Treatment of Patients with Acute Myeloid Leukemia with the Targeted Alpha-Particle Nanogenerator Actinium-225-Lintuzumab

Acute myeloid leukemia (AML) is a cytogenetically and molecularly heterogeneous hematologic neoplasm characterized by lack of differentiation and increased proliferation of early myeloid progenitor cells. Data from the Surveillance, Epidemiology, and End Results (SEER) program of the NCI report 5-year survival rates of 47.5% and 8.2% for patients younger than 65 years and 65 years or older, respectively (1). Despite the U.S. Food and Drug Administration approval of nine new agents for AML since 2017, novel agents are still needed to improve long-term outcomes in this difficult-to-cure disease.

Lintuzumab (HuM195) is a humanized monoclonal antibody that targets CD33, a 67-kDa cell-surface glycoprotein expressed on most myeloid leukemia cells. Previous studies demonstrated that lintuzumab can target leukemia cells in patients without eliciting immunogenicity (2), can eliminate minimal residual disease in acute promyelocytic leukemia (3), and can produce occasional remissions in AML (4–6). When added to standard chemotherapy, however, unconjugated lintuzumab was unable to improve response rate or survival in two randomized trials (7, 8). Therefore, to increase its modest single-agent activity in AML, lintuzumab was initially conjugated to the β particle-emitting isotopes iodine-131 or yttrium-90. Studies with these agents demonstrated that β particle-emitting conjugates of anti-CD33 constructs could eliminate large leukemic burdens but required hematopoietic cell transplantation (HCT) due to profound myelosuppression (9, 10).

In the effort to enhance tumor cell killing and simultaneously reduce the nonspecific cytotoxic effects seen with β-emitters, investigation of α-particle therapy was undertaken. Compared with β-particles, α-particles have a shorter range (50–80 vs. 800–10,000 µm) and a higher linear energy transfer (100 vs. 0.2 keV/µm; ref. 11). As few as 1 or 2 α-particles can kill a target cell. We have previously reported our experience with the α-emitter conjugate bismuth-213 (²¹³Bi)-lintuzumab, which demonstrated significant antileukemic activity, including complete responses following cytoreduction in AML (12, 13). However, the widespread clinical utility of radioimmunotherapy with ²¹³Bi is limited by its short 46-minute half-life and the need for an onsite ²²⁵Ac/²¹³Bi generator.

²²⁵Ac is an isotope of actinium with a half-life of 10 days. In its decay to stable ²⁰⁹Bi, ²²⁵Ac generates francium-221 (²²¹Fr), astatine-217 (²¹⁷At), ²¹³Bi, and lead-209 (²⁰⁹Pb), yielding 4 α-particles (Fig. 1). ²²⁵Ac can be conjugated to various antibodies using the bifunctional chelate p-isothiocyanatobenzyl-1,4,7,10-tetraaczacyclododecane-1,4,7,10-tetraacetic acid (DOTA-SCN). One solution to the limitations of ²¹³Bi is to use lintuzumab as a vehicle to deliver ²²⁵Ac to the target leukemia cell. In this way, ²²⁵Ac serves as an in vivo α-particle generator, such that the decay produces a cascade of 4 α emissions. ²²⁵Ac-labeled immunoconjugates kill in vitro at radioactivity doses at least 1,000 times lower than ²¹³Bi analogues and prolong survival in mouse xenograft models of several cancers (14). Treatment with ²²⁵Ac-labeled prostate membrane-specific antigen (PSMA) ligand has shown significant activity in advanced, metastatic prostate carcinoma (15, 16). To determine the maximum-tolerated dose (MTD), safety, pharmacology, and biological activity of ²²⁵Ac-lintuzumab in AML, we conducted a first-in-human, phase I dose-escalation trial.

²²⁵Ac-lintuzumab is a radioimmunoconjugate composed of humanized anti-CD33 monoclonal antibody lintuzumab (Protein Design Labs, Inc.) linked to ²²⁵Ac, via the bifunctional chelate DOTA-SCN. ²²⁵Ac, supplied by Actinium Pharmaceuticals, was obtained from Oak Ridge National Laboratory. ²²⁵Ac was conjugated to lintuzumab using a 2-step labeling method (17). First, ²²⁵Ac was labeled to DOTA-SCN (Macrocyclics and Dow Chemical Company) species at pH 4.5 to 5 in acetate buffer at 55°C to 60°C for 30 minutes. Then, ²²⁵Ac-DOTA-SCN was mixed with lintuzumab in carbonate buffer at pH 8.5 to 9 at 37°C for 30 minutes. The final product was purified by size exclusion chromatography using a 10-mL Bio-Rad 10DG column and 1% human serum albumin (HSA). Typical reaction yields were 10% ± 5%, which were sufficient for this study. Antibody doses were adjusted to maintain specific activity levels of 1.23 to 5.92 kBq/μg. ²²⁵Ac activity and antibody mass were adjusted per body weight: < 74 kBq/kg were labeled on 15 µg/kg of lintuzumab. ²²⁵Ac activities 74 and 111 kBq/kg were labeled on 20 µg/kg of lintuzumab. ²²⁵Ac activities of 148 kBq/kg were labeled on 25 µg/kg of lintuzumab. Filtered ²²⁵Ac-lintuzumab was diluted in 50-mL normal saline with 1% HSA for intravenous (i.v.) infusion through an online 0.2-µm filter. The final product was infused i.v. over 15 to 30 minutes using an infusion pump.

Eligible patients included those in three restricted categories of AML, chronic myeloid leukemia (CML), or myelodysplastic syndromes (MDS). Eligible patients with AML had relapsed AML or AML refractory to 2 courses of standard induction chemotherapy or to 1 course of high-dose cytarabine-based induction chemotherapy. Eligible patients with CML were in accelerated phase or myeloid blast crisis progressing after imatinib and a second-generation tyrosine kinase inhibitor. Eligible patients with MDS had MDS with excess blasts or chronic myelomonocytic leukemia with International Prognostic Scoring System (IPSS) score ≥ 2.5 and were refractory to or relapsed after a hypomethylating agent. More than 25% of the patient's bone marrow blasts were required to express CD33. No antileukemic therapy was administered for 3 weeks before study entry except for hydroxyurea, which was discontinued before treatment with ²²⁵Ac-lintuzumab. Concurrent use of either oral or i.v. antibiotics was allowed. Entry criteria included adequate renal and hepatic function as demonstrated by a serum creatinine ≤ 1.5 mg/dL, a calculated creatinine clearance > 60 mL/min, < 1 g urinary protein/24 hours, bilirubin ≤1.5 mg/dL, and alkaline phosphatase and aspartate aminotransferase ≤ 2.5 times the upper limit of normal. Patients could not have detectable antibodies to lintuzumab or active central nervous system involvement by leukemia. Patients were treated from February 2006 to March 2012 at Memorial Sloan Kettering Cancer Center on a protocol approved by the Center's institutional review board and registered with ClinicalTrial.gov (NCT00672165). All subjects gave written informed consent according to the Declaration of Helsinki.

To limit “first pass” binding of the radiolabeled antibody in the liver, unlabeled lintuzumab was given before administration of ²²⁵Ac-lintuzumab to improve targeting of isotope to the bone marrow and other disease sites. Patients received unlabeled lintuzumab at a dose of 250 µg i.v. over 15 to 30 minutes. Fifteen to 30 minutes following infusion of unlabeled lintuzumab, ²²⁵Ac-lintuzumab was given as an i.v. infusion over 15 to 30 minutes. Five dose levels of ²²⁵Ac-lintuzumab were administered: 18.5, 37, 74, 111, or 148 kBq/kg. Initially, doses were determined based on actual body weight, but due to concerns for prolonged myelosuppression, the last six patients were treated based on adjusted body weight (ABW) if actual weight was greater than ideal body weight (IBW). IBW was calculated using standard formulas developed by Devine (18). ABW was determined using the following equation: ABW = IBW + 0.4 × (actual body weight – IBW) (19). A standard 3 + 3 dose-escalation design was used with the MTD defined as the highest dose studies for which the incidence of dose-limiting toxicity (DLT) is less than 33%. Therefore, if two or more patients in a cohort experienced DLT, the MTD was exceeded. Two to 7 patients were enrolled at each dose level. Total administered activity ranged from 888 to 14,874 kBq, and total antibody doses from 0.7 to 2.6 mg.

Given the low-level γ-emissions from ²²⁵Ac daughters ²²¹Fr and ²¹³Bi, radiation isolation for patients and precautions for staff were not required. Hematopoietic growth factor support was allowed according to American Society of Clinical Oncology guidelines (20). Complete blood counts and biochemical profiles were performed at least 3 times weekly. Toxicity was assessed according to the common criteria established by the NCI, version 3.0. DLTs were defined as: (i) grade 4 leukopenia lasting 35 days or more from the time of infusion (in patients with a pretreatment WBC > 1,000/μL); (ii) grade 3 or 4 thrombocytopenia that failed to recover to ≤ grade 2 at 12 weeks after treatment (in patients with a pretreatment platelet count > 50,000/μL); (iii), grade 3 elevations of alkaline phosphatase, bilirubin, aspartate aminotransferase, or alanine transaminase lasting ≥ 7 days; (iv) grade 3 elevations of serum creatinine that occurred within 6 weeks of treatment; (v) grade 2 elevations of serum creatinine lasting ≥ 7 days that occurred after 6 weeks from the time of treatment, not related to subsequent therapy; (vi) grade 3 proteinuria; (vii) other grade 3 nonhematologic toxicities lasting ≥ 10 days; or (viii) any grade 4 nonhematologic toxicity.

We evaluated pharmacokinetics of ²²⁵Ac-lintuzumab using the methods described below. To measure the antileukemic effects of ²²⁵Ac-lintuzumab, a bone marrow aspiration and biopsy was performed approximately 4 weeks after treatment. Responses were determined using the International Working Group criteria (21). Antileukemic activity was defined as elimination of peripheral blood blasts or ≥ 10% reduction of bone marrow blasts.

The first patient at each dose level underwent detailed pharmacokinetic studies to determine elimination kinetics in whole blood, plasma, and urine. Blood samples were collected at 5, 10, 30, 60, 90, 120, and 240 minutes after infusion of ²²⁵Ac-lintuzumab. Additional samples were obtained on days 2 and 3, day 6 or 7, and day 9 or 10 if detectable activity remained. Urine samples were obtained approximately 1 hour and 3 to 4 hours after treatment with additional samples collected on days 2 and 3, day 6 or 7, and day 9 or 10. Daughter retention and redistribution were studied by gamma counting at 2 energy windows for ²²¹Fr (185–250 KeV) and ²¹³Bi (360–480 KeV) (Compugamma model 1282; LKB Wallac). The data were decay-corrected to the time of injection, expressed as a percentage of the injected activity per liter, and fit to a 2-exponential decay curve.

The data sets generated in this study and supporting clinical trial documents are available from the corresponding author upon reasonable request.

Eighteen patients (median age, 64 years; range, 45–80) were treated with ²²⁵Ac-lintuzumab (18.5–148 kBq/kg). Eleven patients had relapsed AML and 7 had primary refractory disease (Table 1). According to the Cancer and Leukemia Group B risk classification system, 1 patient (6%) had favorable-risk cytogenetics, 13 patients (72%) had an intermediate-risk karyotype, and 4 (22%) had poor-risk cytogenetic abnormalities (22). Molecular genetic studies were not performed. Three patients each received 18.5, 37, and 74 kBq/kg. After the MTD was exceeded in 2 patients receiving 148 kBq/kg, an additional dose level of 111 kBq/kg was studied. Among the 7 patients enrolled at this dose level, 6 patients received 111 kBq/kg. One patient enrolled at this dose level received only 45% to 50% of the intended dose due to a faulty in-line filter and was excluded from assessment for MTD determination.

Myelosuppression was difficult to evaluate in these patients due to coexisting baseline suppression of hematopoiesis by active AML. Only 7 patients (1 patient at 18.5 kBq/kg, 2 at 74 kBq/kg, and 4 at 111 kBq/kg) entered the study with baseline neutrophil counts of ≥ 500/µL and were considered evaluable. Six of these patients developed grade 4 neutropenia after treatment with 74 or 111 kBq/kg (Table 2). Except for 1 patient, all were red blood cell transfusion dependent at the time of study entry. This subject developed grade 3 anemia after treatment with 37 kBq/kg. Only 6 patients were platelet transfusion independent at the time of study entry (1 patient at 18.5 kBq/kg, 2 at 74 kBq/kg, and 3 at 111 kBq/kg). Grade 4 thrombocytopenia after treatment was seen in all 6 patients. The duration of neutropenia and thrombocytopenia was difficult to assess as all but 3 patients had more than 5% blasts on bone marrow aspirations following treatment. Nevertheless, the median time to resolution of grade 4 leukopenia among 13 patients who could be evaluated across all dose levels regardless of response was 27 days (range, 0–71 days). The duration of myelosuppression was unrelated to the level of CD33 expression (r = 0.067; P = 0.846) or number of prior treatments (r = 0.537; P = 0.089) but correlated with administered activity (r = 0.784; P = 0.007).

As expected, neutropenic fever and infections were commonly seen (Table 2). Febrile neutropenia was seen in 6 patients (33%) at all dose levels except 74 kBq/kg. Six patients (33%) had bacteremia (1 patient after 37 kBq/kg; 4 patients after 111 kBq/kg, including 1 patient who received only a partial dose and another with a catheter-related infection; and 1 patient after 148 kBq/kg). Three patients (17%) developed grade 3 pneumonia after receiving 18.5, 74, and 111 kBq/kg. Other infections included sinusitis in 1 patient receiving 111 kBq/kg, cellulitis in 2 patients treated with 74 and 148 kBq/kg, and colitis in 1 patient after receiving 148 kBq/kg. Death due to sepsis was seen in 2 patients, 1 patient each at the 111 and 148 kBq/kg dose levels 24 and 28 days following treatment, respectively.

One patient with a history of fungal pneumonia and fungal hepatitis prior to treatment on study developed grade 3 elevation in alkaline phosphatase 32 days after receiving 37 kBq/kg of ²²⁵Ac-lintuzumab. This resolved in 6 days after discontinuing voriconazole and micafungin. Additionally, 1 patient treated with 74 kBq/kg experienced grade 2 elevation in alkaline phosphatase (occurring 19 days after treatment and lasting 7 days) and grade 2 increase in bilirubin (occurring 14 days after treatment and lasting 9 days). Grade 3 hyperbilirubinemia occurring 25 days after ²²⁵Ac-lintuzumab treatment was seen in 1 patient after receiving 111 kBq/kg, 3 days prior to death from sepsis and multiorgan failure. Grade 3 hyperbilirubinemia also occurred in 1 patient 24 days following treatment with 148 µCi/kg in the setting of sepsis and treatment with multiple antibiotics. The patient died 2 days later after antibiotic support was withdrawn. Two patients developed atrial fibrillation/flutter in the setting of infection 3 and 25 days following doses of 37 and 111 kBq/kg, respectively. Following treatment with 148 kBq/kg, 1 patient had a syncopal episode (grade 3) likely related to orthostasis, complicated by an asymptomatic subarachnoid hemorrhage (grade 3). Of note, no renal toxicity, manifested by either grade 3 or 4 elevations in serum creatinine or proteinuria, was seen with follow-up between 1 and 24 months (median, 3 months).

Based on these data, we determined that 148 kBq/kg exceeded the MTD of ²²⁵Ac-lintuzumab. DLT occurred in both patients treated at this dose level. One patient died from sepsis as noted above and a second had prolonged marrow aplasia lasting 71 days. We determined the MTD of a single infusion of ²²⁵Ac-lintuzumab to be 111 kBq/kg. Among the 6 patients receiving this dose, DLT was seen in only 1 patient. This patient developed pneumonia, sepsis, and multiorgan failure leading to death 28 days after receiving ²²⁵Ac-lintuzumab. Importantly, this morbidly obese patient received the highest administered activity of ²²⁵Ac in the study, even greater than the 2 patients treated with 148 kBq/kg. This event subsequently led to a protocol modification with dosing based on ABW for patients whose actual body weight exceeded their IBW.

Two-phase elimination kinetics for ²²⁵Ac-lintuzumab in whole blood, plasma, and urine were similar at all dose levels. The mean biological plasma half-lives t_1/2 − α and t_1/2 − β for ²²⁵Ac-lintuzumab were 1.9 and 38 hours, respectively (Fig. 2). The mean percentage clearance for the α half-life was 74.9% ± 5.77%. The mean percentage clearance for the β half-life was 25.1% ± 5.77%. These results are comparable with the results with ¹³¹I-lintuzumab (2). In the case of ²¹³Bi-lintuzumab, the kinetics are dominated by the short physical half-life of ²¹³Bi (46 minutes) that results in a short effective plasma half-life (12). The low administered activities of ²²⁵Ac precluded gamma camera imaging for biodistribution studies.

Patients underwent bone marrow aspirations and biopsies at baseline and approximately 4 weeks after administration of ²²⁵Ac-lintuzumab. Not all patients had evaluable specimens at both time points. Although remissions were not observed, antileukemic effects, defined as elimination of peripheral blood blasts or a ≥ 10% reduction in the percentage of marrow blasts, were seen across all dose levels. Peripheral blood blasts were eliminated in 10 of 16 evaluable patients (63%) but only at doses of at least 37 kBq/kg. Bone marrow blasts were reduced in 10 of 15 patients (67%) who were evaluated (Fig. 3). Eight of these patients (53%) had reductions of 50% or more. Three patients (20%), treated with 37, 111, and 148 kBq/kg, had 5% or fewer bone marrow blasts 1 month after administration of ²²⁵Ac-lintuzumab. The patient receiving 37 kBq/kg achieved a morphologic leukemia-free state. He subsequently underwent allogeneic HCT and died of transplant-related complications, including veno-occlusive disease (VOD). The patient treated with 111 kBq/kg also underwent allogeneic HCT with residual 5% blasts and subsequently died of infectious complications. The patient receiving 148 kBq/kg had marrow aplasia 1 month after therapy and recovered with leukemic blasts 2.4 months following treatment. The percentage reduction in marrow blasts did not correlate with baseline blast percentage (r = 0.392; P = 0.148), CD33 expression (r = −0.282; P = 0.308), or injected activity (r = −0.1; P = 0.738). Median overall survival for all patients was 3 months (range, 1–24 months).

This is the first study to show that targeted therapy with an in vivo α-particle generator is feasible in patients with AML. Although no remissions were seen in this study, ²²⁵Ac-lintuzumab was shown to have antileukemic activity across all dose levels studied, manifested by elimination in peripheral blood blasts or reductions in bone marrow blasts. These results, as well as previous results with ²¹³Bi-lintuzumab, clearly support further investigation of targeted α-particle therapy with ²²⁵Ac-lintuzumab in a variety of clinical settings either alone or in combination with standard chemotherapy and novel targeted agents.

Bone marrow blasts were reduced in 10 of 15 (67%) evaluable patients across all dose levels, 8 of whom had reductions of 50% or more. Although responses did not correlate with injected activities, 10 of 16 evaluable patients (63%) had elimination of peripheral blood blasts, but only at doses of at least 37 kBq/kg. Lack of correlation between injected activity and marrow blast reduction in this study may be explained by the small number of patients with wide variability in tumor burden, CD33 expression, and underlying radiation resistance due in part to prior therapy.

Therapy with high-energy, short-range α-particles can potentially provide for enhanced tumor cell killing and simultaneously reduce the nonspecific cytotoxic effects seen with β-emitters. Despite these advantages, it is not surprising that complete remissions after a single infusion of ²²⁵Ac-lintuzumab were not seen in this study. Assuming a leukemic tumor burden of 10^11–12 cells with an average CD33 density of 10,000/cell results in approximately 10^15–16 binding sites available to lintuzumab in a patient with overt AML. With about 5 DOTA per IgG and 1 in 5,000 DOTA molecules coordinating an ²²⁵Ac molecule, only 0.1% of lintuzumab molecules carried the radiolabel (17). At the administered activities and range of specific activities used in this study including 0.25 mg of cold lintuzumab, the amount of antibody infused was unable to saturate all of the available CD33 sites. Therefore, despite the potency of the isotope, at these low specific activities it was difficult to deliver enough ²²⁵Ac atoms (perhaps 3–5) to every leukemia necessary for efficient cell killing. This supports the study of ²²⁵Ac-lintuzumab at higher specific activities, in combination with other agents, or for elimination of measurable residual disease after cytoreduction.

We determined the MTD for a single infusion of ²²⁵Ac-lintuzumab to be 111 kBq/kg. Myelosuppression and infectious complications, and not extramedullary toxicity, were dose limiting. Prolonged myelosuppression (>35 days) and death due to sepsis within 30 days of treatment were seen in both patients who received a dose of 148 kBq/kg. Of concern prior to this study was the possibility of renal toxicity due to uptake of free ²¹³Bi by the kidneys, which was seen in preclinical testing in nonhuman primates (23). No renal toxicity manifested by either grade 3 or 4 elevations in serum creatinine or proteinuria was seen with follow-up between 1 and 24 months. Because CD33 in nonhuman primates is not recognized by lintuzumab, ²²⁵Ac-lintuzumab had a significantly longer half-life in these animals. Moreover, once internalized, ²²⁵Ac and its daughters are retained by target cells. These factors account for the increased toxicity in animal studies compared with what was observed in this study.

Although only 3 patients (17%) experienced short-lived grade ≥ 3 liver function abnormalities after receiving ²²⁵Ac-lintuzumab, 1 of 2 patients who subsequently underwent HCT developed VOD. Although VOD is a known complication of transplant conditioning regimens, leukemia cells and CD33-expressing reticuloendothelial cells in the liver could potentially result in an increased risk of hepatic toxicity following ²²⁵Ac-lintuzumab. Significant hepatic toxicity has been associated with another CD33-targeting agent, gemtuzumab ozogamicin (GO), an antibody–drug conjugate containing the antitumor antibiotic calicheamicin. In initial studies, treatment with GO resulted in grade 3 or 4 hyperbilirubinemia in 29% of patients and elevated transaminases in 18%, presumably due to dissociation of calicheamicin in the liver. VOD occurred in 5% of patients, predominantly in patients who underwent HCT (24). These safety concerns ultimately led to withdrawal of GO from the market. Additional studies, however, showed the safety and benefit of the drug using lower doses in a fractionated schedule, resulting in its reapproval (25). Given the limited number of patients in this study, particularly those undergoing HCT, it is difficult to assess the risk of VOD after ²²⁵Ac-lintuzumab. Nevertheless, given the potential for hepatic toxicity and VOD associated with this agent, careful monitoring for these events is essential in future studies of ²²⁵Ac-lintuzumab.

The second-generation construct of ²²⁵Ac-lintuzumab was developed to overcome 2 major obstacles for widespread use of ²¹³Bi-lintuzumab, including the need for an ²²⁵Ac/²¹³Bi generator onsite and its short half-life. By directly conjugating the isotope generator to a tumor-specific antibody, ²²⁵Ac serves as an in vivo generator of 4 α-particles. The 10-day half-life and 4 α-particle emission explain the increased potency seen with ²²⁵Ac constructs compared with ²¹³Bi analogues (14). Based on the antileukemic activity seen here and the previously demonstrated benefit to sequential therapy with cytarabine followed by ²¹³Bi-lintuzumab, a phase I/II study of low-dose cytarabine and ²²⁵Ac-lintuzumab in older patients with untreated AML was conducted.

Like most anticancer therapies, we expect that ²²⁵Ac-lintuzmab will be most effective when integrated into rationally designed treatment programs in combination with cytotoxic chemotherapy or other targeted agents. Given the potential for myelosuppression, however, careful phase I studies will be required to optimize dose and treatment schedules in distinct clinical settings. Additionally, lower doses of ²²⁵Ac-lintuzumab as a single agent for elimination of measurable residual disease may mitigate the myelosuppression seen in the setting of overt disease, but dose-finding studies will again be needed to determine the optimal dose that can be given to individuals with normal hematopoiesis at the time of treatment. Finally, ²²⁵Ac-lintuzumab could potentially be used to intensify conditioning before HCT with the possibility of decreased extramedullary toxicity for AML MDS, and in other CD33-expressing malignances. Higher doses may be possible in this setting but careful monitoring for hepatic toxicity, including VOD, will be critical.

M.R. McDevitt reports other support from Pharmactinium during the conduct of the study; in addition, M.R. McDevitt has a patent for Actinium and the use thereof licensed and with royalties paid from Pharmactinium. J.A. Carrasquillo reports grants from Memorial Sloan Kettering Cancer Center during the conduct of the study and has been consultant to Y-mAbs, a company which has been developing antibodies and radiolabeled antibodies for therapy of cancer; those antibodies are unrelated to the antibody used in this trial or to ²²⁵Ac. N. Pandit-Taskar reports grants and other support from Bayer; grants and personal fees from Imaginab, Illumina-Innervate, and Actinium Pharma; and grants from Fusion Pharma, Clarity Pharma, and Janssen outside the submitted work. M.G. Frattini reports other support from Cellectis, Inc. and BMS outside the submitted work. P.G. Maslak reports grants from API during the conduct of the study, as well as grants from BD Biosciences outside the submitted work. J.H. Park reports personal fees from Amgen, Novartis, Kite Pharma, Pfizer, BMS, Kura Oncology, Allogene, Curocel, Artiva, Intellia Therapeutics, Autolus, Servier, Affyimmune, Minerva, and PrecisionBio outside the submitted work. S.M. Larson reports grants and other support from Y-mAbs Therapeutics Inc and Elucida Oncology outside the submitted work. In addition, S.M. Larson has a patent for MSKCC issued and licensed to Y-mAbs Therapeutics and a patent for MSKCC issued and licensed to Samus Therapeutics LLC; please see https://www.mskcc.org/research/ski/labs/steven-larson. D.A. Scheinberg reports grants and other support from Actinium during the conduct of the study, as well as personal fees from Actinium, Eureka, Pfizer, Sellas, and Repertoire outside the submitted work; in addition, D.A. Scheinberg has a patent for Alpha-particle technology issued and licensed to Actinium. J.G. Jurcic reports grants and non-financial support from Actinium Pharmaceuticals, Inc., during the conduct of the study. J.G. Jurcic also reports grants from Astellas Pharma and Forma Therapeutics, Gilead/Forty Seven, Celularity, GlycoMimetics, Arog, PTC Therapeutics, and Genetech; grants and personal fees from Syros Pharmaceuticals, AbbVie, and BMS/Celgene; and personal fees from Novartis outside the submitted work. In addition, J.G. Jurcic is a Clinical Advisory Board member for Actinium Pharmaceuticals, Inc. (without financial compensation). No disclosures were reported by the other authors.

T.L. Rosenblat: Data curation, formal analysis, investigation, writing–original draft, writing–review and editing. M.R. McDevitt: Investigation, methodology, writing–review and editing. J.A. Carrasquillo: Investigation, writing–review and editing. N. Pandit-Taskar: Investigation, writing–review and editing. M.G. Frattini: Investigation, writing–review and editing. P.G. Maslak: Investigation, writing–review and editing. J.H. Park: Supervision, investigation, writing–review and editing. D. Douer: Supervision, investigation, writing–review and editing. D. Cicic: Resources, formal analysis, funding acquisition, project administration, writing–review and editing. S.M. Larson: Formal analysis, supervision, investigation, writing–review and editing. D.A. Scheinberg: Conceptualization, resources, formal analysis, supervision, investigation, writing–review and editing. J.G. Jurcic: Conceptualization, data curation, formal analysis, supervision, investigation, methodology, writing–review and editing.

We thank Renier Brentjens, Ellin Berman, Nicole Lamanna, and Mark Weiss for clinical care; Michael Curcio, Jing Qiao, and Eva Burnazi for laboratory assistance; Suzanne M. Chanel, Amabelle Lindo, and Louise Harris for research nursing support; Kevin Zikaras for data management; Nuclear Medicine Radiopharmacists led by Rashid Ghani for drug dispensing; Saed Mirzadeh (Oak Ridge National Laboratory, Oak Ridge, TN) for supplying ²²⁵Ac; and Mark Berger (Actinium Pharmaceuticals, Inc.) for his helpful comments. This work was supported by NIH grants PO1 CA33049 and RO1 55349, the Lymphoma Foundation, and Actinium Pharmaceuticals, Inc. (New York, NY).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.