Cure of Disseminated Human Lymphoma with [177Lu]Lu-Ofatumumab in a Preclinical Model

Visual Abstract

Non-Hodgkin lymphoma (NHL) is a common hematologic malignancy, with over 80,000 new cases and 20,000 deaths estimated for the United States in 2022 (1). The standard of care for many cases of NHL involves chemotherapy and immunotherapy targeting the CD20 protein, which is highly expressed on most NHL cells, with murine/human chimeric rituximab used most commonly. Although this chemotherapy-with-immunotherapy combination is usually initially effective, many cases are refractory or undergo relapse, indicating the need for improved therapies.
Radioimmunotherapy joined clinical practice 2 decades ago with Food and Drug Administration approval of 2 anti-CD20 radioimmunotherapies for lymphoma: Zevalin ([ 90 Y]Y-ibritumomab tiuxetan; Acrotech Biopharma, Inc.) and Bexxar (tositumomab and 131 Itositumomab; GlaxoSmithKline), which use murine-derived antibodies radiolabeled with b-particle-emitting radioisotopes. Because of potential immune reactions, these antibodies were approved for only a single therapeutic dose. 90 Y (half-life [t1 =2 ], 2.7 d) emits highenergy b-particles (average, 934 keV), whereas 131 I emits lowerenergy b-particles (average, 187 keV), with average ranges in tissue of 3,800 mm and 360 mm, respectively (2), enabling killing over many cell diameters. Thus, in addition to working against individual tumor cells, b-particles may work against larger tumors, tumor-cell aggregates with imperfect antibody access, and heterogeneous tumors, although with potential off-target damage. Despite longterm safety and clinical effectiveness, Bexxar has been discontinued in the United States and Zevalin is applied infrequently (3), in part because of economic and logistic concerns that were present when they were introduced and because of competing nonradioactive therapies (4).
Some concerns that limited the use of Bexxar and Zevalin have been overcome with greater integration of radiopharmaceutical therapy into medicine (5), as exemplified by the Food and Drug Administration approval of 177 Lu-labeled agents for prostate cancer treatment (Pluvicto; Advanced Accelerator Applications (6)) and neuroendocrine tumors (Lutathera; Advanced Accelerator Applications (7)). 177 Lu (t1 =2 , 6.6 d) emits b-particles of 149 keV on average, with an average tissue range of 220 mm. Emission of low-abundance g-particles by 177 Lu permits imaging by SPECT.
Here, we describe the synthesis and evaluation of [ 177 Lu]Luofatumumab. We present in vitro characteristics, dosimetry estimation, and subcutaneous tumor targeting. We also show that [ 177 Lu] Lu-ofatumumab therapy results in long-term survival and elimination of tumor cells in a murine model of disseminated human lymphoma.

Reagents and Cell Culture
Ofatumumab (IgG1 k; Novartis) was purchased from the Washington University clinical pharmacy, and human IgG1 k was purchased from BioXcel. Raji cells and Raji-luc cells stably expressing luciferase (10) were cultured as previously described (9). SCN-CHX-A99-DTPA ({[(R)-2-amino-3-(4 isothiocyanatophenyl)propyl]-trans(S,S)cyclohexane-1,2-diamine-pentaacetic acid}) was from Macrocyclics, size-exclusion chromatography columns from Fisher Scientific, and D-luciferin from GoldBio. Sigma provided human serum, sodium acetate, diethylenetriamine pentaacetate, tetramethylammonium acetate, and L-sodium ascorbate. 177 Lu from the University of Missouri was dissolved in 0.2 M HCl. Silica gel thin-layer chromatography paper was from Agilent, and the 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) salt assay was from Promega. Antibody Conjugation, Radiolabeling, Thin-Layer Chromatography, Mass Spectrometry, and Fast-Performance Liquid Chromatography Antibody was incubated with SCN-CHX-A99-DTPA in 0.1 M sodium carbonate, pH 9.0, at a chelator-to-antibody molar ratio of 8:1 for 1 h at 37 C and purified by size-exclusion chromatography into 0.5 M NH 4 OAc, pH 7.0. A 477-MBq quantity of 177 Lu was added to 400 mg of CHX-A99-DTPA-antibody with 20 mM NH 4 OAc, pH 7.0. After 2 h at 37 C, DTPA was added to 5 mM final concentration, followed by size-exclusion chromatography purification into saline and the addition of a 10 mg/mL concentration of L-sodium ascorbate. Thin-layer chromatography and fast-performance liquid chromatography were done as previously described (9). Radiochemical yield was assayed with a CRC55-tW dose calibrator. Chelate number was determined using a Fisher Scientific Exactive Plus EMR mass spectrometer operating at a mass (m)-to-charge (z) range from 800 to 12,000 and a resolving power of 8,750 or 17,500 at 300 m/z. Data were analyzed using Protein Metric Intact software.  177 Lu were added to 10% human serum in 20 mM NaOAc 150 mM NaCl pH 7.0 with 10 mg/mL L-SA and incubated at 37 C. Another aliquot of [ 177 Lu]-ofatumumab was incubated at 4 C in buffer without serum and with 10 mg/mL L-SA. Aliquots were analyzed by thin-layer chromatography at 0, 1, 5, and 7 d. Immunoreactivity was assayed as previously described (9). To assay affinity, 2.5 3 10 6 Raji cells without or with 10 mg of ofatumumab were incubated with [ 177 Lu]Lu-ofatumumab, washed after 4 h at 23 C, and g-counted. To assay cell killing, 2 3 10 6 Raji-luc cells in 1 mL of RPMI medium with 10% heat-treated fetal bovine serum were exposed to no treatment, ofatumumab, [ 177 Lu]Lu-IgG, or [ 177 Lu]Lu-ofatumumab, with cognate unlabeled antibody added to 20 mg total. After 14 h at 37 C, the cells were washed and 20% were resuspended in fresh medium for an additional 168 h followed by MTS assay.

Dosimetry Estimation
Naïve 5-to 6-wk-old C57Bl6/N mice injected intravenously with 370 kBq (10 mg) of [ 177 Lu]Lu-ofatumumab were killed 4 h, 1 d, 2 d, 5 d, or 11 d later, and tissue and organs were g-counted. Bone was counted after marrow separation. Urine and feces were collected at 4 h, 1 d, and 2 d. Organ residence times were calculated by analytic integration of single or multiexponential fits of the time-activity curve and scaled to human organ weight by relative organ mass scaling (11), which was not applied to the gastrointestinal tract organs. To estimate human radiation dose, residence times were entered into OLINDA, version 2.2, using the MIRD adult-female model and organ weights from International Commission on Radiological Protection publication 106 (12). The calculated radiation dose includes contributions from band g-rays from 177 Lu within the organ, neighboring organs, and remainder of the body.
Therapeutic Studies, Mouse Weight, and Bioluminescent Imaging R2G2 mice (10 per group) injected intravenously with 1 3 10 6 Raji-luc cells and either left untreated or injected 4 d later with ofatumumab, [ 177 Lu]Lu-IgG, or [ 177 Lu]Lu-ofatumumab. When used, 20 mg of antibody were injected per mouse. Bioluminescent images were acquired as previously described (13). Mice were killed if they experienced hind-limb paralysis, lost more than 20% of their body weight, or had other signs of morbidity.
To estimate human dosimetry, integrated time-activity curves for [ 177 Lu]Lu-ofatumumab were calculated (Supplemental Table 2). The longest (59.7 h) was in the blood, with extended time-activity curves seen in the blood-rich heart cavity, lung, and liver. Because of its large mass, muscle had the second longest time-activity curve, at 39 h. The adult human female model (Table 1) showed estimated dosimetry of 0.2-0.5 mSv/MBq in most organs, with the largest dose being to the heart wall (1.02 mSv/MBq) and lesser doses found for liver, spleen, and kidney (0.36, 0.48, and 0.43 mSv/MBq, respectively). Estimated doses to the osteogenic cells (bone surfaces) and red marrow were 0.82 and 0.54 mSv/MBq, respectively. The estimated effective dose was 0.36 mSv/MBq.

Biodistribution of [ 177 Lu]Lu-Ofatumumab in Mice with Subcutaneous Raji-Cell Tumors
Biodistribution was investigated in R2G2 mice with subcutaneous Raji-cell tumors (Fig. 1). These mice are proficient in doublestrand DNA-break repair and are less likely to show artifactual radiation toxicity than are repair-deficient Prkdc SCID mice (16).
[ 177 Lu]Lu-ofatumumab was injected at a low activity (370-444 kBq) to limit therapeutic effect, and biodistribution was determined 1, 3, and 7 d later (3-16 mice per time point). Blood decreased from about 13 to 6 %IA/g, with a similar splenic distribution. Liver levels were about 5%, and marrow was 10 %IA at 1 d and 5 %IA/g at 7 d. Bone distribution was 2-3 %IA/g. Tumor targeting was 11, 15, and 14 %IA/g at 1, 3, and 7 d, respectively. cells were quantified by bioluminescent imaging (13). After injection, these cells disseminate to many organs (10,13,17,18), with hind-limb paralysis being a typical cause for killing of the animal due to growth in and around the spine.
Four days after cell injection, the mice either were left untreated or were treated with native ofatumumab, 8 (Fig. 2), weight (Supplemental Fig. 2), and bioluminescence (Fig. 3A) were tracked for 221 d. Representative bioluminescent images at selected time points are shown in Figure 3B, and images of all mice just before they died or were killed for cause or study termination are shown in Figure 4.
The median survival of untreated mice was 19 d, with none surviving beyond 22 d. Unlabeled ofatumumab yielded a median survival of 46 d, superior to untreated mice (Mantel-Cox, P , 0.0001), with 1 mouse surviving without weight loss or increased bioluminescence. An 8.51-MBq dose of [ 177 Lu]Lu-IgG yielded 0 of 10 surviving mice and a median survival of 25 d, which was not different from that of untreated mice. For all 3 groups, increased bioluminescence and weight loss occurred before death or killing for cause.
A 0.74-MB dose of [ 177 Lu]Lu-ofatumumab yielded median survival of 59 d (9/10 mice not surviving), with increased bioluminescence and weight loss before death or killing for cause. This survival was superior to that of untreated mice (Mantel-Cox, P , 0.0001) but not to that of mice receiving treatment with unlabeled ofatumumab. Hind-limb paralysis was frequently associated with death or killing for cause (Supplemental Table 3).
Notable therapeutic efficacy resulted from treatment with 8.51 MBq of [ 177 Lu]Lu-ofatumumab, with 9 of 10 mice surviving with continuous low bioluminescence (Figs. 3 and 4). This survival was greater than that of untreated mice and of mice treated with unlabeled ofatumumab, 8.51 MBq of [ 177 Lu]Lu-IgG, or 0.74 MBq of [ 177 Lu]Lu-ofatumumab (Mantel-Cox, P , 0.0003 for all comparisons). One mouse succumbed at 117 d, but this death appeared unrelated to tumor burden or therapy as no weight loss or increased bioluminescence occurred. Surviving mice displayed weight loss from 10 to 35 d after cell injection but recovered and gained weight.
To determine how quickly therapy affected tumor cells, bioluminescence slopes from 1 to 18 d after initiation of therapy were compared ( Fig. 5; Supplemental Fig. 3

DISCUSSION
Our preclinical studies add to prior work demonstrating the potential of radiolabeled anti-CD20 antibodies to treat NHL. We show that [ 177 Lu]Lu-ofatumumab can be produced with high radiochemical yield and purity, excellent affinity, good stability and immunoreactivity, and potent cell killing. Additional advances include using a fully human anti-CD20 and 177 Lu, which have broad applicability in radiotherapy of cancer. In a model of rapidly progressing disease, we evaluated [ 177 Lu]Lu-ofatumumab therapy using dose-response studies and bioluminescence monitoring of tumor-cell burden. A single 8.51-MBq dose of [ 177 Lu]Lu-ofatumumab displayed curative efficacy.
Human dosimetry estimates predict that the highest dose from [ 177 Lu]Lu-ofatumumab (1.02 mSv/MBq) will be to the heart wall. The relatively radiation-resistant liver and spleen showed 0.36 and 0.48 mSv/MBq, respectively. The predicted dose to red marrow is 0.54 mSv/MBq, and hematologic toxicity likely will be dose limiting in clinical use, as was found with Bexxar, Zevalin, [ 177 Lu]Lu-J591 (19), [ 177 Lu]Lu-G250 anti-CAIX (20), and [ 177 Lu]Lu-rituximab (21). As 2 Sv is a typical maximal dose for acceptable hematologic toxicity without stem cell support, delivering this radiation to the marrow would be tolerable. As there may be patient-to-patient variability with [ 177 Lu]Lu-ofatumumab due to cross reactivity with normal CD20positive cells, our dosimetry data provide guidance for activity administration to humans. Dosimetric estimation could also potentially be obtained using a PET imaging surrogate, such as [ 89 Zr]Zr-ofatumumab (9,22).
The stable in vivo chelation of 177 Lu by CHX-A99-DTPAofatumumab agrees with the results of others using this chelatorradionuclide combination (23,24). Although it has been suggested that, for stable 177 Lu chelation, macrocyclic DOTA requires high temperatures incompatible with maintaining antibody function (24,25), experiments show that this is not the case (26,27). Thus, CHX-A99-DTPA and DOTA both appear practical for chelation of 177 Lu to antibodies and antibody fragments. After initial weight loss, these mice gained weight, suggesting no or low whole-body toxicity. The internalization of ofatumumab after CD20 binding (30) and the residualization of 177 Lu within the cell may contribute to its therapeutic efficacy. Moreover, the lack of murine sequences in [ 177 Lu]Lu-ofatumumab suggests a potential for fractionated therapy or repeated treatments. In an interesting approach, with relatively small subcutaneous tumors of rituximab-resistant Raji cells, Malenge    a-particle therapy is another potential approach to treating lymphoma. Using a murine Raji-cell disseminated lymphoma model, [ 213 Bi]Bi-rituximab (t1 =2 , 45.6 min) was typically curative when tumor burden was low (4 d after cell injection) but not when it was higher (18), perhaps because of lack of time to target larger tumor masses before decay. Similarly, [ 149 Tb]Tb-rituximab (t1 =2 , 4.2 h) therapy initiated 2 d after Daudi-cell intravenous injection increased survival (31). A 1F5 anti-CD20 antibody with chelated 211 At (t1 =2 , 7.2 h) was 80% curative when injected 6 d after intravenous cell injection with supporting stem-cell transplantation but only slowly reduced the growth of subcutaneous tumors (32). On the basis of these results and on the multiday tumor-targeting pharmacokinetics of intact antibodies, radioimmunotherapy of larger tumor masses with intact antibodies will likely be most successful using radioisotopes that permit tumor localization before decay, including 177 Lu or a-particle-emitting 225 Ac with its 10-d half-life.
Our studies add to the literature demonstrating the effectiveness of 177 Lu-radiopharmaceuticals in cancer therapy. We found remarkable effectiveness in micrometastatic disease, and the 6.6-d half-life and multiple-cell-diameter killing range of 177 Lu suggests that [ 177 Lu]Lu-ofatumumab may be effective against larger tumors.
Although initial anti-CD20 radioimmunotherapies showed limited commercial success for several reasons, we suggest that a reevaluation of next-generation band a-particle therapies is in order. [ 177 Lu]Lu-ofatumumab CD20-targeted radioimmunotherapy may be an effective approach for therapy of NHL or other CD20expressing diseases.

DISCLOSURE
This study was supported in part by the National Institutes of Health (R01CA240711, R01CA229893, and R01CA201035 to Daniel Thorek), the Children's Discovery Institute (MC-II-2021-961 to Diane Abou), and the NIGMS (9995P41GM103422 to the Washington University Biomedical Mass Spectrometry Resource). Richard Wahl is on the scientific advisory board of Clarity Pharmaceuticals, Voximetry, and Seno Medical; has stock options in Clarity Pharmaceutical and Voximetry; receives honoraria from Bristol Myers Squibb, Actinium Pharmaceuticals, Jubilant Draximage, and ITM; and receives research support from Actinium Pharmaceuticals, BMS, Bayer, Siemens, and White Rabbit AI. Diane Abou and Daniel Thorek have an advisory board role for, and own stock in, Diaprost AB and Pharma15. Richard Laforest is a consultant to Curium Pharmaceuticals. No other potential conflict of interest relevant to this article was reported.