The effects of three clinically used antifolates [methotrexate, raltitrexed, and pemetrexed (65), Figure 1] were investigated in combination with a ^99mTc(CO)₃-labeled folate conjugate with the aim to increase the tumor uptake of the radiofolate (62). This study design was based on the observation that incubation of cancer cells with antifolates resulted in an increased uptake of radiofolates in vitro (62, 66).

The in vivo data showed that the accumulation of the radiofolate in KB tumor xenografts remained unaffected but a significantly reduced uptake of radioactivity was found in the kidneys. This resulted in clearly improved tumor-to-kidney ratios, with the best results obtained with pemetrexed (Figure 2) (62, 67).

The same effect was successfully reproduced in different tumor mouse models (KB, IGROV-1, and SKOV-3 tumors in nude mice and M109 tumors in Balb/c mice) and variable folate radioconjugates [^99mTc(CO)₃-folate, ^99mTc-EC20, ¹¹¹In-DTPA-folate, ¹¹¹In/¹⁷⁷Lu-DOTA-click-folate, ⁶⁷Ga-DOTA-Bz-folate (⁶⁷Ga-EC0800), ⁶⁸Ga-NODAGA-folate] (55, 62, 63, 67–69). The injection protocol and the related SPECT/CT images using ¹⁷⁷Lu-EC0800 are shown in Figure 3. The underlying mechanism of this effect is not yet clear but a topic of current investigations in our laboratories. The combination of folate radioconjugates and pemetrexed is, however, appealing in view of a therapeutic application. In this respect, we hypothesized that pemetrexed would have a dual role if it was combined with folate-based radionuclide therapy. On one hand it would reduce undesired renal uptake of radiofolates and on the other hand it may have an effect as a chemotherapeutic (70, 71) or radiosensitizing agent (72–74) on the tumor tissue and thus enhance the therapeutic efficacy on the tumors.

Increasing the serum half-life of pharmaceuticals by serum protein binding may be a measure to improve the pharmacokinetic properties of otherwise rapidly cleared molecules (75). Based on the results which are reported in the literature with antibody fragments that exhibited a non-covalent association with serum proteins (76, 77), it was hypothesized that albumin-binding properties would improve the tissue distribution of radiofolates as well. Therefore, a DOTA-folate conjugate was developed with an integrated small-molecular weight albumin-binding entity which was previously identified from a DNA-encoded chemical library based on the lead structure 4-(p-iodophenyl)butyric acid (78). This novel DOTA-folate conjugate (cm09, Figure 4B) was fully evaluated in its ¹⁷⁷Lu-labeled version and compared with the data obtained with a conventional ¹⁷⁷Lu-labeled DOTA-folate conjugate (¹⁷⁷Lu-EC0800, Figure 4A) (64).

The in vitro cell uptake properties of ¹⁷⁷Lu-cm09 were comparable to those of other folate radioconjugates (64). Different was however, its feature of binding to plasma proteins as demonstrated by an in vitro assay (64). In vivo biodistribution studies were performed with ¹⁷⁷Lu-cm09 and ¹⁷⁷Lu-EC0800 in athymic nude mice bearing KB tumor xenografts (Figure 5). The enhanced blood circulation time of ¹⁷⁷Lu-cm09 compared with ¹⁷⁷Lu-EC0800 resulted in a tumor accumulation of ¹⁷⁷Lu-cm09 which was ∼2.5-fold higher (∼18% ID/g, 4 h p.i.) than the uptake of ¹⁷⁷Lu-EC0800 (∼7.5% ID/g, 4 h p.i.) (64). On the other hand, renal retention of ¹⁷⁷Lu-cm09 was relatively low (∼28% ID/g, 4 h p.i.) compared to other folate conjugates that lack an albumin-binding entity such as ¹⁷⁷Lu-EC0800 (>70% ID/g, 4 h p.i.) (64). These findings resulted in a sevenfold improved tumor-to-kidney ratio (∼0.7, 4 h p.i.) for ¹⁷⁷Lu-cm09 compared to the ratio obtained with ¹⁷⁷Lu-EC0800 (∼0.1, 4 h p.i.). The excellent in vivo properties of ¹⁷⁷Lu-cm09 opened new perspectives for the development of folate-based radionuclide therapy.

To further improve the tumor-to-kidney ratio of the long-circulating ¹⁷⁷Lu-cm09, biodistribution studies were performed with ¹⁷⁷Lu-cm09 and pemetrexed using KB tumor-bearing mice. Pemetrexed was administered according to the same application protocol as it was previously employed with conventional folate radioconjugates (55, 63, 67). Independent on whether or not pemetrexed was pre-injected, the tumor uptake was about ∼18% ID/g 4 h after injection. In contrast, reduced renal accumulation (16.81 ± 2.25% ID/g) of radioactivity was found 4 h after injection of ¹⁷⁷Lu-cm09 if it was combined with pemetrexed compared to the amount of accumulated ¹⁷⁷Lu-cm09 if it was applied as a single agent (28.05 ± 1.35% ID/g, 4 h p.i.) (64). Thus, pre-injected pemetrexed increased the tumor-to-kidney ratio (1.07 ± 0.25, 4 h p.i) compared to control values at the same time after injection of ¹⁷⁷Lu-cm09 (0.65 ± 0.07, 4 h p.i.). However, at later time points after injection of ¹⁷⁷Lu-cm09 the kidney reducing effect of pemetrexed was not observed anymore (64). Most probably, these findings were a result of the fact that pemetrexed was more quickly cleared from the blood circulation than ¹⁷⁷Lu-cm09. Therefore, the pemetrexed related effect to reduce the kidney uptake of ¹⁷⁷Lu-cm09 was not maintained for the fraction of ¹⁷⁷Lu-cm09 which was still in the blood circulation when pemetrexed was already excreted (64).

In a recent study the combined application of ¹⁷⁷Lu-EC0800 (Figure 3A) and pemetrexed (Figure 1D) was investigated using KB tumor-bearing nude mice (79). Four groups of six athymic nude mice with KB tumor xenografts received only saline (group A), pemetrexed at a high dose (2 × 0.8 mg, group B), ¹⁷⁷Lu-EC0800 (1 × 20 MBq), and pemetrexed at a low dose (1 × 0.4 mg) to reduce renal uptake (group C) or ¹⁷⁷Lu-EC0800 (1 × 20 MBq) and a high dose of pemetrexed (2 × 0.8 mg, group D). Application of even a high dose of pemetrexed alone had almost no effect on tumor growth in mice of group B, compared with control animals of group A (Figure 6A). However, the application of ¹⁷⁷Lu-EC0800 resulted in a tumor growth delay (group C) which was increased if it was combined with a high dose of pemetrexed (group D) (Figure 6A). The therapeutic effect of ¹⁷⁷Lu-EC0800 was also reflected by the increased average survival time of mice (group C: 30 days) compared to control mice (group A: 20 days) and mice which received only pemetrexed (group B: 24.5 days) (Figure 6B) (79). However, the clearly most favorable therapy protocol was the combination of ¹⁷⁷Lu-EC0800 with a high dose of pemetrexed (group D) which resulted in an average survival time of 35 days (Figure 6B) (79).

These findings indicated a radiosensitizing effect of pemetrexed as it was previously reported in other in vitro and in vivo studies that combined pemetrexed with external radiation therapy (72, 73, 80, 81). An additional study was conducted with non-tumor-bearing mice over a time period of several months to investigate the potential impairment of the kidneys. It was demonstrated that pemetrexed was able to protect kidneys from radio-nephrotoxicity (79). Further and more detailed studies will be warranted to investigate the utility of this combination in view of a potential clinical application.

The first published therapy experiment with a folate radioconjugate was performed with KB tumor-bearing nude mice using ¹⁷⁷Lu-cm09 (64). Groups of five mice each were treated with only saline (group A) or with the unlabeled DOTA-folate cm09 (group B). One group received ¹⁷⁷Lu-cm09 in a single injection of 20 MBq (group C), another group received two injections of 10 MBq each (group D) and the last group received three injections of 7 MBq each (group E) (Figure 7A). The individual body weight and tumor volume was measured every other day. The results showed a constant tumor growth in mice of groups A and B which did not receive radioactivity (Figure 7B). In all mice of groups C to E tumor growth was clearly delayed with the best effect achieved in mice of group C which had been injected with the whole amount of 20 MBq ¹⁷⁷Lu-cm09 in one single injection (Figure 7B). The average survival time of control mice of group A and group B was 27 and 24 days, respectively, whereas an almost doubled survival time was observed for mice of groups D and E (48 and 46 days, respectively). In mice which received 20 MBq ¹⁷⁷Lu-cm09 in one single injection the tumors disappeared completely in four of five mice and hence these four mice were still alive at the end of the study which prevented the determination of an average survival time (64).

By using the same DOTA-folate conjugate (cm09, Figure 4B) as previously employed with ¹⁷⁷Lu (64) a pilot therapy study was performed with therapeutic terbium radioisotopes (82). ¹⁶¹Tb decays by the emission of β-particles and provides similar decay properties to ¹⁷⁷Lu (Table 1) (82, 83). ¹⁴⁹Tb decays by emission of short-ranged α-particles with a half-life of 4.12 h (Table 1) (82). Since the availability of ¹⁴⁹Tb was limiting in this study the experiments were performed with only a small number of mice bearing KB tumor xenografts.

In the α-therapy study, three mice received only saline (group A) and three mice received twice an injection of 1.1 and 1.3 MBq of ¹⁴⁹Tb-cm09, respectively (group B). In two of the treated mice the tumor growth was clearly delayed and in one mouse the tumors disappeared completely (Figure 8A). Compared to untreated control mice of group A, the average survival time was significantly prolonged (p < 0.05) in mice of group B. The β⁻-therapy study was carried out with five mice which received only saline (group A) and another five mice which received 11 MBq of ¹⁶¹Tb-cm09 (group B). The overall survival time of control mice was 28 days. In four of the five treated mice of group B, KB tumors disappeared completely and hence these mice were still alive at day 56 when the study was terminated (Figure 8B).

The results obtained with ¹⁷⁷Lu-cm09 and ^149/161Tb-cm09 indicated the potential of FR-targeted radionuclide therapy by using folic acid as a targeting ligand. Although the therapy regimes reported in these studies were well tolerated by the test animals, there is a potential risk of damage to the kidneys by particle-radiation. Therefore, kidney function was monitored in a preliminary study over 6 months in non-tumor-bearing mice which received ¹⁷⁷Lu-cm09 (20 MBq/mouse) (unpublished data). Blood plasma parameters were analyzed and quantitative SPECT experiments using ^99mTc-DMSA were performed to estimate potential impairment of kidney function as previously reported (84). Although signs of radio-nephrotoxicity were not observed, more recent investigations indicate long-term impairment of the kidney function if higher amounts of folate radioconjugates are administered. Hence, further preclinical studies with larger cohorts of mice will be clearly necessary to provide more detailed information about the suitability of this therapy concept.

In terms of radionuclide therapy the approach of using liposome- or nanoparticle-based carriers for passive or targeted delivery of particle-emitting radionuclides to cancer cells is scarcely reported in the literature but has recently been reviewed by Sofou (85). It could be in particular attractive for radionuclides that are inappropriate for direct complexation by conventional chelation concepts (86). An example that made use of folic acid decorated liposomes as carriers for the α-emitting radionuclides radium (²²³Ra) and actinium (²²⁵Ac) has been reported by Larsen and Co-workers (87). The radioactive liposome formulations possessed binding properties to FR-expressing cancer cells in vitro and were stable in serum with only low release of radionuclides. Hence, this or similar approaches may have the potential for an in vivo translation in the future.

Folate radioconjugates were found to be very stable in PBS and blood plasma at 37°C with only marginal release of radiometals over time (56). ¹⁷⁷Lu-cm09 showed a particularly high in vitro stability with >99% intact product over 6 days in human plasma (64). Little is known about potential radiolysis of folate radioconjugates (64). However, since there are studies reporting on the degradation of folate molecules exposed to high temperatures, ultraviolet light, and oxygen (88–90) it is likely that radiolytic processes would occur in highly active formulations of radiofolates. In view of a clinical application of therapeutic folate radioconjugates extensive investigations will be required with respect to the stabilization of radioconjugates by using quenchers such as for instance ethanol and ascorbic acid.

Since folic acid and folates are essential nutritions for the human organism and can be administered by supplements in mg-amounts, folate derivatives are expected to be well tolerated even if they were applied for radiotherapeutic purposes, where the injected amount would not exceed about 200 μg. Nevertheless, the high uptake of folate radioconjugates in the kidneys is of concern in terms of long-term radionephropathy caused by particle-radiation if the folate conjugates were used for radionuclide therapy as mentioned above.