Safety and Efficacy of Bendamustine and Rituximab (BR) Regimen in Indian Chronic Lymphocytic Leukaemia Patients

Abstract

We investigated the safety and efficacy of bendamustine-rituximab (BR) in previously untreated symptomatic and advanced CLL patients, as there is no data available on BR from the Indian subcontinent.This retrospective study included 120 consecutive treatment naïve patients with CLL without del (17p), who were registered at the Department of Medical Oncology, AIIMS between January 2010 and July 2018. Bendamustine was given at a dose of 90 mg/m² on days 1 and 2, combined with rituximab 375 mg/m² rituximab on day 1, every 28 days for up to 6 courses. Event-free survival (EFS) was defined as the date of treatment to date of relapse, disease progression, or death due to any cause.The median age was 57 years (range: 30–75 years). As per the clinical Rai stage, 30 (25%) patients were in stage II, 42 (35%) were in stage III and 48 (40%) were in stage IV. ZAP70 was positive (> 20%) in 50%, CD 38 was positive (> 30%) in 33%, and CD49d was positive (> 30%) in 49% of cases. Beta-2 microglobulin (B2M) was elevated (≥ 3.5 mg/L) in 80% of cases. Fifty-five cases (50%, n = 110) were IGHV mutated. The mean number of cycles was 5 (1–6). The overall response rate (ORR) seen with BR was 90% and complete response was 45%. Median progression-free survival was 24 months with a median follow-up period of 29 months. Haemoglobin (< 10 g/dL), elevated B2 M, unmutated IGHV had a statistically significant adverse impact on EFS on univariate analysis but on multivariate analysis, only IGHV mutation status was found to had significance on EFS. The median EFS was 27 months in IGHV mutated versus 18 months in IGHV unmutated-CLL patients (p = 0.001). Grade 3/4 neutropenia, thrombocytopenia, anemia, and infections were observed in 30.6%, 8%, and 12% respectively. The most common non-hematological toxicity was skin rash which was grade 1/2 in 24 (20%) cases and grade 3/4 in 12 (10%) cases. This is the largest study from India to demonstrate the safety and efficacy of BR in symptomatic CLL patients. BR is an effective and safe regimen in the first-line treatment of CLL. Unmutated-CLL patients have inferior EFS than mutated-CLL patients. Skin toxicity was the most common adverse effect seen in our population which was observed in around one-third of cases.

Introduction

Chronic lymphocytic leukemia (CLL) is a lymphoid neoplasm characterized by accumulation of monoclonal, mature, but functionally incompetent, B-lymphocytes in the peripheral blood, bone marrow, and lymphoid organs. Though patients with CLL harboring del(17p) or deleterious mutation in the TP53 gene respond poorly to chemoimmunotherapy and are better treated with targeted agents, those without these abnormalities are often managed with chemoimmunotherapy [1]. The combination of fludarabine, cyclophosphamide, and rituximab(FCR) is a standard first-line treatment option for physically fit patients with advanced CLL after it showed high response rates (including deep responses) and long term remissions in this patient population [2, 3]. Despite its good efficacy, fludarabine-based combination therapy is associated with significant and prolonged myelosuppression, more so in the elderly population [2, 4]. Fludarabine-associated hematological toxicity has, in turn, been associated with a higher need for transfusion, increased infections, and worse overall survival [5]. The cytopenias resulting from FCR are especially problematic in regions that have a high prevalence of multi-drug resistant organisms including tuberculosis. In a previous report from India, approximately one-third of patients with CLL treated with fludarabine-containing regimens developed pulmonary tuberculosis [6].

Chemoimmunotherapy approaches that are less intensive than FCR have been explored by various groups. Combinations of lower dose fludarabine and cyclophosphamide with a higher dose of rituximab have resulted in high overall and complete response rates [7, 8]. Bendamustine–rituximab (BR) is another less toxic combination, which showed good efficacy and safety in the first-line treatment of patients with CLL [9]. Encouraged by these results, the German CLL group compared BR with FCR in a large phase 3 non-inferiority trial; however, the findings of this study failed to show non-inferiority of BR [10]. Of note, BR was associated with significantly less neutropenia, thrombocytopenia, severe infections, and secondary infections. Based on these results, BR has become a preferred regimen for clinicians due to its relative ease of administration and manageable toxicity profile, despite somewhat lower efficacy in comparison to FCR [11, 12].

Most of the evidence pertaining to BR combination chemoimmunotherapy in CLL comes from a clinical trial setting, with sparse real-world practice data. We conducted this retrospective analysis to investigate the safety and efficacy of BR in previously untreated symptomatic and advanced CLL patients from India, as there is little data available on this combination from the Indian subcontinent.

Materials and Methods

This retrospective study included all consecutive treatment-naïve symptomatic patients with CLL without del 17p, who received treatment with BR combination chemoimmunotherapy at the Department of Medical Oncology, All India Institute of Medical Sciences (AIIMS), New Delhi, India, between January 2010 and July 2018. Patient records were accessed to collect the requisite information, including demographic details, results of diagnostic studies, response to therapy, and treatment-related toxicity. Patients with del(17p) were excluded from this analysis as this group of patients was preferentially treated with alternate protocols. The study protocol was approved by the institutional ethics committee.

Diagnosis of CLL was established based on clinical and hematological parameters that included peripheral blood smear examination and flow cytometry. All patients had documentation of an absolute lymphocyte count greater than 5 × 10⁹/L on at least two occasions one month apart. Diagnosis was confirmed by immunophenotyping using a panel of antibodies CD19, CD20, CD5, CD23, CD 79b, CD10, CD25, CD45, CD103, CD3, CD4, CD8, CD11, surface Igkappa, lambda and FMC7.Stage assignment was done using the Rai staging system [13]. An international workshop on CLL criteria was used for treatment decisions and response assessment [14]. Patients were treated if there were persistent fatigue unrelated to other causes, persistent or progressive systemic symptoms (fever, night sweats, weight loss), increasing bone marrow failure (anemia, thrombocytopenia), autoimmune hemolytic anemia or thrombocytopenia not responding to steroids, short lymphocyte doubling time (> 50% over two months), disease-related recurrent bacterial infection or progressive enlargement of the lymph nodes, liver, or spleen. Bendamustine was administered at a dose of 90 mg/m² on day 1 and 2, combined with rituximab 375 mg/m² on day 1, every 28 days for up to 6 courses.

The primary endpoint of the study was overall response rate (ORR); secondary endpoints included event free survival (EFS) and toxicity profile. Endpoints were evaluated in the intent to treat (ITT) population, which included all patients who received at least one cycle BR. EFS was defined as the time from initiation of treatment to relapse, disease progression, or death from any cause, whichever occurred first. Treatment-related non- haematological toxicity was assessed using NCI common terminology criteria for adverse events (CTCAE) version 4 and treatment related haematological toxicity was evaluated by international workshop on CLL criteria [14].Time-to-event survival analysis was done using the Kaplan–Meier method and differences between survival curves were examined using the log-rank test. Cox proportional hazards model was used to identify independent predictors of outcome. All tests were two-sided, and a p-value < 0.05 was considered statistically significant. Statistical analyses were performed using STATA version 13.0 (StataCorp, Texas).

Results

During the study period, 250 treatment-naïve patients with CLL were registered at our institute. Of these, 83 patients are currently under observation whereas 167 required treatment upfront or due course of time. From the subset of patients who underwent therapy for CLL, one hundred and twenty consecutive patients who received BR regimenwere included in this analysis. Patient characteristics are summarized in Table 1. Briefly, the median age of the cohort was 57 years (range 30–75 years) with nearly three-fourths being male. The median total lymphocyte count was 55 × 10⁹/L (range 23–410). Most (40%) patients had clinical Rai stage IV disease.ZAP70 was positive (≥ 20%) in 51 patients, CD 38 was positive (≥ 30%) in 33 while CD49d was positive (≥ 30%) in 47 patients. Beta-2 microglobulin (B2M) was elevated (≥ 3.5 mg/L) in 72 (80%, n = 90) cases. Fifty-five patients (50%, n = 110) had IGHV mutated. Eighteen percent (n = 22) of all patients had comorbid conditions, with type 2 diabetes mellitus and hypertension being the most common. Ninety (75%) patients had eastern cooperative oncology group performance status 1.

Table 1.

Clinico-hematological parameters of patients with chronic lymphocyticleukaemia

Parameter	No. of patients (%) or median (range)
Age (years)	57 (30–75)
Sex
Male	84 (70%)
Female	36 (30%)
Absolute lymphocyte count (× 109/L)	55.9 (23–410)
Haemoglobin (g/dL)	11.1 (4–14)
Platelet count (× 10 9/L)	110 (10–390)
Rai Stage
Stage II	30 (25%)
Stage III	42 (35%)
Stage IV	48 (40%)
Flow cytometry markers
ZAP70positive (n = 102)	51 (50%)
CD38 positive* (n = 99)	33 (33%)
CD49d positive^ (n = 96)	47 (49%)
Beta-2 microglobulin (≥ 3.5 mg/L)(n = 90)	72 (80%)
Mutated IGHV status (n = 110)	55 (50%)
Comorbidities	22 (18.3%)

ZAP70 Zeta-chain associated protein kinase-70 (positivity cut-off 20%), CD Cluster of differentiation, IGHV Immunoglobulin heavy chain variable region

*CD38 positivity cut-off > 20%, ^CD49d positivity cut-off > 30%,

Ninety percent of patients (n = 108) could complete six courses of BR chemoimmunotherapy (median 6 cycles; range 1–6 cycles), with ten patients requiring dose modification for toxicity. Treatment was discontinued during the course of therapy in 12 patients because of adverse events (death = 1, hematological toxicity = 2, skin rash = 4, severe sepsis = 3, progressive disease = 3 patients). In the ITT population, overall response and complete response (CR) were observed in 108 (90%) and 54 (45%) patients, respectively. Seven patients had progressive disease whereas five patients had stable disease as the best response to therapy. With a median follow-up of 29 months, the median EFS for the full cohort was 24 months, while2-year EFS was 70%. A total of45EFS events were recorded during the study period (death = 20, relapse = 15, progressive disease = 10). The cause of mortality was progressive disease in 14 patients, infection in 2 patients, and vascular event in 4 patients. We evaluated various factors (age, sex, total lymphocyte count, hemoglobin, platelet count, ZAP70, CD38, CD49d, B2M, and IGHV mutation status) for their impact on progression-free survival. Of these, hemoglobin less than 10 g/dL, elevated B2M, and unmutated IGHV had been predictive ofEFS on univariate analysis. However, on multivariate analysis, only unmutated IGHV status was found to be associated with significantly worse EFS [hazard ratio 3.6; 95% confidence interval, 1.1–9.2; p = 0.001]. The median EFS was 27 months in IGVH mutated subset and18 months in IGHV unmutated-CLL patients (p = 0.001) (Table 2).

Table 2.

Univariate and multivariate Cox proportional hazard regression model showing the effect of demographic and clinical variables on the events

Variables	Univariate analyses		Multivariate analyses
	HR (95% CI)	p-value	HR (95% CI)	p-value
Age	1.00 (0.98–2.31)	0.878	1.79 (0.70–2.02)	0.576
Gender (Male)	1.11 (1.01–1.68)	0.634	1.70(0.9–2.30)	0.40
Haemoglobin (< 10 g/dL)	1.53 (0.72–2.83)	< 0.001	1.08 (0.72–1.83)	0.356
ALC (× 109/L) (< 50)	1.10 (0.76–1.30)	0.20	1.72 (0.67–2.06)	0.17
Platelet Count (× 109/L) (< 150)	1.05 (0.51–1.62)	0.10	0.68 (0.51–0.92)	0.13
ZAP70 (> 20%)	1.21(0.71–2.1)	0.38	1.11 (0.79–1.84)	0.387
CD38 (> 20%)	1.36(0.94–1.97)	0.10	2.36(0.94–4.97)	0.105
CD49d (> 30%)	1.66(1.14–2.40)	0.08	1.44(0.75–2.40)	0.13
Beta-2 microglobulin (≥ 3.5 mg/L)	2.96 (1.23–7.2)	0.015	1.26 (0.64–2.47)	0.497
Unmutated IGHV status	4.5 (2.55–8.06)	0.01	3.6 (1.10–9.2)	0.001

ALC Absolute lymphocyte count, ZAP70 Zeta-chain associated protein kinase-70 (positivity cut-off 20%), CD Cluster of differentiation, IGHV Immunoglobulin heavy chain variable region

*CD38 positivity cut-off > 20%, ^CD49d positivity cut-off > 30%

The most common toxicity was skin rash, which was grade 1/2 in 24 (20%) patients, and grade 3/4 in 12 (10%) cases. The rash was characteristically pruritic and consisted of polymorphous, erythematous papules and plaques predominantly involving the exposed and unexposed areas of limbs, trunk and back, and rarely the face. The rash typically appeared late during the course of treatment (6 patients: within first 2 cycles, 10 patients: cycles between 2 and 4, 17 patients: between cycles 4 and 6), with four patients developing it in 2 months after completion of treatment. The rash was manageable with interruption of therapy and a brief course of topical emollients with oral antihistamines, but ten patients had reappearance of rash with rechallenge of therapy. Grade 3/4 neutropenia, thrombocytopenia, and anemia were observed in 30.6%, 8%, and 12% of all patients, respectively. Infection and skin rash were the most common nonhematological toxicities. Grade 3 or 4 infections occurred in 12 (10%) patients. Of these, 5 patients developed pneumonia, 5 patients had a neutropenic fever, 3 developed urinary tract infection, and 2 patients developed mucositis.

Discussion

Though novel agents like venetoclax and Bruton’s tyrosine kinase inhibitors are superior to chemoimmunotherapy in terms of efficacy and toxicity, several patients may choose the latter for their time-limited treatment course and lower cost. FCR and BR are both acceptable chemoimmunotherapeutic regimens for young and fit patients with CLL, BR is one of the most frequently used first-line regimens in routine treatment of patients with CLL. In this retrospective analysis of a relatively homogenous CLL patient population without del(17p), chemoimmunotherapy using a combination of bendamustine and rituximab led to high overall and complete response rates, with an acceptable toxicity profile and good treatment completion rates. To our knowledge, this is the largest series of patients treated with BR from the Indian subcontinent. Our analysis showed an ORR of 90%, CR rate of 45%, and 2-year EFS of seventy percent, confirming that BR results in good outcomes among patients with previously untreated CLL. There are two prominent differences between our patient population and most reports from the West. First, the median age in our study (57 years) is more than a decade younger than the median age of 70–72 years seen in Western literature [15, 16]. Our patient demographics closely resemble those seen in another CLL series from India [17]. Second, there is a striking male predominance in our report. This may be attributable to the fact that CLL per se is almost twice as common in males than in females [15, 16]. There may also be a gender difference in healthcare service utilization, that has been observed in India [18, 19].

The outcomes observed in our patient population compare favorably with the results from both clinical trials and real-world settings [9, 20]. Table 3 summarizes the reported efficacy outcome of BR from both trials and various real-world settings. It may be seen from the table that the overall response rate in the Indian population is comparable to that observed in previous studies but the complete response rate is high. We speculate the reason for this high CR rate is that 108 (90%) patients completed the planned 6 cycles of the BR regimen (though 10 of these patients required dose modification for toxicity). In contrast, the completion rates in other series have ranged from 71 to 77% [20, 21]. Even in the phase II German trial that first tested the combination of bendamustine and rituximab, only 73.5% of patients could complete the full 6 cycles of treatment [9]. However, when a similar patient population as ours was treated with near full-course BR, an ORR of 100% and a CR rate of 58% was achieved [22]. Another possible reason for this superior CR rate is that we excluded patients with deletion of the short arm of chromosome 17 from our series, as this aberration is a known predictor of suboptimal response to chemoimmunotherapy [23].Among the various putative factorsaffecting EFS, we found that unmutated IGHV status was associated with a significantly lower EFS. An unmutated IgHV status has been consistently associated with worse progression-free survival among patients treated with chemoimmunotherapy, independent of the specific regimen used (FCR or BR) [10, 24]. For this subset of patients, alternative therapeutic strategies targeting the B-cell antigen receptor and/or B-cell lymphoma-2 pathways could potentially be employed as a more effective approach [25]. Other potentially important predictors of EFS (e.g., beta-2 microglobulin and ZAP70, etc.) may have missed statistical significance due to the relatively small sample size.

Table 3.

Comparison of retrospective and prospective studies on Bendamustine and rituximab

Study (references)	Country	N	Median age (years)	ORR (%)	CR (%)	Median PFS (months)	Grade ¾ hematological toxicity (%) anemia/neutropenia/thrombocytopenia	Cutaneous toxicity all grade (%)
Fischer et al. [9]	Germany	117	64	88	23.1	33.8	19.7/19.7/22.2	n/a
Laurenti et al. [21]	Italy	70	72	88.6	31.4	31.4	2/20/4	3
Gentile et al. [20]	Italy	279	70	86.4	28	28	15.9/25.9/15.4	15.9
Špaček et al. [26]	Czech Republic	83	71	88	20.5	20.5	9.6/48.2/16.9	7.2
Panovska et al. [11]	Czech Republic	257	70.5	85.2	47.5	37 (EFS)	11.7/49.4/7.8	n/a
Tejaswi et al. [17]	India	57	61	91.2	77.2 (CRu)	NR	14/52.6/5.2	n/a
Present study	India	120	57	90	45	24 (EFS)	12/30.6/8	30

N Number of patients, ORR Overall response rate, CR Complete response, PFS Progression free survival, EFS Event free survival

It is probably the more favorable toxicity profile of BR that makes it the preferred chemoimmunotherapy regimen for most clinicians, despite lower efficacy compared to FCR. The toxicity pattern observed in our study was somewhat different from that seen in Western literature (Table 3) [9, 20, 26]. Importantly, the hematological toxicity of BR was much lower, probably due to our younger patient population (median age 57 years) and much fewer (18%) patients having co-morbidities [11]. Despite a lower proportion of patients witnessing hematological toxicity, the incidence of grade 3 or worse infections was comparable to that seen in series from Italy and the Czech Republic [11, 20]. Consequent to its low rate of associated adverse events, many clinicians consider offering BR-based treatment to elderly patients. In our study, the oldest patient was 75 years old, while other authors have treated patients up to the age of 87 years using the BR combination [20, 21].

Curiously, cutaneous toxicity (skin rash) was the most common toxicity observed in our cohort, occurring in one-third of all cases. This cutaneous rash was likely due to bendamustine, as none of these patients had a prior history of skin lesions, there were no concomitant drugs that were thought to have potentially caused the rash, and it subsided within 2 months of completion of chemotherapy in all cases. We performed a biopsy of these lesions in 6 patients, and histopathological evaluation revealed non-specific chronic inflammation. This peculiar rash associated with the use of bendamustine has been seen in a high proportion of patients from the Indian subcontinent, though it has also been reported in Italian patients at a relatively lower frequency [20, 27, 28]. Though the rash is generally mild and manageable with topical steroids and anti-allergic agents, it may occasionally be severe [29, 30]. The pathogenesis of this cutaneous toxicity is unknown, but two postulated hypotheses include hypersensitivity reaction or inflammatory/immune response in patients with specific haplotypes [29], or an exaggerated response to insect bites, which has been reported in patients with CLL [31, 32]. Whether these are more common in Indian patients due to higher prevalence of specific haplotypes or due to insect-bite exposure, remains speculative.

The limitations of our study, besides those inherent to retrospective analysis, include a relatively small sample size, as a result of which we may have missed identification of predictive biomarkers. Additionally, data on minimal residual disease, which would have proved valuable for assessing the depth of response to therapy, was not collected.

In conclusion, this study from India demonstrated that BR is an effective and safe regimen in the first-line treatment of CLL, that may be offered to most patients who prefer a time-limited treatment strategy. Patients who have an unmutated IgHV status have inferior EFS in comparison to those with mutatedIgHV. Skin toxicity was the most common adverse effect observed in our population (seen in around one-third of cases) while hematological toxicity was relatively low.

So BR can be considered as first line chemoimmunotherapy in denovo case of del 17p- CLL in the Indian context. Even the advantages of first line FCR over BR seen in the clinical trial were not seen in the community practise in the CLL registry study [33].

Declarations

Conflict of interest

The authors have no conflict of interest to disclose.