In vitro detection of canine anti-human antibodies following intratumoral injection of the hu14.18-IL2 immunocytokine in spontaneous canine melanoma

Abstract

Background: Canine and human malignant melanoma are naturally occurring cancers with many similarities, making the dog an important parallel patient population to study both diseases. However, development of canine anti-human antibodies (CAHA) needs to be considered when evaluating humanized biotherapeutics in dogs.

Objectives: Characterize CAHA in sera from dogs with spontaneous melanoma receiving radiotherapy and intratumoral immunocytokine (IT-IC) with humanized 14.18-IL2.

Methods: Serum samples were obtained pre-treatment and at several post-treatment times from 12 dogs with locally advanced or metastatic melanoma treated with radiotherapy to the primary site and regional lymph nodes (when clinically involved) followed by IT-IC of humanized 14.18-IL2. Two CAHA assays were developed. A sandwich enzyme-linked immunosorbent assay (ELISA) was developed to detect antibodies against the humanized IgG component of hu14.18-IL2. A flow cytometry assay was developed to determine the ability of CAHA to inhibit binding of a mouse anti-GD2 monoclonal antibody to its target.

Results: Post-treatment sera from 7 of 12 dogs developed CAHA levels over pre-treatment that were identified by ELISA as significant increases at Day 30 and/or Day 60. Day 10, Day 30, and Day 60 post-treatment sera from 10 of 12 dogs significantly inhibited the binding of anti-GD2 monoclonal antibody to its target compared to pre-treatment. Significant binding inhibition was also detected in 2 of 12 dogs after local RT but before IT-IC (Day 1). Normal canine sera did not mediate binding inhibition.

Conclusions: This study advances CAHA detection strategies and reports the kinetics of CAHA following IT-IC in dogs with spontaneous melanoma.

Fig 1. Immunocytokine hu14.18-Il2.

The mAb hu14.18 is specific for GD2 expressed on tumor cells and is linked to two molecules of IL2 at the Fc region which serve to attract immune effector cells to the tumor microenvironment. Figure adapted from [47]. Created in BioRender. Zuleger, C. (2025) https://BioRender.com/yq3u26o.

Fig 2. Treatment Schema.

All dogs receive radiation therapy (RT) to the local site either in 1x 8 Gy (Arm A) or 3x 8 Gy fractions over 1 week (Arm B). RT was delivered on Day −4 (Arm A) or on Days −8, −6, and −4 (Arm B). Intratumoral (IT) injection of IC (IT-IC) is administered on three consecutive days starting five days after completion of RT. All days were allowed ± 1-2 days leeway for flexible scheduling around holidays, subject vacations, and clinic scheduling. ^aArm A, a single 8 Gy fraction; ^bArm B, three 8 Gy fractions delivered on a Monday, Wednesday and Friday schedule.

Fig 3. CAHA is induced in some dogs receiving IT-IC injections.

Mean CAHA levels, measured as O.D. values by ELISA, from sera assayed in quadruplicate collected at Baseline (before radiation), Day 1 (after radiation, but prior to IT-IC), and Days 10, 30 and 60 (after IT-IC) with standard deviation error bars are shown. Day 60 timepoint was not available for two dogs, ITIC-06 and ITIC-12, in Arm A. Increases in absorbance from Baseline to Day 1, Day 10, Day 30, or Day 60 for each dog were analyzed and significance indicated as * p < 0.05; ** p < 0.01; *** p < 0.001.

Fig 4. In vitro inhibition of binding of anti-GD2 antibody to GD2 target.

The % binding inhibition at various timepoints compared to Baseline. Data shown are calculated from the mean of triplicates as described in Materials and Methods. Sera was collected at Baseline (before radiation), Day 1 (after radiation, but just prior to IT-IC), and Days 10, 30 and 60 (after IT-IC). Timepoints for each dog are presented and connected by a line. Day 60 timepoint was not available for two dogs, ITIC-06 and ITIC-12, in Arm A. The indicated significance is based on mean of triplicates at Baseline compared to post-treatment timepoints, * p < 0.05; ** p < 0.01; *** p < 0.001.

Fig 5. The relationship of CAHA and binding inhibition by in vitro flow cytometry assay.

The % binding inhibition related to CAHA at various timepoints is shown. Timepoints for each dog are presented. Timepoints with significant CAHA responses are shown with significance indicated, these timepoints also had significant binding inhibition except for ITIC-14 Day 30. Sera from 10 of 12 dogs significantly inhibited binding of anti-GD2 antibody to GD2 target at Day 10, 30 and 60 with the exceptions of ITIC-13 (shown as open symbols) and ITIC-14 (shown as black symbols). Day 1 sera from two dogs (ITIC-10 and ITIC-15) inhibited binding. A single serum sample, ITIC-14 Day 30, showed a significant CAHA response but did not inhibit binding (shown as the black circle).