Radioimmunotherapy of PANC-1 human pancreatic cancer xenografts in NOD/SCID or NRG mice with Panitumumab labeled with Auger electron emitting, 111In or β-particle emitting, 177Lu

Background Epidermal growth factor receptors (EGFR) are overexpressed on > 90% of pancreatic cancers (PnCa) and represent an attractive target for the development of novel therapies, including radioimmunotherapy (RIT). Our aim was to study RIT of subcutaneous (s.c.) PANC-1 human PnCa xenografts in mice using the anti-EGFR monoclonal antibody, panitumumab labeled with Auger electron (AE)-emitting, 111In or β-particle emitting, 177Lu at amounts that were non-toxic to normal tissues. Results Panitumumab was conjugated to DOTA chelators for complexing 111In or 177Lu (panitumumab-DOTA-[111In]In and panitumumab-DOTA-[177Lu]Lu) or to a metal-chelating polymer (MCP) with multiple DOTA to bind 111In (panitumumab-MCP-[111In]In). Panitumumab-DOTA-[177Lu]Lu was more effective per MBq exposure at reducing the clonogenic survival in vitro of PANC-1 cells than panitumumab-DOTA-[111In]In or panitumumab-MCP-[111In]In. Panitumumab-DOTA-[177Lu]Lu caused the greatest density of DNA double-strand breaks (DSBs) in the nucleus measured by immunofluorescence for γ-H2AX. The absorbed dose in the nucleus was 3.9-fold higher for panitumumab-DOTA-[177Lu]Lu than panitumumab-DOTA-[111In]In and 7.7-fold greater than panitumumab-MCP-[111In]In. No normal tissue toxicity was observed in NOD/SCID mice injected intravenously (i.v.) with 10.0 MBq (10 μg; ~ 0.07 nmoles) of panitumumab-DOTA-[111In]In or panitumumab-MCP-[111In]In or in NRG mice injected i.v. with 6.0 MBq (10 μg; ~ 0.07 nmoles) of panitumumab-DOTA-[177Lu]Lu. There was no decrease in complete blood cell counts (CBC) or increased serum alanine aminotransferase (ALT) or creatinine (Cr) or decreased body weight. RIT inhibited the growth of PANC-1 tumours but a 5-fold greater total amount of panitumumab-DOTA-[111In]In or panitumumab-MCP-[111In]In (30 MBq; 30 μg; ~ 0.21 nmoles) administered in three fractionated amounts every three weeks was required to achieve greater or equivalent tumour growth inhibition, respectively, compared to a single amount of panitumumab-DOTA-[177Lu]Lu (6 MBq; 10 μg; ~ 0.07 nmoles). The tumour doubling time (TDT) for NOD/SCID mice with s.c. PANC-1 tumours treated with panitumumab-DOTA-[111In]In or panitumumab-MCP-[111In]In was 51.8 days and 28.1 days, respectively. Panitumumab was ineffective yielding a TDT of 15.3 days vs. 15.6 days for normal saline treated mice. RIT of NRG mice with s.c. PANC-1 tumours with 6.0 MBq (10 μg; ~ 0.07 nmoles) of panitumumab-DOTA-[177Lu]Lu increased the TDT to 20.9 days vs. 11.5 days for panitumumab and 9.1 days for normal saline. The absorbed doses in PANC-1 tumours were 8.8 ± 3.0 Gy and 2.6 ± 0.3 Gy for panitumumab-DOTA-[111In]In and panitumumab-MCP-[111In]In, respectively, and 11.6 ± 4.9 Gy for panitumumab-DOTA-[177Lu]Lu. Conclusion RIT with panitumumab labeled with Auger electron-emitting, 111In or β-particle-emitting, 177Lu inhibited the growth of s.c. PANC-1 tumours in NOD/SCID or NRG mice, at administered amounts that caused no normal tissue toxicity. We conclude that EGFR-targeted RIT is a promising approach to treatment of PnCa. Supplementary Information Supplementary information accompanies this paper at 10.1186/s41181-020-00111-y.


Background
Pancreatic cancer (PnCa) is one of the most lethal cancers since most patients are diagnosed when there is local invasion or metastasis, rendering these patients ineligible for surgical treatment (Kamisawa et al., 2016). Surgery provides the only opportunity for long-term survival in PnCa (Kamisawa et al., 2016). The median survival for patients receiving surgery is 17-23 months while those with locally advanced PnCa or metastases have a median survival of 8-14 and 4-6 months, respectively (Cleary et al., 2004). Modest improvements in patient outcome have been achieved with the introduction of FOLFIRINOX, a potent chemotherapy regimen that combines oxaliplatin, irinotecan, leucovorin and 5-fluorouracil but this treatment is associated with high toxicity (Conroy et al., 2011). Thus, the outcome for patients with PnCa remains poor and new therapeutic approaches with improved effectiveness and lower toxicity are needed.
Epidermal growth factor receptor (EGFR) overexpression is found in > 90% of cases of PnCa and represents an attractive target for the development of novel therapies (Troiani et al., 2012). However, clinical trials of anti-EGFR monoclonal antibodies, cetuximab (Erbitux; Eli Lilly) (Crane et al., 2011) or panitumumab (Vectibix; Amgen) (Halfdanarson et al., 2019) combined with chemotherapy have not been as encouraging as hoped, possibly due to downstream KRAS mutation in PnCa which obviates the effects of blocking EGFR (Eser et al., 2014). However, this does not preclude EGFR-targeted radioimmunotherapy (RIT) of PnCa, since EGFR overexpression is used only to selectively deliver radiation to tumours. Indeed, we recently reported that panitumumab modified with a metal-chelating polymer (MCP) to complex the β-particle emitter, 177 Lu [Eβmax = 0.498 MeV (78.6%), 0.385 MeV (9.1%), 0.176 MeV (12.2%); t 1/2 = 6.7 days] inhibited the growth of subcutaneous (s.c.) PANC-1 human PnCa xenografts in NRG (NOD-Rag1 −/− IL2Rg null ) mice despite these tumours exhibiting KRAS mutation (Ma et al., 2013), while treatment with unlabeled panitumumab was ineffective . The βparticles emitted by 177 Lu have a maximum 2 mm range in tissues, which results in low linear energy transfer (LET < 0.3 keV/μm). Our group has also been exploring RIT of tumours exploiting the very low energy (< 20 keV) but high LET (LET = 4-26 keV/μm) Auger electrons (AEs) emitted by 111 In (Ku et al., 2019). These nanometer-micrometer range electrons cause lethal DNA double-strand breaks (DSB) in cancer cells, especially if emitted in close proximity to the cell nucleus. The potency of [ 111 In]In-labeled radioimmunoconjugates (RICs) is amplified by modification with nuclear translocation sequence (NLS) peptides which transport the RICs to the cell nucleus (Costantini et al., 2007). Interestingly, the EGFR harbours an endogenous NLS in the transmembrane domain that mediates nuclear importation of EGF (Lo et al., 2006) and cetuximab (Liao and Carpenter, 2009). Thus modification of RICs with an exogenous NLS may not be required for EGFRtargeted RIT. Moreover, recent studies have shown that a local bystander effect extends the therapeutic effects of AEs beyond the physical range of the electrons (Paillas et al., 2016). In addition, 111 In emits two abundant γ-photons [Eγ-171 keV (90%) and 245 keV (94%)] and 177 Lu emits a low abundance γ-photon [Eγ = 208 keV (11%)] which enable single photon emission computed tomography (SPECT), providing an opportunity to combine tumour imaging and RIT (i.e. "theranostic" concept). We previously proposed that panitumumab-MCP dual-labeled with 177 Lu and 111 In could represent a novel theranostic agent that combines SPECT and AE and β-particle mediated RIT .
In the current study, we hypothesized that panitumumab labeled with AEemitting, 111 In or β-particle emitting, 177 Lu would be effective for RIT of s.c. PANC-1 human PnCa xenografts in non-obese diabetic severe combined immunodeficiency (NOD/SCID) or NRG mice, respectively, at administered amounts that are non-toxic to normal tissues. Panitumumab was modified with DOTA (1,4,7,10tetraazacyclododecane-1,4,7,10-tetraacetic acid) or conjugated to a MCP harbouring multiple DOTA to complex 111 In or 177 Lu. We previously reported that conjugation of panitumumab to this same MCP enables high specific activity (SA) labeling with 111 In or 177 Lu (Aghevlian et al., 2018). RIT of s.c. PANC-1 tumours in NRG mice with panitumumab-MCP- 177 Lu was effective at administered amounts that were non-toxic to normal tissues .

Cell lines and tumour xenografts
EGFR positive PANC-1 human PnCa cells (4.0 × 10 5 EGFR/ cell) (Korc et al., 1986) were purchased from the American Type Culture Collection (ATCC) and cultured in Dulbecco's Modified Eagle's Medium (DMEM) with high glucose (Gibco) supplemented with 1% penicillin and streptomycin and 10% fetal bovine serum (Gibco-Invitrogen). Tumour xenografts (7-9 mm diameter) were established at 4 weeks post s.c. inoculation of 3-4 × 10 6 PANC-1 cells in DMEM (100 μL) into the left flank in female NOD/SCID mice (Charles River) or in NRG mice (University Health Network, Toronto, ON). All animal studies were conducted in compliance with the Canadian Council on Animal Care (CCAC) guidelines under a protocol (AUP 2843.3) approved by the institutional Animal Care Committee at the University Health Network.

Clonogenic survival assays and measurement of DNA DSBs
Clonogenic survival assays were performed by exposing 2 × 10 5 PANC-1 cells in 2.0 mL of growth medium in 24-well plates to 2.5 nmoles/L (0.3, 0.6 or 1.2 MBq) of panitumumab-DOTA-[ 177 Lu]Lu,In or panitumumab-MCP-[ 111 In]In for 16 h at 37°C in growth medium or to unlabeled panitumumab-DOTA or medium alone. Approximately 700 cells were then seeded into 6-well plates and cultured for 12 days. Surviving colonies were stained with methylene blue and colonies (> 50 cells) counted. The plating efficiency (PE) was determined by dividing the number of colonies formed by the number of cells seeded. The surviving fraction (SF) was calculated by dividing the PE for treated cells by that for untreated cells. Unrepaired DNA double-strand breaks (DSBs) were assessed in the nucleus of PANC-1 cells by immunofluorescence for phosphorylated gamma histone-2A (γ-H2AX) as previously described (Cai et al., 2008). The images were analyzed for the integrated density of γ-H2AX foci per nucleus area using ImageJ software (Cai et al., 2009).

Subcellular fractionation and cellular dosimetry
The subcellular distribution of activity on the cell membrane (CM), in the cytoplasm (Cy) or in the nucleus (N) of PANC-1 cells incubated with panitumumab-DOTA-[ 177 Lu]Lu were measured to estimate the absorbed doses in the nucleus for panitumumab-DOTA-[ 177 Lu]Lu and panitumumab-DOTA- [ 111 In]In. 111 In and 177 Lu are both strongly bound by DOTA [log K M = 25.4 and 23.9-24.5, respectively] (Baranyai et al., 2020) thus, it was assumed that the subcellular localization of panitumumab-DOTA- [ 111 In]In was predicted by that of panitumumab-DOTA-[ 177 Lu]Lu. Briefly, 2 × 10 5 cells cultured overnight in wells in a 24-well plate were incubated for 1, 4, 8, or 24 h at 37°C with panitumumab-DOTA-[ 177 Lu]Lu (1.2 MBq; 240 MBq/nmole; 2.5 nmoles/L). Incubation was performed in the absence or presence of excess panitumumab, or with non-specific IgG-DOTA-[ 177 Lu]Lu to assess EGFRmediated cellular uptake. The medium was removed and the cells rinsed with phosphate-buffered saline (PBS), pH 7.5. Activity on the CM was displaced with 1 mL of 200 mM sodium acetate/500 mM NaCl, pH 2.5 at 22°C. This procedure was repeated twice and the combined CM fractions measured in a γ-counter. The cells were then lysed with 2 × 500 μL of Nuclei EZ Lysis buffer (Sigma-Aldrich) on ice for 60 min. Lysed cells were transferred to 1.5 mL Eppendorf tubes and the tubes centrifuged at 3000×g for 5 min to separate the N and Cy activity (supernatant), which were measured in a γ-counter. The subcellular distribution of panitumumab-MCP-[ 177 Lu]Lu in PANC-1 cells was previously reported  and was used to predict the subcellular localization of panitumumab-MCP- [ 111 In]In .
The absorbed dose in the nucleus of a PANC-1 cell over the 16 h incubation period used for clonogenic survival assays with panitumumab-DOTA- 177 Lu, panitumumab-DOTA- 111 In or panitumumab-MCP- 111 In, were estimated as P (Gy) from activity in a subcellular or extracellular source compartment (CM, Cy, N or medium),Ã S is the time-integrated activity (Bq × sec) in the source compartment, and S is the Snyder factor (Goddu et al., 1994).Ã S 0 − 16h in each cell compartment or medium was calculated from the area under the curve from 0 to 16 h. Monte Carlo (MCNP Ver. 5.0; Los Alamos National Laboratory) was used to calculate the medium and monolayer S-values for 177 Lu and 111 In using the dimensions of a PANC-1 cell and N (radius = 8.5 μm and 7.0 μm; respectively) (Cai et al., 2017;Cai et al., 2010). The study volume was defined as a cylinder with a radius of 0.77 cm containing 1.05 cm thickness water and a 0.1 cm thick polystyrene bottom on which a monolayer of closely packed cells was attached. The detailed spectra of Auger electron, conversion electron, X-rays and γ-photons of 111 In and 177 Lu, as well as the β full energy spectrum of 177 Lu were taken from the MIRD Radionuclide Data (Eckerman and Endo, 2008) and used in the MCNP5 simulation. MCNP5 uses continuousenergy nuclear and atomic data libraries with the energy cutoff for both electrons and photons at 1 keV. 2 × 10 6 electrons or photons were launched for each calculation on the energy deposition per starting particle per tally volume to reach a statistical relative error lower than 0.05 of one standard deviation. We were not able to estimate the absorbed doses in the nucleus resulting from the 10 days in culture used for the clonogenic survival assays, since we did not fractionate the cells during this period, but estimation of the doses for the 16 h incubation allowed correlation of the effects of absorbed dose with SF after exposure to the RICs. In order to appreciate the relative biological effectiveness (RBE) of the AEs emitted by 111 In compared to β-particles emitted by 177 Lu, the SF was plotted vs. absorbed dose (Gy) for PANC-1 cells exposed to panitumumab-MCP- [ 111

Normal tissue toxicity
Normal tissue toxicity was assessed in non-tumour bearing NOD/SCID mice (n = 5) injected i.v. (tail vein) with 10.0 MBq (10 μg;~0.07 nmoles) of panitumumab-DOTA- [ 111 In]In or panitumumab-MCP- [ 111 In]In. This amount was selected because Ochakovskaya et al. (Ochakovskaya et al., 2001) found that 111 In-labeled anti-CD74 antibodies administered up to 12.9 MBq in SCID mice caused no toxicity. Since NOD/SCID mice, similar to SCID mice harbor a germ line mutation in DNA repair (Biedermann et al., 1991) and panitumumab-DOTA-[ 177 Lu]Lu may be more toxic than panitumumab-DOTA- [ 111 In]In, due to the cross-fire effect of the moderate energy β-particles (Reilly, 2005), we assessed the normal tissue toxicity of panitumumab-DOTA-[ 177 Lu]Lu at a lower activity (6.0 MBq; 10 μg;~0.07 nmoles) in non-tumour bearing NRG mice which are not deficient in DNA repair. Control NOD/SCID or NRG mice received normal saline or panitumumab (10 μg;~0.07 nmoles). Body weight was monitored every 2-4 days for 14 days. The 14 days period for assessment of toxicity was selected based on the requirements of Health Canada for acute toxicity testing of new therapeutic agents (Anonymous, 1995). At 14 days, the mice were sacrificed, and blood samples were collected for measurement of alanine aminotransferase (ALT) and creatinine (Cr). A complete blood cell (CBC) count and hemoglobin (Hb) were also obtained on a HemaVet 950FS instrument (Drew Scientific).

Radioimmunotherapy (RIT) studies
NOD/SCID mice with s.c. PANC-1 tumours were treated with 10 MBq (10 μg) of panitumumab-DOTA-[ 111 In]In (n = 12) or panitumumab-MCP-[ 111 In]In (n = 7) administered i.v. (tail vein) in 100 μL of normal saline every 3 weeks for 9 weeks (Scheme 1). NRG mice with s.c. PANC-1 tumours were injected i.v. with a single amount of 6 MBq (10 μg;~0.07 nmoles) of panitumumab-DOTA-[ 177 Lu]Lu in 100 μL of normal saline. Control NOD/SCID or NRG mice received panitumumab (10 μg; n = 10) or normal saline (n = 10). Tumour length (mm) and width (mm) were measured using calipers every 2-3 days until the humane endpoint was reached. The tumour volume (V = mm 3 ) was calculated as V = length × (width) 2 × 0.5 (Euhus et al., 1986). The tumour growth index (TGI) was calculated by dividing the tumour volume at each observation time point by the volume at the start of treatment. Similarly, the body weight index (BWI) was calculated by dividing the body weight at each time point by the pre-treatment body weight. The TGI vs. time curves were fitted to an exponential growth model by GraphPad Prism Ver. 4.0 and the tumour doubling time (TDT) was calculated.

Tumour and normal organ dosimetry
Biodistribution studies were performed in NOD/SCID or NRG mice with s.c. PANC-1 tumours injected i.v. (tail vein) with panitumumab-DOTA-[ 111 In]In (10 MBq; 10 μg) or panitumumab-DOTA-[ 177 Lu]Lu (6 MBq; 10 μg) to estimate the absorbed doses in the tumour and normal organs. Briefly, at selected times from 24 to 168 h p.i., groups of mice (n = 5) were sacrificed, and the tumour and normal organs collected, weighed and measured in a γcounter. The tumour and normal organ biodistribution of panitumumab-MCP-[ 177 (Bitar et al., 2007). The tumour dose was estimated using the sphere model in OLINDA/EXM radiation dosimetry software (Stabin et al., 2005).

Statistical analysis
Results were expressed as mean ± SD. Statistical significance was tested using an unpaired two-tailed Student's t-test or one-way ANOVA (P < 0.05).

Clonogenic survival assays and measurement of DNA DSBs
Clonogenic survival and DNA DSB assays were performed at constant RIC concentration of 2.5 nM and increasing apparent molar activity from 60 to 240 MBq/nmole at There was no effect of unlabeled panitumumab-MCP or panitumumab-DOTA on the SF of PANC-1 cells ( Fig. 1a; 0 111 In-labeled RICs. Cells exposed to growth medium or unlabeled panitumumab-MCP or panitumumab-DOTA exhibited few γ-H2AX foci, representing sites of unrepaired DNA DSBs (Fig. 2a) and the density of these foci were not significantly different than in untreated cells. Panitumumab-DOTA-[ 177 Lu]Lu caused significantly more γ-H2AX foci in the nucleus of PANC-1 cells than panitumumab-DOTA- 111 In at all amounts (Fig. 2b) and more γ-H2AX foci than panitumumab-MCP- 111 In at 0.6 MBq and 1.2 MBq, but not at 0.3 Significant differences (P < 0.05) are noted by asterisks. b Surviving fraction of PANC-1 cells exposed to RICs vs. absorbed dose (Gy) (Table 1). The integrated density of γ-H2AX foci in the nucleus of PANC-1 cells exposed to RICs vs. absorbed dose to the nucleus (Gy) was fitted by linear regression (Fig. 2c)

Subcellular fractionation and cellular dosimetry
The Integrated density of γ-H2AX foci representing unrepaired DNA DSBs in the nucleus of PANC-1 cells exposed to RICs. Values shown are the mean ± SD (n = 12). Significant differences (P < 0.05) are noted by the asterisks. c Integrated density of γ-H2AX foci in the nucleus of PANC-1 cells exposed to RICs vs. absorbed dose (Gy) in the nucleus. Lines were obtained by linear regression  Lu]Lu to NOD/SCID or NRG mice, respectively, caused no significant increase in the mean serum Cr or ALT compared to mice receiving normal saline (Fig. 3a,b). Administration of these amounts of RICs caused no change in CBC or Hb ( Fig. 3c-f). There were strain differences in normal Cr, ALT and CBC values between control NOD/SCID and NRG mice receiving normal saline.

Discussion
In this study, we evaluated the cytotoxicity in vitro of panitumumab labeled with AEemitting, 111 In or β-particle-emitting, 177 Lu on PANC-1 human PnCa cells and correlated the SF of these cells in clonogenic assays with unrepaired DNA DSBs in the nucleus assessed by immunofluorescence for γ-H2AX as well as the absorbed dose in the nucleus. We further studied the effectiveness of 111 In-and 177 Lu-labeled panitumumab for treatment of s.c. PANC-1 tumours in vivo in NOD/SCID or NRG mice, respectively. Panitumumab was conjugated to DOTA chelators or to a MCP with multiple DOTA to complex 111 In or to DOTA to complex 177 Lu. Tumour-bearing mice were treated with the RICs at amounts that caused no normal tissue toxicity. Estimated using D ¼ PÃ S Â S T←S , whereÃ S is the time-integrated activity (Bq × sec) in the source organ (see Supplementary Fig. S2), and S is the Snyder factor for mouse organs (Bitar et al., 2007). The tumour dose was estimated using the sphere model in OLINDA/EXM dosimetry software and the tumour size (Stabin et al., 2005) b Absorbed dose for NOD/SCID mice injected i.v. (tail vein) with three amounts (10 MBq; 10 μg;~0.07 nmoles) of 111 Inlabeled RICs separated by 3 weeks c Absorbed dose for NRG mice injected i.v. (tail vein) with a single amount (6 MBq; 10 μg;~0.07 nmoles) of 177 Lu-labeled RICs Aghevlian et al. EJNMMI Radiopharmacy and Chemistry (2020)  panitumumab-DOTA-[ 177 Lu]Lu, it is difficult to make direct comparisons of the effectiveness of 111 In or 177 Lu-labeled RICs. PANC-1 tumours grew more slowly in NOD/ SCID mice than in NRG mice, revealed by the slower tumour growth rates in mice treated with normal saline or unlabeled panitumumab (Fig. 4a,b). Thus, it is possible that panitumumab-DOTA-[ 177 Lu]Lu may be even more effective than panitumumab-MCP- [ 111 In]In or panitumumab-DOTA- [ 111 In]In if administered to NOD/SCID mice, but our concern was that panitumumab-DOTA-[ 177 Lu]Lu may be unusually toxic in these mice since NOD/SCID mice harbour a germ-line defect in DNA repair that radiosensitizes normal tissues (Biedermann et al., 1991). Nonetheless, by increasing the amount administered by 5-fold, apparently greater tumour growth inhibition was achieved in NOD/SCID mice treated with panitumumab-DOTA- [ 111  We previously found that 177 Lu-labeled bispecific radioimmunoconjugates (bsRICs) that bind HER2 and EGFR were more effective for RIT of s.c. MDA-MB-231/H2N human breast cancer xenografts in athymic mice than the corresponding 111 In-labeled bsRICs when administered at the same amounts (11.0 MBq) (Razumienko et al., 2016). However, the long range β-particles emitted by 177 Lu increase the risk for hematopoietic system toxicity mediated by a "cross-fire" effect and this has been doselimiting for RIT with 177 Lu-labeled RICs (Vallabhajosula et al., 2016). Administration of higher amounts of 177 Lu-labeled panitumumab to increase the effectiveness for RIT of PnCa would result in increased normal tissue toxicity. Alternatively, higher amounts of 111 In-labeled panitumumab could be administered to achieve greater therapeutic effects, due to the absence of a cross-fire effect in the case of AE (Ku et al., 2019). Interestingly, Behr et al. (Behr et al., 2000) reported that CO17-1A monoclonal antibodies labeled with the AE-emitters, 125 I or 111 In were more effective for RIT of human colon cancer xenografts in mice than CO171A labeled with the β-particle emitters, 131 I or 90 Y, but the RICs were administered at equitoxic and not equal amounts. Notably, a 10-fold higher amount of 125 I-CO17-1A (111 MBq) than 131 I-CO17-1A (11.1 MBq) and a 21-fold higher amount of 111 In-CO-17-1A (85 MBq) than 90 Y-CO17-1A (4 MBq) was administered safely for these RIT studies. Nonetheless, 111 In emits γ-photons [Eγ-171 keV (90%) and 245 keV (94%)] that irradiate normal tissues, and at the high amounts needed for RIT, may cause significant non-targeted normal tissue toxicity. AE emitters with a higher ratio of electrons/photons such as 201 Tl,193m Pt,195m Pt,197 Hg, 119 Sb or 161 Tb may reduce the potential for γ-photon mediated toxicity from RIT (Ku et al., 2019). Aghevlian et al. EJNMMI Radiopharmacy and Chemistry (2020)

Conclusions
Panitumumab labeled with AE-emitting, 111 In was less potent in vitro on a per MBq exposure basis for killing PANC-1 cells in vitro than panitumumab-labeled with βparticle emitting, 177 Lu. This lower cytotoxicity was correlated with fewer DNA DSBs in the cell nucleus and a lower absorbed dose in the nucleus. RIT with panitumumab labeled with 111 In or 177 Lu was effective for inhibiting the growth of s.c. PANC-1 xenografts in vivo in NOD/SCID or NRG mice, respectively, at amounts that caused no normal tissue toxicity, while unlabeled panitumumab was ineffective. However, a 5-fold higher amount of 111 In-labeled panitumumab (total = 30 MBq; 30 μg;~0.21 nmoles) than 177 Lu-labeled panitumumab (6 MBq; 10 μg;~0.07 nmoles) was required to achieve equivalent or greater tumour growth inhibition. Nonetheless, we conclude that RIT of PnCa with 111 In or 177 Lu-labeled panitumumab exploiting EGFR overexpression present in > 90% of cases is a promising approach that could overcome the resistance seen with treatment with anti-EGFR monoclonal antibodies such as panitumumab.