BRIEF REPORT

Late-onset neutropenia following rituximab treatmentfor rheumatologic conditions

Gabriel S. Breuer & Michael Ehrenfeld & Itzhak Rosner &

Alexandra Balbir-Gurman & Devy Zisman &

Shirley Oren & Daphna Paran

Received: 29 December 2013 /Revised: 16 February 2014 /Accepted: 24 February 2014# Clinical Rheumatology 2014

Abstract Rituximab is a monoclonal antibody directedagainst the CD20 antigen on the surface of normal and ma-lignant B lymphocytes. Its use in autoimmune conditions israpidly expanding. Late-onset neutropenia (LON) is a well-recognized side effect of rituximab therapy in lymphomapatients. Only a small number of cases of LON have beenreported in patients with autoimmune disorders. The aim ofthis work is to review cases in Israel and to compare them topublished cases in the literature thus adding to the body ofknowledge regarding this unusual phenomenon. Members ofthe Israeli Rheumatology Association were encountered by e-mail, requesting reports of cases of LON after therapy withrituximab. Submitted cases were reviewed, with demo-graphics and clinical data collated and tabled. Current cases

were compared to previously published rheumatology cases.Twelve episodes of LON following rituximab therapy werereported. All patients were female with an average age of50 years (range 22–78). LON occurred at an average of155 days after therapy (range 71–330). The average leukocytecount was1,456 white cells, with an average of 413 neutro-phils (range 0–1,170 neutrophils). Three of the patientsunderwent bone marrow biopsies which showed white cellline maturation arrest with an increased number of lympho-cytes. No blasts were seen. Our results add support to thegrowing evidence that this adverse event usually follows abenign course and is not an absolute contraindication forrepeat treatment if required in the future. However, vigilanceis recommended with routine periodic blood counts,especially 5 months following rituximab administrationwhen the risk is expected to be the highest.

Keywords Autoimmune disease . Neutropenia . Rituximab

Introduction

Rituximab is a monoclonal chimeric antibody directed againstthe CD20 antigen on the surface of normal and malignant Blymphocytes [1]. It is used extensively in the treatment of Bcell malignancies and moderately to severely active rheuma-toid arthritis (RA) [2, 3] and has recently been FDA-approvedfor the treatment of anti-neutrophil cytoplasmic antibody(ANCA)-associated vasculitis [4, 5]. Its use in autoimmunediseases is rapidly expanding including off-label use for avariety of autoimmune conditions including the following:autoimmune cytopenias, systemic lupus erythematosus(SLE), and other B cell-mediated conditions [6–8]. It is attrac-tive as a therapeutic agent due to its broad efficacy andfavorable toxicity profile with no myelosuppression. Late-onset neutropenia (LON) is a well-recognized side effect of

G. S. Breuer (*)Hebrew University School of Medicine, Shaare Zedek MedicalCenter, P.O. Box 3235, 91031 Jerusalem, Israele-mail: gbreuer@szmc.org.il

M. EhrenfeldChaim Sheba Medical Center, Tel HaShomer, Ramat Gan, Israel

I. RosnerBnai Zion Medical Center, Haifa, Israel

I. RosnerFaculty of Medicine, Technion, Haifa, Israel

A. Balbir-GurmanRambam Health Care Campus, Haifa, Israel

D. ZismanCarmel Medical Center, Haifa, Israel

S. OrenRabin Medical Center, Petah Tikva, Israel

D. ParanTel Aviv Sourasky Medical Ctr, Tel Aviv, Israel

Clin RheumatolDOI 10.1007/s10067-014-2562-x

rituximab therapy in lymphoma patients, defined as unex-plained grade III (0.5–1×109) or grade IV (<0.5×109/L) neu-tropenia occurring at least 4 weeks after the last dose ofrituximab in the absence of an alternative explanation. Itsincidence in the treatment of B cell malignancies is estimatedto be 3 to 27% [9, 10]. However, only a small number of casesof LON have been reported in patients with autoimmunedisorders. A case review analysis from a single center sug-gested an incidence of 6 % [11]. Tesfa et al. recently reportedan overall rate of LON of 5 % in a retrospective study of 209patients with various rheumatologic diseases treated withrituximab and suggested that the incidence varies with auto-immune disease type [12]. The true incidence of this adverseevent in autoimmune diseases, predisposing factors, andmechanisms are still poorly defined. In addition, there havebeen conflicting reports regarding the clinical significance ofthis phenomenon in autoimmune disease.

We performed a national survey of all practicing rheuma-tologists and identified 12 episodes of LON in 11 patientsoccurring after rituximab therapy for an autoimmune disease.We describe the clinical features and course of these cases andreview the literature regarding the characteristics of this ad-verse event as well as the associated risk for infection and theoutcome of this complication.

Methods

The Israeli RheumatologyAssociation includes 150members.All members were contacted by e-mail, to request reports ofcases of LON following therapy with rituximab. In addition todemographic information, collected data included the rheu-matologic diagnosis, current therapy, and number of previousrituximab treatments. Each case was reviewed regarding thelevel of neutropenia, LON-related complications, action takenfollowing neutropenia, and ensuing therapy for the manage-ment of the rheumatologic condition. A literature search wasconducted using PubMed with the following terms: late onsetneutropenia, LON, rituximab, and rheumatology, looking forall reported cases of LON following rituximab treatment forrheumatologic diseases published in the English literature.Our cases were compared to previous published cases inrelation to clinical aspects of the condition.

Results

Twelve episodes of LON following rituximab therapy werereported. All patients were female with an average age of50 years (range 22–78). Ten of the patients were treated withrituximab for rheumatoid arthritis (RA), one patient for mixedconnective tissue disease (MCTD), and one for systemic lupuserythematosus (SLE). In seven cases, patients were treated

with rituximab for the first time. One patient was treated forthe second time; for two patients, this treatment was their thirdtreatment; one patient was treated for the fourth time, and inone case, rituximab was administered for the seventh time(Table 1). In all cases, rituximab treatment resulted in a suc-cessful clinical outcome with control of the rheumatic disease,including the three cases where rituximabwas re-administeredafter the occurrence of LON.

None of the patients received rituximab as monotherapy.Additional treatments included the following: methotrexate(MTX) in seven cases; prednisone in eight cases;hydroxychloroquine in three cases; and leflunomide, cyclo-sporine, cyclophosphamide, azathioprine, and colchicine, inone patient each. LON occurred at an average of 155 daysafter therapy (range 71–330). The average leukocyte countwas1,456 white cells, with an average of 413 neutrophils(range 0–1,170 neutrophils). Six cases were treated with gran-ulocyte colony-stimulating factor (G-CSF), and all individualswho were treated with methotrexate had their treatmentdiscontinued. In terms of associated complications, one pa-tient suffered from arm cellulitis and two patients had fever upto 39 °C with no other signs of infection. Nine of the 11patients had no signs of infection.

Three of the patients underwent bone marrow biopsieswhich showed white cell line maturation arrest with an in-creased number of lymphocytes. No blasts were seen.

Rituximab was re-administered in three cases, one ofwhom developed recurrent LON (Table 1).

Discussion

In this case series, LON occurred in patients treated for rheu-matic diseases an average of 155 days after therapy, similar toprevious reports in both hematologic and rheumatologic dis-eases [10, 12]. In our cohort, 82 % of the patients carried adiagnosis of RA, while only one patient suffered from SLEand one from MCTD. Our findings differ from those reportedby Tesfa et al. where LON occurred more frequently ingranulomatosis with polyangiitis (GPA) and SLE patientsthan in patients with RA. Tesfa et al. suggest that theincidence of LON depends on the clinical setting in whichrituximab is administered, similar to what has been report-ed in hematologic patients [10, 12]. In our series, we didnot corroborate this finding, with the majority of patientsin our series having RA, possibly due to the fact thatrituximab has only recently been approved in Israel for thetreatment of ANCA-associated vasculitis and is used forthe treatment of SLE only as off-label treatment in severerefractory cases. In this series, all cases were females, perhapsreflecting the dominance of RA patients in this series.The size of this series does not allow conclusions regardingthe role of gender in the occurrence of LON. Although LON

Clin Rheumatol

has been considered a self-limiting adverse event, some re-ports have emphasized the increased risk of infection [12]while others suggest the infection risk is low and similar tothat reported in hematologic patients treated with rituximab[10]. In our series, as opposed to the report by Tesfa et al. [12],the risk of infection was low with only one case of infectionwith arm cellulitis which was resolved with a short course ofIVantibiotics and two cases of fever with no overt evidence ofinfection. The course of LONwas benign in all cases reviewedin our series and responded promptly to treatment with G-CSFin five cases or resolved spontaneously in the remaining cases.The lower incidence of infection might be attributed to thehigh percentage of RA patients in our series as opposed to thelarger number of GPA patients in the study by Tesfa et al. [12].GPA patients receiving rituximab usually suffer from amultisystemic disease, including renal failure in some cases,and require additional immunosuppression, including gluco-corticoids which may predispose to infection. This suggeststhat the higher risk of infection in GPA may be attributedchiefly to these additional factors, with a possibly less signif-icant causative role for the episode of LON. An additionalfactor might be the fact that all our patients were treated in anoutpatient setting which reduces the risk of infection.

Our findings suggest that in most cases, LON can bemanaged with a single dose of G-CSF. LON apparently doesnot put the affected patient at significant additional risk;however, additional predisposing factors for infection mayplay an important role in the risk for development/severityof infection. Besada et al. suggest that the risk of LONrecurrence after re-challenge is low with no recurrences inthe six re-treated patients in their cohort [11]. In our series,rituximab was re-administered in three cases, one of whomhad recurrent LON with no evidence of infection. The under-lying disease in this patient was RA as well, with MTX as thesole concomitant antirheumatic drug, possibly explaining thebenign course. Interestingly, Hincks et al. proposed that thedevelopment of rituximab-associated LONmay even be a goodprognostic indicator in lymphoproliferative conditions [13].

With more widespread use of rituximab in rheumatologicconditions as well as an increasing number of patients receivingrepeated courses of rituximab, additional experience with LONis expected to increase and possibly allow prediction of patientsat increased risk for this adverse event as well as aid in formingguidelines for follow-up and treatment. Indeed, Tesfa et al. havesuggested that more pronounced and sustained B lymphocytedepletion and lower IgM levels may be risk factors for LON.

Table 1 Demographic and clinical data for 12 episodes of LON following rituximab in comparison to published cases

Age/gender

Diagnosis No. ofRTXcourses

Concomitantmedications

Days sinceRTX course

Neutrophils(WBC)

RepeatRTXcourses

Infection Action taken Bonemarrowbiopsy

32/F RA 4 MTX, PRED 151 110 (1,100) Yes No G-CSF

34/F MCTD 1 CYC, AZA, PRED 105 420 (1,000) Yes No G-CSF

22/F SLE 1 CYC 240 0 No Cellulitis G-CSF +

38/F RA 1 MTX, PRED 120 0 No No G-CSF +

68/F RA 1 PRED, diclofenac 330 635 (1,270) No No G-CSF

50/F RA 1 MTX, HCQ 138 1,081 (2,300) Yes No MTX D/C

51/F(same patient thefollowing year)

RA 2 HCQ 180 1,170 (2,600) No No None

78/F RA 1 MTX, PRED 150 702 (1,800) No No MTX D/C, increasePRED

+

52/F RA 3 MTX 115 500 (2,400) No No None

66/F RA 3 leflunomide, HCQ,PRED

105 100 (1,500) No Fever, noinfection

G-CSF, tazobactam

85/F RA 1 PRED, colchicine 71 390 (1,500) No No None

50/F RA 6 MTX 150 2,000 (860) No Fever, noinfection

G-CSF

Published casesaverage age: 5217 (60 %) females8 (40 %) males

RA=12GPA=12SLE=3vasculitis=1JRA=1

NA MTX=8MMF=4

148 380 NA Sepsis=5Fever=3

G-CSF=11

All patients were treated with two doses of rituximab of 1,000 mg each, administered 2 weeks apart

MTX methotrexate, PRED prednisone, CYC cyclophosphamide, AZA azathioprine, HCQ hydroxycholroquine, MMF mycophenolate mofetil, D/Cdiscontinued, RA rheumatoid arthritis, GPA granulomatosis with polyangiitis, SLE systemic lupus erythematosus, JRA juvenile rheumatoid arthritis, NAnot applicable

Clin Rheumatol

Our results add support to the growing evidence that thisadverse event usually follows a benign course and is not anabsolute contraindication for repeat treatment if required in thefuture. However, vigilance is recommended with routine pe-riodic blood counts, especially 5 months following rituximabadministration when the risk is expected to be the highest. Werecommend assessment of the infectious risk in individualpatient, taking into account immunoglobulin levels, the un-derlying rheumatologic disease, and additional immunosup-pressive treatment.

Conflict of interest None.

References

1. Coiffier B (2007) Rituximab therapy in malignant lymphoma.Oncogene 26:3603–3613

2. Edwards JCW, Szczepaniski L, Szechinski J et al (2004) Efficacy ofB-cell-targeted therapy with rituximab in patients with rheumatoidarthritis. N Engl J Med 357:2572–2580

3. Leandro MJ, Cambridge G, Ehrenstein MR, Edwards JC (2006)Reconstitution of peripheral blood B cells after depletion withrituximab in patients with rheumatoid arthritis. Arthritis Rheum54:613–620

4. Stone JH, Merkel PA, Spiera R, Seo P, Langford CA, Hoffman GS,Kallenberg CG, St Clair EW, Turkiewicz A, Tchao NK, Webber L,Ding L, Sejismundo LP, Mieras K, Weitzenkamp D, Ikle D,

Seyfert-Margolis V, Mueller M, Brunetta P, Allen NB, FervenzaFC, Geetha D, Keogh KA, Kissin EY, Monach PA, Peikert T,Stegeman C, Ytterberg SR, Specks U, RAVE-ITN Research Group(2010) Rituximab versus cyclophosphamide for ANCA-associatedvasculitis. N Engl J Med 363(3):221–232

5. Jones RB, Tervaert JW, Hauser T, Luqmani R, MorganMD, Peh CA,Savage CO, Segelmark M, Tesar V, van Paassen P, Walsh D, WalshM, Westman K, Jayne DR (2010) Rituximab versus cyclophospha-mide in ANCA-associated renal vasculitis. N Engl J Med 363(3):211–220

6. Berentsen S (2007) Rituximab for the treatment of autoimmunecytopenias. Haematologica 92:1589–1596

7. Carr DR, Heffernan MP (2007) Off-label uses of rituximab in der-matology. Dermatol Ther 20:277–287

8. Dörner T, Radbruch A, Burmester GR (2009) B-cell-directed thera-pies for autoimmune disease. Nat Rev Rheumatol 5(8):433–441

9. Lemieux B, Tartas S, Traulle C et al (2004) Rituximab-related late-onset neutropenia after autologous stem cell transplantation for ag-gressive non-Hodgkin’s lymphoma. Bone Marrow Transplant 33:921–923

10. Wolach O, Bairey O, Lahaw M (2010) Late-onset neutropenia afterrituximab treatment. Case series and comprehensive review of theliterature. Medicine 89:308–318

11. Besada E, Koldingsnes W, Nossent J (2012) Characteristics of lateonset neutropenia in rheumatologic patients treated with rituxi-mab: a case review analysis from a single center. QJM 105(6):545–550

12. Tesfa D, Ajeganova S, Hägglund H, Sander B, Fadeel B, Hafström I,Palmblad J (2011) Late-onset neutropenia following rituximab ther-apy in rheumatic diseases: association with B lymphocyte depletionand infections. Arthritis Rheum 63(8):2209–2214

13. Hincks I, Woodcock BE, Thachil J (2011) Is rituximab-induced late-onset neutropenia a good prognostic indicator in lymphoproliferativedisorders? Br J Haematol 153:411–413

Clin Rheumatol