Abstracts and Commentary

Optimizing the Activityof Rituximab

Abstract #1398

Rituximab Therapy for PatientsWith Newly Diagnosed,Asymptomatic Advanced-StageFollicular Grade I Non-Hodgkin'sLymphoma: A Phase II Trial in theNorth Central Cancer Treatment Group

T. E. Witzig, A. M. Vukov, T. M. Habermann, S. Geyer, W. R.Friedenberg, W. L. White, M. Salim, P. J. Flynn, T. R. Fitch,R. F. MortonHematology, Mayo Clinic, Rochester, Minnesota; OHACI,Peoria, Illinois; Medical Oncology/Hematology, GuthrieClinic, Milan, Pennsylvania; Medical Oncology, Allan BlairCancer Center, Regina, Saskatchewan, Canada; OncologicConsultants, PA, Minneapolis, Minnesota; Hematology/Oncology,Mayo Clinic, Scottsdale, Arizona; Medical Oncologyand Medical Association, Iowa Oncology Research AssociationCCOP, Des Moines, Iowa

Patients with newly diagnosed advanced-stagefollicular grade I non-Hodgkin's lymphoma (NHL)are often asymptomatic and can be observed withoutimmediate chemotherapy without compromising survival.Rituximab (Rituxan) immunotherapy offers anew nonchemotherapy approach to this group ofpatients.Our goal was to assess the rate of complete remission(CR) and partial remission (PR) to rituximab andto determine the time to progression (TTP) and timeto subsequent chemotherapy (TTSC). Eligible patientshad biopsy-proven follicular grade I NHL,measurable disease, stage III or IV, no prior therapy,CD20+, and performance status (PS) 0-2. Patientsreceived rituximab 375 mg/m

²

IV weekly * 4 doses.No maintenance therapy was provided. Patients werefollowed with physical exam, laboratory tests, andCT scans every 3 months for the first year and every6 months thereafter.A total of 37 patients were accrued to this studybetween July 1999 and May 2001. Forty-three percentof patients were male; 89% were PS 0 and 11%were PS 1. The median age was 59 years (range: 29to 83 years). All patients were white except for 1patient, who was Asian. Sixteen percent of patientshad elevated lactate dehydrogenase levels pretreatment.Thirty-eight percent of patients had a lowInternational Prognostic Index (IPI), 46% low intermediateIPI, and 16% high intermediate IPI. Nopatients had B symptoms. All patients completedtreatment as per protocol.The overall response rate (ORR) was 61%, with25% CR. To date, 20 patients remain in unmaintainedremission, 2 died without disease progression, 3 diedof disease progression, and 12 have progressed andare still alive. The median TTP was 1.67 years (range:0.014 to 3+ years); and the TTSC was 1.75 years(range: 0.014 to 3+ years). In general, rituximab waswell tolerated in this patient population. Severe (grade3+) toxicity was seen in 4 patients (11%), where 2patients (5%) had severe hematologic toxicity and 3patients (8%) had severe nonhematologic toxicity.These severe toxicities included rash (2), neutropenia(1), leukopenia (1), infection without neutropenia (1),and urticaria (1).CONCLUSION: Rituximab can be safely administeredto patients with previously untreated advancedstagefollicular grade I NHL with minimal toxicity.This therapy produces an ORR of 61% and prolongedremissions have been documented. Rituximab offersan acceptable alternative to observation in this patientpopulation.

Abstract #604

Prolonged Treatment With RituximabSignificantly Improves Event-FreeSurvival and Duration of Responsein Patients With Follicular Lymphoma:A Randomized SAKK Trial

M. Ghielmini, S.-F. H. Schmitz, S. Cogliatti, G. Pichert,M. Fey, D. Betticher, G. Martinelli, F. Peccatori, U. Hess,R. Stahel, E. Zucca, R. Stupp, T. Kovacsovics, C. Helg,A. Lohri, M. Bargetzi, D. Vorobiof, T. CernyLymphoma Working Group, Swiss Group for Clinical CancerResearch (SAKK), Bern, Switzerland

Clinical and pharmacokinetic data suggest that theeffect of rituximab (Rituxan) could be improved byprolonged exposure to the drug. To verify this hypothesiswe performed a randomized trial of rituximab givenat the standard schedule vs a prolonged rituximabtreatment with additional maintenance cycles.A total of 202 patients with newly diagnosed orresistant/relapsed follicular lymphoma (FL) wereenrolled. After a standard induction treatment withrituximab (375 mg/m

²

weekly * 4), patients respondingor with stable disease at week 12 were randomizedto either observation or maintenance therapy with anadditional rituximab administration (375 mg/m

²

) every8 weeks (at months 3, 5, 7, and 9).In 185 evaluable patients, response to inductiontherapy was 46% (59/128) in pretreated patients and67% (38/57) in chemotherapy-naive cases (

P

< .01).Adverse predictors of response were bulky disease(

P

= .002) and male gender (

P

= .04). The characteristicsof the 151 patients randomized to observationvs maintenance therapy were as follows: median age,57; performance status 0/1, 97%; stage III/IV, 85%;bone marrow involvement, 52%; elevated lactatedehydrogenase, 30%; previous radiotherapy, 18%;previous chemotherapy, 66% (mean 2.3 regimens);all characteristics were equally balanced betweenboth arms.At a median follow-up of 35 months, the medianevent-free survival from the initiation of inductiontreatment was 12 months for the observation arm vs23 months for the maintenance arm (

P

= .02), with ahazard ratio of 0.57, the difference being more pronouncedin the subgroup of chemotherapy-naive patients(19 vs 36 months,

P

= .009). Among patientsresponding at week 12, 56% were still in remission at12 months in the observation arm vs 80% in themaintenance arm (

P

= .01). The median responseduration was 17 vs 36 months (

P

= .0009), with ahazard ratio of 0.45.After induction treatment, 34 patients further improvedthe quality of response, and complete remissionrate increased from 10% at randomization to16% at month 7, 23% at month 12, and 29% (44/151)at further follow-up (no difference between the twoarms). The number of circulating B lymphocytestended to return to pretreatment levels after 1 year ofobservation, while they remained low in patientsreceiving prolonged treatment (65% vs 33% of baselinelevel,

P

= .04). However, in both arms the IgG,IgA, and IgM plasma levels remained stable and theincidence of infections and other side effects was low.CONCLUSION: The study shows that in patientswith FL, stable or responding after standard rituximabtreatment, the administration of an additionalfour single doses of rituximab at 2-month intervalsreduces the risk of progression or relapse by 43%among all randomized patients and by 55% amongresponders, as well as significantly improves thechance of remaining in remission at 1 year, withoutinducing additional toxicity. This effect is obtainedmainly by the capacity of maintenance treatment toprevent relapse in responding patients. The antitumoreffect of rituximab continues long after the end of itsadministration.

Abstract #1390

Rituximab Re-Treatment in B-Cell LymphomaPatients: Efficacy and Toxicityin 59 Patients Treated in One Center

B. Coiffier, F. Bouaffia, C. Thieblemont, O. Hequet,P. Arnaud, C. Dumontet, D. Espinouse, G. SallesDepartment of Hematology, CH Lyon-Sud, HospicesCivils de Lyon, Pierre-Bnite, France

During the past 4 years,483 patients with B-cell lymphomawere treated in ourdepartment with rituximab(Rituxan) alone (166 patients),in combination withchemotherapy (302 patients),or as maintenance therapy(15 patients) in different studiesor as registered treatment.Out of these 483 patients,429 (89%) responded to treatment (complete response[CR] + partial response [PR]): 132/166 (80%)to rituximab alone and 282/302 (93%) to a combinationwith chemotherapy. Out of these 429 patients,117 (27%) had presented a disease progression attime of analysis (median follow-up of around 18months), 54/132 (41%) after rituximab alone, and59/282 (21%) after rituximab in combination withchemotherapy. Out of these 117 patients, 59 (50%)received a second treatment with rituximab alone orin combination with chemotherapy.Rituximab was administered alone at the dose of375 mg/m

²

/wk for 4 weeks or in combination withchemotherapy the same day at 375 mg/m

²

for allchemotherapy cycles. These 59 patients were analyzedfor the response to the second rituximab treatment.For the first treatment, rituximab was givenalone in 30 patients and associated with chemotherapyin 29 patients and for the second treatment, in 32and 27 patients, respectively. All but 4 patientsresponded to this second treatment (CR 42%, PR51%). The median time to progression and time tonext treatment were 12 and 13 months after the firstrituximab treatment and 20 and 21 months after thesecond rituximab treatment, respectively. Eight patientshad a shorter time to progression after thesecond rituximab cycle and in 4 patients it wasidentical.Progression and time to progression after the secondtreatment were not associated with the number oftreatments before rituximab, the association or notwith chemotherapy for the first or second rituximabcycle, the response to the first or second rituximabtreatment, or the status at time of second rituximabtreatment (progressive disease or partial response toprevious cycle). Of the 20 patients who progressedafter the second rituximab treatment, 12 received athird rituximab cycle-6 alone and 6 with chemotherapy;all 12 responded to the treatment and only 3 hadprogressed at time of analysis with a median time toprogression of 13 months. No special adverse eventswere observed with the different cycles of rituximabbut 3 patients had herpes zoster infection during orjust after the second cycle.

CONCLUSION: In conclusion, re-treatment withrituximab alone or in combination with chemotherapywas associated with very good response rates anda longer progression-free survival after the secondrituximab treatment than after the first one. Rituximab,alone or in combination with chemotherapy, isrecommended for re-treatment of patients who progressedafter a first rituximab treatment.

Abstract #3073

Rituximab Plus Chlorambucil in Low-Grade Non Hodgkin's Lymphomas:Clinical Results of a Phase II Study

G. Martinelli, D. Laszlo, P. Mancuso, P. Santoro,G. Pruneri, A. Vanazzi, E. ZuccaHematoncology Division, European Institute of Oncology,Milan, Italy; Medical Oncology Department, OncologyInstitute of Southern Switzerland, Bellinzona, Switzerland

The monoclonal anti-CD20 antibody rituximab(Rituxan) has a well-establishedefficacy in thetreatment of low-grade non-Hodgkin's lymphomas(NHLs), principally the follicularsubtype (60%). Morerecently, the combination ofchemotherapy plus rituximabhas been demonstrated in arandomized study of aggressive NHL patients to producebetter clinical results than chemotherapy alone.Chlorambucil (Leukeran) is considered the drug ofchoice for the treatment of low-grade NHL, mainlyfor follicular: chlorambucil alone induces a responserate of 70% with 30% to 40% of complete remission.No data are available about the combination ofchlorambucil plus rituximab. The aim of this study isto define the feasibility, toxicity, and efficacy ofrituximab plus chlorambucil in low-grade NHL.Twenty-nine NHL patients (20 follicular, 6 marginalzone, 3 chronic lymphocytic leukemia), medianage 40 years, were enrolled in the study betweenNovember 2001 and April 2002. Fourteen patientshad received previous therapy. The treatment planconsisted of chlorambucil at 6 mg/m

²

daily administeredfor 6 consecutive weeks combined with the standard4-weekly rituximab at 375 mg/m

²

administrationschedule. Patients responding to this induction phasereceived further therapy with chlorambucil (2 weeksevery month for four additional cycles) in combinationwith four monthly administrations of rituximab.All patients were monitored weekly with whiteblood cell counts during induction phase and thenonce monthly. Peripheral blood immunophenotypeevaluation was performed at baseline, after the inductionphase, and then at the end of the treatment.Clinical evaluation was performed after the inductionphase, then at the end of the entire therapy program.All patients are evaluable for toxicity and response.The most important hematologic toxicity observedwas National Cancer Institute Common ToxicityCriteria (NCI-CTC) grade 3/4 neutropenia, experiencedby 13% of patients. Only one patient developedsevere anemia and thrombocytopenia. All hematologictoxicity recorded was observed among pretreatedpatients. Twelve patients developed an absolutelymphopenia (NCI grade 3), and significant CD4+reduction; however, no patients had an imbalance inCD4+/CD8+ ratio and no severe infections wererecorded. Among 27 patients evaluable for response,24 (89%) achieved an objective response with approximately53% complete remissions.CONCLUSION: The combination of rituximabplus chlorambucil is well tolerated and active. Clinicalresults achieved seem to confirm the possiblesynergistic or additive effect of the combination, also inlow-grade NHL patients. A randomized trial is nowongoing to further evaluate this promising combination.

Abstract #2246

Phase II Study With Fludarabine andCyclophosphamide Plus Rituximab inRelapsed Follicular Lymphoma Patients

S. Sacchi, A. Tucci, F. Merli, L. Orsucci, G. Cervetti,U. Occhini, M. Liberati, G. Tarantini, V. Callea,M. Brugiatelli, V. M. Lauta, L. Baldini, S. Luminari,M. FedericoDipartimento di Oncologia ed Ematologia, Universit diModena e Reggio Emilia, Modena, Italy; Servizio di Ematologia,Spedali Civili, Brescia, Italy; Servizio di Ematologia,Az. Osp. Santa Maria Nuova, Reggio Emilia, Italy; Divisionedi Ematologia, Az. Osp. Santo Giovanni Battista, Torino, Italy;Divisione di Ematologia, Az. Osp. Pisana Ospedale SantaChiara, Pisa, Italy; Servizio di Ematologia, Az. Osp. Arezzo,Arezzo, Italy; Istituto di Scienze Oncologiche, PoliclinicoMonteluce, Perugia, Italy; Divisione di Ematologia, Osp.Civ. Santa Nicola Pellegrino, Trani, Italy; Divisione di Ematologia,P. Osp. Riuniti "Bianchi, Melacrino, Morelli," ReggioCalabria, Italy; Divisione di Ematologia, Az. Osp. "Papardo,"Messina, Italy; Sezione di Medicina Interna ed OncologiaClinica, Az. Osp. Policlinico, Bari, Italy; Divisione di Ematologia,Ospedale Maggiore, IRCCS, Milano, Italy

Both cyclophosphamide (Cytoxan, Neosar) andfludarabine (Fludara) have individual antilymphomaactivity, and the combination of rituximab (Rituxan)plus fludarabine has been shown to have synergisticactivity against resistant lymphoma cell lines in vitro.In March 2000, we started a phase II multicenterclinical trial to test the safetyand efficacy of the fludarabine/cyclophosphamide plusrituximab combination in patientswith relapsed follicularlymphoma. We haverecently completed the enrollmentof 48 patients in thetrial. Patients received fourdoses of rituximab (375 mg/m

²

/d) in combination with 4cycles of fludarabine (30 mg/m

²

/d for 3 days) andcyclophosphamide (300 mg/m

²

/d for 3 days) every 21days.Patient characteristics: 45% males, 55% females;median age: 62 years (range: 44-71); stage IV, 47%patients; systemic symptoms, 8.5% of patients; elevatedlactate hydrogenase, 15% of patients. Rearrangementof the bcl-2 gene was evaluated withpolymerase chain reaction in 38 patients (79%) onbone marrow or peripheral blood mononucleatedcells; 22 patients showed bcl-2 rearrangement (58%).Median number of previous chemotherapy regimenswas 1.7 (range: 1-4); median duration of last remissionbefore registration into the trial was 12 months(range: 1-68 months).Thirty-nine patients were available for evaluationat the time of current analysis. One patient did notcomplete therapy due to severe cytopenia. The responserate in the intent-to-treat analysis was 97%,with 74% complete responses and 23% partial responses.Out of 18 patients with molecular monitoringof disease before and after chemotherapy, 15patients (83%) obtained molecular remission in thebone marrow.After a median follow-up of 12 months (range: 1-25 months), median duration of remission was 13months; 20% of patients were only recently registeredinto the trial and had a short follow-up (less than4 months). Overall, 8 patients relapsed (21%). Onepatient died due to severe neutropenia that occurredduring chemoterapy; two patients died due to lymphomaprogression. Complete responses are ongoingin 28 patients.As far as toxicity is concerned, World HealthOrganization (WHO) grade III/IV leukopenia wasobserved in 10 patients, with 4 patients presenting aWHO grade IV granulocytopenia. WHO grade IVinfections were observed in only one patient, who hada pulmonary infection after the third cycle. Nonhema-tologic toxicity was minimal. During follow-up, onepatient had a renal cell tumor and one patient hadmyelodysplasia, 3 and 6 months after the end oftreatment, respectively.CONCLUSION: The combination of fludarabine/cyclophosphamide plus rituximab is associated withacceptable toxicity and an excellent response rate inthis group of heavily pretreated patients with relapsedfollicular lymphoma. Further follow-up is required toevaluate response duration and survival in the wholegroup of patients.

Abstract #1401

Combined Fludarabine, Mitoxantrone,and Rituximab Achieves a HighResponse as Initial Treatment forAdvanced Low-Grade Non-Hodgkin'sLymphoma

S. A. Gregory, P. Venugopal, S. Adler, S. Enschede,F. Yunus, T. M. O'Brien, K. F. O'Donnell, E. L. RobinSection of Hematology, Rush-Presbyterian-St. Luke's MedicalCenter, Chicago, Illinois; Boston Baskin Cancer Group,Memphis, Tennessee; The Community Hospital of Munster,Munster, Indiana

This phase II study wasundertaken to evaluate thesafety and efficacy of fludarabine(Fludara) plus mitoxantrone(Novantrone)followed by consolidationwith the anti-CD20 monoclonalantibody rituximab intreating low-grade non-Hodgkin's lymphoma. Thisregimen uses the synergisticeffect of fludarabine and mitoxantrone observed invitro, and combines it with a third agent with acomplementary mode of action and proven activity inlow-grade lymphoma.For eligibility, patients must have previously untreatedCD20-positive low-grade non-Hodgkin's lymphoma,with a Karnofsky performance status of 70.Treatment consists of 4 to 6 cycles of chemotherapy,which includes fludarabine at 25 mg/m

²

on days 1-3plus mitoxantrone at 12 mg/m

²

on day 1 of each 28-day cycle. If patient attains a complete remission(CR) after 4 cycles of chemotherapy, rituximab isadministered at 375 mg/m

²

weekly for 4 weeks, 4-6 weeks after the last dose of chemotherapy. If CR isnot attained after 4 cycles, the patient receives 2 morecycles of chemotherapy and then proceeds to rituximabtherapy after a 4-6 week delay. The totalnumber of patients registered is 41; 33/41(80%) werestage IV, 3 (8%) stage III, and 5 (12%) stage II.The histologic classification of the 41 patients is asfollows: 11(27%) small lymphocytic lymphoma, 18(44%) follicular small cleaved, 6 (15%) follicularmixed, 2 (5%) mantle cell lymphoma, 3 (7%) maltoma,and 1 (2%) marginal zone. Ten patients wereinevaluable for the following reasons: 1 has notcompleted therapy, 2 ineligible, 3 patient removal, 1autoimmune hemolytic anemia, 1 death not related totreatment, and 2 patients were removed due to prolongedcytopenia.Of 31 patients evaluated for response, 30 responded.The overall response rate is 97%(n = 30) (CR45%, n = 14, and PR 52%, n = 16); 50% (n = 7) ofpatients in CR have remained in remission longerthan 24 months (24+, 24+, 30+, 24+, 34+, 24+, 24+months). 31% (n = 5) of patients in PR have remainedin remission for longer than 12 months (12+, 15+,15+, 18+, 28+ months). 30% (n = 9) of patients haveprogressed after achieving a remission (3+, 3+, 5+,7+, 7+, 11+, 13+, 13+, and 25+ months).Serious adverse events included grade 3 and 4neutropenia (n = 64 episodes in 163 cycles of chemotherapy),neutropenic fever (n = 7), shortness ofbreath (n = 2), one epsiode of retinal detachment, andone episode of thrombus formation. Three patientshad 25% dose adjustments for prolonged neutropenia.Two patients were discontinued due to neutropenia.Grade 1/2 neutropenias, anemias, andthrombocytopenias usually resolved during or beforerituximab infusion. Documented infections were rarewith one episode of pneumonia. Antimicrobial anaphylaxiswas not used.CONCLUSION: A regimen of fludarabine plusmitoxantrone followed by rituximab is highly effectiveand well tolerated when used as primary therapyin advanced-stage low-grade non-Hodgkin's lymphoma.Moderate to severe neutropenia occurredduring therapy, but documented infection occurred inonly one patient and neutropenic fevers in sevenpatients. This treatment results in a high rate ofoverall response and durable remissions.

Abstract #1392

High Clinical and MolecularResponse Rate in Elderly PatientsWith Advanced-Stage FollicularLymphoma Treated at DiagnosisWith a Brief ChemoimmunotherapyFND Plus Rituximab

U. Vitolo, C. Boccomini, M. Ladetto, D. Rota-Scalabrini,E. Pogliani, L. Rigacci, A. Carpaneto, A. M. Liberati,M. Bertini, P. Pregno, E. Larizza, C. Castellino, F. Rossini,M. Astolfi, B. Botto, A. Tonso, S. Morandi, C. Tarella,M. Aglietta, E. GalloOn behalf of Gruppo Mutliregionale Per I NHL; GruppoToscano, Hematology ASO S. Giovanni Battista andUniversity, Torino, Italy

The addition of rituximab (Rituxan) to fludarabine(Fludara)-containing regimens may enhance clinicalactivity, yet few data are available on molecularresponse after this combination.From March 1999 to June 2002, 59 elderly patients(age > 60) with advanced-stage follicular lymphomaat diagnosis were enrolled into this study. Treatmentconsisted of 4 courses of FND (fludarabine 25 mg/m

²

days 1-3, mitoxantrone [Novantrone] 10 mg/m

²

day1, dexamethasone 20 mg days 1-3) followed by 4rituximab infusions at 375 mg/m

²

/wk. Patients inpartial remission after this treatment received 2 additionalFND courses and 2 rituximab infusions. Polymerasechain reaction (PCR) molecular monitoringfor the presence of IgH/bcl-2 and/or Ig heavy-chaingene rearrangement was performed at the beginningof the treatment, after FND, after rituximab, andduring follow-up time on bone marrow samples.Median age was 66 years (range: 60-78 years); 30males and 29 females; 13% had stage II, 14% stageIII, and 73% stage IV disease; 66% had bone marrowinvolvement; 47% had bulky disease and 25% had≥ 2 extranodal sites. To date, 36 patients are evaluable.Clinical response to FND was complete response(CR) 31%, CR unconfirmed (CRu) 17%,partial response (PR) 47%, and no response (NR) 5%.Clinical response at the end of the whole treatmentprogram was CR 76%, CRu 14%, PR 5%, and NR5%. With the addition of rituximab, complete response(CR plus CRu) increased from 48% to 90%.Seventy-eight percent of the CRs were achieved witha very brief chemoimmunotherapy (4 FND plus 4rituximab). With a median follow-up of 18 months,3/32 patients relapsed.To date, a molecular marker of disease was detectablein 18 patients. A PCR-negative status wasachieved in 6/18 patients after FND, while PCRnegativity was observed in 15/18 patients after rituximabtreatment; all 15 patients were in clinical CR.Fifteen patients had further molecular monitoring at6 or 12 months after the end of therapy: 13 PCRnegative and 2 PCR positive. Among the 13 PCRnegativepatients, one clinical and one molecularrelapse were observed at 7 and 9 months off therapy.One PCR-positive patient experienced clinical relapse3 months off therapy, while the second PCRpositivepatient became PCR-negative 6 months laterwithout any further therapy with persistent CR.The whole treatment program was entirely performedin an outpatient setting. No toxic deathsoccurred; the only severe toxicity observed was neutropeniain 44% of cases, but only one developedinfection. Rituximab toxicity was as usually expected.CONCLUSION: Elderly patients with advancedstagefollicular lymphoma may be safely and effectivelytreated with a brief chemoimmunotherapy.Clinical and molecular responses are improved by theaddition of rituximab to FND. Further follow-up isneeded to evaluate the duration of clinical and molecularresponse.

Abstract #344

A Multicenter Randomized Trial ofFludarabine and Mitoxantrone PlusRituximab vs CHOP Plus Rituximabas First-Line Treatment in PatientsWith Follicular Lymphoma

P. L. Zinzani, on Behalf of an Italian CooperativeStudy Group on LymphomaInstitute of Hematology and Medical Oncology,"L. e A. Sernoli", University of Bologna, Bologna, Italy

From October 1999 to March 2002, 159 follicularlymphoma patients from 12 Italian centers wererandomized for a comparative study of FM (fludarabine[Fludara], mitoxantrone [Novantrone]) vs CHOP(cyclophosphamide [Cytoxan, Neosar], doxorubicinHCl, vincristine [Oncovin], prednisone) chemotherapywith the addition of rituximab (Rituxan) inselected cases.To be eligible, patients were required to have ahistologically proven diagnosis of CD20 positiveaccording to the REAL classification and a positivepolymerase chain reaction (PCR) analysis (bonemarrow [BM] and/or peripheral blood [PB]) forbcl-2/IgH, age 15 to 70, stage II-IV, and EasternCooperative Oncology Group (ECOG) performancestatus of 0-2. After randomization, patients wereallocated to the FM arm (fludarabine 25 mg/m

²

/dIV on days 1 to 3 and mitoxantrone 10 mg/m

²

IV onday 1) or CHOP arm (doxorubicin 50 mg/m

²

IV onday 1, cyclophosphamide 750 mg/m

²

IV on day 1,vincristine 1.4 mg/m

²

IV on day 1, and prednisone100 mg/d orally on days 1 to 5). In both arms, patientswere assigned to receive 6 cycles of chemotherapy.Thereafter, to be eligible for rituximab treatment,they had to have partial or complete clinical response(PR or CR) and still be PCR positive in the BM and/or PB in two molecular analyses performed 4 and 6weeks after the sixth cycle. These patients wereelected to receive 4 weekly IV administrations ofrituximab (375 mg/m

²

).Clinical response to treatment was defined according3 categories: CR, PR, and failure. From 159patients registered, so far 93 were evaluable forresponse after completion of chemotherapy. CR, PR,and failure were, respectively, 68%, 26%, and 6% inthe FM arm (47 patients) vs 37%, 56%, and 7% in theCHOP arm (46 patients) (

P

= .003 for CR). Themolecular evaluation documented the clearance ofthe bcl-2/IgH chimeric gene in 34% and 20% for FMand CHOP regimen respectively (

P

= .115). Afterrituximab courses, the molecular evaluation documentedthe clearance in 59% and 40% for FM andCHOP subsets, respectively, with an improvement ofthe clinical response of PR patients: 87% and 76% ofCR rate for FM and CHOP subgroups, respectively.CONCLUSION: These results confirm the efficacyof FM for untreated follicular lymphoma patientsin terms of clinical and molecular response, and therole of rituximab in improving molecular and clinicalresponses when added to chemotherapy for first-linetreatment of follicular lymphoma.

Commentary on Abstracts
#1398, #604, #1390, #3073, #2246,
#1401, #1392, and #344
Bruce D. Cheson,MD

Rituximab (Rituxan) has rapidly become thestandard agent for the treatment of relapsedand refractory follicular and low-grade non-Hodgkin's lymphoma (NHL). However, its optimaluse is still being characterized. A numberof approaches are being considered to enhanceits activity. Several investigators have evaluatedthe use of this antibody as initial therapy.Hainsworth et al

(Blood 95:3052-3056, 2000)

treated 39 previously untreated patients with 4weeks of rituximab. The response rate of 54%,including 5% complete responses, was comparableto the results from the pivotal trial inrelapsed and refractory patients

(McLaughlinet al: J Clin Oncol 16:2825-2833, 1998).

On theother hand, results can vary dramatically withdifferences in patient selection. A higher responserate was reported in patients who receivedthe antibody as initial therapy in patientswith a lower tumor burden

(Colombat et al:Blood 97:101-106, 2001).

At the 2002 American Society of Hematology(ASH) meetings, Witzig and coworkers (abstract#1398) described their experience usingrituximab in 38 asymptomatic patients whowere previously untreated. These patients wouldbe considered to be in a low-risk category, with84% in favorable International Prognostic Index(IPI) categories. In this low tumor burdengroup, the overall response rate was 61%, with25% complete remissions. Despite the earlyintervention, the median time to progressionwas only 1.67 years with 20 in an unmaintainedremission out as long as 3+ years. Althoughgenerally well tolerated, there were some seriousadverse events, including hematologic toxicity,rash, infection, and urticaria. Thisapproach needs testing in a randomized trialbefore being accepted as a standard option.

Another possibility has been to evaluate therole of maintenance rituximab therapy.Hainsworth et al

(Blood 95:3052-3056, 2000)

were the first to publish data on this importanttopic. As discussed above, when they initiallyassessed response after completion of therapy,response rates were comparable to those fromthe pivotal study in relapsed and refractorypatients

(McLaughlin et al: J Clin Oncol16:2825-2833, 1998).

However, to consider thepossibility of delayed responses, patients werereevaluated at 6 months. The response categorywas upgraded in five patients: there were twoadditional partial responses and three completeresponses noted. Of the 13 patients whoreceived a second 4-week course, three improvedtheir response: two patients with stabledisease became partial responders and a partialresponse was now complete. The actuarialdisease-free survival at 1 year was 77%; however,the median follow-up was extremely shortat 8 months. Whether a maintenance approachprovides clinically meaningful benefit wouldrequire a randomized trial.

At the ASH meeting the Swiss Group forClinical Cancer Research (SAKK) (abstract#604) presented an update of their randomizedstudy including 202 patients with previouslyuntreated, relapsed or refractory follicular lymphoma.Their schedule differed from that reportedby Hainsworth et al in that the patientswho responded to standard induction with rituximabwere then randomized to either observationor maintenance with rituximab every 8weeks for four courses. There was a higherresponse rate in the untreated cohort (67% vs46%). More importantly, at a median follow-upof 35 months, the event-free survival favored themaintenance arm (23 vs 12 months). Moreover,the percentage of patients still in remission was80% in the maintenance arm vs 12.56% in thecontrol arm. The median response duration of36 months with maintenance was more thantwice that in the observation arm.Although these results with maintenance rituximabare provocative, they do not resolve therelated and important question of whether maintenanceto prevent relapse is preferable toretreatment upon relapse. In a report by Daviset al

(J Clin Oncol 18:3135-3143, 2000),

60patients who had previously responded to rituximaband relapsed at least 6 months later wereretreated with the antibody using the same doseand schedule. The overall response rate was38%, including 10% complete responses. It isnot clear why the response rate was not highersince almost all tumors that were assayed werestill positive for CD20 at the time of relapse. Nopatient developed a human antichimeric antibody(HACA).For unexplained reasons, thetime to treatment failure and time to progressiontended to be longer with the second coursethan with initial rituximab therapy. Other reportshave confirmed the potential for successfulre-treatment

(Igarashi et al: Int J Hematol73:213-221, 2001).

At the ASH meetings, Coiffier et al (abstract#1390) described their institutional experiencewith rituximab retreatment. Of their 483 patientswho had previously received rituximabeither alone or in combination with chemotherapy,117 relapsed, and 59 were retreated withrituximab either alone (n = 30) or in combinationwith chemotherapy (n = 29). The secondresponse rate was 93%, including 42% completeremissions. As was the experience of Daviset al

(J Clin Oncol 18:3135-3143, 2000),

time toprogression was longer with the second coursethan the first in all but 12 patients. Indeed, athird course was uniformly successful in thatsmall group of patients. What was not clearfrom the abstract was the number of patientswho were initially treated with rituximab alonewho responded a second time because they werenow treated with chemotherapy plus the antibody,or whether there were patients who initially received combined-modality treatment whoresponded to single-agent rituximab upon retreatment.Whereas re-treatment is clearly effectivein a subset of patients, the optimalstrategy remains to be defined.

The effect of rituximab appears to be greatestwhen combined with chemotherapy. The scientificrationale for such strategies is supportedby in vitro observations that monoclonal antibodyexposure may sensitize tumor cells tochemotherapy

(Demidem et al: Cancer BiochemRadiopharm 12:177-186, 1997; Johnsonet al: Int J Cancer 85:104-112, 2000; Alas et al:Cancer Res 61:5137-5144, 2001).

Specifically,treatment of lymphoma cells with rituximabappears to sensitize the cells to fludarabine(Fludara), cisplatin, vinblastine, and doxorubicinin association with down-regulation of interleukin(IL)-10 and antiapoptotic cytokines

(Alas et al: Cancer Res 61:5137-5144, 2001;Alas et al: Clin Cancer Res 7:709-723, 2001).

Czuczman et al

(J Clin Oncol 17:268-276,1999)

were the first to publish results of thecombination of CHOP (cyclophosphamide [Cytoxan,Neosar], doxorubicin HCl, vincristine[Oncovin], prednisone) plus rituximab in follicularand low-grade NHL. Their experienceincluded 40 patients, 31 of whom were previouslyuntreated. The response rate was 100%,including 58% complete remissions. The medianduration of response and time to progressionhad still not yet been reached at 45.8+ and47.2+ months, respectively. Almost half thepatients were still in remission from 36+ to54.5+ months. Fludarabine has also been combinedwith rituximab in indolent NHL

(Czuczmanet al: Proc Am Soc Clin Oncol18:17a[abstract 61], 1999).

The overall responserate was 93%, which included 80% completeremissions.

A major problem is that it is difficult toidentify the optimal regimen in the absence ofrandomized trials. Unfortunately, virtually everyregimen that is being used in NHL orchronic lymphocytic leukemia (CLL) is beingcombined with rituximab, and numerous suchabstracts were submitted to the ASH meeting.Martinelli et al (abstract #3073) described theresults of a phase II study of concurrent chlorambucil(Leukeran) with rituximab in 29 patientswith a low-grade histology. The overall responserate was reported as 93%, with 50%complete remissions, with short follow-up. Sacchiet al (abstract #2246) presented their experiencewith the FCR combination (fludarabine,cyclophosphamide, and rituximab) in 48 patientswith relapsed follicular NHL. Althoughthey claim that their patients were heavily pretreated,the median number of prior regimenswas only 1.7. Patients received four cycles ofthe agents, with a response rate of 97%, including47% complete remissions. The median follow-up of 12 months was too short for meaningfulconclusions.

Gregory et al (abstract #1401) gave four tosix cycles of FN (fludarabine, mitoxantrone[Novantrone]) followed by rituximab after a 4-to 6-week delay to a heterogeneous group ofindolent NHL. The overall response rate was97%, with 45% complete remissions. In anotherstudy, FN plus dexamethasone (FND) was deliveredfor four courses followed by rituximab(abstract #1392) in 59 older patients with a95% response rate, including 76% completeremissions (CR), 14% complete remissions/unconfirmed (CRu), and 5% partial remissions(PR). Of the 18 patients who were polymerasechain reaction (PCR) positive and tested aftertreatment, 15 became negative. Whether theseresults are superior to that achievable with thechemotherapy alone cannot be determined onthe basis of these data

(Velasquez et al: ProcAm Soc Clin Oncol 18:9a[abstract 27], 1999; JClin Oncol [in press]).

In one of the few randomized trials, Zinzaniand coworkers (abstract #344) updated theresults of their phase III trial in which 159patients with follicular NHL were randomizedto either fludarabine and mitoxantrone (FM) orCHOP, both regimens followed by rituximab inpartial or complete responders. From the 93patients evaluable for response at the time of thereport, complete response and overall responserates were significantly in favor of FM. Followingrituximab, the response rates were improvedto 87% for FM and 76% for CHOP.There was a trend toward more PCR-negativepatients with FM both before (34% vs 20%) andafter rituximab (59% vs 40%). Unfortunately,no time to progression or survival data are yetavailable and longer follow-up is ongoing.

Chronic Lymphocytic Leukemia

Abstract #771

A High Proportion of MolecularRemission Can Be Obtained Witha Fludarabine, Cyclophosphamide,Rituximab Combination in ChronicLymphocytic Leukemia

M. Keating, T. Manshouri, S. O'Brien, W. Wierda, H. Kantarjian,L. Washington, S. Lerner, M. AlbitarDepartment of Leukemia; Department of Hematopathology,University of Texas M. D. Anderson Cancer Center, Houston,Texas

Definitions of complete and partial remissions(CR + PR) in chronic lymphocytic leukemia (CLL)have become more stringent as treatments have becomemore effective. The National Cancer InstituteWorking Group (NCIWG) categorizes patients ashaving CR, nodular (N) PR (CR except for residuallymphoid aggregates on bone marrow biopsy), andother PRs. Many CR patients have clonal diseasedetectable on flow cytometry and molecular remissions(polymerase chain reaction [PCR]-negative forIgVh gene) have been infrequently reported.A combination chemoimmunotherapy protocol hasbeen developed (FCR) that combines fludarabine(Fludara) at 25 mg/m

²

/d for 3 days, cyclophosphamide(Cytoxan, Neosar) at 250 mg/m

²

/d for 3 days,and rituximab (Rituxan) at 375 to 500 mg/m

²

on day1

(Blood 98:771a [abstract 3210], 2001).

The use ofFCR as initial therapy for 135 CLL patients hasresulted in 67% CR, 19% NPR, and 18% PR. Nineresponders have relapsed clinically and 12/135 patientshave died: 2/19 (2%) CR, 1/19 (5%) NPR, 4/18(22%) PR, and 5/7 (71%) failure. Seventy-sevenpatients had PCR performed on bone marrow at theend of therapy (usually 6 cycles) and 125 patients hadflow cytometry performed. The relationship betweenNCIWG response, PCR, and CD5 + 19 coexpressionon flow, and likelihood of relapse, is shown below.CD5 + 19 coexpression < 1% occurred morefrequently in patients < 70 years of age, spleen size < 5 cm, below the L costal margin, beta-2-microglobulin< 3 mg/L and lower marrow cellularity. Nocharacteristics predicted for PCR negativity. Thedegree of PCR positivity is semiquantitated by comparingthe level of PCR amplification of IgVh to theras gene and developing an IgVh/ras ratio. 13/41PCR-negative patients have become low-level PCRpositive, usually within 6 months of follow-up. Nineof the 36 PCR-positive patients have increased theratio by more than 100% and 13 have had a > 50%decrease. None of the PCR-negative patients havehad a clinical or flow relapse, while three of 36patients have had a flow relapse and two a clinicalrelapse. Two of 85 patients (2%) with CD5 + 19coexpression < 1% have had a flow relapse vs 15/34(44%) with 1% (

P

< .001).CONCLUSION: With longer follow-up, the betterquality remissions obtained with FCR will allowexploration of the characteristics of response, whichare the best predictors of prolonged survival.

Abstract #775

Combined Use of Alemtuzumab andRituximab in Patients With Relapsedand Refractory Chronic LymphoidMalignancies-An Update of theM. D. Anderson Experience

S. Faderl, D. A. Thomas, S. O'Brien, W. Wierda,H. M. Kantarjian, G. Garcia-Manero, F. J. Giles,C. A. Koller, M. Beran, A. Ferrajoli, S. Lerner, M. J. KeatingDepartment of Leukemia, The University of TexasM. D. Anderson Cancer Center, Houston, Texas

Both alemtuzumab (Campath) and rituximab (Rituxan)are active in patients with relapsed/refractorychronic lymphocytic leukemia (CLL) and non-Hodgkin's lymphoma (NHL). However, efficacy insome anatomic disease sites is suboptimal (lymphnodes [LN], bone marrow [BM]) and the antibodiesare not curative by themselves. Combinations ofmonoclonal antibodies with other active agents aretherefore being studied. We explored the safety andefficacy of the combined use of alemtuzumab andrituximab in patients with advanced CLL and otherchronic lymphoid malignancies that coexpress CD20and CD52.Rituximab is given at 375 mg/m

²

IV weekly for 4doses. Alemtuzumab is given at 3, 10, and 30 mg IVon three consecutive days during week 1, and then at30 mg IV on days 3 and 5 for weeks 2 through 4.Patients could be retreated for up to a total of 3courses. All patients received anti-infective prophylaxiswith trimethoprim/sulfamethoxazole and valacyclovir(Valtrex). Cytomegalovirus (CMV)antigenemia testing was done weekly during therapy.Among 48 patients on study (32 CLL, 9 CLL/prolymphocyticleukemia [PLL], 1 PLL, 4 mantle cell lymphoma,2 Richter's), 47 are assessable for response.The median age is 62 years (range: 44-79 yr). Themedian number of prior therapies is 4 (range: 1-9).79% had ≥ Rai stage 3 and 54% were refractory toprior fludarabine (Fludara) with or without alkylatortherapy.Twenty-five patients responded (OR 53%; CR9%, nPR 4%, PR 40%); 20 of the responses occurredamong 32 patients with CLL (OR 63%), including 2CR and 1 nPR. No responses were seen in patients withmantle cell leukemia or Richter's transformation. Responseby site was as follows: peripheral blood 90%CR, BM 37% CR, LN 59% ≥ 50% improvement, andliver/spleen 67% ≥ 50% improvement.Nonhematologic toxicities were mostly infusionrelatedand ≤ grade 2 by National Cancer Institutetoxicity criteria. Most frequently observed were fever(75%), rigors (67%), skin reactions (38%), fatigue(33%), gastrointestinal symptoms (27%), and dyspnea(25%). Twenty-five patients (52%) experiencedat least one infectious episode (CMV reactivation 13/48 [27%], fever of unknown origin 6/48 [13%],pneumonia 5/48 [10%], sinusitis 3/48 [6%]).CONCLUSION: We conclude that the combinationof alemtuzumab and rituximab is feasible. Responserates in this heterogeneous and poor-prognosisgroup of patients appear higher than responses withalemtuzumab alone.

Abstract #772

A Phase II Trial of FludarabineFollowed by Alemtuzumabin Previously Untreated ChronicLymphocytic Leukemia PatientsWith Active Disease: Cancer andLeukemia Group B Study 19901

K. R. Rai, J. C. Byrd, B. L. Peterson, R. A. LarsonLong Island Jewish Medical Center, Albert Einstein Collegeof Medicine, New Hyde Park, New York; Ohio State University,Columbus, Ohio; Duke University, Durham, NorthCarolina; University of Chicago, Chicago, Illinois; Cancerand Leukemia Group B, Chicago, Illinois

Since the demonstrationof efficacy of fludarabine(Fludara) as a single agent infront-line therapy of chroniclymphocytic leukemia(CLL)

(N Engl J Med343:1750, 2000),

the Cancerand Leukemia Group B hasbeen systematically conductingstudies to improve thecomplete response (CR) ratein this disease by combining fludarabine with monoclonalantibodies. Here we report results of a phase IIstudy using sequential administration of fludarabinefollowed by alemtuzumab.Our primary goal was to determine the toxicity andresponse rates in previously untreated CLL patientswith active disease who met the National CancerInstitute Working Group (NCI-WG) 1996 criteria forinitiation of treatment. Treatment included fludara-bine at 25 mg/m

²

/d * 5 IV, monthly * 4, followed byobservation for 2 months when an evaluation ofinitial response was performed. Patients achieving aresponse of stable disease (SD) or better were treatedwith a 6-week course of alemtuzumab. Alemtuzumabwas given at 30 mg IV three times a week after astepped-up dosing during the first week (3 mg as thefirst dose and increasing, as tolerated, to 10 mg, andto 30 mg). Prophylaxis with trimethoprim/sulfamethoxazoleDS and acyclovir was required duringand for 6 months after alemtuzumab therapy.Evaluation of final response status was performed2 months after completion of therapy. Between February2000 and February 2001, 57 patients wereenrolled. One patient died prior to starting fludarabine.Toxicity and efficacy data are based on 56patients who started fludarabine therapy. Eleven patientsexperienced major infections requiring parenteralantibiotics and/or hospitalization during thefludarabine period, and one patient died of septicemiafrom a gram-positive organism. Two patients (4%)achieved a CR and 29 (52%) a partial response (PR)after the fludarabine phase of therapy. Twenty patientseither had progression of disease (n = 10) ordied during the fludarabine therapy phase (n = 1), ordiscontinued treatment due to fludarabine toxicity (n= 3), or withdrew consent (n = 3), or whose data on thepost-fludarabine phase have not been received (n = 3).Thus, data on 36 patients who received alemtuzumabare available.Alemtuzumab-infusion-related reactions occurredin virtually all patients but were manageable withappropriate measures and were grade 1 or 2 in mostcases. Grade 3 or worse major infections (requiringparenteral antibiotics and/or hospitalization) wererecorded in 12 patients during or following alemtuzumabtherapy. Infections with cytomegalovirus(CMV) occurred in 8 patients during or within 4months following alemtuzumab. The outcome ofCMV was fatal in one case, complete or partialresolution occurred in six, and persistent CMV inone. After the fatal CMV event, vigorous surveillancefor CMV by weekly qualitative polymerase chainreaction-based testing was initiated to enable earlyrecognition and prompt discontinuation of alemtuzumaband institution of antiviral therapy.Among 36 patients who entered the alemtuzumabphase of therapy, there were 15 CRs (42%) and 18PRs (50%), with an overall response rate of 92%(NCI-WG 1996 criteria). Out of 11 patients with SDafter fludarabine, 3 achieved CR, and 5 had PRfollowing alemtuzumab. Among all 56 patients (intent-to-treat population), the CR and PR incidencewas 27% and 43%, respectively. After a median of 10months, 87% of 56 patients were alive.CONCLUSION: These early results are encouraging,but a longer follow-up will be necessary todetermine if alemtuzumab given after initial cytroreductionwith fludarabine leads to an improvement inthe natural history of CLL.

Commentary on Abstracts
#771, #775, and #772
Bruce D. Cheson, MD

Chronic lymphocytic leukemia cells provideexcellent targets for monoclonal antibody therapy.Nevertheless, it was not until recently thatmonoclonal antibodies have become availablewith demonstrated activity for this disorder.Alemtuzumab (Campath) was the first antibodyapproved for the treatment of CLL. The pivotaltrial included 93 patients treated at 21 centersin the United States and Europe who had failedprior therapy with fludarabine and who hadalso received alkylating agents

(Keating et al:Blood 99:3554-3561, 2002).

The median agewas 66 years and 76% had Rai stage II or IVdisease. Almost half had never responded to anynucleoside analog. Treatment with alemtuzumabwas continued to a maximum of 12 weeks.

The overall response rate of 33%, including2% complete remissions, was lower in patientswith large lymph nodes. An additional six patientshad clearing of CLL from all sites, butwith persistent anemia or thrombocytopeniathat improved on long-term follow-up. Themedian time to response was 1.5 months, witha median duration of response of 8.7 months.Death was due to or complicated by infection in17 patients. The median survival was 16 monthsfor all patients, and 32 months in responders.Of concern was cytomegalovirus (CMV) reactivationin seven patients.

Recently, Lundin et al

(Blood 100:768-773,2002)

reported their experience with 41 patientswith previously untreated CLL who receivedalemtuzumab by the subcutaneous route.The overall response rate of 87% included 19%complete remissions. Of note was the relativelylow rate of severe infusion-related reactionsmaking this route of administration attractivefor further study. Local reactions, however,were substantial and, perhaps, new diluentsmight ameliorate this problem.

The experience with rituximab as a singleagent in relapsed and refractory CLL has beensomewhat disappointing. A number of investigatorshave reported response rates of 10% to15% in previously treated patients with CLLand small lymphocytic lymphoma (SLL)

(Piro etal: Ann Oncol 10:655-661, 1999; Nguyen et al:Eur J Haematol 62:76-82, 1999; Winkler et al:Blood 94:312a[abstract 1396], 1999; Huhn etal: Blood 98:1326-1331, 2001)

Major increasesin dose or dose intensity have done little morethan induce more partial remissions

(Byrd et al:J Clin Oncol 19:2153-2164, 2001; O'Brien etal: J Clin Oncol 19:2165-2170, 2001).

On theother hand, higher response rates have beennoted when the antibody is used as the initialtreatment of CLL/SLL. In a series of 44 patientswho received 4 consecutive weeks of the antibody,the response rate after the first coursewas 51% including 4% complete remissions

(Hainsworth et al: J Clin Oncol, in press, 2003).

Additional courses added little to this responserate. The median progression-free survival was18.6 months.

Clearly the most interesting data have beengenerated from combinations of rituximab withchemotherapy, such as fludarabine or fludarabineplus cyclophosphamide(

Byrd et al: Blood101:6-14, 2003; Wierda et al: Blood98:771a[abstract 3210], 2001

),with overall responserates of at least 90%, more than half ofwhich are complete remissions.At ASH, Keatingand coworkers (abstract #771) made a casefor changing the criteria for response assessmentin CLL to include molecular responses. Intheir updated experience with FCR in 135 patients,the complete remission rate was 67%with an overall response rate of over 90%.Patients who were PCR negative did not experiencea clinical or flow relapse, although aquarter became PCR positive. In contrast, twopatients who were PCR positive after therapyhad a clinical relapse. They proposed that thequality of response can be better assessed nowwith molecular assessment. Whether the abilityto induce molecular responses with currentlyavailable agents will prolong the survival ofpatients with CLL will require further study.A number of other combinations were discussedat the ASH meeting. The group fromM. D. Anderson (abstract #775) updated theirresults with the combination of alemtuzumaband rituximab in a variety of relapsed andrefractory chronic lymphoid malignancies, primarilyCLL. The rationale was that both wereactive in the disease. However, in vitro synergydata are lacking. Of their 47 patients eligiblefor response, the median number of prior regimenswas 4 and 79% had advanced-stage disease.Actual response rates are difficult toextract from the abstract; however, 90% ofpatients experienced a peripheral blood completeremission. Unfortunately, rates were lowerin other sites. Nevertheless, this regimenwarrants further study to determine if it offersan improvement over alemtuzumab as a singleagent.Using a somewhat different strategy to improvethe activity of alemtuzumab, Cancer andLeukemia Group B (CALGB) investigators administeredfour cycles of fludarabine followedby the antibody after 2 months in respondingpatients (abstract #772). Response rates followingthe fludarabine (4% CR, 52% PR) wereinferior to the results produced in the CALGBled North American trial comparing fludarabinewith chlorambucil(Rai et al: N Engl J Med343:1750-1757, 2000), in which the overallresponse rate was 63%, including 20% completeremissions. The difference between theseresults is probably related to an insufficientnumber of courses in the more recent trial.Nevertheless, upon completion of alemtuzumabthere was an improvement in response rates to42% CR and 50% PR. Unfortunately, the immunosuppressionrelated to the agents resulted ininfections with CMV occurred in eight patientsduring or within 4 months of alemtuzumab, withone fatality.

Waldenstrm'sMacroglobulinemia

Abstract #3211

Pretherapy Serum IgM Levels PredictClinical Response to ExtendedRituximab in Waldenstrm'sMacroglobulinemia

S. P. Treon, C. A. Emmanouilides, E. Kimby, A. Branagan,C. Mitsiades, K. C. Anderson, S. R. FrankelMedical Oncology, Dana Farber Cancer Institute, Boston,Massachusetts; Hematology Oncology, UCLA Medical Center,Los Angeles, California; Hematology Oncology, KarolinskaInstitutet, Stockholm, Sweden; Hematology Oncology,Greenebaum Cancer Center, Baltimore, Maryland

We undertook a prospective,multicenter phase IIstudy to determine the activityof extended rituximab(Rituxan) therapy in Waldenstrm'smacroglobulinemia(WM). Twenty-nine patientswith a median age of 65 years(range: 43-90 years), andmedian prior treatments of 1(range: 0-2) were enrolled.Twenty-six patients completed the intended therapy,which consisted of receiving 4 infusions of rituximab(375 mg/m

²

weekly), followed by 4 additional infusionsof rituximab (375 mg/m

²

weekly) at week 12, ifthey demonstrated a response or had stable disease(SD) following the first course of therapy. Overall,treatment was well tolerated. One patient died ofsepsis and another patient experienced neutropenia,both events deemed by their treating physicians to beunrelated to protocol treatment. One patient experiencedsignificant diffuse joint pains after 4 infusionsof rituximab that resolved after steroid therapy.Best overall responses following rituximab for 26patients who completed protocol therapy are as follows:partial response ( > 50% decline in IgM) in 12patients (46.1%) and minor response ( > 25% declinein IgM) in 7 patients (26.9%). The overall responserate was 73.0% and 3 (11.5%) patients demonstratedSD. The median time to treatment failure for respondingpatients has not been reached with a medianfollow-up of 20 months (range: 7-30+ months).Analysis of baseline serum IgM levels and clinicalresponse to rituximab showed that 18/20 (90.0%)patients with an IgM level under 6,000 mg/dL responded,whereas only 1/6 (16.6%) patients with anIgM level of over 6,000 mg/dL had a response(P = .002).No correlation with pretherapy bone marrowinvolvement and clinical response to rituximabwas seen.CONCLUSION: These data therefore show thatextended rituximab therapy is highly active in WMpatients and produces durable responses. Waldenstrm'smacroglobulinemia patients with serum IgMlevels of < 6,000 mg/dL are more likely to benefitfrom extended rituximab therapy. The mechanism forthis finding is currently under investigation in ourlaboratory.

Abstract #794

Combination Therapy With Rituximaband Fludarabine Is Highly Active inWaldenstrm's Macroglobulinemia

S. P. Treon, P. Wasi, C. A. Emmanouilides, S. R. Frankel, E.Kimby, A. Lister, P. Morel, A. Kelliher, A. Branagan, F.Preffer, K. C. AndersonMedical Oncology, Dana Farber Cancer Institute, Boston,Massachusetts; Hematology Oncology, McMaster UniversityMedical Center, Hamilton, Ontario, Canada; Hematology-Oncology, UCLA Medical Center, Los Angeles, California;Hematology-Oncology, Greenebaum Cancer Center, Baltimore,Maryland; Hematology, Karolinska Institutet, Stockholm,Sweden; Medical Oncology, St Bartholomew'sHospital and Cancer Research, London, United Kingdom;Clinical Hematology, Centre Hospitalier Schaffner, Lens,France; Pathology, Massachusetts General Hospital, Boston,Massachusetts

Rituximab (Rituxan) and fludarabine (Fludara) assingle agents are active in Waldenstrm's macroglobulinemia(WM) with an overall response rate (ORR)of 30% to 70%. In this study, we examined theseagents in combination in WM patients who hadreceived less than 2 prior therapies, and who had notpreviously been treated with any nucleoside analog orrituximab, utilizing the following schema:

• Weeks 1-4: rituximab (375 mg/m²) every week
• Weeks 5, 9, 13: fludarabine (25 mg/m²/d) * 5days
• Weeks 17, 18: rituximab (375 mg/m²) everyweek
• Weeks 19, 23, 27: fludarabine (25 mg/m²/d) *5 days
• Weeks 30, 31: rituximab (375 mg/m²) everyweek

Patients were evaluated at week 12, and if theydemonstrated stable disease (SD) or better were eligiblefor further therapy and were evaluable for response.Dose modifications, and delay in fludarabineadministration, along with use of G-CSF, were permittedbased on hematologic toxicities. 23 WM patientshave been enrolled with a median age of 61(range: 52-75 years), and median prior therapies of 1(range: 0-2).At present, 15 patients are off protocol therapy,including 1 patient who experienced complicationsafter an abrupt surge in her serum IgM and viscositylevels following 4 infusions of rituximab. Two patients(1 partial response [PR], 1 SD) died while offprotocol therapy, and their causes of death werereported by their treating clinicians as unrelated toprotocol therapy or disease. Delays in therapy due toneutropenia were common, and 44.4% of patientsexperienced grade 3/4 neutropenia, which clinicallywas not significant in most patients. However, protocoltherapy was truncated after the fourth (n = 2) and fifth(n = 1) courses of fludarabine for 3 patients who hadpersistent neutropenia and/or thrombocytopenia despitedelays in therapy and/or use of G-CSF support. Of14 evaluable patients, 12 (85.7%) responded (1 completeresponse [CR], 8 PR, 3 minor response [MR]),with response durations ranging from 3 to 11+ months.CONCLUSION: The preliminary results of thisstudy therefore suggest that rituximab in combinationwith fludarabine appears to result in higher ORR thanthose reported with either agent alone, with acceptabletoxicity. The impact on response duration remainsto be defined.

Commentary on Abstracts#3211 and #794Bruce D. Cheson, MD

The standard of therapy for Waldenstrm'smacroglobulinemia has changed in recent yearsfrom alkylating agent-based therapy to the useof a nucleoside analog such as fludarabine,cladribine (Leustatin), or pentostatin (Nipent)(

Zinzani et al: Eur J Haematol 54:120-123,1995; Leblond et al: J Clin Oncol [in press],1998; Dimopoulos et al: Am J Med 95:49-52,1993; Dhodapkar et al: Blood 90:577a[abstract2571], 1997; Dimopoulos et al: Ann Oncol6:49-52, 1995; Dimopoulos et al: J Clin Oncol12:2694-2698, 1994; Betticher et al: Br J Haematol99:358-363, 1997

).More recently, severalgroups have demonstrated a variable level ofactivity for rituximab in this disease (

Byrd et al:Blood 92[suppl 1]:106a[abstract 433], 1998;Colares et al: Blood 94:259b[abstract 4368],1999; Foran et al: J Clin Oncol 18:317-324,2000; Treon et al: Proc Am Soc Clin Oncol19:6a[abstract 13], 2000

),generally in the rangeof 30% or less. Therefore, it would be useful tohave an accurate predictor of response in those patients.At the ASH meeting, Treon and coworkers(abstract #3211) presented their experiencewith 26 patients treated on a protocol involving"extended" rituximab therapy, consisting offour weekly infusions followed by an additionalfour infusions at week 12 for those with at leaststable disease. Responses occurred in 10 of 20patients with a pretreatment IgM concentration< 6 g/dL, and in only 1 of 6 with higher values.In an attempt at improving on the singleagentactivities of both rituximab and fludarabine,Treon and his international colleaguespresented the first data on the regimen of fludarabinealternating with rituximab in patientswith Waldenstrm's macroglobulinemia (abstract#794). Of the 23 patients reported, theclaimed response rate was 85.7%, includingone CR and eight PR. However, when the minorresponses are appropriately excluded from theresponse rate, it is actually 64%, which iswithin the range of the fludarabine alone. Arandomized study is needed to determine if thecombination is superior to the individual singleagents.

Aggressive Lymphomas

Abstract #1396

Long-Term Follow-Up of a Phase IIStudy of Rituximab in CombinationWith CHOP Chemotherapy in PatientsWith Previously Untreated AggressiveNon-Hodgkin's Lymphoma

J. M. Vose, B. K. Link, M. L. Grossbard, M. Czuczman,A. Grillo-Lopez, M. Benyunes, E. Gaynor, R. I. FisherInternal Medicine, University of Nebraska Medical Center,Omaha, Nebraska; Internal Medicine, University of Iowa,Iowa City, Iowa; Hematology/Oncology, St. Lukes-Roosevelt, New York, New York; Oncology, Roswell ParkCancer Institute, Buffalo, New York; IDEC Pharmaceuticals,San Diego, California; Genentech, San Francisco, California;Hematology/Oncology, Loyola University, Chicago,Illinois; Wilmont Cancer Center, Rochester, New York

A phase II study to determine safety and efficacywith the combination of 6 cycles of rituximab (Rituxan)375 mg/m

²

(day 1) and CHOP chemotherapy(cyclophosphamide [Cytoxan, Neosar] 750 mg/m2day 3, doxorubicin HCl 50 mg/m

²

day 3, vincristine[Oncovin] 1.4 mg/m

²

day 3, prednisone 100 mg days3-7) was administered to 33 patients with aggressivenon-Hodkin's lymphoma (NHL) as frontline therapy.Twently four patients were stage III/IV, 8 stage II, and1 bulky stage IE. Ten patients had one single mass> 10 cm. The histologies of the patients includedfollicular large cell (7), diffuse large cell (22), immunoblastic(2), and other aggressive NHL (2). The medianage of patients was 52 years (range: 20-79 years) with15/33 patients having an International Prognostic Index(IPI) of 0 or 1 and 18/33 having an IPI of 2.All 33 patients were able to complete 6 cycles ofCHOP plus rituximab. A previous analysis demonstratedthe overall response rate to be 32/33 (97%),20/33 (61%) CR and 12/33 (36 %) PR. With a medianobservation time of 26 month, 30/33 patients werealive and 29/33 patients were in continued remission(

J Clin Oncol 19:389-397, 2001

). Long-term evaluationof these 33 patients now with a median followuptime of 62 months (range: 28-70 months) revealsthat 29/33 (88%) remain alive and 27/33 (81%) havebeen continuously progression-free.From the time of the first analysis, one patient hadprogressive disease and one patient died withoutprogression of non-small-cell lung carcinoma. Theoverall survival (OS) by Kaplan-Meier plot at≥ 60 mo is 87% and the progression-free survival(PFS) is 80%. Two of 15 patients with an IPI of 0 or1 have failed, with one death, and 4/18 with an IPI ≥2 have failed, with three deaths. Two of the relapsingpatients were able to be salvaged with additionalchemotherapy and/or stem cell transplantation. Nolong-term complications of CHOP plus rituximabhave been reported.CONCLUSION: The long-term results of thiscombination in the phase II trial demonstrate the highpotential for CHOP plus rituxamab to be safe andeffective as frontline therapy for aggressive NHL.

Abstract #603

Rituximab Plus CHOP in theTreatment of Elderly Patients WithDiffuse Large B-Cell LymphomaOvercomes bcl-2-AssociatedChemotherapy Resistance

N. Mounier, J. Briere, C. Gisselbrecht, P. Gaulard, P. Lederlin,C. Sebban, A. Bosly, P. Morel, H. Tilly, R. Bouabdallah,F. Reyes, B. CoiffierInstitut Hematologie, Hopital Saint Louis, Paris, France;Hematologie, Hopital Henri Mondor, Creteil, France; Hematologie,Hopital Lyon Sud, Pierre Benite, France; GELA,Paris, France

In patients with diffuselarge B-cell lymphoma (DLBCL),R-CHOP, the combinationof rituximab (Rituxan)with CHOP (cyclophosphamide[Cytoxan, Neosar],doxorubicin HCl, vincristine[Oncovin], prednisone), hasdemonstrated its benefit incomparison to CHOP alonefor the treatment of elderlypatients. The mechanisms by which rituximab inducesits antitumor activity are not fully understood butinvolve apoptotic pathways. bcl-2 protein expressionhas been associated with poorer prognosis in DLBCLpatients.In order to assess in vivo if rituximab reduces bcl-2-associated failure, we analyzed bcl-2 expressionand clinical outcome from the Group d'Etude desLymphomes de l'Adulte (GELA) LNH-98 5 trial

(NEngl J Med 346:235-242, 2002)

. Patients aged between60 and 80 years with untreated DLBCL wererandomized to receive eight cycles of standard CHOPevery 3 weeks or eight cycles of R-CHOP (CHOPplus rituximab at 375 mg/m

²

the same day) and noradiation. Among the 399 patients included 292 caseswith histologically reviewed DLBCL and materialfor bcl-2 study was available. Staining for bcl-2 wasperformed on deparaffinized sections using an indirectimmunoperoxydase method and a specific monoclonalantibody (bcl-2 124, Dako SA, Glostrup,Denmark). Optimum labeling was obtained by microwavepreatreatment. Scoring for bcl-2 was independentlyperformed with a centralized procedure bytwo of us without knowledge of the clinical data.Tumors were considered positive when at least50% of tumors cells expressed bcl-2 protein. Therewere 193 (66%) bcl-2-positive and 99 (34%) bcl-2-negative patients. Median age was similar at 69 yearsand the male/female ratio was 0.9. The InternationalPrognostic Index (age-adjusted IPI) was identicalscore 0: 0.5%; 1: 30%, 2: 44%, 3: 16.5%. Thedistribution of bcl-2-positive cases and patient characteristicswas similar in both treatment arms: 92(47%) CHOP and 101 (52%) R-CHOP in bcl-2positive; 45 (46%) CHOP and 54 (54%) R-CHOP inbcl-2 negative.Response rate (CR + CRu) at the end of treatmentfor R-CHOP and CHOP was 78% and 61% (

P

= .009)in bcl-2 positive and 76% and 73% (

P

= .7) in bcl-2negative.With 2 years median follow-up, R-CHOPwas significantly associated with a better overallsurvival in bcl-2 positive (67 ± 9% and 48 ± 11% at2 years;

P

= .004) but not in bcl-2 negative (72 ± 12%and 67 ± 14% at 2 years;

P

= .6), as shown in thefigure. In addition, R-CHOP was significantly associatedwith a better event-free survival in bcl-2 positive(58 ± 10% and 32 ± 10% at 2 years;

P

< .0001)but not in bcl-2 negative (60 ± 13% and 40 ± 15% at2 years;

P

= .13). After controlling for heterogeneity,multivariate analysis confirmed the significant survivalbenefit of R-CHOP in bcl-2 positive but not inbcl-2 negative (RR = 1.9 vs 1.2,

P

= .02).CONCLUSION: These results, with 2-yearfollow-up, suggest that rituximab overcomesbcl-2-associated chemotherapy failure, but has lesseffect in bcl-2-negative patients.

Commentary on Abstracts
#1396 and #603
Bruce D. Cheson,MD

The activity of monoclonal antibodies suchas rituximab is certainly not restricted to indolentB-cell malignancies. Early on in the developmentof this agent, responses were observedin patients with aggressive NHL as well. Coiffieret al

(Blood 92:1927-1932, 1998)

conducteda randomized phase I/II study in patients withaggressive NHL and mantle cell lymphoma(MCL) who were refractory to therapy, were infirst or second relapse, following progressionafter an initial partial response, or in those overthe age of 60 years who were previously untreated.Patients received eight weekly infusionsof the antibody at 375 mg/m² or oneinfusion at this dose followed by seven weeklyinfusions of 500 mg/m². Of the 30 patients withdiffuse large B-cell NHL, 37% responded with9% complete remissions, none in patients withbulky disease. The median time to progressionfor the responding patients was 246 days.Based on the impressive results with thecombination of CHOP with rituximab in follicularNHL

(Czuczman et al: J Clin Oncol 17:268-276, 1999),

a similar approach was taken inpreviously untreated patients with large B-cellNHL. Vose and coworkers

(J Clin Oncol 19:389-397, 2001)

published the results of a multicentertrial in which the antibody wasadministered on day 1 of each 21-day cycle,with CHOP beginning on day 3, in contrast tothe schedule reported by Czuczman et al. Theirpatients primarily had diffuse large B-cell NHL,but they also included seven with follicularlarge cell NHL, two with immunoblastic lymphoma,and two others. Of 33 patients evaluablefor response, the overall response rate was98%, including 61% complete remissions, whichis modestly better than would be expected fromCHOP alone

(Fisher et al: N Engl J Med328:1002-1006, 1993).

There was no additionaladditionaltoxicity from CHOP plus rituximab comparedwith what would have been expected fromCHOP alone.This study was updated at the ASH meetings(abstract #1396). Based on long-term followupof 33 patients at a median of 62 months, 81%remain free of progression, with a projectedsurvival of 87%. Patients relapsed in both lowandhigh-risk groups by the IPI, two of whomwere salvageable by stem cell transplantation.This study demonstrates the long-term safetyand efficacy of this program.The first published phase III trial designed todetermine the benefit of adding rituximab tochemotherapy was reported by Coiffier et al ofthe Group d'Etudes des Lymphomes de l'Adulte

(GELA) (N Engl J Med 346:235-242, 2002).

These investigators randomized 399 patientsover the age of 60 with diffuse large B-cell NHLto either CHOP or CHOP plus rituximab. Rituximabwas delivered on day 1 of each of theeight cycles. The CR and CRu rate with thecombined-modality arm was 76% vs 63%, witha significant benefit in event-free survival andoverall survival. A US Intergroup trial completedrandomization of more than 600 patientsonto a similar study, but also with a secondaryrandomization following response to observationor further antibody therapy. Hopefully, theresults of the US study will be available laterthis year and will confirm the French observations.A problem in the interpretation of themultiple trials is that not all of the variousCHOP/rituximab regimens have a similar design

(Czuczman et al: J Clin Oncol 17:268-276,1999; Coiffier et al: N Engl J Med 346:235-242,2002).

Whether one strategy is superior to theothers is not known.An interesting observation was first made byWilson and his National Cancer Institute (NCI)coworkers

(Blood 98:343a[abstract 1447], 2001)

concerning the effect of rituximab on bcl-2-related drug resistance. Patients treated withthe infusional EPOCH (etoposide, prednisone,vincristine, cyclophosphamide, doxorubicin)chemotherapy regimen whose tumors were bcl-2 positive had a poorer survival than theirnegative counterparts. However, when rituximabwas added to this regimen, there was littlechange in the outcome of the bcl-2-negativegroup; however, the outcome of the bcl-2-positive patients was now comparable to theother group. In support of this observation wasan abstract at ASH from Mounier et al of theGELA group (abstract #603). Their study included193 bcl-2-positive patients and 99 bcl-2-negative patients. As with the NCI experience,there was no difference between the groups thatwere bcl-2 negative, but the R-CHOP patientsfared much better if they were bcl-2 positive.

Mantle Cell Lymphoma

Abstract #1399

Fludarabine, Mitoxantrone, andRituxan: An Effective Regimen for theTreatment of Mantle Cell Lymphoma

A. M. Levine, B. M. Espina, A. Mohrbacher, L. H. Buchanan,N. Berman, L. Mark, A. U. Khan, A. Tulpule, D. DouerMedicine, Division of Hematology, Keck School of Medicine,University of Southern California, Los Angeles, California

Mantle cell lymphoma is among the most aggressiveforms of non-Hodgkin's lymphoma (NHL), associatedwith lack of durability of response to standardCHOP or other therapy, and short median survival.Better therapeutic options are currently needed.Patients with newly diagnosed or relapsed mantlecell NHL, with no prior exposure to fludarabine(Fludara) or mitoxantrone (Novantrone), were studied.Patients were treated with 3 cycles of fludarabine(F), 25 mg/m

²

IV days 1-3; mitoxantrone (M) 10 mg/m

²

IV day 1;followed by 3 additional cycles withrituximab (R) 375 mg/m

²

on day 1, fludarabine days3-5, and mitoxantrone day 3. Chemotherapy cycleswere repeated every 28 days. After completion of 6cycles of FMR therapy, single-agent maintenancetherapy with rituximab given weekly * 3 was institutedin 8 (62%) patients.Eight male and five female patients with a medianage of 62 years (range: 43-87 years) were studied.Five patients (38%) had relapsed or had refractorydisease from prior systemic chemotherapy includingstem cell transplantation in one. All patients hadeither stage III (n = 2) or stage IV disease (n = 11),with bone marrow involvement present in 8 (62%).Other sites of extranodal disease included pleuraleffusions in 2, and small bowel and colon in 1 patienteach.Treatment has been well tolerated with transientgrade 3 or 4 neutropenia reported in only 4 (13%)patients. There have been no episodes of neutropenicfever, sepsis, or any hospitalizations due to toxicity.No grade 3 or 4 thrombocytopenia, anemia, or nonhematologictoxicity has been documented. Of the 12evaluable patients, complete remission has been documentedin 11 (92%), including 4 of 5 with relapsed/refractory disease. One patient is too early to evaluatefor response. None of the 11 complete respondershave relapsed, after a median duration of completeremission lasting 14.6+ months (range: 3.3+ to 21.2+months). Two patients have gone on to receive stemcell transplantation after attaining complete remissionto FMR. All patients remain alive after a medianfollow-up of 18.7+ months (range: 1.4+ to 23.6+months).CONCLUSION: Combination therapy with fludarabine,mitoxantrone, and rituximab is very welltolerated, and is highly active with durable completeremission rate of 92% in patients with both newlydiagnosed and relapsed mantle cell lymphoma. Accrualis ongoing.

Abstract #3084

Intensive Chemotherapy andAutologous Stem CellTransplantation Plus RituximabIs Superior to ConventionalChemotherapy for NewlyDiagnosed Advanced-StageMantle Cell Lymphoma:A Matched-Pair Analysis

J. Mangel, H. A. Leitch, J. M. Connors, R. Buckstein,K. Imrie, D. Spaner, N. L. BerinsteinAdvanced Therapeutics Program, Toronto Sunnybrook RegionalCancer Centre, Toronto, Ontario, Canada; Universityof British Columbia and BC Cancer Agency, Vancouver,British Columbia, Canada

We undertook a matchedpairanalysis to compare theoutcomes of patients treatedfor advanced-stage mantlecell lymphoma (MCL) withtwo different strategies.Twenty patients were treatedat the Toronto-SunnybrookRegional Cancer Centre inToronto, Canada, on a prospectivephase II clinical trialof high-dose therapy and autologous stem celltransplantation (ASCT) with rituximab immunotherapy.Forty patients were matched in a 2:1 fashionfrom a historical cohort of patients with MCL treatedwith conventional combination chemotherapy (CC)at the BC Cancer Agency (BCCA) in Vancouver,Canada.The lymphoma database at the BCCA was screenedin order to identify pairs of patients matched with thephase II patients from Toronto for stage of disease,gender and age (± 5 years). Patients treated with ananthracycline (n = 35) or cyclophosphamide/fludarabine(n = 5)-based regimen were preferentially selected.If more than two patients in the BCCA cohort met all thecriteria for matching a patient from Toronto, the twoindividuals with the longest follow-up were chosen.Eighteen of the 20 patients who received up-frontASCT and rituximab remain alive and well withoutevidence of recurrence at a median of 23 months fromdiagnosis; one patient relapsed 2 years post-ASCT,and one died of unrelated causes 7 months postASCT. Of the 40 patients treated with CC, with amedian follow-up of 71 months, 24 have relapsed and26 have died. Median overall and progression-freesurvival were superior in the patients treated withASCT compared to those treated with CC (medianprogression-free survival [PFS]: not reached vs1.3 years,

P

= .0002; median overall survival [OS]:not reached vs 4.02 years,

P

= .058).CONCLUSION: Although follow-up of the ASCTgroup is short, early results suggest that patients withadvanced-stage MCL treated with high-dose therapyand rituximab had statistically significantly betterPFS than patients treated with CC, and there was alsoa trend toward improved OS.

Abstract #606

Marked Antitumor Activity ofRituximab Plus Thalidomide inPatients With Relapsed/ResistantMantle Cell Lymphoma

J. Drach, H. Kaufmann, A. Puespoek, A. Bankier,E. Urbauer, A. Chott, M. RadererDepartment of Medicine I; Department of Medicine IV;Department of Radiology; Department of Pathology,University Hospital, Vienna, Austria

Mantle cell lymphoma (MCL) represents a distinctentity within the non-Hodgkin's lymphomas that isdifficult to treat by conventional treatment approaches.Since the microenvironment plays an importantrole for growth and survival of malignant B cells, wehypothesized that a treatment strategy targeting boththe tumor cell and the microenvironment could beactive in MCL.We therefore initiated a phase II clinical trial toevaluate the toxicity and efficacy of rituximab (Rituxan)plus thalidomide (Thalomid) in patients withCD20-positive MCL who relapsed after or did notrespond to standard CHOP (cyclophosphamide [Cytoxan,Neosar], doxorubicin HCl, vincristine [Oncovin],prednisone) or CHOP-like chemotherapy.Rituximab was administetred at 375 mg/m

²

on days1, 8, 15, and 22; thalidomide was given orally, witha daily dose of 200 mg starting on day 1 and a doseescalation to 400 mg on day 15. Thalidomide wascontinued after completion of rituximab as maintenancetherapy until progression or relapse. Up to now,11 patients are evaluable (7 patients at first relapse, 3patients at ≥ second relapse, and 1 patient with MCLprimary resistant to CHOP). Median age of patientswas 68 years (range: 50-74), and time from initialdiagnosis ranged between 6 and 53 months (median:21 months). According to the International PrognosticIndex, 7 patients were at intermediate high or highrisk.In 5 patients, fever and chills (grade 1/2) associatedwith the administration of rituximab were observed,but no unexpected side effects of rituximab wereencountered by the addition of thalidomide. Adverseeffects of thalidomide included fatigue, constipation,and peripheral neuropathy. Thromboembolic eventswere observed in two patients, and one patient hadsevere neutropenia associated with thalidomide resultingin discontinuation of treatment.Rituximab plus thalidomide induced an objectiveresponse in 10 of 11 patients (3 CR including thepatient who failed to respond to CHOP, 7 PR), and 1patient had stable disease. Tumor responses wereobserved both at nodular and extranodular manifestationsof the disease (gastrointestinal tract). Remissionswere found to be durable: time to progression(TTP) in CR patients was 20.5, 17+, and 10+ months,respectively, which was longer than the precedingTTP after chemotherapy (17, 15, and 3 months,respectively). TTP of PR patients was 24+, 18, 17+,15+, 9+, and 7 months, respectively.CONCLUSION: These results suggest that rituximabplus thalidomide has marked antitumor activityand a favorable toxicity profile in patients with relapsedand CHOP-resistant MCL.

Commentary on Abstracts
#1399, #3084, and #606
Bruce D. Cheson, MD

The therapy of mantle cell lymphoma providesone of the most difficult challenges in themanagement of NHL. Mantle cell lymphomaexhibits the worst characteristics of NHL in thatthese lymphomas are often aggressive, yet theyare incurable with currently available therapy(

Teodorovic et al: J Clin Oncol 13:2819-2826,1995

).Whereas the incorporation of an anthracyclinein the chemotherapy regimen increasedthe overall and complete response rates, theimpact on survival is unclear. The regimen thathas been associated with the best reportedoutcome is the intensive hyper-CVAD regimen(cyclophosphamide, doxorubicin [Adriamycin],vincristine, dexamethasone) followed by stemcell transplantation or combined with rituximab(

Khouri et al: J Clin Oncol 16:3803-3809,1998

).

Rituximab has demonstrated about a 30%response rate as a single agent in mantle celllymphoma; however, these responses are notdurable (

(Foran et al: J Clin Oncol 18:317-324,2000; Ghielmini et al: Ann Oncol 11:S123-S126, 2000

). Nevertheless, the addition of rituximabto the aggressive hyper-CVAD regimenhas shown promising results(

Romaguera et al:Blood 98:726a[abstract 3030], 2001

).

At the ASH meetings, a number of regimensincorporating rituximab for MCL provided interestingresults. Fludarabine has modest single-agent activity in MCL (

Decaudin et al: JClin Oncol 16:579-583, 1998; Foran et al: JClin Oncol 17:546-553, 1999

), yet may be moreeffective in combination with any one of anumber of other agents. Levine and coworkers(abstract #1399) treated 13 patients with fludarabine,mitoxantrone, and rituximab, 38% ofwhom had disease that had relapsed or wasrefractory to prior therapy. Eleven of the 12evaluable patients attained a complete remission.

The German Low-Grade Lymphoma StudyGroup presented the results of two randomizedstudies, one for relapsed and refractory patients,the second as initial therapy. Their chemotherapytreatment regimen consisted offludarabine, cyclophosphamide, and mitoxantrone(FCM) alone or in combination withrituximab (FCM-R). In the former group ofpatients, the response rate with FCM-R was65% vs 33%, with a significantly longer progression-free survival. In the untreated group,the difference was less apparent because theresponse rates in both arms were so high-95%with FCM-R compared with 85% for FCM alone. The progression-free survival had notyet been reached with FCM-R and was 435 dayswith FCM. Because of the relatively short follow-up and the lack of overall survival data, thisapproach cannot yet be accepted a new "standard"for this difficult disease.Mangel and coworkers (abstract #3084)published an analysis of 20 MCL patients whounderwent autologous stem cell transplantationwith rituximab and compared the resultswith 40 patients from a historical cohort ofMCL patients matched in a 2:1 fashion whowere treated with conventional combinationchemotherapy (CC). The median relapse-freesurvival favored the antibody-treated group,with a trend toward a survival advantage.In an interesting study, rituximab was combinedwith thalidomide (Thalomid) in an attemptat treating the tumor as well as themicroenvironment, although without data presentedto support this hypothesis (abstract #606).Of 11 relapsed and refractory patients therewere three complete responses and seven partialresponses, with a time to progression in thecomplete responders from 10+ to 20.5 monthsand 7 to 24+ months for the partial responders.Of note, as has been observed with rituximabretreatment and with tositumomab/iodine-131tositumomab (Bexxar) therapy, response durationwas longer than with the previous treatmentregimen. The explanation for this repeatedobservation is unknown.These results must be interpreted with somecaution because the follow-up is uniformly brief.A recent study combining rituximab with CHOPobserved a high response rate (48% completeresponse, 96% overall response), but of a relativelybrief duration (median progression-freesurvival of 16.6 months), even in patients whoexperienced a molecular response(

Howard etal: J Clin Oncol 20:1288-1294, 2002

).

Hodgkin's Disease

Abstract #3066

Phase II Trial of Rituximab inPatients With Relapsed CD20-PositiveHodgkin's Lymphoma:An Update From the GermanHodgkin's Lymphoma Study Group

H. Schulz, U. Rehwald, M. Reiser, T. Ruediger,F. Morschhauser, V. Diehl, A. EngertDepartment I of Internal Medicine, University of Cologne,Germany; Department of Pathology, University of Wrzburg,Germany; Centre Hospitalier Regional, Universitaire deLille, France

Lymphocyte-predominant Hodgkin's disease(LPHD) presents as early-stage disease with slowprogression and excellent initial response to conventionaltreatment. Nevertheless, many relapses occuryears after the initial diagnosis. While only the minorityof cases of classical Hodgkin's disease (HD) cellsare CD20-positive, all patients with LPHD expressthis B-cell marker in high density on their malignantcell population.With this background the German Hodgkin's LymphomaStudy Group (GHSG) conducted an internationalmulticenter phase II clinical trial to evaluaterituximab in patients with LPHD or other CD20-positive subtypes of HD at first or higher relapse.Lesions were classified as CD20 positive when theCD20 antigen was expressed on more than 30% ofmalignant cells. All histologic slides were reviewedby an independent expert panel consisting of sixreference pathologists. So far 17 patients were treatedwith intravenous infusions of rituximab at 375 mg/m

²

weekly for four doses. All patients had at least oneprior therapy (median: 2). LPHD was diagnosed in 12patients, CD20-positive classical HD in 3, and HDtransformed to T-cell rich B-cell lymphoma in 2patients at study entry. Median age was 40.5 years(range: 18-51 years).The median time since first diagnosis was 9 years(range: 0.5-21 years). Nine patients had stage I/IIdisease; seven were suffering from B symptoms. Sizeof lesions ranged between 1.5 and 4.5 cm in theirgreatest diameter. Adverse events, such as fever,chills, rhinitis, and nausea, were usually transient andof mild to moderate grade allowing outpatient treatmentin all cases. The overall response rate was 88%(15/17 evaluable patients), with 10 complete remissions,5 partial remissions, and 2 progressive disease.Of 15 responders, 10 were in remission at a medianfollow-up of 12 months. The median duration ofresponse amounts to 32 months.CONCLUSION: In this study, rituximab singleagenttherapy proved to be safe and showed significantclinical activity in patients with LPHD. Asfreedom of treatment failure and overall survival ofLPHD patients have not been significantly improvedby intensified chemo- and/or radiotherapy, rituximabmight be a nontoxic and efficient treatment alternativefor this group of patients.

Commentary on Abstract #3066
Bruce D. Cheson, MD

At a previous ASH meeting, there were severalabstracts demonstrating the activity ofrituximab in CD20-positive Hodgkin's disease(

Lucas et al: Blood 96:831a, 2000; Rehwald etal: Blood 96:729a[abstract 3153], 2000; YounesA, et al: Blood 96:733a[abstract 3168], 2000

).The response rates were high; however, responsesdid not appear to be durable even inpreviously untreated patients. At the most recentASH meeting, the German group updatedthe results of their experience (abstract #3066).They had now used rituximab as a single agentto treat 17 patients with CD20-positive lymphocytepredominant Hodgkin's disease failing afterat least one prior treatment. The responserate was 88%, two-thirds of which were completeremissions. The median duration of responsewas now 32 months, which was moreencouraging than with the preliminary data.Studies are in development to incorporate rituximabinto combinations with chemotherapyagents for those CD20-positive patients.

Predictors of Response

Abstract #1222

Predicting Rituximab Response ofFollicular Lymphoma Using cDNAMicroarray Analysis

S. P. Bohen, O. Troyanskaya, O. Alter, R. Warnke,D. Botstein, P. O. Brown, R. LevyOncology Division; HHMI & Department of Biochemistry;Department of Genetics; Department of Pathology, StanfordUniversity School of Medicine, Stanford, California

The recent introduction ofrituximab (Rituxan), a monoclonalantibody targetingthe B-cell surface proteinCD20, has revolutionized thetreatment of follicular lymphoma(FL). The mechanismof rituximab action remainsunclear, and data from in vitrostudies support variouspotential mechanisms, includingcomplement-mediated lysis, antibody-dependentcell-mediated cytotoxicity, and induction ofapoptosis.In this study, we analyzed the expression of morethan 20,000 genes in diagnostic tumor specimens ofpatients who subsequently received rituximab fortreatment of FL to identify genes expressed differentiallyin rituximab responders vs nonresponders. Theresulting gene expression data were analyzed usinghierarchical cluster analysis to examine similaritiesin overall gene expression patterns and statisticaltechniques to evaluate the significance of differencesin expression level of individual genes. We havecompleted analysis of gene expression in FL lymphnodes derived from 24 patients.Hierarchical cluster analysis of the 16 initial patientsamples revealed that gene expression patternsdivide the tumors into two groups. One group clusteredwith normal lymphoid tissues from tonsil andspleen; rituximab nonresponders were contained withinthis group. Conversely, rituximab responders werepredominately found in the second group (

P

= .002).Classification of the genes separating responders andnonresponders by known biologic function revealedthat many of the genes with increased expression innonresponders appear to play a role in cellular immu-nity, suggesting a difference in the antilymphomaimmune response between responders and nonresponders.To extend this observation, gene expression in thefull cohort of 24 patient samples was analyzed on theLymphochip, a cDNA microarray enriched for genesexpressed in lymphoid cells, or genes known orsuspected to be important in the immune response orcancer, as previously described (

Nature 403:503-511, 2000

). The Wilcoxon rank sum test was used toidentify more than 100 genes expressed at significantlydifferent levels (

P

< .005) in rituximab respondersvs nonresponders. Most of the genesidentified showed increased expression in rituximabnonresponders and more than a third of these genesare involved in the cellular immune response.To test the hypothesis that genes involved incellular immunity are more highly expressed in rituximabnonresponders, samples were analyzed by hierarchicalclustering based on expression of genesregulated during activation of macrophages by bacterialpathogens in vitro (

Proc Natl Acad Sci U S A99:1503-1508, 2002

). Surprisingly, hierarchical clusteringbased on the expression of 136 macrophageactivation genes separates rituximab responders andnonresponders (

P

< .013).CONCLUSION: Our findings indicate that biologicdifferences are detectable in lymph nodes of FLpatients using microarray analysis of gene expression.Analysis of a larger, independent cohort isrequired to verify our findings.

Abstract #1400

A Phase II Trial of IndividualizedPharmacokinetic Dosing of RituximabMaintenance for Patients WithCD-20-Positive LymphoproliferativeDisorders

L. Gordan, W. B. Grow, R. C. Braylan, N. P. Mendenhall,V. Douglas, A. Pusateri, L. Balasubramanian, J. W. LynchDepartment of Medicine; Department of Pathology;Department of Radiation Oncology, University of Florida,Gainesville, Florida

Chimeric monoclonal antibody therapy againstCD20 has proven to be an important strategy in themanagement of B-cell lymphoproliferative disorders(LPD), yet the optimal dosing paradigm remainsunclear. The association of serum rituximab (Rituxan)concentration and antitumor response in recurrentlow-grade non-Hodgkin's lymphoma (LGNHL) hasbeen reported. Large phase III clinical trials arecurrently ongoing to discern the utility of maintenancetherapy in B-cell NHL.Our objective was to determine the feasibility ofindividualized pharmacokinetic (PK) dosing of rituximab,evaluate response rates (RR), time to treatmentfailure (TTF), and progression-free survival (PFS),and compare with published historical controls.Patients with B-cell LPD (except small lymphocyticlymphoma) who had failed at least one priortherapy and with Eastern Cooperative Oncology Groupperformance status ≤ 2 were eligible. All patientswere given 4 weekly infusions of rituximab at 375mg/m

²

. Serum levels were obtained immediatelyprior to the first dose, after the fourth dose, andmonthly thereafter for 12 months. A repeat rituximabbolus of 375 mg/m

²

* 1 was given when serum levelsfell to < 25 μg/mL up to 12 months of follow-up.Thirty-one patients were enrolled with the followinghistologies: 18 follicular (16 grade 1, 1 grade 2, 1grade 3), 4 diffuse large cell, 2 transformed low grade,1 mantle cell, 3 mucosa-associated lymphoid tissue(MALT), 2 lymphoplasmacytic, and 1 low grade(LG). One patient withdrew and 2 are not evaluable.The median age was 59 years, 9 patients (29%) weremale, and 26 (84%) had stage III or IV disease.Patients were treated with a median of 2 previoustherapies (range: 1-5).Twenty-eight patients are currently evaluable forresponse. When grouped according to grade therewere 21 evaluable patients with LGNHL and 7 withintermediate grade NHL (INHL). The overall responserate (CR + PR) for the entire group was 57%,stable disease 25%, and primary refractory disease18%. Response rates according to group were 62%for LGNHL and 43% for INHL. To date, 12 patientshave progressed, 7/7 with INHL and 5/21 with LGNHL(

P

= .0009). Seventy-six percent of patients withLGNHL are still progression-free, with median follow-up of 12.5 months.Preliminary longitudinal PK data for at least 6months is available in 10 patients (data collection stillongoing for others). Median rituximab plasma levelswere 407 μg/mL (range: 240-649 μg/mL), 131 μg/mL (range: 28-386 μg/mL), 39 μg/mL (range: 3-131μg/mL), and 17 μg/mL (2.4-78.5 μg/mL) after thefourth dose, 1 month, 3 months, and 6 months followup.Of the 10 evaluable patients, eight required at least 1, 5 required at least 2, and 1 required 3 repeatdoses of rituximab to maintain serum level > 25 μg/mL. Median time for the first repeat dose was 5months (range: 4-9 months), second repeat dose was9 months (range: 6-10 months), demonstrating theefficacy of the single-dose strategy. Further datacollection and maturation will facilitate the determinationof a rational maintenance schedule for subsequentclinical trials.CONCLUSION: Measurement of rituximab serumlevels is feasible in the clinical setting. IndividualizedPK dosing produces excellent objectiveresponse rates and prolonged progression-free survivalin heavily pretreated patients with LGNHL,compared to historical controls. Single-dose re-treatmentprovides adequate serum antibody levels formost patients. Updated PK and response data will beavailable for presentation.

Abstract #1466

Mcl-1/Bax Ratio Is a Predictor forOutcome of Rituximab Therapy inPatients With Chronic LymphocyticLeukemia

S. Kitada, D. Young, M. Pearson, I. W. Flinn, C. A. Shinn,J. C. Reed, J. C. ByrdApoptosis, The Burnham Institute, La Jolla, California;Hematology-Oncology, Ohio State University, Columbus,Ohio; Hematologic Malignancies, Johns Hopkins University,Baltimore, Maryland

Rituximab (Rituxan), achimeric anti-CD20 antibody,is effective as a singleagent in chronic lymphocyticleukemia (CLL) and alsoappears to chemosensitizethese cells to the effectsof chemotherapy. We previouslyreported that rituximabinduces apoptosis ofB-cell CLL (B-CLL) cellsvia a caspase-9 pathway.As part of a single-agent trial of thrice-weeklyrituximab in CLL patients, we obtained pre- andposttreatment blood samples on days 1 and 3 from 11patients from whom adequate posttreatment samplingwas possible, as well as another 10 patientsfrom whom adequate pretreatment sampling waspossible. Real-time processing of cells occurred andprotein cellular lysates were made. Assessment ofserial changes in select apoptosis-related proteinswhose expression previously has been noted to predictchemosensitivity in vivo were performed, andthese data correlated with clinical outcome.In this trial, 10 patients achieved partial response(PR)(10/21, 48%), while 11 patients were nonresponders(NR)(11/21, 52%). Serial expression ofBcl-2, Bax, Mcl-1, and XIAP protein demonstratedthat neither Bcl-2 nor Bax protein levels changedfollowing treatment with rituximab. In contrast, expressionof XIAP declined significantly in 9 of 11patients examined at 1 day posttreatment (

P

= .05)and at day 3 (

P

= .023). Although not significant, asimilar decline was noted with Mcl-1 protein levels in8 of the 11 patients examined. Taken together, thesedata indicate that treatment with rituximab favorablymodulates both Mcl-1 and XIAP protein levels invivo, possibly explaining why this agent sensitizesCLL cells to the cytotoxic anticancer agents.CONCLUSION: Furthermore, pretreatment levelsof Bcl-2, Bax, Mcl-1 and XIAP protein levels wereexamined in a total of 21 patients. A trend toward higherpretreatment levels of Mcl-1 and XIAP with failure torespond was noted (

P

= .133, .137, respectively). PretreatmentMcl-1/Bax ratio was significantly lower inpatients who achieved PR than NR (Mcl-1/Bax ratio =0.39 for PR [with a 95% CI of 0.29] vs 0.82 for NR [with95% CI of 0.28]) (

P

= .016, Wilcoxon rank-sum test),indicating that Mcl-1/Bax ratio may be a predictor foroutcome of rituximab therapy in CLL.

Abstract #2147

Interphase Cytogenetics Are Predictiveof Chronic Lymphocytic LeukemiaResponse to Thrice-Weekly RituximabTherapy

L. L. Smith, N. Heerema, M. L. Hackbarth, I. W. Flinn,D. Young, J. H. Proffitt, J. C. ByrdThe Division of Hematology-Oncology, The Ohio StateUniversity, Columbus, OH, USA; Vysis Incorporated,Downer's Grove, Illinois; Johns Hopkins University,Baltimore, Maryland

Select cytogenetic abnormalities such asdel(17p13.1) and del(11q22-q23) have become recognizedas predictive of rapid disease progression andinferior survival in patients with chronic lymphocyticleukemia (CLL). Similarly, abnormalities ofdel(17p13.1) have been associated with lack of responseto alkylator and fludarabine (Fludara) therapy.We sought to determine the impact of the four mostcommon interphase cytogenetic abnormalities, includingdel(17p13), del(11q22.3), trisomy 12, anddel(13q14.3), in 31 CLL patients relative to responseto thrice-weekly rituximab (Rituxan) therapy. Allpatients receiving this therapy had symptomatic CLL.Response assessment was performed utilizing theNCI 96 criteria. Of these 31 patient samples, successfulhybridization of all the probes was possible in 28(90%). The clinical characteristics of these 28 patientsincluded a median age of 64, with 7 (25%)being female. The patients received a median of 3(range: 0-6) therapies, and 15 (54%) were fludarabinerefractory. Advanced stage (modified Rai 3 or 4)was present in 20 (71%) patients.Interphase abnormalities were noted in 25 of the 28patients and included del(13q14.3) [n=16, 57%],del(11q22.3) [n=10, 36%], +12 [n=6, 21%] ,del(17p13.1) [n=5, 18%], and normal [n=3, 11%].Only a minority of each of these occurred as soleabnormalities. We therefore prioritized the abnormalitiesutilizing the criteria published by Dohnerand colleagues (

N Engl J Med 343:1910-1916, 2000

)[del(17p13.1) > del(11q22.3) > trisomy 12 >del(13q14.3)]. Response to rituximab was noted tovary by cytogenetic group: del(17p13.1), 0% [n = 5];del(11q22.3), 66% [n = 9]; del(13q14.3), 86% [n = 7];+12, 25% [n = 4], and normal, 0% [n = 3]. Responsewas significantly lower (P = .05) in patientswith del(17p13.1) as compared to those with otherabnormalities.CONCLUSION: These data suggest that interphasecytogenetics in CLL may be predictive ofresponse to rituximab therapy with divergent responserates noted in the two high-risk [del(11q22.3)and del(17p13.1)] patient groups. Further studiesexamining the importance of interphase cytogeneticsin predicting response to rituximab combination therapyin CLL are indicated.

Commentary on Abstracts
#1222, #1400, #1466, and #2147
Bruce D. Cheson, MD

Since rituximab is effective in a fraction ofpatients with a CD20-positive lymphoid malignancy,and is expensive, it would be potentiallyuseful to have assays that predict response tothat agent in patients with CD20-positive tumors.The Stanford group (abstract #1222) updatedtheir results on node samples of 24 patientswith follicular lymphoma. They assessed theexpression of more than 20,000 genes in 19patients and found that the nonresponders clusteredwith normal lymphoid tissue, tonsil, andspleen. The responders were in a second genecluster. Genes that were overexpressed in thenonresponding group tended to be those with arole in cellular immunity suggesting a differencein antilymphoma immune response. Theysubsequently evaluated all 24 patients using theLymphochip. Many of the genes in the nonresponderswere those involved in the cellularimmune response.Previous studies have suggested a correlationbetween response to rituximab and serumantibody concentration

(Berinstein et al: AnnOncol 9:995-1001, 1998).

Pharmacokineticswere evaluated in the pivotal trial

(McLaughlinet al: J Clin Oncol 16:2825-2833, 1998).

Thehalf-life of the antibody varied with a number offactors, including the antigen burden, as determinedby the number of circulating CD20-positive cells, tumor bulk, tumor response, andperhaps the location of the tumor, antigen disposition,including the rate of modulation andrate of release, and nonspecific immunoglobulinclearance

(Berinstein et al: Ann Oncol 9:995-1001, 1998).

Berinstein et al found a significantcorrelation between serum levels and responsewith accumulation of antibody in patients whoachieved a CR or PR, but not in the nonresponders.Lower serum concentrations were detected in patients with small lymphocyticlymphoma than those with a follicular histology.To apply such observations therapeutically,Gordan et al (abstract #1400) evaluated 31patients who were heterogeneous with respectto B-cell histology. Additional doses were deliveredto maintain blood levels > 25 μg/mL.Seventy-six percent were still progression-freefrom 3 to 24 months. Whether this approachwill improve patient outcome requires a properprospective trial.In patients with CLL, Kitada et al (abstract#1466) looked at the Mcl-1/Bax ratio. Therationale was that rituximab appears to workthrough a caspase-9-mediated pathway. NeitherBax nor Bcl-2 levels changed during therapy.However, XIAP levels decreased, and, to alesser extent, Mcl-1. The ratio of these factorswas lower in responding patients than nonresponders.Cytogenetics are a potent predictor of outcomein CLL(

(Dohner et al: N Engl J Med343:1910-1916, 2000

).Smith et al (abstract#2147) attempted to correlate response to rituximab-based therapy with cytogenetics. Patientswith del (17p13.1) had a lower response ratethan those with other abnormalities. However,the number of patients in each group was smalland other factors that also predict outcomewere not reported.

Other Monoclonal Antibodies

Abstract #1388

Phase I/II Trial of Epratuzumab(Humanized Anti-CD22 Antibody)in Non-Hodgkin's Lymphoma

J. P. Leonard, M. Coleman, J. C. Matthews, A. Chadburn,W. A. Wegener, H. J. Hansen, H. Ziccardi, H. Kapushoc,U. Gayko, A. Cesano, S. Z. Fields, D. M. GoldenbergCenter for Lymphoma and Myeloma and Division ofHematology and Oncology, Weill Medical College of CornellUniversity and New York Presbyterian Hospital, New York,New York; Immunomedics, Inc, Morris Plains, New Jersey;Amgen Inc, Thousand Oaks, California

Epratuzumab (hLL2 [LymphoCide]) is an unconjugated,humanized IgG

₁

monoclonal antibody againstthe B-cell CD22 antigen (a sialoglycoprotein presenton the cell surface of mature B-cells that is rapidlyinternalized upon binding of ligand or antibody), andis being developed for the treatment of non-Hodgkin'slymphoma (NHL). We report extended follow-upresults from a single-center phase I/II trial examiningthe safety, efficacy and pharmacokinetics of epratuzumabin patients with relapsed/refractory NHL.Epratuzumab was administered at escalating dosesfrom 120 to 1,000 mg/m

²

as an intravenous infusiononce weekly for 4 treatments.Between April 1998 and September 2001, 114patients were enrolled and 111 received ≥ 1 dose ofstudy drug (median age: 61 years; range: 19 to 88years). Approximately half had lesions of ≥ 5 cm, halfhad elevated lactate dehydrogenase (LDH) levels,and half had received ≥ 4 prior regimens, all of whichare poor prognostic features. Overall, epratuzumabwas well tolerated and no dose-limiting toxicity wasobserved. Epratuzumab transiently decreased B-cellbut not T-cell counts. No effect on immunoglobulinlevels was observed. Due to the favorable infusionrelatedsafety profile, 95% of infusions were completedin approximately 1 hour. The mean serum half-lifeof epratuzumab was 23 days.Among 51 indolent NHL patients evaluable forresponse across all dose levels (including level 1) andhistologies, 20 (39%) had stable disease and 9 (17.6%)had objective responses (3 complete responses [CR],6 partial responses [PR]). Of 52 evaluable aggressiveNHL patients, 12 had stable disease (23%) and 5(10%) objective responses were noted (3 CR, 2 PR).All clinical responses were observed at the 240 to 600mg/m

²

/wk levels, and in the follicular NHL anddiffuse large B-cell (DLBCL) histologies only. Six of14 (43%) and 3/11 (27%) of follicular patients treatedat 360 mg/m

²

/wk or 480 mg/m

²

/wk, respectively, hadobjective responses. Median duration of response forthis group was 47+ weeks (11 to 99+ weeks), andmedian time to progression was 103+ weeks (35 to107+ weeks) by Kaplan-Meier estimate. For theaggressive NHL responders, median duration of responsewas 38+ weeks (13 to 38+) and median timeto progression was 35+ weeks (23 to 35+).Responders had lower tumor burden, fewer priortherapies, and normal LDH compared with nonresponders.Notably, however, the DLBCL respondershad a median of 5 prior regimens (range 3-8), andincluded patients who relapsed after high-dose chemotherapyand autologous stem cell transplant (3patientts) or were 80+ years old (1 patient) andtherefore ineligible to receive high-dose chemotherapy.Overall, 5 patients continue to be followed inresponse, with a median follow time of 2.3 years(range: 1.4-3.9 years).CONCLUSION: In summary, epratuzumab therapywas well tolerated in patients with relapsed/refractoryNHL at doses up to 1,000 mg/m

²

/wk ( 4), andproduced clinical activity including durable completeresponses in patients with follicular or diffuse largeB-cell histologies. Ongoing and future studies areevaluating the optimal settings for use of this agent inNHL, including the determination of response ratesin subsets of follicular and DLBCL patients (whereclinical activity appears to be higher), confirmation ofoptimal dosing levels, and combination with otheragents, including rituximab and chemotherapy.

Abstract #610

IDEC-114, an Anti-CD80 MonoclonalAntibody for Relapsed or RefractoryFollicular NHL: Phase I/II Study ofSafety, Efficacy, and Pharmacokinetics

M. S. Czuczman, T. E. Witzig, A. Younes, J. O. Moore,J. M. Vose, T. P. Miller, P. C. Grint, B. R. LeighRoswell Park Cancer Institute, Buffalo, New York; MayoClinic, Rochester, Minnesota; M. D. Anderson CancerCenter, Houston, Texas; Duke University, Durham, NorthCarolina; University of Nebraska Medical Center, Omaha,Nebraska; Arizona Cancer Center, Tucson, Arizona;IDEC Pharmaceuticals, San Diego, California

The success of anti-CD20antibody therapy in lymphomahas sparked widespreadinterest in other monoclonalantibodies as targeted cancertherapies. IDEC-114 is aprimatized anti-CD80 antibodywith human IgG1 constantregions and macaquevariable regions. CD80 is acostimulatory molecule normallyexpressed transiently on the surface of activatedB lymphocytes and constitutively on mostB-lymphoma cells. Prior in vitro and in vivo studieshave demonstrated that IDEC-114 specifically bindsB-lymphoma cells and induces antibody-dependentcell-mediated cytotoxicity (ADCC). Preliminary clinicaldata are now available from an ongoing, openlabel,phase I/II, dose-escalating study in patientswith relapsed or refractory, follicular NHL.The purpose of this study is to evaluate the safety,efficacy, and pharmacokinetics of IDEC-114 therapyin this population. Therapy consists of intravenousinfusions of IDEC-114 once weekly times 4. In phaseI, 3 cohorts of 3 to 6 patients were to receive 125, 250,or 375 mg/m

²

doses to determine the recommendedphase II dose (RP2D). Enrollment into phase I hasbeen completed with 9 patients; 3 in each dose cohort.There are 3 males and 6 females with a median age of54 years (range: 45 to 64 years). Disease stage at studyentry included 1 patient with stage I/II and 8 patientswith stage III/IV. The median number of prior lymphomaregimens was 2 (range: 1 to 5).All 9 patients have completed IDEC-114 therapywith no dose-limiting toxicities reported. The antibodyinfusions were delivered over 1 hour in anoutpatient setting and were well tolerated. Studydrug-related adverse events were limited to grade 1and 2. Reductions in tumor burden have been demonstrated,including one partial response at the highestdose. Analysis of the phase I pharmacokinetic datademonstrated that the C

_max

, T

_1/2

, and are under theconcentration-time curve (AUC) increased proportionallywith the administered dose of IDEC-114.Mean C

_max

was 160, 314, and 430 μg/mL for dosecohorts 125, 250, and 375 mg/m

²

, respectively. MedianT

_1/2

was 12.7, 13.2, and 19.9 days, respectively.A noncompartmental analysis of the serum concentrationdata yielded mean AUC values of 4,107,7,030, and 18,837 μgd/mL.CONCLUSION: In summary, these phase I resultssuggest that IDEC-114 may have potential as a safeand effective treatment with favorable pharmacokineticsin patients with relapsed or refractory, follicularlymphoma. Phase II is ongoing and an update ofthe data will be presented at the meeting.

Commentary on Abstracts
#1388 and #610
Bruce D. Cheson, MD

A number of other unconjugated monoclonalantibodies were discussed at ASH. Leonard andcoworkers (abstract #1388) presented an updateof their phase I/II trial of epratuzumab(hLL2 [LymphoCide]) in patients with indolentand aggressive NHL. This anti-CD22 monoclonalantibody was administered to 114 patientsat a median age of 61 years and half withbulky disease. Half had received at least fourprior regimens. The response rate in the 51patients with an indolent or aggressive histologywas 17.6% and 10%, respectively, across alldose levels. However, 36% of patients with anindolent histology responded at the two highestdose levels. It was difficult to determine theresponse rate for the aggressive NHL at theselevels based on the data provided. This antibodyis of interest to pursue not only as a singleunconjugated antibody, but also conjugated toyttrium-90 and in combination with rituximab(

Leonard et al: Blood 98:844a[abstract 3505],2001

).Apolizumab (Hu1D10) is a monoclonal antibodydirected against a polymorphic determinantof HLA-DR present on normal andmalignant B cells. Impressive activity againstfollicular NHL noted in phase I trials was notreproduced in the phase II setting (

(Link et al:Proc Am Soc Clin Oncol 19:24a[abstract 86],2000; Link et al: Blood 98:606a[abstract 2540],2001

). Investigators at the NCI developed aphase I combination trial based on the factthat both antibodies had exhibited activity inB-cell malignancies (

Hegde et al: Blood100:358a[abstract 1389], 2002

).In their study,6 of 14 evaluable patients responded, includingfour who were rituximab-naive; one of two CLLpatients experienced a CR. Nevertheless, thestudy was closed because of toxicity concerns,notably an arterial thrombosis and hemolyticuremic syndrome. This antibody will have alimited future unless these toxicity issues can beresolved.CD80 is an antigen that is transiently expressedon activated B cells and antigen-presentingcells. Of note is that it is constituitivelyexpressed on malignant B cells and is involvedin the costimulation of T cells. A new monoclonalantibody, IDEC-114, is a primatizedantibody that specifically binds to CD80, mediatingboth ADCC and CDC. The antigen isexpressed on a variety of histologic subtypes ofNHL, although with variable levels of expression.It is important to note, however, that itexhibits at least additive activity with rituximabin a preclinical mouse xenograft lymphomamodel to delay disease progession (

Hariharan etal: Blood 98[abstract 2549]).

At the recent ASH meeting, Czuczman et al(abstract #610) presented the results of the firstphase I trial of IDEC-114 in humans. Theantibody was delivered weekly for 4 weeks inescalating doses from 125 to 375 mg/m². At thetime of the presentation, updated from the publishedabstract, there were 23 patients enrolled-3 each at 125 mg/m² and 250 mg/m², and 17 at375 mg/m². Most (70%) patients were in a lowor low/intermediate IPI category, with a medianof 2 prior regimens, and 83% had stage III/IV disease. There were no dose-limiting toxicities,and the protocol has been amended tocontinue dose escalation to 500 mg/m². Themost common adverse events included fatiguein 26% of patients and nausea in 9%. Of interestwas that of 16 patients at the highest dose levelevaluable for response, there were 3 CR (19%).There is now an ongoing phase I trial of thecombination of rituximab and IDEC-114.

Radioimmunotherapy

Abstract #1381

High Response Rates and DurableRemissions in Patients With PreviouslyUntreated, Advanced-Stage, FollicularLymphoma Treated With Tositumomaband Iodine I-131 Tositumomab

M. S. Kaminski, M. Tuck, D Regan, P. Kison, R. L. WahlDepartment of Medical Oncology, University of Michigan,Ann Arbor, Michigan; Department of Nuclear Medicine,The Johns Hopkins Oncology Center, Baltimore, Maryland

Patients with chemotherapy-relapsed or -refractorylow-grade non-Hodgkin'slymphoma (NHL), who hadfew or no remaining therapeuticoptions, have not onlyresponded well to tositumomab/iodine-131 tositumomab(Bexxar) therapy, but havesustained durable remissions.Therapies that are effectivein relapsed or refractory disease are usuallymore effective when used as frontline therapy. Theseobservations prompted its introduction as initial therapyin patients with previously untreated, stage III/IV, low-grade NHL.Between June 1996 and April 1999, 76 patientsreceived tositumomab/iodine-131 tositumomab therapyon a phase II, single-center study that assessedboth its efficacy and safety. The median follow-up is43 months (range: 8-66 months). The patient-specificdosing regimen for tositumomab/iodine-131 tositumomabhas been previously described (

J Clin Oncol

19:3918-3928, 2001).The median age was 49 years (range: 23-69 years);46% were females. Histologic cell types included71% follicular small cleaved cell, and 29% follicularmixed cell. Poor prognostic factors included 100%stage III/IV disease; 63% bone marrow involvement,and 31% elevated lactate dehydrogenase.An investigator-assessed confirmed response wasobserved in 72 of 76 patients (95%). Fifty-six patients(74%) had a confirmed complete response (CR).Forty-five of these 57 patients (79%) remain in CRfrom 30 to 66 months. The median durations ofresponse and the median progression-free survival(PFS) have not been reached. The 5-year PFS was62.3%. DNA samples from baseline bone marrowwere available for polymerase chain reaction (PCR)detection of the major breakpoint region of the t[14;18]translocation in all but three patients. Ninety-fourpercent of the patients who were PCR positive atbaseline and were followed through B-cell recoverybecame PCR negative.As expected, the primary toxicities were myelosuppresionand hypothyroidism. The median neutrophilnadir was 1,300 cells/μL; the median plateletnadir was 83,000/μL. Grade III/IV neutropenia andthrombocytopenia were noted in 34% and 17%, respectively.One patient had an absolute neutrophilcount of below 100 cells/μL; no patient had a plateletcount below 10,000/μL. No patient required platelet orred blood cell transfusions or granulocyte colony-stimulatingfactor (G-CSF [Neupogen]). The estimatedcumulative 4-year incidence rate for the development ofhypothyroidism was 12%. Sixty-three percent of patientsdeveloped human antimurine antibody followingtositumomab/iodine-131 tositumomab.CONCLUSION: Tositumomab/iodine-131 tositumomabproduces a high confirmed CR rate (74%)in patients with previously untreated, advanced-stage,follicular NHL, with acceptable hematologic andthyroid toxicity.

Abstract #1387

Yttrium-90 Zevalin Phase I SequentialDose Radioimmunotherapy Trial ofPatients With Relapsed Low-Gradeand Follicular B-Cell Non-Hodgkin'sLymphoma: Preliminary Results

G. A. Wiseman, J. P. Colgan, D. J. Inwards,I. N. M. Micallef, L. F. Porrata, D. N. Gansen,N. L. Tuinstra, S. M. Geyer, T. E. WitzigMayo Clinic, Rochester, Minnesota

Ibritumomab tiuxetan (Zevalin) radioimmunotherapy(RIT) consists of a murine anti-CD20 monoclonalantibody chemically linked to the therapeuticbeta-emitting radioisotope yttrium-90. Previous clinicaltrials using a singledose of ibritumomab tiuxetanin patients with relapsedor refractory low-gradenon-Hodgkin's lymphoma(NHL) showed the overallresponse rate (ORR) to be80%, with a 34% completeresponse (CR) rate. Additionaldoses of ibritumomab tiuxetan may be able toincrease response rate and time to progression, butthere are no safety data for two or more doses.The phase I trial was designed to determine themaximum tolerated dose (MTD) of ibritumomabtiuxetan that can be delivered by two doses administered12 to 24 weeks apart for patients with relapsedB-cell low-grade or follicular NHL. All patientsreceived the standard dose of 0.4 mCi/kg (maximum32 mCi for all treatments) for the first dose, with thedose escalation being on the second dose. For thistrial, patients were required to have peripheral bloodstem cells harvested and frozen prior to treatment, abone marrow biopsy with < 25% involvement byNHL, absolute neutrophil count (ANC) ≥ 1,500/μL,and platelets ≥ 150,000/μL.Fourteen patients have been enrolled and havereceived the first dose, with 8 having received thesecond dose of ibritumomab tiuxetan. Three patientsreceiving 0.4 mCi/kg followed by 0.2 mCi/kg had noDLT, but 1 of 5 patients at 0.4 mCi/kg followed by 0.3mCi/kg had a DLT (ANC). No patients have requiredstem cell reinfusion. The biologic clearance of ibritumomabtiuxetan from the blood was longer (up to 2*longer blood T1/2) in all patients on the second dosecompared to the first who had a tumor responsefollowing the first dose.CONCLUSION: Preliminary results of a phase Isequential two-dose ibritumomab tiuxetan therapytrial demonstrate that a second dose of yttrium-90ibritumomab tiuxetan can be safely administered topatients with low-grade and follicular B-cell NHL.However, the MTD for the second dose has not beendetermined. Tumor binding of the radiolabeled antibodyis an important factor in blood clearance ofibritumomab tiuxetan resulting in a slower clearanceof ibritumomab tiuxetan in patients with lower tumorvolumes after the initial treatment.

Abstract #1597

Ibritumomab Tiuxetan Dose EscalationFollowed by High-Dose BEAM andAutologous Peripheral BloodProgenitor Cell Transplant inNon-Hodgkin's Lymphoma:Early Outcome Results

J. N. Winter, D. Inwards, W. Erwin, G. Wiseman,A. Rademaker, D.R. Patton, S. Williams, M. Tallman,J. Mehta, S. Singhal, I. Micallef, P. Multani, M. Zimmer,L. Smith, S. Spies, C. White, L. I. GordonRobert H. Lurie Comprehensive Cancer Center, NorthwesternUniversity, Chicago, Illinois; Division of Hematologyand Division of Nuclear Medicine, Mayo Clinic, Rochester,Minnesota; IDEC Pharmaceuticals, San Diego, California

With the goal of increasingprogression-free survival(PFS) and overall survivalin patients undergoing autologousperipheral blood progenitorcell (PBPC)transplant for relapsed orrefractory non-Hodgkin'slymphoma (NHL), we investigatedthe addition of 90Yibritumomab tiuxetan (Zevalin)to standard high-dose BEAM (carmustine[BCNU], etoposide, cytarabine [Ara-C], melphalan)chemotherapyFourteen patients with relapsed or refractory CD20+NHL in cohorts of 3 to 6 patients were treated at dosescalculated to result in increasing irradiation exposure(100, 300, or 500 cGy) to the critical organ (liver,lung, or kidney). On day -22, rituximab (Rituxan) 250mg/m2 was administered followed by the imagingdose of 111In ibritumomab tiuxetan (5 mCi). Imagingwas performed immediately and at 4, 24, 72, and 144hours postinjection; dosimetry was performed on day-15. On day -14, rituximab at 250 mg/m

²

was administeredfollowed by

⁹⁰

Y ibritumomab tiuxetan at aninitial dose calculated to deliver no more than 100cGy to critical organs. On days -6 through -1, patientsreceived standard high-dose BEAM. On day 0, aminimum of 2.0 * 10

⁶

CD34+ cells/kg was infused.Granulocyte colony-stimulating factor (G-CSF) at5μg/kg was prescribed daily beginning on day 0.The median age was 47 years (range: 25-68 years).Four patients had mantle cell lymphoma, 8 haddiffuse aggressive NHL, and 2 had follicular NHL.The median number of prior therapies was 2 (range:2-4). Seven of 14 patients had never achieved acomplete remission, and 3 of 5 patients who achieveda complete remission with initial therapy had resistantrelapse. Grade III/IV toxicities included fever,infection, hemorrhage, nausea, vomiting, diarrhea,and stomatitis, similar to those seen in transplantpatients receiving BEAM alone. While all but 2patients experienced decreases in percentage predictedDLCO (median: 8%; range: 2%-28%), none wassymptomatic. In only one case did the percentagepredicted DLCO fall below 50% (47% predicted at 3months) and this decrease was transient, with recoveryto 83% predicted occuring by 1 year.Engraftment by Autologous Blood and MarrowTransplant Registry (ABMTR) criteria was at a medianof 9 days (range: 8-17 days) for 1,000 granulocytesand 21 days (range: 12-39 days) for plateletrecovery. The maximum tolerated dose has not yetbeen reached and accrual continues at the 500-cGydose level. The median total dose of

⁹⁰

Y ibritumomabtiuxetan was 21 mCi (range: 2.1-56.7 mCi) or 0.24mCi/kg (range: 0.05-0.63). The median total-bodycGy was 40 cGy (range: 5.3-108 cGy). With amedian follow-up of 15 months (range: 2-25 months),there have been 4 deaths at 7, 8, 10, and 25 monthsresulting in 1- and 2-year survivals of 75%. Six of 14patients have failed therapy, with a 2-year progression-free survival (PFS) of 54%.

The median time for PFS has not yet been reached.CONCLUSION: The combination of

⁹⁰

Y ibritumomabtiuxetan and high-dose BEAM chemotherapywith autologous PBPC transplant is both tolerable

Abstract #679

A Phase I/II Trial of High-DoseRadioimmunotherapy With Zevalin inCombination With High-Dose Etoposideand Cyclophosphamide Followed byAutologous Stem Cell Transplant inPatients With Poor-Risk or RelapsedB-Cell Non-Hodgkin's Lymphoma

A. Nademanee, A. Molina, S. J. Forman, N. Kogut,D. Yamauchi, A. Liu, C. A. White, A. RaubitschekHematology/BMT, City of Hope National Medical Center,Duarte, California; BMT Program, Kaiser Permanente-Cityof Hope, Duarte, California; Radioimmunotherapy, City ofHope National Medical Center, Duarte, California;IDEC Pharmaceuticals, San Diego, Californiaand effective.

Because non-Hodgkin's lymphoma (NHL) is radiosensitive,total-body irradiation (TBI) has been anintegral component of high-dose therapy for NHL. Toreduce the toxicity seen with TBI and take advantageof targeted radioimmunotherapy (RIT), we have aconducted a phase I/II trial incorporating high-doseibritumomab tiuxetan (Zevalin) into the combinationwith high-dose etoposide (40-60 mg/kg) and cyclophosphamide(Cytoxan, Neosar) (100 mg/kg) as anew high-dose regimen for patient with B-cell NHL.Study subjects undergo dosimetry study (day 21)with 5 mCi

¹¹¹

In ibritumomab tiuxetan following 250mg/m

²

of rituximab to confirm favorable localizationof isotope followed 1 week later (day 14) by administrationof ibritumomab tiuxetan (40-100 mCi) todeliver a target dose of 1,000 cGy to highest normalorgan, combined with 5 mCi of

¹¹¹

In ibritumomabtiuxetan. Etoposide is given on day 4 and cyclophosphamideon day 2. Stem cells are reinfused when theradiation dose to the reinfused stem cells is estimatedto be below 5 cGy.Between May 2000 and July 2002, 26 patientswere enrolled, with 8 not able to proceed (5 due tolack of uptake at nodal sites with active disease, 2with progressive disease before ibritumomab tiuxetan,and 1 allergic reaction to rituximab). Eighteenpatients (11 males, 7 females), median age 49 (range:34-58), with follicular (n = 9), diffuse large cell (n =7), and mantle cell (n = 2) lymphoma were treated.Disease status at autologous stem cell transplant(ASCT): first CR/PR = 5 (2 mantle cell, 2 required 2chemo regimens, and 1 with residual bulky mass),second CR = 6, relapse = 6, and IF = 1. Mediannumber of prior chemo regimen was 2 (range: 1-6).All but two patients had received rituximab eitheralone (8), or in combination with chemotherapy (8).Five received etoposide at 40 mg/kg while the restreceived 60 mg/kg. The median dose of ibritumomabtiuxetan delivered was 74.9 mCi (range: 33.6-105mCi). The median number of CD34+ cells infusedwas 7.2 * 106/kg (range: 3.1-33.5 * 10

⁶

/kg).The treatment was well tolerated, with mucositis,neutropenic fever, and skin rash being the mostcommon acute toxicites. One patient who had receivedsix prior chemo regimens developed bacterialsepsis, pneumonia, and endocarditis, and had a delayin recovery of platelet count. One patient with historyof atrial fibrillation (AF) developed recurrent AFafter etoposide. One patient had engraftment syndrome,which resolved after a short course of prednisone.There were no transplant-related deaths. Allpatients engrafted; the median time to reach absoluteneutrophil count > 500/μL and platelets > 20,000/μLwas 10 days (range: 8-17) and 18 days (range: 12-123), respectively.One patient died at 4 months after ASCT fromalcohol-induced liver failure. All 7 patients withactive disease at transplant achieved a complete remission.Three were given consolidative involvedfieldradiation to sites of prior bulky mass > 5 cm aftertransplant. Seventeen patients are alive in remissionat a median follow-up of 8 months (range: 1-24). The1-year estimated overall survival and disease-freesurvival are identical at 92% (95% confidence interval= 77%-100%).CONCLUSION: We conclude that (1) high-doseibritumomab tiuxetan can be safely given; (2) theaddition of high-dose ibritumomab tiuxetan to highdoseetoposide, cyclophosphamide, and ASCT doesnot increase transplant-related toxicity, or delay engraftment;and (3) this high-dose combination iseffective in heavily pretreated patients with refractoryB-cell NHL. Further studies with additional patientsand longer follow-up will be required to document thecurative potential of this approach.

Abstract #1386

Ibritumomab Tiuxetan RadioimmunotherapyIs Not Associated With anIncreased Incidence of SecondaryMyelodysplastic Syndrome or AcuteMyelogenous Leukemia

M. Czuczman, T. E. Witzig, I. Gaston, B. S. Skikne,G. Dimitrov, L. I. Gordon, C. Emmanouilides, C. A. WhiteRoswell Park Cancer Institute, Buffalo, New York; MayoClinic, Rochester, Minnesota; IDEC Pharmaceuticals,San Diego, California; Kansas University Medical Center,Kansas City, Kansas; Northwestern University, Chicago,Illinois; UCLA, Los Angeles, California

Ibritumomab tiuxetan (Zevalin) radioimmunotherapy(RIT) consists of a murine anti-CD20 IgG1 kappamonoclonal antibody covalently linked to tiuxetan,which strongly chelates indium-111 for imaging andyttrium-90 for therapy. Yttrium-90 advantages forRIT include a path length of 5 mm, permittingpenetration into bulky tumors and no penetratinggamma emissions, allowing outpatient administrationand minimizing public and family member exposure.We compared the incidence of secondary myelodysplasticsyndrome (MDS) and acute myelogenousleukemia (AML) in patients treated with ibritumomabtiuxetan with the incidence reported in the literaturein similar non-Hodgkin's lymphoma populations.To date, 10 of 770 patients treated with ibritumomabtiuxetan over the past 9 years have developed MDS orAML. Diagnosis of MDS/AML was 4 to 34 monthsfollowing therapy, and 1.5 to 14 years followingdiagnosis. Annualized rates for the development ofMDS/AML were 0.21% and 0.62% per year frominitial NHL diagnosis and from ibritumomab tiuxetantreatment date, respectively. This incidence issimilar to published reports of MDS/AML in NHLpatients without prior high-dose therapy (4% to 8%,1% to 5% per year;

Semin Oncol 14:435-443, 1987;Ann Intern Med 103:195-200, 1985; Haematologica83:481-482, 1998

)Most patients displayed multiple cytogenetic aberrations.Abnormalities involving chromosomes 5 and7, commonly found in patients treated with priorchemotherapy and/or radiation therapy, were themost frequently observed. Two patients had preexistingmorphologic or cytogenetic abnormalities of bonemarrow.

CONCLUSION: Ibritumomab tiuxetan radioimmunotherapywas not associated with an increasedincidence of secondary MDS/AML, though longerfollow-up is necessary to confirm this initial observation.To more fully elucidate the possible contribution from ibritumomab tiuxetan RIT to the developmentof secondary MDS/AML, cytogenetic testing isrecommended in addition to the required pre-ibritumomabtiuxetan bone marrow studies.

Abstract #1384

Patients With Transformed Low-GradeLymphoma Attain Durable ResponsesFollowing Outpatient RadioimmunotherapyWith Tositumomab/IodineI-131 Tositumomab (Bexxar)

A. D. Zelenetz, M. Saleh, J. Vose, A. Younes, M. S. KaminskiDivision of Hematology/Oncology, Memorial Sloan-Kettering Cancer Center, New York, New York; Division ofHematology/Oncology, University of Alabama-Birmingham,Birmingham, Alabama; Lied Transplant Center, Universityof Nebraska Medical Center, Omaha, Nebraska; Departmentof Lymphoma/Myeloma, M. D. Anderson Cancer Center,Houston, Texas; Department of Hematology and Oncology,University of Michigan Medical Center, Ann Arbor,Michigan

Histologic transformationof indolent lymphoma portendsa poor prognosis andrepresents a major therapeuticchallenge. Conventionaltherapy is often ineffective,and the median survival ofpatients is less than 12months. High-dose therapywith autologous stem cell rescuehas likewise had limitedsuccess in the treatment of these patients.Seventy-one patients with a diagnosis of transformedlow-grade non-Hodgkin's lymphoma (NHL)have been enrolled on five tositumomab/iodine-131tositumomab (Bexxar) studies since 1990. A patientwas classified as having transformed low-grade NHLif a biopsy was consistent with histologic transformationat any point during the course of disease prior tostudy entry. Methodology for patient-specific administrationof tositumomab/iodine-131 tositumomabhas been previously described (

J Clin Oncol 19:3918-3928, 2001).

The median age was 59 years (range: 37-80 years).The median time from diagnosis to study entry was 74months (range: 8-334 months), and that from diagnosisof transformation to study entry was 21 months(range: 0-123 months). The median number of priortherapies was 4 (range: 1-11). Beyond transformeddisease, poor prognostic features included 28% bonemarrow involvement, 70% bulky disease (> 5 cm),57% elevated LDH, and 52% modified InternationalPrognostic Index (IPI) score ≥ 3. The median followupis 19.4 months (range: 0.5-101 months). All results are reported on an intent-to-treat basis. Blindedindependent panel-assessed responses were reportedfor all patients.The overall confirmed response rate was 39%,with a median duration of 20 months (range: 10.8months-not reached [NR]). The complete response(CR) rate was 25%, with a median duration of CR of36.5 months (range: 14.7 months-NR). The mediantime to progression for all patients was 4.3 months(range: 3.2-10.2 months), and that for confirmedresponders was 20.2 months (range: 12.4 months-NR). Complete responses were observed among patientswith adverse prognositic features: 14% ofpatients with an IPI score ≥ 3, 25% of patients withbone marrow involvement, and 24% of patients withfour or more prior therapies. Seventeen of 71 (24%)patients had a durable response, defined as a time toprogression of ≥ 12 months. Five patients remain inremission beyond 40 months, the longest at 66 months.CONCLUSION: These results demonstrate thattositumomab/iodine-131 tositumomab is an effectivetherapy for patients with transformed low-grade NHL.

Commentary on Abstracts
#1381, #1387, #1597, #679, #1386,
and #1384
Bruce D. Cheson,MD

Radioimmunotherapy offers another effectivetherapeutic option for a select group ofpatients with NHL. Data with ibritumomabtiuxetan (Zevalin) and tositumomab/iodine-131tositumomab demonstrate that both agents areactive in relapsed and refractory follicular andlow-grade NHL and transformed NHL, and inpatients who have failed rituximab therapy

(Cheson: Blood 101:391-398, 2003).

However,there is considerable room for improving theirefficacy. Approaches that are being consideredinclude using radioimmunotherapy as initialtreatment. Kaminski et al (abstract #1381) updatedtheir experience with 76 patients who hadreceived tositumomab/iodine-131 tositumomabas their initial therapy with a response rateof 95%, including 74% complete remissions.Tositumomab/iodine-131 tositumomab was alsoassociated with a high level of molecular responsesin that 94% became PCR negativefollowing therapy. Myelosuppression and hypothyroidismwere the most common complications.Human antimouse antibodies (HAMA)developed in 63%. Unfortunately, there was nomention of whether any cases of acute myelogenousleukemia/myelodysplastic syndrome(AML/MDS) developed.Other means to optimize the use of radioimmunoconjugates(RICs) include pretargeting

(Press et al: Blood 98:2535-2543, 2001).

Weidenet al

(Cancer Biother Radiopharm 15:15-29,2000)

administered rituximab conjugated tostreptavidin to bind to tumor-associated antigenreceptors, followed by biotin-N-acetyl-galactosamineto remove nonlocalized conjugatefrom the circulation. The tumor to whole-bodyradiation dose was higher than previously observedwithout pretargeting. The toxicity wasreported to be mild. Further study of this approachis warranted.Other investigators have increased the doseor dose density of the RIC with or without stemcell support. In an earlier experience with tositumomab/iodine-131 tositumomab, Kaminskiet al(

Blood 96:1259-1266, 2000

) found thatrepeated doses of the RIC was not successful inimproving the quality of response in patientswith relapsed/refractory follicular, low-gradeNHL. At the ASH meetings, Wiseman et al(abstract #1387) presented the preliminary resultsof a phase I trial to determine the maximumtolerated dose of a second dose ofibritumomab tiuxetan that could be delivered12 to 24 weeks following the first. The initialdose was the standard 0.4 mCi/kg, and one offive patients who was treated at the second doselevel (0.3 mCi/kg) experienced a dose limitingtoxicity. The trial is ongoing to determine themaximum tolerated dose. One concern is thatthe clearance of the second dose may be slowerdue to reduced tumor binding because of aresponse to the initial dose. The result could belimited activity but excessive toxicity from thesecond dose.Press and coworkers used myeloablativedoses of iodine-131 anti-CD20 antibody eitheralone or in combination with chemotherapy(

Blood 96:2934-2942, 2000

).The progressionfreesurvival was longer than with the patients'longest previous remission or from their lastchemotherapy

Liu et al: J Clin Oncol 16:3270-3278, 1998).

Results from a subsequent phase I/II study of tositumomab/iodine-131 tositumomabwith etoposide and cyclophosphamide

(Presset al: Blood 96:2934-2942, 2000)

comparedfavorably with historical patients who receivedconditioning with total-body irradiation andsimilar chemotherapy. In a recent study, Gopalet al

(Blood 99:3158-3162, 2002)

used highdose radioimmunotherapy with stem cell supportand produced encouraging results for patientswith relapsed MCL. This combined-modalityapproach has the potential to replace totalbodyirradiation-based transplant regimens.

To improve on the efficacy of stem cell transplantation,Winter et al (abstract #1597) addedescalating doses of ibritumomab tiuxetan to theBEAM regimen (carmustine [BiCNU], etoposide,cytarabine, melphalan [Alkeran]), supportedby autologous stem cell transplantationin an ongoing trial. They studied 14 patientswith resistant, aggressive NHL. The 2-yearprogression-free survival was 54%. Toxicitiesof note included a decrease in the diffusingcapacity in almost all patients, although it wasnot symptomatic.Nademanee et al (abstract #679) incorporated40 to 100 mCi of ibritumomab tiuxetan intothe high-dose cyclophosphamide/etoposide regimen.Of 26 patients, eight were not able to gothrough the procedure for a variety of reasons.The regimen was reportedly well tolerated, with a1-year estimated disease-free survival of 92%.The potential toxicity of greatest concernwith radioimmunotherapy is secondary myelodysplasticsyndrome (t-MDS)/acute myelogenousleukemia (AML). This complication hasbeen reported to occur with a frequency ofabout 1.5% of patients treated with ibritumomabtiuxetan at a median follow-up of less than 2years(

Witzig et al: Blood 98:606a[abstract2539], 2001

).This frequency may increase withlonger follow-up. The causal factor for thesesecondary malignancies remains to be elucidatedgiven that the population receiving ibritumomabtiuxetan was pretreated with alkylatingagents with or without external beam irradiation.This complication has been reported tooccur in up to 6.3% of patients who receivedtositumomab/iodine-131 tositumomab for relapsed/refractory disease (

Kaminski et al: Blood96:1259-1266, 2000; Kaminski et al: J ClinOncol 19:3918-3928, 2001

).In contrast, there have been no secondaryhematologic malignancies in patients receivingtositumomab/iodine-131 tositumomab as theirinitial therapy (

Kaminski et al: Proc Am SocClin Oncol 19:5a[abstract 11], 2000

).This findingsupports a contribution of prior chemotherapyto this complication. At the ASH meetings,Czuczman et al (abstract #1386) reviewed thedata on the 10 patients who developed t-MDS ofthe 770 treated over the past 9 years. This eventoccurred 4 to 34 months after ibritumomabtiuxetan therapy, with an annualized rate of0.21%. When they reviewed the literature on thefrequency of AML/MDS in NHL, this figure wasconsistent with prior experiences. In fact, two ofthe patients were found to have a preexistingcytogenetic or morphologic abnormality priorto ibritumomab tiuxetan therapy.One potentially clinically important outcomefollowing radioimmunotherapy is the potentialfor long-term disease-free survival that wouldnot be expected from conventional salvage chemotherapyagents. In the published experiencewith 250 patients with follicular low-grade ortransformed NHL treated with tositumomab/iodine-131 tositumomab, the CR rate was 30%,and 70% of patients with a CR remain in CR upto 7.8+ years (

Kaminski et al: Blood 100:356a,2002; abstract #1384

). Such durable responsesin heavily pretreated patients, especially thosewith transformed NHL, are somewhat unexpected.Similar findings have previously beenobserved with ibritumomab tiuxetan in transformedNHL in which some patients remaineddisease-free out to more than 4 years (

Witzig etal: J Clin Oncol 20:2453-2463, 2002

).

Nonmalignant Indications

Abstract #1722

Rituximab for the Treatment ofRefractory Autoimmune HemolyticAnemia in Children

M. Zecca, B. Nobili, U. Ramenghi, G. Amendola, P. Rosito,M. Jankovic, S. Perrotta, P. Pierani, P. De Stefano,M. Regazzi Bonora, F. LocatelliPediatric Hematology and Oncology, IRCCS Policlinico SanMatteo, Pavia, Italy; Department of Pediatrics, II Universityof Naples, Naples, Italy; Department of Pediatrics, Universityof Turin, Turin, Italy; Department of Hematology, OspedaleNocera Inferiore, Nocera Inferiore, Italy; Department ofPediatrics, University of Bologna, Bologna, Italy; Departmentof Pediatrics, Ospedale San Gerardo, Monza, Italy; Departmentof Pediatrics, University of Ancona, Ancona, Italy

Autoimmune hemolytic anemia (AHA) in childrenis sometimes characterized by an unsatisfactorycontrol of hemolysis, requiring repeated courses ofimmunosuppressive therapies. Rituximab (Rituxan),a chimeric anti-CD20 monoclonal antibody, is able todetermine selective in vivo destruction of B lymphocytes,with subsequent abrogation of antibody production,sparing, at the same time, T lymphocytes.We have evaluated in a prospective study the use ofrituximab for treatment of AHA resistant to conventionalimmunosuppressive therapy.Fifteen patients (5 male, 10 female), with a medianage of 2 years (range: 0.3 -14 years), were treated withrituximab at a dosage of 375 mg/m

²

/dose for a medianof 3 weekly doses (range: 2-4). Nine children hadisolated AHA, 1 AHA and concomitant pure red cellaplasia, and 5 Evans syndrome (AHA and immunemediatedthrombocytopenia). All patients had previouslyreceived 2 or more immunosuppressivetreatments, with various combinations of either oralor intravenous steroids, cyclosporine A, azathioprine,cyclophosphamide (Cytoxan, Neosar), andimmunoglobulins. Two patients had been splenectomizedduring the course of the disease. After completingtreatment with rituximab, all children receivedintravenous substitutive therapy with commercialimmunoglobulin preparation.The median follow-up for the 15 patients is 11months (range: 3-22 months). The treatment wasgenerally well tolerated, and only 3 children presentedfever or cough during drug infusion. One childdeveloped varicella 2 months after rituximab administration,which resolved with antiviral therapy. Thirteenpatients (87%) responded to the treatment,showing at least a 1 g/dL increase in hemoglobin anda 50% reduction of absolute reticulocyte count. Twopatients (13%), both affected by isolated AHA, didnot show any improvement after 3 doses of rituximaband were considered nonresponders. Median hemoglobinlevels increased from a pretreatment value of7.7 g/dL (range: 3.5-10 g/dL) to a 2-month posttreatmentlevel of 12 g/dL (range: 6-14 g/dL) (P < .001).Median pretreatment and 2-month posttreatment absolutereticulocyte count was 236 10

⁹

/L (range: 8-750 * 10

⁹

/L) and 10

⁹

* 10

⁹

/L (range: 35-288 * 10

⁹

/L), respectively (P < .01).In the 13 responding patients the median increasein hemoglobin level 2 months after completion oftreatment was 4 g/dL (range: 1-9 g/dL), while themedian decrease in absolute reticulocyte count was -210 * 10

⁹

/L (range: -22 to -600 * 10

⁹

/L). A concomitantincrease in platelet count was observed in patientswith Evans syndrome as well. Three responder patientsrelapsed 7, 8, and 10 months after the firstrituximab infusion, respectively. All 3 children receiveda second course with rituximab and againachieved disease remission.CONCLUSION: Our data indicate that rituximabis effective in reducing or abolishing hemolysis inmost pediatric patients with AHA, and that a sustainedresponse can be achieved in a majority ofcases. No patient experienced relevant early or lateside effects. Recurrence of the disease may occur, buta second treatment with the monoclonal antibodymay be successful in controlling the disease

Abstract #1027

Rituximab Reduces AntiphospholipidAntibody Titers and ImprovesHypercoagulability in PatientsWith Antiphospholipid Syndrome

M. Yamazaki, A. Takami, H. Asakura, S. NakaoCellular Transplantation Biology, Kanazawa GraduateUniversity School of Medicine, Kanazawa, Ishikawa, Japan

Rituximab (Rituxan) reactsspecifically with theCD20 antigen and inducesB-cell depletion. Administrationof rituximab reportedlyinterferes with theproduction of autoantibodiesin several immune diseasessuch as autoimmunehemolytic anemia and idiopathicthrombocytopenicpurpura, and induces their remission. Antiphospholipidsyndrome (APS) is an autoimmune disordercharacterized by recurrent thromboses, spontaneousabortions, and thrombocytopenia. Rituximab mayalso be useful in treating APS, although there havebeen no reports on successful therapy of APS withthis antibody.We treated two consecutive patients (one male andone female) with non-Hodgkin's lymphoma (NHL)associated with APS with rituximab at a dose of 375mg/m

²

IV on days +1, +8, +15, and +22. Changes inlaboratory data including anticardiolipin antibodies(aCL), lupus anticoagulant (LA), and platelet countswere monitored. Patient 1, a 50-year-old Japaneseman, was in partial remission of follicular lymphomaafter four cycles of chemotherapy consisting of fludarabine(Fludara) and cyclophosphamide (Cytoxan,Neosar), and showed thrombocytopenia (33 * 10

⁹

/L)and prolongation of activated partial thromboplastintime (APTT). He was also positive for aCL-IgM (92MPL), antiphosphatidylserine antibody (aPS)-IgM(76 MPL), and lupus anticoagulant, although no otherautoantibodies, such as antinuclear antibodies (ANA),were detected. Subsequently, he underwent rituximabtherapy for his NHL.After the first rituximab infusion, the patient'splatelet count increased gradually to 71 * 10

⁹

/L, andat the completion of four cycles of rituximab infusions,his platelet count were kept around 100 * 10

⁹

/L and his antiphospholipid antibodies (APA), including aCL, aPS, and LA, disappeared. Plasma levels ofprothrombin fragment 1+2 (F1+2), a marker for thehemostatic condition in patients with APS, decreasedfrom 1.32 nM (normal:< 0.8 nM) to 0.34 nM afterrituximab treatment.Patient 2, a 78-year-old Japanese woman withpolyarteritis nodosa complicated by APS, developedNHL (diffuse large B-cell type). She had developeddeep-vein thrombosis and a stroke, and her aCL-IgGand -IgM titers were 42 GPL and 28 MPL, respectively.The plasma level of F1+2 was elevated at 2.89 nM,although her platelet count was normal. Five coursesof CHOP (cyclophosphamide, doxorubicin HCl, vincristine[Oncovin], prednisone) induced completeremission of her NHL, but her APA levels remainedhigh.A half-year later, her NHL relapsed and she wasprescribed rituximab. Four courses of rituximab infusionsinduced partial remission of her NHL. Sincethen, her aCL-IgG and -IgM levels decreased to 12GPL and 8 MPL, and she has not developed anythrombotic events. Plasma F1+2 levels were normalizedto less than 0.25 nM.CONCLUSION: To our knowledge, this is thefirst report on successful therapy of antiphospholipidsyndrome with rituximab. Administration of rituximabled to a rapid decrease in antiphospholipidantibody titers and improvement of clinical symptomsindependent of the outcome of NHL treatment.Prospective clinical trials on rituximab therapy ofAPS are warranted.

Abstract #2696

Long-Term Remission of RefractoryThrombotic ThrombocytopenicPurpura After Rituximab Therapy

H.-M. Tsai, L. A. Gutterman, B. Kloster, K. Shulman,R. Yomtovian, B. J. Silver, D. H. GallinsonDivision of Hematology, Montefiore Medical Center/AlbertEinstein College of Medicine, Bronx, New York; Mid OhioOncology/Hematology Inc, Columbus, Ohio; Blood Centerof Iowa, De Moines, Iowa; Strauss Oncology Center,Louis A. Weiss Memorial Hospital, Chicago, Illinois;University Hospitals of Cleveland, Cleveland, Ohio;St. Vincent Comprehensive Cancer Center, New York,New York

Plasma exchange (PE) is an effective therapy forthrombotic thrombocytopenic purpura (TTP). However,a subset of patients develop frequent relapses orpersistent disease and require long-term PE that isprohibitively expensive (approximately $2,000 pertreatment), debilitating to the patients, and oftenassociated with serious complications such as thrombosisor infection. Recent studies demonstrate thatautoimmune inhibitors of ADAMTS13, a zinc metalloproteasethat cleaves von Willebrand factor in ashear-dependent manner, causes TTP. Persistent inhibitorspredispose the patients to relapses and providea rationale for immunosuppressive treatments.In this study, we describe the experience of usingrituximab (Rituxan), a chimeric anti-CD20 that depletesthe B cells in the circulation and tissues, in 6consecutive cases of refractory TTP. These 6 femalepatients, ranging from 28 to 62 years of age, had ahistory of recurrent TTP for 7 months to 10 years andhave failed other therapies (ASA-dipyridamole, 6patients; corticosteroids, 6 patients; cyclophosphamide[Cytoxan, Neosar] or azathioprine, 5 patients;splenectomy, 4 patients; cyclosporin A [Neoral, Sandimmune],high-dose intravenous immunoglobulin,and staphylococcal protein A [Prosorba] columns, 1patient each).The indication for rituximab therapy includedlong-term PE (duration 6-9 months, 4 patients),recurrent transient ischemic attack (duration 2 years,1 patient), and frequent relapses (4 times in 6 months,1 patient). Analysis of ADAMTS13 activity prior torituximab therapy demonstrated the presence of inhibitorsin each of the 5 patients investigated.Rituximab at 375 mg/m

²

was administered weeklyfor 4 courses in 2 patients and 8 courses in 4 patients.Five patients achieved a complete clinical response,defined as no plasma therapy, normal platelet counts,and freedom from manifestations of TTP. Inone patient, whose platelet counts were below100 * 10

⁹

/L despite daily 4-liter PE, the plateletcounts increased to over 100 * 10

⁹

/L after fourcourses of rituximab therapy, allowing PE to betapered to 2 liters every 2 to 3 days.Among the 5 patients that achieved completeclinical response, four remain in remission after afollow-up of 6 to 28 months; one relapsed at 17months. Laboratory analysis revealed that rituximabtherapy resulted in a decrease of the inhibitor titers in4 of the 5 cases investigated, and the presence ofmeasurable ADAMTS13 activity (peak levels: 0.18-1.07 U/mL) in 3 cases. In the case that was notinvestigated prior to rituximab therapy, the ADAMTS13level was 0.54 U/mL (normal range: 0.79-1.27 U/mL) immediately after rituximab therapy wascompleted; it increased to a maximum of 0.73 U/mLbefore declining slowly to 0.18 U/mL at 26 months.CONCLUSION: Rituximab is an effective therapyfor refractory TTP caused by ADAMTS13 inhibitors,resulting in long-term remission in 5/6 cases.Strategies that minimize removal of the medicationby PE may be critical for optimal responses. Thepotential role of rituximab for relapsing or refractoryTTP deserves further exploration.

Abstract #2703

Rituximab for the Treatment ofRelapsing/Refractory ThromboticThrombocytopenic Purpura

A. Ahmad, A. Aggarwal, D. Sharma, V. Kinsella, H. P. Dave,G. P. Schechter, M. E. RickHematology/Oncology, Washington Hospital Center C-2151,Washington, DC; Hematology Department, GeorgeWashington University and VA Hospital, Washington, DC;National Institutes of Health-Clinical Center, NIH,Bethesda, MarylandHowever, she remains in clinical remission.

Rituximab (Rituxan), a chimeric monoclonal antibodyagainst CD20, has been increasingly recognizedas a useful therapy for antibody-mediated autoimmunedisease. We recently reported its effectivenessfor patients with Factor VIII autoantibodies (

Wiestneret al: Blood, in press

). The finding of antibodiesto Von Willebrand factor-cleaving protease (VWFCP)in nonfamilial thrombotic thrombocytopenicpurpura (TTP) raises the possibility of rituximab astherapy for TTP.Chemnitz et al reported its successful use in 2patients with severe TTP (ASH 2001, abstract 131).We report here responses to rituximab (375 mg/m

²

weekly * 4 doses) in 3 of 4 patients with recurrentTTP allowing plasmapheresis to be discontinued oravoided. Patient 1 is a 53-year-old male, who was firstdiagnosed in 1990 and responded to plasmapheresis.He relapsed 11 years later and initially responded toplasmapheresis (29 exchanges) and steroids. However,he became refractory to plasma exchanges after 37exchanges, which included the use of cryo-supernatantplasma plus high-dose dexamethasone. He respondedto rituximab and vincristine within 8 daysand is in a complete remission for 4+ months withoutany maintenance treatment.Patient 2 is a 56-year-old female who was diagnosedin 1993. She responded to plasmapheresis andstayed in remission until she relapsed 6 years and 8years later. During the first relapse plasmapheresiswas ineffective until the addition of prednisone andvincristine, which resulted in a complete remission.She required maintenance prednisone for 6 months.The second relapse responded to prednisone alone,but an exacerbation 3 months later was refractory tohigher doses of prednisone and vincristine. She respondedto treatment with rituximab within 9 daysand is in complete remission for 6+ months withoutmaintenance therapy. An inhibitor to VWFCP couldnot be detected 2 weeks after the rituximab course.Patient 3 is a 61-year-old female who was diagnosedwith TTP in 1998. She had a chronic coursewith multiple exacerbations resistant to plasmapheresis,vincristine, and pulse cyclophosphamide (Cytoxan,Neosar). Her last relapse occurred 2 years laterand was treated with weekly dexamethasone andplasmapheresis. The addition of rituximab was withoutbenefit and she continues to need weekly plasmapheresis.Patient 4 is a 57-year-old man who was diagnosedin 1974 and recovered completely following steroids,dextran 70, and splenectomy. TTP recurred 27 yearslater, including cerebral involvement requiring 20plasma exchanges and prednisone. Seven monthslater (off prednisone 2 months), TTP relapsed again.Treatment with prednisone, vincristine, plasma, andrituximab resulted in prompt sustained response withoutplasmapheresis lasting for 10+ weeks. An inhibitorto VWFCP was demonstrated prior to rituximabtreatment but not detected 2 weeks after the rituximabcourse but protease levels remained low.CONCLUSION: We conclude that rituximab mayhave a role in patients with inhibitors to VWFCP anddeserves further study in the treatment of relapsing TTP.

Commentary on Abstracts
#1722, #1027, #2696, and #2703
Bruce D. Cheson,MD

The use of rituximab has been expandingbeyond the limits of B-cell malignancies to thenonmalignant hematologic setting. Previousdata have demonstrated responses to rituximabin autoimmune hemolytic anemia and immunethrombocytopenia in adults and children thathas failed to respond to other therapies(

Hegdeet al: Blood 100:2260-2262, 2002; Herold M, etal: Ann Hematol 79:332-335, 2000; Seipelt G etal: Ann Hematol 80:170-173, 2001; Stasi et al:Blood 98:952-957, 2001; Cooper et al: Blood98:521-522a[abstract 2180], 2001; Lee et al:Blood 96:596a[abstract 2560], 2000; Gupta etal: Blood 98:363a[abstract 1529],2001;abstract#1722).

Activity has been demonstratedagainst a circulating anticoagulant to von Willebrandfactor

(Ege et al: Blood 98:536a[abstract2241], 2001)

and in patients with an anti-factorVIII antibody

(Karwal et al: Blood98:533a[abstract 2232], 2001; Wiestner et al:Blood 98:534a[abstract 2233], 2001).

At the ASH meeting, data were presented tosupport a potential role for rituximab in othernonmalignant indications. For example,Yamazaki et al (abstract #1027) described thefirst successful use of rituximab in patients withhypercoagulability secondary to antiphospholipidantibody syndrome. Both patients whoexperienced this coagulopathy also had activeNHL. Treatment with single-agent rituximabreduced the level of antiphospholipid antibody,the lupus anticoagulant, and fragment 1+2, ahemostatic marker in this condition.Thrombotic thrombocytopenic purpura (TTP)is a disorder thought to be caused by autoimmuneinhibitors of ADAMTS13, metalloproteasethat cleaves von Willebrand factor. Historically,treatment has involved plasmapheresis,and occasionally corticosteroids,immunosuppressive agents, or splenectomy.Nevertheless, recurrence of this condition iscommon. At the ASH meetings, Tsai et al (abstract#2696) reported an anecdotal experiencedemonstrating long-term remissions of TTP infive of six consecutive cases. Levels of theinhibitor were decreased in association withrituximab therapy in four of five cases in whichthey were tested. In a smaller experience, alsopresented at ASH, three of four patients withTTP responded to therapy with this antibody(abstract #2703).Whereas these abstracts are very interesting,a number of issues remain to be addressed.First, these reports are primarily anecdotal.Thus, prospective clinical trials are needed tobetter characterize the response rate. Datafrom such trials need to be carefully analyzed toidentify which patients are most likely to benefitfrom rituximab therapy. The mechanisms bywhich rituximab works in these conditions needto be elucidated. Finally, this well-toleratedagent has the potential to replace other more toxicapproaches to autoimmune conditions such ascorticosteroids, intravenous immunoglobulins,and splenectomy. Therefore, where rituximabshould optimally be positioned relative to traditionaltherapies needs to be determined.

Disclosures:

The author(s) have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.