Radiolabeling of liposomes and polymeric micelles with PET-isotopes

This thesis is divided into three separate chapters that can be read independently. Chapter 1 is a general introduction, touching upon liposomes and polymeric micelles and radiolabeling with 18F and 64Cu. Chapter 2 and 3 address two separate research projects, each described below. A complete reference list is compiled in the end, immediately after the three chapters.

This is followed by the supplementary information, divided into appropriate sections. Finally, the two first‐authored manuscripts are attached as appendices. Chapter 1. The field of nanoparticulate drug delivery has been hailed as a revolution in modern therapeutics, especially in chemotherapy. A major reason is the ability of nanoparticles to accumulate in tumor tissue.

Liposomes are the classic nanoparticle, consisting of a lipid membrane with an aqueous core. Polymeric micelles are made from amphiphilic detergent‐like copolymers, that self‐assemble in water. Therapy with nanoparticles is hampered by often poor tumor accumulation, combined with massive uptake by macrophages in the liver and spleen.

For this reason, visualizing nanoparticle pharmacokinetics in‐vivo is a valuable tool in the on‐going research. Such visualization can be done by labeling with radio isotopes. Isotopes that emit positrons (PET‐isotopes) can be detected by PET (positron emission tomography) technology, an accurate technique that has gained popularity in recent years.

PET‐isotopes of interest include 18F and 64Cu. In addition to being a research tool, radiolabeled nanoparticles hold promise as a radiopharmaceutical in themselves, as a means of imaging tumor tissue, aiding in diagnosis and surgery. Chapter 2. A method for labeling liposomes with 18F (97% positron decay, T½ = 110 min) was investigated. 18F is widely available, but is hampered by a short half‐life only allowing up to 8 hours scans. 18F must be covalently attached to components of the liposome.

By binding to a lipid, it can be stably lodged in the membrane. A glycerolipid and a cholesteryl ether were synthesized with free primary alcohols and a series of their sulphonates (Ms, Ts, Tf) were prepared. [18F]Radiofluorination of these substrates was performed on fully automated equipment using a classic Kryptofix222‐ mediated procedure in DMSO.

Yields were poor, 3‐17% depending on conditions. The [18F]fluorinated probes were purified in‐situ on SEP‐Paks. The cholesteryl ether mesylate performed best. This substrate was radiolabeled and formulated in long‐circulating liposomes by drying the probe and the lipids together, followed by hydration by magnetic stirring.

The liposomes were extruded through 100 nm filter on fully automated equipment. Animal studies were done in tumor‐bearing mice, and PET‐scans were performed over 8 hours. Clear tumor uptake, as well as hepatic and splenic uptake, was observed, corresponding to expected liposomal pharmacokinetics. Tumor uptake was quantifiable (tumor‐tomuscle ratio at 8 h: 2.20), showing that the maximum scan duration with 18F is sufficient for visualizing tumor tissue.

Because of the low [18F]radiofluorination yields obtained, we investigated ways of labeling lipophilic substrates in nonpolar solvents. This involved the transfer of [18]HF gas from a solution of concentrated sulphuric acid into a receiving vial containing the substrate in toluene. A phosphazene base was present to bind [18]HF and mediate fluorination.

This procedure made it possible to fluorinate highly lipophilic substrates in 71% yields. Chapter 3. Radiolabeling of polymeric micelles with 64Cu (18% positron decay, T½ = 12.7 h) was investigated. 64Cu allows longer scans (up to 48 hours), which mirrors the duration of nanoparticle pharmacokinetics.

It is a metal and must be attached to polymeric micelles by covalently conjugated chelators. DOTA and CB‐TE2A are two such chelators, but DOTA is widely believed to be unstable in‐vivo. DOTA and CB‐TE2A were conjugated to triblock polymeric micelles in the shellregion. Here, they were thought to be shielded by the outer PEG‐layer.

The micelles were crosslinked in their coumarin-containing cores by exposure to UV light. Subsequently, the micelles were labeled with 64Cu, followed by removal of unspecifically bound 64Cu by EDTA. Good labeling efficiency was achieved with both chelators (40‐70%). Some of the prepared micelles were found to exhibit gross instabilities, especially with raised temperature, which prevented their in-vivo use.

Other micelles were stable and were investigated in xenographted mice. These micelles were 20‐45 nm. They showed good tumor uptake (4‐5 %ID/g, 48h) and limited uptake in liver (5‐7 %ID/g, 48h) and spleen (3‐6 %ID/g, 48h). It was concluded that there did not seem to be a significant difference between DOTA and CB‐TE2A in‐vivo.

In addition, crosslinked micelles (with 64Cu bound to CB‐TE2A) were compared with non‐crosslinked micelles. To our surprise, we found that the non‐crosslinked micelles exhibited good stability in circulation and obtained a biodistribution very similar to the crosslinked micelles.