Much of my thinking regarding ligands/chelators and GDD are based on the early work that MR contrast agent companies, especially Schering, published on various Gd + ligand combinations and on the importance of thermodynamic and kinetic stability of Gd chelates. This same exercise can also be worked out with all other heavy metals and other cations such as Calcium. So I fundamentally am fond of ligands that have been explored as creating stable GBCA MR agents, and not so fond of agents that have either not been tested as ligands, or have shown poor thermodynamic stability. Stable chelates with high thermodynamic stability, which can be created in vivo, are therefore good chelators. Chelators that have low thermodynamic stability with Gd are poor chelators for Gd. Using this calculation DTPA is a good chelator, and EDTA is a poor chelator. Early work by Schering had also shown this.
The problem with ligands with low stability is that they have the propensity to pick up Gd, and then drop it off somewhere else in the body, before the kidneys can excrete it. DTPA has a far less likelihood of doing this, and in fact has a thermodynamic stability 300,000 times greater than EDTA (that is not a misprint).
We have shown that intravenously DTPA can increase substantially the urine content of Gd in urine. It turns out that other of the ligands used in GBCAs may also be able to recreate Gd-ligand structures in vivo from Ca- or Zn- based ligands. An oral version of DTPA is also in development. Additionally HOPO has been described in the literature as an oral agent with very high affinity for Gd. Ultimately oral chelators may be the best drugs to be used for GDD because of convenience, no pain, and also perhaps the lower concentration but constant presence of chelator in the body may serve to facilitate the constant removal of Gd from tightly bound reservoirs, like bone.
It remains unknown why DTPA also increases elimination of macrocyclic GBCAs, as the macrocycles aught to be intact, and hence should not allow their Gd to combine with the introduced DTPA. We have postulated a couple of explanations: 1. the ligand DTPA is levering fully intact GBCA out of tissues, and/or 2. the host immune system has disrupted the macrocycle so that DTPA can combine with the exposed Gd.
The problem is that all the data on Gd and stability is drawn from in vitro test tube experiments, 'normal' animal experiments, and 'normal' human experiments. The reality is, at the present time we have no idea what GDD patients are doing with these GBCAs. We have to target with experiments GDD patients, and if possible create GDD animals. The closest analog to GDD is auto-immune disease. The essential study is to examine ex vivo peripheral blood mononuclear cells (PBMCs) to compare GDD patients with Gadolinium Storage Condition (GSC, 'normals') to see how their reactions differ in the presence of all the different GBCAs.
The Flare reaction reflects the host reacting to the remobilization of Gd in the body. As all the macrocyclic agents can cause a Flare reaction, and the onset can be instantaneous
after administration of GBCA, it is clear the GDD is determined by two essential factors: 1. host reaction and 2. Gd presence in the body in any of its structural configurations, that is, as: i) intact chelate or as one of its other common speciations: as ii) Gd combined with phosphates or carbonates [forming a salt] or iii) Gd in combination with proteinaceous macromolecules. These are the 3 primary structures.
The Flare is usually intense in patients with early stage disease (< 3 months) as the immune system is still on high alert from the recent GBCA administration. It is not clear if macrocyclic agents may create especially strong Flare responses. It may be imprtant to investigate whether initial chelations should be performed with lower volumes of chelators.
Can Ligands themselves cause disease?
My thinking to the present time is probably not. There is at least one study that looked at animal models that suggested toxicity is possible with DTPA as an actinide metal chelator. What is critical to do in all studies is to analyze the Methodological design. The best studies to evaluate GDD patients, are studies performed directly on GDD patients with conventional GBCAs at standard doses. Many animal studies, create circumstances that do not reflect human conditions: 1) the dose (of whatever) may be exceedingly high, 2. they may be looking at the wrong molecule: studying Gd-Cl (which essentially immediately comes apart) or Gd-citrate (also a poorly bound molecule), 3. repetition may be too high of drug administration, and 4. they may not be looking at the actual clinical setting, to name just 4. The Methodology must be carefully analyzed before drawing any conclusions. This is not to say that even heavily biased studies do not provide useful information, they certainly can.
Why I think ligand is not the problem?
The first point is to recognize the importance of a good ligand. Again using my initial statements in this blog: based on its development as an MR contrast agent and its high thermodynamic stability, DTPA is a good ligand/chelator. EDTA however, based on the fact that it was looked at as a chelator but never developed as a GBCA, and its low thermodynamic stability, is a relatively poor chelator for Gd. I am also very suspect of using other random chelators such as desferoxamine to chelate Gd. I do think that poor ligands may be a problem and may redeposit Gd in the body. So the question is:
Why I think a 'good' ligand is not the problem in GDD.
1. Magnevist (essentially Gd-DTPA) has 140 million + doses administered to humans, and never has an issue with free ligand been described with this agent.
2. The extensive research with NSF, and Magnevist the second most common agent to have caused it. In none of the studies, none of the histological or electron microscopic studies, has the ligand ever been considered the culprit.
3. the extensive animal studies (see Siebert paper) have all described Gd in the tissues and hence Gd as the cause for the skin, etc changes. Never has the ligand been isolated in the tissues and considered responsible.
4. the urine elimination of GBCAs and their calculated decay rates in the body with radiolabelled studies (Eric Lancelot papers) free ligand, as seen with Omniscan, eliminated at the same fast rate as fully intact GBCAs like Dotarem, but the Gd from these linear agents showed much slower decay (reflecting retention).
5. the ligand is the part of the chelate that is designed to make it safe - it is the safe part.
6. Fully intact macrocyclic GBCAs cause GDD.
7. Symptoms from the toxicity of heavy metals are extremely similar to those of GDD, consistent with the heavy metal being the cause of symptoms: that is lead toxicity is very similar to GDD.
Emphasis of this blog:
A) using a good ligand is critical with GDD (thermodynamic stability with Gd, and already extensively tested as stable with Gd - essentially the ligands of all the GBCAs). The latter a personal bias on my part.
B) Research is very much needed to moderate the Flare reaction during chelation.
C) If a 'good' ligand/chelator is used it is almost certainly never the primary problem with GDD.