Two MAD-based DOTA-carrying constructs were synthesized and demonstrated to bind specifically to CD206. The CD206 ligands were structurally similar, differing only in their molecular weights, 8.7 kDa and 22.6 kDa, respectively. Following labeling with [68Ga] and IV injection into Balb/c mice with and without CT26 tumors, biodistributions and TACs were generated. An in vitro plate assay where 50 nmol of [68Ga]MAD-8.7 or − 22.6 illustrated specific binding to surface-adhered CD206 and further illustrated the specificity by blocking with a 100-fold molar (5 µmol) excess of the labeling construct or the HMW agent, MAD-300. Representative MIPs (Fig. 2) show that the predominate sites of localization for either imaging agent in both control and tumor bearing mice were the livers and kidneys, known sites with large numbers of CD206 + Kupffer cells and mesangial cells, respectively. The liver, due to its relatively large size, localizes more imaging agent than other organs. These results are consistent with results obtained from mice injected with commercial Tc99m tilmanocept (Supplemental Fig. S3). Additionally, Fig. 2E–F shows axial views, illustrating tracer localization in tumors for both tracers.

In Fig. 3, the aggregate TAC data for the tumor bearing mice injected with the [68Ga]MAD-8.7 is representative of all time activity curves generated in these studies. Of note, beginning at the first measured time point taken ~ 10 min after injection, the mean SUV at all time points through 90-min generate time activity curves that approximate straight lines running parallel to the x-axis (time). This result has two implications. First, 10 min following injection, all or nearly all the injected dose of the imaging agent that could localize to a particular organ, has localized to that organ, further suggesting that the blood half-lives of these agents are much shorter than 10 min in mice. By comparison, Navidea has reported to the FDA that the blood half-lives of commercial Tc99m tilmanocept in rats and humans are approximately 15 min and 20 min, respectively. Second, once either the 8.7 kDa or 22.6 kDa imaging constructs have localized to an organ, they remain there stably and do not migrate to other sites. This is consistent with the known high affinities of similar MAD constructs to CD206 and their rapid internalization to endosomes [12].

One of the two primary goals of this study was to determine how the biodistribution of tilmanocept-like, DOTA derivatives would be altered by changing their molecular weights. Figure 4 shows biodistributions of the two tracers at the same molarity, in the blood, liver, kidneys, large intestine, and tumors observed after the 90-min imaging studies had been completed. In the livers and kidneys, there were no significant differences in the %ID/g localized to these organs for either MAD construct in either tumor bearing on non-bearing mice. This was as expected since there are no barriers or reduced barriers between the blood flow and the CD206 expressing cells (i.e., mesangial cells and Kupffer cells) in these organs. However, clear differences in localization of the imaging agents in the blood and in the large intestines and tumors were observed. The amounts of [68Ga]MAD-8.7 retained in the blood 90 min post injection were generally lower compared to [68Ga]MAD-22.6. In control mice without tumors, for [68Ga]MAD-8.7 and [68Ga]MAD-22.6, the %ID/g retained in the blood were 0.69 ± 0.08%ID/g and 2.04 ± 0.26%ID/g, respectively (p < 0.001). There were also non-significant differences in the amounts of the [68Ga]MAD constructs retained in the blood of tumor bearing mice. The reasons for these differences are currently not elucidated and will require further investigations to resolve but are probably related to other, non-CD206 mannose binding proteins found in blood, such as the mannose binding lectin [13], and to possible differences in binding affinities of different sized MAD constructs to these proteins.

It had been hypothesized that the smaller, 8.7 kDa construct would exit the blood flow and penetrate tissues, including tumors, more efficiently that the larger 22.6 kDa derivative. More efficient localization of smaller molecules relative to chemically similar, larger molecules is well established in the published literature [14, 15]. The results shown in Fig. 4 demonstrate that for both tumors and the large intestine, [68Ga]MAD-8.7 localized significantly more to these tissues compared to the larger 22.6 kDa construct. For the large intestine, 2.3 (non-tumor bearing, p < 0.0001) to 3.5 (tumor bearing, p < 0.0001) times more [68Ga]MAD-8.7 localized to the large intestines than [68Ga]MAD-22.6. In tumors, 7 times more [68Ga]MAD-8.7 localized to tumors than [68Ga]MAD-22.6 (p < 0.01). These results support the hypothesis that, by varying the sizes of MAD-based constructs, localization of 68Ga or other MAD payloads to tissues like tumors and the large intestine can be increased and optimized. These findings have important implications for designing MAD-based drug delivery vehicles targeting TAM directed cancer immunotherapies. The finding that [68Ga]MAD-8.7 localized more [68Ga]DOTA payload to the large intestines than [68Ga]MAD-22.6 may also have implications for designing MAD-based drug delivery vehicles targeting macrophages and dendritic cells involved in chronic inflammatory conditions of the large intestine or other organs [16,17,18].

Because of superior localizations of [68Ga]MAD-8.7 to large intestines and tumors, studies were performed to determine if competition with unlabeled MAD-8.7 or HMW-MAD could reduce liver localization without significantly reducing localizations to tumors and other potential target tissues as modeled by large intestines. In these studies, 250 μg of unlabeled MAD-8.7 or HMW-MAD were injected IV immediately prior to injection of 5 μg (0.57 nmol) of [68Ga]MAD-8.7. The resulting biodistributions of selected tissues are shown in Fig. 5. As was hypothesized, HMW-MAD significantly reduced the localization of [68Ga]MAD-8.7 to the liver by 46% (p < 0.0001) and did not significantly reduce localization to either the large intestines or tumors. These results have implications towards designing drug dosing protocols with a combination of a MAD-drug construct and an HMW-MAD for delivering small molecule therapeutics to TAMs and other sites of macrophage involved inflammation where off-target liver toxicity is a concern.

Also shown in Fig. 5 were results demonstrating that, similar to observations in the liver, competition with the HMW-MAD reduced retention of [68Ga]MAD-8.7 in the blood by 45% (p < 0.01). HMW-MAD also did not significantly alter the amount of [68Ga]MAD-8.7 that localized to the mesangial cells of the kidneys. This result was expected because HMW-MAD has an Mw of 300 kDa and is excluded from passing across glomeruli membranes due to its large size and, therefore, could not compete with [68Ga]MAD-8.7 for binding to CD206 on mesangial cells [19,20,21].

Contrary to the results returned with competitive blocking with the HMW-MAD, which were all consistent with expectations, competition with unlabeled MAD-8.7 (self-blocking) returned a series of unexpected results. It was expected that self-blocking would reduce the %ID/g retained in the blood similar to what was observed with HMW-MAD competition and that self-blocking would reduce localization of [68Ga]MAD-8.7 to the kidneys. The observed results were that self-blocking did not significantly alter the retention in blood or kidney localization of [68Ga]MAD-8.7. The observation in blood might indicate that [68Ga]MAD-8.7 is binding to an abundant entity in the blood with low affinity. It was also expected that self-blocking would reduce localization to the large intestines and tumors and to similar degrees. Instead, it was observed that self-blocking significantly reduced large intestine localization by 68% (p < 0.0001) but reduced tumor localization by a non-significant 39%. A full explanation of this observation will require further experimentation. However, self-blocking did, as expected, dramatically reduce liver localization of [68Ga]MAD-8.7 by 72% (p < 0.0001). This result might contribute to an explanation of why self-blocking did not reduce [68Ga]MAD-8.7 localization to the kidneys. Self-blocking, by dramatically reducing liver localization, may have permitted [68Ga]MAD-8.7 to stay in the blood circulation longer and allowing it greater opportunity to pass through and localize to the kidneys.

The results of these competitive blocking experiments permitted calculations of tumor to liver localization ratios (Fig. 6). Both self-blocking and HMW-MAD competition significantly increased tumor to liver ratios, but because self-blocking decreased liver localization to a greater extent than it reduced tumor localization, self-blocking increased the ratio more than did HMW-MAD competition. Combined with the results from large intestine localization, these results imply that for therapeutic indications for which liver toxicity is a concern, self-blocking of the liver may be preferred for TAM targeted cancer immunotherapies, while HMW-MAD competition may be preferred for indications related to treating inflammation at other sites such as the large intestine.

This report describes how the biodistributions of MAD-based diagnostic and therapeutic constructs may be beneficially altered by varying the molecular weights of both the MAD-based payload carrying constructs and unlabeled, payloadless MAD-based competitors. However, these results are limited in that neither the full molecular weight spectrum of MAD-payload and MAD-competitor constructs nor the full range of molar doses for the MAD-based constructs were explored. Furthermore, the reported studies revealed several unexpected results. Resolution of these issues will require additional investigations; however, with further investigation, it appears likely that MAD-based constructs and dosing protocols can be developed that optimize delivery of diagnostic imaging agents and/or therapeutic payloads to sites of macrophage involved inflammation while simultaneously minimizing delivery to off-target organs.