The first ( SAINT-1 ) trial, using NXY-059 to treat acute ischemic stroke, worked. However, the second ( SAINT-2 ) trial failed. Our proposals concerning the failure of SAINT-II are published in the journal Stroke as "SAINT-1 Worked, But the Neuroprotectant is not NXY-059"
Basically, we suggest that the positive results of SAINT-1 occurred because of its well-known pharmacologically-active hydrolysis products, NtBHA and MNP. However, we hypothesize, these were below the level of pharmacologic activity in SAINT-II. We also conjecture about how this might have occurred. E.g., a delay in the SAINT-I trial (allowing hydrolysis products to build up in storage) or differential application of AZ's antihydrolysis technology. As the references show, this was not exactly a new idea.Some authors of the SAINT study take issue.
Stated simply, without directly-questioning to the basic proposal itself, Lees et al state that the SAINT trials were not delayed. Likewise, they assert NXY-059 was stabilized against hydrolysis identically in both SAINT-1 and SAINT-2.
Our proposals were clearly conjecture and for the sake of example, though well-based upon the evidence. E.g., SEC filings from Centaur Pharmaceuticals ( e.g.,CENTAUR PHARMACEUTICALS INC, Form:10-K405, 4/2/2001.) clearly imply the the SAINT trials were delayed for at least a year. Another SEC filing from Renovis Pharmaceuticals notes a change in the manufacturer of NXY-059 ( from Renovis to AstraZeneca ) occurring only a few months before the initiation of the SAINT trials.
NXY-059 per se is likely a dead issue, except possibly for prevention of hemorrhagic transformation post-tPA, which it more than halved in SAINT-I. However, if SAINT-I really worked, then neuroprotectants are possible and provable by current methodology.
Other objections are spelled out in Green's review, from an ex-AZ researcher concerned with NXY-059's development. We consider these in detail here.
Uric Acid: Neuroprotective or Neurotoxic speculates on why neuroprotection trials universally fail in humans after successful preclinical studies. Basically, humans have a powerful extracellular antioxidant and antiexcitoxic neuroprotectant missing in the animal models, uric acid. Uric acid is also a messenger in (e.g.) tissue injury. This leaves little "therapeutic room" for analogous extracellular neuroprotective agents like NXY-059.
E.g., normal plasma uric acid levels range from 200-400 micromolar. The target NXY-059 in the SAINT studies was 150 micromolar, with a treatment plateau in rodents of 30 micromolar free drug. Further, NXY-059 is a less potent antioxidant than urate. Uric acid is also antiexcitotoxic. This may be why animal trials with antiexcitotoxic agents also fail human trials. So you-all stop blaming each other for messing up, respectively, the human or preclinical trials. Everybody is right-- it is just that the animal models are unrepresentative.
Simply-stated, the low degree of hydrolysis necessary, the numerous potential variables, and the instability of NXY-059 mean normally insignificant and perhaps cumulative variations in synthesis, storage, or administration might account for the disparate results in SAINT-1 and SAINT-II. This is important because a successful outcome in SAINT-I means that neuroprotection in human stroke is possible and provable by current methods.
After very promising results in the first clinical trial (Saint-1) using NXY-059 for the amelioration of stroke-induced injury and disability, the second NXY-059 trial (Saint-2) failed.
Having ruled out plausible alternatives, the NEJM paper tentatively attributes the positive results in the SAINT-1 trial to "chance". Here "chance" for the SAINT-1 Modified Rankin data means about one chance in 25 (p=.038).
Moreover, a parallel trial, looking at prevention of hemorrhagic transformation secondary to the clot-buster tPA, gave even more striking results in Saint-1, with a p-value of .005, ie., one chance in 200. Further, while the subjective results using the Modified Rankin Scale might be open to question, the tPA results were based upon relatively secure and objective brain-scan evidence. This makes "chance" even more unlikely as an explaination for the positive SAINT-1 results.
However, if it is not "play of chance" or some systemic error, then what is it?
The simplest answer is that the SAINT-1 trial worked exactly as hoped and expected. However, the neuroprotectant in SAINT-1 was not NXY-059, but some unknown agent subsequently absent in SAINT-2.
So, what was this unknown neuroprotectant? Impossible to say for sure. However, there is intriguing precedent-- PBN derivatives such as NXY-059 (PBN-disulfonate) are somewhat unstable. During storage, they spontaneously hydrolyze and/or oxidize to produce N-t-butylhydroxylamine ("NtBHA" or "MNP-OH") and other products. As the last name implies, NtBHA is in redox equilibrium with its corresponding nitroso spintrap, MNP*, to which it readily oxidizes. These compounds are readily taken up and concentrated by cells, particularly in mitochondria.
NtBHA scavenges radicals significantly better than PBN. In doing so it is oxidized to MNP, which, in addition to scavenging radicals itself, can be reduced by (say) vitamin-C back to NtBHA. Under light or oxidizing conditions, MNP releases nitric oxide (NO). Nitric oxide is a natural messenger substance with antiplatelet, antioxidant, neuroprotective and blood-vessel-dialating properties. I had previously noted the relationship between the spintraps and nitric oxide.
Such neuroprotective products may account for the efficacy found in the Saint-I trial. This occurred in previous studies with PBN in which it similarly proved impossible to replicate earlier results produced with "aged" PBN.
That hydrolysis/oxidation products account for much of the pharmacological activity of nitrone spin traps is well-known. The possibility of NtBHA, MNP, etc. contributing to the action of NXY-059 was even hinted at in a Swedish PhD Dissertation. The author, Fredrik Clausen, notes that this might explain why NXY-059 is a neuroprotectant, even though it does not penetrate the blood-brain barrier. Interestingly, he also states, "According to AstraZeneca representatives there is no evidence of any in vivo decomposition of S-PBN or NXY-059 (personal communication)...".
Note the "in vivo" limitation. However, the issue is "decomposition" in vitro.
I attended the 2007 International Stroke conference, at which the results of the Saint-2 trial were announced. Significantly, though I went expecting to discuss this well-known --to free radical researchers at least-- issue, none of the clinical investigators I talked to knew that some breakdown products such as NtBHA or MNP might be responsible for the action of NXY-059. From which I conclude that the AstraZeneca people were also probably unaware of this possibility, since likely they would have informed their clinical investigators of it, had they known. Nor, surprisingly, is it listed as a possibility in the recent report on the SAINT-2 trial, if only to be dismissed. This is the face of good evidence that such products account for much (all?) of the activity of the parent PBN, where this has been examined in detail.
However, an AstraZeneca patent** (the " '527 patent ") describes stabilizing NXY-059 against such breakdown, to which NXY-059 is evidently quite susceptible.Thus, the '527 patent application states "..Standard aqueous formulations of (NXY-059).....suffer from the problem that they readily undergo decomposition. In particular, the shelf life of such formulations is unacceptably short...". Emphasis-added.
Similarly, "... one pathway for the decomposition involves hydrolysis of the nitrone functional group to yield the aldehyde (II) and N-tert-butylhydroxylamine as products. A second pathway involves an autoxidation process, possibly involving a free radical mediated degradation. In this pathway the same two products are formed initially but the N-tert-butylhydroxylamine subsequently undergoes further reactions to give other products..."
The '527 patent gives data on the rate of breakdown of NXY-059 under certain storage conditions and the means to prevent this. E.g., according to the patent, in normal storage NXY-059 breaks down to the corresponding benzaldehyde derivative plus NtBHA and its oxidation products at the rate of a percent or so a year. Again depending, this can approach 1% or so a day(!) in infusion solutions. Some calculations.
Perhaps this patent explains the failure of the Saint-2 trial. E.g., SAINT-1 was delayed for more animal studies. Perhaps increased storage time resulted in more hydrolysis products or increased oxidation of NtBHA to MNP. In another scenario, as the Saint-1 trial progressed, AZ researchers measuring shelf-life became aware of the "unacceptably short" storage life of NXY-059. Perhaps they then attempted to correct this before the Saint-2 trial--- Say, by storing the material under nitrogen, reducing headspace, keeping the pH well above 7, etc., as detailed in the '527 patent. Thus, the scenario goes, much less of the putative "real" neuroprotectants were produced during SAINT-2 than in the less stabilized formulations in the SAINT-1 trial.
Similarly, going by the data in the '527 patent, differences in things as basic as light exposure during infusion or time and conditions of storage might be responsible. All conjecture, naturally. Or it might be something else. The title of our Stroke paper says it in more general terms-- "SAINT-1 Worked, but the Neuroprotectant is not NXY-059".
One investigator states that AstraZeneca explored the possibility of hydrolysis products producing the results in SAINT-1 and that "they were confident there was no breakdown product involved". Unfortunately, no details are given. Similarly, a reply to our letter in Stroke states that there was no delay of the SAINT trials. This is contrary to a Centaur pharmaceuticals SEC filing.
Likewise, they state that both trials took equal advantage of the stabilization technology in the '527 patent. This technology and new synthesis methods specifically-designed to minimize production of MNP ( US patents 6,512,143 and 6,689,911 ) means that the SAINT trials used different material from earlier preclinical and clinical trials.
Similarly, with so many variables, it would be difficult to apply the stabilization technology equally . A simple difference in time from reconstitution to infusion of NXY-059 could account for the difference. For example, in addition to the issue of MNP monomer formation, hydrolysis/oxidation of NXY-059 is orders of magnitude higher in infusion conditions than in storage. The chemistry here is fairly complicated-- involving first hydrolysis, then two single-electron oxidations and (possibly) production of NO.
AstraZeneca could help resolve this issue. E.g., They likely have hydrolysis data for the material in both trials.
Not to belabor the point. However, there are issues here well beyond one more failed clinical trial. For example, the peculiar failure of NXY-059 has generated a nihilistic attitude toward development of neuroprotectants, as well as the therapeutic use of nitrone and nitroso spintraps, an area in which I have some patent claims.
Most important, if our conjecture is correct, the SAINT methodology is quite valid-- it not only showed an unexpected neuroprotectant, but distinguished this from NXY-059.
Phone: 713-960-1616 or firstname.lastname@example.org
Peter H. Proctor, PhD, MD
Who Am I?
Disclosure: I own patents covering MNP and its reduction products, as well as other spin traps and spin labels. And yes, there was a "Failure to communicate" here. In any case, the real losers are stroke victims.
* MNP is also known as "2-methyl,2 nitrosopropane", "t-nitrosobutane" or "t-NB".
** The corresponding Swedish patent (PCT/SE01/01164) lists AstraZeneca as the applicant.
Engine fodder-- pay no attention: NXY-059 astrazeneca stroke saint neuroprotection mnp tertiary-nitrosobutane
2-methyl-2-nitrosopropane n-tertiary-butylhydroxylamine NtBHA preclinical neuroprotectant bruce ames richard cutler.
centaur antioxidant in vitro stroke astrazeneca pasted nonsense for indexing
Stair clinical trial preclinical in vivo t-hba N-tbha neuroprotective deafness
vascular endothelium stroke AstraZeneca. Tempol work for the treatment of fibrocystic disease of breast Clot buster tpa inner ear hair cells stria vascularis deafness neuronal regrowth organic semicoductors ischemic injury neuroprotection saint-1 as well as saint-2 clinical trial loss failure cerovive. Treatment for