Bispecific radioimmunoconjugates exploit receptor heterogeneity for positron emission tomography of tumors expressing HER2 and/or EGFR

Summary HER2 heterogeneity is a challenge for molecular imaging or treating HER2-positive breast cancer (BC). EGFR is coexpressed in some tumors exhibiting HER2 heterogeneity. Bispecific radioimmunoconjugates (bsRICs) that bind HER2 and EGFR were constructed by linking trastuzumab Fab through polyethyleneglycol (PEG24) to EGF. We established s.c. tumors in NOD-SCID mice that homogeneously or heterogeneously expressed HER2 and/or EGFR by the inoculation of HER2-positive/EGFR-negative SK-OV-3 cells, EGFR-positive/HER2-negative MDA-MB-468 cells or mixtures of these cells. [64Cu]Cu-NOTA-trastuzumab Fab-PEG24-EGF were compared to [64Cu]Cu-NOTA-trastuzumab Fab or [64Cu]Cu-NOTA-EGF for the PET imaging of HER2 and/or EGFR-positive tumors. [64Cu]Cu-NOTA-trastuzumab Fab-PEG24-EGF bsRICs imaged tumors expressing HER2 or EGFR or heterogeneously expressing these receptors, while monospecific agents only imaged HER2-or EGFR-positive tumors. Our results indicate that bsRICs labeled with 64Cu are able to exploit receptor heterogeneity for tumor imaging. PET may select patients for radioimmunotherapy with bsRICs complexed to the β-particle emitter, 177Lu or Auger electron-emitter, 111In in a theranostic approach.


INTRODUCTION
The human epidermal growth factor receptor (HER2) is overexpressed on 15-20% of breast cancers (BC) 1 and is the therapeutic target for trastuzumab (Herceptin, Roche), pertuzumab (Perjeta, Roche) and the antibody-drug conjugate (ADC) trastuzumab-emtansine (T-DM1; Kadcyla, Roche). 2 HER2 status is assessed on a tumor biopsy by immunohistochemical (IHC) staining for HER2 protein on the cell membrane or by in situ hybridization (ISH) to identify HER2 gene amplification. 3Only patients with HER2-positive BC are selected for HER2-targeted therapies, although it has been reported that patients with BC exhibiting low (IHC 1+) or moderate HER2 expression (IHC 2+) without HER2 gene amplification may benefit from treatment with trastuzumab deruxtecan, a potent ADC. 4 Nonetheless, despite the careful application of criteria to define HER2-positivity, heterogeneous intratumoral HER2 staining and/or HER2 gene amplification with HER2-positive and HER2-negative regions within the same tumor have been reported in 6-36% of tumors scored as HER2-positive. 5 HER2 heterogeneity is correlated with a poor response to HER2-targeted therapies including trastuzumab, 6 trastuzumab and pertuzumab combined with chemotherapy 7 and trastuzumab emtansine (T-DM1). 8HER2 receptor heterogeneity has further been correlated with shorter disease-free survival in patients with HER2-positive BC. 9 One strategy to overcome and exploit receptor heterogeneity in HER2-positive BC may be to target both HER2 and epidermal growth factor receptors (EGFR).HER2 forms heterodimers with EGFR that initiate a strong tumor growth signal. 10This is the rationale for combining pertuzumab, an inhibitor of HER2 dimerization, with trastuzumab for the treatment of HER2-positive BC. 11 In one study of 119 surgical tumor specimens, 14% of tumors that were classified as HER2-positive were EGFR-positive. 12Coexpression of HER2 and EGFR increased the risk for metastasis and was associated with shorter disease-free and overall survival (OS).In another study of 2,567 tumors, EGFR positivity was found in 18% of cases, but EGFR-positive BC was more likely to be HER2-positive, with 26% of EGFR-positive tumors HER2-positive. 13Furthermore, EGFR/HER2 heterodimerization has been proposed as a mechanism of resistance to HER2-targeted therapies, since trastuzumab does not block the dimerization of HER2 with the EGFR. 14To exploit receptor heterogeneity in HER2-positive BC, we previously reported bispecific radioimmunoconjugates (bsRICs) that recognize HER2 and EGFR. 15These bsRICs were composed of trastuzumab Fab linked through a 24-mer polyethyleneglycol (PEG 24 ) spacer to EGF and modified with diethylenetriaminepentaacetic acid (DTPA) to complex 111 In [t 1/2 = 2.8 days; Eg = 171 keV (90.7%) and 245 keV (94.1%)].At 48 h post-injection (p.i.) of [ 111 In]In-DTPA-trastuzumab Fab-PEG 24 -EGF bsRICs, subcutaneous (s.c.) SK-OV-3 human ovarian cancer xenografts that were HER2-positive but EGFR-negative, MDA-MB-231 human BC xenografts that were HER2-negative but EGFR-positive, and MDA-MB-231/H2N tumor xenografts that coexpressed HER2 and EGFR were imaged in mice by single photon emission computed tomography (SPECT).Furthermore, in a subsequent study, radioimmunotherapy (RIT) with [ 111 In]In-DTPA-trastuzumab Fab-PEG 24 -EGF bsRICs or bsRICs modified with 2,2 0 ,2 00 ,2%-(1,4,7,10-tetraazacyclododecane-1,4,7,10-tetrayl)tetraacetic acid (DOTA) to complex the b-particle emitter, 177 Lu [t 1/2 = 6.7 days; Ebmax = 0.50 MeV (78.6%), 0.38 MeV (9.1%), 0.18 MeV (12.2%)] administered at the No Observable Adverse Effect Level (NOAEL) inhibited the growth of s.c.MDA-MB-231/H2N tumors and trastuzumab-resistant TrR1 tumors that co-expressed HER2 and EGFR. 16These studies demonstrated that targeting HER2 and/or EGFR using bsRICs is a feasible approach for theranostic imaging and RIT of tumors that express these two receptors.
Positron emission tomography (PET) is a more sensitive imaging modality than SPECT.Thus, we constructed analogous bsRICs modified with 1,4,7-triazacyclononane-1,4,7-triacetic acid (NOTA) to complex the positron-emitter, 64 Cu [t 1/2 = 12.7 h; Eb + = 0.7 MeV (19%)] for the PET imaging of tumors that express HER2 and/or EGFR. 17These [ 64 Cu]Cu-NOTA-trastuzumab Fab-PEG 24 -EGF bsRICs imaged s.c.HER2negative/EGFR-positive MDA-MB-468, HER2-positive/EGFR-positive MDA-MB-231/H2N, or HER2-positive/EGFR-negative SK-OV-3 tumors in non-obese diabetic severe combined immunodeficiency (NOD-SCID) mice at 24 or 48 h p.i.An important finding was that the PEG 24 linker prolonged the residence time in the blood compared to bsRICs without a PEG 24 linker, which may contribute to higher tumor uptake of bsRICs compared to monospecific [ 64 Cu]Cu-NOTA-trastuzumab Fab or [ 64 Cu]Cu-NOTA-EGF.These studies were performed in NOD-SCID mice with s.c.tumors formed by inoculating tumor cells that homogeneously express HER2 and/or EGFR, but it would be useful to be able to engineer an experimental tumor xenograft model in mice composed of tumor cells that express HER2 or EGFR or have predefined receptor hetereogeneity by inoculating a mixed population of tumor cells expressing these two receptors.The ability to predefine tumor receptor homogeneity or heterogeneity would enable an investigation of these bsRICs for imaging or RIT under different conditions of receptor heterogeneity.We report here for the first time the establishment and characterization of tumor xenograft mouse models with predefined HER2 and/or EGFR expression, and demonstrate that [ 64 Cu]Cu-NOTA-trastuzumab Fab-PEG 24 -EGF bsRICs were able to exploit receptor heterogeneity for tumor imaging by PET.

Flow cytometry for human HER2-positive and EGFR-positive cells in populations of MDA-MB-468 and/or SK-OV-3 cells
Increasing the proportion of MDA-MB-468 human BC cells (EGFR-positive/HER2-negative) mixed with SK-OV-3 human ovarian cancer cells (HER2-positive/EGFR-negative) from 0% to 100% provided a cell population that was homogeneously HER2-positive, or exhibited an increasing gradient of EGFR-positive cells, or was homogeneously EGFR-positive.Similarly, increasing the proportion of SK-OV-3 cells mixed with MDA-MB-468 cells from 0% to 100%, provided a cell population that was homogeneously EGFR-positive, or exhibited an increasing gradient of HER2-positive cells, or was homogeneously HER2-positive.Flow cytometric analysis of cells immunostained with anti-HER2 Alexa Fluor 488-conjugated trastuzumab characterized cell populations containing 100% SK-OV-3 cells or mixtures of SK-OV-3 cells with 25%, 50%, 75%, or 100% MDA-MB-468 cells as 99.9%, 65.9%, 35.5%, 19.9% and 0% HER2-positive, respectively (Figure 1A).Similarly, flow cytometry of cells immunostained with anti-EGFR Alexa Fluor 647-conjugated panitumumab characterized cell populations containing 100% MDA-MB-468 cells or mixtures of MDA-MB-468 cells with 25%, 50%, 75% or 100% SK-OV-3 cells as 93.6%, 71.3%, 47.2, 20.2% and 0.3% EGFR-positive, respectively (Figure 1B).There was a strong linear correlation between the percent of SK-OV-3 and MDA-MB-468 cells and HER2 or EGFR positivity of cell populations, respectively (Figures 1C and 1D).These results validated the flow cytometric method for measuring the proportion of HER2-positive and EGFR-positive cells in tumor cell mixtures.This method was used to characterize the percent of HER2-positive and EGFR-positive cells ex vivo in s.c.tumor xenografts in NOD-SCID mice formed by the inoculation of 100% SK-OV-3 or MDA-MB-468 cells or a mixture of 30% SK-OV-3 cells and 70% MDA-MB-468 cells.Receptor characterization of these tumors was further used to interpret the results of PET imaging and biodistribution studies.
We assessed the spatial distribution of HER2 and EGFR expression in tumors formed by the s.c.inoculation of SK-OV-3 or MDA-MB-468 cells or a mixture of 30% SK-OV-3 and 70% MDA-MB-468 cells in NOD/SCID mice by immunohistochemical (IHC) staining.SK-OV-3 tumors exhibited strong and homogeneous immunopositivity for HER2 (Figure 4A), while MDA-MB-468 tumors exhibited strong and homogeneous immunopositivity for EGFR (Figure 4B).Replicate tumors (n = 3) formed by the inoculation of a mixture of 30% SK-OV-3 cells and 70% MDA-MB-468 cells exhibited spatially complementary HER2 or EGFR immunostaining (Figure 4C).The percentage of cells that were HER2-positive/ HER2-negative or EGFR-positive/EGFR-negative in these sections was quantified by image analysis (Figure 4D).MDA-MB-468 tumors contained 100 G 0% EGFR-positive cells and 0.7 G 0.1% HER2-positive cells.SK-OV-3 tumors contained 94 G 6% HER2-positive cells and 18 G 4% EGFR-positive cells.Some cells stained positive for HER2 and EGFR, resulting in >100% cell staining for these receptors.In tumors formed by inoculating mixtures of SK-OV-3 and MDA-MB-468 cells, there were 27 G 3% HER2-positive cells and 67 G 7% EGFR-positive cells (Figure 4D) which were significantly different (p < 0.05) than tumors composed only of SK-OV-3 or MDA-MB-468 cells.The proportion of HER2positive and EGFR-positive cells in tumors measured ex vivo by flow cytometry or by IHC staining agreed with the percentage of SK-OV-3 and MDA-MB-468 cells inoculated in NOD-SCID mice to form these tumors.

DISCUSSION
Receptor heterogeneity is a challenge for targeted cancer therapies since it may lead to poor response to treatment and moreover, may apply evolutionary pressure to select tumor cells that do not express the target. 18One approach to overcome receptor heterogeneity may be to use agents that recognize more than one target.Our previous studies showed that SPECT with [ 111 In]In-DTPA-trastuzumab Fab-PEG 24 -EGF 15 and PET with [ 64 Cu]Cu-NOTA-trastuzumab Fab-PEG 24 -EGF 17 bsRICs imaged tumors in mice that express HER2 or EGFR or coexpress these receptors.We further found that the growth of tumors expressing HER2, EGFR or both receptors in mice was inhibited by RIT with Auger electron-emitting [ 111 In]In-DTPA-trastuzumab Fab-PEG 24 -EGF or b-particle emitting [ 177 Lu]Lu-DOTA-trastuzumab Fab-PEG 24 -EGF. 16However, these studies did not examine the ability of bsRICs to localize in tumors with receptor heterogeneity arising from mixed populations of tumor cells that express HER2 or EGFR.We describe here for the first time tumor xenografts in NOD-SCID mice with predefined receptor heterogeneity formed by the s.c.inoculation of mixtures of HER2-positive/EGFR-negative SK-OV-3 human ovarian cancer cells and EGFR-positive/ HER2-negative MDA-MB-468 human BC cells.The tumor localization of [ 64 Cu]Cu-NOTA-trastuzumab Fab-PEG 24 -EGF bsRICs in these tumors with heterogeneous receptor expression was evaluated by PET and biodistribution studies and compared to tumors formed of SK-OV-3 cells or MDA-MB-468 cells that homogeneously express HER2 or EGFR, respectively.In addition, the tumor localization of these bsRICs was compared to monospecific [ 64 Cu]Cu-NOTA-trastuzumab Fab or [ 64 Cu]Cu-NOTA-EGF.Only one other study has reported tumors in mice that heterogeneously express HER2 formed by the inoculation of mixtures of HER2-positive and HER2-negative tumor cells. 19In that study, tumors composed of HER2-positive TUBO murine mammary carcinoma tumors or a HER2-negative subclone, TUBO-P2J or mixtures of these cells were inoculated s.c. in Balb/c mice.Tumors formed from mixtures of these two cell types exhibited HER2 heterogeneity and did not respond to treatment with anti-neu antibodies (analogous to anti-HER2 trastuzumab).However, that study did not evaluate bispecific antibodies that recognize other targets coexpressed with HER2.
SK-OV-3 human ovarian cancer cells and MDA-MB-468 human BC cells were chosen to establish tumors that homogeneously or heterogeneously expressed HER2 and EGFR since SK-OV-3 cells are HER2-positive but have negligible EGFR expression 20 while MDA-MB-468 cells are EGFR-positive but have negligible HER2 21 (Figure 1).Both cell types are tumorigenic in mice.We did not select SK-BR-3 or BT-474 human BC cells as a HER2-positive tumor cell type because SK-BR-3 cells are poorly tumorigenic 22 and SK-BR-3 and BT-474 cells coexpress EGFR at moderate or low levels, respectively. 23Our intent was to establish tumors that heterogeneously expressed HER2 and EGFR on different tumor cell populations.Flow cytometry accurately measured the percentage of HER2-positive SK-OV-3 and EGFR-positive MDA-MB-468 cells in mixtures of these two cell types (Figures 2 and 3).It was important to select the proportion of SK-OV-3 and MDA-MB-468 cells in cell mixtures for s.c.inoculation into NOD-SCID mice to establish tumors with receptor heterogeneity, since there is a risk that one cell type may outcompete the other as tumors grow due to differences in cell doubling times in vivo.Therefore, we first measured the cell doubling times in vitro, which were 19.2 h and 33.5 h for SK-OV-3 and MDA-MB-468 cells, respectively.To avoid overgrowth of SK-OV-3 cells, we inoculated a mixture of 30% SK-OV-3 and 70% MDA-MB-468 cells s.c. in NOD-SCID mice to form tumors with heterogeneous HER2 and EGFR expression.However, analysis of the percentage of HER2-positive and EGFR-positive cells from replicate tumors (n = 3) at 6 weeks post-inoculation ex vivo by flow cytometry revealed 30.7 G 7.4% of cells assigned as HER2-positive (Figure 2C) and 72.1 G 3.2%) EGFR-positive cells (Figure 3C), which was almost identical to the proportion of SK-OV-3 and MDA-MB-468 cells inoculated, respectively.These results were not due to an artifact caused by isolating tumor cells ex vivo, since image analysis of tumor sections immunostained for HER2 and EGFR similarly revealed 27 G 3% HER2-positive cells and 67 G 7% EGFR-positive cells (Figures 4C and 4D).Thus, it appears that in vivo, the cell doubling times of SK-OV-3 and MDA-MB-468 cells were similar.There were spatially complementary HER2-positive and EGFR-positive regions in these All images were adjusted to the same intensity and the intensity scale (ranging from 1.5% ID/g to 12.5% ID/g) is shown at the right of each set of images.Normal organ uptake is visible on the 24 h and 48 h p.i. images but is labeled on the 48 h p.i. images.L: liver; K: kidneys; Sp: spleen.tumors (Figure 4C).In contrast, tumors formed only from SK-OV-3 or MDA-MB-468 cells exhibited homogeneous HER2-positive/EGFR-negative or EGFR-positive/HER2-negative cells, respectively by flow cytometry (Figures 2A, 2B, 3A, and 3B) and the corresponding homogeneous HER2 or EGFR immunostaining (Figures 4A and 4B).
Using in situ hybridization (ISH) techniques to identify HER2 gene amplifications, three distinct patterns of HER2 heterogeneity in human BC tumors have been described: i) ''clustered type'' with regions of HER2 amplified and non-HER2 amplified cells, ii) ''mosaic type'' with interspersed HER2 amplified and non-HER2 amplified cells, and iii) ''scattered type'' with isolated HER2-amplified cells in a tumor composed mainly of non-HER2 amplified cells. 24By combining ISH and IHC staining, two patterns of HER2 heterogeneity were defined: i) ''genetic heterogeneity'' consisting of a mixture of HER2-positive and HER2-negative cells due to corresponding HER2-amplification or non-HER2 gene amplification, and ii) ''non-genetic heterogeneity'' consisting of a mixture of HER2-positive and HER2-negative cells by ICH despite homogeneous HER2 gene amplification. 24SK-OV-3 cells are HER2 gene-amplified 25 while MDA-MB-468 cells are not amplified for the HER2 gene, thus the pattern of HER2 heterogeneity in tumors formed of a mixture of SK-OV-3 and MDA-MB-468 cells in our study could be classified as ''genetic heterogeneity.''It should be noted that in patients with HER2-positive BC, besides intratumoral HER2 heterogeneity, there may be interlesional HER2 heterogeneity.The ZEPHIR trial investigated PET imaging with [ 89 Zr]Zr-DFO-trastuzumab to predict response to treatment with trastuzumab emtansine (T-DM1) in patients with HER2-positive BC. 26 PET revealed no tumor uptake of [ 89 Zr]Zr-DFO-trastuzumab in 29% of patients and heterogeneous interlesional uptake in 46% of patients.Combining PET with [ 89 Zr]Zr-DFO-trastuzumab and [ 18 F]F-2-fluorodeoxyglucose predicted response to T-DM1 and patient survival.In our study, we established two tumors in NOD-SCID mice -one on the right flank that was homogeneously HER2-positive or EGFR-positive and one on the left flank that was heterogeneous for the expression of HER2 and EGFR.We imaged mice with these tumors by PET with bispecific [ 64 Cu]Cu-NOTA-trastuzumab Fab-PEG 24 -EGF or monospecific [ 64 Cu]Cu-NOTA-trastuzumab Fab or [ 64 Cu]Cu-NOTA-EGF (Figures 5 and 6).Imaging of tumors in this model that homogeneously express HER2 or are EGFR-positive but HER2-negative may simulate the scenario of interlesional HER2 heterogeneity in patients with HER2-positive BC.Imaging of mice with tumors that heterogeneously express HER2 and EGFR may simulate intralesional HER2 heterogeneity in patients with HER2-positive BC.
In conclusion, bispecific [ 64 Cu]Cu-NOTA-trastuzumab Fab-PEG 24 -EGF were able to exploit receptor heterogeneity and image tumors in NOD-SCID mice that homogeneously or heterogeneously expressed HER2 or EGFR, while monospecific [ 64 Cu]Cu-NOTA-trastuzumab Fab or [ 64 Cu]Cu-NOTA-EGF were limited to imaging tumors that expressed HER2 or EGFR.We previously reported that these bsRICs labeled with Auger electron-emitting, 111 In or b-particle emitting 177 Lu were effective for RIT of tumors that expressed HER2 or EGFR or co-expressed these two receptors. 16PET with [ 64 Cu]Cu-NOTA-trastuzumab Fab-PEG 24 -EGF bsRICs may identify patients who benefit from RIT with 111 In or 177 Lu-labeled bsRIC that target HER2 and EGFR, enabling a theranostic approach.There was higher kidney uptake of [ 64 Cu]Cu-NOTA-trastuzumab Fab-PEG 24 -EGF (Tables 1 and 2) than we previously reported for [ 177 Lu]-DOTA-trastuzumab Fab-PEG 24 -EGF (8% ID/g) 16 or [ 111 In] In-DTPA-trastuzumab Fab-PEG 24 -EGF (5.8-6.5% ID/g). 15The reason is not known.However, we envision treatment with 177 Lu-or 111 In-labeled bsRICs which were non-toxic to the kidneys in mice. 16In the future, it may be possible to decrease the kidney uptake of 64 Cu-labeled bsRICs by conjugating trastuzumab Fab through a PEG 24 linker to anti-EGFR panitumumab (Vectibix, Amgen) Fab instead of EGF.This would increase the molecular weight of the bsRICs to >100 kDa, which exceeds the renal threshold for the filtration of proteins. 28Finally, our results illustrate that tumor xenografts with predefined heterogeneous receptor expression may be established by inoculating mixtures of tumor cells that express distinct receptors.This approach to establishing tumor xenografts in mice could be very useful to examine the effects of receptor heterogeneity on imaging and treating tumors with multi-targeted theranostic agents.

Limitations of the study
The tumor xenograft models used in this study were composed of human SK-OV-3 ovarian cancer or MDA-MB-468 breast cancer cells which homogeneously and highly express HER2 or EGFR, respectively.The levels of HER2 and EGFR expression by these cells may be greater than found in human tumors.In addition, tumors with homogeneous or heterogeneous HER2 and EGFR expression were formed by inoculating these two cell types or mixtures of these cells.These tumors are experimental models used to compare the imaging properties of bispecific with monospecific agents under conditions of homogeneous or heterogeneous receptor expression but these conditions may not reflect the levels of these receptors or the receptor heterogeneity found in tumors in patients with HER2-positive BC.The tumor xenograft model exhibiting heterogeneous HER2 expression was not intended to incorporate resistance to HER2-targeted therapies, only receptor heterogeneity.Mediastinal activity and liver uptake of the bsRICs may interfere with the imaging of primary breast tumors or metastases of BC to the liver in patients.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:   restrictions to the availability of these radiopharmaceutical agents due to Canadian Nuclear Safety Commission (CNSC) regulations on the transfer of radioactive materials.SK-OV-3 and MDA-MB-468 cells and NOD-SCID mice for tumor engraftment are available commercially.All other methods and data are described in the article and further information is available from the lead contact upon reasonable request.
after PET imaging, mice were transferred to an eXplore Locus Ultra Preclinical CT Scanner (GE Healthcare) for a whole-body CT scan using routine acquisition parameters (16 s anatomical scan at 80 kVp, 50 mA, and voxel size of 150 3 150 3 150 mm.MicroPET and CT images were co-registered using Inveon Research Workplace software (Siemens).Immediately after the 48 h imaging time point, mice were sacrificed and tumors and samples of blood and normal tissues were collected, weighed and 64 Cu measured in a g-counter.Tumor and normal tissue uptake were expressed as percentage injected dose per gram (%ID/g).

QUANTIFICATION AND STATISTICAL ANALYSIS
Statistical comparisons were made by Student's t test (p < 0.05) using Prism Ver. 9 for MacOS (GraphPad Software, San Diego, CA) and are shown in footnotes in Tables 1 and 2 and in Figure 4D.The number of replicate determinations is indicated by ''n''.

ADDITIONAL RESOURCES
None.

Figure 1 .
Figure 1.Flow cytometry for HER2-positive and EGFR-positive cells in populations of SK-OV-3 and/or MDA-MB-468 cells SK-OV-3 or MDA-MB-468 cells or mixtures of these cells immunostained with (A) anti-HER2 Alexa Fluor 488-conjugated trastuzumab or (B) with anti-EGFR Alexa Fluor 647-conjugated panitumumab were analyzed by flow cytometry.The proportion of HER2-positive or EGFR-positive cells was quantified as the area under the relevant peak for SK-OV-3 or MDA-MB-468 cells, respectively and plotted vs. the percentage of (C) SK-OV-3 cells or (D) MDA-MB-468 cells in the cell population.

Figure 4 .
Figure 4. Immunohistochemical (IHC) staining of tumors composed of SK-OV-3 and/or MDA-MB-468 cells for HER2-positive and EGFR-positive cells Representative sections of tumors formed by the inoculation of (A) SK-OV-3 cells or (B) MDA-MB-468 cells or (C) a mixture of 30% SK-OV-3 cells and 70% MDA-MB-468 cells at 6 weeks post-inoculation in NOD/SCID mice.Sections were stained with hematoxylin and eosin (H&E) or immunostained for HER2 or EGFR.Sections are shown at 43 or 203 amplification.(D) The percentage of cells immunopositive for HER2 or EGFR in replicate tumor sections was quantified by image analysis and is shown for SK-OV-3 or MDA-MB-468 tumors or tumors formed by inoculating a mixture of 30% SK-OV-3 cells and 70% MDA-MB-468 cells.Data shown are mean G SD (n = 3).Statistically significant differences (Student's t test; p < 0.05) are indicated by the asterisks.

Figure 5 .
Figure 5. PET/CT images of NOD-SCID mice with s.c.SK-OV-3 tumors with homogeneous HER2 expression or tumors with heterogeneous HER2 expression formed from a mixture of SK-OV-3 and MDA-MB-468 cells Representative images (supine position) of NOD-SCID mice with s.c.SK-OV-3 tumors on the right flank that homogeneously express HER2 (blue arrow) and tumors formed by the inoculation of a mixture of SK-OV-3 cells (30%) and MDA-MB-468 cells (70%) with heterogeneous HER2 and EGFR expression on the left flank (red arrow).Images were obtained at 24 h and 48 h p.i. of (A) [ 64 Cu]Cu-NOTA-trastuzumab Fab-PEG 24 -EGF, (B) [ 64 Cu]Cu-NOTA-trastuzumab Fab or (C) [ 64 Cu]Cu-NOTA-EGF.All images were adjusted to the same intensity and the intensity scale (ranging from 1.5% ID/g to 12.5% ID/g) is shown at the right of each set of images.Normal organ uptake is visible on the 24 h and 48 h p.i. images but is labeled on the 48 h p.i. images.L: liver; K: kidneys; Sp: spleen.

Figure 6 .
Figure 6.PET/CT images of NOD-SCID mice with s.c.MDA-MB-468 tumors with homogeneous EGFR expression or tumors with heterogeneous EGFR expression formed from a mixture of MDA-MB-468 and SK-OV-3 cells Representative images (supine position) of NOD-SCID mice with s.c.MDA-MB-468 tumors on the right flank that homogeneously express EGFR (blue arrow) and tumors formed by the inoculation of a mixture of MDA-MB-468 cells (70%) and SK-OV-3 cells (30%) with heterogeneous HER2 and EGFR expression on the left flank (red arrow).Images were obtained at 24 h and 48 h p.i. of (A) [ 64 Cu]Cu-NOTA-trastuzumab Fab-PEG 24 -EGF, (B) [ 64 Cu]Cu-NOTA-trastuzumab Fab or (C) [ 64 Cu]Cu-NOTA-EGF.All images were adjusted to the same intensity and the intensity scale (ranging from 1.5% ID/g to 12.5% ID/g) is shown at the right of each set of images.Normal organ uptake is visible on the 24 h and 48 h p.i. images but is labeled on the 48 h p.i. images.L: liver; K: kidneys; Sp: spleen.