ANTICANCER RESEARCH 33: 85-96 (2013)

Evaluating Antigen Targeting and Anti-tumor Activity of a New Anti-CD37 Radioimmunoconjugate Against Non-Hodgkin’s

JOSTEIN DAHLE1, ADA H.V. REPETTO-LLAMAZARES1, CAMILLA S. MOLLATT2, KATRINE B. MELHUS2, ØYVIND S. BRULAND3,4, ARNE KOLSTAD4 and ROY H. LARSEN5

1Nordic Nanovector AS, Oslo, Norway; 2Departments of Radiation Biology, Institute for Cancer Research, and 3Oncology, The Norwegian Radium Hospital, Oslo, Norway; 4Insititute for Clinical Medicine, University of Oslo and 5Sciencons Ltd., Oslo, Norway

Abstract. The monoclonal against CD20, , Beta-emitting radioimmunoconjugates have significant alone, or as part of combination therapies, is standard therapy antitumor activity in patients with relapsed or refractory B-cell for non-Hodgkin’s B-cell lymphoma. Despite significantly better lymphoma (1-4), including those refractory to rituximab, a clinical results obtained for beta-emitting radioimmuno- to CD20 (5, 6) and chemotherapy (7). conjugates (RICs), RICs targeting CD20 are not commonly used (RIT) is administered with large in medical practice, partly because of competition for the CD20 quantities of unlabeled “cold” antibody to CD20 one week target. Therefore, novel therapeutic approaches against other and 4 h prior to radiolabeled to CD20. Such a antigens are intriguing. Here, the binding properties of a novel priming dose is necessary to reduce antibody binding to antibody against CD37 (tetulomab) were compared with those normal B-cells by depleting peripheral blood B-cells and of rituximab. The therapeutic effect of 177Lu-tetulomab was lymph node B-cells (8, 9). Thus, sufficient amounts of compared with 177Lu-rituximab on Daudi cells in vitro. The radiolabeled antibody can bypass these sites, penetrate less- biodistribution, therapeutic and toxic effects of 177Lu-tetulomab accessible compartments, such as the lymph nodes, and and unlabeled tetulomab were determined in SCID mice injected target tumor cells. However, both clinical and experimental with Daudi cells. The affinity of tetulomab to CD37 was similar studies in mice have shown that in some circumstances, even to the affinity of rituximab to CD20, but the CD37-tetulomab quite low rituximab concentrations in the blood can reduce complex was internalized 10-times faster than the CD20- tumor cell targeting and thus impair the clinical efficacy of rituximab complex. At the same concentration of antibody, CD20-directed RIT (10). Furthermore, the majority of 177Lu-tetulomab was significantly more efficient in inhibiting cell patients selected for CD20-based RIT have received several growth than was 177Lu-rituximab, even though the cell-bound cycles of “cold” rituximab, which may result in selection of activity of 177Lu-rituximab was higher. Treatment with 50 and tumor cells with low CD20 expression and thus could lower 100 MBq/kg 177Lu-tetulomab resulted in significantly increased the effect of subsequent anti-CD20 treatments (11-13). survival of mice, compared with control groups treated with By targeting B-cell antigens, other than CD20, a better tetulomab or saline. The CD37 epitope recognized by tetulomab response might be achieved. CD37 is a heavily glycosylated was highly expressed in 216 out of 217 tumor biopsies from 40- to 52-kDA and a member of the tetraspan patients with B-cell lymphoma. This work warrants further pre- transmembrane family of (14, 15). CD37 clinical and clinical studies of 177Lu-tetulomab. internalizes and has modest shedding in transformed B-cells expressing the antigen (16, 17). During B-cell development, CD37 is expressed in cells progressing from pre-B to peripheral mature B-cell stages and is absent on terminal This article is freely accessible online. differentiation to plasma cells (18). Given its relative B-cell selectivity, CD37 thus represents a valuable target for Correspondence to: Jostein Dahle, Nordic Nanovector AS, therapies in chronic lymphocytic (CLL), hairy-cell Kjelsåsveien 168 B, 0884 Oslo, Norway. Tel: +47 98458850, Fax: +47 22580007, E-mail: [email protected] leukemia (HCL), B-cell non-Hodgkin’s lymphoma (NHL) and other B-cell malignancies. Key Words: Radioimmunotherapy, CD37, non-Hodgkin’s CD37 has been previously studied as a target for RIT lymphoma, 177Lu. using 131I-labeled MB-1 (16). Recently, there has been a

0250-7005/2013 $2.00+.40 85 ANTICANCER RESEARCH 33: 85-96 (2013) revival in interest in the targeting of CD37. Both an Fc- (26). Cell densities of up to 3×108 cells/ml were used. All conjugates engineered monoclonal antibody and a small modular used in experiments had immunoreactivity greater than 50 %. immuno-pharmaceutical have been developed for the targeting of CD37 in CLL (19, 20). Binding parameters. The association rate constant, ka, and the equilibrium dissociation constant, K , were measured for tetulomab At the Norwegian Radium Hospital a murine IgG1 d and rituximab, and the mean number of binding sites, Bmax, was antibody, HH1 (tetulomab, according to the nomenclature for determined for Raji, Rael and Daudi cells using a one-step curve naming of antibodies), was developed against CD37 in the fitting method (27). Specific binding was measured as a function of 1980s (21). In the present study we investigated binding to time and antibody concentration, and the solution of the differential different subtypes of NHL and compared binding properties equation describing the net rate of formation of the antigen-antibody of tetulomab with the chimeric IgG1 antibody rituximab. complex was fitted to the experimental data points using ka, Kd and Furthermore, the therapeutic effects of 177Lu-tetulomab, Bmax as parameters. Five million cells/ml were used, with four concentrations of “hot” antibody (100 ng/ml, 1000 ng/ml, 5000 ng/ml 177Lu-rituximab, “cold” tetulomab and “cold” rituximab was 177 and 10000 ng/ml) and seven incubation time points (5 min, 10 min, compared in vitro. Increasing concentrations of Lu- 20 min, 30 min, 1 h, 1.5 h and 2 h). After incubation, cells were tetulomab were used to treat SCID mice injected intravenously washed twice with PBS, and then counted in a gamma counter (Cobra with human Daudi lymphoma cells. This animal model (22), is II; Packard Instrument Company, Meriden, CT, USA). a worst-case scenario model of NHL because the tumor cells will, to a large degree, end up in the bone marrow. This Internalization and retention of 177Lu-tetulomab and 177Lu- localization will result in irradiation of bone marrow cells and rituximab. Daudi cells were harvested, counted and diluted to a 6 177 corresponding hematological toxicity. The biodistribution of density of 2×10 cells/ml and incubated with 1 μg/ml Lu- tetulomab, 177Lu-rituximab, 125I-tetulomab or 125I-rituximab. Some antibodies in SCID mice may be different from that in other parallels were pre-blocked with the corresponding “cold” antibody mouse strains because of low concentrations of endogeneous (100 μg/ml). The cells were incubated for 0, 10, 20, 30, 60, 120 and antibodies in the blood (23, 24). In addition, because of the 180 min. Subsequently, the cells were centrifuged and washed twice SCID mutation this mouse strain cannot repair DNA double- with medium, and the supernatant and the wash medium were strand breaks and is thus very sensitive to ionizing radiation collected for counting. Cells were incubated with 1 ml stripping (25). Regardless of the limitations of this model, a significant buffer (150 mM NaCl and 50 mM glycine, pH 2.6) for 10 min, at therapeutic effect of 177Lu-tetulomab was shown at a dosage room temperature. The cells were washed twice and the supernatants and the cells were counted using a calibrated gamma level resulting in tolerable toxicity. detector (Cobra II). In another experiment, 106 Daudi cells/ml medium were Materials and Methods incubated with 1 μg/ml 125I- or 111In-labeled tetulomab or rituximab for one hour, washed twice with medium and incubated further for Cell lines. The CD20- and CD37-expressing B-cell lymphoma cell four days. The cell-bound activity was determined immediately after lines Raji, Rael and Daudi were used in the current study (LGC washing and after four days of incubation by measuring the number Standards, Boras, Sweden). Single-cell suspensions were grown in of cells (Vi Cell Viability Analyzer, Beckman Coulter, Fullerton, RPMI 1640 medium supplemented with 10 % heat-inactivated FCS, CA, USA) and the amount of radioactivity with the gamma detector 1% L-glutamine and 1% penicillin-streptomycin (all from PAA, (Cobra II). Linz, Austria), in a humid atmosphere with 95% air/5% CO2. Effect of 177Lu-tetulomab and 177Lu-rituximab on growth of Daudi Labeling of antibodies with 177Lu. Tetulomab and rituximab were first cells in vitro. One milliliter of Daudi cell suspension (1×106 labeled with a chelator, 2-(4-isothiocyanatobenzyl)-1,4,7,10 cells/ml) was pipetted into each of 24 tubes. Six tubes each were tetraazacycododecane-1,4,7,10-tetraacetic acid (p-SCN-Bn-DOTA, blocked with 100 μg/ml of either tetulomab or rituximab and DOTA) (Macrocyclics, Dallas, TX, USA). DOTA was dissolved in incubated for 30 min at 37˚C. Subsequently, either 177Lu-tetulomab 0.05 M HCl, added to the antibody in a 5:1 ratio and the pH adjusted or 177Lu-rituximab was added to a final concentration of 0, 1, 2.5, 5, to 8.5-9 by washing with carbonate buffer using AMICON-30 10 or 20 μg/ml and the cells incubated further at 37˚C. The specific centrifuge tubes (Millipore, Cork, Ireland). The solution was shaken activity was 91.6 kBq/μg for 177Lu-tetulomab and 136.6 kBq/μg for for 60 min at room temperature, and the reaction was terminated by 177Lu-rituximab. The amount of added activity was measured during adding 50 μl of 200 mM glycine solution (per mg antibody). To the incubation period with a gamma detector (Cobra II). After two remove free chelator, the conjugated antibody was centrifuged 4-5- hours, half of the cells were washed and the cell-bound activity was times with phosphate buffered saline (PBS) (PAA) at 1:10 dilution measured, while the other half of the cells was incubated overnight using AMICON-30 centrifuge tubes (Millipore). Before labeling with before washing and measurement of cell-bound activity. For 177Lu (Perkin Elmer, Boston, MA, USA) the PBS buffer was changed measurement of cell growth, 50,000 cells from each tube were to ammonium acetate buffer, pH 5.4, using the same centrifuge tubes. seeded in three wells in a 12-well plate and the number of cells was 177Lu was then added to 0.5 mg DOTA-Ab, and shaken for 45 min at measured at several time points over the next 14 days using an 37ºC. Specific activity was between 50 and 150 MBq/kg. automatic imaging system (Clone Select Imager, Molecular Devices Ltd., New Milton, Hampshire, UK) that captured images of each Immunoreactivity. The quality of the radioimmunoconjugates was well, analyzed the images and calculated the number of cells in each measured using lymphoma cells and a modified Lindmo method well. Growth delay was estimated as the difference in time between

86 Dahle et al: New Anti-CD37 Radioimmunoconjugate for NHL

exponential growth of the control cells and that of the treated cells. Table I. The mean number of antigens (Bmax) on Raji, Rael and Daudi Survival was estimated by extrapolating the exponential part of the cells, the equilibrium dissociation constant (Kd) and the association rate growth curve to the y-axis (28). Growth delay factor was calculated constant (ka) for tetulomab and rituximab. by dividing the delay in growth to 100,000 cells/ml for cells treated a –1 –1 with 177Lu-tetulomab with the corresponding delay in growth for Antibody Cell line Bmax (Ag/cell) Kd (nM) ka (nM h ) cells treated with 177Lu-rituximab. Tetulomab Raji 146000±7600 6.3±1.7 0.36±0.14 Tetulomab Rael 263000±27000 12.7±5.5 0.07±0.01 Biodistribution of 177Lu-tetulomab in severe combined immune Tetulomab Daudi 340000±5000 2.7±0.3 0.72±0.07 deficient (SCID) mice with intravenously injected Daudi cells. Rituximab Raji 272000±69000 4.8±0.9 0.08±0.006 177 Biodistribution of Lu-tetulomab was determined in female SCID Rituximab Rael 626000±36000 12.0±2.0 0.08±0.007 mice (NOD.CB17/Prkdc scid JHsd; Harlan Laboratories, An Venray, Rituximab Daudi 307000±12000 5.8±1.5 0.14±0.01 The Netherlands) with intravenously-injected Daudi tumor cells. The preparation was administered by tail vein injection of 260-1400 aTetulomab and rituximab bind to CD37 and CD20, respectively. kBq in 100 μl solution in each animal, one week after an injection of 10×106 Daudi cells. Four to five animals were used per time point. Autopsies were performed after cervical dislocation at various time points after injection. The weight of each tissue sample (blood, with the following antibodies (Dako, Glostrup, Denmark): anti-CD3 lung, heart, liver, spleen, kidneys, stomach, small intestine, large (UCHT-1), anti-IgGl [polyclonal rabbit anti-human lambda intestine, femur, muscle, brain, skull, lymph node and neck) was chains, rabbit F(ab’)2], anti-IgGk [polyclonal rabbit anti-human determined, and 177Lu was measured by a calibrated gamma kappa light chains, rabbit F(ab’)2], and anti-CD37 (in-house detector (Cobra II). Samples of the injectate were used as references tetulomab antibody). The expression of the different antigens was in the measurement procedures. All procedures and experiments measured using light microscopy after horseradish peroxidase involving animals in this study were approved by the National visualization (Dako). Animal Research Authority and carried out according to the European Convention for the Protection of Vertebrates used for Results Scientific Purposes. Binding properties of tetulomab and rituximab. The Therapeutic and toxic effect of 177Lu-tetulomab in SCID mice association rate constant, ka, the equilibrium dissociation intravenously injected with Daudi cells. Female, 4- to 8-week-old, constant, K , and the mean number of binding sites, B , SCID mice with body weights in the range of 16- 23 g at the start of d max the experiment, were used. The animals were maintained under were determined for three different cell lines, Raji, Rael and pathogen-free conditions, and food and water were supplied ad Daudi, and for the tetulomab and rituximab antibodies (Table libitum. All mice were ear-tagged and followed individually I). For Raji and Rael cells, the number of CD37 antigens was throughout the study. Mice were injected intravenously with 10×106 approximately half the number of CD20 antigens, while for Daudi cells in 0.1 ml PBS without Ca2+ and Mg2+ (PAA, Paasching, Daudi cells the number of antigens was similar. The Kd was Austria) one week before treatment with NaCl (N=23), 50 μg higher for Rael than for Raji and Daudi cells for both tetulomab (N=14), 50 (N=10), 100 (N=10) and 200 (N=10) MBq/kg antibodies. The K was similar for the two antibodies for 177Lu-tetulomab. A pilot study was performed to validate 100 % d tumor formation following injection of 10×106 Daudi cells per Raji and Rael cells, while for Daudi cells it was significantly mouse. Mice were killed by cervical dislocation, if suffering from lower for tetulomab than for rituximab (t-test, p<0.05), hind leg paralysis (primary end point), body weight decreased by indicating that tetulomab had better affinity for the CD37 20% from baseline, or if they otherwise showed symptoms of illness antigen than rituximab had for the CD20 antigen, in this cell and discomfort. The mice were dissected and histological sections line. For Rael cells, ka was similar for tetulomab and were stained with antibody against CD20 (EP459Y, Novus rituximab, while it was significantly higher for tetulomab Biologicals, Littleton, CO, USA) to locate tumor cells. The mice than for rituximab for Raji and Daudi cells, indicating faster were checked for hind leg paralysis every day and weighed two to three times per week. The different treatment groups were compared binding of tetulomab to CD37 than of rituximab to CD20 in by Kaplan-Meier survival analysis using SPSS version 13.0 (SPSS, these two cell lines. The Daudi cells were chosen for further Chicago, IL, USA). study since they had the highest amount of CD37 antigens: an average of 340,000 CD37 antigens and 307,000 CD20 Toxicity measurements. Prior to the study start and every two weeks antigens per cell. thereafter for up to 12 weeks, 50-75 μl blood was collected from the vena saphena lateralis in 0.5 ml EDTA-coated tubes (BD Internalization of 177Lu-tetulomab and 177Lu-rituximab. Microtainer K2E tubes; Becton, Dickinson and Company, Franklin Tetulomab was internalized faster and to a greater extent than Lakes, NJ, USA). White blood cells, red blood cells and platelets were counted using an automated hematology analyzer (Scil Vet abc rituximab (Figure 1). For the first half hour of incubation the animal blood counter, ABX Diagnostics, Montpellier, France). internalization speed was 0.19 fg/min/cell for tetulomab and 0.02 fg/min/cell for rituximab. There was no significant 125 177 Patient samples. A total of 217 B-cell lymphoma biopsies from difference between experiments performed with I and Lu. patients treated at the Norwegian Radium Hospital were stained Therefore, the results were pooled. In another experiment,

87 ANTICANCER RESEARCH 33: 85-96 (2013)

Table II. Number of 177Lu atoms bound per Daudi cell after two and 18 h of incubation.

Incubation time 2 Hours 18 Hours Antibody dosage (μg/ml) 177Lu-tetulomab 177Lu-rituximab 177Lu-tetulomab 177Lu-rituximab

Unblocked Blocked Unblocked Blocked Unblocked Blocked Unblocked Blocked

0a 12 7 8 53 12 5 10 53 1 8318 449 8554 372 10327 301 12831 356 2.5 9105 720 11629 885 11757 787 18836 1385 5 10025 1837 13658 2019 12123 1857 24097 1871 10 13646 3521 17344 2769 11548 3205 24249 2860 20 16290 8473 30095 9709 15233 5445 26639 5824 aThe different counts for the control samples is indicative of the variation in background radiation of the counter.

however, of longer duration, Daudi cells were incubated with 125I- and 111In-tetulomab for 1 h, washed and cell-bound activity was measured after four days of incubation (Figure 1, insert). The amount of antibody bound to cells was higher for 111In-tetulomab than for 125I-tetulomab.

Cell-bound activity of 177Lu-tetulomab and 177Lu-rituximab. Daudi cells were incubated with increasing concentrations of 177Lu-tetulomab or 177Lu-rituximab for two hours and 18 h before washing. Cell-bound activity was measured and the cells were seeded for further growth. The cell bound activity was lower for cells incubated with tetulomab than for cells incubated with rituximab after two-hour incubation and 18 h incubation (Table II). However, radiolabeled tetulomab saturated the antigen quicker and at lower antibody concentration than did rituximab. The specific activity was 92 MBq/mg for 177Lu-tetulomab and 137 MBq/mg for 177Lu-rituximab, which explains the difference in the number of 177Lu atoms attached to the cells for the two antibodies. Non-specific binding (blocked) was similar for the two RICs. Figure 1. Internalization of tetulomab and rituximab in Daudi cells. The At the 1 μg/ml dosage, there were almost no differences in insert figure shows cell-bound tetulomab 96 h after washing for incubation of Daudi cells with 1 μg/ml 111In-tetulomab or 125I- the number of specifically bound radioactive atoms for both tetulomab. N=3. Error bars=standard deviation. incubation times.

Cell growth experiments. Growth of Daudi cells incubated with RICs for two hours before washing, was followed for antibody, the difference between tetulomab and rituximab 14 days (Figure 2 A and B). There was no effect of unlabeled was even larger, since there was no re-growth of the cells antibodies alone on cell growth (data not shown). The treated with 177Lu- tetulomab. blocked cells treated with the 177Lu-antibody clearly did not Growth of Daudi cells incubated with RICs for 18 h grow as fast as the untreated control cells, indicating that before washing was also followed for 14 days (Figure 2C there was an effect of unbound 177Lu-antibody or a non- and D). There was no effect of unlabeled antibody alone on specific bound 177Lu-antibody on the cells. Treatment of cell growth (data not shown). The inhibition of cell growth of unblocked cells with 177Lu-antibody resulted in an increase the blocked cells treated with 177Lu-antibody was more in growth delay of 44% for cells treated with 10 μg/ml of pronounced than when the cells were incubated for two 177Lu- tetulomab (Figure 2A) and of 31% for cells treated hours before washing, probably because of the 16-h with 10 μg/ml 177Lu-rituximab (Figure 2B), as compared increased incubation time with RIC in the medium. with blocked cells. For treatment with 20 μg/ml of 177Lu- Treatment of unblocked cells with 177Lu-antibody resulted

88 Dahle et al: New Anti-CD37 Radioimmunoconjugate for NHL

Figure 2. Growth of Daudi cells incubated with 177Lu-tetulomab (A, C) or 177Lu-rituximab (B, D) for 2 h (A, B) or 18 h (C, D) before washing. Open symbols: unblocked cells; closed symbols: blocked cells.

in a growth delay of 107% for cells treated with 2.5 μg/ml SCID mice intravenously injected with Daudi cells are 177Lu-tetulomab (Figure 2C) and of 52% for cells treated presented in Figure 3. There was no apparent re-distribution with 2.5 μg/ml 177Lu-rituximab (Figure 2D), even though of nuclide from or to any organs after the initial uptake of cells labeled with 177Lu-rituximab had 1.6-times as much RICs, which indicates the in vivo stability of 177Lu-tetulomab cell-bound activity than did the cells labeled with 177Lu- since free 177Lu tends to relocate to the bone (29). However, tetulomab (Table II). The growth delay factor was 1.4 for the uptake and retention of 177Lu-tetulomab in blood, liver, cells incubated for two hours with 10 μg/ml 177Lu- spleen and kidney was significantly higher than in nude mice tetulomab, as compared with the same concentration of (data not shown). There was no significant difference 177Lu-rituximab and greater than 1.6 for cells treated with between the biodistribution in SCID mice with and without 20 μg/ml of the RICs. For 18-h incubation, the growth delay intravenously injected Daudi cells (data not shown). The factor was 1.6 for cells incubated with 1 μg/ml of the RICs. biodistribution of 177Lu-rituximab was also investigated and the uptake in the spleen was extremely high (data not Biodistribution of 177Lu-tetulomab in SCID mice. Uptake and shown). Although the biodistribution was not ideal, we retention data of 177Lu-tetulomab in normal tissues of female performed therapy experiments with 177Lu-tetulomab.

89 ANTICANCER RESEARCH 33: 85-96 (2013)

Table III. Distribution of Daudi cells in organs of mice with hind leg paralysis.

Organ No. of mice with Daudi cells in organ (%)

Lymph nodes 8 of 28 (28.6) Kidney 14 of 31 (45.2) Fat tissue 12 of 30 (40) Ovaries 17 of 30 (56.7) Liver 1 of 30 (3.2) Spleen 1 of 30 (3.2) Lungs 22 of 31 (71) Spine 30 of 31 (96)

Table IV. Lymphoma biopsies tested for CD37 expression.

Diagnosis No. of CD37-positive samples (%)

B-Lymphoblastic lymphoma 1 100.0 Burkitt’s lymphoma 4 100.0 Figure 3. Biodistribution (% injected dose/gram (ID/g)) of 177Lu- Diffuse large B-cell lymphoma (DLBCL) 25 100.0 tetulomab. The mice were injected with Daudi cells one week before DLBCL transformed from low-grade 19 100.0 injection of RIC. N=5. Data are mean±standard deviation. 92 100.0 Low-grade, unspecified 2 100.0 Marginal zone lymphoma 9 100.0 28 100.0 Small lymphocytic lymphoma 37 97.3

177 Total 217 99.5 groups, except for the 100-MBq/kg Lu-tetulomab-treated group, where the two mice that died from radiation toxicity had significantly decreased body weight. There were no deaths due to the 50-MBq/kg 177Lu-tetulomab treatment. Treated mice died from radiation toxicity before any mice died in the control group, where the first mouse died at 21 Toxic and therapeutic effects of 177Lu-tetulomab in SCID days. Treatments did not have a significant effect on red mice with intravenously injected Daudi cells. SCID mice blood cell counts (Figure 6C). were injected intravenously with 10×106 Daudi cells one The 50 and 100 MBq/kg 177Lu-tetulomab treatments week before administration of 50, 100 or 200 MBq/kg 177Lu- resulted in significantly improved survival as compared with tetulomab, 50 μg tetulomab, or NaCl. The Daudi cells NaCl, with respect to the hind leg paralysis end-point migrated to lymph nodes, kidney, fat tissue, ovaries, lungs (p<0.01, Mantel-Cox log-rank test) (Figure 5B). Treatment and to the spine (Table III and Figure 4). The mice were with 50 and 100 MBq/kg 177Lu-tetuomab was also monitored for hind leg paralysis, radiation toxicity and significantly more effective than treatment with tetulomab bodyweight every other day and blood counts every other (p<0.05), while the tetulomab group was not significantly week (Figures 5 and 6). different from the NaCl group (p>0.05). In Figure 5A, data are analyzed with deaths due to radiotoxicity as end-point. Radiotoxicity was defined as a white blood cell count (WBC) lower than 1.5×109 cells/l, no Expression of the tetulomab CD37 epitope in B-cell hind leg paralysis and decreased body weight. Treatment lymphoma biopsies. Expression of CD37 by binding of with 200 MBq/kg of 177Lu-tetulomab was too toxic for the tetulomab monoclonal antibody was tested in 217 lymphoma mice and they died two to three weeks after injection (Figure biopsies from different subtypes of NHL. All samples were 5A). The WBC decreased below 1×109 cells/l for these mice also stained for CD3, κ and λ light chain expression. In 216 (Figure 6A) and the platelets decreased below 300×109 out of the 217 samples, more than 50% of the κ- and/or cells/l (Figure 6B). The bodyweight of these mice also λ-positive cells (B-cells) expressed CD37 (Table IV). decreased significantly. There were no apparent treatment- Tetulomab did not show binding to the CD3-positive induced changes in bodyweight in the other treatment population (T-cells).

90 Dahle et al: New Anti-CD37 Radioimmunoconjugate for NHL

Figure 4. Microscopy images of sections of organs of SCID mice i.v. injected with Daudi cells. The sections were stained with an antibody against CD20 in order to locate the Daudi cells (Left panels). A: Spinal cord. B: Kidney. C: Ovary. Magnification=10 times. Right panels show corresponding sections stained with hematoxylin and eosin.

Discussion rituximab, due to antigenic drift and possible blockage of the CD20 antigen. Therefore, RIT targeting other antigens could Treatment of patients with lymphoma with current CD20- have some advantages. We have shown that there was a directed RIT is challenging in those previously treated with significantly enhanced inhibition of cell growth after

91 ANTICANCER RESEARCH 33: 85-96 (2013)

treatment with CD37-directed 177Lu-tetulomab as compared with CD20-directed 177Lu-rituximab. The growth delay factor of 177Lu-tetulomab vs. 177Lu-rituximab was 1.6. The uptake of 177Lu-tetulomab in lymphoma cells was lower or similar to the uptake of 177Lu-rituximab directly after labeling. Furthermore, 177Lu-tetulomab was significantly more effective than unlabeled tetulomab in treating SCID mice intravenously injected with Daudi cells. The Kd of tetulomab was similar to that for rituximab. Only one out of 217 clinical biopsies from nine different types of NHL did not express the CD37 antigen epitope targeted by tetulomab. RIT with CD37 as the target has previously been explored using a 131I-labeled murine monoclonal antibody (MB-1), both in a mouse model and in humans (30-35). CD37 antibodies were compared with CD20 antibodies and a higher grade of internalization and de-halogenation of 131I- labeled RIC was found for CD37 than for CD20 (35). Despite clinical responses observed in that study, CD20 was chosen for further development. No subsequent efforts have been made to target CD37 with RICs. In the early studies of CD37 RIT described above, the chloramine-T method of 131I-labeling, was used (35). 131I labeled to antibodies with the iodogen or the chloramine-T method will not be well-retained in the cells if the antigen- antibody complex is internalized (16, 17). The same pattern of de-halogenation has been shown with CD22 antibodies, which also are internalized (36). However, metallic radionuclides labeled to antibodies with chelators were better retained intracellularly when internalized (37). There are several metallic nuclides that can be used for RIT against CD37. Although we demonstrated results in a mouse model with the alpha emitting nuclide 227Th (38), a β-emitting nuclide may be more suitable as therapy for bulky lymphoma. Clinical data indicate that NHL is responsive to low Linear Energy Transfer (LET) β-emitters (1-4), thus in the current work, 177Lu was chosen because of its availability, suitable radiochemistry and half-life, and promising radiation properties. The inhibition of cell growth after treatment with CD37- directed 177Lu-tetulomab was significantly better than that for CD20-directed 177Lu-rituximab, when compared at the same antibody concentration. The differences in cell growth inhibition were higher for 18-h than for two-hour incubation with the RICs and thus might, to some extent, be explained Figure 5. Survival analysis of SCID mice i.v. injected with Daudi cells by the higher internalization of tetulomab than of rituximab. and then treated with 50, 100 or 200 MBq/kg 177Lu-tetulomab, 50 μg The internalization of tetulomab and thus retention of 177Lu tetulomab, or NaCl. A: Survival with radiation toxicity as end-point, i.e. white blood cell count <1.5×109 cells/l and no hind leg paralysis. B: will probably result in a higher absorbed radiation dose to Survival with hind leg paralysis as end point. In panel A no mice in the the cells treated with tetulomab than with rituximab, even NaCl–, 50 μg tetulomab- or the 50 MBq/kg 177Lu-tetulomab-treated though the initial binding was similar or higher for groups died of radiation toxicity, so these lines overlap. The “cold” rituximab. The 1.5- to 2-fold higher binding for rituximab tetulomab-treated group was omitted from this panel. The crosses in than for tetulomab can be explained by the 1.5-fold higher panel A represent animals that were censored because of hind leg paralysis. Similarly, the crosses in panel B represent animals that were specific activity for rituximab than for tetulomab in this censored because they died of causes other than hind leg paralysis. experiment. The growth inhibition induced by 177Lu-

92 Dahle et al: New Anti-CD37 Radioimmunoconjugate for NHL

Figure 6. Hematology of SCID mice i.v. injected with Daudi cells and then treated with 50, 100 or 200 MBq/kg 177Lu-tetulomab, 50 μg tetulomab or NaCl. A: White blood cell counts (WBC). B: Platelet counts. C: Red blood cell counts (RBC).

tetulomab and 177Lu-rituximab was related to selective One would expect, however, that it would be possible to targeting of the CD37 and CD20 antigens, respectively, since attain a similar effect on Raji cells, as on Daudi by blocking with “cold” tetulomab and rituximab significantly increasing the concentration of the antibody by the same reduced the inhibition of cell growth. factor as the difference in the number of antigens they carry. The cell growth experiments were carried out with Daudi The unusually high retention of 177Lu-tetulomab in blood, cells, which had the highest expression of the CD37 antigen, liver, spleen and kidneys and the extremely high uptake of more than two-fold that for Raji cells. Therefore, one might 177Lu-rituximab in the spleen in SCID mice cannot be wonder if the treatment would also be effective for Raji cells. explained by binding to Daudi cells located in the respective

93 ANTICANCER RESEARCH 33: 85-96 (2013) organs, since similar biodistribution without Daudi cells 2 Goldenberg DM, Horowitz JA, Sharkey RM, Hall TC, Murthy being injected was found (results not shown). The reason for S, Goldenberg H, Lee RE, Stein R, Siegel JA and Izon DO: the unusual biodistribution might be the low production of Targeting, dosimetry, and radioimmunotherapy of B-cell with iodine-131-labeled LL2 monoclonal antibody. J endogeneous antibodies in SCID mice. This effect on Clin Oncol 9: 548-564, 1991. biodistribution has been previously described (39). 3 Kaminski MS, Zasadny KR, Francis IR, Milik AW, Ross CW, We report on a similar migration pattern of Daudi cells as Moon SD, Crawford SM, Burgess JM, Petry NA and Butchko the one presented by Gethie et al. in (22). The relatively high GM: Radioimmunotherapy of B-cell lymphoma with [131I]anti- toxicity of 177Lu-tetulomab in this model was probably B1 (anti-CD20) antibody. N Engl J Med 329: 459-465, 1993. related to both the unusual biodistribution and the high 4 Witzig TE, Gordon LI, Cabanillas F, Czuczman MS, radiosensitvity of these DNA double-strand repair-defective Emmanouilides C, Joyce R, Pohlman BL, Bartlett NL, Wiseman mice (due to the SCID mutation). Nevertheless, the GA, Padre N, Grillo-Lopez AJ, Multani P and White CA: 177 Randomized controlled trial of yttrium-90-labeled ibritumomab therapeutic effect of Lu-tetulomab was significantly better tiuxetan radioimmunotherapy versus rituximab immunotherapy than that for the unlabeled antibody. for patients with relapsed or refractory low-grade, follicular, or 125 The binding data for I-labeled tetulomab and rituximab transformed B-cell non-Hodgkin’s lymphoma. J Clin Oncol 20: showed that the antigen-binding properties of tetulomab and 2453-2463, 2002. rituximab were similar, with a Kd between 2.7 and 12.7 for 5 Horning SJ, Younes A, Jain V, Kroll S, Lucas J, Podoloff D and tetulomab and between 4.8 and 12 for rituximab, depending Goris M: Efficacy and safety of and iodine-131 on the cell line used. The reason for the variability in K tositumomab (Bexxar) in B-cell lymphoma, progressive after d rituximab. J Clin Oncol 23: 712-719, 2005. between cell lines might be that the three parameters 6 Witzig TE, Flinn IW, Gordon LI, Emmanouilides C, Czuczman estimated by the curve-fitting method, to some extent may MS, Saleh MN, Cripe L, Wiseman G, Olejnik T, Multani PS and influence each other, or there could be differences in the White CA: Treatment with radio- antigen from cell line to cell line due to mutations or post- immunotherapy in patients with rituximab-refractory follicular non- translational changes. The data for the number of antigens Hodgkin’s lymphoma. J Clin Oncol 20: 3262-3269, 2002. on the Daudi cells fitted well with the measured binding after 7 Vose JM, Wahl RL, Saleh M, Rohatiner AZ, Knox SJ, Radford two hours in the growth inhibition experiment if the JA, Zelenetz AD, Tidmarsh GF, Stagg RJ and Kaminski MS: Multicenter phase II study of iodine-131 tositumomab for difference in specific activity between 177Lu-tetulomab and 177 chemotherapy-relapsed/refractory low-grade and transformed Lu-rituximab is taken into consideration. low-grade B-cell non-Hodgkin’s lymphomas. J Clin Oncol 18: In conclusion, the data presented here indicate that the 1316-1323, 2000. tetulomab antibody is well-suited for RIT of CD37- 8 Maloney DG, Grillo-Lopez AJ, White CA, Bodkin D, Schilder RJ, expressing lymphoma cells. The study of binding of Neidhart JA, Janakiraman N, Foon KA, Liles TM, Dallaire BK, tetulomab to patient biopsies indicates that targeting of the Wey K, Royston I, Davis T and Levy R: IDEC-C2B8 (Rituximab) CD37 epitope by tetulomab could be clinically relevant and anti-CD20 monoclonal antibody therapy in patients with relapsed 90 warrants for future pre-clinical and clinical investigations. low-grade non-Hodgkin’s lymphoma. Blood : 2188-2195, 1997. 9 Reff ME, Carner K, Chambers KS, Chinn PC, Leonard JE, Raab R, Newman RA, Hanna N and Anderson DR: Depletion of B Declaration of Interest cells in vivo by a chimeric mouse human monoclonal antibody to CD20. Blood 83: 435-445, 1994. JD and AHVRL are employed by and have stock options in Nordic 10 Gopal AK, Press OW, Wilbur SM, Maloney DG and Pagel JM: Nanovector AS. RHL owns Nordic Nanovector shares. Rituximab blocks binding of radiolabeled anti-CD20 antibodies (Ab) but not radiolabeled anti-CD45 Ab. Blood 112: 830-835, Acknowledgements 2008. 11 Foran JM, Norton AJ, Micallef IN, Taussig DC, Amess JA, The Authors are grateful to Steinar Funderud and Erlend Smeland Rohatiner AZ and Lister TA: Loss of CD20 expression following who developed the tetulomab monoclonal antibody, to Bjørn treatment with rituximab (chimaeric monoclonal anti-CD20): a Erikstein, Erlend Smeland and Harald Holte for collecting the retrospective cohort analysis. Br J Haematol 114: 881-883, 2001. lymphoma biopsies, and to the Histology Research Laboratory at 12 Sugimoto T, Tomita A, Hiraga J, Shimada K, Kiyoi H, Kinoshita the Norwegian Radium Hospital for preparing the lymphoma slides. T and Naoe T: Escape mechanisms from antibody therapy to This study was supported by a grant from Innovation Norway. lymphoma cells: downregulation of CD20 mRNA by recruitment of the HDAC complex and not by DNA methylation. Biochem References Biophys Res Commun 390: 48-53, 2009. 13 Wada N, Kohara M, Ogawa H, Sugiyama H, Fukuhara S, 1 DeNardo GL, DeNardo SJ, Goldstein DS, Kroger LA, Lamborn Tatsumi Y, Kanamaru A, Hino M, Kanakura Y, Morii E and KR, Levy NB, McGahan JP, Salako Q, Shen S and Lewis JP: Aozasa K: Change of CD20 Expression in Diffuse Large B-Cell Maximum-tolerated dose, toxicity, and efficacy of (131)I-Lym-1 Lymphoma Treated with Rituximab, an Anti-CD20 Monoclonal antibody for fractionated radioimmunotherapy of non-Hodgkin’s Antibody: A Study of the Osaka Lymphoma Study Group. Case lymphoma. J Clin Oncol 16: 3246-3256, 1998. Rep Oncol 2: 194-202, 2009.

94 Dahle et al: New Anti-CD37 Radioimmunoconjugate for NHL

14 Classon BJ, Williams AF, Willis AC, Seed B and Stamenkovic 28 Dahle J, Krogh C, Melhus KB, Borrebaek J, Larsen RH and I: The primary structure of the human leukocyte antigen CD37, Kvinnsland Y: In vitro cytotoxicity of low-dose-rate a species homologue of the rat MRC OX-44 antigen. J Exp Med radioimmunotherapy by the alpha-emitting radioimmuno- 169: 1497-1502, 1989. conjugate Thorium-227-DOTA-rituximab. Int J Radiat Oncol 15 Schwartz-Albiez R, Dorken B, Hofmann W and Moldenhauer G: Biol Phys 75: 886-895, 2009. The -associated CD37 antigen (gp40-52). Structure and 29 Muller WA, Linzner U and Schaffer EH: Organ distribution studies subcellular expression of an extensively glycosylated of lutetium-177 in mouse. Int J Nucl Med Biol 5: 29-31, 1978. glycoprotein. J Immunol 140: 905-914, 1988. 30 Brown RS, Kaminski MS, Fisher SJ, Chang AE and Wahl RL: 16 Press OW, Farr AG, Borroz KI, Anderson SK and Martin PJ: Intratumoral microdistribution of [131I]MB-1 in patients with B- Endocytosis and degradation of monoclonal antibodies targeting cell lymphoma following radioimmunotherapy. Nucl Med Biol human B-cell malignancies. Cancer Res 49: 4906-4912, 1989. 24: 657-663, 1997. 17 Press OW, Howell-Clark J, Anderson S and Bernstein I: 31 Buchsbaum DJ, Wahl RL, Normolle DP and Kaminski MS: Retention of B-cell-specific monoclonal antibodies by human Therapy with unlabeled and 131I-labeled pan-B-cell monoclonal lymphoma cells. Blood 83: 1390-1397, 1994. antibodies in nude mice bearing Raji Burkitt’s lymphoma 18 Moldenhauer G: CD37. J Biol Regul Homeost Agents 14: 281- xenografts. Cancer Res 52: 6476-6481, 1992. 283, 2000. 32 Eary JF, Press OW, Badger CC, Durack LD, Richter KY, 19 Heider KH, Kiefer K, Zenz T, Volden M, Stilgenbauer S, Addison SJ, Krohn KA, Fisher DR, Porter BA and Williams DL: Ostermann E, Baum A, Lamche H, Kupcu Z, Jacobi A, Muller S, Imaging and treatment of B-cell lymphoma. J Nucl Med 31: Hirt U, Adolf GR and Borges E: A novel Fc-engineered 1257-1268, 1990. monoclonal antibody to CD37 with enhanced ADCC and high 33 Kaminski MS, Fig LM, Zasadny KR, Koral KF, DelRosario RB, proapoptotic activity for treatment of B-cell malignancies. Blood Francis IR, Hanson CA, Normolle DP, Mudgett E and Liu CP: 118: 4159-4168, 2011. Imaging, dosimetry, and radioimmunotherapy with iodine 131- 20 Zhao X, Lapalombella R, Joshi T, Cheney C, Gowda A, Hayden- labeled anti-CD37 antibody in B-cell lymphoma. J Clin Oncol Ledbetter MS, Baum PR, Lin TS, Jarjoura D, Lehman A, 10: 1696-1711, 1992. Kussewitt D, Lee RJ, Caligiuri MA, Tridandapani S, Muthusamy 34 Press OW, Eary JF, Badger CC, Martin PJ, Appelbaum FR, Levy N and Byrd JC: Targeting CD37-positive lymphoid malignancies R, Miller R, Brown S, Nelp WB and Krohn KA: Treatment of with a novel engineered small modular immunopharmaceutical. refractory non-Hodgkin’s lymphoma with radiolabeled MB-1 Blood 110: 2569-2577, 2007. (anti-CD37) antibody. J Clin Oncol 7: 1027-1038, 1989. 21 Smeland E, Funderud S, Ruud E, Kiil BH and Godal T: 35 Press OW, Eary JF, Appelbaum FR, Martin PJ, Badger CC, Nelp Characterization of two murine monoclonal antibodies reactive WB, Glenn S, Butchko G, Fisher D and Porter B: Radiolabeled- with human B cells. Their use in a high-yield, high-purity antibody therapy of B-cell lymphoma with autologous bone method for isolation of B cells and utilization of such cells in an marrow support. N Engl J Med 329: 1219-1224, 1993. assay for B-cell stimulating factor. Scand J Immunol 21: 205- 36 Sharkey RM, Behr TM, Mattes MJ, Stein R, Griffiths GL, Shih 214, 1985. LB, Hansen HJ, Blumenthal RD, Dunn RM, Juweid ME and 22 Ghetie MA, Richardson J, Tucker T, Jones D, Uhr JW and Goldenberg DM: Advantage of residualizing radiolabels for an Vitetta ES: Disseminated or localized growth of a human B-cell internalizing antibody against the B-cell lymphoma antigen, tumor (Daudi) in SCID mice. Int J Cancer 45: 481-485, 1990. CD22. Cancer Immunol Immunother 44: 179-188, 1997. 23 Leblond V, Autran B and Cesbron JY: The SCID mouse mutant: 37 Lub-de Hooge MN, Kosterink JG, Perik PJ, Nijnuis H, Tran L, definition and potential use as a model for immune and Bart J, Suurmeijer AJ, de JS, Jager PL and de Vries EG: hematological disorders. Hematol Cell Ther 39: 213-221, 1997. Preclinical characterisation of 111In-DTPA-. Br J 24 Mitchell P, Lee FT, Hall C, Rigopoulos A, Smyth FE, Hekman Pharmacol 143: 99-106, 2004. AM, van Schijndel GM, Powles R, Brechbiel MW and Scott 38 Dahle J, Borrebaek J, Jonasdottir TJ, Hjelmerud AK, Melhus AM: Targeting primary human Ph(+) B-cell precursor leukemia- KB, Bruland OS, Press OW and Larsen RH: Targeted cancer engrafted SCID mice using radiolabeled anti-CD19 monoclonal therapy with a novel low-dose rate alpha-emitting radioimmuno- antibodies. J Nucl Med 44: 1105-1112, 2003. conjugate. Blood 110: 2049-2056, 2007. 25 Fulop GM and Phillips RA: The scid mutation in mice causes a 39 Mitchell P, Lee FT, Hall C, Rigopoulos A, Smyth FE, Hekman general defect in DNA repair. Nature 347: 479-482, 4-10-1990. AM, van Schijndel GM, Powles R, Brechbiel MW and Scott 26 Lindmo T and Bunn PA Jr.: Determination of the true AM: Targeting primary human Ph(+) B-cell precursor leukemia- immunoreactive fraction of monoclonal antibodies after engrafted SCID mice using radiolabeled anti-CD19 monoclonal radiolabeling. Methods Enzymol 121: 678-691, 1986. antibodies. J Nucl Med 44: 1105-1112, 2003. 27 Dahle J, Krogh C, Melhus KB, Kaalhus O, Larsen RH and Stokke T: A one-step method for determining the maximum number of bound antibodies, and the affinity and association rate Received October 18 2012 constants for antibody binding. Nucl Med Commun 28: 742-747, Revised November 7, 2012 2007. Accepted November 8, 2012

95