Sir

We read with interest the article by Koukourakis et al (2000a) describing the use of radiolabelled Caelyx® to study the accumulation of the polyethylene glycol (PEG)-liposomes in glioblastomas and metastatic brain tumours, and would like to take the opportunity to make some principal remarks regarding the applied radiolabelling procedure.

To facilitate imaging of the accumulation of the liposomes in the tumours, the authors labelled the liposomal formulation with technetium-99m-diethylenetriamine pentaacetic acid (99mTc-DTPA). This labelling method was described previously in more detail by the same group (Koukourakis et al, 1999). However, we have some major concerns about the labelling method and imaging protocol used in their studies (Koukourakis et al, 1999, 2000a, 2000b).

The applied labelling method is based on the assumption that adding 99mTc-DTPA to the PEG-liposomes (in the Caelyx® formulation) results in radiolabelled liposomes. This method, which has not been described in the literature before, is expected to yield an unstable radiolabelled product, since there is no driving force that will facilitate the hydrophilic 99mTc-DTPA to pass the lipid bilayer and entrap the radiolabel inside the liposome. Moreover, the described quality control method (Koukourakis et al, 1999, 2000a, 2000b) does not discriminate between free 99mTc-DTPA and liposome-associated 99mTc-activity. The chromatographic method applied can distinguish between reduced 99mTcO2 and unreduced 99mTcO4-, however, it is more essential to discriminate between liposome-associated 99mTc-activity and 99mTc-DTPA in the preparation. The proper way to examine this, would be to elute a sample of the radiolabelled liposomes on a gelpermeation column as described previously (Dams et al, 2000; Harrington et al, 2001). The (radiolabelled) liposomes will elute with the void volume in the early fractions, whereas the non-liposome-associated radiolabel (both as 99mTc-DTPA, 99mTcO2, and 99mTcO4-) will elute in later fractions.

Our attempts to reproduce the labelling method as described by Koukourakis et al (2000a) failed. Analysis of the labelling mixture by instant thin layer chromatography (ITLC) according to Koukourakis et al (2000a) showed the same results as described in their article, suggesting a labelling efficiency of approximately 80%. However, analysis on a Sephadex G25-column revealed that only 1% of the 99mTc activity was associated with the liposomes and the majority of the activity eluted in later fractions (Figure 1). This was confirmed by ITLC on silica strips in sodium citrate, a method to distinguish 99mTc-DTPA and 99mTc-liposomes described previously (Cao and Suresh, 2000).

Figure 1
figure 1

Elution profile of 99mTc-DTPA-Caelyx® (labelled according to Koukourakis et al, 1999) and 111In-oxinate-Caelyx® (labelled according to Laverman et al, 2001) on a 10 ml Sephadex G25 column. A sample of 100 μl was applied. Twenty consecutive fractions of 1 ml were eluted with 5% glucose and the activity of each fraction was counted in a gamma-counter. Note that the radiolabelled liposomes elute in fractions 3–5, whereas the free radiolabel elutes in fractions 6–14.

Full size image

Our second point of concern is the chosen time-point of imaging. The patients were imaged at 2 h after infusion of the liposomes (Koukourakis et al, 2000a). Harrington et al (2001) showed that clear visualization of solid tumours with 111In-labelled long-circulating PEG-liposomes was achieved not earlier than 48–72 h after injection, due to the high blood background signal at earlier time-points. These findings are in line with our clinical study in patients with infectious or inflammatory disease, imaged with similar 99mTc-PEG-liposomes, labelled with 99mTc-hexamethylpropylene-amine-oxime (Dams et al, 2000). Therefore, it appears that the images showing liposome uptake in the tumours (Koukourakis et al, 2000a) represent the accumulation of 99mTc-DTPA, rather than uptake of the radiolabelled Caelyx. 99mTc-DTPA was used in the past for diagnosis of a disrupted blood–brain barrier in brain tumours, before the CT and MRI era (Hauser et al, 1970). Nowadays, 99mTc-DTPA is a well-known agent for evaluation of renal function. It is cleared rapidly and efficiently from the circulation by glomerular filtration. This might explain, for liposomes, the unusual, rapid accumulation of activity in the tumours. In our opinion, the authors should have performed, and shown – at least in some patients – a control 99mTc-DTPA scan, to rule out that the liposome scan represents free 99mTc-DTPA instead of radiolabelled Caelyx®.

A better approach to label Caelyx® would be the labelling with indium-111-oxine (111In-oxine). This easy method will yield radiolabelled liposomes with good radiochemical yield (>80%) and good in vivo stability (Laverman et al, 2001). An additional advantage is the longer physical half-life of 111In, which enables the acquisition of delayed images and thus better visualization of the tumours (Harrington et al, 2001).

In summary, scintigraphic techniques are very helpful in investigating the in vivo distribution of (new) pharmaceuticals, but should only be performed using well-established labelling techniques and quality control methods. The results presented by Koukourakis et al (1999, 2000a, 2000b) should therefore be interpreted with caution.