A Phase 2 Study of Yttrium-90 Ibritumomab Tiuxetan (Zevalin) in Patients With Chronic Lymphocytic Leukemia

Abstract

BACKGROUND

Radioimmunotherapy (RIT) with radiolabeled monoclonal antibodies to CD20 produce a high response rate in patients with relapsed lymphoma. Use of this modality in patients with chronic lymphocytic leukemia (CLL) has been hampered by the extensive marrow involvement seen in patients with CLL, which would produce a high risk for marrow aplasia after treatment with RIT. Patients with lymphoma and marrow involvement have been treated with RIT if involved marrow was less than 25% of the total marrow. Thus, we adapted this approach as consolidation therapy in patients with CLL responding to chemoimmunotherapy.

METHODS

Fourteen patients with relapsed CLL either in partial remission or in complete remission but with disease documented by flow cytometry were treated with ⁹⁰Y ibritumomab tiuxetan.

RESULTS

One patient responded and achieved a complete remission but with residual disease detected by flow cytometry. Of note was that grade 3 or 4 hematologic toxicity was seen in 12 of the 13 (92%) evaluable patients, with grade 3 or 4 thrombocytopenia noted in 11 (85%) of the patients. In addition, myelosuppression was prolonged with a median duration of grade 3 or 4 thrombocytopenia of 37 days. Five patients had persistent thrombocytopenia 3 months post-therapy.

CONCLUSIONS

Even in patients with CLL and limited marrow involvement, the use of RIT results in unacceptable hematologic toxicity.

Advances in chemoimmunotherapy have produced high complete response (CR) rates and durable remissions in patients with chronic lymphocytic leukemia (CLL). However, most patients in CR have detectable disease when evaluated by more sensitive techniques such as flow cytometry or polymerase chain reaction (PCR) assays. This correlates with continuous relapse over time and no suggestion of a plateau in progression-free survival curves. Strategies to eradicate minimal residual disease (MRD) after primary therapy may enhance remission duration and prolong progression-free survival.

Radioimmunotherapy (RIT) refers to the therapeutic use of monoclonal antibodies conjugated via a chelator (tiuxetan) to a radioactive nuclide.¹ Two RIT agents have been approved by the US. Food and Drug Administration for the treatment of relapsed CD20 + non-Hodgkin lymphoma (NHL): yttrium Y 90 (⁹⁰Y) ibritumomab tiuxetan (Biogen Idec, Cambridge, Mass) and iodine I 131 tositumomab (GlaxoSmithKline, Research Triangle Park, NC). Both agents bind to the B-cell CD20 antigen, which is present on most B-cell NHL tumors and CLL cells.² The chelation of the radionuclide results in radiation being delivered to nearby cells in addition to those to which the antibody has bound.³ The ibritumomab tiuxetan regimen also includes the unlabeled antibody rituximab; this is given before the radiolabeled antibody to saturate CD20 binding sites on B cells in the peripheral blood and spleen and potentially improves the biodistribution of the radiolabeled antibody.⁴ As ⁹⁰Y has no γ-emission (⁹⁰Y is a pure β-emitter), dosimetry and tumor images cannot be obtained with it and γ-emitting¹¹¹ Inibritumomab is used for this purpose.¹

The initial phase 1–2 studies in patients with relapsed/refractory low-intermediate grade B-cell NHLs established that pretreatment with rituximab before administration of RIC improved biodistribution and that the maximal tolerated dose of ⁹⁰Y ibritumomab was 0.4 mCi/kg (platelet count ≥150,000/mm³) or 0.3 mCi/kg (platelet count between 100,000 to 150,000/mm³).^2,3 In a phase 3 trial in rituximab naive patients with relapsed or refractory low-grade/follicular/transformed lymphoma, ⁹⁰Y ibritumomab RIT was compared with rituximab (375 mg/m² intravenous weekly for 4 weeks).⁴ In this trial, ⁹⁰Y ibritumomab led to a significantly improved overall response rate (ORR, 80% vs 56%; P = .002) and CR rate (30% vs 16%; P = .04) compared with rituximab. Time to progression was similar in the 2 groups (11.2 vs 10.1 months), although time to next therapy for the nontransformed group was higher in the ⁹⁰Y ibritumomab arm.⁴ In another trial reported by Witzig et al, ⁹⁰Y ibritumomab was used to treat patients with follicular lymphoma who were rituximab refractory; this led to a 74% ORR and 15% CR rate.⁵ On the basis of these trials, the United States Food and Drug Administration (FDA), approved ⁹⁰Y ibritumomab for the treatment of patients with relapsed or refractory low-grade, follicular, or transformed B-cell NHL in February 2002.⁶

Myelosuppression was the main side effect seen with ⁹⁰Y ibritumomab. However, its use has been limited to patients with less than 25% marrow involvement to prevent radiation of neighboring normal stem cells and prolonged myelosuppression. Thus, RIT has not generally been evaluated in patients with CLL who have extensive marrow involvement as part of the disease.

In the early 1990s, DeNardo et al reported a pilot study in which ¹³¹I Lym-1 was used in 5 previously treated patients with CLL.¹⁰ All patients achieved partial remission with >50% decrease in lymphadenopathy. Remission duration was short; thrombocytopenia was the major hematological toxicity and 4 of the 5 patients died of infectious complications. In another study reported in abstract form, Kaminski et al pooled data from 4 phase 1–3 studies of the use of ¹³¹I tositumomab RIT in patients with NHL.⁷ They identified 14 patients with small lymphocytic lymphoma (SLL), 11 of whom had <25% bone marrow involvement. Twenty-one percent of patients achieved CR and an additional 42% achieved partial remission (PR) with hematological toxicity seen in <15% of patients.

The objective of the present study was to evaluate whether ⁹⁰Y ibritumomab given to patients with CLL after maximum response to chemotherapy would: 1) eliminate minimal residual disease in patients who had achieved a CR or 2) convert PR to CR.

MATERIALS AND METHODS

In this single center phase 2 study, patients (≥ 18 years) with CLL and either of the 2 following disease characteristics were eligible: 1) Patients who had achieved a CR with chemotherapy (using NCI-Working Group response criteria)⁸ but with documented residual disease by immunephenotyping (residual population of cells [>10%] that are CD5, CD19, and CD20 positive) or who were PCR positive; 2) Patients with CLL who had achieved a PR after chemotherapy but had <25% bone marrow involvement by morphology. Patients were required to have WHO performance status of ≤2, and normal organ function. In addition, patients with impaired bone marrow reserve as indicated by one of the following were not eligible: prior myeloablative therapies with autologous/allogeneic stem cell transplantation; platelet count <100,000 cells/mm,³ hypocellular bone marrow (<10% cellular), marked reduction in bone marrow precursors of one or more cell lines (granulocytic, megakaryocytic, erythroid), or history of failed stem cell collection. The study was approved by the MD Anderson Cancer Center institutional review board, and all patients signed a written informed consent according to the institutional guidelines.

The patients received an initial infusion of 250 mg/ m² rituximab, followed immediately by a fixed dose of 5.0 mCi of ¹¹¹In ⁹⁰Y ibritumomab. The biodistribution of ¹¹¹In ⁹⁰Y ibritumomab was determined from a visual evaluation of whole body gamma images during the first day (2–24 hours) and the second or third day (48–72 hours) after injection. If the biodistribution was altered such that the safety risk was unacceptable, the patient was taken off study and did not proceed with ⁹⁰Y ibritumomab RIT. For patients with acceptable biodistribution, a second infusion of 250 mg/m² rituximab and either 0.4 mCi/kg (platelet count ≥150,000/mm³) or 0.3 mCi/kg (platelet count between 100,000 to 150,000/mm³) of ⁹⁰Y ibritumomab was given 1 week later. The treatment was administered on an outpatient basis. The dose of ⁹⁰Y ibritumomab was capped at 32 mCi. Once initiated on the treatment protocol, follow-up was done with weekly complete blood counts and monthly basic chemistries. At the end of the 12 week period, physical examination and bone marrow aspiration and biopsy were repeated to assess response. Lymph node response was assessed by physical examination.

The patients were followed at 3-month intervals until alternate treatment or death. Toxicity was graded according to the NCI Common Terminology Criteria, version 3.0. The duration of ≥3 grade hematological toxicity was calculated from the date the ≥3 grade toxicity was first noticed to the date of first value in ≤2 grade toxicity.

A response rate of 20% was considered significant. We were willing to accept a rejection error of 5%, and, therefore, at least 14 patients in each category (CR and PR) were planned initially. If a response was observed, then 35 patients will be entered in each group (CR and PR) to determine an accurate response rate. If none of the first 14 patients had achieved response, the trial will be terminated.

RESULTS

Fifteen patients were registered for the trial between August 2002 and March 2003 at MD Anderson Cancer Center, Houston. Fourteen patients received a therapeutic dose of ⁹⁰Y ibritumomab RIT (1 patient had abnormal biodistribution of ¹¹¹In ⁹⁰Y ibritumomab with intense uptake in liver and therefore was taken off protocol and did not receive ⁹⁰Y ibritumomab). An additional patient had Richter transformation to diffuse large B-cell lymphoma (DLBCL) documented within 1 month of treatment with ⁹⁰Y ibritumomab; it was presumed this was present before treatment but no biopsy had been done so the patient was considered inevaluable. The median age was 59 years (range, 46–79 years) (Table 1). All patients were bone marrow flow-cytometry positive for CLL. Eleven patients had ≥1.0 cm size palpable lymph nodes.

Table 1.

Baseline Characteristics of the Patients

Parameter	Median (range), n=14
Age, y	59 (46–79)
Female sex	4
Duration of CLL, y	7 (2–11)
Hemoglobin, g/dL	13.0 (10.0–15.8)
Platelets, ×109/L	209 (95–318)
No. of prior therapies	5 (3–7)
Duration from prior rituximab use, mo	4 (2–16)
Remission status
PR	13
nPR	1

PR indicates partial remission; nPR, nodular partial remission; CLL, chronic lymphocytic leukemia.

All patients had received rituximab as part of their previous chemotherapy regimens and the median duration between prior rituximab use and RIT was 4 months (range, 2 months to 16 months). Eight patients received 0.3 mCi/kg and 6 patients received 0.4 mCi/kg ⁹⁰Y ibritumomab.

Response

Of the 13 patients evaluable for response, 1 patient achieved morphologic CR but with persistent MRD positive marrow; another patient achieved CR in the bone marrow (flow cytometry negative) but had worsening lymphadenopathy, subsequently diagnosed as Richter transformation. The first patient was a 50-year-old man who had received previous treatment with fludarabine, cyclophosphamide, and rituximab in combination. His pre-treatment bone marrow biopsy showed 30% cellularity with 25% marrow space involvement by CLL (interstitial pattern). He received 0.4 mCi/kg ⁹⁰Y ibritumomab and tolerated it well with no infectious complications/hospitalizations. He had grade 2 thrombocytopenia with a nadir platelet count of 71×10⁹/L, lasting for 6 days. No grade ≥3 hematological toxicity was noted. Repeat bone marrow biopsy at 12 weeks showed 40% cellularity with no morphological evidence of CLL. Flow cytometry was still positive for minimal residual disease. Eight months after completing the RIT, he was noted to have progressive disease and required subsequent chemotherapy 11 months after RIT. The second patient was a 62-year-old man with a 10-year history of CLL, who had received previous treatment with chlorambucil, FCR, and alemtuzumab. Pretreatment bone marrow was hypercellular (70%–80% cellularity) with CLL (nodular and interstitial pattern), replacing 20% of the medullary space. At 3 months, he achieved morphological remission in the bone marrow with negative flow cytometry. However, he continued to have worsening lymphadenopathy, the biopsy of which confirmed Richter transformation. He died 2 months later because of progressive disease. No other patient had a response to RIT, including nodal response.

Toxicity

Grade ≥3 hematological toxicity was seen in 12 of the 13 evaluable patients (Table 2). Grade ≥3 thrombocytopenia was noted in 11 patients (85%), including 5 patients (38%) with grade 4 thrombocytopenia. Median nadir platelet count reached was 27×10⁹/L (range, 12–71×10⁹/L) and the median time to nadir platelet count was 45 days (range, 29–90 days). The median duration of grade ≥3 thrombocytopenia ⁹⁰Y ibritumomab treatment was 37 days (range, 7–57 days). Five patients had persistent thrombocytopenia 3 months post-therapy, including 3 patients with progressive disease.

Table 2.

Hematological Toxicity

S. No	Baseline Platelet Count, x109/L	Baseline Hemoglobin, g/dL	Baseline ANC, x109/L	Nadir Platelet Count	Day of Nadir Platelet Count	Nadir Hemoglobin	Day of Nadir Hemoglobin	Nadir ANC	Day of Nadir ANC	For the first 3 mo
										Days Platelets <100x109/L	Days ANC <1000x109/L	Days ANC <500x109/L
1	245	12.7	1.02	47	56	10.1	71	0.11	29	35	63	49
2	252	11.4	1.79	17	56	7.6	84	0.37	70	55*	50	14
3	163	12.3	2.06	29	29	9.5	64	0.05	29	61	28	14
4	210	13.5	3.36	30	50	10.1	87	0.24	50	55*	28	13
5	284	13.2	5.31	62	45	12.2	58	0.48	50	14	20	7
6	318	11.5	10.16	37	43	9	64	1.2	55	30	0	0
7	256	13.6	1.58	15	39	12.5	50	0.54	42	26	17	0
8	194	12.6	4.01	12	42	8.4	45	0.2	42	49	27	8
9	153	14.8	0.24	12	52	8.6	63	NA	NA	66*	NA	NA
10	95	10	0.98	21	90	6.9	62	NA	NA	60*	NA	NA
11	242	15.8	6.18	71	34	13.2	48	2.05	40	6	0	0
12	111	13.6	3.57	27	33	9.7	54	0.53	47	57	8	0
13	208	14.2	7.68	27	63	10.6	70	3.33	63	27*	0	0

ANC indicates absolute neutrophil count; NA, not applicable as low baseline.

^*Persistent thrombocytopenia after 3 months.

Two patients had low baseline absolute neutrophil count (ANC) requiring G-CSF support before starting ⁹⁰Y ibritumomab therapy. Of the remaining 11 patients, 8 patients (72%) developed grade ≥3 neutropenia, including 6 patients (55%) with grade 4 neutropenia. Seven patients required G-CSF support within 3 months after RIT, including the 2 patients requiring G-CSF at baseline. The median duration of grade ≥3 neutropenia was 27.5 days (range, 8–63 days) and grade 4 neutropenia was 13.5 days (range, 7–49 days). Median nadir ANC was 0.48×10⁹/L (range, 0.05–3.33×10⁹/L) and the median time to nadir ANC was 47 days (range, 29–70 days). Three patients required hospitalization (indication: pneumonia, neutropenic fever, and Escherichia coli sepsis) within 3 months after RIT.

Grade ≥3 anemia was observed in 2 patients (15%), with no grade 4 events. Median nadir hemoglobin was 9.7 g/dL (range, 6.9–13.2 g/dL) and the median time to nadir hemoglobin was 63 days (range, 45–87 days).

DISCUSSION

Achievement of MRD negativity is increasingly being recognized as an important objective of current therapies in patients with CLL. On the basis of the success of RIT in patients with NHL, we explored the safety and efficacy of ⁹⁰Y ibritumomab RIT in patients with CLL and limited marrow involvement. Profound myelosuppression was seen in 12 of the 13 patients, and the study was terminated early because of safety reasons. Myelosuppression is the major toxicity of ⁹⁰Y ibritumomab therapy. In an integrated safety summary of trials using ⁹⁰Y ibritumomab in NHL, Witzig et al reported 63% grade ≥3 thrombocytopenia, 60% grade ≥3 neutropenia, and 17% grade ≥3 anemia.⁹ The higher rate of myelosuppression seen in the current study could be attributed to several reasons. All patients in the present study had received rituximab previously at a median duration of 4 months before RIT. It is possible that the CD20 sites in the patients in the current study were still coated with rituximab and this led to increased ⁹⁰Y ibritumomab circulation in the blood and bone marrow, increasing the potential for myelosuppression. In addition, bone marrow involvement is a recognized risk factor for increased hematological toxicity after ⁹⁰Y ibritumomab treatment. Witzig et al reported that any degree of bone marrow involvement led to a significant increase in grade 4 hematological toxicities and the higher the bone marrow involvement, the higher the risk of hematological toxicity.⁹ All patients in our study had bone marrow involvement. Prior therapy with fludarabine has also been associated with a higher likelihood of grade ≥3 hematological toxicities and longer duration of grade ≥3 thrombocytopenia.⁹ All patients in the present study had received fludarabine in the past, again making them higher risk for myelosuppression. In another pooled data analysis of NHL trials, Emmanouilides et al reported that patients receiving ⁹⁰Y ibritumomab as second-line therapy (number of prior therapies ≤1) had significantly lower incidence of grade 4 thrombocytopenia as compared with patients who had received >1 prior therapies (7% vs 16%, P = .02).¹⁰ Patients in the present study had received a median of 5 prior chemotherapy regimens. The majority of patients experienced grade 3 or greater neutropenia and needed G-CSF support. Both patients starting with ANC ≤1.0×10⁹/L were requiring G-CSF at baseline and continue to receive it while on study for persistent neutropenia. Therefore, caution should be exercised in administering ⁹⁰Y ibritumomab in neutropenic patients.

Responses were seen in only 2 patients, including the patient who became flow cytometry negative in bone marrow but soon thereafter had Richter transformation. The response rate is comparable to that seen with single agent rituximab in previously treated patients with CLL.^111–13 The response rate, however, is much lower than that seen with ⁹⁰Y ibritumomab therapy in NHL.^4–6 Morschhauser et al reported use of ⁹⁰Y ibritumomab in DLBCL patients.¹⁸ In that study, patients who had received rituximab-based chemotherapy previously had a much lower ORR compared with rituximab-naive patients (19% vs 53%, respectively). However, 2 other studies in rituximab refractory follicular lymphoma patients treated with RIT have shown ORR of >65%.^8,19 All patients in our study had received rituximab previously and whether improved response rate would be achieved with RIT in rituximab naïve CLL patients remains unstudied. Emmanouilides et al reported that ORR, CR rate, and duration of response were significantly higher in NHL patients who received ⁹⁰Y ibritumomab as second-line therapy as compared with patients who had received >1 prior therapies.¹⁴ The patients in the current trial were heavily pretreated with a median of 5 prior regimens.

In conclusion, use of ⁹⁰Y ibritumomab in patients with CLL led to significant and prolonged myelosuppression and was of limited efficacy.