Tumour sites and normal tissues were manually delineated on one of the 89Zr- cetuximab PET/CT scans (day 6 or 7) by the same observer (JvL). For one patient, delineations were performed on the FDG PET/CT scan. Tumour sites were delineated based on the CT using FDG PET information when available. To quantify uptake in muscle and liver, a transversal CT slice of the subscapular muscle and the liver was delineated. The resulting regions of interest were subsequently projected onto the other89Zr-cetuximab PET/CT scans and FDG PET scan through coregistration of the corresponding CT images using rigid registration.
2
PET imaging of Zirconium-89 labelled cetuximab: a phase I trial 35
The mean, maximum and peak standardized uptake value (SUVmean, SUVmaxand
SUVpeak, respectively) were determined using in-house developed dedicated software.
The tumour SUVpeakwas defined by calculating the mean SUV in a sphere with a
diameter of 1.2 cm within the tumour region with the highest activity. Tumour-to- background ratio (TBR) was calculated by dividing tumour SUVpeakby SUVmeanin the
aortic arch.
For patients in whom an additional18F-FDG PET scan was performed, the FDG PET and89Zr-cetuximab PET uptake were visually compared.
Results
Patient characteristics are shown in Supplementary Table S2.1, tumour characteristics in Table 2.1. Nine patients, with a median age of 62 years (range: 53–75), were included. The third patient was excluded from the study before injection of89Zr- labelled cetuximab, because of malignant hypertension during administration of the unlabelled cetuximab. The blood pressure normalized within two hours of observation without further consequences. Therefore an extra patient (patient 4) was included in dose step 1. Only two patients were included in step 3 of the study. Given the slow accrual, and since this step was not necessary to reach the primary endpoint, it was decided to close this step prematurely.
From the nine patients included, six had non-small cell lung cancer (NSCLC) and three had head and neck cancer (HNC). All patients had previously undergone anticancer therapy. In one patient, EGFR expression and mutation, and KRAS mutation status of the primary tumour could not be assessed due to insufficient quantity of histological material. In the other patients, six of the seven primary tumours showed high EGFR expression, while none showed a mutation of the EGFR gene. An example of IHC staining in a biopsy with high EGFR expression is shown in Supplementary Figure 2.2 (patient 6). In the only patient with low EGFR expression, the primary tumour showed a mutation in the KRAS gene. One of the two patients included in step 3 showed high EGFR expression on a recent tumour biopsy. The recent specimen of the second patient contained insufficient material to allow a reliable EGFR analysis.
Each patient received 89Zr-cetuximab administration as planned. Patients included in step 1 and 2 underwent the89Zr PET/CT scans at three consecutive days, except for patient 6 who could not undergo the first scan due to pain. Patient 4, included in step 1, was scanned at day 5–7 instead of day 4–6 because of logistical reasons. In two patients (patient 2 and 4) enrolled in the first dose step, the series of PET/CT scans after the second injection could not be performed due to pain and dyspnoea. Patients included in step 3 were scanned at day 6 after injection. Four
36
Ch
apt
er
2
Table 2.1. Tumour characteristics, and FDG PET and89Zr-cetuximab PET uptake in the tumour and normal tissues.
Primary tumour Schedule FDG PET 89Zr-cetuximab PET
Mutation** Tumour Liver Muscle Patient Site Histology KRAS EGFR EGFR IHC score SUVmax TBRpeak SUVpeak SUVmax SUVmean SUVmax SUVmean SUVmax SUVmean
1 Lung SCC NA NA NA step 1 11.3 2.8 4.7 7.1 1.8 NA NA 3.4 0.6 2 Lung LCC + - 80 step 1 NA 3.1 4.3 7.8 2.1 NA NA 2.6 0.6 3* Lung LCC NA NA NA step 1 NA NA NA NA NA NA NA NA NA 4 Lung AC - - 280 step 1 NA 4.6 3.8 7.6 1.3 8.1 5.5 1.4 0.4 5 Lung SCC - - 240 step 2 10.2 1.0 1.2 1.6 0.8 7.8 5.7 1.0 0.2 6 Lung AC - - 290 step 2 7.6 1.7 3.6 6.6 1.9 9.9 9.2 1.3 0.3 7 Oropharynx SCC - - 300 step 2 11.3 3.2 3.0 4.3 2.0 7.0 5.3 1.6 0.3 8 Oropharynx MC - - 270 step 3 NA 1.4 4.4 6.2 3.2 7.1 6.8 1.8 0.8 9 Oropharynx SCC - - 290 step 3 NA 1.5 4.8 6.9 3.0 18.4 12.5 1.2 0.4 *Excluded patient.
**KRAS: codons 12 and 13; EGFR: exons 18–21.
2
PET imaging of Zirconium-89 labelled cetuximab: a phase I trial 37
patients underwent an FDG PET scan within the study period, at an interval of 12 or 13 days after the injection of89Zr-cetuximab.
An overview of the toxicity per patient is presented in Supplementary Table 2.2. No toxicity other than skin rash was observed in any of the patients. Of the eight patients that received the full cetuximab administration, three patients experienced grade one toxicity, and four experienced grade two acne form rash. No changes were observed for haematological, kidney and liver function compared to baseline.
89Zr-cetuximab PET images of all patients are shown in Figure 2.2 (coronal) and
Figure 2.3 (transversal). Tumour and normal tissue uptake values of89Zr-cetuximab are shown in Table 2.1. All but one patient had a peak TBR larger than 1 at any of the imaging time points. For each patient, the89Zr-cetuximab image with the highest TBR was selected for further analyses. For patients in dose step 1, only the scans after the first89Zr-cetuximab injection were analysed, since for two of the three patients PET/CT scans could not be acquired in the second week. The average peak TBR was 2.4 (range: 1.0–4.6). The average tumour SUVmaxand SUVmeanwere 6.0 (range:
1.6–7.8) and 2.0 (range: 0.8–3.2), respectively. The average SUVmaxand SUVmeanfor
the liver were 9.7 (range: 7.0–18.4) and 7.5 (range: 5.3–12.5). For muscle, the average SUVmaxand SUVmeanwere 1.8 (range: 1.0–3.4) and 0.4 (range: 0.2–0.8). Both SUVmax
and SUVmean of the tumour were significantly higher than the muscle SUVmax and
SUVmean.
The FDG PET and 89Zr-cetuximab scans of the four patients with an FDG
PET/CT scan are shown in Figure 2.4. Visual comparison showed a remarkable mismatch between FDG PET and89Zr-cetuximab PET uptake in one patient (patient 1, Figure 2.4). No direct relationship was observed between the EGFR IHC score and TBR.
For the patients in dose step 1, the average peak TBR at day 4, 5, and 6 after the first injection was 1.3 (range: 1.2–1.4), 2.1 (range: 1.7–2.7) and 3.1 (range: 1.7–4.6), respectively (Figure 2.5). As the highest TBR was found at day 6 after injection, imaging in dose step 2 was performed at day 5, 6, and 7. For patients in dose step 2, the average peak TBR at day 5, 6, and 7 after injection was 1.7 (range: 0.9–2.4), 1.6 (range: 1.0–2.7) and 2.0 (range: 0.9–3.2) (Figure 2.5). The peak TBR of patients included in step 3 was 1.4 and 1.5 (day 6). At the regular imaging time points, the highest peak TBR was seen in patient 4 (4.6), at day 6 after injection. In patient 1, an extra PET/CT scan was performed at day 12 after injection. The TBR at this time point was higher than the maximum TBR for this patient at the regular imaging time points (2.8 at day 12 versus 1.7 at day 6).
38 Chapter 2
step 1 - 12 days p.i. - 60 MBq
p1 p2 p4 p5
p6 p7 p8 p9
step 1 - 6 days p.i. - 60 MBq step 1 - 6 days p.i. - 60 MBq step 2 - 6 days p.i. - 120 MBq
step 3 - 6 days p.i. - 120 MBq step 3 - 6 days p.i. - 120 MBq
step 2 - 7 days p.i. - 120 MBq step 2 - 7 days p.i. - 120 MBq
Figure 2.2. 89Zr-cetuximab PET maximum intensity projections of all patients. The GTV of the primary tumour is overlaid in blue. The notes underneath the images indicate in which dose step the patients were included, how many days postinjection the displayed images were acquired and the dose of89Zr-cetuximab (2 × 60 MBq or 1 × 120 MBq) administered. The scans with the highest TBR in the primary tumour were selected. For dose step 1, the images after the first injection were analysed.