Bendamustine-induced rash is associated with a favorable prognosis in patients with indolent B-cell lymphoma.

Bendamustine is now recognized as a key drug for indolent B-cell lymphoma (iBCL), mantle cell lymphoma (MCL) and chronic lymphocytic leukemia (CLL). Skin toxicity associated with bendamustine is one of the characteristic adverse effects. We retrospectively examined the relationship between bendamustine-associated drug rashes and disease prognosis of iBCL and MCL at our institution. Between January 2011 and August 2019, 65 patients (39 men and 26 women, median age 68, range 41-84 years) were treated with bendamustine alone (n=11, 120 mg/m2 on days 1 and 2) or a combination of rituximab and bendamustine (n=54, 90 mg/m2 on days 1 and 2). Of these patients, 47 had follicular lymphoma (FL), 10 had MCL and 8 had other iBCLs. Drug rash occurred in 27 (41.5%). Eight cases (29.6%) were grade 1, 5 (18.5%) were grade 2 and 14 (51.9%) were grade 3. The onset was in the first course in 17 (63.0%), 2nd course in 5 (18.5%), 3rd course in 2 (7.4%), 4th course in 1 (3.7%) and 5th course in 2 (7.4%). No treatment was administered in 1 case (3.7%), topical steroid was applied in 10 (37.0%), antiallergic drug was administered in 2 (7.4%), topical steroid and antiallergic drug were administered in 5 (18.5%), and oral and topical steroid and antiallergic drug were administered in 9 (33.3%). The 3-year progression-free survival (PFS) and overall survival (OS) in patients with rash development were 80.0% and 85.5%, respectively, and those in patients without development were 36.4% and 54.0%, respectively (p=0.009 and 0.02, respectively). By multivariate analysis, the development of rash was associated with a better PFS and a diagnosis of iBCL was associated with a better OS. This study revealed that bendamustine-induced rash is associated with a favorable prognosis among patients with iBCL.

INTRODUCTION

Bendamustine is a cytotoxic agent with a multifaceted mechanism of action. Structurally, it includes both a mechlorethamine group and a benzimidazole ring, which confer properties of both alkylators and purine analogues. Bendamustine directly damages DNA, and induces apoptosis and mitotic catastrophe. Bendamustine is used to treat indolent B-cell lymphoma (iBCL), mantle cell lymphoma (MCL) and chronic lymphocytic leukaemia., Several trials demonstrated high efficacy and manageable toxicity of bendamustine in relapsed/refractory iBCL and MCL patients, both as monotherapy and in combination with rituximab (BR)., The efficacy and safety of BR as first-line treatment for iBCL and MCL patients were also confirmed in several phase III trials.,

Skin toxicity is one of the important adverse effects of bendamustine. Rash is characteristic, often severe, and affects the treatment, but few detailed reports are available.- In this study, we examined the relationship between bendamustine-associated rash and clinical features, including clinical outcome, for iBCL and MCL at our institution.

MATERIALS AND METHODS

We reviewed the medical records of 65 patients with iBCL, consisting of follicular lymphoma (FL), lymphoplasmacytic lymphoma (LPL), marginal zone lymphoma (MZL), mucosa-associated lymphoid tissue (MALT) lymphomas, lymphoplasmacytic lymphoma (LPL) and lymphocytic lymphoma (SLL) or MCL who were treated with bendamustine at our institution as first- or second-line treatment between 2011 and 2019. The diagnosis was made according to the World Health Organization (WHO) 2008 classification. The rashes were graded according to the Common Terminology Criteria for Adverse Events version 4.0. The follow-up period was defined as from the initiation of bendamustine to the last contact. The cut-off date for analysis was August 2020.

Associations between clinical features and skin rashes were analyzed by univariate analysis using the nonparametric Mann–Whitney U test or Fisher’s exact test, as appropriate. p<0.05 was considered significant. Progression-free survival (PFS) was calculated from the date of diagnosis of lymphoma for the patients with initial treatment and date of relapse/progression for salvage therapy, to the date of relapse, progression or death from any cause, and censored at the last verifiable progression-free date. Overall survival (OS) was defined as the time from diagnosis of lymphoma for the patients with initial treatment and date of relapse/progression for salvage therapy, respectively, to death from any cause, and censored at the last verifiable surviving date. Probabilities of PFS and OS were estimated using the Kaplan-Meier method; time-to-event outcomes were compared among risk groups using the log-rank test. Multivariate analyses were performed with a Cox proportional multiple regression model. All statistical analyses were performed with EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Sixty-five patients (39 men and 26 women; median age 68 years, range 41–84 years) with iBCL or MCL who were treated with bendamustine between January 2011 and August 2019 were analyzed (Table 1). The histological diagnoses were 47 FLs, 10 MCLs, 4 MZLs, 2 MALT lymphomas, one LPL and one SLL. At the time of bendamustine treatment, 20 patients (30.8%) were previously untreated; the others had relapsed or become refractory. Eleven patients received bendamustine monotherapy (120 mg/m2) on days 1 and 2 every 3 weeks, whereas 54 patients received a combination of bendamustine (90 mg/m2) on days 1 and 2 and rituximab (375 mg/m2) on day 1 every 4 weeks.

Table 1

Characteristics of 65 patients with indolent B-cell lymphoma and mantle cell lymphoma treated by bendamustine

Age	Range (Median)	41–84 (68)
Sex	Male/Female	39/26
Diagnosis (WHO classification)
	Follicular lymphoma	47 (72.3%)
	Mantle cell lymphoma	10 (15.4%)
	Marginal zone B-cell lymphoma	4 (6.2%)
	MALT lymphoma	2 (3.1%)
	Lymphoplasmacytic lymphoma	1 (1.5%)
	Small lymphocytic lymphoma	1 (1.5%)
Disease status
	Newly diagnosed	20 (30.8%)
	Relapse/Refractory	45 (69.2%)
Chemotherapy
	Bendamustine	11 (16.9%)
	Bendamustine+Rituximab	54 (83.1%)
Response
	CR	47 (72.3%)
	PR	9 (13.8%)
	CR+PR	56 (86.2%)
	PD	9 (13.8%)

MALT, mucosa-associated lymphoid tissue; CR, complete response; PR, partial response; PD, progressive disease

Twenty-seven patients developed drug rash (41.5%; Table 2). Eight cases (29.6%) were grade 1, 5 (18.5%) were grade 2, and 14 (51.9%) were grade 3. Most were erythema with/without concrescence tendency. Thirteen patients had localized patterns (48.4%) and 14 had generalized patterns (51.9%). No toxic epidermal necrolysis was observed. Seventeen patients (63.0%) experienced symptom onset during the first course, 5 (18.5%) during the second, 2 (7.4%) during the third, 1 (3.7%) during the fourth, and 2 (7.4%) during the fifth. Treatment was based on grade and localization, in addition to observation in 1 (3.7%), steroid topical application in 10 (37.0%), antiallergic drugs in 2 (7.4%), topical steroid application and antiallergic drugs in 5 (18.5%), and oral and topical steroid administration and antiallergic drugs in 9 (33.3%). In 5 of 14 cases of grade 3 rashes, bendamustine was discontinued permanently and changed to rituximab monotherapy (n=3) or watchful waiting (n=2). The median follow-up duration was 24 months (range 1-101) in all patients. The 3-year PFS was 80.0% in 27 patients who developed drug rashes and 36.4% in 38 patients without rashes (p=0.009; Fig. 1A). The 3-year OS was 85.5% in patients who developed drug rashes and 54.0% in those without rashes (p=0.02; Fig. 1B). Among iBCL only, excluding MCL, the 3-year PFS and OS were 83.2% and 95.7% in 24 patients with drug rashes, and 46.9% and 56.2% in 31 patients without rashes, respectively (p=0.0166 and 0.0238, respectively; Fig. 2A and 2B). The graphs for FL-only and MCL-only patients are shown in Fig. 3 and 4, respectively. FL patients with rashes had a significantly better PFS and OS rates than those without rashes (82.5% vs. 56.9%, p=0.0413, Fig. 3A, 95.2% vs. 59.8%, p=0.0487, Fig. 3B).

Table 2

Characteristics of drug rash after bendamustine treatment

Drug rash development		27
Grade
		Grade 1		8 (29.6%)
		Grade 2		5 (18.5%)
		Grade 3		14 (51.9%)
Pattern of rash
		Generalized		14 (51.9%)
		Localized		13 (48.1%)
Onset time
		course 1		17 (63.0%)
		course 2		5 (18.5%)
		course 3		2 (7.4%)
		course 4		1 (3.7%)
		course 5		2 (7.4%)
Treatment
	Steroid topical application	10 (37.0%)
	Antiallergic drug	2 (7.4%)
	Steroid topical application and antiallergic drug	5 (18.5%)
	Steroid oral and topical application and antiallergic drug	9 (33.3%)
	Observation	1 (3.7%)

Fig. 1

Outcomes of patients with indolent B-cell lymphoma and mantle cell lymphoma relative to bendamustine-induced rash (A: progression-free survival, B: overall survival)

Fig. 2

Outcomes of patients with indolent B-cell lymphoma relative to bendamustine-induced rash (A: progression-free survival, B: overall survival)

Fig. 3

Outcomes of patients with follicular lymphoma relative to bendamustine-induced rash (A: progression-free survival, B: overall survival)

Fig. 4

Outcomes of patients with mantle cell lymphoma relative to bendamustine-induced rash (A: progression-free survival, B: overall survival)

The relationship between the time of appearance of eruption in BR (course 1 vs. course 2 or later) and treatment response and prognosis was as follows: overall response rate 88.2% vs. 100%, 3-year OS 78.1% vs. 66.6% (p=0.818) and 3-year PFS was 78% vs. 85.7% (p=0.405), and no significant difference was observed.

Patients who developed rashes after bendamustine typically had a history of allergy (p=0.0259), no previous lymphoma treatment (p=0.0147), localized stage (p=0.0398) or combined rituximab treatment (p=0.0197) compared with those without rashes (Table 3). Age, sex, pathological diagnosis of lymphoma, leukocyte counts, eosinophil counts, lymphocyte counts and cycles of therapy did not significantly differ (Table 3). The relationships between the clinical features and prognosis are shown in Table 4. In univariate analysis, iBCL (p=0.02), extranodal involvement (p=0.02), use of rituximab (p=0.01) and rash after bendamustine (p=0.009) were significant factors for PFS. In multivariate analysis, iBCL (p=0.02) and rash after bendamustine (p=0.03) were significant factors for PFS. In univariate and multivariate analysis, rash after bendamustine (p=0.02) was the only significant factor for OS. Patients with 4 or more courses of bendamustine + rituximab therapy and with rash had a 3-year OS of 84.7% vs. 70.6% (p=0.149) and a 3-year PFS of 77.0% vs. 45.3% compared with those without rash (p=0.116).

Table 3

Characteristics of patients who developed rash after bendamustine and those who did not

	Rash+ (n=27)	Rash- (n=38)	P-value
Age Median (range)	67 (41-84)	66 (47-84)	0.547
Sex Male	16	15	0.949
Female	11	23
History of allergy
Yes	4	0	0.0259
No	23	38
Diagnosis (WHO classification)
Follicular lymphoma	21	16	0.574
Mantle cell lymphoma	3	7
Marginal zone B-cell lymphoma	2	2
MALT lymphoma	0	2
Lymphoplasmacytic lymphoma	0	1
Small lymphocytic lymphoma	1	0
Disease status
Newly diagnosed	13	7	0.0147
Relapse/Refractory	14	31
Stage 1-2	8	2	0.0398
3-4	19	36
WBC (x103/μl)
Median (range)	5.12 (2.72-22.9)	5.06 (3.03-79.8)	0.8
Lymphocytes (x103/μl)
Median (range)	1.26 (0.368-19.8)	1.19 (0.33-70.3)	0.599
Eosinophils (x103/μl)
Median (range)	0.23 (0.03-0.58)	0.12 (0.02-0.48)	0.806
Extranodal involvement
Yes	10	24	0.0468
No	17	14
Cycles of therapy
Bendamustine 1~3	1	1	0.182
4~6	0	9
Bendamustine+Rituximab 1~3	4	4	0.918
4~6	22	24

WHO, World Health Organization; MALT, mucosa-associated lymphoid tissue; WBC, white blood cell

Table 4

Clinical features and outcomes of 65 patients with indolent B-cell lymphoma and mantle cell lymphoma treated by bendamustine

Variable		n	PFS				OS
			Univariate		Multivariate		Univariate		Multivariate
			HR (95% CI)	P	HR (95% CI)	P	HR (95% CI)	P	HR (95% CI)	P
Age	≥ 70 years	26	1.4 (0.60-3.3)	0.43			1.5 (0.58-3.8)	0.40
	< 70 years	39
Sex	Male	31	1.1 (0.44-2.5)	0.91			0.93 (0.36-2.4)	0.89
	Female	34
Disease	iBCL	55	3.1 (1.20-8.1)	0.02	3.1 (1.2-8.0)	0.02	2.8 (0.96-8.0)	0.06	2.6 (0.89-7.5)	0.08
	MCL	10
Status	Newly	20	3.8 (0.88-17)	0.07			2.1 (0.45-9.7)	0.35
	R/R	45
Stage	1-2	10	5.3 (0.71-40)	0.10			3.9 (0.52-29)	0.19
	3-4	55
ExtraN	Yes	34	3.2 (1.17-8.7)	0.02	1.9 (0.65-5.5)	0.23	3.1 (1.02-9.5)	0.05	2.3 (0.71-7.1)	0.17
	No	31
Allergy	Yes	4		0.99				0.99
	No	61
Therapy	Bendamustine	11	0.3 (0.12-0.75)	0.01	0.48 (0.18-1.27)	0.14	0.41 (0.16-1.11)	0.08
	BR	54
Rash	Present	27	0.23 (0.08-0.69)	0.009	0.29 (0.09-0.90)	0.03	0.24 (0.07-0.83)	0.02	0.24 (0.07-0.83)	0.02
	Absent	38

R/R; relapse/refractory, iBCL; indolent B-cell lymphoma; MCL, mantle cell lymphoma, ExtraN; extranodal involvement, BR; bendamustine+rituximab

DISCUSSION

This study revealed that after bendamustine treatment for iBCL and MCL, rash development was relatively frequent, and was a better prognostic factor for PFS and OS by multivariate analysis.

The incidence rate of skin and subcutaneous tissue injury was 46% in a pivotal phase 2 clinical trial of bendamustine monotherapy in Japan, and 1% of cases were grade 3 or higher. The onset rate by course was highest (36.2%) after 1 course but occurred at multiple times over 2–6 courses. This is consistent with our study, excluding the high frequency of grade 3 toxicity. Fourteen patients (51.9%) presented with grade 3 toxicity and generalized rash, and were administered prednisolone prophylactically after the rash developed from days 1–7 of bendamustine retreatment. Although rashes relapsed in 7 patients (50.0%), no exacerbation occurred and the bendamustine retreatment was completed. Two patients were followed by watchful waiting after rash development because their lymphomas were in partial remission; they have remained in remission for more than 1 year.

Hypersensitivity reactions associated with cancer chemotherapy include infusion reactions and immediate allergic reactions. Most infusion reactions and anaphylaxis with rituximab are thought to occur at the time of initial infusion and within 24 hours of initiation. Bendamustine-induced rash is considered to be a late-onset hypersensitivity and not due to an acute hypersensitivity reaction such as infusion reaction or anaphylaxis. Delayed hypersensitivity reactions were noted in 6 patients in this study. Bellón proposed T cell-mediated drug-specific immune response due to delayed type IV hypersensitivity reaction. Delayed hypersensitivity reactions have also been reported in patients who received brentuximab-vedotin plus bendamustine.

In our series, more rashes occurred with B-R therapy (26/54) than with bendamustine alone (1/11) (48.1% vs. 9.1%, p=0.0197). However, in the case of B-R therapy, the rash developed on days 3 to 27, which was different from the time of the infusion reaction of rituximab, and rash due to bendamustine was considered.

The rashes were generally not severe, and post-treatment recurrence was frequent but manageable with steroids. In our series, patients with drug rashes typically had a history of allergy, no previous lymphoma treatment, localized stages, extranodal involvement or combined rituximab treatment (Table 3). The ratio and the change in absolute value of absolute lymphocyte counts (ALC) at initial bendamustine administration and the onset of rash in rash-positive patients was not significantly different from those at initial bendamustine administration and final bendamustine administration in rash-negative patients (data not shown).

Cencini E et al. reported the association between a single nucleotide polymorphism of IL2 and skin rash in lymphoma patients treated with bendamustine and rituximab. The TT example reported more skin rashes after bendamustine + rituximab treatment than with GT/GG. However, they reported that they were not associated with the treatment response.

Uchida et al. analyzed 95 patients with non-Hodgkin’s lymphoma who received bendamustine alone or BR therapy and reported that no prior treatment was a significant factor in skin toxicity. In our study, 13/20 (65%) newly diagnosed patients developed rash and 14/45 (31%) relapsed/refractory patients developed rash, demonstrating a significant difference (p=0.0147). The newly diagnosed patients were chemotherapy-naive and the lack of immunosuppression was considered to be related to the development of rash.

Although this study comprised only 65 patients, patients with iBCL and MCL treated with bendamustine who developed rashes had a better 3-year OS and PFS than those without rashes. Several patients experienced continuous remission with watchful waiting after rash developed. To our knowledge, no studies have reported rashes after bendamustine administration in relation to the prognosis of lymphoma. Nishikori et al. found increased CD8+ lymphocytes in the peripheral blood of patients with delayed skin reactions after bendamustine administration. They suggested that inappropriate activation of CD8+ lymphocytes by latently infected pathogens may be one of the triggers of late-onset skin reactions caused by bendamustine. The activated CD8+ lymphocytes may suppress tumor cell growth, which requires further investigation. In addition, the high frequency of grade 3 skin rash may have been associated with better outcomes for iBCL in the current study.

Patients with grade 2 or higher rashes had a 62% response rate, and those without rashes had a 23% response rate in a phase II trial of lenalidomide, an immunomodulatory agent, for adult T-cell leukaemia-lymphoma. Lenalidomide reportedly exerts direct antitumoral effects, and improves T-cell-mediated and NK-cell-mediated anti-tumor immune responses., In solid tumors, epidermal growth factor receptor (EGFR)-mutant non-small cell lung cancer patients with grade 2 or higher skin rashes within 1 week after afatinib treatment alone had higher maximum tumor reduction effects than those without rashes (80% vs. 39%; p=0.077), suggesting that skin rash development is related to the therapeutic response to afatinib monotherapy.

This study has several limitations. It was a retrospective study and limited to single-center patient data. It is difficult to make a firm conclusion in a mixed group of histology and treatment phases. A study consisting of a homogenous group treated with bendamustine as first-line treatment is more suitable for this purpose.

In conclusion, this study suggested that bendamustine-induced skin toxicity is associated with a better prognosis among patients with iBCL.