Radiolabelling and quality control

The standard radiolabelling process of other in-house produced 177Lu-labelled radiopharmaceuticals was adjusted to the automated labelling process of [177Lu]Lu-DOTA-MGS5. Five separate batches of [177Lu]Lu-DOTA-MGS5 were evaluated. Starting activities of 3.2–10.7 GBq were reacted with 100 µg of DOTA-MGS5, resulting in a radioactivity at the end of synthesis ranging from 2.5 to 9.7 GBq. The automated synthesis process lasted around 45 min, and the volume of the final product after synthesis was 14.5 mL with a radioactivity concentration of 175–671 MBq/mL. A graphical representation of the automated synthesis is shown in Fig. 1. HPLC analysis of the final product revealed a radiochemical purity (RCP) of 98.1 ± 0.6% (n = 5), meeting the defined specification of ≥ 95% applied also for other radiopharmaceuticals [10]. In addition, instant thin layer chromatography (iTLC) was performed, confirming the presence of ≤ 1% of free lutetium-177 and ≤ 2% of colloidal lutetium-177. All produced batches fulfilled the defined specifications of the different parameters tested (appearance, pH, radioactivity concentration, radionuclide identity, identity, radiochemical purity, peptide content, apparent specific activity, ethanol content, bacterial endotoxins and sterility) meeting the required acceptance criteria for release. The defined specifications and results of the masterbatches are described in Supplementary Table 1 and Table 1.

Fig. 1

Schematic illustration of the automated synthesis process of [177Lu]Lu-DOTA-MGS5

Table 1 Quality control results of the five masterbatches of [177Lu]Lu-DOTA-MGS5

Repeated HPLC and iTLC analysis was conducted to evaluate the stability of [177Lu]Lu-DOTA-MGS5 stored at room temperature for the time points of 1, 2, 4 and 24 h post preparation (p.p.). A high radiochemical purity with values ≥ 95% was confirmed for [177Lu]Lu-DOTA-MGS5 formulated in physiological saline and < 10% ethanol, containing also sodium ascorbate (150 mg) and calcium trisodium pentatate (~ 0.7 mg). In Table 2 the stability data are summarized for the five masterbatches produced. In Fig. 2 exemplary radiochromatograms of [177Lu]Lu-DOTA-MGS5 at 1, 2, 4 and 24 h p.p. are depicted. For the in-house use, a shelf life of 4 h p.p. was set.

Table 2 Stability of [177Lu]Lu-DOTA-MGS5 stored at room temperature for up to 24 h p.p.Fig. 2

Exemplary radiochromatograms of [177Lu]Lu-DOTA-MGS5 p.p. and at 1, 2, 4 and 24 h p.p.

Non-clinical pharmacology and toxicological dataSaturation binding and cell uptake studies

In the saturation binding studies performed on A431-CCK2R cells as source for the human CCK2R, a dissociation constant (Kd) of 5.25 ± 1.61 nM and a maximum number of binding sites (Bmax) of 1.14 ± 0.32 nM was calculated for [177Lu]Lu-DOTA-MGS5. In Fig. 3 an exemplary binding curve of [177Lu]Lu-DOTA-MGS5 is shown.

Fig. 3

Exemplary saturation binding curve of [177Lu]Lu-DOTA-MGS5 on A431-CCK2R cells

In cell uptake experiments using A431-CCK2R and AR42J cells, increasing cell uptake was observed, staring from values of 22.6 ± 6.2% in A431-CCK2R cells and 21.1 ± 2.8% in AR42J cells at 30 min post incubation, the values increased to 68.0 ± 3.0% in A431-CCK2R cells and 48.6 ± 2.2% in AR42J cells for the last time point of 4 h after incubation. The non-specific uptake in A431-mock cells was < 0.5% for all studied time points. Co-incubation with 1 µM pentagastrin in AR42J cells blocked the receptor-specific uptake resulting in a non-specific internalization of ≤ 1% for all studied time points. The cell uptake of [177Lu]Lu-DOTA-MGS5 over time in both cell lines is shown in Fig. 4.

Fig. 4

Cell uptake of [177Lu]Lu-DOTA-MGS5 in (a) A431-CCK2R, (b) AR42J cells. Uptake in A431-mock cells and blocking with pentagastrin in AR24J cells is additionally shown (dashed lines). The results are shown as mean ± SD based on three independent studies

Specificity of receptor interaction

The specificity of [177Lu]Lu-DOTA-MGS5 for CCK2R was confirmed using CHO-CCK1R and CHO-CCK2R cells. [177Lu]Lu-DOTA-sCCK8 with affinity for both CCK1R and CCK2R was used as control. A very low uptake of 0.06 ± 0.01% and 0.09 ± 0.04%, at 1 and 2 h, respectively, were observed for [177Lu]Lu-DOTA-MGS5 in CHO-CCK1R cells. Similar values of ~ 0.1% were observed also when co-incubating the cells with 1 µM sCCK8, confirming that no receptor-mediated uptake of [177Lu]Lu-DOTA-MGS5 occurred in CHO-CCKR1 cells. In CHO-CCK2R cells, the cellular uptake of 5.9 ± 0.6% and 8.1 ± 0.7% at 1 and 2 h was reduced to uptake values of less than 0.1% at 1 and 2 h by co-incubation with 1 µM sCCK8, confirming the receptor-specificity of [177Lu]Lu-DOTA-MGS5 for CCK2R. In comparative cell internalization studies with [177Lu]Lu-DOTA-sCCK8, a receptor-mediated cell uptake could be confirmed in both cell lines. Uptake values of 0.8 ± 0.3% and 1.3 ± 0.6% at 1 and 2 h were found for CHO-CCK1R cells. A somewhat higher uptake was found in CHO-CCK2R cells, with values of 2.5 ± 0.4% and 3.3 ± 0.9%, at 1 and 2 h, respectively. Co-incubation with 1 µM sCCK8 resulted in abolishment of the receptor-mediated uptake with non-specific uptake values below 0.1% in both cell lines. The cell uptake of [177Lu]Lu-DOTA-MGS5 and [177Lu]Lu-DOTA-sCCK8 in both cell lines is shown in Fig. 5.

Fig. 5 

Cell uptake of [177Lu]Lu-DOTA-MGS5 and [177Lu]Lu-DOTA-sCCK8 in absence and presence of 1 µM sCCK8 for receptor blocking: (a) CHO-CCK1R cells, (b) CHO-CCK2R cells

Biodistribution studies in BALB/c mice

Using peptide derivatives with single substitution with 1-Nal, single substitution with (N-Me)Nle and combined substitution with (N-Me)Nle and 1-Nal the influence of the applied amino acid substitutions on the tumour targeting properties was evaluated. To allow for a direct comparison with biodistribution data available for [111In]In-DOTA-MG11 without substitutions and obtained using the same conditions, this study was performed with the 111In-labelled peptide derivatives. The tumour uptake of the three radiopeptides in A431-CCK2R xenografts at 4 h post injection (p.i.) is shown in Fig. 6 (a) together with the uptake in kidney and stomach. A tumour uptake of 1.2 ± 0.2%IA/g was found for [111In]In-DOTA-[1Nal8]MG11, [11] whereas the uptake of [111In]In-DOTA-[(N-Me)Nle6]MG11 was 12.3 ± 4.2%IA/g. Both values were significantly lower (P < 0.0001 and P = 0.004, respectively) than the uptake of [111In]In-DOTA-MGS5 of 23.5 ± 1.3%IA/g previously reported [6], confirming a synergistic effect of both substitutions in improving tumour targeting. The stomach uptake of [111In]In-DOTA-MGS5 was also increased (8.2 ± 2.4%IA/g versus 1.2 ± 0.3%IA/g of [111In]In-DOTA-[1Nal8]MG11 and 2.9 ± 0.6%IA/g of [111In]In-DOTA-[(N-Me)Nle6]MG11; P = 0.004 and P = 0.014, respectively). Also the kidney uptake of [111In]In-DOTA-MGS5 (3.9 ± 0.5%IA/g) was higher when compared to [111In]In-DOTA-[1Nal8]MG11 (1.1 ± 0.1%IA/g, P = 0.0001) and [111In]In-DOTA-[(N-Me)Phe6]MG11 (3.0 ± 0.3%IA/g; P = 0.039). Still, tumour-to-organ ratios of [111In]In-DOTA-MGS5 with values of 3.1 ± 1.1 for stomach and 6.1 ± 0.6 for kidney were favourable.

Fig. 6

Uptake in kidney, stomach and A431-CCK2R xenografts at 4 h p.i. of (a): [111In]In-DOTA-[1Nal8]MG11 (n = 3), [111In]In-DOTA-[(N-Me)Nle6]MG11 (n = 3) and [111In]In-DOTA-MGS5 (n = 4) (b): [177Lu]Lu-DOTA-MGS5 prepared using manual labelling (n = 5) and cassette-based automated synthesis (n = 3)

Using the same mouse tumour model, the biodistribution of [177Lu]Lu-DOTA-MGS5 prepared using the cassette-based synthesis process was compared with standard manual labelling. The results obtained from this study are shown in Fig. 6 and allow the comparison of the tumour uptake with other MG analogues recently studied in clinical trials [12]. A comparable uptake in A431-CCK2R xenografts was observed, with values of 22.9 ± 4.7%IA/g found for [177Lu]Lu-DOTA-MGS5 (tumour weight: 436 ± 200 mg; n = 5) prepared by manual labelling and values of 31.9 ± 12.8%IA/g for [177Lu]Lu-DOTA-MGS5 (tumour weight: 185 ± 103 mg; n = 3) prepared using the automated synthesis process, at 4 h after injection. A higher variability in tumour uptake was found for A431-CCK2R xenografts of BALB/c nude mice injected with [177Lu]Lu-DOTA-MGS5 prepared using a cassette-based synthesis process, however no statistical significance was found between the two groups for tumour xenografts as well as all other tissues analysed.

Preclinical pharmacokinetics and dosimetry

In the dosimetry study performed for [177Lu]Lu-DOTA-MGS5 intravenously injected into A431-CCK2R xenografted female BALB/c nude mice, adequate tumour targeting was confirmed. A very high accumulation of radioactivity in A431-CCK2R xenografts and good retention of the radioactivity over time was observed. The initial tumour uptake of 68.1 ± 10.0%IA/g at 1 h p.i., decreased to values of 28.9 ± 7.2%IA/g and 12.6 ± 3.3%IA/g, at 24 h and 72 h p.i., respectively. The radioactivity remained detectable in the tumour also at 7 days p.i. (2.0 ± 0.5%IA/g). A rapid washout of the radioactivity was observed from the blood pool, with most organs showing relatively low radioactivity accumulation. The radioactivity in the blood pool decreased from 1.9 ± 0.4%IA/g at 1 h to 0.01 ± 0.002%IA/g at 24 h p.i., resulting in undetectable levels throughout the further course of the study. A low uptake of 3.7 ± 0.4%IA/g was observed in the kidneys 1 h p.i., decreasing to 2.0 ± 0.2%IA/g at 24 h and levels of < 1%IA/g for all later time points studied. The initial uptake in the receptor-positive stomach was 8.1 ± 0.9%IA/g at 1 h p.i., and gradually declined to 4.9 ± 0.4%IA/g, 3.4 ± 0.5%IA/g and 1.3 ± 0.1%IA/g at 24 h, 72 h, and 168 h p.i., respectively. The radioactivity accumulation and washout over time is shown in Fig. 7 for selected organs. In Supplementary Table 2, the uptake values over time are summarized for all dissected tissues and organs analysed. The tumour-to-organ ratios for the most relevant organs are shown in Supplementary Table 3. The absorbed dose found in different tissues and organs are shown in Supplementary Table 4.

Fig. 7

Time-dependent distribution of [177Lu]Lu-DOTA-MGS5 in A431-CCK2R xenografted BALB/c nude mice for selected tissues over time (n = 5, for each time point)

The time-activity curves were fitted to the biodistribution data with single-exponential functions with R2 > 0.7 (Fig. 8). The clearance from the tumour proceeded with a biological half-life of 18.2 h (12.3–27.1 h; 95% confidence interval). The blood clearance proceeded with a half-life of 2.8 h (0–7.6 h; 95% confidence interval).

Fig. 8

Time activity curves in A431-CCK2R xenografts, blood, stomach and kidney p.i. of [177Lu]Lu-DOTA-MGS5 in A431-CCK2R-xenografted BALB/c nude mice. Single-exponential curve fits are shown with 95% confidence intervals

Based on dosimetry extrapolation from mice to humans, the expected absorbed doses of [177Lu]Lu-DOTA-MGS5 in human organs were calculated and are shown in Table 3. An estimated radiation dose for stomach of 0.045 mGy/MBq for males and 0.051 mGy/MBq for females was calculated, taking into account the stomach wall as the source of radiation. A predicted radiation dose of 0.018 mGy/MBq for males and 0.022 mGy/MBq for females was found for kidneys. Assuming an injected radioactivity of 7,400 MBq for each treatment cycle with [177Lu]Lu-DOTA-MGS5, a stomach dose of ~ 354 mGy and kidney dose of ~ 148 mGy can be expected, resulting in a cumulative dose of 1.4 Gy for stomach and 0.6 Gy for kidneys after 4 therapeutic cycles.

Table 3 Expected absorbed dose of [177Lu]Lu-DOTA-MGS5 in humansToxicity study in rats

In a previously performed extended singe-dose toxicity study [13], the intravenous administration of DOTA-MGS5 at the three different dose levels of 0.1, 0.5, and 2.5 mg/kg body weight was well tolerated in Wistar rats. A NOAEL (no observed adverse effect level) of 0.4 mg/kg body weight was established for humans. For the first exploratory clinical trial with [177Lu]Lu-DOTA-MGS5, a starting dose of 100 µg was set. This dose remains well below the 1/100 NOAEL of 4 µg/kg. The performed study justifies the use of [177Lu]Lu-DOTA-MGS5 in a microdose trial without therapeutic intent, investigating the safety of administration and dosimetry aspects (microdose approach 1). When following the microdose approach 2, a maximum of five administrations with each dose ≤ 100 µg and a total cumulative dose of ≤ 500 µg would be acceptable. The performed non-clinical evaluation also forms the basis for a therapeutic trial using [177Lu]Lu-DOTA-MGS5 in patients with advanced cancer and limited therapeutic options [14].

First patient-specific clinical dosimetry

Planar and quantitative SPECT/CT imaging performed after administration of 1.5 GBq [177Lu]Lu-DOTA-MGS5 in a patient with extensive disease small cell lung cancer (ED-SCLC) revealed high uptake in tumour lesions. In Fig. 9 a fused coronal view of the full body SPECT/CT scan as well as axial slices at 24 h p.i. are given. Physiological accumulation was observed mainly in the urinary system and in the gastrointestinal tract (receptor-mediated uptake in stomach; intestinal activity at later time points). For all tumour lesions, clear uptake was confirmed, predominantly in the left-central primary tumour and cervical/mediastinal lesions, both adrenal glands metastases, as well as different bone and soft tissue lesions. The preliminary patient-specific dosimetry study showed a low risk of renal and bone marrow toxicity (absorbed doses 0.28 Gy/GBq and 0.022 Gy/GBq, respectively) and receptor-mediated uptake in stomach (0.42 Gy/GBq). In the dosimetric calculations performed on five well delineable lesions an absorbed dose of 12.5 ± 10.2 (1.2–28) Gy/GBq was found confirming the feasibility of PRRT.

Fig. 9

[177Lu]Lu-DOTA-MGS5 SPECT/CT scan at 24 h p.i. (a) fused coronal view of the full body scan. Dashed areas indicate the level of the axial slices at (b) left-central primary tumour and cervical/mediastinal lesions, as well as a left subpleural lesion, (c) both adrenal glands lesions, (d) right cervical vertebral metastasis and (e) soft tissue lesion in the gastric curvature together with physiological uptake in the gastric wall. The scale is set at 30% of the maximum activity concentration of 845 kBq/ml