Kinetic analysis of HER2-binding ABY-025 Affibody molecule using dynamic PET in patients with metastatic breast cancer

Abstract

Background: High expression of human epidermal growth factor receptor type 2 (HER2) represents an aggressive subtype of breast cancer. Anti-HER2 treatment requires a theragnostic approach wherein sufficiently high receptor expression in biopsy material is mandatory. Heterogeneity and discordance of HER2 expression between primary tumour and metastases, as well as within a lesion, present a complication for the treatment and require multiple biopsies. Molecular imaging using the HER2-targeting Affibody peptide ABY-025 radiolabelled with 68Ga-gallium for PET/CT is currently under investigation as a non-invasive tool for whole-body evaluation of metastatic HER2 expression. Initial studies demonstrated a high correlation between 68Ga-ABY-025 standardized uptake values (SUVs) and histopathology. However, detecting small liver lesions might be compromised by high background uptake. This study aimed to explore the applicability of kinetic modelling and parametric image analysis for absolute quantification of 68Ga-ABY-025 uptake and HER2-receptor expression and how that relates to static SUVs.

Methods: Dynamic 68Ga-ABY-025 PET of the upper abdomen was performed 0-45 min post-injection in 16 patients with metastatic breast cancer. Five patients underwent two examinations to test reproducibility. Parametric images of tracer delivery (K1) and irreversible binding (Ki) were created with an irreversible two-tissue compartment model and Patlak graphical analysis using an image-derived input function from the descending aorta. A volume of interest (VOI)-based analysis was performed to validate parametric images. SUVs were calculated from 2 h and 4 h post-injection static whole-body images and compared to Ki.

Results: Characterization of HER2 expression in smaller liver metastases was improved using parametric images. Ki values from parametric images agreed very well with VOI-based gold standard (R2 > 0.99, p < 0.001). SUVs of metastases at 2 h and 4 h post-injection were highly correlated with Ki values from both the two-tissue compartment model and Patlak method (R2 = 0.87 and 0.95, both p < 0.001). 68Ga-ABY-025 PET yielded high test-retest reliability (relative repeatability coefficient for Patlak 30% and for the two-tissue compartment model 47%).

Conclusion: 68Ga-ABY-025 binding in HER2-positive metastases was well characterized by irreversible two-tissue compartment model wherein Ki highly correlated with SUVs at 2 and 4 h. Dynamic scanning with parametric image formation can be used to evaluate metastatic HER2 expression accurately.

Keywords: Affibody; Dynamic PET; HER2 receptor; Kinetic modelling; Metastatic breast cancer.

Conflict of interest statement

The authors declare that they have no competing interests.

Figures

Fig. 1

Two-tissue compartment kinetic model used in the analysis of dynamic PET studies with 68Ga-ABY-025 in patients with metastatic breast cancer. K1, k2, and k3 represent the transfer rate constants

Fig. 2

Time-activity curves obtained from dynamic ABY-025 PET of the aorta, liver, and metastasis in a patient with metastatic breast cancer

Fig. 3

Correlations between a VOI-based 2TC Ki and Patlak Ki, b parametric 2TC Ki and Patlak Ki, c VOI-based and parametric 2TC Ki values, and d VOI-based and parametric Patlak Ki values (PMI, parametric images)

Fig. 4

Positron emission tomography (PET) images. a In a patient with HER2-positive expression. b In a patient with HER2-negative expression. 18F-FDG PET images to the left followed by SUV static image taken at 2 h time point after 68Ga-ABY-025 injection. 2TC Ki, Patlak Ki, and (2TC) K1 are parametric images created from the 68Ga-ABY-025 dynamic PET study (K1, transfer rate constant from plasma to the tissue compartment; Ki, influx rate of the tracer into irreversible compartment)

Fig. 5

a 2TC-3k Ki correlation with SUV 2 h and SUV 4 h (R2 = 0.87, 0.80 respectively, both p < 0.0001). b Patlak Ki correlation with SUV 2 h and SUV 4 h (R2 = 0.95, 0.90 respectively, both p < 0.0001). Shaded areas represent 95% confidence band

Fig. 6

Reproducibility of parametric images shown as a correlation of a test-retest 2TC Ki values (R2 = 0.85, p < 0.0001) and b test-retest Patlak Ki values (R2 = 0.94, p < 0.0001) for the metastatic lesions in a group of patients rescanned after 1 week (n = 5). The solid line is the Deming line, and the dotted line is the identity line

Fig. 7

The Bland-Altman plot a of two-tissue compartment influx rate (2TC-3k Ki) and b Patlak Ki

Fig. 8

FDG PET and ABY-025 PET SUV 2 h and parametric images of both Patlak Ki and 2TC-3k Ki in a breast cancer patient with multiple small liver metastases