMed] [Google Scholar]
- 6.Pfister V., Marques F.M., Parra F., et al. Lower access to risk stratification tests and drugs, and worse survival of chronic lymphocytic leukaemia patients treated in public as compared to private hospitals in Brazil: a retrospective analysis of the Brazilian registry of chronic lymphocytic leukaemia. EJHaem. 2023;3(3):698–706. doi: 10.1002/jha2.444. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Seymour E.K., Ruterbusch J.J., Beebe-Dimmer J.L., Schiffer C.A. Real-world testing and treatment patterns in chronic lymphocytic leukemia: a SEER patterns of care analysis. Cancer. 2019;125(1):135–143. doi: 10.1002/cncr.31738. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Mato A.R., Ghosh N., Sharman J.P., et al. Real-world prognostic testing and treatment patterns in CLL/SLL: results from 1462 patients in the informCLL registry. Blood Adv. 2023;7:4760–4764. doi: 10.1182/bloodadvances.2022008068. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Mato A.R., Barrientos J.C., Ghosh N., et al. Prognostic testing and treatment patterns in chronic lymphocytic leukemia in the era of novel targeted therapies: results from the informCLL Registry. Clin Lymphoma Myeloma Leuk. 2020;20(1):174–183.e3. doi: 10.1016/j.clml.2019.10.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Mozas P., Rivas-Delgado A., Baumann T., et al. Analysis of criteria for treatment initiation in patients with progressive chronic lymphocytic leukemia. Blood Cancer J. 2018;8(1):10. doi: 10.1038/s41408-017-0044-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Dighiero G., Maloum K., Desablens B., et al. Chlorambucil in indolent chronic lymphocytic leukemia. French cooperative group on chronic lymphocytic leukemia. N Engl J Med. 1998;338(21):1506–1514. doi: 10.1056/NEJM199805213382104. [DOI] [PubMed] [Google Scholar]
- 12.Hoechstetter M.A., Busch R., Eichhorst B., et al. Early, risk-adapted treatment with fludarabine in Binet stage A chronic lymphocytic leukemia patients: results of the CLL1 trial of the German CLL study group. Leukemia. 2017;31:2833–2837. doi: 10.1038/leu.2017.246. [DOI] [PubMed] [Google Scholar]
- 13.Herling C.D., Cymbalista F., Groß-Ophoff-Müller C., et al. Early treatment with FCR versus watch and wait in patients with stage Binet A high-risk chronic lymphocytic leukemia (CLL): a randomized phase 3 trial. Leukemia. 2020;34(8):2038–2050. doi: 10.1038/s41375-020-0747-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Langerbeins P., Zhang C., Robrecht S., et al. The CLL12 trial: ibrutinib vs. placebo in treatment-naïve, early-stage chronic lymphocytic leukemia. Blood. 2022;139(2):177–187. doi: 10.1182/blood.2021010845. [DOI] [PubMed] [Google Scholar]
- 15.van den Broek E.C., Oerlemans S., Nijziel M.R., Posthuma E.F.M., Coebergh J.W.W., van de Poll-Franse L.V. Impact of active surveillance, chlorambucil, and other therapy on health-related quality of life in patients with CLL/SLL in The Netherlands. Ann Hematol. 2015;94(1):45–56. doi: 10.1007/s00277-014-2161-6. [DOI] [PubMed] [Google Scholar]
- 16.Buzaglo J.S., Miller M.F., Zaleta A.K., et al. Chronic lymphocytic leukemia patient-reported outcomes and quality of life: findings from the cancer experience registry. Blood. 2017;130(Supplement 1):2122. [Google Scholar]
- 17.Mato A., Nabhan C., Lamanna N., et al. The Connect CLL Registry: final analysis of 1494 patients with chronic lymphocytic leukemia across 199 US sites. Blood Adv. 2020;4(7):1407–1418. doi: 10.1182/bloodadvances.2019001145. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Seung S.J., Hurry M., Hassan S., et al. Examining treatment patterns and real-world outcomes in chronic lymphocytic leukemia using administrative data in Ontario. Curr Oncol. 2021;28(6):4832–4844. doi: 10.3390/curroncol28060408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Ghosh N., Sharman J.P., Gutierrez M., et al. Real-world treatment patterns and outcomes by line of therapy and race in patients with chronic lymphocytic leukemia/small lymphocytic lymphoma treated in the United States: results from the final analysis of the prospective, observational, informCLL registry. Clin Lymphoma Myeloma Leuk. 2024;24(9):e301–e313. doi: 10.1016/j.clml.2024.05.009. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.