Yttrium-90 anti-CD25 BEAM conditioning for autologous hematopoietic cell transplantation in Peripheral T-cell lymphoma

Most systemic peripheral T-cell lymphomas (PTCLs) have poor prognosis with current therapies, an exception being anaplastic lymphoma kinase (ALK)–positive anaplastic large cell lymphomas (ALCLs) with low International Prognostic Index scores. Allogeneic stem cell transplantation is offered early in the disease course for very aggressive histologies (human T-lymphotropic virus type 1–associated acute T-cell lymphoma/leukemia, gamma delta T-cell lymphoma, and hepatosplenic T-cell lymphoma) preferably in first remission, or in other subtypes in cases of relapse or refractory disease. In the United States, the National Comprehensive Cancer Network guidelines recommend consideration of high-dose therapy and autologous hematopoietic cell transplant (AHCT) for patients with nodal T-cell lymphomas who achieve remission after initial induction therapy to improve outcomes.^1,2 These recommendations are based on 5 prospective studies and many retrospective and meta-analyses that support that AHCT as consolidation is probably the most effective way to improve progression-free survival (PFS) and overall survival (OS) in PTCL.^1-7 For relapsed disease, AHCT could be offered to patients with chemotherapy-sensitive relapse who did not receive AHCT in first remission. However, outcomes may be worse if AHCT is performed beyond complete remission after initial therapy (CR1).⁸ Most of these cases are referred for an allogeneic stem cell transplant, which has been shown to result in long term survival. In an analysis of Center for International Blood and Marrow Transplant Research (CIBMTR) data, the outcomes of allotransplantation in RR PTCL were reported as 3-year OS of up to 60%.⁹ 

The rarity and heterogenicity of T-cell lymphomas make meaningful clinical trials challenging. Most treatment paradigms are borrowed from studies of aggressive B-cell lymphomas. The most-used conditioning regimen for AHCT for lymphomas including PTCL is BEAM (bis-chloroethylnitrosourea [carmustine], etoposide, cytarabine, and melphalan],¹⁰ which lacks any agent directed specifically at malignant T cells and may be the cause of suboptimal outcomes. Recently targeted treatments in PTCL have been shown to improve outcomes¹¹ and their incorporation into conditioning regimens would be expected to improve outcomes for patients with PTCL who receive AHCT.

CD25 (Tac), forms 1 component of the high-affinity heterotrimeric interleukin-2 receptor, and is expressed in 40% to 50% of PTCL specimens.¹² CD25 is also expressed in activated T cells, regulatory T cells, and tumor infiltrating lymphocytes, which populate the tumor microenvironment and may exert immunosuppressive effects.¹³ Interleukin-2–mediated signaling via CD25 leads to proliferation and survival of malignant cells through the JAK/STAT5, PI3K/Akt/mTOR and mitogen-activated protein kinase pathways.¹⁴ Thus, disruption of CD25 is an attractive therapeutic strategy that directly targets tumor cells while simultaneously modifying the tumor microenvironment by disrupting regulatory T cells.^15-18 

The role of CD25 expression in the pathogenesis of PTCL remains unclear, but antibody-based therapy designed to target and disrupt CD25-mediated signaling (anti-Tac) has been attempted in the treatment of T-cell lymphomas.^15-18 The humanized anti-CD25 antibody daclizumab produced partial responses in patients with adult T-cell leukemia-lymphoma (ATLL) with the smoldering subtype.¹⁸ Denileukin diftitox (Ontak) is a CD25-directed immunotoxin with a diphtheria toxin payload, that has produced a response rate of 30 % in patients with relapsed cutaneous T-cell lymphomas (CTCL).¹⁶ This molecule was recently reengineered to have 1.5- to 2-times more bioactivity than denileukin diftitox and renamed E7777. It has shown activity in patients with relapsed/refractory CTCL whose tumors expressed CD25 (overall response rate of 36.2 %; duration of response, 6-12 months) in a multicenter open-label study.¹⁹ ADCT-301, which fuses pyrrolobenzodiazepine with the human immunoglobulin G1 anti-CD25 antibody, has been investigated in both Hodgkin lymphoma (HL) and PTCL. An initial phase 1 study included 22 patients with PTCL; responses were seen in 42% of cases.²⁰ 

Radioimmunotherapy (RIT), in which antibodies are labeled with radioactive conjugates, enables targeted delivery of radiotherapy to tumors. Because of their radiosensitivity, lymphomas have been considered ideal targets for RIT, which has been shown to be safe and effective as a treatment for various lymphomas and has been incorporated into AHCT conditioning regimens.^21-24 Yttrium-90 (⁹⁰Y)-labeled basiliximab (aTac) is a humanized monoclonal antibody designed to deliver targeted radiation to CD25-expressing malignant cells.²⁵ In patients with relapsed ATLL, in which there is high expression of CD25 on malignant ATLL cells compared with normal resting T cells,^26,90Y-labeled aTac administration resulted in responses in 9 of 16 (56%) of patients. The main toxicities were hematopoietic.²⁶ More recent studies have sought to combine ⁹⁰Y-labeled aTac with high-dose chemotherapy as part of conditioning regimens before AHCT with promising results and no major safety concerns.^22,27 Additionally ¹³¹iodide-labeled basiliximab (CHT-25)²⁵ and ⁹⁰Y-daclizumab have been investigated in HL.²⁸ 

Here, we report the results of a phase 1 trial evaluating ⁹⁰Y-aTac with BEAM conditioning for AHCT in patients specifically with PTCL. We tested the hypothesis that using CD25-directed RIT as part of the AHCT regimen would be safe and incur much less radiation dose to nontumor tissues and organs than does total body irradiation (TBI). This will have the advantage of adding a targeted radiation to the “standard” conditioning regimen for PTCL and potentially make it more effective without incurring the toxic effects of TBI, which is associated with increased regimen-related toxicity as well as long-term risk of secondary malignancies.^29,30 

This was a single-center, phase 1, dose-escalation study of ⁹⁰Y-aTac-BEAM for patients with PTCL undergoing AHCT. Eligible patients were aged ≥18 years with a diagnosis of mature T-cell non-HL as per World Health Organization 2008 classification,³¹ for whom AHCT was being considered as a therapeutic option. This included (1) patients who had achieved a CR1 and needed consolidation, and (2) relapsed patients without prior AHCT who had achieved either a partial response or CR after salvage therapy and for whom an allogeneic transplant was not an option. All histologies were allowed except CTCL and ALK-positive ALCL.

To be eligible for AHCT, patients had to meet the following institutionally mandated criteria: Karnofsky performance status of ≥70%; adequate organ function, including cardiac ejection fraction of ≥50% by echocardiogram or multigated acquisition scan; FEV1 of >65% of predicted measured, or DLCO of ≥50% of predicted measured; bilirubin of ≤1.5 the upper limit of normal, and aspartate aminotransferase and alanine aminotransferase of ≤2 × the upper limit of normal; and serum creatinine of ≤1.5 mg/dL, and measured creatinine clearance of ≥60 mL/min. Recovery from nonhematologic toxicities of salvage therapy to grade ≤2 according to the Common Terminology Criteria for Adverse Events (CTCAE) version 4.03 was required, as was collection by apheresis of at least 3.0 × 10⁶ CD34 cells per kg of autologous hematopoietic progenitor cells from a maximum of 10 collections. Supplementation of apheresis by bone marrow harvest was not allowed.

Patients with progressive disease, HIV infection, active hepatitis B or C infection, or evidence of marrow disease by flow and morphology after upfront or salvage cytoreductive therapy and before stem cell mobilization were excluded. Additional exclusions included the following: myelodysplastic syndrome (MDS) or any known MDS–associated cytogenetic abnormality in the bone marrow; prior AHCT or allogeneic HCT; and significant external beam dose-limiting radiation to a critical organ, including >20 Gy to any portion of the lung, >5 Gy to any portion of the kidney, or any prior radiation to the heart.

All patients provided signed informed consent for participation in the clinical trial. The study was approved by the institutional review board and conducted in accordance with the principles of the Declaration of Helsinki, and was registered at www.ClinicalTrials.gov as #NCT02342782.

The anti-CD25 monoclonal antibody basiliximab (5 mg per dose) was conjugated with 1,4,7,10-tetraazacyclododecane tetraacetic acid (DOTA), as previously described.²² Imaging doses prepared with indium-111 (¹¹¹In-basiliximab/DOTA), and therapeutic doses prepared with yttrium-90 (⁹⁰Y-basiliximab/DOTA) were purified and underwent quality control testing, as previously described.²² Administration was performed within 6 hours of radiolabeling.

Unlabeled (“cold”) basiliximab (5 mg) was administered intravenously 21 days before AHCT (ie, on “day –21”; Figure 1) to block circulating soluble CD25 antigen and thereby improve the pharmacokinetics and tumor targeting of radiolabeled basiliximab.^22,26 Premedication (acetaminophen and diphenhydramine) and infusion of unlabeled basiliximab were as previously described.²² Soluble CD25 levels were determined from a blood draw before intravenous administration of cold basiliximab on day –21.

Blood clearance of ¹¹¹In-basiliximab/DOTA and ⁹⁰Y-basiliximab/DOTA was measured in peripheral venous samples drawn before cold dose, before hot dose, and 2 hours, 4 to 6 hours, 1 day, 2 days, 3 to 4 days, and 5 and 6 days after injection. Terminal elimination, serum half-life (t_1/2), and area under the curve were determined as previously described.²² 

To assess biodistribution, ¹¹¹In-basiliximab/DOTA (imaging dose 5 mCi) was administered on day –21. Whole-body planar scans were acquired at 2, 24, 48, 120, and 144 to 168 hours after injection. Single-photon emission computerized tomography (SPECT) images were also obtained at 48 and 72 to 120 hours after injection (Figure 1). To verify scanner sensitivity, a calibrated sample of ¹¹¹In-basiliximab/DOTA was placed adjacent to patients’ legs. Radioactivity quantification was performed using the conjugate-view approach.³² Attenuation correction of the calculated region of interest activity was performed based on organ measurements from a computed tomography (CT) scan of the patient taken close in time to the ¹¹¹In-basiliximab/DOTA scans. Red marrow dose was calculated from blood activity, and urine clearance was used to verify residual activity in the body (samples were collected daily for 6 days, starting on the day of infusion). Absorbed dose estimates were calculated using the standard adult human model in OLINDA/EXM.³³ Biodistribution was determined from both visual and quantitative evaluation of the planar and SPECT images. Visual assessment was performed by the study radiologist (D.Y.). Altered biodistribution was defined as ≥1 of the following: (1) diffuse uptake in normal lung more intense than cardiac blood pool on the 2-hour time point images, or more intense than liver on the 24-hour or 48-hour time point images; (2) kidneys more intense than liver on the posterior view at 24 hour or 48 hours; (3) intense areas of uptake within normal bowel comparable with liver in the 24-hour or 48-hour images.

The therapeutic ⁹⁰Y-basiliximab/DOTA dose was administered on day –14 using the same protocol as on day –21. Three ⁹⁰Y-basiliximab/DOTA dose levels were tested: 0.4, 0.5, and 0.6 mCi/kg. The maximum ⁹⁰Y dose was set at 40 mCi for the 0.4 mCi/kg dose level, 50 mCi for the 0.5 mCi/kg dose level, and 60 mCi for the 0.6 mCi/kg dose level.

The BEAM conditioning regimen was administered according to institutional practice (Figure 1): bis-chloroethylnitrosourea on days −7 and −6, 150 mg/m² IV based on adjusted ideal body weight; cytarabine and etoposide twice daily on days –5, –4, –3, and –2, each dose 100 mg/m² based on adjusted ideal body weight; and melphalan 140 mg/m² on day –1. Stem cells were infused on day 0, followed by 5 μg/kg per day of granulocyte colony-stimulating factor administered starting on day +5 and continued until absolute neutrophil count was >500 for 3 consecutive days. Supportive care (premedications, antiemetics, and infection prophylaxis) was given according to institutional practice.

Safety was monitored continuously according to the CTCAE version 4.03 and the Bearman Toxicity scale.²² Dose-limiting toxicities (DLTs) were assessed from day –21 through day +30. DLTs were defined as any grade ≥3 toxicity according to the Bearman scale, grade 4 neutropenia lasting >42 days, any grade ≥4 infusion-related reaction, or any death unless clearly unrelated to study treatment. Pulmonary function tests and cardiac echocardiogram or multigated acquisition scan were performed on day +100 and 1 year after AHCT. Anti-basiliximab antibodies were assayed at day +100, day +180, and 1 year after AHCT, and post-RIT cytogenetic assessment was performed on day +100, day +180, 1 year, and 2 years after AHCT.

Response to treatment was evaluated according to the 2007 Revised Response Criteria for Malignant Lymphoma.³⁴ Bone marrow biopsies for disease assessment were performed at day +100, day +180, 1 year, and 2 years after AHCT. Extent of disease was monitored by fluorine-18 fluorodeoxyglucose (FDG) positron emission tomography/CT (FDG-PET/CT) scans performed before the first infusion of cold basiliximab (“baseline scan”) and at day +30, day +100, day +180, 1 year, 18 months, and 2 years after AHCT. After documentation of CR, either FDG-PET/CT or CT was allowed.

The primary end point was the safety and feasibility of ⁹⁰Y-basiliximab/DOTA when given to patients with PTCL in combination with standard-dose BEAM as part of conditioning for AHCT. Secondary end points included 1-year and 2-year PFS, OS, nonrelapse mortality, and cumulative incidence of relapse-progression at day +100 (nonrelapse mortality only). Other secondary objectives included time to neutrophil and platelet engraftment as well as estimation of radiation doses to the whole body and organs through serial imaging studies.

Expression of CD25 primary tumor cells was performed by immunohistochemistry, as previously described,²² on pre-AHCT tumor samples.

This phase 1 trial followed a modified, more conservative version of the rolling 6 design of Skolnik et al.³⁵ For a given test dose level, the schema was as follows: at most, 3 patients were under observation for DLT at any time. Patients not evaluable for DLT were replaced. When a patient evaluable for toxicity passed without a DLT, an additional patient could be accrued, up to a maximum of 6 patients. Once 3 patients were evaluated with none experiencing a DLT, up to 3 additional patients could be treated at that dose level, or the dose could be escalated. There was no intrapatient dose escalation.

Although this design does not require that 6 patients be treated per dose level, no more than 6 evaluable patients were accrued to any dose level during this dose-finding study.

All patients provided signed informed consent for participation in the clinical trial. The study was approved by the City of Hope Institutional Review Board and conducted in accordance with the principles of the Declaration of Helsinki.

A total of 23 patients were consented for the protocol. Of the 23 patients, 22 were eligible for treatment between July 2015 and June 2020, but 2 did not receive treatment (Figure 1). Twenty patients received cold basiliximab, an imaging dose of ¹¹¹In-basiliximab-DOTA, a therapeutic dose of ⁹⁰Y-basiliximab-DOTA, and BEAM AHCT. Patient baseline characteristics are shown in Table 1. The median age was 51 years (range, 18-76). The median number of prior therapies was 1 (range, 1-4). Induction therapy consisted of cyclophosphamide (Cytoxan), doxorubicin (Adriamycin), vincristine (Oncovin), and prednisone (CHOP; 6 patients), Cytoxan, Adriamycin, vincristine, etoposide, and prednisone (CHOEP; 8 patients); dose-escalated infusional etoposide, Adriamycin, vincristine, Cytoxan, and prednisone (DA-EPOCH; 4 patients), brentuximab, Cytoxan, Adriamycin, and prednisone (BV + CHP; 1 patient), or other (1 patient). Eighteen (90%) patients were in CR1, and 2 (10%) were in CR2. All patients had chemotherapy-sensitive disease. Salvage regimens consisted of brentuximab vedotin (3 patients) and gemcitabine-based therapy (1 patient). At the time of stem cell transplant, all patients were eligible to receive AHCT as per National Comprehensive Cancer Network guidelines.² 

Of 20 patients who received a therapeutic dose of ⁹⁰Y-basiliximab-DOTA, 4 were treated with 0.4 mCi/kg, 4 with 0.5 mCi/kg, and 12 with 0.6 mCi/kg. At the end of the observation period, 14 patients were alive, and 12 remained progression free (Table 2).

A radioactivity dose of 0.6 mCi/kg was determined to be acceptably safe and tolerable for phase 2 studies of ⁹⁰Y-aTac-BEAM AHCT in PTCL. No DLTs were observed at any dose level, and the incidence of adverse event (AEs) was roughly the same among the 3 dose levels (Table 3).

No patients experienced graft failure. The median time to neutrophil engraftment (absolute neutrophil count of >500) was 10.5 days (range, 10-21), and the median time to platelet engraftment (>20 × 10³) was 13 days (range, 11-92; supplemental Table 1). One patient remained transfusion dependent for platelets for 15 months, and eventually her platelets recovered. At the last follow-up, she was healthy and remained in clinical remission.

Treatment-related AEs are summarized in Table 3 and supplemental Table 2 (all treatment-related AEs are reported.) According to the Bearman criteria (Table 3), the most common side effect was mucositis (experienced by 16/20 [80%] patients), mostly grade 2 that was managed as per standard guidelines and was seen at all dose levels; gastrointestinal tract (GIT) toxicity was seen in 5 of 20 (25%) patients. No grade ≥3 Bearman AEs were noted, and all AEs were managed according to standard institutional supportive measures. According to the CTCAE version 4.03 (supplemental Table 2), the most common AEs observed (all grades) were: anemia, decreased lymphocyte count, nausea, decreased platelet count, decreased neutrophil count, and decreased white blood cell count (all 100%). The most common CTCAE version 4.03 grade ≥3 AEs (supplemental Table 3) were: decreased lymphocyte count, decreased neutrophil count, decreased platelet count, decreased white blood cell count (all 100%), febrile neutropenia (85%), and anemia (55%). No cardiac toxicities or veno-occlusive disease were noted. At last follow-up, no patients had developed MDS or acute leukemia after aTac-BEAM AHCT. Aside from these events, the only other severe AEs in patients who received a therapeutic dose of ⁹⁰Y-basiliximab-DOTA were a grade 3 lung infection that occurred 7 months after AHCT in 1 patient and a re-admission for grade 1 fever and grade 2 nausea on day + 28 after AHCT. There were no treatment-related deaths.

Survival data are shown in Figure 2. As of the end of the observation period, 6 patients (30%) had died at a median time of 17 months after AHCT (range, 9- 21). Median follow-up was 24 months (range, 9-26). Median follow-up time in surviving patients was 24 months (range, 13-26). In all patients who received a therapeutic dose of ⁹⁰Y-basiliximab-DOTA, the 2-year PFS and OS were 59% (95% confidence interval [CI], 34-77) and 68% (95% CI, 42-84), respectively. Five (25%) patients died because of progressive disease, and 1 patient died because of multiorgan failure.

Twenty patients received an imaging dose of ¹¹¹In-basiliximab/DOTA after cold basiliximab. Dosimetry was performed for the 17 patients who had the minimum required imaging data sets (>2 ¹¹¹In-basiliximab/DOTA planar scans, and at least 1 CT scan). Figure 3 shows ¹¹¹In-basiliximab/DOTA SPECT and ¹⁸FDG-PET images of 1 patient (both images superimposed on coregistered CT). No patients had altered biodistribution.

Calculated radiation absorbed doses per unit injected activity from ⁹⁰Y-basiliximab/DOTA are summarized in supplemental Table 4. Effective dose/injected activity was 5.0 ± 0.8 cGy/mCi. Organs with the highest doses/injected activity were the spleen (51 ± 21 cGy/mCi), heart (40 ± 7 cGy/mCi), and liver (25 ± 6 cGy/mCi). Red marrow received 3.1 ± 0.4 cGy/mCi. For the patients in the 0.60-mCi/kg cohort, average doses to these organs were: spleen, 24 ± 12 Gy; heart 17 ± 3 Gy; liver, 10 ± 2 Gy; and red marrow 1.2 ± 0.1 Gy. The effective dose for this cohort was 2.1 ± 0.4 Gy. Tumor doses were not measured because most patients in this HCT population had no macroscopic disease.

Basiliximab/DOTA clearance from blood plasma and serum was observed to conform to a biphasic fit. The mean and standard deviation of the half-lives of the α and β phases were 13 ± 13 hours and 153 ± 82 hours, respectively.

Of 20 patients treated with ⁹⁰Y-aTac-BEAM AHCT, 16 had tumor tissue evaluable for CD25 expression by immunohistochemistry (for the other 4 patients, tumor tissue was either necrotic or unavailable for evaluation.) CD25 expression in malignant T cells was found in 12 of 16 (75%) samples (supplemental Table 5). Data for number of samples with a given percentage of malignant T cells positive for CD25 were as follows: dim to moderate (not measurable), n = 4; 25%, n = 1; 30%, n = 1; 40%, n = 1; 50%, n = 2; and >80%, n = 3. One other case was mostly necrotic, with some CD25⁺ cells. Background tissue and tumor infiltrating lymphocytes had no or low CD25 expression. No correlation was observed between CD25 expression and either treatment response or efficacy in this limited data set.

Our study shows that CD25–directed ⁹⁰Y-basiliximab/DOTA RIT, when given in combination with standard-dose BEAM (⁹⁰Y-aTac-BEAM) as conditioning for subsequent AHCT is safe in patients with PTCL requiring AHCT. The radioactivity dose was escalated from 0.4 to 0.6 mCi/kg without additional toxicity attributable to RIT, including any delays in engraftment or complications of AHCT. Based on biodistribution data, a ⁹⁰Y-basiliximab/DOTA dose of 0.6 mCi/kg delivers lower radiation doses to vital organs including the heart, lungs, kidneys, liver, and spleen compared with more traditional radiation therapy, such as myeloablative TBI.³⁶ 

The targeted RIT approach we evaluated has the potential to provide more robust conditioning for AHCT than do regimens currently used for patients with PTCL. Although this is a small study with varying doses of RIT and different PTCL histologies, we have observed encouraging efficacy for the ⁹⁰Y-aTac-BEAM AHCT regimen. Two-year PFS and OS were 59% (95% CI, 34-77) and 68% (95% CI, 42-84), respectively, which is at least comparable with the data for chemotherapy-only conditioning regimens such as the Nordic Lymphoma Group study, which reported on 160 patients enrolled with intent to transplant; 5-year OS and PFS were 51% and 44%, respectively. The 5-year OS for patients who did not undergo transplant was 28%, clearly showing an advantage for the consolidation approach.³ Note that all our patients had achieved a state of remission before AHCT, and 90% of patients were in CR1. This is in line with the standard practice of AHCT as consolidation after initial chemotherapy. More recently, The Netherlands Cancer Registry analysis of 1427 patients showed improved outcomes with consolidative AHCT in first remission; 5-year survival of patients with ALK-ve ALCL, angioimmunoblastic T cell lymphoma (AITL), and PTCL-NOS was 81% with consolidative transplant vs 39% for those not undergoing ASCT.³⁸ However, outcomes are far from satisfactory, with most patients relapsing within 2 years after transplant.

The use of targeted treatments in PTCL are starting to improve outcomes in specific subgroups. For patients with tumors that express CD30, the phase 3 ECHELON-2 study was the first to demonstrate the potential power of targeted therapies in the treatment of PTCL, showing that brentuximab vedotin plus CHP (A + CHP) improved PFS and OS compared with CHOP alone (5-year OS, 70 % vs 61%).³⁹ This effect was most pronounced in patients with ALCL in which CD30 is uniformly expressed in the tumor cells. The CD30 cutoff in the study was >10%, hence other CD30-expressing histologies were also included. Subgroup analysis confirmed that the difference in outcomes was not as pronounced in non-ALCL subtypes.⁴⁰ A subanalysis of this study suggested that consolidative transplant was associated with improved outcomes compared with patients who did not undergo ASCT. Median PFS was 62 months vs 32 months in patients without ALCL.⁴¹ In another study, the addition of the histone deacetylase inhibitor romidepsin to CHOP (Ro-CHOP) vs CHOP for initial therapy in PTCL failed to show a benefit over CHOP alone.⁴² However, a subanalysis of the study showed an improved PFS in the Ro-CHOP arm in patients with Tfh subtypes (19.5 vs 10.6 months) demonstrating an advantage of adding epigenetic therapy to conventional treatment in specific subtypes⁴³ in which epigenetic alterations play a significant role in pathogenesis.^44,45 Taken together, these studies suggest that in the treatment of PTCL, targeting specific pathways may be more effective than simply intensifying chemotherapeutic regimens.

A meta-analysis to evaluate the efficacy of AHCT as first-line consolidation for nodal PTCL has shown that this approach can improve survival. In the COMPLETE registry that prospectively collected data on the role of AHCT as consolidation in CR1 of PTCLs, the median PFS was 57.6 months.^37,46 However, even after AHCT, up to 50% of patients relapse.⁴⁷ We sought to improve the conditioning regimen for AHCT in PTCL by adding CD25-targeted radioimmunotherapy. External beam TBI combined with high-dose chemotherapy has been part of conditioning regimens for both autologous and allogeneic stem cell transplants. Although inclusion of TBI has enabled better disease control and reduced risk of disease recurrence, increased normal-organ toxicity and risk of secondary malignancies have limited use of TBI in conditioning regimens.⁴⁸ For PTCL, there is 1 prospectively designed study from 2000 to 2006 that used TBI with high-dose cyclophosphamide as conditioning before an autologous transplant.⁴⁹ These patients were either in CR or partial response at the time of transplant, and the 3-year OS rate was 48%, with a treatment-related mortality of 3.6%. These results were considered to be in line with the expected outcomes from other similar studies in B-cell lymphoma but this was the first study of this approach in PTCL.

RIT can be used to deliver radiation directly to tumor sites while simultaneously decreasing radiation dose to normal tissues, and also providing a mechanism for adding antitumor cytotoxicity to monoclonal antibody therapy, hence helping overcome the problem of myelosuppression in AHCT.⁵⁰ Initial trials combining RIT with high-dose chemotherapy and stem cell rescue were conducted using anti-CD20 [¹³¹I]tositumomab (Bexxar) with high-dose etoposide and cyclophosphamide followed by stem cell rescue in relapsed B-cell lymphomas. Comparison with historical controls treated with TBI plus etoposide and cyclophosphamide showed improved PFS and OS.⁵¹ The CD20-chelator fusion molecule, [⁹⁰Y]ibritumomab tiuxetan (Zevalin), was found to be safe with no added toxicity, increased risk of graft failure, or transplant-related toxicity in combination with high-dose chemotherapy for relapsed B-cell lymphoma.^21,23 Importantly, the maximal dose of radiation delivered to critical organs was relatively low (25-27 cGy). Other targets for pretransplant conditioning with combined RIT and chemotherapy that have been explored include CD22 in non-HL⁵² and CD25 in HL.²² 

The results of our small phase 1 trial are encouraging and warrant further study of targeted RIT + BEAM for pre-AHCT conditioning in PTCL. No added toxicity was noted with the addition of CD25-directed RIT to BEAM. Definitive conclusions regarding efficacy are limited by the small number of patients treated, the varying histologies, and varying CD25 expression. However, 3 of 5 patients with progressive disease had no expression of CD25 on tumor cells, an observation that should be explored in a larger cohort. However, the aim of the study was to evaluate safety in the peritransplant period. A sufficiently large patient sample given the recommended 0.6 mCi/kg dose combined with sufficiently long follow-up would enable the efficacy (measured in terms of PFS and OS) of ⁹⁰Y-aTac-BEAM AHCT to be compared with that of BEAM-based AHCT. A larger patient sample would also improve evaluation of the expected correlation between CD25 expression in tumor and response to ⁹⁰Y-aTac RIT. Long-term toxicity and secondary malignancies should be evaluated well beyond the 24-month follow-up interval of this study. All patients in this study were in CR at time of enrollment, with little or no macroscopic residual disease as judged by FDG-PET/CT. We did not include any patients with residual disease because it is not congruent with our clinical practice to offer AHCT to these patients.

This study included various PTCL histologies representative of those seen in standard clinical practice. Future studies may be designed to identify specific histologies for which targeted RIT is especially beneficial and to better evaluate the role of CD25 positivity in relation to response. Patient selection for ⁹⁰Y-aTac-BEAM AHCT might be improved by adding a means for detecting minimal residual disease in the peripheral blood. Other potential targets for RIT in PTCL include CD30, CD70, and CCR4.^53-55 To obtain higher tumor cell kill, basiliximab/DOTA could be labeled with an α-emitting radionuclide such as ²²⁵actinium.

This research was supported, in part, by the National Institutes of Health (grant P30CA033572; City of Hope National Medical Center).

Contribution: J.Z., J.P., and D.C. provided conceptualization and design; E.P.S., J.P., N.-C.T., J.S., J.Y.S., V.A., D.Y., E.K.P., S.D., A.S., R.N., J.Z., A.F.H., N.A.K., and A.P.N. collected and assembled data; P.Y., E.K.P., V.E.B.-A., D.C., and J.E.S. contributed vital new reagents; V.A., N.-C.T., J.P., J.Y.S., J.W., J.Z., A.M.W., D.L.S., J.R.B., and S.J.F provided data analysis and manuscript revision; and J.Z. wrote the first draft of the manuscript.

Conflict-of-interest disclosure: J.Z. is a consultant for Kyowa Kirin, Seattle Genetics, Verastem, Daiichi Sankyo, and Mundi Pharma; serves of the speakers bureau of Seattle Genetics, SecureBio, Daiichi Sankyo, and AbbVie; and reports research support from Seattle Genetics, SecureBio, Daiichi Sankyo, and AbbVie. J.W. reports grant support from RefleXion Inc, Varian Inc, Accuray Inc, Telix Inc, and Blue Earth Diagnostics, Inc. A.F.H. reports research funding from Bristol Myers Squibb, Merck, Genentech, Inc, F. Hoffmann-La Roche Ltd, Gilead Sciences, Seattle Genetics, AstraZeneca, and ADC Therapeutics; and reports consultancy with Bristol Myers Squibb, Merck, Genentech, Inc, F. Hoffmann-La Roche Ltd, Kite Pharma/Gilead, Seattle Genetics, Karyopharm, Takeda, Tubulis, and AstraZeneca. S.D. reports research funding from Bayer. A.M.W. reports consultancy and board membership with ImaginAb; and reports consultancy with AstraZeneca, and Novartis Institute for Biomedical Research. The remaining authors declare no competing financial interests.

Correspondence: Jasmine Zain, Hematology and Hematopoietic Cell Transplantation, City of Hope Hematology, 1500 E Duarte Rd, Duarte, CA 91010; email: jazain@coh.org.