Phase I Study of a CD45-Targeted Antibody–Radionuclide Conjugate for High-Risk Lymphoma

Despite numerous advances, external-beam radiotherapy remains the single most effective agent for the treatment of localized lymphoma. Unfortunately, the ability to safely deliver curative doses of radiation to multifocal disease sites has been limited by toxicity (1).

One potential avenue that has been explored is the use of targeted radiation by means of antibody–radionuclide conjugates (ARC), also known as radioimmunotherapy. CD20 is an effective target in B-cell non-Hodgkin lymphoma (B-NHL) with relatively little nonhematologic toxicity, and prospective clinical trials led to FDA approval of two different agents in this class: yttrium-90 (⁹⁰Y) ibritumomab tiuxetan and iodine-131 (¹³¹I) tositumomab. Both therapeutics emit beta particles that can cross many cell diameters in tissues.

A major barrier to further augmenting the efficacy of antibody-targeted radiation is dose-dependent hematologic toxicity that limits the amount of radioactivity that can be safely delivered. Autologous hematopoietic stem cell transplantation (ASCT) partially abrogates this, allowing for further dose escalation. High-dose CD20-targeted radiation can produce superior outcomes over historic controls receiving traditional high-dose regimens (2, 3).

Despite the clinical efficacy of targeting CD20 with radiation, many patients will relapse. We demonstrated in animal models that circulating rituximab blocks the binding of CD20 in preclinical models of B-NHL, thereby limiting the radiation-absorbed dose to disease sites and yielding inferior survival (4). Targeting the panhematopoietic antigen CD45 could circumvent this obstacle in B-NHL and could also provide antitumor efficacy in models of T-cell NHL (T-NHL), a disease with no currently available ARC (4, 5). We further postulated that targeting CD45 with isotopes that emit long multi-cell-diameter pathlength beta particles would be effective in treating classical Hodgkin lymphoma based on the bystander effect of targeting CD45-bearing cells adjacent to Reed–Sternberg cells, which classically lack CD45.

These rationales served as the foundation for a single-agent phase I study targeting CD45 with ¹³¹I followed by ASCT for patients with high-risk B-NHL, T-NHL, and Hodgkin lymphoma.

Eligible patients were at least 18 years old with a histologically confirmed diagnosis of B-NHL, T-NHL, or Hodgkin lymphoma. Confirmation of CD45 expression in tumor samples (either by IHC or flow cytometry) was required. In Hodgkin lymphoma, observation of CD45 expression on cells adjacent to Reed–Sternberg cells was necessary. Patients must have received at least one prior standard systemic therapy with documented recurrent or refractory disease, with the following exception: high-risk histologies (e.g., mantle cell lymphoma and T-NHL) were eligible while in first complete (CR) or partial remission (PR). Additional eligibility criteria are discussed in the Supplementary Materials and Methods. This study was conducted in accordance with the Declaration of Helsinki, was approved by the Fred Hutchinson Cancer Research Center (FHCRC) Institutional Review Board, and all patients gave written informed consent. The study was registered on clinicaltrials.gov (#NCT00860171; FHCRC protocol 2238).

The anti-CD45 ARC was composed of the antibody BC8, a murine IgG1 that binds to all CD45 isoforms. BC8 was produced and purified in the Biologics Production Facility at the FHCRC and linked to ¹³¹I, as described previously (6). A general treatment schema is depicted in Fig. 1. First, to determine biodistribution and dosimetry, patients received a trace-labeled [∼10 mCi (0.37 GBq)] infusion of ¹³¹I-BC8, followed by serial planar gamma camera images obtained at 0 hours (i.e., after completion of the infusion) and on a minimum of 3 of the subsequent 7 days. Organ uptake, retention, and clearance were determined by region-of-interest delineation and quantitative analysis. Time–activity curves were identified and integrated, from which radiation-absorbed doses (Gy per unit administered of radionuclide activity) were calculated for tissues and organs according to methods recommended by the Medical Internal Radiation Dose (MIRD) Committee of The Society of Nuclear Medicine and Medical Imaging, as described previously (7, 8). Tumor dosimetry was performed in select patients with disease sites amenable to biopsy following trace-labeled infusions. These results were used to optimize biodistribution via a protein-dose escalation schema. Details of these methods, as well as supportive-care measures that were taken for radiolabeled antibody administration are described in the Supplementary Materials and Methods.

The desired ¹³¹I activity for each patient's therapeutic infusion was based on the corresponding data obtained from the lead-in dosimetry study, to administer an absorbed radiation dose to the normal organ that received the highest dose, according to the prespecified radiation-dose escalation schema. Dose escalation was independently performed in 2 Gy increments in two groups: arm A included those with prior ASCT or RT >20 Gy and started at a dose of 10 Gy to the target organ, while arm B was all others and started at a dose of 20 Gy to the target organ. Initial patients also had radiation dose to red marrow capped at 50 Gy.

Adverse events were categorized and graded according to the NCI Common Terminology Criteria for Adverse Events (version 3.0; CTCAE). Response assessments occurred prior to therapy and then approximately 1 month after transplant according to International Working Group criteria (9). Long-term follow-up for efficacy and safety on this phase I trial was based on clinical standard of care.

The primary objective of this study was to estimate the MTD of ¹³¹I-BC8 that could be delivered followed by ASCT in patients with high-risk lymphoma. Dose escalation/deescalation was conducted by a “two-stage” approach (10). Single patients were entered at each dose level in 2-Gy increments until a dose-limiting toxicity (DLT; defined as a therapy-related grade III or IV Bearman toxicity within 30 days following transplant) was observed (11). Once a DLT was observed, a second dose-escalation scheme was to be initiated in cohorts of 4 patients starting at one dose level lower (outlined in the Supplementary Materials and Methods). In addition, antibody dose could be adjusted in patients undergoing serial tumor dosimetry to yield a more favorable biodistribution (also detailed in the Supplementary Materials and Methods). Detailed pharmacokinetic studies of ¹³¹I-BC8 have been reported previously (6).

We enrolled a total of 16 patients, and their characteristics are summarized in Table 1. Histologies included 7 with B-NHL, 6 with Hodgkin lymphoma, and 3 with T-NHL. Among the patients not in CR at enrollment, the median longest diameter of the largest site of measurable disease was 4.0 cm (range: 2.3–15.3 cm). No patients had evidence of bone marrow involvement by morphologic review; however, the 3 patients with T-NHL each had detectable marrow disease by flow cytometry.

The critical nontarget organ that most commonly received the highest absorbed dose was liver at a median of 3.1 cGy/mCi. Patient 2 [mantle cell lymphoma (MCL)] and 3 (Hodgkin lymphoma) underwent serial trace-labeled infusions of ¹³¹I-BC8 at antibody–protein doses of 0.5 mg/kg and 0.75 mg/kg, followed by tumor biopsies for dosimetry. Both patients demonstrated improved tumor uptake of the ARC at 0.75 mg/kg, with an increase in tumor-to-liver dose ratios from <0.66 to 0.66, and from 0.80 to 1.61, respectively. All subsequent patients were treated at this antibody dose. Personalized dosimetry guided the selection of ¹³¹I therapy infusion activities, which were designed to optimize the maximum absorbed dose to liver or marrow (Table 2). The median activity administered of ¹³¹I was 678 mCi (25.1 GBq). Additional dosimetry results are provided in the Supplementary Materials and Methods (Supplementary Tables S1 and S2).

All grade 3 or 4 CTCAE nonhematologic events recorded are shown in Table 3, and no DLTs were observed in any patients. None of these events developed after discharge from our center (i.e., day 30 after ASCT). Infusion-related reactions to BC8 have been described previously (7), and that experience was again observed. These reactions, including two cases of grade 3 serum sickness, were self-limited and resolved with pausing the infusion and/or instituting appropriate supportive care. No patients had to be removed from the trial due to unacceptable toxicity, and adverse events typical for high-dose regimens such as mucositis, diarrhea, and febrile neutropenia were infrequent. Expected grade 3–4 hematologic toxicity was observed in all patients: median time after ASCT to recovery of neutrophils >500/μL was 10 (range: 8–20) days, while for platelets >20,000/μL it was 12 (range: 8–26) days. Seven events were classified as serious adverse events. Secondary malignancies were seen in 6 patients, which included 3 non-melanoma skin cancers, 1 adenocarcinoma of the colon, 1 myelodysplastic syndrome, and 1 follicular lymphoma.

Seven (50%) of 14 patients with measurable disease responded and 6 (42%) achieved a CR at 1-month post-ASCT. The CRs included all 3 patients (100%) with T-NHL, 2 patients (33%) with Hodgkin lymphoma, and 1 patient (20%) with B-NHL (follicular). Among those with chemorefractory disease (n = 9), the overall response rate was 56% (4 CR's and 1 PR). The 2 patients that enrolled while in CR had MCL, but these remissions were after two and three prior chemotherapy regimens, respectively: the first was treated at the lowest dose level (patient 1, 10 Gy) and relapsed 6 months later; the second patient received a higher radiation dose (patient 13, 26 Gy) and was still in CR at last available follow-up (35 months).

The 1-year and 3-year Kaplan–Meier estimates of progression-free survival (PFS) were 50% and 31%, respectively, while the 1-year and 3-year overall survival (OS) estimates were 81% and 55%, respectively (Fig. 2). The median PFS and OS were 11 months [95% confidence interval (CI), 4 months to infinity] and not reached (95% CI, 29 months to infinity). Among the 9 surviving patients, the median duration of follow-up has been 35 months.

Overall, 5 (31%) remain in prolonged remission after ¹³¹I-BC8 at a median of 36 (range: 25–41) months of follow-up. All had received multiple lines of chemotherapy (range: 2–6) and included 2 with Hodgkin lymphoma and 1 each with MCL; follicular; and high-grade B-cell lymphoma (HGBCL) with MYC, BCL2, and BCL6 rearrangements (i.e., triple-hit lymphoma). The 2 patients with Hodgkin lymphoma, who had otherwise been unable to achieve a CR, received a protocol-permitted consolidative ASCT with busulfan, etoposide, cytarabine, and melphalan (BEAM) conditioning following recovery from ¹³¹I-BC8 therapy because CR was induced by the ARC. Both patients with follicular and HGBCL had measurable disease at enrollment and demonstrated remissions of 41 and 35 months, respectively, which are ongoing at last follow-up. Interestingly, the 3 patients with B-NHL who experienced the longest remissions received among the highest radioactivity levels of ARC therapy [ranges: 713–1,014 mCi (26.4–37.5 GBq) and 26–32 Gy to the liver].

The last decade has witnessed improved treatment options for patients with lymphoma. However, even with the most promising recent strategies such as chimeric antigen receptor–modified T cells, many patients do not achieve prolonged remissions and will succumb to their disease (12). Furthermore, despite several new agents for the treatment of T-NHL and Hodgkin lymphoma, very few patients with relapsed or refractory disease achieve prolonged disease control (13–16). These regimens are also notable for protracted treatment courses and cumulative toxicity. As our understanding of the biology of lymphoma advances, many therapies are becoming progressively more refined and, thus, targeting an increasingly narrower spectrum of diseases, further compounding the challenge of drug development. In this study, we specifically designed and evaluated an ARC strategy targeting CD45 that could be broadly applied to B-NHL, T-NHL, and Hodgkin lymphoma, taking advantage of the broad expression of this antigen and the known exquisite radiosensitivity of all lymphoid malignancies.

Nonhematologic toxicity from ARC therapy is generally modest and infrequent. Most of the grade 3 or 4 events observed were due to hypersensitivity or anaphylactoid reactions to the murine anti-CD45 antibody BC8 comparable with other mAbs in routine clinical use. Fortunately, these reactions were self-limited, resolved with slowing or cessation of the antibody infusion along with supportive care, and did not appear to impact the antilymphoma activity of the ARC. Despite increasing the amount of ¹³¹I infused to over 1,000 mCi (37 GBq) with an associated radiation dose of over 30 Gy to the liver, no DLT events were seen, and common adverse events expected with high-dose therapy such as mucositis and febrile neutropenia were rare, emphasizing the safety of this strategy. This level of radioactivity is over 10-fold greater than the median administered in the pivotal trial of ¹³¹I-tositumomab (17). Furthermore, there were also no obvious signals of prolonged effects on hematopoietic recovery.

In addition to a relatively favorable toxicity profile, we documented encouraging signs of activity in patients with high-risk lymphoma, even though many such patients were treated at low dose levels. Objective responses were still seen in this group, where the majority had three or more lines of prior therapy and chemorefractory disease. It is noteworthy that all 3 patients with T-NHL had measurable disease at enrollment and achieved CR about 1 month after anti-CD45 ARC. Although we saw no clear evidence of a dose–response relationship based on 1-month response rate, the 3 patients that experienced the longest remissions without further consolidative therapy received the highest absorbed doses of radiation. At these high doses, the anticipated hematologic toxicity is such that ASCT likely abrogated the potential sequelae that prolonged cytopenias may have caused. These findings also support the hypothesis that high-dose anti-CD45 ARC therapy has antitumor activity even in chemorefractory, high-risk lymphoma, corroborating observations with local control by external beam radiation (1).

In contrast to prior anti-CD20 ARC strategies, targeting CD45 has several potential advantages. Anti-CD45 antibodies can circumvent blocking or CD20 downregulation caused by prior rituximab (4), which may have adversely impacted the results of prior phase III trials of CD20-targeted ARC-based therapy (18, 19). Furthermore, anti-CD45 ARCs can bind to both malignant cells and inflammatory cells in the tumor microenvironment. This provides an opportunity to amplify the delivery of therapeutic radiation to hematolymphoid tumor sites. The pathlengths of the resulting beta particles are typically approximately 1 mm in tissue, cross many cell diameters, and create an optimal radius of potential therapeutic field, a concept referred to as the “crossfire effect.” This is particularly advantageous in Hodgkin lymphoma, where inflammatory infiltrate comprises the vast majority of the tumor volume. The fact that we observed efficacy of this anti-CD45 ARC in patients with Hodgkin lymphoma provides direct evidence of this phenomenon, because the malignant Reed–Sternberg cells classically do not express CD45 and are dispersed among CD45-expressing lymphocytes. This is analogous to observed efficacy of CD25-directed ARC, which is expected to primarily target infiltrating T cells that surround Reed–Sternberg cells (20).

In conclusion, we have shown that anti-CD45 ARC with ¹³¹I at doses from 10 to 32 Gy to liver can be safely administered to patients with a broad range of high-risk lymphomas, yielding a manageable toxicity profile and signs of short- and long-term efficacy. These data, and the additional experience utilizing this approach in acute leukemia and multiple myeloma, also support the combination of anti-CD45 ARC with traditional transplant conditioning regimens for lymphoma, such as our study with ⁹⁰Y-BC8 combined with BEAM followed by ASCT (NCT01921387). Other combinations could be considered where synergy with radiotherapy is expected (e.g., immunotherapy). We anticipate that with these data and results from ongoing trials, anti-CD45 ARC could be broadly applied to treat a wide range of hematologic neoplasms, a therapeutic space that still requires improved options.

R.D. Cassaday is an employee/paid consultant for and has ownership interests (including patents) in Seattle Genetics; is an unpaid consultant/advisory board member for Pfizer, Amgen, Jazz Pharmaceuticals, and Adaptive Biotechnologies; and reports receiving commercial research grants from Pfizer, Amgen, Vanda Pharmaceuticals, Merck, and Kite/Gilead. J.M. Pagel is an unpaid consultant/advisory board member for Actinium Pharmaceuticals. R.S. Miyaoka is an employee/paid consultant for Precision Sensing, LLC, and is an unpaid consultant/advisory board member for MIM Software. B.M. Sandmaier is an unpaid consultant/advisory board member for Actinium Pharmaceuticals. D.J. Green is an employee/paid consultant for Cellectar Biosciences, GlaxoSmithKline, and Celgene. A.K. Gopal reports receiving commercial research grants from Merck, Takeda, Teva, Janssen, Pfizer, Seattle Genetics, Gilead, and Spectrum, and is an unpaid consultant/advisory board member for Seattle Genetics, Gilead, Amgen, Janssen, Adaptive, IMAB, and Sanofi. No potential conflicts of interest were disclosed by the other authors.

Conception and design: O.W. Press, J.M. Pagel, B.M. Sandmaier, D.J. Green, A.K. Gopal

Development of methodology: O.W. Press, J.M. Pagel, J.G. Rajendran, A.K. Gopal

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): R.D. Cassaday, O.W. Press, J.M. Pagel, J.G. Rajendran, L.A. Holmberg, B.M. Sandmaier, D.J. Green, A.K. Gopal

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): R.D. Cassaday, O.W. Press, J.M. Pagel, T.A. Gooley, D.R. Fisher, R.S. Miyaoka, B.M. Sandmaier, A.K. Gopal

Writing, review, and/or revision of the manuscript: R.D. Cassaday, O.W. Press, J.M. Pagel, J.G. Rajendran, D.R. Fisher, L.A. Holmberg, R.S. Miyaoka, B.M. Sandmaier, D.J. Green

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): R.D. Cassaday, J.M. Pagel

Study supervision: A.K. Gopal

This work is dedicated to Oliver W. Press who devoted much of his professional life to the development of high-dose ARC-based regimens. This study was funded by NIH K24 CA184039, P01 CA044991, P30 CA015704, and T32 CA009515; the Lymphoma Research Foundation; and donations from Frank and Betty Vandermeer and Sonya and Tom Campion.

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.